[Wien] Non reproducibility of previous results: spin-orbit/orbital polarization

[Peter Blaha]

On our update page it says somewhere:

SRC_orb: small fix for orbital polarization.

This was done by P.Novak, but I would conclude that the older version had some
(small) bug.

ROBERTO LUIS IGLESIAS PASTRANA schrieb:
> Dear all
> 
> While trying to reproduce some old tests for the bcc Fe system I bumped into a problem. Before I was using Wien2k_06.4 and now I'm using the latest Wien2k_08.3 release. I tried with exactly the same input files which, in a few cases, meant going back to slightly slower mixing schemes (BROYD instead of MSEC1 in case.inm) and with the new and improved ones as well. In a normal spin-polarized calculation there is no problem. However, when spin-orbit coupling and orbital polarization correction in case.inorb is added, I get different results. After executing initso_lapw, and running dstart, dstart up/dn again, the number of plane waves in case.outputd increases from 112 to 286. The case.inso file reads:
> 
> WFFIL
>  4  1  0                      llmax,ipr,kpot
>  -10.0000   7.00000           emin,emax (output energy window)
>    0.  0.  1.                 direction of magnetization (lattice vectors)
>  0                           number of atoms for which RLO is added
>  0 0 0 0 0                    number of atoms for which SO is switch off; atoms
> 
> Subsequently,  I ran
> 
> runsp_lapw -so -cc 0.0001 ec 0.00001
> save_lapw -d SOC
> 
> I build the right case.inorb
> 
>   2  1  0                     nmod, natorb, ipr
> PRATT  1.0                    BROYD/PRATT, mixing
>   1 1 2                          iatom nlorb, lorb
>   1                              nmdep
>   1                              Ncalc
>   0. 0. 1.                    direction of M
> 
> and case.indmc
> 
> -10.                      Emin cutoff energy
>  1                       number of atoms for which density matrix is calculated
>  1  1  2      index of 1st atom, number of L's, L1
>  0 0           r-index, (l,s)index
> 
> And
> runsp_lapw -orb -so -cc 0.0001 -ec 0.00001
> 
> I'm not able to reproduce the original results, specially regarding the orbital moment, which is now:
> 
> :ORB001:  ORBITAL MOMENT:  0.00000  0.00000  0.06208 PROJECTION ON M  0.06209
> 
> and used to be:
> 
> :ORB001:  ORBITAL MOMENT:  0.00000  0.00000  0.07793 PROJECTION ON M  0.07793
> 
> Now, were there any significant changes in lapwdm or orb between these two Wien2k versions? I found some updates in the web page, of course, but I don't know if they would be significant as to change the orbital moment as much as a 20%.
> One thing I noticed is that case.in0 has changed the last line from
>   18  18  18    2.00    min IFFT-parameters, enhancement factor
> 
> to
>   24  24  24    2.00    min IFFT-parameters, enhancement factor
> 
> 
> Please excuse my ignorance, but I don't know if this is relevant or not. I'm using option 13 (PBE) for the XC potential.
> 
> Thanks in advance for any suggestions and merry X-mas to you all!
> 
> Roberto
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------