Astatine (At)

There is no photograph of astatine. There is no description of its colour, no measured boiling point, no confirmed answer to the question of whether it is a metal. This is not an oversight in the literature; it is a physical impossibility.

Astatine is not quite alone in that. Francium has never been collected in a weighable amount either, and everything from rutherfordium upward exists a few atoms at a time and always will. What separates astatine from that company is the reason for it. For the others the binding constraint is supply — nobody can make enough. Astatine's constraint would survive an unlimited supply, because a visible piece of it would take itself apart before anyone could look.

The obstacle is its own radiation. Every astatine isotope is short-lived and every one is a strong alpha or electron-capture emitter, so a sample large enough to see deposits energy into itself faster than it can shed it. Accumulate a visible speck and it heats, vaporises, and radiolytically tears apart whatever it is sitting on. The largest quantities ever assembled are measured in nanograms, held on surfaces and in solution, and studied by tracking their radiation rather than by looking at them.

Alabamine, dakin, helvetium, dor

Element 85 was one of the last holes in the table, and it attracted claims the way empty squares always did.

Fred Allison at Alabama Polytechnic announced it in 1931 and called it alabamine, on the strength of a technique he had developed called the magneto-optic method — an apparatus in which an observer watched for a faint flash at a particular moment. It produced positive results for almost everything Allison pointed it at, including element 87, and none of them survived scrutiny. Rajendralal De claimed element 85 in monazite in 1937 as dakin. Walter Minder in Switzerland proposed helvetium in 1940 and, with Alice Leigh-Smith, anglo-helvetium in 1942. Horia Hulubei and Yvette Cauchois reported it in radon and called it dor.

Every one of these was mistaken, and the pattern is instructive: all of them were looking for a stable or long-lived element in a natural material, and there is no such thing to find.

Bombarding bismuth

The successful route came from an accelerator. In 1940 Dale Corson, Kenneth MacKenzie and Emilio Segrè bombarded a bismuth-209 target with alpha particles in the Berkeley cyclotron. Two neutrons came off and astatine-211 was left behind. For Segrè this was a second missing square filled — he had co-discovered technetium three years earlier by a similar approach — and it established the pattern that the remaining gaps in the table would be closed by machines rather than by mineralogy.

They named it in 1947 from the Greek astatos, unstable, which is the only element name that amounts to an apology for the element's existence.

Astatine does occur in nature, in trace amounts, as minor branches in the uranium and thorium decay chains. Published estimates of how much exists in the entire crust at any instant range from well under a gram to a few tens of grams, and the spread is not sloppiness: the figure depends on branching ratios that are themselves difficult to measure and on assumptions about crustal uranium inventory.

Astatine is conventionally called the rarest naturally occurring element, and if the low end of that range is right it comfortably is. The high end is harder to defend. Francium — the only serious rival, and the only other element in the same territory — is usually put at around thirty grams by an estimate constructed the same way, from branching ratios and an assumed uranium inventory rather than from anything anyone has weighed. Two model outputs that overlap do not support a confident ranking between them. The defensible statement is narrower and more interesting: astatine and francium are the two scarcest elements in nature by a very long way, and both of them are being manufactured and destroyed continuously rather than sitting anywhere as a deposit.

Is it a metal? Nobody has been able to look

Iodine, one row up, is a molecular solid of I₂ units with a lustrous violet-black sheen. The straightforward extrapolation says astatine should be the same, only darker and denser.

Relativistic calculations disagree. Work published in 2013 modelling condensed astatine concluded that relativistic effects on its 6p electrons destabilise the diatomic molecule enough that solid astatine should not be molecular at all, but a face-centred cubic metal with a metallic band structure. If that is right, astatine is a metal sitting in the halogen column, and the periodic table's diagonal metal-nonmetal boundary runs one square further than usually drawn.

There is no prospect of checking it directly. What has been measured is astatine's electron affinity: a team at CERN's ISOLDE facility obtained 2.41578 eV in 2020, working with a beam of astatine atoms rather than a sample. That was the first fundamental atomic property of the element ever measured rather than computed. It confirms astatine as the least electronegative of the halogens and gives relativistic quantum chemistry a hard number to be tested against.

Its solution chemistry already hints at the answer. Astatine forms the expected astatide anion, but it also forms cationic species — At⁺ and complexes built from it — in acidic aqueous solution. No other halogen does that. Chemically, astatine is halfway out of its own group.

Seven hours, and why that is exactly right

Astatine has one serious application, and the isotope that supports it is the same one Corson, MacKenzie and Segrè made in 1940.

Astatine-211 has a half-life of 7.2 hours and, through either of its two decay branches, emits exactly one alpha particle per atom. An alpha particle travels only about five to eight cell diameters in tissue and deposits its energy densely along that path, producing double-strand DNA breaks that cells repair poorly. Attached to an antibody, a peptide or a small molecule that finds a tumour, it delivers a lethal dose to targeted cells while leaving tissue a few tens of micrometres away essentially untouched — which is the entire premise of targeted alpha therapy.

The half-life is what makes astatine-211 unusually well suited to the job. Long enough to label a molecule, inject it and let it distribute; short enough that almost all of the dose is delivered within a day and the patient does not remain a source for weeks. A convenient side effect is that its polonium-211 daughter emits X-rays, so the same dose can be imaged as it goes in.

Clinical trials are running or completed in glioblastoma, in ovarian cancer treated within the peritoneal cavity, in conditioning before stem cell transplant for leukaemia, and in thyroid disease. Astatine follows iodine into the thyroid, though less avidly, which is both a therapeutic opening and the main route by which it would be taken up if released.

The constraint is production. Astatine-211 needs alpha particles of around 28 MeV, which restricts manufacture to a few dozen cyclotrons worldwide, and it cannot be stockpiled or shipped far. Every dose has to be made, labelled and used within the same working day, near a machine that can make it.

Almost everything above is an inference

It is worth being explicit about what the reference data for this element actually is. The melting point quoted for astatine is an extrapolation down the halogen group, not a measurement. There is no boiling point figure at all. The density is calculated. The half-lives, the decay energies and the electron affinity are real measurements; nearly everything else is theory doing its best with an element that will not sit still long enough to be weighed.