A review of criteria functions and response surface methodology for the optimization of analytical scale HPLC separations

Selecting optimum HPLC operating conditions during the development of new analytical separations is difficult due to the high degree of process variable interaction and the lack of robust process models. Traditionally, the methods development strategy in analytical applications involves a trial-and-error grid search method that is both inefficient and costly. Several researchers have investigated more practical and efficient methods for designing optimal HPLC separations at the analytical stage. These strategies typically utilize an efficient design of experiments and response surface optimization techniques. Response or criteria functions are employed to numerically quantify chromatograms and rank them in order of desirability. A crucial step in the optimization problem is the selection of a proper response function. Several such response functions exist and the choice of a proper function is dependent on the overall goal of the separation at hand. The intent of this review is to present and comment on the strengths and shortcomings of several of the more commonly used criteria functions as well as to illustrate basic response surface optimization strategies and techniques as applied to analytical scale HPLC separations.