IRON FOR RADIOCARBON DATING : INFLUENCE OF MODERN STEEL TOOLS ON 14 C DATING OF ANCIENT IRON ARTIFACTS

Before the 17th century, charcoal was regularly used in the production of iron (smelting and forging) and some of this charcoal carbon was incorporated into the iron. Depending on the age of the wood used to produce the charcoal, the age of the carbon incorporated in the iron lattice can reflect the age of manufacture of the iron artifacts. A reliable preparation method allowing for the routine dating of iron artifacts would permit the dating of numerous objects for which now the age can only be estimated. In an earlier work (Hüls et al. 2004), we tested the extraction of carbon from iron samples by closedtube combustion. The samples were cut in small pieces to ease the release of the carbon from the lattice. During the tests, it became clear that the steel tools used to cut the samples can add contamination at the surface. As modern steel is made using coal, this leads to erroneously old ages. We have tested ways to reduce or eliminate this surface contamination from the sampling tools using iron artifacts of known ages. In order to quantify the contamination, we produced standard test materials from pure iron (99.998% Fe) melted with carbon of known 14C content and prepared samples using different cutting tools. The results of these tests indicate that the proper choice of cutting technique and tool, combined with an additional cleaning of the freshly cut surface, reduces sample contaminations to low levels; measured sample 14C concentrations are close to the 14C content of the charcoal used to produce these standard iron samples. INTRODUCTION The dating of iron artifacts using radiocarbon has received considerable attention since accelerator mass spectrometry (AMS) reduced the required sample size to manageable proportions (Cresswell 1992; Harbottle et al. 1993; Nakamura et al. 1995; Beukens et al. 1999; Cook et al. 2001, 2003; Craddock et al. 2002; Enami et al. 2004; Scharf et al. 2004; Oinonen et al. 2009). In Hüls et al. (2004), we reported our ampoule combustion procedure to extract carbon from iron, based on the work of Cook et al. (2001), and our first successful dates on known-age iron artifacts. In a few cases, however, our measured 14C ages were significantly older than expected from the archaeological or historical context of the artifacts. In their review paper, Craddock et al. (2002) discussed the many intricacies of the production and processing of iron over the ages and their implications for the possible influence of old wood, fossil fuel, geological carbonate, or reworking of the iron on the measured 14C age of an iron object. Old carbon contamination may also be introduced into the sample by admixture of “modern” iron from the tools used for sampling (milling, drilling, sawing, etc.). As modern iron and steel is produced with 14C-free coal/coke, tool wear during sampling will lead to erroneously old ages (cf. Scharf et al. 2004). A prerequisite for the reliable 14C dating of iron is thus contamination-free sampling of the samples. In this paper, we report 14C measurements on iron artifacts cut with different tools and methods. Broken pieces of the used milling tools were also measured. The results obtained for the archaeological artifacts provide an indication of the effect of sampling contamination on a selection of real archaeological materials. To avoid problems due to potential age inhomogeneity of our archaeological test samples, we also produced laboratory standards with known carbon contents and 14C concentrations to further test the various cutting techniques. MATERIALS AND METHODS Archaeological Iron Artifacts and Modern Iron Standards We selected 5 iron artifacts (Table 1) with known archaeological ages: #1, an axe (0.5% C) from the 2nd to 3rd century AD; #2, an iron bloom (0.8% C) from the 3rd–2nd century BC; #3 and #4, two 1Corresponding author. Email: mhuels@leibniz.uni-kiel.de. https://doi.org/10.1017/S0033822200034421 Downloaded from https://www.cambridge.org/core. IP address: 54.70.40.11, on 24 Apr 2019 at 09:18:48, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.