A Study Of Thermal Engineering Of Photovoltaic Crystalline Silicon

Solar photovoltaic power is an important solution to the energy and environment crisis of the world. Silicon wafer-based solar cells occupy about88%of the cells installed in solar power systems. Thermal engineering of crystalline silicon refers to enhancing photovoltaic performance of the material through thermal treatments or optimizing thermal processes in fabrication of silicon wafer-based solar cells. It is a potential route to higher energy conversion efficiency at low cost.In the present thesis, effects of thermal parameters on the state and distribution of impurities and defect densities in crystalline silicon, and hence on electrical properties of crystalline silicon, are investigated systematically. Both Cz type mono-crystalline silicon and directionally solidified cast multi-crystalline silicon are involved. The following major results are obtained.1) It has been found that, in continuous cooling of cast multi-crystalline silicon, the dissolved interstitial oxygen has very strong tendency to precipitate, while carbon does not. The former can precipitate even in cooling as fast as10℃/s, while the latter basically does not precipitate even in cooling as slow as0.017℃/s. Calculation shows that the reason for difficulty of carbon precipitation is the high activation energy for diffusion of carbon in crystalline silicon, and its extremely low diffusion rate.2) It has been found that, in both the mono-crystalline silicon and the multi-crystalline silicon, precipitation or dissolving of iron impurity is very sensitive to thermal processes, which worth close attentions. Concentration of dissolved interstitial iron in the two crystalline silicon materials of as-solidified state remarkably increases after heating to300～1050℃followed by a rapid cooling to ambient temperature. The higher the heating temperature, the higher the iron concentration in the rapidly cooled material. For heating to a certain temperature, slower cooling leads to lower concentration of interstitial iron. In both crystalline silicon materials,900～1050℃heated followed by50℃/s quenching to ambient temperature, over90%of iron exists in state of precipitates, with rest in dissolved interstitial state. For multi-crystalline silicon, the supersaturated interstitial silicon will precipitate again after a heating followed by slow cooling procedure, leading to remarkable decrease of its concentration. Higher heating temperature leads to lower level of interstitial iron concentration after slow cooling, down to the as-cast level when the heating temperature reaches900℃.3) Dislocation density in0.2mm thick multi-crystalline wafers significantly decreases after annealing at1320℃or above followed by slow cooling. For instance, in the sample annealed at1340℃for2hours followed by cooling at0.12℃/s, about50%decrease of dislocation density was found. At high temperatures, it is found that generation/disappearing of dislocations in crystalline silicon is very sensitive to thermal stress, and hence to cooling condition and thickness of materials--in the wafers cooled at slightly higher rate (greater than0.13℃/s) after annealing at1320℃/s or above, the dislocation density increased, and in13mm thick sample of multi-crystalline silicon block, increase of dislocation density appeared even when the cooling is as slow as0.03℃/s after the same kind of annealing. It is further shown that annealing at1100℃/s or lower followed by slow cooling results in increase of dislocation density, rather than decrease.4) The concept of thermal degradation of minor carrier lifetime is proposed and supported by experimental studies. Both mono-crystalline silicon and multi-crystalline silicon in their as-solidified state are found to have decreased minor carrier lifetime after300～1050℃heating followed by rapid quench to ambient temperature. Higher the heating temperature, shorter the lifetime after quench. With the same heating temperature, the lifetime decreases as the cooling rate increases. The thermally degraded multi-crystalline silicon can be recovered to various extent by heating followed by slow cooling, with the maximum recovery extent achieved at900℃heating,～92%of the as-solidified multi-crystalline silicon. It was also found that, the thermally degraded multi-crystalline silicon by fast heating-quenching showed better resistance to high temperature thermal degradation than the as-solidified multi-crystalline silicon.5) Based on the phenomena of thermal degradation/recovery, and their correlation with the evolution of the state and distribution of iron, oxygen, carbon, and dislocation density in thermal processes, it is proposed that the major mechanism for thermal degradation is dissolving of the precipitates of iron and other similar metal impurities, which releases iron and other metals, resulting in supersaturated iron and other metal impurities, or fine scattered secondary precipitates in fast cooling. They act as recombination centers of the minor carriers and cause remarkable decrease of the lifetimes. The major mechanism for the recovery of the thermally degraded silicon is then the precipitation of supersaturated iron and other similar metal impurities, or dissolving of the fine precipitates followed by secondary precipitation forming larger size precipitates.6) Thermal donors formed in Cz mono-crystalline silicon can be effectively eliminated by a40min annealing at650℃, which make the measured minor carrier lifetime and resistivity recovered to their real value. In the subsequent heating to high temperature followed by slow cooling, the thermal donors form again, with its concentration increasing with decrease of cooling rate. The re-formation of thermal donors can be avoided with fast cooling. It is found that if the material is slowly cooled to550℃, and then cool quickly, formation of thermal donors can also be avoided. This discovery can be utilized to solve the problem of contradiction between the thermal donor elimination and thermal degradation of minor carrier lifetimes.The results presented above are helpful for optimization of thermal processes in fabrication of solar cells from crystalline silicon wafers, and for designing of added heat treatment of silicon wafers, which in turn are of great value to increase of energy conversion efficiency of crystalline silicon wafer-based solar cells. Part of the results, after publication, has already been used in solar cell manufacturing industry.