Modelling of electronic and structural strain effects in grain boundaries of semiconducting oxides
Synthesis and Properties of Functional Materials
Mechanical Properties of Metallic Materials and their Microstructural Origins
Final Report Abstract
In this project the role of inversion domain boundaries (IDB) in piezoresistive ZnO bicrystals was studied by means of atomistic modelling and electronic structure calculations. The following key results were achieved: • The atomic structure of pristine and doped ZnO{0001} inversion domain boundaries was resolved by extensive DFT calculations and thermodynamics assessment. This structure is characterised as a stoichiometric 2x2 reconstruction where the 4-fold bonding configuration of the bulk is preserved. This finding is of general interest for other wurtzite {0001} or zincblende {111} IDBs such as e.g. GaN, AlN, InN, SiC, etc., including heterojunctions and ternary systems. Concerning solutes and dopants a clear trend for segregation at the IDBs is predicted, in agreement with experiments. • Pristine Pristine GBs show electronic properties similar to the bulk. Solute and dopant segregation leads to the formation of gap states (for TM solutes) that may trap electrons and lead to the formation of electrostatic potential barriers. A strain sensitivity of the electronic structure was, however, not observed in our calculations. To what extend this is related to an insufficient treatment of localisation effects in the standard PBE approximation of the exchange-correlation functional is currently investigated by additional Hubbard-U corrections. • Besides these core results for ZnO, we also developed and implemented a variationcomposition basin-hopping global structure optimiser. The optimiser is a versatile tool to identify, for example, low-energy grain boundaries in large scale polycrystalline material models, in particular, for multi-element systems. It was also successfully applied within another DFG-funded project. Moreover, promising steps towards implementing semiperiodic boundary conditions for electronic-structure calculations were made. Such boundary conditions would be highly appreciable for a wide range of challenges in computational materials science were symmetry constrains prevent direct calculation of properties.
