Carbon as a constituent of some of the source molecules is always present in the metalorganic growth techniques and can potentially be incorporated into the grown layers. Unintentional carbon incorporation as well as intentional carbon doping of III-V compounds has been studied intensely, especially in the case of GaAs. However, a number of experimental findings still is not well understood. Besides the different electrical behavior of carbon in different host materials (acceptor in GaAs, donor in InAs) also the different probability of incorporation into, for example, GaAs compared to InAs is currently not well explained. Model calculations may provide useful hints to enhance the understanding of the experimentally observed trends. In this paper, two different approaches will be reviewed and their results will be discussed.
The first approach is dealing with the expected incorporation site of carbon in III-V“s. The change in total energy associated with substitutional carbon incorporation onto group III or group V lattice sites has been calculated. From this change the conduction type of carbon doped arsenides and phosphides (except for InAs) can correctly be predicted. While this method is dealing with the bulk semiconductor, the second approach attempts to model the binding energy of methyl to atoms on the growing surface. By using appropriate molecules the methyl bond strength can be estimated. Both approaches together provide a basis for a more complete understanding of carbon incorporation behavior in III-V compounds.