This item is available under a Creative Commons License for non-commercial use only
At the nanometer length scale, the size of surface features in crystalline semiconductor systems is of the same order as the electron wavelength. This can result in unusual behaviour in the systems electronic, magnetic and optical properties due to electron confinement effects. Such effects can have practical and commercial applications and are currently the subject of considerable study in the disciplines of theoretical, computational and materials technology within nanoscience. This thesis uses molecular dynamics computational methods to examine such effects in the electronic structure of semiconductor-based crystalline systems. Three unique surfaces were studied in detail - the SiC(111) surface, the SiC(100) surface, and the prototypical In-Si(111) surface. Silicon carbide is of importance in the development of semiconductor technologies due to its physical robustness and relatively high power capabilities. An understanding of surface metallisation in semiconductors is of paramount importance since modern technology relies on the interaction of metals with semiconductors in integrated circuit and device construction. If Mooreʼs Law is to be adhered to, transistors must become smaller and the metal contacts between transistors must likewise shrink. This work explores the possibility that potassium deposited on the SiC(100) surface may provide a solution for nanoscale contacts between devices on this surface. Using modified and highly efficient molecular dynamics code, the energies and reconstructions of a number of possible surface configurations were studied in detail, resulting in proposed new candidates for surface reconstruction for a range of coverages of potassium on the SiC(100) surface. The SiC(111) surface has previously been shown to undergo an interesting metal-insulator transition where the surface band states split. This has been observed by experimentally probing the surface states with scanning tunneling spectroscopy and photoemission techniques. By applying ab-initio molecular dynamics techniques to simulate this surface, this research has found compelling evidence for the actual mechanism that results in this transition. A number of time-dependent simulations of the surface in question were carried out, over ranges of tens of thousands of picoseconds. The results show that the surface is dynamical in nature. Furthermore, the transition is shown to be due to a soft phonon interaction on the surface, and thus surface dangling bonds are seen to split because they are in constant motion. Finally, computational studies of the In-Si(111) surface are also presented. The results indicate a dynamical surface exhibiting surface phonon effects, similar to the SiC(111) surface studied and metallisation in a similar vein to results obtained for the K-SiC(100) surface. The study of the In-Si(111) surface therefore represents a natural progression in studies of this nature. The computational work presented here was carried out using the FIREBALL suite of tools. During the course of this study, the codebase was rewritten and modernised to improve performance and to allow for easier future modification. The extensive changes to the code are discussed, as are its potential future applications in the field of computational solid state physics. Practical methods are presented that allow for the work to progress to the calculation of optical transitions directly in FIREBALL, with a full description of how a reflectance anisotropy spectrum could be calculated as a logical extension of the present work. The calculation of a reflectance anisotropy spectrum would be of considerable interest to experiments in the field.
Haycock, B. (2011). Calculation of the Electronic and Optical Properties of Nanoscale Systems. Doctoral Thesis. Technological University Dublin. doi:10.21427/D7830Q