Use of design of experiments to optimize high-throughput semipreparative LC and LC/MS methods

This paper describes an investigation into decreasing the run time for high-throughput semipreparative RP-HPLC methods without compromising the resolution. Experimental design was used to devise a small set of experiments in which factors, including solvent flow rate, solvent/column temperature, at-column dilution, and run time were varied systematically. The results were analyzed by means of multiple regression and partial least squares to generate a model relating the factors to the results, showing which factors are important. The model was then used to determine the optimal conditions.