How to clean a mineral specimen
Water and a soft brush handle most of it. Quartz will tolerate almost anything; a soft arsenate crystal will not, and the mistakes are permanent.
Most mineral specimens need nothing more than lukewarm water, a soft brush and patience. Identify the species first: anything of hardness 3 or less — calcite, cerussite, erythrite — should not be brushed at all. Never put fluorite or a carbonate in oxalic acid, and never put fluorite, calcite or selenite in an ultrasonic cleaner.
In short
- Water, a soft brush and patience clean the large majority of specimens. Everything past that point carries a risk that is not reversible.
- Check the species before you touch it. Hardness 3 or under — calcite, cerussite, erythrite, gypsum — means no brushing. Halite and chalcanthite dissolve outright in water.
- Oxalic acid lifts iron staining off quartz, but it frosts calcite with an insoluble oxalate crust and etches fluorite dull. Use sodium dithionite on those instead — it is not an acid and does not attack carbonates.
- Ultrasonic cleaners open existing fractures and shake crystals off matrix. Keep them for quartz and away from anything with cleavage.
- Never clean a specimen before it is valued or sold, and never remove an old label, an inked mine number, or the grime of a hundred years from the back of a piece. That is provenance, and it is worth more than a clean rock.
| Specimen | Water | Brush | Ultrasonic | Acid |
|---|---|---|---|---|
| Quartz, agate, cassiterite | Yes | Yes | Usually safe | Oxalic acid is the standard iron-stain treatment |
| Fluorite | Yes, briefly | Soft brush only | No | No oxalic acid — it etches. Use dithionite |
| Calcite, cerussite, other carbonates | Yes, briefly | No | No | None — acid dissolves them |
| Sulphides — galena, pyrite, sphalerite | Brief, then dry thoroughly | Soft brush only | No | No, and no bleach |
| Soft arsenates — erythrite, scorodite, olivenite | Rinse at most | No | No | None |
| Soluble species — halite, chalcanthite, borax | Never | Dry dusting only | No | No |
| Light-sensitive — realgar, proustite, pyrargyrite | Rinse at most | No | No | No — and store them in the dark |
Water first, and usually water only
Lukewarm water, a soft artist's brush or a worn toothbrush, and time. Most field-collected material is carrying clay and dust, and clay releases by soaking rather than scrubbing. Put the piece in a bowl of plain water, leave it a day, change the water, leave it another day. Clay that would not shift under a brush falls away on its own.
Lukewarm, not hot. A specimen with internal fractures — which is most of them — can be made worse by thermal shock, and there is no benefit to heat that patience does not give you for free.
Finish with a rinse in distilled water. Tap water in a hard-water area dries as a white haze of carbonate on crystal faces, which is a self-inflicted problem that then needs solving. Then dry the piece completely: water trapped in a vug will seep out over the following week and spot everything it runs across. A few days somewhere warm and airy, not a hairdryer.
One species-specific note on drying: pyrite and marcasite specimens must be dried thoroughly and stored dry. Reactive material oxidises in humid conditions, producing sulphuric acid and a white efflorescence that destroys the specimen and the label it sits on.
Identify the species before you touch it
There are four ways cleaning goes wrong, and all four are predictable from the species.
Hardness. Anything at 3 or below abrades under a brush. Calcite is 3, cerussite 3 to 3.5, gypsum 2, erythrite 1.5 to 2.5. On those, the brush does more damage than the dirt.
Cleavage. Fluorite has perfect octahedral cleavage, calcite perfect rhombohedral. Cleavage planes are pre-existing weaknesses, and vibration, thermal shock and mechanical pressure all find them. This is why fluorite is on every list of things never to put in an ultrasonic bath.
Solubility. Halite and chalcanthite dissolve in water. Gypsum and selenite are slightly soluble and will lose their surface polish. Some borates and sulphates are worse than they look. If you do not know, do not immerse it.
Light and air. Realgar converts to pararealgar under light, going from red crystal to orange-yellow powder, and the change cannot be undone. Proustite and pyrargyrite darken. These belong in a drawer, and the cleaning question barely arises.
If you cannot identify what you have, our identification guide gets you to a species before you get to the water.
Iron staining: oxalic acid, and where it goes wrong
Orange-brown iron staining on quartz is the commonest cosmetic problem in the hobby, and oxalic acid is the standard treatment. A dilute warm solution lifts the iron oxide over hours to days; the specimen then needs a long rinse in several changes of water, because residues left behind cause their own problems. It is genuinely effective, and it is why so much of the quartz on dealers' tables looks better than it came out of the ground.
The failure is that oxalic acid has a strong affinity for calcium. On calcite it reacts quickly and leaves an insoluble white calcium oxalate crust — worse than the stain you started with, and not removable. On fluorite it will not dissolve the crystal outright but it etches and delusters the faces, turning a glassy purple cube into a dull whitish one. Micas and other calcium-bearing species suffer similarly.
The alternative for those is sodium dithionite, also sold as sodium hydrosulphite and the active ingredient in several household rust removers. It reduces iron oxide rather than dissolving it in acid, so it does not attack carbonates or fluorite. It smells unpleasant and needs ventilation, but it is the right tool for iron stain on anything calcium-bearing.
Whatever you use: test on a worthless piece of the same material first, work dilute, work slowly, and rinse far longer than seems necessary.
Ultrasonic cleaners, air abrasives and damage you cannot undo
An ultrasonic cleaner works by cavitation — microscopic bubbles collapsing against the surface. That energy goes preferentially into flaws, which means healed fractures, cleavage planes and the contact between a crystal and its matrix. On a solid quartz crystal it is usually harmless. On a fluorite, a calcite or a selenite it is a way of turning one specimen into several, and on a plate of crystals it detaches them.
Do not assume your jewellery cleaner is gentler. The question is not the machine's power, it is whether the specimen has flaws for the energy to find, and almost every mineral specimen does.
Air abrasive units — fine aluminium oxide or sodium bicarbonate powder driven by compressed air — are a professional preparator's tool for removing matrix and stubborn coatings. In practised hands they are precise. In unpractised hands they frost crystal faces in seconds, and a frosted face does not come back. If a specimen genuinely warrants that level of work, it warrants paying someone who does it daily.
What you should not clean at all
Anything you are about to have valued or sell. Dealers and auctioneers would far rather see a piece as it came out of the box. Cleaning removes evidence and occasionally removes the specimen; it very rarely adds value, and the person buying it can clean it themselves if they want to.
Old labels, and anything written on the specimen. A mine number inked on the underside of an old British specimen is the thing that connects it to a collection and a date. It is the most valuable square centimetre of the piece. Never wash it off, never wash a label, and keep the label with the specimen when you rehouse it — see buying old collection material for why the paperwork often outweighs the rock.
Patina on old material. A century-old Cornish specimen looks its age, and collectors of classic British material want it to. Stripping it back to bright colour reads as a modern piece and is treated as one.
Anything you cannot identify. The cost of waiting is nothing. The cost of putting an unidentified soft arsenate in an ultrasonic bath is the specimen.