Machine learning meets volcano plots: computational discovery of cross-coupling catalysts

Abstract

The application of modern machine learning to challenges in atomistic simulation is gaining attraction.

Data Mining the C−C Cross‐Coupling Genome

“…Specifically, Figure uses the average value of ΔE RRS ( 2 ) for each of clusters in the Figure sketchmap, along with one additional group, X−M−CO, which represents the combination of one CO ligand with all other ligand types. Overall, our ML predictions for an additional set of ∼18000 catalysts showed only minor differences from DFT computed values (generally less than 3 kcal/mol), which aligns well with the mean absolute error of 2.73 kcal/mol inherent to the ML representation . This indicates that the ML derived data can be used to expand our set of catalysts (from ∼7000 to ∼25000 species), which provides more statistical validity to any chemical trends extracted when unraveling the chemical behavior of catalysts, which to the best of our knowledge, has not been done before.…”

“…Along this line, our research group has done extensive work in crafting a computational toolkit built on volcano plots, commonly used tools for identifying attractive species in heterogeneous/electro‐catalysis, to study homogeneous reactions . Volcano plots are built upon Sabatier's principle, which states that the interaction between a catalyst and a substrate should be neither too weak nor too strong.…”

“…In an attempt to expedite the screening process, we recently turned to ML techniques in order to establish the value of the descriptor variable. This effort produced the desired effect and allowed us to considerably increase the number of species screened, up to 18000 catalysts, for the aforementioned Suzuki C−C cross‐coupling reaction …”

“…[13] Along this line, our research group has done extensive work in crafting a computational toolkit built on volcano plots, [14] commonly used tools for identifying attractive species in heterogeneous/electro-catalysis, [15,16] to study homogeneous reactions. [9,[17][18][19][20] Volcano plots are built upon Sabatier's principle, [21] which states that the interaction between a catalyst and a substrate should be neither too weak nor too strong. In addition to volcano plots, Sabatier's principle has also been used in other ways to understand and ultimately design better catalysts, [22] such as through the use of Kozuch and Shaik's energy span model.…”

Gold‐Catalyzed Cross‐Coupling Reactions: An Overview of Design Strategies, Mechanistic Studies, and Applications

“…However,f uture research should focus on developing selectivea nd scalable reactions that can proceed catalytically under ambient conditions and without the requirements of sacrificialo xidants and stoichiometric amounts of expensive reagents. [422,423] Traditionalt ransition-metal catalysis including crosscoupling reactions [424,425] are already being tested and advancedb ya pplying these modernc omputational tools. The substrate scope for dual Au and visible-light-mediated reactions is ratherl imited because most methodologies require the use of ad iazonium salt as ac oupling partner.S imilarly,o nly af ew examples are availablef or the use of Pd-free plasmonic catalysis by Au for CÀCc oupling reactions.…”