The ABINIT Users Mailing List ( CLOSED )

[Fabien Bruneval <fabien.bruneval@cea.fr>]

Hi again,

Riad or Gian-Marco would provide you a more appropriate answer.

The electronic density is changed only in the case of "self-consistent 
GW". The usual perturbative approach does not affect the density.
Only in the self-consistent case, the density is changed and so is the 
Hartree potential (through the Poisson equation).

In general you can stick to the usual perturbative GW, which gives 
excellent results in most cases.

Cheers,

Fabien




On 09/22/2009 02:42 PM, Rolf Öttking wrote:
> Hello Fabian,
>
> > The options prtpol and prtvxc are meaningless in the context of GW 
> > calculations.
> > The GW self-energy is non-local and dynamic, so it is not a simple 
> > function of the position, like v_xc or v_KS.
> >
> > The code outputs GW-corrected band structure. Does it fit your needs?
>
> Unfortunatley, not quite. In Phys. Rev. Lett. 100, 186401, it says:
> ... In order to compute the QP corrections to the BOs, the many-body 
> effects on Delta V also need to be investigated. This is done by 
> comparing the electronic density and the resulting Delta V calculated 
> within DFT and QSGW...
>
> Would this eventually mean that one computes the electron density in 
> both cases (GW in self-consistency (with QPS), like in the tutorial 
> GW1, chapter 10) and then simply applies Maxwell's equations?
>
> Beste regards
>
> Rolf
>
>
>
>
> > On 09/22/2009 01:09 PM, Rolf Oettking wrote:
> > > Hello
> > >
> > > I have a problem in deriving the total or any other potential in GW
> > > approximation.
> > > Going along the lines of the tutorial in tgw1_9.in and adding
> > > prtpot 1 and
> > > prtvxc 1,
> > > the potentials are printed out correctly in the first and second 
> > > calculation
> > > steps but not in the fourth (or third).
> > >
> > > So, the question is: how would one get the "GW-corrected" potentials
> > >
> > > With best regards
> > >
> > > Rolf Oettking