Page 1 of 1

Post process in TPA

Posted: Wed Dec 25, 2024 8:51 am
by jasonhan0710
Dear Claudio,

I have noticed a Python script that can get the TPA coefficient, TPA coefficient. However, I am confused with the input files

Code: Select all

directory='./'
path_rt_lI_int1='o-RUN1.YPP-X_probe_order_1'
path_rt_lI_int2='o-RUN1.YPP-X_probe_order_2'
path_rt_lI_int3='o-RUN1.YPP-X_probe_order_3'
It seems that the input file is chi(1), chi(2) and chi(3) at the same intensity. According to the formula in https://wiki.yambo-code.eu/wiki/index.p ... note-tpa-1, the TPA coefficient should be calculated with polarization at different laser intensity. So, it seems reasonable to change the input files as

Code: Select all

directory='./'
path_rt_lI_int1='o-RUN1.YPP-X_probe_order_3'
path_rt_lI_int2='o-RUN2.YPP-X_probe_order_3'
path_rt_lI_int3='o-RUN3.YPP-X_probe_order_3'
Is it correct to change the input file name?

Besides, as Yambo has already output the chi(3) with ypp_nl, why should we use Richardson extrapolation to extract the TPA? Can we just calculate the TPA using equation (11) in the paper Phys. Rev. B 98, 165126 (2018)?

Thank you!

Best,
Jason

Re: Post process in TPA

Posted: Wed Jan 22, 2025 9:45 am
by claudio
Dear Jason

> Is it correct to change the input file name?

yes, it is correct, I will fix the wiki webpage.
Moreover we are preparing a new tutorial on TPA with better scripts that will be online in few weeks.

> Besides, as Yambo has already output the chi(3) with ypp_nl, why should we use Richardson extrapolation to extract the TPA?
>Can we just calculate the TPA using equation (11) in the paper Phys. Rev. B 98, 165126 (2018)?

you cannot calculate the TPA directly, as the SHG/THG etc.. because the TPA oscillate at the same frequency of the linear response,
therefore it is impossible to distinguish it in the Fourier analysis of the P(t).
The only case when you can get the TPA directly is if you choose a direction for the TPA where the linear response is zero.
Otherwise you are obligated to perform more calculation to get the coefficient that grow ^3.
best
Claudio