A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis

Abstract

The design of molecular catalysts typically involves reconciling multiple conflicting property requirements, largely relying on human intuition and local structural searches. However, the vast number of potential catalysts requires pruning of the candidate space by efficient property prediction with quantitative structure–property relationships. Data-driven workflows embedded in a library of potential catalysts can be used to build predictive models for catalyst performance and serve as a blueprint for novel catalyst designs. Herein we introduce kraken, a discovery platform covering monodentate organophosphorus­(III) ligands providing comprehensive physicochemical descriptors based on representative conformer ensembles. Using quantum-mechanical methods, we calculated descriptors for 1558 ligands, including commercially available examples, and trained machine learning models to predict properties of over 300000 new ligands. We demonstrate the application of kraken to systematically explore the property space of organophosphorus ligands and how existing data sets in catalysis can be used to accelerate ligand selection during reaction optimization.

Trendbericht Physikalische Chemie 2022: In‐situ‐Spektroskopie und Katalyse

“…10) Dabei wurden neue Intermediate wie die Formylspezies [Pd(dtbpx)(CHO)(H)] (dtbpx = 1,2-Bis(di-tert-butylphosphino)xylene) identifiziert und die Kinetik der Reaktion vollständig analysiert. Die 1 H-und 13 C-NMR-Spektren wurden mit einem Abstand von je zehn Minuten aufgenommen, um ausreichend qualitative Struktur-und quantitative Konzentrationsinformationen zu erhalten.…”

Development and Molecular Understanding of a Pd‐Catalyzed Cyanation of Aryl Boronic Acids Enabled by High‐Throughput Experimentation and Data Analysis

“…For the monophosphine group A, all physicochemical descriptors were extracted from the kraken discovery platform. [64] For the bisphosphine group B, a workflow was initiated by performing conformer searches on the corresponding PdCl 2 adducts using xTB/crest [65 -68] and followed by structural refinement by DFT on the lowest energy conformer. Descriptors were extracted for the complexes and from separate single-point calculations on the stand-alone ligand without PdCl 2 .…”

“…[70,71] The bonding in phosphine ligands is constituted of two major components: σ-donation of the phosphine lone pair to an empty metal orbital and π-backdonation from a filled metal orbital to antibonding σ*-orbitals of the phosphineÀ substituent bonds. The bulkiness of the phosphine ligand is important to modulate the binding to the metal center and facilitate association/dissociation as well as the ligation state (how many ligands are bound to the metal for monodentate phosphine ligands [64] ). We sought to calculate descriptors that would address these features.…”

“…Given the negative coefficient, this steric descriptor correlates lower steric bulk to higher yields, possibly relating to an easily accessible additional coordination to the ligated catalyst. The interaction term E SOMO,rc Boltz :V tot (z +) vburminconf could relate to the σdonor ability of the phosphine (the SOMO energy of the radical cation is largely correlated to the nucleophilicity of a phosphine, as can be assessed by looking at correlations of this descriptor with for instance E HOMO in the larger kraken discovery platform [64] ) as a function of the total volume occupied in the z + hemisphere, which is the part of the metal coordination sphere that points away from the phosphine ligand. Its coefficient indicates that it is the defining parameter of the model and can be viewed as the ability to modulate the fine balance of σ-donation and sterics required for the coordination/decoordination to/from palladium.…”

A reactivity model for oxidative addition to palladium enables quantitative predictions for catalytic cross-coupling reactions

“…Emerging approaches in reactivity prediction that combine high-throughput experimentation [8][9][10][11][12][13][14] with molecular descriptor sets [15][16][17][18][19][20][21][22][23][24] and multivariate statistical analysis including machine learning [25][26][27][28][29][30][31][32][33][34] can accelerate the screening/optimization process and increase success rates; however, predictions generated by these approaches are oen limited to the specic reaction under investigation. Developing and rening the next generation of organic chemistry tools, including computer-aided synthesis design, automated reaction optimization, and predictive algorithms, 35 requires the development of general and quantitative frameworks that rapidly link molecular structure to reactivity for many different reactants and catalysts.…”