[Zurück]

B. Lendl:
"Extracting useful information from evolving data sets using Two Dimensional Correlation Spectroscopy and Multivariate Curve Resolution";
Hauptvortrag: Advanced Spectroscopies on Biomedical and Nanostructured Systems, Cluj (eingeladen); 03.09.2006 - 06.09.2006.

This paper first reviews two different chemometric techniques for analysis of evolving systems namely two dimensional correlation spectroscopy (2DCoS) and multivariate curve resolution alternating least squares (MCR-ALS).
After this introduction examples on two are given where these techniques have been successfully applied to extract meaningful information from the recorded FTIR data sets.
In the first example binary mixtures of methanol and water in three ionic liquids (IL), 1-ethyl-3-methylimidazolium tetrafluoroborate (emimBF4), 1-butyl-3-methylimidazolium tetrafluoroborate (bmimBF4) and 1-butyl-3-methylimidazolium hexafluorophosphate (bmimPF6) have been studied using Fourier transform Infrared Spectroscopy. A sequential injection system (SIA) allowed the reproducible preparation of different concentration mixtures by automated dilution of gravimetrically prepared mixtures. The region of OH stretching (3800 - 3250 cm-1) has been analyzed using 2 Dimensional Correlation Spectroscopy (2DCoS) and Multivariate Curve Resolution (MCR). Both chemometric techniques give evidence that at low concentrations only methanol and water monomers interacting with the anion of the ionic liquid exist. At higher methanol/water concentrations a second species is observed. The concentration profiles and the pure infrared spectra have been recovered using MCR-ALS. According to the observed spectral features the second species can be assigned to water or methanol dimers. In the most hydrophobic ionic liquid bmimPF6 only one water species could be clearly identified. No evidence was found that bigger water/methanol clusters are formed in these IL in the studied concentration range.
The second example reports on the proteolysis of BSA using proteinase K studied at 60°C again using FTIR spectroscopy. The use of Multivariate Curve Resolution-Alternating Least Squares method applied to infrared measurements allowed the recovery of pure infrared spectra and concentration profiles of the different species involved in the reaction. Special attention was paid to the careful inspection of residuals with the use of Two Dimensional Correlation Spectroscopy. In this way, a heat-induced unfolding step previous to protein hydrolysis was identified. The infrared spectra of the intermediate species showed a more disordered structure than native albumin, being especially noticeable the decrease in -helix conformation. The formation of aggregated -sheet aggregates due to heating was detected too.