AIC's 46th Annual Meeting

This paper presents a novel method for conducting wood identification based on chemical analysis using heart-cut pyrolysis gas chromatography/mass spectrometry (HC-Py-GC/MS) to analyze volatile fractions and thermal decomposition products from finely divided wood samples. This method has several advantages over traditional anatomical identification including a significantly reduced sample size (0.3 mg of powder vs. more than 40 mg for traditional thin anatomical sections), and increased ease of sampling. The method also shows promise for successfully discriminating between species that are not separable by anatomical methods. The use of an established analytical technique that is widely found in conservation science laboratories should make this method readily accessible to many researchers in the cultural heritage sector. The use of user-friendly and commercially available software for the evaluation of the GC/MS data also makes it possible to develop a reference database that can be easily shared and referenced by collaborating researchers. Evolved gas analysis (EGA) was used to establish an optimized furnace temperature that minimizes the production of compounds from the pyrolysis of cellulose and hemicellulose while maximizing the contribution of non-cellulosic components such as lignin and extractives, which are more likely to be characteristic of specific species. The use of a selective sampler system further reduces cellulosic contributions to the chromatograms by diverting evolved gases away from the GC column after 30 seconds of sample residence in the pyrolyzer. Results were interpreted through comparison with reference standards utilizing F-Search from Frontier Laboratories, which is software commonly used for the identification of polymeric materials and additives in plastics. The software produces a weighted average of the mass spectra of all integrated components in a chromatogram (an INT-SUM spectrum), which can be matched against an established library of standards. Comparison of the chromatograms and statistical evaluation of the INT-SUM spectra by F-Search provided accurate results and eliminated the need for specific compound identification, thus rapidly increasing the speed of data interpretation. F-Search also allows for the exclusion of peaks, which is a feature used to eliminate problematic peaks produced by contaminants such as glues, varnishes or waxes. For this preliminary study, reference samples of 62 wood species commonly found in decorative arts collections were analyzed with the optimized HC-Py-GC/MS method. The resulting chromatograms and INT-SUM spectra were compiled in a reference library. The method was validated by analyzing samples taken from 17th – 19th century objects within the J. Paul Getty Museum collection and comparing the results to identifications made through traditional anatomical study. All of the samples were correctly identified through the combined use of the F-search ranking system and visual comparison of the chromatograms.