I established first principles calculations, based on density functional theory (DFT), of the structural and electronic properties of the metastable hexagonal lonsdaleite phases of the group-IV semiconductors Si and Ge, and the wurtzite phases of the III-V semiconductors (Al,Ga,In)(P,As,Sb). In addition to the lattice parameters and electronic band structures, DFT calculations of elastic moduli, band edge deformation potentials, band offsets and spontaneous polarisation were also performed.
Based on the DFT-calculated band structures, I established a 10-band k.p Hamiltonian for hexagonal semiconductors. This extends the basis of states included in the conventional 8-band Hamiltonian for wurtzite semiconductors via inclusion of the "pseudo-direct" conduction band, which originates via back-folding of the L-point CB minimum from the conventional cubic (diamond or zinc blende) phase, giving rise to a direct fundamental band gap in Ge and GaP. Parametrisation procedure directly from DFT calculations reproduces the optical matrix elements and selection rules predicted by DFT calculations. This procedure has been applied to several III-V semiconductors, providing k.p parameters suitable for computationally inexpensive analysis of their electronic and optical properties.
I have implemented predictive calculations of spontaneous emission spectra and the radiative recombination coefficient B. Application to lonsdaleite Ge has produced new insights into the nature of direct-gap optical emission, which has driven a recent rapid growth of research interest in this emerging material system. Calculated spontaneous emission corroborates experiment, showing that the observed optical emission from lonsdaleite Ge is associated with a direct fundamental band gap. Prediction of the dependence of the spontaneous emission on carrier density explains recent experimental data pertaining to stimulated emission being associated with a second, higher-energy emission peak. Achieving stimulated emission is a key step towards demonstration of laser action. My calculations provide quantitative understanding of the observed optical emission. Calculated B coefficient for lonsdaleite Ge is approximately three orders of magnitude lower than that inferred based on initial analysis. This demonstrates that the assumption of purely radiative recombination in lonsdaleite Ge nanowires is unlikely, suggesting that non-radiative recombination plays a significant role. I showed that application of strain to lonsdaleite Ge can enhance the B coefficient by over two orders of magnitude, which promises efficient mid-infrared emission. Application of this calculational approach to III-V semiconductors allowed comparative analysis of light emission in zinc blende vs. wurtzite semiconductors, quantifying differences in radiative recombination between crystal phases.
I established DFT calculations for crystal phase heterostructures, a class of quantum-confined heterostructures formed by varying the crystal structure in a single material. Quantum confinement is generated by changes in the electronic structure between the different crystal phases. For several III-V materials, crystal phase heterostructures have been predicted to possess type-II band offsets, spatially separating electrons and holes and reducing the radiative recombination rate. My calculations for Ge confirm the presence of type-I band offsets, with electrons and holes having high spatial overlap and a radiative recombination rate that can be up to an order of magnitude larger than in unstrained bulk lonsdaleite Ge. This identifies nanostructuring via crystal phase engineering as an interesting route for the development of Si-compatible light-emitting materials.
