A 6-Tonne Iron Pillar Has Stood in the Delhi Weather for 1,600 Years. It Still Hasn't Rusted Away

In the courtyard of a ruined mosque in Mehrauli, in south Delhi, stands a shaft of iron a little over seven metres long, weighing about six tonnes. It was forged around the year 400 CE. It has been standing outdoors — through sixteen centuries of monsoon, dust and 45-degree summers — ever since.

In all that time, it has grown roughly 200 micrometres of rust.

A fifth of a millimetre. Two sheets of paper. On a mass of exposed iron that predates the fall of Rome.

The oldest thing in the courtyard by eight centuries

The Iron Pillar is the odd one out in the Qutb Minar complex, the UNESCO World Heritage Site inscribed in 1993 for the Indo-Islamic architecture around it. The famous red sandstone tower is 72.5 metres of 13th-century Ghurid ambition. The pillar in its courtyard is eight hundred years older than the mosque built around it, and made of something else entirely.

It runs about 23 feet 6 inches from bottom to crown, including a decorated “bell” capital, with roughly a foot and a half sunk below ground level.

And it carries a receipt. A six-line Sanskrit inscription in Gupta-era Brahmi characters, set at about face height, records that the pillar was raised as a Vishnudhvaja — a standard of the god Vishnu — on a hill called Vishnupadagiri, in memory of a king named Chandra.

The king, and the hill that isn’t in Delhi

Palaeographers date those characters to roughly 400–450 CE, which puts the pillar squarely in the Gupta period. The consensus identification of Chandra — argued by R. Balasubramaniam of IIT Kanpur from numismatic, archaeological and literary evidence, including the Guptas’ archer-type gold coins — is Chandragupta II Vikramaditya, who reigned from about 375 to 414 CE.

Here is the part most visitors don’t know: the pillar isn’t from Delhi.

Balasubramaniam’s reconstruction places its original home at Udayagiri, in central India near Vidisha and Sanchi — a Gupta religious centre, probably standing before the great Varaha panel carved into the rock there. It was moved north around 1050 CE and erected at Lal Kot by the Tomar king Anangapala. When Qutb-ud-din Aibak took the fortress city and built the Quwwat-ul-Islam mosque between 1192 and 1199 — from the material of demolished temples — the pillar ended up in his courtyard, and stayed.

It has been pulled out of the ground and re-erected at least twice by people who did not make it. It is still standing.

Nobody cast this. They hammered it.

This is the part that deserves more awe than the rust does.

India could not pour liquid iron in 400 CE; furnaces capable of melting it don’t appear there until the 15th century. So the pillar was never cast. It was made by forge welding — iron produced by solid-state reduction of ore with charcoal, coming out of the furnace as lumps, and those lumps hammered onto the growing body of the pillar, by hand, one at a time, until there were six tonnes of it.

Balasubramaniam’s reading of the hammer marks and handling clamps suggests the main body was built up sideways, with the pillar lying horizontal, then finished smooth.

For scale: the West did not forge iron objects this large until the 19th century.

The capital on top isn’t a single lump either. It’s seven distinct parts, individually forge-welded, fitted around a hollow iron shaft and joined with lead solder — with a socket that once held a garuda idol, long gone. Someone planned this.

And it is tough. A cannonball fired at the pillar in the 18th century — by Nadir Shah’s men in 1739, or Ghulam Quadir’s in 1787 — failed to break it.

Now the myth, because it needs killing

The pillar is not rustless. It is not magic. It has not “defeated corrosion.” Even the classic 1970 review of the problem was titled The rustless iron pillar at Delhi — and the title was wrong.

Look closely and the evidence is everywhere. A band just below the decorative capital shows the maximum rusting on the whole exposed surface. At the forge-welded joints the rust runs black — magnetite, forming where the interface accelerates corrosion locally. The 200 micrometres of scale is not zero; it is a very small number achieved very slowly.

And the buried portion is in real trouble. When J.D. Beglar re-erected the pillar in the 1860s and built the stone platform, he coated the underground surface in about 3 mm of lead. Lead is cathodic to iron. The soil around it has been found loaded with soluble sulphates and chlorides. The result is galvanic corrosion: the iron below ground is corroding faster than the iron in the open air.

Worse, when the pillar was relaid on a new foundation in the 1960s, a fresh lead coating went on — against the wishes of the ASI’s own Chief Chemist. Balasubramaniam’s judgement is blunt: the buried region is now under intense galvanic attack, and that coating ought to be replaced.

The pillar is not immune. It is exceptionally resistant, in a specific environment, in a specific way. That’s a better story anyway.

The impurity that modern steelmakers pay to remove

Here’s the composition, by weight, averaged across the published analyses: about 0.15% carbon, 0.25% phosphorus, 0.05% silicon, 0.05% nickel, 0.05% manganese, 0.03% copper, 0.02% nitrogen, 0.005% sulphur. The rest is iron — very pure wrought iron, threaded with slag particles.

That phosphorus number is the whole ballgame.

Modern steels can’t tolerate anything like 0.25% phosphorus. It makes them crack during hot working, as liquid phosphides form at the grain boundaries — the defect is called hot shortness. Phosphorus is an impurity, and modern steelmaking spends real money removing it.

The Gupta smelters didn’t remove it. And the reason is beautifully mundane: they weren’t using limestone.

Modern blast furnaces add lime. Lime (CaO) in the slag is what strips phosphorus out of the metal — far more efficiently than the iron oxide already there. Ancient Indian furnaces added no lime; the slags they left behind are fayalitic — iron silicate, Fe₂SiO₄ — with no CaO in them. No lime in the slag, no efficient phosphorus removal, so the phosphorus stayed dissolved in the iron.

There are hints it wasn’t purely accidental, either. Francis Buchanan, documenting iron working in Karnataka in the 18th century, described a furnace at Devaraya Durga where crucibles were charged with wrought iron and a measured quantity of bark from Cassia auriculata — a plant whose bark happens to be high in phosphorus. Craft knowledge does not have to know chemistry to be reproducible.

What the rust is actually doing

The mechanism, worked out by Balasubramaniam at IIT Kanpur across the 1990s and published in Corrosion Science in 2000, goes in two stages — and it builds on an idea M.K. Ghosh first proposed in 1963.

Stage one. The slag particles in the iron are cathodic to the metal around them, so the pillar actually corrodes quickly at first. That initial attack does something useful: it strips iron away and leaves phosphorus concentrated at the surface. With phosphorus enriched at the metal–rust interface, a compact amorphous layer of δ-FeOOH forms right against the metal — a phase called misawite, after T. Misawa, whose work established its role in weathering steels. In ordinary mild steel this layer forms in patches. With phosphorus present, it forms a continuous film. That buys the initial protection.

Stage two — the real answer. Delhi’s weather cycles wet and dry. Moisture plus surface phosphorus makes phosphoric acid; the acid attacks the iron into soluble phosphates; and as the surface dries, those precipitate. Over centuries the amorphous phosphate crystallises into a thin, dense, low-porosity layer of crystalline iron hydrogen phosphate hydrate — FePO₄·H₃PO₄·4H₂O — sitting directly against the metal.

That layer is the barrier. Balasubramaniam’s rust characterisation — X-ray diffraction, infrared spectroscopy and Mössbauer spectroscopy on rust scraped from the pillar — found exactly that crystalline phosphate, with the iron oxides and oxyhydroxides around it amorphous.

The film grows by parabolic kinetics, which is the signature of a genuinely protective scale: the thicker it gets, the slower it gets. Wranglen predicted about 200 µm after 1,600 years from that model. Bardgett and Stanners went and measured it with a magnetic thickness gauge. The numbers matched.

The pillar built its own armour, out of its own corrosion, and then very nearly stopped.

Delhi helped. It didn’t do the work.

Give the environment its due. Relative humidity in Delhi doesn’t stay above 70% for long stretches, and below roughly 70% the atmospheric rusting of iron is slow — that was the old “environmental” explanation, and it isn’t nothing. Six tonnes of metal also has enormous heat capacity, so the pillar heats and cools out of step with the air around it, sharpening exactly the wet–dry cycling the phosphate film needs.

But climate alone can’t carry the argument, for a simple reason: ancient Indian iron survives where the air is wet. There are iron beams at the Surya temple at Konarak, on the coast of Odisha. There’s an iron pillar at the Mookambika temple at Kollur, in the hills near the west coast. The metal, not just the meteorology, is doing something.

And the clinching test is unkind to the mystics. Immerse samples of the pillar’s iron in dilute solutions and it corrodes briskly — 6 mg/dm²/day in dilute salt, 54 in dilute sulphur dioxide. Had the pillar stood in water rather than Delhi air, it would have been eaten alive.

Craft got there first

None of this is lost super-science. There is no forgotten formula, no anomalous alloy. It’s an ore, a charcoal fire, no limestone, a phosphorus content a modern mill would reject as out of spec, six tonnes of hand-hammered wrought iron, and a dry-ish city — a combination that happened to grow a self-repairing phosphate film a fraction of a millimetre thick.

The Gupta smiths did not know what δ-FeOOH was. They could not have written down FePO₄·H₃PO₄·4H₂O. What they had was a process refined by hand across generations, producing a result they could see and we needed spectroscopy to explain.

That’s the pattern we keep finding in Indian metallurgy. It’s the same story as wootz, the crucible steel of the south that armed Damascus and defeated Faraday’s chemistry: makers reliably producing an effect the science of the day could not account for, because craft was running centuries ahead of theory.

The pillar has outlasted the temple it was made for, the empire that made it, the kingdom that moved it, and the sultanate that built a mosque around it. It now stands behind a protective grille the ASI installed in 1997 — put there because visitors had a habit of standing with their backs to it and clasping their hands around the shaft, and had polished a band of it mirror-bright.

Sixteen hundred years of weather couldn’t get through the metal. Human affection nearly did.