| Boron-doped p-type silicon is the industry standard silicon solar cell substrate. However, it has serious limitations: iron boron (Fe-B) pairs and light induced degradation (LID). To suppress LID, the replacement of boron by gallium as a p-type dopant has been proposed. Although this eliminates B-O related defects, gallium-related pairing with iron, oxygen, and carbon can reduce lifetime in this material. In addition resistivity variations are more pronounced in gallium doped ingots, however Continuous-Czochralski (c-Cz) growth technologies are being developed to overcome this problem. In this work lifetime limiting factors and resistivity variations have been investigated in this material. The radial and axial variations of electrically active defects were observed using deep level transient spectroscopy (DLTS) these have been correlated to lifetime and resistivity variations. The DLTS measurements demonstrated that iron-related pairs are responsible for the lifetime variations. Specifically, Fe-Ga pairs were found to be important recombination sites and are more detrimental to lifetime than Fei.;Typically n-type silicon has a higher minority carrier lifetime than p-type silicon with similar levels of contamination. That is because n-type silicon is more tolerant to metallic impurities, especially Fe. Also, it has no serious issues in relation to lifetime degradation, such as FeB pairs and light-induced degradation (LID). However, surface passivation of the p + region in p+n solar cells is much more problematic than the n+p case where silicon nitride provides very effective passivation of the cell. SiO2 is the most effective passivation for n type surfaces, but it does not work well on B-doped surfaces, resulting in inadequate performance. Al2O3 passivation layer suggested for B-doped emitters. With this surface passivation layer a 23.2 % conversion efficiency has been achieved. After this discovery n-type silicon is now being seriously considered for photovoltaics. The detrimental effects and electrical properties of transition-metal impurities, e.g., iron, and its complexes, such as FeB, in p-type silicon are well-known. However, in n-type silicon wafers, although there is evidence of greater tolerance, the impact of specific metallic impurities on minority lifetimes is not as well established.;The major contribution of this dissertation is to provide new electrical data relating to metallic impurities in n-type silicon, e.g. activation energy, capture cross sections. The injection-dependent lifetimes of intentionally metal contaminated n-type CZ silicon wafers were investigated using resonant-coupled photoconductance decay (RCPCD) and the quasi-steady-state photoconductance technique (QSSPC). Finally, a direct correlation between minority carrier lifetime and the concentration of specific electrically active metallic impurities was established using DLTS. |
