Abstract
In a recent paper, Collins et al. [Phys. Rev. B 26, 6643 (1982)] have shown that a fast process, occurring in aSi:H, and so the slow process of tunneling to defects, in a time scale comparable to that of the radiative recombination, is not dominant, as previously believed from studies on glow-discharge material. We consider their results in terms of the distant-pair model, and we conclude that the fast and slow processes are in fact successive steps in the nonradiative mechanism. Also, we find that a capture cross section for the defects comparable to that of the tail states is consistent with the data.