Moderators: Global Moderator, Moderator
Hello everyone,
The calculation of band structure typically follows these steps:
Step 1: Optimize the structure.
Step 2: Perform an SCF calculation.
Step 3: Read the CHGCAR from the SCF calculation and run a non-self-consistent (NSCF) calculation.
When performing band structure calculations using hybrid functionals, is it acceptable to use PBE for the SCF step? Or must the hybrid functional be employed for both the SCF and NSCF calculations?
Looking forward to your reply. Thanks!
Hello!
The workflow for hybrid functionals is similar but there are some notable differences. Please read up on this topic on our dedicated Wiki page here:
https://www.vasp.at/wiki/index.php/Band ... unctionals
Also, we have a talk about hybrid functionals online at YouTube and there is a section regarding band-structure calculations as well:
https://www.youtube.com/watch?v=zNc0bx8FTlU&t=3563s
Regarding your specific question whether it would be acceptable to use PBE in each of the steps:
Step 1: Of course you could use PBE here to "pre"-optimize your structure but it is system- and functional-dependent whether the converged structure will be close enough to the one corresponding to the hybrid functional.
Step 2: Although it is technically possible to use PBE in this step we would generally recommend to use the hybrid functional here. As in step 1 there is no general advice which pathway is still "acceptable" because the differences will be system- and functional-dependent.
All the best,
Andreas Singraber