The electronic structure and conductivity of large models of amorphous silicon

Holender, J. M.; Morgan, G J

doi:10.1088/0953-8984/4/18/013

Cited by 32 publications

(23 citation statements)

References 27 publications

(2 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such an approach certainly has limitations: for low temperatures in which one expects variable range hopping between defects the correct link to first principle simulation would seem to require solutions of the time-dependent Kohn-Sham equations 14,15 or an approach more akin to a Miller-Abrahams 13 model of the conductivity. There are several studies which use the KGF to compute the static lattice conductivity of amorphous materials 16,17,18 . The computed conductivity vanishes for localized states in a static lattice.…”

Section: A Overviewmentioning

confidence: 99%

Ab initioestimate of temperature dependence of electrical conductivity in a model amorphous material: Hydrogenated amorphous silicon

2007

View full text Add to dashboard Cite

We present an ab initio calculation of the DC conductivity of amorphous silicon and hydrogenated amorphous silicon. The Kubo-Greenwood formula is used to obtain the DC conductivity, by thermal averaging over extended dynamical simulation. Its application to disordered solids is discussed. The conductivity is computed for a wide range of temperatures and doping is explored in a naive way by shifting the Fermi level. We observed the Meyer-Neldel rule for the electrical conductivity with EMNR=0.06 eV and a temperature coefficient of resistance close to experiment for a-Si:H. In general, experimental trends are reproduced by these calculations, and this suggests the possible utility of the approach for modeling carrier transport in other disordered systems.

show abstract

Section: A Overviewmentioning

confidence: 99%

Ab initioestimate of temperature dependence of electrical conductivity in a model amorphous material: Hydrogenated amorphous silicon

2007

View full text Add to dashboard Cite

show abstract

“…Calculations of the electronic properties of a-Si based on a CRN have been performed by several authors. [18][19][20][21] This type of structural model is not able to reproduce the DOS in the vicinity of the Fermi level because it leads to a true gap between the two energy bands and this is not consistent with experimental results. Nevertheless a CRN is a simple model to study the influence of topological disorder-i.e., the presence of five-and seven-membered rings in addition to sixmembered rings characterizing the crystalline state -on the electronic properties of covalently bonded amorphous systems.…”

Section: Introductionmentioning

confidence: 99%

“…19,21 States around the Fermi level are usually believed to be localized. This disorder-induced Anderson localization leads to the formation of a mobility gap containing eigenstates not participating in electronic transport at zero Kelvin.…”

Section: Introductionmentioning

confidence: 99%

Defects ina−Sianda−Si:H: A numerical study

1998

View full text Add to dashboard Cite

We present a numerical study of the electronic properties of various structural models of amorphous silicon and hydrogenated amorphous silicon. Starting from an ideal random network, dangling bonds, floating bonds, double bonds, microvoids, hydrogenated dangling bonds, and hydrogenated floating bonds are introduced. The concentrations of these defects can be varied independently, the amount of disorder introduced to the system is therefore strictly controllable. Two continuous random networks, the vacancy model of Duffy, Boudreaux, and Polk and the bond switching model of Wooten, Winer, and Weaire ͑WWW model͒ are investigated. For the relaxation of the structures the potentials of Keating and of Stillinger and Weber are employed. The electronic structure is described by a tight-binding Hamiltonian; the localized or extended character of the eigenstates is investigated via a scaling approach. The vacancy model shows a band gap for small defect concentrations but this fills up with increasing disorder. Similar behavior is found for the case of the other models. In general defects introduce states into the gap region of a-Si, where the dangling bonds lead to the largest density of states in the gap region for a given defect concentration. This model turns out to be unique. For small system sizes an impurity band results that dramatically changes its character for large systems above 4000 atoms to a nearly uniform density of states as observed experimentally. In a-Si:H the dangling and floating bonds are removed and a mobility gap results with a width in good agreement with experiment. The experimentally observed tailing of the band into the gap region ͑first linear, then exponential͒ is well described only for the a-Si:H model derived from the vacancy model and for very large system sizes above 4000 atoms. The WWW model does not lead to this tail behavior. Localized states are found at all band edges but states at the bottom of the conduction band are more strongly localized than those at the top of the valence band.

show abstract

“…However, in the present calculations there is clear-cut behaviour and in figs. 1, 2 and 3 we show the density of states, oxx and oxy, respectively, as functions of energy E. The density of states and conductivity ox, are calculated as described by Holender and Morgan (1992). The curve for oxy is reasonably smooth and, although the error bars are larger than we would like, there is a clear change from negative values to positive values which then swing negative again at higher energies.…”

mentioning

confidence: 88%

“…We have used both a plane-wave basis for describing the eigenstates (Hickey and Morgan 1986) and a simple tight-binding model due to Chadi (1 984) which has enabled us to investigate very large structures which have been generated using the molecular dynamics method Morgan 1991, 1992).…”

mentioning

confidence: 99%