Barrier Height Prediction by Machine Learning Correction of Semiempirical Calculations

Predicting the enthalpy of formation of energetic molecules via conventional machine learning and GNN

Machine learning (ML) provides a promising method for efficiently and accurately predicting molecular properties. Using ML models to predict the enthalpy of formation of energetic molecules helps in fast screening of potential high-energy molecules, thereby accelerating the design of energetic materials. A persistent challenge is to determine the optimal featurization methods for molecular representation and use an appropriate ML model. Thus, in our study, we evaluate various featurization methods (CDS, ECFP, SOAP, GNF) and ML models (RF, MLP, GCN, MPNN), dividing them into two groups: conventional ML models and GNN models, to predict the enthalpy of formation of potential high-energy molecules. Our results demonstrate that CDS and SOAP have advantages over the ECFP, while the GNFs in GCN and MPNN models perform better. Furthermore, the MPNN model performs best among all models with a root mean square error (RMSE) as low as 8.42 kcal mol-1, surpassing even the best performing CDS-MLP model in conventional ML models. Overall, this study provides a benchmark for ML in predicting enthalpy of formation and emphasizes the tremendous potential of GNN in property prediction.

Structure-Based Reaction Descriptors for Predicting Rate Constants by Machine Learning: Application to Hydrogen Abstraction from Alkanes by CH3/H/O Radicals

Accurate determination of the thermal rate constants for combustion reactions is a highly challenging task, both experimentally and theoretically. Machine learning has been proven to be a powerful tool to predict reaction rate constants in recent years. In this work, three supervised machine learning algorithms, including XGB, FNN, and XGB-FNN, are used to develop quantitative structure-property relationship models for the estimation of the rate constants of hydrogen abstraction reactions from alkanes by the free radicals CH3, H, and O. The molecular similarity based on Morgan molecular fingerprints combined with the topological indices are proposed to represent chemical reactions in the machine learning models. Using the newly constructed descriptors, the hybrid XGB-FNN algorithm yields average deviations of 65.4%, 12.1%, and 64.5% on the prediction sets of alkanes + CH3, H, and O, respectively, whose performance is comparable and even superior to the corresponding one using the activation energy as a descriptor. The use of activation energy as a descriptor has previously been shown to significantly improve prediction accuracy ( Fuel 2022, 322, 124150) but typically requires cumbersome ab initio calculations. In addition, the XGB-FNN models could reasonably predict reaction rate constants of hydrogen abstractions from different sites of alkanes and their isomers, indicating a good generalization ability. It is expected that the reaction descriptors proposed in this work can be applied to build machine learning models for other reactions.

Finding defects in glasses through machine learning

Structural defects control the kinetic, thermodynamic and mechanical properties of glasses. For instance, rare quantum tunneling two-level systems (TLS) govern the physics of glasses at very low temperature. Due to their extremely low density, it is very hard to directly identify them in computer simulations. We introduce a machine learning approach to efficiently explore the potential energy landscape of glass models and identify desired classes of defects. We focus in particular on TLS and we design an algorithm that is able to rapidly predict the quantum splitting between any two amorphous configurations produced by classical simulations. This in turn allows us to shift the computational effort towards the collection and identification of a larger number of TLS, rather than the useless characterization of non-tunneling defects which are much more abundant. Finally, we interpret our machine learning model to understand how TLS are identified and characterized, thus giving direct physical insight into their microscopic nature.