USE OF LASER-INDUCED BREAKDOWN SPECTROSCOPY (LIBS) AND CHEMOMETRICS TO SOLVE COMPLEX GEOCHEMICAL PROBLEMS: THE CASE OF EMERALD PROVENANCE

While many highly accurate and precise analytical techniques are available, many geochemical problems, such as correlation and provenance, are difficult to resolve using traditional techniques such as X-ray Fluorescence (XRF) and Laser Ablation Inductively Coupled Mass Spectrometry (LA-ICP-MS) because the geochemical complexity inherent in rocks and minerals exceeds the number of elements determined by these methods. The determination of provenance (country of origin) for precious gems, including emeralds has been an active area of research for decades because: 1) gem provenance is a factor that affects gem value, and 2) gems have been used for centuries to transport large value in small packages for legitimate and illegitimate purposes. Gem research using traditional techniques employs the concentrations of several (4-15) elements. This number of variables is insufficient to capture the chemical variation in a single gem or in a single gem deposit, leading to provenance models that are unable to uniquely identify individual deposits. Laser Induced Breakdown Spectroscopy (LIBS) is a relatively new geochemical analytical technique that collects the spectra of photon emission from a laser ablation plasma. The intensities of light with wavelengths between 200 and 1000 nm are collected; all elements in the periodic table emit in this region, resulting in a chemically complex fingerprint that records the concentrations of major, minor, and trace elements in the sample, as well as spectral and matrix interferences. LIBS spectra are analyzed using multivariate analysis, which takes advantage of the large amount of information in each spectrum. In this study, 300 emerald samples from 10 unique locations were analyzed by LIBS. Spectra were analyzed by M2S multivariate analysis, which builds a compilation of spectra, or sequence, for each sample. Sample sequences form a reference library, and spectra of unknown samples are compared to the library using signal processing techniques. The emerald provenance determination success rates are 98%, far exceeding attempts at provenance determination using other techniques. This work suggests that LIBS data can be used to provide answers previously unavailable to the geological community.