GW-BSE

[fatimazahra]

Hi
I want to calculate GW-BSE, if I use

Code: Select all

 "E < SAVE/ndb.QP"

, I got that excitation enrgy is lower than GW-gap,
But if I use

Code: Select all

% KfnQP_E
 0.000000 | 1.000000 | 1.000000 |      # [EXTQP BSK BSS] E parameters (c/v)

I got that excitation enrgy is lower than GW-gap.
What is the difference between them ?
Thanks in advance

[Daniele Varsano]

Dear Fatima,
it is not very clear what you mean. In both cases you get excitation lower than GW gap, which is the case when your system has bound excitons.
In the first case you are considering QP corrections previously calculated (at GW level I suppose). In the second case you are considering KS energies. BSE framework requires quasi particle energy, so even you read it as the first case, either you apply a scissor using KfnQP_E, but you set is as 0, so you are not applying it.
Best,
Daniele

[fatimazahra]

Dr Daniele
Tanks for your quick reply.
in this case:

Code: Select all

% KfnQP_E
 3.700000 | 1.000000 | 1.000000 |      # [EXTQP BSK BSS] E parameters (c/v) 

I got that excitation energy is higher than GW-gap. I don't know what is the problem. normally I must find the same result in both cases.

[Daniele Varsano]

Dear Fatima,
I cannot say much as I do not know what system you are dealing with and what 3.7eV is here.
There you should put the QP correction to the KS states (ie E^GW-E0),and not the GW gap. In order to understand a possible difference you can:
1) plot the independent QP spectra (last columns of the BSE output), and compare the BS spectra in both cases
2) Analyze which KS transition participate in the first excitation and look if a scissor operator is a good approximation for it.

Best,
Daniele

[fatimazahra]

Dr Daniele
Thank you very much.You help me a lot!. I understand now