Why are excitons distributed throughout the supercell ??

[Luo Yan]

Dear all,
Thanks for your attentions. I am learning the plotting of the exciton spatial distribution following the website tutorial. However, no matter where the hole position is, the excitons always distribute throughout the supercell. Thanks for your kind helps for me. Some related inputs are attached in the following. The visualization of exciton spatial distribution from o.exc_qpt1_3d_1 dates was done with VESTA .

[Daniele Varsano]

Dear Luo,
either the exciton is extremely delocalized, or you are using a too-small isosurface value.
Try to raise the isosurface value in the software you are using to plot it (Vesta) and see if it localizes.
Best,
Daniele

[Luo Yan]

Dear Daniele,
Thanks very much. I have learned a lot from you !! Here, in this question, the exciton distribution are the same size in whole supercell and don't depend on the isosurface value in VESTA. So, the exciton may be extremely delocalized in my system. In my immature oppion, this extremely delocalized exciton maybe origins from its small binding benergy of 0.44 eV. Here, one question is that what are the physical means about extremely delocalized exciton ? ?

[Daniele Varsano]

Dear Luo,
an extremely delocalized exciton should have a very small binding energy, 0.4 eV it is not that small!
If you send your xsf file together with the input/report file of your BSE calculation we may have a look to inspect if something has gone wrong.

Best,
Daniele

[Luo Yan]

Dear Daniele,
Thanks for your help. The xsf file is too big. Thus, I sent input and xsf files by E-mail (daniele.varsano@nano.cnr.it). Wish you work smoothly !

[Daniele Varsano]

Dear Luo,
if it is too big for an email, you can put in a dropbox folder or alike.
Best,
Daniele

[Luo Yan]

Dear Daniele,
I sent it to you by email (daniele.varsano@nano.cnr.it) yesterday. Didn't you receive it ?
Best wish !!

[Luo Yan]

Dear Daniele,
Some related inputs files are attached via a dropbox folder https://www.dropbox.com/sh/q8dvlhm98hyv ... jq3za?dl=0 . Thank you for your patience and continued support.
Best

[Daniele Varsano]

Dear Luo Yan,

it seems, indeed, that the exc_wf has a problem. A possible reason is that there is no valence density where the hole is placed.
I suggest you make some tests (also using smaller unit repetitions), changing the hole position in the plane.

In order to make some tests, you can also use the 1D plots, in this case, the other two directions will be integrated and the excwf should not be normalized to 1, in this way
you can compare the absolute value for different hole positions.


Best,
Daniele

[Luo Yan]

Dear Daniele,
Thank you for your prompt reply. Before my calculation, I plot the charge densities of VBM and CBM edge states as follow. As shown in Figure, the valence density is localized where the hole is placed in my system. Are there other possible reasons ?? Actually, I have tested some different hole position. Anyway, I will test the hole position again and feed back my tests. Good Luck !!!