The effect of lattice and grain boundary diffusion on the redistribution of Xe in metallic nuclear fuels: Implications for the use of ion implantation to study fission-gas-bubble nucleation mechanisms

A multi-atom gas bubble-nucleation mechanism has been proposed as part of a predictive fission-gas release model for metallic nuclear fuels. Validation of this mechanism requires experimental measurement of fission-gas bubble-size distributions at well-controlled gas concentrations and temperatures. There are advantages to carrying out such a study using ion implantation as the source of gas atoms compared with neutron irradiations. In spite of previous successes using ion implantation to study fission-gas behavior in oxide fuels, there is significant uncertainty about the efficacy of using ion beams for metallic fuel studies. To address the question of the applicability of ion beams in experiments designed to study fission-gas behavior in metallic fuels, we developed and applied an exact model for the redistribution of implanted ions under annealing conditions. The conclusion is that, given the assumptions, the results from implantations at 1MeV or less may be overwhelmed by the surface effects at all relevant temperatures. Implanting at 10 or 80MeV can significantly diminish the influence of the surfaces and the steep concentration gradients. At 80MeV, the location of the peak concentration profile remains stable with annealing time. Thus, it appears that ion implantation can be an appropriate tool to study the size distribution of Xe bubbles in metallic fuels. Of the conditions investigated, the best for the study are to implant at 80MeV and carry out anneals at 773K, 873K, and 973K for times less than 10,000s.