[Pw_forum] A question about PSP Ti.pbe-mt_fhi.UPF and a general question about DFT+U in QE

[evan]

>Dear Paolo

After I have exchanged the occupancies of 4p and 3d states in Ti.pbe-mt_fhi.UPF, I perform the LDA+U calculation with the Ti.pbe-mt_fhi.UPF, the errors appears as:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

     Error in routine offset_atom_wfc (1):

     wrong offset

 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

 

     stopping ...

 

I also try a similar calculation with V.pbe-mt_fhi.UPF, it seems OK. Can you give me some suggestions to solve the problem, the psp and input files are attached.

Thank you in advance.

Evan

USC, China







At 2016-11-27 22:26:19, "Paolo Giannozzi" <p.giannozzi at gmail.com> wrote:

FHI pseudopotentials do not contain information on the reference atomic configuration used in their generation. This is manually provided during the conversion to UPF format. It looks like a mistake. Just exchange the occupancies of 4p and 3d states: they are used only when generating the starting charge for self-consistency, and as a check to prevent usage of non-bound aromic states as Hubbard manifold.


Paolo



On Sun, Nov 27, 2016 at 2:42 PM, Hanghui Chen <chenhanghuipwscf at gmail.com> wrote:

To whom it may concern,
    I noticed an anomaly with the psp Ti.pbe-mt_fhi.UPF in the library. Its valence configuration reads as:


    nl pn  l   occ       Rcut    Rcut US       E pseu
    4s  4  0  2.00      0.000      0.000     0.000000
    4p  4  1  2.00      0.000      0.000     0.000000
    3d  3  2  0.00      0.000      0.000     0.000000
    4f  4  3  0.00      0.000      0.000     0.000000
 I personally think the occupancy for 4p and 3d are incorrect. The correct occupancy might be 4p: 0.00 and 3d: 2.00. I am not sure whether this is indeed the correct psp or a bug in the psp. I want to double check with the community, since it is a psp converted from the Abinit web site. 
   
   A related question to this problem is that in the current implementation of DFT+U in QE, the occupancy for the 'correlated orbital' must have an occupancy larger than 0. Therefore for this psp of Ti, I can not turn on U on its d orbital. I am just curious, is "occupancy of the correlated orbital > 0" a strict condition? Can we make it "occupancy of the correlated orbital >= 0"? I manually changed the source code and it seems that there is no major difference as I have observed. 


  Thank you very much for your time.
  Best regards.


Hanghui Chen
Assistant Professor of Physics
New York University Shanghai
Email: hanghui.chen at nyu.edu

_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum




--

Paolo Giannozzi, Dip. Scienze Matematiche Informatiche e Fisiche,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161210/291e053c/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ti.pbe-mt_fhi.UPF
Type: application/octet-stream
Size: 179544 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161210/291e053c/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tio.opt.in
Type: application/octet-stream
Size: 1536 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161210/291e053c/attachment-0001.obj>