Dear all
Is it technically and physically making sense to use the QPS file obtained by setting gwcalctyp=28 in the standard one-shot gw calculation for band interpolation with MLWFs scheme, instead of using the QPS file from the self-consistent gw. I've tried this and it seems to give the reasonable band structure of Si, with Eg around 1 eV.
Kind regards,
Thanusit.
[SOLVED]GW band structure by standard GW+MLWFs
[SOLVED]GW band structure by standard GW+MLWFs
Last edited by thanusit on Thu Dec 02, 2010 3:06 am, edited 2 times in total.
Re: GW band structure by standard GW+MLWFs
Dear Thanusit
The QPS file contains the quasiparticle amplitudes and energies calculated by diagonalizing the
Hermitian part of the self energy sigma_ij in the bdgw subspace.
These results are close to the perturbative G0W0 output only when self-consistency has been
achieved.
In your case the code is using the self-consistent results of iteration number 1 for the wannier
interpolation hence
you interpolated band structure is not equivalent to the interpolation of the G0W0 energies.
Matteo
The QPS file contains the quasiparticle amplitudes and energies calculated by diagonalizing the
Hermitian part of the self energy sigma_ij in the bdgw subspace.
These results are close to the perturbative G0W0 output only when self-consistency has been
achieved.
In your case the code is using the self-consistent results of iteration number 1 for the wannier
interpolation hence
you interpolated band structure is not equivalent to the interpolation of the G0W0 energies.
Matteo