[Pw_forum] Doubt in spin-orbit coupling results

[Gabriele Sclauzero]

Il giorno 21/set/2011, alle ore 12.10, Padmaja Patnaik ha scritto:

> Thanks a lot GS. 
> I was waiting for some suggestions. As u have advised I need to study more about collinear and non-collinear formalism. About the system I am working on --  SiC doped with Cr. I am taking various supercells and Cr as a dilute impurity, doing relaxed and unrelaxed calcualtions for magnetic properties. Hope this is enough and would expect some suggestions for the magnetic moment problem.
> 

Mmh, not easy to say... if there is a magnetic anisotropy, then you need to choose also the correct orientation of the magnetization (see angle1 and angle2) vector on the impurity in order to be sure that the calculation converges to the ground state. On which atoms do you put the starting magnetization? (what are species 1 and 2?)
Are you using smearing?
It would be easier if you post (or send privately) your input file and the relvant part of the output.

GS


> 
> Thanks and regards,
> Padmaja Patnaik
> Research Scholar
> Dept of Physics
> IIT Bombay
> Mumbai, India
> Message: 2
> Date: Tue, 20 Sep 2011 12:28:58 +0200
> From: Gabriele Sclauzero <gabriele.sclauzero at epfl.ch>
> Subject: Re: [Pw_forum] Doubt in spin-orbit coupling results
> To: Padmaja Patnaik <padmaja_patnaik at yahoo.co.uk>,    PWSCF Forum
>     <pw_forum at pwscf.org>
> Message-ID: <BE7B21AF-2135-4CB9-9A66-6128CBB6A4FE at epfl.ch>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Dear  Padmaja Patnaik,
> 
>     in order to include spin-orbit interaction in the KS equations (lspinorb=.true.) one needs to use a formalism for noncollinear magnetization (noncolin=.true.). This means that there is no preferential axis for the magnetization, which is a vector field, (m_x(r),m_y(r),m_z(r)) in this case.
> Therefore, in general one cannot define "spin up" and "spin down" (unless the ground state of the system is collinear, e.g. m_x=m_y=0 for each r). That's why the code does not give two distinct components of the DOS in the noncollinear formalism.
> 
> What you should do depends on what you are looking for... first I would suggest that you try to understand the difference between the collinear and noncollinear formalisms, in order to be able to interpret the results.
> 
> About the total magnetic moment, I can't say anything because you didn't say a word about your system.
> 
> 
> GS
> 
> Il giorno 12/set/2011, alle ore 07.06, Padmaja Patnaik ha scritto:
> 
> > Hi All
> > 
> > I am doing spin-orbit coupling calculations. To get spin-polarized calculations with spin-orbit coupling I mentioned in the input for scf,
> >  lspinorb=.true.,
> >  noncolin=.true.,
> >  starting_magnetization(1)=0.5,
> >  starting_magnetization(2)=0.5,
> > After this scf calculations I am doing the nscf calculation and finally the density of states and partial density of states. What I expect is it should give me both up and down spin DOS separately but I didnt get that. The input,  nspin=2 in the input and get spin polarized results with both up and down spin density of states given separately which can be plotted. Can anybody please suggest what to do? Another doubt is the total magnetic moment is found to be zero (with spin-orbit interaction) which was earlier (without spin-orbit interaction) found to be 2 BohrMagneton./cell. 
> > 
> 
> 
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum


§ 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/20110921/4e45ccb2/attachment.html>