The Making of Aluminium
From Rock to Lightweight Metal

The Making of
Aluminium

It is the most abundant metal in the Earth's crust, yet it clings so tightly to oxygen that no one could unlock it until the 1880s. Freeing it still takes one of the most electricity-hungry processes in all of industry — and the world now makes about 74 million tonnes of it a year.

Metal Transportation Packaging Recycling uses 5% of the energy

Light enough that you stop noticing it is metal at all.

Drink cans Drink cans
Aircraft Aircraft
Window frames Window frames
Laptop bodies Laptop bodies
Kitchen foil Kitchen foil
EV bodies EV bodies
Once worth more than gold

For most of history, no one could get at it. Aluminium was only isolated as a metal in the 1820s, and prising it loose from oxygen was so hard that the metal was worth more than gold: France put aluminium bars on show beside the crown jewels, and Emperor Napoleon III is said to have saved his aluminium cutlery for his most honoured guests. In 1884 a small pyramid of it — then about as costly as silver — was chosen to cap the Washington Monument. Then, in 1886, two 22-year-olds working independently — Charles Martin Hall in Ohio and Paul Héroult in France — found they could split alumina apart with electricity. Within a generation the price fell from hundreds of dollars a pound to pennies. Aluminium had not become any more common — electricity had.

The journey

From red rock to bright metal — a change of colour

The whole journey reads as a change of colour: a red rock is mined, washed into a white powder, then split apart with electricity into a silver metal — bauxite, to alumina, to aluminium. Tap any stage to look closer.

↓ tap any stage to explore ↓

1 · Bauxite Dug from tropical soils ~250°C 2 · Refining Washed into white alumina ~960°C 3 · Smelting Split apart by electricity Ingot → remelt & cast Billet → extrusions Slab & coil → sheet & foil 4 · Cast Forms

What Aluminium Is Made Of

An abundant metal, locked in rock

The idea

Aluminium is everywhere in the ground, but never as metal — it is always bonded to oxygen. Making it is really about prising those two apart, and oxygen does not let go easily.

Aluminium makes up roughly 8% of the Earth's crust — more than any other metal — yet you will never find a nugget of it. It bonds to oxygen so fiercely that it took chemists until the 1880s to work out how to separate the two, and the method they landed on is still the one used today. Getting to finished metal takes two very different transformations back to back: a chemical wash that pulls a pure white powder out of red rock, and then a torrent of electricity that tears that powder apart. Three ingredients make it possible.

Bauxite

The ore. A reddish rock rich in aluminium hydroxide minerals — gibbsite Al(OH)3 and boehmite AlO(OH). Its rusty colour comes from iron oxide; silica and titania ride along as impurities. Chief sources: Guinea, Australia, China and Brazil.

Caustic Soda

Hot, concentrated sodium hydroxide (NaOH) is the solvent that does the washing. Under heat and pressure it dissolves the aluminium out of crushed bauxite, leaving the iron oxides and other insoluble minerals behind as a residue known as red mud.

Cryolite & Carbon

Alumina melts at 2,072°C — far too hot to handle directly. So instead of melting it, the process dissolves it in molten cryolite (Na3AlF6), a salt bath that is itself liquid at about 960°C. Electrolysis then happens in that bath, not in molten alumina. Carbon blocks carry the current in, and are slowly eaten away as they do.

The real raw material isn't in any of these cards — it's electricity. Smelting one tonne of aluminium draws around 14 megawatt-hours, roughly what a UK home uses in five years.
Two Roads to the Metal

Make it from rock, or melt it from scrap

The idea

Making aluminium from ore is one of industry's most energy-hungry jobs. Making it from old aluminium is one of its cheapest — the same metal, for a twentieth of the energy.

Every piece of aluminium takes one of two routes. The primary route starts from bauxite and runs it through two plants — a refinery that washes out pure alumina, then a smelter that splits that alumina with electricity. It is the only way to make brand-new metal, and it is enormously power-hungry. The secondary route skips all of that and simply re-melts scrap. Because aluminium can be recycled without any loss of quality, remelting needs only about 5% of the energy of the primary route — which is why recycled metal now makes up the majority of supply in places like the United States, even as fresh smelting still dominates worldwide.

Primary · "Virgin" Metal

Bayer Refining → Hall-Héroult Smelting

1
Bayer DigestionCrushed bauxite is stirred into hot, pressurised caustic soda. The aluminium dissolves out as sodium aluminate; the iron oxides and other insolubles are left behind as red mud. The clear liquor is then cooled and seeded so pure aluminium hydroxide crystallises back out.
~150–250°C under pressure
2
Calcination → AluminaThose hydroxide crystals are roasted to drive off their water, leaving a dry white powder: alumina, Al2O3. Around 4 tonnes of bauxite yields 2 tonnes of alumina.
~1,000–1,100°C
3
Hall-Héroult ElectrolysisAlumina is dissolved in a molten cryolite bath and a huge direct current is passed through it. Aluminium metal collects at the bottom of the cell; oxygen is stripped off and burns the carbon anodes away as CO2. Roughly 2 tonnes of alumina makes 1 tonne of metal.
~960°C in the cell
Secondary · Recycled

Remelting Scrap

1
Sort & PrepareUsed cans, extrusions, car parts and offcuts are collected, shredded and sorted by alloy. Coatings and paint are burned off in a clean-up step before melting.
2
Melt It DownAluminium melts at just 660°C — far cooler than it takes to smelt from ore. Scrap is melted in a furnace and skimmed clean, with no bauxite, no refinery and no electrolysis needed.
melts at ~660°C
3
Adjust & CastThe chemistry is tuned by adding a little fresh metal or alloying elements, then the molten aluminium goes to the same casting steps as the primary route — ready to become new product.

Then they converge. Whichever road it came down, molten aluminium ends up in the same holding furnace, where alloying elements are stirred in to fix its recipe. From there it is cast into ingots, billets and slabs, then rolled into sheet and foil, extruded into window frames and tubes, or cast into finished shapes. One metal, two beginnings, a single finish.

Not All Aluminium Is Equal

Families of aluminium alloy

The idea

Pure aluminium is too soft to be much use. Stir in a pinch of another metal and it transforms — and which metal you choose decides everything about the family it belongs to.

On its own, aluminium is soft and easily bent. Its usefulness comes from alloying — dissolving a little copper, magnesium, silicon, manganese or zinc into it. The industry sorts these into numbered families, where the first digit names the main added metal.

+ Copper · high strength

2xxx

Strong as some steels — the alloy that flies. Aircraft skins and structures.

+ Manganese · formable

3xxx

Easy to shape and draw. The beverage-can alloy, also roofing and cookware.

+ Magnesium · corrosion-proof

5xxx

Shrugs off seawater. Boat hulls, pressure vessels and marine structures.

+ Mg & Silicon · all-rounder

6xxx

The workhorse — easy to extrude into long shapes. Window and bike frames, railings.

By the Numbers

The scale of a power-hungry industry

The idea

Bauxite is mined where it lies, but the metal is made where the electricity is cheap — and increasingly, that means one country.

Aluminium's supply chain funnels down at every step: about 440 million tonnes of bauxite become 150 million tonnes of white alumina, which become roughly 74 million tonnes of finished metal. And where each of those steps happens is not the same map — the ore comes from the tropics, but the smelting has migrated to wherever power is cheapest.

74 Mt
Primary aluminium smelted worldwide in 2025
440 Mt
Bauxite ore mined worldwide in 2025 to feed it
~16 t
Tonnes CO₂e per tonne of primary metal, on a coal-heavy grid
5%
Energy to recycle a tonne versus smelting it fresh
61%

The concentration problem, in one number. China smelts about 45 million tonnes of the world's 74 — some 61% of all primary aluminium — despite mining only a fifth of the bauxite it needs. It runs right up against its own self-imposed ceiling: a 45 Mt cap introduced in 2017 to curb overcapacity and emissions, which national capacity finally hit in 2024. No other major metal is made so overwhelmingly in one place.

The same tonne of aluminium, four ways

Coal-powered primary
~16 t CO₂e
Global average primary
~15 t CO₂e
Hydro-powered primary
~4 t
Recycled secondary
0.5 t

Aluminium has no single carbon footprint — it depends almost entirely on where the electricity comes from. The same metal can carry 16 tonnes of CO₂ or 4, decided only by whether its smelter runs on coal or hydropower. Recycling collapses it to about half a tonne. That is why more than 90% of the industry's emissions come from primary production, and why cleaning up aluminium is, first and last, an electricity story.

Two routes to a tonne — and who leads each

A tonne of aluminium can be smelted fresh from ore or remelted from scrap, and the two routes have different geographies. Primary smelting is extraordinarily concentrated in China; recycling — though also China-led — is spread more widely across scrap-rich economies. Japan, Germany and South Korea lead the tier just behind those shown.

Who smelts primary metal

2025 · Mt smelted from ore
China
45
India
4.2
Russia
3.9
Canada
3.3

Who remelts scrap metal

2024 · Mt recycled
China
~16
USA
3.6
India
~2.1
150 MtBauxite · Guinea. The world's largest ore source in 2025 — and nearly all of it is exported, since Guinea barely refines or smelts at home.
93 MtAlumina · China. The world's top refiner in 2025 — well over half the global total, feeding its giant smelter fleet.

China leads both routes and India appears in each, so the two maps aren't opposites — what sets them apart is everything below the top. Primary smelting migrates to wherever power is cheapest, which is why Russia and Canada rank so high, and why China alone makes about 61% of the world's fresh metal; bauxite and alumina then travel oceans by ship to reach those smelters. Recycling instead follows the scrap, taking root in mature, metal-rich economies: the United States now smelts very little, yet more than three-quarters of its aluminium supply comes from remelting rather than smelting.

The Power Problem

Cleaning up the metal that runs on electricity

The idea

Most of aluminium's carbon is just the electricity it drinks — clean the grid and that shrinks. But a stubborn slice is baked into the chemistry itself, and no amount of clean power removes it.

Aluminium's emissions split into two very different problems. The larger part is simply the electricity that electrolysis devours — clean that power and it falls away. The smaller, tougher part comes from the reaction itself: the carbon anodes are consumed and turned into CO2, and that happens no matter how green the grid is. Solving aluminium means tackling both.

The anode is part of the recipe, not just the wiring

In a normal cell, the carbon anodes do two jobs: they carry the current, and they chemically soak up the oxygen freed from the alumina. That second job turns them into carbon dioxide — the anodes are literally eaten away and must be replaced.

2 Al2O3 + 3 C 4 Al + 3 CO2 carbon anode
Today's process. The carbon anode is consumed. Waste product: carbon dioxide.
2 Al2O3 4 Al + 3 O2 inert anode
The goal. Swap in an anode that isn't consumed. Waste product: pure oxygen.
Both are written as simplified overall reactions; inside the cell each runs as separate half-reactions at the cathode and anode.

Those carbon anodes account for almost a sixth of the greenhouse gases behind new aluminium — the part that a clean grid alone can never remove. Replacing them is a chemistry change, not a wiring change.

The inert-anode breakthrough

In November 2025, ELYSIS — a joint venture of Alcoa and Rio Tinto — started up the first commercial-size inert-anode cell, a 450-kiloampere pot slotted into an existing potline at Rio Tinto's Alma smelter in Quebec. It makes aluminium with no direct carbon emissions from smelting, releasing oxygen instead of CO2.

The pilot version has already made metal used in some Apple laptops, Audi wheels and Michelob cans. The next step is a multi-year test to prove the anodes survive full industrial conditions, with the aim of maturing the technology by the end of the decade.

It is aluminium's equivalent of hydrogen steelmaking: not a cleaner fuel poured into the old process, but a genuinely different reaction.

Why is this so hard to shift, when the fix already exists?

What holds the change back
  • Most smelters run on coal. A huge share of the world's aluminium is made in China on a coal-heavy grid, which is exactly what makes the average footprint so high.
  • Smelters last decades. A potline is a vast capital investment built to run for 30–40 years, so there's little appetite to rip one out early.
  • Demand keeps rising. Aluminium demand is projected to grow more than 80% by 2050 for EVs, solar and power grids — if the recycled share stays flat, sector emissions climb with it.
  • Cheap power wins. Smelting migrates to wherever electricity is cheapest, which isn't always where it's cleanest.
What drives it forward anyway
  • The grid is decarbonising. As more clean power comes online, every existing smelter gets cleaner without changing a thing — the single biggest lever, and it's already moving.
  • Inert anodes now work at scale. The 2025 commercial-cell milestone removes the last emissions a clean grid can't, and can be retrofitted into existing plants.
  • Recycling is a free win. Secondary metal costs a twentieth of the energy; as more aluminium reaches end-of-life, the scrap pool grows every year.
  • Carbon costs and buyers. Border carbon levies and customers demanding low-carbon metal create a premium for clean aluminium.

Aluminium's problem was never scarcity — it's the most abundant metal in the ground. It's energy. The metal drinks electricity to be born, so its footprint rises and falls with the power behind the plug: coal-fired smelting emits around sixteen tonnes of CO2 per tonne of metal, hydropower roughly a quarter of that, and recycling barely half a tonne. A clean grid does most of the work, inert anodes now tackle the stubborn sixth baked into the chemistry, and recycling quietly carries the rest. As of 2025 all three are real — what decides how fast they spread is where the world chooses to plug its smelters in.