Yambo Community Forum

Dear Shudong,
It should work by putting:
BSEmod= "coupling" # [BSE] resonant/causal/coupling
BSKmod= "Hartree"

Cheers,

Daniele

Thank you Daniele for your quick reply.
If I only want to get the singlet excitonic absorption,I should set

BSEmod="resonant(or causal) # "causal" # BSE resonant/causal/coupling
BSKmod="BSE" # "" # BSE IP/Hartree/HF/ALDA/SEX/BSfxc

and for triplet shoud set

BSEmod= "coupling" # [BSE] resonant/causal/coupling
BSKmod= "Hartree"

Thanks!

Dear Shudong
sorry forget what I wrote before which is clearly wrong, what I wrote you before is a Time Dependent Hartree (or LF-RPA) calculation.
If you want a singlet calculation you should set:

BSKmod="SEX" #
and
BSEmod="resonant (or coupling)
Resonant or coupling it will depend if the Tamm-Dancoff approximation is valid or not and this is system dependent. In bulk wide gap semiconductor is usually valid, in molecules it is not. In other cases has to be verified. In polymer and GNR I found that it was verified.
You can also use the causal mode, but for the imaginary part should give the same results.

Now: with the new command the triplet it is not straightforward:
A way to do that is to set the same input variable of the singlet but setting:
BSENGexx=0

hoping the code will not complain. Let me now if it works, and I will have a look at that, in order to make the input more intuitive.
Cheers,

Daniele

Thanks Daniele!
I will try it and let you konw soon.
Ciao.

Dear Daniele,
I have got the triplet and singlet spectra but I found it is identical. As you said I set all the parameters the same values except in triplet BSENGexx=0. In the out put file there exist some parameters as following:
BSK|Identifier : 7380
|Dimension : 2268
|Bands : 23 28
|Exchange [res]:yes
|Correlation [res]:yes
|Kernel`s coupling :no
|Exchange [cpl]:no
|W interaction is bare :no
|Correlation [cpl]:no
|ALDA kernel in R-space :no
|RL vectors [exchange]: 1 (for singlet it is 12 Ry)
|RL vectors [correlation]: 1533
You see the "Exchange" is yes instead of "no".
And My input is:
# | Chimod= "Hartree" # [X] IP/Hartree/ALDA/LRC/BSfxc
# | BSEmod= "causal" # [BSE] resonant/causal/coupling
# | BSKmod= "SEX" # [BSE] IP/Hartree/HF/ALDA/SEX/BSfxc
# | BSSmod= "d" # [BSS] (h)aydock/(d)iagonalization/(i)nversion/(t)ddft`
# | BSENGexx= 0 Ry # [BSK] Exchange components (for singlet is 12 Ry)
# | BSENGBlk= 5 Ry # [BSK] Screened interaction block size
# | #WehCpl # [BSK] eh interaction included also in coupling

Does this relates to my using "causal" or any other wrong parameters?
Thanks!

Dear Shudong,
ok evidently it is not working in this way and it is needed to change few lines in the code by hand, before we do it in a clean way from input file.
Anyway let me ask a couple of question: The two calculations are similar or identical?
This make difference as even if it is not working as you have instead of 0.
they should not be identical as it looks that the exchange it is not playing any role and it is converged considering only the first G-vector and it sound me quite strange. Please check from the report if in the triplet calculation yambo is recalculating the BS matrix or it is just reading it (check if there are WARNINGS indicated with (WRN)).
Next I would use:
and uncomment this variable: You can also try to set: instead of Ry, but I do not think it will work.

Finally let me know which release you are using and I will tell which lines have to be changed in the source.
Daniele

Dear Daniele,
In attached is the two spectra, and it seems identical.
There is no WARNING error in output file since I used a new SAVE for triplet calculation.
The version I am using is Version 3.4.0 Revision 3041.
I will try it using new parameters as you said in last reply and let you konw.

Thanks a lot!

Dear Shudong,
OK, Try to use RL instead of Ry, but as i told you before, I do not think that it will be successful. I will send you the patch for the code by mail later on.
In the meanwhile I think it is useful to understand what is happening in your system: can you perform singlet calculations varying the amount of exchange g-vectors, and look if the spectra it is totally unaffected? Moreover please have a look directly at your output instead of plotting them in order to check if they are identical or just very similar.
Thanks,
Daniele

Thanks Daniele!
I will do that.
Ciao.