Modelisation du transport par piegeage multiple de porteurs electroniques dans le silicium amorphe hydrogene.

摘要

We study, using the Monte Carlo (MC) method, multiple trapping (MT) controlled charge transport assuming an exponential band tail in hydrogenated amorphous silicon (a-Si:H). The primary goals of this research are (i) to deepen our knowledge of the microscopic mechanisms for this type of transport and (ii) to develop a unified model able to adequately reproduce the experimental data in all temperature, field and time regimes. First, we review the physics of field assisted detrapping, particularly a high field model, developed earlier. In order to rigorously include the effects of tunneling on the release rate Kr(E), we examine the sum of contributions from all the different tunneling paths Deltax. We apply this idea to the model. We suggest the possibility that tunneling may contribute to carrier trapping. By applying the condition of detailed balance, we find that we may include the effects of tunneling on the capture rate kc(E) by properly defining the Delta x dependent release rate.; We develop a unified model based on MT transport including the Meyer-Neldel effect and field assisted detrapping. By simulating time-of-flight experiments using the MC method, we show that this new model, which includes five fixed parameters, satisfactorily reproduces the temperature and field behavior of all the empirical results. These include the pre-transit dispersion parameter alpha 1 and alternative definitions of the mobility mu. We observe a contradiction between drift mobility muD mesurements made by different laboratories. The data suggests that this inconsistency is due to differences between the samples or to inherent difficulties associated with high field measurements.; We find that our model is able, with a slight modification of the optimal parameters found previously, to reproduce the experimental data measured at all times longer than one picosecond. This shows that MT controlled transport is dominant at short (t ∼ 1 ps) and long (t > 1 ns) times. We also show that picosecond domain data are compatible with the traditional nanosecond domain data. Our calculations suggest that trap-free transport could be experimentally observable with an improvement in time resolution of one order of magnitude. Our model is the first, to our knowledge, to produce predictions in agreement with the experimental results for all temperature, field and time regimes studied to date.