The proper field intensity for THG

Questions and doubts about features of non linear optic in Yambo (yamb_nl)

Moderators: Davide Sangalli, claudio, myrta gruning

Post Reply
Zhengran Wu
Posts: 11
Joined: Tue Jul 27, 2021 9:06 am

The proper field intensity for THG

Post by Zhengran Wu » Tue Nov 30, 2021 2:39 pm

Dear all,

I am calculating THG for Si, GaAs and ZnSe. I am confused with the E field intensity setup.


For Si, the THG spectrum converged with intensity 1E6 ~ 1E8 kW/cm2. When I increase the k mesh, the Im(X3) increases and finally converged. Everything went well.

For GaAs, the THG spectrum can only converge with a narrow intensity range. For a 8*8*8 k mesh, only results with intensity 0.5*E5~E5 kW/cm2 converged. I set the intensity to E5 kW/cm2 and increased the k mesh. Strange things happened. Im(X3) converged for k 8*8*8 and 20*20*20, but not for 16*16*16 and 24*24*24.

For ZnSe, I found no converged results with intensities E1~ E11 kW/cm2.


Here are my questions:
1. Why is the X3 coefficient intensity dependent? Physically, this coefficient should be intensity independent. To my understanding, we can not set the intensity too small because there is numerical noise. If we set it too high, the crystal is damaged. Do I understand correctly?
2. How wide an intensity interval for convergence is reasonable. The convergence interval for GaAs as I calculated is only 0.5*E5 kW/cm2. I do not know if it is a reasonable result.
3. Dose convergence vary at different k mesh? Do I need to do a convergence test at each k mesh?
4. Why is there no proper field intensity for ZnSe?


Result attachment: https://mailsucaseducn-my.sharepoint.co ... w?e=lAqwec
Zhengran Wu
PhD student
Institute of Physics, Chinese Academy of Science

User avatar
claudio
Posts: 448
Joined: Tue Mar 31, 2009 11:33 pm
Contact:

Re: The proper field intensity for THG

Post by claudio » Tue Nov 30, 2021 2:53 pm

Dear Zhengran Wu

may you do some tests using the CRANKNIC integrator instead of INVINT?
this should improve results.

Regarding the correct intensity, I tested it only on Silicon and carbon nanotubes.

For ZnSe probably you should increase the k-points to at least 8x8x8, then try to use longer simulation
and check if this it help

NLtime=75 fs

best
Claudio
Claudio Attaccalite
[CNRS/ Aix-Marseille Université/ CINaM laborarory / TSN department
Campus de Luminy – Case 913
13288 MARSEILLE Cedex 09
web site: http://www.attaccalite.com

Zhengran Wu
Posts: 11
Joined: Tue Jul 27, 2021 9:06 am

Re: The proper field intensity for THG

Post by Zhengran Wu » Sat Dec 18, 2021 3:39 pm

Dear Claudio,

Thanks for your advices. I have used CRANKNIC integrator and increase the simulation time. Now I can find the convergence range of the field intensity :D

But during the test, for the calculation of ZnSe, I found other two problems:
1. the third oder results is converged from 1E^3 to 1E^5 kW/cm, under a 10*10*10 k mesh. But it is not converged under k mesh 26*26*26, when i set the intensity to 2E^4 and 4E^4 kW/cm. Converged results can be found for 0.25E^4 and 0.5E^4 kW/cm. Does a more dense k mesh needs a lower field intensity?
2. I tried to calculate the two photon absorption of ZnSe in the way you instructed before. The o.YPP-X_probe_order_3 is converged and almost the same under three different field intensity. Then I used the script similar to the one in your github (with a little changes) to extract the TPA from _order_1 files. But I found no convergence result. I have no idea what is wrong with it.

results:
https://mailsucaseducn-my.sharepoint.co ... Q?e=zXDWr9
TPA script:
https://mailsucaseducn-my.sharepoint.co ... g?e=QPnOkK

Bests
Zhengran
Zhengran Wu
PhD student
Institute of Physics, Chinese Academy of Science

User avatar
myrta gruning
Posts: 239
Joined: Tue Mar 17, 2009 11:38 am
Contact:

Re: The proper field intensity for THG

Post by myrta gruning » Mon Dec 20, 2021 3:57 pm

Hello

I think the problem is the propagation time.

To get the third harmonic, which signal is tiny, you should propagate about 100-120 fs. The default (which I believe is about 60fs) is OK for the second harmonic but not for the higher harmonics.

The fact that you get a X3 depending on the field strength make me think that there are components of the polarizabilities at 3\omega that are not proportional to the cube of the applied field. Usually this is a linear component from an eigenmode at about 3\omega. The eigenmodes are excited inevitably when switching on the field. They are not visible to the eye, but still larger or comparable in intensity to higher harmonics. This would also explain the system dependence, a system may or may not have strong eigenmode at 3 omega.
Dr Myrta Grüning
School of Mathematics and Physics
Queen's University Belfast - Northern Ireland

http://www.researcherid.com/rid/B-1515-2009

User avatar
claudio
Posts: 448
Joined: Tue Mar 31, 2009 11:33 pm
Contact:

Re: The proper field intensity for THG

Post by claudio » Tue Dec 21, 2021 9:44 am

Dear Zhengran

let me add a couple of comments on this post.

1 ) You are not the only one that reported this kind of problems with the X3,
other people found a similar problem. Honestly we tested the extraction of X3 only in two systems,
Silicon and Carbon Nanotubes, we expected the procedure should work also for other systems, but probably
there are more difficulties to make it converges.

2) Another thing you can try it is to reduce the integration time-step size, and see if simulations become more stable

best
Claudio
Claudio Attaccalite
[CNRS/ Aix-Marseille Université/ CINaM laborarory / TSN department
Campus de Luminy – Case 913
13288 MARSEILLE Cedex 09
web site: http://www.attaccalite.com

Post Reply