Dielectric-dependent hybrid functionals for heterogeneous materials

Abstract

We derive a dielectric-dependent hybrid functional which accurately describes the electronic properties of heterogeneous interfaces and surfaces, as well as those of three-and two-dimensional bulk solids. The functional, which does not contain any adjustable parameter, is a generalization of self-consistent hybrid functionals introduced for homogeneous solids, where the screened Coulomb interaction is defined using a spatially varying, local dielectric function. The latter is determined self-consistently using density functional calculations in finite electric fields. We present results for the band gaps and dielectric constants of 3D and 2D bulk materials, and band offsets for interfaces, showing an accuracy comparable to that of GW calculations.

Structure and Energetics of Dye-Sensitized NiO Interfaces in Water from Ab Initio MD and Large-Scale GW Calculations

“…Coordinate-dependent hybrid functionals, developed using Green's functions formalism, have been recently introduced to account for the spatial dependence of the electron-electron correlation in hybrid materials. 48,49 These functionals are based on the assumption that the screened Coulomb interaction can be approximated through a spatial-dependent local dielectric function and have been only recently applied to heterogeneous systems with good results. [48][49][50] However, this approximation, as all DFT based methods, neglect dynamical polarization or image charge effects, which can be rather pronounced for the dye electronic states in the dye@SC interface and can critically influence the energy level alignment.…”

“…48,49 These functionals are based on the assumption that the screened Coulomb interaction can be approximated through a spatial-dependent local dielectric function and have been only recently applied to heterogeneous systems with good results. [48][49][50] However, this approximation, as all DFT based methods, neglect dynamical polarization or image charge effects, which can be rather pronounced for the dye electronic states in the dye@SC interface and can critically influence the energy level alignment. 46,[55][56][57][58] These higher-order effects, as well as the spatial dependence of the electronic correlation, can be accounted for by using more refined approximations, such as many body perturbation theory (MBPT) based methods, where the dynamical polarization response to the addition or to the removal of a particle are rigorously described.…”

Machine learning dielectric screening for the simulation of excited state properties of molecules and materials

“…In order to interpret our findings, we express Δ τ in terms of projective dielectric eigenpotentials, (PDEP) 99,100 and we decompose f Avg ( r ) into contributions from each individual PDEP, 101 i.e. , where and ϕ i is the i -th eigenpotential of the static dielectric matrix corresponding to the eigenvalue λ i .…”

“…In order to interpret our ndings, we express Ds in terms of projective dielectric eigenpotentials, (PDEP) 99,100 and we decompose f Avg (r) into contributions from each individual PDEP, 101 i.e., f…”

M06-SX screened-exchange density functional for chemistry and solid-state physics

“…3) Dielectric and correlation effects in which the bare exchange hole is screened by electron correlation are not treated properly by the HF approximation (11); in fact the HF approximation has unscreened long-range exchange (i.e., long-range exchange with a unit dielectric constant), which is unphysical for solids, especially for metals (12,13). One way to overcome the third downside is to mix HF and local exchange by using a dielectric constant (14,15), and dielectric-dependent functionals with long-range exchange have been used successfully for condensedmatter applications (14,(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28). In such theories, the ratio of the unscreened exchange to the screened exchange is an effective dielectric constant; this kind of physical consideration shows that the percentage of HFX that provides the best model need not be constant (29).…”