Palladium corrosion of the SiC layer in TRISO particle fuels: A separate effects study to assess mechanisms for palladium attack

Abstract

Palladium (Pd) is a fission product of particular interest for TRISO fuel performance due to its propensity to chemically corrode the SiC layer. Post irradiation examination experiments on irradiated TRISO particles confirm the formation of several palladium silicide intermetallics, such as PdSi and Pd₂Si, in the SiC layer. A corroded SiC layer can lead to TRISO particle failure, resulting in higher fission product release. However, the specific transport mechanisms driving Pd corrosion remain difficult to isolate during irradiation tests. Therefore, this study aims to quantify the thermal diffusion kinetics of Pd through the OPyC layer and its subsequent interaction with the SiC layer under separate effects conditions. The research presented uses surrogate TRISO particles, with Pd-rich coatings on the OPyC layer, and high temperature separate effects testing to examine the behavior of Pd diffusion and chemical interactions on the TRISO particle layers. Physical vapor deposition techniques were used to deposit 200 nm of Pd on surrogate TRISO particles, which were then annealed at various high temperatures. SEM/EDS characterization was used to evaluate the annealed microstructures and revealed a distinct thermal activation threshold. While no Pd diffusion was observed at 850 °C after 10 h, annealing at 1150 °C resulted in complete diffusion through the OPyC layer and Pd-SiC interaction along the OPyC-SiC interface. Pd diffusion was accompanied by OPyC degradation, including OPyC cracking and Si migration to the particle surface. These findings establish 850–1150 °C as a critical transition regime for Pd chemical attack, providing experimentally derived reaction limits necessary for validating fuel performance models and defining safety margins for high-temperature transients. This research characterizes thermally driven Pd diffusion in TRISO fuel layers under non-irradiated conditions, contributing to improved predictive capability for fuel integrity in advanced nuclear energy systems.