My Community

Hi,

I would like to calculate the MBJ band gap with SOC for my structure.
The calculation with PBE-SOC with ZERO KPOINTS runs comfortably in the beginning and then after so many iterations, it crashes or stops. I could not sort out the issue. Please advise.

Dear rishikanta_m,

From your question I understood that you are interested in finding the band gap with MBJ functional, but you did not select the MBJ functional in your INCAR.
Before doing a calculation with zero-weight k-points, you should do a SCF calculation with a regular k-points grid and add zero-weight k-points in a subsequent non-self-consistent calculation (see detailed instructions here and here).
Are you able to reach the convergence with the regular grid?

Yes, I achieved convergence with regular grid.
As you mentioned, the MBJ requires WAVECAR from PBE-SCF calculation (regular k-grid with zero-weighted k) to start the calculation.
The attachment, which I have earlier posted belongs to PBE-SCF regular k-grid with zero-weighted k.
The issue is with this calculation with the inclusion of SOC. The calculation is stopped after 24 hours of calculation due to issue, mentioned in vasp.out. I am not able to comprehend the cause of the error.

So if I understood you correctly, you were able to achieve the convergence for an SCF calculation without zero-weighted k-points. But in the attached files, the PBE calculation with zero-weight k-points starts from scratch (ISTART = 0 and ICHARG = 2). Have you tried starting you calculation from a previously converged charge density?

No, I have not started from a previously converged charge density.
So, to start the PBE calculation with zero-weight k-points from there, will ISTART = 0 and ICHARG =1 work? What else parameters need to be provided?

Since you are interested in the band gap calculated with the MBJ functional, you should restart your calculation from the WAVECAR file, i.e., set ISTART = 1 and ICHARG = 0.