Numerical simulation of metallic impurity distribution in MC-SI crystal growth and subsequent annealing
B. Böttger, I. Steinbach, S. G. Fries, Q. Chen, B. Sundman, D. Franke, M. Apel.
2nd World Conference and Exhibition on Photovoltaic Solar Energy Conversion, 1347-1350, (1998)
Nowadays, metallic impurities are no limiting factor for the minority carrier diffusion length in cast multicrystalline silicon. Diffusion length above 300 µm are frequently observed. However metallic impurities have the ability to enhance the recombination activity of dislocations, even in low concentrations. Nevertheless the recombination activity of dislocations in as grown wafers are generally low, but can increase after further high temperature processing. One possible explanation for the increase is the change of the distribution of trace metallic impurities during annealing. For this reason knowledge about the distribution of trace metallic impurities is helpful to find strategies for a further reduction of their influence. Experimental investigations of the metal distribution are difficult because of their low concentrations. Therefore the aim of our work is to study the distribution of trace metallic impurities by numerical simulations. In this paper we will discuss our calculation model, that combines impurity diffusion and segregation with heterogeneous precipitation in a multicrystalline structure. We have regarded the iron impurities as an example. Calculation results for the precipitation behavior of iron are given for different cooling rates. Low cooling rates in combination with heterogeneous precipitation decrease the iron concentration in large areas. This can reduce the precipitation density and the recombination activity of dislocations. Further results show the influence of precipitate dissolution on the efficiency of external gettering.
Keyword(s): multi-crystalline; silicon; simulation