Oxygen exchange and diffusion in uranium dioxide single crystals

In a talk that promised possible insights into the recent flurry of interest over PuO_2+x (superstoichiometric plutonium oxyhydroxides), Steve Joyce of Los Alamos discussed the issues inherent in the diffusion and incorporation of oxygen into uranium dioxide (UO₂) crystals. These issues included changes in volume and conduction properties in addition to the obvious alterations in stoichiometry, both stoichiometric (U₄O₉, U₃O₇, U₂O₅, and UO₂3) and superstoichiometric (UO_2±x). Joyce emphasized that in its UO2 fluorite crystal structure, the compound is quite amenable to additions of oxygen with shifts in the oxygen positions, even as the uranium lattice is maintained. He offered data that tended to support those contentions.

The basis of these experiments was the exposure of UO2 crystal surfaces to both molecular oxygen and water, at different temperatures, to assess both diffusional penetration and the effect of the diffusing species on the crystal. The temperature variation is important because while UO₂ can be oxidized by water at lower temperatures, O2 is only reactive to the UO2  at higher temperatures, and all previous research has studied the system at temperatures of 950 kelvin (K) and above.

An additional experimental procedure was the use of electric sputter damage to the crystal surface, thereby creating oxygen vacancies within 10 nanometers (nm) of that surface and leading to the evolution of hydrogen gas, H₂. Since each successive exposure to sputter reduced H₂ evolution, the conclusion, confirmed by X-ray photoemission spectroscopy, was that the sputter was accompanied by the dissociation of water molecules. Oxygen was “left behind” to heal those crystal vacancies, and H₂ evolved as a product of that dissociation of water.

Joyce noted that previous studies had indicated that below 1500 K oxygen was found to be the only diffusing species and that uranium diffusion was unimportant. To better assess these processes, water was labeled with ¹⁸O so that the crystal’s surface could be nondestructively labeled with a heavier oxygen isotope. Prior attempts to use isotopically labeled molecular oxygen (¹⁸O₂ ) had not been very successful without sputtering and its concomitant damage, which was undesirable in this instance.

continued...