Predicting Lewis Acidity - Machine Learning the Fluoride Ion Affinity of p-Block-Atom-Based Molecules

Manuel Schmitt, Andreas Albers, Lukas M. Sigmund, Shree Sowndarya S. V., Philipp Erdmann, Robert S. Paton, Lutz Greb

Lewis acids are omnipresent in all branches of chemistry and the strength of a Lewis acid in its thermodynamic sense (global Lewis acidity) is often approached by calculating its fluoride ion affinity (FIA) with quantum chemistry.[1,2] Here, we present FIA49k, an extensive FIA dataset with 48,986 data points calculated at the RI-DSD-BLYP-D3(BJ)/def2-QZVPP//PBEh-3c level of theory, including 13 different p-block atoms as the fluoride accepting site. The FIA49k dataset was used to train FIA-GNN, two message-passing graph neural networks,[3] which predict gas and solution phase FIA values of molecules excluded from training with a mean absolute error of 14 kJ mol⁻¹ (r²=0.93) from the SMILES string of the Lewis acid as the only input. The level of accuracy is notable, given the wide energetic range of 750 kJ mol⁻¹ spanned by FIA49k. The model's value and weaknesses were evaluated and demonstrated with four case studies. FIA-GNN and the FIA49k dataset can be reached via a free web app (www.grebgroup.de/fia-gnn).

Figure 1.

1. K. O. Christe, D. A. Dixon, D. McLemore, W. W. Wilson, J. A. Sheehy, J. A. Boatz, J. Fluorine Chem. 2000, 101, 151.
2. P. Erdmann, J. Leitner, J. Schwarz, L. Greb, ChemPhysChem 2020, 21, 987.
3. P. C. St. John, C. Phillips, T. W. Kemper, A. N. Wilson, Y. Guan, M. F. Crowley, M. R. Nimlos, R. E. Larsen, J. Chem. Phys. 2019, 150, 234111.

 Manuel Schmitt