Handheld X-ray fluorescence in archaeology (2009-2015)

Beyond the Beam. Evaluation and application of handheld X-ray fluorescence in archaeology, PhD research Kaat De Langhe

The rapidly evolving field of XRF (X-ray fluorescence) analysis and the development of a new generation of lightweight, high-performance handheld XRF analysers (hXRF) promised new possibilities for in situ elemental determination of archaeological sites and artefacts. The primary aim of this research project was to evaluate the applicability of such a handheld XRF analyser in archaeological contexts and to develop a comprehensible protocol for data-processing. The first part of this study comprises the evaluation of commercially available handheld XRF analysers and the characterisation of the Olympus Innov-X Delta hXRF analyser. A work methodology was developed applying established lab protocols used in analytical chemistry. This instrument was then employed in three archaeological case studies and combined with different analytical techniques (micro-XRF (µ-XRF), Total Reflection XRF (TXRF) and Raman spectroscopy).

The archaeometrical investigation of two Bronze Age stone settings of the type ‘virtual dwelling’ in the Yustyd valley (Altai Republic, Russian Federation) was the first archaeological case study. The function of this type of structures is unknown. In the summer of 2011 hXRF analysis was used to perform a geochemical survey of the excavated structures in order to see if the soil features, as encountered during excavation, had a different chemical composition than the surrounding soil. Differences in their chemical composition can give information about the use of the site and structures. Aside from the hXRF investigation, soil samples were taken to be investigated in the lab by means of TXRF spectroscopy. The results, however, were not conclusive. Possibly the harsh environmental conditions influenced the results or it might be that the impact of the activities on the site, some four millennia ago, was too limited to leave traces today.

The most extensive case study is the archaeometrical investigation into the origin and distribution of Flemish late and post-medieval stove-tiles by means of hXRF and µ-XRF. Stove-tiles are no uncommon goods in Flemish excavations. Their origin, however, is unknown since no pottery workshops with evidence of stove-tile production have been found in excavation. The goal of the research was to determine the chemical composition of the tiles and to determine their provenance, by comparison with locally produced pottery and raw clay. Furthermore, the impact of tile-stoves and their imagery as a symbol of status and a representation of identity was discussed and interpreted in relation to the archaeometrical results. The hXRF analyses of the whiteware stove-tiles delivered consistent results showing well-defined groups within the research material, corroborating with the results of the µ-XRF analysis. The study clearly set the whiteware stove-tiles apart from the Antwerp maiolica: either the tiles are imported as a finished product, or the clay was imported and locally used for stove-tile production. The results of the redware stove-tiles on the other hand were more complex. The hXRF results diverged from those of the µ-XRF analysis, giving a different view on the relation between the comparative material and the stove-tiles themselves.

The third and last research project consists of a provenance-study of Iron Age red-painted pottery. The goal was to investigate whether the red-painted pottery, found on several Iron Age sites in Belgium and Northern France and referred to as Kemmelware, was actually produced in the Mount Kemmel hill fort. hXRF was employed to determine the elemental composition of the pottery while Raman spectroscopy was used to investigate the pigment used for the red slip decorations. For some sites, i.e. Kooigembos, Houplin-Ancoisne and Elversele, the hXRF data clearly demonstrated connections with the Iron Age settlement of Mount Kemmel, which can support the hypothesis that the pottery was produced in Kemmel and then distributed to other Iron Age sites. The Houplin-Ancoisne dataset gave evidence of two types of red-painted pottery: the second group clearly diverged from the Kemmelware and can be labelled as locally produced red-painted pottery. The pottery from Hove was also chemically and technologically different, again indicating local production.


Prof. dr. Peter Vandenabeele

dr. Kaat De Langhe