analyse real-space exciton wavefunction

[sunxl]

Dear all,

When I calculate the projection of excitons in real space, I use the 2D-BN from the tutorial. When I set Direction = "12" in the input file and import the output file o-2D_WR_WC.exc_qpt1_2d_1.xsf into VESTA, nothing is displayed. What could be the reason for this? However, when I use Direction = "123", a normal projection is displayed, but it seems to differ from what is shown in the tutorial. I have attached both types of output files. Which one is correct?

Thank you in advance for your help.

Best,
sunxl

[Daniele Varsano]

Dear Sunxl,

the 2D plot is essentially the same of a 3D one where you have integrated over the 3rd dimension. If it is different from the tutorial, you can check the position of the hole, or it is just a matter of the considered isosurface. About the 2D plot, we will have a look to why vesta is not displaying it.

Best,
Daniele

[sunxl]

Dear Daniele,

Thank you for your patient explanation.

1、For two-dimensional materials, can Direction = "123" be used when projecting exciton wavefunctions?

2、How is the position of the hole determined? Is it decided by calculating the average hole density and observing which atoms have a higher concentration of holes? I read in the tutorial that it's best not to place the hole exactly on an atom but to slightly offset it from the atomic position. For two-dimensional materials, does this mean that the offset should be applied separately in the x and y directions?

I realize I have quite a few questions. Thank you for your patience in answering them.

Best，
sunxl

[Daniele Varsano]

Dear sunxl,

1) Yes you can, in this case you have a 3D distribution, and you can visualize also the z density.
2) The hole position is arbitrary, usually it is suggested to set it in a symmetric point where there is consitent valence density. The suggestion to not place exactly on the atom is because you can have there a zero density of valence band, e.g. in case of pi orbitals, so it is better to displace along the z direction.
3) About the 2D plot. The structure is not displayed as there is a problem in the origin of the plot:

Code: Select all

BEGIN_DATAGRID_2D
     61   61
     0.00000000      0.00000000      8.73459911

If you replace 8.73459911 with 0.00000 it works.

Anyway the isosurface is not correctly read, you can try to normalize the function value activating the flag #NormToOne in the input., i.e. removing the comment. It seems that while xcrsyden works correctly with E- notation e.g. (0.10073722E-10), vesta can have a problem to read it correctly.

Best,

Daniele