Archaeological reconstruction of medieval lead production : implications for ancient metal provenance studies and paleopollution tracing by Pb isotopes

International audienceThe identification of metal provenance is often based on chemical and Pb isotope analyses of materials from the operating chain, mainly ores and metallic artefacts. Such analyses, however, have their limits. Some studies are unable to trace metallic artefacts or ingots to their ore sources, even in well-constrained archaeological contexts. Possible reasons for this difficulty are to be found among a variety of limiting factors: (i) problems of ore signatures, (ii) mixing of different ores (alloys), (iii) the use of additives during the metallurgical process, (iv) metal recycling and (v) possible Pb isotopic fractionation during metal production. This paper focuses on the issue of Pb isotope fractionation during smelting to address the issue of metal provenance. Through an experimental reconstruction of argentiferous Pb production in the medieval period, an attempt was made to better understand and interpret the Pb isotopic composition of ore smelting products. It is shown that the measured differences (outside the total external uncertainties of 0.005 (2 * sd) for 206 Pb/ 204 Pb ratios) in Pb signatures measured between ores, slag and smoke are not due to Pb mass fractionation processes, but to (1) ore heterogeneity (D 206 Pb/ 204 Pb slag-ores = 0.066) and (2) the use of additives during the metallurgical process (D 206 Pb/ 204 Pb slag-ores = 0.083). Even if these differences are due to causes (1) and/or (2), smoke from the ore reduction appears to reflect the ore mining area without a significant disturbance of its Pb signature for all the isotopic ratios (D 206 Pb/ 204 Pb smokes-ores = 0.026). Thus, because the isotopic heterogeneity of the mining district and additives is averaged in slags, slag appears as the most relevant product to identify ancient metal provenance. Whereas aiming at identifying a given mine seems beyond the possibilities provided by the method, searching for the mining district through analysis of the smelting workshop materials should provide a more appropriate approach in cases where no archaeological evidence of ancient mining is available. Furthermore, smoke Pb isotopic composition does not seem to be significantly affected by the metallurgical process. Paleopollution recorded in peat deposits could help to detect ancient mining production and workshops. Integrated collaboration between mining archaeologists and geochemists appears crucial to achieve this goal