Computational studies for the elucidation of the enantiomer elution order of amino acids in chiral ligand-exchange chromatography

Owing to the exceptional sophistication of chiral ligand-exchange chromatography (CLEC) systems operating in the presence of chiral mobile phase (CMP) additives, only few studies dealing with mechanistic investigations have been presented so far. Nevertheless, dedicated computational protocols applied to simplified models, can furnish valuable information on the factors that mainly affect the overall enantiorecognition event. Accordingly, the extraordinary accordance observed between quantum mechanical (QM) calculations and crystallographic data led us to use optimized ternary complexes carrying the chiral selector O-benzyl-(S)-serine [(S)-OBS], as starting structures to build up a computational model enabling to explain the enantiomer elution order of amino acids with this enantioresolving agent. As a result of the calculation of 113 three-dimensional descriptors on the mixed complexes, and the generation of a decision tree, the delta-Energy of solvation (delta-Esol) was found to correctly classify all the compounds of the training set (20 species) according to the relative chromatographic behaviour. Thus, as a rule of thumb, the diastereomeric couples having a delta-Esol value lower than 5.321 kcal/mol (splitting node) experienced a “canonical” enantiomer elution order while an opposite situation occurred for all the others (reversed elution profile). The profitable predictive power of the developed model was assessed on the selected test set (5 species).