Band structure with hse06

[shishir1]

I am getting band structure shift upward with hase06. Fermi energy is not at zero.
Please help me to get the correct band structure. I have attached the files here.

[henrique_miranda]

Could you share a figure with the band structure so I understand what you mean?
Do you have a semiconductor or a metal?

[shishir1]

Could you share a figure with the band structure so I understand what you mean?
Do you have a semiconductor or a metal?

band_hse2.png

here is the figure. The structure is heterostructure of graphene/Mos2.

[henrique_miranda]

Ok, now I understand.
VASP does not shift the eigenvalues so that the fermi energy is at zero. You need to do that yourself.
The value of the fermi energy is reported in the OUTCAR file.

[shishir1]

I tried the same with pure Graphene but there is the shift of Fermi at 0 with and without hse06. Even calculating Band Structure for Graphene/MoS2 without hse06 the Fermi is at zero.

band.png

[henrique_miranda]

Did you shift the eigenvalues according to the fermi energy?
Which tool are you using for plotting?

[shishir1]

No I didn't shift it for pure graphene. The fermi was exactly at zero for graphene with HSE06. I used origin to plot.

[henrique_miranda]

Ok, could you share the OUTCAR of the two calculations for which you posted the band structure?

[EDIT] OK, I think I know what the problem is: you need to get the fermi energy from an SCF calculation on a regular grid to shift the eigenvalues from your NSCF calculation.
Have a look at the line "Fermi energy:" in your OUTCAR file.

If you don't want to do two separate runs of VASP you might consider using the KPOINTS_OPT file perform an NSCF calculation along a path starting from the density you obtained in the SCF calculation.
The procedure is described on this page: wiki/index.php/Band-structure_calculati ... unctionals

[shishir1]

Here are the BAND.dat and OUTCAR for graphene/MoS2 with and without hse06.

[henrique_miranda]

The tool you are using to produce the BAND.dat file is shifting the eigenvalues according to the fermi energy.

band/OUTCAR

Code: Select all

Fermi energy:         2.0543352579

 k-point     1 :       0.0000    0.0000    0.0000
  band No.  band energies     occupation
      1     -17.0268      2.00000

band/BAND.dat

Code: Select all

#K-Path(1/A) Energy-Level(eV)
# NKPTS & NBANDS:  60 512
# Band-Index    1
   0.00000    -19.081153

This last value -19.081153 is -17.0268-2.0543352579

The same goes for the file in band_hse/OUTCAR:

Code: Select all

 Fermi energy:         0.8727577133

 k-point     1 :       0.0000    0.0000    0.0000
  band No.  band energies     occupation
      1     -19.5942      2.00000

band_hse/BAND.dat

Code: Select all

#K-Path(1/A) Energy-Level(eV)
# NKPTS & NBANDS:  15 294
# Band-Index    1
   0.00000    -20.466959

which again -20.466959 is -19.5942-0.8727577133.

In the case of HSE, you are doing a regular grid + zero weight k-points. The fermi level in this case should be the correct one.
For PBE it looks to me like you are doing a NSCF calculation along a path. In this case, the fermi level is not correct.
You need to compute the fermi level on a regular grid. That's why I suggested using the Fermi level from the SCF calculation.

I hope this helps!