Electron–phonon (e–ph) interactions are central to a wide range of phenomena in condensed matter physics and materials science, including phonon-driven superconductivity, electrical and thermal transport, and optical properties. Over the past decades, density functional theory (DFT), combined with density functional perturbation theory or finite-displacement techniques, has enabled quantitative materials-specific predictions of e–ph interactions from first-principles calculations. Yet, the predictive power of these calculations is highly limited by the choice of exchange–correlation functionals employed in DFT, especially for materials with open-shell d- and f-electrons. In this perspective, we argue that meta-generalized gradient approximation (GGA) functionals, particularly the strongly constrained and appropriately normed (SCAN) functional and its variant r2SCAN, offer a promising route toward more accurate, efficient, and transferable e–ph calculations across diverse classes of materials, surpassing the performance of conventional local density approximation and GGA functionals. We emphasize that the main opportunity is not merely improved electronic structures or phonons in isolation, but a more reliable, self-consistent description of the coupled electronic, magnetic, dielectric, and lattice responses that determine e–ph coupling matrix elements. We discuss why reliable e–ph calculations require better density functionals, why symmetry breaking combined with advanced density functionals is often essential in complex materials, where the present semilocal framework succeeds, where it still falls short, and what future developments will be useful for next-generation e–ph coupling simulations in terms of density-functional development.
Zhang et al. (Mon,) studied this question.