[Wien] on-site hybrid functionals

[Peter Blaha]

 > I stabilize this solution in a separate calculation with fixed
 > occupation matrices run as
 >
 > runsp_lapw -p -orb -so -ec 0.00001 -cc 0.0001

With this you cannot stabilize it.
After you created the dmat files, you need   x orb -up/-dn    and
runsp_lapw -p -orbc -so -ec 0.00001 -cc 0.0001
(maybe you have done it this way anyway ??)

 > I start my hybrid (-eece) calculation from a converged fixed matrices (-
 > orbc) calculation.  The problem is that DNDN occupation matrix changes
 > dramatically in the very first iteration to this:

Your last dmat-dn file looks fully symmetric, i.e. it must have been 
produced without -so
Without SO, it is "natural" to have identical m=+3 and -3 occupation.
You probably forgot  -so  in the runsp command.
(You need a case.indmc  file !)
(Did you get a proper  x lapwdm -up -so   step into the dayfile ?

Regards

Am 31.07.2026 um 16:31 schrieb Kateryna Foyevtsova:
> Dear Wien2k team,
> 
> I am trying to apply the on-site hybrid functional method to treat the 
> Tb 4f electrons in TbMn6Sn6, together with the SO coupling. I use
> 
> runsp_lapw -p -eece -so -ec 0.00001 -cc 0.0001
> 
> My case.ineece is this:
> 
> -12.0  1       Emin natom
> 1  1 3      index of atom, number of l, l
> HYBR          HYBR / EECE
> 0.25          amount of exact exchange
> 
> I would like to have a hybrid solution with the following f-orbital 
> occupation matrices:
> 
> Density matrix UPUP block, real part.  L= 3
>           0.98992  0.00000  0.00000  0.00000  0.00000  0.00000 -0.00047
>           0.00000  0.99177  0.00000  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.99246  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.99251  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.99261  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.00000  0.99181  0.00000
>          -0.00047  0.00000  0.00000  0.00000  0.00000  0.00000  0.99071
>   Density matrix DNDN block, real part.  L= 3
>           0.98393  0.00000  0.00000  0.00000  0.00000  0.00000  0.01313
>           0.00000  0.00237  0.00000  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00251  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00303  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.00240  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.00000  0.00212  0.00000
>           0.01313  0.00000  0.00000  0.00000  0.00000  0.00000  0.00201
> 
> I stabilize this solution in a separate calculation with fixed 
> occupation matrices run as
> 
> runsp_lapw -p -orb -so -ec 0.00001 -cc 0.0001
> 
> I start my hybrid (-eece) calculation from a converged fixed matrices (- 
> orbc) calculation.  The problem is that DNDN occupation matrix changes 
> dramatically in the very first iteration to this:
>           0.49295  0.00000  0.00000  0.00000  0.00000  0.00000  0.01313
>           0.00000  0.00223  0.00000  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00247  0.00000  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00303  0.00000  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.00247  0.00000  0.00000
>           0.00000  0.00000  0.00000  0.00000  0.00000  0.00223  0.00000
>           0.01313  0.00000  0.00000  0.00000  0.00000  0.00000  0.49295
> 
> My mixing factor is small: 0.05 for the Pratt scheme.
> 
> I would appreciate your insights and suggestions.
> 
> Thank you,
> Kateryna
> 
> 
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> https://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:  http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-- 
-----------------------------------------------------------------------
Peter Blaha,  Inst. f. Materials Chemistry, TU Vienna, A-1060 Vienna
Email: peter.blaha at tuwien.ac.at
WWW:   http://www.imc.tuwien.ac.at      WIEN2k: http://www.wien2k.at
-------------------------------------------------------------------------