[Wannier] From wien2wannier to postw90.x

[Marcelo Barbosa]

Dear Sirs,

I’m interested in calculating the shift current using postw90.x with the berry option.
I ran example 25 from the tutorials (shift current in GaAs) and everything worked fine.

For my DFT simulations I usually use WIEN2k, so I decided to compare a calculation starting with GaAs on WIEN2k and then using the package wien2wannier to calculate the wannier functions.
The obtained band structures match, so I believe this worked fine as well.

However, the shift current from this calculation using the berry option in postw90.x is wrong.
Since WIEN2k calculates the Fermi level in Ry, I converted the value to eV and that’s what I used for fermi_energy in the file case.win.
Then, I tested setting fermi_energy = 0 and not only the shift current varied a lot but it’s actually much closer to the value obtained in example 25.

Could you please tell me which value should I use in fermi_energy after converting a simulation from WIEN2k to Wannier90 using wien2wannier?
Should it be the Fermi level from WIEN2k in Ry, eV or zero?

Thanks.

Best regards,
Marcelo