Radiocarbon Dating of Individual Amino Acids from Archaeological Bone Collagen

Abstract

Since the development of accelerator mass spectrometry (AMS) for radiocarbon dating in the late 1970s, its ability to date small samples of bone has been of huge importance in archaeology and Quaternary paleoecology. The conventional approach to sample preparation has been to extract and gelatinize protein, which is then combusted and graphitized for analysis. However, this 'bulk protein' can contain a heterogeneous mixture of non-collagenous molecules, including humic acids and other soil components that may be of a different age than the bone and therefore affect the accuracy of its 14C date. Sample pretreatment methods have been an important area of development in recent years but still show inadequacies for the dating of severely contaminated bone. The idea of isolating and dating individual compounds such as single amino acids, to improve dating accuracy, has been discussed in the literature since the 1960s. Hydroxyproline, for example, makes up over 10% of bone collagen but is extremely rare in most other animal proteins, increasing the chances of its presence being endogenous to the individual being dated. Its successful isolation has therefore been considered a potential 'gold standard' for dating archaeological bone; however, extracting and suitably purifying single amino acids from bone has proved a challenging task. This paper presents a novel method for the compound-specific 14C dating of individual amino acids, including hydroxyproline, from archaeological bone protein. It is based on a preparative, mixed-mode liquid chromatography separation of underivatized amino acids, entirely in aqueous solution and free of organic solvents. The method is presented here in detail including application to standard bone samples establishing its accuracy and background carbon contribution. Results from 14C dating hydroxyproline and other individual amino acids, from both historical and archaeological bone, are shown to provide AMS dates that are statistically indistinguishable from those of the bulk protein.