issue with fermi level value

[bhargabkakati]

Dear experts,
I am trying to calculate the band structure of a system using vasp and quantum espresso. In both the cases, the band structure is almost same but there is a mismatch in the position of fermi level. VASP yields a p-type semiconductor while QE gives an n-type semiconductor. What might be the issue here? I have attached the snaps of band structures along with input files. Any help would be greatly appreciated.

Thanks.

[manuel_engel1]

Hi,

Thanks for reaching out. I'm assuming that these calculations are carried out in the ground state at 0K temperature (no electronic excitations) without any doping. In this case, the Fermi level is not uniquely defined for gapped materials. Any energy inside the gap fulfills "all states below are filled and all states above are empty". Therefore, the Fermi levels produced by both codes are equally valid in this scenario. If, on the other hand, the electronic occupancies in the material are somehow smeared (for example due to finite temperature) or if the system is metallic, then there will be non-empty (usually partially filled) states above the gap. In this scenario, the Fermi level or chemical potential are usually well defined and both codes should give the same result.

To know what is really going on in your calculation I would need to see the output of the calculation. You can check the OUTCAR file and look at the electronic occupancies. You could do the same for the QE output.

[bhargabkakati]

Hello,

My apologies for the late reply. I have attached all the output files of QE and OUTCAR of vasp for your convenience.

Thank you.

[manuel_engel1]

I had a quick look and it has confirmed my initial assumption: You have a semiconductor where the electrons fill all states below the gap and none of the states above the gap. The Fermi level is just not uniquely defined in this case. Any choice that falls inside the gap is a valid choice and so it makes sense that different codes (and even different pseudopotentials) might give different results.

[bhargabkakati]

Thank you so much for your insight into this issue.