- Authorization.
The code "Diectric_function_QE.f90" using the formula in eps_man (manual of epsilon) of Quantum Espresso. The code "Diectric_function-Kresolved.f90" is a modified version. "Diectric_function-projection.f90" is based on the formula in paper 2> listed below. Work done for thes codes was performed in the National University of Singapore (NUS). We ask that users of this code should cite the following reference:
So if you use these codes, please cite
1> QE: 10.1088/0953-8984/21/39/395502
2> Our paper: https://doi.org/10.1002/adfm.202005977 (Light–Matter Interaction in Quantum Confined 2D Polar Metals, AFM)
3> Our paper: https://doi.org/10.1002/adma.202104265 (Tunable 2D Group-III Metal Alloys, AM)
Authors: Kanchan Ajit Ulman (ulman.kanchan@nus.edu.sg), He Wen (e0204973@u.nus.edu)
Contact: Quek Su Ying (phyqsy@nus.edu.sg)
Thank you.
- Pre-process
You need to install a patch of QE called "QE6.3_patch_eps_eigocc.patch".
The output: TDMATRIX.dat and Eig_Occ.dat. These are input of this code.
- Why do we need this code?
Because you can easily adjust parameters includeing interband smearing and intraband smearing without running epsilon.x again.
It saves much time because you don't need to compute dipole matrix again and again.
In addition, you can use according code to compute the contriburions from transitions between specific atomic orbitals or k-resolved epsilon.
See https://doi.org/10.1002/adfm.202005977 (Light–Matter Interaction in Quantum Confined 2D Polar Metals, AFM)
- Change the input.dat as you need.
There are three inputs. Please change the name to "input.dat" when running codes.
The first line is the directory you have TDMatrix.dat and Eig_Occ.dat.
Input.dat should be in the same directory as the code.
- Convergence.
You should make sure your calculated epsilon is converged.
1> k-mesh. (I usually using doubled k-point in each direction.)
2> Bandnumber. For imaginary part, it's dependent on what energy you are looking at. You should include all transitions you need when you choose a
bandnumber. Our code provides a way to check convergence also.
- Input data of codes Diectric_function-projection.f90.
You need an additional file for this code. It is computed from projwfc.x of Quantum Espresso: "projwfc_up.dat".
You need to delete the head of this file until "1 1". Please read mt code "Diectric_function-projection.f90" and have a try.