| The fundamental research addressed in this thesis explores the role of individual defects and impurities in determining the multicrystalline silicon solar cell performance. This approach contrasts that of previous investigations which determined the macroscopic effects of unidentified defects and impurities on the silicon solar cell performance. As part of this research, three tools were developed that helps facilitate the investigation of the impact of individual defects and impurities on silicon solar cell performance. A high energy resolution deep level transient spectroscopy (HER-DLTS) system was developed which, for the first time, provided a means of examining point defects in highly disordered materials such as multicrystalline silicon. This system was used to show that even though an ample amount of hydrogen was introduced using a plasma enhanced chemical vapor deposition (PECVD) system the hydrogen was ineffective in passivating nonmetallic point defects such as oxygen-vacancy and divacancy complexes, as well as, metallic point defects, such as, titanium and vanadium. The HER-DLTS system was also used to show that aluminum treatments getter grown in metallic impurities and that the gettering mechanism is diffusion limited. A second tool, an electron beam induced current (EBIC) system, was developed to examine the recombination properties of extended defects in multicrystalline silicon. This system was used to show that the presence of moderate amounts of oxygen at grain boundaries and dislocations reduces the surfaces recombination velocity of these defects, as well as, the injection level dependence of the surface recombination velocity compared to oxygen lean extended defects. Furthermore, it was found that the PECVD/RTA introduced hydrogen is very effective in passivating oxygen lean grain boundaries, but is much less effective in passivating oxygen rich extended defects. These observations combined with the third tool developed in this research, a grain boundary recombination model, facilitated the development of a theoretical model for describing the difference in density of interface state profiles of oxygen rich versus oxygen lean grain boundaries. The fundamental knowledge obtained in this thesis has been applied to provide guidelines for cost-effective multicrystalline silicon solar cell production. |
