Nanoscale Techniques for Investigating Material Issues in Quantum Dot Based Nanoelectronics

The current scaling of feature size of complementary metal oxide semiconductor transistors has been predicted to reach its limits by around the end of this decade. Therefore, several competing strategies for the post-CMOS era are under investigation.;The focus of this dissertation in on two key materials issues pertaining to semiconductor nanostructures, more specifically Ge-Si quantum dot based nano-electronics. A key issue here is the understanding of local chemistry of these nano-structures because the local chemistry affects the positions of the electronic band edges of these structures relative to that of the surrounding matrix, which in-turn affects the carrier localization properties. While the nano-scale chemistry of the QDs is relatively well understood, the chemistry of QDMs is not as well understood. Therefore, focus of this dissertation is the understanding of detailed nano-scale chemistry of QDMs. Another key issue arises from the use of the focused ion beam (FIB) for controlled delivery of dopant ions into the QDs and the QDMs for creating controlled dopant profiles at the nanoscale, and for templating the growth of these structures. The materials issue associated with this application of the FIB is the damage recovery of ion implanted (FIB) regions.;Although ion implantation damage and recovery of Si implanted using commercial broad area implantation implanters is well understood, the FIB implantation damage and recovery of Si is not as well understood. The focus of this research is thus to understand the effect of high ion implantation current density in the FIB, and the effect of FIB ion species (Si, Ge and Ga) on the damage recovery of Si.;With regards to the first body of research Auger electron spectroscopy (AES) was used for mapping the chemistry of QDMs in the epitaxy of Si 0.7Ge0.3 on Si(100). The AES study shows that the pit bases of QDMs are richest in Ge, which is consistent with one existing paper on composition distribution within these QDMs which employed X-ray diffraction based method. Secondly, our study shows that the Si composition monotonically decreases from the outer edges of the QDMs towards the pit cusps. The segregation of Ge to the pit bases is inconsistent with strain energy minimization, and it is proposed to be due to attachment of Ge to steps within the interiors of the QDMs. The monotonic decrease in Si composition (the total decrease is 12%) across the QDs and towards the pit cusps is inconsistent with literature reports for composition distribution within isolated QD structures. The reason for this observation is unclear at present but it illustrates how the presence of pits adjacent to the QDs can affect the composition of the QDs.;In the second body of research, 30 kV FIB implants of Si2+, Ge2+ and Ga+, and 60 kV broad ion beam Si + implants, in Si(100) were studied using Raman spectroscopy and transmission electron microscopy (TEM). Raman spectroscopy was carried out on the above listed ion implanted Si specimens, both before and after annealing to 730 C, 800 C and 900 C, for different annealing times, using 405 nm and 514 nm wavelengths. This enabled quantification of structural damage and stress using peak height and shift of the crystalline Si Raman peak at 520 cm-1 (for unstressed Si-Si bonds), as the metrics. Raman measurements from the Si-FIB and the Si-broad beam implants in their as-implanted states show a decrease in Raman peak height, and an increase in peak shift, with dose, implying greater structural damage and stress in the FIB implants. Second, in the Si-FIB implants, higher stresses have been observed to lead to accelerated evolutions of category I {311}s and category II defects, when compared to the corresponding evolution in the corresponding Si-broad beam implants. Third, it is observed that the limited solubility of Ga in Si plays a major role in limiting the recovery of Ga-FIB implanted Si. Fourth, it is observed that the damage in the Si-FIB and Ge-FIB implants can be largely recovered by annealing to 800 C (recovery is at least 94%, a recovery of 100% implies that the substrate is as damage free as a piece of unimplanted Si) and 900 C (recovery is at least 95%), whereas the damage in the Ga-FIB implants cannot be fully recovered for Ga doses beyond the solubility limits at the annealing temperatures used for this study. From the above it can be concluded that the higher ion current density (and therefore ion dose rate) in FIB implanters--relative to that of commercial broad beam implanters--and solubility of FIB ion species in Si can strongly affect the damage recovery of ion implanted Si. (Abstract shortened by UMI.).