Copper and tin isotopic analysis of ancient bronzes for archaeological investigation: development and validation of a...

Although in many cases Pb isotopic analysis can be relied on for provenance determination of ancient bronzes, sometimes the use of "non-traditional" isotopic systems, such as those of Cu and Sn, is required. The work reported on in this paper aimed at revising the methodology for Cu and Sn isotope ratio measurements in archaeological bronzes via optimization of the analytical procedures in terms of sample pre-treatment, measurement protocol, precision, and analytical uncertainty. For Cu isotopic analysis, both Zn and Ni were investigated for their merit as internal standard (IS) relied on for mass bias correction. The use of Ni as IS seems to be the most robust approach as Ni is less prone to contamination, has a lower abundance in bronzes and an ionization potential similar to that of Cu, and provides slightly better reproducibility values when applied to NIST SRM 976 Cu isotopic reference material. The possibility of carrying out direct isotopic analysis without prior Cu isolation (with AG-MP-1 anion exchange resin) was investigated by analysis of CRM IARM 91D bronze reference material, synthetic solutions, and archaeological bronzes. Both procedures (Cu isolation/no Cu isolation) provide similar δ (65)Cu results with similar uncertainty budgets in all cases (±0.02-0.04 per mil in delta units, k = 2, n = 4). Direct isotopic analysis of Cu therefore seems feasible, without evidence of spectral interference or matrix-induced effect on the extent of mass bias. For Sn, a separation protocol relying on TRU-Spec anion exchange resin was optimized, providing a recovery close to 100 % without on-column fractionation. Cu was recovered quantitatively together with the bronze matrix with this isolation protocol. Isotopic analysis of this Cu fraction provides δ (65)Cu results similar to those obtained upon isolation using AG-MP-1 resin. This means that Cu and Sn isotopic analysis of bronze alloys can therefore be carried out after a single chromatographic separation using TRU-Spec resin. Tin isotopic analysis was performed relying on Sb as an internal standard used for mass bias correction. The reproducibility over a period of 1 month (n = 42) for the mass bias-corrected Sn isotope ratios is in the range of 0.06-0.16 per mil (2 s), for all the ratios monitored.