EPW Forum

Dear EPW developers,

For the computations of two-dimensional materials, F. Mauri's group proposed the Coulomb cutoff technique [Physical Review B, 96(7), 75448 (2017)],
which has been implemented in QE since version 6.2.1.

Does EPW support the Coulomb cutoff technique for two-dimensional materials calculations?

Best wishes,
Miao Gao
Ningbo University

Dear Miao Gao:

Currently, EPW doesn't support the Coulomb cutoff technique for two-dimensional materials calculations.

Sincerely,

H. Lee

Dear H. Lee,

Thank you very much for your quick reply! Btw, does EPW support the vdW corrections, such as Grimme-D2 (also called DFT-D)? In the phonon calculation, QE supports vdW corrections. In my opinion, EPW actualizes the general Fourier transformation based on the first-principles datum. Thus, if QE supports vdW corrections, so does EPW. Am I right? Similarly, EPW should also support the Coulomb cutoff technique. Because QE already generates the bloch sates, dynamic matrix, and the phonon perturbation potential on the coarse grid with the Coulomb cutoff technique.
Subsequently, EPW just reads these datum and carries out interpolation. Look forward to your reply!

Best wishes,
Miao Gao

Dear Miao Gao:

First of all, I would like to emphasize that I don't know much about both vdW and 2D treatment in EPW.

EPW imports from the phonon code the dynamical matrices (prefix.dyn*), the perturbed Kohn-Sham potentials (prefix.dvscf*), and the pattern files (patterns.*).

Regarding the 2D Coulomb cutoff in EPW:
I can point out the missing correction to the local part of the ionic potential [See "2. Perturbed KS potential." of "E. Phonons and EPC." of "IV. IMPLEMENTATION" in the paper at https://doi.org/10.1103/PhysRevB.96.075448].
Unlike the Hartree and xc potentials included in prefix.dvscf*, the contribution from the ionic potential to the perturbed KS potential is calculated in EPW; it is not imported [See the relevant subroutines in dvqpsi.f90]. I am not sure about the other terms and whether or not I am missing the additional issues.

Regarding the vdW in EPW:
For the nonlocal vdW, looking at the subroutine of dnonloccorr in dnonloccorr.f90 and that of dv_of_drho in dv_of_drho.f90 [See also the paper at https://journals.aps.org/prb/abstract/1 ... .93.235120], it seems to me that the nonlocal correlation term is included in dvscf. But I am not sure whether or not I am missing something.
For the semi-empirical vdW, I think that we can't define the corresponding potential which should be included in dvscf. Looking at the subroutine of dynmat0_new in dynmat0.f90 and that of d2ionq_disp in d2ionq_disp.f90, the correction is added to the dynamical matrices only. But, like the nonlocal vdW, I am not sure whether or not I am missing something.

Sincerely,

H. Lee

Dear H. Lee,

Thanks for the detailed reply! This helps me a lot!

Best wishes,
Miao Gao