Scissor calculation for electron-phonon coupling

Hello users,
I am running DFPT&EPH for calculating the mobility and thermoelectric transport properties of semiconductor. But I want to do this calculation by adding the scissor correction, because GGA not produces correct band-gap. I have searched in abinit input parameters and find a variable mbpt_sciss. Thus,

  1. Can we used this as a scissor value for mobility and others transport, because it also works for optdriver 7 . In the last line of abo, why I got error of Fermi energy.
    mgsi.abo (59.4 KB)
    mgsi.abi (2.8 KB)
    mgsi.log (144.7 KB)

  2. After adding mbpt_sciss of 0.68 eV in my input file, the fundamental gap (in line 734 of abo) is remaining same as DFT value. But direct band-gap changed after this scissor. How this is happening

  3. One thing also, in eph for mobilities - abinit, I have used this scissor correction only in step 5. I think it is not necessary to take this value in step2&3, because they calculates only phonon and electron eigenvalues separately. Am I correct?

Please let me know if I did it wrongly.

Thank you in advance

Hi!

The variable is indeed active, but only in a specific case, for double grid integration and in the eph self energy calculation. Should be generalized, which would not be too hard. If you are doing a multi step calculation, with transport on top of the self energy, then you may be able to impose the sciss by using a double grid as well.

M

In file src/78_eph/m_sigmaph.F90 (eph_task ± 4)

if (new%use_doublegrid) then
! Note that we don’t recompute %fermie and %occ in ebands_dense, only %nelect must be consistent with the
! input ebands to handle possible doping
ebands_dense%nelect = ebands%nelect
ebands_dense%fermie = ebands%fermie
if (abs(dtset%mbpt_sciss) > tol6) then
! Apply the scissor operator to the dense mesh
call wrtout(std_out, sjoin(" Apply the scissor operator to the dense CB with:",ftoa(dtset%mbpt_sciss)))
call ebands_apply_scissors(ebands_dense, dtset%mbpt_sciss)
end if
end if

Thank you sir for replying
But I have applied scissor on a double grid fine mesh and then calculate transport on the top of self-energy (in the previous attached input file).

@gmatteo @gabriel.antonius any ideas here?

@vinod, please ignore my reply in another thread from 2023, this one is up to date.