Optical spectra from molecules to crystals: Insight from many-body perturbation theory

TL;DR

This paper compares the effectiveness of TDDFT with many-body perturbation theory in predicting optical spectra of molecular materials, highlighting the importance of correlation effects and the limitations of simpler approximations like ALDA.

Contribution

It demonstrates that high-level methods including correlation effects are necessary for accurate optical spectra predictions across different molecular environments.

Findings

ALDA fails to accurately predict excitation energies for molecular materials.

Correlation effects significantly influence optical excitation energies and character.

High-level methodologies are essential for precise optical spectra across phases.

Abstract

Time-dependent density-functional theory (TDDFT) often successfully reproduces excitation energies of finite systems, already in the adiabatic local-density approximation (ALDA). Here we show for prototypical molecular materials, i.e., oligothiophenes, that ALDA largely fails and explain why this is so. By comparing TDDFT with an in-depth analysis based on many-body perturbation theory, we demonstrate that correlation effects crucially impact energies and character of the optical excitations not only for molecules of increasing length and in crystalline environment, but even for isolated small molecules. We argue that only high-level methodologies, which explicitly include correlation effects, can reproduce optical spectra of molecular materials with equal accuracy from gas phase to crystal structures.