Separation of ethyl acetate and ethanol azeotrope mixture using dialkylphosphates-based ionic liquids as entrainers

Separation of ethyl acetate and ethanol azeotrope mixture using dialkylphosphates-based ionic liquids as entrainers

Fluid Phase Equilibria 454 (2017) 91e98 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...

888KB Sizes 1 Downloads 241 Views

Fluid Phase Equilibria 454 (2017) 91e98

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

Separation of ethyl acetate and ethanol azeotrope mixture using dialkylphosphates-based ionic liquids as entrainers Zhigang Zhang, Kaifang Wu, Qinqin Zhang**, Tao Zhang, Debiao Zhang, Ru Yang, Wenxiu Li* Liaoning Provincial Key Laboratory of Chemical Separation Technology, Shenyang University of Chemical Technology, Shenyang, 110142, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 July 2017 Received in revised form 28 August 2017 Accepted 19 September 2017 Available online 22 September 2017

Isobaric vapor-liquid equilibrium (VLE) data for the ternary systems of ethyl acetate þ ethanol þ 1-ethyl3-methylimidazolium diethylphosphate ([EMIM][DEP]), ethyl acetate þ ethanol þ 1-butyl-3methylimidazolium diethylphosphate ([BMIM][DEP]) and ethyl acetate þ ethanol þ 1-butyl-3methylimidazolium dibutylphosphate ([BMIM][DBP]) were measured at 101.3 kPa. The results showed that all the three ionic liquids (ILs) produced remarkable salting-out effect, leading to the increase of the relative volatility of ethyl acetate to ethanol. Hence, the azeotropic point of ethyl acetate and ethanol could be eliminated with the addition of a certain content of ILs. The separation effect of the three ILs follows the order of [EMIM][DEP] > [BMIM][DEP] > [BMIM][DBP]. Then, the experimental VLE data were well correlated with the nonrandom two-liquid (NRTL) model. Finally, the s-Profiles of solvents and ILs were used to analysis the micro-mechanism of the different separation performance of the three investigated ILs. © 2017 Published by Elsevier B.V.

Keywords: Vapor liquid equilibrium Ionic liquids Ethyl acetate Ethanol NRTL model s-Profiles

1. Introduction Ethyl acetate is an important solvent and widely used in the chemical industry. In recent years, the Fischer esterification reaction of acetic acid [1] and the ethanol dehydrogenation method [2] have been used to produce ethyl acetate. In those processes, the mixture of ethyl acetate and ethanol would inevitably at atmospheric pressure form a minimum azeotrope [3] which is impossibly separated by conventional distillation. Thus, some special separation technologies have been employed to separate the azeotrope, such as azeotropic distillation, pressure-swing distillation, reactive distillation, extractive distillation, liquid-liquid extraction, and membrane separation. Among them, extractive distillation is the most widely used process, in which a third component (entrainer) is added to enhance the initial relative volatility and to make the separation easier. Therefore, the key to extractive distillation is to select a highly effective entrainer [4e6]. During the past few years, ionic liquids (ILs) have become attractive options (alternatives) for entrainers in extractive

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (Q. Zhang), [email protected] (W. Li). https://doi.org/10.1016/j.fluid.2017.09.016 0378-3812/© 2017 Published by Elsevier B.V.

distillation due to their remarkable physicochemical properties [7e14]. The negligible vapor pressure and high thermal stability make it easy to recover and reuse after the separation process. Existing in liquid state over a wide temperature range allows them to pass through process equipments without clogging up pipes. Last but not the least, ILs can be “designed” by judicious combination of cation and anion, leading a specific and effective entrainer for a given azeotropic system. Obviously, ILs have the unique advantages over conventional organic solvents and inorganic salts. In terms of the ethyl acetate and ethanol azeotropic system, many ILs have been used as entrainers to improve the separation s et al. [15] reported that 1effect of extractive distillation. Orchille ethyl-3-methylimidazolium trifluoromethanesulfonate ([EMIM] [F3CSO3]) could eliminate the azeotropic point at 100 kPa when the mole fraction of ionic liquid in liquid phase was greater than 0.20. Qunsheng Li et al. [16e18] investigated the effects of three tetrafluoroborate-based ionic liquids, namely 1-ethyl-3methylimidazolium tetrafluoroborate([EMIM][BF4]), 1-butyl-3methylimidazolium tetrafluoroborate ([BMIM][BF4]) and 1-octyl3-methylimidazolium tetrafluoroborate ([OMIM][BF4]), on the vapor-liquid phase equilibrium (VLE) behavior of ethyl acetate þ ethanol at 101.32 kPa. It was found that all the three ILs produced a significant salting-out effect with the mole fraction of ionic liquid greater than 0.20. Rui Li et al. [19] measured the VLE

92

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

data of ethyl acetate (1) þ ethanol (2) þ1-ethyl-3methylimidazolium acetate([EMIM][Ac]) (3) at 101.30 kPa, and found [EMIM][Ac] could break the azeotrope as its mole fraction in liquid phase was greater than 0.10. Yingjie Xu et al. [20] compared the influences of [BMIM][BF4] and 1-butyl-2,3dimethylimidazolium tetrafluoroborate ([BMMIM][BF4]) on the VLE behavior of the ethyl acetate (1) þ ethanol (2) system. Andreatta et al. [21] studied the abilities of six ILs to break the ethanol þ ethyl acetate azeotrope at 313.15 K, including 1-ethyl-3methylimidazolium bis(trifluoromethylsulfonyl)imide ([EMIM] [Tf2N]), 1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl) imide ([HMIM][Tf2N]), 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide ([BMPYRR][Tf2N]), 1-ethyl-3methylimidazolium methanesulfonate ([EMIM][MeSO3]), 1-ethyl3-methylimidazolium methylsulfate ([EMIM][MeSO4]), and 1butyl-3-methylimidazolium trifluoromethanesulfonate ([BMIM] [CF3SO3]). And among those ILs, [EMIM][MeSO3] and [EMIM] [MeSO4] exhibited the best performance for breaking the azeotrope with x3 ¼ 0.025 at 313.15 K. As far as we know, the separation performance of ILs is strongly dependent on their structures, including the types of cations and anions, molecular size, side-chain length and so on. They determine the type and strength of the molecular interactions between ILs and solvents. Molecular interactions include electrostatic interaction, hydrogen bond and Van der Waals forces [22,23]. Among them, hydrogen bond is frequently employed to explain the different separation performance of ILs as entrainers in extractive distillation. In the previous work [24e28], the screening charge density profile (s-profile) were applied to vividly analyze hydrogen bond interactions among different molecules in the mixtures of polar solvents, and the results were in good agreement with experimental data. In this work, [EMIM][DEP], [BMIM][DEP] and [BMIM][DBP] were employed as entrainers to separate the ethyl acetate þ ethanol azeotropic system. The VLE data of the ethyl acetate (1) þ ethanol (2) þ IL (3) ternary system were measured at 101.3 kPa. Then, the experimental data were correlated with the nonrandom two-liquid (NRTL) activity coefficient model [29]. Finally, the separation performance of the three ILs was discussed by means of the s-profiles of solvents and ILs, which are able to identify the strength of the hydrogen bond interactions between solvents and ILs.

Ethyl acetate and ethanol were supplied by Sinopharm Group CO. Ltd. The purities of all reagents were confirmed to be analytical grade by chromatography. The ILs [EMIM][DEP] and [BMIM][DBP] were synthesized in our laboratory according to the method mentioned literature [30]. The water mass fraction of ILs were determined by Karl Fischer titration and the mass fraction of ILs were measured by liquid chromatography. [BMIM][DEP] was provided by Yulu Chemical. The specifications of chemical samples used are given in Table 1 and the structures of the three ILs are shown in Fig. 1.

pressure was controlled by a gas buffer connected with the still and measured by a manometer with a standard uncertainty of 0.1 kPa. At the beginning, the vapor-liquid boiling chamber was filled with a sufficient concentration of sample solution. The heating rod was used to heat the liquid to boil, the rough surface of which prevent the liquid from bumping. The vapor is separated from the liquid in the vapor-liquid chamber and condensed in the spherical condenser then refluxed into the vapor-liquid chamber for continuous circulation. The system temperature was measured using a standard and calibrated thermometer with a standard uncertainty of 0.05 K. The system reached equilibrium when the system temperature remained unchanged more than 30 min, then analysis samples can be taken from the liquid phase sampling point and the vapor phase sampling point. All the solutions were gravimetrically prepared with a digital balance (ACCULAB ATL-224-1 China), and the standard uncertainty is 0.0001 g. For the binary system of ethyl acetate þ ethanol, some ethanol was added to the pure ethyl acetate until a very diluted solution was obtained. For each ternary system, the mixture of ethyl acetate and scheduled mole fraction of IL was prepared, where the mixture of ethanol and the same mole fraction of IL was added, trying to keep a constant mole fraction of IL in each series. But for [EMIM][DEP] and [BMIM][DEP], they are not miscible with ethyl acetate at room temperature. Fortunately, adding a very small amount of ethanol could generate a one phase ternary system. So we must prepared the ternary solution of ethyl acetate þ ethanol þ IL firstly and then different quantities of the ethanol þ IL mixture were added to keep constant the mole fraction of IL in each series. In this experiment, the contents of ethyl acetate and ethanol in every gas and liquid samples were placed in headspace sampler (G1888 Network headspace sampler, Agilent Technologies) and then analyzed by the gas chromatography (Agilent GC7890A) with a thermal conductivity detector (TCD), DB-WAXETR capillary column (30 m  0.32 mm, 1 mm). The temperatures of column, injector and detector were 338 K, 473 K and 473 K, respectively. Since the ionic liquid has a nonvolatile property, the content of the ionic liquid in the solution was measured by weighing the mass of the liquid phase before and after dried in the vacuum desiccation at 393 K for 48 h. The standard uncertainty of ILs components in liquid phases was 0.001 in mole fraction. In this work, a quantum chemical calculations based on density functional theory (DFT) was used to obtain the s-Profiles of ethyl acetate, ethanol and the ions of ILs, the procedures were carried out as the following three steps [32e35]. The first step is to derive the optimized molecular geometry in the ideal gas phase with DFT to confirm the configuration in global energy minimum. For this purpose, every compound was optimized with the VWN-BP method at the version 4.0.0 DNP basis set by employing the DMol3 module implemented in the MS. The second step is to proceed “Energy task” for the geometry optimization output files to obtain the COSMO files, which contains the volume of the cavity, screening surfaces and number of surface segments, etc. In the final step, the s-profiles was obtained by averaging the surface charge densities from input the COSMO files to the Fortran program “Sigma-average_v2.exe”.

2.2. Apparatus and procedure

3. Results and discussion

The isobaric VLE data were measured by an all-glass dynamic recirculating still which has been described in detail in a previous literature [31]. The still is mainly composed of six parts: a vapor liquid boiling chamber, a heating rod, a liquid phase sampling point, a spherical condenser, a thermometer and a vapor phase sampling point. The still has a total capacity of about 100 cm3. The system

3.1. Experimental data

2. Experimental 2.1. Chemicals

The binary vapor-liquid equilibrium data for ethyl acetate (1) þ ethanol (2) were measured at 101.3 kPa and compared with the reported data [17,20] for the purpose of validating the accuracy of our apparatus and methods. The experimental data are listed in

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

93

Table 1 Specification of chemical samples. Chemical name

CAS RN

Source

Purity

Purification method

Analysis method

Ethyl acetate Ethanol [EMIM][DEP]a [BMIM][DBP]b [BMIM][DEP]c

141-78-6 64-17-5 848641-69-0 663199-28-8 None

Sinopharm Group Sinopharm Group Prepared in this work Prepared in this work Yulu Chemical

0.995 0.998 0.980 0.980 0.990

None None Vacuum desiccation Vacuum desiccation Vacuum desiccation

GCd GCd KFe and LCf KFe and LCf KFe and LCf

a b c d e f

[EMIM][DEP] ¼ 1-ethyl-3-methylimidazolium diethylphosphate. [BMIM][DBP] ¼ 1-butyl-3-methylimidazolium dibutylphosphate. [BMIM][DEP] ¼ 1-butyl-3-methylimidazolium diethylphosphate. GC ¼ Gas chromatography. KF ¼ Karl Fischer titration. LC ¼ Liquid chromatography.

Fig. 1. The structures of the three ILs in this work.

Table 2 Experimental isobaric VLE data for the binary system of ethyl acetate (1) þ ethanol (2) at 101.3 kPa.a T/K

x1

y1

T/K

x1

y1

350.26 349.17 348.36 347.23 346.26 345.45 344.95 345.00

1.000 0.969 0.939 0.883 0.826 0.735 0.620 0.518

1.000 0.939 0.890 0.812 0.750 0.670 0.591 0.529

345.35 345.83 346.69 347.42 348.45 349.40 350.15 351.50

0.401 0.308 0.226 0.159 0.107 0.067 0.037 0.000

0.459 0.397 0.331 0.265 0.199 0.138 0.082 0.000

a

this work are reliable. The ternary VLE data for ethyl acetate (1) þ ethanol (2) þ [EMIM][DEP] (3) or [BMIM][DEP] (3) or [BMIM][DBP] (3) were determined at 101.3 kPa by keeping the mole fractions of three ILs nearly constant at 0.02, 0.07 and 0.12. The experimental data are listed in Tables 3e5, where T is the equilibrium temperature, x1 is the mole fraction of ethyl acetate in the liquid phase, x3 is the mole fraction of ILs in the liquid phase, x10 is the mole fraction of ethyl acetate in the liquid phase on an IL-free basis, y1 is the mole fraction of ethyl acetate in the vapor phase, and a12 is the relative volatility of ethyl acetate to ethanol.

Standard uncertainties u are u(T) ¼ 0.05 K, u (x1) ¼ u (y1) ¼ 0.001.

3.2. Correlation of VLE data Table 2 and the comparison diagram is shown in Fig. 2. It can be seen that the experimental data are in good agreement with the literature data, indicating that the equipment and method used in

In ternary systems, the assumption of ideal vapor phase can be made due to the total pressure is low (101.3 kPa). Only volatile components are in the vapor phase at equilibrium state because of the nonvolatile nature of ILs. Thus, the activity coefficient of component i and the relative volatility of ethyl acetate (1) to ethanol (2), a12, can be calculated by equation (1) and equation (2), respectively.

gi ¼

Pyi Pi0 xi

a12 ¼

Fig. 2. x1 - y1 diagram for the binary system of ethyl acetate (1) þ ethanol (2) at 101.3 kPa: ( ), this work; (B), ref [17]; (△), ref [20]; solid line, correlated using NRTL model.



y1 =x1 g1 P10 ¼ y2 =x2 g2 P20

(1)

(2)

Where gi represents the activity coefficient of component i, xi represents the mole fraction of component i in the liquid phase, yi represents the mole fraction of component i in the vapor phase, Poi is the saturated vapor pressure of component i and can be calculated by the Antoine equation. In this work, the ternary VLE data were correlated with the most commonly used non-random two-phase (NRTL) model due to its

94

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

Table 3 Isobaric VLE data for the ternary system of ethyl acetate (1) þ ethanol (2) þ [EMIM][DEP] (3) at 101.3 kPa.a T/K

x1

x3

x10

y1

a12

T/K

x1

x3

x10

y1

a12

350.05 349.50 348.56 347.67 346.92 346.08 345.83 345.86 346.20 346.75 347.34 348.10 348.93 350.30 350.70 350.84 350.55 349.72 348.68 347.80 347.13 346.84 347.00

0.949 0.919 0.870 0.811 0.735 0.622 0.511 0.416 0.329 0.257 0.201 0.152 0.109 0.058 0.041 0.870 0.855 0.806 0.731 0.649 0.550 0.449 0.355

0.020 0.021 0.021 0.020 0.022 0.020 0.022 0.020 0.019 0.020 0.021 0.020 0.022 0.021 0.021 0.071 0.071 0.071 0.072 0.071 0.072 0.071 0.071

0.968 0.939 0.889 0.828 0.752 0.635 0.522 0.424 0.335 0.262 0.205 0.155 0.111 0.059 0.042 0.937 0.920 0.868 0.788 0.699 0.593 0.483 0.382

0.956 0.921 0.864 0.796 0.733 0.640 0.573 0.503 0.443 0.386 0.331 0.276 0.219 0.127 0.096 0.962 0.951 0.914 0.858 0.787 0.716 0.633 0.554

0.718 0.757 0.793 0.811 0.905 1.022 1.229 1.375 1.579 1.771 1.919 2.078 2.246 2.320 2.422 1.702 1.688 1.616 1.626 1.591 1.730 1.846 2.010

347.50 348.24 349.28 350.20 351.20 352.08 352.68 353.15 350.75 349.75 348.64 348.00 347.75 348.02 348.53 349.68 350.80 352.00 353.25 354.01 355.17 355.79

0.263 0.203 0.148 0.107 0.073 0.051 0.035 0.026 0.787 0.730 0.640 0.540 0.429 0.343 0.275 0.195 0.146 0.106 0.073 0.054 0.033 0.022

0.070 0.068 0.071 0.069 0.070 0.071 0.069 0.070 0.119 0.120 0.119 0.121 0.121 0.121 0.120 0.118 0.121 0.121 0.121 0.120 0.121 0.120

0.283 0.218 0.159 0.115 0.079 0.055 0.038 0.028 0.893 0.829 0.726 0.614 0.488 0.390 0.312 0.221 0.166 0.121 0.083 0.061 0.038 0.025

0.476 0.395 0.327 0.255 0.196 0.142 0.102 0.079 0.965 0.936 0.880 0.810 0.722 0.642 0.567 0.463 0.393 0.314 0.239 0.186 0.123 0.087

2.301 2.342 2.570 2.634 2.842 2.844 2.876 2.978 3.304 3.017 2.768 2.680 2.725 2.805 2.888 3.039 3.253 3.325 3.470 3.517 3.551 3.716

a

Standard uncertainties u are u(T) ¼ 0.05 K, u(P) ¼ 0.1 kPa, u (x3) ¼ u (x1) ¼ u (x10 ) ¼ u (y1) ¼ 0.001.

Table 4 Isobaric VLE data for the ternary system of ethyl acetate (1) þ ethanol (2) þ [BMIM][DEP] (3) at 101.3 kPa.a T/K

x1

x3

x10

y1

a12

T/K

x1

x3

x10

y1

a12

349.83 349.29 348.47 347.55 346.60 346.19 345.90 345.78 345.92 346.25 346.74 347.45 348.05 348.86 349.51 350.08 351.05 350.75 350.42 349.78 348.90 348.20 347.55 347.16 346.95

0.938 0.911 0.866 0.802 0.711 0.649 0.577 0.485 0.398 0.322 0.253 0.193 0.153 0.110 0.084 0.064 0.032 0.866 0.846 0.810 0.748 0.690 0.612 0.530 0.445

0.021 0.022 0.020 0.021 0.020 0.020 0.022 0.021 0.020 0.021 0.022 0.020 0.021 0.020 0.019 0.022 0.021 0.070 0.071 0.070 0.070 0.072 0.072 0.071 0.070

0.958 0.931 0.884 0.819 0.726 0.662 0.590 0.495 0.406 0.329 0.259 0.197 0.156 0.112 0.086 0.065 0.033 0.931 0.911 0.871 0.804 0.743 0.659 0.571 0.479

0.941 0.908 0.853 0.787 0.707 0.659 0.609 0.547 0.489 0.435 0.379 0.320 0.274 0.217 0.177 0.141 0.079 0.953 0.941 0.911 0.861 0.814 0.751 0.684 0.616

0.699 0.731 0.761 0.817 0.911 0.987 1.082 1.232 1.400 1.570 1.746 1.918 2.042 2.197 2.286 2.361 2.514 1.503 1.558 1.516 1.510 1.514 1.561 1.626 1.745

347.13 347.44 348.15 349.00 349.86 350.66 351.30 352.20 351.47 350.63 349.70 349.10 348.60 348.23 348.03 348.10 348.35 348.93 349.60 350.50 351.67 352.34 353.45 354.15 354.80

0.363 0.302 0.227 0.170 0.123 0.092 0.072 0.043 0.809 0.768 0.711 0.658 0.598 0.527 0.447 0.393 0.327 0.263 0.216 0.168 0.122 0.100 0.070 0.053 0.038

0.069 0.070 0.071 0.071 0.070 0.070 0.070 0.070 0.121 0.120 0.119 0.119 0.120 0.121 0.120 0.120 0.119 0.120 0.121 0.121 0.122 0.122 0.122 0.121 0.121

0.390 0.325 0.244 0.183 0.132 0.099 0.077 0.046 0.920 0.873 0.807 0.747 0.680 0.599 0.508 0.447 0.371 0.299 0.246 0.191 0.139 0.114 0.080 0.060 0.043

0.547 0.493 0.416 0.346 0.276 0.225 0.183 0.119 0.973 0.952 0.918 0.883 0.841 0.787 0.718 0.667 0.603 0.537 0.480 0.410 0.332 0.287 0.219 0.172 0.131

1.889 2.020 2.207 2.362 2.507 2.642 2.685 2.801 3.134 2.885 2.677 2.556 2.489 2.474 2.466 2.478 2.575 2.719 2.829 2.943 3.079 3.128 3.225 3.254 3.355

a

Standard uncertainties u are u(T) ¼ 0.05 K, u(P) ¼ 0.1 kPa, u (x3) ¼ u (x1) ¼ u (x10 ) ¼ u (y1) ¼ 0.001.

good applicability for the correlation of systems containing ILs [11,16,21]. For ternary system, the expression of the model is given as follows:

P3 lngi ¼

j¼1 tji Gji xj P3 k¼1 Gki xk

þ

P3

j¼1 tij xi Gij

 P3

k¼1 xk Gkj

with

3 X j¼1

#

xj Gij

P3

k¼1 xk Gki

ði ¼ 1; 2; 3Þ

"

tij (3)

  Dgij gij  gjj ¼ Gij ¼ exp  aij tij ; tij ¼ RT RT

ði ¼ 1; 2; 3Þ

(4)

Where aij is the non-randomness parameter of the ternary system, △gij is the interaction energy parameter between the two components i and j, R is the ideal gas constant and T is the VLE temperature. In the NRTL model, there are nine adjustable parameters (a12, a13, a23, △g12, △g21, △g23, △g32, △g13, △g31) for a ternary system. Among them, the binary interaction parameters of ethyl acetate (1) þ ethanol (2) (a12, △g12, △g21) were obtained from the

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

95

Table 5 Isobaric VLE data for the ternary system of ethyl acetate (1) þ ethanol (2) þ [BMIM][DBP] (3) at 101.3 kPa.a T/K

x1

x3

x10

y1

350.85 350.25 349.29 348.39 347.52 346.78 346.18 345.90 345.85 346.17 346.68 347.25 348.08 348.65 349.65 350.25 351.01 352.00 351.75 350.95 349.81 348.78 347.75 347.15 347.05 347.10

0.980 0.954 0.908 0.859 0.794 0.724 0.631 0.540 0.445 0.355 0.274 0.219 0.162 0.128 0.079 0.059 0.036 0.929 0.916 0.877 0.813 0.741 0.636 0.519 0.449 0.404

0.020 0.021 0.022 0.021 0.022 0.021 0.022 0.022 0.021 0.020 0.020 0.021 0.022 0.021 0.021 0.020 0.021 0.071 0.070 0.070 0.069 0.070 0.071 0.070 0.071 0.070

1.000 0.974 0.928 0.877 0.812 0.740 0.645 0.552 0.455 0.362 0.280 0.224 0.166 0.131 0.081 0.060 0.037 1.000 0.985 0.943 0.873 0.797 0.685 0.558 0.483 0.434

1.000 0.963 0.904 0.844 0.779 0.716 0.644 0.581 0.518 0.454 0.391 0.341 0.280 0.236 0.163 0.128 0.085 1.000 0.990 0.961 0.909 0.851 0.762 0.663 0.610 0.573

a

a12 0.695 0.731 0.758 0.817 0.886 0.996 1.125 1.286 1.466 1.654 1.795 1.955 2.054 2.209 2.307 2.408 1.508 1.489 1.453 1.455 1.472 1.558 1.674 1.750

T/K

x1

x3

x10

y1

a12

347.40 347.90 348.55 349.05 349.71 350.48 351.45 352.00 352.90 352.96 352.40 351.59 350.65 349.49 348.85 348.38 348.27 348.60 349.00 349.79 350.81 351.80 352.65 353.95 355.00

0.332 0.262 0.210 0.173 0.138 0.102 0.071 0.055 0.033 0.881 0.862 0.827 0.781 0.700 0.627 0.542 0.450 0.358 0.295 0.231 0.169 0.127 0.095 0.054 0.034

0.072 0.070 0.072 0.070 0.071 0.070 0.072 0.070 0.071 0.119 0.119 0.120 0.120 0.119 0.121 0.120 0.120 0.121 0.119 0.120 0.120 0.119 0.121 0.121 0.121

0.358 0.282 0.226 0.186 0.149 0.110 0.077 0.059 0.035 1.000 0.978 0.940 0.887 0.795 0.713 0.616 0.511 0.407 0.335 0.262 0.192 0.144 0.108 0.061 0.039

0.509 0.441 0.382 0.336 0.285 0.227 0.171 0.138 0.086 1.000 0.993 0.979 0.956 0.906 0.857 0.787 0.707 0.621 0.551 0.473 0.386 0.314 0.253 0.162 0.109

1.859 2.009 2.117 2.215 2.277 2.376 2.473 2.553 2.594 3.191 2.976 2.768 2.485 2.412 2.303 2.309 2.387 2.436 2.528 2.646 2.721 2.797 2.976 3.014

Standard uncertainties u are u(T) ¼ 0.05 K, u(P) ¼ 0.1 kPa, u (x3) ¼ u (x1) ¼ u (x10 ) ¼ u (y1) ¼ 0.001.

literature [16] and other six parameters were obtained from the Levenberg-Marquardt method by the minimization of the following objective function.

ARDð%Þ ¼

 exptl    gcalcd 1 X Xgi  i    100 exptl   n n g i

(5)

i

Where n represents the number of points in our experiment, gexptl i and gcalcd represent the experimental and calculated activity coi efficients of component i, respectively. The correlated nonrandomness factors and binary energy parameters for the NRTL model are listed in Table 6. Figs.3e5 are the x10 -y1 diagrams for the ethyl acetate (1) þ ethanol (2) þ IL (3) ternary system. It is obvious that the addition of ILs produces a significant salting-out effect on ethyl acetate, which makes the content of ethyl acetate in the vapor phase increase. The salting-out effect becomes stronger with increasing the contents of ILs. Furthermore, the azeotropic point moves towards the right with the addition of a small quantity of ILs. When the mole fraction of ionic liquids increased to 0.07, the azeotropic point completely disappears. Based on the NRTL model, it was calculated that the minimum mole fractions of [EMIM][DEP], [BMIM][DEP], and [BMIM][DBP] required to break the azetrope are

Fig. 3. Isobaric VLE diagram for the ternary system of ethyl acetate (1) þ ethanol (2) þ [EMIM][DEP] (3) at 101.3 kPa: dotted line, IL-frees system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using NRTL model.

Table 6 Estimated values of non-randomness factors, aij, and energy parameters Dgij and Dgji for NRTL model. Component i

Component j

aij

Dgij/(J/mol)

Dgji/(J/mol)

ARD %

Ethyl acetate Ethyl acetate Ethanol Ethyl acetate Ethanol Ethyl acetate Ethanol

Ethanol [EMIM][DEP] [EMIM][DEP] [BMIM][DEP] [BMIM][DEP] [BMIM][DBP] [BMIM][DBP]

0.300a 0.410 0.015 0.430 0.020 0.355 0.008

971.98a 4076.93 27906.26 3737.14 26736.65 4076.93 36526.52

1638.70a 18928.62 8766.36 21263.51 8766.12 7571.45 23669.18

1.46 1.52

a

Parameters obtained from Ref. [16].

0.84 1.38

Fig. 4. Isobaric VLE diagram for the ternary system of ethyl acetate (1) þ ethanol (2) þ [BMIM][DEP] (3) at 101.3 kPa: dotted line, IL-frees system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using NRTL model.

96

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

Fig. 5. Isobaric VLE diagram for the ternary system of ethyl acetate (1) þ ethanol (2) þ [BMIM][DBP] (3) at 101.3 kPa: dotted line, IL-frees system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using NRTL model.

Fig. 7. Relative volatility of ethyl acetate (1) to ethanol (2) with different mole fraction of [BMIM][DEP] at 101.3 kPa: dotted line, IL-free system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using the NRTL model.

0.039, 0.041 and 0.044, respectively, indicating that the separation performance of the three ILs follows the order of [EMIM] [DEP] > [BMIM][DEP] > [BMIM][DBP]. The relative volatility of ethyl acetate to ethanol (a12) increases with the addition of ILs as shown in Figs. 6e8. Moreover, the higher the content of ILs, the greater the relative volatility of ethyl acetate to ethanol. It can also be seen that [EMIM][DEP] produces the strongest salting-out effect on ethyl acetate giving the highest relative volatility, orderly followed by [BMIM][DEP] and [BMIM] [DBP]. This is in consistent with the results in Figs. 3e5. Therefore, [EMIM][DEP] is superior to the other two ILs. 3.3. Analysis of the s-profiles As reported in literature [24e28], hydrogen bond between ILs and solvents is discussed in this work to explain the different separation performance of the three ILs and to explore the structure-property relationship of ILs by means of the s-profiles of solvents and ILs. The s-profiles of ethyl acetate and ethanol are shown in Fig. 9, in which the two vertical dashed lines represent the cutoff values for the hydrogen bond. s < 0.0082 e/Å2 (negative region) and s > 0.0082 e/Å2 (positive region) represent that molecules have hydrogen bond donator ability and hydrogen bond acceptor ability, respectively [23], while 0.0082 e/Å2 < s < 0.0082 e/Å2 represent the non-polar region. It can be seen from Fig. 9 that ethyl acetate only has the hydrogen bond acceptor ability while ethanol has not only hydrogen bond acceptor ability but also

Fig. 8. Relative volatility of ethyl acetate (1) to ethanol (2) with different mole fraction of [BMIM][DBP] at 101.3 kPa: dotted line, IL-free system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using the NRTL model.

Fig. 9. s-Profiles for ethyl acetate and ethanol.

Fig. 6. Relative volatility of ethyl acetate (1) to ethanol (2) with different mole fraction of [EMIM][DEP] at 101.3 kPa: dotted line, IL-free system; -, x3 z 0.02; C, x3 z 0.07; :, x3 z 0.12; solid line, correlated using the NRTL model.

hydrogen bond donor ability. The s-profiles of the cations and anions of the three ILs are shown in Fig. 10. There are mainly three peaks at the positive region for the anions of the three ILs resulting from the electron lone-pairs of the oxygen atom and indicating their hydrogen bond acceptor ability. Some unresolved and low peaks can be seen at lower values than 0.0082 e/Å2 for the cations. They are related to the hydrogen atoms of the imidazolium ring [36] and indicate their hydrogen bond donor ability. Therefore, both the anions and cations can interact with ethanol forming hydrogen bond. In contrast, ethyl acetate can only form hydrogen

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

97

List of symbols xi xi’ yi T

Dgij ycalcd i P P0i

Fig. 10. s-Profiles for [BMIM]þ, [EMIM]þ, [DBP]- and [DEP]-.

bond with cations. Hence, the interaction between ethanol and ILs is stronger than that between ethyl acetate and ILs, making ethyl acetate more volatile compared to the IL-free system and improving the relative volatility of ethyl acetate to ethanol. When comparing [EMIM][DEP] with [BMIM][DEP], the difference comes from the alkyl chain length on imidazolium ring. It can be seen from Fig. 10 that the two cations have similar peaks in the negative region but [BMIM]þ cation has larger scale peaks in the non-polar region than [EMIM]þ cation owing to the longer alkyl chain of the former, which is probably responsible for the slight decrease in the relative volatility of ethyl acetate to ethanol of the system containing [BMIM][DEP] [37]. As for [BMIM][DEP] and [BMIM][DBP], the difference is from their anions. It is clear that the two anions have similar peaks in the positive region but [DBP]- has stronger peaks in the non-polar region due to its longer alkyl chain than [DEP]-, resulting in the smaller relative volatility of ethyl acetate to ethanol of the system with [BMIM][DBP] [37]. Hence, the separation performance of the three ILs follows the order: [EMIM] [DEP]> [BMIM][DEP]> [BMIM][DBP], showing good agreement with the result of VLE experiment.

4. Conclusions In this work, three ILs, namely [EMIM][DEP], [BMIM][DEP] and [BMIM][DBP], were employed as entrainers for the separation of ethyl acetate and ethanol mixture. The VLE data for the ternary systems of ethyl acetate þ ethanol þ IL were measured at 101.3 kPa. The results indicated that the addition of ILs could significantly increase the relative volatility of ethyl acetate to ethanol. Besides, the more ILs, the larger the relative volatility. All the three investigated ILs could break the azeotropic point of ethyl acetate and ethanol, and [EMIM][DEP] outperforms the other two ILs. The experimental VLE data were well fitted with the NRTL model, and the calculated minimum mole fractions of [EMIM][DEP], [BMIM] [DEP] and [BMIM][DBP] required to break the azetrope are 0.039, 0.041 and 0.044, respectively. By means of the s-profiles of solvents and ILs, it could be deduced that it is the strong hydrogen bonding between ILs and ethanol that contributes to the improvement in the relative volatility of ethyl acetate to ethanol.

Acknowledgments This work is financially supported by National Science Foundation of China (Project No. 21076126), National Science Foundation of China (Project No. 21576166), and Liaoning Province Science Foundation of China (Project No. 2014020140).

Mole fraction of solvent i in the liquid phase The liquid-phase mole fraction of component i excluding IL Mole fraction of solvent i in the vapor phase Equilibrium temperature Binary energy parameter of NRTL model Mole fraction of solvent i in the vapor phase calculated with the NRTL model Total pressure in the equilibrium system Saturated vapor pressure of component i at equilibrium temperature

Greek letters a12 Relative volatility of component 1 to component 2 aij Non-randomness parameter of NRTL model gi Activity coefficient of component i gexptl The activity coefficient of component i measured by i experimental data gcalcd The activity coefficient of component i calculated with the i NRTL model s surface screening charge density, e/Å2 References [1] E. Haslam, Recent developments in methods for the esterification and protection of the carboxyl group, Tetrahedron 36 (1980) 2409e2433, https:// doi.org/10.1016/0040-4020(80)80219-5. [2] E. Santacesaria, G. Carotenuto, R. Tesser, M. Di Serio, Ethanol dehydrogenation to ethyl acetate by using copper and copper chromite catalysts, Chem. Eng. J. 179 (2012) 209e220, https://doi.org/10.1016/j.cej.2011.10.043. [3] L. Berg, Separation of Ethyl Acetate from Ethanol by Azeotropic Distillation, 5993610, 1999. http://patents.justia.com/patent/5993610. [4] Z.G. Zhang, D.H. Huang, M. Lv, P. Jia, D.Z. Sun, W.X. Li, Entrainer selection for separating tetrahydrofuran/water azeotropic mixture by extractive distillation, Sep. Purif. Technol. 122 (2014) 73e77, https://doi.org/10.1016/ j.seppur.2013.10.051. [5] L. Zhang, J. Han, D. Deng, J. Ji, Selection of ionic liquids as entrainers for separation of water and 2-propanol, Fluid Phase Equilib. 255 (2007) 179e185, https://doi.org/10.1016/j.fluid.2007.04.016. [6] J. Fang, R. Zhao, W. Su, C. Li, J. Liu, B. Li, A molecular design method based on the COSMO-SAC model for solvent selection in ionic liquid extractive distillation, AIChE J. 62 (2016) 2853e2869, https://doi.org/10.1002/aic.15247. [7] M. Seiler, C. Jork, A. Kavarnou, W. Arlt, R. Hirsch, Separation of azeotropic mixtures using hyperbranched polymers or ionic liquids, AIChE J. 50 (2004) 2439e2454, https://doi.org/10.1002/aic.10249. [8] C. Jork, M. Seiler, Y.-A. Beste, W. Arlt, Influence of ionic liquids on the phase behavior of aqueous azeotropic systems, J. Chem. Eng. Data 49 (2004) 852e857, https://doi.org/10.1021/je034183r. mez, A. Domínguez, Study of the behaviour of the [9] N. Calvar, B. Gonz alez, E. Go azeotropic mixture ethanol-water with imidazolium-based ionic liquids, Fluid Phase Equilib. 259 (2007) 51e56, https://doi.org/10.1016/j.fluid.2007.03.021. [10] A.B. Pereiro, J.M.M. Araújo, J.M.S.S. Esperança, I.M. Marrucho, L.P.N. Rebelo, Ionic liquids in separations of azeotropic systems e a review, J. Chem. Thermodyn. 46 (2012) 2e28, https://doi.org/10.1016/j.jct.2011.05.026. [11] Z. Zhang, A. Hu, T. Zhang, Q. Zhang, M. Sun, D. Sun, W. Li, Separation of methyl acetateþmethanol azeotropic mixture using ionic liquids as entrainers, Fluid Phase Equilib. 401 (2015) 1e8, https://doi.org/10.1016/j.fluid.2015.04.018. [12] Z. Lei, C. Dai, J. Zhu, B. Chen, Extractive distillation with ionic liquids: a review, AIChE J. 60 (2014) 3312e3329, https://doi.org/10.1002/aic.14537. [13] Z. Zhang, H. Xu, Q. Zhang, A. Zhang, Y. Li, W. Li, Separation of methanol þ dimethyl carbonate azeotropic mixture using ionic liquids as entrainers, Fluid Phase Equilib. (2017), https://doi.org/10.1016/j.fluid.2016.11.026. [14] W. Li, N. Xu, H. Xu, A. Zhang, Z. Zhang, T. Zhang, Isobaric vaporeliquid equilibrium for ternary mixtures of acetone þ methanol þ ionic liquids at 101.3 kPa, Fluid Phase Equilib. 442 (2017) 20e27, https://doi.org/10.1016/ j.fluid.2017.03.014. s, P.J. Miguel, E. Vercher, A. Martínez-Andreu, Isobaric Vapor[15] A.V. Orchille Liquid equilibria for ethyl acetate þ ethanol þ 1-Ethyl-3methylimidazolium trifluoromethanesulfonate at 100 kPa, J. Chem. Eng. Data 52 (2007) 2325e2330, https://doi.org/10.1021/je700322p. [16] Q. Li, J. Zhang, Z. Lei, J. Zhu, J. Zhu, X. Huang, Selection of ionic liquids as entrainers for the separation of ethyl acetate and ethanol, Ind. Eng. Chem. Res. 48 (2009) 9006e9012, https://doi.org/10.1021/ie8017127. [17] Q. Li, J. Zhang, Z. Lei, J. Zhu, F. Xing, Isobaric VaporLiquid equilibrium for ethyl acetate þ ethanol þ 1-Ethyl-3-methylimidazolium tetrafluoroborate y,

98

Z. Zhang et al. / Fluid Phase Equilibria 454 (2017) 91e98

J. Chem. Eng. Data 54 (2009) 193e197, https://doi.org/10.1021/je800175s. [18] Z. JiuJuan, Z. JiGuo, L. QunSheng, X. FengYing, Isobaric vapor-liquid equilibria for ethyl acetate-ethanol-ionic liquid ternary systems, J. Beijing Univ. Chem. Technol. Nat. Sci.Ed.) 36 (15e19) (2009), https://doi.org/10.3969/j.issn.16714628.2009.03.004. [19] R. Li, X. Cui, Y. Zhang, T. Feng, J. Cai, Vapor-liquid equilibrium and liquid-liquid equilibrium of ethyl acetate þ ethanol þ 1-Ethyl-3-methylimidazolium acetate, J. Chem. Eng. Data 57 (2012) 911e917, https://doi.org/10.1021/ je200869q. [20] Y. Xu, T. Li, C. Peng, H. Liu, Influence of C2eH of imidazolium-based ionic liquids on the interaction and vaporeliquid equilibrium of ethyl acetate þ ethanol system: [Bmim]BF4 vs [Bmmim]BF4, Ind. Eng. Chem. Res. 54 (2015) 9038e9045, https://doi.org/10.1021/acs.iecr.5b01325. [21] A.E. Andreatta, M.P. Charnley, J.F. Brennecke, Using ionic liquids to break the ethanoleethyl acetate azeotrope, ACS Sustain. Chem. Eng. 3 (2015) 3435e3444, https://doi.org/10.1021/acssuschemeng.5b01175. [22] M. Diedenhofen, A. Klamt, COSMO-RS as a tool for property prediction of IL mixtures-A review, Fluid Phase Equilib. 294 (2010) 31e38, https://doi.org/ 10.1016/j.fluid.2010.02.002. [23] J. Li, X. Yang, K. Chen, Y. Zheng, C. Peng, H. Liu, Sifting ionic liquids as additives for separation of acetonitrile and water azeotropic mixture using the COSMORS method, Ind. Eng. Chem. Res. 51 (2012) 9376e9385, https://doi.org/ 10.1021/ie3000985. [24] A. Klamt, F. Eckert, COSMO-RS: a novel and efficient method for the a priori prediction of thermophysical data of liquids, Fluid Phase Equilib. 172 (2000) 43e72, https://doi.org/10.1016/S0378-3812(00)00357-5. [25] P. Verdía, E.J. Gonz alez, D. Moreno, J. Palomar, E. Tojo, Deepening of the role of cation substituents on the extractive ability of pyridinium ionic liquids of NCompounds from fuels, ACS Sustain. Chem. Eng. 5 (2017) 2015e2025, https:// doi.org/10.1021/acssuschemeng.6b02922. [26] M.R. Islam, C.-C. Chen, COSMO-SAC sigma profile generation with conceptual segment concept, Ind. Eng. Chem. Res. 54 (2015) 4441e4454, https://doi.org/ 10.1021/ie503829b. [27] E. Ruiz, V.R. Ferro, J. Palomar, J. Ortega, J.J. Rodriguez, Interactions of ionic liquids and acetone: thermodynamic properties, quantum-chemical calculations, and NMR analysis, J. Phys. Chem. B 117 (2013) 7388e7398, https:// doi.org/10.1021/jp402331y.

[28] Z. Zhang, A. Zhang, K. Wu, Q. Zhang, A. Hu, W. Li, Separation of ethyl acetate and 2-propanol azeotropic mixture using ionic liquids as entrainers, Fluid Phase Equilib. 429 (2016) 331e338, https://doi.org/10.1016/ j.fluid.2016.08.011. [29] H. Renon, J.M. Prausnitz, Local compositions in thermodynamic excess functions for liquid mixtures, AIChE J. 14 (1968) 135e144, https://doi.org/10.1002/ aic.690140124. [30] Y. Zhou, A.J. Robertson, J.H. Hillhouse, D. Baumann, Phosphonium and Imidazolium Salts and Methods of Their Preparation 7638636, 2009. http://www. freepatentsonline.com/7638636.html. [31] W. Hunsmann, Vaporization equilibria of formic acid/acetate acid and carbon tetrachloride/perchloroethylene mixtures, Chem. Ing. Tech. 39 (1967) 1142e1145, https://doi.org/10.1002/cite.330391910. [32] E. Mullins, Y. a Liu, R. Oldland, S. Wang, S.I. Sandler, C. Chen, M. Zwolak, K. Seavey, VT-2005 Sigma Profile Database: a Detailed Tutorial for Generating Sigma Profiles Using Accelrys' Materials Studio v3.2 Software Package, Presentation, 2005. [33] E. Mullins, R. Oldland, Y.A. Liu, S. Wang, S.I. Sandler, C.C. Chen, M. Zwolak, K.C. Seavey, Sigma-profile database for using COSMO-based thermodynamic methods, Ind. Eng. Chem. Res. 45 (2006) 4389e4415, https://doi.org/10.1021/ ie060370h. [34] E. Mullins, Y.A. Liu, A. Ghaderi, S.D. Fast, Sigma profile database for predicting solid solubility in pure and mixed solvent mixtures for organic pharmacological compounds with COSMO-based thermodynamic methods, Ind. Eng. Chem. Res. 47 (2008) 1707e1725, https://doi.org/10.1021/ie0711022. [35] L. Yang, S.I. Sandler, C. Peng, H. Liu, Y. Hu, Prediction of the phase behavior of ionic liquid solutions, Ind. Eng. Chem. Res. 49 (2010) 12596e12604, https:// doi.org/10.1021/ie1013647. [36] J. Palomar, J.S. Torrecilla, V.R. Ferro, F. Rodriguez, Development of an a priori ionic liquid design tool. 1. Integration of a novel COSMO-RS molecular descriptor on neural networks, Ind. Eng. Chem. Res. 47 (2008) 4523e4532, https://doi.org/10.1021/ie800056q. [37] M. Taha, Designing new mass-separating agents based on piperazinecontaining goods buffers for separation of propanols and water azeotropic mixtures using COSMO-RS method, Fluid Phase Equilib. 425 (2016) 40e46, https://doi.org/10.1016/j.fluid.2016.05.011.

́