The imaginary part of the absorption spectrum exceeds 100 within GW+BSE level.

[Luo Yan]

Dear Professor,
I used yambo to calculate the absorption spectrum of GW + BSE for a series of two-dimensional materials, but the imaginary part of the absorption spectrum of this structure is more than 100. Other absorption spectra are reasonable. Here, datas used in drawing is "o.alpha_q1_haydock_bse". Thanks for your kind helps.

[Daniele Varsano]

Dear Luo Yan,
please sign your post with your name and affiliation, this is a rule of the forum and you can do once for all by filing the signature in your user profile.

My impression is that the high value of alpha is due to the fact you have an absorption peak at very low energy, the lower the energy the higher is the oscillator strength in
order to fulfil the TRK or f-sum rule.

Best,
Daniele

[Luo Yan]

Thanks a lot. But I don't known how to deal with this case. Could you help me with some solutions ? Another question is that the file of " o.alpha_q1_diago_bse " plotted in absorption spectra is generated sometimes, but sometimes not in different calculations.

[Daniele Varsano]

Dear Luo,
tha alpha is plotted when you use a coulomb cutoff, i.e. when you are dealing with a reduced dimensionality system, as in this case the eps_M is not well defined.

Thanks a lot. But I don't known how to deal with this case. Could you help me with some solutions ?

I do not understand well what do you mean, what's the problem to deal with it?
Please note that in MoS2, which have an absorption onset at higher energy of your system the alpha is larger than 50 A (see. e.g. PHYSICAL REVIEW B 88, 245309 (2013)).

Note also the the absolute eight it is not a physical quantity as it depends on the broadening, the oscillator strength is the area beneath a given peak, so you can raise or lower it assigning a different broadening (damping) in input.

Best,
Daniele

[Luo Yan]

Dear Daniele,
Thanks for your helps. In this calculation, the first peak in absorption spectrum is larger than 100. From your former answers, that's OK ?? If not, how can I solve this problem ? I have done some try, but failed. To be honest, I am a beginner using advanced yambo code.

[Daniele Varsano]

Dear Luo,
I've not arguments to say it is not correct. Again, what matters is the oscillator strengths and not the absolute height of the peak. The eight itself will depend on
the BDmRange you used.

Best,
Daniele

[Luo Yan]

Dear Dr. Daniele
Thank you for your patience. Another problem, tha alpha is plotted when I studied the low-dimensional system, but the "o.alpha_q1_haydock_bse" file didn't generated sometimes. I still have no idea to deal with this situation. One input is attached.

[Daniele Varsano]

Dear Luo,
you need to provide more information otherwise is impossible to spot the problem.
You need to attach (see bottom below) the input/report/output files and any other relevant info useful to spot the problem, in the cases when you have and have not the output file. In order to attach file you need to rename them with an allowed suffix (.zip, .txt etc).
Best,
Daniele

[Luo Yan]

Dear Dr. Daniele
Thank you very much for your prompt reply. Some files have added.

[Daniele Varsano]

Dear Luo,

looking at your input you can recognise;

Code: Select all

CUTGeo= "none"

In this case, even if you have assigned a CutBox values this is ignored by the code and a 3D 1/q^2 potential is considered. This is the reason why you do not have the alpha output.

Other way to check this kind of problem is to look in the report if the ndb.cutoff is written or read from previous calculations. As you can see in your report there is no trace of it.


Best,
Daniele