Application of Solvation Model to Prediction of the Solute Retention in Liquid Chromatography over a Wide Range of Mobile-Phase Compositions

Possibilities have been studied of using a solvation model to predict the retention behavior of solutes in liquid chromatography mobile phases of water-acetonitrile and water-methanol with the organic solvent content varying from 1 to 100 vol%. Twenty-one test solutes, both aliphatic and aromatic compounds, have been selected on the basis of two-level factorial designs. Using the multiple linear regression analysis, regression coefficients, which are characteristics of the stationary and mobile-phase system, were calculated for different mobile phases in the solvation model. Regression coefficients have been used for the prediction of the retention behavior. Unbiased results have been obtained by using two sets, one training and the other the testing set. The predicted retention has been compared with the experimental data. The methanol-water system provided good results at low and medium methanol concentrations; the retention prediction was unsatisfactory for mobile phases containing more than 90% of methanol. The acetonitrile-water system yielded similar results, but the retention prediction ceased to be at acetonitrile concentrations greater than 80%. The retention has primarily been determined by cohesive and acid-base interactions. The dependences of the regression coefficients on the mobile-phase composition were similar for the acetonitrile-water and methanol-water systems.