# Exciton transport in the PE545 complex: insight from atomistic QM/MM-based quantum master equations and elastic network models

## TL;DR

This study combines atomistic QM/MM simulations and elastic network models to analyze environmental noise effects on exciton transport in the PE545 complex, revealing the dominance of site-energy autocorrelations and the role of protein collective motions.

## Contribution

It introduces a parameterization of environmental noise for the HSR model based on atomistic simulations and links low-frequency protein modes to exciton transport pathways.

## Key findings

- Site-energy autocorrelations dominate exciton mobility enhancement.
- Low-frequency protein modes correlate with exciton transfer pathways.
- Environmental fluctuations are faster than exciton transfer timescales.

## Abstract

In this paper we work out a parameterization of the environment noise within the Haken-Strobl-Reinenker (HSR) model for the PE545 light-harvesting complex based on atomic-level quantum mechanics/molecular mechanics (QM/MM) simulations. We use this approach to investigate the role of different auto- and cross-correlations in the HSR noise tensor, confirming that site-energy autocorrelations (pure dephasing) terms dominate the noise-induced exciton mobility enhancement, followed by site energy-coupling cross-correlations for specific triplets of pigments. Interestingly, several cross-correlations of the latter kind, together with coupling-coupling cross-correlations, display clear low-frequency signatures in their spectral densities in the region 30-70 inverse centimeters. These slow components lie at the limits of validity of the HSR approach, requiring that environmental fluctuations be faster than typical exciton transfer time scales. We show that a simple coarse-grained elastic-network-model (ENM) analysis of the PE545 protein naturally spotlights collective normal modes in this frequency range, that represent specific concerted motions of the subnetwork of cysteines that are covalenty linked to the pigments. This analysis strongly suggests that protein scaffolds in light-harvesting complexes are able to express specific collective, low-frequency normal modes providing a fold-rooted blueprint of exciton transport pathways. We speculate that ENM-based mixed quantum classical methods, such as Ehrenfest dynamics, might be promising tools to disentangle the fundamental designing principles of these dynamical processes in natural and artificial light-harvesting structures.