dielectric function don't match literature

[Peace]

Dear all,

I'm a beginner in Yambo and I want to plot the optical properties but i'm facing an issue.

I generated the files bellow:

Code: Select all

 r_optics_dipoles_chi  yambo.in  r_setup  o.eel_q1_ip   SAVE  o.eps_q1_ip 

I know that to plot the imaginary and real part of dielectric function we use o.eps_q1_ip file . The problem is that I found very big values in the y-axis [-20,140 ] as in the attached picture, that don't match the literature [-4,8] . Can you help me figure out where the problem is coming from, please?

[myrta gruning]

Hello,

It is important to make sure that the values that are chosen for the numerical parameters (number of k-points and selected bands) are at convergence, meaning that you have to study the convergence of the spectrum with these parameters and choose the parameters such that the spectrum is not changing (within some tolerance) when further increasing the parameters.

It would be helpful if you attach the report of the calculation. From what I can see, I believe you would need to increase the number of k-points.

Finally, note as well, that the intensity of the spectra depends on the value for the damping (DmRngeXd) though usually the default value should be reasonable and you can leave it as it is for the moment.

[Peace]

Dear Myrta,
For the k-points I used gamma. Isn't that ok?
What is DmRngeXd because I can notice that there is a difference in intensity between my calculations and literature?

[Daniele Varsano]

Dear Salma,
DmRngeXd is the damping range used to calculated the spectrum (Lorentzian peaks). It affects the intensity as the oscillator strength is the are beneath your peaks.
A smaller damping gives you higher intensity and narrower peaks. Providing two values the damping increases linearly between the two values in the range of energies set in EnRnge. If you want a constant damping you need to set the same value for the two entries.

Best,
Daniele

[Peace]

Dear Daniele,

Can you show me how to change those variables through an example please? By the way I'm using Yambo with Quantum ESPRESSO

[Daniele Varsano]

Dear Salma,

For the k-points I used gamma. Isn't that ok?

It depends, if your system is an isolate molecule it is OK, otherwise if it is a bulk or 2D system etc. you need to sample the Brillouin zone.
About the damping:
% DmRngeXd
0.10000 | 0.10000 | eV # [Xd] Damping range
%

In this way, you introduce a damping of 0.1eV along all the energy range of the spectrum. You can reduce or enlarge by changing the two values e.g. (0.05 | 0.05| or 0.2 | 0.2|)

Best,
Daniele