Side by side
Iron vs Steel
The numbers, side by side
| Property | Iron | Steel |
|---|---|---|
| Symbol | Fe | not an element |
| Atomic number | 26 | not an element |
| Atomic weight | 55.845(2) | not an element |
| Category | transition metal | not an element |
| State at 20 °C | solid | not an element |
| Density | 7.874 g/cm³ | not an element |
| Melting point | 1811 K | not an element |
| Boiling point | 3134 K | not an element |
| Electronegativity | 1.83 | not an element |
| Electron configuration | [Ar] 4s2 3d6 | not an element |
| Discovered | antiquity | not an element |
One side of this comparison is not a chemical element, so it has no single set of element data. The written comparison below is where that side is dealt with properly.
Setting these two against each other means setting an element against a recipe. Iron is element 26; steel is iron with a small, deliberate amount of carbon dissolved in it, plus whatever else the specification calls for. There is no such thing as a melting point of steel or an atomic number of steel, and any page that gives you one is quietly comparing iron with iron.
The useful answer is that almost nothing is made of iron and almost everything is made of steel, because a couple of tenths of one per cent of carbon turns a soft, nearly useless metal into the cheapest strong material humanity has.
What two-tenths of a per cent does to a lattice
High-purity annealed iron yields at something in the region of fifty megapascals. Ordinary mild structural steel, which is the same metal with roughly 0.2% carbon in it, yields at around two hundred and fifty. Five times the strength, for an addition you could not see, weigh out by eye or taste.
Carbon atoms are small enough to sit in the gaps between iron atoms rather than replacing them, and in those interstitial positions they obstruct the movement of dislocations — the line defects whose motion is plastic deformation in a metal. Block the dislocations and the metal stops yielding.
Above the solubility limit the carbon precipitates as iron carbide, and the arrangement of that carbide is what steel heat treatment manipulates. Cooled slowly, ferrite and carbide grow as alternating lamellae; quenched fast enough, the carbon has no time to move at all and the lattice distorts around it into martensite, which is glass-hard and about as tough as glass until it is tempered back. Every knife blade, gear tooth and spring is somewhere on that spectrum.
The carbon ceiling for steel is about 2.1% by weight — the most that will dissolve in the high-temperature austenite phase. That number is not a convention. It is where the iron-carbon phase diagram changes character, and it is the line between two entirely different families of material.
Cast iron is the other side of the same line
Push past 2.1% — commercial castings usually sit between 2% and 4% — and the excess carbon has to come out of solution as a separate phase. In grey cast iron it appears as flakes of graphite threaded through the metal, and those flakes behave like a dense population of internal cracks. The result is a material that is weak and brittle in tension, superb in compression, cheap to pour into complicated shapes, easy to machine, and unusually good at damping vibration. Machine tool bases are cast iron for the damping alone.
Treating the melt with magnesium makes the graphite form as spheres instead of flakes, and ductile iron — the same carbon content, a different geometry — is genuinely tough. Most buried water mains are made of it.
So "iron versus steel" is really a three-way split along one axis: not enough carbon to be useful, the right amount, and too much to be ductile.
Wrought iron, and the tower that is not made of steel
The historical material called wrought iron is not pure iron either. It carries under 0.08% carbon and one or two per cent of siliceous slag, drawn out into fibres by hammering, which gives it a grain like timber and a stubborn resistance to crack propagation across that grain.
The Eiffel Tower is built from puddled wrought iron, not steel, even though the Bessemer converter had been in commercial use for three decades when construction began. Eiffel's engineers knew precisely how puddled iron behaved in a lattice structure and how it responded to riveting, and that certainty was worth more than the newer material's numbers. A great many Victorian bridges and train sheds are in the same position, and their survival is partly a corrosion story: the slag inclusions and phosphorus content of old wrought iron often make it rust more slowly than the mild steel that replaced it.
Neither of them resists rust
This is where the comparison is most often got backwards. Steel is not the corrosion-resistant version of iron. Plain carbon steel rusts freely, and adding carbon does nothing whatever for oxidation resistance — if anything the carbide phases give corrosion more to attack.
Rust resistance in steel comes from other alloying elements entirely. At 10.5% chromium and above, a chromium oxide film forms that is thin, transparent, adherent and self-repairing, and the steel is called stainless. Weathering steels take a smaller addition of copper, chromium, nickel and phosphorus and use it to grow a tight, adherent rust layer that seals the surface instead of flaking off it. Both are steels; neither owes anything to its carbon.
The jobs that really do want the metal nearly pure
There is a genuine market for iron with as little in it as possible, and it is magnetic rather than structural. Pure iron has high magnetic permeability and low coercivity — it magnetises easily and lets go easily — which is exactly what a transformer core, a relay armature or a solenoid plunger needs. Carbon ruins that. Transformer laminations are made from electrical steel, which is iron with about 3% silicon added to raise electrical resistivity and cut eddy current losses, accepting brittleness in exchange.
Iron powder is a commodity in its own right for sintered components and for food fortification, and iron oxides are among the oldest pigments in use. None of those applications wants carbon in the iron.
Cost decides the rest. Steel is produced at nearly two billion tonnes a year and is among the cheapest engineering materials per unit of strength ever made. Iron refined to laboratory purity costs orders of magnitude more per kilogram and is bought by the gram.
What to specify, and what to call it
- A beam, a bolt, a car body, a ship — steel, and the only real question is which grade.
- Something that must be poured into an intricate shape and loaded in compression — grey cast iron.
- A buried pipe — ductile iron.
- A magnetic core — the lowest-carbon iron you can get, silicon-alloyed if it will see alternating current.
- Repairing a nineteenth-century structure — match the original wrought iron rather than substituting steel, because the two corrode and move differently.
- Anything sold to you as "iron" in a shop — read it as cast iron, which is a different material from both.