Segregation and binding energetics at grain boundaries in fluorite oxides

Improving oxygen conductivity in fluorite oxides is currently one of the main focus areas in the research of solid electrolytes for solid oxide fuel cells. In the past few years, significant efforts have been made to understand oxygen diffusion mechanisms at grain boundaries to leverage the high conductivity in nanocrystalline oxides. However, the results have remained controversial where both increase and decrease in oxygen conductivities have been observed in nanocrystalline oxides. Recent work shows that dopant segregation could be playing a key role in deciding the ultimate fate of oxygen conductivity at grain boundaries. In this work, using atomistic interatomic potential calculations on six different tilt grain boundaries in three fluorite materials, i.e., CeO2, ZrO2 and UO2, we elucidate dopant segregation and dopant-oxygen vacancy binding energetics at grain boundaries. Using a variety of +3 dopants of different ionic radii, we show that the tendency of dopant segregation could be largely dependent on the difference in the ionic radii of the dopant and the host cation. We find that dopants with the least ionic radius difference could show minimum segregation. We also show that dopant-oxygen vacancy binding at grain boundaries could significantly affect oxygen diffusivity at grain boundaries. Both high and low binding energies are observed at grain boundaries compared to the bulk. We suggest that such wide differences in binding energies possibly explain the wide diversity of oxygen conductivity results that have been previously observed experimentally. Finally, we suggest that dopant-radius based strategies to prevent dopant segregation could help improve oxygen conductivity in the design of fast-ion conducting nanocrystalline oxides.