Ab initio full cell dynamical mean-field theory (DMFT) and GW+DMFT for solids based on PySCF
Authors: Tianyu Zhu (tianyu.zhu@yale.edu), Huanchen Zhai, Zhihao Cui, Linqing Peng, Garnet Chan
Prerequisites
You need to set environment variable PYTHONPATH
to export fcdmft to Python.
E.g. if fcdmft is installed in /opt
, your PYTHONPATH
should be
export PYTHONPATH=/opt/fcdmft:$PYTHONPATH
Full cell G0W0+DMFT and HF+DMFT (mixed MPI and OpenMP parallelization)
Hamiltonian-based impurity solvers
Molecular and periodic G0W0
Molecular and periodic RPA
CAS-CI treatment of the impurity problem
You can find Python scripts for running DMFT calculations in /fcdmft/examples
.
For example, in /fcdmft/examples/Si
, the steps to run a full cell GW+DMFT
calculation are:
Perform DFT and GW calculations by running si_gw.py
(Note: For large systems, GW should be performed separately using multiple nodes,
i.e. MPI, see /fcdmft/examples/NiO
);
Derive impurity Hamiltonian and GW double counting term by running si_set_ham.py
;
Perform GW+DMFT calculation by running run_dmft.py
(serial or MPI/OpenMP).
All DMFT parameters should be set in run_dmft.py
. In this example, CCSD-GF is used
as impurity solver. See run_dmft
for sample Slurm submission script.
(Optional) One may use a CAS-CI treatment when solving the impurity problem. See
CAS-CCSD/run_dmft.py
for setting CAS-related parameters.
Please cite the following papers in publications utilizing the fcdmft package:
T. Zhu and G. K.-L. Chan, Phys. Rev. X 11, 021006 (2021)
T. Zhu, Z.-H. Cui, and G. K.-L. Chan, J. Chem. Theory Comput. 16, 141-153 (2020)
T. Zhu, C. A. Jimenez-Hoyos, J. McClain, T. C. Berkelbach, and G. K.-L. Chan, Phys. Rev. B 100, 115154 (2019)
Cite the following paper if GW code is used:
Cite the following paper if libdmet package is used: