An experimental design based strategy to optimize a capillary electrophoresis method for the separation of 19 polycyclic aromatic hydrocarbons.

Because of their high toxicity, international regulatory institutions recommend monitoring specific polycyclic aromatic hydrocarbons (PAHs) in environmental and food samples. A fast, selective and sensitive method is therefore required for their quantitation in such complex samples. This article deals with the optimization, based on an experimental design strategy, of a cyclodextrin (CD) modified capillary zone electrophoresis separation method for the simultaneous separation of 19 PAHs listed as priority pollutants. First, using a central composite design, the normalized peak-start and peak-end times were modelled as functions of the factors that most affect PAH electrophoretic behavior: the concentrations of the anionic sulfobutylether-β-CD and neutral methyl-β-CD, and the percentage of MeOH in the background electrolyte. Then, to circumvent computational difficulties resulting from the changes in migration order likely to occur while varying experimental conditions, an original approach based on the systematic evaluation of the time intervals between all the possible pairs of peaks was used. Finally, a desirability analysis based on the smallest time interval between two consecutive peaks and on the overall analysis time, allowed us to achieve, for the first time in CE, full resolution of all 19 PAHs in less than 18 min. Using this optimized capillary electrophoresis method, a vegetable oil was successfully analyzed, proving its suitability for real complex sample analysis.