376 Copyright © 2009 The Society of Chemical Engineers, Japan Journal of Chemical Engineering of Japan, Vol. 42, No. 6, pp. 376–380, 2009 Research Paper Recommended NRTL Model Parameters by Simultaneous Correlation of VLE, Infinite Dilution Activity Coefficients and Excess Enthalpy Data Katsumi TOCHIGI 1 , Jürgen RAREY 2 and Jürgen GMEHLING 2 1 Department of Materials and Applied Chemistry, Nihon University, 1-8, Surugadai, Kanda, Chiyoda-ku, Tokyo 101-8308, Japan 2 Department of Industrial Chemistry, University of Oldenburg, D-26111, Oldenburg, Germany Keywords: NRTL Parameter, VLE, Infinite Dilution Activity Coefficient, Excess Enthalpy This paper deals with the determination of recommended temperature dependent NRTL parameters by simultaneous correlation of VLE-, γ γ γ ∞ -, and h E data from the Dortmund Data Bank (DDB) using an interactive program package. Correlation results for selected mixtures are discussed in detail. The pro- gram package RECVAL is directly interfaced to the Dortmund Data Bank, makes extensive use of graphics for data representation and may be extended for the correlation and prediction of further mixture or pure component data. To ensure rapid conversion in case of a large amount of data, simultaneous con- vergence of the outer parameter fitting loop and the inner phase equilibrium iterations is employed. Recommended temperature dependent NRTL parameters have been determined for 133 binary systems. These parameters have been obtained by a simultaneous fit to VLE-, γ γ γ ∞ -, and h E -data stored in DDB. Introduction In order to represent the phase equilibrium behavior of multicomponent mixtures, g E models like Wilson (Wilson, 1964), NRTL (Renon and Prausnitz, 1968), UNIQUAC (Abrams and Prausnitz, 1975) and SILS (Tochigi and Kojima, 1983) are usually employed in the chemical industry. These models allow the prediction of the phase equilibrium behavior of multicomponent sys- tems using two or three binary interaction parameters for each binary combination of components in the mixture. The parameters may be derived either from experimental phase equilibrium data or estimated mixture data via com- mon regression algorithms. Recommended parameters published earlier have been regressed to fully consistent VLE data (Gmehling et al., 1977–2007) only. In order to represent the phase behavior over a wide range of com- position, temperature (pressure) with good accuracy, in addition to VLE data and γ ∞ values for an improved de- scription of the diluted concentration range should be used together with h E data, which supply information about the temperature dependence of the real mixture behavior via the Gibbs–Helmhotz equation. For this purpose, Rarey-Nies et al. (1988, 1989) had developed a program- package (RECVAL) interfaced directly to the Dortmund Data Bank (DDB) (Gmehling, 1985; Gmehling et al., 2008). Newer versions of this software also allow the in- clusion of liquid–liquid and solid–liquid equilibrium data as well as azeotropic compositions and excess heat ca- pacities. In the g E models, NRTL equation is easy to han- dle and has been widely used from practical point of view. This paper deals with the determination of rec- ommended temperature dependent NRTL parameters by a simultaneous correlation of VLE-, γ ∞ -, and h E data stored in the Dortmund Data Bank. 1. Temperature Dependent NRTL Model Param- eters In case of binary mixtures, activity coefficients can be calculated from the NRTL model using Eqs. (1)–(3). ln γ τ τ 1 2 2 21 21 1 2 21 2 12 12 2 1 12 2 1 = +       + + ( )         () x G x xG G x xG ln γ τ τ 2 1 2 12 12 2 1 12 2 21 21 1 2 21 2 2 = +       + + ( )         () x G x xG G x xG where τ τ 12 12 21 21 = = ( ) ∆ ∆ g RT g RT , 3a ln , ln G G 12 12 12 21 12 21 =− =− ( ) ατ ατ 3b Here ∆g 12 , ∆g 21 , α 12 are binary interaction parameters. Received on June 27, 2008; accepted on March 9, 2009. Cor- respondence concerning this article should be addressed to K. Tochigi (E-mail address: tochigi@chem.cst.nihon-u.ac.jp).