[QE-users] Band structure with hybrid functionals

[Vahid Askarpour]

Dear Dominik,

I went back to my old files since I did the HSE band structure a couple of years ago. You are correct that the procedure works for nqx’s =1.  For higher nqx values, I used the wannier code along with the open_grid.x code.

Cheers,
Vahid

On Oct 28, 2019, at 6:19 AM, Dominik Gresch <greschd at gmx.ch<mailto:greschd at gmx.ch>> wrote:

Dear Vahid,

Thank you for your help. Unfortunately this produces the same error, unless the q-grid is set to one -- in which case the results may not be converged. Did you have to choose the band-structure point in any particular way to match with the k-grid?

Best regards,
Dominik

Gesendet: Samstag, 26. Oktober 2019 um 23:33 Uhr
Von: "Vahid Askarpour" <vh261281 at dal.ca<mailto:vh261281 at dal.ca>>
An: "Quantum ESPRESSO users Forum" <users at lists.quantum-espresso.org<mailto:users at lists.quantum-espresso.org>>
Betreff: Re: [QE-users] Band structure with hybrid functionals
To add to my earlier post, and not having seen your input file, it is possible that you set a q-grid (nqx’s) that is not an integer multiplier of the k-grid. So if the k-grid is 8x8x8 and the q-grid is 6x6x6 or 3x3x3, the calculation will crash. A q-grid of 2x2x2 or 4x4x4 would work.

Vahid

On Oct 26, 2019, at 6:34 PM, Vahid Askarpour <vh261281 at dal.ca<mailto:vh261281 at dal.ca>> wrote:

I have calculated HSE band structure using the method you described. The only difference is that for the band-structure k-points, I used a weight of 0.000001 instead of zero.

Cheers,
Vahid

Vahid Askarpour
Department of Physics and Atmospheric Science
Dalhousie University,
Halifax, NS, Canada

On Oct 26, 2019, at 6:21 PM, Dominik Gresch <greschd at gmx.ch<mailto:greschd at gmx.ch>> wrote:

Dear QE users,

I am trying to calculate the band structure of a material with hybrid functionals. Since it is not possible to perform an NSCF run in that case, the desired k-points should be included in the SCF calculation. Since increasing the k-point mesh as a whole is rather expensive, I tried explicitly setting the list of k-points (in "crystal" mode), and then appending the desired k-points for the band structure as points with zero weight [0].

However, when trying this trick, I am faced with the following error:
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine exx_grid_init (117994):
k + q is not an S*k
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

My questions are as follows:
- does anyone have experience in computing the band structure with hybrids, and if so, is this the right approach?
- what is the nature of this error, and is there some work-around?

Thank you and best regards,
Dominik Gresch
--
Microsoft Research

[0] I am familiar with using this technique from experience using VASP (described here: https://cms.mpi.univie.ac.at/wiki/index.php/Si_bandstructure)
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso<http://www.max-centre.eu/quantum-espresso>)
users mailing list users at lists.quantum-espresso.org<mailto:users at lists.quantum-espresso.org>
https://lists.quantum-espresso.org/mailman/listinfo/users
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso<http://www.max-centre.eu/quantum-espresso>)
users mailing list users at lists.quantum-espresso.org<mailto:users at lists.quantum-espresso.org>
https://lists.quantum-espresso.org/mailman/listinfo/users
_______________________________________________ Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso<http://www.max-centre.eu/quantum-espresso>) users mailing list users at lists.quantum-espresso.org<mailto:users at lists.quantum-espresso.org> https://lists.quantum-espresso.org/mailman/listinfo/users
_______________________________________________
Quantum ESPRESSO is supported by MaX (www.max-centre.eu/quantum-espresso<http://www.max-centre.eu/quantum-espresso>)
users mailing list users at lists.quantum-espresso.org<mailto:users at lists.quantum-espresso.org>
https://lists.quantum-espresso.org/mailman/listinfo/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20191028/21b3ebfc/attachment.html>