Temperature Effect on the Interaction Parameters of GC-NRTL Model Using the Genetic Algorithm

Liquid-liquid equilibrium data is imperatively needed for the design of separation processes, particularly solvent extraction. The experimental measuring approach is complex and costly for a great number of chemical systems, and hence the need for reliable predictive models. A great number of thermodynamic models have been developed and are reported in the literature. Among these the NRTL (Non Random Two Liquids) model has shown a great capability for predicting reliable liquid- liquid equilibrium data. However its major drawback is that it requires molecular interaction parameters which are not always available. Consequently, the aim of the present work is to present a new approach where the group contribution concept is incorporated into the NRTL equation, leading to the group contribution NRTL model (GC-NRTL). This GC-NRTL equation was tested by calculating liquid-liquid equilibrium data, particularly distribution coefficients, for a wide variety of liquid binary systems, at different temperatures. The required interaction parameters were calculated, minimizing an objective function using the genetic algorithm (GA) which generally leads to the global minimum. The effect of temperature on the interaction parameter values was also investigated, considering five different temperatures and ten different chemical systems. The agreement between experimental data and the predicted phase equilibrium results was encouraging. Correlations of binary group interaction parameters in term of temperature were proposed and recommended for predicting LLE by means of the GC-NRTL model which had proven to be reliable.