corundum, chemistry.2, synthetic corundum.7, gems

Investigates the use of proton-induced X-ray emission (PIXE) to distinguish natural from synthetic ruby and assess relationships between trace-element composition and source. The study analyzes 160 specimens comprising 133 natural rubies from Afghanistan, Burma, two Indian localities, Kenya, Sri Lanka and Thailand, together with 27 synthetic rubies representing Chatham, Inamori, Kashan, Knischka, Ramaura, Seiko and an unidentified flame-fusion source. Some natural origins were supported by characteristic inclusions, whereas others depended on supplier information because the specimens were insufficiently transparent for microscopic verification. Measurements employed 2 MeV protons from a Van de Graaff accelerator, a silicon-lithium detector and a counting time of 150 seconds per specimen. Spectra were fitted using the AXIL program, with calibration standards enabling quantitative determination of trace-element concentrations relative to aluminum oxide. Natural rubies generally contained a greater variety and higher concentrations of impurities than synthetic specimens. Besides chromium, detected elements included silicon, chlorine, sulfur, potassium, calcium, titanium, vanadium, manganese, iron and gallium. Vanadium was undetectable in every synthetic specimen but occurred in most natural rubies, with exceptions among Thai and Indian samples. The detection limit was approximately 20 ppm in the absence of titanium. Burmese rubies had the highest mean vanadium concentration, with one specimen reaching 1800 ppm, while Thai rubies had the lowest mean. Consequently, detectable vanadium strongly supported natural origin within the studied population, although its absence did not establish synthetic origin. Most synthetic rubies contained little iron, with the Knischka specimen providing an exception; Thai rubies and some Sri Lankan specimens were comparatively iron-rich. Chromium concentrations varied considerably between groups but were relatively consistent within several individual sources, suggesting additional discriminatory value. The authors emphasize that source attribution requires combinations of elements rather than a single concentration. For example, high iron together with low chromium and vanadium favored a Sri Lankan attribution within their dataset. Titanium, calcium and potassium also offered potentially useful information in some cases. The results agree with published analyses obtained by energy-dispersive X-ray fluorescence and neutron activation. PIXE is presented as a rapid and useful identification technique, and the authors recommend developing a larger database of specimens with known origins. They nevertheless caution that synthetic ruby manufacturers could modify product compositions, potentially reducing the future reliability of distinctions based on trace-element chemistry.