Analysis of Experimental Adhesives with Reflectance-mode FTIR: Examining Spectral Reproducibility, Chemometric Approaches, and Archaeological Significance

Reflectance-mode Fourier transform infrared (FTIR) spectroscopy is emerging as a valuable non-destructive tool for studying Paleolithic adhesive technologies. Yet, its broader application requires careful consideration of reproducibility across conditions, methodologies, and time. This study uses an experimental reference set of resins, gums, protein-based glues, waxes, and multi-component mixtures applied to flint and analyzed over an eight-year interval with two different FTIR instruments to assess variability introduced by instrument configuration, sample orientation, and residue composition. Spectral data were examined using both chemometric methods and analyst-defined grouping strategies. A blind validation set of pure adhesives further tested classification accuracy. While key chemical signals were consistent across instruments after standardized preprocessing, subtle distortions affected chemometric clustering, particularly among compositionally similar adhesives such as spruce and pine resins. By contrast, analyst-based groupings following Kramers-Kronig transformation showed greater stability, underscoring the interpretive value of combining visual inspection with statistical analysis. These findings demonstrate that reflectance-mode FTIR can provide reliable insights into adhesive composition, even of complex mixtures, particularly when paired with transparent preprocessing strategies and realistic experimental references. More broadly, the results highlight how the choice of grouping strategy, whether statistical or visual, shapes interpretive outcomes, offering critical guidance for applying FTIR to archaeological residues where destructive sampling is not possible.