Are crucible steel ingots isotopically homogenous? AMS radiocarbon measurements on ingots from Telangana, India

Radiocarbon analysis is increasingly used to directly date archaeological and historical metal objects, ranging from low-carbon bloomery iron to steel and cast iron. However, little is known about the isotopic homogeneity of iron-carbon alloys, particularly relating to the formation of primary cementite during crystallisation. Here, we present 14C measurements for five crucible steel ingots and one crucible steel object from Telangana in southcentral India. Two of the ingots were analysed twice. The results show a very wide scatter of 14C dates, far exceeding the expected age range for this assemblage. The repeat analysis of one of the ingots gave also widely different results, indicating a fundamental problem with the 14C analysis of crucible steel. We discuss the various factors that could have influenced the measured isotopic values, including variability in raw material, sampling contamination, and fractionation during the cleaning of the metal in hot acid leading to excessive sample mass losses prior to the extraction of carbon from the metal. We argue that mass-dependant fractionation of the different carbon isotopes between austenite and cementite during solidification of the ingot, and subsequent selective dissolution of one metal phase over the other, led to a distortion of the 14C signature to seemingly older ages. We recommend further research to explore the compound-specific isotopic signature of high-carbon iron alloys and the effect of selective corrosion on such material, to reduce potential errors in 14C dating of steel and cast iron.

Automated summary

Radiocarbon dating is widely used for metal objects in archaeology and history, but there is limited understanding of isotopic homogeneity in iron-carbon alloys. This study analyzes 14C measurements of crucible steel from India and finds a wide scatter of dates beyond the expected range. Factors including variability in raw material, sampling contamination, and fractionation during metal cleaning are discussed as potential influences on isotopic values. It is argued that selective dissolution of metal phases and mass-dependent fractionation during solidification distort the 14C signature, leading to apparent older ages. Further research is recommended to reduce potential errors in dating steel and cast iron.