BPT10819 PDF

CHEMICAL PROCESS PRINCIPLES
ADVISORY BOARD
For Books in Chemical Engineering
T . H . C H I L T O N , Chem. E.
Engineering Department, Experimental Station,
E. I. du Pont de Nemours and Company
T. B. D R E W , SM.
Professor of Chemical Engineering, Columbia
University
O . A . HouGEN, Ph.D.
Professor of Chemical Engineering, University of
Wisconsin
D . B . :g^YEs; A.D.
Vice President, Heyden Chemical Corporatioti
i
K. M. WATSON, Ph.D.
Professor of Chemical Engineering, University of
Wisconsin
HAROLD C . WEBER, D.SC.

Professor of Chemical Engineering, Massachusetts
Institute of Technology
CHEMICAL PROCESS
PRINCIPLES
A COMBINED VOLUME CONSISTING OF
Part One • MATERIAL AND ENERGY BALANCES

Part Two • THERMODYNAMICS
Part Three • KINETICS AND CATALYSIS
OLAF A. HOUGEN
AND
KENNETH M. WATSON
PROFESSORS OP CHEMICAL ENGINEERINC
UNIVERSITY OF WISCONSIN
NEV YORK
J O H N WILEY & SONS, INC.
CHAPMAN AND HALL, LIMITED
LONDON
e M
}o&i'^
PART I
COPYRIGHT, 1943
BY
OLAF A . HODGEN
AND
KENNETH M . WATSON
PART II
COPYBIGHT, 1947
BY
OLAF A . HOTJGEN
AND
KENNETH M . WATSON
PART III
COPTBIGHT, 1947
BY
OLAF A . HonoEN
AND
KENNETH M . WATSON
All Rights Reserved

Thin hook or any 'part thereof must not
he reproduced in any form without
the written permission- of the publisher.
P R I N T E D If* T H E U N I T E D STATES OF AMERICA

PREFACE
" In the following pages certain industrially important principles of chem-

istry and physics have been selected for detailed study. The significance
of each principle is intensively developed and its applicability and limitations
scrutinized." Thus reads the preface to the first edition of Industrial Chemical
Calculations, the precursor of this book. The present book continues to give
intensive quantitative training in the practical applications of the principles
of physical chemistry to the solution of complicated industrial problems and
in methods of predicting missing physicochemical data from generalized
principles. In addition, through recent developments in thermodynamics
and kinetics, these principles have been integrated into procedures for process
design and analysis with the objective of arriving at optimum economic
results from a minimum of pilot-plant or test data. The title Chemical Process
Principles has been selected to emphasize the importance of this approach to
process design and operation.
The design of a chemical process involves three types of problems, which
although closely interrelated depend on quite different technical principles.
The first group of problems is encountered in the preparation of the material
and energy balances of the process and the establishment of the duties to be
performed by the various items of equipment. The second type of problem
is the determination of the process specifications of the equipment necessary
to perform these duties. Under the third classification are the problems of
equipment and materials selection, mechanical design, and the integration
of the various units into a coordinated plot plan.
These three types may be designated as process, unit-operation, and plant-
. design problems, respectively. In the design of a plant these problems cannot
be segregated and each treated individually without consideration of the
others. However, in spite of this interdependence in application the three
types may advantageously be segregated for study and development because
of the different principles involved. Process problems are primarily chemical
and physicochemical in nature; unit-operation problems are for the most
part physical; the plant-design problems are to a large extent mechanical.
In this book only process problems of a chemical and physicochemical
nature are treated, and it has been attempted to avoid overlapping into the
fields of unit operations and plant design. The first part deals primarily
with the applications of general physical chemistry, thermophysics, thermo-
chemistry, and the first law of thermodynamics. Generalized procedures
for estimating vapor pressures, critical constants, and heats of vaporization
have been elaborated. New methods are presented for dealing with equilib-
rium problems in extraction, adsorption, dissolution, and crystallization.
The construction and use of enthalpy-concentration charts have been extended
to complex systems. The treatment of material balances has been elaborated
to include the effects of recycling, by-passing, changes of inventory, and
accumulation of inerts.
vi PREFACE
In the second part the fundamental principles of thermodynamics are pre-
sented with particular attention to generalized methods. The applications
of these principles to problems in the compression and expansion of fluids,
power generation, and refrigeration are discussed. However, it is not at-
tempted to treat the mechanical or equipment problems of such operations.
Considerable attention is devoted to the thermodynamics of solutions
with particular emphasis on generalized methods for dealing with deviations
from ideal behavior. These principles are applied to the calculation of equilib-
rium compositions in both physical and chemical processes.
Because of the general absence of complete data for the solution of process
problems a chapter is devoted to the new methods of estimating thermody-
namic properties by statistical calculations. This treatment is restricted to
simple methods of practical value.
All these principles are combined in the solution of the ultimate problem
of the kinetics of industrial reactions. Quantitative treatment of these
problems is difficult, and designs generally have been based on extensive
pilot-plant operations carried out by a trial-and-error procedure on succes-
sively larger scales. However, recent developments of the theory of absolute
reaction rates have led to a thermodynamic approach to kinetic problems
which is of considerable value in clarifying the subject and reducing it to the
point of practical applicability. These principles are developed and their
apphcation discussed for homogeneous, heterogeneous, and catalytic systems.
Particular attention is given to the interpretation of pilot-plant data. Eco-
nomic considerations are emphasized and problems are included in estabhshing
optimum conditions of operation.
In covering so broad a range of subjects, widely varying comprehensibility
is encountered. It has been attempted to arrange the material in the order
of progressive difficulty. Where the book is used for college instruction in
chemical engineering the material of the first part is suitable for second- and
third-year undergraduate work. A portion of the second part is suitable
for third- or fourth-year undergraduate work; the remainder is of graduate
level. To assist in using the book for undergraduate courses in thermody-
namics and kinetics those sections of Part II which are recommended for such
survey courses are marked. This material has been selected and arranged
to give continuity in a preliminary treatment which can serve as a foundation
for advanced studies, either by the individual or in courses of graduate level.
The authors wish to acknowledge gratefully the assistance of Professor
R. A. Ragatz in the revision of Chapters I and VI, and the suggestions of Pro-
fessors Joseph Hirschfelder, R. J. Altpeter, K. A. Kobe, and Dr. Paul Bender.
OLAF A. HOUQEN
KENNETH M . WATSON
MADISON, WISCONSIN
August, 194s
CONTENTS
Page
PREFACE . v
TABLE OF SYMBOLS ix
PART I
MATERIAL AND ENERGY BALANCES
Chapter
I STOICHIOMETRIC PRINCIPLES 1
II BEHAVIOR OF IDEAL GASES 27
III VAPOR PRESSURES 53
IV/ HUMIDITY AND SATURATION 89
V SOLUBILITY AND SORPTION Ill
Vr/ MATERIAL BALANCES 167
YJy THERMOPHYSICS 201
viy/ THERMOCHEMISTRY 249
I X y FUELS AND COMBUSTION 323
• X CHEMICAL, METALLURGICAL, AND PETROLEUM PROC-
ESSES 383
PART II
THERMODYNAMICS
XI THERMODYNAMIC PRINCIPLES . . . . . . . . ' 437
XII THERMODYNAMIC PROPERTIES OF FLUIDS . . . . 479
XIII EXPANSION AND COMPRESSION OF FLUIDS . . . . 538
XIV THERMODYNAMICS OF SOLUTIONS 595
XV PHYSICAL EQUILIBRIUM 644
XVI CHEMICAL EQUILIBRIUM 691
XVII THERMODYNAMIC PROPERTIES FROM MOLECULAR STRUC-
TURE 756
vii
viii CONTENTS
PART III
KINETICS AND CATALYSIS
Chapter Page
XVIII HOMOGENEOUS REACTIONS 805
XIX CATALYTIC REACTIONS 902
XX MASS AND HEAT TRANSFER IN CATALYTIC BEDS . . 973
XXI CATALYTIC REACTOR DESIGN . 1007
XII UNCATALYZED HETEROGENEOUS REACTIONS . . . 1049
APPENDIX jcvii
AUTHOR INDEX xxiii
SUBJECT INDEX xxvii
TABLE OF SYMBOLS
A area
A atomic weight
A component A
A total work function
a activity
Om external surface per unit mass
ttp external surface per particle
ttv external surface per unit volume
B component B
B constant of Calingaert-Davis equation
B thickness of effective^ film
C component C
C concentration per unit volume
c degrees centigrade
number of components
c over-all rate constant
c heat capacity at constant pressure
c. heat capacity at constant volume
c. Sutherland constant
c.
c concentration of adsorbed molecules per unit mass of
catalyst
c specific heat
c velocity of light
Cp molal heat capacity at constant pressure
Cv molal heat capacity at constant volume
d surface concentration of adsorbed molecules per unit
catalyst area
D diameter
DAB diffusivity of A and B
Dp effective particle diameter equal to diameter of sphere
having the same external surface area as particle
D', effective diameter equal to diameter of sphere having
, same area per unit volume as particle
d differential operator
E energy in general
E energy of activation, Arrhenius equation
E^ effectiveness factor of catalysis
TABLE OF SYMBOLS
c base of natural logarithms

F degrees of freedom
F feed rate
F force
F, external void fraction
Fi internal void fraction
S friction factor _ |^
/ fugacity ,, j^
/ weight fraction , ' -^^
Q free energy _ ^\
6 mass velocity per unit area „
0 specific gravity < ~- , ^^.
o free energy per mole ,, '. ^Y
5 partial molal free energy ,^^.,_ ^
LQ change in free energy " j -'- 6
(g) gaseous state - . >., ,,,', ,.- v^
9 degeneracy • -^ • - nî •
ffo standard gravitational constant, 32.174 (ft/sec)/sec
H enthalpy
H Henry's constant • .. / v-*!
H humidity . •<
He height of catalytic unit # y
Hi height of mass-transfer unit • -i
HH height of heat-transfer unit ^'
Fp height equivalent to a theoretical plate
H, percentage humidity
HH height of reactor unit ,<
Hr Ti(i relative humidity ; »
Ht height of transfer unit
AH change in enthalpy ,,;. - / ,,
AHc heat of combustion . /
AHf heat of formation , :>•, / |i|
AHr heat of reaction ' ,»
A/^» standard enthalpy of activation '\..(\
H enthalpy per mole r-
H partial molal enthalpy
AH partial molal enthalpy change
ft Planck's constant
h heat-transmission coefficient
I inert component
I integration constant
I moment of inertia
TABLE OF SYMBOLS xi
J Jacobian function
/ mechanical equivalent of heat
jt mass-transfer factor in fluid film n
i.-jisîi . , , t . ' heat-transfer factor in fluid film "-l
'•' K ' " ' " characterization factor
K degrees Kelvin «'
^•^'' '• ;*; ' distribution coefficient '•••'- xtl-nH ,«j
: K equifibrium constant ;i y
K vaporization equilibrium constant ' ^ •
Ka equilibrium constant for adsorption ^
Ke equilibrium constant, concentration units
Kg over-all mass-transfer coefficient, pressure units
Kj, over-all mass-transfer coefficient, fiquid concentration
units
Kp equilibrium constant, pressure units irt?
K' surface equilibrium constant i M,^'
It forward-reaction velocity constant ^ tj,"'
k thermal conductivity "' ?>
ft Boltzmann constant r ,•!•',-• .s ?,
kj^ adsorption velocity constant •' • ?•*
k'ji desorption velocity constant 'S^ ?*
kg mass-transfer coeflScient, gas film ' - 'p\.
kj^ mass-transfer coefficient, liquid film ^^
k' reverse-reaction velocity constant ai*
L mass velocity of fiquid per unit area '?*A
L total molal adsorption sites per unit mass
Lfi molal mass velocity of liquid per unit area
L' active centers per unit area of catalyst
I length
Ip heat of pressure change at constant pressure
l„ heat of expansion at constant temperature
(1) liquid state
]n natural logarithm ''•' "' ^3
1(^ logarithm to base 10 "• ''•' it
M molecular weight ' >' M
Mm mean molecular weight >M ' "i
m mass >' ^
m slope of equilibrium curve dy*/dx ^J
m Thiele modulus
N Avogadro number 6.023(1023) -'^
N mole fraction
Nt number of transfer units
ai TABLE OP SYMBOLS
No number of transfer units, gas film '

Nj^ number of transfer units, liquid film
n number of moles
P pressure (used only in exceptional cases to distinguish
pressure of pure components from partial pressures
of some component in solution
Pf factor for unequal molal diffusion in gas film
Q heating value of fuel ; . '0, ;\
Q partition function ,,- ;i
q heat added to a system ,V
g, rate of heat flow .>\
R component R , ,;A
R gas constant , »
r radius
r,^ rate of reaction or transfer of A per unit area
r,nA rate of reaction or transfer of A per unit mass
r^ rate of reaction or transfer of A per unit volume
S component S
8 cross section
S entropy
8 humid heat
Sp percentage saturation
Sr relative saturation * .• i
(S« space velocity
AiS* entropy of activation
8 molal entropy
8 number of equidistant active sites adjacent to each
other
(B) solid state ,
T absolute temperature, degrees Rankine or Kelvin
t temperature, °F or °C
V internal energy , ; ;;;
U over-all heat-transfer coefficient / ^j
V internal energy per mole ,;;
u velocity
V molecular volume in Gilliland equation
V volume
Vf volume of reactor
V volume per mole ri
w weight
w work done by system
We work of expansion done by system
TABLE OF SYMBOLS xiil
to/ electrical work dme by system

w. shaft work *
X^ activated complex
X mole fraction in liquid phase
X ' mole fraction of reactant converted in feed
. X quality-
y mole fraction in vapor
y* mole fraction in vapor, equilibrium value
elevation above datum plane =' v
z height or thickness of reactor '
zz compressibility factor
z mole fraction in total system !'>'4
., ' DiMENSIONLESS NuMBEHS
DQ
Nj,, Beynold's number —
iVp, Prandtl number —r- '

k
Nst Stanton number 7777
Nsc, Schmidt number •*=?-
«• •rS
• , SUBSCEIPTS
A ' component A •> 1

a air _ • • '••"
B ' component B ' • >-i' , .\
b normal boiling point ' . . * , .;,-:
C component C . 1 • • • ..1
c critical state < ' ~^
D component D >'
D dense arrangement '
e expansion " "
f electrical and radiant , • x.-, (
f formation
f fusion • ' •'' ' " •=
G gas or vapor " ''.• . •- ' '•••' •
H isenthalpic
L hquid • '^
L loose arrangement 'jv-x. *?f i. j .
xiv TABLE OF SYMBOLS
p constant pressure r- : •'

R component R
r reduced conditions
r relative
S component S >
S isentropic
s normal boiling point
s saturation
T isothermal i;?,;
t temperature .
t transition
V constant volume <
V vapor
w water vapor
''^'' G E E B K SYMBOLS
(a) crystal form
a coefficient of compressibility
a proportionality factor for diffusioA
a relative volatility
a thermal diffusivity =.'.-.••...,-'.
(/3) crystal form ^
(3 coefficient of volumetric expansioA ;
7 activity coefficient
(7) crystal form
A finite change of a property; positive value indicates an
increase - .
S change in moles per mole of r e a c t ^ t
S deformation vibration
d partial differential operator ' .~ i ,{
c >. energy per molecule , *-
T) efficiency
0 fraction of total sites covered
K ratio of heat capacities
A heat of vaporization
X heat of vaporization per mole
X wave length
X/ heat of fusion per mole
H chemical potential
11 viscosity
V frequency
V fugacity coefficient of gas
TABLE OF SYMBOLS XV
V n u m b e r of ions „^
V n u m b e r of molecules
V valence or stretching vibration
a expansion factor of liquid
a wave number
V "total pressure of mixture, used where necessary to
distinguish from p „ ,.
P density %, •
PB bulk density 4"
Pp particle density
Pc t r u e solid density :, „-
S summation
<r surface tension
a symmetry number > . :
T time
4> activity coefficient
<f> n u m b e r of phases
SUPERSCRIPTS -
t
*, ideal behavior
* equilibrium state
o standard state
A, pseudo state
/ leverse rate
t standard state of activation
CHAPTER I
STOICHIOMETRIC PRINCIPLES
The principal objective to be gained in the study of this book is the
ability to reason accurately and concisely in the application of the
principles of physics and chemistry to the solution of industrial problems.
It is necessary that each fundamental principle be thoroughly under-
stood, not superficially memorized. However, even though a knowledge
of scientific principles is possessed, special training is required to solve
the more complex industrial problems. There is a great difference
between the mere possession of tools and the ability to handle them
skilfully.
• Direct and logical methods for the combination and application of
certain principles of chemistry and physics are described in the text and
indicated by the solution of illustrative problems. These illustrations
should be carefully, studied and each individual operation justified.
However, it is not intended that these illustrations should serve as forms
for the solution of other problems by mere substitution of data. Their
function is to indicate the organized type of reasoning which will lead to
the most direct and clear solutions. In order to test the understanding
of principles and to develop the ability of organized, analytical rea-
soning, practice in the actual solution of typical problems is indispen-
sable. The problems selected represent, wherever possible, reasonable
conditions of actual industrial practice.
Conservation of Mass. A system refers to a substance or a group of
substances under consideration and a process to the changes taking
place within that system. Thus, hydrogen, oxygen, and water may
constitute a system, and the combustion of hydrogen to form water,
the process. A system may be a mass of material contained within
a single vessel and completely isolated from the surroundings, it may
include the mass of material in this vessel and its association with the
surroundings, or it may include all the mass and energy included in a
complex chemical process contained in many vessels and connecting lines
and in association with the surroundings. In an isolated system the
boundaries of the system are limited by a mass of material and its
energy content is completely detached from all other matter and
energy. Within e given isolated system the mass of the system remains
constant regardless of the changes taking place within the system.
This statement is known as the law of conservation of mxiss and is the
basis of the so-called material balance of a process.
I
2 STOICHIOMETRIC PRINCIPLES [CHAP. I
The state of a system is defined by numerous properties which are

classified as extensive if they are dependent on the mass under consid-
eration and intensive if they are independent of mass. For example,
volume is an extensive property, whereas density and temperature are
intensive properties.
In the system of hydrogen, oxygen, and water undergoing the process
of combustion the total mass in the isolated system remains the same.
If the reaction takes place in a vessel and hydrogen and oxygen are
fed to the vessel and products are withdrawn then the incoming and
outgoing streams must be included as part of the system in applying
the law of conservation of mass or in estabhshing a material- balance.
The law of conservation of mass may be extended and appUed to the
mass of each element in a system. Thus, in the isolated system of
hydrogen, oxygen, and water undergoing the process of combustion the
inass of hydrogen in its molecular, atomic, and combined forms remains
constant. The same is true for oxygen.
In a strict sense the conservation law should be applied to the com-
bined energy and mass of the system and not to the mass alone. By
the emission of radiant energy mass is converted into energy and also
in the transmutation of the elements the mass of one element must
change; however, such transfon;\ations never fall within the range of
experience and detection in industrial processes so that for all practical
purposes the law of conservation of mass is accepted as rigorous and
valid.
Since the word weight is entrenched in engineering hterature as syn-
onomous with moss, the common practice will be followed in frequently
referring to weights of material instead of using the more exact term
mass as a measure of quantity. Weights and masses are equal only at
sea level but the variation of weight on the earth's surface is negligible
in ordinary engineering work.
STOICHIOMETRIC RELATIONS
Nature of Chemical Compounds. According to generally accepted

theory, the chemical elements are composed of submicroscopic particles
which are known as atoms. Further, it is postulated that all of the
atoms of a given element have the same weight,' but that the atoms of
different elements have characteristically different weights.
^ Since the discovery of isotopes, it is commonly recognized, that the individual
atoms of certain elements vary in weight, and that the so-called atomic weight of an
element is, in reality, the weighted average of the atomic weights of the isotopes.
In nature the various isotopes of a given element are always found in the same pro-
portions; hence in computational work it is permissible to use the weighted average
atomic weight as though all atoms actually possessed this average atomic weight. ^
CHAP." I] NATURE OF CHEMICAL COMPOUNDS 3
When the atoms of the elements unite to form a particular com-

pound, it is observed that the compound, when carefully purified, has a
fixed and definite composition rather than a variable and indefinite
composition. For example, when various samples of carefully purified
sodium chloride are analyzed, they all are found to contain 60.6 per
cent chlorine and 39.4 per cent sodium. Since the sodium chloride is
composed of sodium atoms, each of which has the same mass, and of
chlorine atoms, each of which has the same mass (but a mass that is
different from the mass of the sodium atoms), it is concluded that in
the compound sodium chloride the atoms of sodium and chlorine have
combined according to some fixed and definite integral ratio.
By making a careful study of the relative weights by which the
chemical elements unite to form various compounds, it has been pos-
sible to compute the relative weights of the atoms. Work of this type
occupied the attention of many of the early leaders in chemical research
and has continued to the present day. This work has resulted in the
famihar table of international atomic weights, which is still subject to
periodic revision and refinement. In this table, the numbers, which
are known as atomic weights, give the relative weights of the atoms of
the various chemical elements, all referred to the arbitrarily assigned
value of exactly 16 for the oxygen f tom.
A large amount of work has been done to determine the composition
of chemical compounds. As a result of this work, the composition of a
great variety of chemical compounds can now be expressed by formulas
which indicate the elements that comprise the compound and the
relative number of the atoms of the various elements present.
It should be pointed out that the formula of the compound as ordi-
narily written does not necessarily indicate the exact nature of the
atomic aggregates that comprise the compound. For example, the
formula for water is written as H2O, which indicates that when hydrogen
and oxygen unite to form water, the union of the atoms is in the ratio
of 2 atoms of hydrogen to 1 atom of oxygen. If this compound exists
as steam, there are two atoms of hydrogen permanently united to one
atom of oxygen, forming a simple aggregate termed a molecule. Each
molecule is in a st^te of random motion and has no permanent associa-
tion with other similar molecules to form aggregates of larger size.
However, when this same substance is condensed to the liquid state,
there is good evidence to indicate that the individual molecules become
associated, to form aggregates of larger size, (Il20)j:, x being a
variable quantity. With respect to solid substances, it may be said
that the formula as written merely indicates the relative number of
atoms present in the compound and has no further significance. For
example, the formula for cellulose is written CeHioOe, but it should not
be concluded that individual molecules, each of which contains only

6 atoms of carbon, 10 atoms of hydrogen and, 5 atoms of oxygen exist.
There is much evidence to indicate that aggregates of the nature of
(CeHioOs)! are formed, with x a large number.
It is general practice where possible to write the formula of a chem-
ical compound to correspond to the number of atoms making up one
molecule in the gaseous state. If the degree of association in the
gaseous state is unknown the formula is written to correspond to the
lowest possible number of integral atoms which might make up the
molecule. However, where the actual size of the molecule is important
care must be exercised in determining the degree of association of a
compound even in the gaseous state. For example, hydrogen fluoride
is commonly designated by the formula H F and at high temperatures
and low pressures exists in the gaseous state in molecules each com-
prising one atom of fluorine and one atom of hydrogen. However, at
high pressures and low temperatures even the gaseous molecules un-
dergo association and the compound behaves in accordance with the
formula (HF)i, with x a function of the conditions of temperature and
pressure. Fortunately behavior of this type is not common.
Mass Relations in Chemical Reactions. In stoichiometric calcu-
lations, the mass relations existing between the reactants and products
of a chemical reaction are of primary interest. Such information may
be deduced from a correctly written reaction equation, used in con-
junction with atomic weight values selected from a table of atomic
weights. As a typical example of the procedures followed, the reaction
between iron and steam, resulting in the production of hydrogen and the
magnetic oxide of iron, Fe304, may be considered. The first requisite
is a correctly written reaction equation. The formulas of the various
reactants are set down on the left side of the equation, and the formulas
of the products are set down on the right side of the equation, taking
care to indicate correctly the formula of each substance involved in the
reaction. Next, the equation must be balanced by inserting before
each formula coefficients such that for each element present the total
number of atoms originally present will exactly equal the total number
of atoms present after the reaction has occurred. For the' reaction
under consideration the following equation may be written:
3Fe + 4H2O -> Fe304 + 4H2
The next step is to ascertain the atomic weight of each element involved
in the reaction, by consulting a table of atomic weights. From these
atomic weights the respective molecular weights of the various com-
pounds may be calculated.
CHAP. I] VOLUME RELATIONS IN CHEMICAL REACTIONS
Atomic Weights: <• - '..•...,11.,, .. •

Iron 55.84
Hydrogen 1.008
!/ Oxygen 16.00
Molecular Weights: • i ' i'• '• • 1

H2O (2 X 1.008) + 16.00 = 18.02
FeaOi (3 X 55.84) + (4 X 16.00) = 231.5
H2 (2 X 1.008) = 2.016 . , ;
The respective relative weights of the reactants and products may-

be determined by multiplying the respective atomic or molecular
weights by the coefficients that precede the formulas of the reaction
equation. These figures may conveniently be inserted directly below
the reaction equation, thus:
3Fe + 4H2O -* FesOi + 4H2
' *- (3X55.84) (4X18.02) 231.5 (4X2.016)
:-' -(i .; 167.52 72.08 231.5 8.064
Thus, 167.52 parts by weight of iron react with 72.08 parts by weight
of steam, to form 231.5 parts by weight of the magnetic oxide of iron
and 8.064 parts by weight of hydrogen. By the use of these relative
weights it is possible to work out the particular weights desired in a
given problem. For example, if it is required to compute the weight
of iron and of steam required to produce 100 pounds of hydrogen, and
the weight of the resulting oxide of iron formed, the procedure would
be as follows:
Reactants:
Weight of iron = 100 X (167.52/8.064) 2075 lb
Weight of steam = 100 X (72.08/8.064) 894 Jb
Total 2969 lb
Products:
Weight of iron oxide = 100 X (231.5/8.064) '. 2869 lb
Weight of hydrogen 100 lb
Total 2969 lb
Volume Relations in Chemical Reactions. A correctly written reac-

tion equation will indicate not only the relative weights involved in a
chemical reaction, but also the relative volumes of those reactants and
products which are in the gaseous state. The coefficients preceding
the molecular formulas of the gaseous reactants and products indicate
the relative volumes of the different substances. Thus, for the reaction
under consideration, for every 4 volumes of steam, 4 volumes of hy-
drogen are produced, when both materials are reduced to the same
temperature and pressure. This volumetric relation follows from
Avogadro's law, which states that equal volumes of gas at the same
conditions of temperature and pressure contain the same number of
molecules, regardless of the nature of the gas. That being the case,
and since 4 molecules of steam produce 4 molecules of hydrogen, it may-
be concluded that 4 volumes of steam will produce 4 volumes of hy-
drogen. It cannot be emphasized too strongly that this volumetric
relation holds only for ideally gaseous substances, and must never be
applied to liquid or to solid substances.
The Gram-Atom and the Pound-Atom. The numbers appearing in
a table of atomic weights give the relative weights of the atoms of the
various elements. It therefore follows that if masses of different ele-
ments are taken in such proportion that they bear the same ratio to one
another as do the respective atomic weight numbers, these masses will
contain the same number of atoms. For example, if 35.46 grams of
chlonae, which has sa atomic wei^t oi 35.46, are. takea, and if 55.84
grams of iron, which has an atomic weight of 55.84, are taken, there
will be exactly the same number of chlorine atoms as of iron atoms in
these respective masses of material. '
The mass in grams of a given element which is equal numerically to its
atomic weight is termed a gram-atom. Similarly, the mass in pounds of
a given element that is numerically equal to its atoinic weight is termed
a pound-atom. From these definitions, the following equations may be
written:
Gram-atoms of an elementary substance = -—:—r-

Atomic weight
Grams of an elementary substance = Gram-atoms X Atomic weight
Pound-atoms of an elementary substance = — r^ :

Atomic weight
Pounds of an elementary substance = Polind-atoins X Atomic weight
The actual number of atoms in one gram-atom' of an elementary sub-

stance has been determined by several methods, the average result being
6.023 X 10^1 This numbel^ known as the Avogadi-o number, is of con-
siderable theoretical importance.
The Gram-Mole and the Pound-Mole. It has been pointed out that
the formula of a chemical compound indicates the relative numbers and
the kinds of atoms that unite to form a compound. For example, the
formula NaCl indicates that sodium and chlorine atoms are present in
CHAP. I] RELATION BETWEEN MASS AND VOLUME 7
the compound in a 1 : 1 ratio. Since the gram-atom as above defined

contains a definite number of atoms, which is the same for all elementary-
substances, it follows that gram atoms will unite to form a compound in
exactly the same ratio as do the atoms themselves, forming what may be
termed a gram-mole of the compound. For the case under consideration,
it may be said that one gram-atom of sodium unites with one gram-atom
of chlorine to form one gram-mole of sodium chloride.
One gram-mole represents the weight in grams of all the gram-atoms
which, in the formation of the compound, combine in the same ratio as
do the atoms themselves. Similarly, one pound-mole represents the
weight in pounds of all of the pound-atoms which, in the formation of
the compound, combine in the same ratio as do the atoms themselves.
From these definitions, the following equations may be written:
^ , . , Mass in grams
Gram-moles oi a substance Molecular weight
Grams of a substance = Gram-moles X Molecular weight
„ , , , , , Mass in pounds
Pound-moles oi a substance = ^ —
Molecular weight
Pounds of a substance = Pound-moles X Molecular weight
The value of these concepts may be demonstrated by consideration of
the reaction equation for the production of hydrogen by passing steam
over iron. The reaction equation as written indicates that 3 atoms of
iron unite with 4 molecules of steam, to form 1 molecule of magnetic
oxide of iron and 4 molecules of hydrogen. It may also be interpreted
as saying that 3 gram-atoms of iron unite with 4 gram-moles of steam,
to form 1 gram-mole of Fe304 and 4 gram-moles of Ha. In other words,
the coefficients preceding the chemical symbols represent not only the
relative number of molecules (and atoms for elementary substances that
are not in the gaseous state) but also the relative number of gram-moles
(and of gram-atoms for elementary substances not in the gaseous state).
Relation Between Mass and Volume for Gaseous Substances. Lab-
oratory measurements have shown that for all substances in the ideal
gaseous state, 1.0 gram-mole of material at standard conditions {0°C, 760
mm Hg) occupies 22.4 liters.^ Likewise, if 1.0-pound-mole of the gaseous
material is at standard conditions, it will occupy a volume of 359 cu ft.
2 The actual volume corresponding to 1 gram-mole of gas at standard conditions
will show some variation from gas to gas owing to various degrees of departure from
ideal behavior. However, in ordinary work the ideal values given above may be
used without serious error. •
Accordingly, with respect to the reaction equation previously dis-

cussed, it may be said that 167.52 grams of iron (3 gram-atoms) will
form 4 gram-moles of hydrogen, which will, when brought to standard
conditions, occupy a volume of 4 X 22.4 Uters, or 89.6 liters. Or, if
Enghsh units are to be used, it may be said that 167.52 pounds of iron
(3 pound-atoms) will form 4 pound-moles of hydrogen, which will
occupy a volume of 4 X 359 cubic feet (1436 cubic feet) at standard
conditions. . . ? T. .,^^r,';•.!•••• iY>'/';.r^-'''«>
Illustration 1. A cylinder contains 25 lb of liquid chlorine. What volume in
cubic feet will the chlorine occupy if it is released and brought to standard conditions?
Basis of Calculation: 25 lb of chlorine.
Liquid chlorine, when vaporized, forma a gas composed of diatomic molecules, CI2.
Molecular weight of chlorine gas = (2 X 35.46) 70.92
Lb-moles of chlorine gas = (25/70.92) 0.3525
Volume at standard conditions = (0.3525 X 359).... 126.7 cu ft
Illustration 2. Gaseous propane, CsHg, is to be Uquefied for storage in steel
cylinders. How many grams of liquid propane will be formed by the liquefaction of
500 liters of the gas, the volume being measured at standard conditions?
Basis of Calculation: 500 liters of propane at standard conditions.
Molecular weight of propane 44.06
Gram-moles of propane = (500/22.4); 22.32
••' Weight of propane = 22.32 X 44.06 , 985 grams
The Use of Molal Units in Computations. The great desirability of

the use of molal units for the expression of quantities of chemical com-
pounds cannot be overemphasized. Since one molal unit of one com-
pound will always react with a simple multiple number of molal units of
another, calculations of weight relationships in chemical reactions are
greatly simplified if the quantities of the reacting compounds and prod-
ucts are expressed throughout in molal units. This simphfication is not
important in very simple calculations, centered about a single compound
or element. Such problems are readily solved by the means of the com-
bining weight ratios, which are commonly used as the desirable means
for making such calculations as may arise-in quantitative analyses.
However, in an industrial process successive reactions may take place
with varying degrees of completion, and it may be desired to calculate
the weight relationships of all the materials present at the various stages
of the process. In such problems the use of ordinary weight units with
combining weight ratios will lead to great confusion and opportunity for
arithmetical error. The use of molal units, on the other hand, will give
a more direct and simple solution in a form which may be easily verified.
'CHAP. I] THE USE OF MOLAL UNITS IN COMPUTATIONS 9
It is urged as highly advisable that familiarity with molal units be gained

through their use in all calculations of weight relationships in chemical
compounds and reactions.
A still more important argument for the use of molal units lies in the
fact that many of the physicochemical properties of materials are
expressed by simple laws when these properties are on the basis of a molal
unit quantity.
The molal method of computation is shown by the following illus-
trative problem which deals with the reaction considered earlier in this
section, namely, the reaction between iron and steam to form hydrogen
and the magnetic oxide of iron: v i< •• • . i •
Illustration 3. (o) Calculate the weight of iron and of steam required to produce
100 lb of hydrogen, and the weight of the Fe304 formed. (6) What volume will the
hydrogen occupy at standard conditions?
Reaction Equation:
,.: B<, .1. 3Fe + 4H2O —>FesOs + 4Hj
Basis of Calculation: 100 lb of hydrogen. • ~,-.--i

Molecular and atomic weights: ' '-
i' Fe 55.84
!'':' H20 18.02 • '•• •
''""'"'"'*'" FesOi 231.5
. , Hj 2.016
» fX
Hydrogen produced = 100/2.016 49.6 lb-moles
Iron required = 49.6 X 3/4 37.2 lb-atoms ,;,,,
or 37.2 X 55.84 2075 lb
Steam required = 49.6 X 4 / 4 49.6 lb-moles
or 49.6X18.02 894 1b
FeaOi formed = 49.6 X 1/4 12.4 lb-moles
or 12.4X231.6 2870 1b
Total input = 2075 -|- 894 2969 lb
Total output = 2870 4-100 2970 1b - M ».»
Volume of hydrogen at standard conditions =
49.6X359 17,820 cu ft
In this simple problem the full value of the molal method of calcula-
tion is not apparent; as a matter of fact, the method seems somewhat
more involved than the solution which was presented earUer in this
section, and which was based on the simple rules of ratio and proportion.
It is in the more complex problems pertaining to industrial operations
that the full benefits of the molal method of calculation are realized.
10 STOICHIOMETRIC PRINCIPLES [CHIP. I
Excess Reactants, In most chemical reactions carried out in in-

dustry, the quantities of reactants supplied usually are not in the exact
proportions demanded by the reaction equation. It is generally desir-
able that some of the reacting materials be present in excess of the
amounts theoretically required for combination with the others. Under
such conditions the products obtained will contain some of the uncom-
bined reactants. The quantiti,es of the desired compounds which are
formed in the reaction will be determined by the quantity of the limiting
reactant, that is, the material which is not present in excess of that
required to combine with any of the other reacting materials. The
amount by which any reactant is present in excess of that required to
combine with the hmiting reactant is usually expressed as its percentage
excess. The percentage excess of any reactant is defined as the percent-
age ratio of the excess to the amount theoretically required for combina-
tion with the limiting reactant.
The definition of the amount of reactant theoretically required is some-
times arbitrarily established to comply with particular requirements.
For example, in combustion calculations, the fuel itself may contain
some oxygen, and the normal procedure in such instances is to give a
figure for percentage of excess oxygen supplied by the air which is based
on the net oxygen demand, which is the total oxygen demanded for com-
plete oxidation of the combustible components, minus the oxygen in the
fuel.
Degree of Completion. Even though certain of the reacting materials
may be present in excess, many industrial reactions do not proceed to the
extent which would result from the complete reaction of the limiting
material. Such partial completion may result from the estabUshment
of an equilibrium in the reacting mass or from insufficient time or oppor-
tunity for completion to the theoretically possible equilibrium. The
degree of completion of a reaction is ordinarily expressed as the percentage
of the limiting reacting material which is converted or decomposed into
other products. In processes in which two or more successive reactions
of the same materials take place, the degree of completion of each step
may be separately expressed.
In those instances where excess reactants ^re present and the degree
of completion is 100%, the material leaving the process will contain not
only the direct products of the chemical reaction but also the excess
reactants. In those instances where the degree of completion is below
100%, the material leaving the process will contain some of each of the
reactants as well as the direct products of the chemical reactions that
took place.
CHAP. I] BASIS OF CALCULATION 11
BASIS OF CALCtTLATION
Normally, all tKe calculations connected with a given problem are
presented with respect to some specific quantity of one of the streams of
material entering or leaving the process. This quantity of material is
designated as the basis of calculation, and should always be specifically
stated as the initial step in presenting the solution to the problem. Very
frequently the statement of the problem makes the choice of a basis of
calculation quite obvious. For example, in Illustration 3, the weights
of iron, steam, and magnetic oxide of iron involved in the production of
100 pounds of hydrogen are to be computed. The simplest procedure
obviously is to choose 100 pounds of hydrogen as the basis of calcu-
lation, rather than to select some other basis, such as 100 pounds of iron
oxide, for example, and finally convert all the weights thus computed to
the basis of 100 pounds of hydrogen produced.
In some instances, considerable simplification results if 100 units of
one of the streams of material that enter or leave the process is selected
as the basis of computation, even though the final result desired may be
with reference to some other quantity of material. If the compositions
are given in weight per cent, 100 pounds or 100 grams of one of the
entering or leaving streams of material may be chosen as the basis of
calculations, and at the close of the solution, the values that were com-
puted with respect to this basis can be converted to any other basis that
the statement of the problem may demand. For example, if it were
required to compute the weight of CaO, MgO, and CO2 that can be
obtained from the calcination of 2500 pounds of limestone containing
90% CaCOs, 5% MgCOa, 3 % inerts, and 2% H2O, one procedure would
be to select 2500 pounds of the Hmestone as the basis of calculation, and
if this choice is made, the final figures wiU represent the desired result.
' An alternative procedure is to select 100 pounds of hmestone as the basis
of calculation, and then, at the close of the computation, convert the
weights computed on the desired basis of 2500 pounds of Hmestone. In
this very simple illustration, there is Uttle choice between the two pro-
cedures, but in complex problems, where several streams of material are
involved in the process and where several of the streams are composed
of several components, simplification will result if the second procedure is
adopted. It should be added that if the compositions are given in
mole per cent, it will prove advantageous to choose 100 pound-moles or
100 gram-moles as the basis of calculation.
In presenting the solutions to the short illustrative problems of this
chapter, it may have appeared superfluous to make a definite statement
as to the basis of calculation. However, since such a statement is of

extreme importance in working out complex problems, it is considered
desirable to follow the rule of always stating the basis of calculation at
the beginning of the solution, even though the problem may be relatively
simple.
METHODS OF EXPRESSING THE COMPOSITION

•;,i -i. OF MIXTURES AND SOLUTIONS
Various methods are possible for expressing the composition of mix-
tures and solutions. The different methods that are in common use
may be illustrated by considering a binary system, composed of com-
ponents which will be designated as A and B. The following symbols
will be used in this discussion:
m = total weight of the system.

rriA and m^ = the respective weights of components A and B.
MA and MB = the respective molecular weights of components A
and B, if they are compounds.
AA and As = the respective atomic weights of components A and B,
if they are elementary substances. •
y = volume of the system, at a particular temperature and
,, , , ],, pressure. ^,|.^.. ; .,^,,, ,^j,j,. • ^ .
Si VA and VB = the respective pure component volumes of components
A and B. The pure component volume is defined as
the volume occupied by a particular component if it
,_,,,„• ,..— -• . is separated from the mixture, and brought to the
same temperature and pressure as the original
mixture.
Weight Per Cent. The weight percentage of each component is

found by dividing its respective weight by the total wdght of the system,
and then multiplying by 100: . , ^ , . nrs
Weight per cent of A = — X 100

m
This method of expressing compositions is very commonly employed

for solid systems, and also for liquid systems. -It is not used commonly
for gaseous systems. Percentage figures applying to a sohd or to a
liquid system may be assumed to be in weight per cent, if there is no
definite specification to the contrary. One advantage of expressing
composition on the basis of weight per cent is that the composition values
CHAP 1] MOLE FRACTION AND MOLE PER CENT 13
> do not change if the temperature of the system is varied (assuming
there is no loss of material through volatilization or crystallization, and
that no chemical reactions occur). The summation of all the weight
percentages for a given system of necessity totals exactly 100.
Volumetric Per Cent. The per cent by volume of each component
is found by dividing its pure component volume by the total volume of
the system, and then multiplying by 100.
-m
Volumetric per cent of A = ( — j X 100
This method of expressing compositions is almost always used for

gases at low pressures, occasionally for liquids (particularly for the ethyl
alcohol:water system), but very seldom for sohds.
The analysis of gases is carried out at room temperature and atmos-
pheric pressure. Under these conditions, the behavior of the mixture
and of the individual gaseous components is nearly ideal, and the sum of
the pure component volumes will equal the totaj volume. That is,
VA + VB + • • • = V. This being the case, the percentages total ex-
actly 100. Furthermore, since changes of temperature produce the
same relative changes in the respective paxtial volumes as in the total
volume, the volumetric composition of the gas is unaltered by changes
in temperature. Compositions of gases are so commonly given on the
basis of volumetric percentages that if percentage figures are given with
no specification to the contrary, it may be assumed that they are per
cent by volume.
With Uquid solutions, it is common to observe that on mixing the pure
components a shrinkage or expansion occurs. In other words, the sum
of the pure component volumes does not equal the sum of the individual
volumes. In such instances, the percentages will not total exactly 100,
Furthermore, the expansion characteristics of the pure components
usually are not the same, and are usually different from that of the
mixture. This being the case, the volumetric composition of a liquid
solution will change with the temperature. Accordingly, a figure for
volumetric per cent as applied to a Uquid solution should be accompanied
by a statement as to the temperature. For the alcohol :water system,
the volumetric percentages are normally given with respect to a tempera-
ture of 60°F. If the actual determmation is made at a temperature
other than 60°F, a suitable correction is applied.
Mole Fraction and Mole Per Cent. If the components A and B are
compounds, the system is a mixture of two kinds of molecules. The
total number of A molecules or moles present divided by the sum of
the A and the B molecules or moles represents the mole fraction of A
14 STOICHIOMETRIC PRINCIPLES [QnAP. I
in the system. By multiplying the mole fraction by 100, the mole per
cent of A in the system is obtained. Thus,
Mole fraction of A = ——^ ,;,

MA/MA + VIB/MB
J Mole per cent of A = Mole fraction X 100
The summation of all the mole percentages for a given system totals
exactly 100. The composition of a system expressed in mole per cent
will not vary with the temperature, assuming there is no loss of material
from the system, and that no chemical reactions or associations occur.
Illustration 4. An aqueous solution contains 40% Na2C03 by weight. Expre^
the composition in mole per cent.
Basis of Calculation: 300 grama of solution.
Molecular Weights:
NajCOa = 106.0 H2O = 18.02
NasCOs present = 40 gm, or 40/106 = 0.377 gm-moles
H2O present = 60 gm, or 60/18,02 = 3.33 gm-moles
Total 3.71 gm-moles
Mole per cent Na^COa = (0.377/3.71) X 100 = 10.16
Mole per cent H2O = (3.33/3.71) X 100 = 89.8
100.0
Illustration 5. A solution of naphthalene, CioHs, in benzene, CcHc, contains 25
mole per cent of naphthalene. Express the composition of the solution in weight per
cent.
Basis of Calculation: 100 gm-moles of solution.
Molecular Weights:
CioHs = 128.1 , CeHe = 78.1
CioHs present = 25 gm-moles, or 25 X 128.1 = 3200 gm
CeHe present = 75 gm-moles, or 75 X 78.1 = 5860 gm
Total 9060 gm
Weight per cent of CioHs = (3200/9060) X 100 = 35.3
Weight per cent of CcHe = (5860/9060) X 100 = 64.7
• 100.0
In the case of ideal gases, the composition in mole per cent is exactly
the same as the composition in volumetric per cent. This deduction
follows from a consideration of Avogadro's law. It should be emphasized
that this relation holds only for gases, and does not apply to liquid or to
solid systems. '
CHAP. I] ATOMIC FRACTION AND ATOMIC PER CENT 15
Illustration 6. A natural gas has the following composition, all figures being in '
volumetric per cent:
Methane, CH4 83.5%
Ethane, C2H6 12.5%
Nitrogen, N2 4.0%
100.0%
Calculate: .;.-••- .-. • -..•... «.
(a) The composition in mole per cent. M s
(6) The composition in weight per cent. •
(c) The average molecular weight.
(d) Density at standard conditions, as pounds per cubic foot.
Part (a) It has been pointed out that for gaseous substances, the composition
in mole per cent is identical with the composition in volumetric per cent. Accord-
ingly, the above figures give the respective mole per cents directly, with no calcula-
tion.
Part (6) Calculation of Composition in Weight Per Cent.
Basis of Calculation: 100 lb-moles of gas.
Molecular ,-• ^ '-^
Lb-Moles Weight Weight in Pounds Weight Per Cent
CH4 83.5 16.03 83.5 X 16.03 = 1339 (1339/1827) X 100 = 73.3
C2H6 12.5 30.05 12.5 X 30.05 = 376 (376/1827) X 100 = 20.6
N2 4.0 28.02 4.0 X 28.02 = J 1 2 (112/1827) X 100 = 6.1
100.0 1827 100.0
Part (c) The molecular weight of a gas is numerically the same as the weight in
pounds of one pound-mole. Therefore, the molecular weight equals 1827/100, or
18.27.
Part (d) Density at Standard Conditions, as lb per cu ft. J
' Volume at standard conditions = 100 X 359 = 35,900 cu ft
Density at standard conditions =- 1827/35,900 = 0.0509 lb per cu ft
Atomic Fraction and Atomic Per Cent. The general significance of

these terms is the same as for mole fraction and mole per cent, except
that the atom is the unit under consideration rather than the molecule.
Thus, • . .
Atomic fraction of A = ——
' , .•, ymAlAx) + (mB/Afl)
Atomic per cent of A = Atomic fraction X 100
The summation of all of the atomic percentages for a given system is

exactly 100. The composition, expressed in atomic per cent, will not
vary with temperature, provided that no loss of material occurs. The
composition of a system expressed in atomic per cent will remain the
same regardless of whether or not reactions occur within the system.
16 STOICHIOMETRIC PRINCIPLES (CHAP. I
Mass of Material per Unit Volume of the System. Various units

are employed for mass and for volume. Masses are commonly expressed
in grams or pounds and the corresponding gram-moles or pound-moles.
For volume, the common units are liters, cubic feet, and U. S. gallons.
Some common combinations for expression of compositions are grams
per liter, gram-moles per liter, pounds per U. S. gallon, and pound-moles
per U. S. gallon.
This general method of indicating compositions finds its widest appli-
cation in dealing with liquid solutions, both in the laboratory and in
plant work. This is primarily due to the ease with which liquid volumes
may be measured.
Mass of Material per Unit Mass of Reference Substance. One
component of the system may be arbitrarily chosen as a reference
material, and the composition of the system indicated by stating the
mass of each component associated with unit mass of this reference
material. For example, in dealing with binary liquid systems, com-
positions may be expressed as mass of solute per fixed mass of solvent.
Some of the common units employed are: %.
1. Pounds of solute per pound of solvent.
2. Pound-moles of solute per pound-mole of solvent.
3. Pound-moles of solute per 1000 pounds of solvent. , •
The concentration of a solution expressed in the latter units is termed
its molality.
In dealing with problems involving the drying of solids, the moisture
content is frequently indicated as pounds of water per pound of moisture-
free material. In deahng with mixtures of condensable vapors and so-
called permanent gases, the concentration of the condensable vapor
may be indicated as pounds of vapor per pound of vapor-free gas, or as
pound moles of vapor per pound-mole of vapor-free gas.
In all the instances cited, the figure which indicates the composition is,
in reality, a dimensionless ratio; hence the metric equivalents have the
same numerical value as when the above-specified English units are
employed.
For processes involving gain or loss of material, calculations are
simphfied if the compositions are expressed in this nvanner. In-instances
of this kind, the reference component chosen is one which passes through
the process unaltered in quantity. Compositions expressed in these
terms are independent of temperature and pressure.
Illustration 7. A solution of sodiuih qhloride in water contains 230 grams of NaCl
per liter at 20°C. The density of the solution at this temperature is 1.148 gm/cc.
Calculate the following items: .. >
CHAP. I] D E N S I T Y A N D S P E C I F I C GRAVITY 17
(a) The composition in weight per cent. • , .

(6) The volumetric per cent of water. 'm*,^ ' :>
(c) The composition in mole per cent.
(d) The composition in atomic per cent.
(e) The molahty. . . . .
(/) Pounds NaCl per pound H2O. t^>;>t . H.;';;Ui,K\,i'4r :.:IJ - -ri. . , ••
Basis of Calculation: 1000 cc of solution.

Total weight = 1000 X 1.148 1148 gm
' NaCl = 230 gm, or 230/58.5 3.93 gm-moles
H2O = 1148 - 230 = 918 gm, or 918/18.02.... 50.9 gm-moles
Total 54.8 gm-moles
(a) Composition in Weight Per Cent:
Weight per cent NaCl = (230/1148) X 1 0 0 . . . . 20.0
Weight per cent H2O = (918/1148) X 100 80.0
100.0
(6) Volumetric Per Cent W a t e r : , ,
Density of pure water at 20°C = 0.998 gm/cc - -'•*"'-^
Volume of pure water = 918/0.998 = 920 cc
Volumetric per cent of water = (920/1000) X 100 = 92.0
(c) Composition in Mole Per Cent: .;

Mole per cent NaCl = (3.93/54.8) X 100 = 7.17 ., , |
Mole per cent HjO = (50.9/54.8) X 100 = 92.8 " |
I ;. i "^'>, •":•••, 100.0 , j • .| . . ,
(d) The Composition in Atomic Per Cent: .N^ |^ .
Gm-atoms of sodium 3.93
' Gm-atoms of chlorine 3.93
Gm-atoms of hydrogen = 2 X 50.9 101.8
Gm-atoms of oxygen 50.9
i Total 160.6
' Atomic per cent of sodium = (3.93/160.6) X 100 = 2.45
Atomic per cent of chlorine = (3.93/160.6) X 100 = 2.45
Atomic per cent of hydrogen = (101.8/160.6) X 100 = 63.4
Atomic per cent of oxygen = (50.9/160.6) X 100 = 31.7
100.0
(e) The Molahty:
Molality = lb-moles of NaCl/1000 lb H2O >'
= 3.93 X (1000/918) = 4.28 , ,
(/) Lb NaCI/lb H2O = 230/918 = 0.251
DENSITY AND SPECIFIC GRAVITY
Density is defined as mass per unit volume. Density'values are

commonly expressed as grams per cubic centimeter or as pounds per
cubic foot. The density of water at 4°C is 1.0000 gm/cc, or 62.43
Ib/cuft. ,,,iyrj .'.•,r;. t^,'. ...• ^. ^ ^b r ; n , G ; p . m r Vf! J / l j :p ••. - i y
The specific gravity of a solid or liquid is the ratio of its density to the
density of water at some specified reference temperature. The tempera-
tures corresponding to a value of specific gravity are generally symbolized
by a fraction, the numerator of which is the temperature of the liquid in
question, and the denominator, the temperature of the water whidh
serves as the reference. Thus the term sp. gr. 70/60°F indicates the
specific gravity of a hquid at 70°F referred to water at 60°F, or the
ratio of the density of the Hquid at 70°F to that of water at 60°F. It
is apparent that if specific gravities are referred to water at 4°C (39.2°F)
they will be numerically equal to densities in grams per cubic centimeter.
Concentration, ft NaCl by Weight

Concentration, grams NaCl
per 100 cc. of Solution
1.24
0.96
0 5 10 15 20 25 30 35
Concentration, % NaCl by Weight, also grams NaCl per 100 cc.
FIG. 1. Densities of aqueous sodium chloride solutiqps.

The densities of solutions are functions of both concentration and
temperature. The relationships between these three properties have
been determined for a majority of the common systems. Such com-
pilations as the International Critical Tables contain extensive tabula-
tions giving the densities of solutions of varying concentrations at speci-
fied temperatures. These data are most conveniently used in graphical
form in which density is plotted against concentration. Each curve on
such a chart will correspond to a specified, constant temperature. The
density of a solution of any concentration at any temperature may be
readily estimated by interpolation between these curves. In Fig. 1
*.,
CHAP. I] DENSITY AND SPECIFIC GRAVITY 19
are plotted the densities of solutions of sodium chloride at various tem-

peratures.
For a given system of solute and solvent the density or specific gravity
at a specified temperature may serve as an index to the concentration.
This method is useful only when there is a large difference between the
densities of the solutions and the pure solvent. In several industries
specific gravities have become the universally accepted means of indicat-
ing concentrations, and products are purchased and sold on the basis of
specific gravity specifications. Sulfuric acid, for example, is marketed
almost entirely on this basis. Specific gravities are also made the basis
for the control of many industrial processes in which solutions are in-
volved. To meet the needs of such industries, special means of numeri-
cally designating specific gravities have been developed. Several scales
ar^ in use in which specific gravities are expressed in terms of degrees
which are related to specific gravities and densities by more or less com-
plicated and arbitrarily defined functions.
Baume Gravity Scale. Two so-called Baume gravity scales are in
common use, one for use with hquids lighter and the other for hquids
heavier than water. The former is defined by the following expression:
• = - / . . . 140
• »' ' Degrees Baume = — 130
where G is the specific gravity at 60/60°F. It is apparent from this

definition that a liquid having the density of water at 60°F (0.99904
gram per cubic centimeter) will have a gravity of 10° Baume. Lighter
liquids will have higher gravities on the Baum6 scale. Thus, a material
having a specific gravity of 0.60 will have a gravity of 103° Baum6.
The Baume scale for liquids heavier than water is defined as follows:
145
Degrees Baume = 145 —
G
Gravities expressed on this scale increase with increasing density.

Thus, a specific gravity of 1.0 at 60/60°F corresponds to 0.0° Baum6,
and a specific gravity of 1.80 corresponds to 64.44° Baum6. It will be
noted that both the Baume scales compare the densities of liquids at
60°F. In order to determine the Baume gravity, the specific gravity
at 60/60°F must either be directly measured or estimated from specific-
gravity measurements at other conditions. The Baum6 gravity of a
liquid is thus independent of its temperature. Readings of Baum6
hydrometers at temperatures other than 60°F must be corrected for
temperature so as to give the value at 60°F.
API Scale. As a result of confusion of standards by instrument

manufacturers, a special gravity scale has been adopted by the American
Petroleum Institute for expression of the gravities of petroleum products.
This scale is similar to the Baume scale for liquids lighter than water as
indicated by the following definition:
141 5
,ft.,.. ;;. Degrees API = - - ^ - 131.5
G-
As on the Baum6 scale, a liquid having a specific gravity of 1.0 at 60/60°
F has a gravity of 10°. However, a liquid having a specific grayity
of 0.60 has an API gravity of 104.3. as compared with a Baume gravity
of 103.3. The gravity of a Hquid in degrees API is determined by its
density at 60°F and is independent of temperature. Readings of API
hydrometers at temperatures other than 60°F must be corrected for
temperature so as to give the value at 60°F.
API gravities are readily converted by Fig. B in the Appendix.
Twaddell Scale. The Twaddell scale is used only for liquids heavier
than water. Its definition is as follows: .
DegreesTwaddell = 200((r - 1.0) '
This scale has the advantage of a very simple relationship to specific
gravities.
Numeroxis other scales have been adopted for special industrial uses;
for example, the Brix scale measures directly the concentration of sugar
solutions.
If the stem of a hydrometer graduated in specific-gravity units is
examined, it is observed that the scale divisions are not uniform. The
scale becomes compressed and crowded together at the lower end. On
the other hand, a Baume or API hydrometer will have uniform scale
graduations over the entire length of the stem.
For gases, water is unsatisfactory as a reference material for expressing
specific-gravity values because of its high density in comparison with the
density of gas. For gases, it is conventional to express specific-gravity
values with reference to dry air at the same temperature and pressure as
the gas.
Triangular Plots. When dealing with mixtures or solutions of three
Components equivalent values of any particular property of the solution
can be shown related to composition by contour lines drawn upon a
triangular coordinate diagram.
In Fig. 2 is shown such a diagram with contour lines of specific volumes
at 25°C constructed for the system carbon tetrachloride, ethyl dibro-
mide, and toluene. On a triangular chart covering the complete range
CHAP. I] TRIANGULAR PLOTS 21
of compositions, apex A represents pure component A, in this instance
carbon tetrachloride having a specific volume of 0.630, apex B represents
pure component B, in this instance ethyl dibromide having a specific vol-
ume of 0.460, and apex C represents pure component C, in this instance
toluene having a specific volume of 1.159. Any point on the base line
AB corresponds to a binary solution of A and B. For example point
a represents 75% C2H4Br2 and 25% CCU, this composition having a
C=100%C,H8
Percentage Ceurbon Tetrachloride
Fia. 2. Specific volumes of ternary solutions of carbon tetrachloride, ethyl dibromide

and toluene at 25/4°C. (From International Critical Tables, III, 196.)
specific volume of 0.50. Similarly, base lines BC and CA represent

all possible combinations of the binary solutions of B and C, and of C
and A, respectively.
Any point within the area of the triangle represents a definite composi-
tion of a ternary mixture of A, B, and C. For example, point b corre-
sponds to a mixture of 50% A, 35% B, and 15% C. From the scale of
compositions it will be seen that point 6 may be considered as located on
the intersection of three fines, parallel to the three base fines, respectively.
The line parallel to BC passing through h is 50% of the perpendicular.

distance from line BC to A, the line parallel to AC passing through b is
35% of the distance from line AC to B, and the line parallel to AB pass-
ing through b is 15% of the distance from hne AB to C.
Contour lines on a triangular plot represent equivalent values of some
property, in this case specific volume. For example line cd, marked
0.65, represents all possible solutions having a specific volume of 0.65,
Point b lies on this line and corresponds to 50% A (CCI4), 35% B
(C2H4Br2) and 15% C (CvHs), having a specific volume of 0.65.
Triangular co-ordinate charts are useful for following t h e ' chaages
taking place in composition and properties of ternary systems in oper-
ations of extraction, evaporation, and crystallization, as illustrated in
Chapter V. For example, line Bb represents the change in composition
of the solution as component B alone is removed or added to some solu-
tion initially falling on this line. Thus, moving in a straight line from
an initial point within the diagram toward one apex represents the
change in composition as one component only corresponding to the
given apex is added to the initial solution. *»
Illustration 8. It is desired to calculate the final specific volume when 20 grams

of C2H4Br2 are added to 100 grams of solution corresponding to b on Fig. 2.
The resultant solution will contain
20 + 35
JQ = 45.8% C^HiBr,
This point lies on line hB and will be seen to correspond to a specific volume of 0.62
at 12.5% C7H8 and 41.7% CCU.
In the preceding, a triangular chart covering the entire possible

range of percentage compositions is described. Frequently only a por-
tion of the chart is required or the scale on each base line is altered to a
narrow range of composition. Under these circumstances the scale
used defines the significance of new apices and base lines.
Conversion of Units. The conversion of units arid symbols from one
system to another often presents a troublesome operation in technical
calculations. Both the metric and English units are intentionally em-
ployed in this book in order to bridge the gap between scientific and
industrial applications.
In nearly every handbook tables of conversion factors will be found,
and these are recommended for use whenever available and ^.dequate.
A short list of the more important factors is included in the Appendix,
CHAP. I] CONVERSION OF UNITS 23
page 438. A few simple rules will be given for guidance where calcu-
lation of conversion factors becomes necessary.
Most scientific units may be expressed in terms of simple dimensions,
such as length, weight, time, temperature, and heat. In conversion
the unit is first expressed in terms of its simplest dimensions combined
with the known numerical or symboHc value of the unit. Thus, the
viscosity of a liquid is n grams per second-centimeter. In the English
system the value will be expressed in pounds per second-foot. Each
of the dimensions is replaced separately by the dimensions of the desired
system together with its corresponding conversion factor. Thus, since
1 gram = 0.002204 lb and 1 cm = 0.0328 ft
grams 0.002204 lb lb
(sec) (cm) = '^
fi -——rTrT^^^^^^"^^!?
1 (sec) 0.0328 (ft) ~' 0.0670/i
'^ (sec) (ft)
Similarly a pressure of 1 atmosphere = . i^..
14.7 lb 14.7 (453.6) grams _ grams
(in.)2 "" (2.54)2 (cm)2 " (cm)^
since 1 lb = 453.6 grams and 1 in. = 2.54 cm
The gas constant R =
82.06 (atm) (cm)« _ (82.06) (1 atm) (0.0328 ft)^ _ (atm) (ft)'
(gram mole) (°K) ~ (0.002205 lb-mole) (1.8°R) ~ ' '^ (lb-mole)°R
Also since 1 atmosphere = 14.7 lb per in.^ = 14.7 X 144 or 2120 lb
per ft^,
^ ^ (0.729) (2120 Ib/ft^) (ft)^ ^ ^^^^ (ft) (lb)
(lb-mole) (°R) (lb-mole) ("R)
PROBLEMS
Tables of Common Atomic Weights and Conversion Factors will be found in
the Appendix.
While the simple stoichiometric relations included in the following group of
problems may easily be solved by the rules of ratio and proportion, it is nevertheless
recommended that the molal method of calculation be adhered to as a preparation
for the more complex problems to be encountered in succeeding chapters.
In all instances, the basis of calculation should be stated definitely at the start of
the solution.
1. BaClj + NajSOi = 2NaCl -t- BaSO*.
(a) How many grams of barium chloride will be required to react with 5.0
grams of sodium sulfate?
(6) How many grams of barium chloride are required for the precipitation of
5.0 grams of barium sulfate?
24 » > STdiCHlOMETRIC PRINCIPLES [CHAP. I
(c) How many grams of barium chloride are equivalent to 5.0 grams of sodium
chloride?
(d) How many grams of sodium sulfate are necessary for the precipitation of
the barium of 5.0 grams of barium chloride?
(e) How many grams of sodium sulfate have been added to barium chloride if
i 5.0 grams of barium sulfate are precipitated?
(/) How many pounds of sodium sulfate are equivalent to 5.0 pounds of
sodium chloride?
(g) How many pounds of barium sulfate are precipitated by 5.0 pounds of
barliim chloride?
' (h) How many pounds of barium sulfate are precipitated by 5.0 pounds of
sodium sulfate?
(t) How many pounds of barium sulfate are equivalent to 5.0 pounds of sodium
chloride?
2. How many grams of chromic sulfide will be formed from 0.718 grams of chromic
oxide according to the equation:
2Crj08 + 3CS2 = 2Cr5S, + 3COs
3. How much charcoal is required to reduce 1.5 pounds of arsenic trioadel
AsjOa + 30 = 3CO + 2As
4. Oxygen is prepared according to the equation 2KCIO3 = 2KC1 + SOs. What

is the yield of oxygen when 7.07 grams of potassium chlorate are decomposed? How
many grams of potassium chlorate must be decomposed to liberate 2.0 grams of
oxygen?
6. Sulfur dioxide may be produced by the reaction: ,
Cu + 2H2S04 = CuSOi + 2H2O + SOj ' "
(a) How much copper, and (6) how much 94% H2S04 must be used to Obtain
32 pounds of sulfur dioxide?
6. A limestone analyzes CaCOa 93.12%, MgCOa 5.38%, and insoluble matter
1.50%.
(o) How_ many pounds of calcium oxide could be obtained from 5 tons of the
limestone?
(6) How many pounds of carbon dioxide are given off per pound of this lime-
stone?
7. How much superphosphate fertilizer can be made from one ton of calcium
phosphate, 93.5% pure? The reaction is
Cas(P04)2 + 2H2S04 = 2CaS04 + CaHiCPO,),
8. How much potassium chlorate must be taken to produce the same amount of
oxygen as will be produced by 1.5 grams of mercuric exide?
9. Regarding ammonium phosphomolybdate, (NH4)3p04-12MoOs'3H20, as made
up of the radicals NHj, H2O, PjOs and M0O3, what is the percentage composition of
the molecule with respect to these radicals?
10. How many pounds of salt are required to make 1500 pounds of salt cake
(NajSO*)? How many pounds of Glauber's salt (Na2SO4'10H2O) will this amount
of salt cake make? , / '
• * % ••"Hw(i4<
CHAP. I] PROBLEMS ''«ffe 25
11. In the reactions:
2KMn04 + 8H2SO4 + lOFeSOi = SFesCSO,), + K^SO^ + 2MnS0i + 8H2O
K j C r A + 7H2SO4 + GFeSOi = SFcCSOi), + K2SO4 + CrsCSOi)^ + 7HjO
How many grams of potassium diohromate are equivalent to 5.0 grams of potassium
permanganate? How many grams of potassium permanganate are equivalent to
3.0 grams of potassium diohromate?
12. Aoompoundwhosemolecular weight is 103 analyzes: C = 81.5%; H = 4.9%;
N = 13.6%. Whatis its formula?
13. What is the weight of one liter of methane under standard conditions?
14. If 15 grams of iron react thus: Fe + H2SO4 = FeSOi + Ha, how many liters
of hydrogen are liberated at standard conditions?
15. A solution of sodium chloride in water contains 23.0% NaCl by weight.
Express the concentration of this solution in the following terms, using data from
Fig. 1.
(a) Gram-moles of NaCl per 1000 grams of water (molality).
(6) Mole fraction of NaCl.
(c) Gram-moles of NaCl per liter of solution at 30°C.
{d) Pounds of NaCl per U. S. gallon of solution at 40°C.
16. In the following table are listed various aqueous solutions, with the density
at 20°C and 80°C given, as weU as the composition. For one or more of the solu-
tions assigned from the table, report the composition expressed in the following ways:
(a) Weight per cent.
(6) Mole per cent. > < 1 t,. ^ ; ./
(c) Poimds of solute per pound of solvent.
(dl Pound-moles of solute per pound of solvent. 'i '"'"• • .; </
(ef Grams solute per 100 ml of solution at 80°C. ,. , , „ , , , , ,
(/) Grams solute per 100 ml of solution at 20°C. " " ' '
{g) Gram-moles solute per liter of solution at 20°C.
{h) Pounds of solute per U. 8. gallon of solution at 68°P (20°C).
(i) Pound-moles of solute per U. S. gallon of solution at 68°F (20°C).
(j) Molality.
(fc) Normality. -
Density, g/ml
Solute Composition of Solution
20°C 80°C
HCl Weight % HCl = 30 1.149 1.115

HjSOi mole % H28O4 = 36.8 1.681 1.625
HNO3 lb HNOa/lb H2O = 1.704 1.382 1.296
NH4NO3 lb-mole NHiNOs/lb H2O = 0.01250 1.226 1.187
ZnBrs gm ZnBrj/lOO ml of solution at 20°C = 130.0 2.002 1.924
CdClj gm CdClj/lOO ml of solution a t 80°C = 68.3 1.575 1.519
MgCU gm-mole M g C U / l of solution at 20°C = 3.99 1.269 1.245
CaBrj lb CaBrj/U. S. gal at 68°F (20°C) = 4.03 1.381 1.343
SrClj lb-mole SrClj/U. S. gal at 68°F (20°C) = 0.02575 1.396 1.368
LiCl molality = 10.13 1,179 1.158
KQ normality = 2.70 1.118 1.088
17. An aqueous solution of sodium chloride contains 28 grams of NaCl per 100 cc
of solution at 20°C. Express the concentration of this solution in the following
terms, using data from Fig. 1.
(a) Percentage NaCl by weight.
(6) Mole fraction of NaCl.
(c) Pound-moles of NaCl per 1000 pounds of water (molality).
(d) Pound-moles of NaCl per U. S. gallon of solution at 0°C.
18. It is desired to prepare a solution of sodium chloride in water, having a molality
of 2.00. Calculate the weight of sodium chloride which should l5e placed in a 1000-ec
volumetric flask in order that the desired concentration will be obtained by subse-
quently filling the flask with water, keeping the temperature of the solution at 30°C.
19. For the operation of a refrigeration plant it is desired to prepare a solution of
sodium chloride containing 20% by weight of the anhydrous salt.
(a) Calculate the weight of sodium chloride which should be added to one
gallon of water at 30°C in order to prepare this solution.
(6) Calculate the volume of solution formed per gallon of water used, keeping
the temperature at 30 °C.
20. (o) A solution has a gravity of 80° Twaddell. Calculate its specific gravity
and its gravity in degrees Baum^.
(b) An oil has a specific gravity at 60/60°F of 0.651. Calculate its gravjty in
degrees API and degrees Baum6.
21. Make the following conversion of units:
(a) An energy of 8 ft-lb to kilogram-meters.
(6) An acceleration of 32.2 - — - to - — — .
(sec)2 (sec)2
(c) A pressure of 100 mm of Hg to inches of water.
, ,s r^, , , . . ,., Btu kcal
(d) Thermal conductivity of k (hr)rF)(ft)
• to -(hr)(°C)(m)'
, , n^ ^ , 82.06 (atm) (cm)' ^ Btu
(e) The gas constant R, from —(g-mole)(°K) - to •(lb-mole) (°R)
CHAPTER II
BEHAVIOR OF IDEAL GASES

In the preceding chapter consideration was given to problems
pertaining to the transformation of matter from one physical or chemical
state to another. It has been pointed out that matter is essentially
indestructible despite all the transformations it may undergo and that
the mass of a given system remains unaltered. However, in order to
complete the quantitative study of a system it is necessary to consider
one other property.
The properties of a moving ball, a swinging pendulum, or a rotating
flywheel are different from those of the same objects at rest. The differ-
ences lie in the motion of the bodies and in the ability of the moving
objects to perform work, which is defined as the action of a force moving
under restraint through a distance. Likewise, the properties of a red-
hot metal bar are different from those of the same bar when cold. The
red-hot bar produces effects on the eye and the touch very different from
those of the cold bar. The essential property or physicochemical concept
necessary to complete the quantitative description of matter is termed
energy.
Energy. All matter and the properties of matter are manifestations
of energy. Energy is the capacity of matter to perform work and to
affect the senses. For example, the motions of bodies cited above,
the warming of an object by solar radiation, and the change in com-
position of a storage battery when it generates electricity are all mani-
festations of energy. Energy is distributed throughout the universe in a
variety of forms, all of which may be directly or indirectly converted
into one another.
Under the classification of -potential energy are included all forms of
energy not associated with motion but resulting from the position and
arrangement of matter. The energy possessed by an elevated weight,
a compressed spring, a charged storage battery, a tank of gasohne, or
a lump of coal are examples of potential energy. Similarly, potential
energy is stored within an atom as the result of forces of attraction
among its subatomic parts. Thus potential energy can be further
classified as external potential energy, which is inherent in matter as a,
result of its position relative to the earth, or as internal potential energy,
which resides within the structure of matter.
27
28 BEHAVIOR OP IDEAL GASES [CHAP. II
In contrast, energy associated with motion is referred to as kinetic

energy. The energy represented by the flow-of a river, the flight of a
bullet, or the rotation of a flywheel are examples of kinetic energy.
Also individual molecules possess kinetic energy by virtue of their trans-
lational, rotational, and vibrational motions. Similar to the sub-
classification of potential energy, kinetic energy is subclassifiecj as in-
ternal kinetic energy, such as associated with molecular and atomic
structure, and as external kinetic energy, such as associated with the
external motion of visible objects.
In addition to the forms of energy associated with composition,
position, or motion of matter, energy exists in the forms of electricity,
magnetism, and radiation, which are associated with electronic phe-
nomena.
The science pertaining to the transformation of one form of energy
to another is termed thermodynamics. Early studies of the transforma-
tion of energy lead to the realization that although energy can be trans-
formed from one form to another it can never be destroyed and that the
total energy of the universe is constant. This principle of the con-
servation of energy is referred to as the first law of thermodynamics. ]V>.any
experimental verifications have served to establish the^ validity of this
law.
Temperature and Heat. Energy may be transferred not only from
one form to another but also from one aggregation of matter to another
without change of form. The transformation of energy from one form
to another or the transfer of energy from one body to another always
requires the influence of some driving force. As an example, if a hot
metal bar is placed in contact with a cold one, the former will be cooled
and the latter warmed. The sense of " hotness " is an indication of the
internal kinetic energy of matter. The driving force which, even in
the absence of electrical, magnetic, or mechanical forces, produces a
transfer of energy is termed temperature and that form of energy which
is transferred from one body to another as a result of a difference in
temperature is termed heat. /
The Kinetic Theory of Gases. As the basis of the kinetic theory /
it is assumed that all matter is composed of tiny particles which by their jI
behavior determine its physicalând chemical properties. A gas is be- /'
Ueved to be composed of molecules each of which is a material body
and separate from all others. These particles are free to move about
in space according to Newton's laws of the motion of material bodies.
•It is furthermore assumed that each particle behaves as a perfectly
elastic sphere, ^s a consequence of this assumption there is no change
in total kinetic energy or momentum when two particles colhde or when
CHAP. II] THE KINETIC THEORY OF GASES 29
a particle strikes an obstructing or confining surface. On the basis of

these assumptions it is possible to explain many physical phenomena by
considering that each particle of matter is endowed with a certain inherent
kinetic energy of translation. As a result of this energy the particles
will be in constant motion, striking against and rebounding from one
another and from obstructing surfaces.
The energy which is represented by the sum of the energies of the com-
ponent particles of matter is termed the total internal energy. When
heat is added to a gas, additional kinetic energy is imparted to its com-
ponent particles. The average quantity of kinetic energy of translation
which is possessed by the particles of a gas determines its temperature.
At any 'specified temperature the particles of a gas possess definitely
fixed, average kinetic energies of translation which may be varied only
by a change in temperature resulting from the addition or removal of
heat. Thus, an increase in temperature signifies an increase in average
kinetic energy of translation which in turn is accompanied by increased
speeds of translation of the particles. Conversely, when, by any means,
the kinetic energies of translation of the particles of a gas are in-
creased, the temperature is raised.
The theory outlined above accounts for the pressure which is exerted
by a gas against the walls of a confining vessel. The translational mo-
tion of the particles is assumed to be entirely random, in every direction,
and it laaiy be assumed for ordinary cases that the number of particles
per unit volume will be constant throughout the space. These assump-
tions are justified when the number of particles per unit volume is very
large. Then, each element of area of confining surface will be sub-
jected to continual bombardment by the particles adjacent to it. Each
impact will be accompanied by an elastic rebound and will exert a pres-
sure due to the change of momentum involved. In a pure, undissociated
gas all particles may be considered to be of the same size and mass. On
the basis of these assumptions the following expression for pressure
may be derived from the principles of mechanics.'
where y= number of molecules under consideration rt >

V = volume in which v molecules are contained
m = mass of each molecule
u = average translational velocity of the molecules
' This derivation may hp found in simplified form in any good physics or physical
chemistry text or in more rigorous form in the more advanced books deaUng with
kinetic theory.
30 BEHAVIOR OF IDEAL GASES [CHAP. II
From the definition of the molal units of quantity it was pointed out
that one mole of a substance will contain a definite number of single mole-
cules, the same for all substances. Then
v = nN y (2)
where ^
n = number of moles in volume V
N = number of molecules in a mole, a universal con-
stant equal to 6.023 X 10^' for the gram-mole
Combining (1) and (2),
pV = n%N(imu^) = nfu, (3)
where Uj represents the total translational kinetic energy possessed by
one mole of gas.
From extensive experimental investigations the ideal gas law has
been empirically developed. In fact, the definition of the absolute
scale of temperature is based on this relationship.
pv = RT (4)
or
pV = nRT / (5)
where
i2 = a proportionality factor ^
T = absolute temperature .
V= volume of one mole'of gas
n = number of moles of gas
V = volume of n moles of gas
Rearranging (4)
K-f " m
Assuming the validity of the Avogadro principle that equimolal
quantities of all gases occupy the same volume at the same conditions
of temperature and pressure, it follows from Equation (6) that the gas
law factor 7i is a universal constant. The Avogadro principle and the
ideal gas law have been experimentally shown to approach perfect
validity for all gases under conditions of extreme rarefaction, that is,
where the number of molecules per unit volume is very small. The con-
stant R may be evaluated from a single measurement of the volume
occupied by a known molal quantity of any gas al a known temperature
and at a known reduced pressure. ^ ''
CHAP. II] EXTENSION OF THE KINETIC THEORY 31
Combining Equations (3) and (5):
| u . = ijr (7)
or
lmu^-l§T . (8)
Equation (7) states that the average kinetic energy of translation of
a molecule in the gaseous state is directly proportional to the absolute
temperature. The absolute zero is the temperature at which the kinetic
energies of all molecules become zero and molecular motion ceases.
From the fact that R, the gas law constant, and N, the Avogadro num-
ber, are universal constants, it follows that Equation (7) must apply to
all gases. I n other words, the average translational kinetic energy with
which a gas molecule is endowed is dependent only upon the absolute tem-
' perature and is independent of its nature and size. This conclusion is of
far-reaching significance. It follows that a molecule of hydrogen pos-
sesses the same avejage translational kinetic energy as does a molecule
of bromine at the same temperature. Since the bromine molecule has
eighty times the mass of the hydrogen molecule the latter must move at
a correspondingly higher velocity of translation. If the temperature
increases, the squares of the velocities of translation of both molecules
will be increased in the same proportion.
From the theory outlined in the preceding paragraphs it is possible
to form a definite mental picture of the mechanical nature of a gas. The
actual component parts of the gas are invisible and of an abstract and
rather theoretical nature. The kinetic theorj' merely presents a mechan-
ical analogy by which the phenomena of the gaseous state are explained
in terms of the familiar laws of energy and of the behavior of particles of
rigid matter of tangible dimensions. The analogy calls to mind a box
in which are contained energized, elastic marbles which are in constant
motion, colliding with one another and with the confining walls. An
increase in temperature merely signifies an increased velocity of motion
in each marble. A clear mental picture of such an analogy is of great
value in fully understanding the properties of matter and in making use
intelligently of thermodynamic relationships.
Extension of the Kinetic Theory. .Although the kinetic theory was
originally developed to explain the behavior of gases, it has been ex-
tended and found to apply with good approximation wherever small
particles of matter are permitted to move freely in space. It has been
shown that all such particles may be considered as endowed with the same
kinetic energy of translation when at the same temperature regardless of

composition or size. This principle is believed to apply not only to the
molecules of all gases but also to the molecules of all liquids and of sub-
stances which are dissolved in hquids. It has been extended still further
and shown to apply also to particles of solid matter of considerable size
suspended in gases or liquids. Thus, at any selected temperature, a
molecule of hydrogen gas, a molecule of iodine vapor, a molecule of
liquid water, and a molecule of liquid mercury all are supposed to possess
the same translational kinetic energy, indicated by Equation (3).
Furthermore, this same energy is possessed by a molecule of sulfuric
acid in solution in water and by each of the ions formed by the dissocia-
tion of such a molecule. A colloidal particle of gold, containing hun-
dreds of atoms, or a speck of dust suspended in air, each presumably has
the same translational kinetic energy as does a molecule of hydrogen gas
at the same temperature. The larger particles must therefore exhibit
correspondingly slower velocities of translational motion. This general-
ization is of the greatest importance in the explanation of such phenom-
ena as diffusion, heat conduction, osmotic pressure, and the general
behavior of colloidal systems. i.,
The Gas Law Units and Constants. In the use of the gas law equa-
tions great care must be exercised that consistent units are employed
for the expression of both the variable and constant terms. Tempera-
ture must always be expressed on an absolute scale. Two such scales
are in common use. The Kelvin scale corresponds, in the size of its
unit degree, to the Centigrade scale. The zero of the Centigrade scale
corresponds to 273.1 degrees on the Kelvin scale. Thus:
a;°C = (a; + 2 7 3 . 1 )°K (Kelvin) (9)

The Rankine scale of absolute temperature corresponds, in the size of
its unit degree, to the Fahrenheit scale. The zero of the Fahrenheit
scale corresponds to 460 degrees on the Rankine scale. Thus:
x°F = (x + 460)°R (Rankine) (10)
The Avogadro number N denoting the number of molecules in a mole
is one of the most important of physical constants and has been care-
fully determined by a variety of methods. The accepted value is 6.023
X 10^* for the number of molecules in one gram-mole, or 2.73 X 10^
molecules per pound-mole.
Equation (7) may be solved for the gas constant R:
CHAP. II] APPLICATIONS OF THE IDEAL GAS LAW 38
From Equation (11) it is seen that R represents two-thirds of the trans-

lational kinetic energy possessed by one mole per degree of absolute
temperature. The numerical value of R has been carefully determined
and may be expressed in any desired energy units. Following are values
corresponding to various systems of units:
UNITS OF PBBSSURE UNITS OF VOLUME R
Per gram-mole (temperatures: Kelvin)
Atmospheres Cubic centimeters 82.06
Per pound-mole (temperatures: Rankine) •• >

Pounds per square inch Cubic inches 18,510
Pounds per square inch Cubic feet ' 10.71
Atmospheres Cubic feet . • 0.729
APPLICATIONS OF THE IDEAL GAS LAW

When substances exist in the gaseous state two general types of prob-
lems arise in determining the relationships between weight, pressure,
temperature, and volume. The first type is that in which are involved
only the last three variables — pressure, temperature, and volume.
For example, a specified volume of gas is initially at a specified temper-
ature and pressure. The conditions are changed, two of the variables
in the final state being specified, and it is desired to calculate the third.
For such calculations it is not required to know the weight of the gas.
The second, more general type of problem involves the weight of the gas.
A specified weight of substance exists in the gaseous state under condi-
tions, two of which are specified and the third is to be calculated. Or,
conversely, it is desired to calculate the weight of a given quantity of
gas existing at specified conditions of temperature, pressure, and volume.
Problems of the first type, in which weights are not involved, may be
readily solved by means of the proportionality indicated by the gas law.
Equation (5) may be applied to n moles of gas at conditions pi, Vi, Ti
and also at conditions p2, V2, T2.
piVi = nRTi
P2V2 = nRTi
Combining: > • ••
This equation may be applied directly to any quantity of gas. If the

three conditions of state 1 are known, any one of those of state 2 may be
calculated to correspond to specified values of the other two. Any units
of pressure, volume, or absolute temperature may be used, the only re-
quirement being that the units in both initial and final states be the same.
Equation (5) is in form to permit direct solution of problems of the
second type, in which are involved both weights and volumes of gases.
With weights expressed in molal units the equation may be solved for
any one of the four variables if the other three are known. However,
this calculation requires a value of the constant R expressed in units to
correspond to those used in expressing the four variable quantities. So
many units of expression are in common use for each variable quantity
that a very large table of values of R would be required or.,else the vari-
able quantities would have to be converted into standard units. Either
method is inconvenient.
It proves much more desirable to separate such calculations into two
steps. As a primary constant, the normal molal volume is used instead
of R. The normal molal volume is the volume occupied by one mole of a
gas at arbitrarily selected standard conditions, assuming that the ideal
gas law is obeyed. The normal molal volume at any one set of standard
conditions, if the validity of Equation (5) is assumed, must be a universal
constant, the same for all gases. The volume, at the standard condi-
tions, of any weight of gas is the product of the number of moles present
and the normal molal volume. The general type of problem involving
weights and volumes at any desired conditions may then be solved in
two steps. In one, the differences between the properties of the gas at
standard conditions and at those specified in the problem are determined
by Equation (12). In the other step the relationship'between volume
at standard conditions and weight is determined by means of the nor-
mal molal volume constant.
Standard Conditions. An arbitrarily specified standard state of tem-
perature and pressure serves two purposes. It establishes the normal'
molal volume constant required in the system of calculation described in
the preceding section. It also furnishes convenient specifications under
which quantities of gases may be compared when expressed in terms of
volumes. Some such specification is necessary because of the fact that
the volume of a gas depends not only on the quantity but on the temper-
ature and pressure as well.
Several specifications of standard conditions are in more or less com-
mon use but the one most universally adopted is that of a temperature
of 0°C and a pressure of one atmosphere. It is recommended that these
conditions be adopted as the standard for all calculations. Under these
conditions the normal molal volumes are as follows (the abbreviation
S.C. is used to designate the standard conditions):
Volume of 1 gram-mole S.C. = 22.41 liters
Volume of 1 pound-mole S.C. = 359 cubic feet
CHAP. II] GAS DENSITIES AND SPECIFIC GRAVITIES 35
These important constants should be memorized. The conditions of

the standard state may be expressed in any desired units as in the follow-
ing tatfcle:
STANDAED CONDITIONS
Temperature Pressure
0°Centigrade 1 atmosphere
273°Kelvin ' 760 mm of mercury
32°Fahrenheit • 29.92 in. of mercury
492°Raiikine 14.70 lb per sq in.
There are many substances which cannot actually exist in the gaseous
state at these specified conditions. For example, at a temperature of
0°C water cannot exist in a stable gaseous form at a pressure greater
than 4.6 mm of mercury. Higher pressures cause condensation. Yet,
it is convenient to refer to the hypothetical volume occupied by water
vapor at standard conditions. In such a case the volume at standard
conditions indicates the hypothetical volume which would be occupied
by the substance if it could exist in the vapor state at these conditions
and if it obeyed the ideal gas law.
Gauge Pressure. All ordinary pressure gauges indicate the mag-
nitude of pressure above or below that of the atmosphere. In order to
obtain the absolute pressure which must be used in the gas law, the pres-
sure of the atmosphere must be added to the gaiige pressure. The
average atmospheric pressure at sea level is 14.70 pounds per square
inch or 29.92 inches of mercury.
•Gas Densities and Specific Gravities. The density of a gas is ordi-
narily expressed as the weight in grams of one Hter or the weight in
pounds of one cubic foot. Unless otherwise specified the volumes are at
the standard conditions of 0°C and a pressure of 1.0 atmosphere. On
this basis air has a normal density of 1.293 grams per hter or of 0.0807
pound per cubic foot.
The specific gravity of a gas is usually defined as the ratio of its density
to that of air at the same conditions of temperature and pressure.
The gas law expresses the relationship between four properties of a
gas: mass, volume, pressure, and temperature. In order to calculate
any one of these properties the others must be known or specified. Four
different types of problems arise, classified according to the property
being sought. The following illustrations show the application of the
recommended method of calculation to each of these types of problems.
For establishment of correct ratios to account for the effects of pres-
sure and temperature a simple rule may be followed which offers less
opportunity for error than attempting to recall Equation (12). The
ratio of pressures or temperatures should he greater than unity when the
changes in pressure or temperature are such as to cause increase in volume.

The ratios should be less than unity when the changes are such as to cause
decrease in volume.
Illustration 1 (Volume Unknoivn). Calculate the volume occupied by 30 lb of
chlorine at a pressure 743 mm of Hg and 70°F.
Basis: 30 lb of chlorine or 30/71 = 0.423 lb-mole.
Volume at S.C. = 0.423 X 359 = 152 cu ft.
Pa Ti
TO^F = 530°Raiikine.
7 Aft f^Qft
Volume at 743 mm Hg, 70°F = 152 X ~ ^ X — = 167 cu ft.
743 492
Illustration 2 (Weight Unknoim). Calculate the weight of 100 cu ft of water
vapor, measured at a pressure of 15.5 mm of Hg and 23°C.
Basis: 100 cu ft of water vapor at 15.5 mm Hg, 23''C.
15 5 273
Volume at S.C. = 10° ^ X 5 ^ = ^-^S cu ft.
Moles of H2O = 1.88 H- 359 = 0.00523 lb-mole.
Weight of H2O = 0.00523 X 18 = 0.0942 lb.
Illustration 3 {Pressure Unknown). It is desired to compress 10 lb of carbon
dioxide to a volume of 20 cu ft. Calculate the pressure in pounds per square inch
which is required at a temperature of 30°C assuming the appUcability of the ideal
gas law.
Basis: 10 lb of CO2 or 10/44 = 0.228 lb-mole.
Volume at S.C. = 0.228 X 359 = 81.7 cu ft.
From Equation (12):
P. = P i : ^ ^ X -
30°C = 303°K.
0 1 fj QftQ
Pressure at 20 cu ft, 30°C = 14.7 — X -=- = 66.6 lb per sq in.
^0 273
Illustration 4 (Temperature Unknown). Assuming the appUcability of the ideal
gas law, calculate the maximum temperature to which 10 lb of nitrogen, enclosed in
a 30 cu ft chamber, may be heated without the pressure exceeding 150 lb per sq in.
Basis: 10 lb of nitrogen or 10/28 = 0.357 lb-mole.
Volume at S.C. = 0.357 X 359 = 128.1 cu ft.
Pi Vi
Temperature at 30 cu ft, 150 Ib/sq in. =
150 30
^ " ^ 1 4 7 ^ 1281 ^^^^°-^°''^''''°*^
CHAP. II} GASEOUS MIXTURES 37
Dissociating Gases. Certain chemical compounds when in the gas-

eous state apparently do not even approximately follow the relation-
ships deduced above. The tendency of hydrogen fluoride to associate
into large molecules was mentioned in Chapter I. Ammonium chloride,
nitrogen peroxide, and phosphorus pentachloride exhibit an abnormality
. opposite in effect which has been definitely proved to result from disso-
ciation of the molecules into mixtures containing two or more other
compounds. Ammonium chloride molecules in the vapor state separate
into molecules of hydrogen chloride and ammonia:
NH4CI = NHs-f-HCl
Thus gaseous ammonium chloride is not a pure gas but a mixture of three
gases, NH4CI, HCl, and NH3. By decomposition, two gas particles are
produced from one, and the pressure or volume of the gas increases above
that which would exist had no decomposition taken place. For this
reason, when one gram-mole of ammonium chloride is vaporized the
volume occupied will be much greater than that indicated by Equation
(5). However, when proper account is taken of the fact that in the
gaseous state there is actually more than one gram-mole present, it is
found that the ideal gas law applies. Conversely, from the apparent
deviation from the gas law the percentage of dissociation can be calcu-
lated if the chemical reaction involved is known.
Illustration 5. When heated to 100°C and 720 mm pressure 17.2 grams of NjOi
gas occupy a volume of 11,450 cc. Assuming that the ideal gas law applies, calcu-
late the percentage dissociation of N2O4 to NO2.
17.2 ;• •'• '-J
Gram-moles of N2O4 initially present = - — = 0.187
(
Let X = gram-moles of N2O4 dissociated. Then 2x = gram-moles of NOj formed.
Total gram-moles present after dissociation =
„ „„ „ 11,450 273 720
Solving, X = 0.168.
Percentage dissociation = - ' X 100 = 90% n

0.187
' ' ' ' "' ; GASEOUS MIXTURES ^ '

In a mixture of different gases the molecules of each component gas
are distributed throughout the entire volume of the containing vessel
and the molecules of each component gas contribute by their impacts
to the total pressure exerted by the entire mixture. The total pressure
is equal to the sum of the pressures exerted by'the molecules of each
38 BEHAVIOR OF IDEAL GASES [CHAP II
component gas. These statements apply to all gases, whether or not

their behavior is ideal. In a mixture of ideal gases the molecules of
each component gas behave independently as though they alone were
present in the container. Before considering the actual behavior of
gaseous m.ixtures it will be necessary to define two terms commonly
employed, namely, -partial pressure and pure-component volume. By
definition, the partial pressure of a component gas which is present in a
mixture of gases is the pressure that would be exerted by that component
gas if it alone were present in the same volume and at the same temperature
as the mixture. By definition, the pure-component volume of a com-
ponent gas which is present in a mixture of gases is the volume that would be
occupied by that component gas if it alone were present at the same pressure
and temperature as the mixture.
The partial pressure as defined above does not represent the actual
pressure exerted by the molecules of the component gas when present
in the mixture except under certain Knuting conditions. Also, the pure-
component volume does not represent the volume occupied by the
molecules of the component gas when present in the mixtures, for obvi-
ously the molecules are distributed uniformly throughout the volume
of the mixture.
The pure-component volume generally has been termed partial
volume in the past. However, the latter term is currently used to
designate the differential increase in volume when a component is added
to a mixture. Pure-component volumes and partial volumes are not
necessarily the same except under ideal conditions.
Laws of Dalton and Amagat. From the simple kinetic theory of the
constitution of gases it would be expected that many properties of
gaseous mixtures would be additive. The additive nature of partial
pressures is expressed by Dalian's law, which states that the total pres-
sure exerted by a gaseous mixture is equal to the sum of the partial pressures,
that is: .
». p = PA + PB + Pc + • • • (13)
where p is the total pressure of the mixture and PA, PB, pc, etc., are the
partial pressures of the component gases as defined above.
Similarly, the additive nature of pure-component volumes is given by
the law of Amagat, or Leduc's law, which states that the total volume
occupied by a gaseous mixture is equal to the sum of the pure-component
volumes, that is:
V =VA + VB + VC + --- (14)
where V is the total volume of the mixture and VA, VB, VC, etc., are the
pure-component volumes of the component gases as defined above. It
CHAP. II] LAWS OF DALTON AND AMAGAT 39
will be shown later that each of these laws is correct where conditions
are such that the mixture and each of the components obey the ideal
gas law.
Where small molal volumes are encountered, such that the ideal gas
law does not apply either Dalton's or Amagat's law may apply, but
both laws apply simultaneously only for ideal gases. Under such con-
ditions pressures may not be additive, because the introduction of
additional molecules into a gas-filled container may appreciably affect
the pressure exerted by those already there. The presence of new
molecules will reduce the space available for the free motion of those
originally present and will exert attractive forces on them. Similarly,
if quantities of two gases at the same pressure are allowed to mix at
that same pressure, the like molecules of each gas will be separated by
greater distances and will be in the presence of unhke molecules, which
condition may alter the order of attractive forces existing between them.
As a result, the volume of the mixture may be quite different from the
sum of the original volumes. These same effects are present but
negligible under conditions of large molal volumes and wide separation
of molecules.
Where conditions are such that the ideal gas law is applicable: •;
VA = — ^ (15)
where y^.:\ ''•..': ~ •=. ^ . >..
V = total volume of mixture
n^ = number of moles of component A in mixture
Similar equations represent the partial pressures of components B, C,
etc. Combining these equations with Dalton's law, Equation (13):
V = {nA + riB+nc+ • • •)-^ (16)
This equation relates the pressure, temperature, volume, and molal

quantity of any gaesous mixture under such conditions that the mixture
and each of the components follow the ideal gas law and Dalton's law.
By combining Equations (15) and (16) a useful relationship between
total and partial pressure is obtained.
VA = ^ \ P = NAP " (17)

UA + nB + nc • • •
The quantity NA = nA/{nA + UB + nc + • • •) is the mole fraction of
component A. Equation (17) then signifies that, where the ideal gas law
may be applied, the partial pressure of a component of a mixture is equal
to the product of the total pressure and the mole fraction of that component.
40 BEHAVIOR OF IDEAL GASES [CHAr. II
Where conditions are such that the ideal gas law is applicable
"*'•'' VÂ = riART (18)
P F B = UBRT (19)
where YA, ^ B , etc., are the pure-component volumes as defined above.

Adding these equations,
P(FA + 7B H ) = {nA + nB+ •• •)RT (20)
Combining Equations (18) and (19) with Amagat's law (Equation 14),
V .: •^. V nA + nB + --^ ^^^^

or
VA = NAV (22)
Equation (22) signifies that, where the ideal gas law may be applied,
the pure-component volume of a component of a gaseous mixture is equal
to the product of the total volume and the mole fraction of that component.
From Equations (16) and (20) it is evident thtit when the ideal gas
law is valid both Amagat's and Dalton's laws apply, that is, both pure-
component volumes and partial pressures are additive.
Average Molecular Weight of a Gaseous Mixture. A certain group
of components of a mixture of gases may in many cases pass through a
process without being changed in composition or weight. For example,
in a drying process, dry air merely serves as a carrier for the vapor being
removed and undergoes no change in composition or in weight. It is
frequently convenient to treat such a mixture as though it were a single
gas and assign to it an average molecular weight which may be used for
calculation of its weight and volume relationships. Such an fiverage
molecular weight has no physical significance from the standpoint of the
molecular theory and is of no value if any component of the mixture
takes part in a reaction or is altered in relative quantity. The average
molecular weight is calculated by adopting a unit molal quantity of the
mixture as the basis of calculation. The weight of this molal quantity
is then calculated and will represent the average molecular v;eight. By
this method the average molecular weight of air is found to be 29.0.
Illustration 6. Calculate the average molecular weight of a flue gas having the
following composition by volume:
CO, 13.1%
O,.. 7.7%0
N, 79.2%0
100.0%
CHAP. II] DENSITIES OF GASEOUS MIXTURES 41
Basis: 1 gram-mole of the mixture.

CO2 = 0.131 gram-mole or 5.76 grams
O2 = 0.077 gram-mole or 2.46 grams
N2 = 0.792 gram-mole or 22.18 grams
Weight of 1 gram-mole = 30.40 grams, which is the aver-
age molecular weight.
Densities of Gaseous Mixtures. If the composition of a gas mixture

is expressed in molal or weight units the density is readily determined
by selecting a unit molal quantity or weight as the basis and calculating
its volume at the specified conditions of temperature and pressure.
This method may be applied to mixtures which do or do not follow the
ideal gas law. Where the ideal gas law is applicable, a more direct
method is first to obtain the volume of the basic quantity of mixture
at standard conditions by multiplying the number of moles by the normal
molal volume. The volume at the specified conditions is then calcu-
lated from the ideal gas law.
Illustration 7. Calculate the density in pounds per cubic foot at 29 in. of Hg and
30°C of a mixture of hydrogen and oxygen which contains 11.1% Hj by weight.
Basis: 1 lb of mixture.
H2 = 0.111 lb or 0.0555 lb-mole
O2 = 0.889 lb or 0.0278 lb-mole
Total molal quantity 0.0833 lb-mole
Volume at S.C. = 0.0833 X 359 ' 29.9 cu ft
29 92 X 303
Volume at 29 in. Hg, 30°C = 29.9 X —^ —7 = 34.2 cu ft
29.0 X273 ^.^^,,
Density at 29 in. Hg, 30°C = — - 0.0292 lb per cu ft
If the composition of a mixture of gases is expressed in volume units,

the ideal gas law is ordinarily applicable. In this case the volume
analysis is the same as the molal analysis, and the density is readily
calculated on the basis of a unit molal quantity of the mixture. The
weight of the basic quantity is first calculated and then its volume at the
specified conditions.
Illustration 8. Air is assumed to contain 79.0% nitrogen and 21.0% oxygen by
volume. Calculate its density in grams per liter at a temperature of 70°F and a
pressure of 741 mm of Hg.
Basis: 1.0 gram-mole of air.
O2 = 0.210 gram-mole or 6.72 grams
N2 = 0.790 gram-mole or 22.10 grams
Total weight 28.82 grams
Volume at S.C .' 22.41 liters
42 BEHAVIOR OF IDEAL GASES [CHAP. IJ.
Volume, 741 mm Hg, TO^F = 22.41 X y^f^^^- 24.8 liters

28 82
Denaty = —'•— = 1.162 grams per liter (741 mm Hg, 70°F)
The actual density of the atmosphere is slightly higher owing to the

presence of about 1 per cent of argon which is classed as nitrogen in the
above problem. The mixture of nitrogen and inert gases in the atmos-
phere may be termed atmospheric nitrogen. The average molecular
weight of this mixture is 28.2. ^ . '
, ,. • ,1 -. , . - I ; ,-,__. \.,
VOLUME CHANGES WITH CHANGE IN COMPOSITION

Such operations as gas absorption, drying, and some types of evapora-
tion involve changes in the compositions of gaseous mixtures, owing
to the addition or removal oi certain components. In a drying opera-
tion a stream of air takes on water vapor. In the scrubbing of coal gas,
ammonia is removed from the mixture. It is of interest to calculate
the relationships existing between the initial and final volumes of the
mixture and the volume of the material removed or added to the mixture
in such a process. The situation is ordinarily complicated by changes
of temperature and pressure concurrent with the composition changes.
Solution may be carried out by the methods of Chapter I if the quantities
specified in the problem are first converted to weight or molal units.
The quantities which are unknown may then be calculated in these
same units. The last step will then be the conversion of the results
from molal or weight units into volumes at the specified conditions of
temperature and pressure. The relationships between molal units and
volumes under any conditions are expressed by Equations (16) to (22).
This method of solution may be applied with the use of either the ideal
gas law or more accurate equations. The following illustration demon-
strates the method for a case in which the ideal.gas law is applicable.
As in the problems of Chapter I, the calculations must be based on a
definite quantity of a component which passes through the process
unchanged.
Illustration 9. Combustion gases having the following molal composition are

passed into an evaporator at a temperature of 200°C and a pressure of 743 mm
of Hg.
Nitrogen 79.2%
Oxygen '. 7.2%
Carbon dioxide 13.6%
•, . ' '•• 100.0%
CHAP. II] VOLUME CHANGES WITH CHANGE IN COMPOSITION 43
Water is evaporated, the gases leaving at a temperature of 85°C and a pressure of

740 mm of Hg with the following molal composition:
Nitrogen 48.3%
Oxygen 4.4%
Carbon dioxide 8.3%
Water 39.0%
-.••'•^/ , - ' ! ' •:,.,-' •'••.. 100.0%
(o) Calculate the volume of gases leaving the evaporator per 100 cu ft entering.
(6) Calculate the weight of water evaporated per 100 cu ft of gas entering.
Solution: • • ' ' '' " ' '
Basis: 1 gram-mole of the entering gas.
N2 0.792 gram-mole
' ' ' O2 0.072 gram-mole
COj 0.136 gram-mole
Total volume (743 mm Hg, 200°C) calculated from Equations (14) and (20):
p = 743/760 or 0.978 atm
- \ ' ' T = 473°K
iJ = 82.1 cc atm per "K
(UA +nB + nc) RT (0.792 + 0.072 + 0.136) 82.1 X 473
~. p ~ 0.978
= H 2 < | ^ L X i Z 3 = 39,750 cc or 1.40 cu ft

0.978 (743 jjjjjj jjg, 200°C)
This 1.0 g-mole of gas entering forms 61% by volume of the gases leaving the
evaporator.
Gases leaving = —^ = 1.64 gram-moles.

0.61 .-,
Water leaving = 1.64 — 1.0 = 0.64 gram-mole.
Volume of gas leaving, from Equations (14) and (20):
p = 740/760 = 0.973 atm
T = 358°K
R = 82.1 cc atm per °K
(0.792 -t- 0.072 -I- 0.136 -|- 0.64) X 82.1 X 358 '
V =
0.973
1.64 X 82.1 X 358
= 49,500 cc or 1.75 cu ft
0.973
Volume of gas leaving per 100 cu ft entering,
1.75 X 100
125 eu ft (740 mm Hg, 85°C)
1.40
Weight of water leaving evaporator = 0.64 X 18 = il.5 grams or 0.0254 lb
Weight of water evaporated per 100 cu ft of gas entering,
• 2 : 2 ? ^ ^ :: 1.811b
1.40
Pure-component Volume Method. Where the ideal gas law m a y
be applied, t h e above m e t h o d of calculation is unnecessarily tedious.
I n this case the solution m a y be carried o u t without conversion to molal
or weight units b y appHcation of pure-component volumes. T h e total
volume of any ideal mixture m a y be obtained by adding together t h e
pure-component volumes of its components. Similarly, t h e removal
of a component from a mixture will decrease the t o t a l volume b y its
pure-component volume. C a r e m u s t be t a k e n in t h e use of this method
t h a t all volumes which are added together are expressed at the same
conditions of temperature and pressure. A process involving changes in
t e m p e r a t u r e a n d pressure as well as composition is best considered as
t a k i n g place in two steps: first, t h e change in composition a t t h e initial
conditions of t e m p e r a t u r e a n d pressure; and second, the change in
volume of t h e resultant mixture to correspond to t h e final conditions of
t e m p e r a t u r e and pressure. Again t h e entire calculation m u s t be based
on a definite q u a n t i t y of a component which passes through t h e process
without change in q u a n t i t y . This procedure is indicated in t h e follow-
ing illustration: r- r, , /.^
Illustration 10. In the manufacture of hydrochloric acid a gas is obtained which

contains 25% HCl and 75% air by volume. This gas is passed through an absorp-
tion system in which 98% of the HCl is removed. The gas enters the system at a
temperature of 120°F and a pressure of 743 mm of Hg and leaves at a temperature
of 80°F and a pressure of 738 mm of Hg.
(o) Calculate the volume of gas leaving per 100 cu ft entering the absorption
apparatus.
(6) Calculate the percentage composition by volume of the gases leaving the ab-
sorption apparatus.
(c) Calculate the weight of HCl removed per 100 cu ft of gas entering the ab-
sorption apparatus. , , ,: ,, .. / ,
Solution:
Basis: 100 cu ft of entering gas (743 mm Hg, 120°F) containing 75 cu ft of air
which will be unchanged in quantity.
Pure-component vol. of HCl 25 cu ft
Pure-component vol. of HCl absorbed 24.5 cu ft
Pure-component vol. of HCl remaining. . ..' 0.50 cu ft
Vol. of gas remaining:
75 + 0.50 75.5 cu ft (743 mm, 120°F)
Vol. of gas leaving:
743 540
75-5 X r r r X —; 70.8 cu ft (738 mm 80°F)
738 580
CHAP. II] PARTIAL PRESSURE METHOD' 45
Composition of gases leaving: -• i ». •
HCl, 0.5/75.5 0.66%
Air 99.34%
743 492
Vol. at S.C. of HCl absorbed = 24.5 X — X — 20.3 cu ft
HCl absorbed = 20.3/359 = 0.0665 lb-mole or 2.07 lb
Partial Pressure Method. In certain types of work, especially where
condensable vapors are involved, it is convenient to express the composi-
tions of gaseous mixtures in terms of the partial pressures of the various
components. Where data are presented in this form, problems of the
type discussed above may be more conveniently solved by considering
only the pressure changes resulting from the changes in composition.
The addition or removal of a component of a mixture may be considered
as producing only a change in the partial pressure of all of the other com-
ponents. The actual volume occupied by each of these components will
always be exactly the same as that of the entire mixture. The volume of
the mixture may then always be determined by application of the gas law
to any components which pass through the process unchanged in quan-
tity and whose partial pressures are known at both the initial and final
conditions. The use of this method is shown in the following illustration:
Illustration 11. Calcium h}rpoohlorite is produced by absorbing chlorine in milk
of lime. A gas produced by the Deacon chlorine process enters the absorption appa-
ratus at a pressure of 740 mm of Hg and a temperature of 76°F. The partial pres-
sure of the chlorine is 59 mm of Hg, the remainder being inert gases. The gas leaves
the absorption apparatus at a temperature of 80°F and a pressure of 743 mm of Hg
with a partial pressure of chlorine of 0.5 mm of Hg.
(a) Calculate the volume of gases leaving the apparatus per 100 cu ft entering.
(6) Calculate the weight of chlorine absorbed, per 100 cu ft of gas entering.
Solution:
Basis: 100 cu ft of gas entering (740 nmi Hg,75°P). ., : '-•.:,
Partial pressure of inert gases entering = 740 — 59 681 mm Hg
Partial pressure of inert gases leaving = 743 — 0.5 742.5 mm Hg
Actual volume of inert gas entering 100 cu ft
681 540
Actual volume of inert gases leaving = 100 X zrr-;; X -r~ = 92.5 cu ft. This
742.5 535
is also the total volume of gases leaving (743 mm Hg, 80°F).
The actual volumes of chlorine entering and leaving are also 100 and 92.5 cu ft,
respectively.
59 X 492
Volume at S.C. of chlorine entering = 100 — —- . . . . 7.14 cu ft
760 X 635
0 5 X 492
Volume at S.C. of chlorine leaving = 92.5 0.055 cu ft
6-
760 X 540
46 BEHAVIOR OF IDEAL GASES ICHAP. II
Volume at S.C. of chlorine absorbed = 7.14 — 0.055 7.08 cu ft

7.08
Chlorine absorbed = — - = 0.0197 lb-mole or 1.40 lb
359
, , ,,, GASES IN CHEMICAL REACTIONS

In a great many chemical and metallurgical reactions gases are pres-
ent, either in the reacting materials or in the products or in both. Quan-
tities of gases are ordinarily expressed in volume units because of the
fact that the common methods of measurement give results directly on
this basis. The general types of reaction calculations must, therefore,
include the complications introduced by the expression of gaseous quan-
tities and compositions in volume units.
In Chapter I methods are demonstrated for the solution of reaction
calculations through the use of molal units for the expression of quantities
of reactants and products. Where this is the scheme of calculation,
the introduction of volumetric data adds but few comphcations. By
the use of the normal molal volume constants combined with the pro-
portions of the ideal gas law it is easy to convert from molal to volume
units, and the reverse. The methods of conversion have been explained
in the preceding sections.
The same general methods of solution are followed" as were described
in Chapter I. All quantities of active materials, whether gaseous,
solid, or liquid, are expressed in molal units and the calculation carried
out on this basis. Results are thus obtained in molal units which may
readily be converted to volumes at any desired conditions. The most
convenient choice of a quantity of material to serve as the basis of
calculation is determined by the manner of presentation of the data.
In general, if the data regarding the basic material are in weight units,
a unit weight is the best basis of calculation. If the data are in volume
units, a unit molal quantity is ordinarily the most desirable basis.
Illustration 12. Nitric acid is produced in the Ostwald process by the oxidation
of ammonia with air. In the first step of the process ammonia and air are mixed
together and passed over a catalyst at a temperature of 700°C. The following re-
action takes place:
4NH3 + 5O2 = 6H2O •+ 4NO
The gases from this process are passed into towers where they are cooled and the
oxidation completed according to the following theoretical reactions:
2NO +O2 = 2NO2
3NO2 4- H2O = 2HNO3 + NO I
CHAP. II] GASES I N C H E M I C A L R E A C T I O N S 47
The NO liberated is in part reoxidized and forms more nitric acid in successive
repetitions of the above reactions. The ammonia and air enter the process at a
temperature of 20°C and a pressure of 755 mm Hg. The air is present in such pro-
portion t h a t the oxygen will be 2 0 % in excess of that required for complete oxidation
of the ammonia to nitric acid and water. The gases leave the catalyzer at a pressure
of 743 mm of H g and a temperature of 700°C.
(a) Calculate the volume of air to be used per 100 cu ft of ammonia entering the
process.
(b) Calculate the percentage composition by volume of the gases entering the
catalyzer.
(c) Calculate the percentage composition by volume of the gases leaving the cata-
lyzer, assuming that the degree of completion of the reaction is 8 5 % and t h a t no other
decompositions take place.
(d) Calculate the volume of gases leaving the catalyzer per 100 cu ft of ammonia
entering the process.
(e) Calculate the weight of nitric acid produced per 100 cu ft of ammonia entering
the process, assuming t h a t 9 0 % of the nitric oxide entering the tower is oxidized to
nitric acid.
Basis of Calculation: 1.0 lb-mole of NH3.
NH3 + 2O2 = HNO3 + H2O
(a) O2 required 2.0 lb-moles

O2 supplied = 2.0 X 1.2 2.4 lb-moles
2.4
Air supplied = 11.42 lb-moles
0.210
Therefore: •
Vol. of air = 11.42 X (vol. of ammonia at same conditions)
or:
293 X 760
Vol. of NHs = 359 X — z r r = 388 cu ft (20°C, 755 mm Hg)
Vol. of air = 11.42 X 388 = 4440 cu ft (20°C, 755 mm Hg)

4440 X 100
Vol. of air per 100 cu ft of NH3 = — = 1142 cu ft
388
(6) Gases entering process = N2, O2, NHs.

N2 present in air = 0.790 X 11-42 9.02 lb-moles
Total quantity of gas entering catalyzer = 11.42 -1- 1 = 12.42 lb-moles
Composition by volume:
NH3 = 1.0/12.42 8.0%
O2 =2.4/12.42 19.3%
N2 = 9.02/12.42 72.7%
• ^ ....-I ; -'y-"- : '-'Â-^-'-^' 100.0%
(c) Gases leaving catalyzer, N2, NH3, O2, NO, and H2O.
NII3 oxidized in catalyzer , 0.85 lb-mole
NHs leaving catalyzer 0.15 lb-mole
Oa consumed in catalyzer = (6/4) X 0.85 1.06 lb-moles
O2 leaving catalyzer = 2.40 — 1.06 1.34 lb-moles
NO formed in catalyzer 0.85 lb-mole
H2O formed in catalyzer = (6/4) X 0.85 1.275 lb-moles
Total quantity of gas leaving catalyzer = 9.02 -|- 0.15
+ 1.34 + 0.85 + 1.275 12.64 lb-moles
NO = 0.85/12.64 6.7%
H2O = 1.275/12.64 10.1% ^
NH3 = 0.15/12.64 1.2%
O2 = 1.34/12.64 10.6%
N2 = 9.02/12.64 71.4%
Basis of CalculatUm: 100 cu ft of NH3 entering the process.
id) Moles of NH3 = ^ ' ° o ^ / ° ^ 0.258 lb-mole

388
Moles of gas leaving catalyzer = 0.258 X 12.64 3.26 lb-moles
Vol. at S.C. of gas leaving catalyzer = 3.26 X 359..'. 1170 cu ft
Vol. of gas leaving catalyzer = 1170 X r — — — r . . . 4270 cu ft
(700°C, 743 mm Hg) per 100 cu ft of NH3 entering.
(e) NO produced in catalyzer = 0.258 X 0.85 0.219 lb-mole

NO oxidized in tower = 0.219 X 0.90 0.197 lb-mole
HNO3 formed = 0.197 lb-mole or 0.197 X 63 12.4 lb
Range of Applicability of the Ideal Gas Law. The ideal gas law is
applicable only at conditions of low pressure and high temperature
corresponding to large molal volumes. At conditions resulting in small
molal volumes the simple kinetic theory breaks down and volumes
calculated from the ideal law tend to be too large. In extreme cases the
calculated volume may be five times too great, an error of 400 per cent.
If an error of 1 per cent is permissible the ideal gas law may be used
for diatomic gases where gram-molal volumes are as low as 5 liters
(80 cubic feet per pound-mole) and for gases of more complex molecular
structure such as carbon dioxide, acetylene, ammonia, and the lighter
hydrocarbon vapors, where gram-molal volumes exceed 20 liters (320
cubic feet per pound-mole).
The actual behavior of gases under high-pressure conditions is dis-
C H A P . II] PROBLEMS 49
cussed in Chapter XII where rigorous methods of calculation are pre-

sented.
PROBLEMS
Pressures are absolute unless otherwise stated
1. I t is desired to market oxygen in small cylinders having volumes of 0.5 cu ft

and each containing 1.0 lb of oxygen. If the cylinders may be subjected to a maxi-
mum temperature of 120°F, calculate the pressure for which they must be designed,
assuming the applicability of the ideal gas law.
2. Calculate the number of cubic feet of hydrogen sulfide, measured at a temper-
ature of 30°C and a pressure of 29.1 in. of Hg, which may be produced from 10 lb
of iron sulfide (FeS).
3. An automobile tire is inflated to a gauge pressure of 35 lb per sq in. at a tem-
perature of 0°F. Calculate the maximum temperature to which the tire may be
heated without the gauge pressure exceeding 50 lb per sq in. (Assume t h a t the
volume of the tire does not change.)
4. Calculate the densities in pounds per cubic foot at standard conditions and
the specific gravities of the following gases: (o) methane, (b) hydrogen, (c) acetylene,
(d) bromine.
5. The gas acetylene is produced according to the following reaction by treating
calcium carbide with water:
CaCj + 2H2O = C2H2 -t- C a ( 0 H ) 2 '
Calculate the number of hours of service which can be derived from 1.0 lb of carbide
in an acetylene lamp burning 2 cu ft of gas per hour at a temperature of 75°F and a
6. A natural gas has the following composition by volume:
CH4 94.1%
N2 3.0%
' H2 1.9%
• s .' O2 1.0%
100.0%
This gas is piped from the well at a temperature of 20''C and a pressure of 30 lb
per sq in. I t may be assumed t h a t the ideal gas law is applicable. :
(a) Calculate the partial pressure of the oxygen.
(b) Calculate the pure-component volume of nitrogen per 100 cu ft of gas.
(c) Calculate the density of the mixture in pounds per cubic foot at the existing
conditions.
7. A gas mixture contains 0.274 lb-mole of HCl, 0.337 lb-mole of nitrogen, and 0.089
lb-mole of oxygen. Calculate the volume occupied by this mixture and its density in
pounds per cubic foot at a pressure of 40 lb per sq in. and a temperature of 30°G.
8. A chimney gas has the following composition by volume:
CO2 10.5% • '

CO 1.1%
'' - ' O2 7.7% •
•> • '' • N2 80.7% ' •
50 B E H A V I O R O F I D E A L GASES [CHAP. U
Using the ideal gas law, calculate: <

(o) Its composition by weight.
(6) The volume occupied by 1.0 lb of the gas at 67°F and 29.1 in. of H g pressure.
(c) The density of the gas in pounds per cubic foot at the conditions of part (6).
(d) The specific gravity of the mixture.
9. By electrolyzing a mixed brine a mixture of gases is obtained a t t h e cathode
having the following composition by weight:
CU : 67%
Brj 28%
O2 5%
Using t h e ideal gas law, calculate:
(o) The composition of the gas b y volume.
(6) The density of the mixture in grams per liter a t 25°C and 740 m m of H g
pressure,
(c) The specific gravity of the mixture.
10. A mixture of ammonia and air at a pressure of 745 mm of H g and a temper-
ature of 40°C contains 4.9% NH3 b y volume. The gas is passed at a rate of 100 cu
ft per min through an absorption tower in which only ammonia is removed. The
gases leave the tower at a pressure of 740 mm of Hg, a temperature of 20°C, and
contain 0.13% NH3 by Volume. Using the ideal gas law, calculate:
(a) The rate of flow of gas leaving the tower in cubic feet per minute.
j(6) The weight of ammonia absorbed in the tower per minute.
11. A volume of moist air of 1000 cu ft at a total pressure of 740 m m of H g and a
temperature of 30°C contains water vapor in such proportions t h a t ' i t s partial pres-
sure is 22.0 mm of Hg. Without changing the total pressure, the temperature is
reduced to 15°C and some of the water vapor removed by condensation. After
cooling it is found t h a t t h e partial pressure of t h e water vapor is 12.7 m m of Hg.
Using the partial pressure method, calculate:
(o) The volume of the gas after cooling. - .
(6) The weight of water removed.
12. Air is passed into a dryer for the drying of textiles a t a rate of 1000 cu ft per
min. The air enters the dryer a t a temperature of 160°F and contains water vapor
exerting a partial pressure of 8.1 m m of Hg. The temperature of the air leaving
is 80°F, and the partial pressure of the water is 18 mm of Hg. The total pressure
of the wet air may be taken as constant at t h e barometric value of 745 mm, of Hg.
(a) Calculate the volume of gas leaving the,dryer per minute.
(6) Calculate the weight of water removed per minute from the material in the
dryer.
13. A producer gas has the following composition by volume:
CO 23.0%
CO2 4.4%
O2 2.6%
N2 70.0%
(a) Calculate the cubic feet of gas, at 70°F and 750 mm of H g pressure, per
pound of carbon present.
(b) Calculate the volume of air, at the conditions of part (a), required for the
combustion of 100 cu.ft of the gas at the same conditions if it is desired
. C H A P . II] PROBLEMS 51
t h a t the total oxygen present before combustion shall be 2 0 % in excess of

t h a t theoretically required.
(c) Calculate the percentage composition by volume of the gases leaving the
burner of part (6), assuming complete combustion.
(d) Calculate the volume of the gases leaving the combustion of parts (6) and (c)
at a temperature of 600°F and a pressure of 750 mm of H g per 100 cu ft
of gas burned.
14. The gas from a sulfur burner has the following composition by volume:
SO, . 1.1%
SO2 8.2%
O2 10.0%
Ni, 80.7%
(a) Calculate the volume of the gas at 350°F and 29.2 in. of Hg formed per
pound of sulfur burned.
(6) Calculate the percentage excess oxygen supplied for the combustion above
that required for complete oxidation to SO3.
(c) From the above gas analysis calculate the percentage composition b y volume
of the air used in the combustion.
(d) Calculate the volume of air at 70°F and 29.2 in. of H g supplied for the com-
bustion per pound of sulfur burned.
16. A furnace is to be designed to burn coke at the rate of 200 lb per hour. The
coke has the following composition:
Carbon 89.1%
Ash 10.9%
The grate efficiency of the furnace is such t h a t 9 0 % of the carbon present in the coke
charged is burned. Air is supplied in 3 0 % excess of that required for the complete
combustion of all the carbon charged. It may be assumed that 9 7 % of the carbon
burned is oxidized to the dioxide, the remainder forming monoxide.
(o) Calculate the composition, by volume, of the flue gases leaving the furnace.
1 (b) If the flue gases leave the furnace at a temperature of 550°F and a pressure
of 743 mm Hg, calculate the rate of flow of gases, in cubic feet per minute,
for which the stack must be designed.
16. Coke containing 87.2% carbon and 12.8% ash is burned on a grate. It is
found that 6 % of the carbon in the coke charged is lost with the refuse. The com-
position by volume of the stack gases from the furnace is as follows:
CO2 12.0%
CO 0.2%
O2 8.8%
N2 79.0%
(o) Calculate the volume of gases, at 540°F and 29.3 in. of H g pressure, formed
per pound of coke charged.
(6) Calculate the per cent of excess air supplied above t h a t required for com-
plete oxidation of the carbon charged.
(c) Calculate the degree of completion of the oxidation, to the dioxide, of the
carbon burned.
(d) Calculate the volume of air, at 70°F and 29.3 in. Hg, supplied per pound of
coke charged.
17. In the fixation of nitrogen by the arc process, air is passed through a magneti-
cally flattened electric arc. Some of the nitrogen is oxidized to NO, which on cooling
oxidizes to NO2. Of the NO2 formed, 66% will be associated to N2O4 at 26°C.
The gases are then passed into water-washed absorption towers where nitric acid is
formed by the following reaction:
H2O -I- 3NO2 = NO -1- 2HNO3
The NO liberated in this reaction wjU be reoxidized in part and form more nitric acid.
In the operation of such a plant it is found possible to produce gases from the arc
furnace in which the nitric oxide is 2% by volume, while hot. The gases are cooled
to 26°C at a pressure of 750 mm of Hg before entering the absorption apparatus,
(o) Calculate the complete analysis by volume of the hot gases leaving the fur-
nace assuming that the air enteringthe furnace was of average atmospheric
composition.
(b) Calculate the partial pressures of the NO2 and N2O4 in the gas entering the
absorption apparatus.
(c) Calculate the weight of HNO3 formed per 1000 cu ft of gas entering the
absorption system if the conversion to nitric acid of the combined nitrogen
in the furnace gases is 85% complete.
CHAPTER III
VAPOR PRESSURES
Liquefaction and the Liquid State. Molecules in the gaseous state of
aggregation exhibit two opposing tendencies. The translational kinetic
energy possessed by each molecule represents a continual, random mo-
tion which tends to separate the molecules from one another and to
cause them to be uniformly distributed throughout the entire available
space. On the other hand, the attractive forces between the molecules
tend to draw them together into a concentrated mass, not necessarily
occupying the entire space which is available.
The first tendency, that of dispersion, is dependent entirely on the
temperature. An increase in the temperature will increase the trans-
lational kinetic energy of each molecule and will therefore give it an in-
creased ability to overcome the forces tending to draw it toward other
molecules. The second tendency, that of aggregation, is determined by
the magnitudes and nature of the attractive forces between the mole-
cules and by their proximity to one another. These intermolecular
attractive forces are believed to be of such a
nature that they increase to definite maxima
as the distances between molecules are
diminished. This behavior is shown in
Fig. 3, in which are plotted attractive
forces as ordinates and distances of sepa-
ration between two molecules as abscissas.
The greatest attractive force between the
two molecules exists when they are sepa- FIG. 3. Attractive force be-
rated by a relatively small distance 82- If tween molecules.
the distance of separation is diminished below Sj, the attractive force
rapidly decreases and will reach high negative values corresponding to
repulsion. At a distance of separation Si the attractive force becomes
zero, corresponding to a position of equilibrium. If unaffected by other
forces, molecules will group themselves together, separated from one
another by distances equal to Su Any attempt made to crowd them
closer together will meet with repulsive forces. In order to separate
them by a distance greater than S2 it would be necessary to overcome
the maximum attractive force by heating or expansion.
When a gas is isothermally compressed and the distances of separation
between the molecules are decreased, the attractive forces increase
53
64 VAPOR PRESSURES [CHAP III
toward their maximum value. If these attractive forces become so

large that the potential energy of the attraction of one molecule for
another is greater than its kinetic energy of translation, the molecules
will be held together to form a dense aggregation which is termed a
liquid. The characteristic which differentiates a liquid from a gas is the
fact that the liquid possesses a definite volume and does not necessarily
occupy the entire available space. The individual molecules of the
liquid are in motion, owing to their inherent kinetic energies, but this
motion takes the form of vibrations, alternately increasing and decreas-
ing the distances of separation, as though energized, vibrating marbles
were attached to each other by short rods of rubber.
Critical Properties. Whether or not a substance can exist in the
hquid state is dependent on its temperature. If the temperature is
sufficiently high that the kinetic energies of translation of the molecules
exceed the maximum potential energy of attraction between them, the
liquid state of aggregation is impossible. The temperature at which
the molecular kinetic energy of translation equals the maximum po-
tential energy of attraction is termed the critical tem,perature, tc. Above
the critical temperature the liquid state is impossible for a single com-
ponent and compression results only in a highly compressed gas, retain-
ing all the properties of the gaseous state. Below the critical temper-
ature a gas may be liquefied if sufficiently compressed.
The pressure required to liquefy a gas at its critical temperature is
termed the critical 'pressure, Pc- The critical pressure and temperature
fix the critical state at which there is no distinction between the gaseous
and liquid states. The volume at the critical state is termed the critical
volume, Vc. The density at the critical state is the critical density, dc.
In Table XI, page 234, are values of the critical data for the more
common gases.
Reduced Conditions. At conditions equally removed from the crit-
ical state many properties of different substances are similarly related.
This has given rise to the concept of reduced temperature, reduced
pressure, and reduced volume. Reduced^ temperature is defined as the
ratio of the existing temperature of a substance to its critical temper-
ature, both being expressed on ah absolute scale. Similarly, reduced
pressure is the ratio of the existing pressure of a substance to its crittfal
pressure, and the reduced volume the ratio of the existing molal volume
to its critical molal volume. Thus,'
Reduced temperature = Tr = T/Tc
. ._ Reduced pressure = Pr = p/Pe
Reduced volume = y, = v/vc
CHAP. Ill] VAPORIZATION 55
Under conditions of equal reduced pressure and equal reduced temper-

ature, substances are said to be in corresponding states. It will be
later shown that many properties of gases and liquids, for example,
the compressibilities of different gases, are nearly the same at corre-
sponding states, that is, at equal reduced conditions.
Vaporization. As pointed out above, the liquid state results when
conditions are such that the potential energies of attraction between
molecules exceed their kinetic energies of translation. These conditions
are brought about when the temperature of a substance is lowered,
decreasing the kinetic energies of translation, or when the molecules are
crowded close together, increasing the energies of attraction. On the
basis of this theory the surface of a liquid may be pictured as a layer of
molecules, each of which is bound to the molecules below it by the
attractive forces among them. One of the surface molecules may be
removed only by overcoming the attractive forces holding it to the
others. This is possible if the molecule is given sufficient translational
kinetic energy to overcome the maximum potential energy of attraction
and to enable it to move past the point of maximum attraction. Once
it has passed this distance of maximum attraction, the molecule is free
to move away from the surface under the effect of its translational
energy and to become a gas molecule.
In the simple kinetic-theory mechanisms which have been discussed,
it is frequently assumed that all molecules of a substance at a given
temperature are endowed with the same kinetic energies and move at
the same speeds. Actually it has been demonstrated that this is not
the case and that molecular speeds and energies vary over wide ranges
above and below the average values. In every liquid and gas there are
always highly energized molecules moving at speeds much higher than
the average. When such a molecule comes to the surface of a liquid,
with its velocity directed away from the main body, it may have suf-
ficient energy to break away completely from the forces tending to hold
it to the surface. This phenomenon of the breaking away of highly
energized molecules takes place from every exposed liquid surface. As
a result, molecules of the liquid continually tend to assume the gas-
eous or vapor state. This phenomenon is termed vaporization or
euaporation.
When a liquid evaporates into a space of limited dimensions the space
will become filled with the vapor which is formed. As vaporization
proceeds, the number of molecules in the vapor state will increase and
cause an increase in the pressure exerted by the vapor. It will be re-
called that the pressure exerted by a gas or vapor is due to the impacts
of its component molecules against the confining surfaces. Since the
56 VAPOR PRESSURES [CHAP. I l l
original liquid surface forms one of the walls confijiing the vapor, there
will be a continual series of impacts against it by the molecules in the
vapor state. The number of such impacts will be dependent on or will
determine the pressure exerted by the vapor. However, when one of
these gaseous molecules strikes the liquid surface it comes under the
influence of the attractive forces of the densely aggregated liquid mole-
cules and will be held there, forming a part of the liquid once more.
This phenomenon, the reverse of vaporization, is known as condensation.
The rate of condensation is determined by the number of molecules
striking the liquid surface per unit time, which in turn is determined by
the pressure or density of the vapor. It follows that when a liquid
evaporates into a limited space, two opposing processes are in operation.
The process of vaporization tends to change the liquid to the gaseous
state. The process of condensation tends to change the gas which is
formed by vaporization back into the liquid state. The rate of conden-
sation is increased as vaporization proceeds and the pressure of the vapor
increases. If sufficient liquid is present, the pressure of the vapor must
ultimately reach such a value that the rate of condensation will equal
the rate of vaporization. When this condition is reached, a dynamic
equilibrium is estabhshed and the pressure of the vapor will remain un-
changed, since the formation of new vapor is compensated by condensa-
tion. If the pressure of the vapor is changed in either direction from this
equilibrium value it will adjust itself and return to the equilibrium con-
ditions owing to the increase or decrease in the rate of condensation
which results from the pressure change. The pressure exerted by the
vapor at such equilibrium conditions is termed the vapor pressure of the
liquid. All materials exhibit definite vapor pressures of greater or less
degree at all temperatures.
The magnitude of the equilibrium vapor pressure is in no way depend-
ent on the amounts of liquid and vapor as long as any free liquid sur-
face is present. This results from both the rate of loss and the rate of
gain of molecules by the liquid being directly proportional to the area
exposed to the vapor. At the equilibrium conditions when both rates
are the same, a change in the area of the surface exposed will not affect
the conditions in the vapor phase.
The nature of the liquid is the most important factor determining the
magnitude of the equilibrium vapor pressure. Since all molecules are
endowed with the same kinetic energies of translation at any specified
temperature, the vapor pressure must be entirely dependent on the mag-
nitudes of the maximum potential energies of attraction which must be
overcome in vaporization. These potential energies are determined by
the intermolecular attractive forces. Thus, if a substance has high
CHAP. Ill] ' BOILING POINT 57
intermolecular attractive forces the rate of loss of molecules from its

surface should be small and the corresponding equihbrium vapor pressure
low. The magnitudes of the attractive forces are dependent on both
the size and nature of the molecules, usually increasing with increased
size and complexity. In general, among Uquids of similar chemical
natures, the vapor pressure at any specified temperature decreases with
increasing molecular weight.
Superheat and Quality. A vapor which exists above its critical
temperature is termed a gas. The distinction between a vapor and a
gas is thus quite arbitrary, and the two terms are loosely interchanged.
For example, carbon dioxide at room temperature is below its critical
temperature and, strictly speaking, is a vapor. However, such a mate-
rial is commonly referred to as a gas.
A vapor which exists under such conditions that its partial pressure is
equal to its equilibrium vapor pressure is termed a saturated vapor,
whether it exists alone or in the presence of other gases. The tempera-
ture at which a vappr is saturated is termed the dew point or saturation
temperature. A vapor whose partial pressure is less than its equilibrium
vapor pressure is termed a superheated vapor. The difference between
its existing temperature and its saturation temperature is called its
degrees of superheat.
If a saturated vapor is cooled or compressed, condensation will result,
and what is termed a wet vapor is formed. If the vapor is in turbulent
motion considerable portions of the condensed liquid will remain in
mechanical suspension as small drops in the vapor and be carried with it.
The quality of a wet vapor is the percentage which the weight of vapor
forms of the total weight of vapor and entrained liquid associated with it.
Thus, wet steam of 95 per cent quality is a mixture of saturated water
vapor and entrained drops of liquid water in which the weight of the
vapor constitutes 95 per cent of the total weight.
Boiling Point. When a liquid surface is exposed to a space in which
the total gas pressure is less than the equihbrium vapor pressure of the
liquid, a very rapid vaporization known as boiling takes place. Boiling
results from the formation of tiny free spaces within the liquid itself.
If the equilibrium vapor pressure is greater than the total pressure on the
surface of the liquid, vaporization will take place in these free spaces
which tend to form below the liquid surface. This vaporization will
cause the formation of bubbles of vapor which crowd back the surround-
ing liquid and increase in size because of the greater pressure of the vapor.
Such a bubble of vapor will rise to the surface of the Uquid and join the
main body of gas above it. Thus, when a liquid boils, vaporization takes
place not only at the surface level but also at many interior surfaces of
contact between the liquid and bubbles of vapor. The rising bubbles
also break up the normal surface into more or less of a froth. The vapor
once liberated from the liquid is at a higher pressure than the gas in
which it finds itself and will immediately expand and flow away from the
surface. These factors all contribute to make vaporization of a liquid
relatively very rapid when boiling takes place. When the total pressure
is such that boiling does not take place, vaporization will nevertheless
continue, but at a slower rate, as long as the vapor pressure of the liquid
exceeds the partial pressure of its vapor above the surface.
The temperature at which the equilibrium vapor pressure of a liquid
equals the total pressure on the surface is known as the boiling point.
The boiling point is dependent on the total pressure, increasing with an
increase in pressure. Theoretically, any Uquid may be made to boil at
any desired temperature by sufficiently altering the total pressure
on its surface. The temperature at which a liquid boils when under
a total pressure of 1.0 atmosphere is termed the normal boiling point.
This is the temperature at which the equilibrium vapor pressure equals
760 millimeters of mercury or 1.0 atmosphere.
Vapor Pressures of Solids. SoUd substances possess a tendency to
disperse directly into the vapor state and to exert a vapor pressure just
as do liquids. The transition of a solid directly into the gaseous state is
termed sublimation, a process entirely analogous to the vapotization of a
liquid. . A f amifiar example of sublimation is the disappearance of snow
in sub-zero weather.
A solid exerts an equilibrium vapor pressure just as a liquid does; this
is a function of the nature of the material and its temperature. Sub-
limation will take place whenever the partial pressure of the vapor in
contact with a solid surface is less than the equifibrium vapor pressure
of the solid. Conversely, if the equifibrium vapor pressure of the solid
is exceeded by the partial pressure of its vapor, condensation directly
from the gaseous to the solid state will result.
At the melting point the vapor pressures of a substance in the solid
and liquid states are equal. At temperatm'es above the melting gpint
the solid state cannot exist. However, by careful cooling a liquid can
be caused to exist in an unstable, supercooled state at temperatures
below its melting point. The vapor pressures of supercooled liquids
are always greater than those of the solid state at the same temperature,
and the liquid tends to change to the solid.
The vapor pressures of solids, evea at their melting points, are gener-
ally small. However, in some cases these values become large and of
considerable importance. For example, at its melting point of 114.5°C
iodine crystals exert a vapor pressure of 90 millimeters of mercury.
CHAP. Ill] EFFECT OF TEMPERATURE ON VAPOR PRESSURE 59
Solid carbon dioxide at its melting point of — 56.7°C exerts a vapor

pressure of 5.11 atmospheres and a pressure of 1.0 atmosphere at a
temperature of — 78.5°C. It is therefore impossible for liquid carbon
dioxide to exist in a stable form at pressures less than 5.11 atmospheres.
Calculations dealing with the vapor pressures and sublimation of
solids are analogous to those of the vaporization of Uquids. The
principles and methods outlined in the following sections are equally
applicable to sublimation and to vaporization processes.
EFFECT OF TEMPERATURE ON VAPOR PRESSURE

The forces causing the vaporization of a liquid are entirely derived
from the kinetic energy of translation of its molecules. It follows that
an increase in kinetic energy of molecular translation should increase
the rate of vaporization and therefore the vapor pressure. In Chap-
ter II it was pointed out that the kinetic energy of translation is directly
proportional to the absolute temperature. On the basis of this theory,
an increase in temperature should cause an increased rate of vaporiza-
tion and a higher equilibrium vapor pressure. This is found to be uni-
versally the case where vapor pressures have been experimentally in-
vestigated. It must be remembered that it is the temperature of the
liquid surface which is effective in determining the rate of vaporization
and the vapor pressure.
An exact thermodynamic relationship between vapor pressure and
temperature is developed in Chapter XI as
(1)
T(V, - Vt)
where
p = vapor pressure ,
T = absolute temperature '
A = heat of vaporization at temperature T
Vg = volume of gas
Vi = volume of liquid
The above relationship is also referred to as the Clapeyron equation.
It is entirely rigorous, universal, and applies to any vaporization equilib-
rium. Its use in this form is, however, greatly restricted because it
presupposes a knowledge of the variation of A, Vg, and Vi with tem-
perature.
The latent heat of vaporization. A, is the quantity of heat which
must be added in order to transform a substance from the liquid to the
60 VAPOR PRESSURES [CHAP. Ill
vapor state at the same temperature. The heat of vaporization de-

creases as pressure increases, becoming zero a;t the critical point. This
property is fully discussed in subsequent chapters. Values of the
heats of vaporization at the normal boihng point of many compounds
are listed in Tables XI and X I I I , pages 234-237.
By neglecting the volume of liquid and assuming the applicability
of the ideal gas law the above relation reduces to the Clausius-Clapeyron
equation:
dp UT
or ^---l© «)
where
R = gas law constant •
The Clausius-Clapeyron equation in the form written above is accurate

only when the vapor pressure is relatively low, where it may be assumed
that the vapor obeys the ideal gas law and that the volume in the liquid
state is negligible as compared with that of the vapor state.
Where the temperature does not vary over wide Hmits it may be
assumed that the molal latent heat of vaporization, X is constant and
Equation (2) may be integrated, between the limits po, ô, and p, T, to
give, ,^
or > log f = ^ 7 ^ ^ ^ - - - j (4)

' po 2.303J? '
Equation (4) permits calculation of the vapor pressure of a substance

at a temperature T if the vapor pressure po at another temperature To
is known, together with the latent heat of vaporization X. The results
are accurate only over limited ranges of temperature in which it may be
assumed that the latent heat of vaporization is constant and at such
conditions that the ideal gas law is obeyed.
Illustration 1. The vapor pressure of ethyl ethw is given in the International

Critical Tables as 185 mm of Hg at 0°C. The latent heat of vaporization is 92.5
calories per gram at 0°C. Calculate the vapor pressure at 20°C and at 35°C.
Molecular weight 74.
X 6850 calories per gram-mole.
R 1.99 calories per gram-mole per °K.
To ,._ 273°K.
Po 185 mm Hg.
CHAP. Ill] VAPOR-PRESSURE PLOTS 61
when T = 293°K (20°C) • ,
' - 2.30
' ° ^ 185 ^ X^ [ k - i ] ='''' ^°-°°^««^ - °-^^^^3^ = °-^^*
: ^ -2.36 p = 437at20°C
185
when T = 308°K (35°C)
log ^ = 1495 (0.003663 - 0.003247) = 0.621

185
^ = 4.18 . p = 773 mm Hg at 35°C •

185
The values for the vapor pressure of ether which have been experimentally observed
are 442 mm of Hg at 20°C and 775.5 mm of Hg at 35°C.
In the preceding illustration the Clausius-Clapeyron equation yields
results which are satisfactory for many purposes. However, Equation
(3) is only an approximation which may lead to considerable error in
some cases. It should be used only in the absence of experimental
data.
Tables of physical data contain experimentally determined values
of the vapor pressures of many substances at various temperatures.
Because of the frequent requirement of accurate values of the vapor
pressure of water, extensive data are presented in Table I expressed
in English units. -
VAPOR-PRESSURE PLOTS
From experimental data various types of plots have been devised
for relating vapor pressures to temperature. Use of an ordinary uniform
scale of coordinates does not result in a satisfactory plot because of the
wide ranges to be covered and the curvature encountered. A single
chart cannot be used over a wide temperature range without sacrifice
of accuracy at the lower temperatures and the rapidly changing slope
makes both interpolation and extrapolation uncertain.
A better method which has been extensively used is to plot the loga-
rithm of the vapor pressure (log p) against the reciprocal of the absolute
temperature {l/T). The resulting curves, while not straight, show
much less curvature than a rectangular plot and may be read accurately
over wide ranges of temperature. Another method is to plot the loga-
rithm of the pressure against temperature on a uniform scale. These
scales do not reduce the curvature of the vapor-pressure lines as much
as the use of the reciprocal temperature scale but are more easy to
construct and read.
62 VAPOR PRESSURES [CitAP. I l l
TABLE I •*^
VAPOK PRESSURE OP WATER
English units
Pressure of aqueous vapor over ice in 10~' inches of Hg from —144° to 32°F
Temp °F 0.0 2.tf 4.0 6.0 8.0
-140 0.00095 0.00075 0.00063
-130 0.00276 0.00224 0.00181 0.00146 0.00118
-120 0.00728 0.00595 0.00492 0.00409 0.00339
-110 0.0190 0.01,57 0,0132 0.0111 0,00906
-100 0.0463 0.0387 0.0325 0.0274 0.0228
-90 0.106 0.0902 0.0764 0.0646 0.0543
-80 0.236 0.202 0.171 0,146 0.125
-70 0.496 0.429 0.370 0,318 0,275
-60 1.02 0.882 0.764 0,663 0,575
-50 2.00 1.76 1.53 1,33 1.16
-40 3.80 3.37 2.91 2.59 2.29
-30 7,047 6,268 5.539 4.882 4.315
-20 12.64 11.26 10.06 8.902 7,906
-10 22.13 19.80 17.72 15.83 14,17
-0 37.72 33.94 30.55 •27.48 24,65 .
0 37.72 41.85 46.42 51.46 56.93
10 62.95 69.65 . 76,77 84.65 93.35
20 102.8 113.1 124,4 136.6 150.0
30 164.6 180.3 ,
Pressure of aqueous vapor over water in inches of Hg from +4° t5 212''F
Temp °F 0.0 2.0 4,0 6,0 8.0

0 0,05402 0,05929 0.06480
10 0.07091 0.07760 0,08461 0,09228 0,1007
20 0.1097 0.1193 0,1299 0,1411 0,1532
30 0.1664 0.1803 0,1955 0,2118 0,2292
40 0.2478 0.2677 0,2891 0,3120 0,3364
60 0.3626 0.3906 0,4203 0.4620 0.4858
60 0.5218 0.5601 0,600'9 0.6442 0.6903
70 0.7392 0.7912 0,8462 0.9046 0.9666
80 1.0321 1.1016 1,1750 1.2527 1,3347
90 1.4216 1.5131 1,6097 1.7117 1,8192
100 1.9325 2.0519 2,1776 2.3099 2,4491
110 2.5955 2.7494 2,9111 3.0806 3.2589
120 3.4458 3.6420 3,8475 4.0629 4,2887
130 4.5251 4.7725 5,0314 5.3022 5.5852
140 5.8812 6.1903 6,5132 6.850 7.202 •
150 7.569 7.952 8,351 8.767 9.200
160 9.652 10.122 10,611 11.120 11.649
170 12.199 12.772 13,366 ^ 13.983 14.625
180 15.291 15.982 16,699 17.443 18". 214
190 19.014 19.843 20,703 21.593 22.515
200 23,467 24,455 25,475 26.531 27.625
210 28.755 29,922
CHAP. Ill] VAPOR PRESSURE OF WATER 63
' TABLE I ~ (Continued)

Pressure of aqueous vapor over water in Ib/sq in. for temperatures 210-705.4°F
Temp "F 0.0 2.0 4.0 6.0 8.0
210 14.123 14.696 15.289 15.901 16.533
220 17.186 17.861 18.657 19.275 20.016
230 20.780 21.667 22.379 23.217 24.080
240 24.969 25.884 26.827 27.798 28.797
250 29.825 30.884 31.973 33.093 34.245
260 35.429 36.646 37.897 39.182 40.502
270 41.858 43.252 44.682 46.150 47.657
280 49.203 60.790 52.418 64.088 55.800
290 57.556 69.356 61.201 63.091 66.028
300 67.013 69.046 71.127 73.259 76.442
310 77.68 79.96 82.30 84.70 87.15
320 89.66 92.22 94.84 97.52 100.26
330 103.06 106.92 108.85 111.84 114.89
340 118.01 121.20 124.46 127.77 131.17
350 134.63 138.16 141.77 146.46 149.21
360 153.04 156.95 160.93 165.00 169.15
370 173.37 177.68 182.07 186.55 191.12
380 195.77 200.60 206.33 210.25 215.26
390 220.37 225.56 230.85 236.24 241.73
400 247.31 252.9 258.8 264.7 270.6
410 276.75 282.9 289.2 296.7 302.2
420 308.83 315.5 322.3 329.4 336.6
430 343.72 351.1 368.5 366.1 374.0
440 . 381.69 389.7 397.7 405.8 414.2
450 422.6 431.2 439.8 448.7 457.7
460 466.9 476.2 485.6 496.2 504.8
470 514.7 524.6 534.7 644.9 555.4
480 666.1 576.9 687.8 689.9 610.1
490 621.4 632.9 644.6 • 666.6 668.7
500 680.8 693.2 706.8 718.6 731.4
510 744.3 757.6 770.9 784.6 798.1
620 812.4 826.6 840.8 855.2 870.0
530 885.0 900.1 916.3 930.9 946.6
540 962.6 978.7 995.0 1011.5 1028.2
550 1045.2 1062.3 1079.6 1097.2 1115.1
560 1133.1 1151.3 1169.7 1188.5 1207.4
570 1226.5 1245.8 1265.3 1285.1 1305.3
580 1325.8 1346.4 1367.2 1388.1 1409.6
590 1431.2 1463.0 1475.0 1497.4 1520.0
600 1642.9 1566.2 1589.4 1613.2 1637.1
610 1661.2 1686.0 1710.7 1735.6 1761.0
620 1786.6 1812.3 1838.6 1865.2 1892.1
630 1919.3 1947.0 1974.5 2002.7 2031.1
640 2059.7 2088.8 2118.0 2147.7 2178.0
650 2208.2 2239.2 2270.1 2301.4 2333.3
660 2366.4 2398.1 2431.0 2464.2 2498.1
670 2531.8 2566.0 2601.0 2636.4 2672.1
680 2708.1 2745.0 2782.0 2819.1 2867.0
690 2896.1 2934.0 2973.6 3013.2 3063.2
700 3093.7 3134.9 3176.7 3206.2*
• At 705.4**F, the critical temperature.

64 VAPOR PRESSURES > [CHAP. I l l
As a means of deriving consistent vapor-pressure data for homologous

series of closely related compounds Coates and Brown developed a
special method of plotting which has proved particularly valuable for
the hydrocarbons.1 For this plot rectangular coordinate paper is used
with temperatures as abscissas and normal boiling points as ordinates.
Curved lines of constant vapor pressure are then plotted from the ex-
perimental data available for the various members of the series. This
method is particularly well adapted to extrapolating data obtained for
the lower boiling homologs of a series in order to estimate vapor
pressures for the higher boiling homologs.
Reference Substance Plots. The methods of plotting described above
all result in lines having some degree of curvature, which makes neces-
sary a considerable number of experimental data for the complete
definition of the vapor pressure curve. Where only limited data are
available there is great advantage to a method of plotting which wifl
yield straight lines over a wide range of conditions. With such a method
a complete curve can be established from only two experimental points
and erratic data can be detected.
Where an accurate evaluation of a physical property has been de-
veloped over a wide range of conditions for one substance the resulting
relationship frequently may be made the basis of empirical plots for
other substances of not greatly different properties. This general
method may be applied to vapor-pressure data by selecting a reference
substance the temperature-vapor pressure relationship of which has
been evaluated over a wide range. A function of the temperature at
which some other substance exhibits a given vapor pressure may then
be plotted against the same function of the temperature at which the
reference substance has the same vapor pressure. Or, conversely, a
function of the vapor pressure of the substance at a given temperature
may be plotted against the same function of vapor pressure of the
reference substance at the same temperature.
By proper selection of the reference substance and the functions of
the properties plotted, curves which approximate straight lines over
wide ranges of conditions are obtained. The best results are obtained
with reference substances as similar as possible in chemical structure
and physical properties to the compouMs of interest.
Equal Pressure Reference Substance Plots. The first reference sub-
stance plot of vapor-pressure data was proposed by Diihring, who
plotted the temperature at which the substance of interest has a given
vapor pressure against the temperature at which the reference sub-
1 Coates and Brown, " A Vapor Pressure Chart for Hydrocarbons," Dept. Engr.
Research University of Michigan, Circular Series No. 2 (December, 1928).
CHAP. Ill] REFERENCE SUBSTANCE PLOTS 65
stance has the same vapor pressure. Diihring lines of sodium hydroxide
solutions are plotted in Fig. 6, page 83, using water as the reference
substance. Each of these lines relates the temperature of the designated
solution to the temperature at which water exerts the same vapor pres-
sure. Vapor-pressure data for water appear in Table I.
Equal Temperature Reference Substance Plots. If Equation (2) is
divided by a similar equation for a reference substance at the same tem-
perature the following expression is obtained where the primed quantities
indicate the reference substance:^
dlnp _ X
dlnp' X' ^ ^
Integrating,'
* X
logp = ~\ogp' + C (6)
Thus, if the logarithm of the vapor pressure of a substance at a given

temperature is plotted against the logarithm of the vapor pressure of a
reference substance at the same temperature a line will be obtained the
slope of which is equal to the ratio of the heat of vaporization of the
substance of interest to that of the reference substance, provided that
conditions are such that Equation (2) is appUcable. Since the ratio
X/X' is found to be reasonably constant for most substances not close
to their critical points, the lines should be straight, at least at conditions
well removed from the critical point.
This method of plotting was introduced by Cox^ and later more fully
discussed by Othmer.^ Cox found that a wide variety of substances
plotted as nearly straight lines by this method at conditions up to and
including their critical points. Since the ratio of the latent heats of
vaporization of two substances varies rapidly as either one approaches
its critical point, this behavior can be rationalized with Equation (6)
only by assuming that the variations in the ratio of heats of vaporization
are exactly coinpensated by deviations from the Clausius-Clapeyron
equation from which Equation (6) was derived. As was pointed out on
page 60 this equation is vahd only at condition^ of relatively low pres-
sures and large vapor volumes.
Figure 4 is a Cox chart from which, for simplicity in use, the loga-
rithmic scale of pressures of the reference substance has been omitted
and only the auxiliary temperature scale derived from it shows. Such
' E. R. Cox, Ind. Eng. Chem. 15, 592 (1923). Reprinted with permission.
»D. F. Othmer, Ind. Eng. Chem. 32, 841-856 (1940). Reprinted with permis-
sion.
VAPOR PRESSURES [CHAP. Ill
^«
11
5a. o
E ft
^ ^
g
o
M
ft*
Sfi /o 'Uiui 'ftinssaj^

CHAP. Ill] REFERENCE SUBSTANCE PLOTS 67
a chart may be constructed by plotting vapor pressures as ordinates

against reference substance vapor pressures on multi-cycle double
logarithmic paper. From the vapor-pressure data of the reference
substance an auxiliary abscissa scale of temperatures is established.
To extend the range of the chart to temperatures higher than the critical
temperature of this reference substance a second higher boiling reference
substance is selected and its vapor-pressure data plotted over the tem-
perature range of the first reference substance. The vapor-pressure line
of the second reference substance is then extended and from it the exten-
sion of the auxiliary temperature abscissa scale is estabhshed. Figure 4
was developed in this manner using water as the primary reference sub-
stance and mercury for temperatures above the critical of water.
The Cox method of plotting has been studied by Calingaert and Davis*
who found that the data for widely varying types of materials yield
lines with little curvature when plotted on such a chart. Fur-
thermore, it was found that the curves of groups of closely related
compounds converge at single points which are characteristic of the
groups. For example, single points of convergence were found for each
of the following groups: the paraffin hydrocarbons, the benzene mono-
halides, the alcohols, the silicon hydride series, and the metals. For
a member of a group of materials having convergent curves only one
experimental point and the point of convergence of the group are nec-
essary to establish a complete curve.
Calingaert and Davis also found that the method of Cox, when water
is the reference substance, is equivalent to assuming that the vapor pres-
sure of a substance is represented by the following equation:
where
p = vapor pressure
T = temperature, °K ,
A, B — empirical constants
Thus, by plotting log p against 1/(7 — 43) a straight line should be

obtained.
Illustration 2. The vapor pressure of chloroform is 61.0 mm of Hg at 0°C and
526 mm of Hg at 50°C. Estimate, from Fig. 4, the vapor pressure at 100°C.
Solution: The two experimental values of the vapor pressures at 0°C and 50°C
are represented by points on Fig. 4. A straight line is projected through these
two points to the abscissa representing 100°C. The ordinate at this point is approxi-
* Ind. Eng. Chem. 17, 1287 (1925). Reprinted with permission.
mately 2450 mm of Hg, the estimated vapor pressure at 100°C. The experimentally
observed value is 2430 mm of Hg.
Illustration 3. The vapor pressure of normal butyl alcohol at 40°C is 18.6 mm
of Hg. Estimate the temperature at which the vapor pressure is 760 mm of Hg,
the normal boiling point.
Solution: The experimental value of the vapor pressure at 40°C is represented
by a point on Fig. 4. A straight line is drawn from this point to the point of con-
vergence of the alcohol group. This point of convergence is located by extending
the curves for methyl and propyl alcohols. The abscissa of the point at which this
hue crosses the 760-mm ordinate is about 117°C. The experimentally observed
boiling point of normal butyl.alcohol is 117.7°C.
Estimation of Critical Properties. By means of the Cox type of

plot described above the vapor-pressure curve of a substance may be
estimated with fair accuracy from only a fev/ experimental points.
However, to define completely the vapor-pressure characteristics of the
substance, the terminus of the curve, represented by the critical point,
must also be located. Once the vapor-pressure curve is established
this point can be located from knowledge of only the critical temperature.
The critical temperature of a substance is also of great value in
predicting its behavior-in certain processes and in establishing relation-
ships among other of its physical properties. However, the experi-
mental determination of critical temperatures is difficult, and these
data are not available for many substances. It is frequently desirable
to predict the critical temperature of a substance from other more
easily determined properties. Guldberg proposed a rule that the
ratio of the boiling point, under a pressure of one atmosphere, to the
critical temperature is a constant when both are expressed on an abso-
lute scale of temperature. This rule is only a very rough approximation,
which cannot be used where reliable results are required.
In a method proposed by Watson^ the critical temperature of a non-
polar compound is predicted from its boiling point, molecular weight,
and liquid density. A nonpolar compound is one having its atoms
symmetrically arranged in the molecule so that there are no unbalanced
electrical charges which tend to rotate the molecule when in an electro-
static field. Nonpolar compounds are, in general, chemically inactive
and do not ionize or conduct electricity well. For example, the hydro-
carbons of practically all series are relatively nonpolar, whereas water,
alcohol, ammonia, and the like are highly polar. In general, compounds
having symmetrical molecular arrangements such as methane (CH4) or
carbon tetrachloride (CCI4) may be expected to have nonpolar char-
acteristics. Compounds which do not have symmetrical molecular
arrangements such as methyl chloride (CH3CI) or ethyl alcohol
^ Irtd.Eng.Chem. 23, 360 (1931).
CHAP. Ill] ESTIMATION OF CRITICAL PROPERTIES 69
(C2H5OH) or acetic acid (CH3COOH) may be expected to be polar. Polar

compounds do not follow many of the generalizations which apply to
the nonpolar group.
. ,._ ___ ""7"*
y
[_ (T /
J /
, . f
rr,
to ,a molal
m A
vapor volume of 22.4 liters ?
^
Tff- is-ra A'-
y
.r .^
^^
fr? y
'
y .'
^ 40- J
^
/
^ r ^
J ^
>
^
.^
"5
^
^s. -''•
^^ I
^' *•
_c>n
-i _x
100 200 300 - __-500
400 600 xi-700 _i800 ;900
In T e •• • 4.2
FIG. 6 Temperatures of constant vapor concentration.

The following empirical equation was proposed:
where:
| j = 0.283
© (8)
Tc = critical temperature (°K).

Te = the temperature (°K) at which the substance is in equi-
librium with its saturated vapor in a concentration of
1.0 gram-mole in 22.4 liters.
M = molecular weight.
pB = density of the liquid in grams per cubic centimeter at its
normal boiling point. By means of Fig. 109, Chapter XII,
PB can be estimated from density measurements at other
temperatures.
The temperature Te is a function of the normal boiling point of the sub-
stance. In Fig. 5, a curve is plotted relating (Te — T^) to Ts, where
Ts is the normal boiling point in degrees Kelvin. From this curve Te

can be determined for any substance of known boiling point. The
density at the boiling point may be determined or estimated from liquid
density measurements at other temperatures. It was shown that Equa-
tion (8) permits prediction of critical temperatures of nonpolar sub-
stances ranging from oxygen (boiling point 90.1°K) to octane (boiling
point 398°K) with errors rarely exceeding 2.0 per cent. Good results
were also obtained for slightly polar substances such as carbon disulfide
and chlorobenzene, but the method breaks down when applied to highly
polar substances such as water, ammonia, or the lower alcohols.
Illustration 4. Carbon tetrachloride has a normal boiling point of 77°C and a

liquid density at its boiling point of 1.48 g per cc. Calculate the critical temperature.
From Fig. 5
T, = 77 + 273 = 350°K ,
' *" r , = 350 + 10 = 360°K
From Equation (8),
360 //154V18
1 5 4 \ 0.18
— = 0.283; (( r—r : : )) == 00.654
.6
\lA8j
Tc = 550°K or 277''C ,.
The experimentally observed value is 283°C. ' _ ,
The critical properties of some common substances are given in

Table XI, page 234.
H. P. Meissner and E. M. Redding* have developed more generally
applicable methods for estimating all three critical constants. These
methods apply to polar as well as nonpolar substances with the exception
of water. For associated liquids the results are dependable for critical
volume and temperature but not for critical pressure.
The critical volume is obtained from the equation:
; Vc= {0277P + n.0y-^' i • - (9)'
where . • .
Vc = critical volume, cc per gram-mole
P = parachor
The parachor is a measure of the molecular volume of a liquid at a

standard surface tension:
P = (10)
pi ~ Pa
° Ind. Eng. Chem. 34, 521-525 (1942). Reprinted with permission.
CHAP. Ill] ESTIMATION OF CRITICAL PROPERTIES 71
where M = molecular weight

a = surface tension, dynes per cm
pi = density of liquid, grams per cu cm
Pg = density of gas, grams per cu cm
For organic liquids, the parachor can be estimated quite accurately
from the structural formula by use of atomic and structural values listed
in Table II. The value of the parachor of a compound is the sum of the
contributions of its atomic and structural elements,
TABLE II
ATOMIC AND STRUCTURAL PABACHORS
C 4.8 Triple bond ,.; 46.6
H 17.1 Double bond 23.2
N 12.5 3-membered ring 16.7
P 37.7 4-membered ring 11.6
0 20.0 5-membered ring 8.5
S 48.2 6-membered ring 6.1
F 25.7 Gain esters 60.0
• CI 54.3
Br 68.0
1 91.0
S. Sugden, The Parachor and Valency, Rutledge and Sons, London, 1930. Reprinted with pe>
mission.
For a few simple molecules, such ae CO, CO2, SO3, the results of Table II
are not accurate. For water and diphenyl Equation (9) does not apply.
Otherwise, for a group of one hundred compounds selected at random
the deviation never exceeded 5% from experimental values. Equation
(9) is also appUcable to associated liquids, the parachor being based
upon the structure of the nonassociated liquid.
For the critical temperature, Equation (8) is preferable for nonpolar
compounds. For polar substances or where the liquid density is not
known the following empirical formulas have been developed by Meissner
and Redding:
For compounds boiling below 235°K and for all elements.
Tc = 1.70TB - 2.0 (11)
For compounds boiling above 2S5°K

(a) Containing halogens or sulfur:
Tc = IAITB + 66 - IIF (12)
where F = number of fluorine atoms in the molecule

(6) Aromatic compounds and naphthenes free of halogens and sulfur:
Tc = 1.4irB + 66 - K0.383rB - 93) (13)
where
r = ratio of noncyclic carbon atoms to the total number of carbon
atoms in the compound
(c) Other compounds (boiling above 235°K, containing no aro-

matics, no naphthenes, no halogens, and no sulfur): ^ .M,-.
To = 1.027^5 + 159 '• (14)
The above equations for critical temperatures give agreement within
5% with experimental values of nearly all compounds regardless of
degree of association with the exception of water. The equations have
not been tried on substances normally boiling above 600°K,
The critical -pressure in atmospheres may be predicted from the equa^
tion:
20.8r. 20.87, • "
e-)
With the exception of water, maximum errors of 15 per cent are en-
countered and the majority of the results are within 10% of the experi-
mental values.
Meissner and Redding discuss a method of using the preceding equa-
tions when the normal boiling point is unknown but a vapor-pressure
and a liquid density value are available at some other temperature.
Illustration 5. Estimate the critical properties of triethylamine (C3H6)3N, nor-
mal boiling point 362.5°K.
Molecular weight =101.1
Parachor: Prom Table II
Cs 6 X 4.8 = 28.8
" Hl5 15 X 17.1 = 256.5
1 X 12.5 = 12.5
P 297.8
From Equation (9):
fo = [0.377(297.8) + 11.0]i-25 == 416 ocper gram-•mole
Vc (experimental) = 403
Since triethylamine is neither aromatic nor naphthenic and contains no halogens or

sulfur, Equation (14) may be used,
Tc = 1.027(362.5) + 159 = 531°K
Tc (experimental) = 535.2
CHAP. Ill] ' CRITICAL AND VAPOR PRESSURES 73
From Equation (15) using calculated values for Tc and Oc, ,;'i !?;; vf
20.8(531) „,„ ^
Pc (experimental) = 30.0 atm • i
Critical Pressures and Vapor Pressures of Organic Compounds. It

was found by Gamson and Watson' that the vapor pressure data of
all of over forty substances investigated may be represented by the
following equation: . , . „, ,, , „ ,,. : ; ,<:
log p = ^ + B - e-aocr.-w^ • (16)
where A, B, and b are constants characteristic of the substance. In

applying this equation over vapor-pressure ranges from a few tenths of a
millimeter of mercury to the critical point, deviations were found to be
generally less than 3 % .
Because of the difficulty of evaluating the constants, Equation (16)
is not convenient for the extrapolation of fragmentary data. However,
through generalized expressions for the constants of the equation, it
may be used to predict a complete vapor-pressure curve from only a
single measurement. The ranges of the constants are indicated in
Table Ila. * , ,
--.,.,. i • I TABLE Ila \ \ j "
I ' i : "-"•••: VAPOR-PBESSUBB CONSTANTS j
••S.O; ; - , t "i logio p in mm of H g r \~-- ' . ••
A B h p, T°TS.
Methane 2.3383 6.8800 0.000 34,810 190.7
Ethane 2.5728 7.1411 0.088 37,090 305.5
Ethylene 2.5463 7.1269 0.098 38,080 282.8
Propane 2.6606 7.1819 0.125 33,210 370.0
n-Octane 3.2316 7.5034 0.236 18,700 569.4
Water 3.1423 8.3610 0.163 165,470 647.2
Methyl alcohol 3.5876 8.3642 0.243 59,790 513.2
Diethyl ether 2.9726 7.4039 0.204 27,000 467.0
Acetone 3.0644 7.6173 0.180 35,720 508.2
Ammonia 2.9207 7.8519 0.163 85,350 406.1
Methylamine 2.9589 7.7066 0.239 55,940 430.1
Hydrogen cyanide.. . 3.2044 7.7761 0.000 37,300 456.7
Methyl chloride 2.7195 7.4185 0.052 50,010 416.4
Carbon tetrachloride. 2.J989 7.3329 0.158 34,200 556.3
Acetic acid 3.3908 8.0291 0.138 43,480 594.8
' B. W. Gamson and K. M. Watson, presented before Petroleum Div., Am. Chem.
Soc, Pittsburgh meeting, Sept., 1943. To be published Nat. Petr. News, 1944.
74 VAPOR PRESSURES [CHAP. HI
Applying Equation (16) to the critical point, ,

log pc = B ~ A - e^sod-w'
Since th,e exponential correction term is found to be negligible at re-
duced temperatures above 0.8, B == log p^ + A, and Equation (16)
may be written:
't. • ^
log p, = Z ^ ^ I Z I Z L ) _ e-20(T,-»« (17)
Jr
If the methods previously described are used to calculate the critical

temperatures it is possible to estimate a complete vapor pressure rela-
tionship from a single point by use of Equation (17) in conjunction with
generalized expressions for the critical pressure and the constant b.
800
700
^600
6' ,
•g400 0.3 •o
I
3
o
O
300 0.2 f
200 0.1 >
too2 0.0
3 5 6 7 8 9 10 40 50
Number of Carbon Atoms, n
PIG. 5a. Critical pressures and vapor-pressure constants of the paraffins.
It was found that as the number of carbon atoms in a homologous

series of organic compounds is increased above 2, the critical pressmre
progressively diminishes while the constant b increases. These rela-
tionships are shown graphically for the paraffin hydrocarbons in Fig. 5a.
CHAP. Ill] CRITICAL AND VAPOR PRESSURES 75
For series of compounds other than the paraffin hydrocarbons, Fig. 5a

may be used with the following equations for estimating the critical
pressures and constants b:
Pc = p'c + - ^ (18)
nc + c
b=b' +Ab (19)
where pc and b are the values for a compound containing nc carbon
atoms, p'c and b' are the values read from Fig. 5a corresponding to
nc, and Ap,c, andAb are constants characteristic of the homologous
series, given in Table lib.
= TABLE lib 4
CRITICAL AND VAPOB-PBESSURB CONSTANTS
Ap Ib/sq in. c A6
Acids 300 0 0.05
Alcohols 480 0 0.22
Aldehydes 250 0(?) 0.0(?)
Amines, primary 69 —1.3 0.12
" secondary 80 0 0.12
Aromatic hydrocarbons (monocyclic) 830 —3.0 —0.02
Esters 60 -1.6 0.09
Ethers 0 0 0.04
Halogenated paraffins, mono 47 —1.0 0.08
Ketones 180 0 0.05
Naphthenes 970 0 -0.03
Nitriles 0 0 0.02
Phenols 1360 -3.0 0.0(?)
Olefins, mono 63 0 0.01
Illustration 6. The normal boiling point of n-propyl amine is 48.7°C and its
critical temperature 223.8°C. Estimate the critical pressure and the vapor pressure
at a temperature of 0°C.
From Fig. 5a,

Pc = 640 b' = 0.133
From Table lib,
Ap = 69 c= -1.3 A6=0.12
Substituting in Equations (18) and (19),
69
Pc = 640 + - r = 681 Ib/sqin.
(The experimentally observed value is 680.)

b = 0.133 + 0.12 = 0.263
76 . VAPOR PRESSURES [CHAP. Ill
At the normal boiling point,

322 ^,„ 14.7 ,,V' r>^^li '>', •/.-•ut-
TV = - = 0.647; p. = — = 0.0216 _ ,^, ^ „ ^ _ :,,„,,„,,
Substituting in Equation (17),
/ I - 0.647\ ,
-A I ) = log 0.0216 + e-2«(»-"'-o-2")'
A =• 2.9703
AtO°C,
• 273 ' •' • ! • ' ' : , >',-j;iv,'
TB = = 0.549 . • J •
-2.9703(1 - 0.549) „
log pr = —- - e-ax"""-" 263)2 ^ _2.6139
0.549
Pr = 0.00243 . ^
p = 1.661b/sqin. or 85.8 mmHg ..,.,;>.
I t must be emphasized that the above generalizations are not rigorous
and are not supported by experimental data beyond the first few mem-
bers of any series. Where accuracy is important direct experimental
measurements are desirable and generalized methods are to be used for
approximations in the absence of experimental data. However, it
appears that critical pressures derived by this method are somewhat
more reliable than those from the more general method of Meissner
previously described, and that the vapor pressure relations, particularly
in the low ranges, are much better than those obtained from general
reference substance plots or relations.
The most convenient method of using Equations (16) or (17) is through
the derivation of curves on semi-logarithmic paper, plotting vapor pres-
sures on the logarithmic scale and temperatures on the uniform scale.
MIXTURES OF IMMISCIBLE LIQUIDS
Two liquids which are immiscible in each other can exist together only '
in nonhomogeneous mixtures. In such systems, where intimate mix-
ing is maintained, an exposed liquid surface will consist of areas of each
of the component liquids. Each of these components will vaporize at
the surface and tend to establish an equilibrium value of the partial
pressure of its vapor above the surface. As has been pointed out the
equilibrium vapor pressure of a hquid is independent of the relative
proportions of liquid and vapor, but is determined by the temperature
and the nature of the liquid. I t follows from kinetic theory that the
equilibrium vapor pressure of a liquid should be the same, whether it
exists alone or as a part of a mixture, if a free surface of the pure liquid
is exposed. In a nonhomogeneous mixture of immiscible liquids the
vaporization and condensation of each component takes place at the
respective surfaces of tjie pure liquids, independently of the natures or
CHAP. Ill] MIXTURES OF IMMISCIBLE LIQUIDS 77
amounts of other components which may be present. Each component

liquid actually exists in a pure state and as such exerts its normal equi-
librium vapor pressure.
The total vapor pressure exerted by a mixture of immiscible liquids
is the sum of the vapor pressures of the individual components at the
existing temperature. When the vapor pressure of such a mixture
equals the existing total pressure above its surface, it will boil, giving
off a mixture of the vapors of its components. Since each component
of the mixture adds its own vapor pressure depending only upon the
temperature, it follows that the boiling point of a nonhomogeneous mix-
ture must be lower than that of any one of its components alone. This fact
is made use of in the important industrial process of steam distillation of
materials which are insoluble in water. By mixing an immiscible ma-
terial with water it can be distilled at a temperature always below the
boihng point of water corresponding to the existing total pressure. In
this manner it is possible to distill waxes, fatty acids of high molecular
weight, petroleum fractions, and the hke, at relatively low tempera-
tures and with less danger of decomposition than by other methods of
distillation.
The composition of the vapors in equilibrium with or rising from a
mixture of immiscible liquids is determined by the vapor pressures
of the liquids. The partial pressure of each component in the vapor is
equal to its vapor pressure in the liquid state. The ratio of the partial
pressure to the total pressure gives the mole fraction or percentage by
volume, from which the weight percentage of the component in the vapor
may beicalculated. The total vapor pressure exerted by a mixture of
immiscible liquids is easily calculated as the sum of the vapor pressures
of the component liquids. Conversely, the boiling point of the mixture
under a specified total pressure is the temperature at which the sum of
the individual vapor pressures equals the total pressure. This tempera-
ture is best determined by trial or by a graphical method in which a plot
of total vapor pressure against temperature is prepared.
Illustration 7. It is proposed to purify benzene from small amounts of nonvolatile

solutes by subjecting it to distillation with saturated steam under atmospheric pres-
sure of 745 mm of Hg. Calculate (a) the temperature at which the distillation will
proceed and (6) the weight of steam accompanying 1 lb of benzene vapor.
Solution: This problem may be solved by trial, using the data of Fig. 4.
Temp. P.p. C(,H^ v.p. HS Total v.p.

60°C 390 mm 150 mm 540 mm
70°C 650 mm 235 mm 785 mm
65°C 460 mm " 190 mm 650 mm
68°C 610 mm '•• > 215 mm 725 mm
69°C • 520 mm 225 mm 745 mm
The boiling point of the mixture will, therefore, be 69°C. This result could be
obtained graphically by plotting a curve relating temperature to total vapor pressure.
Basis: 1 lb-mole of mixed vapor.
520
Benzene= -— = 0.70 lb-mole or 0.70 X 78 = 55 lb
745
Water = 0.30 lb-mole or 5.4 lb
5.4
Steam per pound of benzene = — = 0.099 lb
oo
In the preceding illustration it has been assumed that the steam
and the Uquid being distilled leave the still in the proportions determined
by their vapor pressures. This will be the case only when the liquids
in the still are intimately mixed and when the steam which is introduced
comes into intimate contact and equilibrium with the liquids. If these
conditions are not realized the proportion of steam in the vapors will be
higher than that corresponding to the theoretical equilibrium.
Illustration 8. It is desired to purify myristic acid (C13H27COOH) by distillation
with steam under atmospheric pressure of 740 millimeters. Calculate the tempera-
ture at which the distillation will proceed and the number of pounds of steam accom-
panying each pound of acid distilled.
Vapor pressure of myristic acid at 99°C = 0.032 millimeter of mercury
The vapor pressure of mjrristic acid is negligible in its effect on the boihng point of
the mixture, which may be assumed to be that of water at 740 millimeters of mercury,
or99''C.
Basis: 1 lb-mole of mixed vapors. •*
0 032
Myristic acid = - i — = 4.3 X lO"* lb-mole or 4.3 X 10-^
740
X 228 = 0.0098 lb
Water = 1.0 lb-mole = 181b
18
Steam per pound of acid = • - = 1840 lb
Vaporization with Superheated Steam. The preceding illustration

deals with an organic compound having a high boiling point which cannot
be subjected to ordinary direct distillation at atmospheric pressure. By
distillation with saturated steam the boiling point of the mixture is
reduced below 100°C, but as indicated by the results of the illustration,
an enormous amount of steam must be used in order to obtain a small
amount of product. An alternative method would be to conduct a
direct distillation under a sufficiently reduced pressure to lower the
boiling point to the desired temperature. However, the maintenance of
high vacua in apparatus suitable for the vaporization of such materials
is difficult and frequently impracticable.
CHAP. Ill] VAPORIZATION WITH SUPERHEATED STEAM 79
These difficulties may be circumvented by maintaining the material

to be vaporized at the highest permissible temperature and introduc-
ing superheated steam or some other inert gas. In this case there
will be no Uquid water in the system, and the superheated steam merely
serves as a carrier which mixes with and removes the vapors of the mate-
rial to be distilled. If the material being vaporized is allowed to reach
equilibrium with its vapor, the partial pressure of the distillate vapor will
belts equilibrium vapor pressure at the existing temperature. The partial
pressure of the steam will be the difference between the existing total
pressure and the partial pressure of the distillate vapor. The amount of
steam required per unit quantity of distillate may, therefore, be di-
minished either by raising the temperature or lowering the total pres-
sure. Distillation with superheated steam is frequently combined with
reduced pressure in order to reduce the steam requirements for the
distillation of high-boiling-point materials which will not withstand
high temperatures.
Ordinarily the mixing of the steam with the material being vaporized
will not be sufficiently intimate to result in equilibrium conditions.
The steam will then leave the liquid without being completely saturated
with distillate vapor.
Illustration 9. Myristic acid is to be distilled at a temperature of 200°C by use
of superheated steam. It may be assumed that the relative saturation of the steam
with acid vapors will be 80%.
(o) Calculate the weight of steam required per pound of acid vaporized if the
distillation is conducted at an atmospheric pressure of 740 mm of Hg.
(6) Calculate the weight of steam per pound of acid if a vacuum of 26 in. of Hg
is maintained in the apparatus.
Vapor pressure of myristic acid at 200°C = 14.5 mm of Hg
Basis: 1 lb-mole of mixed vapors.
(a) Partial pressure of acid = 14.6 X 0.80 = 11.6 mm of Hg
Quantity of acid = -^ = 0.0157 lb-mole or 0.0157 X 228 = 3.58 lb

Quantity of water = 0.9843 lb-mole or 17.7 lb
17.7
Steam per pound of acid = —— = 4.95 lb
0.58
(6) Total pressure = 740 - (26 X 25.4) = 80 mm

11.6
Quantity of acid = -—— = 0.146 lb-mole or 33.1 lb
80
Quantity of water = 0.866 lb-mole or 15.4 lb
, , ., 15.4
Steam per pound of acid = —— = 0.465 lb
80 .!'' VAPOR PRESSURES Ûln'i/ [CHAP. I l l
SOLUTIONS
The surface of a homogeneous solution contains molecules of all its
components, each of which has an opportunity to enter the vapor state.
However, the number of molecules of any one component per unit area
of surface will be less than if that component exposed the same area of
surface in the pure liquid state. For this reason the rate of vaporiza-
tion of a substance will be less per unit area of surface when in solution
than when present as a pure liquid. However, any molecule from a
homogeneous solution which is in the vapor state may strike the surface
of the solution at any point and will be absorbed by it, re-entering the
liquid state. Thus, although the opportunity for vaporization of any
one component is diminished by the presence of the others, the oppor-
tunity for the condensation of its vapor molecules is unaffected. For
this reason, the equilibrium vapor pressure which is exerted by a com-
ponent in a solution will be, in general, less than that of the pure
substance.
This situation is entirely different from that of a nonhomogeneous
mixture. In a nonhomogeneous mixture the rate of vaporization of
either component, per unit area of total surface, is diminished because
the effective surface exposed is reduced by the presence of the other
component. However, condensation of a component can take place
only at the restricted areas where the vapor molecules impinge upon
its own molecules. Thus, both the rate of vaporization and the
rate of condensation are reduced in the same proportion and the equi-
librium vapor pressure of each component is unaffected by the presence
of the others.
Raoulfs Law. The generalization known as Raoult's law states that
the equilibrium vapor pressure which is exerted by a component in a
solution is proportional to the mole fraction of that component. Thus,
VA = PA\-PA — ]^ "":
r—; ^ )-AÂ
= ^J^PA (20)
\WA+ nB + nc^ / /
where
fi, = vapor pressure of component A in solution with
components B, C, . . .
PA = vapor pressure of A in the pure state

riA, ns, nc . •. = moles of components A, B, C, . . .
NA. = mole fraction of A
From the kinetic theory of equilibrium vapor pressures it would be ex-

CHAP. Ill] EQUILIBRIUM VAPOR PRESSURE AND COMPOSITION 81
pected that this generalization would be correct when the following

conditions exist:
1. No chemical combination or molecular association takes place in
the formation of the solution.
2. The dimensions of the" component molecules are approximately
equal.
3. The attractive forces between Uke and unlike molecules are ap-
proximately equal.
4. The component molecules are nonpolar and are not adsorbed at
the surface of the solution.
Few combinations of hquids would be expected to fulfill all these con-
ditions, and it is not surprising that Raoult's law represents only a more
or less rough approximation to actual conditions. Where the con-
ditions are fulfilled, a solution will be formed from its components with-
out thermal change, and without change in total volume. A solution
which exhibits these properties is termed an ideal or perfect solution.
Solutions which approximate the ideal are formed only by liquids of
closely related natures such as the homologs of a series of nonpolar
organic compounds. For example, paraflSn hydrocarbons of not too
widely separated characteristics form almost ideal solutions in each
other. The behavior of the ideal solution is useful a^ a criterion by which
to judge solutions and also as a means of approximately predicting quan-
titative data for solutions which would not be expected to deviate widely
from ideal behavior. For the accuracy required in the majority of
industrial problems, a great many solutions of chemically similar
materials may be included in this class.
Equilibriixm Vapor Pressure and Composition. If the validity of
Raoult's law is assumed, it is necessary to have only the vapor-pressure
data for the pure components in order to predict the pressure ajid com-
position of the vapor in equilibrium with a solution. The total vapor
pressure of the solution will be the sum of the vapor pressures of the com-
ponents, each of which may be calculated from Equation (20). The
partial pressure of each component in the equilibrium vapor will be equal
to its vapor pressure in the solution, thus fixing the composition of the
vapor. • • • •••' "••!•. ^..i. , ' , •• ^-i • .•' .V -I
Illustration 10. Calculate the total pressure and the composition of the vapors in
contact with a solution at 100°C containing 35% benzene (CeHe), 40% toluene
(CeHeCHs), and 25% ortho-xylene (C6H4(CH3)2) by weight. •
Vapor pressures at 100°C:
Benzene = 1340 mm Hg
Toluene = 560 mm Hg
o-Xylene = 210 mm Hg
82 VAPOR P R E S S U R E S [CHAP. I l l
Basis: 100 lb of solution: -••.••.:

35 . ' :. ::
Benzene = 35 lb or — = 0.449 lb-mole
78
40
Toluene = 40 lb or — = 0.435 lb-mole
25
o-Xylene = 25 lb or -— = 0.236 lb-mole
106
Total = lOOlb or 1.120 lb-mole
Vapor pressures: -J* "
0.449
Benzene = 1340 X —— = 1340 X 0.401 = 536 mm Hg
0.435
Toluene = 660 X —— = 560 X 0.388 = 217 mm Hg *.
o-Xylene = 210 X'TTTT = 210 X 0.211 = 44 mm Hg
Total = 797 mm Hg
Molal percentage compositions:
Liquid Vapor !
Benzene... 40.1% 536/797 = 67.3%
Toluene 38.8% 217/797= 27.2%
o-Xylene 21.1% 44/797 = 5.5% '
100.0% 100.0%
In a similar manner the vapor pressure of the solution at any other
temperature might be calculated and a curve plotted relating total vapor
pressure to temperature. From such a curve the boiling point of the
solution at any specified pressure may be predicted. It will be noted
that the composition of the vapor may differ widely from that of the
solution, depending on the relative volatilities. In the special case of a
solution containing a nonvolatile component the vapor will contain
none of this component but its presence in the liquid will diminish the
vapor pressure of the other components in the same proportion that it
reduces their mole fractions.
Nonvolatile Solutes. If one component of a binary solution has a
negligible vapor pressure, its presence will have no effect on the com-
position of the vapor in equilibrium with the solution. The vapor
will consist entirely of,the volatile component but its equilibrium pres-
sure will be less than that of the pure liquid at the same temperature.
Thus, a nonvolatile solute produces a vapor-pressure lowering or a boiling
point elevation in its solvent. If the components possess closely related
characteristics the system may approach ideal behavior. In this case
the total vapor pressure will be the product of the vapor pressure and
CHAP. Ill] NONVOLATILE SOLUTES 83
the mole fractiou of the solvent. With ionizing or associating solutes

the effective mole fraction of the solute is dependent upon the degree
of ionization or association. For these reasons, the theories of ideal
behavior are of little assistance in the estimation of vapor-pressure data
for many solutions, particularly those in which water is the solvent.
wo
60
~1 10 ;15
40 lY >
t" 20
-20
20 40 60 80 100 120
Temperature of NaOH Solutions, °C
FIG. 6 Diihring lines of aqueous solutions of sodium hydroxide.
If the vapor pressure of a solution is known at two temperatures

these data will establish a straight line on a reference substance chart
prepared according to either the method of Cox or that of Diihring, pages
64 and 65. In Fig. 6 are the Diihring lines corresponding to va-
rious concentrations of aqueous sodium hydroxide solutions. Where
sufficient data are available, it is advisable to plot the temperatures of
the solutions against those of the pure solvent, in this case water.
Using this method the curve representing zero concentration of solute
will be a straight line of unit slope. By interpolation between a set of
Diihring lines the boiling point of a solution under any desired pressure
or the vapor pressure at any temperature may be estimated.
Similarly, the Cox type of plot may be applied to solutions, preferably
using one of the components as the reference substance. Such a plot,
developed by Othmer for sulfuric acid solutions is shown in Fig. 7.
This type of plot has the advantage over the Diihring plot of permitting
estimation of thermal data from the slopes of the vapor pressure lines
and Equation (6), and may or may not give closer approximation to
straight Une relationships, depending on the particular system involved.
The difference between the boiling point of a solution and that of the
pure solvent is termed the boiling-point elevation of the solution. It
will be noted that the lines of Fig. 6 diverge but slightly at the higher
temperatures. It follows that the boiling-point elevation of a solution
of sodium hydroxide is practically independent of temperature or pres-
Temperature °C
20 30 40 50 60 70 80 90 100 110 120 130140150160170180
1000 T — •<"
1
1 1T
1 1 '• \ /
800
600
1 i
/\ 1 ]/i 1A
1 1 / 1 I/I 1 1 / 1
400 1 1 1 /
i/i i/i 1
L^^l
1 1 <*
A1
-3 200
1
?^ 1 4
•^
«i 1
100
80 X i 1 \/\ 1/ 1
1
60 1/ 1 1 / \ /
/ 1 1 /I 1 /i
40 1 1 1/1 1 [I
1 t 1 / «^' 1
0< y^* 'A A ?l / ! *
1A
3
\M A 1
/ ' 1 1
1
I /
/ 1
1
1 . 1
i
1
1
> 10
1A 1 '/i 1
10 20 40 60 80100 200 400 600 8001000 2000 4000 6000 10,000
Vapor Pressure of Water in mm Hg
FIG. 7. V a p o r p r e s s u r e of sulfuric a c i d s o l u t i o n s . [ P r o m O t h m e r , Ind. Eng. Chem.,

32, 8 4 7 (1940), w i t h p e r m i s s i o n . ]
sure. Although several systems exhibit this behavior it can by no means

be considered general and the fact that the Diihring lines of sodium
hydroxide solutions happen to be almost parallel is merely a character-
istic of this system. It may be easily demonstrated that the boiling-
point elevation of an ideal solution increases rapidly with an increase
in temperature.
Relative Vapor Pressure. When it is desired to approximate the com-
plete vapor-pressure data for a solution from only a single experimental
observation, a modified form of Raoult's law will frequently give good
results:
p = kpo (21)
CHAP. Ill] PROBLEMS 86
where p = vapor pressure of solution.

po = vapor pressure of pure solvent,
fc = a factor, dependent on concentration.
For an ideal solution the factor k will equal the mole fraction of the sol-
vent and will be independent of temperature or pressure. For non-
ideal solutions k may differ widely from the mole fraction but in many
cases it will be practically independent of temperature or pressure for a
solution of a given composition. The factor k is sometimes termed the
relative vapor pressure of a solution. The value of k for a solution may be
obtained by a single determination of boiling point. Equation (21)
may then be used to estimate the vapor pressures at other temperatures.
Illustration 11. An aqueous solution of sodium chloride contains 5 gram-moles of
NaCl per 1000 g of water. The normal boiUng point of this solution is 106° C.
Estimate its vapor pressure at 25°C. ; ' ,', .
Mole fraction of water = ( 1 ) = 1 - 0.0826 = 0.9174

V 55.5 + 6 /
From Table I, po at 106°C = 940. mm of Hg
Po at 25°C = 23.5 mm of Hg
760
• fc = — =0.81
. . , , , • : 940
Vapor pressure of solution at 25°C = 23.5 X 0.81 = 19.0 mm Hg
Experimentally observed value = 18.97 mm Hg
In the preceding illustration it will be noted that the value of k is

widely different from the mole fraction of the solvent. However, for this
particular system it is, for all practical purposes, independent of tem-
perature. In certain systems, notably aqueous solutions of strong bases,
such constancy does not exist and the relative vapor pressure may vary
considerably. For example, a solution of caustic soda having a molality
of 10 has a relative vapor pressure, k, ef 0.584 at 100°C and only 0.479
at 25°C. When the type of behavior of a system is unknown it is de-
sirable to obtain at least two experimental points and to establish a
Diihring line if it is desired to predict reliable values of vapor pressure.
PROBLEMS
1. (a) Obtaining the necessary data from a physical table, plot a curve relating
the vapor pressure of acetic acid (C2H4O2) in millimeters of Hg to tem-
perature in degrees Centigrade. Plot the curve for the temperature
range from 20° to 140°C, using vapor pressures as ordinates and tempera-
tures as abscissas, both on uniform scales. K; ;«::!«! ;/,
(6) Ethylene glycol (OHCHj-CHzOH) has a normal boiling point of 197°C.

At a temperature of 120 °C it exerts a vapor pressure of 39 mm of Hg.
From these data construct a Diihring line for ethylene glycol using water
as the reference substance. From this line estimate the vapor pressure
at 160°C and the boiUng point under a pressure of 100 mm of Hg.
(c) Ethyl bromide (CjHsBr) exerts a vapor pressure of 165 mm of Hg at 0°C
, and has a normal boiUng point of 38.4°C. From Fig. 4 estimate its vapor
pressure at 60°C.
(d) Nonane (CsHjo) has a normal boiling point of 150.6°C. From Fig. 4
estimate its boiling point under a pressure of 100 mm of Hg.
2. For one of the substances from the following list, prepare a table of data and
graphs as indicated below:
Substance Temperature Range Pressure Units
Ammonia -30°C to24°C atm
Carbon dioxide -38°C to Critical Temp. atm
Carbon disulfide 0°C to 46.3°C mm Hg
Chlorine -34.6°C to 30°C abn,
Sulfur dioxide 0°C to 50°C atm
Acetone 10°C to 60°C mm Hg
Carbon tetrachloride 26°C to 80°C mm Hg
Chloroform 10°C to 70°C mm Hg
Ethyl alcohol 35°C to 80°C mm Hg
Ethyl ether 34.6°C to 120°C atm
Methyl chloride -24.0°C to 40°C atm
(a) From a handbook of physical-chemical data, tabulate the following values.
Indicate source of data.
Column No. 1. Temperature, °C (t)
" " 2. Absolute temperature, °K (T)
' " 3. 1/T
" " 4. Vapor pressure (p) '
(6) Using graph paper with uniform scales, plot vapor pressures as ordinates
and temperature as abscissas. Draw a smooth curve through the plotted
points.
(c) Using semi-log paper, plot vapor pressures as ordinates on the logarithmic
scale versus l/T as abscissas. Choose the scale for 1/T so that the curve
will have a slope as close to 45° as feasible. Draw a smooth curve through
the plotted points.
(d) Using semi-log paper plot vapor pressure as ordinates and temperature
(°C) as abscissas.
3. Construct a Diihring chart and a Cox chart for ethyl alcohol, using water as
the reference substance. Plot the curve using the following data:
Temperature Vapor Pressm-e of JEthyl Alcohol
10°C 23.6 mm Hg
78.3°C 760.0 mm Hg
Using the Diihring line based on these figures, determine;
(a) The vapor pressure at 60°C.
(6) The boiling point under 500 mm pressure
CHAP. Ill] PROBLEMS 87
4. Use the Cox chart (Fig. 4) to determine the boihng point at 2000 mm Hg of one
of the following substances. Use the vapor pressure data that are given below to
establish the line on the Cox chart.
Ethyl Acetate Ethyl Formate Sulfur

0°C 24.2 mm Hg 0°C 72.4 mm Hg 250°C 12 mm Hg
160°C 8.349 atm 200°C 28.0 atm 444.6°C 760.0 mm Hg
5. The boiling point of benzene is 80°C and its liquid density at the boihng point is
0.80 gram per cc. Estimate its critical temperature both by Equation (8) and Equa-
tions (11) to (14).
6. In the following tabulation, the normal boiling point, critical temperature, and
liquid density at the normal boiling point are given for several hydrocarbons. For
one of the materials assigned from this list, calculate the critical temperature, using
Equation (8) and Fig. 5. A comparison of this computed value with the experimen-
tal value will ^ve an indication of the accuracy that may be expected by this method.
Normal Critical Liquid Density
Hydrocarbon Boiling Point Temperature at Normal
°C °C Boihng Point
n-Butane i -0.3 153 0.605 g/cc
n-Pentane 36.1 , 197 0.610
n-Hexane 69.0 235 0.613
n-Heptane 98.4 267 0.613
n-Octane 125.8 296 0.611
Cyclohexane 80.8 281 0.719
7. Using equations (9) to (15) calculate the critical properties of the following
compounds and compare them with the tabulated experimental values:
Compound TB°^ rc°K p^ atm Vc cc/gr mole

Acetone 329.0 508.2 47.0 216.5
Acetic acid 391.0 594.8 57.2 171.0
Ammonia 239.6 405.6 111.5 72.5
Chlorobenzene 405.0 632.2 44.6 308.0
Trifluorotrichloroethane 322.0 487.0 34.0
Ethyl alcohol 351.0 516.3 63.1 167.0
8. Using the data of Fig. 5a and Table lib calculate the necessary data and plot
vapor pressure curves extending from 0.01 lb per sq in. to the critical point for the
following substances. Plot the curves on five cycle semi-logarithmic paper, using
vapor pressures as ordinates on the logarithmic scale and temperatures in °F as
(a) Decane {U = 347°C) {is = 174°C)

(6) Toluene [U = 320.6°C) ((B = 110.8°C)
(c) Propyl ethyl ether «. = 227.4°C) (<B = 64°C)
9. Using the data of Fig. 4 estimate the temperature required for the distillation of
hexadecane (C16H34) at a pressure of 750 mm Hg in the presence of liquid water.
Calculate the weight of steam evolved per pound of hexadecane distilled.
10. A fuel gas has the following analysis by volume (in the third column are the
normal boiling points of the pure components): ^ ''
Components Percentage Boiling Point
Ethane (CsHj) 4.0 -88°C -"
' <;-i| Propane (CsHs) 38.0 v -44°C .•,:,; 0
,'iH Isobutane (C4H10) 8.0 - -10°C . ?;
Normal butane (C4H10)..,. 44.0 0°C
f,!3,,; Pentanes (C6H12) 6.0 ,: -1-30°C (average) .3
100.0 " • '"•"•
It is proposed to liquefy this gas for sale in cylinders and tank cars.
(o) Calculate the vapor pressure of the liquid at 30°C and the composition of the
vapor evolved. (The vapor pressures may be estimated from Fig. 4 and
the normal boiUng points.)
(fc) Calculate the vapor pressure of the liquid at 30°C if all the ethane were
removed.
11. Assuming that benzene (CsHe) and chlorobenzene (CeHsCl) form ideal solu-
tions, plot curves relating total and partial vapor pressures to mole percentages of
benzene in the solution at temperatures of 90°, 100°, 110°, and 120°C. Also plot the
curves relating total vapor pressure to the mole percentage of benzene in the vapor
at 90 and 120°G. The normal boiling point of benzene is 79.6°C. Following are
other vapor pressure data:
Vapor Pressure — mm Hg
Temperature Chlorobenzene Benzene
90°C 208 1013

100 293 1340
s\ 110 403 1744
120 542 2235
132.1 760 2965
12. From the curves of Problem 11, plot the isobaric boiling-point curves of ben-
zene-chlorobenzene solutions under a pressure of 760 mm of Hg. The points on both
the liquid and vapor curves corresponding to temperatures of 90°, 100°, 110°, and
120°C should be used in establishing the curves.
13. An aqueous solution of NaNOs containing 10 gram moles of solute per 1000 g
of water boils at a temperature of 108.7°C under a pressure of 760 mm of Hg. Assum-
ing that the relative vapor pressure of the solution is independent of temperature,
calculate the vîpor pressure of the solution at 30°C and the boiling-point elevation
produced at this pressure.
14. Obtaining the necessary data from the section on " Boiling Points of Mixtures "
of Volume III of the International Critical Tables, plot the isobaric boihng-point
curves for the system acetic acid (CH3COOH) and water under a pressure of 760 mm
of Hg. Both the Uquid and vapor composition curves should be plotted using per-
centages of water by weight as abscissas and temperatures in degrees Centigrade
as ordinatea.
' ! .: , >• CHAPTER IV
HUMIDITY AND SATURATION

When a gas or a gaseous mixture remains in contact with a Uquid sur-
face, it will acquire vapor from the Uquid until the partial pressure of
the vapor in the gas mixture equals the vapor pressure of the liquid at
its existing temperature. When the vapor concentration reaches this
equilibrium value the gas is said to be saturated with the vapor. It is
not possible for the gas to contain a greater stable concentration of vapor,
because as soon as the vapor pressure of the liquid is exceeded by the
partial pressure of the vapor condensation takes place. The vapor
content of a saturated gas is determined entirely by the vapor pressure
of the liquid and may be predicted directly from vapor-pressure data.
The pure-component volume of the vapor in a saturated gas may be
calculated from the relationships derived in Chapter II. Thus, if the
ideal gas law is applicable:
V,= V^ (1)
V
where
Vv = pure-component volume of vapor
Pv = partial pressure of vapor = the vapor pressure of
' the liquid at the existing temperature
V = total volume
p = total pressure
From Equation (1) the percentage composition by volume of a vapor-
saturated gas may be calculated. When the ideal gas law is applicable,
the composition by volume of a vapor-saturated gas is independent of
the nature of the gas but is dependent on the nature and temperature of
the liquid and on the total pressure. The composition by weight varies
with the natures of both the gas and the liquid, the temperature, and
the total pressure.
For certain types of engineering problems it is convenient to use
special methods of expression for the vapor content of a gas. The weight
of vapor per unit volume of vapor-gas mixture, the weight of vapor per
unit weight of vapor-free gas, and the moles of vapor per mole of vapor-
free gas are three common and useful methods of expression. When a gas
89
90 HUMIDITY AND SATURATION [CHAP. IV
is saturated with vapor, the composition expressed by the first method

is independent of both the nature of the gas and the total pressure but
varies with the nature and temperature of the liquid. When the compo-
sition is expressed by the third method it varies with the nature of the
liquid, the temperature, and the pressure but is independent of the
nature of the gas. From a knowledge of the equilibrium vapor pressure
of the liquid the compositions of vapor-saturated gases may be readily
calculated in any of these methods of expression, using the principles
developed in Chapter I I I . |> ,.
Illustration 1. Ethyl ether at a temperature of 20°C exerts a vapor pressure of
442 mm of Hg. Calculate the composition of a saturated mixture of nitrogen and
ether vapor at a temperature of 20°C and a pressure of"745 mm of Hg expressed in
the following terms:
(a) Percentage composition by volume.
(b) Percentage composition by weight. >
(c) Pounds of vapor per cubic foot of mixture.
(d) Pounds of vapor per pound of vapor-free gas.
(e) Pound-moles of vapor pej pound-mole of vapor-free gas.
(a) Basis: 1.0 ou ft of mixture.
Pure-component volume of vapor = 1.0 X
442
rrrr = 0.593 cu ft
745
Ether vapor 59.3%
Nitrogen 40.7%
(6) Basis: 1.0 lb-mole of the mixture.
Vapor present = 0.593 lb-mole or 43.9 lb
Nitrogen present = 0.407 lb-mole or 11.4 lb
Total mixture 55.3 lb
Composition by weight:
Ether vapor 79.4%
Nitrogen 20.6%
(c) Basis: Same as (6).
760 293
Volume = 359 X — X — = 393 cu ft
43.9
Weight of ether per cubic foot = — - = 0.112 lb
ovo
This result is independent of the total pressure. For example, an increase in the
total pressure would decrease the volume per mole of mixture but would correspond-
ingly decrease the weight of vapor per mole of mixture.
(d) Basis: Same as (&).
43.9
Weight of vapor per pound nitrogen = —— = 3.85 lb
CHAP. IV] PARTIAL SATURATION 91
(e) Basis: Same as (6).

0.593
Moles of vapor per mole of nitrogen = = 1.455
Partial Saturation. If a gas contains a vapor in such proportions that

its partial pressure is less than the vapor pressure of the hquid at the
existing temperature, the mixture is but partially saturated. The rela-
tive saturation of such a mixture may be defined as the percentage ratio
of the partial pressure of the vapor to the vapor pressure of the hquid
at the existing temperature. The relative saturation is therefore a
function of both the composition of the mixture and its temperature as
well as of the nature of the vapor.
From its definition it follows that the relative saturation also repre-
sents the following ratios:
a. The ratio of the percentage of vapor by volume to the percentage
by volume which would be present were the gas saturated at the existing
temperature and total pressure.
b. The ratio of the weight of vapor per unit volume of mixture to
the weight per unit volume present at saturation at the existing temper-
ature and total pressure.
Another useful means for expressing the degree of saturation of a
vapor-bearing gas may be termed the percentage saturation. The per-
centage saturation is defined as the percentage ratio of the existing weight
of vapor per unit weight of vapor-free gas to the weight of vapor which
would exist per unit weight of vapor-free gas if the mixture were satu-
rated at the existing temperature and pressure. The percentage satu-
ration also represents the ratio of the existing moles of vapor per mole
of vapor-free gas to the moles of vapor which would be present per mole
of vapor-free gas if the mixture were saturated at the existing tempera-
ture and pressure.
Care must be exercised that the relative saturation and the percentage
saturation are not confused. They approach equality when the vapor
concentrations approach zero but are different at all other conditions.
The quantitative relationship between the two terms is readily derived
from their definition. Thus,
Relative saturation = — X 100 = s, (2)

where
Pa = partial pressure of vapor actually present
p, = partial pressure at saturation Jt:,^
Ua
Percentage saturation = — X 100 = Sp (3)
where
Ua = moles of vapor per mole of vapor-free gas actually present
Ua = moles of vapor per mole of vapor-free gas at saturation
from Dalton's law, : ' , .
• • • • - • •• . M l ' - ' - ,
^ • na Pa , ns Ps ' ,^^
—= and - = (i)
1 P - Pa 1 p - Pa
or
, , . ,. n^^pa/y-pA •• ^
\ n^ p, \p - Pa/ _ ,. • , \;
hence . , : < ,
Sp = Sr (?^^^) (6)
\P - Pa/
where
p = total pressure • , ' ,'
Illustration 2. A mixture of acetone vapor and nitrogen contains 14.8% acetone
by volume. Calculate the relative saturation and the percentage saturation of the
mixture at a temperature of 20''C and a pressure of 745 mm of Hg.
The vapor pressure of acetone at 20°C is 184.8 mm of Hg
Partial pressure of acetone = 0.148 X 745 -.. 110.0 mm of Hg
Relative saturation = 110/184.8 59.7%
Basis: 1.0 lb-mole of mixture. • , .
Acetone 0.148 lb-mole
Nitrogen 0.852 lb-mole
Molesof acetone per mole of nitrogen = 0.148/0.852 0.174
Basis: 1.0 lb-mole of saturated mixture at 20°C and 745 mm of Hg.
Percentage by volume of acetone = 184.8/745 24.8% ,.,,,
Lb-moles of acetone 0.248
Lb-moles of nitrogen 0.752
Moles of acetone per mole of nitrogen = 0.248/0.752 0.329
Percentage saturation = 0.174/0.329 52.9%
As indicated by this illustration, the percentage saturation is always

somewhat smaller than the relative saturation.
The composition of a partially saturated gas-vapor mixture is fixed
if the relative or percentage saturation and the temperature and pressure
are specified. From this information and a knowledge of the equilibrium
vapor pressure at this temperature the composition may be expressed
in any other terms. Conversely, the relative or percentage saturation
may be calculated if the composition, pressure, and temperature are
specified. The temperature required to produce a specified degree of
CHAP. IV] THE DEW POINT 93
saturation may be calculated if the composition at a specified pressure

is known.
Dlustration 3. Moist air is found to contain 8.1 grains of water vapor per cubic
foot at a temperature of 30°C. Calculate the temperature to which it must be
heated in order that its relative saturation shall be 15%.
Basis: 1 cu ft of moist air. • •
81
Water = — ^ = 1.16 X 10"' lb or 6.42 X 10"' lb-mole
Pure-component volume of water vapor = 6.42 X 10"'

X 359 0.0230 cu ft at S.C.
0.0230 303
Partial pressure of water vapor = 760 X X •—•.. 19.4 mm of Hg
Vapor pressure of water at temperature correspond-

19.4
ing to 15% relative saturation = —'-- 130 mm of Hg
0.15
From the vapor-pressure data for water it is found that this pressure corresponds
to a temperature of 57°C.
Humidity. Because of the widespread occurrence of water vapor in

gases of all kinds, special attention has been given to this case and a
special terminology has been developed. The humidity of a gas is gener-
ally defined as the weight of water per unit weight of moisture-free gas.
The molal humidity is the number of moles of water per mole of moisture-
free gas. When the vapor under consideration is water the percentage
saturation is termed the percentage humidity, Hp. The relative satura-
tion becomes the relative humidity, Hr. The relation between these two
humidities follows from Equation (6) as
. . "-'"'{v^J
\P - Pa/ «. iH. <"
Considerable confusion exists in the literature in the use of these terms,
and care must always be exercised to avoid misuse. The terminology
recommended above is an extension of that proposed by Grosvenor.'
The Dew Point. If an unsaturated mixture of vapor and gas is
cooled, the relative amounts of the components and the percentage
composition by volume will at first remain unchanged. It follows that,
if the total pressure is constant, the partial pressure of the vapor will
be unchanged by the cooUng. This will be the case until the temperature
is lowered to such a value that the vapor pressure of the liquid at this
temperature is equal to the existing partial pressure of the vapor in the
mixture. The mixture will then be saturated, and any further cooling
1 Trans. Am. Inst. Chem. Eng., 1 (1908). ;«;-UHHOM'sr-w m' :'=
will result in condensation. The temperature at which the equilibrium

vapor pressure of the liquid is equal to the^ existing partial pressure of
the vapor is termed the dew point of the mixture.
The vapor content of a vapor-gas mixture may be calculated from
dew-point data, or conversely, the dew point may be predicted from the
composition of the mixture. , ——
^
Illustration 4, A mixture of benzene vapor and air contains 10.1% benzene by
volume.
(a) Calculate the dew point of the mixture when at a temperature of 25°C and a
(6) Calculate the dew point when the mixture is at a temperature of 30°C and a
(c) Calculate the dew point when the mixture is at a temperature of 30°C and a
Solution: • • ••
(a) Partial pressure of benzene = 0.101 X 750 = 75.7 mm Hg
From the vapor pressure data for benzene, Fig. 4, it is found that this pressure cor-
responds to a temperature of 20.0°C, the dew point.
(6) Partial pressure of benzene 75.7 mm
Dew point 20.0°C
(c) Partial pressure of benzene = 0.101 X 700 70.7 mm Hg
The temperature corresponding to a vapor pressure of 70.7 mm of Hg is found to be
18.7°C. From these results it is seen that the dew point does not depend on the
temperature but does vary with the total pressure.
VAPORIZATION PROCESSES
The manufacturing operations of drying, air conditioning, and certain
types of evaporation all involve the vaporization of a liquid into a stream
of gases. In dealing with such operations it is of interest to calculate the
relationships between the quantities and volumes of gases entering and
leaving and the quantity of material evaporated. Such problems are of
the general class which WBS discussed in Chapter II under the heading
of " Volume Changes with Change in Composition." The concentra-
tions of vapor in these problems are generally expressed in terms of the
dew points, the relative saturations, or the moles of vapor per mole of
vapor-free gas. The first two methods of expression are convenient
because they are directly determined from dew point or wet- and dry-
bulb temperature measurements. From such data the partial pressures
of vapor may be readily calculated and the partial pressure method of
solution might be used as described in Chapter II,
The vaporization processes all require the introduction of energy in
the form of heat. The effective utilization of this heat is frequently the
most important factor governing the operation of the process, and a
knowledge of the relationships between the quantity of heat introduced
CHAP. IV] VAPORIZATION PROCESSES 95
and that dissipated in various ways is of great significance. The calcu-

lation of such an energy balance is greatly simplified if the quantities of all
materials concerned are expressed in molal or weight units rather than
in volumes. These units have the advantage of expressing quantity
independent of change of temperature and pressure. The same desira-
bility of weight or molal units arises when relationships are derived for
the design of vaporization equipment. For these reasons it has become
customary to express all data in either weight or molal units where ther-
mal calculations are to be made or where design relationships are to be
used. The molal units are preferable. From data expressed in molal
units, volumes at any desired conditions may be readily obtained.
It will be noted that in any mixture following the ideal gas law the
ratio of the number of moles of vapor to the number of moles of vapor-free
gas is equal to the ratio of the partial pressure of the vapor to the partial
pressure of the vapor-free gas. Accordingly, vapor concentration in
moles of vapor per mole of vapor-free gas is obtained by dividing the par-
tial pressure of the vapor by the partial pressure of the vapor-free gas.
Illustration 5. It is proposed to recover acetone, which is used as a solvent in an
extraction process, by evaporation into a stream of nitrogen. The nitrogen enters
the evaporator at a temperature of 30°C containing acetone such that its dew point
is 10°C. It leaves at a temperature of 25°C with a dew point of 20°C. The baro-
metric pressure is constant at 750 mm of Hg.
(a) Calculate the vapor concentrations of the gases entering and leaving the evapo-
rator, expressed in moles of vapor per mole of vapor-free gas.
(6) Calculate the moles of acetone evaporated per mole of vapor-free gas passing
through the evaporator.
(c) Calculate the weight of acetone evaporated per 1000 cu ft of gases entering
the evaporator.
(d) Calculate the volume of gases leaving the evaporator per 1000 cu ft entering.
Solution: The vapor pressure of acetone is:
116 mm of Hg at ICC
185 mm of Hg at 20°C
(a) Entering gases: "• ' ,__ .
Partial pressure of acetone 116 mm of Hg
Partial pressure of nitrogen = 750 — 116 634 mm of Hg
Moles of acetone per mole of nitrogen = 116/634... 0.183
Leaving gases:
Partial pressure of acetone 185 mm of Hg
Partial pressure of nitrogen = 750 — 185 565 mm of Hg
Moles of acetone per mole of nitrogen = 185/565... 0.328
(b) Basis: 1.0 lb-mole of nitrogen.
Acetone leaving the process 0.328 lb-mole
Acetone entering the process 0.183 lb-mole
Acetone evaporated 0.145 lb-mole
(c) Basis: 1.0 lb-mole of nitrogen. < , •,;; ,;.

Total gas entering the process = 1.0 + 0.183 1.183 lb-moles
Volume of gas entering = 1.183 X 359 X — X — 477 cu ft
Molecular weight of acetone 58
Weight of acetone evaporated = 58 X 0,145 8.4 lb
Acetone evaporated per 1000 cu ft of gas entering =
1;^ X 1000 17.6 lb
477
(d) Basis: 1.0 lb-mole of nitrogen.
Total gas leaving the process = 1.0 + 0.328 1.328 lb-moles '
Volume of gas leaving = 1.328 X 359 X •=— X-=- = 526 cu ft
750 273
Volume of gas leaving per 1000 cu ft entering the ' , . . ;«=; s
526
process = 7== X 1000 1102 cu ft
CONDENSATION
The relative saturation of a partially saturated mixture of vapor and
gas may be increased in two ways without the introduction of additional
vapor. If the temperature of the mixture is reduced, the vapor concen-
tration corresponding to saturation is reduced, thereby increasing the
relative saturation even though the existing partial pressure of the vapor
is unchanged. If the total pressure is increased the existing partial
pressure of the vapor is increased, again increasing the relative satura-
tion. Thus, by sufficiently increasing the pressure or reducing the tem-
perature of a vapor-gas mixture it is possible to cause it to become satu-
rated, the existing partial pressure of vapor equaling the vapor pressure
of the liquid at the existing temperature. Further reduction of the
temperature or increase of the pressure will result in condensation, since
the partial pressure of the vapor cannot exceed the vapor pressure of the
liquid in a stable system.
In problems dealing with condensation processses, four interdepend-
ent factors are to be considered: the initial composition, the final tem-
perature, the final pressure, and the quantity of condensate. It may be
desired to calculate any one of these factors when the others are known
or specified. Such calculations are readily carried out by selecting as a
basis a definite quantity of vapor-free gas and calculating the quantities
of vapor which are associated with it at the various stages of the process.
In a condensation process the final conditions will be those of saturation
at the final temperature and pressure. Any one of the three methods of
calculation demonstrated in Chapter II under " Volume Changes with
Change in Composition " may be used. However, for the reasons
CHAP. IV] CONDENSATION 97
given in the preceding section it is generally desirable to express the

vapor concentrations in moles of vapor per mole of vapor-free gas if any
thermal or design calculations are to be carried out.
Illustration 6, Air at a temperature of 20 "C and a pressure of 750 mm of Hg

has a relative humidity of 80%.
(a) Calculate the molal humidity of the air.
(ft) Calculate the molal humidity of this air if its temperature is reduced to 10°C
and its pressure increased to 35 lb per sq in., condensing out some of the water.
(c) Calculate the weight of water condensed from 1000 cu ft of the original wet
air in cooling and compressing to the conditions of part (6).
(d) Calculate the final volume of the wet air of part (c).
Vapor pressure of water:
17.5 mm of Hg at 20°C . ,
9.2 mm of Hg at 10°C =
Solution:
(a) Initial partial pressure of water = 0.80 X 17.5 14.0 mm of Hg
14.0
Initial molal humidity = 0.0190
^ 750 - 14.0
(b) Final partial pressure of water 9.2 mm of Hg
Final total pressure = 36 X -rrz 1810 mm of Hg •'

9.2
Final molal humidity = '—-— 0.0051
^ 1810 - 9.2
Basis: 1000 cu ft of original wet air.
(c) Partial pressure of dry air = 750 — 14 736 mm of Hg
736 273
Partial volume of dry air at S.C. = 1000 X — ; X — = 903 cu ft
Moles of dry air = 903/359 2.52 Ib-molea

Water originally present = 2.52 X 0.0190 0.0478 lb-mole
Water finally present = 2.52 X 0.0051 0.0128
Water condensed = 0.0360 lb-mole or 0.630 lb
(d) Total wet air finally present = 2.52 -1- 0.0128 2.53 Ib-moIea
760 283 • ''
Final volume of wet air = 2.63 X 369 X r r — X r r - 396 cu ft
1810 273
Illustration 7. A mixture of dry flue gases and acetone at a pressure of 750 mm
of Hg and a temperature of 30°C has a dew point of 25°C. It is proposed to con-
dense 90% of the acetone in this mixture by cooling to 5°C and compressing. Calcu-
late the necessary pressure in pounds per square inch.
Vapor pressure of acetone:
at 30°C = 282.7 mm of Hg
- - • • ' ' at 25°C = 229.2 mm of Hg •.' • .«- ,
• v'-' -' at 5°C = 89.1 mm of Hg „ ... , •
Solution: "'"
Basis: 1.0 lb-mole of original mixture. Jr.
Partial pressure of acetone 229.2 mm of Hg
Acetone present = 229.2/750 0.306 lb-mole
Flue gases present 0.694 lb-mole
Acetone present in final mixture = 0.10 X 0.306 0.0306 lb-mole
Final mixture of gas = 0.694 -|- 0.0306 0.725 lb-mole
Partial pressure of acetone in final mixture 89.1 mm
(the vapor pressure at 5°C) . • -,
Mole percentage of acetone in final mixture =
0.0306/0.725 4.22%
89.1
Final pressure = = 2110 mm of Hg or 40.8 lb per sq in.
WET- AND DRY-BULB THERMOMETRY ^

When a liquid evaporates into a large volume of gas at the same tem-
perature, if no heat is supplied from an external source the liquid will be
cooled as a result of the heat removed from it to supply the relatively
large demand of the latent heat of vaporization. If the body of liquid
has a large surface area in proportion to its mass, its temperature will
quickly drop to an equilibrium value. This equihbrium temperature
which is assumed by the evaporating liquid will be such that heat will
be transferred to it from the warmer gas at a rate which is just adequate
to supply the necessary latent heat for the vaporization. The equilibi
rium temperature of a liquid when vaporizing into a gas is termed the
wet-bulb temperature and is always less than the actual dry-bulb tempera-
ture of the gas into which evaporation is taking place. If the gas is
saturated there is neither vaporization nor depression of the wet-bulb
temperature. It is therefore possible to use the depression of the wet-
bulb temperature as a measure of the degree of unsaturation of the mix-
ture of gas and vapor.
Although, by the evaluation of the required equations or charts, it is
possible to extend wet- and dry-bulb thermometry to the determihation
of concentrations of different types of vapors, the only systems for which
the method has been extensively used are those involving water vapor.
This special appUcation of wet- and dry-bulb thermometry is termed
hygrometry or psychrometry. The United States Weather Bureau has
worked out extensive psychrometric tables and charts from which
humidities of air may be obtained with accuracy. For engineering
purposes humidity charts have been developed which permit the deter-
mination of the humidities of any of the common gases from wet- and
dry-bulb thermometric data. The limitation of these charts is that
each chart is applicable only to systems which exist under a single, specified
CHAP. IV] THE HUMIDITY CHART 99
total pressure. For other total pressures it is necessary to develop

other charts. The humidity charts are useful for rapid solutions of
many of the problems of vaporization, condensation, and air-conditioning
processes which occur at substantially constant atmospheric pressure.
Charts exactly similar to the humidity charts may be calculated for
other systems of Hquids and gases from existing data. The construction
of such charts is justified only where a considerable amount of attention
is to be devoted to a single system.
The Humidity Chart. In Fig. 8 is plotted a molal humidity chart
covering a range of molal humidities from 0.00 to 0.34, and a range of
dry-bulb temperatures from 80° to 210°F. A chart of this type was
described in the literature by Hatta.^ A lower range on a semilog scale
is shown in Fig. 9. Values of molal humidities are plotted as ordi-
nates and dry-bulb temperatures as abscissas. The chart is based on a
total pressure of 1.0 atmosphere. Included on the same chart is curve
C showing the relationship between the molal heat of vaporization of
water as ordinates and temperature as abscissas.
Curve A of Fig. 8, termed the saturation curve, is a plot of molal
humidity against temperature for any gas which is saturated with water
vapor. Curves A' express similar relationships between molal humidi-
ties and temperatures corresponding to the specified values of percentage
humidity. These curves are all independent of the nature of the gas
under consideration.
Curves B are fines of constant wet-bulb temperature when the gas is
composed of only diatomic components such as oxygen, nitrogen, carbon
monoxide, hydrogen, air, and the like. One of these curves indicates
the wet-bulb temperature of any mixture whose dry-bulb temperature
and molal humidity determine a point falling on this curve. By inter-
polation between these curves it is possible to estimate any one of the
properties of molal humidity, wet-bulb temperature, and dry-bulb tem-
perature if the other two are known. It will be noted that where the
wet-bulb temperature fines cross the saturation curve A the wet-bulb
temperature must be equal to the dry-bulb temperature.
At wet-bulb temperatures of 100° and 150°F are groups of curves
which indicate the effect on the wet-bulb temperature fines of the pres-
ence of triatomic gases such as carbon dioxide. The symbol x indicates
the mole fraction of triatomic gases in the dry gas mixture. All wet-bulb
lines on the chart for which no value of x is designated correspond to a
composition of x = 0, The wet-bulb temperature fines which corre-
spond to various mixtures of gases must converge at the saturation curve
where the wet-bulb temperature is equal to the dry-bulb, regardless of
« Chem. and Met. Eng., 37, 137 (1930).
100 HUMIDITY AND SATURATION ' [CHAP. IV
the dry gas composition. These lines permit appUcation of the chart
to mixtures of all the common gases.
A line on the chart which is parallel to the temperature axis represents
a change of temperature without change in molal humidity. This fact
0.34
/
•^^g
11
0.32 19,400
^ ^ 1 ^ ^ )
^ ^
0.30
0.28
45i 1/ 1 19,200
- - - - := - 19,000
~-L
0.26
^c , 4. 18,800 Ji
I
^'- 140j
024
0.22
f" >Sf &LU
ft
-i/^' /
1
18,600 h
3
18,400 «
3S1
fh
0.20 18,200 .§
p a
I 0.18 130y 18,000^
o
3
s 17.800 X
1
125> ' '
3 0.16
17,600
120.
0.14
17.400 .
115.
0.12
17,200
110
0.10 105, /90V
^70/* ~60-^ 36 20
ir~ ^1 5 2.5 ?b
17.000
LOO,
0.08
0.06
—
0.04
fca<
1 ^
^x 0.00
/ / /
0.02 p
^ ^ on / /
—T
0.00
80 90 100 110 120 130 140 ISO 160 170 180 190 200 210
Temperature,"?
Fia, 8. Molal humidity chart (high temperature range). To obtain pounds of
water per pound of dry air multiply molal humidity by 0.62.
may be used for estimation of the dew point of a mixture whose proper-
ties are represented by an estabhshed point on the chart. A line is pro-
jected through this point, parallel to the temperature axis, to the satura-
tion curve A. The abscissa of the intersection of this line with the
saturation curve is the dew point of the mixture.
9'
20 30 40 50 60 70 80 90 100 120
Temp«rature°F
uo
Pressure 1 atmosphere = 29.92 in, of mercury
fcii'
X • mole fraction of CO2 in dry gaa
FIG. 9. Molal humidity chart (low temperature range). To use either Fig. 8 or
9 in obtaining the pounds of water per pound of dry air multiply molal humidity,
ordinates of the charts, by 18/29 or 0.62.
101
The use of the chart is demonstrated in the following illustrations:

Illustration 8. Air at a temperature of lOO'F and atmospheric pressure has a
wet-bulb temperature of 85°F.
(a) Estimate the molal humidity, the percentage saturation, and the dew point of
this air.
(6) The air of part (a) is passed into an evaporator from which it emerges at a
temperature of 12(>°F with a wet-bulb temperature of 115.3°F. Estimate the
percentage saturation of the air leaving the evaporator and calculate the weight of
water evaporated per 1000 cu ft of entering air.
Solution: (a) The abscissa representing 100°F is located on Fig. 9 and followed
vertically to its intersection with the 85°F wet-bulb temperature line. This point
represents the initial conditions of the air. The percentage saturation is estimated
from the position of this point with respect to the curves corresponding to various
degrees of saturation. This value would be estimated to be about 52%. The molal
humidity is read from the scale of ordinates as 0.037. The dew point is determined
by following a horizontal line to its intersection with the saturation curve. The
abscissa of this point of intersection is 80.5°F, the dew point.
(b) In the manner described above the percentage saturation of the air leaving the
evaporator is estimated to be 84% and its molal humidity 0.110.
Basis: 1.0 lb-mole of moisture-free air.
Moles of wet air entering = 1.0 -|- 0.037 1.037 lb-moles
Volume of wet air entering = 1.037 X 359 X — 424 cu ft
492
Water evaporated = 0.110 - 0.037 = 0.073 lb-mole or .. 1.31 lb
Water evaporated per 1000 cu ft of entering wet air 1.31 X
1000
—r- 3.1 lb
424
Illustration 9. A combustion gas has the following composition:
CO2 12.1%
CO 0.1%
Oj 7.6%
Ns 80.2%
100.0%
(a) Estimate the wet-bulb temperature of this gas when moistiire-free at a temper-
ature of 200°F and atmospheric pressure.
(b) If the combustion gas has a dry-bulb temperature of 140°F and a wet-bulb
temperature of 9S°F estimate its molal humidity.
Solution: (a) From the group of x-curves corresponding to a wet-bulb temperature
of 100°F, the angle between the curves for x = 0 and x = 0.12 is estimated. A Une
is established at this angle to the 85° wet-bulb line, which is closest to the point repre-
senting the conditions of the gas. A line parallel to this newly established 85° wet-
bulb line, which corresponds to a composition x = 0.12, is projected from the point
representing zero humidity and a dry-bulb temperature of 200°F. This line inter-
sects the saturation curve at a temperature of 87°F, which is the wet-bulb temperature.
These projections may be carried out with the aid of two draftsman's triangles.
(6) In the manner described above, the angle between the curve x = 0 and
X = 0.12 at a wet-bulb temperature of 100° is estimated. A line is established at this
CHAP. IV] THE HUMIDITY CHART 103
angle to the 95° wet-bulb line, determining the 95° wet-bulb line corresponding to
I = 0.12. The intersection of this line with the 140° dry-bulb temperature line
establishes the point representing the conditions of the mixture. The ordinate of
this point is 0,040, the molal humidity of the mixture.
25 26 27 28 29 30
Absolute Pressure, Inches of Mercury
FIG. 9a. Pressure correction to humidity charts.
If many calculations are to be made for a gas having an approximately

constant value of x, it will be convenient to rule wet-bulb lines corre-
sponding to this composition directly onto the chart.
Humidity charts, Figs. 8 and 9, are strictly apphcable only to a baro-
metric pressure of 29.92 in. Hg. Where wet-bulb temperature read-
ings are taken at pressures other than atmospheric a correction should
be applied to the chart readings to obtain the true humidity. In
Fig. 9a the humidity correction to be added to the readings on the at-
mospheric pressure charts is plotted against total pressure for different
wet-bulb readings. It will be observed that the correction increases
with departure from atmospheric pressure and with increase in the
wet-bulb temperature, v.^, • . . . - - .
Illustration 9a. At 26.42 in. Hg the dry-bulb temperature of air is 150°F and
its wet-bulb temperature is 120°F. Obtain the correct humidity from the charts.
From Fig. 8, molal humidity =0.116
From Fig. 9a, molal humidity correction = .020 ' ' i "
Corrected molal humidity / • .136
The error in neglecting the pressure effect is 14.7%.
Adiabatic Vaporization. An adiabatic system or process is one which

is completely prevented from either gaining heat from, or losing heat to,
its surroundings. Frequently the vaporization of a hquid into a gas will
be of this type. As previously explained, the temperature of the vapor-
izing liquid will adjust itself to the wet-bulb value which is determined
by the vapor content and temperature of the gas. As vaporization pro-
ceeds the vapor content of the gas is increased. However, it has been
experimentally demonstrated that the wet-bulb temperature remains
unchanged throughout such an adiabatic vaporization. The heat which
is required for the vaporization of the liquid must be derived from the
gas, reducing its temperature. If such a process is continued until the
gas becomes saturated, the dry-bulb temperature of the gas will equal
its wet-bulb temperature, which will be the same as its initial wet-bulb
temperature. On the humidity chart the changes in humidity and dry-
bulb temperature which take place in an adiabatic vaporization process
are represented by a line of constant wet-bulb temperature. These wet-
bulb temperature lines are also termed adiabatic vaporization or adia-
batic cooling lines.
For example, from Fig. 8 it is seen that dry air at a temperature of
197°F has a wet-bulb temperature of 85°F. Suppose that dry air at
this temperature is brought in contact, in an adiabatic compartment,
with water at a temperature of 85°F. As vaporization takes place the
molal humidity of the air will increase, but if its wet-bulb temperature
is to remain constant the dry-bulb temperature must correspondingly
decrease along the 85°F wet-bulb temperature line. Thus, if vaporiza-
tion continues until the molal humidity of the air becomes 0.020 the dry-
bulb temperature of the air will be reduced to 144°F. If the vaporiza-
tion is continued until the air is saturated, the molal humidity will then
be 0.042 and the dry-bulb temperature will be 85°F. This molal humid-
ity of 0.042 represents the maximum quantity of water which can be
adiabatically evaporated into dry air at an initial temperature of 197°F.
Many types of industrial equipment for drying and evaporation are
practically adiabatic in operation, the heat for the vaporization being
almost entireb' abstracted from the hot gases which enter the process.
The humidity chart permits rapid calculation of the quantities of water
CHAP. IV] PROBLEMS 105
which can be evaporated in such processes and of the temperatures and

humidities throughout.
Illustration 10. Air enters a dryer at atmospheric pressure, a dry-bulb tempera-
ture of 190°F, and a wet-bulb temperature of 90°F. It is found that 0.028 lb-mole of
water is evaporated in the dryer per pound-mole of dry air entering it. Assuming that
the vaporization is adiabatic, estimate the dry-bulb temperature, the wet-bulb tem-
perature, and the percentage saturation of the air leaving the dryer.
Solution: On Fig. 8 it is found that a dry-bulb temperature of 190° and a wet-
bulb temperature of 90° correspond to a molal humidity of 0.011.
Molal humidity of air leaving = 0.011 + 0.028 = 0.039.
If the process is adiabatic the wet-bulb temperature will remain constant at 90°F.
A wet-bulb temperature of 90° and a molal humidity of 0.039 correspond to a dry-
bulb temperature of 116°F, the temperature of the air leaving the dryer. The per-
centage saturation is estimated as 35%.
Illustration 11. Carbon dioxide is saturated with water vapor by passing it
through a wetted chamber. The gas enters the chamber dry, at atmospheric pres-
sure, and at a temperature of 120°F. It may be assumed that the vaporization in
the saturator is adiabatic. Estimate the temperature and molal humidity of the
saturated carbon dioxide leaving the chamber.
Solution: For pure carbon dioxide, wet-bulb curves corresponding to i = 1.0
must be used. In the group of 100° wet-bulb curves the angle is determined between
the curves x = 0 and x = 1.0. A line is estaWished at this angle to the 65° wet-bulb
line, thus determining the 65° wet-bulb line for x = 1.0. A line parallel to this is
projected from the point representing a dry-bulb temperature of 120° and zero humid-
ity to the saturation curve. The intersection is at a temperature of 71°F, the wet-
bulb temperature. The molal humidity at this point is 0.027.
PROBLEMS
Problems 1-19 inclusive should be solved without use of the humidity charts.
1. (a) Calculate the composition, by volumeand by weight, of air which issaturated
with water vapor atapressure of 750mm of Hg and atemperature of 70°F.
(6) Calculate the composition by volumeand by weight of carbon dioxide which
is saturated with water vapor at the conditions of part a. (The necessary
data may be obtained from Table I.)
2. Nitrogen is saturated with benzene vapor at a temperature of 25°C and a pres-
sure of 750 mm of Hg. Calculate the composition of the mixture, expressed in the
following terms:
(a) Percentage by volume.
(b) Percentage by weight.
(c) Grains of benzene per cubic foot of mixture. **
(d) Pounds of benzene per pound of nitrogen.
(e) Pound-moles of benzene per pound-mole of nitrogen.
3. Carbon dioxide contains 0.053 pound-mole of water vapor per pound-mole of dry
CO2 at a temperature of 35°C and a total pressure of 750 mm of Hg.
(a) Calculate the relative saturation of the mixture. ..i,^.
(6) Calculate the percentage saturation of the mixture.
(c) Calculate the temperature to which the mixture must be heated in order that
the relative saturation shall be 30%. „ , , ., ,,. ,,.. , , ,
4. A mixture of benzene and air at a temperature of 24°C and a pressure of 745 mm

of Hg is found to have a dew point of 11 °C. ,
(a) Calculate the percentage by volume of benzene.
(6) Calculate the moles of benzene per mole of air.
(c) Calculate the weight of benzene per unit weight of air.
6. An industrial heating gas is metered at a temperature of 69°F and a pressure
of 752 mm of Hg. The relative humidity in the meter is found to be 47%.
The gas has a heating value of 500 Btu per cu ft measured at 60°F under a pres-
sure of 30 in. of Hg and saturated with water vapor. Calculate the heating value per
cubic foot measured in the above meter.
6. The heating value of an illuminating gas is best determined by means of the
continuous-flow calorimeter. In such a determination a gas sample is burned which
has a volume of 0.2 cu ft. This sample is measured, when saturated with water
vapor, at a total pressure of 29.42 in. of Hg and a temperature of 78°F. What vol-
ume would be occupied by the same quantity of moisture-free gas as contained in
tliis sample, were it at the gas-testers' standard conditions of 60°F, a pressure of 30.0
in. of Hg, and saturated with water vapor?
7. It is desired to construct a dryer for removing 100 lb of water per hour. Air
is supplied to the drying chamber at a temperature of 66°C, a pressure of 760 mm
of Hg, and a dew point of 4.5°C. If the air leaves the dryer at a temperature of 35°C,
a pressure of 755 mm of Hg, and a dew-point of 24°C, calculate the volume of air,
at the initial conditions, which must be supplied per hour.
8. Air, at a temperature of BO^C, a pressure of 745 mm of Hg, and a percentage
humidity of 10%, is suppUed to a dryer at a rate of 50,000 cu ft per hour. Water is
evaporated in the dryer at a rate of 60 lb per hour. The air leaves at a temperature
of 35°C, and a pressure of 742 mm of Hg.
(a) Calculate the percentage humidity of the air leaving the dryer.
(6) Calculate the volume of wet air leaving the dryer per hour.
9. For the best hygienic conditions, air in a living room should be at a tempera-
ture of 70°F and a relative humidity of 62%. It is also desirable that fresh air be
admitted at such a rate that the air is completely renewed twice each hour. This
requires that air be introduced at a volume rate, measured at the conditions of the
room, equal to twice the volume of the room.
(o) Calculate the dew point of the air in the room.
(6) If air is taken in from the outside at a temperature of 10°F and saturated,
calculate the weight of water which must be evaporated, per hour, in order
to maintain the above specified conditions in a room having a volume of
3000 cu ft. (Vapor pressure of water at 10°F, over ice = 1.6 mm of Hg;
barometric pressure = 743 mm of Hg.)
10. Illuminating gks at a temperature of 90°P and a pressure of 760 mm of Hg
enters a gas holder carrying 14 grains of water vapor per cubic foot. If, in the holder,
the gas is cooled to 35°F, calculate the weight of water condensed per 1000 cu ft of
gas entering the holder. The prefsure in the holder remains constant.
11. A gas mixture at a temperature of 27°C and a pressure of 750 mm of Hg con-
tains carbon disulfide vapor such that the percentage saturation is 70%. Calcu-
late the temperature to which the gas must be cooled, at constant pressure, in order
to condense 40% of the CS2 present.
12. A compressed-air tank having a volume of 16 cu ft is filled with air at a gauge
pressure of 150 lb per sq in. and a temperature of 85°F and saturated with water
CHAP. IV] PROBLEMS 107
vapor. The tank is filled by compressing atmospheric air at a pressure of 14.5 lb

per sq in., a temperature of 75°F, and a percentage humidity of 60%. Calculate
the amount of water vapor condensed in compressing enough air to fill the tank,
assuming that it originally contained air at atmospheric conditions.
13. Air at a temperature of 30°C and a pressure of 750 mm of Hg has a percentage
humidity of 60. Calculate the pressure to which this air must be compressed, at
constant temperature, in order to remove 90% of the water present.
14. In a process in which benzene is used as a solvent it is evaporated into dry
nitrogen. The resulting mixture at a temperature of 24°C and a pressure of 14.7
lb per sq in. has a percentage saturation of 60. It is desired to condense 80% of
the benzene present by a cooling and compressing process. If the temperature is
reduced to 10°C to what pressure must the gas be compressed?
15. Acetone is used as a solvent in a certain process. Recovery of the acetone is
accomplished by evaporation in a stream of nitrogen, followed by cooling and com-
pression of the gas-vapor mixture. In the solvent recovery unit, 50 lb of acetone are
to be removed per hour. The nitrogen is admitted at a temperature of 100°F and
750 mm of mercury pressure, and the partial pressure of the acetone in the incoming
nitrogen is 10.0 mm of mercury. The nitrogen leaves at 85°F, 740 mm of mercury,
and a percentage saturation of 85%.
(o) How many cubic feet of the incoming gas-vapor mixture must be admitted
per hour to obtain the required rate of evaporation of the acetone?
(6) How many cubic feet of the gas-vapor mixture leave the solvent recovery
unit per hour?
16. The gas-vapor mixture in Problem 15 leaves the evaporator at 85°F, 740 mm
of mercury, and a percentage saturation of 85%. The mixture is compressed and
cooled to 32°F, after which the condensate is removed. The nitrogen, with the
residual acetone vapor, is expanded to 750 mm of mercury pressure, heated to 100°F,
and re-used in the solvent recovery system. The partial pressure of the acetone in
the gas admitted to the solvent recovery system is 10 mm of mercury.
(o) To what pressure must the mixture be compressed at 32°F?
(6) On the basis of 1 cu ft of the gas-vapor mixture emerging from the solvent
recovery unit, calculate:
(1) The volume of the gas-vapor mixture at 32°F after compression.
(2) The volume of the gas-vapor mixture entering the solvent recovery
unit.
(3) The acetone condensed after compressing and cooling to 32°F.
17. A continuous dryer is operated under such conditions that 250 lb of water are
removed per hour from the stock being dried. The air enters the dryer at 175°F,
and a pressure of 765 mm of mercury. The dew point of the air is 40°F. The air
emerges from the dryer at 95°F, a pressure of 755 mm of mercury, and at 90% relative
humidity.
(a) How many cubic feet of the original air must be supplied per hoiir?
(6) How many cubic feet of air emerge from the dryer per hour?
18. Air at 25°C, 740 mm of mercury, and 55% relative humidity is compressed
and cooled to 0°C.
(o) What pressure must be applied if 90% of the water is to be condensed?
(6) On the basis of 1 cu ft of original air, what will the volume be of the gas-
residual vapor at 0°G in the compressed state?
19. Air at 25°C, 740 mm of mercury, and 65% relative humidity is compressed to
10 atm.
(a) To what temperature must the gas-vapor mixture be cooled if 90% of the
water is to be condensed?
(6) On the basis of 1 cu ft of original air, what will be the volume of the gas-
vapor mixture at 10 atmospheres after cooling to the final temperature?
In working out the following problems, the humidity charts in Figures 8 and 9 are to he
used as jar as possible. The use of the large-scale chart in Figure 9 is preferable within
its range. Unless otherwise specified the pressure is assumed to be one atmosphere.
20. Air at atmospheric pressure has a wet-bulb temperature of 62°F and a dry-bulb
temperature of 78°F.
(o) Estimate the percentage saturation, molal humidity, and the dew point.
(b) Calculate the weight of water contained in 100 cu ft of the air.
21. Hydrogen is saturated with water vapor at atmospheric pressure and a tem-
perature of 90°F. The wet gas is passed through a cooler in which its temperature is
reduced to 45°F and a part of the water vapor condensed. The gas after leaving the
cooler is heated to a temperature of 70°F.
(a) Estimate the weight of water condensed in the cooler per pound of moisture-
free hydrogen.
(6) Estimate the percentage humidity, wet-bulb temperature, and molal
humidity of the gas in its final conditions.
22. It is desired to maintain the air entering a building at a constant temperature
of 75°F and a percentage humidity of 40%. This is accomplished by passing the air
through a series of water sprays in which it is cooled and saturated with water.
The air leaving the spray chamber is then heated to 75°F.
(a) Assume that the water and air leave the spray chamber at the same tem-
perature. What is the temperature of the water leaving?
(6) Estimate the water content of the air in the building in pounds of water per
pound of dry air.
(c) If the air entering the spray chamber has a temperature of 90°F and a per-
centage humidity of 65%, how much water will be evaporated or con-
densed in the spray chamber per pound of moisture-free air?
23. In a direct-fired evaporator the hot combustion gases are passed over the sur-
face of the evaporating liquid. The gases leaving the evaporator have a dry-bulb
temperature of 190°F and a wet-bulb temperature of 145°F. The dry combustion
gases contain 11% COj. Estimate the percentage saturation of the gases leaving the
evaporator, their molal humidity, and their dew point.
24. An air-conditioning system takes warm summer air at 95°F, 85% humidity,
and 760 mm of mercury. This air is passed through a cold-water spray, and emerges
saturated. The air is then heated to 70°F.
(a) If it is required that the final percentage humidity be 35%, what must be
the temperature of the air emerging from the spray?
(6) On the basis of 1000 cu ft of entering air, calculate the weight of water
condensed by the spray.
26. Cold winter air at 20°F, 760 mm pressure, and 70% humidity is conditioned by
passing through a bank of steam-heated coils, through a water spray, and finally
through a second set of steam-heated coils. In passing through the first bank of
CHAP. IV] ' PROBLEMS 109
steam-heated coils, the air is heated to 75°F. The water supphed to the spray-
chamber is adjusted to the wet-bulb temperature of the air admitted to the chamber,
hence the humidifying unit may be assumed to operate adiabatically. It is required
that the air emerging from the conditioning unit be at 70°F and 35 % humidity.
(o) What should be the temperature of the water supplied to the spray chamber?
(b) In order to secure air at the required final conditions, what must be the
percentage humidity of the air emerging from the spray chamber?
(c) What is the dry-bulb temperature of the air emerging from the spray
chamber?
(d) On the basis of 1 cu ft of outside air, calculate the volume at each step of the
process,
(e) Calculate the pounds of water evaporated per cubic foot of original air.
26. At the top of a chimney the flue gas from a gas-fired furnace has a temperature
of 180°F and a wet-bulb temperature of ISST. The composition of the moisture-
free flue gas is as follows:
CO2 14.1%
O2 6.0%
•1 N, 79.9%
X : 100.0%
(a) If the temperature of the gases in the stack were reduced to 90°P, calculate
the weight of water in pounds which would be condensed per pound-mole
of dry gas.
(fe) The gas burned in the furnace is estimated to contain 7% CH4, 27% CO,
and 3% CO2 by volume; it does not contain any carbon compounds other
than these. The gas is burned at a rate of 4000 cu ft per 24 hours meas-
ured at a temperature of 65°F, saturated with water vapor. Estimate the
weight of water condensed in the chimney per day under the conditions
of part (a).
27. A fuel gas has the following analysis: <
Carbon dioxide CO2 = 3.2%
Ethylene C2H4 = 6 . 1
Benzene CsHs = 1 . 0 ^. ^
Oxygen O2 = 0.8 ', ;
Hydrogen H2 • = 40.2
Carbon monoxide.. CO =33.1 |
Paraffins C1.20H4.40 = 10.2
Nitrogen Nj = 5 . 4
100,0%
Assume that this gas is metered at 65°F, 100% saturation, and is burned with air
supplied at 70°F, 1 atm pressure, and 60% humidity. Assume 20% excess air is
supplied and that combustion is complete.
(a) What is the dew point of the stack gases?
(b) If the gases leave the stack at 190°F, what is the wet-btiib temperature?
(c) If the gases are cooled to 85°F before leaving the chimney, how many
pounds of condensate drain down the chimney on the basis of 100 cu ft
of fuel gas metered?
no HUMIDITY AND SATURATION [CHAP. IV
28. Air at 185°F and a dew point of 40°F is supplied to a dryer, which operates
under adiabatic conditions.
(o) What is the minimum temperature to which the air wiU cool in the dryer?
(6) What is the maximum evaporation in pounds that can be obtained on the
basis of 1000 cu ft of entering air?
(c) If maximum evaporation is obtained, what will be the volume of the emerg-
ing air on the basis of 1000 cu ft of entering air?
29. Air enters a dryer at a temperatiire of 240°F with a dew point of 55°F. The
dryer may be assumed to approach adiabatic vaporization in its operation, all the
heat being supplied by the air. If the air leaves the dryer saturated with water
vapor, how much water can be evaporated per 1000 cu ft of entering wet air?
30. A dryer for the drying of textiles may be assumed to operate adiabatically.
The air enters the dryer at a temperature of 160°F with a dew point of 68°F. It is
found that 0.82 lb of water is evaporated per 1000 cu ft of wet air entering the dryer.
Estimate the percentage saturation and temperature of the air leaving the dryer.
CHAPTER V
SOLUBILITY AND SORPTION
Industrial problems involving phase equilibria among solids, liquids,

and gases involve all the mutual equilibrium relations among these
three phases, such as the dissolution of a solid by a hquid, the crystal-
lization of a solid from a liquid, the absorption and desorption of gases
by liquids, the distribution of a solute between immiscible liquids, and
the adsorption and desorption of gases by solids.
j : , •'
DISSOLUTION AND CRYSTALLIZATION

A substance which is a solvent for a solid has an entirely specific effect
upon the distribution of particles between the solid and its dispersed
state. Thus a substance which is an excellent solvent for one sohd
may exert no appreciable influence on another. The solvent action of
a liquid is presumed to result from a high affinity or attractive force
between the liquid and the soUd particles. When a solvent and a solid
are brought into contact with each other, the attractive forces of the
Uquid aided by the thermal motion of the solid particles tend to break
apart the structure of the sohd and to disperse molecules or ions from
its surface in a manner somewhat analogous to the vaporization of a
liquid into a gas. As a result the ions or molecules enter the liquid as
individually mobile units, thus forming what is termed a solution of
the solid in the liquid. The dispersed material in a solution is termed
the solute. This distinction between the-solvent and solute in a solu-
tion is entirely arbitrary since either component may be considered as
either solute or solvent. For example, a solution of naphthalene in
benzene may equally well be considered as a solution of benzene in
naphthalene.
The solute particles of a solution are free to move about as a result of
the kinetic energy of translation which is possessed by each. By bom-
barding the confining walls these particles exert an osmotic pressure
which is entirely analogous to the partial pressure exerted by each
component of a gaseous mixture. Thus, when a solution is in contact
with the solid from which it was formed, there will be a continual
return of dissolved particles to the solid surface. The dispersion of a
solid into a liquid is termed dissolution. The reverse process is known
111
112 SOLUBILITY AND SORPTION [CHAP. V
as crystallization. As dissolution proceeds, the concentration of dis-

persed particles is increased, resulting in an increased osmotic pressure
and a greater rate of crystallization. When the solute concentration
becomes sufficiently high, the rate of crystallization will equal the rate
of dissolution, and a dynamic equilibrium will be established, the con-
centration of the solution then remaining constant. This equilibrium
is analogous to that between a liquid and its vapor. Under these condi-
tions the solution is said to be saturated with the solute and is incapable
of dissolving greater quantities of that particular solute.
The concentration of solute in a saturated solution is termed the solu-
bility of the solute in the solvent. Solubilities are dependent on the
nature of the solute, the nature of the solvent, and the existing tempera-
ture. Pressure also has a small effect on solubility, which ordinarily
may be neglected. This effect is such that, in a solution whose volume
is less than the sum of the volumes of its components in their pure states,
solubility is increased by an increase in pressure. In the opposite case,
in which the total volume is increased in the formation of a solution,
an increase in pressure lowers the solubility.
The simple generalizations which apply to equilibria between liquids
and vapor-gas mixtures are not applicable to liquid solutions. Little
is known regarding the relationships between solubilities and the specific
properties of the solute and solvent. The properties of each.particular
system must be individually determined by experimental means, and it
is impossible to predict quantitatively the behavior of one system from
that of another. This results from the fact that what may be termed
the solution pressure of a solid is entirely dependent on the nature of the
solvent with which it is in contact. In many cases it is impossible to
predict, even qualitatively, the effect of temperature on solubility.
Solubilities of Solids Which Do Not Form Compounds with the Sol-
vent. In general, where no true chemical compounds are formed be-
tween a solute and solvent, the solubility increases with increasing
temperature. The solubility curves of such systems resemble vapor-
pressure curves in general form. In Fig. 10 are plotted the solubility
data of a typical system, naphthalene in benzene.
Curve EB of Fig. 10 represents the conditions of temperature and
concentration which correspond to saturation with naphthalene, of a
solution of naphthalene in benzene. A solution whose concentration
and temperature fix a point on this curve will remain in dynamic equilib-
rium with solid crystals of naphthalene. If the temperature is lowered
or the concentration increased by the removal of solvent, naphthalene
crystals will be formed in the solution. The area to the right of curve
EB represents conditions of homogeneous, partially saturated solutions.
CHAP. V] SOLUBILITIES OF SOLIDS 113
The area between curves EB and ED represents conditions of non-

homogeneous mixtures of crystals of pure naphthalene in solutions of
naphthalene in benzene. If a solution whose concentration and temper-
ature are represented by point xi is cooled without change in compo-
sition, its conditions will vary along the dotted line parallel to the
temperature axis. The temperature at which this dotted line inter-
sects the solubility curve EB is the temperature at which pure naph-
thalene will begin to crystallize from the solution. As the temperature
is reduced further, more naphthalene will crystallize and the remaining
saturated solution will diminish in concentration, its condition always
100
D
80
Solid CloHg
+Soli ition y ^
•3 (S
-to ^-
*i
U °
3 m
40
S
Home geneo us
So ution
(Solid C j H
^ '1'/ L + Solution
C \ A
-10 20 30 40 60
Temperature °C
FIG. 10. Solubility of naphthalene in benzene.
represented by a point on curve EB. When the temperature cor-

responding to line CED is reached the system will consist of pure
naphthalene crystals and a saturated solution whose concentration
corresponds to point E. Further decrease in temperature will cause
this remaining solution to solidify to a mixture of crystals of pure
benzene and pure naphthalene. The line CED represents the comple-
tion of solidification, and the area to the left of it corresponds to com-
pletely solid systems of naphthalene and benzene crystals.
The point E is termed the eutectic point, and the corresponding tem-
perature and composition are, respectively, the eutectic temperature and
the eutectic composition. If a solution of eutectic composition is cooled
it will undergo no change until reaching the eutectic temperature, when
it will completely solidify without further change in temperature. A
solution of eutectic composition solidifies completely at one definite

temperature which is also the lowest solidification point possible for
the system.
The curve EA represents the conditions of temperature and composi-
tion which correspond to saturation with benzene of a solution of
naphthalene in benzene. Whereas curve EB represents the solubility
of naphthalene in benzene, curve EA represents the solubihty of benzene
in naphthalene. The area AEC represents conditions of nonhomoge-
neous mixtures of benzene crystals in saturated solutions of naphtha-
lene in benzene. If a solution whose composition and temperature are
represented by point x^ is cooled, its conditions will vary along a hori-
zontal through X2. At the temperature of the intersection of this line
with curve EA, crystals of pure benzene will be formed. As the tem-
perature is further reduced, more pure benzene will crystallize, and the
remaining saturated solution will vary in composition along curve EA
towards E. When the eutectic temperature is reached, the remaining
saturated solution will be of eutectic composition, represented by point
E. Further decrease in temperature will cause complete solidification
into a mixture of benzene and naphthalene crystals.
The curve EA is frequently termed the freezing-point curve of the
solution representing the temperatures at which solvent begins to freeze
out. The point A represents the freezing point of the pure solvent,
benzene. From the same viewpoint the curve EB might be considered
as a freezing-point curve along which solute begins to freeze out. Thus,
either curve may be considered as a solubihty curve or a freezing-point
curve.
The percentage saturation of a solution may be defined as the percentage
ratio of the existing weight of solute per unit weight of solvent to the
weight of solute which would exist per unit weight of solvent if the solu-
tion were saturated at the existing temperature. The percentage sat-
uration of a solution may be varied by changing either its temperature
or composition. The effects of such changes on the percentage satur-
ation of a solution may be predicted by locating the point representing
the conditions of the solution on a solubility chart such as Fig. 10. A
change in temperature will move this point along a line parallel to the
temperature axis. A change in composition will move it along a line
parallel to the concentration axis. A process in which both composition
and temperature are changed simultaneously is best considered as pro-
ceeding in two steps: a change in composition at constant temperature
and a change in temperature with constant composition.
Solubilities of Solids Which Form Solvates with Congruent Points.
Many solutes possess the property of forming definite chemical com-
CHAP. V] SOLVATES WITH CONGRUENT POINTS 115
pounds with their solvents. Such compounds of definite proportions

between solutes and solvents are termed solvates, or if the solvent is
water, hydrates. Several solvates of different compositions may be
formed by a single system, each a stable form, under certain conditions
of temperature and composition. The presence of such solvates greatly
complicates the solubility relationships of the system.
, 100 1 1 1 1 1 1 1 1 ' 1 1 1 10'

Solid FeClaC?) — •v^^
+ Solution
Solid Mixture FeCla
»90 +FeCl3-2H20 "
-', SolidFeCl3-2H20
+ Solution \
K L-
"s g
• ' • ; • ' , • , • , • , • ^ . •
8 V - > ^ ^ r
Solid Mixture FeClj- 2.5HjO SolidFeCl3-2.5H20
+ FeCl3- 2H2O 6 + Solution Y-^ HjJ. -
Solid Mixture FeClj- 3.5 H2O 'y /If
o 'S + FeCl3-2.5H20 F
4 5J)^--^Solid Fe (Jlj • 3.5 tl2U
70 p4—13 + Solution ~
Solid Mixture FeCla-eHjO 5
FeCl3-3.5H20
1\
> 60
60 2 3 j / c
^ f ~ S o l i d FeCla-eHjO
--Ice+FeCla-eHp 1 + Solution
c
0)
4_
f^ 4 0 1 ''
'S B__.—'— 1
a)
- 1
1
•S,20 \
0)
1
Ice+Solution ^v 1
1
0 1 1 1 1 1 1 A , 1 1 : , 1 ,
-60 -40 -20 0 20 40 60 80
Temperature "C
FIG. 11. Solubility of ferric chloride in water.
The system ferric chloride and water is an excellent illustration of the

effects of the formation of many hydrates. In Fig. 11 are plotted the
solubility curves of this system. Point A represents the freezing point
of the pure solvent, water, and curve AB the conditions of equilibrium
in a solution which is saturated with the solid solvent, ice. This curve
is analogous to curve AE of Fig. 10 and represents the solubility of
water in ferric chloride. The area ABl represents nonhomogeneous
mixtures of pure ice in equilibrium with saturated solutions.
Point B of Fig. 11 is a eutectic point analogous to point E of Fig. 10.

Curve BC represents the conditions of saturation of a solution of ferric
chloride in water with the sohd hydrate, FeCl3-6H2G. If a solution,
whose conditions are represented by point Xi, is cooled, it will become
saturated with FeCl3-6H20 when conditions corresponding to this
concentration on curve BC are reached. Further cooling will result in
separation of crystals of FeCl3-6H20 which will be in equilibrium with
saturated solutions whose compositions correspond to the ordinates of
curve BC at the existing temperatures. If cooling is continued, more
FeCl3-6H20 crystals will be formed and the concentration of the remaia-
ing solution diminished along curve BC until it reaches a value of B,
corresponding to the eutectic temperature. On further cooling, the re-
maining solution will solidify, without change in temperature, into a
mixture of crystals of pure ice and pure FeCls-61120. The area BC2 Wi
represents nonhomogeneous mixtures of pure FeCl3-6H20 crystals and ^-
saturated solutions. The area to the left of curve 1B2 represents mix-
tures of crystals of pure ice and pure FeCl3-6H20, which are not solu-
ble in each other in the solid state.
As the concentration and temperature of a saturated solution are in-
creased along curve BC, a concentration corresponding to point C is
reached, which is the composition of the pure hydrate, FeCl3-6H20. At
this point the entire system is solid hydrate in equilibrium with solution
of the same composition. If a solution of this composition is cooled,
when its temperature reaches that of point C it will completely solidify
into FeCl3-6H20 without change of temperature. This behavior is
analogous to the freezing of a pure compound, and the temperature of
point C represents the melting point of FeCl3-6H20. A point such as
C, at which a hydrate is in equilibrium with a saturated solution of
the same composition, is termed a congruent 'point. The curves ABC21
may be considered as representing the complete solubility-freezing-
point relationship of a system in which water is the solvent and
FeCl3-6H20 the solute. Point A is the melting point of the pure sol-
vent and C that of the pure solute; B represents the eutectic point of
this system.
From the negative slope of curve CD it follows that at concentra-
tions higher than that of point C the concentration of FeCls in a satu-
rated aqueous solution may be increased by lowering the temperature.
This behavior may be readily understood by considering that the curves
SCDE4 represent the solubility-freezing-point data of a new system
in which the solvent is FeCl3-6H20 and the solute is FeCU-S.SHsO. In
this system curve CD is analogous to AB and represents the con-
ditions of equilibrium between pure, solid solvent, FeCls-eHaO, and a
CHAP. V] SOLVATES WITH CONGRUENT POINTS 117
solution of FeCl3-3.5H20 in FeCls-GHaO. If a solution at conditions

represented by point Xi, is cooled, pure FeCl3-6H20 will crystallize out
when a temperature corresponding to this concentration on curve CD
is reached. Further cooling will result in the formation of more pure
solvent crystals, FeCl3-6H20, and the remaining saturated solution
will increase in concentration, along curve CD. When the temperature
of curve ^D3 is reached this remaining solution will have the composition
of point D. Further cooling will result in complete solidification, with-
out further change in temperature, into a mixture of crystals of pure
FeCl3-6H20 and pure FeCl3-3.5H20. Point D represents the eutectic
point of this system and is analogous to point B. The area CDS
represents conditions of nonhomogeneous mixtures of pure crystals of
FeCl3-6H20 and saturated solutions of FeCl3-3.5H20 in FeCU-eHaO.
Curve DE represents the conditions of equilibrium between crystals
of pure solute, FeCl3-3.5H20, and a saturated solution of FeCl3-3.5H20
in FeCl3-6H20. As the concentration and temperature of a saturated
solution are increased along curve DE, a concentration corresponding to
point E is reached which is the composition of the hydrate FeCla-S.SHaO.
At this point the entire system must be solid FeCl3-3.5H20 in equi-
librium with solution of the same concentration. Point E is, therefore,
a second congruent point, representing the melting point of FeCl3-3.5H20.
Area DEl^. represents conditions of nonhomogeneous mixtures of crystals
of pure FeCl3-3.5H20 and saturated solutions. The area to the left of
curve SDlf. represents entirely solid systems composed of mixtures of
crystals of pure FeCU-GHaO and FeCr3-3.5H20.
By similar reasoning, the curves 5EFG6 may be considered as rep-
resenting the solubility-freezing-point data of a system in which the
solute is FeCl3-2.5H20 and the solvent FeCls-S.SHjO. Curves 7GHJ8
represent the system of FeCl3-2H20, solute, and FeCl3-2.5H20, solvent.
Curves 9JKL represent the system of FeCU, solute, and FeCl3-2H20,
solvent. Points F, H, and K are the eutectic points of these systems.
Points G and / are the congruent melting points of FeCl3-2.5H20 and
FeCl3-2H20, respectively.
By means of a chart such as Fig. 11 the changes taking place in even
very compHcated systems may be readily predicted or explained. An
illustration of the peculiar phenomena which may take place in complex
systems is furnished by the isothermal concentration of a dilute aqueous
solution of ferric chloride along the line zy. Such concentration may be
carried out by evaporation at constant temperature. At the intersec-
tion of zy with BC, crystals of FeCl3'6H20 will begin to form. At line
C3 the system will be entirely sohd FeCl6-6H20. Further concentra-
tion results in the appearance of liquid solution, and at curve CD the
system will be once more entirely liquid. As concentration is continued

to curve DE, solidification again begins, becoming complete at line E5.
Further concentration causes the reappearance of liquid, and liquefac-
tion is completed at curve EF. At curve FO solidification begins and
becomes complete at G6. This surprising alternation of the liquid and
solid states while concentration is progressing may be easily demon-
strated.
Many other systems, both organic and inorganic, behave in a manner
similar to that of aqueous ferric chloride and form one or more solvates
and congruent points.
100 5
Solids - Na 2 SO4 + Na 2 SO4 • 10 Hj 0
Solid Na2S04 +Solution

i60
3 2
4
504--
10 HjO NajSOi-lOHjO + Solution E
+ ctic • -
.-^ jr
: —-
^C' î — ^ D
v\»o W^*i
^SMS°'-
^ t.
G - Homogeneous Solution
— Ice
_Eute ctici 3 .
ill «A~lce +Solution
20 30 40 50 60
Temperature °C
FIG. 12. Solubility of sodium sulfate in water.
Solubilities ^f Solids Which Form Solvates without Congruent

Points. In certain systems solvates are formed which are not stable and
which decompose before the temperature of a congruent point is
reached. Such solvates undergo direct transition from the solid state
into other chemical compoimds. These transitions take place at
sharply defined temperatures which are termed transition points. The
system of sodium sulfate and water illustrates this type of behavior. In
Fig. 12 are plotted the solubility and freezing-point data of this system.
Curve AB of Fig. 12 represents conditions of equihbrium between ice
and aqueous solutions of sodium sulfate. Point B is the eutectic point
of the system of Na2SO4-10H2O (solute) and water (solvent). Curve
BC represents the equihbrium between crystals of Na2SO4-10H2O and
saturated solution. As the temperature of the system is increased the
CHAP. V] EFFECT OF PARTICLE SIZE ON SOLUBILITY 119
concentration of the saturated solution increases along curve BC. Nor-

mally this increase would continue until a congruent point was reached.
However, at a temperature of 32.384°C, Na2SO4-10H2O becomes un-
stable and is decomposed into Na2S04 and water. The solubility of
the anhydrous Na2S04 as indicated by curve CD diminishes with
increasing temperature. Points on this curve represent conditions of
equilibrium between crystals of anhydrous Na2S04 and saturated so-
lutions. A solution whose conditions are represented by point Xi will
become saturated if either heated or cooled sufficiently. If cooled,
crystals of Na2SO4-10H2O will form when the conditions of curve BC
are reached. If heated, crystals of anhydrous Na2S04 will form at the
temperature corresponding to composition Xi on curve CD.
The significance of the areas of Fig. 12 is similar to that of the other
diagrams which have been discussed. The area of the small triangle to
the left of line AB represents a region of equilibrium between crystals
of pure ice and saturated solution. Area BC4^ is a region of equilibrium
between crystals of pure Na2SO4-10H2O and saturated solution. The
area to the left of line B2 represents a solid mixture of crystals of ice
and Na2SO4-10H2O. Line C4S indicates the transition temperature at
which the decahydrate decomposes to form the anhydrous salt. Line
S24 indicates the composition of the pure decahydrate, Na2SO4-10H2O.
The area to the right of curve 54CD represents conditions of equilibria
between crystals of pure Na2S04 and saturated solutions. The area
above and to the left of curves 3^4^ is a region of entirely solid mixtures
of Na2S04 and Na2SO4-10H2O. Line 45 therefore represents a tempera-
ture of complete solidification.
Many systems which form several solvates show solubility relation-
ships both with and without congruent points and transition points in
the same system. For example, zinc chloride forms five hydrates.
Four of these hydrates decompose, as does Na2SO4-10H2O, exhibiting
transition points before congruent concentrations are reached. The
fifth hydrate, ZnCl2-2.5H20, exhibits a true congruent point, as do the
hydrates of ferric chloride. The significance of such solubility relation-
ships may be understood from the principles discussed in the preceding
sections.
Effect of Particle Size on Solubility. The solution pressure and solu-
bility of a solid is affected by its particle size in a manner entirely anal-
ogous to the effect of particle size on vapor pressure. The solubility of
a substance is increased with increase in its degree of subdivision. This
increasing solubility with diminishing particle size is demonstrated by
the behavior of crystals which are in equihbrium with their saturated
solutions. Where such an equilibrium exists the total amounts of solid
and liquid must remain unchanged. However, the equilibrium is dy-

namic, resulting from equality between the rates of dissolution and
crystallization. As a result of the effect of particle size on solubihty,
the small crystals in such a solution will possess higher solution pressures
than the large ones and will tend to disappear with corresponding in-
crease in size of the large crystals. This growth of large crystals at the
expense of the small ones in a saturated solution is a familiar phe-
nomenon of considerable industrial importance. For the same reasons,
an irregular crystalline mass will change its shape in a saturated solu-
tion. The sharp comers and points will exert higher sol'ution pressures
than the plane surfaces and will disappear, building up the plane sur-
faces and tending to produce a regular shape. Like vapor pressures,
solubilities are not noticeably affected by particle size until submicro-
scopic dimensions are approached.
Supersaturation. Just as spontaneous condensation of a vapor is
made difficult because of the high vapor pressure of small drops, spon-
taneous crystallization is interfered with by the high solubihty of small
crystals. In order to produce spontaneous crystaUization the concen-
tration of a solution must be sufficiently high that the small crystalline
nuclei which are formed by simultaneous molecular or ionic impacts do
not immediately redissolve. Such a concentration will be much greater
than that which is in equihbrium with large crystals of the same sohd.
Once crystaUization is started and nuclei are formed it will continue and
the nuclei will grow until the normal equilibrium conditions are reached.
For these reasons it is relatively easy to obtain solutions whose con-
centrations are higher than the values normally corresponding to satu-
ration. Such solutions are supersaturated with respect to large crystals
but are only partially saturated with respect to the tiny nuclei which
tend to form in them. Supersaturated solutions may be formed by
careful exclusion of all crystalline particles of solute and by slow changes
in temperature or concentration without agitation.
Because of the phenomenon of supersaturation, it is possible to extend
the curves of solubility diagrams, such as Fig. 12 into regions where the
equilibria which they represent would not normally be stable. The
dotted curves of Fig. 12 represent such equilibria which have been
experimentally observed in supersaturated solutions of this system.
Such equiUbria are termed meiastable and possess the continual tendency
to revert to the normal, stable state corresponding to their conditions
of temperature and concentration.
The dotted curve GE of Fig. 12 represents the metastable equilibrium
between crystals of Na2S04-7H20 and its saturated solutions. .If a
solution at conditions Xi is carefully cooled, normal crystallization of
CHAP. V] DISSOLUTION 121
Na2SO4-10H2O may be prevented and a supersaturated solution pro-

duced at conditions x'\. If cooling is continued the crystallization of
Na2S04-7H20 may be induced at a temperature corresponding to com-
position X\ on curve GE. A supersaturated solution at conditions x'y is
capable of dissolving Na2S04-7H20 and is unsaturated with respect to
this compound. On the other hand, its conditions are unstable, and
any disturbance such as agitation, a sudden temperature change, or the
introduction of a crystal of Na2SO4-10H2O, will cause it to assume its
normal equilibrium conditions with the crystallization of Na2SO4-10H2O.
Metastable equilibria are of little industrial significance, but super-
saturation is very commonly encountered. It may be produced in two
ways. By the exclusion of all particles of solid solute the formation of
crystalline nuclei may be entirely prevented as described above. An-
other type of supersaturation may result from sudden changes of the
conditions of a saturated solution even though crystallization has
started and crystalline solute is present. This results from the fact
that crystallization, especially in certain types of viscous solutions, is a
slow process. Sudden cooling of such a solution will produce temporary
conditions of supersaturation simply because the system is slow in ad-
justing itself to its equilibrium conditions. Agitation of the solution
will hasten this adjustment.
In the majority of crystallization operations it is desirable to avoid
supersaturation. Supersaturation of the type resulting from the ab-
sence of crystalline nuclei is prevented by seeding saturated solutions
with crystals of solute. Spontaneous nucleus formation is also favored
by the presence of rough, adsorbent surfaces. The crystallization of
sugar on pieces of string to form rock candy and the scratching of the
wall of a beaker to cause crystallization of an analytical precipitate are
famiUar illustrations of this principle. Supersaturation due to slow
crystallization rates is avoided by using correspondingly slow rates of
change of the conditions which promote supersaturation and in some
cases by agitation.
DISSOLUTION
Problems arise in which it is required to calculate the amount of solute
which can be dissolved in a specified quantity of solvent or solution, or,
conversely, the quantity of solvent required to dissolve a given amount
of solute to produce a solution of specified degree of saturation. Where
solvates are not formed in the system such calculations introduce no
new difficulties. From the solubility data is determined the quantity
of solute which may be dissolved in a unit quantity of solvent to form
122 SOLUBILITY A N D S O R P T I O N [OHAP. V
a saturated solution at the existing temperature. The amount of solute

which may be dissolved in a solution is then the difference between the
amount already present and the amount which may be present if the
solution is saturated at the specified conditions, both quantities being
based on the same quantity of solvent. ,; .
Illustration 1. A solution of sodium chloride in water is saturated at a tempera-
ture of 15°C. Calculate the weight of NaCl which can be dissolved by 100 lb of
this solution if it is heated to a temperature of 65°C.
SolubiUty of NaCl at 15°C = 6.12 lb-moles per 1000 lb of H2O
SolubiUty of NaCl at 65°C = 6.37 lb-moles per 1000 lb of H2O
Basis: 1000 lb of water.

NaCl in saturated solution at I S ' C = 6.12 X 58.5 = 358 lb
Percentage of NaCl by weight = 358/1358 = 26.4%
NaCl in saturated solution a t 65°C = 6.37 X 58.5 = 373 lb
NaCl which may be dissolved per 1000 lb of H2O = 373 - 358 = 15 lb
Water present in 100 lb of original solution = 100 X (1.0 - 0.264) = 73.6 lb
15 X 73.6
NaCl dissolved'per 100 lb of original solution = — = . .. 1.1 lb
1000
It will be noted t h a t the solubility of NaCl changes but Uttle with change in tem-
perature.
Where the substance to be dissolved is a solvated compound the prob-

lem is complicated by the fact that both solute and solvent are added
to the solution. Such calculations are carried out by equating the total
quantities of solute entering and leaving the process. Algebraic ex-
pressions are formed for the sum of the quantity of solute to be dis-
solved plus that originally present in the solution, and for the quantity
of solute in the final solution. Since the total quantity of solute must
be constant, these two expressions are equal and may be equated and
solved. This method is demonstrated in the following illustration:
Illustration 2. After a crystaUization process a solution of calcium chloride in
water contains 62 lb of CaCU per 100 lb of water. Calculate the weight of this
solution necessary to dissolve 250 lb of CaCl2-6H20 at a temperature of 25°C. Sol-
ubility a t 26°C == 7.38 lb-moles of CaCl2 per 1000 lb of H2O.
Basis: x = weight of water in the required quantity of solution.
250
CaCl2-6H20 to be dissolved = — = 1.14 lb-moles
Total CaCl2 entering process = 1.14 -f- = 1.14 + ' „ - - l b - m o l e s

J-XiXlUU 100
Total water entering process = x + (1.14 X 6 X 18) = x + 123 lb
X -I- 123
Total CaClj leaving process = 7.38 lb-moles
1000
1 : •-
CHAP. V] CRYSTALLIZATION 123
Equating:
'" 1.14+ 0-^^^^ = 7.38^ + ^'^
100 1000
1140 + 5.59a: = 7.38s + 9 0 8
1.79a; = 232
x = 1301b
Total weigl
CRYSTALLIZATION
The crystallization of a solute from a solution may be brought about
in three different ways. The composition of the solution may be
changed by the removal of pure solvent, as by evaporation, until the
remaining solution becomes supersaturated and crystallization takes
place. The second method involves a change of temperature to pro-
duce conditions of lower solubility and consequent supersaturation and
crystallization. A third method by which crystallization may be pro-
duced is through a change in the nature of the system. For example,
inorganic salts may be caused to crystallize from aqueous solutions by
the addition of alcohol. Other industrial processes involve the salting
out of a solute by the addition of a more soluble material which possesses
an ion in common with ihe original solute. The calculations which are
involved in this third type of crystallization processes are frequently
very complicated and require a large number of data regarding the par-
ticular systems involved. Such systems involve more than two com-
ponents and require application of the principles of complex equilibria
which are discussed in a later section.
Where Solvates Are Not Formed. The most important crystalli-
zation processes of industry are those which combine the effect of in-
creasing the concentration by the removal of solvent with the effect of
change of temperature. Where crystallization is brought about only
through change in temperature, the yields of crystals and the necessary
conditions may be calculated on the basis of the quantity of solvent,
which remains constant throughout the process. From the solubility
data may be obtained the quantity of solute which will be dissolved
in this quantity of solvent in the saturated solution which will remain
after crystallization. The difference between the quantity of solute
originally present and that remaining in solution will be the quantity
of crystals formed. Such problems may be of two types: one in which
• it is desired to calculate the yield of crystals produced by a specified
temperature change; and the converse, in which the amount of tem-
perature change necessary to produce a specified yield is desired. The
percentage yield of a crystalhzation process is the percentage which
the yield of crystallized solute forms of the total quantity of solute
originally present. - ;•;% ^ " ' dlt ?} j
Illustration 3. A solution of sodium nitrate in water at a temperature of 40°C
contains 49% NaNOs by weight.
(o) Calculate the percentage saturation of this solution. '
(6) Calculate the weight of NaNOa which may be crystallized from 1000 lb of this
solution by reducing the temperature to 10°C.
(c) Calculate the percentage yield of the process.
J SolubiUty of NaNOs at 40°C = 61.4% by weight
Solubility of NaNOa at 10°C = 44.6% by weight
Basis: 1000 lb of original solution. ^
49 48.6
(o) Percentage saturation = — X 77-7 = 91.0%
61 61.4 * '
(6) Yield of NaNOs crystals = a; lb / ,„ f \ .
From a NaNOs balance
1000(0.49) = (1000 - a;) (0.445) +x
Hence x = 81 lb ' j
81 ' ''
(c) Percentage jdeld = — = 16.5% *' ./
490 /
Illustration 4. A solution of sodium bicarbonate in-water is saturated at 60°C.
Calculate the temperature to which this solution must be cooled in order to crystallize
40% of the NaHCOs. ., . . ' ^
Solubility of NaHCOs at CO^C = 1.96 lb-moles per 1000 lb H2O
Basis; 1000 lb of H2O. <
NaHCOs in original solution = 1.96 lb-moles
NaHCOs in final solution = 1.96 X 0.60 = 1.18 lb-moles
From the solubility data of NaHCOs it is found that a saturated solution con-
taining 1.18 lb-moles per 1000 lb of HjO has a temperature of 23°C. The solution
must be cooled to this temperature to produce the specified percentage yield. /
Calculations of the yields and necessary conditions of crystallization

by concentration may be carried out by consideration of the quantity
of solvent remaining after concentration has taken place. The quantity
of solute which will be dissolved in this quantity of solvent in the satu-
rated solution remaining after crystallization may be calculated from
solubility data. The quantity of crystals formed in the process will be
the difference between the quantity of solute originally present and that
finally remaining in solution. If the concentration is accompanied or
followed by a temperature change the problem is unchanged. It is
only necessary to consider the final temperature in order to determine
the quantity of solute remaining in solution. In such processes three
variable factors are present: the yield, the temperature change, and the
CHAP. V] CRYSTALLIZATION 125
degree of concentration. Problems arise in which it is necessary to
evaluate any one of these factors if the other two are specified.
Ulustration 5. A solution of potassium dichromate in water contains 13% KjCrjO?
by weight. From 1000 lb of this solution are evaporated 640 lb of water. The
remaining solution is cooled to 20°C. Calculate the amount and the percentage
yield of KjCraO? crystals produced.
Solubility of KjCraO? at 20°C = 0.390 lb-mole per 1000 lb HjO
Basis: 1000 lb of original solution.
Water = 870 lb
KzCraO? = 130 lb
Water remaining after concentration = 870 — 640 = 230 lb
K2Cr207 in solution after crystallization at 20°C =
230
X 0.390 = 0.090 lb-mole or 0.090 X 294 = 26.4 lb
• Yield of K j C r A crystals = 130 - 26.4 = 103.6 lb
,, 103.4
Percentage yield = -—— = 79.7%
ioU
Care must always be exercised that the true solvent in a crystallizing

system is recognized. For example, in Fig. 10, curve EB represents
conditions under which naphthalene is the solute and benzene the sol-
vent. However, curve EA represents the solubihty of benzene as
solute in naphthalene as solvent. If a solution having a concentration of
naphthalene less than that of point E is cooled to produce crystalliza-
tion, pure benzene will crystallize as the solute and the naphthalene will
be the solvent, remaining constant in quantity throughout the process.
Similarly, in aqueous solutions of salts, if the concentration is less than
the eutectic value, cooling will produce the crystallization of water as
pure ice and the system may be treated as a solution of water in salt.
Illustration 6. A solution of sodium nitrate in water contains 100 grams of NaNOa
per 1000 grams of water. Calculate the amount of ice formed in coohng 1000 grams
of this solution to a temperature of — 15°C.
Concentration of saturated, water-in-NaNOa solution at — 15°C = 6.2
gram-moles of NaNOa per 1000 grams of H2O .
Basis: 1000 grams H2O.
NaNOa in original solution = 100 grams
100
Per cent NaNOs by weight = — - = 9.1%
1100
Basis: 1000 grams of original solution.
NaNOs = 91 grams or 91/85 = 1.07 gram-moles
Water = . .• 909 grams
Water dissolved in NaNOj in residual solution =
—— X 1.07 = 173 grams
6.2
Weight of ice formed = 909 - 173 = 736 grams
Where Solvates Are Present. Where solvates are involved it becomes

necessary to consider the solvent chemically, combined with the solute
which is removed from solution when solvated crystals are precipitated
or which is added to the solution when solvated crystals are dissolved.
The calculations involved are most easily performed by establishing a
material balance for either component. A binary system of weight W
containing y per cent component A and (100 — y) per cent component
B will be considered. It is assumed that this solution separates under
a given temperature change into two phases, phase 1 having a weight
Wi and a composition of yi per cent component A, and phase 2 having
a weight W — Wi, and a composition of 2/2 per cent component A. A
material balance of component A gives
yW = yiWi + y2(W — wi) (1)
wi y - Vi ._.
or ^ = . (2)
W y^- yi
If separation results from evaporation of a weight wi of pure com-
ponent B, the material balance of component A becomes
yW = y\Wi-\r yi{W - w^ - w-i) / ' ^
Illustration 7. An aqueous solution of sodium sulfate is saturated at 32.5°C.

Calculate the temperature to which this solution must be cooled in order to crystal-
lize 60% of the solute as Na2SO4-10H2O.
From Fig. 12 the solubility at 32.5° is seen to be 32.5% NajSOi. , . j
Basiia: 1000 lb of initial solution.
Na2S04 crystallized = 325 X 0.6 = 195 lb
NajSOi-lOHzO crystaUized = 195/0.441 = 4421b
Water in these crystals = 442 - 195 = 247 lb
Water left in solution = 675 - 247 = 428 lb
NajSOi left in solution = 325 X 0.4 = 130 lb
Composition of final solution = 130/(130 + 428) = 23.3% NaaSOi
From Fig. 12 it is found that this concentration corresponds to a temperature of
27 °C, the required crystallizing temperature.
Illustration 8. A solution of ferric chloride in water contains 64.1% FeCU by I

weight. Calculate the composition and yield of the material crystaUized from
1000 lb of this solution if it is so cooled as to produce the maximum amount of crystal-
lization from a residual liquid.
From Fig. 11 it is seen that, if a solution of this composition is cooled, the hydrate
FeCl3-6H20 will crystallize. The maximum crystalHzation from a liquid residue will
result from cooling to the eutectic temperature, 27°C. Further cooling would cause
complete solidification of the system. From Pig. 11 the solubility of FeCla at the
eutectic temperature is 68.3% by weight.
CHAP. V] CALCULATIONS FROM LINE SEGMENTS 127
Basis: 1000 lb of origmal solution.

162.2
Percentage FeCU in FeCU-eHjO = —— = 60.0%
Let X = pounds of FeCla-CHaO crystallized.
Material balarKe of FeCl 3. • , ' "
Original solution Final solution Crystals
(1000) (0.641) = (1000 - a;) (0.683) + 0.600x
or X = 511 lb FeCl3-6H20 crystals
Illustration 9. A solution of sodium sulfate in water is saturated at a tempera-
ture of 40°C. Calculate the weight of crystals and the percentage yield obtained
by cooling 100 lb of this solution to a temperature of 6°C.
From Fig. 12 it is seen that at a temperature of 5°C the decahydrate will be the
stable crystalline form. The solubilities read from Fig. 12 are as follows:
at40°C:32.6%Na2SO4
at 5°C: 5.75% NajSOi.
Basis: 100 lb of original solution, saturated at 40°C.
142
Percentage Na2S04 in Na2SO4-10H2O crystals = = 44.1%
142 + 180
Let X = poimds of NajSOi-lOHjO crystals formed.
Material balance of Na2S0t.
Original solution Final solution Crystals
0.326(100) = 0.057(100 - x) + 0.441x
or ' , , .• ^ = 69-5 lb Na2SO4-10H2O formed
322
Weight of NazSOi-lOHjO in original solution = 3 2 . 6 — = 74 lb
142
69.5 . • .
Percentage yield = -—- = 94%
Calculations from Line Segments of Equilibrium Diagrams. In the

separation of crystals from a solution the weight ratio of the crystals to
the weight of the original solution is given by Equation (2), page 126.
The ratios of Equation (2) can be obtained directly from the Une seg-
ments on a binary equilibrium diagram when compositions are plotted
in weight percentage. This method is illustrated by Fig. 12 for the
system NaaSO^ and H2O. Starting with a homogeneous solution at a
temperature ti containing y per cent Na2S04 (component A), and cool-
ing, crystals of Na2SO4-10H2O (phase 1) will start to separate at a tem-
perature <2. As the temperature drops crystallization will continue,
the crystals having a uniform composition of 44.1% Na2S04 corre-
sponding to the decahydrate as represented by the upper boundary of
this two-phase field. The percentage of Na2S04 in the remaining liquid
phase becomes progressively less as crystallization proceeds, and its

decreasing composition will be represented, by points on the solubility
curve corresponding to the existing temperature. When the final tem-
perature ti is reached the composition of the residual Kquid will be
represented by ordinate 2/2, where ?/2 represents the percentage of com-
ponent A (Na2S04) in the liquid phase. The weight of the Na2S04-
IOH2O crystals relative to the entire weight of the system as given by
Equation (2) will be seen to be equal to the ratio of the line segments
y - Vi to yi - j/2.
In general, the percentage by weight of any phase in a two-phase
equilibrium mixture at a given temperature may be obtained from the
equilibrium diagram by extending the concentration line across the field
showing the phases present. The weight ratio of phase 1 to the total
weight will then be the ratio of the line segment extending between the
composition of the entire system and the composition of phase 2 to the
line segment extending between the compositions of phases 1 and 2.
Phase 1, represented by segment y — 2/2, always corresponds to that
phase which is richest in component A. For example, in the above
system phase 1 represents the Na2SO4-10H2O crystal phase for com-
positions above the eutectic and represents the liquid phase below the
eutectic composition where ice separates as the solid phase.
Illustration 9 will be solved from line segments on the equilibrium
diagram. By referring to Fig. 12 the value of y corresponding to sat-
uration at 40°C is 32.6 per cent. At 5°C, the composition yi of the
solid phase is 44.1 per cent, and y^ of the liquid phase is 5.75 per cent.
Applying the above rule of segments the percentage weight of crystals
32 6 — 5 75
separated is —7- r r r = 69.5 per cent. By this ratio of segment
lines the composition of any equilibrium mixture of two phases can be
estimated from the equilibrium diagram.
I t must be emphasized that the line-segment rule can be directly
applied only when compositions are expressed in weight percentages
and never when they are expressed in molality or mole fraction.
SOLUBILITIES IN COMPLEX SYSTEMS

FRACTIONAL CRYSTALLIZATION
All the systems thus fat discussed have contained only two primary
components. Where three or more components are involved the solu-
bility relationships become very complex because of the effect which the
presence of each solute has on the solubility of the others. These mu-
tual effects cannot be predicted without experimental data on the par-
CHAP. V] FRACTIONAL CRYSTALLIZATION 129
ticular case under consideration. For salts which are ionized in solu-
tion the following qualitative generalization applies: Where the same
ion is formed from each of two ionized solute compounds, the solu-
bility of each compoimd is diminished by the presence of the other.
For example, the solubility of NaCl in water may be diminished by the
addition of another solute which forms the chloride ion. This principle
is applied in the crystallization of pure NaCl by bubbling HCl into con-
centrated brine. Similarly, the solubility of NaCl in water is reduced
by the addition of sodium hydroxide. This accounts for the recovery
of nearly pure caustic soda by evaporation of cathode liquor in the
electrolysis of sodium chloride.
The mutual solubility relationships in complex systems are of great
importance in many industrial processes. Soluble substances fre-
quently may be purified or separated from other substances by prop-
erly conducted fractional crystallization processes. In such processes
conditions are so adjusted that only certain of the total group of solutes
will crystallize, the others remaining in solution. For the production
of very pure materials it is frequently necessary to employ several
successive fractional crystallizations. In such a scheme of recrystal-
lization the crystal yield from one step is redissolved in a pure solvent
and again fractionally crystallized to produce further purification.
Complete data have been developed for solubilities in many complex
systems. The presentation and use by such data for more complicated
cases are beyond the scope of a general discussion of industrial calcula-
tions. In an excellent monograph^ by Blasdale the scientific principles
involved in such problems are thoroughly discussed. Data for many
systems are included in the International Critical Tables.
In Fig. 13 are the solubihty data of Caspari^ for the system of sodium
sulfate and sodium carbonate in water, presented in an isometric,
three-coordinate diagram. A diagram of this type permits visualization
of the relationship existing in such systems of only three components.
The solubility data determine a series of surfaces which, with the axial
planes, form an irregular-shaped enclosure in space. Any conditions of
concentration and temperature which fix a point lying within this en-
closure correspond to a homogeneous solution. Points lying outside
the enclosure represent conditions under which at least one solid sub-
stance' is present. The surfaces themselves represent conditions of
equilibria between the indicated solids and saturated solutions. The
line formed by the intersection of two of these surfaces represents con-
• W. G. Blasdale, Equilibria in Saturated Salt Solutions, Chemical Catalog Co.,
New York (1927).
2 W. A. Caspari, J. Chem. Soc, 125, 2381 (1924). , .
ditions of equilibria between a mixture of two solid substances and

saturated solution.
For example, curves OABC of Fig. 13 represent the solubility in water
of Na2S04 alone, corresponding to Fig. 12. The addition of Na2C03 to
such a system lowers the solubility of the Na2S04, as indicated by the
surfaces ABRQP and BCUR. The former surface corresponds to
equilibria between crystals of pure Na2SO4-10H2O and solutions con-
taining both Na2S04 and Na2C03. Surface BCUR represents similar
equilibria with anhydrous Na2S04 as the solid. Similarly, surface
GFSQP represents equilibria between crystals of pure Na2CO3-10H2O
«a.
0 G 10 20 30 40 50'
_ Grams Na,CO, per 100 Grams H,0
P I G . 13. Solubility of sodium carbonate-sodium sulfate in water.
and solutions containing both Na2C03 and Na2S04. These solutions

are, therefore, saturated with respect to Na2C03 but unsaturated with
respect to Na2S04. If Na2S04 were added to a solution whose condi-
tions correspond to a point on surface OFSQP, the salt would dissolve
and Na2CO3-10H2O would crystalhze.
The line PQ represents solutions which are saturated with both
Na2C03 and Na2S04. Such solutions are incapable of dissolving greater
quantities of either salt and are in equilibrium with crystals of each.
It so happens that when crystals of Na2CO3-10H2O and Na2SO4-10H2O
are formed together in a solution, each sohd compound is sUghtly soluble
in the other, forming what are termed solid solutions. Therefore, line
PQ represents equilibria between crystals of Na2CO3-10H2O containing
small amounts of Na2SO4-10H2O in solid solution, crystals of Na2S04-

IOH2O containing small amounts of NaaCOs-lOHaO in solid solution,
and liquid solution containing both Na2S04 and Na2C03.
At the higher temperatures the system is complicated by the decom-
position of the hydrates and by the formation of a stable double salt of
definite composition, 2Na2S04-Na2C03. The surface RQSTVU repre-
sents equilibria between pure crystals of this double salt and hquid
solutions. The other surfaces and lines of the diagram may be inter-
preted in a similar manner from the composition of the equilibrium solid
which is marked on each surface.
An isometric diagram such as Fig. 13, though valuable as an aid to
visualization, is not suitable as a basis for quantitative calculations.
Data for calculation purposes are better presented as a family of iso-
thermal solubility curves on a triangular diagram as shown in Fig. 14.
The coordinate scales represent weight percentages of Na2S04, Na2C03
and H2O respectively. Each line of Fig. 14 represents the solubility
relationships at one indicated temperature. By interpolation between
these lines solubilities at any desired temperature may be obtained.
Points along curve AB represent conditions of saturation with both
solutes. The curves running upward to the left from this curve rep-
resent solutions which are saturated with sodium carbonate but only
partially saturated with sodium sulfate. Similarly, the curves running
upward to the right represent saturation with sodium sulfate. The
hydrates, NajSOi-lOHÔ (44.1% NasSOi) and NasCOa-lOHaO (37.1%
Na2C03), are represented by points C and D, respectively.
Point Xi on Fig. 14 represents a composition which at a temperature
of 25°C corresponds to a solution saturated with respect to sodium
carbonate but unsaturated with respect to sodium sulfate. If such a
(Solution is cooled, pure Na2CO3-10H2O will crystallize and the composi-
tion of the residual solution will change along the dotted line Dxi. At
a temperature of 22.5°C, corresponding to the intersection of this
dotted line with curve AB, the remaining solution will become sat-
urated with both sodium carbonate and sulfate. Further cooUng will
result in crystallization of both decahydrates in crystals of the solid
solutions which were previously discussed. The concentration of the
residual solution will then diminish along curve AB. If cooling were
stopped at 22.5°C a jdeld of pure Na2CO3-10H2O crystals would be
obtained, and a separation of one solute from the other would result.
Point Xi on Fig. 14 represents a composition which at a temperature of
25°C corresponds to a solution unsaturated with respect to both solutes.
If water is evaporated from such a solution at a temperature of 25°C
the concentration of each solute will be increased. Since the relative
proportions of the two solutes will remain unchanged, the composition

of the solution will vary along a straight line passing through point G
of the diagram. If evaporation is continued a composition x'2 will be
reached at which the solution is saturated with respect to sodium sul-
fate. Further evaporation will result ia crystallization of pure Na2S04-
IOH2O, and the composition of the residual solution will vary along
the 25°C isothermal line from x'i to B. When the residual solution
reaches a composition B it will be saturated with both solutes. Further
evaporation will produce crystallization of both solutes, and the com-
G 1007» H,0
Weight % NajSO,
F I G . 14. Solubility of the system NazCOs ~ Na2S04 — H2O at low temperatures.
position of the residual solution will remain unchanged at B. By such

isothermal evaporation processes it is possible to separate the original
solution into pure Na2SO4-10H2O and a solution of composition B if
the evaporation is not carried beyond the point of initial saturation with
both solutes.
Points lying below a solubility isotherm on Fig. 14 represent the
compositions of systems which at that temperature are heterogeneous
mixtures. For example, at 25°C, any point lying in the field IBC will
represent a mixture of solid Na2SO4-10H2O and a solution saturated

with sodium sulfate but not with sodium carbonate. Similarly, a
point in the field HBD represents a system of solid Na2CO3-10H2O and
solution. A point in the field EDBCF corresponds to a system sat-
urated with both decahydrates. It will consist of a solution of com-
position B and a mixture of crystals of solid solutions of the two
decahydrates.
With the aid of a chart of the type of Fig. 14 it is possible to carry out
by combined graphical and arithmetical methods the calculations in-
volved in crystallization problems.
Illustration 10. An aqueous solution at a temperature of 22.5°C contains 21 grams

of NazCOa and 10 grams of Na2S04 per 100 grams of water.
(a) Calculate the composition and weight of the crystals formed by cooling 1000
grams of this solution to a temperature of 17.5°C.
(6) Repeat the calculation of part (a) to correspond to a final temperature of 10°C.
Basis: 1000 grams of original solution.
Initial composition, point a on Fig. 14:
NajSO, = 7.64%
~ 131
21
NasCO, = 16.02%
~ 131
H2O = 76.34%
(o) Upon cooling, NajCOa-lOHzO will crystallize out along line aD, giving a com-
position at 17.5°C corresponding to point b.
NaaSOi = 8.5%
,' • NajCOa = 13.8% •' •
Let X = weight of NajCOa-lOHjO crystallized.

Then a material balance for NajCOs gives
(0.1602) (1000) = 0.138(1000 - x) + 0.37i(a;)

X = 98 g NaaCOa-lOHjO
(6) When the original solution cools to 10°C, crystallization of NaaCOa-lOHzO

will proceed along Da until it strikes line AB, after which crystallization of both
decahydrates will foUow line AB.
Composition of final solution at d is
NajSOi = 6.2%
NajCOs = 10.0%
H2O = 83.8%
Let X = weight of NajCOs-lOHzO crystallized,

y = weight of Na2SO4l0H2O crystallized.
From a NaaCOa balance

(0.1602) (1000) = 0.371 (a;) + (0.10) (1000 - x - y)
From a Na2S04 balance
(0.0764) (1000) = 0.441(2/) + (0.062) (1000 - x - y)
Hence a; = 251 grams
';* 2/ =, 79 grams
Where evaporation of solvent is involved in a crystallization process
two types of problems may arise. It may be desired to calculate the
quantity of solvent which must be removed in order to produce a speci-
fied yield of crystals or it may be desired to calculate the yield of crystals
resulting from the removal of a specified quantity of solvent. In the
first type of problem it is necessary to determine the composition of the
solution remaining after the process. The quantity of solvent to be
evaporated will then be the difference between that in the initial and
final solutions. If both solutes crystallize in the process the residual
solution will have a composition corresponding to saturation with both
solutes at the final temperature. If only one solute crystallizes, the
composition of the final solution may be determined graphically.
Illustration 11. An aqueous solution contains 9.8% Na2S04 and 1.8% Na2C03.
How much water must be evaporated from 100 lb of this solution to saturate the
solution with one solute at 20°C without crystallization?
Upon evaporation the composition of the solution will follow along a line }G,
starting from the original composition corresponding to / on Fig. 14.
The line /G crosses the 20°C line at a composition of 15.1% Na2S04 and 2.8%
NaaCOa.
Let X = water evaporated. Then a water balance gives
100(0.884) = X + (100 - x) (0.821)
or i s = 35.2 lb
If a specified quantity of solvent is to be evaporated from a solution
and it is desired to calculate the resultant yield of crystals, the composi-
tion of the entire final mixture of crystals and solution is readily deter-
mined by subtraction of the quantity of evaporated solvent. If, in the
final mixture, the concentration of only one of the solutes is greater than
that corresponding to saturation, only this one. A, will be crystallized.
The entire quantity of the other solute, B, must then be in the residual
solution, fixing its composition with respect to this solute. The com-
plete composition of the residual solution will be that corresponding to
this concentration of solute B and saturation with solute A at the existing
temperature. The quantity of solute A which will be crystallized will
be the difference between the total quantity and that remaining in
solution.
Illustration 12. A solution contains 19.8% NaaCOa and 5.9% Na2S04 by weight.
Calculate the weight and composition of crystals formed from 100 lb of this solution
when
(a) 10 lb of water is evaporated and the solution cooled to 20°C.
(6) 20 lb of water is evaporated and the solution cooled to 20°C.
Basis: 100 lb of solution.
(a) The residue after evaporation consists of
19.8
— - = 22% Na^COa
yu
• ^ = 6.55% NaaSOi
This composition corresponds to h on Fig. 14. Upon cooling, Na2CO3-10H2O

win crystalhze out along line Dh reaching at 20°C a composition corresponding to
point i, 9.4% NazSOi and 15.3% NajCOs.
Let a; = lb NaaCOs-lOHjO crystaUized. i
Then
' 90(0.22) = x(0.371) + (90 - a;)(0.1S3)
or i . X = 27.51b. ^
(b) After evaporation of 20 lb water the residue consists of
^ = 24.8% Na^CO,
|^= 7.4%Na,S04
At 20°C the solution becomes saturated with respect to both solutes, corresponding
to a composition of 14.8% Na2C03 and 11.4% Na2S04.
Let X = lb NajCOs-lOHaO crystallized, ^
2/ = lb Na2SO4-10H2O crystallized.
From a Na2C03 balance
0.248(80) = 0.371a; + 0.148(80 - x - y)
From a Na2S04 balance
0.074(80) = 0.442/ + 0.114(80 - x - y) *"
or X = 38.1 lb NaaCOa-lOHjO crystallized
y = 3.45 lb NajSOi-lOHjO crystaUized
The system of sodium sulfate, sodium carbonate, and water has
been selected as an illustration, not because of its industrial importance
but because it exhibits most of the phenomena to be found in such ter-
nary systems. Many other systems involve similar formations and
decompositions of hydrates, double salts, and solid solutions and may be
dealt with by means of similar diagrams and methods. Several systems
of industrial importance show peculiarities of individual behavior which
136 - SOLUBILITY AND SORPTION [CHAP. V
form the bases of important processes. For example, in the system

NaN03-NaCl-H20 lowering the temperature decreases the solubihty
of NaNOs but increases that of NaCl in solutions which are saturated
with both salts. This peculiarity makes it possible to crystallize pure
NaNOs by cooling a solution saturated with both salts. This prin-
ciple is used on a large scale for the conmiercial production of Chile
saltpetre. In the system NaOH-NaCi-H20 the solubility of NaCI in
solutions containing high concentrations of NaOH becomes very small.
This fact is taken advantage of in the separation of NaCl from elec-
trolytically produced NaOH solutions. The solution is merely con-
centrated by evaporation at a relatively high temperature where the
solubility of NaOH is great. As the concentration increases, NaCl
crystallizes and solutions may be produced containing only traces of
N a C l . • - . ;• . • • -,«.
/ • • •• • •-' . . ^ , 1 •" ,•- ' ^ v ,
VAPOR PRESSURE AND RELATIVE HUMIDITY ABOVE SOLUTIONS

Certain aqueous solutions are used for drjdng gases, dehumidification
of air, and control of humidity in air conditioning. Solutions of lithium
chloride, calcium chloride, glycerol, and triethylene glycol are used for
this purpose in large industrial apphcations, and in addition sulfuric
acid, phosphoric acid, and caustic soda solutions are used as laboratory
desiccants. To simplify calculations it is convenient to show in a single
diagram the relation of vapor pressure, relative humidity, temperature,
and composition of the water-desiccant system.
Such a diagram is shown for the calcium chloride-water system in
Fig. 15. The area to the left of curve ahcde represents unsaturated
solutions of calcium chloride in equilibrium with water vapor. The
curved lines are isotherms showing the decrease in vapor pressure with
increase in concentration. The nearly vertical dotted lines are constant
relative humidity. Curve ahc represents the solubility of the hexahy-
drate in water with corresponding changes in temperature, vapor pres-
sure, and relative humidity. Line cf represents pure hexahydrate cry-
stals containing 50.7% CaCU. In area dbcf, crystals of the hexahydrate
and its saturated solution are present. The horizontal lines inside this
area represent the vapor pressure of the saturated solution at various
temperatures, with corresponding relative humidities.
The vapor pressure of the hexahydrate is lower than that of the satu-
rated solution but any tendency for condensation on the surface of the
hexahydrate is immediately reversed by the formation of a saturated
solution having a vapor pressure equal to that of the system. At point c,
86°F, crystals of hexahydrate are transformed to the tetrahydrate, and
crystals of the hexahydrate do not exist above this temperature. At
CHAP. V] VAPOR PRESSURE AND RELATIVE HUMIDITY 137
point d, 112°F, crystals of tetrahydrate are converted to the dihydrate

and crystals of the tetrahydrate do not exist above this temperature. In
area fchg, crystals of both hexa- and tetrahydrates are present. The
horizontal lines represent vapor pressures of the hexahydrate. Three
phases exist in equilibrium and in mutual contact in this area. The
vapor pressure of the hexahydrate is higher than that of the tetrahydrate
10 20 30 40 50 60 70
Pet Cent CaClj by Weight CaClj CaCl2
6H2O 4HjO
FIG. 15. Vapor pressure and relative humidity over calcium chloride solutions.
but any tendency to condense water on the tetrahydrate is immediately

followed by hexahydrate formation with a vapor pressure equal to that
of the system. In the area cdjh, crystals of tetrahydrate are in equihb-
rium with a saturated solution of tetrahydrate. The horizontal lines
represent vapor pressures of the saturated solutions of tetrahydrate. In
area delm, crystals of dihydrate are in equilibrium with a saturated solu-
tion of dihydrate. The horizontal vapor pressure lines represent the
vapor pressure of the saturated solution at various temperatures. In
the area gjmn two solid phases of tetrahydrate and dihydrate are present.
The horizontal Unas represent the vapor pressure of the tetrahydrate at
various temperatures. The vapor pressure of the dihydrate is lower but
the tendency for water to condense on the dihydrate is reversed by for-
mation of the tetrahydrate with vapor pressure equal to that of the
system. - _ ,
ABEA PHASES PRESENT BESIDES WATER VAPOR PRESSURE LINES

VAPOR REPRESENT
ahcf Saturated solution and crystals of hexahydrate Saturated solution of
hexahydrate
fckg Crystals of hexahydrate and of tetrahydrate Crystals of hexahydrate
cdjh Saturated solution and crystals of tetrahydrate Saturated solution of
tetrahydrate
delm Saturated solution and crystals of dihydrate Saturated solution of di-
hydrate
gjmn Crystals of tetrahydrate and dihydrate Crystals of tetrahydrate
In using the calcium-chloride-water system in a drying and regenerat-

ing cycle, during the drying stage the vapor pressure of water in the gas
to be dried must exceed the equilibrium value of the system and in
regeneration the vapor pressure of water in the hot air blown through
must be less than the equilibrium vapor pressure of the system.
Regeneration of calcium chloride from solution is frequently not con-
sidered feasible because of the relative cheapness of the salt.
I t will be observed that in using a homogeneous solution of CaCU
below 40°F the lowest relative humidity obtainable is 4 5 % whereas
above 90°F the lowest attainable is 19%.
Illustration 13. Ninety pounds of anhydrous CaCU is placed in a room of 90,000 cu
ft capacity. The air is initially at 80°F and 90% relative humidity. Estimate the
conditions of air and desiccant when equilibrium has been attained at 80 °F, assuming
a pressure of I.O atmosphere.
It wiU be estimated that enough water is removed to liquefy the CaCU, leaving a
saturated solution of hexahydrate and air at a relative humidity read from Fig. 15 as
27%. This assumption wUl be verified by the calculations.
Weight of dry air in room = 6560 lb
Initial humidity lb/lb =0.0198
Final humidity = 0.0061
Water removed = 6560(0.0198 - 0.0061) = 89.9 lb.
Let X = crystals of hexahydrate remaining.
Then, (179.9 — x)] = weight of saturated CaCU solution.
From a CetCh balance
90 = 0.507(x) + 0.46(179.9 - x)
X = 153 pounds of CaCh-eHjO
(179.9 — 163) = 26.9 pounds of saturated solution.
CHAP. Y] DISTRIBUTION 139
DISTRIBUTION OF A SOLUTE BETWEEN

IMMISCIBLE LIQUIDS
When a solute is added to a system of two immiscible liquids, the
solute is distributed between the liquids in such proportions that a
definite equilibrium ratio exists between its concentrations in the two
phases. The equilibrium between the solute in the two phases is of a
dynamic nature with solute particles continually diffusing across the
interface from one liquid to the other. At equihbrium conditions the
concentrations adjust themselves so that the rate of loss of solute parti-
cles by each phase is compensated by its rate of gain of particles from
the other phase.
The equilibrium distribution of a solute between immiscible solvents
is expressed by the distribution coefficient, K, which is the ratio of the
concentrations in the two phases. Thus,
where CA, CB = concentrations of solute in phases A and B, respectively.

If sufficient solute is present to saturate the system completely, each
phase must contain solute in the concentration corresponding to its
normal saturation conditions. Therefore, the distribution coefficient
at saturation is merely the ratio of the solubilities of the solute in the two
Hquids.
TABLE III
DISTRIBUTION OF PICRIC ACID BETWEEN BENZENE AND WATER
CA = concentration of picric acid in water, gram-moles per liter of solution
CB = concentration of picric acid in benzene, gram-moles per liter of solution
K = distribution coefficient at 15-18°C = CB/CA
CB K
0.000932 2.23
0.00225 1.45
0.01 0.705
0.02 0.505
0.05 0.320
0.10 T 0.240
0.18 • 0.187
Effect of Concentration. In ideal systems in which dissociation and

association are absent the distribution coefficient is independent of con-
centration. Ordinarily, however, it shows marked variation with con-
centration, as indicated by the values in Table III for the distribu-
tion of picric acid, HOC6H2(N02)3, between water and benzene, CeHe.
Similar data for many other systems may be found in the International
Critical Tables, Vol. I l l , page 418.
Effect of Temperature. The effect of temperature on the distribu-

tion coefficient is generally small if the temperature coefficients of solu-
bility are approximately equal in the two phases. Specific data are
necessary in order to predict the effects of a temperature change. For
many industrial calculations the effects of temperature changes of only a
few degrees may be disregarded.
Distribution Calculations. The distribution of a solute between two
immiscible hquids is of considerable industrial importance in the sepa-
ration and purification of organic compounds. Ordinarily one liquid
will be water or an aqueous solution and the other some immiscible
organic solvent. Equilibrium concentrations of a solute in such systems
may be varied by the addition of a second solute which is soluble in only
one of the liquids. The addition of such a solute is, in effect, a change
in the nature of one of the liquids.
From values of distribution coefficients equilibrium conditions are
readily calculated.
Illustration 14. Picric acid exists in aqueous solution at 17°C in the presence of
small amounts of inorganic impurities whose effects on its solubility may be neglected.
The picric acid is to be extracted with benzene in which the inorganic materials are
insoluble.
(a) If the aqueous solution contains 0.20 gram-mole of picric acid per liter, calcu-
late the volume of benzene with which 1 Uter of the solution must be extracted in
order to form a benzene solution containing 0.02 gram-mole of picric acid per liter.
(Neglect the difference between the volume of a solution and that of the pure solvent.)
(b) Calculate the percentage recovery of picric acid from the aqueous solution.
Basis: 1 liter of original aqueous solution.
Picric acid = 0.20 gram-mole
From Table III, K = 0.505 in final sys-
tem = CB/CA
Final concentration of picric acid in
aqueous solution = 0.02/0.505 = . 0.03% gram-mole per liter
Picric acid in final benzene solution =
0.20 - 0.0396 = 0.16 gram-mole
n 1 fi
Benzene required = -—- = 8.0 liters per liter of aqueous solution
n 1 fi
Percentage extraction of picric acid = --— = 80%
yj.cXj -
In calculations involving concentrated solutions the differences

between the volume of a solution and that of the pure solvent cannot be
neglected as was done in Illustration 14. In such cases it is convenient
to express the concentrations and distribution coefficients in .terms of the
weight of solute per unit weight of solvent. The units in which dis-
tribution data are ordinarily expressed, as in Table III, may be readily
CHAP. V] DISTRIBUTION CALCULATIONS 141
converted into these terms if density-concentration data are available

for both solutions.
If definite quantities of two immiscible solvents and a solute are mixed
together, the final concentration of solute in either solution will be
unknown. If the distribution coefficient varies considerably with con-
centration it will also be unknown. The distribution of the solute in
such a case is best estimated by a method of successive approximations.
A reasonable value of the final distribution coefficient is assumed as a first
approximation. On the basis of this assumed value a first approxima-
tion to the final concentrations is calculated. The distribution coeffi-
cient is then corrected to correspond to these concentrations. On the
basis of the second approximation to the distribution coefficient, a second
approximation to the final concentrations is calculated. Unless the
variation of distribution coefficient with concentration is very marked,
two or three successive approximations of this type will yield results
satisfactory for ordinary purposes.
Illustration 15. One liter of a benzene solution containing 0.10 gram-mole of picric
acid per liter is agitated with 1.0 liter of water. Estimate the final concentration
of picric acid in each solvent.
Solution: As a first approximation, assume from Table III that the final dis-
tribution coefficient will be 0.5.
Let xi = gram-moles of picric acid in final benzene solution.
Picric acid in final aqueous solution = 0.10 — xj.
Xi
0.10 - xi 0.5
Xi = 0.033
The distribution coefiicient corresponding to this concentration is taken from the
data of Table III as a second approximation.
K2 = 0.39 ^ «
X2
= 0.39
0.10 - X2
X2 = 0.028
As a third approximation:
K3 = 0.42 (corresponding to CB = 0.028)
_^1_=0.42
0.10 - X3
X3 = 0.029
This result may be taken as the final concentration of picric acid in the benzene solu-
tion. If greater accuracy is desired, more approximations should be carried out.
Picric acid in final benzene solution = 0.029 gram-mole
Kcric acid in final aqueous solution = 0.071 gram-mole
V , PARTIALLY MISCIBLE LIQUIDS

When two partially miscible liquids are brought together, a range of
compositions occurs in which two liquid phases exist in equilibrium with
each other. This behavior is typified by the water-phenol system illus-
trated in Fig. 16. At a temperature of 50°C mixtures containing more
than 11.8% and less than 62.6% phenol by weight will separate into two
layers, the upper containing 11.8% phenol and the lower 62.6%. Addi-
tion of either component to such a two-phase system will not change the
•~—;d_
r*'*^^^^
10
Upper U er
/)
0
30 40
Tempewtiire.'C
' FIG. 16. Solubility of phenol in water.
composition of either phase but will merely shift the proportions in

which they are present. Two Uquid phases existing in equilibrium with
each other in this manner with their compositions independent of the
composition of the total two-phase mixture are termed conjugate
solutions.
If the temperature of the phenol-water mixture is increased above
50°C the compositions of the conjugate solutions are changed, the phenol
content of the upper layer increasing and that of the lower layer decreas-
ing. When a temperature of 66°C is reached the compositions of the
conjugate phases become equal and the mixture becomes homogeneous.
CHAP. V] TERNARY LIQUID MIXTURES 143
The composition at which equality of composition of the two phases is

reached is termed the critical solution composition and the temperature at
which miscibihty becomes complete is the critical solution temperature.
In Fig. 16, point C corresponds to a critical solution composition of 34%
phenol and a critical solution temperature of 66°C.
The weights of the two phases in a heterogeneous binary system are
readily calculated by component material balances if the composition of
the entire mixture is known, together with compositions of the conjugate
solutions as shown in Fig. 16. It may be demonstrated that if point X
represents the composition of the entire mixture, composed of phases of
compositions A and B, the weight of phase A is proportional to line seg-
ment BX and the weight of phase B is proportional to segment AX.
Ternary Liquid Mixtures. Mixtures of three or more partially mis-
cible liquids may separate into two phases in equilibrium with each other
in a maimer analogous to that of the binary mixtures described above.
In Figs. 17 and 18 are plotted the isothermal solubiHty data of a repre-
sentative ternary system, tetrachlorethylene, isopropyl alcohol and
water at 77°F, as determined by Bergelin, Lockhart and Brown.^ At
this temperature water is completely miscible with isopropyl alcohol.
Similarly, tetrachlorethylene is completely miscible with isopropyl
alcohol. However, water and tetrachlorethylene are almost completely
immiscible with each other.
When the three components are mixed together, homogeneous solu-
tions are formed only at compositions lying above curve ach of Fig. 17.
Mixtures having compositions lying below this curve separate into two
layers, the upper rich in water and the lower rich in tetrachlorethylene.
Curve ad), bounding the two-phase region, is termed an isothermal
solubility curve. Similar curves for other temperatures might be plotted
on the same diagram to define completely the solubility characteristics
of the system,
A mixture having a composition x in the two-phase region separates
into two conjugate phases experimentally found to have compositions
corresponding to points d and e. Thus, an upper layer of composition d
must always be in equilibrium with a lower layer of composition e. As
pointed out in Chapter I it is characteristic of the ternary diagram that
mixtures of two solutions must have compositions corresponding to com-
positions falling on a straight line connecting the composition points of
the original solutions. Thus, all mixtures of the conjugate phases having
compositions d and e must he along line de. This line is termed a tie-line,
connecting the composition points of conjugate phases. It may be
demonstrated from the geometry of the diagram that the weights of the
' Bergelin, Lockhart, and Brown, Trans. Am. Inst. Chem. Eng., 39, 173 (1943).
two phases into which a mixture of composition x separates are propor-

tional to the lengths of the line segments da; and ex. Thus, the weight
fraction of the upper-layer solution is equal to segment ex divided by
segment de.
In order to define completely the solubility and equilibrium relation-
ships of such a ternary system, a family of tie-lines such as fg and hi
must be drawn in, linking together the compositions of conjugate solu-
tions on the solubility curves. Tie-lines lying higher in the diagram than
hi would be progressively shorter until at point c-the length of the tie-
Iso-CsHjOH
A™
k60
\ /(Upper Layer \ /
<&/ \ fd
k50 •,^y
/Jk/ Vi. / \
i.4o\
\sj \ Lower
Sf)\ vC^ Layer
Y \/ \/
•-JAX/XX
*/ Y )/ ^lAtve^-.
HjO A/ / \ '7\ CaCl»
<? ^ <§> •S" <?
F i a . 17. Solubility curve and conjugate line for tetrachlorethylene-isopropyl

alcohol-water system at 77°F in weight per cent. From Bergelin, Lockhart and
Brown, Trans. Am. Inst. Chem. Eng., 39, 173 (1943).
line becomes zero. This point is termed the critical solution point or
plait point, at which the compositions of the upper and lower layers of the
two-phase system become equal.
The locations of all tie-lines can be defined without confusing the dia-
gram by constructing them through use of curve ac, termed the conjiir-
gate line. This line is the locus of the points of intersection of lines
drawn through conjugate compositions points parallel to the base and
right side of the diagram respectively. Thus, point j is located as the
intersection of lines dj and ej. Conversely, point j defines the location of
tie-line de. In a similar manner any other tie-lines on the diagram may
be located from the conjugate line.
It will be noted that for this system the critical solution point does not
CHAP. V] TERNARY LIQUID MIXTURES 145
correspond to the maximum solubility of alcohol. Accordingly, the

concentration of isopropyl alcohol must pass through a maximum in the
upper layer as the isopropyl alcohol content of the system is increased.
This behavior is shown more clearly in Fig. 18 where the concentration
of isopropyl alcohol in the upper layer is plotted as a function of its
concentration in the conjugate lower layer. This behavior is typical
of many ternary systems and clearly indicates the extreme deviations
55
/Wtfo
so
45
Mole %
40
2 35
30 ll
/
10 15 20 25 30 35 40 45
Isopropyl Alcohol in Lower Layer
Fia. 18. Equilibrium distribution of isopropyl alcohol in water and tetrachlorethy-

lene at 77°F in weight per cent. From Bergelin, Lockhart, and Brown, Trans. Am.
Inst. Chem. Eng., 39, 173 (1943).
which may be encountered from the constancy of distribution factor

assiuned in Equation (4).
Methods for the correlation of solubiUty and equilibrium data and the
prediction of tie-line locations have been developed by Othmer and
Tobias^ for systems of two slightly miscible components A and B in
admixture with a third component C which is miscible with both pure
A and pure B. For a large number of such systems the following em-
* Othmer and Tobias, Ind. Eng. Chem., 34, 690, 693, 696 (1942).
146 SOLUBILITY AND SORPTION ^ [CHAP. V
pirical equation was found to relate the compositions of the conjugate

phases: -
1 -ax ^ n - biV
(4a)
where ai = weight fraction of A in the phase rich in A

hi = weight fraction ,of B in the phase rich in B
u, V = constants characteristic of the system and temperature
Since all conjugate compositions lie on the solubility curve knowledge
of ai and 62 completely specifies the compositions of two conjugate
solutions if the solubility curve is known. Knowledge of the com-
positions of two sets of conjugate solutions permits evaluation of the
constants u and v in Equation (4a). Once u and v are established
values of cfi corresponding to assumed values of 62 are readily calcu-
lated from Equation (4a) and the entire system of tie-lines and the
conjugate line are established. These investigators also discuss methods
for developing partial vapor-pressure isotherms which may be plotted
directly on solubility charts such as Fig. 17.
SOLUBILITY OP GASES
Sorption comprises the general phenomenon of the assimilation of a
gas by a solid or liquid. When the sorbed gas forms a homogeneous
solution with the liquid or forms a new chemical compound with the
solid, the transformation is called absorption. When the gas is taken on
ohly at the surface or in the capillaries of the solid to form a surface com-
pound or condensate the phenomenon is designated as adsorption.
When a gas is brought into contact with the surface of a liquid, some
of the molecules of the gas striking the liquid surface will dissolve.
These dissolved molecules of gas will continue in motion in the dissolved
state, some returning to the surface and re-entering the gaseous state.
The dissolution of the gas in the liquid will continue until the rate at
which gas molecules leave the liquid is equal to the rate at which they
enter the liquid. Thus, a state of dynamic equilibrium is established
and no further change will take place in the concentration of gas mole-
cules in either the gaseous or liquid phases. The concentration of gas
which is dissolved in a liquid is determined by the partial pressure of
the gas above the surface.
Henry's Law. For many gases the relationship between the con-
centration of gas dissolved in a liquid aud the equilibrium partial pres-
sure of the gas above the liquid surface may be expressed by Henry's
law. The ordinary statement of this law is that the equilibrium value of
CHAP. V] HENRY'S LAW 147
the mole fraction of gas dissolved in a liquid is directly proportional to the

partial pressure of that gas above the liquid surface, or
1
,iVi (5)
H Pi
where *
pi = equilibrium partial pressure of gas in contact with liquid
iVi = mole fraction of gas in liquid
H = Henry's constant, characteristic of the system
This relationship has been found to be satisfactory at low concentra-
tions, corresponding to low partial pressures of gas and high values of H.
The factor H is a function of the specific nature of the gas and liquid and
of the temperature, in general increasing with increase in temperature.
3.0
1 V
\ \
\
\
0.05 IJ£ 25
\ "^ 4
0.04 > 2.0
0;
0.03
S OfV. ^•^s
Ai
0.02 1.0
H
0.01 0.5
nno
20 40 60 100
Temperaturt, °C
Fio. 19. Solubilities of gases in water. Variation of Henry's constant, H, with

temperature {pressures in millimeters of mercury).
When pressures and concentrations are low the solubility data of a

gas-liquid system are completely expressed by data relating values of
Henry's constant, H, to temperature. In Fig. 19 are curves expressing
this relationship for several common gases in water, the numerical values
of 1/H corresponding to pressures in millimeters of mercury. The data
148 SOLUBILITY AND SORPTION (CHAP. V
for hydrogen sulfide and carbon dioxide will lead to considerable error if
used for pressures above about 1 atmosphere. ' " = ,• •• -
Xllustration 16. Calculate the volume of oxygen, in cubic inches, which may be
dissolved in 10 lb of water at a temperature of 20°C and under an oxygen pressure
of 1 atmosphere. *• ,' ..Y'^^--'
Solution:
From Fig. 19, for oxygen and water at 20°C, 1/ff = 0.033 X lO"'
Mole fraction of Oj = 760 X 0.033 X lO"' = 25.1 X lO"'
Mole fraction of water = 1.00
Pound-moles of dissolved O2 = — X 25.1 X 10-« = 13.9 X IQ-'

18
293
Volume of dissolved O2 = 13.9 X 10-« X 369 X — = • • 5.37 X lO"' cu ft
273
or 9.3 cu in. measured at 20°C, and a pressure of 1 atmosphere
Deviations from Henry's Law. Under conditions of high pressure or
for a gas of relatively high solubility the direct proportionaUty of
Henry's law breaks down. Aqueous solutions of ammonia, carbon
dioxide, and hydrochloric acid are examples of systems whose behaviors
deviate widely from that predicted by Henry's law except at low pres-
sures. This deviation results in part from chemical reaction of the gas
with the liquid and subsequent ionization of the dissolved molecules.
Nemst has pointed out that Henry's law holds closely even for these
cases when applied to the same single species in both phases, that is, to
only the uncombined and unassociated molecules in the liquid phase.
However, in general, experimentally determined data relating tempera-
tures, pressures, and solubilities over the entire desired range are neces-
sary in order to predict solubilities in such systems. These data are
ordinarily expressed by solubility isotherms which relate the concentra-
tion of a dissolved gas to its partial pressure at a constant temperature.
In Fig. 20 are solubility isotherms of ammonia in water at various tem-
peratures. I t will be noted that all the lines have considerable curva-
ture, and that those corresponding to the lower temperatures show points
of inflection at high pressures. The form of the 20°C isotherm is typi-
cal of the behavior to be expected of a gas which is below its critical
temperature and dissolved in a solvent with which it is miscible when
in the liquid state. If measurements were continued at higher pres-
sures the slope of the curve would be expected to increase, becoming
asymptotic to the abscissa corresponding to a pressure of 6420 milli-
meters, the vapor pressure of liquid ammonia at 20°C. Gaseous am-
monia at 20°C could not exist at higher pressures, and at this pressure
an infinite amount of ammonia could be condensed and dissolved in 1
gram of water. The isotherms corresponding to higher temperatures
CHAP. V] ADSORPTION OF GASES 149
should exhibit similar points of inflection if extended to higher pressures.

From the data of Fig. 20 it is apparent that Henry's law should not
be used when dealing with a gas with a considerable affinity for the
solvent or at high pressures. At high pressures the solubility calcu-
lated from Henry's law will, in general, be higher than the correct value.
l£
/'
^
X
O
3! 0,
g I.O
2 ^ oC
C3
v^ *"
g.
m•c
X '
1 0.5
2
w
t
0 1000 2000 3000 4000

/
Pressure of NH.% mm. of Hg
F I G . 20. Solubility of ammonia in water.
In such cases specific experimental data such as those of Fig. 20 are

necessary for dependable calculations. Data for many systems, both
in''aqueous and organic solvents, are contained in the International Crit-
ical Tables (Vol. Ill, pages 255-283). Data are included on the
solubilities of gases in other solutions as well as in pure liquids. In
general, the solubility of a gas in a hquid is diminished by the addition
of a nonvolatile solute with which it does not react chemically.
ADSORPTION OF GASES
The adsorption of gases by solids is of importance in many industrial

operations such as the purification and drying of gases, the recovery c^
solvent vapors, the production of casing-head gasoline, and the oper-
ation of gas masks. An appUcation of increasing importance is in air
conditioning, where dehumidification is often accomplished by adsorp-
tion of water vapor on sohd desiccants.
Two types of adsorption should be recognized, one caused by the
same type of intermolecular forces of attraction that produce normal
condensation to the hquid state, called van der Waals forces, and the
other caused by specific chemical bonds between the atoms on the

surface of the solid and of the gas adsorbed, termed activated adsorp-
tion. Activated adsorption is of interest in connection with catalytic
phenomena but of little interest in processes depending upon readily
reversible adsorption and desorption such as discussed in this chapter.
van der Waals Adsorption. All molecules both hke and unlike are
subject to van der Waals forces of attraction, which bring about the
normal condensation of a vapor at its dew point as well as the adsorp-
tion of gases by solids above the dew point. Where the van der Waals
forces of attraction are greater between a solid and gas than between
the like molecules of the gas, it is possible to condense the gas by ad-
sorption on the solid surface at a temperature above the normal dew
point.
. On a smooth surface van der Waals adsorption is restricted to a
layer of one or a few molecules in thickness. However, on a solid
possessing a minute capillary structure surface adsorption is supple-
mented by capillary condensation which is also brought about by the
van der Waals forces of attraction.
In van der Waals adsorption the union between the surface of the
sohd and the adsorbed molecule is not permanent. Adsorbed molecules
which have acquired sufficient energy to overcome the surface forces
continually evaporate while other molecules are being adsorbed. When
a gas is brought into contact with an adsorbent surface, adsorption takes
place until the rate at which gas molecules strike the surface and are
adsorbed is equal to the rate of evaporation of adsorbed molecules. No
further change will then take place in the concentration of the gas in
either the gaseous or adsorbed phases and a condition of dynamic
equilibrium wiU exist.
The equilibrium between a gas and a solid adsorbent is similar to that
existing between a pure liquid and its vapor or between a gas and its
solutions. The amount of gas which is adsorbed at equilibrium always
increases with increase in partial pressure and decreases with increase
in temperature.
Capillary Condensation. Adsorbents of industrial importance are
generally substances of highly porous structure, which expose enor-
mous interior surfaces. Activated alumina, charcoal, and sihca gel are
familiar examples. In silica gel, for instance, the capillary pores are
about 4 X 10"^ centimeters in diameter, only 10 times the diameter
of simple molecules, and comprise about 50% of the total volume.
The total interior area is about one acre per cubic inch. In such ma-
terials capillary condensation is of great importance.
Capillary condensation can occur only when the soUd is wetted by
CHAP. V] EQUILIBRIUM ADSORPTION 151
the condensate, resulting in concave surfaces of the condensed liquid.

The equilibrium vapor pressure of a liquid having a concave surface is
less than the normal value by an amount depending upon the radius of
curvature. In the submicroscopic capillary pores of a wetted solid the
condensation of liquid will produce concave liquid surfaces of extremely
small radii of curvature and correspondingly low vapor pressures. For
this reason vapors which are at partial pressures much less than the
normal saturation value are condensed, augmenting the adsorption
normally taking place on flat surfaces. The adsorbing capacity of a
material possessing submicroscopic capillarity is considerably greater
than for one having the same surface area but having no capillary
structure.
As the partial pressure of a gas is increased adsorption progressively
increases. At low partial pressures capillary condensation does not
take place. When a pressure sufficient to produce condensation in the
smallest capillaries is reached, capillary condensation begins and the
capillaries fill to levels of greater diameter at higher vapor pressures.
The relationship between pressure and amount adsorbed is, therefore,
dependent on the size distribution of the capillary pores as well as on the
area of exposed surface and the nature of both adsorbent and gas.
Equilibrium Adsorption. The amount of gas adsorbed by a soHd
under equilibrium conditions may be expressed either as percentage by
weight, or as the mass or gaseous volume of adsorbate per unit mass of
gas-free adsorbent. The percentage water adsorbed by a solid can be
plotted against partial pressure oi water vapor in a family of isotherms,
or against temperature in a family of isobars. All the data presented
in such extensive isotherm and isobar plots can be approximately rep-
resented by a single line for a given substance when moisture content
is plotted against relative humidity. In Fig. 21 are plotted equilib-
rium moisture contents against relative humidity for a number of
common materials.
- For silica gel the commercially dry basis is used in expressing moisture
content. The commercially dry gel contains 5% water in chemical
combination which is not removed in desorption or regeneration opera-
tions and hence is not included in reporting moisture content. If this
chemically adsorbed water were removed, the adsorptive capacity of the
gel would disappear. For a particular silica gel the useful range of
moisture content for temperatures up to 100°F may be represented by
the simple equation:
We= 55Hr= 5 5 — ' (6)

where
Hr = relative humidity
We = equihbrium moisture content, pounds per 100 pounds of
commercially dry gel
Pa = partial pressure of water vapor in air -
p, = vapor pressure of pure water at temperature of air
40
/\
35
'•
-30
Q
•& 25
S
,11
^20
IS
yvn
10
-IX
x_
20 40 60 80 lOO
Percentage Relative Humidity
I. Silica gel. VI. Cotton cloth.

II. Leather, chrome tanned. VII. Sulfite pulp, fresh, unbleached.
III. Wool, worsted. VIII. Bond paper.
IV. Activated alumina. IX. Cellulose acetate silk, fibrous.
V. Viscose X. KaoUn (Florida).
PiQ. 21. Equilibrium moisture content of various substances, at 77°F.
The simplicity of presenting equilibrium adsorption against relative

saturation instead of as isotherms or isobars is at once apparent. It
should be noted that relative saturation should be employed and not
CHAP. V] ADSORPTION ISOTHERMS 153
percentage saturation, since the former is quite independent of total

pressure, as shown on page 92.
Illustration 17. Air at atmospheric pressure and 70°F leaves a bed of silica gel
which contains 5% water (commercially dry basis), at the exit end. Assuming
equihbrium conditions, what is the relative humidity of the exit air? It is assumed
that the operation is isothermal.
From Fig. 21 the relative humidity of air leaving at equihbrium with siUca gel of
6% water content is 9% or from Equation (6) the relative humidity is
X 100 = 9.1%
65
Equihbrium moisture content is of importance in the drying of ma-

terials which exhibit adsorptive characteristics. It is evident that water
will not evaporate from an adsorbent sohd into a gas the relative humid-
ity of which is higher than that corresponding to the equilibrium mois-
ture content of the solid. Thus, the equilibrium moisture content repre-
sents the minimum moisture content to which a material can be dried
by a gas of a given relative Humidity. Continued passage of gas will not
result in further dr3dng even though the gas is far from normal saturation
with water.
rv,-Pressure,, mm of Hg
0 5 10 15 20 25
1 •
< ii--
e
O
Q.
T3
<
2 jii
'
^
/ o
>-T"
^
/••
I
/
W
a
2
0 10 20 30 40 50 60_ 19
"^ 1,11, III,-Pressure, cm of Hg
I. COj on silica gel at 0°C. Patrick, III. CiHs on cocoanut charcoal at 59.5°C.
Preston and Owens, J. Phys. Chem., Coolidge, J. Am. Chem. Soc., 46, 596
29, 421 (1925). (1924).
II. SOz on silica gel at 30°C. McGavack IV. H2O on pine charcoal at 25''C. All-
and Patrick, J. Am. Chem. 'Soc, 42, mand, Hand, Manning, and ghiels, J.
946 (1920). Phys. Chem., 33, 1682 (1929).
FIG. 22. Adsorption isotherms.
Adsorption Isotherms. A group of typical adsorption isotherms is

shown in Fig. 22. The quantity of adsorbate is expressed as the volume
of gas measured in millimeters at 0°C and 760 millimeters of mercury,

per gram of gas free solid.
It will be noted from these curves that adsorption is a specific property-
depending upon the nature of the system. Curves I and II are typical
of the adsorption of a gas which is above its critical temperature or far
removed from conditions of normal saturation and are similar in shape
to the curves characteristic of activated adsorption. The adsorbed
quantity increases with increased pressure but at a continually diminish-
ing rate. The adsorption isotherm for this case may be expressed by the
Taylor equation:
1 + ap
where
w = the mass adsorbed per unit weight of adsorbent
p = the partial pressure of the adsorbate in the gas phase
o = a constant
In many cases the adsorption isotherm is satisfactorily represented by
the empirical equation proposed by FreundUcb:
w — fcp" '* ^)
where
h and n = empirical constants
Curve III is typical of the adsorption of a vapor below its critical
temperature at pressures in the region of the normal vapor saturation
pressure. When the pressure of the vapor is sufficiently increased to
equal the normal vapor pressure of benzene at the existing temperature
the vapor becomes saturated and normal condensation to the liquid
state results. The quantity of Hquid in equilibrium with a unit weight
of solid at this pressure may then become infinite.
In Fig. 23 the isotherms of benzene vapor adsorbed on activated
charcoal from the data of CooHdge^ are shown. The adsorbed quantity,
X, in milliliters of vapor, measured at standard conditions, adsorbed per
gram of gas-free or " outgassed " charcoal are plotted as ordinates.
Logarithms of the pressures of benzene vapor in millimeters of mercury
are plotted as abscissas. The logarithmic scale i^ desirable because of
the wide range of pressures required. The characteristic shape of these
curves when plotted in linear coordinates is indicated by Curve III of
Fig. 22. The coconut charcoal used in the experiments on which
Fig. 23 is based had been outgassed by heating to 550°C at reduced
pressure.
» / . Am. Chem. Soc. 46, 596 (1924).
CHAP. V] ADSORPTION ISOTHERMS 155
I t will be noted from Fig. 23 that, like the systems of Curves III and
IV of Fig. 22, when the pressure is sufficiently increased to equal the nor-
mal vapor pressures of benzene at the existing temperature, the vapor
becomes saturated and normal condensation to the Uquid state results.
, 1 11 •••• ~J_"
^ ^ ^ Solid ^='*"---,;"
Z ^/^ ^Z ^^ •'^^^
^'i^ ^ ^-^ ^ ^ ^ ^ ^
^^ ^ ^^ X >^ ^
4>^ y /^ ^ .7 .^
(/' " y / '^ y /
/_ -^JL ~^ ^ /. /.
riC. <^ _ / - A - ^ ^ — /. -/• r —
z ^ zW- TT^ /^ 7 -7
-I ^/ 2 ^ 1^/
V -Xy ^ / / /Z
/ ^ Z I'v'^ ^^ _ ^^
-0 T y /'^ J T^/ • 7 * / T^/ 7
_^' C T -,
4. ^^ /^ y J4 K4 z
^^ ^^ ^ ^ / <^/KJ-
y ^^ y ^ y^ ^^ -/^ ^y vC ^v > ^ j ^ / ^ /
^^ ^^ -""^ ^^ ^"^ ^
^-^^__.^-''^ ^^ ^ -^ y
• -
^---''^
— ' ^
^-^"^ ^'^ ^ — — " * " . — — — " ^ — . • — " ^
-^^
o-_- _---= _-- _ - .- -{-_---
- 3 - 2 - 1 0 1 2 3
Log,„P (millimeters)
FIG. 23. Adsorption of benzene on activated cocoanut charcoal
(outgassed at S50°C.)
The data of Fig. 23 are rigorously applicable only to the particular char-
coal for which they were determined. The quantity adsorbed is depend-
ent on the source of charcoal, the method of its preparation, and its
subsequent treatment. The same limitation applies to adsorption data
for other materials.
Illustration 18. (a) Estimate the number of pounds of benzene which may be
absorbed by 1 lb of the activated charcoal of Fig. 23 from a gas mixture at 20°C in
which the partial pressure of benzene is 30 mm of Hg.
(b) Calculate the percentage of the adsorbed vapor of part (a) which would be
recovered by passing superheated steam at a pressure of 5 lb per sq in. and a tem-
perature of 200°C through the adsorbent until the partial pressure of benzene in the
steam leaving is reduced to 10.0 mm of Hg.
(c) Calculate the residual partial pressure of benzene in a gas mixture treated
with the freshly stripped charcoal of part (6) at a temperature of 20°C.
SohUion: From Fig. 23,
(o) CcHe adsorbed at 20°C, 30 mm of Hg = 110 cc per g
«"• 7;^r:^ X '^^ = 0.382 g per g or lb per lb „ . , , ,,

(6) CeHe adsorbed at 200°C, 10.0 mm of Hg = 22 cc per g

Benzene recovered = 110 — 20 = ,. 90 cc per g
90
Percentage recovery = —— = 82%
(c) Pressure of benzene in equilibrium at 20°C with charcoal containing 20 cc of

benzene per g = antilog of 4.9 = 8 X 10~* mm of Hg, the residual partial
pressure of benzene.
Reversibility of Adsorption: Stripping. As previously pointed out,

van der Waals adsorption is a reversible process and such an adsorbed
gas is vaporized if its partial pressure in the gas phase is reduced below
its vapor pressure in the adsorbed phase. The recovery or stripping of
adsorbed gases may be accomplished in the following ways:
o. The temperature of the solid may be raised until the vapor pressure
of the adsorbed gas exceeds atmospheric pressure. The adsorbate vapor
will then be evolved and may be collected at atmospheric pressure.
b. The adsorbed gas may be withdrawn by applying a vacuum lower-
ing the total pressure below the adsorbate vapor pressure. Enough
heat should be supplied to prevent a drop in temperature as a result of
evaporation. The undiluted adsorbate vapor may then be coUeated at
this low pressure.
c. A stream of an inert, condensible gas may be blown through the
adsorbent, keeping the partial pressure of the adsorbate gas in the gas
stream below the equilibrium pressure of the adsorbate in the solid.
The adsorbate vapor will be evolved in admixture with the inert gas.
By using an easily condensible vapor for stripping, such as superheated
steam, the adsorbed material may be easily recovered by condensing the
stripping vapor only or by condensing the entire mixture and separating
by decantation provided the two condensed vapors are immiscible.
d. The adsorbed vapors may be displaced by treatment with some
other vapor which is preferentially adsorbed.
The various desiccants used for drying gases are usually regenerated
by blowing hot air through the spent adsorbent. For example, silica gel
is regenerated by hot gases at 250°F to 350°F and activated alumina by
hot gases from 350° to 600°F.
Adsorption Hysteresis. Curve IV on Fig. 22 indicates a behavior
which is typical of certain adsorbents which possess a high degree of
capillarity. As indicated by the arrows on the double section of the
curve, the quantity adsorbed in equilibrium with a selected partial pres-
sure is dependent on the direction from which equilibrium is approached.
If equiUbrium is reached by the evolution of adsorbate the upper or
" out " curve is applicable. If the equilibrium conditions are reached by
adsorption, with a continually increasiiig concentration, the lower or
CHAP. V] PREFERENTIAL ADSORPTION 157
" in " curve applies. This behavior is known as adsorption hysteresis

and is exhibited by many vapors when adsorbed in large quantities on
charcoal and other capillary adsorbents. In calculations dealing with
such systems data for both " in " and " out " curves are required.
This phenomenon can be explained by assuming a particular geometric
shape of pore spaces within the solid. In Fig. 24, is shown a pore space
inside a solid undergoing desorption in one case and adsorption in the
other. If desorption is started with a full pore space and unsaturated gas
is passed over the surface with a partial pressure of adsorbate gas of Pa,
evaporation continues until the curvature attained in the upper capillary
corresponds to the equilibrium vapor pressure at the level of liquid indi-
cated. In adsorption, starting with an adsorbate-free soUd, adsorption
stops when the hquid level has reached the point indicated in b, corre-
sponding to the same curvature as in a and exerting a vapor pressure
equal to pa. In each case equilibrium is attained, in desorption with a
high moisture content and in ad-
sorption with a low moisture con-
tent as indicated by shaded portions
in the two figures. The situation
is exaggerated in the illustration.
With interconnecting pore spaces
open to the atmosphere the forces
produced by small capillaries will
cause liquid to flow from large open- w>' «««i>a»> (b) Adw»ptioD
ings into the finer capillaries, thus FIG. 24. Adsorption hysteresis.
emptymg the large pore spaces. A
discussion of these forces is given by Ceaglske and Hougen.^ Other
types of adsorption hysteresis are explained by Emmett and Dewitt.'
Preferential Adsorption. Preferential adsorption has already been
referred to as one of the methods of removing an adsorbate gas. In
Illustration 18 it was assumed that the presence of the gases from which
the benzene was adsorbed had no effect on the equilibrium between the
benzene vapor and the charcoal. This assumption is not necessarily
true because charcoal adsorbs considerable quantities of all the ordinary
gases as well as benzene vapors. It would be expected that when char-
coal is exposed to a mixture of gases a complicated equilibrium would be
reached between each of the gases and its adsorbed quantity. Few
quantitative data are available on the adsorption of mixtures of gases
and vapors, but it is apparent that when several gases are adsorbed the
presence of each must affect the equilibrium concentration of others.
It is an experimentally observed fact that, in general for van der
6 Trans. Am. Inst. Chem. Eng., 33, 283 (1937).
' J. Am. Chem. Soc., 65,1258 (1943).
Waals adsorption, a gas of high molecular weight, high critical tempera-

ture, and low volatility is adsorbed in prefeJ^nce to a gas of low molecu-
lar weight, low critical temperature, and high volatility. Such a pre-
ferentially adsorbed gas or vapor will displace other gases which have
already been adsorbed. The chemical nature of the gas also plays an
20 40 60 80
Per Cent Relative Humidity
FIG. 26. Equilibrium moisture content of various desiccants.
important part, but ordinarily it may be assumed that a heavy vapor of

low volatility will almost completely displace a light gas of high volatility
and similar chemical type. In the experiments of Coolidge it was found
that exposure of the outgassed charcoal to air before treatment with
benzene vapors had no apparent effect on the final equilibrium.
In the absence of definite data it may ordinarily be assumed that when
CHAP. V] PREFERENTIAL ADSORPTION 159
an adsorbent is treated with a vapor of low volatility such as water or
benzene in admixture with a very volatile gas such as air, the adsorption
of the gas will exert only a negligible influence on the normal equilibrium
between the vapor and the adsorbent. When a gaseous mixture having
several components of similar volatiUty is treated with an adsorbent,
Temperature, "F
20 40 60 80 100 120 140 160 180 200 220 240 260 280
10.0 l..,,...l.., 11 , l...^l
8.0
1 1 1' 1 1 1 1 1' ll/
A/ y\ i/' 1 / i 1 11-
6.0
' 1 ' ' ' '' '' X '
^j/f !/] 1 i/i 1 ! ' 11
4.0
! i ! i 'AXo x1 'X ' 1/ 1 1/ 11 1I' 1' 1l/K
1 1 1 , Jr X y
! 'i
!/
2.0
' mvyy l/i ' ' 1 1 ' y i ' 1'
I 1.0
S 0.8
I 0.6
/ \ / \ \/\ \ \ 1 * • ' ' 1 1 ' 1 111
0.4
- i 1 ^f%C
0.2
0.1 !/ ' '<?•/'' 1 '

'I 1 <?/1 1 ; 1 1 ; : i ; ; ; i; i
J» .06 ' i ! 1-oy ' ' t i ( ' 1 f ' I ' ' ( ^
1 1 1 ' ' 1" ] 111 f

i >yf : ' ' ' ;
' 1 ' ' ' 1 1 ' ' * 1 *
1 1 ' 1 1 ' 1 1
1 1 : 1 1 : ! ; 1 1 li i 1 1 ii t 11 I I I ;
.01 \A \ ..II 11 / L _ 1 II 1 1 1 1 II II
0.1 0.2 0.4 0.6 1.0 2 420 640 60 10 100

Vapor Pressure of Water, Inches of Hg
s = slope ol line.
F I G . 26. V a p o r p r e s s u r e of w a t e r a d s o r b e d o n silica gel. -
equilibrium quantities of these components are adsorbed, and general

predictions of the equilibrium conditions cannot be made without
specific data.
The preferential adsorptive properties which many substances exhibit
are of great industrial importance in the selective separation of com-
ponents of gaseous mixtures. An important application of adsorbents
is in the drying of gases. In Fig. 25 are the equihbrium moisture con-
tents of silica gel and activated alumina compared with various solic' and
liquid chemical desiccants. It will be observed that the solid desiccants
compare unfavorably with the liquid desiccants in adsorption capacity.
However, both silica gel and activated alumina are capable of drying air
to much lower dew points than homogeneous solutions of hthium chloride
or calcium chloride.
Cox Chart for Adsorbents. A useful way of presenting the vapor-
pressure relations of adsorbed water is by means of the Cox chart wherein
for the same temperature the vapor pressure of adsorbed water is plotted
against the vapor pressure of pure water for lines of constant moisture
content. Straight lines generally result on a logarithmic plot, as shown
in Fig. 26 for silica gel. The relative humidity at any temperature and
composition is obtained directly from the ratio p/ps, where Ps is the vapor
pressure of pure water. This chart is also useful in estimating heats of
adsorption as discussed in Chapter VIII.
It should be noted that all the discussion in this chapter is restricted
to equihbrium conditions. Calculation of rates of sorption requires
consideration of particle size, diffusion, rates of gas flow and other
factors.
/ - • ;>i
PROBLEMS
1. From the International Critical Tables, Vol. IV, plot a curve relating the sol-
ubility of sodium carbonate in water to temperature. Plot solubilities as ordinates,
expressed in percentage by weights of NasCOs, and temperature as abscissas, ex-
pressed in degrees Centigrade, up to 60°C. On the same axes plot the freezing-point
curve of the solution, or the solubility curve of ice in sodium carbonate, locating the
eutectic point of the system.
2. From the data of Figs. 10, 11, and 12 and Problem 1, tabulate in order the suc-
cessive effects produced by the following changes:
(a) A solution of naphthalene in benzene is cooled from 20° to — 10°C. The
solution contains 0.6 gram-mole of naphthalene per 1000 grams of ben-
zene.
(6) A solution of sodium carbonate in water is cooled from 40° to ~5°C. The
solution contains 4.6 gram-moles of NaaCOa per 1000 grams of water.
(c) A mixture of aqueous sodium sulphate solution and crystals is heated from
10° to 60°C. The original mixture contains 3.3 gram-moles of NazSOi,
per 1000 grams of water.
(d) Pure crystals of Na2SO4-10H2O are heated from 20° to 40°C.
(e) A solution of FeCh in water is cooled from 20° to — 60°C. The solution
contains 2.5 gram-moles of FeCU per 1000 grams of water.
(/) A solution of FeCU is evaporated at a temperature of 34°C. The original
solution contains 5 gram-moles of FeCls per 1000 grams of water, and it is
evaporated to a concentration of 25 gram-moles of FeCh per lOOQ grams
of water.
(g) An aqueous solution of FeCls is cooled from 45° to 10°C. The solution con-
tains 18 gram-moles of FeCls per 1000 grams of water. '•
CHAP. V] PROBLEMS 161
3. In a solution of naphthalene in benzene the mole fraction of naphthalene is 0.12.

Calculate the weight of this solution necessary to dissolve 100 lb of naphthalene at
a temperature of 40°C.
4. An aqueous solution of sodium carbonate contains 5 grams of NajCOs per 100 cc
of solution at 15°C. Calculate the pounds of anhydrous Na2C03 which can be dis-
solved in 10 gal of this solution at 50°C.
5. A solution of sodium carbonate in water is saturated at a temperature of 10°C.
Calculate the weight of Na2CO3-10H2O crystals which can be dissolved in 100 lb
of this solution at 30 °C.
6. A solution of naphthalene in benzene contains 8.7 lb-moles of naphthalene per
1000 lb of benzene.
(a) Calculate the temperature to which this solution must be cooled in order
to crystallize 70% of the naphthalene.
(b) Calculate the composition of the solid product if 90% of the naphthalene is
crystallized.
7. The concentration of naphthalene in a solution in benzene is 1.4 lb-moles per 1000
lb of benzene. This solution is cooled to — 3°C. Calculate the weight and com-
position of the material crystallized from 100 lb of the original solution.
8. A solution of naphthalene in benzene contains 25% naphthalene by weight.
Calculate the weight of benzene which must be evaporated from 100 lb of this solu-
tion in order that 85% of the naphthalene may be crystallized by cooling to 20°C.
9. A batch of saturated NajCOa solution, weighing 1000 lb, is to be prepared at
50°C.
(a) If the monohydrate (Na2C03-H20) is available as the source of Na2C03,
how many lb of this material and how many pounds of water would be
needed to form the required quantity of solution?
(6) If the decahydrate (Na2CO3-10H2O) is available as the source of NajCOs,
how many pounds of this material and how many pounds of water
would be required? By means of a sketch, show how the solubility
chart was used in solving the problem.
10. A system containing 40% Na2C03 and 60% H2O has a total weight of 500 lb.
At 50°C, 35°C and 20°C, report on:
(a) The nature and composition of the phases constituting the system.
(6) The weight of each phase.
By means of a sketch, indicate how the solubility chart was used in solving the
problem.
11. A solution containing 35% Na2C03 weighs 5000 lb.
(o) To what temperature must the system be cooled in order to recover 98% of
the NazCOs?
,(b) What will be the weight of the crystals recovered, and of the residual
mother liquor?
problem.
12. In Fig. 27 the temperature-composition diagram of hthium chloride is pre-
sented, showing the composition of the system as pounds of water per pound of
Uthium chloride. Explain the significance of each separate area, boundary hne. and
point of intersection.
13. A dilute solution of Na2C03 and water weighs 2000 lb, and contains 5%
Na2C03. It is required that 95% of the NaaCOa be recovered as the decahydrate.
The lowest temperature that can be obtained is 5°C. Specify tRe process to be
employed for securing the degree of recovery required. Present calculations with
respect to the quantitative relations involved in the process specified.
problem.
280 T
-
4
4
240
-
; h
200
1
•
^ \
-
\
s
^120
- -
3
80
"
r v •
40 o- o
X
2
1 i 1 e °\.l. . -
0 0.5 1.0 1.5 ,2.0
, Pounds HjO per Pound Li CI
FIG. 27. Solubility of lithium chloride.
14. An aqueous solution of ferric chloride contains 12 lb-moles of FeCU per 1000
lb of water. Calculate the yield of crystals formed by cooling 1000 lb of this solu-
tion to 28''C.
15. A solution of ferric chloride in water contains 15 gram-moles of FeCla per 1000
grams of water.
(a) Calculate the composition of the resulting crystals in percentage of each
hydrate formed when this soJution is cooled to 0°C.
(b) Calculate the percentage of euteotic crystals present in the total crystal
16. An aqueous solution contains 3.44% Na2S04 and 21.0% Na2C03 by weight.
It is desired to cool this solution to the temperature which wiU produce a maximum
yield of crystals of pure sodium carbonate decahydrate. Calculate the proper
final temperature and the yield of crystals per 100 lb of original solution, using the
data of Fig. 14.
17. It is desired to crystallize a maximum amount of pure Na2CO3-10H2O from
the solution of Problem 11 by evaporating water at a temperature of 25°C. Calcu-
late the quantity of water which must be evaporated and the yield of crystals pro-
duced per 100 lb of original solution.
18. A solution contains 25 grams of Na2S04 and 4.0 grams of NaaCOs per 100
grams of water. From 100 lb of this solution 20 lb of water is evaporated and the
residual solution cooled to 20°C. Calciilate the weight and composition of the
crystals formed in the process.
19. Calculate the weight of water which must be evaporated from 100 lb of the
solution of Problem 18 in order to crystallize 70% of the Na2S04 as the pure decahy-
drate at a temperature of 15°C.
20. A solution has the following initial composition:
Na2SO4-10H2O = 80 parts by weight
Na2CO3-10H2O = 60 parts by weight
Solvent water = 100 parts by weight
The solution is at 50°C, and weighs 2500 lb. If this solution is cooled to 5°C:
(a) At what temperature does crystallization start? What phase crystallizes
out first?
(b) At what temperature will a second crystalline phase start to separate?
What is its composition?
(c) Calculate the maximum weight of pure crystals that can be obtained in the
first stage of crystallization, when but one solid phase is separating.
(d) Calculate the respective weights of the two sohd phases that separate
during the second stage of crystallization.
(e) With respect to the residual mother liquid at 5°C, report its total weight,
and the weight of each of the three components that were present in the
original solution.
By means of a sketch, show how Fig. 14 was used in solving the problem.
The solution originally weighs 1500 lb. If evaporation is carried out at 25°C:
(o) How much water must be evaporated before crystallization starts?
(6) What is the first solid to crystallize from the solution? How much of this
pure crystalline material can be removed before a second solid starts to
separate?
(c) How much water must be removed by evaporation before a second crys-
talUne phase starts to form? What is the nature of this second crys-
taUine phase?
By means of a sketch, show how Fig. 14 was used in solving the problem.
The solution weighs 2500 lb. It is required that 80% of the NaaCOs-lOHzO be
recovered in pure form. The lowest temperature that can be obtained is 15°C.
How many pounds of water must be removed by evaporation if 80% recovery is ob-
tained on cooling to 15°C?
By means of a sketch, show how Fig. 14 was used in solving the problem. i;.i
23. A solution has the following initial composition: >' '^ isif,
NajSOi-lOHjO = 80 parts by weight ' -?^
'•' NajCOj-lOHjO = 10 parts by weight ';.':.•'y
The solution weighs 1600 lb. Compare the maximum yield of NajSOj-lOHjO
obtainable by each of the following three processes:
(o) Cooling to 10°C. • •• '" : : •• « '••
(b) Evaporation at a constant temperature of 25°C.
(c) Evaporation of a limited amount of water at 25°C, followed by cooling to
io°c. ^^,
By means of a sketch, indicate how Fig. 14 was used in solving the problem.
24. The residual liquor from a crystallizing operation has the following composi-
tion :
, , Na2CO3-10H2O = 60 parts by weight , .
• Solvent water = 100 parts by weight
This liquor is to be used to extract the soluble material from a powdered mass that
has the following composition: ' -
Na2S04 = 1.25% ' , • »•"'
« Na2C03 = 10.50% "•;
Insoluble matter = 88.25%
(a) What is the minimum weight of the residual liquor required to dissolve the
soluble matter present in 1000 lb of the solid mass, assuming that the
leaching operation is carried out at 25°C?
(6) If the hquor obtained from the leaching operation is subsequently cooled,
what is the maximum possible percentage recovery of pure NajCOs?
By means of a sketch, show how Fig. 14 was used in connection with the solution
of this problem.
25. From the data of International Critical Tables plot a solubihty chart similar
to Fig. 14 for the system NaNOa—NaCl—^H20. Include the solubility isotherms of
15.6°, 60°, and 100°C.
26. A mixture of NaNOs and NaCl is leached with water at 100°C to form a
solution which is saturated with both salts. From the chart of Problem 25, calculate
the weight and composition of the crystals formed by cooUng 100 lb of this solution
to 15.6°C.
27. A solution of picric acid in benzene contains 30 grams of picric acid per liter.
Calculate the quantity of water with which 1 gallon of this solution at 18°C must be
shaken in order to reduce the picric acid concentration to 4.0 grams per liter in the
benzene phaê.
28. One gallon of an aqueous solution of picric acid containing 0.15 lb of picric acid
is shaken with 2 gallons of benzene. Calculate the poimds of picric acid in each solute
after the treatment. . ,,F ,; >«
29. A mixture of phenol and water contains 45% phenol by weight. Calculate
the weight fractions of upper and lower layers formed by this mixture at a tempera-
ture of 45°C.
30. Calculate the weight fractions and compositions of the upper and lower layers
of a mixture of 30% water, 20% isopropyl alcohol, and 50% tetrachlorethylene at a
31. Assuming the applicability of Henry's law, calculate the percentage CO2 by
weight which may be dissolved in water at a temperature of 20 °C in contact with gas
in which the partial pressure of CO2 is 450 mm of Hg.
32. Assuming the applicability of Henry's law, calculate the partial pressure of
HaS above an aqueous solution at 30°C which contains 3.0 grams of H2S per 1000
grams of water.
33. Calculate the volume in cubic feet of NH3 gas under a pressure of 1 atmosphere
and at a temperature of 20°C which can be dissolved in 1 gal of water at the same
temperature.
34. An aqueous solution of ammonia at 10°C is in equilibrium with ammonia gas
having a partial pressure of 500 mm of Hg.
(a) Calculate the percentage ammonia, by weight, in the solution.
(b) Calculate the partial pressure of the ammonia in this solution if it were
warmed to a temperatiKe of 40°C.
36. Assuming that equilibrium quantities of adsorbate are determined only by the
relative saturation of the adsorbate vapor, estimate the quantity of SO2 adsorbed by
the silica gel of Fig. 22 at a temperature of 10°C from a gas mixture in which the
partial pressure of SO2 is 150 mm of Hg.
36. Activated charcoal similar to that of Fig. 23 is to be used for the removal of
benzene vapors from a mixture of gases at 20°C and a pressure of 1 atmosphere.
The relative saturation of the gases with benzene is 83%.
(a) Calculate the maximum weight of benzene which may be adsorbed per
pound of charcoal.
(6) The adsorbed benzene is to be removed by stripping with superheated steam
at a temperature of 180°C. Calculate the final partial pressiu-e to which
the benzene in the steam leaving the stripper must be reduced in order to
remove 90% of the adsorbed benzene.
(c) If the adsorbent is so used that the benzene-bearing gases always come into
equilibrium with freshly stripped charcoal of part b before leaving the
process, calculate the loss of benzene in these treated gases, expressed as
percentage of the total benzene entering the process.
37. Fifty pounds of unsized cotton cloth containing 20% total moisture are hung
in a room of 4000-cu-ft capacity. The initial air is at a temperature of 100°F,
at a relative humidity of 20%, and a barometric pressure of 29.92 in. of mercury.
The air is kept at 100°F with no fresh air admitted and no air vented. Neglect
the space occupied by the contents of room.
(a) Calculate the moistvu-e content of the cloth and the relative humidity of the
air at equilibrium.
(6) Calculate the equilibrium moisture content of the cloth and the correspond-
ing relative humidity of the air if 100 lb of wet cloth instead of 50 lb are
hung in the same room.
(c) What is the final pressure in the room under part (a)?
38. Air at atmospheric pressure is to be dried at 80°F from 70% to 10% relative
humidity by mixing with silica gel, initially dry, and at 80°F. What is the final
equilibrium moisture content of the gel, assuming'a constant temperature of 80°F?
39. If air were to be dried to a dew point of 0°F what would be the equilibrium
water contents of the following desiccants assuming exit temperature of air and
desiccant to be 70°F.? • :
activated alumina :
silica gel J . ! :
glycerin . ,
sulfuric acid
calcium chloride solution (avoid crystallization) v ,
lithium chloride solution (avoid crystallization) , ' ... . r
triethylene glycol ' .:•;
40. How would it be possible to dry air to a dew point 0°F with a homogeneous
LiCl solution?
CHAPTER VI
MATERIAL BALANCES
A material balance of an industrial process is an exact accounting of

all the materials which enter, leave, accumulate, or are depleted in the
course of a given time interval of operation. The material balance is
thus an expression of the law of conservation of mass in accounting
terms. If direct measurements were made of the weight and com-
position of each stream entering or leaving a process during a given time
interval and of the change in material inventory in the system during
that time interval, no calculations would be required. Usually this is
impracticable, and therefore calculations are indispensable.
The generaJ principle of material baknce caJcuktions is to establish
a number of independent equations equal to the number of unknowns of
composition and mass. For example, if two streams enter a process and
one stream leaves, with no change in inventory in the system during
the time interval, the mass and composition of each stream completely
establishes the material balance. For calculating the complete mate-
rial balance the greatest number of unknowns permissible is three,
selected among six possible items. Variations in solving the problem will
depend upon the particular items which are unknown, whether they be
of composition or mass, or of streams entering or leaving. The following
principles serve as guides to direct the course of calculations.
1. If no chemical reaction is involved, nothing is gained by establishing
material balances for the several chemical elements present. In such
processes, material balances should be based upon the chemical com-
pounds rather than elements, or of components of fixed composition even
if not pure chemical compounds.
2. If chemical reactions occur, it becomes necessary to develop mate-
rial balances based upon chemical elements, or upon radicals, compounds,
or substances which are not altered, decomposed, or formed in the
process.
3. For processes wherein no chemical reactions occur, use of weight
units such as grams or pounds is desirable. For processes in which
chemical reactions occur, it is desirable to utiUze the gram-mole or
pound-mole, or the gram-atom or pound-atom.
4. The number of unknown quantities to be calculated cannot exceed
the number of independent material balances available; otherwise, the
problem is indeterminate.
167
168 MATERIAL BALANCES . [CHAP. VI
5. If the number of independent material balance equations exceeds

the number of unknown weights that are to be computed, it becomes a
matter of judgment to determine which of the equations should be
selected to solve the problem. If all the analytical data used in setting
up the equations were perfect it would be inunaterial which equations
would be selected for use. However, analytical data are never free from
error, and a certain amount of discretion is needed to select the most
nearly accurate equations for solving the problem. In general, equa-
tions based upon components forming the largest percentage of the total
mass are most dependable.
6. Recognition of the maximum number of truly independent equa-
tions is important. Any material balance equation that can be derived
from other equations written for the process cannot be regarded as an ad-
ditional independent equation. For example, in the following Illustra-
tion 1 it would be possible to write material balance equations based
upon water, IINO3, HCl, II2SO4, nitrogen, sulfur, hydrogen, oxygen,
chlorine, and overall weights. Of these ten equations only three are
independent. If the equation based upon H2SO4, HNO3 and overall
weights are selected, the other seven equations can be deduced from
computations alone.
7. If any two or more substances exist in 6xed ratio with respect to
one another in each stream where they appear, only one independent
material balance equation may be written with respect to these sub-
stances. Although a balance may be written for any one substance in
question, it is generally best to combine the substances appearing in
constant ratio into a single group and develop a single equation for this
combined group.
8. A substance which appears in but one incoming stream and one
outgoing stream serves as a reference for computations and is termed a
tie-substance. Knowledge of the percentage of a tie-substance in two
streams establishes the relationship between the weights of the streams
so that if one is known the other can be calculated.
Illustration 1. The waste acid from a nitrating proeess contains 23% HNO3,
57% H2SO4, and 20% H2O by weight. This acid is to be concentrated to contain
27% HNO3 and 60% H2SO4 by the addition of concentrated sulfuric aojd containing
93% H2SO4 and concentrated nitric acid containing 90% HNO3. Calculate the
weights of waste and concentrated acids which must be combined to obtain 1000 lb
of the desired mixture. '•""''• •''•
Basis: 1000 lb of final mixture
Let X = weight of waste acid
y = weight of cone. H2S04 . •
2 = weight of cone. HNO3 •
CHAP. VI] PROCESSES INVOLVING CHEMICAL REACTIONS 169
Overall balance: ' . ; .. • !'

x + y +z = 1000 (a)
HiSOt balance:
0.57a; + 0.93?/ = 1000 X 0.60 = 600 (b)
HNOs balance:
0.23x + 0.90Z = 1000 X 0.27 = 270 (c)
Equations (a), (b), and (c) may be solved simultaneously.
Eliminating 2 from (a) and (c): -^'"'H''
270 -0.23a; , , , ,
. ' ' 0.90 +y+- =' ^ ' ' -•«-
2/+ 0.744a; = 700 _ . . . • < (d)
Eliminating y from (d) and (b):
0.57x + 0.93(700 - 0.744a;) =600
0.122X = 5 1
I X = 4181b
From (b) \
600 - 0.57 X 418
= 3901b
" 0.93
From (a)
, . z = 1000 - 390 - 418 = 192 lb
These results may be verified by a material balance of the water in the process:
Water entering = (418 X 0.20) + (390 X 0.07) + (192 X 0.10) = 130 lb
Since the final solution contains 13% H2O, this result verifies the calculations.
More complicated problems may be handled by the same type of
analysis. When more than two simultaneous equations are to be
solved the use of determinants is recommended.
Processes Involving Chemical Reactions. Material balances of
processes involving chemical reactions are of the following two general
classes:
a. The compositions and weight of the various streams entering the
process are known. It is required to calculate the compositions and
weights of the streams leaving the process for a specified degree of com-
pletion of the reaction.
b. The compositions and weights of the entering streams are partially
known. It is required to calculate the compositions and weights of all
entering and leaving streams and to determine the degree of completion
of the reaction.
In these calculations it is desirable to work with molal rather than
ordinary weight units, particularly for components undergoing chemical
170 MATERIAL BALANCES [CHAP. VI
transformation. The limiting reactant should be selected. The quan-

tity of each reacting material may then be specified in terms of the per-
centage excess it forms of that theoretically required. The calculation
is then completed on the basis of the limiting reactant which is present
in a unit quantity of the reactants. The amounts of the new products
formed in the reaction are determined from the degree of completion.
The unconsumed reactants and inert materials pass into the product
unchanged.
Illustration 2. A producer gas made from coke has the following composition by
volume: '
CO 28.0%
CO2 3.5 * ., ' i^?
O2 0.5 . |P"w
• Ns 68.0 ^ ^
100.0%
This gas is burned with such a quantity of air that the oxygen from the air is 20%
in excess of the net oxygen required for complete combustion. If the combustion
is 98% complete, calculate the weight and composition in volumetric per cent of the
gaseous products formed per 100 lb of gas burned.
Discussion. The carbon monoxide is the limiting reactant, while the oxygen is
the excess reactant. The amount of oxygen supphed by the air is expressed as the
percentage in excess of the net oxygen demand, this latter term referring to the total
oxygen required for complete combustion, minus that present in the fuel. Since the
composition of the fuel is known on a molal basis, it is most convenient to choose
100 lb-moles of the fuel gas as the basis of calculation, and at the close of the solution
to convert the results over to the basis of 100 lb ot gas burned. „
Basts of Calculation: 100 lb-moles of producer gas.
Constituent Molecular Weight Mole, % Weight in Pounds
CO 28.0 28.0 2 8 . 0 X 2 8 . 0 = 7841b "^ f
COi 44.0 3.5 3 . 5 X 4 4 . 0 = 1541b
O2 32.0 0.5 0 . 5 X 3 2 . 0 = 161b
N2 28.2 68.0 68.0 X 28.2 = 19171b
100.0 28721b
Oxygen Balance:
O2 required to combine with all CO present 1/2 X 28.0 = . . 14.0 lb-moles
O2 in the producer gas 0.5 lb-mole
Net O2 demand = 14.0 - 0.5 13.5 lb-moles
O2 supphed by air = 13.5 X 1.20 16.2 lb-moles
1^ O2 actually used = 0.98 X 28.0 X 1/2 13.7 lb-moles
fi : O2 in products = 16.2 + 0.5 - 13.7 3.0 lb-moles
or 3.0 X 32 96.0 lb
CHAP. VI] PROCESSES INVOLVING CHEMICAL REACTIONS 171
Carbon Balance
C in fuel gas = 28.0 + 3.5 31.5 lb-atoms
C in CO of products of combustion = 0.02 X 28.0 0.56 lb-atom
C in CO2 of products of combustion = 31.5 — 0.56 30.94 lb-atoms
Carbon monoxide in products 0.56 lb-mole
or 0.56 X 28 15.7 lb
Carbon dioxide in products , 30.94 lb-moles
or 30.94 X 44 1359 lb
Nitrogen Balance
N2 in producer gas 68.0 lb-moles
N2 from air = 79/21 X 16.2 60.9 lb-moles
Nz in products = 68.0 -|- 60.9 128.9 lb-moles
or 128.9 X 28.2 3637 lb
Weight of Products of Combustion
Total weight, based on 100 lb-moles producer gas
96 + 1359 -f 16 + 3637 5108 lb
Total weight, based on 100 lb producer gas
5108 X (100/2872) 1781b
Analysis of Products of Combustion
Total moles in products of combustion ' i
30.9 + 3.0 + 0.56 + 128.9 163.4 lb-moles
Mole or
Volumetric
%
CO2 = (30.9/163.4) X 100 18.92
O2 = (3.0/163.4) X 100 1.84
CO = (0.56/163.4) X 100 0.34
N2 = (128.9/163.4) X 100 78.9
100.00
Illustration 3. A solution of sodium carbonate is causticized by the addition of
partly slaked commercial lime. The lime contains only calcium carbonate as an
impurity and a small amount of free caustic soda in the original solution. The mass
obtained from the causticization has the following analysis:
CaCOa 13.48%
Ca(0H)2 0.28
Na2C03 0.61
NaOH • 10.36
H20 75.27
100.00%
The following items are desired:
(a) The weight of lime charged per 100 lb of the causticized mass, and the compo-
sition of the lime.
(b) The weight of the alkahne liquor charged per 100 lb of the causticized mass,
and the composition of the alkaline liquor. ..
(c) The reacting material which is present in excess, and its percentage excess.
(d) The degree of completion of the reaction.
Discussion. The problem as stated cannot be solved before additional data are
obt^ned. The needed additional information is either the analysis of the lime or the
analysis of the alkaline liquor.
If the analysis of the lime were determined, the problem could be solved by the
following steps: (1) By using calcium as a tie-substance the weight of hme would
be determined. (S) An overall material balance would estabhsh the weight of alka-
line hquor. (3) By a carbon balance, the weight of Na2C03 in the alkaline liquor
would be computed. (4) By using sodium as a tie-substance the weight of
NaOH in the alkaline liquor would be calculated. (6) The weight of water in the
alkahne hquor would be determined by difference.
If the analysis of the alkahne hquor were determined instead, the problem would
be solved according to the following procedure: (1) By using sodium as a tie-
substance the total weight of the alkaline liquor would be calculated. (^) An over-
all material balance would establish the weight of lime. .{3) By means of a carbon
balance, the weight of CaCOa in the lime would be determined. (4) A calcium
balance would give the weight of active CaO [free CaO plus the CaO in the Ca(0H)2]
in the lime. (5) The weight of free CaO and the weight of Ca(0H)2 in the lime
could then be computed from the available values for the weight of lime, the weight
of CaCOs, and the weight of active CaO.
From the foregoing, it is apparent that it is only a matter of convenience whether
the Ume or the alkaline liquor is analyzed. The normal choice would be to analyze
the alkaline liquor, because the analysis is rapid and accurate.
An analysis of the alkaline liquor used in the process gave the following results:
NaOH 0,594%
NajCOa 14.88 %
H20 84.53 %
100.00 %
Basis of Calculation: 100 lb of the causticized mass.
Reactions for the Process: '
CaO-|-H20-^Ca(OH)2
NajCOa + Ca (OH) 2 -^ 2NaOH + CaCO,
Molecular Weights:
CaO = 56.1 NasCOs = 106.0
NaOH = 40.0 CaCO, = 100.1
Ca(0H)2 = 74.1 H2O = 18.02
Conversion into Molal Quantities
AlkalvM lAquor. Basis: 1 lb.
T, Tu n/r 1 Lb-Atoms Lb-Atoms
Lb Lb-Moles ,,T r^
Na C
NaOH 0.00594 0.000140 0.000149
NajCOj 0.1488 0.001404 0.002808 0.001404
H2O 0.8453
1.0000 0.002957 0.001404

CHAP. V I ] PROCESSES INVOLVING CHEMICAL REACTIONS 173
Caitsticized Mass. Basis: 100 lb

<v Tu Tu i\j 1 Lb-Atoms Lb-Atoms Lb-Atoms
% or Lb Lb-Moles ri »T ^
Ca Na C
CaCOa 13.48 0.1347 0.1347 0.1347
Ca(0H)2 0.28 0.00377 0.0038
Na2C03 0.61 0.00575 0.0115 0.00575
NaOH 10.36 0.2590 0.2590
HjO 75.27
100.00 0.1385 0.2705 0.1405

Sodium Balance '
Object: To evaluate weight of alkaline liquor.
Na in caustioized mass = 0.2705 lb-atom
Na in 1 lb alkaline liquor = 0.002957 lb-atom
Weight of alkaline liquor 0.2705/0.002957 = ... 91.50 lb
Overall Material Balance i

Object: To determine the total weight of lime.
Weight of causticized mass = 100.00 lb
Weight of alkaline liquor = 91.50 lb
Weight of lime = 100.00 - 91.50 = 8.50 lb
Carbon Balance
Object: To evaluate the weight of CaCOs in the lime.
C in causticized mass = 0.1405 lb-atom
C in NasCOs = 91.50 X 0.001404 = 0.1285 lb-atom
C in CaCOs = 0.1405 - 0.1285 = 0.0120 lb-atom
Weight of CaCOs = 0.0120 X 100.1 = 1.20 lb
Calcium Balance
Object- To determine the active CaO [that present in the Ca(0H)2 plus the free
CaO] in the lime.
Ca in the causticized mass = Ca in the lime = . 0.1385 lb-atom
Ca present as CaCOa in the lime (see carbon
balance) = 0.0120 lb-atom
Ca present in Ca(OH)2 and in free CaO =
0.1385 - 0.0120 = 0.1265 lb-atom
Overall Balance of Constituents in the Lime

Object: To determine free CaO and Ca(0H)2 in the lime
Total weight of hme charged = 8.50 lb
Weight of CaCOs (see carbon balance) = . .. . 1.20 lb
Weight of CaO + Ca(0H)2 = 8.50 - 1.20 = 7.30 lb
Weight of total active CaO = 0.1265 X 56.1 = 7.10 lb
H2O present in the Ca(0H)2 = 7.30 - 7.10 = 0.20 lb
Ca(0H)2 in lime = 0.20 X (74.1/18.02) = . . . 0.82 lb
Weight of free CaO = 7.30 - 0.82 = 6.48 lb
Results
(a) Weight of lime = 4'.... 8.50 lb
Analysis of lime
• Lb PerCent
CaCOs 1.20 14.1
Ca(0H)2 0.82 9.6
CaO 6 ^ 76.3
8.50 100.0
(6) Weight of alkaline liquor 91.45 lb
Analysis of alkaline liquor was determined experimentally.
(c) Determination of Excess Reactant.
Total active CaO = 0.1265 lb-mole
NajCOs in liquor = 91.50 X 0.001404 = 0.1285 lb-mole
Since, according to the reaction equation, 1 mole of NazCOa requires 1 mole of
active CaO, it is concluded that the NajCOa is present in excess, and that the active
CaO is the limiting reactant.
Excess of NajCOa = 0.1285 - 0.1265 = 0.0020 lb-atom
Per cent excess = (0.0020/0.1265) X 100 = . . 1.6%
(d) Degree of Completion
Ca(0H)2 in causticized mass = 0.00377 lb-mole
CaO -t- Ca(0H)2 in lime charged = 0.1265 lb-mole
Degree of completion of the reaction =
100 - (0.00377/0.1265) X 100 = 97.0%
In each of the preceding three illustrations, there was but one stream
of material emerging from .the process. In many processes, there-are
two or more emergent streams. For example, there may be an evolu-
tion of gas or vapor from the reacting mass, as in the calcination of
limestone, or there may be a separable residue which is removed from
the major product, as when a precipitate is separated from a liquid
solution. While the complexity of the problem tends to increase as the
number of streams of material involved in the process increases, the
general methods of solution are as indicated for the illustrations already
presented. Unknown weights are evaluated through the use of material
balances. Frequent use is made of tie-substances for determining
unknown stream weights. The following illustrative problem is typical
of the treatment applied to a process wherein there is more than one
stream of emergent material.
Illustration 4. The successive reactions in the manufacture of HCl from salt and
sulfuric acid may be represented by the following equations:
NaCl -I- H2SO4 -^ NaHSOi + HCl
NaCl + NaHSOi -» Na^SOi + HCl
In practice the salt is treated with aqueous sulfuric acid, containing 75% H2SO4,
in slight excess of the quantity required to combine with all the salt to form Na2S04.
Although the first reaction proceeds readily, strong heating is required for the second.
In both steps of the process HCl and water vapor are evolved from the reaction mass.
CHAP. VI] PROCESSES INVOLVING C H E M I C A L R E A C T I O N S 175
" Salt cake " prepared by such a process was found to have the following composi-
tion:
NajSO, 91.48%
NaHS04 4.79%
NaCI 1.98%
HÔ 1.35%
HCI 0.40%
100.00%
T h e salt used in t h e process is dry, and m a y be assumed to be 100% NaCl.
(o) Calculate the degree of completion of the first reaction and the degree of com-
pletion of the conversion to Na2S04 of the salt charged.
(6) On the basis of 1000 lb of salt charged, calculate the weight of acid added, the
weight of salt cake formed and the weight and composition of the gases driven off.
Discussion. The logical choice of a basis of calculation is 1000 lb of salt, since the
results are wanted on that basis.
The solution of part (6) is accomphshed through a series of material balances. I t
will be noted t h a t sodium serves as a tie-substance between the salt cake and the
salt charged, and that sulfur serves as a tie-substance between the salt cake and the
aqueous acid. Accordingly, the problem may be solved by the following successive
steps: (1) A sodium balance establishes the weight of salt cake. (2) A sulfur
balance serves to determine the weight of aqueous acid used. (3) A chlorine balance
estabhshes the weight of HCI driven off in the gases. (4) A water balance deter-
mines the weight of H2O in the gases driven off.
Basis of Calculation: 1000 lb salt charged.
Molecular Weights
NaCl = 58.5 HCI = 36.46
H2SO4 = 98.1 ... NaHSOi = 120.1
NasSOi = 142.1
Conversion into Molal Quantities
Salt. Basis: 1000 1b.
NaCl = 1000/58.5 17.08 lb-moles
Sulfuric Acid. Basis: 1 lb. -

H2SO4 = 0.75 lb, or 0.75/98.1 0.00764 lb-mole
Salt Cake. Basis: 1 lb.

T, Ti, A/r 1 Lb-Atoms Lb-Atoms Lb-Atoms
Lb Lb-Moles ^^ g ^j
0.9148
Na2S04 0.9148 = 0.00644 0.01288 0.00644
0.0479
NaHSOi 0.0479 = 0.000398 0.000398 0.000398
0.0198
NaCl 0.0198 T^TT" = 0.000338 0.000338 0.000338
00.0
H2O 0.0135
0.0040
HCI 0.0040 — - — = 0.000110 0.000110
36.46
1.0000 0.01362 0.00684 0,000448
Part (o) can be worked out by focusing attention on the salt cake. Since no H2SO4
is present in the product, the first reaction went to completion. • •
Total Na present 0.01362 lb-atom
NainNajSOi 0.01288 lb-atom
Conversion of NaCl to Na2S04 =
(0.01288/0.01362) X 100 94.5%
Sodium Balance
Object: To determine the weight of salt cake.
Na in 1000 lb salt charged 17.08 lb-atoms
Na in 1 lb salt cake 0.01362 lb-atom
Weight of salt cake = 17.08/0.01362 1253 lb
Sulfur Balance
Object: To determine the weight of aqueous acid used.
S in salt cake = 1253 X 0.00684 8.58 lb-atoms
S in 1 lb of aqueous acid 0.00764 lb-atom
Weight of aqueous acid = 8.58/0.00764 1123 lb
Water Balance , ',

' Object: To determine the weight of H2O in gas evolved.
Water in aqueous acid = 1123 X 0.25 281 lb ;
Water in salt cake = 1253 X 0.0135 17 lb
Water driven of! = 281 - 17 264 lb
Chlorine Balance
Object: To determine the weight of HCl in gas evolved. *•
CI in salt charged 17.O8 lb-atoms
CI in salt cake = 1253 X 0.000448 0.56 lb-atom
CI in HCl driven off = 17.08 - 0.56 16.52 lb-atoms
Weight HCl in gases = 16.52 X 36.46 603 lb
Composition of Leaving Gases
Lb Per Cent
HCl 603 69.6
HiO 264 39.4
867 100.0 .. ;, ,
Overall Balance
Object: To check the accuracy of the calculations. sit* (?
Total weight of reactants = 1000 -|- 1123 2123 lb
Total weight of products = 1253 -f 867 2120 lb
CHAP. VI] STEPWISE COUNTER CURRENT PROCESSING 177
STEPWISE COUNTER CURRENT PROCESSING

The leacMng of solids, the washing of precipitates, the extraction of
oils from crushed seeds, the softening of water, and the sorption of gases
are processes which are often carried out in stages by continuous or by
discontinuous methods. In a continuous counter current extraction or
washing process the precipitate to be washed or solid to be leached is
treated in an agitator with clear solution from tank II as shown in Fig. 28.
The slurry from the agitator is tr'ansferred to tank I where upon slow
agitation and settling the concentrated liquor is decanted off as the
desired product and the precipitate is pumped as a slurry by a sludge
pump to tank II where it is mixed with dilute solution from tank III.
Again upon slow agitation and settling the more concentrated liquid is
pumped into the agitator where it is mixed with the fresh solid. The
Insoluble Solid = m.
FIG. 28. Step-wise countercurreiit extraction.
sludge settling from tank II is pumped to tank III where it is treated

with fresh solvent. Again upon slow agitation and settling the clear
dilute liquid is transferred to tank II and the precipitate, now nearly free
from solute, is pumped out as a residual sludge. As an approximation
for estabUshing a material balance in this type of process it will be
assumed that uniform concentration of solution is attained in each tank,
that the solid pumped out of each tank as a sludge contains the same and
uniform amount of solvent per unit weight of dry soUd, and that the
solution-leaving each tank as clear Uquid is of the same concentration as
that retained by the solid in the sludge pumped out.
A solid containing rriz pounds of insoluble and mi pounds of soluble
material is fed continuously into the agitator and mixed with clear
liquid from tank II, while ms pounds of fresh solvent are fed into tank
III. The sludge pumped from each tank contains c pounds of solvent
per pound of insoluble solid. The concentration of soluble material in
the solution is expressed as pounds per pound of solvent, and is desig-
nated as X in tank I, y in tank II, and z in tank III.
A material balance of the solute for each tank is as follows:
Input Output
First Slurry Clear solution Sludge

tank ' - fill + ym% = (TOS — cm,-i)x + cmix (1)
Second Clear solvAion Sludge Clear solution Sludge

tank TO3Z + cm'iX = may + amy (2)
Third Clear solution Sludge Clear solution Sludge

tank 0 -f- CTO22/ = msz + cmiz (3)
Overall Clear solution Solid Clear solution product Sludge residue

balance 0 ^ TOi = (TOS — cm2)x + cm2Z (4)
The oversimplified assumptions made in this treatment should be

. recognized. A more general solution requires consideration of non-
uniformity of concentration in each tank for both sludge and solution.
Illustration 6. A dry black ash contains 1000 lb of sodium carbonate and 1000 lb
of insoluble sludge. The sodium carbonate is to be extracted from this ash with
10,000 lb of water using three thickeners in series with countercurrent flow of sludge
and water. The fresh water enters the third thickener, overflows to the second, and
is then passed to an agitator, where it is mixed with the black ash. The resultant
sludge from the agitator is passed to the first thickener. The sludge is pumped from
one thickener to the next and discharged as waste from the third thickener. The
sludge holds 3 lb of water for each pound of insoluble matter as it leaves each thick-
ener. The concentrated sodium carbonate is drawn off and recovered from the first
thickener. Calculate the weight of sodium carbonate recovered, assuming that all
sodium carbonate is entirely dissolved to form a uniform solution in each agitator.
Employing the symbols given
OT3 = fresh water supplied = 10,000 pounds s .,i.i
mz = insoluble solid supplied = 1,000 pounds *,
mi = sodium carbonate supplied = 1,000 pounds
c — water per pound of insoluble solid = 3 ;
Substituting these values in Equations (1), (2), (3), and (4) for a balance of sodium
carbonate, there results
First tank 1000 + 10,000?/ - 70002; + SOOOa '
Second tank 10,0002 + 3,000x = 10,000?/ + 30002/
Third tank 3,0002/ = 10,000z + 3000z
Overall balance 1,000 = 7,000x + SOOOz
CHAP. VI] INVENTORY CHANGES IN CONTINUOUS PROCESSES 179
Solving these simultaneous equations, '

x = 0.139
' y == 0.039
s = 0.009
The sodium carbonate recovered in the product is
7000x = (7000) (0.139) = 973 pounds
The sodium carbonate lost in the final sludge residue is
3000? = (3000) (0.009) = 27 pounds
INVENTORY CHANGES IN CONTINUOUS PROCESSES

In steady, continuous processes changes in the inventory, that is,
changes in the quantity of charge, intermediates, and final products con-
tained within the process itself are generally negligible. For the purpose
of design an average material balance is selected ignoring fluctuations in
inventory. However, such variations may become of great significance
in interpreting experimental data obtained during a short period of
operation. Problems of inventory fluctuations arise in preparing daily
performance reports of continuous operations and also in analyzing the
data from special test operations, especially those of short duration.
Inventory changes may result from variations in hquid or solid levels
in the various operating units of the plant, from changes in the quantities
of materials held in intermediate storage between successive units, and
from changes in compositions at various points in the process. Less
commonly inventory changes might be encountered as a result of changes
in the operating pressure. Particular attention must be given to the
separation of soUds from solutions causing a change in liquid level with
or without change in the inventory of the liquid.
Where a change in inventory occurs the material input to the plant is
equal to the output plus the net accumulation of inventory or minus the
net depletion in all parts of the process. A material balance may be
obtained either by adding accumulation items to the output or sub-
tracting them from the input. If the accumulation is an unchanged
charge material or a finished product it is directly subtracted from the
input or added to the output respectively.
If accumulation or depletion of an intermediate product takes place
the problem requires conversion of the quantity of accumulation into
the corresponding quantities of either charge or final product materials.
In order to make this conversion, a set of yield figures must be assumed
relating the intermediate product to the charge materials from which it
originated or to the products to be derived from it, to whichever it is
more similar. Where calculation of yields is the object of the material

balance it is necessary to assume approximate yields on which to base
these corrections. If the inventory changes are small, little error is
introduced by these assumptions. However, the calculation may be
corrected by means of a second approximation based on the yields calcu-
lated from the first assumptions.
Water Spray
mi=12.6 tons m2=n-7 tons ms tons
• ? —
\ / \y \y
Barometric
Condenser
s tons
— 'w^
. Fresh B r i n e ^ ^ ^
29 tons
200° i?
26* NaCl Condensate Condensate Condensate
mi-12.6 tons m2=11.7 tons
Condensate m^
and Condensing
EVAPORATOR I EVAPORATOR H EVAPORATOR III Water
FIG. 29. Inventory changes in triple-effect evaporation.
Illustration 6. A triple-effect, vertical-tube, submerged-type evaporator, Fig. 29,

is fed with 29 tons of salt brine solution during a given test period. The brine is fed
to the first evaporator at 200°F and contains 26% NaCl by weight. The water
evaporated from the first effect is condensed in the steam chest of the second effect
and weighed as 12.6 tons; the water evaporated in the second effect is 11.7 tons.
The salt crystallized during evaporation is allowed to collect in the conical hopper of
each evaporator. During the test period the liquid level was allowed to drop in each
evaporator. A drop of one inch corresponds to 12.1 cu ft of brine.
The following data were obtained:
Feed, 29 tons during period, 26% NaCl, and 200°F. : •
Evaporator I II III
Steam temperature in steam chest 250°F 214°F 173°F

Brine temperature 229°F 187°F 137°F
Level change (inches) 10" drop 12" drop 8" drop
Water evaporated 12.6 tons 11.7 tons
Solubility of NaCl,% by weight 29.0 27.9 27.0
Specific gravity of NaCl crystals 2.15
It is desired to calculate the tons of NaCl crystallized in each evaporator and the
tons of water evaporated from evaporator III.
CHAP. VI] INVENTORY CHANGES IN CONTINUOUS PROCESSES 181
Solution:
In each effect the brine solution is saturated, with a density which may be obtained
from Fig. 1, page 18. The depletion of brine in each evaporator is not only that
corresponding to the drop in liquid level but also that displaced by the salt crystal-
lizing out and accumulating in the hopper. For each ton of salt crystallizing out
-—- tons of brine are displaced, where 2.15 is the Specific gravity of the salt crystals
2.15
and G is that of the brine solution. For example, in the first evaporator,
G 1.140
= 0.53
2.15 2.150
and the total depletion of brine is
(10) (12.1) (62.4) (1.140)

+ 0.53x tons = 4.30 + 0.532
2000
where
J , a; = tons of salt crystallized in Evaporator I
: ! 62.4 = the density of water in pounds per cu ft
The results are tabulated as follows:
Evaporator I II III
Per cent NaCl by weight .29.0 27.9 27.0

Specific gravity of brine (Fig. 1) 1.140 1.145 1.150
Salt crystallized (tons) X y z
Drop in level, inches 10 12 8
Drop in level, tons of brine 4.30 5.18 3.46
Brine displaced by salt 0.53x 0.532?/ 0.5352
Salt in total brine depletion 1,25 + 0.164x 1.45 + 0.149y 0.935 + 0.1452
Water in total brine depletion 3.05 + 0.376X 3.73 + 0.3832/ 2.52 + 0.3902
Water evaporated (tons) 12.6 11.7 W3
From an overall water balance

29(0.74) + 3.05 + 0.376x + 3.73 + 0.3832/ + 2.62 + 0.3902 = 12.6 + 11.7 + ms (a)
or
TOa = 6.45 + 0.376a; + 0.383^ + 0.3902 (b)
From an overall salt balance

29(0.26) + 1.25 + 0.154X + 1.45 + 0.149i/ + 0.935 + 0.145z = » + y + 2 (c)
0.846a; + 0.851?/ + 0.8552 = 11.18 (d)
Let r = tons of brine leaving Evaporator I to Evaporator II

s = tons of brine leaving Evaporator II to Evaporator III
From salt and water balances for each evaporator the following are obtained:
Evaporator I if
Salt 1.25 + 0.154X + 0.26(29) = x + 0.29r (e)
Water 3.05 + 0.376a; + 0.74(29) = 12.6 + 0.71r
X = 3.91
m
1 1 7-= 18.88
Evaporator II
^t 1.45 + 0.149?/ + 0.29(18.88) = 2/ + 0.279s (g)
Water 3.73 + 0.3832/ + 0.71(18.88) = 11.7 + 0.721s (h)
y = 4.82
, s = 10.07
Evaporator III '
Salt 0.936 + 0.145 + 0.279(10.07) = z (i)
z = 4.38
From (b)
ms = 6.45 + 0.376a: + 0.383?/ + 0.390« = 11.47
a)
MATERIAL BALANCE
Input Tone Output Tons

(a) Bnne 29.00 (a) Water from I 12.60
Depletion Items (b) Water from II 11.70
(a) Brine in I i'' ' (c) Water from III 11.47
4.30 + 0.53a; = Accumulation Items
4 . 3 0 + (0.53) (3.91) == 6.37 (a) Salt in I, x 3.91
(b) Brine in II (b) Salt in II, y 4.82
5.18 - 0.532?/ = (c) Salt in III, 2 4.38
5.18 - (0.532) (4.82) = 7.75
(c) Brine in III
3.46 + 0.5352 = . - • . / • - • • • -
3.46+ (0.635) (4.38) 5.80

48.92 48.88
RECYCLING OPERATIONS
The recycling of fluid streams in chemical processing is a common
practice to increase yields or enrich a product. In fractionating columns
part of the distillate is refluxed through the column to enrich the prod-
uct. In ammonia synthesis the gas mixture leaving the converter after
recovery of ammonia is recycled through the converter. In the opera-
tion of dryers part of the exit air stream is recirculated to conserve heat.
CHAP. VI] RECYCLING OPERATIONS 183
Better wetting is provided in scrubbing towers by recirculating part of

the exit liquid. Recycling occurs in nearly every stage of petroleum
processing.
In a recycling operation the total or combined feed is made up of a mix-
ture of the fresh or net feed with the recycle stock. The gross products
of the operation are a mixture of the net products, which are withdrawn
from the system, and the recycle stock. In such an operation two types
of material balances are of interest. In the overall material balance the
net feed is equated against the net products. In a once-through material
balance the total feed is equated against the gross products.
Two methods are employed for expressing the yields of a recycling
operation. The net or ultimate yields are the percentages which the net
products form of the net feed. The once-through yields or yields per pass
are the percentages which the reaction products form of the total feed.
The ratio of recycle stock to fresh feed is termed the recycle ratio
and the ratio of total or combined feed to the fresh feed is termed the
combined or total feed ratio. It is evident that
Total feed ratio = Recycle ratio + 1 , .,
In dealing with operations in which all streams are subject to exact
analysis, design problems are generally approached from data obtained
in laboratory or pilot plant experiments of the once-through type, in
which the conditions resulting from recycling are simulated by merely
varying the composition of the total feed. However, when dealing with
extremely complex mixtures such as encountered in petroleum refining,
data from experiments in which recycling is actually carried out are
necessary as a design basis for the
establishment of ultimate yields and
the characteristics of the recycle
Hi
stock. In the former case the de- ' (i-y)r p-.r lb. air
(dry basis) a-y)r
sign problem is approached by first 1 lb. stock
setting up a once-through material "i (dry basis)*^
balance from which the recycle

quantities and ultimate yields are ^"'- ^°- ^^f^.^^^ôn of air
m arying.
derived. In the latter case both
the overall and the once-through material balances must be based di-
rectly on experimental data. Illustration 7 is typical of the first class
of problem while an example of the second class is shown in Chapter
X, page 421.
A common example of recycling is the recirculation of air in the
drying of solids, shown diagrammatically in Fig. 30. On the basis of
one pound of stock (dry basis) passing through the dryer, r pounds of air
(dry basis) are passed through countercurrently. If y is the fraction of

air (dry basis) recirculated, then (1 — ?/) is the fraction of fresh air
supplied (dry basis) and also the fraction of spent air rejected (dry
basis).
Accordingly, the following material balances result: • ; r.j;;;
Overall balance: i' ';'
^1 - W2*= r(l - y)iH, -Ho) . ;"^ ' p j ;
Once-through balance: '.*:;;
Wi - W2 = riH2 - Hi) ; ' " " *ig^'
where
Wi = moisture content of stock entering, pound per pound dry stock
Wi = moisture content of stock leaving, pound per pound dry stock
Hi = humidity of air leaving, pound per pound dry air
Hi = humidity of air entering dryer, pound per pound dry air ' •^''
Ho = humidity of fresh air supply, pound per pound dry air
From Equations (1) and (2) - '' ' ''^'-^f '"•''
' - • ' iH2-Ho)y = Hi-Ho ' '• ' • -"' (3)
The fraction of air to be recirculated depends upon the relative costs
of dryer, power, and heat, and is hence established by minimizing the
total costs. With increased recirculation for the same drying capacity,
a larger dryer is required with increased costs of equipment, power, and
radiation offset by a reduced cost of heat requirements.
Illustration 7. Propane is cataljîcally dehydrogenated to produce propylene by
quickly heating it to a temperature of 1000-1200°F and passing it over a granular
solid catalyst. As the reaction proceeds carbon is deposited on the catalyst, necessi-
tating its periodic reactivation by burning off the carbon with oxygen-bearing gases.
In a laboratory experiment in which pure propane is fed to the reactor, gases of the
following composition leave the reactor: . n
Gas Mole per cent . !.
Propane 44.5 • i;. ...
Propylene 21.3 ^, j " , ^^ ••;.!.
Hydrogen 25.4
Ethylene 0.3 •
Ethane 5.3
Methane 3.2
100.0
Based on these data it is desired to design a plant to produce 100,000 pounds per
day of propylene in a mixture of 98.8% purity. The flow diagram of the proposed
process is shown in Fig. 31. >.. .
Fresh propane feed, mixed with propane recycle stock is fed to a heater from
which it is discharged to a catalytic reactor operating under such conditions as to
produce the same conversion of propane as was obtained in the laboratory experiment.
The reactor effluent gases are cooled and compressed to a suitable pressure for separa-
tion of the light gases. This separation is accomplished by absorption of the propane
aiid propylene, together with small amounts of the lighter gases, in a cooled absorp-
tion oil which is circulated through an absorption tower. The " fat oil " from the
bottom of the absorber is pumped to a stripping tower where by application of heat
at the bottom of the tower the dissolved gases are distilled away from the oil which is
then cooled and recirculated to the absorber.
f^^ Propylene
FIG. 31. Flow chart of propane dehydrogenation plant.
The gases from the stripping tower are passed to a high pressure fractionating
tower which separates them into propane recycle stock as a bottoms product and
propylene and lighter gases as overhead. The following compositions are established
as preliminary design bases estimated from the vaporization characteristics of the
1. The light gases from the absorber are to contain 1.1 % propane and 0.7% propy-
lene by volume. Substantially all hydrogen, ethylene, and methane leaving the
cooler will appear in the light gases. The ethane, however, wiE appear in both the
light gases and the product.
2. The propane recycle stock is to contain 98% propane and 2% propylene by
weight.
3. The propylene product is to contain 98.8% propylene, 0.7% ethane, and 0.5%
propane by weight.
The total feed is passed over the catalyst at the rate of 4.1 pound-moles per cubic
foot of catalyst per hour. The density of the catalyst is 54 pounds per cubic foot.
It may be assumed that the small amount of propylene in the feed to the reactor
passes through unchanged. «.if'j >
Required:
(a) The amount of carbon formed on the catalyst, expressed as weight per cent
of the propane fed to the catalyst chamber. '
(6) The process period in minutes required in building a carbon deposit equal to
2% by weight of the catalyst. M
(c) An overall material balance.
(d) A once-through material balance.
(e) Ultimate yield of recovered propylene expressed as mole percentage of propane
in fresh feed.
(/) Once-through yield of propylene made in reactor expressed as mole percentage
of total propane entering reactor.
{g) Recycle ratio, weight units.
{h) Total feed ratio, weight units. ' .• ' i . \ ' , , i-ii.
*%
,
Preliminary Calculations . ^.; . ,—'" 1
Gas Formed by Decomposition of Propane ' .*'i;~.. !
Basis: 1 lb-mole of gas.
Lb-Atoms Lb-Atoms
Lb-Moles Mole Wt Lb C H
Propane CaHs 0.445 44.09 19.62 1.335 3.560
Propylene CsHe 0.213 42.08 . 8.96 0.639 1.278
Hydrogen Hj 0.254 2.016 0.512 • 0.508
Ethylene C2H4 0.003 28.05 0.084 0.006 0.012
Ethane C2HS 0.053 30.07 1.594 0.106 0.318
Methane CH4 0.032 16.04 0.513 0.032 0.128
1.000 31.28 2.118 5.804
8.000
Gas formed from 1 mole of propane (by hydrogen balance) = = 1.378
5.804
lb-moles.
Process Design Calculations
Basis of design: 1 hour of operation.
^ , . , , 100,000 ,,„„„ „
Production rate of pure propylene = ——— = 4167 Ib/hr
or = 99.03 Ib-moles/hr
42.08
Evaluation of Unknown Stream Weights ,
Let X = lb-moles recycle
y = lb-moles fresh feed
z = lb-moles light gases
Combined feed
Propane = y + 0.9791x lb-moles
Propylene = 0.02095a; lb-moles
Gases leaving reactor
Gases formed from the
propane entering = 1.378(2/ + 0.97913;)
= 1.378?/-F 1.349a; lb-moles
Propylene passing
through unchanged = 0.0209SX lb-moles
Total gas leaving = 1.378?/ + 1.370a;
Propane leaving = (1.3782/-|-1.349x) (0.445)
= 0.6132/ + 0.600a; lb-moles
Propylene leaving = 0.213(1.3782/ + 1.349x) + 0.02095a;
= 0.29352/ + 0.3083a; lb-moles
.:. MATERIAL BALANCES

(lb-moles)
Leaving Reactor Recycle Product Light Gases
99.03
Propane Balance 0.6132/ + 0.600X = 0.9791X + (0.00476) -l-O.Ollz
0.9855
Propylene Balance 0.29352/ + 0.3083X 0.02095X -t-99.03 + 0.007z
99.03
Total 1.3782/ + 1.370X +z
+ 0.9855
Simplifjdng the above three equations:
-0.379X + 0.613?/ 0.011? = 0.478
0.2873X + 0.29352/ - 0.0072 = 99.03
0.370X + 1.3782/ T « = ^^-^
Solving:
X = 210.8 lb-moles recycle
2/= 134.2 lb-moles fresh feed - ,,, ,
z = 162.9 lb-moles Ught gases
Gases Leaving Realtor
Formed from propane = (1.378) (134.2) + (1.349) (210.8) = 469.3
Propylene from recycle = (0.02095) (210.8) = 4.42
Lb-Moles Mole Wt Lb
C3I18 = (0.445) (469.3) = 208.8 44.09 9,206
CsHe = (0.213) (469.3) + 4 . 4 = 104.4 42.08 4,393
H2 = (0.254) (469.3) = 119.3 2.016 240
C2H4 = (0.003) (469.3) 1.4 28.05 39.2
CgHg = (0.053) (469.3) 24.9 30.07 749
CH« = (0.032) (469.3) 15.0 16.04 241
473.8 14,868
lVrnlp/»nln.T Tvp.ic,M, _ 1^'S68 _ ^R1.38

473.8
Light Gases , ,,.««..„. ,.

^<!i ;..•;:• Lb-Moles Mole Wt Lb
CsHs (0.011) (162.9) • 1 ' , 1.792 44.09 79.0
CsHe (0.007) (162.9) 1.140 42.08 48.0
Hz (from reactor) 119.3 2.016 240.6
CjHj (from reactor) 1.4 28.05 39.2
CHi (from reactor) 15.0 16.04 240.6
C2H6 Leaving reactor 24.9
In product 0.98 •
In light gases 23.9 23.9 30.07 718.7
162.5 1,366.0
. • !,.i-\
Moleciilar weightt == = 8 .41

162.5
Recycle
Lb-Moles Mole Wt Lb
CjHs (0.9791) (210.8) 206.4 44.09 9,100
C3H6 (0.02095) (210.8) 4.42 42.08 186
210.8 ,; - 9,286
9,286
Molecular weight = r r r r = 44.05
210.0
Product
C3H, M229 =4,167 lb •

24 /•
CsHs (0.007) (4,217) = 29.5 lb
CjHe (0.005) (4,217) = 21.11b
Fresh Feed
(134.2) (44.09) j . \ =5,917 lb
Carbon Deposit . , •
Propane to reactor 134.2 + 206.4 = 340.6 lb-moles
C in propane = (3) (340.6) (12) = 12,262 lb
C in gases formed from propane = (1.378) (340.6) (2.118) (12) = 11,929 lb
C deposited on catalyst = 333 lb

Required Results:
(o) Amount of carbon deposited on catalyst, as weight per cent of the propane
fed to the reactor.
333
Caxbon deposit = ^ ; ^ j ^ ^ P ^ = 2.22%
CHAP. VI] RE(:)YCLING OPERATIONS 189
(6) Process Period
Total feed = 210.8 + 134.2 = 345 lb moles/hr
Volume of catalyst 345.0 = 84.2 cu ft
4.1
Weight of catalyst = (84.2) (54) = 4,5471b
Carbon tolerance = (0.02) (4,647) = 90.91b
Rate of carbon deposition = = 3331b/hr
Process period
= 1'"' = 16.4 min
(c) Overall Material Balance ^ if-' ,

••'; '•>. i , , ' f-
Inpvi
Fresh feed 6,917 lb
Output ;,.,,
Light gases 1,366 lb
% Product 4,217 ••• M •
Carbon deposit 333
Total 5,916 5,916 lb

(d) Once-Through Material Balance '< 11
Input 1
V " ]
Fresh feed 5,917 lb

Recycle 9,286
Total 15,203 15,203 lb '

Outprd
\ Light gases 1,366 lb
Product 4,217
Carbon deposit 333 -
Recycle 9,286 ' '
Total 15,202 ,; :, , 15,202 lb

(e) Ultimate Yield of Propylene
i' Propylene in product '" 99.03 lb moles
Propane feed '-• 134.2 lb moles
99 03
Per cent yield = — ^ (100) = 73.8%
(/) Once-Through Yield .• ' ' •. ••

Propylene produced in Reactor = 99.03 -|- 1.14 =
100.2 lb moles
Propane entering reactor = 134.2 4- 206.4 = 340.6 lb moles
inn 9
Per cent yield = —f- (100) = 29.4%
340.6 ^
(g) Recycle Ratio

Weight recycle 9,2861b
Weight fresh feed 5,917 lb
9,286
Recycle ratio = ——-:: = 1.57
5,917
(ft) Total Feed Ratio
Weight recycle 9,286 lb •' '
Weight fresh feed 5,900 lb . '
Total Weight 15,186 1b " "*
Total feed ratio = —'- = 2.57
5,917
Accumulation of Inerts in Recycling. One limitation sometimes
encountered in the recycling of fluid streams is the gradual accumulation
of inerts or impurities in the recycled stock. Unless some provision is
made for removing such impurities they will gradually accumulate until
the process comes to a stop. This problem can be solved by bleeding off
a fraction of the recycle stock. The recycle stock bled off may be passed
through some special process to remove impurities and to recover the
useful components, or discarded Lf such recovery is too costly.
Where processes take place in the medium of a special solvent the
solvent may be circulated throughout the plant in a closed system. As
the product is withdrawn from the solvent some impurities remain in
solution and gradually accumulate. The accumulation is fixed at a
certain limiting concentration by continually withdrawing a definite
fraction of the recycled solvent. This type of control is maintained in
the electrolytic refining of copper wherein the electrolyte is continually
recirculated while a portion is bled off and replaced by fresh electrolyte.
The spent electrolyte is treated separately by a special process for the
recovery of residual copper, for the removal of accumulated impurities
such as nickel sulfate, and the purified sulfuric acid is returned to the
process. In the synthesis of ammonia from atmospheric nitrogen and
hydrogen the percentage conversion of a 1:3 mixture is 25% in a single
pass through the reactor. The ammonia formed is removed by cooling
under high pressure and the unconverted nitrogen and hydrogen are
recirculated to the reactor. The argon from the atmospheric nitrogen
is allowed to accumulate to a fixed upper limit by bleeding off a fraction
of the recycled gas or by leakage from the system.
Illustration 8. In the operation of a synthetic ammonia plant, shown diagram-

matically in Fig. 32, a 1:3 nitrogen-hydrogen mixture is fed to the converter resulting
in a 25% conversion to ammonia. The ammonia formed is separated by condensa-
tion and the unconverted gases recycled to the reactor. The initial nitrogen-hydro-
gen mixture contains 0.20 part of argon to 100 parts of N2-H2 mixture. The tolera-
CHAP. VI] ACCUMULATION OF INERTS IN RECYCLING 191
tion limit of argon in the reactor is assumed to be 5 parts to 100 parts of Nz and Hj
by volume. Estimate the fraction of recycle which must be continually purged.
Fresh Feed
lvww\l
Reactoi Condenser
100 lb. moles Nj+H.
0.2 lb. moles Argon
1
NHjUquid
i..- • Recycle i lb. moles Nj+Hj
i V • ^ y Bled stream
I -\ >>| :\ • ':_• y lb. moles Nj+Hj
Fia. 32. Purging of inerts in recycle stock, •
Basis of Calculations: 100 pound-moles N2-H2 in fresh feed.

Let X = moles N2 and H2 recycled to reactor
y = moles N2 and H2 bled off or purged
Moles N2 and H2 entering reactor = 100 + x
Moles Nj and H2 leaving reactor = 0.75(100 + x)
AT , f iMTi , A 0.25(100 + z)
Moles of NH3 formed =
Moles of argon in fresh feed = 0.20
Moles of argon in total feed = 0.05(100 -\- x)
Moles of argon per mole of N2-H2 mixture leaving condenser =
^ = 0.067
0.75 •
Moles of argon bled ofi = 0.067?/
When a steady state of operation is attained the argon purged is equal to the argon
in the fresh gas supply.
Hence
0.0672/ = 0.20
or
V = 2.98 , '
From an H2-N2" balance around bleed point
0.75(100 +a;) =x + V
Since 2/= 2.98; x = 288 pound-moles
Summary:
Fresh N2 and H2 = 100 pound-moles
Recycle N2 and H2 = 288 pound-moles
Purged N2 and H2 = 3.0 pound-moles
Ammonia formed = 48.5 pound-moles
Argon = 0.20 pound-mole
^ , . 288
2.88
Recycle ratio = ——
100
Purge ratio = — = 0.0104
In the actual operation of a high pressure plant the unavoidable leakage generally
is sufficient to keep argon in the reactor below the toleration limit so that no special
provision need be applied for venting part of the recycle to prevent accumulation.
1.0
/
/
^ J
o 1
SI
91 1 to 1 m 1
°1 « 1
"•3 1
0} 1 SI
u f a> f
•8/ "5/
(5/ <N 1 41 CO 1
to /
c 1
«o 1
Ho
/ j/ /
y 10
7 20
/
30
Pa = vapor pressure of water in atmosphere at temperature of bed

50 TO
w = pounds of dry air passed ttirough bed per square foot /
FIG. 33. Fraction of moisture removed from air by a bed of silica gel.
By-passing Streams. By-passing of a fluid stream by splitting it

into two parallel streams is often practiced where accurate control in
concentration is desired. For example, in conditioning air with a single
stationary bed of silica gel to a lower humidity a more nearly uniform
CHAP. VI] BY-PASSING STREAMS 193
humidity is obtained in the final air mixture if part of the air is dehumidi-
fied to a low moisture content and then mixed with the unconditioned
air to produce the desired mixture. The fraction of the entering humid-
ity remaining in the air after drying in beds of silica gel of various thick-
nesses is shown in Fig. 33 plotted against the product of p^w where
w — the total pounds of dry air passed through gel per
square foot of cross-section
Ps — the vapor pressure of water in atmospheres at
the temperature of the bed
W = GT = Upr :> ;
where '" ' ' '*• .; '
G = pounds of dry air per sq ft-min
T= time in minutes
p= density of entering air, pounds of dry air per
cu ft of mixture
u = velocity of entering air, feet per minute
It will be observed that the humidity of the air leaving the gel is by
no means constant. For an extended period of time, as indicated by the
abscissas, the air leaves with nearly no humidity, then it rises abruptly
as more air is passed through. The bed behaves as though a wedge of
water of the shape of the curves shown in Fig. 33 were advancing through
the bed; at first the advancing wedge is nearly imperceptible, then
suddenly the full wedge appears.
niustration 9. 1000 cu ft of air per minute at 8S°F, 70% relative humidity is to
be conditioned to a relative humidity of 30% and kept at 85°F, using a stationary
bed of silica gel 12 in. thick and 4 ft in diameter. To maintain the relative humidity
constant at 30%, part of the air is by-passed around the dryer and mixed with the
part which is dried. Constant humidity in the final air mixture is controlled auto-
matically by a damper regulating the fraction of air to be by-passed. This is accom-
plished by a constant wet-bulb controller. Assuming isothermal operation,
(a) Construct a plot showing the fraction of air to be by-passed against time.
(b) Calculate the time of operation before regeneration becomes necessary.
Area of bed = —- = 12.52 sq ft

4
1000 492
Total rate of air flow = -—- X •— X 0.9716 X 29 = 70.9 lb per min (dry
oOu o4o
basis)
70.9
Rate of air flow per sq ft of bed = ———: = 5.66 lb per min (dry basis)
12.52
ffo = initial humidity = 0.0182 p, = vapor pressure at 85°F = 0.0406 atm
Hi = desired humidity = 0.0077 partial pressure at 70% r.h. = 0.0284 atm
H = humidity leaving dryer partial pressure at 30% r.h. = 0.0122 atm
Let X = fraction of air by-passed

From a material balance for water M
0.0182X + H{\ -x) = 0.0077
0.0077 - H
(a)
0.0182 - H '
Let Aic = dry air passed through dryer in time interval AT
Aw
Then, = 5.66AT = total dry air delivered at 30% relative humidity.
The calculated results are tabulated at stated intervals of p^w, the

abscissas of Fig. 33. The corresponding values of HJHo are obtained
from Fig. 33 and tabulated in Column 2. From the value of ps = 0.0406
atm at 85°F, the values of w are tabulated in Column 3. The dry air
delivered for each interval of PsW is obtained as the difference of suc-
ceeding items in Column 3 and tabulated in Column 4 as Aw. The
value of the humidity of air H leaving the gel is obtained from Column 2
with /f 0 = 0.0182 and tabulated in Column 5. The average humidity
Ha during each interval is obtained from Fig. 33 as the mean value over
the interval of pjW indicated in Column 1. The average fraction x of
air by-passed during the interval Au; is obtained from the equation
0.0077 - Ha
and tabulated in Column 7. The total air delivered
0.0182 - Ha
1 2 3 4 5 6 7 8 9 10
H/Ho Aw Aw T
p.w (Fig. w inter- H inter- X AT
1 -X min
33) val val Aw
0 0 0 0 0
554.0 0 0.423 960 169.6
22.5 0 554 0 170
61.5 0.00032 0.413 105 18.6
25.0 0.035 616 0.000637 188
61,5 0,00128 0,379 99 17.5
27.5 0.105 677 0,00191 206
61.5 0,00291 0.313 90 15.9
30.0 0.215 739 0,00392 222
61,5 0,00519 0.193 76 13.4
32.5 0.355 800 0,00646 235
24.0 0.00708 0.055 25 4.4
33.5 0.423 824 0,0077 239
2 AT = 239 m in
CHAP. VI] PROBLEMS 195
Aw '
per interval Aw is equal to and is tabulated in Column 8. The
time interval AT corresponding to the air interval Aw is obtained by-
dividing items in Column 8 by 5.66; these are tabulated in Column 9.
The total time elapsed is obtained from a summation of items in column
9, T = SAT, and tabulated in Column 10.
A plot of X against T gives the desired schedule for by-passing air. For
the first 170 minutes, 42% of the air is by-passed, after which time the
by-pass is gradually decreased to no by-pass, according to schedule, until
a total of 239 minutes has elapsed. The silica gel must then be
regenerated.
PROBLEMS I
1. In the mamifacture of soda-ash by the LeBlanc process, sodium sulfate is heated

with charcoal and calcium carbonate. The resulting " black ash " has the following
composition:
NazCOs.....I/.......:....^. 42%
Other water-soluble material 6%
Insoluble material (charcoal, CaS, etc.) 62%
The black ash is treated with water to extract the sodium carbonate. The solid
residue from this treatment has the following composition:
NajCOa 4%
Other water-soluble salts 0.5%
Insoluble matter 85%
Water 10.5%
(o) Calculate the weight of residue remaining from the treatment of 1.0 ton of
black ash.
(6) Calculate the weight of sodium carbonate extracted per ton of black aah.
2. A contract is drawn up for the purchase of paper containing 6% moisture at a
price of 7 cents per pound. It is provided that if the moisture content varies from
5%, the price per pound shall be proportionately adjusted in order to keep the price
of the bone-dry paper constant. In addition, if the moisture content exceeds 5%,
the purchaser shall deduct from, the price paid the manufacturer the freight charges
incurred as a result of the excess moisture. If the freight rate is 90 cents per 100 lb,
calculate the price to be paid for 3 tons of paper containing 8% moisture.
3. A laundry can purchase soap containing 30% of water at a price of $7.00 per
100 lb f. o. b. the factory. The same manufacturer offers a soap containing 5% of
water. If the freight rate is 60 cents per 100 lb, what is the maximum price that the
laundry should pay the manufacturer for the soap containing 6% water?
4. The spent acid from a nitrating process contains 43% H2SO4, 36% HNOs,
and 21 % H2O by weight. This acid is to be strengthened by the addition of concen-
trated sulfuric acid containing 9 1 % H2SO4, and concentrated nitric acid containing
88% HNO3. The strengthened mixed acid is to contain 40% H2SO4 and 43% HNO3.
Calculate the quantities of spent and concentrated acids which should be mixed
together to yield 1000 lb of the desired mixed acid.
5. The waste acid from a nitrating process contains 21% HNO3, 55% H2SO4,
and 24% H2O by weight. This acid is to be concentrated to contain,28% HNO3
and 62% H2SO4 by the addition of concentrated sulfuric acid containing 93%
H2SO4 and concentrated nitric acid containing 90% HNO3. Calculate the weights
of waste and concentrated acids which must be combined to obtain 1000 lb of the
desired mixture.
6. In the manufacture of straw pulp for the production of a cheap straw-board
paper, a certain amount of lime is carried into the beaters with the cooked pulp. It
is proposed to neutralize this lime with commercial sulfuric acid, containing 77%
H2SO4 by weight.
In a beater containing 2500 gal of pulp it is found that there is a lime concentra-
tion equivalent to 0.45 gram of CaO per liter.
(a) Calculate the number of pound-moles of lime present in the beater.
(6) Calculate the number of pound-moles and pounds of H2SO4 which must be
added to the beater in order to provide an excess of 1.0% above that
required to neutralize the lime.
(c) Calculate the weight of commercial acid which must be added to the beater
for the conditions of 6.
(d) Calculate the weight of calcium sulfate formed in the beater.
7. Phosphorus is prepared by heating in the electric furnace a thoroughly mixed
mass of calcium phosphate, sand, and charcoal. It may be assumed that in a certain
charge the following conditions exist: the amount of silica used is 10% in excess of
that theoretically required to combine with the calcium to form the silicate; the
charcoal is present in 40% excess of that required to combine, as carbon monoxide,
with the oxygen which would accompany all the phosphorus as the pentoxide.
(a) Calculate the percentage composition of the original charge.
(6) Calculate the number of pounds of phosphorus obtained per 100 lb of charge,
assuming that the decomposition of the phosphate iby the silica is 90%
* complete and that the reduction of the liberated oxide of phosphorus, by
the carbon, is 70% complete.
8. A coal containing 81% total carbon and 6% unoxidized hydrogen is burned in
air.
(a) If air is used 30% in excess of that theoretically required, calculate the num-
ber of pounds of air used per pound of coal burned.
(b) Calculate the composition, by weight, of the gases leaving the furnace,
assuming complete combustion.
9. In the Deacon process for the manufacture of chlorine, a dry mixture of hydro-
chloric acid gas and air is passed over a heated catalyst which promotes oxidation of
the acid. Air is used in 30% excess of that theoretically required.
(o) Calculate the weight of air supplied per pound of acid.
(b) Calculate the composition, by weight, of the gas entering the reaction
chamber.
(c) Assuming that 60% of the acid is oxidized in the process, calculate the com-
position, by weight, of the gases leaving the chamber.
10. In order to obtain barium in a form which may be put into solution, the natural
sulfate, barytes, is fused with sodium carbonate, A quantity of bar3rtes, containing
only pure barium sulfate and infusible matter, is fused with an excess of pure, anhy-
drous soda ash. Upon analysis of the fusion mass it is found to contain 11.3%
barium sulfate, 27.7% sodium sulfate, and 20.35% sodium carbonate. The remain-
der is barium carbonate and infusible matter.
(a) Calculate the percentage completion of the conversion of the barium sul-
fate to the carbonate and the complete analysis of the fusion mass.
(&) Calculate the composition of the original barytes.
(c) Calculate the percentage excess in which the sodium carbonate was used
above the amount theoretically required for reaction with all the barium
sulfate.
11. In the manufacture of sulfuric acid by the contact process iron pyrites, FeSj,
is burned in dry air, the iron being oxidized to Fe203. The sulfur dioxide thus
formed is further oxidized to the trioxide by conducting the gases mixed with air over
a catalytic mass of platinum-black at a suitable temperature. It will be assumed
that in the operation sufficient air is supplied to the pyrites burner that the oxygen
shall be 40% in excess of that required if all the sulfur actually burned were oxidized
to the trioxide. Of the pyrites charged, 15% is lost by falling through the grate with
the " cinder " and is not burned.
(a) Calculate the weight of air to be used per 100 lb of pyrites charged.
(6) In the burner and a " contact shaft " connected with it, 40% of the sulfur
burned is oonvertfed to the trioxide. Cal«ilatethecompo&itioa,byweight,
of the gases leaving the contact shaft.
(c) By means of the platinum catalytic mass, 96% of the sulfur dioxide remain-
ing in the gases leaving the contact shaft is converted to the trioxide.
Calculate the total weight of SO3 formed per 100 lb of pyrites charged.
(d) Assuming that all gases from the contact shaft are passed through the
catalyzer, calculate the composition by weight of the resulting gaseous
products.
(e) Calculate the overall degree of completion of the conversion of the sulfur in
the pyrites charged to SO3 in the final products.
12. In the LeBlanc soda process the first step is carried out according to the
following reaction:
2NaCl -1- H2SO4 = NaCl + NaHSO, -|- HCl
The acid used has a specific gravity of 58°Baum6, containing 74.4% H2SO4. It is
supplied in 2% excess of that theoretically required for the above reaction.
(a) Calculate the weight of acid suppHed per 100 lb of salt charged.
(6) Assume that the reaction goes to completion, all the acid forming bisulfate,
and that in the process 85% of the HCl formed and 20% of the water
present are removed. Calculate the weights of HCl and water removed
per 100 lb of salt charged,
(c) Assuming the conditions of part (b), calculate the percentage composition
of the remaining salt mixture.
13. In the common process for the manufacture of nitric acid,, sodium nitrate is
treated with aqueous sulfuric acid containing 95% H2SO4 by weight. In order that
the resulting " niter cake " may be fluid, it is desirable to use sufficient acid so that
there will be 34% H2SO4 by weight in the final cake. This excess H2SO4 will actually'
be in combination with the Na2S04 in the cake, forming NaHS04, although for pur-
poses of calculation it may be considered as free acid. It may be assumed that the
cake will contain 1.5% water, by weight, and that the reaction will go to completion,
198 M A T E R I A L BALANCES [CHAP VI
b u t t h a t 2% of the HNO3 formed will remain in the cake. Assume t h a t the sodium
nitrate used is dry and pure.
(o) Calculate the weight and percentage composition of the niter cake formed per
100 lb of sodium nitrate charged.
(6) Calculate the weight of aqueous acid to be used per 100 lb of sodium nitrate,
(c) Calculate the weights of nitric acid and water vapor distilled from the niter
cake, per 100 lb of NaNOa charged.
14. Pure carbon dioxide may be prepared by treating limestone with aqueous sul-
furic acid. The limestone used in such a process contained calcium carbonate and
magnesium carbonate, the remainder being inert, insoluble materials. The acid used
contained 12% H2SO4 by weight. The residue from the process had the following
composition:
CaS04 8.56% "
MgSOi 6.23% ;-
HiiSOi 1.05% ,
Inerts 0.53%
CO2 0.12% . ^
Water 84.51%
During t h e process the mass was warmed and carbon dioxide and water vapor
removed.
(a) Calculate the analysis of the limestone used.
(6) Calculate the percentage of excess acid used.
(c) Calculate the weight and analysis of the material distilled from the reaction
mass per 1000 lb of limestone treated.
16. Barium carbonate is commercially important as a basis for the manufacture
of other barium compounds. In its manufacture, barium sulfide is first prepared
by heating the natural sulfate, barytes, with carbon. The barium sulfide is extracted
from this mass with water and the solution treated with sodium carbonate to precipi-
tate the carbonate of barium.
In the operation of such a process it is found t h a t the solution of bariimi sulfide
formed contains also some calcium sulfide, originating from impurities in the bary-
tes. T h e solution is treated with sodium carbonate and the precipitated mass of
calcium and barium carbonates is filtered off. It is found that 16.45 lb of dry pre-
cipitate are removed from each 100 lb of filtrate collected. The analysis of the pre-
cipitate is:
CaCOa..' 9.9%
BaCOs 90.1% '
The analysis of the filtrate is found to be:
NazS 6.85%
NaaCOa 2.25% ', .
H20 90.90%
The sodium carbonate for the precipitation was added in the form of anhydrous soda
ash which contained calcium carbonate as an impurity.
(a) Determine the percentage excess sodium carbonate used above t h a t required
to precipitate the BaS and CaS.
(6) Calculate the composition of the original solution of barium and calcium
sulfides. {Note: Barium sulfide is actually decomposed in solution, exist-
ing as the compound OHBaSH-5H20. However, in this reaction the
entire calculation may be carried out and compositions expressed as though

the compoimd in solution were BaS.)
(c) Calculate the composition of the dry soda ash used in the precipitation.
16. A mixture of 45% CaO, 20% MgO, and 35% NaOH by weight is used for
treating ammonium carbonate liquor in order to liberate the ammonia.
(a) Calculate the number of pounds of CaO to which 1 lb of this mixture is
equivalent.
(6) Calculate the composition of the mixture, expressing the quantity of each
component as the percentage it forms of the total reacting value of the
whole,
(c) Calculate the number of pounds of ammonia theoretically hberated by 1 lb
of this mixture.
17. A water is found to contain the following metals, expressed in milligrams per
hter: Ca, 32; Mg, 8.4; Fe (ferrous), 0.5.
(a) Calculate the " total hardness " of the water, expressed in miUigrams of
equivalent CaCOs per Uter, the calcium of which would have the same react-
ing value as the total reacting value of the laetals actually present.
(6) Assuming that these metals are all combined as bioarbonates, calculate the
cost of the lime required to soften 1000 gal of the water. Commercial
lime, containing 95% CaO, costs $8.50 per ton.
18. A glass for the manufacture of chemical ware is composed of the siUcates and
borates of several basic metals. Its composition is as follows:
SiOj 66.2% ZnO 10.3%
B2O3 8.2% MgO 6.0%
AI2O3 1.1% NaaO 8.2%
Determine whether the acid or the basic constituents are in excess in this glass, and
the percentage excess reacting value above that theoretically required for a neutral
glass. (Assume that AI2O3 acts as a base and B2O3 as an acid, HBO2.)
19. In Illustration 6, assume that the level is kept constant in each effect with the
same evaporations as noted, namely, 12.6, 11.70, and 11.76 tons from the three
evaporators. How much salt would be crystallized in each effect?
20. In Illustration 6, assume that the salt was removed from the hoppers, that the
drops in liquid levels were the same, and that the water evaporated was 11.1 tons
in the first effect and 10.2 tons in the second effect. Calculate the salt crystallized
in each evaporator and the water evaporated from the .third effect.
21. A mixture of benzene, toluene, and xylenes is separated by continuous frac-
tional distillation in two towers, the first of which produces benzene as an overhead
product and a bottoms product of toluene and xylenes, which is charged to the
second tower. This tower produces toluene as overhead and xylenes as bottoms. A
flow diagram of the operation is shown in Fig. 34. During a twelve-hour period
of operation the following products are removed from the towers:
#1 Tower #2 Tower #2 Tower
. Net overhead Net overhead Bottoms
Gallons at 60''F 9,240 6,390 10,800
Composition % by volume
Benzene 98.5 1.0
Toluene 1.5 98.6 0.7
Xylenes 0.5 r 99.3
During the period the Uquid levels in the accumulator sections of the plant varied, as
shown by level indicators measuring the heights of the levels above arbitrary fixed
points. The hquid volume, corrected to 60°F of the accumulator sections per inch
of height and the initial and final levels were as follows:
Corrected liquid Initial Final level
vol. gal/in. level, in. in.
#1 Tower Reflux Accumulator 24 36 68
#1 Tower Bottom Accumulator i 46
#2 Tower Reflux Accumulator 3$ 86 W
#2 Tower Bottom Accumulator IS 52 M
Benzene
J Condenser No. 2
^y <=s Reflux
^ ^ Accumulator
^ No. 2
_ ^Toluene
^ 1 T —1X1—*-
_j^^jJCylenes
Tower No. 1 Q Tower No. 2 . .
F I G . 34. Inventory changes in three-component distillation.
Calculate:
(a) The volume and composition of the feed during the twelve hour operating
period.
(b) The rate of flow from the bottom of Tower #1.
(c) The rate of production of distillate and bottoms from each tower assuming
the hquid levels had been maintained constant for the same rate of feed.
22. Stock containing 1.562 pounds of water per pound of dry stock is to be dried to
0.099 pound. For each pound of stock (dry basis) 52.5 pounds of dry air pass
through the dryer, leaving at, a humidity of 0.0525. The fresh air is supplied at a
humidity of 0.0152. ,
(o) Calculate the fraction of air recirculated.
(b) If the size of dryer required is inversely proportional to the average wet-bulb
depression of the air in the dryer, calculate the relative size of dryers required
with and without recirculation when operated at a constant temperatxire of
140°F.
23. It is required to condition 1000 cu ft of air per minute from 70°F and 80%
relative humidity to a constant value of 10% relative humidity by means of a station-
ary bed of silica gel. Part of the stream may be by-passed and a damper regulator
automatically controlled to maintain a constant wet-bulb temperature in the final air
mixture. The air velocity through the bed shall not exceed 100 ft per minute (total
cross section basis) and the time cycle, before regeneration is necessary, shall be 3
hours. Assume isothermal operation. Calculate the diameter and thickness of bed
required.
CHAPTER VII
THERMOPHYSICS
In Chapter II, the general concepts of energy, temperature, and heat

were introduced under the broad classification of potential and kinetic
energies. Both these forms were subclassified into external forms deter-
mined by the position and motion of a mass of matter relative to the
earth or other masses of matter and into internal forms determined by
the inherent composition, structure, and state of matter itself, inde-
pendent of its external position or motion as a whole.
Internal Energy. The internal energy of a substance is defined as
the total quantity of energy which it possesses by virtue of the presence,
relative positions, and movements of its component molecules, atoms,
and subatomic units. A part of this energy is contributed by the trans-
lational motion of the separate molecules and is particularly significant
in gases where translational motion is nearly unrestricted in contrast
to the situation in liquids and solids. Internal energy also includes the
rotational motion of molecules and of groups of atoms which are free to
rotate within the molecules. It includes the energy of vibration be-
tween the atoms of a molecule and the motion of electrons within the
atoms. These kinetic portions of the total internal energy are deter-
mined by the temperature of the substance and by its molecular struc-
ture. The remainder of the internal energy is present as potential energy
resulting from the attractive and repulsive forces acting between mole-
cules, atoms, electrons, and nuclei. This portion of the internal energy
is determined by molecular and atomic structures and by the proximity
of the molecules and atoms to one another. At the absolute zero of
temperature all translational energy disappears but a great reservoir of
potential energy and a small amount of vibrational energy remains.
The total internal energy of a substance is unknown but the amount
relative to some selected temperature and state can be accurately deter-
mined. The crystalline state or hypothetical gaseous state at absolute
zero temperature are commonly used as references for scientific studies,
whereas engineering calculations are based upon a variety of reference
conditions arbitrarily selected.
Energy in Transition; Heat and Work, In reviewing the several
forms of energy previously referred to, it will be noted that some are
capable of storage, unchanged in form. Thus, the potential energy of
201
202 THERMOPHYSICS [CHAP. VII
an elevated weight or the kinetic energy of a rotating flywheel are stored

as such until by some transformation they are converted, in part, at
least, to other forms. The internal energy of matter may be considered
as in storage awaiting transformation.
Heat represents energy in transition under the influence of a tempera-
ture difference. When heat flows from a hot metal bar to a cold one
the internal energy stored in the cold bar is increased at the expense of
that of the hot bar and the amount of heat energy in transition may be
expressed in terms of the change in internal energy of the source or of
the receiver. Under the influence of a temperature gradient heat flows
also by the bodily convection and mixing of hot and cold fluids and by
the emission of radiant energy from a hotter to a colder body without
the aid of any tangible intermediary.
I t is inexact to speak of the storage of heat. The energy stored
within a body is internal energy and when heat flows into the body
i't becomes internal energy and is stored as such. Zemansky^ writes
as follows: " The phrase, ' the heat in a body' has absolutely no
meaning. Perhaps an analogy will clinch the matter. Consider a fresh
water lake. During a shower, a certain amount of rain enters the lake.
After the rain has stopped, there is no rain in the lake. There is water
in the lake. Rain is a word used to denote water that is entering-
the lake from the air above. Once it is in the lake, it is"no longer
rain."
Another form of energy in transition of paramount interest is work,
which is defined as the energy which is transferred by the action of a
mechanical force moving under restraint through a tangible distance.
It is evident that work cannot be stored as such but is solely the mani-
festation of the transformation of one form of energy to another. Thus,
when a winch driven by a gasoline engine is used to lift a weight, the
internal energy of the gasoline is transformed in part to the potential
energy of the elevated weight and the work done is the energy trans-
ferred from one state to the other.
Energy Units. The fundamental energy unit is derived from the defi-
nition and concept of work. Thus, in the cgs system the basic energy
imit is the erg, which is the amount of work done by a force of one dyne
acting under restraint through a distance of one centimeter. Since
this unit is inconveniently small, the joule, equal to 10' ergs, is more
commonly used. In Enghsh units there is some confusion in termi-
nology resulting from the fact that the pound is used both as a unit of
mass and as a unit of force. By definition, a force of one pound is the
standard gravitational force exerted on a mass of one pound at sea
1 Heat and Thermodynamics, McGraw-Hill Book Co. (1937). With permission.
CHAP. VII] ENERGY UNITS 203
level. This pound force will accelerate a mass of one pound at a rate
of 32.17 feet per second per second. An alternate unit of force is the
poundal, which is defined as the force which will accelerate a mass of
one pound at a rate of one foot per second per second. Thus, a force
of one pound is equal to 32.17 poundals. The foot-pound is the work
done by a force of one pound acting through a distance of one foot
and the Joot-poundal is the work done by a force of one poundal acting
through a distance of one foot.
Since heat is a form of energy it may also be expressed in ergs, joules,
or foot-pounds. However, in problems dealing with the production,
generation, and transfer of heat it is customary to use special units of
energy called heat units. These units are expressed in various terms
depending upon which temperature scale and system of weights are
employed. The smallest heat unit is the calorie, which is the amount
of heat required to increase the temperature of one gram of water from
15° to 16°C. One calorie is equivalent to 4.185 X 10' ergs. It will be
recognized that the calorie is an arbitrary unit expressed in terms of
the thermal capacity of water and is not based upon the cgs system of
measurements. A calorie has also been defined as one-hundredth of the
amount of heat required to increase the temperature of one gram of
water from 0° to 100°C. This latter unit, called the mean calorie, is no
longer in use. The mean calorie is 0.017 per cent larger than the 15°
calorie. The calorie is also rigorously defined directly in electrical
energy units as S^-Q international watt-hour.
In industrial calculations it is always convenient to use heat units
larger than the calorie, such as the kilocalorie, British thermal unit, or
Centigrade heat unit. The kilocalorie is equal to 1000 calories, or it is
the amoimt of heat absorbed in increasing the temperature of one kilo-
gram of water from 15° to 16°C. When measurements are made in Eng-
lish units, employing pounds and the Fahrenheit temperature scale, the
British thermal unit or Btu is used. This unit is the quantity of heat
absorbed in increasing the temperature of one pound of water from 60°
to 61°F. One Btu is mechanically equivalent to 778 foot-pounds of
energy. Industries of the United States have often been willing to
adopt the Centigrade temperature scale but have refused to abandon
the English system of weights. This condition has given rise to a hybrid
heat unit known as the Centigrade heat unit, or Chu, which is the
amount of heat absorbed in increasing the temperature of one pound of
water from 15° to 16°C. The numerical values of heats of formation
and reaction, as explained later, are the same when expressed in either
kilocalorie per kilogram or Centigrade heat units per pound. For these
reasons the Centigrade heat unit has received some favorable acceptance.
In the Appendix are factors for the conversion from one system of units
to another.
Energy Balances. In accordance with the principle of conservation
of energy, also called the first law of thermodynamics, and referred to in
Chapter II, page 28, energy is indestructible and the total amount of
energy entering any system must be exactly equal to that leaving plus
any accumulation of energy within the system. A mathematical or
numerical expression of this principle is termed an energy balance, which
in conjunction with a material balance is of primary importance in prob-
lems of process design and operation.
In estabhshing a general energy balance for any process, it is conveni-
ent to use as a basis a unit time of operation, for example, one hour for
a continuous operation and one cycle for a batch or intermittent opera-
tion. It is necessary to distinguish between a flow process, which is
one in which streams of materials continually enter and leave the system,
and a non-flow process, which is intermittent in character and in which
no continuous streams of material enter or leave the system during the
course of operation. An ideal flow process is also characterized by
steady states of flow, temperatures, and compositions at any point in
the process in contrast to the changing conditions of composition in a
batch or non-flow process.
In an energy balance the inputs are equated to the outputs plus the
accumulation of energy inventory within the system over the unit period
of time in a flow process, or for a given cycle of operation for the non-
flow process. The separate forms of energy are conveniently classi-
fied as follows, neglecting electrostatic and magnetic energies, which
are ordinarily small:
o. Internal energy, designated by the symbol U per unit mass or mil
for mass m.
b. The energy added in forcing a stream of materials into the system
under the restraint of pressure. This flow work is equal to mpV, where
p is pressure of the system and V is the volume per unit mass. A similar
flow-work term is involved in forcing a stream of materials from the
system. These terms appear only in the energy balance of a flow
process.
c. The external potential energies of all materials entering and leaving
the system. External potential energy is expressed relative to an arbi-
trarily selected datum plane and is equal to mZ, where Z is the height of
its center of gravity above the datum plane.
d. The kinetic energies of all streams entering or leaving the system.
The kinetic energy of a single stream is equal to û^, where u is its
average linear velocity, and energy is expressed in ergs, joules, or foot-
CHAP. VII] ENERGY BALANCES 205
poundals, depending upon the units of m and u. Expressed in foot-

pounds the kinetic energy of a stream is ——, where gc is the standard
gravitational constant and equal to 32.17 feet per second per second.
e. The surface energies of all materials entering and leaving the sys-
tem. Surface energy per unit mass is designated as E^ and is generally
negligible except where large surface areas are involved, as in the forma-
tion of sprays or emulsions.
/ . The net energy added to the system as heat, and designated as q.
This net heat input represents the difference between the sum of stU
heat flowing into the system from all sources and the sum of all heat
flowing out of the system from all sources.
g. The net energy removed as work done by the system, designated as
w. This net work includes all forms of work done by the system, such
as mechanical and electrical work minus all such work added to the
system from all sources.
h. The net change in energy content within the system during the
course of the operation and designated by the symbol AE. In a steady-
state flow process with constant inventory, the energy change term AE
is zero. In a general process AE represents the change in energy con-
tent of the system as a result of any change in inventory of the system,
of change in temperature, of change in composition, of change in po-
tential energy of elevation, or of change in kinetic energy from stirring,
for example. In a non-flow process, the energy terms of the entering
and leaving streams disappear and the term AE becomes of major
importance.
The general energy equation may be set up, designating input items
by the subscript 1 and output items by the subscript 2. All input
energy items are balanced against the output and change items, the
summation symbol X indicating the sum of each form of energy in all
streams entering or leaving. Thus,
ZwiC/i -f- ZmipiFi -f j j ^ + ZmiZi + ^m,E„, + q

2
^gc
+ ZmÊ^^ + w + AE (1)
Simplifications of this general equation result in most specific cases.

In a steady flow process, without fluctuations in temperature, composi-
tion, and inventory, the term AE becomes zero. In a non-flow process,
where no stream of materials enters or leaves the system during the
course of operation, the flow work, kinetic energy, and potential energy
terms of the streams do not appear. For such a non-flow process where
surface energy is negligible the general equation reduces to
q = w + AE (2)
where
1712112 + rrhZi + m2 rriiUi - rriiZi — ) (3)

2gc 2gc /
where the prime values all refer to the properties of the inventory, with
subscript 1 referring to initial conditions and subscript B to final condi-
tions. The kinetic energy term in a non-flow process results from agi-
tation or stirring.
For a non-flow process at constant volume, no work and no potential
energy changes are involved, and if the kinetic and surface energy
terms are absent or negligible, Equation (1) becomes
. q = "^.i^Û^ — JlrriiUi * (4)
For a non-flow process at constant pressure, where changes in kinetic,
potential, and surface energy terms are negligible ^
q = JLmiUi — Y.iniUi -\- w (5)
Where work is performed only as a result of expansion against the con-
stant pressure p
w = "EmipVi — Y^mipVi (6)
Combining Equations (5) and (6),
q = Zm2 (C72 + pVi) - Z ^ i (t/i -|- pFi) (7)
Where only a single phase is present Equation (7) becomes
g = m ([/2 4- pVi) - m (C7i -I- p7i) = mA {U + pV) (8)
In many flow processes of the type encountered in chemical engineering
practice where large internal energy changes are involved and where the
energy associated with work and changes in potential kinetic and surface
energies are relatively small the general energy Equation (1) reduces to
L m i (C/i + piVi) + g = 1:^2 (C/2 4- p^Vi) (9)
Equation (9) may be satisfactorily applied to the great majority of
transformation processes where the primary objective is manufacture
and not the production of power. However, this equation should always
be recognized as an approximation and the significance of the neglected
terms verified when in doubt.
CHAP. VII] ENTHALPY 207
Enthalpy. In the energy equations for both flow and non-flow proc-
esses it will be seen that the term (C7 + pV) repeatedly occurs. It is
convenient to designate this term by the name enthalpy^ and by the
symbol H, thus
H = U + pV (10)
In a flow system the term pV represents flow energy but in a non-flow
system it merely represents the product of pressure and volume, having
the units of energy but not representing energy.
In a non-flow process proceeding at constant pressure and without
generation of electrical energy the heat added is seen from Equation (7)
to be equal to the increase in enthalpy of the system, or
In a flow system where the kinetic energy and potential energy terms
are negligible the heat added is equal to the gain in enthalpy plus the
work done, including both electrical and mechanical work, or
? i q= YlrriiHi — Y^miHi -\-w (11)
Where the work done is negligible in relative magnitude, the heat
added is equal to the gain in enthalpy, or
To smnmarize: For most industrial flow processes, such as the opera-

tion of boilers, blast furnaces, chemical reactors, or distillation equip-
ment, the kinetic energy, potential energy, and work terms are negligible
or cancel out and the heat added is equal to the increase in enthalpy.
Similarly, in non-flow processes at constant pressure where work other
than that of expansion is negligible, the heat added is equal to the
increase in enthalpy. However, where work, kinetic, or potential
energies are not negligible, or in non-flow processes at constant volume,
the increase in enthalpy is not equal to the heat added. The distinc-
tion between increases of enthalpy and additions of heat must be kept
constantly in mind.
Like internal energy, absolute values of enthalpy for any substance
are unknown but accurate values can be determined relative to some
arbitrarily selected reference state. The temperature, form of aggre-
gation, and pressure of the reference state must be definitely specified.
The values so reported are relative enthalpies. For a gas the reference
state of zero enthalpy is usually taken at 32°F and one atmosphere
pressure, and for steam as liquid water at 32°F under its own vapor
' This word was coined by Kammerling Onnes (1909), who purposely placed the
accent on the second syllable to distinguish it in speech from the commonly associated
term, entropy. , •' • <; . ; i
pressure at 32°F. For example, the enthalpy of steam at 200°F and an

absolute pressure of 10 pounds per square inch represents the increase in
enthalpy when liquid water at 32°F and its own vapor pressure is vapor-
ized and heated to a temperature of 200°r while its pressure is increased
to 10 pounds per square inch.
Heat Balance. " Heat balance " is a loose term referring to a special
form of energy balance which has come into general use in all thermal
processes where changes in kinetic energy, potential energy, and work
done are negligible. For such processes, so-called heat or enthalpy
balances are applicable to flow processes at any pressure and to non-
flow processes at constant pressure.
From consideration of the general energy equation it is evident that
neither heat nor enthalpy input and output items balance in the general
case. Although the term " heat balance " has become entrenched in
engineering literature, its use is undesirable because of the misleading
implications of the name. Accordingly, all such balances will be re-
ferred to by the proper term of " energy balance," even where the
kinetic, potential, and work items are neglected.
Water-
PumpFla •
• Hydrogen
High
I Condenser Pressure
Absorber
AE"
Reactor
T
CoDdensate
m "jj initial
mZ. final
64
'"2d
A£'
m[^ initial Compressor
ml final t Turbine
Compressor
Heater
Water
— -<— Steam
FIG. 35. Energy balance of a process for the production of hydrogen

and carbon dioxide from water gas.
Illustration 1. Application of the General Energy Balance. The principle of the

general energy balance of Equation (1) may be illustrated by application to the
recovery of hydrogen and carbon dioxide from water gas (Fig. 36). Dust-free
water gas is compressed, mixed with steam, heated, and passed to a reactor where
CHAP. VII] H E A T BALANCE 209
in contact with a catalyst the CO is converted to CO2. The products from the
reactor are cooled, with resultant condensation of water vapor, compressed further,
and passed into an absorber where CO2 is dissolved in water at high pressure. Hy-
drogen gas is delivered at high pressure in a nearly pure state. The high pressure
carbon dioxide solution generates power in a turbine and the CO2 gas and water
are thereby released at atmospheric pressure, and separated.
The following symbols are used to designate the various streams. All mass
and energy units correspond to the period of time selected.
mio = mass of water gas entering reactor •'i\ -,
mib = mass of steam entering reactor
rriie = mass of water entering absorber
m2a = mass of hydrogen stream leaving absorber
7712!. = mass of CO2 stream leaving turbine
mic = mass of water leaving turbine '
mid = mass of water leaving condenser
niia = initial inventory in reactor
mâ = final inventory in reactor
mjj, = initial inventory in absorber ><
m!tt, = final inventory in absorber
—wia = net work delivered to water gas in compression
—wib = net work delivered to absorber gases in compression
—Wic = net work delivered to water in pumping to absorber
+u)2a = work delivered to turbine by solution leaving absorber
Qia = heat added in heating gas mixture entering reactor
—qia = heat removed from gas stream leaving reactor by condenser
^ — g, = heat lost by radiation from all parts of plant
AE' = change in energy content of mass in reactor over period of run
AE" = change in energy content of mass in absorber over period of run
The markings to U, p, V, u, Z all correspond to those used above.

Applying these conditions to the general equation (1)
'^niiUi = miaUia + mibUib + rriicUic -. (a)
J^niiUi = niiaUia + rtiibUib + micUic + niidU^d - ' (b)
J ^ m i p i y i = miaPiaVia + mnpibVib + muPuVu "' ' (c)
'^niiPiVi = viiaPiaVia + mibPibVu + micPicVu + muPuVu ' (d)

y--~miu\ miaula mibulb micu\c
.,..; . ,.,'a:rf (e)
2gc 2gc 2gc '"'''.
^mtul m2aula mai&> miculc miduld
2gc 2gc 2g^ 2gc
"^rniZi = niiaZia + mibZu + muZi^ (g)
"^rniZi = m2oZ2o + muZ^b + micZi^ + nxuZti (h)
g = gi.. + gaa + gr ' • (i)
to = Wla + Wia + Wib + Wu (j)
AE = A E ' + A E " • • • •• • • (A-.^v , • - ,-r,»v.!- (k)
A£' = ni, (u[a + P L F L + ~ + ^2a) - ml. (u',, + p[aVL + ^ 4 Zja) (1)
AS" = "^^ ( c / ^ + v^Vg + ^ ' + Z ^ ) - »i;^ (u[i + PJ^F;^ + ^ V 2;J ) (m)
In general the inventory of both mass and energy remains constant, the kinetic
energy terms are negligible, the potential energy terms cancel, and the equation
reduces to
Smiffi + g = Sm2-H'2 + w '• ' , , '
It should be noted that the work terms refer to the net work energy added to the
system or supplied to the turbine and as a result of mechanical inefficiencies do not
correspond to the work required to drive the pumps and compressors or to that
generated by the turbine. The heat developed as a result of these inefficiencies
has been neglected. In chemical processing the work terms are usually negligible'
in the total energy balance although they may be of major importance in cost.
HEAT CAPACITY OF GASES

In a restricted sense heat capacity is defined as the amount of heat
required to increase the temperature of a body by one degree. Specific
heat is the ratio of the heat capacity of a body to the heat capacity
of an equal weight of water. Specific heat is a property, characteristic
of a substance and independent of any system of units, but dependent
on the temperatures of both the substance and the reference water.
Water at 15°C is usually chosen as the reference.
The heat capacity of any quantity of a substance is expressed mathe-
matically as
where
C = heat capacity
dq = heat added to produce a temperature change, dT
If a system is heated in such a manner that its volume remains constant,
dq = dU. The heat capacity under these conditions of constant
volume is expressed by
Thus, the heat capacity at constant volume, C„, is equal to the change
of internal energy with temperature.
If a system is heated in such a manner that the total pressure remains
constant and the volume is permitted to change, heat will be absorbed
both to increase the internal energy and to supply the heat equivalent
CHAP. VII] HEAT CAPACITY OF GASES 211
of, the external work which is done by the system in expanding. Thus,
Equation (12) becomes
where Cp is the heat capacity under constant pressure. For an ideal gas,
/aF\ _nR
(15)
Then, the heat capacity at constant pressure, Cj,, of one mole of an ideal
gas is represented by
+ R = c^ + R (16)
or
Cp = Cv + R (17)
where Cp and c» are the molal heat capacities at constant pressure and
volume, respectively.
The heat capacity of a substance which expands with rise in tem-
perature is always greater when heated under a constant pressure than
when heated under constant volume by the heat equivalent of the exter-
nal work done in expansion. For an ideal gas the molal heat capacities
under the two different conditions differ by the magnitude of the con-
stant, R. The numerical value of R is 1.99 calories per gram-mole per
degree Kelvin, or 1.99 Btu per pound-mole per degree Rankine.
For an ideal monatomic gas, such as helium, at a low pressure, it may
be assumed that, as a result of the simple molecular structure, the only
form of internal energy is the translational kinetic energy of the mole-
cules. The translational kinetic energy per mole of gas may be obtained
from Equation 3, Chapter II, page 30, and is equal to the internal
energy V in this particular case. Thus,
J7 = ^nu^ = %RT (18)
Then from Equation (13), for a monatomic gas,

c, = f i? (19)
Since R is approximately 2.0 calories per gram-mole per degree Kelvin,

the molal heat capacity of a monatomic gas at constant volume is equal
to 3.0 calories per gram-mole per degree Kelvin. The molal heat capac-
ity at constant pressure, from Equation (16), will be 3.0 + 2.0 or 5.0
calories. For monatomic gases the ratio of heat capacities, y, is
.-^-l-m (20)
For all gases, other than monatomic gases, the molal heat capacity at
constant volume is greater than 3.0. For a multatomic gas an increase
in enthalpy is used not only in imparting additional translational kinetic
energy, as evidenced by an increase in temperature and an increasing
velocity of translation, but also to impart increased energies of rotation
and vibration of the molecular and submolecular units. , . '«.i ,
TABLE IV
TRUE MOLAL HEAT CAPACITIES OP GASES
Constant Pressure (p = 0 atm abs) in calories per gram-mole per "C*
ec H2 D2 N2 02 CO NO H2O CO2 SO2 Air CHj C2H1 C2H2
0 6.86 6,97 6,96 6,99 6.96 7.16 7.98 8.61 9.31 6.94 8,24 10,02 10.13
100 6.96 6.98 6.98 7,13 7.00 7.15 8.10 9.69 10.17 6.99 9.40 12.42 11.76
200 6.99 7.01 7.05 7.37 7.09 7.25 8.32 10.47 10.94 7.10 10.70 14.74 12.98
300 7.01 7.06 7.16 7.61 7.23 7.42 8.56 11.23 11.53 7.24 12.15 16.73 13.71
400 7.03 7.16 7.31 7,84 7.40 7.62 8.84 11.79 12.03 7.40 13.40 18.42 14.39
500 7.06 7.27 7.47 8,02 7.57 7.79 9.12 12.25 12.38 7.57 14,60 19.90 15.01
600 7.12 7.39 7.63 8.18 7.75 7.95 9.41 12.63 12.65 7.72 15,65 21,17 15.55
700 7.20 7.52 7.78 8.31 7.90 8.09 9.72 12,94 12.86 7.87 16,60 22,30 16.04
800 7.28 7.67 7.91 8.41 8.03 8.22 10.02 13.20 13.02 7.99 17,40 23,30 16,49
900 7.38 7.81 8.03 8.50 8.15 8.32 10.30 13.41 13.15 8.10 18,23 24,17 16,90
1000 7.49 7.94 8.14 8.60 8.24 8.41 10.58 13.60 13.25 8.21 18,93 24,94 17.26
1100 7.59 8.06 8.24 8.66 8.33 8.48 10.84 13.74 13.34 8.29
1200 7.69 8.16 8.32 8.73 8.41 8.55 11.08 13,87 13.41 8,37
1300 7.80 8.26 8.38 8.79 8,47 8.61 11.31 13,98 13.46 8,43
1400 7.89 8.34 8.44 8,85 8,53 8.65 11.52 .14,07 13.51 8,49
1500 7.98 8.43 8.49 8,90 8,57 8.69 11.71 14,15 13.56 8,54
1600 8.08 8.51 8.54 8,96 8,62 8.73 11.88 14,22 13.59 8,59
1700 8.16 8.58 8.59 9,01 8,66 8.77 12.04 14,28 13.62 8,64
1800 8.24 8.64 8.63 9,08 8,69 8.80 12.19 14,33 13.65 8.68
1900 8.32 8.71 8.66 9,14 8,72 8.82 12.33 14,38 13.67 8,72
2000 8.38 8.76 8.70 9.19 8.75 8.85 12.45 14.42 13.69 8,77
2100 8.45 8.82 8.73 9.24 8.78 8.87 12.57 14.46 13.70 8,80
2200 8.51 8.87 8.76 9.29 8,81 8.89 12.68 14.49 13.72 8,83 /,-
2300 8.57 8.91 8.78 9.34 8,83 8.91 12.78 14,52 13.73 8,86
2400 8.62 8.96 8,80 9.39 8,85 8.93 12,87 14.54 13.74 8,89
2500 8.68 9.00 8,83 9.43 8,87 8.95 12,95 14.57 13.76 8,92
2600 8.73 9.04 8,84 9,47 8,89 8.9B 13,02 14,59 13.77 8,94
2700 8.78 9.08 8,86 9,51 8,90 8.98 13,08 14,61 13.77 8,96
2800 8.83 9.11 8.88 9,55 8,91 8,99 13,14 14,63 13.78 8,98
2900 8,88 9.14 8.89 9,59 8,92 9.01 13,18 14.64 13,79 9,00
3000 8,93 9.16 8.90 9,62 8,93 9.02 13,23 14,66 13,79 9,02
Mol. wt. 2,02 4,03 28,03 32,00 28,00 30.01 18,02 44,00 64,07 28,964 16,03 28.04 26.03
• E, JuBti and H. Liider, Forach. Oebiete Ingenieurw., 6, 210-1 (1935).

CHAP. VII] MOLAL HEAT CAPACITIES 213
In Table IV are given values of the true molal heat capacities of
a few common gases at zero pressure and at temperature intervals of
100°C from 0° to 3000°C. These same values are plotted in Fig. 36
against degrees Fahrenheit. These data have been calculated from
spectroscopic data by Justi and Llider and may be safely used at all
pressures up to one atmosphere. At high pressures corrections can
be applied as discussed in Chapter XII.
13
/
12 k//
^
ô-
B 10
//,
9
& 9 1 3- "co^
f KW
"Sj
VVi
600 1000 1500 2000 2500 3000 3500 4000 4500

Temperature, °F,
FiG. 36. True molal heat capacities of gases at constant pressure.
Empirical Equations for Molal Heat Capacities. Rigorous heat ca-

pacity equations such as the Planck-Einstein equation based upon
spectroscopic data and statistical mechanics are too complicated for
engineering use. Many empirical equations have been proposed to
represent the experimental data, the most useful form of which is of the
quadratic type,
Cp = a + bT + cT^ (21)
The constants a, b, and c were established by W. C. M. Bryant from

the best data available up to 1933, and are recorded in Table V for a
number of common gases. The constants for oxygen, nitric oxide, and
air have been recalculated from the more recent data of Justi and Liider.
The values for sulfur trioxide were calculated from vibration frequencies
214 =• ' THERMOPHYSICS [CHAP. VII
by the method described in Chapter XVI. The maximum deviations

from experimental results over the range 0, to 1800°C are given. Data
for hydrocarbon gases are presented in Chapter IX, pages 335 and 336.
TABLE V
EMPIRICAL EQUATIONS FOR MOLAL HEAT CAPACITIES OF GASES* '
Cp = a + hT + cT^ where T is in degrees K
calories per gram-mole per degree Kelvin
Maximum Deviation
Ga8^ 0 b X 10' cXW 0 to 1800°C
tAir 6.27 2.09 -0.459 1.0%
H, 6.88 0.066 +0.279 1.6
Nt 6.30 1.819 -0.345 1.2
to. 6.13 2.99 -0.806 1.5
CO 6.25 2.091 -0.459 1.2
tNO 6.17 2.60 -0.66 <2.0
Ha 6.64 0.959 -0.057 2.0

H20 6.89 3.283 -0.343 1.0
H2S 6.48 5.558 -1.204 <2.0
S02 8 12 6.825 -2.103 2.0
HCN 7.01 6.600 -1.642 <3.0
COj 6.85 8.533 -2.475 <3.0

COS 8.32 7.224 -2.146 2.0
CSj 9.76 6.102 -1.894 2.0
NH3 5.92 8.963 -1.764 <1.0
C2H2 8.28 10.501 -2.644 2.0
JSOs 8.20 10.236 -3.156

* Bryant, Ind. Eng. Chem. S22 25, (1933). With permission,
t Recalculated from 1935 data of Justi and LUder.
X Estimated.
Effect of Pressure on the Heat Capacity of Gases. The heat capaci-

ties of gases are not greatly affected by small variations in pressure at
low pressures. However, at low temperatures, heat capacities increase
rapidly with increase in pressure and become infinite at the critical
point. The effects of high pressures on heat capacities are discussed
in detail in Chapter XII. As the temperature is increased above the
critical, the effect of pressure upon heat capacity gradually diminishes.
Special Units for Heat Capacity of Gases. Industrial gases are usu-
ally measured by volume, in cubic feet or cubic meters. From the
equations given above for the molal heat capacities of gas, values may
be'readily calculated for any other units of weight or volume, such as
CHAP. VII] MEAN HEAT CAPACITIES OF GASES 215
kilocalories per kilogram, kilocalories per cubic meter, Btu per pound,
and Btu per cubic foot.
Whenever heat capacities are based on unit volumes, the basic quan-
tity of gas involved is that contained in a unit volume measured at
standard conditions of temperature and pressure and not at the existing
conditions. The heat capacity per unit volume refers to the heat
capacity of a definite and constant mass of gas, regardless of its tem-
perature and pressure. For example, the heat capacity per cubic
meter of oxygen at 1000°C signifies the heat capacity at 1000°C of
the weight of gas contained in 1 cubic meter a t standard conditions, oi* of
1.44 kilograms of oxygen. It does not signifiy the heat capacity of the
oxygen contained in 1 cubic meter of gas at the given temperature and
pressure. This distinction must always be kept in mind in speaking of
the heat-capacities of unit volumes of gases at various temperatures.
In the fuel gas industries an unusual unit of gas quantity is employed
as the standard. This unit is the quantity of gas contained in 1000
cubic feet, measured at a pressure of 30 inches of mercury, a temperature
of 60°F, and saturated with water vapor. This volume of gas corre-
sponds to 2.597 pound-moles of dry gas containing 0.046 pound-mole
of water vapor, or 2.643 pound-moles of the mixture. The heat capacity
of the mass of gas equivalent to 1000 cubic feet as measured in the gas
industry at 30 inches of mercury, saturated, and at 60°F can be obtained
by multiplying the molal heat capacity of that gas, expressed in Btu per
pound-mole per degree Fahrenheit, by the factor 2.643.
Mean Heat Capacities of Gases. The heat capacity equations given
in Table V represent the values of heat capacities at any temperature,
T°K. In heating a gas from one temperature to another it is desirable
to know the mean or average heat capacity over that temperature
range. The total heat required in heating the gas can then be readily
calculated by simply multiplying the number of moles of gas by the
mean molal heat capacity and by the temperature rise. This method
is easier than employing direct integration of the heat capacity for-
mulas for each problem. For gases of the air group where the tem-
perature coefficient is nearly constant, and over short temperature
ranges for other gases, accurate results may be obtained by simply
employing the heat capacity at the average temperature as the mean
heat capacity. Even for a gas such as carbon dioxide, whose tem-
perature coefficient of heat capacity changes markedly with temperature,
the heat capacity at 500°O is only 0.6 per cent higher than the correct
mean value over the temperature range from 0° to 1000°C.
The mean heat capacity over any given temperature range may be
calculated by integrating the general equation for true values over the
desired temperature range as follows: ^
f CpdT f {a + bT + cT^)dT
'J Ti " Ti
Mean Cp = ^^ _ ^^
T2- T,
(22)
where T% is the higher and T\ the lower temperature.
X TABLE VI
MEAN MOLAL HEAT CAPACITIES OF GASES BETWEEN 18°C AND T C
Constant Pressure (p = 0 atm abs) in calories per gram-mole per °C*
fC H2 D2 N2 O2 CO NO H2O CO2 SO2 Air CH4 C2H4 C2H2
18 6.86 6.97 6.96 7.00 6.96 7.16 7.99 8.70 9.38 6.94 8.30 10.20 10.25
100 6.92 6.97 6.97 7.06 6.97 7.16 8.04 9.25 9.81 6.96 8.73 11.45 11.12
200 6.95 6.98 7.00 7.16 7.00 7.17 8.13 9.73 10.22 7.01 9.48 12.64 11.79
300 6.97 6.99 7.04 7.28 7.06 7.22 8.23 10.14 10.59 7.06 10.20 13.73 12.36
400 6.98 7.02 7.09 7.40 7.12 7.31 8.35 10.48 10.91 7.14 10.88 14.76 12.81
500 6.99 7.06 7.15 7.51 7.20 7.39 8.49 10.83 11.18 7.21 11.53 15.69 13.21
600 7.02 7,10 7,21 7.61 7.28 7.47 8.62 11.11 11.42 7.28 12.15 16.52 13.58
700 7.04 7,15 7,28 7.70 7.35 7.55 8.76 11.35 11.62 7.35 12.74 17.26 13.90
800 7.07 7.21 7.36 7.79 7.44 7.63 8.91 11.57 11.79 7.43 13.28 18.07 14.22
900 7.10 7.28 7.43 7.87 7.51 7.71 9.06 11.76 11.95 7.60 13.79 18.62 14.49
1000 7.13 7.33 7.50 7.94 7.58 7.77 9.20 11.94 12.08 7.57 14.27 19.24 14.75
1100 7.16 7.41 7,57 8.00 7.65 7.84 9.34 12.11 12.20 7.63 '
1200 7.21 7.46 7,63 8.08 7.71 7.90 9.47 12.25 12.29 7.69 tj
1300 7.25 7.52 7,68 8.13 7.77 7.95 9.60 12.37 12.39 7.74 ''-
1400 7.29 7.57 7,74 8.18 7.82 8.00 9.74 12.50 12.46 7.79
1500 7.33 7.63 7,79 8.22 7.85 8.04 9.86 12.61 12.53 7.85
1600 7.37 7.67 7,83 8.25 7.91 8.09 9.98 12.71 12.60 7.89
1700 7.41 7.73 7.87 8.27 7.95 8.13 10.11 12.80 12.65 7.93
1800 7.46 7.78 7.92 8.34 7.99 8.16 10.22 12,88 12.71 7.97
1900 7.50 7.83 7.95 8.39 8.03 8.20 10.32 12.96 12.74 8.00
2000 7.54 7.88 7.99 8.43 8.06 8.23 10.43 13.03 12.80 8.04
2100 7.58 7.93 8.02 8.46 8.10 8.27 10.54 13.10 12.84 8.07
2200 7.63 7.96 8.06 8.49 8.13 8.30 10.63 13.17 12.88 8.09
2300 7.67 8.00 8.09 8.53 8.16 8.32 10.73 13.23 12.92 8.13
2400 7.71 8.04 8.11 8.57 8.19 8.35 10.81 13.27 12.96 8.15
2500 7.75 8.08 8.15 8.60 8.22 8.37 10.89 13.33 12.98 8.19
2600 7.79 8.11 8.18 8.64 8.24 8.39 10.98 13.37 13.01 8.21
2700 7.82 8.14 8.20 8.66 8.26 S.41 11.05 13.42 13.03 8.24
2800
2900
7.86
7.90
8,18
8.21
8.23
8.26
8.69
8.73
8.28
8.31
8.43
8.45
11.13
11.20
13.46
13.51
13.06
13.09
8.26
8.28 f
3000 7.93 8.24 8.27 8.77 8.33 8.46 11.25 13.55 13.12 8.30
Mol. wt. 2.02 4.03 28.03 32.00 28.00 30.01 18.02 44.00 64.07 28.96 16.03 28.04 26.03
* B. Justi and H. Luder, Forach. Gebiete Ingenieurw., 6, 211 (1935).

CHAP. VII] HEAT CAPACITIES OF SOLIDS 217
The available data for the mean heat capacity of gases over the
temperature range from 18°C to t°C are tabulated in Table VI for
temperature intervals of 100°C from 0°C to 3000°C from the data of
Justi and Liider.^ These data are also plotted in Fig. 37 for the tempera-
ture range 65° to i°F. This particular range has been selected because
the lower temperature corresponds to the reference temperature of
thermochemical data.
13
• I
/
/ ^
12
V
&} ^
^ • ^
A,0
//^
9 /
1/
7 3-
^
=J^ N2
— "HT
0 500 1000 1500 2000 2500 30OO 3600 4000 4500

,, Temperature, *F. i
FIG. 37. Mean molal heat capacities oi gases at constant pressure

(mean value 65° to «°P).
HEAT CAPACITIES OF SOLIDS

According to the law of Petit and Dulong the atomic heat capacities
of the crystalline solid elements are constant and equal to 6.2 calories
per gram-atom. This rule applies satisfactorily to all elements having
atomic weights above 40 when applied to constant volume conditions
at room temperatures. From kinetic theory, Boltzmann showed that
the atomic heat capacity of the elements at constant volume reaches a
maximum value of ZB = 5.97 calories per degree. The atomic heat
3 Forsch. Gebiete Ingenieurw., 6, 211 (1935).
218 THERMOPHYSICS [CH-AP. VII
£00
FiQ. 38. Specific heats of common elements and cokes.

CHAP. VII] KOPP'S RULE 219
capacities of elements such as carbon, hydrogen, boron, silicon, oxygen,

fluorine, phosphorus, and sulfur are much lower than the rule indicates.
At increasing temperatures, however, the atomic heat capacities of
these elements also approach the value 6.2. The atomic heat capacities
of all elements decrease greatly with decrease in temperature, approach-
ing a value of zero at absolute zero temperature when in the crystalhne
state.
Kopp's Rule. The heat capacity of a solid compound is approxi-
mately equal to the sum of the heat capacities of the constituent ele-
ments. This generaUzation was first shown by Kopp to be approxi-
mately correct, provided the following atomic heat capacities are
assigned to the elements: C, 1.8; H, 2.3; B, 2.7; Si, 3.8; 0,4.0; F, 5.0;
3K
fjPs..-.
M8£,^«=
^Sô
^
.260
^ 9
-'-^P' 3°!^
.200 / /
—fTo
ov C<iO_- ^ZoO__
.150
SnOj__
.100
PbO
.060
0
800 lOOO 1200 1400
Degrees Kelvin
F I G . 39. Specific heats of some common oxides.
P, 5.4; S, 5.4; all others, 6.2. This rule should be used only where
experimental values are lacking. Since the heat capacities of soUds
increase with temperature it is obvious that the above empirical rule is
inexact. In general, the heat capacities of compounds are higher in the
liquid than in the solid state. At the melting point the two heat capaci-
ties are nearly the same.
The heat capacity of a heterogeneous mixture is a simple additive
property, the total heat capacity being equal to the sum of the heat
220 THERMOPHYSICS [CBAI-. VII
capacities of the component parts. When true solutions are formed,

this simple additive property no longer exists.
The specific heats of various elements and oxides are presented graphi-
cally in Figs. 38 and 39. In Fig. 40 are specific heats of a few calcium
compounds. The heat capacities of many other common solids are
.400 1 1 - 1 " T " ' • • 1
.300
C Ca2^
3^)^::^^-—
CaSi03
CaFj..,-^
X
U.200
c^^ ^ ^ 1 ^ 1 .
//, -
100
ir \ . _ 1 — . 1 1 1 1
200 400 600 800 1000 1200 140O
Temperature," Kelvin
FIG. 40. Specific heats of some calcium compounds.
tabulated in Table VII. It will be noted that no general prediction

can be made as to the quantitative effect of temperature on these heat
capacities. Transition points, TP, indicating changes in crystalline
structure, and melting points, MP, correspond to abrupt changes in the
heat capacity relationships.
CHAP. VII] HEAT CAPACITIES OF SOLIDS 221
TABLE VII
HEAT CAPACITIES OP SOLID SUBSTANCES
Data from the International Critical Tables unless otherwise indicated
Cp = calories per gram per °C or
= kilocalories per kOogram per °C
Inorganic Compounds
Substance Formula fC Cp
Aluminum sulfate AU(S04)3 60 0.184
Aluminum sulfate Al2(S04)3-17H20 34 0.363
Ammonium chloride NH4CI 0 0.357
Antimony trisulfide SbaSa 0 0.0830
(stibnite) " V . , 100 0.0884
Arsenious oxide AS2O3 0 0.117
40 0.122
Barium carbonate BaCOs 0 0.100
, 100 0.110
f ' t
400 0.123
• ' •• -i- - - 800 0.130
Barium chloride BaClj 0 0.0853
100 0.0945
Barium sulfate BaSO, • 0 0.1112
1000 0.1448
Cadmium sulfide CdS 0 0.0881
: ' . " . - • * ,
60 0.0924
Cadmium sulfate CdSOi-SHjO 0 0.1950
Calcium carbonate CaCOa 0 0.182
200 0.230
' ' ' - '<•'',
400 0.270
Calcium chloride CaClj 61 0.164
Calcium chloride CaCla-GHjO 0 0.321
Calcium fluoride CaFs 0 0.204
•. -i' ^ . , S 40 0.212
80 0.216
Calcium sulfate ^ CaS04 0 0.1691
400 0.2275
Calcium sulfate CaS04-2H20 0 0.2650
60 0.198
Chromium oxide CrjOa 0 0.168
50 0.188
Copper sulfate CUSO4 0 0.148
Copper sulfate CuSOi-HjO 0 0.1717
Copper sulfate CuSOi-SHjO 9 0.2280
Copper sulfate CuS04-5H20 0 0.2560
Ferrous carbonate FeCOs 54 0.193
Ferrous sulfate FeSOi 46 0,167
Lead carbonate •' PbCOs 32 0.080
Lead chloride PbCla 0 0.0649
200 0.0704
i! 400 0.0800
Lead nitrate PbCNOs), 45 0.1150
Lead sulfate PbS04 45 0.0838
Magnesium chloride MgCU 48 0.193
Magnesium sulfate MgS04 61 0.222
Magnesium sulfate MgS04-H20 9 0.239
Magnesium sulfate MgS04-6H20 9 0.349
Magnesium sulfate MgS04-7H20 12 0.361
Manganese dioxide Mn02 0 0.152
TABLE VII (Continued)
Substance Formula rc Cp
Manganic oxide MnjOa 68 0.162

Manganous oxide MnO 58 0.168
Mercuric chloride HgCU 0 0.0640
Mercuric sulfide HgS 0 0.0506
Mercurous chloride HgCl 0 0.0499
Nickel sulfide NiS 0 0.116
100 0.128
i • • ' 200 0.138
Potassium chlorate KClOa 0 0.1910
200 0.2960
Potassium chloride KCl 0 0.1626
; 200 0.1725
..( ; : ' 400 0.1790
Potassium chromate KzCrOi 46 0.1864
Potassium dichromate KjCrjO, 0 0.178
400 0.236
Potassium sulfate K2SQ4 0 0.1760
Potassium nitrate KNO3 0 0.2140
200 0.267
300 0.292
Silver chloride AgCl 0 0.0848
200 0.0974
500 0.101
Silver nitrate AgN03 60 0.146
Sodium borate NajBiOr 46 0.234
Sodium borate (borax) NajBiOjlOHaO 35 0.386
Sodium carbonate Na^COs 46 0.266
Sodium chloride NaCl 0 0.204
100 0.217
; ' •
40C 0.229
,600 0.236
Sodium nitrate NaNOs , 0 0.2478
100 0.294
250 0.358
Sodium sulfate Na^SO* 0 0.202
> •
100 0.220
Water H2O -40 0.435
(Ice) 0 0.492
Organic Compounds
Cyanamide CH2N2 20 0.648

Oxalic acid C2H204-2HsO 0 0.338
100 0.416
Tartaric acid CHeOe-HjO 0 0.308
Picric acid C6H3N3O7 0 0.240
100 0.297
Nitrobenzene CeHsNOj 20 0.349
Benzene CsHfi 0 0.376
Benzoic acid C7H60a 20 0.287
Naphthalene CioHs 0 0.281
100 0.392
Anthracene C14H10 50 0.308
Palmitic acid C16H32O2 0 0.382
Stearic acid Cl8H3602 15 0.399
CHAP. V I I ] SPECIFIC HEATS 223
TABLE VII {Continued)
SPECIFIC HEATS OF MISCELLANEOUS MATERIALS*
Specific Heat Temperature Range

Substance Cals/g °C °C
Alumina 0,2 100
0.274 1500
Alundum 0.186 100
Asbestos 0.25
Asphalt 0.22
Bakelite 0.3 to 0.4
Brickwork - Approx. 0.2
Carbon 0.168 26-76
0.314 40-892
0.387 56-1450
Carbon (gas retort) 0.204
Cellulose 0.32
Cement 0.186
Charcoal (wood) 0.242
Chrome brick 0.17
Clay 0.224
Coal 1 0.26 to 0.37 . • : • . '
Coal tars \ 0.35 40

! 0.45 200
Coke , 0.265 21-400
0.359 21-800
0.403 21-1300
Concrete 0.156 70-312
0.219 72-1472
Cryolite - 0.253 16-55
Diamond 0.147
Fire clay brick 0.198 100
0.298 1500
Fluorspar 0.21 30
Glass (crown) 0.16 to 0.20
(flint) 0.117
(pyrex) 0.20
(silicate) 1 0.188 to 0.204 0-100
0,24 to 0.26 0-700
(wool) 0.157
Graphite 0.165 26-76
0.390 56-1450
Granite 0.20 20-100
Gypsum 0.269 16-46
Limestone 0.217
Litharge 0.055
Marble 0.21 18
Magnesia 0.234 100
0.188 1500
Magnesite-brick 0.222 100
0.195 1500
Pyrites (copper) 0.131 19-50
(iron) ' 0.136 15-98
Quartz • 0.17 0
0.28 350
Sand 0.191
Silica 0.316
Steel 0.12
* From Chemical Engineers' Handbook, John H. Perry, McGraw-Hill Book Company, Inc. (1941),
with permiaaion. , . ,
HEAT CAPACITIES OF LIQUIDS AND SOLUTIONS

Few generalizations can be stated regarding the heat capacities of
liquids. The heat capacities of most liquids increase with an increase
in temperature.
The heat capacity of most substances is greater for the Uquid state
than for either the solid or the gas. Where experimental data are lack-
ing Kopp's rule (page 219) may be applied by assigning the following
values of atomic heat capacities to the atoms of the liquid, according to
Wenner^: C = 2.8, H = 4.3, B = 4.7, Si = 5.8, 0 = 6.0, F = 7.0,
P = 7.4, S = 7.4, and to most other elements a value of 8.
1.0
^CHJC :ooH
.95
CHN Os
.90
X .85
'/r^.so.
I. .80
.75
.70
0 20 40 60 80 100 120 140 160 180 200

Moles of Water per Mole of Compound
PIG. 4 1 . Specific h e a t s of a q u e o u s s o l u t i o n s of a c i d s a t 2 0 ° C .
1.0
- — n 1-^==:
.95
LiOH^
NaOH-y^ " ^ Ja^CO1
.90
,<K CO3
.85 /
.80
ri
I /~' :0H
.75
.70
1
0 20 40 60 80 100 120 140 160 180 200

Moles of Water per Mole of Corapound
Fia. 42. Specific heats of aqueous solutions of bases at 20°C.
*Wenner Thermochemical Calculations, McGraw-Hill Book Co. (1941), with

permission. ,
CHAP. VII] H E A T CAPACITIES O F L I Q U I D S A N D SOLUTIONS 225
1.0
NaCl
.95 y ^ .
NH ^KCl
.90 L%
S
a J / /
<^
-CaCl
-Bad
a .85
s / ^//
/
X
i..80
.75 7 / /
''
.70 It ' /
0
1 20
40 60 80 100 120 140 160 180 200
FiQ. 43. Specific heats of aqueous solutions of chlorides at 20°C.
I.O
.95 (NH4:
;
.90
.85
«c:
I .80
NasS
Y/
/A >04
Xcu 304
.75 / MgS
.70
0 20 40 60 80 100 120 140 160 180 200

F I G . 44. Specific heats of aqueous solutions of sulfates at 20°C.
1.0 r 1
Na NOr^
.95 N H,NO 3 ^ ^ - ^ ^
-Cu( TO3)2
-^î (NO3
.90
.85
/ /
1 .80
/ A^ P b ( •^03)2
A
.75
' t
.70
/
0 20
40 60 80 100 120 140 160 180 200
FiQ. 46. Specific heats of aqueous solutions of nitrates at 20°C.
TABLE VIII
HEAT CAPACITIES OF LIQUIDS
Data from International Critical Tables unless otherwise indicated.
s= heatcapacity, calories per gram per °C at t°C.
o = temperature coefficient in equation: Cp + at,
applying over the indicated temperature range
Liquid Formula fC Cp ''Po a Temp. Range

Mercury Hg 0 0.0335
60 0.0330
' 100 0.0329
200 0.0329
280 0.0332
Water* •••.. H2O • 0 1.008
15 1.000
100 1.006
200 1.061 ' %.
300 1.155
Sulfur dioxide SO2 -20 0.3130 y^
0.318 0.00028 11° t o 140°
Sulfuric acid H2SO4 0.339 0.00038 10° t o 45°
Ammonia NH3 -40 1.051
0 1.098
60 1.215 1 • • •
100 1.479
Silicon tetrachloride SiCU 12 to 50 0.200
Sodium nitrate NaNOs 350 0.430
Carbon tetrachloride CCI4 20 0.201 0.198 0.000031 0° to 70°
Carbon disulfide CS2 0.235 0.000246 - 1 0 0 ° t o + 1 5 0 °
Chloroform CHCI3 15 0.226 0.221 0.000330 - 3 0 ° to + 6 0 °
Formic acid CH2O2 0 0.496 0.000709 40° t o 140° •
Methyl alcohol CH4O 0 0.566
40 0.616
Acetic acid C2H4O2 0.468 0.000929 0° t o 80°
E t h y l alcohol CaHeO -50 0.473
0 0.535
25 0.580
50 0,652
100 0.824
150 1.053
Glycol C2H6O2 0 0.544 0.544 0.001194 - 2 0 ° t o + 2 0 0 °
AUyl alcohol CsHsO 0 0.3860
21 to 96 0.665
Acetone CsHeO 0.506 0.000764 - 3 0 ° t o + 6 0 °
Propane CsHs 0 0.576 0.576 0.001505 - 3 0 ° t o + 2 0 °
Propyl alcohol CsHsO -50 0.456
0 0.525
+50 0.654
Glycerol CsHsOs -50 0.485
0 0.540
+50 0.598
+100 0.668
E t h y l acetate C4H8O2 20 0.478 /
n-Butane C4H10 0 0.550 0.550 0.00191 - 1 5 ° to +20°
Ether C4H10O 0 0.529 /.
30 0.548
120 0.802
Isopentane CiHu 0 0.512
8 0.526
* Handbook of Chemistry and Physics (1939), with permission.
CHAP. VII] LATENT HEAT 227
TABLE VII] (Continued)
Liquid Formula fC Cj, Cpo a Temp. Range

Nitrobenzene CeHsNOj 10 0.358
50 0.329
120 0.393
Benzene CeHe 5 0.389
20 0.406
60 0.444
90 0.473
Aniline CeHiN 0 0.478
50 0.521
100 0.547
ra-Hexane CeHu 20 to 100 0.600
Toluene CyHs 0 0.386
50 0.421
100 0.470
n-Heptane C7H16 0 t o 5 0 0.507
30 0.518 0.476 0.00142 30° to 80°
Decane (BP 172°) CioHjj 0 t o 5 0 0.502
n-Hexadeoane C16H34 0 t o 5 0 0.496
Stearic acid CisHaeOa 75 to 137 0.650
Diphenyl* CiaHio 0.300 0.00120 80° to 300°
* Forrest, Brugmann, and Cummings, Ind. Eng. Chem. 23, 340 (1931).
A more nearly accurate method of estimating heat capacities of the

liquid state from data on the gaseous state is discussed in Chapter X I I .
Since, as previously mentioned, heat capacities of gases can be calculated
by generalized empirical relationships based on spectroscopic data, these
methods make it possible to predict the heat capacity of a substance at
any conditions of the liquid or gaseous state from a minimum of experi-
mental data.
Water has a higher specific heat than any other substance, with the
exception of liquid ammonia and a few organic compounds. The heat
capacity of water is a minimum at 30°C. The specific heats of aqueous
solutions in general decrease with increasing concentration of solute.
In dilute solutions the heat capacity of aqueous solutions is nearly
equal to the heat capacity of the water present. The heat capacities of
some common aqueous solutions of acids, bases, and salts are shown
graphically in Figs. 41, 42, 43, 44, and 45. The heat capacities of the
solutions all correspond to a temperature of about 20°C. In Table VIII
are values of the specific heats of some common liquids. Data for
petroleum fractions are presented in Chapter IX, page 334.
LATENT HEAT
Heat of Fusion. The fusion of a crystalline solid at its melting point

to form a liquid at the same temperature is accompanied by an increase
in enthalpy or an absorption of heat. Since the volume changes and
hence the external work in fusion are small, this heat of fusion is largely
utilized in increasing the internal energy through rearrangement of t h e

a t o m s . A t t e m p t s have been m a d e t o establish general relationships
between latent heats of fusion a n d other more easily measured proper-
ties. N o n e of these generalizations are accurate. For most elements^
t h e ratio of — varies from 2 t o 3, for most inorganic compounds from
5 to 7, a n d for most organic compounds from 9 to 11, where
X/ = h e a t of fusion, calories per gram formula weight

Tf = melting point, ° K ;* , ,
T h e r e are a few marked exceptions t o these rules.
TABLE IX •
HEATS OF FUSION*
Xf — heat of fusion, calories per gram-atom or mole' or Chu per pound-atom or
mole. To convert to Btu per pound-mole multiply by 1.8.
t] = melting point, °C.
Tf = melting point, °K.
' -• " Elements •''•'• ' .; • •
X/ tf x//r/
Ae 2800 961 23
Al 2340 657 2.5
Cu 2660 1083 2.0
Fe* • 3660 1535 1.5
Na 629 98 1.7
Ni 4280 1450 2.5
Pb ; H60 327 1.9
s
Sn
300
1600
115
232
0 8
3.2
Zn 1660 419 2.4
Compounds
X/
H2O 1,435 0.0
SbgOa 5,950 640
CO2 2,000 - 56.2
CaCl2 6,040 774
NaOH , 1,600 318
NaCl 7,210 804
Carbon tetrachloride. 640 - 24
Methyl alcohol 525 - 97
Acetic acid 2,690 16.6
Ethyl alcohol 1,150 -114
Benzene 2,370 6.4
Aniline 1,950 - 7.0
Naphthalene 4,550 80
Diphenyl 4,020 71
Stearic acid 13,500
*See also Fig. 89, page 414-
' Wanner, ibid.

CHAP. VII] HEAT OP VAPORIZATION 229
Experimentally determined values of heats of fusion are given in
Table IX.
Heat of Transition. Many crystalline substances exhibit transforma-
tion or transition points at which temperature changes in crystalline
structure take place. The equilibrium temperature of transformation
is nearly constant although the actual temperature of transformation is
frequently a function of the rate at which the substance is heated or
cooled prior to the transformation. The transition usually takes place
at a slightly higher temperature when the substance is being heated
then when it is being cooled.
Crystalline transformations are accompanied by either absorption or
evolution of heat. The transformation of the phase which is stable at
low temperatures into the phase stable at high temperatures requires an
absorption of heat.
Data for heats of transition of a few solids are recorded in Table X.
"^ ' -i TABLE X .-

»• ' HEATS OF TRANSITION *
X( = heat absorbed in transition, calories per gram-atom or Chu per pound-atom
or mole. To convert to Btu per pound-mole multiply by 1.8.
tt = temperature of transformation, °C.
Data from International Critical Tables
Transition X, tt
Sulfur:
rhombic —> monoclinic 7,0 114-151°C
Iron (electrolytic) (see also Fig. 90,
page 419):
a —>/3 363 770
S—>-v 313 910
106 1400
Manganese:
a->^ 1325 1070-1130
Nickel:
78 320-330
Tin:
white —* gray 630 0
• HEAT OF VAPORIZATION
The heat required to vaporize a substance consists of the energy
absorbed in overcoming the intermolecular forces of attraction in the
liquid and the work performed by the vapors in expanding against an
external pressure. The external work performed by one mole in vapor-
izing under a constant pressure is equal to the product of the pressure
230 THERMOPHYSICS [CHAP. V H
and the increase in volume. Thus,

'K = v{vg- vi) / • ' (23)
where
X„ = external work of vaporization
i- p = pressure
y Vg = molal volume of saturated vapor
' i; J = molal volume of saturated liquid '
At ordinary pressures the volume of the liquid may be neglected. If
the vapor follows the ideal gas law the molal external work of vaporiza-
tion is equal to RT or 2T calories per gram-mole, where T is the tempera-
ture of vaporization in degrees Kelvin. The latent heat of vaporization
is much larger than the latent heat of fusion because the forces of molec-
ular attraction must be overcome to a greater extent in vaporization
than in fusion.
As pointed out in Chapter III, page 59, an exact relationship between
heat of vaporization and vapor pressure is expressed by the Olapeyron
equation, derived from the second law of thermodynamics. This equa-
tion permits accurate calculation of latent heats of vaporization a t any
temperature if data on the relationship between vapor pressure and
temperature and on the molal volumes of the liquid and vapor are
available. The results are rigorous at all temperatures and pressures.
However, the necessary data for use of the rigorous equation are available
for only a few substances.
Trouton's Rule. According to a rule proposed by Trouton, the ratio
of the molal heat of vaporization, X, of a substance at its normal boiUng
point to the absolute temperature of the boiUng point T^ is a constant.
Thus T^^ ' • ^^'

where K is termed Trouton's constant or, better, Trouton's ratio. For •
many substances this ratio is equal to approximately 21 where the latent
heat is expressed in calories per gram-mole and the temperature in
degrees Kelvin. However, this ratio is by no means a constant. For
polar hquids Trouton's rule breaks down completely, values of the ratio
being much greater than 21. In nonpolar liquids the variation is smaller
and the ratio increases as the normal boihng point increases.
Kistyakowsky Equation. ' A thermodynamic equation was proposed
by Kistyakowsky for the calculation of Trouton's ratio at the normal
boiling point of a nonpolar liquid:
| - = 8.75-h 4.571 k)gio n / "r '^ (25)

CHAP. VII] HEATS OF VAPORIZATION 231
where
X = heat of vaporization in calories per gram-mole at Ts°K
Tg = normal boiUng point in degrees Kelvin.
This equation is in excellent agreement with experimental results for a
wide variety of nonpolar liquids but is inapplicable to polar hquids.
Heats of Vaporization from Empirical Vapor-Pressure Equations." Al-
though, as mentioned above, the data necessary for the rigorous calcu-
lation of heats of vaporization from the Clapeyron equation are seldom
available, satisfactory approximations at low pressures may be obtained
by combining this equation with the empirical vapor-pressure equation
developed by Cahngaert and Davis, page 67. This method is particu-
larly useful for estimating the heats of vaporization of polar compounds
at their normal boiling points, for which compounds the Kistyakowsky
equation is not applicable.
As a fair approximation the difference between the molal volimies of
the vapor and liquid of any material at its normal boiling point is repre-
sented by
(vg - vi) = 0.95Br!,/pi (26)
where
Tb — normal boihng point ^
» ' pi= pressure at normal boiling point = 1 . 0 atmosphere
The factor 0.95 serves to correct for deviations from the ideal gas law
and for the volume occupied by the liquid. By substituting Equation
(26) in Equation (III-l), page 59, a modified Clausius-Clapeyron equa-
tion is obtained, applicable at the normal boiling point:
Another expression for dp/dT may be .obtained by differentiating Equa-

tion (III-7), page 67. Thus,
dp __ pB /
dT~ (T - 43)2 ^^^^
Combining Equations (28) and (27) for the normal boiling point,
X, = 0 . 9 5 f l B ^ ^ r ^ y (29)
The constant B of the Calingaert-Davis equation is readily evaluated

from vapor pressure values at any two temperatures. The normal
boihng point and the critical point may be used to estabhsh this constant.
^K. M. Watson, Ind. Eng. Chem., 36, 398 (1943), with permission.
However, with accurate vapor pressure data, better results are obtained
by selecting two values relatively close to; the normal boiling point.
Thus,
B = __JE£2Z£L___ (30)
(-1 '—)
\Ti - 43 T2- 4 3 /
Equation (29) may be used with fair accuracy for estimating heats of
vaporization at low pressures other than one atmosphere by substituting
the proper boiling points for Ti,. However^ it is recommended that its
application be limited to estimating values at the normal boiling point.
Values at other temperatures may then be obtained by the empirical
method described later.
Illustration 2. Ethyl alcohol has a normal boihng point at 78.3°C and a vapor
pressure of 15.61 atmospheres at 170°C. Estimate the molal heat of vaporization at
its normal boiling point.
From Equation (30), , _
-r"/ j^l5^ ,,- ; , , , ' „, ' .
^ 7 1 ' r-T-3720 /
, \351.3-43 443-43/ /
From Equation (29) , » r?-
X = (0.96) (1.99)(3720) ( - — ^ ) = 9140 cal per gr-mole

\308.3/
More reliable heats of vaporization may be obtained by differentiating Equation

(III-16) to obtain dp/dT for substitution in Equation (27). Further improve-
ment results from use of the actual compressibility factor of the vapor instead of the
average value of 0.95.
Low Pressure Heats of Vaporization from Reference Substance Vapor

Pressure Plots. From Equation (5) of Chapter III, page 65, it is evident
that the slope of a vapor pressure line on an equal temperature reference
substance plot is equal to the ratio of the heat of vaporization of the
compound in question to that of the reference substance at the same
temperature. Thus, the heat of vaporization at any temperature is
obtained by multiplying that of the reference substance at the same
temperature by the constant value of the slope of the vapor pressure
line. These slopes are indicated on Fig. 7.
A similar but less convenient relationship between the slopes of
equal pressure reference substance plots may be developed by applying
the Clausius-CIapeyron equation to both the compound in question
CHAP. VII] HEAT OF VAPORIZATION AND TEMPERATURE 233
and the reference substance at the same pressure. Thus,

,/TydT'
X= X ' y ^ (31)
where
X = heat of vaporization at temperature T and vapor pressure p
X' = heat of vaporization of reference substance at pressure p
and temperature T'
dT' '
—r;- = slope of Diihring Une ' '
Q/J.
As discussed in Chapter III, the satisfactory linearity of reference

substance plots up to the critical point appears to result from a fortuitous
compensation of errors in the basic assumptions involved. Even though
such a plot may be a straight line the actual ratio of the latent heats
may vary over a wide range and become meaningless as the critical
point of either substance is approached. For this reason great care
must be exercised in deriving heats of vaporization from reference
substance plots at elevated pressures. It is recommended that their
use be restricted to the low pressure range and that the methods of the
following sections be used for all pressures substantially above at-
mospheric.
Empirical Relationship Between Heat of Vaporization and Tempera-
ture. The heat of vaporization of a substance diminishes as its tempera-
ture and pressure are increased. At the critical temperature, as pointed
out in Chapter III, the kinetic energies of translation of the molecules
become sufficiently great to overcome the potential energies of the at-
tractive forces which hold them together and molecules pass from the
liquid to the vapor state without additional energization. At the critical
point there is no distinction between the liquid and vapor states, either
in enthalpy or other physical properties and the heat of vaporization
becomes zero.
It was pointed out by Watson' that if values of Trouton's ratio are
plotted against reduced temperature, curves of the same shape are ob-
tained for all substances both polar and nonpolar. These curves may be
superimposed by multiplying the ordinates of each by the proper con-
stant factor. It was found that the following empirical equation satis-
factorily represents these curves over the entire range of available data:
X_ ^ / I - TrY-
Xi" \ i - TJ (32)
Ind. Eng. Chem., 23, 360 (1931); 36, 398 (1943). • ' • - • ^'
TABLE XI
HEATS OF VAPORIZATION AND CRITICAL CONSTANTS*
X = heat of vaporization at t°C calories/gr-mole or Chu/Ib-mole
t = temperature
4 = critical temperature in °C
Pe = critical pressure ia atm
•—^ ' !
Substance X t°C tc Pc
Air -140.7 37.2

Ammonia — 5,581 -33.4 132.4 111.5
Argon 1,690 -185.8 -122 48.0
Bromine 7,420 68.0 302
Carbon dioxide X 6,030 -78.4 31.1 73.0
Carbon disulfide 273.0 76.0
Carbon monoxide 1,444 -191.5 -139.0 35.0
Carbon oxysulfide 4,423 -60.2 105.0 61.0
Carbon tetrachloride 7,280 77 283.1 45.0
Chlorine 4,878 -34.1 144.0 76.1
D ichlorodifluorome thane 4,888 -29.8 111.5 39.56
(Freon 12)
Dichloromonofluoromethane (F-21) 8.9 178.6 61.0
Helium 22 -268.4 -267.9 2.26 .
Hydrogen 216 -252.7 -239.9 12.8
Hydrogen bromide 4,210 -66.7 90.0 84.0
Hydrogen chloride 3,860 -85.0 61.4 81.6
Hydrogen cyanide 6,027 25.7 183.5 50.0
Hydrogen fluoride 7,460 33.3 230.2
Hydrogen iodide 161.0 82.0
Hydrogen sulfide 4,463 -60.3 100.4 88.9
Mercury 13,980 361 >1550 >200
Nitric oxide /" 3,307 -161,7 -94,0 65.0
Nitrogen . 1,336 -195.8 -147.1 33,5
Nitrous oxide "* 3,950 -88.5 36.5 71.7
Oxygen i 1,629 -183.0 -118.8 49.7
Phosgene t 182.0 66,0
Silicon tetrafluoride ' 6,130 -94.8 -1.5 50.0
Sulfur _ . 20,200 444.6 1,040
Sulfur dioxide 5,960 -5.0 157.2 77.7
Sulfur trioxide 10,190 44.8 218.3 83.6
Trichloromonofluoromethane (F-11) 23.6 198.0 43,2
Water 9,729 100.0 374.0 217.7
* Chemical Engineers* Handbook, John H. Perry, McGraw-Hill Book Company, Inc. (1941), with
permission.
This equation permits estimation of the heat of vaporization of any sub-

stance at any temperature if its critical temperature and the heat of
vaporization at some one temperature are known. In combination with
the methods described in the p r e ^ u s sections for estimating heats of
vaporization sa low pressures it is possible to predict data at all condi-
tions up to the critical with errors generally less than 5%.
Illustration 3. The latent heat of vaporization of ethyl alcohol is experimentally

found to be 204 calories per gram at its normal boiling point of 78°C. Its critical
CHAP. VII] REDUCED REFERENCE SUBSTANCE PLOT 235
temperature is 243°C. Estimate the heat of vaporization at a temperature of 180°C.

Ti/Tc = 361/516 = 0.680 = Tn
T/Tc = 453/516 = 0.880 = Trt
From Equation (32) '
ri - cssoi'-^'
Xj = 204 L _ Q ggql = 140 cal per gram
Reduced Reference Substance Plot. D. H. Gordon' has pointed out
that the limitations of the conventional reference substance plots for the
estimation of heats of vaporization may be greatly minimized by basing
the reference substance plot on equal reditced conditions. The most
convenient form is a logarithmic plot of vapor pressures of the substance
in question and of the reference substance at equal reduced temperatures.
It is found that this method of plotting yields good approximations to
linear curves for a wide variety of substances, both polar and nonpolar
over wide ranges of conditions up to the critical.
Since T = TrTc, the Clausius-Clapeyron equation (page 60) may be
written in terms of reduced temperatures:
Applying Equation (33) to a substance and a reference substance at the

same reduced temperature,
or
, . X = s.^X' (35)
c
where the primed quantities designate the reference substance. The

To
group of terms s, ^ is a constant for any one pair of substances.
The term Sr is the slope of the hne resulting when the logarithm of the
vapor pressure of the given substance is plotted against the logarithm
of the vapor pressure of a reference substance at the same reduced
temperature.
Values of Sr -î are given in Table X I I I for various refrigerants with
c
water as the reference substance. In Table XII are given the heats of
vaporization of water in Btu per pound-mole at various values of reduced
temperature, Tr, and the corresponding vapor pressures.
' Univ. of Wis. Ph.D. Thesis (1942). :
TABLE XII ^
MoLAL HEATS OF VAPORIZATION AND VAPOR PRESSURES OF WATER
Tc = 1165.4°R X = Btu per lb-mole p = pounds per sq in. abs
Tr X V
0.423 . 19,370 0.092
0.44 19,170 0.198
0.46 18,940 0.446
0.48 • ;; 18,700 . I*"V 0.934
0.60 18,460 :;' . 1.824
0.52 18,210 3.365

0.54 17,960 5.896
0.56 17,770 9.871
0.58 17,440 15.871
0.60 17,170 24.613
0.62 16,880 36.959

0.64 '" 16,590 54.006
0.66 16,280 76.785
0.68 15,950 106.65
0.70 15,600 145.08
0.72 15,230 193.68

0.74 14,840 254.1
0.76 14,420 328.3
0.78 13,980 418.4
0.80 13,490 526.1
0.82 12,970 ' 654.2

0.84 12,400 804.5
0.86 11,770 980.3
0.88 11,070 1184.7
0.90 10,290 1419.3
0.91 9,684 1548.7

0.92 9,401 1688.4
0.93 • 8,907 1836.0
0.94 8,365 1995.6
0.95 7,774 2164.3
0.96 7,103 2346.1
0.97 6,336 2538.6
0.98 5,393 2746.8
0.99 4,143 2967.5
1.00 0 3206.2
CHAP. VII] REDUCED REFERENCE SUBSTANCE PLOT 237
TABLE XIII
HEAT OF VAPORIZATION FACTORS AND CRITICAL TEMPERATURES OP REFRIGERANTS*
Critical Temp.
Refrigerant Formula Sr
op
Ammonia ; NHa 0.933 270.4 0.584

Benzene CeHe 0.923 551.3 0.800
Butyl alcohol C4H9OH 1.327 548.6 1.149
Carbon dioxide CO2 0.872 88.43 0.410
Carbon disulfide CS2 0.839 523.0 0.707
Carbon tetrachloride CCL 0.869 541.6 0.746
Chlorine CI2 0.776 295.0 0 502
Chlorobenzene CeHsCl 0.933 678.6 0.912
Chloroform CHCI3 0.916 505.0 0.749
Ethane C2H6 0.810 89.8 0.382
Ethyl alcohol C2H5OH 1.280 469.6 1.022
Ethyl chloride C2H5C1 0.885 369.0 0.644
E t h y l ether (C2H5)20 0.952 380.8 0.687
Freon 11 CCljF 0.897 388.4 0.652
Freon 12 OOlg-T 2 0.869 232.7 0.516
Freon 21 CHCI2F 0.891 353.3 0.621
Freon 113 . • CadsFs 0.962 417.4 0.724
Isobutane C4H10 0.879 273.0 0.551
Methane CH4 0.716 -116.5 0.211
Methyl alcohol CH3OH 1.179 464.0 0.936
Methyl chloride CH3C1 0.845 289.4 0.543
n-Butane C4H10 0.897 307.0 0.591
Nitrogen dioxide NO2 0.851 97.7 0.417
n-Propyl alcohol nC3H70H 1.303 506.8 1.083
Propane CgHg 0.851 206.2 0.486
Sulfur dioxide SO2 0.949 315.0 0.631
Toluene CrHg 0.956 609.1 0.877
* D. H. Gordon, Univ. of Wis., Ph.D. Thesis (1942).

t Reference substance, water.
Illustration 4. Estimate the heat of vaporization of Freon 12 (difluorodichloro-

methane) at 200°F.
From Table XIII the critical temperature of Freon 12 is 232.7°F and the value of
T„
Sr —, is 0.516. The molecular weight is 121.
r^ = ^ = 0.951
692.7
At T, = 0.951 the molal heat of vaporization of water from Table X I I is 7707 Btu
per pound-mole.
From Equation (35),
X = s , - 4 X' = (0.516) (7707) = 3977 Btu per pound-mole

3977
or - r ^ = 32.8 Btu per pound
This is in agreement with published data.

Gordon has studied the appHcation of this method to much of the
available data on heats of vaporization and found good agreement, with
errors generally less than 5%. The reliability of the method is appar-
ently of the same order as the empirical equation discussed in the previ-
ous section. It is particularly convenient to use where a number of
heat of vaporization values at different conditions are required for a
single substance.
Another good method for the prediction of heats of vaporization at
all conditions has been developed by Meissner.^ This method is appli-
cable to all substances and has approximately the same accuracy as the
methods herein described.
EVALUATION OF ENTHALPY
As pointed out on page 207, the absolute enthalpy or energy content
of matter is unknown. However, the enthalpy of a given substance
relative to some reference state can be calculated from its thermo-
physical properties. This state can be taken arbitrarily as a tempera-
ture of 0°C (32°F), atmospheric pressure, and the state of aggregation
normally existent at this temperature and pressure. The reference
state for steam is usually taken as the liquid state, under its own vapor
pressure, at 0°C.
The enthalpy of a substance is calculated as the change in enthalpy
in passing from the reference state to the existing conditions. As
previously pointed out, at constant pressure the increase in enthalpy
is equal to the heat absorbed. Ordinarily at moderate pressures the
effect of pressure on the enthalpy of liquids and sohds may be neglected
except when conditions are close to the critical point. This subject
is discussed in Chapter X I I .
Illustration 6. Calculate the enthalpy of 1 lb of steam at a temperature of 350°F
and a pressure of 50 lb per sq in., referred to the liquid state at 32°F.
Solution. From the vapor-pressure data of water (Table I) it is found that the
saturation temperature under an absolute pressure of 50 lb per sq in. is 281 °P. The
steam is therefore superheated 69°F above its saturation temperature. The enthalpy
will be the heat absorbed in heating 1 lb of liquid water from 32°F to 281 °P, vaporizing
it to form saturated steam at this temperature, and heating the steam at constant
pressure to a temperature of 350°F. The total enthalpy is the sum of the enthalpy
of the liquid, the latent heat of vaporization, and the superheat of the vapors. The
effect of pressure on the enthalpy of the liquid water is neglected.
^ H. P. Meissner, Ind. Eng. Chem., 33,1440 (1941), with permission.
-* •' I • : \
\ -
CHAP. VII] EVALUATION OF ENTHALPY
The mean specific heat of water between 32°F and 281°P is 1.006. The mean •
heat capacity of water vapor between 281°F and 350°F at a pressure of 50 lb per
sq in. is 9.2 Btu per lb-mole per °F. The latent heat of vaporization of water at
281 °F is 926.0 Btu per lb. .'
Enthalpy of liquid water at 281°F =
(281 - 32) 1.006 = 250.3 Btu per lb
Heat of vaporization at 281°F 926.0 Btu per lb
9 20
Superheat of vapor = (350 - 281) — = 35.2 Btu per lb
18
Enthalpy 1211.5 Btu per lb
Extensive steam tables have been compiled giving the enthalpies and
other properties of steam under widely varying conditions, for both
saturated and superheated vapors. In calculating these tables it is
necessary to take into account the variation of the heat capacity with
pressure, as discussed in Chapter XII. Tables and charts of enthalpies
have been worked out for a number of substances for which frequent
thermal calculations are made in engineering practice.
Calculations of enthalpy often include several changes of state. For
example, in calculating the enthalpy of zinc vapor at 1000°C and atmos-
pheric pressure, relative to the solid at standard conditions, it is
necessary to include the sensible enthalpy of the soUd metal at the
melting point, the latent heat of fusion, the heat absorbed in heating the
liquid from the melting point to the normal boiling point, the latent
heat of vaporization, arid the heat absorbed in heating the zinc vapor
from the boiling point up to 1000°C at constant pressure.
Illustration 6. Calculate the enthalpy of zinc vapor at 1000°C and atmospheric
pressure, relative to the solid at 0°C. Zinc melts at 419°C and boils under atmos-
pheric pressure at 907°C.
The mean heat capacities of the soUd and liquid may be estimated from Fig. 38,
page 218.
Mean specific heat of solid, 0=0 to 419''C = 0.105
Mean specific heat of liquid, 419''C to QOT'C = 0.109
From Table IX, page 228, the heat of fusion is 1660 calories per gram-atom.
The heat of vaporization at the normal boiling point may be estimated from
Equation (25)
„ ... , X/1180 = 8.75 -I- 4.571 log 1180 = 22.80
X = 26,900 calories per gram-mole
Since zinc vapor is monatomic its molal heat capacity at constant pressure is con-
stant and equal to 4.97 calories per gram-mole
Heat absorbed by solid = 0.105(419 — 0) = 44 calories per gram
^ ., • 1660 . „, , .
Heat of fusion = = 25 calones per grant
65.4
240 THERMOPHYSICS [CHAP. V H
Heat absorbed by liquid =

0.109(907 - 419) = i>... 53 calories per gram
„ , . . 26,900
Heat of vaporization = = 412 calories per gram
00.4
4.97
Heat absorbed by vapor = -^-— (1000 — 907) = .. 7 calories per gram
65.4
Total enthalpy....; 541 calories per gram
Frequently it is difficult to determine experimentally the individual

heats of transition involved in heating a substance. Under such condi-
tions the enthalpy is measured directly and tabulated as such for various
temperatures. For example, the enthalpy of steel at various tempera-
tures is determined by cooling in a calorimeter from these initial tem-
peratures. This determination includes all heats of transition undergone
in the cooling process.
ENTHALPY OF HUMH) Am
The properties of humid air are conveniently expressed on the basis

of the weight of humid air which contains either 1 pound or 1 pound-
mole of moisture-free air. The enthalpy of the quantity of humid air
containing a unit quantity of moisture-free air is the sum of the sensible
enthalpy of the dry air and that of the water vapor which is associated
with it. The reference states ordinarily chosen are air and liquid water
at 0°C. The water vapor in the air may be coiisidered as derived from
liquid water at 0°C by the following series of processes:
1. The liquid water is heated to the dew point of the humid air.
2. The water is vaporized at the dew point temperature to form
saturated vapor.
3. The water vapor is superheated to the dry-bulb temperature of
the air.
The enthalpy of the water is the sum of the heat absorbed by the
liquid, the heat of vaporization at the dew point, and the superheat
absorbed by the vapor.
Illustration 7. Calculate the enthalpy, per pound of dry air, of air at a pressure
of 1 atmosphere, a temperature of 100°F, and with a percentage humidity of 60%.
Solution. From the humidity chart, Fig. 8, page 100, it is seen that air under
these conditions contains 0.0345 pound-mole of water per mole of dry air or -^ ^^—
= 0.0215 pound of water per pound of dry air. This corresponds to a dew point of
79°F.
From Fig. 36, the mean molal heat capacity of water vapor between 79°F and
100°F is 8.02 and that of air between 32°F and 100°F is 6.95.
CHAB. Y I I ] HUMID HEAT CAPACITY OF AIR 241
The heat of vaporization at 79°F may be estimated from Fig. 8 as 18,840 Btu
per pound-mole or 1046 Btu per pound.
6.95
Sensible enthalpy of air = (100 - 32) — - = 16.3 Btu
Sensible enthalpy of liquid water = (79 - 32)0.0215 = .. 1.0 Btu
Latent heat of water = 1046 X 0.0215 = 22.5 Btu
Superheat of water vapor = (100 - 79) X 0.0215 X
8-02 „ „
-—- = 0.2 Btu
18
Total enthalpy 40.0 Btu per lb
' of dry air
Humid Heat Capacity of Air. It has been pointed out that when
dealing with humid air it is convenient to use 1 pound or 1 pound-mole
of dry air as the basis of calculations, regardless of the humidity of
the air. In problems dealing with the heating or cooling of air where
no change in moisture content takes place the total change in enthalpy
is equal to the sum of the change in the sensible enthalpy of the dry
air and the change in sensible enthalpy of the water vapor. For exam-
ple, in heating 1 pound of dry air associated with H pounds of water
vapor from h to U degrees Fahrenheit, the total heat, q, required is
given by the equation,
•' ' q = CUti -k)+H (C^) ik - h) (36)
where '^ -.»;i j-.-iv! : ',./;,>• w
Cpa = the mean specific heat of air at constant pressure
Cpw = the mean specific heat of water vapor at constant
pressure
Instead of considering the air and water vapor separately it is convenient
to employ a heat capacity term which combines the two.
Thus, ;:,.-:, q = S{t2-ti) (37)
where
S = heat capacity of one pound of dry air and of the
water associated with it, expressed in Btu per
^j pound of dry air per degree F
Combining (36) and (37)
S = Cpa + HCpw (38)
The combined heat capacity, S, is termed the humid heat capacity of the
air. Over the low temperature range from 30° to 180°F the mean heat
capacity of dry air is 0.240 Btu per pound and that of water vapor is
0.446 Btu per pound, from Fig. 36. Accordingly the humid heat
242 ' THERMOPHYSICS [CHAP. VII
capacity of air when expressed in Btu per pound of air per degree Fahren-
heit is given by the equation, 0;,
,S = 0.240 + 0.446Zf ,(39)
Adiabatic Humidification. In the discussion of the humidity chart,

Chapter IV, page 99, it was pointed out that a Une of constant wet-bulb
temperature also represents the relationship between dry-bulb tempera-
ture and humidity in the adiabatic vaporization of water into air. This
relationship may be derived from the thermophysical data of the system
and is used in locating the wet-bulb temperature lines on the humidity
chart. The same procedure may be used to establish wet-bulb tempera-
tures in any other system of liquid and gas in which it is known that the
wet-bulb temperature does not change appreciably during adiabatic
vaporization. The derivation is as follows:
When air is cooled by the adiabatic vaporization of water into it, sensible heat is
a
derived from the humid air to supply the heat necessary in vaporizing the^ water at
the wet-bulb temperature and in heating the evolved vapor to the existing dry-bulb
temperature. Since the total enthalpy of the system remains constant, the heat
lost by the humid air must equal that gained by the water in vaporization and super-
heating. This equality may be expressed mathematically for the evaporation of
AH moles of water into humid air containing 1 mole of dry air. Thus,
dB.{\ + Cp„(< - <„)] = - S d < • ,' (40)
where,
H = molal humidity
X = molal heat of vaporization at temperature <„
Cpw = mean molal heat capacity of water vapor
I = dry-bulb temperature
t„ = temperature of adiabatic evaporation
(S = mean molal humid heat capacity of air
Assuming that the wet-bulb temperature remains constant, as humidification pro-
ceeds the final dry-bulb temperature reached by the entire weight of air will be the
wet-bulb temperature i^, corresponding to saturation and a humidity i7„. In the
temperature range from 32° to 200°F the molal heat capacities of air and water
vapor may be taken from Fig. 36 as constant at 6.95 and 8.04, respectively. Then,
from Equation (38), <S = 6.95 -(- 8.04ff. Substituting these values in Equation (40),
and rearranging,
— ^ ^ _ = ^^ (41)
6.95 4-8.045 X-I-8.04(« - i;„) ^ '
Integrating between the limits B., i and H„, t„,
1 , 6.95 + 8.04ff 1 , X /
8.04• In6.95-(-8.04ff„
——-. „ . „ = 8.04 -—-: InX: -t- 8.04 (« - «„) /
or 6.95X -1- 8.04Xff -1- 8.04 (« - «„) (6.95 -{• 8.04ff) = 6.95X -f- 8.04Xff„
(ff„ - ff)X ,
'=6.95+8.04g+'" . (*2)
CHAP. VII] ENTHALPY OF HUMID AIR 243
The temperature, <„, of adiabatio evaporation corresponds to the experimental

value of wet-bulb temperature provided evaporation from the wet-bulb thermometer
proceeds adiabatically, that is, with no gain or loss of heat by radiation, and also
provided the actual vapor-pressure equilibrium is established at the liquid-air inter-
face. The first condition is realized where the air is passed rapidly over the wet-
bulb thermometer such that radiation errors become negligible; the second condition
is true where the rate of evaporation by diffusion keeps pace with the rate of heat
transfer by convection. This latter condition is realized without appreciable error
at temperatures below 150°F. The lines of adiabatic evaporation are therefore
commonly referred to as wet'bulb temperature lines and will be so designated.
The adiabatic cooling or wet-bulb temperature lines of Fig. 8 were

constructed from Equation (42). Corresponding to a selected value of
tw, values of dry-bulb temperature were calculated to correspond to
various humidities, thus establishing a complete curve. The wet-bulb
temperature lines of Fig. 8 which apply to gases of appreciable carbon
dioxide content were constructed from a similar equation in which the
effect of the carbon dioxide on the humid heat capacity of the gas was
considered. The molal heat capacity of carbon dioxide may be assumed
to be 9.3 (Fig. 36). Then,,
S = 6.95 {I -x) + 9.3x + 8.04/? (43)
where
• X = mole fraction of CO2 in the dry gas
"With this modification Equation (42) becomes
^ ~ 6.95 (l-x)+ 9.Sx + 8MH ^ " ^ ^

This equation permits calculations of adiabatic cooling or wet-bulb
temperature lines to apply to combustion gases or other mixtures con-
taining appreciable amounts of carbon dioxide.
Enthalpy of Humid Air. The ordinary psychrometric chart is limited
to direct use at an atmospheric pressure of 29.92 inches Hg. For other
pressures different sets of wet-bulb temperature and percentage (or
relative) humidity lines are required. W. Goodman^ has designed a
psychrometric chart which covers a range of atmospheric pressures from
22 to 32 inches Hg as shown in Fig. 46. In this chart lines of constant
enthalpy have been constructed instead of the usual wet-bulb tempera-
ture lines, where enthalpy of the humid air is expressed on the basis of
one pound of dry air. The horizontal lines in this chart represent abso-
lute humidities; the diagonal lines, constant enthalpies; the nearly ver-
tical lines, dry-bulb temperatures; and the curved lines, humidities at
saturation for various constant atmospheric pressures. The tempera-
* Air Conditioning Analysis, Maomillan (1943), with permission.
ture lines are given a slight slope to allow for the increase in heat capacity
of air and water vapor with temperature and thus avoid curvature in
the constant enthalpy lines.
Barometer
20 30 40 60 60 70 80 90
Dry Bulb Temperature °P
P I G . 46. Enthalpy chart for water vapor-air system. From W. Goodman, Air
Conditioning Analysis, Macmillan (1943), with permission.
In adiabatic evaporation the wet-bulb temperature is constant and

the system including the humid air and the water to be evaporated is
hence at constant enthalpy. At low temperatures the enthalpy of the
liquid water to be evaporated is negligible so that the enthalpy of the
i
CHAP. VII] . PROBLEMS 246
humid air is nearly equal to that of the system and hence the slope of
the constant enthalpy lines is nearly equal to that of the constant wet-
bulb lines. The constant enthalpy lines instead of wet-bulb tempera-
ture lines have the advantage that the humidity lines are nearly inde-
pendent of atmospheric pressure and may be used directly in establishing
heat requirements in air-conditioning problems. In Fig. 46 wet-bulb
temperature Unes have been constructed for 70°F and for 40°F. It will
be observed that the wet-bulb temperature lines have nearly the same
slope as the enthalpy lines, and become more nearly the same as the
temperature is lowered. It will also be observed that the location of a
given wet-bulb temperature Une depends upon the atmospheric pressure.
The slopes of these particular lines were obtained from Equation (42);
other lines can be drawn in similarly or the slopes may be estimated
from the existing 40° and 70°F wet-bulb temperature lines.
PROBLEMS
1. (a) From the data of Table V, page 214, calculate the mean heat capacity of
.' one of the following gases; oxygen, hydrogen, water vapor, sulfur
dioxide, ammonia:
{1) In kilocals per kilogram per degree Centigrade from 0° to fC.
(^) In Chu per pound per degree Centigrade from 0° to t°C.
(S) In kilocals per cubic meter per degree Centigrade from 0° to t°C.
(,4) In kilocals per kilogram per degree Centigrade from 1000° to 2000°C.
(5) In Btu per pound-mole per degree Fahrenheit from 32° to i°F.
!' • = \ (6) In Btu per cubic foot per degree Fahrenheit from 32° to i°F.
(7) In Btu per pound per degree Fahrenheit from 1000° to 2000°F.
(b) Calculate the heat capacity of the assigned gas in kilocals per kilogram per
degree Centigrade at 1500°C. '. . ,„
2. From the experimental data for the molal heat capacities of oxygen at various
temperatures derive the constants in the following types of empirical equations over
the temperature range from 0°C to 1500°C: >. • > ^ i
c„ = a +bT + cT^ " ".
b e '.'
• • • • ' • • I ' ' * 1
b e ,, ,
' Cp = a +—7= +— •.'...
. ,.-. -^..; i^ ^ V -'^-K^:-: • VT T . s;'
3. Calculate the amount of heat given off when 1 cu m of air (standard condi-
tions) cools from 500° to — 100°C at a constant pressure of 1 atmosphere, assuming
the heat capacity formula of Table V, page 214, to be valid over this temperature
range. ,., ,,
246 THERMOPHYSICS • [CHAP. VII
4. Calculate the number of kcals required to heat, from 200° to 1200°C, 1 cu m

(standard conditions) of a gas having the following composition by volimie:
CO2 '. 20%
N2 77%
O2 2%
H2 1%
6. Calculate the number of Btu required to heat 1 lb each of the following liquids
from a temperature 32° to 100°F:
(a) Acetone.
(b) Carbon tetrachloride.
(c) Ether. '
(d) Propyl alcohol.
6. Calculate the number of calories required to heat 1000 grams of each of the
following aqueous solutions from 0° to 100°C.
(a) 5 % N a C l b y weight. • ;' :, , ; , ; , - , '•.
(b) 20% NaCl by weight. ' '
(c) 20% H2SO4 by weight. '
(d) 20% KOH by weight. :' _
(e) 20% NH4OH by weight. ' .- .
(f) 20% Pb(N03)2 by weight.
7. From Fig. 38 determine the heat required to raise 1 pound of graphite from
32°F to 1450°F. Show how the graph is utilized to determine the quantity of heat
required.
8. Calculate the specific heat at 20°C of MgO, SiOa (quartz), CaO, CuO, PbO
from Kopp's rule and compare with the experimental values.
9. Calculate the heat equivalent in Btu of the external work of vaporization of
1 lb of water at a temperature of 80°F, assuming that water vapor follows the ideal
gas law.
10. Calculate the heat of vaporization in Btu per pound of diethyl ether
(C2H6OC2H6) at its normal boiling temperature by the following methods;
(a) From the equation of Kistyakowsky.
(b) From Equation (29).
(c) From Equation (35).
11. Obtaining the necessary boiling-point and critical data from Fig. 4, estimate
the heat of vaporization, in Btu per pound of n-butane (C4H10) at a pressure ©f 200
lb per sq in.
(a) By Equations (25) and (32).
(b) By Equations (29) and (32).
(c) By Equation (35) and Tables XII and XIII.
12. Using Fig. 4 for the necessary boiling-point and critical data, estimate the heat
of vaporization, in Btu per pound of carbon disulfide (CS2) when under a pressure
of 77.5 lb/in.2 abs using the three methods of Problem 11.
13. Cyclohexane (C6H12) has a normal boiling point of 80.8°C and at this tempera-
ture the density of the liquid is 0.719 grams per cc. Estimate:
(a) The critical temperature.
(b) The boiling point at 10 atmospheres pressure.
(c) The heat of vaporization at 10 atmospheres pressure, expressed as Btu per
pound.
CHAP. VII] PROBLEMS , 247
14. Utilizing the thermal data for diphenyl, (CeHe-CeHs) tabulated below, estimate
the following:
(a) Critical temperature.
(b) Boiling point at 25 lb per sq in. •
(c) Heat of vaporization at 25 lb per sq in., as Btu per lb.
(d) Enthalpy of 1 lb oi saturated diphenyl vapor at 25 Ib/sq in. relative to solid
diphenyl at 32°F.
Data for diphenyl
Normal boiling point 255°C
Density of the liquid at the normal
boiling point 0.75 gram/cc
Melting point 71°C
Specific heat of solid diphenyl 0.385 Btu per lb per deg. F
Heat of fusion 46.9 Btu per lb
Specific heat of liquid diphenyl.... Cp = 0.300 + 0.00120i°C
15. Calculate the heat of vaporization, in calories per gram, of water at a

temperature of 100°C by means of the Clapeyron equation. At this temperature
dp/dt = 27.17, where p is the vapor pressure in millimeters of mercury and t is the
temperature in degrees Centigrade.
16. The vapor pressure of zinc in the range from 600° to 985°C is given by the
equation
6160
log P = - - ^ + 8.10
where.
p = vapor pressure, millimeters of mercury
T = temperature, degrees K
Estimate the heat of vaporization of zinc at 907''C, the normal boiling point.
Compare this result with that calculated from the equation of Kistyakowsky.
17. From the International Critical Tables obtain the following data:
(a) The heat capacity, in calories per gram per degree C, of
(1) Liquid o-nitroaniline (CeHzNaOa) at 100°C.
(2) Liquid SiCL at 25°C.
(S) A solution containing 50 mole per cent ether (C4H10O) in benzene
(CeHe) at a temperature of 20°C.
(4) Solid FeS2 at 100°C. —
(b) The heat of fusion of
(1) BaCla. •
(2) Benzoic acid (C7H6O2). •
(3) Stearic acid (C18H36O2).
(c) The heat of vaporization of v
(;) Nitrogen at -202°C.
(2) SiCU at 57°C.
(5) n-octyl alcohol (CsHuO) at 196°C.
18. Calculate the enthalpy in kcals per kilogram referred to the solid at 0°C, of
molten copper at a temperature of 1200°C.
248 . THERMOPHYSICS [CHAP. V I I
19. Obtaining the latent heat data from t h e steam tables, calculate the enthalpy,
in B t u per pound relative to the liquid at 32°F, of steam at a temperature of 500"?
superheated 200''F above its saturation point.
20. Calculate the enthalpy in Btu per pound relative t o 32°F, of pure molten
iron at a temperature of 2 8 5 0 ^ . I n heating iron from 32°F to its melting point it
vmdeTgoes three transformations, from a t o |3, from. j3 t o 7, and from 7 t o 8 forms.
21. Using the latent heat data calculated in Problem 14, calculate the enthalpy
in Btu per pound relative to the solid at 32°F, of saturated diphenyl vapors under a
pressure of 40 lb per sq in.
22. Calculate the enthalpy in Btu per pound of dry air relative t o air and liquid
water at 32°F, of humid air at a temperature of 150°F, a pressure of 1 atmosphere, and
a percentage humidity of 4 0 % .
23. Humid air at a pressure of 1 atmosphere has a dry-bulb temperature of 180°F
and a wet-bulb temperature of 120°F. This air is cooled t o a dry-bulb temperature
of 115°F. Calculate the heat evolved, in Btu per pound of dry air.
24. Hot gases are passing through a chimney at a rate of 1200 eu ft per minute,
measured at the existing conditions of 600°C and a pressure of 740 mm of Hg. The ^'"' v
gases have t h e following composition by volume on the dry basis: - ''~^-^.
CO2 12% • •• •;*•'!
— •• N2 80% Vi: ;: ;v.A.
O2 8%
The dew point of the gases is 50°C and they contain 20 grams of carbon soot per
cubic meter measured at the chimney conditions. Calculate the enthalpy of t h e
material passing through the chimney per minute, in B t u relative to gases, solid
carbon, and liquid water a t 18°C. ,
CHAPTER VIII
' . THERMOCHEMISTRY
All chemical reactions are accompanied either by an absorption or
evolution of energy, which usually manifests itself as heat. That
branch of science which deals with the changes of energy in chemical
reactions is called thermochemistry.
As discussed in Chapter VII the internal energy of a given substance
is dependent upon its temperature, pressure, and state of aggregation
and is independent of the means by which this state was brought about.
Likewise the change in internal energy, At/, of a system which results
from any physical change or chemical reaction depends only on the
initial and final state of the system. The total change in internal
energy will be the same whether or not energy is absorbed or evolved in
the form of heat, radiant energy, electrical energy, mechanical work,
or other forms.
For a flow reaction proceeding with negligible changes in kinetic
energy, potential energy and with no electrical work and no mechanical
work beyond that required for flow, the heat added is equal to the
increase in enthalpy of the system, i- i)-;:; H, k •-.>.(:! ;
g = AH • ' (1)
For a non-flow reaction proceeding at constant pressure the heat added
is also equal to the gain in enthalpy,
' . ... ^ g-Ai/ • - (2)
For a non-flow reaction proceeding at constant volume the heat added
is equal to the gain in internal energy of the system,
,,,,,,,„^, : , . : . , ^ , : - - , ,,.... g = At/ -,.;•:., ^ 5 . • (3)
Standard Heat of Reaction. The heat of a chemical reaction is the

heat absorbed in the course of the reaction, or, in a more general sense
it is equal to the change in enthalpy of the system for the reaction pro-
ceeding at constant pressure. This heat of reaction is dependent not
only on the chemical nature of each reacting material and product but
also on their physical states. For purposes of organizing thermo-
chemical data it is necessary to define a standard heat of reaction which
may be recorded as a characteristic property of the reaction and from
which heats of reaction under other conditions may be calculated. The
250 THERMOCHEMISTEY [CHAP. V I I I
standard heat of reaction is defined as the change in enthalpy resulting from

the procedure of the reaction under a pressure of 1 atmosphere, starting
and ending with all materials at a constant temperature of 18°C. A tem-
perature of 25°C is also frequently used as a standard reference tem-
perature. /
For example, 1 gram-atom (65.38 grams) of zinc may be allowed to
react with 2073 grams of 1.0 molal aqueous hydrochloric acid containing
2.0 gram-moles of HCl. The reaction may be carried out in a calorim-
eter under atmospheric pressure with all reactants at an initial tem-
perature of 18°C. During the course of the reaction the system will
become heated, hydrogen gas will be evolved, and a 0.5 molal solution
of zinc chloride will be formed. When the reaction is completed the
resultant solution and the hydrogen gas may be cooled to 18°C. If no
evaporation of water takes place it will be found that 35,900 calories
will be evolved by the system. The net result of the reaction is the
conversion of 2 moles of hydrochloric acid in aqueous solution into 1
mole of zinc chloride in aqueous solution and 1 mole of hydrogen gas
at atmospheric pressure, all at a temperature of 18°C. The measured
amount of heat absorbed represents the standard molal heat of reaction
for this particular reaction, proceeding under atmospheric pressure in
an aqueous solution of the specified concentration.
Exactly the same net result is obtained by allowing the above reac-
tion to proceed in an electrolytic cell in which one electrode is zinc and
the other platinum. An electric motor might be connected to the cell
and be permitted to do work as the reaction proceeds. In this case
the amount of heat evolved will be less than 35,900 calories by the heat
equivalent of the electrical energy produced by the cell. However,
the heat of reaction is the same and equal to the algebraic sum of the
amounts of energy absorbed as heat and as electrical energy.
Conventions and Symbols. As pointed out in the preceding section,
the heat of reaction accompanying a chemical change is dependent on
the physical state of each reactant and product as well as on its chemical
nature. For this reason, in order to define a heat of reaction it is neces-
sary to specify completely the nature and state of each material in-
volved. The following system of conventions and symbols, to be used
in conjunction with the conventional chemical equation, is adopted for
this purpose.
The formula of a substance appearing in an equation designates not
only the nature of the substance but also the quantity which is in-
volved in the reaction. Thus, H2SO4 indicates 1 mole of sulfuric acid,
and I j N2 indicates 1^ moles of nitrogen. All equations are written
with the reactants on the left and the products on the right side. The
CHAP. VIII] CONVENTIONS AND SYMBOLS 251
value of the heat of reaction accompanying an equation is the heat of

reaction resulting from the procedure of the reaction from the left to
the right of the equation as written. If the reaction proceeds in the
reverse direction the heat of reaction is of opposite sign.
Unless otherwise specified it is assumed that each reactant or product
is in its normal state of aggregation at a temperature of 18°C and a
pressure of 1 atmosphere.
The state of aggregation of a substance is indicated by a letter in
parentheses following its chemical formula. Thus (g) indicates the
gaseous state, (Q the liquid, and (s) the solid.
Additional information may accompany these letters in parentheses.
Thus, S (rhombic) and C (diamond) indicate sulfur in the rhombic
state and carbon as diamond, respectively, while S (monoelinic) and C
(graphite) indicate monoelinic sulfur and solid graphitic carbon. In
the case of a gas the pressure may be specified. Thus, CO2 (g, 2 atm)
indicates gaseous carbon dioxide under a pressure of two atmospheres.
The concentration of a substance in aqueous solution is indicated by
its molahty (m), by the number of moles of solvent (rii) per mole of
solute; or by the mole fraction of the solute (A^2). Thus, (m = 0.1)
following a chemical formula indicates that the substance is in aqueous
solution with a molality of 0.1. The symbol (ni = 200) indicates an
aqueous solution with 200 moles of water per mole of solute. The
symbol (1^2 = 0.55) indicates an aqueous solution in which the mole
fraction of the solute is 0.55. If the aqueous solution is highly dilute,
such that additional dilution produces no thermal effect, the symbol
(aq) follows the formula of the solute. '
The concentration of a substance in nonaqueous solution is indicated
by the above symbols, accompanied in the parentheses by the formula
of the solvent. Thus, (C2H6O, iVa = 0.55) indicates that a substance is
in alcoholic solution with a mole fraction of 0.55.
)Ionic reactions are indicated in the usual manner, for example, H+
and Ca++ for the positive hydrogen and calcium ions and C\' and SO4 ~
for the negative chloride and sulfate ions, respectively.
As previously pointed out, positive values of q represent an ab-
sorption of heat by the system under consideration, that is, an increase
in enthalpy of the system, q = AH. Where the initial and final tem-
peratures of the system are the same, a subscript may be used to desig-
nate this temperature, thus AHis is the heat of reaction, or change in
enthalpy^ at 18°C and 1 atmosphere pressure.
When heat is evolved in a reaction, corresponding to a decrease in
enthalpy, the reaction is termed exothermic; when heat is absorbed the
reaction is said to be endothermic. 'i -, , , =
252 THERMOCHEMISTRY (CHAP. VIII
With the aid of the above symbols the states of a chemical reaction
are indicated by the following:
Zn(s) + 2HCl(m = 1.0) = ZnCUCm = 0.5) + HzCff, 1.0 atm)
AHii = —35,900 calories per gram-mole
This equation designates the changes occurring in the reaction described
in the preceding section.
Heat of Formation. The heat of formation of a chemical compound
is a special case of the standard heat of a chemical reaction wherein the
reactants are the necessary elements and the compound in question is
the only product formed. Heats of formation are always expressed
with reference to a standard state. The molal heat of formation of a
compound represents, unless otherwise stated, the heat of reaction,
A H / , when 1 mole of the compound is formed from the elements in a
reaction beginning and ending at 18°C and at a pressure of 1 atmosphere
with the reacting elements originally in the states of aggregation which
are stable at these conditions of temperature and pressure. The heat
of formation of a compound is positive when its formation from the
elements is accompanied by an increase in enthalpy. A compound
whose heat of formation is negative is termed an exothermic compound.
If the heat of formation is positive it is called an endothermic compound.
For example, the molal heat of formation of liquid water is —68,320
calories per gram-mole. This means that when 2.016 grams of hydrogen
gas combine with 16 grams of oxygen at a temperature of 18°C and a
pressure of 1 atmosphere to form 18.016 grams of liquid water at the
same temperature, the heat given off to the surroundings is 68,320 calo-
ries and the enthalpy of the system is decreased by 68,320 calories. It is
obvious that this reaction will not proceed at a constant temperature
but during its progress will be at a very high temperature, and the prod-
uct formed will be temporarily in the vapor state. However, upon
cooling to 18°C this sensible and latent heat appearing temporarily in
the system itself is evolved and included in the heat of formation. If
water vapor were the fioial product at 18°C the heat of formation would
be numerically less by an amount equal to the heat of vaporization of
water at 18°C. The heat of vaporization of water at 18°C is 10,565
calories per gram-mole. Therefore the heat of formation of water vapor
at 18°C is -68,320 -|- 10,565 = -57,755 calories per gram-mole.
The basic thermochemical data of inorganic compounds are always
presented in terms of standard heats of formation. In the International
Critical Tables, Vol. V, page 169, are extensive tables giving the heats
of formation of a great variety of inorganic compounds, both in pure
states and in solutions of varying concentrations.
CHAP. VIII] HEAT OF FORMATION 253
TABLE XIV
HEATS OF FORMATION AND SOLUTION
,, Reference Conditions: 18°C and 1 atmosphere pressure

AH/ = heat of formation
AH, = heat of solution
in kilocalories per gram-mole
Multiply values by 1000 to obtain kilocalories per kilogram-mole and by 1800 to
obtain Btu per pound-mole.
Data taken from Bichowski and Rossini, Thermochemistry of Chemical Substances,
Reinhold Publishing Corp. (1936), with permission.
Abbresiations
c = crystalline state dil >= dilute solution
8 = solid a> = iniinite dilution
I = liquid ppt •= precipitated solid
g = gas amorph *= amorphous state
AH/ Moles AH,

Compound Formula State Heat of of Heat of
Formation Water Solution
Acetic acid CH3COOH I -117.4 200 -0.345

A l u m i n u m cliloride AICI3 c -166.8 600 -77.9
Aluminum hydroxide A1(0H)3 ppt -304,9
Aluminum nitride AIN c -80.0
A l u m i n u m oxide AI2OS c -380.0
A l u m i n u m silicate AkOs-SiOa sillimanite -623.7
A l u m i n u m disilicate Al203-2Si02 arnorph -979.0
A l u m i n u m disilicate Al203'2Si02-2H20 amorph -944.0
T r i a l u m i n u m disilicate 3Al203-2SiOa c -1804.0
A l u m i n u m sulfate Al2(S04)j c -770.0 CD
-1.26
Ammonia NH3 0 -11.0 200 -8,35
Ammonia NHs I -16.07 200 -3.28
Ammonium carbonate (NH4)2CO, dil -223.4
Ammonium bicarbonate NH4HCO3 c -202.8 400 +6.81
A m m o n i u m chloride NH4CI c -74.95 » +3.818
Ammonium hydroxide NH4OH -87.67
Ammonium nitrate NH4NO3 c -87.13 00
+6.47
Ammonium oxalate (NH4)2C204 c -266.6 400 +7.90
A m m o n i u m sulfate (NH4)2S04 c -281.46 400 +2.38
A m m o n i u m acid sulfate NH4HS04 c -244.64 800 -0.56
A n t i m o n y trioxide Sb203 c -165.4 00 -0.6
Antimony pentoxide SbaOs c -230.0 CO +2.0
A n t i m o n y sidfide Sb2S3 c -35.7
Arsenic acid H3ASO4 c -214,9 CO +0,4
A t s e n i c trioxide A820j octahedral -154.0 CO +7.5
Arsenic p e n t o x i d e AsjOs c -217.9 CO -6.0
Barium acetate Ba(C2H302)2 c -357.6 CO -5.75
Barium carbonate BaC03 ppt -290.8
Barium carbonate BaCOs c -290.7
B a r i u m chlorate BaCC103)2 c -176.6 CO +6.74
B a r i u m chloride BaCl2 c -205.28 CO -2.45
B a r i u m chloride BaCl2-2H20 c -349.0 CO +4.80
Barium hydroxide Ba(0H)2 c -225.9 400 -11.78
B a r i u m oxide BaO c -133.0 CO -36.1
B a r i u m peroxide Ba02 c -162.4
B a r i u m silicate BaSiOs glass -363.0
B a r i u m sulfate BaS04 -: ppt -349.4 00
+5,24
B a r i u m sulfide BaS c -111.2 133
-7.16
254 THERMOCHEMISTRY [CHAP. VIII
TABLE XIV — Continued
AH/ JMoles AHa

Compound Formula State H e a t of of H e a t of
B i s m u t h oxide BiaOs c -137.1

B o r i c acid HsBOs c -251.6 CO +5.4
B o r o n oxide c -279.9 a -7.3
B r o m i n e chloride BrCI g +3.07
C a d m i u m chloride CdCla c -93.0 400 -3.06
C a d m i u m oxide CdO c -65.2
C a d m i u m sulfate CdSOj c -222.22 400 -10.69
C a d m i u m sulfide CdS c -34.6
Calcium acetate Ca(C2H302)2 -357.6 CO -7.64
Calcium aluminate CaO-AbOa glass -620.0
Calcium aluminate 2CaO-Al203 glass -857.0
Calcium aluminate SCaO-AkOs glass -1098.0
C a l c i u m a l u m i n u m silicate 3CaO-Al203-2Si03 c -1292.0
C a l c i u m a l u m i n u m silicate CaO-Al203-6Si02 c -3287.0
Calcium carbide CaC2 c -14.1
Calcium carbonate CaCOa vvt -287.8
Calcium carbonate CaCOs calcite -289.3
C a l c i u m chloride CaCl2 c -190.6 CO -18.517
C a l c i u m chloride CaCl2-6H20 c -623.45 0= +9.05
C a l c i u m fluoride CaF2 PPt -290.2
Calcium hydroxide Ca(0H)2 c -236.0 CO -3.06
Calcium nitrate Ca(N03)2 c -224.04 400 -3.94
Calcium oxalate CaC204 PVt -329.7
C a l c i u m oxide CaO c -151.7 CO -19.0
Calcium phosphate Ca3(P04)2 c -983.0
C a l c i u m silicate CaO-Si02 glass -377.9
C a l c i u m silicate 2CaO-Si02 glass -432.0
C a l c i u m sulfate CaS04 c -340.4 CO -6.14
C a l c i u m sulfide CaS c -113.4 CO -6.4
C a r b o n )9-graphite C c 0
Diamond C c + .22
Gas carbon C amorph +1.72
C h a r c o a l (Ha tree) C amorph +2.22
Coke C amorph +2.6
Sugar carbon C amorph. +2.389
Carbon monoxide CO a -26.620 sat. -2.76
C a r b o n dioxide CO2 g -94.03 Bat. -4.76
C a r b o n disulfide CS2 a +22.44
C a r b o n disulfide CS2 I +15.84
Carbon tetrachloride CCI4 Q -25.7
Carbon tetrachloride CCI4 I -33.6
C h l o r i c acid HCIO3 dil -20.8
C h r o m i u m chloride(ic) CrCl, c -143.0 CO -40.0
C h r o m i u m chloride (ous) CrCU c -103.1 CO -18.6
C h r o m i u m oxide Cr203 c -273.0
C h r o m i u m oxide Cr203 amorph -266.2
C h r o m i u m trioxidc CrOs c -139.3 80 -2.5
C o b a l t oxide CoO c -57.6
C o b a l t Oxide C03O4 . c -196.5
C o b a l t sulfide CoS ppt -22.3
Copper acetate Cu(C2H302)2 c -216.4 CO -2.4
Copper carbonate CuCOs ppt -141.4
C o p p e r chloride CuCl2 c -53.4 800 -11.2
C o p p e r chloride CuCl c -34.3
C o p p e r oxide CuO c -38.5
C o p p e r oxide CU2O c -42.5
C o p p e r sulfate CUSO4 c -184.7 800 -15.94
TABLE XIV — Continued
AH/ Moles AHe

C o p p e r sulfide CuS c -11.6
C o p p e r sulfide CU2S c -18.97
Copper nitrate Cu(N03)2 c -73.1 200 -10.3
Cyanogen C2N2 0 +71.4 —7.0
H y d r o b r o m i c acid HBr a -8.65 m —20.02
H y d r o c h l o r i c acid HCl a -22.06 CO
—17.627
H y d r o c y a n i c acid HCN e +30.90 100 -6.8
Hydrofluoric acid HF 1 -71.0 600 —4.7
H y d r i o d i c acid HI g +5.91 GO
— 19.28
H y d r o g e n oxide H2O a -57.801
H y d r o g e n oside H2O , I -6S.320
Hydrogen peroxide H2O2 I -45.2 200 —0.46
H y d r o g e n sulfide H2S g -5.3 CQ
—4.6
Iron acetate Fe(C2Hs02)ii dil -355.8
Iron carbonate FeCOs c -172.6
I r o n chloride FeCls c -96.4 1000 —31.7
Iron hydroxide Fe(OH)s ppt -197.3
I r o n oxide ; Fe203 c -198.5
I r o n oxide i FeO c -64.3
I r o n oxide , Fe304 magnetite -266.90
I r o n silicate FeO-Si02 c -273.50
I r o n sulfate Fe2(S04)j dil -653.20
I r o n sulfate FeS04 c -221.30 400 —14.68
I r o n sulfide FeS c -23.1
I r o n sulfide FeS2 pyrites -35.5
Lead acetate Pb(C2Hs02)2 c -233.4 400 -1.40
Lead carbonate PbCOs c -167.8
L e a d chloride PbCb c -85.71 CO
+3.41
Lead nitrate Pb(N03)2 c -106.89 400 +7.61
L e a d oxide PbO c -52.46
L e a d peroxide Pb02 c -65.0
Lead suboxide Pb20 c -51.3
L e a d sesquioxide Pb304 c -172.4
L e a d sulfate PbS04 c -218.5
L e a d sulfide PbS ppt -22.3
L i t h i u m chloride ,LiCl c -97.65 CO —8.665
Lithium hydroxide LiOH c -116.55 00
—4.738
Magnesium carbonate MgCOj c -267. S
M a g n e s i u m chloride MgCl2 c -153.3 800 —35.92
Magnesium hydroxide Mg(0H)2 ppt -218.7
M a g n e s i u m oxide MgO amorph (?) -146.1
M a g n e s i u m silicate MgSiOj c -347.5
M a g n e s i u m sulfate MgS04 c -304.95 400 —20.30
Manganese carbonate MnCOj ppt -207.50
M a n g a n e s e chloride MnCl2 c -112.70 400 — 16.0
M a n g a n e s e oxide MnO c -96.6
M a n g a n e s e oxide Mn304 c -345.0
M a n g a n e s e dioxide MnOj c -123.0
M a n g a n e s e dioxide MnOj amorph -115.6
M a n g a n e s e silicate MnO-SiOa c -301.3
M a n g a n e s e silicate MnO-SiOa glass -292.8
M a n g a n e s e sulfate MnS04 . c -251.2 400 —13.8
M a n g a n e s e sulfide MnS ppt -47.0
M a n g a n e s e sulfide MnS c -59.7
Mercury acetate Hg(C2H302)2 c -197.1 CO
+4.0
M e r c u r y chloride HgCl2 c -53.4 CO
+3.3
M e r c u r y chloride Hg2Cl2 ppt -63.15
TABLE XIV —Continued
AH/ Moles AB,

Mercury nitrate HeCNOsh dil -56.6

Mercury nitrate Hg2(NOs)2-2H20 c -206.5
M e r c u r y oxide HgO c -21,6
M e r c u r y oxide HgjO c -21.6 ' •
M e r c u r y sulfate HgSOi c -166.6

M e r c u r y sulfate Hg2S0i c -176.5
M e r c u r y sulfide HgS amorph -11.0
Mercury thiocyanate Hg(CNS)2 c +52.4
M o l y b d e n u m dioxide M0O2 c -130.0
M o l y b d e n u m trioxide MoO, c -176.5
Nickel c y a n i d e Ni(CN)2 c +232.4
Nickel hydroxide Ni(OH)j ppi -163.2
Nickel h y d r o x i d e Ni(OH)2 ppt -129.S ' • '
Nickel oxide NiO c -58.4

Nickel sulfide NiS ppt -20.4
N i c k e l sulfate NiSOj dil -231.1
N i t r o g e n oxide NO 9 +21.6
N i t r o g e n oxide NjO g +19.65
N i t r o g e n oxide NjO I +18.72
Nitrogen pentoxide N2O6 9 +0.6 a, -29.8
Nitrogen pentoxide N2O5 c -13.1 400 -18.61
Nitrogen tetroxide N2O4 9 +3.06
N i t r o g e n trioxide N2O, g +20.0
N i t r i c acid HNO3 I -41.66 • CO -7.53
Oxalic a c i d HjCzOi-SHaO c -339. S 300 +8.58
Oxalic acid H2C2O4 c -197.2 300 +2.3
P e r c h l o r i c acid HCIO4 I -19.35 500 -20.3
P e r m a n g a n i c acid HMn04 da -122.3
P h o s p h o r i c acid ( m e t a ) HPO3 c -224.8 CO -9.8
P h o s p h o r i c acid (ortho) HsP04 I -300.85 400 -5.31
P h o s p h o r i c acid <pyro) HiP20r I -529.4 03 -10.2
P h o s p h o r o u s acid ( h y p o ) H3PO2 I -139.0 450 -2.41
P h o s p h o r o u s acid (ortho) H3PO3 I -225.86 CO -2.94
P h o s p h o r o u s trichloride PCI3 9 -70.0
Phosphorous pentoxide P2O6 c -360.0
P l a t i n u m chloride PtCU c -62.6 CO -19.4
P l a t i n u m chloride PtCl c -17.0
Potassium acetate KC2H302 c -174.23 400 -3.404
Potassium carbonate K2CO3 c -274.24 400 -6.63
P o t a s s i u m chlorate KCiOj c -91.33 CO +10.31
P o t a s s i u m chloride KCl c -104.361 CO +4.404
Potassium chromate K2Cr04 c -333.4 2185 +4.87
Potassium cyanide KCN c -28.3 200 +3.0
Potassium dichromate K2Cr207 c -488.5 1600 +17,81
P o t a s s i u m fluoride KF c -134.51 400 -3.86
P o t a s s i u m oxide K2O c -86.2 300 -75,00
P o t a s s i u m sulfate K2S04 c -342.66 a +6.32
P o t a s s i u m sulfide K2S c —121.5 CO +10.96
P o t a s s i u m sulfite K2SO3 c -267.7 CO -1.8
P o t a s s i u m thiosulfate KiSiOrHsO c -270.5 <o +4.5
Potassium hydroxide KOH c -102.02 CO -12.91
Potassium nitrate KNO3 c -118.093 CO +8.633
Potassium permanganate KMn04 c -192.9 400 +10.4
Selenium oxide Se02 c -56.36 2000 +0.93
Silicon c a r b i d e SiC c -27.8
Silicon t e t r a c h l o r i d e SiCU I -150.1
Silicon t e t r a c h l o r i d e SiCl4 a -142.5
TABLE 'KlY — Continued
AH/ Moles ASa

Silicon dioxide SiOa quartz -203.34

Silver b r o m i d e AgBr c -23.81
Silver chloride AgCl c -30.30 CO +14.0
Silver n i t r a t e AgNOj ' c -29.4 200 +5.44
Silver sulfide Ag2S c -5.5
Sodium acetate NaC2H302 c -171.12 400 -3.935
Sodium arsenate NasASOi c -366.0 500 -15.65
Sodium tetraborate Na2B407 c -742.6 900 -10.0
S o d i u m boraK Na2B4O7'10H2O c -1453.1 1600 +25.85
Sodium bromide NaBr c -86.73 200 +0.601
Sodium carbonate Na2C03 c -269.67 400 -5.63
Sodium carbonate Na2CO3'10H2O c -975.26 400 + 16.39
Sodium bicarbonate NaHCOs c -226.2 1800 +4.1
Sodium chlorate NaClOa c -83.6 03 +5.37
S o d i u m chloride NaCl c -98.33 CO +1.164
Sodium cyanide NaCN c -22.73 200 +0.37
S o d i u m fluoride NaP c -135.95 (X> +0.271
Sodium hydroxide NaOH c -101.96 CO -10.179
S o d i u m iodide Nal • c -69.28 200 -1,562
Sodium nitrate NaNOs c -111.72 00 +5.051
Sodium oxalate Na2C204 c -314.3 600 +4.12
S o d i u m oxide Na20 c -99.45 CO -66.4
Sodium triphosphate Na3P04 c -457.0 800 -14.0
Sodium diphosphate Na2HP04 c -414.85 800 -5.41
Sodium monophosphate NaH2P04 dil -364.64
Sodium phosphite Na2HP03 c -335.33 800 -9.30
Sodium selenate Na2Se04 c -254.0 CO -7.3
S o d i u m selenide Na2Se c -59.1 440 -18.8
S o d i u m sulfate Na2S04 c -330.48 CO -0.28
S o d i u m sulfate Na2SO4-10H2O c -1033.20 400 +18.9
Sodium bisulfate NaHS04 c -269.04 800 -1.74
S o d i u m sulfide Na2S c -89.8 400 -15.16
S o d i u m sulfide Na2S-4iH20 c -417.54 1000 +5.0
S o d i u m sulfite Na2S03 c -261.2 800 -2.7
S o d i u m bisulfite NaHSOa dil -206.5
S o d i u m silicate Na2Si03 glass -374,0
S o d i u m silicofluoride Na2SiF6 c -669.0 660 +10.0
-70.92 2000 -8.56
Sulfur dioxide SO2 » -93.90 CO -53.58
Sulfur trioxide SO3 g
Sulfuric acid H2SO4 I -193.75 CO -22.05
T e l l u r i u m oxide Te02 c -77.58 CO +0.9
T i n chloride(ic) SnCU I -127.4 200 -29.9
T i n chloride(ous) SnCl2 c -81.1 200 -0.4
T i n oxide(ic) SnOj c -138.1
T i n oxideCous) SnO c -67.7
T i t a n i u m oxide Ti02 amorph -214.1
T i t a n i u m oxide Ti02 c -218.0
c -130.5
T u n g s t e n oxide WO2 /
V a n a d i u m oxide V2O6 c -437.0
Zinc a c e t a t e ZnCCjHsOjJa c -260.2 400 -9.8
Zinc c a r b o n a t e ZnCOs ppt -193.1
Zinc chloride ZnCl2 c -99.55 400 -15.72
Zinc h y d r o x i d e Zn(0H)2 amorph -153.5
Zinc oxide ZnO c -83.5
Zinc s u l f a t e ' ;:;'; ;;. ZnS04 . vr'i ' c -233.4 400 -18.55
Zinc sulfide ZnS c -44.0
Z i r c o n i u m oxide Zr02 glass -253.0
258 THERMOCHEMISTRY [CHAP VIII
In Table XIV are listed a few selected values of heats of formation.

It will be observed in these tables that the heat of formation is made
synonymous with increase in enthalpy. This results in a negative sign
for all exothermic compounds. This practice is in agreement with the
nomenclature of the American Standards Association which is here
adopted. The opposite sign for heats of formation are to be found in
some tables and in previous work of the authors.
When a compound is hydrated, its heat of formation in Table XIV
includes the heat of formation of the water forming the hydrate. For
example, the heat of formation of solid CaCU-GHaO is given as — 623,450.
This represents the heat of reaction accompanying the formation at
18°C of 1 mole of CaCla-GHBO from solid calcium and gaseous chlorine,
hydrogen, and oxygen.
In the sixth column of Table XIV are values of the heats of solution
of a few compounds. The heat of solution represents the change in . ^ i
enthalpy resulting from the formation of a solution of the specified con-
centration from 1 gram-mole of the compound and the number of gram-
moles of liquid water indicated in the fifth column. By means of these
values the total heat of formation of a dissolved material may be calcu-
lated. For example, the heat of solution of 1 gram-mole of CaCU-GHaO
in 400 gram-moles of water is -(-9050 calories. The heat of formation
of the undissolved CaCU-GHaO is —623,450 calories. The combined
heat of formation and solution of 1 gram-mole of CaCl2-6H20 in 400
gram-moles of water is the algebraic sum of these two values,
-623,450 + 9,850 = -614,400 calories.
If an element normally exists in more than one allotropic form at
18°C and atmospheric pressure, one of these forms is selected to serve
as the basis of heats of formation throughout the table. This is equiva-
lent to assigning a heat of formation of zero to the element in the par-
ticular form selected. For example, carbon may exist as graphite,
diamond, or amorphous carbon. The |3-graphite form has been selected
as the basis of the heats of formation of all carbon compounds. This is
indicated by presenting the heat of formation of /3-graphite as zero in
the table. On this basis all other forms of elementary carbon have
positive heats of formation. The heat of formation, for example, of
barium carbonate is the heat of reaction accompanying the formation
of the compound from graphite and the other necessary elements.
When more than one allotropic fonn of an element exists, the particular
form on which the tables are based can be identified as the form for which
the heat of formation is given as zero.
Laws of Thermochemistry. At a given temperature and pressure
the quantity of energy required to decompose a chemical compound
CHAP. VIII] LAWS OF THERMOCHEMISTRY 259
into ita elements is precisely equal to that evolved in the formation of

that compound from its elements. This principle was first formulated
by Lavoisier and Laplace in 1780. For example, the heat of forma-
tion of sodium chloride is —98,330 calories. Theoretically the same
amount of energy is required to decompose sodium chloride into sodium
and chlorine.
A corollary of this first principle of thermochemistry is known as the
law of constant heat summation, which states that the net heat evolved
or absorbed in a chemical process is the same whether the reaction takes
place in one or several steps. The total change in enthalpy of a system is
dependent upon the temperature, pressure, state of aggregation, and
state of combination at the beginning and at the end of the reaction
and is independent of the number of intermediate chemical reactions
involved. This principle is known as the law of Hess, formulated in
1840.
By means of this principle it is possible to calculate the heat of for-
mation of a compound from a series of reactions not involving the
direct formation of the compound from the elements. The majority
of chemical compounds cannot be prepared in the pure state directly
from the elements. For example, the heat of formation of carbon
monoxide cannot be measured directly because it cannot be prepared
in a pure state from the elements without the concomitant formation
of carbon dioxide. However, pure carbon dioxide may be formed from
its elements and the heat of reaction measured. Also, pure carbon
monoxide may be oxidized to form carbon dioxide and the heat of this
reaction measured. Thus, at 18°C,
C(p) + Oiig) = COiig), A H / = -94,030 calories (4)
CO{g) + iOsCg) = C02(ff), AH/ = -67,410 calories (5)
From the law of Hess, the heat of formation of carbon monoxide is
the same as the net heat of reaction accompanying (a) the formation of
carbon dioxide from the elements and (6) the decomposition of this
carbon dioxide into carbon monoxide and oxygen. The first step is
represented by Equation (4) with a heat of reaction, — 94,030. The
second step is the reverse of the reaction of Equation (5) with a heat of
reaction of +67,410. The net heat of reaction of the two processes is
-94,030 -t- 67,410 = -26,620 calories, the heat of formation of carbon
monoxide.
The result of this application of the law of Hess is exactly the same as
that obtained by treating Equations (4) and (5) as algebraic equalities
and combining them as such. If Equation (5) is subtracted from (4),
C (graphite) -t- IO2 = CO; AH/ = -26,620 calories (6)
260 THERMOCHEMISTRY [CHAP. VHl
All thermochemical equations may be treated in this manner and com-

bined with each other according to the rules of algebra. In this way it is
possible to use the principle of Hess effectively in calculating the heat of
a reaction for a complicated series of reactions. The heat of formation
of any compound may be calculated if the heat of any one reaction into
which it enters is known together with the heat of formation of each of
the other compounds present in the reaction.
For example, it is practically impossible to measure directly the heats
of formation of the hydrocarbons. However, a hydrocarbon may be
oxidized completely to carbon dioxide and water and this heat of reac-
tion measured. The heat of formation of the hydrocarbon may then be
calculated from the heats of formation of the other compounds present
in the reaction, namely, carbon dioxide and water. Thus,
CBiig) + 202(5) = C02(g) + 2B.iO{l); AHa = -212,805 cal (7)
C(|3) -H 02(3) = COiig); - >; ' ^* • A^;, = -94,030 cal (8)
2H2(3)-|-02((7) = 2H2O(0; ' AHc= -136,640 cal (9)
Equation (8) + Equation (9) — Equation (7) gives
C0S) + 2H2(g) = CH,ig) ' ' ' ' (10)
AH/ = AHi, + AHc - AHa= -94,030 - 136,640 - (-212,805)
= -17,865 cal '.v,.*5;'. .^ ; • , - .
By the principle of constant heat summation it is thus possible to

calculate heats of reaction which cannot be determined by direct meas-
urement. For example, the oxidation of linseed oil and the souring of
milk are reactions which proceed very slowly. By measuring the heats
of combustion of the initial reactants and final products it is possible
to calculate the desired heat of reaction. Similarly, the heat of transition
of a compound to an allotropic form such as graphite to diamond may
be impossible to measure directly. However, the difference between
the heats of combustion of these two forms of carbon will give the desired
heat of transition. Since this method involves taking the difference
between two large numbers, small errors in either number may result in
a large error in the difference.
Standard Heat of Combustion. The heat of combustion of a sub-
stance is the heat of reaction resulting from the oxidation of the sub-
stance with elementary oxygen. All experimental thermochemical
data on organic compounds are ordinarily expressed in terms of the
heats of combustion of the respective compounds. These data are not
necessarily the results of direct combustion experiments but may be
CHAP. VIII] STANDARD HEAT OF COMBUSTION 261
indirectly obtained from measurements of other heats of reaction which

lead to greater accuracy. Calculated heats of combustion data are
available for many substances which are ordinarily noncombustible, for
example, carbon tetrachloride.
The usually accepted standard heat of combustion is that resulting from
the combustion of a substance, in the state which is normal at 18°C and
atmospheric pressure, with the combustion beginning and ending at a
temperature of 18°C. The data ordinarily presented for standard
heats of combustion correspond to final products of combustion which
are in their normal states at a temperature of 18°C and at a pressure of
1 atmosphere. The major final products are generally gaseous carbon
' dioxide and hquid water.
The standard heat of combustion of a substance is dependent on the
extent to which oxidation is carried. Unless otherwise specified, a value
of standard heat of combustion corresponds to complete oxidation of all
carbon to carbon dioxide and of all hydrogen to liquid water. Where
other oxidizable elements are present it is necessary to specify the extent
to which the oxidation of each is carried in designating a heat of combus-
tion. For example, in oxidizing an organic chlorine compound, either
gaseous chlorine or hydrochloric acid may be formed, depending on the
conditions of combustion. If sulfur is present its final form may be
SO2, SO3, or the corresponding acids.
The situation is further compUcated by the fact that such products
may form aqueous solutions with the water. Standard heats of com-
bustion of compounds which contain elements such as S, CI, I, Br, N,
and F must always be accompanied by complete specification of the final
state of each product.
In Table XV are a few selected values of standard heats of combus-
tion of organic compounds. These values, in all cases, correspond to the
formation of gaseous carbon dioxide from all carbon present in the com-
pound. The hydrogen in the original compound forms hquid water
or may be utilized, in part, to form mineral acids when such elements as
CI, S, or N are present. The final products of combustion which are
formed from other elements are specifically designated. For example,
the heat of combustion of gaseous chloroform is given as —96,250 calories
per gram-mole. In the heading of this section of the table dealing with
halogen derivatives the final state of chlorine is specified as HCl in dilute
aqueous solution. Therefore, the heat of combustion of chloroform
corresponds to the heat of the following reaction:
CHClaCff) -1- IO2 + H20(ag) = CO2 + 3HCl(ag)

AHC = -96,250 calories .-. .» (U)
262 THERMOCHEMISTRY [CHAP. VHI
TABLE XV
STANDARD HEATS OP CoMBitsTioN*
Reference conditions: 18°C, 1 atmosphere pressure
AHS = heat of combustion in kilocalories per gram-mole
Multiply values by 1000 to obtain kilocalories per kilogram-mole and by 1800 to
obtain Btu per pound-mole.
^ Ahhreviations
s = Solid I = liquid g = gaseous
Hydrocarbons
Final Products: COJC?), HaO^).
Heat of
Compound Formula state Combustion
3» = AH»
Carbon (graphite) C s -94.030
Carbon (coke) C s -96.630
Carbon monoxide CO 9 -67.410
Hydrogen H. 9 -68.320
Methane CH4 g -212.805
Acetylene / C2H2 g -310.61
Ethylene CjH^ 9 -337.26
Ethane C2H6 g -372.83
AUylene .; CsHj g -464.78
Propylene CaHfi g -492.01
Trimethylene CgHfl 9 -496.80
Propane CsHs 9 -630.62
Isobutylene C4H8 9 -646.20
Isobutane C4H10 9 -683.37
«-Butane CiHio g -688.00
Amylene CsHio I -811.75
Cyclopentane CsHio I -783.60
Isopentane CiHii 9 -843.40
n-Pentane CsHia 9 -845.33
Benzene CeHs I -782.00
Hexalene CeHij I -963.90
Cyclohexane CeHu I -940.00
n-Hexane CCHH I -1002.40
Toluene CyHg I -934.20
Cycloheptane C7H14 I -1087.30
n-Heptane C7H16 I -1150.77
o-Xylene i CsHio I -1090.90
?re-Xylene ' CsHio I -1090.90
p-Xylene CsHio I -1087.10
n-Ootane CsHis I -1306.8
Mesitylene C9H12 I -1242.80
Naphthalene CioHs s -1231.6
Deoane C10H22 I -1619.4
Diphenyl Cl2Hl0 s -1493.30
Anthracene C14H10 s -1684.75
Phenanthrene C14H10 s -1674.95
Hexadecane Cl6H34 s -2556.1
* Values up to n-pentane taken from F. D. Rossini, Bureau of Standards Journal of Research. Other
values on this page taken from M. P. Doss, Physical Constants of Principal Hydrocarbons Texas Co.
(1942), with permission. Values on following pages taken from International Critical Tables, Vol. V,
162 (1929)."
CHAP. VIII] STANDARD HEAT OF COMBUSTION ' 263
TABLE XV —Continued
Alcohols
Final Products: COi{g), H20(i).
Heat of
Compound Formula State Combustion
Methyl alcohol CH4O -170.9

E t h y l alcohol CiHeO -328.0
Glycol C2H6O2 -281.9
Allyl alcohol CsH.O -442.4
n-Propyl alcohol C3H8O -482.0
Isopropyl alcohol CsHsO -474.8
Glycerol CsHsOs -397.0
n-Butyl alcohol C4H10O -639.0
Amyl alcohol CsHnO -787.0
Methyl-diethyl carbinol C6H140 -927,0
Acids
Final Products: COîg), mOQ).
Formic ; CH2O2 I -62.80

Oxalic C2H2O4 s -60.15
Acetic C2H4O2 I -208.00
Acetic anhydride C4H6O3 I -431.90
Glycolic C2H4O3 s -166.60
Propionic ' C3H6O2 I -365.00
Lactic ' CsHeOs s -326.00
d-Tartaric C4H6O6 s -275.10
n-Butyric C4H8O2 I -620.00
Citric (anhydrous) CeHsOr s -474.50
Benzoic C7H6O2 s -771.84
o-Phthahe C8H6O4 s -771.00
Phthalio anhydride C8H4O3 s -781.50
o-Toluic C8H8O2 s -928.90
Palmitic C16H32O2 s -2380.00
Stearolic , '• C18H32O2 s -2629.00
Elaidic C18H34O2 s -2664.00
Oleic ,^ ^ . C18H34O2 I -2669.00
Stearic • C18H36O2 s -2698.00
Carbohydrates — Cellulose —• Starch, etc.

Final Products: COiig), HjOCZ).
d-Gluoose (dextrose) CeHijOe s -673.0
Z-Fructose C6H12O6 s -675.0
Lactose (anhydrous) O12H22O11 s -1350.8
Sucrose C12H22O11 s -1349.6
Calories
per gram
Starch -4179
Dextrin -4110
Cellulose -4181
Cellulose acetate -4496
TABLE XY —Continued
Other CHO Compouiids
Final Products: C0i(g),'H.20{I).
Heat of
Compound Formula State Combustion
Formaldehyde CH2O 9 -134.0

Acetaldehyde C2H4O g -280.5
Acetone CsHeO I -427.0
Methyl acetate C3H6O2 g -397.7
Ethyl acetate C4H8O2 9 -544.4 •'
E t h y l acetate C4H8O2 I -538.5 '
Diethyl ether C4H1 oO I -651.7
Diethyl ketone CjHi oO I -736.0
Phenol CeHeO s -732.0
Pyrogallol CeHfiOs s -639.0
Amyl acetate C7H14O2 I -1040.0 '
Camphor CioHicO s -1411.0
Nitrogen Compounds
Final Products: CO2, NzCff), HjOCO-
Urea CH4N2O s -152.0

Cyanogen C2N2 9 -260.0
Trimethylamine C3H9N I -678.6
Pyridine CsHsN I -660.0
(1, 3, 5) Trinitrobenzene CeHsNsOe s -664.0
(2, 4, 6) Trinitrophenol CeHsNsOT s -620.0
o-Dinitrobenzene C6H4N2O4 s -703.2
Nitrobenzene CeHsNOj I -739.0
o-Nitrophenol CeHsNOs s -689.0
o-Nitroaniline CsHeNjOi! s -766.0
Aniline CsHvN I -812.0
(2, 4, 6) Trinitrotoluene C7H5N306 s -821.0
Nicotine C10H14N2 I -1428.0
Halogen Compounds
Final Products: COiig), HaOCZ), dil. sol. of HCl.
Carbon tetrachloride ecu 9 -90.08

Carbon tetrachloride CCI4 I -82.18
Chloroform CHCI3 9 -96.25
Chloroform CHCI3 I -89.20
Methyl chloride CH3CI 0 -164.00
Chloracetic acid C2H3CIO2 s -171.00
Ethylene chloride C2H4CI2 9 -271.00
E t h y l chloride CjHsCl 9 -321.80
Sulfur Compounds
Final Products: CO2, SOiig). HzOCO-
Carbonyl sulfide COS 9 -130.5

Carbon disulfide CS2 I -246.6
Methyl mercaptan CH4S 9 -297.6
Dimethyl sulfide C2H6S 9 -455.6
Ethyl mercaptan CgHeS g -452.0
CHAP. VIII] STANDARD HEAT OF COMBUSTION 265
It may seem strange to assign negative values to heat of combustion,

contrary to common usage in engineering. This practice is, however,
consistent with the use of changes of enthalpy as synonymous with heats
of formation, heats of vaporization, and so forth. Since combustion
proceeds with a reduction in the enthalpy of the system, the value of AH
must be negative and hence also the heat of combustion when the term
is used synonymously with the change of enthalpy.
From the heat of combustion of a substance its heat of formation may
be calculated if the heat of formation of each of the other products
entering into the combustion reaction is known. Thus, in order to
calculate the heat of formation of chloroform from Equation (11) it is
necessary to know the heats of formation of CO2, H2O(0, and HCl(ag).
These values may be obtained from Table XIV.
' ; "' ' B.2{g) +^02(3) = HaOCO; AHi = -68,320 cal (12)
C(^) + Oiig) = C02(sr); AHc = -94,030 cal (13)
hB.2ig) + IChig) = HCl(ag); AHa = -39,687 cal (14)
Equation (13) + 3 Equation (14) — Equation (11) — Equation (12)
gives
C(^) + iRîg) + nChig) = CKChig) (15)
or
AH/ = AHc + SAHd - AH a - AHb
AH/ = (-94,030) + 3 (-39,687) - (-96,250) - (-68,320)
= -48,521 cal
Thus the heat of formation of CHCl3(sf) is —48,521 calories per gram-
mole.
In this manner a general equation may be derived for use in calculating
the heat of formation of a compound CaHiBrcCldFel/NgOîSi from its heat
of combustion. If AHC is the heat of combustion of this compound
corresponding to the final products, C02(sf), H2O(0, Br(Z), 012(9^),
HF(ag), I(s), N2(gi), S02(sf), and AH/ its heat of formation, then,
AH/ = - A H C - 94,030a - 34,1606 - 41,540e - 70,920i (16)
If AH'C is the heat of combustion of this compound corresponding to the

final products C02(sr), RÔQ), Brîg), HCl {aq), H F {aq), I(s), HNO3
(aq), H2S04(a5), thdn,
AH/ = -AH'O - 94,030a - 34,1606 + 7,650c - 5,527d - 41,540e
- 1 5 , 0 3 0 3 - 147,480i (17)
266 • THERMOCHEMISTRY [CHAP. V H I
The coefficient of c in Equation (17) is the heat of formation of bromine

vapor from liquid bromine, which is taken aff the basis for the heats of
formation of bromine compounds.
If neither Equation (16) nor (17) is applicable, the heat of formation
of any compound may be calculated from its heat of combustion by the
method demonstrated above in the algebraic combination of Equations
(11), (12), (13), and (14).
Calculation of the Standard Heat of Reaction from Heats of Forma-
tion. The standard heat of reaction accompanjang any chemical change
may be calculated if the heats of formation of all compounds involved in
the reaction are known. If the reference state of enthalpy for a com-
pound at 18°C and 1 atmosphere pressure is taken as its separate com-
ponent elements at 18°C and 1 atmosphere pressure and in their normal
statesof aggregation then the relative enthalpyof the compound is equal to ;?• ,^
its heat of formation. Thus, the standard heat of reaction, or enthalpy •^•-^
change is equal to the algebraic sum of the standard heats of forma-
tion of the products less the algebraic sum of the standard heats of for-
mation of the reactants. When an element enters into a reaction its \
heat of formation is zero if its state of aggregation is that selected as the
basis for the heats of formation of its compounds. /
Illustration 1. Calciilate the standard heat of reaction of the following:
HCl(g) + NHsCg) = NHiCl(s)
From Table XIV the standard heats of formation are for *
HCl(ff), AH/ = -22,060 = Ha '
NHsCs), AH/ = -11,000 = Hj >.:
NHiCKs), AH/ = -74,950 = H.
Substituting these values in the above reaction, • \ 4
Affis = Hr — Ha — H6
AH18 = + (-74,950) - (-22,060) - (-11,000)
or AH18 = -41,890 cal /
Illustration 2. Calculate the standard heat of reaction, Affu of the following:
CsiCis) + 2H20(Z) = Ca(0H)2(s) + CJI;{g)
From Table XIV '
CaCaCs), AH/ = -14,100 cal = Ha -, ., •
H20(Z), AH/ = -68,320 cal = Ht
Ca(0H)2(s), AH/ = -236,000 cal = H,
The heat of formation of acetylene CiH.2{g), is calculated from the heat of combustion
data of Table XV by means of Equation (17)
CiHiig), AH/ = +74,230 cal = H.
CHAP. VIII] STANDARD HEAT OF REACTION 267
The enthalpy change AH, is then equal to H, + H, — Ho — 2H6

Affis + (-14,100) +2(-68,320) = +(-236,000) + (74,230)
Affis = -236,000 + 74,230 - (-14,100) - 2(-68,320)
or Affis = -11,030 cal
lUustiation 3. CalcxAate tiae Btandaxd \ieat ol Teacfion AHjs oi tlie f oUo'wing:
2FeS2(s) + BiOiig) = Fe203(s) + 4S0j
The standard heats of formation of the compounds are obtained from Table XIV,
for
FeS2(s), AH/ = -35,500 cal = Ha
Fe203(s), AH/ = -198,500 cal = H,
SOiig), AH/ = -70,930 cal = H.
Affis = H, + 4H. — 2Ha
= -198,500 + 4(-70,920) - 2(-35,500) = -411,180 cal
Calculation of the Standard Heat of Reaction from Heats of Com-
bustion. For a reaction between organic compounds the basic ttiermo-
chemical data are generally available in the form of standard heats of
combustion. The standard heat of reaction where organic compounds
are involved can be conveniently calculated by using directly the stand-
ard heats of combustion instead of standard heats of formation. An
energy balance is again employed, but in this case the standard reference
state is not the elements but the products of combustion at 18°C and
1 atmosphere pressure and in the state of aggregation specified by the
heat of combustion data. For example, the enthalpy of methane relative
to its products of combustion, gaseous CO2 and liquid HÔ, is equal to
the negative value of its standard heat of combustion, or +212,790
calories per gram-mole. Therefore, in any equation involving com-
bustible materials the formula of a compound may be replaced by its
enthalpy relative to its products of combustion. The enthalpy of the
products minus the enthalpy of the reactants is then equal to the stand-
ard heat of reaction, or the standard heat of combustion of the reactants
minus the standard heat of combustion of the products is equal to the
standard heat of reaction.
Illustration 4. Calculate the standard heat of reaction, AHr, of the following:
CJitOnU) + CH3COOH(0 = C2H600CCHa(i) + UiOil)
(ethyl alcohol) (acetic acid) (ethyl acetate)
From Table XV heats of combustion are as follows:
CjHeOH AH. = -328,000 cal = -Ha
r CH3COOH AH. = -208,000 cal = - H J
•;; CjHsOOCCH, AH, = -538,500 cal = -Hr
Since the heat of reaction is the difference between the enthalpies of the products and
the reactants,
Aflîs = Hr - Ha - Hi = 638,600 - 328,000 - 208,000 = 2500 cal
In general, the heats of formation of organic compounds are small in

comparison to the heats of combustion. Similarly, the heats of reaction
are small in systems involving only combinations of organic compounds.
Since these relatively small quantities can be determined only by the
differences between the large heats of combustion it follows that they are
rarely known with a high degree of accuracy. For example, the small
heat of reaction which is the final result of Illustration 4 is of uncertain
accuracy. An error of only 0.2 per cent in determining the heat of com-
bustion of ethyl acetate results in an error of 40 per cent in the value of
this heat of reaction.
When both, organic and inorganic compounds appear in a reaction it is
best to obtain the heat of reaction by means of heat of formation data.
The heats of formation of the organic compounds may be calculated
from their heats of combustion by means of Equation (16) or (17). The
procedure is then the same as that demonstrated in Illustration 2.
Illustration 6. Calciilate the standard heat of reaction, AHis, of the following:
CHaCKff) + KOH(s) = CHaOHCO + KCl(s)
The heats of formation of the inorganic compounds are obtained from Table XIV.
KOH(s); AH/ = -102,020 cal = HJ
KCl(s); AH/ = -104,361 cal = H.
The heats of combustion of the organic compounds are obtained from Table XV and
their heats of formation calculated by means of Equation (17); thus for CSgClig)
AH. = -164,000 cal
AH/ = 164,000 - 94,030 - 3(34,160) - 5627 = -38,037 cal = Ha
forCH30H(Z), AH. = -170,900 cal !.:•:••,:.
AH/ = 170,900 - 94,030 - 4(34,160) = -59,770 cal = Hr
Affis = Hr + H, - Ha - Hi, = -59,770 + (-104,361) - (-38,037) - (-102,020)
Affis = -24,074 cal
Heats of Neutralization of Acids and Bases. The neutralization of a

dilute aqueous solution of NaOH with a dilute solution of HCl may be
represented by the following thermochemical equation:
NaOH(ag) + HCl(ag) = NaCl(ag) + BÔil) (18)
The heat of reaction, AHr, may be calculated from the respective heats
CHAP. VIII] NEUTRALIZATION OF ACIDS AND BASES 269
of formation in Table XIV. Thus,

NaOH(og'); AH/ = -112,139 cal = Ha
HCl(ag); AH/ == -39,687 cal = H5
-•'' NaCUag); AH/ = -97,166 cal = H,
H2O(0;AH/= -68,320 cal = H .
From an energy balance
AHis = -97,166 + (-68,320) - (-112,139) - (-39,687)
^ ,../,,.- ' Airi8= -13,660 cal
Heats of neutralization may be determined by direct calorimetric meas-
urements in series of solutions of finite concentrations and these results
extrapolated to correspond to infinite dilution. Following are heats of
neutralization based, at least in part, on such measurements:
HCl(ag) + LiOH(a5) = LiCl(ag) + H2O; A ^ = -13,660
HN03(ag) + KOH(ag) = KN03(ag) + HÔ; AH = -13,660
^H2S04(ag) + KOH(ag) = |K2S04(ag) + H2O; AH = -13,660
It will be noted that the heat of neutrahzation of strong acids with
strong bases in dilute solution is practically constant when 1 mole of
water is formed in the reaction. The explanation of this fact is that
strong acids and bases and their salts are completely dissociated into their
respective ions when in dilute aqueous solution. From this viewpoint,
a dilute solution of hydrochloric acid consists of only hydrogen and
chlorine ions in aqueous solution, and a dilute solution of sodium hydrox-
ide consists of sodium and hydroxyl ions. Upon neutrahzation the
resulting solution contains only sodium and chlorine ions. The reaction
may be looked upon as ionic, and Equation (18) rewritten in ionic form.
Na+ (ag) + OH" (ag) + H+ (ag) + CI' (ag)

= Na+(ag)+Cr(ag)+H2O(0 (19)
Canceling the similar terms from Equation (19),
;:hv ;. .' - ^: OH-(ag) + H+(ag) = H2O (20)
AHis = -13,660 cal
Thus, the actual net result of the neutrahzation of dilute solutions of

strong acids and bases is the production of water from hydrogen and
hydroxyl ions. The accepted average value of the heat of neutrahzation
is —13,660 calories per gram-mole of water formed. This corresponds
270 THERMOCHEMISTRY [CHAP. VIE
to the heat of reaction of Equation (20), representing the formation of

water from its ions.
In the neutralization of dilute solutions of weak acids and weak bases,
the heat given off is less than 13,660 cal. For example, in the neutraliza-
tion of hydrocyanic acid with sodium hydroxide, the heat evolved is only
2661 calories per gram-mole of water formed. The unevolved heat may
be considered as the heat required to complete the dissociation of hydro-
gen cyanide into hydrogen and cyanide ions as neutralization proceeds.
As a hydrogen ion from hydrogen cyanide is neutralized by a hydroxyl
ion, more hydrogen cyanide ionizes until neutraUzation is complete.
This ionization of hydrogen cyanide requires the absorption of heat at
the expense of the heat evolved in the union of hydrogen and hydroxyl
ions.
Thennoneutrality of Salt Solutions. When dilute aqueous soluti6ns -•i^v
of two neutral salts are mixed there is no thermal effect provided there
is no precipitation, or evolution of gas. However, upon evaporation of
a mixture of such solutions four crystalline salts will be found, indicating
that double decomposition or metathesis has taken place. For example,
NaCl(ag)-t-KNOaCag) = NaNOsCag)+ KCl(ag); AH = 0 (21)
In dilute aqueous solutions it may be considered that each of these four
salts is completely ionized and Equation (21) written in ionic form,
Na+(ag) + Cl'iaq) + K+{aq) + N03'(ag) = Na+(ag)
-I- N03"(05) + K+iaq) + Cr{aq); AH = 0 (22) .
From this viewpoint it is evident that mixing such systems actually -
produces no change^ the initial and final solutions consisting of the same
four ions. It is only upon concentration of the solution and reassocia-
tion of the ions that the metathesis leads to a definite change in the
nature of the system.
The experimentally observed fact that dilute solutions of neutral salts
of strong acids and bases may be mixed without thermal effect is termed
the thermoneutrality of salt solutions. 'j
Heats of Formation of Ions. Equation (20) represents the formation /
of one mole of water from the combination of hydrogen and hydroxyl '
ions. The average value of this heat of reaction has been determined as / /
—13,660 calories per gram-mole. The heat of formation of water from / -
gaseous oxygen and hydrogen is given in Table XIV as 68,320 calories
per gram-mole.
H2(5) + iOiig) = nÔQ); AHa = -68,320 cal (a)
OR'iaq) + E+iaq) = nÔil); AHi = -13,660 cal (b)
CHAP. VIII] HEATS OF FORMATION OF IONS 271
Equation (a) — Equation (b) gives
HaCff) + |02(?) = B.+{aq) + OB.-{aq)

AH/ = -68,320 - (-13,660) = -54,660 cal (23)
Thus the combined heats of formation of the hydrogen and hydroxyl
ions from elementary hydrogen and oxygen is -54,660 cal. By arbi-
trarily assigning a zero value to the heat of formation of the hydrogen
ion the heat of formation of the hydroxyl ion is obtained. Thus, by
definition,
hUiig) =-R+; AH/ = 0 (24)
^Hiig)+h02ig) = OJI'{aq); A H / = -54,660 cal (25)
On this basis the relative heats of formation of the other ions of highly
dissociated acids and bases may be calculated. For example, from Table
XIV, the heat of formation of NaOH(ag) is —112,139 calories per gram-
mole. Since sodium hydroxide is completely dissociated into so-
dium and hydroxyl ions when in dilute solution, the formation of 1 mole of
NaOH (aq) from the elements is, in actual effect, the formation of 1 mole
of sodium and 1 mole of hydroxyl ions. Thus,
• Na(s) -I- i02(ff) + îig) = Na+(ag) -1- OH-(og)

A H / = —112,139 calories (gram-moles) (26)
' Na(s) = Na+; AH/ = -57,479 cal (27)
Therefore, the heat of formation of the sodium ion is — 57,479 calories per
gram-mole. In a similar manner the heats of formation of other ions
may be calculated, based on the assignment of a value of zero to hydro-
gen. Heats of formation of a few common ions, taken from the data of
Bichowsky and Rossini,^ are given in Table XVI.
The heat of formation in dilute aqueous solution of any compound
which is completely dissociated under these conditions is equal to the
sum of the heats of formation of its ions. From the data of Table XVI
the heats of formation of such compounds may be predicted.
niustration 6. Calculate, from the data of Table XVI, the heat of formation of
barium chloride in dilute solution.
Since BaCU may be assumed to be completely disassociated in dilute solution its
heat of formation in dilute solution is equal to the sum of the heats of formation of
^ Thermochemistry of Chemical Substances, Reinhold Publishing Co. (1936). With
permission. '
272 THERMOCHEMISTRY [CHAP. V I H
TABLE XVI
HEATS OF FORMATION OF IGNS
AH/ = heat of formation, kilocalories per gram-mole
Cations Anions
Ion Formula AH/ Ion Formula AH/
Aluminum A1+++ -126.3 Acetate CH3COO- -117.506

Ammonium NH4+ -31v455 Bicarbonate HCOr -164.6
Barium Ba++ -128.36 Bisulfate HSOr -213.3
Calcium Ca++ -129.74 Bisulfite HSOs" -149.0
Hydrogen H+ 0 Bromide Br" -28.67
Iron Fe++ -20.6 Carbonate CO3- -160.3
Iron Fe+++ -9.3 Chloride -39.687
Lithium Li+ -66.628 Fluoride cr
F" -78.20
Magnesium Mg++ -110.23 Hydroxide OH' -54.66
Manganese Mn++ -49.2 Iodide -13.37
Manganese Mn+++ -25.0 . Nitrite r -25.3
Potassium K+ -60.27 Nitrate Nor
NO3- > -49.19
Sodium Na+ -67.479 Phosphate po,— -297.6
Zinc Zn++ -36.3 Sulfate -215.8
Sulfite sor"
SO3" -148.5
one barium ion and two chlorine ions. From Table XVI,
Ba++; AH/ = -128,360 cal " \ ' \ ]'
•4, - -• • a"; AH/ = -39,687 cal ^ -
Ba+++ 2 C r = BaCUCag) ' ' * '
,^..,, BaClzK); AH/ = (-128,360) +2(-39,687)
or AH/ = -207,734 cal
The heat of formation of a substance which is soluble and highly
ionized in water may be calculated from its heat of formation in dilute
solution from the heat of formation of its ions provided its standard
heat of solution is known. Since experimental determinations of the
heats of solution are relatively easy, a simple method is provided for
estimating the heat of formation of many inorganic compounds. The
heat of formation in infinitely dilute solution is calculated from the ionic
heats of formation and the heat of solution subtracted from this value.
For example, the standard heat of solution of BaCl2(s) in an infinite
amount of water is —2,450 calories per gram-mole. From the result of
Illustration 6, the heat of formation of BaCl2(s) is —207,734 -|- 2,450 or
—205,284 calories per gram-mole.
Chemical reactions which take place between strong acids and bases
and their salts in dilute aqueous solution may be treated as ionic, and
heats of reaction may be calculated directly from ionic heats of forma-
tion. This method is particularly desirable when dealing with analytical
data for complex solutions. By treating the reactions as ionic it is
unnecessary to formulate hypothetical combinations of the analytically
determined ionic quantities.
CHAP. VIII] THERMOCHEMISTRY OF SOLUTIONS 273
Heats of Gaseous Dissociation. At high temperatures the elemen-

tary gases are known to dissociate into their atomic states with absorp-
tion of great amounts of energy. Upon cooUng, these monatomic gases
rapidly recombine to form the original elementary gas. From an indus-
trial standpoint, the most interesting phenomenon of this type is the
dissociation of normal gaseous hydrogen, 'Hiig), to form a monatomic
gas, H (fif). Such molecular dissociations in the gaseous state are accom-
panied by large absorptions of energy. Conversely, the reassociation
of the dissociated products is accompanied by a great evolution of energy.
In Table XVII are data for the heats of reaction of several gaseous
dissociations. ' - -' v-
TABLE XVII
HEATS OF GASEOUS DISSOCIATION

'.vT'
hO,{g} = 0(g); AHI8 = 59,100 cal
* •
i^.y- , iH^Cs) = H(s); AHI8 = 51,900 cal

:',:-••:[:•: • iN,(3) = N(g); AHIS = 85,100 cal
I' ••.; 1 i ; . iCh(g) = CI(9); AHI8 = 28,900 cal
: ,;>^.' ' ih(g) = Kff); AHI8 = 18,140 cal
THERMOCHEMISTRY OF SOLUTIONS
The enthalpy change accompanying the dissolution of a substance is
termed its heat of solution, or better, its heat of dissolution. If chemical
combination takes place between the solvent and the substance being
dissolved, the heat of dissolution will include the heat of solvation or the
heat of hydration accompanying this combination. If ionization takes
place the heat of solution will also include the energy of ionization. The
heat of solution of a neutral, non-dissociating salt is generally positive,
that is, heat is absorbed from the surroundings in the isothermal forma-
tion of the solution, or the solution cools if dissolution proceeds adia-
batically. The dissolution of such a material is analogous to the evapo-
ration of a liquid in that the result of each process is the breaking down
of a condensed structure into a state of great dispersion. In each
process energy is absorbed in overcoming the attractive forces between
the structural units of the condensed state.
Heats of solvation, especially in aqueous systems, are generally nega-
tive and relatively large. For this reason the heat of solution of a sub-
stance which forms a solvate or hydrate has generally a large negative
. value, indicating the evolution of heat when the unhydrated substance
is dissolved.
The enthalpy change when a substance is dissolved depends upon the
amounts and natures of the solute and solvent, upon the temperature,
and upon the initial and final concentrations of the solution. The
numerical value of the heat effect, therefore, requires an exact and
complete statement of all reference conditions.
Standard Integral Heats of Solution. Arbitrarily the standard
integral heat of solution is defined as the change in enthalpy of the system
when one mole of solute is dissolved in ni moles of solvent with the
temperature maintained at IS^C and the pressure at one atmosphere.
The numerical value of the integral heat of solution depends upon the
value of ni. As successive equal increments of solvent are added to a
given mass of solute the heat evolved with each addition progressively
diminishes until a high dilution is attained, usually about 100 or 200
moles of solvent per mole of solute, following which no further heat effect
is perceptible. The integral heat of solution approaches a maximum
numerical value at infinite dilution. This hmiting value is termed the
integral heat of solution at infinite dilution.
In Figs. 47 through 51 are presented the standard integral heats of
solution of common acids, bases, and salts in water. Integral heats of
solution are determined calorimetrically by measuring the heat of
solution of a solute in enough solvent to form a relatively concentrated
solution. The heat of dilution accompanying the addition of solvent to
this concentrated solution is then measured. The heat of solution at any
desired concentration is obtained by adding, algebraically, the ob-
served heats of solution and dilution.
It is evident that the integral heat of solution as defijied above is the
enthalpy of the solution containing 1.0 mole of solute, relative to the
pure components at the same temperature and pressure. Thus, the •
enthalpy of a solution relative to any selected reference state is expressed
by
H, = niHi + naHs -f- naAHji (28)
where - , .
H, = enthalpy of ni + ns moles of solution of components 1 and 2
Hi, Ha = molal enthalpies of pure components 1 and 2 at the tempera-
ture of the solution
AH,; = integral heat of solution of component 2 : »t, <'-,•••
Heat of Formation of a Compound in Solution. By combination of /

the data of Table XIV with those of Figs. 47 to 51 it is possible to calcu-
late the heat of formation of a compound in an aqueous solution of speci- ,
fied concentration. This total standard heat of formation is the sum of
the standard heat of formation of the solute, AH/, and its standard
integral heat of solution, AHsi, at the specified concentration.
CHAP. VIII] HEAT OF FORMATION IN SOLUTION 275
-20,000
—
• ; ^ M- I -HCIv
-HBr- -^
-^
f/ ^"1
/ / ,^ ^ Hjbu.
•12,000 // / /
(// ' / '
/
.
8,000 f
HN O j
o
I - 4,000
H,PO, •~H
UH, (JOO H >aa
FIG. 47. Integral heats of solution of acids in water at 18°C.
-14,000
-12,000 .KOH
N aUl
-10,000
^
/ •*— "=)-
1 -8,000 1-1 L\ I<»COJ
/
Na,C03
-6,000
/ iOFi
- 4,000 \\i'
\\\
-2,000 /
/
0~ 20 30 40 50
10
Fra. 48. Integral heats of solution of alkalies in water at 18°C.

-20,000
o
1
CaCh
^ -15,000
E /
n ZnCI:
• " ' ' '
&-10,000 /-
S -6,000
a. BaCij'
m
—NaCl=„—1
i 0 NH.Cl-s, KCl îl
I +5,000
20 40 60 80 lOO
FIG. 49. Integral heats of solution of chlorides in water at 18°C.
1
-20,000 MgSO.
s. ZnS04 1
1
S -15,000 JusO<
1 Mn SO4''
fe
o -10,000
% -5,000
a. N aiS04
"= n
' 1 5 000
^ NH4)! SO4
M K2SO, ^
1
20 40 60 » 100
Fia. 50. Integral heats of solution of sulfates in water at 18°C.
-15,000
1 1
Cu(N03)2
j ; -10,000
o
s
E -5,000
8'
^^ (NO3; Na N O J V
» +5,000
N H . N O 4?
^
KNOs
§ +10,000
0 20 40 60 80 100
FIG. 51. Integral heats of solution of nitrates in water at 18°C.

C H A P . VIII] H E A T O F SOLUTION O F H Y D R A T E S 277
Illustration 7. (a) Calculate the heat of formation of H2SO4 to form an aqueous

solution containing 5 moles of water per mole of H2SO4.
(6) Calculate the heat of dilution of the solution of part (a) to a concentration of
1 mole of H2SO4 in 20 moles of water.
(a) From Table XIV,
HjSOiCO; AH/ = -193,760 cal
From Fig. 47,

HjSOiCni = 5.0); AH, = - 1 3 , 6 0 0 cal
•' ' Hj(fif) + S(s) + 202(fir) + 6H2O = HjSOiCn, = 5)
'" AHIS = -193,750 - 13,600 = -207,350 cal
Since the same amount of water appears on both sides of the reaction its heat of forma-
tion does not appear in the equation.
(6) From Fig. 47, .,.„../,,,. ' ''; ; •
; I HSSOJCWI = 20); A H . = - 1 7 , 2 0 0 cal = Ha ' •'.'. -.'t,

j HzSOiCm = 5 ) ; A H , = - 1 3 , 6 0 0 c a l = / T r ••: .
< ! HjSOiCni = 5) + I5H2O = HjSOiCni = 20) .J
AH,8 = Hr - Ha = - 1 7 , 2 0 0 - ( - 1 3 , 6 0 0 ) . ,
AH18 = - 3 6 0 0 cal
Heat of Solution of Hydrates. If a solute forms a hydrate, the stand-

ard heat of solution of the hydrate is the difference between the heat of
solution of the anhydrous substance and its heat of hydration. The
heat of hydration is calculated from the data of Table XIV as the differ-
ence between the heat of formation of the hydrated compound and the
sum of the heats of formation of the anhydrous substance and of the
water of hydration.
Illustration 8. Calculate the standard heat of solution of CaCl2-6H20 to form a
solution containing 10 moles of water per mole of CaCU.
From Table X I V ,
• '*••'*" CaClj; AH/ = -190,600 = Ha
H2O; AH/ = - 6 8 , 3 2 0 = H6 •
CaCl2-6H20; A H / = -623,450 = H .
The enthalpy change accompanying hydration is then Hr — H,, — 6HI, or
AHIS = -623,450 - (-190,600) - 6 ( - 6 8 , 3 2 0 )

or lit-ikiJ AHI8 = —22,930 calories per gram-mole = Hi
This heat of reaction represents the molal enthalpy of CaCl2-6H20 relative t o CaClj
and H2O at 18°C. The enthalpy of CaCl2(ni = 10) relative to the same reference
substances and state is obtained from Fig. 49,
CaCl2(rai = 10); AH, = —15,500 calories per gram-mole = HJ

For the reaction
CaCl2-6HjO + iHjO— C&CU{ni' = 10)
AHI8 = H; - H^ = -15,500 -(-22,930)
AHIS = 7430 cal per gram-mole
Heats of Mixing. Heats of solution in a system in which both solute
and solvent are liquids are termed heats of mixing. Heats of mixing are
frequently expressed on a unit weight rather than a molal basis. If 10
grams of glycerin are mixed with 90 grams of water, the heat evolved is
191 calories. Therefore, the heat of mixing water and glycerin to form a
solution containing 10 per cent glycerin is —1.91 calories per gram of
solution formed. The integral heat of solution of either component may
be calculated from heat of mixing data.
Illustration 9. The heat of mixing of water and glycerin to form a solution con-
taining 40% glycerin is —4.50 calories per gram of solution. Calculate the integral
heats of solution of glycerin and of water at this concentration.
Basis 100 grams of solution.
Heat of mixing = —450 calories
Heat of solution of glycerin in water = —TT~ = —11.25 calories per gram
Heat of solution of water in glycerin = —-— = —7.5 calories per gram

60
Enthalpy-Concentration Charts. W. L. McCabe^ has called attention
to the usefulness of the enthalpy-concentration diagram for binary
solutions. In this chart the enthalpy per unit weight of solution is
plotted against concentration for a series of constant-temperature lines.
Once such a diagram for a given binary solution has been constructed,
calculations of the heat effects involved in changing the concentrations
and temperatures of the solution become simple and rapid. The ex-
penditure of the time required in constructing such a diagram is well
justified in case of specialization on a given system.
In constructing an enthalpy-concentration chart it is convenient to
choose as a reference state the pure components at to°C and their re-
spective vapor pressures. With heats of solution measured at 18°C
it is convenient to establish the first enthalpy curve at this temperature
and subsequently from this base line derive all other constant-tempera-
ture lines. This is done directly from Equation (28), employing
standard integral heats of solution together with heat capacity data
on both the pure components and the solutions.
In Figs. 52, 53, and 54 are enthalpy-concentration charts for sys-
tems of hydrochloric acid, sulfuric acid, and calcium chloride in water.
= Trans. Am. Inst Chem. Eng., 31, 129-162 (1935).
CHAP. VIII] ENTHALPY-CONCENTRATION CHARTS 279
Blustaration 10. Calc\jlate the final temperatuie -when 5.0 lb oi 10% HCl at 60°F
are mixed with 8 lb of 30% HCl at 100°F. From Fig. 52
Enthalpy of 10% HCl at 60°F = (5) (-60) = - 300
Enthalpy of 30% HCl at 100°F = (8) (-180) = -1440
Total enthalpy of mixture = —1740
Enthalpy per pound of final mixture = = —134
Xo
Final concentration = (0.5 + 2.4)/13 = 22.3% HCl
From Fig. 52 this enthalpy corresponds to a temperature of 90°F
v\
60 v^
40 \ \
- 20 \ \\ W
I 0 Ai \ \
-20
\ \
iW
\\^
w-80
I-60
-40
\
MVv\
\\
\
O
\ \
§-100 ^^ \
<£
I.-120
\
Nî^
a
<0"-140
V ^^
\ \ \
-160
>^
\
-180 \ ^^^
-200
K\
^ 140°F
-220
120°F
-240 100°F
80°F
\ 60°F
32°F
-300
20 30 50
Weight Per Cent HCl
FIG. 52. Enthalpy-concentration chart of hydrochloric acid solutions relative
to pure HCl(ff) and pure H2O(0 at 32°F and 1 atmosphere.
280 THERMOCHEMISTRY [CHAP. V I I I
In the enthalpy-concentration diagrams for the systems sulfuric acid-

water, and calcium chloride-water, Figs. 53 and 54, steep sloping lines are
shown in the two-phase region of vapor and solution. These lines repre-
sent the enthalpy of the entire system in Btu per pound, including the
liquid and vapor phases for isothermal conditions at one atmosphere
/
40 60 80 100 \
Percentage HjSO, /
FIG, 53. Enthalpy-concentration of sulfuric acid-water system relative to pure / /

components (water and H2S04 at 32°F and own vapor pressures). / '
pressure and meet the saturation curves at the normal boiling points of
the solutions. In the absence of other gases no vapor phase exists
when the total vapor pressure of the system is less than one atmos-
phere.
For the illustrated temperature-concentration range of the sulfuric acid
system the vapors are essentially pure water. Thus, a point on the
CHAP. VIII] ENTHALPY-CONCENTRATION CHARTS 281
chart in the two-phase region indicates a mixture of water vapor and

solution saturated at the indicated temperature and atmospheric pres-
sure. The enthalpy of the mixture is the sum of the enthalpies of the
400
ao 40 60
Percentage of Calcium Chloride
FIG. 54. Enthalpy of calcium chloride-water system. Modiiied from data of

Bosnjakovic, Technische Thermodynamik, Theodor Steinkopff, Dresden and Leipzig
(1937).
saturated solution and the water vapor. For example, a mixture at

250°F which has an overall composition of 4 5 % H2SO4 consists of .06
lb of water vapor plus .94 lb of solution saturated at 250°F and con-
taining 48% H2SO4. The enthalpy of the mixture is the sum of the
enthalpy of the saturated solution at 250°F which is 8 Btu per lb plus
the enthalpy of the superheated water vapor which from the steam
tables is 1169 Btu per lb. The enthalpy of 1.0 lb of mixture is then
(.06) (1169) + (.94) (8.0) = 78 Btu per lb. In this manner the en-
thalpies of other mixtures were calculated and the constant temperature
lines established in the two-phase regions of both Pigs. 53 and 54.
Since both enthalpies and masses are additive in the formation of
mixtures it follows from the principles of the energy and material bal-
ances that the properties of a mixture of two solutions or mixtures
must lie on a straight tie-line connecting the properties of the original
solutions or mixtures on the enthalpy-concentration chart.
If the vapor phase consists of two components, the composition of the
vapor in equilibrium with the liquid solution at its normal boiling point
must be known in establishing the vapor lines and the composition of
both phases.
Illustration 11. One pound of pure H2SO4 at 150°F is mixed vrith one pound
of 20% H2SO4 solution initially at 200°F. Calculate the temperature of adiabatie
mixing and the water evaporated.
From a material balance the resultant mixture contains 60% H2SO4 based upon
the combined liquid and vapor phases. By constructing a tie-line de connecting
the enthalpies of the two initial solutions and noting its intersection with the 60%
abscissa it will be seen that the resultant enthalpy per pound of mixture is 58 Btu.
The temperature of the mixture is 300°F and the corresponding hquid phase has a
composition of 63% H2SO4 with a boiling point of SOOT.
The mass of water evaporated, y, (neglecting the small amount of H2SO4 in the
vapor phase) can Be calculated from either a material or energy balance. From a
material balance,
0.80 = y + 0.37 (2 - y)
y = 0.095 lb water vapor
From an energy balance in which enthalpies at a constant pressure are equated,

Enthalpy of 1 lb H2SO4 at 150°F = 40 Btu
Enthalpy of 1 lb 20% H2SO4 at 200°F = 76 Btu
Enthalpy of water vapor at 300°F and 1 atm = 1194 Btu per lb
Enthalpy of 63% solution at 300°F = 1 Btu per lb
116 = 2/(1194) + (2 - 2/)

y = 0.095 lb water vapor ' /
Maicimum Temperature in Mixing Solutions. The maximum tem-

perature attainable in mixing two solutions of the same components but
of different temperatures and concentrations may be readily obtained
from an enthalpy-concentration diagram by constructing a tie-line con-
CHAP. VIII] SOLID-LIQUID SYSTEMS 283
necting the enthalpies corresponding to the two initial solutions. The

point of meeting or tangency of this line with the highest isotherm on the
diagram is the desired maximum temperature for adiabatic mixing. For
example, in Fig. 53, a tie-Une ab is constructed joining the enthalpy o of a
10% H2SO4 solution at 100°F and the enthalpy b of an 87% solution at
150°F. The maximum temperature attainable occurs at point c, corre-
sponding to the 250°F isotherm, and at a composition of 63%. The
relative weights of the two solutions to give this concentration are ob-
tained from a material balance. Thus, for 1 pound of 10% H2SO4
solution 2.21 pounds of an 87% solution are required to produce 3.21
pounds of a 6 3 % solution.
If the tie-line crosses the region of partial vaporization the maximum
temperature corresponds to the higher temperature of intersection with
the saturation line. Thus in Fig. 53 the tie-line de connects enthalpy d
of a 20% solution at 200°F with the enthalpy e of pure H2SO4 at 150°F.
The maximum temperature occurs at point / corresponding to 400°F
and a solution of 80% H2SO4. The relative amount of the two solutions
to give this concentration is obtained from a material balance. Thus,
for 1 pound of 20% solution, 3.0 pounds of H2SO4 are required to give
4.0 pounds of solution having a concentration of 80%.
Solid-Liquid Systems. The enthalpy composition chart of a com-
plex system is shown in Fig. 54 where the isothermal lines are drawn for
calcium chloride, covering the entire solubility diagram and a region of
partial vaporization. In this diagram line abode represents the solubihty
of the various hydrates of calcium chloride at various temperatures with
corresponding enthalpies. Line efg represents the melting points and
corresponding enthalpies of the various solid hydrates.
a = freezing point of water, 32°F
6 = eutectic of ice and CaCl2-6H20,-67°F
c = transition of CaCl2-6H20 to CaCl24H20 at 86°F
d = transition of CaCl24H20 to CaCl2-2H20 at 113°F
e = transition of CaCl2-2H20 to CaCl2-H20
/ = transition of CaCl2-lH20 to CaCl2
g == melting point of CaCU
In the two-phase region the sloping lines starting at the line marked 1
atmosphere represent the enthalpies of the liquid-vapor system at the
stated temperatures and compositions of combined phases. For exam-
ple, one pound of a vapor-liquid system containing 40% CaCU at 250°F
has an enthalpy of 95 Btu and consists of (0.40/0.42) lb of a 42% solu-
tion, boiling at 250°F at 1 atmosphere pressure and I( 11 - '—
' ^ )I lb of
\ 0.42/
•water vapor at 250°F in equilibrium with the boiling solution. The

value of 42% is obtained as the intersection of the 250°F with the
1 atmosphere line.
The sloping lines running upwards from the dotted melting point line
efg represent the enthalpies of the Uquid-vapor system at the stated
temperature and composition of the combined phases. For example, one
pound of liquid-vapor system containing 80% CaCl2 at 800°F has an
enthalpy of 220 Btu and consists of (0.8/0.84) lb of Uquid containing 84%
CaClz and ( 1 — jr^ j lb. of water vapor at 800°F in equilibrium with
the solid. The vapor pressures corresponding to the latter sloping lines
in the two-phase region are not given.
Illustration 12. To 200 pounds of anhydrous CaCU at 100°F are added 500
pounds of a solution containing 20% CaCU at 80°F.
(a) What is the temperature of the final mixture?
(6) How much heat must be removed to start crystallization?
From Fig. 54,
Enthalpy of CaClj at 100°F = (200) (12) = 2400 Btu
Enthalpy of CaCla solution at 80°F = (500) (-15) = -7500 Btu
Total enthalpy = -5100 Btu
Final composition = ffOO +^100) (100) ^ ^^ ^^^ ^^^^^
Final enthalpy = ^^^^—- - 7 . 3 Btu/lb • , ' '"^
(a) Final temperature is hence 190°F.

(6) This solution must be cooled to 7 0 ^ before crystallization will begin.
^H = 700 [-83 - (-7.3)] = -53,000 Btu
' This problem can also be solved by the construction of a tie-Une as demonstrated
in Illustration 11.
Partial Properties. In deahng with the extensive properties of solu-

tions it is convenient to consider separately the contribution attributable
to each component present. For example, the total volume of a solu-
tion represents the sum of the contributions of all of the components
present. In ideal systems, such as a gas at low pressure, these con-
tributions may be the same as the properties of the pure components
existing separately at the temperature and pressure of the solution.
However, in many systems this additivity of properties does not exist and
the volume of a solution may be quite different from the sum of the pure
component volumes. That portion of the total volume of a solution
which is attributable to the presence of a particular component is termed
CHAP. VIII] PARTIAL PROPERTIES 285
the partial volume of that component. Other partial extensive proper-

ties may be similarly defined such as partial enthalpy and partial heat
capacity.
The partial volume of a component in a solution of given composition
may be experimentally determined by adding a unit mass of the pure
component to a quantity of solution so large that only an infinitesimal
or negligible change in composition results. The increase in total volume
under these conditions is termed the partial specific volume of the com-
ponent added and in nonideal solutions may differ considerably from
the specific volume of the pure component. The partial volume of that
component in the solution is then the product of the partial specific
volume and the mass of the component present. Similarly, the partial
molal volume of the component may be determined as the increase in
volume resulting from the addition of one mole of the component to a
quantity of solution so large that no appreciable change in composition
results.
If the composition of the solution is varied to approach 100% of the
component under consideration it is evident that the partial specific
volume must approach the specific volume of the pure component.
Thus, in nonideal solutions the partial properties vary as functions of the
composition. The concept and use of partial properties have been
developed by Lewis and RandalP whose general methods are followed in
this discussion.
Mathematically, the partial molal volume of a component of a solu-
tion may be defined as
\oni/p,T,ni,n3---
where
V = partial molal volume of component 1
V = total volume of solution
nx = moles of component 1 __
The partial form of the derivative indicates the requirement of constancy

of all intensive properties such as temperature, pressure, and composi-
tion. Thus, Equation (29) is merely the formal expression of the result of
the experimental measurement, described above, of the change in volume
per mole of component added at conditions of constant composition,
temperature, and pressure.
Where M is a single valued, continuous function of several independent
variables x, y, z, the total differential of u can be expressed in terms of its
' G. N. Lewis and M. Randall, Thermodynamics, McGraw-Hill Book Co. (1923).
partial derivative with respect to its independent variables, thus,
du = — ax -\ ay -\ dz (30)
dx dy dz
This principle is illustrated in Fig. 55 for two independent variables x
and y. Applying this mathematical principle to the volume of a solution
in terms of the independent variables of composition.
,„ 37 , dV , dV ^
dV = — dni-\ 0712 H dris + (30a)
aril oui dUi
i|--d..(^)<ix+(f)d,
dy
••^,..
m^''
FIG. 55. Partial and total differentials,

or, combining Equations (29) and (30a)
dV = Vidni + Vidrii + Vsdris + • • • (31)
If the composition of the solution and its temperature and pressure are
held constant, the partial molal volumes of Equation (28) are constant
and the equation may be integrated:
V = Villi + Vjnz + Vzn» -\ (32)
Thus, the total volume of a solution is equal to the sum of the partial
volumes of the components.
Equation (32) may be differentiated to obtain another useful relation-
ship of partial properties. Thus
dV — Vidrii + Tiidvi + V2dn2 + n2dv2 + Vadna + nîiVi + • • • (33)
Combining Equations (31) and (33):
nidvi + n2dv2 + nsdvj + • • • = 0 (34)
This equation relates the changes in the partial molal volume of one
component to changes in the partial molal volumes of the others. i
CHAP. VIII] METHOD OF TANGENT SLOPE 287
Equations (29) to (34) were developed for volumes because of the ease
with which this property may be visualized. However, by parallel
reasoning similar equations may be developed relating any extensive
property to the contributions of the separate components. Thus, the
heat capacity of a solution is equal to the sum of the partial heat capaci-
ties of its components, derived by an equation similar to (32). Similarly,
the total molal enthalpy of a solution is equal to the sum of the products
of the partial enthalpies of the components times their mole fractions,
or the total enthalpy per gram is equal to the sum of the products of the
partial enthalpies per gram times the weight fractions.
It is evident that partial molal volumes or partial specific volumes or
any of the other partial extensive properties per mole or unit weight of
component are themselves intensive properties, independent of the mass
of solution under consideration but varying as functions of composition
as well as temperature and pressure.
Partial Enthalpies. For the establishment of energy balances the
enthalpy of a solution may be calculated either from integral heat of
solution data and the enthalpies of its pure components according to
Equation (28), or directly as the sum of the partial enthalpies of the
components. Thus,
H, = niHi -f n2H2 + W2AHS,-2 = niiii + n2H2 (35)
The latter method is frequently the more convenient and more nearly
accurate, particularly where small changes in composition are involved.
For this reason it is desirable to derive partial from total enthalpy data.
Lewis and Randall* present several methods of calculating partial
properties, two of which are illustrated by determining in one case
partial enthalpies per pound and in the other partial molal enthalpies
of components in solutions. The same procedures are involved in calcu-
lating any other partial extensive properties such as volumes or heat
capacities.
Method of Tangent Slope. If the total enthalpy of a solution is
plotted as ordinates against the moles of solute, component 2, per fixed
quantity of all other components, it follows from an equation similar to
Equation (29) that the partial molal enthalpy of component 2 is repre-
sented by the slope of this curve. If the solution is binary the partial
molal enthalpy of the solvent component then may be calculated from
an equation similar to (31).
For most accurate results covering the entire range of composition
of a binary solution two such plots should be constructed, one cover-
ing the concentration range of solute from 0 to 50% on the basis of
* Thermodynamics, p. 36, McGraw-Hill Book Co. (1923). •
one mole or unit weight of solute and the other covering the concentra-
tion range of solvent from 0 to 50% construc;ted on the basis of one mole
or unit weight of solvent.
Illustration 13. From the following data for the heats evolved when water and
glycerin are mixed to form 1 gram of solution, calculate the partial enthalpies at
1&°C of -water aad glycerin, per gram of each component, at each of the designated
concentrations. Plot curves relating the partial enthalpies of glycerin and of water
to percentage of glycerin by weight. As the reference state of zero enthalpy use the
pure components at 18°C, the temperature of the solutions.
Solution: Method of Tangent Slope. .,,,. , :' ,j ,., •
Percentage
Glycerin AH^ = H w' H' w" H" H' R"
by Weight
10 -1.91 111 -2120 -13.5 -0.6

20 -3.21 250 -4020 -10.6 -1.4
30 -3.90 429 -6560 -8.1 -2.1
40 -4.60 668 -7500 -6.2 -3.4
60 -4.50 1000 -9000 1000 -9000 -4.2 -4.8
60 -4.21 668 -7000 -2.65 -6.5
70 -3.70 429 -5290 -1.7 -8.3
80 -2.61 250 -3660 -0.8 -11.4
90 -1.79 111 -1990 -0.4 -14.3
Aff, I = heat of mixing, = H, the total enthalpy, calories per gram of solution.
w' = grams of glycerin per 1000 grams of water. >
w" = grams of water per 1000 grams of glycerin. .:
H' = total enthalpy of solution per 1000 grams of water. ;
B" = total enthalpy of solution per 1000 grams of glycerin.
H' = partial enthalpy of glycerin, calories per gram.
H" = partial enthalpy of water, calories per gram.
The values of H', the enthalpy of solution per 1000 grams of water, are obtained
by multiplying the heat of solution, per gram of solution, by the number of grams of
the solution (w' + 1000). For a 40% solution of glycerin
H' •= (668 + 1000) (-4.50) = -7500 calories
Values of w' are plotted as abscissas and values of H' as ordinates in Fig. 56. The
slope of a tangent to this curve is the partial enthalpy per gram, of glycerin H' in
a solution of concentration corresponding to the abscissa of the point of tangency.
For a 40% solution;
, .T/ -9600 + 3400
Slope of tangent H' - ' = -6.2
1000
The corresponding values of H", the partial enthalpy of water, are calculated
from the following equation, which is similar to (35):
(wi +W2)H = wiH' + wM"
CHAP. VIII] METHOD OF TANGENT SLOPE 280
For a 40% solution:
Basis: 100 grams of solution.
_„ ^ lOOH _ w^ -450 + (40) (6.2)
60 = —3.4
-10,000r
-10,000
•C -8000
a^
I i
<S ^ -400O
«" 8 • 429 H"
1-2000
•1
•1000 H'
1
400 600 800 1000 200 400 600 800 1000
w'=Grams of Glycerin u;"= Grains of Water
p«r 1000 Grams of Water per 1000 Grams of Glycerin
FIG. 66. Calculation of partial enthalpies FIG. 57. Calculation of partial enthalpie
by the tangent slope method. by the tangent slope method.
The concept of partial enthalpies may be clearly visualized by inspection of Fig. 56.
Point a represents the total enthalpy of a solution containing 668 grams of glycerin
and 1000 grams of water. This total enthalpy is the sum of the partial enthalpy of
the 668 grams of glycerin, equal to o — 6 on the diagram, plus the partial enthalpy
of the 1000 grams of water, equal to b.
It will be noted that at the higher concentrations of glycerin the slope of the curve
of Fig. 56 becomes small and it is difficult to determine a value of H' from it with
sufficient accuracy to permit a reliable calculation of H". For this reason it is
inadvisable to use this curve in the range of concentrations above 50% glycerin.
In this range more accurate graphical results may be obtained by plotting total
enthalpies for solutions containing 1000 grams of glycerin against the weight of
water dissolved. From this curve values of H" are determined directly. The cor-
responding values of H' are calculated from Equation (35). In Fig. 57 values of
w", the grams of water per ICKX) grams of glycerin, are plotted as abscissas against
H", the total enthalpy per 1000 grams of glycerin. The slope of a tangent to this
curve is the partial enthalpy of water per gram in a solution of concentration cor-
responding to the abscissa at the point of tangency. For a solution containing 70%
glycerin:
_„ -10,000 + 1700
Slope of tangent = H" = —— — = —8.3 calories
1000
From Equation (35), on the basis of 100 grams of solution:
_, lOOif wiH" -370 - (-8.3 X 30)
H' = -1.7
Wl IDl 70
In Fig. 57 point a represents the total enthalpy of a solution containing 429 grams
of water and 1000 grams of glycerin. This total is the sum of the partial enthalpy of
the water, equal to o — b plus the partial enthalpy of the glycerin, equal to b.
Method of Tangent Intercepts. In the method of tangent intercepts

the total enthalpy of the solution based upon unit quantity of solution
instead of unit quantity of either component is plotted against the frac-
tion of either component covering the range of concentrations from zero
to unity. If a tangent is drawn to this curve at a concentration corre-
sponding to a desired fraction of solute, then the intercepts of this tangent
line with the ordinates corresponding to zero and unit fractions of solute
give the respective partial enthalpies of the solvent and solute.
This method may be used for obtaining partial enthalpies per unit
weight of either component if compositions are expressed in weight frac-
tions and if the total enthalpies are expressed on the basis of a unit weight
of solution. Partial molal enthalpies are obtained by plotting total
enthalpy per mole of solution against mole fraction.
A proof of this relationship based upon molal units follows:
Let H = enthalpy of a solution containing m moles of solvent and nj
moles of solute relative to the pure components. The
enthalpies of the pure solvent and pure solute are each
arbitrarily taken as zero at the temperature of the solution.
H = enthalpy per mole of solution
Ni = mole fraction of solvent ' /
Ni = mole fraction of solute /
H '
* H= , • ' (36)
rii + 712 ^ ^
AT, = - ^ ' (37)
n i -H ^ 2
In Fig. 58, Curve III, values of enthalpy H in Btu per pound mole of
solution are plotted against mole fraction N-i of the solute. A tangent
is drawn to Curve III at point E corresponding to a mole fraction of
Ni equal to C. This tangent intercepts the ordinate A''2 = 0 at point A
and the ordinate iV2 = 1 at point F. Ordinate OB represents the value
of H at a mole fraction Ni=C. The slope of the tangent at JS is equal to
da
OA = OB - AB = n - Ni-iTT (38)
Variation in the mole fraction of solute, N^, may be produced by addi-

tion or removal of solvent, keeping nz constant. Differentiating Equa-
tion (36) with respect to n\ keeping ni. constant,
, _ ^^ TT rfni .
CHAP. VIII] METHOD OF TANGENT INTERCEPTS 291
Differentiating (37) with respect to rii keeping 712 constant,
dN,= - (40)
(wi + rii)^
-80,000 ]
-70,000
\ Hi
XsOj
-60,000
R. /
\ HjO/ •
-60,000
S • /
1
0,-40,000
\p F
-30,000 -^
•
-20,000
B EL*
• •
-10,000
A y
1 \
0.2 04 06 0.8 1.0
N,= Mole Fraction SO,
FIG. 58. Calculation of differential heats of solution by the method of tangent

intercepts, SO3-H2O system.
Combining (39) and (40)

da _ -dHjni + rn) H
(41)
dNi n%dn\ m
Substituting (41) in (38),
NidHin, + nz) NJI
OA = H + (42)
riidni n%
Substituting (36) and (37) in (42)

^. dH ' dH
OA = K + - H = —- (43)
ani dni
Since rii was taken as constant in all differentiations
i dni
which by definition is the partial molal enthalpy of component 1. Simi-

larly, it may be shown that i / I
0 F = H2 -^^ r : f j' ""' (46)
which is the partial molal enthalpy of the solute.

In Fig. 58 the total and partial enthalpies at 18°C, referred to the '
pure components at 18°C, of liquid SO3 dissolved in Uquid H2O are
plotted. The partial enthalpies for this system were calculated by
R. A. Morgan-^ from the data of Bichowski and Rossini* for values of
SO3 up to N2 = 0.5 and from Hermann' for values of S03fromN2 = 0.5
to 1.0.
For large changes in concentration the use of integral heat of solution
data or an enthalpy concentration chart is sound but more nearly accu-
rate results are obtained from partial enthalpy data, particularly where
sUght changes in concentration are involved. The heat evolved in
any process involving changes in concentration can be obtained by the
usual energy balance method where the enthalpy of the initial system ^
plus the heat absorbed is equal to the enthalpy of the final system. The
following illustration is taken from Morgen.*
Dlustration 14. It is desired to increase the strength of a 23.2% H2SO4 solution

(19% SOs) to 80.6% H2SO4 (65.6% SOa) with an oleum containing 41.2% free SO3
(89.2% SO3).
Calculate the heat evolved at 18°C on the basis of 100 pounds of weak acid. /
From a material balance of SO3 /
Weak Acid + Oleum Strong Acid /
(0.19)(100)+a;(0.892) = (100 + x)(0.656) §Hirn<Jm-='? y,
X = 197.5 pounds oleum /
100 + X = 297.5 pounds strong acid
^ iTid. Eng. Chem., 34, 671, (1942), with permission.

' Thermochemistry of Chemical Substances, Reinhold Publishing Company (1936).
With permission,
' Ind. Eng. Chem., 33, 898 (1941).
' Morgen, ibid.
CHAP. VIII] HEAT OP WETTING 298
Weak Acid Oleum Strong Acid

Pounds SO3 19 176 195
Pounds H2O 81 21 102
Pounds total ,, . i.s 100 197.5 297.5
Pound moles SO3 0.238 2.202 2.44
Partial molal enthalpy, SO3 X 10-' --6677..11 -4.68 -35.82
Pound moles H2O 4.50 1.185 5.686
Partial molal enthalpy of H2O X 10"'-0.14
)-»-0.14 -35.10 -7.16
From an energy balance of the process using partial enthalpies the following
results are obtained.
Weak Acid • Olgum
^rj SO3 H2O SO3 HÔ '
+ 0 . 2 3 8 ( - 6 7 . 1 ) + 4 . 5 0 ( - 0 . 1 4 ) + 2 . 2 0 2 ( - 4 . 6 8 ) + 1.185(-35.10)
^°^ Strong Acid
S02 H20
= 2.440(-35.82) + 5 . 6 8 6 ( - 7 . 1 6 )
I AH = - 59,620 Btu/100 pounds weak acid

or , 1 Heat evolved = 59,620 Btu
The partial enthalpy of a component in a solution referred to the
pure components at the temperature of the solution is termed the differ-
ential heat of solution of the component.
, ^ ^ HEAT OF WETTING
When a solid surface is brought into contact with a hquid in which it
is insoluble the liquid will spread in a thin film over the surface of the
solid provided the solid is wetted by the liquid. This implies that the
surface tension of the hquid relative to air is less than the adhesion ten-
sion between the liquid and the solid. The liquid film may be highly
compressed as a result of attractive forces or a chemical bond may occur.
The formation of such films of hquids is accompanied by an evolution of
heat.
Because of the small amount of hquid affected by the interfacial
forces of wetting, the heat of wetting per square centimeter of inter-
facial surface is a small quantity and is negligible unless a large area
of interface is formed as in wetting a fine powder or a porous material.
However, the thermal effects may then be of considerable importance,
their magnitude depending upon the nature of both sohd and hquid
and on the area of interface formed.
Experimental measurements have been made of heats of wetting in
many systems. Because of the uncertainty of the surface areas of
materials whose heats of wetting are important, such data are ordinarily
expressed as the change in enthalpy when 1 gram of the sohd is wetted
with suiScient liquid so that further addition of hquid produces no

thermal effect. This is termed the heat of complete wetting. Its magni-
tude is roughly proportional to the surface area exposed by a given
mass of solid and is dependent on the quantity of the liquid originally
present in the solid material. In Table XVIII are experimental values
of heats of wetting of dry clay, silica gel, starch, and charcoal by various
liquids. The chemical and, physical structures of the substances are
not specified, and it must be recognized that these data may be used
only for prediction of the order of magnitude of the effect to be expected
in a specific case.
H TABLE XVIII
HEATS OF COMPLETE WETTING OF POWDERS DRIED AT 100°C
Experiments at 12° to IS^C.
Data from International Critical Tables, Vol. V, page 142.
Unit: calories per gram of the dry material.
AH AH
AH AH
Liquid * Amorphous Sugar
Clay Starch
Silica Charcoal
H2O (water) -12.6 -15.3 -20.4 -3.9

CH3OH (methyl alcohol) -11.0 -15.3 -5.6 -11.5
CSHSOH (ethyl alcohol) -10.8 -14.7 -4.9 -6.8
. CsHnOH (amyl alcohol) -10.0 -13.5 -7.0 -5.6
CH3COOH (acetic acid) -9.3 -13.5 -3.0-^.0 -6.0
CH3COCH3 (acetone) -8.0 -13.5 -2.0 -3.6
CJHSOCSHS (ether) -5.8 -8.4 -2.2 -1.2
CeHe (benzene) -5.8 -8.1 -1.2 -4.2
CCI4 (carbon tetrachloride) -1.8 -8.1 -1.7 -1.6
CS2 (carbon disulfide) -1.7 -3.6 -0.5 -4.0
CsHia, CeHu (pentane, hexane) -1.2 -3.1 -0.3 -0.4
As previously mentioned, the heat given off in wetting a material is

diminished if it originally contains hquid. This effect is shown graph-
ically in Fig. 59 by plotting heats of complete wetting, per gram of dry
material, against the milligrams of hquid originally present per gram of
dry solid. The points at which the extrapolated curves approach the
axis corresponding to zero heat of wetting indicate the minimum quan-
tities of liquid required to produce complete wetting. The heat of ;
wetting of an already partially wetted soUd is the difference between the
heats of complete wetting at the final and the initial concentrations.
A more convenient method of presenting heat of wetting data is as
integral and differential heats of wetting. These terms are analogous to
integral and differential heats of solution and are similarly defined, i ,
CHAP. VIII] HEAT OF WETTING 295
Values of integral and dififerential heats of wetting of silica gel with
water are shown in Fig. 60 from the data of Ewing and Bauer.*
The relationships involved are
AH = (W2 + 1)H - WiHi - Hi = W2H2 + F i - W2H2 - Hi (46)
O.IO 0.20 0.30 0.40 0.50 0.60

Initial Grains of Water per Gram of Dry Substance
> . ~-
FIG. 59. Heats of complete wetting with water.
where ,
H = total enthalpy of wetted gel, Btu per pound of wet gel
AH = integral heat of wetting, Btu per pound of commercially dry
gel
El = partial enthalpy of gel, Btu per pound of commercially dry gel
Hi = partial enthalpy of water, Btu per pound of water
Wi = pounds of water per pound of commercially dry gel
Wi = pounds of dry gel equals unity
Hi, Hi = enthalpy of pure water and dry gel, respectively, at the tem-
perature of the system, in Btu per pound. Both values are
given as zero in Fig. 60.
V . Am. Ckem. Soc. 59, 1548 (1937). < , -' * • ' ;
It will be observed that the value of //a is —400 Btu per pound of
water when adsorbed upon dry gel but rapidly falls off when the water
content rises, reaching zero at 40% water (commercially dry basis). The
value of Hi changes in a reverse manner being zero for dry gel and
becoming equal to the integral heat of wetting at a water content of
40% (commercially dry basis). In commercially dry silica gel approxi-
-X60 p _ 1 400
-140
-\ -
•120 300
-iraft-.
- \ -
(S-100
\ -
J'80 200 S
*s - \HJ
1
-
3
I
B< - 6 0
AH^^^
-
(-; AH ^H,
CQ
-40 100
•
-20
i^^ 1 1 — 1 -i »» . 1._..
-6 0 5 10 20 30 40 50
Percentage Water Content (Comniercially Dry Basis)
FiQ. 60. Integral and differential heats of wetting of silica gel. From Ewing and
Bauer, Jour. Am. Chem. Soc, 69, 1548 (1937).
mately 5% of water of constitution is present. This water of constitu-
tion is always present in active gel and represents the minimum content
after regeneration. If this last trace of water were removed the gel would
lose its capacity for adsorption. It is also of interest to note that the
removal of this form of water as liquid would result in evolution of heat
-, rather than in absorption of heat, as indicated by that portion of the
$• AH curve of Fig. 60 below zero useful water content.
CHAP. V I I I ] HEAT OF ADSORPTION 297
Illustration 15. One hundred pounds of commercially dry silica gel are wetted
with water from a 6% to a 20% content (commercially dry basis). Estimate the
heat evolved from (a) integral and (b) differential heat of wetting data.
(a) From integral heat of wetting data (Fig. 60) the enthalpy change is equal to
100[-46 - (-18)] = -2800 Btu
(b) Prom the partial enthalpy data (Fig. 60) the enthalpy change is equal to
100[-26 + 0.20(-100)] - 100[-5 + 0.05(-260)] = -2800Btu
Therefore, the heat evolved is 2800 Btu.
When a liquid is evaporated from a wetted solid adsorbent the heat
required is greater than the normal heat of vaporization of the liquid
removed by the amount of the heat of wetting of the dried material
finally produced. This additional heat which must be supplied corre-
sponds to the work which must be done in removing the adsorbed
molecules away from the attractive forces holding them to the surface.
The heat of wetting is negligible in drying and calcining most soHds. .
V i : HEAT OF ADSORPTION
In Chapter V two types of gas adsorption on sohd surfaces are dis-
cussed, that caused by van der Waals forces and that by activated
adsorption. These two types are also characterized by differences in
their heats of adsorption. The heat of adsorption of a gas caused by
van der Waals forces of attraction and capillarity is the sum of the heat
of normal condensation plus the heat of wetting. In activated adsorp-
tion the gas is adsorbed by formation of a surface compound at tempera-
tures even above the critical temperature of the gas. The corresponding
heat of adsorption greatly exceeds that of normal condensation. The
difference between heat of adsorption and heat of normal condensation
represents the heat of wetting in van der Waals adsorption and the heat
of formation of a surface compound in activated adsorption.
Data on heats of adsorption are presented either as integral or as
differential values. The integral heat of adsorption is the change in
enthalpy per unit weight of adsorbed gas when adsorbed on gas-free or
" out-gassed " adsorbent to form a definite concentration of adsorbate.
The integral heat of adsorption varies with the concentration of the
adsorbate, in general diminishing with an increase in concentration.
The differential heat of adsorption of a gas is the change in enthalpy
when a unit quantity of the gas is adsorbed by a relatively large quantity
of adsorbent on which a definite concentration of the adsorbed gas al-^
ready exists. The differential heat of adsorption is also a function of
concentration, diminishing with an increase in concentration. As com-
plete saturation of an adsorbent is approached the differential heat of
adsorption approaches that of normal condensation. In activated
adsorption the differential heat of adsorption decreases as the more

active centers of the surface become coveretf. In capillary adsorption
the heat of adsorption decreases as the capillaries become progressively
filled. When the gas begins to condense at its normal saturation pres-
sure the heat of adsorption assumes the value of its normal heat of con-
densation.
Integral and differential heats of adsorption bear the same relation-
ship to each other as do integral and differential heats of solution and
may be calculated in a similar manner fro: i total and partial enthalpy
relationships. Heat effects in adsorption are calculated more accurately
from differential heat of adsorption data or partial enthalpies because
concentration changes are usually small and average values of the
differential heat of adsorption may be used.
y ^d
-17,000
I" I
'^-15,000
11
^
-w.ooo - v^ ^.
V
s
^ -n.oooU1
•9000-
\ ^ ^
a ,s \ \ < <
VII " \ *-^—• / •>-».
] •« - 7000 •
1
-5000L.
40 80 120 160 200 240 280 320 360 40O
Concentration of Gas, cc. (S.C.) per Gram of Dry Adsorbent
I. CBHB, 0 ° C on inactive cocoanut carbon, out-gassed at 350°C.
II. CeHe O^C on active cocoanut carbon, out-gassed at 350°C.
III. C2H6OH, 0°C on active cocoanut carbon, out-gassed at 350°C.
IV. H2O, C C on silica gel dried at 300°C for 2 hr and out-gassed at 250°C,
containing 3.5-5.5% HjO.
V. SOs, 0°C on same silica gel.
VI. SOs, -10°C on blood carbon (puriss. Merck), out-gassed at 450°C (ci = 1.63).
VII. NH3, 0°C on cocoanut carbon, heated to 550°C, out-gassed at 40O''C (d = 1.86).
FIG. 61. Differential heats of adsorption.
Experimental measurements have been made of the heats of adsorp-
tion of many of the more common gases on the important adsorbents
such as charcoal, silica gel, and various solid catalysts. In Fig. 61 are
values of differential heats of adsorption in calories per gram-mole ad-
sorbed at 0°C, plotted against concentrations in cubic centimeters of
adsorbed gas, measured at standard conditions of 0°C and 760 mm of
mercury, per gram of adsorbent. Similar data for other systems may
be found in the International Critical Tables (Vol. V, page 139).
lUustration 16. An adsorber for the removal of water vapor from air contains
250 lb of silica gel on which is initially adsorbed 28.0 lb of water. Calculate the
CHAP. VIII] ENTHALPIES OF ADSORBED SYSTEMS 299
heat evolved per pound of water adsorbed at this concentration, assuming that the
characteristics of the gel are similar to that on which the data of Fig. 61 are based.
28
Sohition: Concentration of H2O in gel = —- = 0.112 gram per gram of gel.
0.112
or —— X 22,400 = 139 cc per gram of gel
18
From Fig. 61 AH^ = — 12,500 calories per gram-mole
or -22,500 Btu per lb-mole
' „ , " . . . . , , , 22,500
Heat evolved per pound of water adsorbed = ——— = 1250 Btu
18
The effect of temperature on heat of adsorption is given by the
ClapejTon equation (page 59) and may be evaluated by an equal
temperature reference substance method of plotting equilibrium ad-
sorption pressure similar in principle to that used for normal vapor
pressures and heats of vaporization. By referring the heat of adsorp-
tion of a vapor to the normal heat of condensation of the same vapor at
the same temperature Equation (III-5) (page 65) becomes
:• ' din pi
Xi = X -—^ • (47)
a in p
where at the equal temperatures for each system
X = normal heat of condensation
Xi = heat of adsorption
p = normal vapor pressure of the condensed vapor
pi = actual equilibrium pressure of the adsorbed vapor
Enthalpies of Adsorbed Systems. The enthalpy diagram of a sohd-
adsorbed gas or solid-Uquid system has special advantages in problems
dealing with adsorption. In Fig. 62 the enthalpy of the silica gel-
water system is shown at various temperatures and compositions rela-
tive to dry silica gel at 32°F and liquid water at 32°F. For convenience,
the abscissas represent pounds of water per pound of commercially
dry gel. In constructing this chart the 70°F isotherm was derived first
since heats of wetting are known at this temperature. Thus,
(iU2 + l ) ^ ' = ^ 1 + H2W2 + AHu, (48)
where
Wi =
pounds of water per pound of dry gel
Hi =
enthalpy of dry gel at 70°F, Btu/lb
H2 =
enthalpy of Uquid water at 70°F, Btu/lb
AHV, =
heat of wetting when w^ pounds of water are added
to one pound of dry gel at 70°F
H = enthalpy of wet gel at 70''F, Btu/lb
The enthalpy at any other temperature is then obtained from
if = if70+ / C'dtf (49)
where Cp is the specific heat of the system.

From a combination of the enthalpy chart for air and that of siUca
gel it is easy to calculate the thermal effects in drying air by silica gel.
0 10 20 30 40 50
Water Content of Gel (CommerciaUy Dry Basis)
(Pounds of water per 100 pounds of dry gel)
FIG. 62. Enthalpy of silica gel-water system per pound of commercially dry gel.
(Reference state: Commercially dry gel and liquid water at 32°F.)
Illustration 17. Calculate the heat to be removed when 100 lb of air (dry basis)
at 70°F and 0.010 humidity are dried to a final humidity of 0.001 at 70°F and 1 atmos-
phere pressure starting with 10 pounds of dry gel at 70T. The average temperature
of the gel is initially at 70°F and finally at 80°F.
jo-îl
CHAP. VIII] INCOMPLETE AND SUCCESSIVE REACTIONS 301
Average composition of gel after adsorption =

100(0.010 - 0.001)
= 0.09
10
From Fig. 46 the heat removed from the air = 100(27.7 - 18) = 970 Btu
From Fig. 62 the heat removed from the gel = 10[6 - (-14)] = 200 Btu
,: . . ' ,;i7 Total heat removed = 1170Btu
INCOMPLETE AND SUCCESSIVE REACTIONS

In the preceding sections consideration was given to the thermal
effects accompanying chemical reactions which went to completion and
in which the reactants were present in stoichiometric proportions.
In industrial processes excess reactants are nearly always present and
the reactions are seldom complete.
Frequently the chemical transformations of the same reactant pro-
ceed in successive steps or in divergent stages, and the quantities of
different products which are formed bear no stoichiometric relationship
to one another. For example, in the combustion of carbon, both carbon
monoxide and carbon dioxide are formed in variable proportions which
depend upon the conditions of the reaction. In such reactions the
standard heat of reaction is calculated by exactly the same general pro-
cedure followed when all materials are present in stoichiometric pro-
portions, namely by^subtracting the heats of formation of the quantities
of reactants which actually react from the heats of formation of the
products actually formed. Or the heats of combustion of the products
actually formed are subtracted from the heats of combustion of the
reactants transformed to give the standard heat of reaction.
Illustration 18. In the production of metallic manganese, 10 kg of manganese

oxide, MjiiOi, are heated in an electric furnace with 3.0 kg of amorphous carbon
(coke). The resulting products are found to contain 4.8 kg of manganese metal
and 2.6 kg of manganous oxide, MnO, as slag. The remainder of the products
consists of unconverted charge and carbon monoxide gas. Calculate the standard
heat of reaction of this process for the entire furnace charge.
Solution:
Initial MnaOi = 10.0 kg or 10/229 = 0.0437 kg-mole
Initial C = 3.0 kg or 3.0/12 = 0.250 kg-atom
Mn formed = 4.8 kg or 4.8/55 = 0.0874 kg-atom
JVInO formed = 2.6 kg or 2.6/71 = 0.0366 kg-mole
Unconverted IMnsOi = 0.0437 '• ^-^ = ..., 0.0024 kg-mole
O in final CO = 4(0.0437 - 0.0024) - 0.0366 = 0.1286 kg-atom

CO formed = 0.1286 kg-mole
Unconverted C = 0.250 - 0.1286 = 0.1214 kg-mole
Material balance
Materials entering Materials leaving
Mn304 10.0 kg Mn 4.8 kg
C 3.0 kg MnO 2.6 kg
MnaOj = 0.0024 X 229 = 0.65 kg
Total 13.0 kg CO = 0.1286 X 28 = 3.6 kg
; ;^ C = 0.1214 X 12 = 1.45 kg
^ 13.00 kg
Heats of formation of active reaotants (data from Table XIV, page 253)
MnaOi = -345,000(0.0437 - 0.0024) = -14,200 kcal
C (coke) = +2600(0.250 - 0.1214) = +335 kcal
Total -13,865 kcal
Heats of formation of products actually formed:
MnO = -96,500 X 0.0366 = -3530 kcal
CO = -26,620 X 0.1286 = -3430 kcal -
Mn = 0 kcal
Total -6,960 kcal '
Standard heat of reaction = -6960 + 13,865 = +6905 kcal ,:
EFFECT OF PRESSURE ON HEAT OF REACTION

In Chapters II and VII it was pointed out that the internal energy
and enthalpy of an ideal gas are dependent only on temperature and
independent of pressure. Moreover, changes in pressure have negli-
gible effects on the internal energy contents and enthalpies of solids
and liquids.
Therefore, the heat of reaction at constant pressure is independent of the
pressure provided that in a reaction where gases are involved these gases
behave ideally, and provided that when liquids or solids are present the
changes in volume of the condensed phases are negligible.
Where these assumptions are not justified the effect of pressure on
heat of reaction must be calculated from an energy balance, taking into
account the effects of pressure on enthalpy by methods described later.
Reactions at Constant Volume. From the above section it follows
that the changes in enthalpy and internal energy accompanying a
chemical reaction are the same whether the reaction proceeds under
conditions of constant pressure or under conditions of constant volume
provided the gases present behave ideally. If a reaction proceeds in a
calorimeter at 18°C at constant volume the heat evolved —q can be
measured and is equal to the loss in internal energy.
-q,= -AU, (50)
If there is no change in the product pV during the reaction then the loss
CHAP. VIIIJ HEAT OF REACTION AND TEMPERATURE 303
in enthalpy will also equal the heat evolved. If, however, the reaction
proceeds with an increase in number of gaseous moles then there will
be an increase in p F equal to V Ap at constant volume. The change
in enthalpy then becomes:
AH = AU + VAp (51)
or • AH = qr, + VAp (52)
If An represents the increase in number of gasepus moles, then for
ideal gas behavior
VAp = AnRT (53)
or AH = q,+AnRT (54)
Therefore, the heat evolved in a reaction at constant volume is less
than the enthalpy decrease by the heat equivalent of the increase in pV,
or where ideal gases are involved, by the heat equivalent of the product
AnRT.
The difference between heat evolutions at constant pressure and
constant volume is of particular value in correcting experimental deter-
minations of heats of reaction to constant pressure conditions. Calori-
metric reactions are for the most part conducted at conditions of constant
volume. Standard heats of reaction and formation are calculated from
such data by means of Equation (54).
Illustration 19. In the combustion of 1 gram-mole of benzoic acid CeHsCOGH at
constant volume and a temperature 18°C, forming liquid water and gaseous CO2,
771,550 calories are evolved. Calculate the standard heat of combustion of benzoic
acid.
' CeHsCOOHCs) + 7i02(ff) = 7CO,(g) + 3HaO(Z)
An = 7 -7i = -i ^ / . ., .^
From Equation 54
AH = e„ + AnRT = -771,550 - (| X 1.99 X 291) =
—771,840 calories (gram-moles)
EFFECT OF TEMPERATURE ON HEAT OF REACTION

Standard heats of reaction represent the enthalpy changes during a
reaction at constant pressure in which all reactants are initially at a
selected standard temperature and all products are finally existent at
that same temperature. Such conditions are rarely encountered in
industrial reactions. Various reactants may enter at different tempera-
tures and the various products may each leave at still different tempera-
tures. The heat effects of such reactions may be calculated from data
on standard heats of reaction and thermophysical properties.
Kirchhoff's Equation. A useful analytical relation for the effect of

temperature on the heat of reaction may be derived for the special case
of reactions which begin and end with all materials at the same tempera-
ture.
Let AHT = heat of reaction at temperature T
AHT + dAH = heat (?f reaction at temperature T + dT
Since the enthalpy change in going from reactants at T to products at
T + dT is independent of path, >
dHR + AHT + dAH = AHT + dHp (55)
where
dHs and dHp are the changes in enthalpies of the reactants and
.• .^. products, respectively, corresponding to temperature
change dT
Where no latent heat changes are involved, that is, where there are no
changes in phase of reactants or products,
dHs = CpdT ., , (56)
and ' •••• • dHp = C'jdT • ' (57)
where •.
Cp and C'p are the heat capacities of the reactants and products, re-
spectively ,, ,
dAH-^ (C'p - Cp)dT = ACpdT . ' (58)
where >
ACp = change in heat capacity of entire system at constant pressure.
Equation (58), known as Kirchhoff's equation, may be applied to
the following general reaction at constant pressure: /
UbB + njO - { - • • • = TirR -(- n^S + • • •
where
rib, ric, rir, n, = number of moles of components B,C, R, S • • • j I
For this reaction,
ACp = rirCpr + nfip, 4- • • • — mcp}, — n^Cp, • • • (59)
Where the molal heat capacity of a substance is represented by an
empirical equation of the form
Cp = a + hT + cT^ (60)
CHAP. VIII] ENTHALPY CHANGES IN REACTIONS . 305
each heat capacity tenri in Equation (59) may be replaced by an em-

pirical equation of type (60). Then,
ACp = Aa + AbT + ^cT^' (61)
where
AO = TlrOr + KjOs • • • — fibOft — ricCic' •' (62)
A6 = nrhr + ris&s • •• — tibbb — ricbc • • • (63)
Ac = n,Cr + nfia •' • — nifii — ndCc • - • (64)
Substituting (61) in (58) and integrating,
AHT = In + âT + ÂbT^ + i A c P ' (65)
The constant of integration IB of Equation (65) may be evaluated from
a single value of AHT- This is usually done at the standard conditions
of 18°C. All other constants of Equation (65) are obtained directly
from the empirical heat capacity equations for the reactants and prod-
ucts.
Equation (65) is of thermodynamic importance in determining the
effect of temperature upon chemical equilibria. The utiUty of this
equation is limited to reactions which begin and end at the same tem-
perature where no changes in phase are involved, and where heat
capacities are expressed by the given empirical equations.
Enthalpy Changes in Reactions with Varying Temperatures. The
enthalpy change accompanying any reaction may be expressed in terms
of an overall energy balance. Thus, in constant pressure or flow proc-
esses where changes in kinetic, potential, and surface energies are
negligible and no work is performed,
AH = ZH'p - ZHB =q "• -- "^ (66)
where . "',
^HR = sum of the enthalpies of all materials entering the reaction
relative to the reference state for standard heats of reac-
tion, conveniently 18°C and 1.0 atmosphere
X^Jffp = sum of the enthalpies of all materials leaving the reaction,
referred to the form of chemical combination in which they
entered the reaction at the standard reference conditions
A substance which is produced by chemical reaction has an enthalpy

at 18°C referred to the reactants at 18°C which is by definition equal to
the standard heat of reaction. Thus,
,-,;,, H'p = H^ + AHis -, ,' (67)
f
where
Hp = enthalpy of the product, referred td its standard state at 18°C
AHii = standard heat of reaction
Combining (66) and (67) /
AH = E ^ p + EAJTis - J:HR (68)
In using Equation (68) standard heats of reaction are included in
the summations of the equation only for the products actually resulting
from reactions taking place in the process and to the extent that they are
formed in the process. This term becomes zero for all materials passing
through the process without chemical change.
Illustration 20. Carbon monoxide at 200°C is burned under atmospheric pressure
with dry air at 500°C in 90% excess of that theoretically required. The products
of combustion leave the reaction chamber at 1000°C. Calculate the heat evolved
in the reaction chamber in kcals per kilogram-mole of CO burned, assuming com-
plete combustion.
Basis: 1.0 kg-mole of CO.
CO(ff) + lOj(ff) = CO^to) t '
From the data of Table XV, page 262, /
Affis -67,410 kcal (kg-moles)
O2 required = 0.5 kg-mole
Oj supplied = 0.5 X 1.90 = 0.95 kg-mole
Air supplied = 0.95/0.21 = 4.52 kg-moles
Ns present = 4.52 - 0.95 = 3.57 kg-moles
Unused O2 = 0.95 - 0.50 = 0.45 kg-mole
Enthalpy of reactants (Hg) relative to standard state at 18°C
CO:
c,(18 -200°C) = 7.00 -, , . ' ^
Enthalpy = (7.00) (1.0) (200 - 18) = 1270 kcal ; '
Air: _ /
Cp(18 - 500°C) = 7.20 * '
Enthalpy = (7.2) (4.52) (500 - 18) = 16,690 kcal
J^HR = 16,690 + 1270 = 16,960 kcal
Enthalpy of products (Hp) relative to standard state at 18°C
COj:
Cp(18 - 1000°C) = 11.88 I i
Enthalpy = (11.88) (1.0) (1000 - 18) = 11,670 kcal
c,(18 - 1000°C) = 7.94 •

Enthalpy = (7.94) (0.45) (1000 - 18) = 3510 kcal
CHAP. VIII] ENTHALPY TERMS IN ENERGY BALANCES 307
N,:
Cp(18 - 1000°C) = 7.92 ,
Enthalpy = (3.67) (7.92) (1000 - 18) = 27,750 kcal
Y,Hp = 11,670 + 3510 + 27,750 = 42,930 kcal
From Equation (68)
Aff = 42,930 - 67,410 - 16,960 = -41,440 kcal
Since the reaction proceeds at constant pressure the enthalpy change is equal to the
heat absorbed, or there is an evolution of heat of 41,440 kcal.
Enthalpy Tenns in Energy Balances. As has been pointed out in
Chapter VII, page 207, enthalpy is a convenient property combining
internal energy and the product pV. Although enthalpy has energy
units it cannot in the general case be considered as solely energy. For
example, the enthalpy of an incompressible liquid may be increased by
merely increasing the pressure without doing any work or changing the
energy of the system. In such a case enthalpy is not a measure of
energy and there can be no generalization of " conservation of enthalpy "
parallel to the principle of conservation of energy. In the general case
the enthalpy input and output items of a process do not necessarily
balance and true energy balances should include only energy items.
I t is proper to include enthalpy terms in energy balances under 'two
conditions:
1. In a flow process the enthalpy is the sum of the internal energy and
the flow work and as such is properly included in the general energy
balance represented by Equation (VII-1), page 205.
2. In a non-flow process at constant pressure where work is performed
only by expansion enthalpy is properly included as a term in the general
energy balance of the form of Equation (VII-8), page 206.
Fortunately the great majority of processes of industrial importance
fall in one or the other of these classifications.
General energy balances are of great value in interpreting process
results and a formal tabulation of such a balance serves as a means of
verifying calculations and presenting a compact summary of results.
For such purposes it is desirable to tabulate all energy input items in
one column and all output items in another. For convenience in addi-
tion all entries in each column are arranged to be of positive sign. Thus,
in such a tabulation any work done on the system will appear as an in-
put item while work done by the system will appear as an output item.
Similarly, heat absorbed is an input, and heat evolved an output, item.
For all flow processes and for non-flow processes at constant pressure
the enthalpies of all entering materials are entered as input items and
those of the product materials, referred to their own standard states, as
output items. In such processes enthalpy changes equal heat absorp-

tions or evolutions and may be entered in the energy balance as such.
Standard heats of reaction are input items if they are negative, repre-
senting a contribution of heat to the process and output items if positive.
The overall enthalpy changes of the process are entered as output heat
items if negative and as input- items if positive.
Since energy balances of this type are entirely valid for the specified
types of processes, all forms of energy may be properly included along
with the work, heat, and enthalpy items. Where potential, kinetic,
and surface energies are not negligible, input and output items are
entered as part of the general balance.
Illustration 21. From the results of Illustration 20 prepare an energy balance for
the combustion process under consideration. .».,
Since this is a flow process, enthalpies may be included in the energy balance and '.^^
enthalpy changes are equal to heat absorptions and liberations.
Input Output
Enthalpy of CO 1,270 kcal Enthalpy of CO2 11,670 kcal :.
Enthalpy of Air 15,690 Enthalpy of O2 3,510
Standard Reaction Enthalpy of Nj 27,750
Heat Added 67,410 Heat Evolved 41,440
84,370 kcal 84,370 kcal
TEMPERATURE OF REACTION
Adiabatic Reactions. If a reaction proceeds without loss or gain of
heat and all the products of the reaction remain together in a single
mass or stream of materials, these products will assume a definite
temperature known as the theoretical reaction temperature. In this
particular case, for a flow process or a non-flow process at constant pres-
sure the enthalpy change is zero, and it follows from Equation (68)
that the sum of the enthalpies of the reactants must equal the sum of
the standard heat of reaction and the total enthalpy of all the products.
The temperature of the products which corresponds to this total en-
thalpy may be calculated by mathematically expressing the enthalpy
of the products as a function of their temperature. This requires data
on the heat capacities and latent heats of all products.
The products considered in calculating a theoretical reaction tem-
perature must include all materials actually present in the final system:
inerts and excess reactants as well as the new compounds formed. If
the reaction is incomplete, only the standard heat of reaction resulting
from the degree of completion actually obtained is considered and the
products will include some of each of the original reactants.
The enthalpy, H, of n moles of any material at a temperature r°K,
CHAP. VIII] ADIABATIC REACTIONS 30&
referred to a temperature of 291°K is expressed by:

T
H = nf CpdT + n\ (69)
"291
where
X = sum of the molal latent enthalpy changes in heating from
291°K to T°K at constant pressure
Expressing the molal heat capacity, Cp, as a quadratic function of tem-
perature, Cp = a + bT + cT^,
T
H = nf {a + bT + cT') dT + n\ (70)
Integrating:
H = n\aT-i-ibT' + ^cT'\ '+nX (71)

' L J291 *
The enthalpy of each product of a reaction may be expressed by an
equation of the form of Equation (71) and all these added together to
represent J^Hp of Equation (68), page 306. The theoretical reaction
temperature is then obtained by solution of this equation for T. The
solution is generally best carried out graphically.
Illustration 22. For the production of sulfuric acid by the contact process iron
pyrites, FeS2, is burned with air in 100% excess of that required to oxidize all iron
to Fe203 and all sulfur to SO2. It may be assumed that the combustion of the
pyrites is complete to form these products and that no SO3 is formed in the burner.
The gases from the burner are cleaned and passed into a catalytic converter in which
80% of the SO2 is oxidized to SOs by combination with the oxygen present in the
gases. The gases enter the converter at a temperature of 400°C.
Assuming that the converter is thermally insulated so that heat loss is negligible,
calculate the temperature of the gases leaving the converter.
4FeS2 + IIO2 = 2Fe203 + 8SO2
802 + iOj = SO3
Basis: 4.0 gram-moles of FeS2
Oxygen supplied for 100% excess = 11.0 X 2.0 = . . . . 22 gram-moles
Air introduced = 22/0.21 = 104.8 gram-moles
N2 introduced = 104.8 - 2 2 = 82.8 gram-moles
Excess O2 in burner gases = 22 — 11 = 11.0 gram-moles
SO2 in burner gases = 8.0 gram-moles
Gases entering converter:
SO2 8.0 gram-moles
O2 11.0 gram-moles
„ N2 82.8 gram-moles
f? •
Total , 101.8 gram-moles "• ~

SO3 formed in converter = 8.0 X 0.8 = 6.4 gram-moles

O2 consumed in converter = 6.4/2 = .^ .. 3.2 gram-moles
Gases leaving converter:
SO3 6.4 gram-moles
SO2 = 8.0 - 6.4 1.6 gram-moles
02 = 1 1 . 0 - 3 . 2 7.8 gram-moles
N2 \ 82.8 gram-moles
Total 98.6 gram-moles
Heat capacity of gases entering converter

(Mean molal heat capacities between 18°C and 400°C are taken from Table VI,
page 216.) •
SO2 = (8.0) (10.91) = 8 7
O2 = (11.0) (7.40) = 8 1
N2 = (82.8) (7.09) =587 H ' ^ ,
Total 755 v""^ «
Enthalpy, Y,HR, of gases entering converter = (755)(400 - 18) = 288,000 eal

Standard heat of reaction, AHwi
Heats of formation from Table XIV, page 253,
SO2; ABF = -70,920 cal s
SO3; AHP = -93,900 cal
X)AH29I = (6.4) X (-93,900) - (6.4) X (-70,920) = -147,000 cal

From Equation (68), since AH = 0,
HHp = J^HR - 'Z.AHiii = 288,000 - (-147,000) = 435,000 cal
In order to solve for the temperature of the products ^Hp is expressed as a function
of temperature by the equations of Table V, page 214. Integrating these equations
between the limits of 291 and T°K gives expressions for the enthalpies of the indi-
vidual components.
O2:
r 0.00299 _ 2 0.806 31
H = 7.S\ 6.13(T - 291) H — ( P - 291 ) — (10-^)(T' - 291 )
= 47.8(r - 291) + O.OmiT' - 2 9 1 ' ) - 2.095(10-«) (2" -291^)
r 0.001819 _ —2 0.345 —3 1
H = 82.8 6.30 (T - 291) + {P - 291 ) — (10-^){T' - 291 )
= 522(r - 291) + 0.0753(72 _ ^ ) _ 9.53(io-6)(J'3 _ 291')
SO,:
H = 1.6r8.12(T - 291) -^ — ^ ^ ^ (!P - 29l') - ?—^ (10-«)(r» - 2 9 l ' ) l
= 13.0(r - 291) + 0.00546(r2 - 291") - 1.122(10-8)(T' _ 5§^')

CHAP. V I I I ] NON-ADIABATIC REACTIONS 311
SO3:
r 0.01024 2 3.166 31
H = 6.4 8.20(7 - 291) + (T^ - 291 ) (10-«)(r' - 291 )
= 52.5(r - 291) + 0.0328(T2 - 29l') - 6.73(10-8)(r' - 29l')

Adding these equations
^ f f p = 635.3(r - 291) + 0.1253(T2 - 29T') - 19.48(10-«)(r» - 29l')

= 635.3r + 0.1253^2 - 19.48(10-«)T2 - 195,030
Equating the heat input to the enthalpy of the products
635.37 + 0.125372 - 19.48(10-«)r3 - 195,030 = 436,000
This equation is best solved graphically or by substituting values of T until the
equation is satisfied.
Solving, . .
T = 865°K, or 592°C
This calculation can be verified or can be estimated more readily by assuming
average values of the heat capacities of the products. A slight error in the estimated
final temperature will not alter the values of mean heat capacities appreciably. For
example, let it be assumed that the final temperature is 590°C.
Mean molal heat capacities are obtained from Table VI, page 216,
, 02= 7.60
' ' N2 = 7.20
: SO2 = 11.40
SO3 = 12.46 (Table V, page 214)
X,H, = [(7.8)(7.60) + (82.8)(7.20) + (1.6)(11.40) + (6.4)(12.46)](7 - 291)
= (753.3) (7 - 291) = 436,000
7 = 869°K(596°C).
This value is better than 592°C because experimental values of heat capacities
were used in deriving the mean values of Table VI rather than the empirical equa-
tions of Table V. If the calculated final temperature were considerably different
from that assumed as a first approximation for obtaining mean heat capacities, the
resultant error may be reduced by repeating the calculation with heat capacity data
based on the temperature calculated as a first approximation. Since mean heat
capacities do not vary greatly with temperature successive approximations of this
type rapidly approach the correct solution.
Non-Adiabatic Reactions. If a reaction does not proceed adiabatically

the amount of heat gained or lost from the system during the reaction
must be known in order to calculate the temperature.
If a flow process or a non-flow process at constant pressure is under
consideration the heat absorbed by the system is equal to AH of Equa-
tion (68). Using this value of AH the enthalpy and temperature of the
products are calculated as in Illustration 22.
Theoretical Flame Temperatures. "The temperature attained in the

complete, adiabatic combustion of a fuel which is thoroughly admixed
with air or oxygen is termed the theoretical flame temperature. The
methods developed in the preceding sections may be used to calculate
the theoretical flame temperature of a gaseous, atomized liquid, or pow-
dered solid fuel when burned with air in any desired proportions. The
maximum theoretical flame temperature of a fuel corresponds to com-
bustion with only the theoretically required amount of pure oxygen.
The maximum flame temperature in air corresponds to combustion
with the theoretically required amount of air and is obviously much
lower than the maximum flame temperature in pure oxygen. Because
of the necessity of using excess air in order to obtain complete combus-
tion, the theoretical flame temperatures of actual combustions are
always less than the maximum values.
Illustration 23. Calculate the theoretical flame temperature of a gas containing

20% CO and 80% Nj when burned with 100% excess air, both air and gas initially
being at IS^C.
Basis: 1.0 gram-mole of CO.
Ns in original gas = —— X 0.80 = ;.. 4.0 gram-moles

Oj supplied = 0.5 X 2 = 1.0 gram-mole
Nu from air = — ^ X 0.79 = 3.76 gram-moles
Total Nj = 3.76 + 4.0 = 7.76 gram-moles
Moles of original Nj, O2, CO = 7.76 + 1.0
-f 1.0 = 9.76 gram-moles
' • I •
Combustion products:
CO2 formed = 1.0 gram-mole :
O2 remaining = 1.0 — 0.5 = 0.5 gram-mole
N2 = 7.76 gram-moles
Enthalpy of products, ^Hp (referred to 18°C) ' /
CO2: /' . ,,
r 0.008533 _ 2 2.475 s "1
H = 1.0 6.86(r - 291) + (T^ - 291 ) (lO-'Xr^ - 291 )
Oi
H = 0.6[e.l3iT - 291) + ' - ^ (T' - § 9 1 V ^ (lO-Xr^ - 2 9 l ' ) ]
N .1'
H = 7.76[6.30(r - 291) + ^ ^ (P -2^)-—• (10-)(r' - 29l')]

CHAP. VIII] ACTUAL FLAME TEMPERATURES. 313
Adding:
JI^H, = 58.80(T - 291) + 0.01207(7^ - 291^) - 1.851 (10-«)(r5 - 29i')
Hp = S8.80!r + 0.0120772 - 1.851 X lO-^T^ - 18,095
Since Y^HB = 0 and Aff = 0, from Equation (68), J^Hp = -"£, ^Hin
or 68.802" + 0.012077^ - 1.851 X lO-^r' - 18,095 = 67,410
Solving this equation graphically, T = 1204°K or 931°C.
A better value may be obtained more directly by assuming the final temperature
and then using the corresponding values of mean heat capacities from Table VI,
page 216. Thus the mean heat capacities between 18°C and 930°C are
. , , , . « CO, = 11.81
,, , . ,. .. - , Os = 7.89 , - .
Nj = 7.45
Then,
Y.Hp = [(1)(11.81) + (0.5)(7.89) + (7.76)(7.45)](r - 291) = 67,410
or T = 1206°K(933°C)
The theoretical flame temperature of a fuel is dependent on the initial

temperature of both the fuel and the air with which it is burned. By
preheating either the fuel or the air the total heat input is increased and
the theoretical flame temperature is correspondingly raised.
Illustration 24. Calculate the effect on the theoretical flame temperature of Illus-
tration 23 of preheating both the gas and air to 1000°C before combustion.
Basis: Same as Illustration 23.
Enthalpy of reactants at 1000°C relative to 18°C
CO = (1.0) (7.58)(1000 - 18) = 7,450 . :.,
Os = (1.0)(7.94) (1000 - 18) = 7,800 ;:,',; ' '•,
N, = (7.76)(7.50)(1000 - 18) = 57,100 ' ^•'' • ^ :< ' •
HB = 72,350
Using the values of AH291 and the equation for ^Hp from Illustration 23,
67,410 + 72,350 = 58.80r + 0.012077' - 1.851 X lO'^T^ - 18,095
68.80T + 0.0120772 - 1.851 X IQ-^T^ = 157,855
Solving this equation, .! ; i ;; :>'a>;.:i
T = 2078°K (1805°C)
In Table XXII, page 340, are values of the maximum theoretical flame
temperatures of various hydrocarbon gases when burned with air at 18°C.
Actual Flame Temperatures. A theoretical flame temperature is
always higher than can be obtained by actual combustion under the
specified conditions. There is always loss of heat from the flame, and
it is impossible to obtain complete combustion reactions at high tem-
peratures. The partial completion of these reactions results from the
314 THERMOCHEMISTRY [CHAP. V H I
establishment of definite equilibrium conditions between the products

and reactants. For example, at high temperatures an equilibrium is
estabUshed between carbon monoxide, carbon dioxide, and oxygen, cor-
responding to definite proportions of these three gases. Combustion
of carbon monoxide will proceed only to the degree of completion which
will give a mixture of gases in proportions corresponding to these equi-
librium conditions. The general subject of reaction equilibria is dis-
cussed in Chapter XV.
In Table X X I I are experimentally observed values of the maximum
flame temperatures of various hydrocarbon gases when burned with air
at 18°C. These data are from the work of Jones, Lewis, Friauf and
Perrott.^" These investigators also demonstrated a method of calcu-
lating flame temperatures, taking into account the degree of completion
actually obtained if the combustion proceeds to equilibrium conditions
but neglecting heat loss. The calculated values were ordinarily higher •'
than those experimentally observed. Values of such maximum colcu-
lated flame temperatures are included in Table XXII.
The maximum values of actual and calculated flame temperatures do
not correspond to the air-fuel proportions theoretically required for
complete combustion. Because of the incomplete combustion actually
produced at high temperatures the maximum flame temperature is
obtained with a ratio of air to fuel which is somewhat less than that
required for complete combustion.
It will be noted from Table X X I I that the maximum flame temper-
atures of the various gases vary but little. For example, although '
pentane has twelve times the heating value of hydrogen, its flame tem- _
perature is lower by 110°C. This results from the fact that, in the com-
bustion of the gases of high heating values, correspondingly large
quantities of combustion products are formed with high total] heat
capacities. -
• . • • • • . ^ ^ ' i • ' ' . • '' • • ' • " '-;
i PROBLEMS - /
1. Calculate the heat of formation, in calories per gram-mole, of SOsig) from the
following experimental data on standard heats of reaction: /
PbO(s) + S(s) + iOiig) = PbSOiCs); Aff = -165,500calories (gram-moles)
PbO(s) -I- H2SO4-5H2O(0 = PbS04(s) -|- GHjOy); AH = -23,300 calories /
SOsCs) 4-6H2O(0 = HjSOi-SHjOG); AH = -41,100 calories /
2. From heat of formation data, calculate the standard heats of reaction of the
following, in kcal per kilogram-mole: ' ,
(a) SOJCS) + iOiig) -f H,0(l) = HiSO,(l). -
(b) CaCOals) = CaO(s) + C02(ff).
'»/.^m.C;iem.Soc., 53, 869 (1931). V
CHAP. VIII] PROBLEMS 315
(c) CaO(s)';+ 3C(graphite) = CaCaCs) + CO(g).

(d) 2AgCl(s) + Zn(s) + og = 2Ag(s) + ZnCljCog).
(e) CUSO4H) + Zn(s) = ZnSO^Cog) + Cu(s).
(J) N^Cff) + m,0{g) = 2NH2(s).
(?) N2(ff)+0,(s) = 2 N 0 ( s ) .
(h) 2NaCI03(s) ->2NaCl(s) +302(ff).
(t) CuO(s) + Hsto) ->H2O(0 + Cu(s).
(j) Ca3(P04)2(s) + 38i02(s) + 5C(s) ->3CaO-Si02(s) + 5C0(s) + 2 P ( s ) .
(k) NHsCs) + SOJCS) ->4N0(s) + 6H20(s).
(I) 3N0i{g) + HzOa) ->2HN05(a9) + NO(g).
(m) CaF2(s) + B.SOS)-*Ca.SOîs) + HP(i).
(n) PjOsCs) + SHjOy)-»H3P04(og).
(0) CaC2(s) + 5H20(s) ^ C a O ( s ) + COîg) + SKîg).
(p) AsHîg) + 302(?) -ÂsjOaCs) + SHjOCO.
(Heat of formation of arsine = +43,500 cal per g-mole)
3. Calculate the heats of formation of the following compounds from the standard
heat of combustion data:
(o) Benzene (CeHe)^)-
(b) Ethylene glycol (CsHeOj) (J).
(c) Oxalic acid (C00H)2(s).
(d) Aniline (C6HsNH2)(Z). • :. - v,
(e) Carbon tetrachloride (CCl4)(i).
if) Propane(g). ;. -, ..
(g) Phenol (s).
(h) Carbon disulfide (J).
(i) Urea (s).
(J) Chloroform (I).
Note: The heat of combustion data given in the table for halogen compounds are
based on having all chlorine converted into HCl(ag), by hydrolysis. Accordingly,
the actual reaction to be considered is one of the combined oxidation and hydrolysis.
For CCI4 it is all hydrolysis and no oxidation.
4. Calculate the standard heats of reaction of the following reactions, expressed in
calories per gram-mole:
(a) (C00H)2(s) = H C O O H a ) + C 0 2 ( g ) .
(oxalic acid) (formic acid)
(b) C2H50H(i) +02(^) = CH3C00H(i) + H2O(0. /
•' (ethyl alcohol) (acetic acid)
(c) 2CH3Cl(s) + Zn(s) = CJi,(,g) + ZnCl2(s). /
(methyl chloride) (ethane)
. (d) 3C2H2(s) = C6H6(0. , r _
(acetylene) (benzene)
- (e) (CH3COO)2Ca(s) = CH3COCH3(Z) + CaC03(s).
(calcium acetate) (acetone)
Cf) CH30H(Z) + iO,(.g) = HCHO(s) + H2O(0.
(methyl alcohol) (formaldehyde)
: (g) 2C2H5C1(0 +2Na(s) = C4Hio(g) + 2NaCI(8)
(ethyl chloride) (n-butane)
(fc) C2H2(g) + H2O(0 = CH3CH0(g)
(acetylene) (acetaldehyde) ,',! .
(i) CcHsNOsy) +3Fe(s) +6HCl(o?) = CeHsNHaa) + 3FeClj(og) + HjO(Z).

(nitrobenzene) (anfline)
(Heat of formation of FeCl2(ag) = -99,950 g cal/g-mole)
6. The integral heat of solution of LiCl in water to form a solution of infinite dilu-
tion is —8440 calories per gram-mole. Calculate the heat of formation of LiCl(s)
from the data of Table XVI (page 253).
6. (a) Calculate the number of ,Btu evolved at 18°C when 90 lb of ZnCU are added
to 150 lb of water.
(6) Calculate the number of Btu evolved when 50 lb of CaCU are added to
200 lb of an aqueous solution containing 10% CaCU by weight at 18°C.
7. Calculate the heat evolved, expressed as Btu, when the following materials are
mixed at 18°C:
(a) 50 lb H2SO4 and 50 lb H2O.
(6) 50 lb H2SO4 and 200 lb of a solution of sulfuric acid and water, which con-
tains 50% by weight of H2SO4. ; '
(c) 50 lb H2O and 200 lb of a solution of sulfuric acid and water, which con- ''^
tains 50% by weight of H2SO4.
(d) 60 lb of Na2SO4-10H2O and 100 lb of water.
8. An aqueous sulfuric acid solution contains 50% H2SO4 by weight. To 100 grams
of this solution are added 75 grams of a solution containing 96% H2SO4 by weight.
Calculate the quantity of heat evolved.
9. 100 pounds of an oleum solution containing 15.4% free SO3 is to be diluted
with water to make a 30.8% solution of H2SO4. Calculate the heat evolved.
10. 100 grams of oleum containing 71.0% free SO3 are diluted with a large volume
of water to infinite dilution. Calculate the heat evolved.
11. One ton of 19.1% H2SO4 is to be concentrated to 91.6% H2SO4. Steam is
available to heat the acid to 302°F and a vacuum can be maintained equivalent to
the vapor pressure of 91.6% H2SO4 at 150°C, i.e., 14 mm. Calculate the heat to be
supplied.
12. A batch of dilute sulfuric acid at 90°F, which weighs 250 lb and contains
20% H2SO4 is to be brought to 50% strength by the addition of 98% acid, which is
at 70°F.. How much heat is abstracted by the cooling system if the temperature
of the final acid is 100°F?
By means of a sketch, show how Fig. 53 was used in the solution of this problem.
13. In a continuous concentrating system, dilute acid (60% H2SO4) is concentrated
to 95% strength. The dilute acid enters the system at 70°F, while the water vapor
and the concentrated acid leave the system at the boiling temperature of the latter. /
How many Btu are needed to concentrate 1000 lb of dilute acid?
14. Hydrochloric acid, (?(60°/60°F) = 1.20, is prepared by absorbing HCl gas
at 80°F in water that is admitted to the absorption system at 50°F. If the final
|-.cid leaves the system at 70°F, how much heat is withdrawn from the equipment
per 1000 lb of acid produced?
15. Assume that pure sulfuric acid and water, both at 70°F, are mixed under
adiabatic conditions. If the acid is gradually added to the water, what is the maxi-
mum temperature that can be attained?
16. Two sulfuric acid solutions of 10% and 80% concentrations are to be mixed
under adiabatic conditions. If the stronger solution is gradually added to the
weaker, what is the maximum temperature attained when initial solutions are at 18°C?
17. Calculate the heat evolved when 5 lb HCl gas at 80°F are dissolved in 20 lb
of 10% HCl at 60°F, to form a solution at 60°F (Fig. 52).
18. Calculate the resultant temperature when 10 lb of water at 120°F are added
to 10 lb of 40% HCl at 60°F (Fig. 52).
19. Calculate the heat required to concentrate 40 lb of 5% HCl at 120°F to 8 lb
of 20% HCl at 120°F with the vapors leaving at 120°F (Fig. 52).
20. Calculate the final temperature when 100% H2SO4 at 60°P is diluted with a
solution containing 10% H2SO4 at 100°F to form a solution containing 50% HzSOj
(Fig. 53).
21. In the following table aie values of the standard integral heats of solution, AH„
in calories per gram-mole, of liquid acetic acid in water to form solutions containing
m moles of water per mole of acid. (From International Critical Tables, Vol. V,
page 159.)
m ^H. m AH.
0.25 + 70 sioo + 24
0.58 +126 - 6.19 - 13
1.11 +149 ' 30.00 - 92
1.42 +149 ': '• 63.3 -107
1.95 + 130
(a) Using the method of tangent slope, calculate the differential molal heat of
solution of acetic acid in a solution containing 15% acetic acid by weight.
(6) Using the method of tangent intercepts, calculate the differential molal
heats of solution of acetic acid and of water in a solution containing
50% acetic acid by weight.
22. From the data of Problem 21 calctilate the heat, in Btu, evolved when 10 lb
of acetic acid are added to 1000 lb of a solution containing 50% acetic acid by weight.
Perform the calculation both from integral heat of solution data and from differential
heat of solution data derived in Problem 21.
23. From the integral heat of solution data of Problem 21 calculate the heat, in
calories, which is evolved at 18°C when 1 liter of aqueous acetic acid containing 75%
acid by weight is diluted to 2 liters by the addition of water.
24. The heats of mixing, in calories per gram-mole of solution, of carbon tetra-
chloride (CCI4) and aniline (C6H6NH2) at 25°C are given in the following table.
(From the International Critical Tables, Vol. V, page 155.)
e fraction Mole fraction
CCI4 AH CCI4 AH
0.0942 98 0.6215 288
0.1848 169 „ 0.7175 270
0.3005 237 , 0.7888 246
0.4152 282 . ', 0.8627 188
0.4827 291 0.9092 149
0.5504 298
(a) Calculate the integral and differential molal heats of solution of each com-
ponent in a solution containing 50% CCI4 by weight.
(6) Calculate the heat evolved, in Btu, when 1 lb of aniline is added to a
large quantity of solution containing 40% CCI4 by weight.
318 THERMOCHEMISTRY [CHAP. V I I I
26. (a) For one of the binary systems from the following list, determine the dif-
ferential heat of solution of each component at each of the ooncentra-
tions listed.
From International Critical Tables, Vol. V
Carbon Tetrachloride Carbon Disulfide Chloroform
Benzene Acetone Acetone
Veight % Weight % Weight %
CCI4 AH CS2 AH CHCU AH
10 0.301 10 5.78 . 10 - 4.77
20 0.598 20 11.80 / 20 - 9.83
30 0.816 30 16.45 30 -14.31
40 0.963 40 19.92 40 -19.38
50 1.030 50 20.93 50 -23.27
60 1.030 60 20.80 60 -25.53
70 0.912 70 17.62 70 -25.07
80 0.699 80 16.15 80 -21.55
90 0.452 90 10.80 90 -13.56
t = 18°C t = 16°C t = 14°C
AH = heat of mixing, as joules per gram of solution. (1 joule = 0.2389 calorie.)
(b) Prepare a plot showing the differential heats of solution (calories per gram
of component) and the heats of mixing (calories per gram of solution)
as a function of the composition, expressed as weight per cent.
(c) Calculate the integral molal heat of solution of each component for "a solu-
tion containing 50 weight per cent of each component.
(d) Calculate the Btu evolved when 1 lb of the first-named component is
dissolved in a large volume of solution containing 60% by weight of the
first-named component.
26. Activated charcoal is used for the recovery of benzene (CeHe) vapors from a
nuxture of inert gases. Calculate the heat evolved, in Btu, per pound of benzene
absorbed on a large quantity of charcoal at 0°C, when the charcoal contains 0.25 lb of
benzene per pound of charcoal.
27. Calculate the heat evolved, in Btu, when 1 lb of SO2 gas is adsorbed on 6.0 lb
of outgassed silica gel at 0°C.
28. Estimate the heat evolved, in Btu, in completely wetting 10 lb of dried clay
with water.
29. Silica gel contains 12% H2O by weight. Calculate the heat, in calories, evolved
when 2.0 kg of this material at 0°C are completely wetted with water.
30. Calculate the heat evolved, expressed in Btu, when 100 lb of silica gel (com-
mercially dry basis) adsorbs 25 lb of water at 70''F.
31. Using the data in Table XVIII and Fig. 59, calculate the heat of complete
wetting, expressed in Btu, for the following solids when completely wetted with the
liquid specified.
(o) 100 lb of dry starch, wetted with water.
(6) 100 lb of dry clay, wetted with ethyl alcohol.
(c) 100 lb of animal charcoal containing 8% water when wetted completely
with water. \
32. From International Critical Tables obtain the following data:

(a) Heat of formation of SnBr4(c); calories per gram-mole.
(b) Heat of fusion of SnBr4(c); calories per gram-mole.
(c) Integral heat of solution of SnBr4(i) in a large amount of water; calories
per gram-mole.
(d) Integral heat of solution of OaCff) in water to form a dilute solution; Btu
per pound-mole.
(e) Integral heat of solution of 1.0 mole of B.Cl(g) in 400 moles of water; Btu
per pound-mole.
(f) Heat of transition of rhombic to monoclinic sulfur; calories per gram-mole.
(g) Standard heat of combustion of o-toluic acid (C8H802)(s); calories per
gram.
(h) Heat of mixing of CSiil) and ethyl acetate (C4H8O2)(0; calories per
gram-mole of mixture.
(i) Total heat evolved in the adsorption of 65 co of CO2 on 1 gram of out-
gassed coeoanut charcoal at 0°C; calories.
0') Differential heat of adsorption of CH3OH vapors on activated charcoal
containing 100 cc of vapor per gram of charcoal at 0°C; calories per
gram-mole.
(fc) Heat of wetting of dried bone charcoal with gasoline; calories per gram of
charcoal.
33. 200 grams of 20% CaClj at 40°C and 300 grams of CaCls-OHjO at 20°C are
mixed.
(a) Estimate the temperature of the final mixture.
(6) How much heat must be removed to solidify completely the mixture?
34. Estimate the final temperature when 100 lb of air (dry basis) initially at
70°F and a humidity of 0.010 lb per pound of dry air are mixed with 10 lb of sihca
gel, initially commercially dry and at 70°F. The final humidity of the air = 0.001.
Assume no heat losses from the system.
35. Calculate the heat removed when 100 lb of air (dry basis) initially at 70°F
and a humidity of 0.010 lb per pound of dry air are mixed with 10 lb of silica gel,
initially commercially dry and at 70°F. The final humidity of the air is 0.001.
The gel and air are kept at 70°F.
36. Calculate the heat removed when 100 lb of air (dry basis) initially at 70°F
and 70% relative humidity are mixed with 10 lb of silica gel, initially commercially
dry at 70°F. The final relative humidity of the air is 10% and the final temperature
of the air and gel is 80°F.
37. Solve Illustration 17, assuming the average temperature of air is 85°F and
average temperature of gel is 95°F. (Note: Actually beds of silica gel are operated
under intermittent-flow adiabatio conditions. The treatment of the adiabatic case
presents formidable difficulties because of the fluctuating temperature waves in the
bed.)
38. Calculate the standard heat of reaction, in kcal, accompanying the reduction
of 20 kg of Fe203 by carbon (coke) to form 12 kg of Fe(s). The only other products
leaving the process are FeO(s) and CO{g).
39. The heat requirements are to be estimated for a low-temperature reduction
process applied to a magnetite ore. The ore averages 90% Fe304; the balance is
inert gangue, largely Si02. The process is to be conducted in an externally heated
retort which is closed except for the two openings which are to serve as entry for the
charge and as exit for the product. The opening for admitting the charge also serves
as a vent for the escape of the gas formed by the reaction between the magnetite and
the reducing agent. On the basis of laboratory tests, it is anticipated that 95% of
the iron will be reduced to the metallic state, with the remainder being reduced to
FeO. The reduckig agent to be used is a metallurgical coke containing 85% carbon
and 16% ash, the latter being largely Si02. The coke is to be charged 300%
in excess of the theoretical demand for complete reduction. The solid discharged is
thus a mixture of Pe, FeO, unused coke and Si02 from the ore and from the coke used
up in the reduction process. The gas escaping from the retort will be practically pure
CO, formed by the reaction between the magnetite and the coke.
Analyses - - >
Coke: Carbon = 86% Ore: FesOi = 90%
Ash = 15% • SiOs = 10% ' ;
Temperatures Specific heats

Entering ore = 200°C PesO*: 0 ° C . . . 0.151 ."'••
Entering coke = 200°C 100°C . . . 0.179 v-w»
Leaving solids = 950°C 200°C . . . 0.203
Leaving gas = 960°C . 300°C . . . 0.222 ,^
Calculate the material balance for the process, on the basis of 2000 lb of metallic
iron produced.
Estimate the heat requirements for the process, as Btu per 2000 lb of metallic iron
produced.
40. By the combustion at constant volume of 2.0 grams of 'H.2{g) to form liquid
water at 17°C, 67.45 kcal are evolved. Calculate the quantity of heat which would
be evolved were the reaction conducted under a constant pressure at 17°C.
41. When 1.0 gram of naphthalene (CioHs) is burned in a bomb calorimeter, the
water formed being condensed, 9621 calories are evolved at 18°C. Calculate the heat
of combustion at constant pressure and 18°C, the water vapor remaining uncon-
densed.
42. A fuel oil analyzes 80% C and 20% Ha by weight. The standard heat of
combustion of this oil is determined in an oxygen bomb calorimeter. Calculate the
correction that must be applied to get the heat of reaction at constant pressure.
Which is the greater, the heat of reaction at constant pressure, or the heat of reaction
at constant volume?
43. Calculate the actual heat of reaction in cal per g-mole for each of the follow-
ing reactions. Reactants and products are at a constant pressure of 1 atm. The
temperature of each reactant and product is as indicated. Assume the mean specific
heat of NaHSOi (0 to 250°C) to be 0.23 and NajSOi (0 to 600°C) to be 0.26.
S02(s) + iOaCs) -> SO,(ff) .;. / /
450°C 460°C 460°C / /
NaHS04(s) + NaCl(s) -^ Na^SOiCs) + HClfe) /
.250°C 30°C 600°C 300°C /.
SiOiis) + 3C (coke) - • SiC(s) + 2C0(g)
18°C 18°C 1800°C 1200°C
2NaCl(s) + SO^Csf) + iO,(g) + UÔig) -
400°C 400'C 400°C 400°C
Na2S04(s) + 2HC1(9)
400°C 400°C
44. Derive a general equation for each of the following reactions which will express
the heat of reaction as a function of temperature, with the reactants and products
at the same temperature. Base all equations on the assumption that HjO is in the
vapor state.
CO(s) +iO,{g)-.CO,{g)
. HsO (?) + CO ig) ^ H2 (s) + COj (ff)
N^to) +3Hs,(?)^2NH3(<7)
SOsCff) +i02(ff)-»S03(ff)
HaCff) +a(ff)->2HCl(ff)
46. Sulfur dioxide gas is oxidized in 100% excess air with 80% conversion to SO3.
The gases enter the converter at 400°C and leave at 450°C. How many kcal are ab-
sorbed in the heat interchanger of the converter per kilogram-mole of SO2 introduced?
46. Calculate the heat of neutralization in calories per gram-mole of NaOH (wi = 5)
with HCl(ni = 7) at 25°C where rii = moles H2O per mole of solute.
47. A bed of petroleum coke (pure carbon) weighing 3000 kg, at an initial temper-
ature of 1300°C, has saturated steam at 100°C blown through it until the tempera-
ture of the bed of coke has fallen to 1000°C. The average temperature of the gases
leaving the generator is 1000°C. The analysis of the gas produced is CO2, 3.10%;
CO, 45.36%; H2, 51.53% by volume, dry basis. How many kilograms of steam
are blown through the bed of coke to reduce the temperature to 1000°C? Neglect
loss of heat by radiation and assume that no steam passes through the process
undecomposed.
48. In Problem 47,20% of the steam passes through the coke undecomposed. How
much steam is blown through the bed of coke to reduce its temperature to 1000°C?
Neglect loss of heat by radiation.
49. Steam at 200°C, 50° superheat, is blown through a bed of coke initially at
1200°C. The gases leave at an average temperature of 800°C with the following
composition by volume on the dry basis:
Hs 53.5%
, . CO 39.7%
COs 6.8%
' : 100.0%
Of the steam introduced 30% passes through undecomposed. Calculate the heat
of reaction in kcal per kilogram-mole of steam introduced. Mean specific heat of
coke (0 to 1200°C) = 0.35.
50. Calculate the number of Btu required to calcine completely 100 lb of limestone
containing 80% CaCOs, 11% MgCOa, and 9% H2O. The lime is withdrawn at
1650°F and the gases leave at 400°F. The limestone is charged at 70°F.
61. Limestone, pure CaCOa, is calcined in a continuous vertical kUn by the com-
bustion of producer gas in direct contact with the charge. The gaseous products
of combustion and calcination rise vertically through the descending charge. The
limestone is charged at 18°C and the calcined lime is withdrawn at 900°C. The
producer gas enters at 600°C and is burned with the theoretically required amount
of air at 18°C. The gaseous products leave at 200°C. The analysis of the producer
gas by volume is as follows:
CO, 9.21%
Oa 1.62%
CO 13.60%
N, 75.57%
100.00%
Calculate the number of cubic meters (0°C, 760 mm Hg) of producer gas, required
to burn 100 kg of limestone, neglecting heat losses and the moisture contents of the
air and producer gas.
52. A fuel gas of the following composition at 1600°F is burned in a copper melting
furnace with 25% excess air at 65°F.
CH4 40%
Hi 40%
CO > 4%
COj 3%
Ns 11%
i O2 2%
100%
The copper is charged at 65°P and poured at 2000°F. The gaseous products leave
the furnace at an average temperature of 1000°F. How much copper is melted by
burning 4000 cu ft (32°F; 29.92 in. Hg; dry) of the above gas, assuming that the
heat lost by radiation is 50,000 Btu and neglecting the moisture contents of the
fuel gas and air?
63. One thousand cubic meters of gas, measured at standard conditions (0°C, 760
mm and dry), containing 20 grams of ammonia per cubic meter (as measured above)
are passed into an ammonia absorption tower at a temperature of 40°C, saturated
with water vapor, and at a total pressure of 740 mm. The gas is passed upward
countercurrent to a descending stream of water which absorbs 95% of the incoming
ammonia. The gas leaves the tower at 38°C, saturated with water vapor. Six
hundred kilograms of water enter the top of the tower at 20°C. What is the tem-
perature of the solution leaving the tower, neglecting heat losses? Assume the mean
molal heat capacity of moisture-free, ammonia-free gas to be 7.2.
54. Pure HCl gas comes from a Mannheim furnace at 300°C. This gas is cooled
to 60°C in a silica coil and is then completely absorbed by passing it countercurrent
to a stream of aqueous hydrochloric acid in a series of CeUarius vessels and absorp-
tion towers. The unabsorbed gas from the last CeUarius vessel enters the first
absorption tower at 40°C. Fresh water is introduced in the last absorption tower
at 15°C and leaves the first absorption tower at 30°C containing 31.45% HCl (20° B6
ac'd). This acid is introduced into the last CeUarius vessel and leaves the first
vessel at 30°C containing 35.21% HCl (22° B6 acid). There are produced in this
system 9000 lb of 22° B6 acid in 10 hours. Calculate separately the heat removed
in the cooling coU, CeUarius vessels, and absorption towers, neglecting the presence of
water vapor in the gas stream and assuming complete absorption of HCl.
CHAPTER IX
FUELS AND COMBUSTION
Because of the universal use of the combustion of fuels for the genera-
tion of heat and power, special techniques and methods have been
developed for establishing the material and energy balances of such
processes. Each problem should be pursued independently and as
rigorously as the available experimental data permit, using the chemical
principles involved and not empirical equations.
Heating Values of Fuels. The most important property of a fuel is
its heating value, which is numerically equal to its standard heat of com-
bustion but of opposite sign. This property is usually determined by
direct experimental measurements, although methods are also given for
its estimation.
The major products of complete combustion from practically all fuels
are carbon dioxide and water. Two methods of expressing heating
values are in common use, differing in the state selected for the water
present in the system after combustion. The total heating valiie of a
fuel is the heat evolved in its complete combustion under constant
pressure at a temperature of 18°C when all the water formed and origi-
nally present as liquid in the fuel is condensed to the liquid state. The
net heating value is similarly defined except that the final state of the
water in the system after combustion is taken as vapor at 18°C. The
total heating value is also termed the " higher " or " gross " heating
value; the net is often termed the " lower " heating value. The net
heating value is obtained from the total heating value by subtracting the
latent heat of vaporization at 18°C of the water formed and vaporized
in the combustion.
Coke and Carbon. The combustible constituents of cokes and
charcoals are practically pure carbon. The heating value of such a
fuel may be predicted with accuracy suflacient for most purposes by
simply multiplying its carbon content by the heating value per unit
weight of carbon.
In Table XIV of Chapter VIII, page 253, it will be noted that heats of
formation of carbon compounds are based on a value of zero assigned
to the heat of formation of P graphite. On this basis various other
forms of elementary carbon have positive heats of formation. Several
323
324 FUELS AND COMBUSTION [CHAP. IX
types of " amorphous " carbon are included in the table, each having a
different heat of formation. These differences arise in part from differ-
ences in allotropic forms and in the surface energy of carbon in different
states of subdivision and porosity, and in part from the presence of
hydrocarbon compounds of high molecular weights and low hydrogen
contents.
The heats of combustion of the various forms of amorphous carbon
differ by the same amounts as do their heats of formation. For com-
bustion calculations the value of the heat of combustion of carbon is
taken as —96,630 calories per gram-atom or —14,495 Btu per pound.
This value is the difference between the heat of formation of carbon
dioxide and that of carbon in coke, given in Table XIV, page 253, or
-96,630 = -94,030 - 2600. •
Coal Analyses. Coal consists chiefly of organic matter of vegetable •••^^.
origin which has been altered by decomposition, compression, and heat- •^*—»
ing during long ages of inclusion in the earth's crust. In addition to
organic matter it contains mineral constituents of the plants from which
it was formed and also inclusions of other inorganic materials deposited i
in it during its geological formation.
Two types of analysis are in common use for expressing the composi-
tion of coal. In an ultimate analysis, determination is made of each of
the major chemical elements. In a proximate analysis four arbitrarily
defined groups of constituents are determined and termed moisture,
volatile matter, fixed carbon, and ash. Following are the ultimate and
proximate analyses of a typical Illinois coal :
Ultimate Proximate
Moisture 9.61 Moisture 9.61
Ash (corrected) 9.19 Ash 9.37 f
Carbon 66.60 Volatile matter 30.68 - /
Net hydrogen 3.25 Fixed carbon 60.34 - j
Sulfur 0.49
Nitrogen 1.42 ' - , 100 00 /
Combined H20 9.44 ' /.
100.00 / /
The proximate analysis of coal should be carried out according to an

arbitrarily standardized procedure which has been recommended by the
United States Bureau of Mines. The details of this method are de-
scribed in most books on methods of technical analysis. The determina-
tions may be rapidly and easily carried out, and the majority of the con-
tracts and specifications for the purchase of coal are based on this
analysis. The tedious methods of ultimate analysis are completely
CHAP. IX] COAL ANALYSES 325
carried out only when necessary to serve as a basis for energy and mate-
rial balance calculations. However, the suKur content is of particular
interest, and determination of this element frequently accompanies the
proximate analysis.
In both schemes of analysis " moisture " represents the loss in weight
on heating the finely divided coal at 105°C for one hour. The material
termed " ash " in the proximate analysis is the residue from complete
oxidation of the coal at a high temperature in air. This quantity is
needed for calculating the quantity of refuse formed in the ordinary
combustion of the coal. However, the ash determined in this manner
does not accurately represent the mineral content of the original coal
because of the changes which take place during combustion. An
important mineral component of many coals is iron pyrites, FeS2.
In combustion this is oxidized to form Fe203, which is weighed in the
residual ash, and SO2 gas. In the oxidation of pyrites 4 gram-atoms
(128 grams) of sulfur are replaced by 3 gram-atoms (48 grams) of
oxygen, a loss in weight equal to | times the weight of pjTitic sulfur
present. Thus, in order to determine the actual mineral content of the
coal, including the pyritic sulfur, it is necessary to add to the ash-as-
weighed, a correction equal to f of the pyritic sulfur content. To
determine the actual mineral content, not including the pyritic sulfur,
a correction equal to | of the pyritic sulfur must be subtracted from
the ash-as-weighed. Other less important corrections may also be
applied to the ash. Unless otherwise designated, " ash " refers to ash-
as-weighed.
To obtain the ultimate analysis, direct determinations are made of
carbon, sulfur, nitrogen, and hydrogen by the usual analytical meth-
ods. The moisture and ash are determined by the standardized proce-
dures of the proximate analysis. The percentage oxygen content is
then taken as the difference between 100 and the sum of the percentages
of carbon, hydrogen, sulfur, nitrogen, and corrected sulfur-free ash.
It is recommended that, for this calculation, the corrected ash be esti-
mated by assuming that all sulfur in the coal is present in the pyritic
form. On this basis,
% corrected ash = % ash-as-weighed — | ( % S) =
% mineral content — % S
where % S = percentage sulfur content of coal. This correction
represents only an approximation, since not all sulfur is pyritic and
other changes in the mineral constituents may take place in combustion.
More refined methods for estimating oxygen content are not ordinarily
justified. • J • :
In reporting the ultimate analysis it is convenient to consider that all

oxygen is in combination with hydrogen to, form moisture and " com-
bined water." The surplus hydrogen, above that required to combine
with the oxygen, is termed " net " or " available " hydrogen. This
represents the hydrogen present in the form of hydrocarbons and avail-
able for further oxidation.
Ranji of Coal. The sum of the fixed carbon and volatile matter of a
coal is termed the combustible. The Bureau of Mines has published
extensive tables' of the ultimate analyses of coals representing hundreds
of coal deposits throughout the United States. If the source of a coal
is known the ultimate analysis of its combustible matter can be obtained
with fair reliability from these tables since the composition of combus-
tible material in any one coal bed is nearly constant. In every coal
sample it is necessary, however, to make separate determinations of ash
and moisture contents.
The fuel ratio of a coal is defined as the ratio of its percentage of fixed
carbon to that of volatile matter. The rank of the coal, whether bitumi-
nous, or anthracite, may be estimated from the fuel ratio. The gener-
ally accepted classification of coals and the corresponding ranges of fuel
ratios are as follows: /'
, TABLE XIX /
RANK OP COALS '
Rank Fuel Ratio •

Anthracite between 10 and 60
Semi-anthracite between 6 and 10
Semi-bituminous between 3 and 7
Bituminous between | and 3
Fuels of lower rank than bituminous, namely, sub-bituminous and lig-

nite, may have fuel ratios within the bituminous range but are charac-
terized by higher water or oxygen contents.
The classification of coals on the basis of fuel ratio is not entirely satis-
factory for many purposes. Several other methods^ have been devel-
oped which give more nearly exact differentiation.
Heating Value of Coal. The total heating value of a coal may be
determined by direct calorimetric measurement and is usually expressed
in Btu per pound. The net heating value is obtained by subtracting
from the total heating value the heat of vaporization at 18°C of the
' U. S. Bureau of Mines, Bulletin 123. With permission.
' Haslam and Russell, Fv^ and Their Combustion, McGraw-Hill Book Co.
(1926). 5^' -
CHAP. IX] HEATING VALUE OF COAL 327
water present in the coal and that fonned by the oxidation of the avail-
able hydrogen. Thus,
Net H.V. = Total H.V. - 9 X H X 1056 (1)
where H.V. = heating value, Btu per pound
H = weight fraction of total hydrogen, including available
hydrogen, hydrogen in moisture, and hydrogen in com-
bined water ., ,'...,
When the heating value of coal is determined by burning in a calorim-

eter the sulfur is oxided to form sulfuric acid. Normally the sulfur in
coal burns to form sulfur dioxide only, so that a correction should be
made to the calorimetric value for the heat evolved in forming sulfuric
acid from sulfur dioxide and water.
Many attempts have been made to develop a method of calculating
the heating value of coal from its proximate analysis. None of these
methods are sufficiently reliable to justify their use except as approxima-
tions.
A fair approximation to the heating value of a coal may be obtained
by considering that each of the combustible constituents, carbon, avail-
able hydrogen, and sulfur, is present in its elementary state. On the
basis of this assumption the heating value is the sum of the quantities
of heat evolved in the combustion of each of these elements, using for
carbon the heating value of amorphous carbon and for sulfur the heating
value of FeSa. The respective heats of combustion, in Btu per pound,
may be obtained from the data of Table XIV, page 253. It is assumed
that sulfur is burned to SO2.
Element Heat of Combustion
Carbon,.'.... 14,495 Btu per pound
Hydrogen (total) 61,000 Btu per pound
Hydrogen (net) 51,550 Btu per pound
Sulfur (as FeSz) 6,770 Btu per pound
Then,
Total H.V. = 14,495C + 61,000Ha + 5770S (2)
where
H.V. = heating value, Btu per pound
C, Ho, S = weight fractions of carbon, available hydrogen, and
sulfur, respectively
Equation (2) is known as Dulong's formula. It is not theoretically

sound because it neglects the heats of formation of the compounds of
carbon, sulfur, and hydrogen which exist in the coal. However, as

previously pointed out, the heats of formation of hydrocarbon compounds
are small in comparison to their heats of combustion, and the results of
the above equation are rarely in error by more than 3 per cent. The
experimentally observed total heating value of the coal whose analysis
is given above, page 324, was 11,725 Btu per pound. Applying Equa-
tion (2), the heating value would be predicted as (14,495 X 0.666) +
(61,000 X 0.0325) + (5770 X 0.0049) = 11,920 Btu per pound, an
error of +1.7 per cent.
Because of the fact that the heating value of coal is more easily deter-
mined than its ultimate analysis, the use of Equation (2) for calcula-
tion of heating value is rarely advantageous. I t is more useful as a
means of predicting the available hydrogen content from experimentally
determined values of carbon and sulfur contents and heating value.
A useful relationship has been pointed out by Uehling* between the
heating value of coal per pound of total carhon and its rank. I t was
found that, for each rank of coal, the heating v&lue per pound of total
carbon is nearly constant, rarely varying by more than 2 per cent.
It was also found that the weight of available hydrogen, per pound of
total carbon, varies but little among coals of the same rank. On this
basis, standard average heating values and available hydrogen contents,
per pound of total carbon, were established for the different ranks of
fuel. These values, contained in the following table, were based on the
published results of a large number of analyses carried out by the United
States Bureau of Mines.
TABLE XX
STANDARD HEATING VALUES AND NET HYDROGEI^ CONTENTS OP COAL
H.V.' = lieating value, Btu per pound of total cartjon

Ho = available hydrogen content, pounds per pound of total carbon
Rank H.V.' H^
"• Coke 14,495 0.0
:. Anthracite 16,100 0.029 \ .r"
Semi-bituminous 17,400 0.049
Bituminous 17,900 0.054
Sub-bituminous 17,600 0.045
Lignite 17,100 0.037 /
From the data of Table XX, the heating value of the bituminous coal
whose analysis was given on page 324 would be predicted as 17,900
X 0.666 = 11,920 Btu per pound, in error by only 1.7 per cent as com-
' E . A. Uehling, Heat Loss Analysis, McGraw-Hill Book Co. (1929). With per-
^ mission. , .•.-:
CHAP. IX] CHARACTERIZATION OF PETROLEUM 329
pared to the experimentally determined value of 11,725. The available

hydrogen content would be predicted as 0.666 X 0.054 X 100 = 3.6 per
cent as compared to the experimentally observed value of 3.25 per cent.
Because of the relative difficulty of the total carbon determination as
compared to the calorimetric determination of heating value, the rela-
tionships pointed out by Uehling have their greatest value in predicting
the ultimate analysis from the experimentally determined heating value.
From only the heating value and the data of Table XX, good approxima-
tions to the total carbon and available hydrogen content of a coal may
be predicted. In view of the fact that the composition of a coal sample
will frequently vary by as much as 5 per cent from the true average com-
position of the coal from which it was taken, the accuracy of these
predictions is often as great as is justifiable for calculations of energy
and material balances. ' "A
; , PETROLEUM • .
Petroleum oils are complex mixtures of hydrocarbons including four
important series of compounds: paraffins, naphthenes, olefins, and aro-
matics. These compounds differ in hydrogen content in the order listed,
paraffins having the highest hydrogen content and aromatics the lowest.
In naturally occurring petroleums the first two series predominate; in
cracked products formed by decomposition of natural oils large quanti-
ties of olefins and aromatics may also be present. In addition to hydro-
carbons varying quantities of sulfur, oxygen, and nitrogen compounds
are generally present.
Because of the complexity of petroleum fractions determination of the
actual compounds present is generally impossible. Elementary analyses
may be made, determining carbon, hydrogen, sulfur, and nitrogen as
for coal. Data of this type are available in the publications of the U. S.
Bureau of Mines for many naturally occurring petroleums. However,
such analyses give httle indication of the actual character of an oil and
its thermal properties. Approximate methods have been developed
whereby much of this information may be estimated from easily de-
termined physical properties such as the distillation or boihng range,
the specific gravity, and the viscosity.
Characterization of Petroleum. For general correlation of the aver-
age physical properties of petroleum stocks of different types, it is neces-
sary to develop a means of quantitatively expressing the general char-
acter of the oil. Paraffin hydrocarbons of maximum hydrogen content
may be considered as one extreme and aromatic materials of minimum
hydrogen content as the other.
To serve as a quantitative index to this property, which may be
termed parafEnicity, the U.O.P. characterization factor^ has been

developed and empirically related to six conunonly available laboratory
inspections. Although this factor is not an exact measure of chemical
type and does not show perfect constancy in a homologous series, these
disadvantages are, to a considerable extent, offset by its simpUcity and
convenience of definition and use.
The definition of the U.CXP. characterization factor arose from the
observation that, when a crude oil of supposedly uniform character is
fractionated into narrow cuts, the specific gravity of these cuts is ap-
proximately proportional to the cube roots of their absolute boiling-
points. The proportionality factor may then be taken as an index of
the paraffinicity of the stock. Thus
. ^=V /. : ^'\
where . • . ^ ^^ • - - ,i , , •,-
K = U.O.P. characterization factor i
TB = average boihng point, degrees Rankine
G = specific gravity at 60°F
When dealing with mixtures of wide boiling range a special method qf

obtaining the average boiling point as described by Watson and Nelson
must be used. For narrower cuts the 50 per cent point of the Engler
distillation may be taken as the average boiling point.
The characterization factor shows fair constancy throughout the
boiling range of a number of crude oils and for others may either increase
or decrease in the higher boihng range. In the paraffin series fair con-
stancy for the average of the reported isomers exists up to a boiling tem-
perature of 7G0°F. Values of the characterization factor range as
follows:
Pennsylvania stocks 12.2-12.5
Midcontinent stocks 11.8-12.0 '
Gulf Coast stocks 11.0-11.5
Cracked gasolines 11.5-11.8
Cracking plant combined feeds... . 10.5-11.5
Recycle stocks 10.0-11.0
Cracked residuums 9.8-11.0
The characterization factor is readily calculated from Equation (3)

from only the specific gravity and average boiling point, or it may be
read directly from API gravity and average boiling point by interpola-
* Watson, Nelson, and Murphy, Ind. Eng. Chem., 25, 880 (1933); 27,1460 (1935).
With permission.
CHAP. IX] CHARACTERIZATION OF PETROLEUM 331
tion between the curves of Fig. 63. In this figure, API gravities are
plotted as ordinates and average boiling points as abscissas with lines
of constant K from Equation (3). The relationship between specific
gravity and degrees API is shown by Fig. B in the Appendix.
It has also been found that a fair empirical correlation exists between
FIG. 63. Molecular weights, critical temperatures and characterization

factors of petroleum fractions.
the characterization factor and the viscosity-gravity relationship at a
given temperature. Parafiinic stocks have high viscosities as compared
with aromatic materials of the same gravities.
Because of uncertainties of molecular aggregation at low tempera-
tures, the viscosity measurements used for physical correlations should
be made at as high temperature as possible. In Fig. 64* viscosity in
centistokes at 122°F is plotted against API gravity for stocks of constant
characterization factors. By use of the centistoke scale of viscosity
' W a t s o n et al., ibid. . .' •.
the entire range of fractions from light gasoHnes to heavy residues is

covered in a single relationship. A chart ior conversion of common
viscometer readings into centistokes is included in Fig. C, page 442.
Lines of constant boiling point are plotted on Fig. 64 resulting from
combination of the relationships between characterization factor from
boiling point and viscosity data. These lines permit an approximation
N, 1 1^
1
\
1000
800
-
-
—\ 5
\
14 1—^
kS,
\1
^
\i
- 4 -lobo-
^.
IV
\ 1980
i r \ \^ 4
K 1
1
lioo- 00
IS.^ \
O 80 ~
=fe
T 4
-\
^\ 4 ' 4
4" V^
S^ 40 4 \ ^ rj: K \^
4^
s
0
B \
\
•S 10 -
.\
VT •s•
1 8 - t V\ — > -
n b fc
j ^
\ i
« 'i
3V
• ^ ^ ^
\600 ftv
V-
\ . 500
^40(
1.0 -
.8 - —^
N
s
^ s \ -^V
V (JO
•\ 100
Characterization Factor
1 1 1 1 1 1 1
10.0 10. 5 ll.O
1 1 1
s N ^
1 .5 12.0 1 2.5
-5 0 5 10 15 20 25 3& 35 40 45 50 55 60 65 70 75 80 85
Degrees, A. P. I.
FIG. 64. Characterization factor from viscosity at 122°F.

to the boiling point from only viscosity and gravity data. This rela-
tionship is particularly useful for heavy stocks on which boiling point
data can be obtained only under high vacuum. However, because of
the large change in viscosity with a slight change in the gravity of heavy
stocks, boiling points estimated in this way may be considerably in
error, sometimes as much as 50°F for the heavier residues. Similar
charts were developed on the basis of viscosities at other temperatures.*
8 Watson, Nelson, and Murphy, Ind. Eng. Chem., 27, 1460 (1935).
CHAP. IX] CRITICAL PROPERTIES 333
Molecular Weights of Petroleum Fractions. The average molecular

weights of petroleum fractions may be satisfactorily estimated from
average boiling point and gravity. Aromatic stocks of low characteriza-
tion factors have lower molecular weights than paraffinic materials of
the same average boiling points.
The relationship between molecular weight, characterization factor,
boiling point, and API gravity is included in the curves of Fig. 63.
By interpolation between these curves, molecular weights may be esti-
mated, with errors rarely exceeding 5 per cent. If boiling-point data
are not available, the boiling point may be estimated from viscosity
using Fig. 64. »: • ; * i,,'^ » *
16
^
S
^
U
S ^ ^ <?-
•S12
\
;^^y--^
y^^
'y^
e< ^
f'
& ,,«? oo-

& \ =
y\y^
^ ^
y
,oa? î)^S^H ^ ^ ^ : ^
y
^ ^
P
if'-
> ^
1
10.0 10.5 11.0 11.5 12.0
Characterization Factor K
. F I G . 65. Characterization factor vs. weight % Ha.
Critical Properties. The critical temperature curves of Fig. 63 were

calculated directly from Equation (III-8), page 69, but are in satisfac-
tory agreement with the existing data on petroleum.
Critical temperatures estimated from Fig. 63 are applicable with little
error to pure hydrocarbons, narrow petroleum cuts, or wide-boiling
mixtures if a proper method of obtaining average boiling point is used.
Correct methods of averaging have been developed by Smith and
Watson.' Critical pressures may be estimated by the methods de-
scribed in Chapter III.
» R. L. Smith and K. M. Watson, Ini. andEng. Chem,, 29, 1408 (1937).
Hydrogen Content. The curves of Mg. 65 represent a relationship

between hydrogen content and characterization factor for materials of
constant boiling points.
Figure 65 combined with the preceding charts permits estimation of
hydrogen content from a knowledge of only the specific gravity and one
other property. Ordinarily the error will be less than 0.5 per cent, based
on the total weight of the oil, except for highly aromatic, low-boiling
materials.
Petroleum oils ordinarily contain little ash and in the absence of
specific data may be assumed to be 97 per cent carbon and hydrogen
with the remainder oxygen, nitrogen, sulfur, and ash. This assump-
tion is unsatisfactory for oils of high sulfur content, such as certain
California or Mexican stocks, or where salts are present with water
in partial solution and suspension. Specific data should be obtained on
such stocks.
0.9
/' •\0
^
y^
^ ^
ND K C A ^ ^ ^ ^
^' .^< *•" .^^ ^5/
O ^- ^
A
.'^'^"^
/
Â
^,
^
^ • ^
0 11 12 13
40O 500 600 1000

Temperature ' F
FIG. 66. Specific heat of liquid petroleum oils where K = 11.8 (mid-
continent stocks). For other stocks multiply by correction factor.
Specific Heats of the Liquid State. The following equation was

recommended by Fallon and Watson' for the specific heats of liquid
hydrocarbons and petroleum fractions at temperatures between 0°r
' Jj Fallon and K. M. Watson, presented before Petroleum Div., Am. Chem. Soc,
Pittsburgh meeting, Sept., 1943.
CHAP. IX] SPECIFIC HEATS OF THE VAPOR STATE 335
and reduced temperatures of 0.85.

Cp =[(0.355 + 0.128 X lO-^ °API) + (0.503 + 0.117
X 10-2 °API) X 10-H] [0.05K + 0.41]
where t is in °F and K is the characterization factor.
• Figure 66 is a plot of this relationship together with curves for the
individual light paraffin hydrocarbons as recommended by Holcomb
and Brown.' The curves on the main plot apply directly to Midcon-
tinent stocks whose characterization factors are approximately 11.8.
For other stocks the value read from the main plot is multiplied by a
correction factor derived as a function of K from the small plot in the
lower right-hand corner.
1.0
,
^ •4
I" Aj.
AV
doi
:A.<i
US'.??
pt*
" 0.5
' c^^*"
—
0.4
0.2
0 100 20O 30O 400 600 600 700 800 900 1000 1100 1200 1300 1400
Temperature "F
Fia. 67. Specific heats of paraffin gases at atmospheric pressure.
Specific Heats of the Vapor State. At temperatures below atmos-

pheric the molal heat capacities of complex gas molecules approach a
constant value of 7.95 whereas at high temperatures the curves express-
ing heat capacity as a function of temperature become concave down-
ward. Since this complicated form of relationship is difficult to express
in a single equation, Fallon and Watson* proposed the use of the follow-
ing two forms of equations:
For temperatures from 50 to 1400°F:
Cp = a -I- b r -1- cT^
For temperatures from - 3 0 0 to -f200°F (never above 200°F):
cp = 7.95 -I- uT
where T is in °R and Cj, is the molal heat capacity.
'Holcomb and Brown, Ind. Eng. Chem., 34, 590 (1942). With permission.
In Table X X I are values of the constants of these equations for the

light paraffins and olefins at low pressures. Also included are the
constants of an equation expressing the approximate differences be-
tween the heat capacities of paraffins of more than three carbon atoms
and those of the corresponding olefins.
, «. .,. -ÂBLE XXI
HEAT CAPACITIES OP HYDROCARBON GASES
For 50 to 1400°F, Cp = a + bT + cP
For -300 to 200''F, Cp = 7.95 + uT'
T = "Rankine
Compound a 6 XIO' - c X 10^ M V
Methane 3.42 9.91 1.28 6.4 X 10-" 4.00

Ethylene 2.71 16.20 2.80 8.13 X 10-" 3.85
Ethane 1.38 23.25 4.27 6.20 XlO-5 1.79
Propylene 1.97 27.69 5.25 2.57 X 10-3 1.26
Propane 041 35.95 6.97 3.97 X 10-3 1.25
n-Butane 2.25 45.40 8.83 0.93 X 10-2 1.19
i-Butane 2.30 45.78 8.89 0.93 X 10-2 1.19
Pentane 3.14 55.85 10.98 3.9 X 10-2 1.0
ParafEn-olefin -1.56 8.26 1.72
For specific heats of vaporized petroleum fractions the following

equation is recommended'" for the temperature range of 0-1400°F:
C„ - (0.0450Z - 0.233) + (0.440 + 0.0177X) X lO-î
- 0.1530 X lO'H^
where t is in °F, Cp is specific heat, and K is the characterization factor.
This relationship was found to be independent of gravity or boiling
point and is plotted in Fig. 67 which is appHcable to all petroleum frac-
tions and hydrocarbons containing more than four carbon atoms.
Heat of Vaporization. Heatsof vaporization of hydrocarbons and pe-
troleum fractions under atmospheric pressure were calculated by Fallon
and Watson'" by differentiating Equation (III-16), page 73, and sub-
stituting in the Clapeyron Equation ( I I I - l ) . The resultant correlation
is plotted in Fig. 68 relating heat of vaporization to boiling point and
either API gravity or molecular weight. When working with pure
compounds use of the molecular weight is preferable. Heat of vapori-
zation at other pressures may be obtained by means of Equation
(VII-32), page 233.
'" Fallon and Watson, presented before Petroleum Div., Am. Chem. Soc, Pitts-
burgh meeting, Sept., 1943. To be pubhshed Nat. Petr. News, 1944. ;
CHAP. IX] HEAT OF COMBUSTION 337
The values of Fig. 68 are in close agreement with the Kistyakowsky
equation for low-boihng compounds but are considerably higher for
high-boiling materials. It- is believed that these higher values repre-
sent a more reliable extrapolation than that of the Kistyakowsky
equation.
180
\^^
X^ s*
\o
if ^"''-%
^
B
\ /
§120
a
^
t>'h .•V*
•,n^
R!Jl^
>
CO
"Sioo
i,?0-
>A g .i» ,
)
*)
f>"
80 vi
-?!. >1?
(J h-^
• ' - .
v < % ^0
100 200 300 400 500 600 700 800 900 1000
Average Boiling Point,°P
Pressure = 1.0 atmosphere
FIG. 68. Heats of vaporization of hydrocarbons and petroleum fractions. {Note:

Base on molecular weights rather than API if available.)
Heat of Combustion. Average values of heats of combustion of

petroleum fractions and hydrocarbons are plotted in Fig. 69 as a function
of API gravity and characterization factor. These are total heating
values, corresponding to formation of Uquid water at 60°F.
Ulustration 1. A fuel oil has an API gravity of 14.1 and a viscosity of 150 Say-
bolt Furol seconds at 122^. Estimate the characterization factor, average boiling
point, hydrogen content, specific heat at 200°F, heating value, and average molecular
weight of this oil.
From the conversion chart, Fig. C in the Appendix, it is found that 150 Saybolt
Furol seconds is equivalent to 320 centistokes.
' ' (a) Characterization factor, Fig. 64 =11.35
(b) Average boiling point, Fig. 63 or 64 = 880°F
(c) Hydrogen content, Fig, 65 =11.5
(d) Specific heat 200°, Fig. 66 - 0.485 X 0.975 = 0.473
, (e) Average molecular weight. Fig. 63 = 410
(/) Heating value, Fig. 69 = 18,825 Btu per lb
338 FUELS AND COMBUSTION [CHAI". IX
-20,400 1 •
^12.0
/^.11.9
-20,200
11.8
^11.7
1
O 11.6
11.5
-19,800
^ ^11.4
-19,600 m 11-3 V
^
S -19,400 i U^3^f
v// 2:1^1
= -19,200
û
.r 0.9
.8"~
CO
6
1 1U.7
; -19,000 10.6
/
-18,800
i
-18,600
/
f .10
.- 0.2
-18,400 inr
-18,200
-18,000
V
//
-17300
-17,600
10 20 30 40 60 60 70
A P I Gravity
PIG. 69. Total heats of combustion of liquid petroleum

hydrocarbons.
FUEL GAS
The standard basis which has been adopted for the expression of the
total heating value of a fuel gas is the number of Btu evolved when one
cubic foot of the gas, at a temperature of 60°F, a pressure of 30 inches
of mercury, and saturated with water vapor, is burned with air at the
same temperature, and the products cooled to 60°F, the water formed
in the combustion being condensed to the liquid state. Since gas is
rarely burned under these standard conditions of temperature and
pressure, the heating-value per standard cubic foot is not a convenient
unit for calculations. However, the unit is widely used as a basis for
specifications and legal standards.
CHAP. IX] FUEL GAS 339
Since the vapor pressure of water at a temperature of 60°F is 0.52

inch of mercury, the heating value per standard cubic foot represents
the heating value of 1 cubic foot of moisture-free gas under a pressure
of 30 — 0.52 or 29.48 inches of mercury and a temperature of 60°F.
The number of moles of moisture-free gas in the standard cubic foot is
492 29.48 1
equal to 1.0 X -— X —-— X T— = 0.002597 pound-mole. Con-
520 29.92 o59
versely, r - r r r r r z or 385.5 standard cubic feet of fuel gas contain 1
0.002597
pound-mole of moisture-free gas if the gas behaves ideally.
The heating value of a fuel gas of known composition may be calcu-
lated as the sum of the heats of combustion of its components. The
necessary data may be obtained from Table XV, page 262. The total
heating values of the common combustible gases are also contained in
Table XXII, expressed in Btu per standard cubic foot.
Fuel gases generally contain complex mixtures of both saturated and
unsaturated hydrocarbons. The individual analytical determination of
each component of these mixtures is not feasible for ordinary industrial
purposes. However, Watson and Ceaglske'^ have described a simple
scheme of industrial gas analysis which yields data suitable for ordinary
combustion calculations. In this scheme carbon monoxide and hydro-
gen are separately determined and reported as such. The saturated
paraffin hydrocarbon gases are reported in terms of a hypothetical
compound CnH2n+2, representing the average composition of the mixture
of paraffins in the gas. Similarly, the unsaturated hydrocarbons or
illuminants are reported in terms of a hypothetical compound of average
composition, CJib- For example, the analysis of a gas might be: CO,
40 per cent; H2, 42 per cent; C2.6H4.2 (illuminants), 7 per cent; Ci.2H4.4
(paraffins), 11 per cent.
An analysis of this type may be used as effectively for stoichiometric
calculations as though all components were individually determined.
The heating value of the gas may also be calculated by means of the
following approximate formulas for the total heating values of mixtures
of paraffins and of unsaturated hydrocarbons.
Paraffin hydrocarbons, C„H2n+2:
Calories per gram-mole = 158,100n + 54,700
Btu per cu ft at 60°F, 30 in., sat. = 745w -t- 258 (4)
Unsaturated hydrocarbons, CoH 6:
Calories per gram-mole = 98,200a + 28,2006 + 28,800
Btu per cu ft at 60°F, 30 in., sat. = 459a + 1326 + 135 (5)
" Ind. Eng. Chem., Anal. Ed. (January, 1932).
TABLE XXII
HEATING VALUES AND FLAME TEMPERATURES OP GASES
H.V. = total heating value, Btu per standard cubic foot, measured at 60°F, 30 inches
of Hg and saturated with water vapor (assuming ideal gas behavior)
Maximum Flame Temperatures

with Air at 18°C
Gas Formula H.V. Theoretical Calculated*
(assuming (allowing tor Actual*
*
complete equilibrium
combustion) conditions)
Carbon monoxide CO 315 2440°C

Hydrogen H^ 319 2200
Paraffins:
Methane CH4 994 1980 1918°C 1880°C
Ethane C2H6 1741 2150 1949 . 1895
Propane C3H8 2478 2300 1967 1925
n-Butane C4H10 3212 2080 1973 1900
n-Pentane C6H12 3947 2090 >- ,^'=i-'
Olefins:
Ethylene C2H4 1575 2240 2072 1975
Propylene CsHe 2297 2200 2050 1935
Butylene C4H8 3017 2200 2033 1930
Amylene CsHio 3814 2180
Acetylene C2H2 1544
Aromatics:
Benzene CeHo 3686 2240
Toluene CeHfiCHg 4399 2240
Mesitylene C6H3(CH3)3 5845 2240
Naphthalene CiiHg 5795
* Jones, Lewis, Friauf and Perrott, J . ilm. CAem. Soc, US, 869 (1931).
If the analysis of a gas is carried out carefully, its heating value may
ordinarily be predicted by means of these equations with an error of
less than 2 per cent. Larger errors arise if large quantities of acetylene
are present in the gas.
niustration 2. A city gas has the following composition by volume:
CO2 2.6%
C2 73H4 72 (unsaturateds) 8.4%
O2'. 0.7%
H2 39.9%
CO 32.9%
Ci 14H4 28 (paraffins) 10.1%
N2 5.4%
100.0%
CHAP. IX] I N C O M P L E T E COMBUSTION OF F U E L S 341
(a) Calculate the theoretical number of moles of oxygen which must be suppUed
for the combustion of 1 mole of the gas.
(6) Calculate the heating value of the gas in calories per gram-mole and Btu per
standard cubic foot.
Solution:
(a) Basis: 100 gram-moles of gas.
Oxygen required for:
» Unsaturateds = 8.4(2.73 + 4.72/4) = 32.8
E ,, Hydrogen = 39.9/2 = 19.95
f^ . Carbon monoxide = 32.9/2 = 16.45
Paraffins = 10.1 (1.14 + 4.28/4) = 22.3
Total 9L5
Oxygen to be supplied per mole of gas = 0.915 — 0.007 = 0.908 mole.

(6) Basis: 1.0 gram-mole of gas.
Heating value of:
Hydrogen = 0.399 X 68,320 = 27,250 cal
; Carbon monoxide = 0.329 X 67,410 = 22,180 cal
Unsaturateds =
: • 0.084 [(2.73 X 98,200) -|- (4.72 X 28,200)
• -1-28,800]= 36,100 cal
' Paraffins =
''" 0.101 [(1.14 X 158,100) 4- 64,700] = 23,730 cal
' •• Total 109,260 cal
, , ,. , 109,260 X 1.8
Btu per standard cubic foot = 5^7"^ ~ ^-IÔ-
385.5
Incomplete Combustion of Fuels. The standard heating values of

fuels correspond to conditions of complete combustion of all carbon to
carbon dioxide gas, hydrogen to liquid water, and sulfur to sulfur
dioxide gas. If a fuel is burned in such a manner that complete com-
bustion does not result, the standard heat of reaction may be calculated
by subtracting from its standard heat of combustion the standard heats
of combustion of the combustible products formed.
Illustration 3. A coal having a heating value of 12,180 Btu per lb and containing
6 8 . 1 % total carbon is burned to produce gases having the following composition b y
volume on the moisture-free basis.
CO2 12.4%
CO 1.2%
'« O2 5.4%
N2 81.0%
. ; « j j # . i r : ' - ; . ' " ^ ;'-5^ •-••'•.;;••; 100.0% , ,• ..
Calculate the standard heat of reaction in Btu per pound of coal burned.
Basis: 1.0 lb-mole of flue gas. **"'
C in CO2 = 0.124 lb-atom or 1.49 lb
C in CO = 0.012 lb-atom or 0.14 lb
Total carbon = 1.49 -|- 0.14 = 1.63 lb
Coal burned = 1.63 + 0.681 = 2.39 lb
Heating value of cpal = 2.39 X 12,180 = 29,100 Btu
Heat of combustion of Cb = 0.012(-67,410 X 1.8) = -1460 Btu
Standard heat of reaction = -29,100 — (-1460) =
-27.640 Btu or - 27,640 -r- 2.39 = -11,660 Btu
per lb of coal
MATERIAL AND ENERGY BALANCES

As previously discussed, in establishing an energy balance all sources
of thermal energy are entered on the input side of the balance and all
items of heat utilization and dissipation on the output side. It is
ordinarily desirable to base all thermal quantities on a reference tem-
perature of 18°C, thus permitting direct use of standard thermochemical
data. Other reference temperatures may be used if desired, but in. any
event it is necessary that each complete balance be based on a single
constant-reference temperature.
Where a fuel is used in an industrial reaction two different points of
view are emphasized in establishing an energy balance, depending upon
whether or not the fuel is intended primarily as a source of heat or prin-
cipally as a reducing agent. In the first instance the entire heating
value'of the fuel is listed on the input side of the balance and the entire
heating value of the products resulting from the partial combustion of
the fuel and its reaction with the charge on the output side. In this
instance the utilization of the heating value of the fuel is of principal
interest for heating purposes or in producing a fuel gas which is subse-
quently to be used for heating. In the second instance, where fuel is
used primarily as a reducing agent, as in the reduction of ores, the prin-
cipal interest is in the products of reduction and not in the heating value
of the fuel or of the products of reaction. In this latter instance it is
customary to include on the input side of the energy balance the heat
evolved in the partial combustion of the fuel, which represents the differ-
ence between the heating value of the fuel and the heating value of the
combustible products resulting from the incomplete combustion of that
fuel.
With these two points of view in mind, the input and output items of
an energy balance of a chemical process, based upon a reference temper-
ature of 18°C, are distributed in the following classification:
CHAP. IX] MATERIAL AND ENERGY BALANCES 343
INPUT ITEMS
Group 1. The enthalpy, both sensible and latent, of each material entering the
process.
Group 2. Where fuel is used primarily as a source of heat or in the production
of fuel gases, the total heating value of the fuel.
Group 3. The heat evolved at IS^C in the formation of each final product from
the initial reaetants when such heat effects are exothermic. Where a fuel is used
and its total heating value is included as an input item the products formed from
the fuel by combustion or reaction of the fuel with the charge are not considered
in this group.
Group 4. All energy supplied directly to the process from external sources in the
forms of heat, electrical or radiant energy or work.
OUTPUT ITEMS
Group 1. The enthalpy of each material leaving the process.

Group 2. Where fuel is used primarily as a source of heat or in the production of
fuel gases the total heating value of each product resulting from the partial
combustion of the fuel and its reaction with the charge.
Group 3. The heat absorbed at 18°C in the formation of each final product from
the initial reaetants when such heat effects are endothermic. Where the total
heating value of a fuel is included as an input item the products formed from the
fuel are not considered in this group.
Group 4. AH heat transferred from the process for useful purposes as for the
generation of steam in a boiler furnace.
Group 5. All energy lost from the process in the form of heat, electrical, or radiant
energy or work.
The difference between input group 2 and output group 2 represents
the difference between the heating value oi the fuel and the heating
value of the products of incomplete combustion of that fuel and is con-
sidered as one of the items in input group 3 when the fuel is used pri-
marily as a reducing agent. Separate consideration of the heating values
of the fuel and of its products is desirable to produce an energy balance
of greater economic significance where recovery of heat is a primary
objective.
An energy balance shows how much energy is consumed by necessary
endothermic reactions, how much is transferred to a heat interchanger
or stored in a fluid used for supplying useful heat or power, and how
much heat is wasted owing to incomplete combustion of fuels, to over-
heating of products, and to inadequate thermal insulation.
For example, in a coal-fired boiler furnace an energy balance indicates
the distribution of the chemical energy of the coal into the enthalpy of
steam, the heat lost in the gaseous products due to the presence of com-
bustible gases and to sensible heat, the heat loss due to incomplete com-
bustion of coal as represented by the unburned coke and coal in the refuse,
and the loss of heat by radiation and conduction through the boiler
setting. The justification of further insulation, of increasing the size of

the combustion space, of increasing the draft, and of using automatic
stoking can be answered, at least in part, from the study of such an
energy balance.
Enthalpy of Water Vapor. The enthalpy of superheated water vapor
referred to the liquid at 18°C ia the sum of three separate items:
1. The sensible enthalpy of the water in the liquid state at the satura-
tion temperature of the vapor. This item may be either positive or
negative, depending on whether the saturation temperature is above or
below 18°C.
2. The heat of vaporization of the water at the saturation temperature.
3. The sensible enthalpy of the water vapor referred to the saturation
temperature.
For the sake of consistency the enthalpy of water vapor is determined
in this manner in all energy balances developed in the following pages.
However, where water vapor is highly superheated, as in flue gases, it is
general practice to simplify this calculation by assuming that the
enthalpy of water vapor is equal to the sum of the heat of vaporization at
18°C plus the sensible enthalpy of the vapor referred to 18°C. In effect,
this is assuming a saturation temperature of 18°C for the water vapor.
This assumption introduces negligible errors where the partial pressure
of the water vapor is small, of the order of one atmosphere or less.
THERMAL EFFICIENCY
The thermal efficiency of any process is defined as the percentage of
the total heat input which is effectively utilized in the desired manner.
It is evident that the thermal efficiency of a process may be expressed
in a number of different ways, depending on the method of designating
the total heat input and the effectively utilized heat. In stating a value
of thermal efficiency it is always necessary to specify fuUy the basis upon
which it was calculated.
The total heat input on which the efficiency is based may be taken as
the total of the input items of the energy balance. This would seem to be
the most logical basis for general usage, and unless otherwise specified
thermal efficiencies will be considered as on this basis and termed thermal
efficiencies based on actual heat input. However, many special bases
are in common use to fit the needs of particular processes. In express-
ing the thermal efficiency of a combustion process it is customary to
obtain the total heat input by deducting from the input items of the
energy balance the heat of vaporization, at 18°C, of the water vapor
present in the air used for combustion. On this basis the thermal efiB-
ciency is termed the thermal efficiency based on total heating value.
CHAP. IX] COMBUSTION OF FUELS 345
Because of the fact that in many combustion processes water will not
be condensed from the products of combustion, even if they are cooled
to 18°C, it is sometimes considered that the thermal efficiency of the
apparatus should be based on the net heating Value of the fuel. The
total heat input is then obtained by considering only the net heating
value of the fuel, plus all the sensible heat supplied. A thermal effi-
ciency on this basis is termed the thermal efficiency based on the net
heating value. This method of expression has as its principal advantage
the fact that the percentage efficiencies are higher and appear more
encouraging. However, it is undesirable because combustion apparatus
is available which is capable of recovering some of the latent heat of the
water vapor from the gaseous products. On this basis of expression
such apparatus might have an efficiency above 100 per cent. The
thermal efficiency based on total heating value is a better general cri-
terion for Judging the operation of a combustion process.
Percentage efficiency is also dependent on the quantity of heat which
is designated as effectively utilized. Various interpretations frequently
may be made of this quantity. For example, a furnace and steam boiler
unit used in domestic heating might be considered as effectively utiliz-
ing only the heat represented by the enthalpy of the steam produced.
On the other hand, it might be logical to include as effective heat the
radiation from the furnace itself which is used in heating the room in
which it is situated. The efficiency of the unit might be expressed on
either basis.
A gas producer or water gas generator produces a combustible gas at
a relatively high temperature. If the gas can be utiUzed while hot its
sensible enthalpy as well as its heating value should be included in
the effectively utilized heat of the producer unit. The efficiency of the
unit on this basis is termed the hot thermal efficiency. If the gas must
be cooled before use its sensible heat is not useful and only the heating
value can be classed as heat effectively utilized in the producer. The
efficiency on this basis is termed the cold thermal efficiency.
COMBUSTION OF FUELS
In calculating the material and energy balances of processes involving
the partial or complete combustion or decomposition of fuels, the same
principles are employed whether such fuels are gaseous, liquids, or solid.
The material balance of a simple combustion process includes the weights
of fuel and air supplied and the weights of refuse and gases produced.
This material balance can be calculated fully from a knowledge of
the chemical composition of the four items mentioned without any direct
measurements of the weights except the weight of fuel consumed. The

weights (or volumes) of air and gaseous products are usually not meas-
ured because of the great difficulties involved and because these can
generally be calculated indirectly with greater accuracy than by direct
measurement.
In the burning of coal on a grate as in a boiler furnace the weight and
composition of fuel used and composition of gaseous products are meas-
ured directly. The chemical analysis of the fuel should include the
percentages of carbon, hydrogen, oxygen, moisture, nitrogen, and ash.
It may not be necessary to have a complete ultimate analysis, but in any
event the carbon, moisture, and ash content should be known. A com-
plete analysis of the dry gaseous products is always necessary. The
moisture content in the gaseous products can be calculated, provided
the hydrogen content of the fuel is known, or can be determined by
measuring the dew point of the gas.
The refuse from the furnace may be considered as consisting of ash,
coked carbon, and unchanged combustible matter from the coal. The
composition in terms of these constituents may be estimated from a
determination of ash, fixed carbon, and volatile matter in the refuse,
using the standard scheme of proximate analysis. The weight of refuse
actually formed per unit weight of coal should be calculated on the basis
of the ash contents of refuse and coal, as reported in the proximate
analysis, and not on the basis of the corrected ash reported in the ultimate
analysis.
The air entering the furnace may be assumed to be of average atmos-
pheric composition and its humidity determined by a psychrometric
method. The analysis of the flue gases is ordinarily determined by the
Orsat type of apparatus, yielding the percentages of carbon dioxide,
carbon monoxide, oxygen, and nitrogen in the moisture-free gases. For
more nearly accurate work, determinations of methane and hydrogen
should also be made.
Material and energy balances of combustion processes are based either
upon a unit weight of fuel or upon the weight of fuel used in a given cycle
or unit time of operation. For example, in boiler furnaces operating
continuously the analysis can be based on a period of twenty-four hours
or reduced to a basis of one pound of coal consumed. In operating a
ceramic kiln of the batch type the analysis should be conducted over a
complete cycle of operation including time of preheating and firing and
the final results based on the entire cycle of operation.
When sufficient experimental data are collected the same scheme of
calculations may be employed for all problems in combustion. How-
ever, complete information is seldom available and it becomes necessary
CHAP. IX] COMBUSTION OF COAL IN BOILER FURNACE 347
to devise methods of circumventing these limitations. Extensive in-

formation can often be built up from but few data, and complete material
and energy balances established from a few temperature measurements,
the proximate analysis of coal, and an Orsat analysis of gas.
The various calculations which follow illustrate the modifications in
procedure necessary to make the best use of data available and also to
deal with the special variations in combustion processes represented in
four special cases:
Case 1. Combustion of coal in a boiler furnace where:
0. Complete ultimate analysis of fuel is known.
b. No uncoked coal appears in refuse.
c. Tar and soot are neghgible. ,
d. Sulfur is negligible.
Cose 2. Combustion of coal where:
a. Hydrogen and nitrogen contents are unknown.
b. Uncoked coal drops into refuse.
Case 3. Combustion of coal where sulfur is not negligible.
Case 4. Partial combustion of fuel, as in a gas producer, where:
a. Steam is admitted.
b. Tar and soot are not neghgible.
Case 1. Combustion of Coal in Boiler Furnace. The simplest prob-
lem in combustion calculations exists where complete information is
available or can be directly estimated on the ultimate analysis of coal,
where tar and soot in the gases are negligible, where the sulfur content
of the fuel is negligible, and where no uncoked coal drops into the refuse.
The methods employed in this illustration are general and may be
similarly applied to all problems in the combustion or partial combustion
of a carbonaceous fuel whether solid, liquid, or gaseous.
The material balance of a furnace is represented by the following items:
Input
1. Weight of fuel charged. ;
2. Weight of dry air supplied.
3. Weight of moisture in air supplied.
Output
1. Weight of dry gaseous products.
2. Weight of water vapor in gaseous products.
3. Weight of refuse.
The method of calculating each of these items will be discussed in
detail. General methods for such calculations have already been dis-
cussed in Chapters I and VI.
Illustration 4. From a 12-hour test conducted on a coal-fired steam generating

plant the following data were obtained. jj^ •
Data on Coal Fired
Ultimate analysis
Carbon 65.93%
Available hydrogen 3.50%
Nitrogen 1.30% /
Combined water < 6.31%
Free moisture 4.38% ,>i
Ash 18.58%
Total 100.00%
Total heating value = 11,670 Btu per lb
Total weight of coal fired = 119,000 lb
Average temperature of coal 65°F
Data on Refuse Drawn from Ash Pit . , :•»»,
Ash content = 87.4% •h.'J
Carbon content 12.6%
Average temperature 255°F
Mean specific heat from 65 t o 255°F 0.23
(Estimated from Fig. 39, page 219) , . "
Data on Flue Gas
1. Orsat analysis
Carbon dioxide '... 11.66%
Oxygen 6.52%
Carbon monoxide 0.04%
Nitrogen 81.78%
Total 100.00%
Average temperature 488°F
Data on Air ',
Average dry-bulb temperature 73.0°F
Average wet-bulb temperature 59.4°F
Average barometric pressure = 29.08 in. Hg
Data on Steam Generated ' jr
Average feed water temperature 193°F
Weight of water evaporated 1,038,400 lb j
Average steam pressure = 137.4 lb per sq in. gauge /
Quality of steam 98.3% /
Calculate the material and energy balances for the entire plant. ' / /
MATERIAL BALANCE . /
^' I . • - • • / -
All calculations are based upon 100 lb of coal as fired. :* /
1. Weight of Refuse Formed. '
Where the refuse is not weighed directly its weight can be readily calculated from
its ash content and that of the coal. The following method is correct where mineral
sulfides are not present in the coal. . '
CHAP. IX] MATERIAL BALANCE OP BOILER FURNACE 349
Ash content of coal 18.58 lb

Ash content per pound of refuse 0.8740 lb
Weight of refuse formed = _ = 21.2 lb
2. Weight of Dry Gaseous Products.

A direct measurement of the weight of gaseous products from a com-
bustion process is seldom made because of the many difficulties involved.
Pitot tubes measure inaccurately because of the low velocities encoun-
tered in chimneys and flues. Orifice and Venturi meters are similarly
unreliable because of low-pressure drops encountered and because soot
accumulates in the openings. Electric flow meters read inaccurately
if the composition of the gas varies with respect to carbon dioxide or
water vapor. In any case the direct measurement of gas streams is
made extremely difficult because of variation in temperature and velocity
across each section of the stream. Any accurate measurement must
give a correct integrated value of velocity and temperature over the
entire cross section. Because of these uncertainties and troubles it
becomes easier and more nearly accurate to calculate the weight of
gaseous products from the stoichiometric relationships of combustion.
The complete analysis of the gaseous products includes the percentages
of carbon dioxide, carbon monoxide, oxygen, methane, ethane, hydrogen,
and nitrogen present. Moisture content is not revealed in the usual
gas analysis because the entire analysis is conducted with the gas sample
saturated with water vapor at a constant temperature and pressure.
The general rule is recommended that the weight of dry gaseous prod-
ucts should be calculated from a carbon balance. A carbon balance is
selected as the basis of this calculation for two reasons. In the first
place, carbon is determined with a higher degree of precision in both fuel
and gaseous products than any other elenaent present. Secondly,
carbon is the chief constituent in both fuel and gaseous products so that
a slight error in its determination will not be magnified in subsequent
calculations. To calculate the weight of gaseous products from a
material balance of any other element would invite many additional
sources of error. For example, the hydrogen balance would be out of
the question because of the many sources of hydrogen, its relatively low
percentage content, its several outlets, and its various methods of com-
bination.
Carbon Balance
Carbon gasified. Basis: 100 lb coal fired.
Carbon in coal = 100 X 0.6593 = 65.93 lb or 5.49 lb-atoms
Carbon in refuse = 21.2 X 0.1260 = 2.67 lb or 0.22 lb-atoms
Carbon entering stack gases = 63.26 lb or 5.27 lb-atoms
Carbon in stack gases. Basis: 1.0 lb-mole of gas. •

Carbon in CO2 = j,>' 0.1166 lb-atom
Carbon in CO = 0.0004 lb-atom
Total carbon = 0.1170 lb-atom
Moles of dry stack gas per 100 lb coal fired
= 5.27/0.1170 = 45.1 lb-moles
Total dry gaseous products. Hasis: 100 lb coal fired.
CO2 = 45.1 X 0.1166 = 5.26 lb-moles or X 44 = 231.5 lb IH
CO = 45.1 X 0.0004 = 0.018 lb-moles or X 28 = 0.660 lb
O2 = 45.1 X 0.0652 = 2.94 lb-moles or X 32 = . . . 94.2 lb
N2 = 45.1 X 0.8178 = 36.90 lb-moles or X 28.2 = . .. 1041 lb
Total 45.1 lb-moles 1367 lb ;''
Average molecular weight = 1367/45.1 = 30.3
3. Weight of Dry Air Supplied.
Direct measurement of the weight or volume of air used in combus-
tion is accompanied by the same difficulties as the direct measurement
of gaseous products. Furthermore, air is usually drawn through the
grate by chimney draft so that there is no need for confining the supply
of air in ducts and there is no opportunity for direct measurement of
its flow.
The dry air used in combustion consists of oxygen and inert gases,
chiefly nitrogen. These inert gases also include argon and traces of
rare gases, but because of the small amount present it is customary to
include all the inert gases as nitrogen and assign a molecular weight of
28.2 to atmospheric " nitrogen." ^^ This nitrogen passes through the
furnace unchanged and appears entirely in the gaseous products. Any
nitrogen present in the fuel burned will also appear in the flue gases.
The nitrogen in ordinary solid and liquid fuels burned will usually be
negligible or very small. , However, in the combustion of gases a con-
siderable portion of the nitrogen appearing in the flue gases may come
from the gaseous fuel.
The composition of dry air may ordinarily be taken as constant, con-
taining 21.0 per cent oxygen and 79.0 per cent nitrogen by volume, the
nitrogen content including the argon present. Under certain conditions
the air used in combustion may contain appreciable amounts of car-
bon dioxide, making a separate gas analysis of the air valuable. The
moisture content of air is subject to extreme variations depending upon
weather conditions so that a separate determination of the moisture
content of air is invariably necessary.
Because of the constancy of composition of dry air it is possible to cal-
12 The molecular weight of atmospheric " nitrogen " is taken as 28.2 because of the
argon associated with it. i . v - • i !
CHAP. IX] MATERIAL BALANCE OF BOILER FURNACE 351
culate readily the weight of air used in a combustion process from a knowl-
edge of the nitrogen content of the gaseous products and of the fuel used.
Accordingly, the general rule is expressed that the weight of dry air
actually used in combustion process is calculated from a nitrogen balance.
The chief objection to the use of the nitrogen balance basis is that in
gas analysis, errors resulting from unabsorbed components accumulate
on the nitrogen determination which is always found by difference.
Nitrogen Balance. Basis: 100 lb coal fired.
Nitrogen in gaseous products = 36.90 lb-moles
Nitrogen from coal = 1.30/28.0 = 0.0464 lb-mole
Nitrogen from air = 36.85 lb-moles
Dry air supplied = 36.85/0.79 = 46.6 lb-moles
or 46.6 X 29 = 1354 lb
It will be noted that the nitrogen content of the coal might be neglected without in-
troducing a serious error.
4. Weight of Moisture in Air.

The weight of moisture per mole of dry air depends upon the tempera-
ture, pressure, and relative humidity of the air. From the dew point
the partial pressure of the water vapor is determined, and the moisture
content of the air may be calculated by the methods explained in Chap-
ter IV.
Dry-bulb temperature = 73.0°F
Wet-bulb temperature = 69.4°F
From Fig. 9, the molal humidity of the air is 0.012
Water suppUed with air = 46.6 X 0.012 = 0.559 lb-mole
5. Total Volume of Wet Air Introduced. " •-
Basis: 100 lb coal fired.
Total moles of moist air = 46.6 + 0.559 = 47.2 lb-moles
Volume at 73°F, 29.08 in. Hg =
29 92 533
47.2X359X— X - = 18,870 cu ft
6. Weight of Moisture in Gaseous Products.

To complete the material balance it is necessary to know the weight
of moisture in the gaseous products since this is not obtained by the
ordinary gas analysis. Direct measurement of the moisture content is
difficult. It can also be calculated if the composition of the dry flue
gases, the moisture content of the air used, and the hydrogen and mois-
ture in the fuel burned are known. As a general rule it may be stated
that the moisture consent of the gaseous -products is calculated from a
hydrogen balance. -.... ...
352 FUELS AND COMBUSTION [CHAP. IX -
Hydrogen Balance. Basis: 100 lb coal fired.

Input
From moisture introduced with dry air = 0.559 lb-mole
From combined water in coal = 6.31/18 = 0.351 lb-mole
From free moisture in coal = 4.38/18 = 0.244 lb-mole
From available hydrogen in coal = 3.50/2.016 = .. . 1.738 lb-moles
Total '. 2!i92 lb-moles
Output
The Orsat apparatus used for analyzing the stack gases does not
determine hydrogen or hydrocarbon gases. However, in an efficiently
operated boiler furnace, hydrogen and hydrocarbons are present only in
small quantities. It is therefore assumed that all the hydrogen intro- '
duced into the system leaves as water in the stack gases.
Hydrogen in H2O of stack gases = 2.892 lb-moles
The dew point of the stack gases may now be determined. , '
"2 QC)2
Partial pressure of H2O = ;; ' „^^ X 29.08 = 1.75 in. Hg
45.1 -|- 2.892 •
From Table I, this partial pressure is seen to correspond to a dew point of 36°C or
97°F.
7. Total Volume of Gaseous Products.

Basis: 100 lb coal fired. •' •
Moles of wet gas = 45.1 + 2.892 = 48.0 lb-moles
' Volume at 488°F and 29.08 in. Hg =
,! •• .,1: 29.92 948
48.0X359X^^X-= 34,150 cu ft
Summary of Material Balances. To verify the accuracy of experimental data or
methods of calculation a summary of all material balances is prepared. The
overall material balance is also indicated on the flow chart of Fig. 70.
OVERALL MATERIAL BALANCE /,

Input Output
Coal 1001b 21.2 lb /
Dry air (46.6 lb-moles)... 1354 lb Dry gases (45.1 lb-moles) 1367 lb
HjO in air (0.559 lb-mole) 10.05 lb H2O in stack gases
(2.892 lb-moles) 52.1 lb
Total 14641b Total 14401b
CARBON BALANCE
In coal 65.931b In gases 63,261b

In refuse 2.67 lb
Total 65.931b 65.93 lb
CHAP. IX] SUMMARY OF M A T E R I A L BALANCES 353
NITROGEN BALANCE
I n air (36.85 lb-moles) . . . 1040 lb I n stack gases

In coal 1.30 lb (36.90 lb-moles). 10411b
Total 1041 lb 1 i* Total 1041 lb
Dry Flue Gases 1367 lb

Coal 100 lb (45.1 Pound Moles)
HjO Vapor 52.1 lb
'
: [f
1
-
1 . .' • 1 u
. • » • ' '
\
Dry Air 1354 lb JsTCSfil'

(46.6 Pound Moles) Refuse 21.2 lb
1 >
HjO Vapor in Air 10 lb
(0.56 Pound Mole)
F I G . 70. Material balance of steam generating plant (Illustration 4).

\Ai
HYDROGEN BALANCE
In water vapor of air ,,; In H2O of stack gases

(0.559 lb-mole) 1.127 lb (2.892 lb-moles) 5.83 lb
In combined water of coal
(0.351 lb-mole) 0.708 lb • " - . • , . . ... .
In free moisture of coal , • •
(0.244 lb-mole) 0.491 lb

;; '-vi. i.:!(,- . , 1 ••:)!
Available hydrogen of coal
(1.738 lb-moles) 3.50 lb
Total 5.83 lb Total .. 5.83 lb
ASH BALANCE
18.58 lb .. 18.58 lb
OXYGEI <r BALANCE
In combined water of coal In CO 2 of stack gases

6 31 X 16/18 5.61 lb 5.26 X 32 168.5 lb
In free moisture of coal In O2 of stack gases
4 38 X 16/18 3.89 lb 2.94 X 32 94.2 1b
In dry air In CO of stack gases
36.85 X 21/79 X 32. . . 313.5 lb 0.01804/2 X32 0.289 lb
In water vapor in air In H2O of stack gases
0 559/2 X 32 8.95 lb 2.892/2 X32 46.3 lb
Total 3321b Total 3091b
It will be seen that there is a deficit of 24 lb or 1.6 per cent on the out-
put side of the overall material balance. This discrepancy falls entirely
on the oxygen balance since no direct calculations were made from this
basis. The oxygen content of the fuel is obtained by difference, so that
all errors in any other determination of the coal analysis accumulate
algebraically upon this value. In this particular case the deficit of 24
pounds in the overall material balance indicates that the content of the
carbon in the coal, or the content of the carbon monoxide or carbon
dioxide in the gas, is low.
Other errors or omissions in chemical analysis will be reflected in the
material balance. For example, the oxygen content of the air supply
might be less than 21.0 %, and some carbon dioxide might enter with
the air and would escape consideration unless a special analysis of the
air were made. The presence of any suspended tar or soot in the
gaseous products might introduce a serious error in the overall material
balance.
8. Theoretical Amount of Air Required for Combustion. \i
The weight of air theoretically required for complete combustion de-
pends upon the chemical composition of the fuel and the stoichiometric
relations involved in combustion. Since the one element in common for
all combustion reactions is oxygen, the weight of air required for com-
bustion must be calculated from an oxygen balance. The oxygen already
in the fuel is assumed to be in combination with hydrogen. Hence
only the available hydrogen of the fuel is considered in calculating its
oxygen requirement. , ,!.,..
Per Cent Excess Air « ' ,
OXYGEN BALANCE ' j
Oxygen requirements for combustible constituents of coal charged.
Oxygen Required
Carbon 65.93 lb = 5.49 lb-atoms 5.49 lb-moles
Net hydrogen 3.50 lb = 1.736 lb-moles 0.868 lb-mole
Total 6.358 lb-moles
Air required = 6.358/0.21 = 30.3 lb-moles
Air supplied = 46.6 lb-moles
Excess air = 46.6 - 30.3 = 16.3 lb-moles
Per cent excess air = 16.3/30.3 X 100 = 53.9% /
ENERGY BALANCE
Since it is conventional to utilize total rather than net heating values in energy
balances the reference state for all water involved in the process is the liquid state;
the enthalpies of reactants and products are evaluated on this basis.
CHAP. IX] ENERGY BALANCE OF BOILER FURNACE 355
Reference temperature: 65°F.

Input
1. Heating value of the coal = 100 X 11,670 = 1,167,000 Btu
2. Enthalpy of the coal = 0 Btu
3. Enthalpy of the dry air =
46.6 X 6.96 X (73 - 65) = 2,590 Btu
4. Enthalpy of the water vapor accompanying the dry air.
Dew point of the air from Fig. 9 = 49°F
Heat of vaporization = 19,140 Btu per lb-mole at 49°F
Enthalpy = 0.559[8.00(73 - 4 9 ) + 19,140 - 18(65 - 49)] =^ 10,570 Btu
Total heat input 1,180,160 Btu
" ' Oviput

1. Heating value of refuse.
Weight of carbon in refuse = 2.67 lb
Heating value = 2.67 X 14,550 = 38,850 Btu
2. Heating value of stack gases.
Lb-moles of CO in stack gases = 0.01804 lb-mole
Heating value = 0.01804 X 67,410 X 1.8 = 2,190 Btu
3. Enthalpy of refuse.
Weight of refuse = 21.2 lb . • ;:
Enthalpy = 21.2 X 0.23 X (255-66) = 1,090 Btu
4. Enthalpy of dry stack gases.
Mean heat capacities between 65 and 4 8 8 ^ from Fig. 37.
Enthalpies
COj = • 5.-26 (9.94) (488 - 65) = 22,120 Btu
CO = 0.018 (7.03) (488 - 65) = 50 Btu
O2 = 2.94 (7.22) (488 - 65) = 8,990 Btu
;,>,;,. Nj =36.90 (7.02) (488 - 65) = 109,600 Btu
Total = 140,660 Btu
6. Enthalpy of the water vapor in stack gases.
Dew point of stack gases = 97°F
Heat of vaporization (Fig. 8) =
18,660 Btu per lb-mole at 97°F
Enthalpy of the Uquid = 2.892 X 18 X (97 - 65) = 1,665 Btu
Heat of vaporization = 2.892 X 18,660 = 53,960 Btu
Superheat = 2.892 X 8.21 X (488 - 97) = 9^290 Btu
Total Enthalpy 64,920 Btu
6. Heat utilized in generating steam. .' ;
Pounds of steam generated = ' X 100 = 873 lb

Enthalpy of 1 lb steam as pro- .
duced (151.7 lb per sq in. abso- ., -
lute, 1.7% moisture), relative to ?''••:•:: '" ;',o" ;--'.;!,?JI
32°F =
331.4 + 862.3 - (0.017 X 862.3) = 1179.0 Btu per lb
Enthalpy of feed water at 193°F relative to 32°F = 160.9 Btu per lb
Net heat input into steam produced = 1018.1 Btu per lb
Total heat absorbed by steam produced =
873 X 1018.1 = 890,000 Btu
7. Undetermined losses (by difference) = 42,450 Btu
Siunmarized Energy Balance of Steam Generating Plant:
Reference temperature: 65°F . JI • : , , /. .
Energy Input
.•iiXAmîO Btu Percent
1. Heating value of the coal 1,167,000 98.9
2. Enthalpy of the coal 0 0
3. Enthalpy of the dry air 2,590 0.2
4. Enthalpy of the water vapor accompanying the dry
air 10,570 0.9
,',ao,^ Total... 1,180,160 100.0
Energy Output
1. Heating value of refuse 38,860 3.3
2. Heating value of stack gases 2,190 0.2
3. Enthalpy of refuse 1,090 0.1
4. Enthalpy of the dry stack gases 140,660 11.9
5. Enthalpy of the water vapor in the stack gases.... 69,920 5.5
6. Heat utilized in generating steam 890,000 75.4
7. Undetermined losses (by difference) 42,450 3.6
Total 1,180,160 100.0
This energy balance is summarized in Fig. 71. - / ,
Thermal Efficiency and Economy. The thermal efficiency of a boiler
furnace may be calculated on the total or the net heating value of the
coal. The effectively utilized heat is that which is absorbed in steam
generation.
Based on total heating value of coal,
. 890,0
the thermal efBoienoy is , ,„_' or , ,. 76.3%
1,167,1
Based on net heating value of coal,
890,000
the thermal efficiency is iîgy^QOO - (42) (1060) "' ^^"^^^
CHAP. IX] COMBUSTION CALCULATIONS 367
Case 2. Combustion Calculations Where Ultimate Analysis of Coal
Is Not Completely Known. In the preceding illustration, the calcula-
tions were completed without making any assumptions as to the com-
position of the fuel. However, the determination of the hydrogen
content of coal is a difficult procedure, to be avoided if possible. For
Heatine Value
of Coal 98.9%
Enthalpy of 'a tif .S
Dty Air 0.2% " i - . -" •! ,,]• .
1 ^< :'i ill y

Enthalpy of
Water Vapor
in Air 0.9%
Heating Value
of Refuse 3.3%
Heating Value of 0.2%

i CO in Stack Gas
Enthalpy of ai%
Refuse
^
, Enthalpy of
Dry Stack Gas 11.9%
1
Enthalpy of
3 ater Vapor iin
Water 5.5%
StackI Gas
Gas
V Undetermined
Losses 3.6%
^
Heat Used in
Generating 75.4%
Steam
100.0%
FiQ. 7 1 . E n e r g y balance of s t e a m generating p l a n t (Illustration 4 ) .
this reason it is sometimes desirable to evaluate the material balance of

a furnace or gas producer without data on hydrogen content and to
calculate this quantity from an oxygen balance. In this illustration an
additional complication is introduced in that some of the coal drops
through the grate without coking.
Where the hydrogen content of the fuel is calculated from an oxygen
balance great care must be taken in analyzing the flue gases. The
sampling and determination of oxygen in hot flue gases are particularly
uncertain. For this reason it is frequently preferable to estimate the
hydrogen content of the fuel by empirical methods such as those illus-

trated on page 328 for coal, page 334 for gases, and Fig. 65, page 333
for oils. The calciUations are then carried out as in the preceding
illustration.
In the preceding illustration the nitrogen content of the coal was
known. In making a complete ultimate analysis it is necessary to de-
termine this element in ordei: that the oxygen content may be obtained
by difference. However, from the results of the previous illustration it
is apparent that the nitrogen in the coal might be neglected altogether,
assuming all nitrogen in the flue gases to have come from the air. No
appreciable error will result from neglecting this nitrogen except in
determining the oxygen or combined water content of the coal by differ-
ence. For this calculation it is ordinarily sufficient to assume a nitrogen
content of 1.7 per cent of the combustible in the coal. This assumption
will ordinarily not be in error by more than 0.3 per cent of the weight of
the combustible except in the case of anthracite coals." Greater refine-
ment is not justified because of the uncertainty of the sampling of the
coal and of other data on which the material balance is based.
Illustration 5. Coal-Fired Boiler Furnace. A furnace is fired with a bituminous
material coal having the following proximate analysis:
Moisture 2.9% . -
Volatile matter 33.8%
Fixed carbon 53.1%
Ash 10.2%
,^ <. _ _ * ' 100.0%
The ultimate analysis is known only in part and includes (as-received basis):
Sulfur 1.1%
Carbon 73.8%
The dry refuse from the furnace has the following composition:
Volatile matter 3.1% __ "
Fixed carbon 18.0%
' •>' Ash 78.9%
*'• 100.0%
The Orsat analysis of the flue gases is as follows: /
Carbon dioxide 12.1%
Carbon monoxide 0.2%
Oxygen 7.2%
Nitrogen 80.5%
100.0%
Air enters the furnace at a temperature of 65 °F with a percentage humidity of
55%. The barometric pressure is 29.30 in. of Hg. The flue gases enter the stack at
a pressure equivalent to 1.5 in. of water less than the barometric pressure and at
a temperature of 560°F.
Water is fed to the boiler at a temperature of 60°F and vaporized to form wet
CHAP. IX] MATERIAL BALANCE OF FURNACE 369
steam at a gauge pressure of 100 lb per sq in., quality 98%, at a rate of 790 lb of
steam or water per 100 lb of coal charged.
Compute complete material and energy balances, the volumes of air and flue gases
per 100 lb of coal charged, and the percentage excess air used.
MATERIAL BALANCE
The calculations are similar to Illustration 4 with special methods in parts 1,

4,5.
1. Total carbon content of refuse.
Basis: 100 lb of coal charged.
The weight of refuse is calculated horn the ash contents o{ the coal and oi the
refuse. The weight of ash in 100 lb of coal is 10.2 lb. This weight of ash constitutes
but 78.9% of the weight of refuse.
10.2
Total weight of refuse = r-zr;: = 12.9 lb
0.789
Carbon exists in the refuse as fixed carbon and as volatile matter. The volatile mat-
ter is due to the dropping of uncoked coal through the grate. The combustible of
the uncoked coal in the refuse may be assumed to have the same composition as the
combustible of the coal fired. Therefore, the ratio of combustible to volatile matter
in the uncoked coal in the refuse will be the same as that in the coal.
Ratio of combustible matter to volatile matter in the coal
_ 33.8 + 53.1 _
" 33.8 " 2-^^
Volatile matter in refuse = 12.9 X 0.031 = 0.40 lb
Unchanged combustible in refuse = 0.40 X 2.56 = 1.02 lb
Carbon is also present in the refuse as coked coal accompanied by no volatile matter.
The amount of carbon as coke is the difference between quantities of total combus-
tible and of unchanged coal combustible in the refuse.
Fixed carbon in refuse = 12.9 X 0.18 = 2.32 lb
Total combustible in refuse = 2.32 -\- 0.40 = 2.72 lb
Coked combustible (carbon) = 2.72 - 1.02 = 1.70 lb
The total carbon content of the unchanged combustible in the refuse may be deter-
mined hoia the percentage of carbon in the combustible of the original coa\.
73 8
Total carbon content of combustible of coal = '• = ,.. 85%
33.8 + 53.1 ^^
Carbon in uncoked coal in refuse = (0.85) (1.02) = 0.87 lb
Total carbon in refuse = 1.70 + 0.87 = 2.57 lb
2. Weight of dry flue gases.
Carbon Balance. Basis: 100 lb of coal charged.
Carbon gasified = 73.8 - 2.57 = 71.2 lb or 5.94 lb-atom
1.0 lb-mole of dry flue gases contains:
Carbon dioxide 0.121 lb-mole
Carbon monoxide 0.002 lb-mole
Total carbon per pound-mole of gas = 0.123 lb-atom
5.94
Total dry flue gas = —— = 48.3 lb-moles
Total dry gases:

CO2 = 48.3 X 0.121 = 5.85 lb-moles or. X 44 = . . . . 2581b
CO =48.3X0.002= 0.096 lb-moles or X28 = . . . . 3 1b
O2 =48.3X0.072= 3.48 lb-moles or X32 = . . . . 1111b
Nj = 48.3 X 0.805 = 38.87 lb-moles or X28.2 = . . 1097 lb
Total = 48.30 lb-moles or 14691b
3. Weight of air supplied. , . .|,,.- i

Nitrogen Balance. Basis: 100 lb of coal charged.
N2 in flue gases = 38.87 lb-moles
Assuming all N2 to come from the air: • :
38 87
Dry air supplied = —'-— = 49.2 lb-moles
0.79
or 4 9 . 2 X 2 9 = 14301b -^
From Fig. 9: ,:
Molal humidity of air = 0.012 , s':
Water vapor in air = 0.012 X 49.2 = 0.59 lb-mole
or 0.59 X 18 = 10.6 lb : :;:
Total wet air = 49.2 + 0.59 = 49.8 lb-moles ,,
or 1430 -1-10.6 = 14401b
Volume of air entering at 65°F, 29.3 in. of Hg =
525 29.92
49.8 X 359 X — X ^ ^ = 19,450 cu ft
4, Hydrogen content of coal. •
Oxygen Balance. Basis: 100 lb of coal charged. ,,
Oxygen in dry flue gas = 5.85 -|- -^-— + 3.48 = 9.38 lb-moles
Oxygen entering in dry air = 49.2 X 0.21 = ... . 10.32 lb-moles
Assuming that the oxygen not accounted for in the dry flue gases was consumed in
oxidizing the available hydrogen of the coal: - ; rutjjs.ifu
O2 oxidizing H2 = 10.32 - 9.38 = 0.94 lb-mole
H2 burned = 2 X 0.94 = 1.88 lb-moles or 3.79 lb
The hydrogen burned may be taken as the available hydrogen of the coal, neglecting
the small hydrogen content of the uncoked combustible in the refuse.
6. Complete ultimate analysis of coal. The unknown items of the tiltimate analy-
sis are combined water and nitrogen. As pointed out on page 358, the nitrogen
content may be assumed to be 1.7 X 0.87 = 1.4%. The combined water may then
be determined as the difference between 100 and the sum of the percentages of mois-
ture, carbon, hydrogen, sulphur, nitrogen, and corrected ash.
Corrected ash = 10.2 - 3/8(1.1) = 9.8%
Combined H2O = 100 - (2.9 -|- 73.8 + 3.8 + 1.1 + 1-4
+ 9.8) = 7.2%
CHAP. IX] MATERIAL BALANCE OF FLTINACE 361
Ultimate analysis:
Moisture 2.9%
Carbon 73.8%
Available H 3.8%
Sulfur 1.1%
Nitrogen 1.4%
Corrected ash 9.8%
Combined H2O 7.2%
100.0%
6. Water vapor in flue gases.
Hydrogen Balance. Basis: 100 lb of coal charged.
H2O from air = 0.59 lb-mole
2.9 + 7.2
II2O from coal = — = 0.56 lb-mole
18
H2O formed from H = 1.88 lb-moles
Total = 3.03 lb-moles
or 55 lb
7. Volume of wet flue gases.
Moles of wet flue gas = 48.3 -1- 3.03 = 51.33 lb-moles
15
Pressure in flue = 29.30 - ~— = 29.19 in. of Hg
13.6
Volume at 560°F, 29.19 in. of Hg =
1020 29 92
51.33 X 359 X ^—- X -~-- = 39,200 cu ft
8. Complete material balance.

Inpvi
Coal 100 lb
Dry air (49.2 lb-moles) 1430 lb
Water vapor in air (0.69 lb-mole) 10 lb
.' Total 1540 lb
Output
Dry flue gases (48.3 lb-moles) 1469 lb
Refuse 13 lb
Water vapor in flue gas (3.03 lb-moles) 55 lb
Total 1537 lb
This material balance is summarized in Fig. 72.
9. Percentage excess air.
Total carbon in coal charged = 73.8 lb = 6.15 lb-moles
Available hydrogen in coal charged = 1.88 lb-moles
1 88
Total O2 required = - ^ + 6.15 = 7.09 lb-moles
7.09
Air theoretically required = - ^ = 33.8 lb-moles
Air actually suppUed = 49.2 lb-moles
Percentage excess air = —' = 45.5, based on that required for complete
33.8
combustion of all carbon and available hydrogen in the coal charged and neglecting
that required for sulfur.
The percentage excess air may also be calculated directly from the flue-gas analysis:
Basis: 100 lb-moles of flue gas.
Neglecting Nj from the coal:
Oj introduced in air = — - X 0.21 = 21.4 lb-moles

0.79
O2 in excess = (surplus present in gases minus that
required to complete oxidation of the CO) =
»„ 0-2
7.2--^=. 7.1 lb-moles
Os actually consumed = 14.3 lb-moles
7.1
Percentage excess oxygen = percentage excess air = —— = 49.6.
14.0
3
This percentage excess is based on the oxygen required for all combustible sub-
stances which were actually burned.
Feed Water Steam

BoUer
7901b 7901b ^
Dry Stack Gas

14691b
Ck>al (48.3 lb-moles)
1001b BoUcr HjO in Stack Gas
Furnace 551b
(3.03 lb-moles)
. . ; • :
,, :
Dry Air 1430 lb

(49.2 lb-moles) - L Refiise
HjO in Air 10.0 lb 13jQlb
(0.59 lb-mole) /
FIG. 72. Material balance of a boiler furnace (Illustration 5).
E N E R G Y BALANCE /
The energy balance of this problem is calculated exactly as in Illustration 4, esti-

mating the heating value of the coal by the empirical method discussed on page 328.
//
Effect of Sulfur in Coal. In the prece(iing illustrations the com-
bustion of the suUur in the coal has been neglected. This does not
introduce appreciable error if the sulfur content is low, 1 per cent or
less. It is difficult to take into account the combustion of the sulfur
by any rigorous method because of the uncertainty of the forms in which
it is present in the coal and the difficulty of determining its distribution
in the combustion products. A considerable part of the sulfur which is
CHAP. IX] EFFECT OF SULFUR IN COAL 363
present in the coal in a combustible or available form will appear as sul-

fur dioxide in the flue gases. The remainder will be present in the refuse.
The ordinary scheme of flue gas analysis by the Orsat apparatus, in
which the gas sample is confined over water, does not permit determina-
tion of sulfur dioxide. Because of the high solubihty of sulfur dioxide
in water (about thirty times that of carbon dioxide) the bulk of the SO2
will be absorbed in the water of the sampling apparatus and burette.
Any SO2 which is not removed in this manner will be absorbed and
reported as CO2. The reported analysis will ordinarily represent ap-
proximately the composition of the S02-free gases. Methods are avail-
able by-which separate determination may be made of the SO2, but this
is not ordinarily feasible for performance tests of combustion equipment.
Neglect of the combustion of sulfur in calculations of the type car-
ried out in Illustration 4, in which the oxygen balance is not used,
does not introduce any serious error, even if the sulfur content is relar
tively high. The calculated total weight of dry flue gases will be low
by approximately the weight of SO2 formed. This error is usually neg-
ligible. However, in the method of calculation of Illustration 5, in
which the net hydrogen in the coal is calculated from an oxygen balance,
the errors resulting from neglect of combustion of sulfur may be more
serious. In this type of calculation it is assumed that all free oxygen
not accounted for as CO2, CO, or free oxygen in the flue gas was utiUzed
in the oxidation of the net hydrogen of the coal. This assumption
neglects the oxygen consumed in oxidation of sulfur. As a result the
net hydrogen content calculated by this method will be too high, and
the combined water content, calculated by difference, will be too low.
These errors are particularly high when the sulfur is originally present
as iron pyrites, FeS2, in which case oxygen is consumed in the oxidation
of both the sulfur and the iron. However, because of the relatively
high atomic weight of sulfur it is permissible to neglect these errors
in dealing with coals containing less than 1 per cent of sulfur.
The probable errors involved in neglecting the combustion of the
suHur may be estimated for a typical case by consideration of the data
of Illustration 5. The coal contained 1.1 per cent sulfur. The max-
imum error would result if all the sulfur were in the form of iron pyrites.
In this case, for each pound-atom of suKur present 11/8 pound-moles
of oxygen would be consumed in producing SO2 and Fe203. On the
basis of 100 pounds of coal charged 0.048 pound-mole of oxygen would
be required for combustion of the sulfur of the coal. Introducing this
value into the oxygen balance on page 360 will reduce the unaccounted-
for oxygen, used in oxidizing hydrogen, from 0.94 to 0.89 pound-mole.
The calculated net hydrogen is correspondingly reduced from 3.8 to
3.6 pounds and the combined water content is increased from 7.2 to
7.4 pounds. These errors are not serious and in addition represent
maximum errors because actually not all the gulfur will be in the pjTitic
form and furthermore it will not be completely burned.
Case 3. Method of Calculation Where Sulfur Is High and the
Carbon and Hydrogen Contents of the Coal Are Unknown. Where
sulfur contents of more than about 1 per cent are encountered it is not
ordinarily desirable to use the oxygen balance for computing the net
hydrogen content of the coal. In order to develop an accurate oxygen
balance it would be necessary to have data on the forms in which the
sulfur occurred in the coal and on the sulfur content of the refuse.
Without such data the net hydrogen content is better determined
analytically or estimated by the empirical method of Uehling, discussed
on page 328. This latter method may also be used to estimate the total
carbon in the coal from a determination of the heating value. ' ^
If the sulfur content of a coal is so high that sulfur dioxide con-
stitutes a considerable part of the flue gas it is necessary to obtain data
from which the amount of sulfur actually burned may be calculated.
A determination of either the sulfur in the refuse or of the SO2 in the
flue gases will supply this information. The former determination is
more easily carried out. It may then be assumed that the ordinary flue
gas analysis yields the composition of the S02-free gases and the total
quantity of gases computed on this basis. Direct determination of the
S02 content of the flue gases is more reliable but frequently unwarranted.
In calculating an energy balance involving a coal of high sulfur content
a sulfur correction should be applied to the heating value directly
determined in the oxygen bomb calorimeter. This correction results
from the fact that in the calorimeter the available sulfur is burned
almost entirely to SO3 whereas in ordinary combustion the major part
of it will form SO2. The correction may be taken as 1000 calories per
gram of sulfur, to be subtracted from the observed heating value.
Case 4. Gas Producers. In the operation of a gas producer, a fuel
gas of low calorific value is produced by blowing air, usually accompanied
by steam, through a deep incandescent bed of fuel. Carbon monoxide
and carbon dioxide are formed by partial combustion of the fuel. Hy-
drogen and the oxides of carbon are formed from the reduction of water,
and volatile combustible matter is distilled from the coal without com-
bustion.
Effect of Soot and Tar. In many combustion processes the gases
contain carbon suspended in the form of soot and tar. These forms of
carbon do not appear in an ordinary volumetric gas analysis but must
be determined separately by absorption or retention on a weighed filter.
The tar can then be separated and analyzed for its hydrogen content
CHAP. IX] GAS PRODUCERS 365
although this precision is not usually warranted. It is ordinarily suffi-

cient to assume that the combustible of the suspended tar analyzes 90
per cent carbon and 10 per cent hydrogen and that the combustible of
the soot consists of 100 per cent carbon. Frequently refuse also appears
suspended in the gaseous products so that the ash content of the sus-
pended material should also be measured.
In addition to the tar and soot suspended in the gases a gradual accu-
mulation of these products settles in the flues. The slight correction
necessary for this is usually made by measuring the deposition over a
long period of time, for example, over the usual time interval elapsed
between successive cleanings of the flues. The tar deposit is removed
and weighed and, since the tons of fuel distilled during the time inter-
val is known, a rough approximation can be obtained for the carbon
deposited per unit weight of coal charged.
The following illustration of gas producer operation is of value in
that consideration is given to the tar and soot suspended in the producer
gas and also to the Uve steam which is passed into the producer. The
problem is of added interest in that all experimental data were collected
with unusual care. Analyses of gas and coal and measurements of
temperatures were made at regular intervals over several days of opera-
tion and data carefully weighted to give average results. These data
were taken in part from a report by C. B. Harrop."
Dlustration 6. Air is supplied to a gas producer at 75°F with a percentage hu-
midity of 75% and a barometric pressure of 29.75 in. of Hg. Coal weighing 70,900
lb and having a gross heating value of 11,910 Btu per lb is charged into the pro-
ducer. Tar weighing 591 lb is deposited in the flues prior to the point of gas sam-
pling and contains 93% carbon and 7% hydrogen.
j The water vapor, suspended tar, and soot in the gases were determined experimen-
tally by withdrawing samples of the gas through an absorption train. The results
are expressed in grains per cubic toot of wet, hot gas measured at 1075°F and 29.75
in. of Hg.
Water vapor 3.43 grains per cu ft
Suspended tar 3.31 grains per cu ft
Suspended soot 1.52 grains per cu ft
It is assumed that the suspended tar is 90% carbon and 10% hydfogen and that
the soot is 100% carbon.
Saturated steam at a gauge pressure of 25.3 lb per sq in. is introduced at the bot-
tom of the fuel bed.
Analysis of coal as charged:
Carbon 66.31% Nitrogen... 1.52%, Sulfur 1.44%
Available hydrogen 3.53% Total water 23.16% Corrected ash 4.04%
Total 100.00%
Ash (as weighed) 4.58%
" J . Am. Ceramic Soc, 1, 35 (1918). : , ^ ^ .-.•.n;'~
Analysis of refuse:
Moisture 1.10% Fixed carbon 3.08% Ash 95.82%
Total 100.00%
Analysis of dry, tar- and soot-free gas, by volume:
CO2 7.12% CO 21.85% H2 13.65%
O2 0.90% CH4 3.25% N2 53.24%
Total 100.00%
The gases leave the flue at 1075°F and 29.75 in. of Hg. The refuse leaves the pro-
ducer at 350°F and rnay be assumed to have a specific heat of 0.22. The mean spe-
cific heat of the tar and soot between 75°F and 1075°F may be assumed to be 0.32.
The heating value of the suspended tar was found to be 16,000 Btu per lb, and that
of the deposited tar 15,500 Btu per lb.
Calculate complete material and energy balances for the operation of this gas
producer.
Solution: Since a majority of the details of this problem are similar to those of
Illustration 4, full explanations of all steps will not be repeated. A flow chart of
this process is shown in Fig. 73.
Coal 100 lb.

Dry Gas
(14.18 Pound Moles)
• ^ '
HjO Vapor
(0.45 Pound Mole) 8.1 lb.
Tar Deposited 0.811).
Tar Suspended 7.81b.
Soot Suspended 3.6 !b.
Dry Air (9.48 Pound Moles) 2751b.

HjO Vapor in Air
(0.22 Pound Mole) 3.9 lb. r Refuse 4.81b.
Steam (048 Pound Mole) 8.61b.
FIG. 73. Material balance of a gas producer (Illustration 6).
MATERIAL BALANCE
Special attention should be given to calculations shown in Parts 1 and 3.
1. Weight of gaseous products.
Carbon Balance:
Carbon gasified: Basis; 100 lb. of coal charged.
Carbon in coal = 100 X 0.6631 = 66.31 lb or 5.526 lb-atoms
591
Carbon in deposited tar = —— X 0.93 0.78 lb or 0.065 lb-atom
Weight of refuse = 4.58/0.9582 = 4.79 lb
(calculated on the basis of the ash as weighed in the proximate analysis).
CHAP. IX] MATERIAL BALANCE OF GAS PRODUCER 367
Carbon in refuse = 4.79 X 0.0308 = 0.148 lb or 0.012 lb-atom

Carbon gasified = 5.526 - (0.065 + 0.012) = 5.449 lb-atoms
Carbon in clean gases. Basis: 1.0 lb-mole of dry, tar- and soot-free gas.
C in CO2 = 0.0712 lb-atom
C in CO = 0.2185 lb-atom
C in CHi = 0.0325 lb-atom
Total = 0.3222 lb-atom
Water, tar and soot in gas. Basis: 1000 cu ft of wet gases at 1075°F and 29.75
in. of Hg.
>, , , , 1000
Moles of total gas = — —— = 0.888 lb-mole
looo J9.9J
^^^ >< 152 ^ 2i75
„ „ 3.43 X 1000
H2O present = -——; —- = 0.0272 lb-mole
^ 7000 X 18
Moles of dry gas = 0.8608 lb-mole
Water per mole of dry gas 0.0272/0.8608 = 0.0316 lb-mole
„ . , , 3.31 X 1000
Tarpermoleofdrygas.^Q^^^^gg^g= 0.5501b
1.52 X 1000
Soot per mole of dry gas TT^r:::; = 0.252 lb
7000 X 0.8608
Total carbon in gases. Basis: 1.0 lb-mole of dry, clean gas.
C in clean gases = 0.3222 lb-atom
C in tar = 0.550 X 0.9 = 0.495 lb or 0.0412 lb-atom
C in soot = 0.252 lb or 0.0210 lb-atom
Total 0.3844 lb-atom

Moles of dry gas per 100 lb of coal = 5.449/0.3844 = 14.18 lb-moles.
Total products in gases. Basis: 100 lb of coal charged.
COj = 14.18 X 0.0712 = 1.01 lb-moles or X 44 = 44.5 lb
Oj = 14.18 X 0.0090 = 0.12 lb-mole or X 32 = 3.8 lb
CO = 14.18 X 0.2185 = 3.10 lb-moles or X 28 = 86.8 lb
CHi = 14.18 X 0.0325 = 0.46 lb-mole or X 16 = 7.4 lb
Hj = 14.18 X 0.1365 = 1.94 lb-moles or X 2 = 3.8 lb
N2 = 14.18 X 0.5324 = 7.55 lb-moles or X 28.2 = . . . . 213.0 lb
Total dry, clean gas 14.18 lb-moles or 359.3 lb
Water
14.18 X 0.0316 = 0.45 lb-mole, or X 18 = 8.1 lb
Tar
C = 14.18 X 0.0412 = 0.58 lb-atom, or X 12 = 7.02 lb
'- • Hs = 14.18 X 0.055/2 = 0.39 lb-mole, or X 2 = 0.78 lb
Total tar = 7.8 lb
Soot
C = 14.18 X 0.0210 = 0.30 lb-atom, or X 12 = 3.6 lb
Total products in gasea = 378.8 lb
2. Weight of air supplied.

Nitrogen Balance. Basis: 100 lb of coal charged.
Nj in gas = 7.56 lb-moles
Nj in coal = 1.52/28 = 005 lb-mole
Ns from air = 7.50 lb-moles
Dry air supphed = 7.50/0.79 = 9.48 lb-moles
or 9.48 X 29.0 = 275 lb
Molal humidity of air (Fig. 9) = 0.023
Water in air = 9.48 X 0.023 = 0.218 lb-mole
or 0 . 2 1 8 X 1 8 = 3.91b
Total wet air = 0.218 + 9.48 = 9.70 lb-moles
or 275 4 - 3 . 9 = 278.9 1b
3. Weight of steam introduced.

Hydrogen Balance. Basis: 100 lb of coal charged. / A?,
>' f - •' , : Output ''-•'• 'r •V
591
Hj m deposited tar = -— X 0.07 = 0.0585 lb = 0.029 lb-mole
709
Free Hj in gas = 1.94 lb-mole
H2 in CHi in gas = 0.46 X 2 = 0.92 lb-mole
H2 in H2O in gas = 0.45 lb-mole
H2 in suspended tar = 0.39 lb-mole
Total output of Hj = 3.73 lb-moles
Input
Net H2 in coal = 3.53/2.02 = 1.75 lb-moles
H2 in water in coal = 23.16/18 = 1.28 lb-moles
H2 in water iikair = 0.22 lb-mole '
Total input in addition to steam = 3.25 lb-moles
H2 from steam = 3.73 - 3.25 = 0.48 lb-mole
Steam introduced = 0.48 lb-mole or 8.64 lb
4. Overall material balance. " ,

Input Output
Coal 100 lb Refuse 4.8 lb
Dry air (9.48 lb-moles)... 275 lb Tar deposited 0.83 lb
Water in air (0.22 lb-mole) 3.91b Soot suspended 3.6 lb
Steam (0.48 lb-mole) 8.61b Tar suspended 7.8 lb
Dry clean gas (14.18 lb-moles) 359.3 lb
Total 387.5 lb Water in gas (0.45 lb-moles). 8.1 lb
Total 384.431b
The slight discrepancy in the totals of the material balance results from inaccuracies
of the data and neglect of the sulfur content of the coal. The material balance is
also summarized in Fig. 73.
CHAP. IX] ENERGY BALANCE OF GAS PRODUCER 369
5. Gaseous volumes. Basis: 100 lb of coal charged.

535 29 92
Volume of wet air = 9.70 X 359 X — X — V = 3,800 cu ft
1535 29 92
Volume oi wet gases = 14.63 X 359 X - — X - ~ = . . . 16,500 cu ft
492 29.75
ENERGY BALANCE
Special attention is called to calculations in Parts 2, 3 and 5 of output.

Basis: 100 lb of coal charged. ,
Input
1. Heating value of coal = 100 X 11,910 = 1,191,000 Btu
2. Sensible enthalpy of coal = 0 Btu
3. Enthalpy of steam:
Pressure = 25.3 + 14.7 = 40.0 lb per sq in.
From steam tables, the enthalpy of saturated steam at this pressure referred to
32°F is 1171 Btu per lb
Referred to 75°F this becomes 1171 - 43 = 1128 Btu per lb
Enthalpy of steam = 8.6 X 1128 = 9,700 Btu
4. Enthalpy of dry air = 0 Btu
5. Total enthalpy of water vapor in air.
Dew point (Fig. 9) = 66°F
Enthalpy (as in Illustration 4, page 355) = 4,100 Btu
Total = 1,204,800 Btu
Output
1. Heating value of dry clean producer gas. (Calculated from
composition) 794,000 Btu
2. Heating value of suspended tar = 7.80 X 16,000 = 125,000 Btu
3. Heating value of soot = 0.30 X 96,630 X 1.8 = 52,100 Btu
4. Heating value of carbon m refuse = 0.012 X 96,630
X 1.8 = 2,100 Btu
6. Heating value of deposited tar = 0.83 X 15,500 = .. 12,900 Btu
6. Enthalpy of dry, clean producer gas.
Mean heat capacities between 75°F and 1075°F taken from Fig. 37, page 217.
CO2 (1.01) (11.05) = 11.15
0^(0.12) (7.59) = 0.91
' " " ' i" CO(3.10)(7.26) = 22.50
• "' * • CH4(0.46) (12.03) = 5.54
H2(1.94)(7.01) = 13.60
N2(7.55)(7.27) = 54.89
Total = 108.59
Enthalpy = (108.59) (1075 - 75) = 108,590 Btu
7. Sensible enthalpy of tar and soot.
Suspended tar and soot = 11.4 X 0.32(1075 - 75) = . . . 3,650 Btu
Deposited tar and soot = 0.8 X 0.32(1075 - 75) = 250 Btu
8. Total enthalpy of water vapor in gases.

Total molal enthalpy calculated by method used in Illus-
tration 4 = 27,470 Btu
Total enthalpy = 27,470 X 0.45 = 12,360 Btu
9. Sensible enthalpy of refuse = 0.22 X 4.8 (350 - 75) = . 290 Btu
Total energy accounted for = 1,111,240 Btu
10. Heat losses, radiation, etc. == 1,204,800 - 1,111,240 = .. 93,560 Btu
Summary of Energy Balance:
Input
1. Heating value of coal 1,191,000 Btu 98.9%
2. Sensible enthalpy of coal 0 Btu 0%
3. Sensible enthalpy of air 0 Btu 0%
4. Enthalpy of steam 9,700 Btu 0.8%
5. Total enthalpy of water in air 4,100 Btu 0.3%
Total 1,204,800 Btu 100.0%
Coal t8.9% •V'

0.8% Enthalpy
of Steam
0.3% Enthalpy
of H,0 Vapo>
In'Air
" •'•' H.V. of Carbon in Refuse 0.2%
^ >H. V. of Deposited Tar 1.0%
I Heating Value of
Suspended Tar 10.4%
v::: 3 Heating Value of
Suspended Soot 4.3%
Heating Value of
DTy Broduccr Gas 66.0%
Sensible Enthalpy of Suspended Tar 0.2%

Sensible Enthalpy of Suspended Soot 0.1%
Sensible Enthalpy of Producer Gas 9.0%
3
Total Enthalpy of H , 0 Vapor 1.0%
'Radiation Loss 7.8%

Total 100.0%
FIG. 74. Energy balance of a gas producer (Illustration 6).

CHAP. IX] GRAPHICAL CALCULATION OF COMBUSTION 371
Oidput
1. Heating value of clean gas 794,000 Btu 66.0%
2. Heating value of suspended tar 125,000 Btu 10.4%
3. Heating value of suspended soot 52,100 Btu 4.3%
4. Heating value of carbon in refuse 2,100 Btu 0.2%
5. Heating value of deposited tar 12,900 Btu 1.0%
6. Enthalpy of dry, clean gas 108,590 Btu 9.0%
7. Enthalpy of tar and soot 3,900 Btu 0.3%
8. Enthalpy of water vapor in gases 12,360 Btu 1.0%
9. Enthalpy of refuse 290 Btu 0.0%
10. Heat losses, radiation, etc 93,560 Btu 7.8%
Total 1,204,800 Btu 100.0%
This energy balance is presented in diagrammatic form in Fig. 74.
• ' ' THERMAL EFFICIENCY
Cold gas. The effectively utilized energy includes only the heating value of the
dry, clean gas.
^^ . 794,000 ^
, Efficiency = ' = 66%
./ •^ 1,204,800
Hot gas. The effectively utilized energy includes the total sensible enthalpy of all
materials in the gases and also the heating value of the suspended tar and soot.
Sensible enthalpy of dry, clean gases = 108,590 Btu
Sensible enthalpy of water vapor in gases = 12,360 — (0.45
X 18,870) = 3,870 Btu
Sensible enthalpy of suspended tar and soot = 3,650 Btu
' Heating value of dry, clean gas = 794,000 Btu
Heating value of tar and soot = 177,100 Btu
Effectively utilized energy = 1,087,210 Btu
Efficiency = i^58Zl?H = 90.2%

1,204,800 "
The principal loss in the operation of this particular gas producer is by radiation
and conduction of heat from the apparatus.
GRAPHICAL CALCULATION OF COMBUSTION PROBLEMS

Where short-cut methods of combustion calculations are desired a
simple graphical solution may be resorted to provided that the following
items are negligible: sulfur content of fuel, combustible content of
refuse, water vapor in air, suspended tar and soot in gases, and hydro-
carbons in flue gases. The simplified graphical solution of such com-
bustion problems appears in Fig. 75. This method also serves as an

approximate solution for other combustion problems where the experi-
mental data or time required do not warrant precise methods. From
Percent Excess Air
Fia. 75. Combustion chart for fuels.
Fig. 75 there may be obtained directly the pounds of air used and the
pound-moles of wet flue gas produced and the carbon dioxide content of
the dry flue gas on the basis of 1 pound of combustible in oil, coal, gas,
or coke.
CHAP. IX] PROBLEMS 373
PROBLEMS
1, A Kentucky coal has the following analysis:
Ultimate analysis of combustible
Proximate analysis as received corrected ash-free moisture-free basis
Moisture 2.97% C 84.39%
Ash (uncorrected) 2.94% Net hydrogen 4.81%
Volatile matter 37.75% N 2.00%
Fixed carbon 56.34% S 1.02%
Combined H20 7.78%
100.00% 100.00%
The combustible referred to above includes those portions of coal as received which
are not classified as moisture or corrected ash.
Determine, on the " as received " basis:
(a) The rank of this coal.
(5) The total heating value by Dulong's formula. . ;
(c) The total heating value by Uehling's method. ' ' " •"'"''' ' ' -
(d) The estimated available hydrogen content by Uehling's method.
(e) The net heating value by Uehling's method.
(/) Ultimate analysis (as received, corrected ash).
2. In the following table are the analyses of several typical coals, as given in Bureau
of Mines Report of Investigations, No. 3296, " Classification Chart of Tjrpical Coals
of the United States." Both the ultimate and proximate analyses are on the " as
received " basis, and the hydrogen given in the ultimate analysis includes the hydro-
gen of the free moisture in the coal.
No. 1 No. 2 No. 3 No. 4 No. 5
Proximate Analysis
Moisture 4.3% 2.0% 2.2% 10.6% 42.6%
Volatile Matter 3.0 9.6 21.1 36.4 24.7
Fixed Carbon 82.6 77.8 66.2 42.4 27.0
Ash (uncorrected) 10.1 10.6 10.5 10.6 6.7
Ultimate Analysis
Carbon 82.5% 79.5% 76.1% 62.8% 37.3%
, Hydrogen 2.0 3.7 4.8 5.6 7.3
Nitrogen 0.6 0.9 1.2 1.2 0.5
Sulfur 0.5 0.6 2.6 3.3 0.6
Heating Value
Btu/lb 12,650 13,520 13,530 11,300 6,260
No. 1 Anthracite No. 4 High volatile bituminous
No. 2 Semi-anthracite No. 5 Lignite
No. 3 Medium volatile bituminous
For one of the coals, present the following calculations:

(a) Recalculate the ultimate analysis, reporting:
(/) Carbon (S) Sulfur
(^) Available hydrogen (6) Nitrogen
(S) Free H2O (7) Corrected Ash
M) Combined H2O
(b) Calculate total heating value by Dulong's formula. Compare the result
with the experimental value.
(c) Calculate the total heating value by Uehling's method. Compare the result
with the experimental value.
{d) Assuming that the proximate analysis and heating value alone are available,
calculate the per cent of carbon and of available hydrogen by Ueh-
ling's method.
(e) Calculate the net heating value.
3. A gas has the following composition by volume: •
Illuminants (C2H4 and CeHe) 53.6%
; *• „ O2 1.6%
CH4 16.9%
CzHe 24.3%
N2 .*. 3.6%
The heating value of this gas is 1898 Btu per standard cubic foot. Calculate the
percentages of C2H4 and CeHe in the gas.
4. Calculate the maximum theoretical flame temperature in degree Centigrade
when the following gas is burned with the theoretical amount of dry air starting with
air and gas at 18°C:
CO 30%
Hj 15%
O, 1%
CO2 5%
N2 _49%
100%
5. Calculate the theoretical flame temperature when the above gas is burned with
100% excess air.
6. Calculate the theoretical flame temperature of the above-mentioned gas when
burned with the theoretical amount of air and when both gas and air are preheated to
500°C before combustion.
7. Calculate the theoretical flame temperatm'e of the above gas when burned with
the theoretical amount of air, the combustion of both CO and H2 proceeding to only
80% completion. The gas and air are initially at 18°C.
8. A fuel gas has the following analysis: - /
Carbon dioxide 3.0% '
Ethylene 5.8%
Benzene 0.8% ,
Oxygen 0.9%
Hydrogen 38.8%
Carbon monoxide- 35.2%
Methane 9.5%
Ethane 0.7%
Nitrogen 5.3%
100.0%
C H A P . IX] PROBLEMS 375
Calculate the theoretical flame temperature if this dry fuel gas is burned with the
theoretical amount of dry air. Gas and air are supplied at 18°C. Assume that com-
bustion is complete. Use 1 g-mole of the dry fuel gas as the basis of calculation.
9. For the fuel gas whose analysis is given in Problem 8, calculate the theoretical
flame temperature if the dry gas is burned with the theoretical amount of dry air.
T h e gas is supplied a t 1&°C, b u t the air is preheated t o 700°C. Assume t h a t com-
bustion is complete. Use 1 g-mole of the dry fuel gas as the basis of calculation.
10. For the fuel gas whose analysis is given in Problem 8, calculate the theoretical
flame temperature if t h e gas is burned with 100 % excess air. T h e gas and air both are
dry and are supplied a t 18°C. Assume t h a t combustion is complete. Use 1 g-mole
of the dry fuel gas as the basis of calculation.
11. The dry fuel gas whose analysis is given in Problem 8 is burned with the
theoretical amount of dry air. Both air and gas are supplied at 18°C. If carbon
monoxide and hydrogen burn to 8 0 % completion, what is the theoretical flame tem-
perature? Assume t h a t the combustion of the gases other than CO and H2 is com-
plete. Use 1 g-mole of the dry fuel gas as the basis of calculation.
12. Four thousand (4000) kilograms of coke at an initial temperature of 1400°C
has 360 kg of steam blown through it, forming 448 kg CO, 88 kg CO2, and 40 kg H2
at an average temperature of 1000°C. The steam is supplied a t 120°C, saturated.
The radiation loss is 100,000 kcal. Calculate the final temperature of the bed of coke.
13. A fuel gas has the following composition by volume:
: CO2 2.1%
•> O2 0.5%
C2.BH4.2 (illuminants) 7.0%
CO 33.8%
H2 40.6%
C1.2H4.4 (paraffins) 11.2%
_,,:: Nj... 4.8%
100.0%
(a) Calculate the analysis of the flue gases formed by burning this gas with 3 0 %
excess air, assuming that all combustible components are burned to
CO2 and H2O.
(b) Calculate the heating value in Btu per standard cubic foot.
14. A fuel oil has a specific gravity of 0.91 and a viscosity of 28 Saybolt Furol
seconds at 122°F. Estimate the characterization factor, average molecular weight,
average boiling point, hydrogen content, specific heat at 150°F, and heating value
of this oil:
15. A petroleum fraction has a specific gravity (60/60°F) of 0.88. When tested in
a Saybolt Universal viscometer a t 122°F, a reading of 68 seconds was obtained. Using
the charts in the text, report the following values:
(a) Characterization factor.
(6) Average molecular weight.
(c) Average boihng point.
(d) Hydrogen content.
(e) Specific heat of the hquid oil at 200°P. . •
(/) Specific heat of the vapor at 800°F.
ig) Heating value.
376 F U E L S A N D COMBUSTION [CHAP. I X
16, A Pennsylvania bituminous coal has the following composition:

H 4.71%
C... 69.80%
N 1.42% • ' ' ''
O2 7.83%
Ash 6.73%
H20 9.51%
Total heating vakie 6950 koal per kg
This coal is gasified in a gas producer using air a t 20°C saturated with water vapor.
No additional steam or water is admitted into,^;he producer. The barometric pres-
sure is 740 mm. The resulting gas has the following composition:
: . Hj 0.5% •
; CO 21.2% - : ...
- • N2 64.5% /-t
' CHi 5.8% 'H
,.;.,.;,.,., H. CO2 6.2% - • .'. .Ui • 'f
O2 1.8% -^
I t may be assumed t h a t no tar or soot is present in the producer gas. During a
test period the total coal charged is 10,500 kg.
The dry refuse formed weighs 825 kg and contains 13.3%C.
Temperature of refuse = 220°C. ' ,
Mean specific heat of refuse = 0.25. ) /
Temperature of outgoing gases = 450°C.
(a) Calculate complete material and energy balances of this gas producer on the
basis of 100 kg of coal as charged.
(b) Calculate the thermal efficiency on both the hot and cold bases.
(c) Calculate the total volume of gases leaving the producer
(d) Calculate the heating value of the producer gas in Btu per cubic foot meas-
ured a t 60°F, 30 in. of Hg, saturated with water vapor.
17. A gas producer is charged with bituminous coal having the following compo-
sition :
Proximate analysis: -^ •'
Moisture 2.70% * -• '
•' Volatile matter 25.77% ;
,; Fixed carbon 62.87%
'• Ash 8.66% [•/.,,
100.00%
Ultimate analysis: - ;
Moisture 2.70%
Carbon 78.55%
Net hydrogen 4.13%
Nitrogen 1.58%
Sulfur 0.69%
Corrected ash 8.40%
Combined H2O 3.95%
100.00%
Total heating value = 13,944 Btu per lb.
CHAP. I X ] PROBLEMS 377
Air is supplied at 75°F with a percentage humidity of 90%. The barometric

pressure is 29.65 in. of Hg. Dry, saturated steam is supplied at a gauge pressure of
50 lb per sq in. The producer gas leaves at a temperature of 1220°F and has the
CO 25.0%
H2 22.0%
CH4 3.6%
t , , .' C2H4..; 2.8%
, . CO2 9.2%
. : N2 37.4%
100.0%
A sample of gas is withdrawn and cooled to 100°F for determination of suspended
tar and soot. The tar and soot content is 10 grains per cu ft of gas measured at
barometric pressure, 100°F, and saturated with water vapor. The tar and soot
contain 95% carbon and 5% hydrogen. Its heating value is 17,100 Btu per lb,
and its mean specific heat is 0.34. The dew point of the producer gas is 100°F.
The refuse is discharged at 400°F, moisture-tree, and containing 4.52% carbon.
The specific heat of the refuse is assumed to be 0.23. The meanmolal heat capacity
of CjH, (75° to 1220°F) is 20 Btu per lb-mole per "F.
Neglecting deposition of tar in the flues and presence of sulfur, calculate:
(a) Complete energy and material balances of the producer, based on 100 lb of
coal charged.
(6) The thermal efficiencies on both the hot and cold bases.
(c) The volume of producer gas, measured at 60°F, 30 in. of Hg, saturated with
water, formed per 100 lb of coal charged.
(d) The heating value of the producer gas, per standard cubic foot.
The solution of this problem results in a negative radiation loss. What errors in
experimental data are most likely responsible for this condition?
18. An Illinois bituminous coal is burned in a boiler furnace with air at 78°F, 92%
percentage humidity. The barometric pressure is 29.40 in. of Hg. The furnace
gases leave at 553°F. The refuse is discharged moisture-free, at 440°F. The refuse
as analyzed contains 12.2% carbon as coke and 16.1% moisture. The specific heat
of the dry refuse is 0.25. The proximate analysis of coal is:
Fixed carbon 50.34%
Moisture 9.61%
Ash 9.37%
Heating value of coal = 11,900 Btu per lb.
The ultimate analysis on the moisture-free basis is:
Carbon 73.70% ' ^*'
Hydrogen 4.75%
Oxygen 9.23%
Nitrogen 1.58% ^
Sulfur 0.55%, , •
Corrected ash 10.19%
. .i-s.'..fc~;o7s»i,i .v^ :.•: ••• .K.V1. •«; . . 100.00% .
The flue gas analysis is as follows:

CO2 12.2%
CO 0.2%
Oj 7.0%
Ns 80.6%
100.0%
Calculate:
(a) Total material and energy balances for this process based on 100 lb of coal
charged, neglecting the combustion of sulfur. The heat losses and the
heat efEectively utilized may be considered toîther as a single item of heat
output.
(6) Percentage excess air used in combustion, based on that required for com-
plete combustion of all coal charged.
(c) The dew point of the flue gases.
(d) The actual volumes in cubic feet at the given condition of temperature,
humidity and pressure of the flue gases and air supply, per 100 lb of coal
charged.
19, A bituminous coal is burned in a boiler furnace with air at 85°F, 90% per-
centage humidity. The barometric pressure is 29.20 in. of Hg. The furnace gases
leave at 572°F. The refuse leaves the furnace moisture-free at 520°F and when
analyzed contains 22.3% moisture, 12.3% volatile matter and 41.4% fixed carbon.
The mean specific heat of the refuse is 0.23. The proximate analysis of the coal is:
Fixed carbon 56.34%
Moisture 2.97%
Ash 2.94%
A partial ultimate analysis on the corrected ash- and moisture-free basis is:
Carbon 84.43%
Nitrogen 2.00%
Sulfur '. 0.82%
The total heating value of the coal is 14,139 Btu per lb. Dry, saturated steam
at a gauge pressure of 150 lb per sq in. is produced at the rate of 780 lb per 100 lb
of coal charged. Water is fed into the boiler at 72°F. The Orsat analysis of the flue
gas is: . • ^
CDs ' 12.0%
w= CO 1.2%
• C. O2 6.2% y
N, 80.6%
100.0%
Calculate:
(a) The net hydrogen content of the coal from an oxygen balance, neglecting
combustion of sulfur.
(6) The complete ultimate analysis of the coal.
(c) The complete material balance of the process, based on 100 lb of coal charged.
(d) The complete energy balance of the furnace, based on 100 lb of coal charged.
(e) The thermal efficiencies of the furnace and boiler, based on the total and on
the net heating values.
(J) The percentage excess air used, based on the total combustible charged.
CHAP. IX] PROBLEMS 379
20. A heat at interchanger, used for heating the oil in a circulating hot oil heatihg
system, is fired with coal having the following proximate analysis:
Moisture ;. 12.38%
- Volatile matter 36.88%
Fixed carbon 37.50%
Ash 13.24%
100.00%
The heating value of the coal is 10,361 Btu per lb, and its sulfur content is 5.1%.
The coal is burned with air at a temperature of 70°F and a percentage humidity of
60%. The barometric pressure is 29.3 in. of Hg.
The refuse from the furnace is discharged at a temperature of 600°F and contains
16% fixed carbon and 84% ash. The sulfur content of the refuse is 7.8%. Its
specific heat may be taken as 0.23.
The flue gases leave the furnace at a temperature of 850°F and have the following
composition by volume, on the siilfur- and moisture-free basis.
. :^ ; , CO2 11.50%
* ' • • : ' 1 CO 0.17%
•• 'i '*' O2 7.51%
N2 80.82%
• . /
100.00%
The oil is circulated at a rate of 3800 lb per 100 lb of coal charged and is heated
from 1S5°F to 464°F. The mean specific heat of the oil in this temperature range
is 0.55. Calculate;
(o) The complete ultimate analysis of the coal as estimated from the rank and
heating value.
(6) The complete material and energy balances of the interchanger, based on
100 lb of coal charged.
(c) The complete analysis, by volume, of the wet flue gases leaving the inter-
changer.
(d) The percentage excess air, based on the combustible actually burned.
(e) The volume of wet flue gases leaving the interchanger.
(/) The thermal efficiencies of the interchanger, based on both the total and net
heating values.
21. A brick kiln, of an intermittent type, is fired with 10,420 lb-moles of dry
producer gas. The weight of green ware is 410,000 lb containing 0.52% mechanical
water and 3.02% chemically combined water. The gas enters the kiln at 1220°F
and a pressure of 29.65 in. of Hg, and contains 10 grains of tar (90% C, 10% H)
per cu ft measured at a pressure of 29.65 in. of Hg, 100°F, and saturated with water
vapor. The dew point of the producer gas is 100°F.
During the water-smoking period mechanical water is vaporized and leaves the
kiln at 300°F, and the chemically combined water leaves at 400°F. During water-
smoking the saturation temperature of the gases is 160°F. The flue gas leaves at
an average temperature of 720°F. The average temperature of the ware at the end
of the burn is 2100°F, and its specific heat is 0.23. The producer gas is burned with
air at 75°F, 90% percentage humidity, and a pressure of 29.65 in. of Hg. The average
380 FUELS AND COMBUSTION [CHAP. I X
analysis of gases by volume on the moisture-free basis is:

Producer gas * Flue gas
CO 25.00% CO2 12.20%
H2 22.00% CO 0.16%
CH4 3.60% Ha y... 0.14%
C2H4 2.80% O2 7.10%
COS 9.20% N2 80.40%
N2 37.40% 100.00%
100.00%
Calculate the material balance of the combustion process, the energy balance of the
entire unit, and the thermal efficiency. The heat absorbed by the kiln structure may
be included with the other undetermined heat losses as a single item.
22. Limestone is burned in a continuous vertical kiln which is heated by coal burned
on an external grate located beside the bottom of the kiln shaft. The limestone is
charged at the top of the shaft a t atmospheric temperature and gradually descends
in contact with a rising stream of the fiue gases from the grate. The burned lime is
discharged from the botton\of the shaft at a temperature of 950°F. The flue gases,
mixed with all gases and vapors evolved by the charge, leave the top of the shaft a t
560°F. For each 100 lb of coal burned, 161 lb of burned Ume are produced. The
limestone charged has the following composition:
CaO 51.0%
MgO 2.0%
CO2 42.2%
AI2O3 1.5%
• •'•* Si02 1.2% . -•,^'
H20 2.1% '' '• •' '
100.0%
The ultimate analysis of the coal is as follows:
Moisture 10.69%
C 66.62%
NetH 3.18%
N 1.57%
S 1.91%
Corrected ash 6.41% . Vj
Combined H20 9.62% ,,,:,
100.00%
The total heating value of the coal is 11,805 Btu per lb. The flue gases have the
CO2 16.4%
N2 76.8%
O2 6.8%
100.0%
The coal is burned with air at a temperature of 70°F having a humidity of 80 per
cent. The barometric pressure is 29.4 in. of Hg.
T h e refuse from the grate contains 4.2% fixed carbon and 95.8% ash. I t s sulfur
content is 3 . 1 % .
CHAP. IX] PKOBLEMS 381
It may be assumed that in the burning process all CO2 and water are driven from
the limestone. The heat of wetting of granular limestone is neghgible. It may be
assumed that the sulfur burned forms SO2 which is further oxidized and absorbed
by the hme to form CaS04. The mean specific heat of the burned hme is 0.21.
Calculate the complete energy and material balances of the grate and kiln on the
basis of 100 lb of coal fired.
Calculate the thermal efficiency of the process, considering the effectively utilized
heat to be consumed in the decomposition of the hmestone.
23. A 12-hour test was conducted on a steam-generating plant with four of the
boilers in operation. The data for the 12 hour test are as follows:
Proximate analysis of coal
Moisture 4.38% Fixed carbon 48,98%
Volatile matter 29.93% Ash (uncorrected) 16.71%
100.00%
Half of the sulfur of the coal appears in the volatile matter.
Ultimate analysis of coal
Carbon 65.93% Combined water 6.31%
Available hydrogen 3.50% Free moisture 4.38%
Nitrogen 1.30% Ash (corrected,
Sulfur 2.99% sulfur free) 15.59% "
100.00%
Heating value of coal as fired (total) 11,670 Btu per lb
Weight of coal fired 119,000 lb
Temperature of coal 65°F
Proximate analysis of refuse
Moisture 4.77% Fixed carbon 12.51%
Volatile matter 2.08% Ash (uncorrected) 80.64%
. ^ 100.00%
Flue gas
CO2 11.66% CO 0.04%
O2 6.52% N2 81.78%
100.00%
Temperature 488°F
Ai| . '. ' _
Dry-bulb temperature 73°F
Wet-bulb temperature 59.4°F
Barometer 29.08 in. Hg
Steam.
Feed water 193°F
Water evaporated 1,038,400 lb
Steam pressure 137.4 lb per sq in. gauge
Quality 98.3%
Calculate the complete material and energy balances for this steam-generating
plant.
24. Calculate the complete energy balances for Illustration 5.
26. A steam superheater is to be designed to heat 10,000 pounds per hour of low
pressure process steam from 250°F to 700°F. The fuel is a residual cracked oil having
an API gravity of 8.6 and a viscosity of 200 Saybolt Furol Seconds at 122°F. The
design bases are as foUows: ^^
Percentage excess air 40
Air temperature "F 70
Relative humidity of air 60
Stack temperature °F 600
Assuming a radiation loss of 5% of the beating value of the fuel burned, calculate
complete material and energy balances for 1.0 hour of the combustion operation and
the thermal efficiency based on the total heating value of the fuel. Fig. 75 may be
used for establishing the material balance of the combustion.
CHAPTER X
CHEMICAL, METALLURGICAL, AND PETROLEUM
PROCESSES
The methods applied in calculating material and energy balances are
alike in principle for all industrial processes, differing only in detail.
Three illustrative material and energy balances are presented in this
chapter, representing the procedure applicable to chemical processes
in general. Typical illustrations in the chemical, metallurgical, and
petroleum industries have been selected. The chemical and metallur-
gical illustrations summarize the principles involved in dealing with
complex chemical reactions -where the intermediate courses of the
reactions are unknown. The petroleum process illustrates the use of
an energy balance to establish a material balance and the principles of
recycling to increase yield and effect temperature control. The chemical
and metallurgical processes illustrate the analysis of experimental data
from existing plants while the petroleum process is presented as a
problem in the design of a new plant.
CHAMBER SULFURIC ACID PLANT

The material and energy balances of a chamber process sulfuric acid
plant are selected for examination as representative of the chemical in-
dustries. The operating conditions have been taken from data pub-
lished by Kaltenbach.i In this particular problem iron pyrites is
burned with air in a shelf burner. The burner gases consisting of
sulfur dioxide, oxygen, and nitrogen pass through a dust chamber
where suspended matter is removed, and then into the Glover tower.
In the Glover tower the hot gases meet a descending stream of acids
from the Gay-Lussac tower and chambers. Oxides of nitrogen lost in
the system are replaced by nitric acid introduced at the top of the tower.
In passing through the Glover tower the hot gases are cooled and some
conversion of SOa to sulfuric acid takes place, the chamber acid is con-
centrated, and the oxides of nitrogen are evolved from the acids for
recirculation. The mixed gases leaving the Glover tower pass into a
series of large lead chambers where conversion to H2SO4 is completed.
Finally, the oxides of nitrogen are recovered in the Gay-Lussac tower
' Chemical Age, 28, 295 (1920).
383
384 CHEMICAL, METALLURGICAL, PETROLEUM PROCESSES [CHAP. X
by passing the spent gases from the chamber countercurrent to a/ Stream

of cold sulfuric acid from the Glover tower. These different steps in
the manufacture and the material and energy balances of the burner,
Glover tower, chambers and Gay-Lussac tower are each discussed sepa-
rately. The balances for each unit are based upon 100 kilograms of
dry pyrites charged into the burner. The reference temperature for the
energy balances is taken at 18°C.
100 1 1 1
80 1.8
/
-
60
><=•
ST
« 40
-n. • -
20
^"^ < ' 1 1.0

40 60 100
Per Cent flaSOi
FIG. 76. Density of aqueous solutions of sulfuric acid at 15°/15°C.
The concentration of a strong aqueous solution of sulfuric acid is

usually determined by measurement of its specific gravity. Concen-
trations are usually expressed in terms of specific gravities or of degrees
Baum6 rather than in percentages. In Fig. 76 are curves relating per-
centages of H2SO4 by weight to the degrees Baum6 and specific gravities,
15°/15°C, of the aqueous solutions.
Illustration 1. Material and Energy Balances of a Chamber Sulfuric Add Plant.
Pyrites, containing 85.3% FeSj, 2% H2O, and 12.7% inert gangue, is burned in a shelf
furnace yielding a gas containing 8.5% SO2, 10.0% O2, and 81.5% N2 by volume.
The pyrites is charged at 18°C and the air is supplied at the same temperature with a
percentage humidity of 49% and at a pressure of 750 mm of Hg. The cinder leaves
the burner at 400°C containing 0.42% sulfur as unburned pyrites. The pyrites
burned forms Fe203 and SO2 and the gangue passes into the cinder unchanged. The
mean specific heat of the cinder is 0.18. The gases from the burner, after passing
through the dust chamber, enter the Glover tower at 450°C and leave at 91°C. In
CHAP. X] M A T E R I A L BALANCE O F E N T I R E PLANT 385
the Glover tower 1 6 % of the SO2 in the gas is converted to H2SO4. There are sprayed
into the top of the Glover tower, per 100 kg of moisture-free pyrites charged:
182 kg of aqueous sulfuric acid at 25°C from the chambers, containing
64.0% H2SO4 and 3 6 % H2O.
680 kg of mixed acid at 25°C from the Gay-Lussac tower, containing 7 7 %
H2SO4, 2 2 . 1 % H2O, and 0.885% N2O3.
1.31 kg of aqueous nitric acid a t 25°C, containing 3 6 % HNO3 and 6 4 % H2O.
Acid leaves the bottom of the Glover tower, free from oxides of nitrogen, at a tem-
perature of 125°C, and containing 78.0% H2SO4 and 22%, H2O. This acid is cooled
to 25°C, part of it is returned to the top of the Gay-Lussac tower, and the remainder
is withdrawn as the final product of the plant. The gases leaving the Glover tower
are passed through a series of four chambers and finally enter the Gay-Lussac tower,
at 40°C. Spray water is introduced at the tops of the various chambers at 18°C.
The acid formed in the chambers is withdrawn from the first chamber at 68°C, con-
taining 64.0% H2SO4. This acid is cooled to 25°C and all fed into the top of the
Glover tower. P a r t of the Glover acid after cooling to 25°C is returned to the top
of the Gay-Lussac tower. The Gay-Lussac acid leaves the bottom of the tower a t
27°C and is all fed to the top of the Glover tower at 25°C. The spent gases leave
the top of the Gay-Lussac tower at 30°C. The flow chart of the entire process is
shown diagrammatically in Fig. 77.
HNO3 1.31 Ke
jpent Gases ^
Dry Pyrites 100 Kg- " 4 2 3 Kg *"
Moisture 2.04 Kg
(14.82 Kg Moles)
Di-yAir 506 Kg I T
(17.42 Kg Moles)
Water Vapor 3.16 Kg
(.18 Kg Moles)
-<
178 Kg 78% Acid
Cinders 71.2 Kg
F I G . 77. Material balance of an entire sulfuric acid plant.
Calculate material and energy balances of the entire plant and of each of the follow-
ing units, all based on 100 kg of moisture-free pyrites charged:
(o) The burner. , ^ , 1 .,
(6) The Glover tower.
(c) The four chambers as a single unit. -r-7s;.
(d) The Gay-Lussac tower.
MATERIAL BALANCE OF ENTIRE PLANT

Before discussing the material and energy balances of the individual units in the sul-
furic acid plant it is desirable to calculate the material balance of the entire plant in
order to have in mind a perspective of the whole process. This balance is repre-
sented by the following entries:
Input Oiitput
Dry ore Moisture in air Acid produced
Moisture in ore Nitric acid Dry spent
r D r y air Spray water Cinder
1. Weight of cinder formed.

Basis: 100 kg of moisture-free pyrites charged.
Weight of pyrites as charged = = 102 kg

Weight of PeSj charged = 102 X 0.863 = 87 kg
or 87/120 = 0.726 kg-mole
Weight of gangue charged '= 102 X 0.127 = 13 kg
Percentage S in cinder = 0.42% .-
Percentage FeS2 in cinder = '
~X^X120= 0.78%
Let X — kilograms of FeSs in cinder.

Weight of FeSz oxidized = (87 - x) kg or (87 - x)/l?0 kg-moles
oy ^ 1
FejOs formed = - - X - X 169.7 = (58 - 0.667x) kg
Weight of"cinder = 13.0 + a; + (58 - 0.667a;) = (71 + 0.333a;) kg
Weight of FeS2 in cinder =
a: = 0.0078(71+0.333*) or, X = 0.56 kg
or = 0.56/120 = 0.0048 kg-mole
FejOs in cinder = 58 - (0.667 X 0.56) = 67.63 kg
or = 57.63/159.7 = 0,361 kg-mole
Weight of cinder = 71.2' kg
2. Weight of dry burner gases.

Sulfur Balance, Basis: 100 kg of dry pyrites charged.
FeSz burned = 0.726 - 0.005 = 0-721 kg-mole
S burned = 0.721 X 2 = 1.442 kg-moles
S per kg-mole of burner gas = 0.085 kg-mole
1.442
Dry burner gas = •' • = 16.95 kg-moles
U.UoO
SO2 = 16.95 X 0.085 = 1.441 kg-moles or X 64 = 92.2 kg
O2 = 16.95 X 0.100 = 1.70 kg-moles or X 32 = 54.4 kg
N2 = 16.95 X 0,815 = 13.81 kg-moles or X 28,2 = 389.8 kg
Total dry gases = 16.96 kg-moles or 636.4 kg
3. Weight of dry air used.

Nitrogen Balance:
Nitrogen in burner gas = 13.81 kg-moles
13 81 J
Air introduced = • _• = 17.5 kg-moles /
or 17.42 X 29 = 506 kg /
4. Weight of water vapor in dry air. The air supply enters at 18°C, 49% percent-
age humidity, and 750 mm pressure. From Fig. 9 the molal humidity is 0.0101.
Water vapor in air = 0.0101 X 17.5 = 0.176 kg-mole
or 0.176 X 18 = 3.16 kg
CHAP. X] MATERIAL BALANCE OF ENTIRE PLANT 387
6. H2SO1 produced in system. Sulfuric acid is formed only in the Glover tower
and in the chambers.
SO2 entering Glover tower = 1.44 kg-moles
SO2 converted to acid in Glover tower = 1.44
X 0.16 = 0.230 kg-mole
H2SO4 formed in Glover tower = 0.230 X 98.1 = 22.6 kg
H2SO4 in acid from chambers = 182 X 0.64 = . 116.5 kg
Total H2SO4 formed = 139.1 kg
or = 139.1/98 = 1.42 kg-moles
Total product of aqueous acid, 78% H2SO4 =
139.1/0.78 = 178 kg
6. Weight of spray water in chambers.
Water Balance:
Output
The small amount of water vapor in the gases from the Gay-Lussac tower may be
neglected. On this basis all water leaves the process in the 78% acid product.
H2O used in forming H2S04 = 1.42 kg-moles
or 1.42 X 18 = 25.5 kg
H2O in aqueous acid = 178 X 0.22 = 39.0 kg
Total H2O output = 64.5 kg
Input
H2O in ore = 102 X 0.02 = 2.04 kg
H2O in air = 3.16 kg
H2O from aqueous nitric acid = 1.31 X 0.64 = 0.84 kg
It is assumed that the HNO3 introduced is completely decom-
posed into HjO, NO, and O2.
HNO3 introduced = 1.31 X 0.36 = 0.47 kg or
0.47/63 = 0.0075 kg-mole
H2O formed from HNO3 = 0.0075/2 = 0.0037 kg-
mole or = 0.067 kg
Total H2O input accoimted for = 6.11 kg
Water supplied by sprays = 64.5 — 6.1 = 58.4 kg
7. Gases leaving Gay-Lussac tower. The gases leaving the acid plant consist of
SO2, O2, and N2 from the burner and the oxides of nitrogen which are supphed by
the nitric acid and lost from the system. Most of the SO2 and a corresponding
amount of oxygen are removed from the burner gases to form H2SO4. It is assumed
that no water vapor leaves the Gay-Lussac tower because of the great affinity of
strong, cold sulfuric acid for water.
SO2 from burner = 1.441 kg moles
SO2 forming HaSOi = 1.42 kg-moles
SO2 in gases leaving = 0.021 kg-mole = 1.34 kg
O2 in burner gases = 1.70 kg-moles
O2 used in oxidizing SOj = 1.42/2 = 0.71 kg-mole
O2 from burner in gases leaving = 0.99 kg-mole = 31.9 kg
N2 in gases from burner = 13.81 kg-moles = 389.8 kg
HNO3 decomposed = 0.0075 kg-mole
According to the reaction, •• ' <• ' ,':•:..HnS .h-><,t;-a <-• ;

2HNO3 = H2O + 2N0 + 3/2 O2 ,; • ^
Oj from HNO3 = 0.0075(3/4) = 0.0056 kg-mole = 0.18 kg
NO from HNO3 = 0.0075 kg-mole = 0.22 kg
Total gases leaving: ^ ^^ , ,,.,,,,
SOz 0.021 kg-mole 1.3 kg
O2 V 0.995 kg-mole 31.9 kg
NO 0.0075 kg-mole 0.22 kg
N2 13.81 kg-moles 389.8 kg
Total 14.82 kg-moles 423.2 kg
Material Balance of Entire Plant
Input >
' • .,HsO Output ,
Dry pjTites 100.0 kg Cinder 71.2 kg
H2O in pyrites 2.04 kg Spent gases (14.82 kg-
Dry air (17.42 kg-moles) 506.00 kg moles) 423.2 kg
H2O in air (0.176 kg-mole) 3.16 kg Acid (78% HzSOi).... 178.0 kg
Nitric acid 1.31 kg
Total 672.4 kg
Spray water 68.4 kg
Total 670.9 kg fS'S.i mi
Dry Ore 100 Kg Dry Gas 636.4 Kg ^

•IKHIfeii;! y.
H,0 2.04 Kg SO, 1.441 Moles
<0.113 Kg Mole) 0, 1.695 Moles
'' N, 13.81 Moles
Water Vapor
5.2 Kg
(0.294 Kg Mole)
Air 506 Kg ;.
,5^ (17.42 kg Moles)
HiO 3.16 Kg
(0.175 Kg Mole)
= ••.!./
/
•
'
Cinders 71.2 Kg ", /
FIG. 78. Material balance of pyrites bm-ner.
MATERIAL BALANCE OF BURNER ,1' .'d

Input Ouipvi
Dry ore 100 kg Cinder 71.2 kg
Moisture in ore 2.04 kg Dry gases 536.4 kg
Dry air (17.42 kg-moles) 606. kg Water vapor (3.16 +
Moisture in air 3.16 kg 2.04) = 5.2 kg
Total 611.2 kg Total 612.8 kg
CHAP. X] ENERGY BALANCE OF BURNER 389
ENERGY BALANCE OF BURNER
The energy balance of the burner includes the heat of combustion of FeSs to FesOa
and SO2 and the enthalpy of the water vapor in the air as the only important sources
of heat. The energy output is distributed as sensible enthalpy of the outgoing cin-
ders and dry gases and as enthalpy of outgoing water vapor. The heat losses include
radiation from the dust chamber and flues up to the entrance of the Glover tower
where the temperature of the burner gases, 450°C, was measured.
1. Heat evolved in the combustion of pjrites.

The reaction involved in the combustion of pyrites and the corresponding standard
heat of reaction are as follows:
4FeS2 + IIO2 = 2Fe203 + 8SO2
i- .* Affis = 2(-198,500) + 8(-70,920) - 4(-35,500)
I Affis = -822,360 i " •
FeSj actually burned = 0.721 kg-mole I
QOQ OCA
Heat evolved in combustion of FeSz burned = 0.721 X —'• = 148,200 kcal
2. Enthalpy of water vapor in air.

»
t Prom Fig. 9, dew point = 7°C.
I* Heat of vaporization at 7°C = 10,630 kcal per kg-mole
g'^ Total enthalpy = 0.176 [10,630 - 18(18 - 7) + 8.0(18 - 7)] = 1,850 kcal
3. Enthalpy of cinder = 71.2 X .16(450 - 18) = 4,900 kcal
4. Enthalpy of dry burner gas. | j

Mean heat capacities between 18°C and 450°C taken from Table VI, page 216.
SO2 = (1.44) (11.02) = 15.8
i? -J.';.;. Q^ = (1.70) (7.45) = 12.6 , ,
'• 1 . aawo Nj = (13.81) (7.12) = 98.2 , v^^
Total = 126.6 \
.; ,iKMt.in Li
Enthalpy = (126.6) (450 - 18) = 54,700 kcal ! J •" ,?
6. Enthalpy of water vapor in burner gases.
Water vapor present = 5.2/18 = 0.29 kg-mole.
Molal humidity = 0.29/16.95 = 0.017
From Fig. 9, dew point = 15°C -
' Heat of vaporization at 15°C = 10,580 kcals per kg-mole
Mean molal heat capacity of water vapor (15°C-450°C) = 8.42
Enthalpy = 0.29 [10,580 - 18(18 - 15) + 8.42(450 - 16)] = 4110 kcal
Summary of Energy Balance of Burner.

•I, . 1 Inptit
1. Heat evolved in combustion of pjoites = 148,200 kcal 98.8%

2. Enthalpy of water vapor in air = 1,850 kcal 1.2%
I Total input = 150,050 kcal 100.0%
Ovipul
1. Enthalpy of cinder = ^ 4,900 kcal 3.3%
2. Enthalpy of dry burner gas = .\ 64,700 kcal 36.5%
3. Enthalpy of water vapor in gases = 4,110 kcal 2.7%
4. Heat losses (by difierence) = 86,340 kcal 57.5%
Total output = 150,050 kcal 100.0%
This energy balance is summarized in Fig. 79. -t
96.8* Combustion of Pyrites to F«,0,«nd SO»
1.2* Enthalpy of Water

Vapor in Air
100.0* Total
*\——Sensible Enthalpy
in Cinders 3.3* yw^r*
'^ Total Enthalpy of

Water Vapor 2.7*
Sensible Enthalpy in
Dry Burner Gas 36.S*
^
Radiation Loeses (mm Burner,

Flues, and Dust Chamber 57.5*
Total 100.0*
\ ;•
Fia. 79. Energy balance of pyrites burner.
MATERIAL BALANCE OF GLOVER TOWER
The useful functions of the Glover tower are as follows:

1. Cooling of burner gases before being blown into the chambers. "
2. Conversion of about 16% of SO2 in burner gas to H2SO4.
3. Mixing of Gay-Lussae and chamber acids and nitric acid.
4. Evolution of oxides of nitrogen from Gay-Lussac acid. t ,/
5. Concentration of chamber acid.
At the top of the tower, mixing of the Gay-Lussac and chamber acids takes place with
the release of the oxides of nitrogen from the Gay-Lussac acid when this acid is heated
and diluted. Nitric acid is also added to make up for the losses of the oxides of nitrogen
through leakage in the system and incomplete absorption in the Gay-Lussac tower.
The oxides of nitrogen are present as a mixture of NO and NO2, but since the equilib-
rium mixture gradually changes owing to variation in temperature and in oxygen
content it is customary to assume that the oxides of nitrogen leave the Glover tower
as NO. The released oxides of nitrogen react in the vapor state with SO2, oxygen
and water vapor, forming liquid nitrosyl-sulfuric acid
-• 2N0 -I- fOj -f- 2SO2 -h H2O -> 2NO2SO2OH
I
CHAP. X] MATERIAL BALANCE OF GLOVER TOWER 391
Because of the high concentration of SO2 in the Glover tower, the decomposition of
this nitrosyl-sulfuric acid is complete, forming sulfuric acid according to the equation
2H2O + 2NO2SO2OH + SO2 = 3H2SO4 + 2N0
The second reaction proceeds much more rapidly than the first in moles per unit vol-
ume of space, chiefly because it is a reaction which proceeds in the liquid phase. The
high concentration in the liquid state permits a more rapid rate of reaction, other
conditions being the same. The final concentration of the Glover acid takes place
at the hottest zone near the bottom of the tower and the acid finally leaves as 78%
H2SO4.
The input of the material balance includes the gases from the burner, the chamber
acid, the make-up nitric acid introduced, and the Gay-Lussac acid. The output
includes the Glover acid and the gases leaving to enter the chambers.
Input to Glover Tower
Basis: 100 kg of moisture-free pyrites charged.
1. Dry biu:ner gases = 16.95 kg-moles = 636.4 kg
2. Water vapor in burner gases = 0.29 kg-mole = 5.2 kg
; 3. Chamber acid (64% HaSOi; 36% H2O) = 182 kg
4. Nitric acid (36% HNO3) = 1.31 kg
5. Gay-Lussac acid (77% H2SO4; 22.1% HjO; and 0.885%
. ' N.2O3) = 580 kg
Total input 1304.9 kg
i
Oidput
1. Weight of Glover acid. The H2SO4 leaving the Glover tower in 78% acid in-
cludes that from the chamber and Gay-Lussac acids and that formed by conversion
of SO2 to H2SO4 in the tower already calculated to be 0.230 kg-mole.
H2SO4 from chambers = 182 X 0.64 = 116.4 kg
H2SO4 from Gay-Lussac acid = 580 X 0.77 = 446 kg
H2SO4 formed in tower = 0.230 X 98 = 22.6 kg
I Total H2SO4 585.0 kg
585
Weight of 78% acid leaving Glover tower = —— = 750 kg
0.78
2. Weight of dry gases leaving the Glover tower. The weight and composition of
the dry burner gases in passing through the Glover tower is changed owing to the
disappearance of some SO2 and the corresponding amount of oxygen in the formation
of SO3 and to the evolution of NO and O2 due to the decomposition of nitric acid and
the release of the oxides of nitrogen from the Gay-Lussac acid.
Dry burner gases entering tower:
SO2 = 1.44 kg-moles
02= 1.70 kg-moles
N2 = 13.81 kg-moles
Total = 16.95 kg-moles
SO2 converted to H2SO4 in tower = 0.230 kg-mole
SO2 leaving tower = 1.44 - 0.230 = , 1.21 kg-moles
O2 used in forming SO3 = 0.230/2 = 0.115 kg-mole
Os remaining from burner gases == 1.70 — 0.115 = 1.58 kg-moles
Nitric acid decomposed =

1.31 X 0.36 = 0.472 kg or 0.0075 kg-mole
2HNO3 = H2O + 2N0 + IJOa
O2 from HNO3 = 0.0075 X f = 0.0056 kg-mole
NO from HNO3 = 0.0075 kg-mole
H2O from HNO3 = 0.0075/2 = 0.0037 kg-mole
N2O3 from Gay-Lussac acid = • . :i IJ.T'. ,5.)->
580 X 0.00885 = 5.13 kg or 0.0675 kg-mole
N2O3 = 2N0 + iOj
O2 from N2O3 = 0.0675/2 = 0.0338 kg-mole
NO from N2O3 = 0.1350 kg-mole
Total O2 leaving = 1.58 + 0.0056 -|- 0.0338 = 1.62 kg-moles
Total NO leaving = 0.0075 + 0.1350 = 0.143 kg-mole
Total dry gases leaving . ,
SO2 = 1.21 kg-moles or X 64 = 77.5 kg
O2 = 1.62 kg-moles or X 32 = 51.8 kg
NO = 0.143 kg-mole or X 30 = 4.29 kg
N2 = 13.81 kg-moles or X 28.2 = 390.0 kg
Total 16.78 kg-moles or 523.6 kg
Total weight of dry gases leaving Glover tower = . 523.6 kg
3. Water vapor in the gases leaving the Glover tower. The weight of water vapor
in the gases leaving the Glover tower is calculated on the basis of a water balance.
Water enters as vapor in the gases, as water in the Gay-Lussac and chamber acids
and associated with the nitric acid as HNO3 and as water. The 0.230 kg-mole of
H2SO4 formed in the tower requires 0.230 kg-mole of H2O or 4.15 kg. The 78% acid
leaving requires (750) (0.22) = 165 kg water.
The 1.31 kg of 36% nitric acid charged yields upon dehydration and decomposition
0.84-1-0.07 = 0.91 kg H2O.
Water Balance: ., < -,
i Input
In gas (0.289 kg-mole) = 6.2 kg
In chamber acid (182) (0.36) = 65.5 kg
In Gay-Lussac acid 580 X 0.221 = 129.0 kg
From nitric acid = 0.9 kg
Total 200.6 kg
Output .,! ,, ^,r;•
In78%acid 165.0 kg
In formation of H2SO4 4.1 kg
In water vapor (by difference) 31.5 kg
Total 200.6 kg
Total water vapor in gases leaving = 31.5 kg
or 31.5/18 = 1.75 kg-moles
1.75
Molal humidity of gases leaving = -f— = 0.103
ID.78
Dew point of gases leaving (from Fig. 8) = 113°F or 45°C ,, • ;,. ? ^) '\
CHAP. X] ENERGY BALANCE OF GLOVEE TOWER 393
Summary of Material Balance of Glover Tower.

Input
Dry gases (16.95 kg-moles) = 536.4 kg
Water vapor (0.29 kg-mole) = 5.2 kg
Chamber acid (64% HaSO,) = 182.0 kg
Nitric acid (36% HNO3) = 1.31 kg
Gay-Lussac acid = 580.0 kg
Total = 1304.9 kg
Output
Dry gases = 523.6 kg
Water vapor = . 31.5 kg
Glover acid (78% H2SO4) = 750.0 kg
Total = 1305.1 kg
This material balance is summarized in Fig. 80. j
Gay Lussac Acid

580 Kg (777<,)
Chamber Acid
182 Kg (64%)
Nitric Acid Dry Gas 623.6 Kg ,
y Y Y
J.31 Kg (367o) SO, l.Sll Moles
0, 1.620 Moles
N, 13.810 Moles
NO 0.143 Moles
H.0 31.5 Kg
(1.75 Moles)
Dry Gas 636.4 Kg

(16.95 Kg Moles)
HJSO* 760 Kg (787o)
Water Vapor 52 Kg
(.29 Kg Mole)
Product To Gay Lussac

178Kg 572Kg
FIG. 80. Material balance of Glover tower.
ENERGY BALANCE OF GLOVER TOWER

In addition to the enthalpy of all materials entering and leaving, the input side
of the energy balance of the Glover tower includes heat evolved in the formation of
H2SO4 within the tower. This acid goes into solution but at the same time some
water leaves the solution, the net effect being one of concentration and hence requir-
ing the input of heat. In addition, heat is required to remove the oxides of nitrogen
from solution and to decompose them into NO and O2. It will be assumed that all
oxides of nitrogen, both from the Gay-Lussac and the make-up nitric acids, are de-
composed and leave the Glover tower as NO.
1. Heat evolved in formation of H2SO4. Sulfuric acid is formed from SO2 gas,
liquid H2O, and oxygen. Actually the conversion takes place in two steps with inter-
mediate formation of nitrosyl-sulfuric acid. However, the net effect is the same
as though the reaction proceeded as follows:
SOj(sf) + iOiig) + HjOCO = RSOi
iJIie + (-70,920) + 0 + (-68,320) = -193,760
Affis = -54,510 kcal / : . f »
Heat evolved in formation of HjSO, = 0.230(54,510) = 12,530 kcal
-20,000
-16,000
S
s
e
\AHJ VlH
-10,000
J
-6.000
AHj^
20 40 60 80 100
Per Cent Hj SO,
AH = Integral heat of solution of H2SO4, cal per g-mole

ABI = Diif erential heat of solution of H2O, cal per g-mole
AH2 = Differential heat of solution of H2S04, cal i>er g-mole
FIG. 81. Differential and integral heats of solution of

aqueous solutions of sulfuric acid at 20°C.
2. Heat absorbed in concentrating acid. The sulfuric acid formed in the tower
is diluted by the acid already present to form 78% acid leaving. The chamber acid
entering at 64% is ooncentrated to 78%. The Gay-Lussac acid yields aqueous sul-
furic acid containing 77/0.991 or 77.7% H2SO4. This is concentrated to 78% acid.
The net result of these changes is a concentrating effect.
The net heat effect of the concentration changes in the Glover tower may be cal-
CHAP. X] ENERGY BALANCE OF GLOVER TOWER 395
culated from integral heat of solution data by the method discussed on page 274.
The heat evolved is the difference between the total heat evolved in forming each of
the entering acid solutions from H2SO4 and H2O and that evolved in forming the solu-
tion leaving from H2SO4 and H2O. These thermal effects may be calculated from
the integral heat of solution data plotted in Fig. 81, page 394. This curve was plotted
from data of the International Critical Tables, Vol. V, with permission.
Molal integral heats of solution of H2SO4:
64% H2SO4 = . . . . - n , 8 0 0 kcal per kg-mole

77.7% H2SO4 = .. - 8,700 kcal per kg-mole
r > 78%H2S04 = . . . - 8,600 kcal per kg-mole
„ , , . , , , , 182 X 0.64
Heat of solution of chamber acid = — X —11,800 = —14,000 kcal
98
Heat of solution of Gay-Lussac acid = ——^— X —8700 = —39,700 kcal
Heat of solution of entering acids = —53,700 kcal'

750 X 0 78
Heat of solution of Glover acid leaving = ——^— X —8600 = —51,300 kcal
98
Heat absorbed in concentration of acid = 53,700 — 51,300 = -f-2400 keaJ
The net heat absorbed in concentration is 2400 calories, which should be placed on
the output side of the energy balance. The large amount of heat required for the
concentration of chamber acid is offset by the heat evolved when the HaSO* formed
in the tower is dissolved.
3. Heat absorbed in release and decomposition of N2O3 from Gay-Lussac acid.
The oxides of nitrogen enter the Glover tower as recovered nitrosyl-sulfuric acid
from the Gay-Lussac tower and as make-up nitric acid. The Gay-Lussac acid may
be considered to consist of 0.0677 kg-mole of N2O3 dissolved in 129 kg (7.15 kg-moles)
of water. This concentration corresponds to 106 moles of water per mole of N2O3.
The thermal effects in the evolution of the N2O3 from solution and its decomposition
may be calculated from the following thermochemical data:
Heat of
'initt
Heat 0/ Formation Moles
Formula State
W>IG '"ZZZUZZ"
kcal per kg-mole JZ'^L
of Water Solution,
kcal
NO 9 -1-21,600
NsOs gQ -f 20,000
-f 20,000 100 -28,900
NO2 9 -1-8,030
N2O4 9 -1-3,060 '
N2O6 9 -f600
HNO, I -41,660
-41,660 6.22 -7,000
Heat absorbed in evolution of N2O3 from solution in the Gay-Lussac acid =
<' 0.0677 X 28,900 = 1960 kcal
This entry is not exact since it has been assumed that the heat of solution of N2O3
is the same in 77.7% H2SO4 acid as it is in water, which is a poor approximation.
The N2O3 is assumed to break down entirely to NO and O2; this is also an approx-
396 CHEMICAL, METALLURGICAL, PETROLEUM PROCESSES [CHAP. X '
imation. However, in view of the relatively small heat effects involved these approx-
imations seem justified. > 1> ' • -i i •<>,1
NjOs = 2N0 + i02 ' ' '> ' '''
Affis + 20,000 = 2(21,600) .' ' / . ' , ^,"_ '
AHB = 23,200 kcal „. Vtl-V >'
H e a t absorbed in decomposition of N2O3 = 0.0677 X 23,200 = 1570 kcal
Total heat absorbed in release and decomposition of N2O3 = 1960 + 1570
= 3530 kcal
4. H e a t absorbed in decomposition of nitric acid. The nitric acid consists of

0.0075 kg-mole of HNO3 dissolved in 0.0466 kg-mole of water, corresponding to 6.22
moles of water per mole of HNO3. The acid may be considered to be separated into
its liquid components, HNO3 a n d water, a n d then decomposed according to t h e fol-
lowing reaction:
BNOAD = IH2OW + NO(g) + iOAg)
,:.!(>'iTr.r. - Aff,s+ (-41,660) = -34,160+21,600 ,',.,;•.(,-
AH,8 = 29,100 kcal
,.,.1 .lor; u\ . -.. ;!^
H e a t absorbed in separating HNO3 from solution = 0.0075 X 7000 = 52 kcal
H e a t absorbed in decomposition of HNO3 = 0.0075 (29,100) = . . . . 218 kcal
Total heat absorbed = 270 kcal
The decomposition of the oxides of nitrogen is of particular interest because of their
endothermic heats of formation.
• •
',.'J 1 1 [ 1 1 1
1 • '
__ _
- r^
Cp = Total
r> i Heat
' l T j Capacity,
in •*cal/gi
f u ./ Cn~Ttrrr _r _ ;
12 - - .Cp2 — Partial Heat Capacity of f
,sbi,cai/gr/°c.^ di _ ^__ i:^
:~ ~ - ^ CV^ :
^s__-__rr
= ;^ = = = - V-
•^••^
*N j ^
==5==:F:::::E::::::iE:
^^Pi f- J -
E 0.8 -"*^s. _ i.^ _L_- _C_
- ^r ± :i:
• ^ • ^
^5>^ CD
-^v^^ 1 -
i ~
A
-ju
:!-~>^
^ - __ _ s^ i t __-_ :- ±tI
St--^
^ • ^ ^
* — — •*-••
^™ ^ "^^
a ^'^
^^ s• ^ K
^^ Zm»
g
o - - ae Cp2 ^-*^ *"*
1
««•''
""
o.ot 1111.1.11 11.1-1,1111111 111 U 11:___: ;:_,_;___;;__;::
0. W 20, go 40 50 60 70 go 90 100
Percent H>SO
Pra. 82. Partial and total heat capacities of aqueous solutions of
sulfuric acid a t 20°Cj
V
CHAP. X] ENERGY BALANCE OF GLOVER TOWER 397
6. Heat input in burner gases. This has already been calculated as the heat
output of the burner, page 390, = 58,810 kcal
6. Enthalpy of chamber acid.
Specific heat (from Fig. 82) = 0.60
Enthalpy = 182 X 0.50 X ( 2 5 - 1 8 ) = 640 kcal
7. Enthalpy of Gay-Lussac acid.
Specific heat (Fig. 82) = 0.45
Enthalpy = 580 X 0.45(25 - 18) = 1830 kcal
This calculation neglects the effect of the oxides of nitrogen on the heat capacity
of the Gay-Lussac acid.
8. Enthalpy of nitric acid. Prom Fig. 41, page 224, the specific heat of 36%
nitric acid is 0.70.
Enthalpy = 1.31 X 0.70(25 - 18) = 7 kcal
9. Enthalpy of dry gases leaving.
SO2 (1.21) (9.76) = 11.8 ..".',..
N2 (13.81)(6.97) = 96.4 „ is,,, -.-j "
NO (0.14) (7.16) = 1.0 ! ' " •'"••"'
O2 (1.62) (7.05) = 11.4 •" \
120.6
i
Enthalpy = 120.6(91 - 18) = 8810 kcal
10. Total enthalpy of water vapor leaving Glover tower.
The water vapor leaves at 91°C having a dew point of 45°C.
Heat of vaporization at 45°C (Fig. 8) = 10,280 kcal per kg
Mean molal heat capacity of vapor between 45°C and 91°C = 8.03
Enthalpy = 1.75(10,280 + 18(45 - 18) + 8.03(91 - 45)] = 19,500 kcal
11. Heat absorbed in cooling outgoing acid.
Heat capacity (Fig. 82) = 0.44 kcal per kg per °C, neglecting the temperature
coefficient of the heat capacity.
' Heat absorbed = 750 X 0.44(125 - 25) = 33,000 kcal
12. Enthalpy of outgoing acid = 750 X 0.44(25 - 18) = 2310 kcal
Summary of Energy Balance of Glover Tower.
Input
Enthalpy of dry burner gases 54,700 kcal 74.0%
Enthalpy of water vapor from burner 4,110 kcal 5.6%
Enthalpy of chamber acid 640 kcal 0.9%
Enthalpy of Gay-Lussac acid 1,830 kcal 2.5%
Enthalpy of nitric acid 7 kcal 0.0%
Heat evolved in formation of H2SO4 12,530 kcal 17.0%
Total 73,817 kcal 100.0%
Output
Enthalpy of dry gases 8,810 kcal 11.9%
Enthalpy of water vapor in gases 19,500 kcal 26.4%
Enthalpy of acid leaving cooler 2,310 kcal 3.1%
Heat absorbed in concentration of acid 2,400 kcal 3.3%
Heat absorbed in decomposition of nitric acid.. . 270 kcal 0.4%
Heat absorbed in release and decomposition of
N2O3 from Gay-Lussac acid 3,530 kcal 4.8%
Heat absorbed by outgoing acid cooler 33,000 kcal 44.7%
Heat losses (by difference) 3,997 kcal 5.4%
Total 73,817 kcal 100.0%
This energy balance is summarized in Fig. 83.
74.0fo Sensible Enthalpy in Dry Gas
5.6T» Enthalpy of
Water Vapor
17.09!) Heat Evolved in
Formation
of HjSO*
0.9?ti Enthalpy in
Chamber Acid
2 . 5 ^ Sensible Enthalpy
in Gay Lussac Acid
Sensible Enthalpy _
0.0?& in Nitric Acid
100.0% Total
>5^_,Heat of
Concentration of H2SO4 3.3?^
Total Enthalpy in
Water Vapor 26.49!>
Release and De-
*c ' composition of No O3
from G.L. Acid 4&f,
^ Concentration and De-
composition of HNO3 0.456
1 Sensible Enthalpy
J in Dry Gas 11.9*
_ Enthalpy of Acid
leaving Cooler 3.1*
Heat Absorbed in
Cooling of Acid 44.7*
^
J Radiation 5.4*
Total 100.0*
FIG. 83. Energy balance of Glover tower.

CHAP. X] MATERIAL BALANCE OF CHAMBERS 399
It will be seen that over one-half of the available energy in the Glover tower
is absorbed by the concentrated acid. This acid must be cooled before it can be
used for absorption of gases in the Gay-Lussac tower and before storage. The cool-
ing of this acid represents one of the difficult problems in acid manufacture, in the
development of a heat interchanger which will withstand hot concentrated sulfuric
acid and at the same time permit a rapid transfer of heat.
MATERIAL BALANCE OF CHAMBERS
It would be proper to consider the material and energy balances of each chamber
separately, but to avoid needless repetition all four chambers will be considered as
a unit. In the chambers, H2O, SO2, and O2 are removed from the gases to form
HzSO*.
Basis: 100 kg of dry pyrites charged.
H2SO4 formed in chambers = 116.5 kg or 1.19 kg-moles.
Oas Entering Gas Removed Dry Gas Leaving
SO2 1.21 kg-moles 1.19 kg-moles 0.021 kg-mole
O2 1.62 kg-moles 0.595 kg-mole 1.025 kg-moles
NO 0.143 kg-mole 0.143 kg-mole
N2 13.81 kg-moles 13.81 kg-moles
H2O 1.75 kg-moles 1.19 kg-moles
Total = • 15.00 kg-moles
Water vapor leaving. The water entering with the gases and from the sprays is
used in the formation of H2S04 and dilution of the H2SO4 to form the chamber acid.
Water input
H2O from gases = 31.5 kg
H2O from sprays = 68.4 kg
Total = '. 89I9 kg
Water output
H2O to form H2SO4 = 1.19 X 18 = 21.4 kg
H2O in chamber acid = 182 X 0.36 = 65.5 kg
Output accounted for = 86.9 kg
H2O in gases leaving = 89.9 - 86.9 = .^ 3.0 kg
or 3.0/18 = *.. 0.167 kg-mole
Moles of dry gases leaving = 15.0 kg-moles
MoM humidity = 0.167/15.0 = 0.011
Dew point (Fig. 9) = 46°F or 8°C
Summary of Material Balance of Chambers.
Entering Leaving
SO2 (1.21 kg-moles) 77.5 kg SO2 (0.021 kg-mole) 1.3 kg
O2 (1.62 kg-moles) 61.8 kg O2 (1.025 kg-moles) 32.8 kg
N2 (13.81 kg-moles) 390.0 kg N2 (13.81 kg-moles) 390.0 kg
H2O (1.75 kg-moles) 31.5 kg H2O (0.168 kg-mole) 3.0 kg
NO (0.143 kg-mole) 4.3 kg NO (0.143 kg-mole) 4.3 kg
Spray water 68.4 kg Acid 182.0 kg
Total 613.5 kg Total 613.4 kg
400 C H E M I C A L , M E T A L L U R G I C A L , P E T R O L E U M PROCESSES [CHAP. X
ENERGY BALANCE OF CHAMBERS

The first reaction proceeding in the chambers is. a gaseous reaction between the
SO2, H2O, O2, and NO gases in contact with the water spray forming nitrosyl-sulfuric
acid.
2SO2 + 2NO + IJO2 + H2O -* 2NO2SO2OH
\: .ia hai'.hhk
This reaction is favored by high concentrations of SO2 and NO. The acid spray is
swept against the side walls of the chamber where the spray is condensed owing to
cooUng, and by dilution with water H2SO4 is formed with the release of the oxides
of nitrogen, .jj ^,,j ..„,, „. , „
2NO2SO2OH + H2O = 2H2SO4 + N2O3 . '^'^^s*^
N203 = 2NO + i 0 2 >lo%:\»n -M-^^'S -.
The first of these reactions proceeds in the liquid phase and is favored by a high con-
centration of water brought in by the spray and condensed upon the side walls.
Water Sprayed 58.4 Kg

O/i.
Dry Gas 523.6 Kg r ();H

SO, 1.211 Kg Moles
O, 1.620 Kg Moles Dry Gas 428.4 Kg
N, 13.810 Kjr Moles SO2 0.021 X; Mole
NO 0.143 KE Mole Oj 1.025 Kg Mole
Water Vapor N, 13.810 Kg Moles r!.:/i ntj' OS
31.5 Kg NO O.I4SKgMole
1.7S Kx Moles ^
Water Vapor 3.0 Kg tOrfl
0.167 Kg Mole
Acid 182.0 Kg
F I G . 84. Material balance of chambers.
In calculating the heat evolved in the chamber reactions only the final, net effects
need be considered. It will be assumed t h a t the oxides of nitrogen ultimately leave
the chambers in the same form in which they entered, NO. This assumption is not
exact because some oxidation of NO to N2O3 probably takes place at the relatively
low temperatures of the last chamber. The ultimate effects of the reactions in the
chambers are then the production of H2SO4 from SO2, O2, and H2O and the dissolu-
tion of this acid to form an aqueous solution containing 6 4 % H2SO4.
Reference temperature: 18°C.

T!'••!«(•<
1. H e a t evolved in formation of H2SO1. From item 1 of the energy balance of the

Glover tower, the heat evolved is 54,510 kcal per kg-mole of II2SO4 formed.
H2SO4 formed in chambers = 1.19 kg-moles
Heat evolved = 1.19 X 54,510 = 64,870 kcal
2. H e a t evolved in dissolving H2SO4. Integral heat of solution (Fig. 81) a t a con-

centration of 6 4 % H2SO4 = 11,800 kcal per kg-mole.
Heat evolved in dissolution = 11,800 X 1.19 = 14,050 kcal
jog If
CHAP. X] ENERGY BALANCE OF CHAMBERS 401
3. Enthalpy of dry gases and water vapor entering. This has already been cal-
culated as part of the heat output of the Glover tower = 28,310 kcal
4. Enthalpy of spray water. Since the spray water enters at the reference tem-
perature, 18°C, its enthalpy is equal to 0 kcal
5. Enthalpy of dry gases leaving. i
SO2 (0.021)0.49) = 0.2
O2 (1.025) (7.02) = 7.2
'" N2 (13.81) (6.97) = 96.4 . .„;
• ', NO (0.14)(7.16) = 1.0 'ôâte
« I Total = 104.8 ,_.
Enthalpy of dry gases = (104.8) (40 - 18) = 2310 kcal
6. Enthalpy of water vapor leaving. i '
Dew point = 8°C.
Heat of vaporization at 8''C (Fig. 8) = 10,640 kcal per kg-mole.
Total enthalpy =
0.167[10,640 - 18(18 - 8) + 8.01(40 - 8)] = 1790 kcal
7. Heat absorbed in cooling the acid leaving. Heat capacity (Fig. 82) of acid
containing 64% H2SO4 = 0.50 kcal per kg per °C.
Enthalpy = 0.50 X 182(68 - 25) = 3920 kcal
8. Enthalpy of acid leaving the cooler =
0.50 X 182(25 - 18) = 640 kcal
Sununary of Energy Balance of Chambers. }
Inrivi
^ -l ;w ... •
Enthalpy of dry gases from Glover to*er 8,810 kcal 8.2%
Enthalpy of water vapor in gases entering 19,500 kcaJ 18.2%
Enthalpy of spray water 0 kcal 0%
Heat evolved in forming H2SO4 64,870 kcal 60.5%
Heat evolved in dissolving HzSOj 14,050 kcal 13.1 %
Total 107,230 kcal 100.0%
Ouiyvii
Enthalpy of dry gases leaving 2,310 kcal 2.2%
Enthalpy of water vapor in gases 1,790 kcal 1.7%
Enthalpy of acid leaving the cooler 640 kcal 0.6%
Heat absorbed by cooler 3,920 kcal 3.7%
Heat loss from chambers (by difference) 98,570 kcal 91.8%
Total 107,230 kcal 100.0%
This energy balance is presented diagrammatically in Fig. 85.
It will be seen that nearly the entire source of energy (72.3%) comes from the
formation of H2SO4 and its dilution whereas nearly the entire energy input is lost by
radiation from the extensive lead surfaces of the chambers. More recent develop-
ments in the chamber process have provided for more rapid means of heat removal
with much less floor space and size of equipment' by rapid circulation of both gasea
and acid in packed towers.
60.6^ Formation of HjSO^
&2» Enthalpy oJ |
Dry Gaa '
13.1^ Dissolution
of H , S O ,
18.2?> Enthalpy of
Water Vapor
I00.0?6 Total
sV , Enthalpy of
Dry Ga«e» 2.2*
. Enthalpy of
' Water Vapor 1.7*
1 Heat Removed in
Cooling A d d 3.7*
^^ . Enthalpy of A d d
after Cooling 0.6*
Radiation Loss 913*

Total 100.0*
FIG. 85. Energy balance of chambers.
MATERIAL BALANCE OF GAY-LUSSAC TOWER ;

The purpose of the Gay-Lussac tower is to recover the oxides of nitrogen from the
chamber gases. These oxides are then returned to the system in the Glover tower.
The water remaining in the chamber gases is also absorbed. The conditions favor-
. able for absorption of the oxides of nitrogen are a high concentration of acid, a low
temperature, and a low concentration of SO2 in the residual gas. A high water con-
tent in the gas from the chambers causes objection.able dilution of the acid. Small
amounts of SO2 will cause decomposition of nitrosyl-sulfuric acid with release and
loss of the oxides of nitrogen. The presence of oxygen is essential to effect the oxida-
tion of NO to N2O3 and its absorption in the acid.
The loss of oxides of nitrogen in the gases from the Gay-Lussac tower may be
assumed to be equivalent to the make-up nitric acid introduced. It will be assimied
that these oxides leave in the form of NO.
1. Input. The input to the Gay-Lussac tower consists of the wet gases from the
chambers and the Glover acid which is introduced. All the gases pass through the
tower unchanged with the exception of the NO and O2.
2. NO in gas leaving = 0.008 kg-mole or = 0.2 kg
CHAP. X] ENERGY BALANCE OP GAY-LUSSAC TOWER 403
O2 in gases leaving:
NO oxidized to N2O3 = 0.143 - 0.008 = 0.135 kg-mole
O2 consumed = 0.135/4 = 0.034 kg-moIe
O2 leaving = 1.025 - 0.034 = 0.991 kg-mole = . 31.7 kg
Summary of Material Balance of Gay-Lussac Tower.
Input Output
SO2 (0.021 kg-mole) 1.3 kg SO2 (0.021 kg-mole) 1.3 kg
O2 (1.025 kg-moles) 32.8 kg O2 (0.991 kg-mole) 31.7 kg
N2 (13.81 kg-moles) 390.0 kg N2 (13.81 kg-moles) 390.0 kg
NO (0.143 kg-mole) 4.3 kg Acid leaving 580.0 kg
H2O (0.167 kg-mole) 3.0 kg NO (0.008 kg-mole) 0.2 kg
Glover acid 572.0 kg Total 1003.2 kg
Total 1003.4 kg
This material balance is simimarized diagrammatically in Fig. 86.
Glover Acid 572 Kg
Dry Ga8 423 K g ^

m
SO, 0.021 Kg Mole
Ot 0.991 Kg Mole
Ni 13.810 Kg Moles
KO 0.008 Kg Mole
DiyCas 428.4 Kg
SOt 0.021 ]& Moles
Oi 1.025 Kl Moles
N, 13.810 E : Moles
NO 0.143 Eg Moles
Water Vapor 3.0 Kg
0.168 Kg Moles Acid SeOKg
FIG. 86. Material balance of Gay-Lussac tower.
ENERGY BALANCE OF GAY-LUSSAC TOWER

1. Heat evolved in forming and dissolving N2O3. The N2O3 released from the
Gay-Lussac acid and decomposed in the Glover tower is reformed and recovered in
the Gay-Lussac tower evolving 3530 kcal calculated as part of the energy balance of
the Glover tower, page 398. ,;
2. Heat evolved in the dissolution of the water vapor absorbed. The water vapor
leaving the last chamber is completely absorbedin the Gay-Lussac tower. Since the •
resulting concentration change in this absorption is negUgible, it is necessary to
determine the heat evolved in the dissolution of the water from data on the differen-
tial heat of solution of water in sulfuric acid solutions. From Fig. 81, the differential
molal heat of solution of water in a sulfuric acid solution containing 77.7% H2SO4
is 3100 kcal per kg-mole. This value neglects the effect of the dissolved oxides of
nitrogen.
Heat evolved = 3100 X 0.167 = 620 kcal
3. Enthalpy of Glover acid introduced.
Heat capacity (Fig. 82) = 0.44 kcal per kg per "Q.
Enthalpy = 572 X 0.44(25 - 18) = 1760 kcal
4. Enthalpy of entering gases. Already calciilated as output items in the energy

balance of the chambers = , 4100 kcal
5. Enthalpy of gases leaving.
SO2 (0.021) (9.44) = 0.2
,,\^ O2 (0.99) (7.01) = 6.9
h ^ „ao Sir-
' " '' N2" (13.81)(6.96) = 96.1
NO (0.01) (7.16) = 0.0 -!
Enthalpy = (103.2) ( 3 0 - 1 8 ) = .'.; . . . . 1240 kcal

6. Heat absorbed in cooling acid.
580 X 0.45(27 - 25) = 620 kcal
BJS^ Heat Evolred in Solueion of H,0 tH^:
35.7% Heat Evolved in

Formation and
Dissolution of H2O3
I8.05l> Enthalpy of
HjO Vapor
I 4
t m Sensible Enthalpy ">

in Glover Acid
23.3% Sensible Enthalpy ">

in Dry Gases
100.0% Total
'K !'••>>/'( ts*E -I
Sensible Enthalpy to
^ Gases Leaving 12.6%
1 Heat Removed in
C CooUng Acid 6.2%
Enthalpy of Acid
after Cooling 18.6%
-Radiation Loss 635%

Total 100.0%
FIG. 87. Energy balance of Gay-Lussac tower.
7. Enthalpy of acid leaving cooler.

580 X 0.45(25 - 18) = 1830 kcal
CHAP. X] SUMMARIZED ENERGY BALANCE 405
Summary of Energy Balance of Gay-Lussac Tower.

Input
Enthalpy of dry gases entering = 2310 kcal 23.3%
Enthalpy of water vapor entering = 1790 kcal 18.1%
Enthalpy of Glover acid entering = 1760 kcal 17.7%
Heat evolved in forming N2O3 = 1570 kcal 15.9%
Heat evolved in dissolving N2O3 = 1960 kcal 19.8%
Heat evolved in dissolving H2O = 520 kcal S.2%
Total = 9910 kcal 100.0%
Out-put
Enthalpy of gases leaving = 1240 kcal 12.5%
Heat absorbed in cooling acid = 520 kcal 5-2%
Enthalpy of acid leaving cooler = 1830 kcal 18.5%
Heat loss (by difference) = 6320 kcal 63.8%
Total = 9910 kcal 100.0%
This energy balance is shown diagrammatically in Fig. 87.
SUMMARIZED ENERGY BALANCE FOR ENTIRE PLANT

'•" i ' 1 Input
1. Heat evolved in combustion of pyrites = 148,200 kcal 61.2%
2. Enthalpy of water vapor in air = 1,850 kcal 0.8%
3. Heat evolved in formation of H2SO4 in Glover tower 12,530 kcal 5-2%
4. Sensible enthalpy of nitric acid 7 kcal 0.0%
5. Heat evolved in formation of H2SO4 in chamber... 64,870 kcal 26.8%
6. Heat evolved in solution of H2SO4 in chamber 14,050 kcal 5.8%
7. Heat evolved in solution of H2O in Gay-Lussac tower 520 kcal 0.2%
Total 242,027 kcal 100.0%
Outpitt
1. Enthalpy of cinder 4,900 kcal 2.0%
2. Net heat absorbed in concentrating acid in Glover
tower 2,400 kcal 1.0%'
3. Concentration and decomposition of nitric acid. . . . 270 kcal 0.1%
4. Radiation losses from burners 86,340 kcal 36.7%
Radiation losses from Glover tower 3,997 kcal 1-7%
Radiation losses from chamber 98,570 kcal 40.8%
Radiation losses from Gay-Lussac tower 6,320 kcal 2.6%
5. Cooling of Glover acid 33,000 kcal 13.6%
Cooling of chamber acid 3,920 kcal 1.6%
Cooling of Gay-Lussac acid 520 kcal 0.2%
6. Enthalpy of acid product 178(0.44)(25-18) 650 kcal 0.2%
7. Enthalpy of spent gases leaving Gay-Lussac tower 1,240 kcal 0.6%
Total 242,027 kcal 100.0%
The energy balance for the entire sulfuric acid plant is diown diagrammaticaJly
in Fig. 88.
61.2% Combustion of Pyrites
5.2% Formation of H2SO4 _
in Glover Tower
26.8% Formation of
H2S04in Chambers
6.8% Solution of HjSO,
in Chambers ^
0.2% Solution of H2SO in

Gay Lussac Tower *
0.8% Enthalpy of
Water Vapor
100.0% Total Enthalpy of Product Acid 0.2%
Sensible Enthalpy in
Cinders 2.0%
, Concentration of Acid
in Glover Tower 1.0% y*^*^
3onc. and Decomposition

of Nitric Acid in Glover 0.1%
Radiation from Burner

and Dust Chambers 35.7%
^ Radiation loss from

Glover Tower 1.7%
Radiation from Chambers 40.8%
•ii- Radiation from ^

Gay Lussac Tower 2.6%
) Cooling Glover Acid 13.6*
' Cooling Chamber Acid 1.6%

- Cooling Gay Lussac Acid 0.2%
Enthalpy of Gases Leaving Gay Lussac 0.5%
.' , , .: Total 100.0%
EiG. 88. Energy balance of an entire sulfuric acid plant.
/
THE MATERIAL AND ENERGY BALANCES OF A BLAST FURNACE
A blast furnace is essentially a huge gas producer where, in conjunc-
tion with the partial combustion and distillation of a carbonaceous fuel,
the reduction of ore and the formation of slag occur simultaneously.
The charge, consisting of iron ore, coke, and Hmestone in proper propor-
tions, is fed into the top of the blast furnace. Preheated air, preferably
free from water vapor, is blown through the tuyeres near the bottom of
the furnace into the descending stream of solids. This results in com-
bustion of the coke to carbon dioxide. The carbon dioxide gas in the
presence of excess coke is reduced at the high prevailing temperature to
CHAP. X] BALANCES OF A BLAST FURNACE 407 "
carboni monoxide. A great many chemical reactions occur within the

furnace. As the charge descends the shaft and as its temperature is
gradually increased, dehydration of the ore, coke, and limestone takes
place, followed by distillation of the remaining volatile matter of the
coke, calcination of magnesium and calcium carbonates present in the
limestone or ore, and reduction of the higher oxides of manganese and
iron to manganous and ferrous oxides by the rising stream of reducing
gases. The carbon dioxide formed by reduction of the ore with carbon
monoxide is reduced in the presence of excess coke. As the temperature
of the descending charge becomes still higher the lower oxides of iron
and manganese are reduced to the metallic state.
At the highest temperature of the tuyeres, part of the silica present
is reduced to the metallic state and is dissolved in the molten iron. The
excess sihca and alumina of the charge are fluxed by reaction with the
metallic bases present resulting in the formation of a fusible slag con-
sisting of complex sihcates and aluminates of calcium, magnesium, and
iron. The high temperature at the tuyeres produces a fluid slag and
molten metal which readily flow through the sohd reacting charge, sep-
arate into two layers at the bottom of the furnace, and are periodically
run out in two separate streams as molten pig iron and as molten slag.
A high temperature at the tuyeres favors a ready separation of the slag
and removal, as CaS in the slag, of much sulfur which was originally
present in the coke. The temperature of the blast furnace is too low
and insufficient coke is present to reduce the oxides of calcium, mag-
nesium, and aluminum and the silicates. Hence, these compounds pass
into the slag. Silica is used as a flux in a few exceptional cases where
the alkaline earths and alumina predominate in the gangue present in
the ore.
The purpose of preheating the air used in combustion of the coke is
to permit the attainment of the high temperatures necessary for the final
reduction of the ore and the fusion of the pig iron and slag. Any water
vapor present in the incoming air will lower the temperature in the
fusion zone on account of the heat absorbed in its reduction to hydrogen
and carbon monoxide. For this reason it is desirable to use a blast of
dried air.
The products of the blast furnace consist of molten pig iron, slag, and
blast-furnace gas. The outgoing gas consists essentially of nitrogen,
carbon monoxide, carbon dioxide, and water vapor, with small amounts
of hydrogen and methane, and also carries in suspension a considerable
amount of dust. The heating value of this gas is very low because of
its high content of nitrogen and the small amount of volatile matter in
the coke which is used for reduction. The free moisture in the incoming
charge is distilled off near the top of the furnace and escapes into the
blast furnace gas without reduction. ^
The slag contains all the lime, magnesia, alumina, and alkalies origi-
nally present in the ore and flux, together with most of the silica and some
ferrous and manganous oxides. The exact mineralogical compositions
of ore, flux, and slag are usually not known completely, so that some
uncertainty exists concerning the exact thermal energy involved in reduc-
tion and chemical transformations. The molten pig iron contains, in
addition to iron, some carbon present as cementite and lesser amounts
of silicon and manganese, cooling, the cementite partly decomposes
into graphite and iron.
In order to establish the energy balance of a blast furnace it is neces-
sary to know the masses and chemical compositions of the ore, flux,
dust, and pig iron, and the analysis of the dry blast-furnace gas. The
masses of slag, air, and water vapor can then be calculated.
The material balance includes: • • -i
Input . Output
Iron ore , Dry gases ,
Flux , I . , ,. Water vapor in gases ,< ,.
Coke , ,, Pig iron -.i
,;.,;,;,. pfMr . ,; ,^j Dust ; ^. , , . ;
,,,; ,j Water vapor Slag
As an illustration of the calculations involved in the material and
energy balance of a blast furnace, the data for the reduction of a basic
iron ore with charcoal and an acid flux will be given. An example of the
more usual operation with a limestone flux is given in the problems at the
end of this chapter. The balances are worked out on a basis of 100 kilo-
grams of pig iron produced.
Illustration 2. A blast furnace is charged with 212.7 kg of ore, 110.0 kg of char-
coal, and 13.9 kg of flux per 100 kg of pig iron produced. The compositions of
these materials are as follows:
Ore (212.7 kg) Charcoal (110.0 kg) .^
FesOs 54.93% C 86.89%
/:
FeO 8.48% 0 3.13%,
CaO 9.58% H 0.45%
MnsOi 4.97% N 0.51%
AI2O3 3.00% H20 7.00%
MgO 1.83% Ash 2.00%
SiOj 4.92% 100.00%
H20 4.48%
COj 7.81% /
100.00% • ; '
CHAP. X] MATERIAL BALANCE OF BLAST FURNACE 409
Flux (13.9 kg) Pig Iron (100.0 kg)

SiOj 78.38% C 3.12%
AI2O3 13.99% Si 1.52%
CaO 0.53% Mn 2.22%
PejOa 3.90% Fe 93.14%
H20 3.20% 100.00%
100.00%
The total heating value of the charcoal is 7035 kcal per kg.
The clean gas produced has the following composition by volume on the moisture
free basis:
CO2 ..?.... 12.62%
CO • ; 25.56%
CHi 0.69%
•' ' ' Hs 1.34%
N2 59.79%
:'!•»'• • • ' 100.00% -• = ' - < -
The ore, flux, and charcoal are charged to the furnace at an average temperature
of 18°C. The air blast is dried and enters the tuyeres at a temperature of 300°C
and moisture-free.
The gases leave the furnace at a temperature of 173°C and contain only negligible
quantities of dust.
The slag and pig iron are poured at an average temperature of 1360°C.
In order to cool the outside of the bosh of the furnace and thereby protect the re-
fractories frorn excessive heating, water is circulated in a pipe passing around the
circumference of the bosh. On the basis of 100 kg of pig iron produced 576 kg of
water are circulated and heated through a temperature rise of 13°C.
Calculate the complete material and energy balances of this furnace.
..,.!..,.i..v-J t " i (V - MATERIAL BALANCE -) v f S'.r<?'r,'«",-, -s: .i'Kft
Distribution of Ore Materials. The mineralogical composition of the ore is not

given, although it is customary to assume that the carbon dioxide present is combined
with the hme and the magnesia present as hmestone. This hmestone is calcined at
about 900°C to CaO, MgO, and CO2. The silica is present chiefiy as silicates of
aluminum and magnesium. Any free moisture present in the ore is driven off at
the top of the furnace without reduction. However, the chemically combined water,
as in the minerals kaohnite and limonite, will be retained until the ore reaches a hotter
zone, where it will be partly reduced by coke to hydrogen and carbon monoxide.
It will be assumed that all the elements of the carbon dioxide and water of the ore
leave the blast furnace in the gases.
The alumina, lime, and magnesia pass into the slag without reduction but combine
with the silica of the flux to form complex silicates of calcium, magnesium, and alumi-
num. The exact thermal energy of this latter change is unknown since the miner-
alogical composition of the slag as well as the heats of formation of complex silicates
are unknown.
The higher oxides of iron are reduced at a relatively low temperature to the lower
oxides. In the hot zone of the furnace the oxides are further reduced to the metallic
state to supply the iron requirements of the pig iron. It is assumed that the remain-
ing iron, as ferrous oxide, passes into the slag forming ferrous silicate. This assump-
tion is somewhat inexact because it is kn/ ym that a part of the iron exists as metaUic
particles included in the slag.
Sufficient siUca is reduced to metallic silicon to supply the siHcon content of the
pig iron. The remainder combines with the basic oxides to form silicates in the slag.
The MnaOi of the ore is in part reduced to metaUio manganese, supplying that pres-
ent in the pig iron. The remainder of the manganese is assumed to enter the slag
as MnO, forming silicates.
The oxygen given up in the reduction of the oxides of iron, silicon, and manganese
will be present in the gases as CO, CO2 or II2O. The gases also contain the CO2
and H2O of the ore.
Basis: 100 kg of pig iron produced. /
Distribution of Fe203 and FeO: •
1 1 fi s? •
FejOjinore = 212.7 X 0.5493 = 116.8kgor—-^ = .. 0.732 kg-mole
159.7
18.03 rt-
FeO in ore = 212.7 X 0.0848 = 18.03 kg or •——- = .. 0.2516 kg-mole ; %
71.8 y*-'*'
Total Fe in ore = 2 X 0.732 + 0.2515 = 1.7155 kg-atoms
Fe in pig iron = 93.14 kg or —'-— = 1.669 kg-atoms
55.8
Fe into slag as FeO = 0.0465 kg-atom
FeO into slag = 0.0465 kg-mole or 0.0465 X 71.8 = .. 3.33 kg
Oxygen in iron oxides of ore = (3/2 X 0.732)
+ (i X 0.2515) = 1.223 kg-moles
Oxygen in FeO of slag = ^ X 0.0465 = 0.023 kg-mole
Oxygen into gases = 1.20 kg-moles
or 1 . 2 0 X 3 2 = 38.4 kg
Distribution of Si02.' ' '-
SiOs in ore = 212.7 X 0.0492 = 10.47 kg o r — ^ = . . . 0.174 kg-mole '
,60.1
1.52 /
Si in pig iron = 1.52 kg or ~- = 0.054 kg-atom /
Si into slag as SiOa = , 0.120 kg-atom /
SiOj into slag = 0.120 kg-mole or 0.120 X 60.1 = 7.2 kg /
O2 into gases = 0.054 kg-mole or 0.054 X 32 = . . . , . . . 1.7 kg '
Distribution of MusOi:
MnaOi in ore = 21,2.7 X 0.0497 = 10.57 kg or — ^ = 0,0462 kg-mole

228.8
Mn in ore = 0.0462 X 3 = 0.1386 kg-atom /
2.22 /'
Mn into pig iron = 2.22 kg or •—— = 0.0405 kg-atom ;
54.9 /
Mn into slag = 0.0981 kg-atom
MnO into slag = 0.0981 kg-mole or 0.0981 X 70.9 = . . 6.95 kg ,;:
Oxygen in MnsOi of ore = 2 X 0.0462 = 0.0924 kg-mole
Oxygen in MnO of slag = § X 0.0981 = 0.0491 kg-mole :
Oxygen into gases = 0.0433 kg-mole
or 0.0433 X 32 = . 1.38 kg
Distribution of H2O and CO2:

Both compounds enter the gases in partly reduced forms.
9 52
H2O in ore = 212.7 X 0.0448 = 9.52 kg or - ^ — = . . . . 0.529 kg-mole
18.02
CO2 in ore = 212.7 X 0.0781 = 16.62 kg or —j— = . . . 0.378 kg-mole
44
H2 into gases = 0.629 kg-mole or 0.529 X 2.02 = 1.06 kg
C into gases = 0.378 kg-atom or 0.378 X 12 = 4.54 kg
Total O2 into gases = 0.378 + (i X 0.529) = 0.642 kg-
mole or 20.54 kg
Summary of distribution of ore materials:
Into pig iron:
Fe = 93.14 kg
Mn = 2.22 kg
Si = 1.52 kg
Into slag: -
Feb = T 3.33 kg
SiOz = 7.2 kg
MnO = 6.96 kg
CaO = 212.7 X 0.0958 = 20.4 kg
AI2O3 = 212.7 X 0.0300 = 6.4 kg
MgO = 212.7 X 0.0183 = 3.9 kg
Into gases:
O = 38.4 -t- 1.7 + 1.38 + 20.64 = . . . . 62.0 kg
C = 4.54 kg
H = 1.06 kg
1 Total = '212.6 kg
Distribution of Flux Materials. Since it has been assumed in the preceding calcu-
lations that all the iron, manganese, and silicon of the pig iron were derived from the
ore materials, all the materials of the flux must pass into the slag or the gases. The
silica, lime, and alumina enter the slag unchanged. It will be assumed that the ferric
oxide is reduced to FeO and enters the slag in this form. The oxygen evolved in the
reduction of the iron oxide and the water present in the flux enter the gases.
Basis: 100 kg of pig iron produced.
Distribution of iron oxide:
FeaOs in flux = 13.9 X 0.0390 = 0.542 kg or 0.0034 kg-mole
FeO into slag = 2 X 0.0034 = 0.0068 kg-mole or 0.49 kg
O2 into gases = i X 0.0034 = 0.0017 kg-mole or 0.05 kg
Distribution of water:
H2O in flux = 13.9 X 0.0320 = 0.445 kg or 0.0247 kg-mole
H2 into gases = 0.0247 kg-mole or 0.05 kg
O2 into gases = i X 0.0247 = 0.0123 kg-mole or 0.39 kg
Summary of distribution of flux materials:

Into slag: '• '•!•!"
SiOj = 13.9 X 0.7838 = 10.90 kg
AI2O3 = 13.9 X 0.1399 = 1.95 kg '
CaO = 13.9 X 0.0053 = 0.07 kg
FeO = 0.49 kg . ,0 )
Into gases: ' u^, £,| ,
0 = 0.05+0.39= 0.44 kg ,. ,;\-v
H= 0.05 kg ,"', „
' Total = 13.90 kg '-'.''•'
Distribution of Charcoal. The carbon in the pig iron will be assumed to be derived
from the charcoal. The ash enters the slag while the remainder of the charcoal
constituents passes into the gases.
Distribution of water. ' •' :• - ' •" v s . .;».,
H2O in charcoal = IIO.O X 0.0700 = 7.70 kg or 0.428 fcg-mole '**""
H2 into gases = 0.428 kg-mole or 0.86 kg
O2 into gases = 0.214 kg-mole or 6.85 kg
Summary of distributions of charcoal materials. i . '
Into pig iron:
C = 3.12 kg
Into gases: M- > •
C = (110.0 X 0.8689) - 3.12 = . . . . " . . 92.44 kg •, ' "
H = (110.0 X 0.0045) + 0.86 = 1.36 kg
O = (110.0 X 0.0315) + 6.85 = 10.32 kg
N = 110.0 X 0.0051 = 0.56 kg
! ' • • ' ' ' . •
Into slag: I • » = Tf
Ash = 110.0 X 0.02 = 2.20 kg •
Total = 110,00 kg
Weight and Composition of Slag. Since it may be assumed that the slag contains
only materials derived from the ore, flux, and charcoal, its total weight and compo-
sition may be obtained by adding together the weights of materials entering the slag
from these three sources. Actually these materials are in the slag in the form of
complex compounds.
Component Weight, kg Percentage Kg equivalents
FeO = 3.33 -f-0.49 =... . 3.82 6.0 0.106
SiOj = 7.2 + 10.90 = . . . 18.10 28.4 , , 0.602
MnO = 6.95 10.9 .,, 0.196 /
CaO = 20.4 + 0.07 = . . . 20.47 32.1 ,, 0.730
AI2O3 = 6.4 -1- 1.95 =... . 8.35 13.1 0.246
MgO = 3.90 6.1 * • 0.194
Ash = . 2.20 3.4
Total = 63.79 kg 100.0%
In calculating the kilogram equivalents, alumina is assumed to behave as an acid.

Neglecting the charcoal ash:
Total kilogram equivalents of metallic bases = 1.226
Total kilogram equivalents of acids (SiOz and AI2O3) = 0.848
It is apparent that the slag is distinctly basic in character, despite the fact that an
acid flux was used.
In order to verify the above calculations i.t is possible to determine the weight and
composition of the slag experimentally.
Weight of Dry, Clean Blast Furnace Gas. The direct measurement of the volume
of blast furnace gas is difficult and rarely undertaken. The volume or weight of gases
can be calculated from a carbon balance as in obtaining the material balance of a
combustion process. The carbon in the blast furnace gas comes from the carbon
dioxide of the limestone and the ore and from the charcoal used for reduction.
Carbon Balance. Basis: 100 kg of pig iron produced.
Total C into gases = 4.64 + 92.44 = 96.98 kg or 8.08 kg-atoms
Carbon per kg-mole of dry, clean gas (from gas analysis) =
0.1262 + 0.2556 + 0.0069 = 0.3887 kg-atom
Total dry gas = 8.08/0.3887 = ., 20.78 kg-moles
CO2 = 20.78 X 0.1262 = 2.62 kg-moles or X44 = . .. 115.2 kg
CO = 20.78 X 0.2556 = 5.30 kg-moles or X28 = . . . 148.3 kg
CH4 = 20.78 X 0.0069 = 0.14 kg-mole or X16 = ... 2.2 kg
Hj = 20.78 X 0.0134 = 0.28 kg-mole or X2.02 = . 0.57 kg
N2 = 20.78 X 0.5979 = 12.44 kg-moles or X28.2 = 351.5 kg
Total = 20.78 kg-moles or 617.8 kg
Average molecular weight = 617.8/20.78 = 29.7
Weight of Air Introduced. The weight of dry air introduced is calculated from a
nitrogen balance as in a combustion process. The nitrogen in the gas is derived from
only two sources, the charcoal and the air.
N2 in gases = 351.5 kg
N2 from charcoal = 0.56 kg
N2 from air = 350.9 kg or 12.42 kg-moles
Air introduced = 12.42/0.79 = 15.72 kg-moles
or 15.72 X 29.0 = 456 kg
Since the air is dried, containing no water vapor, this is the total weight of the air
introduced.
Weight of Water Vapor in Blast Furnace Gas. The weight of water vapor in the
blast furnace gas is calculated from a hydrogen balance. All hydrogen of the enter-
ing materials is present in the gas either as H2, CH4, or H2O.
Basis: 100 kg of pig iron produced.
Total H introduced = 1.06 + 0.05 -|- 1.36 = 2.47 kg
Output of H:
As H2 = 0.57 kg
As CH, = 2 X 0.14 X 2.02 = 0.57 kg
"' ';' Total = 1.14 kg
^\.. , As H2O = 2.47 - 1.14 = 1.33 kg or 0.66 kg-mole
HjO in gases = 0.66 kg-mole or 11.9 kg

H2O introduced in charge = 9.52 + 0.44 + 7.70 = 17.66 kg
H2O decomposed in blast furnace = 17.66 — 11.9 = 5.76 kg
or 5.76/18 = 0.320 kg-mole
Overall Material Balance.

Input Output
Ore 212.7 kg Slag 63.8 kg
Plux 13.9 kg Dry gas (20.78 kg-
Charcoal 110.0 kg moles) 617.8 kg
Air (15.72 kg-moles) 456.0 kg Water vapor in gas
Total 792.6 kg (0.66 kg-mole). . 11.9 kg
Pig iron 100.0 kg
Total 793.5 kg
. y^^
Charcoal llO^) Kg
Flux 13.9 Kg 1
Or« 212.7 Kg Dry Gas 618 Kg
(20.7« Moles) .
Air 456 Kg
(16.72 Moles)
FIG. 89. Material balance of a blast furnace.
ENERGY BALANCE /
An energy balance might be estabUshed by considering the enthalpies and heats
of formation of all components of the charge and all components of the slag, pig
iron, and furnace gas together with the heat loss by radiation. Such an energy bal-
ance would be disproportionate since the chemical energies of formation of the oxides
and silicates which pass through the process unchanged, such as the oxides of alumi-
num, siUcon, calcium, and magnesium, are of no interest. More valuable information
is obtained by including in the energy balance only the net chemical and thermal
changes which take place during the process.
During the course of reduction many intermediate chemical reactions take place
each accompanied by a certain thermal change. Examples are the progressive reduc-
tion of the higher oxides of iron and manganese, the oxidation of carbon at the tuyeres
and its subsequent reduction by coke to carbon monoxide, the reduction of metallic
CHAP. X] ENERGY BALANCE OF BLAST FURNACE 415
oxides by carbon and by carbon monoxide, and the reduction of water to carbon
monoxide, hydrogen, and carbon dioxide. However, in any chemical process the
total change in energy is dependent only upon the initial and final states of chemical
constitution, temperature, pressure, and state of aggregation and is independent of
any intermediate state. Hence, in calculating the energy balance of a blast furnace,
the numerous intermediate reactions involved need not be considered. It is sufficient
to know the temperature, state of aggregation, and composition of each material
charged and each product formed, without knowing how the various components of
the products are actually produced.
The oxides of calcium, magnesium, and aluminum pass through the furnace appar-
ently unchanged so that the heats of formation of these oxides need not be considered,
but the state of the oxide is much different in the slag from that in the ore or flux.
For example, in the ore, the oxides of iron and manganese exist as oxides but in the
slag as silicates, so that the net heat effect accompanjdng the formation of sili-
cates from the oxides should be considered. However, accurate calculation of this
quantity requires data which are not ordinarily available.
The energy balance is calculated with a reference temperature of 18°C, based on
100 kg of pig iron produced.
1. Heat absorbed in reduction of iron oxides at 18°C. The heat absorbed is
obtained by subtracting the total heat of formation of the reactants from that of the
products. Iron oxides enter the process in both the ore and the flux. The neces-
sary heat of formation data are obtained from Table XIV, page 253.
nr 1 J IT J t Total Heal of
„ ^ , v , Molal Heat of r, .
Reactants Kg-moles _ ,. Formation,
Formation , , '
kcal
FejOa 0.735 -198,500 -145,900
FeO 0.2515 - 64,300 - 16,170
Total = -162,070 kcal
Products
i Fe 1.669 0
FeO 0.0533 -64,300 -3,430
Oxygen 1.25 0
Heat absorbed in reduction of iron oxides =
-3,430 - (-162,070) = 4-168,640 kcal
2. Heat absorbed in reduction of MnsOt at 18°C.

!<• 7 7 i r i /• Total Heat of
Beactant Kg-moles „ Heat
Molal .. of Formation,
j? , ,-, '
Formation
kcal
Mn,04 0.0462 345,000 -15,940 kcal
Products
Mn 0.0405 0
MnO 0.0981 -96,500 -9470 kcal
Oxygen 0.0433 ' 0
Heat absorbed in reduction of MnsO, = -9470 -f 15,940 = 4-6,470 kcal
416 C H E M I C A L , METALLURGICAL, P E T R O L E U M PROCESSES [CHAP. X
3. H e a t absorbed in reduction of SiOj at 18°C. !'< . * vd J .u; sj.ti.ivs . m

>T . 7 rr 1 t Total Heal of ^'
r, . . IT , Molal Heat of „ ^. •'
Heactant ' Kg-moles _ ^. Formation,
Formation , , ' , i>
kcal
Si02 reduced 0.054 -203,340 - 1 0 , 9 8 0 kcal '
Products Kg-atom 'i '' >'
Si 0.05* -".'"i-tlwT '0 , .It fi,
O 0.054 0 > I ''
Heat evolved in reduction of SiOj = ! +10,980 kcal
4. H e a t absorbed in calcination of carbonates at 18°C. Carbonates are present

only in the ore. I t is assumed that all CaO of the ore is combined with CO2 as the
carbonate and t h a t the surplus CO2 is combined with MgO. .;; -.nU uni r :>
CO2 in ore = 0.378 kg-mole ' ' ,,
CaO in ore = 20.4/56 = 0.364 kg-mole ' .iff.)
, , , MgO in ore = 3.9/40.3 = 0.097 kg-mole '
CaCOa in ore = 0.364 kg-mole , ;:, : :
MgCOa in ore = 0.378 - 0.364 = 0.014 kg-mole i -'
MgO in ore = 0.097 - 0.014 = 0.083 kg-mole '•^'' ''-'''''
M 7 1 Tf f Total Heat of
Reactant Kg-moles „ ^. Formation, •
Formation , , , '
kcal
CaCOs 0.364 -289,500 -105,400
MgCOs 0.014 -268,000 -3,750 -
Total = -109,150 kcal - '
Products '' \ s ,'^ ., • ' s» , t*>itJMnH

C0» 0.378 -94,030 - 3 5 , 5 4 0 kcal
CaO 0.364 -151,700 - 5 5 , 2 2 0 kcal I
MgO 0.014 -146,100 - 2 , 0 4 0 kcal
Total = - 9 2 , 8 0 0 kcal
Heat evolved in calcination = - 9 2 , 8 0 0 -|- 109,150 = +16,350 kcal
6. Heat evolved in the partial combustion of charcoal. The purpose of introduc-

ing charcoal into the charge is to furnish heat for all endothermic reactions involved in
reduction, to supply heat for producing the slag and pig iron in a molten state, and to
supply carbon for the reduction of the various metaDic oxides of the ore. The char-
coil is not burned completely to carbon dioxide and water vapor, and its total heating
value is not rendered available in the blast furnace. The actual products resulting
from the partial combustion of the charcoal include carbon dioxide, water vapor,
carbon monoxide, methane, and hydrogen in the outgoing gases and graphite in the
solidified pig iron.
Since the heats of formation of the materials making up the charcoal are not known
it is necessary to calculate the heat evolved in its partial combustion from standard
heat of combustion data. As pointed out in Chapter V I I I , page 267, the heat evolved
in a reaction is the difference between the sum of the heats of combustion of the
products and that of the reactants. The actual reaotants entering into the com-
bustion of the charcoal include carbon dioxide, water vapor, and oxygen from both
the air and the ore. However, the heats of combustion of these materials are
zero, and they need not be considered in calculating the heat of reaction.
Heat of combustion of charcoal = 110.0 X -7035 = -773,850 kcal
Total Heat of
Molal Heat of
Product Kg-moles ^ i ^. Combustion,
Combustion , ,
kcal
CO 5.30 -67,410 -357,300
CH, 0.14 -212,805 -29,790
Hj 0.28 -68,320 -19,130
Graphite = 0.26 -94,030 -24,450
Total = -430,670 kcal
Heat evolved in partial combustion of charcoal = )ijt ^- ,
-430,670 - (-773,850) = 343,180 kcal
6. Heat evolved in formation of slag. Blast-furnace slags consist of complex mix-

tmes of silicates and aluminates of calcium, magnesium, iron, and manganese. In
order to calculate the heat of reaction accompanying the formation of a slag it would
be necessary to determine the complete mineralogical composition of the ore, flux, and
slag, together with the heats of formation of all the compounds present. Such data
are rarely available, and it is ordinarily necessary to neglect the heat evolved in form-
ing the slag from the oxides. It would not be expected that this thermal effect is
large. The formation of the monosilicate of calcium, manganese, iron, or magnesium
from Si02 and the respective oxide is in every case accompanied by an evolution
of heat. On the other hand, the formation of dicaleium silicate is accompanied
by an absorption of heat. These effects will frequently tend to compensate each
other.
•/: " • ; . ! ; ? , -ill 1-1 'XM! v J i i : ; i l i • :•.•/. • : ' , • ; . . . ' • ' > : ' • • ! • ; -I! •' • • • ! •••
7. Heat absorbed in the formation of cementite. In the blast furnace some

carbon combines with iron to form cementite, FesC, with simultaneous absorption
of heat. The formation of cementite is an endothermic reaction absorbing 4000 kilo
calories per kg-mole of FesC Upon slow cooling of pig iron this cementite is partly
decomposed with liberation of heat. In this instance the carbon present in the pig
iron is assumed to remain present as cementite because of rapid cooling during
solidification.
3.12
Heat absorbed in formation of cementite = 4000 X = . ... 1,040 kcal
This item of heat absorbed is neglected in the output of the present energy balance.
8. Enthalpy of air.
Air supplied = 16.72 kg-moles
Mean molal heat capacity between 18°C and 300°C (from Table VI, page 216) =
7.06. • .,.--:, ,..., -H"-' -
Enthalpy of air = 15.72 X 7.06(300 - 18) = 31,300 kcal
9. Enthalpy of dry blast-furnace gas.

Mean heat capacities between 18°C and 173°C taken from T ibie VI, page 216.
CO2 (2.62) (9.60) = 25.2 kcal
CHi _ (0.14) (9.27) = 1.3 kcal
Hj (0.28) (6.94) = 1.9 kcal
CO (5.30) (6.99) = 37.1 kcal
N« (12.44) (6.99) = 87.0 kcal
152.5 kcal
Total enthalpy = 152.5 (173 - 18) = 23,640 kcal
10. Enthalpy of water vapor in gas.
Moles of H2O in gas = 0.66 kg-mole
Molal humidity = 0.66/20.78 = 0.0318
Dew .point (Fig. 9) = 76°P (24°C)
Heat of vaporization at 76°F = 18,880/1.8 = 10,485 kcal per kg-mole
Total enthalpy of water vapor =
0.66[10,485 + 18(24 - 18) + 8.11 (173 - 24)] = 7790 kcal
11. Enthalpy of slag. S. Umino^ determined the relative enthalpies of various
blast-furnace and open-hearth-furnace slags and has found that the enthalpies of
these two types of slags are nearly the same when measured at the same tempera-
ture. These values are shown graphically in Fig. 90. It wiU be seen that there is
no sudden break in the enthalpy-temperature curve, thus indicating that no sudden
transformations take place in cooling and that the slag exists essentially in the form
of a glass From Fig. 90 it will be seen that at a temperature of 1360°C the enthalpy
of slag is 365 kcal per kg.
Total enthalpy of slag = 365 X 63.8 = 23,290 kcal
12. Enthalpy of pig iron. The average enthalpy of molten pig iron from blast
furnaces has also been determined by S. Umino.' Although the composition of the
pig iron used in his experiments is not identical with the one under discussion, never-
theless the compositions are sufficiently alike to justify use of the same values of
enthalpy. At a pouring temperature of 1360°C this enthalpy is 300 kcal per kg and
includes the sensible enthalpy of the solid and liquid states, the latent heat of fusion,
and the heat evolved in the separation and decomposition of cementite, FeaC, from
solution in the iron.
Total enthalpy of pig iron = 100 X 300 = 30,000 kcal
13. Heat absorbed by cooling-water = 576 X 13 = 7,490 kcal
Summary of Energy Balance, ,
Input •/ /
'/
Partial combustion of charcoal 343,180 kcal 91.6% /
Enthalpy of air 31,300 kcal 8.4%
Total 374,480 kcal 100.0%
2 S. Umino, Sderux Repts. Tdhoku Imp. Unit)., 17, 985 (1928).
• ' Idem, 16, 575 (1927).
1 1
Open Hearth Slags -
400
Blast Furnace Slags - ^
^
//
800
/^(<E. Pig Iron \

3
A4
M.P^
I
200
y^ f*ureIro 1
Jî
W
100
200 400 €00 300 1000 1200 1400 1600 °C

niion of Slags
ûrnace Slags Open Hearth Slags
2 3 4 5 6
SiOj 34.50 37.26 34.22 18.20 18,28 20,28
FeO 1.58 0.82 0.74 13.45 10.27 10.47
FeiOa 0.29 0.84 0,20 2,40 3.63 2.50
CaO 40.92 43.15 41,80 42,63 43.56 44.20
MgO 3.90 1.80 4.56 9.14 11,84 10.47
P trace trace trace 0.33 0.25 0,25
S 0.98 0.11 0.90 0.54 0.45 0.40
MnO 2.24 1.82 1.88 6.97 6.60 6.60
AI2O3 15.48 13.15 15.60 5.00 4.68 4.79
Apparent ep. gr. 20°C 2.80 2.94 2.97 3.46 3,56 3.12
Composition of Pig Iron Thermal Data on Pure Iron

C = 4.31% Heat of transition A2 (a to 0) 6.5 kcal per kg
Si = 1.11% Heat of transition A3 (/S to y) 5.6 kcal per kg
Mn = 0.63% Heat of transition At (r to 5) 1.9 kcal per kg
P = 0.12% Heat of fiaion 65.6 kcal per kg
S = 0.022%
Cu = 0 . 2 1 %
Heat of Fusion = 46.63 kcal per kg
FiQ. 90. Enthalpies of iron and slags, referred to 18°C. Data taken from S.
Umino, Soence iJepte. T6hoku Imp. Univ., 17, 985 (1928) for slag; 16, 675 (1927)
for pig iron; 18, 91 (1929) for pure iron.
Output
Heat absorbed in reduction of iron oxides 158,640 kcal 42.4%
Heat absorbed in reduction of Mn304 6,470 kcal 1.7%
Heat absorbed in reduction of SiOj 10,980 kcal 2.9%
Heat absorbed in calcination of carbonates.... 16,350 kcal 4.4%
Enthalpy of dry gas 23,640 kcal 6.3%
Enthalpy of water vapor...» 7,790 kcal 2.1%
Enthalpy of slag 23,290 kcal 6.2%
Enthalpy of pig iron 30,000 kcal 8.0%
Heat absorbed by cooling-water 7,490 kcal 2.0%
Heat losses unaccounted for (by difference). . . 89,830 kcal 24.0%
Total 374,480 kcal 100.0%
This balance is summarized in Fig 91.
9J.656 Partial Combustion of Charcoal

to COj.HjO, Hj,CH»,CO, and Graphite
6.4% Enthalpy of Air I N

msi Total
O IPal/'infitmn of Carbonatcs U%
Reduction of Oxides of 42.496

Iron in Ore
^
^ 3 Reduction of SiOtto Si 2.9*
L 3.Redaction of Mnj04 1.7%
J Enthalpy of Dry Gas 6.3*
^ = Enthalpy of HjO Vapor 2.1*
J Enthalpy of Slag 6.2*
J Enthalpy of Pig Iron 8.0*
3 Heat Removed by 2.0*

Cooling Water
- Losses Unaccounted for 24.0*
Total 100.0*
FIG. 91. Energy balance of a blast furnace.

CHAP. X] PETROLEUM CRACKING PROCESS 421
PETROLEUM CRACKING PROCESS

In a so-called " vapor phase cracking " process a clean, well frac-
tionated gas oil cut from petroleum, containing no material of gasoline
boiling range, may be decomposed to form gasoline and gas by heating
in the tubes of a furnace designed to provide the necessary reaction tune
at elevated temperatures. In order to arrest the reaction and minimize
coke formation, the hot vapor mixture of gas, gasoline, oil, and tar from
the heater is " quenched " by a relatively cool stream of oU. The result-
ing mixture passes to an evaporator where further cooling and rough
fractionation is accomphshed by means of a reflux stream of cool oil
sprayed in at the top of the vessel. The quantity of reflux is regulated
to obtain such a temperature in the evaporator as to produce the desired
quality of " tar " which is withdrawn from the bottom of the evaporator
and generally sold as heavy fuel oil.
The vapors from the evaporator pass to a fractionating tower the
lower section of which is utilized for preheating the fresh charge by
direct heat exchange. This tower is operated to produce well fraction-
ated gasoUne and gas as the overhead product. All partially decomposed
feed in the boiling range between gasoline and the tar is condensed as
a bottom product. This " recycle stock " mixes in the bottom section
of the tower with the fresh feed to form the " combined feed," part of
which is charged to the furnace, part used for quenching the hot vapors
leaving the furnace, and part used as a reflux in the evaporator. It
may be assumed that the mixture leaving the furnace is completely
vaporized and that negligible condensation is produced by the quench.
It may also be assumed that all the oil used for quench and reflux is
vaporized under the conditions of the evaporator forming no tar.
Illustration 3. The flow diagram of such a process is shown in Fig. 92, on which
are indicated significant temperatures and characteristics of the streams. The
furnace is of the gas-fired type, provided with both radiant and convection heating
sections. An air preheater transfers heat from the stack gases to the air used for
combustion.
The gas used for heating the furnace is supplied at 65 °F and has the following
composition by volume:
CH, 14.6%
CaHe 77.3
CO 1.2
Hj 6.1
Nj 0.8
The temperatures shown on the flow diagram are either arbitrarily set as bases
for design or are derived from previous pilot plant or commercial experience indi-
cating the temperatures necessary for the desired reaction rates and separations.
Pilot plant tests indicate that when cracking a gas oil having a gravity of 29°API
and a characterization factor of 1L8 the following yields are obtained, expressed as
percentage by Hquid volume of the fresh feed: ^j
Gasoline (58°API; K = 11.9) 61.0% .
Tar (2°API; K = 10.2) 24.5 "
It may be assumed that the operation is conducted with 100% material recovery
and that the yield of gas is determined by difference.
Fractionating
Tower 290° F.
J-Crondenser^j^^.j 2 sp-S
Stack Gasea 600'R
4
Gas Separator
Gasoline
C><h*-58°API
U.9 K.
Fresh Feed
•«• 29° API Gas Oil
U.8K.
FIG. 92. Vapor phase cracking unit.
Previous commercial experience has indicated that such an operation may be

carried out satisfactorily under reaction conditions which result in the conversion per
pass into gasoline and gas of 22.0 weight per cent of the combined feed entering the
furnace. The recycle stock from such an operation is found to have a gravity of 19°
API and a characterization factor of 10.7. The standard heat of the endothermic
reaction from liquid charge to liquid gasoline, recycle stock and tar, and gas at 60°F
and atmospheric pressure may be taken as 600 Btu per pound of gasoline plus gas
formed. Thus AH = 600.
On the basis of the above information it is desired to develop material and energy
balances of the separate parts and of the entire plant for the design of a unit to process
5000 barrels (42 gallons) per day of fresh charge. Radiation losses may be neglected
except from the furnace where a loss corresponding to 5% of the fuel is assumed.
Although the pressures throughout will be somewhat above atmospheric it will be
assumed that they are sufficiently low that all enthalpies may be taken as at atmos-
pheric pressure.
In calculating the enthalpies of mixtures, heats of mixing in both gas and liquid
phases may be neglected and also the effect of pressure upon enthalpies and upon the
heat of cracking. It should be noted that the characterization factors and degrees
API are additive on a weight basis.
The following information is required: All flow rates should be expressed in
pounds per hour and in barrels per day (at 60°F) for the hquids and in cubic feet
(at 60°F, 30 inches Hg, saturated) per hour for the gas. Heat rates should be
expressed in Btu per hour.
(a) Production rates of gasoline, gas, and tar.
(6) Flow rate of the combined feed to the furnace.
(c) Properties of the combined feed.
(d) Heat absorbed in heating and cracking oil in the furnace.
(e) Flow rate of the combined feed used for quenching vapors from the furnace.
(/) Flow rate of the combined feed used for the reflux in the evaporator.
(g) Reflux rate of gasoline in the fractionating column.
(h) Temperature of flue gases leaving the convection section of the furnace,
(i) Fuel gas burned in the furnace.
(j) Thermal efficiency of the combined furnace and preheater. •-
(k) Heat transfer duties of the condenser and coolers.
(0 Coohng water required by the condenser and coolers in gallons per minute.
(Assume a 20°F temperatiu-e rise of the water.)
*; SOLUTION
For ready reference the physical and thermal properties of the various petroleum
fractions are tabulated in Tables A and B. The characterization factors and degrees
API of a mixture are additive properties on a weight basis. The average molecular
weights and boiling points are obtained from values of API and K by use of Fig. 63.
Latent heats of vaporization are calculated from Equation (32), Chapter VII,
page 233. The mean specific heats of liquids and vapors are obtained from Figs.
66 and 67. Enthalpies at various temperatures are then calculated from the above
data using 65°F as the reference temperature.
All calculations are based upon 1 hour of operation.
TABLE A
PHYSICAL PROPERTIES or OIL FRACTIONS
K G Av Molecular Av Boiling
"API
60/60° Weight Point — °F
Gasoline 58.0 11.9 0.747 109 240
Tar 2.0 10.2 1.06 320 800
Fresh Feed 29.0 11.8 0.88 300 675
Recycle 19.0 10.7 0.940 220 560
Gas 1.22 (air)
Combined Feed 22.1 11.04 0.922 .240 585
(calculated)
TABLE B •
THEHMAL PEOPEETIBS OF OIL FRACTIONS
Mean Mean
H e a t of Specific H e a t Specific H e a t Enthalpies
Boiling
Vapori- of Liquids of Vapors 65 °F reference
Point
zation (65° to t°F) (240'' to fF)
°F
Btu/lb
fF Op fF tV fF Btu/lb
Gasoline 240 134.8 95 0,475 290 0.464 95(0 14.3

240 0.518 710 0.574 240(0 90,7
800 0.596 240 (y) 225.5
.-u-i >.U 1000 0.640 290 (w) 248.7
710 495.2
800 559.0
":<> .1 •:
1000 711.7
Tar 800 100 (800°F to t°F) 150(0 43,4

150 U.377 800 0.620 745(0 353.6
745 0.620 1000 0.660 800(0 388.1
800 0.528 800(6) 488.1
1000(tf) 620.3
Gas (65°F to f F )
95 0.444 95 13.3
290 0.498 290 112.1
710 0.608 710 392.3
800 0,630 800 463.0
1000 0,676 1000 631.3
Presh Feed 675 88 200 0.464 200 200(0 62.6
Combined 585 104.4 (585°F t o fF)

Feed 300 0.462 710 0.610 300(0 108.6
540 0.523 800 0.630 540(0 248.4
585 0.634 585(0 277.7
685 (w) 382.1
710 (t)) 458.5
800 (w) 517.4
Recycle 560 110.6 560 0.519 710 0.586 560(0 256.9

800 0.605 560(D) 367.5
1000 0.644 710 («) 455.3
800 (») 512.7
1000(f) 650.7
1. Rates of Production. Rates of production are calculated from the yield state-
ment, as follows:
Charge Gasoline Tar Gas
Bbl/day 6000 3050 1225
Gal/hr 8750 5340 2145
Sp. gravity 0.882 0.747 1.06
Lb/hr 64,290 33,210 18,930 12,150
M cu ft/day . ,., .... 3230
2. Flow rate of combined feed to furnace.

Of the combined feed entering the furnace 22% by weight is converted into gas
plus gasoline.
Gas plus gasoline, lb per hour = (33,210 + 12,150) = 45,360
45,360
Combined feed to furnace, lb per hour = "TZT = 206,170
Recycle stock produced in furnace, lb per hour = 206,170 —

(33,210 + 18,930 + 12,150) = 141,880 .... i<Jt .f .
3. Properties of combined feed. '

Recycle Stock Fresh Feed
'• Gravity, "API 19° 29°
Characterization Factor K 10.7 11.8
Lb per hour 141,880 64,290
' ' •'•^^ %bywgt 68.8 31.2
K of combined feed = (10.7) (0.688) + (11.8) (0.312) = 11.04
°API of combined feed = (19) (0.688) + (29) (0.312) = 22.1
From Fig. 63 average molecular weight = 240
Average boiling point = 585°F ••''.•s'*
The thermal properties are tabulated in Table B.
4. Heat absorbed in heating and cracking oil in furnace.

Input
Enthalpy of combined feed at 540°F = (206,170)
(248.4) = 61,212,600 Btu
Heat supplied to oil from combustion of fuel
(by difference) = 111,318,600
I 162,531,200
Ovi-put ..r
Enthalpy of gas = (12,160)(631.3) = 7,670,000 Btu
Enthalpy of gasoline = (33,210) (711.7) = 23,636,600
Enthalpy of tar = (18,930) (620.3) = 11,742,300
Enthalpy of recycle = (141,800)(660.7) = 92,269,300
Energy absorbed in cracking oil (600)(45,360) 27,214,000
Total 162,531,200
6. Weight of combined feed used for quenching hot vapors from furnace.
The energy balance about the quench point is shown diagrammatically in Fig. 93.
The required quantity of quench is fixed by this balance.
Heat • Effluent Evaporator Feed
lOw'F. i »-800°r.
(206,170 lb )H, (206,170 Ha+x Hi
J Ha
300°F
FIG. 93. Flow chart about quenc' point.
Let a; = lb quench supplied to vapor from furnace. " ' ' •
Input
Enthalpy of vapors at 1000°F = 135,317,200 Btu
Enthalpy of quench at 300°F = ' 108.6a:
Total = 135,317,200 + 108.6x
Output • ','.
Enthalpy of gas = (12,150)(462.97) = 5,625,100
Enthalpy of gasoline = (33,210) (559) = 18,564,400
Enthalpy of tar vapor = (18,930) (488.1) = 9,239,300
Enthalpy of recycle vapors = (141,880(512.7) = . . . . 72,741,900
Enthalpy of combined feed quench = (x) (517.42) = . 517.42x
Total = 106,170,700 + 517.42s
From the above balance:
106,170,700 + 517.42a; = 135,317,200 + 108.6a; '
408.82X = 29,146,500
X = 71,294 lb of combined feed for quench
JGas
710* F. 1 Gasoline
""—"—^"^ Reeycle
I Quench
Vpeflux
Vapors • • Combined Feed

800° F. Reflux
Îb.
300" F.
Tar745°F. ^ ' , /
FIG. 94. Flow chart of evaporator. ,'
6. Flow rate of combined feed used for reflux in the evaporator.

Let 2/ = lb of combined feed used for reflux in evaporator.
The reflux rate required in the evaporator is determined by an energy balance
following the flow chart shown in Fig. 94 '. ;
Energy Balance of Evaporator

Input
Enthalpy of vapors a t 800°F = 106,170,700 +
(71,294) (517.42) = 143,059,600
Enthalpy of combined feed for reflux = , 108.6;/
Total = 143,059,600 + lOS.Qy
Output
Enthalpy of gas = (12,150) (392.29) = 4,766,300
Enthalpy of gasoline = (33,210) (495.23) = 16,468,800
Enthalpy of recycle = (141,880) (455.34) = 64,636,400
Enthalpy of combined feed quench = (71,294) (458.45) =* 32,684,700
Enthalpy of combined feed reflux = (y) (458.45) = . . + 458.45?/
Enthalpy of t a r a t 745°F = (18,930) (353.6) = 6,693,600
Total = 125,249,800 + 458.45?/
From t h e energy balance:
125,249,800 + 458.45?/ = 143,059,600 + lOS.By
349.852/= 17,809,800 •^•, -
' I 2/ = 50,906 l b . ;
—»~ Gas=12,140 lb.
290 F. Gasoline-
it33,2101b.
1 RenuxatgS'P.
2 lb. Gasoline
• Feed at 200*p.
Vapors at 710° F.- 64,2901b.
309.4401b.
Combined Feed at 640°F.

328,3701b.
F I G . 95. Flow chart of fractionating tower.
7. Rate of gasoline refltix in fractioning tower.

The reflux rate required in t h e fractionating tower is determined from an energy
balance following the flow chart shown in Fig. 95.
Energy Balance
Let 2 = reflux, lb per hour.
Input
.Enthalpy of gas = 4,766,300 B t u
Enthalpy of gasoline = 16,468,800
Enthalpy of recycle = 64,636,400
Enthalpy of combined feed recycle and quench
(122,200) (458.45) = 56,022,600
Enthalpy of fresh feed = (64,290) (62.57) = 4,022,600
Enthalpy of gasoline reflux = 14.25z
t - 145,916,700 + 14.25Z
Output
Enthalpy of gas = (12,150)(112.0f^ = '.... 1,361,400 Btu
Enthalpy of gasoline = (33,210 + z) (248.66) = . .. 8,267,700 + 248.65?
Enthalpy of combined feed = (328,370) (248.42) = 81,573,700
Total = . . . . ^ 91,192,800 + 248.66Z
From the energy balance:
91,192,800 + 248.66Z i= 146,916,700 + U.25z ,•>,•] '
234.42 = 64,723,900 , .
z = 233,463 lb of gasoline reflux per hour
8. Temperature of stack gases leaving furnace.
The temperature of the stack gases leaving the furnace is calculated from an energy
balance of the air preheater based upon the combustion of 100 lb moles of fuel gas
with 30% excess air.
Basis: 100 lb-moles of fuel gas. .-^^
Heating
0, CO2 H2O
Mol Value Heating value X
Moles Lb required lb- lb-
Wt Btu per mole fraction
lb-moles moles moles
lb-mole
CH, 14.6 16 2.34 29.2 14.6 29.2 383,040 55,930

CjHj 77.3 30 23.19 270.66 164.6 231.9 671,090 518,160
CO 1.2 28 0.34 0.6 1.2 121,340 1,460
H, 6.1 2 0.12 3.05 6.1 122,980 7,510
N, 0.8 28 0.22
26.21 303.40 170.4 267.2 583,060

Btu per lb-mole
Oxygen theoretically required = 303.40 lb-moles

Oxygen actually supphed = (303.40) (1.3) = 394.4 lb-moles
0.79
Nitrogen actually supplied = 394.4 X —— = 1483 lb-moles ofr,>l
Total moles air supplied per 100 moles fuel = 1877 lb-moles
Molal humidity, lb moles water per lb mole dry gas (Fig. 9) = 0.022
Water from air = (0.022)(1877) lb-moles = 41.3 lb-moles
Products of combustion:
Carbon dioxide 170.4 lb-moles
Water = 267.2 + 41.3 308.6
Oxygen = 394.4 - 303.40 91.0
Nitrogen = 0.8 + 1483 1483.8
2063.7 lb-moles
Partial pressure of water vapor = -——'-—- (760) = 114 mm

(2063.7)
Dew point = 130°F
Enthalpy of water vapor in gas streams. Enthalpy

Btu/lb-mole
In entering air, 80°F, dew point 66°P 19,190
In leaving air, 400°F, dew point 66°F 21,820
In waste gases, TF, dew point 130°F 18,500 + 8.30i
In waste gases, 600°F, dew point 130°F 23,490
The temperature of the gases from the furnace is determined by an energy balance
of the air preheater, illustrated in Fig. 96. As a first approximation it may be
assumed that the temperature drop of the gases in the preheater approximately equals
the temperature rise of the air. Mean heat capacities are based on this assumed
temperature which may be corrected by a second approximation if the energy balance
shows it to be seriously in error.
Stack Gases <•••• | |—"*—Air 80° F.

600°F.
' • • f -
-Air 400° P.
t°F.
FiQ. 96. Flow chart of air preheater.
Energy Balance of Preheater

Input
Basis: 1.0 lb-mole of fuel gas
Enthalpy of air (18.75)(80 - 65)(7.0) = 1,971
Enthalpy of water vapor in air (0.413) (19,190) = 7,925
Enthalpy of hot waste gases:
, CO2 = 1.704(J - 65)(10.9) = + 18.57(« - 65)
H2O = (3.085) (18,500 + 8.300 = 57,085 + 25.6<
O2 = (0.910) (t - 65) (7.5) = 6.82« - 65)
N2 = 14.838(J - 65)(7.1) = 105.35(( - 65)
66,981 + 130.74(i - 65) + 25.6«
58,483 + 156.34^
Output
Enthalpy of waste gases leaving (at 600°F)
CO2 = (1.704) (600 - 65) (10.1) = 9,207
H2O = (3.085) (23,490) = 72,466
O2 = (0.910) (600 - 65) (7.3) = 3,554
N2 = (14.838) (600 - 65) (7.1) = 56,362
Enthalpy of air at 400°F
Air = (18.77) (400 - 65) (7.0) = 44,015
Water vapor = (0.413) (21,820) = 9,011
Total = 194,615 Btu/lb-mole of flue gaa
From the energy balance: j^

156.34< + 58,483 = 194,615
t = 871 °F temperature of gas leaving furnace
9. Fuel gas burned in furnace per hour.
The amount of fuel gas burned in the furnace may be calculated from an energy
balance either of the furnace or the furnace plus preheater based upon one hour of
operation. The latter balance is iljugtrated in Fig. 97.
AirSO'F.^ » >AAAA/
Stack Gases-"^ WVNA/^

600°?.
/\AA/V*- - Vapors
at 1000° F.
Combined Feed 640° F.- FuelGasat60°F.

w lb. Moles
FIG. 97. Flow chart of furnace and preheater.
Let w lb-moles of fuel gas busued per hour.
Energy Balance of Furnace plus Preheater
Input
Enthalpy of dry air at 80°F = Ifillw Btu
Enthalpy of H2O in air = 7,925i« Btu
Enthalpy combined feed = + 51,212,600
Heating value of fuel = 583,060«; Btu
Total = 592,956K) + 51,212,600 Btu/hr
Output
Enthalpy of vapors at lOOCF 135,317,200
Enthalpy of waste gases 141,509«)
Heat absorbed in cracking 27,214,000
Heat losses = (0.05) (583,060)u) 29,153u)
Total = 162,531,200 + 170,662«) H.
From the energy balance,
592,956u) + 51,212,600 = 162,531,200 + 170,662w
422,294K) = 111,318,600 ' '
w = 263.6 lb-moles dry fuel gas per hour
Cubic feet per hour of gas, at 60°F and 30 in. Hg saturated with H2O = (263.6)
(385.5) = 101,618 cu ft per hour
10. Thermal efficiency of furnace.
From part (4) the heat supplied to oil = 111,318,600 Btu
Total heating value of fuel gas (583,060)(263.6) = 153,694,600 Btu
111,318,600
Thermal efficiency based on heating value of gas X 100 = 72.4%
153,694,600
C H A P . X] PROBLEMS 431
11. H e a t transfer duties of condenser and coolers.

1. Condenser on fractionation tower
Gasoline product = (33,210) (248.7 - 14.3) = . . . 7,784,420 B t u / h r
Gasoline reflux = (233,463)(248.7 - 14.3) = . . . . 54,723,730
Gas = (12,150)(112.1 - 13.3) = 1,199,690
Total = 63,707,840
2. Tar cooler (18,930) (253.6 - 43.4) = 5,873,200
3. Combined feed cooler (122,200) (248.4 - 108.6) = 17,086,000
Total = 86,667,040 B t u / h r
12. Cooling water required by condenser and cooler.
Heat absorbed per gallon of water = 20 X 8.33 = 166.6 Btu
63,707,840
1. Condenser on fractionating tower = TTTTTTT^ = . . . 6373 gpm
5,873,200
2. Tar cooler = -r r—r- = 587
(166.6) (60)
17,086,000
3. Combined feed cooler = ,. „ „, ,^^s = 1709
(166.6) (60)
Total cooling water required = 8669 gpm
PROBLEMS
1. In a plant for the manufacture of sulfuric acid by the chamber process pyrites
is burned in a shelf burner. The gases from the burner enter t h e Glover tower
a t 480°C and leave this tower at 105°C, entering the first chamber. The gases leave
the last chamber a t 42°C and finally leave the Gay-Lussac tower at 21°C. On the
basis of 100 kg of pyrites, as charged, there are charged into the Glover tower, 175
kg of chamber acid, 65.2% HzSOi (51.8°B6) at 30°C; 610 kg of Gay-Lussac acid
at 23°C, and 1.30 kg of 4 0 % nitric acid at 20°C. The Gay-Lussac acid contains
78.0% H28O4 (60.0°B6), 0.984% N2O3 in solution, and 21.0% H2O.
The analyses of the pyrites, cinder, and the moisture-free gases leaving the burner
are as follows:
Pyrites Cinder Gases (by volume)
FeSj 90.00% FcjOs 89.80% O2 9.32%
SiOz 4.80% FeS2 1.65%, N^ 82.38%
H2O 5.20% SO3 1.93% SO2 8.00%
100.00% Si02 6.62% SO, 0.30%
100.00% 100.00%
The pyrites is charged to the burner at 20°C. The air enters at 20°C, under a
barometric pressure of 722 mm of Hg and with a percentage humidity of 40%. The
cinder is withdrawn a t 320°C.
For each 100 kg of pyrites as charged, 768 kg of acid containing 79.4% H2SO4 leave
the Glover tower at 100°C and are cooled to 23°C. The chamber acid leaves the
first chamber at 65°C and is cooled to 30°C before entering the Glover tower. The
acid leaves the Gay-Lussac tower at 30°C and is cooled to 23°C for recirculation.
The spray water enters the chambers at 20°C.
From the Row chart and assumptions of Illustration 1, calculate individual material
and energy balances, on the basis of 100 kg of pyrites as fired, of:
(o) The burner.
(b) The Glover tovirer.
(c) The chambers.
(d) The Gay-Lussac tower.
(e) The entire plant.
2. The charge delivered to a blast furnace, on the basis of 1000 lb of pig iron, con-
sists of 1810 lb of ore, 361 lb of Umestone, and 892 lb of coke. The analyses of
various components of the charge are as follows:
Ore (1810 lb) Limestone (361 lb) Coke (892 lb)
FeÔa 62.10% CaO 51.12% Carbon 88.20%
FeO 19.07% MgO 2.10% Hydrogen 2.00%
MnaOi 2.12% SiOz 2.89% FeaOa 2.10%
AI2O3 2.89% AI2O3 4.12% Si02 1.98%
Si02 8.62% FeÔs 0.52% CaO 2.32%
H2O 5.20% CO2 35.05% MgO 1.10%
100.00% H2O 4.20% S 0.20%
100.00% H2O 2.10%
100.00%
The total heating value of the coke is 14,200 Btu per lb.
On the basis of 1000 lb of pig iron produced, 51 lb of dust are collected from the
gases leaving the furnace. The analyses of the products are as follows:
Pig iron (1000 lb) Flue dust (51 lb) Gas analysis (by volum
Fe 92.28% FeO 83.2% CH4 0.80%
Si 2.10% C 10.1% CO2 12.10%
Mn 1.38% CaO 3.1% CO 29.30%
S 0.03% SiOj 3.6% H2 2.12%
C 4.21% 100.0% O2 0.20%
100.00% N2 65.48%
; • 100.00%
The surrounding air is at 70°F, 40% percentage humidity and a barometric pres-
sure of 29.2 in. of Hg. This air is heated and supplied to the tuyeres at 850°F.
The ore, flux, and coke are charged at an average temperature of 65°F.
The gases leave the furnace at a temperature of 422°F. The molten slag and pig
iron are tapped from the furnace at a temperature of 2500°F. The sensible enthalpy
of the flue dust is negligible.
Calculate the complete material and energy balances of this furnace, using the as-
sumptions of Illustration 2.
3. In producing 1 ton (2000 pounds) of steel in an open-hearth furnace the follow-
ing charge was supplied: /,
Hot metal from blast furnace (2400°F) 814 pounds
Cold scrap iron 1250 pounds ^
Limestone (95.5% CaCO,, 4.5% H2O) 118 pounds
Iron ore (94% FeiO,, 6% HjO) 56 pounds
Fuel oil 28.2 gallons
CHAP. X] PROBLEMS 433
Air was supplied to the regenerators at 80°F, 60% relative humidity, atmospheric
pressure. The air (85% of total supply) was preheated to 2000°F in the regenerators.
The hot gaseous products of combustion left the hearth at 2860°F, entered the regen-
erators at 2560°F and entered the stack at 1000°F. The average analysis of the flue
gases measured over the nine hour run was as follows:
CO2 = 17.0%
O2 = 0.8%
I N2 = 82.2%
100.0%
' The metals analyzed as follows:
Hot Metal Cold Scrap Steel
f Carbon 4.25% 0.15% 0.15%
'l Silicon 1.92 0.50 0.25
T Manganese 0.32 0.02 0.02
( ; Phosphorus 0.65 0.065 0.02
Iron 92.86 99.20 99.56
The steel and slag were poured at 2800°F. From experience it is known that 15%
of the air used in the furnace leaks in through the doors and brickwork of the hearth.
The fuel oil had the following properties:
Characterization factor 11.1
API gravity 12.0 '
The following information is desired on the basis of 1 ton of steel produced: (Use
80°F as basis of enthalpies.)
1. Material balance of solids charged to the hearth.
2. Weight of flue gas.
3. Weight of dry air used.
4. Weight of water vapor in air supply and in flue gas.
5. Overall material balance of combustion and refining processes.
6. Air theoretically required for combustion.
7. Percentage excess air used based on requirements of fuel oil.
8. Overall energy balance of process.
9. Energy balance of reactions on hearth.
4. It is desired to prepare a preliminary engineering design study of a proposed
catal3îc unit for the " cracking " of higher boiling petroleum fractions into gasohne
and gases rich in recoverable olefins. A flow diagram is shown in Fig. 97a.^
The charge, a well-fractionated gas oil, is pumped through a heater where its tem-
perature is raised to 900°F at a pressure somewhat above atmospheric and passed to a
reactor B where it is contacted with a refractory catalyst in the form of small pellets.
The reactor is provided with a heat exchange system through which a molten salt is
circulated for temperature control and is so designed that heat exchange surface is
uniformly distributed throughout the entire catalyst bed. The decomposition
reaction is endothermic, requiring the supply of heat from the salt system in order to
maintain isothermal conditions.
* The process illustrated in Fig. 97a is not in commercial application as shown but
represents a possible combination of features described in the many patents relating to
this subject.
As the cracking reaction progresses a carbonaceous deposit accumulates on the

catalyst which reduces its activity. In order to maintain continuous operation an
alternate reactor A is provided in which the hydrocarbon stream may be processed
while the catalyst deposit is removed from reactor B by oxidation with air. Thus,
one reactor system is at all times in the process period of the operation while the other
is in the burning or reactivation period. In Fig. 97a the solid lines represent the various
streams while reactor B is processing and the broken lines indicate the changes in
flow when reactor A is processing and B is undergoing reactivation.
Sat. Steam, 400 Ib/sq. in. gage
Gasoline
Fresli Charge
60°F
, "- _ Heavy Light '

, • • ! • - Gas Oil Gas Oil /
FIG. 97a. Flow chart of catalytic cracking process.
The burning of the catalyst deposit is highly exothermic and dmng the reactiva-
tion period the circulating salt heat exchange system serves to cool the catalyst to
prevent destructive overheating. Since the heat liberated in reactivation is gener-
ally greater than that required to maintain isothermal conditions in the cracking
reaction, a waste heat boiler is provided for removing heat from the circulating salt
by the generation of steam. The salt system is provided with switch valves so
arranged that the cooled salt from the waste heat boiler is pumped first to the reactor
under reactivation and then to the reactor which is processing. The temperature in
the processing reactor is controlled by a by-pass valve permitting regulation of the
quantity of salt passed through the heat exchanger. The entire circulating salt
stream passes through the reactor which is undergoing reactivation.
PROBLEMS 435
The hydrocarbon products from the processing reactor pass to a primary fraction-
ating column in which a heavy gas oil fraction is removed as bottoms. The overhead
from this column passes to a secondary fractionating column in which well fraction-
ated gasohne and gases are the overhead products. The bottoms, a light gas oil, is in
part recycled to the cracking process, in part used to reflux the primary column, and
in part withdrawn as final product.
Laboratory tests indicate that at the operating temperatures indicated in Fig. 97a
the following products and yield are obtained from the indicated charging stock:
% by volume
^. • °API K of charge
Charge 30 11.9 100
Gasoline 60 12.0 49
Light Gas Oil 32 11.4 24
, _ Heavy Gas Oil 25 11.5 18
The gas has a specific gravity of 1.6.
The catalyst deposit is found to correspond to 3 % by weight of the charge and con-
tains 4% hydrogen and 96% carbon by weight.
The hydrocarbon combined feed, consisting of fresh charge and recycled light gas
oil, is passed through the processing reactor at a rate of 1.2 volumes of oil, measured
at 60°F, per hour per volume of catalyst. This method of expressing reactor feed
rates is commonly used in cataljrtie processes and is termed the liquid hourly space
velocity. The catalyst volume includes the actual volume of the catalyst pellets and
the voids between the pellets in the catalyst bed. The density of the catalyst bed is
55 pounds per cubic foot.
At the above operating conditions it is found that 42% by weight of the combined
feed is decomposed into gasoline and gas in a single pass through the unit.
The average standard heat of reaction at 60°F and one atmosphere from hquid
combined feed to liquid gasoline and gas oils and gaseous gas is found to be -|-220
Btu per pound of gasoline plus gas formed, corresponding to an endothermic reaction.
The fuel gas burned in the furnace has the following composition:
• Mole per cent
^ Hydrogen 22.4
Methane 26.0
Ethylene 6.8
Ethane 7.2
Propylene 29.5
Propane 8.1
100.0
In accordance with the process information given above, develop the following
design factors and evaluate complete material and energy balances for a plant to
charge 10,000 barrels (42 gal.) per day. Base the balances on one hour of operation
and express rates in barrels per day and pounds per horn- for liquids and pounds per
hour and thousands of cubic feet per day for gases. Heat losses may be neglected
except from the furnace.
(o) The production rates of all net hydrocarbon products and the catalyst deposit.
(6) The rates and properties of the combined feed to the heater and the light gas
oil recycle stream.
(c) The heat absorbed by the oil in the furnace, assuming complete vaporization
but no decomposition in the heater.
(d) The rate of fuel consumption in the furnace, assuming complete combustion
and a radiation loss of 5% of the heating value of the fuel burned.
(e) The thermal efficiency of the furnace.
(/) The reflux rates to the primary and secondary fractionating columns.
(g) The heat transfer duties of the gasoline condenser and the light and heavy gas
oil coolers. >
(h) The cooling water requirements of the plant in gallons per minute with a
30°F temperature rise for the water.
(i) The volume and weight of the catalyst required in each reactor.
(j) The length of the process period in minutes if it is desired to hmit the deposit
on the catalyst to a maximum of 2.5% by weight of the catalyst. —
(k) The rate at which air must be supplied in order that reactivation may be com-,
pleted in the time of the process period. It may be assumed that the oxygen of the
air is 100% utiUzed, going to carbon dioxide and water under the catalytic combus- ,r»,
tion conditions. \.^
(i) The rate at which heat must be removed by the circulating salt from a reactor
under reactivation, neglecting changes in the enthalpy of the catalyst bed.
(m) The rate at which salt must be circulated through a reactivating reactor heat
exchanger. The heat capacity of the molten salt may be taken as 0.25 Btu per pound
per °F.
(re) The temperature of the salt entering the waste heat boiler.
(o) The quantity of steam generated in the waste heat boiler, in M lb per hr.
^ ,! J ^ CHAPTER XI
THERMODYNAMIC PRINCIPLES
In its most practical use thermodynamics deals with the limitations
imposed on the transformation of energy from one form to another and
with the study of the availability of energy for useful work. Thermo-
dynamic principles also provide useful means of predicting with a mini-
mum of experimental data those properties of a substance which are
related to energy and the equilibrium conditions which are approached
in any physical or chemical process.
Thermodynamic potentials constitute the driving forces which cause
all natural processes to progress toward a state of equilibrium. How-
ever, the rate at which the state of equilibrium is approached depends
not only upon the driving force but also upon a resistance or rate factor.
Thermodynamic principles alone do not permit evaluation of th-^se
resistances or rate factors. The combined knowledge of the thermo-
djTiamic "driving forces and of the corresponding resistances or rate
factors constitutes the science of kinetics.
System and Process. In thermodynamic terminology a system refers
to a substance or group of substances set apart for study and a process to
the changes taking place in the system. Thus hydrogen, oxygen, and
water constitute a system, and the combustion of hydrogen to form water
constitutes the process. A system is closed when there can be no ex-
change of matter with its surroundings and open when such exchange is
possible. In a multiphase system each separate phase is an open system,
since material is free to enter and leave individual phases although the
system as a whole may be closed.
A closed system may be free to exchange heat and work with its
surroundings. Heat can be transferred through the walls of the vessel
enclosing the system, and mechanical work can be performed upon or by
the system by means of a piston connecting the vessel with some external
mechanism. Devices can also be arranged for the removal or addition of
electric energy. A closed system is thermally isolated when the enclosing
walls are impervious to the flow of heat, mechanically isolated when it is
enclosed by rigid walls, and completely isolated when neither material nor
energy in any form can be added or removed. A closed system in con-
tact with a h£at reservoir is free to receive or lose energy in the flow of
heat from or to the reservoir. A closed system in contact with a piston
437
438 THERMODYNAMIC PRINCIPLES [CHAP. XI
is free to receive or deliver work energy in the movement of the piston.

The atmosphere surrounding a closed system may also act as a reservoir
for the flow of heat or as a piston for transmitting work from and to the
system. The atmosphere also may be made part of the system as in the
combustion of a fuel in air. In a consideration of the energy relations of
any nonisolated system, the energy of the surroundings must be taken
into account as well as that of the system itself.
Availability and Degradation of Energy. It is generally recognized
that the most useful forms of energy are those which are capable of doing
mechanical work. Thus, the rotation of a shaft or the motion of a pis-
ton represent energies of the highest order, which in the absence of fric-
tion can be completely converted into useful work. Similarly, electric
energy in the absence of electrical resistance and mechanical friction can
be converted entirely into mechanical work. Actually, however, be-
cause of friction part of the mechanical or electric energy of a machine is
always converted into heat. It is also common experience that, whereas
heat can be converted into mechanical or electric energy by means of an
engine, only part of the heat which flows into the engine can be recovered
as work. Thus, although there is an exact quantitative equivalence
among the different forms of energy, there is a marked difference in the
availabiUty of these forms for useful work. Heat represents the least
available form of energy, and the transformation of other forms into
heat represents a degradation of energy.
There are other means of degrading energy and rendering it less avail-
able for useful work. Thus, the internal energy of a gas at high pressure
is more available for performing useful work than that of the sartie gas at
low pressure and at the same temperature. Even though the internal-
energy content of the gas is the same at both pressures, as is the case for
ideal gases, the availability of the internal energy for further work of
expansion is less in the expanded gas. Thus, an isothermal reduction in
the pressure of a gas represents a degradation of its internal energy.
Similarly, the isothermal mixing of two unlike gases represents a degrada-
tion in the internal energy of the system since work is required to separate
them.
Reversibility. A reversible process is defined as one which proceeds
under conditions of balanced forces such that the direction of the process
can be reversed by an infinitesimal change in external conditions. An
example of a reversible process is the vaporization of a liquid under its
own vapor pressure in a cyhnder fitted with a frictionless piston and in
contact with an isothermal heat reservoir. At any stage an infinitesimal
increase in pressure upon the piston will produce condensation, and an
infinitesimal decrease will cause vaporization. It follows that in a
CHAP. XI] THE SECOND LAW OF THERMODYNAMICS 439
reversible process an infinitesimal change will tend to restore the original

conditions, so that no degradation of energy takes place, and the availability
of the energy of the combined system and its surroundings remains constant.
Actual processes are generally to some extent irreversible and result
in a decrease in the availability of energy. The flow of heat from one
body to another is always an irreversible process. It is impossible to
restore heat to the first body except by transfer from a still hotter body or
by the degradation of highly available mechanical energy. Similarly,
the dropping of an object without restraint to a lower level is an irrevers-
ible process.
Another example of an irreversible process is the free expansion of a
gas. If a gas is confined in a closed container of volume Vi and is
allowed to expand into a vacuum chamber such that the new total
volume is V2, and if this system is entirely isolated from its surroundings,
by rigid insulating walls, the gas is said to expand freely. No work is per-
formed upon the surroundings, no heat flows in or out of the system, and
for an ideal gas there is no change in temperature. Hence, the energy
content of the expanded gas is the same as that of the original gas. Ho v-
ever, in order to return the expanded gas to its original condition the
expenditure of external energy as work of compression is required. For
an ideal gas all this work manifests itself as heat in compressing the gas,
and in isothermal compression this heat is transferred to the surround-
ings. Thus, in isothermal compression of an ideal gas the energy content
remains constant but assumes a higher level of availability.
Other examples of irreversible processes are the mixing of hot and
cold fluids, the inelastic deformation of a solid, the magnetic hysteresis of
an iron core, the flow of electricity through a resistance, the dissolution
of a solid, and spontaneous chemical reactions.
The Second Law of Thermodynamics. The concept of reversibility
forms the basis for the second law of thermodynamics. One statement
of this law is that all spontaneous processes are to some extent irreversible
and are accompanied by a degradation of energy. A corollary to this
statement is that it is impossible for any self-acting machine to transfer
energy from a given state to a higher state of availability.
The validity of the second law of thermodynamics is confirmed by
extensive experimental evidence when applied to gross masses of matter
in which large numbers of molecules are present. Under these circum-
stances the laws of probabihty are followed in estabhshing the distribu-
tion of molecular and atomic energies.
An important deduction from the second law is that any machine which
performs work by transforming molecular energy received at fixed intake
conditions and rejecting molecular energy in a further degraded state at
fixed discharge conditions will convert a maximum percentage of the

energy received into work when it operates, reversibly and with no net
degradation of energy. Any actual self-acting machine must operate
with a lower efficiency of transformation than that of a reversible
machine. If it were possible to construct a more efficient engine, it
could be used to drive the reversible machine in reverse to transfer
molecular energy from a given state to the original less degraded state in
contradiction of the second law.
The f amiUar steam engine is an example of such a machine which per-
forms work by receiving steam at a high temperature and rejecting it at a
low temperature. The maximum efficiency possible with given intake
and exhaust conditions is obtained when such an engine operates reversi-
bly. This theoretical maximum efficiency is always less than 100 per
cent because of the energy content of the rejected steam. ' • - • '. • /
ENTROPY ,^ ; .
Before undertaking development of the quantitative relationships
involved in transformations of energy it is desirable to define an addi-
tional thermodynamic function to serve as a measure of the unavailablity
or degradation of the energy of a system. This function is termed
entropy.
The arbitrary mathematical definition of various thermodynamic
functions is an entirely legitimate procedure which may be extended to
create as many additional functions as may prove useful. Enthalpy is
an example of such a function, and numerous others are introduced in the
following pages as opportunities for their uses arise. For mathematical
accuracy it is only necessary that each new function be completely
defined and used in exact accord with the terms of the definition. How-
ever, the intelligent application of such a function is greatly assisted by a
thorough understanding of its physical significance. For this reason the
physical significance of entropy is first developed as a basis for its
mathematical definition. <
The physical concept of entropy may be expressed as follows: Entropy
is an intrinsic property of matter so defined that an increase in the unavaila-
bility of the total energy of a system is quantitatively expressed by a corre-
sponding increase in its entropy. Since entropy is an intrinsic property,
its magnitude is dependent only on the nature of the matter under con-
sideration and the state in which it exists and not upon its external
position or motion relative to other bodies of matter. Thus, the entropy
of an elevated weight is no different from that of the same weight in a
lower position, and the entropy of a rotating flywheel is equal to that of
the same flywheel at rest under the same conditions of temperature and
- ;•" .' • .!?••'' •- i'.' • • \

CHAP. XI] ENTROPY 441
pressure. The same is true for other intrinsic properties such as internal
energy and enthalpy.
Since entropy is a measure of unavailabihty of total energy, it follows
that the entropy of a system is increased by the degradation into heat of
any higher form of energy which it possesses. This is illustrated by
consideration of an isolated system comprising an inelastic weight
suspended above a rigid plate. If the weight is allowed to fall, the total
energy content of the system is unchanged, but the potential energy of
the weight is converted into heat as a result of the inelastic impact on the
plate. The temperature of both the weight and the plate are corre-
spondingly increased by the absorption of this heat in the form of
internal energy. The process of dropping the weight has not changed the
energy content of the system but has degraded it to a less available form
and is therefore accompanied by an increase in entropy. Similarly, if a
rotating flywheel is stopped by a brake, kinetic energy is degraded into
heat, and the entropy of the system is increased. However, if the fly-
wheel were stopped in a reversible manner by means of a frictirnless
transmission which sets another flywheel in motion, there would be no
degradation of energy and no change in entropy.
In both of these illustrative systems an increase in entropy results
from the addition of heat to the system through the degradation of a
higher form of energy. However, since entropy is an intrinsic property
of the matter of the system not influenced by the elevation of the weight
or the movement of the flywheel, it follows that the entropy of a system is
increased by the addition of heat through any mechanism or from any source.
Thus, heat might be added to the system containing the elevated weight
and the rigid plate increasing the total energy content of the system
until a temperature is reached which is equal to that attained after the
weight has been permitted to fall in the system when isolated. Since
entropy is independent of the position of the weight and the plate, the
final entropies of the system are equal, and the increases in entropy
accompanying these two processes are also equal, although heat is added
to the system by the degradation of its own energy in one case and by
increase in its total energy content in the other.
It is evident that the amount of heat added to a system is a partial
measure of the magnitude of its increase in entropy. However, the
quantity of heat added is not the sole measure of increase in entropy.
The unavailabihty of energy and the entropy of an isolated system are
increased by the transfer of heat within the system to a region of lower
temperature. Similarly, the addition of heat to a system at a low
temperature results in a greater degradation of energy than at a high
temperature. It follows that the increase in entropy accompanying the
addition of a given amount of heat to a system is increased by lowering the

temperature at which the heat is added. i)-'
In order to complete the definition of entropy, one more factor must be
considered in addition to the amount of heat added and the temperature
level at which it is added. For example, a gas may be expanded freely
to a lower pressure within a closed system which is completely isolated
both thermally and mechanically. This is an irreversible process result-
ing in degradation of the energy of the system and an increase in its
entropy. However, no heat is added, and no work is done, and for an
ideal gas the process is isothermal. In this irreversible process the
entropy increase is not measured by the addition of heat. The same
final state might be reached by expanding the gas through an engine
within the system and continuously converting the work done into heat
by means of friction. In this case heat is added to the system by the
degradation of the mechanical work. The amount of heat added
increases as the efficiency of the engine is increased and reaches a maxi-
mum when the engine operates reversibly. However, entropy is an
intrinsic property of matter, the change of which in any process is
dependent only upon initial and final states and not upon the path. It is
evident that in the process under consideration the amount of heat
added to the system can be taken as a measure of the increase in entropy
only if the nature of the process is specified. This specification is
logically taken to correspond to the maximum possible degradation of
higher forms of energy into heat. Thus the increase in entropy of a system
is measured by the amount of heat added only when all changes in the
intrinsic states of the matter of the system occur reversibly.
The three requirements for the quantitative definition of entropy are
satisfied by the following equation:
f dS = — (reversible change of state) ' - (1)
where (S = entropy. , ^rtti . ~ / ; ^

The primed differential symbol d' is used for indicating incremental quantities of
heat and work. In contrast to the other terms in the energy equation, neither heat
nor work are properties of the system, nor can they in general be expressed as a func-
tion of the state of the system. An infinitesimal change in a property such as volume
can be expressed as dV which is an exact differential with respect to the state of the
system. However, since neither q nor w are properties an increment of either is
referred to as an inexact differential.
An important corrollary of Equation (1) which follows frora the second

law of thermodynamics and the definition of entropy as a measure of
degradation is that the entropy change accompanying an irreversible
li.v
CHAP. XI] ENTROPY 443
change of state is greater than d'q/T, or

T dS > d'q (irreversible change of state) (2)
Thus, if an isothermal change of state such as the expansion of a gas
occurs reversibly the gain in entropy of the gas is compensated by a
reduction in entropy of its surroundings resulting from withdrawal of
heat, and the combined entropy of the material and its surroundings is
unchanged. If the same change in state is accomplished irreversibly,
the combined entropy of the gas and its surroundings increases although
the increase in entropy of the gas itself is the same as for the reversible
change. Consequently less heat is withdrawn from the surroundings.
It is important that the significance of the restriction to reversibility
in Equation (1) be appreciated. This restriction does not apply to the
manner by which heat is added to the system or the process employed in
the degradation of higher forms of energy. Thus, if a block of steel is
heated by a flame, its increase in entropy is measured by the amount of
heat added even though the addition of the heat from the flame to the
solid is highly irreversible. Similarly, the restriction does n r t apply to
changes involving only the relative positions or movement of bodies of
matter. The process involving the falling weight which was discussed
previously is irreversible, but the accompanying entropy change is
measured by the heat added. It is only processes involving changes in
the intrinsic state of the matter itself which must be reversible for the
application of Equation (1).
Much has been written on the physical significance of entropy. The
concept has already been advanced that an increase in the unavailability
of the energy of a system is accompanied by an increase in entropy. On
this basis entropy may be looked upon as a measure of the unavailabiUty
of the energy of a system of given energy content. However, the entropy
changes which accompany a change in total energy content do not neces-
sarily indicate increased unavailability of energy. For example, if the
temperature of a substance is increased by heating, both the total energy
and the entropy are increased. In this case the increase in entropy of the
substance is accompanied by a decrease rather than an increase in the
unavailability of its energy and the increase in entropy of the substance is
not a measure of the change in unavailability of its energy. However, if
consideration is broadened to include the substance plus its surroundings
from which the heat is withdrawn, the entropy of this entire system of
constant energy content is a measure of the unavailabihty of its energy.
A further physical concept of entropy results from its relationship to
probability. The existence of such a relationship is evident from the
fact that all spontaneous changes are in the direction of maximum proba-
bility and are also accompanied by increases in entropy. It follows from

the theory of probability that the most probable state in which a system
can exist is that having the least order of arrangement or greatest ran-
domness. This concept may be illustrated by consideration of a large
number of black and white balls shaken together in a box. The most
probable average arrangement is a uniform distribution of the black
among the white balls, and the chance of a compartment of the box
containing balls of only one color is remote. Thus, a state of random-
ness, free from orderly arrangement, is the most probable state of any
system, and entropy may be looked upon as a measure of randomness
which is a minimum for systems in orderly arrangement.
It is evident that in any given change in conditions accompanied by an
increase in entropy more heat is added in a system of large mass than in a
smaller one. Thus, like internal energy or volume, entropy is an exten-
sive property the magnitude of which is dependent upon the mass in^
volved. It follows that the entropy of a system is equal to the sum of
the entropies of its separate phases or mechanical parts. u S
In the foregoing discussion no consideration has been given to the scale
or function employed for the quantitative expression of temperature in
the definition of entropy. Although any arbitrary scale may be used to
indicate relative hotness or coldness, it is evident that only one type of
scale or function can be used to express the temperature of Equation (1)
if entropy as defined by it is to represent a true measure of the unavaila-
bility of energy at different temperature levels. This correct function is
termed the thermodynamic temperature scale.
At this point it is difficult to arrive at the nature of the thermodynamic
temperature by direct derivation, and so it will be tentatively assumed
that it is the same as the ideal-gas temperature which is defined in
Chapter II, page 30, as T = pv/R. Rigorous thermodynamic functions
and relations are developed from Equation (1) without consideration of
the nature of the temperature scale. From these relations it is proved
(page 473) that the thermodynamic and ideal-gas temperature scales are
identical. Any size of temperature unit may be used with no change in
the relationships involved except a change in the numerical values of
temperature and entrop^y.
Further confirmation of both the assumed nature of the thermody-
namic temperature scale and the fact that entropy as defined by Equa-
tion (1) is a true measure of the degradation of energy and consistent
with the second law may be developed through consideration of ideal
heat-engine cycles, such as are discussed in Chapter XIII. These proofs
are well summarized by Keenan.'
' J. H. Keenan, " Thermodynamics," John Wiley & Sons, New York (1941). '
•if^^^
\
CHAP. XI] CHANGE OF ENTROPY 445
Change of Entropy with Change of Phase or Temperature. Since

entropy is an extensive property of matter, its magnitude is dependent
on the physical state and mass of the material under consideration as
well as on the temperature and pressure at which it exists. Any change
which occurs in any of these properties is accompanied by a correspond-
ing change in entropy which may be calculated from Equation (1).
The changes in entropy which take place when a substance undergoes
a reversible change of phase at constant temperature and pressure as in
fusion, evaporation, and transition are calculated by dividing the change
in enthalpy by the absolute temperature.
Illustration 1. Calculate the molal entropy changes in the fusion and in the
vaporization of ethyl alcohol at atmospheric pressure.
Ethyl alcohol melts at 159°K with a heat of fusion of 1150 cal per g-mole. Since
fusion and vaporization are reversible isothermal processes, the molal entropy increase
of fusion is ' ' • ' >•; .;'.'"I'J',;
. \ AS = I; = i ^ = 7.23 cal/(g-mole)rK) .•••

i toy
The heat of vaporization of ethyl alcohol at atmospheric pressure is 9400 cal per
g-mole at a normal boiling point of 351°K. Hence, the increase in entropy in vapori-
zation is . ^ ^. ^ iOî-t.Mnû -or--!qf;'V-;
AS = I; = ^ = 26.8 cal/(g-mole) (°K)
It may be observed that the units of entropy are the same as for heat
capacity.
When heat is added with a resultant temperature rise, the increase in
entropy must be calculated by integration of Equation (1) over the
required temperature range, thus
Illustration 2. Calculate the increase in entropy when 1 g-mole of CO2 gas is

heated from 0°C to 1000°C at atmospheric pressure. The molal heat capacity of CO2
expressed in terms of absolute temperature degrees Kelvin is given on page 214 as
Cp = 6.85 4- 0.0085337 - 2.475 (10^) T^
As = r ' ^ = r f^ + 0.008533 - 2.475(10^)r'jdr

JTI i 1/273 \ .( /
1273 2 475
= 6.85ln-—- + 0.008533(1273 - 273) - - — (10-«)[(1273)» - (273)']
= 17.16 cal/(g-mole) (°K) ..-,.'• s. (3)
In calculating the entropy of a substance with reference to any selected

standard conditions, it is necessary to add together the entropy changes
for each transition, for each temperature rise, and for changes in pressure.
The variation of entropy with pressure is developed in a following section.
Third Law of Thermodjmamics. It was proposed by Nernst and
subsequently confirmed by extensive experimentation that at the abso-
lute zero of temperature the entropy of any pure crystalline substance
free of all random arrangement is zero. This principle is known as the
third law of thermodynamics. Accordingly, by extending measurements
of specific and latent heats down to 0°K absolute values of entropy can
be calculated from Equation (1). It is thus evident that the four refer-
ence properties of matter, temperature, pressure, volume, and entropy,
can all be presented as absolute values, whereas energy contents can be
reported only relative to some reference state. It has already been
shown that absolute values of temperature, pressure, and volume are
required in formulating the properties of matter. It is shown later that
the same must be true of entropy.
The utihty of the third law in establishing values of absolute entropy
brings out the importance of calorimetric measurements at temperatures
down to the absolute zero. At these low temperatures the common
empirical heat-capacity equations of Chapter VII do not apply. It is a
corollary of the third law that the heat capacity of a crystalline substance
is zero at the absolute zero temperature. However, the manner in which
the heat capacity diminishes with decreasing temperature in the low-
temperature range is a unique property, varying widely for different
materials. It is impossible to estimate low temperature thermal data by
the extrapolation to zero of high temperature measurements.
FREE ENERGY AND TOTAL WORK FUNCTIONS

A thermodynamic function of great importance is defined by the
following relation and termed the free energy G,
'• •' ' • • •^' G = H - TS • ! (4)
or, for changes taking place at constant temperature, * •

' AG = AH - TAS • * (5)
Where work is available from a process in addition to the mechanical

work of expansion, it follows from the first law discussed in Chapter VII
that for a nonflow process ; J ^, , ._ , - .; .,•
;>;> , q = AU + We + Wf .• , .' . ^: (5a)
' -it-.,,- ¥.,•••"'•

CHAP. XI] FREE ENERGY 447
where • • • • .• - - • ,,, i,.,^
We = work of expansion done by the system
Wf = useful work other than work of expansion, such as electric energy
For a reversible process at constant temperature and pressure
q = T AS and We = p AV. Hence, under these conditions the first law
may be written as
AU+ pAV ^ TAS-Wf = AH (6)
By combination of (5) and (6), it follows that ,
; . .. ' ' ' , AG= -Wf '-,. .: (7)
Thus, in a process which occurs reversibly at constant temperature
and pressure, the useful work is equal to the decrease in the free energy
of the system. When the process occurs irreversibly at constant tem-
perature and pressure, T AS > q, and the useful work is always less than
the decrease in the free energy of the system. Therefore, the decrease in
free energy accompanying a process at constant temperature and press-
ure represents the maximum useful work which becomes available only
when the process is reversible.
Most chemical reactions proceed under no restraint other tha.-- that of
pressure and dissipate the energy of the system as heat with no produc-
tion of useful work. Exceptional cases are found in the operation of
electrolytic cells wherein a part of the energy content of the system is
recovered as useful work. As an example of the significance of free
energy 1 g-atom of zinc may be dissolved at constant temperature and
pressure in excess hydrochloric acid of a given initial concentration, with
the release of hydrogen gas and with no evaporation of water or hydrogen
chloride. If it were possible to connect the zinc as anode through an
external circuit with an insoluble cathode placed in the same solution and
allow the reaction to proceed slowly and reversibly at constant pressure,
the electric energy developed would equal the decrease in free energy of
the system, d'w; = —dG. In the absence of the electrolytic cell, this
reaction would proceed irreversibly with no generation of electric energy.
However, for the same initial and final conditions, the decrease in free
energy of the system is independent of the path of the reaction and
would be the same in both cases.
Another useful thermodynamic function is defined by the following
relation and termed the total work function A,
A = U -TS , (8)
or, at constant temperature,
AA = AU - TAS . ' :•• : .-: (9)
From the first law, for a reversible nonflow process at constant tempera-
ture,
AU = q-w=TAS-w (10)
Combining (9) and (10) for a reversible process at constant temperature

gives
: AA = —W = —{We + Wf) '. (11)
Therefore, the decrease in the total work function accompanying a

process at constant temperature represents,the maximum total work
which becomes available only when the process is reversible. For a
reaction proceeding at constant volume, We = 0, and AA = AG.
There has been considerable confusion between the concepts of free
energy G and total work function A. Although the differences are usu-
ally slight, they are equal in a process at constant pressure only when
there is no change in volume.
The Four Thermodynamic Energy Functions. The internal energy U,
enthalpy H, free energy G and total work A are functions representing
energies which are capable of performing useful work under certain
conditions of restraint. For this reason these functions will be referred
to as the four energy functions. These functions are frequently called
the thermodynamic potentials, but this terminology leads to confusion
with the usual concept of a potential as an intensive driving force. The
significance of the four energy functions can be made tangible by again
referring to the dissolution of zinc metal in a solution of hydrochloric
acid and assuming that the system can be set up as a reversible electro-
lytic cell in a calorimeter. When the reaction proceeds isothermally at
constant pressure and irreversibly with no generation of electric energy,
the heat given up by the system and flowing into the calorimeter is equal
to the decrease in the enthalpy of the system, or — g'p = —AH. If the
reaction proceeds isothermally at constant volume and irreversibly and
with no generation of electric energy, the heat given up by the system
to the calorimeter is equal to the decrease in internal energy of the
system, or — g„ = —AC/.
If the zinc were connected as anode through an external electric circuit
as previously described and the reaction allowed to proceed reversibly at
constant pressure, the electric energy developed would be equal to the
decrease in free energy, or Wf = —AG. The work of expansion against
the atmosphere We in the release of hydrogen would be p AF and the total
work done Wf + We would be equal to the decrease in the total work
potential of the system, or W/ + iWe = — AA.
CHAP. X I ] CRITERIA OF EQUILIBRIUM 449
In summary, for isothermal nonflow processes:

At constant volume AU = q — Wf
At constant pressure AH = q — w/
In a reversible process at constant pressure AG = —Wf
In any reversible process AA = —w/ — We
Although in any reaction the values of AH, AU, AG, and AA depend
only upon initial and final conditions and are independent of the path
pursued, the direct calorimetric and electrical measurements of these
quantities must be made under conditions of change similar to those
previously described. ,
EQUILIBRIUM IN A CLOSED SYSTEM

A system is in equilibrium when its state is such that it can undergo no
spontaneous or unaided changes. Such a condition can result only
when all forces or potentials which tend to promote change are absent or
are exactly balanced against similar opposing forces or potentials. An
iron ball resting at the bottom of a spherical bowl is in mechanical
equilibrium with the bowl, since the mechanical forces acting on the ball
are balanced and the ball is at rest. No alteration in the position of the
ball can possibly take place except at the expenditure of mechanical
energy received from the outside.
Although the iron ball resting at the bottom of the bowl is in a state of
mechanical equilibrium, it is not in a state of chemical or thermodynamic
equilibrium with the surrounding air. Spontaneous, unaided and irre-
versible vaporization and oxidation of the iron can occur to form vapor
and rust. This spontaneous process will proceed, although very slowly,
until all the iron has been vaporized and oxidized to a condition of
complete thermodynamic equilibrium. Thus many conditions of
apparent equilibrium actually represent only partial equilibrium with
respect to all possible changes. In dealing with such partial equilibria
it is necessary that the changes with respect to which equilibrium is
established are clearly recognized.
Criteria of Equilibrium. Since a principal objective of thermodynamic
theory is the development of mathematical expressions for equilibrium
conditions it is important to derive fundamental specifications which
must be fulfilled by any system at equilibrium. Such specifications,
termed criteria of equilibrium, are the foundation upon which complete
relationships among the various properties of a system at equilibrium
are established.
From a thermodynamic standpoint u is necessary that in a system at
equilibrium every possible change which might take place to an infinitesi-
mal extent shall be reversible since any irreversible change would result
in a displacement which would destroy the Original equilibrium. As was
previously pointed out, reversible processes are accompanied by no
change in total entropy of the combined system and its surroundings,
whereas every spontaneous process is accompanied by an increase in
total entropy. Thus, a universal thermodynamic criterion of equilib-
rium is that for any change which takes place the total entropy of the
system and its surroundings shall be constant. In a completely isolated
system the entropy of the system itself is constant at equilibrium. From
Equation (1), defining entropy, it follows that where heat is added to a
system in which all changes of state are reversible, d'q = T dS, where S is
the entropy of the system itself, not including its surroundings. This
expression may also be taken as a criterion of reversibility and equilib-
rium. Since for all irreversible changes of state dS > d'q/T, if any incre-
mental addition of heat to the system is accompanied by an entropy
increase equal to and not greater than d'q/T, all thermodynamic processes
within the system must be reversible, and it follows that the system is in
equilibrium.
A system is in stable equilibrium if after a finite displacement it
spontaneously returns to its original state when the displacing force is
returned to its original value. A round pencil lying in the bottom of a
cylindrical trough is a mechanical example of this type of equilibrium.
If, however, this pencil is carefully balanced on its sharpened point, it
will be in a state of unstable equilibrium such that finite displacement
does not lead to a spontaneous return to its original conditions. Ther-
modynamically, a system is in unstable equilibrium if, although any
infinitesimal change is accompanied by no change in the total entropy of
the system and its surroundings, a finite displacement involves an increase
in total entropy. For example, a finite displacement of the pencil
balanced upon its point results in an irreversible process whereby heat is
developed and the entropy of the pencil and its surroundings increases.
In a system in stable equilibrium no finite change can be accompanied
by an increase in total entropy. Thus, at stable equilibrium, for any
change resulting from a temporarily applied extraneous force, not asso-
ciated with the system or its normal surroundings,
dSi= 0, ASt^ 0 V (12)

where St = total entropy of the system and its normal surroundings.
On the other hand, in unstable equilibrium, for any change
dSt = 0, ASt > 0 (13)
In addition to the general criteria of equilibrium just described, modi-
CHAP. XI] CRITERIA OF EQUILIBRIUM 451
fied criteria are useful to define equilibrium in a system under different

types of specified restraint. For example, liquid ether might be per-
mitted to evaporate in a container initially filled with nitrogen at atmos-
pheric pressure. If the walls of the container are rigid and do not con-
duct heat (and magnetic and electrical effects are absent), the system is
completely isolated, and the process is restrained to conditions of con-
stant volume and constant internal energy. Thus, dV = 0, dU = 0,
q = 0, and w = 0. As vaporization of the ether proceeds spontane-
ously, the enthalpy, pressure, and entropy of the system all increase and
the temperature decreases until equilibrium is reached at a condition
where the entropy reaches a maximum and dS = 0.
Instead of restraining the system to conditions of constant volume and
internal energy, the container might be provided with conducting walls
in contact with surroundings of the same and constant temperature.
The conditions of restraint would then be constant temperature and
volume. Thus, dT = 0, dV = 0, and p dV = 0. By fitting this con-
tainer with a frictionless piston the process could also be restrained to
constant pressure. For all such changes of state under any conditions of
restraint, equilibrium is reached when the fundamental requirement that
dq = T dS is satisfied for the system itself. A combination of this
requirement with the first law jdelds the equation T dS = dU + p dV +
d'wf, which may therefore be taken as a criterion of equilibrium. This
equation is applicable only to reversible changes, and, conversely, if it is
applicable to all possible changes of a system these changes are all
reversible and the system is at equilibrium. If the system approaches
equilibrium under the conditions of restraint of constant entropy and
volume, . " ji._ .
dS =0; dV = 0 and ^
{dU),y = -d'Wf (14)
or, if no means for performing useful work is present, d'Wf = 0
{dU)sv = 0 (15)
Equation (15) is a criterion of equilibrium, for a system restrained to
constant entropy and volume and unable to perform useful work. Such
a system may be visualized by considering ether evaporating in an
atmosphere of nitrogen in a container with rigid conducting walls in con-
tact with a reservoir of heat of variable temperature. As evaporation
takes place, the temperature of the reservoir is so controlled that heat is
withdrawn from the system in such a manner that the reduction in
temperature offsets the increase in entropy resulting from the irreversi-
ble vaporization, and constant entropy is maintained in the system.
Under these conditions vaporization will proceed spontaneously until

equilibrium is reached at a condition of minimum internal energy.
Since U = H - pV = A + TS = G - pV + TS, Equation (5a) may
be written in terms of other thermodynamic energy functions. Thus,
since d'q = TdS and d'we = pdV,
TdS = dH-Vdp + d'wf (16)
It follows that in a system restrained to conditions of constant pressure
and entropy it is a criterion of equilibrium that dH = —d'wj or if no
useful work is possible dH = 0.
idHU = 0 ' " ; - ) ; (17)
Similarly, Equation (5a) may be written
dA + pdV + SdT + d'Wf = 0 (18)
Thus, at conditions of equilibrium restrained to constant volume and
temperature where no useful work is performed ' : • '
idA)^y = 0 ., , (19)
Also, Equation (5a) may be written / ; .^
dG - Vdp + SdT + d'Wf = 0 / (20)
It follows that at conditions of equilibrium restrained to constant tem-
perature and pressure it is a criterion of equilibrium that
{dG)Tp = -d'wf (21)
or if no useful work can.be performed ,
(d(?)rp = 0 '• , (22)
Valid criteria of equilibrium under still other conditions of restraint
might be similarly derived. However, that represented by Equa-
tions (21) and (22) is of the greatest interest because of the fact that the
majority of chemical equilibria are restricted to conditions of constant
temperature and pressure.
From consideration of Equations (14) through (22) it is evident that
the proper criteria of equilibrium in- a process is determined by the con-
ditions of restraint under which it occurs. If the evaporation process
used as an example is started from constant initial conditions but
allowed to proceed under different conditions of restraint, the conditions
at which equilibrium is reached will be different.
The various criteria of equilibrium form the basis for the prediction
of equilibrium conditions in any process. For example, if a chemical
CHAP. XI] THERMODYNAMIC RELATIONSHIPS 453
reaction proceeds spontaneously under restraint of constant temperature

and volume, the total work function A decreases as a function of n, the
number of moles converted. Equilibrium will be reached at a minimum
value of A where dA/dn = 0 if no useful work is performed or where
dA/dn = d'wf/dn if useful work is performed. Since A is a property
determined only by the state of the system, it follows that the state at
which equilibrium is reached is dependent upon d'wf/dn, the amount of
useful work performed per mole of conversion at equilibrium conditions.
Similarly, if a reaction proceeds under the restraint of a constant tem-
perature and pressure, the change in free energy G is the basis of the
criterion of equilibrium from Equations (21) and (22). The most
valuable applications of these energy functions result from their funda-
mental relationships to the important criteria of equilibrium.
Thermodynamic Relationships. The thermodynamic properties of a
system or a process may be classified into four groups as reference proper-
ties, energy functions, derived properties, and path properties.
The reference properties, temperature, pressure, volume, entropy, and
composition, are those required to define completely the state of a
system. Of these properties temperature, pressure, and composition are
intensive; volume and entropy are extensive. The energy functions are
extensive properties whose values are known only relative to some arbi-
trary reference state. Such properties as specific heat, coefficient of
expansion, coefiicient of compressibility, and Joule-Thomson coeflScient
are classified as derived properties.
All of these properties have intrinsic values which are determined by
the existing state of a system and are independent of the course followed
in arriving at that state. Properties such as these in which changes are
dependent only upon the initial and final conditions of the system and are
independent of the path followed in producing the changes are termed
point properties.
As previously pointed out, heat and work including mechanical, radiant
, and electrical forms are not properties of a system but are manifestations
of changes occurring within the system and as such may be considered
properties of a process rather than of a system. From the previous
discussions it is evident that these manifestations are dependent upon the
particular course of the changes and may differ widely in two processes
even though the initial and final properties of the systems are identical.
For this reason these properties of a process are termed path properties.
When any two of the reference properties, temperature, pressure,
volume, and entropy, of a pure substance or of a substance of fixed
composition in all phases are specified, the relative values of other point
properties are definitely established. With reference to one selected
pair of properties, such as pressure and temperature, the magnitude of a

third property such as vohime may be represented as a surface in space.
If extended to all conditions of pressure and temperature of a pure sub-
stance with its three states of aggregation, solid, liquid, and gas, the
surface developed assumes a contour of great complexity. Upon such a
contour surface, for example, with pressure and temperature as inde-
pendent variables and voluine as the dependent variable, equal values
of all other point properties may be designated by lines extending over
the contour surface. Separate sets of lines may be traced respectively for
equal values of entropy, internal energy, specific heats, compressibilities,
and so forth. The relationship of points on one path to those on another
is definitely fixed by the contour of the surface.
If a mathematical equation can be established for the surface in ques-
tion in terms of p, V, and T, then relative values of all related properties
may be established by mathematical methods. If the equation is
unknown or too complex for mathematical manipulation, the desired
relationships may still be established by graphical methods. Obviously,
the number of possible relations among the dozen or more different
properties runs into millions even for first-order differential equations
and becomes unlimited for all possible second-order differentials. Ordi-
narily, only the more important relations are developed in which some
desired property is expressed in terms of properties which are easily
measured, such as temperature and pressure.
In dealing with solutions and mixtures the foregoing discussion must
be expanded to include the variable of composition. To represent the
properties of a solution of fixed composition a contour surface can be con-
structed as for a pure compound, but in covering the entire range of
compositions a graphical presentation by a contour surface is no longer
adequate, and mathematical methods must be employed.
In developing the thermodynamic relations which follow, it will ordi-
narily be futile to attempt to visuahze the relations involved.
Differential Energy Functions. If consideration is limited to systems
of constant mass and composition and to reversible processes in which
the only external force is pressure, differential equations for the four
energy functions are developed as follows from their definitions and from
Equation (1):
dU = d'q - d'We = TdS-pdV (23)
dH = dU + dipV) = dU + pdV + Vdp (24)
dG = dH - d{TS) = dH - TdS - SdT (25)
dA = dU - d{TS) ==dU - TdS- SdT (26)
CHAP. XI] EXACT DIFFERENTIAL EQUATIONS 455
By combination of the foregoing, four basic differential equations are

obtained which express the energy functions in terms of the four reference
properties, p, V, T, and S. Thus,
dU = TdS - pdV (27)
dH = TdS+Vdp (28)
dA = -SdT - pdV (29)
dG = -SdT+ Vdp (30)
Although these relations are developed by consideration of a highly
restricted type of process, it may be noted that-Equations (27-30) con-
tain only terms which are intrinsic point properties of matter. Accord-
ingly, these equations are generally applicable to a mass of matter of
constant composition undergoing any type of physical change. The
relationships between intrinsic point properties are not restricted to
reversible operations and are unaffected by accompanying changes in
external forms of energy.
Exact Differential Equations. Where M is a single-valued continuous
function of several independent variables x, y, z, • • •, the total differen-
tial of u can be expressed in terms of its partial derivatives with respect
to its different independent variables, thus:
du = (f) dx + (f) dy + (f) dz + ... (31)

\dx;y,^... \dy/x,z... \^Z/x,y...
The properties of a substance fulfill the requirements of u, and Equa-
tion (31) can be extended mathematically to include any number of
variables. The value of the integral of du between any two conditions A
and B of a relationship represented by Equation (31) is dependent only
upon the location of these two points and is independent of the path
followed between the two states. For a case involving only three vari-
ables, this situation is evident from consideration of !Fig. 55 (page 286).
Thus,
X du = UB - UA = j{x, y) (32)
Since UA and UB are wholly defined by the corresponding values of x

and y, the value of the integral is unaffected by the path considered in
the integration. An equation of the type of (31) is termed an exact
differential equation.
If, however, any line is drawn on the surface of Fig. 55 connecting A
and B, the length of the line will be dependent upon the path followed.
A differential equation expressing the length of line will involve expres-
sions in addition to those defining the surface and is termed an inexact

differential equation. For the element of Fig. 55 the inexact differential
is represented by the wavy line d'u.
It is a property of an exact differential equation involving two inde-
pendent variables z and y that
dy\dx)y~ dx\dy):, • .^^

or
- ^ = ^ ^ '' • (34)
dy dx dx dy
If more than two independent variables are involved, a relationship
similar to Equation (33) may be \yritten with respect to any pair of
independent variables. Any exact differential equation must satisfy the
relations of Equation (33).
For example, if du is expressed by the relation
du = Mdx + Ndy +Pdz (35)
where M = f(x), N = f{y), and P = f(z), then Equation (35) is an
exact differential equation only provided
\dx/y^' \dy/J \dz/y^

Under these conditions,
/dM\ dû {'dN\ dû
\dyJ:, dxdy \dx )y dxdy \ dy A \dx Jy
Similarly,
/dM\ ^ /dP\ ^ fu_ /dN\ _ /dP\ _ _a2M_
\dz)^~ \dx A dxdz ^^ \dz)y' \dy J^" dydz ^ '
If in the exact differential equation, " / '
du = (~ ) dx + (~) dy (39)
\dx/y \dy/x
the condition is imposed that the property u is constant, the equation
becomes
or
W „ " ~ \^x)y/ \dy}^ ^^^^

CHAP. XI] MAXWELL RELATIONS 457
Equation (39) may be differentiated with respect to x while any other

property such as w is held constant. This gives, upon elimination of
differentials of zero magnitude,
Equations (37), (38), (41), and (42) represent mathematical princi-

ples commonly used in the development of thermodynamic relations.
Maxwell Relations. Since Equation (27) is an exact equation of the
form of (31) the relationships of Equations (37) and (38) permit the
derivation of a differential equation relating the coefficients T and p as
shown in Equation (43). Similar equations result from equating the
partial derivatives of the coefficients of Equations (28), (29), and (30).
Thus,
Equations (43), (44), (45), and (46) are termed the Maxwell relations.
Equation (45) is the basis of the familiar Clapeyron equation. For
the vaporization of a liquid the restriction of constant volume may
be omitted, because the vapor pressure is independent of volume.
Since vaporization is a reversible process at constant temperature,
(a5/aF)r = A / r A 7 and
^ = _A_ = ^
. ^__ (47)
dr T AF r(F« - Fi) ^ '
which is the form of Equation (1), Chapter III, page 59.
Illustration 3. From a plot of the following properties per pound of superheated
ammonia, show that
using the reference conditions, p = 173 lb per sq in. abs, t = 185°F, S = 1.300, and
V = 2.20 cu ft '
AtS = 1.300 At F = 2.20

V ( P S- t P
2.461 165.2 150 1.2466 114.8 150
2.339 174.4 160 1.2705 145.1 160
2.229 183.0 170 1.2935 186.3 170
2,200 185.0 173 1.3000 186.0 173
2.132 191.3 180 1.3154 208.0 180
2.043 199.2 "l90 1.3388 240.3 190
1.960 206.9 200 1.3575 283.1 200
In Fig. 98, p is plotted against S at

S = entropy V = 2.20, and T is plotted against V
1.24 1.28 1.32 1.36 1.40
T
\
• at /S = 1.300.
The value of (dT/dV)s is the slope of
A
r plotted against VntS = 1.300, or
, 200 190
^
160 - 210
= -83.3.
1190 180 h ( dVJs 2.50 - 1.90
S=1.S The value of {dp/d8)y is the slope of p
y=2.2 0 ^ /
g-iso 170 ll' plotted against S a,t V = 2.20. This
Si, slope is (200 - 150)/(1.36 - 1.25) =
/ s = i . 30 \ 454. To convert this slope into the
20 \
'170 F=2 160 same units as (dT/dV)s requires multi-
/// \ pUcation by 144/778, where 144 is the
/I
//
11
\\ conversion factor for pounds per square
\ inch to pounds per square foot and 778
2.0 2.2 2.4 2.6
V=volume represents the number of foot-pounds
equivalent to 1 Btu. The result is 84,
FIG 98. Evaluation of {dTldV)s and
which is in agreement with the Maxwell
{dpldS)v. relation.
Partial Derivatives of Energy Functions. By partial differentiation of

the four differential energy functions Equations (27-30) while one of the
terms appearing as a differential is held constant, there results:
(48) (49)
\dp)s
(50) (51)
\dp/T
/dif
(52) (53)
\dS/pt / p
(54)
(fa=- (55)
CHAP. XI] THERMAL CAPACITY RELATIONS 459
By equating identical properties,

/dU\ /dH\
/dU\ /dA\
By again partially differentiating the four basic differential energy-

function equations with respect to the coefficient terms, another set of
eight relations results:
(a=-(a <-> (a=-<a <-'

' ( a - ( a « (!).=-©.«
Thermal Capacity Relations. One of the most important applications
of thermodynamics is to establish the relations of physical properties
and energy changes from a minimum amount of easily determined experi-
mental data. The thermal capacities of a substance are among those
most frequently measured experimentally. There are four useful types
of thermal capacities, designated mathematically as follows, for opera-
tions in which all changes of state are reversible:
Cp = {—-] = heat capacity at constant pressure (68)
* ' C'« = (-—I = heat capacity at constant volume (69)
Ip = \ — ) = latent heat of pressure change at (70)

constant temperature
latent heat of expansion at constant (71)

\dVjT
temperature
The thermal capacities are independent of the manner in which heat is

added but are restricted by definition to reversible changes of state.
The first two thermal capacities are those most commonly used. The
latent heat of volume change at constant temperature includes the
usual latent heats of fusion and evaporation, and Ip corresponds to the
heat of compression.
Since entropy is a point function, it can be expressed by the following
exact differential equations: -
Combining (72) and (73) with Equations (68), (69), (70), and (71) gives
At constant temperature, from Equations (75) and (77),

KdV/r T\dV/T T ^ ^
h=T[^^) ' (78)
From (46) and (78), / /
• ^^=-KS). '^' - ^''^

From (45) and (79), • ' /• *
For an ideal gas, Equations (80) and (81) reduce to

l*=-V and I* = p
CHAP. XI] (Cj,-C,) and (Cp/C„) 461
Substituting the values of Equations (74-77) into (72) and (73) yields
rfS = ^ c i r + | d p (82)
•• ; V dS = ^dT + ^~dV (83)
If Equation (82) is applied to conditions of constant entropy and Equa-

tion (80) is substituted
\y p Vp
Similarly, from Equations (83) and (81),

C^_ _L (dV\ _ _ /dp\ /dV\
:* T TKdTjs" \dT)v\dT)s ^^^^
{Cp — Cv) and (Cp/Cv). By equating Equations (82) and (83), an
expression for the difference between the heat capacities at constant
pressure and volume may be derived. Thus,
Cp — Cv = ly-pj^ — l'p~pfj (86)
or, in terms of partial derivatives, restricted first to constant pressure

and then to constant volume,
Combining (80) or (81) with (87) gives
By method of Equation (41),
• (a=-(a/(a-
Combining (88) and (89) gives
The ratio of the heat capacities K at constant pressure is obtained by

combining Equations (84) and (85): -•
/dV\ /dp\
Op KdT/AdTh ^ _ {dV\ /dT\ /dp\
{^\ (^J\ \dTJp\dpJv\dVjs ^^
KdT/yKdTJs
Effects of Pressure and Volume on Heat Capacity. The variation with
pressure of the heat capacities Cp and C„ at constant temperature is of
special importance because of the difficulties of calorimetric measure-
ments at pressures other than atmospheric. Apphcation of the princi-
ple of Equations (35-38) to (82) results in ., .,
_ ' •. (92)
/T \dT/p
'" \
Substituting (80) in (92) gives
(tX=-KS); ^ <»?)
Similarly, from (83) and (81), /
PVT Relationships of the Energy Functions. The effects of pressure,

volume, and temperature upon the four thermodynamic energy functions
are among the most useful of relations. The more important relations
are developed herewith. Others may be developed in a similar manner or
more readily by Shaw's method as'discussed on page 465.
A differential expression for internal energy is obtained by combining
Equations (27), (83), and (81): - /
dU=C.dT + ^T[^-^J -pjdV mi-"]

Application of Equation (95) at constant temperature and at constant
(95)
volume, respectively, results in
a-(a- ^
(
CHAP. XI] PVT RELATIONSHIPS OF ENERGY FUNCTIONS 463
Similarly, combining Equations (28), (80), and (82) gives
,H.C.4T + [^T(^1+V].dp (98)
Applying Equation (98) to constant temperature and constant pressure

operations, respectively, results in
/dH\ ^^ ^ /dV\
• • ( ^ . = "•• "°"'
One of the most important relations expresses the variation in free-
energy change with change in temperature at constant pressure. From
Equation (5),
AG AH ^^ •
Is.) -1 (^\ _ ^ _ {^^ (102)

KdT /p~ T\dT )p r KdT/p
By Equation (62),
(103)
combining (102) and (103) gives
— ] - _^ (104)
Illustration 4. From van der Waals' equation of state calculate the increase in
Cp of COj where the pressure is increased from 1 to 100 atm at 100°C. The van der
Waals equation is
(p + ^ ) ( „ _ 6 ) = /jr (a)
where v = molal volume

a,b = van der Waals' constants, characteristic of the gas
For 1 lb-mole, Equation (93) is written
Differentiating the van der Waals equation gives

dv\ R -' /
( dT/p ~ _ ^ , 2ob . ' ^"^
KerOp lip- av-' + 2abv-'yj Kar/p ^^^

Integration of (b) gives
• ''"'--J^ H^J/^ /^
Equation (e) is solved by substitution of (d) and graphical integration where,
a = 3.60(10«); b = 42.8; R = 82.1 (cc-atm)/(g-mole)(°K)
Volume V is first obtained from Equation (a) at various values of p. Equations (c)
and (d) are then solved in turn at various values of p.
p V
atm cc \dTjp Varv

0 00 00 0
1 30,690 82.7 0.64
10 2,985 8.55 0.734
2& 1,150 3.65 0.931
60 530 2.09 1.52
75 320 1.67 2.8
100 213 1.56 5.07
From graphical integration,
Ac„ = 194.20 (cc-atm)/(g-mole)<°C) = 4.7 oal/(g-mole)(°C)
"'' Cp = 9.6 + 4.7 = 14.3^
niustration 6. Prove that Cp — C« = where

a
^ = — ( —— I ' = coefficient of expansion

V \dT/p __ I
Q : = — — ( — ) = coefficient of compressibility
V \dp JT
From this relation calculate C, for mercury, where at 0°C, /3 = 0.00018 per °C,.
a = 0.000 003 9 per atm, Cp = 0.0333 cal/(g) (°C), and the density = 13.596 g per
cc. From (90), /
Substituting a and /3 gives
Cp Cv — -xr
—aV a
CHAP. XI] SHAW'S METHOD OF DERIVATION 465
For mercury at 0°C,

(273) (0.000 18)2
= 0.0040
(13.596)(0.000 003 9)(41.3)
where 41.3 is the conversion fkctor from cubic-centimeter-atmospheres to calories.
Hence,
C, = 0.0333 - 0.0040 = 0.0293/oal/(g)(°K)
Shaw's Method of Derivation. The usual procedure in making ther-

mod3Tiamic transformations to arrive at a usable or desirable relation is
time-consuming and often results in circuitous procedures which return
to the starting point. A. N. Shaw^ has devised a means of circumventing
these difficulties and arriving at the desired result in a systematic manner.
This method employs a system of abbreviated mathematical notations
termed Jacobians. A Jacobian is a determinant, the elements of which
are partial derivatives.
It will be recalled that the solution of simultaneous algebraic equations
can "be accompTis'hed'by the use oi determinants. ? o r example, consider
the two simultaneous equations:
(105)
aix + biy = ci
The values of x and y satisfying these two equations will be found to be

X = C162 — Cibi
ai&2 — 0 2 ^ 1
(106)
aid — UiCi
y
aib2 — aibi
These solutions can be arrived at systematically by expressing the results
in determinant form, thus:
cibi1
C2&2
X =
ai6i
O262
(107)
aiCi
a2C2
y = aibi
a2&2
The lower determinant in each case is the principal determinant con-

2 A. N. Shaw, Phil. Trans. Roy. Soc. A, 334, 299-328 (1935).
sisting of the four coefficients of the unknowns arranged in a square

ai6i
the value of which is obtained by cross-multiplying terms, thus,

aM — a^bi. The determinant in the numerator is derived from the
principal determinant by replacing the coefficient of the unknown sought
with the constant c; thus, for x the a terms are replaced by c terms.
dix tf)
The Jacobian . , '' „; represents the following determinant,
d{a, 13)
/dx\ /dx\
^{x, v)
d{a, &) (dy\ /dy\ \da)s V/SA Va/SA \da)s ^^^^^
\dJs V/3A
where x and y are two dependent variables and a and /? are any two inde-
pendent variables of which x and y are functions. For example, x might
be enthalpy; y, entropy; a, temperature; and |3, pressure. The values
of a and jS may equal x and y. The foregoing Jacobian is also written in
d(x, y)
the form J(a;, y) where J{x, y) = Where this abbreviated nota-
d{a, /3)
tion is used without designation of a and )3 these variables may be any
independent variables of which x and y are functions. However, if
several Jacobians, J{x, y), J(x, z), and J(z, w) appear together in a single
expression the values of a and /? must be the same for all. With this
restriction Jacobians may be combined algebraically in any desired
manner,
The following four properties of Jacobians are the basis of their use in
thermodynamic transformations:
1. Interchanging the order of any two terms in a Jacobian changes its
sign, thus, ' ! . . ' , ' -
d{x^y) _ d{y, x) _ d(y, x) _ dix, y)
a(a, ff) a(a, /3) d% a) m«)
or
J(x, y) -J{y,x) (110)
d{x, x)
2. 0 or J{x, x) =Q (111)
J{y, z) J{z, y) J{y, z) J(z, V)

3. (112)
J{x, z) J(z, x) J{z, x) J{x, z)
4. The exact differential equation,

dx = Mdy + Ndz
may be written as
Jix, a) = MJ{y, a) + NJ{z, a) (113)
The first three relationships can be estabhshed readily by setting up in
determinant form with a and fi as the independent variables and expand-
ing in the usual manner. The fourth relationship may be proved
readily by differentiating the exact differential equation with respect to
/3 holding a constant, thus.
The principle of Equation (112) combined with (114) gives

,f:. J{x, a) _ MJ{y, a) NJ{z, a) }
Equation (115) is identical with Equation (113).

Thermodynamic equations are most frequently expressed in terms of
p, V, T, and >S as the independent variables. Any of the six possible
pairs can be chosen as the independent variables represented by a and /S.
The Jacobians of the six possible pairs of independent variables are
designated as follows:
J{V, T) = a
J{p,V) =b
> • JiP,S) =c [ :_ (116)
• • ' J(P, T) = I
J{T, S) = Jip, V) = b
The proof of this last relationship is derived by writing Equation (43)
in Jacobian form in accordance with Equation (112), thus,
^ ( y , 'S) _ j(p, V) ^ jjp, V)
. ; J{V,S) JiS,V) JiV,S) ^ '
or
J{T, S) = J{p, V) (118)
Using the symbols of Equation (116) Shaw developed Table X X I I I
from which the Jacobian J(x, y) of any pair of thermodynamic proper-
ties X and y may be obtained in terms of the six possible Jacobians of the
four independent variables p, V, T, and S. The properties considered
are p, V, T, S, U, H, A, and G. In addition, to, the work of expansion
468 THERMODYNAMIC PRINCIPLES^ [CHAP. XI
-o
> s s 8 a. fl.
o Si Si.
+e o
1 1 fa. =0
1
1
a.
1 CO
o
j S
f5
a, (/
> E^
o ?i. .* 1 .o + o
E.H EH s
1 CO
&^
1 =0
1
o ft.
EH
;^
1 1
es 1 1 o 1 e
«3
1
1
i - a.
!? 1
A ^
ft.
«2
+
e-i S2 1
« + +
El 8
a, ^ + +a,
'S '^ n. + ++ O
CO 8
ft.
CO
^r 1 <X5 fX3 +
St-. ^1 1 ta 1
&H CO
s 1
in
< y ^
1a.â. rf3 N
« ^ ^ :^ 1 ++
» 0 ~o
1 ^ EH â. H
l-^
tt! E^ 1 1 1 ^PS o ;^ 1
< Co ^
52 + 1^ 1 ^ 1
1-5 tM t s a.
1 + 1 ^
O 1
s
g rO s
o
S1 ^ ^
^ t) + o 1 + s
^ a.
a. 1'
1
si1 1
'l><
+ EN a.
XI a. ^
1 +
i-:i
Vi W s J5 o fi,
ts o 8
ft.
M 1 1
<
e t-
1
55, 8
h ~ s O 1 +
1
^ EH
1 a
1 1
W5
e 8 s
fr- -o o 1 1 ^ +8 o
1 1
1 t L ^ ' ' ^ '"^ C2>
5H f i aa £., CQ
a. ^ ^ 3 ^ U*
+ E^ a.
Sl. o 1 T 1 1 + s§ 1 II 11 II II II
1
,_^
1 s—» Si ^ e^ to to tt3 ^ ;3 &I ^
done by and qr the heat added to a system undergoing a reversible

process are included. The Jacobian J{x, y) of any pair of properties is
obtained by entering the table at the values of x in the left-hand column
and y in the top row. For example, iix = H and y = T,
Jix, y) = J{H, T) = -Tb + Vl= -TJ{p, V) + VJ{p, T) (119)
Table X X I I I was constructed from the differential energy functions,
Equations (27) to (30). For example, by the fourth Jacobian rule
Equation (28) may be written as follows with T replacing a as the inde-
pendent variable:
JiH, T) = TJ{S, T) + VJip, T) = -Tb + Vl (120)
The other Jacobians are developed in a similar manner.
The Jacobians taken from Table X X I I I may be modified to include
any selected pair of independent variables a and /3 in accordance with
Equation (109), or may be transformed according to the four rules of
Equations (109-113) and expanded according to Equation (108) to
obtain any desired relationship. For example. Equation (119) may be
modified to include p and V as independent variables a and /3. Thus,
applying Equation (109) to Equation (119) gives
'MJ^=^T+V'-^^^ (121)
d{p,V) ^ d{p,V) ^ ^
If desired. Equation (121) might be expanded in accordance with
Equation (108) to yield a complete relationship among H, T, p, and V.
A similar relationship among the same variables but of different form
results if p and T are chosen as the independent variables instead of p
and V. In this manner a limitless number of relationships may be
developed. Where a specific relationship is sought in terms of specified
variables, i t is desirable to simplify the expressions and eliminate
unwanted variables before expanding the Jacobians.
Illustration 6. From Table XXIII evaluate (,dG/dT)p. This expression may be
written in Jacobian form in accordance with Equation (112) and the Jacobians
evaluated from the table.
J{G, p) SI
KdTjj,p J{T,p) -I ^
This relationship may be transformed into other variables by replacing S by its
equivalent in Jacobians from Table XXIII. Thus,
J{A, V) __ J{A, V) /dA\
S =
a ~ j(v, T) ~ VarA
In order to eliminate an undesirable independent variable, use is made

of the following important relationship: y . • •,
b^ + ac - In = 0 (122)
This relationship can be proved as follows: From Equation (72),
J(p, T) dS = J{S, T) dp - J{S, p)dT ' (125)

Differentiation with respect to any variable w, with V held constant, gives
Expansion in accordance with Equation (112) results in f

Jip, T)JiS, V) = JiS, T)J{p, V) - J{S, p)J{T, V) (127)
or, substituting symbols for the Jacobians gives /• , - ••
lC-n) = {-b)b - i-cX-a) /
or
• ¥ + ac - In = 0 - . (128)
Illustration 7. From Jacobians, prove tliat
Rewriting in Jacobians gives

TJip, V)J{V, p) TJ{V, pW{p, T)
(b)
J{T, V)J{T, p) J(T, pyj{V, T) 1
1
Substitution of values from Table XXIII gives ' /
Tb{-b) -Ti-b)H /i ,
.(c)
(-a)(-0 (-O'a /
¥T b'T
(d)
al al
Illustration 8. It is desired to evaluate {dp/dV)j} in terms of the independent
variables p and T and if possible to eliminate iS. Equation (112) and Table XXIII
give
<dp\ ^ J{p, H) ^ Tc
KdVjH J{V, H) Tn ~ Vb ^^'
CHAP. XI] SUMMARY OF THE MOST USEFUL RELATIONS 471
The Jacobians c and 6 each include one of the independent variables and may be
expanded directly into a partial derivative in terms of p and T. However,
n = J(V, S) contains neither of the independent variables. It is therefore desirable
to eliminate this quantity by means of Equation (128). Thus,
Tc Tel
Tn-Vb~ Tb^ + Tac - Vbl ^^
The symbols may now be replaced by their corresponding Jacobians, thus,
(dp\ ^ TJjp, S)J(p, T)
KdVjH TJ(p, Vr + TJ{V, T)J{p, S) - VJip, V)J{p, T) ^'^'
Since p and T are the independent variables, this last equation becomes, by
(109),
/dp\ \d{p, T))\d{p, T)J

'ff , /dip, Toy /d(V, T)\ /dip, S)\ _ /dip, V)\ /dip, T)\
KdiV, T)J "^ \d{p, T)J\dip, T)J \dip, T)Adip, T)J
The like terms in (d) are canceled, and the remaining Jacobians written as partial
derivatives according to Equation (112):
(dp\ ^ \dTjp
(e)
KdVjH
KdTjp ^ KdpJAdTjp KdTjp
/dS\ C
Since ( —;;: ) = v ' Equation (e) reduces to
\dl/p 1
Cdp\ ^ Cp
(f)
, , ., , /SwN R / av\ V
For 1 mole of an ideal gas I —-; ) = — ; I—I =
\dTjp V Vp/r P
hence,
(g)
\SVJH
KS-©-"© V
Equation (g) is in agreement with the requirement that for an ideal gas H is a function
of temperature only. For this requirement
\dv/H \dv/T V
Summary of the Most Useful Relations. It cannot be said that one

thermodjTiamic relation is more significant or more important than
another, but it is found that relatively few relations suffice for most
thermodynamic problems. A brief selection of these more useful rela-

tions is presented in Table XXIV. .'
TABLE XXIV
TBEBMODYNAMIC RELATIONS
Equation Number
in Text
)ns
Differential Energy Functions
(a) dU = TdS -pdV (27)
(b) dH ^ TdS + Vdp ' '% (28)
(e) dA = -SdT-pdV /
(d) dG = -SdT + Vdp (30)
Maxwell Relations
'•>(a--(S)K (43) £5
(44)
(45)
*'(a--(i-a : (46)
(56)
(57)
(58)
Energy-Function Derivatives
(59)
(74)
(75)(80)
(76)
Thermal-Capacity Relations
(77) (81)
(82)
CHAP. X I J THERMODYNAMIC TEMPERATURE SCALE 473
TABLE XXIV (cmL)

Equation Number
in Text
(r) dS = - ^ r - + - j T - (83)
»^'-<'--KS);(a=- (90)
(*^ c! = ~ \df)pw)vwjs =" (91)
<»>(t),=-(s;). (93)
M(S^). = K I * ) , • '"-'
(94)
Effect of p, V, T on f/, H, and G •
(-)(a = Kf?).- (96)
(^)(S).^^^ (97)
<^)(f).—(a (99)
(^)(ii) =(^^ (100)
The Thermodynamic Temperature Scale. The properties of an ideal

gas are defined by the requirements that pv — RT and that the internal
energy is a function of temperature only and is independent of volume.
It has been experimentally demonstrated that real gases approach this
behavior at low pressures. The gas-temperature scale is defined from
the behavior of an ideal gas as T" = pv/R and has been numerically
evaluated by the extrapolation to zero pressure of observations with a
constant-volume hydrogen thermometer.
All of the general thermodynamic relations developed in the preceding
sections were defined without reference to the character of the tempera-
ture scale on which the definition of entropy is based. If it can be
shown that any of these relations lead to an expression involving tem-
perature which can also be derived from the ideal-gas law, this fact will
constitute proof of the identity of the thermodynamic and ideal-gas
temperature scales.
From Equation (w) of Table XXIV,
Since Equation (129) relates p, V, and T, it may be termed a thermo-

dynamic equation of state. If this equation is apphed to an ideal gas,
for which (d[//rfF)r = 0
(130)='
Differentiation of the ideal-gas law gives
or
Since Equations (130) and (131) are identical, it follows that the
thermodynamic and the ideal-gas temperature scales are similar func-
tions, both having the same zero point.
Thermodynamic Properties of an Ideal Gas. As previously men-
tioned, the properties of an ideal gas are defined by the requirements that
the pVT relations are exactly expressed by the equation pv = RT and
that the internal energy is dependent only upon temperature and is
independent of pressure and volume. Equations whose applicability is
restricted to ideal gases are designated by asterisks (*) following the
equation numbers. Similarly an asterisk following any thermodynamic
symbol indicates restriction to the ideal-gaseous state.
By differentiation of the ideal-gas law, it follows that
•pdv + vdip = RdT (132)*
or • .j ' ,. . • jv!mw:<qw(ft^v-ni•.;-• :'•:
m ^ ^ ^ l and p ) ^ - ^ and (^\ = ? = ; (133)*

\dT/v V T \dv/T V \dT/p p T \ '
From (s), Table XXIV, ', :
• ;
The differential change in internal energy may be written as //
'iu = [~] dT + [^)^dv " .;;; (135)

\dT/y ^ V
By substitution of Equations (w) and (x) of Table XXIV, there results
dU = C.dT + ^T ( 0 \ - pi dV (136)
CHAP. X I ] THERMODYNAMIC PROPERTIES OF IDEAL GASES 475
/dp\
For an ideal gas P = T I—-] , and Equation (136) becomes for 1 mole
\dl /v
dv = c^dT (137)*
or ,
• ^> u = / c,dT + Vo (138)*
Since for an ideal gas {d^p/dT'^)v = 0, it follows from Equation (v),

Table XXIV, that cj is independent of volume and pressure.
By combination of (98) with (133) and (134) on the basis of 1 mole,
dn = CpdT = cdT + RdT = (c, + R) dT (139)*
or
c,dT + RT + n<i • :'• - '^'• •'' (140)*
e/O
From (n), Table XXIV, and (133),
h= = -V ".' • (141)="
P
Combining (141) with (q) of Table XXIV gives
ds = Cpd In T — R d \n p
or '
As = f 'c^dlnT - R]n- (142)*
Jr, pi
It is evident that Equation (142) cannot be directly integrated with
lower limits of either T = 0 or p = 0, since the integral of each term at
these limits becomes negative infinity. In order to evaluate directly the
entropy of an actual gas in the ideal state from heat-capacity measure-
ments it is necessary to start \vith the crystalline solid at a temperature
near the absolute zero, heat it to its fusion point and then to its boiling
point T' under a pressure p' sufficiently low that ideal-gas behavior is
obtained. The entropy s' of the saturated vapor at this temperature
and pressure is calculated from the low-temperature heat capacities of
the crystalline and liquid state and the heats of fusion and vaporization.
Equation (142) is then integrated from this limit. Thus,
CpdlnT - R\n-,
+ s' (143)*
JT'
p
A perfect gas may be defined as one which behaves ideally at all con-
ditions and hence is incapable of condensation to either a liquid or a soHd
phase. It follows that for the internal energy of such a gas to be inde-
pendent of volume at all conditions of pressure and temperature it can
possess only translational energy of motion and its heat capacity must be
independent of temperature as well as of pressure, as shown by Equa-
tion (VII-19), page 211. Thus, for a perfect gas, c„ = fi2 and Cp = %R,
whereas the heat capacity of what has been termed an ideal gas may
vary with temperature. It is evident that no substance fulfills the prop-
erties of a perfect gas but that perfect behavior is approached by mona-
tomic gases at low pressures.
For a perfect gas, Equation (142) may be integrated as follows:
s = #filnr-i?lnp + 6 (144)*
The integration constant h has been evaluated by methods of statistical

mechanics as (f 72 In ilf — 2.298) where M is the molecular weight. i?
Illustration 9. Assuming that nitrogen behaves ideally over the range indicated,
calculate the values of s*, u*, H*, A*, and Q* for 1 lb-mole at 10 atm, and 100°F, rela-
tive to conditions at 1 atm, and 60°F. Assume c„ = 5.0.
Au* = c^^T = 5.0(40) = 200 Btu
AH* = (c» + R) AT = (5.0 + 1.99) (40) = 279.6 Btu ^
AS* = Cj, In -? - S In — = 6.99 In -— - 1.99 In — = -4.07 entropy units

Fx pi 620 1
It will be evident that numerical values cannot be secured for AA and AG unless
absolute values of entropy are known.
The absolute entropy of nitrogen at 60°F and 1 atm pressure is 45,56 Btu/(lb-mole)
(°R). The final entropy after heating and compressing to 100°F and 10 atm, from
Illustrations, is45.56-h As* = 41.49.
Then, since AA = Au - A(rs) = Au - TaSa -\- TiSi,
AA* = 200 - (560(41.49) -|- (520)(45.56) = 666
Similarly, i' i " I '
AG* = AH* - TasJ -|- TiS? = 279.6 - 23 234 + 23 700 = 746
.i PROBLEMS / '
1. (o) Calculate the increase in entropy when 1 lb of ice at 0°F and atmospheric
pressure is converted into steam at 220°F and 1 atm pressure.
(6) The absolute entropy of water vapor as an ideal gas at 25°C and 1.0 atm
is 45.13 oal/(°K) (g-mole). Using the data of part (a), calculate the absolute entropy
of ice at 3 2 ^ in Btu/(°R) (lb). Steam at 220°F and 1 atm may be assumed to behave
ideally.
2. The volume coefficient of expansion {dV/dT)/V of water at 100°C is 0.00078
per °C. Calculate the change in entropy in oal/(gram)(°C) when the pressure is
reduced from 100 to 1 atm at 100°C, neglecting the effect of pressure on volume.
CHAP. X I ] PROBLEMS 477
3. Derive the differential equations for the following relations:

(a) Change of total work function with temperature at constant volume.
(6) Change of enthalpy with entropy at constant pressure.
(c) Change of internal energy with volume at constant entropy.
(d) Change of temperature with volume at constant entropy.
4. Calculate the latent heat of vaporization per pound of water at 160°F from the
following data:
^ = 0.113 lb/(sqin.)rR)
Volume of saturated vapor = 77.29 cu ft per lb and of liquid = 0.017 cu ft

per lb.
5. From selected data on the properties of superheated steam taken from Keenan
and Keycs, " Thermodynamic Properties of Steam" (John Wiley & Sons, New York,
1936),
(o) Verify the relation dU = T dS - p dV.
(6) Verify any one or more of the four Maxwell relations.
6. Assuming ideal behavior, calculate the entropy of 10 lb of hydrogen gas at
60°r and 10 atm relative to 0°F and 1 atm.
7. Evaluate the following coefficients per pound-mole of an ideal gas:
'"' (1). "> (SOv \dv/T <^' © .

'" (m. '" (g). xdv/r « ©.
«&x «(%X (^) ( ? ) /aT\
\dp/T
» &l
«(t\ (o) Cp — Cv (P) Ip (?)
8. For a van der Waals gas (see Illustration 4, page 463), evaluate the following
coefficients:
« ( I X <« (S). <" CTX «' (£).
9. By combination of Equation (47) with Equation (III-16), page 73, evaluate

the heat of vaporization of methylamine at its normal boihng point, — 6.5°C, assum-
ing that the vapor obeys the ideal-gas law and neglecting the volume of the Uquid.
10. From Table XXIII, express
(a) Cp — C„ in terms of p, V, and T.
(6) ( — ) in terms of V.
\dp/s
( - ) i in terms of G and p.
\dp /S
478 THERMODYNAMIC PRINCIPLES [CHAP. XI.
(d) ( -^ ) _ in terms of p and V.
(e) {-;;;) in terms of Cp and T.

(:dTJp
11. From Table XXIII derive equations for the following quantities in terms of the
indicated variables:
% • \
Quantities Variables
«(a S,p .
»(iX -S ^:'
«o. 4, r
<* Q . S,T
12. From Table XXIII verify the following equations:

/'dp\ (ST.
(a) Cp-Cv =-T\
(b) (Cp - CJ
Vary J, VaF/r
^""^ C, "" " W A W/F wA

13. Calculate the values of As, Au, AH, AA, and AG when 1 lb-mole of Ha gas at
77°F and 1 atin is heated and compressed to 500°F and 100 atm. The initial absolute
entropy of hydrogen is 31.23 Btu/(lb-raole) (°R). Assume ideal behavior.
14. From the relationships of problem 8 and *rable XXIV, calculate the following
changes in the properties of 1 lb-mole of CO2 gas when the pressure is increased iso-
thermally from 1.0 to 100 atm at 100°C. a = 3.60 (lO^) (atm)(ce)V(g-mole)2;
6 = 42.8 cc per g-mole.
(a) AH, (6) Au, . (c) As, (d) ACp, (e) Ac,. /
CHAPTER X I I
THERMODYNAMIC PROPERTIES OF FLUIDS
The ideal-gas law permits satisfactory calculations of pressure-
volume-temperature relationships at low pressures where the volumes
per mole are relatively large and the distances between molecules great or
where the temperatures are relatively high. However, under conditions
of small molal volumes, corresponding to high pressures, the errors in
assuming ideal-gas behavior may be as great as 500 per cent.
Actual Behavior of Gases. Several hundred equations of state have
been proposed to express the pVT relationship of gases, but none has
been found universally satisfactory, and most are applicable only to a
single gas over a limited range of temperatures and pressures. The
most generally satisfactory equation of state is that of Beattie and
Bridgman wherein the various constants must be determined uniquely
for each gas. The van der Waals equation is presented because of its
value in giving an elementary understanding of the reasons for departure
of gases from ideal behavior and because of its support of the theory of
corresponding states which serves as the basis of 'the most useful general-
ized graphical method. However, the van der Waals equation is less
accurate than the generaUzed graphical method, and its application is
rarely warranted.
The actual pVT relationships of carbon dioxide are shown graphically
in Fig. 99. Each curve represents the relationship between the pressure
and the molal volume of carbon dioxide at the indicated temperature.
The critical point is indicated by C. The double-crosshatched area
represents the region of the liquid state. The plain area is the region of a
homogeneous fluid state which at low pressures is recognized as a gas but
at high pressures has continuity with the liquid. The single-cross-
hatched area represents a region in which both the liquid and gaseous
states are present in equilibrium with each other. Thus, following along
the 21.5°C experimental isothermal line from right to left, an increasing
pressure is required to cause a reduction in volume until the saturation
curve CB is reached. At this point the attractive forces between the
molecules become sufficiently greafto start condensation. The volume
may then be diminished without further increase in pressure until the
curve CA is reached. This curve represents the completion of condensa-
tion into the relatively incompressible liquid state. A further decrease
in volume must be accompanied by a large increase in pressure.
479
480 T H E R M O D Y N A M I C P R O P E R T I E S OF F L U I D S [CHAP. X I I
Isotherms of CO2
Experimental
Van der Waals
Simple Gas Law
100 200
Molal Volume, cc. per gram mol
F I G . 99. Isotherms of Carbon Dioxide
Van der Waals' Equation. In the derivation of the simple kinetic

theory, each molecule is considered to have an available free space, in
which it may move about and which is assumed to be equal to the total
volume occupied by the gas. This assumption is not correct except
under such conditions that the volume of the molecules or particles them-
selves is negligible as compared to the total volume occupied by them.
Actually, in each mole of gas there is a space of volume (v — b) available
for free motion, somewhat less than the total volume. The correction b
by which the available free volume per mole differs from the total volume
per mole is dependent on the volume actually represented by the N
molecules themselves. It is this free volume which must be used in the
more rigorous derivation of the kinetic gas laws. ;.
CHAP. XII] VAN D E R WAALS' EQUATION 481
TABLE XXV
VAN DEB W A A L S ' AND CRITICAL CONSTANTS OP G A S E S
From the data of the International Critical Tables
( cc \^
g-mole/
1; 6 =
cc
g-mole
tc = critical temperature, degrees centigrade
Pc = critical pressure, atmospheres
dc = critical vapor density, grams per cubic centimeter
Vc = critical volume, cubic centimeters
lb / ft' \2
T o convert values of a to -(in.;
—- I ) multiply values in table by 0.003776
L.)^ \ l b - m o l e /
ft'
To convert values of 6 to multiply values in table by 0.0160
lb-mole
tc Pc
Argon 35 X 10" 32.3 -122 48 0,531 75,2

Acetylene 37 X 10" 51.2 36 62 0,231 113
Air 33 X 106 36.6 -140 37,2
Ammonia 19 X 106 37.3 132 111.5 0,235 72,4
Carbon dioxide 60 X 10" 42.8 31 73,0 0,460 95.5
Carbon monoxide. 46 X lO" 39.4 -139 35 0.311 90.0
Chlorine 6.50 X 10« 56.2 144 76.1 0,573 124
Ethylene 4.48 X 10" 57.2 60.9 0.22 127
Hydrogen chloride 3.65 X 10" 40.8 51 81.6 0.42 87
Hydrogen 0.245 X 10' 26.6 -239 12.8 0,0310 64.5
Methane 2.25 X 10" 42.8 -82 45.8 0,162 99
Methyl chloride... 7.50 X 10" 65.1 143, 65.8 0.37 137
Nitrogen 1,347 X 10' 38.6 -147 33,5 0,3110 90
Oxygen 1.36 X 10« 31,9 -118,8 49,7 0,430 74
Sulfur dioxide 6.80 X 10<i 57.2 157.2 77.7 0,52 123
Water 5.48 X 10« 30,6 374 217,7 0,4 45
Another factor, neglected in the simple kinetic treatment of a gas, is

the attraction existing between molecules, known as van der Waals'
forces.
These forces tend to draw the molecules together and diminish the
pressure exerted below the value corresponding to ideal behavior. It
may be demonstrated from kinetic theory that this reduction is inversely
proportional to the square of the molal volume. Thus
P ^ -,72
or p' = p + -
where p' = pressure calculated from the simple kinetic theory

•p = actual pressure
482 THERMODYNAMIC PROPERTIES OF FLUIDS [CHAP. XII
The term p' is the internal -pressure of the gas. The corrected equation
of state then becomes ^''
h) = tiVmw^ = RT (1)
('<i)<'
This is the equation of van der Waals, which represents the general
form of the pressure-volume relationships of a gas, even when compressed
to the region of liquefaction. Although this equation is an improvement
over the ideal-gas law, its numerical results represent only a fair approxi-
mation where molal volumes are small. The factors a and h, characteris-
tic of each gas, are termed the van der Waals constants. In the first
two columns of Table XXV are listed values of a and h corresponding to
pressure in atmospheres, volumes per gram-mole in cubic centimeters,
and temperatures in degrees Kelvin.
Beattie-Bridgman Equation of State. One of the most widely used
equations of state is that of Beattie and Bridgman. 1-2
•pv = «.[. + ..(.-^)][i-^]-..(.-^) (2)

where a, 6, 4o, /?o, c = empirically determined constants
V= molal volume, liters per gram-mole
p = pressure, atmospheres
T = temperature, degrees Kelvin
.B = gas-law constant = 0.08206
This equation contains five constants which must be determined for each
particular system. The methods for evaluating the constants are dis-
cussed by Beattie and Bridgman in the papers cited and also by Deming
and Shupe.^
In Table XXVI are values of the constants of Equation (2) for several
common gases. In the ranges of temperature and molal volume indicated
in Table XXVI, the equation was found to jdeld an average deviation
of only 0.18 per cent from the accepted experimental values.
Equation (2) and the constants of Table XXVI may be used to calcu-
late accurately pressure, volume, temperature relationships over wide
ranges of conditions. Such an equation is of value for the extrapolation
of limited experimental data and for highly precise interpolation between
observed values. I t is also valuable in thermodynamic calculations
U. Am. Chem. Soc, 49, 1665 (1927).
2 J. Am. Chem. Soc, 60, 3133 (1928).
5 / . Am. Chem. Soc, 52, 1382 (1930); 53, 843, 860 (1931).
CHAP. XII] BEATTIE—BRIDGMAN EQUATION OF STATE 483
TABLE XXVI
BEATTIE-BRIDGMAN CONSTANTS
Mini-
Tempera-
mum V
Gas Ao a J5O 6 c ture
cc/-g
Range, °C
mole
He 0.0216 0.05984 0.01400 0 . 0 0.004 X 10< 400 to - 2 5 2 100

H2 0.1975 - 0 . 0 0 5 0 6 0.02096 - 0 . 0 4 3 5 9 0.0504 X 10" 200 t o - 2 4 4 100
N2 1.3445 0.02617 0.05046 - 0 . 0 0 6 9 1 4.20 X 10^ 400 to - 1 4 9 180
Oz 1.4911 0.02562 0.04624 0.004208 4 . 8 0 X 10" 100 to - 1 1 7 110
Air 1.3012 0.01931 0.04611 - 0 . 0 1 1 0 1 4.34 X 10" 200 to - 1 4 5 125
CO2 5.0065 0.07132 0.10476 0.07235 66.00 X 10" 100 t o 0 180
CH4 2.2769 0.01855 0.05587 - 0 . 0 1 5 8 7 12.83 X 10" 200 t o 0 166
because it expresses compressibility data in a rigorous equation which

may be incorporated in any desired mathematical operations. I t s use
in compressibility calculations is cumbersome, but once the constants
are evaluated for a particular gas a complete set of compressibility
factors may be calculated and used in the manner described in the
following sections.
Benedict-Webb-Rubin Equation of State. An empirical equation of
state with eight constants has been formulated by Benedict, Webb, and
Rubin* for the lighter hydrocarbons from the experimental data for meth-
ane, ethane, propane, and n-butane; thus, if d is density, ^(-moles/liter.
p = RTd + (BORT - AO - Y2) d' + ( ^ ^ ^ - «)^' + ««^'
(1 + yd') -yd?
+ cd' y2
(2a)
TABLE XXVII
CONSTANTS OF THE BENEDICT-WEBB-RUBIN EQUATION OF STATE
Units: Atmospheres, Liters, Gram-Moles, Degrees Kelvin; R = 0.08207.
Methane Ethane Propane n-Butane
Bo 0.0426000 0.0627724 0.0973130 0.124361

Ao 1.85500 4.15556 6.87225 10.0847
Co 10-8 0.0225700 0.179592 0.508256 0.992830
b 0.00338004 0.0111220 0.0225000 0.0399983
0.0494000 0.345160 0.947700 1.88231
c 10-8 0.00254500 0.0327670 0.129000 0.316400
0.0060000 0.0118000 0.0220000 0.0340000
0.000124359 0.000243389 0.000607175 0.00110132
" M.3enedict, G. W. Webb, and L. C. Rubin, / . Chem. Phys., 8, 334-5 (1940).

Values of the constants for these four gases are recorded in Table XXVII.
This equation holds within 0.34 per cent, eySn at gas densities twice the
critical value.
COMPRESSIBILITY FACTOR
The equation of state may be written
pV = znRT ' (3)
where z is termed the compressibility factor and is a function of pressure,

temperature, and the nature of the gas.
If values of the compressibility factor of a gas are known, all calcula-
tions involving its pVT relationships may be carried out by simple
proportionalities derived from Equation (3). Thus, applying Equa-
tion (3) to a given mass of gas at two different conditions and combining,
gives
- PiVi ziTi :
(4)
P2V2 Z2T2
The correct normal molal volume at 0°C and 1 atm is equal to 3593s
cu ft per lb-mole or 22AlZs hters per g-mole, where z, is the compressi-
bility factor at standard conditions. With a knowledge of compressi-
bility factors it is thus possible to extend the entire system of calculation
which was used with the ideal-gas law to apply at any desired conditions
of temperature or pressure. The ideal-gas law may be considered as
representing a special case in which the compressibiHty factor is equal to
unity.
The nature of the variation of the compressibiUty factor z with pres-
sure, molal volume, and temperature is shown by the curves of Figs. 100,
101, 102 for nitrogen. In the range covered by these three charts, the
compressibility factor of nitrogen may be found at any specified con-
ditions by interpolating on the proper chart. If the pressure and
temperature are specified, the compressibiUty factor is read from Fig. 100.
If the molal volume and temperature or the pressure and molal volume
are specified, Fig. 101 or Fig. 102 is used.
The use of compressibility-factor data is demonstrated by the follow- •
ing illustrations, dealing with nitrogen. /
- • • ••!•• • . , . . •', - /
Illustration 1. One cubic foot of nitrogen at 50°C and 30 atm is compressed to 60

atm and cooled to —50°C. Calculate the final volume.
The compressibility factors at the initial and final conditions may be obtained from
Fig. 100.
CHAP. XII] COMPRESSIBILITY FACTOR 485
1.05 " - _-__— .. _ _ _ _ _— - - —"r-r"7"r- - - — -
" 11 ]"!"
1.00 ^s^M==—^---'^—^=^^^-"-"-—--f°°c.-:i
-~=^^S-?;8-î;---^^-î-;;J^- ^^ ^ ^f
0.95 S'*'!.^"^'^''-*."*'" ^^—••. " 1—1—1—U-y

^.•*."^"*-'^"*-». "" — — ^ '~25°C.'
s"^*""^ '•-.. *"'••. ~"~— III
^s'^ "^-^ '"=^^ '•^^^ ^-^_ 1
^ 0.90 ^ ^^_ __ __^ ^„.^ Jnl r^~~--
S "^1:* *'^ ^'•^ """• 50 C
'>!__ '^ "•|_j ^"-^ ~h"l~î.L
S "^^ "'- ""li R(\<> C~^
*S *'i_ •* • * LI r *1
""s ••- "-^ Ij-kkl

^^ '^V
""^^ ** V
""'TTT nn^ r^^ ^ -- ^
"^ V "'• .
S, "^^
•v
0.75 ^ j^ '^ >- --'..

— 90 = C.-*^
—100° C . ^
0.70 1 1 1 1 1 *'^—. — B = ^
10 20 30 40 50 60 70 80 90 100
\ I Pressure, Atmospheres
FIG. 100. Compreissibility Factors of Nitrogen at Specified Pressures and

Temperatures.
i.os -
1 U 1 _J_
110° c. !
1.00 -
1° f
, •+— - 1 • - - " [ i '•
° c 1
— ^ • " ' " "
L '
I A i\n .,'
'
/ JrX
>o» c.
1
1 AAo
n oc
I A QA /
i
f1
- --
1
/ iL qi
0.70 _ 1 12000
1 2250
0 250 500 750 1000 1250 1500 1750 2500
Volume, cc. per gram mol
FlG. 101. Compressibility Factors of Nitrogen at Specified Molal Volumes and

Temperatures. - ,
Initial conditions: , . '.

pi = 30 atm j / • -j-i--
Vi = 1.0 cu ft
Ti = 50°C =323°K '' •%'',•'
zi = 1.001 "* .
Final conditions:
Pa = 60 atm
Ti = -50°C = 2 2 3 ° K '
22 = 0.930
From Equation (4):

^- VlPlTiZ^
ViTiZi •
/ 3 0 \ / 2 2 3 \ /0.930\
Illustration 2. Calculate the pressure in pounds per square inch to which nitrogen
must be compressed in order that 1.0 kg at 60°C may be contained in a cylinder hav-
ing a volume of 10 liters.
Prom Fig. 100 it is found that e,, the compressibiUty factor of nitrogen at standard
conditions, is equal to 0.999. The volume of 1.0 kg or 35.7 g-moles of nitrogen at
standard conditions is then (0.999) (35.7) (22.41) or 801 liters.
Final conditions:
V = molal volume = 10/35.7 = 0.28 liter
T = 323°K
2 = 1.015 (from Fig. 101) ^ ' . .
From Equation (4): •* .;,,,.;,..; ^•
PiT^.7'2Z2 \ , ^ / 8 0 1 \ / 3 2 3 \ /1.015\ ^ ^, ^ ,^
^^=-T^: = ''•' U ; Km) w ; ='''' '"^ ^^ '^ - -
Illustration 3. A steel cylinder having a volume of 5 liters contains 400 g of nitro-
gen. Calculate the temperature to which the cylinder may be heated without the
pressure exceeding 50 atm.
' • Moles of nitrogen = 400/28 = 14.30 g-moles
Molal volume = 6000/14.30 = 350 cc / \
y> = 50 atm ' , " ' ,
z = 0.945 (estimated from Fig. 102) ,
R = 82.1 . .!
From Equation (3): '• //
T^ (50)(350) ^ ,--. /
(82.1) (0.945) /
A large amount of experimental work has been carried out on the more
common and industrially important gases. In the International Critical
Tables are experimental data for many gases and extensive references to
the literature. ^naicKVmm'fi^ ' .. -^X
• • . • " • • ; ' • ' • K • -
CHAP. X I I ] COMPRESSIBILITY FACTOR 487
Experimental-compressibility data are commonly presented in tables
or curves showing values of the product pV at various values of p for
constant values of T where the product is taken as 1.0 at 1 atm pressure
and 0°C. These data are arranged to show the isothermal variation of
pV with pressure. If Zs is the compressibiUty factor of the gas at 0°C
1.05
1,00
0.95
.0 10 20 30 40 50 60 70 8.0 90 100
Pressure, Atmospheres
FIG. 102. Compressibility Factors of Nitrogen at Specified Pressures and Molal

Volumes.
and 1 atm pressure, then the compressibility factor of the gas for any
other condition may be obtained from the following:
273
J....,.K z = z,{pV) (5)
The value of Zs is nearly unity for the common gases. It may be

obtained accurately from density measurements made at standard con-
ditions from the relation, . ,,,.,:;
M
Zs = (6)
Ps22.41
where M is the molecular weight and Ps is the density in grams per liter
at 0°C and 1 atm pressure.
The compressibility factor of all gases is unity when the pressure is
zero. Thus, values of Zg may also be obtained by plotting the values of
pVagainst p at 0°C and extrapolating the graph top = 0. Then
PsV^
Zs = (7)
POVQ
where poVo is the extrapolated value of p 7 at p = 0.

Experimental-compressibility measurements are also frequently

expressed in the form of isometric data showing the variation of pressure
with change in temperature of a known weight of gas confined in a con-
stant volume. Such isometrics are determined at various selected
volumes. These data yield curves of the type of Fig. 102 directly.
Curves of the types of Figs. 100 and 101 may then be derived by inter-
polation methods. ,
In statistical mechanics values of the compressibility factor are
expressed as a series function in terms of reciprocal molal volume, thus,
, z = 1 + ^(r) i + y{T) (-)' + ••• (7a)
where 0{T), y(T), • • • are functions of T and are designated as the

second, third, • • • virial coefficients. The first virial coefficient is unity
in agreement with the ideal behavior of gases at zero pressure where
virial coefficients above the first disappear.
Generalized Compressibility Factors. In comparing the physical
properties of different substances it has been found that similar behavior
is encountered at the same values of reduced temperatures and pressures.
• The reduced properties are related to the critical properties of a sub-
stance as defined on page 54. Thus,
T V , • V'
?^r = — ) Pr == — and Vr=^~ (8)
where the subscript r indicates the reduced and c the critical properties.
This similarity in the behavior of a substance at equal values of reduced
temperature and pressure is referred to as the theory of corresponding
states and forms the basis of the generahzed procedure described in the
following pages for establishing the deviations of thermodynamic proper-
ties from ideal behavior. The van der Waals and other equations of
state may be shown to be consistent with and to support this theory.
If the theory of corresponding states were rigorous, a single equation
of state would suffice for all gases if expressed in terms of reduced proper-
ties instead of in absolute values. In Fig. 103 are plotted values of
compressibility factors as a function of reduced temperature and reduced
pressure. This chart was derived as an average of data reported in the
literature for carbon dioxide, nitrogen, hydrogen, ammonia, methane,
propane, and pentane, and is not in rigorous agreement with all the data
on any one of these substances. However, these discrepancies, which
result from the fact that the theory of corresponding states is only an
approximation, are small enough so that for many purposes the chart
may be taken as applicable to all pure gases. It then constitutes a con-
CHAP. XII] GENERALIZED COMPRESSIBILITY FACTORS 489
11 -; o „
O
rAi -IIN CO •*tO00,-lrH /

^\^ /
^^ V ^
o
^
v\\\ '
/ oc
11
^
CO
o
0)
i
-V ^ V v|\ \\ //
// 2
9
&^^' \\ i\ a
o
m
m
V
^ A\ i u
1
(LI
1 -
ip
.-H
o
o
-1 -00 I
>
1 " o
« i O
^
o
0 2 ^ 0
00 !:•
o • to 1
'
\ \\ l\ O
' \ ^^
1 \ 4^N^"^ \i O
1 1
(U
2
? 1 \\ CO e,^
o
o
T \^
1
i
r*
CO
E
h /
iH
T3 I
\N O
^ 5 U
/ ^- c \ ^ 3
T3
'1 U5
/r-i
11 1 /<z o
/ c •
\a
J11 '/ / 00
o
I'l '/r^
1 i
I/ »CO
L
U
P
o
III
W
00 <0 Tt CO
o • • • • •
CO 1 O O O O o
490 THERMODYNAMIC PROPERTIES OP FLUIDS [CHAP. XII
venient method for handling pVT calculations wherever the critical

temperature and pressure are known.
Figure 103 is a modification prepared by Watson and Smith^. of a simi-
lar chart presented by Dodge.^ It was pointed out by Newton^ that in
applying relationships of this tj^je to hydrogen and helium better agree-
ment is obtained by using modified reduced conditions calculated from
the equations: '
Tr = T/{T. + 8); pr = p/ivc + 8) (9)
where T is in degrees Kelvin and p is in atmospheres.
In calculations involving gases at high pressure, Fig. 103 is em-
ployed in combination with Equation (3). Values of V/n may be
obtained directly when values of p and T are known. This represents
the usual case. If either temperature or pressure is unknown, the pro-
cedure requires modifications. These three cases are illustrated in the
following problems.
Illustration 4. Volume Unknown. Calculate the volume occupied by 1 lb of

methane at a temperature of 40°F and a gauge pressure of 1000 lb per sq in.
From Table XXV, page 481, the critical temperature of methane is —82.5°C or
— 116.5°F, and the critical pressures 45.8 atm or 673 lb per sq in.
„ , , 40+460 ' ,
Reduced temperature = = 1.40
-116.5+460
^ , , 1000 + 14.7 1
Reduced pressure = = 1.51 '
673
Compressibility factor (Figure 103) = 0.84
Molecular weight ' = 16.00
R 10-1 ^'^^^^*^'
•"""" (in.)Hlb-mole)°(R)
/ \
hence
- (-'L •
nzRT 1(0.84) (10.71) (500)
= 0.278 cu ft
P (16) (1014.7)
Illustration 5. Pressure Unknown. Calculate the pressure necessary to compress
100 Uters of nitrogen at a pressure of 745 mm Hg and 23°C to occupy a volume of 1.0
liter at -110°C.
Pc — 33.5 atm j
Tc = 126°K • ; . .
^ K. M. Watson and R. L. Smith, Natl. Petroleum News, July 1936.

= B. P. Dodge, Ind. Eng. Chem., 24, 1353 (1932).
' R. H. Newton, Ind. Eng. Chem., 27, 302 (1935). i
CHAP. XII] GENERALIZED COMPRESSIBILITY FACTORS 491
0 982
Initial pr == —— == 0.0293
OO.O
296
Initial Tr = = 2.35
'; ' 126 '
1 ' 4i
Initial z == 1.00
163
' Final Tr = = 1.29
126 '
' ''':*• il -'V.' ^ A .
- Pi
Final pr =
' 33.5
From Equation (4): A
;,, PlFl zj\ .,

pzVi Z2T3
Substituting the known values gives
(0.982) (100) 1.00 / 2 9 6 \
• ' 1 ^ P^d) Z2 V163/
V I 22 = o.0185p2 = (0.0185) (33.5)pr, = 0.621p,^
The straight line corresponding to the equation, Zj = 0.621pr, is extended across

the compressibihty chart, Fig. 103. Since Fig. 103 is plotted on double logarithmic
scales, the Une must have a positive slope of 45° and pass through the point z = 0.621
where pr = 1.0. Where the 45° hne intersects T, — 1.29, the value of 22 is found to be
0.80, and the value of pr = 1.30. Hence
p = (1.30) (33.5) =43.6 atm
Illustration 6. Temperature Unknovm. The volume occupied by 1 lb of n-ootane
at 27 atm is 0.20 cu ft. Calculate the temperature.
tc = 565°F (1025°R) pc = 24.6 atm
27 T
• ; ; ' ' • ^' = ^ = 1-^0 ^ = T5^
0.729 (atm) (cu ft) .
R = —7- , ,^„^^ ; molecular weight = 114
(lb-mole) ( R)
From Equation (3):
(27)(0.20)=?^°^
114
845 845 0.825

z = T ~ \Q26Tr ~ Tr
To solve the problem graphically 2 is plotted in Fig. 104 against Tr for a value of
p, = 1.10 from Fig. 103. On the same scale is plotted the equation 2 = -^——•
Where these two curves intersect, 2 = 0.72 and Tr = 1.15.
Hence T = 1025r, = (1025) (1.15) = 1180°R or 720°F.
The generalized compressibility-factor correlation offers the most con-

venient method available for estimating pVT relationships whenever
critical-temperature and critical-pressure values are available. Several
remarkable properties in the behavior of gases are shown by Fig. 103.
The compressibility factor is less than unity for all values of Tr below 2 at
reduced pressures below 8. In this range actual gases are more com-
pressible than ideal. At vaiues of pr above 8 the compressibility factor
of all gases exceeds unity regardless of temperature. In this range all
V'^.^J
NO.6
1.4 1.6
Tr
FIG. 104. Graphical Solution for Tr.
gases are less compressible than the ideal. At a value of p^ = 8, at all

values of Tr the compressibihty factor is approximately unity. The
greatest deviation from ideal behavior occurs near the critical state where
the gas becomes almost five times as compressible as in the ideal state.
Care should be exercised to avoid use of gaseous compressibility factors
at pressures higher than the vapor pressure at reduced temperatures less
than 1.0. Such conditions produce liquefaction to which Fig. 103 is not
appHcable.
Based upon the generalized chart for compressibility factors corre-
sponding charts have been constructed for deviation from ideal behavior
of other thermodynamic properties such as enthalpy, entropy, heat
capacity, and the Joule-Thomson effect.
Effect of Pressure on Enthalpy of Gases. It has been shown that the
enthalpy of an ideal gas is independent of pressure and a function only of
temperature. However, at elevated pressures all actual gases deviate
from ideal behavior and enthalpy changes with change in pressure,
particularly in the region of the critical point.
CHAP. XII] EFFECT OF PRESSURE 493
The enthalpy of a real gas relative to the enthalpy of an ideal gas when
both are at the same temperature may be obtained by integration of
Equation (y) of Table XXIV, between the limits of the existing pressure
and zero pressure where all gases behave ideally and the enthalpy
becomes independent of pressure. Thus,
(^* - ^)^=IKS), - ^]/p (10)

Where the equation of state is known, the values of V and (dV/dT)p
may be expressed as functions of pressure and the integration of (10)
may be carried out analytically or graphically.
The value of (H* — H)T may also be expressed in terms of the com-
pressibility factors as a variable and thus obtained from these factors
for a specific material or from an approximate, generalized chart such as
Fig. 103. Differentiation of Equation (3) gives
•' (a=-:hKa] • <-

Combining Equations (y) of Table XXIV and (11) results in
If experimental-compressibility data are available, values of {dz/dT)p

may be calculated and Equation (12) integrated graphically.
Expressed in reduced properties,
/dK\ 1 ^ _ RTm/ dz\
KdVr/TPc PrVcTAdTr/^ / ^^
Integration, at constant temperature, gives
The derivative {dz/dTr)p is obtained by graphical differentiation of a

cross-plot of Fig. 103, on regular coordinates, and Equation (14) is
then integrated over the pressure range required at various constant
temperatures. Values of (H* — H)/TC are plotted in Figs. 105 and 106
in terms of reduced conditions. These charts permit corrections for
enthalpy to be made readily and with usually sufficient accuracy. An
enthalpy-correction chart similar to Fig. 106 was first developed by
Watson and Nelson.* Figures 105 and 106 have been corrected to repre-
8 K. M. Watson and E. F. Nelson, Ind. Eng. Chem., 25, 880 (1933).
10
9
1 )
8 1.
/
--41 >
7 \
--1.
6 \ — 1.
W lO
-1
1'
o
11h " U
s
OS 2
5 5* / / 'A _4
^ /
— 2.0^^
V
^
t,i 5 i 2.6.^
— 3. 0^
v»..-.>
•"6 0N
-2
N,
] \
-3 \
-4 ^ \
0.1 0.2 0.3 0.4 0.6 0.81.0 2 3 4 6 8 10 20
FIG. 105. Enthalpy Correction for Gases (High Range) '<f'U-'! it

sent an average of the data more recently calculated by Edmister' and
York and Weber. i° Data for the Joule-Thomson coefficient of water
vapor were used to extend the low-temperature ranges of the charts.
York and Weber'" found that in the temperature range above Tr = 1
the enthalpy corrections vary systematically ivith the critical tempera-
ture of the substance when correlated as in Figs. 105 and 106. Accord-
ingly they proposed that the values read from such charts should be
multiplied by a correction factor <^ defined as follows for use with Figs.
105 and 106, based upon pentane, T^ = 470°K:
/
/
^ \47l
470/
y (15)
where Tc is the critical temperature of the substance in degrees Kelvin,

and n is a function of reduced temperature as follows: /.
Tr 1.0 1.05 1.1 1.2 1.3 1.4 1.5 1.6
n 0.37 0.28 0.25 0.20 0.18 0.16 0.15 0.14
5 W. C. Edmister, Ind. Eng. Chem., 30, 352 (1938).
w R. York and H. C. Weber, Ind. Eng. Chem., 32, 388 (1940).
CHAP. XIII EFFECT OF PRESSURE 495
1 1 1 1 1 11 1
M to S
383 14wC
i_ A
\ \ ' Si a 2 i s
I \ ^ - 2 s § 1 i c
" ^71
1 \ \ VC=
c"
11 \ V Y
•^f\ 5 1 " ft fe ^
J] y \ ^ > V
2\l \
- ^
0
^3
K
_
rS
"-r
"-• ^
ft >>
ft
\\ \ A'N^ ^ »^ s ^ s -^
11 ^. <?
JLIv*'^-
^t>h
/^s^
^N
^
U „ 1) .^ -
E4- a,^ e^"
B
s? a, ai
•" g
'>\
:tg
1 ^ x
•?6-p
( ' j n ~ — • O
1-5
"o.
^^
-*?? 3 d
'<?• ?! £
\,^ ^
^cK ^ f i Ci
0 ^0
U
0
^-'•r?
fc
0
°^ 0
t:^
>'. D.
•0 a1
\ 1—1
^C
-t^
o. 0 K
00
<.'?,- 0
0
0
«D T-<
0
^•*3 0 0
(^ 1^
^ ^
0
%' , 0 '
0
1
r>
•<'•p-0
IS
\
\ ^
1 o 10
o o
(ao)(aio«'-3)/(F3-s)HtH
496 THERMODYNAMIC PROPERTIES OP FLUIDS [CHAP. XII
Use of the correction factor (^ results in excellent correlation of data

for several hydrocarbons and steam, and it should be used where a high
degree of accuracy is required. Unfortunately 0 has not been evaluated
for temperatures below the critical. However, in many cases the entire
pressure correction to enthalpy is not large, and the uncorrected values
read directly from Figs. 105 and 106 are satisfactory for general purposes.
niustration 7. Calculate the value of (H* — BP)T for I g-mole of CO2 gas at a
pressure of 100 atm and 100°C.
The critical temperature of CO2 is 304°K, and its critical pressure is 73 atm.
y -373 ; ..
V""**^:*
From Fig. 106:
, H* — H
H* - H = (2.55) (304) = 775 cal per g-mole

Applicationof York and Weber's correction, n = 0.19, gives ''
H * - H = (0.92) (775) = 713 cal perg-mole •
Entropy of Real Gases. In terms of its partial derivatives with

respect to temperature and pressure, the entropy of a substance may be
expressed as follows:
or, combined with Equations (n) and (m), Table XXIV, • /

^„ C^dT /dV\ ^ !
By integration at constant temperature, . //
iS. - S*o)r = - £ {^\clp (18)
Equation (18) is not directly useful, because it expresses the entropy of a

gas relative to its entropy at zero pressure and is equal to negative infinity
at finite pressures. A useful expression is obtained by deriving the
CHAP. XII] ENTROPY OF REAL GASES 497
entropy correction due to deviation from ideal behavior at a given

pressure. For an ideal gas,
(-S*--So*)r=-j['(^pp (19)
The desired entropy correction is obtained by subtracting (18) from

(19):
From the ideal-gas law:

/dv\* ^ R
(21)
Hence, for 1 mole:
Equation (22) may be integrated if {dv/dT)p is evaluated from an equa-

tion or from Fig. 103. Thus, from Equation (3),
/dv\ zR RT/dz\
Substitution of (23) in (22) gives
Writing (24) in terms of reduced properties gives
(s* ~ S^)T = B £ (I - z) d In pr + RTr £ ' { ^ d In p, (25)
(s* - Sp)r = R I (l-z)dlnpr + ^-f-^ (26)
Equation (26) may be evaluated by graphical integration of a plot of

(1 — z) against In pr derived from Fig. 103. Figure 107 was plotted from
values determined in this manner and added to the corresponding values
of (H* - n)/T,Tr obtained from Fig. 106.
Because of the large effect of pressure on entropy and the complex
nature of the deviations from ideal behavior, it is customary to tabulate
THERMODYNAMIC PROPERTIES OF FLUIDS [CHAP. XII
- N :R V* L.^
- rA- 1- - -rJ k. 4
'"' Nj \ \
T \
X i\
^^
1 ^
V
V ^ \^;H
\
^--^
k
96 *(\ —
s^>.
06-n,
V
V
^^lL
^•') e
o
s «^ 1
%>>
•w m ffl CO
0 g £ M
g 0) E at ^ ^^ T)'
IN <ss\
rJ
-3
Tj
b
-d
P
-s
d
ft
(S
t»,
S*
r^
P< "c s
9- r^
"^
t^ ;\
\S
o
t-t
TS
0)
pj
I ". l'.-^ -B & S

t 1 W U 1' 4)
^ •sV
-5?-'
A I A " A 'k
E-i PH ti ?^ m m y-\
Or
p
'T 0 ^ -
?. 7,
p
O^
^ y^ 'ô
-%. -
I'^v \ s-
s N / /
•%> \
K \
(3o)(91o«i-3)/(l«=-3)'"'t*si[s)
O
.4
00
o r
CHAP. XII] EFFECT OF PRESSURE UPON Cp 499
absolute entropy data for gases at a pressure of 1 atm in a hypothetical

ideal state. These values express the entropy Sp=i,o which the gas
would have if it behaved ideally at pressures up to 1 atm. The entropy
of the actual gas at 1 atm pressure must be obtained from Equation (26).,
Illustration 8. The absolute entropy of carbon dioxide in the hypothetical ideal
state at a pressure of 1 atm and 25°C is 51.08 entropy units per mole. Calculate the
molal entropy at a temperature of 100°C and a pressure of 100 atm. The entropy at
100°C in the ideal state at 1 atm is first calculated from heat-capacity data. The
entropy at 100 °C and 100 atm, if ideal behavior is assumed, is then determined, and
from this value is subtracted the correction for deviation from ideaUty read from
Fig. 107.
From Table V, page 214, the molal heat capacity at 1 atm is
K ; r• 4 = 6.85 + (8.533)(10-=')r - {2A75){W-')T'
By substitution in Equation (17) and integration at constant pressure,
s*,^ _ s*r^ = 6.85 In j + (8.533)(10-=) (5^2 - Ti) - 1.237(10-«)(ri - TI) = 2.11
'v,'.|^ s*ioo=C, 1 atm) = 51.08 + 2.11 = 53.19

In cases such as this where relatively small temperature changes are involved, a satis-
factory solution may be obtained by assuming the heat capacity to be its value at the
average temperature. Prom Equation (XI-142) at constant temperature,
sj, - Sp = - B In —
or
SUOO'C, 100 atm) = 53.19 - 1.99 In 100 = 44.00
From Fig. 107, at values of Tr = 1.23 and pr = 1.37,
s* - s = 1.50
hence,
s = 44.00 - 1.50 = 42.50, the entropy of the actual gas at 100°C and 100 atm.
Effect of Pressure upon Cp. The heat capacity at constant pressure
of any substance is represented by the following differential equation:
From Equation (u), Table XXIV: V; , ^,
and r
By integration at constant temperature •;;
(29)
0.01 0.02 0.03 0.06 0.1 0.2 0.3 0.5 0.81.0 2 3 4 56 7

Reduced pressure, p^
FIG. 108. Molal Heat Capacity Correction for Gases.
Since for an ideal gas (dW/dT^)p = 0, the heat capacity of any gas is
independent of pressure at low pressures and Cpo = C*. Thus, for the
molal heat capacity of any gas,
If an equation of state is known, values of Cp — c* can be calculated by

integration of Equation (30). The pressure correction to heat capacity
iV
CHAP. XII] JOULE—THOMSON EFFECT 601
at constant pressure can also be expressed in terms of the compressibility-

factor. Writing Equation (3) in terms of reduced temperature and
pressure and differentiating, gives
(cj, - C%)T = - Tr J ' e dpr (32)
The function 6 can be evaluated in terms of reduced temperatures and

pressures, by graphical differentiation of Fig. 103. Graphical integra-
tion of this function in Equation (32) then gives values of (cp — c*)r as a
function of reduced conditions.
In Fig. 108 values of (Cp — c*)^ are plotted against reduced properties.
This is a modification of a chart developed by Watson and Smith, î
Dlustration 9. Calculate the heat capacity, of CO2 gas at 100°C and 100 atm
pressure.
From Table IV, page 212, cj, the heat capacity in the ideal state and at 100°C,
is 9.7. From Fig. 108, at Tr = 1.23 and pr = 1.37,
Cp — Cp = 0.0
Cp = 8.5 + 9.7 = 18.2
Joule-Thomson Effect. The differential temperature change result-
ing when a gas is expanded freely and adiabatically is known as the
Joule-Thomson coefficient ;u and is designated mathematically aS
M= (dT/dp)^ (33)
This coefficient can be expressed in terms of pVT relations by writing
Equation (XI-98) for conditions of constant enthalpy. Thus,
or
For an ideal gas T{dV/dT)p = V, and from Equation (35) M = 0. From

Equation (35) and Fig. 103 a relationship between fxCp and reduced
temperature and pressure can be developed.
" K. M. Watson and R. L. Smith, Natl. Petroleum News, July 1936.
Integration of (35) applied to 1 mole gives
<'''-«''=rc'.[''(5^),^"']* »«
Combining (36) and (10) gives i ,»; ^ v
' (T — T) = ^ ^ * ' ~ "pl)rCav8) — ( H * - Hp2)r(aYg) . _,

Cp(a.vg)
where Cj,(avg) is the average molal heat capacity over the range of tempera-
ture and pressure involved. Equation (37) is useful only for moderate
changes in temperature. For other cases a rigorous integration of Equa-
tion (36) is necessary. Problems involving estimation of such tempera-
ture changes are most conveniently solved from temperature-enthalpy
charts.
THE LIQUID STATE

All the exact thermodynamic relations developed in the preceding
sections for the gaseous state are equally applicable to the liquid state
and may be used for the calculation of thermodynamic properties and
their variations. Thus, from data relating the volume of a liquid to
pressure and temperature together with heat capacity data, it is possible
to calculate the effects of pressure and temperature on enthalpy, entropy,
and heat capacity by the same methods used for the gaseous state.
At reduced temperatures below 0.8 the molal volumes of liquids are
small and change little with pressure. As a result, the effects of pressure
on the enthalpy, entropy, and heat capacity of liquids are small and
frequently may be neglected. However, at high pressures, particularly
at temperatures approaching the critical, such effects become large and
must be carefully evaluated.
Generalized Liquid Densities. It was pointed out by Watson'^ that
the compressibility and thermal expansion data of a variety of liquids
which were investigated can be satisfactorily represented by the follow-
ing relation: . ,
-©- -
^' ; . .. ::. I.
(38)
or
V = (vicoi)/w (39)
i»K. M. Watson, 7nd. BJnjr. C/im., 35, 398 (1943).
CHAP. Xll] GENERALIZED LIQUID DENSITIES 503
where p or v = density or molal volume, respectively, of liquid at p,
and Tr
Pi or i;i = density or molal volume, respectively, of liquid at p^
and Tn
CO = expansion factor of liquid at pr and Tr
wi = expansion factor of liquid at p , and Tn
0.14 r • * ^
"^^ J*'^'^-*.
0.13 '-=5^^;
0.12
^ N^
^
^
0.11 ^ ^
^ : ^
0.10 ^ ^
0.0
^
g 0.09 ^
I 0.08 ^
f
T) -
' r
H
0.07
3.8 V\
0.06 0 94
1.0
0.05
Critical point
0.04 1^ .044->
0.4 0.5 0.6 0.7 0.8 0.9 1.0
Reduced Temperature
FIG. 109. Thermal Expansion and Compression of Liquids.
The factor co is a dimensionless quantity termed the expansion factor
and when expressed as a function of reduced temperature and pressure is
approximately the same for all liquids. Values of co are plotted in
Fig. 109.
The terms (pi/wi) or (viwi) are characteristic constants for any one
liquid which are established by a single density measurement and the
corresponding value of coi. Thus, from one known value of density
together with the w chart, Fig. 109, the density of that Hquid can be
estimated at any other temperature and pressure by means of Equa-
tion (38).
Illustration 10. The density of ethyl alcohol at 20°C and 1 atm is 0.789 g per ec.
Estimate the critical density and the density at 180°C and 100 atm pressure.
For ethyl alcohol, tc = 243.1°C, andp„ = 63.1 atm.

Temperature, ° C . . . . 20 .243.1 180
Pressure, atm 1.0 63.1 100
Tr 0.568 1.0 0.878
pr 0.0159 1.0 1.59
<o 0.127 0.044 0.0984
Since pi = 0.789 at 20°C anda'atm
^ = 5 : ^ ^ = 6.21 •
0)1 0.127
which is a characteristic constant for ethyl alcohol.
At 243.1°C and 63.1 atm p„ = (6.21)(0.044) = 0.273 g per oc
At 180°C and 100 atm p = (6.21) (0.0984) = 0.611 g per cc
The value of pc reported in the International Critical Tables is 0.275 g per cc.
For most of the substances investigated the deviations from

Equation (38) and Fig. 109 were less than 5 per cent. Larger errors are
encountered for water if the reference value is at a low temperature in the
range where water exhibits anomalous density-temperature relations.
In general the best over-all results are obtained from a reference value
corresponding to a reasonably high temperature, preferably near the
normal boiling point or higher.
At reduced temperatures below 0.65 and at pressures not exceeding
10 atm the expansion factor co is satisfactorily expressed by the equation:
w = 0.1745 - 0.0838^^ (40)
At higher temperatures and pressures Fig. 109 must be used. At the
critical point,
v.^ (z;iWi)/0.0440 (41)
Critical volumes or densities are calculated from Equation (41) with
considerably more accuracy than from Fig. 103 which is based on gas-
phase data.
In Table XXVIII are values of (pi/coi) and (viwi) for the normal
paraffins as developed by Gamson and Watson.'' The corresponding
critical properties, boiling points, and the constants of the vapor pressure
equation (III-16, page 73) are also included. It was noted by these
investigators that the values of (pi/wi) and (wiwi) are substantially inde-
pendent of isomerization. The lower paraffins, both normal and iso-
meric, are satisfactorily represented by the following equation:
" B. W. Gamson and K. M. Watson, Natl. Petroleum News, Tech. Sec, May 1944.
Also " Process Engineering Data," National Petroleum Publishing Company,
Cleveland (1944). ,
CHAP. XII] GENERALIZED LIQUID DENSITIES 505
Paraffins:
(t>iwi) = 1.88 + 2Mnc; cc per g-mole (42)
where
nc = number of carbon atoms
The observed densities for the normal paraffins above tetradecane tend
to be lower than calculated from Equation (42), but it is believed that
these deviations may result from the fact that the densities were meas-
ured close to the melting points. On this basis densities calculated from
the equation should be more reliable for high-temperature applications
than those based on the experimental values.
It was found that the density data of other hydrocarbon series may be
represented by similar equations in which the addition of a CH2 group is
accompanied by an increase of 2.44 cc per g-mole in the value of (îwi).
Thus,
Mono-olefins:
(?;iwi) = 1.1 -f- 2Mnc (43)
Monocyclic aromatics:
(yiwi) = - 3 . 0 -1- 2.44nc * (44)
For homologous series of polar compounds the increment in (wicoi)

for the addition of a CH2 group tends to vary in the lower members of the
series and approach a value of 2.44 with increased molecular weight.
This behavior is expressed by the general relationship,
(wicoi) = a + 2.44nc + m log nc (45)
whei3 a and m are constants characteristic of the series. For the hydro-
carbons m is equal to zero. For other series it may have a finite value,
either positive or negative. Thus, for the alcohols, both normal and
isomeric,
(wicoi) = 2.89 + 2.44nc - 1.23 log nc (46)
Similar equations may be developed for other homologous series.

It was pointed out by Gamson and Watson that equations of the type
of (45) may be used together with the methods discussed in Chapter III
(pages 68-72) for estimating critical temperatures as a means of predict-
ing liquid densities at all conditions without direct experimental-density
data. Conversely, if a single experimental-density value is available,
such equations may be used to estimate the critical temperature of a
homolog. Thus, combination of Equations (39) and (40) and solving for
M'<' rp ^ 0.0838 T-
r. i;'r; ^^yj
0.1745 - (?;iaji)/y
where
V = the molal volume at temperature T under conditions such that Equa-
tion (40) is apphcable ' •
Equation (47) may be used to estimate the critical temperature from
density data when the boiUng point is unknown. The boihng point
may then be estimated from the relationships between boiling and
critical points in Chapter III. . ;• ,ii,;
By means of Equation (III-9), page 70, it is possible to calculate the
critical volume from no data other than the structural formula. This
value may be substituted in Equation (41) and (vxcoi) evaluated. Then
the critical temperature may be calculated from Equation (47) if a single
liquid-density measurement is available. In this manner it is possible to
estimate densities at all conditions and, m t h the relationships of Chap-
ter III, boiling points and vapor pressures from only the structural for-
mula and one liquid-density measurement.
These* methods for estimating physicochemical properties are of
particular importance in dealing with unstable materials under conditions
which preclude direct measurements. However, it must be recognized
that all such methods are approximations to be used only in the absence
of more reliable data. In many cases several alternate relationships
may be used for estimating a property. In the present state of the data
it is best to compare the various methods with what direct measurements
are available on the compound in question and similar compounds, and
in this way select the most consistent and logical value for each particular
problem. The relative reliability of the various methods is to a large
extent dependent upon the type of fundamental data available.
Pressure Correction to Enthalpy of Liquids. Combination of Equa-
tion (38) ^\^th (XI-99) written in terms of reduced temperature and
pressure gives > > , • /•.•'•'
i(^) ,l-,.,(!i)=-fl-T,(i)l (48)

Vc\dVr/T P \dTr/p Pi Leo \dTr/pJ
Equation (48) is conveniently integrated between pr and the critical
pressure, Pr = 1.0, .-I.,-:
-^(fl..-«).= r p + ^ ( ^ ) l * . = * (49)
CHAP. XII] CORRECTION TO ENTHALPY OF LIQUIDS 507
where Hep = enthalpy at temperature T and the critical pressure Pc-

The value of (aco/6r)p is obtained by graphically differentiating Fig. 109.
The results of the integration of Equation (49) are plotted against
reduced temperature and pressure in Fig. 110.
! TABLE XXVIII
PHYSICAL PROPERTIES OF NORMAL PARAFFINS
1 2 -^^^ , 3 4 5 6 7 8 9 10 11
Pf pl/Wi, tiWi, B
C Mol. lb per g per cc per lb per
AUyma Name Wgt. TB°R rc°Rsgiri. cc g-mole h A eqin.
1 Methane 16.04 201.2 343 673 3.679 4.36 0.000 2.338 5.166
2 Ethane 30.07 331.8 550 717 4.429 6.81 0.088 2.573 5.428
3 Propane 44.09 416.2 666 642 4.803 9.18 0.125 2.661 5.468
4 Butane 58.12 491.1 766 544 5.002 11.62 0.166 2.767 5.502
5 Pentane 72.15 557 847 482 5.128 14.07 0.190 2.887 5.570
6 Hexane 86.17 615.7 915 433 5.216 16.52 0.209 2.994 5.631
7 Heptane 100.20 669 972 394 6.285 18.96 0.224 3.124 5.719
8 Octane 114.22 718 1025 362 5.340 21.39 0.236 3.232 5.790
9 Nonane 128.25 763 1073 332 6.382 23.83 0.248 3.298 5.819
10 Decane 142.28 805 1114 308 5.414 26.28 0.258 3.407 5.896
12 Dodecane 170.32 881 1185 272 6.459 31.20 0.275 3.637 6.071
14 Tetradecane 198.38 948 1248 244 6.483 36.18 0.289 3.818 6.205
16 Hexadecane 226.43 1007 1300 221 •5.48 *41.3 0.300 4.006 6.350
18 Octadecane 254.48 1060 1343 202 "5.49 •46.3 0.308 4.195 6.500
20 Eioosane 282.54 1105 1380 187 '5.5 '51. 0.315 4.403 6.675
25 Pentacosane 352.67 1210 156 •5.5 *64. 0.329 4.834 7.027
30 Triacontane 422.80 1305 1545 133 *5.5 ^76. 0.339 5.169 7.293
. 3 5 Pentatriacontane 492.93 1385 1610 120 •S.S •SS. 0.347 5.580 7.660
40 Tetracontane 563.06 1465 1675 108 •3.3 •103, 0.353 6.011 8.045
45 I Pentatetracontane 633.19 1540 1740 100 •5.4 '118. 0.338 6.358 8.358
• Based on experimental density, questionable because near melting point.
Equation (49) is dimensionless, and any consistent set of units may be

used directly with Fig. 110. However, care must be used to recognize
the dimensions of the resultant enthalpy term and apply the proper
conversion factor in order to obtain conventional thermal units. For
example, if p is in pounds per cubic foot, and p is in pounds per square
foot, H will be expressed in foot-pounds per pound and must be divided
by 778 in order to obtain Btu per pound. ,
Illustration 11. The enthalpy of saturated liquid propane at ISOT and an abso-
lute pressure of 522 lb per sq in. is 102.3 Btu per lb." Calculate the enthalpy of the
liquid at this same temperature under critical pressure and also at a pressure of 1500
lb per sq in. At 70°F and 200 lb per sq in. pressure the density of liquid propane is
31.5 lb per cu ft. _^ _j. : ...; ^J
Tc = 672.2°R
: Pc = 643.3 lb per sq in. = 92,600 lb per sq ft
Tn = 530/672.2 = 0.788
Pr, = 200/643.3 = 0.311
130
1 2 3
Reduced Pressure
FIG. 110. Enthalpy Correction for Liquids
CHAP. XII] PRESSURE CORRECTION TO ENTROPY OF LIQUIDS 509
From Fig. 109, on = 0.1052, and pi/wi = 299 lb per cu ft
- ^ = 299/92,600 = 0.00323 lb per fUb

PcOll
At a temperature of 190°F and 522 lb per sq in.

Tn = 0.966 pn = 0.813 Hi = 102.3
From Fig. 110,
{Hep — Hi)T = —7.8

Pcbll
{Hep - H2)r = -7.8/0.00323 = -2420 ft-lb per lb

= -2,420/778 = - 3 . 1 Btu per lb
Hep = 102.3 - 3.1 = 99.2 Btu per lb
At a temperature of 190°F and 1500 lb per sq in.

Trz = 0.966; pr, = 2.34
From Fig. 110,
- ^ (Hep - H,)T = 20
20
^^--^-^--(0.00323)(778)^^-°^^"P^'-^'^ •
Ha = 99.2 - 8.0 = 91.2 Btu per lb
A value of 90.4 was calculated from direct experimental data."
Pressure Correction to Entropy of Liquids. Combination of Equa-

tion (38) with (h) Table XXIV written in terms of reduced temperature
gives
(50)
p. KdpJT To \ dTr )p TcPl \ dTr
Integrating between p , and the critical pressure, pr = 1.0, gives
^(&,-s),=+rife)dp. (51)
where Sep — entropy at temperature T and the critical pressure Pc. The
results of this integration are plotted against reduced temperature and
pressure in Fig. 111.
" B. H. Sage, W. N. Lacey and J. G. Schaafsma, Ind. Eng. Chem., 26, 1218 (1934).
1.0 2.0 3.0 5.0

Reduced Pressure
FIG. Ill Entropy Correction for Liquids. .msjBt^m-
Figure 111 may be used by a procedure similar to that of Illustration 10

to determine the change of entropy accompanying any change of pres-
sure at constant temperature. Similar care must be exercised with
regard to units.
Pressure Correction to Heat Capacity at Constant Pressure of Liquids.
By differentiating Equation (49) with respect to temperature at constant
pressure, an expression relating heat capacity at constant pressure to
heat capacity at the critical pressure is found.
£cPi (53)
(Cpc
'"''' " (arj.
î
where Cpc = heat capacity at constant pressure at temperature T and
4 ;.rt
CHAP. XII] HEAT CAPACITY OF A SATURATED LIQUID 511
800
2.0 3.0 4.0 5.0

Reduced Pressure
PIG. 112. Heat-Capacity Correction for Liquids.
the critical pressure. A generalized evaluation of Equation (53) is ob-
tained by differentiating Fig. 110. The results are plotted in Fig. 112.
By the general procedure of Illustration 10, Fig. 112 may be used to
estimate the change in heat capacity at constant pressure accompanying
any change in pressure at constant temperature.
Heat Capacity of a Saturated Liquid or Saturated Vapor. When the
temperature of a liquid at its boiling point is increased while saturation
is maintained, an accompanying increase in pressure results. The heat
capacity of such a saturated liquid involves the enthalpy change result-
ing from both the temperature and pressure changes. Thus, from Equa-
tion (XI-39),
Differentiation of Equation (54) with respect to T at saturation, if

the subscript s is used to denote saturation, gives
or
(f).=a.-(a(^).
Rearranging and writing in terms of reduced conditions give ,
The diiTerential {dH/dpr)T is expressed in terms of reduced temperature

and pressure by Equation (48). This equation is evaluated by graphical
differentiation of the lines on Fig. 109. The resultant values of
_ -El- ( — ) are plotted in Fig. 113.
Equation (56) may be applied to obtain the heat capacity of a satu-

rated vapor with the proper designation of terms, thus:
From Equation (12), '.

/dH\ RTU dz\ * ,^^, ^
Values of {dz/dTr)p are obtained from a cross-plot of the compressibility

chart (Fig. 103) on regular co-ordinates. The heat^ capacity of a satu-
rated gas assumes a negative value as the critical temperature is
approached.
An approximate value of {dp/dT)s which is satisfactory for many
purposes is obtained by differentiating the Calingaert-Davis vapor
pressure Equation (111-7), /
In P = A - YZr^ (56c)
Thus,
{T - 43)2
where
T = temperature, °K
B = constant of Calingaert-Davis equation
CHAP. X I I ] HEAT CAPACITY OF A SATURATED LIQUID 513
40
0 1 2 . 3
Reduced Pressure
FIG. 113. Differential Effect of Pressure upon the Enthalpy of Liquids.
The constant B may be evaluated from the boiling point and critical
data using Equation (VII-30), page 232. More nearly accurate values
of (dp/dT)s are obtained by differentiating Equation (III-16), page 73.
By the use of Fig. 113 in combination with Equation (57) the heat
capacity at constant pressure of a saturated liquid is calculated from a
_known value of the heat capacity of the saturated liquid. The heat
capacity at any other pressure is then calculated by means of Fig. 112.
This procedure is demonstrated in Illustration 12.
Heat Capacity of a Saturated Liquid from Cp of Its Ideal Gas. Data
on the heat capacities of liquids, particularly at temperatures other than
atmospheric are scanty and frequently unreliable. However, generali-
zations of statistical data discussed in Chapter XVI permit the esti-
mation of heat capacities of ideal gases over wide temperature ranges.
A thermodjTiamic relation exists between the heat capacity of the ideal
gas and the heat capacity of the saturated liquid involving the heat of
vaporization and the effect of pressure on the enthalpy of the gas. From
this relation in generalized form together with Figs. 112 and 113 the heat
capacity of a liquid at any condition may be predicted from the heat

capacity of its ideal gas at the same temperature. Conversely, the
gaseous heat capacities of a high-boiling complex compound may be
calculated from data on the liquid.
There are several methods by which a saturated liquid at tempera-
ture Ti may be converted into a saturated vapor at a higher tempera-
ture Ti- One is to heat the liquid, maintaining saturation, to T2 and
vaporize. Another is to vaporize the liquid at Ti, expand the vapor
isothermally to zero pressure, heat the ideal vapor to T2, and compress
the vapor isothermally to saturation. Since the imtial and final states
are the same in both cases, the enthalpy changes of the two processes
must be equal. Thus,
Xi + (Ht - SsGi) + 4{T2 - TO - ( H | - H.ff2) = X2 + c,LiT2 - Ti) (58)
or, expressed in differential form,
( c s L - c * ) r d r = - d X - d ( H * - H,G) • (59)
or ',,--'.-
( c . . - c > = - - ^^r— (60)
where subscripts, sL = saturated liquid .' ' ,

sG — saturated gas
The total derivative ——^— can be expressed in terms of its partials

in the usual manner, thus, •''-•. •. .,< '"•'[•••- , .-!
[djn* - H.g)-| [djii* - H,g)l ra(H*-H^-| /dp\ ^^^^
Substituting Equation (61) into (60) and writing in terms of reduced

conditions, we have . ,..- ^ "
/
*, i- ( aK\ 1 1 d{YL* - Hs(?j I r ' . r . ^• • -.,; ; ( y
TAdTj TX dTr i.
T.
PcT, -p^m.'-)
The value of d\/dTr is obtained by combining Equations (VII-29 and
32), page 231, and differentiating. Thus, >, i-
d\ V 2 / I „ Trb)-^-^^
T ,\-0.38
n '^^(\^ T?T> I ^ (63)
dTr ;^ \ n - 4 3 >' ( 1 -- r^)0.62
• ' • 0.^ .- ;•
CHAP. X I I ] HEAT CAPACITY OF A SATURATED LIQUID 515
where B — constant of Calingaert-Davis equation

Tb = normal boiling point, degrees Kelvin
The derivatives in Equation (62) which contain enthalpy difference

are evaluated by differentiation of Fig. 106. The results of this opera-
tion are plotted in Figs. 114 and 115 as functions of reduced temperature
and pressure. These figures, together with Equations (57) and (63),
permit evaluation of all terms of Equation (62) and calculation of
70
50
40
30 (
Tl
20 11
-A 1
10
I 7
5
I
bo 4
S //MYA/M
o
I,
bO 2
1,0
'/'ff
col 0,7 %
0.5
0.4 c ^ J^, Yy leduced tennperatui e
0.3
0.2
0.1
0.0004 0.0010.002 0.006 0.01 0.02 0.04 0.070.1 0.2 0.4 0.6 1.0
Reduced Pressure^p^
FIG. 114. Differential Pressure Correction to the Enthalpy of Gases (Constant
Pressure).
{csL — CP)T- For use in Equation (62) it is recommended that vapor

pressures be calculated from the Calingaert-Davis Equation (56c).
However, this equation is not recommended for calculation of vapor
pressures as such where accuracy is required.
Use of Equation (63) and Figs. 114 and 115 is not recommended at
reduced temperatures below 0.55 or above 0.96. At reduced tempera-
tures of 0.55 and below it may be assumed that (CSL ~ c*) r is independent
of temperature.
OJ_ K 7/ /
i f^f
c3 1
11
CO --^
c5 ^
eo-V^ /i / o
B
iî,
11 r
=>>;^^/// o ^
^ f/ 03
o •+->
cd
S-3
%
o
S
11
>o d
o ^^
0)
m /
a
oo
so^*a>" >-.
1
lih
1
1
1 ^ "1/
41 ^/
lO g<
t- h
o c
t^ aS3
•P
c3 J- t3J
(U
T3
o
a
W
1-2
§
b otf O
^
o
lO •
^•^W, lO
oo
to t;
r
O
A<
O
c3' o •^
to
o -a>
s
0)
as
i
iH
^ fi
o ^ iCl
>flv
o
o f-H
^ rH
rt
8 I*.
o
00 C- 50 lO •*
(H „) Olom-3) / aBo-3)^( A ^ ) i£
(ffiH)e' I
CHAP. XII] HEAT CAPACITY OF A SATURATED LIQUID 517
Illustration 12. Propane gas at a temperature of 63°C and atmospheric pressure

has a molal heat capacity'* of 18.8. Calculate for this temperature:
(o) Heat capacity of the saturated hquid.
(6) Heat capacity at constant pressure of the saturated liquid.
(c) Heat capacity of the liquid at the critical pressure.
(d) Heat capacity of the liquid at a pressure of 200 atm.
Physical constants of propane: ,
To = 373.4°K; p, = 43.7;atm; Tt, = 228.6°K
Solution: (a) From Equation (VII-30), page 232,
1600
I L228.6 - 43 373.4 - 43j
From Equation (56c), where p = 1.0,
_ B _ 1600^
8.62
" n - 43 ~ 185.6
• , 1600
p, = 23.5 atm; p, = 23.5/43.7 = 0.54 :

r . = 336/373.4 = 0.90; Tr^ = 0.613
Evaluation of Equation (63) gives
dK „ /228.6\2 (l-0.613)-<''«
- = -(0.361)(1.987)(1600) ( — J -^^3^^=-10.450cal/(g-mole)(OK)
From Fig. 114, • • .
1 /a(H* - H ) \
• ~¥\ dTr ) =25.5cal/(g-moIe)(°K)
From Fig. 115,
- (-^- ) =11.0 cal/(g-mole) (°K)
From Equation (57), _ /'

(23.5) (1600) /
-—— -— = 0.438 atm per K
\dTj, (336 - 43)2 ^
Substitution in Equation (62) gives
* 10.450 , ^, ^ (0.438) (373.4) (11.0)
'"--'^ = ^7^+ ''•' m %;
^ = 2 8 + 2 5 , 5 - 4 1 . 1 = 12.4 '",^\
CsL = 18.8 + 12.4 = 31.2
Direct experimental measurements" give a value of 33.0.
^ B. H. Sage, D. C. Webster, and W. N. Lacey, Ind. Eng. Chem., 29, 1309 (1937).
" B. H. Sage and W. N.Lacey, Ind. Eng. Chem., 27, 1484 (1935). f f,
518 T H E R M O D Y N A M I C PROPERTIES OF FLUIDS [CHAP. X I I
(6) From the data of Illustration 11, PI/MI = 299 lb per cu ft or 4.8 g per cc. From
Fig. 113, •! • ' S •'•••'••' '
1 /SHN (7.8) (44)
The factor 41.3 converts the energy units from cc-atm to calories while 44 is the
molecular weight of the gas. From Equation (56) and the value of (dp/dT), previ-
ously determined, (cp - C,)T = (l.73)(0.438) = 0.755oal/(g-mole)(°K).
Cj, = 31.2 + 0.755 = 32.0, the heat capacity at constant pressure of the saturated
liquid at 63°C
(c) F r o m F i g . 112, . . • - - - - • ; :'-• - 'r''^.V -...:i :,. -" ;/,» :-.^:'U)UA
,..- , _ (110) (43.7) (44)

(c,, c) - (41 3) (373) (4 g) - ^-y
Cpc = 32.0 — 2.9 = 29, the molal heat capacity of the liquid at 63°C and 43.7 atm.
{d) At 200 atm, pr = 4.58.
From Fig. 112,
, , (185) (43.7) (44) _

^'"^ '''' (41.3)(373)(4.8) " *-^ ';•'•
Cp = 29.1 - 4.8 = 24.3, the molal heat capacity of the liquid at 63°C and 200 atm.
DETERMINATION OF THERMODYNAMIC DATA
Considerable attention has been devoted to generalized methods for

estimating thermodynamic properties from a minimum of direct data.
These methods are important because of the frequency with which
process design calculations must be carried out in the absence of direct
thermodynamic data of any type. The generalized relations derived
from the theory of corresponding states are also of value for correlating
the behavior of solutions and mixtures. However, it must be recog-
nized that all such generahzations are approximations, not based on
rigorous principles, and they should be used only in the absence of reli-
able direct data. The errors incurred may not be serious for many
purposes, but for some requirements they render the application of
generalized methods unsatisfactory.
It is difficult to predict the probable errors in the use of the generalized
relations of the preceding sections. For gases the greatest errors are
encountered for the saturated vapors approaching the critical tempera-
ture. The following estimates represent the orders of magnitude of the
maximum errors which may be encountered.
Per Cent
Compressibility of gases. Pig. 103 15
Pressure correction for enthalpy of gas. Figs. 105-106:
Without York and Weber's correction 30
CHAP. XII] THERMODYNAMIC PROPERTIES 519
Per Cent
In the range of T", = 1.0 to 3.0 and p, = 0 to 10 with York and Weber's
correction 5
Nonideality correction for entropy of gas, Fig. 107 30
Pressure correction for heat capacity, Fig. 108 30
Expansion factors of liquids, Fig. 109 5
Pressure correction to enthalpy of liquids. Fig. 110 25
Pressure correction to entropy of hquids, Fig. I l l 15
Pressure correction to heat capacity of hquids, Fig. 112 20
Difference between heat capacity at constant pressure and heat capacity of
saturated liquid. Fig. 113 and Equation (56) 35
Difference between heat capacity of saturated liquid and ideal gas. Figs.
114^115 and Equation (62), in range Tr = 0.55 - 0.96 25
Although these errors appear high, it must be remembered that they

are estimated maximum values and that for many substances and
conditions much better accuracy can be expected. Furthermore,
several of the relations represent correction terms which are relatively
small in themselves, and large errors in the corrections introduce but a
small error in the property sought. In addition, if consistent use is made
of the same series of generalized relations, many of the errors will tend to
cancel in some applications such as energy balances.
It may be concluded that these generalizations are extremely useful
and a great improvement over the assumption of ideal behavior or the
use of relations based on simple corrections such as those of van der
Waals. However, they must be used with care and judgment, with full
realization of the errors which may result.
Thermodynamic Properties from Experimental Data. Thermo-
dynamic properties derived from reliable experimental data are greatly
to be preferred over estimations by any generalized method. The
methods for the use of such data are similar to those employed in deriv-
ing the generalized relations and employ the same thermodynamic rela-
tions. However, the derivations are carried out in terms of absolute
rather than reduced conditions and are restricted to the particular
substance under consideration.
The thermodynamic properties of a substance may be completely
evaluated from the following data covering the range of temperatures
and pressures of interest.
1. PVT measurements of both the liquid and vapor state.
2. Heat capacities of the vapor, either measured at low pressure or
calculated for the ideal state from spectroscopic data.
3. A value of absolute entropy.
4. Vapor pressure measurements.
From the vapor pressure and pVT data heats of vaporization are cal-
culated from the Clapeyron equation. The variations of the enthalpy,

entropy, and heat capacity of the vapors with pressure are calculated
from Equations (10), (20), and (32). Enthalpies of the saturated liquid
may be obtained by subtracting heats of vaporization from the enthal-
pies of the saturated vapor or may be developed from Uquid-heat-
capacity data calculated from Equation (62). If the compressed liquid
region is of interest, calculations similar to those developed on a general-
ized data may be carried out from the experimental pVT data.
Although the foregoing procedure theoretically defines the complete
thermodynamic behavior of a substance, it is generally desirable to verify
the results by calculations from various types of supplementary data to
detect possible inconsistencies in the experimental measurements. This
general procedure is demonstrated by Benning and McHarness and
coworkers""^ 1 in Evaluating the thermodynamic properties of fiuoro-
chloromethanes and ethanes recently developed as refrigerants. Vapor
pressures were determined from 0.1 atm to the critical point.^^ PVT
measurements were made on the vapors over the range of interest," and
the densities of the saturated liquids and vapors were determined up to
the critical point.2° Heat capacities of the vapors were measured^^ at
atmospheric pressure. To verify the results of these data, measuremei^ts
were also made of the heat capacity of the saturated liquid and the
ratio Cp/C.„ for the vapors by velocity of sound measurements.^'
The Uquid-heat-capacity data were used in combination with the
vapor-heat-capacity measurements to calculate heats of vaporization by
means of Equation (60). These results were compared with those
obtained from the Clapeyron equation, using entirely different experi-
mental data. Similarly, the measured Cp/Cv ratios were compared with
ratios obtained from the measured heat capacities of the vapor at con-
stant-pressure and constant-volume values calculated from the pVT
data and Equation (s). Table XXIV, page 472.
For deriving precise thermodynamic data from experi/nental measure-
ments it is desirable to use analytical methods rather than the less accu-
rate graphical methods used in developing the approximate generalized
relations. This requires that all data shall be expressed by accurate
equations the constants of which are evaluated empirically. The follow-
ing form was used by Benning and McHarness. /
/
" A. F. Benning and R. C. McHarness, Ind. Eng. Chem., 31, 912 (1939). /
w/?id. Ens. C/iem., 32, 497 (1940). ^
" / n d . £»9. C/iem., 32, 698 (1940).
»7nd. Eng. Chem., 32, 814 (1940).
" A. F. Benning, R. C. McHarness, W. H. Markwood, and W. J. Smith, Ind. Eng.
Chem., 32, 975 (1940).
CHAP. XII] THERMODYNAMIC CHARTS AND TABLES 521
For vapor pressures, ^
\ogp=^ A+- + ClQgT + DT (64)
This equation was found to deviate less than 0.3 per cent from the data.
It is more convenient and more accurate than the generaKzed Equa-
tion (III-16), page 73, which contains only three constants but is difficult
to evaluate from data.
The heat-capacity data were represented by Equation (VII-21),
page 213. Similar equations were used for liquid densities over limited
ranges.
The pVT data for the vapors were expressed by the Beattie-Bridgman
Equation (2). The constants were evaluated by Benning and McHar-
ness^^ by the method of Buffington and Gilkey^^ which assumes that c is
zero and thus considerably simplifies the operations. This simplified
form is accurate where pT isometrics are linear.
V 1
THERMODYNAMIC CHARTS AND TABLES
For substances which are frequently used over wide ranges of tempera-
ture and pressure in both liquid and vapor phases, it is convenient to
prepare tables giving values of the important properties over this entire
range. Such tables are particularly indispensable for fluids used in
power generation and refrigeration. It is also common practice to pre-
sent such information graphically in the form of charts. Since all the
properties of a pure component may be defined in terms of two inde-
pendent properties, it is possible on a chart with two co-ordinate scales to
construct contour lines for all related'properties. For example, on a
chart with temperature and entropy as co-ordinates, it is possible to
construct additional contour lines for pressure, enthalpy, free energy,
specific heat, and other properties. This procedure is limited by the
confusion resulting from reading values from a multiplicity of inter-
secting hues. Tables have the advantage of permitting great accuracy
in the reproduction of values, whereas charts are more compact and aid
in visualizing the effects of changes.
The familiar steam tables have undergone extension and refinement
with the result that thermodynamic properties are now known with a
high degree of precision over wide ranges of temperatures and pressure
for the saturated vapor and Hquid, the superheated vapor, and the
compressed liquid. Data for saturated vapors are generally tabulated
in terms of the corresponding temperatures or pressures, whereas for the
22 R. M. Buffington and W. K. Gilkey, Ind. Eng. Chem., 23, 254 (1931).
superheated vapors and compressed liquids tables of properties are

required as functions of temperature at vairious constant pressures.
Such tables are published as entire volumes devoted to a single
substance.^'
In thermodynamic calculations the five properties of pressure, volume,
temperature, entropy, and enthalpy are of greatest interest. In the two-
phase region of liquid and vapor, the additional variable of percentage
vaporization or quality is introduced. Four .types of charts, based on
the following co-ordinate scales, are in common use for presenting these
properties:
1. Pressure-volume. "--i..'
2. Temperature-enthalpy. ' ' ' ' '' , '
3. Temperature-entropy '"(i^, L
4. Enthalpy-entropy (MoUier chart). " ' ' '•'
Auxiliary lines may be drawn on each of these charts to show the other
three properties or any other related property. Each chart has special
advantages for certain applications, as is discussed in succeeding chap-
ters, although each chart may be constructed to include all the data of
the others. With known values of absolute entropies, it is also possible
to construct additional charts and tables giving values of free energy
and the total work function. Internal energy can be obtained from
values of enthalpy, pressure, and volume.,
A representative series of diagrams for ammonia is presented in
Figs. 116-119 plotted from tables based on direct experimental data^* up
to 300 lb per sq in. and extended by generalizations above that pressure
as indicated in broken lines. Since ammonia is extensively used as a
refrigerant, values of the enthalpy and entropy are arbitrarily assigned
zero values at the saturated-hquid state at — 40°F. The absolute
entropy of hquid ammonia at this state is 1.207 Btu/(lb)(°R).
The general shape and appearance of Figs. 116-119 are characteristic
of such diagrams for substances of low molecular weight and low molal-
heat capacity. The other common refrigerants and water are repre-
sented by diagrams of similar appearance.
Diagrams from Generalizations. Since few thermodynamic data are
available for compounds of high molecular weight, it is generally neces-
sary to resort to the generalized relations in order to develop diagrams for
such substances. Such a series of diagrams for benzene is shown in
*'J. H. Keenan and F. G. Keyes, " Thermodynamic Properties of Steam," John

Wiley & Sons, New York (1936).
" Natl. Bur. Standards U. S. Circ. 142 (1923). ,. . .,, ,. •,-. ,;,,„„:,. . ,..,,.:.,. -' ,
CHAP. X U ] DIAGRAMS FROM GENERALIZATIONS 523
"
/^
« 4\'
i^N*/<^
'^^J ' •
,Ji 7 r- / /
•
•
^^MM' ^ / ^
'^^" ,> • y
•
^ ^
^5^ ^',.
•
/
^
^t
1 • ^
^/
/ /
•
^ • w
S <!
/ / «* y /
/ •
y .^
•
^' ' / • g /•
03
I'^S^^ . / ** • , y
CO
CD
P.
f
^ : / ^^° t\d 4
N g 5
J yy^j JT ^
' .'V^ , f /
o
^W^' / 4
/ .'1 î.A O
/
y/ji^ Y ' / • 1 y
y
- 1^2 ~ •
y
a
•
d,
/ /'r •'' ;, LL . y ^' a^ , • '
/
/ 'J' ' ' ' 0
<
, A0
J /if • '
kXX-v-'
' • v
lit' •' I' 1 t
HJ-'V '-+{•'--
o y
-'
S^fcT '
•4ll-
S 0 1.' •' 1
'Ut-^-
^ti § ^ . ' ''
g-IC44^ ^ , •
¥5E..^__!
'S ~'~'I.. i-trr-
'a f.'"" I P9:}B J îBc;
+3
~"
u
O
oo o o c> o OOO o o o oooô •^ CO C<1 «-<
O oo o o o o
O OOO » •* CO (M O 00 ;P Tj< CO (M 1-i
t-l sqB "ui bs aad qi 'cf
I!
THERMODYNAMIC PROPERTIES OF FLUIDS [CHAP. XII
-100 -50 0 60 100 ISO 200 250 300 350 400 450
t,°F
FIG. 117. Enthalpy-Temperature Chart for Ammonia
Figs. 120-122. As a minimum requirement direct data must be avail-

able for the normal boiling point, the specific gravity of the Uquid, and
the chemical structure. The heat capacity of the vapor may be esti-
mated by the generaUzed methods discussed in Chapter XVII. The
critical temperature and pressure and the complete vapor-pressure-
temperature relation may be approximated by the methods of Chap-
ter III and the heat of vaporization by the methods of Chapter VII.
In the case of benzene, these data were available from direct experimental
measurements and accordingly were used in preference to estimated
values.
In developing a series of diagrams it is convenient first to establish
enthalpy as a function of temperature for a series of constant pressures.
Enthalpies are generally referred to the saturated liquid at an arbitrary
temperature, in this case 60°F. The enthalpy of the saturated vapor
at the reference state is first calculated by correcting the normal heat of
vaporization for temperature by means of Equation (VII-32), page 233.
The enthalpy of the vapor at this temperature and zero pressure is
CHAP. X I I ] DIAGRAMS FROM GENERALIZATIONS 525
ô'?
qi jad nâ 'jj

CHAP. XII] DIAGRAMS FROM GENERALIZATIONS 527
100 200 300 400 500 600 700

Temperature, t °F.
FIG. 120. Enthalpy-Temperature Chart for Benzene,
800
^
-b^^ \ /
/
600 7 ^ %m-
Critical p oint-; V""^'
•p=600J /
/^ o = 40(
a
.m
^^
..A-*'^
\^^
•
^
Ib/sq in. al
\v- -200-
«•?
y*
200
a&^ « ^
i p=30 3 ^ ^ •^^
w^ \ = 14.7-
•
w
%
N *v 11
/to
/
%
\
O.B 0.6 0.7 0.8 0.9 1.0 1.1
S, Btu/(lb)rR)
PIG. 121. Temperature-Entropy Chart for Benzene.
100
0.70 0.80 0,85 , 0.90
S, Entropy Btu/(lb) (°R)
FIG. 122. Enthalpy-Entropy Chart for Benzene.
CHAP. XII] DIAGRAMS FROM GENERALIZATIONS 529
obtained by adding tiie correction from Fig. 105 corresponding to the

reduced temperature and pressure of the saturated vapor. The line
ab, Fig. 120, representing the enthalpy of the vapor at zero pressure H*
is then established from the heat capacity of the vapor, corresponding to
zero pressure.
Lines representing the enthalpy of the superheated vapors at other
pressures are determined by calculating series of values of {H* — H^T
from Fig. 106 for selected constant pressures and temperatures. These
corrections are subtracted from line ah, establishing other constant
pressure hnes in the superheated-vapor region. The enthalpy of the
saturated vapor is established in a similar manner. Values of
{H* — i?p)r are calculated at selected saturation temperatures and
pressures and subtracted from the corresponding values of H* from
line ah, thus establishing line cd.
Line ef representing the enthalpy of the saturated liquid is established
. by subtracting heats of vaporization from corresponding values of the
enthalpies of saturated vapor. The heats of vaporization are calculated
from Equation (VII-32), page 233.
Constant-pressure lines in the compressed-hquid region are estabhshed
from Fig. 110. It is convenient to establish the enthalpy of the liquid
under the critical pressure over the entire temperature range and use
this as a base to estabhsh lines for other pressures.
This procedure may lead to discontinuities at the critical temperature
for enthalpy lines at pressure higher than the critical pressure. Below
the critical temperature these lines are established from Fig. 110 based
on the liquid-expansion factor correlation, while at temperatures above
the critical they are derived from Fig. 105 based on the gas-compressi-
bility correlation. Discontinuities indicate inconsistencies in these two
correlations and are arbitrarily smoothed, giving the greatest weight to
the values calculated from the liquid correlation which appear to be more
reliable in the critical region.
The temperature-entropy chart, Fig. 121, is established by a similar
procedure. The use of absolute values of entropy instead of relative
values greatly increases the utility of the charts. It has been estab-
lished that the absolute entropy of liquid benzene at 60°F and its own
vapor pressure is 0.530 Btu/(lb)(°R). The entropy of the saturated
vapor at 60°F is obtained by adding the entropy of vaporization at 60°F
to 0.530. The vapor pressure of benzene at this temperature is 1.15 lb
per sq in. The entropy corresponding to ideal behavior at this tempera-
ture and pressure is obtained by adding the generalized entropy correc-
tion for deviation from ideal behavior, from Fig. 107. The entropy of
superheated benzene vapor at 1.15 lb per sq in. for various temperatures.
if ideal behavior is assumed, is obtained from the heat-capacity equation

of ideal benzene vapor. The true entropy values at this pressure are
then obtained by subtracting corrections read from Fig. 107. Entropies
of the vapors at any given temperature and for other pressures are then
determined by calculating the entropy of the ideal vapor at the existing
pressure from that at 1.15 lb per sq in., and then subtracting the addi-
tional correction due to lack of ideal behavior, by use of Fig. 107. The
entropy of the saturated liquid at any temperature is obtained by sub-
tracting the entropy of vaporization from the entropy of the saturated
vapor. Entropies of the compressed liquid are obtained by means of
Fig. 111.
After the constant-pressure hnes of Figs. 120 and 121 are estab-
lished, they may be combined by cross-plotting to determine lines of
constant entropy which are plotted on Fig. 120 and lines of constant
enthalpy which are plotted in Fig. 121. Either of the charts then may
be replotted in the form of Fig. 122.
Illustration 13. It is desired to calculate the enthalpy and entropy of benzene at
491°F and under the following conditions:
(o) The superheated vapor (dew point 60°F).
(b) The superheated vapor at 200 lb per sq in. * > 'i
(c) The saturated vapor. , j
(d) The saturated liquid. t j
(e) The compressed liquid at the critical pressure.
(/) The compressed Uquid at 3000 lb per sq in. '>"•<{' /
The following data arc available: . ,-
Normal boiling point 177°F (637°R)(354°K)
Critical temperature 5 5 r F (1011°R)(561°K)
Critical pressure 704 lb per sq in. (47 9 atm) ' '
Liquid density 0.874 g per cc at 77°F
Molecular weight 78.04 -i< • f '
Molal heat capacity in Btu per Ih-mole at zero pressure: %v.
cl = -5.70 + (54.14)(10-')r - (13.41)(10-8)r2
where T is in degrees Rankine.
Vapor pressure: ^
\ r
2.8096
logio p = 7.3659 e-M(r,H!.i88)» nim Hg
Absolute entropy of saturated liquid at 60°F = 0.53 Btu/(lb)(°R). The reference

state of zero enthalpy is the saturated liquid at 60°F.
Vapor pressure at 60°F;
520
CHAP. XII] THERMODYNAMIC CHARTS AND TABLES 531
logw V = 7.3659 - =^2^^ - e-o.20(5U-o.i88)2 = 1.775

0.514
p = 59.2 mm or 1.15 lb per sq in.
Similarly, at 491°F, the vapor pressure is 470 lb per sq in.

Heat of vaporization at 60°f is calculated from Equation (VII-29), page 231.
From Equation (VII-30), page 232,
n - 43 "" r . - 43 354 - 43 ~ 561 - 43

520 637
^' = H I = °-^^^' ( n ) . = — = 0.631
/ n \2 /354\2
Substitution in Equation (VII-29), gives

/0.486V=«
X = (0.95) (1.987) (3020) (1.295) I ) = 8230 cal per g-mole
\0.369/
Kfi;.!J - 8230
or z-TT (1.8) = 190 Btu per lb
78.04
Similarly, at 491 °F, the heat of vaporization is calculated to be 85 Btu per lb.
Enthalpy of saturated vapor at 60 °F = 190 Btu per lb
190
Entropy of saturated vapor at 60°F = 0.53 H = 0.896 Btu/(lb) ("R)
At the low pressure of 1.15 lb per sq in., ideal behavior may be assumed, and these
values may be taken as representing the ideal state.
(o) At 49l°F and 1.15 (lb)/{sq in.), if we assume ideal behavior;
1 /"gsi
H* = 190 + —— / [-5.70 + (54.14)(lO-^D - (13.41)(10-«)T2] dT
78.04 J520
= 338 Btu per lb
1 /»95i r _ 5 70 "I
S* = 0.896 + r r ^ / ^ ^ + (54.14) (10"^) - (13.41) (10-«)r dT
78.041/520 L i J
= 1.095
(5) At i91°F, 200 lb per sq in.:
951 200 „ „„,
Tr = = 0.94; pr = — = 0.284
1011 ' '^ 704
H* - H 920
From Fig. 105, —^ = 0.91; H * - H = 920 or — — = 11.8 Btu per lb.
1c 7o.U4
ff = 338 - 11.8 = 326 Btu per lb
From Fig. 107, s* - s = 0.74

or
S* - S = 0.74/78.04 = 0.010 Btu/(lb)(°R)
From Equation (XI-142),
1.987 200
S* = 1.095 - ——- In - — = 1.095 - 0.131 = 0.964
78.04 1.15
S = 0.964 - 0.010 = 0.954 Btu/(lb)(°R)
(c) Saturated Vapor at iQVF: -
Pressure = 470 lb per sq in.
Tr = 0.94; pr = 0.67
From Fig. 105,
(H* - H) = (3.10)(1011) = 3130 Btu per lb-mole
or 40.1 Btu per lb
H = 338 - 40 = 298 Btu per lb
From Fig. 107,
s* - s = 2.80 Btu/(lb-mole)(R°) or 0.036 Btu/(lb)(°R) i
1.987 470
S* = 1.095 - —— In — - = 1.095 - 0.152 = 0.943
78.04 1.15
S = 0.943 - 0.036 = 0.907
(d) Saturated liquid at 491°F:
X = 85 Btu per lb
#
ASx = — =0.089
951 I
ff = 298 - 85 = 213
<S = 0.907 - 0.089 = 0.818
(e) Compressed liquid at 491°f and its critical pressure at saturation:
p = 470, pr = 0.67, Tr = 0.94
From Fig. 110,
- ^ (He - H) = - 6 . 0
Pc W
At 77°F, 1.0 atm, Tr = 0.53; pr = 0.0209, pi = 0.874 g per cc

From Fig. 109,
£01 = 0.13
For benzene: ,
pi
— = 6.71 g per cc
-£l- = ( ^ 6.71= 5.8 g per cal

Pcoji (47.9)
CHAP. XII] INTERNAL ENERGY 533
Hence,
o
He - H = -— =- -1.04 cal per g or - 1 . 9 Btu per lb
6.8
He = 213 - 1.9 = 211 Btu per lb
From Fig. I l l ,
^ ( S . - S ) = -12
«iPc 47.9
^ ' - ' = ^ = - ' • ' ' ' '
I -S, = 0.818 - 0.004 = 0.814 Btu/aWCR)
Of) Compressed liquid at 3000 lb per sq in.:

\ Tr = 0.94; pr = 4.26
From Fig^ 110,
- ^ {He -H) = 12.5
12.5
He - H = -—— = 2.16 cal per g or 3.9 Btu per lb
58
H = Hi- 3.9 = 211 - 3.9 = 207 Btu per lb
From Fig. I l l ,
— (Se-S)= 48
aiPe
S = Se- 0.0148 = 0.814 - 0.015 = 0.799 Btu/(Ib)(°R)
The completed enthalpy-temperature, temperature-entropy and enthalpy-en-

tropy charts for benzene, constructed from points calculated by the methods shovra
in this illustration, are shown in Figs. 120, 121, and 122.
Internal Energy, Free Energy, and Total Work Function. As was

pointed out in Chapter XI, absolute values of these three properties
U, G, and A are never known. However, relative values referred to
any selected reference state may be calculated by the methods of the
preceding sections.
The change in internal energy accompanjang any change in conditions
is derived from the corresponding enthalpy change. Thus, from the

definition of enthalpy,
AU = AH ~ AipV) (65)
Applying Equation (65) to determining internal energy relative to a
reference state of zero enthalpy, designated by a zero subscript, gives
U = H - (pV - VoVo) (66) •
Similarly, by definition,
,..,.,,.., AG = AH~A{TS) /:• (67)
or, expressing free energy relative to the reference state, gives
G = H - (TS - ToSo) " '^' (68)
As was pointed out in Chapter XI, it is evident that a knowledge of abso-
lute values of entropy is necessary for the determination of free-energy .
changes in all except isothermal operations. '
Combination of Equation (65) with the definition of the total work
function gives , ,•'!' ;• ' "•'•> ^
AA^ AH - A{pV) - A(TS) , \ (69)
or, expressing A relative to the state of zero enthalpy, ^
A = H-{pV - poYo) - {TS - ToSo) (70)
Illustration 14. The absolute entropy of liquid water at 25°C (77°F) is 16.75 en-
tropy units per mole. Calculate the internal energy, free energy, and total work
function, in Btu per lb of saturated water vapor at an absolute pressure of 125 lb per
sq in., relative to the saturated liquid at 32°F. Use the enthalpy and entropy data of
the steam tables, and assume the heat capacity of liquid water between 32°F and
77°F to be constant at 1.0 Btu per lb per °F, and neglect the effect of pressure on the
entropy of the saturated liquid in this range.
Solution: The absolute entropy of the reference state is calculated from Equa-
tion (17):
^ - 1.0 In Q g ° ^ 33) = 0.9306 - 0.0875 = 0.8431 Btu/(lb) (°R)
From the steam tables, at the reference state, ' " /

p„ = 0.08854 lb per sq in. '•<«'"'"
Vo = 0.01602 cu ft per lb" •' ir.;;:)o;»
For the saturated vapor at 125 lb per sq in. and 344.33°F,
V = 3.587 cu ft per lb
H = 1191.1 Btu per lb
S = 0.8431 + 1.5844 = 2.4275 Btu/(lb)(°R)
CHAP. XII] PROBLEMS 535
Equation (66), converting pV from foot-pounds to Btu gives

U = 1191.1 - [(125)(3.587) - (0.08854)(0.01602)]^
778
= 1191.1 - 83.0 = 1108.1 Btu per lb
From Equation (68),
G = 1191.1 - (804.33)(2.4275) + (492)(0.8431)
: r , - •;. , . = 1191.1 - 1537.7 = -346.6 Btu per lb
From Equation (70) and the preceding calculations of A(pF) and AiTS),
A = 1191.1 - 1537.7 - 83.0 = -429.6 Btu per lb
As is evident from Illustration 14, the relative free energy and maxi-
mum work may decrease with increase in temperature. Complete
tables or charts of the functions could be established either from the
generalized correlations of properties or from experimental data. How-
ever, such data are of limited value, and the principle application of free
energy is in the prediction of equilibrium in chemical reactions or
solutions.
PROBLEMS
1. For the production of liquid oxygen it is desired that the gas be compressed to a
pressure of 100 atm at a temperature of —90°C. Oxygen at a pressure of 14.5 lb per
sq in. and a temperature of 22°C is compressed to these conditions. Calculate the
volume of compressed gas resulting from 100 cu ft of the original: (o) From van der
Waals' equation, (b) From the compressibility chart.
2. Using the Beattie-Bridgman equation, calculate the pressure exerted by 30
liters of air, measured at 20°C under a pressure of 740 mm of Hg, when compressed
into a volume of 200 cc at a temperature of 0°C. It may be assumed that at the lower
pressure the ideal-gas law is applicable.
3. For ethylene gas the value of pV at 20°C and 100 atm is 0.3600 referred to a
value of unity at 0°C and 1 atm. The density of ethylene gas at standard conditions
is 1.2604 g per Uter. Calculate the volume of 1 g-mole at 20°C and 100 atm.
4. Calculate the volume occupied by 1 g-mole of carbon dioxide gas at its critical
state when the product pV = 0.3484 referred to 1.00 at 0°C and 1 atm.
5. From the data of the International Critical Tables, calculate the compressi-
bility factors of ethylene at a temperature of 20°C and at pressures ranging from 0 to
500 atm. Plot these compressibility factors against pressure in atmospheres and also
against molal volumes in cubic centimeters per gram-mole.
6. From the data of problem 5, calculate the volume occupied by 50 g of ethylene
at a temperature of 20°C and a pressure of 5000 lb per sq in.
7. Using the data of Figs. 100 and 102, calculate the density in pounds per cubic
foot of nitrogen at a pressure of 900 lb per sq in. and a temperature of — 40°C.
8. Using the data of Figs. 100-102, calculate the pressure necessary to compress
100 liters of nitrogen at a pressure of 745 mm of Hg and 23°C to occupy a volume of
2.0 liters at a temperature of 40°C.
9. Using the data of Figs. 100-102, calculate the temperature to which 1.2 lb of
nitrogen must be cooled in order that it may exert a pressure of 50 atm when confined
in a volume of 0.27 cu ft.
10. In a multistage compressor, carbon dioxide is compressed from a volume of

2 cu ft at a pressure of 100 lb per sq in. and a temperature of 22°C, to a volume of
0.4 cu ft at a temperature of SCC. Calculate the pressure necessary, using Fig. 103.
11. Calculate the volume occupied by 10 lb of chlorine when compressed to a
pressure of 125 lb per sq in. at a temperature of 30°C, using Fig. 103.
12. Methyl chloride for use in domestic refrigerators is sold in small cyUnders hav-
ing volumes of 0.15 cu ft. Calculate the weight of methyl chloride gas contained in a
cyUnder if the absolute pressure is 68 lb per sq in. and the temperature 20 °C, using
Fig. 103. {tc = 143.rC; Pc = 65.8 atm).
13. For high-pressure distribution in long pipe lines it is proposed to compress
natural gas (methane) to an absolute pressure of 600 lb per sq in. Calculate its
density in pounds per cubic foot at this pressure and a temperature of 80°F, using
Fig. 103.
14. Sulfur dioxide is compressed from a pressure of 40 lb per sq in. and a tempera-
ture of ISCF to a pressure of 190 lb per sq in. at a temperature of 160°F. Calculate
the ratio of the initial to the final volume, using Fig. 103.
15. The absolute entropy of carbon dioxide in the ideal state at 25°C and 1 atm is
61.08 entropy units per mole. For a temperature of 40°C and a pressure of 60 atm,
calculate the absolute entropy, the enthalpy in calories per gram-mole referred to the
ideal gas at 0°C, and the molal heat capacity, using Figs. 106-108.
16. Liquid sulfur dioxide has a density of 92.42 lb per cu ft at a temperature of
0°F and an absolute pressure of 10.35 lb per sq in. Calculate the density of the Uquid
at 150°F and 1000 lb per sq in. from Fig. 109.
17. Saturated liquid sulfur dioxide exerts a vapor pressure of 136.5 lb per sq in. at a
temperature of 130°F and has an enthalpy of 181.24 Btu per lb and an entropy of
0.32472 unit per lb referred to the saturated Uquid at —40°F. Calculate the enthalpy
and entropy of the compressed liquid at 130°F and 3000 lb per sq in., using the data of
problem 16.
18. Sulfur dioxide exists as a saturated liquid at 130°F and 136.5 lb per sq in. At
these conditions calculate the molal heat capacity of the saturated liquid and the heat
capacity of the liquid at constant pressure, using the heat capacity of the ideal gas
(Table V, page 214) and the generahzed correlations. Also calculate the heat
capacity of the liquid at 130°F and 3000 lb per sq in.
19. Construct temperature-enthalpy, temperature-entropy, and enthalpy-
entropy charts for methyl amine in Btu per lb referred to the saturated liquid at
0°F. Evaluate the properties of the saturated vapor and liquids at reduced tempera-
tures of 0.7, 0.8, 0,85, 0.9, 0.95, 0.98, and 1.0. Establish constant-pressure lines at .
absolute pressures of 14.7, 100, 600, 1000, and 2000 lb per sq in. by calculations at the
afore-mentioned temperatures and also at Tr = 1.1 and 1.2.
The properties of methyl amine axe as follows: /
Boiling point, -6.5°C Critical temperature, 156.9°C ' /
Critical pressure, 73.6 atm Density at — 11°C, 0.699 g per cc , /
Molal heat capacity of ideal vapor = 4.02 -t- 30.72 X IQ-^ T - 8.70 X
10-6 fi where T = degrees Kelvin. /^:
Heat of vaporization at —6.5°C, 368 Btu per lb /
The vapor pressure may be obtained from Equation (III-16), page 73, which is
most conveniently used by plotting a curve relating vapor pressure on a logarithmic
scale to temperature on a uniform or reciprocal scale.
20. For methane gas at 122°F and 1000 lb per sq in. abs, calculate the values of
CHAP. XII] PROBLEMS 537
U, H, S, Cp,G, and A, per pound mole relative to the actual gas a t 32°F and 1 a t m
pressure by the generalized method and also by the Benedict-Webb-Rubin equation
of state. The absolute molal entropy of methane gas in the ideal state at 32°F and
1 a t m pressure is 44.5 Btu/(lb-mole) (°R). The other necessary data are contained
in Table X X I , page 336, and X X V I I , page 483.
21. Benning and McHarness" determined the following constants for dichloro-
fluoromethane CHCI2F (molecular weight 102.9): In Equation (2),
A„ = 20.54 a = -0.179
Bo = 0.286 6 = 0.497
c = 0 R = 0.08206
p = atmospheres T — degrees Kelvin
V = liters per gram-mole
In Equation (64),
A = 38.2974 B = -2367.41
C = -13.0295 D = 0.0071731
p = atmospberes T = degrees Kelvin
Heat capacity of liquid at 1 atm,

Cp = 0.2471 -1- 0.000189J
Cp = calories/(gram) (°C) t — degrees centigrade
Heat capacity of vapor a t 1 atm,

Cp = 13.65 + 0.0249J° cal/(g-mole)(°C)
Density of saturated liquid ( - 4 0 ' ' C to -1-70°C),

p = 1.4256 - 2.316 (lO"')* - (2.6)(10-«)P g per cc, where < is in degrees centigrade.
Calculate by rigorous methods
(a) The heat of vaporization at — 40°F.

(6) The enthalpy and entropy, referred to the saturated liquid at — 40°F, of the
saturated vapor and the saturated liquid at -|-120°F.
^
CHAPTER X I I I
EXPANSION AND COMPRESSION OF FLUIDS
The most useful form of energy into which it is generally desirable to
convert other forms is the mechanical work of motion as represented by a
rotating shaft or a moving piston. Mechanical work is defined as the
energy which is transferred by the effect of a force acting through a dis-
tance and is equal to the product of that force times the distance of
action. Thus, when a fluid which is confined under pressure undergoes
a change in volume, work is done as the result of the force of pressure
moving through the distance corresponding to the volume change.
Similarly, a flowing fluid may perform work through changes in volume,
elevation, or external kinetic energy.
As pointed out in Chapter VII, work is a form of energy which is
incapable of storage as such but is in transition from one form of stored
energy to another. Mechanical energy in this transitory form is termed
shaft work, as distinguished from the electrical and other forms of work
which may be included in the work term of a complete energy balance.
Shaft work is capable of transmission as such and may pass from one part
of a system to another through either solid media such as shafts, pistons,
gears, and belts or through fluid media such as hydraulic or pneumatic
couplings and drives. However, no transmission device is perfect, and
some work is always lost by degradation to a lower form of energy, made
manifest as heat resulting from friction.
Differential Energy Balance. In Chapter VII a general energy
balance was estabhshed for a complex chemical process involving several
interconnected systems. By applying this energy equation to an infini-
tesimal change in the energy of a single system of constant mass m and
constant composition, the following differential form results,
d{pV) +d[—-)+ d{mZ) + d{mE,) + dU = d'q - d'w (1)
where the energy terms represent in order, flow work, external kinetic
energy, potential energy, surface energy, internal energy, heat added
from the surroundings, and work done upon the surroundings. The
work term d'w, in general, includes all possible forms of work, such as
638
CHAP. X I I I ] SHAFT WORK IN A FLOW PROCESS 539
shaft, electrical, magnetic, and radiant. Equation (1) may also be

written as
dE = d'q - d'w (2)
where E is the total energy of the system including all the terms repre-
sented by the left side of Equation (1).
In the absence of electrical, magnetic, and radiant forms of work, d'w
represents mechanical or shaft work only. If changes in potential and
external kinetic energy are also negligible, the shaft work results only
from changes in flow work or from work of expansion. The evaluation
of shaft work in such systems requires separate consideration of the
special cases of nonflow and flow processes, in both isothermal and isen-
tropic systems for reversible and irreversible processes.
Shaft Work in Nonflow Processes. In a nonflow process, as defined
in Chapter VII, flow is absent in the initial and final states of the process,
and changes in surface, kinetic, and potential energies are generally
negligible. If only work of expansion is performed, Equation (1) then
becomes
d'q = d'w, + dU '/'t 'A (3)
where d'w, represents only mechanical work accomplished as the result of

a volume change. In a reversible process d'q = T dS and dU =
T dS. — p dV. Hence, for such a process Equation (3) becomes
d'ws = p dV . ,,
r»2
or - w, (nonflow)
' " ^ = I pdV /')G\; (4)
Equation (4) is restricted to mechanically reversible nonflow processes.

Where the expansion is not mechanically reversible J'p dV is equal to
the sum of the useful shaft work plus that lost in friction. In applying
Equation (4) to actual expansion or compression operations care must be
taken that the pressure used in the equation is that actually exerted on
the face of the piston. The type of expansion and the existence of fluid
turbulence and friction during expansion do not affect the validity of
Equation (4) but do affect the amount of work performed in expanding a
fluid between specified initial and final states.
Shaft Work in a Flow Process. In applying Equation (1) to a reversi-
ble-flow process, d{pV) = p dV + V dp; dU = T dS — pdV, and
d'q ~ T dS. Where forms of work other than shaft work are negligible,
d'w, = -Vdp - d[—-]- d{mZ) - d(mE^) (5)
540 EXPANSION AND COMPRESSION [CHAP. XIII
Where changes in kinetic, potential, and surface energies are negUgible,

Equation (5) reduces to d'w, = —V dp ot \
w. (flow) = -J Vdp (6)
Equation (6) represents the shaft work performed by any reversible-flow

process with negligible kinetic-, potential-, and surface-energy changes
and with no electrical, radiant, or magnetic work.
It should be noted that although Equation (4) is applicable to any
type of nonflow process Equation (6) is restricted to processes in which
all changes of state occur reversibly where d'g = T dS.
Shaft Work from the Energy Functions. For reversible processes
under the conditions defined for Equations (4) and (6) the shaft work
performed may be expressed as functions of the four energy functions.
Thus, from the indicated equations of Table XXIV, combined with (4)
and (6), - • .
(J) -)^=-V or -(A4).= j ^ {vdV)r
= {WS)T (nonflow) (7)
(k) O " ^^ "'' ~^^^)^= ~£(Vdph

= {Ws)T{flow) . (8)
(J) (i?I=~^ °' -(Af/). = /(p^F),

= i'Ws)s (nonflow) (9)
(k) 0^)^=7 or -(AH),= -£(Vdp)s ^ -
= (ws)s(flow) ! (10)
From Equations (7-10) it is evident that: /

(a) Under isothermal nonflow conditions, the work of reversible
expansion is equal to the decrease in the total work function.
(6) Under isothermal flow conditions, the work of reversible expansion
is equal to the decrease in free energy.
(c) Under isentropic nonflow conditions, the work of reversible
expansion is equal to the decrease in internal energy.
(d) Under isentropic flow conditions, the work of reversible expansion
is equal to the decrease in enthalpy.
CHAP. XIII] ISOTHERMAL NONFLOW EXPANSION 541
From Equations (7-10) it is evident that the work of reversible expan-

sion of a fluid under isothermal or isentropic conditions may be deter-
mined either by a pressure-volume integration or as the change of the
appropriate energy function. The pV integration must follow the
actual reversible path of the change, whereas the thermodynamic
energy fimctions are state properties which are independent of path and
are completely defined by the reference properties. Either method of
calculating work may be employed with thermodynamic data based on
either experimental measurements or generalized correlations. The
choice of the method is determined by the type of data available. How-
ever, in general, it is more convenient to evaluate the thermodynamic
energy functions at the terminal states of the system than to carry out
integrations from equations of state.
Isothermal Nonflow Expansion. For the reversible isothermal ex-
pansion of any fluid imder nonflow conditions, from Equation (7) and
the definitions of the thermodynamic energy functions
" '• - *
X 2
pdV = -AA = TAS -AU

-. • ;. ^ = T AS - AH + A(pV) (11)
Equation (II) is restricted to reversible isothermal nonflow operations
involving only shaft work and it neglects changes in kinetic, potential,
and surface energies.
For ideal gases, p dV is evaluated by differentiating the ideal-gas law
at constant temperature. Thus, for 1 mole,
• pdv=' -vdp= -RT^ (12)*

P
Substitution of (12) in (11) and integration give
(wf)r (nonflow) = - f l r I n - (13)*

pi
For actual gases, pv = zRT. Differentiation at constant temperature

gives
pdv + vdp = RTdz (14)
Hence
{dw,)T (nonflow) = pdv= -vdp + RTdz= -RTz— + RTdz'jlS)
or
{W,)T (nonflow) = -RT f zdlnp+ RT{Zi - zi) (16)
The integral of Equation (16) can be evaluated graphically by plotting

values of 2 as a function of log p and determining the area under the curve
between the given pressure limits. Values of z may be obtained from
Fig. 103 or from experimental measurements. Where an accurate
equation of state such as the Beattie-Bridgman is available, the inte-
gration may be carried out analytically.
The foregoing graphical integration can be avoided by evaluating the
work from the entropy and enthalpy terms of Equation (11). If the
generalized correlations are employed, Equation (11) may be rearranged
to give, for 1 mole of fluid,
(w),)r(nonflow) = T[{^* - sf) - (sj* - S2) + (sf - Si)]

- [(H* - Hf) - (H* - H2) + (Hi* - Hi)]
"*'^ ' " +RT{z,-z^) • • (17)
or, combining with Equation (XI-142), page 475,
(w,)r(nonflow) = r j i J l n - - ( s * - S 2 ) - l - ( s f - S i ) l -)
R n f - H2) _ (Hi* - Hi)!

-:;'^ •/liJiiioi'l k:i.V,''i L Tc To J
.-; • +RT{Z2-Zi) (18)
The terms of Equation (18) may be obtained directly as functions of

reduced temperature and pressure from Figs. 103, 106, and 107.
Isothermal Expansion, Flow Conditions. For the reversible isother- -
mal expansion of any fluid under flow conditions, from Equations (8),
{W,)T (flow) = - J Vdp = -AG = -AH +TAS (19)
Equation (19) is restricted to reversible isothermal flow conditions

involving only shaft work and neglects changes in kinetic, potential, and
surface energies.
For an ideal gas, per mole, , i; ,
(w*)r (flow) = - B r In ^ , (20)*
Thus for an ideal gas the work of reversible isothermal expansion is the
same for flow as for nonflow conditions. This results from the fact that
at constant temperature the change in flow work is zero for an ideal gas,
or d{pV) = 0.
CHAP. XIII] ISOTHERMAL NONFLOW EXPANSION 543
For an actual gas, per mole, — V dp = —z{RT alp)/p or
(W,)T (flow) = -RT f zdlnp (21)
Equation (21) is evaluated graphically by the method described for

Equation (16) by using generalized compressibility factors or an accurate
equation of state. The graphical or analytical integration can be
avoided by evaluating the work from the entropy and enthalpy terms of
Equation (19). If the generalized correlations of Figs. 106 and 107
are used, Equation (19) is rearranged as follows:
(w,)r(flow) = r T i J l n ^ - (s^ - S2) + (si* - Si)l
- . [ (Hr-HO_(Hr^-| ^^^
Illustration 1. Calculate the work of reversible isothermal expansion when

1 lb-mole of ethylene gas initially at 564°R and 50 atm expands to 1 atm under non-
flow and flow conditions. From Table XXV, tc = 9.7°C; Pc = 50.9 atm
(o) Assuming ideal behavior for nonflow and &ow conditions, from Equation (20),
iw*)T = -RThx- = -I-(1.987)(564)(2.303) log50 = 4,370Btu

Pi
••2
(6) For actiuil-behavior nonflow conditions the value of I z d In p in Equation (16)
is obtained by plotting values of z against values of log p and integrating over the
limits log 50 to log 1. The necessary data, read from Fig. 103 at a reduced tempera-
ture of 1.11 for reduced pressures from 0.0197 (1.0 atm) to 0.983 (50 atm) are tabu-
lated as follows:
P' p (atm) log p z
0.0196 1.0 0.0 1.00
0.1 5.09 0.707 0.98
0.2 10.18 1.008 0.96
0.3 15.27 1.118 0.94
0.4 20.36 1.304 0.90
0.5 25.45 1.406 0.87
0.6 30.54 1.484 0.84
0.7 35.63 1.552 0.805
0.8 40.72 1.610 0.77
0.9 45.81 1.681 0.73
0.982 50.0 1.699 0.71
The area under this curve is determined to be 1.59. Then, substitution in Equa-
tion (16) gives
w, = (1.987) (564) [(2.303) (1.59) + (1.00-0.71)] = 4440 Btu
This problem can be solved without graphical integration by means of Equa-
tion (18). ,„ : . ^ . , , v > -G. , ; . , , , : . . , , ; , ; , . . , : . ,; -.
Prom Pig. 106, at Tri = 1.11; Pr2 = 0.0197; Tn = 1.11; pn = 0.983
(HJ* - H2) = (0.04) (509) = 20; (nf - H,) = (2.30)(509) = 1170
From Fig. 107,

(a* - S2) = 0.03; (sf - Si) = 1.65
Prom Fig. 103, 22 = 1.00; 2i = 0.71

w, = 564(1.987 In 50 - 0.03 + 1.55) - (1170 - 20) + (1.987) (564) (1.0 - 0.71)
= 4,430
This is in agreement with the value obtained by the graphical method.
(c) For actxml-behaviorflowconditions, the work is expressed by Equation (21)
which involves the same integral as evaluated in part b; hence,
w. = (1.987)(564)(2.303)(1.59) = 4115Btu
Similarly, from Equation (22), the data developed in part 6, give
w. = 564(1.987 In 50 - 0.03 + 1.55) - (1170 - 20) = 4105 Btu
Where thermodynamic charts are available for any particular system,

the reversible work can be obtained directly. The applications of these
charts for evaluating the reversible work term under isothermal, isen-
tropic, fiow and nonfiow conditions are shown in Illustrations 2 and 3.
Illustration 2. From the data of Figs. 116-119, calculate the work performed in
the reversible isothermal expansion of 1 lb of ammonia initially at 200 lb per sq in.
and 150°F to 15 lb per sq in. under nonfiow andflowconditions.
From Fig. 116,
Fi = 1.740; F2 = 25.5cuftperlb
PiVi = (200) (144) (1.740) = 50,100 ft-lb per lb
P2F2 = (15) (144)(25.5) = 55,000 ft-lb per lb
A(,pV) = 4,900 ft-lb per lb
From Fig. 117,
Hi = 671 Btu per lb
Hi = 697 Btu per lb
Aff = 26 Btu per lb or 20,200 ft-lb per lb
From Fig. 118,
Si = 1.244 Btu/(lb) (°R)
& = 1.575 Btu/(lb) (°R)
AS = 0.331 Btu /(lb) (°R) = 257.5 ft-lb /(lb) (°R)
TAS = (610) (257.5) = 157,000 ft-lb per lb
Under ncmflow conditions, by Equation (11),
w, -= T AS - AH + A{pV) = 157,000 - 20,200 + 4900 = 141,700 fUb per lb

Underflowconditions, by Equation (19),
w, = -AH + TAS = -20,200 + 157,000 = 136,800 ft-lb per lb
CHAP. XIII] ISENTROPIC NONFLOW EXPANSION 545
Isentropic Nonflow Expansion. For reversible isentropic expansion

under nonflow conditions, from Equation (9),
{w,)s (nonflow) = £ iP dV)s = - ( A t / ) s = [ - A ^ + A(,pV)]s (23)
Equation (23) applies to isentropic nonflow conditions where changes

in kinetic, potential, and surface energies are negligible and where only-
mechanical work is involved. Calculation of the work is complicated
by the fact that the final temperature or pressure after expansion is
unknown and must be evaluated. For an ideal gas, from Equation
(XI-137), page 475,
du* = c * d r (24)*
The ideal-gas law may be written in differential form as pdv + v dp ==

R dT. Combining this expression with Equations (23) and (24), gives
V i r pdv= -vdp + RdT = -ctdT (25)*
Eliminating v from (25) and rearranging and combining with Equa-

tion (XI-134), page 474, gives
• V .^ -^dp=-{c: + R/-^=-c*/-^ (26)*
Where moderate temperature changes are involved, it is generally satis-

factory to assume a constant mean value of (c*)m at the arithmetic mean
temperature. Then Equation (26) is integrated to give
ii!ln(^)=(c:).ln(g) : ., (27)^
or
K
T2 r^^VsL ' • ;(28)*
From Equations (23), (28), and (24),

R
(«>r)s (nonflow) = {C*)UT. - T,) = [{cX - R]Ti [ l - ( ^ ) ^ ' ' ' ^ " ]
. :. (29)*
It should be noted that (Cj,)^ and (Cj,)m are different averages which
result from differential equations involving dlnT and dT, respectively.
The corresponding volume changes are obtained by combining the

ideal-gas law with Equation (28). Thus,
Equations (28), (29), and (30) are frequently written in terms of K, tlie
ratio of the heat capacities at constant pressure and constant volume.
For an ideal gas,
c* ct + R u;n !:;M P
(31) *
Cn — R c„
Combination of (31) with (28), (29), a n d (30), respectively, gives
T,= TJ^) :/':•': • (32)*

t-1
(«,.*)« ( n o n f l o w ) = - ^ [ l - ^ ) " ] (33) *
Equations (27) to (34) are restricted to expansions of ideal gases

involving small changes in heat capacity. Where the heat-capacity
terms are not satisfactorily represented by constant mean values,
empirical equations relating heat capacity to temperature should be used
in the integrations of Equations (25) and (26), as shown in Equation (37).
For actual gases work is evaluated more readily from the terminal state
properties using Equation (23) than from the pV integration. Since the
expansion is isentropic. As = 0, or
As = 82 - si = (s* - st) - (st - S2) + (sf - si) = 0 (35)
From Equation (XI-142), page 475, for 1 mole
/ • • • . ,
s* - si* = f \ldhxT - R\n- •. (36)"
where c,;, the molal heat capacity for ideal behavior, is a function of
temperature only.
Combination of (35) and (36) gives
^' • P2
/ cld\n T - R\n — - {4 - Ss) -1- (sf - Si) = 0 (37)
r. Pi
CHAP. X I I I ] I S E N T R O P I C N O N F L O W EXPANSION 547
The integral of Equation (37) is evaluated analytically from the empiri-

cal heat-capacity equation for the ideal gas. The equation is then
solved for the final temperature corresponding to a specified final pressure
and initial state. This is best done by assuming a series of final tem-
peratures and plotting the corresponding values of the left-hand side of
the equation. For ideal gases the two (s* — s) terms become zero.
Once the initial and final states are defined the reversible work is
calculated directly from Equation (23) from the terminal enthalpy and
pV values.
If condensation occurs during the expansion, Equation (37) is appli-
cable only until conditions of saturation are reached. A final tempera-
ture calculation from Equation (37) which is lower than the boiling point
at the final pressure is proof of condensation and establishes the correct
final temperature as the boiling point under the final pressure.
Where a chart of the thermodynamic properties of the system is
available, the final temperature is obtained directly by the following:
( X2\ X2
S2, - Y)^ ®^' ~ (^ ~ ^2) — (38)

or
A s = S2 - Si = (S2* - Si*) - (S*, - S2s) + (sj* - Si)
- (1 - X2) ^ = 0 (39)
where S2, is the entropy per mole of the saturated vapor and X2 is the
molal heat of vaporization at temperature T2 which is calculated by the
methods of Chapter VII. Equation (39) may be solved for quality X2 by
use of Equation (36), Fig. 107, and the temperature of condensation
corresponding to the final pressure p^.
Once the final quahty is evaluated, the work is calculated from the
enthalpy and pV changes of Equation (23). The enthalpies and
volumes of the saturated hquid and vapor are separately evaluated by
the methods previously demonstrated. If the generalized correlations
are used, the complete expression for the shaft work per mole is as follows:
(..)s(nonflow) = f \ U T - T. [ ^ - ^ ^ ^ ]
+ (l - x,)CK2 + P2V2) (40)

V2 = molal volume of the liquid phase at the final conditions.
Ordinarily the volume of the liquid phase at the final conditions is
negligible and the term p2f 2 may be omitted.
I
Isentropic Expansion, Flow Conditions. For an ideal gas, since heat

capacity is independent of pressure, Equation (10) may be written
{w*)s (flow) = - / Vdv= -AH =. - J^cldT (41)*
Equation (41) applies to the work in reversible isentropic expansion under

flow conditions and neglects changes in external kinetic potential, and
surface energies. Combination of Equations (12) and (41) for 1 mole
gives .
^ - £ Rd]iip= - £ 4d]nT , ^ , (42)* '
If a constant mean value of heat capacity is assumed. Equation (42)

integrates to give Equation (28), showing that for ideal gases the tem-
perature changes in isentropic expansion are the same for flow and non-
flow processes.
Combination of Equations (28) and (42) for 1 mole gives j
R ^ 1
{wt)s (flow) = {4)UTi - T,) = {ct)iT, [ l - (?)^"*^"'] (43)*
It may be noted that Equation (43) differs from (29) in the appearance of
(Cp)^ instead of (c, ),J, as the multiplying coefficient. /
The temperature and volume changes of isentropic flow expansion are
expressed in terms of the heat-capacity ratio by Equations (32) and
(34). By a similar development
t-i
(«^.%(flow) = ^ [ l - ^ ) " ] (44)*

For an actual gas under flow conditions the procedure is similar to that
followed in the isentropic nonflow case. Equations (35-39) are all
directly applicable to the flow case since they express the condition of
constant entropy. Thus, it may be concluded that for anyfluidthe
temperature changes in isentropic expansion are the same forflowand
nonflow processes.
If the final temperature is known, values of w, follow from Equa-,
tion (10). Thus, if the generalized correlations are used, the work per
mole is expressed by
-Hx H*- H,
(w,)s (flow) = fj>, dT - r . [—
r„ J
+ (1 - Xi)\2 (45)
CHAP. XIII] ISENTROPIC EXPANSION 549
When Equations (23) and (41) are compared, it is evident that the
work in isentropic flow expansion is greater than that of a nonflow opera-
tion by the amount of — A(pF).
Illustration 3. Calculate the work of isentropic expansion of 1 lb-mole of ethylene
gas when expanded from an initial pressure of 60 atm and 564°R to a final pressure of
1 atm under nonflow and flow conditions.
For nonflow conditions assuming ideal behavior, an equation for the heat capacity of
ethylene, at zero pressure in the temperature range below 200°F is given in Table XXI,
page 336.
4 = 7.95 + 8.13 X 10-"(r)'-»
Combining this equation with Equation (26) and integrating gives

C^' r7.95 1 1
J
I - ^ + 8.13(10-")(r)ii-85 \dT = R In— = 7.95
564 L -i J 50
+ 2.11(10-")[(r2)5-85 - (564)=-8=^]
This equation is solved graphically for Ti giving T2 = 234°R. The reversible work
of expansion is obtained by combination of Equations (23) and (25) with the heat-
capacity equation:
(^234 /»504
J CvdT = \ [7.95 + 8.13(10^")(D'-ss _ n] dT

564 1/234
= 2334 Btu per Ib-
/>564
(w*) s(flow) = / [7.95 + 8.13(10-")(r)3-s5] dT = 2990 Btu per lb-mole
«/234
In actual behavior, condensation occurs at Ti = 304°R, the boiling point at pressure
of 1 atm. The properties at the initial and final state are as follows:
I 2
Temperature, °R 564 304
Reduced temperature, T, 1.11 0.598
Pressure, p, atm 50 1.0
Reduced pressure, pr 0.982 0.0196
(H* - H) IT, (Fig. 106) 2.3 0.39
(s* - s) (Fig. 107) 1.55 0.64
« (Fig. 103) 0.71 0.98
The molal heat of vaporization at 1 atm is estimated from the available experimen-
tal data as 6200 Btu per lb-mole. From Equation (36),
s* _ s,* = 7.95 In ^ -F 2.11(10-")[(304)'-» - (564)'-85] - 1.99 In i

5D4 50
= -4.91 - 0.76 + 7.76 = 2.09
' • • . 304
X = 1 - g ^ 12.09 - 0.64 -f 1.551 = 0.85
\
The terms in Equation (40) may be evaluated individually, the volume of the liquid
being neglected. Thus,
• • " ' • * »
/»564
J
f
304
'304
[(7.95 + 8.13(10-")r3-86)] dT = 2417 Btu per lb-mole.
R{zJ'i - KiZjTa) = 1.987[(0.71)(564) - (0.85) (0.98) (304)1

= 290 Btu per lb-mole
- Hi VL* — H2^
508(2.3 - 0.39) = 972
(1 - x ) X = (0.15)(6200) = 930
• ' , - • • I ' .' . =,
Substituting in Equation (40)
{w,)s (nonflow) = 2417 - 972 - 290 -|- 930 = 2085 Btu per lb-mole '
{w.)s (flow) = 2417 - 972 + 930 = 2375 •
Illustration 4. Calculate from Figs. 116-119 the work of isentropic expansion of
1 lb of ammonia from an initial absolute pressure of 200 lb per sq in. and 150°F to a
final absolute pressure of 15 lb per sq in. under nonflow and flow conditions.
From Fig. 118, following a vertical line from the initial conditions to the final
pressure gives
„^„„ 1.243 - 0.032 „
Similarly, from Fig. 119 it is found that ', , .

Rx = 671 Btu per lb »
Hi (saturated gas) = 602.4 Btu per lb
Hi (saturated liquid) = 13.6
ffj (mixture) = x(602.4) -|- (1 - x)(13.6) = 537.6 Btu per lb
From Fig. 116, at the final temperature, pressure, and quality:
Y% (saturated gas) = 17.6 cu ft per lb
xpiVi = (0.89)(15)(144)(17.6) = 33,800 ft-lb per lb
The volume of the liquid phase may be neglected.
PiFi = (200) (144) (1.74) = 50,100 ft-lb per lb ;
Using the data of Illustration 2 for the initial state, we get
T
Aff = 538 - 671 = - 1 3 3 Btu per lb or -103,300 ft-lb per lb
A(p7) = 33,800 - 50,100 = -16,300 ft-lb per lb
For nonflow conditions, from Equation (23), - f,',.
M). = 103,300 - 16,300 = 87,000 ft-lb per lb
For flow conditions, from Equation (40),
to. = 103,300 ft-lb per lb
CHAP. XIII] JOULE-THOMSON EXPANSION 661
FREE EXPANSION
The unrestrained expansion of a gas is known as free expansion.
Under conditions of no restraint no work is done, and under adiabatic
conditions no heat is added. Free expansion under flow conditions is
commonly known as throttling or as the Joule-Thomson effect referred
to in Chapter X I I .
The Joule-Thomson effect is measured experimentally by expanding
the gas slowly and in steady flow through a well insulated porous plug;
in this way potential work is lost, and no heat is allowed to enter or leave
the system through the walls. The fluid flows reversibly into and out
of the process, but the expansion step is completely irreversible. Since
no heat is added or lost, the process takes place imder conditions of con-
stant enthalpy, according to Equation (VII-11), page 207. However a
change in internal energy results if any change in flow energy pV occurs.
For an ideal gas, since enthalpy is independent of pressure, no change
in temperature occurs in free expansion imder either nonflow or flow
conditions, and values of AH, A(pV), and AU are zero. However, an
increase in entropy results which is equal to —J? In P2/P1 by Equa-
tion (XI-142), and the values of A A and AG are each equal to —T AS.
For nonideal gases, the temperature change in Joule-Thomson expan-
sion is of great experimental and practical value. Such data are used for
establishing the deviations from ideal behavior of real gases, particularly
the effects of pressure on enthalpy and entropy. Free expansion under
flow conditions is made use of industrially in the cooling and liquefaction
of gases. ' , : '.
Free expansion under nonflow conditions, known as the Maxwell
effect, is of little practical value. Because of the small heat capacity of a
gas compared with that of the vessel, adiabatic conditions cannot be
approached, and experimental measurements are of uncertain accuracy.
In the foregoing discussion it was assumed that kinetic energy changes
were negligible as is generally the case. Where this is not true, the
general energy-balance equation including kinetic-energy changes must
be employed. Changes in both internal energy and temperature may
result from the free expansion of an ideal gas at high velocities in either a
flow or nonflow system.
Joule-Thomson Expansion. The Joule-Thomson coefficient is given
by Equation (XII-35), page 501. Temperature changes in free expansion
where only small changes are involved and where the heat capacity may
be taken as constant may be calculated from Equation (XII-37).
Where large temperature changes are involved, these calculations are
more conveniently carried out through differences in state properties.
A Joule-Thomson expansion is represented on a thermodynamic chart

by a line of constant enthalpy. By following such a line from the initial
state to the final pressure, the final temperature and state are determined.
The state changes accompanying Joule-Thomson expansion also may
be calculated from the generalized correlations of thermodynamic proper-
ties. Since the expansion is isenthalpic, where changes in kinetic energy,
potential head, and surface energy are negligible,
AH = £ ' . ; . r - r . [ s J ^ - ! ^ ] = o (46)
Equation (46) may be solved by assuming values of T2 until the con-
ditions of the equation are satisfied by graphical or trial methods.
If condensation occurs during the expansion, the final temperature
will be the boiling point at the final pressure. Equation (46) must be
modified to include the enthalpies of the liquid and vapor phases and
the quality X2. Thus,
AH = £ C U T - T . [ 5 ^ _ 5 t ^ ] _ (1 _ , , ) x , = 0 (47)
where X2 is the heat of vaporization at the final conditions and H2s is the
enthalpy per mole of the saturated vapor at state 2. Equation (47) may
be solved for xa since the final temperature T2 is known.
After the final temperature T2 is evaluated, the changes in the other
state properties follow. Thus, .-.
A{vv) = XiZiRTi + (1 - xi)p2VLi - ZiRTi (48)
where VL2 is the molal volume of the liquid phase.
As = / c* d In r - i? In — — (s^ - Ss,) + (sf - Si)

JT, • Pi
, ! -^(l~x,) (49)
Au = AH — Aipv) = ZiRTi — X2Z2RT2 — (1 — X2)P2VL2 (50)

AA = ziRTi — X2Z2RT2 - (1 — Xi)p2Vi — T2&2S . ;.
+ X2(1 -X2) +Ti&i • : ' - . <51)
AG = -r2S2, + X2 (1 - Xi) + TiSi ,, - ; (52)/
If continuous free expansion is carried out at high velocities, the
changes in kinetic energy are not negUgible. In this case, from the^
general energy-balance Equation VII-1, page 205, it is evident that if
potential and surface energy changes are negligible a .decrease in enthalpy
CHAP. XIII] MAXWELL EXPANSION 553
occurs which is equal to the increase in kinetic energy. Thus, from an

energy balance where no condensation occurs:
AH = Z ! ^ ( | Z 1 ^ = (H* - Hf) - T . ( ^ )
^ + T . ( ^ ^ (53)
The final temperature T'2 may be evaluated from Equation (53), and
the other state properties may be calculated from Equations (48-52).
Maxwell Expansion. As previously pointed out, adiabatic free expan-
sion under nonflow conditions is characterized by constant internal
energy. Thus, ^
AU = 0 = AH - A{pV) (54)
Equation (54) may be expressed in terms of generalized corrections and

ideal behavior to give, in the absence of condensation,
A. = f \ t dT - T. ( ^ ) + r . ( 5 ^ ) - Z.RT.
+ ziRTi = 0 (55)
The final temperature T2 is determined by trial and error or by
graphical solution of Equation (55). In the case of condensation.
Equation (55) is modified by subtracting the term (1 — Xi) (X2 + PiVn)
and replacing z^RT^ by xa2RT2.
ThermodjTiamic charts of the type of Figs. 116-119 do not permit
direct solution of problems in Maxwell expansion. A trial-and-error
procedure is necessary in which a final state is assumed, and the assump-
tion is tested by means of Equation (54).
From a knowledge of T2, the changes in the other state properties
may be evaluated from equations similar to (48) to (52).
niustration 5. Estimate t h e changes in state properties in t h e free expansion of
ethylene gas from 564°R and 50 a t m to 1 a t m under flow and nonflow conditions.
The absolute entropy of ethylene at 77°F and 1 a t m is 52.48 Btu/(lb-mole) (°R).
T h e initial absolute entropy of the ethylene gas is 43.72 B t u / (lb-mole) (°R).
Ti = 564°R; Tri = 1.11

pi = 50 a t m ; p,i = 0.982; zi = 0.71
From Fig. 106, H^ - Hi = 1218
From Fig. 107, s* - Si = 1.55
(o) Flow conditions {Joule-Thomson effect). The final temperature is evaluated by

graphical solution of Equation (46). A trial value of T2 — 460°R is assumed. Then
.. 7,2=0.905; pr2 = 0.0196
0.0196
From Fig. 106, HJ - HJ - (0.1) (508) = 33
U.Uo
The term 0.0196/0.03 represents the hnear factor for pressures which are off the
chart. Atpr = .03; (H* - a)/T^ = 0.10. From the data of Illustration (3),
' /^460
J '
564
17.95 + (8.15) 10-"r3-85]dr = -1060
, •
Substitution of these values into Equation (46) gives

AH = -1060 - 33 + 1218 = 125 Btu per lb-mole
By repeating the preceding procedure a value of AH = 0 is obtained when Tj =
448°R. At this temperature there is no condensation and Xi = 1. The change in
entropy is evaluated from Equation (49).
At state 2,
448 -<i«>i.
^ ^ = S;^ = 0-882; Pr2 = 0.0196; zj = 0.99 " £."!

608 r*ma,'
From Fig. 107, sj - S2 = -j—— (0.1) = 0.07

\
r^^ r7.95 1 1
J ' \-i;r + (8.13)10-"r2-85 \dT -Rln—
564 L r J 50
= 6.45
Substituting in Equation (49), yields
As = 5.45 - 0.07 + 1.55 = 6.93
Hence, S2 = 6.93 + 43.72 = 50.65 Btu/(lb-mole) (°R)
The change in internal energy is evaluated from Equation (50)
Au = (0.71) (1.987) (564) - (0.99) (1.987) (448) = - 8 8 Btu per lb-mole
A(rs) =448(50.65) - (564) (43.72) = -1975 Btu per lb-mole
AA = Au - A(rs) = - 8 8 + 1975 = 1887 Btu per lb-mole
AG = AH - A(rs) = 1975 Btu per lb-mole
u, = 0; g = 0; v^ = (0.99)(0.729)(448) = 322 cu ft per lb-mole
(6) Nonflow conditions {Maxwell effect). The final temperature is determined by
graphical solution of Equation (55) by the same procedure followed in solving Equa-
tion (46) in part a. Thus, if 460°R is taken as a trial value of T2, from part a,
H2 - H* = 1060 and -(H^ - H2) + (H* - HJ = - 3 3 + 1218
Values of z from Fig. 103 give
ZiRTi = (0.99) (1.987) (460) = 9 0 3 ' '
Ziflri = (0.71)(1.987)(564) = 793 ,. '
Substitution in Equation (55) gives Au = -1060 - 33 -|- 1218 - 9 0 3 - 1 - 793 = 15
Since Au > 0, the value of 460°R is too high. At Au = 0, the value of T2 is found
to be 458°R. Then
Tri = 0.90; p,2 = 0.0196; Z2 = 0.99
CHAP. XIIIl EXPANSION OF ETHYLENE 555
o o o o o o o o o o
i> r^ lo >o t^ t - o o
CO CO 0 5 OS CO CO CO OS
lOiO
^T" ^ ^ ^ ^ ^f^ ' ' ^ ^ ^
.—(
I—1 I—i
11 1 T 7 7
o o o o o o o o
<1
r>-1>» CO ^ t ^ 1>"
c o c o OS c : ^ ^ ^
1 1 ;:^;^ 1 1
o o
Tt< - I *
11
o
o>
o
o> 1 OS
1-^
O O o "Ô O O oco
i > . i>-,—1 >—(i> t ^ 00 00 00
CO CO CD CD CO ^7 ccco CO CO CO 00
< 03 C33
1 1 22 1 1 11
03
o o o o
cnro o o t» o
w o o g g o o <N (N CO
IN
CO
(N
o
<]
< 1 1 1 1
n
m ^
(N
lO
t. 10 COCO 00 00
V. TtH p COCO 00
o o o
< 00 00
11 1 1 1
toco coco o o o
O O O C D O O
coco
coco
o o §8
11 1 1
o
T3 II coo
I? a t^lr^cooi — _ 00 »C 00
(jQ
coo o o
X g" cococoo>°° o CO
IN
o o o o
So Oi ^ J ^ o o o o o o o o
lOiO
o o CO CO
m coco o o OS
<l 003 1> CO
(D tl-l
^ Ol OS lO l O O l C i 03 05 o o o
00 ^
p.
o o o d o o o OS (N
^ ^ ^ | > , | > ^ r-H oo 00 IN N
o w
+
C5 a;
o -yl ^ ^ l—t I—* ^ T^
I—1
CO CO
t-^ ,-1 ^ P ^ ^ ^ ^ '^T' ^ ^ ^^ 00

r-> CO O CO CO O ;D CO CD o o
inio CO CO
o
o rt "-i
IS CO .
« o —'
aa
oj o ft S
« ir ^ !>>
S&^-s S
C3
I
.a
JS ^ i N co-*mco
/
From Pig. 106, HJ - Hz = 33
From Pig. 107, s' - s^ = ( ^ | j ^ ) (0.10) = 0.06

From Equation (49),
As = / -^ B In - - 0.06 + 1.55 = 7.;.13

i/564 ^ 50 ,
Si = 7.13 + 43.72 = 50.85 Btu/(lb-mole) (°R)

A(pV) = ziRT, - ziRTi = (0.99) (1.987) (458) - (0.71) (1.987)(564) = 110 Btu per
lb-mole
AH = ATT + A(pi;) = 0 + 110 = 110 Btu per lb-mole
A(rs) = TiSt - TiSi = (458) (50.85) - (564) (43.72) = -1370 Btu per lb-mole
AA = Au - A(rs) = 1370 Btu per lb-mole
AG = AH - A(7's) = 110 + 1370 = 1480 Btu per lb-mole
U2 = (0.99) (0.729) (458) = 329 cu ft per lb-mole f
In Table X X I X are summarized the changes in point and path

properties when ethylene gas is expanded from 50 atm and 564°R to
1 atm under various conditions of restraint and no restraint when both
ideal and actual behaviors are assumed. It may be observed that for
both ideal and actual behavior the changes in point properties under
reversible conditions are identical for both flow and nonflow conditions,
but work and heat terms are dependent upon the path and conditions
of flow. For all reversible conditions q = J'TdS, and each of the
work terms is identified with one of the energy functions. In free expan-
sion, the heat and work terms are zero, and the changes in point proper-
ties for actual behavior are different for flow and nonflow conditions. _
It is interesting to note that free expansion cools ethylene gas to 458°R
without condensation, whereas by isentropic expansion the gas is cooled
to 304°R with 15 per cent condensed.
s.
i , * ' ; ; - ; • — - • . '
CYCLIC PROCESSES
A series of operations so conducted that all changes are periodically
repeated in the same order is a cyclic process. This may comprise a
sequence of either nonflow or flow operations or a combination of the '
two types, either repeated on the same substance or involving a new
mass of substance for each cycle. For example, a reciprocating air
compressor produces a repeated series of nonflow operations in which the
conditions at a selected point of any cycle duplicate those at the same
point of any other cycle. However, a new mass of gas is involved in
each cycle, and a flow result is produced by the sequence of nonflow
cycles.
CHAP. XIII] THE CARNOT CYCLE 557
On the other hand, in a steam-power plant the change may be con-

tinually recirculated through a series of operations such as feed-water
heating, evaporation, superheating, expansion, and condensation. Such
a cycle is essentially composed of flow operations but may involve com-
ponent nonflow steps such as the expansion of steam in a reciprocating
engine.
Nonflow cyclic operations repeated on the same mass of substance
have no present practical applications but are of great value in estab-
lishing the theoretical behavior of cyclic processes in general.
Reversible Cycles. A cycHc process which is wholly reversible must
of necessity be composed of individual steps each of which is in itself
reversible. This requires that at no point in the cycle may there be
degradation of higher forms of energy to heat through mechanical or
fluid friction, nor may there be any other irreversible steps such as free
expansion of a fluid or the transfer of heat under a finite temperature
difference.
The concept of reversibihty as discussed in Chapter X I requires that a
reversible cycle may be operated in either a forward or a reverse direc-
tion and must return to any given initial state without change in any of
the state properties of the system. Thus, it follows that in any reversible
cycle the algebraic sum of the changes in any state property for all steps
must be equal to zero. Thus, for any reversible cycle,
HAS /V'-» w
The Camot Cycle. It has been previously stated that the maximum
shaft work accomplished as a result of any change in state of a fluid is
obtained when the change takes place reversibly without friction or
turbulence. The Carnot engine has been conceived as a hypothetical
device which represents this maximum achievement in a cyclic process
and hence is used as a standard in evaluating the efficiency and per-
formance of all actual mechanical cycles for transforming heat energy
into mechanical energy.
In the Carnot engine, a gas is contained in a cylinder equipped with a
frictionless piston connected to a frictionless receiver of mechanical
energy. The cylinder walls and piston are impervious to heat, but the
cylinder head is interchangeable alternately to an impervious or to a
highly conducting plate. In the first stage of the cycle, that of iso-
thermal expansion, heat is supplied to the gas through the conducting
cylinder head from a source of heat maintained at a constant temperature
Ti and flows into the fluid as a result of an infinitesimal temperature
drop. The gas is allowed to expand isothermally and reversibly with an
influx of heat equal to gi. The reversible work of expansion from state a
to state b ia I p dV.
The first stage is followed by a reversible adiabatic or isentropic
expansion which is accompUshed theoretically by attaching the non-
conducting head to the cylinder and permitting expansion to continue.
Work of adiabatic expansion is performed at the expense of the internal
energy of the fluid and in expanding from state h to state c is equal to
/ p dV. In this stage the temperature decreases to T2 + dT.
A receiver of heat is now provided by replacing the impervious cylinder
head with a conducting plate in contact with the receiver at a constant
temperature T2. In the third stage, an isothermal compression is per-
formed wherein work of compression / pdV is performed on the gas, and
an amount of heat —52 flows into the receiver at Tt. In the fourth and
last stage, the cylinder head is made nonconducting, and compression is
continued under isentropic conditions from state d to the original state a
with a reversible work of compression equal to / pdV.
Since by completing the cycle the fluid is returned to its original state,
its net internal energy change is zero. The same is true of all other point
properties. Since no heat was received or lost during either adiabatic
change, it follows from the law of conservation of energy that the net
work done, Wnet, must equal qi + q-i. This is equal to the difference
between the amount of heat received from the reservoir at temperature
Ti and the amount rejected to the receiver at temperature T2. Thus,
w^net = £vdV + £vdV + £pdV + £vdV / ' -
\ dV = qi+ q, (57)
- / '
The change in entropy of the fluid during isothermal expansion is

A»Si = Sb — Sa = qi/Ti and in isothermal compression is ASi =?= Sd —
Sc = q^/Ti. Since the change in entropy over the entire cycle is zero,
and the change during each reversible adiabatic stage is zero, it follows
that ASi = —AS2. Hence, ;
Qi + q2-= AS,{Ti ~ T2) (58)

CHAP. XIIIl THE CARNOT CYCLE 559
Since qx + q^ = Wnet and qx = Tx AtSi, it follows that
Wnet ^ gl + g2 ^ Ti - Tg
(59)
gi ~ 21 ~ Ti
The ratio Wnet/gi expressed by Equation (59) is termed the thermody-
namic efficiency of • he cycle. It will be recognized that an efficiency of
100 per cent can be realized only when the temperature of the receiver is
at absolute zero.
The operation and efficiency of the Carnot cycle is shown graphically
by straight lines on the temperature-entropy diagram, Fig. 123. Dur-
ing the first stage, the reversible iso-
thermal expansion at temperature Ti is -
'
represented by the horizontal line ab;
the heat removed from the source is
represented by the area SaObSb or TiAS.
During the second stage, the reversible " a/l.V
adiabatic expansion is represented by 9i
the vertical line be with no change in / . '\
entropy and with no heat added. In / d 1 J- c \
the third stage, heat is rejected at Ti
along horizontal line cd and is repre- -92
sented by area Sa dcSb or T2 ASi. In 1 1 1 1
the last stage the adiabatic compression Sa

moves along a vertical line da of constant
entropy and with no heat rejected. The FIG. 123. The Carnot Cycle.
net work done Wnet is represented by
area abed equal to gi + q^. The thermodynamic efficiency of the cycle
may be visualized as the ratio of area abed to area abcSbSad.
The Carnot cycle also may be operated by starting with a saturated
liquid and vaporizing it during the first stage. This condition is indi-
cated on Fig. 123 by the curved line which represents the temperature-
entropy relationship of the saturated vapor and liquid of the working
fluid. On such a vapor cycle two phases exist in the cylinder at all
points except a where the system is completely liquefied and b where it is
completely vaporized. Other methods of operation might involve
varying degrees of hquefaction at the various points of the cycle, but in
all cases the energy transformations are expressed by Equations (57-59).
The concept of the Carnot eng'ne permits visualization of the signifi-
cance of the Carnot principle and the second law of thermodynamics.
The Carnot principle involves two basic conclusions which follow directly
from Equations (56-59) and the second law of thermodynamics.
1. No self-acting engine which operates by the absorption of heat at

one constant temperature and by the rejection of heat at another lower
constant temperature can be more efficient than an engine operating on a
reversible cycle with absorption and rejection of heat at only these same
two temperature levels. If such an engine of higher efficiency existed,
it could be used to drive a reversible Carnot engine operating as a heat
pump. The net result of the two engines would then be the transfer of
heat from a lower to a higher temperature, in contradiction of the second
law.
2. The thermodynamic efficiencies of all reversible cycles which absorb'
heat at the same constant temperature and reject heat at the same con-
stant lower temperature must be equal, regardless of the nature of the
system or the fluid. This conclusion follows from Equation (59) which
expresses efficiency, independent of system or fluid characteristics.
Although this equation was derived by consideration of the Carnot
engine, it is applicable to any other engine operating between two con-
stant temperatures as required by Equation (55).
The Availability of Thermal Energy. The Carnot principle furnishes
a basis for quantitatively expressing the availability of thermal energy
in performing mechanical work. Thus, Equation (59) expresses the
maximum fraction of the heat energy gi which can be converted into
work by the most efficient machine possible when receiving heat from a
temperature Ti and rejecting heat at temperature T^. The quantity
(?! + 92) may thus be termed the available portion of qi with respect to
temperature T2. Similarly, ga is the unavailable portion with respect to
this same temperature of rejection. An expression for the fractional
unavailability of energy is obtained by rearranging Equation (59).
—?2 T2
Fraction unavailable = = — , (60)
The temperature of rejection is ordinarily fixed by the temperature of

the atmosphere or the available cooling water. As the temperature level
of the energy source is reduced, its availability is decreased, and, when
the temperature of the source becomes equal to the rejection tempera-
ture, heat energy becomes wholly unavailable, even in an ideal engine.
POWER-PLANT CYCLES
The production of shaft energy from the great store of chemical
energy in fuels is accomplished by the highly irreversible process of com-
bustion, whereby heat is released at high temperatures and transferred
to a working fluid at a much lower temperature. In a steam-power plant
CHAP. XIII] POWER-PLANT CYCLES 561
part of the heating value of the fuel is transferred to water in a boiler;
the steam generated expands against the piston of a steam engine, or
through a turbine and is then condensed with rejection of heat at a lower
temperature to a cold fluid flowing in a condenser; the condensate is
pumped back into the boiler. A flow diagram of these operations is
Superheater
Engine
^Boilen^
, Pump
6 „ a Condenser
FIG. 124. Rankine-Cyole Power Plant.
shown in Fig. 124. When both expansion of steam and pumping of the
liquid condensate occur reversibly, the Rankine cycle is followed. This
cycle differs from the Carnot in that each stage is carried out in a differ-
ent apparatus and particularly in that condensation is completed at the
h 5 Pi \
c' of \d pm d
\ e
P, « a
by e \ \
/r 1
\
1
b'a' d'
V
FIG. 125. Rankine-Cycle Diagrams.
low temperature and pressure of the condenser instead of by isentropic
compression. The condensate is pumped into the boiler to mix with the
hot liquid.
The changes taking place in the Rankine cycle are shown graphically
in the HS, TS, and pV diagrams in Figure 125. In the first stage the
feedwater is pumped from pressure p^to pi with an increase in enthalpy
from HA to HB, as shown on HS diagram, and with a negligible rise in
562 EXPANSION AND COMPRESSION [CHAP. X H I
temperature, shown as a to b, on the TS diagram. The water enters the

boiler subcooled, that is, at a temperature below that corresponding to its
equilibrium pressure. The entropy remains constant during pumping,
since the process is assumed to be reversible and adiabatic, as indicated
by the vertical line a6 on t h e / f 5 diagram.
The work of pumping is equal to / F dp. Since water is nearly j
incompressible, the work input becomes
-Wl = (pa - Pb)Vf <61)
where v/ is the average specific volume of the feed. The work of pumping
is represented on the pV diagram as the area mban. If the pumping
operation is a reversible process, and changes in kinetic and potential
energy are negligible, the work done upon the fluid is also equal to the
increase in enthalpy, or
-wi = Hi-Ha ,: , - (62)
In the second stage the water is heated along line bed, becomes satu-
rated at point c and temperature Ti and then vaporizes at constant
temperature Ti and pressure pi to form saturated vapor at d. The TS
diagram shows the wide departure from isentropic conditions in heating
the liquid from Ti to Ti, as represented by the slope of the line be and the
area bee'. The gain in enthalpy is represented by Hd — Hb on the HS
diagram. If kinetic-energy terms are neglected, the heat supplied is
equal to Hd — Hb or I T dS, which is represented on the TS diagram
by the area SiobcdeSi. Thus, > ,
qi = Hd-Hb= f'TdS • . • (63)
' • • i " *
In the third stage the vapor expands isentropically from d to e. If the

expansion is a flow process with negligible changes in kinetic energy, the
work done is equal to the decrease in enthalpy as represented by Hd — He
on the HS diagram. On the pV diagram the work done during this
stage of isentropic expansion is represented by the area m den.
In the condensation stage the enthalpy loss is represented by He — Ha
on the HS diagram and as area SiaeSi on the TS diagram, both of which
equal the heat ~q2 absorbed by the condenser.
The net work of the cycle represents the engine work minus the small
work input to the pump. Since the process is cyclic with no change in
any point property, it follows from (XI-2) that the net work done is
CHAP. XIII] POWER-PLANT CYCLES 563
equal to the heat absorbed less the heat rejected if kinetic energy changes
are negligible. Thus,
np2
J ' Vdp- {pi- p2)Vj = {Hi - H,) - (Hi - Ha)

= qi+Q2 (64)
The net work done during the cycle is represented by (Hd — He) —
(Hb — Ha) on the HS diagram, by the area abde on the pV diagram and
by the area abcde on the TS diagram.
The ideal thermodynamic efficiency of the Rankine cycle is repre-
sented by
Waei 9l + Qi (Hd - H,) - (HI, - Ha)
(65)
; 2i 2i . Hd — Hb
or, if pump work is neglected,

, i •"'net Hd - He
(66)
qi Hd — Ha
The significance of the efficiency of the Rankine cycle is best visualized

from a study of the TS diagram where it is represented by the ratio of the
area abcde to area SibcdeS2. Any change in conditions which will enlarge
the former area with respect to the latter will improve the efficiency. It
is evident that the efficiency of the Rankine cycle is lower than that of a
Carnot cycle operating between the same extreme temperatures. This
results from the fact that in the Rankine cycle a portion of the heat
absorbed is delivered to the process at intermediate temperatures along
the line be representing the heating of the liquid to saturation at tem-
perature Ti. This difference is. shown in Fig. 126a where the efficiency
of the Carnot cycle is represented as —- and of the Rankine
area (A + C)
areas (A + B) :. .' ,., ./
cycle as ——;—„ , r^ i ^x *
•' area (A + B + D + C)
The efficiency of the Rankine cycle can be increased by increasing the
temperature and pressure of the saturated steam supply or by lowering
the temperature at which condensation takes place. For the same
temperature change it is evident from Fig. 126b that a decrease in the
condensation temperature is more effective than an increase in the
temperature of the saturated steam. Increased efficiency obtained by
lowering the temperature of condensation is limited by the temperature
of the cooling-water supply. For boilers operating at the same pressure,
greater efficiency can be expected from this source in cold than in hot
climates. Several degrees temperature difference must be allowed for the

flow of heat from condensate to cooling water to permit economical
designs of condensers at the expense of reduced efficiency of conversion of
heat into work.
An increase in efficiency produced by increasing the temperature of
saturated steam is limited by the convergence of the saturation lines as
shown on the TS diagram. It is evident that the gain due to increased
temperature and pressure diminishes as the temperature approaches the
\d
r,--
r^
II 9\\
B iri \
IL
Si S4S3S2 £
( a ) Camot and Eankine Cycles (c) Effect of Superheating (e) EfEect of Eeeenerative
Heating
T
Ts •-/•-n
T.
•'I
7;.
-/ ^ "a
3 ,Q
T2
S4=
-i
/!
III m^
/1
1
Sy S2S3S4 g Si S3S2 g
(6) Effect of Increasing High (d) Effect of Superheating (/)Effect of using
Pressure and Decreasing and Heheating Process Steam
Lew Pressure
Fio. 126. Improvements in Cycle EfEciencies.

critical. The maximum useful temperature and pressure are also
limited by the cost and strength of available materials of construction.
To obtain the advantage of high temperature without the disadvantage
of high pressures, some fluid other than water may be resorted to, such as
mercury or diphenyl. A dual power plant employing both mercury and
water has been used for increased efficiency of stationary power plants.
Improvements over the Rankine Cycle. Improvement in efficiency
over the Rankine cycle is obtained by superheating the vapors before
expansion as shown in Fig. 126c. This also increases the quality of
steam after expansion and thus reduces erosion and friction due to
condensate in the engine or turbine. The temperature of superheating is
limited by the available materials of construction. Higher efficiencies
are also obtained by reheating in combination with multistage turbines or
CHAP. XIII] HEAT EXCHANGERS 565
engines. In this scheme the superheated steam at high pressure partially

expands in the first stage and is then again superheated in an auxiliary
heat exchanger in the furnace. The reheated steam is expanded in a
second stage as shown in Fig. 126d. This reheating increases efficiency
and eliminates condensation in the expansion stages of the engine or
turbine. The advantage of reheating is partly offset by the cost of the
auxiliary heat-exchange surface required which operates at a lower
pressure than the primary steam supply.
Regenerative heating of the feedwater to the boiler is another means of
increasing efficiency where multistage turbines or engines are used. In
this scheme steam is bled from the turbine progressively as expansion
proceeds, and this stream of fluid of steadily decreasing temperature is
used to heat the accumulating condensate progressively and counter-
currently as it is returned to the boiler. If this heat exchange could be
accomplished with a minimum temperature drop, the efficiency of the
Rankine cycle could be made to approach that of the Carnot cycle. This
hypothetical continuous preheating scheme is designated by dotted line
dg in Fig. 126e. The efficiency in the limiting case will approach the
ratio of A/A -f- B which is that of the Carnot cycle. Actually the results
are limited by the finite steps in multistage expansion, as represented by
the four stages in Fig. 126e. By regenerative heating more steam is
actually supplied to the turbines per horsepower-hour developed, but less
heat is rejected to the condenser. Again the advantage of regenerative
heating is offset by the cost of auxiliary heat-exchange equipment, but
the gain in efficiency is real, and distinct gains in economy of operation
are achieved in large power plants.
Process Steam. Of particular interest in chemical manufacture is the
use of process steam. Where multistage power production is available,
process steam can most economically be suppHed by generating steam
at the highest pressure available and bleeding off the desired amount at
the temperature required for the process, from the lowest stage per-
missible. Where a quantity of heat represented by area C on Fig. 126f
is desired for process heating at temperature T', an amount of work
energy equivalent to area D can be obtained by generating the steam
first at the higher temperature Ti. This work can be achieved at httle
additional fuel cost and little increased boiler capacity. The work
accompanying the generation of process steam in this manner is virtually
obtained-by the absorption of an equivalent amount of heat. Obviously,
to generate process steam at temperature T' would permit no power
generation, and all heat would be rejected to the process.
Heat Exchangers. The heat removed from any fluid in cooling loses
least in entropy and hence retains highest availability if this transfer is
made under the smallest temperature drop. The transfer of heat
approaches reversibility as the temperature difference diminishes

towards zero, whereas direct mixing of a hot stream with a cold stream is
most inefficient and wasteful. Ideal heat transfer is approached by
countercurrent exchangers. The actual temperature drop for the trans-
fer of heat will be fixed by the cost of the exchanger.
A comparison of the advantage of countercurrent heat exchange with
that of parallel flow or mixing can be illustrated by considering a given
mass of a hotter fluid of a given heat capacity exchanging heat with the
same mass of a colder fluid of the same heat capacity. If the hotter
fluid cools countercurrently from temperature Ti to T2, and the colder
fluid is heated from temperature T2 to Ti, then the loss in entropy of the
hot fluid is exactly compensated by the gain in entropy of the cold fluid,
and there is no net loss in entropy. The heat exchange is reversible in
this limiting case. In the case of parallel flow or mixing, the hotter
fluid is cooled to temperature Ts, and the colder fluid is heated to tem-
perature Ta. The hotter fluid loses entropy to the extent of
{Cp)m In — , and the colder fluid gams in entropy by the amount
-t 1 ;
(Cp)m In - ^ • The net gain in entropy is the algebraic sum of the two
-'2
changes or / T^ \
It can be shown mathematically that for all possible values of Ts in

Equation (67), A^ > 0.
Engine Efllciency and Performance. The gross work performed by a
Rankine cycle results from the flow and expansion of the fluid through
the engine. As previously pointed out, the net work is slightly lower
than this, because of the work required for pumping. Maximum gross
work is obtained by ideal isentropic expansion as expressed by Equa-
tion (45). However, actual engines as a result of mechanical friction
and deviations from isentropic conditions of the fluid deliver as shaft
work an amount of energy less than that calculated from Equation (45).
The ratio of the shaft work actually delivered to that ideally obtainable
by isentropic expansion between the existing terminal conditions is
termed the engine efficiency. Thus, from Equation (45),
„ . actually delivered Ws
Engme efficiency = — — — (67a)
Hi — H2
Engine efficiencies vary widely with size and design, ranging from as
high as 80 per cent for large well-designed turbines and reciprocating
engines down to 30 per cent or less for small units of inferior design.
CHAP. XIII] ENGINE EFFICIENCY AND PEIRFORMANCE 56?
Engine performance is commonly expressed by the steam rate or water

rate which is defined as the pounds of steam used per horsepower-hour
produced. One horsepower-hour equals 2545 Btu. It is evident that
the steam rate is affected both by the operating conditions and the
engine efficiency. The ideal water rate for any engine may be calculated
from Equation (45). The actual steam rate is then the ideal rate
divided by the engine efficiency.
Ideal steam rates vary from as low as 5 to 10 lb with superheated high-
pressure steam and vacuum exhaust to more than 40 lb where high
exhaust pressures are used for the production of process steam. With
saturated steam at 100 t o 200 lb per sq in. in exhausting at atmospheric
pressure ideal steam rates vary from 18 to 24 lb.
The factors of mechanical design and operation which determine engine
efficiency and performance are discussed in texts on mechanical engi-
neering thermodynamics.!"'
Illustration 6. Calculate the ideal thermodynamic efficiency of a steam-power

plant operating on the Rankine cycle in which dry saturated steam is generated at a
gauge pressure of 150 lb per sq in. and expanded through an engine exhausting at
atmospheric pressure. Compare this efficiency with that of a Carnot cycle operating
between the same temperatures, and calculate the ideal water rate of the engine
Solution: The following data are taken from the steam tables:^
Gauge pressure, lb per sq in. 0 150

Absolute pressure, lb per sq in. 14.7 164.7
Saturation temperature, T, °F 212 365.9
Enthalpy of saturated vapor, Btu per lb 1150.4 1195.6
Enthalpy' of saturated hquid, Btu per lb 180.1 338.4
Entropy of saturated vapor, Btu /(lb) (°R) 1.757 1.562
Entropy of saturated Hquid, Btu /(lb) (°R) 0.312 0.523
Specific volume of saturated liquid, cu ft per lb 0.0167 0.0182
From Equations (38) and (39), the nomenclature of Fig. 125 is used,
Sd - S. = 0 = 1.562 - [1.757* + (1 - x)(0.312)l = 0 • " /.

X = 0.865, the quality after expansion
H. = (1150.4)(0.865) + (180.1)(0.135) = 1019.5 Btu per lb
From Equation (61), w , • ,; : h ; . ; • ; :
Work of pumping = {m - Ha) = (164.7 - 14.7) (144) (0.0167)/778 •

= 0.465 Btu per lb
1 J. H. Keenan and F. G. Keyes, " Thermodynamic Properties of Steam," John.

Wiley & Sons, New York (1936).
* P. J. Kiefer and M. C. Stuart, " Principles of Engineering Thermodynamics,"
John Wiley & Sons, New York (1930).
' J. H. Keenan, " Thermodynamics," John Wiley & Sons, New York (1941).
From Equation (65),

Wnet 1195.6-1019.5-0.465 „ ,„„ , . , , .u j • ^-
—— = = 0.173 or the ideal thermodynamic efficiency is
g, 1195.6-(180.1 +0.465) 17.3 per cent.
From Equation (59) the corresponding Caraot cycle efficiency is
u ^ ^ 365.9 - 212 ^ or 18.6 per cent

gi 365.9 + 460
The ideal work obtained from the steam in the engine is given by Equation (10),
(u).),(flow) = 1195.6 - 1019.5 = 176.1 Btu per lb.
2545
Ideal water rate of engine = r z r r = 14.6 lb per hp-hr
Thus, if the design of the engine were such as to result in an engine efficiency of 60 per
cent the over-all thermodynamic efficiency would be 10.4 per cent, and the actual
water rate of the engine 24.2 lb per hp-hr.
COMPBESSIOI^ 0 ¥ GASES ;
Gases are compressed for transportation, storage, and liquefaction, or
to maintain either high or subatmospheric pressures in chemical and
metallurgical processes. Although compression is accompanied by an
increase in temperature, this temperature rise is rarely the object of the
operation, except in special applications such as the ignition of fuel in the
Diesel engine or in vapor compression evaporation. Where a high
temperature of the compressed gas is desired, the temperature rise in
compression is useful but must be considered as incidental to the opera-
tion, since a temperature increase is produced more economically through
direct heating than by the dissipation of valuable shaft work.
Mechanical gas compressors are of three general types:
1. Reciprocating compressors.
2. Positive-displacement rotary blowers. /
3. Turbocompressors and fans. - /
The first two types operate with a sequence of intermittent operations
including a nonflow process. Thus, a charge of gas is taken in at the
suction pressure pi and compressed to the final pressure p2 along be on
the pV diagram "of Fig. 127, if the operation is adiabatic. This step is
nonflow in character but is immediately followed by the discharge of the
compressed gas along line ed. Since all the compressed gas flows into
the apparatus at the suction conditions and out of the apparatus at the
discharge conditions, the net effect of the over-all operation is a flow
process even though an intermediate nonflow step is involved. The
over-all work requirements and changes in thermodynamic properties
CHAP. XIII] TURBOCOMPRESSORS AND FANS 569
Pj d •lAe Isothermal e
V—Isentropic Intercoolins / ,
P TI
\ ,
IsentropiC
\ *<;*^rntercooling
\ f (2) ,.
" ^s^Îsentropic
K -^Z ' l*-Isentropie
H
yj ^ I >. K-lsentropic
I 'p'j^ \y^ rÎsothermal
LIL X
y m g, n
(a) (6)
'"'~ •• FIG. 127. Compression of Gases.
accordingly follow the equations developed for expansion or compression
under flow conditions.
Turbocompressors and Fans. Equipment of this general type oper-
ates essentially in a continuous manner. Although within the machine
high velocities and kinetic energies may be developed, the design is
generally such that these intermediate kinetic energies are largely trans-
formed to flow energy at the suction and discharge. Accordingly, from
an over-all standpoint, kinetic energies are generally negligible as com-
pared with the work of compression if the ratio of discharge to suction'
pressure is greater than 1.1. Under such conditions the theoretical
work requirements may be calculated from Equation (22) or (45) if con-
ditions of reversibility are approached.
Where small pressure changes are involved, changes in the kinetic
energies of the suction and discharge streams may constitute an appreci-
able portion of the work requirement. These effects are taken into
account in isentropic compression by combination of Equations (5)
and (41), with changes in potential and surface energies being neglected.
m{ul — ul)
-(w.)s = jT Ydv^
2gc (68)
miul — Ml)
= {H2 - H,)s +
Similarly, for isothermal compression, by combining Equations (5) and

(19), m(u^ — Ml)
-{w;)r = {m - HOT - T{S2 - S,) + (69)
2^^
Thus, the work input is increased by the difference between the kinetic
energies of the discharge and suction streams.
The kinetic-energy terms in Equations (68) and (69) may be omitted if
the effective pressures in the suction and discharge streams are both
taken as the sums of the respective static pressures and impact pressures
of the streams. The impact pressure is equal to mv^llg^ which corre- '
spends to the complete conversion of kinetic to -pY flow energy without
performance of work. Combined static and impact pressures are
measured directly by an impact-pressure gauge.
In Fig. 127 are diagrams showing the thermodynamic changes accom-
panying various types of compression between fixed suction and dis-
charge pressures. The ideal>work of compression is represented by the
area to the left of the compression curve on the pV diagram. Thus, for
isentropic compression the ideal work corresponds to area abed. For
isothermal compression the corresponding work requirement is repre-
sented by area abed. It is evident that isothermal operation is highly
advantageous as regards power consumption. In general, it is attempted
to design compressors to approach ideal isothermal operation at the low-
est feasible temperature. In this way power requirements are reduced ' .
by expenditures for cooling. The optimum design is that corresponding •-;-''
to minimum over-all costs of compression and coohng.
Single-stage Compressors. Large-size compressors and fans ordi-
narily approximate adiabatic operation and would behave in accordance
with Equation (68) if the operation were reversible. However, irreversi-
bility results from fluid friction and turbulence as well as from mechanical
friction in the equipment, both of which increase the work requirement.
Fluid friction also increases the final temperature of the fluid above the
value calculated from Equation (68). The departure of the operation
from ideal reversibility is expressed by an isentropic compression efficiency
factor tjs which is defined as the percentage ratio of the ideal work
requirement, calculated from Equation (68) for the pressure increase
obtained, to the total shaft work input to the compressor. Thus, . ,, .
[Mi — ni)s -\ ~ ,.„ . - (

i vs = ,, , ,, ^' X 100 - (70)
ws (total)
where H^ is the enthalpy of the discharged gas at the calculated tempera-
ture resulting from isentropic compression. Inasmuch as work of com-
pression is a minimum in a reversible isothermal operation, another index
of performance, termed the isothermal compression efficiency, TIT is also
used. '/
work of reversible isothermal compression
VT — total shaft-work input to the compressor
m(u2 — ul)
((?2 ~ Gi)r -\-
-^ - ;. : : (71)
ws (total)
CHAP. XIII] RECIPROCATING COMPRESSORS 571
The over-all efficiency includes compression efficiency and the effec-

tiveness of the cooling system in approaching isothermal operation.
Compression efficiencies of commercial machines range from 40 to 65
per cent for fans and from 65 to 85 per cent for turbocompressors. The
power required for a specified compression duty is determined by calcu-
lating the work per unit quantity of fluid from Equation (68) and multi-
plying by the rate of flow per unit time. If the work is in foot-pounds
per pound-mole and the flow in poimd-moles per minute, the result is
divided by 33,000 to obtain the required horsepower.
Reciprocating Compressors. As previously pointed out, an ideal
reciprocating compressor would operate on the cycle abed of the pV
diagram of Fig. 127 if the compression were isentropic. In such a case,
the work requirement would be expressed by Equation (68). Actual
compressors of industrial sizes closely approximate adiabatic operation
but deviate from reversibility as a result of fluid and mechanical fric-
tion. A further compUcation is introduced by the necessity of providing
a certain clearance volume between the cylinder head and the piston at its
extreme position. This clearance volume is represented by dj in Fig. 127.
At the end of the stroke the clearance volume is filled with compressed
gas which expands to k, reducing the effective suction volume to kb.
The volumetric efficiency -QV is defined as the percentage ratio of the
volume of gas, measured at pi, actually taken into the cylinder during the
suction stroke to the volumetric displacement of the piston. The volu-
metric efficiency determines the capacity of a machine for a specified
duty while the compression efficiency determines the total work input
required per imit quantity of gas compressed. The total power required
for a machine of given displacement operating between specified pressiu-es
is increased by increased volumetric efficiency and reduced by increased
compression efficiency.
Compression efficiencies are reduced by mechanical friction and by the
pressure drops through the suction and discharge valves. As a result of
these pressure drops the actual suction pressure is somewhat below pi
and the discharge pressure above pt, as indicated by the dotted lines
between ah and je on Fig. 127. The increased work resulting from the
throttling through valves represents a major loss, and proper design of
valve openings is of great importance. Compression efficiencies are
increased by any cooling of the gas during compression which will cause
the operation to approach isothermal conditions, as indicated in the pV
diagram of Fig. 127. Cylinders of small compressors are frequently
water-cooled, but water jackets are not highly effective on large machines
because of the small ratio of surface to volume. In some designs cooling
water is injected into the gas during compression. :^,, ,
Volumetric efficiency is reduced by pressure drop through the suction

valves and by increased clearance volume. The problems of compressor
design have been discussed in detail by York* who concludes that the
gas retained in the clearance volume of large machines approximates
isentropic expansion as the piston starts on the suction stroke. In this
case the ideal work of compression per unit quantity of gas discharged is
imaffected by the clearance volume which may be looked upon as con-
taining an isolated quantity of gas alternately compressed and expanded
in a reversible manner without any net consumption of power. Reduc-
tion in volumetric efficiency due to clearance can be calculated if isen-
tropic expansion of the clearance volume on the intake stroke is assumed,
but further reduction due to pressure drop through valves and parts can
be determined only experimentally.
Variation of the clearance volume offers an effective means of control-
ling the capacity of a compressor driven at a constant speed. Machines
are constructed with clearance " pockets " of variable volume which
permit operating the machine at reduced capacities without waste of
power. An increase of the clearance volume greatly reduces the volu-
metric efficiency with relatively little effect on the compression efficiency.
Volumetric efficiencies of industrial compressors range from 50 to
90 per cent, whereas compression efficiencies, based on the work require-
ment of isentropic compression, vary from 75 to 90 per cent.
Illustration 7. A single-stage compressor is to compress 800 cu ft per min of
ammonia gas at 0°F and an absolute pressure of 15.0 lb per sq in. to 75 lb per sq in.
Calculate the actual horsepower required to drive the compressor and the piston dis-
placement in cubic feet per minute, assuming a compression efficiency of 76 per cent
and a volumetric efficiency of 85 per cent.
From Fig. 119 and Equation (10), for isentropic compression at the special con-
ditions,
- {w.)s = (H2 - Hi)s = 715 - 617 = 98 Btu per lb
The ideal discharge temperature is 200°F.
At 0°P, 15 lb per sq in., the specific volume of ammonia gas from Fig. 116 is 18.92
cu ft per lb.
Compression rate = 800/18.92 = 42.3 lb per min
(42.3) (98) (778) , ,
Theoretical horsepower = = 97.8 hp
(00,000)
97.8
Actual horsepower = —zr = 129 hp .-
800
Piston displacement = -—- = 940 cu ft per mm
0.80
Multistage Compression. Because of mechanical difficulties associ-
ated with high ratios of discharge to suction pressure a single-stage recip-
* R. York, Jr.,/nd. £7jiff. CAew., 34, 355 (1942). '
CHAP. XIIIJ MULTISTAGE COMPRESSION 673
locating compressor is ordinarily not designed for a compression ratio

greater than 4 or 5 to 1. Turbocompressors are limited to much lower
ratios per stage. In order to develop high overall compression ratios it is
necessary to use multistage compression in which the one stage discharges
to the suction of the next.
An important advantage of multistage operation is that it permits
intercooling of the gas between stages and thus approaches the same
suction temperature for each stage. This results in a reduction of the
work of compression and approaches isothermal conditions where
several stages of small compression ratios are employed. A two-stage
operation with intercooling is indicated in Fig. 127. Isentropic com-
pression in the first stage along bf is followed by .cooling at constant
pressure p' along/^r. The final pressure p^ is reached by isentropic com-
pression in the second stage along gh with a final temperature intermedi-
ate between that of e resulting from single-stage isentropic compression
and the isothermal value of c.
It is evident that the work required for the two-stage operation is less
than that of the single stage by the area, fghe. If more stages with inter-
cooling were employed, the work would be still further reduced,
approaching the ideal isothermal case. For multistage compression, the
equivalent compression efficiency is defined as the ratio of the work for
multistage isentropic compression with complete intercooling between
stages divided by the actual shaft-work input.
The total work in a multistage compression is the sum of the work in
the individual stages. Thus, from Equation (44), if a constant value of K
and a constant suction temperature Ti are assumed and kinetic energy
change is neglected, the total work per mole for two-stage compression
of an ideal gas becomes
If the intermediate pressure is p', the work required will be a minimum

value when dwt/dp' is zero. The optimum intermediate pressure is
determined by equating dw*/dp' from Equation (72) to zero. Thus,
2(«-l) t-1
, p' ' = ipiPi) '
p' = V^2 (73)*
Rearrangement of (73), gives
^' = ^ = JS^ • . ' " -^ (74)*

: Pi V \Pi ; h,. :
Thus, for best operation, each stage should have the same compression
ratio, equal to the square root of the over-all compression ratio. Where
s stages are employed it may be demonstrated similarly that optimum
results are obtained when the compression ratio in each stage is equal to
If (75) is combined with an equation of the form of (72), it follows that .,*,
the work per stage is constant, and the over-all work per mole is given by ;,<3
-^ = 0^[©''-'] .•:: ™'

By means of Equation (76) the number of stages for optimum economy
of operation may be estimated by consideration of the increased- equip-
ment and maintenance costs which accompany an increased number of
stages with a resultant reduction in power requirement.
For compression of nonideal gases the problem of calculating optimum
intermediate pressures involves trial-and-error methods. A simplified
method was proposed by York* which may be used where economy is of
great importance. However, for many cases the over-all work require-
ment is not greatly affected by small variations of the intermediate
pressures from the optimum values, and Equation (75) may be used for
establishing the intermediate pressures. Once these pressures are
estabhshed, the work and volumetric capacity for each stage are calcu-
lated from Equation (45) and the p 7 data. Equation (75) may lead to
serious departure from optimum pressure conditions when vapor mixtures
are dealt with if large amoimts of condensate are separated in the inter-
coolers. Where this condition is encountered, the intermediate pressures
may be adjusted to keep the work in different stages approximately /
equal. If only two stages are involved, a series of intermediate pressures •'
may be assumed and the corresponding over-all work requirements calcu-
lated; the optimum intermediate pressure corresponds to the minimum
of a curve relating work to intermediate pressure.
In the isentropic compression of gases of low molecular weight and low
molal heat capacity there is no tendency for condensation in the com-
CHAP. XIII] REFRIGERATION 675
pressor. This is evident from the TS diagram for ammonia, Fig. 118.
Adiabatic compression of the saturated vapor results in superheating.
However, gases of high molal heat capacity behave quite differently, as
may be seen from the TS diagram for benzene in Fig. 121. If satu-
rated benzene vapor at 100°F is isentropically compressed, it is first
superheated, and then at a pressure of 655 lb per sq in. it becomes satu-
rated, and condensation begins. Formation of condensate in compres-
sors is frequently hazardous, and great care must be exercised in designing
compression systems for vapors of this type. Formation of condensate
in the intercoolers of multistage compressors is common when vapor
mixtures are being compressed, and faciUties for its separation from the
uncondensed gas passing to the next stage are readily provided.
REFRIGERATION
The purpose of refrigeration is to produce a region of temperature
below that of the atmosphere. The over-all result is the extraction of
heat from the low-temperature region and its rejection at the higher
temperature of the surroundings or of the available coohng medium.
Thus, in effect a refrigeration machine is a heat pump performing the
reverse function of an engine. Refrigeration is of constantly increasing
importance in the preservation of foods, in air conditioning, and in
chemical engineering operations where low temperatures are required for
the control of reactions, the condensation of vapors, or the crystalUzation
of solids.
Two general types of refrigeration cycles are in common use. Both
depend upon the attainment of a cold region by the vaporization of a
refrigerant fluid from an evaporator at a low temperature and pressure.
In compression refrigeration, shown diagrammatically in Fig. 129,
mechanical energy is used to compress the vapor leaving the evaporator
so that the heat absorbed at the low temperature of the evaporator may
be rejected at the temperature level of the condenser. A refrigerant
fluid is employed to absorb heat at low pressure and temperature and is
then compressed to a higher pressure and temperature where it gives off
heat, usually with condensation. Expansion to the lower pressure and
temperature completes the cycle. This latter is usually free expansion;
isentropic expansion has rarely proved feasible.
In absorption refrigeration, shown diagrammatically in Fig. 131, heat
which is absorbed at a low temperature and pressure is rejected at an
intermediate temperature and high pressure after its temperature level
has been increased by addition of heat from a high-temperature source
such as the combustion of a gas. In this scheme no mechanical energy is
required except in some systems for pumping the liquid absorbent, and
the refrigerant is compressed from the low pressure at which it absorbs
heat to the high pressure at which it rejects heat through intermediate
absorption or adsorption at a low-pressure level. The refrigerant is
released from the absorbent or adsorbent at a high pressure through the
addition of heat from the high-temperature source. This method has the
advantage of minimizing or eliminating moving mechanical parts and
may effectively utilize low-pressure exhaust steam or fuel gas as the
source of heat.
Compression Refrigeration. A Carnot engine operating in reverse
represents an ideal reversible refrigeration cycle. The TS diagram of
Condenser P-i
n a At id
/W\AWW\A^
T2
/ b
-I1
t
\A 1 Pi Storage
Ti
Expansion
Compressor
- W !•
/ i '^ i \ Evaporator ^2
MA/WWW^
' 1 1 \
SB SC
T
?2
FIG. 128. Eeverse Carnot Cycle. FIG. 129. Vapor Compression Refrigeration.
such a cycle operating with saturated vapor is shown in Fig. 128. A

quantity of heat ga is absorbed in increasing the quahty of the vapor at
temperature T2 along be. Isentropic compression along cd raises the
temperature to Ti where a quantity of heat —qi is rejected through con-
densation of the vapor along da. The liquid is isentropically expanded
and partially vaporized along ab to temperature T2 thus completing the
cycle.
As was developed in consideration of the Carnot engine, page 557, the
net work done by the system is the difference between the heats absorbed
at the higher and lower-temperature levels, and Equations (58) and (59)
are directly applicable to the reverse cycle operating as a heat pump.
However, in this case both Ws and Sh — Sc are negative. It follows that / /
the amount of heat rejected at the high temperature is greater than that
absorbed at the low temperature by the amount of the work done on the
system.
For comparing the performance of refrigeration machines a coefficient
of performance is used as an index. This is defined as the ratio of the
CHAP. XIII] COMPRESSION REFRIGERATION 577
heat absorbed in the evaporator to the work done on the fluid. Thus,
from Equation (59) the coefficient of performance
TiAS2
(77)
where AS2 = <Sic — St- According to the Carnot principle, the coefficient
of performance of any reversible refrigeration machine operating between
temperatures Ti and Tj is expressed by Equation (77), regardless of the
nature of the refrigerant or the steps of the cycle.
FIG. 130. Refrigeration Cycle.
The capacity of a refrigeration machine is ordinarily expressed in tons

of ice-making capacity per 24 hr. One ton of refrigeration capacity is
capable of absorbing in 24 hr the heat evolved in freezing 1 ton of water at
32°F. On this basis, 1 ton refrigeration capacity = 200 Btu per min =
42.4 hp. To define completely the characteristics of a refrigeration
operation, it is necessary to specify the coefficient of performance, the
refrigeration capacity, and the two temperature levels of heat absorption
and rejection.
Actual compression refrigeration cycles differ from the ideal Carnot
heat pump in that each step is carried out in a separate piece of equip-
ment with the refrigerant fluid circulating from one to another as indi-
cated in the flow diagram of Fig. 129. The corresponding TS and HS
diagrams are shown in Fig. 130. The liquid refrigerant leaves the stor-
age tank at temperature Ti and pressure pi and is expanded through a
valve along ab. This expansion is substantially isenthalpic free expan-
578 EXPANSION AND COMPRESSION " [CHAP. XIII
sion and is hence highly irreversible and accompanied by a large increase

in entropy, Sb — Sa. Partial vaporization and reduction in temperature
to T2 occurs in this expansion. Heat is absorbed at temperature T2
along be until vaporization is completed. The saturated vapors are
then isentropically compressed to pi along cd. The superheated vapor
at d is cooled at constant pressure, becoming saturated at e and con-
densing along ea.
This cycle differs from the ideal Carnot cycle in the irreversible nature
of the expansion and the fact that saturated vapor leaving the evaporator
is compressed to a superheated state. Further deviations result from
the inefficiency of actual compressors, as previously discussed.
By neglect of the changes in kinetic, potential, and surface energies, the
energy balances for the individual steps may be written as follows; with
the'nomenclature of Fig. 130: ^
Expansion ; Ha = Ht (78)
Evaporation •S , Hb + q^ = He (79)
Compression Ho — w, = Hd (80)
Since qi in going from d to a is negative,
Cooling and condensation Hd = Ha. — qi (81)
For the complete cycle,
- w , = - ( ? ! + ^2) =^Hd-Ho (82)
The ideal coefficient of performance then becomes, from .(78), (79), and
(82),
— =f ^ " (83)
— Wa Hd — He
where —Ws represents the net work done on the fluid and not the total
work input to the compressor.
The coefficient of performance may be visualized from the TS diagram
of Fig. 130 in which the double-crosshatched area represents q-i. and the
single-crosshatched area represents — w,. The coefficient of performance
is the ratio of these two areas. It is evident that the coefficient of
performance would be greatly improved if the free expansion along ah
were replaced with isentropic expansion in an engine, the work from
which would be transmitted to the compressor. This scheme is rarely
used because of mechanical complications.
In developing Equation (83) it was assumed that the compression step
is isentropic; actually some heat is generally absorbed in the piping
CHAP. X I I I ] COMPRESSION REFRIGERATION 679
leading to the compressor, and during compression heat is at first gained

by conduction through the walls of the compressor. As the temperature
is increased in the compression stroke, the temperature difference between
the gas and its surroundings is reduced. As a result the compression
foUows some curved path such as the broken line cjd'. The final tem-
perature reached may be either higher or lower than that resulting from
isentropic compression. Similarly, the work of compression may be
either increased or decreased by such deviations from adiabatic con-
ditions.
The coefficient of performance of a vapor compression cycle may be
improved by two-stage compression with intercooling. However, the
selection of the optimum intermediate pressure is different from that
employed in ordinary compression in which the suction temperatures of
all stages are the same.
In order that intercooling may be carried out with the same higher-
temperature cooling medium that is used to maintain temperature Ti,
it is necessary that compression be carried out adiabatically to a tem-
perature greater than Ti. The gas is then cooled to approximately Ti,
and compression is continued in another stage. A two-stage operation
of this type is indicated by lines cfghe of Fig. 130. In this way work is
saved corresponding to the area fghd on the TS diagram. A modification
of this scheme is to cool during the second stage of compression by inter-
jecting liquid refrigerant from the reservoir.
Refrigeration cycles are also operated with a gas such as air as the
refrigerant without condensation at any stage. A gas refrigeration cycle
comprises isentropic compression, isobaric cooling, isentropic expansion,
and isobaric absorption of heat corresponding to the cycle of Fig. 127b.
The power generated in the isentropic expansion is utilized to aid in
driving the compressor. Despite its approach to Camot efficiency, this
type of refrigeration has been outmoded because of its mechanical com-
plexity and the relative difficulty of heat transfer to a gas.
Many different fluids are used in vapor-compression cycles.
Anunonia, sulfur dioxide, carbon dioxide, methyl chloride, and the chloro-
fluoroparaffins, the so-called Freons, are the most popular. Selection of
the refrigerant is governed by the working temperatures, the limitation
regarding the bulk of the compressor, the hazards of very high pressures,
and many other considerations such as toxicity, inflammability, and
corrosiveness.
Illustration 8. A compression refrigeration plant using propane is to be designed

for a capacity of 50 tons of refrigeration per day at an evaporator temperature of
30°F. Cooling water is available at 90°F, and the condenser is to be designed for a
condensation temperature of 100°F. It is assumed that the compression is adiabatic
580 . EXPANSION AND COMPRESSION [CHAP. XIII
with a compression efficiency for the compressor of 85 per cent and a volumetric
efficiency of 75 per cent. , , ' ^
Calculate
(a) The pounds of propane circulated per hour.
(6) The heat removed in the condenser, Btu per hour.
(c) The power required to drive the compressor.
(d) Volumetric displacement of compressor, cubic feet per minut«.
(e) The coefficient of performance. .
In the absence of tables or charts on the thermodynamic properties of the refriger-
ant, this problem can be solved by the following generalized procedure.
Basis of calculations: 1 lb-mole of propane circulated
Let Ha = enthalpy of liquid propane at 100°F •
Hb = enthalpy of liquid propane at 30°F
He = enthalpy of saturated propane vapor at 30°F
Hd = enthalpy of propane vapor leaving compressor »
He = enthalpy of saturated propane vapor at 100°F
Corresponding subscripts are used for entropies. For the reference state, use
liquid propane at 30°F; Sb = 0; Hb = 0. Heat capacity of propane vapor, low-
temperature range (page 336):
4 = 7.95 + 0.00397ri-25 Btu/ab-moIe)(°R)

Pc = 43.0atm; Tc = 666°R
Conditions in . Condenser / Evaporator

From Table Ha p = 13.69 atm / 4.83 atm
T = 660°R ' . 490°R
Pr = 0.318 0.113
r . == 0.840 0.735
From Fig. 103: .., 2 = 0.78 0.91
From Fig. 107: -S* - S = 1.75 Btu/(lb-mole) (°R) 1.10
From Fig. 106: {H* - H) jTe = 1.8 0.88
From Table XIII: X = 6040 Btu per lb-mole 7250
For isentropic compression from Equation (39):
Si-Se = iS*a - St) - {S*a ~ S,) + (S* ~ Sc) (a)
S*-St= r - r ! : 9 5 +0.00397(^)^^1 ^ ^ ^ ^ ^ 13.69

T J 4.83
Si — Sd depends upon the temperature Td leaving condenser S* — Sc = 1.10
By trial-and-error solution of Equation (a), Td = 566B

Hence, Trd = 0.85 prd = 0.318
St-Sd = 1.62 {Ht ~ Hd) ITc = 1.7
From Equation (45),
' (7.95 -I- 0.00397rii») dT - 666(1.7 - O.i
= 822 Btu per lb-mole ' ' '*:-'' i

CHAP. XIII] ABSORPTION REFRIGERATION 681
Since He = 7250, H^ = 822 + 7250 = 8072 Btu per lb-mole
H.-H. = i H t - m ^ - T . \ ^ - ^ ^ - ^ J ^ ^ ]
•»560
(7.95 + 0.00397ri-2') dT - 666(1.8 - 1.7) = - 1 8 0
i/see
/666
He = 8072 - 180 = 7892
Ha = 7892 - 6040 = 1852
Heat absorbed in evaporator = (50) (60) (200) = 600,000 Btu per hour
600,000
(a) Propane circulation rate = —— —— =111 lb-moles per hr or 4890 lb
i ZsM — l o O ^
per hr
(b) Heat transfer duty of condenser = (111) (8072 - 1852) = 691,000 Btu per hr
Ideal work of compression = 111(822) = 91,300 Btu per hr
91 300
(c) Power to drive compressor = ,^^,'^ . , , = 42.2 H.P. at 85 per cent com-
(0.85) (2545)
pression efficiency
490 1
Volumetric intake = (111)(359) -—— —— (0.91) = 7650 cu ft per hr
(492) 4.83
(d) Piston displacement of compressor at 75 per cent volumetric efficiency =

7650
170 cu ft per min
(60) (0.75)
(e) Coefficient of performance = '•— = 5.58

^ (91,300)
An ammonia compression refrigeration system operating between the same
temperature levels and with the same compression efficiency requires 46.8 hp and
gives a coefficient of performance of 5.04. This compression indicates that use of
propane as the fluid results in shghtly improved performance as compared to
ammonia, although the difference indicated is almost within the possible error of the
geherahzed correlations. Propane has the advantage of a relatively high heat
capacity which results in little superheating in the compression step.
Absorption Refrigeration. The most widely applied absorption

refrigeration cycle uses ammonia as the refrigerant and water as the
absorbent. Aflowdiagramof such a process is shown in Fig. 131. This
flow diagram indicates a type of operation which may be used for large-
scale process refrigeration. Many other modifications of the basic
ammonia absorption cycle are employed to meet particular specific
situations. Liquid ammonia from the accumulator is expanded freely
through an expansion valve, where it is partly vaporized andthen passes
to the evaporator. The ammonia vapors leave the evaporator to an
absorber where they are absorbed by weak aminonia liquor. In order to
reduce the amount of weak liquor circulated per unit mass of ammonia
absorbed, it is necessary to cool the absorbent close to the temperature
of the available cooHng water and to provide intermediate cooling surface
in the absorber.
The strong ammonia liquor on its way from the absorber to the stripper
is passed through a heat exchanger where it is heated coimtercurrently
by the hot weak liquor on its way from the stripper to the absorber.
The weak liquor after cooling in the heat exchanger enters the top of the
absorber. The strong liquor is fed to the stripping column where heat is
added by means of a reboiler to evaporate the ammonia. The vapors
Condenser
waterr#2
Cooline
Accumulate
ti
Evaporator,, -
^Steam q^
Reboiler - :- -
Heat Ezchanser
FIG. 131. Ammonia Absorption Refrigeration.
from the top of the stripping column are condensed, and a portion of the
liquid is returned as reflux to the top of the stripping column which con-
tains bubble decks or packing to effect fractional distillation. It is
evident that the stripper must be operated at a pressure high enough to
produce condensation of the ammonia at the temperature attainable with
the available cooling water.
Absorption refrigeration cycles are also in common use in which a
simple still, with or without a partial condenser or dephlegmator,
replaces the stripping column. In other modifications even the pump
used for pumping liquid to the still is eliminated. This is accomplished
by mixing with the refrigerant an inert gas of low density sUch as hydro-
gen which permits the entire system to be kept at a constant total
pressure. However, different partial pressures of the refrigerant are
CHAP. X I I I ] ABSORPTION REFRIGERATION 583
maintained for condensation and for evaporation. Circulation is main-

tained by thermal convection induced by the still, and absorption pro-
ceeds by the partial-pressure gradient of the refrigerant from vapor to
liquid phase. Descriptions of such units which are used largely in house-
hold refrigeration are given by Keenan.'
The problems of design and analysis of absorption refrigeration
systems involve a series of energy balances embracing the entire cycle
and each of its parts. Thus, for the over-all process, with reference to
Fig. 131, if changes in kinetic, potential, and surface energies are
neglected,
q2 + q, + qi + qa- w, = 0 (84)
where 52 = heat absorbed in evaporation I

—gi = heat removed in the condenser
qa = heat added to system in the stripper reboiler
—qa = heat removed in the absorber
—Wg = work done upon the system by the pump
Energy balances for individual steps involve only enthalpy and heat
terms except where the pump is included. Thus, for example, for the
stripper,
msHs + maHa + q, = msHs + mJIs (85)
where wis is the mass of the stream 5 in Fig. 131, and Hi is the corre-
sponding enthalpy per unit mass. The other symbols have similar
significance.
Since mixtures of varying concentrations are involved throughout the
cycle, energy balances are most readily evaluated by means of an
enthalpy-concentration chart, such as described in Chapter VIII.
Fig. 132 is such a chart for the enthalpies of the ammonia-water system
at various pressures and temperatures and for all compositions and
phases. Enthalpies are based upon 1 lb of the fluid and are referred to
Uquid ammonia at — 40°F and to liquid water at 32°F, each at its respec-
tive vapor pressure at these temperatures. This chart is a modification
of one prepared by Bosnjakovic* by the methods described in Chap-
ters VII and XIV.
The lowest group of lines on the chart represents the enthalpies of
various sohd phases of ice and ammonia with the indicated eutectic and
congruent points. The next lowest group of curves represents the
enthalpies of saturated solutions over the entire range of composition for
various temperatures and pressure. The third group from the bottom is
' " Technische Thermodynamik II," Theo. Steinkopf, Dresden and Leipzig (1935).
EXPANSION AND COMPRESSION [CHAP. XIII
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Concentration—Weight Fraction NH3
FIG. 132. Enthalpy Concentration of the Ammonia - Water System.
(Reference: Water at 32°F, liquid ammonia at — 40°F).
CHAP. XIII] ABSORPTION REFRIGERATION 585
construction lines for obtaining vapor-liquid equilibrium relations The

next higher group of Unes represents the enthalpies of saturated vapors
covering the entire range of composition at various temperatures and
pressures.
Conditions of vapor-Hquid equilibria may be established by selection
of points of equal temperature and pressure in the saturated liquid and
vapor curves. Thus, point b represents saturated liquid containing
16 per cent NH3 at 300°F and 200 lb per sq in. By following a vertical
line from b to c and a horizontal line from ctod the composition of the
vapor in equilibrium with the liquid is found to be 60 per cent ammonia,
on the corresponding pressure line. The enthalpies required in dealing
with problems in absorption refrigeration may be read directly from
Fig. 132.
Illustration 9. An ammonia-absorption refrigeration plant of the type shown in
Fig. 131 is to be designed for the same duty as the propane-compression plant of Illus-
tration 8. The capacity is to be 50 tons per day with an evaporator temperature of
30°F and cooling water such that condensation and cooling can be carried out at a
temperature of 100°F. In order that the evaporator pressure may be high and that
no difficulty due to formation of ice in the low-temperature sections of the plant will
ensue, it is desired that the ammonia from the stripper shall contain only 0.5 per cent
by weight of water. It may be assumed that the absorber produces contacting of
such effectiveness that the ammonia concentration of the strong liquor is 95 per cent
of the equiUbrium value. It also may be assumed that vaporization is completed in
the evaporator under conditions such that the total vapor formed is at all times in
equihbrium with the residual liquid.
The design study is to be based on the following arbitrary specifications:
(1) Temperature of weak liquor leaving the heat exchanger, 120°F.
(2) Reflux ratio of stripper [weight of stream (9) /weight of stream (1)] = 0.50.
(3) Ammonia content of weak liquor, 20 per cent by weight.
(4) Temperature of strong Uquor leaving absorber, 100°F.
These specifications might be varied over wide ranges with resulting changes in the
design of the plant for the given duty. These changes would affect both the capital
and direct operating costs. A series of design studies in which such specifications are
varied is necessary in order to arrive at an optimum design which results in minimum
over-all operating costs.
In this design study it is permissible to neglect:
(1) Heat exchange between the equipment and its surroundings.
(2) Variations in pressure in both the high- and low-pressure sections of the plant.
(3) Variation in temperature of the ammonia-water mixture in the evaporator.
The evaporator pressure may be conservatively established as corresponding to com-
plete vaporization at the specified temperature.
On this basis calculate a complete material and energy balance of the plant together
with the temperatures and pressure at all points. Also calculate:
(a) The teat-transfer duties of the condenser, absorber cooler, heat exchanger, and
reboiler, in Btu per hour.
(6) The hydraulic horsepower of the pump.
(c) The pressure and quantity of dry saturated steam required for heating the
reboiler with a 10°F temperature difference.
586 EXPANSION AND COMPRESSION [CHAP. XlII
Solution: The material and energy balances are established simultaneously from
the data of Fig 132, starting with the evaporator. These calculations are sum-
marized in Table A where the streams are identified by the numbers of Fig 131.
The results in this table are reported in more significant figures than are justified by
the accuracy of the data in order that the numerical accuracy of the calculations may
be readily verified by over-aU balances.
(1) The pressure of stream (1) is read from Fig. 132 as 210 lb per sq in.; the equilib-
rium pressure of liquid containing 99.5 per cent NHa at 100°F.
(2) The pressure of stream (2) is the equilibrium pressure of saturated vapor con-
taining 99.5 per cent NH3 at 30°F, which is read from Fig. 132 by a trial-and-error
procedure. A trial value of the equilibrium pressure is assumed, and a horizontal
Mne is followed from the point representing saturated vapor at this pressure and
99.5 per cent NH3 over to the corresponding construction line. A vertical line is
followed downward from this point to the 30°F hne of the saturated Uquid curves, and
the corresponding equilibrium pressure is read. This second approximation pressure
is used to repeat the operation until consistent pressures are obtained. In this man-
ner it is found that the evaporator pressure is approximately 45 lb per sq in. and that
the liquid in equilibrium with the vapor leaving the evaporator contains 77 per cent
NH,.
TABLE A
MATERIAL AND ENERGY BALANCES OF
AMMONIA ABSORPTION REFRIGERATION PLANT >
NH3
• • /
Per Cent Abs Pressure, H m

Stream byWgt <°F lb per sq in. Btu per lb U) per hr Btu per hr
1 99.5 100 210 152 1229.5 186,880
2 99.5 30 45 640 1229.5 786,880
3 40.0 100 45 -5 4887.3 -24,440
4 40.0 100+ 210 -4.4 4887.3 -21,660
5 40.0 217 210 128.8 4887.3 629,432
6 20.0 290 210 220 3657.8 804,720 •
7 20.0 120 45 42 3657.8 153,628
8 99.5 110 210 650 1844.3 1,198,800
9 99.5 100 210 152 614.8 93,450
The rate of stream (2) is fixed by the heat-transfer duty of the evaporator and the
differences in specific enthalpies of streams (1) and (2). Thus,
I
nil = (50)(200)(60)/(640 - 152) = 1229.5 lb per hr ^" ''" /
(3) The ammonia content of the strong liquor is 95 per cent of the equilibrium
concentration at 100°F and 45 lb per sq in., which is read from Fig 131. Thus,
NH3 in stream (3) = 0.95 X 42 = 40 per cent. It may be assumed that the enthalpy
of a liquid which is not completely saturated with NH3 is the same as that of a satu-
rated solution of the same NH3 content and at the same temperature. On this basis
H, = - 5 Btu per lb
The rate of stream (3) is determined by the over-all material balance and the
ammonia balance of the absorber. Since from the over-all material balance m? =
m, - 1229.5,
O.4OOT3 = (ms - 1229.5)0.20 -|- (1229.5) (0.995)
ms = 4,887.3 rm = 3,657.8 i
CHAP. X I I I ] ABSORPTION REFRIGERATION 587
(4) The work done by the pump is calculated from. Equfl.tion (61) The data of
the International Critical Tables indicate a density of approximately 0.86 for a
solution at these conditions.
( 2 1 0 - 4 5 ) (144) (4887.3)
-"" = ^ (eSik^) - = 2,160.000 ft-lb per hr
= 2,780 Btu per hr
2 160 OflO
Power input to fhiid = -—^:—'- = 1.09 hydraulic hp
^ (60) (33,000) •^ ^
In order to determine the power required for the pump drive, the hydraulic horse-
power must be divided by the fractional efficiency of the pumps.
2780
Enthalpy increase of the fluid in the pump = 0.57 Btu per lb
4887.3
(5) The temperature of the feed to the stripper is determined by an energy balance
around the heat exchanger. The quantity of stream (7) is determined and its
enthalpy may be evaluated from the specified temperature of 120°F. The tempera-
ture of stream (6) is the equiUbrium temperature of a solution containing 20 per cent
NHs at a pressure of 210 lb per sq in., which is read from Fig. 132 as 290°F with an
enthalpy of 220 Btu per lb. The enthalpy of stream (5) is then calculated from the
specific enthalpies of Table A and the energy-balance equation niilli = nitHi +
mjlt — VlyHTOT ,1./., n
msffs = 804,720 - 21,660 - 153,628 = 629,432 Btu per hr

ffs = 629,432/4887.3 = 128.8 Btu per lb
This specific enthalpy corresponds to a temperature of approximately 217°F on

the saturated hquid Unes of Fig. 132 which is above the saturation temperature of
202°P for a 40 per cent solution at 210 lb per sq in. Thus, the feed to the stripper is
partially vaporized by the heat exchanger. The heat-transfer duty of the exchanger
is 804,720 - 153,628 = 651,092 Btu per hr.
(6) The heat input to the reboiler is determined by an over-all energy balance
around the stripping column. The rate of stream 9 is fixed by the specified reflux
ratio andTOS= ms + mi. The corresponding enthalpies are read from Fig. 132 as
listed in.Table A. The temperature of stream (8) is the temperature of the saturated
liquid which is in equilibrium with vapors containing 99.5 per cent NH3 at 210 lb per
sq in. From Fig. 132 it is seen that this dew-point liquid has a composition of
approximately 85 per cent NH3 and a temperature of 110°F. From Table A,
g, = mjla + nuMt - nnHi - m,Hi = 1,198,800 + 804,720 - 629,432 - 93,450
= 1,280,638 Btu per hr
The heat removed by the condenser is expressed by
-qc = mgffs - MiHi - tnJI, = 1,198,800 - 186,880 - 93,450 = 918,470 Btu per hr
The heat removed in the absorber is expressed by • *' '• • ' ^- '
-5a = m7H7+m2H,-m3ff3 = 153,628-I-786,880 -(-24,440) = 964,948 Btu per hr
These results may be substituted in Equation (84) as a verification of the calculation.
Thus,
600,000 + 1,280,638 - 918,470 - 964,948 + 2780 = 0
(7) The steam used for heating the reboiler must be at a temperature of 300°F
which corresponds to an absolute pressure of 67 lb per sq in. or a gauge pressure of 62
lb per sq in. If it is assumed that the condensate is withdrawn at 300°F, heat to the
reboiler per pound of steam is the heat of vaporization which from the steam table is
910 Btu per lb.
1,280,638 , , „ , „ ^ , .
Reboiler steam rate = • — — — = 1407 lb per hr.
If the results of Illustrations (8) and (9) are compared, it is seen that with the
particular specifications employed identical refrigeration duties are performed by a
44.3-hp mechanical drive in the one case and by 1,407 lb per hr of 52-lb steam in the
other. If the heating steam were used to operate the drive of the compression ma-
chine an engine with a steam rate of 34 lb. per hp-hr would be required. This corre-
sponds to an engine efficiency of 65 per cent if the exhaust is at atmospheric pressure.
The heat to coofing water in the absorption cycle is (1.96)10« Btu per hr as compared
to (0.696)10^ in the compression machine. This difference in cooUng-water require-
ment is greatly reduced if a steam drive with a condenser is used for the compressor.
It may be concluded that the relative desirability of compression and absorption
refrigeration cycles is largely dependent upon particular local conditions. The
absorption operation is favored by the availabihty of low-cost exhaust steam for
heating the reboiler. The choice of a refrigeration cycle thus involves consideration
of the entire power-steam balance of the plant. As previously mentioned, many
modifications of the absorption cycle are possible in order to adapt it to local con-
ditions.
Other binary systems have been used for absorption refrigeration, but
the ammonia-water is the most common. The chief difficulty in the
development of other binary systems is in the selection of an absorbent of
high absorption capacity and ease of separation, and one that is non-
toxic, noncorrosive, and not decomposed by prolonged use at the stripper
temperatures.
Refrigeration cycles based upon the adsorption of refrigerant vapors
upon solid adsorbents have been used to a limited extent and may
develop to greater importance. Such schemes operate on an inter-
mittent cycle, adsorbing vapors at low temperature and pressure in one
chamber, while another chamber is evolving previously adsorbed vapors
at a higher temperature and pressure. Adsorbents of high capacity
such as sihca gel, activated alumina, or charcoal may be used with a
variety of fluids. The disadvantages of these methods are the'inter-
mittent nature of the operation and the difficulty of transferring heat to
a solid material. The thermodynamic problems of adsorption refrigera-
tion may be handled by the same general principles as demonstrated
here for the absorption systems. It is desirable to develop an enthalpy-
composition'chart for the adsorbed system to facilitate calculations.
Such a chart is shown in Fig. 62, page 300, for water vapor on silica gel.
CHAP. XIII] LOW-TEMPERATURE LIQUEFACTION OF GASES 589
Low-Temperature Liquefaction of Gases. It is frequently desired to

liquefy gases such as air which are normally far above their critical tem-
peratures. The object may be to separate components of a gaseous mix-
ture by subsequent fractional distillation or to produce very low-tem-
perature refrigeration, below the practical working ranges of the
common refrigerants.
In order to obtain liquefaction of such gases, combined cooling and
compression must be produced. For example, air must be cooled to
— 140.7°C before it can be Uquefied even at high pressures. When
working with gases of very low critical temperatures, it may be necessary
Multistage compressor
Inlet gas and intercoolers
A 1
<:J Heat exchanger

®
Expansion valve
[ Liquid
FIG. 133. Low-Temperature Liquefaction of Gases.
to precool by means of conventional refrigeration. The lower tempera-
tures required for liquefaction are then produced by the self-cooling
cycle shown diagrammatically in Fig. 133.
The gas to be liquefied enters the system at temperature Ti and pres-
sure pi and is combined with recycled gas which drops the temperature
slightly to T2. The gas is then compressed to ps with an increase in
temperature. Multistage compression is generally employed with rejec-
tion of heat — gi in the intercoolers. The compressed gas is cooled at the
interstages with water or a cooling fluid circulated by a conventional
refrigeration system. The compressed gas is passed to a heat exchanger
where it is cooled by a return flow of the same gas after its expansion.
The cold compressed gas is then expanded to a lower pressure with a
corresponding drop in temperature. This expansion may be carried out
isentropically, and the power developed may be used in the compressor >.
drive or for other purposes. This scheme is termed the Claude process.
More commonly, however, the simple Linde process involving free
expansion through a throttling valve is used, even though the tempera-
ture drop is much less, and no work is recovered. The cold expanded
mixture of liquid and gas passes to a separator where the liquid is
collected and the uncondensed gas is returned to the heat exchanger and
thence to the compressor inlet, completing the cycle.
The design or performance of a gas-liquefaction plant may be devel-
oped without difficulty from principles already demonstrated. With
reference to the stream numbers of Fig. 133, it is desirable to base the
design on specified values of pi, ps, ps, Ti, T3, and T?. Energy balances
involving x, the fraction of the compressed stream which is liquefied,
may then be estabUshed. Thus, if free expansion is used and changes of f
kinetic and potential energy and the exchange of heat with the surround- »*
ings are neglected,
- Hs = xHs + (1 - x)Hj (86)
Since Hi is fixed by the saturation conditions at ps. Equation (86)

may be solved directly for x. Evaluation of x permits establishment of a
complete material balance to correspond to the specified rate of produc-
tion of liquid. The intermediate temperatures and the heat-transfer and
compression duties are then readily calculated. If absorption of heat
from the surroundings is appreciable, these items should be included in
the energy balance.
It is evident that the Linde process is apphcable only to gases which -
cool in free expansion. With reference to Fig. 105, it is seen that free
expansion cooling is not obtained at reduced temperatures above 6.
Hence, for example, in order to liquefy hydrogen, it is necessary to pre-
cool to 100°K before the Linde process may be employed.
i
VAPOR-RECOMPRESSION EVAPORATION
The evaporation of water from nonvolatile solutes or suspensions is /
generally accomplished by multiple-effect evaporators, such as illus-
trated in Fig. 29, page 180. In this scheme several pounds of water may •I'i
be evaporated by the use of 1 lb of heating steam, depending upon the

number of effects used. A somewhat similar result can be produced in
more compact and simple equipment by the use of the vapor-recom-
pression cycle, as shown diagrammatically in Fig. 134.
A dilute solution at a temperature Ti is charged to the evaporator con-
taining the boiling concentrated solution at temperature T2 and pressure
CHAP. XIII] VAPOR-RECOMPRESSION EVAPORATION 591
Pi. As a result of the elevation of the boiling point of the solution over
that of pure water, the vapors evolved are superheated. These vapors
are isentropically compressed to pt, the pressure of the fresh steam
supply, and combined with this supply to the evaporator. Condensate
is withdrawn substantially at the saturation temperature of the heating
steam. Concentrated solution is continuously withdrawn from the
bottom of the evaporator to maintain a constant liquid level and con-
centration.
By this method it is possible to evaporate several pounds of water per
pound of steam supply through the addition of a relatively small amount
V a p o r p., Tj
<D
r'n-T"^ Compressor
—-Ws
©" Compressed
Feed
vapors
®
M_ ©
Saturated
steam
Condensate
Concentrated Bolution
FIG. 134. Vapor Recompression Evaporation.
of energy as shaft work in the compressor. However, the cost per unit of
energy as shaft work is much higher than as heat. The following equa-
tion expresses the energy balance of an evaporator in which mi is the feed
rate in pounds per hour, ma is the rate of evaporation, and m& is the rate of
supply of fresh steam.
miHi + m&Hi = (m2 -f mi)H7 + (wii — m^Hi + w. (87)
where the H values represent enthalpies per pound corresponding to

the designations on Fig. 134. For specified values of mi, m2, Pi and PA,
the work done on the vapors in the compressor is calculated by the
methods previously demonstrated. The heating steam requirements mf
is then obtained directly from Equation (87) and the ratio of evaporation
to steam mi/mi may be calculated. ••
This scheme of evaporation is of interest where power costs are low and
fuel costs high. Such a situation might exist in the vicinity of large
hydroelectric-power developments.
- • , ' ^- . \ - . ' • • • ..
- • ' • « • , • - , . . ; • : / • • . .:.
PROBLEMS
1. Calculate from the changes in the energy functions, the v/ork done, the heat
added, and the changes in T, v, s, u, H, A, and o in the reversible expansion to 1 atm
by each of the following methods of 1 lb-mole of chlorine originally at an absolute
pressure of 500 lb per sq in. and 250°F. The initial absolute entropy of chlorine
is 48.76 Btu/(lb-mole)(°R). Carry out these calculations (I) assuming ideal behav-'
ior, (II) taking into account deviations from ideal behavior by the generalized correla-
tion. Express work in foot-pounds and the energy functions in Btu.
(o) Nonfiow isothermal. •,'
(b) Flow isothermal. ;,
(c) Nonflow isentropioi , ' ' j
(d) Flow isentropic.
(e) Maxwell expansion (irreversible).
(/) Joule-Thomson expansion (irreversible). *
The physical properties of chlorine are as follows*. /
c* = 8.28 + O.OOOSir Btu/(lb-mole) (°R) [T = °R]6
Normal boiling point = 430°R
X = 8780 Btu per lb-mole at 430°R
T, = 750°R; p, = 76.1 atm
Molecular weight = 70.9; pz, = 1.56 g per cc at 430°R
-1
2. Repeat the calculations of parts oil and bll of problem 1 by graphical integra-
tion of the pV changes based on the generalized compressibility-factor correlation.
3. Propane at an absolute pressure of 200 lb per sq in. and 120°F is expanded
through a valve to an absolute pressure of 39 lb per sq in. Using the generalized
thermodynamic correlations and the vapor-pressure equation from Table Ila, page 73,
calculate the temperature on the downstream side of the valve arid the quaUty or
degrees of superheat of the vapor above saturation before and after expansion.
4. The expansion referred to in problem 3 might also be carried out isentropically
in an engine. Calculate the temperature and degrees of superheat or quality after
expansion in this manner and the ideal work performed per pound-mole.
5. A Rankine-cycle power plant generates dry saturated steam at a gauge pressure
of 400 lb per sq in. which is isentropically expanded through engines to an exhaust
gauge pressure of 10 lb per sq in. for use in process heating. It may be assumed that
the condensate is returned to the boilers at its saturation temperature under the
exhaust pressure by an isentropic pump.
(a) By means of the data of the steam tables, determine the number of pounds of
dry saturated low-pressure steam available for process heating per 100 lb of high-
pressure steam generated.
^ K. K. Kelley, U. S. Bur. Mines. Bull. 371 (1934).

CHAP. XIII] PROBLEMS 593
(6) Calculate the ideal thermodynamic efficiency of the cycle. Compare this
result to the Carnot efficiency.
(c) Calculate the ideal water rate of the engine.
6. Determine the effect on the results of problem 5 of superheating the steam 100°F.
7. A pump is driven by a steam engine which deUvers 20 hp and operates on dry
saturated steam at 400 lb gauge pressure and exhausts at an absolute pressure of
2.0 lb per sq in. The engine efficiency is 70 per cent. Using data from the steam
tables, calculate the pounds per hour of steam required and the heat-transfer duty of
the condenser in Btu per hr. Compare the ideal steam rate of the engine with that of
problem 5c.
8. A single-stage compressor is required to compress 450 cu ft per min of CO2
measured at 60°F and 14.5 lb per sq in. abs from an absolute pressure of 5 lb per sq in.
and 80°F to an absolute pressure of 20 lb per sq in. Assuming isentropic compression,
calculate the horsepower required to drive the compressor, the required piston dis-
placement in cubic feet per minute and the discharge temperature, assuming a volu-
metric efficiency of 77 per cent and a compression efficiency of 83 per cent.
9. The compression duty of problem 8 might also be handled by a two-stage
machine with intercooling to 80°P. Calculate the horsepower required to drive the
compressor in this operation and the discharge temperature, assuming the efficiencies
given in problem 8.
10. Anhydrous HCl {ta = -85°C; tc = 51.4°C; p« = 81.6 atm) at an absolute
pressure of 30 lb per sq in. and 80°F is to be compressed to 450 lb per sq in. abs at a
rate of 100 cu ft per min measured at 60°F and 14.5 lb per sq in. A two-stage ma-
chine is to be used with intercooling to 80°F. The compression efficiency based on
isentropic compression in each stage is 80 per cent and the volumetric efficiency
65 per cent. The heat capacity of HCl in the ideal state is given in Table V. The
intermediate suction pressure may be calculated from Equation (73).
Assuming isentropic compression in each stage, calculate:
(o) The horsepower required.
(6) The required piston displacement, cubic feet per minute each stage.
(c) The discharge temperature.
11. A compressor having a 75 per cent mechanical efficiency is to compress 1.5
lb-moles of ammonia per minute from an absolute pressure of 15 lb per sq in. and
100°F to 200 lb per sq in. abs. . The operation is to be in two stages, each of which
may be assumed to be isentropic with intercooling to 115°F. By use of Pigs. 116-119
calculate the power required to drive the compressor when designed for the optimum
intermediate pressure. Calculate the discharge temperature from each stage and the
heat-transfer duty of the intercooler, and compare the optimum intermediate pressure
with that calculated from Equation (73).
12 (o). The refrigeration duty of Illustration 8 is to be performed by a plant using
ammonia as the refrigerant, and with the same temperature levels. Assuming the
same compressor efficiencies, calculate the power input, piston displacement, coeffi-
cient of performance, and heat input to the condenser for the ammonia machine.
Compare these values and also the operating temperatures and pressures with those
of the propane system.
(fc) Repeat the design for an evaporator temperature of 0°F.
13. Design an ammonia-water absorption refrigeration plant for the conditions
specified in Illustration 9 except that the evaporator temperature is to be maintained
at 0°F and the reflux ratio increased to 0.9 because of the more dilute solutions
handled. Compare the heat-transfer duties, circulation rates, and steam requirement
with those of Illustration 9.
14. It is desired to design a plant for the liquefaction at atmospheric pressure of

2000 lb per day of ethylene. The ethylene is available at a pressure of 1 atm and
90°F. Develop a process design for the plant on the basis of compression to 50 atm
in a four-stage compressor with intercooling to 90 °F. The compressed ethylene is
cooled to 90°F, and the recycle gas is heated to 70°F in the heat exchanger. Gain
or loss of heat by the equipment may be neglected and isentropic compression assumed
in each stage of compression. Equation (76) may be used to estabhsh the inter-
mediate pressures in multistage compression. In liquefaction the gas is expanded
irreversibly through a free expansion valve with self-cooling as shown in Fig. 133.
Calculate:
(a) The fraction of the compressor discharge which is liquefied. - ,a
(6) The temperatures at all points of the cycle.
(c) The horsepower input to the gas in each stage of the compressor.
(d) The heat^transfer duties of the cooler and heat exchanger in Btu per hour.
(e) Repeat calculations for isentropic expansion during liquefaction.
16. It is desired to concentrate continuously 1000 lb per hr of a 10 per cent calcium
chloride solution in a vapor-recompression evaporator to 40 per cent concentration.
The dilute solution enters at 100°F. The evaporator is operated at an absolute
pressure of 3.0 lb per sq in. Dry saturated heating steam is supplied at an absolute
pressure of 20 lb per sq in. : 1.1 fj v.
Calculate: ' i ''i
(a) The steam required from outside the cycle in pounds per hour.
(b) The pounds of water evaporated per pound of outside steam.
(c) The horsepower input to the compressor at 75 per cent compression efficiency.
In solving this problem the enthalpy-concentration chart for the calcium-chloride-
water system, Fig. 54, page 281, may be used.
CHAPTER XIV !
THERMODYNAMICS OF SOLUTIONS
The thermodynamic principles which are developed in previous chap-
ters are applied in Chapters X I I and X I I I to problems involving essen-
tially pure substances. The general relations of Chapter XI are in no
way restricted to pure substances and may be validly applied to solu-
tions and mixtures of all types. Such apphcations are comphcated by
the introduction of composition as an added variable.
ii COMPRESSIBILITY OF GASEOUS MIXTURES

The laws of additive pressures and of additive volumes discussed in
Chapter II, page 37, when considered separately, do not necessarily
imply ideal gases. In general, these laws apply with much greater ac-
curacy than the ideal-gas law and are reasonably satisfactory even where
pressures are moderately high. As previously pointed out, either Dal-
ton's or Amagat's law may hold for actual gases, but together they hold
rigidly only for ideal gases. Which of these laws gives the better ap-
proximation depends upon the conditions and the nature of the gaseous
mixture. For example, for mixtures of argon and ethylene Dalton's law
of additive pressures is the more nearly accurate; for mixtures of nitro-
gen and hydrogen the law of additive volumes holds better.
Mean Compressibility Factors. A simple empirical equation of state
for a mixture of gases may be written in a form similar to Equa-
tion (XII-3), thus, U ..* r, :r^
where
p = total pressure of mixture ' ' ;>•:?;>;' ;;u; M I ; ; ' *
F = total volume of mixture
, J I, 2„, = mean compressibility factor of mixture
' ' nt = total number of moles of gases in mixture '
Equation (1) permits calculation of the pressure, temperature, and vol-

ume relationships of gaseous mixtures of known compositions if the mean
compressibility factors are known. The mean compressibility factor is
595
596 THERMODYNAMICS OF SOLUTIONS [CHAP. XIV
a function of the pressure, temperature, and composition of the mixture.

By the same methods which were described for a pure gas, the compressi-
bility factors of a mixture of fixed composition may be determined and
expressed graphically as a function of pressure, temperature, or molal
volume. In order to obtain complete information regarding a particular
system, these measurements must be repeated with mixtures of different
compositions. Because of the great amount of experimental work in-
volved, only a few simple systems of binary gaseous mixtures have been
investigated and these only over relatively narrow ranges of conditions.'
J.i 1
^
/ /
1/ )<A
W
/ /
/ ^1 ^
•i?
./ 1 *•
E /^
S
/.
A
//• /• ''
/ V
^/
A ^
4'
y ^,
0.9
200 400 1
600 800 1000
\ Pressure, Atmospheres
FIG. 135. Compres.sibility Factors of Mixtures of Hydrogen and Nitrogen at 0°C.
Bartlett^ has carried out a series of careful determinations for mixtures

of hydrogen and nitrogen at a temperature of 0°C. The results of this
investigation are shown graphically in Figs. 135 and 136. In Fig. 135
is shown the variation of the mean compressibility factor with pressure
for mixtures of various molal compositions. In Fig. 136 is shown the
variation of the compressibility factor with the composition of the mix-
ture at various pressures. If Amagat's law is assumed to apply, an ex-
> "International Critical Tables," HI, 17.
2 E. P. Bartlett, / . Am. Chem. Soc, 49, 687-701, 195&-1958 (1927).
CHAP. XIV] MEAN COMPRESSIBILITY FACTORS 597
pression for the mean compressibility factor follows:
ZAUA + ZBUB + zcnc
*'m (2)
n«
where
ZA = compressibility factor of component A at the temperature
and total pressure of the mixture
This equation permits the approximate calculation of the mean com-
pressibility factor of a mixture, at any conditions of temperature and
pressure, from the compressibility factors of the component gases at the
^ .<;
^
1.9
^ ,}cnr
^ 0 ,
,1.7
~~^^=^
"^ = 5 - A00
1
b
1"
%
"- - • 600
1
1.3 fin
300
-— — ?,00
1.1
rT=- —
— —— 100
"^ = ^ ^ = •
=^ ^ ^ ^n. -6 0 -
0.9
20 40 60 100
Mole Per Cent of Hydrogen
FIG. 136. Compressibility Factors of Mixtures of Hydrogen and Nitrogen at C C .
same conditions. The dotted lines on Fig. 136 represent the results ob-
tained by the appHcation of Equation (2) to the mixtures of hydrogen
and nitrogen at the conditions investigated. It may be noted that for
this system the validity of Amagat's law is satisfactory, the maximum
error being of the order of 2.5 per cent.
Equations (1) and (2) may be used in conjunction with compressibility
data on the pure components obtained either by direct measurement or
from Fig. 103. This procedure yields fairly satisfactory results where
the compressibility factor of none of the individual components is smaller
598 THERMODYNAMICS OP SOLUTIONS [CHAP. XIV
than 0.5. Serious errors result if the conditions of the mixture are close
to the critical points of any of the components. This method involves
questionable extrapolations of Fig. 103 for components whose liquid-
vapor pressure at the temperature of the mixture is less than the total
pressure of the mixture. , .; ,;,,.-,f .o ,/);>ni • ; ? ' ; . i ^ , . . •
Illustration 1. A gaseous mixture containing 6 lb of methane and 4 lb of ethylene
is compressed to 30 atm at 100°F. Calculate the volume occupied by the compressed
gas.
Basis: 10 lb of mixture: • -i.t
• Methane Ethylene
Pound-moles of gas ' 0.374 . 0.143
Critical temperature, °R 343 510
Critical pressure, atm 45.8 50.9 i
Reduced temperature • 560/343 = 1.63 560/510=1.10 ,*,
Reduced pressure 30/45.8 = 0.654 30/50.9 = 0.59 £.5
Compressibihty factor (Fig. 103) 0.95 0.83
.....^ , ^ , .^ (0.95) (0.374)+ (0.83) (0.143) „ „ , ^
Mean compressibihty factor of mixture = = 0.916
0.517
(0.916)(0.517)(359)( ^ ^ ) ( — ] = 6.46 cu ft
\492A30/
Beattie-Bridgman Equation for Mixtures. Where a high order of

precision is required in dealing with a gaseous mixture, the Beattie-
Bridgman Equation (XII-2) may be applied to the entire mixture using
the following average constants derived from the individual constants of
the components as proposed by Beattie.*
VIZ = ^NVAO ' • . • "^
"/ I • - 1, 1 B^= l^NBo . «
where -= f '- • ' -••• ' •• •' " •

Ami, o-m, etc. = average constants for the mixture
^Ny/A~o, ^Na, etc. = summations of the products of the mole fractions
times the indicated functions of the constants
of each of the component gases •/
Beattie found that equations with average constants derived in this
manner were in good agreement with experimental data on the com-
pressibility of mixtures. For the systems investigated the maximum
error encountered was 0.55 per cent.
' J. A. Beattie, J. Am. Chem. Soc, 51, 19 (1929).
ta-
CHAP. XIV] CRITICAL PHENOMENA OP MIXTURES 599
This method requires a knowledge of t h e Beattie-Bridgman constants
of each component gas a n d is of Umited utility for t h a t reason.
Critical Phenomena of Gaseous Mixtures. T h e behavior of a mixture
in t h e region of t h e critical point is best understood b y reference to the
pressure-temperature diagram of Fig. 137.
Curve AC represents t h e vapor-pressure of a pure compound having
the critical temperature a n d pressure corresponding t o point C". F o r a
single-component system t h e area to t h e left of this curve represents t h e
region of t h e hquid phase, a n d t h e area below it is t h e region of t h e
vapor phase. • •, • •.
Second type
/ / / /
L
3»-
^^^^/ X
^^^>/First
Liquid ©/
c'B
/
/ Vapor and
y liquid
Isometric lines
/ %^
y
Vapor ,,. ;.
Ay/
: ; < ! , • ; !
Temperature *-
FIG. 137. Critical Phenomena of Mixtures.
The sloping straight hnes on Fig. 137 are lines of constant volume
for the single-component system and are termed isochors or isometrics.
It has been found that for most substances the isometrics are approxi-
mately straight, represented by the equation,
p = mT-b '' (3)
where m and b are constants dependent on the volume and the substance
Curve BDECFGH represents what is termed the border curve of a mix-
600 THERMODYNAMICS OF SOLUTIONS [C&AP. XIV
ture having the same average volatility as that of the pure compound
represented by line AC, but made up of two or more substances of dif-
ferent volatilities. The area enclosed by the border curve represents a
two-phase region in which both hquid and vapor are present in equilib-
riiun. Line BDE represents conditions of initial vaporization, to the
left of which is the region of complete liquefaction termed the bubble-
point line. Line HGF termed the dew-point line represents conditions
of initial condensation below which is a region of complete vaporization.
Point C represents the critical point of the mixture. This critical
point does not necessarily correspond to a maximum temperature at
which the hquid phase can exist as in the case with a pure component,
but rather the particular point on the border-line curve where the vapor
and hquid phases become indistinguishable, or where the bubble-point
and dew-point lines meet. In general, both the critical temperature and
the critical pressure of a mixture are higher than those of a pure com-
pound having the same average volatility. |
It may be noted that the dew-point line passes through a maximum
temperature at F. Thus, in the case of a mixture," tiquid may exist at a
temperature higher than the critical temperature. This maximum tem-
perature F on the border curve is termed the critical-condensation tem-
perature. Similarly, in the case of many mixtures the bubble-point
line passes through a point of maximum pressure E, higher than the
critical pressure. The areas FJCM and EKCL represent regions of
retrograde condensation. If the mixture at the condition of point 1 is
compressed at constant temperature, a more dense phase appears at
point G on the dew-point line. As the pressure is further increased, the
quantity of this more dense phase increases to a maximum at point J
and then diminishes, disappearing entirely when point M is reached.
This type of retrograde condensation occurring in area FJCM is called
the first type. If the liquid mixture at conditions of point 3 is heated
at constant pressure, a less dense phase appears at point D on the bubble-
point line, reaches a maximum at point K, and then diminishes, disap-
pearing entirely at point L. The type of retrograde condensation which
occurs in area EKCL is called the second type. • ,
The entire area outside the border curve is a region of homogeneous
fluid in which no phase separations occur. Moving clockwise in the
region about the border curve, the liquid phase merges imperceptibly
into the vapor phase.
Retrograde condensation is of particular interest in petroleum pro-
duction. When natural gas is withdra\vn from high-pressure wells, liq-
uid gasohne condenses upon release of pressure. The residual gas may
then be recompressed and recycled to the underground oil-bearing for-
lo^ff
CHAP. X I V ] CRITICAL PHENOMENA OF MIXTURES 601
mation where it again reaches equilibrium with the liquid oil. As a
result low-boiling fractions of oil enter the gas phase and are subsequently
recovered by the retrograde condensation accompanying the expansion
step. A full discussion of this operation is given by Katz and Kurata.^
It may be noted from Fig. 137 that retrograde condensation of the
first type occurs when a constant-temperature line twice crosses the dew-
point Une (broken line), and retrograde condensation of the second type
uoo
100 200 300 400 500 600

Temperature, °F.
FIG. 138. Pressure-Temperature Diagrams for Mixtures of Ethane and n-Heptane
at Various Compositions.
occurs when a constant-pressure line twice crosses the bubble-point line

(soUd line). Both types of retrograde condensation occur in a given
mixture when the critical point C lies on the border curve between the
maximum pressure 1^ and the maximum temperature F. When the
border curves of a binary system are obtained for the entire range of
compositions, it is found that the critical temperatures occur at diêrent
points on the border-line curves relative to E and F.
These effects are illustrated in Fig. 138 showing the border curves
of various mixtures of ethane and heptane as determined by Kay.°
«D. L. Katz anj F. Kurata, Ivd. Eng. Chem., 32, 817 (1940).
' W. B. Kay, Ind. Eng. Chem., 30, 459 (1938), reprinted with permission.
Line ACa is the vapor-pressure curve of pure ethane, and point Ca is

its critical point. Line BCb has the same significance for pure heptane.
The lines 2-6 are the border curves for five different mixtures containing
the indicated mole percentages of ethane. The corresponding critical
points of the mixtures are designated as CTC^, respectively, the corre-
sponding maximum pressure as E^-Ef,, and the maximum temperatures
as Fi-Fi. It may be noted that for compositions 2 and 3 only retro-
grade condensation of the first type can occur, whereas for compositions 5
and 6 both types can occur. Line CJJi • • • C%Ch represents the envelope
of all critical points for all possible mixtures of ethane and heptane.
600
s^
\ " ' ' •
600
400 •^
&
\r
^
^
.300
Hi
u
V\ V \ ^
&200
S
100 \ \
H
^
^
^J(^
^
'•^?x-- '•*••--*
^
-100
20 40 60 80 100
Mole Per Cent Ethane
Fia. 139. Temperature-Composition Diagrams for Mixtures of Ethane and //
n-Heptane at Various Pressures. w. / '
If the envelope curve were of such shape that it approached point Ct,
with a positive slope, it would indicate the existence of mixtures having
critical temperatures higher than that of the higher boiUng component.
These mixtures would have critical points lying in a clockwise direction
beyond the corresponding points E and F, both of which would be on
CHAP. XIV] CRITICAL PHENOMENA OF MIXTURES 603
the bubble-point line. Such mixtures could exhibit only retrograde
condensation of the second type. This situation is possible but unusual.
v...J,
20 40 60 80 100
Liquid, Mole Per Cent Ethane
FIG. 140. Vapor-Liquid Mole Fraction Relations for Ethane-re-Heptane at Various
Pressures.
Isobaric boiling-point curves and Uquid-vapor composition curves of

the ethane-heptane system^ at different pressures are plotted in Figs. 139
and 140.
Each envelope on Fig. 139 corresponds to the indicated constant pres-
sure The upper curve relates temperature to vapor composition and
604 THERMODYNAMICS OP SOLUTIONS [CHAP. XIV
the lower to liquid composition. Thus, at a pressure of 100 lb per sq in.

and a temperature of 190°F liquid of the eomposition corresponding to
point 1 is in equilibrium with vapor of the compqeition of point 2. It
may be noted that as the pressure is increased the range of the two-
phase region is reduced. This behavior is of great importance in the
design of distillation equipment. At pressures higher than 396 lb per
sq in. pure heptane cannot' exist in the liquid phase in contact with
ethane gas, and this represents the highest pressure at which pure hep-
tane can be separated by distillation. At pressures between 396 and
712 lb per sq in., distillation can theoretically separate mixtures into
pure ethane and liquid solutions of ethane in heptane. At pressures
above 712 lb per sq in. pure ethane caimot be separated, and distillation
can produce separation only into two mixtures, one rich and one poor
in ethane. The maximum pressure at which a two-phase system can
exist is 1263 lb per sq in., the maximum of the envelope curve of Fig. 138.
This is the maximum pressure at which any type of distillation of the
system can be conducted.
Pseudocritical Point. It was shown by Kay* that the pVT data of a
gaseous mixture may be satisfactorily correlated by the generalized com-
pressibility-factor curves of Fig. 103 if pseudocritical temperature and
pressure are used for the calculation of reduced temperatures and pres-
sures. For all except pure compounds or mixtures of compounds differ-
ing little in physical properties, the pseudocritical temperature and pres-
sure are less than the true critical temperature and pressure.
It was found that the pseudocritical properties are approximately
equal to the molal averages of the critical properties of the components.
Thus, the pseudocritical temperature T'c of a mixture is approximately
the molal average of the critical temperatures of the components, and
the pseudocritical pressure p'c is approximately the molal average of the
component critical pressures.
Kay's measurements indicated some differences between correct
pseudocritical properties and the molal averages of the properties of
the components, but an improved general method of estimating pseudo-
critical properties has not yet been developed. Kay^ and later Smith
and Watson' proposed more accurate methods for use with hydrocarbon
mixtures.
Once the pseudocritical point is established, pVT calculations on a
mixture are carried out exactly as for a pure compound, the average
molecular weight of the mixture being used. Compressibility factors
are determined from Fig. 103, the pseudoreduced temperature and pres-
«W. B. Kay, Ind. Eng. Chem., 28, 1014 (1936).
' R. L. Smith and K. M. Watson, Ind.'Eng. Chem., 29, 1408 (1937). !
CHAP. X I V ] COMPRESSIBILITY OF LIQUID SOLUTIONS 605
sure being used. However, it must be emphasized that this method ap-
plies only to the vapor phase and breaks down at conditions within the
border curve of the mixture. The pseudocritical point itself is located
in this region.
Illustration 2. Calculate the volume occupied by l i b of a mixture of 59.9 mole

per cent ethylene and 40.1 per cent argon at a temperature of 25°C and a pressure
of 100 atm.
Ethylene Argon
Critical temperature, °C 9.7 —122
Critical pressure, atm 50.9 48
Molecular weight 28 39.9
Pseudocritical temperature T ; = (0.599) (283) + (0.401) (151) = 230°K

Pseudocritical pressure p'^ = (0.599) (50.9) + (0.401) (48) = 49.7 atm
Average molecular weight = (0.599) (28) + 0.401(39.9) = 32.7
Pseudoreduced temperature T, = 298/230 = 1.3
Pseudoreduoed pressure p'r = 100/49.7 = 2.01
CompressibiUty factor (Fig. 103) = 0.69
(359)(0.69)/298\/ 1 \ „ „„„ ,^
Volume = ^^ 32.7
' _ -r-—
\273/ -1=
\ looj 0.083 cu ft
The experimental measurements of Masson and DoUey indicated a

compressibility factor of 0.712 as compared to the value of 0.69 de-
rived from Fig. 103 and the pseudocritical point in the preceding il-
lustration.
It is believed that use of the pseudocritical point together with re-
duced compressibility factors is the most satisfactory generalized method
for handling pVT calculations on gaseous mixtures, particularly at con-
ditions near the critical. In this region it is more reliable than either
Dalton's or Amagat's laws.
COMPRESSIBILITY OF LIQUID SOLUTIONS

When two miscible liquids at the same temperature are mixed together,
the resulting solution may exhibit ideal behavior, that is, possess prop-
erties which are additive functions of the properties of the pure compo-
nents. Thus, the volume and enthalpy of the solution may be the exact
sum of the volumes and enthalpies of the components. In this case the
density of the solution is the sum of the products of the densities of the
components multiplied by their respective volumetric fractions, and the
specific heat of the solution is the sum of the products of the specific
heats of the components multipUed by their respective weight fractions.
Nearly ideal solutions are formed by closely related homologs of a

series of organic compounds such as heptane and octane or benzene and
toluene at atmospheric temperatures. However, if the temperature is
increased and the liquids are maintained at saturation, the deviations
from ideal behavior increase. Finally, conditions are reached at which
the more volatile component is above its critical temperature, and large
deviations from additivity of volumes result, depending upon the pres-
sure under which this component is confined. Thus, even adjacent
homologs form ideal liquid solutions only at low temperatures corre-
sponding to low vapor pressures for both the solution and the individual
components.
Ideal Systems. An ideal system is defined' as a group of components
which tend to form ideal liquid solutions at low temperatures where the
saturated vapor of each component behaves as an ideal gas. A nonideal
system does not show ideal behavior as a liquid in any temperature
range. For example, chemically dissimilar materials such as alcohol and
benzene never form ideal liquid solutions. Chemical homologs generally
behave as approximately ideal systems but do not form ideal solutions
at all conditions. Thus, solutions of liquid ethane and heptane at at-
mospheric temperatures depart widely from ideal behavior but at lower
temperatures approach ideality. Many somewhat dissimilar materials
.but of low polarity may be treated as ideal systems with accuracy satis-
factory for many purposes. The greatest deviations from the behavior
of the ideal system are encountered with highly polar compounds such as
water, ammonia, the lower alcohols, and ketones.
Density of Liquid Solutions. It was found by Gamson and Watson*
that the empirical correlation of Fig. 109 may be applied to solutions in
ideal systems through the use of pseudoreduced temperatures and pres-
sures. If components A, B, C, • • • may be considered as an ideal sys-
tem, at some low temperature Ta they will form an ideal liquid solution
in which the volumes of the components are additive. Thus,
^ 0 = 7x0 + 7 . 0 + F c o - - - = ^ + ^ + ^ + --- (4)

PAa PBO PCO
Combining (4) with Equation (XII-38), page 502, gives
Ko = + • + (5)
PAIUAO PBiO^Ba PCi^Ca
' B . W. Gamson and K. M. Watson, Nail. Petroleum News, Tech. Sec, 36, fl554
(Aug. 2, 1944). Also "Process Engineering Data," National Petroleum Publishing
Company, Cleveland (1944).
CHAP. XIV] DENSITY OF LIQUID SOLUTIONS 607
where Vo = total volume of solution at the low temperature To for

the respective pure components:
VAO, VBO, VCO = volume at temperature To
mA, inB, mc = masses
PAO, PBO, Pca = densities at temperature To
pAi, PBi, Pci = densities at reference conditions
oiAt, WBo, (^ca = expansion factors at temperature To
WAi, <^Bi, coci = expansion factors at the reference conditions of tempera-
ture Ti and pressure pi where the density is pi
The value of WAI is determined from Fig. 109 by expressing TAI and
PAi in reduced terms based on the critical properties of component A.
It is evident from consideration of Fig. 109 that as To is decreased coô,
WBo, ojco • • • approach equality and it is reasonable to assume that if the
relationship were extrapolated to the absolute zero the values of u for
all compounds would become equal to coo. If To is taken as the ab-
solute zero,
1 /wAiftiA , oiBims . oiamc
Ko = — I i H h•••I (6)
(ô\ PAi PBI PCi
)\ - -
The volume V of the solution at any temperature T and pressure p is
obtained by applying Equation (XII-38) to the entire mixture and com-
bining it with (6),
F =-
where V = volume of liquid solution of mass m^ + m^ -(- mc + • • • at
temperature T and pressure p
w' = expansion factor for entire mixture at T and p
The expansion factor w' of the solution is derived from Fig. 109 by ex-
pressing T and p in pseudoreduced units based on the pseudocritical
properties of the mixture.
Similarly, in molal units,
^, ^ t ^ ^ J_ [-(y^^^)^^^ ^ (VIÔBNB + iviw,)cNc + • • •] (8)
CO CO
where v' = volume per mole of mixture

VIA etc. = molal volume of component A at the reference condition lA
coiA = expansion factor of component A at the reference condition lA
NA = mole fraction of component A
By means of Equations (7) and (8) it is possible to calculate the volume
of an ideal liquid system at any temperature and pressure within the
range of Fig. 109, even though some of the components may be above
their critical temperatures. Care must be exercised that this equation is

applied only to the liquid state. Where vaporization occurs, the vol-
umes of the Hquid and vapor phases must be considered separately by
methods developed in later sections.
Gamson and Watson* also developed methods for calculating partial
molal volumes of the components of solutions if ideal systems in both
the gaseous and liquid states prevail over the wide ranges of conditions.
Illustration 3. The reflux to a fractionating column which is depropanizing nat-
ural gasohne has the following composition in mole per cent:
Methane ' 1.5
Ethane 12.0
)^ ^ Propane 86.5
100.0
To serve as a basis for the design of the pump, meter, valves, and pipe line for this
stream, it is desired to calculate the density in pounds per cubic foot oi this liquid
solution at 100°F and an absolute pressure of 300 lb per sq in.
From Table XXVIII, page 507:
A^ Tc NTo Vc Np, fiwi A'Ci'iui) M NM
°R lb per lb per cc per cc per
sq in. sq in. g-mole g-mole
CH4 0.015 343 6.1 673 10.6 4.36 0.07 16.04 0.24
C2H6 0.120 550 66.0 717 86.1 6.81 0.82 30.07 3.60
CsHs 0.865 666 576.1 642 555.3 9.18 7.96 44.09 38.06
= 647.2 Pc = 652.0 = 8.85
n-
At 100°F and 300 lb per sq in.
{ViUiY Afavg •= 41.90
r; =560/647.2 = 0.865 ^"Z'-

V, = 300/652 = 0.460 _
a' (Fig. 109) = 0.096
From Equation (8),
8 85
: —— = 92.0 cc per g-mole
0.096 ^ ^
^ . (41.90) (62.4) ' i
Density = -— = 28.4 lb per cu ft '
Nonideal Systems. Many solutions, particularly those involving
highly polar components, deviate widely from Equations (7) and (8) and
at low temperature have volumes either greater or less than the su'rn of
the volumes of the pure components. Direct experimental data are
necessary in order to determine the behavior of such systems. Gener-
ally densities or specific volumes of the solutions are determined and em-
pirically expressed as functions of composition at a constant temperature
where the vapor pressures are less than 1 atm. Pressure variations in
this range have a negligible effect on the volume of hquid solutions.
The partial volume of each component may be derived from such data
CHAP. XIV] ENTHALPIES OF GASEOUS MIXTURES • 609
by the methods developed in Chapter VIII, page 287, and individually

expressed as a function of composition.
The deviations from ideal solution behavior which are exhibited by
ideal systems may be designated as deviations due to differences in
molecular size. Nonideal systems show this same type of deviation
and in addition deviations due to differences in molecular type or chem-
ical dissimilarity ^{fhich characterize the behavior of such systems at
low temperatures. In general, it is observed that deviations from ideal
behavior which are due to chemical dissimilarity decrease with increased
temperature, whereas deviations due to differences in molecular size in-
crease as the temperature is increased. The opposite temperature ef-
fects result in a temperature range of minimum deviation from the laws
of ideal solutions for some nonideal systems. At low temperatures large
deviations due to chemical dissimilarity are encountered which are re-
duced as the temperature is increased. Still further increase in temper-
ature leads to important deviations due to dissimilarity in size. It is
possible that Equation (7) is applicable without serious errors to non-
ideal systems in this higher-temperature range where deviations due to
chemical dissimilarity are neghgible in comparison to those due to dif-
ferences in molecular size.
ENTHALPIES OF SOLUTIONS
Where an ideal solution is formed, the enthalpy of the solution is equal
to the sum of the enthalpies of the components at the existing tempera-
ture and pressure. In the general case of a nonideal solution in either
an ideal or nonideal system, a heat of mixing is involved in the formation
of the solution, as discussed in Chapter VIII. The magnitude of the
heat of mixing is a measure of the extent of the deviation from ideal
behavior.
Enthalpies of Gaseous Mixtures. Under conditions of nonideal gas
behavior the enthalpy of a gaseous mixture is best determined by first
obtaining its enthalpy in the ideal state at the existing temperature and
applying'an isothermal correction for the effect of pressure on the en-
thalpy of the entire mixture based upon its pseudocritical properties.
This procedure is straightforward where only the gaseous state is in-
volved. The enthalpy of the mixture at zero pressure is obtained as the
sum of the enthalpies of the components at the existing temperature and
zero pressure. The pseudoreduced conditions of the mixture are then-
calculated and the corresponding value of -—,— for the mixture is
determined directly from Figs. 105 or 106. .^^^ _
Enthalpies of Liquid Solutions. In dealing with solutions of liquids

in an ideal system under such conditions that no component is at a re-
duced temperature above 0.9, it is generally satisfactory to assume that
the enthalpy of the mixture is equal to the sum of the enthalpies of the
components at the temperature and pressure of the mixture. For non-
ideal systems the heat of mixing must be added, as discussed in Chap-
ter VIII. If some components of the Uquid are at reduced temperatures
above 0.9, a more nearly accurate procedure is to calculate the enthalpy
of the liquid from the enthalpy of the ideal-gas mixture at the existing
temperature. It is assumed that the liquid is formed by compressing
the gas to the pseudo-vapor pressure of the mixture which may be con-
sidered as the vapor pressure of the equivalent pure compound whose
critical properties, liquid density, and heat of vaporization are equal to
the corresponding pseudo properties of the mixture. Condensation to
the liquid state is considered as occurring at this constant pressure ac-
companied by what may be termed a pseudo heat of condensation, —X'.
The liquid is then compressed to the pressure of the mixture. Actually
condensation occurs over a range of pressures.
At low temperatures where the saturated vapor of each component
behaves as an ideal gas and the enthalpy of the vapors is independent
of pressure, the heat of vaporization of a mixture of an ideal system at
a temperature f is equal to the sum of the heats of vaporization of the
individual components at this same temperature. Thus, the pseudo heat
of vaporization of a mixture may be estimated by assuming that at the
absolute zero the heats of vaporization of the components are additive.
This value of AJ is corrected to any desired finite pseudoreduced tem-
perature by Equation (VII-32), page 233. Thus for a mixture of WA,
TIB, nc • • • moles of components A, B, C • • •
where A' = pseudo heat of vaporization of the mixture at pseudore-

duced temperature T'r
hiA, XftB = molal heats of vaporization of components A, B, • • '• at their
normal boiling points
TrbATrbB = Teduced temperatures of the normal boiling points of com-
ponents A, B, based on their individual true critical
„ temperatures
T'r = pseudoreduced temperature of the mixture
In view of the other approximations involved, it is generally satisfac-
tory to assume, for the calculation of liquid enthalpies, that the pseudo-
reduced vapor pressure at which condensation of the mixture is assumed
CHAP. XIV] ENTHALPIES OF LIQUID SOLUTIONS 611
to occur is a unique function of pseudoreduced temperature. This

assumption is not accurate and should not be used for other purposes but
leads to little error in calculating enthalpies. Such pseudoreduced vapor
pressures based on the data for n-pentane are given in Table XXIXa.
/ H * — H,V
On the basis of this assumption unique values of f , ' j for the gas-
eous state and xf/'s for the liquid state have been calculated for saturation
conditions as functions of reduced temperatures only and are plotted in
Fig-141. . .-,u-..>.r:l
10 +5
iitf 9
s
8 3
7
^^;=
t^\^ •jp'Xla/j'
I>3
m I
O
6 -5 §•
6^ ^ ah"
4
' ^ • ^
"^•"v..^ P^T...
^ 1 ^ 'c'p'^s'T A §
-10-J:
//
/
'
-F
-15 15?
^^^'
^ {s*-s,\ P,T -1"
ss^"^
= ^ -20
0.65 0.7 0.75 0.80 0.85 0.90 ,0.95
Pseudo reduced Temperature, Tj.
FIG. 141. Enthalpy and Entropy Corrections of Gases and Liquids for Saturation
at Various Reduced Temperatures.
The enthalpy HL of the liquid mixture comprising ut moles at tempera-

ture t and pressure p is then
HL = H*- nXc {~Y^ - ^' + "'^'^ ('^"i)' (^^' - ^') (10)

612 ' THERMODYNAMICS OF SOLUTIONS [CHAP. XIV
where H* = enthalpy of mixture as an ideal gas at temperature t

TT* __ XT
——— = enthalpy correction due to nonideal behavior of gas from

" Fig. 141 corresponding to the pseudoreduced tempera-
ture of the mixture
A' = pseudo heat of vaporization at temperature t
yp'a = enthalpy cor;;ection for liquid from Fig. 141 corresponding
to the pseudoreduced temperature of the mixture a t
saturation
(yicoi)' = defined by Equation (8)
i/*' = value from Fig. 110 corresponding to the final pseudore-
duced temperature and pressure of the liquid
Care must be taken that consistent units are used in Equation (10).
^ TABLE XXIXa » ^ i
PSETTDOREDUCED V A P O R PRESSURES
Tr v'r Tr
0.60
V'r
0.0110
n
0.85 0.310 0.96
v'r
0.759
0.65 0.027 0.90 0.478 0.97 0.815
0.70 0.057 0.92 0.561 0.98 0.873
0.75 0.109 0,94 0.655 0.99 0.935
0.80 0.189 0.95 0.699 0.995 0.990
1.000 1.000
As previously pointed out, it is possible for liquids to exist at tem-

peratures above the pseudocritical temperature of the mixture. Where
only a single phase is present, such liquids may be treated as a highly
compressed gas to which Fig. 106 is applicable. The enthalpy of the
Hquid is then • ^ ,., r •
HL = H*-n,n(^~^) (where ^ > 1.0) (11)

where _, j• "• •-../ , •• • /
——— = value from Fig. 106 corresponding to the pseudoreduced tem-
' perature and pressure of the mixture
Equations (10) and (11) are applicable only to single-phase conditions.
Where two phases exist the enthalpy of each must be considered sep-
arately. • /
' Illustration 4. Calculate the enthalpy of the mixture of Illustration 3 when com-
pletely vaporized (o) at a pressure of 175 lb per sq in. and 100°F and (6) when ex-
isting as a liquid at 100°F and 300 lb per sq in. Express the enthalpies in Btu per
pound relative to the ideal gaseous state at 60°F.
CHAP. XIV] ENTHALPIES OF LIQUID SOLUTIONS 613
Basis: 1 lb-mole of mixture = 41.9 lb

Enthalpies: Ideal gas at 100°F
The enthalpy of the mixture in the ideal gaseous state at 100°F is calculated from
the heat-capacity equations of Table XXI, page 336. Over the small temperature
range involved it may be assumed that the mean heat capacities are the values at
the mean temperature of 80°F or 540°R.
CH4 = 40[3.42-1- (9.91)(0.540) - (1.28)(5.40)2(10-2)](0.015) = 5.0 Btu
CjHe = 40[1.38 + (23.25)(0.540) - (4.27)(5.40)2(10-2)](0.120) = 60.9 Btu
C3H8 = 40[0.41 + (35.95)(0.540) - (6.97)(5.40)2(10-2)](0.865) = 615.5 Btu
Total enthalpy = 681.4 Btu per lb-mole
(a) In the gaseous state at 100°F and 175 lb per sq in., use of the pseudocritical
pressure of 652 lb per sq in. from Illustration 3 gives,
p'r = 175/652 = 0.268; ^ = - ^ = 0.865
647.2
From Fig. 106,
(H* - H)/r; = 1.25 .::
H = 681.4 - (1.25)(647.2) = -127.6 Btu per Ib-iaole • ; , , »;;
Or, since the average molecular weight of the mixture is 41.9,
• ' 7 I i3r=-127.6/41.9 = -3.05 Btu per lb
(b) In the liquid state at 100°F and 300 lb per sq in., from Illustration 3,
r ; = 0.865; p'r = 0.460
The following heats of vaporization at the normal boiling points are recommended
by Doss.'
Xi cal
per \b Nh,
N g-mole nR TIR Tri 1-TA (1 - r,i,)»-^ (1 - rrt,)«-!«
CH4 0.015 2,218 201.2 343 0.587 0.413 3,100 46.5
C2H6 0.120 3,515 331.8 550 0.603 0.397 5,000 600
CsHs 0.865 4,493 416.2 666 0.625 0.375 6,530 5650
' ,-|'.. - '1
' i
Xo cal per g-mole = 6,296
By using the pseudoreduced temperature of 0.865 calculated in Illustration 3 the
foregoing data may be substituted in Equation (9) and the pseudoheat of vaporiza-
tion at 100''F calculated:
X' = (1 - 0.865)''=8(6,296) = 2,940 cal per g-mole
From Fig. HO, ^' = -1.5
From Fig. 141, •/-', = - 2 . 2 ; (H* - HSY/T', = 1.75
In Illustration 3 (ficoi)' was found to be 8.85 cc per g-mole, which is equal to
(8.85) (454)/28,320 = 0.1417 cu ft per lb-mole. Substitution in Equation (10), for
1 lb-mole, gives
H = 681.4 - 647.2(1.75) - (2,940)(1.8)+ (652)(144)(0.1417)(-2.2 + 1.5)
778
= 681.4 - 1132 - 5,290 - 12 = -5,753 Btu per lb-mole or -5,753/41.9
= -137 Btu per lb
' M. P. Doss, " Physical Constants of thie Principal Hydrocarbons, " 3d Ed., Texas
Company, New York (1942). _ ,,,,_.„,..,.:.,.., ,
614 THERMODYNAMICS OF SOLUTIONS ' [CHAP. XIV
ENTROPIES OF SOLUTIONS
Problems involving the expansion or compression of a mixture may be
handled directly by the methods developed in Chapter X I I I if the ther-
modynamic properties of the mixture can be calculated. Volumes and
enthalpies may be obtained by the methods demonstrated in the pre-
ceding sections and entropies are calculated by a procedure parallel to
that developed for enthalpies. If pseudoreduced terms are employed,
Figs. 107 and 111 are applicable to mixtures. Thus, the entropy of a
gaseous mixture is obtained by subtracting the correction of Fig. 107
from the entropy which it would possess were it an ideal gas at the tem-
perature and pressure of the mixture. Similarly, the entropy of a
pseudosaturated liquid mixture is obtained by subtracting a pseudo
entropy of vaporization from the entropy of the pseudosaturated vapor.
On the assumption of a pseudoreduced vapor pressure corresponding
to a given pseudoreduced temperature as given in Table XXIX, unique
values of (s*-Ss)p r for the gaseous ^ a t e and — — (Scp-Ss)?- for the
liquid state have been calculated and are plotted in Fig. 141. The
values of {S*-SS)J,,T for the gaseous state give the corrections in entropy
per mole in going from ideal conditions to conditions of the pseudosatu-
rated vapor at the same temperature and pressure.
Because of the entropy changes which accompany the formation of
even ideal solutions, entropies of mixtures of fixed composition are most,
conveniently expressed relative to the mixture at some reference state
rather than relative to the pure components.
i,; CRITERIA OF COMPLEX EQUILIBRIA

The discussion on page 449 of equilibrium in a closed system is restricted
to consideration of the conditions attained by the system considered as
a whole. However, for complete equilibrium it is necessary that equilib-
rium also be maintained between all parts- mthin the system. In the
example discussed, two phases, liquid ether and a gaseous mixture of
ether and nitrogen, are present, and together constitute a closed system
in which the phases must reach complete equilibrium with each other.
However, each individual phase considered by itself constitutes an open
system which can undergo changes in composition and mass. For this
reason the criteria of equihbrium must be extended to include the effects
of these variables in dealing with such systems.
Chemical Potentials. In order to define a single-phase open system,
it is necessary to specify mass and' composition and two independent
variables such as temperature, pressure, volume, and entropy; for a
CHAP. XIV] CHEMICAL POTENTIALS 616
system of given mass and composition specification of two of the last

four variables serves to fix the other two. Any infinitesimal change tak-
ing place in such a system can be mathematically expressed in terms of
all the infinitesimal changes in its independent properties in accordance
with the properties of a continuous function represented by Equa-
tion (XI-31), page 455. Thus for an infinitesimal change in internal
energy, if volume, entropy, and composition are selected as the independ-
ent variables, the general differential equation is as follows:
^^ = (7?) '^^+iW) ^^+(^) '^"i + (!rV"^+• • • (12)

\0b/v,ni,n2...\aV /s,ni,n2...\0ni/v, S, n 2 . . . \ o n 2 / s , F, m...
where the subscripts 1, 2, • • • refer to the individual components of the
phase which undergo change in mass.
Where composition is constant, all except the first three terms of Equa-
tion (12) become zero and in accordance with the equation
dU^TdS-vdV,
it is evident that \—^) = — P and I —r ) = T when reversible

changes are under consideration. The differential coefficients of mass
are designated by the sjmibol pi. Thus,
Oii)vs=(^ ; in2)vs = (^ (13)

\dni/v,s,7is... \on2/v,s,fii...
and Equation (12) may be written for reversible changes as
« dU=TdS-pdV+(fii)v,sdni+(n,)r.sdn2+--- • (14)
It may be noted from Equation (14) that a change in the total internal
energy of a system is the sum of the changes of a number of energy
terms, each of which comprises ah intensive and an extensive factor.
Thus, for mechanical work of expansion pressure is the intensive and
volume the extensive factor. Similarly temperature is the intensive
factor of heat energy and entropy the extensive factor. Since ni and 112
are extensive factors, it follows that (fii)v,s and (fi2)v,s represent in-
tensive factors of energy associated with the masses of components 1,
2 • • • in solution and hence, like temperature and pressure, may be desig-
nated as potentials.
Expressions similar to Equation (14) may be developed for the other
three energy functions. Thus, for reversible changes
dH = TdS+Vdp+ (Mi)s,^c^ni+ {,x,)s.^dn,+ - • • (15)
dA=-SdT-pdV+ (fiih. rdni+ (fi^h. rdm-l (16)
dG = -SdT+Vdp+ (Mi)r, p dm + (^2)7-. pdn2+• • • (17)
From the definitions of the energy functions,

U = H-pV = A + TS = G+TS-pV (18)
Differentiation yields
dU = dH-pdV-Vdp = dA + TdS + SdT
= dG+TdS+SdT-pdV-Vdp (19)
Adding (—T dS + p dV) to each of the equalities, gives ,,
dU-TdS + pdV = dH-TdS-Vdp = dA + pdV+SdT
= dG+SdT-Vdp (20)
If Equation (20) is compared with Equations (14-17) it is evident that
(MI)F, S = (MOS.P = (MOT.F = (MOT.P = Ml (21)
and ,.= f?£) J'-E) f?A)

\dnjv,s,n2... \oni/s,p,m... \dni/T,v, n2..
/dG\
= (— =Gi (22)
As previously pointed out m is the intensive factor of the internal en-

ergy associated with the mass of component 1, but it may be defined in
terms of any of the energy functions by Equaôn (22). This intensive
property is designated as the chemical potential. From the definition of
partial molal quantities (page 285) and Equation (22), it is evident that
the chemical potential of a component is equal to its partial molal free energy.
The other partial derivatives in Equation (22) are equal to the chemical
potential but are not partial molal quantities in the restriction of con-
stant temperature and pressure.
If a closed system consisting of several phases, each of which is an
open system, is at equilibrium, its properties as a whole must satisfy the
criteria established on page 449. Thus, if the conditions of restraint are
constant temperature and pressure, at equilibrium, dG = 0, ii no means
of performing useful work is present. Also, any change in the free energy
of the system must equal the sum of the corresponding changes of its
parts. If the properties of the different individual phases of the system
are identified by prime markings and the different components of the
various phases by subscripts, from Equation (17),
dG=i-S' dT+V dp + ti[ dn[ + ^'^ dn'^ + • • •) '.;
-\-i-S" dT-^V" dp + ^[' dn{' -I-M2' dn',' + • • •) //
-1- ( - S ' " dT+V" dp + n[" dn[" + M2" dn'," + . . . ) + ..-. (23)
At equilibrium under conditions of constant temperature and pressure
dG = 0,dT = 0, dp = 0, and
(jx'i dn'i + ix'i'dn['+ ix'i" dn'i" •] )
+ (ni dn'2 + Ma' dn'/ + Ms" dn'^" + ••.) + ••• = 0 (24)
CHAP. XIV] CHEMICAL POTENTIALS 617
If the total number of moles of each component in the system remains

constant,
dn[ + dn[' + dn["-\ = 0 and dn^ + dn'i' +dn'2" =0 etc. (25)
From comparison of Equations (24) and (25), it is evident that those
two equalities are satisfied only if
Ml = Ml = Ml = •
l,_"'\ (26)
M2 = M2' = M2, — • • • J
In the development of Equation (26) it was assumed that the total

number of moles of each component in the system remains constant.
This is true in the case of a system in physical equilibrium not involving
chemical reactions. Where chemical reactions take place, the total num-
ber of moles of a component in the system may vary as a result of certain
types of possible changes. However, the system must be in equilibrium
wdth respect to all possible changes, both physical changes involving no
change in the number of moles of the components as well as chemical
changes. Therefore, Equation (25) represents one type of change which
must be considered, and for this type Equation (24) is satisfied only if
(26) is true. Thus, in a system, at equilibrium the chemical potentials of
any component must he equal in all phases in which it can be present.
This criterion is of far-reaching significance and applies to equilibria
under any conditions of restraint. For example, instead of considering
a system restrained to constant temperature and pressure, the foregoing
analysis may be applied to a system at constant volume and tempera-
ture. Such a system might be achieved by means of a container of rigid
conducting walls in contact with a constant-temperature heat reservoir
and divided into two parts by a semipermeable membrane. In one com-
partment might be placed a solvent to which the membrane is permeable
and in the other compartment a solution in this solvent of a solute to
which the membrane is not permeable. If the compartments are ini-
tially filled at constant pressure, solvent will diffuse into the compart-
ment containing the solution and increase its pressure until an equilib-
rium is reached with different pressures in the two compartments but
with equal values for the chemical potential of the solvent. This fol-
lows from Equation (XI-19) which states that the criterion of equilibrium
in a system at constant temperature and volume is that dA = 0. On
this basis an expression for dA analogous to Equation (23) may be writ-
ten, and Equation (26) may be derived by the same procedure as before.
Systems involving other conditions of restraint are more difficult to
visualize but are subject to the same criteria regarding equihbrium be-
tween phases. „ i-i„ . - , : . . ,
Gibbs Phase Rule. In any system comprising several phases existing

at equilibrium there is a limited number .of intensive properties which
can be freely varied without causing a change in the number of phases
or the number of components in some phase. For example, in a system
comprising a pure liquid and its pure vapor at equilibria, temperature
can be varied over a wide range without affecting the number of phases.
However, it is impossible to vary both temperature and pressure freely
without causing the disappearance of one phase or the other. The num-
ber of intensive properties which can be varied without changing the
number of phases or the number of components in any phase is termed
the number of degrees of freedom of the system.
Since in addition to temperature and pressure the chemical potential
of each component is an intensive property, the tcnal number of intensive
properties subject to variation in a single phase containing C components
is C + 2. If this phase is a part of a system in equilibrium at constant
temperature and pressure, a differential expression for the free-energy
change, which is the criterion of equilibrium, may be written in the form
of Equation (23). Thus, for each phase,
c^(?' = f ( 7 ' , p , M i , M 2 , - - - ) = 0 ; .; *•
•"i V, aG"=^f"(T,p,^^,t,2,---)-0 ' "• '
i.i :.: • Etc.

where the functions /', / " , / ' " are characteristic of the respective phases.
In this manner, if <j) phases are present, simultaneous equations may be
written in terms of C -f 2 variables. Since for determination of the
variables in simultaneous equations one equation is required for each
variable, it follows that the number of variables not fixed by the equa-
tions is C -\-2 — cj) or
F = C+2-,j> • (27)
where '
F = degrees of freedom " / / ';•
C = number of components
(t> = number of phases
Equation (27) is the famous phase rule of Gibbs, developed in 1875.
The foregoing derivation follows that presented by Keenan^";
As an example of the application of the rule, the equilibrium among
the liquid, vapor, and solid states may be considered. Thus, if only a
pure Uquid, for example water, is in equilibrium with its vapor, C = 1,
<^ = 2, and F = 1; one intensive property, either temperature or pressure,
» J. H. Keenan, "Thermodynamics," John Wiley & Sons, New York C1941).
CHAP. XIV] FUGACITY 619
but not both may be freely varied. If, however, ice is also present in
the system, C = 1, <^ = 3, and F = 0. Under such conditions no condi-
tions can be varied and the specification of three phases fixes both tem-
perature and pressure. If instead of pure water a binary solution of
water and alcohol is in equihbrium with its vapor, C = 2, = 2, and
F = 2. Thus, two properties may be freely varied in such a system.
For example, both temperature and pressure may be varied freely over
restricted ranges with corresponding changes in composition of the
phases, but all three variables cannot be independently varied. A speci-
fied composition and temperature fixes the corresponding pressure.
When used in connection with the phase rule, the number of compo-
nents is the least number of independently variable chemical substances
from which the system in all its variations can be produced. Elsewhere
in this text the term component is not used in this restricted sense.
FUGACITY
The concept of the chemical potential is of value in establishing the
fundamental thermodynamic requirements which must be satisfied by
a complex system when equilibrium is reached. Since at equilibrium
the chemical potentials of a component must be equal in all phases in
which it can occur, differences in the chemical potentials must be equal-
ized by redistribution of the component as a system approaches equihb-
rium. If the chemical potential is high in one phase and low in another
material escapes from the first to the second phase until equality is
reached. Thus, the chemical potential may be considered as a measure
of what may be termed the escaping tendency.
Since the chemical potential is a measure of the escaping tendency,
it follows that any other property which is a unique function of the chem-
ical potential is also a measure of the escaping tendency. Such a function,
termed the fugacity, was so defined by Lewis" as to simplify the
mathematical relationships of the equilibrium. By definition,
{dÂ = dGA = RTd\nJA)T (28)
where
/A = fugacity of component A at temperature T
HA = chemical potential of component A at temperature T
GA = partial molal free energy of component A at temperature T
From this definition it follows that in any system which is restrained to
a constant temperature at equilibrium the fugacities of any component must
he equal in all phases in which it appears. For a pure component the par-
" G. N. Lewis, Proc. Am. Acad. Arts, Sd., 37, 49 (1901). ' ' •'
tial molal free energy GI is equal to the molal free energy G and Equa-
tion (28) reduces to • ,
(dG = RTd In f)T (29)
Equations (28) and (29) do not serve to define the numerical values of/
but do establish the changes in / which correspond to specified changes
in G or G at constant temperature.
From Equation (k), Table XXIV, page 472,
\dp/T
Combining this with Equation (29) gives I
' • / a In A ^ _ v _
(30)
\ ap IT RT
In appljang Equation (30) to a gas v may be replaced by zRT/p.

T^^^' [d \n f=^zd In P)T (31)
From Equation (31) it is evident that for an ideal gas, where 2 = 1.0
the fugacity is proportional to the pressure. The definition of fugacity
is completed, and numerical values are assigned to it by arbitrarily set-
ting the fugacity of an ideal gas equal to its pressure. Thus fugacity
has the units of pressure and is numerically equal to pressure in the ideai
gaseous state. Under other conditions pressures and fugacities are not
equal, and fugacity is sometimes referred to as a corrected or thermo-
dynamic pressure. However, from the definition it is evident that the
corrections involved are a function of free-energy changes and the fu-
gacity is useful only for equihbrium calculations where free-energy
changes are the criterion.
Fugacities of Pure Gases. An expression for the fugacity of a gas at
any conditions is obtained by rearranging Equation (31) in terms of the
ratio of fugacity to pressure f/p which by definition is equal to 1.0 at
zero pressure where ideal behavior is realized. Thus s , ,; /
d\nf/p = zdlnp — d\ap= {z--l)d\np • (321 /
Integration from p = Otop gives / -
\nl= r{z-l)dlnp= r^^^dp . (33)

p Jo Jo P /'
Equation (33) may be integrated graphically for any substance for which
compressibility data are available. The integration is more accurately
carried out by analytical methods if an equation of state is evaluated.
On the basis of the theorem of corresponding states a generalized in-
tegration of Equation (33) is obtained from the compressibility-factor
CHAP. XIV5 FUGACITIBS OF PURE LIQUIDS AND SOLIDS 621
relationship of Fig. 103. Various investigators have carried out this in-
tegration which is conveniently presented by plotting the ratio of fu-
gacity to pressure (//p) against reduced temperature and pressure. This
ratio / / p = V will be termed the fugacity coefficient and is equal to unity
where the ideal-gas law is vaUd. This ratio is frequently termed the
"activity coefficient," but it is believed preferable to reserve this latter
term for expressing relationships between fugacities and composition in
solutions.
Average values of the fugacity coefficients of gases are plotted in
Fig. 142.'-^ This chart was derived from the same data on which
Fig. 103 was based. For many purposes these charts may be taken as
appHcable to all gases. The errors involved are generally less than
10 per cent.
The fugacity of any gas or vapor at specified conditions is readily ob-
tained when its critical temperature and pressure are known.
Illustration 5. Calculate the fugacity of methane at 122°F and 1000 lb per
sq in. abs.
From Table XXVIII, Tc = 343°R and p, = 673 lb per sq in.
^ , , . 122 + 460 , „„
s ; 343
Reduced pressure = -VT# = 1-49
. :.* Fugacity coefficient (Fig. 142) = 0.94 '•
Fugacity = (1000) (0.94) = 940 lb per sq in.
Fugacities of Pure Liquids and Solids. From the basic concept of
fugacity as a measure of escaping tendency, it follows that the fugacity
of a liquid or solid must be equal to that of its vapor in equilibrium with
it. Thus, when a liquid is in equilibrium m t h its pure vapor, the fugac-
ity of the liquid is determined by calculating the fugacity of the vapor
at the equilibrium temperature and pressure.
Illustration 6. Calculate the fugacity of liquid benzene in equilibrium with its
pure vapor, at a temperature of 428°F. The critical temperature of benzene is 550°F,
and its critical pressure is 700 lb per sq in. The vapor pressure at 428°F is 281 lb
per sq in.
650 + 460
Reduced pressure = 281/700 = 0.40
Fugacity coefficient (Fig. 142) = 0.79
Fugacity of vapor = fugacity of liquid = (281) (0.79) = 222 lb per sq in.
The fugacity of the liquid state is a function of total pressure as expressed
by Equation (30). This equation may be integrated by expressing v as
" B. W. Gamson and K. M. Watson, Natl. Petroleum News, Tech. Sec, 36, /J623
(September 6, 1944). Also "Process Engineering Data," National Petroleum Fub-
Ushing Company, Cleveland (1944). - •
l O CO CO CO lO Tl> CO
CO
o o o o o o o
u 'luaioijjaoQ iIq.tDBSi\j
d/S
CHAP. XIV] EFFECT OF TEMPERATURE ON FUGACITY 623
a function of pressure by means of Fig. 109. Since the volume of a liquid

is little affected by pressure at reduced temperatures below 0.9, an ap-
proximate integration assuming v constant at an average value is gen-
erally satisfactory in this range. Thus,
where
/ff = fugacity at total pressure ir
/p = fugacity at normal vapor pressure P
Equation (30) permits calculation of the fugacities of liquids at any
total pressure but the integrated form, Equation (34), is useful only at
reduced temperatures below 0.9.
Illustration 7. Calculate the fugacity of liquid benzene at 428°F if the liquid is
in an atmosphere of hydrogen such that the total gauge pressure is 2000 lb per sq in.
The average density of liquid benzene at these conditions is 0.63 g per cc.
f=• = 1.98 cu ft per lb-mole
(0.63) (62.4) ^
Substitution in Equation (34), gives
Xnl^= 1.98(2015-281) ^
222 (10.71) (428+ 460)
log/,r/222 = (0.361)(0.434) = 0.157 or /,,/222 = 1.43
f^ = the fugacity at 2015 lb per sq in. = (1.43) (222) = 318 lb per sq in.
Effect of Temperature on Fugacity. The following equation is derived

by the method of Lewis and Randall'' to relate fugacity to temperature.
A substance in a given state at temperature T, pressure p, molal free
energy G, fugacity/, and molal enthalpy H is compared to the same sub-
stance at the same temperature but at a low pressure where the free
energy is G*, and the fugacity /* is equal to the pressure. Under these
conditions the enthalpy H* is independent of pressure. From Equa-
tion (29),
e*-G = i2r(ln/*-ln/) (35)
By differentiation with respect to temperature at constant pressure,
Since /* is equal to pressure.

( dT A
"Levris and Randall, "Thermodynamics," McGraw-Hill Book Company, New
York (1923).
Combining (35) and (36) results in
(f)r(fa4*-l-(m
Combining (37) with Equation (XI-4) and (1) of Table XXIV, gives
^
'"
âin/\ H*-H
(-
where /
H = molal enthalpy at an elevated pressure p and temperature T
H* = molal enthalpy at the same temperature T but at a pressure suffi-
ciently low so that the gas behaves ideally
The quantity H* — H may be obtained from the generalized correlation
of Fig. 105 or may be calculated from compressibility data by the meth-
ods developed in Chapter X I I . Charts such as Figs. 117 or 120 or
thermodynamic tables permit direct evaluation of H* — H for either the
liquid or vapor state as a function of temperature and pressure.
Fugacities in Solutions. The partial molal extensive thermodynamic
properties of a component in solution are related to each other by expres-
sions parallel to those developed for pure substances in Chapter XI.
For example, by Equation (XI-4), page 446, for any system,
G = H-TS (39)
By differentiation of this equation with respect to the number of moles
of component 2, all other conditions being held constant, there results
\dn2/T,p,m... \dn2/T,p,m... \dn2J T,p,m'...

Each of the derivatives of Equation (40) represents a partial molal quan-
tity. Therefore,
G2 = S 2 - r s 2 (41)
Equation (41) is entirely parallel in significance to (39). In a similar
manner it may be shown that the other relationships of Chapter X I
which involve extensive properties are valid for partial mo'lal quantities.
On this basis, from Table XXIV, Equation (k),
iiôm Equation (1),

/dG2\ G2—Ha ,,
—TT- •r'-": i (43)
CHAP. XIV] FUGACITIES IN SOLUTIONS 625
Combining (28) with (42) gives
\ dp JT RT ^^^
Combining (28) and (43) in the manner used in developing (38), gives
/aln/zN H|-H
(45)
\ dT ) ~ RT^
In using Equation (45) care must be taken that the partial molal en-
thalpy H2 is expressed with respect to the same reference state as the
molal enthalpy of the ideal gas, H | .
Equations (28), (44), and (45) permit calculations of the fugacities of
components in solutions from a variety of data such as vaporization,
solubility, or distribution equilibrium measurements. For example,
the fugacity of a pure solid at a specified temperature may be deter-
mined from its vapor pressure, and this value is equal to the fugacity
of the solid as a solute in any saturated solution at this same temperature.
The variation with temperature and pressure of the fugacity in solution
can then be calculated by means of Equations (44) and (45), if partial
volume and enthalpy data are available.
If Equation VIII-34 is written for free energies, at constant tempera-
ture and pressure,
rii dGi +112 dSi + ws doz -f • • • = 0 (46)
or, considering 1 mole of solution, at constant temperature and pressure,
NxdGi^-N2dG2-\-N3dG3-V--- =Q (47)
Equation (47) may be written as partial derivatives with respect to the
mole fraction of any selected component, temperature and pressure being
kept constant
NJ^\ +NJ'^\\NJ'^ +... = 0 (48)

\5iV2A.p V^Wr.p VaWr,/ , ^'
Combining (28) and (48) gives '
u,mi) +N,c^\ +N,{'-m +...=0 m

If Equation (49) is applied to a binary solution, dNi = —dN2 and
This important relation between the fugacities of the components of a

solution is termed the Gibbs-Duhem equation. ,. ,„
Fugacities in Ideal Solutions. In an ideal solution, either gaseous or

liquid, the fugacity of a component is proportional to its mole fraction.
/2 = ^ 2 / ; / (51)
where • ' •^''^'''"^' '''•''"'' ' •''-•''"> ;4H»VV ...v .
/ j = fugacity of component 2 in the solution
/ j = fugacity of pure component 2 at the temperature fnd pressure of
the solution . , ,
N2 = mole fraction of component 2 . •
• I 'i • ! !-> .' -• 'i • ;
Lewis and RandalP' define the ideal solution solely by Equation (51)
and point out that all other properties attributed to ideal solutions such
as additivity of volumes and enthalpies follow from this definition.
For gaseous mixtures at low pressures Equation (51) is equivalent to
Dalton's law while for liquid solutions whose vapors are ideal gases it is
equivalent to Raoult's law. However, at conditions resulting in large
deviations from ideal-gas behavior, fugacities are not equal to partial
pressures, and Equation (51) may be looked upon as an improved state-
ment of Dalton's and Raoult's laws which is generally applicable over
a much wider range of conditions.
The assumption that gases form ideal solutions is of a much higher
order of accuracy than the ideal-gas law and may be applied with accu-
racy satisfactory for many purposes to gaseous mixtures at pseudore-
duced pressures less than 0.8. If the pseudoreduced temperature is low,
of the order of 1.0 or less, difficulty is encountered in handling the high-
boiling components of the mixture, which, if they existed alone at the
temperature and pressure of the mixture, would be liquefied. In such
cases the term f'2 of Equation (51) may be estimated from the broken-
line extrapolations of the curves of Fig. 142 into the two-phase region.
Thus, hypothetical fugacity coefficients for the pure gases atf pressures
above their vapor pressures are obtained. This procedure is not satis-
factory where extended extrapolation is required. For mixtures at
higher temperatures, above the critical temperatures of all components,
the assumption of ideal solutions is more satisfactory and may be ex-
tended to higher pressure ranges. j ! . /
Illustration 8. A mixture of gases has the following composition expressed in

mole per cent:
Methane 17 /
Ethane 35 I '
Propane 48 . "'
106
Assuming an ideal gaseous solution, calculate from Fig. 142 the fugacity of each
CHAP. XIV] STANDARD STATES 627
component when the mixture is at an absolute pressure of 300 lb per sq in. and a
Solution: Methane Ethane Propane
Tin (Table XXVII) 343 550 666
Pc lb per sq in. (Table XXVII) 673 717 642
Tr 1.63 1.02 0.84
Pr 0.446 0.418 0.466
;- (Fig. 142) 0.98 0.87 0.73
/ ' = 3001- ' ,' , 294 261 219
/ = A7' 50.0 91.5 105
ACTIVITY
For treatment of problems involving solutions and chemical equilibria
it is convenient to define another thermodynamic property which is di-
rectly related to fugacity and hence also to free energy and the chemical
potential. This property, called activity, a, is defined as the ratio of the
fugacity of a component in a given state to its fugacity in an arbitrarily
defined standard state at the same temperature. Thus,
-(fX <=^)
where
a = activity
/ = fugacity in the given state
f° = fugacity in the standard state at the same temperature
Combining Equations (28) and (52) gives
i RTlna = G-G° (53)
where
G = partial molal free energy in given state at temperature T
G° = partial molal free energy in the standard state at the same
temperature T
The concept of activity is particularly useful in dealing with liquid
solutions. For example, if the standard state of a component which is
in solution is taken as the pure component at the temperature and pres-
sure of the solution, the activity becomes a function of the concentration
or fraction of the component in the solution. Thus, the activity provides
a basis for the thermodynamic expression of concentrations or composi-
tions in terms directly related to free energy by Equation (53).
Standard States. The choice of the standard state necessary to com-
plete the definition of activity is arbitrary since it affects only the nu-
merical magnitude of the function and not its relationship to other prop
erties. It is, however, desirable to select standard states such that the
resulting numerical scales will be convenient and correspond to common
expressions of composition in ideal systems. Different standard states
may be selected for the same substance for use in different types of re-
lationships, the choice being dictated by convenience. It is important
however that the standard state of each component be kept the same
throughout any one series o£ relationships or calculations. It is evident
that the numerical value of activity is without significance unless the stand-
ard state is specified.
The following standard states have been found convenient for certain
types of calculations and are in more or less general use.
Components of Gaseous Mixtures, (a) It is frequently convenient to
define the standard state as the state of unit fugacity, orf°— 1.0. With
this choice of standard state the activity of a gaseous component is equal
to its fugacity and has the same numerical value. In mixtures of
ideal gases the activity of each component is equal to its partial pressure.
(6) In other cases it is more convenient to define the standard state
as the pure component at the temperature and the pressure of the mix-
ture. With this choice of standard state the activities become equal to
the mole fractions in mixtures which form ideal solutions. It may be
noted that with this definition the standard state changes with change
in pressure whereas standard state (a) is independent of the pressure of
the system.
Pure Liquids and Solids. When a liquid or solid is involved in a
process in its pure state, it is customary to designate its activity as
unity under a specified pressure. Three choices have been used in the
specification of the pressure of the standard state.
(a) The pure component under a pressure of 1.0 atm is widely used
as the standard state for nonvolatile substances.
(6) For volatile substances whose vapor pressures exceed 1.0 atm it
is convenient to define the standard state as the pure substance under
its own vapor pressure.
(c) In certain calculations it is convenient to define the standard state
as the pure substance under the pressure of the system. With this
definition the standard state varies with change in pressure whereas
states (a) and (b) are independent of pressure.
Components in Nonelectrolytic Solutions. The standard state for either
a solid or liquid component in a nonelectrolytic solution may be taken
as the pure component in either the solid or liquid state under one of the
three pressure designations enumerated in the preceding paragraph.
Pressure designations (a) and (6) have the advantage of defining the
standard state independent of pressure. Use of designation (c) leads
CHAP. XIV] STANDARD STATES 629
to the convenient relationship that activities are equal to mole fractions

in ideal solutions under constant pressure.
An alternate standard state which is sometimes desirable for sparingly
soluble materials is so defined that the mole fraction is equal to the ac-
tivity as the mole fraction approaches zero. Thus,
Lim ^ = 1.0 (54)
It may be experimentally demonstrated that as infinite dilution is ap-

proached the fugacity of a solute becomes proportional to its mole
fraction,
f2 = KN2 as N2—>-0 or (7^7) =K (55)
Combining (52), (54), and (55) for an infinitely dilute solution gives
or !* '. /^' = fc2 (66)

Thus, the standard state, according to the definition of Equation (56),
is a hypothetical state in which the fugacity is equal to A; 2, the factor re-
lating fugacity to mole fraction at infinite dilution. It should be noted
that the standard state is not the state of infinite dilution but is defined
by reference to this state. A standard state defined in this manner will
be termed a standard state referred to infinite dilution. The pressure at
which such a standard state is defined may be either (a) 1 atm, (b) the
vapor pressure of the solvent, or (c) the pressure of the system. If desig-
nation (c) is employed the activity is equal to the mole fraction at all
concentrations in an ideal solution at constant pressure exactly as when
the pure component is chosen as the standard state.
I t is important to recognize that the standard state referred to infinite
dilution is a hypothetical pure state of unit activity derived by linear
extrapolation of the ideal fugacity-composition relationship at infinite
dilution. This may be obtained by drawing a tangent to the fugacity-
mole-fraction plot at zero mole fraction of solute. The intersection of
this tangent with the abscissa of unit mole fraction N2= 1.0 gives the
fugacity fl' of the standard state. Since partial molal volumes, en-
thalpies, and heat capacities are independent of composition in an ideal
solution, it follows that these properties of the hypothetical reference state
are equal to the corresponding actual partial properties of the solute in in-
finite dilution.
An alternate standard state referred to infinite dilution is so defined
that molality is equal to activity as the molality approaches zero. Thus,
Lim — = 1 . 0 as ni2—>-0 (57)
As infinite dilution is approached the fugacity becomes proportional

to molality, or
f^^k'^'mt as m2—»-0 or (^\ = fcj' (58)
Combining (52), (57), and (58) for an infinitely dilute solution gives
..a^'=m2=A = _ ^ or Jl" = K' , (59)
The hypothetical standard state can be obtained by drawing a tangent

to the fugacity-molality plot at zero molality of solute. The intersection
of this tangeat with the abscissa of unit molality m2 = 1 gives the fu-
gacity/^',' of the standard state.
Gases in Ldquid Solutions. The activity of a gas dissolved in a liquid
may be referred to a standard state of unit fugacity, making activities
equal to fugacities. More frequently it is desirable to refer the activities
to infinite dilution so that the activity equals the mole fraction as they
both approach zero. The pressure of the standard state may be desig-
nated by any of the conventions discussed in the preceding paragraphs.
For gases below or not greatly above their critical points a special
standard state is useful in which the standard state of the gas in the
Hquid solution is taken as the hypothetical pure component in the liquid
state at the temperature and pressure of the solution. The fugacity of
such a hypothetical reference state may be estimated by extrapolation
of the vapor-pressure curve if the temperature is above the critical.
This standard state is extensively used for vaporization equilibrium
calculations.
As previously mentioned, the choice of the standard state is based on
convenience and does not affect the relationships involved. However,
it is important that the selected standard state be clearly defined and
consistently adhered to in any specific problem. The concept of the
activity tends to be confusing, and its value and use will be better under-
stood as the applications of the follo\ving sections are developed.
Activity Coefficients. The concept of the activity is useful because it
provides a thermodynamically defined quantity which with proper choice
of standard state is a simple measure of composition in ideal solutions.
Because of its thermodynamic definition it is possible to develop exact
relationships between the activities of components in complex systems
at equilibrium under varying conditions. Where ideal solutions are in-
CHAP. XIV] ACTIVITY COEFFICIENTS 631
volved or approximated, these relationships between activities are di-

rectly translatable into ordinary compositions. If ideal behavior cannot
be assumed, it is necessary to introduce an empirical factor which relates
activity to composition. This factor is termed the activity coefficient y,
which is the ratio of activity to a numerical expression of composition.
The numerical value of the activity coefficient is dependent both upon
the standard state of the activity and upon the units of expression of
composition. Both of these factors may be arbitrarily selected, and the
activity coefficient has no significance unless both are specified.
The most commonly used activity coefficients relate activities to mole
fractions or molalities. The following symbols are used for the three
coefficients of most importance in dealing with systems at low pressures.
72 = a^/Ni , (60)
* ''^ '"' ] y2 = a'JN2 ' '•' ^, (61)
T2' = 027^2 (62)
where
02 = activity of component 2 referred to the pure component
a'2 = activity of component 2 referred to infinite dilution where
02 = N2
a'2' = activity of component 2 referred to infinite dilution where
a'2' = m2
The activity coefficient of a component in a nonideal solution is a

function of composition, temperature, and pressure, which ordinarily
must be derived from direct experimental data on the specffic system.
A variety of data on systems at equihbrium may be used for this pur-
pose. For example, the composition of the vapor in equilibrium with a
liquid solution may be determined. The fugacities of the components
in the hquid solution are then equal to those of the same components
in the gaseous phase and may be calculated by the methods of the pre-
ceding sections. By determining such vapor-Hquid equilibrium data
over a range of compositions the fugacities of the liquid components
may be evaluated. The activities and activity coefficients are then cal-
culated from the fugacities of the standard states. In such investiga-
tions it is necessary to evaluate experimentally the fugacities of all but
one component. The fugacity of the remaining component can then be
calculated by Equation (49) or (50) if a single known value is available
in the range of concentrations covered by the data.
If the unknown component is a nonvolatile material of unknown fu-
gacity or if it is desired to express activities in terms referred to infinite
dilution, it is more convenient to rewrite Equations (49) and (50) in
terms of activities or activity coefficients. The limits of integration of
these expressions are fixed by the concentration of the state of unit ac-
tivity or of unit activity coefficient, respectively, even though the fugacity
is unknown. Since a = f/f°, Equation (47) may be written for changes
in activity due to changes in composition at constant temperature and
pressure. '>
Nidlnai + N2d\na2 + Nsdlnas-i = 0 (63)
or integration, for a binary solution, gives • .yif'i};;
—•dlnai (64)
where x - . '
Nl = mole fraction of the standard state where 02 = 1.0
Equations (63) and (64) are convenient where the pure liquid or solid
component 2 is chosen as the reference state. Since 02 = JV272 Equa-
tion (63) may also be written in terms of activity coefficients. Thus,
NidlnNi + Nidlnyi+N2d\nN2+N2dlny2-\ = 0 (65)
also ^ dNi + dN^ + dNs-] = 0 • '/' , •^~ ^, !
Then since Nid In Ni = dNi, it follows that / , :|
NidlnNi + N2dlnN2 + N3dlnNs-i = 0 (66)
and Nidlnyi + N2dlny2 + Nadlny3-{ = 0 (67)
Integration of (67) for a binary solution gives
I 2ldlny,+
—( C ;.•, . (68)
w! f2
A/2 •
where C is the value of In 72 at A7'2.

Equation (67) may also be written in the form of (50) for binary
solutions: . * /
-^ / a In 7 i \ ,^ / 9 In 72\
Equations (63) through (67) may also be written in terms of common

logarithms and then integrated graphically by plotting log ai, or log 71,
against Ni/N2- The equations in terms of the activity coefficients are
preferable for graphical manipulation, because in general the coefficient
varies less with change in composition than does the activity.
For solutions of more than two components the procedure is more
complicated since an additional integral is introduced for each additional
component. Such integrations are best carried out with respect to a
constant number of moles of all components other than the one under
CHAP. XIV] ACTIVITY COEFFICIENTS 633
consideration. Thus,
In as = / —dlnai+ I — d In as -| (70)
where n\, n^, • • • are constant.

Equations (63-70) are rigorous thermodynamic relations which are
valid for conditions of constant temperature. They are of great value
in minimizing the experimental data necessary to evaluate the properties
of a system and for detecting inconsistent or erroneous measurements.
Where activities are referred to molalities at infinite dilution the ac-
tivity coefficients y" are not constant or equal to unity, even in ideal
solutions except in the dilute range. Their principal value is in dealing
with dilute aqueous solutions in which ionization occurs. Equations (89)
and (70) are not applicable to activity coefficients defined in this manner.
From the definition of activity it is evident that the activities of a
solute when expressed on the basis of the three different'standard states
used in Equations (60), (61), and (62) must be related to each other
by constant factors, independent of concentration. Thus,
, f, = a,fl = a'JV=a','fl" (71)
The conversion of activities from one standard state to another for a
given system is obtained by rearranging Equation (71) to give
It is thus evident that the ratio of activities relative to different stand-

ard states is given by the ratio of the fugacities of the standard states
according to Equation (72) and that these conversion factors are con-
stant over the entire range of concentrations.
A further relation between activities in aqueous solutions results from
consideration of the infinitely dilute solution where, since 1000 g of
water is equivalent to 55.51 g-moles, it follows that,
^ nii = 55.51 iV2 or dm2 = 55.51 dN2 as m2 —>• 0 (73)
and where, from Equations (55, 56),
• (î.=r^" . ^ <'«
and where, from Equations (58, 59),
(^) =fl" (75)
Combining (73-75) gives / r = 55.51/2°" -' (76)


02' = 55.51 a ^ (77)
A comparison of activities and of activity coefficients referred to the
three different standard states is illustrated diagrammatically in Fig. 143
in which fugacities, activities, and activity coefficients of a solute in a
nonideal solution are plotted against mole fraction. In this example,
fl = 0.5 and 02= 1.0 at N2 = 1.0. The value of fl' is obtained by
drawing a tangent to the fugacity curve at zero mole fraction and ex-
tending it to A^2 = 1.0 w h e r e / j ' = 1-3- The value oi fl" is obtained
by constructing a tangent to a fugacity-molality curve at zero molaUty
1.3
and extending it to m2 = 1.0, or from Equation (76),fl" ~ -z^r-^ — 0.0234.
55.51
The ratios of activities referred to the different standard states are given
by Equation (72) as
For mole fractions of zero and unity the following values are obtained-.
AtiV2 = 0 AtiVa=1.0(m2 = 00)
' " 02 0 1.0
ai. 0 0.384
oj' 0 V 21.35
72 2.60 ^ 1.0
72 1-0 0.384
. . . :i- y'i 1-0 ^z-.-,:; 0 f .^ > -• ?:
Where fugacities of component 2 are unknown the ratio of standard

fugacities in Equation (72) can be obtained from the activity coefficient
of component 2 at infinite dilution (72)0- By combination of Equations
(52) and (60) and Equations (61), (57), and (73), and, since at infinite
dilution 72 and 72' are each equal to unity, Equation (72) becomes
02 = (72)0 ^2 = (72)0 a 2 ' / 5 5 . 5 l .^ (77a)
Illustration 9. From the data for the vapor pressure of water above sugar
(C12H22O11) solutions taken from the International Critical Tables and summarized
in Table XXX, calculate the corresponding
(o) Fugacities of the water in solution.
(6) Activitiesffliand activity coefficients 71 of the water referred to pure liquid water.
(c) Activities 02 and activity coefficient 72 of the sugar referred to the pure sohd.
(d) Activities 02 and activity coefficients 72 of the sugar referred to mole fractions
at infinite dilution.
(e) Activities oj' and activity coefficients 72' of the sugar referred to molaUty
at infinite dilution.
FIGS. 143O and 6. Comparison of Activities and Activity CoefBcients for Different
Standard States.
' 6 3 5
TABLE XXX
ACTlVITy OF ^^ A T E K I N SUGAR SOLUTIONS AT 25°C 1
Vapor
Mole Fractions pressure
Mohdity C i j H j j O i i H2O mmHg
m Ni iVi Pi ffli 71 log 71 Ni/Ni
0 0 1.0 23.756 1.0 1.0 0 00
0.1 0.00180 0.99820 23.714 0.99823 1.0000 0.0000 554.56
0.4 0.00715 0.99285 23.585 0.99280 0.99995 -0.00002 138.86
1.0 0.01770 0.98230 23.302 0.98089 0.99856 -0.00063 55.497
2.0 0.03478 0.96522 22.762 0.95816 0.99269 -0.00319 27.752
3.0 0.05127 0.94873 22.166 0.93307 0.98349 -0.00723 18.505
4.0 0.06722 0.93278 21.521 0.90592 0.97120 -0.01269 13.877
5.0 0.08263 0.91737 20.846 0.87750 0.95654 -0.01930 11.102
6.0 0.09755 0.90245 20.20 0.85031 0.94222 -0.02585 9.251
6.18 (Sat) 0.10018 0.89982 20.08 0.84526 0.93937 -0.02716 8.9820
Solution: (a) At the low pressures involved it may be assumed that the fugacities
of the water vapor/i are equal to partial pressures pi.
(6) T?he activities Oi of the water
'70 n 1 1 \ 1 I 1 1 in the solution are obtained by di-
viding j)i by 23.756 the vapor pres-
60 sure and fugacity of the pure water
at 25''C which is chosen as the
50 standard state. At the low pressures
involved the pressure designation of
40 the standard state is of no conse-
^1
quence since pressures of this order
N,230 have a neghgible effect on the fugac-
ity of the liquid. The activity co-
20 efficients 71 are equal to ai/Ni.
(c) The activity coefficients of
10 the sugar 72 are determined by
-^6| ^6 I -'14 I A^ I A, graphical integration of Equation
(68). Values of -log 71 and Ni/Ni
8 12 16 20 24 28
from Table XXX are plotted in Fig.
-log (TiXlO"')
144. Since the fugacity of the sugar
FIG. 144. Evaluation of the Activity Coeffi- in the saturated solution is equal to
cients of Sucrose that of the pure solid, the activity
02 of the sugar in the saturated so-
lution is 1.0 referred to the pure solid. The corresponding activity coefficient 72 =
l.O/ATj = 9.9820 where Ni/Ni = 8.9820. These values fix the lower limit of the inte-
gral which may be written
rNi/N2 Ni I
log 72 = - I —- d log 71 + log 9.9820
^8.9820 AT
>/8.9820 Ni
The incremiptal evaluation of the integral is indicated in Fig. 144, and the resultant
values of 72 are given in Table XXXI. Corresponding values of a^ are obtained by
multiplying A'^2 by 72.
(d) In Fig. 145 values of 72 are plotted as ordinates on a logarithmic scale against
y .. \
CHAP. XIV] IONIC ACTIVITIES 637
Ni and the resulting curve extrapolated to N2 = 0. Values of oj and ai' are then
calculated from Equation (72). The corresponding activity coefficients are by defi-
nition yi = ai/Ni and 72' =a^'/m. These results are summarized in Table XXXI
and Fig. 145.
TABLE XXXI
ACTIVITIES OF SUGAR IN AQUEOUS SOLUTIONS AT 2 5 ° C
m N^ 72 az 02 72 ai' 72
0 0 3.2800 0 0 1.000 0 1.000
0.1 0.00180 3.2899 0.00592 0.00180 1.000 0.1000 1.000
0.4 0.00715 3.3696 0.02409 0.00734 1.027 0.4074 1.019
1.0 0.01770 3.8539 0.06821 0.02080 1.175 1.155 1.155
2.0 0.03478 4.7222 0.16424 0.0501 1.400 2.779 1.390
3.0 0.05127 5.7860 0.29665 0.0904 1.764 5.020 1.673
4.0 0.06722 7.0175 0.47172 0.1438 2.139 7.983 1.996
6.0 0.08263 8.3880 0.69310 0.2113 2.557 11.730 2.346
6.0 0.09755 9.7225 0.94843 0.2899 2.964 16.05 2.675
6.18 0.10018 9.9820 1.00000 0.3049 3.043 16.92 2.738
From inspection of Fig. 145 it is evident that the activity coefficients 72 and yi
are widely different in numerical values but are related by a constant factor. The
coefficient 72' approaches 72 at low
concentrations but is not related to 10 r
9
the other two coefficients by a con- 8-
stant factor. 7-
It may be noted that the graphical
integration shown in ,Fig. 144 be-
__^J
comes uncertain at low concentra-
tions .where Ni/Nt approaches
infinity. Special graphical methods T2
have been developed by Lewis and
Randall" to circumvent this diffi- yj^
culty and permit accurate integra-
^72"
tions with zero concentration as one
limit.
Although determination of
0 0.02 0.04 0.06 0.08 0.10
activities and activity coefH-
cients from vapor-pressure data FIG. 145. Activity Coefficients of Sucrose
is most direct, in principle the in Aqueous Solutions at 25 °C.
results are frequently less accu-
rate than those obtained by other methods involving different types of
equilibrium.
Lewis and Randall'^* discuss in detail the evaluation of activity coeffi-
cients from data on solubilities, distribution coefficients, freezing points,
boiling points, and electromotive-force measurements.
Ionic Activities. The expression of activities in solution of electrol5rtes
is complicated by the dissociation of the components into ions. Tbia
effect may be disregarded, and the apparent molecular activity, frequently

termed the activity of the undissociated electrolyte, may be determined
from vapor-pressure data by the procedure of Illustration 9. For ex-^
ample, the molecular activities of sodium chloride in water may be de-
termined in this manner from data in the literature. It is found, how-
ever, that the molecular-activity coefRcient 72 = a2/Ni varies widely
with concentration and approaches zero instead of a constant value at
zero concentration. This behavior results from the fact that the effec-
tive mole fraction of the binary solute is increased by dissociation of
one molecule into two ions, each of which behaves as a solute, and the
actual molal concentration of the solute approaches twice the stoichio-
metric value based on the undissociated salt. Thus, for determining
molecular activities in a binary solution Equation (64) may be written
rii 55.51
dlog as = d\oga\= '•— dXogai ' ' ' (78)
rii JUi
Actually, however, the number of moles of effective solute is equal to

n-iiy- +»'+) or n^v^, where ;'_ and v+ are the numbers of negative and
positive ions formed by the dissociation of one molecule of the solute
and v^ = V--\-VJ^. On this basis an expression may be written for what
is termed the mean ionic activity a±2-
n\ 55.51
d log a±2 = d log ai = '-— d log oi (79)
n2V2± m-iVi,
By combination of Equations (78) and (79) it follows that
.; u ^' '-• a, = kia^r- = ^±^ .:". ; ' ,' '. (80)
where v^, v_ = number of positive and negative ions, respectively,

formed from the dissociation of one molecule,
K = the equilibrium constant of ionic dissociation

fc = a proportionality factor, the value of which is deter-
mined by the standard states chosen for a^ and a±2
Rearrangement of Equation (80) gives, '*'•'" ;/:"^''"'= '• f4- v!>-'«•>i
I"* - -^ (80a)
- \kK J
The standard state of ionic activity is conventionally based upon the
mean ionic molality which takes the same form as Equation (80a); thus:
m^2 = mÛ+pl-y^ (81)

CHAP. XIV] IONIC ACTIVITIES 639
where Wj = stoichiometric molality

For example, calcium chloride dissociates into one positive and two nega-
tive ions or v+ = 1, ><_ = 2, v^ = 3, and m^2 = m2(4)*.
The standard state for the expression of the activities of electrolytes
is so chosen that 0*2 = w±2 as infinite dilution is approached. Thus,
j_
' Lim a±2 = 'm^2 - "mii/^v"-)"^ (82)
m±2=0 ""''fl
The mean ionic-activity coefficient 7^2 is defined by "
..• •-7.2 = ^ . . . . . . ^ . \: • (83)

m^2 ; : :
I t is evident that 7±2 = 1.0 at zero concentration.
For the expression of the apparent molecular activities of electrolytes,
a special standard state is commonly selected in order that the following
relationship shall be valid.
V'i = a'," = (a.2)'- = (^//1"')a2 '^ • -" ''''' - (84)
where ctj" = the apparent molecular activity referred to what may be
termed the electrolytic standard state having a fugacity/^"'.
Oi = the apparent molecular activity referred to the pure
component having a fugacity/2. The apparent molecular
activity coefficient y'^" may be defined as follows:
r 7 2 " = — •"'- (85)

mi
It should be noted that this definition of standard state does not require
that 7 2 " be imity at infinite dilution.
By combination of Equations (81), (82), and (84) mean ionic activities
may be determined from vapor-pressure measurements which permit
calculation of the apparent molecular activity 02 referred to the pure
component or some other convenient state. Thus, / ••'
fc27=.2=^^- ;. (86)
mi > U-îiiHth
where
• '•. - " • • 1
h = ^^ '-Y (87)
Values of /c27±2 from Equation (86) are plotted against m2 and extrapo-
lated to zero molality where by definition 7±2 = 1-0. This intercept is
equal to ki from which values of 7±2 at all concentrations may be cal-
culated by Equation (86). Corresponding values of a^2 and a'/' are

then obtained from Equations (83) and-(84).
It is of interest to develop the relationship between 7^2 and the ap-
parent molecular-activity coefficients 72" = a'2"/m,2 and 72 = 0,2/^2.
10
8
6
6
4
T2^
'2
1.0 '^\
,0.8 « l
CO l±'i .2
ID"- 3|
•30. 01
So. 0)1
=3
do.
ml
Ml
3 <fa 2/111 1 _ ^ ^ —
0.1
0.08
0.06
0.05
0.04
0.03
0.02
0.01
0 1 2 3 4 5 6
Molality, rrii '. '
FIG. 146. Activity Coefficients of Aqueous Sodium Chloride Solutions at 25°C.
By combination of Equations (81), (82), and (84) with the equations
of definitions of the coefficients:
T2" = 7 l t < * - ' ' ( v > l - ) (88)
It is evident from Equations (88) and (89) that both 72"and 72 must
approach zero at zero concentration where 7*2 = 1.0 and m2 = 0. The
CHAP. XIV] PROBLEMS 641
relationship between the various activity coefficients is shown graphi-

cally in Fig. 146 for NaCl in water at 100°C. Values of 02 and 72 re-
ferred to pure NaCl(s) may be calculated from vapor-pressure data by
the method of Illustration 9. In accordance with Equation (86) the
quantity \ailm2 is plotted against rui and extrapolated to ma = 0.
The value of fca is equal to this intercept, since 7^2 = 1-0. The corre-
sponding values of 7 2 " are obtained from Equation (88). This method
is presented to clarify the relationships between the various standard
states and activity coefficients. Actually it is not a practical means of
evaluating activity coefficients with a high degree of accuracy because
of the difficulty of obtaining precise vapor-pressure data for very dilute
solutions. More accurate experimental methods are discussed in detail
by Lewis and Randall.'^
PROBLEMS 1
1. A mixture of hydrocarbon gases has the following composition in mole per cent:
CH4 38.38
CaHe 7.56
CsHg 7.05
nC4Hio 11.29
nCsHij 35.72
100.00
Calculate the density of this mixture in pounds per cubic foot in the gaseous state
at a temperature of 200°F and an absolute pressure of 400 lb per sq in.
2. Calculate the density in grams per cubic centimeter of the mixture of prob-
lem 1 in the liquid state at a temperature of 100°F and an absolute pressure of
1300 lb per sq in,
3. Calculate the enthalpy in Btu per pound of the mixture of problem 1 in the
gaseous state at the specified conditions. As the reference state of zero enthalpy
use the ideal gaseous state at 60°F for the propane and lighter constituents. For
butane and the heavier constituents use the saturated liquid state at 60°F as the
reference. It may be assumed that the enthalpies of the components are additive
under these reference conditions. Heats of vaporization of the three light gases are
given in Illustration 4. Doss' recommends the following values in calories per gram
at the normal boiling point: n-butane 92.0; n-pentane 85.5.
4. Calculate the enthalpy in Btu per pound of the mixture in the liquid state
at the conditions of problem 2, using the reference states specified in problem 3.
5. From the data of Table XXVIII and Fig. 142 calculate the fugacity in pounds
per square inch of pure ethane gas at a temperature of 200°F and an absolute pressure
of 1500 lb per sq in.
6. Calculate the fugacities of the following gases;
(a) Air at GOT and 100 atm.
(b) Ammonia at 80°C and 40 atm.
{<•) Carbon dioxide at 150°F and a gauge pressure of 2,000 lb per sq in.
7. Calculate the fugacity of the Uquid ethylene in contact with its saturated
vapor at 0°C and 40.6 atm.
642 , THERMODYNAMICS OF SOLUTIONS [CHAP. XIV
8. Calculate the fugacity of liquid chlorine in contact with a mixture of hydrogen

and its own vapors at a temperature of 122°F and an absolute pressure of 1,000 lb
per sq in. The vapor pressure of hquid chlorine at this temperature is 14.1 atm, and
its density is 1.557 g per co at —33.6°C.
9. Assuming that an ideal solution is formed, calculate the fugacities of the com-
ponents of the mixture of problem 1 under the specified conditions.
10. The gas from the converter of a synthetic ammonia plant has the following
composition in mole per cent:
Ns = 20.2
' ^ -•'-•', •'• ' H2=60.8 : • • : : ; . •' , . . , , ; .
NH3= 19.0
Calculate the fugacity of each component of the mixture at 800°F and an absolute
pressure of 4,500 lb per sq in., assuming that an ideal solution is formed.
11. The International Critical Tables give the following data for the lowering of
the vapor pressure of water at 0°C by urea [CO(NH2)2]:
m 100 5
1 1.52
2 1.49
4 1.46 (Po - v)
6 1.45 R--
mPo
10 1.43
where m — molality; Po = vapor pressure of the pure water; p = vapor pressure of

H2O above the solution. The solubiHty of urea at 0°C is 67.1 g per 100 g of H2O.
Calculate and plot as functions of molaUty the activity of the water Oi and of the
urea 02 referred to the solid at 0°C, and estimate the activities of the urea a^ and ai'
referred to infinite dilution. Also calculate and plot the corresponding activity co-
efficients, 71, 72, 72, and 72'.
12. The following data on the vapor-pressure lowering of aqueous Nal solutions
at 100°C are expressed in the terms defined in problem 11:
TO 100 B m 100 J?
0.7 3.25 7 4.91
1.5 3.62 9 5.00
4.0 4.46 11 4.89
5.0 4.67 17 4.25
6.0 4.82 20 3.85
The solubility of NaT at lOO'C is 20.24 moles per 1000 g of H2O.
Calculate and plot as functions of the molality: ai; 71; 02 (referred to sohd Nal);
72; o±2; T±2; 0 2 " ; 7 2 " . -
13. The following data from the International Critical Tables show the partial
pressures in milhmeters of Hg of toluene and acetic acid in solutions at a tempera-
ture of 69.94°C:
CHAP. XIV] PROBLEMS 643
Xi = mole per cent toluene Pi (toluene) Pz (acetic acid)
0 0 136
12.50 54.8 120.5
23.10 84.8 110.8
31.21 101.9 103.0
40.19 117.8 95.7
48.60 130.7 88.2
53.49 137.6 83.7
59.12 145.2 78.2
66.20 155.7 69.3
75.97 167.3 57.8
82.89 176.2 46.5
90.58 186.1 30.5
95.65 193.5 17.2
100 202 0
Calculate the activity coefficients 71, 72 from these data. Plot the activity coeffi-
cients as ordinates on a logarithmic scale against the mole per cent toluene on a
uniform scale and check these cui'ves for consistency with the Gibbs-Duhem equation
at compositions of 10, 30, 50, 70, and 90 per cent toluene.
CHAPTER XV
,j PHYSICAL EQUILIBRIUM
The concepts of fugacity and activity are of particular value in prob-
lems of equilibrium in both physical and chemical processes. Where
only pure components or ideal solutions are involved, such problems are
readily solved by apphcations of the principles developed in the preced-
ing chapters. If a system under consideration involves a nonideal solu-
tion, the evaluation of the activity coefficients of the components is
frequently the most difficult problem. Some direct experimental data
combined with empirical or semiempirical relationships are generally
necessary.
VAPOR-LIQUID EQUILIBRIUM
The evaluation of activity coefficients in solutions of miscible liquids
is of particular importance for correcting data on vapor-liquid equilib-
rium. In working with such mixtures it is convenient to choose the
pure components at the pressure of the solution as the standard states.
With this convention activity is equal to mole fraction in an ideal solu-
tion at constant temperature and pressure. Then,
N, N^fl • ^^^
where
7i = activity coefficient of component 1
Ci = activity of component 1 in solution referred to the pure
component
/ i = fugacity of component 1 in solution
fl = fugacity of pure component 1 at the temperature and pres-
sure of the solution
If it is assumed that the vapors form an ideal solution, the fugacity / i
of Equation (1) is readily calculated as in Illustration 8 (Chapter XIV)
from the composition, temperature, and pressure of the vapor in equilib-
rium with the solution. The fugacity of the standard state is calculated
from the vapor pressure of the pure compound and the total pressure
of the solution by the method of Illustration 7, page 623. In this man-
ner, if sufficient data are available, activity coefficients may be evaluated
in any given system as a function of composition, temperature, and total
pressure. The results may be verified for thermodynamic consistency
644
CHAP. XV] VAPOR-LIQUID EQUILIBRIUM 645
by means of the Gibbs-Duhem Equation (XIV-67). This equation

must be satisfied by activity coefficients at a constant temperature and
pressure.
Application of these methods to binary solutions is complicated by
the fact that when a vapor phase is present such a system possesses only
two degrees of freedom and it is not possible to vary composition at con-
stant temperature and pressure. For rigorous application of the Gibbs-
Duhem equation, fugacities/i determined at a constant temperature and
varying total pressures should be corrected to a constant-pressure basis
by means of Equation (XIV-45). However, this correction is small for
moderate pressure changes and generally may be neglected except when
one is working with solutions near their critical points where partial
molal volumes and pressures are both high. Similarly, at moderate
pressures and small liquid volumes the effect of pressure on the fugacity
of the liquid is negligible, and fl may be taken as the fugacity of the
liquid under its own vapor pressure. Under conditions such that the
vapors may be treated as ideal gases still further simplification is possible
and Equation (1) may be written:
^, = _ ^ = .î!l (2)
where
Xi = mole fraction of component 1 in the liquid phase
2/1 = mole fraction of component 1 in the vapor phase
pi = partial pressure of component 1 in vapor
P i = vapor pressure of pure component 1 at the temperature
I of the solution
IT* = total pressure of the system
Component 1 is usually taken as the component which in the pure state
has the lower boiling point.
In Fig. 147 are plotted the activity coefficients of benzene and toluene
(system I) of water and n-butanol (system II) of isopropyl ether and
isopropyl alcohol (system III) and of acetone and chloroform (system IV)
all at 760 mm of Hg. For the nearly ideal benzene-toluene system, the
activity coefficients of both components are the same and equal to 1.0.
The corresponding vapor-liquid composition curves of these four systems
are plotted in Fig. 148 and the boiling-point curves in Fig. 149.
It should be noted that the activity coefficients of Fig. 147 are re-
stricted to a constant total pressure and therefore correspond to vary-
ing temperatures. Similar curves for the same midrange conditions
* Indicates a range of pressures sufficiently low so that the ideal-gas law is
applicable.
646 PHYSICAL EQUILIBRIUM [CHAP. XV
might be plotted for activity coefficients at constant temperature and

varying pressure. In general, the differences between such plots are
negligible for systems whose boiling points vary only a few degrees with
change in composition. The constant pressure plots are in general more
useful because most vaporization operations are conducted at substan-
10
' / ••
1
.
s .-/
I'
V * /
4)
0
i / 7
?>..
^s^JV
;^
a
t/
c
A
/
\X
r HI
^
" ^ ^ '
>>
1.0
1^^ ^ ^
0.9 .*Benzene - tolu sneJL -^ —
0.8:
0.7 Q
^
0.6 <^ f5f°^
^
0.5 —*\ . ^ ^
0.4
0.3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
a;=Mole Fraction of Lower Boiling Component (*)
1 in
:_ Liquid
T: ;A Phase
T>I ^
FIG, 147. Activity Coefficients of Typical Binary Systems. ''

tially constant pressure. On the other hand, activity coefficients at
constant temperature are more accurately rationalized by the Gibbs-
Duhem equation.
From Fig. 149 it may be noted that the isopropyl alcohol-isopropyl
ether system forms a solution having a minimum boiling point at 78 per
cent ether, whereas the acetone-chloroform system forms a solution
having a maximum boiling point at 40 per cent acetone, and the water-
CHAP. XV] FREE ENERGY OF MIXING 647
butanol system has a minimum boiling point at 75 per cent water.

Such solutions which have maximum or minimum boiling points and
which evolve vapors of the same composition as the liquid are termed
azeotropic solutions or azeotropes. By comparison of Figs. 149 and 147
it is seen that for a system forming an azeotrope of minimum boiling
point the activity coefficients of the separate components are each
greater than unity over the entire range of compositions, whereas for
1.00
13 0.
I
§
S â / \\
o
U '/o
V
B
^ ^x ^^
fy
i> *
if
o>0,
.2-S
8 0,
i i-A VA /^
^
a
•>
S.
r /,A
o 0,
IL 0, /y^
/ V
0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
»i=Mole Fraction of Lower Boiling Component
(*) in Liquid Phase
Fro. 148. Vapor-Liquid Equilibria of Typical Binary Systems.
the system forming an azeotrope of maximum boiling point the activity
coefficients are less than unity. This behavior is general for many non-
ideal systems. Activity coefficients greater than 1.0 are much the more
common.
If the effects of the small variations in temperature are neglected, the
slopes of the curves of Fig. 147 for any system at constant temperature
and low pressure are related to each other -by Equation (XIV-69), and
if one curve is known the other is readily calculated. For the prediction
of both activity curves from a minimum of experimental data a general
relation between activity coefficients and mole fraction is necessary.
Free Energy of Mixing. Thermodynamically consistent expressions
for the activity coefficients of components in solutions are most con-
veniently derived by consideration of the free-energy changes accom-

panying the formation of a solution from its components in their stand-
ard states. This free-energy change of mixiri^ MXM is given by the
following general equation where the standard state is taken as the pure
120 r
0 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 O.QO 1.00
a;=Mole Fraction of Lower Boiling Component (*)
in Liquid Phase
FIG. 149. Saturation Temperatures of Typical Binary Systems at l.OAtm Pressure,
components at the pressure and temperature of the system:
AGjif = SniGi — SriiGi, (3)
where _ _ _ , ; •
SwiGj = niGi-|-?i2G2 + WaGs-|-• • • = G
./'
2niGi = riiGl + n^Gl + U^GI -\
Combining (3), (XIV-53) and (XIV-60) and using Xi for mole fraction
instead of Ni for the liquid phase gives
A(?M = RTÛi In Xi + RTSm In ji (4)
where
Ui In Xi = n-i In Xi -\- n^ In x^ + wa In xs -|-
w
M^
CHAP. XV] FREE ENERGY OF MIXING 649
For an ideal solution all activity coefficients are equal to unity and the
term RTllrn In ji reduces to zero. This term is designated by Scatchard
and Hamer' as the excess free energy G^ of the solution. Thus, the free
energy G of any solution may be written as
G = -SniGl + RTl;ni hi Xi + G'^ (5)
where
G^ = RT'Srii In yi = the excess free energy
The chemical potential of any component is obtained by differentiating
Equation (5), thus for component 1,
dSnilnXi dG^
, Mi = G, = Gl + i ? r — — ^ + — (6)
dni dn
The expansion of the second term combined with the principle of
Equation (XIV-66) gives
RT^- '=RT\liixi + ni—-^ + n,---^ + --- \ = RTlnxi{7)

Then,
; '. .• ixi = Gi = Gl + RTlnxi+—- (8)
dni
By comparison of Equations (8) and (XIV-53) it may be seen that
RT\ny, =- - (9)
dni
Similarly,
/2rinT. = — (10)
Equations (9) and (10) permit derivation of thermodynamically consistent

expressions for the activity coefficients of all components of a solution
from an empirical expression relating the excess free energy to the com-
position of the solution.
It was pointed out by WohP that the empirical equations which have
been commonly used for correlation of activity coefficients represent
special cases of the following general equation for the excess free energy:
• = 'EiihZi?naih + îhiZighZflihi + îhiiZiZhîZiaihii + ••• (11)

RTiSqiXi)
where
Xi = mole fraction of component i
Qi = effective molal volume of component i
' * Zi = effective volume fraction of component i
' G. Scatchard and W. J. Hamer, / . Am. Chem. Soc, 57, 1805 (1935).
»K. Wohl, Trans. Am. Inst. Chem. Engrs. 42, 215-250 (1946).
Subscripts i, h, j , I each m a y correspond to any component of t h e mix-

ture in the terms of the indicated summations.
Uih etc. = empirical constants corresponding t o t h e indicated groups
of components in the summations.
Equation (11) is designated a, four-suffix empirical equation as charac-
terized by the last term. T h e first summation term m a y be considered
as indicating the effect on t h e excess free energy of the interactions be-
tween unlike molecules of t h e components in pairs. T h e second term
indicates the contributions from interactions of groups of three mole-
cules while the third takes into account interactions of groups of four.
If the last summation is omitted t h e remainder is a three-suffix equation,
whereas if the last two terms are omitted it becomes a two-suffix equa-
tion. E a c h additional summation introduces added constants to permit
improved representation of increasingly complex relationships.
E a c h summation in E q u a t i o n (11) represents the sum of all possible
combinations of molecules in unlike groups of the indicated size. Thus,
in a mixture of components 1, 2, 3, • • •, component 1 is first t a k e n as i,
and all possible groups are expanded by successively designating each
component of the mixture as h, j , and I. For example, for a ternary
system the second summation is expanded to 24 t e r m s as follows for all
possible triple combinations among two and three unlike species.
liihiZi^hZiaihj = ZiZiZâiu -\- ZiZizâni -{- ZiZiîUm + ZiZîam

-\- ZiZ'îai23-\- ZiZsgiai3i-{- ZiZ3Z2au2-{- ziZ3Z3au3-\- zîZittin .„•
-^ ZiZiZ2a2u-\- zîZ3a2i3-\- Z2Zîa2ii-\- Zi?^za22s-\- ZiâZiUisi
-\- Z2ZdZ2(t23i + 22^3230233 + Zs^l^ 1^311 + Z^lZiUsu "H 232123^313
-|- 2322210321 + 2322220322 "f" 2322230323 + 232321O33I "f 2323220332 (12)
Since the order of multiplication is immaterial, t e r m s such as 2122O112

and 2i222iai2i are identical a n d m a y be grouped together. T h u s , E q u a -
tion (12) reduces to seven t e r m s : / ,•. , . •
SiA,-2i2A2,-Oa; = 212280112 + 2f2330ii3-}-Zi2|3ai22 + 21222360123

-|-2i2|3ai33 + 212330223 + 222I30233 ; ,A. , ., (13)
T h e other summations m a y b e expanded in a similar manner for a n y

system of any number of components.
T h e effective molal volume q in Equation (11) serves to relate t h e
effective volume fractions 2 to the mole fractions x. Thus,
2l: (14)
, ?2 93
21 + — 2:2-H - a;3 + '
Si Qi
CHAP. XVJ ACTIVITY COEFFICIENTS IN BINARY SYSTEMS 651
and
! 22= (15)
. •: . . a;i + — X2 + — X3-\
21 3i
Activity Coefl5cients in Binary Systems. For a binary system of
components 1 and 2, a three-suffix equation of the form of (11) may be
expanded as follows:
= ZiZ22an + ZiZ2Saiu + Z\z\âxi.2 (16)

RTiqiXi + q^Xi)
or, since Zi + Za = 1.0,
— = hri + - a;2 piZ2[2iQ'i(2ai2 + 80112) + 22«f(2ai2 + 80122)] (17)
This equation may be written with the following abbreviations:

A' = 5i(2ai2 + 3ai22) (18)
B'=g2(2ai2 +80112) (19)
Then
-'IXI + ^^XXIZLI-)B' + Z2A'~\ (20)
V^ ;vuV\ RT
The term In 71 may be evaluated by partial differentiation of Equa-

tion (20) with respect to wi in accordance with the relationship of Equa-
tion (9). This is done by first replacing x and z by n by means of the
following relations:
Qi
rii —
xi = — , — ; X2 = — • — ; zi = — ; 22 — (21)
ni + n2 ni + n2 ni + ^^2 n, + ^-^ n^
r gi 91
•• , a;i-|-a;2 = 1.0; 2i-|-Z2=1.0
Also on the molal basis, = G^

ni + n2
By substitution in Equation (20),
M)t si ^p nln2B'+ninl(-) A'

\qi/ (22)
I n i + -712)
\ gi /
Upon differentiation with respect to ni, where A = 2.303A' and

B = 2.303B',
nl(^^y[-n,A+2n.B(^^ynA{f)]
^^" = log 71 = -^^^^
= I I IM - V l =
-
7 rri ^ ^ ^ ^ (23)
2.303dni{RT)
(
Replacing values of Zi and Za gives
log 71 = z| U + 2(3 ^ - A \ ^ (24)

Similarly, :^ ,• •
logy, = zllB + 2(A^-B\^ (25)
These general equations developed by WohP involve three constants,

A, B, and gi/gs, which must be empirically determined for each system
under consideration. By the use of various assumptions regarding the
ratio of the effective molal volumes 32/51 the number of empirical con-
stants may be reduced to two.
Margules' in effect assumed that qi/qi = 1.0. With this assumption
Equations (24) and (25) reduce to the familiar Margules equations as
modified by Carlson and Colburn.*
logyi = xllA + 2{B-A)xi]= (2B -A)xl + 2{A - B)xl (26)
log 72 = z\ [S + 2(A - B)X2] = (2A - B)x\ + 2{BÂ)xl (27)
Scatchard' and co-workers'^'') took the effective molal volumes as equal
to the actual molal volumes of the pure components. On this basis
Qi/qi = F2/F1, and Equations (24) and (25) reduce to what is termed
the Scatchard-Hamer equations. Thus,
log 71 = zllA + 2(^By^- A V ] • , (28)
log 72 = z\[B + 2(Ây^- BV] (29)
In these equations 2 represents the volume fraction based on the vol-

umes of the pure components and is the true volume fraction if volume
changes in mixing are negligible.
If it- is assumed that 51/52 = A/B Equations (24) and (25) reduce to
' M. Margules, Sitzber. Akad. Wiss. Wien. Math, naturw. Klasse, II, 104, 1243-
78 (1895).
* H. C. Carlson and A. P. Colburn, Ind. Eng. Chem., 34, 581 (1942).
» G. Scatchard, Chem. Rev., 8, 321 (1931). \
6 G. Scatchard and S. S. Prentiss, J. Am. Chem. Soc, 56, 1486 (1934). i
' G. Scatchard and W. J. Hamer, J. Am. Chem. Soc., 57, 1805 (1935). '
CHAP. XV] ACTIVITY COEFFICIENTS IN BINARY SYSTEMS 653
the form developed by van Laar^-^ as rearranged by Carlson and

Colburn.^
log7i = 42i = T l — — - , (30)
(|xx + x.J
log y, = Bzl = ^ — - (31)
(.x+fx.y
Differentiation proves that Equations (24-31) satisfy the general rela-
tion of Equation (XIV-69). The constants A and B are characteristic
of each particular system. In the van Laar equations (30) and (31)
it is evident that A is the terminal value of log 71, where xi = 0, and B
is the terminal value of log 72, where X2 = 0.
Equations (24-31) are basically restricted to conditions of constant
temperature. However, as previously pointed out, little error is gen-
erally encountered in applying them to activity coefficients at constant
pressure in close boiling systems.
If the validity of the van Laar Equations (30) and (31) is accepted,
determination of 71 and 72 at a single known composition permits evalu-
ation of A and B and calculation of the complete 7 curves. Measure-
ment of a single set of equilibrium liquid and vapor compositions, to-
gether with knowledge of the vapor pressures of the pure components,
suffices for calculation of 71 and 72 from Equation (2). Where an azeo-
trope is formed the composition of only one phase need be measured
since the liquid and vapor compositions are the same. By rearrange-
ment of Equations (30) and (31) the constants A and B are obtainec?
directly.
A = log7Jl + ^ - ^ T (32)
L a;ilog7iJ
£=iog7.ri+^r (33)
L Xi log 72J
Illustration 1. The azeotrope of the ethanol-benzene system has a composition
of 44.8 mole per cent ethanol with a boiUng point of 68.24°C at 760 mm Hg. At
68.24°C the vapor pressure of pure benzene is 517 mm Hg, and that of ethanol is
506 mm Hg. Calculate the van Laar constants for the system, and evaluate the
activity coefficients for a solution containing 10 mole per cent ethanol.
Solution: Ethanol is designated as component 1. At the azeotropic composition,
from Equation (32), since y = x,
71 = fSt = 1-502; log7i = 0.177
y,^m = lA7; log 72 = 0.167
8 J. J. van Laar, Z. phys. Chem., 72, 723 (1910).
»J. J. van Laar, Z. phys. Chem., 185, 35 (1929).
From Equation (32),

A = 0 . 1 7 7 U + ^0-^^2)(0.167) = 0.830
= 0.177] 1
(0.448) (0.177).
Sidiilarly, from Equation (33), »' '
B = 0.577
For a solution containing 10 per cent ethanol, from Equations (30) and (31),
(0.830) (0.90)- ^ , , = 4.14
r(0.830)(0.10) ^^p-l^
L 0.577 ^ J ,
I,.,... log 72 = 0.0109; 72=1.026 :.,„,,,;,,•:.,.., :!^ (
By this procedure the activity coefficients of the entire range of compositions may
be calculated.
It was pointed out by Carlson and Colbum^ that the van Laar con-
stants may be evaluated without any liquid-vapor equilibrium composi-
tion data if a series of isothermal vapor-pressure or isobaric boiling-point
measurements are available for solutions of known, composition over the
entire composition range. It is frequently desirable to work from such
data because of the difficulty of securing a reliable analysis of vapor
mixtures and the ease of making up liquid solutions of a desired com-
position. Since, in Equation (2),
PI = TT* - p; = TT* - y,x^l / V .,; , , (34)
as x% approaches 1.0 and 72 is approximately 1.0, ' jn ^ s1',}« . t<!!>v
* ™ p* * ' '
Lim 7i = - ^ -. (35)
Similarly,
r*-x^P\ , '
Lim 72 = —,— (36)
X2=0 Xir2
In working from isothermal total vapor-pressure data the observed

values of TT* are substituted in Equation (35) with the corresponding
mole fractions, and vapor pressures and apparent values of 71 are cal-
culated. By plotting log 71 against Xi and extrapolating to a;i = 0 the
value of A is determined. Similarly, when working from isobaric boiling-
point data, values of the vapor pressures of the pure components at the
observed temperatures are substituted in Equation (35) with the cor-
responding total pressure and mole fractions. The calculated apparent
values of 71 are extrapolated to a;i = 0 as before. This latter procedure
involves the assumption that the activity coefficients are independent
of temperature over the range covered.
Illustration 2. In Table XXXII are data for the boiling points of ethanol-
benzene solutions at 750 mm Hg and the vapor pressures of pure ethanol and ben-
CHAP. X V ] ACTIVITY COEFFICIENTS IN BINARY SYSTEMS 655
zene at these temperatures. Calculate the van Laar constants from these data,
assuming the activity coefficients to be independent of temperature.
Solution: The values of apparent 71 and 72 shown in Table XXXII are calculated
from Equations (36) and (36) and plotted in Fig. 150 against xi. Extrapolating the
71 curve to ii = 0 gives the value of 71 = 5.7 01 A = log 71 = 0.756. Similarly,
72 is determined to be 4.06 by extrapolating the 72 curve to Xi = 1.0 and B = log 72
= 0.608. These constants are in fair agreement with those derived from the
azeotrope composition.
TABLE XXXII
BOILING POINTS (DF ETHANOL-BENZENE SOLUTIONS AT 750 MM HG
Bnil-
ing Vapor Pressure of Appar- Appar- True True
Mole Fraction Point Pure Components ent 7: ent 72 71 72
Ethanol Benzene Mm of Hg at ta'C
Xl Xi <B°C Ethanol Pi Benzene P2
0.0 1.0 79.7 804 mm 750 mm — 1.00 5.70 1.00
0.04 0.96 75.2 671 648 4.74 1.16 4.20 1.00
0.11 0.89 70.8 560 562 4.07 1.38 3.66 1.05
0.28 0.72 68.3 507 618 2.66 1.63 2.11 1.16
0.43 0.57 67.8 497 609 2.16 1.84 1.58 1.34
0.61 0.39 68.3 507 618 1.77 2.18 1.22 1.84
0.80 0.20 70.1 645 549 1.47 2.86 1.05 2.64
0.89 0.11 72.4 698 592 1.285 3.33 1.02 3.14
0.94 0.06 74.4 650 632 1.165 3.64 1.00 3.58
1.00 0.0 78.1 750 711 1.00 — 1.00 4.06
Data from "International Critical Tables," III, pp. 217-21, 313.
Improved accuracy in the determination of van Laar constants from
boiling-point or total-pressure data is obtained by first calculating the
constants from Equations (34) and (35) as in Illustration 2. Values of
72 are then calculated from the van Laar equation with these approxi-
mate values of A and B substituted j into Equation (34) instead of
72 = 1-0 being assumed as in the first approximation. Similarly, cal-
culated values of 71 are substituted in Equation (35), and the extrapola-
tion of the two equations to zero concentration is repeated to obtain
second approximations of A and B- In the second approximation of
Illustration 2 the same values of A and B Avere obtained. The corrected
values of 71 and 72 are plotted in Fig. 150.
The preferred method for evaluating the constants of the van Laar
equation depends on the nature of the system and the data available.
Reliable liquid-vapor composition data in the dilute ranges are most
desirable. A generally applicable method for obtaining such data by
equilibrium condensation has been developed by Colburn, Schoenborn,
and Shilling.'" Where such data are not available the composition of
i» A. P. Colburn, E. M. Schoenborn, and G. D. ShiUing, Ind. Eng. Chem., 36, 1250
(1943).
the azeotrope generally furnishes the best basis if the azeotrope is in the
middle of the composition range, between xt = 0.25 and zi = 0.75. If
the azeotrope composition is outside this range, boiUng-point or vapor-
pressure data provide a better basis.
Once the activity coefficients of a system are established Equation (2)
permits calculation of equilibrium compositions and vapor-pressure and
6
Vl°
W
.2 3 J^ .p
X
o
^WW
w
o
*00.2 0.4 0.6 0.8 1.0

a;pmole fraction of ethanol
FIG. 150. Apparent and True Activity Coefficients of the Ethanol-Benzene
System at 750 mm Hg.
boiling-point curves at low pressures where the ideal-gas law may be
assumed. For a binary system of components 1 and 2,
m
Combining (2) and (37), since x^ = 1.0 — xi, gives
m
Combming (2) and (38) gives
PI
^t7i/ x^ \
f Pi
2/1 = (39)
x,Phi+{l-x,)Ph2
Equations (38) and (39) permit direct calculation of isothermal vapor-

pressure and Vapor-liquid composition curves. For deriving the more
CHAP. XV] PARTIALLY MISCIBLE LIQUIDS 657
valuable isobaric boiling-point curves and the corresponding isobaric

liquid-vapor composition curves it is necessary that Pi and Pi be
known as functions of temperature. Equation (39) may then be solved
by a trial and error method for the boiling points and vapor compositions
corresponding to selected liquid compositions. As a first approximation
the boiling point is assumed; this fixes a trial value of the relative volatility
Pi/Pi which is substituted in Equation (39), and a first approximation
2/1 is calculated. Then, from Equation (2), Pi is equal to ynr*/xiyi,
and a corrected boiling point is obtained as the temperature correspond-
ing to this value of Pi. Since Pi/Pi varies but little with tempera-
ture, the value of yi calculated from Equation (39) with the corrected
temperature being used is generally satisfactory.
Illustration 3. From the van Laar constants of Illustration 1, calculate the boil-
ing point at 750 mm Hg of a solution containing 28 mole per cent ethanol in ben-
zene and the equilibrium composition of the vapor evolved. The vapor pressures of
the pure components may be estimated from the data of Illustration 2.
Solution: As a first approximation assume is = 70°C. Then,
• L ^ ^ = o.99
• ' , • PJ 548
From Equations (30) and (31), the values of A and B from Illustration (1) give
71 = 2.19; 72 = 1.18
By substitution in Equation (39), a first approximation of 2/1 is obtained:
Substitution of this value in Equation (2) gives

(0.416) (750) ' „ „
. ^ ^ * = ^^'"^^"'"'=(0.28) ( 2 . 1 9 ) - = ' ° ' ' " ' " ^ ^
This vapor pressure corresponds to a temperature of 68.4°C, the corrected boiling
point. At this temperature P* = 520 mm: Pf/PJ = | | § = 0.978. Substitution in
Equation (39) gives
0.706 .,,„
2/1 = = 0.413
' ^ 1.706
In this manner complete boiling-point and liquid-vapor composition curves may be
derived. The differences between the values calculated in this manner from the
data of Table XXXII are attributable to the inadequacy of the van Laar equation
and the uncertainty of experimental data of this type.
Partially Miscible Liquids. As the chemical dissimilarity of the com-
ponents of a solution increases, the deviations from unity of the activity
coefficients increase until finally miscibility is incomplete and two liquid
phases result. This behavior is illustrated by the aqueous solutions of
the alcohols. The ethanol-water system shows considerable deviation
658 PHYSICAL EQUILIBRIUM ; [CHAP. XV
from ideal behavior with activity coefficients greater than 1.0. The
water-n-propanol system has higher activity coefficients but complete
miscibility at its normal boiling points. The water-butanol system
has still higher activity coefficients and is only partially miscible at its
normal boiling temperatures. . ix :';/»:;>; • - rfb
The properties of the water-butanol system at atmospheric pressure
are showii in Figs. 147, 148, and 149. It may be noted that two liquid
phases exist in the composition range from a to 6 corresponding to 66
to 98 mole per cent water. Between these limits the fugacity of each
component is constant and equal in the two phases. As a result the
boiling point is constant, and the vapor evolved is of constant composi-
tion c. The apparent activity coefficients in this range are inversely
proportional to the mole fractions based on the combined phases.
Although the vapor evolved from any liquid mixture of composition
between a and b is of constant composition c, the liquid and vapor
compositions are equal only at point c, which may be termed the hetero-
azeotropic composition. However, in each of the individual phases the
activity and composition are constant as indicated in Fig. 147. Since
the fugacities in the two phases are equal at equilibrium, in the region of
the two liquid phases,
«i/i = a[n' (40)
where the primed quantities represent one liquid phase and the unpriraed
the other. Since the standard state is the pure liquid component,
/t=/rand ,..• :^r^.--^ ^
XiTi = XJTI or -^ = — (41)
The ratio xi/xj is the distribution coefficient of component 1.

Carlson and Colburn^ pointed out that Equation (41) permits evalua-
tion of the constants of the van Laar equation for a binary system from
liquid-liquid solubiUty data. By combining Equations (41) with (30)
and (31), the follo^\ing equations were developed which may be solved
simultaneously for A and B:
'4: Hi:
A _ \X2 X2/\\0gX2/X2/
(42)
B Xi x'l 2xix[ log x[/xi
Xi X'i XiX'i log X^/x'i
, • \o&,x'/xi
A= J L_Z ^ (43)
(-fi) (-©•
CHAP. XV] TERNARY SYSTEMS G59
I t was found that the van Laar equation with constants derived from
Uquid-solubility data in this manner did not accurately represent activity
coefficients in solutions far removed from the two-phase region. For
example, for the water-re-butanol system the constant A based on liquid
solubility was found to be 0.334, whereas that based on vapor-composition
data in the dilute range is 0.61. Similarly, 5 is 1.60 from solubility and
1.34 from vapor compositions. Equations (42) and (43) are useful prin-
cipally as a means of obtaining rough approximations for activity co-
efficients where no vapor-composition data are available.
As the chemical dissimilarity of the components of a system increases,
the deviation of the activity coefficients from unity increases and the
range of mutual solubility decreases. This effect corresponds to length-
ening the horizontal portions between a and h on the curves of Figs. 147,
148, and 149. As complete immiscibility is approached, point a ap-
proaches zero, point b approaches 100 per cent composition, and the ac-
tivity coefficients and v£tn Laar constants become very large. To serve
as a rough guide to the relative magnitudes of the activity coefficients
corresponding to various degrees of miscibility Colburn" prepared
Table XXXIII from Equations (42) and (43) for symmetrical systems
in which the mutual solubilities of the two components are equal.
, , TABLE XXXIII
ACTIVITY COEFFICIENTS AT VARIOUS MUTUAL SOLUBILITIES IN SYMMETRICAL SYSTEMS
Solubility Limits Van Laar Constants 7 at 7 at
X, Mole Fraction A and B s=0 x = x,
0.5 0.875 . - i ,,;;7;6.' v l ^ i.i ....ilT
0.2 1.0 lao 6.0
' ' 0.1 1.2 15.8 9»
0.05 1.43 « 2b
0.02 'i ' • ' 1.77 ' ' • ! » • ' ! ' '•' to-
il]'-: :'. - 0.01 >•; 2.03 106 94
Ternary Systems. Few reliable data are available from which activity
coefficients in ternary systems can be calculated. The general problem
of the correlation of activity coefficients in such systems has been re-
viewed by Wohl^ who has developed a series of equations analogous to
those for binary systems which have already been discussed. The ac-
curacy of any of these equations depends to a large extent on the number
of experimentally determined constants involved. The most useful form
requiring minimum data is probably the ternary van Laar equation
which involves only six constants, all of which can be evaluated from
data on the three individual binary systems. The ternary van Laar
" A. P. Colburn, private communication (1942). •. , - , , , - ; . ; , . .•
equation for a system of components 1, 2, and 3 was developed by Wohl

by applying the van Laar type of relationship that qi/qi = Ai^^/A-^î
to a two-suffix form of Equations (13-15). Thus,
..x,.,(^)V.iA...(fe-;)' ' •:
log Ti =
a;2X3-;^ -. ( Ai-2 + ^1-3—Aa-a-; I

• ' + Ai_2Ai_3\ ^3-1/ .
/ _^ A2_, la-iV
V 4l_2 .4l_3/
In equation (44) the constants î_2 and A2_i correspond to the van
Laar constants A and B of the binary system of components 1 and 2.
Similar equations for log 72 and log 73 are obtained by changing the sub-
scripts on all terms in Equation (44) in accordance with the following
schedule of rotation:
Subscript in Equa-
tion (44) for Subscripts in Equation for
log 71 log 72 log 73
1 2 3
2
3 1 * . 5
Thus, in the equation for log 72, Xi of Equation (44) becomes x^; x^ be-
comes 0:3; X3 becomes :ci; and 41_2 becomes 42-3 etc. Similarly, in the
equation for log 73, Xs becomes X2, and A1-2 becomes A^î.
By means of Equation (44) complete activity coefficient relationships
may be estimated from the coefficients of the three binary systems which
in turn may be approximated from the binary liquid-solubility data by
Equations (42) and (43). Colburnî tested this general method on the
methyl cyclohexane-n-heptane-aniline system and found good agree-
ment with ternary solubility data. However, it is impossible to predict
the errors which may be encountered in other systems.
The most generalized equation for ternary systems recommended by
Wohl is the four-suffix form and involves 14 constants, 11 of which may
be evaluated from data on the three binary systems.
Knowledge of the activity coefficients in a ternary system permits cal-
culation of the distribution of any one of the components between the two
liquid phases in a region of partial miscibility by means of Equation (41)
which is valid for any component regardless of the number of others
CHAP. XV] EFFECT OF TEMPERATURE ON COEFFICIENTS 661
present. These relationships are of fundamental importance in problems

of solvent extraction and extractive or azeotropic distillation.
Effect of Temperature on Activity Coefficients. An equation for the
variation of the activity coefficient with temperature at constant com-
position results from combination of Equation (XIV-45) and (1):
/d In 7 i \ ^ / 3 1 n / i \ _ / a i n / ; \
\ dT J^ \ dT )^ \ dT A
K - Hi) - (Ht - Hj) (li-H?)
(45)
RT^ RT^
The quantity HI-H" is the partial molal enthalpy relative to the pure
components at the temperature of the solution or the differential heat
of solution. If Hi-Hi is positive, y is reduced by an increase in tempera-
ture. Carlson and Colburn^ observed that in general if the activity
coefficients of a system are greater than 1.0 the heat of solution is nega-
tive, whereas if the coefficients are less than 1.0 the heat of solution is
positive. Thus, in either case increasing the temperature causes the
activity coefficients to approach 1.0. Conversely, lowering the tem-
perature tends to increase deviation from ideal behavior, and many sys-
tems miscible at high temperatures separate into two phases at low
temperatures.
Use of Equation (45) for calculating the effects of temperature is com-
plicated by the fact that values of (1 — Hi) vary considerably with tem-
perature and are rarely known over any extensive range.
An approximate method for estimating the effect of temperature on
activity coefficients is obtained by extension of a method suggested^by
Carlson and Colburn for estimating the effect of temperature on azeotrope
composition. It is assumed that the ratio of the activity coefficients
7 I / T 2 is independent of temperature at a given composition. It follows
from Equation (2) that at the azeotropic composition where Zi = yi,
a Pi
P\ (46)
The variation of the azeotropic composition with pressure or temperature
is estimated from Equation (46) by plotting 71/72 against Xi from the data
at the known temperature. On the same scale the ratio of the vapor pres-
sures Pl/PX is plotted against temperature. The azeotropic composition
at a selected temperature is then the composition at which 71/72 is equal
to the value of PyP*i at this temperature. This procedure is demon-
strated in Fig. 151 developed by Carlson and Colburn* for the ethyl
acetate-ethyl alcohol system. The 71/72 curve was derived by evaluat-
ing the van Laar equation from the composition of the azeotrope at
atmospheric pressure and 71.8°C. To estimate the composition of the

azeotrope at 91.4°C the broken lines on the diagram are followed from
this temperature vertically to the Pl/P\ curve, then horizontally to
the 71/72 curve, and vertically to the composition axis giving a value of
Temperature, "C
0 20 40 60 80 100
L2 1
\
<—
1.0
0.8
t \\ p*
\
TiV
0.6
0.4
20 40 60 80 100
Mole Per Cent Ethyl Acetate
Fio. 151. Effect of Temperature on the Composition of the Ethyl
Acetate-Ethyl Alcohol Azeotrope.
48 mole per cent ethyl acetate. In Table XXXIV are values estimated^
in this manner compared with the experimental measurements of
Merriman.i^
The total pressures P^ corresponding to the various azeotrope compo-
sitions and temperatures may be approximated by the following em-
pirical equation,
P^(a;iPi + X2F2) , ,.^
P.= (47)
{x\P\ + x^PD
where
Pz, Pi, P2 = vapor pressures of the azeotrope, component 1, and
component 2 at temperature t
P'i, P'l, P2 = vapor pressures of the azeotrope, component 1, and
component 2 at the reference temperature t'
Xi, X2 = mole fractions of components 1 and 2 in the azeotrope
at temperature t
x'l, x'i = mole fractions of components 1 and 2 in the azeotrope
at temperature t'
From the azeotrope compositions and pressures calculated in this
manner corresponding constants for the van Laar equation may be
derived by means of Equations (32) and (33). In this manner coipplete
u R. W. Merriman; / . Chem. Soc, 103, 1801-16 (1913).
CHAP. XV] HIGH-PRESSURE VAPOR-LIQUID EQUILIBRIUM 663
activity coefficient and vapor-liquid composition curves may be esti-

mated over limited ranges of temperatures or pressures. This method
has not been sufficiently explored to permit estimation of the errors
which may be encountered.
- TABLE XXXIV
EFFECT OF PRESSURE ON THE ETHYL ACETATE-ETHYL ALCOHOL AZEOTROPE*
mpera- Mole Fraction Ethyl Acetate Total Pressure
rc, °C P*/P* Calculated Measured nrniHg
ia7 0.598 0.787 0.734 77.4
40.5 0.725 0.677 0.660 220
^3 0.827 0.602 ' 0.601 423
71.8 0.921 0.539 0.539 760
83.1 0.994 0.498 0.490 1121
91.4 1.049 0.480 0.451 1476
HIGH-PRESSURE VAPOR-LIQUID EQUILIBRIUM

It has been pointed out that in any heterogeneous system it is a gen-
eral criterion of equilibrium that the chemical potential ixi of any com-
ponent must be equal in every phase in which it is present. For systems
restricted to constant temperature this requirement of equality of chem-
ical potential may be replaced by the requirement of equality of the more
convenient function, fugacity. Since fugacity is equal to the product
of activity times the fugacity of the standard state, the criterion of
equilibrium at constant temperature is expressed by the following
equation.
(ai)L(/;)L = iai)Un)L' = (ai).(fi). = (ai).(/Ds- (48)
where
(ai) L, (fli) L', etc. = activities of component L in the phases L, L', v, s—
(fi)ii(/i)i') etc. = standard states for activity of component 1 in
the phases L, L', v, s--
For a system comprising a single liquid phase L and a vapor phase v
Equation (48) may be written
where Ki is termed the vaporization equilibrium constant of component 1.

It is evident that the value of the equilibrium constant is determined
by the definitions of the standard states for both phases. It is conven-
ient to choose as the standard state for each component in the vapor
phase the pure component in the gaseous state at the temperature and
pressure of the system. Similarly, the standard state for each compo-
nent in the liquid phase is chosen as the pure component in the liquid
state at the temperature and pressure of the system. These definitions

of standard states have the advantage that where the Uquid and vapor
approximate ideal solutions activities are equal to mole fractions. It
follows from these definitions of the standard states that Ki is a function
of both the temperature and pressure of the system as well as of the na-
ture of the component.
The vaporization equilibrium constant as defined by Equation (49)
provides a thermodynamically rigorous relationship between the activ-
ities of a component in vapor and Uquid phases which are iii equilibrium.
However, before this relationship is useful it is necessary to relate ac-
tivities to mole fractions through the introduction of activity coefficients.
Thus, in the vapor phase,
(aO» = ^ = <^«2/i (50)
\Jl)v I
where ' '
(ai)s = activity of component 1 in the vapor phase
t/i = mole fraction of component 1 in the vapor phase
0V1 = activity coefficient of component 1 in the vapor phase '
The activity coefficient ^ is dependent on the properties of both com-
ponent 1 and the other components with which it is mixed and also upon
the temperature and pressure. Under conditions of moderate pressures
and elevated temperatures gaseous solutions tend to be ideal if the com-
ponents are not too widely different in properties. In such cases 4>^
may be taken as 1.0. Maximum deviations from 1.0 occur in the region
of the critical point of the solution.
In the hquid phase the relationship between activity and mole fraction
is complicated by the two types of deviation from the laws of ideal solu-
tions which are discussed on page 609. Because of the widely different
contributing causes and relationships it is convenient to consider these
two types of deviation separately and assign an activity coefficient lo
express each. Thus,
{ax)L = Li'i\Xi ^,j (51)
where
(oi)i = activity of component 1 in the liquid phase
<t>Li = activity coefficient of component 1 in the liquid phase which
takes into account deviations from the laws of ideal solu-
tions resulting from differences in molecular size or vola-
tility
7i = activity coefficient of component 1 in the liquid state which
takes into account deviations from the laws of ideal solu-
tions resulting from differences in molecular or chemical
type
Xi = mole fraction of component 1 in the liquid phase
CHAP. XV] VAPORIZATION EQUILIBRIUM CONSTANTS 665
For ideal systems 71 = 1.0 at all conditions. For nonideal systems 71

may be determined by the methods of the preceding sections which are
applicable at low pressures where the effects of differences in molecular
size or volatility are small and <J)LI may be taken as 1.0. As the tempera-
ture and pressure in such a system are increased 71 tends to approach
1.0 while deviations of <J)LI from 1.0 tend to increase.
Combining Equations (49), (50), and (51) gives
yj = t^K, = Ki (52)
It was suggested by Gamson and Watson'^ that the ratio of the mole
fractions in the vapor and liquid phases be termed the vaporization
ratio K'l of component 1. It is evident from Equation (52) that the
vaporization ratio is equal to the thermodynamic equilibrium constant
only in ideal solutions where the activity coefficients are each equal
to 1.0.
Equation (52) provides a sound basis for calculations involving vapor-
liquid equilibria. In many cases it is satisfactory to assume that the
coefficients (f>Li and 0„i are unity, and, as previously pointed out,
7 i = 1.0 for all ideal systems. Under other conditions it is necessary
to evaluate the coefficients as functions of the properties of the com-
ponent and the system, either from experimental data or generalized
methods of calculation.
At low temperatures and pressures where Dalton's law and the ideal-
gas law are applicable CJ>L and 4>v become unity and Equation (52) re-
duces to
•Kî = — = liKi = - — - (53;
Xi r
where P ' is the vapor pressure of component 1 at the temperature of
the system.
Vaporization Equilibrium Constants. The definition of the standard
states of unit activity which determine the value of the vaporization
equilibrium constant is complicated by the fact that for a heavy com-
ponent of a mixture the standard state for the vapor phase is a hypo-
thetical state in which the pure component cannot actually exist as a
gas. Similarly, for a light component the standard state for the liquid
phase is a hypothetical state in which the pure component cannot actu-
ally exist as a liquid. The fugacities of the standard states under these
' conditions must be arrived at by arbitrary extrapolation of the fugacities
" B. W. Gamson and K. M. Watson, Nat. Petroleum News, Tech. Sec, 36, R623
(Sept. 6, 1944). Also "Process Engineering Data," National Petroleum Publishing
G66 PHYSICAL EQUILIBRIUM «•"'< [CHAP. XV
of the pure components under real conditions. However, these arbitrary-

extrapolations in no way interfere Avith the fundamental validity of
Equations (49) and (52). If the method of extrapolation is changed,
the values of Ki are affected, but corresponding changes result in the
activity coefficients so that it is only necessary that the values of (t>n,
4>Li, Ti; 3'Dd Ki all be referred to consistently defined standard states.
The concept of the vaporization equilibrium constant was introduced
by Souders, Selheimer and Brown" who expressed it by/ the following
equation resulting from combination of Equations (XIV-34) and (49)
with the definition of the fugacity coefficient: '
VmJT - P i )
^ , = (Z?£ = £^!-L^!_ (54)

Pi = normal vapor pressure oi component 1 at the temperature T
• of the system
Vp = fugacity coefficient of the vapor of component 1 at pressure Pi
and the temperature of the system'
T = total pressure of the system -
Vm = mean molal volume of liquid component 1 between Pi and jr
at the temperature of the system
Vn = fugacity coefficient of component 1 in the vapor state at the
temperature and pressure of the system
In order to evaluate cî extrapolation in Fig. 142 is required where ir
is greater than Pi.
As previously mentioned this extrapolation may be arbitrary, but it is
desirable that it be so carried out that activities in the vapor phase are
equal to mole fractions over as wide a range as possible. In the higher-
pressure ranges where deviations from the laws of the ideal solution be-
come large it is desirable that the extrapolation be such as to result in
simple expressions for the activity coefficients. As a basis for such an
extrapolation Gamson and Watson" replotted the data of Fig. 142 in the
form shown in Fig. 152 AAith f/pc = vpr as ordinates. The lines corre-
sponding to low reduced temperatures were extrapolated on the basis
of the available experimental data on the fugacities of hydrocarbons in
vapor phase mixtures in the region where the vapors form approximate^
ideal solutions. At higher pressures the extrapolations were extended to
approach the horizontal. Actual fugacities of components under these
conditions tend to reach a maximum with increase in pressure and then
diminish as the pressure is further increased. However, for any compo-
« M. Souders, Jr., C. W. Selheimer, and G. G. Brown, Ind. Eng. Chem., 24, 517
(1932). •,•!
' y»') •
03
^dc^'d/f
nent at a given temperature the pressure at which this maximum is

reached and the variation with pressures beyond that of the maximum
fugacity are entirely dependent on the composition of the mixture in
which the component is present. The horizontal extrapolations of
Fig. 152 are believed desirable in order to simplify the relationships be-
tween activities and mole fractions in this region of extreme departure
from ideal solutions. The extrapolations at the higher reduced tem-
peratures approaching 1.0 were carried out to yield a continuous rela-
tionship between the extrapolated lines and the real values correspond-
ing to reduced temperatures above 1.0. The broken lines on I^g. 142
are derived from those of Fig. 152, and the two charts may be used to-
gether to cover conveniently the entire range of conditions and thus
define {fl)y for any material under any conditions.
The evaluation or definition of (/I)L in Equation (54) is complicated
by uncertainty as to the proper value of Vm. At low temperatures where
molal volumes are small and little affected by pressure Vm is taken as the
actual molal volume of the liquid. At higher temperatures the molal
volume of the liquid becomes highly dependent upon pressure and above
the critical temperature of the component has no real significance. The
vapor pressure of the liquid in this region offers a further problem.
Because of these uncertainties Souders, Selheimer, and Brown calculated
equilibrium constants by Equation (54) at low temperatures and pres-
sures and then graphically extrapolated the constants into the high-
temperature and pressure range to conform to experimental vapor-
liquid equilibrium data.
In order to be more specific, Gamson and Watson^' proposed that the
standard state of a component in the liquid phase be taken as a hypo-
thetical incompressible liquid whose molal volume is expressed as a func-
tion of temperature by the following equation;
v,n=={viiî)i5.7 + S.0Tr) / '"~ (55)
where (wiui) is the product of the molal liquid volume and w from Fig. 109
at any selected conditions. By comparison of Equation (55) with
(XII-39), page 502, it is seen that the term (5.7 -|- S.OTr) corresponds to
1/co for the hypothetical incompressible liquid reference state. This'
value is a linear extrapolation of l/w from the absolute zero and thus
defines the liquid standard state as having a constant volumetric coeffi-
cient of expansion equal to that of the real liquid at the absolute zero.
At low reduced temperatures values of Vm calculated from Equation (55)
differ little from actual molal volumes. As the critical point is ap-
proached Vm is smaller than the actual molal volume except at high
pressures. ; . • . .,
The definition of (/i)t at all conditions is completed by means of the

vapor-pressure Equation (III-16), page 73, which extrapolates logically
above the critical point. Thus, from (55) and the numerator of (54),
(yicoi)(5.7 + 3 . 0 r , ) ( 7 r - P i ) ,
l o g ( / D i = log(Pi) + log.p, + ^ 2 303fir • ^^^)
where
' : l o g P i = - — + B-e-2(Xrr-6)= (57)
Equations (54), (56), and (57) together with Figs. 142 and 152 completely
define the fugacities of the Uquid and vapor reference states and the
equilibrium constant in an arbitrary but logical manner which is definite,
reproducible, and apphcable to any material at any conditions.
Illustration 4. Calculate the vaporization equilibrium constant of methyl chlo-
ride at 200°F and an absolute pressure of 600 lb per sq in. The density of liquid
methyl chloride is 0.920 g per cc at 18°C, the critical temperature is 750°R, the
critical pressure 967 lb per sq in., and the molecular weight 50.5.
Solution: At 200° .and 600 lb per sq in., Tr = 660/750 = 0.880; p, = 600/967
= 0.620
At 18°C (64.4°F), Tr = 524.4/750 = 0.699
From Fig. 142, w = 0.68; (/°), = (0.68) (600) = 408 lb per sq in.
From Fig. 109 at 18°G, ui = 0.1146
(viui) = (50.5) (0.1146)/(0.920) = 6.290 cc per g-mole = 0.1007 cu ft per lb-mole
From Table Ila, page 73:
log Pi = - ^ ^ ^ ^ -I- 7.4185 - c- 20(0.880 - 0.062)2

^ 0.880
= -3.0903 + 7.4185-10-5.95=4.3282
Pi = 21,290 mm or 21,290/51.70 = 412 lb per sq in.
At 412 lb per sq in., Pr = 412/967 = 0.426
From Fig. 142, >.p = 0.77
log i n . = log 412 + log 0.77 + (0-1007)[5-7 + (3.0)(0.880)][600-412]
iogu;i B -r 6 -r (2.303) (10.71) (660)
= 2.6149 - 0.1135 + 0.0097 = 2.5111
( D t = 324.41bpersqin.
From Equation (54),
K = 324.4/408 = 0.795
In this manner charts may be developed which express the vaporiza-
tion equiUbrium constant of a specific compound as a function of tem-
perature and pressure. Series of such charts have been developed by
several investigators for many of the lower-boiling hydrocarbons. An
alternate method of graphical presentation which is convenient when
working with mixtures of homologs of a single series is to restrict each
PHYSICAL EQUILIBRIUM [CHAP. XV
-300 -200 -100 0 100 200 300 400 500 . 600

Normal Boiling Point, °F
FIG. 153. Vaporization Equilibrium Constants of Hydrocarbons. (Absolute Pres-
sure n = 100 lb per square inch.)
chart to a constant pressure and plot curves each of which^ expresses the
equilibrium constant of a particular compound as a function of tempera-
ture at the pressure of the chart. With this scheme a series of charts
each corresponding to a different pressure is required.
Another method of presenting equilibrium-constant data for an ho-
VAPORIZATION EQUILIBRIUM CONSTANTS 671
-300 -200 -100 0 100 200 300 600

Normal Boiling Point, °P
FIG. 154. Vaporization Equilibrium Constants of Hydrocarbons, (Absolute Pres-
sure n = 200 lb per square inch.)
mologous series is shown in Figs. 153 and 154 for the paraffin hydro-
carbons. These charts were calculated by Equations (54) and (56)
from the data of Table XXVIII. Although based on the data for paraf-
fins they may be used for olefins with little error. The products of KTT
in pounds per square inch are plotted as ordinates with the normal boil-
ing points of the homologs as abscissas. Each curve on the chart cor-
responds to a constant temperature, an(i the chart is restricted to a
constant total pressure x. It will be noted that for the low-boiling com-
pounds there is little difference between the values of Kir at 100 and 200 lb
per sq in., from Figs. 153 and 154, respectively. For such compounds
in the moderate-pressure range the problem of interpolating between
charts to obtain values of >K at some intermediate pressure is simpUfied
by plotting Kir rather than K. In the high-pressure range or for high-
boiling compounds K-ir varies with pressure more than K, and it is prefer-
able to plot K directly. The choice of the method of plotting iâ matter
of convenience dependent upon the type of problem. Kirkbride^^ has
developed a method of simplifying dew-point and bubble-point calcula-
tions by plotting all equilibrium constants relative to those of ethane at
the same temperature and pressure. Because of insensitivity of these
relative equilibrium constants to temperature and pressure this method
reduces the labor of trial-and-error calculations.
Bubble-Point Equilibria. A liquid at its bubble point must be in
equilibrium with the first trace of vapor formed. In a mixture of com-
ponents 1, 2, 3 • • • the mole fractions in the combined phases of the total
mixture are represented by zi, Zi, Zz, • • -, respectively. The mole frac-
tions in vapor phase are yi, 1/2, ya, • • •, and in the hquid phase xi, Xa,
Xi, • • •. Under all conditions
Zl + 22 + 2 3 + - - - = 1.0 (58)
• ^. yi + y2 + y3+--- = 1.0 (59)
• . f^xi + X2 + xs+--- = 1.0 (60)
When the mixture is at its bubble point Xi = zi, X2 = zi, etc. Also, by
Equation (52), t/i = K{x\, y-i = ^^'2X2, etc. Combining these relation-
ships with Equation (59), at the bubble point, gives
KUi + X^Z2 + -f:;33+'-- = 1.0 "V - (61)
or SK'z = 1.0 and i:,{K'T^Z) I-K = 1.0
If the vaporization coefficients K' are known for all components as func-
tions of temperature and pressure. Equation (61) permits complete cal-
culation of bubble-point conditions by a trial-and-error procedure.
If it is desired to calculate the bubble-point pressure of a mixture at a
specified temperature, a pressure is assumed as a first approximation.
Corresponding values of K[, K'^, X3 • • • or K'^^K, KW Kiv • • • are deter-
mined, and the left-hand side of Equation (61) is evaluated. If the
summation equals 1.0 the assumed pressure is correct. If not, other
values are assumed, and the corresponding summations are plotted as a
1' C. G. Kirkbride, Petroleum Refiner, 24, 99 (1945). ^
CHAP. XV] PARTIAL VAPORIZATION 673
function of pressure. The pressure at which this curve crosses 1.0 is

the correct solution. The vapor composition is then calculated from
Equation (52), the equilibrium constants being used at the correct pres-
sure. A similar procedure is followed for calculating the bubble-point
temperature at a specified pressure.
At low pressures Equations (53) and (61) may be combined to permit
direct calculation of bubble-point pressures.
.' •• yiP*iZi + y2Plz2 + y3Plzi+--- = w* (62)
Dew-Point Equilibria. A vapor at its dew point is in equilibrium
with the incipient liquid having a composition Xi, x^, Xa- • •. Then,
since 2/1 = 21 and Xi = yi/K'i, Equation (60) becomes
2i Z2 23
^ + ^ + ^ + ••• = 1-0 (63)
or l(|,)=1.0 and ^z(£)'1.0
Dew-point conditions may be evaluated by a trial-and-error solution of

Equation (63) by the same procedure described for bubble points. At
low pressures Equations (53) and (63) may be combined to permit direct
solution for dew-point pressures.
''* = - 7 r- (64)
Zi , Z2 . Z3
^ ;:•.., _,--'-,:,, TiPt T2P2 TsP; . .^
Partial Vaporization. If a solution comprising F moles is partially
vaporized to form L moles of liquid and V moles of vapor, all components
will be in equilibrium concentrations in the two phases, as required by
Equation (52). From an over-all material balance, following the method
of Katz and Brown, ^^ there results
V + L =F .. ' (66)
Similarly, for any component, ,'
Vyx^hx^. = Fzx ' (66)
where Z\ = mole fraction of component 1 in the F moles of total mixture.
Combining (52) and (66) and solving for X\ yields
Fzx F / zi \ - : ,
w D. L. Katz and G. G. Brown, Ind. Eng. Chem., 25, 1S73 (1933J.

Since
X1 + X2 + X3 + = lix=\.0
_ F
(68)
" F
-+f
Similarly, solving (66) for 2/1 gives
Fzi
yi = (69)
and
V
(70)
F
1 + VK'
Either Equation (68) or Equation (70) may be used in a trial-and-error
solution to establish the relationship among temperature, pressure, and
percentage vaporization. Values of the quantity sought are assumed,
and the corresponding summations are evaluated. The correct value is
that producing a summation equal to V/F. The relationship between
V/F and L/V is obtained by rearranging Equation (65):
^ = ^-1.0 (71)
V V
If Equation (70) is used in the summation, values of vapor compositions
are obtained directly.
Zl
yx = (72)
['/(-^)]
Equation (72) may be used with little error even if the summation does
not exactly equal V/F. Corresponding values of Xi are then obtained
from Equation (66). ^ '
Fz,-Vy, A , F \ V
Xi = (73)
If Equation (68) is used for the summation, values of x are obtained di-
rectly by an expression similar to Equation (72). It is evident that in a
partially vaporized mixture at equilibrium the vapor phase exists at its
dew point while the liquid phase exists at its bubble point.
Illustration 5. A hydrocarbon mixture contains 25 mole per cent propane, 40 per
cent n-butane, and 36 per cent n-pentane. Calculate the bubble-point temperature,
CHAP. XV] PARTIAL VAPORIZATION 675
the dew-point temperature, and the temperature of 45 mole per cent vaporization
at an absolute pressure of 200 lb per sq in. Also, calculate the compositions of the
vapor formed at the bubble point, of the liquid formed at the dew point, and of the
hquid and vapor resulting from 45 mole per cent vaporization. It may be assumed
that ideal solutions are formed.
Solution: (a) Bubble-Point Temperature and Vapor Composition. Values of K,
which are assumed equal to K', are obtained from Fig. 154 for substitution in Equa-
tion (61).
Assumed Tem- •,:^ .:• .
perature, 180°F 200°F 220°F
Component z KV K' K'z X V K' K'z X'T K' K'z 2/
CBHS 0.25 374 1.87 0.468 420 2.10 0.525 462 2.31 0.578 0.496
C4H10 0.40 164 0.82 0.328 196 0.98 0.392 226 1.13 0.452 0.359
CsHij 0.35 74 0.37 0.130 92 0.46 0.161 110 0.55 0.193 0.145
1.00 0.926 1.078 1.223 1.000
By graphical interpolation the temperature where tK'z = 1.0 is found to be 190°F.

For a rigorous calculation of vapor compositions y the equilibrium constants should
be evaluated at this temperature. A good approximation is obtained by calculating
K'lZÎ'ZK'z at 180° and 200° and interpolating hnearly.
(6) Dew-point Temperature and Liquid Composition. Substitution in Equation (63),
gives
Assumed Temperature, °F 220 240 260
Component z K'rr K' z/K' K'T K' z/K' K'17 K' z/K' X
CaHs 0.25 462 2.31 0.108 510 2.50 0.100 546 2.72 0.092 0.104
C4H11) 0.40 226 1.13 0.354 254 1.27 0.315 290 1.45 0.276 0.329
CBHU 0.35 110 0.55 0.637 132 0.66 0.531 156 0.78 0.449 0.567
S 1.00 1.099 0.946 0.817 1.000
Graphical interpolation fixes the temperature at which SZ/z = 1.0 at 233°F. A good
approximation to the hquid composition is obtained by calculating (zi/Kd/Zz/K' at
230° and 240° and interpolating.
(c) Temperature and Composition of 45 Per Cent Vaporization.
V/F = 0.45
From Equation (71), L/V = 1/0.45 - 1.0 = 1.22
or y/L= 0.82
Equation (70) is used for the trial summations. <
Assumed Temperature, 200°F 220°F 240°F
Com- z z z
ponent z K' 1+L/VK' l+L/VK' K' l+L/VK' l+L/VK' K' l+L/VK' l+L/VK'
CaHs 0.25 2.10 1.581 0.158 2.31 1.528 0.164 2.50 1.488 0.168
C4H10 0.40 0.98 2.245 0.178 1.13 2.080 0.192 1.27 1.961 0.204
C5H12 035 0.46 3.655 0.096 0.55 3.220 0.109 0.66 2.850 0.123
2 1.00 0.432 0.465 0.495
By graphical interpolation the temperature at which S«/(l + L/VK') = 0.45 is
determined as 211°F. Satisfactory values of y are obtained by interpolating between
values calculated from Equation (72) at 200 and 220°P. Corresponding values of z
are calculated from Equation (73). J
Component z y 1.822! 0.821/ x
CsHg 0.25 0.358 0.455 0.294 0.161
C4H10 0.40 0.413 0.728 0.339 0.389
CsHij 0.35 0.229 0.637 0.187 0.450 ;;
LQO 1.000 1.000 , •. ;
Activity Coefficients. In Illustration 5 it is assumed that all activity

coefficients are equal to 1.0 and that the vaporization coefficients are
equal to the equilibrium constants. This assumption is satisfactory for
many calculations for ideal systems at moderate pressures if the com-
ponents are not widely different in properties. In working with nonideal
systems under similar conditions it is satisfactory to assume that L and
<j>v are unity. Values of 7 may be derived by the methods described in
the preceding sections. At elevated pressures where the ideal-gas law
is not valid, the following relation is used in preference to Equation (2).
71 = ; ^ % ^ " ' (74)

XiKi • I
At high pressures, approaching the critical of the mixture, the assump-
tion that <^L and 0„ are unity is unsatisfactory for both ideal and nonicjêal
systems. Serious errors are also encountered at moderate pressures for
components widely different in volatility from the average of the mix-
ture. An example is methane dissolved at moderate pressure in ab-
sorption oil. Much effort has been directed toward the evaluation of
these corrections, but no entirely satisfactory and general method has
yet been developed.
On the basis of the pseudocritical concept Gamson and Watson'' de-
veloped general methods for calculating fugacities in both the liquid and
vapor states at all cohditions. The accuracy of these methods is limited
by the lack of sound definitions of the pseudocritical properties and by
the inaccuracies inherent in the generalized compressibility-factor re-
lationship. These difficulties require the introduction of empirical cor-
rection factors which bring the calculated results into fair agreement with
the available data over wide ranges of conditions. The method of cal-
culation is too tedious to be of practical value in the direct calculation
of vapor-liquid equilibria but may be used to derive values of 0^ and <^„
from Equations (50) and (51) which for a particular system can be re-
lated graphically to the properties and conditions of the respective phases.
Graphical correlation of (^L and 0„, either from experimental data or
from the generalized method of calculation of Gamson and Watson, is
CHAP. XV] SOLUBILITY OF SOLIDS AND LIQUIDS 677
complicated by the large number of variables entering into the relation-

ships. The most important are:
(a) Tci/T'c = the ratio of the critical temperature of the component to
the pseudocritical temperature of the Tphase under con-
sideration
(6) Pr = the pseudoreduced pressure of the phase under consider-
ation
(c) T'T = the pseudoreduced temperature of the phase under con-
sideration
For a system composed of homologs, <^„ and ^z, may be expressed in terms
of only these three variables over limited ranges of variation of the aver-
age properties of the mixtures. In order to obtain more general rela-
tionships additional variables must be considered, even when dealing
with homologs.
Consideration of the activity coefficients considerably complicates
vaporization calculations. Since the activity coefficients are functions
of the properties of the individual phases, it is necessary to assume trial
phase compositions, and then evaluate approximate activity coefficients
and vaporization coefficients. The phase compositions are then cal-
culated and corrected by a repetition of the procedure.
SOLUBILITY
The distribution of a component among the phases of any system at
equilibrium under a constant temperature is expressed by Equation (48).
As has been pointed out, this general relationship may be used to esti-
mate activity coefficients and vapor-liquid equilibria from solubility or
distribution data. Conversely, these same methods may be used to
estimate distribution and solubility data from vapor pressure or vapor-
liquid equilibrium measurements.
Solids and Liquids. For solutions of nonelectrolytes Hildebrand^'
recommends that the activity in the liquid phase of a sofid solute be
referred to the pure solute in a hypothetical liquid state at the tempera-
ture and pressure of the system. The activity in the solid state is taken
as unity for the pure solid at the temperature and pressure of the sys-
tem. With these conventions Equation (48) for the solubility of a solid
reduces to : . ^
•;— TW"'^'^ (75)
" J. H. Hildebrand, "Solubility of Non-Electrolytes," Reinhold Publishing Cor-

poration, 1936. i . • • . • .,
678 • PHYSICAL EQUILIBRIUM [CHAP. XV
where
S2, 72» = the mole fraction and activitjr coefficient of component 2
in the solid phase
Xi, 72i = the mole fraction and activity coefficient of component 2
in the liquid phase
(Jl)s = fugacity of component 2 in the solid state at the tempera-
ture and pressure of the system
(JI)L = fugacity of component 2 in the hypothetical liquid state at
the temperature and pressure of the system
Ki = solubility constant of component 2, referred to the pure
component
An expression for the solubility constant as a function of temperature
is obtained by combining Equations (XIV-38) and' (75):
^ ^ (76)
where HJL and the molal enthalpies of pure component 2 in the

hypothetical liquid and solid states, respectively, at the temperature
and pressure of the system. If it is assumed that the heat of fusion is
independent of temperature Hsi — HJS may be taken as the normal heat
of fusion and Equation (76) integrated with K2 = 1.0 at the melting
point Tm where {fl),= {f^)L. Then"
^^'- R\T Yj--- RTT„ (^^)

where
V = molal heat of fusion of component 2
T — temperature of the system (
= melting point of component 2
Similarly, the effect of pressure upon the solubility equilibrium con-
stant is obtained by combining Equations (XIV-30) and (75) to give
\ dp l~ RT (^^)
i"g^--^'""^;"^^"(x-p.) (7^)
where
v°,2 = molal volume of the pure solid solute at the temperature and
pressure of the system
flu = molal volume of the hypothetiqal Hquid solute at the tempera-
ture and pressure of the system
Pm = equilibrium pressure corresponding to melting point T
CHAP. XV] SOLUBILITY OF SOLIDS AND LIQUIDS 679
I t may be noted from Equations (77-79) that for an ideal solution

where K2 = x^ the solubility of a solid decreases with increase in molal
heat of fusion, increases when the solid contracts upon melting, and when
expressed as mole fraction is independent of the nature of the solvent;
on a weight basis the solubility decreases with increase in molecular
weight of the solvent for ideal solutions.
Equation (75) furnishes a sound basis for the evaluation of the cor-
responding activity coefficients. These coefficients approach unity in
systems which do not involve polar components, large differences in
molecular size, or large differences between Tm and T. If the activity
coefficients are assumed to be unity, the ideal solubility^'' of a pure solid
is equal to K2, or x^ = K2.
Equation (77) may be used for predicting the solubility of a solid from
its melting-point and heat-of-fusion data or conversely for calculating
heats of fusion from solubility data where conditions are such that ideal
solutions are approximated. The usefulness of Equation (75) is re-
stricted by the extensive experimental data required for evaluation of
the activity coefficients.
Illustration 6. From Fig. 10, page 113, it may be seen that the solubility of
naphthalene in benzene is 60 per cent by weight at 45°C and 21 per cent at 0°C.
Assuming the solution to be ideal, calculate heats of fusion of naphthalene from these
two solubility values. The melting point of naphthalene is 80.1°C.
Solution:
Basis: 100 g of solution (component 2 is naphthalene).
0''C 45°C
Molecular Mole Mole
Weight g g-moles Fraction g g-moles Fraction
Naphthalene 128 21 0.164 0.14 60 0.469 0.478
Benzene 78 79 1.013 0.86 40 0.512 0.622
100 1.177 1.00 100 0.981 1.000
Log K2 = logxj -0.854 -0.321
Afn<.o^ (2.303) (0.854) (1.99) (273.1) (353.2)
At 0 C, V = (353.2,273.1) == ^™ ""^ ^'' ^'"^"^^
. ..or. , (2.303)(0.321)(1.99)(318.1)(353.2) ^^,„ ,
at45°C,X, = (353,2,273.1) ^^ = 4710 cal per g-mole "
These results are in fair agreement with the observed value of 4,550.
The fair agreement between the heats of fusion calculated at the two
temperatures of Illustration 6 indicates that the solutions of benzene in
naphthalene closely approximate ideal behavior. In other systems,in-
volving less similar components much larger discrepancies can be ex-
pected and the activity coefficients of Equation (75) must be evaluated
for satisfactory results. These deviations in a number of systems are
discussed by Hildebrand."
For solutes having melting points far. removed from the temperature
of the system or for electrolytes whicn dissociate into ions in solution,
it is not advantageous to use the hypothetical standard state of Equa-
tion (75) for the hquid phase. A more convenient convention is to ex-
press the activity of the solute in the liquid referred to the infinitely
dilute solution at the pressure of the system. As is mentioned on page 629
this convention corresponds to the selection of a hypothetical standard
state the fugacity of which is the value which would exist if the solution
behaved ideally at unit mole fraction (or unit molality) as it does at zero
concentration. The enthalpy and volume of this hypothetical standard
state are therefore equal to the partial molal enthalpy and volume, re-
spectively, at zero concentration. The activities in the solid phase are
best referred to the pure solid at the pressure of the system as the stand-
ard state. On this basis the solubility of a solid is expressed by
_ 7^7- - -- A2 (.SU)
«23 J2L S272J
where
Si, T23 = mole fraction and activity coefficient of component 2 in
the solid phase
2^2, 72 L = mole fraction and activity coefficient of component 2 in
the given solution
jla = fugacity of the pure solid at the temperature and pressure
of the system
y^L = fugacity of the hypothetical standard state for the liquid
phase referred to infinite dilution
K'i = solubility equilibrium constant of component 2 referred to
infinite dilution
The eifect of temperature on the solubility constant is given by an
equation expressing the effect of temperature on the fugacities of ^the
standard states. From Equations (XIV-45).and (80),
(d\nK',\ JBli^-nl,)
V- dT A RT^~ ^^^'
where
H2i = partial molal enthalpy of component 2 at infinite dilution at
the temperature and pressure of the system
Has = mo}al enthalpy of component 2 in the solid state at the tem-
perature and pressure of the system
The quantity Elf^ — nla is the differential heat of solution at infinite
dilution.
Similarly, by combining Equations (XrV-44) and (80) the effect of
CHAP. XV] SOLUBILITY OF GASES 681
pressure on solubility is derived,

/d\nK'2\ ^ ivlL-vl.) ^g^^
\ dp )T
where
^2L = partial molal volume at infinite dilution at the temperature
and pressure of the solution
vis = molal volume of the pure solid at the temperature and pres-
sure of the system
I t may be noted from Equation (81) that the solubiUty of a solid is
increased by increased temperature if its differential heat of solution is
positive, corresponding to absorption of heat in dissolution. Similarly,
the solubility is increased by increase in pressure if the partial molal
volume in dilute solution is less than the volume of the solid.
The practical value of Equations (80-82) is restricted by the great
variations >vith concentration of the activity coefficients in liquid solu-
tions. Except in dilute solutions these coefficients must be evaluated
from empirical data which are seldom available.
Equations (80-82) are directly applicable to electrolytes with activi-
ties expressed in either molecular or ionic terms. However, the use of
mean ionic activities and molalities for these materials has the advantage
that the activity coefficients approach unity in dilute solutions, as dem-
onstrated in Fig. 145, whereas the molecular activity coefficients become
very small.
Solubility of Gases. The solubility in a hquid of a gas, either in the
pure state or from a mixture, is expressed thermodynamicaUy by Equa-
tion (49). If the temperature of the system is below or not greatly above
the critical temperature of the gas, the standard state for the gaseous
component in the liquid phase is conveniently taken as the pure com-
ponent in a hypothetical incompressible liquid state at the temperature
and pressure of the system, exactly as described for handling vapor-
liquid equilibria. If it is assumed that ideal solutions are formed in the
gaseous phase and the standard state for the gaseous phase is taken as
the state of unit fugacity. Equation (49) combined with (51) reduces to
— - {Ji)!' - J^2 (83)

a;272<P2z,
where
TT = total pressure of the system
V2w = fugacity coefficient of pure gaseous component 2 at the
temperature and pressure of the system
72, <^2t = activity coefficients in the liquid state as defined by Equa-
tion (51), page 664
Comparison of Equation (83) with (V-5), page 147, in which the par-
tial pressure p2 is equal to iry^ results in a thermodynamic expression for
Henry's constant H2. Thus ,
m = (84)
Equations (83) and (84) ihay be used to predict what may be termed
ideal solubilities if 72 and <^2t are assumed to be unity; v^rr is derived
from Fig. 142, and (JDL is calculated from Equation (56). As pointed
out by Hildebrand," this method gives fair results in systems not in-
volving highly polar materials if the temperature is below the critical
temperature of the gas. At higher temperatures the calculation of (fl)L
becomes uncertain. Arbitrary specification of (fî in such ranges re-
quires consideration of corresponding values of (^21, which must be based
on experimental data. Such data are necessary in any case for the pre-
diction of 72 in systems involving high polarity or chemical reaction and
association.
Illustration 7. Calculate the ideal solubility of chlorine in carbon tetrachloride at
a pressure of 1.0 atm and a temperature of 0°C. The density of hquid chlorine is
1.561 g per cc at -34.6°C. The vapor pressure at O'C is 3.65 atm. Tc is 417°K and
Pc is 76.1 atm.
Solution: At -34.6°C, Tr = 238.6/417 = 0.571
From Pig. 109, wi = 0.1266
(fiwi) = (71)(0.1266)/(1.561) = 6.758 cc per g-mole
At 0°C and'3.65 atm, T^ = 0.655 and p, = 3.65/76.1 = 0.048. '
Then, from Pig. 142, ,>P = 0.95.
From Equation (56), at 0°C,
log (/D. = log 3.65 + log 0.95 + 5.758[5.7 + (3.0)(•655)j(1.0 - 3.65)
susyi B -re -r (2.303) (82.06) (273.2)
= 0.6623 - 0.0223 - 0.0023 = 0.5377
(y?)£ = ^2 = 3.45
Prom Equation (83), neglecting vir at the low pressure involved and assuming that
72 and <t>iL are unity, we have
X2 = (1.0)(1.0)/(3.45) = 0.290
The experimentally determined value is 0.298.
It may be noted that with this definition of the standard state in the
liquid-phase gas solubility calculations are identical with the vapor-
liquid equilibrium calculation demonstrated in preceding sections.
For systems involving gases far above their critical temperatures or
where dissociation or chemical reaction is involved, the standard state
used in Equation (83) for the liquid phase loses its advantage. It is
CHAP. XV] SOLUBILITY OF GASES 683
frequently more convenient to express the activities of the solute gas

referred to infinite dilution. With this convention Equation (83)
becomes
^ " = ( / r ) x = 2f^ (85)
where 72L, {JI')L and K'2 all correspona to activities referred to infinite
dilution. For such systems the single coefficient (72) i may be used to
take into account all types of deviation from ideality. In dilute solu-
tions or when deaUng with nonpolar nonreacting solutes it is frequently
satisfactory to assume that (72 )L is unity. With this assumption a
single experimental solubility value serves to determine K'^. The varia-
tion of the solubility constant with temperature is derived from Equa-
tion (XIV-45)
\r^T~), Rf^ ^^^)

where
H2G = enthalpy of component 2 in the ideal gaseous state at tem-
perature T
H2i = partial molal enthalpy of component 2 in solution at infinite
dilution at the temperature and pressure of the system
The effect of pressure on K'2 is derived from Equation (XIV-44),
/d In K'A ^
... si »')
where
VIL — partial molal volume of component 2 in the liquid phase at
infinite dilution
Equation (86) may be used to calculate the effect of temperature on
solubility from heat-of-dissolution data or conversely to calculate the
differential heat of dissolution at infinite dilution — (H2ff —H2L) from
solubility data at two temperatures if the ideal solution behavior is
approximated.
Illustration 8. At a partial pressure of 1.0 atm pure CO2 is in equilibrium with
'aqueous solutions containing 0.096 mole per cent CO2 at 10°C and 0.0294 mole per
cent CO2 at 60°C.
(a) Assuming the solution to behave ideally, calculate the mole fraction of CO2 in
an aqueous solution saturated with CO2 at 60°C and 100 atm. The effect of pressure
on K'^ may be neglected.
(6) Calculate the average differential heat of dissolution of CO2 in water at infinite
dilution in the range of 10-60°C.
Soluiidh:
(a) For COa, Tc = 304.3°K; p» = 73.0 atm
At 60°C, 1 atm, Tr = 1.10; p, = 0.014

From Fig. 142, v„ = 0.995 »*
Substitution in Equation (85), if it is assumed that 72L = 1.0, gives
X2 = (1.0) (0.995)/(0.000294) = 3380
At 60°C, 100 atm, T, = 1.10; Vr = 1.37; vi, = 0.63
From Equation (85), ,
X2 = (100)(0.63)/3380 = 0.0186 or 1.86 per cent
(6) At 10°C, and 1.0 atm, Tr = 0.93; p , = 0.014; v = 0.995
Ki = (1.0)(0.995)/(0.00096) = 1036
Substitution in Equation (86), after integration, gives
(Hfe - 3 . ) = - [(1.99)(2.303) log | | ] / ( ^ - ^ ) = 4,431 calperg-mole
The differential heat of dissolution of the ideal gas at infinite dilution is therefore
—4,431 cal per g-mole.
The experimentally observed mole percentage of CO2 in water at 60° and 100 atm
is approximately 1.8 per cent.'' Assumption of Henry's law in this case would pre-
dict a value of 2,94 per cent, an error of over 60 per cent as compared to an error
of less than 5 per cent for Equation (85). The good agreement with Equation (85)
indicates that the activity coefficients in the hquid phase are substantially unity
as assumed, in spite of the chemical reaction occurring between carbon dioxide and
water and the accompanying ionization. This may be interpreted as indicating that
only a small fraction of the dissolved CO2 reacts with the water. Where higher so-
lute concentrations or more extensive reactions are involved, the assumption of ideal
behavior may lead to large errors which require evaluation of the activity coefScientS,
The solubiUty of gases with accompanying chemical reaction is discussed in Chap-
ter XVI, page 742.
Gas-Solid Adsorption Equilibria. Equilibrium conditions in adsorp-
tion of gases on solids are discussed in Chapter V, pages 149-160, and
the enthalpy of such systems in Chapter VIII, pages 297-300. From
Equation (VIII-48),
(M)2 + 1)H = Hi + H2W2 + AH = Hi + W2H2 (88)
or AH = Hi~Hi+ WzC-ff2- H^) = Affi-1- WiAH^ (89)
where '
AHi = the differential heat of adsorption per unit mass of adsorbftnt
AHz = the differential heat of adsorption per unit mass of adsorbate
AH = the integral heat of adsorption per unit mass of adsorbent;
Wi — mass of adsorbate per unit mass of adsorbent
From Equation (89), at constant mass of adsorbent,
AH, = (--,) (90)
f "International Critical Tables," McGraw-Hill Book Company, New York.

CHAP. XV] GAS-SOLID ADSORPTION EQUILIBRIA 685
Since the differential heat of adsorption varies with the mass adsorbed,
the total heat of adsorption in adsorbing a mass W2 of gas is obtained by-
integration, thus
X U!2
AH2 dw2 (91)
From Equation (g) page 472, applied to gas adsorption under conditions
of constant composition,
/3AS\ ^ /dp\ •
Since adsorption at Constant temperature and composition also takes

place at constant pressure T AS = AH and Equation (92) may be di-
rectly integrated, _
^.-^.. = (|)jF.-T^..) =^ ^ (93)
If the partial molal volume of the adsorbed gas is neglected and ideal-
gas behavior is assumed,
/dlnp\
H2-H2, = -RTnnp = R —f-\ (94)
and
Illustration 9. In the adsorption of NO2 gas by silica gel the following equilib-
rium isotherm equations were derived from the experimental work of Foster:"
At 16°C, W2 = 0.330 pi'''
At25°C, Wi = 0.115 p\'^
At 35° C, W2 = 0.022 pi«™
where pi = partial pressure of NO2 in millimeters of Hg
W2 = grams of NO2 adsorbed per 100 g of siUca gel
Calculate the differential heat of adsorption per gram-mole of NO2 when Wi = 5.0.
From a cross-plot of In p against 1/T at w^ = 5.0,
/ d\np\
\S|
From Equation (94),
AH2 = H2 - H25 = -3760(1.987) = -7490 cal per g-mole NO2
" Gordon Poster, Ph.D. thesis, University of Wisconsin (1944).
To obtain the integral heat of solution of 5 g of NO2 on 100 g of silica gel it is nec-
essary to obtain the differential heat of solutionjor different values of Wi from 0
to 5 and then integrate according to Equation (95).
MINIMUM WORK AND ENTROPY OF SEPARATION

When two dissimilar substances are mixed together to form a solution
the escaping tendency and fugacity of each is diminished. As a result,
the operation of mixing is accompanied by a decrease in free energy
of the system. If the mixing could be carried out in a reversible manner,
the process would be capable of performing a corresponding maximum
quantity of useful work. Conversely, in order to separate the compo-
nents of a solution, useful work must be done on the system. The
useful work expended in separation is a minimum when the" operation
is conducted reversibly and in this case is equal to the increase in4free
energy accompanying the separation. The total work expended in the
separation is the sum of the useful work plus the work of expansion which
accompanies any changes in volume.
The employment of a reversible method for separating the components
of a solution can conceivably be devised by the use of a semipermeable
membrane which will permit one component to flow through under pres-
sure but not the other. Such membranes are rarely used in industrial
processes because of the extreme slowness of separation but ar^ widely
distributed in biological processes, such as metabolism and osmosis.
Some electrochemical processes approach reversibility in the separation
of components from solution. Actual separations normally consume
. energy far in excess of the reversible work requirements, which serve
as a criterion for judging the thermodynamic efficiencies of separation
processes.
The free energy of a solution is related to the partial free energies of
its components by an expression similar to Equation (VIII-32), page 286.
Thus, when a solution is forme(| from the standard states of its purft
components, the free-energy change is represented by
AG = {UAGA + + ncGc + • • •)
UBGE
— (W^GA + WSGB + WcG? -\ ) = —Wf (96)
AG = RT(nA In aA+ns In as+ nc In ac-{ ) = —Wf (97)
or
AG = RT{nA In T^JV^ + UB In yeNe + ric In -ycNc -\ ) = -Wf (98)
Where all activities are referred to the pure components, Equation (98)
represents the free energy of 'formation of the solution from its compo-
nents, each in its specified standard state. "^
CHAP. X V ] ENTROPY OF DISSOLUTION 687
Illustration 10. From the data of Fig. 146, calculate the minimum useful work
required to separate 100 lb of a 1.0 molal solution of sodium chloride at 100°C
into pure water and solid salt at the same temperature and pressure.
Solution:
Basis: 100 lb of solution.
Per cent NaCl by weight = (58.5/1058.5)100 = 5.52
Lb-moles NaCl = 5.52/58.5 = 0.0945
Lb-moles H2O = 94.48/18 = 5.26
Mole fraction NaCl = 0.0945/6.35 = 0.0177
Mole fraction H2O = 0.9823
From Fig; 146,
^NaCl= 0-593 ^ 7H,O = 0.9832

"NaCi ^ (0-0177) (0.593) = 0.0105 " H J O " (0.9823) (0.9832) = 0.964
From Equation (97),

AG = (1.99) (672) (2.303) [(0.0945) log 0.0105 -t- (5.26) log 0.964]
= —833 Btu in the formation of 100 lb of solution
Therefore the minimum useful work done on the solution in separating it into its
components is -}-833 Btu or, for the separation process, a/ = —833.
Entropy of Dissolution. The entropy changes which accompany dis-

solution may be calculated from the enthalpy and free-energy changes
by Equation (XI-5). Thus,
„ AH-AG
A'S = J, (99)
In'Equation (99), AH is the integral heat of solution and AG the free

energy of solution calculated from Equation (98). For ideal solutions
AH = 0 and the activity coefficients of Equation (98) become unity.
Illustration 11. Calculate the minimum useful work and entropy change of sepa-
ration of 1 lb-mole of air at 65°F and 1.0 atm pressure into oxygen and nitrogen each
at the temperature and pressure of the mixture.
Solution: For air at the given conditions mole fractions may be substituted for
activities referred to the pure components at the temperature and pressure of the
solution. From Equation (98),
AG = (1.99) (525) (2.303) [(0.21) log 0.21 -1- (0.79) log 0.79] = -540 Btu
Therefore, the free-energy change of separation is —540 Btu per lb-mole of air, which
is the minimum work of separation done on the system, or w/ = —540 for the sepa-
ration process.
From Equation (99), for the separation process,
AG 540
AS = = = —1.03 entropy units
T 525
PROBLEMS
1. The system toluene-acetic acid forms an azeotrope containing 62.7 mole per
cent toluene and having a minimum boiling point of 105.4:°C at 760 mm Hg. The
vapor-pressure data from the International Critical Tables are as follows:
fC Toluene Acetic Acid

70 202.4 136.0
80 ' 289.7 202.3
90 404.6 293.7
100 557.2 417.1
110 — 580.8
120 — 794.0
Normal boiling point, °C 110.7 118.5
Calculate the van Laar constants A and B for this system, and plot n and yi as
ordinates on a logarithmic scale against the mole fraction of toluene (component 1)
on a uniform scale, neglecting the association of acetic acid in the vapor phase.
2. The following data on the vapor pressures in millimeters of Hg of carbon-
tetrachloride-ethyl-alcohol solutions are from the International Critical Tables:
Weight Per Weight Per

Cent e c u 34.8°C 66°C Cent CCI4 34.8°C 66°C
0 103 462 58.25 206 752
10.23 122 520 71.68 220 782
20.02 142 576 76.69 223 789
27.13 156 614 84.25 226 788
39.06 179 677 92.98 225 780
43.87 187 700 97.43 221 741
48.86 193 716 98.6 SIS 717
99.9 187 693
100.0 173 544
From these evaluate the van Laar constants for the system at 34.8°C and 66°C.
3. From the results of problem 2 calculate vapor compositions corresponding to
Uquid compositions of 10, 20, 30, 40, 50, 60, 70, 80 and 90 mole per cent CCU SH,
pressures of 760 and 200 mm of Hg. Plot these results as y-x diagrams similar to
Fig. 148. Assume that the activity coefficients are independent of temperature over
the hmited ranges involved. The vapor pressures of the pure component are as
follows:
VAPOE PKBSSTJEBS, MM OF Hg
i°C CCI4 C2H5OH ec CCI4 CjHsOH
20 91 43.9 55 379.3 280.6
25 114.5 59.0 60 450.8 352.7
30 143.0 78.8 65 530.9 448.8
35 176.2 103;7 70 622.3 542.5
40 215.8 136.3 75 — •
666.1
45 262.5 174.0 80 843 812.6
'50 317.1 222.2 Normal boiling point, °C 76.75 78.32
CHAP. XV] PROBLEMS 689
4. The composition of the azeotrope formed by carbon tetrachloride and ethyl

alcohol at 760 mm of Hg is given in the International Critical Tables as 61.3 mole
per cent CCU and the boiUng point as 64.95°C. Using the vapor-pressure data of
problem 3, calculate the composition of the azeotrope at a temperature of 34.8°C
and its total vapor pressure, using Equations (46) and (47). Evaluate the van Laar
constants for the system at 34.8°C, and compare these results with those of problem 2.
6. Methyl.cyclohexane and 7i-heptane are each partially miscible with aniUne at
25°C. The equilibrium compositions in mole per cent are as foUows:*"
Methyl cyclohexane-Aniline Heptane-Aniline
Hydrocarbon layer 11.38% aniline 7.55% aniline
Aniline layer 82.35% aniline 93.91% aniline
Calculate the van Laar'constants for the methyl cyclohexane-aniline system and
for the n-heptane-aniUne system at 25°C. Plot curves relating the activity coeffi-
cients of each of the respective hydrocarbons to its mole fraction in aniline.
6. (o) Using the results of problem 5 and Equation (44), calculate the activity
coefficients of methyl cyclohexane and n-heptane in a solution of the following com-
position at 25°C, using the indicated liquid densities of the pure components:
Liquid Density (Pure)
Mole Per Cent (g per cc)
n-Heptane 3.6 0.68
Methyl cyclohexane' 5.4 0.76
AniUne 91.0 1.02
It may be assumed that the n-heptane and methyl cyclohexane form ideal binary
solutions with each other.
(6) The vapor pressures at 25°C of n-heptane and methyl cyclohexane may be
taken as 46 and 37 mm of Hg, respectively. Neglecting the vapor pressure of
aniline, calculate the composition of the vapors and the total vapor pressure in
equilibrium with the solution of part (a). Compare the effective relative volatilities
[a = Pi-Yi/P^yi) of the heptane and methyl cyclohexane in the presence and absence
of the aniline.
7. Calculate the vaporization equilibrium constant of acetone at a temperature
of 300°F and an absolute pressure of 250 lb per sq in. The density of acetone at
20°C is 0.7915 g per cc. The vapor-pressure equation is given in Table Ha, page 73.
8. A stream of gas in a natural gasoline plant has the following composition in
mole per cent: Ethane 10
Propane 14
Isobutane 19
n-Butane 54
Isopentane 3
100
(a) Calculate the pressure necessary to condense this gas completely at a tem-
perature of 100°F.
(6) For a condenser operating at the pressure of part (a), calculate the temperature
at which condensation will start and the temperature of 50 mole per cent condensa-
tion. Also calculate the composition of the first liquid to condense and the com-
positions of the hquid and vapor phases at 50 per cent condensation.
2°K. A. Varteressian and M. R. Fenske, Ind. Eng. Chem., 29, 270 (1937).
9. A fractionating column is to produce overhead and bottoms products having

the following compositions in mole per cent:
Overhead Bottoms
Propane 23
Isobutane 67 2
7i-Butane 10 46
Isopentane 15
n-Pentane 37
100 100
(a) Calculate the pressure at which the column must operate in order to condense
completely the overhead product at 120°F.
(6) Assuming that the overhead product vapors are in equilibrium with the liquid
on the top plate of the column, calculate the temperature of the overhead vapors
and the composition of the liquid on the top plate when operating at the pressure
of part (a). .
(c) Calculate the temperature of the liquid leaving the reboiler. of the column
under the pressure of part (a), assuming equilibrium conditions to exist in the re-
boiler. Also calculate the composition of the vapors leaving the reboiler.
10. The heat of fusion of a-naphthylamine (CioHsN) at its melting point of 49°C
is 3,150 cal per g-mole. Calculate the solubility at 30°C, assuming ideal behavior.
11. Calculate the ideal solubility of hydrogen cyanide gas in acetonitrile at a tem-
perature of 100°C and a partial pressure of 5 atm. The density of liquid hydrogen
cyanide is 0.699 at 20°C. The vapor pressures at 100°C are 9.2 and 1.89 atm.
respectively.
12. Calculate the ideal solubility of chlorine in carbon tetrachloride at a partial
pressure of 10 atm and a temperature of 100°C. The vapor pressures at 100°C
are 37.6 and 1.92 atm. respectively.
13. The solubiUty of carbon monoxide in benzene at 20°C under a partial pressure
of 1 atm is reported^ as 0.1533 cc of gas measured at 0°C and 1 atm, per cc of liquid
benzene. Assuming that the activity coefficients are unity, calculate the solubility
in mole per cent at 20°C and 50 atm. Neglect the vapor pressure of benzene.
14. The following data for the solubility of carbon dioxide at a partial pressure of
1 atm in methyl alcohol are reported'' in cubic centimeters of gas measured at 0°C
and 1 atm per cc of hquid.
fC Solubility
15 4.366
20 3.918
25 3.515
(a) Assuming that ideal solutions are formed, calculate the average differential
heats of dissolution of CO 2 in methyl alcohol in the temperature ranges of 15°-20°
and 20°-25°C. Neglect the vapor pressure of alcohol.
(6) Prom the results of part (a), calculate the solubility at a temperature of SS'C
and a pressure of 100 atm.
15. Calculate the minimum total work in Btu in separating 100 lb of a solution
at 80°F containing 30 mole per cent benzene, 25 per cent toluene and 45 per cent
xylene into the three pure components. Also calculate the entropy change per pound
of solution.
16. Calculate the minimum work and the entropy change in separating 100 lb of
a 2-molal aqueous solution of sucrose into its components at 26°C. Compare this
•^ result with the heat required to vaporize the water at the same temperature.
' CHAPTER XVI
CHEMICAL EQUILEBRIUM
As explained in Chapter XI the criterion of equilibrium in a chemically
reacting system at constant temperature and pressure is that the change
in free energy of any possible reaction shall be zero. A negative free-
energy change may be looked upon as the driving potential which is
directing the reaction towards a state of rest and as a direct measure
of the departure of the reacting system from its equilibrium state. From
a kinetic viewpoint the reaction at equilibrium may still be considered
as proceeding reversibly, but with equal rates in opposite directions and
with no net change in composition.
By evaluation of the free-energy changes of the reactions of a chemical
process it is possible to calculate the composition of the equilibrium
mixture and to determine the extent of conversion of the initial reactants.
Such considerations are essential in determining the most favorable con-
ditions of temperature, pressure, composition, and ratios of reactants to
obtain the greatest conversion of reactants and the highest yield of
products.
THE EQUILIBRIUM CONSTANT
For developing the equations of chemical equilibrium the general re-
action represented by the following stoichiometric equation is considered:
bB + cC+---=rR + sS+--- (1)
where b, c, r, s are the number of moles of reactants and products B, C,
R, and S, respectively.
When this reaction proceeds isothermally at any temperature T,
starting with each reactant in its standard state of unit activity and
ending with products each at unit activity, the accompanying change of
free energy is represented by the symbol AG°. The corresponding ac-
tivity a j , ac, a|, a% of each component in the standard state is unity.
When this reaction proceeds isothermally until equilibrium is estab-
lished all the activities adjust themselves to new values at which the
change in free energy is zero. At equilibrium the activities of the sep-
arate components are represented by symbols, as, ac, aR, and as-
Since the free-energy change is the difference between the free energies
of the products and reactants,
AG° = rG^ + sG^-{ bm-cG^ (2)
AG = rGij -|- sGs -1 bGB — cGc (3)
where 5 = partial molal free energy
691
692 CHEMICAL EQUILIBRIUM [CHAP. XVI
Gombining (2) and (3) gives

AG-AG° = r(GB-Gk)+siGs-G;) + b(GB-QB)~c(Gc-Gh) (4)
Equation (4) may be written in terms of activities by being combined
with Equation (XIV-53). Thus,
AG - AG° = rRT In an + sET In as^ hRT In as - cRT In ac
( ttR al • • A
-T^ ) (5)
At equilibrium, AG. = 0, and '
(6)
T \al a§---J
where K is the equilibrium constant at the temperature corresponding to
the temperature of AG°. Thus,
r , -AG° AH° AS°
j ^ ^ ^ o ^ ^ - ^ ^ - — ^ - ^ (7)
fljB a c • • •
where AG°, AH°, and AS° are the changes in free energy, enthalpy, and
entropy, respectively, which accompany the stoichiometric reaction with
all reactants and products in their standard states.
From Equation (7) it is evident that the equilibrium constant is en-
tirely determined by the temperature and the free-energy change which
would accompany the reaction of the indicated numbers of moles if each
reactant were initially in its standard state and each product finally in
its standard state of unit activity at the temperature of the system. This
free-energy change AG" is termed the standard free-energy change of the
reaction. The standard free-energy change depends upon the tempera-
ture, the definition of the standard state of each component, and the
number of moles entering into the stoichiometric equation under con-
sideration. Accordingly, the mmierical value of an equihbrium constant
is without significance unless it is accompanied by specification of these
three factors.
The effect on the equilibrium constant of the form of the stoichio-
metric equation is illustrated by consideration of the synthesis of
ammonia. This reaction may be designated as
N2 + 3H2 = 2NH3
The corresponding equilibrium constant is then
CHAP. XVI] STANDARD FREE-ENERGY CHANGES 693
The same reaction may be written
and K = —, ;
aL AHI
I t is evident that, in this case, K' = K^ and specification of the number

of moles involved in tlie stoichiometric equation is essential to proper
interpretation.
r Effect of Pressure on the Equilibrium Constant. The effect which the
pressure of the systen^ has on the reaction equilibrium constant is de-
pendent on the definition of the standard states of the components. As
pointed out in Chapter XIV, the choice of the standard states is entirely
arbitrary and may be on any convenient basis. When working with
reacting systems which involve only pure solids or liquids and gases
which may be assumed to approximate ideal solutions it is convenient
to refer the activities of the gaseous components to the ideal-gaseous
state at one atmosphere, making activities equal to fugacities in atmos-
pheres. The activities of the pure soUds and Hquids are referred to the
fugacities of the pure components under their own vapor pressures or
under a pressure of 1 atm at the temperature of the system. With stand-
ard state defined in this manner the reaction equilibrium constant is in-
dependent of the pressure of the system which by virtue of the definitions
can have no effect on the standard free-energy change.
When working with reacting systems which involve nonideal solutions,
either gaseous or liquid, it is more convenient to select as standard states
the pure components at the temperature and pressure of the system. As
is developed in Chapter XIV, this basis has the advantage of permitting
useful correlations between mole fractions and activity coefficients.
Similarly, if it is desired to refer the standard state to infinite dilution,
it is convenient to choose the dilute solution at the pressure of the sys-
tem. It should be noted that standard states defined in this manner are
dependent upon the pressure of the system and that the corresponding
equilibrium constants vary as a function of pressure.
I t is of primary importance that the exact definition of the standard
state be clearly recognized in all equilibrium calculations.
STANDARD FREE-ENERGY CHANGES

I t is pointed out in Chapter X I that free energy is an extensive prop-
erty and that the free-energy change in any process is determined by
the final and initial conditions and not by the intermediate path. Thus,
free-energy changes may be treated in the same manner as enthalpy
changes and are functions of temperature, pressure, and composition at

the beginning and end of the change.
The free-energy change accompanying a reaction as it actually pro-
ceeds under the conditions of the reaction process is of little interest
except in the consideration of electrochemical reactions where the de-
crease in free energy of the system represents the electric energy released
under reversible conditions lat constant pressure. Also, it has been,
pointed out in Chapter X I I I that the decrease in free energy of a flow
process proceeding reversibly at constant temperature without genera-
tion of electric energy is equal to the mechanical work done. Usually
chemical reactions proceed in a highly irreversible manner \vithout gen-
eration of either electric energy or other useful^work, and the loss in free
energy is not accompanied by a corresponding release of useful work.
The greatest value of the free-energy concept is in the calculation of
compositions of reaction systems at equilibrium under which conditions
the actual free-energy changes are zero. These compositions are, how-
ever, related by Equation (6) to the standard free-energy changes which
are not equal to zero at the standard states. In considering any changes
which take place in a system actual and standard free-energy changes
must not be confused. Calculation of the standard free-energy change
is important, because from it the equilibrium constant is obtained.
Standard Free-Energy Change at 25°C. The standard free-energy
changes and the corresponding standard enthalpy and entropy changes
which accompany a reaction at an arbitrary reference temperature may
be calculated by the summation principles demonstrated in Chapter VIII
for heats of reaction. The accepted reference temperature which has
been extensively used for thermodynamic data is 25°C (298.1°K)
instead of the 18°C standard frequently employed for heats of reaction.
It has been common practice in the past to present standard free-
energy data in the form of tables of standard free energies of formation
at 25°C. From such tables a standard free energy of reaction is cal-
culated as the algebraic sum of the standard free energies of formation
of the products less the algebraic sum of the standard free energies of
formation of the reactants. When an element enters into a reaction,
its standard free energy of formation is zero if its state of aggregation
is that selected as the basis for the standard free energy of its compounds.
Illttstration 1. Calculate the standard change of free energy at 298.1°K in the
gas-phase alkylation of isobutane with ethylene to form neohexane. Free energies
of formation at a fugacity of 1 atm and 298.1°K (cal per g-mole) '
C4H10, isobutane (gas); Aa} = —4,160 = «„
C2H4, ethylene (gas); AG/ =i + 16,300 = Gj
~C6Hi4, neohexane (gas); AG/ = —2,300 = a.
For the reaction,
CiHio (g) + C2H4 (g) = CeHn (g)
CHAP. XVI] EFFECT OF TEMPERATURE ON AG" AND K 695
Aff^., = G r - G „ - G j , = - 2 , 3 0 0 - ( - 4 , 1 6 0 ) - 1 6 , 3 0 0 = -14,400 cal per g-mole in

the ideal state at 1 atm.
Effect of Temperature on AG° and K. A general differential relation-
ship between AG° and temperature is expressed by Equation (XI-104).
Combination of this wdth Equation (6) yields a similar expression for the
effect of temperature on the equilibrium constant. Thus
/a I n / A _ ^H°_
\ dT ) ~ T^
Equation (8) may be used to predict the effect of temperature on the

equilibrium constant or conversely to determine the standard heat of
reaction from equilibrium constants known at different temperatures.
Such a determination is carried out graphically by determining the slope
of a curve relating fl In E^ to 1/T.
' A more generally useful integrated relationship between the standard
free-energy change and temperature is obtained by expressing the stand-
ard heat of reaction and entropy change in terms of empirical heat-
capacity equations. If heat capacities are represented by the general
empirical] expression cl = a+bT -{- cT^ + d/T^, Kirchoff's Equation
(VIII-65), page 305, may be written:
AH'r = lH + AaT + ÂbT^ + iAcT'-Ad/T (9)
The integration constant Is is determined from a single value of AH°
at a temperature within the range of applicability of the heat-capacity
equations.
Since at constant pressure dS = Cp dT/T, it follows that
d{AS°)=Ac°dT/T or ASl = Is + Aa\nT + AhT + iAcT^-~ (10)
The integration constant / » i s determined from a single value of AS° at
any temperature within the range of the accuracy of the heat-capacity
equations.
Since AG° = AH° - TAS°, from Equations (9) and (10),
AG° IH , , , , , ^ AhT AcT^ Ad
— = -+(Aa-7s)-Aolnr ^ " 2 2 ^ dD
The integration constants 7^ and 7s are determined from a single value
of AH° together with a single value of either AS° or AG°. These reference
values may be at different temperatures.
The forms of Equations (9), (10), and (11) depend upon the empirical
equations used for expressing heat capacities. Other equations may be
used with corresponding changes in the form. The standard free energy
of formation of a compound from its elements at any temperature level

is a special modification of Equation (11) a's illustrated in the following
illustration:
Illustration 2. Derive a general equation for the standard free energy of formation
ofiNH3(g)Iat temperature r ° K with the standard state of each component being the
ideal gas at 1 atm. The absolute entropies in the ideal-gaseous state at 298.1°K and
1 atm are as follows: * -
N2 (g); 45.79 = s^ cal/(g-raole)(°K)
H2 (g); 31.23 = Hi cal/(g-mole)(°K)
NHs (g); 46.03 = s? cal/fe-mole)(°K)
From Table V, page 214,
N2; cp = 6.30 + 1.819(10-')r - (0.345)(10-<)!r2
H2; Cp = 6.88 + (0.066)(10-s)r + (0.279) (lO-*)^^
NH3; Cp =-5.92 + (8.963)(10-5)r - (1.7&i)(10-')T^
For the reaction,
JN2(g)-f-|H2(g)->NH,(g)
AH291 = -11,000 cal per g-mole NH3
Aa = 5.92 - K6.30) - 1(6.88) = -7.55
Ah = [8.963 - i(1.819) - f (0.066)]10-3 = (7.955) (lO"')
Ac = [-1.764 + ^(0.345) - f (0.279)]10-« = (-2.010)(lO"')
Ad = 0
IH = -11,000 + 7.55(291) - 3.977 X 10-3(291)2 + 0.670(10-«)(291)' = -9130
A^sa.i = s? - K - fsb = 46.03 - ^45.79) - 1(31.23) = -23.72 ,
From Equation (10),
AS2°s8.i = / s + Aa In r + M)T + iAcT^
-23.72 = Is -7.55(5.6971) + 7.955 X 10-H298.1) -K2.01)(10-«)(298.1)2
or Is = 17.00
(Aa - Is) = -7.55 - 17.00 = -24.55
AG° -9,130
— = —~- + 7.55 In r - (3.977)(.10-3)r + (0.335)(10-<)r2 -24.55
The standard free-energy change of any reaction at any temperature

level may be obtained by the method of Illustration 2 in which Equa-
tion (11) is applied directly to the reaction combined with a knowledge
of heat-capacity equations of the components and single values of AH°
and AG° or AS° for the-ieaction, each at some known temperature. An
alternate method is to evaluate the free energies of formation as functions
of temperature and calculate the free-energy change of the reaction by
combining the standard free energies of formation of the separate com-
ponents. There is little point in this latter procedure except where
numerous different reactions which invplve the same components are
CHAP. XVI] FREE ENERGIES OF HYDROCARBONS 697
For reactions in which ACp = 0 both AH° and A»S° are independent of
temperature. Equation (11) then reduces to
AG° = AHl - T AS°i (12)
From Equation (12) it is evident that the standard free-energy change
of a reaction may vary widely with temperature, even though the stand-
ard heat of reaction and entropy change are constant.
Where the heat-capacity data for a reaction are not conveniently ex-
pressed by empirical equations the standard free-energy change at any
temperature may be calculated by correcting separately the standard
enthalpy and entropy 'changes by graphical integrations. Thus, if ACp
is the difference between the heat capacity of the products and the re-
actants, and it is desired to obtain the standard free-energy change at a
temperature Tz from values of standard enthalpy and entropy changes
AHl and AJS2 at temperature Ti and T2, respectively.
AC; dT - Tz ^Sl -Tz ~dT (13)

Ti JTI 1
The integrations of Equation (13) may be carried out graphically by
plottmg ACl and AC°p/T against T.
Standard Free Energies of Formation of Hydrocarbons. Standard free
energies of formation of a few hydrocarbons are shown in Fig. 155, modi-
fied from a similar figure of Parks and Huffman.^ In this figure the
molal free energies of formation are expressed per carbon atom in order
that the figure shall show visually the relative stabilities of the different
compounds in their standard states with respect to the elements. For
example, the fact that the acetylene curve (C2H2) is positive indicates
that over the entire temperature range this compound is thermodynami-
cally unstable in its standard state and tends to decompose spontane-
ously to an equilibrium mixture containing large proportions of carbon
and hydrogen. Similarly, since the curve for propane (CsHg) is higher
than that for ethane (CsHe), the latter compound is the more stable.
Propane becomes relatively unstable at temperatures above 450°K.
The standard free energies of formation of the hydrocarbons are ob-
tained by multiplying the values from Fig. 155 by the number of carbon
atoms. The standard free-energy changes of reactions involving these
hydrocarbons may then be calculated by summations of the standard
free energies of formation.
> Parks and Huffman," The Free Energies of Some Organic Compounds," Reinhold
Publishing Corporation, New York (1932).
Illustration 3. Calculate the standard free-energy change of the following reaction

at 1000°K from the data of Fig. 155:
2C2H4 = iC^Ha
iC4H8; AG; = (4) (15,300) = 61,200 cal per g-mole
CjH,; AG; = (2)(14,000) = 28,000 cal per g-mole
For the reaction at 1000°K,
AG" = 61,200 - 66,000 = 5,200 cal
35r
"^300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Temperature, ° K
(Revision of chart from Parks and Huffman, " Free Energies of Some Organic Compounds,"
with permission of the Heinhold Publishing Corporation.)
FIG. 155. Standard Free Energies of Formation of Hydrocarbons.

Experimental Free-Energy Data. Four common methods are avail-
able for determining standard free energies of formation and the stand-
ard free-energy change of a chemical reaction.
1. For reactions which proceed reversibly in a galvanic cell at constant
pressure and temperature, the free-energy change can be obtained di-
rectly from measurements of electric energy generated.
2. The standard free-energy change may be calculated from meas-
urements of temperature, pressure, and composition of each phase in a
system at equilibrium by means of Equation (6).
3. The standard change in free energy can be obtained directly from
measurements of AH° and AS° by the equation-AG° = AH°-TAS°.
CHAP. XVI] PRESENTATION OF FREE-ENERGY DATA 699
Values of ^H° are determined from thermochemical data, as described

in Chapter VIII. Values of A(S° are obtained from the absolute en-
tropies of the products and reactants. These absolute entropies are
calculated either from low-temperature heat-capacity measurements and
the third law of thermodynamics, as discussed in Chapter XI, or from
spectroscopic data and statistical calculations, as outlined in Chapter XVI.
4. Free energies of formation and reaction may be obtained through
combination of free energy data for other series of reactions by the pro-
cedure demonstrated in Chapter VIII for the indirect evaluation of heats
of reaction.
Presentation of Free-Energy Data. Three methods are in common
use for the presentation of standard free-energy data. The earliest
method was the tabulation of. standard free energies of formation at an
arbitrary base temperature of 25°C together with the constants for the
empirical heat-capacity equation of each compound. From these data
standard free-energy changes may be calculated at any desired tempera-
ture by the methods demonstrated in Illustration 1.
An alternate method is to present tables of heats of formation and
absolute entropies at a selected base temperature. By combining these
values the standard free energies of reaction at the base temperature
are obtained. Free energies at other temperatures are then calculated
from the heat-capacity equations. This method has the advantage of
separating the heat of formation and entropy data on which the free-
energy calculations are based. Changes are constantly being made in
the accepted values, and tabulation of separate heats of formation and
entropies minimize the confusion resulting from such changes. Further-
more, the entropy values are frequently useful for other calculations.
In Table XXXV are values of absolute entropies at 25°C. Similar
data for many other compounds are to be found in the literature. The
use of such data in conjunction with the heats of formation and reaction
of Tables XIV and XV, pages 253 and 262, is demonstrated in Illustra-
tion 2.
With the extension of calorimetric measurements into the region of
low temperatures and the development of statistical methods based on
spectroscopic data it has become possible to express values of enthalpy
and free energy of many substances relative to their values at the ab-
solute zero of temperature. A further improvement has been introduced
by reporting values of (H? —HO) and (GJ —Go)/r as direct tabular
or graphical functions of temperature rather than formulating the results
in terms of the constants of empirical heat-capacity equations. Con-
sistent with this practice heats of formation (AH/)O at 0°K are now being
reported relative to the elements at 0°K instead of at some arbitrary
temperature level. By this scheme the agreement of different values
can be judged directly by numerical values rather than by the com-

parison of different equations which appear in as many forms as there
are empirical heat-capacity equations for the separate components. The
relative agreement of several forms of equations can be judged only by
repetition of calculations and usually not by inspection.
The standard heat of formation of a compound relative to the elements,
all in their standard state of unit activities at any temperature level, is
expressed by the following equation:
(AH?)J. = [(H^ - nl) + {An})o]a - S ( H ? - H^)^ (14)

where
(AH/)r = standard heat of formation at temperature T
Hr = enthalpy of the compound or element at temperature T
Ho = enthalpy of the compound or element at 0°K
(AH/)O = standard heat of formation at C K
The subscripts C and E refer to the compound and to the element,
respectively. At the absolute zero where the temperature, entropy,
and vapor pressure all become zero the four energy functions of a
substance in a specified state become equal. Thus
vl=sl = Gl=Al (15)
As a result it is common practice to report data in terms of the exactly
equivalent functions (G^ —Go)/r, {GT — 'BD/T, or (Oy — U Q ) / ! " , and
(AH/)O or (AG/)O.
The standard free energy of formation of a compound at a tempera-
ture T from the elements at the same temperature is expressed by an
equation similar to (14).
Similarly, the standard free-energy change and standard heat of re-

action of any reaction at any temperature may be expressed in terms
of heats of formation at the absolute zero. Thus,
A H ; = S [ ( H ; - H°o) + (AH?)o]i> - 2[(H? - HS) + (AH})O]« (17)
/AG°\ _ jG°r - Hg (AH;)O-| JG^-Hg (AH?)o1 ....
\T)T' I T + T jp~\ T + ~ r J« ^ ^
where the subscript P indicates the products and R the reactants.
Standard enthalpy and free-energy changes are readily calculated
by Equations (17) and (18) from values of (AH/)O and tables or
curves which express (H? — HQ) and (Gr — ^D/T as functions of
temperature. It will be noted that for any compound the group
\o^tf.
CHAP. XVI] STANDARD MOLAL ENTROPIES 701
TABLE XXXV
STANDAED MOLAL ENTROPIES AT 298. 1°K
(cal)/(g-mole)(°K)
Standard States:
Gases: The ideal-gaseous state at 1 atm
Liquids: The liquid state at 1 atm
Solids: The solid state at 1 atm
Aqueous Solutions: The hypothetical ideal 1.0 molal solution in which o/m = 1.0
INORGANIC COMPOUNDS
From K. K. Kelley, U£. Bur. Mines Bull. 434 (1941).

t
Aluminum
Al(8) 6.75 AI2O3 12.5
Al (g) 39.31 AljSiOe (Sillimanite) 27.0
A1+++ (aq) -76
Antimony
Sb(s) 10.5 SbaO, (s) 30.4
Sb(g) 43.07 SbiOs (s) 29.9
Sb, (g) 60.9 SbjSs (s) 39.6
SbjOa (s) 29.4 SbCl3 (g) 80.9
SbiO, (g) 102 SbCl, (s) 44.8
Argon
A(g) 36.99
Arsenic
As(s) 8.4 AszOs (s) 25.2
A8(g) 41.62 AsCla (g) 78.2
As2 (g) 67.3 AsCla (1) 55.8
AS4 (g) 69 AsF3 (g) 69.2
AS2O3 (s) 25.6 AsFa a) 50.1
AS4O6 (g) 101
Barium
Ba(g) 40.67 B a F j (s) 23.0
Ba+j+ (aq) 2.3 BaCOs (s) (Witherite) 26.8
BaO (s) 16.8 Ba(N03)2 (s) 51.1
BaO (g) 56.0 BaS04 (s) 31.6
BaCU • 2H2O (s) 48.5
Beryllium
Be(s) 2.28 BeO (s) 3.37
Be(g) 32.56 BeO(g) 47.2
Bismuth
Bi(s) 13.6 BijOa (S) 36.2
Bi(g) 44.68 BiCls (g) 86.5
BU(g) 65.4 BiCls (s) 45.8
Boron
B(s) 1.7 BCI3 (g) 68.6
B(g) 36.66 B F , (g) 61.1
BBra (g) 76.7 B i C (s) 6.47
TABLE XXXV {Coniirwed)

Bromine
Br2 (1) 36.7 Br- (aq) 19.7
Br^Cg) 58.63 BrCl (g) 57.4
41.81 BrOr (aq) 38.5s
Br(g)
Cadmium
12.3 CdCU (s) 31.2
CdSg) 40:07 Cd(OH).(s) f
Cd-^(aq) • -15.6 CdCO, (s) 25.2
CdO(s) 13.1 CdS04(s) 31.3
Cdoi) 46.9 CdSO.-H.O(s) 39.7
Cdsg IV CdSO.-8/3H,0(s) 57.5-
Calcium
Ca (s) \ 9.95 Ga(OH)j (s) 17.4
37.0 CaCOa (s) (Calcite) 22.2
Ca(g)
Ca++(m) -11.4 CaCOa (s) (Aragonite) 21.2
CaOC) 9-5 CaC.O.-H.O(s) 37.4
CaO g 52.3 Ca3(P0,).(s)(a) 57.6
P_o fg^ • 13.5 CaSiOa (s) (WoUastonite) 19.6
Q^Y^ls) 16.4 CaSOi (s) (Natural Anhydrite) 25.5
CaHj (8) 9-9
Carbon
C (Graphite) 1.36- CF^ (g) 62.7
C (Diamond) 0.585 COCU 67.2
37.77 CH4 44.5
C(g) CN (g) 48.42
47.89
CO (e) 47.32 C2N2 (g) 57.9
Z% 51.08 CNBr(g)
CS(g) 50.3 CNCKg) 56.4
CS.a 36.2 CNI(g) _5-3
OS, (a) 56.84 CO3- (aq) -1^-"
cost) 55.37 HCOr(aq) 22.2
CBr.fe) 85.5 H.CO, (aq) 45.1
CCU(g) 74.2 CN (aq) 25
e c u (1) 52.2 CsOr (aq) «-°
Chlorine
CU(g) 53.31 ClO-(aq) 10.0
C1(g) • 39.47 ClOr(aq) 24.1
01-(iq) • 13.5 ClOr(aq) 39-4
CUO(g) 63.7 ClOr(aq) 43.6
CIO2 (g) 59.6
Chromium
Cr(8) 5.68 CrCU(8) 27.7
Cr(g) 41.64 CrCl3(s) \ 31.0
CrjOs (s) 19.4 CrOr (aq) 19-5
CHAP. XVl] STANDARD MOLAL ENTROPIES 703
TABLE XXXV {Continued)
Cobalt
Co (s) 6.8 C0CI2 (S) 25.4

Co(g) 42.89
Copper
Cu (s) 7.97 CuBr (s) 22.8

Cu(g) • 39.75 CuBr (g) 59.9
Cu^ (g) 58.9 CuCl (s) 20.8
Cu++ (aq) -26.5 CuCl (g) 57.3
CusO (s) ,24.1 C u F (g) 54.9
CuO (s) 10.4 C u l (s) 23.1
CujS (s) 28.9 C u l (g) 61.6
CuiS (s) (CovelUte) 15.9 C u H (g) 46.9
CuCOs (s) 17.7 CUSO4 (S) 25.3
Fluorine
F2(g) 48.58 ¥- (aq) -2.3

F(g) 37.93 F2O (g) 58.95
Helium
He 30.13
Hydrogen
H2 31.23 H2S (g) 49.15

D, 34.62 H B r (g) 47.48
H(g) 27.40 H C l (g) 44.66
H+ (aq) 0 HF(g) 41.53
H2O (1) 16.75 HI(g) 49.36
H2O (g) 45.13 H C N a) 26.96
D2O (1) 18.08 HCN:(g) • 48.23
DÔCg) 47.38
rt Iodine
12(8) 27.9 I B r (s) 33.0

12(g) 62.29 ICl(g) 59.11
1(g) 43.19 ICI3 (S) 41.1
I - (aq) 25.3 lOi" (aq) 28.0
IBr(g) 61.84
Iron
Fe(8) 6.47 FeS (s) 16.1

Fe(g) 43.12 FeSj (s) (Pyrite) 12.7
Fe++ (aq) -25.9 FeCU (s) 29.4
Fe+++ (aq) -61.0 F e , C (s) 24.2
FeO (s) 13.4 FeCOs (s) (Siderite) 22.2
FejOs (s) 21.5 FeiN 37.4
Fea04 (s) 35.0
CHEMICAL EQUILIBRIUM [CHAP XVI
704
TABLE XXXV {Cmtinued)
Lead
50.5
15.49 Pb304 (S) 21.8
Pb(s) 41.90 PbS (s) 61.2
Pb(g) 3.9 PbS (g) 32.6
Pb++ (aq) PbCU
PbO (s) (YeUow) 16.6
16.6 ^^^^ (s) 75.9
5t.4 PbCU (g)
PbO(g) ^'-^ '-
PbSOi (s) ^^-^^
18.3
PbOj (s)
lAthium
9.7
6.70 LiF (s) 56.5
Li(8) 33.15 Lil (g)
Li(g) 5.9
47.0 LiH (s) 40.8
Lij (g) LiH (g)
4.7 12.8
Li+ (aq) LiOH (s)
49.9
LiCl (g)
LijCOj 21.5
Magnesium
15.09
7.77 Mg(0H)2 (s) 15.7
Mg(s) 35.51 MgCOs (s) (Magnesite) 15.4
Mg (.g) -_3q1i. 66 ivigoiws (s)
MgSiOs v>j on
Mg^Maq) 26.55
^1, M,S04(s)
MgS04 (s) 2°
MgO (s)
50.7
MgO (g)
Manganese
22.9
7.61 MnjOs (s) 35.5
Mn(s) Mn304 (s)
41.50 18.7
Mn (g) MnS (s)
14.4 25.9
MnO (s) MnCU (s)
54.0
MnO(g) ^-g MnCOa (Rhodochrosite) ^"-^
13.9 MnCOa (Rhodochrosite)
MnOj (s) ^„ ^ ivrnP,
MnC
46.7
MnOr (aq)
Mercury
74.7
18.5 HgBra (g) 38.9
HgO) 41.8 HgBrj (s) 70.4
Hg(g) 17.7 HgCU (g) 34.6
Hg2 l,aqj 16.6 xigvia (s)
HgCU vn; 707
17.6 Hgla (g)
26.4 H g l j (s)
HgBr (s) 48.0
23.5 HgjSOi (s)
HgCl (s)
Molybdenum
15.0
6.83 M0S2 (s) 15.9
Mo (s) M0S3 (s)
43.47
Mo (g)
\ Neon
34.96
Ne(g) Nickel
25.6
7 12 NiClj (s) 97
Ni (8) 43 53 Ni(C0)4 (g)
Ni (g) n r,
NiO (8) ^-^
TABLE XXXV [Coniinued)
Nitrogen
N2(g) 45.79 NOCl (g) 63.0
N(g) 36.62 NH3 (g) 46.03
NsOfe) 62.58 NH4CI (S) 31.8
NO(g) 50.34 NH4HCO3 (S) 28.3
^ O , (g) 57.47 NH4HS {%) 27.1
NÔi (g) 72.7 NH4OH (aq) 42.8
N2O5 (S) 36.6 NHi" (aq) 26.4
N2O6 Gg) 82 NO7 (aq) 29.9
NOBr (g) 65.2 NOr (aq) 35.0
HNO3 (liq) 37.19
Oxygen
0^(8) 49.03 OH(g) 43.9
0(g) 38.48 OH- (aq) -2.49
03(g) 57.1
Phosphorus
62.1 PCls (s) 40.8
P«(g)
P4(g) 66.8 PF3 (g) 64.2
P (s) (White) 10.6 PH3 (g) 50.35
38.99 PN(g) 50.44
P(g)
83.4 H3PO4 (aq) 44.0
PBra (g)
74.7 HiPOr (aq) 28.0
PCI3 (g)
52.2 HPOr (aq) -2.3
PCI, (1)
87.7 POf (aq) -45
PCl6 is)
Potassium
K(s) 15.2 KCl (g) 57.7
K(g) 38.30 KF(g) 56.7
K,(g) 59.5 ' KI (s) 24.1
24.2 KH(g) 47.3
K+ (aq)
22.6 KCIO3 (S) 34 2
KBr (s)
60.4 KNO3 (S) 31.8
KBr (g)
19.76 KSOi. (s) 44.8
KCl^(s)
Silicon
Si(s) 4.50 Si02 (Cristobalite) 10.35
40.13 SiOa (Tridymite) 10.5
Si(g)
10.1 SiOj (Glass) 11.2
SiOz (Quartz)
79.2 SiH4 (g) 48.7
SiCU (g)
57.3 SiC (s) 3.95
SiCU 0)
SiF4(g) 68.0 SisNi (s) 22.8
Silver
10.20 AgBr (g) 62.1
Ag(s)
41.33 AgCl (g) 59.8
Ag(g)
17.54 AgCl (s) 23.0
Ag+ (aq)
29.1 AgNOa (S) 33.7
AgiO (s)
35.0 AgsSOi (s) 47.9
AgaS (a)
25.60 Ag(NH3)r (aq) 57.8
AgBr (s)

Sodium
Na (s) 12.2 NaH (g) 45.0
Na(g) 36.72 NaOH (s) 14.2
Nas (g) 54.9 NajCOs (s) 32.5
Na+ (aq) 14.0 NaHCOs (s) 24.4
NaÔ (g) 17 , NaNO, (s) 27.8
NaBr (g) 58.1 NaiSiOi (s) 46.8
NaCl (s) 17.3 NajSiOs (s) 27.2
NaCl (g) 55.5 NasSijOs (s) 39.4
NaF (s) 13.1 NaiiSOi (s) 35.7
NaF (g) 53.8 n&sOi • 10H2O (s) 140.5
Nal (g) 60.0 NaK 68.6
Sulfur
S (rh) 7.62 SOs (1) 31.7
S (mono) 7.78 SFe (g) 69.6
S(g) 40.1 H2S (aq) 29.4
S^Cg) 54.41 HS- (aq) 14.9
S6(g) 92 • H2SO3 (aqi 54.7
SsCg) 109 HSOr (aq) 32.6
SO(g) 53.07 SOT (aq) 3
BO, (g) 59.24 HSOr (aq) 30.6
SO, (g) 63.8 SOf (aq) 4.4
Tin
Sn (white) 12.3 Sn(g) 40.24
Sn (gray) 10.7 Sn++ (aq) -4.9
SnO (s) 13.5 SnCU (g) 87.2
Sn02 (s) 12i5 SnCU (l) 62.1
Titanium
Ti(s) 6.6 TiCU (g) 84.4
Ti(g) 43.07 Ticu a) 60.4
TiOj (Rutile) 12.4
Vajiadium
V(s) 7.0 V2O4 (S) 24.5
V(g) 43.55 V2O5 (S) 31.3
V2O3 (S) 23.5
Zinc
Zn (s) 9.95 ZnBrj (s) 33.0
Zn(g) 38.46 ZnCU (s) 25.9
Zn++ (aq) -25.7 Znlj (s) 38.5
ZnO (s) 10.4 ZnCOa (Smithsonite) 19.7
ZnO (g) 54.1 ' ZnSOt (s) 30.7
ZnS (s) 13.8
Zirconium'
Zr(8) 9.5 Zr(g) 43.33
ORGANIC COMPOUNDS
Ideal-gaseous state at 1 atm unless otherwise noted

Hydrocarbons
5° Ref. S' Ref.
Methane 44.46 28 Styrene (1) 56.78 11
Acetylene 48.08 28 Ethyl benzene (1) 61.3 26
Ethylene 52.75 28 Nonane (1) 94.0 26
Ethane 54.86 28 Mesitylene 92.34 29
Methyl acetylene (allylene) 56.83 20 n-Butyl benzene (1) 76.9 26
Propadiene (allene) • 66.04 22 Neopentane 73.23 28
Propylene 65.06 28 Isopentane 82.03 28
Propane 64.69 28 Cyclopentane 70.70 4
Cyclopropane 56.79 21 Cyclopentane (1) 48.87 4
Dimethyl acetylene 67.93 33 «-Hexane 92.45 28
Dimethyl acetylene (1) 46.63 33 2-Methyl pentane 90.15 28
Isobutene 71.28 28 2-Methyl pentane (1) 69.9 28
cis-2-Butene 73.58 28 2, 2-Dimethyl butane 85.75 28
frons-2-Butene 72.98 28 2, 2-Dimethyl butane 64.4 28
1-Butene 75.38 28 2, 3-Dimethyl butane 86.49 28
Butane 74.21 28 3-MethyI pentane 89.96 28
Butane (1) 55.2 28 Benzene 64.39 29
Isobutane 70.51 28 Benzene (1) 41.49 29
Isobutane (1) 52.09 28 Cyclohexane 71.41 6
Trimethylethylene 79.7 18 Cyclohexane (1) 49.3 6
Pentane 83.23 28 Methyl cyclopentane 0) 59.3 26
2-Methyl hexane 99.48 28 n-Heptane 101.57 28
Methyl cyclohexane (1) 59.4 26 Naphthalene (s) 39.9 26
2, 2, 3-Trimethyl butane 92.31 28 Decane (1) 102.7 26
Toluene 76.44 29 |3-Methyl naphthalene (s) 48.8 26
Toluene (1) 62.40 29 Duodecane 147 16
Ethyl cyclopentane (1) 67.1 26 Diphenyl (s) 49.2 26
n-Octane 110.69 28 n-Heptyl cyclohexane (1) 106.8 26
2, 2, 4-Trimethyl pentane 101.35 28 Anthracene (s) 49.6 26
2, 2, 3, 3-Tetramethyl butane 94.05 28 Phenanthrene (s) 50.6 26
o-Xylene 84.50 29 Pyrene (s) 51.4 26
o-Xylene (1) 68.80 29 ji-Dodecyclohexane (1) 147.5 26
m-Xylene 86.60 29 1, 3, 6-Triphenyl benzene (s) 87.9 26
jn-Xylene (1) 60.42 29 2-Methyl 2-butene 82.8 10
p-Xylene 84.27 29 3-Methyl 1-butene 82.0 10
p-Xylene (1) 59.20 29 2-Methyl 1-butene 81.0 10
Styrene 82.07 11
Alcohols CHO Compounds
Methyl alcohol (g) 57.72 13 Formaldehyde (g) 52.42 32
Ethyl aclohol (g) 66.39 2 Acetone (g) 72.7 31
Ethyl alcohol (1) 38.4 31 Acetone 0) 47.9 31
Isopropyl alcohol (1) 43.0 31 Dimethyl ether (g) 63.72 19
Isopropyl alcohol (g) 73.4 31 Dimethyl ether (1) 44.98 19
708 CHEMICAL EQUILIBRIUM [CHAP. X V I
TABLE XXXV (Continued)

Halogen Compounds Nitrogen Compounds
S° Re}. S° Bef.
Chlorofomi (g) 70.82 32 Methylamine (g) 67.73 5
Methylene chloride (g) 64.68 32 Methylamine (1) 35.90 5
MethyL chloride (g) 65.94 23 Dimethylamine (g) 65.24 3
Methyl chloride (1) 36.74 23 Dimethylamine (1) 43.58 3
Fluorotrichloromethane (g) 74.07 25 Methyl cyanide (g) 68.02 9
Fluorotrichloromethane (1) 63.92 25 Methyl isocyanide (g) 68.70 9
Ethylene dichloride (1) 49.84 27 Bromoform (g) 79.14 32
Ethylene dichloride (g) 72.74 27 Methylene bromide (g) 70.16 7
Phosgene (g) 67.24 32 Methyl bromide (g) 58.74 8
A.cids Ethylene dibromide (1) 53.37 27
Ethylene dibromide (g) 79.37 27
Formic (g) (monomer) 60.0 12 Methyl fluoride (g) 53.30 7
Formic (g) (dimer) 83.1 12 Methyl iodide (g) 60.85 7
Acetic acid (g) 70.1 12
Lactic acid (s) 34.30 15 Sulfur Compounds
Hippurio acid (s) 67.2 14 Methyl mercaptan (g) 61.02 30
Dimethyl sulfide (g) 68.28 24
Carbohydrates Dimethyl sulfide (1) 46.94 24
1-Sorbose (s) 52.8 17
/3-Maltose monohydrate (s) 99.8 1
a-Lactose monohydrate (s) 99.1 1
/3-Lactose (s) 92.3 1
a-d-Galactose • (g) 49.1 17
EEFERENCES
1. ANDERSON, A. G., and G. STEGEMAN, / . Am. Chem. Soc, 63, 2120 (1941).
2. ASTON, J. G., Ind. Eng. Chem., 34, 614-21 (1942).
3. ASTON, J. G . , M . L . E I D I N O F F , and W. S. FOBSTER, / , Am. Chem. Soc, 61,1539-43
(1939).
4. ASTON, J. G., H . L. F I N K , and S. C. SCHUMANN, / . Am. Chem. Soc, 65, 341-6
(1943).
5. ASTON, J. G., C. W; SILLER, a n d G. H . MBSSEBLY, J. Am. Chem. Soc., B9,
1743-51 (1937).
6. ASTON, J. G., G. J. SZASZ, and H . L. F I N K , J. Am. Chem. Soc, 65, 1135-9 (1943).
7. EDGELL, W . F . , and G. GLOCKLER, J. Chem. Phys., 9, 484-5 (1941).
8. EGAN, C . J., and J. D . K E M P , / . Am. Chem. Soc, 60, 2097-2101 (1938).
9. EWELL, R . H . , and J. F . BOURLAND, J. Chem. Phys., 8, 635-6 (1940).
10. EWELL, R . H . , and P . E . HARDY, J. Am. Chem. Soc, 63, 3460-5 (1941).
11. GuTTMAN, L., E . F . WESTHUM, J R . , and K. S. PITZER, / . Am. Chem. Soc, 65,
1246-7 (1943).
12. HALFORD, J. O., J. Chem. Phys., 10, 582-4^(1942).
13. HALFORD, J . O., and B. PECHERER, J. Chem. Phys., 6, 671-5 (1938).
14., HUFFMAN, H . M . , / . Am. Chem. Soc, 63, 688-9 (1941).
C H A P . XVI] PRESENTATION OF FREE-ENERGY DATA 709
15. H U F F M A N , H . M . , E . L . E L L I S , and H . BOBSOOK, J. Am. Chem. Soc, 62, 297-9

(1940).
16. HuGGiNS, M . L., / . Chem. Phys., 8, 181 (1940).
17. JACK, G . W . , and G. STEGEMAN, / . Am. Chem. Soc, 63, 2120 (1941).
18. KASSEL, L . S., J. Chem. Phys., 4, 435-41 (1936).
19. K E N N E D Y , R . M . , M . SAGENKAHN, and J. G. ASTON, / . Am. Chem. Soc., 63, 2268
(1941).
20. KisTiAKOwsKY, G. B., and W. W. R I C E , / . Chem. Phys., 8, 610-18 (1940).
21.'LiNNETr, J. W., J. Chem. Phys., 6, 700 (1938).
22._LINNETT, J. W., and W. H . AVEEY, / . Chem. Phys., 6, 690 (1938).
23. MESSEELY, G . H . , and J. G. ASTON, J. Am. Chem. Soc., 62, 886-90 (1940).
2 4 . ~ O S B O B N E , D . W . , R . N . DOESCHEE, a n d D . M . YOST, / . Am. Chem. Soc, 64, 165
(1942).
25. OsBOENB, D . W., C. S. GAENEE, R . N . DOESCHEE, and D . M . Yost, J. Am.
Chem. Soc, 63, 3496-9 (1941).
26. PAEKS, G . S., Chem. Rev., 27, 76-83 (1940).
27. PiTZEE, K . S., J. Am. Chem. Soc, 62, 331-5 (1940).
28. PiTZEE, K . S., Chem. Rev., 27, 39-57 (1940).
29. PiTZEE, K. S., and D . W. SCOTT, / . Am. Chem. Soc, 65, 803-29 (1943).
30. R U S S E L L , H . , J E . , D . W . OSBOENE, and D . M . YOST, J. Am. Chem. Soc, 64.
165-9 (1942).
31. SCHUMANN, S . C . , and J . G. ASTON, / . Chem. Phys., 6, 485-8 (1938).
32. STEVENSON, D . P., and J. Y. BEACH, / . Chem. Phys., 6, 25-9, 108 (1938).
33. Y O S T , D . M . , D . W . OSBORNE, and C. S. G A E N E E , / . Am.'fihem. Soc, 63, 3492-6
(1941).
[ ( H ^ — H?) + (AH/)O] represents its enthalpy in the standard state

at temperature T relative to the elements in their standard states a t
0°K. Similarly, [(G^ — HQ) + (AH/)O] is the standard free energy of
the compound at temperature T relative to the elements at 0°K. It is
suggested by Aston^ that these quantities be termed "single' thermody-
namic functions" and tabulated as such. Thus,
Hjy = ( H j - H ° o ) + (AH?)o (19)
G j r = ( G r - H ° o ) + (AH?)o (20)
where
H/r and G/T = the relative enthalpy and relative free energy, respec-
tively, of the compound at temperature T referred
t)o the elements at 0°K
> J. G. Aston, Chem. Rev., 27, 63 (1940).
TABLE XXXVI
/GJ.-H5\
FOB THE IDEAL-GASEOUS STATE AT 1 ATM
\ T ) (AH;)O
cal/(g-inoIe)C'K) kcalper
r, °K 298.1. ..400. . . . 600,. . . 600. . . 8 0 0 . . . 1 0 0 0 . . . 1500 g-mole
Methane 36.42 38.82 40.72 42.36 45.18 47.62 52.81 -15.96
Ethane 45.25 48.20 50.72 53.06 57.28 61.12 69.49 -16.48
Ethylene 44.05 46.70 48.8 60.8 54.4 57.5 64.2 14.51
Propane 52.83 66.62 59.98 63.13 69.00 74.44 86.3 -19.44
Propylene 54.3 67.6 60.6 63.3 68.3 72.9 82.8 8.68
n-Butane 58.54 63.56 68.02 72.16 79.93 87.12 102.65 - 23.25
Butene-1 62.0 66.3 70.2 73.9 80.6 86.8 100.3 6.49
as-2-butene 60.0 64.3 68.2 71.7 78.4 84.4 97.5 3.65
</-ans-2-butene 59.4 63.7 67.6 71.1 77.8 83.8 96.9 2.70
isobutane 56.14 60.85 65.13 69.21 76.90 83.99 99.54 - 24.52
isobutene 57.0 61.6 65.6 69.4 76.3 82.5 95.9 1.64
n-Pentane 64.19 70.32 75.80 80.89 90.51 99.37 118.49 -27.03
2-Methyl butane 64.70 70.40 76.64 80.63 90.12 98.87 118.01 -28.45
Tetramethylmethane1 56.28 61.87 67.13 72.18 81.71 90.55 109.96 -31.07
n-Hexane 69.86 77.12 83.63 89.69 101.16 111.68 134.40 - 31.06
2-Methyl pentane 69.7 88.7 110.4 -32.2
3-Methyl pentane •69.7 88.6 110.4 -32.0
2,2-Dimethyl butane 66.0 &4.6 106.4 -34.6
2,3-Dimethyl butane 66.3 85.4 . 107.3 -34.0
Acetylene 40.01 42.49 44.56 46.38 49.60 62.14 67.43 64.34
Graphite 0.545 0.854 1.180 1.610 2.164 2.798 4.206 0
Hydrogen H j (g) 24.436 26.438 27.965 29.218 31.204 32.752 35.605 0
HjO (g) 37.191 39.529 41.316 42.789 46.153 47.039 50.647 - 57.108
CO 40.364 42.408 43.963 46.238 47.271 48.876 51.880 -27.18
CO2 43.578 45.848 47.681 49.261 51.921 54.137 58.513 -93.949
0, 42.081 44.127 46.691 46.984 49.062 50.715 53.826 0
N2 38.834 40.877 42.431 43.705 45.729 47.322 50.301 0
NH, 43.338 45.205 46.782 49.402 61.580 65.968
NO 42.985 45.140 46.769 48.100 50.214 51.878 54.979 21.579
TABLE XXXVII
B'J. —;
B.I FOR THE IDEAL-GASEOUS STATE AT" 1 ATM
kcal per g-mole
r, "K 298.1 . . 4 0 0 . . . . 500 . . . . 6 0 0 . . .
. 8 0 0 . . . .1000.. .1500
Methane 2.397 3.31 4.35 8.3 5.55 11.4 21.4
Ethane 2.865 4.27 6.02 8.03
12.78 18.37 34.56
Ethylene 2.59 4.75* 7.24* 10.15* 16.87* 24.64*
Acetylene 2.41
Propane 3.535 5.59 8.08 11.06 17.91 25.92 48.77
Propylene 3.20 6.36* 10.14* 14.49* 24.75* 36.62*
n-Butane 4.66 7.43 10.77 14.63 23.68 34.13 63.74
isobutane ' 4.29 7.08 ' 10.46 14.39 23.50 34.00 64.00
1-Butene 3.99
cts-2-butene 4.06
trans-2-buiene 4.06
isobutene 4.25
n-Pentane 5.68 9.12 13.26 18.04 29.21 42.08 78.42
2-Methyl butane 5.17 8.63 12.84 17.70 28.98 41.91 78.69
Tetramethyl methane 5.05 8.56 12.84 17.81 29.37 42.54 79.50
n-Hexane 6.71 10.83 15.77 21.46 34.76 50.04 93.10
2-Methyl pentane 6.07 21.1 . 49.9
3-Methyl pentane 6.04 21.1 50.1
2, 2-Dimethyl butane 5.87 20.9 50.2
2, 3-Dimethyl butane 6.020 21.3 50.3
Graphite 0.251 0.51 0.83 1.20 2.07 3.07 6.0
Hydrogen, Hj (g) 2.023 2.731 3.430 4.128 5.537 6.966 10.696
Water, HÔ (g) 2.365 3.190 4.019 4.874 6.669 8.583 13.89
CO 2.073 2.784 3.490 4.209 5.701 7.258 11.363
CO2 2.240 3.197 4.227 6.328 7.697 10.233 17.02
O2 2.069 2.789 , 3.524 4.280 5.855 7.499 11.77
N2 2.074 2.772 3.485 4.188 5.672 7.230 11.220
NO 2.193 2.900 3.670 4.344 5.872 7.460 11.610
*CalcuIated from data of Table XXI, page 336.
Because of its smaller variation with temperature it is convenient to
tabulate values of G}T/T. In these terms Equations (17) and (18) re-
duce to
Am = S(H;7-)P - 2(H/r)ij (21)
-i.<|.X-<fX
T \ i /P
(22)
It is evident that values of the single functions relative to the elements

at 0°K are the most convenient form for the presentation and use of
thermodynamic data. The disadvantage of this method at present is
that the term (AH/)O is frequently based on relatively uncertain calori-'
metric determinations of heats of reaction or combustion whereas the
groups (Hr — HS) and (G^ — HQ) may be derived with considerable
accuracy from spectroscopic and low-temperature heat-capacity data.
For this reason it is considered desirable to maintain separate tabulations

of the component groups in order that in^roved calorimetric data may
be more readily utilized.
In Table XXXVI are values of (GJ- — ^D/T and (AH/)O for a number
of compounds'' * and elements in the ideal-gaseous state at 1 atm for
seven temperatures. In Table XXXVII is a corresponding summary of
values of (H^ — HJ). By interpolation between the temperatures of
the tables standard enthalpy and free-energy changes are obtained by
Equations (17) to (22) with no uncertainties regarding the accuracy of
empirical heat-capacity equations. Values of {AG°JT)T calculated from
+4
/ f
+3
/
+2 '^''^ Oi#^"
t< +1 4-CO
Hi+S2^252^
3 0 p/
S5
-1
^:*^
^/
-^^Q*^
-2 l-Oi'*
• ^
-3
200 400 600 800 1000 1200 1400 1600 1800 2000
Temperature, "Kelvin
FIG. 156. Equilibrium Constants of Chemical Reactioi^.
Equations (18) or (22) are directly related to the equilibrium constant.

Thus rearranging Equation (6) gives
4.576 log K
=-(f).
where {LG°/T)T is expressed in cal/(g-mole)(°K) or Btu/(lb-mole) (°R).
(23)
In Fig. 156 are values of log K for several reactions plotted as func-
. tions of temperature. These values were calculated by Equation (23)
using the data of Tables XIV, XV, XXXV, and XXXVII.
' K. S.-Pitzer, Chem. Rev., 27, 39 (1940).
* F. D. Rossini, E. J. R. Prosen, and K. S. Pitzer, / . Research Natl. Bur. Standards,
27, 529-41 (1941).
Illustration 4. From the data of Tables XXXVI and XXXVII calculate the
values of AH/ and hG}/T for the formation of isobutene from the elements at
298'. 1°K
4 C ( s ) + 4 H 2 ( g ) - - > C4H8(g)
From Tables XXXVI and XXXVII

iC^HsCg) H2(g) C(s)
-57.0 -24.436 -0.545
\ J- / 298.1
(AH;)O 1.64 .0 0
(HJ> — H 0)298.1 4.25 2.023 0.251
From Equation (14), >
(AH;)298.I = [4.25 + 1.64] - [4(0.251) + 4(2.023)] = -3.206 kcfil per g-mole
From Equation (16),
= 48.43 cal/(g-mole) (°K)
Illustration 5. From the data of Tables XXXVI and XXXVII calculate the values
of AH° > and the equiUbrium constant K for the dehydrogenation of propane to
propylene at 800°K.
From Table XXXVI at 8 0 0 %
3H8(g) H2(g) C3H6(g) •
(^") 69.00 -31.204 -68.3

(AH;)O 19.440 0 8.580
From Table XXXVII,
(HV-HS) 17.910 5.537 24.750
For the reaction,
C3H8(g) = C3He(g) + H2(g)
From Equation (17),
AH'm = [24.750 + 8.580 + 5.537] - [17.910 - 19.44] = 40.397 kcal per g-mole
From Equation (18),
f ^ ^ = [-68.3 -f M!2 - 3I.204I - [-69.00- IM!?]

\T jsoa I 800 J L 800 J
= _68.3 - 31.204 + 69.00 + ^'580 + 19,440 ^ ^ ^î cal/(g-mole) (°K)
From Equation (23), . <
log K = -4.521/4.576 = -0.9880 = -0.988
K = 0.103
EQUILIBRIUM COMPOSITIONS
B. F. Dodge* has aptly discussed the fe&ibility of a chemical reaction
thus:
"The statement is sometimes made that a given reaction is thermo-
dynamically impossible. This is a loose statement which has no
meaning in the absence oj qualifying statements. For example, any
reaction, starting with pure reactants uncontaminated by any of the
products, will have a tendency to proceed to some extent even though
this may be only infinitesirfial. Thus, the reaction
H20(g) = H2(g) + K)2(g)
proceeds to some extent at 25''C, and we can even calculate with con-
siderable assurance the percentage of water vapor that would be de-
composed. From the accurately known value of AG° of_, this reaction
at 25°C the equilibrium constant is about 1 X 10"*°, and the extent of
decomposition is infinitesimally small but definite,"
"From the value of the standard free-energy change for any reaction
we can form an opinion about the feasibility of the reaction without
further calculation. Thus, if AG° = 0 at a given temperature, then
K = I, and it is obvious that the reaction must proceed to a consider-
able extent before equilibrium is reached. The situation becomes less
favorable as AG° increases in the positive direction but there is no defi-
nite value that one can choose as clearly indicating that the reaction is
not feasible from th'e standpoint of industrial operation. At 600°K the
AG° for the methanol synthesis reaction is +11,000 calories per gram-
mole, and yet the reaction is certainly feasible at this temperature. In
this case the imfavorable free energy change for the standard state is
partially overcome by utilizing high pressure to displace the equilibrium.
Other means can also be used such as changing the ratio of reactants or
removing one of the reaction products."
"For the purpose of Ascertaining quickly and only approximately if any
given reaction is proinising at a given temperature, the following rough
classification may be useful:
AG° < 0 Reaction is promising
AG° > 0 but < +10,000 Reaction is of doubtful promise but warrants
further study
AG° > +10,000 Very unfavorable, would be feasible only under
unusual circumstances
It should be understood that these are only approximate criteria that
are useful in preliminary exploratory Work."
" B. F. Dodge, Trans. Am. Inst. Chem. Engrs., 34, 640 (1938). '
CHAP. XVI] GASEOUS SYSTEMS 715
From Equation (7) compositions of systems at equilibrium can be

calQulated from a knowledge of the standard free-energy change and the
relations of activities to stoichiometric compositions. These relation-
ships involve the activity coefficients discussed in Chapter XIV. If
these coefficients are included, Equation (7) becomes, for the general case,
(yRNRyiysNs)' • • •
K=
(yaNBYiycNcy-
• N'^Ns---_-^yiyc--- (24)
Gaseous Systems. 'As previously mentioned, in a system which in-

volves only gaseous components which behave as approximately ideal
solutions, it is convenient to choose as the standard state the component
gases each at unit fugacity expressed in atmospheres. This state is
hypothetical at temperatures where saturation pressures are below unit
fugacity. With this definition, the activity of component B in ideal
solutions of gases is
aB=fB = Nsfs,, = NBVBITT!- (25)
where
/BIT, VBn = the fugacity and fugacity coefficient, respectively, of pure
component B at the temperature and total pressure x
of the system
^ = (N}m^\(±tliy+'^-''-^-c--- (26)
Viviiva • • • y w ^ • •/
In a system containing ni moles of insert gases not entering into the
reaction,
NB= , , , ""^ , 7— (27)
nB + nc+- • •+nB + ns + \- - + 711
where UR = moles of R in the equilibrium mixture.
The ratio of the fugacity coefficients in Equation (26) is constant for
a given temperature and pressure and is designated by Kv.
^^ = ^ 1 4 ^ (28)
VBVC • • •
However, unlike the equilibrium constant K, the term K^ is affected by

changes in pressure as well as in temperature.
For reactions at low pressures, in the neighborhood of atmospheric or
below, the fugacity coefficients may be taken as unity and the term Ky
neglected. At higher pressures this term may have a marked effect on
/
t h e equilibrium calculations a n d m a y be. determined from Fig. 142,

page 622. •^
Combining (26), (27), a n d (28), gives
-<! + •••)
(29)
[nine ••J "L^B+ncH -\-nB-\-na+ • • • +ni]
E q u a t i o n (29) is t h e mostv useful form of the equilibrium equation from
which, if t h e equilibrium constant K is known, t h e composition of a re-
acting system a t equilibrium m a y be directly calculated. This calcula-
tion is best carried o u t b y expressing t h e number of moles of each active
material present a t equilibrium in terms of t h e equilibrium conversion x
of a given reactant and t h e numbers of moles of components in t h e origi-
nal unreacted mixture. These values are substituted in E q u a t i o n (29),
which is t h e n solved for the equilibrium conversion. If t h e final equation
t o b e solved is of a complicated form, graphical methods m a y offer t h e
most convenient solution.
Illustration 6. The gases from the pyrites burner of a contact sulfuric acid plant
have the following composition by volume:
SO2 7.8%
O2 10.8%
N2 81.4%
100.0%
This gaseous mixture is passed into a converter where in the presence of a catalyst
the SO2 is oxidized to SO3. The temperature in the converter is maintained at
500°C and the pressure at 760 mm of Hg. Calculate the composition of the gases
leaving the converter, assuming that equilibrium conditions are reached. It may
be assumed that the fugacity coefficients and hence Kv are equal to unity.
Solution:
Basis: 100 Ib-moles of the original mixture.
SO.(g)+iO.(g)=S03(g)
From Fig. 156, at 500°C (773''K), log K = 1.93; K = 85
TT = 1.0 atm
Limiting reactant = SO2
Let X = equilibrium conversion of SO2. Composition of equilibrium mixture:
502 = 7.8 - 7.8r lb-moles = ngoj
503 = 7.8x = ngo^
O2 = 10.8 - 7.8 f - ) Ib-molea = no.
N2 = 81.4 lb-moles = n
Na
From Equation (29):
\ =K (a)
(«S02)('^02) ' "-"SO,
L«s02 ++ »S03
MSO3 ++ »0»
»02++ "NsJ
"NsJ
CHAP. XVI] GASEOUS SYSTEMS 717
Substituting in this equation gives
(7.8-7.8x)(10.8-3.9x)*Ll00-3.9xJ " *^
Since ir = 1.0 and K = 85, for graphical solution:
X /lOO - 3.9xy
1.0-x\10.8-3.9x/ ~ ^ 5 - ^
The solution is obtained by assuming a series of values of x and calculating the cor-
responding values of A. These results may be plotted, the value of x corresponding
to A = 0 being the correct solution. Since the quantity in parentheses ^varies but
little when x is close to 1.0, the correct solution also may be readily obtained by trial
without a plot. The use of this method gives
X = 0.9585 or 96.85 per cent

Composition of gases leaving the converter:
SO2 ='7.8-(7.8) (0.9585) = 0.32 lb-mole 0.3%
S03= (7.8) (0.9585) = 7.49 lb-moles 7.8%
O2 = 10.8 - (3.9) (0.9585) = 7.07 lb-moles 7.3%
Nj= 81.40 lb-moles 84.6%
Total = 96.28 lb-moles 100.0%
Thissame procedure may be followed to calculate the degree of com-

pletion in any system at conditions for which the necessary data are
available. By repeating the calculations of Illustration 6 to correspond
to other conditions of temperature and pressure, the effects of varying
these conditions may be quantitatively predicted.
Under conditions oi high pressure the fugacity coefficients cannot be
neglected, but the general method of solution is the same. Values of Kv
may be calculated directly from fugacity coefBcients of the individual
components obtained from Fig. 142. Where a number of calculations
are to be made for a particular system of reaetants and products a chart
may be prepared relating Kv to temperature and total pressure. Fig-
ure 156a is such a chart for the ammonia-synthesis system where
•^ ''NH,
Kv =
i a
2 y2
N2 Ha
Illustration 7. Calculate the equilibrium percentage conversion of nitrogen to
ammonia at 700°K and a pressure of 300 atm if the gas enters the converter with a
composition of 75 mole per cent H2 and 25 mole per cent N2.
iNj 4- f Ha = NH3
At 300 atm, 700°K, K = 0.0091 (Fig. 156), and K. = 0.72 (Fig. 156a).
Let X represent the number of moles of ammonia formed at equilibrium starting

with J mole Nj and I moles Hj. Then
"NHa 0.77a;
4 A (4:-4x)*(f - fx)* (1-^)'
and 2 — x = total number of moles at equilibrium
1.0
_ TNES
tLy"
0.8 'MJ Ha
0.6
^
0.4
500'
0.2 ^ * > x ^
450°
^
400 "K
100 200 300 400 600 600 700 800 900 1000
Pressures in Atmospheres
FIG. 156O. Ratio of Fugacity Coefficients for Ammonia Synthesis.
Substitution of the preceding values in Equation (9) gives
L(l-i)d \{2-x))
or
X = 0.589
The percentage conversion of Nz to NHs is hence 58.9 per cent, and the composition
of the equihbrium mixture in mole per cent is
Ni = 14.6 per cent
Ha = 43.8 per cent
NH3 = 41,6 per cent
Efifect of Reaction Conditions on Equilibrium Conversion. Equa-
tion (29) may be rearranged:
(30)
{nBy{ncY-~ K\_ TT J
From inspection of this equation the effects which are produced by
changes in the conditions of the reaction on the equihbrium conversion
can be predicted. Any change which will increase the right-hand side
of Equation (30) will tend to increase the ratio of products to reactants
in the equilibrium mixture and correspond to an increased conversion.
Effect of Temperature. From Equation (30) it is apparent that an
CHAP. XVI] EQUILIBRIUM CONVERSION 719
increased value of the equilibrium constant K must correspond to an

increased conversion. The value of the equilibrium constant for the
commonly chosen standard state depends only upon temperature, as
already discussed. It follows that the equilibrium conversion is increased
by a rise in temperature in the case of an eTidothermic reaction and decreased
in the case of an exothermic reaction.
Effect of Pressure. It has been pointed out that the equilibrium con-
stant K is independent of pressure with the standard states for which
Equation (30) was derived. The value of Ky is affected by pressure in
a manner for which generalized predictions cannot be made. When the
compressibility of the products is greater than that of the reactants, an
increase in pressure \vill decrease K^ and hence increase the conversion.
In addition, from inspection of Equation (30) it is apparent that the
pressure under which a reaction proceeds will affect its equilibrium con-
version in case the reaction produces a change in the total number of
moles of gaseous components present in the system. If there is no
change in the number of moles of gases, the exponent (r -t- s -f- • • •)
— (6 + c + • • •) will equal zero, and the magnitude of the pressure TT
will have no effect on the extent of the reaction except as it affects K^.
However, if a reaction produces a decrease in the total number of moles
of gaseous components the equilibrium degree of completion is increased by
an increase in pressure. If the total number of moles of gases is increased
as a result of the reaction, an increase in pressure reduces the equiUbrium
degree of completion. This is in accord with the classical Le ChateUer-
Braun principle.
Effect of Dilution with an Inert Gas. Dilution of a reacting system
with an inert gas corresponds to an increase of ni of Equation (30).
The effect produced is similar to that of a decrease in pressure. Hence,
if a reaction produces an increase in the number of moles of gaseous com-
ponents the equilibrium degree of completion is increased by dilution with
an inert gas. If no change in the total number of moles of gases accom-
panies a reaction, the presence of inert gases has no effect on the equilib-
rium conversion.
Effect of Excess Reactants. If component B of Equation (30) is the
limiting reactant, an increase in the number of moles of the other reac-
tants C • • • increases the number of moles of products, R, S, • • •, and also
the degree of conversion of reactant B at equilibrium. Therefore, the
presence of excess of one reactant tends to increase the equilibrium conversion
of the other reactant.
Effect of Presence of Products in Initial Reacting System. I t is apparent
from Equation (30) that the presence in the original unreacted system
of any of the compounds which are products of a reaction reduces the
amounts of these compounds which are formed by the reaction in pro-

ceeding to equilibrium conditions. Therefore, the addition of reaction
products to the initial reaction system reduces the equilibrium conversion
of any reactant.
Equilibrium-Conversion Charts. Where numerous calculations are to
be made of equilibrium composition corresponding to various conditions
of temperature and pressure for a reacting system of constant initial
composition it is frequently convenient to prepare charts expressing the
equilibrium degree of completion as a function of temperature and pres-
sure. Such a chart obviates the necessity of frequent repetitions of
tedious graphical solutions of the type of Illustration 6.
Even in a complicated system the data for an equilibrium-conversion
chart may be readily calculated by arbitrarily selecting a series of values
of conversion. From Equation (29) the equilibrium constant is cal-
culated corresponding to each of these selected degrees of completion
and to a selected constant pressure. The temperature corresponding to
each equilibrium constant may then be obtained from Fig. 156. These
results are plotted, temperatures being used as abscissas and fractional
conversion as ordinates. By repeating these calculations to correspond
to other selected pressures a complete chart may be constructed. These
calculations are best tabulated. In the following illustration tîs method
is applied to the system discussed in Illustration 6.
Illustration 8. For the reacting system discussed in Illustration 6, calculate and
plot curves relating the equilibrium conversion to temperature at pressures of 1.0
and 2.0 atm, respectively, assuming K, = 1.0.
1 TABLE A
1 2 3 4 5 6 7 8
X 1.0-X 3.9x 10.8 - 3.9a; 100 - 3.9x [(5)/(4)]^ (l)/(2) KM^

0.50 0.5 1.95 8.85 98.0 3.33 1.0 3.33
i 0.70 0.30 2.73 ' 8.07 97.3 3.47 2.33 8.09
0.90 0.10 3.51 7.29 96.5 3.64 9.0 32.8
.0.95 0.05 3.70 7.10 96.3 3.68 19.0 70.1
0.98 0.02 3.82 6.98 96.2 3.71 49.0 181
Basis: 100 lb-moles of the original mixture.

Let X = fractional conversion of SO2 at equilibrium
From Equation (b) of Illustration 6: '
(x) (100 • • 3.9x)^
^TW* = .
(1.0 - x) (10.8 - 3.9x)
A series of values of x are selected and the corresponding values of K calculated, first
with V equal to 1.0 and then to 2.0 atm. The corresponding temperatures !r°K are
CHAP. XVI] EQUILIBRIUM-CONVERSION CHARTS 721
then obtained from Fig. 156. The calculations are summarized in Tables A and B.
In the column headings the numbers in parentheses represent the results contained
in the columns bearing these numbers.
TABLE B
X = 1.0 . x = 2.0
X ,^^ K logK r°K K logX T'K
0.5 3.33 0.523 978 2.35 0.371 1005

0.7 8.09 0.908 910 5.71 0.757 935
0.9 32.8 • 1.516 822 23.2 1.366 841
0.95 70.1 1.846 782 49.5 1.695 800
0.98 181 2.258 737 128 2.107 752
These calculations may be continued to correspond to other pressures. The results

are plotted in curves E of Fig. 157 relating the equilibrium degree conversion as ordi-
100
V^s^
^^
I 90
% 7
gg80
S«
.3^70
#7
1
/ i/ ^ >
& 60
/l f
m hT jF
sA^ \E
50
400 450 500 650 600 650 700 750 800
Temperature, °C
(Composition of Entering Gas; 7.8% SO2, 10.8% 02, 81.4% Nj)
FIG. 157. Equilibrium Conversion and Adiabatic Temperature

Relations for the Oxidation of SO2.
nates to temperature in degrees centigrade. This chart is applicable only to systems
of the particular initial composition here considered. Change in relative proportions
of any of the original reaotants renders the chart inapplicable.
A chart of the type of curves E of Fig. 157 may be prepared for any
reacting system and from it equihbrium conversion may be readily esti-
mated by interpolation. For example, at a temperature of 600°C and
a pressure of 1.5 atm it is estimated from the E curves of Fig. 157 that
the equilibrium conversion is 82 per cent. If the initial composition of
the reacting system undergoes change a new chart must be prepared to
correspond to each different composition.
Heterogeneous Reactions. Equation (29) may be applied to any

gaseous system if the gas mixture behaves as an ideal solution. When
a component of a reaction is involved in a heterogeneous reaction as a
pure liquid or pure solid, its activity may be taken as unity provided
the pressure on the system does not differ much from the chosen stand-
ard state. The effect of pressure upon the activity of a solid or liquid
may be calculated from Equation (XIV-30); this effect is negligible at
moderate pressures.
If in the reaction represented by Equations (1) and (26) there are, in
addition to gaseous components B, C, R and S, d moles of a solid or liquid
reactant D and t moles of a solid or liquid product T, Equation (26)
becomes
Where the standard state for solids and liquids is taken at atmospheric
pressure or at low equilibrium vapor pressures, the activitieb of pure
liquids and pure solids may be taken as unity at all moderate pressures,
and the composition of the gaseous phase at equilibrium will not be af-
fected by the presence of the solid or liquid. However, at high pressures
the activities of pure solids and liquids are affected by pressure, and the
composition of the gaseous phase at equilibrium is affected by the pres-
ence of the liquid or solid. When solid or liquid solutions are formed
the activities of the components in solid or liquid solutions are no longer
unity even at moderate pressures, and the equilibrium composition of the
gas is greatly affected by the presence of the solid or liquid phase. In
such cases the activities in the solid or Uquid solutions are expressed in
terms of mole fractions and activity coefficients as discussed in Chap-
ter XIV.
Illustration 9. Ferrous oxide, FeO, is reduced to metallic iron by passing a mixture
of 20 per cent CO and 80 per cent N2 over it at a temperature of 1000°C under a
pressure of 1 atm. Assuming that equilibrium is reached, calculate the weight of
metalUc iron produced per 1000 cu ft of gas admitted measured at 1 atm and 1000°C. •
Fugacity coefficients may be assumed equal to unity. The reaction taking place is
as follows:
FeO (s) + CO (g) ^ Fe (s) + COj (g)
At 1000°C the value of K for thig reaction is 0.403.

Basis: 1.0 lb-mole of entering gas.
Since the activities of FeO and Fe equal 1.0, Equation (31) may be arranged in
the form of (29). Then
/^CoA K or ——• = it
\ « c o / "CO + "COs + "NJ "CO
CHAP. XVI] DECOMPOSITION AND VAPORIZATION 723
Let X = fractional conversion of CO at equilibrium. At equilibrium:

nco, = 0.20X
Jico = 0-20(1 - x)
" °-^°^ =0.403

0.20(1 - x)
or x = 0.287
CO2 produced = (0.287) (0.20) = 0.0574 lb-mole
Pe produced = 0.0574 lb-atom = 3.2 lb
Volume of entering gas at 1000°C = (359) (Wir) = 1673 cu ft
Fe produced per 1000 cu ft of gas = —-^—— = 1.9 lb
> 1673
Pressures of Decomposition and Vaporization. Many solid com-
pounds decompose to yield another solid and a gas as in the calcination
of calcium carbonate to form lime and carbon dioxide. Such decom-
positions will proceed only when the activity of the gaseous product in
contact with the sohd is less than the equilibrium value determined by
the temperature and the nature of the reaction.
A solid decomposition reaction may be represented by the following
equation:
65 (s) ^ ri?(s)-f sS (g) (32)
The activities of the solids are approximately unity at moderate pres-
sures so that as long as any of compounds B and R are present Equa-
tion (7) may be applied to this reaction as follows:
K = a% = e «^ = e «^ « (33)
At low pressures, where the activity of a gas may be taken as equal to its
partial pressure, a*s = Ps and
I t may be noted that if AH° and LS° are independent of temperature the
form of Equation (34) is similar to that of the Clausius-Clapeyron
vapor-pressure equation. In general both AH° and A»S° vary somewhat
with temperature but to a much smaller extent than AG°.
Illustration 10. Calculate the decomposition pressure of limestone at 1000°K and
the temperature necessary to produce a decomposition pressure of 1.0 atm.
CaCOs (s) ^ CaO (s) -f- CO2 (g)
From Table XXXV,
AS^.I = 9-5 + 51.1 - 22.2 = 38.4
From Table XIV,
Affî = (-151.7) -t- (-94.03) - (-289.3) = 43.57 kcal per g-mole
From Table V, for CO2:

cp = 6.85 + (8.533) (10-8) r - ^.475) (lO-')?^
From the data of Kelley/
108 000
for CaO: Cp = 10.00 + (4.84)(.IQ-^T ^
0(Y7 AAA
for CaCOa: Cj, = 19.68 + (11.89)(10-^)T j ^
Using the symbols of Equation (9), yields
Aa = 6.86 + 10.00 - 19.68 = -2.83
A6 = (8.533 + 4.84 - 11.89)10"' = (1.483) (lO"")
Ac =-(2.475) (10-«)
Ad = (-108,000) - (-307,600) = 199,600
Substitution in Equation (9) gives V
Aff29i = !« + AaT + iAbT' + iAcT" - y
199,600
43,570 = !„+ (-2.83) (291) + i(1.483) (10-^) (291)'+K-2.475) (10-«) (291)' ^
or IH = 45,037 and A^rooo = 41,924
™ Ad
A^jj.i = Is + AalaT + AbT + iAcT^
^ 1
38.4 = Is+ (-2.83) (5.697) + (1.483) (10"') (298.1)
+ «--«<'«-'<-'>-3(S-)
or Is = 55.31 and ASjooo = 35.90
From Equation (34),
, ^ I f -41,924 35.90] , „.„
log Ps = —— = —1.316
2.303 L(1.987) (1000) 1.987J ^
or p, = 0.0483 atm at lOOO^'K
At a decomposition pressure of 1 atm, from Equation (34) AG^ = 0. The cor-
responding temperature T is calculated from Equation (11); .
AGT = 7H + (Aa -I,)T - AaT In T - ÂbT^ - iAdP - —
or, substitution of known values gives
AG°I. = 45,037 - 58.14r + 6.517r log T - 0.7415(10"') T*
+ (0.4125) (lO-*)?' - ? 5 ^ = 0
This equation is solved graphically or by trial and error to give

T = 1180°K, the decomposition temperature of CaCOs at 1.0 atm
Equations (23) and (34) are directly applicable to the equilibrium be-
tween a pure solid or liquid and its vapor. In this case AG°, Aff°, AS°
° K. K. Kelley, " High-Temperature Specific Heat Equations for Inorganic Sub-
stances," V.S. Bur. Mines Bull, 371 (1934).
CHAP. XVI] ADIABATIC REACTION TEMPERATURES . 725
represent t h e changes in free energy, enthalpy, and entropy, respectively

in t h e isothermal transformation of t h e soUd or hquid t o t h e ideal vapor
a t a pressure of 1.0 a t m . I n vaporization a t low pressures Aff° becomes
equal t o t h e h e a t of vaporization a t t h e given temperature.
Illustration 11. Calculate the vapor pressure and latent heat of vaporization of
molten copper at 3,000°K from the following data:
For soUd copper at 298.1°K: Gu(l) ^ Cu(g)
H° — Ho = 1199 cal per g-atm
s° = 7.92
For copper vapor in the ideal state at 298.1°K and 1.0 atm:
H° - Ho = 82,720
8° = 39.75
Atomic heat capacities (above 298°K):
Solid copper, Cp = 5.44 + 0.0014627
Liquid copper, Cp = 7.50
Gaseous copper, Cp = 4.97
Heat of fusion at 1357°K (melting point) = 3,110 cal per g-atom
For liquid copper at 3,000°K:
H° - HS = 1199 + / (5.44 + 0.001462r) dT + 3110 + 7.50(3000 - 1357) = 23,670

.^298
H° - H; for copper vapor at 3000°K = 82,720 + 4.97(3000 - 298)
= 96,120 cal per g-atom
For hquid copper at 3,000°K:
ô ^ 7.92 + p r (5.44+ 0.0014627) , 3 1 1 0 3000 ^
^298 T ^1357 1357
s° for copper vapor at 3,000°K = 39.75 + 4.97 hi Ws"- = 39.75 + 11.45
= 51.20 cal/(g-atom) (°K)
In vaporization at 3,000°K:
AH° = 96,120 - 23,670 = 72,450 cal per g-atom
As° = 51.20 - 26.02 = 25.18 cal/(g-atom) (°K)
AG° = AH° - TAa° = 72,450 - 3000(25.18) = -3090 cal per g-atom
From Equation (23),
logiS: = ?^ = 0.225
(4.57) (3000)
or K = Pv= 1.68
Since at this moderate pressure v is approximately unity the vapor pressure P is
1.68 atm.
ADIABATIC REACTION TEMPERATURES
I n Chapter V I I I methods were demonstrated whereby t h e tempera-
t u r e attained b y a reacting system m a y be calculated if t h e reaction
goes t o completion without loss of h e a t from t h e system. I n a reaction
reaching equilibrium conditions which do not approach 100 per cent

conversion, the temperature attained depends upon the degree of con-
version actually produced, which in turn depends on temperature. The
equilibrium temperature attained by such a reaction when proceeding
adiabatically may best be determined by a graphical calculation, making
use of an equilibrium-conversion chart for the system.
On an equilibrium-conversion chart of the type of curves E of t i g . 157,
•page 721, a curve is plotted relating caclulatedadiabatic reaction tempera-
ture to conversion. This curve may be established by selecting a series
of values of degrees of conversion and calculating the reaction tempera-
ture corresponding to each by the method discussed in Chapter VIII,
1.0 Isothermal equilibrium ,vxtve
Temperature —>• Temperature—^

Exothermic Reaction Endothermic Reaction
FIG. 158. Equilibrium Conversion in Exottiermic and Endotliermic Reactions for
Isothermal and Adiabatic Operations.
page 308. The values on this curve are independent of pressure if it is

assumed that enthalpies are independent of pressure. However, a new
curve must be plotted to correspond to each change in the initial tem-
peratures of any of the reactants.
In reactions proceeding adiabatically the characteristic behaviors of
exothermic and endothermic reactions are shown diagrammatically in
Fig. 158. In exothermic reactions the equilibrium conversion decreases
with increase in temperature whereas the temperature steadily rises as
the reaction proceeds. The adiabatic reaction obviously cannot proceed
beyond the intersection of these two lines. In endothermic reactions
the reverse is true: the equilibrium degree of conversion is favored by a
rise in temperature whereas the temperature falls as the reaction pro-
ceeds. It may be noted that isothermal operation leads to a higher
equilibrium conversion in either an, endothermic or exothermic reaction.
Illustration 12. ^ The mixture of 7.8 per cent SO2, 10.8 per cent O2, and 81.4 per
cent N2 discussed in Illustrations 6 and 8 enters^the converter at a temperature of
400°C. Calculate the equilibrium temperature attained in the converter, assuming
that it is thermally insulated so that the heat loss is negligible.
CHAP. X V I ] A D I A B A T I C REACTION TEMPERATURES 727
Solution:
Basis: 100 g-moles of the original gaseous mixture.
Let X = degree of completion
From the heat-of-formation data, Table XIV, page 257,
AH = (-93,900) - (-70,920) = -22,980 cal per g-mole
SO2 converted = 7.8i g-moles
Total standard heat of reaction = (7.8x) (22,980) = 179,240a; cal
Enthalpy of Reactants: Mean heat capacity between 18° and 400°C is obtained
from Table VI, page 216.
SO2 = (10.91) (7.8) (400 - 18) = 32,510
02>= (7.40) (10.8) (400 - 18) = 30,530
N2 = (7.09) (81.4) (400 - 18) = 220,460
Enthalpy of reactants = 283,500
Enthalpy of Products Relative to 18°C:
AH= f ncj, dT; T is in °K
•^291
502 = (7.8 - 7.8a;) f [8.12 + (6.825) (10-^) 7 - (2.103) (10-6) r^] ^T
503 = 7.8a; f^ [8.20 + (10.236) (10-3) r - (3.156) (10-6) r^] dT

-'291
O2 = (10.8 - 3.9x) r^[6.13 + (2.99)(10-')r - (0.806)(10-«)r21 dT
•/291
N2 = 81.4 r^[6.30 + (1.819) (10-3) T" - (0.345) (10-«)r2] dT

"'291
By integration and addition the total enthalpy of products
= [643.4(r - 291) + (116.7) (10-3) (r^ - 29P) - (17.74) (10-«)(J" - 291')]
- x[23.3(r - 291) - (7.5)(10-3)(T^ - 29V) + (1.69)(10-«)(r' - 29P)]
From an energy balance, solving for x, there results
-480,170 + 643.47 + (116.7)(10-3)r» - (17.74)(10-°)f'
^ ~ 173,060 + 23.3r - 7.5{m-')T' + (1.69)(10-«)r'
This equation permits direct calculation of values of x corresponding to Selected
values of T.
fC X
519 0.5
570 0.7
627 0.95
These values are plotted in the F curve of Fig. 157, page 721, corresponding to an
initial temperature ta, of 400°C. This curve crosses the E curve corresponding to a
pressure of 1.0 atm at the point corresponding to a temperature of 598°C and a degree
of completion of 82.5 per cent. These are the equilibrium conditions of the reaction.
The F curves of Fig. 157 relate calculated reaction temperatures to

conversion, each corresponding to a different initial temperature of the
reactants entering the converter. These curves were established by the
f
method demonstrated in Illustration 13. Such a chart may be con-
structed in a similar manner to'apply to any system for which the ,
necessary data are available. From Fig. 157 the equilibrium tempera-
ture and degree of completion corresponding to any selected adiabatic ,
operating conditions may be readily estimated. For example, if the re-
acting gases enter the converter at a temperature of 375°C and a pressure
of 1.5 atm, an equilibrium ^temperature of 585°C and a degree of com-
pletion of 87 per cent will be attained.
If a reaction cannot be assumed to proceed under adiabatic conditions,
a chart of the type of Fig. 157 can be constructed which takes into ac-
count the loss of heat from the reacting system. Such losses in no way
affect the E curves. However, the heat losses, expressed as a function
of the temperature of the system, must be included in the energy-balance
equation from which the data for the F curves are calculated.
Equilibriiim-Reaction Temperatures at High Pressures. When reac-
tions proceed at high pressures, the method described for obtaining the
equilibrium-reaction temperature should be modified to take into con-
sideration the effect of pressure upon enthalpies, heat capacities, and
activities of the component gases. For example, if a 1:3 mixture of
nitrogen and hydrogen enters a catalytic reaction chamber at 400°C and
reacts adiabatically to' equilibrium at 300 atm, the temperature leaving
will be 590°C with a corresponding percentage conversion of nitrogen to
ammonia of 25 per cent. If the results were calculated with the effect
of pressure upon heat capacities and activities neglected, the calculated
results would be 600°C with a corresponding conversion of 21 per cent.
The smaller temperature rise under actual conditions is due to the in-
creased heat capacity of ammonia at 300 atm, and the increased per-
centage conversion is due to the decrease in activities of ammonia as^.
compared to nitrogen and hydrogen.
Equilibrium-reaction temperatures are readily calculated by the gen-
eral methods already demonstrated if ideal solutions are formed. This /
condition is generally satisfactorily approximated where all reactants and
products are well above their critical temperatures. For other cases in-
volving reactions at conditions in the vicinity of the critical point of the
mixture, the precise methods for estimating activities described in Chap-
ter XIV must be used.
Illustration 13. It is desired to calculate the equilibrium percentage conversion \
of SO2 to SO3 and the corresponding temperature when the system described in Il-
lustration 12 reacts adiabatically at 100 atm and the entering temperature is 400°C.
The equilibrium-temperature line E, corresponding to 100 atm, is constructed as
in Illustration 8 except that the deviation of K„, from unity is calculated from Fig. 142
as in Illustration 7. The adiabatic heating line "F is constructed by the method em-
ployed in Illustration 12, except that allowance is made for the effect of the pressure
CHAP. XVI] EQUILIBRIUM OF SIMULTANEOUS REACTIONS 729
of 100 atm upon the enthalpies of the entering and reacting mixtures. These correc-
tions are obtained from Fig. 106. The procedure is as foUows:
The critical constants of the component gases are:
TlK Pcsim
S02 430.2 77.7
Ns 125.9 33.5
O2 154.2 49.7
SO3 491.3 83.6
For the entering gas, which contains 7.8 per cent SO2, 81.4 per cent N2, 10.8 per
• cent O2:
Pseudocritical temperature = 0.078(430.2) + 0.814(125.9) + 0.108(154.2) = 152°K
Pseudocritical pressure = 0.078(77.7) + 0.814(33.5) + 0.108(49.7) = 38.6 atm
Pseudoreduced temperature, T^ = = 4.40
Pseudoreduced pressure, Pr = = 2.59

38.6
From Fig. 106, ° ~ " =0
To
The effect of the elevated pressure on the enthalpy at 400°C is negligible. When
the gas has been heated adiabatically by reaction to equilibrium, the approximate
conversion is 96% at 631 °C.
The corresponding composition is
502 = 7.8(1 - 0.96) = 0.31 moles 0.3%
N2 . = 81.40 moles 84.6%
O2 = 10.8 - .^^-^ - Q-^l) ^ 7 05 moles 7.3% '
503 = 7.8 - 0.31 = 7.49 mole? 7.8%

•96.25 moles 100.0%
The correction of enthalpy for pressure may be shown by Fig. 106 to be negligible
and the adiabatic line at 100 atm coincides with the adiabatic line for 1 atm.
EQUILIBRIUM OF SIMULTANEOUS REACTIONS

Where several reactions occur among a given group of reactants, the
composition of the resultant products at equihbrium depends upon the
• simultaneous equilibria of all the separate reactions. If large numbers
of products are formed as in the pyrolysis of heavy hydrocarbons, or of
cellulose, the resultant composition cannot ordinarily be calculated.
However, even in such cases where many reactions are possible, restric-
tion to a single course may be accomplished by the choice of a specific
catalyst which promotes one selected reaction. This selection is de-
pendent on kinetic rather than on thermodynamic considerations, and
' i t is common for equilibrium to be approached with respect to one or
two reactions while many other reactions which are thermodynamically
possible do not occur to any appreciable extent. Under such circum-
stances thermodynamic methods are only partly successful and must be

supplemented by knowledge of the actual course of the reaction and na-
ture of the products formed before equilibrium calculations can be
carried out.
For example, in the reaction of carbon monoxide with hydrogen the
number of products theoretically possible is almost unlimited and the
free energy in the formation of many of these products may be extremely
favorable; yet it is possible by the selection of a special catalyst to ex-
clude consideration of nearly all possible products save methanol. An
apparent equilibrium yield can be. established by considering only the
reaction CO + 2H2 "^r^. CH3OH, and ignoring all other possible reac-
tions. However, such an equilibrium is valid only with respect to the
one reaction and is not a true equilibrium for all possible reactions. If
the reacting system is permitted to remain at reaction conditions for an
indefinite time the other relatively slow reactions •will take place with
resultant changes in composition and ultimately a true equilibrium in-
volving all possible products will be reached. The calculation of such
true-equilibrium compositions for complex systems is of little practical
significance, and consideration is ordinarily limited to the relatively few
rapid reactions. I
Isomerization. The simplest case of simultaneous reactions is en-
countered in the isomerization of organic compounds, as, for example/the
isomerization of normal butane to isobutane. Similarly, normal pentane
may be isomerized to either isopentane (2-methyl butane) or neopentane
(2-dimethyl propane). Isomers in equilibrium proportions are produced
in many reactions, particularly in the presence of catalysts.
Calculation of equilibrium compositions in isomerization is simplified
by the monomolecular nature of the reactions and by the fact that the
physical properties of the isomers are so similar that deviations from ideal
behavior cancel. The composition of the equilibrium mixture of isomers
is calculatfed on the basis of a unit quantity of the starting compound in
the equilibrium rnixture. The corresponding quantity of each isomer is
then equal to the equilibrium constant of the reaction by which ittis
formed.
Illustration 14. Prom the data of Table XXXVI calculate the composition of the
mixture obtained by isomerizing n-pentane to equilibrium at 400°K.
n-Pmtane Isopentane Neopentane
-70.32 -70.40 -61.87
T A
(AH;)O -27.030 V -28.450 -31.070
(AQ7r)4oo 0 -3.63 -1.65
K 1 6.21 2.30
CHAP. XVl] COMPLEX REACTIONS 731
The standard free-energy change (AG°/7')4OO accompanying the isomerization of the
normal compound is calculated by the method of Illustration 5. The composition
of an equilibrium mixture containing 1 mole of n-pentane is obtained directly from
Equation (29) to give the following:
Moles Mole per Cent
n-Pentane 1.0 10.5
Isopentane 6.21 65.4
Neopentane 2.30 24.1
Total 9.51 100.0
By repeating the calculations of Illustration 14 at other temperatures the complete
isomerization equilibrium diagram is developed as shown in Fig. 159. From this
diagram it is evident that low temperatures are required for maximum production
of neopentane while isopentane is at a maximum at approximately SOO'K. Similar
diagrams for more complex systems are readily developed once the basic thermo-
dynamic constants are evaluated.
327 427
600
IF. D . Eossini, E . J. R. Presaen and K. S. Pitzer, J. Nat.

Bureau of Standards, 27, 629 (1941), with permission!
FIG. 159. Equilibrium Concentrations of Pentanes.
Complex Reactions. Where a limited number of simultaneous reac-
tions are known to proceed, a generalized procedure for equilibrium cal-
culations may be illustrated by considering the following three reactions
from initial reactants B and C:
(1) bB + cC:?±rR + sS
(2) bB + rR:^tT+uU
(3) cC + sS^vV
If each reaction proceeded to completion, the over-all equation would be
(4) 2bB + 2cC = tT + uU + vV
If the foregoing are all gaseous reactions the respective equilibrium
compositions are expressed by Equation (29):
A X \ ^ / ^ \ ( r + .)-(6 + 0
Ki
\nUcJ Â0 (35)
where 'Sn = nB-\-nc-VnB-\-ns-\-nT + nu-\-nv + ni

From Equation (6): i
AGl = -RT\nKi (38)
AGl^-RThiKi (89)
AGl=-RT\nK^ (40)'
Combining (38), (39), and (40) gives
^G\ = AGl + AGl + AGs = -RT In ^1^2X3 = -RT In K4 (41)
or Ki = KiK^Ki (42)
Thus, the equilibrium constant of the over-all reaction is the product
of the equilibrium constants of the intermediate reactions. By substi-
tution of (35), (36), and (37) in (42)
-=(tF)-'(ir"""''' •<-'
Equation (43) is appHcable to any equilibrium mixture of the system
under consideration, though the intermediate products R and S are
present but are not indicated by the equation. However, by itself it
does not completely define the composition of the equilibrium mixture
and fixes only the relative proportions of components B, C, T, U, and V..
Thus, an over-all equilibrium equation representing the result of a
sequence of reactions is useful for the calculation of complete equiUb-"^
rium compositions only when intermediate products are not present in ,
significant quantities.
In the general case where all intermediate and final products must be
considered, it is necessary that the equilibrium equations of all reactions
be satisfied by the composition of the system at equilibrium. Deter-,
mination of the equilibrium composition therefore requires simultaneous
solution of the independent equilibrium equations. The number of
equations to be solved is equal to the number of independent equations
involved. For example, in the system just considered there are three
independent reactions, (1), (2), and (3). Reaction (4) is not independent
since it results from combination of the others. Thus, simultaneous
solution of the equilibrium Equations (35), (36), and (37) completely
establishes the equihbrium composition) and this composition will of
necessity satisfy Equation (43). I t is a general rule that the number
CHAP. XVI] COMPLEX REACTIONS 733
of independent reactions which must be considered in equilibrium calcula-

tions is equal to the least number of equations which includes every reactant
and product which is present to an appreciable extent in the equilibrium
mixture including all phases.
For simultaneous solution of the equilibrium equations a definite quan-
tity of the initial system is selected as a basis, and the number of moles
converted by each reaction in proceeding to equilibrium is designated
as an algebraic unknown. For example, in the system considered above
an equilibrium calculation may be based on 100 moles of the initial re-
actants. The number of moles of B converted by reaction (1) in going
to equihbrium may b6 designated as x. Similarly, y may represent the
moles of B converted by reaction (2), and z the moles of C converted by
reaction (3). The number of moles of all components in Equations (35),
(36), and (37) may be expressed in terms of these three variables x, y,
and z, which are then evaluated by simultaneous solution. The solution
may reqiire a graphical or trial-and-error procedure where the equations
are complex.
These principles are well illustrated by consideration of the catalytic
process for the production of hydrogen and carbon monoxide by the re-
action of steam and methane at high temperatures. This process is ex-
tensively used for obtaining both pure hydrogen and mixtures of hy-
drogen and carbon monoxide for use in synthesis. A thermodynamic
analysis of the operation involved was presented by Dodge. ^ The kinetic
mechanism of a catalytic reaction of this type may involve several inter-
mediate steps which are of importance in determining the rate of reac-
tion, as discussed in Chapter XVIII. However, a sound thermodynamic
treatment is possible without consideration or knowledge of the true
mechanism."" It may be assumed that the principal reaction proceeds
in two stages with intermediate formation and removal of carbon.
Thus, these reactions with the corresponding equilibrium constants at
600°C from the data of Tables XV and XXXV are as follows:
(1) CH4 = C + 2H2; Ki = -^^2.1Z
iar,r) (flp )
(2) C-|-H20 = CO-t-H2; K^^- - = 0.269
%.o
The over-all reaction is
(<^TT,) idrc)
(3)' CH4 + H2O = CO 4- 3H2; K, = ' — V ^ = 0.574
In order that no carbon may appear in the equilibrium mixture repre-

sented by these three reactions it is necessary that sufficient steam be
added so that the ratio a^j^/ocH, ™^y ^^ equal to or greater than Ki

and that the ratio (aco)('^Hj)/^H20 ^® equal to or less than K2- If the
first ratio is greater than Ki, then carbon added to such a system can
react with hydrogen and form methane until the ratio is reduced to Ki.
If the second ratio is less than K2, then carbon added to such a system
can react with steam until the ratio rises to K^.
Many other reactions are possible in this system. A few, including
the more likely, follow, together with the corresponding equilibrium con-
stants at 600°C.
At eoo^c
(4) CO + H20 = C02 + H2; K4 = 2.21
(5) 2C0 = C + CO2; Ki = 8.14
(6) CO2 = CO + iO^; K, = 4.95(10-1')
(7) H z O - H a + iOa; K, = 1.12(10-12)
(8) 2CH4 = C2H6 + Ha; Kg = 5.51(10-«)
Consideration of the values of these equilibrium constants indicates
that at 600°C reactions (6), (7), and (8) can proceed only to negligible
extents and hence that O2 and CaHg cannot be appreciably present at
equilibrium. When the ratio of steam to methane in the feed is suffi-
ciently high so that carbon cannot be present at equilibrium, the equilib-
rium composition may be calculated from consideration of only Equa-
tions (3) and (4) which involve all the significant reactants in the absence
of carbon. In order to determine the minimum steam ratio required for
freedom from carbon, the equilibrium compositions corresponding to a
series of steam ratios are calculated on the assumption of no carbon for-
mation. The resultant ratios a^JciQ^^ and (fflco^C'^Hj/C^Hao) ^ ^ plotted
against the steam-methane ratio. The minimum steam ratio is that
where aîjacn, = -î ^^^ (aco)(%2)/(%2o) = J^z-
This procedure is demonstrated in the following illustration: {
Illustration 15. Calculate the composition of the equilibrium mixture obtained
when 5 moles of steam react with 1 mole of methane at 600°G and 1.0 atm, assuming
that no carbon is present. Also determine the minimum ratio of steam required for
freedom from carbon.
Basis: 1 mole CH4, 5 moles H2O.
Let X = moles CH4 converted by reaction (3)
y = moles CO converted by reaction (4)
At equilibrium,
Moles CH4 = 1 - X
Moles H2O = 5-x-y
Moles CO = x — y
Moles H2 = 3a; + 2/
Moles CO2 = y
Total moles = 6 + 2a;
CHAP. XVI] COMPLEX REACTIONS • 735
Since the pressure is 1.0 atm, K, may be taken as 1.0. Substitution of the fore-
going values in Equation (29) gives
Reaction (3):
r(x-.)(3. + .)Mr_i^r ^^^^^
Reaction (4): L(l-x)(5-a:-!/)JL6 + 2xJ
r (.)(3x + .) 1,;,^,,.,,
L(a;-'!/)(5-x-y)J
These two equations are solved by assuming several values of x and calculating
the corresponding values of y from each equation. Because of the comphcated forms
of the individual equations the solutions are best carried out graphically. The values
of y from each equation are plotted against the assumed values of x. The intersec-
tion of these two curves gives the correct solution of the two equations. Thus
X = 0.9124
y = 0.633
Equilibrium, Mixture:
Mole per cent
Moles Moh per Cent [Moisture Free)
l-y = 0.0876 1.12 2.01
HjO -x-y = 3.4546 44.15
CO x-y = 0.2794 , 3.57 6.39
H, Sx + y = 3.3702 ' 43.07 77.11
COj y = 0.6330 8.09 14.49
Total 6 + 2x = 7.8248 100.00 100.00
ty ratios for reactions (1), (2), and (5) are
- ^ = 16.5 > (Xi = 2.13)
- 0.0348 < (.Ki = 0.269)

"H2O
^ = 63.4> (,K, = 8.U)

"CO
It follows from these ratios that with 5 moles of steam per mole of methane no carbon
is present at equihbrium. This calculation is repeated for steam ratios of 2 and 1.25.
The corresponding values of Oco'^H2/''H20 ^^^ plotted in Fig. 160, from which the mini-
mum steam ratio is determined as 1.38. The limiting steam ratio can be calculated
directly from reactions (1) and (4), which embrace all components including carbon.
It must be emphasized that the solution of Illustration 15 is based
on the assumption that equilibrium is attained. In actual operations
which do not reach equilibrium quite different results might be obtained,
depending on the rates of the various reactions. Thus, if reaction (1)
were fast and (2) were slow, carbon might form even with high steam
ratios. Conversely, if reaction (3) were primarily a catalytic reaction
and reactions (1) and (5) were relatively slow, operation might be pos-
sible at low steam ratios \vithout carbon formation. These effects are
determined by the kinetics of the reactions and the types of catalysts

present.
Where three significant and independent reactions are involved, the
solution of three simultaneous equations is required with the evaluation
of three unknowns, x,'y, and z.
100 - - Where algebraic methods of so-
"•COt ' ^ „ - ^ ' lution fail or become too com-
o-co ^ ^ ^ plex, a more general graphical
'
solution involves finding the si-
multaneous intersection of three
10 -1^6=8.14 - lines in space; this is equivalent
_
y ^ ^ ^ _ to finding the point of intersec-
/ ^.^^ ClcH, tion of two three-sided pyramids
Ki=2.lZ
- which meet at a common apex.
/
-^1.0 I Carbon This can be done by assuming"
No carbon present - various values of z and plotting
O .present
.2 V the three equations correspond-
w ^2= 0.269 _ ing to reactions I, II, and III in
S^ Ico'^Hj terms of x and y as shown'in
0.1 i tX3CO ^ V ^ ttfljO - Fig. 161. In this figure the in-
01 »H
B II
tersection of the three lines gives
2 a triangular area A at z = 5.0;
- 'Si '•+3c4h „ at another value of 2 = 4.0 the
1 1 1
area of intersection is reduced
0.01 to B; and by further decrease
1 2 3 4
Steam Ratio—H2O/CH4 m z, the area of intersection di-
FIG. 160. Minimum Steam Ratio for miaishes to zero at z = 2 and in-
Absence of Carbon. creases again as values of z are
further decreased.
It is fortuitous that in industrial processes rarely more than three in-
dependent and significant reactions are involved under equilibrimn con-
ditions at a given temperature level.
Metallurgical Reactions. The methods previously developed are dir
rectly apphcabie to the calculation of equihbrivuns in metallurgical proc-
esses such as calcining, roasting, and smelting. The majority of these
reactions are heterogeneous and complex. Such problems may be han-
dled by extension of the procedure demonstrated in Illustration 15.
An interesting and unique case is encountered where the primary re-
actants are all solids while the products include substances in the gaseous
and liquid phases. In such cases a reaction pressure is developed, analo-
gous to the decomposition pressures previously discussed. The reaction
pressure is a result of the nature of the reactants and the temperature,
CHAP. XVI] METALLURGICAL REACTIONS 737
and it is not possible to vary independently temperature and pressure

as is the case where a gaseous reactant is present.
When a reaction starts out in a heterogeneous system any or all of the
solid or liquid phases may be absent when equilibrium is reached.
A typical metallurgical reac-
tion in which all reactants are
solids is the reduction of zinc
oxide by carbon. This problem
has been studied in detail by
Maier.''
Illustration' 16. Zinc <oxide (ZnO)
is reduced by roasting it with carbon
in a closed retort from which the gase-
ous and liquid products of reaction
may be continuously removed. The
atmosphere is excluded from the retort
so that the reaction is free to proceed
under its own equilibrium pressure.
(a) Calculate the equilibrium reac-
tion pressure as a function of the
roasting temperature.
(6) Determine the temperature at
which the operation must be conducted
in order that the products may be
withdrawn at a pressure of 1.0 atm.
(c) CalculateHhe^'minimum temper-
ature and the corresponding pressure
at which zinc is produced as a liquid. FIG. 161. Solution of Three Complex
(d) Calculate the temperature and Simultaneous Equations.
pressure at which the operation must
be conducted in order that 50 per cent of the zinc may be produced directly in the
liquid state.
Solution: Although the over-all effect is the reaction of the two solids, it may be
assumed, in order to account for all the products, that the actual reaction proceeds
in two stages:
(1) ZnO (s)-h CO (g) = Zn (g) + CO2 (g)
(2) C (s) + CO, (g) = 2C0 (g)
The ovSr-all reaction becomes
(3) ZnO (s) + C (s) = Zn (g) + CO (g)
In order to define completely the equilibrium composition, including the CO2

resent, reactions (1) and (2) must be considered simultaneously.
The standard free-energy changes of the two reactions may be expressed as func-
' C . G. Maier, "Zinc Smelting from a Chemical and Thermodjmamic Viewpoint,"
U.S. Bur. Mines Bull., 324 (1930).
tions of temperature by means of the data of Tables V, XIV, and XXXV, together
with the data of Kelley*-' on the heat capacities and heats of vaporization:
For reaction (1),
iSGl = -RT In Ki = 49265 - 70.50r + 5.83T In T - 2.496(10-3)7"'
+ 0.336(lb-<')r3-5:î?^ (a)
For reaction (2), ,
AGJ = -RT ki Kj = 40608 - 24.967 - 2.9777 hi T + 3.484(10-')r2
-0.259 (10-«)r3 _ 58450/r (b)
Basis of Calculations. 1 lb-mole of ZnO reduced which Uberates 1 lb-atom of oxy-
gen appearing as either CO or CO2 in the products of the over-all reaction.
(a) Let X = lb-moles of CO produced
Then (1 — x) lb-atoms of oxygen is present in CO2 or
^ 1 — x) = lb-moles of CO2 formed
The total gas formed is the sum of the CO, Zn, and the CO2, or f
2(3 + x) = total gaseous moles formed
These values are substituted in Equation (29) with the assumption that all zinc is
vaporized and that activities are equal to partial pressures.
K,=l(lr.i) ' (e)

x(3 +x)
K, = ~ (d)
(3-(-x)(l-x)
Combining (c) and (d) by eliminating v gives
4^1x3 = Ki{l - xY (e)
For any given temperature, values of Ki and K2 are obtained from Equations (a)
and (b), and corresponding values of x are calculated from Equation (e). The re-
sultant pressures are then obtained from Equations (c) or (d). These results are
shown in the first four columns of Table A. In Fig. 162 the equiUbrium pressure is
plotted against temperature.
(5) From Fig. 162, where ir = 1.0 atm, T = 1170°K. From Table A, at 1.0 atm,
X = 0.978. The composition of the gaseous products,is obtained directly from this
value. Thus, the total number of moles of products is 3.978/2 = 1.989.
Moles Partial Pressure Mole per Cent
Zn 1.0 0.603 atm 50.3
CO 0.977 0.491 49.1
CO2 0.012 0.006 0.6
1.989 1.000 100.0
For each lb-atom of carbon consumed, 0.503/0.497 or 1.012 lb-atoms of zinc are pro-
duced, or 5.51 lb of zinc are formed per lb of carbon consumed.
(c) Minimum Temperature for Producing Liquid Zinc. If operating conditions are
* K. K. Kelley, "The Free Energies of Vaporization and Vapor Pressures of In-
organic Substances," U.S. Bur. Mines Bull., 383 (1935).
CHAP. XVI] METALLURGICAL REACTIONS 739
such that Uquid zinc may be produced directly at equilibrium conditions, the vapori-
zation equilibrium of zinc must be considered as a simultaneous reaction. Thus
(4) Zn (1) ^ Zn (g)
The equilibrium constant for this vaporization is equal to the vapor pressure of the
liquid zinc if ideal behavior is assumed, or Kt = P^n- 1° ^'g- 162 are plotted vapor
pressures as a function of temperature taken from the data of Maier.'
1200 1300 1400 1500

Temperature, °K
FIG. 162. Equilibrium Conditions in the Reduction of Zinc Oxide.
The minimum temperature at which hquid zinc can be produced occurs when the
partial pressure of the zinc vapor is just equal to the vapor pressure of molten zinc.
This corresponds to the dew point of zinc vapor. The partial pressure of zinc vapor
P2n ^t different reduction temperatures is calculated by the method of part (b) and
tabulated in column 6 of Table A. In column 7 are the vapor pressures Pgn °f liquid
zinc. The partial pressure of zinc vapor p^n ^°<1 ^^^ vapor pressure of molten zinc
Pgn are plotted against temperature in Fig. 162. The intersection point A of these
740 , CHEMICAL EQUILIBRIUM [CHAP. XVI
two curves indicates that the minimum temperature at which molten zinc can occur
is 1292°K, where the partial pressure of zinc vapor is 2.8 atm and the total pressure
is 5.6 atm. However, under these conditions a negligible amount of zinc is formed,
and higher temperatures and pressures must be employed to obtain appreciable yields
of the liquid.
TABLE A
COMPLETE VAPOHIZATION OP ZINC ASSUMED
< ' Pzn ^^'»
yK Ki K, X TT Atm PZn ^<™ = Ki
1100 0.001666 16.35 0.980 0.322 0.162 0.43
1200 0.009014 70.8 0.978 1.607 0.808 1.20
1300 0.03853 255.3 0.976 6.31 3.174 3.06
1400 0.1300 760.4 0.976 20.01 10.07 6.33
1500 0.3740 1940 0.973 54.25 27.31 12.2
By application of the phase rule, Equation (XIV-27), to this system point B is
identified as an univariant point. At this point two solids, one liquid, and one
vapor phase can exist simultaneously. At temperatures below. B two solids &nd
one gas phase exist. At temperatures above 1292°Ksthe system is univariant;
thus, fixing the temperature automatically fixes the pressure.
(d) Yield of Liquid Zinc by Direct Reduction. When liquid zinc is present at equi-
librium conditions, the partial pressure of the zinc vapor must equal the vapor pres-
sure of the liquid zinc at the equilibrium temperature.
Since Equation (29) is applicable only where each component is present in only
one phase, Equation (6) must be used as the basis for developing equilibrium expres-
sions. On the basis of 1 lb-mole of ZnO reduced, let x represent the lb-moles of CO
formed and y the lb-moles of zinc vapor present at equilibrium. Then the lb-moles
of CO2 are ^ 1 — x), and the total moles of gaseous products are J(l -}- a; -|- 2y).
Equilibrium paitial pressures, if ideal behavior of the gases is assumed, are as
follows:
2x7r
(1 - X)T
Pco. = 1 + ^ ^ 2 2 , ^^
or, from (h), ir = (i)

2y
These values are substituted in Equation (16) with the assumption that partial pres-
sures are equal to activities.
Reaction (1), ZnO + CO —>• Zn + COj
2x = ^t (J)
Reaction (2), C + COj —>- 2C0

0
4irx' ^
= Kj • (k)
{l-x){l+z+2y)
CHAP. XVI] EQUILIBRIA IN LIQUID SOLUTIONS 741
Substitution of (i) in (k) gives
Solving (j) for X gives

PZn
(m)
Substitution of (m) in (1) gives
P3
= K2 (n)
2/Xi(Pz„ + 2X1)
p3
or «= (o)
{Pz^ + 2Xi)XiK2
From Equation (o) y is evaluated directly for any selected temperature which
fixes the values of Pgni î> ^^'^ ^2. The corresponding compositions and total pres-
sures are evaluated from Equations (m) and (i). These results are tabulated in
Table B together with the percentages of the total zinc formed which is present in
the liquid state.
TABLE B
THE PRESENCE OP LIQUID ZINC CONSIDEBED
Per Cent
Tempera- Kt K, Pzn V X 5r Liquid Tin
ture °K 100(1-^)
1290 0.0339 226 2.80 1.00 0.976 5.555 0
1353 0.074 464 4.6 0.603 0.970 12.1 39.7
1400 0.1300 760.4 6.33 0.390 0.961 22.2 61.0
1500 0.3740 1940 12.2 0.192 0.941 83.9 80.8
The values of ir in Table B are plotted in Fig. 162 for comparison with those of
Table A which were based on the assumption of complete vaporization. From in-
spection of Fig. 162, 50 per cent liquefaction of zinc wiU occur at 1373°K, 16.3 atm
pressure.
It is evident that the direct production of liquid zinc in yields approaching com-
plete recovery would require operating conditions which cannot be attained with the
structural materials at present available.
EQUILIBRIA IN LIQUID SOLUTIONS

T h e fundamental thermodynamic principles of equilibrium are uni-
versally applicable, and Equation (6) is rigorous for reactions involving
liquid or solid solutions of all types. However, the application of
thermodynamic methods t o such systems has not been particularly
fruitful because of the complex relationships frequently existing be-
tween aativities and concentrations. This situation is illustrated b y
t h e activity coefficients developed in Chapter X I V for N a C l in aqueous
solutions. W h e n more t h a n one solute is involved, as is commonly t h e
case where reactions occur in solution, the complexity of the relation-
ships is compounded to such a n extent as t o render the methods of little
practical value except where ideal behavior is approximated and where

activity coefficients can be assumed to be constant.
Constant activity coefficients can be satisfactorily assumed when one
is dealing with very dilute solutions or in systems involving only closely
related types of materials such as organic homologs. In such cases the
calculation of equilibrium compositions is the same as for gaseous sys-
tems except that pressure is not an appreciable factor. Thus, from
Equation (24),
or, in a form similar to Equation (29),
-|-... + „^)fc + 6 + ...)-(r + . + - ) (45)

In working with liquid solutions it is necessary that the activities and
standard free-energy data used in equilibrium calculations be based on
the same standard states. The entropy of a substance in solution varies
widely with change in concentration. Hence, the standard free erlergy
of formation based on Oi/Aî = 1 when A''i = 0 will be very different from
that based on ai/mi = 1 when mi = 0, even though both are referred to
the state of infinite dilution as explained in Chapter XIV. With this
precaution the procedures in calculating equilibrium compositions are
the same as those demonstrated for the gaseous systems.
Solubility of Carbonates. The application of thermodynamic prin-
ciples to the solubility of the metal carbonates has been developed by
Kelley and Anderson.' Th6se principles are of importance in connection
with studies of the recovery and purification of carbonate ores by leach-
ing operations.
When a metal carbonate MeCOj is dissolved in water in the presence
of carbon dioxide, the reactions shown in Table XXXVIII may take
place. '
TABLE XXXVIII
REAcmoNS IN THE DissoLnTiON OF A METALUC CARBONATE
(1) MeCOs (8) = MeCOs (aq)
(2) MeCOs (aq) = Me+ + + c o r "
(3) MeCO, (s) + H2O = Me++ + OH" -|- H C O r
(4) MeCOs (s) + H2CO8 (aq) = Me++ -|- 2HC0r
(5) MeCOs (s) + 2H2O = Me(0H)2 (s) + H2CO3 (aq)
' K. K. Kelley and C. T. Anderson, "Contributions to the Data on Theoretical
Metallurgy; IV Metal Carbonates-Correlation and Applications of Thermodynamic
Properties," U.S. Bur. Mines Bull, 384 (1935).
CHAP. XVI] SOLUBILITY OF CARBONATES 743
(6) Me(0H)2 (s) = Me(0H)2 (aq)

(7) Me(0H)2 (aq) = Me++ + 2(0H)-
(8) H2O = H+ + OH-
•(9) H2CO3 (aq) = H+ + HCOr
(10) H2CO3 (aq) = HjO + COj (g)
(11) HCO3"" + H2O = OH- + H2CO3 (aq)
(12) HC03- = H+ + C03--
(13) Me(HC03)2(3) = Me(HC03)2 (aq)
(14) Me(HC03)2 (aq) = Me++ + 2(HC03)-
(15) MeC03 (s) + H2CO3 (aq) = Me(HC03)2 (s)
All of these possible reactions must be in equilibrium at saturation

conditions. Since reaction (10) produces COjCg) as a product, it follows
that a definite equilibrium CO2 pressure will result from dissolving a car-
bonate in pure water. Dissolution will at first take place by reactions
(1-5), and carbon dioxide gas will tend to be formed as a result of reac-
tion (10). If the solution is in contact with an atmosphere containing
no partial pressure of carbon dioxide, the gas will be evolved with re-
sultant formation of Me(0H)2(aq), Me"''"'', and 0 H ~ . This reaction
\vill proceed until the Me(0H)2(aq), Me"*""*", and 0 H ~ concentration
reaches saturation and Me(0H)2(s) is formed. The reaction then pro-
ceeds with the over-all formation of Me (OH) 2(3) and C02(g) by reac-
tions (5) and (10) until the MeC03(s) is completely consumed.
If dissolution were to take place in contact with a CO2 free atmosphere
as previously described, the equilibrium CO2 pressure of the solution
would progressively diminish until the first Me(0H)2(s) forms and then
remain constant at the value fixed by the equilibrium of reactions (4)
and (10). Thus, the partial pressure of CO2 corresponding to equilibrium
in reactions (5) and (10) is the minimum equilibrium CO2 pressure ex-
erted by a solution in equilibrium with MeC03(s) and corresponds to an
invariant point at which two solids; one liquid, and one gas phase are in
equilibrium.
If a carbonate is dissolved in water in contact with an atmosphere
containing a partial pressure of CO2 greater than minimum value pre-
viously discussed, dissolution will proceed to an equilibrium fixed by
reactions (1-4) and (7-15). If the CO2 pressure in the atmosphere is
less than the equilibrium CO 2 dissolution pressure of the pure carbonate,
CO 2 will be evolved as equilibrium is approached, and the final solution
will be basic. If the CO2 pressure in the atmosphere is greater than the
dissolution pressure, the solution will absorb CO2 and be acidic when
equilibrium is reached. As the CO2 pressure in the atmosphere is further
increased, the acidity of the solution increases until the concentration
of Me(HC03)2(aq) reaches the saturated value and Me(HC03)2(s) be-
gins to form. The CO2 pressure at which this occurs is determined by
the equilibrium of reactions (16) and (10) and is the maximum pressure
which can be in equihbrium with a solution which is in equihbrium with
MeC03(s). Higher CO2 pressures will result in disappearance of the
solid carbonate and formation of solutions whose compositions are fixed
by the CO2 pressure and the equihbria of reactions (2) and (8-14).
In dealing with the sparingly soluble carbonates the situation is sim-
plified by the fact that th^ salts are completely dissociated in solution.
Thus, reactions (1) and (2), (6) and (7), and (13) and (14) maybe com-
bined into single reactions and the dissolved salts, MeCOs, Me(0H)2,
and Me(HC03)2, disregarded. Furthermore, the solid forms of the acid
carbonates of the metals which form sparingly soluble carbonates are
not known to exist, and reactions (13) to (15) need not be considered.
Additional simplification results from the fact that in dilute solutioils
the activity of water is constant and equal to 1.0.
The equilibrium constant of a reaction in dilute solution is determined
from, its standard enthalpy and entropy changes. In working With
dilute aqueous solutions it is customary to express compositions in
terms of molalities and to select the standard state such that ajm = 1 . 0
where m = 0. The standard enthalpy change in dilute solution is ob-
tained from Tables XIV, XV, and XVI. The standard entropy change
is determined from Table XXXV. I t may be noted that values of zero
are arbitrarily assigned to the heat of formation and the entropy of the
hydrogen ion in dilute solution. The corresponding free-energy change
of formation is then 4,655 cal per g-mole at 298.1°K. All other ionic
heats of formation and entropies are thus expressed relative to the hy-
drogen ion since the individual absolute values are unknown.
Illustration 17. Calculate the equilibrium constant of the following reaction at
298°K:
CaCOs (s) (ppt) ?=i Ca++ (aq) +.COr- (aq)
From Tables XIV and XVI, pages 253 and 272,
Affwi = -160.3 - 129.74 - (-287.8) = -2.24 kcal per g-mole
From Table XXXV,
AS29S = -11.4 - 13.0 - 22.2 = -46.6 cal/(°K) (g-mole)
Neglecting the difference between the heats of reaction at 291° and 298°K yields'
/A(A -2240
-V 46.6 = 39.1
298
—39 1
log K298 = -rzh- = -8.548
4.574
X298= (2.83) (10-^)
In calculating equilibrium constants for ionic reactions great care must
be exercised to use heat of formation and entropy data which are con-
sistent with each other. The entropies of the ions are determined from
CHA-f. XVI] SOLUBILITY OF CARBONATES 745
equilibrium measurements and heat of reaction data. When such en-

tropies are used for the reverse operation of calculating the equilibrium
constant, exactly the same heat-of-reaction data must be used. This
fact results in some confusion in the use of the heats of formation of
H20(l) and COaCg) from Table XIV. The values in this table are re-
cent determinations by Rossini which differ slightly from the values re-
ported by Bichowsky and Rossini. However, the entropies of the ions
and aqueous compounds in Table XXXV are all based on the original
Bichowsky and Rossini values of —68.37 and —94.23 for the heats of
formation of H20(l) and C02(g), respectively. Accordingly, these values
should be used for tlê calculation of equilibrium constants in aqueous
solutions rather than the more accurate values of Table XIV.
The equilibrium composition of a system involving ions is calculated
by the procedure demonstrated in the preceding sections dealing with
complex reactions. In order to define the equilibriimi, it is desirable to
consider the smallest number of reactions which include all reactants
and products present at equilibrium. An algebraic equation relates the
concentrations of the components of each reaction to its equilibrium con-
stant. An additional algebraic equation expresses the electrical neu-
trality of the solution, that is, the equality of the positive and negative
charges of all of the ions. A further equation which may prove useful
expresses the material balance of the dissolution. The equilibrium com-
position is evaluated by simultaneous solution of these equations.
Illustration 18. Calculate the complete compositioa of the solution obtained when
precipitated calcium carbonate is dissolved in water in contact with an atmosphere
containing a partial pressure of CO2 of 3 X 10~^ atm.
Solution: It may be assumed that the specified partial pressure of CO2 is above
that of the invariant point and that the system at equihbrium contains CaCOa (s),
Ca-^, CO3—, HCO3-, H+, OH", and H2CO3 (aq). There are therefore six unknown
compositions to be evaluated, which requires five independent equihbrium equations
in addition to the equation of electrical neutrahty. The material-balance equation
cannot be used in this case, because CO2 may be gained or lost during the dissolution.
Although other selections are possible, it is convenient to consider the following
reactions for the estabhshment of the equilibrium equations. The corresponding
equiUbrium constants are calculated by the method of Illustration 17.
22^298
(a) CaCOa (s) = Ca++ + COr" (2.9) (10-=) = (Ca++) (CO3-")

(b) H2O (1) = H+ + OH- (5.5) (lO-i') = (H+) (OH-)^
(c) H2COa(aq)=H^ + HC03- (9.3) (10-) = f ] l | ^
(d) H2CO3 (aq) = H2O + CO2 (g) 31.7 = - — f M _ ^

liiKj'Ji (aq;
(e) HCOr = H+ + CO3- - (1.4) (10-") = ^^^^gg"^

The equation of electrical neutrality may be written as follows if the chemical for-
mulas are used to denote molahties:
(f) 2Ca+ + + H+ = 2C0r - + HCO3" + OH-
It should be noted that (f) is not a stoichiometric equation.
Since the CO2 pressure is specified, it follows from (d) that the molality of
H2CO3 (aq) is represented by
H2CO3 (aq) = (3)(10-i)/31.7 = (0.95) (10-=) g-moies per 1000 g HÔ
If y is used to represent the molality of the HCOr ion, the following expressions may
be written from Equations (c), (b), (e), and (a) for the molalities of the other ions:
H+ = (0.95) (10-5) (9.3) (10-«)/2/ = (8.8) 10-"/^
OH- = (5.5)(10-'5)i//(8.8)(10-") = (6.25) (10-5)2/
COr~= (1.4)(10-")j/V(8.8)(10-i3) = 15.9J/2
Ca++= (2.9)(10-»)/(15.9)!/2=(1.82)(10-i»)/2/!' /
These values may be substituted in Equation (f), resulting in an equation which con-
tains only y as an unknown. Thus,
(3^64)00-) ^ (8.8)(10->5) ^ 3^_g^, _^ ^^^ ^ ^_^^^^^_^^^ ' •
y^ y
This equation is solved graphically or by trial and error.
y = HCO3- = (7.1) (10-^) g-mole per 1000 g H2O
The concentrations of the other ions are evaluated from the equations relating them
to y. Thus
H+ = (1.2)(10-») g-moIe per 1000 g H2O
OH- = (4.4)(10-«) g-mole per 1000 g H2O
C O r - = (7.9)(10-«) g-mole per 1000 g H2O
Ca++ = (3.7) (10-*) g-mole per 1000 g HjO
The. procedure of Illustration 18 may be repeated for other CO2 pres-
sures and curves plotted relating the composition of the solution to the
equilibrium partial pressure of CO2 above it. Such curves for the solu-
bility of CaCOs in water are shown in Fig. 163. It may be noted that
the concentration of the calcium ion passes through a minimum and -is
increased by either increasing or decreasing the CO2 pressure from this
point. As the CO2 pressure is increased, carbon dioxide is absorbed by
the solution, but as the pressure is reduced CO2 is evolved and the con-
centration of the 0 H ~ ion increases until solid Ca(0H)2 is formed. The
appearance of the added phase results in an invariant point at which the
composition of the solution and the partial pressure of the CO2 are fixed
as long as both solid phases are present. The conditions at the invariant
point are established by the following equation in conjunction with those
of Illustration 18:
(g) CaCOa (s) + 2H2O ^ Ca(OH)2 (s) -f H2CO3 (aq);
K298 = (9.3)(10->^)=H2CO3(aq)
CHAP. XVI] SOLUBILITY OF CARBONATES 747
If calcium carbonate is dissolved in pure water in a closed liquid sys-

tem which permits no gain or loss of carbon dioxide, an equilibrium par-
tial pressure of CO 2 is established which is determined by the equations
of Illustration 18 together with the following material-balance equation
in which the formulas indicate molalities:
(h) Ca++ = CO3— + HCO3- -I- H2CO3
This is not a stoichiometric equation and merely indicates that each
atom of calcium entering the solution is accompanied by an atom of
-14
-13- -12 -10 - 8 - 6 - 4 - 2 0 +2
Logjo CO2 Partial Pressure, Atmospheres
FIG. 163. Solubility of CaCOa in Water.
carbon. The solubility of CaCOs in pure water without gain or loss of

CO2 is determined by simultaneous solution of Equations (a-f) and (h).
It may be noted from Fig. 163 that minimum solubility is obtained under
these conditions, in the case of calcium.
These same principles may be applied to calculating the compositions
obtained in selective leaching and precipitation operations where several
sparingly soluble metals or salts are involved. For each added con-
stituent an additional equation is added to the group which must be
solved simultaneously in order for the equihbrium composition to be
evaluated. However, simpUfication is possible in many instances by
neglect of components which are present in the solution in relatively
small quantities. For example, from Fig. 163 it is evident that at CO2
pressures above 10"^ atm the concentrations of 0 H ~ , C O j " , and H"""
are negligible in comparison to the Ca+"'", HCOJ, and H2CO3. In this

range the dissolution may be considered as proceeding by the following
stoichiometric equation:
CaCOs (s) + H2O (1) + CO2 (g) = Ca++ + 2HC07;
K... = ( 6 . 0 ) ( 1 0 ' ) = — p ^ ^ ^ - ^ ^ ^
Since it is assumed that HCOs" = 2Ca''"+ the foregoing equilibrium

equation may be written
4(Ca++)3
K298 = (6.0) (10-^) =
CO^Cg)
or /
Ca++ = (5.3)(10-')[CO2(g)r
where the chemical formulas indicate activities in molalities and atmos-
pheres, respectively. In this manner the molalities of the Ca^^ &nd
HCO7 ions are determined directly, and the molalities of the other com-
ponents may be individually evaluated from the equations of Illustra-
tion 18. This procedure is generally applicable to the sparingly soluble
carbonates, either singly or in complex mixtures if the CO2 pressure is
equal to or greater than that of the normal atmosphere.
The application of these methods to complex metallurgical problems
is discussed in detail by Kelley and Anderson.^ By assuming that the
activity coefficients cancel out of the equilibrium equations it is possible
to solve effectively many such problems by thermodynamic principles.
PROBLEMS
1. Normal butane is isomerized to isobutane by'the action of a catalyst at mod-
erate temperatures. It is found that equilibrium is reached at the following com-
positions:
Temperature, °C Mole per Cent n-Butane Mole per Cent Isobutane
44 31 - 69
118 43 57
Assuming that activities are equal to mole fractions, calculate the standard free^
energy change of the reaction at each temperature and the average values for the heat-
of-reaction and entropy change over this temperature range.
2. From the data of Tables V, XIV, and XXXV derive equations relating AG°/T
to temperature for the following reactions:
(a) SO2 (g) 4 iO, (g) = SO3 (g)
(b) im (g) + fH, (g) = NH3 (g)
(c) C (s) + H2O = H2 + CO
(d) CO2 -f C = 2 CO
CHAP. XVI] PROBLEMS 749
The atomic heat capacity of graphitic carbon,^" related to T°K, is 2.673 + 0.0026177
- 116,900/72.
3. From the liquid-state values in Table XXXV estimate the entropies in the
ideal-gaseous state at 298.1°K for the following compounds:
Density, 20°C, g per cc
(a) n-Decane
nc
174 0.747
(b) n-Butyl benzene 180 0.862
(o) n-Heptyl cyclohexane 223 0.801
In the absence of data on the physical properties of these compounds use Equa-
tions (III-16) and (III-l) for estimating vapor pressures at 298.1°K and heats of
vaporization at the normal boiling points and Equation (VII-32) for the effect of
temperature on heat of vaporization. Critical temperatures should be estimated by
Equation (III-8), critical pressures by. (III-18), and liquid densities by Equation
(XII-38).
i. From the data of Tables XXXVI and XXXVII calculate the equilibrium con-
stants and heats of reaction of the following teactions at temperatures of 298,600,800,
and 1000°K:
(a) C.He (g) = C,H4 (g) + Ha (g)
(b) 2C2H4 (g) = i-CiHs (g)
(c) CsHs + C2H4 = CsHu (tetramethylmethane)
6. The following values of (H° — HQ) for the ideal-gaseous state were calculated
by Pitzer and Scott," and the corresponding heats of vaporization at the normal boil-
ing point and critical temperatures are from the International Critical Tables. From
the data of Tables XVI and XXXVII calculate the values of (AH/)o the heat of
formation at 0°K in the ideal-gaseous state for each of these compounds. The dif-
ference between the heats of combustion at 291 °K and 298°K may be neglected as
less than the probable errors of the measuremeiits.
( H ° — Ho)298kcal
per g-mole XB cal per g • nc T'cC
(a) Benzene 3.415 94.4 80.1 289
(b) Toluene 4.314 86.5 111 321
(c) Ortho-xylene •5.604 82.8 144.5 358.3
6. Nitrogen tetroxide dissociates into nitrogen peroxide according to the following
reaction:
N2O4 (g) = 2NO2 (g)
The standard free-energy change in calories per gram-mole of this reaction in the ideal
state at 1 atm is represented by the following equation from the International Critical
Tables:
- A G ° =-13,600-|-41.6r°K ^
Calculate the equilibrium composition of the mixture formed from the dissociation
of pure N2O4 under the following conditions, assuming K, = 1.0;
(a) At a temperature of 273°K and 1 atm
(b) At a temperature of 400°K and 1 atm
, i» K. K. Kelley, U.S. Bur. Mines Bull, 371 (1934).
" K. S. Pitzer and P. W. Scott, J. Am. Chem. Soc, 65, 803-29 (1943).
7. Isobutane is alkylated by ethylene to form neohexane, 2, 2-dimethyl butane,

at elevated temperatures and pressures. From the data of Table XXXVI calculate
the composition of the equilibrium mixture resulting from reacting a mixture of 4
moles of isobutane and 1.0 mole of ethylene at 700°K and 100 atm.
8. From the data of Table XXXVII calculate the heats of reaction in the ideal-
gaseous state of the following reactions at 600°K and 1000°K:
(a) CsHs —> CH4 4- C2H4
(b) * C3H8 - » C3H6 + Ha
(c) CO + JOi -> CO2
(d) CH4 + H2O -^ CO + 3H2
9. Benzene may be produced by the catalytic dehydrogenation of n-hexane.
From the results of problem 5 and the data of Tables XV and XXXV calculate the
composition of the equilibrium mixture and the percentage conversion of hexaiie at
a temperature of 1050°F and 1 atm. The following heat-capacity equations may
be used.
Benzenei" 4 = 0.23 -|- 77.83 X IQ'^T - (27.16){\(i-^)T^
n-Hexane" 4 = 4.296 -|- 118.661 X lO-^r - (42.13)(10-«)7'2
10. Propane may be dehydrogenated catalytically to form propylene. From the
data of Table XXXVI develop curves relating the percentage of propane dehydro-
genated at equilibrium to temperature in the range of 800-1200°F and at pressures
of 0.5, 1.0, and 2.0 atm.
11. Water and chlorine react at elevated temperatures according to the following
equation:
4CI2 (g) -1- iHÔ (g) = HCl (g) -f 4O2 (g)
Using the data of Tables V, XIV, and XXXV and the heat-capacity equation of
problem XIII-1, page 692,
(a) Calculate an equation for AG°/T for this reaction.
1(6) Calculate the composition of the equilibrium mixture at SOO°C starting with
equal volumes of chlorine and water and assuming that Kv — 1.0.
(c) Repeat the calculation of part (6) for a pressure of 100 atm, evaluating Kp from
Fig. 142 and the critical data of Table XI, page 234.
12. Sulfur dioxide is reduced by hydrogen according to the following equation:
3H2 (g) + SO2 (g) = H2S (g) + 2H2O (g),
(a) From the data of Tables V, XIV, and XXXV derive an equation for the stand-
ard free-energy change of this reaction.
(6) Calculate the equilibrium composition of the. mixture obtained at 1200°C and
atmospheric pressure, starting with three parts of hydrogen and one part of SO2.
(c) Repeat the calculation of part (6) at 100 atm, using Fig. 142 and the critical
data from Table XI, page 234,
13. Methyl alcohol is synthesized by passing a mixture of CO and H2 over a cata-
lyst according to the following equation:
CO (g) + 2H2 (g) = CH3OH (g)
" J. W. Andersen, G. H. Beyer, and K. M. Watson, Natl. Petroleum News, Tech.
Sec, 36, fl476 (July 5, 1944). Also "Process Engineering Data," National Petroleum
Publishing Company, Cleveland (1944).
'3 H. M. Spencer and G. N. Flannagan, J. Am. Chem. Soc, 64, 2511 (1942).
CHAP. XVI] PROBLEMS 751
From the data of Tables V, XV, and XXXV calculate the composition of the equi-
librium mixture obtained at a temperature of 300°C and a pressure of 240 atm, start^
ing with two parts of H2 and one part of CO, assuming that only this reaction takes
place. The critical temperature of CH3OH is 240°C and the critical pressure 79 atm.
The heat capacity is given by the following equation (Chapter XVII, page 801) with
r in °K: . ' ''
Cp = 5.72 + (24.85) (10-3)r - (8.167)(10-*)r2
14. Methyl alcohol is oxidized by air to formaldehyde at 550°C in a catalyst
chamber at atmospheric pressure. Calculate the percentage yield of formaldehyde
using the theoretical air supply and assuming no further oxidation.
CH3OH (g) + iO, (g) ^ HCHO (g) + H2O (g)
Entropy and heat-of-reaction data are given in Tables XXXV and XV. Heat ca-
pacities may be taken from Table V and problem 12 and the following equation for
HCHO12 with T in °K:
Cp = 4.50 + (13.95) (10-5) r - (3.73) (10^)^2
15. In the Birkeland-Eyde process the nitrogen of the atmosphere is oxidized in
a long flaming electric arc:
1N2 + 4O2 = NO
Assuming that only this reaction takes place, calculate the percentage conversion"of
nitrogen to NO in air of average atmospheric composition at a pressure of 1.0 atm
and at temperatures of, respectively, 2000°K and 3000°K.
16. Carbon dioxide is reduced by graphite according to the equation:
C (graphite) + CO2 (g) = 2C0 (g)
Assuming that equiUbrium is attained, calculate the degree of completion of the re-
duction of pure CO2 under the following conditions, using the data of Table XXXVII:
(a) A temperature of 1000°K and a pressure of 1.0 atm, assuming Kv = 1.0
(6) A temperature of 1500°K and a pressure of 1.0 atm, assummg Kv — 1.0
(c) A temperature of 1000°K and a pressure of 100 atm
17. A mixture of 79 per cent N2 and 21 per cent CO2 by volume is passed over
hot carbon (graphite) at a temperature of 1000°K and a pressure of 1.0 atm. Using
the data of Table XXXVI calculate the equilibrium composition of the gases, and
compare this result with that of part (a) of problem 16.
18. Carbon monoxide is burned with pure oxygen'in the theoretically required
proportions. Calculate the degrees of completion of the oxidation if equilibrium is
attained at temperatures of, respectively, 1000°K and SOOO'K, under a pressure of
1.0 atm. Evaluate iiG°/T at 3000°K by extending the data of Tables XXXVI and
XXXVII from 1500°K through graphical integration of the heat capacities of
Table IV, page 2^2.
19. Water gas leaves a generator containing 51.1 per cent H2, 2.3 per cent CO2,
and 46.6 per cent CO by volume on the dry basis. Ten per cent of the steam which
was introduced into the bottom of the generator passed through the bed of hot coke
without decomposition and is present in the gases. This gaseous mixture is passed
into a reaction chamber under a pressure of 1.0 atm in contact with a catalyst of
chromium oxide and allowed to attain equiUbrium at a temperature of 700°K (423°C).
Calculate the equilibrium composition of the gaseous mixture, using the data of
Fig. 156.
20. One volume of the initial wet water gas described in problem 19 is mixed with
three volumes of additional water vapor. THis mixture is passed into the reaction
chamber operated at the conditions described in problem 19 and allowed to reach
equilibrium.
(a) Calculate the equilibrium composition of the gaseous mixture.
(b) Calculate the composition of the residual gas if the CO2 and H2O are removed
from the gaseous mixture of, part (a) after equihbrium is attained.
21. The hydrate of sodium carbonate decomposes according to the following
equation:
Na2C03 • H2O (s) = NaaCOa (s) + H2O (g)
The equilibrium pressure in atmospheres of water vapor in this reaction is given by
the following equation:
log p = 7.944 - 3000.0/r°K /
Derive an expression for the standard free-energy change of this reaction.
22. Zinc oxide is reduced with carbon monpxide, under a pressure of 1.0 atm ac-
cording to the following reaction:
ZnO (s) + CO (g) = Zn (g) + CO2 (g)
Calculate the degree of completion of the oxidation of CO at atmospheric pressure
under the following different conditions, assuming that equilibrium conditions are
attained and that ZnO is always present:
(a) At a temperature of 1000°C, using pure CO
(b) At a temperature of 1500°C, using pure CO
(c) At a temperature of 1000°C, using a mixture of 27.5 per cent CO, 4.3 per cent
CO2, and 68.2 per cent N2 by volume
(d) At a temperature of 1500°C, using a mixture of 27.5 per cent CO, 4.3 per cent
CO2, and 68.2 per cent N2 by volume
23. Sodium bicarbonate is calcined according to the following equation:
2NaHC03 (s) = Na2C03 (s) + H2O (g) + CO2 (g)
Calculate the pressure of the equimolecular mixture of H2O and CO2 in equilibrium
with NaHCOa at a temperature of 100°C.
24. (a) Calculate the boiling point of aluminum at atmospheric pressure from the
following data:'
G°-H; G°-Hg
Tempera- T T
ture, °K Cos I Solid or Liquid
298.1 -33.766 - -3.13
1000 -40.217 -9.09
2000 -43.762 -14.64
2100 -44.009 -15.01
2200 -44.245 -15.37
2300 -44.470 -15.71
2400 -44.681 -16.04
The value of AHJ for vaporization is 66.920 kcal per g-atom,
(b) Calculate the vapor pressure at 200G''K.
25. Considering each of the following reactions:
Sec,, S02(g) + i02(g) = SOs(g)
Publishu . iN2(g) + iOiCg) = NO(g)
" H . M. C(s)+C02(g) =2C0(g)
CHAP. XVII PROBLEMS 753
Tabulate the effects of the following changes upon (a) the velocity of reaction (moles
transformed per unit time per unit volume), (6) the equilibrium degree of completion,
and (c) the actual degree of completion obtained in a specified time interval:
(1) Increase of temperature
(2) Increase of pressure
(3) Provision of a positive catalyst
(4) Dilution with an inert gas
(5) Agitation of the reacting system
Tabulate each effect as an increase, decrease, no effect, or indeterminate.
1 26. In the American process of synthesizing ammonia a mixture of three volumes
of hydrogen and one volume of nitrogen is passed into a reaction chamber in contact
with a catalyst of granular iron oxide combined with oxides of potassium and alumi-
num. Using the data of Illustration 7 and Fig. 156:
(a) Plot curves relating the equiUbrium degree of completion of this reaction to
temperature at pressures of 1.0, 100, and 300 atm. The temperature range from
400°K to 800°K should be included in the calculations.
(6) Calculate the equilibrium degree of completion of the afore-mentioned reaction
at a pressure of 200 atm and a temperature of 750°K.
27. In the synthesis of ammonia described in problem 26 the mixture of N^ and
Hj is introduced into the reaction chamber under a pressure of 300 atm and a tem-
perature of 400°C. Assuming that heat loss from the reaction chamber is negligible,
calculate the equilibrium-reaction temperature, using the curves of problem 26.
28. From the results of problem 10 and the data of Table XXXVII calculate the
equilibrium temperatures and conversions reached when propane is dehydrogenated
in an adiabatic reactor with an inlet temperature of 1200°F and pressures of 1.0 and
2.0 atm, respectively.
29. From the data of Table XXXVI calculate the complete compositions ofthe
equilibrium mixtures resulting from the polymerization of pure ethylene to the bu-
tenes if it is assumed that no other reactions occur at the following conditions:
Temperature, °C Pressure, lb per sq in.
(a) 150°C 300
(&) 150°C 600
(c) » 200°C 300
(d) 200°C 600
30. From the data of Table XXXVII calculate the composition of the equilibrium
mixture of the five isomeric hexanes at a temperature of 600°K. Calculate the heat
of reaction in converting pure normal hexane to the equilibrium mixture in the ideal-
gaseous state at this teiSperature.
31. When propane is pyrolyzed the following reactions take place:
CsHg —> CsHe -|- Hj
CgHa —> C2H4 -f- CHi
Assuming that only these two reactions occur, calculate from the data of Table
XXXVI the composition of the equilibrium mixture formed by heating propane
at 1400°F and 1 atm.
32. Propane also decomposes to form carbon and hydrogen at high temperatures.
Calculate the composition of the equilibrium mixture of probleni 29 if this reaction
is considered.
774 THERMODYNAMIC PBÊRTIES [CHAP. XVII
Since entropy is independent of pressure for the contributions under

consideration, from Equation (o) of Table XXIV, page 472
dSrn = c^^dT/T^c^md In T (26)
^1 r,d}n_Q
(s° - &l)m = RlnQ (27)
0 dlnT
From Equation (20) defining the partition function it is evident that at
the absolute zero Qo = go, which is the statistical weight of the ground
energy level or the number of different molecular states possessing
substantially the ground energy level. Since Qo is finite it follows
that d In Q/d In T is equal to zero where T" = 0, and Equation (27)
becomes:
sl = R\nQ + R^^+{sl,n-Rlngo) (28)
However, the most general statement of the third law of thermody-

namics derived by statistical mechanics states that the entropy of a sub-
stance at the absolute zero is equal to R times the logarithm of the statistical
weight of the ground energy level, or si = R In go. For substances in the
crystalline state at 0°K the statistical weight of the ground level is gen-
erally unity, leading to the more common statement of the third law.
Equation (28) then becomes
s'L = R\nQ + R—-^ 29)
d In 1
Combining Equations (24) and (29) gives expressions for the m contribu-
tions to G° and A°, thus,
(G° - H°)„ (A°-HIU ,„„,
• = - f t In Q = (30)
By adding the m contributions calculated from Equations (24), (25),

(29), and (30) to the translational contributions from Equations (8) to
(17) the thermodynamic functions c°, c°, s°, u° — n°, H° — HQ, G° — H",
and A° — HQ can all be completely evaluated if the partition function is
known as a function of temperature. If the contributing molecular
energy levels and their statistical weights are known, Q™ may be evalu-
ated by the summation of Equation (20). Similarly, from Equation (22),
d In Q _-2iEigie RT
d In T RTQ ^ '
CHAPTER XVII .
THERMODYNAMIC PROPERTIES FROM MOLECULAR .
' STRUCTURE
The application of thermodynamic principles is in many cases ham-
pered by lack of data which are sufficiently reliable to be useful for engi-
neering design. The accuracy with which standard free energies mustf
be known is evident from consideration of the error AK, in the equiUb-
rium constant which corresponds to an error A(AG°) in the standard free-
energy change. Thus, from Equation (XVI-6) at constant temperature,
-AiAG°) = RTln(^+l\ (1)
An error of 10 per cent in the equilibrium constant {AK/K = 0.10) at a

temperature of 300°K corresponds to an error in the standard free-energy
change of only 0.057 kcal per g-mole. Such an error results from an
equal error in the standard heat of reaction or an error of 0.2 entropy
unit. Since the heats of combustion of even simple organic compounds
generally exceed 200 kcal per g-mole it follows that for such a compound
an error in the heat of combustion of only 0.03 per cent may result in
an error of 10 per cent in the equilibrium constant based upon it. An
error in heat of reaction of only 0.42 kcal per g-mole or 1.4 entropy units
will result in a 100 per cent error in the equilibrium constant at 300°K.
It is evident that, where equilibrium is a limiting factor in a process,
measurements of the highest accuracy are necessary in order to develop
thermodynamic properties which are satisfactsry for engineering cal-
culations. Such measurements have been carried out on relatively few
materials, and in the present state of the basic data it is commonly neces-
sary to resort to every available device for extrapolation and interpola-
tion from available measurements.
Developments in spectroscopy and statistical mechanics together with
improved methods of low-temperature calorimetry have resulted in re-
liable heat-capacity and entropy values for a variety of compounds, as
indicated by Table X X X V . ' The corresponding heats of formation are
generally less reliable. The equilibrium constants calculated from much
of the available heat-of-formation data are useful only for predicting the
general feasibility of reactions and the ranges of operating conditions in
which equilibrium effects may be neglected.
756
CHAP. XVII] EMPIRICAL CORRELATION OF ENTROPIES 757
Where reasonably reliable entropy and heat-capacity data are avail-

able or can be estimated by the procedures outlined in this chapter, a
complete expression for the equilibrium constant which is satisfactory
for many purposes may be developed from a single experimental measure-
ment of the compositions at equilibrium under conditions not far re-
moved from the range of interest. The standard heat of reaction i^°
is calculated from Equation (XVI-7) from the kno\vn entropy and equi-
librium constant. Equilibrium constants at other conditions may then
be calculated from Equations (XVI-6) and (XVI-11). In this manner a
value of AH° is obtained which is consistent with the value of /^S° in the
expression for the equilibrium constant, and moderate errors in the en-
tropies used will not seriously affect equilibrium calculations over a
limited range of conditions.
EMPIRICAL CORRELATION OF ENTROPIES

Methods o"8^forreIating the entropies of inorganic compounds are re-
viewed by Wenner,^ who points out that the entropies of similar com-
pounds in the solid state vary with molecular weight in accordance
with the following equation:
s, = A\ogM + B (2)
where s = molal entropy
M = molecular weight
A, B = constants characteristic of the type of compound
On this basis a group of similar compounds should yield a straight line
when molecular weight is plotted on a logarithmic scale against entropy
on a uniform scale. Data on two compounds serve to establish such a
line, or the constants of Equation (2), from which entropies of similar
materials may be estimated.
Wenner found that the points for individual compounds in each group
fall fairly well on a straight line for each of the following groups of metal-
lic compounds in the solid state: Me02, MeO, MeS, MeNOa, and metallic
haUdes, MeX and MeX2. A further generalization pointed out is that
the addition of 1 mole of H2O to a soHd compound causes an entropy
increase of 10 units in forming a crystalline hydrate, 8.2 in forming an
inorganic base from its oxide, and 7 units in forming an organic acid
from its corresponding anhydride.
In the gaseous state Wenner shows that molal entropies vary with
molecular weight according to the equation:
logs(g)=Alogilf-l-B (3)
* R. R. Wenner, " Thermochemical Calculations," McGraw-Hill Book Company
(1941).
758 THERMODYNAMIC PROPERTIES [CHAP. XVII
where A and B are constants determined by the number of atoms in the

molecule. Thus, all monatomic gases, fall reasCnably on a straight Hne
•when entropy is plotted against molecular weight on logarithmic scales.
Another line is formed by diatomic gases. This relationship is applicable
to gases containing as many as nine carbon atoms. However, errors
ranging up to six or seven units are encountered, and Wenner recom-
mends that entropies estimated from these relationships should be used
only as rough approximations.
GROUP CONTRIBUTIONS
Various methods for empirically correlating heats of formation, en-
tropies, and heat capacities of organic compounds were reviewed by
Andersen, Beyer, and Watson^.who proposed a scheme whereby these
properties for the ideal-gaseous state are resolved into contributions
attributable to atomic groups. From the restllting tables the properties
of complex molecules are readily estimated by summation of the contri-
butions of their component groups. The corresponding properties in
the liquid and nonideal-gaseous states may then be calculated by the
methods of Chapter XII.
The molal values for the ideal-gaseous state of heats of formation at
25°C, entropies at 25°C, and 1 atm, and the constants a, b, and c of the
three-term heat-capacity equation (VII-21), page 213, were resolved
into group contributions on the following basis:
Each compound is considered as composed "of- a basic group which is
modified by the substitution of other groups for atoms comprising it.
For example, all paraffin hydrocarbons may be considered as derived
from methane by successive substitution of CH3 groups'for hydrogen
atoms. Similarly, any secondary amine can be considered as derived
from the base group NH(CH3)2. The contributions of nine base groups
are given in Table XXXIX.
The contributions resulting from the primary substitution of a methyl
group for a hydrogen atom in any one of the base groups is given in
Table XL. In the cases of benzene, naphthalene, and cyclopentane, the
base group contains several carbon atoms, and successive substitutions
on different carbon atoms involve different contributions depending upon
the number and position of the substituted groups. For the naphthenes
the terms ortho, meta, and para are taken as indicating a minimum sep-
aration of the two substituted groups by respectively 0, 1, and 2 inter-
mediate carbon atoms in the ring. The heat-capacity coefficients, a, b,
^ J. W. Andersen, G. H. Beyer, and K.^M. Watson, Nat'l Petroleum News, Tech.
Sec., 36, fi476 (July 5,1944). Also " Process Engineering Data," National Petroleum
CHAP. XVII] PROPERTIES FROM PARTITION FUNCTIONS 775
Also, since d} In Q/dQxi Ty = Td{d In Q/d In T)/dT

-3- -ii -Ii
d{\nTy~ R'T^Q RTQ \ RQ^ )\dT)

-li.
^ ZjE^gie RT^ d\nQ _ /d In QV
R'T^Q dlnT \d In Tj ^ '
From Equations (31) and (32) it is seen tliat all the derivatives of the
partition function which are required for the evaluation of thermody-
namic properties can be calculated from two summations of all significant
energy levels, one SjBigfiC ^^ and the other Sifif giC «'^. For simple mole-
cules the individual energy levels can be evaluated by application of
Equation (6) to spectroscopically determined frequencies of emission or
absorption.
Illustration 7. T h e development and application of these methods for calculating
thermodynamic properties is outhnod by Kelley'", who developed the following cal-
culation of the thermodynamic properties of nickel vapor in the state of an ideal gas
a t 298°K and 1 atm. This is an entirely hypothetical state in which nickel vapor
cannot exist but is legitimately used as a reference from which values for actual
states may be derived by the temperature and pressure relations developed in Chap-
ters X I and X I I .
Nickel vapor, Ni (g), is a monatomio gas possessing only translational and elec-
tronic-energy contributions. In Table A the spectroscopic notations for the various
levels of electronic energy are given in column 1 and the corresponding frequencies
in terms of wave numbers and statistical weights, derived from the spectrum, in
columns 2 and 3, respectively. The molal energy levels in column 4 are calculated
from Equation (6) in calories per gram-mole. The three summations required for
calculation of thermodynamic properties are given in columns 5, 6, and 7. The
electronic contributions to the thermodynamic functions are calculated directly from
these summations. Thus, from Equations (20), (31), and (32), and Table A,
Q = 11.690
d In Q 1753
= 0.2532
d\nT (1.987) (298.1) (11.69)
dMnQ 1,533,000
-0.2532 -(0.2532)2 = 0.0565
rf(lnr)2 (1.987)^298.1)2(11.69)
TABLE A
1 2 S 4 B e 7
State 0) 9i Ei ff-ie RT QiEie RT giE\e~'Sf
^Ti 0.0 9 0 9.0000 0 0

3r.3 204.82 7 585.27 2.6058 1525.1 892,610
3n. 879.82 6 2513.98 0.0717 180.3 453,200
Fa 1332.15 7 3806.62 0.0113 43.0 164,060
1713.11 3 4895.21 0.0008 3.9 18,500
3D,
2216.55 5 6333.79 0.0001 0.6 4,551
%a
Q = 11.690 1753 1,533,000
CHAP. XVII] GROUP CONTRIBUTIONS 759
and c, apply to ideal gases where temperature is expressed in degrees

Kelvin.
TABLE XXXIX
BASE GROUP PBOPEBTIES
AH;298.I (g) Swa.i (g) a b(lO') c(,W)

Group ical per g-mole cal/(g-mole) (°K) Ideal Gas at T'K
Methane -17.9 44.5 3.42 17.85 -4.16
Cyclopentane -21.4 70.7 2.62 82.67 -24.72
Benzene 18.1 64.4 0.23 77.83 -27.16
Naphthalene 35.4 80.7 3.15 109.40 -34.79
Methylamine -7.1 57.7 4.02 30.72 -8.70
Dimethylamine -7.8 65.2 3.92 48.31 -14.09
Tfknethylamine -10.9 % 3.93 65.85 -19.48
Dimethyl ether -46.0 63.7 6.42 39.64 -11.45
Fonnamide -49.5 — 6.51 25.18 -7.47
TABLE XL
CONTBIBUTIONS OF PRIMARY C H 3 SUBSTITUTION GROUPS REPLACING H T D B O G E N
A(AH;298.I) (g) AS29S.1 (g) Aa Ab(lO') Ac(10«)

Base Group kcal per g-mole cal/ (g-mole) (°K) Ideal Gas at fK
1. Methane -2.2 10.4 -2.04 24.00 -9.67
2 Cyclopentane
(o) Enlargement of ring —9.3 0.7 -1.04 19.30 -5.79
(b) First substitution —5.2 11.5 -0.07 18.57 -5.77
(c) Second substitution:
ortho -12.2
meta —8.4 -0.24 16.56 -5.05
para —7.1
(d) Third substitution - 7 . 0 — — — —
3. Benzene and naphtha-
lene
(a) First substitution —4.5 12.0 0.36 17.65 -5.88
(6) Second substitution:
ortho —6.3 8.1 5.20 6.02 1.18
meta —6.5 9.2 1.72 14.18 -3.76
para —8.0 7.8 1.28 14.57 -3.98
(c) Third substitution
(sym) — 8.0 0.57 16.61 -5.19
4. Methylamine —5.7 —
5. Dimethylamine —6.3 — -0.10 17.52 -5.35
6. Trimethylamine —4.1 '~~ .
7. Fonnamide
Substitution on C atom —9.0 — 6.11 -1.75 4.75
If more than one substitution is made on a single carbon atom of a
base group the additional contributions are treated as secondary sub-
stitutions and evaluated from Table XLI.
7G0 THERMODYNAMIC PROPERTIES [CHAP. XVll
The contributions resulting from the secondary substitution of methyl

groups for hydrogen atoms are classified in Table XLI according to A,
the type of the carbon atom on which the substitution is made, and B,
the highest type number of an adjacent carbon atom. The carbon-atom
types are defined on the basis of the number of hydrogen atoms attached.
Thus:
Type 1 , —CH3
-i IHj
3 —CH
I
1
4 —C—
I
5 C atom in benzene or naphthalene ring

Two special secondary substitutions are defined in Table XLI for use
in calculating the properties of esters and ethers. One is the substitution
of a methyl group for the hydrogen of a carboxyl group to form a methyl
ester. The other is the substitution of a methyl group for one of the
hydrogens of a methyl ester or ether to form an ethyl ester or ether. If
additional substitutions are made on this same carbon atom, the con-
tributions are evaluated from the corresponding values of A and B based
on the type numbers of the carbon atoms involved.
In Table XLII are the contributions resulting from the substitution
of multiple bonds for single bonds between two carbon atoms of types A
and B, respectively. An additional contribution must be added for each
pair of conjugated double bonds formed by such substitutions.
In Table XLIII are the incremental contributions resulting from the
substitution of various groups for one or two methyl groups. Thus, if
a methyl group is replaced by an OH group the contribution to the heat
of formation is —32.7. If two methyl groups are replaced by an oxygen
atom to form an aldehyde the contribution is —12.9. It may be noted
that the phenyl group is included in Table XLIII in addition to being
designated as a base group in Table XXXIX. This substitution con-
tribution is used in calculating the properties of complex compounds in
which several base groups are combined, as, for example, in polybasic
aromatic acids.
The contributions to heats of formation resulting from substitution of
chlorine for methyl groups vary with the number of substitutions made
on a single carbon atom. Corresponding variations were not found for
the contributions to entropy or heat capacity or for the substitution of
the other halogens. As noted in the table a correction must be apphed
to tê entropies calculated for the halogenated methanes. In general
TABLE XLI
SECONDARY METHYL SUBSTITUTIONS REPL.ACING HYDROGEN
A(AH;)J9S.I (g) AS°298.1 (g) Aa A6(103) Ac(10»)
A B kcal perg-mole cal/(g-mole) (°K) Ideal Gas at T'K
1 •1 -4.5 9.8 -0.97 22.86 -8.75
1 2 -5.2 9.2 1.11 18.47 -6.85
1 3 -5.5 9.5 1.00 19.88 -8.03
1 4 -5.0 11.0 1.39 17.12 -5.88
1 5 -6.1 10.0 0.10 17.18 -5.20
2 1 -6.6 5.8 1.89 17.60 -6.21
2 2 -6.8 7.0 1.52 19.95 -8.57
2 3 -6.8> 6.3 1.01 19.69 -7.83
2 4 -5.1 6.0 2.52 16.11 -5.88
2 6 -5.8 2.7 0.01 17.42 -5.33
3 1 -8.1 2.7 -0.96 27.47 -12.38
3 2 -8.0 4.8 -1.19 28.77 -12.71
3 3 -6.9 6.8 -3.27 30.96 -14.06
3 4 -5.7 1.7 • -0.14 24.57 -10.27
3 5 -9.2 1.3 0.42 16.20 -4.68
1 —0—in ester
or ether -7.0 14.4 -0.01 17.58 -5.33
Substitution of H
ofOHgrou pto form
ester +9.5 16.7 0.44 16.63 -4.95
TABLE XLII
MULTIPLE-BOND CONTRIBUTIONS REPLACING SINGLE BONDS
Type oj A ( A H ; ) 2 9 8 . 1 (g) AS^gS.! (g) Aa Ab(lO') Ac(106)
A Bond B Iccal per g-mole cal/ (g-mole) (°K) Ideal Gas at T°K
1 = 1 32.8 -2.1 1.33 -12.69 -1-4.77
1 = 2 30.0 0.8 1.56 -14.87 -f5.57
1 = 3 28.2 2.2 0.63 -23.65 -f-13.10
2 = 2 28.0 -0.9 0.40 -18.87 -f9.89
2 = 2 cis 28.4 -0.6 0.40 -18.87 -t-9.89
2 = 2 ttans 27.5 -1.2 0.40 -18.87 4-9.89
2 = 3 26.7 1.6 0.63 -23.65 +13.10
3 = 3 ,25.5 — -4.63 -17.84 +11.88
Additional correction for

each pair of conjugated
double bonds -3.8 -10.4 Approximately zero
1 ^ 1 74.4 -6.8 5.58 -31.19 +11.19
1 = 2 69.1 -7.8 6.42 -36.41 +14,53
2 = 2 65.1 -6.3 4.66 -36.10 +15.28
Correction for double
bond adjacent to aro-
matic ring -5.1 -4.3 Approximately zero
TABLE XLIII
SUBSTITUTION G B O U P CONTRIBUTIONS REPLACING CHa1 G R O U P
A(A°H/)29a.i (g) AS°298.1 (g) Aa Ab(lO') Ac(10«)

Group kcal per g-mole cal/(g-mole)(°K) Ideal Gas at T°K
-OH (aliphatic,
meta, para) —32.7 2.6 3.17 -14.86 5.59
-OHortho -47.7 ) — — — —
-NO2 1.2 2.0 6.3 -19.53 10.36
•CN 39.0 4.0 3.64 -13.92 4.53
•CI 0 for first CI on
a carbon; 4.5 for 0 2.19 -18.85 6.26
each additional
-Br 10.0 3.0t 2.81 -19.41 6.33
-F -35.0 -i.ot 2.24 -23.61 11.79
•I 24.8 5.0t 2.73 -17.37 4.09
: 0 aldehyde -12.9 -12.3 3.61 -55.72 22.72
: 0 ketone -13.2 -2.4 6.02 -66.08 30.21
•COOH -87.0 15.4 8.50 -15.07 7.94
•SH 15.8 5.2 4.07 -24.96 12.37
•CSHB 32.3 21.7 -0.79 53.63 -19.21
•NH2 12.3 -4.8 1.26 -7.32 2.23
t Add 1.0 to the calculated entropy contributions of halides for methyl derivatives;
for example, methyl chloride = 44.4 (base) + 10.4 (primary CH3) — 0.0 (CI substitu-
tion) + 1.0.
the calculated results tend to be uncertain for single carbon-atom

compounds.
The suggested sequence of operations in estimating the properties of
a complex compound is as follows:
1. Select the base group and determine its properties from Table
XXXIX. Where a choice of base group is possible, select the group
having the largest entropy. Proceed to build up the desired compound
with as few substitutions as possible.
2. Add the contributions given in Tables XL and XLI which result
from all CH3 substitutions replacing hydrogen required to establish the
carbon skeleton of the compound. In this operation the longest straight
chain should be built up first and then the longest side chain. Where
the same compound may be arrived at by alternate substitutions an
average result is used.
3. Add the contributions from Table XLI which result from addi-
tional CH3 substitutions replacing hydrogen in the positions occupied
in the compound by other groups which are listed in Table XLIII.
4. Add the contributions given in Table XLII which result from mul-
tiple bonds.
5. Add the incremental contributions given in Table XLIII which
result from replacement of CH3 groups by substitution groups.
Illustration 1. . Approximate tiie standard iieat of formation AH/JSS.I (g) of 2,2,4
— trimetliylpentane at 25°C.
Base group (methane) Table XXXIX -17.9
Primary CH3 Table XL -2.2
Secondary CH3 substitutions by successive replacement of hydrogen (Table XLI)
A B
1 1 -4.5
Secondary methyl groups
1 2 -5.2
in 5 carbon chain
1 2 -5.2
2 2 -6.8
Side methyl'groups 2 2 -6.8
3 2 -8.0
56.6
The value found by Rossini and co-workers^ is 56.2 kcal per g-mole.
Illustration 2. Approximate the value of AH/293.I (g) for dimethyl phthalate.
C O Base group (benzene) Table XXXIX 18.1
/ ^ Primary CH3 replacing H Table XL -4.5
C C—C -0CH3 Ortho CH3 replacing H Table XL -6.3
II I -COOH replacing CH3 Table XLIII -87.0
C C—C—OCH3 -COOH replacing CH3 Table XLIII -87.0
\ ^ II CH3 replacing H of-COOH Table XLI 9.5
C O . CH3 replacing H of-COOH Table XLI 9.5
AH/298.1 = -147.7
The value of AH/298, i calculated from the heat of combustion listed by Kharasch* is
—147.1 kcal per g-mole.
Illustration 3. Approximate the entropy of 2,2,3,3-tetramethyl butane
(a) Base group (methane) Table XXXIX 44.5
Primary CH3 Table XL 10.4
Secondary CH3 substitutions replacing hydrogen Table XLI
A B
Secondary CH3 groups in 4 C chain
1: 1
2
2
9.8
9.2
7.0
3 6.3
Side methyl group clockwise
3 5.8
4 1.7
94.7
(6) Base Group (methane) Table XXXIX 44.5
Primary CH3 Table XL 10.4
Secondary CH3 substitutions replacing hydrogen Table XLI
3 K. S. Pitzer, Chem. Rev., 27, 39 (1940).
* M. S. Kharasch, Bur. Standards J. Research, 2, 359 (19^9).
764 THERMODYNAMIC PROPERTIES [CHAP. XVll
B
1
Secondary CH3 groups in 4 C chain
II 9.8
2
9.2
2
7.0
2
3 2
4.8
Side methyl groups counterclockwise 4
i2 6.0
'3 4
1.7
93.4
The average result for the entropy of 2,2,3,3-tetramethyl butane is 94.05 cal/
(g-mole) (°K) which is in agreement with the value found by Pitzer.'
Illustration 4. Approximate the entropy of 1,2-dibromoethane.
Base group (methane) Table XXXIX 44.5
Primary CH3 replacing hydrogen Table XL 10.4
Secondary CH3 substitutions replacing hydrogen:
A B
1 1 Table XLI 9.8
1 2 9.2
Br substitution replacing CH3 Table XLIII 3.0
Br substitution replacing CH3 Table XLIII 3.0
79.9
This result is in agreement with the value reported by Pitzer.'
Illustration 5. Calculate the heat required to raise 1 mole of 2,3-dimethylpentane
in the ideal-gaseous state from 298.1°K to 1000°K.
Ao A6(102) AcClO")
Base (Table XXXIX) 3.42 17.85 -4.16
Primary (Table XL) -2.04 24.00 -9.67
A B
1 1 -0.97 22.86 -8.75
5 C chain (Table
1 2 1.11 18.47 -6.85
XLI)
1 2 1.11 18.47 -6.85
2 2 1.52 19.95 -8.57
Side methyl groups
2 3 1.01 19.69 -7.83
5.16 141.29 -52.68
-1000
Hiooo — H29S.1 =I (5.16 + 141.29(10-=)r . ,52.68(10-«)r2) d r
-'298.1
lO.L
r niooo
= 5.16? + 70.64(10-')7'2 _ (17M){10-^)T'
= 60,890 cal per g-mole
Pitzer' reports a value of 50,550 cal per g-mole.
Extensive comparisons with the data from the literature lead Ander-
sen, Beyer, and Watson to conclude that in general molal heats of forma-
tion and entropies calculated by the "group-contribution method differ
from the better experimental values by less than 4.0 kcal and 2.0 entropy
units, respectively. Calculated heat capacities appear to be within
5 per cent of the accepted values, although serious discrepancies exist
between the experimental values of different investigators. In many
CHAP. XVII] STATISTICAL METHODS 765
cases, particularly for heats of formation, it is believed that the cal-

culated valuQS may be preferable to the experimental.
The contributions in Tables X X X I X to XLIII in some cases are based
on data of questionable accuracy and will be subject to revision as better
experimental values become available. This method of calculation may
be extended to other Series of compounds if data are available to estab-
lish the necessary base-group contribution. In many cases the prop-
erties of such base groups may "be estimated by the theoretical methods
described in the following sections. Properties of higher homologs or
derivatives .may then be estimated from the group contributions. It is
believed that this procedure in many cases is more reliable than direct
application of the simplified forms of statistical calculations to the
complex molecules.
Heats of formation and heat capacities estimated from group con-
tributions are sufficiently accurate for use in energy balances except
where heats of reaction are very small or must be known with unusual
accuracy. Equilibrium constants based on the estimated entropies and
heats of formation are satisfactory for predicting the feasibility of a re-
action but not for accurate calculation of equilibrium compositions.
For accurate equilibrium constants the estimated entropy and heat-
capacity data are best combined, as previously outlined, with a meas-
ured equilibrium constant to establish a consistent heat of reaction.
STATISTICAL METHODS
In recent years development of the principles of statistical mechanics
has lead to methods for calculating precise heat-capacity and absolute
entropy data for the ideal-gaseous state from spectroscopic measure-
ments. Further generalization of the fundamental data has resulted in
means for estimating these properties for many compounds even in the
absence of direct spectroscopic measurements, from a knowledge of the
structure of the molecule. A few of these developments which are of
practical applicability are presented in the foUomng pages, and their
uses are illustrated. Derivation of the relations employed and complete
discussions of the principles may be found in standard texts.^"^ The
5 " Statistical Thermodynamics," J. Mayer and ]M. Mayer, John Wiley & Sons
(1940).
*" Statistical Thermodynamics," R. H. Fowler and E. A. Guggenheim, Cam-
bridge Press (1939).
' " Statistical Mechanics," R. C. Tolman, Oxford Press (1938).
* " Introduction to Quantum Mechanics," L. PauUng and F. B. Wilson, McGraw-
Hill Book Company (1935).
»H. Eyring, J. Walter, and G. E. Kimball, " Quantum Chemistry," John Wiley
& Sons (1944).
development presented represents an amplification of the outline pub-

lished by Kelley'" with an extensive bibliography of current literature.
Molecular Energy of an Ideal Gas. The internal energy of a molecule
is present in kinetic and potential forms resulting from the motion and
attractive forces of the molecule and its parts., These .forms of energy
may be classified in the following five groups:
1. Energy of translation of the molecule as a whole designated as Vt
per mole or et per molecule. This form of energy was discussed in Chap-
ter I I and is equal to (3/2)i?r per mole or {3/2){RT/N) = (3/2)fer per
molecule where N is the Avogadro number and k is termed the Boltz-
mann constant.
2. External rotational energy resulting from rotation of the molecule
as a whole, designated as Vr or e,. Just as in the rotation of large bodies,
such rotational energy may be resolved into components about each of
three perpendicular axes passing through the center of gravity of the
molecule and is a function of the frequency of rotation and the moments
of inertia about these three axes.
3. Internal rotational energy resulting from the rotation of groups of
atoms in the molecule with respect to other groups in the same molecule
and designated as Ur' or e,-. A simple example is the ethane molecule
in which one CH3 group rotates with respect to the other about the
C-C bond. Significant internal rotational energy is believed to result
only from the rotation of groups of atoms, not single atoms, about single-
valence bonds. It is a function of the, moments of inertia of the rotating
groups about the axis of rotation, the frequency of rotation, and any at-
tractive forces between the rotating groups which hinder rotation. Ro-
tation which occurs without such restrictive forces is termed free rotation,
and, if significant attractive forces are encountered, it is termed hindered
rotation. For example, in the ethane molecule as a CH3 group rotates
about the C-C bond, at three points in each revolution its hydrogen
atoms attain minimum distances from the hydrogen atoms of the other
CH3 group. At these three positions the restrictive force hindering ro-
tation is a maximum. The energy required to rotate such a group from
a position of minimum restrictive force to a position of maximum force
is termed the potential barrier restricting rotation and is expressed in
molal energy units. The energy associated with hindered rotation is
less than the energy of free rotation of the same group by an amount
depending on the magnitude of the potential barrier and the number of
times it is encountered per revolution. In complicated molecules ro-
" K . K. Kelley, "Contributions to the»Data of Theoretical Metallurgy; IX The
Entropies of Inorganic Substances," Revision (1940) of Data and Methods of Cal-
culation, U.S. Bur. Mines Bull., 434.
CHAP. XVIl] MOLECULAR DEGREES OF FREEDOM 767
tation of all possible combinations of groups and subgroups must be

considered. It is assumed that rotation does not occur about multiple-
valence bonds and that the energy associated with movements of groups
about such bonds can be classified as vibrational in character.
4. Vibrational energy designated as u„ or £„ results from deformation of
valence bonds. In order to simplify consideration of the many possible
types of vibration about an elastic bond, all vibrational contributions
are resolved into those which result from vibrations along the axis of the
bond, termed valence or stretching vibrations ;', and those which occur
perpendicular to the axis of the boi\d, termed deformation 6 or bending
vibrations. Deformation vibrations change the angle between two ad-
jacent bonds. The corresponding energy contributions are designated
respectively as u^ and us per mole or €„ and es per molecule.
5. Electronic energy designated as Ue or «« results from the motion and
forces of attraction of the electrons and nuclei comprising the molecule.
These energy contributions are relatively very large, and changes in
electronic arrangements are accompanied by large energy changes.
However, except at high temperatures or in the presence of electric
forces or radiation, it may be assumed that the electronic arrangements
are stable and remain in a state of constant electronic energy. On this
basis electronic energy frequently can be neglected in considering the
changes of thermodynamic functions resulting from changes of state.
This assumption may lead to serious errors at high temperatures,
600°C or above. In this range corrections based on spectroscopic
measurements are required.
Molecular Degrees of Freedom. Based upon the foregoing classifi-
cation, the number of independent component forms of internal energy
other than electronic which can be possessed by a molecule in the ideal-
gaseous state is termed its number of degrees of freedom. It may be
shown that a molecule comprising n atoms has 3n degrees of freedom.
For example, a monatomic gas molecule can possess energy only as a
result of translation which can be resolved into three directions; hence,
it has three degrees of freedom. A diatomic molecule such as nitrogen
similarly has three degrees of translational freedom and in addition two
degrees of external rotational freedom and one degree of vibrational
bond-stretching freedom, a total of six. In such a linear molecule the
moment of inertia is negligible about the axis joining the atoms, and ex-
ternal rotational energy results only from rotation about the other two
perpendicular ,axes. There is no energy of internal rotation, because
there is no group having a moment of inertia, and there is no bending
or deformation energy, since there is only one bond. In a nonlinear
triatomic molecule such as water there are six degrees of translational
•
and rotational freedom, two degrees of vibrational bond-stretching free-
dom, and one degree of bend-bending freedom.
The degrees of freedom of a molecule exclusive of its translational
freedom are termed its internal degrees of freedom, which include all free-
doms for external and internal rotation, bond stretching, ^nd bond
bending. From the foregoing discussion it is evident that any molecule
possesses 3n — 3 internal degrees of freedom.
If only the degrees of freedom for internal rotation and vibration are
considered, for a linear molecule,
Ur- + n„ + ns = 3w — 5 (i)
For a nonlinear molecule,
nr' + nv + ns — Sn—Q (5)
where _ ' .
nr' = degrees of freedom for internal rotation
n„ = degrees of freedom for stretching vibration
ns = degrees of freedom for bending vibration
The Quantum Theory. According to the quantum theory variation

of the amount of energy possessed by a molecule in any degree of free-
dom does not occur as a continuous function but takes place in steps
corresponding to the absorption or release of definite increments of en-
ergy termed quanta. This principle of discontinuity corresponds to the
establishment of definite energy levels in each degree of freedom, which
differ from each other by finite increments. The lowest possible level
in each degree of freedom is termed the ground level. In a change of
state from one energy level to another one quantum of energy is ab-
sorbed or released. The quantum does not correspond to a fixed quan-
tity of energy, however, but varies widely in energy content depending
upon the difference of the energy levels involved.
The levels of translational energy are so closely spaced that it may be
considered to vary continuously. However, all rotational, vibrational,
and electronic energies can exist only at definite levels differing from each
other by finite increments. The different energy ,levels in a particular
degree of freedom are identified by quantum numbers which are integers
starting with 1 for the lowest energy level. When a change in energy
level occurs in any degree of freedom, one quantum of energy is emitted
or absorbed as one photon of radiation. The frequency or wave length
of this radiation is related to the energy content of the quantum by the
following equation:
Ae = hv = hc/X = hc« (6)
or AE = 2.857a) cal per g-mole (7)
CHAP. XVII] TRANSLATIONAL CONTRIBUTIONS 769
where
e = energy per quantum, ergs per molecule
h = Planck's constant = (6.624)(IQ-^') erg sec
V = frequency, reciprocal seconds
c = velocity of light = (2.99776) (10i») cm per sec
X = wave length, cm
CO = wave number, 1/X = reciprocal cm
Because of their convenient numerical magnitude, wave nimibers w
are commonly used to designate the frequency or wave length of radia-
tion. Wave numbers are often referred to as frequencies, whereas actu-
ally the wave number has the dimension of reciprocal centimeters. The
distinction between wave numbers w and true frequencies v should be
carefully noted.
From Equation (6) it is evident that changes between widely separated
energy levels emit or absorb radiation of high frequency and wave num-
ber or short wave length. By spectroscopic measurement of the fre-
quency of emitted or absorbed radiation it is possible to evaluate the
various levels of molecular and atomic energy. Thus, the large energy
changes resulting from shifts of electrons correspond to radiation of short
wave length in the ultraviolet or visible regions of the spectrum. The
radiation of medium wave length corresponding to changes in vibra-
tional energy is in the visibleor infrared region, while the frequencies of
rotational energy changes are detected by Raman spectra.
Translational Contributions to Thermodynamic Properties. In the
development of relationships between molecular energies and thermo-
dynamic properties it is convenient to consider separately the contribu-
tions resialting from the various forms of internal energy. It has been
pointed out that the total internal energy of a gas is the sum of the in-
dividual contributions resulting from the five forms of internal energy.
Since entropy is an extensive property, determined by energy content
and its availability, the total entropy can also be expressed as the sum
of the entropy contributions resulting from the five forms of internal
energy. Similar summations represent the total enthalpy, free energy,
and total work function. Derived properties such as the heat capacities
can also be expressed as summations of the contributions due to each
form of energy.
It was pointed out in Chapter II that the pressure and volume rela-
tions of an ideal gas are completely determined by the translational en-
ergy of its molecules. I t was also shown in Chapter XI that, although
the internal energy of an ideal gas is dependent only upon temperature,
its entropy is a function of pressure or volume as well as temperature.
It follows that for an ideal gas the variation of entropy and othfer thermo-
dynamic properties with pressure and .volume result solely from transla-
tional contributions and that contributions from the other four forms of
energy are independent of pressure or volume. For this reason it is con-
venient to develop separately all translational contributions to the
thermodynamic properties.
The translational contribution to internal energy is given by Equa-
tion (II-7), page 31, as
v1 = f i j r • (8)
Combination of Equations (XI 134, 137, 139), page 475, with.Equa-
tion (8) gives
clt = i B (9)
c;, = 1/2 ^ (10)
Hi = fiJT (11)'
These are the properties of a perfect monatomic gas which possesses only
translational energy. As was pointed out in Chapter XI, the absolute
entropy of such a gas cannot be arrived at by simple integration of the
thermodynamic relations because of the indeterminant lower limits en-
countered at the absolute zero. However, the translational, contribu-
tions to entropy have Seen derived by statistical methods and are ex-
pressed by the Sackur-Tetrode equation which has been verified by
experimental measurements on monatomic gases.
s1 = | f i l n M - | - f i 2 h i r - Z i : h i p - 2 . 2 9 8 (12)
or at 298.1''K and 1 atm, ^
si = 6.861 log M + 25.98 ' (iS)
where M is molecular weight, and p is expressed in atmospheres and T
in °K. From Equations (8),' (11), (12), and (XI-4 and 8),
• ^ = E | ^ - - l n M - - l n r + hi'p + 3.664j ^ (14)
= i^F-^ In M - ^ In r-I-hi p + 2.6641 (15)

T
or at 1 atm and 25°C, in calories per gram-mole, '
G1298.16 = - 2045.7 log M - 6266.5 (16)
A1298.16 = - 2045.7 log M - 6858.8 (\7)
niustration 6. Calculate the translational contributions to the heat capacities,
internal energy, enthalpy, entropy, free energy, and total work function of metl^jl
alcohol in the ideal-gaseous state at temperatures of 298.1°K, 500°K, and 1,000°K
and a pressure of 1 atm.
CHAP. XVll] THE PARTITION FUNCTION 771
1. Heat Capacities: From Equations (9) a n d (10)

clt = iR = 2.9798 cal/(g-mole)(°K)
Cpj = | « = 4.9663 cal/(g-mole)(°K)
For a perfect gas the heat capacities c« and Cp are independent of the temperature.
2. Interned Energy: From Equation (8), u j = 2.980T
3. Enthalpy: From Equation (11), B°, = 4.966r
4. Entropy: Since M = 32.043 and p = 1, from Equation (12),
si = IB hi 32.043 + iRhiT - Rlnl - 2.298 = 8.035 + 11.437 log T
5. Free Energy: From Equation (14),
^ = i ? [ - f hi (32.043) - I hi r + hi 1 + 3.664] = - 3 . 0 5 5 - 11.437 log T
6. Total Work Function: From Equation (15),
— = i J C - f In 32.043 - f h i 2 ' + h i l + 2.664] = - 5 . 0 4 1 - 11.437 log T
A?
T til Ht 8°, T T
298.16°K 888.4 1480.7 36.315 -31.349 -33.336
500°K 1489.9 2483.1 38.882 -33.916 -35.903
1000°K 2979.8 4966.3 42.325 -37.359 -39.346
All t h e preceding values are expressed in caI/(g-mole)(°K).
The Partition Function. One mole of an ideal gas comprising N mole-

cules contains individual molecules possessing many varying energy
levels, which result from various combinations of different levels in the
several degrees of freedom other than translation. Of the N molecules,
NNa may have an energy level a, NNb an energy level h, and so on,
where Na, Nh, • • • represent the fractions of the total number of mole-
cules having energy levels o, &, • • •. Then
N = NNa + NNb + NNo+ 2NNi (17a)
where HiNN, represents the summation of the numbers of molecules of
all energy levels.
Since the total internal-energy content of matter is unknown, it is
possible to express an energy level only relative to a selected reference
or ground state. I t is customary to choose the ground state for a par-
ticular molecule as its lowest possible energy level. The energy of any
other level resulting from any mode of motion or degree of freedom is
then designated as e ergs per molecule or E cal per g-mole, referred to this
ground state. The internal energy of the molecule as an ideal gas in
its ground state at 0°K is designated as UQ cal per g-mole. On this basis
the molal internal energy is represented by
U° - uS = TJ? + NaEa + NiEi + NcEc + • • • = U? + XNiEi (18)
u° = the translational contribution to internal energy from Equa-

tion (8)
Na = the fraction of the molecules existing in state a with an en-
ergy level Ea
Useful application of summations such as Equation (18) is restricted
to the ideal-gaseous slâte in order to avoid the complications which are*
introduced if the energy of 1 molecule is influenced by the proximity of
others and cannot be considered individually.
It frequently occurs that two or more different molecular states will
result in substantially the same energy level. In such cases this mul-
tiple contribution to a given energy level b is designated by a factor gt
representing the number of different molecular states which result in the
single energy level 6. This factor is termed the statistical weight of the
energy level or sometimes the quantum weight or degeneracy and is an
integer.
According to the Maxwell-Boltzmann distribution law the fraction of
molecules Nt having any specified energy level EI, is given by the following
expression,
N, = !^.?^ (19)
where Q = go + gae i^'^ + gte R'^ + Qde «r = Sgie RT (20)

i
The derivation of Equation (19) is based on the laws of probability and
may be found in the standard texts on statistical mechanics. The
group e~ ^b/R^ is termed the Boltzmann factor for the particular b energy-
level.
The summation in the denominator of Equation (19) includes the
Boltzmann factors for all possible energy levels, including the ground
level, in all degrees of freedom and is termed the partition function Q.
Since by definition the relative energy E of the ground level is zero, its
Boltzmann factor is unity, and the corresponding contribution to the
partition function is go, the statistical weight of the ground level. It
may .be noted that the form of the Boltzmann factor is such that its
magnitude is greatest for low energy levels. Consequently Equation (19)
denotes that the states of lowest energy level are most probable, but that
the probability of occurrence of high energy levels increases as the tem-
perature is increased. Similarly, the greatest contributions to the sum-
mation of the partition function aê from low energy levels. Molecular
states corresponding to very high energy levels are negligible in the de-
termination of the partition function except at very high temperatures.
CHAP. X V I I ] PROPERTIES FROM PARTITION FUNCTIONS 773
An expression for molal internal energy is obtained by combining

Equations (18) and (19):
l/ - 3 . _A _ i \
v° — vl = v°t + j.\Eagae ^'^+Bigie ^ ^ + M « e *^H )
»,
= ^, + _ _ _ (21)
Also, differentiation of Equation (20) with respect to temperature gives
dT RT'' (22)
Equation (23) permits calculation of internal energy from the variation

of the partition function with temperature.
Thermodynamic Properties from Partition Functions. The contribu-
tions of the thermodynamic properties which result from energy forms
other than translation are readily expressed in terms of their individual
partition functions by combination of Equation (23) with the thermo-
dynamic relations of Chapter XI. Since rotational, vibrational, and
electronic contributions are independent of pressure or volume, these
variables need not be considered in the development. The subscript m
is used to designate contributions resulting from both external and in-
ternal rotations, vibrations, and electronic energy. The relationships
are developed in terms of the combined group of contributions but are
equally applicable to any one type of contribution considered individu-
ally. The total thermodynamic properties are obtained by adding the
m group of contributions to the translational values of the preceding
section.
Equation (23) is directly applicable to any type of contribution to in-
ternal energy. Thus, since the m contributions to internal energy are
independent of pressure, u^ = Hm, and
(u° - H S ) . = (H" - H°o)„ = ^ r ( ^ ) (24)
Because the m contributions are independent of pressure or volume, the

heat-capacity contributions at constant pressure and volume are equal.
Then, since c„ = dv/dT = {l/T)(dv/d In T]
,0 = , . ^RiM + R^^M. (25)
Since entropy is independent of pressure for the contributions under

consideration, from Equation (o) of Table XXIV, page 472
ds,n==Cr,„,dT/T = c^d\nT (26)
T
(s° - sS)m = R\nQ % K ^ « (27)
0 dhiT 0
From Equation (20) defining the partition function it is evident that at

the absolute zero Qo =fifo,which is the statistical weight of the ground
energy level or the number of different molecular states possessing
substantially the ground energy level. Since Qo is finite it follows
that d In Q/d In T is equal to zero where T = 0, and Equation (27)
becomes:
sS. = filnQ + i 2 ^ ^ + ( s ? „ - i 2 1 n g o ) (28)

(t In i V
However, the most general statement of the third law of thermody-
namics derived by statistical mechanics states that the entropy of a sub-
stance at the absolute zero is equal to R times the logarithm of the statistical
weight of the ground energy level, or si = R In go- For substances in the
crystalline state at 0°K the statistical weight of the ground level is gen-
erally unity, leading to the more common statement of the third law.
Equation (28) then becomes
s°^^BlnQ+R^^ (29)
aln r ,
Combining Equations (24) and (29) gives expressions for the m contribu-
tions to G° and A°, thus,
(G°-Hg),„^_^^^^^(A°-^Hg). ^^^^
By adding the m contributions calculated from Equations (24), (25),

(29), and (30) to the translational contributions from Equations (8) to
(17) the thermodynamic functions c°, c°, s°, u° — H°, H° — H?, G° — H°,
and A° — Ho can all be completely evaluated if the partition function is
known as a function of temperature. If the contributing molecular
energy levels and their statistical ^weights are known, Qm may be evalu-
ated by the summation of Equation (20). Similarly, from Equation (22),
3.
dhiT RTQ ^^
CHAP. X V I I ] PROPERTIES FROM PARTITION FUNCTIONS 775
Also, since d^ In Q/dQn T)^= Td{d In Q/d In T)/dT
d"^ In Q îE^gie '^^ îEfgiC

d(ln Ty R^T^Q RTQ \ RQ^ )\dT)
R^T^Q dhiT mi
From Equations (31) and (32) it is seen that all the derivatives of the
partition function which are required for the evaluation of thermody-
namic properties can be calculated from two summations of all significant
energy levels, one 'ZnEiQie ^ ^ and the other "LiElgie «r. For simple mole-
cules the individual energy levels can be evaluated by appUcation of
Equation (6) to spectroscopically determined frequencies of emission or
absorption.
Illustration 7. The development and application of these methods for calculating
thermodynamic properties, is outlined by Kelley^", who developed the following cal-
culation of the thermodynamic properties of nickel vapor in the state of an ideal gas
at 298°K and 1 atm. This is an entirely hypothetical state in which nickel vapor
cannot exist but is legitimately used as a reference from which values for actual
states may be derived by the temperature and pressure relations developed in Chap-
ters XI and XII.
Nickel vapor, Ni (g), is a monatomic gas possessing only translational and elec-
tronic-energy contributions. In Table A the spectroscopic notations for the various
levels of electronic energy are given in column 1 and the corresponding frequencies
in terms of wave numbers and statistical weights, derived from the spectrum, in
columns 2 and 3, respectively. The molal energy levels in column 4 are calculated
from Equation (6) in calories per gram-mole. The three summations required for
calculation of thermodynamic' properties are given in columns 5, 6, and 7. The
electronic contributions to the thermodynamic functions are calculated directly from
these sumlmations.. Thus, from Equations (20), (31), and (32), and Table A,
Q = 11.690
d In 0 1753
= 0.2532
dhiT (1.987) (298.1) (ll.e
d2 In Q 1,533,000
-0.2532 -(0.2532)2 = 0.0565
d(hi TY (1.987)^298.1)^11-69)
TABLE A
1 2 S 4 5 e 7
State Cl) 9i Ei 9,je BT QiEiB RT giE\e~RT
^y^ 0.0 9 0 9.0000 0 0

3n3 204.82 7 585.27 2.6058 1525.1 892,610
3ns 879.82 5 2513.98 0.0717 180.3 453,200
3^3 1332.15 7 3806.62 0.0113 43.0 164,060
3ni 1713.11 3 4895.21 0.0008 3.9 18,500
3F, 2216.65 5 6333.79 0.0001 0.6 4,551
Q = 11.690 1753 1,533,000
776 THERMODYNAMIC PROPERTIES [CHAP XVIi
The complete thermodynamic functions are obtained as follows:

Eq. (8 and 21),
(3)(1.987)(298.16) , 1753 ,.__ , , . »
u° — Ho = >- = 1038 cal per g-mole
2 11.69
Eq. (11 and 21),
H» - HS = C5)(1.987)(298.16) ^ ^^^ ^ ^^^^^^^ ^^^ ^_^^,^
Eq. (9 and 25), '
d = -^^^^^„^ + 1.987(0.2532 + 0.0565)"= 3.595 cal/(g-mole)(°K)
Eq. (10 and 25),
c; == .(^)(1-Q^''') ^ 0.6152 = 5.581 cal/(g-mole)(°K)
Eq. (13 and 29),
s° = 6.861 log 58.68 + 25.98 -1- 1.987(hi 11.69 -f 0.2532)
= 43.49 cal/g-mole (°K)
Eq. (17 and 30),
A" - Ho = -2045.7 log 58.68 - 6858.8 - 1456 = -11,933 cal per g-mole
Eq. (16 and 30),
G° - H^ = -2046.7 log 58.68 - 6266.5 - 1456 = -11,340 cal per g-mole
THERMODYNAMIC DATA FROM MOLECULAR CONSTANTS

The method described in the preceding section requires extensive
spectroscopic data and becomes tedious and complicated for the more
complex molecules although it is rigorous if the significant energy levels
can be identified and evaluated. Considerable simplification with little
loss of accuracy results from approximate methods for calculating ro-
tational and vibrational contributions from constants of molecular struc-
ture. Electronic contributions can be evaluated only by summations
based on spectroscopic data as in Illustration 7. However, for most'of
the complex molecules it may be assumed that all molecules are in the
ground state of electronic energy at all except very high temperatures.
For organic compounds the statistical weight of the ground state is gen-
erally 1, and accordingly electronic contributions to the partition func-
tion may be safely neglected in calculations dealing with complex or-
ganic molecules.
It has been demonstrated that as an approximation which is satisfac-
tory at temperatures above about 50°K the summation represented by
the partition function may be divided into groups representing the con-
tributions of the five different types of internal energy previously dis-
cussed. Thus,
\îgHe'"^)\lîgH&~'^)\îg.ie~'^) (33)
CHAP. XVII] EXTERNAL ROTATION 777
or Q=(Qr)(Qr')(Q.)(Qa)(a) (34)
•where Qr, QH, BH represent, respectively, the partition function, the

statistical weight, and the molecular energy contributions to the ith level
which are due to external rotation, and the other symbols have similar
significance for the contributions of internal rotation, stretching and
bending vibrations, and electronic energy. Thus, the total partition
function is the product of the individual partition functions of the con-
tributing forms of energy. As previously mentioned, electronic energy
levels are relatively high, and at ordinary temperatures the Boltzmann
factors for this form < of energy approach zero and the partition func-
tion Qe approaches unity.
From Equations (23) to (30) it is seen that the thermodynamic prop-
erties are all expressed in terms of the logarithm of the partition function.
I t follows from Equation (34) that each of these thermodynamic prop-
erties may be evaluated as the sum of the individual contributions of
the various forms of energy.
iSm = (S? -f- S°T' -{- SI -\- S^-\- 8% (35)
Each of the individual contributions with the exception of that resulting

from electronic energy may be evaluated from molecular constant data
which permit calculation of the Corresponding partition functions of the
individual forms of energy.
External Rotation. The energy possessed by a molecule as a result of
its rotation as a whole is a function of its moments of inertia and its
speed of rotation. As a result of X-ray and electron diffraction measure-
ments the distances between the atoms and the angles formed by inter-
atomic bonds have been established with sufficient accuracy to permit
quantitative prediction of the spatial arrangement of a molecule. Mo-
ments of inertia are calculated from such molecular dimensions and from
the atomic masses by the methods of classical mechanics. Quantum me-
chanics has shown that molecular rotation is quantized in series of defi-
. nite rotational states whose energy levels differ from each other by finite
increments. For all complex molecules, however, it is found that at
temperatures above 50°K the energy levels become so closely spaced that
they can be treated as a continuous distribution. On this basis it has
been shown that the partition function for the external rotation of a
linear molecule at temperatures above 50°K is represented approximately
by the following expression:
«' = ^ ^ = 2 - 4 8 4 ( 1 0 + ' « ) ^ (36)

mh2
778 THERMODYNAMIC PROPERTIES (CHAP. XVII
or
In Qr = In IT - ' I n a + 88.39 (37)
dlnQr , J dHnQr „ • , ,
The corresponding rotational contributions to thermodynamic function?

of Unear molecules are obtained as follows:
From Equation (25),
(cDr = (cl)r = R (39)
From Equation (24),
iv" - uDr = (H° - HS), = RT (40)
From Equation (29),
s? = fi(ln r + In Z - In <r + 89.39) (41)
From Equation (30),
( ^ } = ( ^ - ^ ) = - ^ a n IT + In a + 88.39) . (42)
where
I = moment of inertia for rotation about the center of gravity in
the plane of the line through the atoms (gr)(cm^)
k = Boltzmann constant = (1.3805)(10-") erg/(molecule)(°K)
h = Planck's constant = (6.624) (IQ-^^ (erg) (sec)
T = degrees Kelvin
<T = the symmetry number
The symmetry number is defined as the number of different positions
into which the molecule can be rotated with identical appearance from
every point of view. Thus, a diatomic molecule composed of like atoms
can be rotated into two positions of identical appearance and therefore
has a symmetry number a = 2. However, if the atoms are dissimilar,
cr=l.
The symmetry number a of complex nonlinear molecules may vary
widely with the configuration of the molecule. Many such molecules
are unsymmetrical and in such cases o- = 1.0. However, for H2O,
(T = 2; for NH3, 0- = 3; for C2H4, <T = 4; for CH4, a = 12; for CaHe,
(T = 6; and for CeHe, a = 12. Symmetry numbers are most easily estab-
lished by inspection of three-dimensional molecular models.
By a similar analysis it has been shown that at temperatures above
50°K the partition function of a nonlinear molecule, having moments of

inertia about three axes, is represented by the following:
a
or
hi Q, = f In T +
+ iili]
In (IJyL) - In (7 + 133.18 (44)
d l n Q , 3 dHnQr
= 0 • (45)
d In r 2 ' d(ln r)2
From these relations the contributions of external rotation to the ther-
modynamic properties of nonlinear molecules are obtained, thus:
{V°-Kl)r=iRT=iB°-Hl)r (46)
c;_. = fi? = c°, -(47)
s° = %R\nT + iR In (IJyL) - i? In o- + 267.54 (48)
Equations (30) and (44): •
(G°-HS). (A°-HS)r = i 2 [ - | l n r - 4 ( l n 7 , 7 / . ) + ln <T- 133.18] (49)
T
It is evident that the rotational contributions to internal energy,
enthalpy, and heat capacity are easily calculated for any compound,
whereas determination of the entropy and free-energy contributions re-
quires evaluation of the product of the principal moments of inertia.
Calculation of the moments of inertia h, ly, and h from data on atomic
masses, angles, and distances is relatively simple for symmetrical mole-
cules whose center of gravity may be located by inspection. A diagram
or model of the molecule is prepared with the center of gravity located
at the intersection of the three perpendicular axes, x, y, and z. The
moment of inertia about each axis is then the summation of mdk for
all the atoms, where m is the weight of the atom and d< its distance from
the axis.
For unsymmetrical molecules the calculation of the moments of in-
ertia is more complicated. The following method was suggested by
Hirschfelder. ^1 A model or diagram is prepared and placed in a con-
venient orientation, generally with one atom at the origin and as many
other atoms as possible lying on the co-ordinate axes. The position of
each atom is then expressed in terms of its Cartesian co-ordinates,
n J. O. Hirschfelder, J. Chem. Phys., 8, 431 (1940),
Xi, yt, Zi, where i designates a particular atom. Then the product of
the three principal moments of inertia is equal to a determinant:
A-D-E
IJyh = -D B-F = ABC - AF^ - CD^ - 2DEF - BE^
-E-F C
m m
B = Simi(x? + z?) - - (S.-miXi)' (S.-miZi)^

m m
(50)
m m
D = S,-miXi2/i (Sim<Xi)(Simi2/i)
m
£' = S.-miXiZi (S,-mia;i)(S,-mi2i)
m
F = îMiyiZi (S£m,-j/.)(S,-w-<z,)
m
ilf = SiWij = mass per molecule
For convenience in calculating it is customary to express interatomic
distances in Angstrom units equal to 10~^ centimeter and to use grams
per mole M as the unit of atomic mass instead of grams per molecule m.
If the moment of inertia calculated in these units is designated as I',
T' T
' 6.023(102»)(10»)2 (6.023)(W) ^ '
and
I'xl'l'z
E In J . = i? In 7; - 181.96 (53)
R In 7 . 7 / . = Z2 In 7;7i7^ - 545.87 (54)
Substituting Equation (53) into (41) and (42) yields:
For linear molecules:
s? = 4.575 log I'T - 4.575 log a - 4.39 (55)
^" ~^'''" = -4.575 log 7 ' r + 4.575 log <j + 6.37 (56)
Substituting Equation (54) into (48) and (49) gives:

For nonlinear molecules:
s° = 2.287 log I'J'yn f 6.861 log T - 4.575 log a - 5.40 (57)
(G°-HS),
• = -6.681 log T- 2.287 log I'J'yK + 4.575 log a + 8.38 (58)
The data necessary for estimating the interatomic distances and bond
angles required for calculating values of / ' are summarized in Table
XLIV for many of the atoms in the more common configuration. In
column I are values of covalent radii recommended by Hirschfelder'^
from the electron diffraction measurements of Pauling^' and others. I t
is assumed that internuclear distances are the sum of the two covalent
bond radii. Thus the internuclear separation for C-H bond is always
0.30A + 0.77A=1.07A.
TABLE XLIV
ATOMIC COVALENT RADII AND BOND ANGLES
Covalent Radius Angstroms Bond Angles between Bonds
I. Hydrogen 0.30 A
II. Carbon
1. Single-bond carbon 0.77 A Regular tetrahedral an-
gles, 108° between bonds
2. Double-bond carbon 0.67 A (for double bonds) Bonds all he in one plane
0.77 A (for single bonds) 124°
112°^'C=
124°
3. Triple-bond carbon 0.60 A (for triple bond) Linear
0.77 A (for single bond)
4. Benzene carbon 0.695 A (for each of the Planar
two C-C bonds) 120°
0.77 A (the bond extend- 120°'>C—
ing outward) '/
120°
III. Oxygen
1. Single-bond oxygen 0.66 A 111
o/
2. Double-bond oxygen 0.57 A =0
IV. Nitrogen
1. Amino nitrogen 0.70 A (flat pyramid with
three bonds mak- 108° 1108°
ing tetrahedral an- / N \
108°
gles witheachother)
2. Nitrate nitrogen 0.65 A (double bond) Planar
0.70 A (single bond) 120
\N^
120^120
^' J. O. Hirschfelder, private communication (1942).
" L . PauUng, Proc. Nat. Acad. Sci. U.S., 18, 293 (1932).
TABLE XLIV —Continued

ATOMIC COVALBNT RADII AND BOND ANGLES
CovalerU Radius Bond Angles between Bonds

3. Isonitrile nitrogen 0.55 A (double bond) Linear
0.70 A (single bond) —N==
4. Cyanide nitrogen 0.65 A =N
V. Sulfur
1. Single-bond sulfur 1.04 A 105°
2. Double-bond sulfur 0.95 A

—s/
=S
3. Sulfate sulfur 0.95 A (double bond) Tetrahedral angles
1.04 A (single bond) lOS'lUOS"
108°^ 1 K.'ios"
VI. Sodium 1.81 A
VII. Chlorine 0.99 A
fill. Bromine 1.14A
IX. Iodine 1.33 A
The angles between the bonds are given in the last column of Table
XLIV. These data permit construction of molecular diagrams or models
from which values of / ' are calculated by application of Equation (50).
Illustration 8. Calculate the rotational contribution to the entropy of NaCl (g)
at 298.1°K and 1.0 atm.
Solution: From Table XLIV the interatomic distance, d = 1.81 + 0.99 = 2.80 A.
The moment of inertia / ' of a diatomic molecule is calculated as follows:
Let xi, Xi = distances from center of gravity of atoms 1 and 2 of
masses mi and mj, respectively
d = distance between atomic centers
d = Xi + X2 = 2.80 (a)
From the principle of moments,
miXi = mzXs (b)
Also r = miXi + mixl (c)
Combination of Equations (a), (b), and (c), and elimination of xi and xa gives
„ mim^' (23) (35.5) (2.80)^ , „ „ .
1 = = = 109.O
(mi + mi) (23 -|- 35.5)
Substitution in Equation (55), since a- = 1, gives
s; = 4.575 (log 109.5 + log 298.1) - 4.39 = 16.25
Illustration 9. Calculate the rotational contribution to the entropy of ethylene,
HjC = CH2 at 25°C and 1.0 atm pressure.
Solution: Reference to Table XLIV shows that the bonds of a double-bond carbon
atom are in one plane; therefore the molecular model of ethylene may be represented
by a two-dimensional diagram with interatomic spacings, as shown in Fig. 164, and
• the center of gravity located midway between the carbon atoms. The co-ordinates
of the atoms are, then,
H': xi = -(1.07 cos 56° + 0.67) = -1.27 A;
j/i = 1.07 sin 56° = + 0.89 A; Zi = Q
w 12 = - 1 . 2 7 A; y% = - 0 . 8 9 A; 22 = 0
B? X3 = +1.27 A; 2/3 = +0.89 A; Z3 = 0
H< i 4 = +1.27 A; 2/1 = - 0 . 8 9 A; 24 = 0
Ci; Xi = - 0 . 6 7 A; 2/6 = 0; Z6 = 0
C: X6 = +0.67 A; ^6 = 0; Z6 = 0
By using these co-ordinates the product of the moments of inertia could be obtained
from Equation (50). Since, the origin is located at the center of gravity, however,
if is simpler merely to sum up the moments of inertia of each of the atoms about
each axis. Thus:
7x' = 4(1.008)(0.89)2 = 3.19
ly = 4(1.008)(1.27)2 + 2(12.00) (0.67)2 = 6.61 + 10.78 = 17.29
/; = 4(1.008) [(0.89)2 + (1.27)2] + 2(12.00) (0.67)^ = 9.70 + 10.78 = 20.48
It may be noted as a general rule that when
aU atoms lie in the plane of the x, y axes y
The symmetry number of the molecule «0.67(

is arrived at by study of Fig. 164. The
molecule as drawn can be rotated about
the X axis into two positions of identical FIG. 164. The Ethylene Molecule.
appearance. Two more positions of iden-
tical appearance can be obtained by the further rotation about the y axis. There-
fore, the symmetry number cr is 4. Substituting in Equation (57) gives
s; = 2.287 log [(3.19) (17.29) (20.48)] + 6.861 log 298.1 - 4.575 log 4 - 5.40
= (2.287) (3.0529) + (6.881) (2.4744) - (4.576) (0.6021) - 5.40
= 15.80 cal/(g-mole)(°K)
From a study of the bsDd spectrum of ethylene Badger'* has been able to determine
the three moments of inertia directly. These results are as follows, compared with
the calculated values of Illustration (9) which are converted to (g)(cm2) by Equa-
tion (51).
Calculated Experimental
Ix(g)(cm2) 5.3 (io-«) 5.7(10-*')
ly (g)(cm2) 28.7 (10-«) 27.5 (10-*i)
7, (g)(cm2) • 34.0(10-") 33.2 (lO"*")
(U/.) 5.17(10-'") 5.20(10-"')
The agreement in the product of the moments of inertia is satisfactory and indicates
that the methods used may be expected to yield reliable results for more complex
molecules.
Illustration 10. Calculate the external rotational contribution to the heat capac-
ity, enthalpy, entropy, and free energy of methyl alcohol at 1.0 atm and 298.1, 500,
and 1000°K, for the ideal gaseous state.
" R . M. Badger, Phys. Rev., 45, 648 (1934).
784 THERMODYNAMIC PROPERTIES [CHAP. X V I I
Solution: The first step in dealing with an unsymmetrical molecule is the construe- •
tion of a diagram or model to serve as a guide in determining the co-ordinates of the
atoms. A diagram for methyl alcohol is shown in Fig. 165 in three projected views
using values from Table XLIV. For convenience the carbon atom is placed at the
origin and the molecule and its parts so oriented that as many atoms as possible lie
in the xy plane. The molecule could equally well be drawn with other relative posi-
tions of orientation of the oxygen with respect to the carbon atom, but the product
of the three moments of inertia is neghgibly affected by such changes. The inter-
atomic distances from Table XLIV are C-H = 1.07; C-0 = 1.43, ând O-H = 0.96.
From the geometry of the diagram the co-ordinates of the atoms are calculated as
follows:
ff: x = -1.07sinl8° i / = - 1 . 0 7 sin 72° 41 = 0
= -0.331; =-1.018;
H': X = -0.331; y = 1.018 sin 30° z = -1.018 sin 60°
= 0.509; = -0.88
H': X = -0.331; y = 0.509; g = 0.88
H*: 3; = 1.43 4-0.96 sin 21° ?/ = 0.96 sin 69° z= 0
= 1.77; = 0.894;
0: 2 = 1.43; y = 0; 2=0
These co-ordinates and the corresponding atomic weights M are substituted in Equa-
tion (50):
Summations M{y^ + z^) My Mz Mx
IP 1.0446 -1.0261 0 -0.3336
ff 1.0453 -f0.5131 -0.8890 -0.3336
IP 1.0453 +0.6131 -1-0.8890 -0.3336
H« 0.8092 +0.9032 0 +1.7882
O 00 0.0 0 +22.8800
Sum 3.9444 0.9033 0 23.6674
M(xy) M(x^ + y^) Mxz M(x^ + z^) Myz
W +0.3397 1.1551 'o 0.1105 0
H» -0.1698 0.3716 +0.2943 0.8946 -0.4515
H» -0.1698 0.3716 -0.2943 0.8946 +0.4515
H* +1.6022 3.9815 0 3.1723 0
O 0 32.7184 _0 32.7184 _0
Sum 1.6023 38.5982 0 - 37.7904 0
From Equation (50):
A = 3.9444 ?— (0.9033)2 = 3.9189
32.042
B = 37.7904 i— (23.6674)2 = 20.3088
32.042
C = 38.5982 - — i ^ (23.6674)2 _ _ ! _ (0.9033)'' = 21.0911
D = 1.6023 ^— (23.6674) (0.9033) = 0.9351

32.042
S=0-0=0
^=0-0=0
7^/;/; = ABC - CD' = 1660
Substitution in Equation (57), since a = 1.0, gives
T = 298.16°K s; = 1.96 + 16.98 = 18.94
T =-- 60D°K s; = 1.96 + 18.52 = 20.48
T = 1000°K s; = 1.96 + 20.59 = 22.55
The other thermodynamic'properties are similarly calculated from Equations (46),
(47), and (58), and the results are summarized in Table LI, page 802.
Complex molecules in general are most conveniently treated by pre-
paring an accurate large-scale diagram similar, to Fig. 165, using pro-
FiG. 165. The Methyl Alcohol Molecule.

jected angles and dimensions. The co-ordinates of the individual atoms
are then determined by direct measurement. The projected angles
necessary for such graphical construction are shown for the methyl
alcohol molecule in Fig. 165. It may be noted that the projected angles
of the carbon atom may also be used for sulfate sulfur while the angles
of the group CH'H^H' of Fig. 165 are the same as those of the amino
nitrogen atom.
Many different diagrams are possible for a complex molecule, depend-
ing on the positigns into which the various atoms are rotated. As is
discussed in the next section, these different possibilities are represented
in the actual molecules as a result of either continuous or periodic rota-
tion of the atoms and groups of atoms about the bonds. However, the
resulting changes in configuration have only small effects on the product
of the three principal moments of inertia of .the molecule as a whole.

Satisfactory results are obtained by assViming positions of rotation cor-
responding as closely as possible to a linear structure where straight
chains are involved and approaching a single plane for chain structures.
Thus, if Fig. 165 were extended to represent ethyl alcohol, hydrogen
atom H^ would be replaced with a carbon rotated with one bond pointing
upward in the xy plane. Normal propyl alcohol would be represented
by attaching another carbon to this bond, so .oriented that one- of its
bonds would lie along the x axis. Normal butyl would result from adding
a carbon atom to this bond with one of its bonds extending upward in
the xy plane to which another carbon atom would be added to form nor-
mal amyl alcohol. This construction would result in all carbon atoms
lying in the xy plane, both above and below the x axis. Such a configura-
tion is intermediate between the extreme deviations resulting from in-
ternal rotation and may be taken as satisfactorily representing the aver-
age product of the moments of inertia.
Internal Rotation. In the preceding section the rotation of the mole-
cule as a whole about its center of gravity was considered, treated as a
rigid structure of an average configuration. As previously discussed, it
is recognized that in addition to the energy of this external rotation there
are also energies of rotation of groups of atoms about the bonds joining
them, each contributing to the partition function of the molecule.
In the early treatments of this problem it was recognized that a double
or triple bond exerts definite resistance to rotation which may reduce the
motion to oscillation with only periodic rotation when the energy level
of the restrictive barrier is exceeded. However, it was assumed that
completely free rotation could exist about single bonds and energy and
entrppy contributions were calculated on this basis. The results ob-
tained were too high to agree with experimental data and Pitzer and
coworkers'^^' *^ extended these methods to allow for a potential barrier
restricting every internal rotation.
The partition function Q/, resultmg from the free rotation of a group
with respect to another group or with respect to the molecule as a whole,
is given by the following equation:
^' hivT" ^^^^

or, if /red is cxpressed in (gram) (cm) ^ and T in degrees Kelvin,
_ 2.7934[(/redr)(10^^)]*
Q, = Jl (60)
« K . S. Htzer, J. Chem. Phys., 5, 469, 473, 752 (1937).
i«K. S. Pitzer and W. D. Gwinn, / . Chem. Phys., 10, 428 (1942).
€HAP. XVII] INTERNAL ROTATION 787
where
/red = the reduced moment of inertia of the internal rotation
Nh = the symmetry number of the internal rotation, the number of
points of maximum attraction existing per revolution between
the atoms of the rotating groups.
For example, in the ethane molecule at three positions in each revolu-
tion the hydrogen atoms of one group reach points of minimum separa-
tion from those of the other. Hencef for this rotation Nh = 3.
. The reduced moment of inertia of one group about another is given
by the following expression if the axis of the internal rotation coincides
•with one of the principal axes of the molecule as a whole.
lied IA IB
where I A, IB are the moments of inertia of the A and B groups, respec-,
tively, about the bond joining them. Equation (61) may be applied as
an approximation where two or more internal rotations' are involved if
the axis of rotation, approximately coincides with a principal axis of the
molecule. A more exact relationship which is applicable to the rotation
of several groups with respect to a rigid molecular framework is proposed
by Pitzer and Gwinn.^"
The contributions of free internal rotation to the thermodynamic prop-
erties are evaluated by combining Equation (60) with Equations (24),
(25), (29), and (30), thus:
(cDf = ict)f = iR (62)
(u° - HS)/ = (H° - Hl)f = iRT (63)
Sf = iR + RlnQ/ (64)
^5l^--«tae, (66)
The contributions of hindered internal rotation to the thermodynamic
properties have been evaluated by Pitzer and Gwinn" and are presented
in Tables XLV to XLVII as functions of the partition function of free
rotation Q/ and the potential barrier Vh cal per g-mole, which hinders
the rotation Vh times per revolution. In Table XLVI the entropy con-
tributions of hindered rotation s°' are tabulated directly for values of
1/Qf greater than 0.30. For lower values of 1/Qf the table gives the
value of (s/ —s°'). The corresponding values of s?- are calculated
from Equation (64). Thus:
s?' = s ^ - ( s ? - s°') =iR + RhiQf-{s}-s?') (66)
TABLE XLV
±IN
0.0 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80

VH/RT
0.0 0.9934 0.993 0.993 0.993 0.993 0.993
0.2 1.1822 1.106 ;.050 1.022 1.008 1.000
0.4 1.3513 1.249 1.151 1.073 1.036 1.015
0.6 1.5011 1.374 1.251 1.138 1.072 1.030 ^
0.8 1.6324 1.482 1.340 1.211 1.114 1.048
1.0 1.7460 1.576 1.418 1.275 1.155 1.065
1.5 1.9607 1.753 1.561 1.385 1.230 1.103
2.0 2.0934 1.854 1.636 1.440 1.265 1.120
2.5 2.1657 1.900 1.662 1.448 1.260 1.104
3.0 2.1971 1.909 1.651 1.426 1.224 1.060 0.92
3.5 2.2030 1.893 1.621 1.382 1.176 1.006 0.88
4.0 2.1944 1.864 1.577 1.329 1.121 0.947 0.82
4.5 2.1788 1.829 1.529 1.273 1.061 0.884 0.73
5.0 2.1607 1.794 1.481 1.218 1.002 0.824 0.67 0.59
6.0 2.1261 1.727 1.392 1.115 0.893 0.714 0.56 0.47
7.0 2.0984 1.670 1.315 1.029 0.802 0.624 0.482 0.38 0.31
8.0 2.0781 1.623 1.251 0.955 0.725 0.549 0.418 0.31 0.25
9.0 2.0634 1.583 1.196 0.892 0.661 0.488 0.363 0.269 0.20
10.0 2.0526 1.548 1.147 0.838 0.608 0.437 0.319 0.231 0.17
12.0 2.0382 1.492 1.067 0.745 0.519 0.356 0.244 0.169 0.116
14.0 2.0292 1.441 0.997 0.672 0.450 0.295 0.195 0.128 0.084
16.0 2.0229 1.401 0.937 0.613 0.394 0.249 0.157 0.099 0.063
18.0 2.0182 1.363 0.886 0.561 0.347 0.211 0.128 0.077 0.047
20.0 2.0147 1.329 0.841 0.516 0.307 0.181 0.105 0.061 0.036
Values of (G° — nl)r'/T are readily calculated from the entropy and
enthalpy contributions taken from the tables. Thus:
(«°-HSV (H°-Hg).>
y = y Sr' (67)
The potential barriers VK which hinder internal rotation are evaluated
by comparison of calculated with experimentally determined entropies
or free energies from equilibirum or third-law measurements. For this
reason all errors in any of the calculated contributions and in the experi-
mental values result in corresponding errors in the apparent potential
barrier, and the reported values for given combinations of groups vary
widely. No satisfactory method of generalizing potential barriers has
been developed, but approximations can be arrived at from the values
of Table XLVIII taken from a summary prepared by Aston." The po-
tential barrier is largely determined by the atoms directly attached to'
each of the atoms joined by the bond under consideration.
" J . G. Aston, Ind. Eng. Chem., 34, 514 (1942).
CHAP. XVII] INTERNAL ROTATION 789
TABLE XLVI
INTERNAL ROTATIONAL CONTRIBUTIONS TO ENTROPY, cal/(°K)(g-mole)
^ 1• - - C s / - 8°,')
Of!}
y ^ 9
0]
0.0 0.10 >-
1/Q/ 0.20 0.30 0,30 0.40 0.50 0.60 0.70
0.80
I
VH/RJ
0.0 0.0000 0.000 0.000 0.000 3.386 2.814 2.371
0.2 0.0049 0.004 0.004 0.004 3.382 2.811 2.369
0,4 0.0198 0.018 0.018 0.016 3.370 2.801 2.359
0.6 0.0440 0.043 0.040 •0.039 3.347 2.780 2.340
0.8 0.0771 0.077 0.072 0.068 3.318 2.750 2.315
1.0 0.1185 0.117 0.112 0.107 3.279 2.714 2.279
1.5 0.2527 0.250 0.242 0.230 3.156 2.600 2.173
2.0 0.4182 0.415 0.402 0.382 3.004 2.458 2.048
2.5 0.6001 0.594 0.577 0.550 2.836 2.303 1.907
3.0 0.7856 0.777 0.757 0.719 2.667 2.138 1.756 1.45
3.5 0.9660 0.957 0.929 0.886 2.500 1.978 1.610 1.34
4.0 1.1356 1.126 1.094 1.043 2.343 1.834 1.475 1.22
4.5 1.2918 1.280 1.244 1.187 2.199 1.698 1.348 1.10
5.0 1.4339 1.421 1.380 1.318 2.068 1.579 1.233 0.97 0.81
6.0 1.6781 1.662 1.616 1.542 1.844 1.370 1.040 0.79 0.63
7.0 1.8783 1.860 1.807 1.721 1.665 1.204 0.891 0.665 0.50 0.39
8.0 2.0447 2.024 1.962 1.867 1.519 1.071 0.770 0.664 0.41 0.31
9.0 2.1864 2.163 2.095 1.989 1.397'»0.962 0.674 0.482 0.348 0.25
10.0 2.3095 2.284 2.208 2.091 1.295 0.872 0.596 0.418 0,295 0.21
12.0 2.5155 2.485 2.394 2.261 1.125 0.728 0.476. 0.315 0,212 0.143
14.0 2.6847 2.650 2.547 2.392 • 0.994 0.620 0.388 0.247 0,158 0.102
16.0 2.8289 2.788 2.674 2,496 0.890 0.533 0.322 0.196 0,120 0.075
18.0 2.9545 2.910 2.781 2,585 0.801 0.464 0.270 0.158 0,093 0.056
20.0 3.0659 3.017 2.872 2.659 0.727 0.405 0.228 0.129 0.073 0.042
In general the barriers are greater for large atoms. Approximation for
specific groupings can be estimated from analogy to those of Table
XLVIII.
Where long chains are involved, as in the higher normal paraffins and
their derivatives, the rotational contributions of the intermediate groups
are difficult to evaluate, and the entire problem is complicated by the
effect of internal rotation on the external-rotational contributions. Im-
proved methods for handhng this problem have been proposed by Craw-
ford'* and Pitzer,'^ but in the present state of development these schemes
require a high degree of judgment or are too tedious for general applica-
tion. For the present the most practical procedure for developing data
for a new compound is to work out in detail the lowest, or better the next
to the lowest, homolog with the best possible assumptions as to potential
18 B. L. Crawford, / . Chem. Phys., 8, 273 (1940).
1' K. S. Pitzer, Chem. Rev., 27, 39 (1940).
TABLE XLVII
INTEENAL ROTATIONAL CONTRIBUTIONS TO HEAT CAPACITY, cal/(°K) (g-mole)
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
Vh/til
0.0 0.9934 0.993 0.993 0.993 0.993 0.99
0.2 1.0033 1.003 1.001 0.999 0.998 1.00
0.4 1.0326 1.032 1.028 1.024 1.019 1.02
0.6 1.0799 1.079 1.073 1.065 1.056 1.05
0.8 1.1433 1,141 1.133 1.121 1.106 1.09
1.0 1.2201 1.217 1.206 1.190 1.169 1.14 ,
1.5 1.4506 1.444 1.423 1.391 1.348 1.30
2.0 1.6975 1.687 1.655 1.606 1.541 1.469
2.6 1.9211 1.908 1.866 1.801 1.717 1.623
3.0 2.0986 2.082 2.033 1.952 1.846 1.738 1.7
3.5 2.2223 2.204 2.146 2.054 1.934 1.803 1.7
4.0 2.2986 2.276 2.213 2.110 1.980 1.834 1.69
4.5 2.3354 2.312 2.238 2.129 1.990 1.832 1.67
5.0 2.3443 2.318 • 2.241 2.120 1.972 1.808 1.62 1.4
6.0 2.3155 2.283 2.192 2.059 1.893 1.711 1.51 1.33
7.0 2.2647 2.228 2.126 1.973 1.787 1.588 1.394 1.21 1.0
8.0 2.2157 2.174 2.058 1.888 1.684 1.468 1.260 1.07 0.91
9.0 2.1759 2.130 1.999 1.808 1.587 1.362 1.149 0.955 0.79
10.0 2.1454 . 2.094 1.951 1.745 1.507 1.262 1.047 0.853 0.68
12.0 2.1050 2.043 1.877 1.636 1.365 1.107 0.877 0.683 0.519
14.0 2.0810 2.009 1.814 1.546 1.254 0.978 0.744 0.555 0.408
16.0 2.0654 1.983 1.764 1.468 1.156 0.873 0.639 0.457 0.321
18.0 2.0544 1.961 1.717 1.397 1.070 0.780 0.549 0.378 0.256
20.0 2.0462 1.944 1.678 1.333 0.991 0.701 0.477 0.316 0.207
TABLE XLVIII
POTENTIAL, B A R R I E R S H I N D E R I N G INTERNAL, ROTATION
Group A Group B F A I eal per g-ioaole

CHs —CHs 3150
CH3 CH2CII3 3300
CH3 —CH(CH3)ii 3870 ,
CHa —C(CH3)3 4540
CH, — C = C CH3 0 ••. •
CH3 —CH2OH 3000

GHs —C0(CH3) 1000
CH3 —NH(CH3) 3460
CH3 —0—CH3 3100
OH -CH3 2700»
OH CH2CH3 10,000
OH --GH(CH3)2 5000
^^ Revised in accordance with spectroscopic data of Borden and Barker, / . Chem.
Phys., 6, 553 (1938), and the third law value of s^gg = 56.63 calculated by L. 8 . '
Kassel, J. Chem. Phys., 4, 493 (1936), from the low-temperature measurements of
K. K. Kelley, J. Am. Chem. Soc, 51, 181 (1929).
CHAP. XVII! VIBRATIONAL CONTRIBUTIONS 791
barriers and then extend these results to higher homologs by means of

the group contributions of Tables X X X I X to XLIII.
Illustration H. Calculate the contributions of internal rotation to the heat capac-
ity, enthalpy, entropy, and free energy of methyl alcohol in the ideal-gaseous state at
1 atm and 298.1, 500, and 1000°K.
Solution:
n. _ f8.Î^T\i 2.794iIT(m)'
—"--'''" -' VhHA^*)V NH
Equation (61), - = - + -
^ ^methyl ÔH
\
Equation (51) and Table XLIV,
/.et..l = '^'-°°'^^'-''^"''°^^=(5.203)(10-)
methyl 6.023(10'")
1(1.008)(0.96 sin 6 9 T _ 1 3 ^ ( 1 0 ^ ^
"" 6.023 (10=»)
7 = (1.068) (10-*"); NK = 3
10.39
Qf = (0.09625)T^ i - =
From Table XLVIII,
Vh = 2700 cal per mole
w mi
MQt n/Kr
298.16 17.267 0.602 4.56
500 22.361 0.465 2.72
1000 31.623 0.329 1.36
Interpolation in Tables XLV to XLVII and with Equation (67) gives
(Hy — Ho)r» (G°j, - Ho),/
T T e;, cr. T (H°3,-H;),,

298.16 0.74 1.08 1.67 0.34 221
500 1.14 1.97 1.72 0.83 570
1000 1.31 3.02 1.32 1.71 1310
Internal rotation about a multiple bond such as exists in ethylene or

acetylene is so highly restrained that it is generally treated as an oscilla-
tion or vibration by the method of the following section.
Vibrational Contributions. As previously discussed, molecular vibra-
tional energies result from movements of the atoms which stretch in-
dividual bonds, change the angles between adjacent bonds, or result in
the twisting of multiple bonds. Each such mode of motion corresponds
to a single vibrational degree of freedom. The number of such vibra-
tional degrees of freedom possessed by a molecule may be calculated from
Equation (4) or (5).
If complete spectroscopic data are available the vibrational contribu-
tions to the partition function might be evaluated through summation

of the Boltzmann factors of all energy levels in all vibrational degrees of
freedom by the procedure demonstrated in Illustration 7. A more con-
venient method results from treatment of each vibrational degree of
freedom as a harmonic oscillator having a fundamental frequency v' cor-
responding to a wave number w'. The partition function for such a
single harmonic vibration has been derived from the principles of quan-
tum mechanics.
1 1 1_
Q"= E7 = i : i ^ = T—r; (68)
1 — 6"="
where
v' and w' = the fundamental frequency and wave number, respec-
tively, of the oscillator
T = degrees Kelvin
a; = 1.4384 w ' / r
Equation (68) may be differentiated to obtain d In Q„/d In T and
d^ In Qu,/d (In TY for evaluation of the thermodynamic properties from
Equations (24), (25), (29), and (30). Thus, smce
' d\RQ/dhiT =TdQ/QdT
and X = 1.4384w7r,
d.ln Q„ X
(69)
d^ In Q„ a; V
(70)
d(lnT)2 ( e ^ ~ l ) 2 e^-1
The total vibrational contributions to the thermodynamic properties are
obtained as summations of the contributions of all individual degrees
of vibrational freedom, each oi which corresponds to particular vklues
of / , w', and x. Thus
(H°-HS)„ = i 2 T S - ^ (7*1)
e^— I
( c ; ) . = (c°). = 5 ! S - ^ ^ (72)
(e^— 1)''
• hi(l - e--) (73)
(^)r flSbi(l-e-) (74)
I t is evident that each of the contributions may be evaluated from a

knowledge of the fundamental frequencies of all vibrational degrees of
CHAP. XVII] VIBRATIONAL CONTRIBUTIONS 793
freedom. These fundamental frequencies and the corresponding statis-

tical weights are determined from spectroscopic data or may be esti-
mated from generalized bond assignments, as discussed in the following
section. To facilitate solution of Equations (71-74) extensive tables
have been prepared relating the various so-called Einstein fimctions to x.
In Table XLIX are values of the vibrational contributions to heat ca-
pacity, enthalpy, and free energy expressed as functions of JjT. This
table originally developed by H. L. Johnston^' and revised by Schu-
mann and Schwartz, 22 is more convenient to use than tables of the
Einstein functions but has the disadvantage of being dependent on the
values of the physical'constants. Contributions to the entropy are ob-
tained from these tables by means of the equation:
Illustration 12. From spectroscopic data the fundamental vibrational frequencies

of methyl alcohol have been evaluated by Borden and Barker^" as follows:
WaveNo.jia Statistical Wgt., g 298.16 500 1000

1 3683 1 12.352 7.366 3.683
2 2978 2 9.988 5.956 2.978
3 2845 1 9.542 5.690 2.845
4 1477 2 4.954 2.954 1.477
5 1455 1 4.880 2.910 1.455
6 1340 • 2 4.494 2.680 1.340
7 1034 1 3.469 2,068 1.034
8 1030 1 3.455 2.060 1.030
Calculate the contributions of vibration to the heat capacity, enthalpy, entropy, and
free energy in the ideal-gaseous state at 1 atm and 298.16, 500, and 1000°K.
The contribution corresponding to each vibrational degree of freedoi\j is read from
Table XLIX and multiplied by the corresponding statistical weights. Thus, at
298.16°K,
1
(^).
0.00000
-(^l — —
{c°)»
2 0.00004 — —
3 0.00003 — —
4 0.02342 0.00330 0.166
5 0.01282 0.00183 ' 0.090
6 0.04120 0.00636 0.264
7 0.06920 0.01386 0.346
8 0.07030 0.01413 0.350
S 0.21701 ' 0.03948 1.216
^ E. B. Wilson, Jr., Chem. Rev., 27, 17 (1940).
*" S. C. Schumann and M. L. Schwartz; Taylor and Glasstone, " Treatise on
Physical Chemistry," Vol. I, D. Van Nostrand Co., New York (1942) with permission.
TABLE i L I X
VIBRATIONAL CONTRIBUTIONS TO THERMODYNAMIC PROPERTIES I
cal/(°K)(g-mo]e)
(H°-HS (G°-H;) (H°-HD (O°-H;)

(J/T) c W/T) c
T T •T
> ^
0.10 1.9834 1.8473 3.9942 0.45 • 1.9188 1.4128 1.4726

0.11 1.9827 1.8338 . 3.8186 0.46 1.9159 1.4016 1.4416
0.12 1.9819 1.8203 3.6596 • 0.47 1.9128 1.3904 1.4117
0.13 1.9810 1.8068 3.5143 0.48 1.9097 1.3793 1.3826
0.14 1.9801 1.7935 3.3807 0.49 1.9066 1.3682 1.3543
0.15 1.9791 1.7801 3.2572 0.50 1.9033 1.3573 1.3267

0.16 1.9780 1.7670 3.1428 0.51 1.9001 1.3464 1.3001
0.17 1.9768 1.7538 3.0361 0.52 1.8966 1.3354 1.2741
0.18 1.9756 1.7407 2.9362 0.53 1.8933 1.3248 1.2488
0.19 1.9744 1.7276 2.8424 0.54 1.8898 1.3140 1.2242
0.20 1.9731 1.7147 2.7542 0.55 1.8863 1.3034 1.2001

0.21 1.9717 1.7018 2.6710 0.56 1.8827 1.2929 1.1766
0.22 1.9703 1.6890 2.5920 0.57 1.8791 1.2823 1.1539
0.23 1.9687 1.6763 2.5142 0.58 1.8754 1.2719 1.1316
0.24 1.9671 1.6636 2.4662 0.59 1.8718 1.2616 1.1099
0.25 1.9655 1.6510 2.3786 0.60 1.8680 1.2513 1.0887

0.26 1.9638 1.6384 2.3140 0.61 1.8641 1.2411 1.0681
0.27 1.9620 1.6260 2.2525 0.62 1.8602 1.2308 1.0481
0.28 1.9601 1.6135 2.1935 0.63 1.8562 1.2207 1.0284
0.29 1.9583 1.6012 2.1372 0.64 . 1.8522 1.2107 1.0093
0.30 1.9563 1.5889 2.0831 0.65 1.8481 1.2007 0.9905

0.31 1.9542 1.5767 2.0311 0.66 1.8441 1.1908 , 0.9722
0.32 1.9521 1.5645 1.9813 0;67 1.8398 1.1810 0.9545
0.33 1.9500 1.5525 1.9333 0.68 1.8357 1.1711 0.9370
0.34 1.9478 1.5404 1.8872 0.69 1.83i5 1.1614 0.920*
0.35 1.9454 1.5285 1.8427 0.70 1.8272 1.1517 0.9033

0.36 1.9430 1.5167 1.7998 0.71 1.8227 1.1420 0.8870
0.37 1.9406 1.5048 1.7584 0.72 1.8184 1.1325 0.8713
0.38 1.9381 1.4931 1.7185 0.73 1.8140 1.1230 0.8557
0.39 1.9356 1.4814 1.6799 0.74 1.8093 1.1135 0.8405
0.40 1.9330 1.4698 1.6425 0.75 1.8047 1.1041 0.8256

0.41 1.9303 1.4583 1.6063 0.76 1.8002 1.0948 0.8110
0.42 1.9275 1.4468 1.5713 0.77 1.7955 1.0856 0.7968
0.43 1.9247 1.4354 1.5374 0.78 1.7908 1.0764 0.7828
0.44 1.9218 1.4240 1.5045 0.79 1.7863 1.0674 0.7692
CHAP. X V I I ] VIBRATIONAL CONTRIBUTIONS 795
TABLE XLIX (Continued)

VIBRATIONAL CONTRIBUTIONS TO THERMODYNAMIC PROPERTIES II
(H°-Ho°) (G°-H;) (H°-H„°) (O°-HS)

(co'/r) c (f^'/T) c
T T T T
0.80 1.7813 1.0582 0.7558 1.20 1.5592 0.7427 0.3894

0.81 1.7765 1.0492 0.7437 1.21 1.5531 0.7359 0.3832
0.82 1.7716 1.0403 0.7299 1.22 1.5468 0.7292 0.3773
0.83 1.7667 1.0314 0.7174 1.23 1.5406 0.7225 0.3713
0.84 1.7618 1.0226 0.7052 1.24 1.5343 0.7159 0.3656
0.85 1.7568 1.0138 0.6931 1.25 1.5281 0.7093 0.3598
0.86 1.7517 1.0050 0.6812 1.26 1.6217 0.7027 0.3542
0.87 1.7466 0.9964 0.6697 1^27 1.6156 0.6963 0.3486
0.88 1.7415 0.9878 0.6583 1.28 1.5091 0.6898 0.3432
0.89 1.7364 0.9793 0.6473 1.29 1.5030 0.6834 0.3379
0.90 1.7312 0.9708 0.6363 1.30 1.4966 0.6771 0.3326
0.91 1.7259 0.9620 0.6256 1.31 1.4902 0.6708 0.3275
0.92 1.7206 0.9541 0.6151 1.32 1.4839 0.6646 0.3225
0.93 1.7154 0.9458 0.6049 1.33 1.4774 0.6585 0.3175
0.94 1.7099 0.9377 0.5948 1.34 1.4710 0.6523 0.3126
0.95 1.7045 0.9295 0.6848 1.35 1.4646 0.6462 0.3078
0.96 1.6991 0.9212 0.6751 1.36 1.4583 0.6402 0.3030
0.97 1.6935 0.9132 0.5657 1.37 1.4520 0.6342 0.2983
0.98 1.6882 0.9051 0.5563 1.38 1.4455 0.6283 0.2937
0.99 1.6826 0.8971 0.5471 1.39 1.4390 0.6223 0.2893
1.00 1.6770 0.8893 0.5382 1.40 1.4324 0.6164 0.2848
1.01 1.6714 0.8815 0.5294 1.41 1.4259 0.6107 0.2804
1.02 1.6668 0.8737 0.5208 1.42 • 1.4194 0.6049 0.2761
1.03 1.6601 0.8659 0.6123 1.43 1.4129 0.5998 0.2719
1.04 1.6544 0.8582 0.5040 1.44 1.4065 0.5935 0.2677
»
1.05 1.6487 0.8507 0.4959 1.45 1.3999 0.5879 0.2636
1.06 1.6429 0.8430 0.4877 1.46 1.3934 0.5823 0.2595
1.07 1.6371 0.8355 0.4800 1.47 1.3868 0.5768 0.2556
1.08 l.iB312 0.8281 0.4722 1.48 1.3802 •0.5713 0.2518
1.09 1.6254 0.8207 0.4647 1.49 1.3736 0.6658 0.2471
1.10 1.6196 0.8133 0.4572 1.50 1.3671 0.5604 0.2441
1.11 1.6137 0.8060 0.4499 1.51 1.3605 0.5550 0.2404
1.12 1.6077 0.7988 0.4426 1.52 1.3538 0.5493 0.2367
1.13 1.6017 0.7916_ 0.4357 1.53 1.3473 0.5445 0.2332
1.14 1.5957 0.7845 0.4287 1.54 1.3406 0.5392 0.2297
1.15 1.5897 0.7774 0.4218 1.55 1.3340 " 0.5340 0.2262
1.16 1.5837 0.7703 0.4151 1.56 1.3274 0.5288 0.2227
1.17 1.5776 0.7634 0.4086 1.57 1.3208 0.5238 0.2193
1.18 1.5715 0.7564 0.4021 1.58 1.3143 0.5187 0.2161
1.19 1.5653 0.7495 0.3957 1.59 1.3077 0.5138 0.2129
THERMODYNAMIC PROPERTIES [CHAP. XVII
796
VIBRATIONAL CONTRIBUTIONS TO THERMODYNAMIC PROPERTIES III
(Q°-H;)
(H°-Ho°) _ (g°-Ho°) („,/y)
wm
1.30.0 o.5c^ o.»r js ; s s o.;s
l-S "o-S S IS S S
1.2943
1.2878
0.5039
0.4988
.:S5
! S St IS ;S S
1.2812
12746
0.4942
0.4893
°o:S
5s E E JE E E =
IS s i S -s = s s:=
;.^« 0.«,3 0.IS02 J» a»«; 0573 0035^6
\T-, °„1S 0S S oS 0.2659 0.08084
\fS, oafs 0 TS 2.26 0.8799 0.2605 0.07850
S 0:«35 0:1093 2.28 0.8682 0.2551 0.07622
«.nni niR79 2^0 0 8570 0.2497 0.07402
;r,l VZ 0S Sz 0;8455 0.2145 0.07189
;-S S'S aS S °0:^S S.2295 0.«535

nAi77 01551 2.40 0.8000 0.2246 0.06395
1.1625 0.4177
0:4135 u.iooi
0.1528 ^2.42
-1" . _„„,
0.7885 . „n,„o nriROif!
^•^^^^ °-^°?^^ S:S 2.46 S S S 0:05859

1-;S S:S VZ tZ = 0:20. 0.05690
i:i363 0.3972 0.1440 2.50 0.7453 0.2017 0.05526
1-1298 0.3932 0.418 2.52 0.7346 0 197^ ^^^^,^^
M S 0 5 0 77 2: 6 S7135 0.1891 0.050^5

;-;;o2 ssfe 0:1356 2.58 0.7031 0.1850 0.04919
,,038 0.3777 0..1336 2.60 0.6928 0.1810 " 0.04778
iS oS 0 2S S:6?25 0:i732 0.04507
^•°Tc nffiS 0 1278 2 66 0.6625 0.1695 0.04379
S 0.f627 S 2.68 , 0.6527 0.1656 0.04253
r^\nAn o7n 0 6429 0.1622 0.04131
'•T^ 0?S 01222 272 06332 0.1587 0.04014
•nf«o 0 3?18 oSS ' 2 74 0.6236 0.1553 0.03901
^•°^^^ n.fi 0 187 2 76 0 6140 0.1519 0.03788
IS S 0.nS S 0.6046 0.1483 0.03679
CHAP. XVII] VIBRATIONAL CONTRIBUTIONS 797
VIBRATIONAL CONTRIBUTIONS TO THERMODYNAMIC PROPERTIES IV
(H°-H3 (G°-HS) (H°-HO) (Q°-Ho)

c^'IT) c WIT) c
T T T
2.80 0.5953 0,1454 0.03572 5.00 0.0775 0.01074 0,00150

2.82 0.5862 0,1422 0.03471 5.10 0.0697 0.00948 0,001290
2.84 0.5771 0,1391 0.03371 5.20 0.0628 0.00836 0,001116
2.86- 0.5680 0,1361 0,03275 5.30 0.0563 0.00740 0,000965
2.88 0.5591 0,1331 , 0,03181 5.40 0.0506 0.00652 0,000837
2.90 0.5502 0.1301 0.03089 5.50 0.0454 0.00575 0.000727
2.92 0.5417 0.1272 0.03003 5.60 0.0409 0.00507 0.000629
2.94 0.5331 0.1244 0,02917 5.70 0.0367 0.00447 0.000545
2.96 0.5246 0.1216 0,02834 6,80 0.0330 0.00393 0.000472
2.98 0.5163 0.1189 0.02752 5.90 0,0295 0.00347 0.000409
3.00 0.5080 0,1162 0.02673 6.00 0,0265 0.00305 0.000353

3.05 0.4876 0,1099 0,02486 6.10 0,0236 0.00269 0.000306
3.10 0.4680 0,1038 0,02311 6.20 0,021^ 0.00236 0.000266
3.15 0.4489 0,0981 0,02150 6.30 0,0189 0.00208 0.000231
3.20 0.4305 0.0927 0,02000 6.40 0,0169 0.00183 0.000200
3.25 0.4127 , 0.0875 0,01861 6.50 0,0151 0.00161 0.000173
3.30 0.3956 0.0826 0,01731 6.60 0,0134 0,00142 0.000151
3.35 0.3787 0,0780 0,01608 6.70 0,0120 0.00125 0.000131
3.40 0.3627 0,0736 0,01498 6.80 0.0107 0.00110 0.000114
3.45 0.3471 0.0696 0,01393 6.90 0.00960 0.000962 0.0000993
3.50 0.3323 0,0656 0,01297 7.00 0.00852 0.000845 0.0000865

3.55 0.3177 0,0619 0,01207 7.10 0.0075& 0.000742 0.Q0Q075Q
3.60 0.3038 0,0584 0,01124 7.20 0.00674 0.000651 0.0000651
3.65 0.2903 0,0550 0.01045 7.30 0.00600 0.000571 0.0000565
3.70 0.2775 0.0519 0,00972 7.40 0,00534 0.000502 0,0000491
3.75 0.2651 0.0489 - 0.00904- - 7.50 0,00474 0.000441 0,0000427
3.80 0.2532 0,0461 0.00842 7.60 0.00423 0.0003868 0,0000369
3.85 0.2417 0,0435 0,00783 7.70 0.00376 0.0003497 0,0000320
3.90 0.2307 0.0410 0,00729 7.80 0.00332 0.0002979 0,0000278
3.95 0.2201 0,0387 0,00678 7.90 0.00298 0.0002618 0,0000241
4,00 0,2099 0,0364 0,00631 8.00 0.00265 0.0002292 0,0000209
4.10 0.1907 0,0322 0,00547 8.10 0.00234 0.0002008 0,0000181
4.20 0.1732 0.0286 0.00473 8.20 0.00208 0.0001759 0.0000156
4.30 0,1571 0.0253 0.00409 8.30 0.00185 0.0001542 0.0000135
4.40 0.1424 0.0225 0,00354 8.40 0.00164 0.0001352 0.0000117
4.50 0.1286 0.0200 0,00306 8.50 0.00145 0.0001187 0.0000101
4.60 0.1166 0.0176 0.00285 8.60 0.00129 0.0001038 0.0000088
4.70 0.1054 0.0156 0.00230 8.70 0.00113 0,0000908 0.0000077
4.80 0.0952 0.0138 0.00199 8.80 0.00101 0,0000795 0.0000067
4.90 0.0859 0.0122 0.00173 8.90 0.00089 0,0000896 0.0000058
798 THEEMODYNAMIC PROPERTIES [CHAP. XVII

VIBRATIONAL CONTRIBUTIONS TO THERMODYNAMIC PROPERTIES IV
, ,m^ (H -H„) (O - H „ ) (H°-H;) (G°-H;)

(o>'/T) c ^ T WIT) c T T
9.00 0.000792 0.0000610 0.0000050 9.50 0.000431 0.0000312 0.0000024

9.10 0.000701 0.0000535 0.0000043 9.60 0.000380 0.0000273 0.0000021
9.20 0.000620 0.0000466 0.0000037 9.70 0.000335 0.0000239 0.0000018
9.30 0.000549 0.0000408 0.0000032 9.80 0.000297 0.0000209 0.0000016
9.40 0.000486 0.0000357 0.0000028 9.90 0.000262 0.0000184 0.0000014
10.00 0.000232 0.0000161 0.0000012
From Equation (75),
S298.1(; = 0.2170 + 0.03948 = 0.2665
Calculations at other temperatures are carried out by the same procedure. The
results are summarized in Table LI, page 802.
Generalized Bond Frequencies, Useful approximations to the vibra-

tional contributions may be obtained by assuming that a particular type
of bond possesses two characteristic fundamental frequencies, which are
independent of the nature of the molecule in which the bond occurs.
Generalized average values of the stretching and bending frequencies
may be arrived at on this basis from compounds for which spectroscopic
data are available. This general procedure was developed by Benne-
witz and Rossner^* and modified by Dobratz.^^ A complication arises
from the fact that there are generally more degrees of freedom than are
accounted for by assuming two vibrational degrees per bond. Dobratz
proposed the following equation for handling this problem for calculating
heat capacities of nonlinear molecules,
/;3w — 6 — rir- — 2gi'
•jS(giCi» ) (76)
where
CM = total vibrational contribution to the heat capacity
2gi = total number of bonds in the molecule
c„^ = contribution to the heat capacity from the stretching vibra-
tions of bond i
c« = contribution to the heat capacity from the bending of bond i
n = number of atoms in the molecule
rir' = number of single bonds about which internal rotation of
groups can take place
" Bennewitz and Rossner, Z. physikChem., 393, 126 (1938).
" C. J. Dobratz, Ind. Eng. Chem., 33, 769 (1941).
CHAP. XVll] GENERALIZED BOND FREQUENCIES 799
Equation (76) assumes that all degrees of freedom not accounted for
by internal rotation, stretching or bending, make the same average con-
tribution to the heat capacity as do the recognize#bending contributions.
This same assumption can be made for calculating the contributions to
the other thermodynamic properties and equations developed similar
to (76).
In Table L are the fimdamental frequencies of various bonds expressed
in wave numbers, as recommended by Bennewitz and Rossner^* and
modified by Stull and Mayfield^ for hydrocarbons.
TABLE L
FUNDAMENTAL BOND FHEQUENCT ASSIGNMENTS
Stretch- Deforma- Stretch- Deforma-

Bond ing ul tion u'^ Bond ing al tion w'g
C—^H (aliphatic) 2914 1247 C==S 1050 530
C—C (aliphatic) 989 390 S—S 500 260
C = C (aliphatic sym.) 1618 599 S—H 2570 ,1050
C ^ C (aliphatic unsym.) 1664 421 C—N 990 390
C ^ C (aliphatic) 2215 333 C=N 1620 845
C—H (aromatic) 3045 1318 N—N 990 390
G—C (aromatic) 989 390 N—H 2920 1320
C = C (aromatic) 1618 844 N—0 1030 205
C—I 500 260 N=0 1700 390
C—Br 560 280 C—0 1030 205
C—CI 650 330 C=0 1700 390
C—F 1050 530 0—H 3420 1160
C-S 650 330
For the calculation of heat capacities above 300°K Dobratz assumed

that each degree of internal rotation contributed its full value of R/2.
The complete expression for heat capacity at constant pressure then
becomes
c; = 4R + nr.{R/2) + S,giC., + 1 ~ ^-hqiCsi (77)
Equation (77) may be evaluated by summation of the contributions from

Table XLIX corresponding to the various frequencies at a selected tem-
perature. The procedure is simplified by use of Fig. 166, in which the
contributions corresponding to various wave numbers are plotted as a
function of temperature.
Illustration 13. From Fig. 166 and the frequencies of Table L calculate the
heat capacity at constant pressure of methyl alcohol as an ideal vapor at 298. l°i
500°K, and 1000°K.
» D. R. Stull and F. D. Mayfield, Ind. Eng. Chem., 35, 639 and 1303 (1943).
800 THERMODYNAMIC PROPERTIES [CHAP. X V I I
250 300 350 400 500 600 700 800 9001000 1600
Temperature, °K
Fia. 166. Contributions of Bonding Frequencies to Molal Heat Capacities.
lo^lf
CHAP. XVII] TOTAL THERMODYNAMIC PROPERTIES 801
Solution: The frequencies of the bonds are obtained from Table L, and the cor-
responding contributions to heat capacity at 298,l°Kâre read from Fig. 166:
Bond 3< "^ OJj •-• •^<&;(298.1) giC5j(298.1)
C—H 3 2914 1247 0.000 0.529
C—0 1 1030 205 0.346 1.833
0—H 1 3420 1150 0.000 0.240
s 6 0.346 2.602
Substitution in Equation (77), since n, = 1, gives
c;(29s.i) = 7.95 + 0.99 + 0.346 + A^) (6) - 6 - 1 - 5\ ^^^^^ ^ ^^ 41
At other temperatures the same procedure is followed. Thus:

T c;
298.16 12.4
600 16.1
1000 22.4
A three-term equation is readily derived to express heat capacities calculated in this
manner, as a function of temperature. This method may be used for estimating the
heat capacities of a variety of compounds. The results become uncertain for complex
molecules, and, as previously mentioned, it is behoved that this method is best ap-
plied only to small molecules. The results obtained may then be extended to larger
homologs by means of the group contributions of Tables XXXIX-XLIII.
Total Thermodynamic Properties. Total values of (H° — HQ), cl, s°,

and (G° — Ho)/r are obtained by summation of the contributions result-
ing from translation, external and internal rotation, and vibration each
of which is calculated by the methods demonstrated in the preceding
sections. Electronic contributions may be neglected when one is dealing
with organic compounds at moderate temperatures.
Illustration 14. Using the results of Illustrations 6,10, 11, 12, calculate the values
of (H°— HO), Cp, s°, and {Q° — nD/T for methyl alcohol va the ideal-vapor state at
a pressure of 1 atm and temperatures of 298.1, 500, and 1000°K.
The results of the individual calculations from the preceding illustrations are sum-
marized in Table LI.
Where reliable values for potential barriers and vibration frequencies
can be obtained, it is desirable to calculate tables of thermodynamic
properties as functions of temperature, as is done in Illustration 14.
However, where potential barriers are uncertain and generalized bond-
frequency assignments are used, this rather tedious procedure is not jus-
tified, and it is more convenient to derive an empirical equation for heat
capacity from values at three temperatures and calculate the entropy
at 298.1°K. Free-energy changes at any temperature may be calculated
from these data and standard heats of reaction by the methods demon-
strated in Chapter XV.
TABLE LI
(H° - HS)
298.1''K 500°K lOOCK
Translation 1481 2483 4966
External rotation 888 1490 2980.
Internal rotation 221 570 1310
Vibration , 65 685 5220
Total 2655 5228 14476
Translation 4.97 4.97 4.97

External rotation 2.98 2.98 2.98
Internal rotation 1.67 1.72 1.32
Vibration 1.22 5.04 12.38
Total 10.84 14.71 21.65
S
Translation 36.316 38.883 42.326
External rotation 18.941 20.482 22.647
Internal rotation 1.080 1.97 3.02
Vibration 0.257 1.76 7.75.
Total 66.594 63.095 75.643
(G° - H;)/r
Translation -31.349 -33.916 -37.359
External rotation -15.961 -17.502 -19.667
Internal rotation -0.339 -0.820 -1.700
Vibration -0.040 -0.390 -2.530
Total -47.689 -62.628 -61.156
Illustration 15. Calculate the entropy of methyl alcohol as an ideal vapor at
298.1°K and 1 atm using the moment of inertia calculated in Illustration 10, the in-
ternal rotational contributions of Illustration 11, and the generalized bond, frequencies
of Table L for the vibrational contributions. Compare this result with that of Il-
lustration 14.
Solution: AU contributions are the same as those in Table LI except those of vi-
bration. These are read from Table XLIX with Equation (75) to correspond to the
values of [oi'JT) and {oi'JT) calculated in Illustration 13.
-< Stretching- ->- -<- -Bending-
Bond u\, T T
C—H 2914 1247 9.78 4.18 0.0000 0.0000 0.0000 0.0295 0.0048 0.0343
C—H 2914 1247 9.78 4.18 o.oboo 0.0000 0.0000 0,0295 0.0048 0.0343
C—H 2914 1247 9.78 4.18 0.0000 0.0000 0.0000 0,0295 0.0048 0.0343
C—O 1030 205 3.45 0.688 0.0696 0.0139 0.0835 1.1634 0.9234 2.0868
O—H 3420 1150 11.48 3.86 0.0000 0.0000 0.0000 0.0430 0.0077 0.0607
S = 0.0835 22404
The vibrational contribution, by an equation similar to (76), is then
sL(298.i) = 0.0835-t- (6/6) (2.2404) = 2.772
CHAP. XVll] PROPERTIES OF LIQUIDS AND SOLIDS 803
Comparing this value with that of 0.257 in Table LI indicates an error of 2.515 in
the total entropy as a result of the use of the generalized bond-frequency assignments.
The large error in the vibrational contribution to the entropy of methyl
alcohol calculated from the generalized bond frequencies results from the
low frequency of 205 for the deformation of the C-0 bond. The actual
spectrum of methyl alcohol summarized in Illustration 12 indicates that
no frequency actually exists in this range. The effect of this erroneous
frequency on the calculated heat capacity is shown by comparison of the
values of Illustration 13 with those of Table LI. The heat capacity at
298°K is too high by 14 per cent. At higher temperature the agreement
becomes better in this case.
It is thus evident that the bond-frequency assignments are compromise
values which may lead to serious errors particularly when one is working
with the first member of a series. This method appears to be best suited
for handUng second or third homologs. As further data are accumulated
it is hoped that improved generalizations of fundamental frequencies
may be developed:
Thermodynamic Properties of Liquids and Solids. The statistical
methods of the preceding sections yield results which are directly ap-
plicable only to the ideal-gaseous state. However, by the methods de-
veloped in Chapters VII and XII these data are readily converted to
apply to the liquid state at any conditions and also to the solid state if
data on the heat of fusion and heat capacity of the solid are available.
To obtain the properties of the saturated liquid, state, at a specified tem-
perature T, the properties of the ideal-gaseous state at 1 atm are first
corrected to the ideal state at the vapor pressure of the liquid at tem-
perature T. Corrections for deviations from ideality are then applied,
and the changes accompanying vaporization are subtracted. This pro-
cedure is described in detail for enthalpy and entropy calculations in
Chapter XII. It should be remembered that the free-energy change in
vaporization at saturation is zero. The effect of pressure on the free
energy of an ideal gas is obtained by integrating Equation (k) of
Table XXIV.
G* = G° -t- RT In p (78)
where
G* = free energy of ideal gas at pressure p atm
G° = free energy of ideal gas in the standard state of 1 atm
Corrections to free energy for deviation from ideal behavior may be ob-
tained by the combined use of Figs. 106 and 107. Thus,
(2!^=(5!^_(s*_s) (79)
PROBLEMS
1. Estimate from the group contributions of Tables XXXIX-XLIII the heat of
formation, entropies, and heat-capacity equations for the following compounds in
the ideal-gaseous state:
(a) Isoprene (2-methyl-l, 3-butadiene)
(6) Meta di-isopropyl benzene
(c) Iso butyronitrile
(d) Diphenyl
(e) o-Nitrobenzoic acid
2. The boiling point of isobutyro nitrite is 108°C, and its liquid density at 20°C
is 0.773 g per oc. From the result of problem Ic and the generalizations of Chap-
ters III, VII, and'XII calculate the entropy and heat of formation of hquid isobutyro
nitrile at 25°C under its vapor pressure.
3. From the covalent radii and bond angles of Table XLIV and the generalized
bond frequencies of Table L, calculate values of s°, (H° — HO), (a° — nl)/T, and
Cp at temperatures of 298.16, 500, and 1000°K for dimethyl sulfide.
4. From the bond-frequency assignments of Table L derive three-term equations
for the heat capacity at constant pressure in the range 298-1500°K for the following
compounds in the ideal-gaseous state:
(a) CCI3F
(c) Dimethyl amine

CHAPTER XVIII
HOMOGENEOUS REACTIONS
The economy of cnemical manufacture depends largely upon estab-
lishing high rates of production of desired products and minimizing the
production of undesired by-products. Thermodynamic principles may
permit selection of conditions which will result in favorable production
if equilibrium is reached but give no clue as to the rates at which the
various reactions occur. Thus, the displacement of a system from
equilibrium conditions may be considered as a potential which may
cause a reaction to proceed if other conditions are favorable. This dis-
placement is analogous to an electric voltage which may cause,a flow of
electricity. However, the rate of current flow is dependent on the con-
ductance of the system as well as on the voltage. Similarly, reaction
rates depend on the kinetic properties of the system as well as on the
driving potential. Quantitative knowledge of these kinetic properties
is desirable in the design and intelligent operation of a chemical process.
Industrial processes rarely reach equihbrium and in general are desired
to proceed toward only a partial equilibrium among the many reactions
possible in a complex system. In many cases if complete equilibrium
were reached undesirable by-products would predominate. However,
by proper selection of conditions the rate of formation of such by-
products may be suppressed and good yields of the desired product
obtained at conditions far removed from complete equilibrium. It is
thus evident that the kinetic properties of a system may determine not
only the size and capacity of process equipment but the yields and
qualities of the products as well. All three factors are of vital economic
significance.
Classification of Reactions. Chemical reactions may be classified as
homogeneous if only one phase is involved or heterogeneous if more than
one phase actively participates in the reaction. Where catalytic sur-
faces are present, a reaction may proceed simultaneously by a homo-
geneous mechanism in the main body of the fluid and by a heterogeneous
mechanism at the catalytic surfaces. In many cases even the walls of a
reactor exert catalytic effects to such an extent that the rates of pre-
sumably homogeneous reactions are found to vary with the ratio of
surface to volume.
A further classification is as flow and nonflow or batch processes. Both
805
806 HOMOGENEOUS REACTIONS [CHAP. X V I I I
types are of industrial importance, and it is frequently necessary to

interpret data obtained in the laboratory under nonflow conditions to
serve as a basis for the design of a commercial process under flow con-
ditions.
On the basis of the conditions of restraint a reaction may be classified
as isothermal, adiabatic, constant pressure, or constant volume. All
four types are of industrial importance, and the general case of a non-
adiabatic process occurring under simultaneously varying conditions of
temperature and pressure is frequently encountered.
Order of Reaction. For the general reaction aA + bB + cC • • • ^
rR -\- sS • • • which results from the simultaneous combination of a
molecules of A, b molecules of B, and c molecules of C • • • the reaction
rate for the forward reaction may be written as
r„ = kaâWc • • • (1)
Similarly, for the reverse reaction,
rL = k'aWs--- (2)
where
û> *•« = unidirectional rates of reaction in moles of a particular
compound formed or consumed per unit volume of
reacting system per unit time
k, k' = the reaction velocity constants
OA, CIB' • ' = activities of components A and B, etc.
The net rate of the reaction as written is then
r = r„- r[, = ka''â%ac •• - k'a'jfl's • •/ (3)
A t equilibrium, where r = 0,
aual••• k
where K = the equilibrium constant

Combination of (3) and (4) gives
r = /f I a » ' c , • • • ——j
I t is mathematically permissible to write any form of equation such

as (1) or (2) which involves an otherwise undefined proportionality
factor. The choice of the particular form of Equations (1) and (2) is
justified by experimental evidence that under many ideal conditions the
proportionality factors are constants which depend only on temperature
CHAP. XVIII] ORDER OF REACTION 807
and the nature of the reaction. Thus, Equations (1) and (2) serve to
define the reaction velocity constants k and k' which must be recognized
as inherently different from the classical reaction velocity constants of
chemical kinetics. Because many of the early laboratory studies were
carried out under conditions of constant volume it was convenient to
define reaction rates as rates of change of concentration expressed in
moles per unit volume. Since concentrations are changed by change of
volume as well as by reaction, the rate of change of concentration is a
proper expression for rate of reaction only in a system of constant volume.
Similarly, the classical law of mass action was formulated in terms of
concentrations, and kinetic equations were developed to relate rate of
change of concentration to the concentrations of the reactants. As
commonly written, such equations are restricted to conditions of con-
stant volume, and the reaction velocity constants so defined have differ-
ent dimensions and significance from those of Equations (1) and (2).
The relationships between reaction velocity constants expressed in terms
of activities and those in terms of concentrations are discussed on
page 820.
The order of the forward reaction is equal to the sum of the exponents
a, 6, c • • • of Equation (1). This sum is equal to 1.0 for a first-order
reaction, 2.0 for a second-order reaction, and 3.0 for a third-order reac-
tion. In reactions of simple order the exponents of the activity terms iij
the fundamental rate equation are equal to the minimum number of
reactant molecules which must simultaneously combine in order to
effect the reaction. This minimum number of combining molecules is
termed the molecularity of the reaction. For reactions of simple order
the order is equal to molecularity. Thus, a simple first-order reaction is
unimolecular, and a simple second-order reaction is bimolecular.
I t is important to recognize that the exponents a, h, c in Equation (1)
and r and s in Equation (2) correspond to the number and kind of mole-
cules which actually simultaneously combine and do not necessarily cor-
respond to the molecular proportions used iri the stoichiometric equation
representing the over-all reaction. For example, the synthesis of methanol
from carbon monoxide and hydrogen is represented by the stoichiometric
equation CO + 2H2 —> CH3OH. However, it is probable that in a
homogeneous reaction the alcohol results from secondary hydrogenation
of formaldehyde which is formed as a primary product. Similarly, the
combustion of a gas such as benzene actually proceeds by a succession
of low-order reactions and never by the mechanism suggested by stoi-
chiometry. Thus, Equations (1) and (2) are applicable only to individ-
ual chemical steps in one direction and not to over-all stoichiometric
results. In this respect kinetic equations differ from thermodynamic
808 HOMOGENEOUS REACTIONS [CHAP. XVIII
equilibrium relationships which are independent of the mechanism of the

reaction and dependent only upon the Over-all results.
In the application of Equation (1) or (2) to a single chemical step the
exponents a, b, c- • • are integers whose sum is probably never greater
than three.
Theory of Absolute Reaction Rates. Eyring and coworkers' have
developed from the principles of statistical mechanics a theory which
permits calculation of reaction velocity constants from a knowledge of
chemical structure and energy distribution without actual rate measure-
ments. However, in its present state of development rates calculated
by this method may be in error by a factor of from 10 to 100 or more.
In view of the wide variations of reaction rates this degree of agreement
furnishes confirmation of the theory but is not sufficient to serve as a
basis for process design. For this reason the engineering utility of the
method is limited to rational extension of fragmentary rate data rather
than to the prediction of rates.
According to the Eyring theory a chemical reaction is preceded by the
fomiation from the reactant molecules of an activated complex which then
decomposes to form the final products. The activated complex is con-
sidered as present in the reacting system in continual thermodynamic
equilibrium with the reactants. Thus, if reactants A and B combine to
form products R and S, the reaction is considered as proceeding in two
steps as follows,
aA + bB:^X^ -*rR + sS • • • (5)
where X^ represents the activated complex formed by A and B. Since
equilibrium is maintained between the reactants and the activated com-
plex, the rate of the reaction is determined by the rate of decomposition
of the complex. If the properties of the activated complex are
identified by the superscript *, there results
^ - K ^ - (6)
where
o , aA, and as = activities of the activated complex and re-
actants A and B, respectively
K^ = the equilibrium constant between the react-
ants and the activated complex
By statistical mechanics it was developed by Eyring that in the
majority of cases the rate of decomposition of the activated complex
^ S. Glasstone, K. J. Laidler, and H. Eyring, " The Theory of Rate Processes,"
McGraw-Hill Book Company, New York (1941).
CHAP. X V I I I ] THEORY OF ABSOLUTE REACTION RATES 809
and, therefore, the rate of the reaction, is given by the following equation,
r. = ^c' (7)
where r« = rate of decomposition, moles/(sec) (unit volume)

c* = concentration of activated complex, moles per unit
volume = N^/vm
k = Boltzmann constant = (1.3805)10-" erg/(°K) (molecule)
h = Planck constant = (6.6240)10-^^ erg-sec per molecule
^ = (2.0842)101" per (°K)(sec)
n '
iV* = mole fraction of the activated complex
Vm = average molal volume of the system
In dealing with liquid systems or with gases which do not form ideal
solutions, the standard state for the expression of activities is conven-
iently taken as that of the pure components at the temperature and
pressure of the system:
a* = y^N^ = Y^cW (8)
7*. = activity coefficient of the activated complex
Combining Equations (7) and (8), with this choice of standard states
gives
k r g^
'•« = -h- U -
yhm '. (9)
When ideal solutions of gases are dealt with, where it is convenient to

select as the standard state the pure components at the temperature of
the system and a unit fugacity of 1 atm,
a} = KW* (10)
and'
'•=sS? ''^'
where
v* = fugacity coefficient of the activated complex
Zm = mean compressibility factor of the mixture
w — total pressure, atmospheres
R' = gas law constant in units consistent with ir,
Vm, and T

k a*'
'•" = r ^ T ~ (13)
hR' vhm
Expressions for the reaction velocity constant are obtained by combining
Equations (1) and (6) with (9) or (13). Thus, where unit activity is
equal to unit mole fraction^
kT a* k r K^
For gaseous systems approximating ideal solutions,

k _a} k K^
hR'zy alal ~ hR' z^ ^ ^^
I t is evident from Equations (14) and (15) that for a given reaction the
values of k and K^ are dependent upon the choice of the standard state.
Thus the k and K^ of Equation (15) are numerically different from those
of Equation (14) and must be used with care to maintain a consistent
choice of standard state.
Equations (14) and (15) express the reaction velocity constant in
thermodynamic terms involving K^ the equilibrium constant relating
the activated complex and the reactants. Thus, by evaluation of the
thermodynamic properties of the activated complex and the reactants
the rate of a reaction is established. Since K^ is a true equilibrium con-
stant expressing the ratio of the activities of products to reactants, it
may be expressed in terms of the standard free-energy change in form-
ing the activated complex. Thus,
K* = e - BT = e - B r + - r (16)
where AG*, AH^, A(S* = the free energy, enthalpy, and entropy changes,
respectively, accompanying the formation of the activated complex from
the reactants all in their standard states. It should be noted that K
is dimensionless only for first-order reactions. Combining Equations
(16) and (14), for liquid systems yields
kr 4Ht ASt ^ ^
k = --r-e--Sf + -R- (17)
hy*Vm
or for gases, if (15) and (16) are combined,
k . AH* , ASt
CHAP. X Villi THEORY OF ABSOLUTE REACTION RATES 811
The Eyring equation of absolute reaction rates may be applied to all

types of reactions and permits direct calculation of the reaction velocity
constant if the enthalpy and entropy of activation are known or if they
can be calculated. It should be noted that the time unit in the reaction
velocity constant of Equations (17) and (18) is in reciprocal seconds
since it is based upon the frequency expressed in the ratio of k/h.
Illustration 1. In the dimerization of ethylene at 673 °K Glasstone, Laidler, and
Eyringi report a value of AH* equal to 32.73 kcal and AjS^-equal to —35.00 entropy
units per mole of butene-1 for standard states of unit fugacity. Assuming ideal
behavior:
(o) Estimate the value of the reaction velocity constant.
(b) Estimate the instantaneous rate of reaction in pound-moles of ethylene con-
verted per cubic foot per hour at 673 °K and with ethylene at 2 atm partial pressure.
(c) Calculate AS' when the standard state is taken as unit mole fraction at 673°K
and 2 atm pressure.
When Equation (18) is used, R' = 0.08206 (liters) (atm)/(g-mole)(°K). Assume
v^ = 1,0 and Zm = 1-0, then
-32730 35.00
(2.084)10"" (1.987)(673) 1.987 ioor/,«-.\ i u x //i-j. ^, v
(a) k = . „ - p - - - , e = 13.35(10-8) g-mole butane/(hter)(sec)
(0.08206)
(atm) 2
This constant, expressing the rate of formation of butene, is compared by Glass-
tone, Laidler, and Eyring with an experimental value of k" = 0.0374 liter/(atm) (hr)
for the rate of conversion of ethylene. The units of this constant result from the
expression of reaction rate in terms of the rate of reduction of pressure in atmospheres
per hour as the reaction proceeds at constant temperature in a system of constant
volume. Thus, if ideal-gas behavior is assumed, rj/ = A;"p^ atm per hr. This
rate r J,' is related by the ideal-gas law to r„ the rate in moles/ (liter) (hr). Thus, since
n/V = p/R'T, r„ = r'u/R'T and k = k"/R'T. Since 2 moles of ethylene are
converted to 1 mole of butene, the experimental value offc"is converted to the units
of the preceding illustration by the following expression:
; ^ - * = (673)(0.08206K3600)(2) = '-'^''''^ g-mole/(Uter)(sec)(atm)»
(6) ru = fca2 = (13.35)(10-«)(2)2 = 53.4(10-8) g.jjôle butene-1/(sec)(liter) or

(53.4)(10-8) ( — ) (28.25)(2) = 2.39(10-^) lb-mole ethylene/(hr)(cu ft)
\ 454 /
(c) If the standard state of each component is taken as unit mole fraction then
the pressure of each component in its standard state is 2 atm. If ideal behavior is
assumed, application of Equation (XI-142) to the reactants and activated complex
gives
AS? = AS* - AnRln- = -35.00 - (-1)1.987 In 2 33.67
^ PI
For ideal-gas behavior the value of Aff* is not altered with change in standard state.
Ordinarily in the application of Equations (17) and (18) the enthalpy
and entropy of activation are assumed to be independent of temperature
since the change in heat capacity which accompanies the activation is

small. For unimolecular reactions this change is negligible, and AH^
and A/S* may be treated as constants over wide ranges of temperature.
This assumption is also satisfactory for other reactions over moderate
ranges of temperature. In such cases the values of AH^ and AS^ are
average values for the temperature range involved.
Where rate data are t6 be extended over wide temperature ranges the
effect of ACl, the change in heat capacity accompanying the formation
of the activated complex should be considered. Where AC^ is constant,
Equations (XVI-9 and 10) become
AH^ = 4 + AClT (19)
AS^ ^ II + AClln T (20)
Equations (19) and (20) are ordinarily used only where rate data are
translated to a widely different temperature range, as, for example, if
low-temperature rate data in a forward reaction are combined with
thermodynamic data in order to estimate rates of the reverse reaction at
high temperature. For rate equations in either temperature range AS^
and AH^ may be taken as constant at average values.
If the Eyring equation is used for the interpretation of experimental
data it is necessary that rate measurements be available at two tempera-
tures in order to evaluate AJtf* and AS^. If AS^ can be approximated
by theoretical calculations, rate measurements at one temperature
suffice to evaluate reaction rates at other temperatures. By application
of the statistical methods developed in Chapter XVI, values of AS* can
be estimated which in many cases are probably as accurate as the average
experimental data. Eyring and co-workers^ have also devefoped
theoretical methods for estimating AH^, but these are uncertain for
engineering application at present.
According to the Eyring theory an activated complex must be formed
for the reverse reaction which has the same thermodynamic properties
as that formed in the forward reaction. Since the forward complex is in
equilibrium with reactants and the reverse complex is in equilibrium
with the products it follows that the actual over-all rates are determined
by the rate of conversion of one complex to the other. Since these two
complexes are identical in thermodjmamic properties and differ only in
the direction in which they are decomposing, it is customary to use but
one symbol for both. Thus if Equation (17) is apphed to the reverse
reaction and the reverse properties are designated with primes, there
results,
fr' = - 7 - T ^ - = r r - ^ '''' '^ (21)

CHAP. XVIII] THEORY OF.ABSOLUTE REACTION RATES 813
By combination of (17) and (21) with (4),
AH° AS°
k
K (22)
k'
where
MI° = AH^ - AH^' (23)
AS° = AS^ - AS*' (24)
AH", AS° = standard enthalpy and entropy changes of the over-all re-
action
Equation (22) permit^ evaluation of the velocity constant of the reverse
reaction as a function of temperature if that of the forward reaction
together with the over-all standard heat and entropy changes of the
reaction are known.
The significance of the activated state is clarified by Fig. 167 which
indicates the enthalpy changes when reactants at A change to products
at B. In order that the reaction
may proceed the reactants must be
energized to the level corresponding
to H^ by the acquisition of enthalpy
of activation AH^. When the complex
decomposes energy AH^' is released H
to form the products at level B. In
order that the reverse reaction may
take place an energy of activation
AH^' must be acquired so that the
molecules may pass over the energy
barrier represented by level H^. The
FIG. 167. Enthalpy of Activation.
heat of reaction is the difference be-
tween the enthalpies of the final and initial states which in Fig. 167 is
negative for the forward and positive for the reverse reaction.
Illustration 2. Nitric oxide is decomposed according to the foUowing second-order
reaction:
2N0 ?± Na + O2
The reaction velocity constants as converted from values given in the International
Critical Tables are
at y = 1620°K, k = 0.0108 g-mole/(liter) (sec) (atm)>
at T = 1525°K, k = 0.0030
For the reaction AC| may be taken as —1.0 cal per "K per g-mole of Nj formed.
The fugacity coefficients and compressibility factors may be assumed to be unity.
Calculate:
(a) The values of AH* and AS* at the average experimental temperature of
1572°K.
(6) The values of ^H^, AS\ AH^', and AS^' at 2500°K.

(c) The values offcand fc'at 2500°K.
(a) The enthalpy and entropy of activation of the decomposition reactions are
evaluated by substitution in Equation (18). Thus, in the specified units,
k (1.3805) (10-")
l o g — = log ^^ — = log (2.540) (10")-= 11.40479
^hR' ^(6.62i0)(10-") (0.08206) S >.^->'^"M " ; x . ^.r,
Substituting in Equation (18) gives
-logO.0108 = 11.40479 - ^ : ^ ^ ^ + ^ , = -1-96658
Afl* AS'
log 0.0030 = 11.40479 + = -2.52288
^ 4.576(1525) 4.576
If these are solved simultaneously,
AH' = 66,200; AS* = -20.323
(6) From Equation (19) at 1572°K,

66,200 = J ^ - 1.0(1572) or I ^ = 67,772
From Equation (20),
-20.323 = I%- 1.0 In 1572 or 1% = -12.963
At 2500°K.
AHi = 67,772 - 1.0(2500) = 65,270
Afit = -12.963 - 1.0 In 2500 = -20.788
From the data of Tables V, XIV and XXXV (pages 214, 253, and 701)
AH" = -43,200
r»2500
J I
291
[(0.09 - 0.00039ir + (0.169) (10-«)r2] dT = -43,327
AS" = - 5 8 6 + r ^ ^^"""^ ~ 0.0003917- + (0.169)(10-°)?"] dT ^ _^^^

J298 T
AG'JT = (-43,327/2500) +6.009 = -11.322
AG"
log K = :—- = 2.47421
^ 4.576r
K = 298
From Equations (23) and (24),

AH*' = AH* - AH" = 65,270 - (-43,327) = 108.597
AS*' = AS* - AS° = -20.788 - (-6.009) = -14.779
From Equation (18),
65,270 20 788,
A; = (2.540)(10")e B(2500) R = (2.540) (10") (10-«>-2«»)
= 14.346
CHAP. X V I I I ] ARRHENIUS EQUATION 815
108,597 14.779
jfc'= (2.540) (10")e K(2500) R = (2.540) (10") (10-«-"«)
= 0.0477
For verification of agreement with K obtained from 4G°,
k' 0.0477
In the present state of development of the theory there is room for
considerable speculation as to the exact structure of the activated com-
plex in a given reaction. Where the complex is formed from a single
molecule it is assumed to be an abnormal structure which has acquired
additional bonding energy resulting in deformation and changes of the
normal energy levels. The formation of the complex is accompanied by
a large absorption of energy and increase in enthalpy, but since the
general configuration of the complex is similar to that of the molecule
the corresponding change in standard entropy is generally small and
may be either positive or negative.
Where the activated complex is formed from two or more molecules
new bonds are established with generally a large increase in enthalpy.
Since two or more molecules go to form the complex there is a loss in
translational entropy, and the over-all standard entropy change accom-
panying formation of the complex is generally negative. If a single
product molecule results, the entropy of the complex is similar to that
of the product. Thus, in the dimerization of ethylene discussed in Illus-
tration 1, Glasstone, Laidler, and Eyring^ assume that the activated com-
plex is a linear molecule having the same entropy as butene-1 but with an
enthalpy greater by approximately 57 kcal per g'-mole. In the oxidation
of nitric oxide by the termolecular reaction 2 N 0 -j- O2 = N2O4 the acti-
vated complex is assumed to have the structure 0 = N—0—0—^N = O
with the bonds of the oxygen atoms at 90° angles and with free rotation
about the 0 — 0 bond.
Arrhenius Equation. Previous to the development of the Eyring
theory of absolute reaction rates the Arrhenius equation was generally
used to evaluate the" effect of temperature upon the reaction velocity
constant. According to Arrhenius a reaction is able to take place when
the requisite molecules which collide with each other possess an abnor-
mally high energy content greater than the average content of the
corresponding temperature level by an amount termed the energy of
activation, E. In accordance with the Maxwell-Boltzmann distribution
law the fraction of molecules energized to this activated state is equal to
E
e ^^, and the following expression results
k = AeRT (25)
816 HOMOGENEOUS REACTIONS " [CHAP. XVIII
where E = the molal energy of activation

A = proportionaUty factor characteristic of the system and
termed the frequency factor
Equation (25) is designated as the Arrhenius equation and has been
found to represent the temperature variation of the reaction velocity
constants of single reattions with satisfactory accuracy. By comparison
of Equations (18) and (25) it may be noted that for gaseous reactions
the Arrhenius equation is in agreement with the Eyrmg equation where
k ^
E = AH*, and A = —z—— e B . In the general case for nonideal
hv ZmR
solutions represented by Equation (17) the frequency factor, is not
temperature independent as assumed in the Arrhenius equation, but
there is ordinarily little numerical difference between AH^ and E.
DIFFERENTIAL RATE EQUATIONS

For the first-order reaction A ^R, Equation (1) for the forward rate
may be written as
r„ = kaA (26)
Where activities are unity at unit mole fraction ax = JANA =
yA{nA/n,), and Equation (26) becomes
r„ = kyA— = JCJANA (27)

nt
in which 7^ = activity coefficient of component A
UA = moles of reactant A present in the system under
consideration
ni = total number of moles in the system under con- ^
sideration
For gaseous reactions in ideal solutions and where activities are unity
at unit fugacity, a^ - VVA1!Â = WAinA/nt), and x = (zmntB'T)/V.
Substituting in Equation (26) gives
kvAUATT hvAZmR'TUA ,„„,
r„ = ^ ^ = (28)
For ideal gases VA and Zm are unity, and Equation (28) becomes
* kuAT* ' * hR'TuA .„„,^
r„ = —^ = kvA = — ^ (29)*
CHAP. XVIII] DIFFERENTIAL RATE EQUATIONS 817
, , —dnA
or in a batch system where r„ = ——— and T = time
Var
- ^ = kR'Tn^ (30)*
Integration gives
In—= kR'Tr (31)*
UA
where TIAO = moles of A initially present when T = 0. It is evident
from Equations (29-31) that for a first-order reaction proceeding in a
mixture of ideal gases the extent of conversion of the reactant in a given
time interval at constant temperature is independent of pressure Or
volume while the rate of reaction per unit volume is directly proportional
to the first power of the pressure.
For second-order reactions of the type A •\- Bv^R- • • the rate of the
forward reaction is expressed by
Tu = kaAaa (32)
Where activities are unity at unit mole fraction Equation (32) becomes
kyAnAyariB .„„,
r„ = J (33)
For gaseous reactions in ideal solution where activities are equal to
fugacities,
kvAVBTlAnBT^ JcVAVB(ZmR'T)ÂnB
For ideal gas behavior Equation (34) re'duces to

* knAnB{Tr*Y J * * kjR'T'PnAnB
r„ — 2 — KJIAVB — ^T; C^SJ
n, V
From Equation (35) it is evident that in mixtures of ideal gases the
rate of a second-order reaction is proportional to the square of the
pressure or inversely proportional to the square of the volume of the
system.
For third-order reactions of the type A -{• B -{• C^R • • • the rate of
the forward reaction is expressed by
r„ = kaAttBac (36)
Where activities are unity at unit mole fraction the rate equation
becomes
kjAnAyBnBycnc ,__.
r„ = 3 (37)
For gaseous reactions which approximate ideal solutions
kvAfsvcnAnsncTT^ kvAVRVciZmR'TYnAUBnc ,„„.
r„ = 3 = zr, (38)
If the gases behave ideally,
r* = ^ ^ ^ ^ = lvA% = ^^^^^^^^"-"^ (39)*

From Equation (39) it is evident that in ideal-gaseous systems the
rate of a third-order reaction at constant temperature is proportional to
the cube of the pressure or is inversely proportional to the cube of the
volume of the system.
Homogeneous reactions of higher than the third order are improbable,
and it is doubtful that any have been observed; only a iew third-order
reactions are known.
Simultaneous, Consecutive, and Reverse Reactions. The rate equa-
tions thus far discussed are apphcable to the effects of single reactions
without consideration of other reactions which may be occurring in the
system simultaneously. Any one of the typical reactions considered
might be accompanied by one or more simultaneous reactions, such as
B + F—> S, involving one or more of the same reactants. In such cases
the results of the simultaneous reactions are additive, and the rate at
which component B is consumed is the sum of the rates of all the reac-
tions in which it is involved. Similarly, if the product of a primary
reaction enters into a secondary or consecutive reaction the net rate at
which it is formed is the diiference between the rates of the primary and
secondary reactions. In this manner over-all rate equations may be
developed for complex reacting systems. For example, the over-all
rate of reaction of reactant A in the consecutive reactions A -\- B -^ R
and A + R—> S is given by
Tu = kiaAttB - f kiQAaB (40)
As developed in Chapter XVI, all reactions are to some degree reversi-

ble, and if conditions are such that the rate of the reverse reaction is
important its effect must be included in the over-all rate equation.
Thus, if we consider a reversible first-order reaction of the type A'^R,
the net rate of reaction of A is expressed by
r ^kaA — k'aa (41)
where
k' = reaction velocity constant of the reverse reaction
CHAP. X V I I I ] REACTIONS OF COMPLEX ORDER 819
At equilibrium the net rate of reaction must be zero, and
where
K — equilibrium constant
r = A-(a^-^) (43)
Similarly, for a reactidn of the type A + B ?^ i2 + >S,
r = k UiAaB ^ j (44)
I t should be noted that the equilibrium constant is equal to the ratio of

the reaction velocity constants only where these constants refer to the same
chemical steps. For a complex reaction the over-all equilibrium constant
is not related by any simple ratio to the reaction velocity constants of the
separate steps. <
In Equations (43) and (44) the terms {aA—as/K) and {aAas—asas/K)
express quantitatively the displacement of the reactions from equilibrium
and may be looked upon as the driving forces or potentials which cause
the reaction to proceed toward equilibrium. These terms are analogous
to voltage as the potential causing an electric current to flow while the
reaction velocity constant is analogous to the electrical conductance
which determines the rate of flow of electricity under a given voltage.
Reactions of Complex Order. By application of Equation (1) to
experimentally determined rate data it has been found that many
apparently simple reactions do not correspond to a simple order and that
exponents other than integers must be employed in the rate equations.
Thus, a reaction may have' an experimentally determined order of 0,
0.5, or 1.5. Reactions of this type are classified as of complex order.
It is difficult to assign a definite molecularity to reactions of complex
order, and molecularity and order bear no simple relationship to each
other. This distinction between order and molecularity must be clearly
understood, and it should be recognized that exponents in a rate equa-
tion can be determined only from experimental measurements of rates
at varying concentrations.
Many theories have been advanced as to the mechanism of reactions
of complex order. One of the most useful views is that siich a reaction
actually consists of a chain of reactions involving the formation of inter-
mediate free radicals or atoms which do not appear in the final products
in appreciable proportions. For example, the formation of hydrogen

iodide from iodine and hydrogen is a simple second-order reaction
corresponding to the equation H2(g) + h{g) <=i 2HI, However, the
parallel formation of hydrogen bromide from hydrogen and bromine is
of complex order. This may be explained by intermediate dissociations
to form atoms resulting in the following six simultaneous and consecu-
tive reactions: '
(1) Bra -t- H2 ?=^ 2HBr
(2) Bra ^ 2Br
(3) H2 ^ 2H
(4) " H2 -F Br <=4 HBr -I- H
(5) , H -f Br : ^ HBr
(6) Brj + H ^ HBr + Br
A complete rate equation for the over-all reaction including the

effects of all six contributing reaction rates would be so complicated as
to be of little value. However, because of the wide variations in mag-
nitude of specific reaction rates, one or two of the reactions in such a
sequence may proceed so slowly as to determine the rate of the over-all
reaction while others may proceed so rapidly that equilibrium may be
assumed to exist at all times. For this reason the over-all rate is gen-
erally determined by the rate of one or two of the intermediate reactions
and can frequently be expressed by an equation based on these rate-
determining steps only. The subject of complex chain reactions is
discussed in more detail on page 843.
Rate Equations in Concentration Units. Equation (1) expresses
reaction rates in terms of thermodynamically defined activities which are
applicable to all types of systems under all conditions. The rate equa-
tions developed on this basis have the advantage of a simple relationship
to the thermodynamic equilibrium constant which permits the establish-
ment of a single equation which is applicable to the net rate of a reaction'
when approaching equilibrium from any direction. These equations are
recommended for all problems in commercial reactor design where con-
ditions are generally variable and nonideal.
Most of the reaction-rate data which have been gathered in the past
have been correlated in terms of the classical law of mass action proposed
by Guldberg and Waage in the following form:
r = k^^BCc ••• (45)

where
CA = HA/V = concentration of A, moles per unit volume
kc = reaction velocity constant in concentration units
CHAP. XVIIl] RATE EQUATIONS IN CONCENTRATION UNITS 821
For systems in which activities are taken as unity at unit mole
fraction Equation (1) may be written as
r = KjACAYiyBCByilcCc) '••• M^^"^' • •' (46)
where Vm — average molal volume of the reacting system
Combining (45) and (46) with this reference state for activities and
using consistent units throughout gives
•h = Hyly%7y)---(v^y^''- (47)
For ideal solutions of gases where activities are taken as unity at unit
fugacity, Equation (1) becomes ••
r = H''A.CA.n''BCB)\''cCcy • • • (zJi'T)-^' (48)
Combining (45) and (48), for gases in ideal soluijon, and using consis-
tent units throughout yields
h = kvlvlvl • • • (zJi'TY+^' (49)
Equations (47) and (49) relate the two reaction velocity constants k
and kc and may be used to calculate one from tho other. Equilibrium
constants in concentration units corresponding to k^ are also in use to
some extent. Thus,
îfis ' ' ' k„
J^^ = ^T~.-f'
a b c
CAOBPC '
(50)
This equilibrium constant Ke is related to the true thermodynamic
equilibrium constant K by the following equation where activities are
unity at unit mole fraction:
Ko=^K '^""I'^f M <"+*+= • • •'-''+' • • •> (51)

T«7s • • •
Where activities are unity at unit fugacity,
K, = K "^yf {ZrJt'T) '"+»+= • • •'-''+• • • •' (52)

VRVS • • •
In many cases it may be assumed that ideal conditions exist, and the
values of y, v, and z in Equations (46-52) may be taken as unity.
The widespread use of concentration units in kinetic relationship has
resulted from the fact that many laboratory rate measurements have
been made at conditions of substantially constant volume. Under these
conditions data are determined directly in concentration units, and the
resulting rate equations are more simply integrated than if activities are
employed. It is convenient to use rate equations in concentration units
for the analysis of such laboratory data at constant volume. Thermo-

dynamic reaction velocity constants may then be calculated from
Equations (46-52).
In dealing with gaseous-reaction-rate data in concentration units the
normal molal volume is frequently used as a unit of volume. Thus, in
the metric system concentrations and rates are expressed in terms of
gram-moles per 22.41 liters. Concentrations in these units are numeri-
cally equal to pound-moles per 359 cu ft. If pressures are expressed in
atmospheres, the values of the gas-law constant R' is 1/273 in (atm)
(22.41 liters)/(°K)(g-mole) or 1/492 in (atm)(359 cu ft)/(°R)(lb-mole).
The normal molal volume may be used exactly like any other unit of
volume, but careful designation of units is necessary to avoid confusion.
I t is a good plan to verify the dimensional consistency of any rate equa-
tion by writing it in terms of the units employed and making sure that
all cancel.
Illustration 3. A second-order reaction in a system of ideal gases at a temperature

of 910°K has a reaction velocity constant kc = 70 liters/(g-mole) (sec). Calculate
the reaction velocity constant fc in g-moles/(liter) (atm)* (hr).
Solviion: From Equation (49), if kc is converted to consistent time units,
fc«(3600) (70) (3600) ,, g-moles

fc = (B'T)2
,„,„,„ = [(0.08206)
,, ,(910)]2
,,. = 45 (liter)
- (atm)2(hr)
INTEGRAL CONVERSION EQUATIONS
Equations (1) and (45) express the instantaneous rate at which a reac-'
tion proceeds as a function of the conditions in the system for a particular
infinitesimal interval of time dr in a batch process or for an infinitesimal
increment of volume dV in a flow process. As the reaction proceeds
these conditions change with time in a batch process and with distance
traveled in a flow process; as a result the rate of reaction continually
changes. In order to evaluate the effects produced during a finite
interval of time or in a finite flow reactor it is necessary to integrate the
instantaneous rate equations with due consideration of the changes of all
variables.
Nonflow Reactions. It is convenient to express the results of a reac-
tion in terms of the extent of conversion of the limiting reactant A.
The conversion XA is defined as the moles of A transformed per unit mass
of original reactor charge or feed. Thus,
nA = riiio - ZA (53)
—driA = dxA (54)
CHAP. XVIIl] NONFLOW REACTIONS 823
where nÔj-riA = the number of moles of A per unit mass of reacting sys-
tem initially and at any time r, respectively.
From the definition of reaction rate, in a batch process,
- + ^ (.5,
where r = moles of A converted per unit time per unit volume of reactor
occupied by the reacting system
Vc = volume occupied by the reacting system per unit mass
The integration of Equation (55) requires complete expression of both
r and F„ as functions of XA- The instantaneous rate of reaction r is
related to XA through the differential rate equations of the preceding
section.
For a batch or nonflow reaction Equation (55) is integrated directly
to obtain the time r required for a specified extent of conversion XA-
Thus,
r^dxA
(56)
Jo Vcr
If the reaction is restrained to constant volume and temperature the
integration of Equation (56) is readily carried out for reactions of simple
order in systems which form ideal solutions. For a first-order reaction
in an ideal-liquid system in which the reverse reaction is negligible the
rate is expressed by Equation (27) which may be combined with Equa-
tions (47) and (56). Thus,
rÂ dXATlj ^ r^'A dXA .
Jo VMUAO - XA) JO JC/JIAO - XA)
where Ut = total number of moles per unit mass of reacting system =

YJvm- As previously pointed out, the expression in terms of A-^ is more
e3,sily integrated. Thus
r = - In (58)
ko UAO — XA
By the same procedure integrated equations may be derived from

unidirectional second- and third-order reactions and for simple reversible
reactions. The results of these integrations, in concentration units, are
summarized in Table LIT. These equations are directly applicable to
either ideal-liquid or gaseous systems under the specified conditions.
Integrated equations of this type may also be applied satisfactorily to
nonideal systems in which the term involving the activity coefficients does
not vary greatly.
TABLE LII
HOMOGENEOUS REACTION R A T E S AT CONSTANT VOLUME
FOK IDEAL SOLUTIONS AT CONSTANT TEMPERATURE
r = moles of component A converted per unit time per unit volume
T = time elapsed
fee = reaction velocity constant in concentration units
Ke = equilibrium constant in concentration units
riAo, nBo = moles A, B, inilially present per unit mass of charge
XA = moles of A transformed per unit mass of charge
Vc = volume of reacting system in reactor a t time T per unit mass of initial
charge
riA = wxo — XA = moles of A present a t time T per unit mass of charge
dUA
General rate equation: r = — •
VcdT
Unidirectional Readions
First Order: A —> yP
hUA 1 , UAO , ,
r = -rj^; T = - In (a)
Second Order: 2A —> vP
-W-- hriAa («Ao — XA)

Second Order: A + B —» yP
Ve . rrefio(»AO — XA)~\
) — UBO) \,nAo{nBa — XA)J
Third Order: 3A - ^ vP
{^\ _ ^ ^ r 2nAo -XAI
r
~ '\ y j ' ^ ~ 2hn\, L(nxo - XA^J ^'^^
Third Order: 2A+B-^vP
, n\nB yl n \ rXA ,
r = kc —r;r":
Vl ' T
' = kc(,nBti — ^nAo)LnAo{nAo
• ; — XA) +' nô — 2WB
, rnAoiuBi, -^XA)! , .
In -. ^ (e)
L TlBo (JIAO — XA) J
Third Order: A + B + C - ^ i-P
_ UAnBric
1' — 'Ô y i V
^ r ^ 1 '^•^'' , ^ 1 »g» , ^ 1 »»o 1 /n

T =
Kc L - A C n ^ o — XA A B nsa — XA B U rico — XAJ
where A = TIBO — UAD; B = nao — nco; C = nco — TIAO
Reversible Reactions
First Order: A *^ R
kc ( ncN Kc , r
r = - ( n . - - j ; r = ^ ^ ^ . . - ^ In \_^^^^^ _ ^^^ _ ^^^ ^ ^^^J (g)
CHAP. XVIII] NONFLOW REACTIONS 825
For various types of second-order and mixed second-first order reversible reactions
T may be expressed by Equation (h). The constants o, 6, c, and q are defined in each
case.
q = iac — 6^
where
3 < 0; T= -
First and Second Order:

A •\-Evi-R
a = VcKcTiM — nlio ^ a — KcnAoriBo — VCURO
b = —KeVe — inm (>) b = —KcTiBo — KcnAo — Vc (J)
c 4
2A*^B
a = VeKcnAo — waowso a - KCIIAI — VeURU
6 = — VcKc — nsa — WHO (k)
c = -I h = — 2Xc»iAo — ~ 0)
c =Z.
Second Order:
A+B^R+S A+B^2R
a = KcUAonBo — WBOWSO a = ifcriAoiBo i l o
(m) 6 = —KCUM — KcUm ~ 4n.Bo (n)
— nso c = iiTo - 4
c = Kc - 1
24 ?± K + iSr 2A?=i2i?
a = X„»ô " B O
J. or^ '*'"' h = —2KcnAii — 2nBo (P)
6 = -iKcUAD — (o)
c = X, - 1
c = Kc . 1
In using rate equations care must be exercised to use consistent units

throughout. Thus, in the equations of^Table LII it is important that
the unit of volume on which the reaction rate is based is the same unit of
volume on which concentrations are based and that the value of kc is
consistent with this volume.
Illustration 4. Sulfuryl chloride vapors are heated in a closed vessel for 30 min at
610°F and 1 atm initial pressure. The dissociation of SO2CI2 to SO2 and CI2 is a
reaction of fiie first order and not influenced by the walls of the vessel. The reverse
reaction is negligible. The reaction velocity constant h is 0.00132 per min.
Calculate:
(o) The percentage decomposition of SO2CI2.
(6) The weight of SO2 formed from 1 cu meter of SO2CI2 at the initial conditions in
30 min.
(c) The time required to decompose 90 per cent of the SO2CI2.
(o) Let njLo = kg-moles of SO2CI2 initially present in the constant volume of
reacting system
XA = kg-moles of SO2CI2 transformed in time r
From Equation (a), Table LII,
1 , nAo
T = —In-
fo nAo — XA
0.00132 WAO - XA
1 - — = 0.9612
7140
Hence the percentage decomposition of SO2CI2 = (1 — 0.9612)100 = 3.88

per cent
1 (273)
(6) SO2CI2 initially present per cubic meter = = 0.0205 kg-mole
ZlAl (593)
Weight of SO2 formed = (0.0388) (0.0205) (64) = 0.0512 kg
(c) For 90 per cent decomposition, XA = 0.9»4o
I n - ^ ^ ^ = 0.00132T
O.lnô
Hence, T =? 1740 min.
Illustration 6. The decomposition of nitrous oxide proceeds as a second-order
reaction, 2N20-»2Nj + Oj, The reverse reaction is negligible. At SQS'C the value
of kc is given by the International Critical Tables as 977 where concentrations are
expressed in gram-moles per cubic centimeter and time in seconds'. If pure N2O at an
initial pressure of 1 atm is heated at 895°C in an autoclave of constant volume, calcu-
late the time required for 90 per cent decomposition.
In Equation (b), Table LII, 7 . = 22,410 (1163/273) = 95,500 cc
UAti = 1.0; XA = 0.9
/95,500\ / 0 . 9 \ „„„
In Illustrations 4 and 5 it was assumed that the rate of the reverse

reaction was in all cases negligible. Where this is not the case the net
rate of reaction is obtained as the difference between the rates of the
forward and reverse reactions. In Table LII are classified the differ-
ential and integrated rate equations for several types of simple reversible
reaction at constant temperature and volume. For a reversible reaction
which is first order in both directions volume terms do not enter indicat-
ing that the extent of conversion in a given time is independent of pres-
sure and volume. The integrated equations for reversible reactions
which proceed with no change in number of moles apply to all isothermal
conditions whether at constant volume or constant pressure, whether
batch or flow types.
CHAP. XVIII] NONFLOW REACTIONS 827
Illustration 6. The hydrolysis of methyl acetate is a reaction of the second order,

proceeding according to the equation:
CH3COOCH3 + HjO £5 CH3COOH + CH3OH
This reaction is catalyzed by the presence of hydrogen ions. In a normal solution of
HCl at 25°C the velocity coefficient of the forward reaction is kc = 0.0001482, and
that of the reverse reaction isfee= 0.000677 liter/(g-mole)(nun). The initial con-
centration of such a solution is UAO = 1.151 g-moles of methyl acetate per liter and
nso — 48.76 g-moles of water per liter. No alcohol nor acetic acid is initially present.
On the basis of 1.0 liter of solution, calculate:
(o) The moles and percentage of ester transformed in 60 min.
(b) The moles and percentage of ester transformed at equilibrium.
(c) The moles and percentage of ester transformed in 60 min if the reverse reaction
and the change in concentration of water are neglected. This in effect is treating the
hydrolysis as though it were a unidirectional first-order reaction.
(o)_In Equations (h) and (m) of Table LII:
^, k, 0.0001482 „ „
kc 0.000677
kc = 0.0001482 liter/(g-mole) (min)
TiAQ = 1.151 g-mole per Uter
nBo*= 48.76 g-mole per liter
TlRO = 0
nso 0=
XA =
g-moles ester converted
Vc =
1.0 Uter
o =
KcnAoriBo - JiBonso = (0.2189) (1.151) (48.76) = 12.29 (g-moles) V (liter)*
6 =
-KcUA'i - KCUBO - WBo - nso = (0.218g;(-1.151 - 48.76) = -10.93
g-moles per Uter
c = Z» - 1 = -0.7811
g = 4ac - 6« = (4) (12.29) (-0.7811) - (10.93)^ = 157.9 (g-moles)V(Uter)'
V—g = 12.57 g-moles per liter
Substituting in Equation (h) gives
(0.2189) [2.303 T /2(-0.7811)a:A - 10.93 - 12.58\
~ 0.0001482 [12.57 L °^ V2(-0.781 Izx) - 10.93 -t- 12.58/
/ ( - 1 0 . 9 3 + 12.58)\11
\ ( - 1 0 . 9 3 - 12.58)/Jj
^ (-1.562x4 + 1.65)
or
2.712XA = 1.099
XA = 0.405
Percentage conversion of ester = (0.405/1.151)(100) = 35.2 per cent
(6) At equiUbrium:
^- - =0.219
(1.151 -x^)(48.76 -XA)
XA = 1.04
Percentage conversion = (1.04/1.151) (100) =90.5 per cent
(c) If the reverse reaction and removal of water are neglected, the differential
rate equation may be written as (driA) KTIA) = (Jcnadr) / (Vc) or
dxA knBodr
UAO — XA VC
Integrating yields
, UAO kUBoT
In =
, riAo — XA VC
1 151
In ^ = (48.76) (0.0001482) (60)
(1.151 — XA)
XA = 0.405 g-moles
/0.405\
Percentage conversion = I 1(100) = 35.2 per cent
From the results of parts (a) and (c) of Illustration 6 it may be noted
that little error is involved in assuming that the reaction follows a first-
order mechanism. This results from the fact that the water is present
in such great excess that changes in its activity or concentration are
negligible. Reactions of higher order under conditions which permit
their treatment as first order are termed pseudo-first-order reactions.
Many systems of industrial importance are of this type and may be
handled with satisfactory accuracy on the basis of assumed pseudo
mechanism. This procedure frequently results in considerable simphfi-
cation of calculations but must be used with great care and recognition
of the possible errors involved.
Simultaneous Reactions. The simplest case of simultaneous reac-
tions is that of an irreversible first-order reaction in which the reactant A
simultaneously reacts in two ways. Thus,
(1) A-^R
(2) A-^S
The over-all rate of decomposition of A is the sum of the rates of the two
reactions which may be considered separately. If yA and ZA represent
the moles of A decomposed per unit mass of feed by reactions 1 and 2,
respectively, and XA = VA + ZA, from Equation (45),
dyA = kciiuAo — yA - ZA) dr (59)
dzA = kciiuAo — yA — ZA) dr (60)
dxA = dyA + dzA = (^ci + A-c2)(nAo — x^ dr (61)
Combining (59) and (60), where yA and ZA are 0 at T = 0 gives
VA '^ci XA .„„,
— = — or 2/A = (62)
ZA kci ' 2i£ 4_ 1
A-ci
CHAP. XVIII] CONSECUTIVE REACTIONS 829
By integrating (61) an expression is obtained for the time r corresponding

to a total conversion XA in a reaction a t constant temperature and
volume.
^ = (r4¥-)ln-^^^ (63)
\hi + Keif riAa — XA
Equations (62) and (63) permit complete calculation of the decomposi-
tion products formed at any time. *
Similar simple integrated equations can be developed for simultaneous
second- or third-order irreversible reactions in which the reactants for
both reactions are the .same and they combine in the same proportions.
However, in the general case where the simultaneous reactions involve
different reactants, complex simultaneous equations are obtained which
are generally best integrated graphically.
Consecutive Reactions. The simplest example of consecutive reac-
tions is the successive unimolecular conversion of reactant A, first to
product R and then to S.
A-^R-^8 (64)
Where a group of simultaneous or consecutive reactions involves no

reactant which is common to all reactions, it is not possible to set up
rate equations in terms of the extent of conversion of a hmiting reactant,
as was done in the development qf the preceding rate equations. Instead
the equations must be expressed in terms of the concentrations of the
individual reactants present at any time. Thus, for the reaction of
Equation (64),
— — = KcinA (65)
dnR
= koitiR — kcinA (66)
~d7
dns
17 = kciUB - (67)
where UA, n^, ns = moles o{ A,R, and S present per unit mass of original
charge
An expression for UA in terms of T and UAO, is obtained by rearranging
Equation "(a) of Table LII. Thus:
UA = nôe"*"'' (68)
- ^ = h^riB - hiTiA^-'"'^ (69)

(XT -
830 HOMOGENEOUS REACTIONS [CHAP. XVIIl
Solution of this linear differential equation gives^-*
Equations (68) and (70) permit evaluation of the quantities of A and R

present at any time T. The corresponding quantity of S is obtained from
the material balance.
ns = riAo — UA — UR (71)
If the reactions considered are reversible to a significant extent, Equa-
tion (64) is written
id ka •
A^R^S (72)
Then, from Equation (43),
TAV = - - — = hilnA - -^j (73)

UT \ Kcl/
dns ( ns\ . ( ns\

- —= ,.,[nR-—yjc..[nA--^
= riB \kci + ^ ) - -^ riAkoi (74)
Differentiating (73) with respect to time gives

d'ÛA . , dnA , kcl driB ,„^^
rfr^ dr Kcl dr
substituting E q u a t i o n (74) in (75) yields
d^'UA , duA kcl / , , kci\ , kcikci

~d!T2 dr Kci\ KciJ KciKci
+ ^nA (76)
From Equation (73), JIB may be expressed in terms of n^, and, from a
material balance,
ns = riAo — riB- riA (77)
' I. S. and E. S. Sokolnikoff, " Higher Mathematics for Engineers and Physicists,"
McGraw-Hill Book Company (1934).
' T. K. Sherwood and C. E. Reed, " Applied Mathematics in Chemical Engineer-
ing," McGraw-Hill Book Company (1939).
CHAP. XVIII1 CONSECUTIVE REACTIONS 831
If values of ns from (79) and nn from Equation (73) are substituted in

(74),
UAo = 0 (78)
This second-order differential equation may be solved^'' to give
n, =.'C,e-^- + Cê-^- + - ^ p - (79)

AcliVc2 •C'2
for the boundary conditions that T = 0, n^ = 1, n« = 0, ns = 0,

—— = 0 where,
dr
Ci =
:lA"<;2
Kci-
JC2 =
-02
1
-^1 + V^f- -4:E2

A = 2
-^x- Vfi!- ~4E2
E, = ka (l +
^)+^"^0+^; )
Ei
Although the reaction under consideration is of a simple type, Equa-

tion (79) is complex and difficult to apply. For the more complicated
combinatiqns of higher-order consecutive and simultaneous reactions
which are commonly encountered in industrial operations analytical
integration may not be possible or practicable. For these problems,
which are generally complicated by varying reaction conditions, graphi-
cal or trial integration procedures are preferable, as discussed in follow-
ing sections.
Flow Reactions. The time of reaction in a flow system may be

visualized as the time required for an infinitesimal mass of charge to pass
through the reactor in an imaginary compartment which moves with the
stream and adjusts itself to the temperature, pressure, and composition
of the stream. Although in a nonflow process time of reaction is a
significant and convenient variable which can be directly measured by a
chronometer, it is evident that a flow reaction presents an entirely
different problem in which time of reaction cannot be directly measured
and is not a convenient variable for correlation of rate data. Much con-
fusion has resulted from improper attempts to interpret data from flow
systems on the basis of reaction times which are frequently calculated
on the basis of unsoimd approximations. A rigorous treatment, with all
reference to time of reaction omitted, is both more simple and more
accurate whenever only a single fluid phase is present. This treatment
assumes that diffusion in the direction of flow is negligible and that the
stream of reactants moves progressively through the reactor without
longitudinal mixing. This condition is closely approached in all com-
mercial and properly designed laboratory reactors except in relatively
large-diameter adiabatic reactors, such as are discussed on page 864.
The pertinent problem in the design or analysis of a flow process is the
relationship between the reactor volume and the feed rate. This may be
arrived at by consideration of an elementary reactor volume dVr having
the cross-sectional area of the reactor and a length dZ in the direction
of flow.
Under steady-state conditions the number of moles of A converted in
this elementary section per unit time is constant and equal to rdVr.
The material balance of the section requires that this conversion equals
the reduction in the nmnber of moles of A in the stream passing through
the section. Thus,
rdVr = F dxA (80)
where
r = rate of reaction, moles of A converted per unit volume of react-
ing system per unit time
Vr = reactor voliune occupied by reacting fluid system
F = reactor feed rate, mass per unit time
XA = moles of A converted per unit mass of reactor feed
The reactor voltune required for a specified conversion XA is obtained
by integration of (80) for a constant feed rate F.
CHAP. XVIII] FLOW REACTIONS • 833
The reaction rate r for a feed of fixed composition processed at constant

conditions of temperature and pressure is a function only of conversion x.
Thus, under these conditions Equation (81) expresses a relationship
between Vr/F and x which is independent of reactor size, shape, or feed
rate as long as longitudinal diffusion is negligible. For this reason it is
applicable either to following the progressive changes in composition
through a reactor operating at steady-state conditions or to correlating
data from a series of runs in which Vr/F is varied by varying the feed
rate to a reactor of fixed volume operating at constant temperature and
pressure.
The integral of Equation (81) may be evaluated analytically for simple
reactions under constant pressure and temperature. For example, the
rate of a unidirectional first-order reaction in an ideal solution of gases is
given by Equation (28). Combining this with (81) gives
T^ _ 1 P'A. Ut dXA
1 p!î^ (82)
F irk Jo VAfiA
rk Jo VAnA
where,
Tfit = total moles of reacting system per unit mass in the section dVr
nA — moles of reactant A per unit mass of reacting system in the sec-
tion dVr
Before the integration can be completed nt and UA must be expressed as
functions of XA. If 5 is the increase in the number of moles in the reacting
system per mole of reactant A converted.
nt- nail -\ -] = no(l + (Â) (83)

\\ nno/
o/
where
no = total moles of initial feed per unit mass, or the reciprocal of
the average molecular weight of the feed
5 = increase in the number of moles of the system per mole of A
converted = [(r + s • • •) — (a + 6 • • •)]/« for the general reac-
tion of Equation (5)
CO = 5/no, a constant for a particular system
Combining (53), (82), and (83), gives
Vr _ no rÂ (1 + MA) ,
r irkvAô {nAa — Â)
Integrating where VA is constant yields
•^ = ~~-\ -UXA + (1 + unAo) hi —— (85)

834 HOMOGENEOUS REACTIONS (CHAP. XVIII
By this procedure the integrated equations summarized in Table LIII

were developed. As written, the equ"ations are directly applicable to
systems of ideal gases. If the term v is omitted they are applicable
\
TABLE LIII
HOMOGENEOUS REACTION RATES AT CONSTANT PKESSUHE AND TEMPEBATUBE
Flow Reaclions
aA + bB = rR + sS
r = moles of component A converted per unit time per unit volume

riA = moles of A unconverted per unit mass of feed
no = moles of feed per unit mass of feed
«( = moles of reacting system per unit mass of feed = 'no(l + USA)
Vr = volume of reactor occupied by reacting system
,F = feed rate, mass per unit time
riAo, nao, nso, nso = moles of components A, B, R, S, respectively, per unit mass of
entering feed
XA = moles of A converted per unit mass of feed
»A = nAQ — XA
a = ratio of moles of B to moles of A in stoichiometric equation
V — moles of product in stoichiometric equation
r -V s — a — h
S =
a
S
0) = —
no
k = reaction velocity constant
K = equilibrium constant
jT = total pressure
General rate equation:
dVr
Irreversible Reaclions
First Order: A~*vR; 5 =v — 1
rii t fcir L "Ao — XAJ
Second Order: General Solution:
^^fXH.A +J - - ^ ^+Mln(^^^)+iVln(-^^^)l (b)

fcff* L nAo(nAo — XA) \nô — XA/ \nBo — XAJA
CHAP. XVIII] FLOW REACTIONS 835
v-2
Case 1. 2A -^vB; 5 = —-—
H = 0,^
' =< - ) '
J ~ (1 + uriAo)^ (o)
M = —2M (I + uriAo)
AT = 0
Case 2. A+B-^vR «.= » ' - 2

fl' = co2
J =0
^j (I + conô)' (d)
^40 ~ n s o
„ (1 + UHBo)'
Third Order: General Solution:

„3
XA
HXA+J + M'In' + A'^ In •
F k%^ L flAnfjlAO — XA) riAo — XA nBo — <TXA
UcO xx(2nô — XA)'\
+ OIn •\-P (e)
<r =1
r - 3 H = -a,» iV = 0
Case 1. ZA-*vR S=
3 1 (f)
/ = -3&)(1 + wnô)* Q = 0
r-3
Case2. 24+B-»>'if; «=
'r=h
2(1 +2anBo)'
H = -2a.' Ar = -
(2»B0 — UAH)'
•..(=?f=)l 2(1 + iiUAoY

2nBi> — UAO
Q=0 (g)
2(1 + o)nô)H2wrexo — GoiUBti — 1)

Af = — — P = 0
, I (2nBti - TlAor
Case 3. A + B + C -^ vR; 6 = »- 3
"^ ~ \ w! /
When nô = nao, use Equation (g).
Whenriô = nso = nco, use Equation (f).

<r = 1 (1 + wnso)'
N =
H = -co' (UAO — nso) (jiBo — nco)
J =0 (1 + tonco)' (h)
(1 +<inAo)' (iiAo— nco) (nso — nco)
M =
(nô ~- nBoJinAo — nco) P = 0
Reversible Reactions
First Order: A^R (w = 0)
Vr _ Kno , P
-In • (i)
F ' (1 + K)k7r '" P -XA
KnAo — nna
where P =
1 +^
Mixed First and Second Orders: T h e constants o, 6, c, and g are defined in each case,
g = 52 _ 4oc, and g < 0
General Case.
•=• = -7— s^ + 17 In ( )
V -? L(2cx4 + 6 + V - 2 ) (6 - V-g)J
Forg<0
ca(bo> — 2c)
?7 = y = ^ _ (1 + 6t7) (k)
2c2 c
Casel, A-^R-^B
a = irTiKonso — KnanAn
;T-5- ) 6& = XX«o(l
n o ( l — oinAo)
0)1 + 7r(nso + nso) (1)
n, Kn\ / C = KlMo + X
Case 2. 4 + B <=» B
.J . aa — noJifio
= WoWfio — — iiTrnxoiBo
KirnAonBo
= fc I ; -r— ) b6 = Kir +
= KTriuAo + UBD) + ni)(l + omBo)
{UAO (m)
\ n'. Km/ c = wwo — ^^ XTT
Case's. A^2R
r = fc ( —-— ) 6 = KnoCl — toriAo) + 4ffnji (n)

\ni Kn\J rr JL A
Case 4. 2.4 ?=± B

a = noTiso — Kimô
= k ( —^ —— 1 6 = ax'imô
lAo ++ n o ( i + wnô) (o)
c = ôirio — Kir
CHAP. XVIII] FLOW REACTIONS 837
Second Order: (w = 0). The solution is given by equations (i) and (k). To deter-
mine the constants a, b, c, use the negatives of the values of a, b, c given in Table LII
for the similai cases as follows:
A +B :^B + S Equations (m) of LII
A +Bi±2R Equations (n) of LII
2A^B +S Equations (o) of LII
2A ^ 2R Equations (p) of LII
or use the solution (h) of LII directly, with +.he values of a, b, c given by equations
(m), (n), (o), or (p), Table LII, and with r in (h) replaced by — ^ and kc = k (R'T)'.
to liquid systems forming ideal solutions. For nonideal solutions the

products of the fugacity or activity coefficients must be included as in
Equation (84).
The equations of Table LIII may be used for the evaluation of reaction
velocity constants from experimental data or for the design of reactor
systems, at constant temperature and pressure, without the considera-
tion of reaction time. Velocity constants evaluated in this manner from
flow data may be used for the design of batch processes, or, conversely, •
velocity constants derived from laboratory batch experiments may be
used for the design of flow processes. If for any reason it is desired to
calculate the time of reaction in a flow process a separate integration of
Equation (56) under suitable conditions of restraint is required. In the
special case of a reaction at constant temperature and pressure which
produces no change in the number of moles of the system, T = VrPf/F, and
the equations of both Tables LII and LIII are apphcable.
It should be noted that Vr is the volume of the reactor actually occu-
pied by the reacting fluid. Where the inside of a reactor is partly occu-
pied by inert space, then Vr = BV'r where Vr is the total inside volume
based upon over-all inside dimensions, and B is the fraction of this space
available to and occupied by the reacting fluid.
Illustration 7. It is desired to decompose sulfuryl chloride into sulfur dioxide and
chlorine in a flow process. From the data of Illustration 4 calculate the reactor
volume'required to produce 90 per cent decomposition when operating at a constant
temperature of 610°F and at atmospheric pressure with a feed rate of 50 lb per hr. It
may be assumed that the system follows the ideal-gas law.
Solution: Basis: 100 lb of SO2CI2
From Egu5,tion (49),
fc = kcKR'T) = (0.00132) (60)/(0.729) (1070)
= (101.5)(10^) lb-mole/(cuft)(hr)(atm)
In Equation (a), Table LIII,
s in
tiAo = no = 100/135 = 0.741 lb-mole; „ = — = -£— = 1.35
no 0.741
838 HOMOGENEOUS REACTIONS ICHAP. XVIII
XA = (0.9) (0.741) = 0.667; S = 1.0; UXA = 0.9

on ' "^» 0.741
1 + uriAo = 2.0; = = 10
Substituting,
Vr 0.741
[-0.9 + 2.0 In 10] = 27,100 (cu ft)(hr)/100 lb
F (0.1015) (10-') (1)
For a plant to process 50 lb per hr of feed, F = 50, and
Vr = (27,100) (0.5) = 13,600 cu ft
This large reactor volume would be required for the uncatalyzed decomposition at this
temperature.
Illustration 8. Iodine may be produced by the decomposition of hydrogen iodide
according to the following reaction:
2HI(g)±5H2(g)+Ij(g)
The reaction velocity constant K of the forward reaction at 393 °C and 1 atm is given
in the International Critical Tables as 0.000588 (22.41 liters) /(g-mole) (min). The
corresponding equilibrium constant K^ is 0.0167.
A decomposition process is to be carried out under these conditions in a flow-type
reactor with a feed of 2.0 Ib-moIe of HI per hour. The unconverted hydrogen iodide
is separated and recycled. Ideal-gas behavior may be assumed.
Calculate:
(a) The percentage deomposition of the HI at equilibrium.
(b) The reactor volumes required for conversions of HI equal to 80 per cent and
90 per cent of the equilibrium value.
(c) Compare the net production of iodine in pounds per hour per cubic foot of
reactor volume for the two cases of part (6).
Since this reaction is at constant pressure and does not involve a change in the total
number of moles in the system, the design of the flow process is most conveniently
handled by means of the integrated equations for constant-volume operations in
Table LII. The reactor volume is then obtained from the time of reaction by the
relationship V, jF = VfT, where V, is the reactor volume, F is the feed rate per unit
time, and Vt is the specific volume of the reacting mixture.
Basis: 1.0 lb-mole HI x = moles of HI converted
fc = 0.000588 UAO = 1.0
K, = 0.0167 riA ^1 -X
F = 0.0333 lb-mole HI per min n^, = n/^ =
2
(a) At equiUbrium:
(«g,) (nxt)
Kc = ; = T7, TT = 0.0167
nh 4(1 - xy
X = 0.205
(fe) At 80 per cent of equilibrium conversion, x = (0.8) (0.205) = 0.164. This
reaction is a reversible second-order reaction of the type represented by Equation (o),
Table LII.
riAo = 1.0; a = KcJiô = 0.0167; c = Kc - i =- -0.2333
CHAP. xvin] SPACE Vl )CITY 839
T 666
njjo = 0; 6 = —2KcnAo = -0.0334;
^ ' = 2^3 = i7-3 = 2 - ^ (^^^^" '']
per lb-mole
nso = 0; 6' = 0.001166 t Vj = 5.954
g = 4oc - 6' = 4(0.0167) (-0.2333) - 0.001166 = -0.01674

V^ = 0.1294
y
From Equation (h), Table LII, if T is replaced by — ^
Vr ^ VjKc 2.303
L\2ca; + b + V-q/\b- V^/J
Vr ^ (5.954)(0.0167) (2.303) r / 2 ( - 0 . 2 3 3 3 ) (0.164) - 0.0334 - 0.1294\
F (0.000588) (0.1294) °^L\2(-0.2333) (0.164) - 0 . 0 3 3 4 + 0 . 1 2 9 4 /
, / - 0 . 0 3 3 4 + 0.1294\1
\ - 0 . 0 3 3 4 - 0.1294/J
=[(3010)(0.86) = 2590(359cu ft)(min)/lb-mole
Vr = (2590) (0.0333) = 86.2(359 cu ft) or 30,900 cu ft
Similarly, at 90 per cent of equilibrium conversion, z = 0.185.
y , = 36,500 cu ft
(c) Netrateof production of iodine in pounds per hour per cubic foot:
. (2) (126.9) (0.164)
For 80 per cent conversion „„ " = 0.00135 Qb I,)/(lir)(cu ft)
oU,9(X)
^ „ . (2) (126.9) (0.185) „„
For 90 per cent conversion— „^ ' ' • = 0.00129 (\h Ij) /(hr)(cu ft)
00,500
Space Velocity. It may be noted that all of the equations of Table

LIII express extent of conversion as a function of Vr/F, the reactor
volume per unit feed rate. A convenient unit for expressing this rela-
tionship between feed rate and reactor volume in a flow process is
the space velocity which is defined as the volumes of feed, measured at
standard conditions, per unit time per unit volume of reactor. This unit
has the dimension of reciprocal time (1/r)" which is commonly expressed
in hours.
(The space velocity has gained widespread use, because it directly
expresses the volumetric feed capacity of a reactor and is independent of
all imits__except time. However, care must be taken that the standard
conditions at which tne feed is measured are clearly designated. In
dealing with liquid-phase reactions the space velocity is commonly
based on the volume of the liquid feed measured at 60°F and is desig-
nated by the symbol Si expressed in reciprocal hours. For vapor-phase
reactions it is convenient to use the symbol S, to denote the space
velocity in reciprocal hours, based on the feed in the ideal-gaseous state

at 32°F and 1 atm. In processes in which the feed enters as a liquid but
is vaporized before reaction either method of expression may be used,
and care is necessary to avoid confusion.
From the definitions of space velocity it follows that
where V° = the volume of a unit mass of feed at the selected standard

conditions, expressed in the same units as Vr, the reactor volume.
In dealing with vapor-phase reactions a convenient simpUfication
results if F is expressed in moles per hour and Vr is expressed in normal
molal volumes. In these units
S. = y^ (87)
where
F = feed rate in pound-moles or gram-moles per hour
Vr = reactor volume in 359 cu ft or 22.4 Uters
REACTIONS OF COMPLEX ORDER

As is pointed out on page 819, reactions which from their stoichio-
metric equations might be expected to be of simple order frequently
yield rate equations of complex order. Although it is believed that any
reaction of complex order can be expressed in terms of a succession of
reactions of simple order, the resulting'expressions are frequently so
complex that it becomes expedient to employ empirical rate equations.
Such empirical equations may be used satisfactorily where the reverse
reaction is negligible and equilibriiun is not a consideration.
An empirical rate equation of complex order for an irreversible reac-
tion is developed frpm experimental data by graphical differentiation of
curves relating conversion to time or volume. The resulting reaction
rates are then empirically related to activities or concentrations. Such
rate equations may be applied to problems of reactor design by the same
procedures used for reactions of simple order.
Where reverse reactions and equilibria are involved, the problem is
much more difficult. If it is attempted to develop empirical equations of
complex order for both the forward and reverse reactions, their forms
must be such as to satisfy the thermodynamic requirements of equilib-
rium. It is difficult to accompHsh this result by a purely empirical
procedure, and it is preferable to develop a form of equation based on a
CHAP. XVIII] REACTIONS OF COMPLEX ORDER 841
simplified sequence of intermediate reactions which can be combined

into a consistent over-all equation. Such equations may be classed as
semiempirical. They are thermodynamically consistent and may be
applied at equilibrium conditions. However, because of the necessary
simphfying assumptions as to mechanism they may not be valid over
wide ranges of conditions. Where wide ranges of conditions are in-
volved it may be necessary to develop different sets of constants for
different ranges! Such difficulties are avoided if fundamentally sound
equations can be developed by consideration of the rate-controlUng steps
of the sequence of reactions actually taking place.
The formation of hydrogen bromide from hydrogen and bromine is a
classical example of a reaction of complex order which has been exten-
sively studied*. The experimental measurements indicate that the
initial rate is proportional to the first power of the hydrogen concentra-
tion and to the square root of the bromine concentration. However, a
rate equation of this form is not vahd except at initial conditions, and as
conversion proceeds the rates are much lower than calculated. This
indicates the effect of a reverse reaction, even though the over-all thermo-
dynamic equihbrium constant is very large and the reaction should go
substantially to completion at the conditions investigated.
Fundamental equations in agreement with this behavior may be
developed by consideration of the series of possible reactions given on
page 820. The fact that the initial rate is proportional to the square root
of the bromine concentration indicates that the bimolecular reaction
between bromine and hydrogen (1) is not the predominant mechanism
so it will be assumed that this reaction is slow and may be neglected.
The dissociation of hydrogen into atomic hydrogen is accompanied
by a large absorption of energy and is known to be slow at low tempera-
tures; accordingly this reaction rate will be neglected. The reaction
between hydrogen atoms and bromine atoms should require practically
no energy of activation, and its reaction velocity constant should be high.
However, the concentrations of both atomic forms should be extremely
.low, and it will be assumed that the rate of formation of HBr by this
reaction is negligible. With these three reactions eliminated, only (2),
(4), and (6) remain for consideration.
(2)' _ - Br2 ±^2Br

(4) Br-fHa ?=> HBr -1- H
(6) H-FBra ^ HBr -1- Br
* R. N. Pease, "Equilibrium and Kinetics of Gas Reactions," Princeton University
Press (1942).
These three reactions constitute a chain in which atomic bromine is-
consumed in reaction (4) and a corresponding amount is produced in
reaction (6). Since there is no net consumption of atomic bromine it is
reasonable to assume that equilibrium is maintained in reaction (2).
Then, if the chemical symbols are taken as representing concentrations,
' Br = VKciBr, (88)
where
Kc2 ~ the equiUbrium constant of reaction (2)
If only reactions (2), (4), and (8) are significant, the rates of (4) and (6)
must be equal, and r the over-all rate of formation of HBr is equal to 2r4
or 2r6, where
n = rate of removal of H2 by reaction (4)
n = rate of removal of Br2 by reaction (6)
Treating (4) and (6) as reactions of simple order, gives
n = /•c4(Br)(H2) - A-'4(HBr)(H) • (89)
re = A-o6(H)(Br2) - fr6(HBr)(Br) (90)
where ^c4 and k^ are the forward and kU and ^^ the reverse reaction
velocity constants of reactions (4) and (6), respectively. Since the two
rates are equal, Equations (89) and (90) may be equated and com-
bined with Equation (88).
A-c4VKe2Br2(H2) - /f^(HBr)(H)
= ^.6(H)(Br2) - kiyK,2BT2(HBt) (91)
Solving for H yields
hiVK^îm) + kU^K,2Br^{BBT)
/f.4(HBr) + kc,{Bv,) ^ ^
ir = ?-4 = kcyKc2Br2i'H2)
, ' ruT. ^ r^-4^g^2Br2(H2) + kU^K,,Bv2{RBv)'\
- A- 4 HBr) , / m p A -L- I. /p ^ (^^)
L k'ci (HBr) + /(•c6(Br2) J
where r = the rate of formation of HBr
Equation (93) is in agreement with the experimental data and indi-
cates that formation of HBr can appreciably affect n even though the
over-all equiUbrium constant Ki may be very large. This results from
the fact that the relative forward and reverse rates of Equatioii (93) are
CHAP. XVIII] CHAIN REACTIONS 843
dependent on the equilibrium constant of reaction (4) which is not

necessarily large.
Equation (93) is complicated but with the constants evaluated could
be used in graphical-reaction design methods without difficulty. The
evaluation of the constants and the deviations of the equation from
experimental data are shown in Illustration 10, page 854.
Chain Reactions. The hydrogen bromide reaction is a simple example
of a type of chain reaction commonly encountered particularly in the
high temperature decomposition of organic compounds and in oxidation
reactions of many types. Rice and coworkers* have demonstrated the
existence of free radicals in" many reacting systems and have proposed
various reaction mechanisms involving the continuous formation and
removal of these substances. Thus, in the high-temperatxu-e decomposi-
tion of ethane the controlling reaction is beUeved to be the formation of
methyl and ethyl radicals and atomic hydrogen which may then react
with each other and with undecomposed ethane as follows:
C2H6 *=^ H + C2H6
C2H6 <=^ 2CH3
CaHe + CH3 ^ CH4 + C2H5
^ C2H6 + H <=^ C2H5 + H2
C2H5 <=i C2H4 + H
H + C2H5<=tC2H4 + Ha
H + CH3 ?^ CH4
CH3 + C2H5 ?=^ C3H8
2C2H6 ^ C4H10
According to one interpretation of the free-radical theory of chain

reactions the concentrations of free radicals in the reacting system reach
constant values after an immeasurably short period. The rates of the
other reactions and the rates of formation of ultimate products are then
dependent on the concentrations of free radicals as well as those of the
other reactants.
An alternate approach to the problem is that used in the development
of Equation (93) in which the atoms which are responsible for the chain
reaction are assumed to exist in an equilibrium concentration which
varies as the reaction progresses. This method has the advantage of
leading to thSrmodynamically consistent equations.
It must be recognized that in the present state of knowledge determina-
tion of the mechanism of a chain reaction and the development of rate
equations from it must be based on extensive experimental data. With
such data available the analysis becomes a problem of testing various
»F. 0. Rice and K. F. Herzfeld, J. Am. Chem. Soc, 56, 286 (1934)
844 HOMOGENEOUS REACTIONS [CHAP. X V m
plausible mechanisms until one is found which is in satisfactory agree-

ment with the data. The selection of "probable mechanisms may be
effectively guided by the theory of absolute-reaction rates. ^
Convincing proof of the existence of free radicals has been furnished
by various experimental methods, and the chain-reaction theory is
supported by the action of inhibitors which greatly reduce the rate of a
reaction even though present in minute amounts. For example, small
traces of nitric oxide are found to reduce markedly the rate of decom-
position of ethyl ether. According to the chain-reaction theory such an
inhibitor functions by reducing the concentration of free radicals in the
system. An inhibitor initially present in a reaction may be consumed
by irreversible reactions and finally disappear. Such a situation fur-
nishes the explanation of the frequently observed induction period
during which a reaction proceeds at an abnormally low rate and then
abruptly increases to its normal value. Similarly, there is evidence that
the rate of chain reactions can be increased by the presence of small
amounts of accelerators which serve to increase the free-radical concen-
trations.
Another experimentally observed phenomenon which supports the
chain mechanism is the retarding effect of solid surfaces on many nor-
mally homogeneous reactions. It is frequently found that the rate of
reaction is increased merely by enlarging the reactor with a resultant
reduction of solid surface per unit volume. This effect is explained by
the adsorption of free radicals on thei solid surfaces which reduces the
effective free-radical concentration in the homogeneous phase.
INTERPRETATION OF LABORATORY RESULTS

The laboratory and pilot plant data on which reactor designs must be
based are in general of three types:
1. Measurements of composition as a function of time in a batch
reactor of constant volume at a substantially constant temperature.
2. Measurements of composition as a function of feed rate to a flow
reactor of constant volume operated at constant pressure and sub-
stantially constant temperature.
3. Measurements of composition as a function of time in a variable-
volume batch reactor operated at constant temperature and substan-
tially constant pressure.
The third type of data is much less common than the other two, and
the experimental technique is more difficult. A variable-volume reactor
generally depends on varying the level of a confining liquid such as a
molten metal in order to maintain constant pressure. This method is
CHAP. XVIII] COMPLEX REACTIONS 845
desirable where the volume of the reacting system is difficult to calculate

as when a change of phase accompanies the reaction. It has the advan-
tage of yielding positive rate measurements and at the same time data
on the specific volume of the reacting system.
Data of the second type are generally the most dependable and simple
to obtain. This method has the advantage of direct apphcability to
flow-type reactors. Data of the first type at constant volume are
satisfactory except where added moles, of gas result from the reaction., In
such cases the varying pressures make the data difficult to interpret for
complex systems.
Simple Reactions. 'It has been pointed out that the order of a reac-
tion cannot ordinarily be predicted from its stoichiometry but must be
determined experimentally from the relationships of reaction rate to
pressure and composition. For reactions proceeding at constant tem-
perature this determination can be based upon either differential or inte-
grated rate equations such as appear in Tables LII and LIII. The
integrated forms are particularly convenient for the relatively simple
types of reactions to which they are applicable.
If data at constant volume and temperature are available the equa-
tions of Table LII are used. If one of these equations properly repre-
sents the experimental data a plot of r against the variable expression
on the right side of the equation will yield a straight fine having a slope
equal to l/kc. Equations based on plausible mechanisms are tested in
this manner to confirm the mechanism and order of reaction. For flow
reactions the same procedure is foUowed, the equations of Table LIII
being used and Vr/F oj: 1/Sv being plotted instead of T.
In any case it is desirable to secure data for a wide range of concen-
trations and conversions for positive determination of the proper rate
equation. Such data are obtained by varying the initial compositions of
reactants and in gaseous systems by operation at different pressures.
From determinations of reaction velocity constants at two temperatures
in this manner, the entropy and enthalpy of activation are calculated by
the method shown in Illustration 2, and the general equation for the
reaction velocity constant in terms of temperature is estabfished.
Complex Reactions. Many processes of commercial importance in-
volve simultaneous and consecutive reactions to such an extent that the
development of integrated rate equations for the over-all reactions is
impracticable. The analysis of data on such processes may be handled
by determining differential rates of reaction r by graphical differentiation
of curves relating conversions to time for nonflbw reactions or to Vr/F
for flow reactions. The differential rates of the individual reactions
may be determined in this manner from data on the over-all process.
T h e individual rates are then correlated with t h e corresponding concen-

trations or activities t o determine t h e proper rate equations. T h i s
procedure is demonstrated in t h e following illustration:
Illustration 9. Diphenyl is produced by the pyroiytic dehydrogenation of benzene

according to tlie reaction:
(1) 2CeH6 ?^ C,2H,„ + Hj
Triphenyl is also formed by the secondary reaction:
(2) CcHe + CinHio ?± CisHi4 + Hj

Reliable rate data are not available on these reactions, but the following approximate
data, based on estimates prepared by L, S. Kassell, were presented by Murphy, Lamb,
and Watson'.
In a laboratory investigation of the pyrolysis of benzene to diphenyl, liquid benzene
is vaporized, heated rapidly to the desired reaction temperature, and fed continu-
ously through a reactor tube which is maintained at constant temperature and
pressure. Under experimental conditions no appreciable decomposition occurs in the
preheater. The reactor tube is 37.5 in. long and 0.5 in. in diameter. The product is
withdrawn continuously from the reactor and cooled rapidly to condense the vapor
and to avoid further decomposition, and the liquid product is analyzed for benzene,
diphenyl, and higher polybenzenes which are assumed to be tripbenyls. .
The compositions of the liquid product for different feed rates are tabulated in
Table A for experiments run at 1265°F and 1 atm and for one run at 1400°F and 1 atm.
From these data it is desired to develop complete reaction-rate equations.
Basis of Interpretation. Reaction rates are expressed in pound-moles per hour per
359 cu ft of reactor volume and feed rates F in pound-moles per hour. On this basis,
F IVr = Sv in accordance with Equation (87). The following symbols are used to
denote the compounds involved: B = benzene, CeHe; D = diphenyl, C12H10; and
T = triphenyl, CisHu.
, 37.5ir
Yr = volume of reactor = ,• = 0.004263 cu ft
Since the density of liquid benzene at 60°F is 879 g per liter, if F' is the feed rate in
liters per hour.
p = ^'^^^ = 0.0248^' lb-moles per hr

(454)(78) ^
and
S . = (-°:°?^^^^2(35?)=2089F'reciprocal hr
0.004263
The complete composition of the product from each experimental run is calculated
from the stoichiometry of the reactions. A typical calculation based on 100 lb of
liquid products, is shown in Table B.
' G. B. Murphy, G. G. Lamb, and K. M. Watson, Trans. Am. Inst. Chem. Engrs,
34, 429 (1938).
CHAP. XVIII] PYROLYSIS OF BENZENE 847
TABLE A
PYROLYSIS OF BENZKNE AT 1 ATM PRBSSUBB
F' = feed rate of liquid benzene in liters per hour (measured at 60°F)
Percentage Composition of Liquid Product by Weight
F' Per cent B Per cent D Per cent T

At 1265°F
0.846 83.0 14.6 2.4
0.423 70.8 22.4 6.8
0.282 62.5 26.2 11.3
0.212 56.8 27.8 15.4
0.14^ 50.3 29.2 20.5
0.121 48.6 29.4 22.0
0.106 47.4 29.4 23.2
0.0282 44.7 • 29.4 25.9
0.0141 44.7 29.4 25.9
1
At 1400°F
2.75 84.3 13.7 2.0
TABLE B
COMPLETE ANALYSIS OP REACTION PRODUCTS
F' = 0,846 liter of liquid benzene per hr

Benzene Diphenyl Triphenyl Hydrogen Total-
Molecular Weight 78.1 154.2 230.3 2.016
Composition of liquid prod-
uct per cent by weight 83.0 14.6' 2.4 100
Pound-moles per 100 lb of liq-
uid product 1.0627 0.0947 0.01042 .... 1.16782
Equivalent pound-moles of
benzene in feed 1.0627 0.1894 0.03126 1.28336
Pound-moles of H2 formed.... 0 0.0947 0.02084 .... 0.11554
Pound-moles of product per
pound moje of feed 0.8281 0.0738 0.0081 0.0900 1.000
ras no riT UH tit
These calculations are summarized in Table C as the compositions of the products

leaving the reactor expressed in moles per mole of feed.
In Fig. 168 values of UB, UD, nr and ng are plotted against the reciprocal of space
, velocity. Ftom the slopes of these curves the differential reaction rates and reaction
velocity constants at 1265°F are determined by meansof the following rate equations
for reactions (1) and (2) in which the fugacity coefficients are assumed to be 1.0 at the
low pressure involved:
1.0 r 1 1 ... 1 , 1 1
0.9
0.4
gO.7 \ 1 -
y ^ n r
yC Wr+Mu
f 0.6 0.3
o
y' ^^^-^..^^B _ 01
0.443^ M
0.2
WH D.322 -<-
^0.3
II .•sS
8 0.2 0.1
nv' S'a
Ui 0.1476*
0.1 .^0.087 i
0 r—-t""^ 1 1 1 1
0 10 20 30 40 50 60
ixlO*
S.
FIG. 168. Pyrolysis of Benzene.
TABLE C
SUMMARY OF COMPOSITIONS OF TOTAL PRODUCTS
F' — feed rate, liters liquid benzene per hour (measured at 60°F)
Sv = 2089i^' = gas space velocity at 32°F (1 atm.), in reciprocal hours
F = 0.0248F' = feed rate, pound moles per hour
n = moles per mole of feed
TlT
S, llSviW) UB no tlT njcf
no + nr
At 1265°F
0.846 1766 5.63 0.828 0.0737
0.00812 0.0900 0.099
0.423 883 11.32 0.704 0.113
0.02297 0.1590 0.169
0.282 688 16.97 0.622 0.1322
0.03815 0.2085 0.224
0.212 442.5 22.62 0.665 0.1400
0.0519 0.2440 0.270
0.141 294.5 34.0 0.499 0.1468
0.0691 0.2847 0.320
0.121 252.7 39.7 0.482 0.1477
0.0740 0.2960 0.337
0.106 221.2 45.2 o.4ro 0.1477 0.0781 0.3040 0.346
0.0282 68.9 169.7 0.3220 0.370
0.443 0.1476 0.0870
0.0141 29.47 339.3 0.3220 0.371
0.443 0.1476 0.0870
CHAP. XVIII] PYROLYSIS OF B E N Z E N E 849
At 1400°F
2.75 5750 1.735 0.841 0.0695 0.0068 0.0830
Ti. =
d{llS,} n\
dnD2
rj = —
dns
ri + rj (c)
~d{llSr)
dno Ti
5 ( 1 7 ^ = -'•= + 2 (^^
driT
where
ri, rj = rates of reactions ( l ) a n d (2), respectively, lb-moles/(hr) (359cuft)
ki, ki = reaction velocity constants of reactions (1) and (2), lb-moles/(hr)(atm)*
(359 cu ft)
Sv = gas space velocity at 32°F and 1 a t m (1 /hr)
UB, no = moles of benzene and diphenyl, respectively, per mole of feed
nt = total moles per mole of feed = UB + TID + nr + UH
a- = total pressure, atmospheres
Ki, Ki = equilibrium constants of reactions (1) and (2), respectively
Values of and ,,, ,„ . are obtained by graphical differentiation of the

0,(1/bv) a{llii„)
riB and nr curves of Fig. 168 at various space velocities. These values are sum-
marized in columns 3 and 4 of Table D .
Values of the equilibrium constants Ki and Ki at 1265°F for the two reactions are
obtained from the equilibrium compositions approached a t low space velocities.
Since each reaction proceeds with no change in number of moles, and fugacity coeffi-
cients may be neglected a t the low pressures employed, the values of Ki and K^ may
be estimated from the data of Table C a t the lowest space velocity.
nonii (0.1476) (0.3220)
^ ' = l i = (0.443)^ -'•''' . (S)
nTfin (0.0870) (0.3220)
^ - = ; ^ - (0H43) (0.1476) - » - ^ ^ ^ (^)
T h e reaction velocity constants fci and ki may now be obtained from Equations (a)
and (b) where ni the total number of moles of reacting system per mole of feed is con-
stant and equal to 1.0. For example, for a space velocity of 883 reciprocal hr from
Equation (a),
(146.4)
fci
= 347 lb-mole/(hr) (359 cu ft) ( a t m ) '

From Equation (b),

r, (28.5)'
= 400 (lb-mole) /(hr) (359 cu ft) (atm)»
TABLE D
StJMMABT OP CALCULATED REACTION RATES AND VELOCITT CONSTANTS
r = lb-mole/(359 cu ft)(hr)
k = lb-moles/(359 cu ft)(hr)(atm)2
s. llSr(.W) Ti ri ki k2
1766 5.63 100.0 22.1 244.2 370 372

883 11.32 44,70 28.5 146.4 347 400
588 16.97 20.20 26.2 93.6 342 411
442.5 22.62 9.30 22.0 62.60 351 442
294.5 34.0 3.18 10.1 26.36 342 370
252.7 39.7 0.962 8.1 18.12 350 406
221.7 45.2 0 0.6 1.20 338 430
58.9 169.7 0 0 0
44.25 226.2 0 0 0
Average 348 404
The constancy of the reaction velocity constants in Table D confirms the assump-
tion that the reactions are consistently of the second order. A more conclusive test
would be to operate the reaction under widely different pressures and redetermine the
reaction velocity constants. These constants are unaffected by pressure if the proper
rate equations are used and deviations from ideal-gas behavior are considered.
Comparison of the composition of the products in Table C obtained in the single
run at 1400°F and a space velocity of 5750 with the curves of Fig. 168 indicates that
at 1265°F this same composition is obtained at a space velocity of 1815. Since the
conversion is small, this indicates that the rates of the forward reactions are equally
accelerated by increase in temperature. With this assumption the rates of the for-
ward reactions at 1400°F are 5750 /1815 or 3.16 times faster than at 1265°F. Thus, at
1400°F (1033°K),
fci = (3.16) (348) = 1100 lb-moles/(359cu ft) (hr)(atm)»
ki = (3.16) (404) ^ 1276 lb-moles/(359 cu ft) (hr) (atm)»
The enthalpies of activation AHi, of the two forward reactions ase equal with the
assumption made and may be calculated from Equation (18). Thus,
AH^ = 30,190 cal per g-mole

Equations for the velocity constants of the forward reactions are completed by
evaluating the group of terms (kJhR')e ^ '^ of Equation (18). If this group is
CHAP. XVIII] PYROLYSIS OF BENZENE 851
denoted hy A, k = - j — e and, if it is assumed that v*Zm = 1.0,

V Zm
30190 ' -I
[g (1.987)(1033)J = (1100) (2.44) (10)8 = 2.68(10')
A, = (1276) (2.44)(10)8 = (3.11) (10') (lb-mole)/(hr) (359 cuft)(atm)»
-30190 -6597
Jfcivfa^ = (2.68)(10')e «^ = (2.68) (10') (10) ^
-6597
kiP^Zm= (3.11)(10»)(10) ^
In order to complete the over-all reaction rate equation it is necessary to evaluate

the equilibrium constants as functions of temperature. The best available data for
the heat capacities and absolute entropies are used to calculate the standard entropy
changes of the reactions and Ivaluate'the constant I, of Equation (XVI-10). From
the experimentally determined equilibrium constants at 1265°F the corresponding
values of AG"/Tare calculfi!Sd from Equation (XVI-6) and the constant Ig evaluated
from Equation (XVI-11). Thus, complete equations are obtained expressing
AH°, AS°, and AG° IT as functions of temperature which are consistent with the
measured equihbrium data.
Reliable data for the entropies of hydrogen and benzene (g) are given in Table
XXXV, page 701. The heat-capacity constants for hydrogen are given in Table V,
page 214, and those for benzene (g) in Table XXXIX, page 759. These latter values
are based directly on the data of Pitzer and Scott from which the entropy in Table
XXXV was obtained. The entropies and heat-capacity constants for diphenyl and
triphenyl may be estimated from the group contributions of Tables XXXIX-
XLIII. The triphenyl is assumed to be a mixture of the ortho, meta and para forms
in equal proportions. These results are summarized in Table E.
TABLE E
THERMODYNAMIC PEOPEBTIES
S298.2 a (6) (10') (C)(10»)

Hydrogen (g) 31.23 6.88 0.066 0.279
Benzene (g) 64.39 0.23 77.83 -27.16
Diphenyl (g) 98.1 -0.20 149.11 -52.25
Triphenyl (g) 128.2 1.74 214.33 -73.65
AS298.2 Aa A6(10') Ac(10«)
Reaction (1) 2B = D + H 0.55 6.22 -6.48 -h2.35
Reaction (2) B-{-D = r + ff - 3 . 0 6 8.69 -12.54 -t-6.04
From Equation (XVI-10) and the data of Table E:

Reaction (1),/. = 0.55 - (6.22) (In 298.2) + (6.48) (298.2) (10"') - (1.17)
(298.2)2(10-') = -33.05
Reaction (2), I, = -3.06 - (8.59) (In 298.2) + (12.54) (298.2) (10"') - (3.02)
(298.2)»(10-«) = -48.53
852 HOMOGENEOUS REACTIONS [CHAP. X V I H "
. From Equation (XVI-23) and the experimental equilibrium constants of Equations

(g) and (h),
(1) (AG°/r)a58 = -4.576 log 0.242 = 2.82 cal/(g-mole)(°K)
(2) (AG°/r)958 = -4.576 log 0.428 = 1.69 cal/(g-mole)(°K)
From these results the constants Ijj in Equation (XVI-11) are evaluated.
(1) /j? = 958[3.82 - (6.22 + 33.05) + (6.22) (In 958) - (3.24) (958) (10"')
+ (0.39)(958)ni0"«)] = 3363 cal per g-mole
(2) IR = 958[1.69 - (8.59 + 48.53) + (8.59) (In 958) - (6.27) (958) (lO"')
+ (1.01) (958)2(10-')] = 1475 cal per g-molo
The standard heats of reaction at 298°K may be calculated from (XVI-9):
(1) Aff298 F= 3363 + (6.22) (298.2) - (3.24)(298.2)'(10-^) + (0.78)(298.2)»(10-«)
= 4950 cal per g-mole
(2) Affjss = -1475 + (8.69) (298.2) - (6.27) (298.2)^(10-') + (2.01) (298.2)'
(10-«) = 580 cal per g-mole
The available calorimetric data indicate a heat of reaction at 298°K of 7,250 cal
per g-mole for reaction (1). This is in poor agreement with the value previously
calculated from the estimated entropy change and equilibrium data although the
difference is only 0.15 per cent of the heat of combustion of diphenyl. The dis-
crepancy may result from errors in any of the three sets of data involved. Such
inconsistencies are constantly encountered in the present state of development of
thermodynamic data and must be rationalized as well as possible. In this ease the
discrepancy in heats of reaction will not seriously affect the equilibrium or kinetic
calculations over the temperature range of interest, and the calculated values will be
used to maintain consistency with Ihe equilibrium measurements. In the design of
the reactor, to which heat must be supplied for the endothermic reaction, the calori-
metric heat of reaction may be used to insure a more conservative design.
An alternate procedure would be to accept the calorimetric and estimated heats of
reaction at 298°K and the heat-capacity equations and calculate the standard entropy
changes from the equilibrium measurements. In, general, however, as a result of
spectroscopic and statistical methods, absolute entropy data are more reliable than
calorimetric heats of reaction.
The equilibrium constants are expressed by the following equations obtained by
substituting the calculated values of IH, IS, Aa, A6, and Ac in Equations (XVI-11
and 23).
-4.576 log Ki =
3363
— + 39.27 - (14.32) (log T) + (3.24) (lO"')? - (0.39)(10-«)T» (i)
-4.576 log Ki= *
—1475
— — , + 57.12 - (19.78) aog T)''+ (6.27)(10-')r - (1.01)(10-«)r» (j)
Reaction rates in lb-moles/(hr) (359fcuft) are calculated by substituting the fore-

going values for the equilibrium constants in the following equations:
_ (2.68) (10)V [(„,,,). l(!^£i:£^] (k)
*•! ~ 8597
(10) ^ ph„n^
CHAP. XVIII] COMPLEX REACTIONS 853
(3.11) (10)V r . w , (nr.>r)(nHVg)1

ft = 6897 y{nBVB){nDVD) ^^ J Q,)
(10) ^ ;.*2„nJ
where T = temperature, degrees Kelvin
X = total pressure, atmospheres
v^ = fugacity coefficient of t h e activated complex
Zm = compressibilitj' factor of the reaction mixture
TiB, riD = moles of components B , Z ) , . . . in a reaction mixture of nt total moles
VBVD = fugacity coefficients of components B, D . . .
At moderate pressures deviations from ideal behavior may be neglected b y assum-

ing t h a t the fugacity coefficients in the numerators of Equations (k) and (1) approxi-
mately cancel the products v^Zm in the denominators.
The method of graphical differentiation demonstrated in Illustration 9

is theoretically sound but frequently is difficult to apply satisfactorily to
the scattered and erratic data which are typical of kinetic measurements.
Small changes in the shape of the curv^ drawn through such data have
relatively large effects on the measured slopes and result in widely vary-
ing rates. In such cases it may be necessary to consider the average
rates which are evaluated by differentiation as merely first approxima-
tions which are adjusted by trial until the conversion curves calculated
by their graphical integration show satisfactory agreement with the
experimental data. Methods of integration are discussed in the follow-
ing section on reactor design.
A more direct and accurate method of evaluating velocity constants
from complex integral conversion data was pointed out by Reiser and
Watson^ and Myers and Watson^ This method is applicable where the
number of moles converted per mole of feed by the reaction in question
can be established from the disappearance of some reactant or the
appearance of a product. Where several simultaneous reactions are
involved it may be necessary to evaluate successively the conversions
attributable to the individual reactions by a series of material balances,
starting with reactions which are solely responsible for the disappearance
of some reactant or the appearance of some product. For example, in
Table C of Illustration 9, page 848, at a feed rate of 0.0141, there is 0.0870
mole of triphenyl produced which establishes the conversion of reaction
(2) as corresponding to 0.0870 mole of benzene. The conversion by
' C. 0 . Reiser and K. M . Watson, Natl. Petroleum News, Tech. Sec. 38, K260
(April 3, 1946). " Principles of Reactor Design," National Petroleum Publishing
8 P . S. Myers and K. M. Watson, Natl. Petroleum News, Tech. Sec. 38, fl388
(May 1, 1946), and 38, K439 (June 5, 1946). " Principles of Reactor Design,"
National Petroleum PubUshing Company, Cleveland (1946).
reaction (1) is then 1 - 0.443 - 0.0870 = 0.470 mole of benzene.

This same procedure may be followed in more complex systems.
Once the conversion attributable to a reaction is established at several
different values of Fr/jP for a given feed at constant temperature and
pressure, the reaction velocity constant corresponding to each experi-
mental point may be established by graphical integration. Thus, if the
reaction is of the fitst order in the forward direction and of the second
order in the reverse direction, the following expression results from
combining the basic rate equation with Equation (81) and rearranging:
A-x = — I (94)
riA nBHsir
nt n^K
where
^1 = reaction velocity constant of reaction (1)
xî = moles of component A converted by reaction (1) per mole of
feed
nA,nB,ns = moles of components A, B, and S present in there acting
system per mole of feed
nt = total moles of reacting system per mole of feed
Each of the quantities in the integral of Equation (94) may be evalu-

ated for each experimental point. A curve is then plotted relating
XAI as abscissas to l/{nA/nt — riRnsT^/'n^t^) as ordinates. The area
under this curve between xî = 0 and any selected experimental value
of a;î is equal to the product of the reaction velocity constant and the
value of -KVT/P corresponding to that value of XAI- In this manner a
series of values of ki is established which may be averaged.
This same method of evaluation may be applied to reactions of higher
order, regardless of the complexity of the system if the conversion
attributable to the reaction and the numbers of moles of all reactants
and products actually present in the system are known for a series of
runs at different values of Vr/F on the same feed at a constant tempera-
ture and pressure.
Illustration 10. Interpretation of Chain-Reaction Data. In Table A are data on
the chain reaction involved m the hydrogenation of bromine gas which have been
selected from the experimental work of Bodenstein and Lind.' All experiments were
conducted at constant volume. An independent analysis of these data is reported
by Pease.'"
»Bodenstein and Lind, Z. Physik. Chem., 5J, 168 (1907).
w R. N. Pease, " Equilibrium and Kinetics of Gas Reactions," Princeton Univer-
sity Press (1942).
CHAP. XVIII] HYDROG^NATION OP BROMINE 855
ioioioio»oioioo
a.
tOOO«Ort<0500i-l
t003(N5DOC0>0tD
OOT-irt<N(NC<IN
CO
o o o o o o o o
m
I < N - * c O 0 0 O O O ' f
OOtSlMtÔO'Ô
—, e < 3 ' O c o o o ( N t o e < j t ^
3a '^ t-lNCC-^COOO'-HM
o
D3 OT-<r-i(N(N(NC<IN.
I a. _
o o o o o o o o o
l O O O O O O O O O
6 1—i f—I I—t CO ' ^
00^<OM'OC0r-lt»
o
1 n
M r~, IMa00OC<3Ot>.i-iTj(
odidoddiodid)
I o
o O Q O O O O
•*S55oc^.oow
I-H (M M CO • * t o OJ
O
O
lomomioeocoM
Q
^ (N <N CO •* t o C3
oeoootoifliNio-*
tOt-Or-HCOCOINO
O rt(N-*iotOtÔTj<
O O O O O O ' - i ' - i
dddddddd
M
856 HOMOGENEOUS REACTIONS [CHAP. X V I H
In each run the initial concentrations a and b expressed in gram-moles per 22.4
liters were as follows:
For H2, o = 0.5637 '
For Br2, b = 0.2947
In Table A, T indicates time in minutes and x is the gram-moles of Hj or Brj reacted
per 22.4 liters.
The mechanism of this chain reaction is discussed on page 842. In terms of the
initial concentrations o and 6 add the conversion x, Equation (93) may be written to
express the reaction rate as follows,
^r = u = Ti = kctvKci'Vb — x{a — x) — kci{2x)
["fcîVxTaVb - x(o - i ) -1- fertVg^V6~^(2x)"|
(a)
kU2x) + k,,(b - x)
The thermodynamic properties of the reactants. and products are summarized in
Table B, taken from the data of Tables V, XIV, XVII, and XXXIV.
TABLE B
o o o
Brj(g) 7650 58.63 8.58 + (0.30) (lO"')?

Br(g) 26,880 41.81 4.97
H2(g) 0 31.23 6 . 8 8 + (0.066)(10-')r-f- (0.279)(10-»)r2
HBr(g) -8,650 47.48 6.30 + (1.819)(10-')r - (3.45)(10-«)T2
H(g) 51,900 27.40 4.97
From these data the thermodynamic changes and equilibrium constants accompany-
ing the individual reaction steps are calculated as follows:
(1) H2(g)+Br2(g)-^2HBr(g)
AHr = -24,234 - 2.867 + (1.636) (lO-')?^ - (0.323) (IQ-^)?'
AST = 20.462 - 2.86 In r + (3.272) (lO-')r - (0.485) (lO-')?^
Run I Run II Run III Run IV
r , °K 298.1 498.8 524.5 550.6 574.4
AH" -24,950 -25,244 -26,331 -25,367 -25,399
AS" 5.10 .4.21 4.13 4.07 4.01
AG" -27,394 -27,498 -27,608 -27,701
Ki 1.004(10'^) 2.876(10") 9.106(101°) 3.51(101°)
Kci 1.004(1012) 2.876(10") 9.106(101°) 3.51(101°)
(2) Br2(g)->2Br(g)
AHT = 45,718 + 1 . 3 6 r - 0.00015r2
AST = 17.33 + 1.36 In r -• o.ooosr
Run I Run II Run III Run IV

!r,°K 298.1 498.8 524.5 550.6 674.4,
AH" 46,110 46,359 46,390 46,421 46,444
AS" 24.99 25.63 25.69 25.74 25.799
(\G° 33,575 32,900 32,249 31,640
K, 1.951(10-15) 1.96(10-") 1.579(10-") 9.03(10-")
Kci 1.068(10-") 1.02(10-") 7.835(10-") 4.29(10-")
(4) Hs + Brfc;HBr + H
CHAP. XVIII] HYDROGENATION OF BROMINE 857
AHr = 16,471 - O.SSr + 0.876(10-')r2 - 0.208(10-8)73
AST = 4.650 - 0.58 In T + 1.753a0-')r - 0.312(10-»)r2
Run I Run II Run III BwnlV
r, °K 298.1 498.8 524.5 550.6 574.4
AH" 16,370 16,374 16,377 16,382 16,393
AS" 1.84 1.84 1.852 1.860 1.863
AG° 15,456 15,406 15,358 15,322
Kt 1.692(10-') 3.8(10-') 8.006(10-') 1.512(10-«)
Kct 1.692(10-') 3.8(10-') 8.006(10-') 1.512(10-«)
(6) H(g) + Brj(g) -^HBi:(g) + Br(g)
AHr = - 40,677 - 2.28r + 0.759(10-')r2 - 0.115(10-^')r'>
AST = 15.813 - 2.28 In T + 1.519(10-5)T - 0.172(10-^)72
^ Bun I Run I I Run III Run TV
T, °K 298.1 498.8 524.5 550.6 574.4
AHr -41,320 -41,668 -41,709 -41,722 -41,786
ASJI 3.26 2.36 2.284 2.208 2.15
AG°T -42,845 -42,907 -42,938 -43,020
K, 5.902(10«) 7.08(10") 1.108(10") 2.38(10")
JKCS 5.902(1018) 7.08(10") 1.108(10") 2.38(10")
With these thermodynamic constants the rate data are analyzed by the following
scheme which is demonstrated for Run I at 498.8°K.
1000 2000 3000 4000

Time=minutes
FIG. 169. Hydrogenation of Bromine.
The gram-moles of hydrogen consumed x are plotted against time T in Fig. 169.
Prom this plot on a large scale values of reaction rates rt were obtained by graphical
differentiation and plotted on a logarithmic scale against time in Fig. 170. By exten-
sion of this plot to T = 0 the value of u = kci'vKa'Vba = 1,95(10-') is obtained
directly, since at zero time the second term of Equation (a) disappears and the value
of X is zero. Since K^, a, and b are known, the value of kc4 is obtained.
1.95(10-5)
''-' = (3.27) (10-) (0.542) (0.5637) = '-^^^^'"'^ '
Values of kct are calculated from Kd'.

• , J , = ^ = 2:955(101;^
'* Kci 1.692(10-') ^ "^ ^^" •'
Values offccemay now be calculated by rearrangement of Equation (a) and the rela-
tion fce/fc'e = Kc6. Thus:
kct = ^
rikei(2x)
h,VK^^Vb - x{a - x)K,^{b - x) - uKMb - x) - fc.4(2x)2V^vT^ (b)
Values offee6 are calculated from Equation (b) for different times of reaction and
tabulated in Table C. It may be observed that there is no definite trend in the calcu-
lated values of kce- The average value ig taken as
kU = 1.753 (10-«)
kct = k'cf Kce = 1.763(10-8) (5.90) (10") = 1.035(10")
2
1000 2000 3000 4000 5000 6000

Time=minutes
FIG. 170. Evaluation of fcc4
(Equation (a), Illustration 10].
To test the vaUdity of the equation, the rates of, Run I were calculated from Equa-
tions (a) and the establtshed constants. The time-conversion relation was then
evaluated by graphical integration of the expression
-£ dx /r, as shown in Fig. 171.
The calculated values for Run I are plotted in Pig. 169 and show agreement with
experimental data.
In ordSr to establish the validity of Equation (a) at other temperatures, the reac-
tion velocity constants were similarly established for three other runs. The results
are summarized in Table D.
.In Fig. 172 these reaction velocity constants are plotted on a logarithmic scale
against 1 /T. It may be observed that nearly linear plots are obtained with maximum
deviations for Run II. From the lin^ of Fig. 171 values of the enthalpy and entropy
of activation of Equation (18) were calculated.
The following results were obtained:
(4) Hj + Br ?s HBr + H
AH* = 17,720; AS* = -9.035
CHAP. X V I I I ] HYDROGENATION OF BROMINE 859
O y-^ ^ r~l T-i Ci ^ T^
l£3 t ^ C^ t * 05 r)( CO
O ^ " 5 05 CO Co ®
O O O O i - i O i - H i - l
>n CO
tj- °'-IO3r-CC0<n-.^t»
o oocico-^tdrrt
«
coost^-î—(oi>co
COIN<M<N(N(MiHrH
o cootîoeo^HOi-^
^coeO(N(NlM(N'-i'-'
OtO«5(NO-*Oo O
(N-*^COt>tO>g O
_COU5000lM-*gO O ,_
<=>000.-irH'-ilNl M S^
d o o d o ' o o o p
CO
H I>
i>t-ThcnîN'n(Nco
cot^cOINlNOO'iCO
USTJICOIN'HOOICDN
d d d d d d d o d
I
•*(N'-'OOiOOt-COO>
u5U5iiOiCTj<Tt(TtHrt<CO
1 d d d d d d d o o 4
H t^t--*0!'-i<NiOlNCO
inooiveocoi-ii-iiM-*
1 rot^to'OTjtcocôi'o
e
,_( lO -^ 05 t^ CD 00
"-So t~ C2 <N O m C» rH
d d i': 00 i-H I N i-^
i-H r l 1-1 (M rq cvi
II
-4 o
lON'^lM'NOOeO'H
OiOOt^tOiO'^COi-l
oeoootoic(NiO'<j(
5Dt>-Oi-lCOCO(MO
° O O C > O O O i - i i - l
d d d o d d o o
o o o o o o o o o
OOiONOIîOCDO
i-i N CO CO • * CO ca
(6) H(g) + B r , ( g ) - > H B r ( g ) + B r ( g )

Afl* = 7200; AS' ..= 5.24
TABLE D
SUMMARY OF REACTION VELOCITT CONSTANTS
Run I Bun II Run III Run IV
498.8 ' 524.5 550.6 574: A
1.955(10') 5.77(10') 1.037(10*) 2.39(10*)
Lisaaoio) 1.52(10i») 1.28(101°) 1.58(10")
kcs 1.035(10") 1.035(10") 2.29(10") 3.106(10")
kct 1.753(10-8) 1.46(10^) 2.085(10-«) 1.305(10-5)
From these values of AH* and AS* and the equilibrium constants values of con-
version at various times of reaction were calculated for all four runs and plotted in
Fig. 169. It may be seen that the calcu-
lated values are in close agreement with
experimental data for all runs except Run
II. It is possible that there was some ex-
perimental irregularity in this run.
The values of AH^ calculated by this
analysis are in only fair agreement with the
experimental and calculated values cited by
Glasstone, Laidler, and Eyring for the same
reactions.
Product Distribution and Selec-

tivity. In complex systems the over-
all effect may be the result of many
primary, secondary, and tertiary re-
actions. .Since these individual reac-
tions may exhibit^different responses to
variations in temperature, pressure,
and composition, their combined re-
sults frequently present a complex
pattern of relationships which can be
rationalized only by being resolved
into the contributions attributable to
individual reaction steps.
0.0016 0.0018 0.0020 In presenting data on such systems
1 it is convenient to designate as con-
yog- version the decomposition or dis-
FIG. 171 Effect of Temperature on appearance of the key or limiting
the HBr Reaction Velocity Constants. reactant. Thus, percentage conver-
sion is the percentage of the key reactant in the feed which is con-
verted into other products. A tabulation of number of moles of each
CHAP. XVIII] DISTRIBUTION AND SELECTIVITY 861
product formed per mole of key reactant decomposed is termed the

product distribution of the reacting system. The selectivity of the reac-
tion is the percentage ratio of the moles of desired product formed
per ,mole of converted key reactant to the number of moles wliich would
have resulted were there no undesired side reactions.
In laboratory studies of such complex systems it is desirable to select
a fixed feed composition, temperature, and pressure and to cover vnde
-ago
20 80 40 50 60 70 80 90
Per Cent Propane Converted
Fia. 172. Distribution of Products from the Pyrolysia of Propane, (at 725°C and
1 atm pressure).
ranges of conversions by operations at widely different space velocities.

The reôn of low conversions is of importance in identifying the primary
reactions, whereas data at high conversions are necessary for the evalua-
tion of secondary reactions.
.In analyzing such data it is convenient to plot first the complete
product distribution of each series of runs as ordinates against percentage
conversion. Extrapolation of these curves to zero conversion fixes the
product distribution of the primary reactions ahd gives a basis for identi-
fying these reactions.
In Fig. 172 are the product distribution curves for the pyrolysis of
propane at 650°C and 1.0 atm, as plotted by Myers and Watson' from
the data of Schneider and Frolich." Inspection of these curves shows
the primary reaction products to be propylene, hydrogen, ethylene,
methane, and butane.
"V. Schneider and P. K. Frolich, Ind. Eng. Chem., 23, 1405 (1931).
862 HOMOGENEOUS REACTIONS {CHAP. XVIII
This result may be expressed by the following stoichiometric equa-
^^°°^- C3H8?:iH2+C3H6
CaHg ?=i CH4 + C3H4
2C3H8 ^ CjHe + CsHs + CH4
2C3H8 ^ C2H6 + C4H10
There is evidence that all of these reactions actually occur through

chain-reaction mechanisms involving free radicals of the general type
analyzed in Illustration 10. However, in such complex systems this
type of fundamental kinetic analysis is difficult, and it is at present more
fruitful to adopt a semiempirical approach. Each individual net re-
action is assumed to occur in accordan.ee with its simple stoichiometric
equation, but with an empirically determined apparent order of reaction
which is evaluated from the effect of pressure on the rate. Many pri-
mary pyrolytic reactions satisfactorily approximate simple first-order
results. However, careful analysis'" of data over wide ranges of condi-
tions indicates that large errors may result from extended extrapolation
based on such approximations.
Myers and Watson followed this procedure in analyzing the data from
the literature on propane pyrolysis and arrived at the product distribu-
tion and rate-of-reaction relationships which are summarized in Table
LIV. These reactions are assumed to proceed independently at the
indicated rates. However,, it should be recognized that these rate
equations do not represent truly independent reactions of simple order
and that they are rehable only over restricted ranges of conditions in the
system for which they were developed.
TABLE L l V
EFFECTIVE OVER-ALL REACTIONS IN PKOPANE PYBOLYSIS'
r = rate of reaction, lb-mole/(cu ft)(hr)
V = total pressure, atm.
B_ AH" .AS"
k = Ae ^^; E = ca\ per g-mole; T = degrees Kelvin
(1) CsHs -* CiUt + CH4
r = kv'—*-*; (moles of propane) A = (3.0158)10"; E. = 66,500
(2) 2CsH8 —* CaHg -f- CaHc ~i- CH4

r = fcir —*-5; (moles of propane) A = (3.6745)101*; E = 66,000
n,
(3) C H s 7± C,H. + H2
•
Jes of nronjiTifi^
A = (9.989)10"; E = 60,000; AHISOOT = 30,605; ASISOO^F = 32.85

CHAP. XVIII] PYROLYSIS O F P R O P A N E 863
(4) 2C3H8 ?^ CjHe + C4H10
r = fcir 1 —^-5 — \ / — I ' ; (moles of propane)

L i( '' niK J
A = (2.490)101"; E = 54,000;
K = 1.30 (substantially independent of temperature)
(5) C H e + H s - • CHi + C2H4
r = t j T ^ S ^ ; (moles of propylene) A = (5.266)10'; E = 38,000

n<
(6) C , H , -> 0.149CHi + O.O64C2H2 + O.2555C2H4 + 0.085CjH8 + 0.2555aH8
+ O.O745C6H10 + O.O745H2 + O.O53C2H6 + O.O53C4H.
r = fcTT^^^^; (moles of propylene); A = (2.746)10"; E =56,000
nt
(7) G H i o ^ O.I2H2 + 0.49CH4 + O.SOCjHi + O.SSCjHe + O.49C3H6 + O.OlCHs
VO.I2C4H8
r = fcx^^^^ ; (moles of butane); A = (4.78)10"; E = 63,000
nt
(8) CsHs •^ C2H4 + H2
r = kir
["n^H. _ n c a " a ^ 1 . (jj^^jgg ^f ethane); A = (9.83) 10«; £/ = 63,000
L nt n,K J
AHuooT = 34,316; A>Si3oo»r = 32.12
(9) C H s + H2 ^ CH4 + CsHe
J. = fcJ"°'°'; (moles of butylene); A = (4.743)10"; E = 67,000

n<
(10) O.287C2H2 + O.333C6H10 + O.SSCsHs -^ 0.472C6H6 + O.giCHi + O.333C2H4
+ 0.178C
r = kv^ (^^^^f^); (moles of total reactants); A = (9.34)10»; E = 35,000
WHHC = combined moles of reactants
For example, the apparent rate of pyrolysis of butenes as a secondary-

reaction in the pyrolysis of propane may follow a rate equation which
would not be applicable to the pyrolysis of pure butenes. Ultimately it
is to be hoped that a practical method can be devised for handling such
systems by fundamental reactions involving free radicals similar to those
of Illustration 10.
By consideration of the reactions of Table LIV it is possible to predict
the conversion and product distribution obtained when propane is
pyrolyzed under specified conditions and thus to arrive at optimum pro-
cedures. Myers and Watson demonstrated this method by carryiag out
a design study on a large-scale tubular reactor operating with widely
varying temperatures and pressures.
GRAPHICAL REACTOR DESIGN

Industrial reactions are usually complicated either by simultaneous
reactions or by varying operating conditions or both. For this reason
it is rarely possible to employ integrated rate equations in problems of
reactor design. Graphical methods are more generally useful and permit
the solution of all types of problems.
Adiabatic Reactors. Large-scale processes are frequently carried out
in a flow-reactor system consisting of a preheater in which the pharge is
heated to the reaction temperature and an adiabatic-reaction chamber
which is designed to produce the extent of conversion desired in addition
to that obtained in the preheater.
If the flow through a reaction chamber is progressive, without signifi-
cant turbulence in the direction of flow, the temperature in an endother-
mic reaction will progressively drop from a maximum at the inlet to a
minimum at the outlet. This condition is approached in a reactor
having a diameter which is small in comparison to its length. In a
reactor having a diameter of the same order as the length of the direction
of flow there is opportunity for extreme longitudinal turbulence under
certain flow conditions. Such turbulence tends to mix the contents of
the reactor to a state of uniform temperature and composition, and the •
operation becomes both isothermal and adiabatic. Under these con-
ditions the uniform average temperature of the reactor is lower than the
temperature of the inlet stream by an amount depending on the endo-
thermic heat of reaction and the conversion produced in the reactor. In
effect, the inlet stream is abruptly quenched to a lower uniform tempera-
ture upon entering the reactor. ,
The first condition of progressive temperature change is favored by any
factor which diminishes convection, such as shiall reactor diameter, low.
heat of reaction, and a small volume change accompanying the reaction.
The second condition is formed by all factors which promote convection,
such as large diameter of the reactor, high heat "of reaction, and a large
volume change. This latter condition is approached where an ej:otherT
mic reaction with a large volume increase is conducted in a vertical
reaction chamber with flow in the downward direction. In such a
system all changes accompanying the reaction make the mixture less
dense which results in extreme turbulence and convection in a downflow
system. A similar condition of adiabatic and isothermal reaction is
frequently produced by means of mechanical agitation in liquid-^hase
reactors.
Many reactors operate afconditioiis intermediate between the two
extremes. Since temperature distribution in such reactors is difficult to
CHAP. XVIII] ADIABATIC REACTORS 865
predict it is desirable to develop a design on both bases and select a con-

servative intermediate value.
Case 1. Progressive Changes. For a simple reaction where the course
can be represented by a differential equation involving only one variable
conversion term XA, a direct method of reactor design results from combi-
nation of the rate equation with an equation relating temperatures to
conversion, such as is developed in Illustration 22 of Chapter VIII,
page 309. Such an equation expresses the temperature at any point in
an adiabatic reactor as a fimction of the conversion XA- Thus:
T = UXA) (95)
The rate of the typical reaction, A -{- B = R + S, is given by Equa-
tion (44) in which k and K are functions of temperature represented by
Equations (18) and (22), respectively. Thus:
^ A- = MT) (96)
K = MT) (97)
By combination of Equations (95), (96), and (97) with the rate equation
it is possible to calculate the reaction rate r corresponding to any selected
conversion XA for a particular system at fixed initial conditions. The
temperature is first calculated from Equation (95) and the velocity and
equilibrium constants from (96) and (97). From a series of such calcu-
lations corresponding to different values of XA a curve may be plotted
relating 1/r as ordinates to XA- From Equation (81) it is evident that
the area under this curve between XAO and XA is equal to Vr/F for a flow
system. A s>imilar integration for obtaining time of reaction in a batch
system is made by plotting l/rVc against XA and determining the area
under the curve in accordance with Equation (56). Where a large
pressure drop occurs through the reactor the general method of integra-
tion discussed on page 880 is followed.
Case 2. Uniform Reactor Conditions. If conditions of extreme turbu-
lence exist in a reactor the average composition and temperature of the
material in the reactor must be the same as those of the product dis-
charged from it. For any selected over-all conversion the outlet tem-
perature is calculated from Equation (95). This temperature fixes the
average reaction velocity constant and equihbrium constant for the
entire reactor. The average rate of reaction per unit volume of reactor is
then calculated directly from Equation (44) the composition of the
product being employed. The required reactor volume is then equal to
F{XA - XAo)/r.
Illustration 11. The decomposition of phosphine is afirst-orderreaction which
proceeds according to the following stoichiometric equation:
4PH3(g)-^P4(g)+6H2(g) (a)
In the International Critical Tables the reaction velocity constant of this reaction is
expressed by the following empirical equation with temperature in degrees Kelvin
and time in seconds:
log kc = — + 2 log r + 12.130 (b)
It is proposed to produce phosphorous by the decomposition of phosphine in an

adiabatic-flow reactor. Pure phosphine is fed to the reactor at the rate of 500 lb per
hr at atmospheric pressure and at a temperature of 672°C which is the highest tem-
perature permitted by the available materials of construction. The decomposition
produced in the preheater and the rate of the reverse reaction may be neglected.
(a) Assuming that the reactor is of such shape that the temperature changes
progressively without longitudinal turbulence, calculate the reactor volume per pass
for various degrees of conversion.
(b) Repeat part (a) assuming that the adiabatic reactor is of such a shape that the
temperature and composition are uniform.
(c) For comparison with the results of'parts (a) and (6) plot a curve relating
conversion per pass to reactor volume if an isothermal reactor at 672°C is employed.
Solution: (a) The standard heat of reaction (a) at 18°C per g-mole of phosphine
gas is given by Bichowsky and Rossini^^ as 5665 cal.
The molal heat capacities of phosphine gases were calculated by Hirschfelder''
from estimated vibration frequencies.
P4(g); cp = 5.9 + 0.0096r
PHsCg); Cj, = 6.70 + 0.00637 '
HjCg); cp = 6.88 + 0.066(10-')r + 0.279(10-«)T2
The following energy balance may be written on the basis of 1 mole of phosphine fed
where x is the moles of phosphine decomposed.
£
45 nT
(6.70 + 0.00637) dT = 0.25z / (5.9 + 0.00967) dT

_I •/291
+ 1.5x / [6.88 + 0.066 (10-') 7 + 0.279(10-8)72] dT
«/291
+ (1 - x) / (6.70 + 0.00637) dT + 5665a; (c)

/291
t/29]
Upon simplification, this becomes

9145 - 6.707 - 0.0031572
(d)
5.0957 - 0.001972 + 0.14(10-«)7' + 4341
Solving for x at various values of 7 yields the following values:
T X T X
945 0 600 0.5917

900 0.0742 500 0.7787
800 0.2432 400 0.9799
700 0.4145
" " Thermochemistry of Chemical Substances," Reinhold Publishing Company
(1936).
' ' J. O. Hirschfelder, private communication (1942).
CHAP. X V I I I ] DECOMPOSITION OF PHOSPHINE 867
By plotting T against x, values of T are obtained for regular increments of x.

The corresponding values of kc are calculated from Equation (b).
TABLE A
X fcc(l/hr) X T fe(l/hr)
0 U^ 37.19 0.6 596 2.627(10-")
0.1 885 1.422 0.7 542 1.469(10-")
0.2 826 , 3.654(10-^) 0.8 489 1.913(10-18)
0.3 768 6.832(10-^) 0.9 440 3.133(10-")
0.4 708 • 4.603(10-«) 1.0 394 5.598(10-")
0.5 652 1.667(10-8)
The total pound-moles of product per pound-mole of feed for any degree of con-
version X is equal to (1 + 0.75x4), and the volume Vm of product in cubic feet per
pound-mole of feed is given as
359
f„ = (1 + 0.75x4) — T (e)
FdxA
From Equations (47) and (80), r = —r-~
aVr
F = (500)/(34.12) = 14.65 Ib-mMes per hr
273 Jo «c(l — XA) JO KC

For an adiaba^c reactor without longitudinal turbulence, Equation (f) is evaluated
19.32 IY(x)
from a plot of —-—; against x, as shown in Fig. 173. For selected values of x,
k
f(x)
values of T and kc are obtained from Table A, and values of T —— are calculated from
k-
Equation (f). The total reactor volume required for a given degree of conversion is
obtained by summation of the volumes for small Ax increments from 0 to x. The
volume required for any degree of conversion is plotted in Fig. 174.
(b) Adiabatic Reactor with Uniform Mixing. If uniform conditions prevail in the
reactor because of mixing, the differential form of the rate equation can be used to
calculate F , directly. Thus, from Equation (f) it follows that,
Values of T and kc are obtained from Table A, and the corresponding values of V,
from Equation (g) are plotted in Fig. 174.
(c) Isothermal Conditions. Where the reaction proceeds isothermally at 945°K the
reaction velocity constant is 37.19 per hr, and Equation (f) may be integrated
directly to give
Vr = 504[-0.75x - 1.75 In (1 - x)] (h)
In Fig. 174 values of Vr are plotted for various degrees of conversion under the three
stated conditions of operation. Comparison of the curves of Fig. 174 shows the
excessive size of reactor required for the adiabatic operation of this particular endo-
thermic reaction as compared with constant temperature operation. It is evident
that longitudinal mixing in an adiabatic reactor for an endothermic reaction is
unfavorable.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

a;= fractional conversion
PIG. 173. Integration of Equation (f).
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

a;, fractional conversion
FIG. 174. 'Volume of Reactor for the Decomposition of Phosphine.
CHAP. XVIII] VISCOSITY 869
Because of the temperature drop resulting from a highly endothermic reaction it

may not be possible to obtain high conversion per pass in an adiabatic reactor of
reasonable size. Turbulence or mixing in the reactor is disadvantageous for the
endothermic reactions, whereas for most exothermic reactions it is advantageous. In
general, adiabatic operation is employed for endothermic reactions where the addition
of heat is difficult or where secondary reactions which may accompany high conver-
sion must be avoided. It is sometimes advantageous to add inert components such as
nitrogen or steam to increase the heat capacity of the system and thus reduce the
temperature drop.
I
Pressure Drop. The calculation of pressure drops in reactors is
frequently a difficult problem, particularly if a change of phase occurs in
the reactor; For tubular reactors in which gases are flowing at rela-
tively high velocities pressure drops are satisfactorily represented by the
following equation:
dp _ 0^235 / w
dL~ L 77;;^ -"^ (98)
7 P
where
dp/dL = pressure drop per foot of equivalent pipe length, pounds per
square inch
D = internal diameter, inches
w = mass rate of flow, pounds per hour
fi = viscosity, micropoises
p = density, pounds per cubic foot
Equation (98) is derived from the general Fanning equation which is

treated at length in the standard texts on unit operation and fluid flow.
The general equation should be used in deahng with problems involving
the flow of Kquids or of gases at low velocities.
The problem of pressure drops of gases flowing through beds of granu-
lar materials is treated in detail in Chapter X X I , page 1015.
Viscosity. Viscosity is a measure of the resistance to flow offered by
a fluid in straight line or laminar motion without turbulence. It may be
visualized as resulting from the friction of infinitesimally thin layers of
fluid moving over each other. The force required to maintain such rela-
tive motion is proportional to the area of contact between the layers and
the relative velocity of movement. Thus, viscosity is defined as the
force per unit area which is required to maintain a unit velocity gradient,
or
H = F/(A){du/dx) (99)
where
fi = viscosity
7^ = force in the direction of flow
A = area of surface in relative motion
u = velocity of flow
X= distance, perpendicular to the direction of flow
(du/dx) = velocity gradient
If consistent units are employed, it is evident that viscosity has the
dimensions of (force) (time)/(length)2. In metric units the unit of
viscosity is termed the poise which is 1.0 (dyne) (sec)/sq cm. The
English unit of viscosity is the (poundal) (sec)/sq ft, or, since force equals
mass times acceleration, lb/(ft) (sec).
1.0 poise = 0.0672 (poundal) (sec)/sq ft or lb/(ft)(sec)
The most commonly used units of viscosity are the cmtipoise which is
0.01 poise and the micropoise which is 10~^ poise. The viscosity of water
at 68°F is 1.0 centipoise. For this reason viscosities in centipoises are
frequently referred to as viscosities relative to water.
Kinematic viscosity is defined as the absolute viscosity divided by the
density. In the metric system the unit of kinematic viscosity is the
stoke which has the dimensions of (dyne) (sec) (cm)/g or (sq cm)/sec.
The centistoke is 0.01 stoke.
Numerical values of viscosities vary over wide range, depending on
the nature of the substance and its temperature and pressure. Experi-
mental measurements of viscosity are presented for many substances in
the standard tables of physical data. For the most part such data are
restricted to pure substances at temperatures near atmospheric and to
moderate pressures. Where viscosities must be known at high tempera-
tures or pressures or for fluid mixtures it is frequently necessary to resort
to generaUzed methods of approximation which should be rationalized
with whatever direct data are available.
A universal viscosity correlation based on the theorem of correspond-
ing states was proposed by Uyehara and Watson" who developed the
graphical relationship between reduced viscosity (jur = n/nc) and reduced
temperature and pressure shown in Fig. 175. This relationship is based
on consideration of all available data on the variation of viscosity with
temperature and pressure in both the liquid and gaseous states. It is
a,pproximately the same for all substances.
" O. A. Uyehara and K. M. Watson, Natl. Petroleum News, Tech. Sec., 36, iJ764
(October 4, 1944). Also, " Process Engineering Data," National Petroleum Pub-
lishing Company, Cleveland (1944).
CHAP. XVIII] VISCOSITY 871
By means of Fig. 175 the viscosity of a substance at the critical point

He may be estimated from a single viscosity measm-ement on either the
WW \
\U\\ \^^\ \
\Ss
m^ ^^ \\
10 ^^ \ \ \
vv \\\^^\
\\\ \\\\ \ \
8
• \ \ ^ s\V
6 1 \ > ^w
\\ \\\ Â\ s \
\\
A^
\i^\S\
5 Vi \ ^ '
\ \\\^ sV \
4 ^\\ v\N^ \ C\ , \ \
'•m \^^V s\ ^ \ \ ^ ,
3 W1^w \ \ f
d \ ' ^"^ ^ < . ^ ,^
2
0.7
0.8 1W
m
\^
\^^
\> \
v^l \\
^
\
^ ^-
•<
fck^
1^^^
^,
- J : ; — 1.-
^ '^. ?J
-^' ^
U*y
r
f ^
0< M \ -6
6^
'î ^ ;>'
\v \
.
v^
w
1 0
c5
o1
\ •s?'
V ''/& ^
w i ^ ^
0.8 Poo • ^
d•iy
0.6
?'\
tv ^/
"^
l4
0.5 i?t\^
•cPX;
0.4 0.2 ^^
y y
0.3 ^^
0.2
0.4 0.5 0.6 0.8 1.0 2.0 3 4 5 6 10
Reduced Temperature, T^ = T/r^
Fia. 175. Generalized Reduced Viscosities.
Hquid or the gas at any known reduced temperature and pressure. The
viscosity at any other temperature and pressure may then be calculated
from the equation: i-.r.i^s
fi = Hriic (100;
where
iir = reduced viscosity read from Fig. 175
lie = viscosity at the critical point
In using Fig. 175 in conjunction with experimental viscosity data it is
desirable to calculate the value of Hc used in Equation (100) from data at
conditions as close as possible to the range of interest. Thus, although
it is possible to estimate the viscosity of a gas from a measurement on
•the liquid, more reliable results are obtained if calculations for the gas
phase are based on gas-phase data wherever possible. The critical
viscosities of various substances based on experimental data for the
indicated phase are summarized in Table LV.
If no direct viscosity data are available, useful approximations may be
calculated from the Uquid density or structueal parachor together with
molecular weight and critical temperature by the following equation:".
jUc = 61.6 . , | = 7.70 rr- = 7.70 •—rr—i (101)

(fc)' (uiwi)* {My
where nc = critical viscosity in micropoises
M — molecular weight
Tc = critical temperature, degrees Kelvin
Vc = critical volume, cubic centimeters per gram-mole
(fiwi) = product 6i expansion factor (Fig. 109) and mblal volume,
cubic centimeters per gram-mole
(pi/«i) = liquid density in grams per cubic centimeter divided by
the expansion factor
A similar expression, based oh a knowledge of molecular weight, criti-
cal temperature, and critical pressure was derived from an equation
proposed by Licht and Stechert.'^
.. = 7 . 7 0 % ^ (102)
Uyehara and Watson found both Equations (101) and (102) to be in good
agreement with over 50 compounds on which data are available. In
general the errors were less than 10 per cent. Exceptions are water,
hydrogen, helium, and bromine. Equation (101) was found to be more
reliable for complex molecules for which good Uquid-density data are
available, whereas Equation (102) giveis better results for the light gases.
" W. Licht, Jr., and D. G. Stechert, /. Phys. Chem., 48, 23-47 (1944).
CHAP. X V I I I ] VISCOSITY 873
TABLE LV
CEITICAL ViscosmES
lie = micropoises
Phase Mo Phase fc
Acetic acid g 376 Zso-Butane g 239

Acetone g 285 Zso-Butene g 250
Acetylene g 237 Zso-butyl alcohol... 1 319
Air g 193 7so-butyl formate . g 310
Ammonia . g 309 /so-pentane g 240
Argon g 264 /so-propyl alcohol.. g 297
Benzene g 312 Krypton g 396
7i-Butane g 239 Methane g •
159
Carbon dioxide g 343 Methyl acetate . . . . g 307
Carbon disulfide g 404 Methyl alcohol g 375
Carbon monoxide . . . g 190 Methyl chloride . . . g 338
Carbon tetrachloride . g 413 Methyl propionate g 301
Chlorine g 420 Naphthalene 1 340
Chloroform g 410 Neon g 156
Cyclohexane g 284 Nitric oxide g 258
Diethylamine g 276 Nitrogen g 180
g 210 Nitrous oxide g 332
Ethane
g 299 n-Nonane 1 265
Ethyl acetate
g 334 n-Octane 1 259
Ethyl alcohol
g 345 Oxygen g 250
Ethyl chloride 238
Ethyl ether g 268 n-Pentane g
g g 274
Ethyl formate 282 Phosphine
g g 228
Ethyl propionate 284 Propane
g g 298
Ethylene 215 Propyl alcohol....
g g 296
Helium 25 .4 Propyl formate . . .
1 g 233
n-Heptane 254 Propylene
g 411
n-Hexane g 248 Sulfur dioxide . . . . 306
Hydrogen g 34 Toluene g
g 495
Hydrogen chloride . . . g 353 Water 490
Hydrogen cyanide . . . g 316 Xenon g
Illustration 12. Estimate the viscosity of toluene vapors at a temperature of

700°F and a pressure of 400 lb. per sq in. from the following data:
Tc = 320.6°C Pc = 41.6 atm
At 20°C, p = 0.866 g per cc ilf = 92
Solution:
Tr = (293.2)/(593.8) = 0.494
From Equation XII-40,
£01 = 0.1745 - (0.0838) (0.494) = 0.1331
pi/wi = (0.866)/(0.1331) = 6.506
From%quation (101),
(593.8)* (6.506)1 _
From Equation (102),

^y„ (92)^41.6)?
' ' ° - ^ - ^ ° (593.8) * - ^ ° ^
At 700°F and 400 lb per sq in.,
T, = (1160)/(593.8) (1.8) = 1.085; ' p, = (400)/(41.6)(14.7) = 0.654
From Fig. 175, ^r = 0.54.
1
From Equation (100),
f. = (0.54) (310) = 168 micropoises
Viscosities of Mixtures. No entirely satisfactory method is available
for estimating the viscosities of mixtures. Uyehara and Watson recom-
mend the assumption that Fig. 175 is appKcable to mixtures if the
reduced values are based on pseudocritical properties. The pseudo-
critical viscosity, pressure,'and temperature are each taken as equal to
the molal average of the true critical properties of the components of the
mixture. Thus
Mc = •^''iMci + A'2' M<:2 + iVsMcs • • •• (103)
where
jUj = the pseudocritical viscosity of a mixture containing iVi mole
fraction of component 1 whose critical viscosity is ]i.c\, etc.
From the pseudocritical viscosity calculated from Equation (103)

the viscosity at any selected pseudoreduced conditions is obtained as the
product of Mc times the pseudoreduced viscosity ii\ read from Fig. 175.
This method shows fair agreement with data on the viscosity of gaseous
mixtures and is believed to be more simple and reliable than methods in-
volving the viscosities of the pure components at the conditions of the
mixture.
Nonadiabatic Reactors. In the general case of a process which in-
volves a combination of consecutive and simultaneous reactions, a
differential equation may be written for the net rate of disappearance or
formation of any reactant or product. This procedure is demonstrated
in the development of Equations (a-f) of Illustration 9, page 849 for the
production of diphenyl from benzene.
By means of such differential rate equations the over-all results of an
operation producing finite changes may be calculated by a stepwise
summation of the changes produced in successive small increments of
time or reactor volume, starting from the conditions of the original
charge. If the incremental steps are small, it may be assumed that the
average velocity constants, volumes, and numbers of moles for an incre-
ment are the arithmetic means of its terminal values. This method of
integration is not unduly tedious and has the advantage of appUcabiUty
CHAP. X V I I I ] NONADIABATIC REACTORS 875
to the most complex systems under conditions of varying temperature,

volume, and pressure. Equations similar to (a-f) of Illustration (9) can
be written for any combination of reactions.
The summation procedure is best understood by study of Illus-
tration 13 which is a problem in the design of a reactor for a compex
system in which temperature and pressure vary simultaneously over
wide ranges. The illustration deals with a flow system, but a similar
procedure is applicable to batch reactors.
niustration 13. On the basis of the fundamental data developed in Illustration 9,

it is proposed to design a plant for the production of diphenyl by the continuous
pyrolysis of benzene in a flow system. A preheater and reactor is desired to produce
35,000 lb per operating day of a mixture of diphenyl and triphenyl containing approxi-
mately 30 per cent by weight of triphenyl. The benzene is to be supplied from a
vaporizer coil at 700°F. It may be assimied that no conversion takes place in the
vaporizer. A trial design is to be based on the use of alloy steel tubes having an
inside diameter of 2.0 in. for both the preheater and reactor. The tube materials
• limit the safe operating temperature of the reactor to a maximum of 1250°F. To
obtain maximuni capacity per unit volume of reactor the feed will be heated to a
temperature of 1250°F and held constant at this temperature in a sufficient number
of tubes to produce the desired conversion. A constant heat input rate of 10,000 Btu
per hr per sq ft of inside tube surface will be assumed for the preheating section, and
the reactor section will be placed in a separate zone in which the heat input may be
controlled as necessary to maintain a constant temperature. In order to minimize
tar formation the gauge pressure at the reactor outlet wiU be controlled at 15 lb per
sq in.
Calculate the number of tubes 10 ft in length which are required for the preheater
and reactor sections, and determine the distribution of temperature, pressure, and
composition throughout. The tubes will be connected in series with 180° bends each
of which produces a pressure drop equivalent to a length of 50 diameters of straight
pipe. The volume of one tube and its return bend is 0.23 cu ft, and the heated length
of each tube is 9.2 ft.
' The physical data in Table A for the components are taken from the International
Critical Tables.
rTABLE A
Mokeular Diphenyl
Weight Benzene Diphenyl Benzene Hydrogen
78 154 230 2.0
te'C 79.6 255 427 -252.7
tc°C 288.5 ,,' ,, -239.9
Pc atm 47.7 ,. ,, 12.8
n (gas) at (fC)
(10~* poise) 123(212.5°C) 88.2(23°C)
S^hition: The specifications of the problems fix the temperature and pressure at the
outlet of the reactor and the temperature and conversion at the inlet of the preheater.
The inlet pressure is unknown. Since the enthalpies of activation of the forward
reactions are approximately equal, and there is no change in the number of moles in
either reaction, the distribution of the products is little affected by changes in either
reaction temperature or pressure. For this reason the conversion at the reactor out-
let may be closely approximated from the specification of the ratio of diphenyl to
triphenyl in the product.
' A design of this type requires a trial-and-error procedure in which the graphical
integration may be started from either end of the reactor. Thus, a first approxima-
1000
—
4_
600
^
/C;
100
" ^
50
Y
kyz'
10
5
1.0
0.5
Kz = —
KA,
- ^
0.1
I—•
0.05
0.01
700 800 900 1000 1100 1200 1300 1400
Temperature, °F
FIG. 176. Reaction Velocity and Equilibrium Constants of the Diphenyl Reaction
tion to the inlet pressure may be assumed, and temperature, pressure, and con-
version through the preheater and reactor may be evaluated on this basis. If the
calculated outlet pressure, corresponding to the desired conversion, does not agree
with the specified value, the integration must be repeated with a new trial inlet
pressure. Two or three trials of this type generally serve to estabhsh the correct
conditions throughout.
An alternate procedure is to start the integration from the outlet of the reactor, and
integrate back toward the inlet. In this method it is necessary to assume a trial
value of the conversion at the outlet of the preheater. On this basis the conversion,
temperature, and pressure throughout the preheater are calculated. If the calculated
conversion at the inlet of the preheater does not agree with the specified value, the
CHAP. XVIII] DIPHENYL REACTOR 877
calculation must be repeated with an adjusted trial value of the conversion at the
outlet of the preheater.
The choice between these two procedures is determined largely by personal prefer-
ence and the conditions of a particular problem. If pressure drop is relatively small
and unimportant, the first procedure based on an assumed inlet pressure is generally
xuu 1 1 1 1 1 1 ^ 150 -22,000
90 140 -21,000
80 - 130 -20,000
120 -19,000-0
3 S1
M -0
u"60 - 110 "I--18,000 %
<N
3
w
a>
1 0 0 ^ -17,000 n
UA
S4-> 50 / / ^^^^^ c-
0
oK! 4J
P^
d
O 40 -
a u
90 (3
„ -16,000 (§
<v ' 0 0
W / c.«^-<^>-^ %
3 30
/^-JK'^'^ - 80 -15,000 «J
•M
ei
<5
20 y/y - 70 --14,000
10 ^ ^— C'p\ H2 - 60 -13,000
' 1 1 1 1 '1 1 50 -12,000

200 400 600 800 1000 1200 1400
Temperature, °F.
FIG. 177. Thermal Properties of the Diphenyl System.
more convenient. If pressure drop is of considerable importance, and relatively little

conversion occurs in the preheater the second alternate is desirable. In questionable
cases a combined method may be used in which the first trial integration is carried
back from the outlet and a trial inlet pressure is calculated. The second integration
may then be carried out in the forward direction, starting with an inlet pressure based
on the first calculation. Such a combined procedure will be followed in this illus-
tration.
In carrving out the integrations of a complex problem of this type, values of""
reaction velocity constants, equilibrium constants, heat capacities, and heat of

reaction are required at many different temperatures. Curves expressing these
properties as functions of temperature are plotted in Figs. 176 and 177 based on the
data of Illustration 9. In order to insure a conservative reactor design the heats of
reaction are based on the values calculated from the group contributions of Tables
XXXIX and XLIII rather than on the lower values indicated by the equilibrium
data.
On this basis: '
(AHI)298 = 9,700; Ijii = 8,113 cal per g-mole of diphenyl
i,AHi)m = 7,200; In, = 5,143 cal per g-mole of triphenyl
In the absence of experimental data the critical temperature and pressures of
diphenyl and triphenyl are estimated from the boiling points by means of Equa-
tions (III-9, 13-15), page 71. The critical viscosities of hydrogen and benzene are
estimated from the data of Table A and Fig. 175, while the critical viscosities of
diphenyl and triphenyl are calculated from Equation (101). These results, together
with the available experimental data, are summarized in Table B.
TABLE B
SUMMARY OF CHITICAI. PROPERTIES
Benzene Diphenyl Triphenyl Hydrogen
T° R 1011 1458 1895 59.9
Pc lb per sq in. 701 461 385 188
iJLc poise (10-8) 324 317 326 35
SPECIFIED COMPOSITION OP PRODUCT

Weight Per Cent Mole Per Cent
Diphenyl 70 77.70
Triphenyl 30 22.30
From Fig. 168 may be read the complete composition of the reacting system corre-
sponding to this product composition if the reaction proceeded isothermally at 1265°F.
In the plant the average reaction temperature will bo lower than 1250°F which will
slightly increase the proportion of triphenyl formed with a given conversion of ben-
zene. This qualitative trend results from the fact that, although the enthalpies of
activation of the two forward reactions are taken as equal, the equilibrium constant
of the first reaction is increased more by an increase in temperature than is that of the
second reaction. Accordingly, increased temperature favors'the first reaction, and
the design will be based on a conversion of benzene slightly lower than that corre-
sponding to the desired product composition at 1265°F. This is accomplished by
assuming 12.8 per cent diphenyl in the reactor-outlet mixture instead of the 13.2 per
cent indicated by Fig. 168. The proportions of the other components are adjusted to
conform to the specified product composition and the stoiohiometry of the reactions.
Thus, nB — 2hT + nz>.
Weight of product, lb per hr = 35,000/24 = 1,458
Weight of reactor feed, lb per hr = (1458) /(0.2526 + 0.1086) = 4,036
Molal reactor feed rate F lb-moles per hr ='4036/78.1 = 51.68
It is convenient to consider integral numbers of tubes in carrying out the reactor
CHAP. XVIII] DIPHENYL REACTOR 879
REACTOB OUTLET COMPOSITION FOB SPECIFIED PRODUCT COMPOSITION

• At 1265''F from Fig. 168 Design Basis
Mole Per Cent Mole Weight
Per Cent PerCerU
Benzene 62.2 63.36 63.36
Diphenyl 13.2 12.80 25.26
Triphenyl 3.8 3.68 10.86
Hydrogen 20.8 20.16 0.52
100.0 100.0 100.0
design. In these calculations expressions for the values of Ap, AVr — A(1/<S„) and
AA, the heated surface area, per tube are required.
Equivalent length of tube, feet, = 10 + (50) (2) /12 = 18.33. From Equation (98),
(18.33) (0.0235) (4.036)'-^ (359) (zT) M"-^ (14.7) '
Ap, lb per sq in., per tube — ^ g
2*-* (78.1) (492) (p)
0.0262Z'T(A.')°-^
= z ' (a)
where 2' = mean compressibility factor
T = absolute temperature, degrees Rankine
p = absolute pressure, pounds per square inch
/i' = mean viscosity, micropoises
AF,, (359 cu ft) per tube = 0.23/359 = 0.641 (10-']
A(l/S,) per tube = AVr IF = (0.641) (IQ-^)/(51.68) = 0.124(10-<)
AA sq ft, per tube = (9.2)(27r)/(12) = 4.82
The calculations for the first trial analysis of the reactor and the preheater, working
back from the reactor outlet, are summarized in Table C. In order to develop a
basis for estimating the changes in conditions per tube, the first reaction section con-
sidered comprises only the outlet tube. In this analysis the tubes are numbered
back from the outlet.
In lines 5-10 are entered the specified reactor-outlet conditions based on 1.0 lb-
mole of feed while in lines 11-16 are estimated average conditions in the section. The
average composition is arrived at by first estimating average values of no and nr.
Then nn = 2nr + no, and ns is determined by difference.
On the basis of the estimated average composition, the pseudocritical tempera-
ture, pressure, and viscosity and the corresponding pseudoreduced temperature and
pressure are calculated. Values of the compressibility factor, z , {Cp ~ c'p), and M' are
read from Figs. 103, 108, and 175. Ideal molal heat capacities and heats of reaction
at the average temperature of the section are obtained from Fig. 177. Heat-capacity
yalues are not required in the reactor section where there is no change in temperature.
Values of kvh and K are read from Fig. 176. At the moderate pressures of this
process the fugacity coefficients of both the activated complex and the reactants and
products may be neglected. On this basis the reaction rates are calculated from the
following equations:
fci^V(7r) =
ri =
z'
A;2>'V(ir)2r nrnffl
n=
0
Thus, in the first section,

(300) (32)^ [(0.64P_(0i26^:15S0)J^^^
ri = •
(14.7)
Then, since from Equation (87), •
Ml IS,) = (AV)KF)
Ani = - [ ( n + r2)][A(l/S.)] (d)
A»B = [r,/2-r-2][A(l/S„)] (e)
Anr =r2A(l/S„) (f)
An^=[(r,/2)+rd[A(l/S.)] (g)
The incremental conversions calculated from these equations are subtracted from
the corresponding numbers of moles at the outlet of the section in order to obtain the
compositions at the inlet, which are entered in the next column as the outlet con-
ditions of the next section. The estimated average conditions of the section are
compared with these inlet and outlet values, and if any of the estimated values are
badly in error the calculation is repeated.
The heat of reaction per mole of feed of each of the reactions is calculated in
lines 44 and 45. In the reactor section where there is no change in temperature the
sum of these heats of reaction times F the molal feed rate equals q the heat absorbed
per hour in the section. Where there is a temperature change,
q = F[{AHO (|) (-Ans - Any) + (AH,) (Anr) + c'ÂT] (h)
The temperature change corresponding to a specified heat-input rate is calculated by
rearranging equation (h).
qlF - (Agi)(l)(-AWfl - Anr) + (AH;)(Arar)
Ar = 7 (i)
Cp
The pressure drop in the section is calculated from Equation (a). Thus, for the first
section,
(0.02622) (1710) ,„„„,„, , ,
Ap = ^ —" (208)''-= = 4.1
Based on the conversion and pressure drop in the first tube, the average conditions
in the section comprising the next five tubes are estimated, it being remembered that
the rate of conversion is markedly affected by the increase in pressure. The conver-
sion and pressure drop,in these tubes are then calculated, and the same procedure is
repeated for another section of five tubes.
It may be noted that in the first 11 tubes of the reactor the mole fraction of benzene
is increased from 0.83 to 0.86. Since the reaction rates are high in tubes 7-11 it is
evident that another section of five tubes at the constant temperature of 1250°P
would account for substantially all of the desired conversion, leaving none to be
accomphshed in the preheater. Accordingly, it will be assumed that tube 11 is the
last tube of the reactor section and that following tubes are heated at a uniform rate
of 10,000 Btu /(sq ft) (hr). On this basis calculations are carried out for tube 12 and
then for the section comprising tubes 13-15. The results of these' calculations are
summarized as they are carried out in Fig. 178 in which conditions at the tube outlets
are plotted against tube numbers. The cuhres representing these first calculations
for tubes 12-15 are shown by broken lines. It may be noted that these broken lines
for the moles of diphenyl and triphenyl are rapidly approaching zero at temperatures
\
CHAP. X V I I I ] DIPHENYL REACTOR 881
far above the specified temperature at the inlet to the preheater. This situation
indicates that the number of tubes assumed for the reactor section is too large.
As a second approximation, it is assumed that only six tubes are required for the
reactor section. On this basis the changes in tubes 7-9 are recalculated to correspond
to the heat-input rate of 10,000 Btu/(sq ft)(hr) specified for the preheater section.
These calculations are repeated, as indicated in Table C and Fig. 178 for successive
sections of 3, 3, 5, 5, and 10 tubes until the conditions at the inlet of tube 35 or the
0.16
40 30 20 10 0
Tube Numbers from Reactor Outlet Tube 1
FIG. 178. Conditions at Tube Outlets of the Diphcnyl Reactor and Preheater.
outlet of tube 36 are evaluated. It may be noted that at these conditions the rates of
reaction are very low and that the number of moles of diphenyl is close to zero while a
shght negative conversion to triphenyl is indicated. These differences from zero
are less than the probable errors of the calculations and indicate that the assumption
of six tubes for the reactoCpection is satisfactory.
By extrapolating the temperature curve of Fig. 178 to 700°F the specified inlet
temperature, it is seen that a total of 36 tubes is required. Thirty of these tubes are
used in the preheater and six in the reactor sections. Extrapolation of the pressure
curve indicates a pressure of 96 lb per sq in. at the inlet of tube 36 which is the inlet
of the preheater.
The fortuitous choice of six tubes tor the reactor section made further calculation
unnecessary. If five or seven tubes had been assumed for this section, the corre-
sponding calculated conversions at the preheater inlet would have been, respectively,
iOt*iO OS 1-H CO
A : : •^TjJo) o
ci<S3ai o
S
o
8
o
o o o o
<005t» CO «-i to
• '»-« ifS »-i CO
OMO O O O
•âOOi O O O
O ' - ' b - O to CO
d di ci ii ddSo '^
CJtOOO O O O
^ dodo
111 i piiqps ^ps22iis§§iPi i I p ^ gas
d ddd dddd "^
Biiii dddd d d d d ' ^

'"J'^ t ^ CO ' ^
— - : 2 ' - ' ^ c»
CIOSO t-( <M U5
- " - " t * 1-) O i-t
ri i i peiBp^s sîn|dSi^llP| l i p s ip:^
^liiii dddd o d d d " ^
III
oo>«
^cooo
d6do S <6 <6 d d d <6 <D ^
S*3
ail riilpsislpS :::::::iPî!Piiiiigll
*? . H U J O O
côso
•5
d odd dddd oddd
Qoocq
ill i\° i i fiSg^gpS : : : : : : ^^^1^^ f § f l 11°|

o o o o dddd dddd '^
05 - c o
NOiCO
ill dddd dddd
CiOi
C0P5
II
11;^
11 •d
if.e S
«S2"
g 3 3 O.
I
•IgiJ Hi
l aWgflj • • • ^_ • •».
mifff
'.ffl ** ^ ^ ' i ^ l i , O
?S
eel l|i
'S ^::
CHAP. XVIII] OPTIMUM REACTOR DESIGN 883
greater or less than zero. In this case the final design could have been arrived at by
repeating the calculations, starting from the preheater inlet with a pressure assump-
tion based on the first calculations.
The problem of Illustration 11 is solved by a relatively compUcated procedure
which insures accurate results if reasonably small increments are considered. In
many cases this degree of accuracy may be greater than justified, and many simplifi-
cations are possible, particularly for the preliminary designs necessary to explore
various ranges of operating conditions. After experience is gained in such calcula-
tions, it is possible to estimate average conditions in succeeding sections with such
accuracy that corrections are rarely required.
Optimum Reactor Design. The optimum reactor design for a given

process is that which results in the manufacture of the desired product at
the lowest cost. In arriving at the optimum design, costs and credits
from all sources must be considered", including all fixed and operating
charges, raw-material costs and by-product credits. A series of designs
may be developed in which the basic design factors are systematically
varied and the costs of the product from each design are calculated. The
combination of design factors which produces the lowest over-all cost is
then determined by graphical interpolation.
The principal design factors generally considered are composition,
temperature, pressure, or concentration, and space velocity or time.
Since the optimum value of any one of these factors is dependent on the
values of the others, itjs theoretically necessary to explore all possible
combinations. This majy be done by plotting production costs against
space velocity at several different temperatures, all corresponding to a
constant pressure. This procedure is then repeated for other pressures.
The lowest product cost for each pressure is then plotted against pressure
to determine the over-all optimum. The corresponding values of
temperature and space velocity are determined by interpolation between
the values at the optimum for each pressure.
Such a complete analysis is rarely carried out because of the great
amount of work involved. A minimum of 27 design studies and corre-
sponding product costs are required for exploration of the economic
effects of all three variables. Instead of carrying out such detailed
calculations it is common practice to establish approximate relationships
between the optimum values of the different factors and to fix as many as
possible by considerations of practical expediency. For example, in the
reaction of Illustration 11, the temperature is fixed at the highest possible
in the available materials. The space velocity corresponding to a given
temperature and pressure is fixed by the specification of product purity
which limits the conversion of benzene to approximately 37 per cent.
With these factors fixed only the effects of different operating pressures
need be investigated from an economic standpoint. A higher pressure is
helpful in reducing the size of the reactor but introduces added costs of
pumping and increases the cost per unit volume of the reactor. In
order to determine the optimum pressure, designs are worked out for a
series of pressures, and the costs are evaluated. In general, such
economic optimum conditions are not critical, and operations can be
carried on over a wide range of conditions with little change in costs.
For this reason it is not ne'cessary that the optimum be evaluated with a
high order of accuracy. The general effects of the individual operating
variables are as follows:
Temperature. In general the highest possible operating temperature
is the most economical for endothermic reactions or for any operation in
which reverse reactions are negligible. For exothermic reactions which
approach equilibrium the problem is more complicated. In such case
increasing the temperature increases the rate of reaction but at the same
time makes the equilibrium composition less favorable. As a result a
definite optimum temperature exists which for a specified space velocity
and pressure will result in maximum conversion. Calculations of such'
optimum temperatures for catalytic systems are discussed on pages 1020
to 1028.
Pressure. Increased pressure increases the production rate or con-
version of any reactor if reverse reactions are negligible. Where a
decrease .in number of moles accompanies the reaction, increased pres-
sure also favors the equihbrium composition. However, where an
increase in the number of moles is involved, an optimum pressure will
result in maximum conversion if equilibrium is approached. This effect
is similar to that of temperature.
(Space Velocity. In a reactor operating at fixed temperature and
pressure the extent of conversion is determined by the space velocity in a
flow process or the time in a batch process. Reduction of space velocity
increases conversion and reduces costs of recycling. On the other
hand, the cost of the reactor is increased and frequently an increased
formation of undesirable by-products results in a poorer ultimate jaeld
of desired products.
Equivalent Reactor Volume. • In experimental operations it is fre-
quently difficult to maintain a constant temperature throughout a flow
reactor or during the entire cycle of operation of a batch reactor. The
rigorous interpretation of such, data becomes comphcated if large tem-
perature differences are involved, particularly where reverse reactions are
significant. For this reason it is important that isothermal conditions be
approximated as closely as possible.
Where relatively small temperature differences exist in a flow reactor
, satisfactory results can be obtained by calculating the equivalent volume
CHAP. XVIII] EQUIVALENT REACTOR VOLUME 885
of the reactor which at a constant reference temperature Ti, would pro-

duce the same conversion as the actual reactor with its varying tempera-
tures. In order to calculate such an equivalent reactor volume Vh, it is
necessary to assume an approximate value of the average enthalpy of
activation A ^ ^ of the reactions taking place. If an intermediate refer-
ence temperature is selected somewhat lower than the maximum tem-
perature, the value of the calculated equivalent volimie is not greatly
affected by moderate errors in Aff^. Where experimental data at two
different temperature levels are available, the equivalent reactor volimies
at the two reference temperatures are calculated from a trial or assumed
value of A^J,. Rate equations may then be developed by treating these
data as though they were obtained in isothermal reactors operating at the
reference temperatures, and a corrected value of A/?|, is then calculated.
If the assumed values were grossly in error the calculation is repeated.
The same principles^jnay be followed in calculating the equivalent
time of reaction in a batch process.-*. From Equation (55) the time drb
required at reference temperature Tb t o produce the conversion dxA
resulting from time dr at the existing temperature T is given by the
following equation
dn = dT{h/h) (104)
where h and hh are the reaction velocity constants at the existing and the
reference temperatures, respectively. Similarly, from Equation (80)
for a flow process,
d{Vr/F)b = d{Vr/F)(k/h) • (105)
Combining (104) and (105) with (18) and neglecting deviations from
• ideal-gas behavior gives
dn = dr Le « vr nJ j (i06)
d(Vr/F)i = d{Vr/F) le « Vr TJ] (IO7)
Equations (106) and (107) may be integrated for a reactor or process of

known temperature distribution. Thus,
I e R KT Ti)dr (108)
0
{VrIF)i= / e-~BrKT-Yj d{Vr/F) (109)

t/0
In using Equations (108) and (109) it is convenient to plot a curve

relating the faxitor e ^ v^ î"' on a logarithmic scale to temperature

on a uniform scale for the particular reaction under consideration. The
total equivalent time or volimie in any reactor is then readily calculated
from the known relationship between temperature and either T or Vr
by graphical integration of Equation (108) or (109). This is accom-
Afl^m / I 1 \
plished by plotting e ^ v^ nJ as ordinates against r or F , as
abscissas and determining the area under the curve between the terminal
values of T or F,. In this manner experimental data may be corrected
to a base temperature to facilitate interpretation where temperatures
were not' constant. However, this procedure is subject to errors where
several reactions occur simultaneously with different enthalpies of activa-
tion. For such systems a close approach to isothermal conditioris is
desirable.
Severity Factors. In dealing with reactor data and designs where
both temperature and pressure vary it is frequently convenient to corre-
late conversions and yields in terms of a seventy factor which is a measure
of the severity of the over-all reaction conditions. This semiempirical
approach is particularly desirable for complex reacting systems such as
the cracking of petroleum where many reactions occur simultaneously
and the complete compositions of both the charge and the products are
generally unknown. Although there is much evidence that cracking
occurs through complex chain reactions, the over-all result may be
treated as approximately equivalent to a first-order reaction in which the
reaction velocity constant diminishes as conversion of the charge
progresses. Thus,
k = A-o(l - ax) (110)
where
k = effective average reaction velocity constant corresponding to the
range of fractional conversions from 0 to x
ko = effective initial reaction velocity constant for the unconverted
charge
a = a proportionality factor depending upon the characteristics of the
charge
Equation (110) is frequently satisfactory only over moderate ranges of

conversion. Where high conversions are involved the following alter-
nate form,
k = /fo/(l -i a'x) ' (111)
may give better correlations.

CHAP. XVIII] SEVERITY FACTORS g87
Combining Equations (110) and (84), taking nô as 1.0 and assuming
VA = 1.0 gives
r^r, dVr C- (1 + wa;) dx
/ hT—-^ ,, ,,/ r (112)
Jo rioF Jo (1 - ax)(l — x) ^
If it is assumed that the fugacity coefficient of the activated complex v^
is constant, ko may be expressed in terms of (^-0)6, the initial reaction
velocity constant at the reference conditions of temperature Tb and
pressure TB where the compressibility factor of the feed is z^. Thus,
from Equation (18) written for the two temperatures,
ko= {ko)b-ê~~ir\.T-rJ (113)
Combining (113) and (112) and evaluating the second integral, we have
{ko)b I — e B KT n/—-
Jo z' noF
= -a(l
J T—^ a) 1" ^1 - «^) - T(1T —
^ l a)
' ^ (1 - ^) = (^0)/.
* (114)
The first integral of Equation (114) is a severity factor F, referred to the

base conditions Tb and unit pressure. This severity factor represents
the reactor volume operating at unit pressure and the selected base
temperature Tb which is equivalent to the actual reactor volume per unit
molal feed rate. Its dimensions are (pressure) (vol) (time)/moles. It is
evident that the severity factor is increased by increased pressure p
ajj^d temperature T and is reduced by increase in the mean compressi-
bihty factor z or the molal feed rate per unit volume of reactor, rioF/Vr.
If the reaction occurs under such conditions that variations in the com-
pressibility factor may be neglected, the severity factor is readily deter-
mined from the distribution of temperature and pressure throughout the
reactor. This procedure is indicated diagrammatically in Fig. 179
where reactor temperature and pressure are respectively plotted against
VT. From these two curves and the enthalpy of activation AH^ are
AHt / I 1 \
derived a curve relating xe R \^ Tb) to Vr/nJP. The area under
this curve is equal to the severity factor.
If the compressibility factor z' of the reacting mixture caimot be
assumed to be constant, the severity factor must be evaluated by means
of a stepwise integration similar to that demonstrated in lUustration'll.
In this manner variations of the compressibility factor with temperature,
pressure, and conversion are taken into account. These variations are
particularly important if a portion of the reaction occurs with the charg-
0 1 2 3 4 6 6 7
Volume of Reactor, Vr
FiQ. 179. Evaluation of Severity Factors where Compressibility Factor Clfanges
are Negligible.
1.0
/ ^ =0
02
§0.4
T/r
u ^
; ^ ^
01=2.0
nj__
,0.2
0 2 4 6 8 10 12 14
Pio. 180. Relation between Conversions and Severity Factors.

CHAP. XVIII] PREDICTION OF ENTROPY AND ENTHALPY 889
ing stock only partially vaporized. Under such conditions the com-
pressibility factor z' is equal to pv/RT for the entire two-phase mixture
and accordingly may vary widely with the degree of vaporization.
The severity factor may be employed as a basic factor for the correla-
tion of conversions, yields, product distributions, and qualities in com-
plex reacting systems. The relationship between the severity factor and
conversion x may be obtained by evaluating the integral on the right side
of Equation (112). In this manner a family of curves may be con-
structed such as is shown in Fig. 180. Each curve corresponds to a
different class of feed stocks for which values of co and a are specified.
Where Equation (114)'is not applicable, curves relating conversion to the
severity factor may be empirically evaluated.
Similar severity factors may be defined and used for semiempirical
correlations of rate and yield data in reaction systems which do not
approximate first-order behavior. Such factors will generally involve a
pressure term with an empirically determined exponent.
PREDICTION OF ENTROPY AND ENTHALPY OF ACTIVATION

Eyring and coworkers have developed detailed methods for predicting
both the entropy and the enthalpy of activation. The enthalpy calcu-
lation is complicated and uncertain in its present state of development.
The. entropy of the activated complex is determined by the statistical
methods outlined in Chapter XVII. This calculation requires that a
completely defined structure be assigned to the activated complex with
all atomic distances and vibration frequencies evaluated. General rules
are not yet available for this procedure, and considerable latitude is
exercised in the assignment of structures which result in calculated
entropies and enthalpies in agreement with experimental values.
In view of the complexity and the judgment required in detailed
calculations it appears that the best immediate engineering appUcation
of the theory of absolute reaction rates is in the rough approximation of
entropy and enthalpy of activation. With an estimated entropy the
enthalpy of activation may be calculated from a single determination of
the reaction velocity constant where only one chemical step is involved.
The resulting equation is useful for estimating the effect of temperature
on the rate of reaction over limited temperature ranges, even though the
entropy and energy of activation are considerably in error. For example,
at 1000°K an error in entropy of activation of 10 units corresponds to an
error in enthalpy of activation of 10 kcal per g-mole. The enthalpies of
activation of high-temperature reactions range from 30 to 100 kcal per
g-mole, and in several instances the disagreement between experimen-
tally determined enthalpies of activation is greater than 10 kcal.
Similarly, if the enthalpy of activation is estimated from an empirical

rule, the corresponding entropy of activation may be calculated from a
single experimental result. If rules of the same degree of reliability are
available for estimating both entropies and enthalpies of. activation, the
following procedure may be followed for arriving at a complete equation
from a single experimental-reaction velocity constant:
1. Estimate the entrop;^ of activation by the empirical rule.
2. Calculate the corresponding enthalpy of activation from the experir
mental-reaction velocity constant and Equation (17).
3. Estimate the enthalpy of activation by the empirical rule.
4. Average the enthalpies of steps 2 and 3.
5. Using the average enthalpy of activation calculate the correspond-
ing entropy of activation from Equation (17).
Another useful application of approximate rules for estimating both
entropy and enthalpy of activation is in the estimation of approximate
relative reaction rates to serve as guides in selecting the rate-determining
steps of chain reactions. Even though the absolute rates estimated in
this manner are greatly in error, they may serve to eliminate the neces-
sity of further consideration of reactions which are extremely fast or slow
in comparison to the order of magnitude of the over-all rate. Methods
for predicting reaction rates are reviewed by Daniels.'^
Empirical Rules for Entropy of Activation. The following rules may
be used for the rough estimation of the entropy of the activated complex
at 298°K and 1 atm.
Reactions Unimolecular in Both Directions. Isomerization or re-
arrangement reactions of complex molecules which do not result in the
formation or breaking of a ring appear generally to form an activated
complex having a lower entropy than the reactant or product molecules.
This is presumed to result from the formation of temporary multiple
bonds in the complex which eliminate opportunities for internal rotation.
The resulting loss in entropy is dependent on the moments of inertia and
potential barriers of the groups whose rotation is stopped and also on the
vibrational contributions added by the temporary bonds. For complex
molecules containing more than ten atoms it may be assumed that an
average loss in entropy of ten units accompanies the formation of the
complex. The entropy of the activated complex is determined by
averaging the results obtained on this basis from consideration of the
forward and reverse reactions. Thus, for the reaction,
S' = ^ ^ ^ - 10 (115)
" F. Daniels, Ind. Eng. Chem., 35, 504 (1943).

CHAP. X V I I I ] PREDICTION OF ENTROPY AND ENTHALPY 891
Then, for the forward reaction,

AS* = S^-SA (116)
and the reverse reaction,
AS* = S^-SB (117)
Unimolecular reactions resulting in ring closure would be e?cpected to
proceed at an abnormally slow rate because of the high decrease in
entropy due to restrictions in internal rotations, whereas ring-opening
reactions would be expected to proceed at an abnormally rapid rate for
opposite reasons. For example, the decomposition of ethylene oxide
results in ring breakmg in the formation of the activated complex with
an increase of 7.5 entropy units. Similarly, reactions which proceed by
a chain mechanism proceed rapidly, in part owing to the increased en-
tropy in the formation of free radicals.
Unimolecular-Bimolecular Reactions. Where a simple unimolecular
reaction produces two product molecules, it appears that the entropy of
the activated complex may be taken as equal to that of the reactant
molecule. Thus, for the reaction,
Av±X^v±R + 8 (118)
For the forward reaction,
AS* = 0 (119)
For the reverse reaction,
AS*' = SA- SB-SS^ -AS (120)
This rule is based on the finding of Daniels^^ who pointed out that the
frequency factor kT/h at 300°K is equal to 0.63(10i'). From a survey of
28 of the most reliable unimolecular reactions Daniels^^ found that for
82 per cent the value of -— e R ranged from lO^^ to 10" at 300°K,
h
indicating that AS* was very small.
Reactions Bimolecular in Both Directions. When two molecules form
an activated complex which decomposes to form two other molecules
without the formation or destruction of a ring, it may be assmmed that
for complex molecules the external rotational contribution to the entropy
of the complex is equal to the sum of the rotational contributions of the
reactants. On this basis the entropy changes accompanying the for-
mation of the complex result from changes in the translational contribu-
tions and the vibrational and internal rotational contributions. The
changes in the translational contributions are readily calculated from
Equation (XVII-12), and it will be assumed that the formation of the
complex is accompanied by an increase of four units in vibrational and

internal rotational entropy. The entropy of the activated complex is
taken as the average of the values obtained by applying these rules to
the forward and reverse reactions. Thus, for the reaction,' with each
component in the ideal-gaseous state at 1 atm and 298°F
Si' = SA+SB + (fi2 In Mx + 26.00) - (|i2 In MA + 26.00)

- i§R In MB + 26.00) + 4: = SA + SB
Similarly,
>SL = 5ij + & + | / 2 In ^ r ^ - - 22.0 (122)
where
Mx, MA, MB • • • = molecular weights oi X^, A, B • • •
Si = {Sh + Sh)/2 (123)
For the forward reaction,
AS* = SIC-SA-SB (124)
For the reverse reaction,
AS* = Si ~SE-SS (125)
Empirical Rules for Enthalpy of Activation. Empirical rules for

• estimating enthalpies of activation have been proposed by Hirschfelder"
on the basis of the assumption that the energy of formation of a chemical
bond in the ideal-gaseous state from the component elements as ideal
monatomic gases is in general independent of the nature of compound in
which it exists. On this basis values of bond energies have been devel-
oped which are given in Table LVI.^^ From this table approximate
values of heats of reaction in the ideal-gaseous state may be obtained by
subtracting the sum of all the energies of bond formation of the reactants
from the corresponding, sum of the products. However, the values
obtained in this manner are less reliable for complex compounds than
those obtained by the method of group contributions described in
Chapter XVII for heats of formation.
Hirschfelder proposed the following rules relating to bond energies and
I enthalpies of activation. These rules must not be applied to the over-all
1 result of a chain reaction but are applicable to any of the intermediate
single steps.
" J. O. Hirschfelder, J. Chem. Phys., 9, 645 (1941),
CHAP. XVIII] ENTHALPY OF ACTIVATION 893
Dissociation Reactions. In general, it may be assumed that the

enthalpy of activation of the combination of two free radicals or atoms is
zero under conditions which are thermodynamically unfavorable for
their existence in appreciable concentrations. On this basis, for the
dissociation of a molecule AB into atoms or radicals A and B according
to the reaction,
AB:Â+B
Aff* = AH (126)
AH*' = 0 (127)
Reactions Bimol^lar in Both Directions. Reactions of this type may

be represented by the general equation:
AB + CD:ÂC + BD .
TABLE LVI
APPROXIMATE BOND ENERGIES^'
Kcal per G-Mole Absorbed in Bond Rupture
C—H 92 C=G 122 = N — N = 33
C—C 79 C=C 200 C — 0 82
C—Br 58 H—H 102 C = 0 188
C—I 44 0—H 113 Br—Br 45
C—C 77 N—H 96
According to Hirschfelder the enthalpy of activation of such a reaction

in the exothermic direction is equal to 28 per cent of the energy required to
dissociate the reacting molecules. Thus,
AH^ = 0.28{EAB + ECD) when AH is negative (128)
AH^' = -AH + AH^ (129)
Atomic-Molecular Reactions. Reactions between a free radical or an
atom and a molecule may be represented by the general equation:
A+BCÂB +C
Hirschfelder proposes that for exothermic reactions of this type
AH^ = 0.05EBC when AH < 0 (130)
AH^' = -AH + AH^ (131)
These rules give results in fair agreement with the relatively few reliable
experimental data which are available.
Illustration 14. Estimate the entropy of activation AS* in concentration, units
of the formation of HI from gaseous hydrogen and iodine at a temperature of 298°K.
Solution: The entropies of the reactants and products at 298°K and 1 atm are
obtained from Table XXXV, page 701.
Sm = 31.23 Mm = 2.0
Sij = 62.29 iWij, = 254
SHI = 49.36 Mm = 128
Hj + Ij ?± (2HI)* 5=i 2HI
F r o m E q u a t i o n (122), ^
S^zi = 31.23 + 62.29 + 2.98 In ( - S T T ) - 22.0 = 69.5

\(2)(254)/
>SL = (2) (49.36) + 2.98 In ( T ^ ) - 22.0 = 64.3
S^x = (69.5 + 64.3)/2 = 66.9

A<S{298'>Ki = 66.9 - 31.23 - 62.29 = - 2 6 . 6
Glasstone, Laidler, and Eyring^ report a value of AS^ (300°K) =
— 20.3 for t h e reaction of Illustration 14. I n this case t h e simple rule
for estimating t h e entropy of activation led t o a n error of 6.3 units. F o r
more complex molecules this error might be reduced somewhat by im-
proved additivity of t h e rotational contribution, b u t it should be recog-
nized t h a t rules of this t y p e m a y lead t o errors of t e n entropy units or
more for simple reactions a n d are n o t applicable t o t h e over-all results of
chain reactions.
PROBLEMS
1. Calculate the reaction velocity constants from the following data of Schu-
macher"^' for the designated unimolecular reactions, assuming ideal behavior. Ex-
press rates in g-moles /(liter) (sec) with activities in atmospheres.
AH^ AS*
T KCal Iper G-Mole {Entropy Units)
(a) Decomposition of di-
methyl ether, CH3OCH3 780°K 56.95 2.5
(6) Decomposition of methyl
azide, CH3N3 500°K 42.5 8.2
(c) Decomposition of tetra-
methyl silicane Si (CHs)^ 950°K 76.9 2.2
(d) Decomposition of azo
methane CH3N NCH, 600°K 48.8 10.8
Calculate the percentage error resulting from assuming that AS* = 0 in each of
the afore-mentioned cases.
2. The kinetic and thermodynamic behavior of a reaction of the type A + B =
B -f S in an ideal-gaseous system is expressed by the following data:
Reaction velocity constants:
fc450°F = 0.0061 g-mole/(liter) (atm^) (see)
fcsoo-F = 0.0239 g-mole/(liter) (atm^) (sec)
" H. J. Schumacher, " Chemische Gasreaktionen," T. Steinkopf (1938).
CHAP. X V I I I ] PROBLEMS 895
The standard enthalpy and entropy changes may be assumed to be constant over the
temperature range of interest.
AH" = 8,400 cal per g-mole
AS° = -2.31 cal per (g-mole) (°K)
(o) Derive an equation for the reaction velocity constant of the forward reaction
as a function of temperature.
(b) Evaluate AH* and AS*.
(c) Evaluate an expression for the reaction velocity constant of the reverse reac-
tion as a function of temperature.
3. The rate of a reaction is reported as 2.46(10"^) lb-mole/(hr) (cu ft). Express
the rate in
(a) Molecules/(cu cm) (sec).
(6) G-moles/(Uter)(sec).
4. In the rate equation,
r = kA
kc = 0.0031 at 700°K
where -r = g-moles/(min) (22.41 liters)
c = g-moles/(22.41 liters)
Calculate the value of k for the same reaction when r = fcp^.
r = lb-moles/(cu ft)(min)
PA = atmospheres
6. A first-order reaction in an ideal-gaseous system at 680°R has a reaction
velocity constant of 0.14 lb-mole /(359 cu ft) (atm) (hr). Calculate the corresponding
value of kc in 1 /sec.
6. A third-<)rder reaction in an ideal-gaseous system has a reaction velocity con-
stant kc = 140 (hters)2/(g-moIe)^/(sec). Calculate the value of k in lb-moles/
(359 cu ft) (atm)' (hr) as a function of T°R.
7. The reaction velocity constant for the dimerization of butadiene is given by
23,900
Vaughan" as fcc = 9.2(10')e *^ cc/(g-mole) (sec). Estimate the values of AH*
and AJS* at 600°K, assuming ideal behavior and a standard state of unit fugacity.
8. Glasstone, Laidler, and Eyriag^ report the reaction velocity constant for the
48,000
isomerization of isostilbene as 6.0(10'2)« ^^ reciprocal seconds. Calculate the
values of AS* and AH* at 600°K for a standard state of 1 atm fugacity, assuming ideal
behavior.
9. In the dimerization of ethylene, ACp for the reaction is —6.0 cal per g-mole of
complex. From Illustration 1 the value of AH* is 32.73 kcal and that of AS* is
—35.00 both at 673°K and a standard state of 1 atm fugacity. Calculate the values of
AH* and AS* at 1000°K and the corresponding value of the reaction velocity constant.
10. Hydrogen peroxide in aqueous solution is decomposed by the homogeneous
catalytic action of alkaUes with Uberation of oxygen. The reaction is of the first
order, and the reverse reaction is negligible. In a 0.04 normal NaOH solution at
60°C, h is 0.0389 per min; at 20°C, h = 0.00106.
A bleach bath of 2,000 kg contains H2O2 equivalent to an available oxygen content
i« W. E. Vaughan, J. Am. Chem. Soc, 64, 3863 (1932).
of 40 per cent by volume and NaOH equivalent to 0.04 normality. The density of
the bath may be taken as 1.0.
The percentage available oxygen content refers to the volume of oxygen gas at
1 atm, 0°C., which may be liberated from the hydrogen peroxide expressed as a per-
centage of the volume of the liquid. For a period of 1 hr at 60°C calculate:
(o) The percentage decomposition of hydrogen peroxide.
(6) The grams of H2O2 decomposed.
11. Five hundred grams of phosphine gas are heated in a tube at 672°C., 1 atm
pressure, for 200 sec. Dissociation proceeds according to the reaction, 4PHj(g) —•
PiCg) + SHjCg), and is of the first order. The reverse reaction is negligible. The
reaction velocity constant kc per sec is expressed as a function of temperature accord-
ing to the following equation given by the International Critical Tables:
18963
logiofcc= - + 2 logio T + 12.130 where T i§ in degree Kelvin.
T
Calculate:
(a) The grams of phosphorus formed.
(b) The time required to decompose 95 per cent of the phosphine at a temperature
of 672°C.
12. The saponification of ethyl acetate by alkalies in aqueous solution is a rela-
tively rapid reaction; the reverse reaction is negligible. For a second-order reaction
the reaction velocity constant is given by the International Critical Tables where the
initial normality of caustic soda is 0.05, as
1780
logio kc = + 0.00754r + 5.83
where kc is the reaction velocity constant in liter / (g-mole) (min) and T is in degrees
Kelvin. Calculate the time required to saponify 98 per cent of the ester at 40°C
when the initial concentration of ethyl acetate is 2.0 g per liter, and the initial nor-
mality of NaOH is 0.05.
13. The decomposition of nitrous oxide proceeds as a second-order reaction,
2N2O —> 2N2 + O2. The reverse reaction is negligible. At 895''C, the International
Critical Tables give the value of % as 977 (cc) (g-mole) /sec. Calculate the time
required to decompose 90 per cent of the N2O at 895°C:
(o) When the volume is kept constant, initial pressure 1 atm.
(6) When the pressure is kept constant at 1 atm.
14. One hundred grams of acrolein are treated with 120 g of 1,3-butadiene in a
5-liter vessel to form 1, 2, 3, 6-tetrahydrobenzaldehyde. The weights of the alde-
hyde formed in various time intervals are given herewith for constant temperatures
Grams Grams Grams
of of of
Tern- Alde- Tem- Alde- Tern- Alde-
peror hyde Time, pera- hyde Time, pera- hyde Time,
lure Formed Min ture Formed Min ture Formed Min
200°C 31 4
150°C 43 100 44 8 250°C 45 1
73 200 57 12 63 2
90 300 68 16 83 3
109 400 73. 20 96 4
111 500 83 24 113 5
CHAP. XVIII] PROBLEMS 897
150, 200, and 250''C, respectively. Evaluate the reaction velocity constants for this
reaction and the enthalpy of activation, AH^.
15. From the work of Cain and NicoU^" the following rate data were obtained on
the decomposition of diazobenzene chloride in aqueous solution at 50°C:
C6H5N2CI -> CeHsCl + N2
The initial concentration of the diazo salt in solution is 10 g per liter, and the sample
chosen liberates 58.3 cc of nitrogen at 60°C and atmospheric pressure when com-
pletely decomposed. Calculate the reaction velocity constant kc for the first-order
reaction:
Time (min) 6 9 12 14 18 20 22 24 26 30
Gas liberated (cubic centimeters)
19.3 26.0' 32.6 36.0 41.3 43.3 46.0 46.5 48.4 50.35
From the following temperature data calculate the value of AH^:
fC 20 30 40 50 60
h 0.00166 0.00678 0.0210 0.0688 0.251
The reverse reaction is neghgible.
16. From the data of Table C, Illustration 9, page 848, evaluate the average reac-
tion velocity constants of reactions (1) and (2) by graphical integration using equa-
tions similar to (94). Compare these results with those of Table D, page 850.
17. By reference to Illustration 10 calculate the values of the reaction velocity
constants, kct,feci,fcceandfccsfor the chain reaction involved in the hydrogenation of
bromine gas from the following experimental data from Bodenstein and Lind.^'
At 277.5°C, a = 0.2881; 6 = 0.1517.
a = initial concentration of hydrogen, gram-moles per 22.4 liters
6 = initial concentration of bromine, gram-moles per 22.4 liters
T X T X
120 0.0359 840 0.1227
240 0.0632 1020 0.1308
360 0.0811 1320 0.1389
480 0.0957 1680 0.1442
X = gram-moles of hydrogen or bromine reacted per 22.4 liters,
18. Dimethyl ether upon heating without catalysis decomposes by a substantially
irreversible first-order reaction into CH4, CO, and H2. The reaction velocity con-
£8,500
stant is given in the International Critical Tables as fc = 1.55(10i')e ^^ (1/sec).
(0) One hundred grams of dimethyl ether are heated at 500°C at constant volume
conditions in a closed retort of 120 liters capacity for a period of 30 min.
(1) Calculate the percentage conversion of the ether.
(2) Calculate the final pressure in the closed retort.
(6) One hundred grams of ether are heated in the same retort open to the atmos-
phere. The ether which escapes is not subject to further decomposition, and decom-
position during preheating to 500°C may be neglected. The heating is continued for a
period of 30 min at 500°C.
"1 J. C. Cain and F. Nicoll, Proc. Chem. Soc., 24, 282 (1909).
" Bodenstein and Lind, Z. Physik. Chem., 57, 168 (1907).
(1) Develop a differential equation expressing the rate of disappearance of ether

from the retort, both by reaction and expulsion, as a function of the number of gram-
moles of ether Ue present at any time T after the reaction temperature is reached.
(2) Calculate the percentage decomposition of the ether originally charged.
(3) Calculate the percentage of undecomposed ether lost by escape from the open
retort.
(c) Dimethyl ether is passed at 500°C and 1 atm pressure in a continuous stream
through a reacting cyUnder of 120Ûters capacity at a constant entering rate of 5 g per
min. Calculate the percentage of decomposition of the ether.
19. From the data of Illustration 5 calculate the reactor volume required to pro-
duce 90 per cent decomposition of a stream of NjO in a flow system operating at a
temperature of 895°C and a pressure of 2.0 atm with a feed rate of 650 lb per hr.
20. From the data of Illustration 6 calculate the volume of the flow reactor
required to produce 90 per cent of the equilibrium hydrolysis of the methyl acetate
in the specified solution when fed at a rate of 150 gal per hr at a temperature of 25°C.
21. Ratchford and Fisher^^ found that methyl acetox3T)ropionate decomposes on
heating to form methyl acrylate and acetic acid in accordance with the following
equation:
CH3COOCH(CH3)COOCH3 -> CH3COOH + CHjiCHCOOCHa
The pyrolysis was found to approximate closely the behavior of a first-order uni-
directional reaction having the following reaction velocity constant {T in degrees
Kelvin):
38,200
h = 7.8(10')e" «^ (1/sec)
Calculate the volume of the flow reactor required to produce 65 per cent decomposi-
tion in a feed of methyl acetoxypropionate flowing at 650 lb per hr at a pressure of
4 atm and a temperature of 1100°F.
22. The viscosity of liquid benzene at 40°C and atmospheric pressure is given by
the International Critical Tables as 0.492 centipoise. Estimate the viscosity of the
liquid at 70°C. Estimate the viscosity of the vapor at 300°C and 3 atm pressure.
23. The viscosity of CO2 gas is given by the International Critical Tables as
0.0167 centipoise at 40°C and 1 atm pressure. From Fig. 175 estimate the viscosity
of CO2 gas at 500°C and 6 atm pressure.
24. Estimate the viscosity of propyl butyrate vapors at a temperature of 250°C
and a pressure of 1.0 atm from the following data:
Molecular weight 130
Critical temperature, °C 327
Liquid density, 15°C, g per cc 0.879
25. From the following data, estimate the viscosity of pyridine vapors at a
temperature of 450°C and a pressure of 30 atm.:
Molecular weight 79
Critical temperature, °C • 344
Critical pressure, atm 60
26. The International Critical Tables give the viscosity of H2 as 0.0084 centipoise
and of NH3 gas as 0.0092 centipoise at 0°C and 1 atm pressure. Estimate the vis-
cosity of a gas mixture containing 20 per cent Hj and 80 per cent NH3 at 12°C and
22 W. P. Ratchford and C. H. Fisher, Ind. Eng. Chem., 37, 382 (1945).
CHAP. XVIII] PROBLEMS 899
1 atm. Compare this result with the International Critical Tables value of 0.0104
centipoise for this mixture.
27. Vapors of sulfuryl chloride at 200°F and 1.2 atm are fed to a l^-inoh reactor
tube at a rate of 418 lb per hr. The reactor tube is heated at a rate of 5,000 Btu/
(hr) (sq ft) based on the internal area. At elevated temperatures the SO2CI2 decom-
poses by a first-order reaction to form SO2 + CI2. It is desired to produce 98 per cent
decomposition of the SO2CI2 fed. The pressure drop in the reactor and the effect of
the reverse reaction may be neglected. From the following data calculate the length
of reactor tube required, and plot a curve relating temperature to tube length:
Dimensions of tube:
Nominal size IJ in.
Inside diameter . 1.334 in.
Outside diameter 1.50 in.
Inside cross-sectional area • 1.398 sq in.
Surface per foot of length
Inside 0.3491 sq ft
Outside 0.3925 sq ft
For the uncatalyzed reaction, values of fc are reported by Smith.''
50,610
kc = (6.427) 10"e «^ (1/sec)
The heats of formation are reported by the U.S. Bureau of Mines Bulhtin 406
(1937).
Components SO2CI2 SO2 CI2
AH (formation) (gas) 298°Kj
cal/g-mole -82,040 -70,920 0
Molecular weight 135 64 71
For the molal heat capacity of Cl2(g), Spencer and Justice''^ give
Cp = 7.5755 + (2.4244)(10-3)r - (0.965)(10-«)r2; Tin °K
No data are available for the heat capacity of sulfuryl chloride at high temperatures.
The following approximate equation is estimated from low-temperature data by the
methods of Chapter X^VI and may be used in the range from 300 to 900°K:
Cp = 13.00 + (24.0) (10-') r - (14.4)(10-«)r2
28. Calculate the volume of the adiabatic reactor operating at a pressure of 1.2 atm
with an outlet temperature of 7 5 0 ^ which would be required to produce a conversion
of 49 per cent in the reaction of Problem 27, neglecting decomposition in the preheater
and assuming:
(a) Uniformity of temperature as a result of mixing in the reactor.
(6) A progressive temperature change with neghgible longitudinal mixing in the
reactor.
Develop an equation expressing temperatures throughout the reactor as a function
of conversion.
29. It is desired to design a tubular heater and reactor for the pyrolysis of propane
to produce a mixture of ethylene and propylene. Substantially pure propane is to
» D . F. Smith, J. Am. Chem. Soc, 47, 1862 (1925).
" H. M. Spencer and J. L. Justice, J. Am. Chem. Sac., 56,2311 (1934).
be charged through a preheater which delivers it at a temperature of 600°F and an

absolute pressure not in excess of 60 lb per sq in. to the inlet of the reactor-heater. In
order to avoid secondary and reverse reactions the absolute pressure at the outlet of
the reactor is maintained at 20 lb per sq in. The gases from the reactor are com-
pressed and fractionated to recover the desired products and substantially pure pro-
pane which is recycled to the heater and reactor. The design is to be based on a total
propane feed to the reactor of 7000 lb per hr and a conversion per pass of 80 per cent,
corresponding to a net fresh charge of 5600 lb per hr.
A trial design is to be prepared on the basis of passing the entire heater and reactor
charge in a single stream through a series of uniformly sized tubes each 30 ft long and
connected together with 180° return bends. Each return bend has a volume equal to
3.W where d is the inside diameter of the tube. The equivalent length of a return
bend is 60 diameters of straight pipe, and the heated length of each tube is 28.0 ft.
Tubes are available with inside diameters varying by |-in. increments from 1 in.
Since the maximum operating temperature permitted by the tube material is 1400°P,
the heater will be designed to raise the charge to this temperature with a uniform heat
input rate of 8000 Btu/(hr)(sq ft) based on the actually heated internal tube area.
In order to obtain favorable heat-transfer conditions the tube diameter should be as
small as possible without exceeding the allowable inlet pressure. In the reactor
section a constant temperature of 1400°F will be maintained by varying the heat-input
rate.
It is required to determine the diameter and numbers of tubes for the specified
service and also the temperature, pressure, conversion, product distribution, and heat-
input-rate distribution curves throughout the heater and reactor, using the data of
Table LIV.
For determination of the proper tube size simplified prehminary analyses should be
made in which secondary reactions are neglected and only the combined effects of the
four primary reactions are considered. The tube size and inlet pressure arrived at in
this manner are used as the basis of the detailed analysis which considers all reactions.
For both the prehminary and the detailed analyses it is convenient to plot working
curves relating the density, viscosity, and specific heat of the reactant mixture to
temperature and conversion. It is satisfactory for this purpose to plot one curve for
pure propane and one for 80 per cent conversion, assuming the product distribution
indicated in Fig. 172. Values for other conversions can be estimated by linear
interpolation between these curves.
30. The total feed to the heater of a single-coil cracking unit similar in flow to
Fig. 92, page 422, is a light gas oil having a gravity of 38° API and a characterization
factor of 11.3.
This stock is charged to thelinlet of the radiant section of the heater at a rate of
65,000 lb per hr, a temperature of 730°F, and a gauge pressure of 630 lb per sq in.
The products leaving the heater enter the top of a vertical adiabatic-reaction chamber
where a small amount of cold quench oil is introduced to prevent coke formation on,
the walls of the chamber.
The ra,diant section of the heater comprises 66 tubes each 3 i in. internal diameter
and 30 ft long connected with 180° return bends each having a volume of 0.06 cu ft.
The volume of the reaction chamber is 600 cu ft, and its ratio of diameter to height is
such that it may be assumed to operate with uniform conditions of temperature and
pressure throughout. •
The distribution of temperatures and pressures at'the outlets of the indicated
/
lo^ll
CHAP. X V I I I ] PROBLEMS 901
tubes are as follows. The tubes are numbered from the inlet of t
of the heater.
Tube No., Temperature, op Pressure, Lb 3
Furnace Inlet . 730 • 630
10 795 615
20 840 595
30 875 570
40 905 640
50 922 500
60 942 448
66 950 400
Reaction chamber 875 400
The cracking of this charging stock is found to produce 2.9 moles of products per
mole of net charge converted. The reaction may be assumed to approximate a first-
order reaction in accordance with Equation (111). At a temperature of 900°F the
initial-reaction velocity constant in concentration units (kc) is 0.0015 (1 /sec), and
the enthalpy of activation is 55,000 cal/(g-mole). The value of a in Equation (111)
is 1.6. Neglecting changes in the compressibility factor of the feed, calculate:
(a) The severity factors in (equivalent cu ft) (hr) /lb-mole of the heater and the
reaction chamber using a base temperature of 900°F with pressures in atmospheres.
(6) The relationship between severity factor and conversion x:
(c) The conversion produced in the heater alone and in the heater plus the reaction
chamber.
31. Estimate the value of AS* in the formation of NOCl(g) from NO(g) and
Cl2(g) at 298.1°K. The absolute molal entropies at 298.rK and 1 atin are as
follows:
NOCKg), s" = 63.00
NO(g), s° = 50.34
Cl2(g), s° = 53.31
CHAPTER XIX
CATALYTIC REACTIONS
By definition, a catalyst is a substance which influences the rate of a
reaction but is not one of the original reactants or final products. The
catalyst must participate in intermediate steps in such a manner as to
facilitate the over-all course of the reaction. In terms of the theory of
absolute reaction rates the function of the catalyst is to reduce the
positive free-energy change accompanying the formation of the acti-
vated complex. This may be brought about by the substitution of a
sequence of steps, each having a low free energy of activation for a single
step involving a high free energy of activation in the uncatalyzed
reaction. ,
Although a catalyst may greatly alter the free-energy changes accom-
panying intermediate steps of a reaction, the over-all free-energy change
is not so influenced but is dependent only upon the initial state of the
reactants and the final state of'the products. Thus, although the cata-
lyst may change the rate of reaction it cannot influence the over-all free-
energy change and, hence, the equilibrium conditions approached,
except as it may influence the nature of the terminal reactants and
products. SimUarly, a catalyst is unable to cause any reaction to pro-
ceed under conditions not in agreement with the principles of thermo-
dynamics. For example, by means of a catalyst it is possible to acceler-
ate the low-temperature combustion of hydrogen with oxygen to form
water. However, it is impossible for a catalyst to restore such a system
to its original state by reversing the reaction at the same conditions.
It follows from thermodynamic principles and the theory of absolute-
reaction rates that if a catalyst increases the rate of reaction in the for-
ward direction a corresponding increase is produced in the rate of the
reverse reaction. Thus, the net rate of change of the system in either
direction is increased, but it is always in the direction of thermodynamic
equilibrium.
It is believed that in general catalysts function through forming defi-
nite intermediate compoimds with the reactants. The high degree of
specificity of chemical combination is in agreement with the selectivity
exhibited by catalysts in accelerating certain reactions in preference to
others. This selective behavior is illustrated by the influence of differ-
ent catalysts upon the decomposition of ethyl alcohol. In the presence
902
CHAP. XIX] SOLID CATALYSTS 903
of an alumina catalyst, ethylene and water vapor are formed, whereas,

with a catalyst of metallic copper, acetaldehyde and hydrogen predomi-
nate. Thus, 364°C
CaHeOH ^ C2H4 + H2O (1)
AUOJ
CsHsOH ^ CH3CHO + H2 (2)

Cu
The selectivity of a catalyst is also dependent upon temperature. For

example, from the measurements of Alvarado^ the decomposition of
ethyl alcohol in the presence of activated alumina follows Equation (1)
at 354°C, whereas at a lower temperature, 269°C, ether is the principal
product according to the following reaction:
269''C
2C2H6OH ^ (C2H5)20 + H2O (3)
A1»0.
In many processes the selectivity of the catalyst is its most important

property. By proper choice of catalyst it is possible to accelerate only
the desired reactions and to minimize the formation of undesired by-
products which would result if all reactions were equally increased in
rate. The principles governing the selection and preparation of cata-
lysts for specific purposes are not yet well understood, and many of the
developments in this field are the result of elaborate exploration pro-
grams involving the trial of coimtless materials. This type of approach
is being supplemented by more scientific methods as the understanding of
catalysis increases and new data on the nature of catalysts are made
available through developments such as true surface-area measurements.
X-ray and electron diffraction patterns, and electron-microscope studies.
The published theories and results of catalysis of all types have been
summarized in considerable detail by Berkman, Morrell, and Egloff.^
The discussion herein presented is directed toward the development of
methods for quantitatively correlating the behavior of a given catalyst as
a function of the conditions imder which it is used and for arriving at
process designs which lead to optimum results for that particular cata-
lyst. The principles of the selection and development of catalytic
materials are treated only in a general manner, incidental to this princi-
pal objective.
Solid Catalysts. Solid catalysts which increase the rates of reactions
whose reactants and products are either liquids or gases are extensively
used in many important processes. It is beUeved that such solids func-
1 Alvarado, / . Am. Chem. Soc., 60, 790 (1928).
' S. Berkman, J. C. Morrell, and G. EglofiF, " Catalysis," Keinhold Publishing
Corporation, New York (1940).
904 CATALYTIC REACTIONS [CHAP. XIX
tion through the occurrence of intermediate reactions on their surfaces.

As a result the extent and character of the surface are of primary im-
portance in determining catalytic effectiveness. In general it is desir-
able that the catalyst shall have a large surface area per imit mass or
volume and that this surface be relatively accessible to the fluid-reactant
mixture through interconnected pores and openings. In addition to
having a large accessible aiea, the surface itself must exhibit the desired
selective activity.
It was proposed by Taylor that reactions which are catalyzed by solids
actually occur on the surfaces of the solids at points of high chemical
activity which are termed active centers? On this basis the activity of a
catalytic surface is proportional to the number of active centers per unit
area. There is evidence that in many cases this concentration of active
centers is relatively low, as indicated by the extremely small quantities
of " poisons " which are sufficient to destroy the activity of a catalyst.
The exact nature of an active center and the conditions which must be
fulfilled in order that a point on the surface may become an active cente*
remains the subject of much speculation. There is evidence that the
interatomic spacing of the solid structure is important as well as its
chemical constitution and lattice structure. Catalytic activity also has
been found to increase with increased magnetic susceptibility in some
cases. Other specific indexes to catalytic activity; such as various
chemical tests and analyses; and physical characteristics; such as true
density, diffraction patterns, magnetic properties, dielectric constant,
and structure as revealed by the electron microscope; have been found
useful. Methods developed by Emmett' and coworkers for the reliable
measurement of surface areas are of great importance to these studies in
distinguishing changes in activity due to alterations in area from changes
attributable to the nature of the surface itself. At present, however, the
only general index to specific catalytic-activity is an, actual performance
test.
Since catalytic-reaction rates are dependent upon the extent of accessi-
ble surface, much attention is given to increasing both the area and
accessibility of catalyst surfaces. One method of achieving this which
currently is undergoing intensive development is the use of finely divided
particles of catalysts suspended as a dust or slurry in the reacting stream.
By use of high concentrations of suspended catalyst it is possible to
obtain large effective areas on the basis of both volume of reactor and
weight oif catalyst. This method is particularly well adapted to processes
requiring frequent regeneration of the catalyst. Another widely used
' National Research Council, " Twelfth Report of Committee on Catalysis,"
John Wiley & Sons (1940).
CHAP. XIX] SOLID CATALYSTS 905
method involves the formation of the catalyst into pellets having a

porous spongy structure. In order to be most effective the pellets should
have relatively large pores with thin walls continuously interconnected
to provide free access of reactants to the interior surfaces.
In so-called unsupported catalysts, the entire particle consists of the
catalytic material which may be made by direct precipitation and drying
of the catalyst to produce in its final form a highly porous solid gel struc-
ture. Silica gel is characteristic of this type. Another method of mak-
ing unsupported catalysts is to grind the material to a fine powder and
with the aid of binders to compress it into pellets. The noncatalytic
material is then removed by a subsequent operation, such as heating,
burning, or treatment with chemicals. Another method is to prepare
pellets or crystals of a substance which is later transformed to a catalyst
by chemical treatment. Metallic catalysts are produced ia this maimer
by the low-temperature reduction of sulfides, oxides, and nitrates.
It is sometimes possible to obtain catalysts of high activity by produc-
ing a structure having an abnormal atomic spacing. This is accom-
plished by chemical treatment of a substance with a normal structure in
order to remove certain atoms and leave a solid lattice with unusual
atomic separations. For example, activated charcoal is prepared by
charring peach pits, crystals of organic acids, and the hke. A zinc-
chromate catalyst is prepared by heating zinc-ammonium chromate,
driving off ammonia and water, and leaving the zinc and chromium
atoms as widely separated as in the original compound. In a similar
manner copper-chromate catalyst is prepared from copper-ammonium
chromate. Nickel catalysts are prepared by dissolving out the alumi-
num from a nickel-aliuninum alloy leaving the nickel widely spaced.
When such a catalyst is used industrially only a small amount of the
almninum may be removed at a time. A fresh catalytic surface is then
subsequently formed by each further treatment with alkali. A remark-
able lattice was produced by Langmuir and Blodgett^ from barium
stearate. Stearic acid was dissolved from a mixture leaving the
barium stearate intact with its initial atomic displacement.
Supported catalysts, as their name implies, consist of a carrier sub-
stance having a suitable form and a large surface upon which the catalyst
is deposited. Substances such as silica gel, activated aliunina, pumice,
charcoal, and kieselguhr are widely used for supports. The catalytic
material may be deposited by impregnation with a solution, followed by
treatment with other reagents to cause precipitation or modification by
drying and reduction by heating or oxidation or other chemical treat-
* Langmuir, " Recent Advances in Surface Chemistry and Chemical Physics,"
Science Press (1939).
ment. The catalyst may be deposited either directly on a porous pellet

of the support or on a powder which is subsequently compressed into a
pellet, generally with a binder which is removed in a subsequent
operation.
The activity of a catalyst surface is permanently destroyed by sinter-
ing or overheating which permit the atoms to lose their favorable separa-
tion and to attain normal spacing.
Mechanisms of Reactions Catalyzed by Solids. The general quanti-
tative relationships involved in the catalysis of fluid reactions by solids
are summarized in a recent paper by the authors* from which much of the
following discussion is taken. As a basis for the development of rate
equations, it is postulated that, when catalyzed by a solid, a liquid- or
gas-phase chemical reaction actually occurs on the surface of the catalyst
and involves the reaction of molecules or atonas which are chemically
adsorbed on the active centers of the surface. From this viewpoint the
catalyst increases the rate of reaction through its ability to adsorb the
reactants in such a form that the activation energy necessary for reaction
is reduced far below its value in the uncatalyzed reaction.
In order that a reactant in the main fluid phase may be converted
catalytically to a product in the' main fluid phase, it is necessary that the
reactant be transferred from its position in the fluid to the catalytic
interface, be activatedly adsorbed on the surface, andjundergo reaction
to form the adsorbed product. Tile product must then be desorbed and
transferred from the interface to a position in the fluid phase. The rate
at which each of these steps occurs influences the distribution of concen-
trations in the system and plays a part in determining the over-all rate.
Because of the differences in the mechanisms involved, it is convenient to
classify these steps as follows when dealing with catalysts in the form of
porous particles:
1. The mass transfer of reactants and products to and from the gross
exterior surface of the catalyst particle and the main body of the fluid.
2. The diffusional and flow transfer of reactants and products in and
out of the pore structure of the catalyst particle when reaction takes place
at interior interfaces. ^
3. The activated adsorption of reactants and the activated desorption
of products at the catalytic interface.
4. The surface reaction of adsorbed reactants to form chemically
adsorbed products.
It is evident that the rates of these four types of operations are
dependent on widely different factors in addition to the concentrations
or concentration gradients involved. Type 1 is determined by the flow
«0. A. Hougen and K. M. Watson, Ind. Eng. Chem., 35,529 (1943).
CHAP. X I X ] MECHANISMS OF REACTIONS 907
characteristics of the system, such as the mass velocity of the fluid

stream, the size of the particles, and the diffusional characteristics of the
fluid. Type 2 is determined by the degree of porosity of the catalyst, the
dimensions of the pores, the extent to which they are interconnected, the
size of the particles, the diffusional characteristics of the system, and the
rate at which the reaction occurs at the interface. Type 3 is determined
by the character and extent of the catalytic surface, and by the specific
activation energies required for the adsorption and desorption of each
of the components of the fluid. Type 4 is determined by the nature and
extent of the catalytic surface and by the activation energies required for
the reaction on the surface.
The relative importance of these four operations in determining the
over-all rate varies widely. Type 1 is important only when rapid reac-
tions are dealt with or where flow conditions are unfavorable. Since
the rate of this operation is little affected by temperature, its relative
importance tends to vary for a particular system; it is frequently
negligible at low and highly important at high temperatures. Type 2 is
frequently negligible for catalysts of low activity in small particles with
large intercoimected pores and ceases to be a factor for nonporous
catalysts having no internal surface. However, in the general case of an
active catalyst in moderately large particles having large internal sur-
faces with re^ricted capiUarity it may be of major importance.
Operations of types 3 and 4 are chemical phenomena generally involv-
ing relatively large enthalpies of activation and are therefore highly
sensitive to temperature. The actual chemical transformations fre-
quently proceed by several successive stages, each with its own charac-
teristic rate. This is particularly true where several molecules are
involved. Since chemical rates vary over wide ranges, it is improbable
that the rates of any two steps of types 3 and 4 will be of equal order in
any given system. For this reason in many cases it is permissible to
consider only the slowest single step of types 3 and 4 and to assume that
equilibrium is maintained in all other steps of these tjrpes. Under such
conditions the slowest activated step may be termed the " rate-control-
ling step," and the over-all rate is determined by consideration of it in
combination with the physical steps of types 1 and 2.
In order to calculate the rate at which a cataljîc reaction proceeds it is
necessary to develop quantitative expressions for the rates of each of the
individual steps which contribute to the mechanism. This requires
consideration of the fundamental principles of activated adsorption,
surface reactions, mass and heat transfer, and diffusion in porous solids.
ACTIVATED ADSORPTION
In Chapter VII, page 149, it is pointed out that there are two types of
adsorption, one referred to as van der Waals or physical adsorption and
the other as activated adsorption or chemisorption. In the van der Waals
type, adsorption results from physical-attractive forces similar to those
causing the condensation lof a vapor. Chemisorption, as the name
implies, is believed to involve definite electron bonds corresponding to
the formation of a chemical compound between the adsorbate and the
surface. Like a chemical reaction, chemisorption is a highly specific
phenomenon depending upon the chemical natures of the adsorbate and
adsorbent.
The concept of chemisorption was introduced by Taylor in 1930.
Since that time much attention has been devoted to it in the literature as
summarized by Taylor,' and Glasstone, Laidler, and Eyring.* I t is
currently believed that activated adsorption occurs only on specific
active centers which, as previously pointed out, may represent only a
small fraction of the total surface. On this basis the maximmn capacity
of a surface for a specific chemisorption is frequently much less than the
amount of adsorbate required to form a monomolecular layer. By
van der Waals adsorption it is possible to adsorb much larger quantities
which form layers several molecules in thickness or result in capillary
condensation.
Because of the larger energy changes involved in the formation of
valence bonds the enthalpy changes accompanying chemisorption are
generally highly negative, in the range of —10 to —100 kcal per g-mole.
The enthalpy change of van der Waals adsorption is of the same order as
the heat of condensation, in the range of — 5 to —10 kcal per g-mole.
I t follows that the effect of temperature in diminishing the quantity of
adsorbate at equilibrium is greater for activated than for van der Waals
adsorption.
The most significant difference between van der Waals and activated
adsorption is in the rate at which equilibrium is approached. Acti-
vated adsorption, as the name implies, requires a definite energy of acti--
vation which corresponds to relatively slow rates of adsorption. These
rates are greatly affected by changes in temperature, just as is the case
of any chemical reaction.
Although van der Waals adsorption may be considered as requiring an
activation energy' its value is so low that in general it may be assumed
that equilibrium is reached instantly at the interface and that observed
8 " Theory of Rate Processes," Glasstone, Laidler, and Eyring, McGraw-Hill Book
Company, New York (1941).
CHAP. XIX] ACTIVATED ADSORPTION 909
rates of physical adsorption are largely determined by the accompanying

diffusional phenomena.
Both activated and van der Waals adsorption may be encountered in a
single system, although they are frequently important in different ranges
of temperature. In general, van der Waals adsorption becomes very
small at temperatures not greatly above the critical temperature of the
adsorbate. Much higher temperatures may be required in order that
activated adsorption may proceed at a significant rate. This effect is
shown diagrammatically in Fig. 181 in which quantity adsorbed at
equilibrium under a constant pressure is plotted against temperature.
For the system represented, the quantity which is chemisorbed at equilib-
rimn approaches a maximum, corresponding to reaction with all avail-
able centers, at a relatively high temperature where the amoimt of
van der Waals adsorption is negligible.
At lower temperatures van der Waals
adsorption becomes increasingly im- Equilibrium
Van der Waals
portant. 'adsorption .
Because of the low rates of activated Equilibrmm
activated
adsorption at low temperatures, appar- âdsorption
ent total-adsorption curves of the form Apparent
indicated by the broken line of Fig. dsorption
181 are frequently observed. Such
Temperature -
regions of minimum adsorption are
believed to result from failure to attain F I G . 181. Van der Waals and Acti-
equilibrium chemisorption. The posi- vated Adsorption.
tion of the broken-line curve of Fig. 181 is therefore dependent upon the
time allowed for adsorption.
It is evident that measurements of either adsorption-equilibrium
quantities or rates of total adsorption may represent a complicated
combination of activated and van der Waals effects, unless there happens
to be a wide temperature range in which van der Waals adsorption is
negligible in quantity and chemisorption is negligible in rate. In the
general case where such a separation does not exist, the interpretation of
activated-adsorption data is difficult. This problem is further compli-
cated by the observation that more, than one type of activated adsorp-
tion of a single component may occur simultaneously on a given surface,
presumably because of the existence of different types of active centers.
The different types of adsorption may be effective in different tempera-
ture ranges. In such cases it is possible to obtain apparent adsorption-
equilibrium curves which show two or more minima similar to t h a t
of Fig. 181.
In summary, van der Waals adsorption is characterized by high rates
at low temperatures, large quantities of adsorbate, low heats of adsorp-

tion, and a relatively low degree of specificity. Activated adsorption on
the other hand is characterized by low rates and high energies of activa-
tion, low quantities of adsorbate, high heats of adsorption, and a high
degree of specificity.
Chemisorption Rates and' Equilibria. In the development of rate
equations for simple activated adsorption, it may be assumed that a
unit area of surface offers L' active centers on which adsorption,may
occur and that all of these centers behave similarly. The rate of adsorp-
tion per unit area of a component A from a fluid in contact with the sur-
face is then proportional to its activity UAI in the fluid at the interface
and to the concentration c^ of vacant active centers per unit area of sur-
face. Since the actual surface area of porous materials is generally
unknown, it is convenient to express reaction rates in moles formed per
unit time per imit mass of solid having a surface area am. It is also con-
venient to express the number of active centers per unit mass in molal
rather than molecular units. Thus L, the nmnber of molal active centers
per imit mass is equal to L'um/N where N is the Avogadro number.
Similarly the concentration of vacant active centers may be expressed
in molal imits per imit mass and designated as Cj. On this basis,
rA = kAaAiCi (4)
where TA = rate of adsorption of A, moles per unit time per imit mass
kA = adsorption velocity constant of A
CiAi = activity of A in the fluid at the interface, in concentration
units
Ci = molal concentration of vacant active centers per unit mass
Since chemisorption is a reversible phenomenon, component A is also
desorbed from the surface at a rate proportional to the concentration of
adsorbed molecules on the surface. Thus,
T'A = KCA (5)
where CA = molal concentration of adsorbed A, moles per unit mass.
The net rate of adsorption is then the difference between the rate of
adsorption and desorption, or
r = kAaAiPi ~ TCACA (6)
When adsorption equiUbrium is reached the net rate of adsorption
becomes zero, arid
- ^ 4 f = K. (7)
where KA = adsorption-equilibrium constant of A.
CHAP. XIX] CHEMISORPTION RATES AND ÊQUILIBRIA 911
If (6) and (7) are combined, an expression is obtained for the net rate
of adsorption where all sites are equally accessible.
r = fc^ (aAiCi - ^ ) = kACiiuAi - al,-) . (7a)
where a*j = activity of component A in equilibrium with the activatedly

adsorbed molecules
The fraction of available sites ci/L can be expressed in terms of the
nimiber of moles adsorbed per unit mass, thus;
£^ =
- 1
1 _- !J1
l l ^ 9, (7b)
T ~ *
where UA = moles of A adsorbed per imit mass of catalyst

n* = moles of A adsorbed per unit mass of catalyst when all sites
are covered
ei = fraction of available sites not covered
Where all the active sites are not equally accessible as when the
adsorbent is in pellet form, an effectiveness factor EA should be included,
as discussed later.
/ If component A is in admixture with other components B, B, S, and I
which are also adsorbed on active centers of the same type, rate and
equilibrium equations similar to (6), and (7) may be written for each
component. Then,
ci = L - (CA + CB + ct • • •) ' (8)
At equilibrium conditions each of the adsorbate concentration terms
in Equation (8) may be replaced by an expression similar to that obtained
by solving Equation (7) for c^:
ci -^ L - ciiuAiKA + asiKs + anKi • • •) (9)
(1 + UAiKA + ttBiKs + aiiKi • • •)
An expression for the equilibriuni surface concentration of A in terms

of interfacial fluid activities is obtained by combining Equations (7)
and (10):
aAjK-AL ^
(1 + ttAiKA + asiKB + aiiKi • • •)
Similar equations may be written for the equilibrium surface concentraf-

tions of the other components of the mixture. .
.*
The theory of absolute-reaction rates is applicable to chemisorption in
the same manner as to homogeneous reactions. It is assumed that
chemisorption takes place through the formation of an activated complex
with the active center. Thus, if I represents an active center the
adsorption of A is represented by
A + l:Âl^ = Al
If it is assumed that the activities of the adsorbed constituents are equal
to their molal concentrations per unit mass of adsorbent, the reaction
velocity constant of this reaction per unit mass of adsorbent is,
= -7- ^
KA (12)
n
where Ai?i and A<Si = the enthalpy and entropy of activation of the
adsorption reaction.
Similarly, for desorption,
_AH^' AS^
kl=ê ^ ^ " ^ (13)
Combining (7), (12), and (13) gives
KA = e ^'' '^ = e ^^ * (14)
where AH^ = AH\ — AHA' = the standard enthalpy change of adsorp-

tion
A/S^ = ASA — AiSi' = the standard entropy change of adsorption
Chemisorption and Dissociation. There is evidence that certain
species of molecules are dissociated during chemisorption. For example,
hydrogen and nitrogen molecules are dissociated to atoms when ad-
sorbed on the surfaces of some materials.
Although other mechanisms are possible, Glasstone, Laidler, and
Eyring^ suggest that, where a molecule is dissociated during adsorption,
it is commonly first adsorbed on a dual adsorption site consisting of two
adjacent active centers. This adsorbed molecule then decomposes, and
the atoms jump to two vacant active centers. Thus, if A 2 is a molecule
which dissociates into two A atoms and ^2 is a dual adsorption site,
A2 + kÂih
A2l2 + 2l:p±2Al + k
It is believed that the first step is rate-controlling and that equilibrium is
CHAP. XIX] CHEMISORPTION AND DISSOCIATION 913
maintained in the second. Then if r is expressed in moles of A 2,

r = kAfiAîCi^ — ^'ASPAA (15)
Since the second ste'p maintains equilibriumj
<Â2k = -^r^ (16)
where X^ = equilibriimi constant of the dissociation step.

In order to develop an expression for ci^, the molal concentration of
dual adsorption sites, it is assumed that the active centers are arranged
on the surface in a regular geometric pattern such that each site is sur-
rounded by s equidistant'adjacent sites. Thus, if the active centers are
situated at the corners of squares, s = 4, whereas, if they are at the
comers of equilateral triangles, s = 6.
If 01 is the fraction of active centers unoccupied, an average center has
adjacent to it Bis vacant centers. The total concentration of pairs of
adjacent v z ^ n t active centers which are dual adsorption sites is then
one-half the product of the concentration of vacant centers and the
average number of adjacent vacant sites. The factor of one half results
from the fact that, in the summation represented by the product of the
concentration and fraction of vacancy terms, each pair of vacant active
centers is coimted twice.
where Cj^ = concentration of dual sites per unit area

61 = fraction of active centers unoccupied
s = number of equidistant active centers adjacent to each center
Equation (17) also may be written in terms of molal concentrations per
unit mass.
(18)
Combining (15), (16), and (18) yields
T = — (fcâflAjiCj — k^^—r) (19)
At equilibrium,
CAI kAji-A jT
(20)
aAîCi liAi
CA = CA.1 = CiâAîKî (21)

If A is adsorbed with dissociation from a mixture also containing com-

ponents B and I, an expression similar to Equation (11) may be devel-
oped for adsorption-equilibrium conditions:
c^ = âA,iK,,L ^22)
1 + Va^jiiÂj + ttBiKs + aiiKx + • • •
It is evident by comparing Equations (11) and (22) that, if a molecule
dissociates as it is adsorbed, the square root of the product of its activity
times its adsorption-equilibrium constant appears in all adsorption-
equilibrium expressions in place of the first power of this product.
The net rate of adsorption on dual active sites in the absence of disso-
ciation and where all sites are equally accessible may be written from
Equation (19) as
J- = ^ kAAio-Aî - a*J) (22a)
* . "
where a^î is the activity of component A 2 at the interface in equilibrium
with the adsorbed molecules.
The fraction of available single sites Bi can be expressed in terms of the
number of moles adsorbed per unit mass, thus:
;-' = 1 - ^ ^ = e , (22b)
where w^j = mole of A 2 adsorbed per unit mass of catalyst

w^j = moles of 4 2 adsorbed per unit mass of catalyst when all sites
are covered
From Equations (22a) and (22b;,
r = - kA^Leâ^î - a!,.-) (22c)
Limitations of Simple Adsorption Theory. In the derivation of

Equations (4) to (22) it was assumed that all active centers have the same
activity and imdergo the same reaction in chemisorptlon. It is also
assumed that the enthalpy and entropy changes accompanying chemi-
sorptlon are constant, independent of the amount adsorbed. Experi-
mental data on adsorption rates and equilibria indicate considerable de-
viation from both of these assumptions. The heat of adsorption is
frequently found to become less negative as the amount adsorbed in-
creases. There is also evidence of varying activity of the centers.
The observed variations in heat of adsorption are attributed both to
variation in the character of the active centers and to interaction between
CHAP. XIX] SURFACE REACTIONS 915
the adsorbed molecules. Numerous modifications®'^ of the fimdamental

adsorption equation? which have been proposed take these effects into
accoimt and give improved agreement with experimental data. How-
ever, no simple form has been developed which has proved applicable to
all systems. This difficulty probably results from the fact previously
mentioned that observed adsorption data generally represent a complex
combination of several simultaneous phenomena. Thus a single adsorb-
ate may undergo simultaneously two or more types of activated adsorp-
tion together with some van der Waals adsorption. For this reason
adsorption data for a particular system frequently are in better agree-
ment with various empirical equations of logarithmic or exponential form
than with the simple equations given.
The failure of Equations (5) to (22) to represent accurately data on the
adsorption of ûre components does not of necessity invalidate their
application to the adsorption step of chemical reactions. It is probable
that such an adsorption step in a particular reaction involves only active
centers of a restricted degree of activity. The fundamental equations
may accurately represent this single type of chemisorption while failing
to apply to the composite effects of it and other types. The rate and
equilibrium constants of such a single effective or apparent adsorption
step must be evaluated from reaction-rate data by methods described in
following sections. Such constants are effective average values for the
range of conditions on which their evaluation is based and may show
some lack of constancy over wide ranges. However, iiseful rate equa-
tions can be developed in this manner, with the assumption of simple
adsorption relationships.
Furthermore, the distinction should be made between reaction rates
proceeding at a steady state as do most reactions in flow systems and
reaction rates of adsorption where no other reactions are involved. In
the latter case, a nonsteady state is realized as the gas is gradually
adsorbed with a rate diminishing to zero, whereas in the former case the
reaction rates are constant requiring also that the amount adsorbed is
not undergoing change.
SURFACE REACTIONS
When a reaction is catalyzed by a solid it is presiuned that the actual
combination of the reactants occurs on the surface of the soUd. Such
surface reactions may take place either between an adsorbed reactant
molecule and a molecule in the fluid phase or between adsorbed mole-
cules on adjacently situated active centers. With the latter mechanism
' S. Brunauer, K. S. Love, and R. G. Keenan, J. Am. Chem. Soc., 64, 751 (1942).
a reaction proceeds at rates proportional to the concentrations of

adjacently adsorbed reactants. Thus, if-adsorbed molecule A reacts
with adsorbed molecule B, the rate of the reaction is proportional to the
number of pairs of adjacently adsorbed A and B molecules per unit area
of surface. Similarly, if in a monomolecular reaction, adsorbed mole-
cule A reacts with a vacant active center to form a complex which subse-
quently breaks down to form two adsorbed product molecules, the rate
of the reaction is proportional to the number per unit area of A mole-
cules adsorbed adjacent to vacant active centers.
If the arrangement of active centers is such that each is surroimded by
s equidistant centers, an average adsorbed molecule of A has adjacent to
it sei vacant centers, where 0i is the fraction of the total centers which is
vacant. Similarly, SBB molecules of B are adjacent to each adsorbed A
molecule, where BB is the fraction of the total centers which is occupied
by adsorbed B molecules. Then by the type of summation obtained in
deriving Equation (22), the surface concentration of A molecules and
vacant active centers adjacent to each other becomes sCjêi, and the con-
centration of A and B molecules adsorbed in adjacent positions becomes
sc^^js. Since 8 equals q/L' and BB equals C'B/L',
CM = ^,CÔI (23)
C'AB = J, CAC'B (24)
where c^j = surface concentration of pairs of adsorbed A molecules and

vacant centers in adjacent positions
<^'AB — surface concentration of pairs of adsorbed A and B mole-
cules in adjacent positions
Equations (23) and (24) may be similarly written with concentrations

CA and CAB in moles per unit mass.
The rate of a monomolecular surface reaction involving interaction of
an adsorbed A molecule and a vacant active center then becomes
s
r = liCAi = kyCACt (25)
Similarly, for a surface reaction between adsorbed A and B molecules:

fcs
r = kcAB = J- CACB (26)
If products R and S are formed, the net rate of the forward reaction on
CHAP. XIX] GENERAL SURFACE RATE EQUATIONS 917
the surface is expressed by
^= -(^0.^-—j (27)
where K' = the equilibrium constant of the surface reaction =

CRCS/CACB at equilibrium
If the surface reaction results from direct combination of an adsorbed
molecule A with a molecule of B from the fluid phase, the rate is pro-
portional to the concentration of adsorbed A molecules and to the activ-
ity of 5 in the fluid phase at the interface. Thus, in this case,
r = kasiCA (28)
General Surface Rate Equations. On the basis of the principles
discussed in the preceding sections, a general equation inay be developed
for the following surface reaction:
A+B^R
In the general case the system will include inert components in the fluid
phase represented by 7. . . .
The net rate of the forward reaction is the difference between the
rates of the forward and reverse reactions which are represented by
Equations (26) and (25), respectively. Thus,
r = — {kcACB - k'cRCi) (29)

LI
The surface concentrations may be expressed in terms of the activities '
in the fluid at the interface from Equations (4) and (5). Thus, the net
rate of adsorption of component A is equal to the over-all rate of reac-
tion. This net adsorption rate is the difference between the rates of
adsorption and desorption, or
r = kAttAiCi — k'^CA = k'^{KAaAiCi — CA)
w n' (3")
CA = JÂdAiCi — r/kj^
Similar expressions may be written for the other reactant and product
adsorptions, and substituted in Equation (29):
r = — fc (KAaAiCi — -jTji^BaBiCi — -rr)
- k' \KnaBiCi + ^ j c J (31)
The concentration of vacant centers also may be expressed in terms of

fluid-phase activities at tlie interface:

Ci = L — (fiA + CB + CB + Ci • • •)
= L — CiiKAttAi + KsdEi + KsaBi + Kjan • • •)
\toA kg k'J
(32)
(1 + KAttAi + Ksasi + KBam + Kiau • • •)
Combining Equations (31) and (32) yields a complete expression for the
rate of reaction in terms only of activities in the fluid at the interface and
the constants of the system. If the constants were all evaluated, the
equation could be solved graphically for the rate at specified conditions.
However, in general tljis complete equation involving the rates of all
adsorption steps as well as the rate of the surface reaction is so cumber-
some as to be of little value.
As previously mentioned, in many cases it is satisfactory to assume
that the rate is controlled by a single slow step and that all other steps
are so fast that equilibrium may be assumed. The slow step may be
either the surface reaction or the adsorption of any one reactant or the
desorption of any product. Simplified useful equations may be devel-
oped on this basis.
Monomolecular Reactions. Two mechanisms may be postulated for
reversible surface reactions which are monomolecular in both directions,
such as isomerization reactions. The reaction might proceed through an
adsorbed molecule, acquiring sufficient energy to cause it to react and
form a product molecule (all the time confined to a single active center).
In the other mechanism an adsorbed molecule becomes sufficiently
energized to form a complex with an adjacent active center, which then
decomposes to form an adsorbed product molecule.
If the first mechanism is followed-, the rate of reaction is proportional
to the concentration of adsorbed reactant molecules:
r = kcA (33)
If the second mechanism is correct, the rate of reaction is proportional to
the concentration of pairs of adsorbed molecules and adjacent vacant
active centers. This rate is expressed by Equation (25).
If it is assumed that only a single activated step is rate-controlling,
simpUfied over-all equations may be derived according to either mecha-
nism for the reaction:
A:t±R
CHAP. XIX] MONOMOLECULAR REACTIONS 919
Surface Reaction Controlling. If the surface reaction is rate-control-

ling, it is assumed that all adsorption steps are in equiHbrium, and the
surface concentration terms in Equation (33) may be replaced by
Equations ( U ) :
MECHANISM 1
_ ^^ { TT - ^ 5 £ i ^ \ (QA\
' l + a^,K^ + auû + a,J^,--X^'^^ K' ) ^^^^
where K' = surface equilibrium constant = k/k'
At equihbrium the net rate becomes zero, and
^* = | i K ' = K (35)
Substituting Equation (35) in (34) gives
^ kLKA / _ ORA
'• •l + aAiKA + aRiKR + aiiKi-'-V''' K) ^^^^
MECHANISM 2, Equation (25) combined with Equations (10) and

(11) gives
ICSLKA ( ORA
(37)
(1 + ttAiKA + aBiKu + aiiK
The essential difference between Equations (36) and (37) is the

squared term in the denominator. According to these equations the
initial rate, when am = 0, of a reaction following mechanism 1 is pro-
gressively increased by increased activity of the reactant in the fluid
phase. If mechanism 2 is followed, the initial rate may increase with
increased activity of A at low activities, reach a maximum corresponding
to an optimum activity, and then diminish as the activity of A is further
increased. Effects of this type have been experimentally observed in
varying the pressure on monomolecular catalytic reactions.
Adsorption of Reactant Controlling. If the adsorption of the reactant
is the rate-controlling step, it is assumed that the surface reaction and the
other adsorption steps are in equilibrium. The rate is expressed by the
following:
r = kAiaAiCi - CA/KA) (38)
In the application of Equation (38) all surface concentrations may be

expressed in terms of the equilibrium of Equation (7), with the exception
of CA. This surface concentration must be arrived at from the condition
of equilibrium in the surface reaction:

CRKA aRiCiKA ,__,
Ci = L — icA+ CB -\- ci- • •) = L — ci{aRiKA/K

+ aBiKn + aiiKi • • •)
1 + amKA/K + asiKit + anKi • • •

Combining Equations (39) and (40) with (38) gives
^ ÂL / _ OgA
'" 1 + aEiKA/K + as,KB + aiiKi • •-y^' K) ^ '
Equation (41) is independent of the mechanism of the surface reaction.

If the form of Equation (41) is compared with (36) and (37), it may be
noted that, when adsorption of the reactant is controlling, the initial
rate (where aRi — 0) is directly proportional to the activity of the
reactant; if the surface reaction is controlling, the initial rate increases
less than in proportion to increased activity of the reactant.
Another characteristic of reactions controlled by adsorption is evident
by consideration of the net rate of the reverse reaction of Equation (41).
If the over-all equilibrium constant K is small, the net rate of the reverse
reaction becomes independent of the activity of A. This behavior is
characteristic of reactions in which the rate is controlled by the rate of
desorption of the product.
Bimolecular-Monomolecular Reactions. The class of reaction which
is bimolecular in one direction and monomblecular in the other may be
represented by the equation:
If it is assumed that the reaction takes place between adjacently ad-

sorbed molecules, it is necessary that the monomolecular reverse reac-
tion follow the mechanism expressed by Equation (25). By the same
procedures followed in the preceding section, the following rate equa-
tions are developed:
Surface Reaction Controlling
ksLKAKs
r = (1 + ttAiKA + ttBiKs + aBiKR + aiiKr • --y
12
X (aAiaBi - Y) (42)
CHAP. XIX] BIMOLECULAR IN BOTH DIRECTIONS 921
Adsorption of A Controlling
UAL
y = ^
1 + aBiKa + ^ , / + aBiKu + anKi H
Adsorption of R Controlling
kuLK
r = ; ••
1 + ttAiÂ + as^B + aAidBiK-RK- + auKi + . . .

xLiaBi-^) (44)
In Equation (42) the initial rate of the reaction is increased by in-

creased activity oi component B; in Equation (43) increasing the
activity of B reduces the initial rate where the activity of R is zero.
Thus, it is characteristic of a bimolecular reaction in which the rate is
controlled by the adsorption of a reactant that the initial rate is reduced
by increased activity of the other reactant.
In Equation (44) the over-all equilibrium constant of the reaction
appears in both the numerator and denominator of the multiplying
fraction. Under conditions favorable for the forward reaction where
K is large, the rate becomes independent of the activity of the product
and may approach ICKL/KR, independent of the activities of both react-
ants and product.
If the reaction on the surface occurs between an adsorbed A molecule
and an unadsorbed B molecule in the fluid phase, the rate is propor-
tional to the product of the surface concentration of A and the interfacial
fluid activity of B. Equations similar to (42), (43), and (44) result,
except that the adsorption-equilibrium constant of B does not appear.
If the surface reaction is controlling, the term s does not appear, and the
denominator group enters as the first power instead of the square.
Reactions Bimolecular in Both Directions. Reactions of this type are
represented by the following equation:
A+B^R+S
By extension of the methods demonstrated in the foregoing, the following

rate equations may be derived, if the surface reaction is assumed to take
place between adjacently adsorbed molecules:
Surface Reaction Controlling

ksLK^Ke
r = (1 + GAiKA + asiKs + aniKs + asiKs + auKr H ^
X [aAiaei —j (45)
Adsorption of A Controlling
UAL
r =
1 -\ '—^ h aBiKs + aRiKii-\- asiKs + anKi +
XiaAi — ] (46)
Adsorption of R Controlling
kRLK
r =
1 + ttAiKA + asiKs + — h asiKs + anKi +
asi
/aAjaBi __ ORA
X \P asi
^ - ^ K)
(47)
Comparisons between Equations (45), (46), and (47) are similar to
those between (42), (43), and (44). A reaction whose rate is controlled
by the adsorption of a reactant is characterized by an adverse effect on
the initial rate resulting from an increase in activity of the other reactant.
A reactant whose rate is controlled by adsorption of a product is charac-
terized by the initial rate being independent of reactant activities under
conditions of large over-all equilibrium constants and negligible reverse
reaction.
Reactions Involving More than Two Molecules. The adsorption
theory of catalysis does not preclude reactions involving the simultane-
ous interaction of several molecules, and no highly improbable mecha-
nism such as the simultaneous collision of all molecules is involved. The
rates of such reactions should be proportional to the concentration of
groups of the required number of molecules adsorbed on adjacent active
centers. Rate equations might be built up by extension of the pro-
cedures used in developing Equations (46) and (47) and the following
relations.
I t would be expected that the rates of such reactions would be low
because of the low concentrations of properly adsorbed groups except in
the case of one reactant molecule or atom A reacting with several other
molecules or atoms B, all of the same species. In such a case high rates
CHAP. XIX] EFFECT OF DISSOCIATION 923
are obtained with a catalyst which strongly absorbs reactant B so that

the majority of the surface is covered with adsorbed B molecules or
atoms, and most of the A molecules or atoms which are adsorbed are
surrounded by the requisite number of B units on adjacent centers.
Effect of Dissociation on Rate Equations. In the development of
Equations (29) to (47) it was assumed in every case that all reactants
and products are adsorbed without dissociation. From Equation (22)
it is evident that, where a molecule dissociates as it is adsorbed, the
square root of the product of its activity times its adsorption-equiUbrium
constant appears in all adsorption-equilibrium expressions in place of the
first power of this product. Thus, Equations (36), (37), (42), and (45),
in which all adsorption steps are at equilibrium, may be modified to
apply where a component is dissociated and one-half molecule partici-
pates in the reaction merely by raising to the one-half power the activity
and adsorption-equilibrium constant of the component which is disso-
ciated wherever either appears in the equation.
The effects of dissociation where adsorption rate is controlling may be
demonstrated by consideration of the effect on Equation (46) of the
dissociation of B first and then A. If A is not dissociated the rate is
expressed by Equations (4) and (5):
= KA \aAfii — — I (48)
If B is dissociated and only ^B molecule enters into the reaction, an

expression for CA is derived from the equilibrium of the surface reaction;
CRCS ciaRiKnasiKs ciamasiKA .^„,
CA = = — ^ ^ = =^ (49)
cisK' V^^^K' VasiK
ci = — (50)
1 H -=z h VasiKs + ajtiKR + asiKs + auKi -{
VttBiK
Substituting Equations (49) and (50) in (48) gives
kAL
r = O-RidSiKA
1 H j=^ h VaBiKs + ORiKR + asiKs + auKi +
VasiK
{ aRidsi \
X I aaAi
, , - J ^7=—
^^ I (51)
When Equations (51) and (46) are compared, if the component whose
adsorption rate is controlling is not dissociated, the rate equations are
modified for application where a component is dissociated, and only

one-half molecule enters the reaction, by raising the activity and adsorp-
tion-equilibrium constant of the dissociated component to one-half
power wherever they appear in the equation.
If component A of Equation (46) is dissociated and only one-half
molecule enters the reaction, it is necessary to express the rate of reaction
by Equation (19). From t\ie equihbrium of the surface reaction,
CBK' UBiK
«' = aRiasiVEl ^^^^

1 -\ '„ ' + asiKa + aBiKa + asiKs + anKi ^
aBvK
Substituting Equations (52) and (53) in (19), gives
r =
iM + ^' ^' -f- QBiKa + aRiKn + asiKs + anKx -j j
This equation is quite different in form from Equation (46), and it is also
dependent on the mechanism assumed for the adsorption and dissociation
of A. By assuming a different mechanism (6), an equation can be
obtained in which the initial rate is proportional to the one-half power
of the activity of A instead of the first power of Equation (54).
If reactant A is dissociated into two chemisorbed half molecules, both
of which enter into the reaction, a still different series of equations
results. If reactant B is adsorbed without dissociation, the surface
reaction in this case involves three active centers. Thus,
2 ( U ) + 5 4- 3Z = 72 -1- S -t- 3J (55)
This mechanism, which appears somewhat improbable, requires that the

reverse reaction proceed by forming an activated complex involving an R
and an S molecule and three active centers. The rate of the surface
reaction is proportional to the concentration of adsorbed B molecules
with two adsorbed Â molecules adjacent to them. An average B mole-
cule is surrounded by s active centers of which a fraction e\A. is occupied
by adsorbed | A molecules. The average concentration of pairs of
adjacently adsorbed B and \A molecules is thus CBSBU- The B molecule
CHAP. XIX] EFFECT OF' DISSOICATION 925
in each of these pairs is in turn surrounded by (s — 1) centers in addition

to that occupied by the | A molecule. A fraction eÂ of these centers is
also occupied by Â molecules, and the concentration of B molecules
with two §A molecules adjacently adsorbed becomes
CBHÂ) = CBseÂis ~ l)0Â = — Yi v^^)
where e^^ — fraction of total centers L which is occupied by § 4 mole-

cules = CÂ/L.
The rate of the surfaqe reaction is expressed by a combination of Equa-
tion (56) with a similar expression for the reverse reaction.
s(s - l)kA r , . , cncscil

r = Y^ ]^B{C\AY - - ^ J (57)
If equilibrium is maintained in the adsorption steps, Equations (10),

(11), and (22) may be combined with (57):
s(s - \)L'kAKAKB
r = .
(1 H- VaAiKA + aBiKa + amKii + asiKs + aAiKiY
[aAiUBi — \ (58)
Equations may similarly be derived for cases in which this mechanism is

followed by an adsorption rate as the controlling step. If the adsorp-
tion of A is controUing, the rate is expressed by Equations (19) and (20):
Since the surface reaction and the adsorption of B, R, and S are assumed
to maintain equilibrium,
, .„ CRCSCI aAiasiCJKRKs amds^KA ,„„,

{cur = — ^ = F~F7~ = IP— (60)
ci = L — cA •\j-^^~^jr^ + o-BiKB + aBiKB + asiKs + anKi
or
ci = — . , (61)
1 + J ^' ^' ^ + OBiKs + OBiKR + asiKs + auKi
Substituting (60) and (61) in (59), gives

SKAL-
r =
! 1 + \ / " ~ F ^ + (^BiKa + aBiKs + asiKs + anKi
lo.Ai —
If the adsorption of B, R, or S is contrpUing, the rate equations are simi-

lar in form to Equations (46) and (47), except that in the denominator
group the product of the activity times the adsorption-equihbrium con-
stant of the dissociated molecules is raised to the one-half power.
POISONS
The term " poison " has been loosely used to denote any material
which retards the rate of a catalytic reaction. Thus, the products of the
reaction are sometimes referred to as poisons because increasing their
concentration reduces the rate. However, it is beheved more logical to
eliminate all reactants and products from classification as poisons and
reserve the term for other components which retard the reaction by
reducing the number of active centers available for the reactants. Such
poisons may be classified as either temporary or permanent.
From inspection of the rate equations it is evident that any inert com-
ponent / which is adsorbed on the active centers promoting the reaction
serves as a poison; the extent of the retardation depends on the concen-
tration of the inert and its adsorption-equilibrium constant. If the
adsorption of the, inert is reversible with a moderate equilibriimi con-
stant, the poisoning effect quickly reaches a constant value which is
temporary and is eliminated by removal of the component from the
reactant fluid. Thus, the nitrogen in air might serve as a temporary
poison in a catalj^tic oxidation. The effects of such inerts are included
in the rate equations.
Another type of poisoning frequently encountered results from the
permanent substantially irreversible adsorption of components present
in small quantities in the reactant fluid. The effect of this type of
poisoning is progressively to reduce L', the number of active centers per
imit area, as the catalyst continues in service. The frequently ob-
served loss of activity of a catalyst in service may result from poisoning
of this type, from structural changes affecting the area and L, or from
physically coating the active surface with a solid or semisolid material
either present in the reactants or formed by secondary reactions. In
high-temperature organic reactions, catalysts are frequently coated
with high-molecular-weight compounds approaching pure carbon in
composition.
CHAP. XIX] CONVERSION EQUATIONS 927
Limitations of Surface Rate Equations. The rate equations derived

in the foregoing are ideal forms involving several assumptions which are
not rigorous. The derivation implies that all active centers on the cata-
lytic surface behave similarly, but there is evidence that this is not the
case. Frequently the energy of activation gradually increases as active
centers are progressively occupied. It is well known that surface activ-
ity depends upon spatial arrangement of active centers which may not
be uniform over the entire surface. Also it is assumed that each com-
ponent is adsorbed independent of interaction between the adsorbed
molecules of like or unlike species. Again there is evidence that such is
not the case, and corrections are developed^ for such interaction effects
in simple systems.
Although nmnerous attempts have been made to establish more rigor-
ous adsorption relations, there is httle evidence that their use is justified
in rate equations for complex reaction systems. It is believed more
generally satisfactory to use the equations previously developed and
evaluate the constants from direct-reaction rate measurements at con-
ditions representing the ranges of interest of the variables. Constants
directly evaluated from experimental data in this manner represent
average apparent or effective values over the range considered, and the
resultant equations cannot be expected to be entirely rigorous.
As in all theoretical rate calculations, the form of the relation is
dependent on the mechanism. This situation is entirely different from
that encountered in thermodynamic equilibrium calculations in which
the mechanism is immaterial. Furthermore, the correct mechanism
cannot be selected by consideration of stoichiometric equations but must
be determined by comparison of the trends of experimental data with the
trends manifested in the various rate equations.
Conversion Equations. In a steady-state flow system the relationship
between space velocity and conversion is obtained by consideration of an
elementary section of reactor containing a mass of-catalyst dW in which
a conversion dx is produced. Then
Fdx = rdW (63)
where F = feed rate, mass per unit time

W = mass of catalyst in reactor
T= reaction rate, moles/(mass of catalyst) (time)
X= conversion, moles per unit mass of feed
Integration yields
F (So pj
f - (64)
Jo r
where Sv = space velocity, volume of feed/(volume of catalyst bed)

(time)
PB = bulk density of catalyst, mass/volume
pf = density of feed, mass/volume
The integration of Equation (64) can be completed only when r is
expressed as a unique "function of x. In all except the most simple cases
of reactions at constant temperature and pressure this integration is best
carried out graphically by plotting x against 1/r. Where temperature
and pressure vary throughout the reactor, a step-by-step procedure simi-
lar to that-of Illustration 13, Chapter XVIII, page 875, must be employed.
Useful integrated equations may be developed from (64) for simple
reactions in which diffusional effects are negligible and temperature and
"pressure are constant. For example, Equation (36) represents the rate
of a reversible reaction which is monomolecular in both directions and is
controlled by a surface reaction involving single active centers. For a
feed composed of pure reactant A, if diffusional effects are neghgible and
activities are equal to partial pressures,
aAi = (1 — a;)ir
asi = XTT
where x = moles of A converted per mole of feed
v = total pressure
Substituting in Equation (36) and combining with (64) gives
W ^ l /*^[1 - K l - x)7rKA + XirKn] dx .
(65)
where C = kLK^, a constant at constant temperature

Rearranging gives
W ^ l r 1- r- [(I/TT
[(I/TT ++ KA)
KA) +
+ '{Kn
(KE -- KA)X]
KA)X] dx
dx
F CJo
cJo l/K)x\
[1 - (1 + 1/K)z] ^ '
Integration gives
W r ( l / x -f- KA)iX + l/K) -H {Kn - KAT\
In
L C(l + IIKY J
(67)
CHAP. XIX] CONVERSION EQUATIONS 929
Equation (67) contains three constants, C, KA, and KB, in addition to

the over-all thermodynamic equilibrium constant K. All constants are
assumed to be functions of temperature only for a given system. Once
these constants are evaluated, a complete relationship is obtained among
conversion, space velocity, and pressure, at conditions of constant
temperature and pressure.
In Table LVII are conversion equations for a few simple reactions in
which diffusional effects are neghgible. Similar equations for other
cases may be developed by the general procedure followed in deriving
Equation (67). Such equations are useful primarily for the interpreta-
tion of laboratory data in which conditions of constant temperature and
pressure may be closely approximated. Many of the equations involve
differences between logarithmic terms which require a high order of
numerical accuracy for satisfactory computations. For this reason
graphical integration of Equation (64) is frequently more convenient
than use of an integrated equation even where the latter is available and
applicable.
TABLE LVII
CATALTnc Co^fVEBSION EQUATIONS
Asstiming: Ideal-gas behavior
No reaction products in the feed
Surface reaction controlling
Nomenclature at end of table
A^R; X — moles A reacted per mole of feed
Single-Site Mechanism:
C (^^ - " " + PÂo f WAO \ _ ^ 1
\F)~ a' ^ \nAo -ax) a I (a)
o = (1 + \/K); a = (I/TT + KAn^ 4- Km:); P= {KR - KA)\
Dual-Site Mechanism:
w\F J \a a^ a^ JxuAo- ax) a 2a a^ ^'

where
a = l + l/K; a= {X/T^ + KAnAo + Km); ^ =- (KB - KA)
A'^R + S; X = moles A reacted per mole of feed
DwaZ-Site If ecâmsTn with a feed of pure A;raô= 1.0
v\F / 2Va \ a/ 1 _ xVa « " > (c)

where
a = l+T/K; « = (I/TT + KA); P= aZ-^ + Ks + Ks-KA))
Dual^Site Mechanism with a diluent / present in such large proportions that the
change in the number of total moles is negligible:
v\F J ~ \2a2 a/ \ riAo ) a

//3'(1 +2reôa) _ °g , . V j (2ai + 1 + 7 ) ( 1 - 7 )
\ 20' o / 7 '^(2ax + l - 7 ) ( 1 + T )
(d)
where
a = It IK; a = (I/TT + K^nô + X^/n/)
0=^ {KR + KS- Ki); 7 = V4nAoa + 1
4 + B ?± 2R
Dual-Site Mechanism; equimolal proportions of A and B in feed:
a; = moles of R formed per mole of feed
no = moles of A + B originally present per mole of feed
[ax"^ — 2ni)X + ri?)
"0
1 r 2a0no , j 3 2 n ? ( 2 - a ) 1 , ( o x - n o - 7) ( - n o + 7)
27 L a a2 J (ax-no+ 7)(-no-7) ^^'
o = (1 - 4/X); « = 1 /^ + (KA + i?B)n„/2 + X/n/
e = [Kii- {KA + X B ) / 2 ] ; 7 = VnS - ar^
Nomenclatuie
a = constant, defined for each equation
C = combined proportionality factor in the differential rate equation
written in terms of activities referred to unit fugacity
F = feed rate, moles per unit time
K = over-all thermodynamic equilibrium constant of the gas-phase reac-
tion
KA, KB, etc. = adsorption-equihbrium constants of components A, B, etc.
nxo = moles of A originally present per mole of feed
n/ = moles of inert component /, per mole of feed
W = mass of catalyst in the reactor
«> ft 7 = constants, defined for each equation
ir = total pressure in units used for expressing fugacity in the differential
rate equation
CATALYST FOULING AITO DEACTIVATION

In many high-temperature organic reactions in the presence of a solid
catalyst the conversion rapidly declines owing to the accumulation of a
carbonaceous catalyst deposit as processing is continued under constant
operating conditions. Frequently this fouling of the catalyst and the
necessity of removing the deposit causing ii^ constitute the most serious
problems encountered in the design of a process. Three fundamentally
different solutions to this problem are in commercial tj.se:
CHAP. XIX] CATALYST FOULING AND DEACTIVATION 931
1. Two or more alternate reactors are provided to permit continuous

operation in one while another is being regenerated by removal of cata-
lyst deposit. Regeneration is most commonly accomplished by oxida-
tion. In this case removal of the heat liberated becomes a major prob-
lem which may be handled by:
(a) Use of a reactivation gas containing so small a quantity of oxygen
that the heat of combustion may be removed with but a small ri:;3 in
temperature. This scheme normally involves recirculation of hot oxy-
gen-free combustion gases to which a small quantity of air is added.
(b) Indirect cooling of the catalyst bed by circulating a gaseous or
hquid heat-transfer mbdium. This requires that the catalyst be con-
tained in a specially designed reactor with extensive heat-transfer
surface.
(c) Storage of the heat liberated in regeneration in the catalyst bed
for' use in the subsequent processing period. This so-called " regenera-
tive adiabatic " operation is feasible only if the catalytic reaction is endo-
thermic and an exact balance can be maintained between the heat
evolved in regeneration and that absorbed in processing.
Equipment for the operation of alternating reactors with process
periods as short as 5 min has been highly developed in the Houdry
Process for the catalytic cracking of petroleum* in processes for the de-
hydrogenation of butane and butenes and other high-temperature
reactions.
2. The catalyst may be continuously removed from the reactor and
transferred to a regenerator where the deposit is burned. This principle
is employed in the TCC catalytic cracking process' in which catalyst
pellets are mechanically transferred from the reactor to a kiln in which
the deposit is burned. It is also followed in t h e " Fluid " catalytic
cracking process^" which employs finely divided catalyst suspended in a
" fluidized " state in the hydrocarbons being cracked and in the air used
for regeneration. In such " moving-bed " operations the time during
which the catalyst is in the reactor as determined by the rate of catalyst
circulation is analo§o*us to the length of the process period in an operation
employing a fixed catalyst bed.
3. The catalyst deposit may be continuously removed by addition of a
regenerating constituent to the reactant stream. This method has been
successfully used in dehydrogenating butene to butadiene. In this
8 R. H. Newton and H. G. Shimp, Trans. Am. Inst. Chetn. Engrs., 41,197 (1945).
' R, H. Newton, G. S. Dunham, and T. P. Simpson, Trans. Am. Inst. Chem. Engrs.,
41, 215 (1945). .
i» E. V. Murphree, C. L. Brown, F. J. Gohr, C. E. Jahmg, H. Z. Martin, and C. W.
Tyson, Trans. Am. Inst. Chem. Engrs, 41,19 (1945).
process" steam is the regenerating' agent and removes carbonaceous

material by the familiar water-gas reaction', producing a mixture of CO,
CO2, and hydrogen.
Quite different from the relatively rapid fouling of a catalyst by
carbonaceous deposits is its progressive deactivation with continued use.
This phenomenon of long-time deactivation is encountered either with or
without short-time fouling. Thus, each regeneration to remove catalyst
deposit may restore the catalyst to a slightly lower activity than the
preceding one. Such progressive loss in activity may result from irre-
versible poisoning, from structural changes of the catalyst, or from
gradual loss of some constituent of the catalyst. Continued service at
high temperature is known to reduce the surface area of some catalysts
and also to produce changes in lattice structure which are evidenced by
changed diffraction patterns. In many cases it appears that these struc-
tural changes are accelerated both by increased temperature and by
contact with water vapor. The effect of water vapor in promoting such
rearrangements of solid inorganic catalysts is particularly marked at
temperatures of 1000°F or higher.
It is common practice to compensate for deactivation of a catalyst by
progressively increasing the temperature or pressure of the reaction.
Loss of production may be avoided in this maimer for a time, but even-
tually the activity of the catalyst reaches a level at which further opera-
tion is uneconomical, and it must be either replaced by new catalyst or
subjected to some reactivation operation. Reactivation may involve
chemical treatment to remove poisons or to restore some lost constituent,
or may consist of alternate oxidation and reduction or solution and pre-
cipitation to restore the surface structiu-e of the catalyst.
Considerable confusion exists in the use of the terms " regeneration "
and " reactivation." It is behoved preferable to designate the removal
of catalyst deposit formed by short-time fouHng as regeneration to dis-
tinguish this operation from the periodic restoration of catalyst activity
by reactivation.
Fouling Factors. The effect of fouling on the results produced by a
catalyst is best studied by plotting the duration of processing time against
the conversion produced at fixed operation conditions. The conversion
plotted may be either the differential or instantaneous conversion being
produced at the time in question or the integral or average conversion
from the begirming of the process period to the time in question. The
former type of curve may be established experimentally from a single
long process period byanalysis of a series of " spot " samples t a k e n a t
'^R. P. Russell, E. V. Murphree, and W. C. Asbury, Trans. Am. Inst. Chem.
Engrs., 42, 1-14 (1946). I
CHAP. XIX] FOULING FACTORS 933
intervals. The latter type of curve is established by analyzing the total
product from a number of process periods of different length. Once
either curve is established the other can be derived by graphical inte-
gration or differentiation.
In Fig. 182 are curves from the data of Johanson and Watson'^ relat-
ing both instantaneous and average conversions to process-period length
for the production of toluene from an equimolal mixture of benzene and
50
1 1 1 1 1 L 1 1 1
40
^
Y>
30
k^-^- -
^>»
•«». " ^ A ^S. ^V '^ ***'^ -968° F. -
^ ^ \'--*^^*V^
•S^^>.1020°F. "~ -932 °F.
0)
glO _
M 9 Z'860°F.--
-
g 6 ^ ; ' ^ 6 8 ° F.
u 788 "F.
\ 1020°F.
2 -- -
1 .F
1 1 \ 1 I 1 1
100
200 300 400 BOO
Process period T (min)
FIG. 182. Effect of Process Period upon the Production of Toluene.
(Pressure = 31.5 atmospheres: W/F = 66.)
xylene passed at a liquid hourly space velocity of 1.0 over a silica-alumina

cracking catalyst at a gauge pressure of 450 lb per sq in. and at the vari-
ous -indicated temperatures. The instantaneous conversions were
determined and the average conversions calculated.
It may be noted that the rate of fouling of the catalyst, as indicated by
the slope of the curves of Fig. 182, increases rapidly with increased
temperature. Thus, although for short process periods conversion
increases progressively with increased temperature, with longer process
'" L. N. Johanson and K. M. Watson, Natl. Petroleum News, Tech. Sec. (August
and September 1946). Also " Principles of Reactor Design," National Petroleum
934 CATALYTIC REACTIONS [CHAP. X I X
periods the conversion at 1020°F drops below that obtained at much

lower temperatures.
It is evident that the complete rate equation for a catalyst which
undergoes fouling and periodic regeneration must include the effect of
process-period length, or apparent catalyst residence tirhe in the case of
moving-bed operations. There are not sufficient data available on any
one system to estabhsh with certainty the best procedure for introducing
such a factor into a fundamental rate equation. A " fouUng factor "
might be introduced as a multiplying factor in the numerator of the rate
equation with the philosophy that folding merely reduces L, the total
number of active centers on the catalyst surface. An alternate pro-
cedure is to introduce the fouling factor as an added term in the group
in the denominator of the rate equation with the view that fouling has
the same effect as the equihbrium adsorption of a constituent on the
active centers. In either case the fouling factor would be a function of
temperature, pressure, conversion, and process-period length.
Johanson and Watson obtained fair correlation of their data with a
rate equation in which a fouhng factor was introduced in the latter form.
The principal reaction under consideration was as follows:
CeHe -t- C6H4(CH3)2 ^ 2C6H5CH3 (68)

There is evidence that toluene is also simultaneously produced by dis-
proportionation of xylene to form trimethyl benzene and toluene. How-
ever, for the restricted case of a feed of equimolal proportions Johanson
and Watson assumed that the over-all rate of production of toluene
could be expressed by the empirical evaluation of the rate equation of
reaction (68). On this basis it was found that the average conversion
data over wide ranges of conditions were fairly well represented by the
following basic rate equation:
r = ^^''^'''^ ~ « « / - ^ (QQ)
(I + OAKA + GBKB + UBKB + Biry ^
where subscripts A, B, R refer to benzene, xylene, and toluene, respec-
tively, and
BIT = the fouling factor = ^(MT* + VT'X) (70)
T= process-period length
u,v = functions of temperature
b,c = constants
X= average conversion, moles of toluene per mole of
feed of equimolal composition produced from
the beginning of the process period to time T
CHAP. XIX] ACTIVITY FACTORS 935
In Figs. 183, 184, and 185 are average conversions calculated by inte-
gration of Equation (69) to show the effects of varying pressure, space
velocity, and temperature, respectively, at process-period lengths T of 0,
40
I 30-
1 1 1 1 1
.."^ ^T^BO
^^^ -
9. ^ T=300
1
II
10 -
» 1 1 1 1 1
0 10 15 20 25. 30 35
PressTire TT, Atm
FIG. 183. Effect of Pressure on the Production of Toluene from an Equimolal Mix-
ture of Benzene and Xylene. WjV = 66; temperature = 932° F.)
50, and 300 min. It may be noted that the length of the process period
is an important variable which must be carefully considered in the engi-
neering analysis of such a process.
The methods followed in evaluating the constants of Equation (69)
are discussed on page 938.
50 T • I - 1 1 1 1
•§
§40-
&
o .^v^'^
I 3 0 ~-"
^/J^--^ _
% 20 - .^^—- -^—'-
a
ft-10

•3
& ^ 1 1 1 1 1 1
II 0 20 40 60 • 80 100 120 140
W/F^Mss.s. Catalyst) (Hr) / (Mole Feed)
FiQ. 184. Effect of Catalyst-Feed-Rate Ratio on the Production of Toluene.
(Temperature = 932° P; pressure = 31.5 atm.)
Activity Factors. The constants in catalytic rate equations are

strictly applicable only to the catalyst for which they were evaluated and
are rendered invalid by any changes in catalyst activity. In some cases
a change in the catalyst due either to changed preparation or to deactiva-

tion in use may markedly change both the'adsorption-equilibrium con-
stants and the over-all rate factor. In this event all constants must be
re-evaluated at different levels of catalyst activity in order to maintain a
800 900 1000

Temperature, °F
FIG. 185. Effect of Temperature on the Production of Toluene.
{W/F = 66; pressure = 31.5 atm.)
valid rate equation. However, where changes in catalyst activity are

not too great it is frequently possible to take them into account merely
by including a catalyst-activity factor in the over-all rate factor of the
equation. For example, Equation i 69) might be written:
LrC'iaAaB - a%/K)
r = (1 + aAKA + asKB + aRKs + Bir^ (71)
where L. the catalyst-activity factor, a constant depending only on

catalyst activity
It is evident that the factor L, will appear,as a multipHer in any inte-
grated-conversion equation of the type shown in Table LVII. Thus,
if the same conversion is produced by catalysts of different activities
they must operate at different space velocities, and, under these
conditions,
(72)
If catalyst 1 of Equation (72) is adopted as an arbitrary standard of com-

parison and assigned an activity factor of 1.0, the activity factor of any
other catalyst of generally similar type may be determined by experi-
mentally evaluating the conversion produced at a selected space veloc-
ity. If the conversion produced by the standard catalyst at the same
conditions is known, the activity factor is calculated directly from
Equation (72).
CHAP. XIX] EVALUATION OF RATE EQUATIONS 937
Ordinarily catalyst activities are compared at standardized test con-
ditions. Frequently activity ratings have been based on merely a ratio
of the conversions obtained in such a test. Such a scale may give a dis-
torted picture of the relative activities as indicated by Fig. 186 in which
conversion x is plotted against W/F for a standard catalyst. A catalyst
producing a conversion represented by point B has an activity factor
equal to the ratio of the abscissas at point C to that at B. Similarly, a
W/F
FIG. 186. Catalyst Activity Factors.
catalyst producing the conversion indicated by D has an activity factor

equal to the ratio of abscissas E/D. Such factors are much more
descriptive of the change in properties of the catalyst than would be
ratios of the ordinates B/A and D/A, and they may be used directly in
rate and conversion equations in ranges where the adsorption-equilibriimi
constants are not seriously changed. The only additional work involved
in expressing activities on this basis is the evaluation of the relationship
between x and W/F for the standard catalyst.
EVALUATION OF RATE EQUATIONS

Each of the equilibrium and velocity constants of the rate equations
is a function of temperature in accordance with the following equations
discussed in Chapters XVI and XVIII.
(73)
1- = p BT "^ R (74)
where «* = TW
7* = activity coefficient of the activated complex
Vm = molal volume of the activated complex.
The incorporation of (74) in a rate equation is simplified by combining

all the multiplying constants into an over-all rate constant C. Thus,
for Equation (45),
ah
and this equation may be written
e «•'
r = (1 + aAiKA + aBiKs + auiKR + asiKs + anKi -\ Y
( aAittBi aRiasi\
—j
,__ .
(75a)
It is evident from consideration of Equations (73) to (75a) that each of

the constants in the catalytic rate equations can be expressed as a func-
tion of temperature by evaluating it at two different temperatures.
Thus, in order to establish completely the kinetic-behavior of a system,
sufl&cient rate data must be obtained to permit evaluation of each con-
stant of the fundamental rate equation at two temperatures. Further-
more, the data must be sufficient to establish the mechanism of the
reaction and to identify the rate-controlling step in order that the proper
equation form may be used.
In order that the constants of a rate equation maiy be evaluated at a
selected temperature, it is necessary that rate data shall be available
which include at least two different values of each variable so that a
number of independent equations may be established which is equal to
the number of unknown constants. These equations are then solved
simultaneously to determine the constants. Ordinarily, more than this
theoretical minimum of data is required in order to establish the correct
mechanism and equation form.
Evaluation of Constants by Least Squares. In developing the con-
stants of an equation to represent experimentally observed data it is
desired to arrive at the most probable or " best " relationship which will
represent all of the data with a minimum average error or deviation.
Where errors are small and only a single unknown constant is involved,
it is generally satisfactory to calculate the constant from each of the
available experimental observations and to adopt the arithmetic average
value. Where large deviations are encountered, this procedure does not
lead to the most probable solution. Furthermore, it is not applicable to
the simultaneous evaluation of several constants.
The mathematical theory of probability has resulted in the develop-
CHAP. XIX] METHOD OF LEAST SQUAEES 939
ment of the Gaussian law of error which leads to the conclusion that the
most probable value of a quantity which is derived from a number of
experimental observations of equal absolute precision is that value from
which the sum of the sqvnres of the deviations of the individual measure-
ments is a minimum. The derivation of this " principle of least
squares " may be found in standard mathematical texts which deal with
probabihty.
The principle of least squares is particularly valuable for evaluating
equations which can be written in a form which is Unear with respect
to each unknown constant. This general form isl
». y = a + bxi + cxc (76)
where y, Xt, Xc- • • = experimentally determined quantities which do
not involve unknown constants
a, b, c • • • = unknown constants
The deviation D of any observed value of y from the most probable value
calculated from the corresponding values of Xb, a;,, • • • is expressed by
D = a + bxi, + cxc- •• -y (77)
According to the principle of least squares the most probable values of
a,b,c--- are those which result in a minimum value of SD^ for all of the
n experimental observations involved. Thus,
1:D^ = Dl + Dl-i +Dl = minimum (78)
Since SZ)^ is a function of the unknowns a, b, c- • • its minimum corre-
sponds to a zero value for the derivative of SZ)^ with respect to each of
the unknowns. Thus, with respect to a,
da -[-f--^---^-]- (79)
Similar expressions may be written with respect to each of the other

unknowns. The terms dD/da, dD/db, etc., may be evaluated from
Equation (77). Thus, dD/da = 1, and
diZD^)
•—-— = 2(a + bxbi + cXci-\ -yi
da
-\ a + bX6„ + CXcn +. 2/n) = 0 (80)
Similarly, since dD/db = Xj,
——— = 2(0x1,1 + bxli + cxciXbi H -Xiiyi

do
-\ aXbn + bx^tn + CXcnXbn + • •• —XbnVn) = 0 (81)
Similar equations may be written with respect to c and any other

unknown constants in the basic equation. These equations may be
written in summation form as follows :
na + bUXi + cSXc + • • • — Sy = 0 (82)
oSx6 + bi:xl + cXXcXb + . • • — 2yxb = 0 (83)
aZXc + hI,Xi;Xc + cSx^ + • • • — 'LyXc
0 (84)
Each of the summations of Equations (82-84) may be evaluated from the
n experimental values of Xh, x^, • • • and y. Thus three equations are
obtained which may be solved simultaneously for the most probable
values of the three unknowns, a, b, and c. If more constants are in-
volved, additional equations may be developed by the same procedure.
Illustration 1. In the life test of a catalyst weekly determinations are made of the
activity factor L,. These data for a period of 9 weeks are shown in Table A and
plotted in Fig. 187. Assuming a linear relationship between the activity factor and
time, estimate its value after 180 days of service.
TABLE A
Service Time., Activity Factor,
Observation days, T T» TLr
1 0 39 0 0
2 7 40 49 280
3 14 38 196 632
4 21 38 441 798
5 28 35 784 980
6 35 38 1225 1330
7 42 37 1764 1654
8 49 35 2401 1715
9 56 3136 2072
S 252 337 9996 9261
The relationship assumed is L, = o + br. Thus, in applying Equations (82-84).
y = L, and Xb = T while the other variables are all zero. The two equations required
for evaluation of o and 6 follow from (82) and (83). '
na + 6ST - SL, = 0 (a)
OST + bSr' - SrLr = 0 (b)
The summation in Equations (a) and (b) are evaluated in Table A.
With these substitutions, since 7» = 9,
9o + 2626 - 337 = 0 (c)
252o + 99966 - 9261 = 0 (d)
Multiplying (e) by 262 /9 to eliminate o, we get '
252a + 70566 - 9436 = 0 (e)
CHAP. XIX] METHOD OP LEAST SQUARES 941
Combining (d) and (e) gives
2940b = - 1 7 5 ; b = -0.05952
From (c), a = (337 + 15.00) /9 = 39.11
The most probable equation is therefore Lr = 39.11 — 0.05952T, and, when T = 180,
Lr = 39.11 - 10.71 = 28.40. The curve in Fig. 187 is a plot of this equation.
50 r
10 20 30 40 60 60
Catalyst Service, r (Days)
FIG. 187. Most Probable Life of Catalyst.
If more extensive life data were available, a better basis for prediction would be a
three-term equation of the forni Lr = a + 6r + vr\ The constants of this equation
could be determined by simultaneous solution of three equations developed from
(82-84).
Use of the method of least squares requires a high order of numerical

accuracy in the calculations. Where three constants are involved, these
are most conveniently evaluated by determinants.
It may be noted that,' although Equation (76) is linear with respect to
the constants a,b,c---, the variable terms y, Xb, x^- • • may be complex
functions of any form. For example, the initial rate of a unimolecular
reaction controlled by a dual site with surface reaction in a pure feed
containing only reactant A is expressed by the following equation:
ro = ,. . ^ : ., (85)
(1 + KA^Y
1 , KAB
(86)
In this equation y = Vir/ro; a = 1/VC; b = KA/VC, and xi, = v.

The constants may be evaluated from experimental data for ro and ir by
the procedure of Illustration 1. Similarly, Equation (73) may be
rewritten in a form which is linear with respect to the unknown constants

Ai?°andAS°. Thus,
, ^ AS° AH°
-AH°
In this case y = In E^; o = AS°/R; b= , and Xi = \/T. The
method of least squares is readily applicable to determination of
A/S° and AH.° from data for K and T. In all such rearrangements
it is necessary that the form of Equation (76) be obtained with y,
a dependent variable or group of variables which contains no unknown
constant.
It is evident that the method of least squares may be applied to the
evaluation of constants in relationships of many different forms. How-
ever, it should be remembered that the basic assumption oi the laethod is
that the absolute precision of measurements is constant over the entire
range. Thus, in applying the method to Equation (86) it is assumed
that the precision of determination of Vir/ro is constant, while for
Equation (87) it is assumed that the absolute error of measurement of
In K is constant. This latter assumption corresponds to a constant
'percentage error in the measurement of K. The method of least squares
is not rigorous where the methods of measurement are such that this
fundamental assumption is not correct.
The method of least squares may be used to determine the best con-
stants in a given form of equation and also for establishing the best form
of equation for a given set of data. The best constants are first evalu-
ated for each form of equation under consideration. The best form of
equation is then that for which the average of the squares of the devia-
tions (SD2)/n is the least.
Differential Rate Data. Where accurate methods of chemical
analysis are available, rate equations are best established from direct
differential rate measurements in a differential reactor containing a
bed of catalyst so small that relatively small changes of composition
are obtained. Such apparatus should be designed to produce the
smallest composition changes that permit accurate evaluation of the
rate of reaction with the available analytical methods. Because of
the exponential influence of temperature, it is desirable that the dif-
ferential bed be at as imiform temperature as possible, requiring a
small-diameter reactor with good heat-transfer provisions in the walls
of the container.
In a differential reactor the average activity of each component may be
taken as the arithmetic average of the inlet and outlet values for small
CHAP. XIX] DIFFERENTIAL RATE DATA 943
changes or, better, derived from the logarithmic mean of the differences
between the activities at the inlet and the outlet for the reactants. The
mean activity of a product R may be taken as Ofl, minus the log mean
value of {uRe — o«) at the terminal conditions where age is the activity of
R when the reaction proceeds to equiUbrium. This use of log mean
activities is merely an improved approximation which in general is not
satisfactory except where concentration changes are small.
The adfeorption-equihbrium constants which appear in the rate equa-
tions theoretically might be independently determined by adsorption
measurements on the individual components. Such measurements may
prove to be a valuable adjimct to direct catalytic rate measurements but
are open to question because of the interaction effects and variability of
active centers previously mentioned. Until these relations are better
understood, these constants in the rate equations are best determined
in the specific reaction under study by making rate measurements over a
wide range of concentrations.
To minimize tedious algebraic solutions of simultaneous equations, it is
desirable in complex systems to determine the effective average value of
the absorption-equihbrium constant of each component by a constant-
temperature series of at least three differential rate measurements in
which the activity of that component is varied over a wide range while
the activities of all other components are held constant. A convenient
procedure is to make an initial run with the activity of each component
at approximately the middle of the range of interest. Additional series
of two runs each are then made in which the activity of each component
is varied from a minimum value in one to a maximum in the other while
the activities of all other components are held constant. This procedure
may require two or more trials at some of the conditions with interpola-
tion in order to obtain satisfactory constancy of the grouped activities
but is not difficult with a flexible differential-reactor apparatus which
permits wide variation in feed rates and depths of catalyst bed.
Once the adsorption-equilibrium constants are all evaluated, each
single rate measurement permits calculation of the over-all rate con-
stant C. These methods are demonstrated in the following illustration.
Illustration 2. Mixed iso-octenes, commercially known as codimer, are hydro-
genated in the vapor phase to the corresponding iso-octanes according to the fol-
lowing reaction:
CsHiaCg) + HaCg) ^ CsHisCg)
The reaction as it takes place on a supported nickel catalyst was studied in a pilot
plant which was operated with a small bed of catalyst to approximate differential
conditions. The experimental data are tabulated in Table A. These data were
obtained when the catalyst was used at sufficiently high mass velocity that diffusional
gradients were shown to be negligible, and the activities at the surfaces of the catalyst
pellets may be assumed equal to those in the fluid stream. From these data it is
desired to determine the probable mechanism of the reaction and to evaluate the con-
stants of the corresponding rate equations. These experimental data and their inter-
pretation are taken from the work of Tsehernitz, Bornstein, Beckmann, and Hougen."
The experimental procedure was planned to estabhsh the effect of the average
partial pressures of each component while holding the average partial pressures of the
other two components nearly constant. However, the exact attainment of such con-
ditions is not practical, since this would predicate advance knowledge of the exact
effect of all the variables controlling the reaction rates.
In the operation of the experimental plant the rate of Uquidflowwas measured from
. TABLE A
EXPBBIMENTAI. D A T A
Temperature = 200°C
r
average values lb-mole /
Run V PB Pa PS ab)(hr)
Id 1.09 0.482 0.100 0.508 0.00353
3c 3.50 2.459 0.527 0.515 0.0250
3d 3.49 2.450 0.530 0.515 0.0320
10a 3.51 0.477 0.494 2.538 0.00553
11a 1.51 0.514 0.540 0.455 0.00870
lid 1.50 0.473 0.552 0.473 0.01392
lie 1.50 0.470 0.558 0.473 0.00960
12ab 1.105 0.104 0.562 0.440 0.00514
14a 3.52 0.450 2.840 0.230 0.01920
14b 3.51 0.409 2.810 0.289 0.0206
25a 2.50 0.484 1.075 0.942 0.0131
28b 2.10 0.357 1.590 0.153 0.0186
Temperature =
-- 275°C
f
average values lb-mole /
Run T PH Pv PS Ob)(hr)
2a 1.105 0.478 0.102 0.525 0.00298
4d 3.520 2.505 0.518 0.497 0.0389
4e • 3.500 2.500 0.517 0.485 0.0450
5b 3.510 0.425 2.770 0.310 0.0206
5d 3.500 0.433 2.800 0.270 0.0185
6bc 1.104 0.489 0.562 0.051 0.0180
7b 1.500 0.546 0.506 0.446 0.0103
7cd 1.500 0.469 0.556 0.475 0.01215
8bc 1.109 0.100 0.540 0.464 0.00705
9a 3.500 0.422i 0.462 2.615 0.00824
9b 3.610 0.467 0.485 2.555 0.00734
26ab 2.500 1.495 0.511 0.495 0.0319
29ab 2.100 1.222 0.776 0.103 0.0435
*' J, Tsehernitz, S. Bornstein, R. B. Beckmann, and O. A. Hougen, Trans. Am.
Inst:Chem. Engrs., 42, 883-903 (1946).
CHAP. XIX) HYDROGENATION OF CODIMER 945
Temperature = 325°C
average values lb-mole/

Run w PH Pa Ps ab)(hr)
15a 3.50 0.500 2.735 • 0.270 0.0201
16a 1.10 0.482 0.660 0.062 0.0U4
17b 1.50 0.501 0.533 0.466 0.0134
17c 1^50 0.475 0.653 0.471 0.00946
17e 1.50 0.476 0.656 0.478 0.00871
18a 1.10 0.101 0.548 0.451 0.0025
19b 3.50 2.335 0.550 0.610 0.0338
19c 3.50 2.540 0.452 0.510 0.0286
19d 3.50 2.405 0.624 0.572 0.0297
20ab 1.10 0.489 0.098 0.513 0.0044
21a 3.51 0.480 0.480 2.550 0.01106
24a 2.50 0.555 0.299 1.645 0.00533
27a 2.50 0.599 0.996 0.907 0.01885
30a 2.10 0.438 1.523 0.136 0.01414
30b 2.10 0.478 1.480 0.144 0.0159
the volume and density of the outgoing liquid, the outgoing hydrogen by a displace-
ment meter, and the degree of conversion from the index of refraction of the ingoing
and outgoing liquids. The composition of the liquid was related to the refractive
index by previous calibration pf known samples analyzed for unsaturates primarily
by hydrogen absorption and secondarily by standard chemical analysis by the
bromine method. The following ranges of variables were investigated: pressures
from 1.0 to 3.5 atm, temperatures from 200 to 325°C, and feed compositions from 10
to 90 mole per cent of each component. The five different feed stocks used were
previously purified by steam distillation to separate polymers and by passing through
a bed of nickel catalyst in the absence of hydrogen to remove sulfur. Temperatures
above 350°C could not be investigated because of catalyst disintegration. The calcu-
lations involved in a specific run are illustrated herewith.
Blank tests were made in the pilot plant in the absence of the catalyst to determine
the extent of the uncatalyzed reaction. This was not detectable at 200°C and 275°C.
At 325°C a slight reaction took place amounting to 1 to 5 per cent of the catalyzed
reaction depending upon feed composition. This correction was made to establish the
rate of the catalytic reaction at 325°C.
RUN 3C:
Feed stock II, 54.1 mole per cent unsaturates
Average molecular weight 112.9
Temperature: Entering 195°, leaving 206°, average 200.5°C
Catalyst, 0.0440 lb; density = 86.4 lb per cu ft; volume = 0.000509 cu ft
Pressure, 3.500 atm
Liquidflowleaving reactor, 1206 milliliters per hr; specific gravity 0.710
Index of refraction at 25.00°C of Uquid entering 1.40580
Index of refraction at 25.00°C of Uquid leaving 1.40453
Difference 0.00127
Factor for conversion to percentage change = 5160
Percentage change in composition of unsaturate = (5150) (0.00127) = 6.54
Liquid leaving = .,,„;,^ " .^^ = 0.01671 lb-mole per hr

(112.9; (454j
Conversion = (0.01) (0.01671) (6.54) = 0.001093 lb-mole per hr
0.001093
'' "^ nn44n = 0.02484 lb-mole / (hr) (lb catalyst)
Hydrogen leaving = 14.02 cu ft per hr (standard conditions)

14.02 1
= —— = 0.03905 lb-mole per hr
MATERIAL BALANCE
lb-moles Mole Partial
Entering per hr Per Cent Pressures, atm
Unsaturates = (0.541) (0.01671) =0.00904 15.90 0.557
Saturates = (0.459) (0.01671) =0.00767 13.49 0.472
Hydrogen = 0.03905 + 0.00109 = 0.04014 70.61 2.471
Total 0.05685 100.0 3.500
Leaving
Unsaturates = 0.00904 - 0.00109 = 0.00795 14.26 0.499
Saturates = 0.00767 + 0.00109 = 0.00876 15.71 0.550
Hydrogen = 0.03905 70.03 2.451
Total 0.05576 100.0 3.500
1^
Since the change in composition is small, the arithmetic average partial'. pressures may
be taken. Thus,
Pv = 0.628
PS = 0.511
PH = 2.461
X = 3.500
Cross-sectional area of bed (1.05 in. diameter) = 0.00601 sq ft
Enteringflowrate = 0.00905(112) -1-0.00770(114) +0.0401(2) = 1.972 lb per hr
1 972
G = mass velocity = ^^~l = 328 lb/(hr) (sq ft)
At this mass velocity and reaction rate r, the transverse partial-pressure gradients due
to mass transfer are negligible.
Gaseous space velocity of entering feed = —' — = 40,100 cu ft (at CC,
latm)/(cuft)(hr)
Over-All Equilibrium Constant. Before the mechanism of this reaction was
studied, thermodynamic calculations were made to determine the reversibility of the
reaction within the experimental range of pressures and temperatures. The equili-
brium degrees of conversion in the hydrogenation of codimer are plotted against tem-
perature for various total pressiu'es in Fig. 188. These data were calculated from the
standard free-energy equation based upon the hydrogenation of 2,4,4-trimethyl
pentene-1:
ACTT = -RT \n K = -25,289 + 11.217 logwT - 0.001267r» + (0.0465) l O T *
- 4.599r (a)
CHAP. XIX] HYDROGENATION OP CODIMER 947
This equation waSj^stablished from standard absolute entropy values supplied by
Parks." The standard heat of formation of the paraffin was measured by Rossini,"
and heat capacity of the paraffin was taken from the data of Pitzer and Scott.''
The heat of formation and heat capacity of the olefin were estimated from Tables
XXXIX-XLIII.
«1.0
I 0.8
m
I 0.6
a S.V
A \ * 3
6 0.4
Y.
.2 0.2
•§0.0 700 900 1100 1300 1500
Temperature, °K
Fia. 188. Equilibrium Conversion of Codimer to Hydrogenated Codimer.
It may be observed from Fig. 188 that up to 650°K the reverse reaction is negligible.
It was also found experimentally by Beckmann"that no decomposition occurred when
vapors of iso-octane were passed through the catalyst at 300°C. Within the range of
these experimental studies the effect of the reverse reaction can be neglected.
Determination of Mechanism of Reaction. Determination of the most plausible
mechanism in the hydrogenation of codimer is based upon the assumptions that only
one chemical step is rate-controUing. Upon this assumption 17 different mechanisms
seem possible for the catalytic hydrogenation of codimer in the vapor phase. Diffu-
sion of the gaseous components occurs independently and simultaneously with the
chemical steps, but under the selected experimental conditions these steps offered
negligible resistance. The possible chemical steps, together with their corresponding
rate equations, expressed in convenient forms for comparison, are fisted in the follow-
ing pages. In comparing these equations with those presented previously (pages
911-926) it should be noted that the terms involving the reverse reaction drop out
because of the high equifibrium constant.. The subscript H refers to hydrogen, U to
the unsaturated component, and S to the saturated component.
TABLE B
POSTULATED MECHANISMS
I. Reaction between molecularly adsorbed hydrogen and adsorbed codimer:

(o) Adsorption of hydrogen controlling:
R = — = - (1 + Kupu + Ksps) = o + bpu + cpa

r a
" G. S. Parks, Private correspondence.

" F . D. Rossini, Chem. Rev., 27, 1-16 (1940).
" K. S. Pitzer and D. Scott, J. Am. Chem. Soc, 63,2419 (1941).
" R. B. Beckmann, Ph.D. thesis. University of Wisconsin (1944).
948 CATALYTIC REACTIONS [CHAP XIX
(b) Adsorption of codimer controlling:

Vv 1
B = = - (1 + KHVH + Ks^pa) = a + fvn + cpa
T a
(c) Desorption of hydrogenated codimer controlling:
VHVu 1
R r/r „ •. (1 + KHVH + Kxfpu) = a+fpH + bpa
r
(d) Surface reaction controlling:
R = ^/ =^ „ „ (1 + ii^Hyi/ + Kupu + Ksps) = a +fpB + bpv + cps
II. Reaction between atomically adsorbed hydrogen and adsorbed codimer:

(e) Adsorption of hydrogen controlling:
R = J y = J - (1 + ûPu + Ksps) =a + bpu + cps

•\l r yjct
(/) Adsorption of codimer controlling:
/2 = — = - (1 + VKHPH + Ksps) = o +fV^ + cps

r a
(ff) Desorption of hydrogenated codimer controUing:
R = ^ ^ ^ = - ~ - (1 + ^KHPH + Kvpu) = a +fV^ + bpt.

r UKHKXJ
(Ji) Surface reaction controlling:
R
= a + fVp^ + bpu + cps
III. Reaction between codimer in the gas phase and molecularly adsorbed hydro-
gen:
(t) Adsorption of hydrogen controUing:
fi = — = i ( 1 + Ksps) = a + cps
r a
(j) Desorption of hydrogenated codimer controlling:
R = ^^^ = ^ ( 1 + KHPH) = a +fpB
(k) Sm-face reaction controlhng:

PHPU 1
R = = - ^ (1 + KHPH + Ksps) = a + fpa + cpa
r ati-H
CHAP. XIX] «? HYDROGENATION OF CODIMER 949
IV. Reaction of codimer in the gas phase and atomically adsorbed hydrogen:
(0 Adsorption of hydrogen controlling:
=^/? = ^^^•
{m) Desorption of hydrogenated codimer controlling:
R = mSR = - ^ (1 + y/KaVH) = a + / \ ^
r octiit
(n) Surface reaction controlling:
g = J ^ ^ = / ^ (1 + VKJIPH + Ksps) = a+fV^ + cps
V. Reaction between hydrogen in the gas phase and adsorbed codimer:

(o) Impact of hydrogen upon adsorbed codimer controlling;
VHVU 1
R = = —^ (1 + KvPu + Ksps) = a + bpu + cps
r UKH
(p) Desorption of hydrogenated codimer controlling:
R = iSJPu^ - ^ (1 + KuPu) = a + 6pw

r aKu
(g) Adsorption of codimer controlling:
1
fl = ' — = - (1 + Ksps) = a + cps
r a
VI. Uncatalyzed reaction in the gas phase:
(r) B = —^ = a
PHPJ/
Interpretation of Data. The constants of the equations for the various proposed
mechanisms were evaluated by the method of least squares. The four constants,
a, h, c, and / were evaluated for each mechanism. It is a requirement of the theory
upon which these equations are based that all constants must have positive or zero
values. As an illustration, the method of evaluating the constants for mechanism d
are given for the 12 runs made at 200°C.
The four simultaneous equations for evaluating the four constants by the method
of least squares are as follows, where K = / and n = 12.
an + fcSpu + c2ps + /2pH = 2/2 (b)

aSpH + VLp'HPu + cSpHPs + /2p/7 = 2p^B (c)
o2pc7 + &Spu + cZpups + f^PuPH - ^PuR (d)
o2ps + bXpspu + cSp% + fipsPH = SpsB (e)
950 CATALYTIC REACTIONS [CHAP. X I X
£5 ffd
00
05
in
"5
00
CO CD
i n N-
>-<
w
CO
c^ ?!g CO
05
CO
O
CO
t - CO
% ^
O
03
(N
05
IN
CO r ^
IN IN
to
N
to in
IN O
«> Ml
CO •-H
CO
in
CD
m
in
CO
a O 1—1 'H O O o o O >-i - H O O t^
CO t - in lO IN t^ Oi CO m
CO to <N 1^ CO o O CO t ^ - * b - O
C<3
IM
•*
O
O
O
IN
C^
S
IN
N
c^
IN
cq
rt
O
O CO CO CO
CO I-H C I - - l
CO
O
1
O CD to o O o oo o O O O CO
I-H
t3
.-1 M N • ^ •* «^ W CD O i i n T j i •*
s ^'C iO CO CD
T-H
• *
t~
o
CO CO
IN IN
CO CO CO t -
C5 i n 0 0 0 0
CO
C»
in
CO
at N IN (N IN I N C<) i-i O O 00 O rH
^
• *
o o o to o o o o o O O O 05
O 00 Ol o to 1> TjH 0 0 CD 1—1 CO T-^

S t - o in m CD i n 0 0 1-1
1-1 1 >
O IN
00
IN
^
^ »-( ?^^
n>
IN
^ H CO c:2 i n CO
CO CO CO O 0 0 1-1 i n
00
to
o o oo O o o O 00 t - 1-1 CO i-H
CO
I
00 00 in o to IN O 00 ^ CO O O CO
CD »-l CO i n to CO CO CO <-< T j i CO i n CD
C5 CO t ~ CD CO m •* i n CO CO t ^ CO i n 05
o CD
o
& to 03 CD i n C3
T_t
t- f-i '(I i n 00 CD CO CO
§ ^ M
1-1
i-H
lO
O IN
•* • *
00
CO 1-1
t ~ 1-H
CO
CD i n
in
cn o
CO CO
o
CO
u S5.
•*
o
o /-^ O CO CO CO • *
I
•2 .3 m o o in r^ c» to ^ CO CO 1-1 CD
o ^ CO i n IN in m CO 05 in 03 o CO 0 0
O C<1 CO o •-I O O —1 CO 1-H 1-1
o o o o 8 O O O O O O O
•<
o o o o o O O O O O O O
IN Ci O r^ • * CO o T)l O 05 ^ t^ C5
b 00 in i n r^ t^ t - o in o 00 in CI
a. Tti Tji - ^ 'ti in T t l TJH T—1 T j l M l •* CO 1-1
O IN IN o o o o o o O O O OS
00 in i n no lO CO CO o o C2 C? CO tH
w O ^-1 ^H in « > lr~ T h CO 0 0 • * in CO
sa. in in in in Tti Ttl T j l - * CO CO 0 3 i i in
O O O IN o o oo o O O O t^
o t - o • * o CO 00 CI O o in o 00
O N CO m Ml in i n CD T * 1-1 w as t -
>3 rH in in in in in in CO 0 0 o in 1-1
a o o o
-* o o o o O CO CO 1-1 1-1 CO
rH
s '^ O T 3 oii en T ) O -a « x> la XI ^

•—1 CO CO o T-i i-H CO •*
c=5
CHAP. X I X ] HYDROGENATION OF CODIMER 951
~ 00 CD CO i n lO TH lO CO o i OS l O •* -* •«•
u 00 •* o 00 O <3S f-H CO CO CO CO o 1* M
O oj
K
tt
*"
t^ >*
1
00 >-< 0 0
•"^
CO i-H
I—1
" 1
CO
1
111
»
-H î
a. 1 1 45,
•J?
^ (N o n*
CO o o
O 1-1 N
rt
O
•*
O
CD O
C 3 CO ^
O CO CD
O
03
O
CO o
C d CO
" - I 1-1
CO
CD
O
i
CO
o o o o •-I O o o o o o o II
Ol ( M T-l M (N O on o o m O U3
(N CO l O rt O
ta
«o
CO >n CO CO CO TjH eq
O O O O .-1 O o o o o o o
< 1 1 1 1 1
(a. CO 0 0
CO r H
lO
CO
O i n
l O CO
(M o CO • * CO o o
CO CO 1-1 • * CO in
f«[§ CO » CO CD CO 0 0 w CO in
t» CO CD ( N CO a <N 00 •* 1-1 O lO
oi 00 0 0 CO CO CD CO rH 00 0 0 rjH t -
^<3 CO CO CO CD • * • * • * •* t^ l > i n i n
-* CD t - i-i T H CO Tt. i r j »o O O r f i t < » 0 CD l O TjH CD in
t * CD »C T-i 0 > O • * CO CO O
^ o en in
I « •-I i >
r-1
l o CO ( N
T-l
(M I N O CO CO CO 1-1
i n
d
C32 0 0 O
CD 0 2 t -
t - —1 CO o
i >
a s t - - i * CO 0 0
oq
t^
CD l O
m CO
N
•* s
9 a.
ft?
0 0 CD ( N
i-i CO CO CO ( M
^ lO O Tfl " *
IM IN
00
1-1 1-1
r H OS
C^ m o
00 t>.
a
0 0 - ^ CO
CO CO CD
i-t
^ lO 00 8 2§
0 3 1-1 0 3
CO m CO
CO
a CO 1 > CO (N O CO 0 5 t > i > i n
B5 O CO CO CO CO <N I N r-l CO
IN
O CO
IN
00 o
00
00 •<*< o 00 t~- --I Oi CO I N 1-1 b - CO

O —1 CO CO XO ^H CO l > CO I N C<1 CO 00
lO t> t^ » 0 T)( O CO • * m 1-1 1 - 1 - ^ 00
% o cd N e-H o o O o o o O O i n
05 • ^ 00 CD 0 3 1 ^ CO 0 0 CD CO 05 CO
M< CO ^ O CO CO I N U 5 CO 00 i n •* CO
TJH CO CO i-l CO C ^ I N ^ O 1-1 i n i n •<ii
« IM ( N IM (N IN (N O I-I 1-1 1 * O
O TH r-1 »-( o o O o o o o o m
•*
J3
II
J "B CJ 1 3 o5 a
r~t
T3 o
I-* I-H N iJH
d
•* i n
N
^
00
CO H
The indicated summations are summarized in Table B. Introduction of these

values gives
I2.OOO0 + 12.1786 + 7.531c + 9.129/ = 69.660 (f)
9.1290 + 7.2536 + 5.442c + 13.960/ = 57.558 (g)
12.178a + 21.6826 + 5.588c + 7.253/ = 80.566 (h)
7.53I0 + 5.5886 + 9.125c + 5.442/ = 44.700 (i)
. By the elimination of a three equations result for further evaluation by determinants.
0.22036 + 0.0315c - 0.7685/ = -0.6001 (j)
-0.98606 + 0.1372c + 0.9336/ = -0.3112 (k)
1.0385b - 0.7527c - 0.1269/ = 0.6808 (1)
Solving the last three equations by determinants gives
6 = 1.526
c = 1.010
/ = 1.129
By substitution in any of the foregoing four-term equations, o = 2.764.
Similar calculations were made for all 17 mechanisms and for all experimental data
at the other two temperature levels. The resultant values of a, 6, c, and/ are summar-
ized in Tables D, E, and F. From inspection of Table D for runs at 200°C it may be
observed that all mechanisms except d and h may be rejected, because "certain con-
stants are negative or are not zero when demanded. For example, mechanism r,
corresponding to the uncatalyzed reaction, is unsatisfactory for four reasons, since a is
negative, and 6, c, and / are not zero. Rejection of this mechanism would be expected
from the very function of the catalyst. The acceptable mechanism d corresponds to
the reaction between activatedly adsorbed molecular hydrogen and codimer to give
the activatedly adsorbed product; mechanism h is similar except that atoniic hydro-
gen is involved. The experimental fit is better with mechanism d. Mechanism d
seems more logical than h, since no dissociation of molecular hydrogen is required by
the hydrogenation reaction. The acceptable equation for the rate of hydrogenation
at 200°C is therefore
JPHPI
g = JESES = 2.764 + 1.526pu -1- l.OlOps -|- 1.129pH (m)
The accuracy of this equation for any particular experiniental run is established by
comparing calculated with experimental values by the standard deviation squared,
5\ where S is the difference between calculated and experimental 1 values of R. The'
percentage deviation of individual runs is given as 1003 /R where R is the experimental
value. These values are recorded in the last three columns of Table C. It may be
observed that the average percentage deviation for individual runs is ±8.4 per cent;
the experimental accuracy of individual runs is no better than this because of the small
changes in composition obtained in the small catalyst beds. Runs'lla and 12ab show
large deviations. Greater degrees of .conversion can be obtained by using larger beds,
but this introduces even greater errors due to uncertainties in average values of
temperatures and partial pressures. ,
Evaluation of Constants in the Rate Equations. The experimental values of a, 6, c,
and/for the accepted mechanism d are recorded in Table G for th^ three temperature
levels and plotted in Fig. 188a on log constant-reciproqal temperature co-ordinates.
CHAP. XIX] HYDROGENATION OF CODIMER 953
Since the several constants a, b, c, and/ are each products and quotients of adsorption-
equilibrium constants and one reaction velocity constant as shown in Equation (d)
of Table B, they would be expected to plot as nearly straight lines on log constant-
reciprocal temperature plots. In this respect it may be observed that the values of a
and b plot well, / plots fairly well, but c plots poorly. The best straight lines on this
logarithmic reciprocal plot were established by the method of least squares for each
constant. The corrected values obtained from the straight lines are recorded in
Table G for the three temperature levels.
TABLE D
CONSTANTS FOR VARIOUS MECHANISMS AT 2 0 0 ° C
Mech-
a 6 c / Reasons for Rejection
anism
a 64.6 - 1 9 . 4 12.1 16.3 f should be 0 6 should be +

h 44.0 35.4 14.8 -22.6 b should be 0 f should be +
c -101 • 44.0 117 24.3 c should be 0 a should be +
d 2.76 1.53 1.01 1.13 Acceptable
e 6.64 - 1 . 3 1 0.89 2.73 / should be 0 b should be +
f 66.9 36.5 15.6 -61.8 b should be 0 / should be +
9 -115 42.4 114 45.0 c should be 0 a should be +
h 0.736 0.781 1.29 1.09 Acceptable
i 64.6 - 1 9 . 4 12.1 16.3 h should be 0
/ should be 0
J -101 44.0 117 24.3 6 should be 0 a should be +
c should be 0
k -101 44.0 117 24.3 6 should be 0 a should be +
I 5.64 -1.31 0.89 2.73 b should be 0
/ should be 0
m -116 42.4 114 45.0 & should be 0
c should be 0
n 1.74 1.50 0 95 2.46 6 should be 0
0 -101 44.0 117 24.3 / should be 0 a should be +
P -101 44.0 117 24.3 c should be 0 a should be +
/ should be 0
5 44.0 36.4 14.8 - 2 2 . 6 , & should be 0
/ should be 0
r -101 44.0 117 24.3 6 should be 0 a should be +
c should be 0
/ should be 0
From the corrected constants a, b, c, and / the equilibrium-adsorption constants

KH, KV, and Ks and the over-all surface-reaction rate constant a = Ek were obtained
from the following relations*
Ku = -; Ks = - ; KH = - and a = (n)
a a a a'KnKu
These fundamental constants are recorded in Table G for the three temperatures.
The adsorption-equilibrium constants can be expressed in terms of temperature by

the following general relationship:
1 ^ Aff AS
(o)
Provided the catalyst activity remains unaltered with change in temperature the con-
stants AH and AS correspond to the effective enthalpy and entropy changes, respec-
tively, in activated adsorption; otherwise they are empirical only.
TABLE E
CONSTANTS FOB VAEIOUS MECHANISMS AT 2 7 5 ° C
Mech-
a 6 c / Reasons for Rejection
anism
a 63.0 - 1 9 . 0 2.87 1.0 f should be 0 b should be 4-
b 31.3 41.8 8.01 - 2 1 . 0 6 should be 0 '/ should be +
c 3.98 18.9 4.45 5.97 c should be 0
d 2.84 1.60 0.494 0.687 Acceptable
e 7.12 - 1 . 6 1 0.370 0.810 / should be 0 b should be +
f 13.0 67.4 15.4 -31.7 b should be 0 / should be +
9 -311 22.2 7.65 40.8 c should be 0 a should be +
h 1.74 0.606 0.216 0.644 Acceptable
i 63.0 - 1 9 . 0 2.87 1.0 b should be 0
/ should be 0
J 3.98 18.9 4.45 5.97 6 should be 0
c should be 0
k 3.98 18.9 4.45 5.97 b should be 0
I 7.12 -1.61 0.370 0.810 6 should be 0
/ should be 0
m -311 22.2 7.65 40.8 6 should be 0
c should be 0
n 2.12 1.60 0.500 1.53 b should beO
0 3.98 18.9 4.45 5.97 / should be 0
P 3.98 18.9 4.45 5.97 c should be 0
9 / should be 0
5 31.3 41.8 8.01 -21.0 b should be 0
/ should be 0
r 3.98 18.9 4.95 5.97 ft should be 0
c should be 0
/ should be 0
The term effective is used here to distinguish the values of AH and A<S effective in
catalysis from those values obtained by separate adsorption of the pure component
gases in the absence of a catalytic reaction. The values of AH and AS for each com-
ponent are recorded in Table G.
The rate constant a is made up of the surface-reaction velocity constant k arid the
CHAP. XIX] HYDROGENATION OF CODIMER 956
effectiveness factor E where a = Ek. The constant a can be written in the usual
Arrhenius form, thus
A B
In this investigation A was found to be +1740, and B was found to be 2.82.

If the effectiveness factor were assumed to be unity, then according to the Eyring
theory (page 808), A could be taken as the enthalpy of activation in forming the acti-
TABLE F
CONSTANTS FOB VARIOUS MECHANISMS AT 3 2 5 ° C
Mech-
a h c / Reasons for Rejection
anism
a 63.0 - 2 3 . 3 2.35 11.7 f should be 0 6 should be +

b 54.0 33.8 -8.35 -10.7 b should be 0 c should be +
c 9.97 20.1 0.623 8.11 c should be 0 / should be +
d 3.75 1.60 0.039 0.697 Acceptable
e 7.01 - 1 . 6 4 0.108 1.77 / should be 0 h should be +
f 124 26.7 - 1 1 . 4 - 8 4 . 1 b should be 0 c should be +
9 2.29 20.2 0.376 17.3 c should be 0 / should be +
h 2.17 0.593 0.010 0.598 Acceptable
i 63.0 - 2 3 . 3 2.35 11.7 6 should be 0
/ should be 0
i 9.96 20.1 0.623 8.11 b should be 0
c should be 0
k 9.96 20.1 0.623 8.11 6 should be 0
I 7.01 -1.64 0.108 1.77 b should be 0
f should be 0
m 2.29 20.2 0.376 17.3 6 should be 0
c should be 0
n 3.00 1.61 0.0114 1.60 b should be 0
0 9.96 20.1 0.623 8.11 / should be 0
V 9.96 20.1 0.623 8.11 c should be 0
/ should be 0
9 53.9 33.8 -8.35 -10.7 6 should be 0 c should be +
/ should be 0
r 9.96 20.1 0.623 8.11 b should be 0
c should be 0
/ should be 0
vated complex, AH^ = 1740, and B could be taken as the entropy of activation,
A(S* = 2.82. The low value of AH* and the positive value of AS* are both consistent
with the properties of an effective catalyst such as the one under investigation. It
should be noted that if the catalyst activity does not remain constant with tempera-
ture change the adsorption values of &.H and AS for different gases become empirical
constants and do not correspond to the true enthalpies and entropies of adsorption.
~o •"
a / -
2
b
1 f •> —
0.8 — —
0.6
- _a- A
—
- 1 V C j / -
0.4 -
0.2 -
0.1 — J,
—
0.08
- / -
0.06
- -
0.04
-
0.0016 0.0018 0.0020 0.0022
1/T°K,
FIG. 188a. Experimental Values of Constants.
The final recommended equation for the vapor-phase hydrogenation of codimer by

the given catalyst is as follows:
EkKnKv [pHPu - ^]
[1 + Ktipu + Kvvv + KsPs]"

, „ , 3110 8.49
where
940 3.08
(q)
13,700 30.96
In Xs = + RT R I
, „, 1740 , 2.82
K is given by Equation (a).
From anjnspection of Equation (q) and of Fig. 189 it may be observed that an
increase in temperature has opposing effects on the reaction and adsorption constants.
The surface-reaction velocity constant k increases with rise in temperature, whereas
the adsorption-equilibrium constants decrease with rise in temperature. The net
result of these conflicting effects is that the reaction rate increases with rise in tem-
perature at a low-temperature range and decreases with rise in temperature at
a higher-temperature range and yet at temperatures far below the region at which the
reverse rate becomes appreciable. Above the optimum temperature the rate dimin-
ishes owing to desorption of reactant gases which more than offsets the accelerating
CHAP. XIX]
HYDROGENATION OF CODIMER
957
o.ooso 0.0022
i/r°K
FIG. 189. E^te-Equation Constat
nts for the Hydrogenation of Codimer.
TABLE G
SUMMABT OP T H E B M O D V N A M I C CONSTANT,
t t t
200°C 275°C
•a (experimental) 325°C
6 (experimental) 2.764 2.84 3.763
c (experimental) 1.526 1.60 1.603
/ (experimental) 1.010 0.494 0.0393
1.129 0.687 0.697
o (corrected)
6 (corrected) 2.64 3.14 3.43
c (corrected) 1.53 1.58 1.61
/ (corrected) 1.29 0.203 0.0762
1.01 0.774 0.653
a = Ek
Kv 0.644 0.830 0.954
Ka 0.580 0.603 0.469
Kn 0.489 0.0646 0.0222
KsKrj 0.383 0.246 0.190
0.222 0.123 0.0891
0.143 0.102 0.0850
-940
ASjj = - 3 . 0 8
-13,700
A^H = ASa = -30.46
-3110
A •-1740 A*SH -8.49
=
A + AHB + AHy = -2310 B = 2.82
B+ASs + ASjj = - 8 . 7 6
effect of temperature on the reaction among the molecules remaining adsorbed. This
retardation of a reaction rate with increase in temperature is characteristic of many
reactions which are catalyzed by solid surfaces. At low-temperature ranges an
increase in temperature causes an increase in reaction rate owing to increased activa-
tion, whereas at high-temperature mnges the reaction is slow due to decreased adsorp-
tion of the reactants.
The general methods denionstrated in Illustration 2 are also applica-

ble to moderately complex reacting systems in which several simultane-
ous and successive reactions may occur. It is only necessary that all
products be clearly defined compounds which can be introduced into feed
mixtures to simulate in the differential reactor conditions at various
points of an integral reactor. The rates of the individual reactions are
calculated from stoichiometric balances, and the rate equations are
evaluated from the variation of these rates with varying activities of the
components of the system.
Integral-Conversion Data. Wherever the precision of chemical
analysis permits its use, the direct measurement of differential-reaction
rates, as discussed in the preceding section, is the most satisfactory
method for studying the kinetics of a not too complex system of catalytic
reactions. However, there are many cases where the differential reactor
is not satisfactory, due either to the inaccuracy of the available analytical
methods or to the formation of complex products of side reactions which
are not readily reproduced as components in the synthetic feed to a
differential reactor. Under such circumstances it is necessary to work
with integral-conversion data where large changes in composition are
produced with low space velocities.
Differential rates can be obtained by plotting conversion x against
W/F for a series of runs on a feed of given composition processed at
constant temperature and pressure. If the diffusion of components to
and from the catalyst particles is of importance, W/F is preferably
varied by W being changed instead of F. From Equation (63) it is
evident that the slope of such a curve is equal to r, the differential
reaction rate. Thus a differential reaction rate may be determined
either by a single run in a differential reactor or by a series of runs in an
integral reactor. Once the differential rates are established, the method
of analysis is the same as that discussed in the preceding section.
As pointed out in Chapter XVIII, pages 853-4, the evaluation of the
constants in a differential-rate equation by graphically differentiating
integral-conversion curves, although theoretically sound, is generally
subject to large errors, depending on the procedure followed in drawing
the curves and estabUshing the slopes. For this reason it is necessary
CHAP. XIX] INTEGRAL-CONVERSION DATA 959
to average a number of determinations corresponding to different con-

versions in order to obtain constants which satisfactorily represent the
data. After such average values are determined, they are used to
calculate an integral-conversion curve from Equation (64) for com-
parison with the experimentally determined conversions. This inte-
gration is readily carried out by plotting from the experimental data the
activities of the various components of the system as functions of WIF.
From these curves and the rate constants being tested reaction rates r
are calculated, and a curve is plotted relating r as ordinates to WjF.
The area under this curve between zero and any value of WiF gives
the calculated conversion. The constants of the rate equation are
adjusted by successive trials until satisfactory agreement is obtained
between the experimental conversions and those calculated by in-
tegration. This method is tedious but may be appUed to very com-
plex systems.
The constants of a rate equation may be calculated by the direct appli-
cation of an integral-conversion equation of the type shown in Table
LVII if only a single reaction is involved or if the over-all results of more
than one reaction may be treated as equivalent to a single reaction.
The curves in Fig. 184 relate conversion to catalyst/feed ratio W/F, for
the catalytic production of toluene from benzene plus xylene. For com-
plete and direct evaluation of the rate equation such curves should be
available at series of different temperatures, pressures, and feed com-
positions. However, where minimum data are available and the system
is sufficiently simple to permit use of an integral-conversion equation
it may be possible to evaluate the constants of the equation from merely
three sets of curves such as Figs. 183, 184, and 185 in which pressure,
space velocity, and temperature are independently varied. This pro-
cedure was followed by Johanson and Watson, who assiuned that
Equation (69) may be used to express the rate of production of toluene
from an equimolal "feed even though some toluene results from dispro-
portionation of xylene as well as its reaction with benzene. This equa-
tion form was selected on the basis of the shape of the pressure relation-
ship shown in Fig. 183. It may be noted that at low pressures the
conversion is greatly increased by increase in pressure, whereas at high
pressures further increase has little effect. Since thermodynamic data
indicate that equilibrium is not approached at the high-pressure con-
ditions, it was concluded that this form of variation is consistent with
control of the reaction rate by a surface reaction involving dual active
centers. The rate of such a reaction is proportional to the concentra-
tion of adjacently adsorbed xylene and benzene molecules and is
expressed by Equation (69), including the fouling factor. At low pres-

sures the denominator is substantially constant at 1.0, and the rate is
proportional to the square of the pressure. At high pressures the
denominator becomes proportional to the square of the pressure and the
rate becomes independent of pressure.
The constants in Equation (69) were evaluated for a pure feed by use
of the integrated form. Equation (e) of Table LVII. Consideration was
first given to the data of Figs. 182 and 183 for short process periods where
T = 0 and the fouling factor BT is zero. As a first approximation it was
assumed that because of the chemical and physical similarity of the
reactants and products KR = {KA + KB)/2. With this assumption
/3 of Equation (e) becomes zero, leaving the following simplified equation :
C /W\ «=> , ( a x - l - 7 ) ( - l + 7 )
(88)
(ax-l + 7)(-l-7)
Equation 88 contains only C and a as unknown constants if K is taken

from the thermodynamic data of Pitzer and Scott. '^ These constants are
evaluated by the calculation of values of a / V C at various pressures from
the T = 0 curve of Fig. 183. The slope of the best straight line through
a plot of a / V C against I/TT is equal to 1/VC, and the intercept is equal
to {KA + KB)2/VC or KAB/VC.
With these two constants evaluated the assumption that KR = KAB
was verified by using Equation (88) to calculate a curve relating W/F to
X for comparison with the r = 0 curve of Fig. 184. The agreement was
satisfactorily within the probable accuracy of the data. Had the calcu-
lated curve given conversions not in agreement with the data at high
values of W/F it would have been necessary to evaluate a first approxi-
mation value of KR by applying Equation (e) to experimental data in the
high-conversion range and solving, for /3. A finite value of fi would then
require re-evaluation of KAB and C and then calculation of a second
approximation of KR.
The effects of temperature on C and KAB were determined from the
T = 0 data of Fig. 185. These constants may be expressed as functions
of temperature by the following equations involving the values already
determined at 932°F (773°K):
C = Cme «v^ 773; (89)
KAB = {KAB)me « v^ ^^s^ (90)

J» K. S. Pitzer and D. W. Scott, / . Am. Chem. Soc, 65,803 (1943).
CHAP. XIX] EMPIRICAL CONVERSION PLOTS 951
By combining (88), (89), and (90) an equation is obtained which con-

tains only two unknown constants, be and AHAB- These are evaluated
by applying the equation to a high-temperature and to a low-temperature
point on the T = 0 curve of Fig. 185. The resulting two equations are
simultaneously solved by assuming values of be and calculating the corre-
sponding values of AHAB from each equation. The intersection of the
two curves relating AHAB to h evaluates the constants.
The fouling factor terms of Equation (69) enter into both a and ^ of
Equation (e), Table LVII. Values of UT* at 932°F are obtained by esti-
mating the initial reaction rates where a; = 0 by measuring the slopes of
the curves of Fig. 184 at the origin. Where a; = 0, Equation (69)
reduces to the following:
r c
ro-— ^ = ^, (91)
: F + KAB + Mr»J
The constants u and h are readily calculated from values of ro for differ-
ent values of T. Since an unguided graphical determination of slope at
the origin is difficult, better results are obtained by calculating ro for
r = 0 from the known constants C and KAB and then assuming that the
values of ro are proportional to the values of x read from the curves
where W/F = 10. This same method is used for evaluating u at the
lowest temperature of Fig. 185, it being assumed that at the low con-
versions involved the initial rates of reaction are proportional to con-
version. It is assumed that 6 is independent of temperature and that u
is related to temperature by an equation of the form of (89).
Values of v and c are obtained by solving Equation (e) of Table LVII
for (8 at different values of r at each of two temperatures where high
conversions are obtained. In this manner all of the constants are evalu-
ated as functions of temperature as summarized in Table LVIII.
This method of evaluating a rate equation has the advantage of per-
mitting accurate determination of constants from a minimum of experi-
mental data. However, the integrated-conversion equations are almost
invariably of such complex form that the calculations are very tedious.
A high order of numerical accuracy with 6 to 8 significant figures is neces-
sary in most of this work, because the result is generally a small differ-
ence between two large logarithmic terms of opposite sign.
Empirical Conversion Plots. Simplified approximate correlations of
kinetic data which have no sound theoretical bases are dangerous and
must be used with constant consideration of the assumptions and possi-
ble errors involved, particularly in extrapolating away from the ranges
TABLE L V m
RATE-EQUATION CONSTANTS
]^ _ g(682.3/flr)+C1.539/«)
(Temperature range of 700-1050° F)
-13,000
C = 7.56 e ^^
(mole)/(piass of catalyst) (hr) (atm)^
K^=KB=KJ,= ^CS,000/flr)-(0.54/B)
(1/atm)
-16,691
u = 131 e "^
l/(atm) (min)"-^^
6 = 0.63
-23,149
V = 9820 e "^
l/(atm) (min)
c = 1.0
of experimental experience. However, in many cases such approximate
empirical methods can be effectively used with great savings in time.
A most useful basis for correlating and predicting the effects of varia-
tions in space velocity is to assume that even complex catalytic systems
approximate a pseudo-first-order relationship when only space velocity
or feed rate is varied, keeping the temperature, pressure, and the propor-
tions of reactants all constant. Consideration of the curves of Fig. 190
indicates that this assumption is at least qualitatively correct for this
system. As the amount of catalyst is increased, the conversion increases
toward equilibrium at a progressively diminishing rate, exactly as in a
simple first-order reaction where the rate of reaction is proportional to the
activity of the unconverted reactant. The conversion W/F curves of
many catalytic reactions follow this general form at any one set of con-
ditions, even though the effects of varying temperatures, pressures, and
compositions are most complex.
Assumption of pseudo-first-order behavior leads to the following rate
equation for a pure reactant A:
r. = . ( l - . . - | ) = ^ ( l - | ) (92)
where r^ = rate of reaction of A, mole/(mass of catalyst) (unit time)

k = apparent-reaction velocity constant
XA. = mole fraction of A converted
K = equilibrium constant of reaction
x\ = mole fraction of A at equilibrium = 1/(1 + 1/-K)
CHAP. XIX] EMPIRICAL CONVERSION PLOTS 963
0.60 0.50 0.40 0.30 0.20 0.10 0.00

u.ou 1 A
1 1 y '/ y
/ 1 ^7 —65
0..45 // // /V/
\ " / / / /
0.40 / / // 60
/ / //
/ / / /
0.S6 / / / AS" / 55
R
'c / / / ^ /
0.30 - 50 ..r.
//W /
^pO.25 j\ 45 1
40 u
>ci
foô -
1-1
X "^^^ —
35 O
/ / 30
0A5 25
• 0.10 20
" / 11 \ .
r 15
0.05 10
6
noo 1 1 1 1 1 1
20 40 60 80 100 120 140
W/F
FIG. 190. Linear Co-ordinates for Pseudo-First-Order Kinetics (Low Conversion
Range).

W 1 PÂ dx _ 1 / !_
(93)
It is evident from Equation (93) that where pseudo-first-order behavior

is approximated linear relationships should result from plotting W/F or
reciprocal space velocity against log 1/(1 — XA/X*), or F/W or space
velocity against;—--— --r- • A convenient method of arranging
log 1/(1 - XA/X*)
such plots on rectangular-co-ordinate paper is shown in Figs. 190 and 191.
In each plot a nonuniform ordinate scale for conversion XA/X* is estab-
lished by means of the curve relating conversion on the upper scale of
abscissas to conversion on the ordinate scale. As an example of the use

of such a plot consider a conversion of 54 per cent with a value of WjF
equal to 55, \FjW = (1.82)10-2]. The point C is plotted by entering
either chart at point A on the upper scale of abscissas, following down to
the curved line at B, and then horizontally to the proper abscissa value
»3
0.40 0.50 0.60 0.70 0.80 0.90 1.0
0.0
w' ' A1 I I I ;
98
95
1.0 — \ \ ^ ^ ^ •' ' •>" ^^^''"^ 90
80
2.0 70
65
/^ \ ^NC " 60
S.Oh \ ^ ^ \. ^>.. > 55
^ < ^ 50
,4.0 45
X >
40
- /
\\
35
iJ 6.0
\A
\ \ ^\ '- ^.
\ ~
\
30
7 . 0 -- X
\ \
\ \
\ \ V* -^25
8.0 -
9.0-
10
0.0 0.4
I . I
0.8
(F/W) X 10^
1
1.2 1.6
v,^N
2.0 2.4 2.8
FIG. 191. Linear Co-ordinates for Pseudo-First-Order Kinetics (High Conversion

Range).
at C. A truly first-order reaction plotted in this manner will yield

straight lines, the slopes of which are proportional to the reaction veloc-
ity constant. Figure 191 is useful for the high-conversion low-space-
velocity range, because all curves must extrapolate to XA/X\ = 1.0
where the space velocity is zero. Figure 190 is useful for the low-con-
version high-space-velocity range where all curves converge at zero
conversion and zero W/F or reciprocal space velocity.
These charts may be used for empirically plotting either the ratios of
CHAP. XIX] COMPLEX REACTIONS 965
conversion to equilibrium conversion or the ratfo of yield of a product to

its ultimate yield at 100 per cent conversion. If pseudo-first-order
behavior is approximated, a single experimental point serves to estab-
lish an approximate relationship between space velocity and conversion
over the entire range.
To test the applicabiUty of this approximation, points from the curves
of Fig. 184 are plotted on Figs. 190 and 191. The broken lines are
average curves through the points, whereas the straight hnes are drawn
from the extreme points to the origin on each chart. It may be noted
that the points from the curve for short process periods (T = 0) are in
excellent agreement with the first-order approximation over the entire
range. However, the points corresponding to process periods of 50 and
300 min show increasing deviation from the hnear relationships, indicat-
ing that the effective first-order-reaction velocity constant diminishes
progressively with increased conversion. In these cases considerable
error would result from assumption of a hnear relationship based on a -
single point. However, a fair approximation may be obtained from two
results, one at high and one at low space velocity. The low-space-
velocity point is used to estabhsh a line on Fig. 191 and the high-space-
velocity point to estabhsh a fine on Fig. 190. The high-range line is then
plotted on the low-range chart and the reverse, and the two hnes are
joined by smooth curves similar to the broken-line curves of Figs. 190
and 191.
This method of plotting is useful for correcting experimental results
to a common space-velocity basis for correlation and for making prelimi-
nary extrapolations of space-velocity relationships. It is also con-
venient for comparing different catalysts and estimating the different
space velocities required for a given conversion. However, it must be
emphasized that, even though such empirical methods may give excellent
agreement with observed conversion-«pace-velocity relationships, this
gives no basis for assuming that the effects of pressure, temperature, and
composition variations bear any similar relationship to first-order
behavior.
Complex Reactions. As previously mentioned the general methods
developed for analysis of data for the relatively simple systems discussed
in the preceding sections also may be apphed to complex systems involv-
ing several simultaneous reactions. This situation is common for high-
temperature catalytic reactions of organic compounds. In addition to
several catalytic reactions there are frequently important imcatalyzed
reactions which occur in the void spaces of the catalyst bed and in the
preheating and cooling sections of the reactor. It is evident that reac-
tions of such widely different mechanisms may be affected quite differ-
CATALYTIC REACTIONS [CHAP. X I X
ently by changes in operating conditions, and the resulting relationships

may be very complex. Under such conditions the inherent inaccuracy
of catalytic data may make it diiEcult to estabUsh firmly even the direc-
tion of trends of considerable commercial significance.
There are many, advantages to be gained by resolving a complex
reacting system into its component individual reactions and evaluating a
quantitative expression for the rate of each. It may be necessary to
resort to semiempirical methods such as are discussed in Chapter XVIII,
but the resulting equations will be relatively simple in form. The
individual reactions are best studied over wide ranges of conditions which
are representative of those in the complex system. By combining
individual rate equations properly evaluated in this manner it may be
possible to predict by calculation the behavior of the complex system with
greater accuracy than by direct experimental measurements over limited
ranges.
The dehydrogenation of n-butane to form butenes over a chromia-
alimiina catalyst at temperatures of the order of 1100°F is an example of
such a system. The principal reaction is the removal of 1 mole of hydro-
gen according to the following equation:
C4H10 «^ C4H8 + H2
Secondary dehydrogenation of butene to butadiene also occurs.

C4H8 ?=^ C4H6 + H2
At the high temperatures used these reactions take place to an appreci-

able extent by pyrolysis in the homogeneous phase as well as by catalysis.
In addition, the following reactions occur by both pyrolytic and catalytic
mechanisms:
(o) The dealkylation or cracking of butane to form methane, ethane,
ethylene, and propylene. ^
(b) The dealkylation of butenes to form methane, ethane, propane,
ethylene, propylene, and carbon.
(c) The dimerization of butadiene to form 4-vinyl cyclohexene-1.
(d) The decomposition of butadiene to form hydrogen, methane,
ethylene, acetylene, and carbon.
A complete kinetic analysis of this system should include measure-
ments of the rates of all of these reactions, both catalyzed and uncata-
lyzed. The uncatalyzed reactions are evaluated by replacing the cata-
lyst bed by a noncatalytic material such as fused quartz. The effects of
catalytic reactions are then segregated by subtracting from the over-all
results those which may be attributed to pyrolysis. In this maimer, by
working first with the reactions of the terminal products and then of the
CHAP. XIX] COMPLEX REACTIONS 967
intermediate products, it is possible to evaluate by difference the results

of reactions which carmot be isolated.
A preliminary study of the catalytic dehydrogenation of butane was
carried out by Dodd and Watson" who developed a simplified analysis
in which only three rate equations were used to represent the behavior
of the system. The secondary formation of butadiene and its reactions
were neglected, and it was assumed that the cracking of butane and
butenes occurred by pseudo-first-order reactions having the same veloc-
ity constants. Product distributions were assigned to these reactions
on the basis of data in the literature for uncatalyzed pyrolysis. Thus,
the three reactions considered were:
(a) C4H10 : C4H8 + H2
(b) C4H10 : O.IC4H8 + O.IH2 + 1 . 8 (dealkylation products)
(c) C4H8 i O.IH2 + 1 . 8 (dealkylation products)
Although the data were not conclusive, it was indicated that the rate of
the principal reaction (a) is controlled by a surface reaction between
adsorbed butane molecules and adjacent vacant active centers. On this
basis the following equations were developed.
,/ VRPS\
(94)
""' (1 + PAKA + PBsKns)'
where rma = rateof reaction, moles/(mass of catalyst) (hr)

C = over-all rate constant, moles/(mass of catalyst) (hr) (atm)
K = over-all gas-phase equilibrium constant, atm
KA = effective adsorption equilibrium constant of butane, 1/atm
KBS = effective average adsorption constant of hydrogen and
butenes
PBS = average partial pressure of hydrogen and butene =
(PB + ps)/2.
The constants are related to temperature {T°K) by the following equa-
tions:
18 073
^"^^^ = l575r-^-^^^^ (^^^
" R. H. Dodd and K. M. Watson, Trans. Am. Inst. Chem. Engrs., 42,263 (1946).
Also " Principles of Reactor Design," National Petroleum Publishing Company,
Cleveland (1946).
20,092
log ^«« = i j ^ - 5 - 0 2 7 3 (97)
log X = ^ ^ +6.7532 (98)

4.575r
The cracking reactions (b) and (c) were assumed to follow the following
equation:
Tnib = kpA) Tmc = k-pB (99)
where
Tmb and fmo = rates of reactions (b) and (c), respectively, moles/(mass of
catalyst) (hr)
VA, PR = partial pressures of butane and butene, respectively
k = apparent reaction velocity constant
(100)
Space Velocity=(V)/(y) (Hr)

400 200 120
40 60 80 100 120 140 160 180 200

W^ Grams of Catalyst
F G-Mole of n-Butane Fed per Hour
PIG. 192. Butene Production by the Catalytic Dehydrogenation of n-Butane.
(p = 1.0 atm; t = 1060°P.)
The results of simultaneously integrating Equations (94) and (99)

for selected isobaric-isothermal reactors are shown in Figs. 192 and 193.
These integrations were carried out by the progressive stepwise proced-
CHAP. XIX] COMPLEX REACTIONS
969
ure demonstrated in Chapter XVIII which is equally applicable to con
ditions of varying temperature and pressure. In Fig. 192 it may be
seen that the yield of butenes per mole of butane fed tends to pass
through a maximum as space velocity is reduced. This maximum
shifts mth temperature and pressure in a complex fashion. Similarly
in Fig. 193 it is seen that the selectivity of the operation follows a comph-
100
20 30 40 50 60 TO
Per Cent Conversion
Fia. 193. Selectivity in the Catalytic Dehydrogenation of n-Butane.
cated pattern with low temperatures most favorable at low conversions,
whereas the reverse is true at high conversions.
Because of the several admittedly unsound assumptions involved in
the development of Equations (94-100), it is probable that the relation-'
ships of Figs. 192 and 193 are not very accurate. However, it'is be-
lieved that the trends are qualitatively correct. It is of interest that so
complete a pattern can be developed from only a few experimental runs.
It would be extremely difficult to evaluate the trends shown in the lower
conversion range of Fig. 193 by experimentation alone.
More complete evaluations and more complex systems require more

data than were used for Equations (94-100). Such data are preferably
of the differential type, but frequently difficulties of chemical analysis
interfere, and integral conversions must be determined.
, PROBLEMS
1. The equilibrium constant for the chemisorption of a gas A without dissociation
on a catalytic surface is 10.0 with activities referred to partial pressures in
atmospheres. The number of molal adsorption sites L, corresponds to 0.5 g-mole
per kg of catalyst.
(o) Plot a curve relating the equilibrium concentration of the adsorbed gas A in
gram-moles per kilogram as a function of partial pressiu'e over the range of 0-10 atm,
assuming that no other component is adsorbed.
(6) Replot the curve of part (a) for the adsorption of gas A from a mixture con-
taining 40 mole per cent of A and 60 mole per cent of component B which has an
adsorption equilibrium constant of 5.0.
(c) Replot the curve of part (b) for the case in which the partial pressure of com-
ponent B is held constant at 2.0 atm and the partial pressure of A is varied from
0 to 10 atm.
2. Repeat problem 1 for the case in which component A is dissociated on adsorp-
tion into two JA atoms.
3. Develop derivations for the following rate equations in Chapter XIX:
a (42) b (43) c (44)
d (45) e (46) / (47)
4. Derive the following equations of Table LVII:
a (a) 6 (b) c (c)
6. From the following experimental data relating j factors for mass transfer and
modified Reynolds numbers DpG/ii, establish the best value of constants a and b in the
relationship, j = o ( 1, using the method of least squares. '
j Dj^G/n j D/J/n
0.0928 332 0.0636 970
0.0840 493 0.0538 1160
0.0690 595 0.0536 1371
0.0693 734 0.0482 1555
0.0662 809 0.0453 1655
0.0637 914 0.0476 1858
6. From the data in Table A of Illustration 2 evaluate by the method of least

squares the constants of the rate equation corresponding to mechanism d at a tempera-
ture of 275°C.
7. Toluene is produced in accordance with Equation (68) by passing an equimolal
mixture of benzene and xylenes over a catalyst at a temperature of 450°C. The
CHAP. XIX] PROBLEMS
971
catalyst has a bulk density of 45 lb per cu ft. In order to investigate the kinetics of
the reaction pilot plant runs were made at the conditions summarized in Table A
In all cases the process periods were so short that fouling of the catalyst may be
neglected. The equihbnum constant of the reaction at this temperature may be
taken a^ 3.5. Assummg that the rate equation is of the form of Equation (69):
J. = CjaAaB - a%/K) •
(1 + UAKA + GBKB + UBKR) 2 '
(a) Graphically evaluate the initial rates of toluene production in lb-moles/fhr)

(lb) at zero conversion for the various pressures of operation.
Compare the results ob^tained by measuring the slopes of the origins of the curves
relating x to W/F with those obtained by plotting a curve relating x/(IF/f) to W/F
and extropolating to W/F = 0.
(6) From the results of part (o) evaluate the constants {KA + KB) and C of
Equation (69), using the method of least squares. Activities should be referred to
unit fugacity in atmospheres.
(c) Determine the differential-reaction rates corresponding to various degrees of
conversion by plotting and graphically differentiating the data at 465 lb per sq in
Conversions expressed in moles of toluene per mole of feed should be plotted as ordi-
nates against W/F Ob) (hr) /lb-mole and values of r evaluated at abscissas of 50 75
100,125, and 150. ' '
(d) Using the values of C and {KA -f KB) determined in part (6) evaluate the
arithmetic average KR corresponding to the differential rates of part (c).
(e) Verify the value of KB determined in part (d) by plotting calculated values of
1 /r as ordinates against x as abscissas and graphically integrating to establish the
relationship between x and W/F for comparison with the experimental values
(/) Using Equation (e) of Table LVII evaluate the constant KR from the data at a
pressure of 465 lb per sq in. and space velocities of 1.0 and 0.5. Compare the average
of these values with that of part [d).
(g) Using the constants determined in parts (b) and (/) in Equation (e) of Table
LVII, plot a curve relating x to W/F for comparison with the data of Table A for a
pressure of 465 Ib/sq in.
TABLE A
Pressure, . Si, Toluene, Pressure, Si, Toltiene,
lb per sq in. 1/hr Mole Per Cent lb per sq in. 1/hr Mole Per Cent
20 0.5 4.5 315 3.0 10.7
20 0.25 8.3 315 6.0 5.6
65 1.0 12.5 465 6.0 5.9
65 2.0 6.6 465 3.0 11.0
115 2.0 8.6 465 1.0 26.5
115 4.0 4.4 465 0.5 35.0
Si = liquid hourly space velocity referred to 60°F.

8. Dodd and Watson" found the following relationship between butene production
and space velocity in the catalytic dehydrogenation of n-butane at a temperature of
lOeO^F and a pressure of 1.0 atm.:

Butene produced, • W/F
moles per mole fed (lb catalyst) Qir)/lb-mole butane
0.05 1.0
0.10 3.2
0.17 8.1
0.22 11.5
0.27 • 16
0.34 32
*
The equilibrium constant of the dehydrogenation reaction at this temperature may be
taken as 0.2 atm.
(a) Plot these data on charts of the type of Pigs. 190 and 191.
(6) Determine the error which would result in estimating the conversion where
W/F = 5 and 30, respectively, by assuming that this complex reaction follows a
conversion-space-velocity relationship of a pseudo-first-order form based on the
experimental point at 25 per cent conversion.
9. Assuming the apphcabiUty of Equations (94-100) calculate the initial selectivity
obtained at zero conversion in the dehydrogenation of n-butane to butene at the
following conditions of temperature and pressure.
(o) 3.0 atm, llOCF (d) 1.0 atm, 1150°F
(&) 1.0 atm, 1100°F (e) 1.0 atm, 1050°F
(c) 0.3 atm, 1100°F
CHAPTER X X
MASS AND H E A T ' T R A N S F E R IN CATALYTIC BEDS

The reaction rate equations developed in the preceding chapter all
involve activities or partial pressures of the reacting components at the
stirface of the catalyst. As previously pointed out, these interfacial
activities may be significantly different from the average activities of the
main fluid stream because of the gradients required for the diffusion of
reactants toward and products away from the interface. These activity
differences due to diffusion may be neglected under conditions of rela-
tively slow reaction rates combined with favorable conditions for mass
transfer such as result from high velocities of flow and from small parti-
cles. In general, the calculation of heterogeneous reaction rates should
include evaluation of the diffusional gradients of the system.
A further complication in heterogeneous catalytic rate calculations is
introduced by heat transfer. Since the actual reaction occurs on the
surface, the heat of reaction is absorbed or released at this point, and
temperature differences exist between the reaction surface and the fluid.
Where thermal effects are large, the control of temperature by suitable
heat-transfer arrangements becomes the most important problem of
reactor design.
Since the transfer of mass and heat are closely related phenomena of
great importance in heterogeneous reactions, these principles are
developed concurrently in considerable detail.
Laminar and Turbulent Flow. It has been experimentally demon-
strated that when a fluid flows at low velocities over the surface of a solid
or the interface of another immiscible fluid, the flow tends to be laminar.
At the interface the velocity relative to the other phase is substantially
zero while at increasing distances from the interface the velocity pro-
gressively increases. In effect the fluid may be considered as composed
of thin layers which follow the contours of the surface without mixing
with the adjacent layers except as mingling results from molecular diffu-
sion. Even in laminar motion the flow is not necessarily linear but
follows the contours of the restraining surfaces as though a bundle of silk
filaments or ribbons were being pulled over the surface, each maintaining
its same relative position, regardless of the irregularity of the path. In
fluids flowing around bends in pipes laminar flow may take a spiral path
not conforming to the shape of the bend. Laminar flow is also referred
to as viscous or streamline.
973
974 MASS AND HEAT TRANSFER [CHAP. XX
If the velocity of laminar flow is increased, a condition is finally

reached at which mixing of the adjacent layers of fluid begins and cross-
^ w currents and eddies are established in the moving stream. The
velocity at which such cross currents and turbulence are est-iblished is
termed the critical velocity marking the transition from laminar to what
is termed turbulent flow. For isothermal flow in a path of uniform cross
section such as in a pipe the critical velocity is sharply defined, whereas
in &0W through irregular passages as in a bed of granular material the
transition from laminar to turbulent ôvf occurs gradually over a range
of velocities. Also in nonisothermal ôw when heat is being transferred
through a fluid stream a gradual transition from laminar to turbulent
flow results because of mixing induced by thermal convection.
The Film Concept. Even under conditions of turbulent flow there is
always a layer of fluid at the interface in which laminar flow is main-
tained. The thickness of this laminar layer or film depends upon the
conditions of flow, and becomes less as the velocity is increased. At
sufficiently low velocities laminar flow may include the entire stream.
The term film as used influidflowshould not be confused with adsorbed
films which may be only a single molecule in thickness.
For engineering problems deaUng with mass and heat transfer in
turbulent fluid streams it has proved convenient to assume that condi-
tions of concentration and temperature are uniform in the main body of
the stream and that the differences between the average conditions of the
stream and those at the interface all occur across an effective boundary
film. Actually small fractions of the total gradients of concentration and
temperature occur in the turbulent zone so that the effective film thick-
ness represents an arbitrary extension of the actual laminar layer to
include the entire gradient. For this reason the effective thickness of
thefilmmay be slightly different for heat transfer than for mass transfer.
On the basis of this^Zm concept mass transfer is considered as resulting
from molecular diffusion across the effective film under the influence of a
concentration gradient. Similarly, heat transfer is considered as taking
place by thermal conduction through thefilmas a result of a difference in
temperature'between the interface and the average temperature of the
fluid stream.
More detailed general discussions offluidflow,heat transfer, and mass
transfer are to be found in the standard texts on unit operations, absorp-
tion,' and heat transmission.^
' T. K. Sherwood, " Absorption and Extraction," McGraw-Hill Book Company,
New York (1937). ^
' W.H.McAdams, "Heat Transmission," McGraw-Hill Book Company, New York
(1942).
CHAP. XX] DIFFUSION g^g
DIFFUSION
If two gases A and B at the same pressure are initially separated into
two adjoining compartments, removal of the partition will initiate trans-
fer of component A into B and B into A by diffusion. The theory of
diffusion in gases is based chiefly on kinetic theory and was developed
principally by MaxwelP and Stefan.^ A summarized discussion is pre-
sented by Sherwood/ and a more extensive development is given by
Jeans.^
According to the classical kinetic theory of gases the diffusion of com-
ponent A in a gaseous mixture of components A and B results from a
driving force which is equal to the partial pressure or concentration
gradient —dpA/dL which exists at any instant in the direction of diffu-
sion. This driving force is expended in overcoming a resistance to
diffusion which is proportional to the product of the concentration of the
gases. Mathematically,
TT = dABCACBiUA — UB) = UAB -TJ- {UA — UB) (1)
where PA = partial pressure of A

L = distance in the direction of diffusion of 4
ciAB = a proportionahty factor
CAJ CB = molal concentrations of A and B, respectively
UA, UB = linear velocities of diffusion of A and B in the direction of
diffusion of A
Wherie the ideal-gas law is applicable, Equation (1) may be written
dpA VAPB , - ,„.
also
TAJJT
UA = rAaVA = (3)
PA
where rô = the molal rate of diffusion of 4 , in moles per unit time per
unit of cross-sectional area of diffusional path.
VA = molal volume of component A.
Substituting (3) in (2)
dpA ClAB , .
--dL=W ^''^' - ''^'^ ^^^
' J. C. Maxwell, Set. Papers, 2, p. 5, Cambridge University Press (1890).
* J. Stefan, Siisber. Akad. Wiss. Wien, 63 (2), 63 (1871); 65 (2), 323 (1872).
' J. H. Jeans, " Dynamical Theory of Gases," 3d edition, Chapter XIII, Cam-
bridge Umversity Press (1921).
In the counterdiffusion of two components it follows that the two rates

must be equal and opposite since the average molecular concentration
and total pressure must be unchanged. Thus^
TAa = -TBa (5)
. VA -^VB=^ T^ (6)
Combining (5) and (6) with (4) and rearranging gives
^r dVA .-.
rAa = Tf- (7)
OiABir aiJ
or, since PA = CART,
{RTy dCA ,_,

l-Aa = -TT- (8)
«ABT ah
Equation (8) forms the basis for the definition of the coefficient of diffu-
sion DAB of components A and B.
dCA DAB dpA ...
r„.-r..--D„ — = - _ — (9)
OiABTT-
Equation (9) is a fundamental equation for the diffusion of gases but is

directly applicable only to a binary mixture of A and B in which equi-
molal counterdiffusion exists. It is evident that the diffusion coefficient
DAB is a physical constant having the dimensions of (length) V (time)
which is characteristic of the pair of components A and B.
Unidirectional Diffusion in Binary Gas Mixtures, If a mixture of
components A and B at constant pressure is brought into contact with an
interface at which component A is removed from the gaseous phase but
component B is not, diffusion of A will result and a concentration of B
will be maintained in a stagnant film. Since diffusion of A requires a
concentration gradient of ^ a similar gradient must exist for component
B in order to meet the requirement of a uniform total pressure. Thus,
dCA/dL = -dCm/dL.
Since, in the case under consideration, no component B is removed
from or supplied to the interface under steady-state conditions it follows
that the net'Tnte of movement of B is zero with respect to the fixed point
in space represented by the interface. However, the two components
must maintain the diffusion velocities relative to each other which are
called for by Equation (1). If velocities are expressed relative to the
CHAP. XX] DIFFUSION IN COMPLEX SYSTEMS 977
fixed interface, both UB and TB become zero, and Equation (4) becomes
dp A dps aABPBrAa RT
-dL^lL^-RT-^Dl^P^'^" (11)
Rearrangingând integrating between the hmits Li and L2 gives
DABTT . p£2 _ DABvjPBi — Psi)
'''"' " RTiL, - L^) ^"^ pBi ~ RTBaVf ^ ^
where Ba = effective film thickness = L^ — Li
Pf = the logarithmic mean of PB2 and PBI which is defined as
PB2 - PBI , . ._^
Pf = ; — r - = KPB2 — PBi)im (13)
-CT:)
Since pB2 — PBI = PA\ — PA2, Equation (12) may also be written
Equation (14) is the fundamental form for unidirectional diffusion of

component A through a stagnant film of component B.
Diffusion in Complex Systems. In heterogeneous chemical reactions
two or more components may diffuse to and from an interface in the
presence or absence of a film of inert gas. The transfer may be equi-
molal in both directions or may correspond to an increase or decrease in
the number of moles.
Gilliland^ has developed a rigorous treatment of the simultaneous
diffusion of two gases in the presence of a stagnant film which leads to
complex equations for even this relatively simple case. However, useful
approximations are obtained by simple relationships which are based on
the assumption that, in a complex system of diffusing gases, the diffu-
sional gradient established for any component A is equal to the sum of
the gradients which would result from the separate diffusion of A with
each of the other components in separate binary systems in which the
concentrations and rate are the same as the complex system.
For the general gaseous reaction, aA + 5B ?^ riJ + sS proceeding in
contact with a solid catalyst in the presence of an inert gas I, Equation
(1) may be expanded on this basis to include the effect of all components
present. Thus, for the diffusion of .4,
~ = aAsCACB{UA — UB) + aABCACB{UA — UB)

dh
+ <XASCACS{UA — Us)+ aAiCACi(uA) (15)
or, in the form of Equation (4) combined with (10),
~ ^ ^ " J T = T T " V ^"P^ ~ ruaPA) + -p— {rA.aPR - TRaPA)

HI aJ-i -UAB JJAR
• + 7 — {rAaPs - rsaPA}-^- f^- TAOPI. (16)

UAS ÂI
The terms ui and ri do not appear since there is no net diffusion of inert
gas.
Because of the uncertainty of the proportionaUty factors a and the
corresponding diffusivities in mixtures of several components and
because of the comphcations resulting in attempting to apply different
values for each binary pair, an average diffusivity DAm will be used to
replace the separate values recorded in Equation (16). The average
value of the diffusion coefficient will be taken as the weighted mean of
the values for each pair. Thus, for component A,
(1 - NA) DAm = NBDAB + NBDAR + NSDAS + NIDAI (17)
where NA, NB • • • = average mole fractions of components A, B • • •

in the diffusional film
From the stoichiometry of the general reaction,
TBa = f-jÂa; TRa = " ( " j râj Tsa = - f - j TAa (18)
Substituting Equation (18) in (16) and replacing each diffusion coeffi-

cient with the average value DAm gives
TT dpA 1 r , (^ — '• "~ s ) l ^„v

-RTdL^K'J''^'"'^'''-^'''-^'"'''' a J ^'')
or, since :r = p^ + PB + PB + ps + pi,
dpA
TAadL — —
RT (a -\-h — r — i\
RT V + PASA) ^ ^
where S^ = ^
a
Integrating between the film boundaries L and Li, corresponding to an
CHAP. XX] DIFFUSION IN COMPLEX SYSTEMS 979
effective film thickness Bg gives
T / 1 \ /ff + PASA \ .-,,

TAa =
RTBQ
Equation (21) also may be written in the form of Equation (14):
•DxmT
where
(x + SAPA) - (TT + SAPAJ) ,„„,
p/ = — . (23)
Equation (23) may be recognized as the logarithmic mean value of

(w + SAPA) over the boundary limits of the gas film. Thus, for diffusion
oiA,
P/ = h + SAPA^M ' (24)
Similarly, for S
P/ = [IT - SsP.lim (25)
r + s--- — a—&•••
where 8s =
Where the ratio of (ir + 5^pA)/(ir + SAPAI) is small, for example, less
than 1.2, the arithmetic mean is in close agreement with the logarithmic
mean and may be used for most purposes.
For equimolal diffusion it may be recognized from Equation (24) that
Pf = TT and Equation (22) reduces to the form of Equation (9). For the
diffusion of only one component in the presence of a stagnant gas 5^ in
Equation (23) is equal to — 1, and the value of pf becomes that given by
Equation (13).
Equations (22) and (23) also can be applied where multicomponent
diffusion takes place without stoichiometric relations. For example, in
the countercurrent adsorption of one gas and desorption of another the
constants a and s may be taken as the actual molal rates of adsorption
and desorption, respectively, and the ratio (s — a)/a may assume any
value. Equation (15) indicates that for a given concentration gradient
the rate of diffusion of A is favored by the presence of another component
diffusing in the same direction and retarded by another component diffus-
ing in the opposite direction. Approximate relative values of a and s
may be obtained in such a case by calculating the rate at which each
component would diffuse under the conditions of the system if the other
were absent. This procedure is demonstrated in the following illus-

tration:
Illustration 1. Sherwood' presents an illustration in which ammonia is diffusing
from an air-ammonia mixture into water under a total pressure of 0.2 atm. The
stagnant gas layer is assumed to be at 55°C. At a given point in the apparatus the
partial pressure of water and ammonia in the main gas stream are respectively zero
and 0.006 atm. The partial pressure of ammonia in the solution is considered as zero.
The vapor pressure of water at 55°C is 0.0727 atm. The diffusion coefficients are
given as follows:
DAB, ammonia in air = 1.075 cm''/sec
DsB, water vapor in air = 1.245 cm''/sec
DAS, ammonia in water vapor = 1.47 cm^/sec
It is desired to calculate relative values of (a) the rate of diffusion of ammonia,
neglecting the effect of the diffusion of water; (6) the rate of diffusion of water,
neglecting the effect of the diffusion of ammonia; (c) the approximate rates of diffu-
sion of water and ammonia, in the presence of each other.
(a) If the air and water are considered as a stagnant component B, the diffusion of
ammonia is expressed by Equation (14) where p/ = (0.2 — 0.194) /In (0.2 /0.194) =
0.197. Thus,
TAaRTBa = (1.075) (0.2/0.197) (0.006) = 0.00655
(6) If the air and ammonia are considered as a stagnant component, the diffusion
of water is similarly expressed by Equation (14) with jjy = (0.2 —0.1273)/In
(0.2/0.1273) = 0.1609.
TsaRTBa = (1.245) (0.2)/(0.1609) (-0.0727) = -0.1125
(c) The arithmetic mean composition of the gas film in mole fractions is: NHs =
0.015; H2O = 0.182; air = 0.803. The average diffusion coefficients of ammonia,
DAm, and water, Dsm, are calculated from Equation (17) using this mean composition:
(0.803) (1.075) + (0.182) (1.47) , ,^^
^ ^ = (1-0.015) = '-^'^
(0.803) (1.245) + (0.015) (1.47) ^ ^^„
^'- = (1-0.182) = ^-^^
The values of p/ are obtained from Equations (24) and (25) by using for a and s the
values of TAH and —rsa calculated for the individual diffusion of the components in
binary systems.
For anamonia: VA = 0.006; pxi = 0
(a - s)/o = (0.00655 - 0.11251)/(0.00655) = -16.18
IT - VA{p, - s)/a = 0.2 + (0.006) (16.18) = 0.2971
p/ = (0.2971 - 0.2)/2.3026 log (0.2971/0.2) = 0.2454
For water: vî = 0.0727; PA = 0
(s - a)ls = (0.11251 - 0.00655)/0.11251 = 0.94178
T - PAi{.s - a)/s = 0.2 - (0.0727) (0.94178) = 0.13153
Pf = (0.2 - 0.13153)/2.3026 log (0.2/0.13153) = 0.1634
The rates of diffusion of the two components in the presence of each other are obtained
by substitution in Equation (22):
CHAP. XX] DIFFUSION IN THE LIQUID STATE 981
For ammonia:
TAaRTBa = (1.148) (0.2/0.2454) (0.006) = 0.00561
For water:
rsJtTBa = (1.249) (0.2/0.1634) (-0.0727) = -0.1111
By use of the more rigorous equations developed by Gilliland, Sherwood calculated

values of 0.0057 and —0.1116 for these rates. The results of neglecting the effect of
the diffusion of one gas in calculating the rate of diffusion of the other are summarized
as follows for this illustration:
Considering Negkding Percentage
Counterdiffusion Cwnterdiffusion Error
TAaRTBa (ammonia) ' 0.00561 0.00655 +16.7
rsaffirBo (water) -0.1111 -0.1125 +1.3
Diffusion in the Liquid State. The rate of molecular diffusion of two

components A and B in the liquid state is given by Fick's law Vfhicli is
analogous to Equation (9) for gases;
TAa = -DAB - ^ (26)
where
rAa = rateof diffusion of A, moles/(area) (time)
CA = concentration in moles per unit volume
DAB = diffusion coefficient
Since CA = —
where
XA = mole fraction
Vm = average molal volume of solution
DAB dxA ,„„^
rAa= -TT- • (27)
Vm dL
Equation (27) may be integrated over a film thickness BL to give

DAB , \ ,„„,
TAO. == —^ (XA - aiAi) (28)
For the general case of diffusion in a complex system undergoing the
reaction aA + 65 <=± r72 + s8, a development analogous to that used for
Equation (22) leads to
TAa = - p"* (a;^ - XAi) (29)
where Xs is the logarithmic mean value of the term, (1 + S^rr^) in the

liquid film. For equimolal counterdiffusion; x/ = 1.0, and for the case
of the unidirectional diffusion of A in a stagnant medium x/ is the loga-
rithmic mean of the values of (1 — CCA) at the film boundaries.
TRANSFER OF HEAT
The equations for the transfer of heat in fluid streams are similar in
general form to those for diffusion and involve a parallel concept of a
driving force, in this case a temperature difference, causing flow against a
resistance concentrated in a fluid film. The heat-transfer problem is
simplified by the fact that it may always be treated as a unidirectional
single phenomenon. The basic equation of heat transfer may be written
in the form of Equation (9).
.ga= -kdt/dL (30)
or, in integrated form,
Qo = -J 7 - = ^~{ti- ti) (31)
where qa = heat-transfer rate per unit area

Bha — effective film thickness for heat transfer
k = thermal conductivity Btu/(hr)(ft)(°F)
TRANSFER COEFFICIENTS, FACTORS, AND UNITS

The equations developed in the preceding section for diffusion and heat
transfer are limited in their usefulness, because the value of the effective
film thickness B is generally not known. For this reason it is convenient
to group this term with the constants of the equations to form what are
termed transfer coefficients. Thus, for the general case of mass transfer,
Equation (22) may be written as
TAa = ko(pA - PAi) (32)
where fcg = — = the mass-transfer coefficient of the gas film. An

RTBaPf
analogous expression may be written from Equation (29) in terms of fci,
the mass-transfer coefficient of the hquid film.
Similarly, for heat transfer, Equation (31) may be written as
Qa = Kiti - ii) (33)
where hg = fc/Bse = the heat-transfer coefficient of the gas film.
CHAP. XX] TRANSFER COEFFICIENTS 983
For the calculation of mass and heat transfer it is necessary that the
transfer coefficients kg, fc^, hg, and hj, be empirically correlated in terms
of the determining variables. These variables include the physical
properties and conditions of the system which determine the diffusivities
and thermal conductivity as well as the conditions of flow which deter-
mine the effective film thickness. In order to facilitate the plotting of
experimental data for widely varying systems and conditions Colburn^
and Chilton^ combined dimensionless groups of the variables to define
what they termed a mass-transfer factor jd and a heat-transfer factor jh-
where
Hmn • '^-
kg = mass-transfer coefficient of the gas film
Mm = mean molecular weight of the gas stream
Pf = film pressure factor, defined by Equation (23)
G = mass velocity, mass/(area) (time)
H = viscosity of gas
..^ p = density of gas
^"^^^--J^^m = average diffusivity of component A
(jj,/pDAm)f = "Qie diinensionless Schmidt number
hg = heat-tr\nsfer coefficient of the gas film
Cp = specific neat of the gas
k = theririal conductivity of the gas
(Cp/x/k) = the dimensionless Prandtl number
Subscript / = properties at average condition of the film
Equations (34) and (35) are also directly appUcable to Hquid films.
The mass- and heat-transfer factors have the advantage of correlating
as unique functions of the dimensionless Reynolds number (DG/ju) in a
form analogous to that of the correlation of the friction factor of the
Fanning equation for pressure drop in fluid flow. Thus,
Jd = <!> ( - ^ j ; j^ = <^' (^ (36)

where
(f) and (j)' denote unique functions for a particular mechanical
system of flow
= A. P. Colbum, Trans. Am. Inst. Chem. Engrs., 29,174 (1933).
; T. H. Chilton and A. P. Colbum, Ind. Eng. Chem., 26,1183 (1935).
D = effective diameter of the path of fluid flow

G = mass velocity
n = viscosity of the fluid
{DG/ix) = Reynolds number at the average conditions of the fluid
stream
Chilton and Colburn pointejl out that <j) and <^' are numerically similar
and that as an approximation for turbulent flow in a circular cylinder
jd may be taken as equal to jh and to / / 2 where / is the friction factor of
the Fanning equation. These analogies lead to serious errors in other
cases, and correlations based on measurements of the separate phenom-
ena of interest are required.
Although the J transfer factors are most convenient for the correlation
of experimental data they are somewhat cumbersome in application to
process design and analysis problems. The basic transfer coefiicients
kg and hg are more convenient to use but have the disadvantage of vary-
ing over wide ranges with changes in mass velocity G, the film pressure
factor pf, and the specific heat Cp. For convenience of application in
packed towers and granular beds Chilton and Colbm-n^ developed a
method of expressing the mass- and heat-transfer characteristics of a
system in terms of the height of a transfer unit Ha or ff/,. As its name
implies, the height of a transfer unit has the dimension of length and is a
measure of the height of packing or bed required for a specified transfer
service. Accordingly it must vary inversely with the corresponding
transfer coeSicient. The physical significance of the transfer unit is
developed on page 991. By definition,
ri
Hdo = ;—— (37)
koMrnP/av
Sw = ^ (38)
where
Hig = height of gas-film mass-transfer unit
H},g = height of gas-film heat-transfer unit
a„ = interfacial area per unit volume of packing or bed
Similar definitions may be applied to Uquid films.
By combining (34) with (37) and (35) with (38); .
1/ u \t G
JdSplJAm.// KgMmVt
8 T. H. Chilton and A. P. Colburn. Ind. Eng. Chem., 27, 255 (1935).

CHAP. XX] TRANSFER COEFFICIENTS 935
„ l(C,^\^ GC,
Rates of heat and mass transfer in gases flowing through beds of uni-
formly sized granular solids were investigated by Gamson, Thodos, and
Hougen' and by Wilke and Hougen.^" These investigators correlated
heat and mass-transfer factors jh and jd against a modified Reynolds
number in which the diameter term is taken as the diameter of a sphere
which has the same surface area as an average particle in the bed.
Thus, ^
Dp = Vap/ir (41)
where
Up = average surface area per particle
For cylinders of height and diameter each equal to D, Equation (41)
reduces to
Dp = DVh5 • (42)
For spherical particles of diameter D or for cylinders whose height and

diameter are each equal to D the surface area per unit volume is repre-
sented by the following equation:
6p, 6(1 - F.)
where
Pb = bulk density of catalyst bed, mass/volume
Pp = actual density of catalyst particle, mass/volume
Fe = fraction of external void volume in catalyst bed
On this basis, Gamson, Thodos, and Hougen developed the following
expression for j ^ and jd in the range of Reynolds numbers above 350:
jd = 0.99(Z>p(?/M)-o-"] (44)
\ (DpG/fi) > 350
yA = 1.06(DpG/M)^-«J (45)
Wilke and Hougen recommended the following equations for Reynolds
numbers less than 350.
ja = 1.82(I>P(?/M)^-«0 (46)
\ (J)^G/n) < 350
J , =1.95(Dp(?/M)^-5iJ (47)
' B. W. Gamson, G. Thodos, and 0 . A. Hougen, Trans. Am. Inst. Chem. Engrs.,
39, 1(1943).
1° C. R. Wilke and O. A. Hougen, Trans. Am. Inst. Chem. Engrs., 41, 445 (1945).
where
G = mass velocity of flow based on the total cross-sectional area
of the bed
Dp = effective diameter of particle from Equation (41)
No significance is attached to the value of D^G/ix at 350. The mass-
and heat-transfer factors form a smooth and continuous curve over the
entire range of values, but for convenience the relationship may be
expressed on logarithmic paper by two straight lines which intersect at
350.
For convenience in use, the relationships expressed by Equations (45)
and (46) are combined with Equation (39) and plotted in Fig. 194 which
shows values of aâa as a function of the modified Reynolds and the
Schmidt numbers. Values of {aMha/i-07) may also be read as ordinates
from the chart by substituting the Prandtl number for the Schmidt
number.
Equations (44-47) and Fig. 194 are based on data for the vaporization
of water from completely wetted spherical and cyhndrical particles of
various materials ranging in diameter from 0.09 to 0.74 in. It is be-
Ueved, however, that they may be applied to general problems of mass
and heat transfer in gaseous systems at widely varying conditions if
reliable values for the various physical constants are available.
No similar data are as yet available for heat and mass transfer in
liquids flowing through granular beds. It seems reasonable to expect
that relationships of the form of Equations (44) to (47) should represent
this case, and in lieu of direct data it may be assumed that these equa-
tions are applicable to the liquid state.
. For the calculation of diffusional pressure and temperature differences,
from transfer unit data, Equations (39) and (40) may be combined with
(32) and (33). Thus, if ApA = PAt — PA, it follows that,
ApA. = - ! - = " TT = Tn— (.(^"Hjo) (48)
Similarly, if At = U — t, it follows that

At = -Qm/haam = ^ „ ^ iflvHm) (49)
fliCrCp
where
TA = rate of reaction or mass transfer of A per imit mass of catalyst
qm = rate of heat transfer per unit mass of catalyst
Om = particle area per imit mass of Qatalyst
a„ = particle area per unit volume of catalyst bed = CUPB
AHA — heat of reaction per mole of A reacted
CHAP. XX] TRANSFER COEFFICIENTS
987
t-w^m ^^55 \ •b jr
i t 4 X y i J >5 \
\
3 :;i' • -J7^
- ^^î
^3-4l t ^
— ^<A\
s \ . . sCO ^Q|k ^.u»;^^
I H .-^ rm. K O^ '
- %
1 _ o
- w T\\\\ -J' 02
-
\aV
* ^ ^V'
i W^-^
\ ô- 1. ° ^
VÔ
-t"^
^ -ti
jt 3
7j
- \\
L \
\
\\\V Ok
P« ,
tI i a
S
-
7 v\- ^) O
\ \ ^
^^_ 7
\ '\
A ^ •*
rl\ k;—
^
»l
-nlZ/i 3
^_ \ \ '^^ ^ c ^ J . -
t\ c\ 2
\^;\* A
/. \
_ / _.:: \ ^ L_ \Y^ f|||5^§2§§i|-
< \ \ Âo\ (
^ 3 I V JA 5|lb§§HI||:
C »\ " 1 ru
T V V \w ? ^ ^ J
^ ^^ ll^îssisil- >
01
G
?
'y
"^ a.
^^
\vN\
!t
\
iKiliil: -a
o
\^ \ s^ " : S a
3 V ^ > \ !5V 03
1 ^v^^N
t ^ o^ 3
0 r-l ^- o li t - A ^ S \ !V- -
Q V SS A
a \>V \
^v
'/
\
\ A \
' \\ ^ \ \
•
'V L VVV
•B
13
1 1 03
o '^V 1 1 \ \^\ ^
// \ V \
-v/ ' V ji:

3
Iw.- t2
// ^^ \ \^'^
^^ f:x^~\S
1 f^ 1
. tA ^ V ^\
•
\ ^ V - t A ^ ^'\
1 ^fe
JX ÂA ^^
^s^s_S \^; ^
o ___ IL 00 CO 1IN
C
1"^ O 00
° "=
U> •* 25 O
o o
o c5 c5 o o o o
rm O 00 C0 c> =5
8 •»t CO
MASS AND HEAT TRANSFER [CHAP. XX
PHYSICAL CONSTANTS
In order to use Fig. 194 or Equations (45-47) for the evaluation of
transfer data it is necessary tliat the physical constants p, Cp, n, DAm,
and k be known for the system. Generalized methods for estimating
the first three constants in the absence of experimental data are dis-
cussed in Chapters XII, XVII, and XVIII, respectively.
Diffusion Coefficient. The diffusion coefficient DAB for the inter-
diffusion of two gases, A and B, may be estimated from the following
empirical relationship developed by GilUland:''
3 /
D AB T' / 1 1 (50)
0.0043 —; r - M r r + TT
DAB = diffusion coefficient,Tîvl+viyyMA
cmVsec. MB
T = temperature, degrees Kelvin
MA, MB = molecular weights of A and B
TT = total pressure, atmospheres
VA, VB = molecular volumes of A and B which may be taken from
Table LIX for simple gases or estimated as the sum of the
indicated atomic volumes for complex molecules
TABLE L I X
ATOMIC VOLUMES FOB COMPLEX MOLECULES
Bromine 27.0 Oxygen in:

Carbon 14.8 Aldehydes and ketones 7.4
Chlorine 24.6 Methyl esters 9.1
Hydrogen 3.7 Methyl ethers 9.9
Iodine 37.0 Higher esters and ethers 11.0
Nitrogen, double-bonded 15.6 Acids 12.0
In primary amines 10.5 Sulfur 25.6
In secondary amines 12.0
In aromatic compounds deduct:
For benzene ring 15
For naphthalene ring 30
Molecular volumes of simple gases r'^
H, 14.3 NHa 25.8
0, 25.6 H2O 18 9
N, 31.2 H2S 32.9
Air 29.9 COS 51.5
CO 30.7 CI2 48.4
COj 34.0 . Brj 63.2
SOj 44.8 Is 71 5
NO 23.6
NjO 36.4
" E. R. Gilliland, Ind. Eng. Chem., 26, 681 (1934).
" J . H. Arnold, Ind. Eng. Chem., 22, 1091 (1930).
f
Ciap. XX] THERMAL CONDUCTIVITY ggQ
Illustration 2. Calculate the diffusion coefBoient of gaseous benzene in air at
80°C and 1 atm. From Equation (50) and Table LIX:
For benzene, v = 6(14.8) + 6(3.7) - 15 = 96.
(353.1)2 I 1
D = 0.0043 ^—-—r-.. \-^ + 7;R 29
1.0(96i+29.9*)2\78 29
= 0J051 sq cm per sec or (0.1051) (3.87) = 0.407 sq ft per hr
Experimentally determined diffusion coefficients for many systems of

both gases and liquids, are included in the " International Critical
Tables," Volume V. In general the coefBcients for gases range from
0.05 to 10 sq cm per sec while those for liquids are much lower, generally
falling in the range of 0.3 to (5) 10^^ sq cm per sec. An empirical method
for the estimation of diffusion coefficients of liquids in the absence of
experimental data was developed by Arnold.'^
Thermal Conductivity. Few reliable data are available for the thermal
conductivity k of fluids. The thermal conductivities of Uquids are in
general relatively little affected by temperature and fair approximations
for many substances can be arrived at by study of tables^ of experimental
values for similar materials.
In heat-transfer equations for gases the thermal conductivity appears
only as a part of the dimensionless Prandtl number {Cpn/h). For ideal
gases this number shows httle variation with temperature, pressure, or
type of gas. Prandtl numbers of various gases are given in Table LX
for a temperature of 212°F and 1 atm. The effects of temperature and
pressure are uncertain and in general may be neglected except in the
critical region.
TABLE' LX
PBANDTJJ NUMBEBS OP GASES AT 2 1 2 ^ AND 1 ATM
k \k J
Airj CO, Hj, Ni ,,02 0.74 0.818
Na 0.78 0.848
CO2, SO2 0.80 0.862
Ethylene 0.83 0.883
H28 0.77 0.840
Methane 0.79 0.855
H2O 0.78 0.848
Taken from " Heat Transmission," by W. H. McAdams, McGraw-Hill Book

Company (1942), New York (with permission).
" J. H. Arnold, / . Am. Chem. Soc, 52, 3937 (1930).

McAdams^ recommends the following equation based upon the work of

Maxwell and Eucken for estimating the value of the Prandtl number
PROCESS RATES CONTROLLED BY DIFFUSION

(
Many precises in heterogeneous systems proceed at rates which are
almost entirely determined by rates of mass and heat transfer through a
single laminar film at the interface. This situation results when all
reactions occurring at the interface proceed at such high rates that con-
ditions of equilibrium may be assumed to exist. Heterogeneous reacting
systems tend to approach this condition at high temperatures where the
rates of chemical combination become very high without proportional
increases in the rates of mass transfer which are httle affected by temper-
ature. In catalytic operations an analysis based on the assumption that
the rate is controlled by mass transfer is frequently of interest as indi-
cative of the minimum reactor size that may result from the ultimate in
catalyst activity.
A useful general equation for such a process in which interfacial
partial pressures may be assumed constant or may be calculated from
equilibrium relationships may be derived by consideration of the general
reaction aA -\-hB ^ rR •{• sS occurring in a gas stream flowing through
a granular bed at substantially constant temperature and total pressure.
In an elementary section of bed of thickness dZ and unit cross-sectional
area the quantity of A transferred to the interface in a unit time must
equal the quantity lost from the gas stream if steady-state conditions
prevail. Expressing this material balance mathematically for ideal gas
behavior, . „ .
koAdvipA - pAi)dZ = - d l - ^ ) (52)
Before integration can be carried out it is necessary to express the molal

rate of flow G/Mm in terms of PA. The change in number of moles per
mole of A reacting is expressed by 5A = (r + s — a — b)/a. Then, in
terms of XA, the moles of A reacting per mole of feed,
Also, since XA = TIAO — nA and PA = Wx5r/(1 + ÂSA),
^^^B^_P^^±^^ (64)
TT IT
CHAP. X X J PROCESS RATES CONTROLLED BY DIFFUSION 991
where the subscript o denotes the properties of the feed. Combining (53)
and (54) and rearranging gives
G _ G /TT + PAO8A\ ^ / g \ / I + PAo5A/Tr\
Mm " Mmo \^ + VASA ) \MJ \l + VA5A/^ ) ^ '
Substituting (55) in (52) yields

(?(1 + VAO5AM
MmoT^kaAttv (PA - VAij \ 1 + VA-Â/Tt/
GO- + PAOSAM dp A
(56)
MmoTrkaAO-v (PA - PAi^{l + VAÂHY
Replacing G/Mmo by its value from Equation (55) and introducing the
film pressure factor pj in both the numerator and denominator results in
a z = - - ^ — - ^ (57)
MJcoAPfav {TT + PA5A){PA — PAO
Equation (57) cannot be directly integrated for the general case but
may be handled by an approximate graphical integration and averaging
procedure which is satisfactory for many purposes. Thus, an inte-
grated form may be written as follows:
Z,-Z.= -(--^—) r\ _, ^^f (58).

XMrnKGAPfar/a.yg'^PA, (^ + PAÂjipA ~ PAO
The term inside the parenthesis of Equation (58) may be recognized as
the height of a transfer unit Hda defined by Equation (37). Although
the terms Mm, koA, and pf in this group all vary as functions of PA or Z
in a given reactor and properly should be included in the integral, the
variation of the entire term is generally not large, and the arithmetic
average of the terminal values at Zi and Zi may be used as a good
approximation.
The integral of Equation (58) may be evaluated graphically by plotting
PA against p//(7r + PAÂ){PA — PAI)- In simple cases where pAi is
constant and ps may be expressed as the arithmetic mean, analytical inte-
gration is possible. The value of this integral is a measure of the diffi-
culty or duty of the transfer operation which is determined by the partial
pressure change produced and the driving force causing it. Chilton and
Colburn^ have termed this integral the number of transfer units required
for the operation. Where variations in the ratio p//(7r + PAÂ) are
small, it is evident that a single transfer unit produces a partial pressure
992 MASS AND HEAT TRANSFER XX
change in the gas stream which is equal to the average driving force.
With this concept Equation (58) may be written as /
Zi — Zi — {Hda)a.ve^da (59)
where ^ T -.J)
(Jfdo)avg = average height of transfer unit

N'da = number of transfer units
P/dpA
PA, (J + PA5A){PA PAi)
Ai VfdyA
(60)
'lÂi (1 + yA5A)(yA ~ VAi)
where yA = mole fraction of A
Vf = Vsl'^ = log mean of (1 + j/^S^) and (1 + yAiÂ)
Similar expressions may be derived for the liquid state.
By an analogous development equations for heat transfer may be
obtained in the general forms of Equations (58-60). Thus,
z._z..-(2£.) r-^=fl^„ (61)
For the case of equimolal counterdiffusion, where 5 = 0, and pj

Equation (58) reduces to
7 7 - m \ ^'A^_dpA_^ •dpA
rÂ,i
(62)
•'PA, PA — PAi '^PA, ApA
'•
This situation is approximated in fractional distillation columns.
Evaluation of the integral of Equation (60) for the general case where
the reverse reaction is of importance is complicated by uncertainty as to
the value of pAi- When systems are dealt with in which the molecular
weights of the reactants and products are not widely different and where
Pf is not greatly different from :r, the following approximate relationship
may be assumed to exist between the partial pressure differences of the
various components.
(PA< - PA) = ApA = Apfi f - j = -Apij f - j = - A p s f - j (63)
Then, if equilibrium exists at the interface,

l/a
(p«-Apîy(p,-Ap.i)
ApA = -PA (64)
iT^PB + A p A ^ '
CHAP. XX] PROCESS RATES CONTROLLED BY DIFFUSION 993
In the general case Equation (64) may be solved graphically for ApA
to correspond to selected values of PA, and the corresponding values of
PB, PB, and ps which are determined by the stoichiometry of the system.
The errors involved in use of Equation (64) are generally negligible if
K is large but may become serious where gases of widely different charac-
teristics are involved. In such cases the more rigorous stepwise inte-
gration discussed in the following section should be used.
Where Pf or yf do not vary greatly they may be assumed to be constant
at the arithmetic mean of the terminal values. Then, for cases where
yAi = 0, Equation (60) reduces to
Ar f \ rvM^dyA_^ r2Mi(l+2/425^)1 .__,

^vAi (1 + VASAjyA LyA2{i + VAISAJJ
lUusttation 3. A mixture of 38 mole per cent propylene and 62 mole per cent pro-
pane is charged to a vapor-phase catalytic polymerization reactor at a temperature of
450°F and an absolute pressure of 225 lb per sq in. The polymerization reaction is
substantially unidirectional and produces a mixture of olefins having an average
molecular weight of 105, an average boiling point of 220°F, and a specific gravity of
0.71 at 60°F. The catalyst is in the form of -j- X -j-in. cylinders with 50 per cent
external voids. Assuming that the catalytic activity is so great that the concentra-
tion of propylene is zero at the interface, calculate the space velocity (per hour)
required for 98 per cent polymerization of the propylene at a mass velocity of 200
lb/(hr)(sqft).
The effects of variations in pressure throughout the reactor may be neglected.
Solution: The reaction under consideration is represented stoichiometrically by the
equation:
2.5C3II6 <=i Gi.iH.is
t
Basis: 100 lb-moles of feed, 98 per cent conversion of propylene.

Entering Leaving
Propylene 38 moles Propylene 0.76 mole; 1.0%
Propane 62 moles Propane 62.00 moles; 79.8%
Polymer 14.90 moles; 19.2%
The temperature of the stream leaving the reactor may be calculated by the
methods of Chapter VIII. However since temperature has relatively little effect on
diffusional rates an approximate calculation is satisfactory in which the heat of reac-
tion is assumed independent of temperature and the specific heat of the reactant
stream taken as constant at that of the feed. The standard heat of reaction may be
estimated from the group contributions of Table XL, page 759, if the polymer is as-
Bimied to have the properties of an equimolal mixture of 2,3-dimethyl butene-l and
2,3,3,4-tetramethyl pentene-1.
Heats of formation k-cal per g-mole
Propylene 4-5.4
2,3-dimethyl butene-l -15.2
2,3,3,4-tetramethyl pentene-l -32.8
994 MASS AND HEAT TRANSFER [CHAP-: X X
"F
Heat of reaction, AH°k-cal per g-mole propylene = [( — 15.2 — 32.8)0.5 — (2.^)5.4]/
2.5 = -11.76 .
AH°, Btu per lb-mole propylene = -(15,000) (1.8) = -27,000
The molal heat capacity of the feed at 1000°R is calculated from Table XXI, page 336.
Propane 29.39
Propylene 24.41
Feed = (0.38)(29.39) + (0.62)(24.41) = 26.30
Approximate temperature rise = (27,000)(0.38)(0.98)/26.30 = 382°F
Approximate outlet temperature = 832 °F
From Equation (42),
Z),(ft) = (0.25/12) (1.225) =0.0225
From Equation (43),
n/u, (0-5) (6) (12) ^^.
"•^'/^*^ = (0.25) =''^
P H Y S I C A I , PBOPEBTIES OF CoMPONEliTS
Propane Propylene Polymer

Critical temperature
Tc'R 666 (Table XXVII) 658 (ICT) 984 (Eq. III-14)
Critical pressure Pc,
lb per sq in. 642 (Table XXVII) 661 (ICT) 390 (Eq. III-15)
coiKi ccper g-mole 9.18 (Table XXVII) 8.17 (Fig. 109) 19.23 (Fig. 109)
Critical viscosity fic,
micropoises
(Equation
XVIII-105) 224 230 259
Molecular weight 44 42 105
Molecular volume
(Table LIX) 74.0 66.6 166.5
PHYSICAL PROPERTIES OP MIXTURE

Entering Leaving
Pseudocritical properties
• Temperature, T'c°R 663.5 727
Pressure, p'c, lb per sq in. 650 593
Viscosity, /c, micropoises 230 233
Pseudoreduced properties
Temperature, T'r 1.371 1.78
Pressure, pr 0.35 0.380
Viscosity (Fig. 175), nj 0.63 0.80
Viscosity, micropoises 144.8 186.
Viscosity, lb/(hr)(ft) 0.0346 0.045
Compressibility factor (Fig. 103) 0.975 1.0
Molecular weight 43.3 55.8
Density, lb per cu ft 1.02 0.905
CHAP. XX] DIFFUSION IN POROUS CATALYSTS 995
PHTSICAL, PROPERTIES OP MIXTTOIE (Continued)

Entering Leaving
Diffusion coefficients, Equation (50) (sq ft per lir)
DAB (propylene-propane) 0.0392 0.0665
DAC (propylene-polymer) 0.0247 0.0419
DAm [Equation (17)] 0.0392 0.0618
Schmidt number, tt /pDAm 0.875 0.787
Reynolds number DpG/;u 147 113
ji [Equation (46)] 0.142 0.163
o.Hdff [Equation (39)1 6.4 5.3
(fl'<iG)avg, ft = (6.4 + 5.3)/(2)(144) = 0.0405
Nda is evaluated from Equation (65):
SA= (r - a)/a = (1 - 2.5)/2.5 = - 0 . 6
At the entering conditions:
(1 + VAiW - (1 + VASA) 1.0 - 0.772
, / I + VAih^ . /T 1.0 \
For the low concentration of the outlet gas the arithmetic mean is satisfactory:
(2//)2 = (1.0 + 0.994)/2 = 0.997
(^/)avg = (0.881 + 0.997)/2 = 0.939.
; , , , = 0.939 l n l # M : ^ l = 3.66
L(o.o
1.01) (0.772) J
The thickness of the bed is obtained by Equation (59):
Zj - Zi = (3.66)(0.0405) = 0.148 ft or 1.78 in.
The hourly space velocity, based on total gas at 60°F and 30 in. Hg (saturated) is
obtained by consideration of 1 sq ft of bed:
^ (200)(385.5) ^
(43.3) (0.148) '
It is evident that a perfect catalyst under these conditions would result in very high
reaction rates.
DIFFUSION IN POROUS CATALYSTS
In chemical reactions catalyzed by solid surfaces the reaction rate per
unit mass of catalyst is influenced by the size and shape of the catalyst
particle. In general an increase in gross external surface area or decrease
in particle size for given surface conditions of temperature and com-
ponent activities increases the rate. For a completely impervious
catalyst the reaction is confined to the external surface, and the rate is
hence directly proportional to the external surface area. In pprmeable
catalysts the reaction extends to the interior surfaces, and the gross
external area is generally a negligible fraction of the total effective inter-
facial area.
Commercial catalysts are manufactured in the forms of spheres, cylin-

ders, irregular granules, and hollow cylinders. In the so-called fluid
catalyst system the solid is suspended as a fine powder in a gas stream
and thereby displays maximum exterior surface and effectiveness. In a
stationary bed the pellets or granules are rarely used in sizes much less
than i in. in height and diameter because of excessive pressure drops,
manufacturing difficulties, and the fact that further reduction in size
frequently results in httle gain in effectiveness.
The availability of the interior of the pellet for catalysis depends upon
size, shape, and permeability of the pore structure. For high effective-
ness it is required that the pores and capillaries be of large and reason-
ably uniform cross section and be interconnected with the external sur-
face of the pellet. The effectiveness of the interior surface also depends
upon the rate and nature of the reaction. A given pellet is relatively
less effective for a gaseous reaction which proceeds at a high rate with an
increase in number of gaseous moles than for one proceeding at a low
rate with a decrease in moles and is also influenced by the effects of acti-
vated adsorption and the rapidity of the reverse reaction.
In considering the properties of a catalyst bed the external void space
which surrounds the pellets should be distinguished from the internal
voids within the particles. This distinction is established from the
measurements of three densities, bulk density PB expressed as mass per
unit volume of bed; particle density pp, mass per unit volume of particle:
and solid density pc, mass per unit volume of solid free from all voids,
external and internal. For many catalysts the particle density is con-
veniently determined by displacement of mercury while the solid density
is measured by hehum displacement. The external void fraction of the
bed is given by the relation,
Fe = 1 - - (66)
Pp
and the internal void fraction by
Fi = l - ^ (67)
I t is evident that in a fluid reaction catalyzed by a porous solid the

concentration of the fluid reactant at the interface will be lower at the
interior smâces than at the gross surface of the particle and that the rate
of reaction per unit interfacial area will be lower at the interior surface.
The ratio of the actual rate of reaction per unit mass of solid to the rate
which would exist if the concentrations at all interior interfaces were the
same as those at the gross exterior interface has been termed^ the effective-
CHAP. XX] DIFFUSION IN POROUS CATALYSTS - 997
ness factor of the particle. A general equation for the rate of a reaction
catalyzed by a porous solid may be written in the following form,
rA = CAai = E^JAai (68)
where
TA = reaction rate per unit mass
C = observed over-all rate factor of the reaction
AOi = driving force of the reaction in terms of activities at the external
surfaces of the particle
EA = effectiveness factor
J = the rate factor of the catalytic reaction which when multiplied by
the driving force Aa,- gives the rate of reaction per unit mass of
catalyst if the driving force is uniform throughout
For example, in Equation (XIX-36) the driving force, Aai is equal to

(uAi — asi/K) while the rate factor J is equal to (kLKA)/(l + aAiKA +
ctRiKB + aiiKj • • •)• For other mechanisms corresponding values for
Afflj and J can be obtained directly from the rate equations developed
previously.
With an effectiveness factor of 1.0 the reaction rate at all interior
surfaces is the same as that at the exterior surfaces. It is evident that
this situation is approached when (1) the particle size is small, (2) the
pores are large and well interconnected, (3) the rate factor of the reaction
is relatively low, (4) the diffusion coefficients of reactants and products
are high. The relations among these factors were developed by Thiele'*
by analyses of several simple cases which with the aid of reasonable
simplifying assumptions could be treated mathematically. Thiele
found that in the cases studied for a first-order reaction in which the rate
is proportional to the first power of the concentration at the interface the
effectiveness factor is a function of a modulus which is defined as follows:
"42Wci)J
s^=/(mr)=/hr\-7r (69)
where
niT = Thiele modulus = -^•x ~^

2 \cDv
Dp = effective particle diameter for Thiele modulus
c = average radius of pores in the particle
Dv = diffusion coefficient
" E. W. Thiele, Ind. Eng. Chem., 31, 916 (1939).
MASS AND HEAT TRANSFER
Rigorous expansions of this equation involving complex rate equatigns

are difficult to develop and cumbersome to use.
It is evident from Equation (69) that Thiele's modulus is dependent
upon the diffusional and reaction characteristics of the system, the struc-
ture of the solid, and the particle size. Both the reaction velocity con-
stant and Dv are functions of temperature. The effective pore radius c is
difficult to evaluate but for pa'rticles formed by the pelleting or extrusion
of a given catalyst should be approximately proportional to the square
root of the fraction of internal void volume. On this basis the modulus
may be expressed in the following empirical form,
Wr = (70)
where a' and h' are empirical constants, characteristic of the system
Fi ~ fraction internal voids from Equation (67)
0.5 0.81.0 2.0 3.0 4.0 6.0 8.010

Modulus, m-
Pia. 195. Effectiveness Factor for Catalyst Pellets.
Thiele found that for the several simple cases studied the effectiveness
factor is related to the modulus by curves which do not differ greatly in
general form. As an empirical approximation it was suggested that the
curve derived by Thiele for spherical particles be assumed applicable in
all cases and that D'J, be taken as the diameter of a sphere having the same
surface area per unit volume as the particle in question. This curve is
plotted in Fig. 195. I t will be noted that at low values of the modulus
J)
CHAP. XX] DIFFUSION IN POROUS CATALYSTS 999
the effectiveness factor approaches 1.0 indicating that the entire internal
surface is reacting at a uniform rate. At high values of the modulus the
effectiveness factor approaches inverse proportionaUty to the first power
of the modulus. In this range the internal area is ineffective, and the
reaction occurs only on the gross exterior surface of the particle. At low-
values of the modulus the reaction rate per unit mass of solid is inde-
pendent of particle size, whereas in the high modulus range the rate is
inversely proportional to the particle diameter.
The effective particle size D^ for the Thiele modulus must be clearly
distinguished from the effective particle size Dp used for mass-transfer
correlations and defined by Equation (41). By definition,
^.^G7,^6.(l-F.)^_^ (71)
fflp a I, ClmPp
where Vp = average volume per particle

Cj, = average gross exterior area per particle
Fe = fraction external void volume
a» = surface area of particles per unit volume of bed
dm = surface area per unit mass of particles
Pp = density of particles, mass per volume
For a cylinder whose diameter D is equal to its length. Dp = D.

If Fig. 195 and Equation (70) are accepted for empirical application
the constants of Equation (70) may be evaluated for a specific catalyst
by comparing the reaction rates produced by two catalyst beds operated
at identical conditions of temperature, pressure, and composition, but
composed of particles having different known values of Dp/(Fi)*. The
ratio of the two rates is equal to the ratio of the effectiveness factors,
EA\/EAi of the two different pellets and the ratio of their moduli,
mri/mra, is equal to D'jii(Fi2r/Dp2{Fa)^. It is evident from inspection
of Fig. 195 that evaluation of both EAI/EA2 and the corresponding
nvn/viTi serves to evaluate all four terms. On the logarithmic scales of
this figure the ratio rriTi/mTi corresponds to a fixed distance on the scale
of abscissas, and the ratio EAI/EA2 corresponds to a fixed distance on the
ordinate scale. By constructing a right triangle whose base is equal to
log (wn/mrz) and whose altitude is log EA\/EAI and moving the triangle
along the curve with its hypothenuse as a chord, the properties of the
two catalysts would correspond to the position at which the base of the
triangle is parallel to the abscissa scale. This calculation is simpUfied
by Fig. 196 which is derived from the mathematical equation of the
curve of Fig. 195. This chart permits direct reading of m n from experi-
1000 MASS AND HEAT TRANSFER [CHAP. X X \
mentally observed ratios of mn/mTi and EAi/EAi at a constant tempera-'

ture. In order to evaluate the constant' b' of Equation (70) the entire
procedure is repeated at another temperature.
100
0.0 0.1 0.2 0.4 0.5 0.6 0.7 0.8 0.9 1.0
FIG. 196. Determination of the Effectiveness-Factor Modulus.

niustration 4. In order to evaluate the effectiveness factor of a nickel hydro-
genation catalyst in the operation discussed in Illustration XIX-2 the following data
were taken in a differential reactor. At each teniperature the average activities of
the reactants and products were adjusted to identical values, for the two runs on the
different catalysts.
Catalyst Size No. 1 2
Pellet diameter, Dp, ft 0.0304 0.00425
Pellet density, pp, lb per cu ft 123 124
Solid density, pc, lb per cu ft 231 231
Temperature, °C 200
r, lb-moles/(lb) (hr) 0.0364 0.0710
Temperature, °C 250
r, lb-moles/(lb) (hr) 0.0458 0.0765
(o) Calculate the effectiveness factor and modulus of each catalyst size at the two
temperatures.
(b) Evaluate the constants of Equation (70) for this catalyst.
CHAP. X X ] DIFFUSION IN POROUS CAT'ALYSTS 1001
(c) Calculate the effectiveness factor of this catalyst in the form X h" cylin.
ders with 40 per cent internal voids a t a, temperature of 275°C.
Solution:
Catalyst Size No. 1 2
Pellet area, sq in. 0.1338 0.00820
Pellet, volume, cu in. 0.00811 0.000695
D'J,, Equation (71), ft: 0.0304 0.00425
Fi, Equation (67) 0.466 0.463
(^.O' 0.826 0.825
D'„/(Fi)^ 0.0368 0.00514
Temperature, °G 200 250
EAI/EA2 = n/ra 0.513 0.606
niTilmTi =
1 0.139 0.139
iP'vi){Fi,)^
rriTi (Fig. 196) 4.8 3.7
rriTi 0.669 0.615
EAI (Fig. 195) 0.50 0.60
EA^ (Fig. 195) 0.96 0.98
(6) Substituting in Equation (70) for catalyst size 1 gives
log 4.8 = log a' + log 0.0368 +

(2.303) (473)
V
log 3.7 = log o' + log 0.0368 +
(2.303) (523)
a' = 8.57; 6' = 1290
(c) For 1 " X 5 " cylinders:
Dp 0.0418
D ; = 5:^ = 0.0418 ft = 0.053
^/¥i -vôiio
Fi = 0.40
1290
mr. Equation (70) = 8.57(0.053)eS*8 = 4.71
EA, Fig. 195, = 0.50
It is frequently experimentally difficult to duplicate exactly the con-

ditions of operation in a differential reactor when itiaking comparisons of
two different catalyst sizes by the procedure outlined in Illustration 4.
Close duplication is desirable because of the irregularities generally
encountered in kinetic relationships, but the effects of small variations in
compositions or activities may be eliminated if a rate equation is avail-
able which is apphcable to one size of catalyst. Thus, from Equa-
tion (68), where variations in / are negligible,
EAI ^ ri(Aai)2
(72)
EAi r2(Aai)i
1002 MASS AND HEAT TRANSFER ' [CHAP. XX
In evaluating effectiveness factors from integral conversion data it is

necessary to evaluate the space velocities which wiU result in the same
conversion with different catalyst sizes when operated at the same feed
composition, temperature, and pressure. The ratios of these space
velocities at a constant conversion are then the ratios of the correspond-
ing effectiveness factors. It is generally necessary to operate each
catalyst size at two or more'space velocities and determine the space
velocity required for a given conversion by graphical interpolation.
It should be recognized that the effectiveness-factor relationship
expressed by Fig. 195 is based on the assumption that the reaction
approximates pseudo-first-order behavior over the limited range of con-
ditions under consideration. This approximation may be far from
correct where strong adsorption effects result in rates which pass through
maxima with increases in partial pressures or temperatures. In such
cases extrapolation of effectiveness-factor results is dangerous except for
systems whose composition, pressure, or temperature are reasonably close
to the range at which the empirical evaluation of the effectiveness factor
was carried out. I t is to be hoped that a more rigorous method of
handUng this problem may be developed.
Where consideration is restricted to a single type of catalyst of fixed
particle size and density the effectiveness factor may be entirely ignored
in developing rate equations. In such cases the variation of the effec-
tiveness factor with temperature is included in the over-all rate constant
C. However, an effectiveness factor far from unity tends to confuse the
significance of the empirically determined effective adsorption-equilib-
rium constants in the rate equation and may result in poor agreement
between observed rates and the simple forms of rate equations.
LONGITUDINAL DIFFUSION
In all of the developments of Chapters XIX and XX for flow reactors
it is assumed that the longitudinal diffusion of reactants and products in
the direction of flow is negligible. Diffusion of this type results from the
concentration gradients established by the conversion of the reactants.
Thus, products tend to diffuse back against the stream while reactants
tend to diffuse forward. The relative importance of these effects de-
pends on the magnitudes of the concentration gradients which in turn
are determined by the depth of the catalyst bed and by the velocity of the
fluid stream. In a reactor operated at a fixed space velocity the effects
of longitudinal diffusion are negligible with a deep catalyst bed having a
small cross-sectional area, but they may be of considerable importance
in a shallow bed of large cross section. The situation is similar to that
CHAP. XX] LONGITUDINAL DIFFUSION 1003
discussed in Chapter XVIII, page 865, for homogeneous reactions except
that in fixed catalyst beds there is not the opportunity for convective
circulation and large-scale eddy currents that exist in a vessel with no
obstructions.
In the absence of convective circulation in the stream it may be con-
sidered that in a fixed catalyst bed operating adiabatically the net flow
of all components at all points occurs in the direction of the main flow
after steady-state conditions are reached. Thus, although a product
may tend to flow back against the stream, no such Aow actually occurs
with respect to a fixed point in space under steady conditions of opera-
tion. The distribution of reactants and products must be such that the
resulting concentration gradients will conform to the requirement of a
uniform flow in one direction and at the same time satisfy Equation (1)
for the diffusion of reactants and products relative to each other.
The mathematical requirements to be satisfied by the steady-state
concentration gradients may be established by consideration of a down-
flow reactor and unit cross-sectional area to which the mass feed rate is F.
At any cross-sectional plane at a distance Z from the inlet the net down-
ward rate of massflowmust also equal F since there is no accumulation
in the reactor. The net rate of flow of any individual reactant or prod-
uct at position Z is then equal to the rate at which this material enters
with the original feed plus the rate at which it is produced in the reactor
section from the inlet to Z. Considering the reactant A, this net rate of
flow past position Z is the sum of the rate at which A is carried downward
by the movement of the stream and the rate at which it tends to diffuse
downward in accordance with Equation (1) or (20). Thus, the net
downward flow of A at position Z is equal to
RTiv + PAÂjnt dZ Jo
where
F = feed rate, mass per unit time, per unit cross-sectional area
nA,nAo = moles of A per unit mass of feed at section Z and at the inlet,
respectively
Tit = total number of moles per unit mass of feed
DAm = mean diffusion coefficient of A in the stream at position Z
SA = {r + s — a — b)/a .
TA = rateof reaction of A, moles/(mass of catalyst) (time)
PB = bulk density of catalyst bed
A second relationship may be established by means of a material

balance for component A in an elementary volume dZ.
Net flow of A into element = Fn^ — ^ ^ . • f —- —— (74)

Net flow of A out of element =
F(nA + dn^ - rT-(3^ + 3^^^) (^5)
The difference between (74) and (75) is the rate of reaction of A in the
element.
i2r(x + pA8A)nt \ dZ^ I
Integration of Equation (76) is difficult because it is a second-order

equation which requires two sets of boundary conditions to define a
unique solution. These boundary conditions are fixed by the inlet and
outlet conditions of the reactor at which discontinuities may be con-
sidered to exist both as regards reaction and diffusion. Progressive step-,
wise integration is not possible because conditions in even the first ele-
ment of volume at the reactor inlet are dependent on the gradients
throughout the entire bed. This becomes evident by consideration of
the extreme case in which diffusion is so relatively rapid that the gradi-
ents throughout the bed are negligible except at the inlet. In this situa-
tion the conditions at the outlet may be assumed'to prevail throughout
the entire bed, as was done in Illustration 11 of Chapter XVIII, page 862.
The general problem of longitudinal diffusion has been analyzed by
Hulburt'^ who developed integrated expressions for simplified cases.
Even where these expressions are applicable their usefulness is limited by
uncertainty as to the proper value of the diffusion coefficient. It is
evident that the effective or apparent coefficient for diffusion in a stream
flowing through a granular bed is quite different from that determined
from static conditions in the absence of a granular soUd.
In general, the mass velocities in commercial scale reactors with fixed
catalyst beds are such that longitudinal diffusion is neghgible. In the
corresponding pilot-plant operations it may be a serious factor which
should be carefully considered. Since the effects are difficult to analyze
or to translate to large-scale operations it is desirable if possible to design
the pilot plant .so that they are neghgible. This requires a catalyst bed
sufficiently deep to minimize the gradients dn^ldZ. Fortunately the
problem is not serious in an experimental differential reactor in which
only small incremental conversions are produced. Even though large
concentration gradients are established in such a reactor their effect is
•» H. M. Hulburt, Ini. Eng. Chem., 36,1012 (1944); 37,1063 (1945).

CHAP. X X ] F L U I D I Z E D CATALYST BEDS lOQS
small because the outlet composition differs but little from that at the
inlet.
For a small experimental reactor of the integral type where large over-
all conversions are produced the design may be verified by applying
Equation (73) to a small section at the inlet of the reactor.
-FAn^ = rAPBAZ - —— - — (77)

RTir + 'PAÂ)nt AZ
I t is first assumed that diffusion is negligible, and An^ is calculated by
omitting the last term of Equation (77). The last term is then evaluated
with this value of AUA And the diffusion coefficient from Equations (17)
and (50). If the last term is negligible in comparison with the second,
longitudinal diffusion may be neglected. This calculation should be
made for the reactant or product having the largest coefficient of
diffusion.
Fliiidized Catalyst Beds. In the so-called " fluid " catalytic process^^
the catalyst is carried upward as a fine powder into a large reactor vessel
where a relatively dense fluidized bed of catalyst and reactants is main-
tained in a state of turbulence. Because of the limited velocity of flow
which is permissible without excessive carry-over of catalyst the reactor
beds are generally relatively short in comparison to their diameters. As
a result of the turbulence inherent to the fluidized bed and this low ratio
of length to diameter it is probable that the composition throughout the
bed is approximately uniform, corresponding to the outlet conditions.
This situation is comparable to that which would exist in a fixed bed
operating at an extremely low mass velocity.
It is evident from the results of Illustration 11 of Chapter XVIII that
this longitudinal mixing is undesirable in that a larger reactor or catalyst
bed is required for a given conversion. A still more serious disadvantage
is that longitudinal mixing may greatly increase the extent of secondary
reactions because of the fact that the high concentration of products
which exists at the outlet also prevails throughout the entire bed. As a
result it would be expected that a fluidized catalyst bed will generally
produce a lower conversion and a poorer selectivity than a fixed bed
operated with the same weight of catalyst in the reactor per unit feed
rate. This disadvantage may be in part offset by the higher effective-
ness factor of -the finely divided catalyst. However, the fluidized bed
has the further disadvantage of a larger fraction of void volume which
increases the extent of uncatalyzed side reactions in the homogeneous
phase.
« E. V. Murphee, C. L. Brown, E. J. Gohr, E. C. Jahnig, H. Z. Martin, C. W.
Tyson, Trans. Am. Inst. Chem. Engrs., 41, 19 (1945).
The ease of regeneration of afluidizedcatalyst outweighs the inherent

disadvantages of the fluidized bed where catalyst fouling is an important
. factor, and the longitudinal mixing in the bed may be minimized by vari-
ous baffling arrangements which are imdergoing considerable current
development. Slowly moving catalyst beds of the TCC type" have flow
and diffusional characteristics similar to those offixedbeds.
If the rate equations for all of the reactions involved are known the
results of an unbaffled fluidized bed may be compared with those of a
fixed bed by first calculating the conversion and selectivity corresponding
to the operation of thefixedbed at a given space velocity. The product
distribution obtained in this calculation is used as a basis for a first
approximation to the product distribution from the fluidized bed. The
space velocity of the fluidized operation and a corrected product dis-
tribution are then calculated by assuming that the outlet conditions pre-
vail throughout the bed. On this basis the rate of each reaction per unit
mass of catalyst may be directly calculated. The procedure is repeated
until agreement is obtained between the assumed and calculated product
distributions.
PROBLEMS
1. Estimate the diffusion coefficients of the following pairs of gases at 500°C and
atmospheric pressure:
(a) CO2—CO; (6) COr-02; (c) CO2—N,
(d) Estimate the diffusion coefficient of CO2 in a gas mixture containing 15 per
cent CO2,1 per cent CO, 80 per cent N2, and 4 per cent O2, at 500°C and atmospheric
pressure.
2. Estimate the diffusion coefficients of n-butane, n-butene, and hydrogen in a
stream of n-butane, 10 per cent of which has been dehydrogenated to n-butene and
hydrogen at a temperature of 1075°F and a gauge pressure of 10 lb per sq in.
3. Estimate the value of -pf for the following reactions under the specified con-
ditions:
(a) For the transfer of CO2 in the reaction of CO2 with carbon at atmospheric
pressure with a main fluid composition of 15.3 per cent CO2, 7.7 per cent CO, and
77 per cent N2, where the partial pressure of CO2 at the interface is zero.
(b) The same as (a) for a composition 1.7 per cent CO2, 30.5 per cent CO, and
67.8 per cent N2.
(c) The same over a reactor where terminal conditions correspond to (a) and (6)
(d) In the adiabatic evaporation of water in a stream of air at atmospheric pres-
sure, 100°F and 50 per cent humidity.
4. A bed of cyhndrical clay pellets having 39 per cent void volume is dried by
through circulation of air. The surface area of the pellets per cu ft is 82.8 sq ft, and
their equivalent diameter is 0.0542 ft. The air enters the bed at 0.985 atm, 160°F,
wet bulb, 115°F, and a mass velocity of 1720 lb /(sq ft) (hr).
" R. H. Newton, G. S. Dunham, and T. P. Simpson, Trans. Am. Inst. Chem.
Engrs., 41, 215 (1945).
CHAP. XX] PROBLEMS 1007
Calculate the rate of drying at the inlet conditions expressed in lb/(hr)(cu ft) of
bed during the so-called " constant-rate " period when the particles may be con-
sidered as coated with liquid water.
6. Ammonia is produced at a rateof 7240 g per liter of catalyst per hour at 300 atm
and 380°C, in a 1: 3 molal mixture of Ns and H2 entering a bed of a granular iron
catalyst. The gas flows through the catalyst at a rate of 600 lb /(sq ft) (hr). The
effective particle diameter Dp is 0.028 ft and the external surface area of the particles is
145 sq ft per cu ft. Estimate the temperature and partial pressure of nitrogen at the
surface of the catalyst.
6. Normal butane is dehydrogenated by passing it through a catalyst bed con-
tained in vertical tubes 2.5 in. in internal diameter and 12 ft long. The depth of the
catalyst bed is 11 ft. The catalyst particles are -g-" X -f" cylinders arranged with
38 per cent void volume in Ihe bed which has a bulk density of 52 lb per cu ft. The
gaseous hourly space velocity is 1400 at standard conditions of 60°F and 29.5 in. Hg.
Neglecting the effect of secondary reactions and pyrolytic side reactions and assum-
ing the applicabihty of Equation (XIX-94-98), calculate the partial pressures and
temperature of the butane, butenes, and hydrogen at the surface of the catalyst
pellets at the point in the reactor where a conversion of 10 per cent is produced at a gas
temperature of 1075°F and a gauge pressure of 10 lb per sq in. The results of prob-
lesm 2 axe applicable to tiiis problem.
Estimate the percentage error resulting from neglect of heat^ and mass-transfer
effects under these conditions.
7. In the dryer described in problem 4 it is desired that the air shall leave the bed
during the constant rate period with a percentage saturation of 90 per cent. Assum-
ing that steady-state conditions are maintained and neglecting the heat capacity of
the bed and the pressure drop through it, calculate:
(o) The temperature of the air leaving the bed.
(6) The height of a transfer unit at the average conditions of the bed.
(c) The number of transfer units required for the specified vaporization of
water.
(d) The depth of the bed.
8. The bulk density of PB of a catalyst bed, determined by direct weighing is
0.64 g per cc. The density of the bed with the external void spaces filled with
mercury is 5.46 g per cc. When the catalyst bed is evacuated and filled with helium
it is found that 0.756 milliliter of the gas is required per cc of bed at a pressure of
1.0 atm and a temperature of 74°F. Calculate:
(o) The fraction external voids.
(6) The pellet density pp.
(c) The true density of the catalytic solid Pc.
(d) The fraction of internal void volume.
9. .Butane is dehydrogenated at a high space velocity in a short catalyst bed to
determine the effectiveness factor of the catalyst. The rates of reaction in the experi-
mental reactor with a constant mass of catalyst were as follows with different sizes of
catalyst in the form of cylinders whose heights and diameters are equal:
Diameter, ID. Internal Vmd Û050°P AU100°F

' Fraction Fi
0.125 0.51 2.68 5.52
0.250 • 0.40 1.90 3.41
1008 ' MASS AND HEAT TRANSFER [CHAP. XX
(a) Calculate the apparent modulus and effectiveness factor of each catalyst size
at each temperature.
(b) Estimate the effectiveness factor of a catalyst having a diameter of 0.080 in.
and'36 per cent internal voids at a temperature of 1100°F.
10. An experimental reactor for the hydrogenation of butene codimer is operated
isothermally at a temperature of 300°F, a pressure of 100 lb per sq in. and a liquid
hourly space velocity of 0.5 at 6Q°F. The feed is pure codimer with hydrogen in
10 per cent excess of that required for complete saturation. The depth of the cata-
lyst bed is 14 in. From Equation (77) estimate the relative importance of longi-
tudinal diffusion in this reactor.
11. Normal butane is dehydrogenated in a catalyst bed having the properties of
that described in problem 6. The operation is substantially isothermal at a tem-
perature of 1060°F and a pressure of 1.0 atm. Integration of Equations
(XIX-94-100) for these conditions as shown in Figs. 192 and 193 indicates that with
an hourly WIF of 30 a conversion of 39 per cent and a selectivity of 89 per cent is
obtained. The moles of products per mole of butane charged are as follows: Butenes,
0.346; hydrogen 0.39; dealkylation products, 0.09.
Assuming that the operation is carried out at the same temperature and pressure
in a fluidized bed of the same catalyst in which the same fraction of void volume is
maintained, calculate the W IF ratio and the product distribution corresponding to
39 per cent conversion, if Equations (XIX-94-100) are applicable.
(Actually the fraction of void volume in the fluidized bed would be considerably
greater than in the fixed bed. However, Equations (XIX-94-100) do not properly
separate uncatalyzed from catalyzed reactions and for this reason are restricted to a
substantially constant fraction of void volume.)
CHAPTER "XXI
CATALYTIC REACTOR DESIGN
The ultimate objective in the process design of a catalytic reactor is to

arrive at the arrangement which will result in the most economical con-
struction and operation. In general, the economic considerations are
most complex and involve careful balancing of the yields and product
distribution of the process against the capital costs of the plant and the
operating costs represented by labor, utilities, and maintenance require-
ments. However, in any event, the first requisite for the inteUigent
selection of a plant design is the abihty to predict the results which will
be obtained with any specific arrangement proposed for study. These
predictions must be based on the best possible interpretation of labora-
tory, pilot-plant, and any available commercial-scale data. In general,
these data are best correlated as semiempirical rate equations and effec-
tive transfer coefficients, as discussed in the preceding sections. It is
seldom practicable or possible to develop complete kinetic data prior to
the design of a plant, and the engineer is forced to improvise and extrapo-
late fragmentary data to conditions far outside the range of his past
experience. It is in such cases that sound fundamental principles are of
the greatest value in breaking down the available information into com-
ponent effects and contributions of such simplicity that their individual
trends can be estimated with fair assurance.
After working correlations between operating conditions and results
are established thedesign of the plant resolves itself into a series of system-
atically planned " case studies." A set of operating and design con-
ditions which appears attractive is selected, and the complete results of
the proposed operation are calculated in order to estabhsh the ultimate
cost of the desired product. This procedure is then repeated for another
case in which operating conditions or design factors are varied in the
direction expected to reduce costs. The optimum design leading to the
lowest costs is arrived at by consideration of the results of the entire
series of cases. It is within the scope of this treatment to consider only
the problem of predicting the conversions, product distributions, and
complete operating conditions resulting from any primary reactor specifi-
cations which are selected for study.
1009
1010 CATALYTIC REACTOR DESIGN [CHAP. XXI
GENERAL STEPWISE PROCEDURE

The fundamental catalytic rate equations developed in pages 910 to
926 are all in terms of the activities at the catalytic surface. In the case
of a gaseous system in which the components form ideal solutions, the
surface activity of a reactant is related to that in the fluid stream by the
following expression:
aAi = aA + VA ApA (1)
where OA.- = activity of A at the surface of the catalyst pellet

ttA = activity of A in the fluid stream
ApA = PAi — PA = partial pressure difference required for the
diffusion of A to the catalyst pellet, when ideal-gas
behavior is assumed
VA = fugacity coefficient of A at the temperature and pressure of
the system
Although there has been httle investigation of diffusion of gases under
nonideal conditions, it appears that the activity or fugacity gradient
should be the correct driving-force term. Equation (1) is based on the
assumption that with activities as driving forces diffusional data for ideal
gases may be used where the ideal-gas law is not applicable. This
assumption is in lieu of a more rigorous development for diffusion under
nonideal conditions. The errors involved are rarely of any significance,
and in most cases the fugacity coefficientf is taken as 1.0.
In the case where mass transfer and chemical rates are both of impor-
tance, the rate of the general catalytic reaction aA +bB ^rR + sS
must be determined by successive approximations. A rate is assiuned
as a first approximation, frequently based on the assumption that mass
transfer is negligible. On the basis of this rate the partial pressure
difference between the fluid stream and the interface is calculated for
each component by means of Equation (XX-48). Similarly, the tem-
perature difference between the stream and the catalyst particles
{ti — t) = At may be calculated from Equation (XX-49). The rela-
tionship between At and ApA is obtained by combining (XX-48 and 49)
/AHAApA\/H,o\.^/jA/AHÂpA/C,,.Yf M Y
\C^PSMJ\HÂ) \jJ\C,pfM,n/\ k J \PDAJ ^^
By means of Equations (XX-48 and 49) the temperature and the partial
pressure of each component are calculated at the surface of the cal^lyst
particle. A second approximation to the rate of reaction is then calcu-
lated on the basis of these pressures and temperature. The evaluation
CHAP. XXII GENERAL STEPWISE PROCEDURE ^Q^.
of the partial pressure and temperature differences is then repeated if

necessary in order to establish an accurate rate.
It is theoretically necessary that the algebraic sum of the partial
pressure differences of all components in a complex system shall be zero.
Where an inert gas is present it may be considered that any unbalanced
partial pressure difference between the reactants and products corre-
sponds to a partial pressure difference of the inert component. Where
an inert component is not present it may be necessary to adjust the
partial pressures to an algebraic balance in order to compensate for
inaccuracies in the approximate diffusional treatment which is used.
A complex problem' in reactor design, involving simultaneous reac-
tions, together with temperature and pressure variations may be worked
out by combining this method for evaluating differential rates with the
stepwise method of integration demonstrated in Illustration XVIII-13,
page 875. Small sections of the reactor are considered successively and
the conversion in each calculated on the basis of an average differential
rate. Thus, / \ ATIT
where
Aa;A = conversion of A moles per unit of initial feed
F = rate of feed per unit time
W = mass of catalyst
('*^)avg ~ ^^ts of conversion of A at the average conditions of the section
In each small section the partial pressure and temperature differences
must be evaluated if they are of significance in determining the average
rate. This operation is not difficult after the first two or three sections,
because the partial pressure and temperature differences are all propor-
tional to the rate of reaction and may be estimated from a projected plot
of reaction rate against W/F.
As a part of such a progressive stepwise integration changes in total
pressure and temperature in an adiabatic reactor are readily calculated
from each section as in Illustration XVIII-13. In this manner complete
temperature and pressure distribution curves are obtained. The rela-
tionships necessary for calculating pressure drops in granular beds are
presented on page 1015. The case of temperature gradients in a nonadia-
batic reactor where transverse temperature differences exist is discussed
on page 1033. In an adiabatic reactor the temperature change in the
fluid stream is calculated from the heat of reaction by the following
differential energy-balance equation
dZ ~ GC„ ^^
where Z = distance in the direction of flow

AH4 = heat of reaction per mole of A reacted
Cp = heat capacity of fluid stream per unit mass
DIRECT GRAPHICAL INTEGRATION

Where variations in totdl pressure are not important direct graphical
integration of the rate equations may be employed for predicting the
performance of either an adiabatic
reactor or one in which tempera-
ture is controlled in some known
relationship to conversion. This
procedure is illustrated diagram-
matically in Fig. 197 for an exo-
thermic reaction of the general
type A + £ ? ^ i ? + S in an adia-
batic reactor in which variations in
total pressure are small.
From the stoichiometry and
thermochemistry of the reaction
together with the specified initial
conditions, the mole fractions, ac-
tivities, and temperature of the
reactants are plotted as functions
of the conversion x^- These rela-
tionships are plotted.in the lower
section of Fig. 197. From the ac-
tivities and temperatures the cor-
responding rates of reaction r\
are calculated, with mass and
heat-transfer effects in the catalyst
bed neglected. The corrected rate
TA and true catalyst temperature
at the initial conditions are then
calculated by the methods dis-
Graphical Calculation of cussed in the preceding sections.
Reactor Size. If the effects of mass and heat
transfer are not large the entire corrected rate curve may be approxi-
mated from this calculation at initial conditions by means of the follow-
ing equation:
TA TA VAO ~ '"AO) (5)
r'Aoidr'Ao/dxAo)
CHAP. XXI] THE REACTOR UNIT CONCEPT 1013
The ratio of derivatives in Equation (5) serves to improve the ap-

proximation where owing to adsorption effect the reaction rate is little
influenced by conversion. If the mass transfer effects are large, trial-
and-error corrections should be estabhshed at several conversions.
The mass of catalyst per unit feed rate which is required for a specified
conversion XA is obtained by determining the area imder the curve relat-
ing 1/rA to XA, as indicated in the upper section of Fig. 197.
THE REACTOR UNIT CONCEPT

For simple reacting'systems at constant temperature and pressure,
analytical integration of Equation (XIX-64) may be expanded to include
mass-transfer effects. In the case of a reaction which is unimolecular in
both directions, represented hjA^R, there is frequently a range of con-
ditions over which the effects of the adsorption terms in the denominator
of Equation (XIX-36) are negligible and the rate equation may be
written as
TA^^C \vAi - - ^ j = C PA + APA -g J (6)
where Ap^, Api? = differences between partial pressures of A and U,

respectively, at the pellet surface and in the fluid stream. If the diffu-
sional characteristics of the product and reactant are similar, as is gener-
ally the case in isomerization reactions of this type, it may be assumed
that ApA and Aps are equal. Thus, from Equation (XX-48), if the mass
transfer coefficient fc^ is assumed to be the same for A and R, it fol-
lows that
Ap = typA - -Apfl = - r (7)
--(--i)-ie('4) <«)
or, solving for TA,
^ („ NBO + XA\
TA = -p = TT (9)
KACIV KACIV
Substituting (9) in (XIX-64) and integrating results in
or, if F is expressed in moles of feed per unit time, F = SG/Mmp
PsS ICT{1 + 1/K)M^B M^fc^Tra J

rNAo - NRO/K - xi(l + l/K)-\
\.NAO-NBO/K-X,{1 + 1/K)\ ^ ^
where
Z = height of catalytic reactor
S = cross-sectional area of reactor
Mmo = mean molecular weight of the feed
PB = bulk density of catalyst
G = mass velocity of feed .
It may be noted that the second term in the brackets of Equation (11)
is the height of a transfer unit for the mass transfer of A as defined by
Equation (XX-37), neglecting changes in Mm- The other term in the
brackets also has the dimension of length and is determined largely by the
catalytic rate constant C, the mass velocity G, and the pressure x. A
term of this general form has been designated by Hurt' as the height of a
catalytic unit He, while the entire term in brackets is termed the height of a
reactor unit Hn. Then by analogy to Equations (XX-58-59), times the
logarithmic term of Equation (11) represents the number of reactor units
NB required for the specified degree of conversion. Thus,
Z={H, + Ha)NR =HRNB (12)
where
He = height of a catalji;ic unit = (j/[Cx(l + l/K)MncPB]
Hd = height of a transfer unit
NB = number of reactor units = (1 + i/K) I r-
yA = mole fraction of A in the gas stream

J/A^= VR/K. = the equilibrium value of yA
' Hurt pointed out the convenience of Equations (11) and (12) both for
the correlation of rate data and the design of reactors under conditions of
constant temperature and pressure. From measurements of conversion
1D. M. Hurt, Ind. Eng. Chem., 35,522 (1943).
CHAP. XXI] PRESSURE DROP IN GRANULAR BEDS 1015
in a catalytic reactor to which these equations are applicable, the height

of the catalytic unit H^ is readily calculated if Equations (XX-32-47) are
used for the evaluation of Hd. The resultant values of HCTT/G may be
empirically plotted on a logarithmic scale against l/T, yielding an
approximately straight line.
The development and application of Equations (6-12) are rigorous
only for the assumed conditions of a reaction which is unimolecular in
both directions and in a range of conditions in which the adsorption
terms of the rate equation are neghgible. The integrated equations are
further restricted to conditions of constant temperature and pressure.
Development of analogous treatments for more complex cases leads to
cumbersome expressions.
Hurt showed that Equations (6-12) may be applied with fair accuracy
to other than unimolecular reactions if all but one reactant are present in
large excess, leading to a " pseudounimolecular behavior " at a constant
total pressure. Where the adsorption of a product is important in
retarding the rate of the reaction it was proposed that the height of the
reactor unit be corrected by a " poisoning factor," Thus, for such a case
where product R is strongly adsorbed and retards the reaction,
Hn = {l+VBknW,+Ha) (13)
where
hn = " coefficient of poisoning " by product B
TpB = partial pressure of R
The coefficient fcfl may be empirically correlated as a function of tem-

perature.
It is evident that simplified treatments involving approximate Hnear
relationships of this type may lead to large errors in some cases if they
are apphed to combinations of conditions not included in the range on
which the empirical evaluations of the heights of the catalytic units and
the coefficients are based. The assumptions involved are similar to
those on which Figs. 190 and 191 are based. For this reason considerable
discretion is necessary in their use. For general problems of reactor
design which involve important variations in temperature and total
pressure httle saving in labor is effected as compared to more rigorous
methods.
PRESSURE DROP IN GRANULAR BEDS
Extensive studies which have been carried out on the pressure drops of
fluids flowing through conduits of uniform cross section are reviewed in
detail in the standard texts on unit operations and hydraulics. These
results are correlated by the dimensionless Fanning equation:

dp 2/(?2
(14)
dZ gj)p
where
p = pressure in head of fluid
Z = length of path of flow
/ = dimensionless friction factor
G = mass velocity per unit cross-sectional area
p = density, mass per unit volume
g^ = standard gravitational constant = 32.2 ft/(sec)*
D = diameter for circular pipes = 4mA for noncircular conduits
m* = hydraulic radius = (cross-sectional area)/(perimeter)
The friction factor / is correlated as a function of the dimensionless

Reynolds number Nse which is defined as imkOf/j, or for circular pipes
DG/fi. The relationship between / and NRS varies somewhat with the
character of the surface but is approximately represented by the follow-
ing equations^ for flow in conduits:
NB. < 2100; / = 16/NB, (16)
NBe > 2100; / = 0.00307 + O-lSQiVif^ (16)

Equation (XVIII-98), page 869, is a modified form of the Fanning equa-
tion applicable in the region of Reynolds numbers above 100,000.
For flow through granular beds the Fanning equation may be modified
to incorporate terms for the effective free cross-sectional area which is
available for flow and the effective hydraulic radius of the path between
the particles. This problem was studied by Oman and Watson' who
pointed out the importance of the type of arrangement of the particles in
the bed. If the bed is formed by adding particles at so slow a rate that
each comes to rest before another falls on top of it, minimum bridging
occurs and a high bulk density results. A bed formed in this manner is
referred to as being in random dense arrangement. The opposite extreme
termed random loose arrangement is obtained by simultaneously dumping
all the particles of the bed into position. A stable arrangement results
which shows no appreciable change in density with vibration. However,
the bulk density may be as much as 20 per cent less than that of a random
2 T. B. Drew, E. C. Koo, and W. H. McAdams, Trans. Am. Inst. Chem. Engrs.,

28, 56 (1932).
^ A. O. Oman and K. M. Watson, Natl. Petroleum News, Tech. Sec. 36, B795
(November 1, 1944). Also " Process Engineering Data," National Petroleum Pub-
lishing Company, Cleveland (1944).
dense arrangement. It was found that the pressure drop through a

dense bed may be nearly three times greater than through a loose bed of
the same particles at the same superficial mass velocity of flow.
Either type of arrangement is fairly reproducible if care is used in
forming the bed. A particularly undesirable type of bed which will
produce nonuniform flow distribution or " channeling " results when
both loose and dense arrangements are present in different sections.
This situation can result if a large reactor is filled by dumping bags of
pellets through a manhole and raking the bed level. Based on measure-
ments of the pressure drop of air flowing through various types of granu-
lar packing materials and a review of the data of the hterature, Oman
and Watson proposed the following dimensionless equation for beds in
random dense arrangement:
dpD _ 2foG"'a,
rTl.7 \'-'}
dZ pg.F'B'
where
VD = pressure in a bed in dense arrangement expressed in force
(mass units) per unit area
Z = length of packed bed
JD = friction factor for dense arrangement as a function of the
modified Reynolds number A^'ijep = G/a^^
Qc = standard gravitational constant = 32.2 ft/(sec.) ^
G = mass velocity per unit of total cross-sectional area of the
bed
o, = surface area of particles per unit volume of bed
FD = fraction of void space in the bed in dense arrangement
= (1 - PB/PP)
p = density of fluid
pp = average density of solid particles
PB — apparent bulk density of the granular bed
The recommended relationship between fo and the modified Reynolds

number Nsep is approximately represented by the following two equa-
tions,
fo = 2.60(iVflJ-»-' [for values of Nnep from 10 to 150] (18)
fo = 1.23(iVfiep)-<'-" [for values of Ngep from 150 to 300] (19)
where NRCP = G/avH
It may be noted that the modified Re3Tiolds number used for pressure-
drop calculations is quite different from that found best for heat- and
mass-transfer correlations. This is due to the situation that in fluid
flow, pressure drop is affected by the configuration of void spaces whereas

the rates of mass and heat transfer are controlled principally by the
superficial area of the particles.
Correlation of pressure drops in beds in loose arrangement was less
satisfactory than for dense beds. For the general case Equation (17) is
modified to include a factoi; which is a function of the diiference between
the fraction of voids in the bed under consideration and that in a bed of
the same particles in random dense arrangement. Thus,
where
dZ Kpg.Ft' )© (20)
JL = friction factor for the loose arrangement

Fx, = fraction of voids in the loose arrangement
fo = friction factor from Equations (18) and (19) corresponding
to the Nnep calculated from the value of a„ for the loose
arrangement
The ratio /L//D is expressed as an approximate function of (Fi — FD) by
the following equation:
h/fD = 1.0 - 3.5(Fi - FD) (21)
Equations (18-21) were found to be apphcable with errors generally
less than 20 per cent to a variety of shapes of particles including spheres,
cylinders, rings, saddles, and crushed fragments. For irregular crushed
fragments the surface area per unit volume o„ was calculated by the
following expression which assumes a surface-volume relationship mid-
way between that of a cube and a tetrahedron,
' o. = 6.6iNP(l - Fe9 (22)
where
N = number of particles per unit volume
Fe = fraction of voids volume
Where the total pressure drop in a reactor is small in comparison to the
total pressure. Equations (17) and (20) may be written in incremental
form and Ap calculated directly from the average conditions. For
larger pressure drops a graphical integration is desirable because of the
variation of p with both pressure and molecular weight. Ordinarily it is
satisfactory to use an average value of the friction factor. For ideal-
gaseous systems p is given by the following expression
^ ^-P (23)
(1 + S^XA)RT
where
XA, = moles of A converted per mole of feed
8A = increase in number of moles per mole of A converted
Combining (23) with (17) and integrating gives
Vl - P^ = ^ 1 ^ r (1 + ^lxA)TdZ (24)
Equation (24) frequently must be used by successive trials. A prelimi-

nary relationship of conversion, temperature, and bed depth is calcu-
lated as indicated in Fig. 197, an approximate pressure distribution being
assumed. A corrected pressure distribution is then calculated by graphi-
cal integration of Equation (24). This revised pressure curve may then
be used for an improved relationship between bed depth and conversion.
It is sometimes desired to carry out an operation to produce a specified
conversion with a specified pressure drop through the reactor. If mass-
transfer effects are not large an approximate space velocity for the
required conversion may be obtained by an integration" of the type of
Fig. 197 with an assumed pressure distribution curve. It is then neces-
sary to calculate the bed depth Zi which corresponds to the required space
velocity and the specified pressure drop. The relation between mass
velocity and space velocity is as follows
G = S,pfZi (25)
where p/ is the density of the feed at the standard conditions for the
expression of space velocity.
Combination of Equations (23) and (25) with (17) and (18) and in-
tegration gives, for modified Reynolds numbers less than 150,
{S.Pfy-K'Rt^''''Jv {i+dAXA)T ^^'
Equation (26) corresponds to a fixed space velocity with mass veloci-

ties varying in accordance with Equation (25). It is not applicable to
calculations of the pressure drop distribution in a bed in which the mass
velocity is constant. In the use of Equation (26) approximate relation-
ships between p, XA and T may be assumed for the first integration. An
improved pressure distribution relationship may then be calculated by
Equation (24) and used as a basis for a new relationship between bed
depth and conversion by an integration of the type of Fig. 197. A few
trials of this type will establish all factors within the accuracy ordinarily
justified by the data.
Optimum Pressure Drop. The selection of an optimum pressure drop

is generally a most complicated problem'. As is pointed out in Chapter
XX, it is desirable to operate at high mass velocities in order to minimize
both the resistance to mass transfer to and from the catalyst particles
and the undesirable effects of longitudinal diffusion. Howeyer, high
mass velocities result in Jiigh pressure drops which may be undesirable
from three viewpoints: '
1. A high pressure drop may increase the rates of undesirable reactions
and reduce the selectivity corresponding to a given outlet pressure.
2. Excessive pressure drops may result in the crushing and disintegra-
tion of fragile catalysts.
3. Frequently a high pressure drop corresponds to a significant power
cost for compression.
All of these factors must be considered in arriving at the mass velocity
and pressure drop which result in the most economical over-all operation,
.OPTIMUM REACTION TEMPERATURES

Since reaction velocity constants are always increased by increased
temperature there are many cases in which reaction rates increase pro-
gressively with temperature and the maximum rate corresponds to the
highest temperature attainable. In such cases the optimum tempera-
ture of operation is determined by such factors as the physical and
mechanical properties of the catalyst and the materials of construction,
the effect of temperature on undesirable side reactions, the problem of
preheating to high temperatures without decomposition, and the prob-
lems of heat transfer which accompany high reaction rates.
Increased temperature may lead to diminished rates of reaction in two
general cases. The first is encountered in both homogeneous and
heterogeneous systems when an exothermic reaction approaches equilib-
rium conditions. Under these conditions, the decrease in the equiUb-
rium constant with increased temperature may more than offset the
increase in the reaction velocity constant so that a reduction in rate
results. The other case, which is encountered in heterogeneous catalytic
reactions, occurs when the adsorption-equilibrium constants are reduced
by temperature at a greater rate than the reaction velocity constant is
increased. For example, in Equation (XIX-42) the group of terms
kKAKs may decrease with increased temperature, even though k itself
always increases. In such a case the rate of reaction may increase at low
temperatures but decrease at high temperatures where the adsorption
terms in the denominator of the rate equation become small.
In all cases where reaction rates may be diminished by increased
CHAP. XXI] OPTIMUM REACTION TEMPERATURES 1021
temperature, a well-defined optimum temperature exists which will
result in a maximum rate of reaction at specified conditions of composi-
tion, pressure, ariS mass velocity. The minimum reactor size is obtained
when the temperature is controlled at the optimum at all points. Since
the optimum temperature generally varies as conversion proceeds, elabo-
rate cooling or heating provisions may be required to approach this
condition.
The optimum temperatures for a reaction at constant pressure and
mass velocity may be determined by using the methods described in the
preceding sections to calculate reaction rates at various temperatures and
conversions. The resultant data may be plotted as lines of constant
conversion or degree of completion with reaction rates as ordinates and
temperatures as abscissas. Such a plot is shown in Fig. 198 for the
catalytic oxidation of SO2 in the system described in Illustrations XVI-6,
8, 12, and 13, pages 716, 720, 726, and 728. This chart is restricted to
a pressure of 1.0 atm and a mass velocity of 600 lb/(hr)(sq ft). Similar
charts for other conditions may be developed by the methods demon-
strated in the following illustration.
niustratioa 1. The kinetics of the oxidation of SO2 on a platinum catalyst were
reviewed by Uyehara and Watson* and Hurt.' The general rate equation proposed
by Uyehara and Watson was subsequently modified to agree with the temperature
effects indicated by the new data of Hurt. This modified equation is as follows:
8000 , , . . . .
— ^ + 14.154
; PisOzVpich ^ J
(1 + VpioiKoi+ Piso^KsozV I
where
r = reaction rate, lb-moles SO2 oxidized/(hr) (lb catalyst)
T — degrees Kelvin at catalyst surface
PiSOa PiOsi Piso, = the respective partial pressures of the active components
at the catalyst surface, atmospheres
20,360 23.0
16,800 17.51
Kso, = e fi^ «
K = the over-all gas-phase equilibrium constant (Fig. 156)
The gas mixture specified in Illustration XVI-6, page 716, enters the reaction at a
temperature of 400°C, 1.0 atm, and a mass velocity of 600 lb /(hr) (sq ft). The cata-
lyst is in the form of | " X f" cylindrical pellets of such porosity that it may be
assumed that the effectiveness factor is 1.0 and that the foregoing rate equation is
applicable. The bed is in random dense arrangement with a void fraction of 0.35
* 0. A. Uyehara and K. M. Watson, Ind. Eng. Chem., 35,541 (1943).

1022 CATALYTIC REACTOR DESIGN [CHAP. X X I
and a bulk density of 50 lb per cu ft. At the conditions of the reactor inlet it is
desired to calculate:
(a) The rate of conversion of SO2 in lb-moles/(hr) (lb).
(6) The temperature of the catalyst pellets.
(c) The pressure drop per inch of bed depth.
0.01
I
O
M
a 0.001
1 I
a
T
0.0001
0.00001
700 800 900 1000 1200
Gas Temperature, °K
FIG. 198 Effect of Temperature and Conversion on the Rate of Catalytic Oxidation of
Sulfur Dioxide. (Composition of Entering Gas, 7.8%S02, 10.8%Oj, 81.4%Nj).
G = 6001b/(hr)(sqft)
(i) The rise in temperature of the gas stream/inch of bed depth at entrance
conditions.
Solution: As a first approximation the rate of reaction is calculated with the
assumption that the partial pressures and temperature at the interface are the same as
CHAP. XXI] OXIDATION OF SULFUR DIOXIDE 1023
those of the gas stream. At 4{)0°C,
20 360
4.576 log Ko, = - ^ ^ - 23; KQ, = 38.46; VK^ = 6.20
16,800
4.575 log Kso, = - ^ - 17.51; Kso, = 42.46
From Fig. 156, K = 930

2.3026 log C = - ^ ^ + 14.154; C = 9.663
9.663 [(0.078)V(0.108) - —^
= 0.0268
[1 + 6.20\/0.108 + 42.46(0)]"
For calculating the partial pressure and temperature differences across the gas film
the physical properties of the system must be evaluated. The properties of the pure
components are summarized in Table A. The critical viscosity of SO3 was calculated
from Equation (XVIII-102).
TABLE A
Component SO, 0, SO3 N»
M 64 32 80 28
Tc, °K 430.3 154.3 491.4 126.0
Pc atm 77.7 49.7 83.6 33.5
jic, micropoises 411 250 469 180
In order to be rigorously correct the viscosity of the mixture should correspond to

the average conditions in the film for calculating the Schmidt number and the
average conditions of the fluid stream for calculating the Reynolds number. How-
ever, the differences are ordinarily not large, and it is generally satisfactory to base the
Schmidt number on the main stream conditions. On this basis:
Mr, = (0.078) (64) + (0.108) (32) + (0.814) (28) = 31.23
r ; =^(0.078) (430.3) + (0.108) (154.3) + (0.814) (126) = 152.5
•p'c = (0.078) (77.7) + (0.108) (49.7) + (0.814) (33.5) = 38.64
li'c = (0.078) (411) + (0.108) (250) + (0.814) (180) = 205.3
T'r = 673.1/152.5 = 4.41; p'r = - ^ = 0.0259

00. o4
p!r = (Fig. 175) = 1.51
n = (1.51) (205.3) = 310 micropoises = (310) (0.000242) = 0.0746 lb/(ft)(hr)
(31.23) (273.1) ..„,.,. ,,
p= — — — = 0.0354 lb per cu ft
(359) (673.1) ^ '•
'T\XQ diffusivities of the individual pairs are calculated from Equation (XX-50).
The molecular volume of SO3 is estimated as 47.8.
(0.0043) (673.1)2 h 1
Osoi-o. = ; i r;* /;;;; + :;;; = Q.Zm sq cm per sec
[(47.8)* + (25.6)if\80 32
Similarly*
flso3-N2 = 0.362 .
£>sOs-so, = 0.244
I>so.-o. = 0.385
i^sos-Nj = 0.383
D02-N, = 0.525
The mean diffusivity of each diffusing component in the mixture is calculated
from Equation (XX-17) and converted to square feet per hour by the factor 3.88.
Dao^ = (0.108) (0.366) + (0.814) (0.362) + (0.078) (0.244)
= 0.354 cmVsec or 1.37 sq ftperhr
(0.108) (0.385) + (0.814) (0.383) „ „„„ ,,
Dso,m = ( 1 - 0 078) • " ^ ^ ^ ^ ^ ° ""^ ^•^'^ ^ *• P ^ ' ^
(0.078) (0.385) + (0.814) (0.525) „ . , , , ,
00^=-^ (1 - 0 108) ^ • =0-511cmVsecorl.98sqftperhr
(0.078) (0.383) + (0.108) (0.525) „ , „ ,,
r>N^ = -^^ _Q8i4) • = 0-463 cm= /sec or 1.80 sq ft per hr
The corresponding Schmidt numbers follow directly from the mean diffusivity of each
component together with the average viscosity and density of the mixture. The
effective diameter for the modified Reynolds number for mass and heat transfer is
calculated from Equation (XX-42) and the transfer factors from Equations
(XX-46-47).
Dp = (I) V T S = 0.458 in. or 0.0383 ft
" = ^^Sf^='''•• ^^='-''^^'^-'-'^='-^^ '

H 0.0740
JH = 2.00 (308)-•*•" = 0.1080
The average molal heat capacity 01 the gas is obtained from Table IV, page 212.
Thus at 400''C,
Cp = (0.078) (12.03) + (0.108) (7.84) + (0.814) (7.31) = 7.74
or, on a pound basis,
7.74
C„ = -±- = 0.248
31.23
From Equation (XX-51),
4
0.830
,-6^'-''^
7.74
The value oi'avHha is then calculated from Equation XX-40:
1 / C p A i • 0.830 _ _
ajlho = — I -^ 1 = = 7.70
jAkJ 0.1080
The gross exterior catalyst surface per unit volume of bed is obtained from Equa-
tion (XX-43):
6(0.65) (12) ,„, '^ -
f^^ — = 125 sq ft per cu ft
CHAP. XXI] OXIDATION OF SULFUR DIOXIDE 1025
7.70
Thenj HM = — - = 0.0616 ft
The values of arHdo and Hag for the diffusion of the individual components are calcu-
lated from Equations (XX-39) and summarized in Table B.
TABLE B
- li
OeHda Hda
pDAm \pDAmJ
SO, 1.54 1.334 13.50 0.1080
S02 1.435 1.272 12.90 0.103
02 1.062 1.041 10.50 0.084
N2 1.19 1.116 11.30 0.0905
Since the mole fractions of all the diffusing components are small the values of p/
are satisfactorily approximated by the arithmetic mean rather than the logarithmic
mean of Equation (XX-25). From the stoichiometry of the reaction,
Ssoj = - 0 . 5 So, = - 1 . 0 Sso, = - 0 . 5
In the first approximation p/soj may be taken as 1.0 and the factors for SO2 and O2
based on the main stream partial pressures. Thus
P/SO2 = 1.0 - (0.5) (0.078) = 0.961 atm
. P/02 = 1.0 - (0.108) = 0.892 atm
The partial pressure differences for each component may be calculated from Equa-
tion (XX-48) using the rate arrived at by neglect of diffusion. Thus, since G/ilf„ =
600/31.23 = 19.18,
rPBPMo,H,so,M„ (0.0268) (50) (1.0) (0.1080) „^^^^^
Apso, = ^ = j^3^ = 0.00756
rPi)?)/so.g,i3o.M-„ (0.0268) (50) (0.961) (0.103) ^^^^„

^ Apso. = - ^ = - ^^^ = -0.00693
rPBPfo,Hao,M„ (0.0268) (50) (0.892) (0.084) „^„^^
^p°» = - - — w — ^ (^)Wm = -°-^2'^
The first approximation partial pressures at the interface are then
PBo,i =0+ APS03 = 0 + (0.0076) = 0.0076
Pao,i = 0.078 + APS03 = 0.078 - 0.0069 = 0.0711
po,i = 0.108 -t- Apoj = 0.108 - 0.0026 = 0.1054
An expression for the heat of reaction at any temperature is obtained by substituting
the thermochemical data of Illustration XVI-12, page 726, in Equation (VIII-65),
page 305.
At 673.1 °K, AHso, = - 23,832 cal per g-mole
Since the heat of reaction is liberated at the interface it is transferred to the gas stream
as a result of a temperature difference which is expressed by Equation (2).
^ AHAAPAHHO ^ (-23,832)(-0.00693)(0.0616) ^ ,32=0
CJ^mPiHia (0.248) (31.23) (0.961) (0.1105) '
The second approximation to the temperature of the catalyst is then 673.1 + 13.2 =
686.3°K. The constants of the rate equation at this temperature are
C = 12.13; VK^ = 5.35; Kso, = 33.19

K = 660; AHsoi = -23,860; Cp = 7.76
A second approximation to the rate is calculated with these constants
, r„so, = 0.0312
Second-approximation partial pressure and temperature differences may be calculated
from this corrected rate. Since the values of aMdo vary but little with changes in the
temperature or composition of the system they need not be corrected. Corrected
values of the film pressure factor are:
P/so, = [1.0 + 1.0 + (0.5) (0.0076)] (I) = 1.002
VfBCH = [l-O - (0.5) (0.078) + 1.0 - 0.5 (0.0711)](^) = 0.963
p/o, = [1.0 - (1.0) (0.108) + 1.0 - (1.0) (0.1054)](i) = 0.893
Then
Apso, = 0.0088 piso, = 0.0088
Apsoj = -0.0081 pisoi = 0.0699
Apoj = -0.0031 pi02 = 0.1049
At = 15.3 Ti = 688.4°K
A third-approximation rate calculated on the basis of these interface pressures and
temperature is 0.0312 lb per mole/(hr) (lb) which gives results in agreement with the
second approximation.
The pressure drop per inch of bed depth is calculated from Equation (17).
G* 600
a,ii (125) (0.0746) = 64.3
From Equation (18),
U = 2.60(64.3)-«-3 = 1 ^ = 0.745
3.48
dp^2J^^ 2(0.745)(600)^(125) ^ 37.2 ib/(sq ft) (ft)
dZ QcpFJi' (32.2) (0.0354) (0.35)'-'(3600)2 '*• ^ -"•'"-' •
27 2
: ^ = 0.0158 lb/(sq in.) (in.)
The temperature rise at the inlet of the reactor is calculated from Equation (4):
dt > TAAHAPB -0.0312(-23,858)(50)
= 20.8°C per in.
dZ GCr, 7.74
( 6 0 0 ) — (12)
A rigorous procedure is followed in Illustration 1 to demonstrate the

principles involved. In many cases much simplification is justified. If
diffusional effects are not large and the reactants and products are of
similar physical properties an average diffusivity and height of transfer
unit may be used for all. If the equilibrium constant is large the diffu-
CHAP. XXI] OPTIMUM REACTION TEMPERATURES 1027
sional gradients of the products may be neglected. In many cases varia-
tions in the film pressure factor pf may be neglected, and in dilute systems
it may be assumed to equal the total pressure.
The result of Illustration !• represents one point on Fig. 198. Other
points corresponding to other conditions of temperature and conversion
may be calculated in the same
manner. The effects of ignor- 10 — I i 1 1—i^-' r\
10% conversion y >
ing the diffusional and tempera-
ture difference corrections in I uncorrected for mass yr
scd heat transfer •
^-
such calculations are shown

graphically in Fig. 199. The
solid^line curves in this figure
correspond to those of Fig. 198.
The broken-line curves are the
results obtained if the indicated
corrections are neglected. It
may be noted that the correc-
tions are very important where
the rates are high, but become
insignificant at low rates.
Illustration 1 may also be con-
sidered as the initial step of a
progressive integration for the 700 900
Temperature," K
design of an adiabatic reactor.
By similar calculations apphed PIG. 199. Effect of Mass and Heat-Transfer
Gradients on the Rate of Catalytic Oxidation
tr succeeding increments of of Sulfur Dioxide.
reactor volume the variation
of composition, pressure, and temperature may be predicted. If
change in total pressure is negligible the curve representing the adiabatic
operation of a reactor may be plotted on Fig. 198 from the adiabatic
.heating Hues of Fig. 157, page 721. Such a curve is plotted for the
reactor of Illustration 1. It may be noted that the rate of reaction rises
to a maximum at approximately 30 per cent conversion and falls toward
zero at approximately 80 per cent.
It is evident from Fig. 198 that if high conversions are to be reached
some means of cooling the reactor must be provided. One scheme is to
use two reactors, of which the first is adiabatic and the second is pro-
vided with cooling coils so arranged that the temperature may be con-
trolled to follow the curve of maximum rate. For example, the first
reactor with the initial conditions of Illustration 1 might be designed to
produce a conversion of 52 per cent, where the adiabatic operation curve
of Fig. 198 intersects the maximum rate curve. Controlled cooling in
the second reactor will permit operation to high conversions along the
maximum rate curve. The amount of catalyst required per unit of feed
WIF is readily calculated by graphical integration of Equation (XIX-64).
Values of 1/r read from Fig. 198 are plotted against x to correspond to the
selected scheme of temperature control.
An alternate scheme which simplifies the mechanical problems of
reactor construction and the replacement of catalyst is the use of a series
of adiabatic reactors with controlled intercoolers. For example, as indi-
cated on Fig. 198, the adiabatic operation might be continued to 70 per
cent conversion with Httle loss in rate. An intercooler might then be
used to cool to 400°C before entering a second adiabatic reactor which
would produce a conversion of 95 per cent operating along the indicated
cvuê. By using more reactors and intercoolers a closer approach to
maximum rate operation and higher conversions may be obtained.
MULTISTAGE OPERATION
There are frequently advantages to be gained in carrying out a reac-
tion in two or more reactors or " stages," In the preceding section the
use of multiple reactors simply as a means of temperature control is dis-
cussed. In such operations there is no addition or removal of reactants
or products between reactors, and the series of reactors may be con-
sidered as constituting only a single stage of operation.
True multistage operation may be of either a concurrent or counter-
current type. Concurrent multistage operation is advantageous for
polymolecular reactions in which one reactant undergoes undesirable
side reactions if it is permitted to exist in high concentrations. For
example, in the catalytic alkylation of butane'with butenes to produce
isomeric octanes the alkylation reaction is accompanied by undesirable
polymerization of the olefins. Such side reactions are effectively con-
trolled by successive introductions of the olefin into the isobutane stream
so that at no point is there a high concentration of olefins which will
produce a high polymerization rate. Multiple-point introduction of a
reactant in this manner may be practiced in a series of reactors or by
providing multiple-feed points in a single reactor, but in either case it
may Jse considered as a concurrent multistage operation.
Prediction of the operating results of concurrent multistage reactor
systems involves no new principles if rate equations are available for all
important reactions. Calculations' are started from the reactor inlet
and carried through the successive stages, each addition of reactants
being taken into account.
Where it is desired to produce a high degree of completion of a poly-
CHAP. XXl] MULTISTAGE OPERATION 1029
molecular reaction with a minimum of excess reactants it is frequently
economical to employ countercurrent multistage operation. For exam-
ple, in the reaction A -{- B ^ R it may be desired to obtain maximum
conversion of the limiting reactant A with minimum excess of reactant B.
This would be accomplished in a reactor in which A is continuously con-
tacted with B in counterflow. With this arrangement a maximum
concentration of A is present at one end where 5 is at a minimum, while
at the other end £ is at a maximum where A is at a minimum. Such a
continuous counterflow operation is possible in a single reactor if B is
gaseous while A and R are Hquid under the conditions of reaction. In
the general case where the' reactants and products are in a single phase at
«doWj!oM/' W B O % WÔ^BOWJ' WBO% nMUBan,! riBinii
ITo 3: ITo 1. TTi
o
ni si
« K
TCBIW, >
1^ »B0M,
To
«fii%' Separator Separator WAIMJJI ""
I \-^ I II Wj'
•R.11 71,'
Single Stage Two Stage Counter Flow

PIG. 200. Flow Diagrams for Multistage Reactors. '
reaction conditions it is possible to approach continuous counterflow

operation by a series of separate reactor stages with separation of the
reactants between stages. The practicability of the counterflow opera-
tion depends largely upon the difficulty of this separation. Ordinarily
it must involve only cooling followed by a simple phase separation of a
gas or solid from a liquid or the separation of two immiscible liquids.
In Fig. 200 are flow diagrams representing single-stage, two-stage
counterflow, and continuous counterflow operations for the general reac-
tion A + B-?±R. A larger number of multiple stages, approaching the
effect of continuous counterflow, is obtained by adding intermediate
reactors and separators to a two-stage arrangement.
The symbol n on Fig. 200 represents the number of moles of each com-
ponent, per unit mass of total feed to the system. The diagrams repre-
sent operations which are possible if A and R are liquids, associated with
a liquid inert component l' and if B is a gas associated with a gaseous

inert component / . In such systems the pressure drops in the reactors
are determined by the direction of gas fiow. Thus, in the two-stage
operation TTQ > 7r2 > vi, and the indicated pump is necessary to transfer
the hquid phase from state I to II. By the addition of suitable heaters
and coolers this two-stage operation is feasible with single-phase con-
ditions of the fluid streams in the reactors. An infinite number of such
stages would yield the counterflow operation.
If XA and x^ represent the conversion of A per mole of total feed to the
system in the first and second reactors, respectively, the composition of
the system at any point in a reactor is represented by the material-
balance equations in Table LXI.
TABLE LXI
MUI/TISTAGE-REACTOB EQUATIONS
Counterflow
Single Stage Two Stages
(Infinite
First Second
Stages)
nA IIAO—XA nAU—XA riAi —X'A UAO—XA
UB WflO—XA riBi—XA nBO—XA nBl+XA
nB nito+XA WBO+XA UBI+XA nso+XA
nt nô+JiBo+JiEo WAo+nBz+riBo n^l+rtBO + Jlfil nAa+nBi-\-nRo
+n[+ni'—XA +ni+n'i-XA +ni+n'[-x'A +ni+nj+XA
Overall nBi=nBo—XAi tlAl — nAO—XAl UBi^flBtl—XAZ nBi = nBa—XAi
balances nRi=nBo+XAi nAi=nAit—XAi XAl=nAO—XAl
1
XA^
If the pressure drops in the reactors do not significantly affect the reac-
tion rates the amounts of catalyst required in the reactors of a two-stage
system are readily calculated by either analytical or graphical integra-
tion of/Equation (XIX-64) in combination with the reaction rate equa-
tions and the material balances of Table LXI. For specified values of
XAi. and X'AI, the mole fractions of the rea'ctants in each reactor are ex-
pressed as functions of the conversions XA and X'A. The corresponding
activities are then calculated and the integrations completed by the usual
methods, thus determining the amount of catalyst required in each
reactor. Variations in temperature and pressure in each of the reactors
may be taken into account if the inlet conditions are known. If no
compressor is used between the stages on the gas stream the pressure at
the inlet to the first reactor must be determined by a trial-and-error
procedure in which pressure drops are first assumed and then verified by
the reactor integration.
This same procedure may be repeated for different values of xî at a
CHAP. XXI] REACTOR TEMPERATURE CONTROL 1031
constant over-all conversion XAI + Â2- The results of such a series of
calculations may be plotted with weights of catalyst per unit mass of
feed W/F as ordinates against XAI, the conversion in the first reactor. A
typical set of curves is shown diagrammatically in Fig. 201. It may be
noted that the minimum catalyst requirement corresponds to a first-
stage reactor approximately 50 per cent larger than the second. How-
ever, there are frequently mechanical and operating advantages in build-
ing reactors of equal size. In
this case the increase in over-
all catalyst I'equirement would
be neghgible.
The equations corresponding
to an infinite number of stages
or continuous counterflow are
readily integrated if constant
temperature and pressure are
assumed. The mole fractions
throughout the reactor are ex-
pressed as functions of XA by
the equations of Table LXI.
First Stage Conversion, xî-
The size of the reactor is ob-
FiG. 201. Catalyst Requirements for a Two-
tained by determining the area
Stage Reactor System.
under a curve of l/r^m plotted
as ordinates against XA- Where constant pressure and temperature are
not assumed the calculations become difficult. In most cases the ad-
vantages of more than two counterflow stages are too small to justify
the added complexity of the equipment. This conclusion is readily
verified by comparing the results of an optimum two-stage system with
the hypothetical infinite stage case.
The economics of multistage operation are reviewed by KasseP who
developed design charts for the hydrogenation of butene codimer on the
basis of a simplified rate equation which permitted analytical integration.
REACTOR TEMPERATURE CONTROL

In many catalytic operations the most serious design problem is that
of providing suitable control of the temperatures in the reactor to
approximate optimum conditions. Six fundamentally different methods
of temperature control are in more or less common commercial use.
1. The method of dividing thg reactor into adiabatic sections with
intermediate heat exchangers has already been mentioned. This scheme
«L. S. Kassel, Ind. Eng. Chem., 31, 275 (1939).
is effective but leads to mechanical complications, particularly in high-

temperature reactions.
2. Temperature may be controlled by the addition of an inert heat
carrier to the reactant stream. Where it can be used steam is a most
effective heat carrier because of its high heat capacity and the ease
with which it may be separated from gaseous systems by condensation.
In some operations the molal ratio of steam to reactants is ten or greater.
Under such conditions even a large heat of reaction produces only a small
temperature change.
3. Temperature may be controlled* by the injection of reactants at
various points along the length of the reactor. For example, an exo-
thermic reaction may be controlled by successive additions of a cold
reactant.
4. The catalyst may be contained in a heat exchanger so arranged that
confining surfaces which are in contact with a heating or cooling medium
are in close proximity to all catalyst particles. This method of tempera-
ture control depends upon lateral transfer of heat through the catalyst
bed at right angles to the direction of flow of tlie reactants. Because the
conditions for heat transfer are relatively poor, it is generally necessary
that no catalyst particle be more than 2 in. from a cooling surface if its
temperature is to be appreciably affected. The problem of calculating
the temperature gradients resulting from such lateral heat transfer is
discussed on page 1033.
5. The catalyst bed may be moved in a steady stream through the
reactor. Two types of moving-bed reactors are in common use. In the
" Thermofor "^ or TCC catalytic reactor the catalyst is in the form of
cylindrical or spherical particles which arei permitted to flow slowly
downward as a moving sohd bed. In the " Fluidized " ' reactor the
catalyst is in the form of a powder which is carried by the reactant stream
and forms a relatively dense bed in the reactor in a turbulent fluidized
state. In both methods the heat capacity of the catalyst stream is of the
same order of magnitude as that of the reactants, and heat may be added
to the reactor by preheating the catalyst. Heat may be removed by
passing all or part of the catalyst through a cooler and returning it to
the reactor or by placing cooling coils in the path of flow in the reactor.
6. What may be termed a regenerative reactor has been developed for
use where a carbonaceous catalyst deposit from an endothermic reaction
frequently must be removed by oxidation. In this operation the catalyst
' R. H. Newton, G. S. Dunham, T. P. Simpson, Trans. Am. Inst. Chem. Engrs.,
41, 215 (1945).
' E. V. Murphree, C. L. Brown, E. J. Gohr, C. E. Jahnig, H. Z. Martin, C, W.
Tyson, Trans. Am. Inst. Chem. Engrs., 41, 19 (1945).
CHAP. XXI] LATERAL HEAT TRANSFER 1033
bed is used as a heat reservoir to store up heat from the exothermic regen-
eration period. This heat is then given up to the reactants in the subse-
quent process period. Good temperature control is possible by this
method, but the cycle of operation must be short, and a perfect heat
balance must be maintained between the regeneration and process
periods.
The design problems in the first three types of systems may be handled
by the principles developed in the preceding sections. Where pressure
changes are significant the progressive stepwise procedure described on
pages 875 to 883 permits simultaneous development of temperature, pres-
sure, and conversion relationships in any adiabatic reactor. If pressure
changes are negligible direct graphical integration as indicated in
Fig. 197 estabUshes both conversions and temperatures throughout the
reactor. This operation is direct and simple where a chart is available
which expresses reaction rate as a function of both temperature and con-
version. Integrations for adiabatic reactors are readily carried out from
adiabatic operating lines plotted on such a chart.
The problems of lateral heat transfer in the fourth type of reactor are
discussed in the following sections. The heat-transfer and temperature-
distribution problems of the last two types of reactors involve specialized
data or procedures which are not yet generally available.
LATERAL HEAT TRANSFER

For control of the temperature of a catalytic reaction by lateral heat
transfer through the bed four different types of reactor arrangement have
been used.
1. The catalyst is placed in vertical cylindrical tubes which are sur-
rounded by the heat-exchange medium.
2. The catalytic bed is formed around tubes which carry the heat-
exchange medium. Elaborate fins are welded to the tubes to improve
heat transfer into the grandular bed.
3. The catalyst is placed in flat slabs in the alternate envelopes of a
flat-plate heat exchanger of the type commonly used for preheating air
for combustion.
4. The catalyst is placed in the aimular space between two cylindrical
tubes. Heat-exchange medium circulates through the inner and around
the outer tube.
Various types of heat-exchange medium may be iised with any of these
reactor types. Boiling liquids, condensing vapors, circulating liquids,
molten salts, and circulating gases have all been used.
For the design of any of these reactors it is necessary to predict the
lateral temperature distribution at all cross sections. The magnitude of

the lateral temperature differences fixes the .diameter or distribution of
the tubes, the arrangement of the fins or the thickness of the envelope
which is required for specified Umits of temperature variation.
The problem of calculating the temperature gradients accompanjdng
lateral heat transfer is made difficult by the complex combinations of
mechanisms involved. Some> heat is transferred by conduction through
the catalyst particles and the intervening fluid, some by the convection
and turbulence of the fluid, and some by radiation between the particles.
It is evident that what may be termed the effective thermal conductivity
of the bed is a function of the conductivity of the catalyst particles; their
shape, size, and arrangement; the properties of the fluid, its mass velocity,
and the temperature; and emissivity of the catalyst. If it is assumed
that such an effective conductivity is known in terms of lateral heat-
transfer rate per unit area and unit temperature gradients, differential
equations may be developed to represent the temperature distributions in
catalytic beds of simple geometric shapes. General correlations for'pre-
dicting effective thermal conductivities must be developed before these
relationships become quantitatively useful.
Catalyst in Thin Slabs of Infinite Eirtent. This case is approximated
in the flat-plate type of reactor previously described. It is assumed that
the temperatures of the confining surfaces on each side of the bed are
equal and known at each lateral cross section. The
dg^
conditions in any lateral section at a distance Z from
<-X <dX the inlet are shown diagrammatically in Fig. 202 for
dZ a reactor in which the flow of the fluid is downward,
Ix
^z \cLq,^> parallel to the confining walls. An elementary sec-
tion dZ in height and dX in thickness is located at
%i a distance X from the center line of the bed. For a
FIG. 202. Lateral bed section of unit width the volume of this ele-
Heat TransferinThin jênt is dX dZ. If the longitudinal transfer of heat
^ ^' by any means other than the heat capacity of the
flowing fluid is neglected, the rates of heat transfer to and from the
element are represented by the following equations corresponding to the
symbols on Fig. 202.
dqi = GC^dX (27)
dqi = GCp (t + ^ d7\ dX , (28),
dq,= -hdZ^ (29)

CHAP. XXI] CATALYST IN CYLINDRICAL TUBES 1035
dqs = ~kdZ •
dX
dqz = - VAPB^HA dX dZ (31)

where
G = mass velocity
Cp = mean heat capacity per unit mass of fliiid
k = effective lateral thermal conductivity of the catalyst-bed
fluid system
TA = rate of reaction, moles of A per unit mass of catalyst
AHA = heat of reaction per mole of A
PB = bulk density of catalyst bed
The energy balance is expressed by

dqi + ofg'2 + dqi = dqi + dq^^ (32)
Substituting Equations (27-31) in (32) gives
where
(33)
A = k/GCp
TAPBAHA
B= -
uCj,
Direct integration of Equation (33) is made difficult by the complex

variation of B with temperature and conversion XA which in turn vary
with both Z and X.
Catalyst in Cylindrical Tubes. The conditions in a catalyst bed inside
a cylindrical tube are shown diagrammatically in Fig. 203. This case
differs from that of the flat slab in that the heat flow is radial, through a
cross section of varying area. The following equations similar to
(26-33) represent the rates of heat transfer into and out of an elementary
ring of height dZ, radial thickness dX, and radius X.
dqi = GCpt 2'KX dX (34)
dqi = GCj, (t + ^ dZj 2-KX dX (35)

dqi = -fc(^)2.ZrfZ (36)
dqs, = -k -^ L 2x(X + dX) dZ (37)

dX
dqi = - rAPBAHA2TX dX dZ (38)
dqi + dq2 + dqz = dqi + dqs (39)
Combining and simplifying jaelds
where A and B are the same as in Equation (33).

Equations similar to (27-40) can also be devel-
oped for mass transfer. Thus, for a cylindrical
reactor,
\dz )x \P-VA)\RT) \_dx^ ^ X
FIG. 203. Lateral

Heat Transfer in a
Cylindrical Reactor.
m\ G
Use of Equation (41) is made difficult by uncer-

(41)
tainty as to the proper effective diffusivity D^m to be used for a fluid

moving through the voids of a catalyst bed.
Double Stepwise Integration. Grossman^ has presented a double step-
wise method of integrating Equations (33) ai\d (40) which is based upon
the use of small but finite increments. The equations are written in incre-
mental notation as follows:
For thin slabs:
AJ, "^^^-{AU) + B AZ (42)
(Axy
For cylindrical reactors:
AJ.
AAZ
{Axy G--f") + BAZ (43)
where Aji denotes the change of temperature in the longitudinal incre-

mental distance AZ and AJ, denotes the change of temperature in the
lateral or radial distance AX. i
^ ' L. A. Grossman, Trans. Am'. Inst. Chem. Engrs., 42, 535 (1946).
CHAP, XXI] DOUBLE STEPWISE INTEGRATION 1037
The term B of Equations (33-43) involves the reaction rate r^ which
is determined by the temperature and activities at the interface of the
catalyst particle. Where diffusional and temperature differences
between the fluid and the catalyst are negUgible a general relationship
between B and the average temperature and conversion in the fluid
stream is readily derived from the rate equation. Where these gradients
cannot be neglected it is necessary to develop a chart similar to Fig. 198
which expresses reaction rate as a function of the average properties of
the fluid stream but is restricted to a single mass velocity. It is thus
A
• 0
yy
*n, * + l
y' X
i B'LZ 1 y^ /
B ^ .'-''
u tn.k
S
*n-l. * R
H
In (n—1) In n In (w+1)
FIG 204. Graphical Calculation of Temperature in a Cylindrical Reactor.
possible to present graphically both A and B as functions of the average

temperature and conversion XA in the fluid stream at any point, even
though the relationships may be very complex. Such charts are shown
in Figs. 205-207 for the system of Illustration 2. The values of B and A
in Equations (42) and (43) are properly based on the mean temperature
and conversion existing over the increment AZ at the position under con-
sideration.
The conversion AJSIA produced in any increment of length Z is expressed
by
rkPB^ZS
^ZXA = — (44)
where S = total cross-sectional area of reactor

F = reactor feed rate
TA = rate of reaction at the average temperature and conversion
of the increment AZ, moles/(mass of catalyst) (time)
The stepwise solution of Equations (42) and (43) is facilitated by

maintaining the following definite relationship between the lateral and
longitudinal dimensions of each incremental volume considered:
AZ =
(Axy
(45)
2A
0.86
0.30-
0.25
0.20-
t- -
0.15-
0.10
0.05
200 240
FIG. 205.V Effect of Temperature and Conversion on a Specific Catalytic Reaction

Rate.
With this restriction Equations (42) and (43) become
(Slabs) A,i = - (AÔ +BAZ (46)
(Cylinders) A,i = i Uy, + ^ A.A + B AZ (47)

CHAP. XXI] DOUBLE STEPWISE INTEGRATION 1039
In applying Equations (46) and (47) the lateral or radial dimeiîon of
the reactor is divided into n small but finite equal increments AX. Then,
X = nAX (48)
The longitudinal dimension of the reactor is divided into k small incre-
ments AZ whose lengths are not necessarily equal but.are related to the
0.0038
200 240 280 320 360 400 440
Temperatures, °C
FIG. 206. Effect of Temperature on Factor A in a Specific Catalytic Reaction.
corresponding values of AZ through Equation (45). Then, if the lateral

gradients about any point Z, X corresponding to increment k, n, are
considered:
A,t = fn.m-l-in.fc ^ (49)
Ax< = K W i . k - tn-i, k) (50)
Alt = (in+l. k — in, k) — (tn.k — tfl-l. k) (51)
where <„. t+i indicates the temperature at a point (Z + AZ), X and
tn+i.k indicates the temperature at a point Z, {X + AX). Combining
Equations (48-51) with (46-47) gives:
For slabs:
A J, = K W i . k + t„-i. k - 2tn. k) +BAZ (62)
For cylinders:
Aji=
\zt— h]
2I ^<n+l.
5 * + 'n-l, k — 2i„, * + 5~ (Wl. k tn-1, k) I + BAZ (53)
If the lateral temperature distribution is known at some starting point

such as the reaction inlet Equations (52) and (53) permit direct calcula-
tion of the successive increments in longitudinal temperature corre-
8000
7000
6000
6000
014000-
8000
2000
1000
FIG. 207. Effect of Temperature and Conversion on Factor B in a Specific

Catalytic Reaction.
\
spending to the lateral position X of each of the selected increments AX.
However, the longitudinal temperature increments so calculated do not
correspond to uniform increments in, AZ,' and Equation (45) must be
successively used to establish the value of Z corresponding to each
temperature.
Grossman developed the successive solution of Equations (52) and

(53) for both slabs and cyHnders by a graphical procedure similar to the
general method of Schmidt''i" for heat-transfer calculations in the un-
steady state in slabs. This procedure is demonstrated in Fig. 204 for
the case of a cylindrical reactor involving Equation (53). The radius of
the reactor is divided into equal increments AX and on the abscissa scale
of Fig. 204 are plotted values of In n where n is the number of the incre-
ment, starting with zero at the center. Thus, the center of the reactor
corresponds to negative infinity on this scale while the zero point repre-
sents the first increment at a distance AX from the center. Figure 204
illustrates the gradients kbout a typical increment number n for an endo-
thermic reaction under conditions such that the temperature of the fluid
stream is being increased by heat transfer from the walls of the reactor.
It may be shown from the geometry of Fig. 204 that the distance PN
is equal "to one half of the terms inside the brackets of Equation (53)
and that QP is therefore equal to B AZ. Thus,
In ^
V
BP _ M ^ _ In (n) - In (n - 1) ^ n - 1^ 1 J_ ' ,^.
S0~ MS~ In (n + l) - In (n - 1) ^^ n + 1 2 4w
n — 1
Then,
PN = RP-RN = SO (i + ^)-RN v (55)
\2 4n/
or
PN = (Wl. ft - <n-l. *) ( 2 + i;^) ~ (*»• * ~ *"-!• *)
= - tn+1, k + in-1, k — 2 i „ , k -\- — (.tn+1, k — < n - l . k) I (56)
Comparison of Equations (56) and (54) shows that point Q on Fig. 204
may be established by graphically locating point P and adding the tem-
perature increment B AZ = PQ. Since the value of B should be based
on the average temperature and conversion over the increment AZ it may
be necessary to evaluate it by successive trials, if the changes per incre-
ment are large. The change in conversion is expressed by a combination
of Equations (44) and (45). '
r^PB(AX)^^ •
^^^^ = 2AF ^^^)
• » E. Schmidt, Beitr. Tech. Mechanik und Tech. Physik, pp. 179-89 (1924).
" E. Schmidt, Forsch. Gehiete Ingenieur-w., 13,177-84 (1942).
This procedure is repeated for each increment across the reactor, thus
establishing a temperature-distribution curve corresponding to the
longitudinal increment fc + 1. The corresponding AZ values are calcu-
lated from Equation (46) to permit a plot of Z against temperature and
conversion for each radial position.
It is readily demonstrated that Equation (52) may be solved by a
graphical procedure similar to that of Fig. 204, except that the abscissas
are increment numbers n plotted on a uniform scale instead of In (n).
At the center of either a slab or a cyhndrical reactor where n = 0 the
temperature i(ft+i)(„=o) at a value of Z corresponding to increment (k + 1)
is approximated by adding the proper mean value of B AZ to the tem-
perature <4(„_i). For obtaining the value of <(A+i)(n=i) in a cylindrical
reactor the appropriate value of B AZ is added to the value of iA(„=2).
This is equivalent to making line 0PM of Fig. 204 horizontal from In (2)
to negative infinity in order to locate point P where n = 1.
This method of calculation neglects lateral mixing of the fluid flowing
through the bed as well as all diffusion or, mixing in a longitudinal direc-
tion. It is also assumed that the mass velocity is uniform across the
entire cross section of the reactor. Although these assumptions are not
rigorous it is probable that the error they contribute is small in compari-
son to the uncertainty of the effective thermal conductivity k.
Illustration 2. In a hydrogenation operation an olefinic vapor is charged to a

cataljrtic reactor with 1.2 moles of hydrogen per mole of olefin. The mixture enters
the reactor at 200°C and 5 atm at such a mass velocity that partial pressure and
temperature differences between the fluid stream and the catalyst are negligible. The
effective thermal conductivity k of the gas-solid systeni is estimated to be 0.5 Btu /
(hr) (ft) (°F). The catalyst is contained in cylindrical tubes having an internal diame-
ter of 1.25 in. The walls of the tube are surrounded by boiling water at a temperature
of 200°C. It is desired to calculate the temperature and conversion gradients
throughout a representative tube.
From the catalytic rate equation the rates of reaction in pound-moles of olefin
reacted per hour per pound of catalyst were calculated at various temperatures and
conversions and plotted in Fig. 205. It may be noted that this reaction rate passes
through a maximum as temperature is increased. This retardation of the reaction
rate occurs at temperatures far below that at which the reverse reactive becomes
appreciable.
Based on the rate relationships of Fig. 205, the thermochemistry of the reaction,
and the physical properties of the system, the values of A and B in Equations (33)
and (40) were calculated at various temperatures and conversions and plotted in
Figs. (206) and (207), respectively. The effect of conversion on A is negligible.
The feed rate and catalyst-bed density are such that Equation (57) becomes
31.8 (AX) VA
ASXA = ;; (a)
The method of graphical calculation based on Equation (63) and Fig. 204 is
demonstrated in Fig. 208 for the increments k = 1, 2, and 7, corresponding to five
equally spaced radial positions extending from the center of the tube where n = 0
to the wall where n = 6. The lines used in the calculations for the longitudinal incre-
ments fc = 3, 4, 5, and 6 are omitted from Fig. 208 for simplification, although it is
ln{l-AX) ln(2-AX) ln(3-AX) ln(5-AX)

ln{i-AX)
- Radial Positions »-
FIG. 208. Graphical Method of Calculating Temperature and Conversion Gradients
in a Cylindrical Reactor.
evident that they must be carried out in order to establish the gradient at A; = 6.
Since A Z = (1.25)/(12) (5) (2) = 0.0104 ft, Equation (45) becomes.
(0.0104)2 54.1 (10-6)
AZ = (b)
2A A
0.00344 TA
and Equation (a) becomes AÂ = (o)
At the reactor inlet Z = (i,k — 0,f = 200, and XA = 0 for all radial positions. Over
the interval from fc = 0 to fc = 1 a mean temperature of 211°C and a mean conversion
of 0.026 are assumed for a first trial calculation at all radial positions except at the wall
(n = 5) where t is kept constant at 200°O and a mean conversion of 0.019 is assumed.
Values of TA, A, and B are read from Figs. 205-207 for these conditions.
n = 0, 1, 2, 3, 4 n = 5
TA ' 0.078 0.055
A ' 0.00522 0.00528
B 2140 1500
AXA, Equation (a) 0.0514 0.0358
AZ, Equation (b) 0.0103 0.0103
BAZ 22.3
The temperatures at fc = 1 are hence 200 + 22.3 = 222.3°C for all radial positions
except at the surface which is at 200°C, and conversions are 0.051 for all radial posi-
tions except at the surface which is 0.036. The mean temperatures for all posi-
tions except the surface are 211 in agreement with the assumed value, and the mean
conversion is 0.025 at all positions except at the surface which is 0.019, both in agree-
ment with the assumed values; the calculated values are hence correct. These calcu-
lations are summarized in Table A.
Similar calculations are repeated for the second increment of AZ from fc = 1 to
fc = 2. Since values of temperature at fc = 1 and radial positions n = 1, 2, 3, and 4
are the same A^t = B AZ in the first three radial positions. At the radial positions 4
and 5 the temperatures are determined by adding values of B AZ from Table A to the
points determined by the intersections of the lines on Fig. 208. These calculations
are repeated for succeeding increments as indicated in Table A and Fig, 208.
Values of temperature and conversion for given radial positions in Table A are
plotted against longitudinal distance Z in Fig. 209. In this figure the mean and adia-
batic lines as well as all radial lines are shown for both temperature and conversion.
The mean values of t and XA were obtained by graphical integration from 0 to X^
from a cross-plot of the six radial positions against X":
The foregoing procedure may seem extremely laborious, especially in securing
agreement between assumed and calculated mean values of temperature and con-
version; however, the first trial values can be made satisfactory after the first few
increments of AZ have been passed by plotting (and XA againstfcand extrapolating to
the next half interval of AZ. The greatest difficulty occurs, where the temperature
gradients begin to reverse in direction as shown in the range of fc = 6.
Unexpectedly high radial temperature gradients are shown in Fig. 209. The mean
temperature line occurs at a radial position of about 0.7 of the radius and a mean
conversion at a radial position of about 0.5. It is particularly interesting to note that
the maximum conversion occurs at a radial position of 0.8 rather than at the center of
the tube.
In this analysis no allowance was made for radial diffusion of gases, and the value of
thermal conductivity is uncertain. The method can be extended to include diffusion
effects. At present there are no published data on either thermal conductivity or
diffusion of fluid streams flowing through granular solids. Much experimental work
is needed in this field.
.00 0.04 0.08 0.12 0.16 0.20 0.24 0.28 0.32 0.36
Z=Feet
FIG. 209. Conversion and Temperature Gradients in a Cylindrical Reactor.
TABLE A
CAIICULATION OF TEMPERATURE AND CONVERSION GRADIENTS IN A
CYLINDRICAL REACTOR
Radial
Positions tv, BUZ
&Z
(.trial) Itrid) rA (trial) AxA (calc^)(.cak.)
= Otol
Center 0 200° 0 0
211° 0.025 0.00522 0.0103 2140 0.078 22.3° 0.0514 211° 0.026
1 200° 0 0
211° 0.025 0.00522 0.0103 2140 0.078 22.3° 0.0514 211° 0.026
2 200° 0 0
211° 0.025 0.00522 0.0103 2140 0.078 22.3° 0.0514 211° 0.026
3 200° 0 0
211° 0.025 0.00522 0.0103 2140 0.078 22.3° 0.0514 211° 0.026
4 200° 0 0
211° 0.025 0.00522 0.0103 2140 0.078 22.3° 0.0514 211° 0.026
Surface 5 200° 0 0.019 0.00528 0.0103 1500 0.055 0.0358 0 200° 0.019
& = 1 to2
Center 0 222° 0.051 — 0.0103

251° 0.125 0.00493 0.0110 5350 0.210 59.0° 0.146 251° 0.125
1 222° 0.051 0.0103
251° 0.125 0.00493 0.0110 5350 0.210 59.0° 0.146 251° 0.125
2 222° 0.051 0,0103
251° 0.125 0.00493 0.0110 5350 0.210 59.0° 0.146 251° 0.125
3 222° 0.051 0.0103
251° 0.125 0.00493 0.0110 5350 0.210 59.0° 0.146 251° 0.125
4 222° 0.051 0.0103
235° 0.097 0.00506 0.0107 3490 0,135 37.2° 0.092 235° 0.097
Surface 5 200° 0.036 0.054 0.00528 0.0103 1500 0.057 0.037 0.0103 200° 0.054
= 2 to 3
Center 0 281° 0.197 0.0213

1 281° 0.197 0.0213
2 281° 0 197 0.0213
3 281° 0.197 0.0213
4 24r 0.143 0.0210
Surface 5 200° 0.073 0.0206
* =6to7
Center 0 380' 0.480 0.0590

384° 0.503 0.00417 0.0130 950 0.058 12.4° 0,0479 382' 0.504
1 37r 0.491 0.0584
382° 0.515 0.00417 0.0130 1000 0.058 13.0° 0.0429 382' 0.515
2 374' 0.499 0.0583
375° 0.52 0.00420 0.0129 1050 0.049 13.5° 0.040 375' 0.618
8 352" 0.533 0.0567
351° 0.56 0.00433 0.0125 1450 0.068 18.2° 0.054 351 0.660 '
4 303° 0.555 0.0547
301° 0.695 0.00458 0.0118 2550 0.106 29.7° 0.0792 301° 0.595
Surfaces 200' 0.182 0.198 0.00528 0.0103 1450 0.050 0.0325 0.0516 200° 0.198
CHAP. XXI] PROBLEMS 1047
ft = 7 t o 8
Center 0 389° 0.528 0.0720

1 38r 0.525 0.0714
2 375° 0.539 0.0712
3 351° 0.687 0.0692
i 299° 0.644 0.0665
Surface 5 200° 0.214 0.0618
PROBLEMS
1. The nickel catalyst of Illustration 2, Chapter XIX, is to be used for the hydro-
genation of butene codimer in 4 single-stage isothermal reactor operating at a gauge
pressure of 150 lb per sq in. Pure hydrogen is fed in 40 per cent excess of that
required for complete reaction, the remainder being recycled. Neglecting mass and
heat-transfer effects and pwessure drop, calculate by graphical integration the weight
of catalyst which must be supplied per pound-mole of codimer fed per hour in order to
produce 99.8 per cent hydrogenation at an inlet temperature of:
(o) 250°F.
(6) 350°F.
The reverse reaction may be neglected.
2. Repeat problem 1 for an adiabatic reactor in which hydrogenated product is
recycled for temperature control in the ratio of two moles of hydrogenated product
per mole of codimer feed. Hydrogen is supplied in 40 per cent excess of that
required to hydrogenate the olefins in the feed. It is desired that the total prod-
uct shall be 99.8 per cent hydrogenated codimer. Determine the catalyst-feed
ratio for the following inlet temperatures:
(a) 275°F
(6) 300°F
3. Hurti found that the oxidation of a gas containing 7.8 per cent SO2, 10.8 per
cent O2 and 81.4 per cent N2 over a catalyst in the form of | " X | " pellets is repre-
sented by the following equation for the height of a reactor unit HB in inches,
HB = [1+ h(.Psoz)]iHi + He)
where HD — height of a transfer unit at average conditions, inches,

pso, = average partial pressure of SO3, atm.
1^22-16.89
He = e ^ (in.) = height of a catalytic unit at a mass velocity of
600 lb / (sq ft) (hr) and a pressure of 1.0 atm
'"'" -5.84
fci = e (1 /atm)
Calculate the depth of the catalyst bed required to produce 95 per cent oxidation
of the SO2 in an isothermal reactor operating at a temperature of 450°C and 1.8 atm
with a mass velocity of 500 lb/(hr)(sq ft). The pressure drop in the bed may be
neglected, and the average height of a transfer unit may be calculated by consideration
of the diffusion of only SO2, if zero concentration is assumed at the interface.
'4. A catalyst in the form of ^ " X x%" cylinders is found to pack with a void frac-
tion of 0.345 in random dense arrangement. A mixture of 10 mole per cent butene—2
with 90 per cent steam is passed at a mass velocity of 620 lb/(sq ft) (hr) through the
bed at an inlet temperature'of 1200°F and an absolute pressure of 24 lb per sq in.
(o) Calculate the pressure drop in pounds per square inch per foot of bed depth at
the inlet conditions if the bed is packed in a dense arrangement.
(5) Repeat the calculation of part (a) for a bed in a loose arrangement having void
fraction of 0.42. .
6. Calculate the pressure drop in pounds per square inch per foot of bed depth at
the inlet conditions of the reactor of problem 5 of Chapter XX if the bed is in random
dense arrangement with a 36 per cent void volume.
6. Calculate the pressure drop in pounds per square inch per foot of bed depth in
the reactor of problem 6 of Chapter XX if the catalyst when in dense arrangement has
36 per cent void volume.
7. It is proposed to dehydrogenate w-butane to produce 35 per cent conversion in a
catalytic reactor operating at a temperature of 1050°F with an absolute pressure of
200 mm of Hg at the outlet and an inlet pressure of 375-425 mm of Hg. It is known
that under these conditions considerable yields of butadiene are obtained, but this
secondary reaction and the accompanying pyrolysis may be neglected and the reac-
tion assumed to follow Equations (XIX-94-98). The catalyst bed has the same
density and void volume as that of problem 6, Chapter XX. '
(a) Assuming isothermal operation and neglecting mass-transfer effects, determine
the space velocity required by a graphical integration of the type of Fig. 197, based on
an assumed pressure distribution.
(6) Calculate the depth of bed corresponding to the specified pressures and the
space velocity of part (o).
(c) Based on the bed depth and mass velocity of parts (o) and (6), calculate the
pressure distribution in the reactor.
{d) Correct the space velocity calculated in part (a) by a second integration based
on the corrected pressure distribution of part (c). If a large correction is involved,
parts (b) and (c) should be repeated.
8. Consider the reactor of problem 2 which corresponds to an inlet temper-
ature of 275°F as built in the form of a vertical cylindrical bed 10 ft in depth. The
pellet density of the catalyst is 3.5 g per cc, and it is placed in the bed in a random
dense arrangement with 36.5 per cent void volume. Retaining this bed depth and with
the mass velocity fixed by these conditions, develop a corrected determination of
catalyst-feed ratio which takes into account the effects of heat and mass transfer, and
pressure drop.
The pressure drop per foot of bed depth may be calculated for both the inlet and
outlet conditions of compositon and temperature, if a mean gauge pressure of 150
lb per sq in. is assumed. The total pressure drop may then be based on an average of
these two values. From the inlet and outlet pressures calculated on this basis con-
struct an approximate curve which relates pressure to bed depth and conversion and
conforms,to the calculated pressure drop per unit length at the inlet and outlet.
The rate of reaction based on correct interfacial partial pressures and temperatures
should be calculated at the inlet conditions and at the conditions corresponding to
approximately' the maximum rate encountered in the reactor. From these points a
complete corrected rate curve is estimated by the principle of Equation (5).
9. Calculate the rate of reaction for the system'of Illustration 1 and Pig. 198 at a
temperature of 825°K and a conversion of 70 per cent.
CHAP. XXI] PROBLEMS 1049
10. From the data of Fig. 198 plot curves relating conversion to the catalyst-
feed ratio W IF in lb / (lb-mole) (hr) for reactors operating under the following con-
ditions:
(a) Temperature control to produce maximum rate at all conversions.
(6) Adiabatic operation from 0 to 52 per cent conversion with an inlet temperature
of 673°K followed by temperature control to maintain maximum rate at higher
conversions.
(c) Isothermal operation at 750°K.
(d) Single-stage adiabatic operation with an inlet temperature of 673°K.
(e) Two-stage adiabatic operation with an intercooler. The inlet temperature for
the first stage is 673°K, whereas that for the second stage is 700°K with a conversion
of 70 per cent.
11. If the catalyst of probleni 10 has a bulk density of 52 lb per cu ft and is arranged
in a bed having a random dense arrangement with 36 per cent void volume, calculate
the dimensions of the vertical cylindrical beds required for the two-stage operation of
part (e) with an ultimate conversion of 95 per cent. Calculate the pressure drop
through each bed.
12. It is desired to hydrogenate butene codimer over the nickel catalyst of Illus-
tration 2 of Chapter XIX at a gauge pressure of 100 per sq in. and a temperature of
350°F. The catalyst is placed in small-diameter tubes which are surrounded by
boiling water held under a suitable pressure for temperature control. The operation
may be treated as isothermal, and pressure drops, diffusional gradients, and tempera-
ture differences between the catalyst and vapors may be neglected.
The liquid charge is 100 per cent butene codimer while the source of hydrogen is a
methane-hydrogen mixture containing 80 mole per cent H2. It is desired to produce
99.8 per cent saturation of the codimer with an excess of hydrogen of 5 per cent above
that required for complete saturation.
Calculate the weight of catalyst for a codimer feed rate of 1 bbl (42 gal) per hr in
each of the following systems:
(o) A single-stage reactor.
(6) A hypothetical counterflow or infinite-stage operation.
(c) A two-stage counterflow operation, if complete separation of the hydrogen and
methane from the hquid in the separators is assumed. Calculate the weights of
catalyst in each stage and the combined weight with conversions of 20, 40, 60, and
80 per cent in the first stage. Graphically estimate the optimum first-stage conver-
sion, and compare the corresponding minimum quantity of catalyst with the results
of parts (o) and (6).
13. Normal butane is to be dehydrogenated in the reactor described in problem 6
of Chapter XX. Assuming that tlê catalyst bed is at a uniform temperature of
1075°F throughout, calculate by a stepwise integration from Equations (XIX-94-100)
the conversion and selectivity resulting from the operation with an inlet temperature
of 1075°F and an inlet gauge pressure of 20 lb per sq in. The effects of mass and
heat transfer between the-fluid and the catalyst particles may be neglected.
14. It is desired to investigate the possibiUty of dehydrogenating butane in a three-
stage adiabatic reactor with interheaters between the stages. Each stage is to pro-
duce 10 per cent conversion of the butane in the original feed. * The temperature at
the inlet to each stage is to be maintained at 1125°F. The gauge pressures at the
inlets to the reactors are, respectively, 23, 13, and 3 lb per sq in. Neglecting the
effects of mass and heat transfer and assuming applicability of Equations (XIX-94-
100), calculate the catalyst-feed ratio, W /F, for each stage and the conversion-
selectivity relationships throughout. The density and void volume of the catalyst
bed are the same as those specified in problem 6 of Chapter XX.
15. The catalyst tubes of problem 13 are surrounded by circulating flue gas which
may be assumed to maintain the tube walls at a temperature of 1100°F. By a step-
wise graphical integration establish the radial and longitudinal temperature and con-
version distributions throughout the bed. Also calculate the average product dis-
tribution and pressure in the fluid stream as a function of longitudinal position,
assuming uniform flow distribution. The effective thermal conductivity of the bed
may be taken as 0.9.
16. It is proposed to place the catalyst of problem 7 in an envelope type'of heat
exchanger so arranged that the catalyst is in the form of vertical flat slabs 1 in. in
thickness. The retaining walls are maintained at 1075°F, the temperature of the
entering butane, by a circulating fluid. Using the space velocity and bed depth
finally determined in problem 7, calculate by a stepwise integration the lateral and
longitudinal distribution of temperature and conversion. Also calculate the longi-
tudinal variation of the average conversion and product distribution of the fluid
stream. The effective thermal conductivity of the bed may be taken as 0.9.
CHAPTER X X I I
UNCATALYZED HETEROGENEOUS REACTIONS
Many of the catalytic reactions discussed in Chapter XIX are
properly classified as heterogeneous, because more than one phase
participates. However, in such catalytic reactions the initial reactants
and final products are present in a single phase, and the net result is the
same as though a homogeneous reaction had occurred. Uncatalyzed
heterogeneous reactions are characterized by the presence of initial
reactants or final products in more than one phase. Such a reacting
system may include a gas phase and several distinct liquid or soKd
phases. For example, in the blast furnace, there may be present in one
zone a system of gas, molten metal, molten slag, soHd carbon, and several
solid oxides, all participating in a complex series of reactions.
Many organic processes involve heterogeneous reactions of various
types. The nitration and sulfonation of hydrocarbons are frequently
carried out in Mquid systems comprising two phases, one of which is pre-
dominantly acid and the other organic. The reaction products may be
distributed between both phases. Alternately, these same reactions
may be carried out in a homogeneous vapor phase or in a two-phase liquid
and vapor system. The principles goveriiing a reaction vary widely
with the conditions under which it is conducted, and there is little logic
in designating such a process as a " unit " for study on the basis of its
reactants or products. •
It is evident from the variety of types of heterogeneous reactions that
many different mechanisms are possible. The actual reaction may occur
in a homogeneous phase, on a solid surface, or at an interface separating
two liquid phases or a gas and a liquid phase. In any case, however, a
problem of mass transfer of reacting materials from one phase to another
phase or to an interface is involved. The rates of these mass-transfer
steps are governed by the principles discussed in Chapter XX. The
net rate of reaction is then determined by the rate of the chemical
change itself, by the rates of mass transfer, and in some cases by rates of
adsorption on solid surfaces. Thus, as in the case of heterogeneous cataly-
tic reactions, the rate is controlled by a series of steps, each of which may
be considered as offering resistance to the progress resulting from the
existing driving forces.
Two-phase uncatalyzed reactions are conveniently grouped in the
following classifications:
1051
1052 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. XXII
1. Gas-liquid reactions typified by gas-absorption operations in which

chemical change occurs. The absorption of carbon dioxide gas in caustic
solution is an example.
2. Liquid-liquid reactions such as the nitration of toluene with mixed
acid.
3. Gas-solid reactions in which either all products or all reactants are
gaseous. The oxidation of carbon is an example.
4. Liquid-solid reactions in which either all products or all reactants
are liquid as in the reaction of lime with hydrochloric acid. i
If more than two phases are present many additional types of reactions
are possible, of which the following are of particular interest:
5. Reacting systems composed of a gas and two soUd phases. The
low-temperature reduction of iron oxide with hydrogen or the absorption
of CO2 by lime are examples.
6. Reactions involving two liquid phases and a gas phase, as, for
example, when nitric acid reacts with mercury with evolution of gases.
7. Solid-soUd reactions typified by the sintering of powders. Some
compounds have been produced only by this type of reaction.
Many other classifications could be cited, each of which would involve
different combinations of the fundamental principles governing process
behavior. However, for any one classification these principles should
apply in a similar fashion, regardless of the particular compounds under
consideration. For this reason it is beheved desirable to study process
problems from the viewpoint of " unit principles " rather than " unit
processes " grouped together on the basis of similar chemistry.
Unfortunately at present there are sufficient data to permit sound
kinetic analyses of only a few reaction types. Rate measurements are
badly needed on other typical systems in order to segregate and evaluate
the individual controlhng principles. Of the seven types of reactions
listed here, data are available only for the partial analysis of systems
illustrating types 1, 2, and 3. Of these, type 1 is extensively treated
in the standard texts on unit operations and gas absorption. No satis-
factory fundamental method has; been developed for problems involving
chemical reaction in absorption., and at present such processes are
designed from empirical over-all transfer coefficients. More data- are
needed on which to base a sound analysis of such problems,
LIQUID-LIQUID REACTIONS
Reactions between two relatively immiscible liquids are favored by
intimate intermixing to produce a highly extended interfacial area in the
nature of an emulsion and to approach an equifibrium distribution of all
CHAP. XXll] HOMOGENEOUS REACTION IN SEPARATE PHASES 1053
components in the two phases. With such interminghng the over-all

reaction rate is influenced by the rates of difi'usion to the interface, reac-
tion at the interface, and homogeneous reactions in the two separate
phases. The resistance to diffusion can be made neghgible by effective
agitation which results in a very large interfacial area per unit volume.
Where two reactants are completely immiscible the chemical reaction
step may be confined to the interface; as mutual solubility is increased
the chemical steps become progressively more prominent in one or both
of the separate phases.
Homogeneous Reaction in Separate Phases. In a frequently en-
countered Kquid-liquid reaction, component A, which is highly soluble
in phase a, reacts with component B, which is highly soluble in phase b.
The two phases are of hmited miscibihty, and product R is highly soluble
in phase a while product ;S is a highly soluble in phase b. The reaction
may occur by a homogeneous mechanism in both phases. Equihbrium
distribution of all components in both phases may be accomplished by
effective agitation. For the general reaction A -\- Bv^R + S, the rate
in each phase is expressed by the following equations if a simple second-
order mechanism is followed:
Phase o: »•„ = fco I aôOsa ^ ^ ) (1)
Phase b: n = hlaAbasb ^^~) (^)
where a = activity
r = moles of A transformed per unit time per unit volume of an
individual phase
K = reaction equihbrium constant
Subscripts A,B,R,S refer to components A,B,R, and S
Subscripts a, b refer to phases a and b, respectively
If the agitation of the system is sufficient to maintain equilibrium dis-
tribution, the activity of each component in the phase in which it is less
soluble may be expressed in terms of its activity in the phase in which it is
more soluble. Thus, if i? is more soluble in the a phase and S more solu-
ble in the b phase,
[• (3)
ajtb = KRCRa', O-Sa = KsdSb J
where KA, KB, • • • = distribution equilibrium constants of A, B, etc.

Equation (3) may be substituted in (1) and (2) and each activity ex-
pressed as the product of the mole fraction and the corresponding activity
coefficient.
Ta = Ka{ XAayAaXBbyBbl^B — I (4)
, / „ xnayBaXsbyshKRK
n = Kb ixAayAo.XBbyBJ<-Ab — I (5)
The over-all rate of reaction is the sum of the reactions in the two phases.
Thus,
r = XAayAaVBbyBb(VakaKB + VbkbKA)
fVakJCs^VbkbKn\ ,„-
— XRayRaXsbysb I r= 1 — 1 (6)
where r — rate of reaction of A, moles per unit volume of total system

per unit time
Va, Vb = fractional volumes of a and b phases, respectively
Thus,
Va + Vb= 1.0
Application of Equation (6) is comphcated by the extreme variations
of the activity coefficients in liquid systems. In the general case it is
necessary to express empirically each activity coefficient as a function of
the composition of the phase to which it applies. Fortunately, in many
systems some of these variations are negligible, and approximate methods
may be used for the expression of others. A large over-all reaction
equiUbrium constant Ka or Kb simplifies the equation by eliminating the
terms representing the reverse reaction. It is also common for the rate
of reaction in one phase to be negligible in comparison to that in the other.
niustration 1. Nitration of Aromatics. The nitration of aromatic hydrocarbons
such as benzene, toluene, and xylene is an important heterogeneous Uquid-phase
reaction which has been extensively investigated. The process has been carried out
for the most part in batch-type reactors equipped with agitators and cooling surfaces.
The acid employed in such operations is a concentrated mixture of nitric and sulfuric
acids containing 30-40 mole per cent H2SO4 and 10-30 mole per cent H2O. It is
generally considered that the function of the sulfuric acid is to maintain the nitric
acid in a " dehydrated " state as a result of its strong affinity for water. If the nitra-
tion is carried out with strong nitric acid alone the rate of reaction is rapid at the start
but drops off quickly as the water formed by the reaction dilutes the acid. This
reduction in rate as the reaction proceeds is greatly lessened by the presence of sulfuric
acid or some other " dehydrator."
Rates of mononitration when mixed nitric and sulfuric acid are employed have
been measured by Lewis and Suen' on benzene and by McKinley and White^ on
1W. K. Lewis and T. J. Suen, Ind. Eng. Chem., 32,1095 (1940).
2 C. McKinley and R. R. White, Trans. Am. Inst. Chem. Engrs., 40, 143 (1944).
CHAP. XXII] HOMOGENEOUS REACTION IN SEPARATE PHASES 1055
toluene. Both groups of investigators made measurements in continuous equipment

under steady-state conditions with high degrees of agitation to insure distribution
equilibrium.
Thermodynamic and experimental data indicate that the equilibrium constant is
very large for the nitration of aromatics under liquid-phase conditions. Accordingly,
the reverse reaction may be neglected and the rate equation (6) written as follows if it
is assumed that agitation is sufficient to approximate equiUbrium distribution between
the phases:
r = kxA^yAMBbysbiVa + VtK') (7)
where k = kaKs
K> = kiKAlkoKa
XAa — mole fraction of HNO3 in acid phase
XBi = mole fraction of aromatic in organic phase
The over-all reaction velocity constant ki^ a, function of temperature which may be
expressed by the Arrhenius equation.
in* = - ; ^ + ^ (8)
In appljring Equation (7) the mutual solubiUty of the phases may be neglected and
XAa based on the total HNO3, H2SO4, and H2O content of the system while xsb is based
on the total benzene plus nitrobenzene.
The data of McKinley and White indicate that for the mononitration of toluene the
activity coefficient of the toluene 76 may be taken as unity over a wide range of con-
ditions. It was also found that K' = 0, indicating that the reaction taking place in
the organic phase is negUgible. With these terms evaluated, rate data at a constant
temperature may be used for empirical evaluation of the relative activity cofficients
of the nitric acid yAa-
The measurements of Lewis and Suen on the mononitration of benzene indicate
that the activity coefficient of the benzene jBb is increased by increased concentration
of both nitrobenzene in the organic phase and nitric acid in the acid phase. An inde-
pendent analysis of these data indicates that they are approximately represented by
the following empirical expression for ysb over a range of nitric acid concentrations
from 2 to 10 mole per cent.
7B!, = 1 -I- 62a;Aaa:s/l+*°^^''' (9)
where xsb = mole fraction of nitrobenzene in the organic phase
XAa = mole fraction of nitric acid in the acid phase
Equation (9) is of uncertain accuracy for high nitric acid concentrations.

Values of jAa, the relative activity coefficient of the nitric acid in the acid phase,
were evaluated from the measurements of both groups of investigators'- ' and are
plotted as a function of the acid-phase composition in Pig. 210. An arbitrary numeri-
cal scale has been assigned to these vMues. It should be noted that these coefficients
express the activity of the particular form of nitric acid which is effective in the
nitration reaction. This active form of nitric acid is beheved to be a " dehydrated "
or " pseudo " nitric acid, and the activity coefficients of this material do not neces-
sarily bear any simple relationship to activity coefficients of HNO3 as calculated from
vapor pressure measurements by the methods of Chapter XIV. Great reliance should
not be placed on the data of Fig. 210 in the regions where the broken Unes indicate
extrapolation of the experimental data.
With the assumption that the relative activity coefficients of Fig. 210 are ipde-
pendent of temperature, the other constants of Equations (7) and (8) for the nvono-
nitration of benzene and toluene in the temperature range from 15 to 50°C are as
follows:
Benzene Toluene
ysb 1 + Q2xAa{xsi.)^^+'^'''"> 1.0
K' 0.14 0
E cal per g-mole 14,000 14,000
A (rates in g-moles per hr
per liter of combined phases) 26.22 27.58
Per Cent Hz SO4

5 10 15 20 26 SO S5 40 46
#30
cf 25,
-^*20
10 15 20 25 30 35 40 45
Per Cent H2 SO4
FIG. 210. Activity Coefficients of Mixed Acids.
As previously pointed out, these data are uncertain because of the limited ranges
covered experimentally. If they are assumed to be correct the rates of processing
under various conditions may be mathematically evaluated and compared to deter-
mine the optimum procedure where distribution equilibrium is maintained. This
procedure is illustrated by the following problem.
It is desired to produce 10,000 lb per day of mononitrobenzene by nitration of
benzene with mixed acid. In a batch operation a mixed acid of the following compo-
sition is prepared by mixing cycle acid from a previous run with fresh acid:
Mok Per, Cent Weight Per Cent

HNO, 33.0 32.63
HjS04 39.0 69.61
HsO 28.0 7.86
100.0 100.0
CHAP. XXII] NITRATION OF BENZENE 1057
This'acid is mixed with benzene to provide 2.5 per cent excess HNGj above that
required to complete the formation of nitrobenzene. The densities of mixed nitrat-
mg acids are plotted in Fig, 211.
(a) Assuming that the acid and beêne are mixed together and maintained with
adequate agitation at 35°C, calculate the minimum time required to obtain 99.5 per
cent conversion to nitrobenzene.
(6) Calculate the volume of the reactor required for a capacity of 10,000 lb per day
of nitrobenzene, allowing 2 hr for charging and discharging the reactor,
HNO3 (1.502)
(1.00) (1.835)
From Pascel and Gamier, Bull, de la Society Ckemique, 25,145 (1919).

PIG. 211. Densities of Mixed Acids at 18°C. Compositions are given in
per cent by weight.
(c) Calculate the volume required in a single-stage continuous reactor designee
to produce the same conversion of benzene and the same waste acid composition as in
parts (a) and (6). It may be assumed that sufficient agitation is provided to insure
distribution equiUbrium.
From the definition of reaction rate in a batch process
VT
= -P
where na — moles of benzene per unit mass of original feed
(10)
V = total volume of mixture

The solution of Equation (10) requires a graphical integration with the aid of
Figs. 210 and 211. For progressive conversion the changing compositions of the acid
and benzene phases are calculated from a material balance, as summarized in Table A.
The corresponding values of 7x0 are obtained from Fig. 210. Values of TBI, are calcu-
lated from Equation (9). Reaction rates r are then obtained by substitution in
Equation (7) with the assumption that the term (F, + VtK') is constant at the
arithmetic average of the initial and final values. This assumption neglects the small
variations in volume accompanying the reaction as a result of changes in densities or
miscibility. Sample calculations of the initial and final rates are as follows:
Basis: 100 g-moles benzene chained requires 100 g-moles HNO3 or 6320 g
HNO3 charged = 1.025(6320) = 6478 g
fi4-7S
Weight of mixed acid - = 19,948 g
O.o25o
From Fig. 211, density = 1.75
19,900
Acid volume = — - — = 11.39 liters
1750
Density of benzene = 0.879
7800
Volume of benzene = = 8.87 liters
879
= 0.662
Total initial volume = 11.39 + 8.87 20.26
= 20.26 liters
Vt = 11.39 0.438
Initiallyfc(7a + K'Vb) = 29.08[0.562 + 0.14(0.438)] = 18.13
As the reaction proceeds the nitrobenzene formed dissolves in the benzene phase
and is but slightly soluble in the acid phase at 99.5 per cent conversion.
Vi = volume of benzene + volume of nitrobenzene
^^ (0.005) (7800) , (12,300) (0.995)
^' = m + ^5^ = '"-'^""''''
In acid phase:
H2S04 = (19,948) (0.596) = 11,900 g. 77.05% wt
H N 0 3 = 6478 - (99.5) (63.2) 190 g 1.23
H20 = (19,948) (0.0786) + 99.5(18) = 3,358 g 21.72
15,448 g 100.00
Prom Figure 211I,
Density of acid = 1.695
15448
Acid volume = = 9.11 liters
1695
y= •• 9.11 + 10.17= 19.28 Uters
9.11
= 0.473
19.28
0.527
k(.Va + K'Vi) = 29.0'8[0.473 +0.14(0.527)] = 15.90
18.13 + 15.901
Avo •KV. -i- TT'V.I = = 17 m
CHAP. XXII] NITRATION OF BENZENE 1059
The reaction rates are calculated by substituting this average value ol'
k(Va + K'Vb) in Equation (7).
Initially xsb = 1.0; XAa — 0.33; xsb = 0

From Fig. 210, since the mole fraction of H2SO4 is 0.39, jAa = 33
From Equation (9), jsb = I + (62)(0.33)(0)i+(«)(°=3) = i.O
From Equation (7), r = (17.01) (0.33) (33) (1)(1) = 185.3
Finally, XBb = 0.005; XA<, = 0.0097; xsb = 0.995 -•
From Fig. 210, corresponding to 39 mole per cent H2S04, 740 = 140
From Equation (9), ysb = 1.0 + (62)(0.0097)(0.995)i+«»W»»n = 1.60
From (7), r = (17.01) (0.0097) (140) (0.005) (1.60) = 0.184
TABLE A
NITRATION OF BENZENE
Basis: 100 g-moles benzene
r = (g-moles benzene)/(liter) (hr)
Per Cent
Con-
verted XBb XAa TAa a p. JBb r 1/r
100X51 Q2xAa ]I + 40XAa 1 + axsb
0 1.0 0.33 33 • 20.46 14.20 1.0 185.3 0.0054
5 0.95 0.3139 33.5 19.46 13.56 170 0.00588
10 0.90 0.2978 34 18.46 12.91 155 0.00645
15 0.85 0.2817 34.5 17.47 12.27 140.5 0.00711
20 0.80 0.2656 35.5 16.47 11.62 128.5 0.00778
25 0.75 0.2495 37 15.47 10.98 118 0.00848
30 0.70 0.2334 38.5 14.47 10.34 107 0.00935
35 0.65 0.2173 39.5 13.47 9.69 1.0 94.9 0.01054
'40 0.60 0.2012 40.1 12.47 9.05 1.003 82.4 0.01213
43 0.55 0.1851 41.5 11.48 8.40 1.017 73.0 0.01370
60 0.50 0.1690 42.8 10.48 7.76 1.048 61.5 0.01626
55 0.45 0.1529 44.4 9.48 7.12 1.135 69.0 0.01695
60 0.40 0.1368 47 8.48 6.47 1.311 67.4 0.01742
65 0.35 0.1207 49 7.48 5.83 1.607 56.5 0.01770
70 0.30 0.1046 51 6.49 5.18 2.02 65.0 0.01816
75 0.25 0.0885 56 5.49 4.54 2.49 52.5 0.01904
80 0.20 0.0724 57 4.49 3.90 2.88 40.4 0.0248
85 0.15. 0.0563 64 3.49 3.25 3.06 28.1 0.0356
90 0.10 0.0402 72 2.49 2.61 2.89 14.25 0.0702
95 0.05 0.0241 86 1.49 1.96 2.35 4.15 0.241
96 0.04 0.0209 90.5 1.30 1.84 2.21 2.84 0.352
97 0.03 0.0177 96 1.10 1.71 2.04 1.77 0.565
98 0.02 0.0145 110 0.90 1.58 1.872 1.016 0.985
99 0.01 0.0113 120 0.70 1.45 1.690 0.389 2.57
99.1 0.009 0.01094 125 0.678 1.438 1.669 0.35 2.86
99.2 0.008 0.0106 128 0.657 1.424 1.65 0.305 3.28
99.3 0.007 0.0103 132 0.639 1.412 1.633 0.265 3.77
99.4 0.006 0.0100 138 0.620 1.400' 1.614 0:.227 4.41
99.5 0.005 0.0097 140 0.601 1.388 1.597 0.184 5.44
1060 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. X X U
Similar calculations for intermediate conversions are summarized in Table A. The

resulting values of 1 jr are plotted in Fig. 212 against lOOxm, the moles of unconverted
benzene left at any time. Because of the wide range of the variables the curve is
plotted in three sections. The sum of the areas under these curves is the integral of
Equation (10). Thus,
7 T = 7.13 or T = (7.13)(60)/19.77 = 22.1 min
(6) Total batch-reactor volume required to produce 10,000 lb of nitrobenzene per
day, allowing 2 hr for filling and emptying.
Operating cycle = 2 + 22.1 /60 = 2.368 hr
Charges per day = 24 /2.368 = 10.1 "
Nitrobenzene per charge = 10,000/10.1 = 990 lb
4.5
4.0 0.08
8.5 0.07
I 3.0 |0.D6
6
«2.5 ^0.05
M ,
|o.04
\
V r = ] .69 lite -hi
6.> •t.0.03 \
\
1.0 0.02
0.5 0.01
1^T = 1.4 5 liter-1ir "5^
0.6 0.6 0.7 0.8 0.9 1.0 2

Vr,=i ,15 iter
3 4 5
-Ts6 7 8 9 10
0.00
10 20 30 40 60 60 70 80 90 100
n^—Wix^, G-Moles of Benzene Unconverted
(Basis: 100 g-moles of benzene charged)

Prom part (a) an initial reactor volume of 20.26 liters is required for 12,240 g
nitrobenzene
^T^ -A (990) (454) (20.26) „„ „ ,^
Volume required = —--: — = 26.3 cu ft
^ (12,240) (28.32)
(c) Reactor Volume for Continuous Operation. In a continuous operation the
average conditions in the reactor are the^same as those at the end of the batch opera-
tion. Under these conditions, from Table A, r = 0.184 g-moles/(hter)(hr) or
(0.184) (28.32)/(454) = 0.01145 lb-moles/(cu ft) (hr).
Hourly production =. (10,000) /(24) (123.1) = 3.38 lb-moIe3
Reactor volume = 3.38/0.01145 = 295 cu ft
Comparison of the results of parts (6) and (c) of Illustration 1 indicates
CHAP. XXII] CONTACTOR EFFICIENCY 106I
that under the specified conditions the reactor volume required for con-
tinuous operation is over ten times that for the batch operation. How-
ever, the cost per unit volume of continuous reactor is relatively low
because of the low maximum rate of release of heat of reaction. The
cooling system for the batch reactor must be designed to handle the high
rates of heat Uberation at the start of the cycle where the reaction rate per
unit volume is as much as 1000 times as great as in the continuous reac-
tor. The continuous reactoi- thus has the advantage of permitting pre-
cise control of temperature with only a small fraction of the cooling
surface required in the batch operation. In the actual operation of
batch equipment of thi^ type the high initial heat release is partially
controlled by gradual addition of the reactants to retard the rate of
reaction.
The large reactor volume required in a single-stage continuous operar
tion may be avoided by using multiple stages in counterflow. In such an
operation the hydrocarbon is charged to the first reactor where it is con-
tacted with partially spent acid from the second reactor. The efHuent
from the first reactor is passed through a separator from which the waste
acid is discarded and the hydrocarbon phase charged to the second con-
tactor. The total reactor volume is progressively reduced by this type
of multiple staging, and the optimum number of stages is determined by
an economic analysis. A phase separator is required for each stage. A
disadvantage of multiple-stage operation in the nitration of aromatics
is an increased tendency toward secondary reactions forming di- and
frinitro compounds.
Contactor Efficiency. Where agitation is not thorough, the distribu-
tion equilibria represented by Equation (3) may no.t be assumed. Thus,
the actual activity of A in the b phase is less than the equilibrium value,
KAdAa by an amount which is a function of the rate of reaction in the b
phase, the over-all transfer coefiicient of component A from the a to the b
phase, and the extent of the interfacial area. Since relatively high con-
centrations of A and R exist in the a phase it may be assumed that the
mass transfer of these components is controlled by a diffusional film at
the interface in the b phase. Then
ttAb = KAttAa — I (11)

« KAO^V
where n = molal rate of reaction of A per unit volume of b phase

Vb = fractional volume of b phase
kA — mass-transfer coefficient of A, moles per unit time per unit
activity difference per unit interfacial area
Uv = interfacial area per unit of total volume
An expression similar to Equation (11) may be written for anb and sub-
stituted in Equation (5).
or
nil + — 1- — \ = h[ KAttAaaBb j ; ) (13)
It may be noted that the right-hand side of Equation (13) is the same as
that of Equation (2). Accordingly, the rate of reaction in this phafee
may be written as
n = EA iaAaasbKA ^^—-) (14)
where Ecb is the contactor efficiency for the 6-phase reaction defined as
follows:
1
Ed, =
^^(e^+ax^)
A similar expression may be written for the contactor efficiency for the
a-phase reaction. The contactor efficiency is increased by improved agi-
tation which'increases the transfer coefficients and interfacial area and is
reduced by an increase in the reaction velocity constant or increases in the
activities of the nondiffusing reactants or products. For the empirical
evaluation of the contactor efficiency it is convenient to write Equa-
tion (15) in terms of diffusion film thickness 4>b and diffusion coefficients
DAV, and Dsm- Thus, since, in accordance with Equation (XX-26), page
981,
_, (a,A — aAi) , , V
TA = l>Am = KA (.a-A — aAi),
•Pi
it follows t h a t
, DAM
KA = ——
<p5
and
Ecb =
CHAP. XXII] REACTIONS WITH GASEOUS PRODUCTS 1063
where
DAM DRm = over-all diffusion coefficients of A and R from phase o to 6
<j}i = effective diffusional film thickness in phase b
ttv = interfacial area, per unit total volume
The reaction velocity constants and activity coefficients of Equa-
tion (16) may be evaluated from rate measurements under conditions of
extreme agitation where E^ = 1.0. Then, if the relative values of the
diffusion coefficients are estimated by the methods outhned in Chap-
ter XX, the ratio <^6/a„ may be evaluated as a function of the power
expended in agitation from a series of rate measurements at varying
degrees of agitation. If the rate of reaction is important in both phases
such tests must be repeated with different phase ratios in order to deter-
mine independently the contactor efficiencies for the two phases.
GAS-SOLID REACTIONS
Reactions between a gas and a solid may form either a gaseous or a
solid product or both. If a soHd product is formed three phases partici-
pate in the reaction and the mechanism is somewhat uncertain, but
probably involves the formation of an activated complex involving both
solid phases at a point of mutual contact between all three phases.
If only gaseous products are formed at least three mechanisms are
possible..
1. The solid may sublime and the actual reaction occur in the homo-
geneous vapor phase.
2. A sufficiently energized gaseous-reactant molecule reacts upon
impact with an active cefnter on the soHd surface to form a chemisorbed
product molecule which is then desorbed. Only one active center is
involved in this reaction, and if more than one product gas is formed
only one is chemisorbed while the others are evolved directly with the
gas phase.
3. The reactant gas is chemisorbed on the solid surface and then reacts
with an adjacent active center to form chemisorbed product molecules
which are then desorbed. Two adjacent active centers are involved in
this mechanism, and if two product molecules are formed each may be
chemisorbed on a separate site.
Reactions, with Gaseous Products. A common general case is repre-
sented by the following equation:
A{s) + B{g)^R{g) + S{g) (17)
Rate expressions which are developed for this general reaction are
readily modified to apply to the less complicated cases where the two
product molecules are alike or where "only a single product molecule is

formed.
Sublimation Mechanism. The sublimation mechanism is possible
even though the vapor pressure of the solid is very low. In such a case
the rate of reaction may be pontroUed either by the rate of the homo-
geneous reaction or by the rates of sublimation and condensation. If the
rate of the homogeneous reaction is controlling and mass-transfer effects
are not important, it may be assumed that the activity of the gasified
solid is constant at the equilibrium value and the rate expressed by the
equations of Chapter XVIII. In such a case the rate of reaction per
unit mass of solid is independent of the surface area exposed and pro-
portional to the volume of the gas phase.
If the rate of sublimation is controlling it may be assumed that
chemical equilibrium is maintained in the vapor phase. The rate of the
forward reaction per unit mass of solid is then expressed by
r = A„,{ksA — k'sAttAi) = A^ iksA — k'sA j ~ — ^ j (18)

where
Am = interfacial area, per unit mass of solid
ksA, ksA = velocity constants for the sublimation and condensation of A
aAi, asi = activities in the gas phase at the interface of components A,
B, etc.
Kg = gas-phase-reaction equilibrium constant
When equilibrium is reached between the solid and gas phases,
{aAi)e - ^ = KSA (19)

ksA
where KSA is the sublimation equilibrium constant of A which is equal to
its vapor pressure where the ideal-gas law is applicable. Combining
(18) and (19) gives
r = AmkAs(l - ^ ^ ) (20)
\ KsAKgaBi/
Data are not available to establish whether or not the sublimation
mechanism is followed in any of the industrially important reactions of
this type. It is possible that carbon and hydrogen are formed from
methane at high temperatures by the reverse of such a reaction, with
the rate of the homogeneous reaction controlling at the lower tempera-
tures and the rate of sublimation or condensation controlling at higher
temperatures. In considering such reactions, it is important to recog-
CHAP. XXII] REACTIONS WITH GASEOUS PRODUCTS 1065
nize that the activity of a very small aggregation of a solid is higher than
the normal value of an extended surface. For this reason a condition of
high supersaturation may exist in the gas phase if no solid phase is
initially present. Once condensation of the solid is initiated by the
formation of condensed nucleii these small particles will grow as the
reaction proceeds.
Absorption Mechanism. There is evidence that in many gas-soUd
reactions one or more of the gaseous components is chemisorbed on the
solid surface and the reaction actually occurs on the surface. Thus the
reaction of Equation (17) might proceed in the forward direction by
chemisorption of B on the surface of A followed by reaction to form
chemisorbed molecules of R and S which are then desorbed. Since the
rates of such activated phenomena differ widely it is general for one step
of a sequence of this type to be so slow that equilibrium may be assumed
in the others. The rate of the reaction is then controlled by the slow
step which may be any one of the four.
Where two adsorbed product molecules are formed it is probable that
the forward reaction occurs by an adsorbed B molecule reacting with
an adjacent vacant adsorption center of the sohd reactant A. The
reaction rate is then expressed by
r = A, ,(A-CBC{ — k'c'nc's) = AJc {C'BC'I —•] (21)
where
r = rate of reaction per unit mass of solid
Am = interfacial area per unit mass of solid
CB, CE, CS = concentration of adsorbed B, R, S molecules per
unit area
c^ = concentration of vacant reaction sites per unit area
k, k' = velocity constants of forward and reverse surface re-
actions, respectively
K' = surface-reaction equilibrium constant = CRCS/CBC'I
If the surface reaction is the slow rate-determining step surface con-

centrations in Equation (21) are the adsorption-equilibrium values
expressed by Equations (XIX-10 and 11), page 911. Combining these
equations yields
AîL'Yk
r = (1 -f- aaiKa + amKB -f- asiKs + anKiY
{a,,Ks-''-^^§P^) (22)
When equilibrium is reached,
(^-^ =§ ^ =K (23)
\ ttBi / KRKS
where K = the over-all thermpdynamic equilibrium constant.

A^L'^Ksh ( aBiasi\ .„,,

r =
(1 + QBiKB + amKn + asiKs + ai
where, KB, KB, etc. = adsorption-equilibrium constants.

I t is evident from Equation (24) that where this mechanism is followed
the rate is proportional to the interfacial area and to the number of
active centers for reaction per unit area. The relationship with the
activities of the gaseous reactants may be complex if the adsorption-
equilibrium constants are large, corresponding to extensive coverage of
the surface with adsorbed molecules. The rate may be diminished by
increased activity of the reactant B if its adsorption-equilibrium constant
is large.
If the adsorption of the reactant B is the rate-determining step the
surface concentration of B is fixed by the equilibrium of the surface
reaction and the other adsorptions. The rate of reaction is expressed
by a combination of Equations (XIX-6 and 7), page 910.
G^'^^'S)
r = AJCB loBiCi - ^ j (25)
where. JCB = adsorption velocity constant of B

KB = adsorption-equilibrium constant of B
From the equilibria of the surface reaction and the adsorption of R

and S:
/ CRCS ttBiCiKRasiCiKs amasiKBCi .„^.
CB = rp-, = zp-, = —- (26)
Kci Kci K
c'l = L' - [C'B + 4 + c'^ + c'j] (27)
Combining (26) and (27) results in

,
Cj =
£
1 + ^ ^ i ^ + UBiKn + asiKs + at^Kt (28)
K.
CHAP. X X I I ] REACTIONS WITH GASEOUS- PRODUCTS 1067
Substituting (28) and (26) in (25) gives

AJCBL' ( amasA ,_„.
( l + ' ' - ^ ^ + aniK^ + as^s + aîK^
It may be noted that aBi, the activity of B at the interface, appears only
once in Equation (29). Where this mechanism is followed the rate is
increased over the entire range by increased activity of the gaseous
reactant B.
If two like product molecules R are formed in the reaction, Equations
(24) and (29) are modified by omitting the term asiKs and replacing
aRittsi by {HR^^. If only a single product molecule R is formed all terms
pertaining to component S are omitted.
It is evident that where' chemisorption is involved in uncatalyzed
heterogeneous reactions the rate equations are of the same general forms
as those developed in Chapter XIX. By these same methods rate equa-
tions are readily derived for cases where the rate-controlling step is the
desorption of a product gas or where dissociation of one of the reactant or
product gases is involved. Similar equations may be derived for the case
where reaction occurs between a gas-phase molecule and a single active
center.
As in the case of catalytic reactions the proper mechanism and the rate-
determining step can be identified only by the analysis of rate measure-
ments over a sufficient range of conditions to test the applicability of the
various equations. With such data available many of the possible rate
equations may be eliminated merely by comparison of the trends of the
data with those of the equations. The final determination frequently
requires the evaluation of the constants in two or more equations in order
to determine which properly represents the data.
The constants of these rate equations can only be evaluated empirically
from rate measurements, and it has not been proved feasible to employ
adsorption-equilibrium constants determined by adsorption measure-
ments made on individual components. However, such adsorption
measurements may serve to indicate that some of the equifibrium con-
stants are negligible.
The similarity between rate equations for gas-solid reactions and gas
reactions catalyzed by solids is apparent. The distinction becomes evi-
dent upon appfication where in the gas-solid reactions the solid dis-
appears and the surface area per unit mass of solid changes as the reac-
tion proceeds, whgreas in catalytic reactions these quantities remain
constant.
1068 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. X X n
Mass-Transfer Effects. In the rate equations developed in the pre-

ceding section the activities are in all cases those existing at the inter-
faces. The interfacial activities of the gaseous components differ from
the activities in the main gas stream as a result of the gradients required
to cause diffusion to and from the interface. These activity differences
may be calculated from the mass-transfer relationships developed in
Chapter XX.
As in the case of catalytic reactions the importance of mass-transfer
effects varies widely. In many systems where the chemical rates are low
and mass velocities are high the mass-transfer resistances may be ne-
glected. Since mass-transfer rates are little affected by temperature it is
possible for the over-all rate of a reaction to be controlled entirely by
chemical steps at low temperatures and by mass transfer at high tem-
peratures where the chemical rates are relatively high. In such a case
the temperature coefficient of the over-all reaction rate diminishes as
temperature is increased, and where mass-transfer resistances are domi-
nant, temperature has little effect on rate of reaction.
Illustration 2. Extensive tests on the combustion in air of coke on grates' indicate
that at the high temperatures encountered the rate of the reaction is primarily limited
by mass transfer. Although the experimental errors in the work were large, the data
obtained are in general consistent with the following mechanism:
1. Oxygen diffuses from the air stream to the surface of the coke particle where it
reacts to form CO or CO2. It may be assumed that at the high temperatures involved
the chemical reactions on the surface are so fast that chemical equilibrium is main-
tained among C, CO, CO2, and O2. Reference to Fig. 156, page 712, shows that at
temperatures above 1300°K the equilibrium concentrations of both CO2 and O2 are
substantially zero.
2. Carbon monoxide diffuses from the interface into the gas stream where if oxygen
is present it is ojcidized to carbon dioxide in a homogeneous gas-phase reaction:(i
2C0 + O2 f i 2CO2
3. As the carbon dioxide content of the gas stream is built up by steps 1 and 2,
COj diffuses to the interface where it is reduced to CO in a-surfaoe reaction;
C02(g) + C ( s ) ^ 2 C 0
In a deep fuel bed this reaction continues after all oxygen is consumed, and the final
products approach an equiUbrium composition containing substantially no CO2.
On the basis of the assumption of equilibrium at the high-temperature interface,
the rates of steps 1 and 3 are controlled entirely by diffusion and may be calculated
from the data of Fig. 194. Step 2 is a homogeneous reaction which has been studied
by Haslam* with the conclusion that it approximates a third-order mechanism in
accordance with the stoichiometric equation. Measurements of spontaneous ignition
temperatures by Falk* indicate that the rate of reaction is increased by a factor of
5 Kreisinger, Ovitz, and Augustine, U. S. Bur. Mines Tech. Paper 137 (1916).
* R. T. Haslam, Ind. Eng. Chem., 15, 679 (1923).
s K. G. Falk, / . Am. Chem., Soc, 29, 1536 (1907).
CHAP. XXII] MASS-TRANSFER EFFECTS 1069
1.14 as a result of a 10°C increase in temperature at 1000°C. This corresponds to an

apparent energy of activation of approximately 42,000 cal /(g-mole) (°K). Since the
equilibrium constant (Fig. 156) corresponds to a negligible reverse reaction at tem-
peratures up to 2000°K the rate of this reaction may be expressed by the following
equation,
»• = fc(pco)^P02 (30)
where
r = rate of consumption of oxygen based on gas-phase volume, lb-moles/(cu
ft)(hr)
Pco, POi = partial pressures of CO and O2, respectively, atmospheres
Analysis of the data of Kreisinger, Ovitz and Augustine indicates that the reaction
velocity constant k is expressed by the following equation, which includes the energy
of activation derived from Falk's measurements,
42,000
]nk = ^ + 24.74414
or 9179.7
log fc = ~ + 10.7462 (31)
where T = degrees Kelvin
Because of the experimental errors inherent in combustion measurements, the data
of Kreisinger, Ovitz, and Augustine show inconsistencies when single runs are com-
pared with each other. A better picture of the combustion process is obtained by
considering the average results of a series of similar runs. In one series air was
passed at rates varying from 380 to 756 lb/(sq ft) (hr). The average rate for the
series was approximately 600 lb /(sq ft) (hr). The temperatures in the lower sections
of the fuel beds averaged approximately 2600°F. In each test coke which was sized
between 1|- and 1-in. screens was continuously fed to the top of the fuel bed, and the
grates were kept clean and free from all but a thin layer of ash. On the basis of
Equations (30) and (31) and the mass-transfer relationships of Chapter XX it is
desired to calculate the composition of the gas phase throughout the fuel bed in this
series of runs.
Solution: In order to calculate mass-transfer rates some assumptions are necessary
regarding the particle sizes in the bed. Actually the particle size diminishes from a
maximum at the top to a minimum at the bottom. However, as the size is reduced,
the particle becomes coated with ash which increases the resistance to mass transfer,
tending to offset the effect of reduced particle size. Accordingly, it will be assumed
that the particle sizS is constant at the size of the coke fed. If it is assumed that the
particles are approximately spherical and pack with a void space of 50 per cent, the
superficial surface area a per cu ft of bed, is calculated as follows:
/1.25V ^
Volume per particle = I "T^" J -^ ~ (0.59) (10"') cu ft
/1.25\2
Area per particle "= I "T^ J '^ ~ (3.4) (10~') sq ft
(0.50) (3.4) (10-2)

Area per cu ft = —(0 59)(10-°)— ^ ^^'^ ^^ ^* ^^^ "^ ^*
1070 UNCATALYZED HETE ROGENEOl IS REACTIONS [CHAP. XXII
TABLE A—
Basis: 1 sq ft of grate area;
Oxygen'. entering = 4.36 lb-moles per hr;
Section, in. from grate 0-0.25 0-0.25 0.25-0.5 0.5-0.75
Volume, Av, cu ft 0.0208 0.0208 0.0208 0.0208
Trial no. • 1 2 1 1
Est. avg. partial pressure, atm.
O2 0.200 0.205 0.195 0.176
COj 0.007 0.001 0.007 0.021
CO 0.006 0.007 0.016 0.0195
Reaction rates, lb-mole/(hr) (cu ft)
n 8.00 8.20 7.80 7.05
rt 0.80 1.11 5.65 7.45
J-j 0.25 0.04 0.25 0.76
n + rj 8.80 9.31 13.35 14.50
2r2 — Ti 1.35 2.18 10.85 14.14
2(ri + ra - n) 14,90 14.26 5.00 0.72
Incremental changes, lb-moles per hr
AO2 -0.183 -0.194 -0.278 -0.301 '
ACO2 0.028 0.043 0.226 0.294
AGO 0.310 0.297 0.104 0.015
Gas leaving section, lb-moles per br
Oj 4.167 4,156 3.878 3.577r
CO2 0.028 0.043 0.269 0.563
CO 0.310 0.297 0.401 0.416
Ns 16.350 16.350 16.350 16.350
Total 20.855 20.846 20.898 20.906
Composition of gas leaving, mole per cent
0, 20.00 20.00 18.52 17.1
CO2 0.13 0.20 1.29 2.69
CO • 1.49 1.42 1.92 1.99
In view of the approximations required with respect to area, it is satisfactory to

neglect variations in mass velocity, viscosity, density, and diffusion coefficients
throughout the bed and assume constant average values.
Mass velocity of air entering = 600 lb /(hr) (sq ft)
This mass velocity is 20.7 lb-moles / (hr) (sq ft), containing 4.35 lb-moles of oxygen.
Mass velocity of gas leaving with oxygen as CO = 600 + (4.35) (24) = 704
Average mass velocity = 652 lb/(hr)(sq ft)
652
Average gas density, p ~ o i'7\^QKfi\^Qnftn Mno\ ~ 0.0128 lb per cu ft
(20.7 +,2.17) (369) (3060/492)
Average molecular weight 652/22.87 = 28.5
' The average viscosity may be taken as approximately that of air which is given in
the International Critical Tables as 368.3 micropoises at 600°C and atmospheric
pressure. Based on the critical temperature of air as — 140.7°C, its reduced tempera-
COMBUSTION OP COKE IN AIB
air rate = 600 lb/(lir)(sq ft);

N2 entering = 16.35 lb-moles per hr
0.75-1.0 5.0-5.5 5.5-6.0 6.0-6.5 6.6-7.0 7.0-8.0 8.0-9.0 9-10
0.0208 0.0416 0.0416 0.0416 0.0416 0.0832 0.0832 0.0832
1 1 1 1 1 1 1 1
0.163 0.008 0.003 0.0013 ......

0.037 0.166 0.164 0.155 0.144 0.128 0.108 0.096
0.0205 0.059 0.070 0.089 0.110 0.140 0.166 0.184
6.53 0.32 0.12 0.05

7.60 3.09 1.63 1.14
1.34 6.02 5.95 5.61 5.21 4.65 3.90 3.46
14.13 3.41 1.83 1.19
13.86 0.16 -2.69 -3.33 -5.21 -4.65 -3.90 -3.46
0.54 6.50 8.88 9.04 10.42 9.30 7.80 6.92 .
-0.295 -0.142 -0.076 -0.050

0.291 0.006 -0.112 -0.138 -0.216 -0.386 -0.319 -0.282
0.011 0.270 0.370 0.375 0.435 0.772 0.637 0.665
3.282 0.122 0.046 -0.004

0.854 3.533 3.421 3.283 3.067 2.681 2.362 2.080
0.427 1.381 1.751 2.126 2.661 3.333 3.970 4.635
16.350 16.360 16.360 16.360 16.350 16.360 16.350 16.360
20.913 21.386 21.568 21.755 21.978 22.364 22.682 22.966
15.7 0.67 0.21

4.08 16.60 15.90 15.0 14.0 12.0 10.4 9.1
2.04 6.49 8.16 9.8 11.6 14.9 17.5 19.8
ture at 600'55 is 5.86, whereas the bed temperature of 2600°F corresponds to a reduced
temperature of 12.9. Extrapolating Fig. 175 yields
Viscosity at 2600°F = 358.3(2.9/1.8) = 577 micropoises

or . (2.42) (10-^) (577) = 0.14 lb/(ft)(hr)
The components whose diffusion limits the rate of combustion are Oj and COj.
Since nitrogen represents a large proportion of the gas stream, a good approximation
may be obtained by assuming that these components are each diffusing in an atmos-.
phere of nitrogen. On this basis the diffusion coefficients are calculated from Equa-
tion XX-50), page 988.
For oxygen, at 2600°F (1700°K):
(0.0043) (1700)?
Uv —
[(25.6)* + (31.2)^]22 ^ 3 2 ^ 2 8
= 2.07 sq cm per sec or (3.88) (2.07) = 8.02 sq ft per hr
For COj:
(0.0043) (1700)
.ip-^44^28
[(34)3 4- (31.2)3]"
= 1.77 sq cm per sec or (3.88) (1.77) = 6.9 sq ft per hr
In order to obtain a^Hdo from Fig. 194, page 987, the dimensionless groups DpG/ix
and nlpDv must be evaluated. On tlie assumption of spherical particles,
Dp = 1.25/12 =0.104 ft.
DpG (0.104) (652) ^^^
~ = 0.14 = ^^ C
For O2:
M 0.14
—— = = 1 36
pD„ (0.0128) (8.02)
For CO,:
M _ • 0.14 = 1.58
pD, ~ (0.0128) (6.9)
From Fig. 194 or Equations (XX 39 and 44),
For O2: a^Hdo = 15.7; Hda = 15.7/28.8 = 0.545 ft
For CO2: a^Hda = 17.3; Hda = 17.3/28.8 = 0.600 ft
The value of p/ for the diffusion may be calculated from Equation (XX-23),
page 979. Since the variation of (T + SPA) is small the arithmeticmean may be used.
It is assumed that PA is zero at the interface for both O2 and CO2 while in the main
gas stream it will vary from zero to 0.21. Since S — 1.0, p/ varies from 1.0 to 1.105.
An average value of 1.05 may be used.
By substituting in Equation (XX-48), page 986, the rates of diffusion of O2 and CO2
are obtained on the basis of 1 cu ft of fuel bed.
For O2 (step 1 of the combustion process):
_ ApG _ 652po2 _. lb-moles O2
'"' M„VfHdo ~ (28.5) (1.05) (0.545) " ^ ' ^"^ (cu ft) (hr)
For CO2 (step 3 of the combustion process):
,652pco2 „„ - lb-moles CO2
'"' ~ (28.5) (1.05) (0.600) ~ ' ^^"^ (cuft)(hr)
where the partial pressures are expressed in atmospheres.
The reaction velocity constant of the gas-phase oxidation of CO is given by Equa-
tion (31).
9179 7
log fc = - - — ^ -f 10.7462 = 5.3464
k = 0.222 (10'«) /
Since the bed is assumed to contain 60 per cent voids, the rate of step 2, the gas-phase
reaction, per cubic foot of bed is given by the equation:
r2 = 0.111 (10)Hpco)npo.) ^ ^ ^ ^ ^ ^
1 (cu ft) (hr)
On the basis of the assumptions established these three rate equations may be
applied to an incremental section of bed having a volume Aw to determine the rates of
disappearance of O2 and formation of CO and CO2. Thus:
Disappearance of oxygen = —AO2 = {n + ri)&.v lb-moles per hr

Formation of CO2 = ACO2 = (2r2 — r3)Ati lb-moles per hr
Formation of CO = ACO = 2(ri + T} — ri)Av lb-moles per hr
These three relationships may be applied successively to small increments of bed

depth, starting from the grate and developing curves relating the composition of the
gas stream to position in the/uel bed. The procedure is that of the typical stepwise
integration in which the rates in each increment are based on arithmetic average
28
24
*
Q I20
\
^ 2 C(3 2 /
;i2
o
8
4 UU
0
0 3 24 5 6 7 8 10
Distance from Grate, In.
FIG. 213. Composition of Gas in the Combustion of Coke on a Grate.
\
Air rate = 600 lb/(sq ft)(hr)
Average temperature = 2600°P.
partial pressures which are adjusted by successive approximations to correspond to

the calculated composition changes. These calculations are shown in Table A for bed
depths from 0 to 1.0 and from 5 to 10 in. The intervening calculations from 1 to 5 in.
were carried out by the same procedure. The resulting curves for the percentages of
O2, CO2, and CO are plotted as a function of bed depth in Fig. 213.
The curves of Fig. 213 are in fair agreement with the average experimental results'
with the exception of the CO concentrations near the grate. The experimentally
observed CO concentrations at a distance of 1.5 in. from the grate were sUghtly less
than 1 per cent, whereas the calculated value at this point is 2.2 per cent. These low
initial CO concentrations have been generally interpreted as indicating that the
primary combustion reaction is the formation of CO2 rather than CO at the surface.
However, it is believed that the differences between the observed CO values and those
calculated on the basis of the assumptions of Fig. 213 are within the probable error of
the measurements The problem of obtaining a representative sample of CO in the
presence of a large excess of oxygen at high temperatures is extremely difficult and it
would be expected that the CO concentrations would tend to be too low.
The assumption of chemical equilibrium at the interface in Illustration 2 would not

be satisfactory at low temperatures or when dealing 'with a particularly unreactive
type of coke. Under such conditions the rate of reduction of CO2 would have a
significant effect on the over-all rate, and it would be improper to assume an inter-
facial CO2 concentration of zero. As a result, higher maximum CO2 concentrations
would be reached in the gas stream. It seems probable that the rate of combination
of O2 with carbon can be neglected 9ver wide ranges of conditions, but under extreme
conditions it would also have to be considered, with a finite O2 concentration at the
interface. In such cases calculations of rates and composition changes would require •
evaluation of rate equations for the interfacial reactions. The general procedure
followed would be that developed in Chapter XIX for cases where both mass-transfer
and surface-reaction rates are of importance.
LIQUID-SOLID REACTIONS
Many intermittent chemical reactions are carried out in granular beds
where some component in the fluid phase reacts with a component in the
sohd phase and is held there until recovered or eliminated by regeneration
of the soKd. Such processes are intermittent, involving a progressive
change with time in both sohd and fluid at each cross section of the bed
and requiring regeneration when the capacity of the solid for further
reaction is exhausted or when the efiluent fluid begins to show insufiicient
recovery or change. One of the most common and familiar examples of
this type of process is in the softening of water by a granular bed of a
cation exchanger. In addition to the removal of calcium and magnesiiun
ions from water such exchangers have been developed for the removal of
iron from water for textile mills, hydrogen from water for beverage pur-
poses and in the recovery of nickel and copper from waste-refinery solu-
tions. The gradual poisoning of a catalyst by adsorption of some
impiu-ity may be considered in this same category. A pattern for the
mathematical treatment of the rate equations in this type of process
cannot be prescribed. Unique treatment is required for each different
case arising. Only in a few cases have successful mathematical devices
been developed for solving the unique differential equations estabUshed.
An illustration of this type of process is presented here for the removal of
minerals from water by a granular cation exchanger.
Water Softening. The softening of water is accomplished by the
removal of metallic ions from the water in exchange for sodivun ions from
the exchanger. The reverse reaction predominates in the regeneration of
the spent exchanger with sodium salt solution. The most plausible
theory of the rate of cation exchange consistent with experimental data is
that the instantaneous rate of softening at .constant temperature and con-
stant flow rate is proportional to the product of the concentration of cal-
cium ions in the water and the square of the concentration of available
CHAP. X X I I ] WATER SOFTENING 1075
sodium in the exchanger. The rate equations consistent with this theory
and experimental evidence are as follows:
dp
Forward reaction rate: ri = — = Jciuv^ (32)
9T
dp
Reverse reaction rate: r^ = = k2(yq)'^ (33)
or
Net reaction rate: r„ = kiuv'^ — kîyq)^ (34)
where p = calcium content of exchanger in equivalents per unit mass

V = sodium content of exchanger in equivalents per unit mass
q = sodium content of water, equivalents per million parts by
weight of water
u = calcium content of water, equivalents per million parts by
weight of water
ri = equivalents of calcium removed by exchanger/(unit mass)
(unit time)
ri — equivalents of calcium removed from exchanger/(unit mass)
(unit time)
Tn = net rate of softening, equivalents/(unit mass) (unit time)
y = activity coefficient for sodium ion in solution; at low con-
centration, 7 = 1
Du Domaine, Swain, and Hougen^ found that the reaction velocity

constants in the units previously defined were independent of fluid veloc-
ity but increased with decreasing particle size of exchanger as follows:
Mesh Size h
28-35 2.86
20-28 2.29
14-20 * 1.91
2-14 1.62
The equilibrium constant for the reaction, K = hi/hi was found to be

241,000. In the softening period the reverse reaction was found to be
negligible in a flow system where the water enters with no sodium
content.
In the softening period the variables u, v, p, and q are functions of dis-
tance X and time T. For an element of the bed, at depth x, dx in thick-
ness and unit area in cross section normal to the direction of water flow,
»J. Du Domaine, R. L. Swain and O. A. Hougen, Ind. Eng. Chem., 35, 546-54
(1943).
the following differential calcium balance can be set up for the interval
of time dr.
.PwVfdu
Gd- (— -—d^) = pdx[ — dr) + TTT T" "^ dr (35)
10« \\ dxdx JJ •'^ \dT \dr JJ W dT
The last term in this equation represents the change in calcimn content in
the water in the free space of the bed which is negligible compared to the
total amount of water passed through. Neglecting this term and simpli-
fying gives ^^
where p = mass of exchanger per unit total volume of bed, bulk density
Pw = density of water, mass per unit volume
G = mass velocity, mass per unit area per unit time
Vf = volume fraction of water space in bed
Since each gram equivalent of calcium ion added to the exchanger is
replaced by one gram equivalent of sodium,
p - Po - vo — V (37)
dv dp
or —= (38)
or OT
Combining Equation (32) with (36) and (38) gives

du
dx = auv
(39)
dv
buv^
where a = —10^ —;- and b = —fci

Gr
For the usual boundary conditions that the entering water has a constant
composition with time, and that the initial composition of the exchanger
is uniform, Swain* has developed the following solution to Equation (39):
ln2 + z = l n s + s — r
vo 1+2
(40)
i +i
u s
1
1+
CHAP. XXII] WATER SOFTENING 1077
•where
„ Wpkivh
r = -avlx = - ~
Cr
s = —but^VoT = kiUoVar (41)
2 = 1
V
A detailed study of the application of these equations to water soften-

ing shows the surprising results that, when hard water which enters free
of sodium flows through a bed of exchanger, the reverse rate is nearly
negligible under condition of steady fluid flow, and the exchanger be-
comes completely spent before chemical equilibrium is attained.
To expedite calculations, performance charts have been constructed
from Equation (40) expressing the effects of all variables in terms of the
dimensionless groups, r, s, s/r and U/UQ.
In Fig. 214 u/uo is plotted against s/r for constant values of r. In
Fig. 215, values of r are plotted against s for such conditions that the
residual hardness of the water leaving the bed is seven parts of calcium
carbonate per milHon parts of water.
Illustration 3. It is desired to estimate the residual hardness of water after

steady operation of a softener for 16 hr, with water flowing through a bed 3 ft thick at
a rate of 4 gal per sq ft per min. The initial hardness of the water corresponds to
300 parts calcium carbonate per million parts of water, which is equal to 6.0 lb equiva-
lents of calcium ion per million lb of water. The initial replaceable sodium content
of the exchanger is equivalent to 25,000 grains of calcium carbonate per cu ft. The
density of the mineral is 28 lb per cu ft, and the grain size is 14 to 20 mesh. Under
these conditions:
p = 28 lb per cu ft
X = 3.0 ft
ki = 1.91 per min
O = (4) (8.33) = 33.32 lb/(sq ft) (min)
Vo = I —'• )I 1 ( — 1 = 0.00255 lb equivalent Na per lb exchanger
\ 100 / \ 7 0 0 0 / \ 2 8 /
Uo = 6.0 lb equivalents Ca ion per million lb water
T = (16) (60) = 960 min
Prom Equation (41),
lO^pkivlx 108(28) (1.91) (0.00255)2(3)

31.3
G 33.32
s = kiUoVoT = (1.91) (6.0) (0.00255) (960) = 28.1
1078 UNCATALYZED HETEROGENEOUS REACTIONS ICHAP. XXII
CHAP. XXII] JDSE OF PERFORMANCE CHARTS 1079
From Equation (40),

l n 8 + z = In s + s - r = In 28.1 + 28.1 - 31.3 0,14
or z = 0.62
1+^
u s 1.0356
= 0.397
2.61
1 +
Hence u = (0.397) (6.00) = 2.38 lb equivalent of Ca''"+per million lb of water leaving
softener at the end of 16 hr
50
45
40 4^^
35
30
»-25
w
20 ^
4-^^^
15
10
5
0
0 10 15 20
25 30 35 40 45
s
FIG. 215. Performance Chart for Softening Water to a Residual Hardness of
7 Parts per Million.
In this manner, values of residual hardness can be calculated for any time, thick-
ness of bed, rate of water flow, cation-exchange capacity of exchanger, and reaction
velocity constant.
Use of Performance Charts. The preceding problem can be solved
directly from the performance chart. Since r — 31.3 and s = 28.1,
sjr = 0.898. The value of u/Uo corresponding to these values of r and
s/r is read from Fig. 214 as 0.36.
Illustration 4. In water softening the exchanger should be regenerated when the
effluent water reaches a hardness of seven parts per million of calcium carbonate.
Estimate the amount of water which can be treated in the system of Illustration 3
under this requirement.
From Fig. 215, when r = 31.3, and MO = 6.0, s = 24.6 or T = 24.6/0.0292 =
844 min (14.1 hr)
1080 UNCATALYZED .HETEROGENEOUS REACTIONS [CHAP. XXII
The water softened by 3 cu ft of mineral during this period is 844(4) = 3376 gal;
the calcium carbonate is
(3376)(8.34)(300-7) „ . . ,. ^ ^ ^
— = 8.24 lb CaCOs
or 19,200 grains of CaCOs are remoyed per cu ft of exchanger. The water softened by
1 cu ft of mineral under the aforementioned conditions is (4) (844) /3 = 1130 gal.
The reaction velocity constant of a cation exchanger can be obtained

from the commercial performance of a thick bed by use of Fig. 214. It is
necessary to start' with a fully regenerated exchanger and to know its
density, maximum cation-exchange capacity, and thickness. The
initial hardness MO of the water, rate of water-flow G, and exit hardness u,
at end of a given time T, must be measured.
Illustration 5. It is desired to estimate the reaction velocity constantfcof a cation**
exchanger. A bed 3 ft in thickness is used, having a density of 28 lb per cu ft and a
maximum cation-exchange capacity of 25,000 grains calcium carbonate per.cu ft.
Water having an initial hardness of 300 parts of calcium carbonate per million is passed
through the exchanger at a rate of 4 gal per sq ft per min for 14.1 hr when the water
leaving has a hardness of seven parts per million of CaCOj.
Mo = 6.0
Wo = 0.00255
G = 33.32
- = 0.0233
Mo
y = 846 min
s 2(33.32) (3.0) (846)
0.79
10H28) (0.00255) (3)
From Fig. 214, the*intersection of s/r = 0.79 with u/u^= 0.0233 gives an r value
of 32. '
rG: (32) (33.32) '
Thinj Jc, = •— = — — = T Q.=>
^^"^ "' lO^pvi, 10«(28)(0.00255)'(3) ^'^^
The mathematical treatment and performance chart of cation exchang-
ers can be extended to similar heterogeneous reactions in a batch process
where one atom or ion in the fluid phase reacts with two atoms or active
spots in the solid.
GAS ADSORPTION
The problem of estimating the changes in concentration in a fluid
stream and in a bed of adsorbent where a, process of a;dsorption proceeds
intermittently presents formidable mathematical difficulties which have
not been completely solved for the general case. For the special case
where adsorption takes place from a dilute solution and where the equilib-
CHAP. XXII] GAS ADSORPTION logl
rium adsorption is directly proportional to the concentration, Marshall

and Hougen' developed the following procedure.
The material balance over an elementary section of thickness dZ
measured in the direction of fluid flow and of unit cross-sectional area for
an elementary period of time dr may be written as follows;
Gydr=G[y + (^^ dZ^r + PB dZ ^ ^ ) dr
+ PGFe(~)dTdZ (44)
where
6 = mass velocity of adsorbate-free fluid, lb/(sq ft) (hr)
PB = bulk density of solid, lb per cu ft
pa = density of gas, lb per cu ft
Fe = external void fraction of bed
T = time, hr
w = adsorbate content of solid, lb per lb of solid (adsorbate-free
basis)
y = adsorbate content of fluid stream, lb per lb of fluid
Z = distance in bed measured in direction of flow, ft
The adsorbate is designated as component A.
The first term in Equation (44) represents the mass of component A
entering the section in time dr, the second term the mass leaving the
section, the third the change in the mass of component A present in the
solid portion of the elementary volume, and the fourth term the change
in the content of component A in the fluid portion of the elementary
volume all in the element of time dr.
Equation (44) reduces to
In a nonflow process where no fluid is being added or removed, the first

term of Equation (45) is zero, and the material balance becomes
In a steady-flow process the last term of Equation (45) becomes negli-

gible if the fluid content of the bed is negligible compared with the volume
' 0. A. Hougen and W. R. Marshall, Jr., Chem. Eng. Progress 43,197 (1^47). This
mathematical development was originally prepared by these authors in a report by
O. A. Hougen and F. W. Dodge entitled, "Drying of Gases," prepared for NDRC of
OSRD, June 30, 1941 and now declassified for distribution through the Office of
Technical Services, PB 2301, U.S. Department of Commerce (1946).
1082 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. XXU
of fluid passed through. In this case,
In gas adsorption where mass transfer through a gas film is the control-
ling factor, the rate of adsorption of component A can be expressed as
follows:
TA = KaaîpA - pf) (48)
where
TA = lb-moles A adsorbed/(cu ft) (hr)
PA = partial pressure of A in main gas stream, atmospheres
PA = partial pressure of A in equilibrium with the adsorbate
content of the solid, atmospheres
a„ = external area of solid particles, sq ft per cu ft
Kg = mass-transfer coefficient as defined by Equation (48)
For ideal-gas behavior, by Dalton's law,
_z_ = yp (49)
IT - p MA
where IT = total pressure
Ma = molecular weight of gas (adsorbate-f ree basis)
MA = molecular weight of adsorbate gas A
At low concentrations of A, Equation (49) becomes
p= - ^ (50)
Equation (48) may then be written for these conditions as '^
TA = Xca, —^ ir(2/ - / ) (51)

m.A
Equation (51) may also be written in terms of the height of transfer unit
Rio, Equation (XX-37), page 984. Thus, for dilute gases where p/ = ir.
The changes in adsorbate content of the adsorbent and of the gas may
be expressed in terms of the rate of adsorption given by Equation (52).
Thus, for" the solid adsorbent, expressing rate in pounds rather than in
pound moles,
^ (2/-/) (53)
H,do
CHAP. XXII] ADSORPTION WITH LINEAR EQUILIBRIUM 1083
Similarly, for the gas,
Equations (53) and (54) may be written as follows:
( g ) = ^(2/-2/*) (56)
and
(S)+^(S)=-°<-^ «
where
G
|3 =
paHdo
paPe
0 ™
1
a =
The development of Equations (55) and (56) is restricted to ideal-gas

behavior and low concentrations for which Equation (50) is satisfactory.
In the general case the temperature in the bed may vary and y* bears a
complex relationship to w. Under these conditions an analytical solu-
tion is not available, and it is necessary to resort to the tedious graphical
method discussed on page 1094.
Isothermal Adsorption with Linear Equilibrium. For the restricted
case of isothermal adsorption analytical integration of Equations (55)
and (56) is possible if it may be assumed that y*, the equihbrium adsorb-
ate content of the gas, is proportional to w the adsorbate content of the
soUd. This situation is approximated in many systems. Thus,
y* = cw + a (58)
where a = value of y* where the linear portion of the equilibrium
adsorbate content curve is extrapolated to a zero value
of w.
Where a is not zero the follo\ving mathematical development holds
strictly only for operation over the linear part of the equilibrium curve
and does not extend to zero adsorbate content of solid.
For the majority of adsorption systems of industrial importance the
absorbate content of the gas present in the bed at any instant is negligible
in comparison to the total'quantity of adsorbate passed through in an
operating cycle. Under these conditions the term yidy/dr) is negligi-
ble, and Equations (55) and (56) may be combined with (58) and written
as follows:
(") = '(!-») (-)

where
. b = 0c
-(I) aiy - cw) (60)
The solution for this pair of partial-differential equations was

developed by Hougen and Marshall' for the special condition where y is
constant at Z = 0 for all values of r, and w is zero at r = 0 for all values
otZ.
This solution is in terms of the four dimensionless groups, (y — y*)/
iVo - 2/*)> (w - w*)/{wo - w*), aZ and hr.
Thus,
* f*ciZ
y
2/0 -
- 2/0
*
2/0 J 0
e-«2 Joi2iVh^)d{aZ) (61)
w • * f,br \
- Wo
Wo — W* J ' e-*^ J"o(2i\4Z6^)d(6T) (62)
0
where yo = mass of adsorbate per unit mass of adsorbate-free gas in
the entering stream
Wo = mass of adsorbate per unit mass of adsorbent in equilibrium
with the entering gas
w* = adsorbate content of bed, lb per lb solid (adsorbate-free
basis) at T = 0. When the solid is free of adsorbate at
zero time, w* = 0
y* = equilibrium adsorbate content of the fluid stream, lb per lb
gas (adsorbate-free basis) at T = 0. When the solid is
adsorbate free at zero time, ?/* = 0
and Jo = Bessel function of the first kind.*
Equations (61) and (62) are similar in form to those developed by
Anzelius' for the analogous problem in heat transfer. For this latter
problem gchumann^" presented the solution in charts which were
extended by Furnas^^ to high ,values of 2iVhraZ. Special forms of
' I. S. and E. S. Sokolnikoff, " Higher Mathematics for Engineers," McGraw-Hill
Book Company.
» A. Anzelius, Z. Angew. Math. Mech., 6, 291 (*1926).
» T. E. S. Schumann, J. Franklin Inst., 208,405 (1929).
" C. C. Furnas, Trans. Am. Inst. Chem. Engrs., 24,142 (1930).
CHAP. XXll] HEIGHTS OF TRANSFER UNITS 1085
these charts appHcable to the special case of gas adsorption as well as to

heat transmission are shown in Figs. 216, 216a, and 217.
Figures 216 and 216a show the composition of the gas stream as it
progresses through the bed at different time intervals, and Fig. 217 shows
the corresponding fractional saturation of the bed under the same circum-
stances. In Figs. 216 and 216a values of 2//2/0 and in Fig. 217 values of
M/IWO are plotted against br for various values of aZ.
Heights of Transfer Units in Gas Adsorption. If mass transfer in the
gas film were the only resistance to adsorption the value of if d, the height
of mass-transfer unit, could be calculated from Equations (XX-39-47),
page 985. Actually, the over-all values of Ha and Hh are larger than given
by these equations, because the entire superficial surface may not be
available and the solid itself offers resistance to transfer.
From the experimental data of Ahlberg,^^ on the adsorption of water
vapor from air by silica gel, Hougen and MarshalP found the over-all
heights of mass- and heat-transfer units, Hdo and Hka could be expressed
by the following relationships.
a« \ M /
1 4Q / O G\+'>-^'-
av \ n /
These values are 3.57 times greater than the values calculated from the
gas film alone. In estabUshing Equations (63) and (64) account was
taken of the adiabatic operation of the bed with calculations of solid and
air temperatures from the heats of adsorption and the energy balance of
the system.
For Mlica gel the values of Dp and a„ given in Table LXII have been
estimated for various Tyler standard mesh sizes.
TABLE LXII
PROPERTIES OF SILICA GEL
Mesh Size D„ ft • sq ft per cu ft

2-4 0.0220 117
4-6 0.0128 202
6-8 0.00909 284
8-10 0.00634 407
10-12 0.00576 448
12-14 0.00446 576
14-20 0.00320 805
20-28 0.00226 J 1140
" J. E. Ahlberg, Ind. Eng. Chem., 31,988-92 (1939).
SuiuieniaJ uofq.DBai= «.
CHAP. XXII] HEIGHTS OF TRANSFER UNITS 1087
The equilibrium moisture content of a typical silica gel is given by the
following relationship:
where
w = 0.55
© or p* _ 1.82u;p, (65)
p = partial pressure of water vapor in gas in equilibrium with the

water adsorbed by the gel
Ps = vapor pressure of water at the existing temperature
0.2 0.4 0.6 0.8 1.0 1.2 1.4

bT
FIG. 216a. Fraction of Adsorbate Remaining in a Fluid Stream after Adsorption by
a Stationary Granular Bed.
(Low Time Range)
The ratio p*/ps is the relative humidity of the gas in equilibrium with the
gel. For the drying of air at atmospheric pressure, from Equation (50),
P*=V^=1.62/ (66)
O O i O O t - CO W3 CO (M 00500t-tD 1^ -^^ CO (M
• _: ^ ^ r - l O O O O O O o O
> =• =" = o o d d o o- d o* d § C5
ov
CKAP. XXII] DRYING OF AIR 1089
Combining Equations (65) and (OG) gives

., 1.82
where
c = 1.122pj,
For the special case of isothermal adsorption of water vapor on silica

gel, from Equations (57) and (63),
= 0.703aJ^j (68)
From Equations (57), (63), and (67),

Gc _ 1.122Gp, _ 0.789 Gpâ^ (D^G\-°-^^
i, = - - = - - — ^ = —
psHd PaHd pB
^ ^
m (69)
Illustration 6. Air is blown through a bed of initially dry silica gel, 1 ft in thick-
ness. The air enters at 80°F and 80 per cent relative humidity (J/O = 0.0179) at a
linear velocity of 100 ft per min based on the total cross section. The gel is 6-8 mesh
in size with a l)ulk density, pg of 39 lb per cu ft. It is assumed that isothermal con-
ditions are maintained by means of cooling coils in the bed. The following data are
desired:
(a) The humidity of the air leaving the bed at the end of 2 hr.
(b) The average humidity of the air leaving the bed over a period of 2 hr.
(c) The moisture distribution in the bed at the end of 2 hr.
(d) The average moisture content of the bed at the end of 2 hr.
Data for calculations:
Density of gel, p„ ~ 39 lb per cu ft
Density of air entering, pg = 0.0715 lb per cu ft
G = (100)(0.0715) = 7.15 lb/(sq ft)(min)
M = 74.5(10-') lb/(ft) (min)
From Table LXII,
Dp = 0.00909 ft
ffli> = 284 sq ft per cu ffc
//)pG\ _ (0.00909) (7.15)105
87.4
V M/ 74.5
p, (80°F) = 0.0345 atm

From Equation (68),
a = 0.703 (284) (0.1025) = 20.5/ft
From Equation (69),
6 = 0.789 ' ^ " ^ (284) (0.1025) = 0.145/min
(a) Humidity of air leaving at end of 2 hr:

hr = (0.145) (120) = 17.4
aZ = (20.5) (1) =20.5
From Fig. 216
- = 0.34
2/0
y = (0.94)(0.0179) = 0.0061
(6) To determine the average humidity of the air leaving the bod it is necessary to
integrate the outlet humidity over the 2-hr period, Values of y at different times are
calculated as in part (a) and tabulated as follows:
niin br 2/2/0 y
0 0 0 0
20 2.9 0 0
40 6.8 0.003 0.00005
60 8.74 0.016 0.00029
80 11.6 0.065 0.00116
100 14.5 0.185 0.00331
120 17.4 0.340 0.0061
Graphical integration by plotting y against T gives the average value of y as 0.0010

(Fig. 218).
(c) The moisture distribution in the bed at the end of 2 hr is determined from
Fig. 217.
T = 120 min, br = 17.4, yo = 0.0179
From Equation (65),M)o = 0.55(0.80) = 0.44
Values of w Iwo corresponding to selected values of az are read from Fig. 217 along a
constant abscissa of br = 17.4 and
tabulated as follows:
0.006 aZ w/wo
- 0 0 1.00 0.44
•i! 0.005 0.2 4.1 1.00 0.44
< 0.4 8.2 0.92 0.405
^0.004 0.6 12.3 0.78 0.344
-3 0.8 10.4 0.57 0.25
'E 0.003 1.0 20.5 0.30 0.132
s
W The moisture distribution in the bed
is shown graphically in Fig. 218a
;» 0.002 by plotting w against Z.
(d) The average moisture content
0.001 of bed at end of 2 hr is obtained by
a graphical integration of the rela-
20 40 ]B0 80 100 120 tionship between w and Z from part
•r=time in minutes (c) and found to be 0.348
FIG. 218. Humidity of Air Leaving a Bed of For the adsorption of gases in
Silica Gel, 1 Foot Thick. highly dilute mixtures, nearly iso-
thermal conditions prevail, and the
assumptions made in Illustration 6 are satisfactorily accurate for many purposes.
Ik
CHAP. .XXII] ADgORPf ION OF BENZENE 1091
Illustration 7. Benzene vapor is to be adsorbed .from air in a solvent recovery

plant by the air being passed downward through a bed of silica gel at 70°F and atmos-
pheric pressure at a mass velocity (solvent-free) of 7.5 lb / (sq ft) (min). It is desired
to operate for an adsorption period of 60 min with a minimum benzene recovery from
the air of 9D per cent. The entering, air contains 0.90 per cent benzene by volume.
The siUca gel is 4-6 mesh in size, having an external void fraction of 50 per cent and a
bulk density of 39 lb per cu ft. For adsorption up to 20 per cent by weight (adsorbate-
free basis) the weight of benzene adsorbed is directly proportional to the partial
pressure of the benzene in the air in accordance with the equation w = l.&2p*Ip,.
The vapor pressure of benzene p« at 70°F is 0.125 atm.
(a) Calculate the depth of bed required.
(b) Calculate the concentrations of benzene in the silica gel at the top, middle, and
bottom of the bed at the end of 60 min.
(a) Depth of bed required:
From Table LXII, £)p = 0.0128 ft.;
o, = 202
n = 74.5(10-5) lb/(ft) (min)
G = 7.51b/(sqft) (min)
DpG. (0.0128) (7.5) (10=^)
= 129.
74.5
cf)'
From Equation (63),
0.0825
0.4 0.6 0.8
H.
1
= 0.703O,
m
= (0.703) (202) (0.0825) = 11.8
Bed thickness
FIG. 218a. Moisture Distribution in a
Bed of Silica Gel at the End of 2 Hr.
In order to evaluate 6 it is necessary to calculate c of Equations (58) and (59). Since

tromEquation (50) and since t« = 1.67p*/p., andp* = 2/*x(29/78), = wpjl.67
y* p. (78)
= 1.60p./^
w 7r(29)(1.67)
Then, from Equations (57) and (59):
1.60g?.' ^ (1.60) (7.5) (0.125)
0.455/min
~ PBH^GT (39)(l/n.8)a)
br = (0.455) (60) = 27.2
y
= 0.10 (90% removal)
2/0
From Fig. 216, where 6T = 27.2 and 2//yo = 0.10; aZ = 38

aZ 38 „ „„ ,
Z =— = —- = 3.22 ft
a 11.8
(b) Concentration of benzene in silica gel at top, middle, and bottom of bed:
pi 1.67(0.009)
"'" = ^ • ' ^ ^ 7 . " 0.125 - ° - ' ^ °
26 r y
9R
/ / I
9A
. /
23 / 1
22 /1
21 / /'1
90 / /
/ / 1 1/
1R / /
17 - ^ !_ J 1
Ifl 1 '1
L lA
IK _ 1
'
•o
//
1 1 1I
O AO
11
'î'^1
/ 1 *
1 1k
/ r<
\ J~
Q
Q 4 -> 7 • ^j
^ / /
7
A ^ i / JT
R /
A -
/ / y '^
s ^y'^y%< ^' >
% î^- ^
9 9^^ = : ^
'
1 -
0 ^ _...
2 3 4 5 6 8 10 20 80 40 60 100 200 800 500 1000
aZ
FIG. 219. Rate of Composition Change of a Flujd ^Stream Leaving a Bed of a
Granular Adsorbent.
CHAP. XXII] EXPERIMENTAL EVALUATION OF CONSTANTS 1093
From part (o), a = 11.8; 6T = 27.2
Top Middle Bottom
zaZ 0 1.61 3.22
0 19.0 38.0
w/wo (from Rg. 217) 1.0 0.83 0.079
w 0.120 0.099 0.0095
Experimental Evaluation of Constants a and b. Where uncertainty

prevails in the evaluation or calculation of the height of a transfer unit,
values of a and b in Figs. 216 and 217 can be obtained directly from
experimental data by measuring the adsorbate content of the effluent gas
from a given bed over an extended period of time, provided the bed is
S.0
2.6
2.0
M
.a 4^
,^L0 at
0.6 L0 QI
0 6 6 8 10
-aZ-
Fia. 220. Rate of Composition Change of a Fluid Stream Leaving a Bed of a
Grajiular Adsorbent (At Low Values of aZ).
initially free of adsorbate, nearly isothermal conditions are maintained,

and the gas is supphed at a steady rate and fixed composition. By
graphical differentiation of Fig. 216 the values of (d log y)/id log hr)
were obtained and plotted in Figs. 219 and 220 as functions of ax at
different values of y/yo.
In establishing values of a and b from experimental data a plot of y
against T is made on logarithmic paper, and values of idlogy)/{d\ogbT)
are obtained at values of y/yo in the experimental range by graphical
differentiation. From Fig. 219 or 220 the corresponding value of ax is
determined and the value of a calculated from the known thickness of the
bed. From the known values of y/yo and aZ, the values of br are
obtained from Fig. 216, and from known values of time the values of h
are calculated. From the known values of a„, and DpG/n the constant in
Equation (63) can be evaluated for the given experimental system. By
employment of this equation values of a and h can be calculated for

other conditions of operation from Equations (57) and (59).
Illustration 8. Air containing 0.58 mole per cent A is passed through bed of silica a
gel kept at 25°C at a rate of 44.8 standard cu ft /(minute) (sq ft). The gel is 10-14
mesh and has a density of 44 lb /(cu ft). The bed is 2.0 ft deep. From measurements
of the exit concentration of A at *-arious intervals of time the values of time were
obtained for given values of y jys,, as shown in the first two columns of Table A. The
values of {d log y) j{d log T) were obtained from a plot of y ly^ against T on logarithmic
paper. Values of aZ were obtained from Fig. 219 and a calculated from the thickness
of the bed. From the values of y /ya and aZ, values of 6T were obtained from Fig. 216
and values of 6 calculated. These values are summarized in the following table:
TABLE A
ADSOKPTION OF A BY SIUCA GEL
y__ Time, dhgy

yo min d log T aZ a, ft IT h, min
0.1 11.2 1.80 6.0 3.0 1.9 0.170
0.3 21.4 1.66 7.1 3.65 4.6 0.215
0.5 31.2 1.32 7.4 3.7 6.8 0.218
0.7 42.4 0.86 6.0 3.0 7.4 0.175
Avg = 3.4 Avg = 0.195
Isothermal Adsorption with Complex Equilibrium. Where the rela-

tionship between the equilibrium adsorbate content of the adsorbent and
that of the gas is not linear, or where the gas in the void of the bed is not
negligible in comparison with the amount flowing through, the assump-
tions on which Equations (61) and (62) and Figs. 216-220 are based may
lead to serious errors. In the former case it is necessary to resort to
graphical integration of Equations (55) and (56).
By a change in variables the form of Equations (55) and (56) can be
simplified to eliminate all constant terms and one partial derivative.
The time variable r is modified to allow for the time elapse from entrance
to a position Z in the bed. As the front of a gas stream which enters the
bed moves to a position Z units distant from the inlet, the time increased
by the time required to reach that position. This time elapse can be
calculated from the average linear velocity through the voids of the bed.
Since the average fraction of free cross-sectional area is equal to the
fraction of void volume, the average linear velocity is equal to G/paFe,
and the time elapsed in advancing from the entrance to position Z is
PaFeZ/G or yZ. The modified time variable then becomes T — yZ. By
this expedient one partial derivative and the constant y are eliminated.
The remaining constants a and P can then be eliminated by defining the
CHAP. XXII] ADSORPTION WITH COMPLEX EQUILIBRIUM 1095
new variables thus:

u = aZ (70)
v = Kr- yZ) (71)

From Equation (70) it is evident that, when Z is zero, u is zero, and, from
Equation (71), when T = yZ, v is equal to zero.
Since w and y are functions of Z and T, they are also functions of u
and V and from Equation (XI-31):
»dw = I — ) du + l-— ) dv (72)

\du/.o \dv/u
dy^CAdu + Mdv (73)

Then
/dw\ /dw\ /du\ /dw\ /dv\ (74)
W / z "" \du), XdrJz \d^)u W z
(dy\ _(dy\ (du\ ,(dy\ /dv\
(76)
\dz), W . WA \dv)\dz).
(dy\ _• /dy\ /du\ /dy\ /dv\
(76)
From Equation (70):
^ m.-" • (77)
(78)
(JX-O
From Equation (71):
/ '\ \
(11=-^ (79)
(a- (80)
By substituting Equations (80), (78), and (55) into (74), there results
(S).=<-''-' (81)
1096 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. XXIl
By substituting Equations (79), (80) into (75) and (76), and combining
with Equation (56), there results
(|^)__ = -(2/-2/*) (82)
A graphical method of double stepwise integration of Equations (81)

and (82) has been developed after the pattern of Grossman.'' Thus in
terms of finite increments,
(g). = -&.-»:> • w
where mean values are taken over the interval AM and
(^:)^ = (.:-.r, - m
where mean values are taken over the interval Av. The relationship
between y* and w is obtained from isothermal equilibrium-adsorption
measurements which may be expressed graphically for any system.
The graphical solution for isothermal adsorption follows in three stejis.
The adsorbate content of the gas stream as a function of time must he
known at Z = 0 and hence at M = 0. The adsorbate content of the bed
must be known when T = 0. Since the initial adsorbate content of the
bed at any location is unchanged during the lapse of time yZ, the time
required in passing to location Z, it follows from Equation (72) that the
adsorbate content of the bed is the same at t; = 0 as when r = 0.
The first step of the solution is to establish the adsorbate content of the
gas at zero value of i; as a function of M; the second step is to establish the
adsorbate content of the bed at zero value of u, as a function of i;. In
the third step all other values of y and w are obtained for successive incre-
ments of Aw and Ai;.
In designating values of w, y, and y* at different intervals of dis-
tance and periods of time double subscripts are used, the first subscript
designates increments of u and the second increments of v. Thus, the
symbol 2/01 indicates the value of y at 0 AM and 1 Au, the symbol w^
indicates the value of lu at 1 Aw and 2 A?;.
The general objective is shown in Fig. 221, where values of y, if', and w
are plotted against increments of w for different intervals of v. The
mean values over the interval Aw are represented as y^, J/^ and Wm, and
" L. V. Grossman, Trans. Am. Inst. Chem. Engrs., 42, 535, (1946).
CHAP. XXII] ADSORPTION WITH COMPLEX EQUILIBRIUM 1097
the mean values over the interval Au by the period values y^, y*^ and w'„.
The circled points represent the initial given conditions.
Ste'p I. In Fig. 222, the initial distribution of A in the bed is known
and plotted against Au. The corresponding values of y* are obtained
from adsorption-equilibrium data. The composition of the entering gas
stream is known and designated as 2/00 and in this case is taken as con-
stant with y at w = 0. From the value of Wm at |Aw, the corresponding
value of y* is obtained from adsorption-equilibrium data. From this
point a horizontal line is projected of unit value in u to point B. For
example, if Au = 0.2 the length of the horizontal line is 5AM. A straight
line is then drawn connecting 2/00 with B. The intersection of this line
Given
Stepn
OAu Att lAu 2Au

0 lAu 2Au 2
FIG. 221. Calculation of Concen- FIG. 222. Step I. Calculation of
tration Gradients in Adsorption y at Zero Value of v.
of Gases by a Granular Bed.
with the lAw line gives the desired value of 2/10, and y^ is its intersection
with the AM/2 vertical line. The validity of this procedure is evident
from the geometry of Fig. 222, where, from the relationship of similar
triangles,
2/00 — 2/10 Ay
(85)
Au 1 Au
in agreement with Equations (83).
Successive values of 2/20, ^so. • . are obtained in a similar manner with
successive increments of Au.
For the special case where the solid is initially adsorbate-free,2/* = 0
at t; = 0, and Equation (82) is integrated to give
In - = M (at t) = 0) (86)
y
%
Step II. In step II it is desired to establish the adsorbate content of

the solid at the entrance of the bed (u = 0) for various values of v.
The adsorbate content of the entering gas is plotted against v in Fig. 223.
In this case a constant value of y is assumed at u = 0 for all values of v.
The value of Woo is known and plotted. A mean value y'„ is obtained
at Au/2. A horizontal line of length equal to unity in v is extended
horizontally from y'„ to C. Parallel lines are drawn through C and woo
such that at Ay/2, w'„ and yt! a^re in equilibrium. The intersection of the
w line with the Ay vertical line gives the desired value of Woi. The valid-
1
J <— Unity m » -
'"oj
1
< /c
Way
1/ "''my/
y'Jl
0A» Av_ lAs 2A»

2
FIG. 223. Step II. Calculation of FIGS. 224-225. Step III. Evalu-
10 at Zero Value of u. ation of y and w Except at
Zero Values of u and v.
ity of this procedure is evident from the geometry of Fig. 223, where,
from the relations of similar triangles,
Woi — Woo t/„, Aw

(87)
Ay Au
in agreement with Equation (84). Succeeding values of w at zero values

of u are obtained in the same manner for successive increments of Ay.
In the usual case where the gas enters the bed at constant composition
and where equilibrium adsorption can be expressed by an equation,
step II can be solved analytically. At u = 0, y = ya, and where
y* — mw", Equation (81) becomes
(where « = 0) (88)
0 yo — mW^
A graphical integration is required where n is not an integer.

CHAP. XXIIl UNSTEADY HEAT TRANSMISSION 1099
Step III. With the estabUshment of the boundary conditions for both
gas and solid the concentration of both gas and solid for succeeding incre-
ments not at zero values of either M or v is obtained by a combination of
the methods illustrated, in Figs 222, and 223, and shown in Figs. 224,
and 225. For example, the next point to establish is Wn and yn.
On Fig. 224, values of yoi and Wai and on Fig. 225 values of yw and Wio
are located from previous steps I and II, respectively. The values of yn
and Wn are established on each chart by the respective methods used in
Figs. 221 and 222. Adjustments of the two figures must be made
according to their geometry until points yn and Wn agree on both charts.
A mechanical device fulfilling the geometry of the two figures may be
used to adjust the two figures readily. It is required in Fig. 224 that the
linkage of lines pivots on yoi and Woi, that Wm and ?/* are in equilibrium
and are free to move along the vertical Aw/2, and that line y*B is hori-
zontal and of unit length in u. Points yn and Wn are free to move along
the vertical IAM. It is required in Fig. 225 that the linkage of lines
pivots on 2/10 and Wio, that w^ and yl, are in equilibrium and are free to
move along the vertical Ay/2, that Une yl,C is horizontal and of unit
length in v and that Unes Wio, Wn, and y*J,C are parallel.
It should be noted that this graphical method was developed for
dimensionless time and distance parameters. However, the solution
arrived at for a specific system and boundary conditions is a general
solution for other variable conditions of operation.
Unsteady Heat Transmission in Granular Beds. As originally
developed by Schumann^" and Furnas" charts similar to Figs. 216 and
217 were used for calculating temperature changes in both gas stream
and bed where heat was transferred between a flowing gas and a station-
ary granular bed in the absence of heat effects due to chemical reactions
and where constant values of heat-transfer coefficients and thermal
properties could be assumed. To use these charts for heat transmission
the following replacements of symbols should be made.
Substitute, t — To y
to — To yo
T -To w_
to — To Wo
(89)
Ua. / 1
—- = — for a
UpLr rlho
where t = temperature of gas

to = temperature of gas entering
T = temperature of solid
To = initial temperature of solid ^
U = over-all heat transmission coefEcient, per unit area
Hho = over-all height of heat-transfer unit
Cp = specific heat of gas
Cs = specific heat of solid
Where the thermal resistance of the solid is negligible, compared with

the gas, and where the total superficial area of the granular solid is
available for heat transfer, the over-all heat-transmission coefEcient is
equal to that of the gas film, U = ha. The heat-transmission coefficient
ho of the gas film can be estimated from the Equations (XX-40, 45, 47),
pages 985.
NONISOTHERMAL REACTIONS IN THE UNSTEADY STATE

Grossman^' has developed a graphical method of establishing the
unsteady-state time-temperature gradients in a granular bed through
•which a reacting fluid is flowing. This method is applicable to general
heterogeneous reactions involving a fixed bed of solid reactant and to
such problems as the temperature fluctuations in a checkerbrick or
pebble-bed regenerator, to the period required for heating or cooling a
reactor to a steady state, or to the burning of the carbonaceous deposit in
the regeneration of a catalyst. Adiabatic conditions and a uniform
velocity throughout any cross section of the bed are postulated. On this
basis temperature-time position relationships may be developed to corre-
spond to any specified initial temperature distribution of the solid and
time-temperature relation for the entering fluid. The effects of varia-
tions in the thermal properties of the solid and fluid are readily taken
into account.
In the development given herewith the chemical reaction is assumed
to take place on the surface of the solid so that for an exothermic reaction
the temperature of the soUd is greater than that of the fluid. A differen-
tial energy balance is set up for the transfer of heat from the surface of the
soUd to the fluid in an element of the bed of unit cross-sectional area and
of thickness dZ. For thefluid,the temperature changes are expressed by
' CoPoF, ( ^ ) ^ + (C«GO ( g ) = habits - g (90)

CHAP. XXUl NONISOTHERMAL REACTIONS HQl
Similarly, for the solid,

/dts\
CsPb I — ) = ~hav{ta -Q+r AHn (91)
where
Cg = specific heat of fluid
P(5 = density of fluid
tg = temperature of fluid
is = temperature of solid
Cs = specific heat of solid
Z = distance irom the inlet of the bed
h = heat-transmission coefficient per unit area of solid
CB = external area of solid per unit volume
G = mass velocity of fluid stream, mass/(area) (time)
r = rate of reaction, adsorption, or desorption, moles of A con-
verted, adsorbed, or desorbed per unit time per unit mass
of solid
AH = heat of reaction per mole of reactant A
Pb = bulk density of solid bed
Ft = external void fraction of solid bed
Let
httv J , hUv G
Cspb CQPBF, HHPQF,
where H^ = height of a heat-transfer unit, Equation (XX-38), page 984.

Then Equations (90) and (91) become
ana©.-*-'-'
O.=-«'»-«+ rAH
-F- (94)
Equations (93) and (94) may be simplified by transforming them in

terms of two new variables u' and v' to replace Z and T. By the pro-
cedure followed in the development of Equations (81) and (82) it is
found that these variables are properly defined as follows:
W = — ^ =— (95)
G Eh
r/-k.{r-'-fz) (56)
In terms of these variables Equations (93) and (94) reduce to
For finite increments Equations (97) and (98) become
where B„ = ( ' " ^ f g ^ ' ) ^ (lOD
Reaction Rates Independent of Conversion. Equations (99) and

(100) can be solved by a stepwise graphical procedure similar to that
discussed for Equations (83) and (84) in the restricted case where both
the rate of reaction r and the heat of reaction LH can be assumed to be
independent of conversion and functions only of temperature. Such an
assumption with respect to r is justified only in the case of a fluid stream
carrying a large excess of reactants or where as a result of the reaction
mechanism the rate is substantially independent of composition. In
such a case the solution is carried out in three steps similar to those of
Figs. 221-225. The initial temperature distribution in the soUd (at
V = 0) is known. The temperature of the entering gas stream as a func-
tion of time is also known. The heat of reaction and the rate of reaction
are known as functions of temperature. The first step is to establish the
temperature of the gas at zero value of v' as a function of u', the second
step is to estabUsh the temperature of the bed as a function of u' at zero _
value of u'. In the third step other values of temperature are obtained
for successive increments of AM' and Av'.
In designating values of IQ and ts at different intervals of distance and
periods of time double subscripts are used; the first subscript designates
increments of v! and the second increments of v'. Thus, the symbol
<Goi indicates the temperature of thci gas at 0 AM' and 1 At;', the symbol
<si2 indicates the temperature of the solid at 1 AM' and 2 AD'.
The general objective of the calculations is shown in Fig. 226 where
ts and to are plotted against increments of u' for different time intervals
v'. The mean values over the interval Aw' are represented as tg^ and
CHAP. XXII] RATES INDEPENDENT OP CONVERSION 1103
is and the mean values over the interval Av' as <« and ti . The
circled points represent the initial given conditions.
Step I. In Fig. 227 the initial known temperature of the bed'is
plotted against AM'. The temperature of the entering gas stream is
known and designated as taoo- The first step is to establish the tempera-
ture of the gas stream as it first progresses through the bed (at «' = 0).
The establishment of the first point toio is indicated in Fig. 227. From
the mean temperature of the solid ts„ at the interval Au'/2 a horizontal
line is projected of unit length in terms of u' to point B. (If AM' = 0.1,
then the length of the horizontal line is 10 AM') . A straight luie is drawn
connecting B with ieoo- The intersection of this line with the lAu' line
Stepll
0 I lAMt' 2 A » ' 3 A u ' 4Att' OAtt' lAtt' 2Att'

AM' AM'
2
FIG. 226. Calculation of Tem- PIG. 227. Step I. Calcula-
perature Gradients in a Granular tion of Gas Temperatures
Bed. at Zero Values of «'.
gives the desired temperature toio. The validity of this procedure is

evident from the geometry of Fig. 227 where, from the relationship of
similar triangles,
toio — iooo
-L (102)
AM' ~ AM' ~ '•''"
in agreement with Equation (99). Successive values of toto, toso are

obtained in a similar manner with successive increments of AM'.
For the common case where the solid is initially at a uniform tempera-
ture, Equation (97) can be integrated directly to give
tso ~ ig
In u', (at v' = 0) (103)
tsa ~ taa
Step II. In Step II it is desired to establish the temperature of the
solid at the entrance of the bed (M' = 0) for various values of v'. The
temperature of the entering gas is plotted against v'. The values of tsoo
are known and plotted. A mean value of t'^^ is obtained at At;'/2.
The value of Bm is calculated at the average gas temperature t'a^ and

added to i^^to give point C. A horizontal" line CD is extended equal
in' length to a unit value of v'. A straight line tsooD is constructed.
The intersection of this line with the Av' line
gives the desired value of tsoi. The validity of
this procedure is evident from the geometry of
Fig. 228 where, from the relation of similar
triangles,
tso
+ £m = (104)
Av' 1 Av'
in agreement with Equation' (100). Succeeding
values of ts at zero values of u' are obtained in
A»' the same manner for successive increments
of At;'.
FIG. 228. Calculation of Step III. With the establishment of the
Temperatures of the Solid boundary, conditions for both gas and solid
at Zero Values of u.
temperatures of both gas and solid for suc-
ceeding increments not at zero values for either u' or v' are obtained by
a combination of the methods illustrated in Figs. 227 and 228 and
shown in Figs. 229 and 230. For examples, the next points to be es-
tablished are tsu and ton-
On Fig. 229, values of fcoi and tsai and on Fig. 230 values of taio and
tsia are located from previous steps I and it, respectively. The values of
ton and tsn are estabKshed on each
chart by the respective methods of
Figs. 227 and 228. Simultaneous
adjustment of the two figures must
be made according to their geom-
etry until points ton agree on both
charts, and similarly tsn- A me-
chanical device fulfilling the ge-
ometry of the two figures may be
used to obtain agreement readily.
It is required in Fig. 229 that the
linkage of lines pivots on points
tsn and ton, and that the line ts„B PIGS. 229-230. Step III. Evaluation of
is horizontal and of unit length in to and ts at Zero Values of M' and v*.
u'. Points ^511 and tan are free to
move along the vertical line Au'. I t is required in Fig. 230 that the
linkage of hnes pivots on points ioioand tsio, and that the points C, t'sâie
spaced at an interval B^, depending upon the rate and heat of reaction.
CHAP. XXII] GENERAL CASE 1105
and are both free to move along the vertical Av/2, that line CD is hori-
zontal and of unit length in v'. Succeeding values of gas and solid tem-
perature are obtained in a similar manner.
General Case. The procedure illustrated in Figs. 227-230 required
knowledge of the reaction rate r at all points in the reactor. In the
general case r is a function of conversion and composition as well as of
temperature. Under such conditions it is necessary to evaluate simul-
taneously the distribution of conversion and composition along with
temperature as the stepmse procedure is carried out. This involves
combination of equations of the form of (83) and (84) for the evaluation
of conversions with Êquations (99) and (100) for temperature. The
simultaneous solution of the four equations is carried out by a stepwise
procedure similar to that used for either pair individually but compli-
cated by the necessity of additional successive approximations which
result from the fact that the temperature changes and conversion changes
in any increment are interdependent.
A common example is in the adiabatic adsorption of a gas by a station-
' ary bed. The rate of adsorption is a function of the adsorbate content
of both the gas and the solid as well as of the temperature of each and
can be determined experimentally and expressed graphically for even
the most complex cases. Similarly, the differential heat of adsorption is
a function of temperature and the adsorbate content of the adsorbent
which may be expressed in either graphical or analytical form. From
these two relationships working charts may be prepared which for a
specific problem relates B^ of Equation (101) to temperature and the
adsorbate content of both the gas and the adsorbent. A similar pro-
cedure may be followed for a chemical reaction whose rate equation and
thermochemistry are known.
The evaluation of the complete distributions of temperature, reac-
tants, and products as functions of time and position is a straightforward
but tedious operation following the patterns previously described. In
this operation the increments of time and distance used in establishing
temperature variations must be related to the time and distance incre-
ments used in establishing variations in composit/on. For simplifica-
tion the increments of u and v should be related as follows:
AM' = - AM = fi AM (105)
and Ay' = ^ Av = ^ ^ Av (106)

1106 UNCATALYZED HETEROGENEOUS REACTIONS [CHAP. XXll
PROBLEMS
1. It is desired to design a plant for the production of 15,000 lb der day of mono-
nitrotoluene from toluene. A design study is to be made on the basis of a fresh mixed
acid of the following composition in mole per cent:
HNOj 45
H2SO1 30
H2O 25
100
It is desired to produce" a conversion of 99.0 per cent with an excess of 3 per cent
HNOs above that theoretically required.
(o) Using the data of Illustration 1, calculate the minimum reactor volume
required in a batch operation at 30°C in which 2 hr are required for charging and
emptying the reactor.
(6) Calculate the reactor volume required for a single-stage continuous operation
under the conditions of part (a),.
(c) Calculate the volumes of the reactor required for a two-stage counterflow
operation in which 50 per cent of the total conversion is produced in each reactor.
Neglect the formation of di- and trinitro compounds.
(d) Calculate the reactor volume required if the process is carried out in a two-
stage counterflow operation with reactors of equal size.
2. For the reaction represented by Equation (17) derive rate equations correspond-
ing to the following mechanisms:
(o) Reaction between an adsorbed B molecule and an adjacent active center with
the rate of desorption of product B as the rate-determining step.-
(b) Reaction on a single active center between gaseous B and solid A to form a
chemisorbed R molecule and a gaseous S molecule. The rate-controlling step is the
surface reaction.
3. Using the data and assumptions of Illustration 2, develop curves similar to those
of Fig. 213 for combustion of the coke with an air rate of 150 lb/(sq ft) (hr) at an
average temperature of 2400°F. Compare these results with Fig. 213.
4. Hard water is softened at a rate of 20 gal per min in an exchanger bed 2 sq ft
in area and 2 ft thick. The initial hardness of the water is equivalent to 200 parts of
calcium carbonate per milhon parts of water. The exchanger has an initial replace-
able sodium content equivalent to 28,000 grains of calcium carbonate per cu ft. The
density of the mineral is 30 lb per cu ft, grain size 20-28 mesh, and with a value of k
equal to 2.29.
(a) Calculate with and without use of Fig. 214, the residual hardness of the water
after 5 hr of steady flow.
(b) Plot the residual hardness of water against time for a period of 24 hr by use of
Fig. 214.
(c) Calculate with and without the use of Fig. 215 the time required to reach a
residual hardness of seven parts per milUon and the gallons of water softened.
6. In a cation exchanger water is softened at the rate of 3 gal per sq ft per min in a
bed 4 ft thick, having a density of 30 lb per cu ft and an exchange capacity of 20,000
grains of calcium carbonate per cu ft. The water has an initial hardness of 200 parts
equivalent calcium carbonate per million. After operation for 8 hr, the water leaving
has a residual hardness of 12.8 parts per million of calcium carbonate.
CHAP. XXII] PROBLEMS 1107
(o) Calculate the value of the reaction velocity constant ki as defined by Equation
(32).
(6) Plot the residual hardness against time over a period of 16 hr.
6. Air at 70°F and 90 per cent relative humidity is passed isothermally through a
bed of 2-4 mesh activated alumina 2 ft in thickness at a velocity of 50 ft per min
(based upon total cross section of bed).
The bed is initially free of moisture. An average value of the adsorption equilib-
rium ratio m over the entire relative humidity range is 0.20.* The following proper-
ties of the bed are assumed: PB = 50 lb per cu ft, Dp = 0.0216 ft,' a, = 181 sq ft per
cu ft 50 per cent external void space. Calculate, over a drying period of 4 hr and
at the end of 1, 2 and 4 hr,
(o) The humidity of the air leaving the bed.
(6) The average humidity of the air leaving the bed.
(c) The moisture content gradient in the bed.
(d) The average moisture content of the bed.
(e) The total water adsorbed by a bed 10 sq ft in area.
7. Air at 80°F and '80 per cent relative humidity is passed through a bed of silica
gel 6.8 mesh, 12 in. thick, at an entering rate of 136 ft per min based on total cross
section. The drying of the air takes place isothermally. The bulk density of the gel
is 39 lb per cu ft,'and the value of m is 0.55.* At the end of 1 hr the relative humidity
of the air leaving is 10 per cent. Calculate the value of Hda^ for the bed.
8. Benzene vapor present to the extent of 1.0 per cent by volume in hydrogen is to
be removed by passing the gas mixture downward through a bed of silica gel at 80°F
and 2 atm pressure at a linear velocity of 200 ft per min (based on total cross section).
It is desired to operate for a period of 90 min with a minimum benzene recovery from
the gas of 90 per cent. The silica gel is 2-4 mesh, with a bulk density of 39 lb per cu ft,
and witt the following estimated properties, Dp = 0.0220 ft, a„ = 117 sq ft per cu ft.
The adsorption-equilibrium factor m for benzene on silica gel is 1.67.* Over the 1.5
hr period calculate:
(a) The depth of bed required.
(6) The concentration gradient of benzene in the bed.
(c) The total benzene removed by a bed 8 sq ft in area.
* In problems 6, 7, and 8, m is defined by the relation w = mp*/p, as used in

Equation (65).
APPENDIX
ATOMIC WEIGHTS OF THE MORE COMMON ELEMENTS
Aluminum Al 26.97 Manganese Mn 54.93

Antimony Sb 121.76 Mercury Hg 200.61
Argon A ' 39.944 Molybdenum Mo 95.95
Arsenic As 74.91 Neon Ne 20.183
Barium Ba 137.36 Nickel Ni 58.69
Bismuth Bi 209.00 Nitrogen N 14.008
Boron B 10.82 Oxygen 0 16.000
Bromine Br 79.916 Phosphorus P 30.98
Cadmium Cd 112.41 Platinum Pt 195.23
Calcium Ca 40.08 Potassium K 39.096
Carbon C 12.010 Selenium Se 78.96
Chlorine CI 35.457 Silicon Si 28.06
Chrotjiium Cr 52.01 Silver Ag 107.880
Cobalt Co 58.94 Sodium Na 22.997
Copper Cu 63.57 Strontium Sr 87.63
Fluorine F 19.00 Sulfur S 32.08
Gold Au 197.2 Tellurium Te 127.61
Helium He 4.003 Tin Sn 118.70
Hydrogen H 1.008 Titanium Ti 47.90
Iodine I 126.92 Tungsten W 183.92
Iron Fe 55.85 Uranium U 238.07
Lead Pb 207.21 Vanadium V 50.95
Lithium Li 6.940 Zinc Zn 65.38
Magnesium Mg 24.32 Zirconium Zr 91.22
Source: J. Am. Chem. Soc, 63, 850 (1941).

xviii APPENDIX
CONVEKSION FACTORS AND CONSTANTS

Analysis of Air
By weight: oxygen, 23.2%; nitrogen, 76.8%
By volume: oxygen, 21.0%; nitrogen, 79.0%
Average molecular weight of air 29
Average molecular weight of atmospheric nitrogen 28.2
Physical Constants
The Gas Law Constant R
R = 1.987 (calories)/(g-mole)C'K)
R = 82.06 (cu cm)(atm)/(g-mole)(°K)
R = 10.71 (Ib/sq in.)(cu ft)/(lb-mole) (°R)
R = 0.729 (atm)(cuft)/(lb-mole)(°R)
1 faraday = 96,500 coulombs

Avogadro constant = 6.023 X 10" per gram-atom
Density
1 gram-mole of an ideal gas at 0°C, 760 mm of mercmy = 22.414 liters
1 pound-mole of an ideal gas at 0°C, 760 mm of mercury = 359.0 cubic feet
Density of dry air at 0°C, 760 mm of
mercury = 1.293 grams per liter or.. 0.0807 pound per cubic foot
1 gram per cc 62.4 pounds per cubic foot
1 gram per co 8.337 pounds per U. S. gallon
Energy K
/ Foot- Kilogram-
Calories Btu Joules pounds meters
Calorie 1 3.968 X 10-' 4.185 3.087 0.4267
Btu 252 1 1055 777.9 107.5
Joule 0.2389 9.482X10-4 1 0.73756 0.1019
Foot-pound 0.3240 1.286X10-3 1.356 1 0.13826
Kilogram-meter 2.343 9.298 X 10-' 9.806 7.2327 1
Liter-atmos 24.21 9.607X10-2 101.32 74.733 10.333
iChu 453.6 1.8 1899 1400 193.5
Liter-atm Cuft- Foot- Horsepower

atm poundals hours
Calorie 4.130 X 10-2 1.459 X 1 0 - ' 99.31 1.5591 X 10-«
Btu 10.41 0.3676 25030 3.929 X 10-4
Joule 9.869 X 1 0 - ' 3.485 XlO-4 23.73 3.725 X 1 0 - '
Foot-pound 1.3381X10-2 4.7253 X 10-* 32.174 6.0505 X 10-'
Kilogram-meter 9.678 X 10-2 3.4177 X 10-5 232.7 3.6529 Xl0-<
Liter-atmos 1 3.5319X10-2 2403.8 3.7734X10-5
IChu 13.74 0.6617 45054 7.072 XlO-4
* From Perkins, iTitroduciion to General Thermodynamicit John Wiley & Sons. Inc., Publishere, with
permission.
APPENDIX xix
Length
1 inch 2.540 centimeters
1 micron 10~* meter
1 Angstrom unit 10""" meter
Mass
1 poimd* 16 ounces*
1 pound* 7000 grains
1 poimd* 453.6 grams
1 ton (short) 2000 pounds*
1 grata 15.43 grains
1 kilogram 2.205 pounds*
* Avoirdupois.
Mathematical Constants
e 2.7183
IT .' 3.1416
In N 2.303 log N
Power
1 kilowatt 56.92 British thermal units per minute
1 kilowatt 1.341 horsepower
1 horsepower 550 foot-pounds per second
1 watt 44.24 foot-pounds per minute
1 watt 14.34 calories per minute
Pressure
1 pound per square inch 2.04 inches of mercury
1 pound per square inch 2.31 feet of water
1 atmosphere 14.7 pounds per square inch
1 atmosphere 760 mm of mercury
1 atmosphere 29.92 inches of mercury"
Temperature Scales
Degrees P 1.8 (degrees C) + 32
Degrees K degrees C + 273.15
Degrees R degrees F -1- 459.7
Volume
1 cubic inch 16.39 cubic centimeters
1 Uter 61.02 cubic inches
1 cubic foot 28.32 Uters
1 cubic meter 1.308 cubic yards
1 cubic meter 1000 liters
1 U. S. gallon 4 quarts
1 U. S. gallon 3.785 Uters
1 U. S. gallon 231 cubic inches
1 British gallon 1.20094 U. S. gallons
1 cubic foot 7.481 gallons
1 liter 1.057 quarts
1 U. S. fluid ounce 29.57 cubic centimeters
APPENDIX
-240 -875
160-j; 460-j: 1600-::
-70 800 — :
— 420 -850
iBo-j: 440-: 1550-::
775-j
-: -: - 2 2 0 -825
140—: — 60 420-j " 1500-j
J 750-^ — 400
- |-800
130-j 400-= - 1450-|
J 7-200 726-j : ^775
|-50 ^380
120-4 380-j 1400-j
700-j ^750
110-4 360 - :
-^360 1350 — ^ 7 2 5
j - 180
-: ^ 4 0 675-;
100-4 840 - j 1300-j ^700
- 650-j 1
i-340 1250 — — 675 .
90-i 3 2 0 - i f-160
: t5 u. -: U
: o
,4 625 — «i
be : • &
JsoH L 1*300-1 - ^1200-; peso's
1 SL ^320
1-140 600 — ^625
TO-i 280-j z. 1150-;
^20
-- 575-; ^600
260-t E-300 1100-j
60-i
I
[-120 — 675
650-]
60-j ^ 1 0 240-i 1060-^
- ^280
^550
*0- 220 — 625-5 -_ 1000-: :
^100 ^525
:
80 — i-0 200 — 600 — ^ 2 6 0 • 950-
—500
-
^ 900-
:
20- 180 — 475-
\'—&) :^475
'• 10 : 1-240
I
XO- 160- 4 5 0 - I: 8 5 0 - 1^450
:-
0- i- 140-^ 6 0 •t-220 800-t.425
FIG. A. Temperature conversions.
/
APPENDIX
10- 1.00 0.85

1.16
36
0.99
1.15 12- 0.84
-8- 0.98 38 •
— 1.14 14- 0.83
-7- 0.97 40 •
-6- r-1.13 0.82

16- 0.96 42-
-5- ^ 1.12 0.95

18- 44 - t-0.81
-4- 1.11 0.94
46- 0.80
20-
-3- £-0.93
— 1.10 48-
— 0.79
.-2- 22- 0.92
. — 1.09 . ^'60-
*" u.
h- 1.08-5
24—1 - 0.91 , 3
S 52-
^0.78
3 a.' 0.77-
-0.90 ,
E-1.07« < 54 -
1 •
(0
< +- tj)
-0.89 ' 0.76'
Q 28 ° 56 -
1.06
•0.88 0.75
3 ' 1.06 30
58-
4 •0.87 60-
1^0.74
- 1.04 32 —^
S 62-
0.86
6 • 1.03 34 - ^ 0.73
64-
0.85
7—1 • 1.02 36 0.72
66-
8 0.84
1.01 38 — £ 68 - 0.71
9—1 0.83
10 70-
1.00 40 —f- 0.70
Fia. B. Gravity conversions.

xxu APPENDIX
2500.0- 23000- -10000

Zt-10000
2000.0- 20000-
-8000
15000-^
lOOO.O — I -5000
-4000 -6000
-3000 lOOOO-
-6000
500.0-
400,0 - -2000 8000- -4000
300.0 ~
-1500
6000- -3000
200.0-
-1000
-2000
-750 40OO-
i5o;o -
3000- -1600
-500
100.0-
-400
75.0- 2000- -1000
-300 -800
1600-
50.0- -600 „
S 40.0-
-200 I -500 •§
g 1000-^ RO-
s -150 I =-400 • 120
= 800 — 7.0 - • 110
•3 30.0-
c H 600-^
-300 I 6.0- - 100
<3 -100 S 6.5 - -90
•3 20.0 -
-200 f 5.0- •80
-90 I s .i 400- 4.5 • -70
1 15.0- -«> s 1 300-
-150 ^ 4.0-
-60 Sc
I
> E o
-70 &
t
2 200- -50,g_
10.0- •I -80 o
9.0- fe-60 I 150- -40 I
-55 -60
8.0- -35 §
I",
— 50 100- -50
7.0- -40
80- -30 1
6.0-
— 45 -25 ^
60-
5.0 — 50- -30
;=;;20 ,
40- lE-18 •
— 40
4.0 —
30- -20 i.o-i ^ 1 6
0.9 4
20-3 0.8 - i
r- 12
3.0- 0.7 -4
0.6 -1 1-10
Z4- -35 10—r^'^ O.S> — = - 8
FIG. C. Viscosity conversions.

AUTHOR INDEX
Part I, pages 1-436 Part II, 437-804 Part III, 805-1107
Ahlberg, 1086 Carlson, 652, 653, 654, 668, 661

Alvarado, 903 Caspari, 129
Amagat, 38, 695 Ceaglske, 157, 339
Andersen, 758, 764 Chilton, 983, 984, 991
Anderson, A. G., 708 Clapeyron, 60, 457
Anderson, C. T., 742, 748 Clausius, 60
Anzelius, 1084 Coates, 64
Arnold, 988, 989 Colburn, 652, 653, 654, 665, 658, 659,
Asbury, 933 661, 983, 984, 991
Aston, 708, 709, 788 Coolidge, 164, 158
Augustine, 1068, 1069 Cox, 65, 83
Avery, 709 Crawford, 789
Avogadro, 28 Cummings, 227
Badger, 783 Dalton, 38, 595, 626

Barker, 790 Daniels, 890
Bartlett, 596 Davis, 67, 512
Bauer, 295 Deming, 482
Beach, 709 Dewitt, 157
Beattie, 482, 598 Dobratz, 798
Beckmann, 944, 947 Dodd, 967
Benedict, 483 Dodge, 490, 714, 733
Bennewitz, 798, 799 Doescher, 709
Benning, 520 DoUey, 605
BergeUn, 143, 144, 145 Doss, 262,. 613
Berkman, 903 Drew, 1016
Beyer, 758, 764 Du Domaine, 1075
Bichowsky, 253, 2?», 292, 745, 866 Duhem, 625
Blasdale, 129 Duhring, 65, 83
Blodgett, 905 Dulong, 217
Bodenstein, 854, 897 Dunham, 931, 1006, 1032
Boltzmann, 217
Borden, 790 Edgell, 708
Bornstein, 944 Edmister, 494
Borsook, 709 Egan, 708
Bosnjakovic, 281, 683 Egloff, 903 •
Bourland, 708 Eidinoff, 708
Bridgman, 482 Ellis, 709
Brown, 931, 1005 Emmett, 157, 904
Brugmann, 227 Eucken, 990
Brunauer, 915 Ewell, 708
Bryant, 213 Ewing, 295
Buffington, 521 Eyring, 765, 808, 815, 894, 895, 908, 912
Cain, 897 Falk, 1068

Calingaert, 67, 512 Fallon, 336, 339
XXIV AUTHOR INDEX
Fink, 708 Kaltenbach, 303

Fischer, 898 Kassel, 709, 790, 1031
Forest, 227 Katz, 601, 673
Forster, 708 Kay, 601, 604
Foster, 685 Keenan, 444, 522, 567, 583, 618, 915
Fowler, 765 KeUey, 692, 724, 738, 742, 748, 766, 775,
Freundlich, 154 790
Friauf, 314 Kemp, 708
FroUch, 861 Kennedy, 709
Furnas, lOSJ^ Keyes, 522, 667
Kharasch, 763
Gamson, 73, 604, 606, 621, 665, 676, 985 Kiefer, 567
Kimball, 765
Garner, 709 Kkchhoff, 304
Gibbs, 618, 625 Kirkbride, 672
Gilkey, 521 Kistyakowsky, 230, 231, 709
Gilliland, 677, 981, 988 Koo, 1016
Glasstone, 793, 808, 815, 894, 895, 908, Kopp, 219
912
Glockler, 708 Kreisinger, 1068, 1069
Gohr, 9S1, 1032 Kurata, 601
Goodman, 243
Gordon, 235, 238 Lacey, 509, 517
Grossman, 10S6, IO4I, 1095, 1100 Laidler, 808, 816, 894, 895, 908, 912 "
Grosvenor, 93 Lamb, 846
Guggenheim, 765 Langmuir, 905
Guldberg, 68 Laplace, 269
Guttman, 708 Lavoisier, 259
Gwin, 786, 787 Lewis, G. N., 285, 287, 314, 619, 623,
626, 637, 641
Lewis, W. K., 1064, 1056
Halford, 708 Licht, 872
Hamer, 649 Lind, 854, 897
Hardy, 708 Linnett, 709
Harrop, 365 Lockhart, 143, 144, 145
Hart, lOU, 1015, 1021 Love, 916
Haslam, 326, 1068 Luder, 212, 216, 217
Hatta, 99
Henry, 146 Maier, 737
Hermann, 292 Margules, 652
Herzfeld, 843 Markwood, 520
Hess, 259 Marshall, 1080, 1084, 1086
Hildebrand, 679, 682 Martin, 931, 1005, 1032
Hirschfelder, 779, 781, 866, 892, 893 Masson, 605
Holcomb, 335 Maxwell, 457, 975, 990
Hougen, 157, 906, 944, 985, 1075, 1080, Mayer, J., 765
1084, 1086 Mayer, M., 765
Huffman,'697, 708 Mayfield, 799
Huggins, 709 McAdams, 974, 989, 990, 1016
Hulburt, 1004 McCabe, 278
McHarness, 520
Jack, 709 McKinley, IO64, 1055
Jahnig, 1005, 1032 Meissner, 70, 236
Jeans, 975 Merriman, 662
Johanson, 933, 934, 959 Messerly, 708, 709
Johnston, 793 Morgen, 292
Jones, 314 Morrell, 903
Justi, 212, 216, 217 Murphree, 931, 932, 1005, 1032
Justice, 899
AUTHOR INDEX XXV
Murphy, 330, 332, 846 Shupe, 482

Myers, 8S3, 861 Simpson, 931, 1006, 1032
Smith, D. F., 899
Nelson, 330, 332, 493 Smith, R. L., 333, 490, 601, 604
Newton, 490, 931, 1006, 103S Smith, W. J., 620
NicoU, 897 Sokolnikoff, E. S., 830, 1084
Sokolnikoff, I. S., 830,1084
Oman, 1016, 1017 Senders, 666
Onnes, 207 Spencer, 899
Osborne, 709 Stfiohert, 87S
Othmer, 65, 83, 84, 145 Stefan, 975
Ovitz, 1068, 1069 Stegeman, 708, 709
Parks, 697, 709, 949 Stevenson, 709
PauUng, 765, 781 Stuart, 667
Pease, 841, 854 Stull, 799
Pecherer, 708 Suen, 1054, 1055
Perkins, 438 Sugden, 71
Perrott, 314 Swain, 1075, 1076
Perry, 223 Szasz, 708
Petit, 217
Pitzer, 708, 709, 712, 763, 786, 787, 789, Taylor, 793, 904, 908
949, 960 Thiele, 997, 998
Prentiss, 652 Thodos, 985
Prosen, 712 Tobias, 145
Tolman, 765
Randall, 285, 287, 623, 626, 637, 641 Trouton, 230
Raoult, 80, 626 Tsohernitz, 944
Ratchford, 898 Tyson, 931, 1006, 1032
Reed, 830
Reiser, 853 Uehling, 321, 328, 329
Rice, F. O., 843 Umino, 418, 419
Rice, W. W., 709 Uyehara, 870, 1021
Rossini, 253, 262, 271, 292, 712, 745,
868, 943 van der Waals, 463, 480
Rossner, 798, 799 van Laar, 653, 664, 660
Rubin, 483 Vaughan, 895 '
Russell, H., 709
Russell, R. P., 326, 93S Walter, 766
Watson, K. M., 68, 73,231,233,330,331,
Sage, 509, 517 332, 333, 334, 336, 339, 490, 493,
Sagenkahn, 709 501, 602, 604, 604, 606, 621, 665,
Scatchard, 649, 652 676, 758, 764, 846, 853, 861, 870,
Schaafsma, 509 906, 933, 934, 959, 1016, 1017,
Schmidt, 1041 1021
Schneider, 861 Webb, 483
Schoenborn, 655 Weber, 494
Schumacher, 894 Webster, 617
Schumann, S. C, 708, 793 Wenner, 224, 228, 757
Schumann, T. E. S., 1084 Westrum, 708
Schwartz, 793 White, 1054
Scott, 709, 949, 960 Wilke, 985
Selheimer, 666 Wilson, 766, 793
Setler, 708 Wohl, 649, 652, 669, 660
Shaw, 462, 465
Sherwood, 830, 974, 980, 981 York, 494, 572
ShiUing, 655 Yost, 709
Shimp, 931 Zemansky, 202
SUBJECT INDEX
Part I, pages 1-436 Part II, pages 437-804 Part III, pages 805-1107
Absolute pressure, 35 Activity factors, 935-937

Absolute reaction rates, 808 Actual gases, behavior of, 479-B02
Absolute zero, 31 Adiabatic combustion, 308, 312
Absorption, 146 Adiabatic conversion, 721
Absorption refrigeration, 575, 681-588 Adiabatic coofing lines, 104
Accelerators, S44 Adiabatic free expansion of gases, 501
Accumulation of inventory, 179-180 Adiabatic humidification, 242
Acetone-chloroform, activity coefficients Adiabatic reactions, 308, 864
of, 645-647 Adiabatic reaction temperature, 725-729
saturation temperature of, 648 Adiabatic reactors, 864, 1011, 1027
Acids, specific heats of, 224 Adiabatic vaporization, 104
Activated adsorption, 150, 908-910 Adsorbate gases, recovery of, 156
Activated alumina, 158, 902, 905 Adsorbed system, enthalpy of, 299-300
Activated charcoal, 153, 155, 905 Adsorbents, 158
Activated complex, 808 Adsorption, activated, 149-150, 906-910
Activation, energy of, 816, 908 by silica gel, 1089-1091
enthalpy of, 810, 955 controlfing, 922, 948-949
prediction of, 889 of atomic hydrogen, 948-949
entropy of, 810, 956 equilibrium, 151, 1096
prediction of, 889 constants of, 943, 956, 1020
Active centers, 904, 910 heat of, 297-299
equidistant, 913 in granular bed, adsorbate in bed, 1085
fraction occupied, 913 adsorbate in fluid stream, 1085-1086
fraction unoccupied, 913 concentration gradients in, 1097
Activity, 627-641, 715, 911, 1010 isothermal, with complex equilibrium,
Activity coefficients, 630-641, 676, 715, 1094
1055 mechanism of, 1065
apparent, 656 of benzene by silica gel, 1091
effect, of chemical type on, 664 of gases by solids, 149, 1080-1105
of molecular size on, 664 of reactant, 919
of temperature on, 661 preferential, 157
in acetone-chloroform system, 645- rate-determining step, 1067
647 , reversibiUty of, 156
in aqueous sodium chloride solutions, theory of, limitations of simple, 914
640 • van der Waals, 150, 908-910
in benzene-toluene system, 645-647 velocity constant of, 910
in binary systems, 651-659 with linear equilibrium, 1083
in different standard states, 635 Adsorption calculations, 153
in isopropyl ether-isopropyl alcohol Adsorption capacity of desiccants, 160
system, 645-647 Adsorption equilibria, effect of tempera-
in sucrose solutions, 636-637 ture and pressure on, 153-155
in symmetrical systems, 659 gas-solid, 684
in water-w-butanol system, 645-647 Adsorption hysteresis, 156-157
ionic, 637-638 Adsorption isotherms, 151-155
of mixed acids, 1056 chart, 153
XXVUl SUBJECT INDEX
Adsorption refrigeration, 688 Benedict-Webb-Rubin equation of state,

Air, 240, 243, 244 483
humid heat of, 241 Benzene, adsorption, by sihca gel, 1091
theoretically required for combustion, on activated cocoanut charcoal,
354 154-156
used in combustion, weight of, 354 diffusion coefficient of, 989
work of separation of, 687 MoUier chart, 628
Alcohol, catalytic decomposition of, 903 nitration of, 1054-1060
Alumina, activated, 90S, 905 pyrolysis of, 84S, 876
Amagat's law, 38, 595 solubihty in naphthalene, 112-113
Ammonia, diffusion of, 980 temperature-enthalpy chart, 627
Moliier, chart, 526 temperature-entropy chart, 627
pressure-volume chart, 623 toluene system, activity coefiicients,
solubility in water, 149 645-647
synthesis, equilibrium conversion, 718 saturation temperature of, 648
fugacity coefficient ratio, 718 • • Bimolecular-monomolecular reactions,
temperature-entropy chart, 525 920
Ammonia-water, enthalpy-composition Bimolecular reactions, 921
diagram, 684 Binary systems, activity coefficients in,
Amorphous carbon, 324 661-659
Analyses of coal, 324r-325 Blast furnace, 406
API gravity scale, 20, 331-332, chemical reactions in, 407
441 distribution, of charcoal in, 412
Aromatics, nitration of, 10S4-1066 of flux in, 411
Arrhenius equation, 816, 955,1065 of ore in, 409
Ash content, corrected, 346 energy balance of, 406-409, 414-420
in coal, 325 material balance of, 406-414
Atomic covalent radii, 781-782 Boiler furnace, combustion of coal in,
Atomic fraction, 15 347-362
Atomic heat capacities, 217 Boiling points, 57, 77, 84, 647
Atomic per cent, 15 estimation of, from hquid density, 506
Atomic volumes, 988 normal, 58
Atomic weights, 3, 437 of ethanol-benzene, 665
Attractive force between molecules, 52- of paraffins, 607
53 of petroleum, 332
Availability of energy, 438-444, 660 Boltzmann factors, 772, 792
Available hydrogen in coal, 329 Bond angles, 781-782
Available sites, 911 Bond bending, 767-768
Avogadro number, 28, 30, 910 Bond frequencies, 798
Avogadro principle, 28, 30 Bond frequency assignments, 799
Azeotropes, 647, 661 Bond stretching, 767-768
effect, of pressure on, 663 British thermal unit, 203
of temperature on, 662 Brix gravity scale, 20
Azeotropic ethyl acetate-ethyl alcohol, Bromine, hydrogenation of, 8S0, 841-855
663 Bubble-point equiUbria, 672
Azeotropic solutions, 647, 656-656 Bubble-point fine, 600
Bulk density, 996
Base group properties, 769 Bureau of Mines method of coal analysis,
Bases, specific heats of, 224 423
Basis of calculation, 11 Butadiene, 966
Baum6 gravity scale, 19 decomposition of, 966
Beattie-Bridgman constants for gases, dimerization of, 966
483 Butane, alkylation of, 969
Beattie-Bridgman equation of state, 479, cracking of, 966
482-483 dehydrogenation of, 967
for mixtures, 698 n-Butane dehydrogenation, 966
SUBJECT INDEX XXIX
Butene production, 968 Catalytic dehydrogenation, 185

Butenes, dealkylation of, 966 of n-butane, 969
By-passing fluid streams, 193 Catalytic oxidation of sulfur dioxide,
1021-1026
Calcium balance, 1076 Catalytic reactions, 90S
Calcium carbonate, solubility of, 747 mechanism of, 906
Calcium chloride, as a desiccant, 158 problems of, 970-972
solutions of, 137 Catalytic reactor design, 1009
enthalpy of, 281 Catalytic unit, height of, IOI4
relative humidity of, 137 Cation exchange, 1074-1080
vapor pressure of, 137 Cementite, heat of formation of, 417
Calorie, 203 Centigrade heat unit, 203
"Calorimetry, low temperature, 699 Centigrade temperature scale, 32
Capillary condensation, 150 Centipoise, 870
Carbon, amorphous, 324 Centistoke, 570
heat of combustion of, 262, 323 Chain reactions, 819, 843
Carbonaceous deposit, 1032 Chamber sulfuric acid, 383, 399
residue, 1100 energy balance of, 400-402
Carbonates, solubility of, 742-748 material balance of, 399
Carbon balance, 349 Channeling, 1017
Carbon dioxide, pVT relations of, 479- Characterization factor, 329
480 from viscosity, 332
solubility of, 683 of petroleum, 330-333
Carnot cycle, 667-560 versus weight percentage of hydrogen,
reverse, 576 333
Carnot principle, 559-560 Charcoal, activated, 153-155, 90S
Catalyst, 902 adsorption on, 155
activity factor of, 936-937 distribution in blast furnace, 412
deactivation of, 930 Checkerbrick performance, 110
diffusion in, 995-1002 Chemical compound, 2
fouling of, 9S0, 1008 Chemical equilibrium, 691-755
life test of, 940 Chemical potentials, 614-617, 663
porosity of, 907 equality in phases, 617
probable life of, 940 relation to energy functions, 616
reactivation of, 932 Chemical reactions, 4
regeneration of, 904, 931, 1008 equilibrium constants of, 712
selectivity of, 903, 1006 feasibility of, 714
sintering of, 906 Chemisorption, 908, 1063, 1067
structure of, 907 and dissociation, 912-913
Catalyst bed, cooling, 931 equilibria of, 910
fluidized, 1005 rates of, 910
Catalyst deposit, 930 Chlorides, specific heats of, 225
removal of, 931 Chlorine, solubility of, 682
Catalyst requirement, 1031 Clapeyron equation, 59, 457
Catalysts, in cylindrical tubes, 1035 Claude process, 690
in thin slabs, 1034 Clausius-Clapeyron equation, 60-61
. shape of, 996 Clearance volume, 571
solid, 906 Closed system, 437
supported, 905 Coal, analyses of, 324
unsupported, 905 Bureau of Mines method, 423
Catalytic alkylation, 1028 proximate, 324
Catalytic beds, heat transfer in, 97S ultimate, 324
mass transfer in, 973 ash content of, 325, 346
Catalytic conversion equation, 929 available hydrogen in, 329
Catalytic cracking, 433-436 classification of, 326
Catalytic decomposition of alcohol, 90S combustion of, 326
XXX SUBJECT INDEX
Coal, composition of, 324 Compressibility, of liquid solutions, 605-

heating value of, 326 609
by Dulong's formula, 327 Compressibility factor, 484-492
from total carbon, 328 chart, 489
net, 327 use of, 490-492
moisture in, 324-325 expressed as virial coefficients, 488
net hydrogen content of, 328 for mixtures, 595-598
oxygen in, 354 for nitrogen, 484-487
rank of, 326 and hydrogen mixtures, 696-597
Coal-fired furnace, 347 generalized, 488-492
energy and material balance of, 348- interpretation of experimental data,
356 487-488
carbon and hydrogen content un- mean, 595-598 *
known, 364 Compression, multistage, 672-576
not neglecting sulfur, 327 of fluids, 538-594
Codimer, hydrogenation of, 9^3-968 of gases, 568-576
Coefficient of performance, il^STI of ideal gas, 439
Coke, 323 refrigeration, 575-581
kinetics of combustion of, 1088-1073 single-stage, 570-671
specific heat of, 218 Compression efiiciency, 673
Combined feed ratio, 186 isentropic, 570
Combustion, adiabatic, 721 isothermal, 670 '
gases of, weight of, 354, 366 over-all, 571
heat of, 260 volumetric, 571
see aho Heat of combustion Compressors, 668
of coal, 326 turbocompressors and fans, 669-670
in boiler furnace, 347-362 Concentration units in rate equations,
of coke, kinetics of, 1068-1073 820
on grate, composition of gas leaving, Condensation, 56, 96
1073 capillary, 150, 157
of fuels, 338, 345 retrograde, 600-601
calculations, chart for, 372 Configuration of void space, 1018
incomplete, 341 Congruent points, 116
where ultimate analysis is unknown, Conjugate line, 144
357 Conjugate solutions, 142
problems of, graphical calculation, 371 Consecutive reactions, 818, 829
processes of, 342-371 Conservation, of energy, 28
Complex, activated, 808 of mass, law of, 1
Complex equilibria, criteria of, 614-619 Constant vapor concentration, tempera-
Complex molecules, structure of, 785 ture of, 69
Complex order, 819, 840 Constant volume, heat of reaction at, 302
Complex reaction, 701, 731-748, 963, 966 Contactor efficiency, 1061-1063
Complex systems, diffusion in, 977 Continuous counterflow operation, 1029
Component, in phase rule, 619 Continuous processes, 179
Composition, by volume, 13 inventory changes in, 179
changes in, 42 ControUed-cooling reactor, 1027
of coal, 324 Conventions, thermochemical, 250-261
of gases, 13 Conversion, adiabatic, 721
of mixtures, 12 of energy units, 438
of solutions, 12 of symbols, 22, 23
units of, 13 of units, 22
Compounding mixtures, 168-169 Conversion equations, 927
Compressibility, of gaseous mixtures, Conversion gradients in reactors, 10iS~
59e-«05 1047
of gases, 479-492 Conversion plots, empirical, 961-962,
of liquids, 602-506 Cooling of catalyst bed, 931
SUBJECT INDEX XXXI
Copper, heat of vaporization calcula- Cylindrical reactors, 1035

tion, 72B lateral heat transfer in, 1036
vapor-pressure calculation, 725
Corresponding states, theory of, 488 Dalton's law, 38, 92, 595, 626, 665
Cotton, equilibrium moisture content of, Data, thermodynamic, experimental de-
152 termination of, 518-521
Countercurrent heat exchange, 566 presentation of, 521-535
Countercurrent processing, 177 Deactivation of catalyst, 930, 932
Counterdiffusion, 982 Dealkylation of butenes, 966
Covalent radii, atomic, 781-782 Decomposition pressure, 723
Cox chart, 65-66, 160 of butadiene, 966
Cracking, catalytic, 433-i36 Definite proportions, law of, 3
of butane, 966 Degeneracy, 772
of petroleum, 421 Degradation of energy, 438-444
process of, 931 DegreCj of completion, 10
Cracking reactions, 967 of superheat, 57
Criteria, of complex equilibria, 614-619 Degrees of freedom, 618
of equilibrium, 449-453 internal, 768
Critical composition of solution, 143 molecular, 767
Critical-condensation temperature, 600 Dehydrogenation, 184-189
Critical constants, of gases, 481 of butane, 966-967
table, 234, 237 Densities, of gaseous mixtures, 41
of paraffins, 74 of liquids, 18-20, 502-506
Critical density, 54 of mixed acids, 1057
estimation of, 604 of sodium chloride solutions, charts, 18
Critical phenomena, of gaseous mix- of sulfuric acid, chart, 384
tures, 599-605 of ternary solutions, plot, 21
Critical point, for mixtures, 600 Density, 17
pseudo-, 604-605 bulk, 996
Critical pressure, 54, 72-73 critical, estimation of, 604
and vapor pressure, 234, 237 estimation of liquid, 505-506
of paraffins, 507 generahzed, for liquids, 502-505
Critical properties, estimation of, 68, 504 of gases, 35
Critical state, 54 of liquid solutions, 606-608
Critical temperature, 54, 71 of particle, 996
estimation from boiling point, 69 of solid, 996
of paraffins, 507 Density constants for paraffins, 507
of petroleum, 70, 73, 333 Depletion of inventory, 179-182, 200
oi refrigerants, 237 Derived properties, 453
of solution, 143 Dcsiccants, 158
Critical velo^-'ty, 974 activated alumina, 158
Critical volume, 70 adsorption capacity of, 160
estimation of, 604 calcium chloride, 158
Crystallization, 111-112, 123 equilibrium moisture content of, 158
fractional, 128 glycerine, 158
kinetic theory of, 111-112 hthium chloride, 158, 162
with no solvates, 123 relative humidity of, 158
with solvates, 126 silica gel, 158
Cycle, Carnot, 557-560 sulfuric acid, 158
reverse, 576 triethylene glycol, 158
power-plant, 560-668 vapor pressure of, 159
llankine, 561-565 Desorption controlling, 948-949
improvements on, 564-565 Determinants, 465, 962
refrigeration, 576-579 Dew point, 57, 93-94, 600
reversible, 567 Dew point equiUbria, 673
Cychc processes, 666-592 Dielectric constant, 904
xxxu SUBJECT INDEX
Difference of heat capacities, 461 Effectiveness factor, 996-998, 1000

Differential energy balance, 638 ratio, 999
functions, 454-155, 472 variables affecting, 997
Differential equations, exact, 455-457 Effectiveness of catalyst, 996
and inexact, 442 Efficiency, of compression, 573
Differential heats, of adsorption, chart, isentropic, 670
298,684-686 ' isothermal, 670
of solution, 29i, 293 over-all, 671 •
of wetting, 294-297 volumetric, 571
Differential rate data, 942 of engine, 566
equations, 816 thermodynamic, 569
Differential reactor, 980 for Carnot cycle, 559
Diffusion, 906, 976-982 for Rankine cycle, 563
coefficients, 976, 982-991 Electrolytes, activities in, 638
equimolal, 979 dissociation in, 638
in complex systems, 977 Electron diffraction, 777
in liquids, 981-982 patterns of, 902 <•
in porous catalysts, 995-1002 Electronic energy, 767
in process rates, 990-993 Electron microscope, 904
longitudinal, 1002-1005 Elements, specific heats of, 218
proportionality factor in, 978 Empirical conversion plots, 961-962
unidirectional, 976 Endothermic compounds, 252
Diffusion velocities, relative, 976 Endothermic reactions, 251
Dimerization, of butadiene, 966 Energy, 27
of ethylene, 811 availability of, 438-444
Diphenyl, formation of, 8^6 thermal, 560
production of, 875 conservation of, 28
Dissociation and chemisorption, 912-913 degradation of, 438-444
effect of, on rate equations, 923-924 external potential, 27
of electrolytes, 638 free, 446-448
of gases, 37 internal, 201, 448
Dissolution, 111, 121, 123 of activation, 816, 908
see also Heat of solution of rotation, 766
entropy of, 687 of translation, 766
heat of, 243-247, 273-282 of vibration, 767
kinetic theory of, 111~112 potential, 27
processes of, 111-112, 121-123 Energy balance, 95-204
Distillation, inventory changes in, 200 differential, 538
Distribution, calculations of, 140-141 enthalpy terms in, 307-308
of a solute between immiscible liquids, input items, 307-308
139-140 of acid chambers, 400
product, 860 of a process for the production of
Distribution coefficients, 139 hydrogen, diagram, 208
Double stepwise integration, 1036,1096- of blast furnace, 414-420
1097 of chemical processes, 204-206
Driving force, 28 of coal-fired boiler furnace, 348-356
Dry-bulb temperature, 98 of combustion processes, 342-344,355-
Dual adsorption sites, 913 356, 369-371
Dual-Bite mechanism, 929-930 of gas producer, 369-371
Diihring lines, 65, 83 of Gay-Lussac tower, 403
chart, 83 of Glover tower, 393
Dulong's formula, 327 of intermittent processes, 369-371
of metallur^cal processes, 406-408,
Economic design, 1009 414-420
Eddy currents, 974 of pyrites burner, 389
Effective diameter, 986-987 of sulfuric acid plant, 405-406
SUBJECT INDEX XXXlll
Energy functions, 448-449, 453 Entropy, change in heat exchangers, 566

change in isothermal nonflowprooesses, changes, with phase change, 445
448-449, 540 with temperature change, 445
differential, 454-455, 472 empirical correlation of, 757
partial derivatives of, 458-469, 472 internal rotational contributions,
relation to pressure, volume, and 781
temperature, 462-463, 473 of activation, 810, 965
shaft work from, 448-449, 540 • prediction of, 889 \
Energy levels, 768 of an ideal gas, 475
Energy units, 202-203 of benzene, chart, 627
conversion of, 438 of dissolution, 687
Engine, Carnot, 557-660 of ethylene, 783
efficiency of, 566 of gaseous mixtures, 614
performance of, 567 of gases, effect of pressure, 496-499
Enthalpy, 207, 448, 711 of hquid solutions, 614
as a criterion of equilibrium, 462 of rotation, 787
charts, 244, 291 of separation, 686
composition diagrams, 279-281, 300 of solutions, 614
of air-water vapor, 243-245 . physical concept of, 440
of ammonia-water, 584 quantitative definition of, 442
of calcium chloride solutions, 281 relationship to probability, 443
of hydrochloric acid, 279 Entropy correction for nonideality of
of silica gel, 300 gases, 496-499
of sulfuric acid, 280 chart, 498
evaluation of, 238-239 Entropy-pressure correction, for liquids,
internal contributions to, 788 509-610
of activation, 810, 965 chart, 610
prediction of, 889 for saturated fluids, 611
of adsorbed systems, 299-300 standard molal, 701-709
of an ideal gas, 475 temperature charts, 526
of benzene, chart, 527 Equations of state, 479-492
of calcium chloride, chart, 281 Beattie-Bridgman, 479, 482-483
of gaseous mixtures, 609 for mixtures, 598
of gases, effect of pressure, 492-496 Benedict-Webb-Rubin, 483
of humid air, 240, 243-245 van der Waals, 463, 479-482
of iron, 419 , , virial coefficients, 488
of liquids, effect of pressure, 606-509 Equilibria, bubble-pointj 672
of silica gel, chart, 300 criteria of complex, 614-619
of solutions, 609-613 dew-point, 673
of sulfuric acid, chart, 280 in liquid solutions, 741
of water vapor, 344 in phases, 617
relation to pressure, volume, and Equilibrium, adsorption, 151, 1096
temperature, 463, 473 ^calculation from line segments, 127
relative, 207, 709 criteria of, 449--453
Enthalpy pressure correction, for gases, in closed system, 449-453
492-i96 in complex reactions, 701
chart, 494-496 in simultaneous reactions, 729-741
differential at constant pressure, 516 physical, 644
differential at constant temperature, stable and unstable, 460.
616 vapor-liquid, 644
for liquids, 506-609 for ammonia-water, 684
chart, 608 for ethane-heptane system, 601-604
differentia], 612-613 for mixtures, 600
for saturated fluids, 611 Equilibrium composition, 714-725
Entropy, 440-446 Equilibrium concentrations of pentanes,
as a criterion of equilibrium, 450 731
XXXIV SUBJECT INDEX
Equilibrium constant, 691-713 Evaluation of constants in rate equation,

effect, of pressure on, 693 962-9BS
of small errors on, 7S6 Evaporation, 55
of temperature on, 695 inventory changes in, 180
in ionic dissociation, 638 multiple effect, 590
of chemical reaction, 712 vapor recompression, 590-592
over-all, 9^6 Exact differential, 442
vaporization, 663 equations, 455-457
Equilibrium conversion, charts, 720- in Jacobian notation, 467
721 Excess reactants, 10
effect, of dilution, 719 Exothermic compounds, 252
of excess products, 719 reactions, 251
of excess reactants, 719 equilibrium conversion in, 726
of pressure, 719 Expansion, adiabatic, 501, 551-553
of reaction conditions, 718 fiovr, 651-553
of temperature, 718 nonflow, 553-556
in endothermio reaction, 726 free, 439, 442, 551-556
in exothermic reaction, 726 Joule-Thomson, 661-663
of ammonia, 718 Maxwell, 553-556
Equilibrium distribution, 145, 106S of fluids, 538-594
Equilibrium moisture content, 158 of liquids, 502-606
chart, 152, 158 of solutions, 606-609
of activated alumina, 158 work of, 640-656
of cotton, 152 isentropic flow, 540, 548-560
of glycerin, 158 isentropic nonflow, 640, 545-547
of kaolin, 152 isothermal flow, 540, 542-644
of leather, 152 isothermal nonflow, 640, 641-642
of paper, 152 Experimental data, thermodynamic
of pulp, 162
properties from, 518-521
of silica gel, 152, 138, 1088 Extensive property, 2, 444-445
of silk, 152 External potential energy, 2
of triethylene glycol, 168 External rotation, 283, 777
of viscose, 162 Eyring equation, 809
of wool, 152 Eyring theory, 956
Equilibrium-reaction temperature, high
pressure, 728 Fahrenheit temperature scale, 32
Equilibrium vapor pressure, 81 Fanning equation, 984, 1016
Equivalent compression efficiency, 573 Fans, 669-570
reactor volume, 88J^ Feasibility of reactions, 714
Erg, 202 Ferric chloride water system, 118
Errors of generalized charts, 518-519 Film concept, 974
Escaping tendency, 619 Film thickness, 974
Estimation of physioochemical proper- effective, 977
ties, 605-506 for heat transfer, 982
Ethane, decomposition of, 8^3 Fine powders, heat of wetting, 293
Ethane-heptane phase diagrams, 601- Fixed carbon in coal, 324
604 Flame temperature, actual, 313
Ethanol-benzene, boiling points, 655 effect of preheating upon, 309, 311-313
Ethyl acetate-ethyl alcohol azeotrope, in air, 312
663 maximum calculated, 314
Ethylene, dimerization of, 811 of gases, table, 336
entropy of, 783 theoretical, 312
moment inertia of, 783 Flow processes, 204, 832
Eutectic composition, 113 shaft work in, 539
Eutectic point, 113 isentropic, 640, 548-560
Eutectic temperature, 113 isothermal, 640, 642-544
SUBJECT INDEX XXXV
Fluid catalytic cracking process, 931 Fugacity, as related to pressure, 620

Fluidized catalyst bed, 1005 coefficients, 621, 716
Fluidized reactor, 1032 chart, 622
Fluidized state, 931 ratio, 715-718
Fluids, expansion and compression of, effect of pressure and temperature on,'
638-592 623
thermodynamic properties of, 479-535 Furnaces, coal-fired, 354-356
Fltix, in blast furnace, 411 energy and material balances of,
Fouling, 932 354-356
factors of, 932-9S4, 959, 961 Fusion, heat of, 227
of catalyst, 930, 1006 table, 228
Fractional crystallization, 128
Fraction of moisture removed in bed of Gas, adsorption of, 1080-1105
silica gel, 192 by solids, 149
Free energy, 446-447 evaluation of constants in, 109S
as a criterion of equilibrium, 452-453 initial distribution, 1096
change, effect of temperature at con- rate of composition change in, 1092-
stant pressure, 463, 473 1093
data, presentation of, 699 constants of, 32-33
measurement of, 698 applications of, 33
of activation, 902 units of, 33, 438
of formation of hydrocarbons, 698 densities of, 35
of mixing, 647-651 ideal, compression of, 439
relative, 709 perfect, properties of, 476-476
Free expansion, 439, 442, 551-656 •producers of, 364
adiabatic, 601, 561-563 energy and material balances of,
flow, 501, 551-663 366-371
nonflow, 563-656 temperature scale, 444, 473-474
Free radicals, 843 Gaseous dissociation, heat of, 273
Free rotation, 766 Gaseous fuels, 338, 341
Freezing-point curve, 114 energy and material balances of, 342-
Freon, 234, 237 344
Frequencies of vibration, 769 Gaseous mixtures, 37
Frequency factor, 816 * Beattie-Bridgman equation for, 598
Freundlich isotherm, 154 changes of composition and volume,
Friction factor, 1016 42
Fuel gas, 338 composition by volume, 13
heating value of, 338 compressibility of, 595-605
table, 340 critical phenomena of, 599-605
hypothetical composition of, 339 critical point, 600
Fuels, 323 density, 41
combustion of, 338 enthalpy of, 609
chart, 372 entropy of, 614
energy balance of, 342 in chemical reactions, 46
heating value of, 327-328, 340 molecular weight, average, 40
heat of incomplete combustion of, 341 partial pressure of, 45
material balance of, 342 standard states, 628
Fugacities, 715 Gases, actual, behavior of, 479-502
at equilibrium, 619 average molecular weight, 40
in ideal solutions, 626 Beattie-Bridgman constants for, 483
in solution, 621 Benedict-Webb-Rubin constants for,
of gases, 619, 667 483
of pure liquids and solids, 621 compressibiUty of, 479-492
related to partial molal free energies, compression of, 568-675
'625 critical constants of, 234, 237, 481
Fugacity, 619-627 dissociating, 37
XXXVl SUBJECT INDEX
Gases, enthalpy correction due to pres- Glover tower, energy and material bal-
sure, 492-496 ances of, 390-398
differential at constant pressure, 515 Glycerin, as a desiccant, 158
differential at constant temperature, heat of mixing, 288-289
516 Glycerin-water, partial enthalpies, 289
enthalpy of, 475 Gram-atom, 6
entropy correction due to nonideality, Gram-mole, 6
496-499 Granular beds, adsorption in, 1085
entropy of, 475, 496-499 concentration gradients in, 1097
fugacities of, 619, 667 pressure drops in, 1015-1019
heat capacity of, 213, 217, 474-513 unsteady heat transmission in, 1099
ideal behavior of, 33 Graphical calculation of combustion
deviation from, 37 problems, 371
in liquid solution, standard state, 630 Graphical integration, 1012
in reaction processes, 46 Graphical reactor design, 884
internal energy of, 475 Ground level, 768
liquefaction of, 589-590 Ground state, 771
pressure correction to heat capacity of, Group contributions, 758-765
499, 501 to thermodynamic properties, 758
pVT relations of, 479-492
solubility of, 146-147, 681 Harmonic oscillation, 792
thermodynamic properties of, 474-476 Heat, 28
van der Waals constants for, 481 of adsorption, 297-299, 914
Gas law, satisfactory use of, 479 chart, 298
Gas-liquid reactions, 1053 differential, 298, 684^686
Gas-soUd adsorption equilibria, 684-686 integral, 297-300, 684-686
Gas-solid reactions, 1053, 1063-1067 effect of temperature on, 299
Gauge pressure, 35 of combustion, 337
Gaussian law of error, 939 effect of allotropic forms, 324
Gay-Lussac tower, 383 effect of surface, 324
energy and material balances of, 403- in calculating heat of formation,
405 265, 267
GeneraUzed bond frequencies, 798 of petroleum, 337-342
charts, errors of, 518-519 of hydrocarbons, 337-342
thermodynamic diagrams from, 622- of various forms of carbon, 262, 323
535 standard, 260-261
Generalized compressibility factors for tables, 262-264
gases, 488-492 of condensation, pseudo-, 610
chart, 489 of dilution, 274
use of, 490-492 of formation, 252-253
GeneraUzed enthalpy corrections, for from heats of combustion, 265, 267
gases, 492-496 of allotropic elements, 258
differential, 512-513 of carbon compounds, 260
at constant pressure, 515 of cementite, 417
at constant temperature, 516 of compound in solution, 274
for liquids, 506-509 of ions, 270-272
GeneraUzed entropy corrections, for table, 272
gases, 496-499 of slags, 417
for Uquids, 509-510 tables, 253-257
GeneraUzed heat capacity corrections, of fusion, 227-228
for gases, 496-501 tables, 228
for Uquids, 510-511 of gaseous dissociation, 273
GeneraUzed liquid densities, 502-506 table, 273
Gibbs-Duhem equation, 645 of hydration, 277
Gibbs phase rule, 617 of mixing, 278, 282-283
Glover tower, 383 of neutralization, 268-270
SUBJECT INDEX xxxvu
Heat, of reaction, 249-252 Heat capacity, of gases, chart, 600
at constant volume, 302 mean molal, 217
calculation of, 266-268 true molal, 213
effect of pressure on, 302-303 of hydrocarbons, 334-336
effect of temperature on, 803 of ideal gas, 513
standard, 250 of liquids, 224-227, 510-518
under changing pressures and vol- chart, 511
umes, 302 pressure correction for, 510-611
of sofution, 253-257, 273-282 ta6fe, 226-227
charts, 276-276 of monatomic gases, 211
differential, 291-293 of nitric acid, 224
integral, 274 of organic vapors, 335
of hydrates, 277 of refractories, 219, 223
of sulfuric acid, chart, 275 of saturated liquid, 611-513
tables, 253-257 from Cp of its ideal gas, 613-618
of transition, 229, 419 of saturated vapor, 611-613
table, 229, 419 of solids, 217-223
of vaporization, 229-238, 337 tables, 221
at critical temperature, 233 of solutions, 224
at normal boiUng point, 230 charts, 224-225
effect of pressure on, 232-238 partial, 396
effect of temperature on, 233 of sulfides, chart, 225
from Clapeyron equation, 230 of sulfuric acid, chart, 396
from empirical vapor-pressure equa- ratio of, 462, 473
tions, 231 thermodynamic relations of, 459-462
from Kistyakowsky equation, 230 472
from reduced reference plots, 235 types of, 459, 472
from reference substance plots, 232 Heat carrier, WSS
from Trouton's rule, 230 Heat exchangers, 666-566
of hydrocarbons and petroleum finned, 10S3
fractions, 336-337 flat-plate, 1038
of refrigerants, 237 Heating value, 323
of wetting, 293 of coal, 326
differential, 294-297 of paraffin hydrocarbons, 338
integral, 294-297 of unsaturated hydrocarbons, 262
table, 294 relation to flame temperature, 314,
units, 203 340
with water, figure, 295 Heat reservoir, 437-438
Heat capacity, 210-212 Heat transfer, 982
at constant pressure, 211-217, 459,472 coefficient of, 1099
at constant volume, 210, 459, 472 in catalyst beds, 973
atomic, 217 in granular beds, 1099
differences of, 461, 473-474 lateral, 1033-1036
effect, of pressure on, 214, 462, 472 Height, of catalytic unit, IOI4
of temperature on, 213 of reactor unit, IOI4
of volume on, 462, 472 of transfer unit, 98Jt, 987, 992
English units of, 213-217 in gas adsorption, 1086
equations, English units, 213 Henry's constant," 147, 682
metric units, 214 chart, 147
internal rotational contributions to, deviations from, 148
790 law, 146-148
mean values, English units, 214-217 Heterogeneous reactions, 805
metric units of, 216 thermodynamics of, 722
metric units of, 214 with gaseous product, 1063, 1068
of gases, pressure correction to, 499- High-pressure vapor-liquid equilibrium,
601 663-677
XXXVUl SUBJECT INDEX
Hindered rotation, 766, 787 Ideal gas, properties of, 474-476

Homogeneous reactions, 805 temperature scale of, 444, 473-474
in separate phases, 1063 thermodynamic properties of, 474-476
Houdry process, 931 Ideal solubility, 682
Humid air, enthalpy of, 240-245 Ideal solutions, fugacities in, 626
heat capacity of, 241 Ideal system, 606
Humidification, adiabatic, 242 Immiscible liquids, 76, 139
Humidity, 89, 93 Incomplete reactions, thermochemistry
chart, construction of, 98, 99 of, 301
pressure correction to, 103 Independent reactions, 733
eSect of carbon dioxide vapor on, 243 Induction period, 844
high-temperature range, 100 Industrial reactions, thermochemistry of
low-temperature range, 101-103 323, 389, 414
molal, 99 Inexact differential, 442
percentage of, 93 Infinite dilution, standard state, 629
relative, 93, 136 Inhibitors, 844
Hydrates, 115 Integral and differential heat of wetting
heat of solution of, 277 of silica gel, figure, 296
Hydration, heat of, 277 Integral-conversion data, 968-969
Hydraulic radius, 1016 Integral-conversion equations, 822
Hydrocarbons, free energies of formation Integral heat, of adsorption, 297-300,
of, 697-698 684-686
heat capacities of, 334-336 of solution, 274
vaporization of, 336-337 charts, 275-276
heating value of, 338 of wetting, 294-297
specific heats of, 335 Intensive property, 2
vaporization equilibrium constants of, Interatomic spacing, 904
670-671 Intercooler, 1028
vapor pressure of, 66, 73, 76 Intercooling, 573
Hydrochloric acid, enthalpy of, 279 Interface, 973
Hydrogen, atomic activated adsorption Internal contributions to enthalpy, 788
of, 94S-949 Internal energy, 201, 488
Hydrogenation, effectiveness factor in,' as a criterion of equilibrium, 451
1000-1001 kinetic theory of, 201
of bromine, 8W, 841-865 of an ideal gas, 476
of codimer, 9Jfi-958 relation to pressure, volume, and
of iso-octene, 943 temperature, 462, 473
of olefins, 1042 Internal rotation, 766, 786-791
Hydrogen balance, 349 contribution, to entropy, 789
Hydrogen bromide, formation of, 820, to heat capacity, 790
841, 866 hindered, 787
Hydrogen content, net of coal, 328 potential barriers hindering, 790
of petroleum, 333-334 Intermediate product, effect on equilib-
Hydrogen iodide, formation of, 893 rium conversion, 732
Hydrolysis of methyl acetate, 8Z7 Interpretation of rate data, 949
Hygrometry, 98 Intrinsic property, 440
Hysteresis, adsorption, 156-157 Invariant point, solubility of carbonates,
745
Ice, vapor pressure of, 62 Inventory changes, 179-182
Ideal behavior of liquids, 605 Ionic activity coefficients, 637-640
Ideal gas, 27, 33 Ionic dissociation equilibrium, 638
compression of, 439 Ionic molality, 638
heat capacity of, B13 Ions, heat of formation of, 270-272
law, range of appUcability, 48, 89 Iron, enthalpy of, chart, 419
satisfactory use of, 479 Iron oxide reduction, 722
molecular energy of, 766 Isentropic compression efficiency, 570
SUBJECT INDEX XXXIX
Isentropio compression efficiency, work Kirchhoff's equation, 230

of expansion, flow, 540, 548-550 Kistyakowsky equation, 230
nonflow, 540, 545-547 Kopp's rule, 219
Isochors, 599
Isolated system, 1, 437 Laboratory results, interpretation of, 844
Isomerization of pentanes, 730 Laminar flow, 973
Isomerization reaction, 913, 1013 Latent heat, 227-229
Isometrics, 599 see also Heat of transition; of vapori-
Iso-octane, hydrogenation of, 9 ^ zation
Isopropyl alcohol in water, equilibrium of expansion at constant temperature,
distribution of, chart, 145 459, 472
Isopropyl ether-isopropyl alcohol, activ- of pressure change at constant temper-
ity coefficients, 645-647 ature, 459, 472
saturation temperature >of, 642 Lattice structure, 932
Isothermal adsorption, 1083 Least squares, method of evaluation, 938,
with complex equilibrium, 1094 941
Isothermal compression efficiency, 570 principle of, 939
Isothermal solubility curve, 143 Leather, 152
Isothermal work of expansion, flow, 540, Leduc's law, 38
542-544 Limestone, decomposition pressure and
nonflow, 540-542 temperature of, 723-724
Isotherms, adsorption of, 151-155 Limiting reactant, 10
chart, 153 Linde process, 590
of carbon dioxide, 480 Linear molecules, 768, 778
Line segments, calculations from, 127
j factor, 983, 987 Liquefaction, 53
Jacobians, 465-471 of gases, 589-590
properties of, 466 Liquid-liquid reactions, 105S-1053
table of values of, 468 Liquids, compressibility of, 502-606
use of, in deriving relations, 469-471 densities of, 18-20, 502-506
Joule-Thomson effect, 501 diffusion in, 981-982
Joule-Thomson expansion, 551-563 enthalpy of, correction due to pressure,
605-509
•Kaolin, 152 differential, 512-513
Kelvin temperature scale, 32 entropy of, correction due to pressure,
1. eselguhr, 906 509-510
Kilocalorie, 203 expansion of, 502-506
Kinematic viscosity, 870 generalized densities of, 502-506
Kinetic energy, external, 28 heat capacity of, correction due to
internal, 28 pressure, 510-511
molecular, 28 ideal behavior of, 605
Kinetics, 437 properties of, 803
of combustion, 1068-1073 saturated, heat capacities of, 511-513
Kinetic theory, 28 from Cp of their ideal gases, 513-518
extension of, 31 solubility of, 677
of adsorption, 149-150 specific gravity of, 17-22
of capillary condensation, 150 standard states of, 628
of condensation, 56 thermodynamics of, 502-518
of dissolution, 111-112 Liquid-solid reactions, 1062,1074
of distribution, 139 Liquid solutions, 142-146
of gases, 28-32 compressibility of, 605-609
of Uquids, 53-54 densities of, 606-608
of solubiUty, 111-112 enthalpy of, 610-613
of gases, 146-149 entropy of, 614
of translation, 30 equilibria in, 741
of vaporization, 55-57 Liquid state-'' , 502, 518 .
xl SUBJECT INDEX
Liquid state, diffusion in, 981-982 Miscible hquids, partially, 142

Lithium chloride, solubihty of, 158, 162 Mixed acids, 168-169
Longitudinal diffusion, 1002 activity coefficients of, 1056
Longitudinal mixing, 1005 densities of, 1057
Mixing, entropy change in, 666
Magnetic susceptibility, 604 free energy of, 647-650
Mass, and heat transfer, coeiEcients, heat of, 278, 282-283
982-987 Modulus ratio, 999
effects of, 1068 Thiele, 997-998, 1000
in catalytic beds, 973 Moisture, in coal, 324-325
problems, 1008-1008 in combustion gases, 351, 361
conservation of, 1 removed from air by bed of silica gel,
relations of, 4 plot, 192
transfer of, 906 Molal heat capacity, 211-217
volume relations of, 7 see also Heat capacity
Material balance, 1, 167 bonding frequencies, 799-800
in gas adsorption, 1081 Molal humidity, 99
of acid chambers, 399 charts, 100-101
of blast furnace, 406^14 Molal units, 68
of chemical processes, 2S&-310 Molal volume, 34
of coal-fired furnace, 348-356 at standard conditions, 484
of fuels, 342 Molecular activity, apparent, 638
of gas producer, 366-369 Molecular configuration, 786
of Gay-Lussac tower, 402 Molecular constants, 776
of Glover tower, 390 Molecular contributions, 776
of metallurgical processes, 407-420 Molecular degree of freedom, 767
of pyrites burner, 388 Molecular energy, 766
of sulfuric acid plant, 484-488 Molecularity of reaction, 806
Maximum conversion, 1029 Molecular speed, 29-32
Maxwell-Boltzmann distribution law, Molecular volume, 988
772 Molecular weight, 40
Maxwell expansion, 653-556 average, 40
Maxwell relations, 457-458, 472 of paraffins, 607
_ Mean compressibiUty factors, 595-598 of petroleum, 331-332
Mean ionic molality, 638 Mole fraction, 13
Mechanical work, 538 Mole percentage, 13
Mechanism of reactions, catalyzed, 906 Mollier chart, 522
determination of, 947 for ammonia, 526
Metallurgical reactions, 736-741 for benzene, 528
energy balance of, 407-420 Moments of inertia, 779-780, 783, 787
thermodynamics of, 736 Monomoleoular reactions, 918-919
Metastability, -120 Moving bed, 931
Methane decomposition, effect of steam Multiple-bond contributions, 761
upon, 734-735 Multiple-effect evaporation, 590
equilibrium in, 734 Multistage compression, 572-676
Methanol conversion, 730 in counterflow, 1061
Method, of least squares, 949 operation of, 1028
of tangent intercepts, figure, 291 reaction, 1029-1031
Methyl acetate, hydrolysis of, 827
Methyl alcohol, configuration, 785 Naphthalene solubiUty, 679
heat capacity of, 799 in benzene, 112
internal rotational contributions to, chart, 113
791 Net heating value, of fuels, 327
properties of, 783, 802 of hydrogen in coal, 238
total properties, 801 Neutralization, heat of, 268-270
Methyl substitution, 759 Nickel catalyst, 905
SUBJECT INDEX xli
Nitration, of aromatics, 1054-1058 Paraffins, vapor-pressure constants of,
of benzene, 1054-1060 figure, 74
Nitric acid, heat capacity of, chart, 224 table, 75
Nitric oxide, decomposition of, 813 Parallel heat exchange, 566
Nitrogen, and hydrogen, compressibility Partial derivatives of energy functions,
of, 596-697 458-459, 472
atmospheric, 42 Partial differentials, 286
compressibility factors of, 484-487 Partial enthalpies, 285
Nitrogen balance, 351 calculations of, 287-293
Nitrogen peroxide, adsorption of, 685 of glycerin in water, 289
Nitrous oxide, decomposition of, 8S8 of water in glycerin, 289
Nonadiabatic reactions, 311 Partial heat capacities, 285
Noneleotrolytes, solubility of, 677 Partial heat of adsorption, 298
Nonelectrolytic solution, standard state, Partially miscible liquids, 142-146, 661,
724 688
Nonflow processes, 204,1081 Partial molal free energy, 616
reactions of, 822 Partial pressure, 45
shaft work in, 539 Partial vaporization, 673, 675
isentropic, 540, 545-647 Partial volume, 285
isothermal, 640-542 Particle density, 996
Nonideal system, 606, 608-609 Particle size, D'p, 999
Nonisothermal reactions, unsteady state, effect of on solubility, 119
1100-1105 Partition functions, 771-776, 786, 792
Nonlinear molecules, 768 thermodynamic properties from,
Nonpolar compounds, 68 773
critical temperature of, 68, 71-76 Path properties, 453
Nonvolatile solute, 82 Pebble-bed regenerators, 1100
Normal molal volume, 34 Pentane isomerization, 730
at standard conditions, 484 Pentanes, equilibrium concentrations,
Number, of reactor units, 1014 731
of transfer units, 991 Percentage excess reactant, 10
Percentage humidity, 93
Olefins, hydrogenation of, 1043 Percentage saturation, of gases, 91
Open system, 437 of solutions, 114
Optimum intermediate pressure for Perfect gas, properties of, 475-476
multistage compression, 673-574 Performance chart, in water softening,
Optimum reactor design, 883 1078-1079
Order, complex, 840 use of, 1079
of reaction, 806 tests, 904
complex, 819 Petroleum, 329-337
Ore. in blast furnace, 409 boiling points of, 332
Oscillators, 792 characterization factor of, 329
Osmotic pressure. 111 cracking of, catalytic, 433-436
Outgassed carbon, 154-155 thermal, 421
Over-all reaction rate, 1054 critical temperatures of, 333
Oxides, specific heats of, 219 heat capacity of, 334-336
Oxygen balance, limitations of, 354 combustion of, 337-342
in coal, 354 vaporization of, 336-337
heating value of, 338
Paper, 152 hydrogen content of, 334
Parachor, 70-71 molecular weights of, 331, 333
Paraffins, boiling points, 607 specific gravity of, 330-331
critical constants of, 74-75 specific heat, of hquid, 334
heating values of, 338 of vapor, 335
physical properties of, 507 viscosity of, 332
vapor-pressure constants of, 144 p/ factor, 977, 987, 992
xlii SUBJECT INDEX
Phase change, entropy variation with, Pressure, of decomposition, 723

445 of solutions, 136
Phase equilibria, 617 by Cox method, 65-66, 83, 160
Phase rule, component in, 619 by Dilhring method, 65, 83
Gibbs, 617 of vaporization, 723
Phenol, solubility of, 142 optimum, lOW
Phosphine, decomposition of, 865 reduced, 488
Photon, 768 Primary methyl group contributions, 769
Physical equilibrium, 644-690 Probability, 443
Pig iron, enthalpy of, 418 Problems, behavior of ideal gases, 49-52
chart, 419 catalytic reactions, 971-972
metallurgy of, 406-420 catalytic reactor design, 1047-1050
Planck's constant, 769 chemical equilibrium, J48-755
Point properties, 453 metallurgicaland petroleum, 431~436
Poise, 870 expansion and compression, 692-594
Poisoning factor, 1015 fuels and combustion, 373-382
of catalyst, lOTJi. homogeneous reactions, 894-901
Poisons, 90/^, 926 humidity and saturation, 105-110
Polar compounds, 68 mass and heat transfer in catalytic
Polar liquids, vaporization of, 233 beds, 1006-1008
Pore radius, effective, 998 material balances, 195-200
Porosity of catalyst, 907 physical equilibrium, 688-690
Porous catalyst, diffusion in, 995-1002 solubility and sorption, 160-166
Potential barrier, 766, 786-788, 801 stoichiometric principles, 23-26
hindering internal rotation, 790 thermochemistry, 314^322
Potential energy, 27 thermodynamic principles, 476-47§
Potentials, chemical, 614-616 thermodynamic properties, from mo-
thermodynamic, 437, 448 lecular structure, 8O4
Pound-atom, 6 of fluids, 635-637
Pound-mole, 6 of solutions, 641-643
Power-plant cycles, 660-568 thermophysics, 245-248
Prandtl number, 98Ji., 986, 989-990 uncatalyzed heterogeneous reactions,
Preferential adsorption, 157 1106-1107
Pressure, absolute, 35 vapor pressures, 85-88
correction of, for enthalpy of gases, Process, 437
492-496 cyclic, 656-592
at saturation, 611 flow, 204
differential, at constant pressure, 615 shaft work in, 639
differential, at constant temperature, isentropic, 540, 548-650
616 isothermal, 640, 642-644
for enthalpy of liquids, 606-509 irreversible, 439
at saturation, 611 nonflow, 204
- for entropy of gases at saturation, shaft work in, 539
611 isothermal, 540-542
for entropy of liquids, 609-510 isentropic, 540, 545-547
at saturation, 611 reversible, 438-139, 447
for heat capacity of gases, 499-501 spontaneous, 439
for heat capacity of liquids, 510-511 steam, 565
distribution of, 1019 Process rates, diffusion in, 990-993
drop of, 869 Producer gas, 364-371
in granular beds, 1015-1019 Product distribution, 860
effect of, on density of liquid, 502-506 Propane dehydrogenation, 189-190
on equilibrium conversion, 728 Properties, derived, 453
on heat of reaction, 302 extensive, 444-445
factor of, in diffusion, 979, 987 intrinsic, 440
gauge for, 35 of fluids, thermodynamic, 479-535
SUBJECT INDEX xliii
Properties, of paraffins, 507 Reaction temperatures, adiabatic, 725-

path, 463 729
point, 453 optimum, 1020
reduced, 488 Reaction velocity constant, 806
reference, 446, 463 Reactivation of catalyst, 932
Proportionality factor in diffusion, 978 Reactor, adiabatic, 864, 1011, 1027
Proximate analysis of coal, 324 catalytic, 1009
Pseudocritical point, 604-606 isothermal, 867
Pseudo-firs1>order behavior, lOOS nonadiabatic, 874
Pseudo-firstôrder reactions, 962-964, Reactor design, graphical, 864
967 Reactor size, 1012, 1021
Pseudo heat of condensation, 610 Reactor unit, concept of, lOlS
Pseudounimolecular behavior, 1015 height of, 1014
Psychrometric chart, 244 number of, IOI4
Psychrometry, 98 Reciprocating compressors, 571-672
Pulp, 152 Recirculation of air in drying diagram,
Pumice, 905 183
Pure-component method, 44 Recycle ration, 186-190
Pure-component volume, 38 Recycle stock, purging of impurities in,
Purging of process, 197 191
•pVT relations, of carbon dioxide, 479- Recycling fluid streams, 82-90
480 Reduced conditions, 54
of energy functions, 462-463, 473 Reduced pressure, 54, 488
of gases, 479-492 from reference substance plot, 232,
of nitrogen, 484-487 235-238
Pyrolysis, 966 Reduced temperature, 54, 488
of benzene, 848, 876 Reduced volume, 54, 488
Reference properties, 446, 463
Quantum mechanics, 777 Reference substance, 16
Quantum numbers, 768 plot of, 64
Quantum theory, 768 equal pressure, 64
Quantum weight, 772 equal temperature, 65
Reference temperature, 200, 342, 355
Random dense arrangement, 1016, 1018 Reflux accumulation, 200
Random loose arrangement, 1016, 1018 Refractories, heat capacity of, 219, 223
Rankine cycle, 661-565 Refrigerants, 237
improvements on, 564-665 critical temperatures of, 237
Rankine temperature scale, 32 heats of vaporization of, 237
Rank of coal, 326 vaporization factors of, 235
Raoult's law, 80, 626 vapor pressure of, 235-238
Rate-controlhng step, 907 Refrigeration, 575-590
Rate data, differential, 942 absorption, 676, 681-588
Rate equations, constants for, 957, adsorption, 588
962 capacity, 677
evaluation of, 937 compression, 575-681
differential, 816 cycle of, 576-579
in concentration units, 820 work in, 678
Rate of catalytic oxidation, effect of mass Refuse in combustion, 348, 359
and heat transfer, 10^7 Regeneration of catalyst, 1006 .
Rates, independent of conversion, 1102 Regenerative adiabatic operation, 931
Ratio of heat capacities, 462 Regenerative heating in cyclic processes,
Reaction, heat of, 249-252, 266-268, 565
302-303 Regenerators, pebble-bed, 1100
Reaction rates, absolute, 808 Reheating, in cyclic processes, 564
effect, of conversion on, 1038 Relative humidity, 93, 136
of temperature on, 1038 of solutions, 137
xliv SUBJECT INDEX
Relative saturation, 91 Single-site mechanism, 929

Relative volatility, 657 Single-stage compression, 570-571
Retrograde condensation, 600-601 Sintering of catalyst, 906
Reverse Carnot cycle, 576 Slab, catalysts in, lOS^
Reverse reactions, 818 Slag, enthalpy of, 416, 418
Reversibility, 438-440 chart, 419
of adsorption, 156 heat of formation of, 417
of gaseous adsorption, 146 Sodium carbonate-sodium sulfate-water
Reversible cycle, 557 system, 129-133
Reversible process, 438-439, 447 chart, 130, 132
Reynolds number, 986, 987, 1016-1017 Sodium chloride, density of, 18
Rotational energy, 766 entropy of separation, 687
solutions of, activity coefficients in,
Sackur-Tetrode equation, 770 640
Salting out, 123 work of separation, 687
Saturated liquid, heat capacity of, 511- Sodium hydroxide, Duhring lines for, 83
513 Sodium sulfate, solubility of, 118
from Cp of its ideal gas, 513-518 Softening of water, 1074-1080
Saturated solutions, 112 SoUd catalysts, 903
Saturated vapor, 57, 89 Solid-liquid systems, 283
heat capacity of, 511-513 Solids, densities of, 996
Saturation, partial, 91 heat capacities of, 217, 223
percentage of, 91, 114 properties of, 803
temperature of, 121 solubilities of, 677
of binary systems, 648 standard states of, 627-629
Schmidt number, 986-987 vapor pressures of, 58
Secondary methyl substitutions, 761 Solid-solid reactions, 1062
Second law of thermodynamics, 439 Solubility, 111-112, 677-684
Seeding, 211 charts, 113, 115, 118, 130, 132
Selectivity, 860, 98§, 1020 diagram of, isometric, 132
of catalyst, 903, 1006 effect, of pressure on, 683
Separation, entropy of, 686 of temperature on, 683
work of, 686 in complex systems, 129-136
Severity factors, 885 liquid-liquid, 142-146
Shaft work, 538 of calcium carbonate, 747
from energy functions, 540 of carbonates, 742-745
in flow processes, 539 of chlorine, 682
isentropic, 540, 548-550 of gases, 146-147, 681
isothermal, 640, 542-544 chart, 147
in nonflow processes, 539 of hthium chloride, 158, 162
isentropic, 540, 545-547 of naphthalene in benzene, chart, 113
isothermal, 540-542 of nonelectrolytes, 677
Shaw's method of derivation, 465-471 of phenol, 142
Silica-alumina catalyst, 9S3 of soUds, factors affecting, 679
Silica gel, 158, 906 with congruent points, 114
adsorption by, 1089-1091, 1094 without congruent points, 118
enthalpy of, 300 SolubiUty isotherms, 148
equilibrium moisture content of, 152, Solute, 111
158, 1088 Solution pressure, 112 *
heat of wetting, 296 Solutions, 80, 111-145
partial, 299 densities of, 606-608
vapor pressure of, 159 enthalpy concentrations, 278-282
Silk, 152 entropy of, 614
Simultaneous equations, solution of, 737 expansion of, 605-609
Simultaneous reactions, 818, 828 fugacities of, 621
equilibrium of, 729 solid, 130
SUBJECT INDEX xlv
Solutions, thermochemistry of, 273-274 Stepwise countercurrent extraction, proc-
thermodynamics of, 596-641 essing of, 177
vapor pressures of, 84, 136 Stoichiometric molality, 639
Solvates, 115 Stoichiometry, 1-2
Solvent, 111 Stoke, 870
Soot and tar in combustion, 364-371 Streamline flow, 978
Sorption, 111 Stripping of adsorbate gases, 156
Space velocity, 839, 928, 964, lOOZ, 1006 Structure of catalyst, 907
Spatial arrangement, 9S7 Sublimation, 58
Specific gravity, 17 mechanism of, 1064
see also Densities of solid, 106S
of gases, 35 Successive reactions, effect of, in com-
of liquids, 17-22 bustion calculations, 362-364
of petroleum, 330-331 thermochemistry of, 301
Specific heat, see also Heat capacity Sucrose solutions, activity coefficients in,
of acids in aqueous solutions, 224 636-637
of bases, 224 Suffix equation, 649-650
of calcium compounds, 220 Sulfates, specific heats of, 225
of chlorides, 225 Sulfur dioxide, catalytic oxidation of,
of coke, chart, 218 lOZl-1026
of elements, charts, 218 conversion, adiabatic equilibrium of,
of hydrocarbon gases, figure, 335 727-728
of nitrates, 225 \ oxidation of, conversion charts, 721
of oxides, chart, 210 equilibrium of, 716
of petroleum, liquid, chart, 334 Sulfuric acid, chamber plant, 383-406
vapor, 335 energy and material balance of, 385-
of sulfates, 225 388, 405-406
Specificity of catalyst, 902 density of aqueous solutions, chart, 384
Spectroscopic data, 791, 793 desiccant, 158
Spectroscopic notations, 775 enthalpy of, 280
Spontaneoul*process, 439 heat capacity of, chart, 396
Stable equihbrium, 460 solution, 275
Stagnant film, 976 vapor pressure of, 84
Standard conditions, 34 Sulfur trioxide, heat of solution, 291
Standard free energy, 693-713 Sulfuryl chloride, decomposition of, 825,
change of, 692, 694 837
effect of temperature upon, 695 Summary of thermodynamic relations,
Standard heat, of combustion, 260-266 472-473
of formation, 252-258 Superheat, 57-58
of reaction, 249 degrees of, 57
Standard states, 627 in cyclic processes, 664
for gases in liquid solution, 630 Superheated vapor, 57
for pure liquids and solids, 628 Supersaturation, 120-121
in gaseous mixtures, 628 Surface area measurements, 903-904
in infinite dilution, 629 Surface rate equations, 917
in nonelectrolytio solution, 628 limitations of, 927
Statistical methods, 765-798 Surface reaction, 908, 915
Statistical weight, 772, 793 controlling, 920, 9S2, 948-949
Steady-flow process, 1081 mechanism of, 920
Steam, minimum ratio, to methane, 736 monomolecular, 916
process, 665 velocity constant of, 966
superheated, vaporization with, 78 Surface tension, 71
Steam power-plant cycles, 660-568 Surface-volume relationship, 1018
Steam rate, 667 Suspended catalyst, 904
Stepwise countercurrent extraction, 177 Symbols, conversion of, 23
procedure for, 1010-1011 thermochemical, 250-252
xlvi SUBJECT INDEX
Symmetrical systems, activity coeffi- Thermodynamic charts, for generaUza-

cients in, 669 tions, 522-535
Symmetry number, 778, 783, 787 minimum data required for, 524
Systems, 1, 437 procedure, 524-536
closed, equilibrium m, 449-463 Thermodynamic data, experimental de-
ideal, 606 termination of, 518-521
isolated, 1 presentation of, 521-535
nonideal, 606, 608-609 Thermodynamic efficiency, 559
. for Carnot cycle, 559
Tangent intercepts, method of, figure, for Rankine cycle, 663
291 Thermodynamic energy functions, 448-
TCC catalytic cracking process, 931 449, 453
TCC catalytic reactor, lOSS Thermodynamic functions, 710
Temjjerature, 28 Thermodynamic potentials, 437, 448
change of, entropy variation with, 44B Thermodynamic properties, 453
flame, 313 from molecular constants, 776
see also Flame temperature from molecular structure, 756-803
ideal-gas scale, 444, 473-474 from partition functions, 773
of constant vapor concentration, 69 group contributions to, 758
of reaction, 308 of fluids, 479-535
adiabatic, 308-309 of an ideal gas, 474-476
nonadiabatic, 311 of a perfect gas, 475-476
reduced, 488 <« translational contributions to, 769
thermodynamic scale of, 444, 473-474 Thermodynamic relations summarized,
Temperature gradients, graphical calcu- 472-473
lation of, 1037 Thermodynamics, first law of, 28
in reactors, 1043-1047 of liquid state, 602-518
Ternary equilibrium, 144 of solutions, 595-641
Ternary liquid mixtures, 143 principles of, 437-476
Ternary solubility, 145 second law of, 439
Ternary systems, 659-660 third law of, 446, 774, 788
Tetrachlorethylene-isopropyl alcohol- Thermodynamic tables, 621-622
water system, chart, 144 Thermodynamic temperature scale, 444,
Theoretical reaction temperature, 308 473-474
Theory, of absolute reaction rates, 90B Thermofor catalytic reactor, 103S
of corresponding states, 488 cracking, 1008
Thermal capacity, see Heat capacity Thermoneutrality of salt solutions, 270
Thermal conductivity, 983, 989 Thermophysics, 201-248
effective, 1034 Thiele modulus, 997-1001
Thermal efEcienoy, 344, 371 Throttling, 551
based on net heating value, 345 Tie hne, 143, 145
based on total heating value, 344 Tie substance, 168, 174
of gas producer, 345 .Toluene production, 933, 935-938
cold, 345 Total heating value, of, fuels, 327
hot, 345 of hquid petroleum hydrocarbons, 338
Thermal energy, availability of, 660 Total thermodynamic properties, 801
Thermal properties, see Heat Total work function, 447-448
Thermochemical equations, 259-261 as a criterion of equilibrium, 462-463 ^
Thermochemistry, at standard condi- Transfer coefficient, 982-988
tions, 249-277 Transfer factor, 982-988
laws of, 268-260 Transfer of heat, 982
of solutions, 273-274 Transfer unit, 984-988, 1086
symbols and convention^ of, 25 number of, 991
Thermodynamic charts, 521-635 Transformation of variables, 1096
for ammonia, 523-526 Transition, heat of, 229, 419
for benzene, 527-528 chart, 229
SUBJECT INDEX xivu
Transition points, 118 Vapor pressure, of soUds> 58
Translational contributions, 766 of solutions, 84, 136
Translational energy, 766 of sulfuric acid, 84
Translation of energy, 29 of vater, t . - 6 3 , 236
Triangular charts, 21, 114, 132 table, 236-237
for densities, 21 of water on silica gel, 159
for solubilities, 114, 132 by Cox chart, 66
Triethylene glycol, 158 by Duhring method, 64, 83
Trouton's ratio, 230 tables, 62-63
Turbocompressors, 569-570 Vapor-pressure constants, 73
Turbulent flow, 973-974 for paraffins, 507 -^
Twaddell gravity scale, 20 Vapor-recompression evaporation, 6 "
Tyler standard mesh sizes, table, 10S6 692
Variables, transformation of, 1095 .Q
Ultimate analysis of coal, 324 Velocity, constant for, in adsorption, "
Uncatalyzed heterogeneous reactions, reaction of, 806
1050 in diffusion, 976
Uncatalyzed reactions, 968 of hght, 769
Unit principles, 1058 Vibrational contributions, 791-798
Unit processes, 105^ Vibrational energy, 767
Units, conversion of, 22 Vibration frequencies, 769, 801
Unsteady state, nonisothermal reactibil, Virial coefficients, 488
1100 "^ Jfiscose, 152
Viscosities of mixtures, 874
van der Waals adsorption, 150, 908-910 Viscosity, 869
van der Waals constants for gases, 481 of petroleum, 332
van der Waals equation of state, 463, Viscous flow, 973
479-482 Void fraction, external, 996
Vapor, saturated, heat capacity of, 5 1 1 - internal, 996
513 Volatile matter in coal, 324r325
superheated, 58 Volume changes, by partial pressU''®
Vaporization, 55 method, 44
equilibrium constant of, 663, 665 reduced, 488
heat of, 229, 238, 337 with change in composition, 42
of hydrocarbons, 670-671 Volumetric efficiency, 571
Vaporization pressure, 723 per cent, 13
Vaporization processes, 94
Vapor-liquid equilibria, for ammonia- Water, diffusion of, 980
water system, 584 enthalpy of, 240
for ethane-heptane system, 601-604 heat of vaporization of, table, 236
for mixtures, 600 pressure of, table, 62-63, 236
high-pressure, 663 reduced temperature of, 236
Vapor-liquid equilibrium, 644, 647 softening of, 1074-1080
Vapor phase cracking, 420-431 performance chart, 1078-1079
Vapor pressure, 53, 56, 73, 136 residual hardness in, 1077
calculations of, by Clapeyron equa- Wave number, 769, 793
tion, 59-61 Weber and York correction, 494
by Cox chart, 65-66 Weight, 12
by Diihring lines, 65 per cent, 12
effect of temperature upon, 59 Wet-bulb temperatures, 98
estimation of, from liquid density, 506 Wetting, heat of, 293-294
kinetic theory of, 53-54 Wool, 152
of hydrocarbons, 66, 73-76 Work, 27
of ice, 62 in multistage compression, 673-674
of organic compounds, 73 in refrigeration, 578
of refrigerants, 235-238 in single-stage compression; 670
xlviii SUBJECT INDEX
Work, in turbocompressors and fans, Water-butanol, activity coefficients ofj

. 669 645-647
mechanical, 538 saturation temperature of, 648
of reversible expansion, 640-566 Water rate, 667
of separation, 686
shaft, 538 X-ray measurements, 777
from energy functions, 640 X-ray pattern, 903
in flow processes, 639
isentropic, 540, 548-650 York and Weber correction, 494
isothermal, 540, 542-544
in nonflow processes, 539 Zinc, production of liquid, 739
isentropic, 540, 545-647 Zinc-ammonium chromate catalyst, 90S
isothermal, 540-542 • Zinc chromate catalyst, 905
total, function of, 447-448 Zinc oxide reduction, 737-741

BPT10819 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BPT10819 PDF

Uploaded by

Copyright:

Available Formats

CHEMICAL PROCESS PRINCIPLES

For Books in Chemical Engineering

HAROLD C . WEBER, D.SC.

Part One • MATERIAL AND ENERGY BALANCES

All Rights Reserved

P R I N T E D If* T H E U N I T E D STATES OF AMERICA

" In the following pages certain industrially important principles of chem-

c base of natural logarithms

No number of transfer units, gas film '

to/ electrical work dme by system

., ' DiMENSIONLESS NuMBEHS

iVp, Prandtl number —r- '

Nst Stanton number 7777

Nsc, Schmidt number •*=?-

A ' component A •> 1

p constant pressure r- : •'

The state of a system is defined by numerous properties which are

Nature of Chemical Compounds. According to generally accepted

When the atoms of the elements unite to form a particular com-

be concluded that individual molecules, each of which contains only

Atomic Weights: <• - '..•...,11.,, .. •

Molecular Weights: • i ' i'• '• • 1

The respective relative weights of the reactants and products may-

Volume Relations in Chemical Reactions. A correctly written reac-

Gram-atoms of an elementary substance = -—:—r-

Pound-atoms of an elementary substance = — r^ :

The actual number of atoms in one gram-atom' of an elementary sub-

the compound in a 1 : 1 ratio. Since the gram-atom as above defined

Accordingly, with respect to the reaction equation previously dis-

The Use of Molal Units in Computations. The great desirability of

It is urged as highly advisable that familiarity with molal units be gained

Basis of Calculation: 100 lb of hydrogen. • ~,-.--i

Excess Reactants, In most chemical reactions carried out in in-

as to the basis of calculation. However, since such a statement is of

METHODS OF EXPRESSING THE COMPOSITION

m = total weight of the system.

Weight Per Cent. The weight percentage of each component is

Weight per cent of A = — X 100

This method of expressing compositions is very commonly employed

This method of expressing compositions is almost always used for

Mole fraction of A = ——^ ,;,

Atomic Fraction and Atomic Per Cent. The general significance of

The summation of all of the atomic percentages for a given system is

Mass of Material per Unit Volume of the System. Various units

(a) The composition in weight per cent. • , .

Basis of Calculation: 1000 cc of solution.

(c) Composition in Mole Per Cent: .;

DENSITY AND SPECIFIC GRAVITY

Density is defined as mass per unit volume. Density'values are

Concentration, ft NaCl by Weight

Concentration, % NaCl by Weight, also grams NaCl per 100 cc.

FIG. 1. Densities of aqueous sodium chloride solutiqps.

are plotted the densities of solutions of sodium chloride at various tem-

where G is the specific gravity at 60/60°F. It is apparent from this

Gravities expressed on this scale increase with increasing density.

API Scale. As a result of confusion of standards by instrument

Percentage Ceurbon Tetrachloride

Fia. 2. Specific volumes of ternary solutions of carbon tetrachloride, ethyl dibromide

specific volume of 0.50. Similarly, base lines BC and CA represent

The line parallel to BC passing through h is 50% of the perpendicular.

Illustration 8. It is desired to calculate the final specific volume when 20 grams

In the preceding, a triangular chart covering the entire possible

4. Oxygen is prepared according to the equation 2KCIO3 = 2KC1 + SOs. What