Research Journal of Chemical Sciences ______________________________________________ ISSN 2231-606X Vol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 64 Vapor-Liquid Equilibrium Data Prediction by Advanced Group Contribution Methods for a Binary System of Cyclopentyl methyl ether and Acetic acid at Atmospheric PressureParsana V.M. 1,2,* and Parikh S.P.School of Engineering, RK University, Rajkot, INDIAChemical Engineering Department, V.V.P. Engineering College, Gujarat Technological University, Ahmedabad, INDIAAvailable online at: www.isca.in, www.isca.me Received 7th June 2015, revised 13th June 2015, accepted 17th June 2015 AbstractThe isobaric vapour-liquid equilibrium data predictions for the binary system of cyclopentyl methyl ether and acetic acid have been obtained using UNIFAC method and modified UNIFAC Dortmund method. Group identification is done with Dortmund Data Bank and the same has been confirmed by artist free software. The interaction parameters in the UNIFAC method and modified UNIFAC Dortmund method, for the ether group (-CHO) and acid (-COOH), are used to predict VLE data. Thermodynamic consistency of the predicted VLE data has been checked by the Herington method. The predicted data have been correlated with Van Laar, Wilson and NRTL activity coefficient models. The binary interaction parameters of models had been obtained by regression. The predicted VLE data of UNIFAC method fit much more accurately than that of modified UNIFAC Dortmund method by these activity coefficient models Van Laar, Wilson and NRTL. Keywords: Vapor-liquid equilibrium, cyclopentyl methyl ether, acetic acid, UNIFAC method, modified UNIFAC Dortmund method. Introduction Traditional solvents play a key role in the chemical process industries and hence they are major contributor to the concerns related to their impact on environment, health and safety because most of the solvents are flammable, volatile and toxic. The solvents that have reduced or no toxicity to health and environment compared to the traditional solvents are called green solvents. These green solvents may provide an attractive alternative to the traditional solvents. Cyclopentyl methyl ether (CPME) is considered to be one of the green solvents which has high boiling point (379.15 K) and preferable characteristics such as low peroxide formation, high hydrophobicity, relative stability under acidic and basic conditions, high boiling point and low melting point, low heat of vaporization, narrow explosion area and low solubility of salts. Due to such characteristics CPME is preferred as an alternative to other ethereal solvents such as tetrahydrofuran, 2-methyl tetrahydrofuran, dioxane (carcinogenic), and 1, 2-dimethoxyethane, which are hazardous to human health and environment. The recovery of acetic acid from water has become industrial problem of public concern because this separation process has a major influence on economy of products, resource utilization and important meanings for environmental protection. Hongxun Zhang, Guangyu Liu, Chen Li, et al. measured liquid-liquid equilibria of water + acetic acid + Cyclopentyl methyl ether (CPME) system at different temperatures and concluded that CPME would be a good substitute for conventional organic solvents to separate acetic acid from water by the method of liquid liquid extraction followed by heteroazeotropic distillation. After literature survey it is found that vapour-liquid equilibrium data for CPME + Acetic acid system which is essential for the design of distillation column for separating CPME and acetic acid from their mixture does not exist in the literature. So investigation on VLE data of this binary system becomes necessary. The experimental determination of VLE data requires sophisiticated and suitable VLE apparatus and composition measurement instruments such as gas chromatograph, refractometer, spectrophotometer, etc. So this procedure is very costly and time-consuming. Numerical simulations using group contribution methods provide an alternative to experimental measurement of VLE data. The aim of this paper is to predict VLE data for CPME with acetic acid at atmospheric pressure. Advanced Group Contribution Methods: Reliable values of the properties of materials are necessary for the design of industrial processes. The knowledge of physical properties of fluids is essential in the design of many kinds of products, processes, and industrial equipment. The vapour-liquid equilibrium (VLE) data are essential for the design of chemical and separation processes. When experimental binary data are available, phase equilibrium behaviour is easily modelled with the help of cubic equation of state (using fugacity coefficient data) and local composition g models (using activity coefficient data). When little or no experimental data are available, group Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 65 contribution (GC) methods can be employed to predict the phase equilibrium under specified conditions of temperature and composition5,6. So prediction of thermodynamic properties is important in chemical process and product design. Various GC methods are available for the prediction of VLE data. Some examples of GC methods which have been developed for the estimation of properties of pure compounds include those published by Joback and Reid, Lydersen, Ambrose, Constantinou and Gani10 and Marrero and Gani11,12. On the other hand, many GC based property models have also been developed to predict properties of mixtures mainly to predict the non-ideality of the liquid phase using activity coefficients which includes ASOG13,14, Original UNIFAC15, Modified UNIFAC Dortmund16 and PSRK17. In the present work, well known and established group-contribution methods such as UNIFAC method and modified UNIFAC Dortmund method are employed to predict liquid phase activity coefficients for binary mixtures of CPME and acetic acid.UNIFAC method and modified UNIFAC Dortmund methods: The general UNIFAC equation is as follows with the combinatorial and residual contributions: ()lnlnlnresidualialcombinator(1) The combinatorial part,  F -+F q + F ialcombinator(ilnlnln (2) Where, ()() 2 (3) (4) (5) )i(  (6) )i(  (7) 17 . 15 wk(8) wk 10 x 5 . 2 (9) Where: )i(, always an integer, is the number of groups of type k in molecule i. Group parameters R and Q are obtained from the van der Waals group volume and surface areas Vwk and Awk, given by Bondi. The value of parameter Z is taken as 10. And the residual part, ( ) )i()i(residual(ilnlnln  (10) Where, nmkmmklnln (11) Where, the group area fraction q and group mole fraction Xare given by the following equations: (12)   jn)j()j( (13) Where, the group-interaction parameter mn Y is given by the following equation: RTmnnnmnmnexpexp (14) Where: Umn is a measure of the energy of interaction between group m and n. Note that amn has unit of Kelvin and amn anm. In the Original UNIFAC model, the interaction parameters are considered to be independent of temperature. Therefore, quantitative predictions of excess enthalpies, H could not be obtained. In order to improve this and other things, the modified UNIFAC Dortmund method was developed. The usage of modified UNIFAC Dortmund method leads to much better results. This means that the introduction of temperature-dependent parameters allows a more reliable temperature extrapolation and the extension of the range of applicability18,19. In both UNIFAC method and modified UNIFAC Dortmund method, there is a difference in both combinatorial and residual part. These differences are given in the following equations. In modified UNIFAC Dortmund method, equation-2 and equation-14 of UNIFAC method are replaced by equation-15 Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 66 and equation-19 as described below. ialcombinator(ilnq5lnln (15) Where, (16) (17) (18) and mnmnmnmnexp (19) In addition to that, in the residual part, temperature dependent interaction parameters are used where they have a logarithmic and quadratic dependency towards temperature. Due to this temperature dependency, the predictions of VLE, H and have improved since it is based on more experimental data. The modified UNIFAC Dortmund method can also extrapolate reliably the predictions of VLE at higher temperatures compared to the Original UNIFAC19. Group identification of the compounds: Group contribution methods predict properties of pure compounds or mixtures based on the groups existing in the compounds so correct identification of groups is very essential. Group identification for UNIFAC method is done using the data given in the literature and for modified UNIFAC Dortmund method it is done using the data given in the literature18. The identified groups are presented in table-1 and table-2 respectively. The identified groups have been verified with Dortmund data bank by using artist free software. Binary interaction parameters (BIPs): Binary interaction parameters (amn) for UNIFAC method have been taken from the literature and (amn, bmn, and cmn) for modified UNIFAC Dortmund method have been taken from the literature18,19,20 which are presented in table-3 and table-4 respectively. Calculation of VLE data using group contribution methods: The VLE data for binary system CPME and acetic acid are calculated through a spread sheet in which temperature T and are given as input and g and g are calculated using group contribution methods as described in the precious sections. Using Antoine equation-20, satand satare calculated, then total pressure P is calculated and correct temperature T is found out by regression using equation-23. The calculated data are presented in table-5 and table-6 for UNIFAC method and modified UNIFAC Dortmund method respectively. Table-1 Group identification for CPME and acetic acid for UNIFAC method Molecule (i) Name Group no.(i) Q M S CPME (1) CH 2 1 2 4 0.6744 0.540 CH 1 3 1 0.4469 0.228 CH 3 O 13 24 1 1.1450 1.088 Acetic acid (2) CH 3 1 1 1 0.9011 0.848 COOH 20 42 1 1.3013 1.224 M=Main Group no., S=Secondary Group no. Table-2 Group identification for CPME and acetic acid for modified UNIFAC Dortmund method Molecule (i) Name Group no.(i) Q M S CPME (1) c-CH 2 42 78 4 0.7136 0.8635 c-CH 42 79 1 0.3479 0.1071 CH 3 O 13 24 1 1.1434 1.6022 Acetic acid (2) CH 3 1 1 1 0.6325 1.0608 COOH 20 42 1 0.8000 0.9215 M=Main Group no., S=Secondary Group no. Table-3 BIPs for CPME and acetic acid for UNIFAC method Group CH 3 CH 2 CH CH 3 O COOH CH 3 0.0 0.0 0.0 251.5 663.5 CH 2 0.0 0.0 0.0 251.5 663.5 CH 0.0 0.0 0.0 251.5 663.5 CH 3 O 83.36 83.36 83.36 0.0 664.6 COOH 315.3 315.3 315.3 -338.5 0.0 The Antoine equation, satln (20) Where pressure is in kPa and temperature is in Kelvin21. The constants A, B, and C of Antoine equations of CPME and acetic acid are listed in table-722. Thermodynamic Consistency Test: The thermodynamic consistency of the predicted VLE data for the binary system is checked by semi-empirical Herington method. In this method, the values for D and J are found out by equation-21 and equation-22 respectively. If the value of D – J is not larger than 10 then the predicted VLE data are said to be thermodynamically consistent. The values of D – J for the Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 67 binary system are listed in table-8. dxlndxln100 (21) minminmax150 - (22) Data Reduction Using g Models: The predicted VLE data are correlated by various models such as Van Laar, Wilson and NRTL24,25. The vapor pressures of pure components are calculated by equation-20. By the minimization of the objective function %AAD P), the binry interaction parameters are obtained for these models which are used to minimize error by the regression procedure. (%AAD = absolute average deviation and n represents no. of predicted data points). Similarly AAD T) and AAD y) are calculated by equation-24 and equation-25 respectively. The “pre” and “cal” subscripts represent the predicted and calculated values respectively. %AAD P) pre,ical,ipre,i100 (23) AAD T) cal,ipre,i (24) AAD y) cal,ipre,i (25) Table-4 BIPs for CPME and acetic acid for modified UNIFAC Dortmund method Group m n a mn b mn c mn a nm b nm c nm CH 3 CH 3 O 1 13 -9.654 -0.03242 233.1 -0.3155 0 0 CH 3 COOH 1 20 2017.7 -9.0933 1182.2 -3.2647 0.009198 0.01024 CH 3 c-CH 2 , c-CH 1 42 1020.8 -6.0746 -680.95 4.0194 -0.006878 0.01015 CH 3 O c-CH 13 20 -310.82 0 521.48 0 0 0 CH 3 O c-CH 2 , c-CH 13 42 -86.6 0.9724 251.4 -1.021 0 0 COOHc-CH 2 , c-CH 20 42 582.81 1.4976 1169.3 -3.0737 0 0 Table-5 VLE data for CPME and acetic acid binary system at atmospheric pressure by UNIFAC method T/K x 1 y 1  1  2 379.15 1.0000 1.0000 1.0000 --- 378.57 0.9500 0.9367 1.0025 1.8825 378.24 0.9000 0.8850 1.0094 1.7298 378.07 0.8500 0.8406 1.0200 1.6074 378.03 0.8000 0.8009 1.0339 1.5079 378.08 0.7500 0.7643 1.0508 1.4258 378.21 0.7000 0.7296 1.0706 1.3571 378.42 0.6500 0.6960 1.0934 1.2990 378.69 0.6000 0.6628 1.1192 1.2494 379.03 0.5500 0.6295 1.1484 1.2067 379.44 0.5000 0.5956 1.1813 1.1697 379.92 0.4500 0.5605 1.2185 1.1375 380.49 0.4000 0.5238 1.2606 1.1094 381.14 0.3500 0.4847 1.3087 1.0848 381.90 0.3000 0.4424 1.3640 1.0635 382.80 0.2500 0.3958 1.4282 1.0453 383.85 0.2000 0.3432 1.5036 1.0299 385.11 0.1500 0.2824 1.5930 1.0175 386.66 0.1000 0.2097 1.7007 1.0081 388.60 0.0500 0.1191 1.8324 1.0021 391.12 0.0000 0.0000 --- 1.0000 Table-6 VLE data for CPME and acetic acid binary system at atmospheric pressure by modified UNIFAC Dortmund method T/K x 1 y 1  1  2 379.15 1.0000 1.0000 1.0000 1.6644 378.97 0.9500 0.9465 1.0015 1.5714 378.90 0.9000 0.8987 1.0058 1.4910 378.92 0.8500 0.8551 1.0126 1.4216 379.01 0.8000 0.8143 1.0219 1.3613 379.18 0.7500 0.7757 1.0334 1.3090 379.40 0.7000 0.7383 1.0471 1.2634 379.68 0.6500 0.7015 1.0630 1.2236 380.02 0.6000 0.6649 1.0810 1.1888 380.42 0.5500 0.6279 1.1012 1.1583 380.88 0.5000 0.5901 1.1237 1.1 314 381.40 0.4500 0.5510 1.1488 1.1078 381.99 0.4000 0.5102 1.1769 1.0870 382.66 0.3500 0.4670 1.2085 1.0686 383.41 0.3000 0.4209 1.2442 1.0524 384.26 0.2500 0.3710 1.2854 1.0383 385.23 0.2000 0.3163 1.3336 1.0261 386.35 0.1500 0.2552 1.3912 1.0158 387.65 0.1000 0.1852 1.4619 1.0077 389.20 0.0500 0.1025 1.5510 1.0021 391.12 0.0000 0.0000 1.6667 1.0000 Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 68 Table-7 Antoine equation constants Compound Antoine constants temperature range/K A B C CPME 15.0255 3798.52 -14.2 357 to 395 AA 15.0694 3580.79 -48.5 375 to 395 Table-8 Thermodynamic consistency check D J D-J Method 22.01 4.74 17.27 UNIFAC 26.46 4.74 21.72 modified UNIFAC Dortmund The binary interaction parameters, correlated from predicted VLE data by UNIFAC method and modified UNIFAC Dortmund method, are shown in table-9 and table-10 respectively. which is a characteristic constant of the non-randomness for the binary system is recommended as 0.3 for this binary system because it belongs to type I system according to the definition given in the literature25. The comparison of predicted data by UNIFAC method and modified UNIFAC Dortmund method with calculated T-x-y data by Van Laar, Wilson, and NRTL models for the binary system CPME (1) + acetic acid (2) at atmospheric pressure is given through figure-1 to figure-6. Table-9 Correlated model BIPs from predicted data by UNIFAC method Model Binary Parameter AAD T) AAD y) Van Laar 12 A 21 0.0601 0.0023 0.6358 0.6760 Wilson 12 a 21 0.0564 0.0020 -1253.267 3534.491275 NRTL 12 b 21 0.0648 0.0023 1311.714 883.945 Table-10 Correlated model BIPs from predicted data by modified UNIFAC Dortmund method Model Binary Parameter AAD T) AAD y) Van Laar 12 A 21 0.0399 0.0021 0.4714 0.4819 Wilson 12 a 21 0.0363 0.0018 -1428.309 3044.001492 NRTL 12 b 21 0.0394 0.0019 778.801 795.580 Figure-1 T-x-y diagram calculated by Van Laar and predicted by UNIFAC method Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 69 Figure-2 T-x-y diagram calculated by Wilson and predicted by UNIFAC method Figure-3 T-x-y diagram calculated by NRTL and predicted by UNIFAC method Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 70 Figure-4 T-x-y diagram calculated by Van Laar and predicted by modified UNIFAC Dortmund method Figure-5 T-x-y diagram calculated by Wilson and predicted by modified UNIFAC Dortmund method Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 71 Figure-6 T-x-y diagram calculated by NRTL and predicted by modified UNIFAC Dortmund method From figure-1 to figure-6, it can be seen that isobaric VLE data predicted by UNIFAC method and modified UNIFAC Dortmund method are very well represented by Van Laar, Wilson and NRTL models. Conclusion The VLE data for the binary system CPME with acetic acid have been predicted at atmospheric pressure using UNIFAC method and modified UNIFAC Dortmund method. The activity coefficient models Van Laar, Wilson and NRTL have been found capable of accurately fitting the predicted VLE data by UNIFAC method and modified UNIFAC Dortmund method. However, they fail the consistency test by Herington. Azeotrope formation is found for this system. Nomenclature P - Absolute pressure, kPa T - Absolute temperature, K q - Surface area fraction of compound i F - Volume fraction of compound i - Relative Van der Waals volume of compound i - Relative Van der Waals surface area of compound i - Relative Van der Waals surface area of subgroup k - Relative Van der Waals volume of subgroup k G - Temperature dependant integration constant q - Surface area fraction of subgroup m - Mole fraction of subgroup m Y - Group-interaction parameter - Volume/mole fraction of compound i in the mixture - Surface area fraction of compound i in the mixture - Modified volume/mole fraction of compound i in the mixture (modified UNIFAC Dortmund method) ln - Natural logarithm (base e) log - Logarithm (base 10) - Liquid phase mole fraction of ith species - Vapor phase mole fraction of ith species g - Activity coefficient of ith species - Excess enthalpy ij - Adjustable parameter (Van Laar Model) ij l - Interaction parameter (Wilson Model) ij L - Adjustable parameter (Wilson Model) ij a -The non-randomness of the fluid empirical parameter ij t - Adjustable parameter (NRTL Model) A, B, C- Antoine equation constants Superscripts E - Excess property sat - Saturated property value ¥ - Property at infinite dilution concentration Subscripts 1 - Component 1 Research Journal of Chemical Sciences ___________________________________________________________ ISSN 2231-606XVol. 5(6), 64-72, June (2015) Res. J. Chem. Sci. International Science Congress Association 72 2 - Component 2 i - Property of ith species j - Property of jth species References 1.Kiyoshi W., The Toxicological Assessment of Cyclopentyl Methyl Ether (CPME) as a Green Solvent, Molecules,18(3), 3183-3194 (2013)2.Kiyoshi W., Noriyuki Y. and Yasuhiro T., Cyclopentyl Methyl Ether as a New and Alternative Process Solvent, Organic Process Research and Development,11(2), 251-258 (2007)3.Zhang H., Liu G., Li C. and Zhang L., Liquid Liquid Equilibria of Water + Acetic Acid + Cyclopentyl Methyl Ether (CPME) System at Different Temperatures, Journal of Chemical and Engineering Data,57(11), 2942-2946 (2012)4.Poling B.E., Prausnitz J.M. and O'Connell J.P., The Properties of Gases and Liquids, 5th ed., The McGraw-Hill Company limited, New York, pp-8.23 (2012)5.Gmehling J., From UNIFAC to Modified UNIFAC to PSRK with the Help of DDB, Fluid Phase Equilibria,107(1), 1-29 (1995)6.3rd International Conference on Medical Sciences and Chemical Engineering (ICMSCE'2013), Bangkok (Thailand), Dec. 25-26, (2013)7.Joback K.G. and Reid R.C., Estimation of Pure-Component Properties from Group-Contributions, Chem. Eng. Commun.,57(1-6), 233 -243 (1987)8.Lydersen A.L., Estimation of Critical Properties of Organic Compounds, College Engineering University Wisconsin, Engineering Experimental Station Report 3, Madison, WI, April (1955)9.Ambrose D., Correlation and Estimation of Vapor-Liquid Critical Properties. I. Critical Temperatures of Organic Compounds, National Physical Laboratory, Teddington, UK, NPL Report Chem., 92, September (1978)10.Gani R. and Constantinou L., Molecular Structure Based Estimation of Properties for Process Design, Fluid Phase Equilibria,116(1-2), 75- 86 (1996)11.Marrero J. and Gani R., Group-Contribution Based Estimation of Pure Component Properties, Fluid Phase Equilibria,183-184, 183-208 (2001)12.Hukkerikar A.S., Sarup B., Kate A.T., Abildskov J., Sin G. and Gani R., Group-Contribution+ (GC+) Based Estimation of Properties of Pure Components: Improved Property Estimation and Uncertainty Analysis, Fluid Phase Equilibria,321, 25-43 (2012)13.Derr E.L. and Deal C.H., Analytical Solutions of Groups: Correlation of Activity Coefficients Through Structural Group Parameters. 1, Chem. E. Symp., Ser. No. 32, Instn. Chem. Engrs., London, , 88 (1969)14.Ronc M. and Ratcliff G.A., Prediction of Excess Free Energies of Liquid Mixtures by an Analytic Group Solution Model, Can. J. Chem. Eng.,49(6), 825-830 (1971)15.Fredenslund A., Jones R.L. and Prausnitz J.M., Group Contribution Estimation of Activity Coefficients in Nonideal Liquid Mixtures, AIChE Journal,21(6), 1086-1099 (1975)16.Weidlich U. and Gmehling J., A Modified UNIFAC Model. 1- Prediction of VLE, h, and , Ind. Eng. Chem. Res., 26(7), 1372-1381 (1987)17.Holderbaum T. and Gmehling J., PSRK: A Group Contribution Equation of State Based on UNIFAC, Fluid Phase Equilibria,70(2-3), 251-265 (1991)18.Gmehling J., Li J. and Schiller M., A Modified UNIFAC Model. 2. Present Parameter Matrix and Results for Different Thermodynamic Properties, Ind. Eng. Chem. Res., 32(1), 178-193 (1993)19.Lohmann J., Joh R. and Gmehling J., From UNIFAC to Modified UNIFAC (Dortmund), Ind. Eng. Chem. Res.,40(3), 957-964 (2001)20.Jakob A., Grensemann H., Lohmann J. and Gmehling J., Further Development of Modified UNIFAC (Dortmund): Revision and Extension 5, Ind. Eng. Chem. Res., 45(23), 7924-7933 (2006)21.Smith J.M., VanNess H.C. and Abbott M.M., Introduction to chemical engineering thermodynamics, 7th ed., Tata McGraw Hill Education private limited, New Delhi, (2010)22.Modi C.K., M.Tech. Thesis, Determination Of VLE Data For System Containing CPME, Institute of Technology, Nirma University, May (2014)23.Wisniak J.A., The Herington test for thermodynamic consistency, Ind. Eng. Chem. Res., 33(1), 177-180 (1994)24.Wilson G.M., A New Expression for the Excess Free Energy of Mixing, J. Amer. Chem. Soc.,86(2), 127-130 (1964)25.Renon H. and Prausnitz J.M., Local compositions in thermodynamic excess functions for liquid mixtures, AIChE Journal, 14(1), 135-144 (1968)26.Gadhiya P.M., Parsana V.M., Parikh S.P. and Joshipura M.H., Vapor-Liquid Equilibrium Data Prediction by Advanced Group Contribution Methods for a Binary System of Cyclopentyl Methyl Ether and Cyclopentanol at Atmospheric Pressure, International Journal of Advance Engineering and Research Development,2(2), 260-269 (2015)