Category: Care and maintenance

Renovating marble means reworking its surface with abrasives of increasing grit, from coarse to fine, then bringing back the shine. It removes material: it is irreversible and limited by the thickness of the slab. Before sanding, you therefore need to identify how deep the damage goes and what the original finish was.

Key points

– The dulling is not a stain: the ICOMOS-ISCS glossary classes it among losses of material. It cannot be cleaned; it is repolished.

– Calcite is at degree 3 on the Mohs scale and “is easily scratched with a knife blade” (University of Waterloo). That is what makes sanding possible, and risky.

– Grit progression cited by Google's AI Overview: 30 to 50, then 120 to 220, then 400, then 800, then 1000 and above. No standard publishes it,

– Repolishing a floor makes it more slippery when wet : under NF EN 14231, a polished finish measures 10–15 SRV, honed under 35, bush-hammered above 45.

– The chemistry of the crystallisation is published by no accessible source. The only page that describes the process speaks of an “acid powder”.

This article deals with the work itself, not the service. It continues our complete guide to marble care and maintenance, picking up where cleaning can do no more.

Where cleaning stops and renovation begins

The boundary comes down to a single criterion: has the material gone, or is it still there under a deposit? Dulling is the increase in roughness of an originally polished surface, placed by the ICOMOS-ISCS glossary in the family of features induced by loss of material, not among stains. The glossary adds that this loss of polish produces an increase in lightness, and that dark marbles show it visibly.

The direct consequence: a matt area on a polished marble cannot be cleaned. No product puts calcite back. It is repolished, or it stays. That is exactly the case covered in our article on white marks on black marble, where the white mark is a rough surface that diffuses light instead of reflecting it.

The good news is documented by a producer: natural stone “can always be restored by repolishing the surface, which restores the original finish” (Levantina, data sheet, March 2025). The bad news lies in the word “always”, which applies to stone in general, not to your slab at its current thickness.

Sanding marble: the method and the grit sequence

Grinding is the abrasion stage before polishing: a series of increasingly fine abrasives is used, each removing the marks left by the previous one, until the surface is even enough to take a shine. The term belongs to the trade vocabulary and does not appear in the ICOMOS glossary.

Why it works on marble: calcite, which makes up most of it, is at grade 3 on the Mohs scale and “is easily scratched with a knife blade or a geologist’s hammer” (University of Waterloo, Earth Sciences Museum). This softness is why marble can be reworked by hand, and also why too coarse a grit digs in quickly. Note that the figure comes from the mineralogy of calcite and not from a standardised test on the stone: NF EN 14205 on Knoop hardness was withdrawn in September 2013 (CTMNC, July 2021).

The grit progression

• Grit 30 to 50 : tackling a heavily scratched or uneven surface. This is the stage that removes the most material.
• Grit 120 to 220 : removing the marks of the previous stage.
• Grit 400, then 800, then 1000 and above : preparing the surface for the shine.
• Work wet or on a dampened surface, to avoid overheating and scratches.
• Vacuum or rinse between each change of grit : coarse abrasive left on the slab produces micro-scratches at the next stage.

Caution. This sequence is the one cited by Google’s AI Overview for the query “poncer du marbre” (sanding marble), as recorded on 25 September 2026. Its nine sources are tool retailers, three videos and a directory. It is therefore an aggregate of commercial pages, not a standard, and no tier 1 or 2 source consulted publishes a quantified progression. To be checked against the technical documentation of the abrasive used.

No rotation speed, contact pressure or time per stage is given here: these values are documented in no source consulted, and they depend on the machine.

What sanding removes from the slab

Sanding does not erase a defect: it lowers the surface below the defect. The thickness lost does not come back, and that is the information neither renovation articles nor tool tutorials give. The available margin therefore depends on the thickness of the product laid, not on the state of its surface.

On a thin product, the margin is small. A modular tile under NF EN 12057 structurally leaves less material to sacrifice than a thick slab. It is a selection criterion at purchase as much as a care criterion: our marble tiles come in several thicknesses, and that thickness determines how many times they can be reworked over the life of the floor or the work plan.

Caveat: the number of possible reworkings per thickness. No standard, data sheet or professional document consulted publishes the thickness removed per pass or the maximum number of repolishings. This figure does not exist in the accessible sources, and inventing it would be a lie that helps nobody.

There is, however, a marker for the limit of surface abrasion against deep colouring. On a marble stained by iron, iron remains the dominant element down to about 800 µm and calcium only becomes predominant again at about 1.7 mm depth (Reale et al., Molecules, 2023). In other words, colouring of this kind lies below the zone a finishing sand reaches: it has to be extracted, not abraded.

What cannot be sanded

A finish is not a state of wear: it is a product. A honed, brushed, sandblasted, bush-hammered or tumbled marble was made that way. Reworking it with a fine grit does not “restore” it; it turns it into a polished finish, and therefore into a different product. The result is not a renovation but an irreversible change of appearance on the piece concerned.

Two other cases call for diagnosis before any work: a crack running through the slab, and a slab whose joints already stand proud through wear. In both situations, abrasion moves the problem without treating it.

Repolishing a floor makes it more slippery

This is the safety warning that the pages ranking for these queries do not give. Slip resistance is measured with the pendulum friction tester under NF EN 14231, and it changes with the finish. On the data sheets of a black limestone producer, the wet polished state is declared at 10 SRV and under 15 SRV, honed under 35 SRV, bush-hammered above 45 SRV, the same stone reaching 70 SRV when dry.

Bringing a bathroom, shower or entrance floor back to a high shine therefore lowers its grip when wet. These values come from producer data sheets for a given batch; they are not a universal property of marble: ask for the declaration of performance for your batch and your finish. If the floor is regularly wet, the question to ask before sanding is not “how do I make it shine” but “which finish do I want to live with”.

Crystallisation, and the contradiction Google produces

Crystallisation is a shine-restoring process applied wet to a marble that has already been ground, using a specific product and a buffing machine. This definition describes a method, not a reaction: it stops exactly where the accessible sources stop.

A survey of the search results reveals a contradiction produced by Google itself. On one query, its generated summary explains that an acidic product attacks the limestone of marble and leaves a matt haze. On the neighbouring query, the same system concludes that you should “finish with crystallisation” to protect the stone against stains. And the only page in the corpus that describes the process calls it “acid powder for wet polishing”. Nobody reconciles the two.

Caution, without exception. None of the nine source pages of Google’s summary names the product’s composition. None gives its pH or the mechanism by which an acid process would make calcium carbonate shine. This information is not published in the accessible sources, and saying so is better than an invented explanation. The only document that would contain it is the safety data sheet of the crystallising product used, to be requested from whoever applies it.

Two useful clarifications in the meantime. Crystallisation is not a protective treatment: a water-repellent or impregnating sealer follows a different logic and is a different product. And “crystallisation” has nothing to do with “washing soda crystals”, an alkaline household product whose French name causes frequent confusion.

Three widespread tips best avoided

These three claims appear in the top ten French results on marble renovation. None holds up.

• “Making marble less porous, that is to say more permeable”. The two halves of the sentence contradict each other. A less porous stone lets less water through, so it is less permeable, not more.
• Buffing with petroleum jelly. A greasy substance deposited on the surface is not a polish. The ICOMOS glossary calls it a film : a thin, usually organic covering that follows the surface without penetrating it, changes its appearance and “may, as it ages, lose its transparency”.
• Buffing with olive oil. Same mechanism, and it is exactly the description of the whitish haze seen a few months later. The shine obtained is that of the film, not of the stone.

Caution. No standardised test consulted concerns petroleum jelly or olive oil applied to marble. The reasoning above relies on the ICOMOS definition of a film, not on a measurement. It stands as a probable mechanism, not as proof.

What you can do yourself and what you cannot

The dividing line is simple: as long as no material is removed, the attempt is reversible. As soon as you abrade, it is not. Reasonably within a homeowner’s reach are cleaning with a neutral detergent, removing a film, and restoring the shine of a small area with a hand polishing product. A floor, a large visible continuous surface, any finish other than polished, a thin slab, and anything requiring a machine are for a professional.

Sanding dust: a point not to be glossed over

Sanding stone produces dust, and that subject should not disappear from a tutorial. Caveat: no tier 1 or 2 source consulted for this article deals with the respiratory risk of sanding marble or the protective equipment to wear. The information is not published in the accessible documents, and it is not our place to establish it. Two points can nevertheless be checked. All the methods found work wet or on a dampened surface. And the safety data sheet of the abrasive and the product used is the only authoritative document on this point. In an occupied home, calling in a professional equipped with on-tool extraction is the prudent answer.

When to call a professional

Four signs are enough: the surface is a floor, the finish is not polished, the available thickness is small, or the damage is structural. In these cases, trying it yourself costs more than calling someone in. Sagit supplies the stone and has it shaped to the dimensions required, without providing on-site renovation services: for a question about thickness, finish or compatibility before work, write to us.

Frequently asked questions

Can marble be sanded?

Yes, because calcite is soft, at Mohs 3. But not all marble can be sanded: it depends on the original finish, the available thickness and the depth of the damage. A polished marble can be reworked; a honed or brushed marble is ruined.

Which sandpaper should be used for marble?

The rule is not a single grit but a progression: from the coarsest needed to the finest useful, without skipping a stage. The progression cited by Google’s AI Overview runs from 30–50 to 1000 and above, via 120–220, then 400 and 800. It is not standardised: refer to the abrasive’s documentation.

Can marble be sanded by hand?

On a small area and with a fine grit, yes. On a large surface, no: a hand cannot hold a flat plane, and uneven reworking shows more than the original defect in raking light. The risk is not breakage; it is hollowing.

How do you restore dulled marble?

First, by identifying the cause. Uniform dulling with no relief calls for polishing, without prior grinding. A matt area with the sharp outline of a spilt liquid is dulling, a loss of material that can only be treated by repolishing. A regular, slightly greasy haze is more likely a film to removed by cleaning.

How do you strip marble?

Stripping, strictly speaking, means removing an added layer: wax, film, treatment residue. It is done with a stripper suited to the film in place, not with an abrasive. If the surface remains matt after stripping, the problem was not the layer but the stone itself.

Which products should not be used on marble?

Anything acidic, since calcite dissolves: vinegar, lemon, descaler, rust stain remover not approved for calcareous stone. Also avoid scouring powders, which are abrasive on a stone at Mohs 3, and greasy substances presented as polishes. One producer recommends sticking to detergents with pH 7 to 10.

Key points

A successful marble renovation starts with a decision, not a machine. The useful question is not “which grit” but “have I lost material, and how much more can I lose”. A polished marble can be restored, provided there is thickness left below the surface and the finish you are aiming for is indeed the original one. If the doubt concerns the thickness or finish of your stone, it is a supply question before it is a care question: our marble tiles are described thickness by thickness.

Sources

• University of Waterloo, Earth Sciences Museum, Calcite, limestone and marble. Online data sheet (consulted on 25/09/2026).
• Vergès-Belmin V. (ed.), ICOMOS-ISCS: Illustrated glossary on stone deterioration patterns, Monuments and Sites XV, 2008, ISBN 978-2-918086-00-0. Open-access glossary (consulted on 25/09/2026). Definitions of increase in roughness and dulling (p. 30), discolouration (p. 46) and film (p. 52).
• NF EN 12057, Natural stone products, modular tiles, requirements, April 2015. AFNOR listing (consulted on 25/09/2026).
• NF EN 14231, Natural stone test methods, Determination of slip resistance by means of the pendulum tester, December 2003. Exact title according to CTMNC, Principales exigences techniques applicables aux produits de construction en pierre naturelle, July 2021, p. 15. All rights reserved to CTMNC, www.ctmnc.fr.
• Levantina, Negro Marquina, ficha técnica, March 2025: SRV values by finish under NF EN 14231, care instructions pH 7–10, and restoration by repolishing. Additional value of 10 SRV wet polished: TINO Stone & Surfaces, Marquina, ficha técnica, November 2024. 70 SRV value when dry: Canteras Olaspe.
• Reale R., Andreozzi G.B., Sammartino M.P., Salvi A.M., Analytical Investigation of Iron-Based Stains on Carbonate Stones, Molecules, 28(4):1582, 2023: iron diffusion depths (≈ 800 µm, ≈ 1.7 mm).
• Withdrawal of NF EN 14205 on Knoop hardness in September 2013: CTMNC, July 2021, p. 15.

A white mark on black marble is almost never a stain: it is dulling, the local loss of polish caused by an acidic product. The ICOMOS-ISCS glossary classes this phenomenon as an “increase in lightness” and states that dark marbles often undergo it. The roughened surface diffuses light instead of reflecting it.

The distinction changes everything: a loss of material cannot be cleaned. Before choosing a product, you therefore need to know whether the mark is recessed or raised. The finish chosen at the outset also matters, as with our marble worktops.

Key points

– The mark is almost always a dulling by acid, not a deposit: ICOMOS classes “stain” and “dulling” in two distinct families of deterioration.

– Four mechanisms produce white: dulling, efflorescence, encrustation, film. Only one can really be cleaned off.

– Three actions settle it: raking light, fingernail, clean water.

– Dulling cannot be cleaned off, it must be repolished.

– Vinegar is the reagent geologists use to identify a limestone; bicarbonate, for its part, scratches: calcite is at degree 3 on the Mohs scale.

Why this white mark is not a stain

Standardised vocabulary exists, and it settles the debate. The illustrated ICOMOS-ISCS glossary (2008, 86 pages, bilingual, with LRMH and CICRP among the contributors) names what the market calls “white haze”. The dulling is the French name for what the glossary classes under roughening, a sub-type of erosion: “a selective loss of small particles from an originally smooth surface”, the substrate remaining sound. The same source explains why the mark looks white: loss of polish produces an “increase in lightness”, classed as discolouration, and “dark marbles and those of bright colours often undergo discolouration”. Nothing has been deposited: a micro-roughness diffuses light instead of reflecting it, and the contrast is greatest on a polished black background.

The classification is the real argument: ICOMOS places “stain” among the chromatic alterations and dulling among the losses of material. A chromatic alteration can be removed; a loss of material cannot. The mechanism is simple: a commercial marble is overwhelmingly composed of calcite. The University of Waterloo points out that calcite defines grade 3 on the Mohs scale and dissolves on contact with an acid, releasing carbon dioxide. The Natural Stone Institute confirms it: some products “etch the finish without leaving a stain”.

Acid etching or limescale deposit: the decision tree

Three checks are enough to tell the possibilities apart, with no product at all. Raking light first, with your eye level with the surface: a matt area with a sharp outline, following the shape of the spilt liquid, indicates dulling; a diffuse or milky haze points to a deposit. The fingernail next: if recessed, material has gone; if raised, the texture tells you more, powdery for salts, hard for precipitated limescale, soft and greasy for product residue. Clean water finally: soluble salts partly redissolve, dulling does not change.

Definitions and families: ICOMOS-ISCS glossary, 2008.

Theefflorescence refers to a crystallisation of salts on the surface, whose constituents come from the stone’s pore network: it is not a deposit in the strict sense. When these salts crystallise under the surface, ICOMOS speaks of subflorescence, a frequent consequence of which is loss of stone through scaling. A mark that comes back after every wipe is not a cleaning problem; it is a water problem.

Three levels of severity, three responses

Level 1: a film or deposit on the surface

This is the only case that can be removed without touching the stone. ICOMOS defines a film as a thin, homogeneous, usually organic covering that follows the surface without penetrating it, and notes that a film may, as it ages, lose its transparency. That is the mechanism behind the whitish haze left by some waxes: a suitable stripper, then rinsing. For routine care, the Natural Stone Institute recommends a neutral cleaner or a mild washing-up liquid in warm water, and Levantina requires detergents with a pH 7 to 10.

Level 2: salts or encrustation

The damage is no longer merely cosmetic. Efflorescence is cleaned dry and then with clean water, but it will return as long as the source of moisture is not dealt with.encrustation is a compact, hard, adherent layer resulting from precipitation, in ICOMOS terminology, and it follows a different logic. The glossary warns that “the removal of encrustations may be accompanied by detachment of stone”. Scraping and aggressive descalers are therefore ruled out, in favour of specialist intervention.

Level 3: dulling

Dulling cannot be cleaned; it is repolished. No cleaner rebuilds a polish, which is not a layer but a surface state. Levantina writes that natural stone can always be restored by repolishing the surface, and the Natural Stone Institute refers readers to a professional restorer for repolishing etched areas. The work is done with abrasives of increasing grit, then polishing; (not verified): no standard consulted sets the grit sequence. There is one objective limit, since repolishing removes material: the available thickness, which is very small on a thin tile. No source gives a depth threshold for irreversibility, and none will be given here.

What you must not use

White vinegar is not a cleaning product for a calcareous stone: it is an identification reagent. Thethe Earth Sciences Museum at the University of Waterloo describes the geologist’s test: a dilute acid is poured cold onto the rock, “10% hydrochloric or sulphuric acid, or vinegar”. A limestone then gives off bubbles of carbon dioxide. Pouring vinegar on a black marble means running that test on your own worktop.

Bicarbonate poses a different problem, mechanical rather than chemical. Being alkaline, it does not attack calcium carbonate. But calcite is at Mohs 3 and “is easily scratched with a knife blade” (University of Waterloo). Scouring powders, for their part, contain abrasives liable to scratch some stones (Natural Stone Institute). Vinegar dissolves the stone, bicarbonate scratches it, and neither cleans a limestone. Also to be avoided: bleach, ammonia, descalers, metal utensils and rust stain removers, which contain traces of hydrofluoric acid.

One last point nobody mentions: susceptibility to staining is measured. The standard NF EN 16301 (March 2021) concerns the determination of resistance to accidental staining, and a test result can be requested from the supplier before purchase. None of the sources consulted, however, presents a water-repellent or impregnating sealer as protection against acid etching.

Two questions to see where you stand

Which product touched the surface, and how long ago? Is the area rough to the touch, or simply matt? Rough and recessed is level 3; smooth and raised, levels 1 and 2. For a laid floor, see made-to-measure marble floors and tiles and marble tiles.

Frequently asked questions

How do you restore black marble?

It depends on what has been lost. If the black is dulled by a film or a deposit, a neutral cleaner followed by a suitable stripper is enough. If the polish has gone, no product will bring it back: Levantina states that natural stone can always be restored by repolishing the surface.

How do you remove limescale marks from black marble?

A hard limescale deposit is what ICOMOS calls an encrustation : a compact, adherent layer resulting from the precipitation of carbonates. The glossary warns that removing it may be accompanied by detachment of stone, which rules out scraping and descalers. Wiping and drying after each contact with water remains more effective than any treatment.

Which products should not be used on marble?

Vinegar, lemon juice and any acidic product, descalers and limescale removers, bleach, ammonia, scouring powders, metal utensils, and rust stain removers, which contain traces of hydrofluoric acid according to the Natural Stone Institute. Levantina’s instruction is simple: neutral detergents, pH 7 to 10, and nothing else.

Does white vinegar damage marble?

Yes, and the mechanism is documented: vinegar is one of the reagents in the limestone identification test described by the University of Waterloo, alongside hydrochloric and sulphuric acids diluted to 10%. There is no pH threshold below which the attack would “start”: calcite dissolves as soon as the solution is undersaturated; only the speed varies. Below about pH 3.5, that speed is limited only by diffusion (Penn State, course CE 574, after Plummer and Wigley, 1978).

How do you rescue a stained marble?

Through diagnosis, above all. A real stain, that is a chromatic alteration, is treated by cleaning or with a poultice suited to the staining agent. A loss of material can only be treated by refinishing. Applying a stain-removal protocol to dulling achieves nothing.

How do you remove white vinegar marks?

There is nothing to remove, and that is the key to this page. A mark left by vinegar is dulling, a dissolved surface, not deposited matter. No cleaning will make it disappear; only repolishing restores the original surface.

Key points

On a polished black marble, the contrast forgives nothing: damage invisible on a light stone becomes an obvious white mark. The right reflex is not to look for a product but to qualify the mark in three checks, then choose between cleaning, treating the moisture and repolishing.

Sources

• Vergès-Belmin V. (ed.), ICOMOS-ISCS: Illustrated glossary on stone deterioration patterns, Monuments and Sites XV, ICOMOS, 2008, 86 p., ISBN 978-2-918086-00-0.
• University of Waterloo, Earth Sciences Museum, “Calcite, limestone and marble”.
• Natural Stone Institute, “Care & Cleaning”.
• Levantina, Negro Marquina, ficha técnica, March 2025, p. 5.
• CTMNC, Principales exigences techniques applicables aux produits de construction en pierre naturelle, July 2021, p. 15, for the title of the NF EN 16301 standard. All rights reserved to CTMNC, www.ctmnc.fr.
• Penn State, College of Earth and Mineral Sciences, course CE 574, Mineral Reaction Kinetics.

All sources were consulted on 25/09/2026.

See also: Nero Marquina: material guide to the black marble from Markina

A rust stain on marble has two possible origins, and they are not treated in the same way: a deposit of iron from outside, or the oxidation of iron-bearing minerals contained in the stone itself. In both cases, the iron is extracted with a poultice, never with acid.

Key points

– Iron does not stay on the surface: it migrates into the pore network and precipitates inside the stone. Wiping does nothing.

– A second origin, overlooked by French-language sources: the oxidation of iron sulphides, carbonates and silicates contained in the stone, visible 8 to 12 months after laying.

– Reference protocol, U.S. General Services Administration: sodium citrate poultice, about 48 h. It can dull a polished surface.

– A conservation laboratory has excluded oxalic acid on calcareous stone, pH too low. The solubility of calcium carbonate increases down to pH 5, and below that decomposition becomes irreversible.

– Reversibility depends on the depth, not of time: diffusion measured to ≈ 1 mm, iron predominant to ≈ 800 µm.

Where this rust comes from

Three practical origins, only two of which are documented by the sources used here. The diagnosis governs the rest: the same poultice succeeds in one case and has no effect in the other.

A deposit of iron from outside

La imported rust is a deposit of iron from outside: a furniture leg, a tin can, iron-rich water. The stain is not deposited; it migrates. The iron goes into solution as soluble Fe²⁺ ions, water carries it into the pore network, then it oxidises to Fe³⁺ and precipitates as insoluble oxyhydroxides inside the stone. In marble, diffusion goes deeper than elsewhere and follows the grain boundaries (Reale et al., Molecules, 2023).

Iron-bearing minerals contained in the stone

La native rust corresponds to the oxidation of iron-bearing minerals contained in the stone itself: a stone laid new and sound can develop its own stains. The oxidation affects accessory minerals in the mass: iron sulphides (pyrite, marcasite), iron carbonates (siderite), ferromagnesian silicates (biotite, hornblende), dispersed or concentrated along veins. Water carries this iron to the surface of the slab, where it precipitates as limonite.

The decisive fact is chronological: these stains develop after laying, within 8 to 12 months. And the damage is not only cosmetic, since cracks appear, attributed to thevolume expansion of the oxidised and hydrated minerals (Bams and Dewaele, Materials Characterization, 2007). Reale et al. (2023) confirm the mechanism on a different corpus, archaeological marbles against marbles stained in the laboratory.

Two reading precautions. These statements come from the abstract of the 2007 paywalled article: nothing is extrapolated beyond it. And the criterion is mineralogical, not commercial. The sources reason in terms of pyrite or siderite, never in quarry names: no trade name can be presented as “prone to pyrite”.

Iron rising from the substrate

A third possibility, to be treated as a diagnostic hypothesis : metal fixings, reinforcement, iron-bearing laying mortar. No institutional source used here documents it, but the symptom is clear: the stain reappears after each successful treatment. This is decided at the laying stage and by the thickness, well before maintenance: see our marble tiles and our marble worktops.

The extraction poultice, step by step

Le poultice is a white absorbent filler, impregnated with a solution that makes the iron mobile, applied to the stain and then removed. The reference procedure is public: Poulticing Rust Stains from Limestone and Marble, from the U.S. General Services Administration. It gives the formulations, the permitted fillers and the application times.

1. Prepare a white, inert absorbent filler : modelling plaster, untreated white flour, kitchen paper, powdered chalk, talc, Sommières earth.
2. Mix the solution : 7 parts glycerine, 1 part sodium citrate, 6 parts warm water. The glycerine keeps the poultice moist while it is in place.
3. Apply thickly, then cover with a film to prevent drying.
4. Leave for about 48 h. The time depends on the product: 30 minutes per application for sodium hydrosulphite, 24 h maximum for a commercial rust remover.
5. Remove, rinse immediately with the sodium citrate solution, then dry.
6. Check the finish. The GSA warns that the treatment may dull polished surfaces, which can be corrected by repolishing. Its warning concerns sodium hydrosulphite and sodium hypochlorite, which are not acids: it is indeed a loss of polish, not acid etching. On a thin slab, the margin is small.

Three families of products, three chemical roles, and that is what explains the differences in application time.

Warning: what not to apply

Read before buying a product

A marble is made of calcium carbonate. Any sufficiently strong acid dissolves it: that isacid etching, with loss of polish, a whitish matt finish and micro-relief. No vinegar, no lemon, no descaler, no rust remover not approved for calcareous stone.

The best-established point concerns oxalic acid. A conservation laboratory withdrew it from its tests on calcareous stone, its pH being too low and damaging the stone surface (Cezar, 1998). The same article gives a quantified marker: the solubility of calcium carbonate increases down to pH 5, and below that threshold decomposition becomes irreversible.

This acid does more than dissolve. It reacts with calcite and precipitatescalcium oxalate, which is very poorly soluble and therefore impossible to rinse away. You swap a brown stain for an irreversible white haze, plus an etched surface.

Why are these products tolerated on granite or glazed ceramics? Because granite is made of silicates and contains no calcite accessible to a weak acid. But tolerance is not harmlessness : the Natural Stone Institute states that hydrofluoric acid, present in traces in many commercial rust removers, also attacks silicates, and that all stones including granite and quartzite are attacked if exposed to it. On a limestone the damage is immediate and visible; on a granite it is slower and less spectacular. Neither stone is a good candidate.

No source makes it possible to generalise about the composition of consumer rust removers sold in France. The only reliable information is the data sheet and safety data sheet of the specific product.

When the stain will not come out

No institutional source consulted gives a threshold in days or weeks, and the GSA procedure states no case of irreversibility. The criterion is not a time; it is a depth and a mineral state.

The measurements by Reale et al. (2023) give two markers. On an artificially stained Carrara marble, iron diffusion reaches about 1 mm at most. On a historic sample, iron remains the most abundant element down to about 800 µm, and calcium only becomes predominant again at about 1.7 mm. The iron also evolves into more stable phases, which are harder to extract.

A poultice works on the first few tenths of a millimetre. Reversibility therefore depends on the depth reached and the stability of the mineral formed.

Can a rust stain be prevented?

On imported rust, a water repellent limits liquid water that carries the iron. The only experimental study available here concerns ornamental slates with sulphide inclusions, treated with silane-siloxanes: the products tested significantly increased the sulphides’ resistance to oxidation, without significant colour change (Rivas et al., 2011). These are slates, neither a marble nor a limestone. The result is indicative; it proves nothing about a carbonate stone, and it speaks of increased resistance, never of elimination.

For native rust, a water-repellent cannot prevent the phenomenon. This is reasoning, not a quotation: the iron is already in the stone, and a surface treatment does not remove a pyrite inclusion.

The upstream lever is a standard, and nobody uses it: the mineralogical identification of a stone falls underEN 12407:2019, petrographic examination. That is where the question of accessory minerals arises, before purchase or before laying.

Frequently asked questions

How do you remove ingrained rust stains?

By extraction, not by attack: a white absorbent poultice loaded with a solution that makes the iron mobile, left in place, then rinsed off. About 48 hours with the GSA’s mixture of glycerine, sodium citrate and warm water. Several short applications are better than one prolonged one.

How do you remove a stain ingrained in marble?

The same logic applies to any stain that has migrated into the pore network : you do not rub the surface, you draw the substance out. The product depends on the nature of the stain, and on a limestone it rules out acids. If the colouring comes from within the stone, no poultice will remove it.

Does white vinegar damage marble?

Yes. Vinegar is an acid, and marble is calcium carbonate: the calcite dissolves and the surface dulls. Vinegar is even used as a reagent to identify a calcareous rock. On a rust stain, it extracts nothing and adds acid etching to the problem.

Which products should not be used on marble?

Anything acidic: vinegar, lemon, descaler, rust stain removers not approved for calcareous stone. Oxalic acid deserves a special mention, a conservation laboratory having excluded it from its tests on limestone because its pH was too low (Cezar, 1998). Also avoid scouring powders, which are abrasive.

How do you clean marble the old-fashioned way?

The recipes in circulation are almost all based on vinegar or lemon, and therefore on an acid: the wrong choice on a limestone. The only traditional method validated by an institutional procedure is the poultice, a white absorbent filler and a non-acidic solution left to act.

Key points

Before choosing a product, establish the origin. Two questions are enough: did the stain follow contact with a metal object, or did it appear with no visible cause several months after laying? And does it come back in the same place after treatment? In the first case, the sodium citrate poultice is the safe route. In the second, the iron comes from the stone or from the substrate.

Sources

• U.S. General Services Administration, Poulticing Rust Stains from Limestone and Marble, Historic Preservation Technical Procedures. (consultée le 25/09/2026).
• Cezar T.M., Calcium Oxalate: A Surface Treatment for Limestone, Journal of Conservation and Museum Studies, no. 4, May 1998. Online text (consulted on 25/09/2026).
• Reale R., Andreozzi G.B., Sammartino M.P., Salvi A.M., Analytical Investigation of Iron-Based Stains on Carbonate Stones: Rust Formation, Diffusion Mechanisms, and Speciation, Molecules, 28(4):1582, 7 February 2023. Open-access article (consulted on 25/09/2026).
• Bams V., Dewaele S., Staining of white marble, Materials Characterization, vol. 58, no. 11-12, November 2007, pp. 1052-1062, DOI 10.1016/j.matchar.2007.05.004. Paywalled article; the points cited come from the abstract.
• Rivas T., Iglesias J., Taboada J., Vilán J.A., Sulphide oxidation in ornamental slates: protective treatment with siloxanes, Materiales de Construcción, vol. 61, no. 301, 2011 (CSIC). Points cited taken from the abstract.
• EN 12407:2019, Natural stone test methods, Petrographic examination (CEN). Title and scope checked; full text not consulted.