Three nanometres of oxide is the only thing between stainless steel and rust — and it needs oxygen to survive

The only thing standing between a stainless steel sink and a rusty one is a layer of oxide roughly two to three nanometres thick — a few dozen atoms. You cannot see it, feel it, or buy it separately. It formed by itself the moment the steel met air.

An entire material class, used in every kitchen, hospital, chemical plant and railway coach, exists because of a film that thin. And unlike paint or plating, this film has a property no applied coating has: when you scratch it, it comes back.

It also has a condition that almost nobody is told about. It needs oxygen to repair itself — and when it cannot get oxygen, or when chloride interferes, it stops repairing. That single dependency explains nearly every case of stainless steel rusting, including the ones that look impossible.

A barrier that rebuilds itself

Add roughly 11% or more chromium to steel and something useful happens. Chromium has a stronger affinity for oxygen than iron does, so when the surface meets air the chromium reacts first, forming a dense, tightly bonded, chromium-rich oxide across the whole surface.

That film is a barrier, but the barrier is not the remarkable part. Plenty of things are barriers. The remarkable part is that it is self-repairing: cut the surface, grind it, scratch it with a knife, and fresh chromium is exposed to oxygen and a new film forms within seconds.

This is the whole difference between stainless steel and painted steel. As our article on rust-preventing coatings explained, a scratch through a barrier coating is worse than no coating, because it concentrates attack onto a tiny exposed area. A scratch through a passive film is a non-event — the film simply re-forms. Stainless steel is not protected by a coating; it is protected by a process that keeps running.

Aluminium works the same way, and it is the more startling case. Aluminium is thermodynamically very reactive — it should corrode rapidly in air — and the only reason a window frame lasts decades is a two-to-four-nanometre alumina film that forms instantly and re-forms when damaged. Anodising is simply growing that same film deliberately, to micrometres instead of nanometres.

The proof that the film is doing all the work is what happens when you disable it. Amalgamate aluminium with mercury and the oxide can no longer adhere and re-form, and the metal then oxidises rapidly and visibly — which is a genuine reason mercury is treated as serious contamination around aircraft structures. Remove the film's ability to regenerate and aluminium behaves like the reactive metal it actually is.

Passivation is not a coating and not really a property of the metal. It is a dynamic equilibrium — a film being continuously maintained against damage, using oxygen from the environment. Anything that interrupts the supply of oxygen or the chemistry of repair converts a corrosion-resistant metal back into an ordinary one.

Starve it of oxygen and it fails where you cannot look

Follow that logic and the first failure mode appears immediately.

Under a gasket. Under a washer or a bolt head. Under a deposit of scale, dirt or biofilm. Inside a lap joint. In a narrow crack. In all of these, water can get in but fresh oxygen cannot circulate, and the oxygen present is consumed. The passive film in that crevice can no longer repair itself, and the local chemistry drifts acidic and chloride-enriched as corrosion proceeds.

This is crevice corrosion, and its signature is cruel: the freely exposed surfaces, which have all the oxygen they need, stay bright and perfect, while the metal in the gap you cannot inspect is being eaten. A stainless bolt in a stainless flange can be severely attacked under its head while the visible hardware looks brand new.

The design consequence is specific and worth carrying: stainless steel is at its best fully exposed and washed, and at its worst tucked under something. Good design avoids tight crevices, allows drainage, avoids trapped deposits, and does not rely on visual inspection of the accessible surfaces to reveal the condition of the hidden ones.

Chloride, and the pit that digs itself

The second failure mode is the one that dominates coastal India.

Chloride ions have a specific ability to break down the passive film locally, displacing oxygen at a point and preventing repassivation there. Once that happens at a single spot, the process becomes autocatalytic, which is what makes it dangerous.

Inside the tiny pit, metal dissolves. The dissolving metal ions hydrolyse, which makes the solution inside the pit acidic. Chloride migrates in to balance the charge, so the pit interior becomes an acidic, concentrated chloride environment — precisely the conditions under which the film cannot re-form. The pit therefore maintains its own aggressive chemistry and drives itself deeper, while the surrounding surface, sitting in ordinary aerated water, stays passive and shiny.

The result is a failure mode that defeats casual inspection completely. A pit can penetrate the full thickness of a sheet while over 99% of the surface looks flawless. Leaks appear with no visible corrosion around them. This is why pitting is assessed by deliberate inspection and, where it matters, by testing — not by whether something looks rusty.

Everything that raises chloride concentration makes this worse: sea air, marine splash, swimming pool atmospheres, de-icing salt, and many cleaning chemicals. Higher temperature makes it worse. Standing water that evaporates and concentrates salt makes it much worse.

Why 304 and 316 are not interchangeable

This is the practical payoff, and it is a fact most buyers of stainless steel do not know.

The common grades look identical, cost differently, and behave very differently in chloride. 304 stainless, the standard kitchen and architectural grade, contains chromium and nickel. 316 adds roughly 2 to 3% molybdenum, and molybdenum substantially improves resistance to chloride pitting and crevice attack.

The industry quantifies this with a pitting resistance equivalent number, PREN, calculated roughly as the chromium content plus 3.3 times the molybdenum plus 16 times the nitrogen. It is an index rather than a guarantee, but it captures the ranking correctly and explains why 316 is specified for marine, coastal, pool and food-processing service while 304 is adequate inland.

Two consequences follow. Specifying 304 for a coastal application is a common and expensive mistake, and it will look fine for a while before pitting appears. And because the two grades are visually indistinguishable, substitution is easy — which is why material certificates exist, and why a portable analyser is standard equipment for anyone who cares.

The three ways it gets ruined in the workshop

Most stainless steel that rusts in service was compromised during fabrication, in one of three recognisable ways.

Heat tint at welds. Welding heats the surrounding metal enough to grow a thick, visibly coloured oxide scale, and that scale is not the protective passive film — it is chromium-depleted and the metal beneath it is impoverished of chromium too. Those straw-and-blue bands beside a weld are where stainless rusts first. The remedy is a specified process: pickling to remove the scale and depleted layer, followed by passivation to restore a clean film. This is a manufacturing step, not cosmetic finishing, and skipping it is the single most common cause of new stainless rusting within months.

Sensitisation. Held for too long in a temperature range around 450 to 850 °C, chromium and carbon combine into chromium carbides at grain boundaries, which robs the metal immediately adjacent of the chromium it needs to stay passive. The result is corrosion that runs along grain boundaries. This is exactly why low-carbon "L" grades such as 304L and 316L exist, and why welding procedures control heat input and time.

Free iron contamination. Grinding or brushing stainless with a disc or wire brush previously used on carbon steel embeds ordinary iron particles in the surface. Those particles rust, and the result looks precisely like the stainless steel failing. It is not — it is carbon steel rusting on top of perfectly good stainless. This is why dedicated stainless-only abrasives and brushes are a rule in good fabrication shops, and why the fix is cleaning and passivating rather than replacing the part.

Why the sink in your bathroom rusts and the railing outside does not

Domestic stainless rusts for exactly the reasons above, recombined.

Bathrooms and kitchens supply chloride from cleaning products, salt and food residues. They supply standing water that evaporates and concentrates that chloride. They supply deposits — soap film, hard-water scale, tea stains — under which oxygen is depleted. And they often supply steel wool or a carbon-steel scourer, which leaves iron particles behind.

An exposed outdoor railing, by contrast, is rinsed by rain, dried by air and washed clear of deposits, and has unlimited oxygen. The same grade of steel is in a much easier situation.

The practical advice follows directly and is unglamorous: rinse and dry surfaces rather than leaving salty water to evaporate on them, do not leave chloride-based cleaners in contact, do not leave steel wool or a wet carbon-steel pan sitting on a stainless surface, and remove deposits rather than working around them.

Nanotechnology that arrived a century early

It is worth noticing what stainless steel actually is, in the vocabulary this site normally uses.

It is a bulk structural material whose entire defining property comes from a two-to-three-nanometre oxide layer whose composition differs from the metal beneath it, engineered into existence by choosing the alloy content so that the right oxide forms spontaneously. That is surface nanoengineering, practised industrially since the early twentieth century, decades before the language existed — much as the Lycurgus cup was plasmonics long before anyone could say so.

Where modern nanoscience genuinely contributes is in seeing and extending it. The composition and thickness of a passive film are measured by surface-sensitive techniques — X-ray photoelectron spectroscopy and related methods — which is the only way to know what is actually there, and the same problem of measuring a nanoscale object indirectly applies. And deposited oxide layers, sol-gel films and atomic-layer coatings can add barrier protection to surfaces that cannot passivate on their own.

But there is an honest limit, and it is the interesting part. A deposited film, however perfect, does not regenerate. Scratch it and the scratch stays. The property that makes stainless steel and aluminium so useful is not that their oxide is good — many oxides are good — but that it is continuously rebuilt from the material underneath. Replicating that self-healing behaviour with an engineered coating remains a genuinely hard problem, and it is why a hundred-year-old alloy strategy has not been displaced by better deposition.

Why it matters for students and researchers

The central lesson is one that transfers everywhere: "corrosion resistant" is not a property of a material. It is a property of a material in an environment, and the environment includes oxygen availability, chloride concentration, temperature, whether water can evaporate and concentrate, and whether there is a crevice. A datasheet that says "corrosion resistant" without naming an environment has said very little.

It is also a good lesson in inspection bias. Both of the failure modes that matter here — crevice corrosion and pitting — are localised and preferentially occur where you cannot see, while leaving the visible surface in excellent condition. Any inspection regime that relies on what looks rusty will systematically miss the failures that actually happen, which is an uncomfortable and general point about condition monitoring.

The open problems are practical and active. Genuinely self-healing engineered coatings — films that regenerate from a reservoir when damaged — would extend passive behaviour to materials that cannot passivate. Predicting pitting initiation remains statistical rather than deterministic, so design still leans on conservative grade selection. Lower-nickel and lower-molybdenum alloys with equivalent chloride resistance matter because both elements are expensive and supply-constrained. And for India specifically, coastal-atmosphere performance data on the grades actually sold in the domestic market is thinner than the length of the coastline warrants.

Frequently asked questions

Why is my stainless steel sink or tap rusting?

Almost always chloride, oxygen starvation, or contamination — often together. Cleaning products and food residues supply chloride; standing water evaporates and concentrates it; soap film and scale create low-oxygen patches where the film cannot repair; and steel wool leaves iron particles that rust on top of perfectly sound steel. Rinsing and drying, removing deposits, and never using carbon-steel scourers fixes most domestic cases.

Is 316 stainless worth the extra cost near the sea?

Yes, and this is one of the clearer material choices there is. The 2 to 3% molybdenum in 316 substantially improves resistance to chloride pitting and crevice corrosion, which is exactly the attack coastal air delivers. The grades look identical, so ask for a material certificate — and note that even 316 is not immune in genuinely marine splash conditions, where higher alloys are used.

Does scratching stainless steel ruin its corrosion resistance?

No, in normal circumstances — that is the point of passivation. Fresh chromium is exposed, oxygen reaches it, and a new film forms within seconds. What does cause trouble is scratching it with a tool contaminated with carbon steel, which embeds iron particles, or creating a rough, crevice-rich finish that traps deposits and starves the surface of oxygen. The scratch is not the problem; what the scratch leaves behind can be.

Why does stainless rust at the weld and nowhere else?

Two reasons, both fixable. Welding grows a thick coloured oxide scale, and both that scale and the metal under it are depleted of chromium, so the protective film cannot form properly. And if the heat was held too long in the sensitising range, chromium carbides form at grain boundaries and rob the adjacent metal of chromium. The specified remedy is pickling to remove the affected layer followed by passivation, and using low-carbon grades where welding is extensive.

Can I clean stainless steel with steel wool?

No. Ordinary steel wool is carbon steel, and it leaves iron particles embedded in the surface which then rust, producing what looks like the stainless steel failing. Use a nylon pad, a stainless-specific scourer or a proprietary stainless cleaner, and always rub in the direction of any existing grain finish. If iron contamination has already occurred, a passivating cleaner will usually remove it and restore the surface.