Problems of A and B excitonic peak in 2D MoS2 absorption spectrum

[pandachang97]

Hello Daniele,
I hope you are doing well. Recently, I am running BSE calculation of absorption spectrum of 2D MoS2. The DFT cal is done in the QE (lattice vector a=3.168 A, c=20 A, ecut=60 Ry, nbands=500, k-grid=27*27*1). Then, I have G0W0 and G1W1 calcultions (see the gw_try1.in in the attachment). Later, I followed the tutorial listed in the webcite of Yambo tutorial and three input files are attached. My problem is that I can not reproduce the A or B peak in the absorption spectrum at around 2.0 eV(I know I did not include spin-orbit coupling, there should be split of peaks at 2.0 eV, but there should be a single peak at around 2.0 eV. ). I can reproduce the absorption spectrum at around 3.0 eV. The convergence test of paprameters is the same as the nature paper mentioned. The absortion spectrum is below. Please feel free to let me know the potential problems of my input files (e.g. the electric field direction or the GW calculations). Thank you so much and best regards,
Xin

[lyzhao]

Dear Chang,
What is the number of electron?
% BSEBands
10 | 20 | # [BSK] Bands range
%
means there is at least 20 valence electrons in your 2D MoS2 system

Best regards.

[Daniele Varsano]

Dear Xin Chang,

We should look at the report file to spot if something went wrong.

In general, check your DFT band structure and GW calculations band gap. ie if a reasonable gap is obtained. Next, check your convergence parameters. When choosing the q direction, you can set the z direction, ie. the direction perpendicular to the 2D system to zero.

here you can find a tutorial (not prepared by us):
https://www.paradim.org/sites/default/f ... _YAMBO.pdf

Anyway, take it with caution as for instance having em1s and em1d in the same run is redundant.


Best,

Daniele