COMPUTATIONAL STRATEGY FOR SOLVENT STRENGTH OPTIMIZATION IN REVERSED-PHASE LIQUID-CHROMATOGRAPHY

Computer algorithms for iterative solvent strength optimization are presented together with a practical development strategy consistent with other widely accepted schemes. This optimization strategy appears to be flexible enough to make use of all the experiments realized during the mobile phase development. The computer algorithm accepts any case of solute elution, during or after the gradient rise for linear or multi-linear gradients, as input or output parameters. The computer routine is able to compute solute retention from two gradients differing in their slopes and also slight changes in solvent composition. The iterative computation of the linear model allows a further minimization of the error in prediction. If some additional experiments are provided, the retention model can be extended to a more accurate quadratic form. From its computation, which uses the Nedler and Mead simplex controlled by an error reduction criterion, this model allows a further improvement in retention time prediction. This permits low prediction errors in the case of difficult developments that need several chromatographic runs.