To the best of our knowledge, the prediction of the enantioresolution ability of polysaccharides-based stationary phases in liquid chromatography for structurally unrelated compounds has not been previously reported. In this study, structural information of neutral and basic compounds is used to model their enantioresolution levels obtained from an immobilised cellulose tris(3,5-dichlorophenylcarbamate) stationary phase in reversed phase conditions. Thirty-four structurally unrelated chiral drugs and pesticides, from seven families, are studied. Categorical enantioresolution levels (RsC, 0 = no baseline enantioresolution and 1 = baseline enantioresolution) are established from the experimental enantioresolution values obtained at a fixed experimental conditions. From 58 initial structural variables, three topological parameters (two of them connected to the chiral carbon), and six molecular descriptors (one of them also related with the chiral carbon), are selected after a discriminant partial least squares refinement process. The molar total charge of the molecule at the working pH is the most important variable. The relationships between RsC and the most important structural variables and the drug/pesticide family are evaluated. An explicit model is proposed to anticipate the RsC levels, which provides 100% of correct anticipations. A criterion is introduced to alert about the compounds that should not be anticipated.
Keywords: Cellulose tris(3,5-dichlorophenylcarbamate) stationary phase; Discriminant partial least squares; Enantioresolution modelling; Enantioseparations; Reversed phase liquid chromatography.
Copyright © 2018 Elsevier B.V. All rights reserved.