Measurement of the chemical composition of ancient coins with portable XRF

Abstract

The elemental composition of antique coins is of great interest in the field of numismatics. An obvious quick and easy method to investigate these artifacts is X-Ray fluorescence, especially using portable spectrometers, although these spectrometers are limited in terms of precision and accuracy of the measurement results. In this work some shortcomings of XRF spectrometry were investigated. A crucial point is that enough single measurements should be taken to obtain meaningful averaged elemental compositions, leading to an increase in precision. Further, correction factors were derived to improve accuracy, so that overall, the set of data will be improved regarding both precision and accuracy. These correction factors were based either on reference samples or on simulations, thus demonstrating the feasibility of deriving meaningful correction parameters for defined sets of coins. Although portable XRF can possibly not replace high-precision/high accuracy measurements like ICP-MS, we show that nevertheless results, especially when obtained from a set of samples and from multiple measurements of one and the same sample, can be used to differentiate the material nature of uniform groups of coins, e.g. to find correlations between trace element concentrations or to identify clusters. Thus, portable XRF provides easy access to measurement of high sample numbers with the potential to discover yet unknown trends in composition, e.g. in numismatic collections.