Nuclear Fuel Corrosion over Millennia Interpreted Using Geologic Data

Corrosion of nuclear fuel over the 10,000 year regulatory period in a geologic repository will be a function of physical characteristics (e.g., crystallinity, crystal sizes, crystal forms) and chemical characteristics (e.g., crystal composition, compositional variability, accessory phases). Natural uraninite (nominally UO_2+x) which has undergone long-term corrosion can be studied to infer the long-term behavior of nuclear fuel. Previously, uraninite from the Nopal I deposit, Peña Blanca district, Chihuahua, Mexico, has been shown to constitute an outstanding analog material for comparison with nuclear fuel. Similarities between Nopal I uraninite and nuclear fuel have been shown to include bulk composition, general crystal structure, and total trace element content.

In this paper, the crystallographic and compositional comparisons between Nopal I uraninite and nuclear fuel are examined in more detail using data from high-resolution transmission electron microscopy. These measurements indicate good correspondence between Nopal I uraninite and spent nuclear fuel for most of the factors listed above, but also the existence of notable differences. Both nuclear fuel and Nopal I uraninite have bulk compositions consisting largely of U and O and the general crystal structures of the materials may both be approximated by cubic UO₂. Structural and compositional differences between the materials, however, are clear at the scale of the crystal lattices (e.g., <100 nm). The crystal lattice boundaries of Nopal I uraninite appear to be physically stronger and more resistant to oxidant penetration than the grain boundaries of irradiated nuclear fuel. There is limited evidence for preferential oxidation of Nopal I uraninite along crystal boundaries, but no zoning of corrosion products from a crystal boundary inward has been observed. Although crystals of Nopal I uraninite (e.g., < 100 nm) are much smaller than irradiated fuel grains (e.g., < 25 μm), because of the boundary differences described here, the effective surface area/volume ratio for Nopal I uraninite may be much smaller than that of nuclear fuel, and consequently the rates of reaction may be different. Micro-compositional measurements show that the content and distribution of trace elements in Nopal I uraninite is different from irradiated fuel. High-resolution (10 Å spot) EDS measurements of individual crystals of Nopal I uraninite indicate that some areas consist of only U and O but that other uraninite lattices incorporate Al and Si as well as U and O. Irradiated fuel includes many elements in addition to U and O, but Al and Si are not significant among those impurities. Nopal I uraninite has a composition close to U₄O₉, whereas irradiated fuel is close to UO₂. Lattice mismatches within individual Nopal I uraninite crystals suggest that in some limited areas a more reduced phase persists. On a compositional basis, Nopal I uraninite may, therefore, correspond more closely to irradiated fuel which has undergone dry oxidation to UO_2.25, than to fresh (unoxidized) irradiated fuel.

Data presented here suggest that, as a bulk material, Nopal I uraninite compares favorably with irradiated nuclear fuel. Nevertheless, some fine-scale differences are noted between Nopal I uraninite and irradiated nuclear fuel with respect to both internal structures and compositions. These observations suggest that whereas the long-term responses of the two materials to oxidative alteration in a geologic repository may be similar, the detailed mechanisms of initial oxidant penetration and the short-term oxidative alteration of Nopal I uraninite and irradiated nuclear fuel are likely to be different.