Modify the static dielectric function to run the BSE calculation

[lz dong]

Hi everyone,

I would like to know if we can make some modifications to the static dielectric function (epsilon^-1_GG'(q) ), such as adding or subtracting a value, to alter the static screening. Could we then use the modified static dielectric function to run BSE calculations ? If possible, we will be able to investigate the effect of the static dielectric function on the exciton binding energy.

Thank you in advance for your time and assistance.
Best regards,

[Daniele Varsano]

Dear lz Dong,

In principle, you can do this, but not directly as an input variable.
You can alter the values of the screening when building the BSE kernel, looking at BS_W arrays in src/bse/K_screened_interaction.F
or you can do an external script e.g. in python which read, change and overwrite the screening database ndb.em1s

Best,
Daniele

[lz dong]

Dear lz Dong,

In principle, you can do this, but not directly as an input variable.
You can alter the values of the screening when building the BSE kernel, looking at BS_W arrays in src/bse/K_screened_interaction.F
or you can do an external script e.g. in python which read, change and overwrite the screening database ndb.em1s

Best,
Daniele

Dear Daniele,

Thanks for your reply. I have seen the rim.F file under the /src/coulomb path, and is em1_anis in this file referring to the static dielectric function (epsilon^-1_GG'(q))?

[Daniele Varsano]

Dear lz Dong,

em1_anis refers to the anisotropy and it is not the array you should look at.
As reported above, the screening potential for the BSE is evaluated in src/bse/K_screened_interaction.F

The dielectric matrix is stored in X_mat(ig1,ig2,1)

Best,
Daniele

[Quxiao]

Dear Daniele:

Hello, I have been following the calculation code of Yambo recently. I would like to know which Fortran file is used to calculate the value of X_mat? Or rather, where is the code part corresponding to the" [05] Static Dielectric Matrix "module in the log file of yambo running yambo - X s?

1.After checking the code, I found that the calculation of X_mat seems to be in the relevant file of dielectric_matrix. F in/src/pol_function. I wonder if I understand it correctly?
2.Why BS-W=Xmat * isc% gamp?
(Xmat is a pole, and isc%gamp is a residuals)
I saw your response under this post that the dielectric matrix is stored in Xmat, but the dielectric matrix multiplied by the residual is the shielding matrix, which I don't quite understand here.

Best Wishes！
Quxiao

[lz dong]

Dear lz Dong,

em1_anis refers to the anisotropy and it is not the array you should look at.
As reported above, the screening potential for the BSE is evaluated in src/bse/K_screened_interaction.F

The dielectric matrix is stored in X_mat(ig1,ig2,1)

Best,
Daniele

Dear Daniele,

Thank you very much for your reply，it helped me a lot. Since static dielectric function (epsilon^-1_GG'(q) ) is proportional to the screened Coulomb potential W (J. Phys.: Condens. Matter 31 (2019) 325902, equation (10)), if I want to adjust the epsilon^-1_GG'(q) to be twice of the original value, do I need to multiply all BS_W arrays in src/bse/K_screened_interaction.F by 2? I don't know if I understand correctly.

Best,

[Daniele Varsano]

Dear Quixao,

1) Correct
2) BS-W=Xmat * isc% gamp is the screened potential W=eps*V_coulomb

Best,
Daniele

[Daniele Varsano]

Dear lz Dong,

Yes you can do as you wrote, then I do not know if you aim to rescale only the diagonal part or the entire matrix, this depend on the model you have in mind.

Best,
Daniele