Modelling retention in liquid chromatography as a function of solvent composition and pH of the mobile phase

The aim of this work was to develop a model that accurately describes retention in liquid chromatography (LC) as a function of pH and solvent composition throughout a large parameter space. The variation of retention as a function of the solvent composition, keeping other factors constants, has been extensively studied. The linear relationship established between retention factors of solutes and the polarity parameter of the mobile phase, E-T(N), has proved to predict accurately retention in LC as a function of the organic solvent content. Moreover, correlation between retention and the mobile phase pH, measured in the hydroorganic mixture, can be established allowing prediction of the chromatographic behavior as a function of the eluent pH. The combination of these relationships could be useful for modelling retention in LC as a function of solvent composition and pH. For that purpose, the retention behavior on an octadecyl silica column of a group of diuretic compounds covering a wide range of physico-chemical properties were studied using acetonitrile as organic modifier. The suggested model accurately describes retention of ionizable solutes as concomitant effects of variables included and is applicable to all solutes studied. We also aimed to establish an experimental design that allows to reproduce to a good approximation the real retention surface from a limited number of experiments, that is from a limited number of chromatograms. Ultimately, our intention is to use the model and experimental design for the simultaneous interpretive optimization of pH and proportion of organic solvent of the mobile phase to be used in the proposed separation. (C) 2000 Elsevier Science BN. All rights reserved.