An image processing approach to approximating interface textures of microcrystalline silicon layers grown on existing aluminum-doped zinc oxide textures.

We present an algorithm for generating a surface approximation of microcrystalline silicon (μc-Si) layers after plasma enhanced chemical vapor deposition (PECVD) onto surface textured substrates, where data of the textured substrate surface are available as input. We utilize mathematical image processing tools and combine them with an ellipsoid generator approach. The presented algorithm has been tuned for use in thin-film silicon solar cell applications, where textured surfaces are used to improve light trapping. We demonstrate the feasibility of this method by means of optical simulations of generated surface textures, comparing them to simulations of measured atomic force microscopy (AFM) scan data of both Aluminum-doped zinc oxide (AZO, a transparent and conductive material) and μc-Si layers.