Multilayer Multistate Reaction Coordinate Model for Nonadiabatic Dynamics in Open Quantum Systems
- 1New York University Shanghai
- 2New York University
Abstract
Simulating nonadiabatic quantum dynamics in complex condensed-phase systems is computationally challenging due to the high dimensionality of the nuclear environment. To address this, we introduce the multilayer multistate reaction coordinate (ML-MRC) model, a general hierarchical transformation of the multistate harmonic (MSH) Hamiltonian that maps the complex bath into a structured chain of collective modes. We demonstrate that the ML-MRC framework allows the MSH model to be reformulated into hierarchies of arbitrary depth, from single-layer reaction coordinates to full-chain representations, while maintaining rigorous mathematical equivalence. This flexibility is validated across diverse physical systems, including the spin-boson model, the Frenkel exciton model for the photosynthetic light-harvesting Fenna-Matthews-Olson complex, and a realistic photoinduced charge transfer model for a carotenoid-porphyrin-fullerene triad in tetrahydrofuran. Furthermore, this hierarchical structure enables a robust dimensionality reduction strategy, where truncating the chain yields low-dimensional effective Hamiltonians that accurately reproduce the nonadiabatic dynamics of the untruncated model at a reduced computational cost. Finally, the ML-MRC model reveals valuable mechanistic insights through reaction coordinate trajectories, which are shown to synchronize remarkably with electronic population dynamics across multiple systems.
Related
- Non-Markovian exceptional points and information backflow: a Heisenberg reaction-coordinate approach
- Energy conserving variational dynamics with Gaussian wavepackets in the nonadiabatic regime
- Nonadiabatic molecular dynamics simulation with external field
- Nonadiabatic vibrational dynamics of CO on Pt(111) studied with first-principles many-body perturbation theory
- Analytic Derivative Couplings for MRSF-TDDFT: Applications to Nonadiabatic Molecular Dynamics and Conical Intersections
BibTeX
@misc{zeng2026multilayer,
title = {{Multilayer Multistate Reaction Coordinate Model for Nonadiabatic Dynamics in Open Quantum Systems}},
author = {Hao Zeng and Xiang Sun},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15007616/v2},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15007616/v2},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Multilayer Multistate Reaction Coordinate Model for Nonadiabatic Dynamics in Open Quantum Systems
AU - Hao Zeng
AU - Xiang Sun
PY - 2026
DA - 2026/10/14
PB - ChemRxiv
DO - 10.26434/chemrxiv.15007616/v2
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15007616/v2
AB - Simulating nonadiabatic quantum dynamics in complex condensed-phase systems is computationally challenging due to the high dimensionality of the nuclear environment. To address this, we introduce the multilayer multistate reaction coordinate (ML-MRC) model, a general hierarchical transformation of the multistate harmonic (MSH) Hamiltonian that maps the complex bath into a structured chain of collective modes. We demonstrate that the ML-MRC framework allows the MSH model to be reformulated into hierarchies of arbitrary depth, from single-layer reaction coordinates to full-chain representations, while maintaining rigorous mathematical equivalence. This flexibility is validated across diverse physical systems, including the spin-boson model, the Frenkel exciton model for the photosynthetic light-harvesting Fenna-Matthews-Olson complex, and a realistic photoinduced charge transfer model for a carotenoid-porphyrin-fullerene triad in tetrahydrofuran. Furthermore, this hierarchical structure enables a robust dimensionality reduction strategy, where truncating the chain yields low-dimensional effective Hamiltonians that accurately reproduce the nonadiabatic dynamics of the untruncated model at a reduced computational cost. Finally, the ML-MRC model reveals valuable mechanistic insights through reaction coordinate trajectories, which are shown to synchronize remarkably with electronic population dynamics across multiple systems.
ER -