Experimental heats of formation and enthalpies obtained from G4 calculations both find that the resonance stabilization of the two unpaired electrons in triplet O2, relative to the unpaired electrons in two hydroxyl radicals, amounts to 100 kcal/mol. The origin of this huge stabilization energy is described within the contexts of both molecular orbital (MO) and valence-bond (VB) theory. Although O2 is a triplet diradical, the thermodynamic unfavorability of both its hydrogen atom abstraction and oligomerization reactions can be attributed to its very large resonance stabilization energy. The unreactivity of O2 toward both these modes of self-destruction maintains its abundance in the ecosphere and thus its availability to support aerobic life. However, despite the resonance stabilization of the π system of triplet O2, the weakness of the O–O σ bond makes reactions of O2, which eventually lead to cleavage of this bond, very favorable thermodynamically.
29 January 2020 in Journal of the American Chemical Society
There is a perceived dichotomy between structure-based and descriptor-based molecular representations used for predictive chemistry tasks. Here, we study the performance, generalizability, and explainability of the quantum mechanics-augmented graph neural network (ml-QM-GNN) architecture as applied to the prediction of regioselectivity (classification) and of activation energies (regression). In our hybrid QM-augmented model architecture, structure-based representations are first used to predict a set of atom- and bond-level reactivity descriptors derived from density functional theory calculations. These estimated reactivity descriptors are combined with the original structure-based representation to make the final reactivity prediction. We demonstrate that our model architecture leads to significant improvements over structure-based GNNs in not only overall accuracy but also in generalization to unseen compounds. Even when provided training sets of only a couple hundred labeled data points, the ml-QM-GNN outperforms other state-of-the-art structure-based architectures that have been applied to these tasks as well as descriptor-based (linear) regressions. As a primary contribution of this work, we demonstrate a bridge between data-driven predictions and conceptual frameworks commonly used to gain qualitative insights into reactivity phenomena, taking advantage of the fact that our models are grounded in (but not restricted to) QM descriptors. This effort results in a productive synergy between theory and data science, wherein QM-augmented models provide a data-driven confirmation of previous qualitative analyses, and these analyses in turn facilitate insights into the decision-making process occurring within ml-QM-GNNs.
29 May 2019 in Journal of the American Chemical Society