Computational methods for fundamental studies of plasma processes

We present a combination of a wide range of computational methods that permits us to perform in-depth numerical studies of processes taking place in silicon/hydrogen plasma reactors during the fabrication of solar cells by means of Plasma Enhanced Chemical Vapor Deposition (PECVD). Notably, our investigations are motivated by the question under which plasma conditions hydrogenated silicon SinHm (n <= 20) clusters become amorphous or crystalline. A crystalline structure of those nanoparticles is crucial, for example, for the electrical properties and stability of polymorphous solar cells. First, we use fluid dynamics model calculations to characterize the experimentally employed hydrogen/silane plasmas. The resulting relative densities for all plasma radicals, their temperatures, and their collision interval times are then used as input data for detailed semi-empirical molecular dynamics (MD) simulations. As a result, the growth dynamics of nanometric hydrogenated silicon SinHm clusters is simulated starting out from the collision of individual SiHx (x=1-3) radicals under the plasma conditions derived above. We demonstrate how the plasma conditions determine the amorphous or crystalline character of the forming nanoparticles. Finally, we show a preliminary absorption spectrum based on ab initio time-dependent density-functional theory (DFT) calculations for a crystalline cluster to demonstrate the possibility to monitor cluster growth in situ.