In its high-temperature cubic δ-phase, pure bismuth oxide demonstrates the highest oxide ion conductivity among all solids, reaching approximately 1 S cm^-1 just above 1000 K. However, a phase transition occurs below this temperature, causing a sudden decline in conductivity. Substantial scientific efforts have been directed towards developing a bismuth oxide-based compound that maintains the high conductivity of the δ-phase at lower temperatures. Various approaches, such as doping or creating materials where bismuth is not the predominant cation, have been explored. In this study, we illustrate how a comprehensive understanding of conduction processes in new solid electrolytes can be attained by combining experimental methods with theoretical modeling, employing a molecular dynamics approach

We utilize diverse Density Functional Theory (DFT) methods, primarily employing versions of the general gradient approximation in molecular dynamics, especially at elevated temperatures reaching up to 1200 K. Depending on the specific issue under investigation, we adjust the conventional ab initio approach by incorporating van der Waals interactions or Hubbard corrections as necessary. All simulations are performed using Vienna Ab initio Simulation Package. Calculations are typically conducted on systems ranging from as much as three hundred atoms, especially when modeling lighter species, to as few as one hundred atoms for compounds predominantly composed of heavy ions.

In the perovskite sodium bismuth titanate, Bi_0.5Na_0.5TiO₃, we elucidate the modulation of ionic conductivity through the composition-dependent existence of polarons, resulting in the peculiar formation of clusters of oxide ion vacancies. This phenomenon has been employed to account for the observed sudden decrease in ionic conductivity when subtle adjustments are made to the Bi/Na ratio, as observed in experiments. In a similar pervoskite system, 0.2(Ba_0.4Sr_0.6TiO₃)-0.8(Bi_0.5Na_0.5TiO₃), the integration of high-resolution powder neutron diffraction, impedance spectroscopy, and ab initio calculations unveils that titanium makes a contribution to the overall polarization that is less than one-third in magnitude. Notably, the displacements of oxygen ions and A-site cations, especially bismuth, play a crucial role. In the case of the rhombohedral form of praseodymium doped Bi₂O₃ we show how the experimentally observed phase transition of the anion sublattice can be explained by the rotation of BiO₃ quasi-molecule, driven by the presence of the 6s² electron lone pair of the bismuth cation.