Investigating a Novel Non‐Destructive Identification Technique for Prehistoric Adhesives With Dynamic Headspace Coupled to Comprehensive Two‐Dimensional Gas Chromatography—Mass Spectrometry

Adhesives were used by prehistoric humans for attaching a handle to a stone tool, to improve tool use. Remains of these adhesives preserve on stone tools until today. Chemical analysis of these residues is essential for an improved understanding of how humans exploited their natural environment, stone tool manufacturing and use. However, chemical analysis is not straightforward, the highly degraded residue and the precious artefacts impose limitation. In this study a novel (semi‐) non‐destructive identification technique for prehistoric hafting adhesives is reported; dynamic headspace sampling coupled to comprehensive two‐dimensional GC‐MS. The dynamic sampling results in a full characterization of the volatile profile of the adhesives. A major advantage is that the whole stone tool, with the adhering adhesive, can be analyzed. Moreover, good results are obtained using only slightly elevated temperatures, which avoids heat damage to the stone tools. Nonetheless, the established biomarkers for prehistoric adhesives are not extracted with this method. Therefore, a non‐targeted analytical approach combined with multivariate analysis is utilized. In this approach, the chromatogram of an unknown sample is compared to a database of known samples. In this study a start of an adhesive database is made with 14 different adhesives divided over 4 adhesive classes. The identification capability of this technique is further evaluated using six experimental stone tools, with different adhesives adhered to them and subjected to UV‐induced degradation. The large stone pieces could not fit in the automated sampling station, thus, a manual sampling set‐up was build. It was found that the sampling strategy did not affect the volatiles extracted and that comparison with the database was possible. The tar samples were the least affected and could be easily identified while the resin samples were more degraded and identification was difficult. This technique is promising for non‐destructive adhesive identification on prehistoric stone tools.