Advanced discriminating criteria for natural organic substances of cultural heritage interest: spectral decomposition and multivariate analyses of FT-Raman and FT-IR signatures.

Natural organic substances are involved in many aspects of the cultural heritage field. Their presence in different forms (raw, heated, mixed), with various conservation states, constitutes a real challenge regarding their recognition and discrimination. Their characterization usually involves the use of separative techniques which imply destructive sampling and specific analytical preparations. Here we propose a non destructive approach using FT-Raman and infrared spectroscopies for the identification and differentiation of natural organic substances. Because of their related functional groups, they usually present similar vibrational signatures. Nevertheless the use of appropriate signal treatment and statistical analysis was successfully carried out to overcome this limitation, then proposing new objective discriminating methodology to identify these substances. Spectral decomposition calculations were performed on the CH stretching region of a large set of reference materials such as resins, oils, animal glues, and gums. Multivariate analyses (Principal Component Analyses) were then performed on the fitting parameters, and new discriminating criteria were established. A set of previously characterized archeological resins, with different surface aspects or alteration states, was analyzed using the same methodology. These testing samples validate the efficiency of our discriminating criteria established on the reference corpus. Moreover, we proved that some alteration or ageing of organic materials is not an issue to their recognition.