Advances in Chemical Engineering and Science, 2012, 2, 504-507
http://dx.doi.org/10.4236/aces.2012.24061 Published Online October 2012 (http://www.SciRP.org/journal/aces)
An Equation of State for Nonaqueous Electr olyte Solutions
Zheng Han
The Laboratory of Molecular and Materials Simulation, Department of Chemical Engineering, College of Chemical Engineering,
Beijing University of Chemical Technology, Beijing, China
Received August 29, 2012; revised September 30, 2012; accepted November 11, 2012
ABSTRACT
A two parameters equation of state (EOS) for nonaqueous electrolyte solutions system has been developed. The equa-
tion is in terms of Helmholtz free en ergy and incorp orated with results of low density expansion of non-primitive mean
spherical approximation. The EOS was tested for experimental data reported in literatures of 9 nonaqueous single elec-
trolyte solutions of which the temperature was 298.15 K, and it also has a good predictive capability for nonaqueous
electrolyte solutions at different temperature in this work. The comparisons with EOSs published earlier by other re-
searchers in literatures are carried out in detail.
Keywords: EOS; Nonaqueous; Electrolyte Solutions
1. Introduction
Electrolyte solutions are encoun tered in a wide variety of
industrial processes, for example, wastewater treatment,
extraction, seawater desalinization, distillation and geo-
logical processes. It is very important for us to describe
the thermodynamic properties of such systems accurately.
Phase equilibrium in electrolyte systems is very impor-
tant to chemical industry.
Past a few decades, people made a lot of progress on
describing thermodynamic properties of electrolyte sys-
tems with some models [1-4]. But most of studies in lit-
eratures were restricted in aqueous electrolyte systems.
There is little attention on nonaqueous electrolyte sys-
tems until now. Although we can get some data of prop-
erties from literature, the data about nonaqueous electro-
lyte systems is much less than the one about aqueous
electrolyte systems. So in engineering, we need a simple
predictive model in order to describe phase behavior of
nonaqueous electrolyte systems.
EOSs of nonaqueous electrolytes have been developed
successfully since the late 1970s. Pitzer’s models [5,6]
have also extended to nonaqueous electrolyte solutions
and the adjustable parameters are needed in all of these
models. But up to now, there are still few models to rep-
resent phase equilibria properties of nonaqueous electro-
lyte solution.
In general, EOS can be derived by differentiating the
Helmholtz free energy with respect to the density. The
Helmholtz free energy is divided into several contribu-
tions, typically inclu ding solvent-so lvent, ion -solv ent and
ion-ion terms. In this work , we expanded Helmholtz free
energy as several contributions (including electrostatic
contribution and association contribution) according to
perturbation theory. On the other hand, the EOS pro-
posed is tested for 9 nonaqueous solutions of alkali metal
halides. The parameters can be obtained by fitting the
vapor pressure of solvents. In addition, we also compared
our results with the results of Mock et al., Youxiang Zuo
and Tzujen Chou.
2. Model and Theory
The systems of interest in this work are limited in non-
aqueous solu tions (methanol solvent) of alkali metal hal-
ides. Since the salts are fully dissociated, there are three
components in the solution, including cation, anion and
methanol solvent respectively. The ions are treated as
charged Lennard-Jones (LJ) spheres with additional as-
sociating sites forming h ydrogen bond s with methano l. A
methanol molecule is regarded as the LJ sphere with
embedded a point dipole and three associating sites, two
of which represent lone pair electrons and the others rep-
resent prot ons.
At temperature T and volume V, the system consists of
N particles, and the number of species i is Ni. By using
the perturbation theory [7-10], the differences of the
Helmholtz free energies (A – Ahs) between the syste m and
the corresponding hard sphere system can be expanded as
elect assochs LJ
AA A AA
NkTNkT NkTNkT
 (1)
where k is the Boltzmann constant. The superscripts hs,
LJ, elect and assoc represent the contributions from hard
C
opyright © 2012 SciRes. ACES

Z. HAN 505
sphere, Lennard-Jones, electrostatic (including ion-ion,
ion-dipole and dipole-dipole terms) and association in-
teractions, respectively. Then we can get this equation:
elect assoc
L
Jhs
A
AA
NkTNkT NkT
 A A
NkT NkT
  (2)
The equation of state expressed as the compressibility
factor can be derived from above equations by differenti-
ating the free energ y with respect to the density,
A
NkT 


 


Z

 (3)
The chemical potential of species k is derived from the
following differentiation,
,Tk
A
k
kT NkT
k



 










 (4)
Note although the differentiation s of Equations (3) and
(4) can be derived analytically, for convenience, the nu-
merical ones are used for our calculations in this work
directly.
3. Results and Discussion
As all the ions are removed, the system is regressed to
pure methanol, i.e., LJ spheres with a point dipole and
three associating sites. Its Helmholtz free energy can be
expressed by equations. The dipole moment is set as 2.49
Debye to reproduce the experimental dielectric constant
of methanol, which is predicted as 32.49 for our EOS and
is very close to the experimental value 32.70 at 298.15 K.
There are still two kinds of parameters need to be
fixed. The first one is the effective average ionic diame-
ter (σi), which is assumed adjustable here for each salt.
Another one is the ion-methanol association parameter
for each ion. Th is parameter can be obtained by fitting it
as a salt-dependent parameter. Furthermore, it is found
that the association of the anion and methanol can be
ignored in this model without notable losses of accura-
cies. The anion-methanol association term was therefore
removed from our EOS either. Consequently, only two
parameters are required in our model.
The EOS proposed was tested for 9 nonaqueous elec-
trolyte solutions of alkali metal halides. The parameters
were obtained by fitting the experimental data of the va-
por pressure and activity of electrolyte solutions at 1 bar
and 298.15 K. The regressed parameters and the average
absolute deviations (AADs, see definitions in Table 1) of
the vapor pressure data are listed in Table 1.
As can be seen from Table 1, our EOS gives a good
correlation of vapor pressure and activity with an average
AAD of 1.120% and 0.106%. Mean while, it is o bviou s to
see that the predicted activities are in good agreement
with the experimental d ata over from low molality ran g es
to high molality ranges. And the agreement with experi-
mental data is very good when the maximum molality up
to 4.58 mol/kg methanol. So it reveals that our EOS is
very successful in activity calculation over molality
range about 0 - 5 mol/kg methanol in general although
the AADs about vapor pressure are little higher than the
nes obtaine d by Z uo . o
Table 1. Regressed parameters for EOS in this work and the average absolute deviations (AADs) in the vapor pressure (P)
and activity (a), from this work and other models at 1 bar and 298.15 K.
EOS parameters1 AAD%2
This work
Salt σi (Å) εassoc/k (K) Zuo, P3 Mock et al., P4 Chou, P5 P a
Molarity range
(mol/kg)
LiCl 5.326 3215.46 2.33 2.90 0.42 1.825 0.387 0 - 4.580
LiBr 5.316 3137.97 1.99 3.17 0.59 1.800 0.282 0 - 4.345
NaCl 6.126 2106.30 0.17 0.19 0.01 0.942 0.023 0.041 - 0.219
NaBr 5.651 3623.06 0.36 0.22 0.08 0.839 0.053 0.042 - 0.649
NaI 5.111 2717.21 0.84 0.84 0.26 0.907 0.063 0.024 - 0.755
KBr 6.992 4471.59 0.19 0.12 0.00 0.961 0.016 0.044 - 0.134
KI 6.008 5129.05 0.24 0.25 0.06 0.889 0.066 0.022 - 0.735
RbI 6.473 5241.23 0.20 0.24 0.01 0.945 0.045 0.02 - 0.436
CsI 7.421 5787.42 0.21 0.16 0.00 0.975 0.020 0.033 - 0.130
Average 0.99 1.63 0.26 1.120 0.106
1There are two p arameters for each salt . One is the ef fectiv e average io n diamet er, σi, and the other is the cation-methanol associating parameter, εassoc. The two
parameters are all salt dependent. 2cal exp
exp
1
100
%NP
i
f
f
AAD NP f


, where NP is the number of the experimental points and f is the property of interest (P and a).
The supers cripts cal and ex p indicate th e value is fr om the calcul ation and exp eriment, respecti vely. 3The AADs% were reported for the electrolyte EOS pro-
posed by Julia n Youxiang Zuo, Dan Zhang and Walter Furst [8] . 4The AADs% were reported for the electrolyte NRTL model proposed by Mock et al. [9]. 5The
AADs% were reported for the two-parameter ACM proposed by Tzu-Jen Chou and Akihiko Tanioka [10].

Z. HAN
506
02
10000
15000
20000
25000
30000
35000
40000
45000
4
298.15K
318.15K
308.15K
vapor pressure (pa)
molality (mol/kg methano l)
(a) 024
10000
15000
20000
25000
30000
35000
40000
45000
298.15K
318.15K
308.15K
vapor pressure (pa)
molality (mol/kg methanol) (b)
Figure 1. Predicted vapor pressure of (a) LiCl and (b) LiBr nonaqueous electrolyte solutions as a function of salt molality.
The lines are calculated from equation of state with the parameters in Table 1, which were obtained by fitting the experi-
mental data at 298.15 K. The points represent the experimental data. For average absolute deviations (ADDs), see Table 2.
Table 2. The average absolute deviations (AADs) about the vapor pressure (P) for same regressed parameters of EOS in dif-
ferent temperature, from this work at 1 bar.
EOS parameters1
Salt σi (Å) εassoc/k (K) Molarity range (mol/kg) AAD%2 for P T (K)
1.825 298.15
2.802 308.15 LiCl 5.326 3215.46 0 - 4.580
4.045 318.15
1.800 298.15
2.733 308.15 LiBr 5.316 3137.97 0 - 4.345
3.630 318.15
1There are two parameter s for each salt . One is th e effecti ve averag e ion d iameter σi, and the other is the cation-methanol associating parameter: εassoc. The two
parameters are all salt dependent. 2cal exp
exp
1
100 NP
i
f
%f
NP f



AAD , where NP is the number of the experimental points and f is the property of interest (P). The
superscript cal and exp indicate the value is from the calculation and experiment, respectively.
The predictive capability of EOS in this work can be
demonstrated by extrapolating the temperature to a little
higher value. For example, Figures 1(a) and (b) show the
predictive vapor pressures by using the parameters given
in Table 1, which are correlated from experimental vapor
pressures with a temperature of 298.15 K. Strikingly,
even up to 308.15 K, our EOS can still accurately repre-
sent the non-ideality of the nonaqueous electrolyte solu-
tions and the AADs are shown in Table 2.
4. Conclusion
A fundamental two-parameter equation of state for non-
aqueous electrolyte solutions is proposed by incorpora-
tion of low density expansion of nonprimitive mean
spherical approximation and statistical associating fluid
theory. The EOS has been tested for 9 nonaqueous alkali
halide solutions at ambient condition . The parameters are
obtained by fitting the vapor pressures and activities with
the average absolute deviation (AAD, see definition in
Table 1) of 1.120% and 0.106%. With the parameters
given by 298.15 K, the EOS can also well predict the
vapor pressure data of nonaqueous electrolyte solutions
at different temperature points and over the same mo-
lality range accurately.
REFERENCES
[1] A. Anderko, P. Wang and M. Rafal, “Electrolyte Solutions:
From Thermodynamic and Transport Property Models to
the Simulation of Industrial Processes,” Fluid Phase
Equilibria, Vol. 194-197, 2002, pp. 123-142.
doi:10.1016/S0378-3812(01)00645-8
[2] J. R. Loehe and M. D. Donohue, “Recent Advances in
Modeling Thermodynamic Properties of Aqueous Strong
Electrolyte Systems,” AIChE Journal, Vol. 43, No. 1, 1997,
pp. 180-195. doi:10.1002/aic.690430121
[3] Y. Liu a nd S. Wata nair i, Chemical Engineering Progress,

Z. HAN 507
Vol. 95, 1999, pp. 25-42.
[4] H. L. Friedman, “Electrolyte Solutions at Equilibrium,”
Annual Review of Physical Chemistry, Vol. 32, 1981, pp.
179-204. doi:10.1146/annurev.pc.32.100181.001143
[5] K. S. Pitzer, “Thermodynamics of Electrolytes. I. Theo-
retical Basis and General Equations,” The Journal of
Physical Chemistry, Vol. 77, No. 2, 1973, pp. 268-277.
doi:10.1021/j100621a026
[6] K. S. Pitzer and G. Mayorga, “Thermodynamics of Elec-
trolytes. II. Activity and Osmotic Coefficients for Strong
Electrolytes with One or Both Ions Univalent,” The Jour-
nal of Physical Chemistry, Vol. 77, No. 19, 1973, pp.
2300-2308. doi:10.1021/j100638a009
[7] J. A. Barker and D. Henderson, “What Is ‘Liquid’? Un-
derstanding the States of Matter,” Reviews of Modern
Physics, Vol. 48, No. 4, 1976, pp. 587-671.
doi:10.1103/RevModPhys.48.587
[8] J. Y. Zuo, D. Zhang and W. Furst, “Extension of the Elec-
trolyte EOS of Fürst and Renon to Mixed Solvent Elec-
trolyte Systems,” Fluid Phase Equilibria, Vol. 175, No.
1-2, 2000, pp. 285-310.
doi:10.1016/S0378-3812(00)00463-5
[9] Y. X. Zuo and W. Furst, “Prediction of Vapor Pressure
for Nonaqueous Electrolyte Solutions Using an Electro-
lyte Equation of State,” Fluid Phase Equilibria, Vol. 138,
No. 1-2, 1997, pp. 87-104.
doi:10.1016/S0378-3812(97)00145-3
[10] T. J. Chou and A. Tanioka, “A Vapor Pressure Model for
Aqueous and Nonaqueous Solutions of Single and Mixed
Electrolyte Systems,” Fluid Phase Equilibria, Vol. 137,
No. 1-2, 1997, pp. 17-32.
doi:10.1016/S0378-3812(97)00091-5