Abstract
We compare experimental data for the saturated light‐induced defect density Nsat in hydrogenated amorphous silicon with results obtained by a quasi‐equilibrium model. If the model draws on a limited pool of defects the results agree with the experimental data, but if the model relies on the conversion of valence‐band‐tail states they do not. The model reproduces all three regimes of Nsat: a constant, maximum value of Nsat at high carrier generation rate G and low‐temperature T; the dependence of Nsat on both G and T at intermediate temperature; and independence of G coupled with dependence on T at high temperature.