| dc.description.abstract | Materials of high purity are required for the production of crystalline silicon for solar cells. Impurities are indeed very detrimental to solar cell efficiency. Contamination can occur at many stages during the processing, such as directional solidification of silicon ingots. Directional solidification is performed in quartz crucibles coated with silicon nitride. The impurity transport between the crucible, coating and silicon is not well understood and is difficult to predict. This thesis attempts to evaluate the diffusivity and segregation coefficient of the impurities, when transported through the crucible, coating, solid silicon and liquid silicon during directional solidification of multicrystalline silicon. The quantification of these parameters will allow for a better understanding of how contamination from the crucible and coating occurs, and thus contribute in the design of higher purity silicon solar cells.
Three sets of samples have been analyzed; coated quartz samples, coated silicon wafers and liquid silicon on a silicon nitride substrate. The samples were heat treated at different temperatures and for different durations, after which the chemical composition was analyzed. The coated quartz samples and the coated silicon wafers were analyzed with inductively coupled plasma mass spectroscopy (ICP-MS). The liquid silicon and silicon nitride samples were analyzed by an electron probe microanalyzer (EPMA). A one-dimensional model was developed using the finite element method to simulate the transport of impurities across the interfaces.
High contamination levels in the acid solutions used for the ICP-MS sample preparation lead to high background noise, thus the impurity concentration for the coated quartz and coated silicon samples was underestimated. Nevertheless, data for chromium were suitable to be used. The finite element model was fit to the experimental data and was used to predict the diffusivity and segregation coefficient of chromium. The segregation ratio of chromium between silicon and silicon nitride and the diffusivity of chromium in silicon nitride were found. The sensitivity of the EPMA appeared to be too low to detect contamination.
Even though the results did not allow to estimate the transport mechanisms of several impurities initially aimed, the method developed in this thesis is good and can be further used with some refinement. Improvements should be made to the experimental method in order to obtain more reliable results, and this is discussed with some details in the thesis. Finally, it is expected that the results of this thesis will be used in a national project and can have positive impact in the PV industry. | en |
