Fluid Phase Equilibria 313 (2012) 102106

Contents lists available at SciVerse ScienceDirect

Fluid Phase Equilibria

Salting effect on the liquidliquid equilibrium for the ternary system water + methanol + methyl acetate at 283.15 K
Xiaomei Chen, Xu Gao, Huidong Zheng , Suying Zhao , Liangen Wang
College of Chemistry and Chemical Engineering, Fuzhou University, Fuzhou 350108, China

a r t i c l e

i n f o

a b s t r a c t
Liquidliquid equilibrium (LLE) data for the ternary system of water + methanol (MeOH) + methyl acetate (MA) and the quaternary system of water + MeOH + MA + calcium chloride (CaCl2 ) have been measured at 283.15 K under atmospheric pressure. The liquid-phase splitting of water + MeOH + MA can be enhanced by the addition of CaCl2 . Separation factors at different CaCl2 concentrations were calculated and compared. In addition, the experimental data were correlated with the improved EisenJoffe equation. 2011 Elsevier B.V. All rights reserved.

Article history: Received 3 August 2011 Received in revised form 23 September 2011 Accepted 23 September 2011 Available online 1 October 2011 Keywords: Salting effect Liquidliquid equilibrium Methyl acetate Calcium chloride

1. Introduction In the production of polyvinyl alcohol (PVA), MA is produced as a byproduct at a ratio of 1.68 ton of MA per ton of PVA [1]. While the purity of MA is not high enough for industrial application because of containing MeOH and some other impurity, an effective and economical method needs to be developed for the separation and purication of MA from the alcoholysis liquid, which is industrially and environmentally signicant. Since there is an azeotrope of MeOH + MA with a minimum boiling point at 326.65 K and 0.6618 mol fraction MA in the vapor at 101.32 kPa [2], the mixture cannot be separated completely through common distillation. MeOH is removed from MA with the extractant of water by extractive distillation in industry production. However, the production contains about 6 wt% of water, which is not high enough in purity for industrial application and needs further purication and treatment. This method has some disadvantages, such as long process, complicated operation, and high energy consumption. Much attention should be paid on that, on one hand, liquid extraction plays an important role as a separation process in chemical engineering [3]. On the other hand, previous works [48] have demonstrated that the distribution coefcient and

Corresponding author. Tel.: +86 13905906759; fax: +86 59123502179. Corresponding author. Tel.: +86 13705061396. E-mail addresses: youngman@fzu.edu.cn (H. Zheng), zhaosuying@fzu.edu.cn (S. Zhao). 0378-3812/$ see front matter 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.uid.2011.09.025

selectivity increased with the addition of salt while the characteristics as extractant of the solvents improved considerably. Therefore saline extraction was put forward to improve the purication of MA. To understand and provide further information about the phase behavior by addition of salts, a large amount of investigations have been carried out in the recent years on LLE measurements. Pai and Rao [9] and Lin et al. [5] measured the mutual solubility of water + ethanol + ethyl acetate with a hydrophilic agent (glycerol) or an electrolyte compound (potassium acetate, or sodium chloride, or calcium chloride and so on). Rajendran and Srinivasan [10] determined the experimental LLE data of ethyl acetatewater in the presence of dissolved inorganic salts. Based on the research mentioned above and the theory of salting effect, the salt addition as extractant could considerably improve the characteristics of the solvent [11], including the saturated vapor pressure, the boiling point, and the solubility and so on. On the microcosmic side, the main functional mechanism of the salts in extraction process is that the salt ions form complex with the methanol molecules and the intermolecular hydrogen bonds are formed between the molecules of salts and water [12]. The mutual solubility of water + MeOH + MA with fourteen kinds of salts has been compared by Yang et al. [12], which has shown that the capability of CaCl2 for the elimination of MeOH and water was excellent. In this work, LLE data for the ternary system water(1) + MeOH(2) + MA(3) and the quaternary system water(1) + MeOH(2) + MA(3) + CaCl2 (4) at 283.15 K were reported. No literature data were found at comparable conditions. Separation factors at different CaCl2 concentrations were calculated and compared. These ternary and quaternary LLE data are correlated with

X. Chen et al. / Fluid Phase Equilibria 313 (2012) 102106

103

the improved EisenJoffe equation. In general, the model reproduces reasonably for the LLE phase envelops. 2. Experimental 2.1. Materials MeOH (99.99%), MA (99.95%) and hydrochloric acid (HCl, 36%) were purchased from the Sinopharm Chemical Reagent Co. Ltd. and used without any purication. CaCl2 (>96%), ethylene diamine tetraacetic acid disodium salt (EDTA, >99%), sodium hydroxide (NaOH, >96%), sodium chloride (NaCl, 99.5%), calcium carbonate (CaCO3 , 99%) and calcium carboxylic acid (C21 H14 N2 O7 S, 99.6%) were purchased from the Sinopharm Chemical Reagent Co. Ltd. and dried in an oven until the weight remained as a constant before being used. The purity of the organic compounds was checked with gas chromatographic analysis. Deionized distilled water was prepared in our laboratory. 2.2. Apparatus and procedure LLE data for the ternary and quaternary system were measured by an apparatus similar that of Zhao [13]. A jacketed equilibrium glass vessel (an intracavity volume of 40 cm3 ) was maintained at the constant temperature within 0.1 K by circulating water with a thermoelectric controller (model-CS501, Yuhua Instrument Co., China). Continuous stirring was achieved with a magnetic stirrer. A precision mercury-in-glass thermometer with an accuracy of 0.05 K was inserted into the inner chamber of the vessel for the measurement of the temperature of mixture. At the beginning of the experiment, water, MeOH, MA and CaCl2 were added to the cell by mass at known ratios. The weights of these reagents were determined by an analytical balance (Model-AE200, Mettler Toledo, Switzerland), with an accuracy of 1 104 g. The heterogeneous mixtures were agitated vigorously at least 2 h to sufciently mix and then settled at least 3 h for complete phase separation. The sample of organic-rich phase was carefully taken from the top sampling port of the cell with a syringe, and that of waterrich phase from a bottom sampling port of the cell. This operating method can avoid cross-contamination by the other phase during the sampling process. 2.3. Analysis The composition of MeOH and MA in both phase was analyzed by a gas chromatograph (Model-GC2014, SHIMADZU Co., Japan), which equipped with a hydrogen ame ionization detector and a capillary column Rtx -Wax (Restek Co., USA. 30m 0.25 mm ID 0.25 m). To prevent salt entering into the capillary column, a glass insert tube packed with glass wools was used as a lter and was connected at the entry of the column. The lter should be replaced when a signicant amount of salt was trapped. To obtain quantitative results for the organic phase, an internal standard method was used with the standard compound of ethanol. The area fraction was converted into mass fraction by the calibration equations. Two calibration lines were previously constructed with gravimetrically prepared samples in two composition ranges, in accordance with those in the organic-rich and water-rich phases. The deviations of the calibration curves from the actual values are tabulated in Table 1. The composition of water in organic-rich phase was determined by a Karl Fischer coulometer (model-KF756, Metrohm, Switzerland), which has an accuracy of 3 106 g. Water in the aqueous phase was determined from the material balance in that phase.

Table 1 Average deviations of GC calibration curves. Mixture MeOH + ethanol MA + ethanol

Phase Organic Aqueous Organic Aqueous

ARDa 0.0008 0.0001 0.0006 0.0004 , when N is the number of cal-

Average relative deviation =

ibration points and w is the mass fraction of the minor constituent compound. The superscripts cal and act represent the calibrated values and actual values, respectively.

The concentration of calcium ion (Ca2+ ) in each phase was determined by EDTA titration method. CaCO3 and HCl were used to prepare the standard solution for the EDTA compleximetry. Aqueous solution of NaOH was used to regulate pH value at about 14. Calcium carboxylic acid mixed with NaCl in the weight proportion of 1/99, was employed as indicator of the titration. The experimental mass fractions for each phase were obtained by averaging the results from ve replications. 3. Results and discussion 3.1. Experimental results The experimental apparatus was rstly employed to measure the LLE data for the ternary system of water(1) + MeOH(2) + MA(3) at 283.15 K under atmospheric pressure. Table 2 reports the experimental results. Since this ternary system behaves as a type-I LLE (water/MA is the only pair that is partially miscible), the effects of adding an auxiliary agent (an electrolyte) on the mutual solubility of water + MA can provide useful indication for the liquid-phase splitting enhancement. The electrolyte used in this study is CaCl2 . The LLE data of the quaternary system of water(1) + MeOH(2) + MA(3) + CaCl2 (4) have been measured at 283.15 K under atmospheric pressure while the composition of CaCl2 in the initial salt solution was 5 wt%, 10 wt%, 20 wt%, and 30 wt%, respectively. The experimental data are listed in Table 3, while the superscript I represents the organic phase and II the aqueous phase. The salts were found to be sparingly soluble in the organic phase, that is, most of the salt dissolves in the aqueous phase. The difference in behavior of the LLE data can be seen from the phase diagram shown in Fig. 1. The immiscibility gaps

0.0 1.0 0.2 0.4 0.8 0.6

w'

w 2'

0.6 0.8 1.0 0.0

0.4 0.2 0.0

0.2

0.4

w1'

0.6

0.8

1.0

Fig. 1. Comparison of LLE phase diagrams for water(1) + MeOH(2) + MA(3) + CaCl2 (4) at different mass fractions of CaCl2 and T = 283.15 K: ( ) ws = 0 (salt free); ( ) ws = 0.05; ( ) ws = 0.10; ( ) ws = 0.20; ( ) ws = 0.30.

104 Table 2 LLE data for water(1) + MeOH(2) + MA(3) at 283.15 K. Organic phase
I w1 I w2

X. Chen et al. / Fluid Phase Equilibria 313 (2012) 102106

Aqueous phase
I w3 II w1 II w2 II w3

0.0774 0.0776 0.0949 0.1080 0.1454 0.1582 0.1743 0.2312

0.0025 0.0054 0.0137 0.0222 0.0362 0.0418 0.0514 0.0642

0.9201 0.9170 0.8914 0.8698 0.8184 0.8000 0.7743 0.7046

0.7450 0.7296 0.6967 0.6378 0.5277 0.5069 0.4642 0.3996

0.0064 0.0149 0.0366 0.0564 0.0796 0.0826 0.0853 0.0827

0.2486 0.2555 0.2667 0.3058 0.3927 0.4105 0.4505 0.5177

are substantially enlarged as CaCl2 was introduced in the aqueous solutions. The areas of immiscibility increase with increasing concentration of CaCl2 . To estimate the capability of water for separating MeOH from MA at different CaCl2 concentrations, the distribution coefcients (Di ) and the separation factors (S23 ) were calculated as follows: Di = wiII wiI D2 D3 (1)

of MeOH (D2 ) increases while the distribution coefcient of MA (D3 ) decreases with the increasing of water in aqueous phase. The distribution coefcient of MeOH (D2 ) has an obvious ascend trend in the high concentration of CaCl2 . Fig. 4 shows the variation of the separation factor at different CaCl2 concentrations. It can be seen that the separation factor increases with the increasing of the salt concentration. The consistency of the equilibrium data was veried with the OthmerTobias equation and the Bachman equation on a salt-free basis [1417] the equations, depicted as follows: ln
II 1 w1 II w1

S23 =

(2)

where wiI represents the composition of component i in organic phase and wiII represents the composition of component i in aqueous phase. The distribution coefcient of MeOH (D2 ) and MA (D3 ) are shown in Figs. 2 and 3, respectively. The distribution coefcient
Table 3 LLE data for water(1) + MeOH(2) + MA(3) + CaCl2 (4) at 283.15 K. Organic phase
I w1 I w2 I w3 I w4

= n + m ln
II w1 I w3

I 1 w3 I w3

(3)

II w1 = n + m

(4)

Aqueous phase
II w1 II w2 II w3 II w4

5 wt% CaCl2 0.0541 0.0630 0.0785 0.0991 0.1190 0.1299 0.1491 0.1715 0.2258 10 wt% CaCl2 0.0479 0.0592 0.0653 0.0843 0.1085 0.1400 0.1579 0.1819 20 wt% CaCl2 0.0331 0.0418 0.0490 0.0567 0.0664 0.0728 0.0906 0.1034 30 wt% CaCl2 0.0141 0.0185 0.0234 0.0279 0.0379 0.0383 0.0520 0.0676

0.0027 0.0131 0.0252 0.0393 0.0535 0.0641 0.0766 0.0933 0.1113 0.0020 0.0164 0.0252 0.0480 0.0789 0.1094 0.1231 0.1387 0.0020 0.0218 0.0390 0.0562 0.0819 0.0967 0.1197 0.1455 0.0012 0.0140 0.0319 0.0530 0.0771 0.0855 0.1179 0.1405

0.9433 0.9238 0.8963 0.8616 0.8275 0.8060 0.7731 0.7323 0.6558 0.9501 0.9244 0.9095 0.8677 0.8126 0.7479 0.7125 0.6694 0.9649 0.9364 0.9120 0.8871 0.8518 0.8288 0.7845 0.7400 0.9847 0.9675 0.9447 0.9191 0.8844 0.8745 0.8258 0.7806

0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0012 0.0029 0.0071 0.0000 0.0000 0.0000 0.0000 0.0000 0.0027 0.0065 0.0100 0.0000 0.0000 0.0000 0.0000 0.0000 0.0017 0.0052 0.0111 0.0000 0.0000 0.0000 0.0000 0.0007 0.0017 0.0044 0.0113

0.7828 0.7473 0.6915 0.6562 0.6164 0.5775 0.5214 0.4679 0.3972 0.7722 0.7267 0.6897 0.6363 0.5617 0.4930 0.4113 0.3410 0.7339 0.6715 0.6171 0.5665 0.5164 0.4526 0.3933 0.3218 0.6518 0.6075 0.5394 0.4677 0.4273 0.3645 0.3282 0.2363

0.0088 0.0372 0.0669 0.0888 0.1085 0.1185 0.1335 0.1433 0.1442 0.0071 0.0528 0.0730 0.1145 0.1578 0.1833 0.1878 0.1977 0.0069 0.0638 0.1080 0.1461 0.1793 0.2008 0.2199 0.2300 0.0064 0.0524 0.1043 0.1708 0.2097 0.2435 0.2512 0.2666

0.1674 0.1757 0.2034 0.2171 0.2401 0.2704 0.3133 0.3588 0.4341 0.1311 0.1359 0.1496 0.1700 0.2061 0.2583 0.3492 0.4127 0.0717 0.0919 0.1068 0.1374 0.1617 0.2166 0.2742 0.3555 0.0565 0.0705 0.1040 0.1354 0.1715 0.2180 0.2760 0.3814

0.0411 0.0398 0.0382 0.0379 0.0349 0.0336 0.0319 0.0300 0.0246 0.0895 0.0845 0.0877 0.0791 0.0744 0.0655 0.0517 0.0487 0.1875 0.1729 0.1682 0.1501 0.1426 0.1300 0.1126 0.0926 0.2853 0.2695 0.2523 0.2260 0.1914 0.1741 0.1446 0.1158

X. Chen et al. / Fluid Phase Equilibria 313 (2012) 102106 Table 4 Constants of the OthmerTobias equation and the Bachman equation. ws OthmerTobias n 0 0.05 0.10 0.20 0.30 1.3023 0.8966 1.1348 1.8864 2.3953 m 0.9729 0.868 0.9933 1.2514 1.1594 R
2

105

Bachman n 0.3627 0.5916 0.401 0.2237 0.1188 m 1.3672 1.6192 1.3788 1.191 1.104 R2 0.9965 0.9938 0.9833 0.9973 0.9987

0.9923 0.9898 0.9819 0.9966 0.9912

Table 5 Parameters of the EisenJoffe equation for water(1) + MeOH(2) + MA(3) + CaCl2 (4) at different CaCl2 concentrations. ws 0 0.05 0.10 0.20 0.30 a 0.4817 0.0310 0.0083 0.0038 0.0094 b c 1.0259 0.0255 0.0113 0.0012 0.0008 d R2 0.9959 0.9974 0.9964 0.9995 0.9964 ARD 0.0178 0.0153 0.0273 0.0096 0.0284

II where w1 is the mass fraction of water in the aqueous phase and I w3 is the mass fraction of MA in the organic phase. The parameters of the equations are listed in Table 4. All of the squares of the liner correlation coefcients (R2 ) are greater than 0.98, which suggests that it is reasonable to accept the LLE data of the considered systems as reliable.

D2

3.2. Data correlations To extend the application to salt solution systems, Rajendran [18,19] somewhat improved the EisenJoffe equation [20]. The revised equation is as follows: log
1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
I w2 I w3

= (a + bws ) + (c + dws )log

II w2 II w1

(5)

w1
Fig. 2. Distribution coefcient of MeOH (D2 ) at different mass fractions of CaCl2 and T = 283.15 K: ( ) ws = 0 (salt free); ( ) ws = 0.05; ( ) ws = 0.10; ( ) ws = 0.20; ( ) ws = 0.30.
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.2

I I where w2 and w3 represent the mass fraction of MeOH and MA II II in the organic phase, respectively, w1 and w2 represent the mass fraction of water and MeOH in the aqueous phase, respectively, ws represents the mass fraction of CaCl2 in the initial salt solution. The letters a, b, c and d are parameters of the EisenJoffe equation, which depend upon the nature of the system components, including that

100 80 60

D3

S23
40 20 0 0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.4

0.6

0.8

w1
Fig. 3. Distribution coefcient of MA (D3 ) at different mass fractions of CaCl2 and T = 283.15 K: ( ) ws = 0 (salt free); ( ) ws = 0.05; ( ) ws = 0.10; ( ) ws = 0.20; ( ) ws = 0.30.

w1
Fig. 4. Separation factor of MeOH/MA at different mass fractions of CaCl2 and T = 283.15 K: ( ) ws = 0 (salt free); ( ) ws = 0.05; ( ) ws = 0.10; ( ) ws = 0.20; ( ) ws = 0.30.

106

X. Chen et al. / Fluid Phase Equilibria 313 (2012) 102106

of the salt and temperature. Furthermore, b and d are relevant to the salt concentration, while a and c are not. Parameters of the improved EisenJoffe equation for water(1) + MeOH(2) + MA(3) + CaCl2 (4) are listed in Table 5. All linear correlation coefcients (R2 ) are greater than 0.99, and average relative deviations (ARD) are less than 0.0284. A good agreement can be observed for all the data. 4. Conclusions LLE data for the ternary system of water(1) + MeOH(2) + MA(3) and the quaternary system of water(1) + MeOH(2) + MA(3) + CaCl2 (4) at 283.15 K have been measured under atmospheric pressure. All the investigated systems formed a type-I phase diagram of LLE. The results showed that the two phase region and separation factor increased with the increasing concentration of CaCl2 . The consistency of the results was tested with OthmerTobias equation and the Bachman equation on a salt-free basis. The experimental data could be correlated well with a modication of the EisenJoffe equation. The results were very satisfactory in accordance between experimental and calculated compositions of both equilibrium phases, while the average relative deviations were always below 2.8%. The concentration of MA in the aqueous phase decreased and the immiscibility regions enlarged with the addition of CaCl2 , while it can be concluded that CaCl2 can be used as a saline extractant to recycle the alcoholysis liquid.

Acknowledgement Financial support from the Research Start-up Fund of Fuzhou University is gratefully acknowledged. References
[1] C.B. Dirk-Faitakis, W.Z. An, T.B. Li, K.T. Chuang, Chem. Eng. Process. 48 (2009) 10801087. [2] J. Gmehling, R. Blts, J. Chem. Eng. Data 41 (1996) 202209. [3] G.H. Bochove, G.J. Krooshof, T.W. de Loos, Fluid Phase Equilibr. 171 (2000) 4558. [4] V. Gomis, F. Ruiz, N. Boluda, Ind. Eng. Chem. Res. 37 (1998) 599603. [5] H.M. Lin, C.E. Yeh, G.B. Hong, M.J. Lee, Fluid Phase Equilibr. 237 (2005) 2130. [6] J.T. Chen, M.C. Chen, Fluid Phase Equilibr. 266 (2008) 17. [7] J.T. Chen, C.D. Lin, J. Chem. Eng. Data 54 (2009) 12201224. [8] M.A. Pereira, M. Aznar, J. Chem. Thermodyn. 38 (2006) 8489. [9] M.U. Pai, K.M. Rao, J. Chem. Eng. Data 11 (1966) 353356. [10] M. Rajendran, D. Srinivasan, Fluid Phase Equilibr. 44 (1988) 5357. [11] V. Gomis, F. Ruiz, N. Boluda, M.D. Saquete, Ind. Eng. Chem. Res. 37 (1998) 599603. [12] D.J. Yang, H.M. Lou, Y.X. Pang, X.Q. Qiu, J. Chem. Eng. Chin. Univ. 21 (2007) 239244 (in Chinese). [13] S.Y. Zhao, H.J. Chen, X.M. Chen, J.Y. Zhou, L.E. Wang, J. Chem. Eng. Data 55 (2010) 52765279. [14] D.F. Othmer, P.E. Tobias, Ind. Eng. Chem. 34 (1942) 693700. [15] S. Lee, H. Kim, J. Chem. Eng. Data 43 (1998) 358361. [16] G.S. Ladha, T.E. Degaleesan, Transport Phenomena in Liquid Extraction, McGraw-Hill, New York, 1978. [17] G. Saravanan, D. Srinivasan, J. Chem. Eng. Data 30 (1985) 166171. [18] M. Rajendran, D. Srinivasan, Fluid Phase Equilibr. 44 (1988) 5375. [19] M. Rajendran, S. Renganarayanan, D. Srinivasan, Fluid Phase Equilibr. 50 (1989) 133164. [20] E.O. Eisen, J. Joffe, J. Chem. Eng. Data 11 (1966) 480484.