New approach for simultaneous spectral analysis of a complex mixture using the fractional wavelet transform

Fractional wavelet transform (FWT) is a powerful toot in order to extract largest analytical information from the spectral bands. For the above reason, the combined applications of fractional wavelet transform with principal component regression (PCR) and partial least squares (PLS) calibrations are performed for the quantitative analysis of a pharmaceutical mixture containing trimethoprim (TMP) and sulphamethoxazole (SMX). One interesting example of complex pharmaceutical mixture is given by the multi-component pharmaceutical sample having constant matrix content. In this study, the absorption spectra of the related compounds and their samples were processed by fractional wavelet transform. In the following procedure, FWT-coefficients obtained in the above step are used to construct the PCR and PLS calibrations. These hybrid mathematical approaches are denoted as FWT-PCR and FWT-PLS calibrations. FWT-PCR and FWT-PLS obtained by using the FWT-coefficients for the quantitative analysis of TMP and SMX gives the higher accurate and precise results than classical PCR and PLS based on the use of the original absorbance values. (C) 2009 Elsevier B.V. All rights reserved.