Dependence of impact ionization coefficients on electric-field profile and injection mode approximations

Abstract

The validity of the determination of the ionization coefficients deduced from the experimental measurement of photocurrent multiplication Mph in GaAlSb diodes, in which the electric-field profile is nonlinear and successive pure hole and pure electron injections cannot be made, are compared and discussed. Three methods for determining ionization coefficients kp and kn are discussed: (i) adjusting kp and kn in order to fit the experimentally observed Mph values using an approximated linear electric field; (ii) the same method applied to a more realistic electric-field profile; (iii) the Grant method in which kp and kn are calculated from the results of multiplication measurements in the case of pure injections and using this last electric-field profile. It is concluded that when the condition of pure injection is not fulfilled the Grant method may lead to erroneous k values and the adjustment method is more valid.