negative excitation energies

[martinspenke]

Dear Developers,

I am studying a 2D system with a relatively small band gap of 0.95 eV.
The problem is that inclusion of excitonic effects redshifts the spectrum to the negative region (negative x-axis)
(negative excitation energies can be seen by exact diagonalization of the BS matrix).

Therefore, the BSE spectrum lacks absorption onset. I have no clue how to interpret the BSE spectrum. Does this mean that the BSE spectrum is meaningless ?

Best wishes
Martin

[amolina]

Hello,

if it is a 2D material and the gap is 0.95 eV, it is possible that the exciton binding energy is bigger than 0.95 eV. My advice would be to run a GW calculation to obtain the GW gap and apply the scissor operator in yambo accordingly. This will fix the issue.

Alejandro.

[Daniele Varsano]

Dear Martin,
from your plot it looks that you already considered GW corrections, so the 0.95eV gap is it already at GW level?
Being a 2D system have you considered coulomb cutoff potential, or at least the RIM when considering coulomb integrals?

Best,

Daniele

[martinspenke]

Dear Alejandro and Daniele,

from your plot it looks that you already considered GW corrections, so the 0.95eV gap is it already at GW level?

yes, 0.95 eV is the QP gap.

Being a 2D system have you considered coulomb cutoff potential, or at least the RIM when considering coulomb integrals?

yes, i considered both the cutoff and RIM, since i was using the box.

However, it seems that there are such massive excitonic effects, that it simply shifts the spectrum into the negative regions !
I also saw this behavior in other 2D systems. Whenever there is a small gap, there is a good chance that the absorption onset shifts to the negative area.
I can not understand this.

Bests
Martin