Abstract
The anomalous dispersion of hydrophobic, highly retained analytes at high electric field strengths E in micellar electrokinetic chromatography (MEKC) is experimentally and theoretically quantified. A simple model for plate number N is proposed, in which only plug size and longitudinal diffusion govern dispersion. This model describes well the dispersion of acetone, 2′‐deoxyadenosine, and 2′‐deoxyadenosine‐5′‐monophosphate (dAMP) in MEKC buffers and also dAMP in capillary‐electrophoresis buffers. However, the model describes the dispersion of 1‐nitropyrene, pyrene, perylene, and the 4‐chloro‐7‐nitrobenzofuran (NBD) derivative of cyclohexylamine only at low E's; for these analytes, a nonequilibrium‐like dispersion causes N to decrease rapidly at high E's. Furthermore, the dispersion of NBD‐cyclohexylamine has a dependence on capillary diameter. Theoretical arguments are proposed, which show that this additional dispersion cannot be explained quantitatively by mass‐transfer kinetics, micellar polydispersity, hydrodynamic flow, Joule heating, or the radial variation of partition coefficient. In addition, experiments are performed suggesting that wall adsorption and sample inhomogeneity are not responsible for the efficiency loss. Other dispersion sources are suggested for future exploration.

This publication has 63 references indexed in Scilit: