GEOCHEMICAL EVALUATION OF PETRIFIED WOOD AS A PROVENIENCE TOOL IN ARCHAEOLOGY

More than 100 samples of petrified wood from 7 different source localities within Colorado have been studied with SEM, and subjected to chemical analyses, including ED-XRF, ICP-AES, ICP-MS and LA-ICP-MS, in an attempt to define one or more chemical parameters which would be useful in assigning a source locality for archaeological artifacts. Abundance data for as many as 30 elements have been obtained for some pieces.

The SEM work indicates that original structures are preserved only in partially silicified pieces; pieces sufficiently silicified to be useful as raw material for artifacts exhibit few if any original structures.

The abundances of these elements tend to be low (< 200 ppm, most < 60 ppm) due to their residing in minor interstitial phases (as opposed to SiO2). Order-of-magnitude variations in abundance were observed for most elements, even within a single piece; LA-ICP-MS analysis of one piece indicates that this variation can be on a millimeter scale. Significant overlap in the abundances of virtually all the elements in samples from different localities is observed. In some cases, the differences in chemical abundances were relatable to color differences; in others, no visible differences were apparent. For most elements, there is nearly as much variation within a single piece as there is between different sites.

To evaluate the efficacy of applying 87Sr/86Sr ratios, ED-XRF Rb and Sr abundances were obtained for a subset of 60 samples from all 7 localities. The abundances of both Rb & Sr tend to be low (< 20 ppm), but are highly variable; calculated 87Rb/86Sr ratios (0.1 to 7.7, average 1.5) require that significant correction to the measured 87Sr/86Sr for in-situ decay would need to be applied.

The financial support of the Colorado State Historical Fund, Project #2002-02-039, is gratefully acknowledged.