The ABINIT Users Mailing List ( CLOSED )

[Xavier Gonze <gonze@pcpm.ucl.ac.be>]

Dear David,

prendergast2@llnl.gov wrote:
> Can anyone give a clear explanation of ABINIT's approach to dealing with 
> charged systems, (i.e.
> input variable: charge!=0) ?
>
> Basically, I need to know how nonneutral systems are treated in terms of the 
> potentials used? Does
> the code just modify the periodic summations/integrals to include a uniform 
> background charge
> density of equal but opposite magnitude to the value given by the variable 
> 'charge', or is there
> more to it? For example is there screening between neighbouring unit cells to 
> remove charge-charge
> interactions that would lead to finite size errors.

Concerning this, the answer of Vincenzo Fiorentini is quite accurate : one 
should
add the Makov and Payne correction to the total energy obtained by ABINIT.

> This question is relevant to the determination of quasiparticle gaps for 
> finite systems using gamma
> point calculations (E_QP = E_N+1 + E_N-1 - 2 E_N).

Well, you should use TDDFT ! This is the typical case (finite system) where 
TDDFT works
quite well to find the gaps.
In ABINIT, this means that you have to use iscf -1 .
Examples of the use of TDDFT are given in Test_v3 : t54.in, t55.in, and the 
quite
detailed README (especially about the N2 molecule).

Good continuation,
Xavier