Abstract
Today, silicon-wafer-based technology dominates the photovoltaic (PV) industry because it enables high efficiency, is produced from abundant, non-toxic materials and is proven in the PV marketplace.[1] However, costs associated with the wafer itself limit ultimate cost reductions.[1,2] PV based on absorber layers of crystalline Si with only 2 to 10 µm thickness are a promising route to reduce these costs, while maintaining efficiencies above 15%.[3-5] With the goal of fabricating low-cost film crystalline Si (c-Si), recent research has explored wafer peeling,[6,7] crystallization of amorphous silicon films on glass,[4,8-10] and “seed and epitaxy” approaches.[3,5,11] In this third approach, one initially forms a seed layer that establishes the grain size and crystalline order. The Si layer is then grown heteroepitaxially on the seed layer, so that it replicates the seed crystal structure. In all of these film c-Si approaches, the critical challenge is to grow c-Si with adequate material quality: specifically, the diffusion length (LD) must be at least three times the film thickness.[12] In polycrystalline Si films, grain boundaries (GBs) are recombination-active and significantly reduce LD. This adverse effects of GBs motivates research into growth of large grained c-Si [13,14] (for a low density of GBs) and biaxially-textured c-Si [11] (for low-angle GBs).
