My Community

Dear fellows,

I am trying to calculate the band structure of GaAs using the HSE06 hybrid functional using the KPOINTS_OPT file and Vasp 6.4.0. When I plot the band structure using py4vasp, I only get the energies on the automatic k-points mesh of the SC calculation, not in the high symmetry points along the path specified by the KPOINTS_OPT file (image below. The amount of k points is small to reduce the run time for testing). The input and output files of the calculation are in the zip file attached.
Could you help me with this?

Best regards,
Renan Lira.

First problem is that KPOINTS_OPT does not work together with NSW>0. It simply runs the calculation but won't use the KPOINTS_OPT file. In a successful run you should see the following lines in the stdout after the regular SCF convergence:
k-point batch [1-3\24]
...

So I suggest first optimize your structure with IBRION and NSW>0. After that remove the IBRION and NSW line (or set NSW=0) and run with the KPOINTS_OPT file.

For the plotting with py4vasp you have to specify that the KPOINTS_OPT file is used:
mycalc.band.plot(source="kpoints_opt")
In order to find this you could also write help(mycalc.band.plot).
It is also covered in this talk:
https://www.youtube.com/watch?v=zNc0bx8FTlU

Just to add to my initial post, I have tried performing a SC calculation with PBE, copying the WAVECAR file into a new folder and performing the calculation with HSE06 in the new folder. The output is different now and the PROCAR_OPT file was successfully generated; however, when plotting the band structure I get the same results as before.

The INCAR files for the PBE run and HSE06 run, along with the OUTCAR and PROCAR_OPT files for the HSE06 run are attached below. The rest of the input files are the same as posted before.

Best regards,
Renan Lira.

Thank you very much for your reply.

Optimizing the structure first with IBRION and NSW>0 and then running again without IBRION and NSW tags produced the PROCAR_OPT file and the lines
k-point batch [1-3\24]
...
were present in the stdout file.

Specifying the KPOINTS_OPT when plotting the bands on py4vasp worked too. However, the command I used was mycalc.band.plot("kpoints_opt") without the "source=", which produced an error (see image below). The help(mycalc.band.plot) command was not very helpful, as you can see on the image.

I was not able to use the Microsoft teams link to check on the talk.

Best regards,
Renan Lira.

Sorry, I opened it in teams and sent you the wrong link.
Here is the corresponding youtube link:
https://www.youtube.com/watch?v=zNc0bx8FTlU

I am also correcting the link in the original post.