Machine learning meets volcano plots: Computational discovery of cross-coupling catalysts
- Laboratory for Computational Molecular Design, Institute of Chemical Sciences and Engineering, École Polytechnique Fédérale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland;
- Institute of Physical Chemistry, Department of Chemistry, University of Basel, Klingelbergstrasse 80, CH-4056 Basel, Switzerland
(version: v1, submitted on: 01 August 2018)
How to cite this entry
Benjamin Meyer, Boodsarin Sawatlon, Clémence Corminboeuf, Stefan Niklaus Heinen, O. Anatole von Lilienfeld, Machine learning meets volcano plots: Computational discovery of cross-coupling catalysts, Materials Cloud Archive (2018), doi: 10.24435/materialscloud:2018.0014/v1.
The application of modern machine learning to challenges in atomistic simulation is gaining attraction. We present new machine learning models that can predict the energy of the oxidative addition process between a transition metal complex and a substrate for C-C cross-coupling reaction. In turn, this quantity can be used as a descriptor to estimate the activity of homogeneous catalysts using molecular volcano plots. The versatility of this approach is illustrated for vast libraries of organometallic catalysts based on Pt, Pd, Ni, Cu, Ag, and Au combined with 91 ligands. Out-of-sample machine learning predictions were made on a total of 18,062 compounds leading to 557 catalyst candidates falling into the ideal thermodynamic window. This number was further refined by searching for candidates with an estimated price lower than 10 US$/mmol. The 37 catalyst finalists are dominated by palladium phosphine ligand combinations but also include earth abundant (Cu) transition metal with less common ligands. Our results indicate that modern statistical learning techniques can be applied to the computational discovery of readily available and promising catalyst candidates.
Materials Cloud sections using this data
No Explore or Discover sections associated with this archive entry.
|30.9 MiB||The overall 25,116 generated structures of each catalytic intermediates.|
|10.6 MiB||The overall 7,054 optimized geometries at the B3LYP-D3/3-21G level of each catalytic intermediates.|
|782.4 KiB||The single point energies computed at the B3LYP-D3/def2-TZVP level, the corresponding binding energies and the 18,062 out-of-sample machine learning predicted binding energies using the Coulomb Matrix, Bag of Bonds and SLTAM representations|
01 August 2018 [This version]