QE wedge versus EPW wedge

[Vahid]

Dear EPW users,

When I run graphene with a 500x500 k-grid, QE reduces the job to 21084 k-points in the wedge as expected. However, EPW creates a 41917 k-point wedge with the same k-grid. The issue is likely due to the use of a Gamma-centred BZ by QE versus a [0:1] k-grid by EPW. Are there any unwanted consequences with getting the grid.f90 to generate Gamma-centred k-grids particularly when mp_mesh_k is set to .true.?

Thank you,
Vahid

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

[Shashi]

Dear Vahid,

EPW should generate the same grid as QE. Does your system break inversion symmetry but preserve time-reversal symmetry? If so, I would suggest trying the QE development version, where this has already been implemented.

Would it be possible for you to share your input and output files? That would help me better understand the issue and provide a more accurate solution.

Regards,
Shashi

[Vahid]

Dear Shashi,

Attached please see the EPW input files, and the epw.out file. The scattering rate includes 41917 kpoints.

Attached is also the file mesh_k output by PW/tools/kpoints.x. There are 21084 wedge points.

It is possible that the issue is related to the use of two different carbon atoms, C and C1 in order to generate the Wannier functions for a larger energy range.

Thanks,
Vahid

[Shashi]

Dear Vahid,

I tested your input files and found that EPW correctly generated 21,084 wedge points, consistent with the QE results.

I did not make any changes to your original input files; I only split the EPW calculation into two separate inputs for testing purposes.

Please find the attached input and output files. Let me know if you have any questions.

Best regards,
Shashi

graphene_shashi_debug.tar

(208 KiB) Downloaded 8 times

[Vahid]

Dear Shashi,

It seems that the reason we are getting two different wedge points is the EPW version. I was using QE-7.0 and in that version of EPW, wedge points are not the same as the latest development QE version. Perhaps this was a bug which got fixed at some point but there is no documentation on how it got fixed.

Best,
Vahid

[Shashi]

Dear Vahid,

You are right. The earlier version of EPW did not include the implementation of time-reversal (TR) symmetry with broken inversion symmetry.
In your system, the TR is preserved, but inversion is broken.

In the latest version, time-reversal symmetry with broken inversion is implemented. Please use the latest version.


Regards,
Shashi