Aluminum (Al)

When the Washington Monument was capped in December 1884, the builders chose for its apex the largest single piece of aluminum ever cast — about a hundred ounces of it. The metal was selected partly because it would not corrode and partly as a display of wealth: it cost roughly $225 at a time when a labourer might earn a dollar a day. Within fifteen years the same quantity of aluminum was worth a few dollars. Nothing about the metal changed. What changed was that two young men independently worked out how to make it.

Abundant, and impossible to get at

Aluminum is the most common metal in Earth's crust and the third most common element there after oxygen and silicon. It is also, in geological terms, never found as metal. The bond between aluminum and oxygen is strong enough that no natural process on this planet reduces it, and no furnace of the kind that has been smelting copper for six thousand years and iron for three thousand will touch it. Charcoal takes oxygen away from iron oxide readily; it cannot take oxygen away from alumina.

This is why an element making up something like eight per cent of the crust by mass was unknown as a substance until the nineteenth century. Hans Christian Ørsted produced an impure sample in 1825 and Friedrich Wöhler an aluminum powder in 1827, both by chemical routes using potassium. Henri Sainte-Claire Deville improved the method enough to make ingots, and bars of it were exhibited at the Paris Exposition of 1855 under the heading of silver from clay. It was a curiosity for jewellers and for the imperial court, and the widely retold story that Napoleon III reserved aluminum cutlery for his most honoured guests while everyone else made do with gold is one of those anecdotes that is repeated everywhere and sourced nowhere very convincingly.

Two men, one year, one solution, and the same lifespan

In 1886 Charles Martin Hall, working in a woodshed behind his family home in Oberlin, Ohio, dissolved alumina in molten cryolite and passed a current through it. That same year, Paul Héroult in France did the same thing. Neither knew of the other. Both were twenty-two, both had been born in 1863, and — as if the coincidence needed a final flourish — both died in 1914.

The insight they shared was about the solvent rather than the electricity. Alumina melts far too high to electrolyse directly, but dissolved in molten cryolite it forms a conductive bath at around 960 °C, which is manageable. That process, universally called Hall-Héroult, is still how essentially all primary aluminum is produced today, with the notable modern change being that almost all of the cryolite is now synthetic, since the natural Greenland deposit was worked out.

Karl Josef Bayer supplied the other half of the industry two years later, with a process for extracting pure alumina from bauxite using hot caustic soda. Bauxite and Bayer at one end, cryolite and Hall-Héroult at the other, and the price of the metal fell by something like two orders of magnitude inside a generation.

Solid electricity

Smelting aluminum takes on the order of thirteen to fifteen kilowatt-hours of electricity per kilogram of metal. That figure dominates everything about the industry's geography.

Aluminum smelters are not built near bauxite mines, which are mostly tropical. They are built wherever electricity is cheap and, critically, continuous — beside hydroelectric dams in Iceland, Quebec, Norway and the Pacific Northwest, or beside dedicated coal plants elsewhere. The pots cannot be switched off casually; if the bath freezes, the cell is wrecked. This makes smelters both enormous baseload customers and, in some grids, paid participants in demand response, since a smelter can shed load briefly in a way a household cannot.

The corollary is recycling. Remelting scrap aluminum needs roughly five per cent of the energy required to win it from ore, because the hard part was never the melting — it was the electrolysis. That ratio is unusually favourable and it is why aluminum recycling was economically self-sustaining long before recycling anything else was.

Alumium, aluminum, aluminium

Humphry Davy, who never succeeded in isolating the metal, named it three times. He tried alumium in 1808, revised it to aluminum in 1812, and was then criticised in a British review for the unclassical ending — -ium being the pattern set by sodium, potassium and magnesium. Aluminium took hold in Britain.

America went the other way, not by nationalism but by accident of reference works and trade usage: Webster's dictionary carried aluminum in 1828, and when Hall began marketing the metal he used that spelling on his handbills. The American Chemical Society formally adopted it in 1925. IUPAC settled on aluminium as the international standard in 1990 and then, in 1993, listed aluminum as an acceptable variant, which is why both forms appear in current chemical literature and why this page is filed under the American one.

The film four nanometres thick

Aluminum is thermodynamically eager to react with oxygen — that eagerness is precisely why it was so hard to isolate — and yet aluminum window frames survive decades of rain. The reason is that the oxide it forms is dense, tightly bonded to the metal beneath, and almost exactly the same volume as the metal it replaced, so it grows to a few nanometres and then stops. Scratch it and it reforms within milliseconds.

Anodising is the deliberate exaggeration of this: an electrolytic process that grows the film to tens of microns, with a porous structure that can be dyed before being sealed. The colour on an anodised camera body or a bicycle component sits inside the oxide, not on top of it.

Aluminum's usefulness in aircraft depends less on the pure metal than on what happens when small amounts of copper, magnesium and zinc are added. Alfred Wilm discovered age hardening in 1906, apparently after leaving a test piece over a weekend and finding it had strengthened by itself in the interim; the alloy became Duralumin, and precipitation hardening remains the basis of the 2000- and 7000-series alloys used in airframes.

The isotope that melted the first planetesimals

Only one isotope of aluminum is stable, but a radioactive one matters a great deal to geochronology. Aluminum-26 has a half-life of about 717,000 years, which is far too short for any primordial supply to have survived.

It was present in the early solar system nonetheless. Calcium-aluminum-rich inclusions in primitive meteorites — the oldest solids known, at around 4.567 billion years — contain excess magnesium-26, the decay product, in proportion to their aluminum content. The inference is that live aluminum-26 was incorporated when they formed, which requires a source of freshly synthesised material shortly beforehand, most plausibly a nearby supernova or an evolved massive star. Its decay also released enough heat to melt bodies tens of kilometres across, which is how asteroids came to differentiate into cores and mantles so early.

Aluminum-26 is produced today in a thin layer at the top of the atmosphere and in exposed rock surfaces struck by cosmic rays. Because it is generated alongside beryllium-10 in a known ratio but decays faster, the changing ratio between the two acts as a clock for how long a rock surface has been shielded from the sky — the standard technique for dating cave sediments, and the method behind several of the contested early dates for hominin remains in South Africa.