COMPUTER-AIDED MODELS FOR OPTIMIZATION OF ELUENT PARAMETERS IN CHIRAL LIQUID-CHROMATOGRAPHY

The application of a central composite design to the enantiomeric separation of the antifungal drug tioconazole is investigated. The design involves application of a mathematical model to the data to model the response in regions of the factor space not investigated in the experimental design. The significance of the variable terms in the model is a statistically and those terms declared not significant are removed from the model. The statistical adequacy of these reduced models is discussed, together with an examination of the prediction errors of the models. Three-dimensional predicted response surfaces for the complete models are presented and the predictive performance assessed.