Actinides

Until the middle of the Second World War there was no actinide series. Thorium sat beneath hafnium, protactinium beneath tantalum, uranium beneath tungsten, and the arrangement was defensible: uranium forms a volatile hexafluoride and a trioxide much as tungsten does, and its highest oxidation state is +6, exactly where the column predicts. Mendeleev had placed them that way and nothing in ordinary laboratory chemistry contradicted it for seventy years.

The arrangement collapsed as soon as anyone tried to work with elements beyond uranium. Neptunium and plutonium were not behaving like heavier rhenium and osmium. They were behaving like something with no obvious precedent, and the failure to predict their chemistry was holding up a programme that had a war attached to it.

The hypothesis he was advised not to publish

In 1944 Glenn Seaborg proposed that the whole placement was wrong: that a second f-block series begins in the region of actinium, parallel to the lanthanides, and that the transuranium elements are therefore homologues of the rare earths rather than of the third transition row. Colleagues warned him that publishing it would damage his standing, because it required moving four well-studied elements out of columns where their chemistry appeared to fit.

The prediction it generated was specific and immediately testable. If elements 95 and 96 were the analogues of europium and gadolinium, they should be stubbornly tripositive and should refuse the oxidation that had been used, unsuccessfully, to separate them from the surrounding mess. That is exactly what they turned out to be — so exactly that the two elements had already acquired the laboratory nicknames pandemonium and delirium before the hypothesis explained why they were so difficult. Once the analogy was assumed rather than fought, their elution position on an ion-exchange column could be predicted in advance, and the technique that the lanthanide chemists had built as a separation tool became, for the actinides, an identification tool: an element was recognised by where in the sequence it came off the column.

Seaborg made the naming carry the argument. Element 95 became americium, after a continent, because its lanthanide counterpart europium is named after one; element 96 became curium, after the Curies, because gadolinium honours Johan Gadolin. Anyone looking at the two rows was meant to see the parallel before reading a word of the justification. The announcement itself was famously disorderly — Seaborg mentioned both elements in answer to a child's question on a radio quiz programme in November 1945, days ahead of the formal disclosure.

Why 5f behaves nothing like 4f

The parallel with the lanthanides is real, but it only holds for the second half of the row, and the reason is a difference in where the f orbitals sit.

In a lanthanide the 4f orbitals are tucked inside the filled outer shells and take no part in bonding. The 5f orbitals of the early actinides are not: they are spatially more extended, and their energies lie close to those of the 6d and 7s levels. Several electrons are therefore available for bonding rather than one fixed set, and the early actinides consequently show a range of oxidation states as wide as anything in the d-block. Thorium works almost exclusively at +4, protactinium at +5, uranium at +6, and both neptunium and plutonium can be pushed to +7 in strongly alkaline solution — the highest oxidation state anywhere in the series that has been firmly established.

From americium onwards the 5f orbitals contract and drop below the 6d in energy, and the series settles into the +3 uniformity that the lanthanides show throughout. This is the structural fact that distinguishes the two f-block families most sharply. One of them is homogeneous from end to end. The other spends its first half behaving like a transition series, changes character around element 95, and only then starts to resemble its counterpart.

Plutonium sits precisely at the transition and is the most awkward element in the periodic table because of it. Four of its oxidation states can coexist in the same aqueous solution at the same time, each with a distinct colour, held in a disproportionation equilibrium that no other element manages. Warm the metal from room temperature to its melting point at ordinary pressure and it passes through six distinct crystal structures on the way — more solid phases than any other element goes through over that interval — and the densest of them is around a quarter heavier than the least dense, so modest heating swells and shrinks the piece by amounts a machinist cannot ignore. The δ phase, the one that is workable, has to be held in place with a few atomic per cent of gallium.

Six were already here

The common assumption that this is a synthetic family is wrong in a way that matters. Thorium occurs in the crust at roughly the same level as lead, and uranium outweighs silver, tin and mercury there. Both are still present because their principal isotopes have half-lives comparable with the age of the Earth — around fourteen billion years for thorium-232 and four and a half billion for uranium-238 — so a substantial fraction of the original supply has not yet decayed. Actinium and protactinium persist only as short-lived links in those decay chains, and traces of neptunium and plutonium form continuously in uranium ores by neutron capture.

Uranium-235 decays more than six times faster than uranium-238, which is why it now constitutes only 0.72 per cent of natural uranium. In the deep past that fraction was much higher, and about 1.7 billion years ago, at Oklo in Gabon, groundwater-moderated deposits of natural uranium sustained nuclear chain reactions for a period measured in hundreds of thousands of years. The evidence is isotopic: the uranium at Oklo carries measurably less U-235 than any other natural sample on Earth, and the products of that fission have never left the rock that made them.

Einsteinium and fermium arrived by the least orthodox route of any elements ever discovered. In November 1952 the Ivy Mike thermonuclear test at Enewetak subjected uranium-238 to a neutron flux so intense that individual nuclei captured a dozen or more neutrons before they had time to decay, and the resulting nuclides beta-decayed up to atomic numbers 99 and 100. They were found on filter papers flown through the fallout cloud and in debris collected from the atoll, and the results stayed classified for several years, which is why the published discovery record for these two elements is later than the work.

Where the neutron route runs out

That accident also marks a boundary. Building heavier nuclei by soaking lighter ones in neutrons works only while each new isotope survives long enough to catch the next neutron, and at fermium-257 the chain stops: the isotopes immediately beyond it fission almost as fast as they form. No reactor, however intense, gets past element 100 this way.

Everything heavier has to be assembled by firing one nucleus at another, which is a different discipline with different laboratories and a different set of arguments, treated at superheavy elements. The practical consequence for this family is a sharp gradient in availability: uranium and thorium by the tonne, plutonium by the tonne in inventories, americium and curium by the kilogram, californium by the milligram, and the last four actinides in quantities that have never been seen as a bulk solid. Einsteinium's chemistry is still being established from samples of a few hundred nanograms.

What they are actually used for

Reactor fuel is the obvious answer, and it accounts for essentially all actinide tonnage. The less obvious answers are more revealing about the range of the family.

• Plutonium-238 heats radioisotope thermoelectric generators. Its 88-year half-life and a decay mode that produces little penetrating radiation make it the only practical choice for a power source that must work unattended for decades, which is why it is aboard both Voyagers, Cassini, New Horizons and the Mars rovers.
• Americium-241 is in domestic smoke alarms of the ionisation type, where its alpha emission maintains a small current across an air gap that smoke particles interrupt. It is far and away the actinide most people live with.
• Californium-252 decays partly by spontaneous fission and is therefore a portable neutron source, used for starting reactors and for neutron logging of boreholes. It is among the most expensive substances on Earth by mass.
• Uranium and thorium underpin radiometric dating. Uranium's two independent decay chains run at different rates to two different lead isotopes, so a single mineral grain carries its own internal cross-check — the property Clair Patterson used in 1956 to fix the age of the Earth at about 4.55 billion years.
• Actinium-225 is being developed for targeted alpha therapy, where a short-range emitter is attached to a molecule that binds to a tumour. Supply, not medicine, is the limiting factor.

Names that were never fully sorted out

The disputes over the transfermium elements are usually told as a story about elements 104 and upwards, but the trouble starts inside this series. Nobelium was announced in 1957 by a Stockholm-based collaboration whose result nobody could reproduce, including the original group. Berkeley claimed the element the following year and Dubna in the 1960s, and when the joint IUPAC/IUPAP working party reported in the early 1990s it credited Dubna. The name proposed by the team whose experiment failed had been in print for thirty-five years by then, and it stayed. Lawrencium's credit was likewise divided between Berkeley and Dubna, decades after both had claimed it.

Two misconceptions are worth naming directly. "Actinide" and "transuranic" are not synonyms: the series starts at actinium or thorium, four elements below uranium. And depleted uranium is less radioactive than natural uranium, not more — it is what remains after the more active isotope has been extracted, and the hazard associated with it is chemical and physical rather than nuclear.