dephasing and time step

[sohan]

Dear Sir,
Where to specify the parameters 'dephasing' and 'time step' in Yambo SHG input ?

the equation of motion
(EOM)isintegrated using the Crack-Nicholson method with a
time step of 0.02 fs [35]. A finite dephasing time of about 6.5
fs is added in the EOM Hamiltonian amounting to filter out
spurious initial nonlinearities in the macroscopic polarization
function caused by the sudden switching-on of the external
f
ield. After five dephasing time constants, all the spurious
components in the polarization can be ignored with respect
to the SHG signal.

regards
sohan

[claudio]

Dear Sohan

in the SHG input the time step appear if you create input with the "nl" verbosity, using the command: yambo_nl -u n -V nl

and then the parameters are:

NLstep= 0.010000 fs # [NL] Time step length
NLDamping= 0.200000 eV # [NL] Damping (or dephasing)

a damping of 0.2 eV is equivalent to 6.5 fs dephasing, a damping of 0.1 eV to 13 fs.

best
Claudio

[sohan]

Dear sir,
Thank you for the response. Sir there are two methods of GW-BSE for linear response one is diagonalization (solving kernel and all) and second is by Real-TIme simulation in which the process is same as of GW-BSE SHG just change of field to 'DELTA' and adding damping in pp. Do these two produce identical results? Can I prefer the RT-BSE over the first one?
regards
sohan

[sohan]

Dear sir,
This SHG spectra is at IPA approximation where 2 valance and 2 conduction bands been considered. Why there's fluctuation at the end of spectra? (damping is set to default)

ip.png

Also should I go for more dense grid as 30x is not converging?

[claudio]

Dear Sohan

try to use a longer simulation time. Plase read in the report the total similation time and try to increase it.
For example set 80 fs and use the last 10 fs for the analysis.
I also I advise you to use YamboPY for the analysis of the response

let us know
best
Claudio

[sohan]

Dear Sir,
Increasing simulation time helped. Sir in your paper abt SHG in GaSe and InSe 2D monolayers you have used shifted grid instead of gamma centered, why and how? Is there any significant difference in spectra when using shifted or gamma centered grid?
regards
sohan

[claudio]

Dear Sohan

in our paper on GeSe and InSe we used shifted grids only for the groud state density, while we used gamma-centered grid for the linear and non-linear response

best
Claudio