[Pw_forum] [Fwd: paramagnetic structure]

[Davide Ceresoli]

Subhradip Ghosh wrote:
> I would like to know how the latest version of espresso can be used to do
> paramagnetic calculation? Would nspin=1 be enough?

Dear Subhradip,
     a "material" can be paramagnetic, not a "calculation". Actually, materials
can be paramagnetic, diamagnetic, ferromagnetic, antiferromagnetic, ...

Calculations can be either "spin polarized" (nspin = 2) or "not spin
polarized" (nspin = 1).
Putting nspin=1 would mean to neglect spin polarization effects at all.
This is a good approximation in a closed shell system, where all spin
up and spin down levels are the same and there are no unpaired spins.

Regards,
     Davide