Chiral Separations by High‐Performance Liquid Chromatography

Abstract

Abstract The word “chiral” is derived from the Greek word “cheir”, which means hand. Chiral molecules are molecules that are related to each other in the same way that a left hand is related to a right hand. These molecules are mirror‐images of each other and are nonsuperimposable. Chiral separations have been considered among the most difficult of all separations since enantiomers have identical chemical and physical properties in an achiral environment. In this chapter we will focus on techniques used in high‐performance liquid chromatography (HPLC). Most chiral separations by HPLC are accomplished via direct resolution using a chiral stationary phase (CSP). In this technique a chiral resolving agent is bound or immobilized to an appropriate support to make a CSP, and the enantiomers are resolved by the formation of temporary diastereomeric complexes between the analyte and the CSP. Various types of CSPs have been developed, including ligand exchange, protein‐based, carbohydrate‐based, Pirkle‐type, cyclodextrin‐based, and macrocyclic antibiotic‐based CSPs. Ligand exchange phases are used with aqueous buffer mobile phases in which enantiomers are separated based on the differences in their charge and ionization constants. Limitations are that only ionized analytes can be separated using this technique and the copper‐salt containing mobile phases used absorb in the ultraviolet (UV) region, decreasing detection sensitivity. Protein‐based CSPs comprise a number of commercially available columns. These CSPs can be used in the reversed‐phase mode with aqueous buffers and there are a limited number of variables to control in developing a separation method. Advantages of protein‐based CSPs include low column capacity, limited solvent options and the inability to reverse the elution order of the analyte. The carbohydrate‐based CSPs consist of derivatized cellulose and amylose phases and are generally used in the normal phase mode, with the exception of two derivatized phases which are conditioned for the reversed‐phase mode. The main disadvantages of these phases are the limitations in pressure and solvent used since these phases are not covalently bonded but merely adsorbed on the silica. These phases may not be used with solvents of intermediate polarity, for example, methylene chloride, acetone, tetrahydrofuran, and acetonitrile. The Pirkle‐type CSP typically uses nonpolar organic mobile phases such as hexane, with 2‐propanol or ethanol as organic modifiers. Under these conditions, retention of the solutes decreases as the mobile phase polarity increases, following the normal phase mode behavior. The Pirkle‐type columns are generally employed in separating compounds containing a π‐acid or π‐basic moiety, or both. The cyclodextrins can be used with either aqueous buffers or in the polar organic mode. Generally analytes separated using the cyclodextrins require formation of an inclusion complex with the cyclodextrin. Separation is most favorable when the analyte interacts with the mouth of the cyclodextrin molecule via hydrogen bonding to the functional groups present. The macrocyclic antibiotics can be used in either the reversed‐phase or the normal phase mode. Enantioselectivities in each mode have been shown to be different. Furthermore, the macrocyclic antibiotics can be derivatized to alter their selectivity. The bonded macrocyclic antibiotic CSPs resemble the protein‐based CSPs in many ways. However, the macrocyclic antibiotic CSPs are more stable and have greater capacities than the protein‐based CSPs. Numerous analytes are separated using this class of CSP.