LCAO-TDDFT-$k$-$\omega$: Spectroscopy in the Optical Limit

TL;DR

This paper introduces an efficient LCAO-TDDFT-$k$-$$ method for modeling optical absorption in low-dimensional systems, providing physical insights and matching experimental and $G_0 W_0$-BSE results.

Contribution

It develops a semi-quantitative, robust computational approach using LCAO-TDDFT-$k$-$$ with derivative discontinuity correction for optical spectra prediction.

Findings

Accurately models optical spectra of diverse low-dimensional systems.

Provides physical insight via electron-hole density analysis.

Demonstrates reliability and efficiency of the new method.

Abstract

Understanding, optimizing, and controlling the optical absorption process, exciton gemination, and electron-hole separation and conduction in low dimensional systems is a fundamental problem in materials science. However, robust and efficient methods capable of modelling the optical absorbance of low dimensional macromolecular systems and providing physical insight into the processes involved have remained elusive. We employ a highly efficient linear combination of atomic orbitals (LCAOs) representation of the Kohn--Sham (KS) orbitals within time dependent density functional theory (TDDFT) in the reciprocal space ($k$) and frequency ($\omega$) domains, as implemented within our LCAO-TDDFT-$k$-$\omega$ code, and apply the derivative discontinuity correction of the exchange functional $\Delta_x$ to the KS eigenenergies. In so doing we are able to provide a semi-quantitative description of the optical absorption, conductivity, and polarizability spectra for prototypical 0D, 1D, 2D, and 3D systems within the optical limit ($\|\bf{q}\|\to0^+$) as compared to both available measurements and from solving the Bethe$-$Salpeter equation with quasiparticle $G_0 W_0$ eigenvalues ($G_0 W_0$-BSE). Specifically, we consider 0D fullerene (C$_{60}$), 1D metallic (10,0) and semiconducting (10,10) single-walled carbon nanotubes (SWCNTs), 2D graphene (GR) and phosphorene (PN), and 3D rutile (R-TiO$_2$) and anatase (A-TiO$_2$). For each system, we also employ the spatially resolved electron-hole density to provide direct physical insight into the nature of their optical excitations. These results demonstrate the reliability, applicability, efficiency, and robustness of our LCAO-TDDFT-$k$-$\omega$ code, and open the pathway to the computational design of macromolecular systems for optoelectronic, photovoltaic, and photocatalytic applications $in$ $silico$.