This research proposes a multi-scale methodology combining QM/MM simulations, machine learning potentials, and nonadiabatic surface hopping dynamics to investigate how solvent environments modulate photoprotective mechanisms in chromophores.
Key findings
Photoprotective chromophores dissipate harmful UV radiation through ultrafast non-radiative decay pathways.
The efficiency of these mechanisms is influenced by solvent polarity, hydrogen bonding, viscosity, and temperature.
The framework integrates explicit solvent representation with neural network potentials for enhanced sampling.
Time-dependent density functional theory (TD-DFT) is used for excited-state characterization.
Trajectory surface hopping is employed for nonadiabatic dynamics.
Limitations & open questions
Current continuum solvation models cannot capture specific hydrogen-bonding interactions and dynamical solvent effects.
Explicit QM/MM simulations of nonadiabatic dynamics remain computationally prohibitive for statistically meaningful sampling.