Atomistic Simulation of Defective Oxides

Abstract

Atomistic simulation calculations based on energy minimization techniques and classical pair potentials are used to study several inorganic materials. Chapter 1 gives an overview of the relevant literature and chapter 2 describes the methodology. In chapters 3–5 the lattice parameter variation arising from non-stoichiometry and doping is predicted for three binary oxides. It is shown that the progressive expansion or contraction of the lattice is a direct consequence of the defects and the relaxation of the surrounding lattice. Thus it is possible to test proposed defect models against experiment. The systems investigated are non-stoichiometric wüstite and M2O3-doped c-ZrO₂ and CeO₂.

The extensive non-stoichiometry exhibited by wüstite is accommodated by large defect clusters. Due to the size and high concentration of defects found in wüstite predictions are made at the dilute limit and at finite concentrations via the large unit cell method. In the cases of the M₂O₃-doped c-ZrO₂ and CeO₂ the introduction of a lower valent cation creates oxygen vacancies and particular attention is paid to the interaction of the dopant with its compensating defect.

Chapter 6 investigates La₂NiO_4+δ. The diffusion properties of La₂NiO_4+δ indicate that it would be a good material for use as a cathode in electrochemical devices. This study attempts to gain a better understanding of the underlying defect processes. The predominant intrinsic disorder reaction and the mechanism by which excess oxygen is accommodated are established. Furthermore, the most favourable migration mechanism and pathway for oxygen ions is predicted.

Chapters 7 and 8 investigate pyrochlore oxides. These materials are candidates for solid oxide fuel cell components and as actinide host phases. Such applications require a detailed understanding of the defect processes. The defect energies, displayed as contour maps, are able to account for structure stability and, given an appropriate partial charge potential model, to accurately determine the oxygen positional parameter. In particular, the dependence of the positional parameter on intrinsic disorder is predicted. It is demonstrated, by radiation damage experiments, that these results are able to predict the radiation performance of pyrochlore oxides.