This is the data set for *Detecting electron-phonon coupling during photoinduced phase transition*, [Phys. Rev. B **103** (2021) L121105](https://doi.org/10.1103/PhysRevB.103.L121105).
We perform a four-step protocol to obtain phonon modes of Ta_{2}NiSe_{5} (TNS) that
1. (Optional) Getting optimized atomic position with experimental lattice shape, unless the experimental atomic position is employed
2. Self-consistent field (SCF) solution for the given atomic positions, output the electron density with the primitive cell for band calculations and the wave-function with the supercell for phonon mode calculation
3. (Optional) Band calculation with the primitive cell using the density from the SCF calculation
4. Phonon mode calculation with the supercell using the wave-function from the SCF calculation.
The phonon mode calculation relies on density-functional perturbation theory (DFPT) derived from the density-functional theory[^Hohenberg-Kohn] [^Kohn-Sham] (DFT).
We employ two kind exchange-correlation functionals for DFT, local-density-approximation (LDA) and generalized-gradient-approximation (GGA), specifically Perdew-Wang[^PW-LDA] (PW-LDA) and Perdew-Burke-Ernzerhof[^PBE-GGA] (PBE-GGA) functionals respectively.
We have two types of DFPT calculations according to atomic positions, DFT-optimized and experimental[^Sunshine] ones.
Then, we have four type calculations (PW-LDA, PW-LDA optimal), (PBE-GGA, PBE-GGA optimal), (PW-LDA, Experimental), and (PBE-GGA, Experimental), for (functional for DFT, atomic positions).
We use abinit[^abinit] code throughout all protocols written above, atomic position optimization, SCF, band calculation, and DFPT calculation.
The primitive cell contains 4 Ta, 2 Ni, and 10 Se atoms. The proper spatial symmetry, C2/c (No. 15), is not recognized by abinit with the primitive cell.
We only use the primitive cell to obtain band structure. Double size supercell, a monoclinic Bravais lattice, is properly recognized as the by abinit.
We use the supercell for atomic position optimization and DFPT because expected symmetry is broken when the symmetry is not explicitly enforced. The criteria for the residual force at the optimized atomic position is $5.0 \times 10^{-5} \ \mathrm{a.u.}= 0.0257 \ \mathrm{eV/nm}$.
We choose Gaussian occupation, `occopt = 7`, with a smearing parameter, `tsmear = 1.0e-3`, corresponding to 27.2 meV energy width.
[^Hohenberg-Kohn] P. Hohenberg and W. Kohn, Phys. Rev. 136, B864 (1964).
[^Kohn-Sham] W. Kohn and L. J. Sham, Phys. Rev. 140, A1133 (1965).
[^PW-LDA] J. P. Perdew and Y. Wang, Phys. Rev. B 45, 13244 (1992).
[^PBE-GGA] J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996).
[^abinit] X. Gonze, F. Jollet, A. F. Abreu, D. Adams, B. Amadon, T. Applencourt, C. Audouze, J.-M.Beuken, J. Bieder, A. Bokhanchuk et al., Comput. Phys. Commun. 205, 106 (2016).
[^Sunshine] S. A. Sunshine and J. A. Ibers, Inorg. Chem. 24, 3611 (1985).
**NOTE** for the input files:
We did not modify the input files at all from the original ones we actually used. Some misleading comments are in them. DON'T TAKE THE COMMENTS SERIOUSLY AT ALL but just trust the actual values in them.
**NOTE** for the missing wave-function file `*_WFK`:
We did not store WFK on a desktop computer for analysis but keep on a supercomputer because the file size is so huge around 30 GB. Then, the `*_WFK` files are not included here. To perform DFPT calculation, once `*_WFK` should be created in advance by the SCF calculations.
## Pseudopotentials
We use two kinds of pseudopotentials provided by the web page, [LDA][fhiLDA] and [GGA][fhiGGA], of abinit.
These are generated by fhi98pp[^fhi98pp] code with PW-LDA and PBE-GGA functionals for Ta, Ni, and Se atoms.
The same functional is chosen for DFT computations.
Relevant files are put in the directory `${file_ROOT}/pseudopotentials`. You can also find them in other directories.
[fhiLDA]: https://www.abinit.org/sites/default/files/PrevAtomicData/psp-links/lda_fhi "Pseudopotentials for the ABINIT code” (ABINIT Group)"
[fhiGGA]: https://www.abinit.org/sites/default/files/PrevAtomicData/psp-links/gga_fhi "Pseudopotentials for the ABINIT code” (ABINIT Group)"
[^fhi98pp]: M. Fuchs, M. Scheffler, Comput. Phys. Commun., Comput. Phys. Commun. 119, 67-98 (1999)
## Crystal structure and SCF
Dependent files: just input files, input.in, files.files and the pseudopotentials
We put the files, input files, and corresponding output files, for SCF calculation of the primitive cell in `${file_ROOT}/SCF_primitive`.
One should note that the residual force is over the criteria, $5. 0 \times 10^{-5}\ \mathrm{a.u.}$, because the atomic position optimization is performed under the supercell instead of the primitive cell. There are small errors.
`${file_ROOT}/SCF_super`
### PBE-GGA optimal , super cell, ecut=50, Nk=24 x 6 x 6: **Default**
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_DFT_Geometry_PBE/NK_24_6_6-ecut_50-tsmear_0.001`
### PBE-GGA with the experimental geometry, super cell, ecut=50, Nk=24 x 6 x 6
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_Experimental_Geometry_PBE/NK_24_6_6-ecut_50-tsmear_0.001/`
### PBE-GGA optimal, super cell, ecut=60, Nk=24 x 6 x 6
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_DFT_Geometry_PBE/NK_24_6_6-ecut_60-tsmear_0.001`
### PBE-GGA optimal, super cell, ecut=50, Nk=36 x 9 x 9
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_DFT_Geometry_PBE/NK_36_9_9-ecut_50-tsmear_0.001`
### PW-LDA optimal, super cell, ecut=50, Nk=24 x 6 x 6
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_DFT_Geometry_PW/NK_24_6_6-ecut_50-tsmear_0.001`
### PW-LDA with the experimental geometry, super cell, ecut=50, Nk=24 x 6 x 6
PATH: `${file_ROOT}/SCF_super/monoclinic_SCF_Experimental_Geometry_PW/NK_24_6_6-ecut_50-tsmear_0.001/`
### PBE-GGA optimal, primitiv cell, ecut=50, Nk=24 x 24 x 12
PATH: `${file_ROOT}/SCF_primitive/mono_SCF_DFT_Geometry/NK_24_24_12-ecut_50-tsmear_0.001_PBE`
### PW-LDA optimal, primitive cell, ecut=50, Nk=24 x 24 x 12
PATH: `${file_ROOT}/SCF_primitive/mono_SCF_DFT_Geometry/NK_24_24_12-ecut_50-tsmear_0.001_PW`
### PBE-GGA with the experimental geometry, primitive
PATH: `${file_ROOT}/SCF_primitive/mono_SCF_EXP_Geometry/NK_24_24_12-ecut_50-tsmear_0.001_PBE`
### PW-LDA with the experimental geometry, primitive
PATH: `${file_ROOT}/SCF_primitive/mono_SCF_EXP_Geometry/NK_24_24_12-ecut_50-tsmear_0.001_PW`
## Bandstructure calculation for primitive cell
Dependent files: input files and density `*_DEN`
### PBE-GGA optimal
PATH: `${file_ROOT}/Band_primitive/mono_Band_DFT_Geometry/ecut_50-SCFNK_24_24_12-SCFtmear_0.001_PBE`
### PW-LDA optimal
PATH: `${file_ROOT}/Band_primitive/mono_Band_DFT_Geometry/ecut_50-SCFNK_24_24_12-SCFtmear_0.001_PW`
### PBE-GGA with the experimental geomerty
PATH: `${file_ROOT}/Band_primitive/mono_Band_EXP_Geometry/ecut_50-SCFNK_24_24_12-SCFtmear_0.001_PBE`
### PW-LDA with the experimental geomerty
PATH: `${file_ROOT}/Band_primitive/mono_Band_EXP_Geometry/ecut_50-SCFNK_24_24_12-SCFtmear_0.001_PBE`
### Script for plotting the band dispersion
PATH: `${file_ROOT}/Band_primitive/Scripts`
## DFPT with the supercell
Dependent files: input files and wave-function `*_WFK`
### PBE-GGA optimal
PATH: `${file_ROOT}/DFPT_super/monoclinic_DFPT_DFT_Geometry_PBE/NK_24_6_6-ecut_50-tsmear_0.001`
### PW-LDA optimal
PATH: `${file_ROOT}/DFPT_super/monoclinic_DFPT_DFT_Geometry_PW/NK_24_6_6-ecut_50-tsmear_0.001`
### PBE-GGA with the experimental geomerty
PATH: `${file_ROOT}/DFPT_super/monoclinic_DFPT_Experimental_Geometry_PBE/NK_24_6_6-ecut_50-tsmear_0.001/`
### PW-LDA with the experimental geomerty
PATH: `${file_ROOT}/DFPT_super/monoclinic_DFPT_Experimental_Geometry_PW/NK_24_6_6-ecut_50-tsmear_0.001/`
### Script for the diagonalization of the dyanmical matrix and assignment of irreducible representation for the eigenmode
PATH: `${file_ROOT}/DFPT_super/Scripts`