A simple relation to predict or to correlate the excess functions of multicomponent mixtures

• Published in: Fluid phase equilibria Vol. 62; no. 3; pp. 173 - 189
• Main Authors: Hwang, Chih-An, Holste, James C., Hall, Kenneth R., Ali Mansoori, G.
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 1991; Elsevier Science
• Subjects: Chemical thermodynamics; Chemistry; Exact sciences and technology; General and physical chemistry; Intermolecular interaction; Gibbs free energy; Multicomponent mixture; Thermodynamic activity; Activity coefficient; Theoretical study; Excess parameter; Chemical potential; Thermodynamic properties; Forecasting
• ISSN: 0378-3812; 1879-0224
• DOI: 10.1016/0378-3812(91)80009-K

Semi-theoretical relations for the excess functions (e.g. excess Gibbs energies G E, excess chemical potentials) developed previously for binary mixtures have been extended to multi-component mixtures. We postulate that contributions from two-body and three-body interactions are significant, and we propose an expression relating unlike three-body interactions to binary interactions. We have tested the relation with ternary vapor-liquid equilibria (VLE) having various chemical interactions and have found good agreement between the experimental excess functions and the predictions from the relation based solely upon binary data. Predicting fluid phase equilibria of multicomponent mixtures using existing binary data is relatively simple and, for the system tested, appears to be considerably better than the NRTL model for VLE systems having partially or wholly negative G E. For the systems tested having wholly positive G E, the new model (H 3M) is superior to the NRTL model except for ethanol-water. While the model is less satisfactory for liquid-liquid equilibria (LLE) than for VLE, it is significantly better than the NRTL or UNIQUAC model for the (randomly selected) system tested. The current model is also extremely flexible for either correlating or predicting multicomponent data, and it always converges to a solution.