Yambo Community Forum

Dear Daniele,
I have checked my "o-BSE.exc_E_sorted" file and find the positive and negative value are almost symmetrical. In my corresponding TDA calculation, the lowest exciton energy is negative and shows the excitonic instability.
Could you please tell me how can I (or in which output file) find "the negative eigenvalues having v-->c components in their eigenvectors" in full BSE calculation?

Thanks so much!
Best,
Jing

Dear Jing,
in TDA only v-->c transitions are taken into account.
In order to inspect the transitions forming the exciton you need to print the weights !Acv!^2 as explained in the tutorials e.g:

http://www.yambo-code.org/wiki/index.ph ... 021_school

Daniele

Dear Daniele,
Thanks for your kind reply!

Best,
Jing

Dear Daniele,
I have calculated the weight of different exciton states after the full BSE calculation.

1. In order to look at exciton instabilities, I chose some of the negative excitations to check whether there are v-->c components in their eigenvectors.
I have uploaded the "o-BSE.exc_E_sorted" file and also the "o-BSE.exc_weights_at_110" file.
I found there are v-->c components in this negative excitations. Does it means that there exists exciton instability in this system? The corresponding energy is "-0.2704 eV". Does it means that the lowest exciton transition energy is "-0.27 eV"? (My TDA calculation shows the lowest exciton energy is -0.27 eV).
But however, I also did a test of the weight of state "175" (the corresponding energy is -1.019 eV), there are also v-->c components.
So I am confused that how to verify the lowest exciton transition energy?

2. I also did a full BSE test in a semiconductor system (not an intrinsic excitonic insulator predicted by TDA). But I found the weight of several negative excitations also have the v-->c components, does it means that there are also exciton instabilities in this system? Not as predicted as TDA?

Thanks so much!

Best,
Jing

Dear Jing,

if the coupling has an effect of course v->c and c->v do mix (by definition!) so you will have always a small component from both transition.
My warm suggestion is to stay in TDA approximation (as you also verified that coupling does not have great effect) where the picture is more clear for interpretations.

Best,
Daniele

Dear Daniele,
If the "coupling" has a great effect, how to judge the excitonic instability?
It seems that the "weight" can not give a specific criterion.

Thank you!

Best,
Jing

Dear Jing,
if the system is unstable wrt e-h formation it will appear as negative eigenvalue in the TDA.
Coupling can be seen as a perturbation to the resonant and anti-resonant part.

If unstable you will have negative eiegenvlues with mainly v->c components (resonant) and positive eigenvalues with mainly c-->v (antresonant) component.

Daniele

Dear Daniele,
Thanks so much!

Best,
Jing