Reexamination of Dependence of Plate Number on SDS Concentration in Micellar Electrokinetic Chromatography

Chromatograms of hydrophilic, hydrophobic, and intermediate-polarity analytes were developed in 50-μm capillaries by micellar electrokinetic chromatography at field strengths less than 31 kV/m. The analytes were solubilized by phosphate/borate buffers containing 15, 50, and 100 mM sodium dodecyl sulfate (SDS). The plate numbers N of the analytes, as well as those of the electroosmotic flow and micellar markers, were compared to predictions of N estimated by a simple model based on longitudinal diffusion and plug size. Good to fair agreement between theory and experiment was obtained for the hydrophilic and intermediate-polarity analytes in all buffers over the entire field strength range. Good agreement between theory and experiment was obtained for the hydrophobic analyte and micellar marker in all buffers at low field strengths; however, these compounds were subject to dispersion at higher field strengths by what appears to be Joule heating. The magnitudes of other, closely related Joule heating losses are quantified here using temperature profile measurements by Morris and co-workers and Taylor dispersion calculations. In contrast to the commonly reported increase of N with media concentration, the Ns of the hydrophilic and intermediate-polarity analytes were found to be essentially independent of SDS concentration over the investigated SDS range, and the Ns of the hydrophobic species were found to be independent of SDS concentration until (what appears to be) Joule heating became significant. These results were compared to those of Sepaniak and Cole. A critique of some previous studies of N vs SDS concentration is presented, in which quantitative explanations for some dispersions are offered as alternatives to surfactant concentration effects.