Merge fo excitonic states

[acastanedam]

Dear Yambo team

I've seen nearly degenerate excitonic states are "merged" when are post-processed to produce a wavefunction output. What is meant exactly by "merged?. I guess, a linear combination
of degenerate states is made, but, what are the weights?, the oscillator strength of each state as it's given in the o.exc_E_sorted?.

Thanks in advance for your help,

Arcesio Castañeda Medina
Max-Planck Institute for Microstructure Physics
Weinberg 2
06120 Halle
Germany

[Daniele Varsano]

Dear Arcesio,
yes, it is a linear combination,
The weight of the transitions of degenerate states are summed up.
The excitonic wavefunction is constructed as sum of the excitonic wavefunctions of the
degenerate states, which are defined as the sum of the \pis_v\psi_c weighted with the
coefficients of the suolution of the Bethe Salpeter equation. If they have big or small
oscillator strength (optical allowed or forbidden) does not infer in the character of the
excitonic wave function.
If you do not want to cast the degenerate wfs because you think they have different
character you can change the threshold of degenearcy changing the "deg_energy" variable
in ./src/ypp/ypp_init.F which is fixed to 0.01 eV.

Hope it helps,

Daniele

[acastanedam]

Dear Daniele

Just to clarify, in the LiF case (tutorial) which has a three fold degenerate excitonic eigenvalue around 12.0 eV, when I request for the excitonic wavefunction
what I get is:

\psi_exc=\psi_1+\psi_2+\psi_3

provided the indices for the degenerate states are 1, 2 and 3 (from the o.exc_E_sorted file)?. Of course, if I understood well, each one of these states are constructed by the procedure you've described.

Best regards and thanks for your help,

Arcesio

[Daniele Varsano]

Dear Arcesio,
yes, it is right.

Daniele