The deposition of with TEOS is a complex pyrolytical chemical reaction. In this process TEOS is transported via a carrier gas to the hot surface of the wafer, where TEOS is dissociated [25]. A certain amount of the decomposition products sticks on the surface and build a layer while the other particles are reflected from the surface. Those are in general highly reactive by-products of TEOS decomposition. In particular, more than 40 secondary reactions have been reported in this complex reaction [328]. A rigorous simulation would cover all possible by-products and their secondary and ternary reactions but it would also require a considerable amount of computational power and memory to calculate and investigate this TEOS reaction.
However, one of the industrial requirements is to provide the engineers with rather fast and sufficiently accurate simulation results. Therefore, this project has focused on developing certain models to predict the TEOS deposition for a certain series of test trenches, where the characteristic aspect ratio (AR) is used to determine or estimate the impact on the chemical reaction behavior.
In order to obtain quantitatively accurate simulation results for deposition processes rather complex chemical models are required to describe the chemical reactions mechanisms. The computational effort is too high to include these rigorous models to software tools for industrial use [329]. Therefore, simplified model have been developed to speed up the simulation time [330] to obtain industrial-ready simulators. However, these simplified models have to be calibrated for each particular deposition process separately. The overall goal for this project was to find an appropriate deposition model and a certain parameter set which can be applied to all trenches of these test series to sufficiently predict the shape of the TEOS deposition.