In Silico Modeling of Hundred Thousand Experiments for Effective Selection of Ionic Liquid Phase Combinations in Comprehensive Two-Dimensional Gas Chromatography

The selection of the best column sets is one of the most tedious processes in comprehensive two-dimensional gas chromatography (GC × GC) where a multitude of choices of column sets could be employed for an individual sample analysis. We demonstrate analyte/stationary phase dependent selection approaches based on the linear solvation energy relationship (LSER), which is a reliable concept for the study of interaction mechanisms and retention prediction with a large database pool of columns and compounds. Good correlations between our predicted results, with experimental results reported in the literature, were obtained. The developed approaches were applied to the simulation of 157 920 individual experiments in GC × GC, focusing on the application of 30 nonionic liquid and 111 ionic liquid (IL) stationary phases for separation of some example sets of model compounds present in practical samples. The best column sets for each sample separation could then be extracted according to maximizing orthogonality, which estimates the quality of separation