A test to determine the nature and presence of the memory effect columns packed with the amylose tris(3,5-dimethylphenylcarbamate) stationary phase

Acid/base mobile phase modifiers affect enantioseparations in ways that are not fully understood yet, for the lack of systematic studies. This makes chiral analysis of some pharmaceuticals difficult to reproduce. Once a column has been exposed to a modifier, the selectivity of certain pairs of enantiomers may change, for the better or the worse. We study the behavior of five enantiomeric pairs, three which are highly sensitive to the addition of certain modifiers and two that have little sensitivity to these modifiers. Their use permits the determination of the extent of the memory effect response on individual columns. The selectivity of 4-chlorophenylalanine methyl and ethyl ester, and of ketoprofen improve as a solution of ethanesulfonic acid is percolated through the column. As a result, these pairs are most useful for the determination of the extent of acid memory effect on a column. The selectivity of propranolol HCl and, to a lesser degree, Tröger's base increases as a solution of diisopropylethylamine is percolated through the column. The separation of each one of these five pairs is inversely affected by the percolation of the opposite acid/base solution. We used trans-stilbene oxide (TSO) as a ‘standard’ to determine the column stability because no memory effect is observed for it (its retention, enantioselectivity, and resolution remain constant). Understanding whether a column is under the influence of the memory effect is critical to both the analysis of pharmaceutical ingredients and to the development of preparative purification techniques for racemic mixtures. Thus, columns that were unreliable for method development and method transfer, due to the memory effect and a lack of proper solvent exposure records, can now be used.