A and B Excitons

[ahsan_javed]

I tried to redo the work done by R. Ashwin in his paper "Large excitonic effects in monolayers of molybdenum and tungsten dichalcogenides" (2012).

I am able to obtain the optical spectrum for NON-SOC as well for SOC on, but i am not getting two peaks known as A and B excitons.

The relevant files are attached for WSe2 monolayer, the absorption spectra i obtain is part of BSE folder.

[alexey.tal]

Dear ahsan_javed,

It can be quite tricky to correctly set up a GW-BSE calculation.

Here are some issues that I noticed in your setup:

1. The main issue in your calculation seem to be the G0W0 step. In the OUTCAR file, there is a warning that the WAVEDER file was not found. It could be due to the fact that WAVEDER was indeed not present or just not compatible. The latter can be due to the fact that the number of bands in your GW calculation (252) exceeds the number of bands in Step 3 (231).
2. You use SCAN functional, but LOPTICS should not be used with METAGGA functionals without LPEAD. You can see a corresponding warning in the OUTCAR file of Step 3.
3. From your OUTCAR files I can see that you used the standard potentials. However, for the GW/RPA calculations, the recommended potentials are suffixed with "_GW" (see wiki). The reference paper used such potentials.

General recommendations:

• Always read the warnings as they can notify you of a problem and often suggest a solution.
• If you are trying to reproduce the results from the reference paper, you should use the same xc functional.
• If you don't have much experience running GW - BSE calculations in VASP, a good starting point would be to run the tutorials GW and BSE.

[ahsan_javed]

This message pops up in start of gowo calculation step. How to verify this i.e., that we are doing it right?


The Fermi energy was updated, please check that it is located mid-gap
values below the HOMO (VB) or above the LUMO (CB) will cause erroneous energies
E-fermi : -1.4308

[alexey.tal]

You can find the VBM and CBM by looking at the KS-energies printed in the OUTCAR file. According to OUTCAR from your G0W0 calculation, VBM is at -2.0918 and CBM at -0.7698, so the Fermi level of -1.4308 is correct.

[ahsan_javed]

Now on the BSE part, i am getting this error:

"The number of conduction band and valence band pairs is zero. This |
| indicates that NBANDSV or NBANDSO are not properly set up. The VASP |
| default is to set them to the number of occupied orbitals. Try to |
| set them manually."

I tried different values and also recommended by ashwin in paper i.e., 4 & 8 but error remains there. Then i commented out the NBANDS(V, O) both for VASP default values and error is now:

| CALCULATE_BSE (AMAT_SCALA) is not able to allocate 2.617e8 kB of |
| data on MPI rank 0. |

The updated files according to previous discussion for GoWo and unoccupied orbitals is uploaded here. POTCAR is _sv_gw for GoWo and BSE (but not for unoccupied case).

[alexey.tal]

I see that you have used different POTCAR in Step 3 and Step 4 and 5. This is incorrect. The potentials in all steps must be the same.

| CALCULATE_BSE (AMAT_SCALA) is not able to allocate 2.617e8 kB of |
| data on MPI rank 0. |

This error suggests that you cannot allocate 260 Gb. It is probably due to the fact that your BSE matrix is too large.