I just noticed that EFERMI = MIDGAP is not the default in v6.4.1, but EFERMI = LEGACY is.
this is not what is stated in this wiki.
wiki/index.php/EFERMI
Is there a specific reason why the default value was changed again?
EFERMI = MIDGAP is not the default in v6.4.1?
Moderators: Global Moderator, Moderator
-
alpinnovianus
- Newbie
- Posts: 43
- Joined: Tue Dec 17, 2019 7:56 am
-
martin.schlipf
- Global Moderator
- Posts: 542
- Joined: Fri Nov 08, 2019 7:18 am
Re: EFERMI = MIDGAP is not the default in v6.4.1?
Thank you for catching that. I fixed the wiki entry to correct that.
We reverted to the legacy behavior because for some materials EFERMI = LEGACY would yield different results than EFERMI = MIDGAP. The typical case is a semiconductor, where the user forgot to change the default ISMEAR. Then Methfessel-Paxton smearing is used, which has multiple possible solutions for the Fermi energy because it uses a non-monotonous function. In this cases, the new behavior would be preferred but we did not want to break backwards compatibility. Therefore we recommend to use EFERMI = MIDGAP.
We reverted to the legacy behavior because for some materials EFERMI = LEGACY would yield different results than EFERMI = MIDGAP. The typical case is a semiconductor, where the user forgot to change the default ISMEAR. Then Methfessel-Paxton smearing is used, which has multiple possible solutions for the Fermi energy because it uses a non-monotonous function. In this cases, the new behavior would be preferred but we did not want to break backwards compatibility. Therefore we recommend to use EFERMI = MIDGAP.
Martin Schlipf
VASP developer