Use of Global Analysis for Studying Drying Oils
Global analysis demonstrates several advantages over single wavelength analysis for studying drying oils. Firstly, as demonstrated by the 2.72 mM Co(hfac)2 and 2.04 mM Co(oct)2 data, diffraction and wavelength dependent baseline drift are intrinsically addressed through the fitting procedure, allowing for analysis of data without the need for pre-treatment and simplifying the overall fitting procedure. Secondly, as demonstrated by the quality of fit to theA B C model, including the fingerprint region of the IR in the modeling allows for the identification of intermediate species formed during the drying process that may be difficult or impossible to detect in other regions of the IR spectrum. Finally, the use of global analysis provides a visual representation of the quality of the final fit in the form of residuals, which aids in the identification of places where the model can be further improved.
Both A and C in these models correspond to those previously identified as the starting and ending species in previous IR studies of linseed oil drying. Specifically, A corresponds to bis-allylic C-H sites found in linoleic and linolenic fatty acids (Fig. 1) as evidenced by the peaks at 3010 and 723 cm-1, andC corresponds to isolated trans-alkenes formed after isomerization as evidenced by the peak at 970 cm-1. The intermediate species B has peaks at 1117 and 985 cm-1, neither of which are present in speciesA or C . These peaks are consistently present regardless of the catalyst used but have varying estimated molar absorptivities depending on the identity and amount of catalyst used. This results from limitations of global analysis as described above in the section “Selection of the Model for Global Analysis”.