[Pw_forum] about U value determination in DFT+U calculation

[Gabriele Sclauzero]

Il giorno 23/lug/2010, alle ore 18.02, jia chen ha scritto:

> Dear Gabriele Sclauzero,
> 
> Thank you so much for you help, it took me a while to test and to do
> calculations. Now, I think, at least, pw can choose 3d instead of 4d
> without any problem. Now I can get U values and other properties which
> are reasonable even though different from results calculated from
> ultrasoft pseudo-potential. I really appreciate your help.

Yes, you will get different results because the two PP have been generated with different atomic configurations. About the TM PP, it looks strange to me that if you sum the occupations you get 15, while Z=17. Maybe is one of those special PPs used for XPS calculations.

> 
> Just one thing I would like to make sure. If offset_atom_wfc.f90 is
> the only place we need to change something, I am not familiar with pw
> code, I still not 100% sure everything is all rigorous right now.

I think it's OK like that. The "index" of the atomic wavefunctions is calculated only once at the beginning of the run.

> One
> thing I noticed is that the initial guess is not correct, because the
> occupation number after the first step is always 0, anyway, it is not
> a big problem as long as we can get correct ground state. Thank you in
> advance.

Occupations can fluctuate during self consistency because the charge and the wavefunctions do. The case of the isolated atom is quite "unstable" (and difficult to converge in general, you may need to fix occupations of the KS levels from input). The initial guess is built according to the (experimental) occupations of the electronic states in the isolated atom (and on the starting magnetization on that atom, for spin polarized calculations)


Regards,

GS

> 
> Bests
> Jia
> 
> --
> Jia Chen
> 
> Dept of Chemistry
> Princeton University
> Princeton, NJ 08544
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum


§ Gabriele Sclauzero, EPFL SB ITP CSEA
   PH H2 462, Station 3, CH-1015 Lausanne

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20100723/5aa8b789/attachment.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1753 bytes
Desc: not available
Url : http://www.democritos.it/pipermail/pw_forum/attachments/20100723/5aa8b789/attachment.bin