An efficient implementation of two-component relativistic exact-decoupling methods for large molecules
Daoling Peng, Nils Middendorf, Florian Weigend, and Markus Reiher

TL;DR
This paper introduces an efficient algorithm for large-scale two-component relativistic calculations that incorporate spin-orbit coupling, utilizing local approximations and symmetry to significantly speed up computations for large molecules.
Contribution
The paper presents a new, optimized implementation of relativistic exact-decoupling methods that reduces computational cost for large molecules by combining local schemes and symmetry considerations.
Findings
Successfully applied to silver clusters with up to 309 atoms
Achieved significant speed-up in relativistic density functional calculations
Demonstrated accurate extrapolation of cohesive energy to bulk materials
Abstract
We present an efficient algorithm for one- and two-component relativistic exact-decoupling calculations. Spin-orbit coupling is thus taken into account for the evaluation of relativistically transformed (one-electron) Hamiltonian. As the relativistic decoupling transformation has to be evaluated with primitive functions, the construction of the relativistic one-electron Hamiltonian becomes the bottleneck of the whole calculation for large molecules. For the established exact-decoupling protocols, a minimal matrix operation count is established and discussed in detail. Furthermore, we apply our recently developed local DLU scheme [J. Chem. Phys. 136 (2012) 244108] to accelerate this step. With our new implementation two-component relativistic density functional calculations can be performed invoking the resolution-of-identity density-fitting approximation and (Abelian as well as…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
