Capturing Excited State Proton Transfer Dynamics with Reactive Machine Learning Potentials DOI
Umberto Raucci

The Journal of Physical Chemistry Letters, Journal Year: 2025, Volume and Issue: unknown, P. 4900 - 4906

Published: May 9, 2025

Excited state proton transfer is a fundamental process in photochemistry, playing crucial role fluorescence sensing, bioimaging, and optoelectronic applications. However, fully resolving its dynamics remains challenging due to the prohibitive computational cost of ab initio simulations need for ultrafast experimental techniques with high temporal resolution. Here, we tackle this challenge by using machine learning-driven excited molecular simulations. We propose an active learning framework powered enhanced sampling constructing high-quality training set potentials, which then use map reaction free energy landscape capture photorelaxation dynamics. Using 10-hydroxybenzo[h]quinoline as test case, our reveal barrierless occurring within ∼50 fs, accompanied significant red shift emission (∼1 eV), agreement findings. Furthermore, results highlight strong coupling between charge redistribution, facilitates rapid tautomerization process. These findings showcase power accurately capturing photochemical while enabling large-scale statistical sampling.

Language: Английский

Capturing Excited State Proton Transfer Dynamics with Reactive Machine Learning Potentials DOI
Umberto Raucci

The Journal of Physical Chemistry Letters, Journal Year: 2025, Volume and Issue: unknown, P. 4900 - 4906

Published: May 9, 2025

Excited state proton transfer is a fundamental process in photochemistry, playing crucial role fluorescence sensing, bioimaging, and optoelectronic applications. However, fully resolving its dynamics remains challenging due to the prohibitive computational cost of ab initio simulations need for ultrafast experimental techniques with high temporal resolution. Here, we tackle this challenge by using machine learning-driven excited molecular simulations. We propose an active learning framework powered enhanced sampling constructing high-quality training set potentials, which then use map reaction free energy landscape capture photorelaxation dynamics. Using 10-hydroxybenzo[h]quinoline as test case, our reveal barrierless occurring within ∼50 fs, accompanied significant red shift emission (∼1 eV), agreement findings. Furthermore, results highlight strong coupling between charge redistribution, facilitates rapid tautomerization process. These findings showcase power accurately capturing photochemical while enabling large-scale statistical sampling.

Language: Английский

Citations

0