Systematic screening ionic liquid as extractant for benzene-cyclohexane separation

In the separation of aromatic-alkane systems, ionic liquids (ILs) demonstrate superior performance over organic solvents. In this work, a rational screening strategy is proposed for tailoring ILs for the benzene-cyclohexane extraction. The reliability of COSMO-RS and UNIFAC-IL for this system are first assessed by comparing experimentally reported and model-predicted ternary liquid-liquid equilibria. The key thermodynamic evaluations of ILs are predicted by COSMO-RS and UNIFAC-IL separately while their important physical properties and commercial availability are estimated by a deep learning model and prepared python program. The screened promising IL [EPY][Tf2N] is tested by liquid-liquid equilibrium experiments and the NRTL parameters are regressed. The process simulation with the chosen IL is conducted using Aspen Plus and then compared with that of a benchmark IL, affirming the exceptional efficacy of the identified [EPY][Tf2N] as a prospective solvent for the benzene-cyclohexane separation. This study can be a crucial reference for guiding screening of ILs for various separation applications.