UNIFAC group interaction prediction for ionic liquid-thiophene based systems using genetic algorithm

Abstract

The group interaction parameter prediction of Ionic Liquids(IL's) with thiophene (C4H4S) and other hydrocarbons are essential to generate (Liquid Liquid Equilibria) LLE through UNIFAC model. UNIFAC model is highly non-convex and can have several local extrema. In this work, the structural group interaction parameters have been calculated for [OMIM][BF 4] + thiophene + hydrocarbons and [OMIM] [BTI] + thiophene + hydrocarbons systems through regression using GA.The obtained LLE data has been correlated with reported values and it was observed that the cumulative RMSD(root mean square deviation) of ten ternary systems used for regression were 3.01% and 3.65% for [OMIM][BF4] and [OMIM]BTI] based system respectively. Further, the obtained interaction parameters were used to correlate the experimental LLE data for four ternary systems which were not used for regression. These systems having a total of 40 tie lines gave a very satisfactory RMSD of 1.76 to 3.99% between reported and predicted composition. © 2010 Springer-Verlag.