Cobalt (Co)

Cobalt is the rare element whose name most people first encounter as a colour. That colour is genuinely ancient — Egyptian and Mesopotamian glassmakers were producing deep blue with cobalt ores by the middle of the second millennium BCE, and none of them had any conception of a metal being responsible. A fraction of a per cent of cobalt in a glass melt is enough to saturate it, which is why the pigment spread so widely and so early: a very small quantity of an unremarkable-looking mineral transformed an entire batch.

The blue that dates a porcelain

Chinese blue-and-white porcelain, made from the Yuan dynasty onwards, is painted with cobalt oxide under the glaze and fired once. The finest early wares used cobalt imported from Persia, known in Chinese sources as Muhammadan blue, which is high in iron and low in manganese and fires to an intense violet-tinged blue that tends to pool darkly where the brush lingered.

Domestic Chinese cobalt ores are the reverse — manganese-rich — and produce a softer, greyer, somewhat muted blue. When the supply of imported cobalt was interrupted, potters had to work with the local material, and the change is visible in the finished pieces. Specialists routinely use the character of the blue as one of the criteria for assigning a piece to a period, because the pigment records a trade route.

The European counterpart came much later and by government commission. Ultramarine, ground from lapis lazuli, was ruinously expensive, and in 1802 Louis Jacques Thénard produced cobalt aluminate — a stable, brilliant, comparatively cheap blue that took his name and quickly became a painter's standard.

Not bismuth after all

The ore that yielded this pigment in the German mining districts had an evil reputation. It looked like a valuable ore, produced no useful metal, and made smelter workers sick, because the minerals involved are cobalt arsenides and roasting them releases arsenic oxide. Miners blamed the kobold, the goblin of the workings, and the name stuck to the substance.

The prevailing assumption among chemists was that the blue colour came from bismuth. Georg Brandt, working in Stockholm, disproved that in 1735 and isolated a new semi-metal responsible for it. Cobalt is consequently the first metal unknown to antiquity to be discovered and identified by a named individual — a distinction it is often not given, partly because Brandt published in Latin in a Swedish journal and partly because it took decades for the finding to be generally accepted.

The only organometallic bond in human biochemistry

Every atom of cobalt in your body is in vitamin B12, and every molecule of B12 has a cobalt ion at the centre of a corrin ring. It is the only place cobalt appears in human biology, and it does something no other biological cofactor does: in the active forms of the vitamin, the cobalt is bonded directly to a carbon atom. That is the sole naturally occurring metal-carbon bond in human metabolism.

The vitamin was found by working backwards from a disease. Pernicious anaemia was reliably fatal until George Minot and William Murphy showed in 1926 that feeding patients large quantities of raw liver reversed it; they shared a Nobel prize for a treatment whose active ingredient nobody could name. B12 was finally isolated in 1948, and Dorothy Hodgkin determined its structure by X-ray crystallography in 1956 — a molecule of around a hundred atoms, far larger than anything previously solved, worked out with punched-card computing and awarded the 1964 chemistry Nobel. Robert Burns Woodward's total synthesis took roughly a hundred collaborators eleven years.

Neither plants nor animals can make B12. Only bacteria and archaea can. Ruminants obtain theirs from the microbial population in the rumen; humans obtain it from animal products, or from supplements, which is why it is the one nutrient a purely plant-based diet cannot supply from food alone.

Cobalt-60, in a clinic and in a thought experiment

Cobalt-60 has a half-life of about 5.27 years and emits two gamma rays of high and precisely known energy on each decay. That combination — long enough to be practical, energetic enough to penetrate, simple enough to calculate — made it the workhorse radiation source of the twentieth century.

Harold Johns in Saskatoon put the first cobalt-60 teletherapy unit into clinical use in 1951, and for decades afterwards cancer radiotherapy in most of the world meant a cobalt unit. Linear accelerators have replaced them in wealthy countries; cobalt units remain in service elsewhere precisely because they need no reliable electricity supply. The same isotope sterilises single-use medical equipment, irradiates spices and inspects welds.

The phrase "cobalt bomb" is unfortunately ambiguous, and both meanings are real. One is the therapy machine. The other is Leo Szilard's 1950 radio proposal for a weapon jacketed in cobalt, designed so that neutron activation would spread cobalt-60 fallout of exactly the wrong half-life — long enough to persist for years, short enough to be intensely radioactive. Szilard described it in order to argue that a genuinely world-ending device was technically possible, not to advocate building one. None has ever been built.

Cobalt-60 also settled a question in fundamental physics. In 1956 Chien-Shiung Wu cooled cobalt-60 to a hundredth of a kelvin and aligned the nuclear spins in a magnetic field, then measured the direction in which the beta particles came out. They came out preferentially in one direction relative to the spin, which meant the weak interaction distinguishes left from right — parity is not conserved. Tsung-Dao Lee and Chen Ning Yang, who had proposed the test, received the Nobel prize the following year. Wu did not.

Magnetic at temperatures that demagnetise everything else

Cobalt's Curie temperature — the point above which it loses its ferromagnetism — is about 1115 °C. Iron's is around 770 °C and nickel's around 354 °C. Cobalt holds its magnetic ordering hundreds of degrees beyond either.

That single property is why cobalt appears in every permanent magnet expected to work hot. Alnico alloys, samarium-cobalt magnets in aerospace and military hardware, and the cobalt added to neodymium magnets to raise their operating temperature all exist for this reason. It is also the basis of the superalloys used in the hot sections of jet engines and gas turbines.

A by-product that cannot respond to its own price

Around seventy per cent of mined cobalt comes from the Democratic Republic of the Congo, and almost all cobalt anywhere is recovered as a by-product — of copper in the Congolese Copperbelt, of nickel in Canada, Russia and Australia.

That structure is the central fact of the cobalt market. A copper mine's output is decided by the copper price; no operator expands a copper mine because cobalt has become expensive. Supply is therefore close to unresponsive to cobalt demand in the short run, which is why cobalt prices swing so violently when battery demand moves.

The remainder comes from artisanal digging, estimated at somewhere between a tenth and a third of Congolese production, worked by hand and documented by Amnesty International in 2016 as including substantial child labour. That report is a large part of why battery manufacturers have pushed so hard toward low-cobalt and cobalt-free cathode chemistries — a commercial motive and a reputational one arriving at the same answer.

Cobalt's occupational hazards are well characterised. Inhaled cobalt-tungsten carbide dust causes hard metal lung disease, a fibrosis specific to the combination rather than to either component alone, and metal-on-metal hip implants that shed cobalt debris have produced systemic cobalt toxicity in a documented series of patients.