Arsenic (As)

Arsenic never becomes a liquid unless you force it. Heated at ordinary pressure it goes straight from grey brittle solid to vapour at around 614 °C, and only under roughly 40 atmospheres will it consent to melt — which is why the melting point listed above is higher than the boiling point, a reversal that looks like a typing error and is not. The number quoted as a melting point is a pressurised measurement; the one quoted as a boiling point is really a sublimation temperature.

That behaviour is a consequence of how the grey allotrope is built. Its atoms form puckered layers held together sideways by strong covalent bonds and stacked with only weak forces between sheets. Heat separates the sheets long before it disrupts the bonding inside them, so whole molecular fragments leave the surface while the lattice below is still intact.

Yellow stone, and a word that changed meaning

The name arrives through Greek arsenikon, which came from Persian zarnikh, the yellow mineral orpiment — arsenic trisulfide, a pigment traded across the ancient world. Greek speakers heard arsenikon as related to arsenikos, "masculine" or "potent", and that folk etymology stuck to the word long enough that early metallurgists thought of arsenic compounds as virile substances that strengthened other metals.

The compounds were known for millennia; the free element was not. Its isolation is conventionally given to Albertus Magnus in about 1250, on the strength of a description of heating orpiment with soap, though the account is compressed and several Arabic alchemical texts describe something similar earlier.

Inheritance powder, and the test that ruined the trade

White arsenic — the trioxide — is the compound that gave the element its reputation. It has almost no taste, dissolves in hot liquid, and produces symptoms that a nineteenth-century physician read as cholera or gastric fever. In seventeenth-century Italy it was sold as Aqua Tofana, and in France it acquired the name poudre de succession, inheritance powder. The practical problem for prosecutors was that nobody could prove it had been used.

James Marsh changed that. He had testified in an 1832 murder trial where his precipitate test degraded before the jury saw it and the accused was acquitted; by 1836 he had built an apparatus that converted any arsenic present into arsine gas, decomposed it on a heated surface, and deposited a mirror-bright metallic film that could be weighed and shown in court. The Marsh test was sensitive to fractions of a milligram, and it turned arsenic from an undetectable poison into a detectable one almost overnight. Its most famous outing was the 1840 trial of Marie Lafarge in France, where Mathieu Orfila's testimony over the results effectively founded forensic toxicology as a discipline.

A green people lived inside

Carl Wilhelm Scheele produced a vivid copper arsenite pigment in 1775, and a brighter variant, copper acetoarsenite, followed as Paris green or emerald green. Victorian consumers loved it. It went into wallpaper, artificial flowers, ball gowns, children's toys and confectionery decoration, and it went into them by the tonne at a time when nothing regulated what a colour was made of.

Damp rooms made it worse: moulds growing on arsenical wallpaper metabolised the pigment into volatile trimethylarsine, filling bedrooms with a garlic-smelling vapour. The chronic illnesses associated with green rooms were real, and the theory that the wallpaper at Longwood House contributed to Napoleon's decline on Saint Helena rests on genuine arsenic in genuine samples, though not on any way of separating that from the arsenic in everything else he encountered. Paris green's second career was as an insecticide, and it stayed in orchards well into the twentieth century.

The largest mass poisoning of a population

The worst arsenic exposure in history was an accident of public health policy. From the 1970s, aid agencies drilled millions of tube wells across Bangladesh and West Bengal to move villagers off bacterially contaminated surface water, and the programme succeeded — infant mortality from waterborne disease fell sharply. Nobody tested the groundwater for arsenic, because there was no reason to expect it.

The aquifer sediments of the Ganges–Brahmaputra delta carry arsenic bound to iron oxide grains, and in the reducing conditions at depth that arsenic is released into the water. Tens of millions of people drank it for two decades before the pattern of skin lesions and cancers was traced to its cause in the 1990s. The World Health Organization has described the episode as the largest mass poisoning of a population in history. Mitigation now turns on well-by-well testing, painting spouts red or green, and deeper or shallower drilling — unglamorous work at enormous scale.

The poison that became a medicine, twice

Paul Ehrlich's laboratory screened hundreds of organoarsenic compounds against syphilis and, in 1909, arrived at compound 606, arsphenamine, marketed as Salvarsan. It was the first chemical designed to kill a pathogen inside a patient without killing the patient, and it established the whole idea of chemotherapy decades before the word meant cancer treatment.

Arsenic's second medical career is stranger. Arsenic trioxide, the classic poison, induces remission in acute promyelocytic leukaemia by degrading the fusion protein that drives the disease. Traditional Chinese medicine practitioners in Harbin were using arsenic preparations against leukaemia in the 1970s; Western trials followed, and the compound was approved in the United States in 2000. Combined with a vitamin A derivative it has turned one of the most rapidly fatal leukaemias into one of the most curable, in many cases without chemotherapy at all.

Where the tonnage actually goes

For most of the late twentieth century the dominant use of arsenic was chromated copper arsenate, a timber preservative that made softwood decking and fence posts resistant to rot and insects. Residential use was withdrawn in the United States from 2003 and restricted across Europe, and demand fell steeply.

What remains splits between two very different markets. Small percentages of arsenic harden lead alloys, which is why it appears in lead-acid battery grids and in shot. And high-purity arsenic goes into gallium arsenide and indium gallium arsenide semiconductors — the radio-frequency chips in phones and the detectors in fibre-optic receivers. Most arsenic reaching the market is recovered from copper smelter flue dust rather than mined deliberately, so it is a byproduct of demand for something else entirely.

One isotope, and one retracted claim

Arsenic is mononuclidic: every arsenic atom in nature is arsenic-75. That is why its atomic weight is quoted to eight significant figures while elements with several isotopes carry ranges and uncertainties — there is no isotopic mixture to vary between samples.

Arsenic sits directly below phosphorus, which prompted the most publicised biology paper of 2010: a claim that a bacterium from Mono Lake, GFAJ-1, could substitute arsenic for phosphorus in its DNA. Replication attempts in 2012 found the organism was merely arsenic-tolerant and was scavenging trace phosphate. Science formally retracted the paper in July 2025, fifteen years after publication and over the objections of its authors.