Fluid Phase Equilibria 419 (2016) 50e56

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 equilibria for the extraction of phenols from alkane

using ethylene glycol
Fangfang Dai a, Kun Xin a, Yuhe Song b, Midong Shi a, Haipeng Zhang a, Qingsong Li a, *
a
The State Key Lab of Heavy Oil Processing, College of Chemical Engineering, China University of Petroleum e East China, Qingdao, Shandong 266580, China
b
College of Pharmacy and Chemistry & Chemical Engineering, Taizhou University, Taizhou, Jiangsu 225300, China

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

Article history: The separation of low temperature coal tar can be carried out using extraction technology, especially for
Received 10 January 2016 the collection of phenols. Liquideliquid equilibria data for the ternary mixture of 1-
Received in revised form Heptane þ phenols þ ethylene glycol were obtained at 323.15 K and 343.15 K under atmospheric
3 March 2016
pressure, where phenols refer to phenol, o-cresol, m-cresol and p-cresol. The tie lines were displayed in
Accepted 3 March 2016
Available online 4 March 2016
ternary phase diagrams. The distribution coefficient and selectivity were calculated according to the
measured tie-line data. The calculation results revealed that the selectivity was large enough for the
extraction application. The root-mean-square deviation (RMSD) value of the NRTL model was a little
Keywords:
Liquideliquid equilibria
lower than that of UNIQUAC, demonstrating NRTL can better fit the experimental data.
Coal tar © 2016 Elsevier B.V. All rights reserved.
1-Heptane
Phenols
Ethylene glycol

1. Introduction
Coal tar, as a byproduct of coal coking, is an important feedstock
for the coal chemical industry due to its numerous and valuable
components. In the chemical industry, distillation is widely applied
to separate light components of coal tar. After distillation, several
fractions of high temperature tar are roughly obtained: BTX,
carbolic oil, naphthalene oil, wash oil, anthracene oil and coal-tar
pitch [1]. However, the distillation process demands high energy
at a high cost. Therefore, it is necessary to find a lower consumption
process like the extraction technology to obtain phenols in low-
temperature tar. Phenol and phenol derivatives mainly obtained
from coal tar are widely used in the production of polycarbonates
and phenolic resins [1].
An optimized planning and realization of extraction plants in
the chemical industry will require comprehensively qualitative and
quantitative phase equilibrium data [2]. Due to the complicated
components of tar, it would be an insurmountable task to
completely obtain the liquideliquid equilibrium data of every in-
dividual compound and solvent. Therefore, a simplification method
is used in this work, as referred to that in Catherine A. Peter's report
[3]. 1-Heptane and phenols are used as the main representative
components in low temperature tar. Considering the polarities of
these components are different, the polar solvent may be used as
the extractant to separate coal tar. Ethylene glycol, as a typical polar
solvent, has been used for phenol extract and arenes separation
from nonaromatic compounds [4,5].
At present, the liquideliquid equilibria data of a coal tar system
is rarely reported in literature, especially for the system containing
phenol. According to Zhang's report [6], phenol, o-cresol, m-cresol,
and p-cresol are the main phenols in low temperature coal tar. So,
the LLE data for the ternary mixture of 1-Heptane þ phenols
þ ethylene glycol were measured at 323.15 and 343.15 K under
atmospheric pressure. Tie-lines are determined for the ternary
systems according to the LLE data. The distribution coefficient and
separation factor were calculated according to the LLE data and
used as the standard to evaluate the separation efficiency. The
experimental data are also correlated with Non-Random Two Liq-
uids (NRTL) [7] and Universal Quasi-Chemical (UNIQUAC) [8] ac-
tivity coefficient models.
2. Experimental

2.1. Materials
* Corresponding author.
E-mail address: licup01@163.com (Q. Li). The chemicals used were 1-Heptane, phenol, o-cresol, m-cresol,

http://dx.doi.org/10.1016/j.fluid.2016.03.003
0378-3812/© 2016 Elsevier B.V. All rights reserved.
 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56 51

Nomenclature s(Xi ) experimental standard deviation of input mean Xi

equal to the positive square root of s2 ðXi Þ
g, u interaction parameters S the separation factor
K the distribution coefficient u(xi) standard uncertainty of input estimate xi that
M number of tie lines estimates input quantity Xi
OF objective function wi mass fraction composition of liquid phase
q area parameter in UNIQUAC equation or randomly xi estimate of input quantity Xi
varying quantity described by a probability Xi i th input quantity on which the measurand depends
distribution Xi estimate of the value of input quantity Xi
q arithmetic mean of n independent repeated
observations qk of randomly-varying quantity q Greek letter
qk kth independent repeated observation of randomly- a non-randomness factor in NRTL equation
varying quantity q
r volume parameter in UNIQUAC equation Superscripts
RMSD root mean square deviation ^ calculated
sðqÞ experimental standard deviation of the mean q, equal
to the positive square root of s2 ðqÞ Subscript
s(qk) experimental standard deviation, equal to the positive i, j component identification: 1, 2, 3
square root of s2(qk) k number of the tie line

p-cresol and ethylene glycol, which were purchased from Sino-

pharm Chemical Reagent. The details about the chemicals are
presented in Table 1. The purities of these components were tested
by gas chromatography and no appreciable peaks of impurities
were found, which indicated no purification methods were needed
for the LLE measurements.

2.2. Apparatus and procedure

LLE data for the studied ternary system were obtained at 323.15
and 343.15 K under atmospheric pressure. The details about
experimental equipment have been presented in our previous work
and the reliability of the experimental system has been evaluated
[9]. The set-up is shown in Fig. 1. The agitation and settling time
were investigated, demonstrating that the equilibrium could be
attained in 1 h. So, this mixture was agitated vigorously for at least
1 h and left for 1 h before GC measurements. Meanwhile, the air in
the still was replaced by N2 in order to prevent the oxidation of
phenol. The evaporated compounds were completely condensed by
the condenser to ensure the mass balance.
After 1 h of settlement, the liquideliquid equilibrium was
formed. Afterwards, the samples were taken from the still and then
analyzed by GC. Agilent GC6820 gas chromatograph equipped with
a thermal conductivity detector and Porapak N column
(3 mm  3 m) was used. The carrier gas was hydrogen with a flow
rate of 60 mL/min. The temperature of the injector and detector was
523.15 K. The column temperature was kept at 423.15 K for 0.5 min,
then increased to 523.15 K at a rate of 20  C/min and maintained at
this temperature for 6.5 min. The calibration area normalization
method was used to obtain quantitative results in the analysis. Each

Table 1
Materials description. Fig. 1. (Liquid þ liquid) equilibrium still. 1. thermometer; 2. circulating water chamber;
3. equilibrium cell chamber; 4. circulating water inlet; 5. condenser; 6. circulating
Component CAS Source GC purity (mass%) water outlet; 7.light phase port; 8. heavy phase port 9. magnetic stirrer.
1-Heptane 142-82-5 Sinopharm 99.97
Phenol 108-95-2 Sinopharm 99.90
o-Cresol 95-48-7 Sinopharm 99.26 sample was analyzed at least three times and the average value was
m-Cresol 108-39-4 Sinopharm 99.19 used as the sample composition. According to GUM standard [10],
p-Cresol 106-44-5 Sinopharm 99.37
the standard uncertainty of the liquid phase compositions was
Ethylene glycol 107-21-1 Sinopharm 99.59
52 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56

Table 2 describing the variability of the observed values qk or more spe-

UNIQUAC structural parameters (r and q). cifically, their dispersion about their mean q. Thus, for an input
Component r q quantity Xi, the standard uncertainty u(xi) of its estimate xi ¼ Xi is
1-Heptane 5.1740 4.3960
u(xi) ¼ s(Xi). Generally, u(xi) calculated by this method is some-
Phenol 3.5465 2.7160 times called a Type A standard uncertainty.
o-Cresol 4.2867 3.2480
m-Cresol 4.2867 3.2480
p-Cresol 4.2867 3.2480 2.4. Separation efficiency calculation
Ethylene glycol 2.4086 2.2479
The separation efficiency of solvents is evaluated by the distri-
bution coefficient (K) and separation factor (S). K and S are calcu-
calculated. A series of LLE data were obtained by changing the feed lated from the tie-line data according to the equations as follows.
composition.
w2b
K¼ (4)
w2a
2.3. Uncertainty calculation

w2b w1b
There are two categories for calculating the uncertainty, named S¼ (5)
w2a =w1a
as “A” and “B” [10]. In this work, Type A was used to evaluate the
uncertainty of every quantity q. The corresponding calculation
where w2b and w1b are the concentration of solute and diluent in
equations are given as follows.
the solvent phase, w2a and w1a are the mass fraction of solute and
n 
X 2 . diluent in the dilute phase, respectively.
s2 ðqk Þ ¼ qj  q ðn  1Þ (1)
j¼1
3. Thermodynamics modeling
.
s2 ðqÞ ¼ s2 ðqk Þ n (2) The NRTL and UNIQUAC models were applied to correlate the
experimental mole fractions by using the Aspen Plus 7.2 software.
  There are three adjustable parameters gij, gji, and aij in the NRTL
uðxi Þ ¼ s Xi (3)

The experimental standard deviation (s(qk)) is calculated, Table 4

Experimental LLE Data (mass Fraction) for 1-Heptane (1) þ Phenols (2) þ Ethylene
glycol (3) system at 343.15 K under atmosphere pressure.a
Table 3
Experimental LLE Data (mass Fraction) for 1-Heptane (1) þ Phenols (2) þ Ethylene Solute 1-Heptane phase Ethylene glycol phase K S
glycol (3) system at 323.15 K under atmosphere pressure.a w1 w2 w3 w1 w2 w3
Solute 1-Heptane phase Ethylene glycol phase K S Phenol 0.9980 0.0018 0.0002 0.0033 0.0383 0.9584 21.28 6435
0.9958 0.0042 0.0001 0.0124 0.1574 0.8302 37.48 3010
w1 w2 w3 w1 w2 w3
0.9929 0.0070 0.0000 0.0182 0.2248 0.7570 32.11 1752
Phenol 0.9979 0.0018 0.0002 0.0026 0.0538 0.9435 29.89 11472 0.9888 0.0112 0.0001 0.0217 0.3290 0.6493 29.38 1339
0.9959 0.0039 0.0002 0.0057 0.1712 0.8232 43.90 7670 0.9852 0.0146 0.0002 0.0308 0.3981 0.5711 27.27 872
0.9934 0.0063 0.0003 0.0084 0.2497 0.7418 39.63 4687 0.9786 0.0212 0.0002 0.0395 0.4779 0.4826 22.54 558
0.9908 0.0090 0.0003 0.0125 0.3195 0.6680 35.50 2814 0.9715 0.0283 0.0002 0.0522 0.5406 0.4072 19.10 356
0.9877 0.0121 0.0003 0.0174 0.3892 0.5934 32.17 1826 0.9585 0.0405 0.0011 0.0714 0.5938 0.3348 14.66 197
0.9813 0.0184 0.0002 0.0263 0.5027 0.4710 27.32 1019 0.9386 0.0599 0.0015 0.0716 0.6616 0.2668 11.05 145
0.9678 0.0316 0.0006 0.0486 0.6263 0.3251 19.82 395 o-Cresol 0.9958 0.0041 0.0001 0.0030 0.0462 0.9508 11.27 3740
0.9591 0.0404 0.0005 0.0673 0.6835 0.2493 16.92 241 0.9829 0.0168 0.0004 0.0059 0.1522 0.8418 9.06 1509
o-Cresol 0.9980 0.0020 0.0001 0.0021 0.0294 0.9685 14.70 6986 0.9779 0.0220 0.0001 0.0103 0.2347 0.7550 10.67 1013
0.9941 0.0058 0.0001 0.0038 0.1053 0.8909 18.16 4749 0.9703 0.0295 0.0001 0.0166 0.3090 0.6744 10.47 612
0.9892 0.0106 0.0002 0.0065 0.1834 0.8101 17.30 2633 0.9628 0.0371 0.0002 0.0229 0.3693 0.6078 9.95 419
0.9847 0.0151 0.0002 0.0122 0.2834 0.7044 18.77 1515 0.9515 0.0478 0.0007 0.0315 0.4205 0.5480 8.80 266
0.9768 0.0231 0.0001 0.0225 0.3821 0.5955 16.54 718 0.9385 0.0609 0.0006 0.0458 0.4842 0.4701 7.95 163
0.9675 0.0324 0.0001 0.0351 0.4525 0.5124 13.97 385 0.9231 0.0754 0.0015 0.0624 0.5270 0.4106 6.99 103
0.9553 0.0446 0.0001 0.0537 0.5240 0.4223 11.75 209 0.9049 0.0928 0.0023 0.0778 0.5613 0.3609 6.05 70
0.9268 0.0730 0.0002 0.0913 0.5990 0.3098 8.21 83 m-Cresol 0.9964 0.0035 0.0001 0.0032 0.0440 0.9528 12.57 3914
m-Cresol 0.9968 0.0021 0.0011 0.0027 0.0415 0.9558 19.76 7296 0.9909 0.0086 0.0004 0.0062 0.1366 0.8572 15.88 2539
0.9939 0.0053 0.0008 0.0055 0.1454 0.8491 27.43 4958 0.9837 0.0156 0.0007 0.0102 0.2203 0.7694 14.12 1362
0.9935 0.0064 0.0001 0.0113 0.2352 0.7536 36.75 3231 0.9829 0.0169 0.0002 0.0156 0.2889 0.6955 17.09 1077
0.9921 0.0078 0.0001 0.0185 0.3230 0.6585 41.41 2221 0.9758 0.0234 0.0008 0.0229 0.3582 0.6189 15.31 652
0.9836 0.0163 0.0001 0.0314 0.4186 0.5500 25.68 804 0.9693 0.0300 0.0007 0.0344 0.4135 0.5521 13.78 388
0.9828 0.0171 0.0001 0.0425 0.4728 0.4847 27.65 639 0.9649 0.0344 0.0007 0.0465 0.4693 0.4842 13.64 283
0.9784 0.0215 0.0001 0.0560 0.5131 0.4309 23.87 417 0.9617 0.0381 0.0002 0.0616 0.5130 0.4254 13.46 210
0.9698 0.0301 0.0001 0.0736 0.5588 0.3676 18.56 245 0.9353 0.0632 0.0015 0.0974 0.5752 0.3274 9.10 87
p-Cresol 0.9979 0.0019 0.0002 0.0025 0.0328 0.9647 17.26 6891 p-Cresol 0.9974 0.0025 0.0002 0.0034 0.0450 0.9516 18.00 5280
0.9958 0.0041 0.0001 0.0047 0.1202 0.8751 29.32 6211 0.9925 0.0071 0.0004 0.0061 0.1365 0.8573 19.23 3128
0.9937 0.0060 0.0002 0.0072 0.1963 0.7965 32.72 4515 0.9887 0.0109 0.0004 0.0075 0.1702 0.8224 15.61 2058
0.9926 0.0072 0.0001 0.0131 0.2673 0.7196 37.13 2813 0.9847 0.0152 0.0001 0.0116 0.2419 0.7465 15.91 1351
0.9900 0.0099 0.0001 0.0224 0.3443 0.6333 34.78 1537 0.9781 0.0216 0.0003 0.0189 0.3142 0.6669 14.55 753
0.9882 0.0116 0.0002 0.0327 0.4160 0.5513 35.86 1084 0.9741 0.0252 0.0007 0.0273 0.3735 0.5992 14.82 529
0.9800 0.0197 0.0003 0.0462 0.4743 0.4795 24.08 511 0.9715 0.0283 0.0002 0.0528 0.4769 0.4702 16.85 310
0.9709 0.0280 0.0011 0.0716 0.5470 0.3814 19.54 265 0.9602 0.0397 0.0001 0.0850 0.5464 0.3687 13.76 155
a a
Standard uncertainties u are u(T) ¼ 0.1 K, u(p) ¼ 0.3 kPa and u(w) ¼ 0.003. Standard uncertainties u are u(T) ¼ 0.1 K, u(p) ¼ 0.3 kPa and u(w) ¼ 0.002.
 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56 53

model. The recommended aij values for mixtures of different types

are commonly between 0.1 and 0.47.
There are two contributions to the excess Gibbs energy and the
X
M X 3 
2 X 2
activity coefficient in UNIQUAC model [8]. They are a combinatorial
OF ¼ b ijk
wijk  w (6)
term related to differences in size and shape between the compo-
k¼1 j¼1 i¼1
nents and a residual (energetic) term accounting for energy dif-
ferences between the molecules. The structural parameters (r and
where M is the number of tie lines, w is the experimental value, and
q) of the pure components are measures of the molecular volume
c is the calculated value. The subscripts i, j and k refer to the
w
and area, respectively. The r and q values of pure components are
components, the phases and the tie line, respectively. The function
recommended by DECHEMA [11] and given in Table 2.
describes the differences between the experimental and calculated
The interaction parameters between components are essential
equilibrium data for each of the components over all the experi-
when NRTL and UNIQUAC models are applied. They can be obtained
mental tie lines. The experimental data was correlated at each
by minimizing the objective function (OF) of the composition. The
temperature respectively. Meanwhile, the quality of the correlation
objective function is presented as follows:
is assessed by root-mean-square deviation (RMSD) calculated ac-
cording to the following equation:

0.00
1.00
(a)

0.25
0.75
l
yco

Ph
eno
Gl

0.50
0.50
l

0.75
0.25

1.00
0.00
0.00 0.25 0.50 0.75 1.00

1-Heptane

0.00
(b) 1.00

0.25
0.75
l
yco

Ph

0.50
eno
Gl

0.50
l

0.75
0.25

1.00
0.00
0.00 0.25 0.50 0.75 1.00
1-Heptane
Fig. 2. Ternary phase diagram for 1-Heptane þ Phenol þ Ethylene glycol system, (a) at Fig. 3. Ternary phase diagram for 1-Heptane þ o-Cresol þ Ethylene glycol system, (a)
323.15 K (b) at 343.15 K, ( ) feed composition; (-) experimental data; (B) NRTL at 323.15 K (b) at 343.15 K, ( ) feed composition; (-) experimental data; (B) NRTL
▵
model; ( ) UNIQUAC model. ▵
model; ( ) UNIQUAC model.
54 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56

fraction. 1-Heptane, phenols and ethylene glycol are identified as

8  2 91=2
>
<X M X 2 X 3 c ijk >
wijk  w = component 1, 2 and 3 respectively. The phase diagrams for the
RMSD ¼ 100* (7) ternary mixtures are presented in Figs. 2e5. The tie-lines and feed
>
: k¼1 j¼1 i¼1 6M >
; compositions are also plotted in Figs. 2e5. As displayed in these
figures, the points of the feed compositions agree the tie line with
c ,i, j, k are the same with those in the OF great accuracy, which is in accord with the lever rule. The result
the corresponding M, w, w
reveals that the mass balance is satisfied during the experimental
equations.
operations, demonstrating the reliability of experimental data [12].
As shown in these figures and tables, it can be found that phenols
4. Results and discussion have a higher solubility in ethylene glycol than in 1-Heptane.
The values of K and S calculated from the experimental data are
4.1. Experimental data given in Tables 3 and 4 As presented in these tables, it can be
observed that the distribution coefficients tend to decrease when
The LLE data for the ternary systems (1-Heptane þ phenols the concentration of phenols increases. Simultaneously, the tie line
þ ethylene glycol) at temperatures of 323.15 and 343.15 K are slope factor reflects the same tendency. It is known that the dis-
shown in Tables 3 and 4 with all concentrations expressed in mass tribution coefficient is related with the solubility of solute which

Fig. 4. Ternary phase diagram for 1-Heptane þ m-Cresol þ Ethylene glycol system, (a)
at 323.15 K (b) at 343.15 K, ( ) feed composition; (-) experimental data; (B) NRTL
▵
model; ( ) UNIQUAC model.
Fig. 5. Ternary phase diagram for 1-Heptane þ p-Cresol þ Ethylene glycol system, (a)
at 323.15 K (b) at 343.15 K, ( ) feed composition; (-) experimental data; (B) NRTL
▵
model; ( ) UNIQUAC model.
 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56 55

same, selectivity for phenol extraction is the highest. That's related

with the higher polarity of phenol in ethylene glycol in comparison
of cresols.

4.2. Data correlation

The data correlations were carried out by using NRTL and

UNIQUAC models. The interaction parameters of these models are
obtained after data regression when the OF value is the smallest.
Thus, the value of RMSD is the lowest, which is a measure of models
correlation.
The results of data regression are listed in Table 5. Generally,
since all the RMSD values are below 1%, the NRTL model fits the
experimental data better than UNIQUAC model. The calculated LLE
data from NRTL and UNIQUAC models are also plotted in Figs. 2e5,
demonstrating that the two models can correlate well with the
experimental data.

5. Conclusions

The liquideliquid equilibria for 1-Heptane þ phenols þ ethylene

glycol ternary system were investigated at 323.15 and 343.15 K
Fig. 6. Experimental separation factor versus the phenols mass fraction (x2b) in the under atmospheric pressure. It was found that phenols have a
Ethylene glycol phase for the ternary system at 323.15 K. higher solubility in ethylene glycol than in 1-Heptane. The sepa-

Table 5
Binary energy parameters of models for 1-Heptane (1) þ Phenols (2) þ Ethylene glycol (3) system.

T/K i-j NRTL parameters RMSD UNIQUAC parameters RMSD

gij-gjja gji-giia a uij-ujja uji-uiia

Phenol 323.15 1e2 5170.86 3425.74 0.28 0.84 3407.62 549.70 0.85
1e3 16242.92 13584.30 0.20 8652.14 2075.51
2e3 5447.22 9985.80 0.28 3176.17 5138.14
Phenol 343.15 1e2 6028.66 5881.72 0.25 0.78 3923.35 1153.95 1.36
1e3 21955.67 6566.77 0.10 9941.23 1184.75
2e3 35140.26 9985.80 0.47 0.00 1292.35
o-Cresol 323.15 1e2 8966.06 1100.67 0.20 0.68 5279.51 1526.35 0.46
1e3 18516.07 14206.03 0.20 9337.19 1815.41
2e3 5145.79 1206.29 0.47 161613.73 1608.14
o-Cresol 343.15 1e2 729.00 5537.08 0.25 0.62 2613.04 587.08 0.60
1e3 21662.67 7892.44 0.10 11702.26 1546.08
2e3 22042.26 17999.24 0.37 364.80 206.18
m-Cresol 323.15 1e2 14261.18 2617.17 0.16 0.92 5097.97 1586.03 0.89
1e3 17298.29 14242.12 0.20 8902.25 1883.69
2e3 31900.06 570.58 0.20 259043.46 2641.21
m-Cresol 343.15 1e2 1945.89 4578.37 0.16 0.83 4729.82 1511.68 0.88
1e3 18179.17 15529.51 0.2 8406.66 1670.93
2e3 8090.77 16490.25 0.2 4280.63 1590.99
p-Cresol 323.15 1e2 76956.00 1408.90 0.20 0.81 3620.87 995.04 1.06
1e3 19247.42 13478.68 0.20 9675.88 1796.42
2e3 733.41 83140.00 0.30 2483.13 4741.47
p-Cresol 343.15 1e2 14383.65 2195.91 0.20 0.92 5876.32 1555.59 0.81
1e3 17366.58 13932.10 0.20 9098.87 1699.81
2e3 17397.96 2226.25 0.30 225787.46 1226.32
a
The unit of parameters is J/mol.

has an effect on the tie line slope [13,14]. As presented in Tables 3
and 4, the separation factors in all cases are more than one,
meaning that the extraction process is possible. Meanwhile, the
separation factor decrease as the concentration of phenols in-
creases. That's because the two-phase region shrinks when the
concentration of phenols increases, demonstrating that the sepa-
ration capacity of the solvent is reduced [15]. The experimental
separation factor versus phenols mass fraction in the ethylene
glycol phase at 323.75 K is demonstrated in Fig. 6. When the con-
centrations of phenols are low, the values of S are much higher,
indicating ethylene glycol is an ideal extractant for separating
phenols. Meanwhile, when the concentrations of phenols are the
ration factors are much larger than one, implying the feasibility of
ethylene glycol to extract the phenols from low-temperature coal
tar. NRTL and UNIQUAC models are applied to correlate experi-
mental data. The corresponding optimized interaction parameters
of models were obtained. Compared with the UNIQUAC model, the
RMSD value of NRTL was a little smaller, indicating that the NRTL
activity coefficient model can better correlate the LLE experimental
results.
Acknowledgments
This study is supported by Qingdao postdoctoral fund (NO.
56 F. Dai et al. / Fluid Phase Equilibria 419 (2016) 50e56
T1504117) and the Project of Natural Science Research of Higher
Education Institutions of Jiangsu Province (NO. 15KJD530001).
Harold G. Sherrard is greatly acknowledged for his attribution to
the manuscript's language modification.
References
[1] M. Granda, C. Blanco, P. Alvarez, J.W. Patrick, R. Mene ndez, Chem. Rev. 114
(2014) 1608e1636.
[2] A. Martin, M. Klauck, A. Grenner, R. Meinhardt, D. Martin, J. Schmelzer,
J. Chem. Eng. Data 56 (2011) 741e749.
[3] C.A. Peters, R.G. Luthy, Environ. Sci. Technol. 27 (1993) 2831e2843.
[4] B. Semeniuk, Fluid Phase Equilib. 115 (1996) 193e204.
[5] A.A.S. Taher, K. Emina, Fluid Phase Equilib. 118 (1996) 271e285.
[6] T.T. Jiao, M.M. Gong, X.L. Zhuang, C.S. Li, S.J. Zhang, J. Ind. Eng. Chem. 29 (2015)
344e348.
[7] H. Renon, J.M. Prausnitz, AIChE J. 14 (1968) 135e144.
[8] D.S. Abrams, J.M. Prausnitz, AIChE J. 21 (1975) 116e128.
[9] F.F. Dai, K. Xin, Y.H. Song, M.D. Shi, Y.M. Yu, Q.S. Li, Fluid Phase Equilib. 409
(2016) 466e471.
[10] BIPM, IEC, IFCC, ILAC, ISO, IUPAC, IUPAP and OIML, Evaluation of Measurement
Data- Guide to the Expression of Uncertainty in Measurement, 2008.
[11] J. Gmehling, U. Onken, VaporeLiquid Equilibrium Data Collection, DECHEMA,
Chemistry Data Series, Frankfurt, Germany, 1977.
[12] Y.X. Li, Q. Xu, S.L. Liu, H.M. Li, F.B. Zhang, G.L. Zhang, Q. Xia, Fluid Phase Equilib.
412 (2016) 199e204.
[13] J.S. Santos, A.P.L. Craig, J.M.O. Santana, A.F. Santos, M.F. Heredia, M.L. Corazza,
E. Franceschi, C. Dariva, J. Chem. Eng. Data 60 (2015) 2050e2056.
[14] A. Chafer, J. de la Torre, E. Lladosa, J.B. Monton, Fluid Phase Equilib. 361 (2014)
23e29.
[15] M.S. Al-Tuwaim, K.H.A.E. Alkhaldi, M.S. Fandary, A.S. Al-Jimaz, Fluid Phase
Equilib. 315 (2012) 21e28.