A graphical peak analysis method for characterizing impurities in SiC, GaN and diamond from temperature-dependent majority-carrier concentration

A method for uniquely determining the densities and energy levels of impurities from the temperature dependence of the majority-carrier concentration in wide band gap semiconductors (e.g., SiC, GaN, and diamond) is discussed. It is demonstrated that the proposed graphical peak analysis method can evaluate the number of impurity species and can determine those densities and energy levels uniquely and accurately, while fitting a simulation to the experimental temperature-dependent majority-carrier concentration leads to less reliable densities and energy levels of impurities. In the case that the Fermi levels in p-type SiC, GaN and diamond are located between the acceptor level and the valence band maximum, the excited states of acceptors strongly affect the hole concentration. This indicates the distribution function including the influence of the excited states should be applied to determine the densities and energy levels of acceptors from the temperature-dependent hole concentration.