In capillary electrophoresis (CE), resolution (Rs) and selectivity (α) are criteria often used in practice to optimize separations. Nevertheless, when these and other proposed parameters are considered as an elementary criterion for optimization by mathematical maximization, certain issues and inconsistencies appear. In the present work we analyzed the pros and cons of using these parameters as elementary criteria for mathematical optimization of capillary electrophoretic separations. We characterized the requirements of an ideal criterion to qualify separations within the framework of mathematical optimizations and, accordingly, propose: -1- a new elementary criterion (t') and -2- a method to extend this elementary criterion to compose a global function that simultaneously qualifies many different aspects, also called multicriteria optimization function (MCOF). In order to demonstrate this new concept, we employed a group of six alkaloids with closely related structures (harmine, harmaline, harmol, harmalol, harmane and norharmane). On the basis of this system, we present a critical comparison between the new optimization criterion t' and the former elementary criteria. Finally, aimed at validating the proposed methods, we composed an MCOF in which the capillary-electrophoretic separation of the six model compounds is mathematically optimized as a function of pH as the unique variable. Experimental results subsequently confirmed the accuracy of the model.
Keywords: Capillary electrophoresis; Harmala alkaloids; Optimization; Quality criteria.
Copyright © 2016. Published by Elsevier B.V.