Heavy elements are one thing, but it is lighter uranium that is the worry

A wrong hypothesis is seldom completely wrong – a consoling thought for me and many other chemists. My favourite wrong chemical hypothesis is William Prout’s 19th century notion that atomic weights should be whole numbers. Atomic weights measured to higher accuracy soon showed that Prout’s claim was simply wrong. The atomic weight of silver, for example, as measured to high accuracy by electrochemistry, was 107.870. And yet his idea did not go away.

Many atomic weights were oddly close to whole numbers: hydrogen, 1.008; carbon, 12.011; nitrogen, 14.007. (I am quoting modern values.) The matter was largely cleared up by the discovery of isotopes. Many elements exist in nature as mixtures of atoms, each with its own number of neutrons in its nucleus. Those neutrons have no effect on the chemistry of the element: but they do affect its atomic weight, which chemists measure as an average over many billions of atoms. Atoms which only have one dominant isotope in nature (hydrogen, carbon and nitrogen, for example) have atomic weights close to a whole number.

The idea of isotopes grew gradually in the scientific mind; H G Wells was an early speculator. When the mass spectrograph came along in the 1920s, and single atoms could be weighed, isotopes were made plain. Modern chemical atomic weights are derived from natural samples; but good tabulations warn of small variations due to isotopic differences, arising perhaps in nature and perhaps from technology – the inadvertent or unacknowledged partial separation of atomic isotopes. 


Credit: M H Jeeves

Isotopes added a new dimension to chemistry. The best known isotope, deuterium, ‘heavy hydrogen’, has a single neutron attached to its proton nucleus, and is stable. Heavy water is well known, and so are many other ‘heavy’ compounds.

The most sinister isotope, central to the atom bomb, is 235U. It amounts to about 0.7% of natural uranium, most of the rest being 238U. Natural uranium can explode – indeed, a site in Africa may in the distant past have suffered a geological nuclear explosion from uranium ore. But a weapon light enough to be lifted by an aircraft or a rocket needs almost pure 235U. Accordingly, much of the effort of developing the atom bomb went into the separation of uranium isotopes.

That effort settled on UF6, which is conveniently gaseous, and less reactive than many other compounds of uranium. Its isotopic molecules have slightly differing weights: 235UF6 = 349 and 238UF6 = 352, their ratio 1.0086. The diffusion rate of those two weights differs slightly. As Thomas Graham showed, diffusion rates through a porous membrane vary as the square root of the molecular-weight ratio. In this case, the ratio was very close to unity and so the isotopic separation of UF6 (which used many porous diffusion membranes of PTFE) was extremely slow and difficult. It was made worse by the very small proportion of the desired 235UF6. Conversely, ‘depleted uranium’, 238U from which 235U has been removed, became relatively cheap, and has several uses.

These days, UF6 is still the chemical of choice for the separation of uranium isotopes. The method still depends on that tiny difference of weight, though centrifugation is now preferred to diffusion. Even so, the process is still fearsomely slow and troublesome. Currently, the state of Iran is trying to do it, allegedly for peaceful nuclear power. Few believe this claim – nuclear power has many troubles of its own, and Iran has plenty of oil. Tom Lehrer’s song about nuclear bomb proliferation urged its audience to ‘try to stay serene and calm, when Alabama gets the bomb’. I suspect that Iran will precede Alabama.