Chemical Engineering

Thermodynamics
CHE F213

Solution Thermodynamics
 Contents
• Chemical Potential
• Partial Properties
• Ideal Gas Mixtures
• Fugacity and Fugacity Coefficient: Pure
species
• Fugacity and Fugacity Coefficient: Species
in solution
• Ideal Mixtures
 Solution Thermodynamics
• Real system in Chemical, Petroleum and
Pharmaceutical industries usually contains multi-
component gases and liquids undergoing composition
changes
• Develop the theoretical foundation for applications of
thermodynamics to gas mixtures and liquid solutions
• Introducing
– Chemical Potential
– Partial Properties
– Fugacity
– Excess Properties
– Ideal Solution
 Fundamental property relation
• The basic relation connecting the Gibbs energy to the
temperature and pressure in any closed system:
 (nG)   (nG) 
d (nG)    dP    dT  (nV )dP  (nS )dT
 P  T ,n  T  P ,n
– applied to a single-phase fluid in a closed system wherein no
chemical reactions occur.
• Consider general case of a single-phase, open system:

 (nG)   (nG)   (nG) 

d (nG)    dP    dT     dni
 P  T ,n  T  P ,n i  ni  P ,T , n
j
   (nG ) 
Define the chemical potential: i   
 ni  P ,T , n j

The fundamental property relation for single-phase fluid systems of variable

mass and composition:
d (nG)  (nV )dP  (nS )dT   i dni
i

When n = 1, dG  VdP  SdT    dx

i
i i G  G( P, T , x1 , x2 ,..., xi ,...)

 G   G 
V   S  
 P T , x  T  P, x

Solution properties, M
Partial properties, M i
Pure-species properties, Mi
 Chemical potential & Phase Equilibria
• Consider a closed system consisting of two phases (α
and β)in equilibrium:
d (nG)  (nV ) dP  (nS ) dT   i dni
i

d (nG)   (nV )  dP  (nS )  dT   i dni

i

nM  (nM )  (nM ) 

d (nG)  (nV )dP  (nS )dT   i dni   i dni

i i

Since two-phase system is closed d (nG)  (nV )dP  (nS )dT

Mass balance: dni  dni i  i
Multiple phases at the same T and P are in equilibrium when chemical
potential of each species is the same in all phases.
 Partial properties
• Define the partial molar property of species i:
  (nM ) 
Mi   
 ni  P ,T ,n j
– the chemical potential and the partial molar Gibbs
energy are identical: i  Gi

• Relation between Molar and Partial Molar

Properties nM  M ( P, T , n , n ,..., n ,...)
1 2 i

 (nM )   (nM ) 
d (nM )  
  dP   T  dT   M i dni
 P  T ,n   P,n i
  M   M 
d (nM )  n    T  dT   M i dni
 dP n Since ni = n xi
 P  T ,x   P, x i dni = xi dn + n dxi
d(nM) = n dM +M dn

 M   M 
ndM  Mdn  n 
  dP  n  T  dT   M i ( x i dn  ndx i )
 P  T ,x   P, x i

 
n and dn are
 M   M   
 dM  
 

 T ,x
dP  
 

 P, x
dT   M i dx i n   M   x i M i  dn  0 independent
 P T i   i 

 M   M  M   xi M i
dM    dP    dT   M i dx i  0 and
 P  T ,x  T  P, x i
i

dM   xi dM i   M i dxi
i i
nM   ni M i
i

 M   M 
  dP    dT   x i dM i  0 Calculation of mixture
 P  T ,x  T  P, x i properties from partial
properties
The Gibbs/Duhem equation
 Gibbs Duhem Equation
• General form:
 M   M 
  dP    dT   xi dM i  0
 P T ,n  T  P ,n i

• For the changes at constant T and P,

 x dM
i
i i T ,P 0

 n dM
i
i i T ,P 0
 A rationale for Partial Properties
• A solution/ mixture property is represented by
the sum of its parts (partial properties M i )
• Partial properties like solution properties are
function of compositions
– In the limit as a solution becomes pure i species,
both M and M i approaches to pure-species
property Mi 
lim M  lim M i  M i
xi 1 xi 1

– In the infinite-dilution limit i.e., mole fraction i

approaches zero, no general statement is valid
  
lim M i  M i
xi  0
Partial properties in binary solution
• For binary system
M  x1M1  x2 M 2

dM  x1dM1  M1dx1  x2 dM 2  M 2 dx2

Const. P and T, using Gibbs/Duhem equation

dM  M1dx1  M 2 dx2

x1  x2  1

dM
 M1  M 2 M 1  M  x2
dM
M 2  M  x1
dM
dx1 dx1 dx1
 Partial properties in binary solution
• Gibbs/Duhem Equation for binary System:
x1dM1  x2 dM 2  0

dM 1 dM 2
x1  x2 0
dx1 dx1
dM 1 x2 dM 2

dx1 x1 dx1
• Given an expression for M 1 ( x1 ) , integration of above
equation yields an expression for M 2 ( x1 ) that
satisfy Gibbs/Duhem equation.
Graphical representation of Partial
properties
 Illustration 1

If the molar density of binary mixture is given

by the empirical expression:
  a0  a1 x1  a2 x12
Find the corresponding expressions for V 1 and V 2

Example 11.2, 3 and 4

 Illustration 2
The molar volumes of a binary solution at 25˚C are
measured as given below:
X1 0 0.2 0.4 0.6 0.8 1.0
v x 106 20 21.5 24 27.4 32 40
m3/mol

Calculate the partial molar volumes of components

1 and 2 at x1=0.5 and x1=0.75

Solution: Draw a tangent to the curve at x1=0.5,

intercept at x1=0 gives v2 =17x10-6 m3/mol and x1=1
gives v1 =33.8x10-6 m3/mol
 Relations among partial properties
d (nG)  (nV )dP  (nS )dT   Gi dni
i

• Maxwell relation:
 V   S   Gi    (nS )   Gi   (nV ) 
               
 T P ,n  P T ,n  T P ,n  ni  P ,T , n j  P T ,n  ni  P ,T , n j

 Gi   Gi 
    Si    Vi
 T  P ,x  P T ,x

In a constant-composition solution, G i is a function of P and T, and

therefore:
 G i   G i 

dGi    dP    dT
   
 P T , x  T  P, x dGi  Vi dP  Si dT
• Consider equation that defines Enthalpy:
H  U  PV
nH  nU  P(nV )
Differentiation with respect to ni at constant T, P, and nj yields:
 (nH )   (nU )   (nV ) 
     P 
  ni  P ,T ,n j   ni  P ,T ,n j  ni  P ,T , n j

This becomes: H i  U i  PVi

Every equation that provides a linear relation among thermodynamic

properties of a constant-composition solution has its counterpart an
equation connecting the corresponding partial properties of each
species in solution
 The Ideal Gas Mixture Model
• The ideal gas mixture model provides a conceptual
basis to build the structure of solution thermodynamics
– It is simple and has a molecular basis
– Applied in well defined limit of zero pressure
– In an ideal gas mixture, the density is so low that it’s
molecules do not appreciably interact
– The molar volume V=RT/P, regardless of nature of
the gas (whether pure or mixed)
 ig  (nV ig )   (nRT / P)  RT  n  RT
Vi        
 ni  P ,T , n j  ni  P ,T , n j P  ni 
 n, j P

 ig RT
V i  Vi ig  V ig 
P
 Ideal-gas mixture
The partial pressure of species i in an ideal-gas mixture,
denoted by pi is define as the pressure that species I would
exert if it alone occupied the molar volume of the mixture.
pi (n, V t , T )  Pressure (ni , V t , T )
For the ideal gas mixture,
y i RT
pi  ig
 yi P (i  1,2,....,N )
V
• Gibbs’s theorem
– A partial molar property (other than volume) of a
constituent species in an ideal-gas mixture is equal to the
corresponding molar property of the species as a pure
ideal gas at the mixture temperature but at a pressure
equal to its partial pressure in the mixture.
M iig  Vi ig M iig (T , P)  M iig (T , pi )
Ideal Gas mixture Enthalpy and Internal Energy
H iig (T , P)  H iig (T , pi )  H iig (T , P)
Enthalpy and internal energy of ideal gas is independent of pressure
H ig   yi H iig U ig   yiU iig For ideal gas mixture, enthalpy
i i change of mixing is zero

Ideal Gas mixture Entropy

dSiig   Rd ln P (const.T )

P P
Siig (T , P)  Siig (T , pi )   R ln   R ln  R ln yi
pi yi P
M iig (T , P)  M iig (T , pi )

Siig (T , P)  Siig (T , P)  R ln yi S ig   yi Siig  R yi ln yi

i i
1
S ig   y i S iig  R y i ln Entropy change of mixing for
i i yi
ideal gases, always positive
Ideal Gas mixture Gibbs Energy
Giig  H iig  TSiig

H (T , P)  H (T , P)
ig ig Siig (T , P)  Siig (T , P)   R ln yi
i i

Giig  H iig  TSiig  RT ln yi

iig  Giig  Giig  RT ln yi G ig   y i Giig   y i Giig  RT  y i ln y i

i i i

An alternative dGiig  Vi ig dP 
RT
dP  RT ln P Constant T
expression, P
Giig  i (T )  RT ln P

iig  i (T )  RT ln yi P

G ig   yi i (T )  RT  yi ln yi P
i i
 Ideal Gas Mixture
U ig   yiU iig
i

H ig   yi H iig
i

  yi C p i
ig ig
Cp
i

S ig   yi Siig  R yi ln yi
i i

G ig   yi Giig  RT  yi ln yi
i i

 ig
 iig  G i  i (T )  RT ln y i P
 Illustration (11.1)
What is the change in the entropy when 0.7 m3 of CO2
and 0.3 m3 N2 each at 1 bar and 298.15 K blend to
form a gas mixture at the same conditions. Assume
ideal gas mixture.
 Illustration 2 (Problem 11.4)
What is the ideal work for the separation of an
equimolar mixture of methane and ethane at
175˚C and 3 bar in a steady-state flow process
into product streams of the pure gases at 35˚C
and 1 bar if Tσ = 300 K?
 Fugacity and fugacity coefficient
• Fugacity:
– The concept by introduced by G.N. Lewis and is used in
solution thermodynamics to represent the behavior of the
real fluid
– a quantity that takes the place of μi to facilitate the
criteria for equilibrium
For a constant temperature process, the property relation :
dG  VdP
For 1 mol of an ideal gas, V may be replaced by V=RT/P
dP
dG  RT  RTd (ln P)
P
 For a real fluid, the true pressure (for ideal gas case)
can be replaced by an ‘effective’ pressure, which is
called ‘fugacity’ of gases
dG  RTd (ln f ) With units of pressure

Gi  i (T )  RT ln f i
Giig  i (T )  RT ln P

fi Subscript i, for
Gi  G  RT ln
i
ig
pure species i
P

Residual Gibbs energy G  RT ln i

i
R

fi
Fugacity coefficient i 
P
  fi   fi 
lim ln i  lim ln    0 and lim i  lim    1
P  0 P  0
P P  0
 
P  0 P

• Pressure Explicit Form

fi P dP
ln  i  ln   ( Z i  1) (Constant T)
P 0 P

• Volume Explicit Form

d i
R
f i (T , P) Gi 
ln    ( Z i  1)  Z i  1  ln Z i
P RT 0 i

Like departure functions, there are different ways to estimate

fugacity and fugacity coefficient
 Fugacity from Generic Cubic
Equation of State (for pure species)
Evaluation of fugacity coefficients through cubic EoS
(e.g., van der Waals, R/K, S/R/K, P/R)
R
Gi
 ln  i  Z i  1  ln(1   i bi ) Z i  q i I i
RT
1  1   i bi 
Case 1 :    I ln  
    1   i bi 
 i bi
Case 1 :    I
1   i bi

Application of these equations at a given T and P

requires prior solution of an equation of state for Zi
 Generalized correlations for the
fugacity coefficient (for pure species)
Pr dPr
ln i   ( Z i  1) (const. Tr )
0 Pr

Z  Z 0  Z 1
Pr dPr Pr dP
ln    ( Z  1)
0
  Z1 r (const. Tr )
0 Pr 0 Pr
or
ln   ln  0   ln  1 with

Pr dPr Pr dP
ln   
0
( Z  1)
0
ln    Z 1 r For pure gas
1
0 Pr 0 Pr

Table E1:E4 or Table E13:E16

Fugacity Coefficient from Virial EoS

Bii P
Zi  1 B P

RT
ln  i  ii
RT 
0
dP (Constant T)

Bii P
ln  i 
RT
 VLE for pure species
• Saturated vapor: Giv  i (T )  RT ln f i v

• Saturated liquid: Gil  i (T )  RT ln f i l

fi v
G  G  RT ln l
i
v
i
l

fi
VLE
fi v
G  G  RT ln l  0
i
v
i
l

fi

f i v  f i l  f i sat iv  il  isat

For a pure species coexisting liquid and vapor phases are in equilibrium when they
have the same temperature, pressure, fugacity and fugacity coefficient.
 Fugacity of a pure liquid
• The fugacity of pure species i as a compressed liquid:
f i v ( Pi sat ) f i l ( Pi sat ) f i l ( P)
f i ( P) 
l
Pi sat
Pi sat f i v ( Pi sat ) f i l ( Pi sat )

(A) (B) (C)

Ratio (A) is  isat and given as

Pi s at
dP
ln  i   ( Z i  1)
sat v
(Constant T)
0
P

Ratio (B) is unity at equilibrium

 Fugacity of a pure liquid
• Ratio (C) reflects the effect of pressure on fugacity
of pure liquid i,
fi P
Gi  G i
sat
 RT ln Gi  G i
sat
 V l
i dP (isothermal process )
f i sat
sat
Pi

fi 1 P Vi l ( P  Pi sat )
f i ( P)
 P
l
ln Vi dP ln sat sat 
fi sat
RT i
s at
Since Vi is a weak f i ( Pi ) RT
function of P

• Substituting three ratios in initial equation :

Vi l ( P  Pi sat ) The exponential is
f i ( P)   isat Pi sat
l
exp
RT known as Poynting
factor
For H2O at a temperature of 300°C and for pressures up to 10,000
kPa (100 bar) calculate values of fi and φi from data in the steam
tables and plot them vs. P.
From steam table at T=300°C, Psat = 8592.7 kPa
Gi  i (T )  RT ln f i
For a low pressure reference state: Gi*  i (T )  RT ln f i*
fi 1
fi 1  Hi  H *
* 
ln  (Gi  G *
i )
  
*
ln * 
i
( S i S i ) Gi  H i  TSi fi RT
fi R T 

The low pressure (say 1 kPa) at 300°C: f i  P  1 kPa

* *

H i*  3076.8 J Si*  10.3450 J

g gK

For saturation pressure P = 8592.7 at 300°C H i  2751.0 J S i  5.7081 J

g gK
 isat  0.7843 f i sat  6738.9kPa

For different values of P up to the saturated pressure at 300°C, one obtains the
values of fi ,and hence φi . Note, values of fi and φi at 8592.7 kPa are obtained

Values of fi andφi at higher pressure: Vi l ( P  Pi sat )

fi   i
sat
Pi sat
exp
RT

Vil can be taken as molar volume of saturated liquid water at 300°C

Vil = 25.28 cm3/mol,
fi= 6789.8 kPa
φi= 0.6790

Fig 11.3
 Fugacity and fugacity coefficient:
species in solution

• For species i in a mixture of real gases or in a solution of

liquids:

i  i (T )  RT ln fˆi
Fugacity of species i in solution
(replacing the partial pressure)
• Multiple phases at the same T and P are in equilibrium when
the fugacity of each constituent species is the same in all
phases:
fˆ  fˆ  ...  fˆ
  
i i i
 i  i (T )  RT ln fˆi iig  i (T )  RT ln yi P

fˆi
i    RT ln
ig
i
yi P

The residual property: G R  G  G ig

The partial residual property: Gi R  Gi  Giig
Gi  RT ln ˆi
R
  (nG ) 
i     Gi
 ni  P ,T ,n
ˆ
j

f
ˆi  i The fugacity coefficient of species i in solution
yi P
fˆi
For ideal gas, Gi R  0 ˆ
i  1 fˆi  yi P
yi P
 Fundamental residual-property relation
 nG  1 nG
d  d (nG)  dT
 RT  RT
2
RT

d (nG)  (nV )dP  (nS )dT   i dni

i

G  H  TS

 nG  nV nH Gi
d  
 RT  RT
dP 
RT 2
dT  
i RT
dni nG
 f ( P, T , ni )
RT

G/RT as a function of its canonical variables allows evaluation of all other

thermodynamic properties, and implicitly contains complete property information.
The residual properties:
 nG R  nV R nH R Gi
R
 nG R  nV R nH R
d    dP  2
dT   dni or d    dP  2
dT   ln ˆi dni
 RT  RT RT i RT  RT  RT RT i
  nG R  nV R nH R Gi
R
 nG R  nV R nH R
d    dP  2
dT   dni d    dP  2
dT   ln ˆi dni
 RT  RT RT i RT  RT  RT RT i

Fix T and composition: V R   (G R / RT ) 

  
RT  P T , x

Fix P and composition: HR  (G R / RT ) 

 T  
RT  T  P, x

Fix T and P:
ˆ  (nG R / RT ) 
ln i   
 n i  P ,T ,n j
 Illustration (11.20)
Justify the following equations

G R  RT ln 
 Illustration
• Estimate the fugacity of gaseous mixture
consisting of 30% component 1 and 70%
component 2 by mole ,given that at 100oC and
50 bar the fugacity coefficients of components
1 and 2 in mixture are 0.7 and 0.85
respectively.
ln Ø =∑ xi ln Øi
and Ø = f/P
Answer: 40.124 bar
Develop a general equation for calculation of ln ˆi values form compressibility-
factor data.

  ( nG R
/ RT ) 
ln ˆi   
 ni  P ,T ,n j

nG R P dP
  (nZ  n)
RT 0 P

P   ( nZ  n)  dP
ˆ
ln i    
0
 ni  P ,T ,n j P

n  (nZ )
1  Zi
ni ni

ˆ P dP
ln i   ( Z i  1)
0 P Integration at constant temperature and composition
 Properties of real gas mixtures
• So far we have dealt with pure species (focus on
gases)
• Methods discussed earlier need to be modified by
including an additional composition variable to
predict properties of real gas mixtures
• Constants which appear in EOS which
characterize pure species behavior can be
averaged to get mixture constants
• Equations which express mixture constants in
terms of pure species constants are called
MIXING RULES
 Mixing rules
A mixing rule expresses a mixture constant am in terms
of composition expressed in mole fraction yi and the pure
component constant ai, i.e. am = yiyjaij
aii = ai, ajj = aj, (based on interaction of like pairs of
molecules) and aij is based on interaction of unlike
pairs of molecules
Equations which provide the interaction constant in terms
of pure component constants are called combining rules
If aij = (aii + ajj) / 2 = (ai + aj) / 2 ----- sqrt(aii ajj) =
sqrt(aiaj)

am = yiai am = [yisqrt(ai)]2
 Mixing Rules
For Cubic Equation of State (RK, SRK, PR)
n n
a mix    y i y j a ij
i 1 j 1
n
bmix   y i bi
i 1

where a ii and b i are the parameters of pure components i,

and combining rule
a ij  a ii a jj (1  k ij )  a ji

• Here kij is known as the binary interaction parameter, which results in

more accurate mixture equation of state calculations.
• It is found by fitting EoS to mixture data (VLE data). In the absence of
data, kij can be set equal to zero
 Fugacity coefficient in mixture from
the virial E.O.S
• The virial equation: Z  1  BP
RT
– the mixture second virial coefficient B: B  
i j
yi y j Bij

– for a binary mixture:

B  y1 y1B11  y1 y2 B12  y2 y1B21  y2 y2 B22

For species i in mixture,

 (nZ )  P  (nB) 
Zi     1   
 n i  P ,T , n RT  n i  P ,T , n
j j

Fugacity coefficient of species i in mixture, dP

ˆ P
ln i   ( Z i  1)
0 P
 Fugacity coefficient from the virial
E.O.S
Fugacity coefficient

ln ˆ1 
P
RT

B11  y2212 
12  2B12  B11  B22
ln ˆ2 
P
RT

B22  y1212 

For multi-component gas mixture, the general form:

P  1 
ˆ
ln k  
 Bkk   yi y j (2 ik   ij )  where  ik  2Bik  Bii  Bkk
RT  2 i j 
 The Ideal Solution Model
• Serves as a standard to which real-solution behavior
can be compared
From relation for ideal-gas mixture model
iig  Giig  Giig  RT ln yi
G id   xi Gi  RT  xi ln xi
  Gi  Gi  RT ln xi
id
i
id
i i
M id
  xi M i
id

i From fundamental property relation

 Gi id   G  id
  xi Si  R xi ln xi
S id
     i   R ln xi S id
 Si  R ln xi S
 T  T
i i
 P, x P i i

 Gi id   G    xiVi
     i 
id
Vi id
Vi id
 Vi V
 P T , x  P T i

H id
 Gi  TS
id id
 Gi  RT ln xi  TSi  RT ln xi H id
 Hi H id
  xi H i
i i i
i
Ideal Solution/Mixture
V id   xiVi
i

H id   xi H i
i

G id   xi Gi  RT  xi ln xi
i i

S id   xi Si  R xi ln xi
i i
 The Lewis/Randall Rule
• The composition dependence of the fugacity of a
species in an ideal solution is simple.
i  i (T )  RT ln fˆi iid  Giid  i (T )  RT ln fˆi id

Comparison with the equation Giid  Gi  RT ln xi where Gi can be

replaced by, G   (T )  RT ln f
i i i

fˆi id  xi f i Division both ˆiid  i The Lewis/Randall rule

sides by xiP
The fugacity coefficient of species i in an ideal solution is equal to the fugacity
coefficient of pure species i in the same physical state as the solution and at the
same T and P.
 Excess properties
• The mathematical formalism of excess properties is
analogous to that of the residual properties:
M E  M  M id
– where M represents the molar (or unit-mass) value of any
extensive thermodynamic property (e.g., V, U, H, S, G, etc.)
– Similarly, we have:
The fundamental excess-property relation

 nG E  nV E nH E Gi
E
d    dP  2
dT   dni
 RT  RT RT i RT
 The excess Gibbs energy and the
activity coefficient
• The partial excess Gibbs energy is of particular
interest: G E  G  Gid
i i i

Gi  i (T )  RT ln fˆi

Giid  i (T )  RT ln xi f i
fˆi
Gi  RT ln Gi E  RT ln  i Gi R  RT ln ˆi
E

xi f i
fˆi
The activity coefficient of species i in solution.  i 
xi f i
A factor introduced into Raoult’s law to account for liquid-phase non-idealities.
For ideal solution, Gi  0,  i  1
E
  nG E  nV E nH E Gi
E
d    dP  2
dT   dni
 RT  RT RT i RT

 nG E  nV E nH E
d    dP  2
dT   ln  i dni
 RT  RT RT i

V E   (G E / RT ) 
  
RT  P T , x
Experimental accessible values:
H E
 (G / RT ) 
E
 T   activity coefficients from VLE data,
RT  T  P, x VE and HE values come from mixing experiments.
 (nG E / RT ) 
ln  i   
 ni  P ,T ,n j

GE
  xi ln  i
RT i Important application in phase-equilibrium
 x d ln 
i
i i  0 (const. T , P) thermodynamics.
 The nature of excess properties
• GE: through reduction of VLE data
• HE: from mixing experiment
• SE = (HE - GE) / T

• Fig 11.4

– Excess properties become zero as either species ~ 1.

– GE is approximately parabolic in shape; HE and TSE exhibit
individual composition dependence.
– The extreme value of ME often occurs near the equimolar
composition.
Fig 11.4
 Models for the Excess Gibbs Energy
• In general GE/RT is a function of T, P and compositions,
but for liquids at low to moderate pressure, it is very
weak function of P.
• Therefore, for data at constant T:
GE
 g ( x1 , x2,........xn ) (const T)
RT

• Margules Equation is an example of this functionality

GE
 ( A21x1  A12 x2 ) x1 x2
RT
• For binary system, the function more often
represented by an equation is GE/x1x2RT,
which may be expressed as a power series :
GE
 a  bx1  cx12  .....
x1 x2 RT

• Redlich/Kister expansion
GE
 A  B( x1  x2 )  C ( x1  x2 ) 2  .....
x1 x2 RT
• Another well known equation results when the
reciprocal expression x1x2RT/GE, is expressed
as a linear function of x1:
x1 x2
E
 A' B' ( x1  x2 )  .....
G / RT

• van Laar equation

GE A12' A21
'
 '
x1 x2 RT A12 x1  A21'
x2