pXRF: a study of inter-instrument performance Nathan Goodale a, * , David G. Bailey b , George T. Jones a , Catherine Prescott a , Elizabeth Scholz a , Nick Stagliano a , Chelsea Lewis a a Anthropology Department, Hamilton College, Clinton, NY 13323, USA b Geosciences Department, Hamilton College, Clinton, NY 13323, USA article info Article history: Received 20 April 2011 Received in revised form 25 August 2011 Accepted 10 October 2011 Keywords: X-ray fluorescence (XRF) Portable XRF (pXRF) Great Basin Lithic analysis abstract Integrating portable X-ray fluorescence (pXRF) instrumentation into archaeological investigations has generated as much interest as skepticism because several characteristics of the technique limit analytical accuracy and precision. This paper seeks to explore inter-instrument performance in order to assess the potential pXRF instrumentation has in examining the elemental composition of lithic raw materials. Two pXRF instruments, Delta and Omega models both manufactured by Olympus Innov-X Systems, were utilized to obtain the trace element signatures of five fine-grained volcanic sources and four obsidian sources in the Great Basin. We compared the two portable instruments to a wavelength-dispersive instrument extensively calibrated with geologic standards and to a commercial benchtop energy- dispersive system. Our results suggest that the relationships between the data produced by these different instruments are complex, and as we look to the future of pXRF use in archaeological analysis we argue that that the same strict protocols applied in laboratory-based XRF analyses, involving instrument calibration, evaluation of inter-instrument performance, and comparison to accepted geologic standards, need to be applied in pXRF analysis. Ó 2011 Elsevier Ltd. All rights reserved. Portable XRF (pXRF) technology is becoming commonplace in archaeological lithic sourcing studies (Craig et al., 2007; Gliozzo et al., 2011, 2010; Jia et al., 2010; Jones et al., in press.; Kato et al., 2009; Nazaroff et al., 2010; Phillips and Speakman, 2009; Sheppard et al., 2010; Tykot, 2010; Tykotet al., 2011). While there are benefits to pXRF technology, for example non-destructive analysis in the field and considerable saving of analysis time, we have not fully assessed what ramifications this technology has for the comparability of results with those produced by conventional energy-dispersive (edXRF) and wavelength-dispersive (wdXRF) systems. In this paper we tackle some of the issues involved in integrating pXRF instruments into archaeological sourcing studies. In this study, we compare data produced by two pXRF units to data from wdXRF and edXRF systems that were calibrated using geologic standards and which produce composition estimates that align with results from other laboratories. Our data demonstrate the complexity of XRF analysis, in terms of intra-instrument results, inter-system results, and instrument drift. As we look into the future of lithic studies there is no question that XRF analysis will continue to be an important technique for elucidating cultural processes, but productive use of pXRF for archaeological analyses will require that operators follow systematic protocols for instru- ment calibration using international rock standards, and develop methods for on-going assessment of instrument stability. 1. Importance of sourcing stone Stone tool analyses shed light on many of the cultural, social, and ecological patterns of prehistoric and historic societies (Jones et al., 2003). Among the broad array of technologies and methods avail- able for studying stone tools and the byproducts of stone tool production, source provenance analysis has proven especially effec- tive for interpreting land-use and mobility strategies of past peoples. Provenance analysis relies on accurate estimates of elemental composition to differentiate among geologic sources of raw material and for attribution of artifacts to those sources. A range of techniques can be applied to this task; pXRF is a relatively new approach that offers an efficient, non-destructive means to analyze the elemental composition of lithic artifacts. Our study seeks to determine the cost of this efficiency on analytical accuracy and precision. 2. Current research The upsurge in use of pXRF instrumentation has generated as much skepticism as it has interest. Several studies exploring the * Corresponding author. E-mail address: ngoodale@hamilton.edu (N. Goodale). Contents lists available at SciVerse ScienceDirect Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas 0305-4403/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2011.10.014 Journal of Archaeological Science 39 (2012) 875e883