[Wien] Treating 4f states as core

[Ali Baghizhadeh]

Dear Users of Wien2K

I am doing spin polarized calculation of h-YbFeO3 with Fe ions as antiferromagnetic. I used LDA+U on both Fe and Yb to open the band gap, even with U =7 ev, the band gap is very narrow and MMT becomes as large as 2 (no matter if two Yb ions are AFM or FM). As I read in mailing list, I may force Yb 4f orbitals to be core state to resolve large magnetic moments.
To do so I assume I have to modify case.in1c (or case.in1_st in w2web?).  The line:  3    0.30     0.0010  CONT  1  should be modified e.g. as   3    -3.00     0.0000  CONT  1. I got the value (-3) from mailing list, but in my case, the threshold between core and valance states in Yb is -4.5 ev as we see during initialization in the following. I do not know how to come up with proper value to force Yb 4f to become Core state? and then shall I do LDA+U only for Fe or for both Fe (3d) and Yb (4f)?

Thank you very much.
Ali  Baghi zadeh

Program input is: "13  -7.0 "
Atomic configuration for atom: Yb1   Z=  70.00
  .
  .
  .
  4D*     -14.018197    -13.950573  2.00  2.00    1.0000  T
  4D       -13.341524    -13.274425  3.00  3.00    1.0000  T
  5S        -4.359687     -4.309111     1.00  1.00    0.9745  F
  5P*      -2.627456     -2.577135    1.00  1.00    0.9328  F
  5P        -2.153318     -2.103138    2.00  2.00    0.8939  F
  4F*      -0.774283     -0.709142    3.00  3.00    0.9875  F
  4F        -0.671996     -0.607142    4.00  3.00    0.9853  F
  5D*     -0.162143     -0.120441    1.00  0.00    0.3841  F
  6S        -0.346341     -0.313221    1.00  1.00    0.0802  F

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20200407/348c7f6c/attachment.html>