Gettering of impurities in silicon: delivering quantitative understanding to improve photovoltaics

点击翻译按钮获取中文摘要

Photovoltaics have the potential to supply all the world's energy needs. The market for photovoltaics is dominated by cells made from crystalline silicon, which account for more than 80% of today's production. Whilst other technologies are being researched, silicon's abundance, chemical stability, density, band gap and non-toxic nature mean that is certain to play a leading role in at least the medium term. More than half of bulk silicon solar cells are fabricated from multicrystalline silicon (mc-Si) wafers. Although mc-Si photovoltaics have lower efficiencies than their single-crystal counterparts, their substantially lower production costs means the technologies have equal commercial viability at present. Mc-Si is produced by casting, often using a low grade feedstock, and is consequently packed with extended defects (dislocations, grain boundaries and precipitates) and transition metal impurity point defects. Recombination of photogenerated charge carriers at such defects is a major reason for the reduced efficiency of mc-Si cells. Gettering processes are routinely used either to redistribute the defects or remove them from the material. However, such processes are not completely effective. One of the major reasons for this is that the interaction between defects prevents them being gettered. This project aims to further the fundamental understanding of defect interactions in mc-Si. The thermodynamics of interactions between transition metals (particularly iron) and extended defects (particularly dislocations and oxide precipitates) will be studied experimentally. Passivation of key extended defects will also be investigated. The fundamental knowledge obtained should allow the development of new or modified gettering processes with the ultimate aim of facilitating the use of dirtier (hence cheaper) feedstocks for silicon photovoltaics.