Thermodynamic and QSRR Modeling of HPLC Retention on Modern Stationary Phases

This paper investigates correlations between the standard chromatographic parameter, log k(w), obtained experimentally (classical thermodynamic) and obtained through the use of quantitative structure-retention relationships (QSRR) models (extrathermodynamic approach). QSRR models were created using descriptors from calculation chemistry, selected a priori on the basis of chemical intuition. Four modern analytical columns (packed with C18), often used in pharmaceutical analysis, were selected for the study. The simple extrathermodynamic model was demonstrated to be superior in terms of retention description with relation to rather complex models based on thermodynamic hermeneutics. However, all the QSRR equations derived are of limited predictive value as regards prediction of retention on the basis of the structure of the analyte. Nonetheless, they allow the differentiation of mechanism of separation operating on individual tested stationary phases. The descriptors, which are most significant in QSRR equations, appear to characterize better the net effect of interactions of the analytes with the stationary phases and the mobile phase containing acetonitrile as compared to the HPLC systems comprising methanol. These observations suggest that retention properties of new stationary phases studied in this work differ.