Thorium (Th)

Berzelius named thorium twice. In 1815 he announced a new element in a mineral from Falun and called it thorium after the Norse god; a few years later he established that his sample was yttrium phosphate and quietly withdrew the claim. The name sat unused until 1828, when a Norwegian mineralogist sent him a heavy black stone found on Løvøya in the Langesundsfjord. That one really did contain a new element, Berzelius reissued the name he had already spent, and the mineral became thorite.

For seventy years thorium was a curiosity in a cabinet. Then it became a lamp.

The mantle that built a rare-earth industry

Carl Auer von Welsbach spent the 1880s trying to make gas lighting compete with the electric lamps that were about to destroy it. His answer, perfected in 1891, was a knitted cotton mesh soaked in a solution of metal nitrates and burned off, leaving a fragile ceramic skeleton of 99% thorium dioxide and 1% cerium dioxide. Held in a gas flame, it glows brilliantly white.

The mixture is not arbitrary. Thoria is a poor thermal radiator across the infrared, so it does not waste the flame's energy on heat it cannot use, while the small cerium fraction emits strongly in the visible. The Welsbach mantle produced far more light per unit of gas than any open flame, and it sold in the billions — it is still the standard for camping lanterns, though now usually made with yttrium instead.

To supply it, companies went looking for monazite sand: beach placers in Brazil, in Travancore in southern India, and in North Carolina. Monazite is a phosphate of cerium, lanthanum, neodymium and thorium, and the whole extraction industry existed to get the thorium out. The lanthanides were the leftovers — sold cheaply, dumped, or turned into lighter flints because nobody had a better idea.

Byproduct, then liability

The relationship has since inverted completely. The rare earths are now the product, and thorium is the reason rare-earth refining is difficult to site anywhere.

Monazite still carries its thorium, and separating the lanthanides concentrates it into the residue. That residue is legally radioactive waste in most jurisdictions, and disposing of it is a permitting problem that has repeatedly outweighed the chemistry. It is the core of the long-running dispute over the residue storage at the Lynas refinery at Gebeng in Malaysia, and it is one of several reasons Western rare-earth projects tend to prefer bastnäsite, which carries much less thorium, even when monazite deposits are richer.

The element that once paid for the industry is now the thing that makes the industry hard.

Fertile is not the same as fissile

Thorium's reputation as a nuclear fuel needs one distinction stated plainly, because most popular accounts skip it. Thorium-232 does not fission usefully. It is fertile: absorb a neutron and it becomes thorium-233, which beta-decays through protactinium-233 to uranium-233, and uranium-233 is an excellent fissile material. A thorium reactor is therefore a uranium-233 reactor with a thorium feedstock, and it needs something else — enriched uranium or plutonium — to start.

The advantages are real: thorium is three to four times more abundant in the crust than uranium, all of it is the useful isotope so there is no enrichment step for the feed, and the fuel cycle produces far less long-lived transuranic waste. The disadvantages are equally real. Breeding uranium-233 also makes traces of uranium-232, whose decay chain includes thallium-208 and a 2.6 MeV gamma ray hard enough to require heavy shielding. That is often described as a proliferation safeguard, and it is; it is also the reason fabricating recycled thorium fuel is expensive.

Seventy years of almost

Oak Ridge's Molten-Salt Reactor Experiment ran from 1965 to 1969 and in 1968 became the first reactor anywhere fuelled with uranium-233. It demonstrated that a liquid fluoride fuel salt works; it never bred thorium in anger, and the programme was cancelled in favour of liquid-metal fast breeders.

India built an entire national strategy around thorium, because it has one of the world's largest monazite reserves and, for decades, restricted access to uranium. Homi Bhabha's three-stage plan — heavy water reactors, then fast breeders, then thorium — has completed stage one and is still working through stage two, sixty years on.

The only reactor currently demonstrating continuous thorium-to-uranium conversion under irradiation is Chinese. The TMSR-LF1, a 2-megawatt thermal molten-salt unit in the Gobi desert near Wuwei in Gansu, reached first criticality in October 2023 and its operators reported successful thorium-uranium conversion, followed by online refuelling of the running reactor. It is a research device with a thorium inventory measured in tens of kilograms, not a power plant. It is also the first physical evidence in half a century that the concept works outside a paper study.

The rock clock and the star clock

Thorium-232's half-life is about 14 billion years, comfortably longer than the age of the universe, which is why essentially all the thorium that ever existed on Earth is still here and why thorium supplies a large share of the radiogenic heat driving mantle convection.

Its shorter-lived relative does the precision work. Thorium-230 — long enough ago called ionium to have earned its own name before anyone realised it was thorium — grows into any material that incorporates uranium but excludes thorium when it forms. Coral skeletons and cave carbonates do exactly that, so measuring the thorium-230 that has accumulated dates them. This is the backbone of the last half-million years of sea-level history and of the speleothem records that anchor palaeoclimate chronology.

Astronomers run the same idea on stars. The thorium and uranium lines in extremely metal-poor halo stars give an independent age for the material those stars formed from, providing a lower bound on the age of the galaxy that does not depend on stellar models.

A clock inside a nucleus

Every atomic clock ever built works on a transition between electron energy levels. Nuclear transitions are typically thousands to millions of times more energetic — well out of reach of any laser — which is a pity, because a nucleus is shielded by its own electron cloud and would make an exceptionally stable oscillator.

Thorium-229 is the exception. It has an excited nuclear state just 8.4 electronvolts above its ground state, the lowest known anywhere in the nuclear chart, and that lands in the vacuum ultraviolet where a laser can reach it. Finding the precise frequency took decades. In 2024 a group at TU Wien, using a 148-nanometre laser source developed at PTB, drove the transition in thorium-229 embedded in a calcium fluoride crystal, and a JILA and NIST team measured its frequency against an optical frequency comb.

A working nuclear clock would be less perturbed by stray fields than today's best optical clocks, and would be extraordinarily sensitive to any drift in the fine-structure constant or in the strength of the strong interaction — because nuclear and atomic energies depend on those constants in different ways. Thorium is the only element that offers the experiment.

Where you have already met it

Thorium's older consumer uses lingered long enough to be collectable. Thoriated tungsten TIG welding electrodes, typically 2% thoria, are still sold though lanthanum and cerium versions have largely displaced them. Optical glass doped with thorium dioxide has a high refractive index and low dispersion, and camera lenses from the 1940s to the 1970s made liberal use of it — the Aero-Ektars and several Takumars among them. Those lenses turn visibly yellow-brown over the decades, as the glass accumulates colour centres from its own radiation.