cosidering nonlocal effect in dielectric response

[m.abasifard]

Hi all

I'm trying to study plasmonics in the semi metallic graphene nanoribbons considering nonlocal effect. I'm trying to calculate dielectric response (cosidering nonlocal effect) using yambo. I would really appreciate it, if anyone would answer below questions.
1- Is the gw+bse a good strategy for this goal? or should I use rpa or some thing else?
2- how can I include nonlocal effect in my calculation? I found some parameters like BSEQptR and QpntsR. Are they the parameters I should consider? if yes, which one and how?

Best regards,
Mostafa

[Daniele Varsano]

Dear Mostafa,
What do you exactly mean for "non local effects"?
If you mean local field effects, they are contained both in the RPA approximation and in the Bethe-Salpeter equation.
I suggest you to have a look to the two methods in order to understand what calculation is more suitable for your needs: you can have a look to this page for some lectures and reviews.
In order to learn how build up input files and run yambo, please have a look to the tutorials first.
The variable you mentioned are related with the transferred momentum in the response functions.
Best,
Daniele

[m.abasifard]

Dear Daniele

Thanks for your quick reply.
Sorry for no clarification in the previous post.
I want to calculate the epsilon(q,w), i.e the epsilon which is not just a frequency-dependent one and it is'nt local. Indeed in local response, the q tends to zero.
Let me say, the purpose is considering the spatial dispersion.
I would really thank you if you would answer the questions in the my previous post, considering this clarification.

Best,
Mostafa

[Daniele Varsano]

Dear Mostafa,
Im not sure I've totally understood. From your post I think you want to calculated the dielectric function for finite q, i.e. for instance to simulate an electron energy loss spectra (plasmons). If this is the case, you can do calculations with Yambo in RPA approximations (yambo -o c), considering the q of the Brillouin zone you have sampled. The variable to selelct the q point is:

Code: Select all

%QpntsRXd				
1|5|		#	(Xd)	Transferred momenta
%

this calculate the response for the first 5 q points and so on.

At the moment the BSE solution is implemented for q=0 only (optics).

Hope it helps,
Daniele

[m.abasifard]

Dear Daniele

Thank you for your so helpful answers. I'll try to apply what you said.

Best regards,
Mostafa