No abstract

No abstract

An extension of the model proposed by Stutzman, Jackson, and Tsai [Phys. Rev. B 32, 23 (1985)] for the kinetics of light-induced metastable defect creation in a-Si:H is presented. We assume a limited density of defect centers and take into account the attenuation of the incident light into the sample. The experimental data of the work mentioned above are well fitted with our modified model. This fit allows for a quantitative estimation of the number of defect centers related to the Staebler-Wronski effect.Many experimental works have been done in order to clarify the origin of the Staebler-Wronski effect (SWE) (Refs. 1 -8), in a-Si:H. Among them, one of the most complete sets of measurements is that presented, in recent papers, ' by Stutzmann, Jackson, and Tsai (SJT). According to the SJT microscopic model for the SWE, the light-induced defects are assumed to be single dangling bonds, and consequently the density of light-induced metastable defects can be measured through the induced spin density. The kinetic model proposed by SJT (Refs. 7 and 8) predicts that the density of light-induced defects is a sublinear function of illumination time and light intensity. This prediction is in agreement with their experimental data for the time evolution of spin density and photoconductivity. ' However, this model is not able to predict the saturation behavior physically expected at long times. As mentioned by SJT, saturation could take place due to the depletion of accessible metastable sites. In spite of the fact that the typical increase in the spin density (= 10' cm ) measured by different authors is well below the value estimated for saturation (-=10' cm ), it is not obvious whether or not the existence of a limited number of centers could affect the kinetics of spin generation for relatively short times of exposure.In this work, we have added two hypotheses to the SJT kinetic model. (a) In the annealed state, only a limited number N of defect centers are able to be transformed into metastable centers by illumination. Thus, the limiting condition is not only that the induced defects provide alternative paths for electron-hole recombination, but also that the number of centers available is depleted with time. (b) The N centers available in the annealed state are homogeneously distributed through the sample. Light attenuation into the sample is taken into account, instead of assuming a two-phase model as proposed by SJT. These hypotheses allow for a good fitting of the SJT experimental data, which in turn provides a quantitative estimation for N, a useful magnitude to investigate in order to clarify the microscopic origin of the SWE and to know more about the material structure.Starting from the equation given by SJT for the spingeneration rate, and assuming that this rate is proportional to the number of centers available to be -5.I E CJ 4.I Eh 3 a 40 80 I20 I 6 0 200 240 ILluminotion Time (min ) FIG. 1. Effective dangling-bond density as a function of illumination time, for various light intensities. The experimen...

We measured the effects of light soaking on the extended state electron mobility in intrinsic and n-type doped hydrogenated amorphous silicon samples. We obtained the temperature dependence of the mobility in the range 0–80 °C, using a recently proposed method [Dawson et al., Appl. Phys. Lett. 63, 955 (1993)]. We found a decrease of the mobility as the degree of light-induced degradation increased. We suggest that these changes in the extended state transport are caused by an enhancement in the magnitude of the potential fluctuations introduced by the extra created charged defects.