How to remove iron stain from a specimen
Oxalic acid strips iron stain and attacks 43 of the 109 species here, which contain carbonate, lead or calcium. Read the formula, not the stain on the crystal.
Oxalic acid is the reagent, and the first job is deciding whether your specimen can survive it. Of the 109 species sheets in this site's reference set, 43 contain carbonate, lead or calcium — the three chemistries oxalic acid dissolves or precipitates on. Read the formula before you open the bottle, because the damage is not reversible.
In short
- 43 of the 109 species covered here contain carbonate, lead or calcium. Twelve are carbonates, which dissolve. Sixteen are lead minerals, on which insoluble white lead oxalate forms. Nineteen are calcium-bearing, and the acid-soluble ones among them acquire a white calcium oxalate crust.
- The risk is readable from the formula, not from the stain. A quartz specimen with iron staining is a candidate. A calcite specimen with iron staining is not, and no amount of dilution changes that.
- Oxalic acid works by complexing iron(III), converting the insoluble iron oxide and hydroxide of the stain into soluble ferrioxalate. It is not scrubbing the stain off; it is taking it into solution.
- The rinse is the operation, not the soak. Ferrioxalate left in the pores of a specimen decomposes in light and leaves a yellow deposit that is harder to remove than the original stain. Rinse far longer than feels necessary, in changes of clean water.
- It is a poison and it is not a mild household chemical. See the PubChem record for oxalic acid for the substance record. Gloves, eye protection, ventilation, and nothing decanted into a drinks container — and nothing on this page is safety advice.
| Group | Count in the reference set | Examples | What happens | Verdict |
|---|---|---|---|---|
| Carbonates | 12 | Calcite, aragonite, dolomite, magnesite, siderite, smithsonite, cerussite, malachite, azurite, rosasite, aurichalcite, hydrozincite | Dissolve or etch; surfaces go dull and detail is lost | Never |
| Lead minerals | 16 | Cerussite, wulfenite, pyromorphite, mimetite, vanadinite, anglesite, crocoite, galena, descloizite, mendipite | Insoluble lead oxalate forms as a white bloom that will not rinse off | Never |
| Acid-soluble calcium species | Part of 19 Ca-bearing | Fluorite, gypsum, scheelite, autunite | Etching, and an insoluble white calcium oxalate crust | Never |
| Sulphides | Several | Pyrite, marcasite, pyrrhotite, chalcopyrite | Acid exposure accelerates the decay reaction already under way | Never |
| Calcium silicates and zeolites | Part of 19 | Epidote, zoisite, vesuvianite, heulandite, stilbite, scolecite | Silicates resist; zeolites are decomposed by acids | Test first, or avoid |
| Quartz and resistant oxides | — | Quartz, cassiterite, rutile, corundum | No reaction. This is what the method is for | The candidate group |
A count of the species this treatment should never touch
Method. On 20 September 2026 the chemical formula on the Handbook of Mineralogy sheet for every species this site covers — 109 sheets after removing duplicates, mirror copies and non-species files — was parsed and tested for a carbonate group, for lead and for calcium.
Result: 12 carbonates, 16 lead-bearing species, 19 calcium-bearing species; 43 distinct species in the union. That is 39% of the list.
The three mechanisms are different and it is worth keeping them apart.
Carbonates dissolve. Oxalic acid is a stronger acid than most organic acids — its first dissociation constant is comparable to some mineral acids — and carbonates go into solution in it. A calcite matrix under an iron-stained quartz is a matrix that will be gone.
Lead forms an insoluble oxalate. Lead oxalate has very low solubility, so the reaction does not simply attack the surface; it deposits a white bloom on it. That bloom does not rinse away, and removing it means another treatment with its own risks.
Calcium does the same thing. Calcium oxalate is likewise insoluble, and it is the reason oxalate crusts are a familiar problem on marble. Where the calcium species is acid-soluble — fluorite, gypsum, scheelite, autunite — you get both failures at once: etching, and then a white crust in the etched surface.
The rule that falls out is short: read the formula, not the stain. If the specimen or anything it is attached to contains CO3, Pb or an acid-soluble Ca, the answer is no.
The procedure, for the specimens that can take it
This is the treatment most likely to destroy a specimen, and the steps that matter most are the ones before and after the acid.
1. Identify everything on the piece, including the matrix. Not just the showy crystal. A quartz group on a carbonate matrix fails the test even though the quartz would be fine. Identification methods are on how to identify a mineral specimen.
2. Clean mechanically first. Remove clay and loose material before any chemistry — see how to remove clay from a mineral specimen. Acid working through mud is acid wasted and stain left behind.
3. Test on an inconspicuous area. The back, the base, a broken edge. Wait, rinse, dry, look. An hour spent here is the cheapest hour in the process.
4. Warm dilute solution, and patience rather than strength. A weak solution working for hours does what a strong one does in minutes, with a fraction of the risk to anything you have misidentified. Never boil it.
5. Rinse until you are bored, then rinse again. Several changes of clean water over days for a porous specimen. Ferrioxalate remaining in the pores decomposes in light to a yellow iron oxalate deposit, and the specimen you have just cleaned acquires a new stain that is worse than the first.
6. Dry slowly and completely, and if the specimen contains any sulphide, get it back to 45% relative humidity promptly — the reasoning is on how to clean pyrite safely.
7. Record what you did. Chemical cleaning is an alteration, and a later owner is entitled to know about it. That is the same principle as declaring a repair, argued on how to tell if a specimen is repaired.
Why the specimen sometimes comes out worse
Four failure modes account for nearly all of it, and all four are avoidable.
Yellow staining after treatment. The commonest complaint, and it is under-rinsing. The iron is in solution as ferrioxalate; if it stays in the specimen's pores it decomposes and deposits. The cure is prevention, because the second stain is harder to shift than the first.
A white bloom. Lead or calcium oxalate, which means the specimen contained something the treatment should never have met. There is no gentle reversal — this is why step one is identification.
Dull, etched crystal faces. The lustre has gone. On carbonates this is dissolution; on fluorite and gypsum it is etching. Lustre is not restorable by any method a collector should be using.
Disintegration a month later. A sulphide-bearing specimen that was treated, imperfectly rinsed and returned to a humid drawer. Acid residue plus humidity plus pyrite is the decay reaction with a head start — see why is my pyrite crumbling.
The honest alternative to all of this is to leave the stain. Iron staining is part of what an oxidised-zone specimen looks like, and a collector who removes every trace of it is removing evidence of where the piece came from. For a specimen with real locality value, no treatment is a legitimate choice and frequently the right one. The Care and Conservation of Geological Specimens leaflet exists for the cases that genuinely need intervention, and a conservator is the right call for anything important.
What else to try before reaching for acid
Mechanical removal. A dental pick, a stiff brush, a wooden skewer, an air abrasive unit if you have one. Slow, controlled, and it stops the moment you stop — which is the opposite of a soak.
Plain water and time. A surprising amount of what looks like iron stain is iron-stained clay, and a week in distilled water with occasional brushing takes it off with no chemistry at all.
Sodium dithionite. A reducing agent used in conservation as an alternative to oxalic acid on some materials. It has its own hazards and its own species restrictions, and it is not a beginner's reagent. Mentioned here for completeness rather than recommended.
Nothing at all. See above. It is a real option.
General cleaning that applies to everything else is on how to clean mineral specimens, and the rest of this section is indexed at collecting guides. We hold no stock and sell nothing; the only commercial route here is the wanted list.