ABINIT Discussion Forums

I calculate Band structure for Ag in both ELK and ABINIT codes. Though the band structure is similar in all directions but surprisingly the Ef is shifted upward ~6.3 eV in ABINIT as compared to ELK. The actual Fermi energy in ABINIT is as follows.

Fermi (or HOMO) energy (hartree) = -0.03070 Average Vxc (hartree)= -0.49252

So any idea why this 6.3 eV mismatch between these two code ?

Thanks

Raj

NB : Ag.in is also attached.

Note that ELK is an all-electron code while with ABINIT you are computing only the valence electrons (as defined by the pseudopotential file), therefore the "zero of energy" may be different between the two codes. The Fermi level in both calcs appears to be at about the same place with respect to the band features so I think the two calculations are consistent with each other, however, I think your abinit calc may not be well converged because it appears to be so jagged.