[Pw_forum] electrical polarization calculations for non-collinear magnetism

[Xun-Wang Yan]

Dear Paolo,

Thank you for your reply, your comment is very useful to me. I shall
have a try to change some part of the code 'bp_c_phase.f90' to include the
noncolin=.true. case.

Best regards

Xun-Wang

2009/3/28 Paolo Giannozzi 

>
> On Mar 28, 2009, at 6:25 , Xun-Wang Yan wrote:
>
> >  nspin=1 and nspin=2 cases is supported,   there is not
> > noncolin=.true.  case.
> > how to modify the bp_c_phase.f90 to include the noncolin=.true.
> > case for electric
> > polarization calculation?
>
> I know nothing about polarization and noncolinear calculations, but
> basically
> what you need in a polarization calculation is a sequence of scalar
> products
> between electronic states at different k-points. In the case of
> noncolinear
> calculations, electronic states are spinors and are expanded into plane
> waves with both spin-up and spin-down components. Coefficients 1 to npw
> are for spin-up plane waves, coefficients npwx+1 to npwx+npw are for
> spin-down plane waves. Scalar products are summed over the two sets.
>
> Paolo
> ---
> Paolo Giannozzi, Democritos and University of Udine, Italy
>
>
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090329/d44033bf/attachment-0001.htm
 

[Pw_forum] electrical polarization calculations for non-collinear magnetism

[Paolo Giannozzi]

On Mar 28, 2009, at 6:25 , Xun-Wang Yan wrote:

>  nspin=1 and nspin=2 cases is supported,   there is not   
> noncolin=.true.  case.
> how to modify the bp_c_phase.f90 to include the noncolin=.true.  
> case for electric
> polarization calculation?

I know nothing about polarization and noncolinear calculations, but  
basically
what you need in a polarization calculation is a sequence of scalar  
products
between electronic states at different k-points. In the case of  
noncolinear
calculations, electronic states are spinors and are expanded into plane
waves with both spin-up and spin-down components. Coefficients 1 to npw
are for spin-up plane waves, coefficients npwx+1 to npwx+npw are for
spin-down plane waves. Scalar products are summed over the two sets.

Paolo
---
Paolo Giannozzi, Democritos and University of Udine, Italy

[Pw_forum] electrical polarization calculations for non-collinear magnetism

[Xun-Wang Yan]

Dear Gabriele Sclauzero,
Thanks for your reply,
 the electric polarization calculation for noncolinear magnetism is not
available for espresso-4.0.4 version, which can be found from the code
'bp_c_phase.f90'.
  For this calculation , the nelec bands(in my calculation,87 bands) should
be  dealt with.
I found the 'nscf' calculation is finished, and next step to calculate
electric polarization with berry phase method is not implemented,
i.e.the running process stop at the code 'bp_c_phase.f90' .
In bp_c_phase.f90, the  nspin=1 and nspin=2 cases is supported,   there is
not  noncolin=.true.  case.
how to modify the bp_c_phase.f90 to include the noncolin=.true. case for
electric polarization calculation?
your suggestion and comment are expected.

Bests

Xun-Wang

2009/3/27 Gabriele Sclauzero 

>
> Xun-Wang Yan wrote:
> > &system
> >ibrav=14,
> >celldm(1)=9.73748132, celldm(2)=1.155067564,celldm(3)=1.00,
> >celldm(4)=-0.144291712,celldm(5)=-0.50,celldm(6)=0.288583423,
> >nat=12, ntyp=5,
> >ecutwfc=30, ecutrho=300,
> >report=10,
> >occupations='fixed',degauss=0.00,
> >noncolin=.true.
> >starting_magnetization(2)=0.5
> >angle1(2)=90.0
> >angle2(2)=0.0
> >starting_magnetization(3)=0.5
> >angle1(3)=90.0
> >angle2(3)=120.0
> >starting_magnetization(4)=0.5
> >angle1(4)=90.0
> >angle2(4)=-120.0
> >nbnd=87
>
>
>
> >
> %%
> >  task # 3
> >  from c_phase : error # 1
> >  Polarization only for insulators and no empty bands
> >
>  
> %%
> >
> > I look over 'bp_c_phase.f90' , and find that electrical polarization
> > calculation is applicable for nspin=1 and nspin=2.
>
> I don't know if this calculation is available for noncolinear calculation,
> but if you pay
> attention to the error message, what it says is different. You need to deal
> with an
> insulator (and you are) and you must compute only nelec bands, i.e. NOT
> adding extra bands
> through the nbnd keyword, as you're doing.
> Retry after removing the nbnd line.
>
> GS
>
>
> > I want to know to how to modify the bp_c_phase.f90 to implement the
> > electrical polarization calculation for non-collinear magnetism.
> > Any comments are welcome , many thanks in advance.
> >
> > regards,
> > Xun-wang Yan
> > 
> > Institute of Theoretical Physics, CAS
> > Address: No.55, Zhong-Guan-Cun East Road
> > Beijing, China
> > Email: xwyan at itp.ac.cn
> > 
> >   yanxunwang2008 at gmail.com 
> > ===
> >
> >
> > 
> >
> > ___
> > Pw_forum mailing list
> > Pw_forum at pwscf.org
> > http://www.democritos.it/mailman/listinfo/pw_forum
>
> --
>
>
> o  o
> | Gabriele Sclauzero, PhD Student  |
> | c/o:   SISSA & CNR-INFM Democritos,  |
> |via Beirut 2-4, 34014 Trieste (Italy) |
> | email: sclauzer at sissa.it |
> | phone: +39 040 3787 511  |
> | skype: gurlonotturno |
> o  o
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090328/65ba7b23/attachment.htm
 

[Pw_forum] electrical polarization calculations for non-collinear magnetism

[Gabriele Sclauzero]

Xun-Wang Yan wrote:
> &system
>ibrav=14,
>celldm(1)=9.73748132, celldm(2)=1.155067564,celldm(3)=1.00,
>celldm(4)=-0.144291712,celldm(5)=-0.50,celldm(6)=0.288583423,
>nat=12, ntyp=5,
>ecutwfc=30, ecutrho=300,
>report=10,
>occupations='fixed',degauss=0.00,
>noncolin=.true.
>starting_magnetization(2)=0.5
>angle1(2)=90.0
>angle2(2)=0.0
>starting_magnetization(3)=0.5
>angle1(3)=90.0
>angle2(3)=120.0
>starting_magnetization(4)=0.5
>angle1(4)=90.0
>angle2(4)=-120.0
>nbnd=87



> %%
>  task # 3
>  from c_phase : error # 1
>  Polarization only for insulators and no empty bands
>  
> %%
> 
> I look over 'bp_c_phase.f90' , and find that electrical polarization 
> calculation is applicable for nspin=1 and nspin=2.

I don't know if this calculation is available for noncolinear calculation, but 
if you pay 
attention to the error message, what it says is different. You need to deal 
with an 
insulator (and you are) and you must compute only nelec bands, i.e. NOT adding 
extra bands 
through the nbnd keyword, as you're doing.
Retry after removing the nbnd line.

GS


> I want to know to how to modify the bp_c_phase.f90 to implement the 
> electrical polarization calculation for non-collinear magnetism.
> Any comments are welcome , many thanks in advance.
> 
> regards,
> Xun-wang Yan
> 
> Institute of Theoretical Physics, CAS
> Address: No.55, Zhong-Guan-Cun East Road
> Beijing, China
> Email: xwyan at itp.ac.cn 
> 
>   yanxunwang2008 at gmail.com 
> ===
> 
> 
> 
> 
> ___
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum

-- 


o  o
| Gabriele Sclauzero, PhD Student  |
| c/o:   SISSA & CNR-INFM Democritos,  |
|via Beirut 2-4, 34014 Trieste (Italy) |
| email: sclauzer at sissa.it |
| phone: +39 040 3787 511  |
| skype: gurlonotturno |
o  o

[Pw_forum] electrical polarization calculations for non-collinear magnetism

[Xun-Wang Yan]

Dear pwscf users,

I want to calculate the electrical polarizaiton for CuCrO2 with
non-collinear antiferromagnetic ground state.After a 'scf' calculation, I
run 'nscf' calculation. the input file is following,
-
&control
   calculation = 'nscf'
   prefix='120-AFM-12at-tricli',
   pseudo_dir ='/home/xwyan/Pseudo/'
   outdir='/home/xwyan/tmp-120-0/'
   lberry=.true.
   gdir=3
   nppstr=8
/
&system
   ibrav=14,
   celldm(1)=9.73748132, celldm(2)=1.155067564,celldm(3)=1.00,
   celldm(4)=-0.144291712,celldm(5)=-0.50,celldm(6)=0.288583423,
   nat=12, ntyp=5,
   ecutwfc=30, ecutrho=300,
   report=10,
   occupations='fixed',degauss=0.00,
   noncolin=.true.
   starting_magnetization(2)=0.5
   angle1(2)=90.0
   angle2(2)=0.0
   starting_magnetization(3)=0.5
   angle1(3)=90.0
   angle2(3)=120.0
   starting_magnetization(4)=0.5
   angle1(4)=90.0
   angle2(4)=-120.0
   nbnd=87
/
&electrons
   electron_maxstep=300
   mixing_beta = 0.3
   conv_thr =  1.0d-10
/
&electrons
   electron_maxstep=300
   mixing_beta = 0.3
   conv_thr =  1.0d-10
/
ATOMIC_SPECIES
Cu   63.55  Cu.pz-d-rrkjus.UPF
Cr1  52.00  Cr.pz-rrkjus-5-1.UPF
Cr2  52.00  Cr.pz-rrkjus-5-1.UPF
Cr3  52.00  Cr.pz-rrkjus-5-1.UPF
O16.00  O.pz-rrkjus.UPF
ATOMIC_POSITIONS {angstrom}
Cu   0.0   0.0  -0.00127
Cu   2.576425576   0.0   1.487499959
Cu   0.0   0.0   2.97591
Cr1  0.858807500   2.849329739   1.497499627
Cr2  3.435229613   2.849329739   0.00479
Cr3  3.435229613   2.849329739   2.985001062
O2.576426737   1.824279229   1.487500064
O1.717608630   3.874380147  -0.00400
O4.294034914   3.874380249   1.487500064
O5.152850596   1.824279331  -0.00400
O1.717611367   3.874380065   2.975000291
O0.03312   1.824279413   2.975000291
K_POINTS {automatic}
1 1 8 0 0
-


%%
 task # 3
 from c_phase : error # 1
 Polarization only for insulators and no empty bands
 %%

I look over 'bp_c_phase.f90' , and find that electrical polarization
calculation is applicable for nspin=1 and nspin=2.
I want to know to how to modify the bp_c_phase.f90 to implement the
electrical polarization calculation for non-collinear magnetism.
Any comments are welcome , many thanks in advance.

regards,
Xun-wang Yan

Institute of Theoretical Physics, CAS
Address: No.55, Zhong-Guan-Cun East Road
Beijing, China
Email: xwyan at 
itp.ac.cn
  yanxunwang2008 at gmail.com
===
-- next part --
An HTML attachment was scrubbed...
URL: 
http://www.democritos.it/pipermail/pw_forum/attachments/20090323/cb44b9c6/attachment.htm