Use of principal component analysis for studying the separation of pesticides on polyethylene-coated silica columns

The retention time of 27 pesticides was determined on an polyethylene-coated silica high-performance liquid chromatography column (PEEsil) using water-methanol mixtures as eluents. Linear correlations were calculated between the logarithm of the capacity factor and the methanol concentration in the eluent (C) and the relationship between the slope and intercept values of the above correlation, the hydrophobicity, specific hydrophobic surface area of the pesticides and their retention parameters on porous graphitized carbon column was elucidated by using principal component analysis followed by two-dimensional nonlinear mapping. The most polar pesticides showed irregular retention behaviour of the PEEsil column, their retention decreased with increasing concentration of methanol in the eluent. This irregular retention was tentatively explained by the silanophile effect: at higher methanol concentrations the polar substructures of pesticides have a higher probability of binding to the polar adsorption centers on the silica surface not covered by the hydrophobic ligand increasing in this manner the retention capacity of the support Principal component analysis showed that PEEsil support has retention characteristics different from those of traditional reversed-phase supports. (C) 1998 Elsevier Science B.V.