How to tell if a crystal is dyed
Dye pools in fractures and pores rather than following growth zones. Check colour distribution under a lens first: the destructive tests are a last resort.
Look at where the colour sits before you do anything else. Natural colour in a crystal follows growth zoning and crystal faces; dye pools in fractures, pores and drill holes. A 10x lens answers it on most specimens in a minute. Howlite at hardness 3.5, dyed blue for turquoise, is the commonest case in the mineral trade.
In short
- Colour distribution is the whole test on most dyed material. Dye enters through fractures and porosity and concentrates there, giving darker lines along cracks and a paler body colour. Natural colour follows growth zoning, which is geometric and related to crystal faces.
- The fractures are sometimes deliberate. GIA records that fractures are sometimes purposely induced by heating a gem so that otherwise non-porous material can more readily accept dye. That is how ‘cherry quartz’ and crackle-dyed quartz are made.
- Howlite and magnesite dyed blue are the commonest imitations in the hobby, both standing in for turquoise. Both are white and porous with dark veining, which is what makes them convincing and also what gives them away.
- The solvent test works and it is destructive. Many dyes are removable with alcohol or acetone. It is a last resort, done on an inconspicuous spot, and it can harm perfectly natural material too.
- Treatment is not the problem; undisclosed treatment is. In the UK, failing to disclose a treatment that would change a buyer’s decision is a misleading action under consumer protection law, not merely bad manners.
| Material | Sold as | How the dye behaves | The check | Confidence |
|---|---|---|---|---|
| Howlite | Turquoise | Soaks the whole porous body; veins stay dark | Hardness 3.5 against turquoise 5–6. A steel point marks howlite easily | High |
| Magnesite | Turquoise | As howlite; colour often too even and too bright | Hardness 3.5–4.5; effervesces in warm dilute acid as a carbonate | High |
| Agate and chalcedony | Agate, in colours that do not occur | Follows banding, but with unnatural saturation in the porous bands | Look for colour confined to alternate bands with hard edges | High |
| Quartz, crackled | ‘Cherry quartz’, ‘strawberry quartz’ | Pools in induced fractures, leaving a network of coloured veins | The fracture network is internal and shows no surface expression | Very high |
| Marble and calcite | Various trade names | Enters along grain boundaries | Colour follows grain boundaries, visible under a lens | High |
| Quartz geodes and druses | ‘Aura’ and coated quartz | Not a dye — a thin metallic coating | Iridescent metallic sheen; scratches off to reveal clear quartz | Very high |
| Genuinely coloured species: amethyst, fluorite, dioptase | Themselves | No dye involved | Colour follows growth zoning and crystal faces | — |
Check one: where the colour sits
This resolves most cases and requires nothing but a 10x lens and a strong light, ideally held at the side so it rakes across the surface and also shines through any translucent edge.
Natural colour in a crystal is structural. It comes from trace elements or colour centres distributed as the crystal grew, so it follows growth zoning: bands parallel to crystal faces, colour concentrated at the termination or at the core, sharp geometric boundaries that correspond to something the crystal did. Amethyst shows this beautifully, as does zoned fluorite.
Dye is a fluid that went where fluids go. It enters through fractures, pores, grain boundaries and drill holes, and it concentrates there. The result is the opposite pattern: darker along cracks and paler in sound material, colour that stops at a grain boundary rather than at a crystal face, and pooling in surface pits.
On a bead or a polished piece, check the drill hole and the edges. Dye concentrates at both, because that is where the surface area is. On a cabochon, look at the back, which is usually less well finished and shows the untreated body colour where the polish has not sealed it.
Check two: is the colour possible at all
A large proportion of dyed material announces itself by being a colour the species does not produce, or a saturation it does not reach. This requires knowing the species, which is why identification comes before treatment assessment.
The clearest examples in the hobby are the uniform, opaque, slightly plastic blues and greens: bright turquoise-blue howlite, acid-green “quartz”, magenta agate slices, and the tumbled material sold in bowls in every gift shop. Nature makes saturated colours, but it very rarely makes them perfectly even across a whole piece, because the trace-element concentration that causes the colour varies as the crystal grows.
The GIA reference on gem treatments lists the materials historically dyed: pearls, coral, turquoise, lapis lazuli, howlite, nephrite jade, chalcedony, quartz, emerald and ruby. That list is a good prior. If you are looking at one of those in a startling colour, the burden of proof is on the colour.
One caution in the other direction. Plenty of minerals really are that colour — dioptase is emerald-green, crocoite is orange, erythrite is crimson, vanadinite is bright red-orange — and dismissing a good natural specimen as dyed is a mistake collectors make almost as often as the reverse. Our species reference records the actual colour ranges from the Handbook of Mineralogy, which is the check to make.
Check three: the tests that cost something
When the visual checks are inconclusive, there are three further tests, in increasing order of what they cost you.
Hardness, which costs a small mark in a hidden place. This is the decisive test for the turquoise imitations: howlite is hardness 3.5 and magnesite 3.5 to 4.5, against turquoise at 5 to 6. A steel point marks the first two easily and the third with difficulty. Do it on the base, under magnification, and read the method on testing mineral hardness first.
Acid, which costs a fragment. Magnesite is a carbonate and effervesces in warm dilute acid; turquoise, a phosphate, does not. Test a detached fragment on a watch glass, never the specimen.
Solvent, which is genuinely destructive and should be the last thing you try. GIA notes that many dyes are susceptible to removal through contact with solvents such as alcohol or acetone, and that dyes in porous material can leak from fractures. A cotton bud dampened with acetone, rubbed briefly on an inconspicuous area, will pick up colour from most dyed material. Three warnings. It will not work on every dye. It can damage natural material — some minerals are solvent-sensitive, and any old repair adhesive will dissolve. And if the piece is not yours, do not do it at all: an unauthorised test on a dealer’s stock is criminal damage, not diligence.
Ultraviolet is worth a mention and not much more. Some organic dyes fluoresce and their natural counterparts do not, so a UV lamp occasionally shows dye in fractures as a glowing network. It is a useful confirmation when it works and proves nothing when it does not — see do I need a UV lamp.
Disclosure, and what the law in Britain actually requires
The problem is never that a stone has been treated. It is that a treated stone is described as though it has not been. Dyeing is an ancient and legitimate practice, disclosed dyed material has an honest place in the trade, and a dyed agate slice sold as a dyed agate slice is not a scandal.
GIA’s position is that disclosure of treatments is necessary and legally required for anyone selling a gem, and that non-disclosure misleads buyers about a stone’s natural quality and value. GIA’s reference on gem treatments sets out how dyes are introduced, which materials are affected, and the durability problems that follow.
In the United Kingdom the relevant law is consumer protection law rather than gem-trade rules. Under regulation 5 of the Consumer Protection from Unfair Trading Regulations 2008, a commercial practice is a misleading action if it gives false information, or if its overall presentation deceives the average consumer, in relation to the main characteristics of a product — and causes that consumer to take a decision they would not otherwise have taken. Selling dyed howlite as turquoise fits that description exactly. We are setting out what the regulation says rather than giving legal advice; if you have bought something on a false description and want to act on it, Citizens Advice and Trading Standards are the right first calls.
The practical collector’s position is simpler. Ask the seller directly and in writing whether the piece has been dyed, coated or otherwise treated. A straightforward dealer will tell you, and a written answer is worth having whatever it says. The wider set of questions to ask is on buying minerals online, and the related problem of outright fakes is on how to tell if a crystal is fake. The rest of the collecting guides cover condition, provenance and the questions worth putting to a seller. If you would rather have someone look for undyed, documented material on your behalf, that is what the wanted list does.