Integrating radioactive material into mineral-based ceramics is a leading contender for the disposal of nuclear waste. Some of these ceramics, such as "zircon" (ZrSiO4), already occur naturally with slowly-decaying radioactive isotopes incorporated into their crystalline structure. Nevertheless, they have remained intact over billions of years despite the damage caused by the onslaught of high-energy alpha particles produced in the decay process.

Some scientists had hoped that zircon could withstand much higher doses of the radioactive plutonium isotope 239Pu, which is found in spent nuclear fuel. The risk is that increased exposure to alpha particles would displace too many atoms and damage the crystalline structure irrevocably. But this damage had been difficult to measure and in the past scientists relied on vague empirical calculations based on the assessment of large defects to predict how long the ceramics would last.

Mineral physicists Ian Farnan and colleagues at the University of Cambridge may now have the answer, however. They used a technique called "magic-angle spinning" NMR on zircon, showing that each alpha-particle displaces up to 5000 atoms in the crystal lattice, rather than the 1000 to 2000 estimated before. The technique enhances the resolution of the NMR spectrum by spinning a sample at high speeds and at a certain angle to the applied magnetic field. This is the first time individual damage events have been witnessed, and could put an end to the "back of the envelope" calculations that had prevented scientists from accurately determining a material's lifespan.

Unfortunately this means that zircon containing 10% of 239Pu (roughly the dose required for radioactive waste storage) would break down after just 1400 years – nowhere near the 250 000 years that regulation dictates. Although the technique has ruled-out zircon, it could pave the way for characterizing other materials over long timescales.

"The main issue with siting a nuclear waste repository is that there are many uncertain factors," said Farnan. "When you extrapolate these into the future you get a very large uncertainty, which can make the idea of a repository intractable. But we feel that by working on the material itself, that's where you are going to get the biggest effect."