Hello
I am trying to plot dos for pbe0 calculations as it has been explained in the example of "Fcc Ni DOS with hybrid functional". I have used NKRED=4. Checking my OUTCAR for PBE0 calculations I have a complet different band gap comparing the dos for pbe0 calculations. Infact, the dos calculations show a smaller band-gap. Is it because it does not consider the NKRED?
How can I plot the correct dos for the PBE0?
Best Regards,
Asiyeh
DOS using PBE0
Moderators: Global Moderator, Moderator
-
martin.schlipf
- Global Moderator
- Posts: 542
- Joined: Fri Nov 08, 2019 7:18 am
Re: DOS using PBE0
Did you modify the input compared to the tutorial other than the NKRED flag?
If yes, please provide the input and output of your calculation, so that I can inspect the calculation for possible issues.
If you used the tutorial as is, you cannot add NKRED. The 5x5x5 mesh is already so coarse that any further downsampling will introduce large errors. You would probably need to increase the k-point mesh for a production calculation. Also there should not be a gap, because Ni is a metal. That example concerns being cautious about applying hybrid functionals to 3d metals.
Finally, regarding the gap in the DOS vs the gap in the OUTCAR. Please carefully check the smearing ISMEAR used. In the tutorial a tetrahedron method is used for which you would expect any gaps in the OUTCAR to be reproduced in the DOS as well. However, if you changed that to any other smearing method, you expect to see tails in the gap where the width of these tails is dictated by the SIGMA tag.
If yes, please provide the input and output of your calculation, so that I can inspect the calculation for possible issues.
If you used the tutorial as is, you cannot add NKRED. The 5x5x5 mesh is already so coarse that any further downsampling will introduce large errors. You would probably need to increase the k-point mesh for a production calculation. Also there should not be a gap, because Ni is a metal. That example concerns being cautious about applying hybrid functionals to 3d metals.
Finally, regarding the gap in the DOS vs the gap in the OUTCAR. Please carefully check the smearing ISMEAR used. In the tutorial a tetrahedron method is used for which you would expect any gaps in the OUTCAR to be reproduced in the DOS as well. However, if you changed that to any other smearing method, you expect to see tails in the gap where the width of these tails is dictated by the SIGMA tag.
Martin Schlipf
VASP developer