EPW: a program for calculating the electron-phonon coupling using maximally localized Wannier functions (Q546804)

The paper presents a computer code (electron-phonon coupling using Wannier functions, EPW) for calculating the electron-phonon coupling from the first principle using the density functional perturbation theory and maximally localized Wannier functions. It is shown that EPW enables extremally accurate and highly efficient calculations of the mode-resolved and total electron-phonon coupling strength as well as phonon and electron line-widths. EPW exploits the real-space localization of Wannier functions to generate large numbers of first-principles electron-phonon matrix elements through a generalized Fourier interpolation. This communication outlines the functionality of EPW and the details of the technical release. In particular, the quantities which can be computed using EPW are stated and include: (i) the total electron-phonon coupling strength, (ii) the phonon self-energy associated with the electron-phonon interaction within the Migdal approximation, (iii) the electron self-energy associated with the electron-phonon interaction within the Migdal approximation, (iv) the Eliashberg electron-phonon spectral function, (v) the transport electron-phonon spectral function. EPW is a program written in Fortran90, it uses information provided by Quantum ESPRESSO and Wannier90, and runs as a post-processing tool. The paper reviews the individual sections of the code and describes its parallel structure. In particular, the parallelization is achieved through the MPI library specification for message passing. EPW can be executed in parallel by splitting the calculation over electron or phonon wave vectors. The parallelization strategy in EPW is tailored to each step of the computation. The capabilities of EPW are illustrated through example calculations by describing three systems: (i) lead being as a simple metal as a strong-coupling superconductor, (ii) graphene being a 2D zero-gap semiconductor with linear quasi-particle dispersion close to the Fermi-level, (iii) silicon carbide being a wide-gap semiconductor which becomes a low-temperature superconductor upon boron doping. Finally, there is a discussion on some future directions for additional development.