A novel strategy for systematic optimization of micellar electrokinetic chromatography separations

In this paper, a novel strategy for systematic optimization of micellar electrokinetic chromatography (MEKC) separations has been developed. Statistical methods were extensively used in this strategy. A new statistical experiment-arrangement method called uniform design was employed to set optimization experiments to examine the effects on separations of multiple experimental factors at their ''multiple levels'' (multiple levels here means multiple test points enough to cover a vast range of an experimental factor). By statistical regression analysis and diagnostic method, a new optimization function called ''uniformity of peak distribution'' was selected from a series of definitions. Furthermore, this optimization strategy is an empirical and experiments-based approach, little information about the physical or chemical properties of the analytes is needed for optimization. When the resolution between the least resolved peak pair reaches zero in most cases of the optimization experiments, this strategy still works, while other reported optimization schemes such as overlapping resolution map and minimum resolution methods might fail. Finally, this strategy has been experimentally confirmed by the successful prediction of the separation conditions for seven barbiturates. This novel strategy might be applicable to separations of other kinds of compounds, and to capillary zone electrophoresis (CZE) separations as well.