Identification & instruments

Magnification is the first test and the one used most. The standard loupe is a 10× triplet: three cemented lenses that correct color fringing (chromatic aberration) and edge blur (spherical aberration). Ten power is the reference magnification for diamond clarity grading. The US Federal Trade Commission ties the word “flawless” to a stone that shows no blemishes when examined under a corrected magnifier at 10 power, with adequate illumination, by a person skilled in diamond grading. Stronger hand lenses have narrower fields of view and shallower focus, so they are harder to use.

The gemological microscope is a binocular stereo microscope with lighting designed for gems. In darkfield illumination a baffle blocks direct light so inclusions shine against a black background. Brightfield shows low-relief features and curved growth lines as dark shapes, and overhead or fiber-optic light reveals polish marks and pits. Immersing a stone in liquid cuts surface reflections and makes color zoning easier to see. GIA notes that what the microscope yields depends on the operator, who needs years of training.

Gemologists look for diagnostic features: curved striae and gas bubbles in flame-fusion synthetic corundum, flux fingerprints and platinum platelets in flux-grown stones, doubled facet junctions in synthetic moissanite seen when the stone is tilted, and flash effects along filled fractures in treated diamonds.

The refractometer measures refractive index (RI) by critical angle. A flat facet rests on a dense glass hemicylinder, joined by a drop of contact liquid; the boundary between light and dark on the internal scale marks the stone's RI, read under yellow light. A doubly refractive stone gives two readings as it is turned; their difference is the birefringence, and the way they move shows whether the stone is uniaxial or biaxial. Curved cabochons give a blurred spot reading instead. GIA's contact liquid has an RI of 1.81, and nothing above it can be read, so demantoid, high zircon, cubic zirconia, diamond and garnet at the top of its range show no edge and are recorded as over the limit.

The polariscope places a gem between crossed polarizing filters. A doubly refractive stone goes dark every 90°; a microcrystalline aggregate such as chalcedony stays light. A stone dark through one full turn is not proved singly refractive: a doubly refractive stone looked at down its optic axis does the same, so it must be dark in three positions. Internal strain can make a singly refractive stone show anomalous double refraction, which turning the analyzer separates from true double refraction. A conoscope adds interference figures that tell uniaxial stones from biaxial ones.

The dichroscope shows pleochroic colors side by side: two in a uniaxial stone, three in a biaxial one, none in a singly refractive one.

The hand spectroscope spreads light that has passed through a gem into a spectrum, where absorbed wavelengths appear as dark lines and bands. Prism models stretch the blue end; diffraction-grating models spread the visible range more evenly. Several patterns are diagnostic: chromium gives ruby a narrow line at 694 nm, iron in corundum produces features near 377, 388 and 450 nm once the iron content passes about 150 ppma, and zircon shows a series of lines gemologists call the organ-pipe spectrum. GIA calls the hand spectroscope one of the hardest portable instruments to master.

The Chelsea filter passes light in two narrow bands and is used to screen certain green, red and blue stones and to detect dyes in chalcedony. Filters of this kind assist identification but are not used for a final determination.

Ultraviolet lamps test fluorescence and phosphorescence, long-wave at 365 nm and short-wave at 254 nm. An inert response does not prove a stone untreated, and luminescence is only semiquantitative.

The hydrostatic balance measures specific gravity by Archimedes' principle: weight in air divided by loss of weight in water. A drop of detergent lowers surface tension. The method is slow and unreliable for very small or porous stones, but it separates look-alikes of different density.

Handheld thermal testers are usually said to measure thermal conductivity, but what they read is thermal inertia, the rate at which a material draws heat from the hot tip. Cubic zirconia, glass, YAG, GGG, strontium titanate and colorless sapphire take up heat slowly and fail; diamond passes.

Synthetic moissanite defeated the test in the late 1990s: GIA reported that thermal probes react to it as if it were diamond. GIA gives diamond a thermal inertia of 0.55 to 1.7 and moissanite 1.6 to 4.8 in cal/cm C, ranges that overlap. Later multi-testers added electrical conductivity, since moissanite is a semiconductor. That cross-check has its own trap: boron-bearing type IIb diamond conducts too, and GIA reports that over 80% of colorless HPHT laboratory-grown diamonds submitted to it since 2020 are type IIb, so a conductive reading is not proof of moissanite. A loupe helps only if used correctly. Moissanite is cut with its c-axis perpendicular to the table, so face-up it shows no doubling; tilt it and look in at an angle and the back facet junctions double. It is also lighter, 3.22 against diamond’s 3.52.

No probe separates natural from laboratory-grown diamond, since both are the same carbon crystal. Spectroscopic screening is needed. GIA’s iD100 tests loose or mounted colorless to near-colorless stones of 0.9 mm and larger in under two seconds and returns “pass” or “refer”; GIA states that no rapid, low-cost tool identifies every laboratory-grown diamond.

Major laboratories add instruments that probe structure and chemistry, most of them nondestructive.

FTIR (Fourier-transform infrared) spectroscopy records infrared absorption. It establishes diamond type by detecting nitrogen and boron, and reveals polymer impregnation in jade and evidence of some treatments in corundum. UV-Vis-NIR spectroscopy measures absorption from ultraviolet to near infrared to establish the cause of color; stones are cooled with liquid nitrogen to sharpen their spectra. Raman spectroscopy aims a laser at a spot and reads small energy shifts in the scattered light, a vibrational fingerprint that identifies a gem or an inclusion below the surface, and it works best on transparent to translucent material. Photoluminescence (PL) spectroscopy, also laser-excited and usually run at liquid-nitrogen temperature, detects defects present in parts per billion and is central to separating natural, laboratory-grown and treated diamonds. The DiamondView images luminescence under ultraviolet light below 225 nm to show growth patterns.

For chemistry, EDXRF (energy-dispersive X-ray fluorescence) reads the characteristic X-rays of each element but cannot detect anything lighter than sodium, so it misses beryllium in corundum. LA-ICP-MS (laser ablation inductively coupled plasma mass spectrometry) vaporizes a speck, leaving a crater about 50 µm across, usually on the girdle, and measures most trace elements below 1 ppm and heavy elements to parts per billion. For pearls, X-ray radiography and micro-CT image internal structure, though mounted pearls generally have to be removed first.

A gem's trace elements reflect the rocks and fluids it grew from, so laboratories compare a stone's chemistry, inclusions and spectra with reference samples of known origin. The approach expanded in the early 2000s, when laboratories adopted LA-ICP-MS to detect beryllium diffusion in corundum and found the same data useful for telling deposits apart.

For blue sapphire GIA uses magnesium, titanium, vanadium, iron and gallium. It screens first with the 880 nm absorption band, which separates metamorphic from basalt-related stones, then plots the unknown against reference stones of similar composition, and accepts an origin only when inclusions, chemistry and spectra all agree. For emerald the element set GIA works with is lithium, potassium, vanadium, chromium, iron, rubidium and caesium. Peucat and colleagues showed in 2007 that magmatic sapphires carry high gallium and low magnesium, giving a Ga/Mg ratio above 10, while metamorphic sapphires fall below it.

Inclusions add a time dimension. SSEF's GemTOF, a time-of-flight LA-ICP-MS installed in July 2016, records nearly the whole periodic table at once and can date zircon inclusions by uranium and lead. Raman spectra of those inclusions help as well: zircons in geologically young Kashmir sapphires give narrow peaks, while the older, more metamict zircons in Madagascar sapphires give broad bands.

The method has real limits. GIA reports that stones from different localities often overlap so far that origin cannot be determined, and issues an inconclusive result rather than guess.