About dark exciton

[sdwang]

Dear users and developers；
I want to kown how can I distinguish the dark excitons in BSE absorption? I kown ypp can plot exciton wavefuncitons and ypp -e s can sort excitation energy by energy and stenghth, does this results can deduce the dark exciton? If not so , waht should I do?
Another queation: what can I deduce through analyzing excitation with >ypp -e a?
Thanks!

S D Wang

[Daniele Varsano]

Dear Shudong Wang,
as you say, with ypp you can sort excitations by energies or intensities.
A dark exciton is an excitation with zero or very small intensity.
So you can recognize it, analyze it (ypp -a) and plot its wave function.

what can I deduce through analyzing excitation with >ypp -e a?

You can look at the KS contributions to the excitations and their relative weight.


Cheers,

Daniele

[sdwang]

Thank you for your reply so soon!
I have another question about singlet and triplet excitons, and how can I distinguishi them? Can >ypp deal with this?

Thanks!

[Daniele Varsano]

If you want to look at triplet excitations please have a look at this reference:
Rohlfing and Louie, Phys. Rev. B 62, 4927 (2000). According to that you can not
see them simultaneously. This subject has been previously discussed in the forum and
you can have a look to this thread

Cheers,

Daniele

[sdwang]

Thanks!
And, is the position of holein >ypp -e w is random? What should I do when I decide which position is the right one? Is there some rules for deciding the hole position?

SDWang

[Daniele Varsano]

Dear Shudong,
you can decide the hole position by input. Typing ypp -e w you can edit the
input file ypp.in. There you can decide the exciton you want to plot (index),
unit cell repetition and the hole position.
The hole position will be the most near to what you choose in input file compatible
with the fft grid. Where to put the hole, you can choose looking at the symmetries of
your system, or analyzing before the KS state with highest weight, you can put the hole
in a point with enough density of the occupied states that do participate in the excitation.

Hope it helps,

Cheers,

Daniele

[sdwang]

Dear Shudong,
you can decide the hole position by input. Typing ypp -e w you can edit the
input file ypp.in. There you can decide the exciton you want to plot (index),
unit cell repetition and the hole position.
The hole position will be the most near to what you choose in input file compatible
with the fft grid. Where to put the hole, you can choose looking at the symmetries of
your system, or analyzing before the KS state with highest weight, you can put the hole
in a point with enough density of the occupied states that do participate in the excitation.

Hope it helps,

Cheers,

Daniele

I am very sorry, but I still confused about"or analyzing before the KS state with highest weight, you can put the hole
in a point with enough density of the occupied states that do participate in the excitation." When I do >ypp -e s and >ypp -e a, I can get o.exc_E_sorted, o.exc_I_sorted and
o.exc_amplitude_at_index number, o.exc_weights_at_index number,and how can I decide the hole position from these files as you mentioned above ?Please show me more detail.
For example, in the o.exc_weights_at_1 file, it is:
# Electron-Hole pairs that contributes to Excitonic State 1 more than 5%
#
# K-point [iku] Weight
# 0.000000 0.000000 0.000000 0.533268
# 0.08000 0.00000 0.00000 1.00000
# 0.160000 0.000000 0.000000 0.828685
# 0.240000 0.000000 0.000000 0.618436
# 0.320000 0.000000 0.000000 0.425414
# 0.400000 0.000000 0.000000 0.277074
# 0.480000 0.000000 0.000000 0.175084
# 0.560000 0.000000 0.000000 0.110540
# 0.64000 0.00000 0.00000 0.07007
#
# Band_V Band_C K ibz Symm. Weight Energy
#
34.00000 35.00000 1.00000 1.00000 0.09333 0.15424
34.00000 35.00000 2.00000 2.00000 0.08733 0.15551
34.00000 35.00000 2.00000 1.00000 0.08733 0.15551
34.00000 35.00000 3.00000 2.00000 0.07165 0.15938
34.00000 35.00000 3.00000 1.00000 0.07164 0.15938
34.00000 35.00000 4.00000 2.00000 0.05246 0.16549
34.00000 35.00000 4.00000 1.00000 0.05246 0.16549 ,
Thanks!

[Daniele Varsano]

Dear Shudong,
from o.exc_E_sorted (or I_sorted) you can look at the index of your excitation.
Typing ypp -a you will indicate the index you want to analyze.
Next you will have o.exc_weights_at_index. Here you can see the KS state that
participate to the excitation (more than 5% or 10%).
Then you can plot your KS wavefunction (ypp -s w), or from post processing of the code
you used to calculate your ground state (abinit or pwscf). Here you can see were the
charge is, in the valence bands that do participate to your excitations, in general you do not
want to put the hole in a zero density point.
If you have a symmetric molecule, or unit cells, may be you can put the hole in some symmetry
center. In summary, you have freedom on where to put the hole, and it is up to you
to place it where you can get more physical insight.

Hope it helps,

Cheers,

Daniele

[sdwang]

Thank you very much! I kwon it.

[sdwang]

Dear Shudong Wang,
as you say, with ypp you can sort excitations by energies or intensities.
A dark exciton is an excitation with zero or very small intensity.
So you can recognize it, analyze it (ypp -a) and plot its wave function.

what can I deduce through analyzing excitation with >ypp -e a?

You can look at the KS contributions to the excitations and their relative weight.


Cheers,

Daniele

Dear developers:
I sorted the excitations by energies as following；
# E [ev] Strength Index
#
3.715367 0.982025 1.000000
3.961412 1.000000 2.000000
4.209363 0.118684 3.000000
4.261 0.5183E-4 4.000
4.423 0.2095E-2 5.000
4.446 0.2610E-5 6.000
4.639 0.2353E-3 7.000
4.67909 0.06153 8.00000
4.82617 0.01846 9.00000
4.845 0.9682E-7 10.00
...
as can be seen, some intensities are strong and others are very low with 1.0E-7 ... I want to kown, what can I think it as dark exciton, for examples, can I believe the 4th index with strength 0.5183E-4 as a dark one?? Are there some rules to sort the dark excitons?
Thanks!