Track synthesis, materials and catalysis as it is published.
Search papers, labs, and topics across Lattice.
This paper introduces Local-Resolvent Density-Functional Theory (LRes-DFT), a novel approach to linear-scaling Kohn-Sham calculations for insulators by utilizing the diagonal of the one-electron resolvent at complex energies, thereby avoiding explicit orbital calculations. The method achieves linear scaling by approximating orbital occupations with a smooth step function derived from summing resolvents, and its accuracy is validated against standard Kohn-Sham calculations for molecular crystals, demonstrating high precision. While efficient for large systems, LRes-DFT exhibits reduced accuracy in disordered or compressed systems and with diffuse basis functions.
Linear-scaling DFT calculations for insulators are achieved without orbitals or kinetic energy functionals by employing a local resolvent formulation with complex energies, matching standard Kohn-Sham accuracy for large molecular crystals.
Kohn-Sham density-functional theory is expensive because it needs orbitals that extend over the whole system. For insulators this can be avoided by using only the diagonal of the one-electron resolvent, $g(\mathbf{r},z)=\langle\mathbf{r}|(z-\hat h)^{-1}|\mathbf{r}\rangle$, at a few complex energies $z$. We call the scheme local-resolvent density-functional theory (LRes-DFT). The density and the band energy are linear in $g$, so neither orbitals nor a kinetic-energy functional are needed. The complex energies are $M$ equally spaced points on a circle that encloses the occupied levels, offset by half a spacing so that no point lies on the real axis. Summing the resolvent over them is equivalent to occupying each level with $1/(1+u^M)$, where $u$ is the level's energy measured from the circle's center in units of its radius. This is a smooth step that is near 1 inside the circle and near 0 outside, and it stays between 0 and 1 for every $M$. If the circle moves with the chemical potential, forces at fixed electron number are exact. Gauss-Legendre and Zolotarev rules need fewer points when the gap is small. In three dimensions we use eigenpairs of small windows around each molecule instead of resolvents. This is divide-and-conquer (DC) with a different occupation function, and it is as accurate as DC with Fermi-Dirac occupations. For H$_2$ molecular crystals of up to 1024 electrons it agrees with Kohn-Sham to $2.5\times10^{-7}$ Ha per electron. On one core it is slower than dense diagonalization below about 500 electrons and 3.5 times faster at 1024 electrons. It loses accuracy for disordered or compressed systems and for diffuse basis functions.