Modeling the size distribution of iron silicide precipitates in multicrystalline silicon

Precipitation of iron in multicrystalline (mc) silicon during typical solar cell processing is investigated with an advanced model that has been previously demonstrated to be able to simulate the iron distribution in two dimensions. The improving detection limits of precipitate measurements with x-ray fluorescence microscopy (mu XRF) allow a first comparison of measured and simulated precipitate densities in mc solar cell material. We focused on simulating the size distribution of iron-silicide precipitates at a grain boundary for a better comparison with experimental data. The simulated line densities of detectable iron precipitates as well as the size distribution along a grain boundary are in good agreement with experiments. These results verify the iron kinetics model, independent of the interstitial iron concentration. The dependency of the assumed total iron concentration and the model structure is investigated by varying these input parameters. The model predicts a fast dissolution of small and medium iron precipitates during inline phosphorus diffusion, resulting in an increase of the mean size of precipitates. Low-temperature anneals effectively reduce the interstitial iron concentration, but influence only the density of small precipitates, which are below experimental detection limits. The simulations demonstrate that iron precipitates become more important for the charge-carrier lifetime during solar cell processing.