THE EFFECTS OF ALTERATION ON SOURCING ARCHAEOLOGICAL TURQUOISE

Turquoise was an important commodity in prehistoric Southwestern and Mesoamerican cultures. If the source of archaeologically recovered turquoise could be determined, prehistoric trade networks could be reconstructed. Previous attempts to identify turquoise source regions using trace element analyses have only been partially successful because some of the blue-green minerals recovered from archaeological sites and identified as turquoise are other materials (e.g. chrysocolla) and turquoise alters to clay minerals such as pyrophyllite, which could affect its trace element content. To avoid these problems we developed a technique using SIMS to source turquoise on the basis of H and Cu stable isotopes and the Fe/H ratio of the sample. To test our technique we analyzed turquoise samples from three source regions in the Southwestern United States and four variably altered samples from a single turquoise mine. Evaluation of EMPA analyses of these same samples shows that there is a strong correlation between charge balance and Fe content in unaltered turquoise, and altered turquoise plots off of this trend. These data suggest that as the Fe content of turquoise increases, the hydrogen content decreases because the O^2- substitutes for OH^- to neutralize the excess charge. Thus, most Fe substitution cannot be by charge neutral exchange, such as Cu²⁺ for Fe²⁺ and Al³⁺ for Fe³⁺, but Fe³⁺ must substitute for Cu²⁺ in the octahedral site. FTIR spectra of Fe-rich and Cu-rich turquoise also clearly show this variation in OH content between the two minerals. Analyses of the four variably altered samples of Fe-rich turquoise from the Castillian Mine, NM, show that as the turquoise alters to clay minerals, the Si content increases and the Cu and P contents decrease. In addition their δD values are significantly affected by alteration compared to the δ⁶⁵Cu values. However, the Fe and Al contents are less affected by this alteration. FTIR spectra of altered and unaltered turquoise also show that the H content of these turquoises is relatively unchanged. Therefore, some overlap between source regions occurs when only using the Cu and H isotopic data to source turquoise, whereas a combination of Cu isotopic analyses and Fe/H ratios provides a useful sourcing technique for archaeological turquoise that is relatively impervious to alteration.