An algorithm for the regression of the UNIQUAC interaction parameters in liquid–liquid equilibrium for single- and multi-temperature experimental data

• An algorithm for the computation of the UNIQUAC parameters from LLE data is proposed.
• The algorithm checks the common tangent to the change of the Gibbs free energy of mixing.
• The problem is ill conditioned.
• The global minima of the total Gibbs free energy depend on the values of the model parameters.
• Only one set of parameters collocates the total Gibbs free energy at its lowest level.

This paper describes an algorithm for the computation of the UNIQUAC interaction parameters from liquid–liquid experimental data. The algorithm comprises two separate levels. The inner level is devoted to the calculation of the interaction parameters, minimizing an objective function which is function of the activities. The outer level uses the parameters by the inner level and aims to minimize the error between experimental and calculated molar fractions through an adjustment of the experimental molar fractions, provided that the condition of common tangent to the change of the Gibbs free energy of mixing is matched.The algorithm is applied to seven binary systems for the evaluation of single temperature parameters and to the tetrahydrofuran/water system for the evaluation of the temperature dependent parameters. In both cases, the calculated parameters provide results with lower error than the previously published parameters. In all cases thermodynamically consistent and precise results are obtained in terms of common tangent of the Gibbs free energy of mixing and molar faction errors.

Figure optionsDownload as PowerPoint slide