[Pw_forum] Magnetic susceptibility in GIPAW calculation

[Davide Ceresoli]

Dear Peng Tao,
     let's take PRB 63, 245101 (the famous GIPAW paper by
Pickard and Mauri). Eq. 65. is used to calculate the
magnetic susceptibility, and the sequence of operators is:

     <wfc| momentum * green's function * velocity |wfc>

hence the name 'pGv'. However, the PAW contributions to
the susceptibility are very complicated and have not been
implemented, neither in QE-GIPAW nor in PARATEC. A good
estimate is 'vGv', that is velocity*green's function*velocity.

It is expected that the true susceptibility is bounded
between the two values printed by the code (pGv and vGv).
You can see that the difference in not huge.

Regarding time evolution of susceptibility, I think you mean
during a molecular dynamics run. Well, the only way is to
extract snapshots from MD and computed the susceptibility
for each of them.

HTH.

Best wishes,
     Davide



On 05/07/2012 08:24 AM, Peng Tao wrote:
> Dear all,
>
> I find it is difficult to understand the magnetic susceptibility in GIPAW output files. It is written like this:
>
>       chi_bare pGv (HH) in 10^{-6} cm^3/mol:
>             -17.5504          0.0000          0.0000
>               0.0000        -18.2449          0.0000
>               0.0000          0.0000        -24.3766
>
>
>
>
>       chi_bare vGv (VV) in 10^{-6} cm^3/mol:
>             -17.8523          0.0000          0.0000
>               0.0000        -15.4914          0.0000
>               0.0000          0.0000        -25.0791
>
>
> What is "pGv" or "vGv"?
>
> And, I want to investigate how the susceptibility varies with the time. Is it possible for GIPAW to realize this?
>
> Thank you very much and many regards.
>
> Plato Tao
>
>
> --
> -------------------------------------------------------------------
> PH.D. candidate Peng Tao
> Magnetism and Magnetic Materials Division
> National Laboratory for Material Science
> Institute of Metal Research, Chinese Academy of Sciences
> Phone  +86-024-83978751
> -------------------------------------------------------------------