Excited‐State Distortions Promote the Photochemical 4π‐Electrocyclizations of Fluorobenzenes via Machine Learning Accelerated Photodynamics Simulations

Li, Jingbai; Lopez, Steven A.

doi:10.1002/chem.202200651

Cited by 13 publications

(20 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The training set contains the 901 structures from Wigner sampling and geometrical interpolation of the C1−C4 bond formation. 3 The control set contains 100 Dewar-HFB structures with the C1−C4 bond. The predictions show similar error distributions regardless of the permutations (Figure 6c).…”

Section: Data Augmentationmentioning

confidence: 99%

“…The [2 + 2] cycloaddition of [3]-ladderdienes can access cubanes. The quantum yield of cubane increases in the [3]- ladderdienes octasubstituted with methyl (CH 3 ), trifluoromethyl (CF 3 ), and cyclopropyl (cPr) groups.…”

Section: Substituent Effects On the Photochemical Reactions Of [3]-la...mentioning

confidence: 99%

“…96 The ML photodynamics simulations enable us to model the excited-state dynamics of HFB in 4 ns at the XMS-CASPT2(6,7)/aug-cc-pVDZ level. 3 The ML photodynamics simulations show that most HFBs reform the reactant. Figure 9b illustrates one of the HFB trajectories in the first 600 ps (0.6 ns).…”

Section: Photocyclization Of Hexafluorobenzenesmentioning

confidence: 99%

“…In this subsection, we use ML photodynamics simulations to explore the influence of the integration time step and number of trajectories on simulation of photochemical reactions. The model system is octacyclopropyl [3]-ladderdiene (A″).…”

Section: Remarks On ML Photodynamics Simulationsmentioning

confidence: 99%

See 3 more Smart Citations

A Look Inside the Black Box of Machine Learning Photodynamics Simulations

Lopez

2022

Acc. Chem. Res.

Self Cite

View full text Add to dashboard Cite

Metrics & MoreArticle Recommendations * sı Supporting Information CONSPECTUS: Photochemical reactions are of great importance in chemistry, biology, and materials science because they take advantage of a renewable energy source, mild reaction conditions, and high atom economy. Light absorption can excite molecules to a higher energy electronic state of the same spin multiplicity. The following nonadiabatic processes induce molecular transformations that afford exotic molecular architectures and high-energy-isomers that are inaccessible by thermal means. Computational simulations now complement time-resolved instrumentation to reveal ultrafast excited-state mechanistic information for photochemical reactions that is essential in disentangling elusive spectroscopic features, excitedstate lifetimes, and excited-state mechanistic critical points. Nonadiabatic molecular dynamics (NAMD), powered by surface hopping techniques, is among the most widely applied techniques to model the photochemical reactions of medium-sized molecules. However, the computational efficiency is limited because of the requisite thousands of multiconfigurational quantum-chemical calculations multiplied by hundreds of trajectories. Machine learning (ML) has emerged as a revolutionary force in computational chemistry to predict the outcome of the resource-intensive multiconfigurational calculations on the fly. An ML potential trained with a substantial set of quantum-chemical calculations can predict the energies and forces with errors under chemical accuracy at a negligible cost. The integration of ML potentials in NAMD dramatically extends the maximum simulation time scale by ∼10 000fold to the nanosecond regime.In this Account, we present a comprehensive demonstration of ML photodynamics simulations and summarize our most recent applications in resolving complex photochemical reactions. First, we address three fundamental components of ML techniques for photodynamics simulations: the quantum-chemical data set, the ML potential, and NAMD. Second, we describe best practices in building training data and our procedure toward training the ML photodynamics model with our recent literature contributions. We introduce a convenient training data generation scheme combining Wigner sampling and geometrical interpolation. It trains reliable and effective ML potentials suitable for subsequent active learning to detect undersampled data. We demonstrate how active learning automatically discovers new mechanistic pathways and reproduces experimental results. We point out that atomic permutation is an essential data augmentation approach to improve the learnability of distance-based molecular descriptors for highly symmetric molecules. Third, we demonstrate the utility of ML-photodynamics by showing the results of ML photodynamics simulations of (1) photo-torquoselective 4π disrotatory electrocyclic ring closing of norbornyl cyclohexadiene, which reveals a thermal conversion from experimentally unobserved intermediates to the reactant in 1 ns; (2) [2 + 2] ph...

show abstract

Section: Data Augmentationmentioning

confidence: 99%

Section: Substituent Effects On the Photochemical Reactions Of [3]-la...mentioning

confidence: 99%

Section: Photocyclization Of Hexafluorobenzenesmentioning

confidence: 99%

Section: Remarks On ML Photodynamics Simulationsmentioning

confidence: 99%

See 2 more Smart Citations

A Look Inside the Black Box of Machine Learning Photodynamics Simulations

Lopez

2022

Acc. Chem. Res.

Self Cite

View full text Add to dashboard Cite

show abstract

“…19 Exceptions are rare and limited to systems with ≲ 10 atoms. 17,20,21 Finally, achieving the desired accuracy when training reactive MLPs is challenging because the energy scale required for accurate dynamics is much larger than for non-reactive MLPs (where sampling around the minima is often sufficient). As such, more training data is required to reach the same accuracy.…”

Section: Introductionmentioning

confidence: 99%

Reaction dynamics of Diels–Alder reactions from machine learned potentials

Young

Johnston-Wood

Zhang

et al. 2022

Phys. Chem. Chem. Phys.

View full text Add to dashboard Cite

show abstract

Meta‐Ortho Effect on the Excited State Pathways of Chloroanilines

Nitu,

van der Wal,

Kaul

et al. 2023

Eur J Org Chem

View full text Add to dashboard Cite

Direct excitation of aromatic compounds grants access to high‐energy intermediates that can be utilised in organic synthesis. Understanding and predicting the substituent effects at the excited state for aromatic molecules remains challenging for the synthetic photochemist. In this work, we present an experimental and computational investigation of the excited state of the isomeric chloroanilines, which promptly react by losing the chloride when the amino group is in para position, but are non‐reactive and non‐emissive in the meta and ortho isomers. XMS‐CASPT2//CASSCF computations explain this apparent contradiction of the meta‐ortho selectivity rule of Zimmerman, which originates from the substituent effects lowering to a different extent the barrier to populate the prefulvenic conical intersection that deactivates non‐radiatively the singlet excited state of the chloroanilines.

show abstract

Excited‐State Distortions Promote the Photochemical 4π‐Electrocyclizations of Fluorobenzenes via Machine Learning Accelerated Photodynamics Simulations

Cited by 13 publications

References 49 publications

A Look Inside the Black Box of Machine Learning Photodynamics Simulations

A Look Inside the Black Box of Machine Learning Photodynamics Simulations

Reaction dynamics of Diels–Alder reactions from machine learned potentials

Meta‐Ortho Effect on the Excited State Pathways of Chloroanilines

Contact Info

Product

Resources

About