Multivariate curve resolution alternating least square (MCR-ALS) analysis on total synchronous fluorescence spectroscopy (TSFS) data sets: Comparing certain ways of arranging TSFS-based three-way array

摘要

Total synchronous fluorescence spectroscopy (TSFS)-based three-way array sample x excitation wavelength x Delta lambda has inherently complex data structure. Multivariate curve resolution alternating least square (MCR-ALS) analysis is a soft mathematical modelling technique that can be used to retrieve pure spectral and contribution profile of the fluorophores of a multifluorophoric mixture provided it is applied to bilinear data sets. In the present work, possibilities of generating bilinear two-way arrays by (i) unfolding and (ii) augmenting the TSFS-based three-way array were studied. Bilinear characteristics of the obtained two-way arrays were tested by subjecting them to MCR-ALS analysis. It was found that lack of analytical-mathematical relationship between spectral profile along the excitation wavelength mode and the offset mode do not allow the application of MCR-ALS analysis on augmented TSFS data set. Moreover, the augmentation procedure and chemical-mathematical rank of the subsequently obtained data sets were found to be 'not independent' of each other. Two-way array sample x (excitation wavelength x Delta lambda) obtained by unfolding TSFS-based three-way array along the sample mode was found to have symmetrical structure, i.e., both sample mode and spectral mode have equal mathematical and chemical rank making them highly attuned to MCR-ALS analysis. (C) 2015 Elsevier B.V. All rights reserved.