question about local PBE0 (useexexch) ?

[pipidog]

Hello abinit users,

I have a few question regarding local PBE0, i.e. the function controled by "useexexch"

1. I found the band structure can be calculated directly by using standard nscf calculation
when local PBE0 is truned on.Is it the correct way to obtain band structure and the fatbands?

I know most hybrid functional calculations require much more complicated way to obtain
band structure because the wave functions are also involved. So standard nscf run where
only density will be loaded won't work. If so is it okay to calculate band structure for local
PBE0 using standard nscf run? Is local PBE0 irrelevant to wave functions?

2. I did some paper survey on local PBE0 in abinit, such as PRB 80 235109. It looks like
local PBE0 is just an improved approach to parameter-free DFT+U. If so, if my local PBE0
calculation gives me a very different band structure than DFT+U, such as metal in DFT+U
but a large gap insulator in local PBE0 (on Cu-oxide 5d10 compounds, fully occupied in
DFT+U but partially occupied in local PBE0), should I believe it?

Any suggestion would be helpful.

Thanks,
Shu-Ting Pi
UC Davis