Research: Science and Education

1348 Journal of Chemical Education Vol. 84 No. 8 August 2007 www.JCE.DivCHED.org

The main concern of chemists is the control of product
yields and reaction rates, but the number of ways we can in-
fluence the outcome of a particular reaction is surprisingly
small. Moreover, the methods that are available have limited
applicability owing to the physical properties of the system
under study. Temperature, for example, is one of the most
commonly controlled parameters, but the temperature range
available for monitoring is often severely limited by the sta-
bility and phase of reacting substances. Changes in pressure
are effective only in systems with at least one gaseous com-
ponent. Other perturbations, including radiation, mechani-
cal and acoustic actions, and so forth, have limited
applicability and are specific to the process under investiga-
tion. Any additional method that can be used to control
chemical reactivity is particularly welcome.
Solvent-control of chemical reactions is one method that
has gained popularity in recent years, since a judiciously se-
lected solvent can alter the equilibrium or kinetics of a reac-
tion, even changing its direction entirely. Also, since the
overwhelming majority of known chemical processes take
place in solution or liquid phase, solvent-control is a conve-
nient means of influencing reactions. The comprehensive
development of solution chemistry has allowed chemists to
put forward and prove the idea that the solvent acts simulta-
neously as a medium and a reactive agent in a chemical sys-
tem (13).
A number of books (49) are devoted to detailing the
power that solvents have over the kinetics of chemical pro-
cesses (10, 11) and the issue is also addressed in nearly every
physical chemistry textbook. The influence of solvents on
chemical equilibriaparticularly the role of solvent in alter-
ing product yields and reaction branchinghas received less
attention (1218) and consequently is not included in many
textbooks. The didactic presentation of solvent effects on
chemical equilibria given below is intended to provide the
basis for teaching physical chemistry of solutions. Following
a general review of equilibrium thermodynamics and quan-
tification of the universal and specific solutesolvent inter-
actions, the solvent-control is illustrated with examples from
homo- and hetero-molecular associations and conformational
equilibria.
The material given in this article is appropriate for pre-
sentation following a review of the general ideas underlying
chemical equilibrium and an introduction of different types
of solvent effects. The students should be familiar with the
equilibrium constant and its relationship to the equilibrium
free energy (Gibbs energy at constant pressure and tempera-
ture); then the concepts of universal and specific solvation
may be introduced. Solvents such as hexane and carbon tet-
rachloride are excellent examples of universal solvents, since
they provide inert environments for carboxylic acid dimer-
ization. The specific solvents such as water and dioxane are
capable of hydrogen bonding with the acid. The concept of
a dielectric medium that screens electrostatic solutesolvent
interactions should be discussed in terms of the dielectric con-
stant.
Examples of nonpolar solvents with small dielectric con-
stants, such as hexane, and polar solvents with large dielec-
tric constants, such as acetonitrile and water, should be
considered. The instructor may elect to quiz the students on
chemical equilibrium and the types of solutesolvent inter-
actions at this point. At this point, the discussion of solvent
effects on chemical equilibrium may be terminated in a gen-
eral chemistry curriculum. A physical chemistry class can fur-
ther explore the detailed examples included below and in the
Supplemental Material.
W
These descriptions of homo- and
hetero-molecular association and conformer equilibria could
be used to develop illustrative laboratory experiments. In such
case, and especially if the laboratory experiment does not fol-
low the lecture immediately, the instructor can provide a to-
review list in the week preceding the lab. The to-review list
should be based on the lecture material and should give sev-
eral examples of solutesolvent pairs and types of interactions.
Finally, students should be required to hand in a complete
laboratory report.
General Considerations
Chemical equilibrium (15) in solution can be represented
in general as
B A
(1)
where A are reactants and B are products formed from the
reactants as a result of a chemical reaction. Consideration of
the thermodynamic cycle gives the Gibbs energy of the reac-
tion
! ! G = GG G
solvA
v
" + G
solvB
! ! (2)
where !G
solvA
and !G
solvB
are the Gibbs solvation energies
of reactants and products, respectively, and !G
v
is the Gibbs
energy of the reaction in vacuum.
The solutesolvent interactions can be separated into
universal and specific components, giving a sum (19):
G G G
solv
univ spec
= +
( ) ( )
! ! !
(3)
Control of Chemical Equilibrium by Solvent: W
A Basis for Teaching Physical Chemistry of Solutions
Oleg V. Prezhdo* and Colleen F. Craig
Department of Chemistry, University of Washington, Seattle, WA 98195-1700; *prezhdo@u.washington.edu
Yuriy Fialkov
Department of Physical Chemistry, Ukrainian National Technical University, Kiev, Ukraine
Victor V. Prezhdo
Institute of Chemistry, Jan Kochanowski University, 25-020 Kielce, Poland
Research: Science and Education
www.JCE.DivCHED.org Vol. 84 No. 8 August 2007 Journal of Chemical Education 1349
In !G
(univ)
the solvent acts as an inert medium, whereas
!G
(spec)
includes chemically active interactions such as hy-
drogen bonding. Substituting eq 3 in eq 2 leads to a general
equation that describes the effect of a solvent on chemical
equilibrium:
G G
v
= ! ! ++ "
+ "
) ( )
( )
!
!
G
G
(
!G
solvA
univ
solvB
univ
solvA
spec (
!G
solvB
spec ))
(4)
!G
v
, of course, does not depend on the solvent. The differ-
ence in the first bracket in eq 4 characterizes universal solva-
tion, and the difference in the second bracket accounts for
specific solvation. Thus, the solvent effect on chemical equi-
librium is determined by the solvation energies of each of
the chemical species involved. Note that for universal, chemi-
cally-inert solvents where !G
(univ)
>> !G
(spec)
, we may re-
write eq 4 as
( ) ( )
= + " G G G G ! ! ! !
v
solv B
univ
solvA
univ
(5)
The equilibrium Gibbs energy is directly related to the
equilibrium constant, K, at temperature T. As in eq 4 for the
Gibbs energy, one can distinguish between the universal and
specific contributions of a solvent to the magnitude of K
" =
+
K
G !
v
ln
+
RT
"
( ) ( )
! ! G G
solvB
univ
solvA
univ
"
( ) (
! ! G G
solvB
spec
solvA
spec))
(6)
or
" =
+ +
K
G G G
RT
ln
! ! !
v (spec) (univ)
# #
(6a)
As before (cf. eq 5), the equations simplify for universal sol-
vents:
= "
+
K
G
ln
!
(univ)
v
RT
"
( ) ( )
! ! G G
solvB
univ
solvA
univ
(7)
or
"
! !
K
G G
univ
v univ
( )
=
+
ln
#
(( )
RT
(7a)
We can develop a practical representation for #!G
(univ)
by considering the screening of the electrostatic solutesol-
vent interactions in terms of solvent dielectric permittivity, $.
Then eq 6a may be expressed (17) as
( )
= + + "
"
"
=
K
RT
G G
RT
1
1
ln ! !
v
solv
spec
solvB solvA
#
% %
$
! ! G G
v
solv
spec
solv
+ +
( )
#
# %
$
(8)
where the %
i
represent the main types of electrostatic inter-
actions, such as dipoledipole, iondipole, ionion interac-
tions, and so forth. Equation 8 simplifies for the case of a
universal solvent to
" ! K
R T
G
univ v sol
= +
( )
# % 1
ln
vv
$
(9)
where #%
solv
equals the universal solvation component ex-
trapolated to vacuum ($ = 1), that is, #%
solv
= #!G
$=1
(univ)
. In
addition to characterizing a solvent, the permittivity $ pro-
vides a convenient handle for hypothetically turning the sol-
utesolvent interaction on and off. For instance, in the limit
of perfect screening, $ & ' and the 1$ interaction terms in
eqs 8 and 9 disappear.
An analysis of eq 8 allows us to draw essential conclu-
sions about the influence of solvent on the thermodynamic
characteristics of chemical processes taking place in solution.
It follows from eq 8 that at fixed temperature, the equilib-
rium constant changes with solvent properties because of a
change in either a specific interaction, #!G
solv
(spec)
, or a uni-
versal electrostatic interaction, #%
solv
. This leads to several sce-
narios for the functional dependence of lnK on 1$,
depending on the behavior of #!G
solv
(spec)
and #%
solv
.
Generally, if both #!G
solv
(spec)
and #%
solv
vary between the
solvents, ln K = f (1$) is a nonlinear function. Even in sol-
vents where the specific interaction #!G
solv
(spec)
is constant,
and only the universal interaction #%
solv
changes, ln K =
f (1$) is still nonlinear in general. If both the #!G
solv
(spec)
and #%
solv
, values are constant for different solvents, that is,
in a conditionally universal medium (16), the ln K = f (1$)
dependence becomes linear (cf. eq 9)
univ
b
$
= +
( )
K a ln
(10)
where a and b are the coefficients of the linear equilibrium
equation. In the absence of specific solutesolvent interac-
tions #!G
solv
(spec)
( 0, the ln K = f (1$) dependence is also
linear and is described by eq 10.
Extrapolation of the ln K = f (1$) dependence to $ &
', that is, putting the system in a hypothetical medium in
which there are no electrostatic interactions, reduces eq 8 for
a solvent with specific interactions to
v
solv
spec
#
$
= +
& '
( )
K
RT
G G
1
ln ! ! "
(11)
Similarly, eq 9 becomes
=
& '
( )
K
G
RT
ln
$
univ
v
!
"
(12)
from which we can find the vacuum component of the Gibbs
energy of a chemical equilibrium process:
=
& '
( )
G RT K ln
$
v univ
! "
(12a)
Finally, we can determine the energy of specific solva-
tion by comparing the equilibrium constants of chemical pro-
cesses taking place in specific and universal solvents with
approximately the same dielectric constant $
(spec)
= $
(univ)
:
RT K ln
sppec univ ( ) ( )
" = lnK G #!
( )
solv
spec
" (13)
Research: Science and Education
1350 Journal of Chemical Education Vol. 84 No. 8 August 2007 www.JCE.DivCHED.org
Thus, analysis of the ln K = f (1$) dependence allows one
to characterize all three basic components of Gibbs energy
of a chemical process in solution: !G
v
, #!G
solv
(spec)
, and
#!G
$=1
(univ)
.
To compare the equilibrium constants of eq 1 in two
solvents S
(1)
and S
(2)
, we start by writing eq 8 for each sol-
vent:

=
( )
lnK
RT
S
1
1
)
v
solv S
(spec
(1)
+ G G # ! !
))
solv S
1
+
(
#%
$
"
(14a)

S
2
=
( )
K
RT
1
ln
v
solv S
(spec)
(2)
+ + # G G ! !
##%
$
solv S
2 ( )
"
(14b)
Solving eqs 14a and 14b for 1$ and equating the results,
we obtain
S
I II
S
1 2
const const
( ) ( )
= + K K ln ln
(15)
where
connst
I
v
solvS
spec
v
solv S
1
2
= +
" +
( )
( )
( )
G G
G G
! !
! !
#
#
sspec ( )
RT
1
"
#%
solv S
1 ( )
#%
solv S
2 ( )
(15a)
II
solvS
solv S
solvB
const
1
2
= =
" ( )
( )
# %
# %
% %%
% %
solvA
S
solvB solvA
S
1
2
[ ]
" [ ]
( )
( )
(15b)
An examination of eq 15 leads to two important conclusions.
First, the logarithm of the equilibrium constant of a chemi-
cal process in one solvent, S
(1)
, is linearly related to the loga-
rithm of the equilibrium constant of the same process in
another solvent, S
(2)
, Figure 1. Second, if const
II
> 1, as rep-
resented in Figure 1, the equilibrium constant in solvent S
(1)
is increased relative to solvent S
(2)
. Solvent S
(1)
shifts the equi-
librium towards products, so it can be thought of as having
a differentiating effect with respect to solvent S
(2)
. Con-
versely, solvent S
(2)
has a leveling effect with respect to sol-
vent S
(1)
, since it tends to shift the equilibrium towards the
middle.
In this section we have shown that solvent contributions
to the equilibrium constant produce two major factors that
affect chemical equilibrium: the Gibbs energies of specific
and universal solvation. The former occurs by a donorac-
ceptor interaction between solute and solvent. The latter origi-
nates due to universal electrostatic interactions between solute
and solvent molecules and is often taken into account by con-
sidering dielectric permittivity of the solvent.
Several methodologically interesting examples of solvent
influence on chemical equilibrium are presented in the fol-
lowing sections. The examples cover a wide range of phe-
nomena, including homo- and hetero-molecular association
and conformational equilibrium.
Homo-Molecular Association Processes
Dimerization is the most common homo-molecular as-
sociation process:
A
2
2A
(16)
The solvent dependence of the equilibrium constant for the
dimerization process (eq 16) can be obtained through appli-
cation of eq 8:
= K
RT
G
1
ln
dim d
!
iim
v
solv A
spec
solv A
spec
solv A solvA
2
2
+ "
+
"
( ) ( )
! ! G G 2
2 % %
$
"
(17)
The solvent interaction factors %
solvA
2
and %
solvA
can be cal-
culated using eq 5. For a dimerization process (eq 16) taking
place in a universal medium, we write eq 7a as:
= + K
RT
G G
1
ln
dim
! !
dim
v
dim
( ) univ ( ) univ
# "
(18)
We can specify the form of #!G
dim
(univ)
in terms of dielectric
permittivity as in eq 9

= +
RT
G
1
dim
!
v
%% %
$
solv A solvA
2
" 2
" K ln
dim
( ) univ
(19)
obtaining a linear relationship
$
= + a
b
K ln
dim
( ) univ
(20)
where !G
dim
*
= RTa and !G
dim
(univ)
= RTb$.
Equation 20 agrees well with experimentally determined
K
dim
values, as illustrated in Table 1 for solutions of carboxy-
lic acids. The correlation coefficient r presented in the table
is close to one. The linear correlation between the logarithm
Figure 1. Dependence of the equilibrium constants, K, of processes
taking place in a solvent S
(1)
on the equilibrium constants of the
same processes taking place in another solvent S
(2)
.
e h T . 1 e l b a T a d n a b g n i b i r c s e D s t n e i c i f f e o C
K 5 1 . 8 9 2 t a s d i c A c i l y x o b r a C f o n o i t a z i r e m i D
d i c A a b r
c i t e c A 2 6 . 4 0 1 . 5 2 8 9 . 0
c i t e c a o r o l h c o n o M 1 9 . 3 5 1 . 5 7 9 9 . 0
c i t e c s o r o l h c i D 5 3 . 3 7 9 . 5 8 9 9 . 0
c i t e c a o r o l h c i r T 0 6 . 5 5 3 . 4 8 9 9 . 0
Research: Science and Education
www.JCE.DivCHED.org Vol. 84 No. 8 August 2007 Journal of Chemical Education 1351
of the equilibrium constant and the polarity parameter (20,
21) E
T
, which is directly related to the solvent dielectric per-
mittivity $, gives reasonable estimates for K
dim
in universal
media (Figure 2).
Table 2 presents the vacuum and universal solvent in-
teraction components of the equilibrium Gibbs energy of car-
boxylic acid dimerization. The data show that in the series
from acetic to trichloroacetic acid, the vacuum interaction
component !G
dim
*
decreases and the solvent interaction com-
ponent #!G
dim
(univ)
increases. While the !G
dim
*
and
#!G
dim
(univ)
values are comparable in weakly-polar solvents
(small $), the #!G
dim
(univ)
component rapidly decreases with
increasing $; in chlorobenzene #!G
dim
(univ)
drops to about 20
30% of #!G
dim
*
. This trend is even more pronounced in the
highly-polar nitrobenzene. Thus, the dependence of ln K (or
!G) on 1$ in universal media provides important informa-
tion about the thermodynamics of homo-molecular associa-
tion.
The ln K
dim
= f ($, T) dependence for acetic acid dimer-
ization (CH
3
COOH, Figure 3) in a mixture of the universal
solvents CCl
4
and C
6
H
5
Cl can be approximated (22) by the
following
K
T
1 72
1817 1 0 92 15
$
= " + " +
dim
ln .
. . 553 4 .
$T
The dimer concentration (x, mol L
1
) is related to initial con-
centration of the monomer (c) and the equilibrium constant
(K
dim
) by
=
+ " 4 1 8
x
K c K
dim dim
cc
K
+ 1
8
dim
(21)
Equation 21 predicts that the concentration of the acetic acid
dimer in hexane (n-C
6
H
14,
, K
dim
( 1.5 10
3
) equals half that
of the monomer, whereas the dimer concentration in ni-
trobenzene (C
6
H
5
NO
2
, K
dim
( 10
2
) equals one third. Using
selected mixtures of the two solvents, one can adjust the ace-
tic acid dimer content from 33 to 50%. Similarly, dissolving
phenol in mixtures of CCl
4
and C
6
H
5
NO
2
with varying com-
position one can control the concentration of the phenol
dimer within a 2-to-12% interval relative to the monomer
concentration. More examples can be found in refs 17, 23
25.
In a specific medium, the interaction between some par-
ticipants of a chemical equilibrium and solvent molecules can
be exceptionally strong. In such a case, homo-molecular
dimerization is described by
A
2
+ (S) 2AS (22)
which assumes that the dimer is solvated to a lesser degree
than the monomer. In order to relate the homo-molecular
dimerization in the presence of specific solutesolvent inter-
action (eq 22) to that in a universal solvent (eq 16), we must
consider the solvation process
AS A + S
(23)
s d i c A c i l y x o b r a C f o n o i t a z i r e m i D f o s e i g r e n E s b b i G f o s t n e n o p m o C . 2 e l b a T
K 5 1 . 8 9 2 t a s t n e v l o S l a s r e v i n U n i
t n e v l o S
$
!G
m i d
) v i n u (
l o m J k ( /
"1
)
c i t e c A
d i c a
- o r o l h c o n o M
d i c a c i t e c a
- o r o l h c i r T
d i c a c i t e c a
- l y h t e m i r T
d i c a c i t e c a
e n a x e H 9 8 . 1 7 . 6 8 . 6 9 . 7 3 . 6
e n a h t e m o r o l h c a r t e T 3 2 . 2 7 . 5 7 . 5 6 . 6 3 . 5
e d i f l u s i d n o b r a C 2 6 . 2 8 . 4 9 . 4 6 . 5 5 . 4
m r o f o r o l h C 2 7 . 4 7 . 2 7 . 2 1 . 3 8 . 2
e n e z n e b o r o l h C 2 6 . 5 2 . 2 3 . 2 6 . 2 1 . 2
e n e z n e b o r t i N 8 . 4 3 6 3 . 0 7 3 . 0 3 4 . 0 4 3 . 0
m u u c a V 4 . 1 1 7 . 9 3 . 8 9 . 3 1
N E T O : s a d e s s e r p x e s i m u u c a v e h t r o f y g r e n e s b b i G e h T !G
m i d
v
l o m J k ( /
"1
. )
Figure 2. Dependence of the dimerization constants for acetic acid
(solid line) and trichloroacetic acid (dashed line) on the polarity
parameter (E
T
) of the following solvents: 1n-C
6
H
14
; 2CCl
4
; 3
CS
2
; 4CHCl
3
; 5C
6
H
5
Cl; and 6C
6
H
5
NO
2
.
Figure 3. Molecular models of the acetic acid dimers: (A) cyclic
dimer and (B) linear dimer.
Research: Science and Education
1352 Journal of Chemical Education Vol. 84 No. 8 August 2007 www.JCE.DivCHED.org
Comparing eqs 22 and 23 with eq 16 indicates that
homo-molecular association in a specific medium is compli-
cated by the necessity to break the AS bond first. The en-
ergy required to break the bond is subtracted from the energy
gained during dimerization, resulting in a decrease of K
dim
compared to the corresponding value in a universal medium
with the same dielectric permittivity. Mathematically, the
decrease in K
dim
follows from eq 17: the [!G
solvA
2
(spec)
"
2!G
solvA
(spec)
] term becomes negative in a specific medium,
thereby lowering the total energy of the association process
relative to the Gibbs energy for a universal medium, eq 19.
This analysis is supported by the experimental data presented
in Table 3, which shows a comparison of K
dim
values for
dimerization of acetic acid in various universal and specific
solvents.
The role of specific solvation is illustrated particularly
well through consideration of K
dim
values in specific and uni-
versal solvents with similar dielectric primitivities. We con-
sider here the specific solvents benzene and 1,4-dioxane, and
the universal solvent CCl
4
, which all have $ + 2.5. The K
dim
value for acetic acid dimerization in benzene is an order of
magnitude lower than in CCl
4
. The role of specific solvation
is even more dramatic in 1,4-dioxane, where the equilibrium
constant is almost three orders of magnitude lower than its
value in CCl
4
. 1,4-Dioxane shows significantly stronger spe-
cific solvation than benzene because it is much more basic.
Equations 8 and 9 were used to calculate the energies of spe-
cific solvation of acetic acid by benzene and 1,4-dioxane, giv-
ing 5.8 and 16.6 kJ mol
1
, respectively.
In water, the Gibbs energies of both specific and uni-
versal solvation of acetic acid are large. Owing to waters high
dielectric permittivity, the K
dim
value is insignificant.
Binary mixtures of universal and specific solvents allow
one to control the degree of homo-molecular association over
a wide range of concentrations (17). For instance, by vary-
ing the composition of the n-hexane1,4-dioxane mixture it
is possible to change the content of the acetic acid dimers
from 50% to 8%. The dimer concentration can be further
decreased below the 8% mark, all the way down to 0.5%, by
using a mixture of 1,4-dioxane and water.
It happens quite often that suitable solvent selection pro-
vides the only practical mechanism for changing the degree
of molecular association.
Solvent Effect on Conformer Equilibrium
A conformational equilibrium process can be expressed
in general as
conformer II conformer I
(24)
or for the special case of cistrans isomerization as
trans-isomer cis-isomer (25)
The experimental approaches for determining conformer con-
centrations are not always sufficiently accurate (17); as a rule
the difference in conformer energies is relatively small, rang-
ing from 0.1 to 1012 kJmol. This is similar to the ener-
gies of dipoledipole interaction and specific solvation in
moderately active solvents, so it is difficult to discriminate
between the conformer equilibrium and other processes.
Since dipole moments are often very different between
the conformers (26), dielectric permittivity of the solvent
plays a crucial role in the equilibrium processes (22, 25). The
conformer transformation energies are inversely proportional
to permittivity in universal solvents
con
G !
ff
A
B
= +
$
(26)
where A and B are the coefficients of the linear energy equa-
tion. Correspondingly, ln K
conf
depends on 1$ linearly
conf
= +
$
K a
b
ln
(27)
and the equilibrium constant shows an exponential depen-
dence on 1$. Here, a = ART and b = BRT, cf. eq 6.
These equations are sufficiently accurate to describe experi-
mental data for conformation and isomerization equilibria.
We will illustrate the argument above with the isomer-
ization of 1,2-disubstituted ethane: XCH
2
CH
2
Y. The po-
tential energy of the molecule depends on a number of
geometric parameters, the most important of which is the tor-
sion angle , (Figure 4). Stable conformers include the syn-
clinal (sc) conformation characterized by ,
sc
and the
anti-periplanar (ap) conformation characterized by ,
ap
, which
exactly equals 180 owing to symmetry. The key parameters
of the potential energy profile include the difference !G in
the Gibbs free energies of the isomers and the heights of the
two types of maxima !G
#
1
and !G
#
2
. Given an explicit form
n o i t a z i r e m i D r o f s t n a t s n o C m u i r b i l i u q E . 3 e l b a T
K 5 1 . 8 9 2 t a s t n e v l o S s u o i r a V n i d i c A c i t e c A f o
s t n e v l o S l a s r e v i n U $ K
m i d
e n a x e H 8 8 . 1 0 5 2 1
e n a h t e m o r o l h c a r t e T 3 2 . 2 0 9 9
e d i f l u s i d n o b r a C 2 6 . 2 0 0 7
m r o f o r o l h C 2 7 . 4 0 9 2
e n e z n e b o r o l h C 2 6 . 5 0 4 2
e n e z n e b o r t i N 8 . 4 3 0 1 1
s t n e v l o S c i f i c e p S
e n e z n e B 3 . 2 5 9
e n a x o i D - 4 , 1 2 . 2 2 . 1
r e t a W 3 . 8 7 5 0 . 0
Figure 4. The potential energy (!G
#
1
) of molecules of the
XCH
2
CH
2
Y type (see main text for detailed explanation).
Research: Science and Education
www.JCE.DivCHED.org Vol. 84 No. 8 August 2007 Journal of Chemical Education 1353
for the shapes of the minima and the maxima, for instance, a
cosine function, one can uniquely determine the values of the
angles ,
sc
and ,
ap
. The Gibbs energy difference !G between
the stable rotamers follows the inverse dependence on solvent
dielectric permittivity. For example, eq 26 for 1-fluoro-
2-chloroethane in mixtures of alkane and chloralkane solvents
becomes (27)
2 86
8
. G !
conf
= " +
..
.
69
0 963
$
r =
Considering eq 5 in the context of eq 26, it is clear that
an increase in solvent dielectric permittivity decreases the
absolute value of the electrostatic components of the con-
former transformation free energies in universal solvents. This
stabilizes the isomerization equilibrium. For instance, the
!G
conf
for bromocyclohexanone in cyclohexane ($ = 2) is
5.2 kJmol, while in acetonitrile ($ = 36) the energy equals
0.3 kJmol.
The stabilization of polar conformers offered by solvents
with higher dielectric permittivity affects not only the final
equilibrium composition of the conformer mixture, but also
the rate of approach to equilibrium. When the transition state
between the conformers is more polar than the stable con-
formersas is often the caseequilibration occurs faster in
polar solvents. We see this phenomenon, for instance, with
the rotation around the formally double CC bond in sub-
stituted ethylenes
In this example, the isomerization transition state (II) is sig-
nificantly more polar than the stable conformer (I) (28).
The influence of specific solvation on conformational
equilibrium is widely known for solvents capable of inter-
molecular hydrogen bonding. A good illustration is given by
the tautomerization of imidazoletetrahydropyran (III), which
can form a hydrogen bond or even abstract a proton (IV) in
such solvents as chloroform (28):
However, correlating specific solvation effects on con-
formational equilibrium with the physical and chemical prop-
erties of the solvent is extremely difficult. Attempts to
determine simple and reliable connections between !G and
ln K
conf
values and the donor and acceptor energy levels do
not succeed in general. Significantly better results have been
obtained with pairwise solutesolvent interaction parameters
(29).
As in the molecular association example above, often-
times the solvent effect is the only method available to effect
drastic change in the relative concentrations of different con-
formational forms. Consider, for example, dichloro-
acetaldehyde conformers in cyclohexane and DMSO solvents.
At room temperature, the axial rotamer is favored in cyclo-
hexane; the equilibrium constant for the formation of this
rotamer is K
conf
= 0.79. However, the corresponding K
conf
value in DMSO is just 0.075. To obtain a similar equilib-
rium in cyclohexane, one would have to cool the solution to
64 K (209 C); this is entirely impossible experimentally,
since cyclohexane freezes at +6.5 C. Similarly, to achieve K
conf
= 0.79 in DMSO, one would have to heat the solution to
435 K (162 C).
Conclusions
We have presented a didactic description of solvent ef-
fects on chemical equilibria, illustrated by several examples
of basic processes occurring in solution. This material is in-
tended to provide the basis for teaching physical chemistry
of solutions, assuming the students are familiar with the gen-
eral ideas underlying chemical equilibrium: the equilibrium
constant and equilibrium free energy. The universal and spe-
cific types of solvation have been introduced, followed by
detailed examples of homomolecular association and con-
former equilibrium. Further examples involving hetero-
molecular association may be found in the Supplemental
Material.
W
This article shows the students that solvent can
be used to influence chemical equilibrium, thereby altering
the yield of reaction products.
To determine the influence of solvent on chemical equi-
librium, it is necessary to consider the donoracceptor prop-
erties of the solvent as well as its polarity (i.e., permittivity
as a microscopic property or the dipole moment and polar-
izability as a molecular property). Using mixed solvents al-
lows one to widely vary these properties.
We have not considered chemical processes taking place
in water, because they have been studied more extensively
than those taking place in nonaqueous environments. More-
over, owing to its singular physical-chemical properties, wa-
ter in no way provides a representative system for the study
of nonwater solvents. It is certainly necessary to give atten-
tion to mixtures of water and nonwater solvents, which find
greater application in research and in industry (dyeing, elec-
trochemical processing of materials, etc.). However, areas of
science and engineering based on nonwater solvents are ad-
vancing no less intensely. Clearly the development presented
in this article is urgently needed and will provide a valuable
addition to the physical chemistry curriculum.
Acknowledgments
Financial support from the National Science Founda-
tion of the United States and Petroleum Research Fund of
the American Chemical Society is gratefully acknowledged.
W
Supplemental Material
Mathematical derivation of the temperature dependence
of the Gibbs energy as well as a detailed example of the sol-
vent effect on heteromolecular association are available in this
issue of JCE Online.
Research: Science and Education
1354 Journal of Chemical Education Vol. 84 No. 8 August 2007 www.JCE.DivCHED.org
Literature Cited
1. Lei, Z. G.; Li, C. Y.; Chen, B. H. Sep. Purif. Rev. 2003, 32,
121.
2. Noeres, C.; Kenig, E. Y.; Gorak, A. Chem. Eng. Process. 2003,
42, 157.
3. Kapucu, N.; Guvenc, A.; Mehmetoglu, U.; Calimli, A. Rev.
Chem. Eng. 1999, 15, 233.
4. Amis, E. S.; Hinton, J. F. Solvent Effect on Chemical Phenom-
ena; Academic Press: New York, 1973.
5. Entelis, S. G.; Tiger, R. P. Kinetic of Reactions in Liquids,
Khimiya: Moscow, 1973; in Russian.
6. Palm, V. A. The Foundations of Quantitative Theory of Organic
Reactions; Khimiya: Leningrad, 1977; in Russian.
7. Reichardt, C. Solvents and Solvent Effects in Organic Chemis-
try; VCH: Weinheim, Germany, 1988.
8. Connors, K. A. Chemical Kinetics; VCH: New York, 1990.
9. Pross, A. Theoretical and Physical Principles of Organic Reac-
tivity; John Willey & Sons, Inc.: NewYork, 1995.
10. El Seoud, O. A.; Bazito, R. C.; Sumodlo, P. T. J. Chem. Educ.
1997, 74, 562.
11. Mazzacco, C. J. J. Chem. Educ. 1996, 73, 254.
12. Pietro, W. J. J. Chem Educ. 1994, 71, 416.
13. Leenson, I. A. J. Chem. Educ. 1986, 63, 437.
14. Butler, J. N. J. Chem. Educ. 1984, 61, 784.
15. Honig, J. M.; Amotz, D. B. J. Chem. Educ. 2006, 83, 132.
16. Fialkov, Yu. Solvent as an Agent of Chemical Process Control;
Khimiya: Leningrad, 1990; in Russian.
17. Fialkov, Y. Y.; Chumak, V. L. Mixed Solvents: in Handbook of
Solvents; Wypych, G., Ed.; ChemTec Publ., William Andrew
Publ.: Toronto, 2001; pp 505564.
18. Smith, J. V.; Missen, R. W. Chemical Reaction Equilibrium
Analysis. Theory and Algorithms; Wiley Interscience: New York,
1982.
19. Silla, E.; Arnau, A.; Tunon, I. Fundamental Principles Solvents
Use. In Handbook of Solvents; Wypych, G., Ed.; ChemTec
Publ., William Andrew Publ.: Toronto, 2001; pp 736.
20. Dimroth, K. Liebigs Ann. 1963, 57, 576.
21. Johnson, D. A.; Shaw, R.; Silversmith, E. F.; Ealy, J. J. Chem.
Educ. 1994, 71, 517.
22. Fialkov, Yu. Ya.; Barabash, V. A.; Bondarenko, E. S. Ukr. Khim.
Zh. 1987, 53, 490.
23. Barcza, L.; Buvari-Barcza, A. J. Chem. Educ. 2003, 80, 822.
24. De Leon, D. G. L.; Guidote, A. M. J. Chem. Educ. 2003, 80,
436.
25. Silverstein, T. P. J. Chem. Educ. 1998, 75, 116.
26. Minkin, V. I.; Osipov, O. A.; Zhdanov, Yu. A. Dipole Moments
in Organic Chemistry; Plenum Press: New York, 1970.
27. Orville-Thomas, W. J.; Redshaw, M. Internal Rotation in Mol-
ecules; Wiley-Interscience: London, 1974.
28. Vereshchagin, A. I.; Kataev, V. E.; Bredikhin, A. A.; Timosheva,
A. P.; Kovylyaeva, G. I.; Kazakova, E. H. Conformation Analysis
of Hydrocarbons and Their Derivatives; Nauka: Moscow, 1990;
in Russian.
29. Prezhdo, V. V.; Shevchenko, T. N.; Prezhdo, O. V. J. Mol. Str.
1994, 318, 243.
Control of Chemical Equilibrium by Solvent: A Basis for
Teaching Physical Chemistry of Solutions
Supplementary Material

Oleg V. Prezhdo

Department of Chemistry, University of Washington, Seattle, WA 98195-1700 USA;
E-mail: prezhdo@u.washington.edu; Tel: +1-206-221-3931; Fax: +1-206-685-8665
Colleen F. Craig
Department of Chemistry, University of Washington, Seattle, WA 98195-1700 USA;
E-mail: cfcraig@u.washington.edu; Tel: +1-206-543-2738
Yuriy Fialkov
Department of Physical Chemistry, Ukrainian National Technical University, Kiev, Ukraine
Victor V. Prezhdo
Institute of Chemistry, Jan Kochanowski University, 5 Chciska str., 25-020 Kielce, Poland
E-mail: victor@pu.kielce.pl

Temperature Dependence of the Reaction Gibbs Energy.

The temperature dependence of the reaction Gibbs energy can be approximated
by a linear function
G = G
0
+ TG
1
. (S1)
Equation (S1) almost always gives a valid description, provided the temperature range is
such that the solvent remains liquid at normal pressure (1). In the ideal case, G
0
= H is
enthalpy of reaction and -G
1
= S is reaction entropy. Substituting (S1) into (6a) in the
main text gives the temperature dependence of the equilibrium constant
( )
) (
1
) (
0
) (
1
) (
0 1 0
) / 1 ( ln
univ
solv
univ
solv
spec
solv
spec
solv
v v
G T G G T G G T G RT K + + + + + = . (S2)
The terms in (S2) either depend inversely on temperature or have no explicit temperature
dependence.
A similar representation of the temperature dependence of the coefficients a and b
in equation (10)
a = a
0
+ a
1
/T, b = b
0
+ b
1
/T,
leads to a convenient approximate form for the lnK = f(T, ) functional dependence,
which is valid for conditionally-universal and universal media:
ln K = a
0
+ a
1
/T + b
0
/ +b
1
/T. (S3)
The coefficients a
0
, a
1
, b
0
, and b
1
in (S3) are taken to be independent of both and T. For
a specific medium or an equilibrium constant considered over a wide temperature
interval, the lnK = f(T, ) dependence takes a more general form:

=
+ + + = + + + = + + + + + =
n
n
n
n
f
T
b
T
a
a
b
T
b
T
a
a
T
b
T
b b
T
a
a K
1
0 1
0
0 1
0 2
2 1 0 1
0
) (
1 1
ln

.
(S4)

Solvent Effect on the Processes of Heteromolecular Association
In universal media, equilibrium constants for heteromolecular adduct formation
mA + nB A
m
BB
n
(S5)
depend exponentially on solvent permittivity:
ln K
add
= a + b/, (S6)
cf. Eq (27) in the main text. In the case m, n = 1, the concentration c
M
of adduct AB may
be related to equilibrium constant K
AB
and initial concentration of the components by:
0
M
c
( ) [ ]
AB M AB M AB M
K c K c K c 2 / 1 4 1 2
2 / 1
0 0
+ + = . (S7)
If at least one of the stoichiometric coefficients m, n is greater than one, the
expression gives an equation of higher degree than (S7):
n m
B A
K
( ) ( )
m
M B M
n
M A M M B A
mc c nc c c K
n m
=
0
,
0
,
/ . (S8)
Let us consider, for example, the interaction between acetic acid (HAc) and
tributyl phosphate (TBP) to form the adduct lHAcTBP, where l = 1...4; the trend
predicted by (S6) is nicely illustrated by lHAcTBP formation in universal solvents, such
as n-C
6
H
14
, CCl
4
, n-C
4
H
9
Cl, and C
6
H
5
NO
2
, and in mixed solvents (2).
Equilibrium constants for lHAcTBP formation in one- and two-component
solvents are presented in Table S1. Note that the magnitude of K
add
decreases with
increasing permittivity. In this way, changing the permittivity of the universal solvent
allows us to change the output of the reaction. The percent output of complexes is listed
in table below for initial concentration of 0.1 mol/l:
0
M
c
HAcTBP 2HAcTBP 3HAcTBP 4HAcTBP
% Complex in hexane
% Complex in nitrobenzene
84
43
46
13
32
4.5
22
2

The larger the electrostatic component of the process Gibbs energy ( ), the larger
the relative concentration change. Just as with our consideration of conformer
equilibrium, the solvent used here is an effective means of process adjustment. To reduce
the output of adduct HAcTBP in nitrobenzene to that in hexane, the nitrobenzene
solution would have to be cooled to 78
el
G
0 =

0
C (taking into account that the enthalpy of
adduct formation is 15 kJ/mol) (3). Clearly, this process is not possible because the
melting point of nitrobenzene is +5.8
0
C.

Table S1. Equilibrium constants (K
add
) of adduct (nHAcTBP) formation
in n-hexane (H), nitrobenzene (NB) and the mixed solvent (H + NB)
at 298.15 K, and coefficients of equation (S6)
K
add
Solvent
HAcTBP 2HAcTBP 3HAcTBP 4HAcTBP
Hexane ( = 2.23)
H + NB ( = 9.0)
H + NB ( = 20.4)
Nitrobenzene ( = 34.8)
327
28.1
19.6
17.4
8002
87.5
45.0
36.2
6.1910
5
275
88.1
60.8
1.4610
7
1935
518
337
Coefficients of equation (S6)
a
b
2.69
5.82
3.28
10.73
3.58
18.34
5.21
21.22


To demonstrate the effect of specific solvation on the equilibrium constant for
heteromolecular association, we begin by setting m = n = 1 in (S5):
A + B AB (S9)
Generally both A and B are subject to specific solvation:
A + S AS; B + S BS, (S10)
where AS and BS are solvated solute molecules. Note that the exact number of solvent
molecules interacting with the solutes is not important for the following.
Assuming that the specific solvation of the adduct AB is negligible in comparison
to the specific solvation of initial components, A-B interaction in the solvent S can be
represented by the scheme:
AS + BS AB + 2S. (S11)
A and B are initially solvated by S, but when AB is formed, the specific interactions
between A and B replace those formerly between A and S, B and S. In this way,
heteromolecular association in a specific medium is a resolvation process, since each
initial component changes its solvation structure. Clearly, equilibrium constants of
heteromolecular association that are calculated without consideration of this circumstance
will not accurately represent process (S9), since there is a contribution to the equilibrium
from process (S11).
However, we can still determine the correct equilibrium constant for (S9)
K
AB
= [AB]/[A][B] (S12)
by removing the contribution of the resolvating process (S11)
K
RP
= [AB]/[AS][BS]. (S13)
In dilute solutions, the concentration of the specific solvent S or the specific component
of a solvent mixture is much higher than the initial concentration of the equilibrium
species [A]
0
and [B]
0
. In this case it is possible to use the approximation that the
activities of the equilibrium participants are equal to their concentrations, and then the
equations of mass balance for components A and B may be obtained (4).
The equilibrium constant for adduct AB formation in a dilute solution was
obtained in terms of the solvating (S10) and resolvating (S11) processes (4):
K
AB
= K
RP
(1 + K
AS
)(1 + K
BS
). (S14)
A similar form for equation (S14) has been proposed elsewhere (5).
If the specific solvent concentration is not much larger than the concentration of
A and B, equation (S14) changes to
K
AB
= K
RP
(1 + K
AS
[S]

)(1 + K
BS
[S]). (S15)
Let A be an acidic (electron-pair acceptor) reagent in reaction (S9), and B a basic
(electron-pair donor) reagent. Then, if S is an acidic solvent, one can neglect specific
solvation of reagent A, and the equations for the equilibrium constants of process (S9)
can be written in the form
K
AB
= K
RP
(1 + K
BS
); K
AB
= K
RP
(1 + K
BS
[S]). (S16)
If, on the other hand, A is the basic component, equation (S16) can be re-written as
K
AB
= K
RP
(1 + K
AS
); K
AB
= K
RP
(1 + K
AS
[S]). (S17)
This analysis in the case of basic S will lead to a similar set of equations. Equations
(S16) and (S17) for specific solvation.were developed in Ref.s (6-8).
The constant K
AB
is identified in the literature as calculated by taking into
account the specific solvation (9), but to account for specific solvation one must obtain
the equilibrium constants for several processes. Process (S11) delivers K
RP
, (S10) gives
K
AS
and K
BS
.
Fig. S1 provides a practical illustration of the approach described above by
presenting the equilibrium constants for the heteromolecular association of DMSO and o-
creosol
(CH
3
)
2
SO + o-CH
3
C
6
H
4
OH (CH
3
)
2
SOHOC
6
H
4
CH
3
(S18)
in various binary solvent mixtures at 298.15 K: CCl
4
heptylchloride (10), two universal
solvent components (open circles); CCl
4
nitromethane (11), universal component (CCl
4
)
and specific acceptor component (nitromethane) (closed circles); and CCl
4
ethyl acetate
(12), universal component (CCl
4
) and specific donor component (ester) (crosses). Note
that the linear dependence of ln K on 1/ is supported by fig. S1.
Lines 1-2 represent the initial solvation processes (cf. (S10)) for o-creosol in ethyl
acetate (K
AS
), and DMSO in nitromethane (K
BS
), respectively; lines 3-4 represent
resolvation (K
RP
) of the o-creosol DMSO complex (cf. (S11)). Line 5 gives the
heteromolecular-association equilibrium constants (K
AB
) calculated from lines 1 and 3, 2
and 4, as per equations (S16) and (S17). These derived K
AB
are free of specific solvation
effects, and compare well with K
AB
for process (S18) occurring in the mixture of CCl
4

and heptylchloride universal solvents.
Fig. S1 indicates that specific solvation of one equilibrium participant decreases
the equilibrium constant compared to processes taking place in an isodielectric universal
solvent.

Figure S1. Equilibrium constant (lnK) vs. solution permittivity () for the heteromolecular
association of DMSO with o-cresol (process 45). A range of solution permittivities were
obtained by varying the amount of heptylchloride (), ethyl acetate (), and nitromethane
(). Lines 1 and 2 correspond to the initial specific solvation of a component, cf. process
(S10). Lines 3 and 4 correspond to resolvation of the component, cf. process (S11). Line 5
presents a comparison of K
AB
values: in the universal solvent mixture CCl
4
-heptylchloride
(); in the mixed universal/specific solvent mixtures CCl4-ethyl acetate () and CCl4-
nitromethane (), calculated as per equation (S16).
1. K
BS
for solvation of o-cresol in ethyl acetate;
2. K
BS
for solvation of DMSO in nitromethane;
3. K
RP
for resolvation of o-cresol in DMSO. (DMSO in a CCl
4
ethyl acetate mixture is
added to the o-cresol ethyl acetate mixture.)
4. K
RP
for resolvation of DMSO in o-cresol. (o-cresol in a CCl
4
nitromethane mixture is
added to DMSO nitromethane mixture.);
5. Comparison of K
AB
values for process (S18).

One method of studying the effect of specific solvation on the process of
heteromolecular association was described in Ref. (11), which focused on the equilibrium
process
C
6
H
3
(NO
2
)
3
+ C
6
H
5
N(CH
3
)
2
C
6
H
3
(NO
2
)
3
C
6
H
5
N(CH
3
)
2
(S19)
in binary mixtures of universal solvent heptane with a specific component:
trifluoromethylbenzene, acetophenone or p-chlorotoluene. Trinitrobenzene is initially
solvated in these binary mixtures, so its interaction with the electron-donor component of
the mixture represents the resolvation process. A charge transfer complex is formed as a
result:
C
6
H
3
(NO
2
)
3
S + C
6
H
5
N(CH
3
)
2
C
6
H
3
(NO
2
)
3
C
6
H
5
N(CH
3
)
2
+ S (S19a)
We can determine K using equation (14) written for heteromolecular association:
( ) .
1
ln
) (
B
) (
A
) (
AB

+ + =

solv spec
solv
spec
solv
spec
solv
v
G G G G
RT
K (S20)
For equilibrium (S19a) in mixed solvents we can assume 0 and 0.
Therefore, in the case of conditionally universal media, equation (14) can be converted
to:
) (
B
spec
solv
G
) (
AB
spec
solv
G
.
1
ln
) (
A

+ =

solv spec
solv
v
RP
G G
RT
K (S20a)
It follows from equation (S20a) that lnK
RP
vs. 1/ will give a rectilinear plot.
Experimentally determined equilibrium constants for process (S19a) support this
conclusion (11).
The possibility for adjustment of the product yield for homomolecular
association is more dramatic in the case of the interaction between o-nitrobenzene and
triethylamine. The equilibrium constants for this process in several solvents are
determined in Ref. (12). The data indicate a 93% product yield in n-hexane, compared to
6% in 1,2-dichloroethane (DCE). To increase product yield in DCE to 93% it would be
necessary to cool the solution approximately by 120
0
C, which is below the melting point
of DCE. On the other hand, the temperature would have to be increased to nearly 320
0
C
to observe a product yield of 6% in n-hexane. This would require the use of high external
pressure or an autoclave.
Keto-enol tautomerism (2,13,14), ionization processes (4,15-23), acid-base
equilibia (4,23,24), transport and non-equilibrium processes (4), complex formation (23-
26), and solubility (4,27-31) provide further examples of the solvent control influence on
chemical processes; these are unfortunately beyond the scope of this paper. One article
cannot address every situation in which solvent may be employed to adjust chemical
equilibria, but the framework we have presented above provides a useful starting point
for investigating the role of solvent in various chemical processes.

Literature cited.

1. Modell, M; Reid, R. Thermodynamics and its Applications. Prentice-Hall,
Englewood Cliff: New Jersey, 1974.
2. Fialkov, Yu. Solvent as an Agent of Chemical Process Control, Khimiya:
Leningrad, 1990 (Rus.).
3. Chrisian, S; Lane, E., in Solutions and Solubilities, V.1, John Wiley&Sons: New
York, 1975.
4. Fialkov, Y.Y.; Chumak, V.L. Mixed Solvents: in Handbook of Solvents, Ed. G.
Wypych, ChemTec Publ., William Andrew Publ.: Toronto-N.Y., 2001, p. 505-
564.
5. Bishop, B; Sutton, L. J. Chem. Soc. 1964, 6100.
6. Bhownik, B; Srimani, P. Spectrochim. Acta 1973, A29, 935.
7. Ewal, R; Sonnesa, A. J. Am. Chem. Soc. 1970, 92, 3845.
8. Drago, R; Bolles, T.; Niedzielski, R. J. Am. Chem. Soc. 1966, 88, 2717.
9. Fialkov, Yu.; Barabash, V. Theor. Exp. Chem. (Rus.) 1986, 22, 248.
10. Barabash, V; Golubev, N; Fialkov, Yu. Doklady AN USSR, 1984, 278, 390.
11. Borovikov, A.; Fialkov, Yu. Russian J. Gen. Chem. 1978, 48, 250.
12. Pawelka, Z; Sobczyk, L. Roczn. Chem. 1975, 49, 1383.
13. Reichardt, Ch. Solvents and Solvent Effects in Organic Chemistry, VCH:
Wainheim, 1988.
14. Prezhdo, V.V.; Khimenko, N.L.; Surov; Yu.N. Ukr. Khim. Zh. 1986, 52, 58.
15. Person, R.G.; Vogelgong, D. J. Am. Chem. Soc. 1963, 85, 3533.
16. Falkenhagen, H.; Ebeling, W. in: Ionic Interaction. V.1. S. Petrucci, ed. Academic
Press: New York, 1971.
17. Barthel, J.; Kreinke, H.; Kuntz, W. Physical Chemistry of Electrolyte Solutions.
Modern Aspects, Steinkopff, Dermstadt and Springer: New York, 1998.
18. Cox, B.J. Ann. Rept. Progr. Chem. 1974, 70A, 249.
19. Boyer, F.L.; Bircher, L.J. J. Phys. Chem. 1960, 64, 1330.
20. Sakamote, N.; Masuda, R. Electrochim. Acta, 1981, 26, 197.
21. Schmelzer, N.; Einfeld, J.; Grigo, M. J. Chem. Soc. Faraday Trans. I, 1988, 84,
931.
22. Fialkov, Yu.Ya.; Gorbachev, V.Yu; Kamenskaya, T.A. J. Mol. Liq. 2000, 89, 159.
23. Gutmann, V. Coordination Chemistry in Non-Aqueous Solvents, Springer-Verlag:
Wien-N.Y., 1968.
24. Complexing in Non-Aqueous Solution. Ed. G. Krestov, Nauka: Moscow, 1989 (in
Rus).
25. Bak, M. Chemistry of Complex Equilibria, Akademiai Kiado: Budapest 1970.
26. Burger, K. Solvation, Ionic, and Complex Formation in Non-Aqueous Solvent:
Experimental Methods for the Investigation. Elsevier Scientific Pub. Co.:
Amsterdam-New York, 1983.
27. Hildebrand, J.H.; Scott, R.L. The Solubility of Non-Electrolytes, Reinold Publ.
Corp.: N.Y. 1950.
28. Hildebrand, J.H; Pransnitz, J.M.; Scott, R.L. Regular and Related Solutions, Van
Nostrand: N.Y. 1970.
29. Senichev V.Yu., Tereshatov V.V. General Principles Govering Dissolution of
Materials in Solvents, in: Handbook of Solvents, Ed. G. Wypych, ChemTec Publ.,
William Andrew Publ.: Toronto-N.Y., 2001, p. 101-132.
30. Gordon, J. The Organic Chemistry of Electrolyte Solutions. Wiley: New York,
1975.
31. Marcus, Y. Ion Solvation, Wiley Limited: Chichester-New York, 1985.