Application of FT-Raman and FTIR measurements using a novel spectral reconstruction algorithm

Recently, an advanced spectral reconstruction algorithm based on information entropy was developed to identify individual compounds contained in mixture spectra without recourse to any library or any a priori knowledge. In this study, standard mixtures containing various polycyclic aromatic hydrocarbons (PAHs) and alpha,omega-dicarboxylic acids were measured by solid-state FT-Raman and FTIR spectroscopy, and this was followed by the application of the aforementioned algorithm to recover all pure component spectra contained therein. The results demonstrate that the developed algorithm successfully recovered the spectra of individual species even though only a very limited number of mixture spectral measurements were made. FT-Raman measurements coupled with chemometric analysis provides satisfactory pure spectral results for both PAHs and alpha,omega-dicarboxylic acids. Additionally, FT-Raman measurements together with FTIR measurements provided even better spectral recovery for some of the alpha,omega-dicarboxylic acids when using band-target entropy minimization (BTEM). This study demonstrates that chemometric analysis without a priori information provides a promising solution and simplified approach to identify individual components from samples containing a complex composition. In particular, this is desired for a better understanding of environmental samples. Copyright (C) 2003 John Wiley Sons, Ltd.