Technetium (Tc)

Between molybdenum and ruthenium there is a hole in the periodic table that nature genuinely left empty. Every other element up to bismuth exists in the ground somewhere. Element 43 does not — not because it is rare, but because no arrangement of 43 protons and any number of neutrons produces a nucleus that lasts. Chemists hunted for it in ores for sixty years and kept announcing it, and every one of those announcements was wrong. It had to be manufactured before anyone could hold it.

Why 43 protons cannot settle down

The absence is not bad luck. It follows from a regularity in nuclear stability first stated by Josef Mattauch: two isobars — nuclei with the same mass number — cannot both be stable if they sit on adjacent squares of the table. For every mass number technetium could plausibly take, one of its neighbours, molybdenum or ruthenium, already occupies the stable slot for that mass. Technetium is squeezed out at every single value.

What is left is a set of long-lived but doomed isotopes. The most durable, technetium-97 and technetium-98, survive on the order of four million years. That sounds enormous until you set it against 4.5 billion years of Earth history, which is roughly a thousand times longer. Any technetium present when the planet formed decayed away before the first cells appeared. Promethium, element 61, is excluded from the crust by the same argument, and those two are the only gaps below uranium.

Masurium, and a claim nobody could reproduce

Mendeleev left the square empty and predicted its occupant as ekamanganese, expecting something that would behave like the element directly above it. Filling it became a minor obsession. Japanese chemist Masataka Ogawa announced "nipponium" in 1908; the sample turned out to be rhenium, an honourable error, since rhenium was also genuinely undiscovered at the time.

The claim that mattered came in 1925, when Walter Noddack, Ida Tacke and Otto Berg reported two new elements from X-ray analysis of columbite and platinum ores. One was rhenium and it was correct. The other they called masurium, after Masuria in East Prussia. No other laboratory could reproduce it, and the trio never withdrew the claim, which cost them credibility they did not deserve to lose on the rhenium work. The story has an awkward coda: in 1999 a Los Alamos group led by David Curtis re-examined the 1925 method and argued the Noddack team's signal was not obviously impossible, since uranium minerals do contain minute quantities of technetium from spontaneous fission. Whether they saw something real remains genuinely unsettled, and the element's name reflects the outcome rather than the argument.

A strip of foil posted from Berkeley

The successful route was not mineralogical. Ernest Lawrence's cyclotron at Berkeley used a molybdenum deflector foil that had spent months being hit by deuterons, and Emilio Segrè asked for the discarded part. In 1937 in Palermo, Segrè and the mineralogist Carlo Perrier separated from that foil an activity whose chemistry sat between manganese and rhenium exactly where it should. The data card above credits Segrè alone, which is the conventional shorthand; Perrier did much of the radiochemical separation and the announcement carried both names.

They deferred naming it for over a decade. When they did, in 1947, they chose technetium, from Greek tekhnetos, "artificial" — a deliberate statement that this was the first element made by people rather than found. That claim is narrowly true and often overstated: several elements beyond uranium were synthesised soon afterwards, but technetium is the first synthetic element that occupies a square below uranium, a gap in ordinary chemistry rather than an extension past its end.

The lines in a red giant that ended an argument

In 1952 Paul Merrill, working at Mount Wilson, found absorption lines of technetium in the spectra of S-type red giant stars. Because the element cannot survive four million years, the technetium in those stars could not have been there since the star formed. It was being built inside the star, in real time, by slow neutron capture.

That single observation converted stellar nucleosynthesis from a plausible theory into something directly witnessed. Astronomers still use technetium lines as a marker for a star currently dredging up freshly made material from its interior — an element defined by its instability turned into a clock for stellar evolution.

Forty thousand scans a day

Almost all of the technetium ever produced has been used inside people. The reason is a metastable nuclear state, technetium-99m, which sheds its excess energy as a single gamma photon of 140 keV and drops to technetium-99 with a half-life of six hours.

Those numbers are close to ideal for imaging. The photon energy passes through tissue but is efficiently caught by a sodium iodide detector; six hours is long enough to make a scan and short enough that essentially all of it has decayed within a couple of days. Technetium's chemistry is unusually accommodating too — it will sit in dozens of different molecules, so the same isotope can be attached to a bone-seeking phosphonate, a red blood cell, or a compound that lodges in heart muscle. One isotope covers cardiac perfusion, bone metastasis, kidney function, thyroid, lung ventilation and sentinel lymph node mapping. Estimates of its share of nuclear medicine range from about 75% to 85% of all diagnostic procedures, tens of millions of scans a year, some 40,000 a day in the United States alone.

A supply chain measured in hours

An isotope with a six-hour half-life cannot be stockpiled, which gives technetium one of the strangest logistics problems in medicine. Hospitals do not buy technetium; they buy its parent. Molybdenum-99, half-life 66 hours, is loaded onto an alumina column, and as it decays the technetium-99m that forms is washed off with saline. The device is universally called a generator and, less formally, a technetium cow: you milk it each morning and it refills itself overnight.

Molybdenum-99 itself comes from a handful of ageing research reactors — Petten in the Netherlands, and a small number of others in Belgium, South Africa, Australia and elsewhere. When one goes down for repair, imaging departments across whole continents reschedule. Shortages in 2009–2010 and again in the 2020s pushed producers off weapons-grade uranium targets and toward low-enriched alternatives and accelerator routes, driven as much by non-proliferation policy as by supply security.

What happens to the rest of it

Technetium-99 is also one of the more troublesome components of spent nuclear fuel. Fission of uranium-235 yields it in high proportion, its half-life is 211,000 years, and in oxidising conditions it exists as the pertechnetate ion, which is soluble and does not stick to rock. Most fission products are immobilised by the geology around them; this one travels with the groundwater, which is why it features so heavily in long-term repository modelling.

The same pertechnetate ion is a remarkably effective corrosion inhibitor for steel, at concentrations of a few parts per million. It is a real chemical property with essentially no application, confined to closed-loop systems where the radioactivity is contained anyway — an element whose best non-medical trick is one almost nobody can use.