Passivation
A Socratic walk-through of passivation — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does stainless steel stay bright when it is mostly the same iron that rusts away in a nail?
A stainless sink and an iron nail are both mostly iron. Left in the same damp kitchen, the nail turns to flakes and the sink does not. The usual explanation — that stainless "does not oxidise" — cannot be right, because the alloying element that saves it, chromium, is more reactive than iron, not less. Aluminium makes the point sharper still: it is thermodynamically eager to oxidise, energetic enough to be used as rocket fuel, and yet a window frame of it survives decades of rain.
So the puzzle is not why some metals avoid reacting with oxygen. It is why the ones that react most eagerly are sometimes the ones that stop.
Reasoning it through
REASONING #Take the question apart. Every one of these metals oxidises on first contact with air. The difference must lie not in whether an oxide forms, but in what the oxide then does.
Consider the companion case in this collection, why iron rusts. Iron's rust is a hydrated oxide occupying far more room than the metal it came from, with no crystallographic relationship to the iron beneath. It grows, builds stress, cracks, and flakes away, exposing fresh metal. The oxide is a product that does nothing to hinder the reaction, so corrosion proceeds at a roughly steady rate until there is no iron left.
Now imagine the opposite. Suppose an oxide formed as a continuous film, tightly bonded, with no cracks or pores. Any further reaction would need oxygen or metal ions to cross it — so the rate would fall, and keep falling, because each new layer lengthens the crossing. The film would grow quickly and then almost stop, self-limiting by construction. Aluminium's natural film is a few nanometres thick, forms in milliseconds, and there it stays. That is passivation: a negative feedback in which the product of a reaction throttles the reaction.
What must an oxide satisfy to behave like that? One condition is geometric and can be written down. Compare the volume of oxide produced with the volume of metal consumed in making it — the Pilling-Bedworth ratio. Below one, the oxide cannot cover the surface it came from and grows in porous islands with bare metal between them; magnesium sits at about 0.81, and corrodes badly. Far above two, the oxide is so much bulkier than the metal beneath that compressive stress buckles it and it spalls off as scale, which is what tungsten does at temperature. Roughly between one and two, coverage is achievable: aluminium is about 1.28 and chromium about 1.99.
But the ratio is necessary, not sufficient — the honest complication. Iron's own oxides fall in that window too, near two, and iron plainly does not passivate in damp air. The film must also adhere, match the metal well enough not to crack under stress, and pass ions poorly. Chromium's oxide is dense and tenacious and admits ions extremely slowly; iron's hydrated rust is loose, permeable and mechanically weak.
Now the sink resolves. Add enough chromium — conventionally at least about 10.5 per cent — and the film that forms is chromium-oxide-rich rather than iron-oxide-rich, because chromium oxidises preferentially and its oxide is the stable one at the surface. The bulk is still mostly iron. One to three nanometres of skin is doing all the work.
The best evidence that it is a skin is what happens when you cut it. Scratch stainless and the exposed metal reoxidises within moments — the protection is restored rather than lost. Passivation is not armour applied once, but a repair loop that runs whenever oxygen is available.
The analogy
THE ANALOGY #Think of a wound that scabs. The scab is made of the injury — it is the reaction's own product — and once formed it slows what created it, and if you knock it off it forms again from the fresh surface underneath. What matters is not that the tissue is unreactive but that the response seals rather than crumbles.
A scab is built by an active biological system that senses the damage, whereas an oxide film is entirely passive chemistry — there is no repair mechanism beyond the plain fact that bare metal meets oxygen, which is why removing the oxygen removes the healing.
Clarifying the model
THE MODEL #Two clarifications follow from the loop, and both matter in practice.
Passivity needs an oxidiser. The film is maintained by access to oxygen, so stainless steel in an oxygen-starved crevice — under a gasket, beneath a deposit, inside a tight lap joint — cannot repair itself and corrodes there while the exposed surface a centimetre away stays bright. Crevice corrosion is not an exception to passivation; it is passivation deprived of its input.
The other failure mode is chemical, and chloride is the culprit. The chloride ion is small and adsorbs strongly at defects in the film, where it helps convert the oxide into soluble metal chlorides and opens a tiny hole. Inside that hole the loop reverses sign: metal ions dissolve and hydrolyse, the local water turns acid, the acid attacks the film from within, and chloride migrates in to balance the charge. Instead of a feedback that shuts the reaction down, you have one that accelerates it, concentrated on a spot a fraction of a millimetre across. That is pitting, and it is more dangerous than general rusting precisely because it is localised: a component can be perforated while losing almost no mass and looking, from outside, clean. Molybdenum and nitrogen make the film markedly more chloride-resistant, which is why marine and food-processing grades contain them — though the formula that ranks alloys on their content is an empirical ordering, not a prediction of service life.
A picture of it
THE PICTURE #How to readEach point is a metal-and-environment pairing rather than a metal alone. Read across for whether the oxide actually covers the surface it grew from, and up for whether it comes back after damage; only the top-right corner is genuine passivity, because a film needs both properties at once. The instructive comparison is the two stainless points: nothing about the alloy differs between them, only the presence of chloride, which drags the same steel out of the passive corner. Rust sits low and left because it neither covers nor reseals, which is the whole difference from the nail.
What became clearer
WHAT CLEARED #Passivation is a feedback loop, not an absence of reactivity. The most eager metals form an oxide instantly, and where that oxide happens to be continuous, adherent and slow to pass ions, it throttles the very reaction that made it — and rebuilds itself when scratched. Iron's oxide fails that test by flaking, chromium's passes it, and a few per cent of chromium in the alloy is enough to decide which oxide covers the surface. The loop's dependence on oxygen and its vulnerability to chloride are then not quirks but direct consequences of the same mechanism.
Where to go next
ONWARD #- Why anodising thickens aluminium's natural film deliberately, and what that buys beyond the natural few nanometres.
- How galvanising protects steel by the opposite strategy — sacrificing zinc electrochemically rather than sealing the surface.
Key terms
TERMS #| Term | What it means |
|---|---|
| Passivation | the formation of a surface film that slows or halts further corrosion of the metal beneath it. |
| Pilling-Bedworth ratio | the volume of oxide formed divided by the volume of metal consumed, a necessary but not sufficient test of whether a film can cover the surface. |
| Pitting corrosion | localised breakdown of a passive film, typically initiated by chloride, producing deep narrow holes with little overall mass loss. |
| Crevice corrosion | attack in an oxygen-starved gap where the passive film cannot be maintained. |
Every term the collection defines is gathered in the glossary.