Abstract
One of the key objectives in developing solid oxide fuel cells and oxygen membranes is the improvement of ionic conductivity in electrolyte materials. Optimization of such materials relies on the understanding of the oxygen−ion diffusion mechanisms at the atomic scale. Getting a clear physical−chemical picture is thus the prerequisite to make an educated guess of the best choices of dopant. We highlight in this work some of the most salient recent advances in point defects and oxide ion conductivity studies obtained from atomic-scale simulations performed in the framework of the density functional theory using a supercell approach. First principles calculations have been performed on ceria doped with three kinds of trivalent cations (La3+, Y3+, and Lu3+) to probe the incidence of both dopant size and distribution on relative phase stability and oxygen diffusion efficiency. Ionic relaxation patterns indicate that the crystal structure reorganization after introduction of defects (DCe′, VO..) involves both electrostatic and steric parameters. A clear dopant site selectivity is evidenced for Y- and Lu- doping cases, while much less selective situation of dopants positioning characterizes lanthanum-doped ceria. The study of oxygen mobility has been extended to all possible successive atomic jumps within the supercell, along the three main directions. The set of energy barriers to diffusion can be rationalized in terms of stabilizing and destabilizing Coulomb interactions, elastic energy loss, and steric factors in link with the match between dopant size and its coordination number. On the basis of maximum energy barriers to diffusion and site selectivity features, yttrium and lanthanum dopants appear to be the most appropriate choices provided that a low to medium dopant concentration is preserved. Lutetium doping is clearly less favorable. At a higher doping rate, the increase of probability of occurrence of a very unfavorable configuration in La-doped ceria should be detrimental on the migration viewpoint.