AMS radiocarbon dating of iron artefacts

Abstract

Iron is combusted with CuxO as an oxygen donor in a sealed quartz tube to extract its carbon for radiocarbon dating. We used “modern” iron and steel samples with carbon contents between 0.07% and 3.8%, to test the efficiency of carbon extraction varying: (a) the grain size of the sample aliquot, (b) temperature (1000 and 1050 °C), (c) duration (10 and 24 h) and (d) oxygen excess of the combustion. The amount of oxygen had a clear effect on the carbon yield, requiring a stoichiometric excess. Incomplete carbon extraction has, however, no significant effect on the measured 14C-concentration and radiocarbon age. No significant difference in carbon yield was observed when varying the other combustion parameters or the carbon content of the iron. Reproducibility of the carbon yield is between ±0.1%C for high carbon iron (>3%C content) and ±0.02%C for low carbon iron (∼0.5%C content). For routine AMS sample preparation we have chosen a combustion of the iron samples at 1000 °C in sealed quartz ampoules for 24 h with a CuxO/Fe ratio >5 (by weight). Modern iron samples give apparent ages from 16 000 to 33 000 yr BP, reflecting a mixture of 14C-free coal carbon with small amounts of modern carbon (between 2% and 12%). Preliminary dates for seven iron artefacts agree well with “known” archaeological ages in the range 800 BC to 1200 AD.