Gemstone Treatments

Heat treatment, filling, diffusion, coating, and detection methods for treated gemstones – with per-species diagnostic depth.

By gemmology.dev editors Last updated
heat-treatment market/disclosure detection laboratory

Introduction

Most coloured gemstones on the market today have undergone some form of treatment
to enhance their appearance. Understanding treatments is essential for accurate
identification and proper disclosure.

Heat Treatment of Corundum

The most common treatment. Heat permanently alters colour and clarity by modifying
chromophore chemistry or dissolving silk inclusions. Plain heating (H) and flux-assisted
heating with borax (H(a)) are now classified separately by Gem-A and major laboratories.

Common Heat Treatments
Gemstone Effect Temperature Range
Blue sapphire Improves blue colour, dissolves silk 1400–1800 °C
Ruby Improves red colour, dissolves silk 1200–1800 °C
Tanzanite Changes brown to blue-violet 450–650 °C
Aquamarine Removes green tint (yellow component) 400–450 °C
Citrine Heat-treated amethyst 450–500 °C
Zircon Creates blue colour from brown 900–1000 °C

Diagnostic Inclusion Changes (Darkfield 40–60×)

  • Dissolved silkrutile needles partially resorbed; appear as dotted trails
    or "fingerprint" remnants along former needle orientation; diagnostic for heat
    above 1400 °C
  • Discoid (halo) fractures – disc-shaped stress fractures around zircon,
    apatite, or calcite inclusions from thermal expansion; highly diagnostic for
    heat above 1600 °C
  • Flux residue droplets – colourless rounded glass blebs at fracture mouths
    (borax); show anomalous birefringence under crossed polars; diagnostic for
    H(a) flux-healing treatment
  • Partially healed fractures – feathered, partially recrystallised fractures
    with residual fluid inclusions; distinct from natural growth features

FTIR Evidence for Unheated Sapphire

Goethite (dehydrates 300–325 °C) and diaspore (dehydrates 525–550 °C) have
diagnostic OH absorption peaks at ~3300–3400 cm⁻¹ that vanish on heating.
Their absence in stones from origins known to contain them is strong evidence
for heat treatment (Krzemnicki et al. 2023, 10.3390/min13121557, [VERIFIED]).

A 3232 cm⁻¹ FTIR band identifies heat-treated metamorphic-type blue sapphires
(Delaunay 2024, 10.15506/jog.2024.39.1.33, [VERIFIED]).

Disclosure Status

  • H (plain heat): widely accepted; CIBJO and Gem-A require disclosure but value
    impact is minimal for corundum
  • H(a) (flux/fracture healing): must be disclosed separately; value impact
    is greater; GIA and Gem-A treat this as distinct from simple heating
  • "No evidence of heat treatment" (NTE) on a laboratory report commands a
    significant premium; verify by the inclusion criteria above

Beryllium Diffusion – Lattice Diffusion (Sapphire)

Commercialised around 2001. Beryllium (Be²⁺) ions diffuse through the full corundum
lattice at 1700–1800 °C producing intense orange, yellow, or padparadscha colours.
Distinct from titanium surface diffusion in penetration depth.

Process Overview

Rough is packed in BeO or chrysoberyl powder and fired at 1700–1800 °C in an
oxidising atmosphere for 12–100 hours. Be²⁺ (ionic radius 0.27 Å) penetrates
the full lattice, unlike Ti or Cr which remain within ~50–100 µm of the surface.
Colour is lattice-deep – re-cutting does not remove it.

Sources: Emmett et al. 2003 (10.5741/gems.39.2.84, [VERIFIED]);
Emmett et al. 2023 (10.5741/gems.59.3.268, [VERIFIED])

Key Detection Points

  • 10× loupe: colour may appear suspiciously homogeneous; immersion spider-web
    is faint (if present) – less pronounced than Ti surface diffusion
  • LA-ICP-MS: Be >1–2 ppm in orange/yellow corundum is diagnostic; natural
    corundum contains <0.1 ppm Be
  • SIMS: diffusion gradient (high at surface, declining inward) confirms treatment
  • Stability: permanent; colour survives repolishing, ultrasonic, and steam
  • Disclosure: AGTA code U; GIA reports "lattice diffusion – beryllium present"

Surface Diffusion vs Lattice Diffusion – The Depth Distinction

The critical diagnostic difference between titanium surface diffusion (Ti, pre-2001,
now rare) and beryllium lattice diffusion is colour penetration depth and the resulting
immersion microscopy appearance.

Surface vs Lattice Diffusion – Key Differences
Property Ti Surface Diffusion Be Lattice Diffusion
Penetration depth 50–100 µm (skin only) Full lattice depth (mm scale)
Immersion spider-web pattern Strongly visible under darkfield Faint or absent
Re-cutting effect Removes colour on any recut facet No effect on colour
Definitive detection Immersion microscope (practical) LA-ICP-MS / SIMS required
Stability Fragile to repolishing Permanent
AGTA code U U

Spider-Web Test (Ti Surface Diffusion)

Immerse the stone in di-iodomethane (or water) and observe under darkfield illumination
at 40–60× magnification. In titanium surface-diffused sapphire, colour accumulates as
a visible network following facet edges – the "spider-web". Where two surfaces meet,
the diffused zone doubles in thickness and appears darker.

Scratches, chips, or abrasions on facets will expose colourless material beneath.
Any repolishing through a facet removes colour entirely from that facet.

Source: Kane et al. 1990 (10.5741/gems.26.2.115, [VERIFIED])

Fracture Filling of Emerald

Emeralds typically contain abundant surface-reaching fractures ("jardin"). Cedar oil
(RI ~1.51) and Opticon resin (RI ~1.55) are the most common fillers. They are detected
by the flash effect under darkfield and FTIR spectroscopy.

Detection Methods

Emerald Fracture Filling Detection
Method Finding
10× loupe, reflected light Lustre difference at fracture mouths; slightly vitreous vs adamantine
Darkfield microscope (40×) – flash effect Orange/yellowish-orange iridescent flash at filler–crystal interface when tilted (primary in-lab test)
Gas bubbles / flow structures Spherical bubbles or swirling patterns in polymer; absent in natural fluid inclusions
UV fluorescence (LWUV/SWUV) Opticon and resins: chalky-blue or yellowish-green glow; cedar oil: negligible fluorescence
FTIR spectroscopy Resins: C–H stretching ~3000–3050 cm⁻¹; C=O stretching ~1700–1730 cm⁻¹; unfilled emerald shows no organic absorptions – definitive

LMHC Fracture-Fill Disclosure Scale

The Laboratory Manual Harmonisation Committee (LMHC) has established a harmonised
grading scale for emerald fracture filling [PARTIALLY_SUPPORTED – institutional document;
not API-retrievable as peer-reviewed paper; cite via https://www.lmhc-gemmology.org]:

  • F1 (None): no filler; unfilled stone
  • F2 (Minor/Insignificant): trace filler; no material effect on appearance
  • F3 (Moderate): filler clearly present; requires disclosure
  • F4 (Significant): substantial filling; strongly impacts apparent clarity
  • F5 (Prominent/Prominent+): extreme filling; stone may approach composite character

Cedar oil at minor level is widely accepted but Gem-A and CIBJO now require disclosure
of all fillings. Polymer (Opticon, epoxy) filling always requires disclosure.

Stability and Care

  • Cedar oil: temporary; dries out over months to years; avoid ultrasonic, steam, solvents
  • Opticon/epoxy: more stable; dissolved by acetone, alcohol; avoid ultrasonic, torch
  • Advise all clients to avoid ultrasonic, steam, and alcohol-based cleaners for filled emeralds

Lead Glass-Filled Ruby (Composite Ruby)

Key Properties and Detection Summary

  • SG: depressed from natural ruby ~4.00 to composite values as low as 3.60–3.80
  • Blue/orange flash effect (darkfield, 40–60×): vivid blue flash one way, orange
    flash the other, at the glass–ruby interface – the most reliable in-lab indicator
  • Gas bubbles: spherical bubbles in glass fill; never in natural growth features
  • SW UV: lead glass fluoresces chalky greenish – abnormal for ruby inclusions
  • EDXRF: elevated Pb at surface is diagnostic; non-destructive, rapid
  • Acid test (destructive): lemon juice etches glass within minutes; corundum unaffected

Source: McClure et al. 2006 (10.5741/gems.42.1.22, [VERIFIED])

HPHT Diamond Colour Treatment

HPHT annealing (5–6 GPa, 1700–2100 °C) can decolourise brown Type IIa diamonds to
near-colourless (D–H) or create fancy colours in other types. Cannot be detected by
standard tools – laboratory testing is required.

Key Signatures

  • Absent LWUV blue fluorescence: ~80% of natural gem diamonds show blue LWUV;
    HPHT-treated type IIa are typically inert – primary screening trigger
  • FTIR: type IIa signature (no N >5 ppm); 270 nm brown absorption absent after
    treatment; 1344 cm⁻¹ isolated-N peak may appear in partially decoloured stones
  • DiamondView: cross-hatched or irregular green sectors (modified octahedral growth);
    distinct from CVD columnar striations
  • PL at 77 K: NV⁻ (637 nm) strong relative to NV⁰ (575 nm) in treated type IIa;
    H3 (503 nm) may be prominent in type IaB treated stones
  • Stability: permanent; disclose with AGTA code HPHT

Sources: Fisher & Spits 2000 (10.5741/gems.36.1.42, [VERIFIED]);
Eaton-Magana et al. 2017 (10.5741/gems.53.3.262, [VERIFIED])

CVD Diamond Detection

CVD synthetic diamonds grow in columnar layers from carbon-bearing gas plasma at low
pressure and 700–1000 °C. They are typically Type IIa and require specialist laboratory
tools to distinguish from natural or HPHT-treated stones.

Key Diagnostic Features

  • DiamondView (225 nm SW UV): orange-red phosphorescence (diagnostic when present);
    columnar/striated growth threads – no octahedral growth sectors
    (Zhang et al. 2024, 10.3390/cryst14090804, [VERIFIED] – resolves VERIFIED.md F-03 flag)
  • PL at 77 K: SiV⁻ doublet at 736.9/736.6 nm characteristic of CVD growth;
    rarely present in natural or HPHT-treated stones
  • FTIR: Type IIa signature; NV-H (3123 cm⁻¹) in some N-doped CVD samples
  • Disclosure: must be disclosed as "synthetic diamond" or "laboratory-grown diamond";
    "cultured" or "cultivated" are not acceptable per CIBJO

Sources: Martineau et al. 2004 (10.5741/gems.40.1.2, [VERIFIED]);
Zhang et al. 2024 (10.3390/cryst14090804, [VERIFIED])

Pearl Colour Treatment

Pearl colour treatments include gamma irradiation (darkening the bead nucleus via MnCO₃
oxidation) and dyeing with organic dyes or silver nitrate. Irradiated Akoya pearls are
the most commercially significant treatment requiring detection.

Gamma Irradiation (Akoya)

Pearls are exposed to ⁶⁰Co gamma-ray sources at 0.1–100 kGy. Radiation oxidises
MnCO₃ in the bead nucleus to MnO₂ (dark brown/black), creating a dark body colour
visible through the nacre. The radiation also denatures conchiolin, detectable by ESR.

ESR (Electron Spin Resonance) detects the CO₂⁻ radical (g-factor 2.001 ± 0.002)
from radiation damage to carbonate – the gold-standard diagnostic for irradiation.
LWUV fluorescence: irradiated Akoya show no/very weak fluorescence (radiation
damages conchiolin fluorophore); natural black Tahitian show reddish-pink/red glow.
EDXRF: elevated Ag = silver-nitrate dye; Mn/Fe nucleus profiling for irradiation.
X-radiography: dark nucleus in irradiated Akoya vs involvement of nacre and nucleus
in naturally dark Tahitian pearls.

Source: Kim et al. 2012 (10.5741/gems.48.4.292, [VERIFIED])

Silver Nitrate Dyeing

Pearls are immersed in AgNO₃ solution then exposed to H₂S gas or sunlight,
precipitating Ag₂S (silver sulphide) in the nacre layers, producing blue-grey to
black colour. [PARTIALLY_SUPPORTED – Ag₂S mechanism is established trade knowledge;
the specific precipitation chemistry details are not independently confirmed to
peer-reviewed paper level at this time; present as trade knowledge.]

Detection: EDXRF detects elevated Ag; drill-hole inspection shows dye concentration
at and around the entry point.

Source: Karampelas et al. 2007 (10.12681/bgsg.16720, [VERIFIED])

Disclosure and Stability

  • Disclosure: CIBJO and Gem-A require disclosure of all colour treatments in pearls;
    irradiated Akoya sold as Tahitian substitutes constitutes fraud
  • Irradiation stability: permanent at normal temperatures; some fading in prolonged
    strong UV; nucleus darkening is stable
  • Organic dyes: fade in light, heat, sweat, cosmetics; variable stability
  • Silver nitrate: Ag₂S relatively stable; altered by strong reducing agents
  • Avoid ultrasonic, bleaches, steam for all treated pearls

Jade Treatments (A-jade, B-jade, C-jade)

Jadeite treatments range from traditional waxing (A-jade, accepted) to aggressive
bleaching, polymer impregnation (B-jade), and dyeing (C-jade). The A/B/C classification
is standard in the trade.

Jadeite Treatment Classification
Type Treatment Status
A-jade Untreated (waxing accepted) Natural; highest commercial value
B-jade Bleached in acid + vacuum polymer impregnation Treated; must be disclosed
C-jade Dyed (Cr-green or other pigments) Treated; must be disclosed
B+C-jade Both bleaching/impregnation and dyeing Treated; must be disclosed

Detection

B-jade process: jadeite is immersed in dilute HCl or H₂SO₄ for days to weeks
(bleaching), then vacuum-impregnated with epoxy resin (restores structural integrity).
The acid permanently damages the interlocking grain structure.

Detection methods:

  • 10× loupe: polymer filaments or threads along grain boundaries (reticulate pattern
    absent in A-jade); C-jade shows colour concentrated in fractures and grain boundaries
  • SW UV: B-jade shows chalky blue fluorescence from polymer; A-jade inert or faint
    whitish glow
  • Chelsea filter: C-jade (Cr-green dye) reacts red; A-jade (Fe-green) does not
  • FTIR (definitive): C–H stretching at 2800–3000 cm⁻¹; C=O at ~1700 cm⁻¹;
    absent in untreated jadeite

Source: Fritsch et al. 1992 (10.5741/gems.28.3.176, [VERIFIED])

Disclosure and Stability

  • CIBJO: "jade" without qualification implies A-jade; B and C must specify treatment
  • AGTA codes: B (bleaching), D (dyeing), F (filling) as applicable
  • Gem-A: "treated jadeite" with treatment type specified
  • B-jade: avoid ultrasonic (acid-damaged matrix permanently weakened); solvents dissolve
    polymer; C-jade organic dyes may fade in light

Thin-Film Coating (Mystic Topaz)

Colourless topaz is coated with a multilayer TiO₂/SiO₂/Nb₂O₅ thin-film stack by PVD
(Physical Vapour Deposition) at low temperatures. The multilayer interference film produces
vivid iridescent "rainbow" colours varying with viewing angle.

[PARTIALLY_SUPPORTED – mystic topaz TiO₂/SiO₂ multilayer PVD specifics are from industry
literature; Shigley et al. 2012 (10.5741/gems.48.1.18, [VERIFIED]) confirms the detection
principle and coated gem methodology via study of coated CZ (Diamantine).]

Detection and Stability

Detection:

  • 10× loupe: visible chipping or abrasion at facet junctions; underlying colourless
    topaz visible through scratches
  • Angle-dependent colour shift: colour pattern shifts predictably with angle
    (thin-film interference) – different from random play-of-colour of precious opal
  • Acetone swab (cautious, on pavilion): some coatings lift on contact
  • EDXRF: anomalous Ti, Nb, or Si at the surface – topaz itself contains neither

Stability:

  • Poor wear resistance; coating hardness lower than topaz (Mohs 8);
    daily wear damages coating at exposed facet edges; avoid ultrasonic, steam, solvents
  • Disclosure: CIBJO requires "coated"; AGTA code C; GIA: "surface coating present"

Irradiation

Exposure to radiation (gamma rays, electrons, neutrons) creates or modifies colour.

Gemstone Starting Colour Result
Blue topaz Colourless Blue (followed by heat annealing for stability)
Fancy diamond Various Green, blue, yellow, pink
Kunzite Pink Deeper pink (may fade in light)
Smoky quartz Colourless Brown/smoky
Akoya pearl Light cream Dark grey to black (nucleus darkening)

Quench-Crackling of Quartz

Rock crystal (colourless quartz) is heated to 300–400 °C then rapidly quenched in cold
water or a dye solution. Thermal shock creates a dense internal fracture network
which absorbs dye. Sold as "cherry quartz", "blue quartz", or "crackled quartz".

Detection

  • 10× loupe: dense, uniform, pervasive fracture network with no preferred
    crystallographic orientation – diagnostic; contrast with sparse, irregular,
    curved natural quartz fractures
  • Transmitted light: dye visible only in fractures; quartz crystals between
    fractures remain colourless; no natural quartz variety has colour only in fractures
  • UV fluorescence: organic dyes may fluoresce; natural quartz inclusions do not
  • FTIR / Raman: organic C–H peaks at ~3000 cm⁻¹; dye-specific peaks; pure
    quartz shows no organic absorptions

This material is an imitation/simulation: CIBJO requires "dyed crackled quartz"
or "quench-crackled quartz" – not a natural quartz variety designation.

Treatment Acceptance Summary

Widely Accepted

  • Heat treatment (corundum)
  • Oil in emerald (cedar oil, minor)
  • Waxing (turquoise, jade)

Accepted with Full Disclosure

  • Resin filling (polymer) – emerald
  • Beryllium lattice diffusion
  • Irradiation (most types)
  • HPHT diamond (permanent – must disclose)

Controversial / Explicit Disclosure Required

  • Lead glass filling (composite ruby)
  • Ti surface diffusion
  • Thin-film coatings
  • B-jade / C-jade

Treatment Stability Chart

Treatment Stability and Care Requirements
Treatment Stability Sensitive To Care Notes
Heat (corundum – plain H) Permanent N/A No special care needed
Flux heating H(a) Permanent colour; flux residues stable N/A Disclose separately from plain H
Heat (zircon) Stable Extreme heat Avoid jeweller's torch
Oil (emerald) Temporary Heat, solvents, time Re-oil periodically; no ultrasonic
Resin (emerald) Semi-permanent Heat, strong solvents No ultrasonic; no steam
Lead glass (ruby) Fragile Heat, acids, ultrasonic Extreme care; no ultrasonic or acids
Beryllium diffusion Permanent N/A No special care needed
Ti surface diffusion Fragile Repolishing removes colour Avoid recutting
Irradiation (topaz) Usually stable Strong UV light (some types) Stable once processed
Coating (mystic topaz) Fragile Abrasion, solvents, ultrasonic Will wear off at facet edges
HPHT (diamond) Permanent N/A No special care needed
B-jade polymer Moderate; matrix permanently weakened Solvents, ultrasonic, impact No ultrasonic; handle carefully
Pearl irradiation Stable (normal conditions) Prolonged UV Avoid bleaches; gentle care only

Treatment Stability Warning

Laboratory Report Terminology

Treatment Codes

Code Meaning Details
N or NTE No treatment evidence No indication of any treatment
H Heat treatment Evidence of heating detected
H(a) Heat with flux residue / fracture healing Borax or similar flux – disclosed separately from plain H
H(b) or H(Be) Beryllium diffusion Lattice diffusion; LA-ICP-MS required
O(minor) Minor oil Light oiling; typical for emerald
O(moderate) Moderate oil Moderate enhancement; requires disclosure
O(significant) Significant oil Heavy treatment; strongly impacts clarity
R Resin Polymer-filled fractures
F Filled Fracture or cavity filling (glass, resin)
C Coated Surface coating present
HPHT HPHT treatment High pressure high temperature colour modification

Key Report Phrases

  • "No indication of heat treatment" – stone appears unheated; high commercial value
  • "Lattice diffusion treatment – beryllium present" – Be-diffusion; not plain heating
  • "Composite ruby" – glass-filled composite; not a standard ruby report
  • "HPHT Processed" – diamond colour modified by HPHT treatment
  • "Filler detected" – fracture filling present; type will be specified

Trade Organisation Standards

CIBJO Blue Books

CIBJO (World Jewellery Confederation) requires full disclosure of all treatments:

  • Natural: formed in nature without human intervention
  • Treated: natural material altered beyond cutting/polishing
  • Synthetic: man-made with same properties as natural
  • Imitation: any material resembling another

AGTA Enhancement Codes

Code Enhancement
N Not enhanced
H Heating
O Oiling/resin infusion
F Filling
D Dyeing
B Bleaching
C Coating
U Diffusion (surface or lattice)
R Irradiation
HPHT High pressure high temperature
ASBL Assembled (composite)

Disclosure Chain

Treatment information must pass through the entire supply chain:

  1. Treater → dealer (must disclose)
  2. Dealer → retailer (must disclose)
  3. Retailer → consumer (must disclose)

Failure at any point constitutes fraud or misrepresentation in most jurisdictions.