Data from: Anomalous thermal transport in Eshelby twisted van der Waals nanowires
Data files
Dec 30, 2024 version files 277.27 MB
-
GeS_lattice_thermal_conductivity_theory_data_repository.zip
277.26 MB
-
README.md
7.78 KB
Abstract
Dislocations in van der Waals (vdW) layered nanomaterials induce strain and structural changes that substantially impact thermal transport. Understanding these effects could enable the manipulation of dislocations for improved thermoelectric and optoelectronic applications, but experimental insights remain limited. In this study, we use synthetic Eshelby twisted vdW GeS nanowires (NWs) with single screw dislocations as a model system to explore the interplay between dislocation-induced structural modifications and lattice thermal conductivity. Our measurements reveal a monoclinic structure stabilized by the dislocation, leading to a substantial drop in thermal conductivity for larger diameter NWs (70% at room temperature), supported by first-principles calculations. Interestingly, we also find an anomalous enhancement of thermal conductivity with decreasing diameter in twisted NWs, contrary to typical trends in non-twisted GeS NWs. This is attributed to increased conductivity near the NW cores due to compressive strain around the central dislocations and aligns with a density functional theory-informed core-shell model. Our results highlight the critical role of dislocations in thermal conduction, providing fundamental insights for defect and strain engineering in advanced thermal applications.
README: Anomalous Thermal Transport in Eshelby Twisted Van der Waals Nanowires: Open Source Data
The raw data used in lattice thermal conductivity calculations is made publicly available. The dataset includes crystal structure files (POSCAR), dielectric tensor and Born effective charges (BORN), second-order interatomic force constants (fc2.hdf5), and third-order interatomic force constants (fc3.hdf5) for GeS in both orthorhombic and monoclinic structures. The files are organized into the following directories:
- GeS_orthorhombic: Contains the structure (POSCAR), Born effective charges (BORN), and second- and third-order interatomic force constants (fc2.hdf5 and fc3.hdf5) for GeS in the orthorhombic structure.
- GeS_monoclinic: Contains the structure (POSCAR), Born effective charges (BORN), and second- and third-order interatomic force constants (fc2.hdf5 and fc3.hdf5) for GeS in the monoclinic structure.
- GeS_monoclinic_normal_compression: Contains the structure (POSCAR), Born effective charges (BORN), and second- and third-order interatomic force constants (fc2.hdf5 and fc3.hdf5) for GeS in the monoclinic structure at 0%, 2%, 4%, and 6% normal compression.
Descriptions
POSCAR File Description
The POSCAR
file is a input for Vienna Ab initio Simulation Package (VASP), which is used for density functional theory (DFT) simulations. It defines the atomic structure of a material, including lattice parameters, atomic positions, and other essential data required for computational modeling. The POSCAR file is formatted in plain text. This file is used to parse the stucture information to the phonopy/phono3py code.
Structure and Content of the POSCAR File
The file is organized into specific sections: a descriptive comment line, a scaling factor, lattice vectors defining the unit cell, atomic species and their counts, and the atomic coordinates in either Cartesian or fractional form.
For more information about the structure of POSCAR file please check: https://phonopy.github.io/phonopy/input-files.html#vasp-poscar-like-format
BORN File Description
The BORN
file is used by Phonopy/Phono3py tools for phonon and lattice thermal conductivity calculations. This file provides the dielectric tensor and the Born effective charges, which are essential for incorporating non-analytical term corrections (NAC) in phonon dispersion calculations and lattice thermal conductivity.
Structure and Content of the BORN File
The BORN
file is a plain text file with a specific format, organized as follows:
- The first line contains a header with the atom indices for which the Born effective charges are provided.
- In the second line, dielectric constant is specifed in Cartesian coordinates. The nine values correspond to the tensor elements of xx, xy, xz, yx, yy, yz, zx, zy, and zz.
- From the third line, Born effective charges for the independent atoms in the primitive cell have to be written in Cartesian coordinates. The nine values correspond to the tensor elements of xx, xy, xz, yx, yy, yz, zx, zy, and zz. For the orthorhombic cell, there are two non-equivalent atomic sites (site 1 and site 5). For the monoclinic cell, there are four non-equivalent atomic sites (site 1, site 3, site 5, and site 7). Therefore, the BORN file for the orthorhombic cell contains 4 lines, and the BORN file for the monoclinic cell contains 6 lines.
For more details about the structure of born file please check: https://phonopy.github.io/phonopy/input-files.html#born-file
fc2.hdf5 File Description
The fc2.hdf5
file is used by Phonopy/Phono3py to store the second-order force constants in a structured format. These force constants describe the harmonic interactions between atoms in a crystal and are used for phonon calculations, such as phonon dispersion relations, density of states, and lattice thermal conductivity.
Structure and Content of fc2.hdf5\
The fc2.hdf5
file is written in the HDF5 (Hierarchical Data Format version 5), a versatile format for storing and organizing large amounts of numerical data. Second-order force constants are stored in units of eV/Ų. The fc2.hdf5
file contains three objects: ['force_constants', 'p2s_map', 'physical_unit']
.
- The
force_constants
object is a NumPy array with dtype='double' and order='C', with the shape: (num_patom, num_satom, 3, 3).num_patom
refers to the number of atoms in the POSCAR file (here,num_patom = 8
).num_satom
refers to the number of atoms in the 4x4x2 supercell of the POSCAR file, so the shape of the array is (8, 256, 3, 3).- Each atom pair has a 3x3 matrix of force constants corresponding to the Cartesian directions.
- The
p2s_map
object provides the indices of atoms in the primitive cell relative to the supercell index system. - The
physical_unit
object specifies the physical unit of the interatomic force constants, which is eV/Ų.
For more details about the structure of fc2.hdf5
file please check: https://phonopy.github.io/phonopy/input-files.html#format-of-force-constants-hdf5
fc3.hdf5 File Description
The fc3.hdf5
file is used by Phono3py to store the third-order force constants in a structured format. These third-order force constants describe the anharmonic interactions between atoms in a crystal lattice, which are critical for understanding thermal transport and phonon lifetimes.
Structure and Content of fc3.hdf5
The fc3.hdf5
file is written in the HDF5 format, with third-order force constants stored in units of eV/ų. The fc3.hdf5
file contains three objects: ['fc3', 'p2s_map', 'version']
.
- The
fc3
object is a NumPy array with dtype='double' and order='C', with the shape: (num_satom, num_satom, num_satom, 3, 3, 3).- The first three
num_satom
values correspond to the atom indices in the 3x4x1 supercell based on the POSCAR file. - Therefore, the shape of the array is (96, 96, 96, 3, 3, 3).
- For each atom triplet, there is a 3x3x3 matrix of force constants corresponding to the Cartesian directions.
- The first three
- The
p2s_map
object provides the indices of atoms in the primitive cell relative to the supercell index system. - The
version
object specifies the version of the Phono3py software used (2.6.0 in our case).
For more details about the structure of fc3.hdf5
file please check: https://phonopy.github.io/phono3py/input-output-files.html#fc3-hdf5
Key Information Sources
ll datasets presented here were generated using a combination of Phonopy and Phono3py, alongside density functional theory (DFT) as implemented in the VASP code. Detailed information regarding the DFT calculations can be found in the manuscript.
Code/Software
To examine the HDF5 files, we recommend using the h5py
package. The provided files require the use of the Phonopy and Phono3py packages. Both Phonopy (https://phonopy.github.io/phonopy/) and Phono3py (https://phonopy.github.io/phono3py/) are open-source Python tools developed by Dr. Atsushi Togo at Kyoto University, Japan. Installation instructions for these packages can be found on their respective websites. For further details on the methods implemented, please refer to the following pages: