Gems, Minerals, and Artefacts: Geological Provenance

Gemstone provenance rests on two lines of evidence used in combination. The first is the inclusion assemblage: the mineral grains, fluid pockets, and negative crystals trapped inside the gem during growth. The minerals that were stable at the same time and place as the gem crystallised are geologically specific. Mogok rubies from the Mandalay region of Myanmar contain calcite, rutile silk, and spinel inclusions in a marble host-rock geochemistry; Vietnamese Luc Yen rubies contain different inclusion types reflecting a different metamorphic environment; Thai/Cambodian rubies from the Pailin region are basalt-hosted and contain zircon, corundum, and ilmenite inclusions that marble-type deposits lack.

The second line is trace-element profiling by LA-ICP-MS. A laser pulse ablates a 50-micrometre spot on the gem surface, and the plasma is measured for a suite of elements: chromium, iron, vanadium, gallium, titanium, magnesium, and others. Each deposit type has a characteristic multi-element pattern that differs from other deposits in the same gem species. Combining the two lines of evidence allows a gemmologist and geochemist working together to assign a ruby, sapphire, or emerald to a deposit of origin with a stated confidence level.

Diamonds pose the hardest provenance problem in gemstone forensics because they are elemental carbon, nearly pure, with very few trace elements to measure. The main diagnostic approaches are: nitrogen aggregation state (the proportion of nitrogen in A-form versus B-form aggregates, which reflects the temperature and duration of mantle residence), inclusion mineralogy (eclogitic versus peridotitic diamond types have different inclusion suites), carbon isotope ratios (d13C), and physical morphology (alluvial versus kimberlite-mined diamonds have different surface textures).

The Kimberley Process Certification Scheme covers rough diamonds only. Once a diamond is cut and polished, the surface texture record of its mining history is largely lost, and the internal inclusion pattern is the only remaining geological evidence. The scheme has documented weaknesses: large producing countries can certify parcels that include artisanal diamonds of uncertain origin, and internal conflict stones in some countries escape the certification net entirely. Forensic geology provides an independent provenance check, but the reference databases for diamond provenance are less complete than those for coloured stones.

Gold and silver artefacts and bars carry two types of geological information. First, trace-element profiles: the concentrations of platinum-group elements (Os, Ir, Ru, Rh) and other siderophiles in native gold vary with the mineralisation style and host geology. Orogenic gold deposits in Archaean greenstone belts have different PGE ratios from epithermal gold deposits in Tertiary volcanic arcs. These ratios are preserved through smelting as long as no alloying metals are added.

Second, lead isotope ratios: both gold and silver ores contain galena (lead sulfide) or lead-bearing phases whose isotope ratios reflect the geological age and uranium-thorium content of the ore-forming environment. Different mining districts around the world occupy distinct positions in 207Pb/204Pb versus 206Pb/204Pb space. A smelted bar from artisanal gold mining in eastern DRC falls in a different isotope field from gold smelted in South Africa or Australia. This is the basis of provenance attribution for conflict gold that has been refined and is now trading as apparently legitimate bullion.

The meteorite trade includes both genuine objects and fraudulent terrestrial rocks sold as meteorites. Authentication uses a combination of physical, petrographic, and geochemical evidence. Physical features include fusion crust (a glassy quench layer from atmospheric entry), regmaglypts (thumb-print-like surface depressions from ablation), and for iron meteorites, the Widmanstatten pattern of intergrown kamacite and taenite lamellae visible after polishing and etching with dilute nitric acid.

In practice, nickel content is the quickest screen: virtually all genuine meteorites contain nickel above levels found in any terrestrial iron oxide mineral (magnetite, hematite, goethite) because planetary differentiation concentrated siderophile elements in metallic phases. A handheld XRF reading of less than 1% Ni in a supposed iron meteorite is a strong warning flag. Definitive authentication requires thin-section petrography (for chondrule identification and mineralogy) and ICP-MS or INAA for platinum-group element ratios.

Obsidian is volcanic glass with a trace-element composition that is specific to each volcanic flow at each source. Trace elements (especially the high-field-strength elements: Zr, Nb, Y, Rb, Sr, Ba) are set at the time the lava cooled and do not change with subsequent weathering or burial. XRF or ICP-MS measurement of a small sample produces a multi-element fingerprint that can be compared against a global database of obsidian source geochemistry.

In a forensic context, obsidian sourcing arises in cultural property cases involving lithic artefacts claimed to have been legally exported from one country but actually originating from protected sites in another. The same sourcing methods apply to flint and chert, though these are less precisely discriminated because flint-forming environments are more geographically variable and less geochemically distinctive than volcanic flows. Silicon isotopes and strontium isotope ratios provide supplementary discrimination for flint provenance.

All gemstone and mineral provenance methods rely on comparison against a reference database of well-characterised samples from known localities. The quality of that database determines the quality of the attribution. For rubies and sapphires, commercial laboratories (Gübelin Gem Lab in Switzerland, SSEF Swiss Gemmological Institute, GIA in the United States) maintain proprietary reference collections built over decades. For gold, academic and government-funded databases cover major mining districts in Africa, South America, and Australia. For obsidian, the global database maintained at multiple research institutions now covers most known source flows in the Americas, Mediterranean, and Pacific.

Gaps in the database are the main limitation. Newly discovered or artisanal deposits may not be in the reference set, which means an unusual sample falls outside all known clusters. The analyst can say where the object is inconsistent with, but cannot positively assign it to an unlisted source. Expanding reference databases for commercially exploited gemstone deposits is an active research area with direct forensic applications.