A computer-aided ionic liquid design (CAILD) study is presented for the frequently encountered alkane/cycloalkane separations in petrochemical industry.Exhaustive experimental data are first collected to extend the UNIFAC-IL model for this system,where the proximity effect in alkanes and cycloalkanes is considered specifically by defining distinct groups.The thermodynamic performances of a large number of ILs for 4 different alkane/cycloalkane systems are then compared to select a representative example of such separations.By applying n-heptane/methylcyclohexane extractive distillation as a case study,the CAILD task is cast as a mixed-integer nonlinear programming (MINLP) problem based on the obtained task-specific UNIFAC-IL model and two semi-empirical models for IL physical properties.The top 5 IL candidates determined by solving the MINLP problem are subsequently introduced into Aspen Plus for process simulation and economic analysis,which finally identify 1-hexadecyl-methylpiperidinium tricyanomethane ([C16MPip][C(CN)3]) as the best entrainer for this separation.