Nouvelles du nucléaire

CONTROVERSES NUCLEAIRES !
2007
janvier
Le zircon fait faux bond: la gestion à très long terme des déchets nucléaires
ne sera pas un long fleuve tranquille.
La preuve par la saga du zircon.
Sources ADIT/NATURE

Le magazine Nature en apporte encore une confirmation sous la forme d'une sorte de douche froide: les matériaux que l'on espérait utiliser pour immobiliser les substances radioactives dont la durée de vie est la plus longue (plutonium, neptunium, américium, curium, etc.), ces matériaux risquent de ne pas résister bien longtemps...

    Jusqu'ici le consensus s'était fait dans la communauté scientifique sur la nécessité de séparer les diverses composantes du combustible usagé des centrales nucléaires (retraitement) et de fixer les isotopes les plus dangereux dans des matériaux qui devraient résister à l'épreuve du temps. On pensa d'abord à les vitrifier, mais tout porte à croire que le verre ne résisterait pas suffisamment à l'humidité et que les eaux présentes dans les couches géologiques où l'on enfermerait les déchets finiraient par se charger de radioactivité.
     Une alternative fut proposée, qui consiste à fixer les radioéléments dans des zircons, c'est-à-dire des céramiques composées d'oxygène, de silicium et de zirconium. Ces matériaux sont bien connus des scientifiques dans la mesure où ils servent souvent à dater des couches géologiques.
     Pratiquement, en effet, ils sont aussi vieux que la Terre elle-même.
     Un gage de stabilité donc, encore renforcé par le fait qu'il existe des zircons naturels contenant des matériaux radioactifs comme l'uranium et le thorium.
     On se livra toutefois à des études poussées sous la forme de simulations afin d'établir le risque de voir le matériau détruit par les radiations. Il faut savoir en effet que le phénomène fut observé dans les réacteurs américains destinés à produire les composantes des premières bombes atomiques. On disait alors que les matériaux souffraient de la maladie de Wigner, du nom du physicien qui la décrivit.
     On estima ainsi que chaque particule alpha émise par le déchet radioactif déplacerait quelque 1.000 à 2.000 atomes, un risque que l'on estima tout à fait acceptable. Mais il ne s'agissait que d'une estimation. En collaboration avec des collègues du Pacific Northwest National Laboratory (USA), le Pr Ian Farma, de l'université anglais de Cambridge décida de procéder à une vérification expérimentale en introduisant une forte dose de plutonium 239 dans un zircon et en observant celui-ci dans un appareil de résonance magnétique.
     Une expérimentation très difficile à mettre en place, mais le jeu en valait la chandelle: il s'avéra que les particules alpha émises par ce produit ainsi que par l'uranium 235 dans lequel il commence par se transformer avant de perdre sa radioactivité, ces particules donc déplacent chacune près de 5.000 atomes. Soit significativement plus que prévu. Pour paraphraser l'éditorial de Nature, c'était comme envoyer un boulet de canon sur une piste de danse bondée.
     Avec pour conséquence que le zircon devrait perdre sa structure cristalline parfaite après 210 ans et sa stabilité après quelque 1.400 ans. Ces délais peuvent sembler longs.
     Mais il faut savoir que le plutonium (qui n'est pas nécessairement le déchet le plus dangereux, mais c'est celui qui est le plus politiquement sensible, en raison sans doute de son passé militaire) a une demi-vie de 24.110 ans. Qu'en d'autres termes, il lui faut toutes ces années pour perdre la moitié de sa radioactivité.
     A cette échelle le zircon fait donc pâle figure. Les techniciens du nucléaire pourraient donc bien revoir une fois de plus leur copie.

ANGLAIS

Published online: 10 January 2007; | doi:10.1038/news070108-6

Canned nuclear waste cooks its container. Estimates of radiation damage to materials have been too low.

     The atomic order of a ceramic is muddled into a glassy mess by radiation.
     Storing high-level nuclear waste without any leakage over thousands of years may be harder than experts have thought, research published in Nature today shows.
     Ian Farnan of Cambridge University, UK, and his co-workers have found that the radiation emitted from such waste could transform one candidate storage material into less durable glass after just 1,400 years — much more quickly than thought1.

Current plans for disposal of some of the most dangerous material generated in nuclear power plants, such as radioactive elements extracted from spent fuel rods, differ from one country to another. A common strategy being explored is to encase the waste in a hard, crystalline ceramic material — a kind of synthetic rock — and then put it in steel canisters and bury them in cavities excavated underground.
Because many radioactive substances continue emitting radiation for a very long time, the containment must persist for an awesome duration. Plutonium-239, one of the most deadly by-products of nuclear power, has a half-life of 24,000 years, meaning that only half of any initial batch has decayed over this time. Ideally it should stay put for about ten times as long: a quarter of a million years.

Candidate ceramic
     Farnan and colleagues have investigated one candidate material hoped to do the job, called zircon (zirconium silicate). The plan is that this ceramic material will hold on fast to the radioactive atoms and stop them from finding their way into the environment — for example by being dissolved and dispersed in ground water.
     The problem is that the radioactive waste damages the matrix that contains it. Many of the waste substances, including plutonium-239, emit alpha radiation, which travels for only very short distances (barely a few hundredths of a millimetre) in the ceramic, but creates havoc along the way.
     A fast-moving alpha particle knocks into hundreds of atoms in its path, scattering them like skittles. Worse still, the radioactive atom from which the particle comes is sent hurtling in the other direction by the recoil. Even though its path is even shorter than that of an alpha particle, the atom is much heavier, and can knock thousands of atoms out of place in the ceramic.
     All this disrupts the crystalline structure of the ceramic matrix, jumbling it up and turning it into a glass. That can make the material swell and become a less secure trap. Farnan says that some zircons that have been heavily damaged in this way by radiation have been found to dissolve hundreds of times faster than undamaged ones. So if the ceramic gets wet, there could be trouble.

Hit and run
     Previous estimates of the radiation damage to waste-storage ceramics have relied largely on calculations and computer simulations. Now Farnan and colleagues have measured it directly.
     They used a technique called nuclear magnetic resonance spectroscopy — similar to the method of magnetic resonance imaging (MRI) used in biomedicine — to measure the relative amounts of crystalline and glassy material, both in artificial zircon containing plutonium and in naturally occurring mineral zircon, which commonly contains radioactive uranium. They estimate that each alpha-decay event of a radioactive atom displaces around 5,000 atoms in the zircon - between 2.5 and 5 times more than predicted previously.
     "There's more damage than we thought," says Rod Ewing, a specialist in nuclear-waste disposal at the University of Michigan in Ann Arbor.
     There are other materials that may fare better than zircon, including other zirconium minerals. But Farnan's work implies that we probably don't yet fully understand how well any of these materials might stand up to the battering of radiation. He thinks the findings should encourage engineers to think very carefully about the matrix encasing the radioactive waste, rather than focusing on the geological characteristics of the burial site. Ideally, the best material would be able to heal itself, with the atoms displaced by alpha decay moving back slowly into their crystalline positions.
     Ewing notes that the technique used in this study could be used to investigate these alternative materials, hopefully to find a longer-lived candidate.

References
1. Farnan I., et al. Nature, 445. 190 - 193 (2007).

Letter
Nature 445, 190-193 (11 January 2007) | doi:10.1038/nature05425; Received 30 June 2006; Accepted 3 November 2006
Quantification of actinide alpha-radiation damage in minerals and ceramics, Ian Farnan1, Herman Cho2 and William J. Weber2

1. Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge, CB2 3EQ, UK
2. Pacific Northwest National Laboratory, Richland, Washington 99352, USA

Correspondence to:
Correspondence and requests for materials should be addressed to Ian Farnan (Email: ifarnan@esc.cam.ac.uk).
There are large amounts of heavy alpha-emitters in nuclear waste and nuclear materials inventories stored in various sites around the world. These include plutonium and minor actinides such as americium and curium. In preparation for geological disposal there is consensus that actinides that have been separated from spent nuclear fuel should be immobilized within mineral-based ceramics rather than glass because of their superior aqueous durability and lower risk of accidental criticality. However, in the long term, the alpha-decay taking place in these ceramics will severely disrupt their crystalline structure and reduce their durability. A fundamental property in predicting cumulative radiation damage is the number of atoms permanently displaced per alpha-decay. At present, this number is estimated to be 1,000–2,000 atoms/alpha in zircon4. Here we report nuclear magnetic resonance, spin-counting experiments that measure close to 5,000 atoms/alpha in radiation-damaged natural zircons. New radiological nuclear magnetic resonance measurements on highly radioactive, ²³⁹Pu zircon show damage similar to that caused by ²³⁸U and ²³²Th in mineral zircons at the same dose, indicating no significant effect of half-life or loading levels (dose rate). On the basis of these measurements, the initially crystalline structure of a 10 weight per cent ²³⁹Pu zircon would be amorphous after only 1,400 years in a geological repository (desired immobilization timescales are of the order of 250,000 years). These measurements establish a basis for assessing the long-term structural durability of actinide-containing ceramics in terms of an atomistic understanding of the fundamental damage event.