[Wien] Anyone done Silicium yet?

[Gilles Hug]

On Sat, 10 Jan 2004, Michael Frotscher wrote:

> Hello, Wien-ML
>
> > Yes Wien (DFT-LDA)  calculates only the ground state. I you want the
> > correct gap you
> > need to calculate correctly the excited states. Much more complicated and
> > more sophisticated is required..
> >
> > I got .56 eV approx with optimized cell (10.5202 a. u.)  and 2000 kpts.
>
> I thought it might be something like that, although I learned that the
> allocation of excited states is certainly temperature-dependent, the
> gap-energy is not. So that should be the same at absolute zero (which is
> calculated by WIEN, if I'm correct here) or at room temperature.

Temperature has nothing to do here. Well it might change a little bit at
Fermi level the distribution of electron but this is extremely small.
The problem is that the width of the gap involves conduction bands and
that these are not calculated correctly by Wien. Wien calculates the
ground state in which all electrons are in the valaence band. To measure
the gap on need  to promote an electron in the conduction band. This
exited state has a hole (electron missing) in the valence band and it is
not stable : after some life time the electron which was promoted to
excited falls back in CB. The hamiltonian for that is more complicated, it
is not the one Wien solves. You can simulate the excited wih supercells
but it is an approxiamtion.
Best
Gilles