Vapor pressure measurement for water, methanol, ethanol, and their binary mixtures in the presence of an ionic liquid 1-ethyl-3-methylimidazolium dimethylphosphate

Vapor pressure measurement for water, methanol, ethanol, and their binary mixtures in the presence of an ionic liquid 1-ethyl-3-methylimidazolium dimethylphosphate

Fluid Phase Equilibria 255 (2007) 186–192 Vapor pressure measurement for water, methanol, ethanol, and their binary mixtures in the presence of an io...

266KB Sizes 9 Downloads 283 Views

Fluid Phase Equilibria 255 (2007) 186–192

Vapor pressure measurement for water, methanol, ethanol, and their binary mixtures in the presence of an ionic liquid 1-ethyl-3-methylimidazolium dimethylphosphate Jun-Feng Wang a,b , Chun-Xi Li a,b,∗ , Zi-Hao Wang b , Zi-Jia Li b , Yan-Bin Jiang b a

State Key Lab. of Chem. Resource Eng., Beijing University of Chem. Tech., Beijing 100029, PR China b College of Chem. Eng., Beijing University of Chem. Tech., Beijing 100029, PR China Received 29 January 2007; received in revised form 2 April 2007; accepted 8 April 2007 Available online 13 April 2007

Abstract Vapor pressure data were measured for water, methanol and ethanol as well as their binary mixtures with an ionic liquid (IL) 1-ethyl-3methylimidazolium dimethylphosphate ([EMIM][DMP]) at varying temperature and IL-content ranging from mass fraction of 0.10–0.70 by a quasi-static method. The vapor pressure data for the IL-containing binary systems were correlated using NRTL equation with average absolute relative deviation (ARD) within 0.0076, and the binary NRTL parameters was used for predicting the vapor pressure of the IL-containing ternary systems with reasonable accuracy. In addition, the infinite activity coefficients of solvents in [EMIM][DMP] and isobaric vapor–liquid equilibrium for IL-containing ternary systems at 101.325 kPa and mass fraction of IL being 0.5 were predicted with the regressed NRTL parameters. The results indicate that ionic liquid [EMIM][DMP] can depress the volatility of the solvents of water, methanol and ethanol but to a varying degree, leading to the variation of relative volatility of a solvent and even removal of azeotrope for water–ethanol mixture. © 2007 Elsevier B.V. All rights reserved. Keywords: Vapor pressure; Measurement; Ionic liquid; NRTL equation; Vapor–liquid equilibrium

1. Introduction Vapor pressure data for ionic liquid (IL) containing systems is of great importance in assessing intermolecular interactions between IL and solvent, screening appropriate entrainer for special distillation as well as developing thermodynamic models specific to IL-containing systems. Among various ILs reported in the literature [1], the ILs with alkyl-substituted imidazolium cation and dialkylphosphates is of potential applications and worthy of development for their good stability and hydrophilicity, less toxicity and corrosiveness and more importantly their ease of production and low cost for industrial applications. In our previous work [2–4], vapor pressure of water, methanol, ethanol and their binary mixtures with an alkyl-substituted imidazolium dialkylphosphate IL have been measured, for ∗ Corresponding author at: College of Chem. Eng., Beijing University of Chem. Tech., Beijing 100029, PR China. Tel.: +86 10 64410308; fax: +86 10 64410308. E-mail address: [email protected] (C.-X. Li).

0378-3812/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.fluid.2007.04.010

example, 1-methyl-3-methylimidazolium dimethylphosphate ([MMIM][DMP]), 1-ethyl-3-methylimidazolium diethylphosphate ([EMIM][DEP]), 1-butyl-3-methylimidazolium dibutylphosphate ([BMIM][DBP]), and 1-ethyl-3-ethylimidazolium diethylphosphate ([EEIM][DEP]). The objective of this series of work is to find an efficient and feasible entrainer for the separation of water methanol and ethanol mixture via extractive distillation, and have an insight into the effect of alkyl substitutes in an IL on the interaction between IL and specific solvent and accordingly the relative volatility of the solvent. Toward this end, vapor pressure data of water, methanol and ethanol as well as their binary mixtures with a new IL 1-ethyl-3-methylimidazolium dimethylphosphate ([EMIM][DMP]) were measured using quasi-static ebulliometer method. The experimental data of vapor pressure for ILcontaining binary systems were correlated using non-electrolyte NRTL equation, and interaction between IL and solvent molecules were discussed in terms of infinite activity coefficients of solvent in [EMIM][DMP] derived from the regressed NRTL parameters.

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192

187

2. Experimental

Table 1 Vapor pressure data of binary system water (1)–[EMIM][DMP] (2)

2.1. Materials

T (K)

Pexp (kPa)

PNRTL (kPa)

γi

γiNRTL

x1 = 0.9915 323.76 331.49 337.42 341.67 346.27 349.73 354.06 357.98

12.609 18.256 23.924 28.972 35.221 40.683 48.559 56.785

12.546 18.197 23.884 28.844 35.158 40.649 48.527 56.740

1.0026 1.0010 0.9996 1.0024 0.9998 0.9990 0.9989 0.9990

0.9976 0.9978 0.9979 0.9980 0.9981 0.9981 0.9982 0.9983

x1 = 0.9683 329.14 338.86 342.95 345.87 349.44 353.59 356.92 360.27

15.383 24.089 28.840 32.706 38.006 45.046 51.484 58.781

15.487 24.286 29.100 33.007 38.373 45.522 52.054 59.395

0.9651 0.9662 0.9662 0.9667 0.9670 0.9669 0.9670 0.9682

0.9716 0.9740 0.9749 0.9756 0.9763 0.9771 0.9777 0.9783

x1 = 0.9291 331.82 337.61 342.89 347.54 351.65 356.18 360.46 363.71

15.082 19.715 24.888 30.483 36.194 43.556 51.763 58.804

15.387 20.172 25.584 31.340 37.294 44.924 53.323 60.545

0.8689 0.8714 0.8716 0.8750 0.8760 0.8783 0.8822 0.8847

0.8864 0.8916 0.8960 0.8996 0.9027 0.9059 0.9088 0.9109

x1 = 0.8489 338.84 343.48 347.73 351.54 356.93 363.48 367.26

15.216 18.825 22.775 26.726 33.494 43.556 50.463

14.739 18.301 22.180 26.235 33.016 43.201 50.207

0.6968 0.7033 0.7100 0.7112 0.7175 0.7237 0.7273

0.6750 0.6837 0.6914 0.6981 0.7072 0.7178 0.7236

The chemical reagents used in this study were ethanol, methanol, redistilled water and IL [EMIM][DMP]. AR grade methanol and ethanol with purity of 0.997 were purchased from Beijing Red Star Reagents Company, China. The purity of reagents was checked by gas chromatogram (GC2010, Japan). The IL used was prepared and purified in the laboratory according to the literature method [5,6], and the purity was more than 0.98 in terms of NMR analysis. Furthermore, the IL was treated before use by vacuum evaporation to remove the residual volatile impurities, and the water content was within 500 ppm as measured by Karl–Fischer method (CBS-1A). [EMIM][DMP] is insensitive to air and water, and mutual soluble to water with thermal decomposition temperature being expected ca. 275 ◦ C suggested by the decomposition temperature of 275.0 and 278.7 ◦ C for [BMIM][DBP] and [MMIM][DMP], respectively. 2.2. Apparatus and procedure The details of the experimental apparatus and the operation procedure was described elsewhere [2]. The apparatus was composed of two ebulliometers in series with a buffer, viz. a working ebulliometer filled with liquid mixture and a reference one filled with pure water. The equilibrium pressure of the system was determined by the boiling temperature of pure water in the reference ebulliometer in terms of the temperature–pressure relation represented by Antoine equation [7]. The equilibrium temperature of the ebulliometers were measured by two four-wire 25- calibrated platinum resistance thermometers (type CST6601) with an uncertainty of 0.02 K, connected to a two-channel standard digital thermometer (CST6502). The uncertainty of the vapor pressure arising from the uncertainty of temperature measurement is estimated within ±0.04 kPa, and the vapor pressure reproducibility for a replicate sample is within ±0.07 kPa. The condensers were cooled with chilling glycol aqueous solution at 275 K to minimize the vapor phase loss during the measurement and hence the concentration variation of the solution. The uncertainty of the mole fraction in the liquid phase is estimated within 0.002. 3. Results and discussion The vapor pressure data for water, methanol and ethanol as well as their binary mixtures, i.e. water–methanol, water–ethanol and ethanol–methanol in the presence of IL [EMIM][DMP] with IL mass fraction ranging from 0.10 to 0.70 (mole fraction from 0.0085 to 0.3119) were measured and listed in Tables 1–6, respectively. The effect of IL on the nonideality of a solution can be expressed by activity coefficient of component i, γ i , which can be calculated by the phase equilibrium equation [8]: γi =

yi φˆ i P xi φis Pis

(1)

n  NRTL exp  exp P −P /P i=1 ARD(P) = 0.014 = , n  n 2 exp cal RMSD = 0.018 =

i=1

/P

(P

n

−1)

exp

.

where P and Pis are vapor pressure of liquid mixture and pure component i at system temperature, respectively, and the latter can be calculated by Antoine equation with Antoine constants taken from literature [7] and were listed in Table 7. yi and xi represent mole fraction of component i in the vapor phase and liquid phase, respectively. φˆ i is the fugacity coefficient of component i in the vapor mixture, and φis is the fugacity coefficient of pure component i in its saturated state. For an IL-containing binary system, i.e. solvent (1)–IL (2), the vapor phase is fully composed of solvent vapor due to the negligible volatility of IL [9], and thus y1 = 1. Since the vapor phase composition for such binary system and for the pure solvent is same, and the pressure difference between them is relatively small, the fugacity coefficient correction can be cancelled out. Therefore, Eq. (1) is reduced to Eq. (2): γ1 =

P P1s x1

(2)

188

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192

Table 2 Vapor pressure data of binary system methanol (1)–[EMIM][DMP] (2) exp

Table 3 Vapor pressure data of binary system ethanol (1)–[EMIM][DMP] (2) exp

T (K)

Pexp (kPa)

PNRTL (kPa)

γi

γiNRTL

T (K)

Pexp (kPa)

PNRTL (kPa)

γi

γiNRTL

x1 = 0.9851 299.20 303.04 306.80 310.45 313.84 317.74 321.12 324.50

17.460 21.243 25.597 30.556 35.819 42.854 49.870 57.859

17.474 21.250 25.595 30.520 35.795 42.797 49.763 57.653

0.9910 0.9919 0.9926 0.9941 0.9939 0.9948 0.9959 0.9976

0.9918 0.9922 0.9926 0.9929 0.9932 0.9935 0.9938 0.9940

x1 = 0.9787 309.95 314.00 320.12 323.38 328.35 330.72 334.42 339.27

14.808 18.301 24.953 29.189 36.872 41.146 48.576 60.073

14.801 18.291 24.881 29.144 36.859 41.100 48.535 59.980

0.9991 0.9992 1.0017 1.0004 0.9993 1.0001 0.9999 1.0006

0.9986 0.9987 0.9988 0.9988 0.9989 0.9990 0.9990 0.9991

x1 = 0.9451 300.94 303.56 306.15 309.90 315.14 319.87 323.49 326.99

17.001 19.448 22.184 26.684 34.161 42.447 49.844 58.025

17.013 19.462 22.175 26.662 34.212 42.495 49.939 58.137

0.9198 0.9221 0.9253 0.9287 0.9304 0.9340 0.9356 0.9377

0.9205 0.9228 0.9249 0.9279 0.9318 0.9350 0.9374 0.9395

x1 = 0.9229 310.50 316.95 320.67 325.84 330.66 333.90 337.47 340.98

14.165 19.761 23.785 30.469 38.080 44.095 51.571 60.003

14.127 19.727 23.751 30.469 38.133 44.163 51.689 60.148

0.9845 0.9850 0.9855 0.9851 0.9846 0.9850 0.9848 0.9852

0.9819 0.9833 0.9841 0.9851 0.9859 0.9865 0.9870 0.9876

x1 = 0.8806 304.51 309.18 313.42 317.39 321.86 326.01 329.69 333.06

15.969 20.307 25.144 30.280 37.098 44.681 52.477 60.419

15.905 20.192 24.898 30.123 37.080 44.719 52.545 60.679

0.7750 0.7851 0.7959 0.7990 0.8024 0.8078 0.8128 0.8151

0.7719 0.7806 0.7881 0.7948 0.8020 0.8084 0.8139 0.8187

x1 = 0.8368 313.46 319.68 325.34 329.31 333.55 337.28 340.29 343.58

14.187 19.458 25.652 30.948 37.546 44.529 50.755 58.368

14.138 19.453 25.680 30.988 37.667 44.504 50.747 58.415

0.9319 0.9337 0.9364 0.9388 0.9396 0.9454 0.9468 0.9476

0.9287 0.9334 0.9374 0.9400 0.9427 0.9449 0.9466 0.9484

x1 = 0.7597 314.99 318.98 324.41 327.59 332.42 336.61 340.40 343.86

16.761 20.289 26.098 30.280 37.434 44.663 52.360 60.419

16.776 20.412 26.423 30.601 37.996 45.592 53.496 61.707

0.5718 0.5777 0.5858 0.5935 0.6008 0.6056 0.6123 0.6190

0.5723 0.5813 0.5931 0.5998 0.6098 0.6182 0.6256 0.6321

x1 = 0.6881 320.08 323.71 332.30 334.95 340.75 344.88 347.49 350.67

14.187 17.108 25.652 29.013 37.546 45.188 50.755 58.368

14.176 17.032 25.775 29.148 37.831 45.259 50.559 57.714

0.8118 0.8206 0.8258 0.8297 0.8351 0.8454 0.8533 0.8635

0.8112 0.8170 0.8298 0.8335 0.8414 0.8467 0.8500 0.8539

ARD(P) = 0.0050; RMSD = 0.0081.

ARD(P) = 0.0023; RMSD = 0.0032.

According to Eq. (2), the experimental activity coefficient of the solvent in an IL-containing binary liquid mixture can be calculated from the vapor pressure data, which were listed in Tables 1–3, respectively. In order to correlate the experimental vapor pressure data, the NRTL equation for non-electrolyte solution is used to describe the activity coefficient of solvent as rigorous model specific for IL-containing systems has been not available. The NRTL parameters for three binary systems, i.e. αij and (gij − gjj ) as listed in Table 8, were obtained by fitting the experimental vapor pressure data in the whole temperature and composition range using a least square method. As shown in Tables 1–3, the experimental vapor pressure can be well correlated by NRTL equation with average absolute relative deviation (ARD) less than 0.0076 for the three binary systems studied. Based on the fitted NRTL parameters, infinite dilution activity coefficients of solvent in [EMIM][DMP] at varying temperatures were calculated, as shown in Table 9.

The representative variation trend of vapor pressure versus temperature at different IL-content was shown in Fig. 1 for methanol–[EMIM][DMP] binary system. It is seen that the log(P/kPa) against 1/(t + C) for a given concentration is linear over the pressure and temperature range studied, which is similar to the vapor pressure behavior of the pure solvent and C is the corresponding Antoine constant for the corresponding pure solvent. It is apparent that ionic liquid always decreases the vapor pressure of solvent due to its diluting effect and affinity to the solvent, and the higher the IL-content, the lower the solvent vapor pressure. Moreover, the vapor pressure of solvent shows a negative deviation from the Rault’s law as the activity coefficient of solvent nearly always smaller than unity, as shown in Tables 1–3. The vapor pressure for ternary systems water–ethanol– [EMIM][DMP], water–methanol–[EMIM][DMP] and ethanol– methanol–[EMIM][DMP] at varying liquid composition and temperature was predicted using the binary NRTL parameters listed in Table 7 and compared with the experimental

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192 Table 4 The experimental and predictive vapor pressure and activity coefficient of solvents for ternary system water (1)–methanol (2)–[EMIM][DMP] (3)

Table 5 The experimental and predictive vapor pressure data and activity coefficient of solvents for ternary system water (1)–ethanol (2)–[EMIM][DMP] (3) PNRTL (kPa)

γ1NRTL

γ2NRTL

x1 = 0.8782, x2 = 0.0859 317.84 15.234 323.20 19.936 328.15 25.172 333.29 31.397 337.69 38.182 341.56 45.076 345.30 52.965 349.11 62.117

15.475 20.084 25.325 31.954 38.737 45.673 53.323 62.203

0.9811 0.9826 0.9839 0.9852 0.9862 0.9870 0.9878 0.9885

3.7733 3.7159 3.6644 3.6125 3.5693 3.5323 3.4973 3.4624

1.1002 1.0999 1.0995 1.0992 1.0987 1.0983 1.0978 1.0972

x1 = 0.7601, x2 = 0.1985 315.36 15.930 320.71 20.822 324.70 25.264 328.62 30.386 333.15 37.291 337.18 44.559 340.87 52.241 344.34 60.504

16.455 21.459 25.973 31.168 38.262 45.655 53.448 61.745

1.0205 1.0224 1.0237 1.0249 1.0261 1.0271 1.0279 1.0287

2.5226 2.4983 2.4806 2.4635 2.4441 2.4273 2.4121 2.3981

0.9708 0.9739 0.9769 0.9790 0.9806 0.9821 0.9834 0.9845

x1 = 0.5993 x2 = 0.3518 316.63 18.392 319.92 21.652 324.54 27.061 329.30 33.784 333.10 40.192 336.87 47.425 340.27 54.897 343.81 63.675

19.037 22.520 28.144 35.138 41.750 49.296 57.048 66.199

1.0834 1.1077 1.1100 1.1120 1.1135 1.1148 1.1159 1.1169

1.7673 1.7507 1.7417 1.7328 1.7258 1.7190 1.7129 1.7068

x1 = 0.2062, x2 = 0.7263 316.09 19.076 318.88 21.888 322.27 25.687 325.40 29.809 330.02 36.960 334.39 44.991 337.83 52.303 340.89 59.676

19.299 22.203 26.228 30.471 37.804 46.039 53.557 61.104

1.0761 1.0842 1.0937 1.1021 1.1141 1.1248 1.1329 1.1398

1.1502 1.1492 1.1479 1.1467 1.1450 1.1434 1.1422 1.1411

PNRTL (kPa)

γ1NRTL

γ2NRTL

T (K)

x1 = 0.6960, x2 = 0.2685 310.90 15.628 317.91 21.984 321.90 26.594 326.37 32.585 329.48 37.406 333.48 44.506 336.79 51.100 340.04 58.425

14.315 20.052 24.132 29.508 33.817 40.143 46.095 52.663

0.9434 0.9470 0.9489 0.9509 0.9522 0.9538 0.9551 0.9563

1.1721 1.1694 1.1678 1.1660 1.1648 1.1632 1.1618 1.1606

x1 = 0.5196, x2 = 0.4380 308.83 17.519 312.59 21.001 317.11 26.274 320.53 30.774 324.92 37.529 328.56 43.882 331.92 50.661 336.47 61.268

16.693 20.038 24.813 29.026 35.316 41.389 47.735 57.627

0.9161 0.9191 0.9226 0.9251 0.9281 0.9305 0.9326 0.9354

x1 = 0.1569, x2 = 0.7920 302.15 16.221 307.60 21.319 313.64 28.527 318.17 35.166 321.81 41.387 325.54 48.635 328.78 55.795 331.88 63.700

16.530 21.744 29.085 35.906 42.299 49.835 57.269 65.225

0.8672 0.8752 0.8835 0.8893 0.8937 0.8980 0.9016 0.9049

T (K)

Pexp (kPa)

189

ARD = 0.057; RMSD = 0.064.

values as listed in Tables 4–6. It is seen that the agreement between the experimental and the predicted values is fairly good with average deviation ARD of 0.0295 and the maximum deviation of −0.0986. By comparing the results as listed in Tables 4–6, it is seen that the highest prediction deviation was found for water–methanol–IL system, which was likely

Pexp (kPa)

ARD = 0.024; RMSD = 0.027.

Fig. 1. The experimental and correlative vapor pressure data of binary system methanol (1)–[EMIM][DMP] (2) at different mass fraction of [EMIM][DMP]. Legend: (- - -) pure methanol; (—) calculated by NRTL equation; symbols are experimental data at different mass fraction of [EMIM][DMP]: () 0.10; () 0.30; (䊉) 0.50; () 0.70.

related to the difference of ionization degree of IL in pure solvents, viz. water and methanol, and their mixtures, and accordingly the variation of binary NRTL parameters from pure solvent to solvent mixtures. Considering the fact that the ionization degree of IL in different solvents followed the order water > water + methanol > methanol, and the IL component was treated as a neutral molecule in the NRTL model, it can be inferred that the water–IL interaction and methanol–IL interaction were overestimated and underestimated respectively in the ternary system, leading to a negative and positive deviation respectively for the prediction of partial pressure of water and methanol component. Therefore, in water-rich concentration range, negative deviation of the predicted pressure prevails and vice versa. However, from the point of view of practical application, the conventional NRTL model for non-electrolyte solution is applicable for representing the vapor–liquid equilibrium of IL-containing multi-component systems, as indicated by Doker and Gmehling and Shi et al. [10,11].

190

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192

Table 6 The experimental and predictive vapor pressure data and activity coefficient of solvents for ternary system ethanol (1)–methanol (2)–[EMIM][DMP] (3) T (K)

Pexp (kPa)

γ1NRTL

PNRTL (kPa)

γ2NRTL

x1 = 0.6831, x2 = 0.2455 309.93 16.265 312.42 18.516 316.61 22.840 320.34 27.380 324.33 33.033 328.62 40.075 332.60 47.781 336.84 57.340

16.117 18.336 22.638 27.166 32.854 40.045 47.861 58.131

1.0367 1.0370 1.0376 1.0380 1.0384 1.0387 1.0390 1.0392

0.7237 0.7279 0.7347 0.7406 0.7467 0.7530 0.7586 0.7644

x1 = 0.4767, x2 = 0.4569 308.08 17.277 311.55 20.569 315.85 25.394 320.48 31.624 325.16 39.143 329.39 47.135 333.10 55.259

17.267 20.606 25.487 31.806 39.502 47.747 56.134

1.0488 1.0495 1.0503 1.0510 1.0517 1.0522 1.0526

0.8185 0.8230 0.8285 0.8340 0.8393 0.8439 0.8477

x1 = 0.2974, x2 = 0.6405 303.95 16.079 305.80 17.729 309.20 21.053 313.82 26.381 318.70 33.153 323.60 41.204 327.77 49.410 331.23 57.252

16.067 17.679 21.009 26.378 33.266 41.662 50.153 58.255

1.0471 1.0477 1.0487 1.0500 1.0512 1.0522 1.0530 1.0536

0.8687 0.8708 0.8745 0.8793 0.8840 0.8885 0.8921 0.8949

x1 = 0.0717, x2 = 0.8720 303.51 18.507 305.22 20.237 309.17 24.599 313.75 30.657 318.50 37.970 323.25 46.729 327.31 55.658 331.07 65.190

18.560 20.241 24.640 30.726 38.336 47.470 56.686 66.501

1.0366 1.0374 1.0391 1.0410 1.0427 1.0443 1.0456 1.0467

9.1539 9.1688 9.2016 9.2374 9.2721 9.3046 9.3306 9.3535

ARD = 0.0076; RMSD = 0.0094.

Table 7 Antoine vapor pressure constants of pure compounds Component

Ethanol Methanol Water

Antoine constants A

B

C

8.1122 8.08097 8.07131

1592.864 1582.271 1730.63

226.184 239.726 233.426

Antoine equation log psat = A − B/(t + C), where psat is in Torr and t is the temperature in ◦ C.

Table 8 The NRTL parameters fitted for IL-containing binary systems and taken from literature [7] for the vapor pressure prediction of the IL-containing ternary systems System

α

g12 − g22 (J mol−1 )

g21 − g11 (J mol−1 )

Water–[EIM][DMP] Ethanol–[EMIM][DMP] Methanol–[EMIM][DMP] Ethanol–water [7] Ethanol–methanol [7] Methanol–water [7]

0.3594 0.6319 0.9610 0.3008 0.3053 0.3013

12.560 5277.6 34.770 −510.82 1580.2 −172.12

−8824.4 −3925.0 −4162.2 5612.1 −1292.9 768.56

Fig. 2. Isobaric VLE diagram for water (1)–ethanol (2)–[EMIM][DMP] (3) ternary systems at 101.325 kPa. Legend: (- - -) IL-free mixture of water and ethanol; () water–ethanol mixture at mass fraction of [EMIM][DMP] of 0.50.

In order to show the salt effect of [EMIM][DMP] on the distillation separation of three binary mixtures, water–ethanol, water–methanol and ethanol–methanol, isobaric vapor–liquid equilibrium for ternary mixtures of solvent (1) + solvent (2) + IL (3) at fixed mass fraction of IL of 0.5 were predicted using Eq. (3): ps xi γi yi =  i psj xj γj

(3)

j

The results were plotted in Figs. 2–4, respectively, on a salt free basis for liquid composition, and compared with the VLE curves in the absence of IL. The results suggest that the affinity between ionic liquid and the solvents basically follows

Table 9 Infinite dilution activity coefficient of solvent γ1∞ (cal) at varying temperature for binary mixtures solvent (1)–[EMIM][DMP] (2) Solvent

Water Methanol Ethanol

γ1∞ T = 310 K

T = 320 K

T = 330 K

T = 340 K

T = 350 K

T = 360 K

0.0327 0.2016 0.3823

0.0364 0.2119 0.4030

0.0403 0.2221 0.4231

0.0443 0.2322 0.4428

0.0484 0.2421 0.4620

0.0652 0.2716 0.5199

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192

191

sure of water, methanol and ethanol but to different extents due to the affinity difference between [EMIM][DMP] and different solvents. The vapor pressure data of binary systems can be well correlated with NRTL equation, and the NRTL parameters obtained can be applied for the prediction of vapor pressure of multi-component systems with fair accuracy. The affinity between ionic liquid and solvents follows the order [EMIM][DMP]–water > [EMIM][DMP]–methanol > [EMIM][DMP]–ethanol, and hence the relative volatility of ethanol in water–ethanol and methanol–ethanol binary mixtures is enhanced and even the azeotrope in the first mixture is destroyed.

Fig. 3. Isobaric VLE diagram for water (1)–methanol (2)–[EMIM][DMP] (3) ternary systems at 101.325 kPa. Legend: (- - -) IL-free mixture of water and methanol; () water–methanol mixture at mass fraction of [EMIM][DMP] of 0.50.

Fig. 4. Isobaric VLE diagram for ethanol (1)–methanol (2)–[EMIM][DMP] (3) ternary systems at 101.325 kPa. Legend: (- - -) IL-free mixture of ethanol and methanol; () ethanol–methanol mixture at mass fraction of [EMIM][DMP] of 0.50.

the order [EMIM][DMP]–water > [EMIM][DMP]–methanol > [EMIM][DMP]–ethanol, as a result the relative volatility of ethanol in water–ethanol and methanol–ethanol binary mixtures is enhanced and even the azeotrope in the first mixture is disappeared, see Figs. 2 and 4. In contrast, a complicated salt effect was observed for the VLE of water–methanol system manifested by a salting-in effect for methanol in the water-rich region and a salting-out effect for methanol in the water-lean region. 4. Conclusions Vapor pressure data for three binary and three ternary ILcontaining systems at varying temperature and IL-content were measured using a quasi-static method. The results indicate that ionic liquid [EMIM][DMP] can reduce the vapor pres-

List of symbols gij − gjj binary interaction parameters of the NRTL equation n number of data points P vapor pressure (kPa) Pcal the calculated vapor pressure (kPa) Pexp experimental vapor pressure (kPa) PNRTL vapor pressure calculated by NRTL equation (kPa) vapor pressure of pure component i at system temperPis ature (kPa) t temperature (◦ C) T temperature (K) xi mole fraction of component i in liquid phase xi mole fraction of component i in liquid phase on a salt free basis yi mole fraction of component i in vapor phase Greek letters non-random parameter of the NRTL equation αij the activity coefficient of component i γi exp γi the activity coefficient of component i determined by experimental vapor pressure data γiNRTL the activity coefficient of component i calculated with the NRTL equation φˆ i the fugacity coefficient of component i in the vapor mixture φis the fugacity coefficient of pure component i in its saturated state Acknowledgement The authors are grateful to the financial support from National Natural Science Foundation of China (No. 20376004) that allows the authors to accomplish the research presented herein. References [1] S.J. Zhang, X.M. L¨u, et al., Ionic Liquids: From Fundamentals to Applications (in Chinese), Science Press, 2006. [2] J. Zhao, X.-C. Jiang, C.-X. Li, Z.-H. Wang, Vapor pressure measurement for binary and ternary systems containing a phosphoric ionic liquid, Fluid Phase Equilib. 247 (2006) 190–198. [3] J. Zhao, C.-X. Li, Z.-H. Wang, Vapor pressure measurement and prediction for ethanol + methanol and ethanol + water systems containing ionic liquids, J. Chem. Eng. Data 51 (2006) 1755–1760. [4] X.-C. Jiang, J.-F. Wang, C.-X. Li, L.-M. Wang, Z.-H. Wang, Vapour pressure measurement for binary and ternary systems con-

192

J.-F. Wang et al. / Fluid Phase Equilibria 255 (2007) 186–192

taining water methanol ethanol and an ionic liquid 1-ethyl-3ethylimidazolium diethylphosphate, J. Chem. Thermodyn. 39 (2007) 841– 846. [5] Y.H. Zhou, A.J. Robertson, J.H. Hillhouse. Phosphonium and imidazolium salts and methods of their preparation. Patent WO2004/016631, February 26, 2004. [6] X.C. Jiang, C.Y. Yu, C.X. Li, Z.H. Wang, Preparation of a novel ionic liquid 1-butyl-3-methyl imidazolium dibutylphosphate, J. Beijing Univ. Chem. Tech. (Sci. Ed.) 33 (1) (2006) 5–7. [7] J. Gmehling, U. Onken, Vapor–Liquid Equilibrium Data Collection, DECHEMA, Frankfurt, 1977, p. 41, 154.

[8] S.I. Sandler, Chemical and Engineering Thermodynamics, John Wiley & Sons, Singapore, 1989, p. 382, 372. [9] M.J. Earle, J.M.S.S. Esperance, M.A. Gilea, J.N.C. Lopes, L.P.N. Rebelo, J.W. Magee, K.R. Seddon, J.A. Widegren, The distillation and volatility of ionic liquids, vapor pressure of ionic liquid, Nature 439 (2006) 831–834. [10] M. Doker, J. Gmehling, Measurement and prediction of vapor–liquid equilibria of ternary systems containing ionic liquids, Fluid Phase Equilib. 227 (2005) 255–266. [11] Q.B. Shi, F.C. Zheng, C.X. Li, Calculation of vapor–liquid equilibrium for ionic liquid-containing systems with NRTL equation, J. Chem. Ind. Eng. (China) 56 (2005) 751–756.