Chemometric modeling of neurotransmitter amino acid separation in normal and reversed migration micellar electrokinetic chromatography

A chemometric experimental design has been applied for the optimization of neurotransmitter amino acid separation in capillary electrophoresis. The op timizations were carried out for normal micellar electrokinetic chromatogra phy (N-MEKC) and reversed migration micellar electrokinetic chromatography (RM-MEKC). In order to optimize three separation factors and study the inte raction between factors, a response function was optimized via searching it s optimum (minimum/maximum). Far this purpose a central composite design wi th multivariate linear regression (MLR) analysis was utilized. Modeling wit h good regression coefficients from the MLR adequately described the intera ction of factors such as background electrolyte and sodium dodecylsulfate c oncentrations which had a large impact on selectivity and migration behavio rs. Similar optimal conditions regarding resolution and number of theoretic al plates but different retention behaviors as a function of background ele ctrolyte and micellar concentrations were observed for N-MEKC and RM-MEKC. Improved overall performance from the RM -MEKC separation of five neurotran smitter acids, superior to N-MEKC, is demonstrated in terms of repeatabilit y, peak symmetry, sensitivity, and in particular, impurity determination in an overloaded separation system. (C) 2001 Elsevier Science B.V. All rights reserved.