[Pw_forum] Total magnetization and local magnetic moment

[Gabriele Sclauzero]

On 07/06/2010 11:34 AM, Padmaja Patnaik wrote:
> First of all thanks a lot for your suggestions.
> As you mentioned that the super cell atomic positions are not correct 
> then why the code is not giving error?
>

I said "not correct" in the sense that they do not describe any polytype 
of SiC. I was meaning that your atomic structure does not represent 
3C-SiC. The code does not know what kind of system you want to describe 
since it cannot read into your mind and say "hey, stop! this is not the 
system you have in mind!".

It gives however an error message when atoms are (almost) overlapping, 
since this would lead it into serious troubles when computing the energy 
and the potential.

Then I don't know if with the other supercells you entered the atomic 
positions correctly or not, but if those describe correctly 3C-SiC, then 
of course you'll find different results compared to this supercell.

>
> Coming to the magnetic moment querry, as you said the d electrons get 
> filled and thus loosing the magnetic moment. But then one can expect 
> that the moment may be trasffered to some other atom in the cell. But 
> this also in not true in my result. Can this error be because i am 
> doing the scf calculations for gamma point only and then the DOS 
> calculations for 2*2*2 k points mesh? This is the only thing coming to 
> my mind where i can be wrong. Is the difference in k point mesh give 
> such a big difference in the magnetization value, total magnetization 
> is > 1 and local moment is zero?
>

It is important to converge total energy (and magnetization, for 
spin-polarized systems) in the first place. Then you may need more 
k-points in the following nscf calculation in order to converge the DOS 
(also depending on the smearing you choose for projwfc).


HTH

GS

> Please suggest.
> Regards
>
> Padmaja Patnaik
> Research Scholar
> Dept of Physics
> IIT Bombay
> Mumbai, India
>


-- 

Gabriele Sclauzero, EPFL SB ITP CSEA
PH H2 462, Station 3, CH-1015 Lausanne

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20100706/7007083f/attachment.html>