Normal and anti Meyer–Neldel rule in conductivity of highly crystallized undoped microcrystalline silicon films

Abstract

We have studied the electrical conductivity behavior of highly crystallized undoped hydrogenated microcrystalline silicon (lc-Si:H) films having different microstructures. The dark conductivity is seen to follow Meyer-Neldel rule (MNR) in some films and anti-MNR in others, which has been explained on the basis of variation in the film microstructure and the corresponding changes in the effective density of states distributions. A band tail transport and statistical shift of Fermi level are used to explain the origin of MNR as well as anti-MNR in our samples. The observation of MNR and anti-MNR in electrical transport behavior of lc-Si:H is discussed in terms of the basic underlying physics of their origin and the significance of these relationships.

The statistical shift of the chemical potential causing anomalous conductivity in hydrogenated microcrystalline silicon

“…The behavior of the activation energy and 0 as a function of crystallinity within the model is in qualitative agreement with experimental observations, like those of Kočka et al 10 and Ram et al, 28 if we assume that the condition , which is lower than reported oxygen concentrations in some device grade materials. 11 The defect density derived in this way is 10 16 cm −3 , which is a typical value for spin densities in ESR experiments, 31 but we have to keep in mind that the spread in properties of these materials can be large and it is suggested that high spin densities at high X c are related to surface effects.…”

“…For X c Ͼ X s 0 increases again, presumably because of a fast growing doping/dangling bond ratio. In experiments of Kočka et al 10 and Ram et al 28 E a is also decreasing for X c Ͼ X s , which leads to the experimentally observed inverse Meyer-Neldel rule ͑MNR͒. The lower E a could be explained by a decrease in the a-Si transport barriers between the nanocrystals.…”

“…The very high values of 0 ͑Ͼ1000 ⍀ −1 cm −1 ͒ are found ͑in Fig. 9͒ only for the data of Ram et al 28 The reason for this is not known. It is very remarkable that there is an increase in 0 at low E a for high N d .…”

“…The higher the crystallinity the larger the drop in 0 , which is in agreement with experimental results. Ram et al 28 records a drop in 0 of four orders of magnitude at a crystalline ratio of close to one, while Kočka et al 10 observe a drop of three orders of magnitude at a crystallinity of 0.9.…”

“…However, it seems possible that there are constraints in the relation between 0 and E a that are not obvious from the model, leading to anti-MNR behavior or an inverse MNR. Ram et al 28 report an anti-MNR for sample with very high crystallinity.…”