Yambo Community Forum

Dear all,

is the RPA response function valid to be used in the case of semi-metals?
If not, is it somehow possible to treat the screening for semi-metallic systems using yambo?

Best,
Franz

Dear Franz,

All the contribution at finite momentum q are correctly taken into account in the RPA calculation. The limit of vanishing q is only partially considered, as the intra-band contribution is missing. You can anyway model it using a Drude model with empirical parameters.

Best,
Daniele

Dear Daniele,

what are the keywords I would need to add to my input file for the static/dynamic screening calculation using the Drude model?
Also, will there be some problems using the "slab z" truncation? As this basically puts the the q=0 (G,G' || z and even) components of eps^{-1} to 1

Best,
Franz


PS:
I am using yambo version 5.2.0.

Dear Franz,

sorry, I did not understand that you are dealing with a 2D system. In this case, it is a bit more complicated.
We are currently working on an implementation for 2D metals, but still we need a comprehensive testing and validation before realising the algorithm.

You can have a look at this recent paper for 3D metals:
https://journals.aps.org/prb/abstract/1 ... 107.155130
and explore if for your system some of the proposed strategies can be applied, e.g. looking at finite, but small q dielectric screening.
Eventually, you can also think about hard-coding the head of the dielectric matrix extracted from the behavuoir at finite q.

Please note that not all the features reported in that paper are released, in particular concerning the multipole sampling (it will be soon available).

If not, currently, I'm afraid the only solution is the brute force convergence with respect k points, and it is not guaranteed that it is computationally affordable.

Best,
Daniele

Dear Daniele,

thanks a lot for the suggestions!

Best,
Franz