The use of heterojunctions between hydrogenated amorphous silicon (a-Si:H) and crystalline silicon (c-Si) has attracted much attention in recent years for the fabrication of solar cells with efficiencies up to 23%. Furthermore, the possibility of doping a-Si and a-Si:H and their lower production cost with respect to c-Si have made them the materials of choice for many different applications in optoelectronics and photovoltaics. Yet, these promising perspectives are highly limited by the large defect density present in bulk a-Si and at the a-Si:H/c-Si interfaces, which reduces the doping efficiency and increases the interface carrier recombination losses in solar cells.
We have used ab initio simulations to study the relevant defects states in a-Si and at its interface with c-Si, and how these defects interact with charge carriers. We have identified intrinsic hole traps in a-Si associated to locally strained regions, and we have analyzed their interaction with boron atoms. We have found that the low doping efficiency in the case of B is an intrinsic property of amorphous silicon since, even if it is well relaxed, locally strained regions exist. This fact limits the application of amorphous silicon in devices that require higher carrier densities.
DFT simulations have shown that the hole spatial localization in a-Si:B (and therefore the dopping efficiency) is highly influenced by the presence locally strained regions. These strained regions induce around them the spatial localization of holes, independently of the position and concentration of B atoms (a, b, c), and are inherent to the a-Si matrix (I. Santos et al., Phys. Rev. B 81, 033203 (2010) - Self-trapping in B-doped amorphous Si: Intrinsic origin of low aceptor efficiency).