Resolvability indices and variable selection in factor analysis of diode-array high-performance liquid chromatography: development of an approach for multi-peak clusters and application to mixtures of chlorophyll degradation products

Abstract

Resolvability indices for clusters of peaks in diode-array detector high-performance liquid chromatography are described. Indices can be defined for different resolution numbers of components within a cluster. Using a double window, wavelengths that are most useful for resolution can then be selected, and factor analysis performed on a subset of the over-all data. The method is applied to a real case study, namely mixtures of degradation products of chlorophyll a.