Study of the mechanism of enantioseparation of macrocyclic glycopeptide-based chiral stationary phases

Abstract

<p>The purpose of this research has been to investigate the mechanisms of chiral separations in HPLC. The goal was to develop a method of interrogating chiral separations that is applicable to all CSPs. This dissertation focuses on the development and application of this approach. The macrocyclic glycopeptide CSPs were used to experimentally test our model/approach and to provide proof of principle;The linear solvation energy relationship (LSER), developed by Kamlet, Taft, and Abraham, was selected for its ability to de-convolute the interactions a solute experiences in a biphasic system. The LSER model has a term for each type of interaction that a solute can experience or participate in. These include interactions through polarizable n and pi electrons (eE), dipolar interactions (sS), hydrogen bond acceptance ( aA) and donation (bB), and dispersion forces ( vV). Each term has a component for the system's ability to participate in the specified interaction, the lower case variables which are called system constants, and a component for the solutes ability to participate in the specified interaction, the upper case variables called solute descriptors;We obtained system constants for the macrocyclic glycopeptide CSPs in the reverse phase mode and the normal phase mode. With knowledge of the system constants, it is possible to determine the solute descriptors of each enantiomer. This is achieved using multiple linear regression analysis, with the system constants of a CSP as the independent variables, and the logarithm of the retention factor of one enantiomer as the dependent variable. The solute descriptors for each enantiomer are determined, and the differences in the solute descriptors reveal the relative importance of each intermolecular interaction in generating enantioselectivity;We applied this method of analysis to the reverse phase mode separations with the teicoplanin CSP. Several neutral compounds and amino acids were eluted and the solute descriptors for each enantiomer were determined. The statistical fits of the regression were excellent. It was found that steric repulsions and ion-dipole interactions had the strongest influence on enantioselectivity, with both types of hydrogen bonding having a weak influence on enantioselectivity. Dipolar interactions were found to be unimportant in generating enantioselectivity.</p>