This thesis investigates the relationship between the structure of silicon (100) twist boundaries and their electrical activity. A long standing question is whether the localized defect states which cause the electrical activity are intrinsic to the grain boundary structure, i.e. related to dangling or distorted bonds, or extrinsic, resulting from impurities or stray dislocations.; A series of bicrystals was produced, with misorientation angles spanning the range from {dollar}sim{dollar}0{dollar}spcirc{dollar} to 40{dollar}spcirc{dollar}, by hot-pressing single crystals under ultra-high vacuum conditions. Measurements were made of the transport across the potential barriers in our series of bicrystals. Analysis using a combined drift-diffusion and thermionic emission transport model revealed a small, consistent density of mono-energetic defect states near midgap, for all the boundaries. The integrated number of states was found to be too small (on the order of 3 {dollar}times{dollar} 10{dollar}sp{lcub}10{rcub}{dollar} cm{dollar}sp{lcub}-2{rcub}){dollar} to be related to the boundary dislocation structure. The consequent conclusion that the origin of the boundary electrical activity must therefore be extrinsic to the structure is supported by the lack of dependence on misorientation angle of the defect state densities. While this is consistent with what is known about symmetrical tilt boundaries, it has never before been explored for twist boundaries.; Evidence for an inhomogeneous distribution of charge in the boundary plane was found from the analysis. The transport model was modified to allow for the resulting fluctuations in the boundary potential barrier, but the modifications are shown not to affect the above conclusion.; Minority carrier properties of the twist boundaries were also explored in a subset of the bicrystals, using the electron beam induced current (EBIC) technique. These measurements reveal marked variations in contrast along the length of the boundary planes, suggesting that the misorientation angle (and therefore structure) does not determine the recombination rate. This further supports the conclusion drawn from the majority carrier measurements that the structure of the boundary is not responsible for its electrical activity.
展开▼
