Time-resolved role of coherence and delocalization in photosynthetic energy transfer from an extended exciton model

Abstract

Photosynthetic antenna complexes achieve high quantum efficiency through exciton transport in coupled pigment networks. Conventional Frenkel-exciton models treat each chromophore as a structureless site and neglect internal electronic degrees of freedom that can influence coherence and delocalization. Here we develop an extended excitonic network model that preserves the pigment-pigment coupling topology while introducing tunable intrachromophoric electronic mixing within the single-excitation manifold. Using a Lindblad open-quantum-system framework, we quantify coherence, delocalization, and trapping efficiency across parameter space. We show that intrachromophoric mixing plays a time-dependent role: enhanced mixing on the antenna side promotes short-time coherent delocalization and improves excitation injection, whereas excessive mixing near the trapping site induces persistent delocalization and suppresses transfer efficiency. Simulated two-dimensional electronic spectra reveal enhanced cross peaks and systematic blue shifts, providing spectroscopic signatures of coherence-modulated transport. These results establish a microscopic connection between internal electronic structure and quantum transport performance in excitonic networks.

Figures (7)

• Figure 1: (a) Schematic of the minimal extended FMO network along a representative energy‑transfer path: each chromophore is split into two intrachromophoric sites coupled by $U_i$; solid arrows denote conventional inter‑chromophore couplings, while dashed arrows indicate additional pathways introduced by the internal sites. The inset shows the spatial distribution of the 7 site FMO pigments, with red arrows indicating the main energy transfer pathway. (b) Corresponding energy‑level diagram: the intrachromophoric electronic mixing $U_i$ splits and mixes local excitonic states, reshaping the energy landscape and transport routes.
• Figure 2: Pearson correlation matrices for (a) intrachromophoric electronic mixing $U_i$ versus dynamical observables (left) and (b) for the observables among themselves (right). The observables include the RC population; the peak and final delocalization lengths $(L_{\text{peak}}, L_{\text{end}})$; the peak and final relative entropy of coherence $(E_{\text{peak}}, E_{\text{end}})$; and their differences $\Delta L = L_{\text{peak}} - L_{\text{end}}$ and $\Delta E = E_{\text{peak}} - E_{\text{end}}$. Colour denotes the strength and sign of the correlation.
• Figure 3: SE-path 2DES spectra for the standard FMO (a) and extended FMO (b) models at waiting times 0, 300 and 1000 fs (300 K). Diagonal peaks reflect excitonic populations, and cross peaks indicate coherence-assisted energy transfer. Panel (c) plots the time dependence of diagonal-peak intensities (top: FMO; middle: E-FMO) and the difference between cross peaks D and C (bottom), which quantifies net energy flow.
• Figure 4: SE-path 2DES difference spectra of the E-FMO model at 300 K for two parameter-scan configurations. The difference signal is defined as $\Delta S \equiv S(T=1~\mathrm{ps})-S(T=0.3~\mathrm{ps})$, so that positive (negative) features indicate spectral components that grow (decay) from 0.3 to 1 ps. In the Top row, $U_1=0.006$ and $U_2=0.016$ are fixed while $U_3$ is varied. And the Bottom row, $U_3=0.006$ is fixed while $U_1$ and $U_2$ are tuned as labeled. Diagonal peaks mainly track population redistribution during the incoherent-transfer stage, whereas cross peaks capture changes in coherence-assisted energy-transfer pathways.
• Figure 5: SE-path 2DES spectra of the extended FMO model at 300 K and $T=0.3$ ps. With $U_3=0.006$, the left and right panels compare low and high intrachromophoric electronic mixings $(U_1,U_2)$. The stronger cross-peak intensity at this early time reflects enhanced coherent phase correlations and greater exciton delocalization.