Accelerated dinuclear palladium catalyst identification through unsupervised machine learning

Science, Journal Year: 2021, Volume and Issue: 374(6571), P. 1134 - 1140

Published: Nov. 25, 2021

Although machine learning bears enormous potential to accelerate developments in homogeneous catalysis, the frequent need for extensive experimental data can be a bottleneck implementation. Here, we report an unsupervised workflow that uses only five points. It makes use of generalized parameter databases are complemented with problem-specific silico acquisition and clustering. We showcase power this strategy challenging problem speciation palladium (Pd) catalysts, which mechanistic rationale is currently lacking. From total space 348 ligands, algorithm predicted, experimentally verified, number phosphine ligands (including previously never synthesized ones) give dinuclear Pd(I) complexes over more common Pd(0) Pd(II) species.

Language: Английский

---

Importance of Engineered and Learned Molecular Representations in Predicting Organic Reactivity, Selectivity, and Chemical Properties

Accounts of Chemical Research, Journal Year: 2021, Volume and Issue: 54(4), P. 827 - 836

Published: Feb. 3, 2021

ConspectusMachine-readable chemical structure representations are foundational in all attempts to harness machine learning for the prediction of reactivities, selectivities, and properties directly from molecular structure. The featurization discrete structures into a continuous vector space is critical phase undertaken before model selection, development new ways quantitatively encode molecules an active area research. In this Account, we highlight application suitability different representations, expert-guided "engineered" descriptors automatically "learned" features, tasks relevant organic organometallic chemistry, where differing amounts training data available. These include statistical models stereo- enantioselectivity, thermochemistry, kinetics developed using experimental quantum data.The use provides opportunity incorporate knowledge, domain expertise, physical constraints modeling. applications stereoselective catalysis, sets may be relatively small 3D-geometries conformations play important role, mechanistically informed features can used successfully obtain predictive that also chemically interpretable. We provide overview several recent approach quantitative reactivity selectivity, topological descriptors, mechanical calculations electronic steric properties, along with conformational ensembles, feature as essential ingredients used.Alternatively, more flexible, general-purpose such attributed graphs approaches learn complex relationship between target. This has potential out-perform traditional representation methods "hand-crafted" particularly set sizes grow. One large train structure–property relationships. A general toward curating useful highly accurate graph neural network discussed context bond dissociation enthalpies, strategy outperforms regression precomputed descriptors.Finally, describe how predictions incorporated selectivity. Once trained, avoids expensive computational overhead associated calculations, while maintaining interpretability. illustrate examples which fast enthalpy identities radicals formed through cleavage molecule's weakest simple site-selectivity reactivity.

Language: Английский

---

Predictive chemistry: machine learning for reaction deployment, reaction development, and reaction discovery

Chemical Science, Journal Year: 2022, Volume and Issue: 14(2), P. 226 - 244

Published: Nov. 28, 2022

This review outlines several organic chemistry tasks for which predictive machine learning models have been and can be applied.

Language: Английский

---

Embracing data science in catalysis research

Nature Catalysis, Journal Year: 2024, Volume and Issue: 7(6), P. 624 - 635

Published: April 23, 2024

Language: Английский

---

In Silico Screening of P,N-Ligands Facilitates Optimization of Au(III)-Mediated S-Arylation

Chemical Science, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

In silico examination of 13 P , N -ligated Au( iii ) OACs determined the key mechanistic factors governing )-mediated S -arylation. Three complexes were synthesized which exhibited bimolecular coordination rate constants as high 20 200 M −1 s .

Language: Английский

---