[Pw_forum] polarization spin versus non-spin

Andrei Malashevich andrei.malashevich at yale.edu
Fri Feb 20 16:45:45 CET 2015


Dear Javier,

For spin-polarized case, you have 0.6698526-1.0038673=-0.3340147,
value very similar to the spin unpolarized case.
Polarization is defined only modulo a quantum, and in the above, I just
subtracted the quantum.

Andrei


On Fri, Feb 20, 2015 at 10:27 AM, Nossa, Javier <jnossa at carnegiescience.edu>
wrote:

> Dear pwscfers,
> I am calculating the polarization of BaTiO3 and found a different value of
> it when implementing spin-polarized, nspin=2.
>
> non-spin-polarized gives P=-0.3340199 (mod 2.0077346) C/m^2, which is
> correct.
> spin-polarized gives P=0.6698526 (mod 1.0038673)  C/m^2, which is not
> correct.
>
> Although it seems to be just a factor of two difference in the absolute
> value of the polarization, it is not exactly a factor of two. It is
> 2.0054...
> In addition, the sign of the polarization is opposite.
> On the other hand, the mod is exactly a factor of two.
>
> Why there is such a difference in the polarization?
>
> I am doing this as a test because my real system is magnetic and bigger
> than BaTiO3.
>
> Here are the inputs for the nscf calculation:
>
> spin-polarized:
> &control
>     calculation='nscf'
>     restart_mode='from_scratch',
>     prefix='job',
>     lberry        = .true.
>     gdir          = 3
>     nppstr        = 10
>  /
> &system
>     ibrav= 0,
>     celldm(1)=1.0,
>     nat=  5, ntyp= 3,
>     ecutwfc = 120,
>     !occupations='smearing', smearing='gauss', degauss=0.003,
>     !occupations='fixed',
>     input_dft='wc'
>     nspin=2,
>     tot_magnetization= 0.0,
> /
> &electrons
> /
> ATOMIC_SPECIES
>  Ba  137.327 Ba.wc-n-nc.UPF
>  Ti  47.867 Ti.wc-n-nc.UPF
>   O  15.9994 O.wc-nc.UPF
> CELL_PARAMETERS (alat=1.0)
>  7.54662 0.0 0.0
>  0.0 7.54662 0.0
>  0.0 0.0 7.74409
> ATOMIC_POSITIONS {crystal}
> Ba      0.00000 0.00000 0.00000
> Ti      0.50000 0.50000 0.48186
> O       0.00000 0.50000 0.51791
> O       0.50000 0.00000 0.51791
> O       0.50000 0.50000 0.03133
> K_POINTS {automatic}
> 4 4 4 0 0 0
>
>
> non-spin-polarized:
> &control
>     calculation='nscf'
>     restart_mode='from_scratch',
>     prefix='job',
>     lberry        = .true.
>     gdir          = 3
>     nppstr        = 10
>  /
> &system
>     ibrav= 0,
>     celldm(1)=1.0,
>     nat=  5, ntyp= 3,
>     ecutwfc = 120,
>     !occupations='smearing', smearing='gauss', degauss=0.003,
>     !occupations='fixed',
>     input_dft='wc'
> /
> &electrons
> /
> ATOMIC_SPECIES
>  Ba  137.327 Ba.wc-n-nc.UPF
>  Ti  47.867 Ti.wc-n-nc.UPF
>   O  15.9994 O.wc-nc.UPF
> CELL_PARAMETERS (alat=1.0)
>  7.54662 0.0 0.0
>  0.0 7.54662 0.0
>  0.0 0.0 7.74409
> ATOMIC_POSITIONS {crystal}
> Ba      0.00000 0.00000 0.00000
> Ti      0.50000 0.50000 0.48186
> O       0.00000 0.50000 0.51791
> O       0.50000 0.00000 0.51791
> O       0.50000 0.50000 0.03133
> K_POINTS {automatic}
> 4 4 4 0 0 0
>
>
> Pseudopotentials are taken from THEOS PSLibrary 0.3.1.
>
> Outputs are attached.
>
>
> Thank you very much for your help.
>
> --
> With best regards,
> Javier Francisco Nossa
>
> Postdoc at Geophysical Laboratory
> Carnegie Institution of Washington
> 5251 Broad Branch Road, N.W.
> Washington, DC 20015-1305
> Tel.: 1.240.476.3993
> E-mail: jnossa at carnegiescience.edu <dappleby at ciw.edu>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
>
> https://urldefense.proofpoint.com/v2/url?u=http-3A__pwscf.org_mailman_listinfo_pw-5Fforum&d=AwICAg&c=-dg2m7zWuuDZ0MUcV7Sdqw&r=iaCUSkNPjaS0OZmx8CL3zuFJHmomB6kEDsD3Du2Xz38&m=jHc1e4BMDezKEN4w-4Sz3WYY43oEOCVzdofkatuhMr0&s=ptunJ7qbTZCjdsr3TWNH3-Yh3KVY76v7PTd5Aj-blsQ&e=
>



-- 
Andrei Malashevich
Postdoctoral Associate
Center for Research on Interface Structures and Phenomena
Department of Applied Physics
Yale University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150220/35b6c8e9/attachment.html>


More information about the users mailing list