- Have CASTEP compiled and
castep.mpi
in the$PATH
base script.sh
- Run this notebook or
python post-process.py
Adjust the HUBBARD_ALPHA
value and record the initial and final occupations of the Mn d levels when continuing from the unperturbed run (base_calc
).
The value of U is extracted using the gradient of the two response curves.
The example in http://hjkgrp.mit.edu/content/calculating-hubbard-u is reused here and the same methodology is followed.
The more self-consistent
U can be computed by repeating the procedure with different input value for the U.
Reference:
Cococcioni, M.; de Gironcoli, S. Linear Response Approach to the Calculation of the Effective Interaction Parameters in the LDA+U Method. Phys. Rev. B 2005, 71 (3), 035105. https://doi.org/10.1103/PhysRevB.71.035105.
DISCLAIMER: Use at your own risk. Please raise an issue if you think the method here is not correct due to implementation details.