[Wien] 4f and LDA+U

[laurent.baudery]

Dear stephane,

Once again thank you for your help.

I checked my files and actually before lunching runsp_lapw-orbc the case.dmat is empty. I thought the script runsp_lapw execute automatically lapwdm due to the presence of the inorbc  option and the appropriate case.indmc file. I will recalculate by adding lapwdm before executing  runsp_lapw -orbc.
For the Ef, I do simply a grep command and  I get:

:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.60908
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.60908
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54087
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54275
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.53488
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54824
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54019
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54627
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.53275
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.54530
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.49869
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.53494
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48346
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68766
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48440
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.66995
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48226
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.65435
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48261
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.66488
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48292
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.65511
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.49169
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.67817
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48419
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.67826
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48373
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68742
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48172
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68709
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48223
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68729
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48286
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.67837
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48289
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68727
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48254
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68725
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48251
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68725
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48270
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68720
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48279
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68726
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48245
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68731
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48230
:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68735
 
When I look in the  case.scf2up/dn files, I get also:

case.scf2up  :

       GAUSS-SMEARING WITH    0.00100 Ry

         BANDENERGY CORRECTION:  0.000000
         Bandranges (emin - emax) and occupancy:
        Energy to separate low and high energystates:    0.11380


:NOE  : NUMBER OF ELECTRONS          = 376.000

:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.48230

case.scf2dn  :

       GAUSS-SMEARING WITH    0.00100 Ry

         BANDENERGY CORRECTION:  0.000000
         Bandranges (emin - emax) and occupancy:
        Energy to separate low and high energystates:    0.12854


:NOE  : NUMBER OF ELECTRONS          = 376.000

:FER  : F E R M I - ENERGY(GAUSS-.M.)=   0.68735


Laurent



> Regarding the two different values of  Ef, I checked  case.inorb and it seems ok. U is applied only on the  f states of Nd:

Yes, that file seems to be fine.

>  1  2  0                    nmod, natorb, ipr
> PRATT  1.0                    BROYD/PRATT, mixing
>  1 1 3                          iatom nlorb, lorb
>  2 1 3                          iatom nlorb, lorb
>    1                              nsic 0..AFM, 1..SIC, 2..HFM
>    0.625 0.00        U J (Ry)  Note: we recommend to use U_eff = U-J and J=0
>    0.625 0.00        U J (Ry)  Note: we recommend to use U_eff = U-J and J=0

> the calculation are performed as follow :
>
> runsp_lapw -p -fc 0.1
> save_lapw case1



> runsp_lapw -orbc -p -fc 0.1
> save_lapw case2
> runsp_lapw -orb -p -fc 0.1.
>
> x lapw2 -qtl -c -up  -p
> x lapw2 -qtl -c -dn  -p

This is OK, but I'm not sure whether your -orbc step has any effect. With orbc, the orbital potential is calculated from the existing density matrix (case.dmat*), without updating these later on. In your description, there are no dmat files present yet (I did not verify how wien2k deals with that). You could run lapwdm (with the same options as lapw2) once at that stage.

Can you show by output where you get this information on 'two different fermi energies'? You have to look in case.scf, not elsewhere.

Stefaan

---------------------------------------------------------------------------- 
Laposte.net fête ses 10 ans ! 

Gratuite, garantie à vie et déjà utilisée par des millions d'internautes... 
vous aussi, pour votre adresse e-mail, choisissez laposte.net. 

Laposte.net, bien + qu'une messagerie 
----------------------------------------------------------------------------
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20091030/b4a26297/attachment.htm>