[Pw_forum] Doubts in U (Hubbard parameter) calculation

Phanikumar Pentyala phani12.chem at gmail.com
Sun Oct 16 11:54:34 CEST 2016


Hello QE users

I want to find U value for CeO2 which is having 81 atoms. I already
calculated this value by using linear response (
http://hjklol.mit.edu/content/calculating-hubbard-u). I got the U value
initially in negative numbers in this case *diago_thr_init* value was
default, so that I changed *diago_thr_init* to 10^-11 then I got U value as
0.00881. From literature I found this value should be around 4.5 - 7eV
(Phy. Rev. B 75, 035109, 2007͒). I have some doubts regarding this
calculation

Is that correct, occupations means I took value in the name of *"N of
occupied +U levels" *(i.e sum of Tr[ns(na)] values on all 27 Ce atoms)?

What is the unit for value I got (0.00881)? (Ry or eV or any other )

Is U parameter value depend on number of atoms in unit cell?

Is there any other mistake I done in my input file ? (input file attached)

*Conditions I applied in my calculations:*
Surface calculation with vacuum 15A
First three layers relaxed (two oxygen and one ceria layers) remaining are
frozen


Thanks in advance and I will be greatly appreciated for your help

(input file and results of my calculation attached)


Phanikumar
Research Scholar
Indian Institute of Technology Kharagpur
India
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161016/62b14fb6/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: CeO2.in
Type: application/octet-stream
Size: 5664 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161016/62b14fb6/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Results_u_cal.xlsx
Type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
Size: 4879 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161016/62b14fb6/attachment.xlsx>


More information about the users mailing list