[Pw_forum] magnetization and spin contamination

[Paolo Giannozzi]

On Sunday 30 October 2005 22:58, Silviu Zilberman wrote:

> I have a few questions related to spin-polarized calculations in PWSCF:
>
> 1. Is there a way to get the spin contamination information?

not presently, but in CP there is an estimation (or maybe two)
of spin contamination that could be easily adapted to PWscf

> 2. One can constrain the total magnetization. The INPUT_PW file provides 
> some information on how it is done in the non-collinear case, but the 
> collinear case (nspin=2) is more obscured. Could someone explain it 
> some more or point me to a reference?

see example 13

> 3. Can one remove the spin-contamination by constraining to total
> magnetization to the correct value s(s+1)?

I think that the constraints are on S_z (for a generic direction of z),
not on S^2, but I might be wrong

Paolo

-- 
Paolo Giannozzi             e-mail:  giannozz at nest.sns.it
Scuola Normale Superiore    Phone:   +39/050-509876, Fax:-563513 
Piazza dei Cavalieri 7      I-56126 Pisa, Italy