DETERMINING THE PROVENANCE OF GEMSTONES USING
THE MATERIALYTICS SEQUENCING SYSTEM (M2S)
A new quantitative process, the Materialytics Sequencing System (M2S), increases accuracy in identifying the provenance of certain geomaterials, including (but not limited to) beryl, corundum, gold, and tourmaline. Using laser-induced breakdown spectroscopy (LIBS) and unique pattern-recognition software, this system can match these geomaterials to their country of origin with greater than 95% accuracy.
An extensive inventory of specimens—more than 42,000 and growing—has been collected and maintained; the integrity of this sample set is crucial to provenance identification. The MaterialyticsSM inventory contains samples from six continents, more than 60 countries, and 225 localities in all.
The M2S system consists of dual Nd:YAG lasers (266 and 1064 nm), an intensified CCD camera, an Echelle spectrometer, and computer software designed to maximize the signal obtained from a sample while minimizing the traces of testing. The laser ablates a small area of the sample, producing a microplasma that emits light in the 200–900 nm range. This light is collected by more than 40,000 channels of the spectrometer. These 40,000 dimensions are then analyzed using pattern recognition software that compares the signature of an unknown specimen to a large database of signatures from thoroughly documented specimens.
The interplay between the variability and commonality of the data from crystal to crystal, mine to mine, mining region to mining region, and country to country requires a combination of statistical pattern recognition methods (principal components analysis, maximum redundancy analysis, partial least squares, and Bayesian networks) to achieve consistent results over a wide range of samples. M2S integrates these various methods into a single computer program that adapts to each sample, based on the nature of the data, and does not require any human intervention.
Date Presented: May, 2011
Conference Presented: GIA Symposium 2011
1. Catherine McManus, Tristan Likes
2. Nancy McMillan, Kristen Yetter
3. Jim Dowe
4. Michael Wise
5. Steve Buckley, Gregg Lithgow, Chris Stipes
6. Peter Torrione
1. Materialytics, LLC, P.O. Box 10988, Killeen, TX 76547
2. Geological Sciences, New Mexico State University, Las Cruces, NM 88003
3. Analytical Data Services, Harker Heights, TX 76548
4. Smithsonian Institution National Museum of Natural History, Washington, DC 20560
5. Photon Machines Inc., Redmond, WA 98052
6. New Folder Consulting, Durham, 27712