A retention model for the chiral separation of an uncharged solute, felodipine, on CHIRAL-AGP, using a micellar mobile phase is proposed. The model assumes the presence of two stereoselective sites and each enantiomer was found to interact with different sites. Addition of a chiral aliphatic alcohol, (+)-(S)-2-octanol, preferentially interacted with the binding site for (-)-(S)-felodipine. The monomeric form of the micellar agent (Tween@ 20) competed with the enantiomers for the adsorption sites, and the formation of a 1: 1 complex between the enantiomers and the micelles was assumed. The retention of the solutes was effectively controlled by adding small quantities (c1.63 x lop3 M) of the nonionic detergent Tween 20 to the mobile phase. Baseline separation was achieved by addition of 1.0 mM n-octylamine to the mobile phase; 8.14 x M Tween 20 in phosphate buffer pH 7.0. The separation factor (a = 1.74) was unaffected by the detergent concentration in the presence of 1.0 d n-octylamine. 0 1995 Wdey-Liss, Inc.
KEY WORDS: felodipine, retention model, micellar mobile phases, chiral resolution, CHIRAL-AGPThe enantiomeric resolutions of three hydrophobic amines' and an acid2 using Tween@ 20 as an organic modifier along with different charged additives on a,-acid glycoprotein columns have been described previously. The experiments were typically performed by decreasing the retention time by addition of increasing concentrations of a detergent, and thereafter adding a suitable counterion in order to increase the enantioselectivity. This protocol proved to be particularly useful for the hydrophobic enantiomers studied as it gave short retention times with maintained stereoselectivity. Organic solvents like methanol or acetonitrile have been used extensively as mobile phase modifiers in order to shorten retention times andor give better peak shape. Unfortunately, the presence of high concentrations of uncharged modifiers usually ruins the stereoselectivity, and thus the usefulness of this approach is limited when deahng with highly retained solutes.Although the mechanism of c k a l recognition by the protein is largely unknown, a retention model study can give a general conception of the main types of solute-micelle-protein interactions involved. In order to simplify the system, an aprotic solute, felodipine, was chosen as a test substance for evaluation of the retention model.In 1964 Hemes et al. developed an equation which described the elution of a small solute on a Sephadex G-25 column with a micellar mobile phase.3 The derivation outlined closely followed the classical work of Martin and Synge on partition chromatography4 and also accounted for molecular sieving and micelle partitioning effects. In 1981 Armstrong and Nome introduced the three-phase model for micellar liquid chromatography and also treated the relationship between 0 1995 Wdey-Liss, Inc.solute retention and micellar mobile phase cornpositi~n.~ Later Arunyanart and Cline Love' modelled the retention equation in the form of binding constants and ...