In this thesis we develop and apply a variety of first-principles computational methods to the study of complex oxide heterostructures.
Firstly, we introduce a new set of partially localised orbitals that offer a natural description for 1D- and 2D-periodic systems,
providing the same accuracy as traditional plane-wave DFT-based approaches but with a computational cost that increases only linearly with the size of the system along the non-periodic directions, thus alleviating the unfavourable scaling of traditional DFT calculations. The method was implemented in the ONETEP code, which is a leading large-scale DFT code that was originally designed to work with fully localised orbitals and Gamma-point sampling only of the Brillouin zone. We show that the use of our new basis set in conjunction with the implementation of the Brillouin zone sampling at arbitrary k-points significantly extends the systems that can be studied with a DFT-based approach. We then discuss the performance of a force field parametrised to reproduce DFT energies, stresses and forces in bulk barium titanate (BTO), when it is applied to the study of BTO surfaces and interfaces. We find that, upon structural relaxation, the displacements of the surface
atoms are qualitatively incorrect when compared with DFT predictions because of the different local environment with respect to the bulk, and that the transferability of the force field can be greatly improved by introducing a handful of extra parameters for the surface atoms only.
We then explore interesting phenomena at the interface between two different complex oxides. We first analyse the possibility to induce oxygen octahedral distortions from a perovskite substrate into an originally undistorted film. We observe that for
the system of interest (barium zirconate grown on top of strontium zirconate) the induced distortions have a non-negligible decay length of roughly 25 angstrom, and the
gradients associated with the decaying modes can induce novel physical properties in the film.
We finally explore the application of complex oxides as solid electrolytes in next-generation Li-ion batteries, a promising path towards the development of efficient, light and safe energy storage systems to be used in the transportation industry. We develop a custom framework to perform a preliminary screening of potential candidates, focusing on the requirement of the electrolyte to be stable
against commonly used electrodes, and we find materials that are stable in contact with pure Li metal at significantly higher voltages than the traditional organic solvents used in current Li-ion batteries.