Search for: All records

DFT-level descriptor libraries were constructed to train 2D and 3D graph neural networks for on the-fly the prediction of carboxylic acid and alkyl amine descriptors suitable for statistical modeling of medicinally relevant molecules. 

“How strong is this Lewis acid?” is a question researchers often approach by calculating its fluoride ion affinity (FIA) with quantum chemistry. 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 differentp‐block atoms as the fluoride accepting site. The FIA49k dataset was used to train FIA‐GNN, two message‐passing graph neural networks, 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⁻¹spanned by FIA49k. The model's value was demonstrated with four case studies, including predictions for molecules extracted from the Cambridge Structural Database and by reproducing results from catalysis research available in the literature. Weaknesses of the model are evaluated and interpreted chemically. FIA‐GNN and the FIA49k dataset can be reached via a free web app (www.grebgroup.de/fia‐gnn).

 

Skeletal modifications enable elegant and rapid access to various derivatives of a compound that would otherwise be difficult to prepare. They are therefore a powerful tool, especially in the synthesis of natural products or drug discovery, to explore different natural products or to improve the properties of a drug candidate starting from a common intermediate. Inspired by the biosynthesis of the cephalotane natural products, we report here a single-atom insertion into the framework of the benzenoid subfamily, providing access to the troponoid congeners — representing the reverse of the proposed biosynthesis (i.e., a contra-biosynthesis approach). Computational evaluation of our designed transformation prompted us to investigate a Büchner–Curtius–Schlotterbeck reaction of ap-quinol methylether, which ultimately results in the synthesis of harringtonolide in two steps from cephanolide A, which we had previously prepared. Additional computational studies reveal that unconventional selectivity outcomes are driven by the choice of a Lewis acid and the nucleophile, which should inform further developments of these types of reactions.

 

Our self-evolving graph neural networks for predicting solubilities reconciled different magnitudes of errors and uncertainties of experimental and computational databases, maximizing the database size and the model’s prediction accuracy.

 

Benzo[ghi]perylene monoimides (BPIs) have recently been employed as organic photocatalysts for challenging reductions. In probing their function, we identify a thermal degradation product involving imide ring opening, and this in turn motivates the development and synthesis of a high-symmetry model systema benzo[ghi]perylene diester (BPDE)whose structural simplicity is useful for mechanistic exploration relevant to the broader photocatalyst class. Using electrochemical and spectroscopic tools, we probe both the singly and doubly reduced states of BPDE and report the generation of [BP-H]−, a twoelectron, one-proton activated closed-shell super-reductant. This catalytically active species, after visible photon absorption, operates from its singlet excited state, where the motions of the added proton are coupled to an electron transfer event, which enables direct reduction of inert substrates like benzene and fluorobenzene. Traditional Birch chemistry on benzene has been previously realized only by solvated electrons or electrochemistry. The function of this model system uncovered in these mechanistic explorations suggests modes of operation for this photocatalyst class that will enable future optimizations. 

Arynes are highly reactive and versatile intermediates for the functionalization of aromatic rings that are often generated using strong bases or fluoride sources, which, in some cases, can limit functional group tolerance. Here we demonstrate that triaryloxonium ions can be transformed into arynes through treatment with solid potassium phosphate at room temperature. A substantial range of functional group-bearing arynes, including 4,5-pyrimidynes, may be generated and trapped using cycloaddition reactions with high yields. Other arynophiles including nitrones, alkenes and azides are compatible with these conditions. Quantum computation in conjunction with an intramolecular kinetic isotope study is consistent with an elimination, unimolecular, conjugate base-like mechanism of elimination to form the aryne. These investigations demonstrate that the oxonium ion is a powerful electron-withdrawing group and a particularly effective leaving group. We anticipate that this study will stimulate further investigations into the synthetic utility of aryl oxonium ions.