U-series dating of bone using the diffusion-adsorption model

U-series dating of bone has suffered problems of reliability since its inception because bone remains an open system with respect to uranium. Commonly applied a priori assumptions of U uptake, such as early uptake or linear uptake, are inadequate because they have no physical or chemical bases, no means of demonstrating which model is suitable for a particular bone, and no intrinsic tests of reliability. Despite this and numerous examples of anomalous U-series dates, such assumptions are still routinely applied. We address this problem using the diffusion-adsorption (D-A) model of U uptake (Millard and Hedges, 1996), which incorporates a physicochemical description of U uptake. Using this model, we show how the U uptake of a bone responds to geochemical changes in the burial environment, which can lead to phenomena such as the removal of U from bones ("leaching") or U uptake late in their burial history ("recent uptake"), and we show how the overall uptake history is reflected in distributions (profiles) of U and U-series isotopes across a bone section. We present measurements of U concentration profiles, and Th-230/U-234 profiles on archeological bone from a number of different sites and burial environments and compare the results to profiles predicted by the D-A model. Bones that have undergone complex uptake histories (which include U leaching or recent uptake) are identified on the basis of these profiles and rejected as unsuitable for dating. For bones that appear to have undergone uptake under constant geochemical conditions, the D-A model is applied to calculate U-series dates, with much improved reliability. Copyright (C) 2002 Elsevier Science Ltd.