Fluid Phase Equilibria 412 (2016) 199e204

Contents lists available at ScienceDirect

Fluid Phase Equilibria

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / fl u i d

Liquideliquid equilibrium for the ternary system of isopropyl

acetate þ 2-propanol þ glycerol at different temperatures under
atmospheric pressure
Yuan-Xiao Li, Qin Xu, Shi-Liang Liu, Hai-Ming Li, Feng-Bao Zhang, Guo-Liang Zhang,
Qing Xia*
School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, PR China

a r t i c l e i n f o a b s t r a c t

Article history: The aim of this work is to analyze the possibility of separating isopropyl acetate and 2-propanol, an
Received 12 September 2015 azeotropic system, by using glycerol. Thus, liquideliquid equilibrium (LLE) data for the ternary system of
Received in revised form isopropyl acetate þ 2-propanol þ glycerin have been determined at various temperatures T ¼ (298.15,
25 December 2015
308.15, 318.15) K under atmospheric pressure. The consistency of the experimental data was verified by
Accepted 28 December 2015
Available online 30 December 2015
the mass balance. The extraction capacity was evaluated throughout the distribution coefficient and
selectivity parameters. Phase diagrams have been obtained at each temperature. To facilitate the
manipulation of the obtained data for industrial applications, the NRTL and the UNIQUAC models have
Keywords:
Liquid-liquid equilibrium
been applied to correlate the experimental data and low deviations were obtained.
Phase diagram © 2015 Elsevier B.V. All rights reserved.
Selectivity
Glycerol
NRTL
UNIQUAC

1. Introduction phase equilibrium of the quaternary system water þ 2-

Isopropyl acetate is an important material with a wide variety of
manufacturing uses in the chemical industry. It has really good
miscibility with other solvents like alcohol, ether and ketone, and is
also described as an effective solvent for many synthetic and nat-
ural resins [1]. As a result, isopropyl acetate has been widely used in
the fields of coating, printing ink, adhesive agents, etc. [2].
Generally, isopropyl acetate was produced from the esterifica-
tion of acetic acid and 2-propanol [3]. Isopropyl acetate and the
residual 2-propanol form an azeotropic mixture in this process,
which is difficult to be separated by the traditional distillation.
Several methods, including reactive distillation, extractive distilla-
tion, azeotropic distillation, liquideliquid extraction and mem-
brane separation, have been proposed to separate the azeotrope or
close boiling mixtures [4e7]. As for the azeotropic mixtures con-
taining isopropyl acetate and 2-propanol, Hong et al. [8] separate
the ternary mixtures water þ isopropyl acetate þ 2-propanol
through reactive distillation. Teodorescu et al. [9] studied the
* Corresponding author.
E-mail address: xiaqing@tju.edu.cn (Q. Xia).
propanol þ acetic acid þ isopropyl acetate. Andreatta and co-
workers [10,11] used the ionic liquid as an azeotrope breaker in
extractive distillation. In comparison to other separation processes,
liquideliquid extraction shows the advantage of being less energy
intensive. Moreover, high separation efficiency can be reached with
a suitable solvent [12]. However, no report about the extraction
separation of isopropyl acetate and 2-propanol can be found so far.
Selecting a suitable and applicable extractive solvent is of great
importance for a cost-effective separation process. Preliminary
studies showed that glycerol could be a potential solvent because it
was completely miscible with 2-propanol but almost immiscible
with the other one. This feature is mainly due to the strong polarity
and the ability to act as a proton receptor that can form hydrogen
bonds with 2-propanol.
To evaluate the feasibility of glycerol as an extractive solvent, the
liquideliquid equilibrium (LLE) data for isopropyl acetate þ 2-
propanol þ glycerol were measured at various temperatures
T ¼ (298.2, 308.15, and 318.15) K under atmospheric pressure. The
mass balance was applied to ascertain the consistency of the
experiment. The NRTL and the UNIQUAC models were used to
correlate the experimental data for the studied ternary system.

http://dx.doi.org/10.1016/j.fluid.2015.12.050
0378-3812/© 2015 Elsevier B.V. All rights reserved.
200 Y.-X. Li et al. / Fluid Phase Equilibria 412 (2016) 199e204
Additionally, the capacity of glycerol as an extractive solvent was
analyzed by the solute distribution coefficient and the selectivity.
2. Experimental
2.1. Materials
In this work, isopropyl acetate, 2-propanol and glycerol were
used. The source, CASRN, mass fraction purity and analysis method
of all chemicals are listed in Table 1. Purities of the materials were
determined by a gas chromatography (GC) equipped with a flame
ionization detector. In addition, the water contents were measured
by Karl Fischer titration and no appreciable water was detected. All
the materials were used without further purification.
2.2. Apparatus and procedure
The LLE experiment was carried out in a 30 mL glass cell with a
magnetic stirrer in it. The experimental method was described in
literature [13e15]. In this study it was specified as follows: The
temperature of the system was controlled by a refrigerated heating
circulator (Julabo FP45-HF, German, temperature stability ±0.01 K)
and measured by a platinum resistance thermometer Pt-100
(calibrated with an accuracy of 0.01 K). In each measurement of
the LLE data, 10 ml mixture of isopropyl acetate and 2-propanol
plus 10 ml glycerol were added into the glass cell and then stir-
red vigorously by the magnetic stirrer. Through a series of
comparative tests, the stirring time was chosen to be 2 h to get a
sufficient mixing for the extraction equilibrium. The mixture was
settled for 5 h to ensure a complete separation of the two phases.
Then the samples of the upper and lower phase (isopropyl acetate-
rich and glycerol-rich phase) were carefully collected by syringes at
least three times for the composition analysis.
2.3. Sample analysis
The collected samples were analyzed by gas chromatography
(Agilent Technologies 6890N). The gas chromatography was
equipped with a flame ionization detector and a HP-5 capillary
column (30 m  0.32 mm  0.25 mm). The carrier gas was nitrogen
with the purity of 0.99999 provided by Liufang industrial gases co.,
LTD, Tianjin. The operation condition was as follows: The injector
temperature and detector temperature were 573 K and 583 K,
respectively. The column temperature was programmed at 310 K
for 3 min firstly, then increased to 563 K at a rate of 45 K min1 and
held for 1 min. The calibration factor was obtained to correct the
measured values. The accuracy of the analytical method was tested
by known samples of mixtures and the maximum relative error of
the mass fraction did not exceed 0.003.
3. Results and discussion
3.1. Experimental data
The LLE data for isopropyl acetate (1) þ 2-propanol
Table 1
Experimental chemicals.
Chemical Source CASRN
Isopropyl Acetate Yuanli, China 108-21-4
2-Propanol Yuanli, China 67-63-0
Glycerol Yuanli, China 56-81-5
a
Gas chromatography.
b
Karl Fischer titration.
(2) þ glycerol (3) system at various temperatures T ¼ (298.15,
308.15, and 318.15) K under atmospheric pressure are listed in
Table 2. Triangular phase diagrams with tie-lines are shown in
Fig. 1. The experimental data show that the ternary system is a type-
I system. The large two-phase region in Fig. 1 enables a large
operation range for the extraction process. The mutual solubility of
isopropyl acetate and glycerol increases with the addition of 2-
propanol. The increasing temperature from 298.15 K to 318.15 K
does not show obvious effect on the equilibrium of the ternary
system. But it is worthy of mention that increasing temperature will
decrease the viscosity of glycerol significantly. According to the
research of Secur et al. [16], the viscosity of glycerol decreases from
1005 mPa s to 550 mPa s as the temperature increasing from
293.15 K to 323.15 K, which provides convenience for operation.
The consistency of the LLE data was verified by the method
proposed by Marcilla et al. [17]. According to the mass balance
between the initial mixture and the two conjugated phases (extract
phase and raffinate phase), three independent mass balance
equations can be written as follows:
0 0
M ðx1 Þ ¼ MI ðx1 ÞI þ MII ðx1 ÞII (1)
0 0
M ðx2 Þ ¼ MI ðx2 ÞI þ MII ðx2 ÞII (2)
0 0
M ðx3 Þ ¼ MI ðx3 ÞI þ MII ðx3 ÞII (3)
where M0 is the mole of initial mixture, MI and MII are the mole of
the isopropyl acetate-rich phase and the glycerol-rich phase
respectively. (xi)’ is the mole fraction of component i in the initial
mixture, while (xi)I and (xi)II are the mole fraction of component i in
the isopropyl acetate-rich phase and the glycerol-rich phase
respectively, i ¼ 1, 2, 3.
The values of MI and MII were calculated by solving these
overdetermined linear equations with least-squares method. The
relative error RE of the mass balance was determined by the
following equation:
0
MI þ MII  M
RE ¼ 0 (4)
M
The maximal RE is less than 1%. From Fig. 1, one can also
conclude that the experimental data is reliable because the points
of the overall compositions agree the tie line with great accuracy.
3.2. Evaluation of the extractive solvent
The distribution coefficient (b) and the selectivity (S) are used to
evaluate the extraction capacity of glycerol. These parameters are
defined as follows:
xІІ
2
b¼ (5)
xІ2
W(Mass fraction) Analysis method
0.990 GCa, KFb
0.999 GCa, KFb
0.995 GCa, KFb
 Y.-X. Li et al. / Fluid Phase Equilibria 412 (2016) 199e204 201

Table 2
The experimental LLE data of isopropyl acetate (1) þ 2-propanol (2) þ glycerol (3), together with the distribution coefficient b and selectivity S.a

Overall composition Isopropyl acetate-rich phase (I) Glycerol-rich phase (II) b S

x1 x2 x3 x1 x2 x3 x1 x2 x3

T ¼ 298.15 K
0.388 0.000 0.612 0.995 0.000 0.005 0.010 0.000 0.990
0.342 0.057 0.601 0.850 0.130 0.020 0.012 0.049 0.939 0.373 26.784
0.298 0.113 0.589 0.782 0.198 0.020 0.013 0.071 0.916 0.359 21.650
0.256 0.165 0.579 0.633 0.324 0.043 0.020 0.121 0.859 0.374 11.785
0.216 0.216 0.568 0.545 0.390 0.065 0.021 0.129 0.850 0.332 8.586
0.196 0.241 0.563 0.508 0.418 0.074 0.024 0.153 0.823 0.365 7.649
0.177 0.265 0.558 0.447 0.447 0.106 0.029 0.173 0.798 0.388 6.044
0.157 0.290 0.553 0.372 0.483 0.145 0.037 0.209 0.754 0.434 4.336
0.138 0.313 0.549 0.319 0.493 0.188 0.047 0.225 0.728 0.457 3.116
0.120 0.337 0.543 0.265 0.500 0.235 0.045 0.246 0.709 0.491 2.854
T ¼ 308.15 K
0.388 0.000 0.612 0.994 0.000 0.006 0.011 0.000 0.989
0.342 0.057 0.601 0.857 0.114 0.029 0.013 0.043 0.944 0.378 24.517
0.298 0.113 0.589 0.785 0.192 0.023 0.014 0.063 0.923 0.328 18.111
0.256 0.166 0.578 0.634 0.316 0.050 0.022 0.115 0.863 0.364 10.784
0.216 0.216 0.568 0.564 0.372 0.064 0.024 0.129 0.847 0.348 8.205
0.196 0.241 0.563 0.491 0.414 0.095 0.025 0.155 0.820 0.373 7.265
0.177 0.265 0.558 0.420 0.457 0.123 0.027 0.165 0.808 0.360 5.680
0.157 0.290 0.553 0.361 0.480 0.159 0.036 0.221 0.743 0.459 4.570
0.132 0.320 0.548 0.305 0.490 0.205 0.039 0.249 0.712 0.507 3.928
0.120 0.336 0.544 0.260 0.499 0.241 0.040 0.275 0.685 0.552 3.581
T ¼ 318.15 K
0.388 0.000 0.612 0.989 0.000 0.011 0.012 0.000 0.988
0.342 0.057 0.601 0.840 0.132 0.028 0.013 0.043 0.944 0.327 21.747
0.298 0.113 0.589 0.795 0.180 0.025 0.016 0.065 0.919 0.360 17.515
0.256 0.165 0.579 0.649 0.296 0.055 0.023 0.097 0.880 0.326 9.301
0.216 0.216 0.568 0.566 0.362 0.072 0.026 0.133 0.841 0.367 7.982
0.196 0.241 0.563 0.495 0.415 0.090 0.032 0.158 0.810 0.380 5.824
0.177 0.265 0.558 0.421 0.450 0.129 0.033 0.181 0.786 0.402 5.061
0.157 0.290 0.553 0.358 0.468 0.174 0.045 0.216 0.739 0.462 3.649
0.120 0.336 0.544 0.292 0.483 0.225 0.051 0.259 0.690 0.537 3.089
0.111 0.349 0.540 0.217 0.490 0.293 0.055 0.301 0.644 0.614 2.412
a
All compositions are expressed as mole fraction, the standard uncertainties u are u(T) ¼ 0.01 K, u(x1) ¼ u(x2) ¼ u(x3) ¼ 0.001.

isopropyl acetate and 2-propanol, which makes it easy to be

xІІ xІ
S ¼ 2І ІІ1
x2 x1
where x is the mole fraction. The values of b and S are listed in
Table 2, from which one can note a high selectivity.
The selectivity is directly related to the extraction capability of
the solvent. Fig. 2 presents the variations of the selectivity with the
content of the 2-propanol in the isopropyl acetate-rich phase. The
solid lines in Fig. 2 represent the fitting curves of the experimental
data using the following equation:

. 
S ¼ e þ f exp x12 g
recovered by distillation. (The boiling points of glycerol, isopropyl
(6)
acetate and 2-propanol are 563.00 K, 362.15 K and 355.60 K,
respectively [3].) Despite the high viscosity of glycerol is unfavor-
able for the extraction process, its viscosity can be decreased effi-
ciently by increasing the operating temperature. Another drawback
is the low distribution coefficient. A low distribution coefficient
suggests that a large amount of extractive solvent is needed to
obtain a satisfactory extraction effect, which means high costs
considering the size of equipment and the energy to posteriorly
separate solvent and solute.
(7) 3.3. Correlation of experimental data

The equation has been used to correlate the variation of selec-
tivity with a good degree of accuracy in the published studies
[18e20]. The parameters e, f, g in Eq. (7) and the standard de-
viations d are listed in Table 3.
As demonstrated in Fig. 2, the selectivity decreases with the
increase of the 2-propanol content in the isopropyl acetate-rich
phase, and slightly decreases with temperature. It should be
noted in Fig. 2 that the values of S are much higher than the unity,
especially at low content of 2-propanol.
As a potential solvent, glycerol has many attractive features
including good chemical stability, non-toxicity, non-corrosivity,
and relatively low price, besides its high selectivity. Moreover,
compared to isopropyl acetate, its higher density improves the
separation efficiency of the two phases in the extraction process.
(The densities of glycerol and isopropyl acetate are 1.261 g/cm3,
0.888 g/m3, respectively [3].) Its boiling point is higher than that of
Andreatta et al. summarized thermodynamic models describing
the ternary system alkyl ester þ alcohols þ glycerol [21]. Several
models have been previously tested to correlate these kinds of
mixtures: NRTL model, UNIQUAC model, traditional UNIFAC and its
versions, Wilson activity coefficient model, GCPPCSAFT model
CPA EoS model and GCAEoS model.
Except Andreatta, many other researchers also spent much
effort on the LLE of alkyl ester þ alcohols þ glycerol system. For
instance, Basso et al. [22] published the LLE data of
glycerol þ ethanol þ distilled fatty acid ethyl esters from crambe oil
in the range of 298.2Ke338.2 K. The data were correlated with the
NRTL and the UNIFAC models, and the NRTL model gave a better
result. Mesquita et al. [23] measured the LLE data of
glycerol þ ethanol þ coconut biodiesel system. The NRTL and the
UNIQUAC models were used to correlate the LLE data and satis-
factory results were obtained. De Azevedo Rocha et al. [24] studied
202 Y.-X. Li et al. / Fluid Phase Equilibria 412 (2016) 199e204

the LLE for ternary systems containing ethylic palm oil

biodiesel þ ethanol þ glycerol. The experimental data were
correlated by the NRTL model with a global deviation below 1%.
In the present study, both NRTL [25] and UNIQUAC [26] models
were applied to correlate the experimental data. These models
were successfully used in correlating the LLE data of many kinds of
mixtures, including the strongly non-ideal systems, such as the
associated systems which contain carboxylic acids [27e29]. In
terms of the systems of alkyl ester þ alcohols þ glycerol, several
authors correlated the experimental data with the NRTL or the
UNIQUAC model as mentioned above [22e24]. Moreover, Machado
et al. [30] correlated the LLE data for the ternary systems of
biodiesel þ ethanol þ glycerol with the NRTL model and the
calculation results agreed well with the experimental data. Franca
et al. [31] obtained satisfactory results when the authors correlated
the experimental data of castor oil biodiesel þ glycerol þ alcohol
system with the UNIQUAC model.
For the system of isopropyl acetate (1) þ 2-propanol
(2) þ glycerol (3), the non-randomness parameter aij among all
the components is set to 0.3 in the NRTL model, and the UNIQUAC
structural parameters of the pure components are listed in Table 4.
The binary interaction parameters for the two models were ob-
tained by minimizing the objective function OF [32]:

2 !2 !2 3
X
N X2 X 3 exp
Tk  Tkcal xexp  xcal
OF ¼ 4 þ
ijk ijk 5 (8)
j¼1 i¼1
sT sx
k¼1

where N is the total number of the tie-lines. The subscripts i and j

denote the component and different phase, respectively. The
subscript k represents the tie-line. The superscripts exp and cal
mean experimental value and value calculated by the models. In
the regression, sT and sx are 0.01 K and 0.001, respectively.
The correlation results were ascertained by the corresponding
root-mean-square deviation (RMSD).

vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi
u
u N 2 3 xexp  xcal 2
uX X X ijk
RMSD ¼ t
ijk
(9)
6N
k¼1 j¼1 i¼1

The values of the NRTL and the UNIQUAC model parameters are
shown in Table 5. The calculated data from these parameters are
plotted together with the experimental data in Fig. 1. Through Fig. 1
one can note that both models are in good agreement with the
experimental results, and the NRTL model makes a better correc-
tion. From the correlated results in Table 5, it can also be inferred
that the phase equilibria of the studied system can be accurately
described by the two models, and the NRTL model has a smaller
RMSD than the UNIQUAC model does.
Some authors correlated the LLE data of the system of alkyl
ester þ alcohols þ glycerol with other thermodynamic models.
Oliveira et al. [34] correlated the experimental data of this system
with the CPA-EoS model. The average deviations between the
experimental data and modeling results exceeded 2%. Andreatta
[21] applied the GCA-EoS and the A-UNIFAC model in this kind of
system. The standard deviations in the prediction of methanol
distribution coefficients were 9.5% and 3.93% for the two models,
respectively. Kuramochi et al. [35] applied different versions of the

Fig. 1. Triangular phase diagram for the system isopropyl acetate(1) þ 2-

propanol(2) þ glycerol(3) with tie-lines of the experimental and calculated data at
different temperatures under atmospheric pressure. (a) 298.15 K (b) 308.15 K (c)
▫
318.15 K. ( ) overall composition; (C) experimental tie line data; dash line, data
calculated by the NRTL model; dot line, data calculated by the UNIQUAC model.
 Y.-X. Li et al. / Fluid Phase Equilibria 412 (2016) 199e204 203

to correlate the experimental data and the RMSD values of both

models were reasonable small. The ternary system of isopropyl
acetate þ 2-propanol þ glycerol is a type-I system with a large two-
phase region, which provides a large operation range for the
extraction process. The increasing temperature from 298.15 K to
318.15 K showed no obvious effect on the equilibrium of the ternary
system. The selectivity (S) was much higher than unity. However,
the low distribution coefficient may pose difficulties for its appli-
cation in large-scale industrial production.

References

[1] W. Qi, M.F. Malone, Semibatch reactive distillation for isopropyl acetate syn-
thesis, Ind. Eng. Chem. Res. 50 (2011) 1272e1277.
[2] J.A. Riddick, W.B. Bunger, T.K. Sakano, Organic Solvents, fourth ed., Wiley-
lnterscience, New York, 1986.
[3] W. Bao, H. Si, Z. Liu, Organic Chemical Material, Hand Book of Chemical
Products, third ed., Chemical industry press, Beijing, 1991.
[4] D.J. Upadhyay, N.V. Bhat, Separation of azeotropic mixture using modified PVA
membrane, J. Membr. Sci. 255 (2005) 181e186.
Fig. 2. The variations of the selectivity coefficient with the content of 2-propanol in [5] Z. Guo, J. Chin, J.W. Lee, Feasibility of continuous reactive distillation with
azeotropic mixtures, Ind. Eng. Chem. Res. 43 (2004) 3758e3769.
the isopropyl acetate-rich phase. ,, T ¼ 298.15 K; B, T ¼ 308.15 K; △, T ¼ 318.15 K,
[6] Z. Zhang, A. Hu, T. Zhang, Q. Zhang, M. Sun, D. Sun, W. Li, Separation of methyl
solid lines represent the fitting curves using Eq. (7).
acetate plus methanol azeotropic mixture using ionic liquids as entrainers,
Fluid Phase Equilib. 401 (2015) 1e8.
[7] S. Yuan, C. Zou, H. Yin, Z. Chen, W. Yang, Study on the separation of binary
Table 3 azeotropic mixtures by continuous extractive distillation, Chem. Eng. Res. Des.
The parameters and the standard deviations in Eq. (7). 93 (2015) 113e119.
[8] G.B. Hong, M.J. Lee, H.M. Lin, Multiphase coexistence for mixtures containing
T (K) e f g d water, 2-propanol, and isopropyl acetate, Ind. Eng. Chem. Res. 42 (2003)
937e944.
298.15 20.879 60.875 0.538 0.242
[9] M. Teodorescu, K. Aim, I. Wichterle, Isothermal vapor-liquid equilibrium in
308.15 14.044 47.766 0.505 0.326
the quaternary water þ 2-propanol þ acetic acid þ isopropyl acetate system
318.15 3.821 40.909 0.277 0.486 with chemical reaction, J. Chem. Eng. Data 46 (2001) 261e266.
[10] A.E. Andreatta, M. Francisco, E. Rodil, A. Soto, A. Arce, Isobaric vapour-liquid
equilibria and physical properties for isopropyl acetate þ isopropanol þ 1-
butyl-3-methyl-imidazolium bis(trifluoromethylsulfonyl)imide mixtures,
Table 4 Fluid Phase Equilib. 300 (2011) 162e171.
The UNIQUAC structural parameters of the pure component.a [11] A.E. Andreatta, A. Arce, E. Rodil, A. Soto, Physical properties and phase equi-
libria of the system isopropyl acetate þ isopropanol þ 1-octyl-3-methyl-
Component ri qi
imidazolium bis(trifluoromethylsulfonyl)imide, Fluid Phase Equilib. 287
Isopropyl acetate 4.152 3.652 (2010) 84e94.
2-Propanol 2.914 2.528 [12] P. Lv, Z. Chang, S. Jin, S. Wang, H. Liu, Progress and expectation of mechanisms
Glycerol 3.386 3.060 research on liquid-liquid phase extraction and novel three-liquid-phase
extraction, Chem. Ind. Eng. Prog. 28 (2009) 1507e1512.
a
Taken from Aspen Plus V 7.2 physical properties data bank [33]. [13] F. Cai, J.J. Ibrahim, L. Niu, W. Xu, G. Xiao, Liquid-liquid equilibrium for ternary
system methanol plus methyl acetateþ1,3-dimethylimidazolium dimethyl-
phosphate at several temperatures and atmospheric pressure, J. Chem. Eng.
Data 60 (2015) 57e64.
Table 5
[14] Y. Feng, E. Yang, L. Dang, H. Wei, Liquid-liquid phase equilibrium for ternary
The NRTL and the UNIQUAC model parameters of the system isopropyl Acetate
mixtures of formamide (or ethylene glycol, or monoethanolamine) þ
(1) þ 2propanol (2) þ glycerol (3). indoleþ2-methylnaphthalene at 308.15 K, Fluid Phase Equilib. 398 (2015)
10e14.
i-j NRTL UNIQUAC
[15] S.-H. Chen, L.-J. Chen, Liquid-liquid equilibria for the ternary system water þ
aij Dgij (J/mol) Dgji (J/mol) uij (J/mol) uji(J/mol) octaneþ2-butyloxy-ethanol, Fluid Phase Equilib. 399 (2015) 16e21.
[16] J.B. Secur, H.E. Oberstak, Viscosity of glycerol and its aqueous solutions, Ind.
1-2 0.3 51.9 1259.2 2039.7 369.0 Eng. Chem. 43 (9) (1951) 2117e2120.
1-3 0.3 9862.2 9248.1 3084.7 1519.9 [17] A. Marcilla, F. Ruiz, A.N. García, Liquid-liquid-solid equilibria of the quaternary
2-3 0.3 2457.2 2395.7 1130.7 4282.5 system water-ethanol-acetone-sodium chloride at 25  C, Fluid Phase Equilib.
RMSD 0.0097 0.0157 112 (1995) 273e289.
[18] A. Arce, M.J. Earle, H. Rodriguez, K.R. Seddon, Separation of benzene and
hexane by solvent extraction with 1-alkyl-3-methylimidazolium bis{(tri-
fluoromethyl)sulfonyl}amide ionic liquids: Effect of the alkyl-substituent
UNIFAC models to correlate the methanol þ biodiesel þ glycerol, length, J. Phys. Chem. B 111 (2007) 4732e4736.
and the average relative deviation was more than 10%. The RMSD [19] A.B. Pereiro, A. Rodriguez, A study on the liquid-liquid equilibria of 1-alkyl-3-
methylimidazolium hexafluorophosphate with ethanol and alkanes, Fluid
values of the NRTL and the UNIQUAC models in the present study Phase Equilib. 270 (2008) 23e29.
were 0.97% and 1.57% separately, suggesting that these two models [20] E.J. Gonzalez, N. Calvar, B. Gonzalez, A. Dominguez, (Liquid plus liquid)
can describe the studied mixtures very well. equilibria for ternary mixtures of (alkane plus benzene plus EMpy ESO4 ) at
several temperatures and atmospheric pressure, J. Chem. Thermodyn. 41
(2009) 1215e1221.
[21] A.E. Andreatta, Liquid-liquid equilibria in ternary mixtures of methyl oleate
4. Conclusions plus ethanol plus glycerol at atmospheric pressure, Ind. Eng. Chem. Res. 51
(2012) 9642e9651.
[22] R.C. Basso, A.J. de Almeida Meirelles, E.A. Caldas Batista, Liquid-liquid equi-
LLE data for the ternary system of isopropyl acetate þ 2- librium of pseudoternary systems containing glycerol plus ethanol plus
propanol þ glycerol were measured at various temperatures ethylic biodiesel from crambe oil (Crambe abyssinica) at T/K ¼
T ¼ (298.15, 308.15, 318.15) K under atmospheric pressure. The (298.2,318.2,338.2) and thermodynamic modeling, Fluid Phase Equilib. 333
(2012) 55e62.
experimental data showed good consistency by checking the mass [23] F.M.R. Mesquita, N.S. Evangelista, H.B. de Sant'Ana, R.S. de Santiago-Aguiar,
balance. The NRTL and the UNIQUAC models were successfully used Liquid-liquid equilibrium for the glycerol plus alcohol plus coconut biodiesel
204 Y.-X. Li et al. / Fluid Phase Equilibria 412 (2016) 199e204
system at different temperatures and atmospheric pressure, J. Chem. Eng.
Data 57 (2012) 3557e3562.
[24] E.G. de Azevedo Rocha, L.A. Follegatti-Romero, S. Duvoisin Jr., M. Aznar,
Liquid-liquid equilibria for ternary systems containing ethylic palm oil bio-
diesel plus ethanol plus glycerol/water: experimental data at 298.15 and
323.15 K and thermodynamic modeling, Fuel 128 (2014) 356e365.
[25] H. Renon, J.M. Prausnitz, Local compositions in thermodynamic excess func-
tions for liquid mixtures, AlChE J. 14 (1968) 136e144.
[26] D.S. Abrams, J.M. Prausnitz, Statistical thermodynamics of liquid mixtures: a
new expression for the excess Gibbs energy of partly or completely miscible
systems, AlChE J. 21 (1) (1975) 116e128.
[27] H. Zhang, Liquid-liquid phase equilibria of the quaternary system {water (1) þ
acrylic acid (2) þ acetic acid (3) þ cyclopentyl methyl ether (4)}: measure-
ment, correlation, and comparative study, Fluid Phase Equilib. 403 (2015)
23e29.
[28] A. Samarov, M. Toikka, A. Toikka, Liquid-liquid equilibrium and critical states
for the system acetic acid þ n-butanol þ n-butyl acetate þ water at 308.15 K,
Fluid Phase Equilib. 385 (2015) 129e133.
[29] Z. Shen, Q. Wang, B. Shen, C. Chen, Z. Xiong, Liquid liquid equilibrium for
ternary systems water plus acetic acid plus m-Xylene and water plus acetic
acid plus o-Xylene at (303.2 to 343.2) K, J. Chem. Eng. Data 60 (2015)
2567e2574.
[30] A.B. Machado, Y.C. Ardila, L.H. de Oliveira, M. Aznar, M.R. Wolf Maciel, Liquid-
liquid equilibrium study in ternary castor oil biodiesel plus ethanol plus
glycerol and quaternary castor oil biodiesel plus ethanol plus glycerol plus
NaOH systems at (298.2 and 333.2) K, J. Chem. Eng. Data 56 (2011)
2196e2201.
[31] B.B. Franca, F.M. Pinto, F.L.P. Pessoa, A.M.C. Uller, Liquid-liquid equilibria for
castor oil biodiesel plus glycerol plus alcohol, J. Chem. Eng. Data 54 (2009)
2359e2364.
[32] Y. Lei, Y. Chen, X. Li, Y. Qian, S. Yang, C. Yang, Liquid-liquid equilibria for the
ternary system 2-methoxy-2-methylpropane plus phenol plus water, J. Chem.
Eng. Data 58 (2013) 1874e1878.
[33] Aspentech, Physical Property Data Reference Manual: Databank 11.1, Aspen
Technology, Inc., Burlington, M.A, 2001.
[34] M.B. Oliveira, S. Barbedo, J.I. Soletti, S.H.V. Carvalho, A.J. Queimada,
J.A.P. Coutinho, Liquid-liquid equilibria for the canola oil biodiesel plus
ethanol plus glycerol system, Fuel 90 (2011) 2738e2745.
[35] H. Kuramochi, K. Maeda, S. Kato, M. Osako, K. Nakamura, S.-i. Sakai, Appli-
cation of UNIFAC models for prediction of vapor-liquid and liquid-liquid
equilibria relevant to separation and purification processes of crude bio-
diesel fuel, Fuel 88 (2009) 1472e1477.