Ab initio complex band structure of conjugated polymers: Effects of hydrid DFT and GW schemes

TL;DR

This paper compares computational methods for calculating the complex band structure of conjugated polymers, showing hybrid DFT and GW schemes improve accuracy in predicting charge transfer properties.

Contribution

It systematically evaluates local, semilocal, hybrid DFT, and GW methods for complex band structure calculations, highlighting the improvements of hybrid and GW approaches over standard functionals.

Findings

Local and semilocal DFT underestimate the eta parameter.

Hybrid-exchange functionals partially correct this underestimation.

GW calculations and experiments show better agreement with hybrid DFT results.

Abstract

The non-resonant tunneling regime for charge transfer across nanojunctions is critically dependent on the so-called \beta{} parameter, governing the exponential decay of the current as the length of the junction increases. For periodic materials, this parameter can be theoretically evaluated by computing the complex band structure (CBS) -- or evanescent states -- of the material forming the tunneling junction. In this work we present the calculation of the CBS for organic polymers using a variety of computational schemes, including standard local, semilocal, and hybrid-exchange density functionals, and many-body perturbation theory within the GW approximation. We compare the description of localization and \beta{} parameters among the adopted methods and with experimental data. We show that local and semilocal density functionals systematically underestimate the \beta{} parameter, while hybrid-exchange schemes partially correct for this discrepancy, resulting in a much better agreement with GW calculations and experiments. Self-consistency effects and self-energy representation issues of the GW corrections are discussed together with the use of Wannier functions to interpolate the electronic band-structure.