The problem is quite confusing to me. I calculated the band of graphene using both vasp and dmol and got two different band structures. They are almost the same below the fermi energy, while having some significant difference above the fermi level.
To make things more interesting, I try to project the band structure into atomic orbitals (e.g. s, p, d, f) using PROCAR and find some bands have a total projection parameter MUCH less than 1. Here "total" means a sum over the projection parameter of every atom's every orbital, which I suppose should be 1 (or at least almost 1).
Coincidently, those bands with a small total projection (actually near 0) are just the difference between vasp and dmol. That is to say, if I delete these bands, the band structure of vasp and dmol would be quite alike.
I have two question about this:
1. Should the total projection of every atom and orbital be 1? What contributes to the rest part of the band, if all atom's contribution doesn't equal 1?
2. Is the bands with small total projection reliable? Should I exclude them in analysis?
I'm sorry if I didn't make the problem clear, for I'm still learning English. Any response is appreciated!
