The Fakes: Diamond Simulants — and Why “Real vs. Fake” Is the Wrong Frame

Change log: Wave 4 first pass. A18 is the most “practical” of GIA’s assignment lists; every entry is an identification-method paper, so the module is organized as a property-ladder rather than an article-by-article recap.

Sixty years of look-alikes, in five sentences

The A18 timeline, read as a story of escalation: 1957 — Benson catalogs “diamond substitutes” (glass, strontium titanate, YAG-era materials: the soft-optics zoo the pre-CZ market traded); 1978 — Liddicoat & Koivula describe cubic zirconia arriving as a production material (“synthetic cubic stabilized zirconia”), not a lab curiosity; 1981 — Nassau updates CZ and Hobbs offers “a simple approach” to separation, the trade’s first mass-detection doctrine; 1990s — Kammerling 1991 examines non-transparent “CZ” from Russia (i.e., fakes outside the normal spec), and 1997 — Nassau et al. publish synthetic moissanite as a new diamond substitute: the first simulant to beat the thermal tester; 2000s–2010s — the arms race turns combinatorial: Shigley et al. 2012 characterize coated CZ (“Diamantine”) — a soft material wearing a thin diamond-like film — while Choudhary 2013 flags boron carbide imitating black diamond; Zhou 2015 and Song et al. 2016 document synthetic moissanite in melee — including moissanite made to imitate lab-grown colored diamonds, the substitution hiding inside a substitution. The pattern every one of these papers encodes: every detector is a property test, and every new material picks off the detectors one by one.

The property ladder (what a professional actually uses)

Simulant separation is not tricks; it’s a decision tree over physical properties (textbook values; ranges are standard-reference level — any number you print on a card):

Test	Diamond	CZ	Moissanite	What the A18 papers add
Read-through (facet junctions)	single, sharp	single, soft, heavy dispersion	doubles at table facets (0.02–0.05 DRL)	Nassau 1997’s shock: moissanite passes thermal probes → the doubling rule returns
Luster/edge wear	adamantine, crisp	greasier, abrades	harder than CZ, near-crisp	Hobbs 1981 “simple approach”: edges + wear before any electronics
Specific gravity	3.52	~5.6–6.0 (heavy in hand)	3.22 (light)	hydrostatic/thermal balance era
Electrical conductivity	insulator except blue IIb (M08)	insulator	semiconductor — most moissanite conducts	the tester-generation response; today’s multi-testers
UV fluorescence	varied	often strong chalky blue	often none/weak	supplementary only
Spectroscope/polariscope	single-refraction (except strain)	isotropic + Fe lines common	doubling, strong DR	classic separation before electronics (Benson 1957 era methods)

The screening narrative you can trust in a store (from GIA’s own verified video narration, GIA-03): “before even stepping into the four Cs, the diamond is tested to determine whether it is natural or lab-grown” — and every report system sits behind that gate. For simulant control the same logic applies one level down: thermal/electrical probes at intake, loupe double-check on anything the probe calls “pass,” and the polariscope/doubling rule as the tie-breaker when moissanite is in play.

The disclosure conversation: retire “fake,” install “different material”

The vocabulary problem is a liability problem. “Fake” accuses; it also teaches the customer that the category CZ exists to deceive — which your own jewelry case disproves (most CZ sold is disclosed fashion jewelry). The honest, sales-safe frame:

Simulant = a material that simulates diamond appearance (CZ, moissanite, glass, coated goods). It is a product, not a crime.
Synthetic (lab-grown) = a real diamond, made by another process (M11). Never call it “fake”; that’s legally imprecise and commercially wrong.
Natural diamond = the M02/M03 story; its trade advantage is provenance and resale norms, not “realness.”

So: the buyer of a CZ ring is buying look; the buyer of a lab-grown is buying material (identical physics, different market); the buyer of a natural is buying material + history. Every complaint the A18 papers exist to answer — undisclosed moissanite melee in a bracelet (Zhou 2015), a coated stone wearing two identities (Shigley 2012) — is a disclosure failure, not a material failure. That’s also the FTC-frame (Overton 2004, A19 list — M12). Your job on the floor is to make the three sentences above automatic.

The melee problem (what the 2013–2016 papers are really about)

Single-stone detection is solved; parcel and repair detection is the open wound the later A18 entries document: loose melee runs through the market untested; a repair or re-polish can swap one stone into a tray of twenty; the “Diamantine” and “synthetic moissanite” waves specifically target the colored melee niche where certificates never existed. Practice consequences:

Intake screening on parcels (Dieck et al. 2016, “Analysis of Yellow Diamond Melee for Color Treatment and Synthetics,” G&G 52:1 — title-level here): if your shop repairs and a tray looks too clean/too sparkly for its era, screen it.
The loupe at 10× finds the two big tells: facet-junction doubling (moissanite), and lead-light orange-flash in fracture/rim (filled or coated). Both are A18/A19 paper abstracts’ exact themes.
Tell the repair story: customers should hear “we screen every stone we touch,” which is precisely the promise the GIA video’s gate makes at lab scale.

Self-check

Order the five properties that separate CZ from diamond by the loupe route (no electronics).
Why did moissanite “break” the industry default tool, and what replaced it? (thermal → conductivity; multi-testers)
Name the two A18-era combination frauds and their telltales. (coated CZ — flash/abrasion, interferometry at lab; moissanite-as-synthetic-colony — polarization/doubling + tester cross-check)
Give the three-sentence disclosure frame in your own words (simulant/synthetic/natural).
Which is riskier for a repair shop: a loose 2 ct round or a sealed bracelet of melee — and why? (the melee: no cert, untested intake, swap opportunity)