[Pw_forum] experimental Dit values versus DFT values

[hannu.komsa at epfl.ch]

Hello

Dit is the number of trapped charges at the interface divided by  
energy. See for example:
http://dx.doi.org/10.1016/S0038-1101(00)00119-2
You need to have electron/hole traps at the interface in order to have  
non-zero Dit and I guess it only makes sense to look at this within  
the band gap. It is not the density of interface states per se that  
originate just from the disruption of the bulk periodicity. The latter  
can be traps or they can be resonant with the bulk states or whatnot.  
Anyway, what you should do is to look at the DOS only within the band  
gap. If you have traps at band gap energies also within the bulk  
regions, then perhaps local DOS for some region near the interface can  
be sufficient. If this is indeed what you already did, then perhaps  
your model needs improvement. It should be possible to construct  
Si/SiO2 interface model without any trap/gap states, I think.


Regards,
   Hannu-Pekka Komsa
   EPFL


>   1. I was wondering whether the sum of all LDOS calculated from projected
>   wavefunction on the Si at the interface can give me the density of
> interface
>   states. The question is whether the sum of LDOS of projected wavefunction
> on
>   the inter-facial atoms is equivalent to planar local density of state at
> the
>   interface.
>   2. If positive, the sum of LDOS of all projected wavefunction on all
>   atoms at the interface divided by the surface of the interface is by far
>   higher than the Dit reported from experiment. Actually, I am thinking, no
>   matter how big the structure is, the Dit is still very high. I was
> wondering
>   whether I miss any point to consider for calculation.