dftBephy

Calculating electron-phonon couplings (EPCs) with DFTB.

For more information see our article:

Croy, A., Unsal, E., Biele, R., Pecchia A. DFTBephy: A DFTB-based approach for electron–phonon coupling calculations. J Comput Electron (2023). doi:10.1007/s10825-023-02033-9 full-text access

Prerequisites

• numpy and scipy

• dftb+

• phonopy (conda install -c conda-forge phonopy)

• cython (pip install Cython) for faster routines.

• Other packages might be necessary: mpi4py, spglib, h5py, hsd

Installation

• Pull / download latest version from github.
• Run python setup.py build_ext --inplace in the terminal to build faster routines.
• Run pip install -e . in the terminal to install package but keep it editable in the current directory.

Running calculations

• Starting point for all dftBephy calculations is a finished phonopy calculation of the force constants (e.g. FORCE_SETS and phonopy_disp.yaml).
• The working directory should contain a dftb_in.hsd file, which reads the geometry from geo.gen (will be written by dftBephy) and contains the option WriteHS = Yes (to be used by dftBephy). The directory may also contain charges.bin from a previous SCC run.
• See the examples/ directory for more details. (It's recommended to copy one of the examples and adapt it to your needs.)
• Detailed information about DFTBephy input (dftbephy_in.hsd) can be found here and in the wiki.

What you can get

The main purpose of dftBephy is the calculation of electron-phonon couplings. Apart from that, the package also allows the calculation of the electronic band-structure and the electron relaxation-time (at the moment only within SERTA). The latter can be used as an input for BoltzTrap2 to calculate transport properties. The scripts/ directory contains some programs and templates for computing

• EPCs and relaxation times along a band path (q and k paths, respectively) -- graphene-ephline.py and graphene-lws.py -- the results are stored in a json file;
• EPCs at k-point on a (fine) q-mesh -- dftbephy-epc.py -- the result is stored in a hdf5 file;
• Relaxation times on a (fine) k-mesh -- dftbephy-relaxationtimes-mpi.py -- the results are stored in a hdf5 file;
• Relaxation times as input for Boltztrap2 -- graphene-mobility-bt2.py -- the results are stored in a hdf5 file. For this script, ase and spglib are required;
• Conductivity tensor -- dftbephy-mobility-mpi.py -- the results are stored in a json file. For this script, spglib is required.

See jupyter notebooks in notebooks/ for how to read, use and visualize the output.