Promethium (Pm)

Promethium's defining property is an absence. It is one of only two elements lighter than bismuth with no stable isotope at all — technetium is the other — and unlike technetium it sits in the middle of a row where every one of the other fourteen members has been sitting in the ground since the Earth cooled. Not all fourteen are strictly stable: lanthanum, neodymium, samarium, europium, gadolinium and lutetium each carry a primordial isotope that decays, on timescales running from tens of billions of years upward. But a nuclide that takes longer than the age of the universe to halve is, for the purpose of finding some, indistinguishable from one that never decays. The line that separates promethium from its neighbours is not stability; it is whether any original atoms are left, and promethium is the only square in the row where the answer is none. Neodymium on one side, samarium on the other, and a hole between them.

Why the gap is there

The explanation is a rule about isobars: nuclei that share a mass number but differ in proton count. For an odd-numbered element, there is at most one stable isobar at any given mass, and if a neighbour has already claimed it, the odd element has nothing left.

Promethium's atomic number, 61, is odd, and it is wedged between two even-numbered elements that are unusually well supplied with stable isotopes. Neodymium at 60 and samarium at 62 between them occupy every mass number promethium could plausibly use. At each mass, promethium finds a lower-energy configuration one square to either side and decays into it. There is no mass number where a promethium nucleus is the most stable option, so there is no stable promethium.

The longest-lived isotope, promethium-145, has a half-life of 17.7 years. Nothing that brief could still be here from the birth of the solar system, so the element is genuinely absent from the Earth as inherited. What little exists is made continuously: spontaneous fission of uranium-238 produces promethium in uranium ores at a level estimated to amount to a few hundred grams in the entire crust at any moment.

Curiously, it has also been detected in a star. Spectral lines of promethium appear in HR 465, one of a class of chemically peculiar stars, where something must be producing it faster than it decays.

Illinium, florentium, cyclonium

An empty square in the middle of the rare earths, in an era when separating rare earths was the hardest analytical work in chemistry, was irresistible — and unfalsifiable enough that several groups convinced themselves they had it.

• Florentium. Luigi Rolla and Lorenzo Fernandes in Florence believed they had found element 61 in monazite in 1924. They deposited a sealed note with the Accademia dei Lincei to establish priority and published two years later.
• Illinium. B. Smith Hopkins and colleagues at the University of Illinois announced element 61 in rare-earth residues in 1926, and the two claims immediately collided.
• Cyclonium. In the early 1940s a group at Ohio State bombarded neodymium and praseodymium in a cyclotron, produced activities they attributed to element 61, and proposed a name referring to the machine that made them.

The first two were looking for a natural element that does not exist in any recoverable quantity, and neither result was reproducible. The third group had genuinely made promethium atoms; what they had not done was isolate them chemically and prove what they were.

Who actually had it

Conclusive proof came out of wartime uranium fission work. In 1945 at Clinton Laboratories in Tennessee — later Oak Ridge — Jacob Marinsky, Lawrence Glendenin and Charles Coryell separated promethium-147 and promethium-149 from the mixture of fission products and identified them using ion-exchange chromatography, a technique then new and the first that could reliably pull the rare earths apart from one another. Secrecy delayed publication until 1947.

The data card above credits Chien-Shiung Wu and the year 1945, reflecting earlier accelerator work in which element 61 was produced without being chemically isolated. That is a defensible position and it is not the standard one: most reference works give the credit to Marinsky, Glendenin and Coryell. The disagreement is really about what counts — making atoms of an element, or demonstrating beyond argument that you have them in a bottle.

The name came from outside the laboratory. Grace Mary Coryell, wife of one of the three, suggested Prometheus, who stole fire from the gods and was punished at length for it. The element that came out of a reactor's fission products, in a programme born of the atomic bomb, was named for the theft of fire. Coryell said afterwards that the name was meant to carry both the daring and the penalty.

What a 2.6-year half-life is good for

Promethium-147 is the isotope that has ever been useful, and its properties are unusual in a helpful way: it emits soft beta particles with essentially no penetrating gamma radiation, and its 2.62-year half-life is long enough to build a product around.

• Thickness gauges. A beta source on one side of a moving web of paper, plastic film or sheet metal and a detector on the other measures thickness continuously and without contact, because the fraction of betas transmitted depends on how much material is in the way. Promethium-147 gauges were standard in paper mills for decades.
• Luminous paint. When radium was abandoned for watch and instrument dials, promethium was one of the replacements: it excites a phosphor without emitting penetrating radiation, so a dial does not irradiate its wearer. Tritium later displaced it.
• Betavoltaic cells. Promethium batteries — beta particles generating current in a semiconductor — powered some early cardiac pacemakers in the 1970s and were studied for spacecraft. Lithium chemistry made them unnecessary for medicine.

Promethium salts glow with a pale blue-green light of their own, produced by the beta radiation exciting the surrounding air and the compound itself.

Nobody had measured its chemistry until 2024

Because promethium has to be made in a reactor, separated from an intensely radioactive fission mixture, and used before it decays, its basic chemistry went unmeasured for nearly eighty years. Its properties in reference tables were interpolated from its neighbours on the assumption that it must behave the way a lanthanide between neodymium and samarium ought to.

In 2024 a team at Oak Ridge finally tested that assumption. They prepared a promethium complex with an organic chelating ligand in aqueous solution and used X-ray absorption spectroscopy to measure the length of the bond between the promethium ion and the oxygen atoms around it. It came out at about 2.09 ångströms — placing promethium exactly where the lanthanide contraction says it should sit, and confirming by measurement what had for three generations been an educated guess.

The work also matters practically. Separating the rare earths from one another depends on the small, systematic differences in ionic radius across the series, and promethium was the one point on that curve nobody had ever pinned down experimentally.