Prediction of liquid-liquid equilibria in ternary mixtures from binary data

Abstract

Liquid-liquid equilibria in type-I and type-II ternary mixtures are predicted from binary data alone by two methods: (1) Statistical Associating Fluid Theory (SAFT) which models molecular association based on the first-order perturbation theory developed by Wertheim; (2) a hybrid model in which a modified form of the Panagiotopoulos-Reid mixing rule is applied to the physical contribution terms of SAFT. The proposed mixing rule reduces to the original Panagiotopoulos-Reid mixing rule for binary mixtures while simultaneously satisfying the invariant condition when a component is divided into sub-components and hence is consistent for application to multicomponent systems. Although the Panagiotopoulos-Reid mixing rule provides accurate representation of phase equilibria in type-II ternary systems, the prediction of ternary phase behavior in type-I systems is not satisfactory. The association model, however, predicts phase behavior in both type-I and type-II ternary mixtures quite accurately. The significance of inclusion of hydrogen bond acceptor groups on aromatic rings, due to the presence of π-electrons, is elucidated in the representation of phase equilibria in binary and ternary mixtures.