HSE + spin orbit coupling

Question on input files/tags, interpreting output, etc.

Please check whether the answer to your question is given in the VASP online manual or has been discussed in this forum previously!

Moderators: Global Moderator, Moderator

Post Reply
Message
Author
vrinsom
Newbie
Newbie
Posts: 5
Joined: Sat Dec 14, 2019 4:23 am

HSE + spin orbit coupling

#1 Post by vrinsom » Wed Aug 03, 2022 10:15 pm

Hello,
I am trying to carry out spin orbit coupling calculations with HSE. I generate the WAVECAR and CHGCAR from a collinear run, and then follow the guidelines given in https://www.vasp.at/wiki/index.php/LSORBIT to run the non-collinear calculation. However, my job fails with the error

ERROR: while reading WAVECAR, plane wave coefficients changed 30310
15155

I was wondering why this error occurs and how to correct it? I use ISYM = 0 for both the collinear and non-collinear runs. I am running only on one k-point (gamma).

Any guidance would be greatly appreciated. Thank you!

alexey.tal
Global Moderator
Global Moderator
Posts: 228
Joined: Mon Sep 13, 2021 12:45 pm

Re: HSE + spin orbit coupling

#2 Post by alexey.tal » Thu Aug 04, 2022 2:37 pm

Hello,

The gamma version of VASP only requires half of the plane waves due to the symmetry at the Gamma point and hence the std and ncl versions cannot read the WAVECAR produced by the gamma version.

A similar issue has been discussed here.

vrinsom
Newbie
Newbie
Posts: 5
Joined: Sat Dec 14, 2019 4:23 am

Re: HSE + spin orbit coupling

#3 Post by vrinsom » Thu Aug 04, 2022 2:50 pm

Thank you for the clarification! Yes, I realized that the ncl version cannot read the WAVECAR generated by the gamma version due to the reduced number of plane waves in the gamma version.
But when I generate the WAVECAR using the std version (with one k-point), I still face this issue. I wonder why is this the case? Thank you!

alexey.tal
Global Moderator
Global Moderator
Posts: 228
Joined: Mon Sep 13, 2021 12:45 pm

Re: HSE + spin orbit coupling

#4 Post by alexey.tal » Fri Aug 05, 2022 8:48 am

I see what you mean now. If the spin-orbit coupling is included, the orbitals are represented by spinors with up and down components, hence the number of required plane waves is doubled. In the LSORBIT wiki page, it is implied that the collinear calculation should be spin-polarized then the basis is consistent between the collinear and non-collinear calculations. Also, be aware that you should switch off the symmetry (ISYM=-1), otherwise the number of k-points in the non-collinear case will be different.

Post Reply