Heat treatment of Kalahari and Cape silcretes: impacts upon silcrete chemistry and implications for geochemical provenancing

Recent studies in southern Africa and eastern Australia have demonstrated the feasibility of using a geochemical fingerprinting approach to determine the source locations from which silcrete raw materials were procured prior to their use in stone tool manufacture. The impact of intentional heat treatment of silcrete upon its chemistry, however, is unknown, meaning that heat-treated silcrete artefacts have to date been excluded from provenancing studies. This investigation presents the first high-resolution experimental analysis of the impacts of heat treatment upon the chemical composition of silcrete. The study compares the composition of unheated control samples against samples heat-treated to target temperatures of up to 600°C taken from four silcrete blocks from South Africa and Botswana. Chemical compositions of samples are determined using ICP-MS and ICP-AES. Experimental results indicate that heat treatment has a limited impact upon silcrete chemistry. Only 7 out of 65 minor, trace and rare earth elements analysed (Al2O3, Fe2O3, K2O, As, Cr, Cs and Cu) were depleted beyond expected error limits following controlled heating. There was no consistent pattern of elemental depletion across the four silcrete samples, although a greater number of elements were depleted from chalcedony-cemented Kalahari silcretes compared to microquartz-cemented Cape silcretes. We conclude that it is safe to use chemical data from heat-treated artefacts from the Cape as part of geochemical fingerprinting studies; however, we recommend that Cu and Cs concentrations be omitted from any statistical analyses until the effects of heat treatment upon these elements are fully understood. We echo the conclusions of previous studies by recommending that chalcedony-cemented silcrete artefacts that show signs of burning or intentional heat treatment be excluded from provenancing studies in the Kalahari and potentially elsewhere.