Machine-learning-accelerated discovery of single-atom catalysts based on bidirectional activation mechanism

Mutual modulationviacharge transfer and unpaired electrons of catalytic sites for the superior intrinsic activity of N2reduction: from high-throughput computation assisted with a machine learning perspective

“…It is defined as the highest free energy change at which the pathway is endothermic. In NRR reaction, the first hydrogenation process has been identified as the most common PDS for NRR catalysts 28,53,54 . Therefore, herein the ΔG of the first hydrogenation smaller than 0.55 eV is adopted as the second criterion for screening.…”

“…Recently, data-driven method brought about great success in computer vision 20 , natural language processing 21 and speech recognition 22 . With the inspiration and opportunities of machine learning (ML), amounts of related studies using MLassisted method have emerged for prediction of formation energy of crystals 23 , superhard compounds 24 and catalytic activity [25][26][27][28][29][30][31] . In general, material features should be properly selected for descriptor-to-property mappings in ML.…”

“…The boron can not only be active anchoring site but also substitutional doped element into the substrate 27 . Substitutional doping of boron can largely retain the host lattice unchanged and simultaneously 4 induce tailored electronic structure which triggers the catalytic activity 28,29,43,44 . Kumar et al demonstrated that boron-doped MoS 2 nanosheet is an effective strategy for tailoring the physical properties of MoS 2 and fabricated it successfully in experiment 43 .…”

Learning the Use of Artificial Intelligence in Heterogeneous Catalysis

“…Discovering new materials is of the utmost importance from both technological and scientific points of view. That is why, in recent years, the use of artificial intelligence techniques to speed up their search has proliferated (Goldsmith et al, 2018;Gupta et al, 2018;Gómez-Peralta and Bokhimi 2020;Chen et al, 2021;Cho and Lin, 2021;Konno et al, 2021). Of particular interest is the finding of nanoporous materials because they have a large surface area that is attractive in heterogeneous catalysis (Cho and Lin, 2021).…”

Applications of Machine Learning in Alloy Catalysts: Rational Selection and Future Development of Descriptors

“…[ 18 ] With the enrichment of material databases and the enhancement of computing power, ML techniques have been widely used in the prediction of the activity, selectivity and stability of catalysts. [ 19 ] According to the well‐known Sabatier principle, moderate binding strength of key intermediates should be carefully tuned to achieve the highest activity. [ 20 ] The adsorption energy of key intermediates is thus usually used as the numerical target variable in the ML scheme of heterogeneous catalysis.…”