Towards standardised development of palaeofire charcoal reference sets via muffle furnace pyrolysis for Fourier-transform infrared spectroscopy

Background The application of Fourier-transform infrared spectroscopy (FTIR) to subfossilised charcoal preserved in natural archives has been shown to be a useful technique to better understand historical fire (i.e. palaeofire) conditions. The reliability of FTIR analyses in palaeoecological studies is contingent on the robustness of charcoal reference sets, many of which are produced via muffle furnace pyrolysis under varying methodologies. Aim We tested how various methodological approaches influence FTIR spectral signatures, using dried plant material (leaves and branches) derived from Eucalyptus camaldulensis. Methods Methods evaluated included the preparation (e.g. milled or unmilled), containment (e.g. aluminium foil or container) and cooling (e.g. internal or external) of the modern reference material. Key results We found that methodological approaches substantially influence FTIR signatures and subsequent classification accuracies, determined via leave-one-out centroid matching, varying by nearly 30% between approaches. Conclusions We suggest that loosely containing samples in aluminium foil, using sand mediums (e.g. 50.00 ± 0.05 g and a grain size of 250–500 μm) for thermal buffering and immediately removing samples from the muffle furnace for cooling are preferred protocols for generating reference sets via muffle furnace pyrolysis. Implications This study highlights the critical importance of methodological transparency and standardised methodological protocols in developing charcoal reference sets for palaeofire analyses.