How to fix the hole position?

[Dean]

Dear all,
I want to see the spatial character of the exciton using YPP. However, when I fix the hole at different positions, I got different spatial characters of the exciton.
So, how to understand these results and is there a principle to fix the hole position?
Thank a lot.

[Daniele Varsano]

Dear Yimin,

However, when I fix the hole at different positions, I got different spatial characters of the exciton.

Yes, this is expected as the excitonic wavefunction is a 2 variable function.
Reasonable criteria to fix the hole position is to choose a point possibly in a symmetric point of the system where you have significant electron density in the orbitals participating in the excitation (if possible). Which are the orbitals (or Bloch states) contributing to an exciton can be inspected looking at the eigenvectors of the BSE by using ypp (ypp -a).
Best,
Daniele

[Dean]

Dear Daniele,
Thanks a lot.

[sdwang]

Dear Daniele,
I found the hoke position in the plot are not exactly designated in the input. And My input of ypp is
excitons # [R] Excitons
wavefunction # [R] Wavefunction
Format= "x" # Output format [(c)ube/(g)nuplot/(x)crysden]
Direction= "123" # [rlu] [1/2/3] for 1d or [12/13/23] for 2d [123] for 3D
FFTGvecs= 8009 RL # [FFT] Plane-waves
States= "161 - 161" # Index of the BS state(s)
Degen_Step=0.1000E-4 eV # Maximum energy separation of two degenerate states
% Cells
17 | 17 | 1 | # Number of cell repetitions in each direction (odd or 1)
%
% Hole
0.00000 | 1.87000 | 16.00000 | # [cc] Hole position in unit cell
%
The hole position using Cartesian coordinates, and in the plot I found the z direction coordinate is far less than 16. What problem could be?
Thanks!

Shudong

[Daniele Varsano]

Dear Shudong,
please post your report and xsf file as I can have a look.

Daniele

[sdwang]

Dear Daniele,
Attached is the input, report and xsf format plot where you can find the hole is not in the upper layer around Mo atom. In the input file, I fixed the
hole around Mo atom in the upper layer, but it is far away it in the plot. (I reduced the cell numberc and neglect the charge distribution to lower the file zise!)
By the way，'Degen_Step' would induce a different excition distribution with various values, and how could one choose it reasonably?
Thanks!

Shudong

[Daniele Varsano]

Dear Shudong,
1) attached there is a cif file and not an xsf file, so I cannot say anything on what is going on.
Looking at the report file the z coordinate has been changed from 9.4 au in the input to 9.7 in the report, this is because this is the closest point in the real space grid given by the FFT. In order to have a closer point you need a finer grid, so you need to raise the FFTGvecs variable.
What yambo should do is just to convert the position in input file given in atomic units, to Angstrom in the xsf file.

2) Degen_Step group excitons that are close in energy up to a threshold given by this variable. If you want to plot excitonic wfs of a given only excitation just set it to 0. In the case of degeneracy anyway, it is needed to make a linear combination of them in order to have the symmetry of the system respected, so you need to choose this threshold accordingly to what you want to plot.

Best,
Daniele

[sdwang]

Dear Daniele,
I have increased the FFTGvecs variable to its highest value (80 Ry), the the problem is still there. Could I used another coordinate system in the input file for hole position?

Thanks!

SD

[Daniele Varsano]

Dear Shudong,
input file needs a hole position in atomic units.
Actually, I've not clearly understood what is your problem, if there is something that went wrong or if it is just a matter of grid discretization as you did not post the xsf output. Have you tried to look at the gnuplot format (in that case you do not have any conversion and the real space grid is reported in the outpu file.).

Daniele