[salmon-user:00376] ovlp output
Elena Kachan
Dear all,
I have several questions.
What is exactly written in the file _ovlp.data? Are these the occupations in the end of the tddft_pulse simulation?
How is the number of excited electrons calculated (the data in _nex.data)? I am doing a TDDFT simulation in a metal and I'm wondering to what the number of excited electrons corresponds in this case.
Thank you,
Elena
Télécom Saint-Étienne
18 rue Pr Benoît Lauras
42000 Saint-Étienne
France
Mitsuharu UEMOTO
Thank you for reaching out.
The ovlp file represents the projection of the time-dependent wavefunction onto the ground-state wavefunction at each time step. We believe this is effectively equivalent to the occupation probability of the ground-state band.
For example, if we denote the time-dependent wavefunction at time t, Bloch wavevector k, and band b as |u_bk(t)>, and the corresponding ground-state wavefunction is |u_{b,k+A/c}(0)>.
Thus, ovlp is defined as: | <u_b,k(t)|u_{b,k+A/c}(0)> |^2
The _ovlp file contains the following information at every time intervals,
* Output time step
* k-point index, Occupation probability of each band at the same k-point
Furthermore, _nex represents the excited carrier density, which is obtained by integrating the deviation of ovlp from its initial value over all bands and the entire Brillouin zone.
Please note that this definition of carrier density is specific to band semiconductors.
For metals, it is generally difficult to define the carrier density in the same manner
(although a careful integration of ovlp in k-space might still provide an estimate).
Sincerely,
----
Mitsuharu UEMOTO (Assist.Prof. / Ph.D.)
E-mail: uem...@eedept.kobe-u.ac.jp Phone: 078-803-6497
Department of Electrical and Electronic Engineering,
Graduate School of Engineering, Kobe University
________________________________________
差出人: Elena Kachan <elena....@univ-st-etienne.fr>
送信日時: 2025年9月24日 22:28
宛先: salmo...@salmon-tddft.jp
件名: [salmon-user:00376] ovlp output
Dear all,
I have several questions.
What is exactly written in the file _ovlp.data? Are these the occupations in the end of the tddft_pulse simulation?
How is the number of excited electrons calculated (the data in _nex.data)? I am doing a TDDFT simulation in a metal and I'm wondering to what the number of excited electrons corresponds in this case.
Thank you,
Elena
--
Elena Kachan
Télécom Saint-Étienne
18 rue Pr Benoît Lauras
42000 Saint-Étienne
France
+33 4 77 91 58 46
Elena Kachan
Thank you very much for your fast reply. For the excited electron density in _nex, do you integrate only over the initially occupied bands? Otherwise if you integrate the deviation over all bands (occupied and empty) it seems that it should sum up to zero.
Elena
Mitsuharu UEMOTO
My previous email lacked sufficient detail and may have caused some misunderstanding.
You were absolutely correct in your observation.
The integration is performed only over the valence bands (or only over the conduction bands) for the hole (or electron density). Because of this formalism, _nex is defined only for band semiconductors.
In the case of metals, there are partially occupied bands below the Fermi level, which makes such an numerical integration difficult to carry out.
Sincerely,
Mitz.
送信日時: 2025年9月24日 23:55
宛先: salmo...@salmon-tddft.jp; Mitsuharu UEMOTO
件名: Re: [salmon-user:00377] Re: ovlp output
Thank you very much for your fast reply. For the excited electron density in _nex, do you integrate only over the initially occupied bands? Otherwise if you integrate the deviation over all bands (occupied and empty) it seems that it should sum up to zero.
Elena
Le 24/09/2025 à 16:18, Mitsuharu UEMOTO a écrit :
Dear Elena,
Thank you for reaching out.
The ovlp file represents the projection of the time-dependent wavefunction onto the ground-state wavefunction at each time step. We believe this is effectively equivalent to the occupation probability of the ground-state band.
For example, if we denote the time-dependent wavefunction at time t, Bloch wavevector k, and band b as |u_bk(t)>, and the corresponding ground-state wavefunction is |u_{b,k+A/c}(0)>.
Thus, ovlp is defined as: | <u_b,k(t)|u_{b,k+A/c}(0)> |^2
The _ovlp file contains the following information at every time intervals,
* Output time step
* k-point index, Occupation probability of each band at the same k-point
Furthermore, _nex represents the excited carrier density, which is obtained by integrating the deviation of ovlp from its initial value over all bands and the entire Brillouin zone.
Please note that this definition of carrier density is specific to band semiconductors.
For metals, it is generally difficult to define the carrier density in the same manner
(although a careful integration of ovlp in k-space might still provide an estimate).
Sincerely,
----
Mitsuharu UEMOTO (Assist.Prof. / Ph.D.)
Graduate School of Engineering, Kobe University
________________________________________
送信日時: 2025年9月24日 22:28
宛先: salmo...@salmon-tddft.jp<mailto:salmo...@salmon-tddft.jp>
Dear all,
I have several questions.
What is exactly written in the file _ovlp.data? Are these the occupations in the end of the tddft_pulse simulation?
How is the number of excited electrons calculated (the data in _nex.data)? I am doing a TDDFT simulation in a metal and I'm wondering to what the number of excited electrons corresponds in this case.
Thank you,
Elena
--
Elena Kachan
Télécom Saint-Étienne
18 rue Pr Benoît Lauras
42000 Saint-Étienne
France
+33 4 77 91 58 46
--
Elena Kachan
[cid:part1.wRYQR...@univ-st-etienne.fr]<https://www.univ-st-etienne.fr>
Elena Kachan
Dear Mitsuharu,
Thank you for the details. So, I performed the tddft_pulse calculation of Ni and ovlp.data shows unexpected values. Here is the first line
# Projection
# ik: k-point index
# ovlp_occup: Occupation
# NB: Number of bands
# 1:ik[none] 2:ovlp_occup(1:NB)[none]
1
1 0.891582317583915E+002 0.890919036230466E+002
0.890919036281576E+002 0.890919036766141E+002
0.889090560931560E+002 0.889090556605696E+002
0.889090556941627E+002 0.889154354023737E+002
0.889154354634873E+002 0.889137383207967E+002
0.889137392847177E+002 0.889137385957607E+002
0.898213031659518E+002 0.898213025449411E+002
0.898213027203811E+002 0.907053132992068E+002
0.278341918505815E+002 0.278341906161199E+002
0.205389897237870E+002 0.205389882430709E+002
0.205389858992120E+002 0.857951865074214E+000
0.857951928235807E+000 0.857951969145173E+000
0.114173688092622E+001 0.114173692277768E+001
0.114173684230435E+001 0.114875404994021E+001
0.418741449200821E-001 0.408292453122853E-001
0.427236531020625E-001 0.208165212541855E-001
0.171567808729755E-001 0.172013535886113E-001
0.173806031902087E-001 0.217868614992323E-001
0.225036660524854E-001 0.220380914065979E-001
0.376906486809465E-002 0.658726607597087E-002
The maximum occupation here is 89. However, the ground state calculation at 0 K shows the correct maximum occupation of 2. So, I suppose that something is wrong in the calculation of ovlp for Ni. I did the same calculation for Si and the ovlp data are correct and evolve reasonably from the ground state Fermi-Dirac occupation.
I also have a feeling that something is wrong with dnsdiff_*.cube data since it does not evolve much during the calculation and seems quite high from the very beginning of the laser pulse. I attach the input file and cube data results.
Do you know what can be the issue?
Thank you,
Elena
Yamada Shunsuke
送信日時: 2025年9月30日 20:44
宛先: uemoto <uem...@eedept.kobe-u.ac.jp>; salmo...@salmon-tddft.jp <salmo...@salmon-tddft.jp>
件名: [salmon-user:00380] Re: ovlp output
Elena Kachan
Dear Shunsuke,
Thank you very much for the files. I also realized that I made a mistake in the units, I used the different ones in the ground state and TDDFT calculations. So now everything works fine.
Best,
Elena