Scope of partial least-squares regression applied to the enantiomeric composition determination of ketoprofen from strongly overlapped chromatographic profiles

PADRÓ, J. M.; OSORIO GRISALES, J.; ARANCIBIA, J.; OLIVIERI, A.; CASTELLS, C. B.; CASTELLS, C. B.

JOURNAL OF SEPARATION SCIENCE

WILEY-V C H VERLAG GMBH

Lugar: Weinheim; Año: 2015 p. 2423 - 2430

1615-9306

Valuable quantitative information could be obtainedfrom strongly overlapped chromatographic profiles of two enantiomers by usingproper chemometric methods. Complete separation profiles where the peaks are fullyresolved to the baseline are difficult to achieved in chiral separationmethods, and this becomes a particularly severe problem in case that theanalyst need to measure the chiral purity, i.e., when one of the enantiomers is present in the sample invery low concentrations. In this report, we explore the scope of a multivariatechemometric technique based on unfolded partial least-squares (U-PLS)regression, as a mathematical tool to solve this quite frequent difficulty.U-PLS was applied to obtain quantitative results from partially overlapped chromatographicprofiles of R- and S-ketoprofen, with different values of enantioresolutionfactors (from 0.81 down to less than 0.2 resolution units), and also at severaldifferent S:R enantiomeric ratios. The profiles were acquired at multipleUV-visible absorption wavelengths by means of a diode array detector. Enantiomericpurity below 1% could be determined with excellent precision even from almostcompletely overlapped signals. All these assays were tested on the mostdemanding condition, i.e., when the minor peak elutes immediately after thepeak for the main enantiomer. The results were validated using univariatecalibration of completely resolved profiles and the method applied to thedetermination of enantiomeric purity of a commercial pharmaceuticals.