THE CONTRIBUTION OF MAJOR AND TRACE ELEMENTS IN THE EMERALD-DEPOSITS CLASSIFICATION SCHEME: STATISTICAL TREATMENT OF EMPA AND SIMS DATA

Since ancient times, the mining and exchange of emeralds have been playing an important role in the history of civilizations. The widespread diffusion of this gem very far from the deposits highlights the historical importance of emeralds in the economic and religious structures of ancient cultures. Therefore, studies aimed at connecting each gemstone to its source (i.e., its provenance) continue to be a central topic for archaeologists, historians, geologists, gemologists, and so on. Today, the scientific community is involved to solve the existing debate about the emerald-deposits genetic classification schemes, which is based on several genetic models (Gavrilenko and Dashevsky, 1998; Dereppe et al., 2000; Schwarz and Giuliani, 2001; Schwarz et al., 2001; Barton and Young, 2002). Some authors classified emerald deposits in two, three, five or more different categories (Gavrilenko and Dashevsky, 1998; Dereppe et al., 2000), or suggested artificial neural networks (ANN, based on 450 electron microprobe analyses) to group the worldwide deposits. The genetic model for emerald formation involves the interaction of Be-bearing magmatic hydrothermal fluids, related to granitic-pegmatitic complex, with Cr- and V-bearing mafic and ultramafic metamorphic rocks within localized deformation zones. However, Zwaan (2006) noticed that a significant number of emerald deposits cannot be unambiguously classified using the existing schemes, and suggested that a future classification scheme should be based on the trace elements geochemistry of the gemstone. In order to individuate useful geochemical markers for provenance purpose, we present and discuss for the first time the chemical composition of both major and trace elements of selected emerald samples from some of the most important worldwide deposits. Electron Microprobe (EMP) and Secondary Ion Mass Spectrometry (SIMS) investigations allowed to determine major and trace elements concentrations used for binary and spider diagrams along with statistical analysis, i.e., Principal Component Analysis (PCA). Crossing the different results, we observed that major or trace elements considered separately did not give useful results in term of discrimination; indeed, we were able to discriminate each deposit with high reliability when both groups of elements were considered at once. In particular, PCA results identified different groups on the basis of their content of SiO 2, Al 2 O 3 , V, Sc, B, Li. Moreover, spider and binary diagrams involving Cs, Rb, B, Li, Cr, V, Sc highlighted peculiar differences inside the emerald deposits, and in particular for those deposits not discriminated by the PCA. References Barton M.D. and Young S., 2002. In: Grew, E.S. (Ed.), Beryllium: Mineralogy, Petrology, and Geochemistry. Reviews in Mineralogy and Geochemistry 50, 591–691. Dereppe J.M., Moreaux C., Chauvaux B., Schwarz D., 2000. Journal of Gemmology 27, 93–105. Gavrilenko E.V. and Dashevsky D.M., 1998. Proceedings of the Russian Mineralogical Society 127, 45–57. Schwarz, D. and Giuliani, G., 2001. Australian Gemmologist 21, 17–23. Schwarz D., Giuliani G., Grundmann G., Glas M., 2001. In: Schwarz, D., Hochlitner, R. (Eds.), Smaragd,derkostbarste Beryll, der teuerste Edelstein. ExtraLapis 21, 68–73. Zwaan J.C., 2006. Scripta Geologica, 131–211.