The ABINIT Users Mailing List ( CLOSED )

[Valerio Olevano <valerio.olevano@polytechnique.fr>]

Dear all,

>  but what I was wondering was if it was possible to use it on 
> semiconductor surfaces (specifically, of those semiconductors which are 
> known to be problematic for regular old LDA/GGA-DFT).  I realize that 
> GW calculations tend to be rather difficult and are quite 
> computationally expensive, so I wouldn't be surpris!
> ed if the GW implimentation was inadequate (although I'd be quite happy 
> if it was possible).

I know for sure that you can do also
semiconductor surfaces with the GW part of ABINIT.
There are many papers in the literature
(check the names O. Pulci and M. Palummo)
that show calculated GW corrections in surfaces
by using a code which is not far from that one in abinit.

> I would certainly encourage you to write up some documentation on the 
> GW implementation besides the GW variables. I know of several groups 
> looking forward to using GW.

I agree that the documentation is not complete.
We wrote recently a tutorial for the GW part of abinit
which very early will be available.
However, before to learn how to use the code,
it's the case to learn
what is Many-Body  Quantum Field or Perturbation Theory,
what is a Green function,
what is a Self-Energy, and only at the end what is
the GW approximation. You cannot wait that the ABINIT
site will learn you on this, as well as you have to learn at least the 
DFT
before to use the basic abinit kernel.
In my opinion it's not a good aptitude to use the codes
without having understood what's the theory behind.
It's a bit what the users of another code, the quantum chemistry
Gaussian, very often do.
If you do that, you will learn also which answers can give
you the GW, which others the TDDFT and what you cannot ask
to GW or TDDFT.

best regards

Valerio Olevano