Chlorine (Cl)

In September 1908 John L. Leal began adding chloride of lime to the water supplied to Jersey City, New Jersey — around two hundred thousand people — without telling the public, without a permit, and in the middle of litigation over whether the supply was clean enough to satisfy his employer's contract. He had three months and no precedent at municipal scale.

Typhoid rates in the city collapsed. When the case returned to court, the judge accepted the evidence and ruled the water pure and wholesome. Within a decade chlorination was standard across American cities, and the pattern repeated wherever it was adopted. Economists David Cutler and Grant Miller later attributed roughly half of the total mortality decline in major US cities between 1900 and 1936 to clean water technologies, with an outsized effect on infant mortality. Few interventions in the history of medicine can show a comparable figure.

Thirty-six years mistaken for a compound

Carl Wilhelm Scheele made the gas in 1774, noted its suffocating smell and its ability to bleach plant material, and described it as dephlogisticated marine acid air. The name reflects the theory he was working in, not what he had.

The next generation did no better, for a subtler reason. Lavoisier's chemistry held that oxygen was the essential component of every acid, so a gas obtained from muriatic acid was assumed to contain oxygen, and Claude Louis Berthollet named it oxymuriatic acid on that basis. It was regarded as a compound of oxygen with some unidentified radical for the better part of four decades.

Humphry Davy tried repeatedly to strip the oxygen out and could not. In 1810 he made the argument that no oxygen was there to remove, that the substance was elementary, and named it chlorine after chloros, the yellow-green of new foliage. The name records nothing but its colour, which was the only property he could be sure of. Chlorine is thus the element whose stubbornness demolished the theory that acids are defined by oxygen.

Langemarck, 22 April 1915

Late in the afternoon, along roughly six kilometres of the Ypres salient, German troops opened the valves on several thousand steel cylinders and released something on the order of 150 tonnes of chlorine into a favourable wind. It drifted across French colonial and territorial positions as a visible yellow-green cloud. Being denser than air, it settled into the trenches, which were the one place a soldier could not leave.

The operation was directed by Fritz Haber, in uniform, in person. Ten days later his wife Clara Immerwahr — the first woman to take a chemistry doctorate at Breslau — shot herself with his service pistol. The connection between the two events is widely asserted and genuinely uncertain: some historians treat her death as a protest against the gas programme, others point to years of documented frustration at a scientific career subordinated to his. The evidence does not settle it.

Chlorine was quickly superseded as a weapon by phosgene and mustard agent, both more effective and neither so easily seen. Its brief primacy is the reason it is remembered.

Two products, one reaction, and no way to want them separately

Chlorine is manufactured by electrolysing brine, and the reaction produces chlorine at one electrode and sodium hydroxide at the other, in a fixed molar ratio, along with hydrogen. There is no operating adjustment that yields more of one and less of the other.

This creates an economic structure found in very few industries. Demand for chlorine — driven largely by plastics and construction — and demand for caustic soda — driven by alumina refining, pulp and paper, and soaps — move independently, and the market is permanently out of balance in one direction or the other. Producers speak of being chlorine-driven or caustic-driven depending on which product is currently paying, and the co-product's price adjusts until someone will take it away. Occasionally that price approaches zero.

The technology has also had to change for reasons unconnected to chemistry. Mercury-cell plants dominated for most of the twentieth century and leaked mercury persistently; the European industry completed its phase-out in 2017 under the framework that became the Minamata Convention, replacing them with membrane cells that are both cleaner and substantially more energy-efficient.

Where the chlorine atoms end up

The largest single destination is polyvinyl chloride. PVC is around 57% chlorine by mass, which is the underappreciated reason it is so cheap: more than half the material is derived from salt rather than from oil, and salt does not track the crude price. Pipes, window frames, cable insulation and flooring absorb an enormous tonnage.

Much of the rest never appears in the final product at all. Chlorine is an exceptionally useful intermediate — it activates molecules for further reaction and is then removed — so a large fraction of pharmaceutical and agrochemical syntheses pass through a chlorinated step even when the finished compound contains no chlorine. Titanium dioxide pigment is made by a chloride route that recycles its chlorine continuously.

The number that killed Prout's hypothesis

In 1815 William Prout noticed that the known atomic weights were close to whole-number multiples of hydrogen's and proposed that hydrogen was the fundamental building block of all matter. It was an inspired guess a century ahead of its evidence.

Chlorine ruined it. Measured with increasing care through the nineteenth century, its atomic weight sat obstinately near 35.45 — not close to any integer, and precisely determined enough that experimental error could not be blamed. Jean Servais Stas's meticulous measurements were taken as decisive against Prout.

The resolution, when it came from Frederick Soddy's concept of isotopes and Francis Aston's mass spectrograph around 1920, vindicated Prout in a form he could not have imagined. There is no atom of mass 35.45. There are two chlorine isotopes, at mass numbers 35 and 37, both very nearly integers, mixed in a ratio of about three to one. The weight in the table is a population average, and the element that appeared to refute the whole-number rule turned out to be its best illustration.

Dating water that fell before the last ice age

Chlorine-36 has a half-life of about 301,000 years, produced in the atmosphere by cosmic rays and in rock by neutron capture. That timescale is inconveniently long for archaeology and exactly right for hydrology.

Groundwater in the Great Artesian Basin of Australia and in the Nubian aquifer beneath the Sahara has been dated by its declining chlorine-36 content to hundreds of thousands of years, which establishes that these are fossil waters recharged under past climates rather than resources being replenished today. A second, artificial pulse is also in circulation: atmospheric nuclear tests over the Pacific in the 1950s irradiated seawater chloride and roughly doubled global chlorine-36 deposition for a few years, leaving a dated marker layer that hydrologists use to identify water that entered the ground after 1952.

Not a purely synthetic idea

Organochlorine compounds have a reputation as industrial artefacts, and several of the most persistent pollutants are. The blanket assumption is nevertheless wrong: several thousand naturally occurring organohalogen compounds have been catalogued, produced by marine algae, fungi, lichens and bacteria, and by forest fires and volcanoes. The antibiotic vancomycin is a chlorinated natural product, and so is the epibatidine found in the skin of an Ecuadorian frog.