Bandstructure by hybrid functional

Questions will be moved to this forum when we consider them out of scope for support from our side: for instance when we do not have the necessary expertise to come up with an answer.
Another user still might, though!

Moderators: Global Moderator, Moderator

Post Reply
Message
Author
IBRAHIM
Jr. Member
Jr. Member
Posts: 82
Joined: Mon Jan 30, 2017 6:00 pm
License Nr.: 5-2551

Bandstructure by hybrid functional

#1 Post by IBRAHIM » Tue Dec 21, 2021 1:20 pm

Dear All,

For calculation of bandstructure using hybrid functional, I follow the procedure 2 of vasp wiki wiki/index.php/Si_HSE_bandstructure

1- standard DFT
2-hybrid calculation with wavecar (rocommend to be from step and not from hybrid calculation) and modified KPOINTs file

My question is that: is it possible in step 2 to allow also ionic relaxation by provide NSW>1. Most researcher use NSW=0. Howvere, aditional ionic relaxation by hybrid functional may also affect.

Many thanks in advance,
Ibrahim

User avatar
SKM
Full Member
Full Member
Posts: 125
Joined: Wed Oct 30, 2019 5:39 am
License Nr.: 5-516

Re: Bandstructure by hybrid functional

#2 Post by SKM » Fri Dec 24, 2021 2:25 pm

hi
i did not do this way, as i mostly link hybrid functional i.e. HSE06 with Wannier90 option. So, i cannot give much tech input for your specific problem. Based on the method i follow, i can guess some answer to your query.

in my view: [someone correct me, if my perception is wrong]
if you want to do ionic relaxation, either you may do the standard DFT itself with hybrid functional tags. or after standard DFT, repeat the run with Hybrid functional, taking the CONTCAR from previous DFT run.

But you should NOT combine, band structure calculation with NSW>0 and its wasting resources.

Regards
Regards
SKM

fabien_tran1
Global Moderator
Global Moderator
Posts: 316
Joined: Mon Sep 13, 2021 11:02 am

Re: Bandstructure by hybrid functional

#3 Post by fabien_tran1 » Fri Dec 24, 2021 6:38 pm

Hi,

To optimize the geometry also with HSE or not, depends on the importance of geometry for your final goal (does the considerd property depend sensitively on the geometry) and the computational cost it involves (can you do it in a reasonabe amount of time?).

From the technical point of view, NSW>1 should work.

IBRAHIM
Jr. Member
Jr. Member
Posts: 82
Joined: Mon Jan 30, 2017 6:00 pm
License Nr.: 5-2551

Re: Bandstructure by hybrid functional

#4 Post by IBRAHIM » Mon Dec 27, 2021 6:31 am

Thank you very much.

Post Reply