Dislocations

In previous sections, we have discussed how to decrease excess carrier recombination by using various kinds of gettering of transition-metal impurities in dislocation-free crystalline-Si-solar cells. In this case, present understanding of gettering processes has reached a quantitative level that allows predictive computer simulations of gettering [13, 20, 21, 66, 73]. Such gettering simulations serve as tools to optimize gettering conditions,
e. g., temperature-time schemes suitable to achieve a certain diffusion length and obeying the requirements of typical solar-cell manufacturing.

A considerable fraction of present-day silicon photovoltaics, however, uses cheaper multicrystalline silicon materials produced by casting or other techniques [90]. Such materials inherently contain grain boundaries and dislocations that are inhomogeneously distributed and whose density may considerably vary in different grains, typically from 102 to 107 cm-2. Obviously, the presence of dislocations should influence kinetics and efficiency of gettering processes and, also the minority carrier diffusion length. It is well known that dislocations often act as efficient recombination centers [148].

However, it is also well known that the effect of dislocations on the minority carrier diffusion length is significantly influenced by processing conditions and can be strongly modified by gettering. This has been taken as evidence that recombination at dislocations is not governed by intrinsic properties but mainly governed by interaction of dislocations with point defects of intrinsic or extrinsic origin. In order to understand the specifics of gettering and to extend gettering simulations to dislocated Si solar-cell materials we have to – as far as possible – quantitatively model excess carrier recombination at dislocations and its dependence on thermal history and impurity contamination. In addition, we need to know how dislocations influence gettering processes and the behavior of transition metal impurities. Hence, to make progress on the problem we must understand at least (i) the mechanism of e-h recombination at dislocations, (ii) the electronic properties of impurity atoms and precipitates at dislocations, and (iii) the interaction of impurity atoms and intrinsic point defects with dislocations, including their precipitation at dislocations.