Development and application of an analytical method for radiocarbon dating bones using the amino acid hydroxyproline

Archaeological bones are usually dated by radiocarbon measurement of extracted collagen. However, low collagen content, contamination from the burial environment or museum conservation work have previously lead to inaccurate results, especially for old bones, compromising the ability to reconstruct reliable past chronologies. It is reported, for example, that up to 70% of Palaeolithic radiocarbon dates on bones are likely to be underestimates of the real age, blurring the picture of modern human dispersals and Neanderthal extinction.In this thesis, a method for isolating and radiocarbon dating the collagen amino acid hydroxyproline is described. Hydroxyproline consists of about 10% of bone collagen but is not found in significant amounts elsewhere in nature. The hydroxyproline dating method uses a mixed-mode (i.e. ion-exchange combined with hydrophobic chemistry), semi preparative HPLC methodology. The amino acids do not require derivatisation, and no organic solvents are used, thereby avoiding addition of carbon. The hypothesis of this thesis is that the hydroxyproline can be used as a bone specific biomarker, improving dating accuracy and making it possible to obtain radiocarbon determinations where previously it has been impossible.It was calculated that on average 3.3±1.4μg of contaminant carbon are added to each sample in the process of isolating the hydroxyproline, a low level suitable for 14C dating. It was investigated whether a deliberately contaminated bone and 'naturally' contaminated archaeological bones, yielding erroneous dates when dated using the normal pretreatment method, could be dated accurately using this method. In addition, a hydroxyproline date was obtained for a bone with too little surviving collagen to be dateable by the bulk collagen method. Finally, using the hydroxyproline dating method, the earliest direct ages for the presence of anatomically modern humans on the Russian Plain were obtained. The method proved to be a powerful tool that can help resolve longstanding archaeological questions.This thesis is not currently available via ORA