Selection of entrainers for the separation of the binary azeotropic system methanol + dimethyl carbonate by extractive distillation

Selection of entrainers for the separation of the binary azeotropic system methanol + dimethyl carbonate by extractive distillation

Fluid Phase Equilibria 310 (2011) 166–181 Contents lists available at ScienceDirect Fluid Phase Equilibria journal homepage: www.elsevier.com/locate...

2MB Sizes 4 Downloads 124 Views

Fluid Phase Equilibria 310 (2011) 166–181

Contents lists available at ScienceDirect

Fluid Phase Equilibria journal homepage: www.elsevier.com/locate/fluid

Selection of entrainers for the separation of the binary azeotropic system methanol + dimethyl carbonate by extractive distillation Hiroyuki Matsuda a,∗ , Hideyuki Takahara a , Satoshi Fujino a , Dana Constantinescu b , Kiyofumi Kurihara a , Katsumi Tochigi a , Kenji Ochi a , Jürgen Gmehling b a b

Department of Materials and Applied Chemistry, Nihon University, 1-8 Kanda Surugadai, Chiyoda-ku, Tokyo 101-8308, Japan Carl von Ossietzky Universität Oldenburg, Technische Chemie, D-26111 Oldenburg, Germany

a r t i c l e

i n f o

Article history: Received 7 May 2011 Received in revised form 4 August 2011 Accepted 8 August 2011 Available online 16 August 2011 Keywords: Entrainer Vapor–liquid equilibria Dimethyl carbonate Activity coefficient model Residue curve map Solvent effect

a b s t r a c t The objective of this study is the selection of a suitable entrainer for the separation of the binary azeotropic mixture, namely methanol + dimethyl carbonate (DMC) by extractive distillation. In this study, 2-ethoxyethanol and 4-methyl-2-pentanone were considered as entrainer candidates for the separation of methanol and DMC. Isobaric vapor–liquid equilibria (VLE) were measured for the ternary mixtures methanol + DMC + 2-ethoxyethanol and methanol + DMC + 4-methyl-2-pentanone, and for their constituent binary mixtures, under reduced pressures of 66.66 kPa and 93.32 kPa. Experimental VLE data for the constituent binary mixtures were represented by the Wilson and the NRTL model. The VLE predictions for two ternary mixtures were compared with the experimental VLE data on the basis of the Wilson or NRTL parameters obtained from the constituent binary VLE data. The modified UNIFAC (Dortmund) model was also used for the prediction of the binary and ternary mixtures. The solvent effects of the two entrainer candidates investigated are discussed using the predicted results for the ternary mixtures and both 2-ethoxyethanol and 4-methyl-2-pentanone were found to be suitable entrainers for the separation of methanol and DMC. When the selectivity of 2-ethoxyethanol and 4-methyl-2-pentanone are compared, it could be concluded from the separation factors that 4-methyl-2-pentanone shows a higher selectivity. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Dimethyl carbonate (DMC) is a non-toxic substance that is widely used as a replacement for dimethyl sulfate, methyl halides, and phosgene in methylation and carbonylation reactions because it is considered to be an “environmentally benign building block” [1–3]. In recent years, therefore, DMC has become attractive as an environmentally benign solvent. DMC is also important as an excellent co-solvent for non-aqueous electrolytes in lithium batteries [4], and as an intermediate material in an environmentally benign process for polycarbonate production [5–8]. DMC has also been attracting increasing attention as a promising candidate for an oxygen-containing fuel additive to replace methyl tert-butyl ether (MTBE) [9]. There are at present several methods for the synthesis of DMC [10], for example, phosgenation of methanol using phosgene and hydrogen chloride [11], and oxidative carbonylation of methanol with oxygen and carbon monoxide [12]. However, the phosgene-based process has a number of environmental draw-

∗ Corresponding author. Tel.: +81 3 3259 0793; fax: +81 3 3293 7572. E-mail address: [email protected] (H. Matsuda). 0378-3812/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.fluid.2011.08.007

backs, particularly the use of phosgene, which is highly toxic, as a raw material. The direct synthesis of DMC using carbon dioxide and methanol at supercritical conditions [13–18] is also attractive because carbon dioxide is an environmentally benign and thermodynamically stable compound, and this process does not use or form toxic and corrosive compounds such as phosgene and hydrogen chloride. However, as has been reported in previous papers [13,18], the yield of DMC is low because of the limitations of the reaction thermodynamics. A more promising approach for the synthesis of DMC is based on transesterification of cyclic carbonates with methanol [19–21]. The cyclic carbonate used as a raw material in this transesterification reaction can be synthesized by cycloaddition of carbon dioxide to epoxides [11]. DMC synthesis by transesterification is therefore attractive from the viewpoint of sustainable green chemistry because carbon dioxide, which is a greenhouse gas and an abundant carbon resource, can be used as a starting material for DMC production [20,22]. In this transesterification reaction of cyclic carbonates with methanol, the product contains DMC, ethylene glycol, and two unreacted compounds, namely ethylene carbonate and methanol. There is a separation problem because methanol and DMC form an azeotrope [23–34]. A process for separating DMC from this mixture is therefore needed. Several methods for the separation of

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

167

Table 1 Normal boiling points Tb , densities  at 298.15 K, and liquid molar volumes vLi at 298.15 K, of pure components used in this study. Component

 (298.15 K) (kg m−3 )

Tb (K)

vLi × 106 (m3 mol−1 )

Experimental

Literature

Experimental

Literature

Methanol DMC

337.65 363.39

786.67 1063.38

786.37a 1063.28a

40.73 84.71

0.999 0.999

2-Ethoxyethanol 4-Methyl-2-pentanone

407.98 389.10

337.696a 363.60b 408.8a 407.95c 390.6a 389.45d

925.22 796.02

925.20a 796.25e

97.40 125.82

0.998 0.998

a b c d e

Ref. [56]. Ref. [24]. Ref. [57]. Ref. [58]. Ref. [59].

Table 2 Experimental vapor pressures of pure components used in this study. P (kPa)

T (K)

40.00 53.33 66.66 79.99 93.32 101.3

Methanol

DMC

2-Ethoxyethanol

4-Methyl-2-pentanone

315.74 322.17 327.36 331.75 335.56 337.65

336.61 344.33 350.64 356.01 360.72 363.39

379.75 387.90 394.56 400.26 405.25 407.98

359.61 368.13 375.07 380.98 386.25 389.10

DMC from methanol + DMC mixtures have been proposed: lowtemperature crystallization [35], high-pressure distillation [36], azeotropic distillation [37], extractive distillation [38–41], membrane separation [42–44], and adsorption [45]. In this study, we focus on extractive distillation as the separation method for this mixture. For the design of extractive distillation processes, a reliable knowledge of vapor–liquid equilibrium (VLE) data is required. In particular, the knowledge of the enhancement of the separation factor of the binary azeotropic mixture – methanol + DMC – by addition of an entrainer is important. Some entrainers, namely furfural [39], aniline [40], ethylene glycol [40], and phenol [41] have been investigated, but ethylene glycol was found to be an unsuitable entrainer for the separation of this mixture [40]. In this study, 2-ethoxyethanol and 4-methyl-2-pentanone (methyl isobutyl ketone, MIBK) were investigated as entrainer candidates for the separation of the binary azeotropic mixture methanol + DMC. They were selected on the basis of the following criteria for the selection of entrainers for extractive distillation shown by Gmehling and Möllmann [46]: (a) the entrainer is sup-

Table 3 Determined Antoine equation constantsa of pure components used in this study. Component

A

B

C

  Pis 

Methanol DMC 2-Ethoxyethanol 4-Methyl-2-pentanone

6.68015 5.78894 5.94768 5.85876

1289.885 1049.375 1197.630 1197.160

−61.736 −85.977 −104.155 −78.373

0.10 0.32 0.22 0.29

a

b

Purity (mass faction)

posed to exhibit zeotropic behavior with all components of the system to be separated; (b) the entrainer should alter the activity coefficients of the components to be separated to different extents, in order to achieve separation factors far from unity; and (c) for convenient recovery of the entrainer, its boiling point must usually be sufficiently higher (e.g., T = 40 K) than that of any of the components of the mixture to be separated. In this study, VLE data for two ternary mixtures containing the entrainer candidate were investigated: methanol + DMC + 2-ethoxyethanol and methanol + DMC + 4-methyl-2-pentanone; their five constituent binary mixtures, i.e., methanol + DMC, methanol + 2ethoxyethanol, DMC + 2-ethoxyethanol, methanol + 4-methyl-2pentanone, and DMC + 4-methyl-2-pentanone, were measured at reduced pressures of 66.66 kPa and 93.32 kPa using a modified Rogalski–Malanowski equilibrium still. Isobaric VLE data at 101.3 kPa for the constituent binary mixtures methanol + DMC and methanol + 4-methyl-2-pentanone have been reported by several authors [29,31,34,47,48]. Thus, in this study, the measurements of VLE data at two reduced pressures were performed for ternary mixtures. The experimental constituent binary VLE data were correlated by the Wilson [49] and the NRTL [50] model. The VLE for the two ternary mixtures were predicted using binary Wilson or NRTL parameters obtained from the constituent binary VLE data, and the predicted ternary VLE data were compared with the experimental results. In addition, the modified UNIFAC (Dortmund) model [51–55] was also tested for the prediction of these binary and ternary mixtures. Finally, the solvent effects of the two entrainers investigated were evaluated by analysis of the residue curves and calculation of the separation factors ˛12 using the predicted results for the ternary mixtures.

av.

(kPa)b

Literature (Antoine constant) [60] [61] [57] [62]

  Pis  0.16 0.42 0.16 0.19

av.

(kPa)c

Literature (vapor pressure) [63] [61] [57] [62]

log(Pis /kPa) = A − B/[(T/K) + C].

  Pis 

av.

=

NDP    s s s s Pi,exptl  /NDP, where Pi,exptl − Pi,calcd is the experimental vapor pressure, and Pi,calcd is the calculated vapor pressure using Antoine constants in the k

k=1

literature. c

  Pis 

av.

=

NDP    s s s s Pi,lit  /NDP, where Pi,lit − Pi,calcd is the literature vapor pressure, and Pi,calcd is the calculated vapor pressure using the determined Antoine constants. k

k=1

168

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Table 4 Experimental isobaric VLE data for the binary system methanol(1) + DMC(2). x1

y1

T (K)

1

2

(GE /RT)/x1 x2

0.000 0.059 0.079 0.096 0.118 0.173 0.251 0.287 0.318 0.352 0.443 0.500

0.000 0.306 0.358 0.401 0.451 0.514 0.591 0.614 0.629 0.642 0.680 0.699

350.64 341.62 340.04 338.57 337.32 334.44 331.92 330.95 330.24 329.87 328.50 328.15

– 2.932 2.721 2.655 2.550 2.221 1.949 1.842 1.754 1.642 1.463 1.352

1.000 1.018 1.020 1.025 1.010 1.065 1.094 1.127 1.166 1.202 1.322 1.405

– 1.437 1.337 1.334 1.141 1.331 1.248 1.274 1.306 1.287 1.312 1.283

0.000 0.045 0.065 0.079 0.106 0.131 0.151 0.170 0.202 0.254 0.337 0.377

0.000 0.230 0.300 0.348 0.402 0.456 0.479 0.509 0.552 0.608 0.652 0.671

360.72 354.20 352.11 350.51 348.86 347.15 345.93 344.85 343.34 341.25 339.25 338.55

– 2.563 2.490 2.515 2.297 2.243 2.138 2.100 2.027 1.922 1.677 1.585

1.000 1.000 0.996 0.995 0.996 0.990 1.013 1.016 1.018 1.029 1.108 1.144

– 0.977 0.918 0.941 0.891 0.851 0.982 0.984 0.974 0.988 1.083 1.096

y1

T (K)

1

2

(GE /RT)/x1 x2

66.66 kPa 0.597 0.651 0.668 0.765 0.802 0.814 0.843 0.885 0.905 0.920 1.000

0.736 0.756 0.760 0.800 0.820 0.827 0.839 0.866 0.879 0.893 1.000

327.49 327.13 326.97 326.55 326.48 326.43 326.41 326.47 326.52 326.69 327.36

1.226 1.172 1.156 1.082 1.061 1.056 1.036 1.016 1.006 0.998 1.000

1.571 1.702 1.772 2.123 2.274 2.331 2.573 2.916 3.181 3.316 –

1.261 1.272 1.293 1.318 1.322 1.333 1.343 1.343 1.340 1.278 –

93.32 kPa 0.451 0.575 0.667 0.716 0.787 0.843 0.862 0.896 0.921 0.946 1.000

0.700 0.743 0.771 0.785 0.816 0.841 0.855 0.878 0.899 0.924 1.000

337.35 336.11 335.33 335.17 334.86 334.67 334.76 334.79 334.89 335.02 335.56

1.448 1.266 1.168 1.115 1.067 1.035 1.025 1.011 1.004 0.999 1.000

1.239 1.438 1.686 1.867 2.157 2.547 2.634 2.937 3.188 3.492 –

1.150 1.187 1.249 1.254 1.282 1.326 1.302 1.312 1.303 1.303 –

y1

T (K)

1

2

(GE /RT)/x1 x2

66.66 kPa 0.408 0.448 0.468 0.568 0.590 0.628 0.665 0.700 0.786 0.861 0.926 0.970 1.000

0.912 0.923 0.929 0.952 0.957 0.964 0.969 0.973 0.983 0.990 0.995 0.998 1.000

348.04 346.09 345.01 340.77 339.83 338.30 336.59 335.89 333.10 330.96 329.19 328.12 327.36

0.996 0.986 0.988 0.979 0.982 0.986 1.000 0.981 0.987 0.990 0.995 0.996 1.000

0.910 0.935 0.942 0.963 0.952 0.948 0.989 0.997 1.017 1.032 1.066 1.116 –

−0.238 −0.175 −0.150 −0.117 −0.128 −0.123 −0.016 −0.069 −0.041 −0.039 0.000 −0.019 –

93.32 kPa 0.604 0.652 0.675 0.726 0.781 0.834 0.855 0.868 0.899 0.935 0.962 0.975 0.989 1.000

0.952 0.961 0.965 0.973 0.980 0.986 0.988 0.989 0.992 0.995 0.997 0.998 0.999 1.000

348.32 346.33 345.39 343.42 341.73 340.06 339.47 339.09 338.21 337.25 336.71 336.36 336.00 335.56

0.974 0.979 0.983 0.991 0.989 0.993 0.993 0.994 0.996 0.997 0.992 0.993 0.994 1.000

1.025 1.040 1.045 1.050 1.057 1.059 1.070 1.099 1.091 1.112 1.174 1.211 1.401 –

−0.027 −0.001 0.011 0.036 0.020 0.028 0.031 0.059 0.053 0.064 −0.059 −0.092 −0.229 –

x1

Table 5 Experimental isobaric VLE data for the binary system methanol(1) + 2-ethoxyethanol(2). x1

y1

T (K)

1

2

(GE /RT)/x1 x2

0.000 0.020 0.044 0.074 0.139 0.170 0.203 0.218 0.258 0.275 0.295 0.316 0.352 0.371

0.000 0.179 0.349 0.499 0.678 0.723 0.772 0.789 0.826 0.839 0.853 0.869 0.887 0.895

394.56 389.60 384.35 378.75 369.45 366.36 362.54 361.47 358.15 356.56 355.18 353.68 351.43 350.25

– 1.071 1.100 1.100 1.056 1.016 1.028 1.013 1.002 1.008 1.001 1.003 0.995 0.993

1.000 0.988 0.963 0.935 0.919 0.927 0.928 0.915 0.915 0.928 0.925 0.908 0.914 0.923

– −0.510 −0.759 −0.804 −0.543 −0.427 −0.331 −0.389 −0.341 −0.261 −0.262 −0.301 −0.263 −0.227

0.000 0.024 0.045 0.060 0.097 0.123 0.232 0.244 0.258 0.289 0.381 0.475 0.504 0.550 0.577

0.000 0.184 0.309 0.384 0.520 0.593 0.768 0.780 0.794 0.821 0.881 0.918 0.928 0.941 0.947

405.25 399.52 395.43 392.78 386.53 383.11 370.78 369.64 368.30 365.92 359.72 354.29 352.68 350.54 349.39

– 0.984 0.982 0.983 0.978 0.970 0.963 0.963 0.967 0.963 0.961 0.966 0.974 0.976 0.975

1.000 0.998 0.985 0.973 0.975 0.960 0.987 0.994 0.999 0.996 0.983 1.009 1.007 1.002 1.009

– −0.085 −0.354 −0.472 −0.282 −0.369 −0.107 −0.075 −0.047 −0.065 −0.111 −0.047 −0.040 −0.051 −0.044

2. Experimental 2.1. Materials All chemicals were supplied by Wako Pure Chemical Industries, Ltd., Osaka, Japan. Special grade methanol, 2-ethoxyethanol, and 4methyl-2-pentanone were dried with 3A molecular sieves, and first grade DMC was dried with 4A molecular sieves. The purities of the materials were checked by gas chromatography and were found to

x1

be better than 0.999 mass fraction for methanol and DMC, and 0.998 mass fraction for 2-ethoxyethanol and 4-methyl-2-pentanone. The purities were further confirmed by measuring normal boiling points (Tb ) and densities () at 298.15 K. Tb were measured using a modified Rogalski–Malanowski equilibrium still (details are given in next section).  at 298.15 K were measured using a vibrating tube density meter (SS-D-200-Exp. Type, Shibayama Scientific Co. Ltd., Tokyo, Japan), with a reproducibility of 10−2 kg m−3 . Experimental boiling points and density data, and purity of the chemicals used in

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

169

Table 6 Experimental isobaric VLE data for the binary system DMC(1) + 2-ethoxyethanol(2). x1

y1

T (K)

1

2

(GE /RT)/x1 x2

0.000 0.046 0.093 0.137 0.157 0.182 0.241 0.294 0.314 0.334 0.363 0.421 0.441 0.470 0.494

0.000 0.237 0.401 0.503 0.541 0.582 0.658 0.711 0.727 0.743 0.764 0.799 0.808 0.821 0.832

394.56 387.75 382.38 378.81 377.24 375.47 372.18 369.13 368.28 367.40 366.29 364.12 363.44 362.38 361.60

– 1.676 1.622 1.525 1.496 1.461 1.373 1.332 1.308 1.291 1.264 1.219 1.202 1.185 1.171

1.000 1.005 1.001 0.993 0.995 0.997 0.996 1.019 1.024 1.028 1.032 1.056 1.074 1.103 1.120

– 0.651 0.547 0.439 0.445 0.448 0.402 0.469 0.468 0.467 0.455 0.471 0.491 0.529 0.542

0.000 0.040 0.105 0.168 0.255 0.293 0.341 0.386 0.422 0.450 0.478 0.511 0.542 0.597

0.000 0.178 0.385 0.522 0.644 0.682 0.721 0.755 0.779 0.794 0.810 0.826 0.838 0.860

405.25 400.16 393.42 387.99 382.23 380.14 378.02 376.13 374.62 373.65 372.68 371.66 370.74 369.04

– 1.477 1.441 1.406 1.335 1.304 1.257 1.228 1.210 1.190 1.176 1.156 1.136 1.114

1.000 1.002 0.999 1.002 1.020 1.035 1.052 1.064 1.078 1.096 1.105 1.124 1.157 1.215

– 0.457 0.399 0.424 0.464 0.492 0.497 0.493 0.509 0.519 0.519 0.524 0.548 0.593

y1

T (K)

1

2

(GE /RT)/x1 x2

66.66 kPa 0.543 0.584 0.607 0.632 0.655 0.685 0.764 0.792 0.823 0.854 0.885 0.918 0.985 1.000

0.854 0.868 0.877 0.885 0.893 0.901 0.925 0.934 0.942 0.952 0.961 0.972 0.994 1.000

360.24 359.10 358.58 357.96 357.47 356.78 355.15 354.57 353.97 353.38 352.80 352.21 350.98 350.64

1.142 1.120 1.107 1.095 1.083 1.069 1.039 1.032 1.022 1.015 1.008 1.003 0.997 1.000

1.141 1.189 1.199 1.230 1.246 1.301 1.412 1.446 1.534 1.579 1.672 1.728 2.140 –

0.534 0.569 0.558 0.573 0.568 0.596 0.612 0.615 0.640 0.636 0.652 0.633 0.577 –

93.32 kPa 0.628 0.695 0.732 0.753 0.774 0.795 0.817 0.839 0.861 0.888 0.914 0.942 0.988 1.000

0.872 0.895 0.908 0.915 0.922 0.929 0.935 0.943 0.950 0.960 0.968 0.978 0.995 1.000

368.20 366.70 366.00 365.56 365.12 364.65 364.22 363.77 363.35 362.82 362.35 361.71 360.82 360.72

1.101 1.069 1.052 1.045 1.038 1.034 1.026 1.022 1.016 1.012 1.007 1.007 1.005 1.000

1.244 1.321 1.354 1.382 1.411 1.443 1.506 1.529 1.581 1.604 1.704 1.784 2.033 –

0.606 0.619 0.604 0.607 0.611 0.622 0.640 0.640 0.648 0.643 0.661 0.737 1.137 –

y1

T (K)

1

2

(GE /RT)/x1 x2

0.881 0.897 0.914 0.920 0.923 0.934 0.948 0.962 0.982 0.993 1.000

332.29 331.30 330.66 330.28 329.76 329.19 328.66 328.18 327.70 327.51 327.36

1.193 1.145 1.106 1.083 1.063 1.039 1.017 1.006 1.001 0.998 1.000

1.431 1.535 1.576 1.683 1.912 2.150 2.468 2.823 3.187 3.438 –

1.038 1.048 0.992 1.013 1.110 1.150 1.170 1.208 1.216 1.101 –

0.890 0.898 0.900 0.912 0.922 0.931 0.936 0.947 0.960 0.975 1.000

340.46 339.89 339.78 338.98 338.27 337.79 337.53 336.99 336.54 336.10 335.56

1.170 1.138 1.117 1.079 1.059 1.039 1.031 1.015 1.008 1.005 1.000

1.416 1.497 1.553 1.716 1.884 2.063 2.165 2.491 2.700 2.911 –

0.977 0.994 0.995 1.023 1.072 1.094 1.111 1.159 1.182 1.218 –

x1

Table 7 Experimental isobaric VLE data for the binary system methanol(1) + 4-methyl-2-pentanone(2). x1

y1

T (K)

1

2

(GE /RT)/x1 x2

0.000 0.072 0.141 0.192 0.225 0.275 0.358 0.415 0.463 0.507 0.559

0.000 0.448 0.628 0.696 0.736 0.769 0.812 0.832 0.844 0.855 0.870

375.07 359.35 350.63 346.42 343.60 341.06 338.00 336.24 334.88 334.08 333.05

– 1.870 1.809 1.714 1.717 1.615 1.474 1.396 1.339 1.279 1.230

1.000 1.000 0.997 1.016 1.026 1.061 1.104 1.164 1.246 1.304 1.366

– 0.681 0.671 0.748 0.811 0.877 0.879 0.935 1.018 1.023 1.027

0.000 0.035 0.096 0.123 0.210 0.266 0.303 0.358 0.402 0.461 0.564

0.000 0.261 0.501 0.571 0.711 0.758 0.779 0.808 0.822 0.847 0.874

386.25 377.15 366.70 363.12 354.38 350.80 349.10 346.77 345.66 344.08 341.71

– 1.786 1.726 1.724 1.687 1.610 1.544 1.475 1.392 1.326 1.222

1.000 1.005 1.012 1.012 1.028 1.056 1.082 1.115 1.158 1.175 1.313

– 0.746 0.729 0.714 0.793 0.855 0.884 0.910 0.918 0.872 0.943

this study are shown in Table 1 together with the literature values [24,56–59]. As shown in Table 1, the value of  at 298.15 K in the present study approximately agrees with the literature values. Also the experimental normal boiling points Tb of methanol and DMC agree well with the published values given Refs. [24,56]. On the other hand, for 2-ethoxyethanol and 4-methyl-2-pentanone, good agreement between the experimental and literature values (Refs. [57,58]) were obtained, while the experimental Tb were somewhat lower than those given in Ref. [56]. Tables 2 and 3, respectively, show the experimental vapor pressures and the Antoine constants of the compounds investigated

x1 66.66 kPa 0.602 0.665 0.720 0.752 0.785 0.832 0.882 0.923 0.967 0.988 1.000 93.32 kPa 0.629 0.667 0.684 0.740 0.784 0.822 0.841 0.883 0.917 0.951 1.000

in this study. The average deviations between the experimental and calculated values using the Antoine constants in the literature [60,61,57,62] are within 0.32 kPa. The average deviations between the literature [61,57,62,63] and calculated values using the determined Antoine constants are within 0.42 kPa at reduced pressures. 2.2. Experimental apparatus and procedures A modified Rogalski–Malanowski equilibrium still (Hiaki et al. [64]; Kurihara et al. [65,66]; Tochigi et al. [67,68]; Matsuda et al. [69]) was used for the VLE measurements. The equilibrium still

170

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Table 8 Experimental isobaric VLE data for the binary system DMC(1) + 4-methyl-2-pentanone(2). x1

y1

T (K)

1

2

(GE /RT)/x1 x2

0.000 0.044 0.071 0.102 0.131 0.168 0.236 0.293 0.358 0.405 0.438 0.506

0.000 0.107 0.166 0.227 0.279 0.340 0.440 0.510 0.581 0.624 0.651 0.702

375.07 372.84 371.61 370.24 369.04 367.51 365.13 363.38 361.49 360.36 359.68 358.21

– 1.199 1.196 1.186 1.176 1.171 1.160 1.144 1.132 1.115 1.099 1.076

1.000 1.002 1.002 1.004 1.006 1.012 1.012 1.015 1.019 1.026 1.032 1.056

– 0.243 0.219 0.229 0.234 0.258 0.244 0.240 0.247 0.246 0.241 0.256

0.000 0.051 0.106 0.144 0.198 0.265 0.306 0.352 0.408 0.491

0.000 0.120 0.231 0.295 0.379 0.469 0.518 0.567 0.616 0.686

386.25 383.60 381.04 379.46 377.35 374.95 373.65 372.29 370.98 368.99

– 1.198 1.191 1.170 1.160 1.149 1.141 1.130 1.101 1.082

1.000 1.001 1.003 1.007 1.010 1.015 1.017 1.022 1.034 1.049

– 0.2187 0.2206 0.2320 0.2354 0.2465 0.2452 0.2495 0.2454 0.2520

y1

T (K)

1

2

(GE /RT)/x1 x2

66.66 kPa 0.537 0.574 0.613 0.655 0.722 0.756 0.790 0.827 0.873 0.948 1.000

0.726 0.751 0.776 0.801 0.842 0.862 0.880 0.899 0.926 0.970 1.000

357.50 356.81 356.11 355.42 354.28 353.77 353.28 352.79 352.12 351.22 350.64

1.073 1.062 1.052 1.040 1.030 1.024 1.018 1.010 1.008 1.003 1.000

1.062 1.075 1.092 1.115 1.145 1.160 1.194 1.242 1.270 1.300 –

0.265 0.268 0.274 0.279 0.294 0.296 0.307 0.317 0.336 0.328 –

93.32 kPa 0.556 0.614 0.668 0.701 0.753 0.789 0.833 0.876 0.948 1.000

0.734 0.769 0.807 0.827 0.857 0.878 0.901 0.925 0.969 1.000

367.56 366.39 365.35 364.81 364.09 363.57 362.94 362.32 361.33 360.72

1.068 1.050 1.046 1.038 1.024 1.018 1.009 1.005 1.004 1.068

1.068 1.109 1.115 1.130 1.158 1.178 1.233 1.285 1.311 1.068

0.2649 0.2935 0.2968 0.2996 0.2924 0.2912 0.3065 0.3242 0.3544 –

x1

Table 9 Results of thermodynamic consistency tests for the constituent binary mixtures. System

P (kPa)

Point  testa

Methanol(1) + DMC(2)

66.66 93.32 66.66 93.32 66.66 93.32 66.66 93.32 66.66 93.32

0.009 0.008 0.009 0.003 0.004 0.002 0.008 0.008 0.001 0.001

methanol(1) + 2-ethoxyethanol(2) DMC(1) + 2-ethoxyethanol(2) methanol(1) + 4-methyl-2-pentanone(2) DMC(1) + 4-methyl-2-pentanone(2) a b

y1 

Area testb D−J

av.

+ + + + + + + + + +

−1.49 −5.90 – – −17.81 −8.82 −4.14 −4.70 −6.97 −10.26

+ + – – + + + + + +

Ref. [73]. Ref. [74].

and experimental procedure are described in detail in our previous work by Hiaki et al. [64]. The measurement system consisted of the equilibrium still, a pressure controller, a platinum resistance thermometer, and a computer to analyze the data. The pressure in the apparatus was established by a pressure controller (DPI145, Druck Co., Kirchentellinsfurt, Germany) with an accuracy of ±0.005%. The equilibrium temperature was measured with a calibrated platinum resistance thermometer with an accuracy of ±0.01 K. A buffer tank and vacuum pump were connected to the pressure control system to allow setup of reduced pressure conditions. The reproducibility of the equilibrium temperatures was within 0.01 K over the entire mole fraction range. 2.3. Analysis Vapor and liquid samples were analyzed using a gas chromatograph (GC-1700, Shimadzu Co. Ltd., Kyoto, Japan) equipped with a flame ionization detector (FID). ULBON HR-20 M (50 m × 0.25 mm I.D., 0.50 ␮m film thickness, Shinwa Chemical Industries Ltd., Kyoto, Japan) was used as the column packing. Injections (0.5 ␮L) were performed in the split mode at a split ratio of 60/1. Helium was used as carrier gas at a flow rate of 2.0 mL min−1 . The detector temperature was maintained at 513 K. Compositions were determined by the relative area method using a calibration curve. The accuracy was ±0.001 mol fraction.

3. Results and discussion 3.1. Binary systems In this study, the VLE data for five constituent binary mixtures, i.e., methanol + DMC, methanol + 2-ethoxyethanol, DMC + 2ethoxyethanol, methanol + 4-methyl-2-pentanone, and DMC + 4methyl-2-pentanone, were measured at pressures of 66.66 kPa and 93.32 kPa. The experimental VLE data together with the values of the activity coefficients of component i ( i ) and the excess Gibbs energies, (GE /RT)/x1 x2 , delivered from  i for these mixtures are summarized in Tables 4–8.  i were calculated using Eq. (1) assuming ideal gas behavior, since no second virial coefficients were available and, the parameters of the Tsonopoulos method [70,71], that are required for the calculation of the vapor phase fugacity coefficients of DMC were missing: i =

Pyi Pis xi

(1)

Pis in Eq. (1) was expressed by the Antoine equation with the Antoine constants listed in Table 3. Diagrams of the experimental temperatures,  i , and (GE /RT)/x1 x2 versus the composition are shown in Figs. 1–4. For the system methanol + 2-ethoxyethanol, nearly ideal solution behavior was observed from the values of  i in Table 5 and Fig. 2. Furthermore, from Fig. 2, a variation of ln  i ,

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

171

Fig. 1. Temperature–composition relationships: 䊉, experimental T – x1 ; , experimental T – y1 ; ⊗, azeotropic points: —–, Wilson; ——, NRTL; —·—, modified UNIFAC (Dortmund). (a) methanol(1) + DMC(2) at 66.66 kPa; (b) methanol(1) + DMC(2) at 93.32 kPa; (c) methanol(1) + 2-ethoxyethanol(2) at 66.66 kPa; (d) methanol(1) + 2ethoxyethanol(2) at 93.32 kPa; (e) DMC(1) + 2-ethoxyethanol(2) at 66.66 kPa; and (f) DMC(1) + 2-ethoxyethanol(2) at 93.32 kPa.

and a non-regular distribution of the values for (GE /RT)/x1 x2 were found. The experimental binary VLE data were tested for thermodynamic consistency using the point test of Van Ness et al. [72], modified by Fredenslund et al. [73], and the Herington’s area test [74]. These point and area tests have been described by Gmehling and Onken [75]. For the system methanol + 2-ethoxyethanol, only

the point test was performed because the experimental VLE of this system showed nearly ideal behavior. The results of the two thermodynamic consistency tests are given in Table 9. As can be seen in this table, all the experimental VLE at the two pressures are thermodynamically consistent. The binary system methanol + DMC forms a minimum-boiling azeotrope. The binary azeotropic points were determined using a

172

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

graphical method [76] on the basis of the experimental VLE data. The determined binary azeotropic compositions and temperatures are shown in Table 10. The pressure dependences of the temperatures and methanol compositions of the determined azeotropic points are shown graphically in Fig. 5, together with the literature values [23–33], including our determined azeotropic points from the experimental boiling points measured by ebulliometry [25].

Table 10 Azeotropic points for the system methanol(1) + DMC(2). P (kPa)

y1,az

Taz (K)

66.66 93.32

0.838 0.848

326.38 334.69

Fig. 2. Activity coefficient and excess Gibbs energy–composition relationships: 䊉, experimental ln  1 –x1 ; , experimental ln  2 –x1 ; , experimental (GE /RT)/x1 x2 ; —, Wilson; ——, NRTL; —·—, modified UNIFAC (Dortmund). (a) methanol(1) + DMC(2) at 66.66 kPa; (b) methanol(1) + DMC(2) at 93.32 kPa; (c) methanol(1) + 2-ethoxyethanol(2) at 66.66 kPa; (d) methanol(1) + 2-ethoxyethanol(2) at 93.32 kPa; (e) DMC(1) + 2-ethoxyethanol(2) at 66.66 kPa; and (f) DMC(1) + 2-ethoxyethanol(2) at 93.32 kPa.

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

173

Fig. 3. Temperature–composition relationships: 䊉, experimental T–x1 ; , experimental T–y1 ; —, Wilson; ——, NRTL; —·—, modified UNIFAC (Dortmund). (a) methanol(1) + 4methyl-2-pentanone(2) at 66.66 kPa; (b) methanol(1) + 4-methyl-2-pentanone(2) at 93.32 kPa; (c) DMC(1) + 4-methyl-2-pentanone(2) at 66.66 kPa; and (d) DMC(1) + 4methyl-2-pentanone(2) at 93.32 kPa.

Fig. 5 shows that a reasonable agreement between the experimental and the literature azeotropic points was obtained. 3.2. Ternary systems Tables 11 and 12 and Figs. 6 and 7 show the experimental VLE data for the ternary mixtures of methanol + DMC + 2-ethoxyethanol and methanol + DMC + 4-methyl-2-pentanone at 66.66 kPa and 93.32 kPa. The tails of the solid arrows in Figs. 6 and 7 represent the experimental liquid compositions, and the heads of the arrows show the experimental vapor compositions on the same tie line. As Figs. 6 and 7 indicate, the vapor–liquid tie lines of these mixtures turn toward the azeotropic point of the binary mixture methanol + DMC. In addition, these figures show that the two ternary mixtures do not form ternary azeotropic points. 4. Data reduction and prediction 4.1. Binary systems The experimental VLE data for the five constituent binary mixtures were represented using the Wilson and the NRTL model. The liquid molar volume for the pure component, vLi , in the Wilson

model was determined from the density measured at 298.15 K. The values of vLi are shown in Table 1. The non-randomness parameter, ˛12 , in the NRTL equation was set to 0.3. The binary parameters of the Wilson and the NRTL model were evaluated by the Marquardt method [77] for each system and pressure. In the binary mixtures, except for the mixture methanol + 2-ethoxyethanol, the following objective function (Fobj ) was minimized during optimization of the binary Wilson and NRTL parameters. Fobj =

 2 NDP   1,exptl − 1,calcd k=1

1,exptl

k

 +

2,exptl − 2,calcd 2,exptl

2  (2) k

In the mixture methanol + 2-ethoxyethanol, which shows nearly ideal behavior, the following objective function was used: Fobj =

NDP 

(Texptl − Tcalcd )2k

(3)

k=1

The estimated values of the parameters for the Wilson and the NRTL model, and the absolute deviations between the experimental and calculated boiling points and vapor phase mole fractions at each experimental pressure are listed in Tables 13 and 14. Similar correlation accuracies were obtained for the Wilson and the NRTL

174

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Fig. 4. Activity coefficient and excess Gibbs energy–composition relationships: 䊉, experimental ln  1 –x1 ; , experimental ln  2 –x1 ; , experimental (GE /RT)/x1 x2 ; —, Wilson; ——, NRTL; —·—, modified UNIFAC (Dortmund). (a) methanol(1) + 4-methyl-2-pentanone(2) at 66.66 kPa; (b) methanol(1) + 4-methyl-2-pentanone(2) at 93.32 kPa; (c) DMC(1) + 4-methyl-2-pentanone(2) at 66.66 kPa; and (d) DMC(1) + 4-methyl-2-pentanone(2) at 93.32 kPa.

model, but the Wilson model provides slightly better results for some VLE datasets. The VLE predictions for the five constituent binary mixtures were performed using the modified UNIFAC (Dortmund) model. The van der Waals volumes Rk and surface areas Qk of the functional groups, and the group interaction parameters anm , bnm , and cnm were taken from the papers by Gmehling et al. [52–55]. Some group interaction parameters were obtained with the help of the UNIFAC Consortium [78]. Deviations between the experimental and predicted data are listed in Tables 13 and 14, and the predicted diagrams are shown in Figs. 1–4. Comparisons between the experimental and predicted VLE data show that the modified UNIFAC (Dortmund) model provides reasonable prediction accuracy, while some prediction errors can be observed in the mixtures methanol + 4-methyl-2-pentanone and DMC + 4-methyl2-pentanone.

tion results, and the prediction accuracy between these models was almost the same. The predicted results for the two ternary mixtures using the Wilson model are compared with the experimental data in Figs. 6 and 7. The modified UNIFAC (Dortmund) model was also tested for the VLE predictions of the two ternary mixtures. The predicted results are listed in Tables 13 and 14. The prediction accuracy obtained with the modified UNIFAC (Dortmund) model was almost similar to those obtained using the Wilson and the NRTL model. 4.3. Solvent effects of entrainers In this study, the solvent effects of the two entrainers investigated were examined by two approaches, i.e., analysis of the residue curves and calculation of the separation factors. The residue curves are described by the following set of differential equations:

4.2. Ternary systems

dxi = xi − yi d

The VLE predictions, obtained using the binary Wilson and NRTL parameters, for the two ternary mixtures are shown in Tables 13 and 14. The results are also summarized in these tables. Both the Wilson and the NRTL model provides reasonable predic-

where  is a dimensionless measure of time. In this study, xi was solved using the Runge–Kutta fourth-order method, coupled with the Wilson model with the constituent binary parameters listed in Tables 13 and 14. Figs. 8 and 9 show the calculated residue

i = 1, . . . , n − 1

(4)

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181 Table 11 Experimental isobaric VLE data for the ternary system methanol(1) + DMC(2) + 2ethoxyethanol(3). x1

x2

y1

y2

0.036 0.071 0.079 0.084 0.086 0.110 0.161 0.179 0.183 0.241 0.261 0.317 0.337 0.340 0.346 0.351 0.369 0.369 0.380 0.400 0.412 0.416 0.417 0.424 0.426 0.433 0.433 0.448 0.450 0.467 0.483 0.485 0.496 0.509 0.514 0.520 0.532 0.548 0.557 0.564 0.568 0.585 0.587 0.608 0.608 0.621 0.643 0.654 0.677 0.707 0.737 0.739 0.765 0.790 0.797 0.814 0.862 0.885

0.564 0.832 0.441 0.224 0.904 0.053 0.748 0.481 0.807 0.016 0.730 0.022 0.581 0.024 0.259 0.603 0.622 0.064 0.391 0.030 0.450 0.473 0.086 0.032 0.188 0.254 0.491 0.509 0.095 0.525 0.294 0.039 0.224 0.042 0.322 0.235 0.334 0.348 0.128 0.258 0.361 0.268 0.372 0.279 0.384 0.148 0.058 0.299 0.168 0.176 0.184 0.070 0.074 0.077 0.200 0.080 0.086 0.090

0.158 0.339 0.318 0.370 0.417 0.550 0.516 0.513 0.566 0.786 0.615 0.829 0.641 0.841 0.685 0.655 0.667 0.805 0.672 0.858 0.680 0.679 0.807 0.866 0.739 0.719 0.688 0.693 0.811 0.702 0.722 0.876 0.748 0.879 0.728 0.749 0.728 0.734 0.813 0.754 0.736 0.757 0.742 0.759 0.744 0.813 0.887 0.765 0.811 0.812 0.813 0.888 0.887 0.886 0.815 0.886 0.886 0.885

0.742 0.643 0.583 0.444 0.581 0.152 0.473 0.452 0.433 0.043 0.385 0.050 0.352 0.053 0.275 0.342 0.333 0.119 0.309 0.060 0.310 0.313 0.137 0.063 0.226 0.254 0.307 0.304 0.143 0.297 0.261 0.070 0.230 0.073 0.261 0.234 0.261 0.260 0.163 0.235 0.259 0.234 0.256 0.235 0.256 0.172 0.089 0.232 0.180 0.181 0.183 0.099 0.102 0.103 0.184 0.107 0.111 0.114

0.012 0.013 0.024 0.027 0.069 0.084 0.108 0.117 0.135 0.147 0.196 0.199 0.212 0.233 0.235

0.031 0.608 0.271 0.291 0.109 0.029 0.635 0.691 0.766 0.803 0.021 0.189 0.023 0.534 0.025

0.092 0.090 0.133 0.140 0.359 0.450 0.384 0.414 0.461 0.493 0.691 0.569 0.708 0.565 0.740

T (K)

66.66 kPa 355.06 344.09 354.44 363.53 340.21 370.73 338.00 345.22 334.36 358.28 331.75 352.62 332.36 350.66 341.86 330.97 329.53 346.61 336.05 347.07 332.69 331.71 343.53 345.51 339.12 337.31 330.58 329.43 341.51 328.29 333.99 342.50 335.08 337.20 331.90 333.84 330.93 329.90 335.74 331.64 328.97 330.64 328.05 329.67 327.16 332.92 335.23 327.81 330.48 329.36 328.30 331.56 330.72 329.90 326.37 329.16 327.70 327.03 93.32 kPa 0.139 396.41 0.814 366.92 0.599 377.72 0.612 376.33 0.291 383.09 0.092 386.83 0.571 355.58 0.555 353.06 0.525 349.08 0.501 346.91 0.052 373.59 0.301 363.31 0.054 371.45 0.406 350.01 0.056 369.27

1

2

3

1.526 2.460 1.430 1.153 2.893 1.040 2.082 1.416 2.320 1.016 1.964 0.990 1.546 1.003 1.109 1.605 1.650 1.024 1.240 0.990 1.324 1.362 1.018 0.998 1.079 1.108 1.389 1.418 1.023 1.445 1.138 0.988 1.099 1.158 1.173 1.103 1.179 1.204 1.036 1.119 1.211 1.128 1.227 1.133 1.233 1.040 0.999 1.147 1.052 1.056 1.060 1.009 1.008 1.008 1.066 1.009 1.013 1.013

1.132 0.974 1.161 1.297 0.934 1.506 0.999 1.137 0.976 2.080 1.064 2.123 1.192 2.206 1.452 1.180 1.181 2.141 1.346 2.265 1.338 1.337 2.049 2.357 1.821 1.622 1.322 1.323 2.085 1.313 1.638 2.397 1.816 2.832 1.625 1.848 1.629 1.623 2.195 1.845 1.619 1.841 1.613 1.847 1.621 2.237 2.697 1.836 2.274 2.284 2.307 2.874 2.897 2.906 2.312 2.995 3.069 3.097

1.115 1.368 0.945 0.837 1.781 0.832 1.202 0.719 1.199 0.893 0.000 0.910 1.136 0.905 0.832 0.935 0.000 0.877 0.911 0.921 0.948 0.993 0.853 0.899 0.853 0.888 0.963 1.087 0.845 2.072 0.931 0.903 0.907 1.107 0.915 0.854 1.179 0.876 0.850 0.855 1.125 0.893 0.819 0.816 0.000 0.839 0.920 1.086 0.855 0.936 0.839 0.945 0.993 1.256 6.146 1.044 0.988 0.711

1.068 2.274 1.305 1.273 1.047 0.969 1.700 1.846 2.052 2.180 0.940 1.055 0.951 1.409 0.960

1.633 1.104 1.326 1.312 1.378 1.481 1.065 1.035 1.013 0.995 1.673 1.469 1.689 1.087 1.720

1.060 0.963 0.955 0.962 0.881 0.938 1.069 1.101 1.155 1.084 0.963 0.935 0.990 0.974 0.955

175

Table 11 (Continued) x1

x2

y1

y2

T (K)

1

2

3

0.238 0.243 0.248 0.257 0.258 0.259 0.265 0.327 0.333 0.338 0.342 0.347 0.359 0.369 0.374 0.401 0.412 0.426 0.428 0.432 0.446 0.453 0.458 0.462 0.466 0.479 0.479 0.500 0.509 0.513 0.524 0.526 0.536 0.541 0.541 0.546 0.551 0.562 0.565 0.567 0.570 0.576 0.586 0.596 0.606 0.615 0.625 0.634 0.661 0.668 0.682 0.683 0.711 0.734 0.750 0.762 0.792 0.807 0.816 0.827 0.847 0.855 0.869 0.879 0.902 0.926

0.462 0.632 0.667 0.045 0.027 0.699 0.089 0.472 0.499 0.064 0.526 0.553 0.578 0.607 0.017 0.048 0.390 0.021 0.052 0.437 0.460 0.059 0.485 0.110 0.024 0.280 0.280 0.085 0.325 0.129 0.345 0.082 0.141 0.362 0.255 0.097 0.212 0.377 0.176 0.227 0.104 0.203 0.186 0.216 0.121 0.228 0.129 0.205 0.123 0.144 0.067 0.130 0.159 0.166 0.171 0.086 0.168 0.038 0.175 0.098 0.102 0.042 0.106 0.043 0.045 0.047

0.577 0.582 0.593 0.729 0.764 0.605 0.701 0.639 0.641 0.784 0.654 0.645 0.665 0.673 0.857 0.835 0.691 0.873 0.841 0.692 0.696 0.843 0.702 0.805 0.883 0.736 0.736 0.834 0.734 0.808 0.733 0.844 0.807 0.736 0.757 0.838 0.778 0.744 0.795 0.777 0.840 0.788 0.795 0.788 0.835 0.787 0.832 0.797 0.843 0.831 0.889 0.842 0.829 0.829 0.829 0.882 0.833 0.933 0.834 0.879 0.877 0.933 0.877 0.933 0.933 0.932

0.384 0.405 0.398 0.088 0.059 0.391 0.149 0.340 0.330 0.107 0.335 0.334 0.331 0.325 0.034 0.081 0.292 0.038 0.084 0.298 0.298 0.092 0.295 0.145 0.043 0.243 0.243 0.116 0.254 0.156 0.257 0.112 0.162 0.257 0.227 0.125 0.203 0.252 0.183 0.208 0.129 0.195 0.187 0.198 0.142 0.202 0.148 0.191 0.141 0.156 0.089 0.145 0.163 0.165 0.167 0.107 0.164 0.056 0.166 0.117 0.120 0.060 0.122 0.062 0.064 0.067

352.51 345.62 344.03 366.54 367.09 342.23 363.69 345.89 344.46 359.80 343.11 341.85 340.36 339.02 359.10 356.01 343.70 355.86 354.35 341.19 339.84 352.76 338.52 350.48 353.54 344.09 344.09 349.48 341.13 347.32 339.88 348.48 345.78 338.69 341.74 346.93 342.99 337.56 343.32 341.77 345.74 342.20 342.14 341.03 343.46 339.91 342.40 339.86 341.35 340.31 342.63 340.32 338.37 337.46 336.83 339.03 335.68 339.29 334.86 336.67 335.99 337.76 335.32 337.03 336.35 335.67

1.290 1.632 1.729 0.943 0.967 1.807 0.964 1.319 1.370 0.961 1.431 1.459 1.538 1.595 0.972 0.981 1.228 0.970 0.980 1.289 1.322 0.981 1.366 0.996 0.972 1.108 1.108 0.988 1.163 1.009 1.183 0.985 1.020 1.205 1.103 0.997 1.061 1.225 1.045 1.079 1.000 1.059 1.053 1.070 1.018 1.081 1.023 1.064 1.020 1.035 0.993 1.025 1.045 1.049 1.052 1.012 1.048 1.000 1.052 1.018 1.019 1.002 1.020 1.003 1.004 1.003

1.091 1.070 1.055 1.631 1.792 1.056 1.526 1.191 1.151 1.724 1.165 1.157 1.159 1.140 2.110 1.970 1.340 2.123 1.993 1.338 1.337 2.030 1.319 1.855 2.271 1.531 1.531 1.988 1.537 1.901 1.535 2.061 1.907 1.530 1.712 2.053 1.759 1.504 1.887 1.760 2.062 1.816 1.905 1.810 2.119 1.824 2.153 1.921 2.236 2.197 2.472 2.261 2.236 2.245 2.259 2.647 2.361 3.105 2.368 2.779 2.811 3.189 2.823 3.310 3.352 3.449

0.909 0.998 1.096 1.012 0.935 1.075 1.006 0.990 1.751 0.928 0.902 2.416 0.786 1.102 0.939 0.913 0.903 0.971 0.929 0.907 0.811 0.921 0.714 0.895 0.969 0.899 0.899 0.966 0.861 0.891 0.968 0.942 0.914 0.971 0.907 0.935 0.873 0.934 0.910 0.841 0.907 0.870 0.895 0.892 0.896 0.887 0.911 0.946 0.873 0.858 0.971 0.862 0.841 0.859 0.748 0.957 1.175 0.926 0.000 0.794 0.907 0.958 0.639 0.938 0.857 0.581

176

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Table 12 Experimental isobaric VLE data for the ternary system methanol(1) + DMC(2) + 4methyl-2-pentanone(3). x1

x2

y1

y2

0.041 0.045 0.045 0.056 0.064 0.071 0.085 0.085 0.093 0.114 0.119 0.129 0.131 0.136 0.146 0.162 0.181 0.189 0.200 0.205 0.215 0.220 0.225 0.236 0.258 0.288 0.309 0.323 0.355 0.370 0.386 0.386 0.406 0.449 0.461 0.474 0.490 0.499 0.538 0.568 0.570 0.570 0.609 0.635 0.637 0.670 0.703 0.760 0.786 0.819 0.840 0.856 0.867 0.873 0.894

0.524 0.214 0.566 0.646 0.687 0.724 0.279 0.795 0.840 0.350 0.207 0.374 0.139 0.147 0.399 0.666 0.453 0.251 0.481 0.731 0.262 0.508 0.766 0.113 0.298 0.129 0.325 0.140 0.362 0.156 0.070 0.149 0.400 0.168 0.440 0.176 0.456 0.252 0.090 0.094 0.275 0.151 0.100 0.164 0.026 0.108 0.180 0.194 0.203 0.092 0.094 0.099 0.038 0.104 0.042

0.229 0.275 0.243 0.283 0.306 0.326 0.413 0.365 0.388 0.479 0.519 0.495 0.541 0.573 0.513 0.520 0.554 0.613 0.570 0.577 0.636 0.594 0.592 0.687 0.664 0.715 0.679 0.732 0.691 0.750 0.787 0.760 0.703 0.771 0.712 0.777 0.717 0.750 0.823 0.828 0.770 0.801 0.831 0.805 0.878 0.836 0.807 0.809 0.807 0.863 0.866 0.867 0.916 0.867 0.914

0.564 0.311 0.577 0.591 0.591 0.591 0.314 0.589 0.588 0.328 0.223 0.332 0.161 0.163 0.335 0.429 0.336 0.214 0.338 0.406 0.215 0.333 0.406 0.108 0.223 0.113 0.229 0.118 0.240 0.127 0.063 0.122 0.251 0.132 0.265 0.138 0.271 0.186 0.080 0.085 0.191 0.127 0.090 0.141 0.027 0.102 0.158 0.177 0.190 0.106 0.110 0.117 0.048 0.125 0.060

0.023 0.033 0.039 0.046 0.056 0.122 0.141 0.146 0.149 0.152 0.156 0.170 0.180 0.181 0.185 0.188 0.210 0.214

0.059 0.262 0.603 0.673 0.747 0.608 0.254 0.668 0.512 0.343 0.163 0.286 0.557 0.062 0.383 0.183 0.609 0.112

0.168 0.201 0.201 0.228 0.290 0.461 0.533 0.497 0.511 0.539 0.579 0.561 0.557 0.654 0.570 0.610 0.580 0.677

T (K)

66.66 kPa 350.44 356.07 349.25 346.44 344.94 343.63 349.11 341.13 339.81 344.79 347.59 343.27 348.99 347.05 341.59 336.50 338.52 340.85 337.07 333.54 338.99 335.68 332.39 340.76 336.35 338.03 334.38 336.91 332.46 334.76 335.57 334.25 330.70 332.60 329.08 331.81 328.33 330.62 331.98 331.29 329.09 330.41 330.64 329.19 331.34 329.45 328.04 326.97 326.48 327.69 327.36 327.01 328.04 326.68 327.57 93.32 kPa 0.116 377.16 0.386 368.39 0.624 360.11 0.642 357.87 0.628 355.48 0.449 350.73 0.245 354.56 0.444 348.02 0.381 349.56 0.295 351.94 0.167 355.29 0.252 351.60 0.369 346.72 0.066 354.76 0.296 349.16 0.172 352.24 0.370 344.24 0.105 351.13

1

2

3

2.284 2.052 2.304 2.387 2.387 2.407 2.083 2.474 2.528 2.109 1.976 2.039 1.778 1.947 1.989 2.212 1.949 1.888 1.920 2.182 1.850 1.922 2.136 1.701 1.784 1.611 1.647 1.537 1.576 1.496 1.457 1.483 1.506 1.382 1.437 1.363 1.404 1.312 1.263 1.238 1.256 1.237 1.190 1.174 1.168 1.143 1.115 1.082 1.065 1.039 1.030 1.027 1.026 1.022 1.013

1.083 1.209 1.069 1.060 1.051 1.046 1.186 1.041 1.033 1.152 1.198 1.153 1.226 1.257 1.160 1.078 1.149 1.210 1.151 1.043 1.249 1.132 1.042 1.361 1.260 1.383 1.280 1.389 1.300 1.458 1.561 1.496 1.320 1.532 1.353 1.577 1.377 1.558 1.776 1.857 1.560 1.790 1.898 1.923 2.129 2.090 2.058 2.236 2.341 2.741 2.822 2.892 2.961 2.981 3.415

1.104 1.055 1.122 1.141 1.182 1.218 1.046 1.273 1.254 1.035 0.988 1.062 1.000 0.970 1.089 1.189 1.109 1.037 1.129 1.205 1.031 1.113 1.059 1.061 1.027 1.110 1.101 1.101 1.158 1.118 1.149 1.117 1.215 1.196 1.277 1.186 1.263 1.321 1.264 1.286 1.383 1.339 1.395 1.470 1.405 1.511 1.722 1.837 1.683 2.036 2.157 2.142 2.181 2.127 2.388

1.749 1.909 2.114 2.192 2.484 2.141 1.872 2.125 2.026 1.925 1.792 1.813 2.026 1.777 1.846 1.742 1.981 1.767

1.198 1.162 1.057 1.047 0.999 1.030 1.182 1.019 1.081 1.151 1.225 1.193 1.064 1.295 1.139 1.245 1.066 1.290

1.023 1.016 1.121 1.147 1.124 1.071 1.024 1.128 1.069 1.010 1.014 1.069 1.051 1.025 1.057 1.053 1.136 1.024

Table 12 (Continued) x1

x2

y1

y2

T (K)

1

2

3

0.215 0.221 0.243 0.261 0.261 0.277 0.284 0.305 0.306 0.312 0.323 0.348 0.374 0.376 0.384 0.388 0.416 0.434 0.445 0.481 0.510 0.555 0.573 0.587 0.623 0.638 0.673 0.675 0.724 0.725 0.766 0.766 0.778 0.826 0.872 0.892 0.944

0.419 0.070 0.354 0.452 0.123 0.714 0.240 0.087 0.520 0.138 0.395 0.560 0.294 0.102 0.187 0.602 0.319 0.204 0.485 0.512 0.372 0.397 0.153 0.257 0.163 0.279 0.173 0.072 0.184 0.076 0.195 0.028 0.083 0.087 0.092 0.037 0.044

0.597 0.700 0.622 0.621 0.705 0.628 0.680 0.746 0.646 0.728 0.665 0.664 0.711 0.779 0.742 0.680 0.721 0.753 0.703 0.709 0.739 0.746 0.804 0.772 0.809 0.776 0.813 0.858 0.818 0.864 0.818 0.897 0.868 0.873 0.877 0.919 0.932

0.299 0.068 0.262 0.301 0.111 0.369 0.187 0.077 0.309 0.116 0.260 0.314 0.204 0.085 0.143 0.317 0.213 0.152 0.280 0.289 0.234 0.243 0.123 0.187 0.133 0.202 0.144 0.070 0.155 0.077 0.171 0.034 0.086 0.097 0.109 0.049 0.063

346.47 351.97 345.96 344.16 348.51 340.05 345.65 347.30 341.19 346.24 342.27 339.44 341.88 344.22 343.12 337.84 340.33 341.41 337.72 336.62 337.61 336.41 339.01 337.35 337.99 336.21 337.02 338.48 336.11 337.61 335.24 338.21 336.80 336.01 335.25 336.40 335.42

1.834 1.717 1.723 1.712 1.657 1.905 1.630 1.568 1.699 1.555 1.590 1.641 1.490 1.487 1.445 1.604 1.441 1.384 1.453 1.415 1.338 1.301 1.227 1.227 1.182 1.187 1.142 1.135 1.107 1.101 1.083 1.057 1.064 1.039 1.019 0.998 0.994

1.156 1.299 1.221 1.172 1.360 1.058 1.300 1.392 1.166 1.373 1.241 1.175 1.327 1.463 1.398 1.172 1.353 1.450 1.291 1.316 1.412 1.439 1.712 1.650 1.806 1.715 1.912 2.112 2.004 2.275 2.157 2.665 2.400 2.662 2.913 3.115 3.498

1.072 1.004 1.107 1.121 1.043 1.618 1.088 1.064 1.199 1.079 1.182 1.190 1.155 1.072 1.152 1.593 1.196 1.206 1.296 1.595 1.227 1.291 1.349 1.424 1.430 1.505 1.534 1.472 1.674 1.589 1.668 1.752 1.834 1.975 2.299 2.540 2.445

Fig. 5. Azeotropic temperature Taz and composition y1,az as mole fraction, on the basis of the experimental VLE data and literature values, for the system methanol (1) + DMC (2): 䊉, this work; , Chen et al. [23]; , Comelli and Francesconi [24]; , Fukano et al. [25]; , Gmehling [26]; ♦, Han and Park [27]; 夽, Lecat [28]; , Luo et al. [29]; , Ma et al. [30]; , Rodríguez et al. [31]; , Shi et al. [32]; ×, Zhang et al. [33].

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Fig. 6. VLE tie lines (tails of arrows represent liquid-phase mole fractions x1 and x2 , and heads of arrows represent vapor-phase mole fractions y1 and y2 ) for the system methanol(1) + DMC(2) + 2-ethoxyethanol(3): ⊗, azeotropic point of , Wilson. (a) 66.66 kPa, and methanol(1) + DMC(2); →, experimental values; (b) 93.32 kPa.

curves at 93.32 kPa for the ternary systems methanol + DMC + 2ethoxyethanol and methanol + DMC + 4-methyl-2-pentanone. In Figs. 8 and 9, the contour lines (separation factor ˛12 = 1) are also shown graphically. As can be seen in these figures, the azeotropic point for the mixture methanol + DMC is an unstable node (residue curves begin), 2-ethoxyethanol and 4-methyl-2-pentanone are stable nodes (where residue curves end), and methanol and DMC are saddle points (no residue curves enter or exit). All residue curves approaching the 2-ethoxyethanol or 4-methyl-2-pentanone (entrainer) vertex are inflected toward the DMC + entrainer face. DMC and the entrainer could therefore be obtained as a bottom product, and methanol could be obtained as a distillate. Moreover, as described by Laroche et al. [79], if the isovolatility lines intersect (1)–(3) (entrainer) edge, component 1 is recovered as

177

Fig. 7. VLE tie lines (tails of arrows represent liquid-phase mole fractions x1 and x2 , and heads of arrows represent vapor-phase mole fractions y1 and y2 ) for the system methanol(1) + DMC(2) + 4-methyl-2-pentanone(3): ⊗, azeotropic point of , Wilson. (a) 66.66 kPa, and methanol(1) + DMC(2); →, experimental values; (b) 93.32 kPa.

a distillate in an extractive distillation column, and component 2 is recovered as a distillate in the entrainer-recovery column. This is the case for the isovolatility lines in both of the ternary mixtures, so methanol could be obtained as a distillate, and DMC could be obtained as a distillate in the entrainer-recovery column. The separation factors were calculated using the Wilson model with the binary parameters of the constituent binary mixtures listed in Tables 13 and 14. The separation factor, ˛12 , is defined as follows [46]: ˛12 =

K1 y1 /x1 = K2 y2 /x2

(5)

The separation factor ˛12 was calculated by predicting ternary VLE with several liquid mole fractions, x3 , of the entrainers. The

178

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

Fig. 8. Residue curve map and isovolatility line for the system , residue methanol(1) + DMC(2) + 2-ethoxyethanol(3) at 93.32 kPa: curve using Wilson model; ——, isovolatility line; ⊗, unstable node [azeotropic point of methanol(1) + DMC(2)]; 䊉, stable node; , saddle node.

Fig. 10. Predicted separation factors ˛12 for the system methanol(1) + DMC(2) + 2ethoxyethanol(3) by the Wilson model with several liquid mole fractions of , x3 = 0.462; , 2-ethoxyethanol x3 : (a) 66.66 kPa, ——, x3 = 0.000; , x3 = 0.700, and (b) 93.32 kPa, ——, x3 = 0.000; , x3 = 0.423; x3 = 0.500; , x3 = 0.500; , x3 = 0.700.

Fig. 9. Residue curve map and isovolatility line for the system , residue methanol(1) + DMC(2) + 4-methyl-2-pentanone(3) at 93.32 kPa: curve using Wilson model; ——, isovolatility line; ⊗, unstable node [azeotropic point of methanol(1) + DMC(2)]; 䊉, stable node; , saddle node.

calculated results for ˛12 versus the entrainer-free liquid phase mole fraction, x1s , with several x3 values for the two entrainers are shown graphically in Figs. 10 and 11. Lines with x3 = 0.000 are the results for the binary mixtures methanol + DMC, and show that the binary system cannot be separated by ordinary distillation, intersecting with a line ˛12 = 1 at the azeotropic point of this mixture. However, when 2-ethoxyethanol is added as an entrainer, the apparent azeotropic point disappears, and the value of ˛12 is larger than 1 over the entire mole fraction range of x1s , for x3 = 0.462

and 0.423 at pressures of 66.66 kPa and 93.32 kPa, respectively. When 4-methyl-2-pentanone is used as entrainer, a similar disappearance of the azeotropic point is observed for x3 = 0.179 and 0.170 at pressures of 66.66 kPa and 93.32 kPa, respectively. Therefore, 2-ethoxyethanol and 4-methyl-2-pentanone can be used as entrainers for extractive distillation of the binary azeotropic mixture methanol + DMC. When the performances of 2-ethoxyethanol and 4-methyl-2-pentanone as entrainers are compared, the disappearance of the azeotropic point occurs with a smaller amount of 4-methyl-2-pentanone than of 2-ethoxyethanol. In addition, a larger value of ˛12 can be obtained over the entire mole fraction range if excess 4-methyl-2-pentanone is added. From these results, it can be concluded that 4-methyl-2-pentanone is a more selective entrainer than 2-ethoxyethanol for the separation of methanol and DMC by extractive distillation.

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181 Table 13   Determined parameters and deviations between the calculated and experimental temperaturesa T 

179



av.



, and vapor-phase mole fractionsb yi 

av.

, for the Wilson, NRTL,

and modified UNIFAC (Dortmund) models for the ternary system methanol(1) + DMC(2) + 2-ethoxyethanol(3) and its constituent binary systems.

Wilson parameters 12 − 11 (J mol−1 ) 21 − 22 (J mol−1 ) Deviation  

T  (K)  av. y1   av. y2 

Methanol(1) + DMC(2)

Methanol(1) + 2-ethoxyethanol(2)

DMC(1) + 2-ethoxyethanol(2)

Methanol(1) + DMC(2) + 2ethoxyethanol(3)

66.66 kPa 3751.341 607.532 66.66 kPa

93.32 kPa 3136.792 1208.753 93.32 kPa

66.66 kPa 4977.917 −4334.977 66.66 kPa

66.66 kPa 483.676 1651.584 66.66 kPa

93.32 kPa 265.440 1844.061 93.32 kPa

66.66 kPa 93.32 kPa No ternary parameters 66.66 kPa

93.32 kPa

0.20

0.32

0.16

0.16

0.36

0.18

0.34

0.60

0.008

0.005

0.010

0.004

0.003

0.002

0.010

0.007













0.008

0.006

av.

NRTL parameters g12 − g22 (J mol−1 ) g21 − g11 (J mol−1 ) ˛12 Deviation  

T  (K)  av. y1   av. y2 

66.66 kPa 2540.354 1407.497 0.30 66.66 kPa

93.32 kPa 3571.763 463.577 0.30 93.32 kPa

0.20

0.32

av.

mod. UNIFAC (Do) Deviation  

av.

a

b

  T 

66.66 kPa 1071.853 −1019.935 0.30 66.66 kPa

93.32 kPa 1860.520 −1778.002 0.30 93.32 kPa

0.23

66.66 kPa 2602.086 −470.762 0.30 66.66 kPa

93.32 kPa 2945.868 −798.664 0.30 93.32 kPa

66.66 kPa

93.32 kPa

0.36

0.19

0.45

0.59

0.12

66.66 kPa 93.32 kPa No ternary parameters

0.008

0.005

0.011

0.004

0.002

0.002

0.008

0.007













0.007

0.006

66.66 kPa

T  (K)  av. y1   av. y2 

93.32 kPa 1794.559 −1924.698 93.32 kPa

66.66 kPa

93.32 kPa

0.48

93.32 kPa 0.29

66.66 kPa 0.66

93.32 kPa 0.21

66.66 kPa 0.24

93.32 kPa 0.31

0.45

0.62

0.012

0.007

0.016

0.004

0.008

0.006

0.016

0.014













0.012

0.012

  Texptl − Tcalcd  /NDP. NDP

av.

  yi 

=

k

k=1 NDP

av.

=

  yi,exptl − yi,calcd  /NDP, where NDP is the number of data points. k

k=1

Table 14   Determined parameters and deviations between the calculated and experimental temperaturesa T 



av.



, and vapor-phase mole fractionsb yi 

av.

, for the Wilson, NRTL,

and modified UNIFAC (Dortmund) models for the ternary system methanol(1) + DMC(2) + 4-methyl-2-pentanone(3) and its constituent binary systems.

Wilson parameters 12 − 11 (J mol−1 ) 21 − 22 (J mol−1 ) Deviation  

T  (K)  av. y1   av. y2 

Methanol(1) + DMC(2)

Methanol(1) + 4-methyl-2pentanone(2)

66.66 kPa 3751.341 607.532 66.66 kPa

93.32 kPa 3136.792 1208.753 93.32 kPa

66.66 kPa 3582.389 −31.192 66.66 kPa

93.32 kPa 3425.672 227.443 93.32 kPa

66.66 kPa

93.32 kPa

0.20

0.32

0.40

0.46

0.04

0.05

0.18

0.18

0.008

0.005

0.005

0.007

0.001

0.001

0.008

0.009













0.006

0.007

av.

NRTL parameters g12 − g22 (J mol−1 ) g21 − g11 (J mol−1 ) ˛12 Deviation  

T  (K)  av. y1   av. y2 

93.32 kPa 4295.536 −687.762 0.30 93.32 kPa

66.66 kPa

93.32 kPa

0.20

0.32

0.45

0.49

0.04

0.05

0.18

0.18

0.008

0.005

0.006

0.008

0.001

0.001

0.009

0.010













0.007

0.007

66.66 kPa

av.

b

93.32 kPa

66.66 kPa

av.

  yi 

=

66.66 kPa

93.32 kPa

0.29

0.77

0.79

0.29

0.35

0.84

0.76

0.012

0.007

0.009

0.009

0.005

0.006

0.010

0.010













0.008

0.007

  Texptl − Tcalcd  /NDP. k

k=1 NDP

av.

=

  yi,exptl − yi,calcd  /NDP, where NDP is the number of data points. k

k=1

66.66 kPa

93.32 kPa

66.66 kPa 93.32 kPa No ternary parameters

0.48

NDP

93.32 kPa

93.32 kPa 2579.996 −1372.202 0.30 93.32 kPa

66.66 kPa 93.32 kPa No ternary parameters

66.66 kPa 3793.206 −331.109 0.30 66.66 kPa

T  (K)  av. y1   av. y2 

66.66 kPa 2508.057 −1336.698 0.30 66.66 kPa

93.32 kPa 545.809 530.531 93.32 kPa

93.32 kPa 3571.763 463.577 0.30 93.32 kPa

av.

  T 

66.66 kPa 526.380 517.423 66.66 kPa

Methanol(1) + DMC(2) + 4methyl-2-pentanone(3)

66.66 kPa 2540.354 1407.497 0.30 66.66 kPa

mod. UNIFAC (Do) Deviation  

a

DMC(1) + 4-methyl-2pentanone(2)

180

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

UNIFAC (Dortmund) model. Modified UNIFAC provided reasonable prediction results for both the ternary and the constituent binary systems. Finally, the selectivity of the two entrainers investigated were evaluated by two approaches, i.e., residue curve maps and the calculation of the separation factors on the basis of the binary Wilson parameters. The calculated results for the residue curve maps and the separation factors at several mole fractions x3 of the entrainers show that both 2-ethoxyethanol and 4-methyl-2pentanone can be used as entrainer for extractive distillation of the binary azeotropic mixture methanol + DMC. Comparison of the performances of 2-ethoxyethanol and 4-methyl-2-pentanone as entrainers showed that 4-methyl-2-pentanone is the more selective entrainer for this separation since the calculated separation factors with 4-methyl-2-pentanone are higher than those with 2-ethoxyethanol, and the disappearance of the azeotropic point can be achieved with smaller amounts of 4-methyl-2-pentanone than of 2-ethoxyethanol.

Fig. 11. Predicted separation factors ˛12 for the system methanol(1) + DMC(2) + 4methyl-2-pentanone(3) by the Wilson model with several liquid mole fractions of 4, x3 = 0.179; , methyl-2-pentanone x3 : (a) 66.66 kPa, ——, x3 = 0.000; , x3 = 0.500; , x3 = 0.700, and (b) 93.32 kPa, ——, x3 = 0.000; x3 = 0.300; , x3 = 0.170; , x3 = 0.300; , x3 = 0.500; , x3 = 0.700.

5. Conclusions The isobaric VLE data of two ternary mixtures, namely methanol + DMC + 2-ethoxyethanol and methanol + DMC + 4methyl-2-pentanone, and their five constituent binary mixtures were measured at 66.66 kPa and 93.32 kPa. The results of thermodynamic consistency tests showed that the VLE data for the five constituent binary mixtures are thermodynamically consistent. Experimental VLE data for the two ternary mixtures showed that the vapor–liquid equilibrium tie lines turned toward the azeotropic point of the binary mixture methanol + DMC. The experimental VLE data of these binary mixtures were correlated using the Wilson and the NRTL model with good accuracy. The ternary VLE data were compared with predicted data using the binary Wilson and NRTL parameters, and reasonable prediction accuracy was obtained. VLE predictions for the constituent binary mixtures and the two ternary mixtures were also performed using the modified

List of symbols Antoine constants A, B, C anm , bnm , cnm group interaction parameters of modified UNIFAC (Dortmund) model Fobj objective function GE excess Gibbs energy (J mol−1 ) gij − gjj binary interaction parameter of the NRTL model (J mol−1 ) number of data points per system NDP P pressure (kPa) s P   saturated vapor pressure (kPa) P s  average absolute deviation between experimental and i av. calculated vapor pressure of component i (kPa) Qk surface area of a functional group in modified UNIFAC (Dortmund) model R genera gas constant = 8.314 (J mol−1 K−1 ) Rk van der Waals volume of functional group k in modified UNIFAC (Dortmund) absolute temperature (K) T   T  average absolute deviation between experimental and av. calculated boiling points (K) liquid molar volume (m3 mol−1 ) v xi liquid phase mole fraction of component i liquid phase mole fraction of component i (solvent-free xis basis) vapor phase mole fraction of component i yi yis vapor phase mole fraction of component i (solvent-free   basis) yi  average absolute deviation between experimental and av. calculated vapor phase mole fraction of component i Greek letters ˛12 non-randomness parameter of the NRTL model ˛12 separation factor activity coefficient  ij − ii binary interaction parameter in the Wilson model (J mol−1 ) density (kg m−3 )   dimensionless measure of time Superscripts L liquid s saturated s solvent-free Subscripts 1, 2, 3, i, j components 1, 2, 3, i, and j az azeotrope

H. Matsuda et al. / Fluid Phase Equilibria 310 (2011) 166–181

calcd exptl lit

calculated value experimental value literature value

References [1] Y. Ono, Appl. Catal. A 155 (1997) 133–166. [2] P. Tundo, M. Selva, Acc. Chem. Res. 35 (2002) 706–716. [3] P. Tundo, P. Anastas, Green Chemistry: Challenging Perspectives, Oxford University Press, Oxford, UK, 2001. [4] S. Matsuta, Y. Kato, T. Ohta, H. Kurokawa, S. Yoshimura, S. Fuhitami, J. Electrochem. Soc. 148 (2001) A7–A10. [5] S. Fukuoka S., M. Kawamura, K. Komiya, M. Tojo, H. Hachiya, K. Hasegawa, M. Aminaka, H. Okamoto, I. Fukawa, S. Konno, Green Chem. 5 (2003) 497– 507. [6] J. Haubrock, M. Raspe, G.F. Versteeg, H.A. Kooijman, R. Taylor, J.A. Hogendoorn, Ind. Eng. Chem. Res. 47 (2008) 9854–9861. [7] J. Haubrock, W. Wermink, G.F. Versteeg, H.A. Kooijman, R. Taylor, M. van Sint Annaland, J.A. Hogendoorn, Ind. Eng. Chem. Res. 47 (2008) 9862–9870. [8] J. Gong, X. Ma, S. Wang, Appl. Catal. A: Gen. 316 (2007) 1–21. [9] M.A. Pacheco, C.L. Marshall, Energy Fuels 11 (1997) 2–29. [10] B. Schäffner, F. Schäffner, S.P. Verevkin, A. Börner, Chem. Rev. 110 (2010) 4554–4581. [11] A.-A. Shaikh, S. Sivaram, Chem. Rev. 96 (1996) 951–976. [12] S. Uchiumi, K. Ataka, T. Matsuzaki, J. Organomet. Chem. 576 (1999) 279–289. [13] S. Fang, K. Fujimoto, Appl. Catal. A 142 (1996) L1–L3. [14] N. Isaacs, B. O’Sullivan, C. Verhaelen, Tetrahedron 55 (1999) 11949–11956. [15] T. Sakakura, J.-C. Choi, Y. Saito, T. Masuda, T. Sako, T. Oriyama, J. Org. Chem. 64 (1999) 4506–4508. [16] T. Sakakura, J. Choi, Y. Saito, T. Sako, Polyhedron 19 (2000) 573–576. [17] G. Chu, J. Park, M. Cheong, Inorg. Chim. Acta 307 (2000) 133–137. [18] X.L. Wu, M. Xiao, Y.Z. Meng, Y.X. Lu, J. Mol. Catal. A: Chem. 238 (2005) 158–162. [19] J.A. Gilpin, A.H. Emmons, United States Patent 3,803,201 (1974). [20] S.R. Jagtap, M.D. Bhor, B.M. Bhanage, Catal. Commun. 9 (2008) 1928–1931. [21] B.M. Bhanage, S. Fujita, Y. Ikushima, K. Torii, M. Arai, Green Chem. 2003 (5) (2003) 71–75. [22] H. Cui, T. Wang, F. Wang, C. Gu, P. Wang, Y. Dai, Ind. Eng. Chem. Res. 43 (2004) 7732–7739. [23] B. Chen, G. Li, Y. Zheng, J. Chem. Eng. Chin. Univ. 11 (1997) 193–196. [24] F. Comelli, R. Francesconi, J. Chem. Eng. Data 42 (1997) 705–709. [25] M. Fukano, H. Matsuda, K. Kurihara, K. Ochi, J. Chem. Eng. Data 51 (2006) 1458–1463. [26] J. Gmehling, Unpublished Data, 1972–2004. [27] K.-J. Han, S.-J. Park, J. Korea Inst. Chem. Eng. 43 (2005) 387–392. [28] M. Lecat, Tables Azeotropiques; Monograph, L’Auteur, Bruxelles, 1949. [29] H.-P. Luo, W.-D. Xiao, K.-H. Zhu, Fluid Phase Equilib. 175 (2000) 91–105. [30] X.-B. Ma, Z.-H. Li, Q. Xia, B.-W. Wang, Shiyou Huagong 30 (2001) 699–702. [31] A. Rodríguez, J. Canosa, A. Domínguez, J. Tojo, Fluid Phase Equilib. 201 (2002) 187–201. [32] L. Shi, H. Liu, K. Wang, W. Xiao, Y. Hu, Fluid Phase Equilib. 234 (2005) 1–10. [33] L.-Q. Zhang, X.-L. Zhu, M.-H. Zhu, Comp. Appl. Chem. 18 (2001) 285–286. [34] B. Chen, G. Li, Y. Zhang, Gaoxiao Huaxue Gongcheng Xuebao 11 (1997) 193–196. [35] G.A. Passoni, DE Patent 2,450,856 (1973). [36] V.M. Shah, C.R. Bartels, M. Pasternak, J. Reale Jr., AIChE Symp. Ser. 85 (1989) 93–97. [37] K. Nishihara, S. Yoshida, S. Tanaka, United States Patent 5,292,917 (1994).

181

[38] W. Himmele, K. Fischer, G. Kaibel, K. Schneider, R. Irnich, United States Patent 4,162,200 (1979). [39] G.X. Li, G.X. Xiong, Chem. Eng. (Chinese) 28 (2000) 12–13. [40] K.-Y. Hsu, Y.-C. Hsiao, I.-L. Chien, Ind. Eng. Chem. Res. 49 (2010) 735–749. [41] S.-J. Wang, C.-C. Yu, H.-H. Huang, Comp. Chem. Eng. 34 (2010) 361–373. [42] W. Won, X. Feng, D. Lawless, Sep. Purif. Technol. 31 (2003) 129–140. [43] J.H. Chen, Q.L. Liu, A.M. Zhu, Q.G. Zhang, J. Fang, J. Membr. Sci. 315 (2008) 74–81. [44] L. Wang, J. Li, Y. Lin, C. Chen, Chem. Eng. J. 146 (2009) 71–78. [45] I. Janisch, H. Landscheidt, W. Struver, A. Klausener, United States Patent 5,455,368 (1994). [46] J. Gmehling, C. Möllmann, Ind. Eng. Chem. Res. 37 (1998) 3112–3123. [47] D.W. Hill, M. van Winkle, Ind. Eng. Chem. 44 (1952) 208–210. [48] G.H. Eduljee, K.K. Tiwari, J. Chem. Eng. Jpn. 9 (1976) 319–321. [49] G.M. Wilson, J. Am. Chem. Soc. 86 (1964) 127–130. [50] H. Renon, J.M. Prausnitz, AIChE J. 14 (1968) 135–144. [51] U. Weidlich, J. Gmehling, Ind. Eng. Chem. Res. 26 (1987) 1372–1381. [52] J. Gmehling, J. Li, M. Schiller, Ind. Eng. Chem. Res. 32 (1993) 178–193. [53] J. Gmehling, J. Lohmann, A. Jakob, J. Li, R. Joh, Ind. Eng. Chem. Res. 37 (1998) 4876–4882. [54] J. Lohmann, J. Gmehling, J. Chem. Eng. Jpn. 34 (2001) 43–54. [55] J. Gmehling, R. Wittig, J. Lohmann, R. Joh, Ind. Eng. Chem. Res. 41 (2002) 1678–1688. [56] J.A. Riddick, W. Bunger, T.K. Sakano, Organic Solvents Physical Properties and Methods of Purification, 4th ed., John Wiley & Sons, New York, 1986. ˜ [57] E. Lladosa, J.B. Montón, M.C. Burguet, R. Munoz, Fluid Phase Equilib. 255 (2007) 62–69. [58] A. Tamir, J. Wisniak, J. Chem. Eng. Data 23 (1978) 293–298. [59] R. Malhotra, L.A. Woolf, J. Chem. Thermodyn. 28 (1996) 1411–1421. [60] B.E. Poling, J.M. Prausnitz, J.P. O’Connel, The Properties of Gases and Liquids, 5th ed., McGraw-Hill, New York, 2001. [61] A. Rodríguez, J. Canosa, A. Domínguez, J. Tojo, Fluid Phase Equilib. 198 (2002) 95–109. [62] N.F. Martínez, E. Lladosa, M.C. Burguet, J.B. Montón, M. Yazimon, Fluid Phase Equilib. 277 (2009) 49–54. [63] M. Broul, K. Hlavaty, J. Linek, Collect. Czech. Chem. Commun. 34 (1969) 3428–3435. [64] T. Hiaki, K. Yamato, K. Kojima, J. Chem. Eng. Data 37 (1992) 203–206. [65] K. Kurihara, T. Minoura, K. Takeda, K. Kojima, J. Chem. Eng. Data 40 (1995) 679–684. [66] K. Kurihara, T. Oshita, K. Ochi, K. Kojima, J. Chem. Eng. Data 48 (2003) 102–106. [67] K. Tochigi, C. Kikuchi, K. Kurihara, K. Ochi, J. Mizukado, K. Otake, J. Chem. Eng. Data 50 (2005) 784–787. [68] K. Tochigi, H. Takahara, Y. Shiga, Y. Kawase, Fluid Phase Equilib. 260 (2007) 65–69. [69] H. Matsuda, K. Yokoyama, H. Kyuzaki, K. Kurihara, K. Tochigi, J. Chem. Eng. Jpn. 43 (2010) 247–252. [70] C. Tsonopoulos, AIChE J. 20 (1974) 263–272. [71] C. Tsonopoulos, AIChE J. 21 (1975) 827–829. [72] H.C. Van Ness, S.M. Byer, R.E. Gibbs, AIChE J. 19 (1973) 238–244. [73] Aa. Fredenslund, J. Gmehling, P. Rasmussen, Vapor–Liquid Equilibria Using UNIFAC, Elsevier, Amsterdam, The Netherlands, 1977. [74] E.F.G. Herington, J. Inst. Petrol. 37 (1951) 457–470. [75] J. Gmehling, U. Onken, Vapor–Liquid Equilibrium Data Collection, Chemistry Data Series, DECHEMA, Frankfurt, starting 1977. [76] T. Hiaki, K. Tochigi, K. Kojima, Fluid Phase Equilib. 26 (1986) 83–102. [77] D.W. Marquardt, Soc. Ind. Appl. Math. 11 (1963) 431–441. [78] UNIFAC consortium, Oldenburg, 2011, http://www.UNIFAC.org. [79] L. Laroche, N. Bekiaris, H.W. Andersen, M. Morari, Can. J. Chem. Eng. 69 (1991) 1302–1319.