CVD a-Si has been given less attention than glow discharge a-Si. The reason is the large density of native dangling-bond defects which make it less suitable for electronic applications. This situation may be modified in the future by use of the post-hydrogenation technique and/or extension of the low temperature deposition methods, using higher silanes or HOMOCVD. In the long term, specific advantages of CVD (or LPCVD) systems for designing large production systems may become important. In the near term, it is not likely to be a strong competitor for those applications which require the lowest density of gap states, e.g., solar cells. By other criterion, however, the material appears to be superior, for instance, in the ability to obtain high conductivity n+ or p+ layers. It may find its niche in specific applications.
From a fundamental point of view it can be considered as a reference material, relatively independent of preparation details, and a good vehicle to investigate basic questions concerning a-Si. In this respect, the problem of the relationship between doping and defect formation appears to be central at the time of writing and deserves further investigations.
Another question is that of the microstructure. Perhaps the most striking observation for CVD material is the fact that hydrogen diffusion coefficients appear to be independent of hydrogen content and similar to those observed in glow discharge samples. This gives a strong presumption that the hydrogen microstructures, e.g., clustered and diluted regions, do not depend on whether the hydrogen is grown into the film as in glow discharge deposition or introduced a posteriori as in plasma post-hydrogenated CVD films. This is an incentive to pursue the idea that these microstructures are intrinsic to the amorphous network and a basic property of tetrahedral materials.
Electron Spin Resonance Chemical Vapor Deposition Glow Discharge Hydrogen Content Electron Spin Resonance Signal
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.