Abstract:Accurately predicting activation energies is crucial for understanding chemical reactions and modeling complex reaction systems. However, the high computational cost of quantum chemistry methods often limits the feasibility of large-scale studies, leading to a scarcity of high-quality activation energy data. In this work, we explore and compare three innovative approaches—transfer learning, delta learning, and feature engineering—to enhance the accuracy of activation energy predictions using graph neural networks, specifically focusing on methods that incorporate low-cost, low-level computational data. Using the Chemprop model, we systematically evaluated how these methods leverage data from semiempirical quantum mechanical (SQM) calculations to improve predictions. Our findings reveal that delta learning, which adjusts low-level SQM activation energies to match high-level CCSD(T)-F12a targets, offers the greatest accuracy improvement. However, this approach also requires more computational resources due to the complexity of transition state searches for each query of activation energy during model application. Transfer learning, which pretrains models on large datasets of low-level data, provided mixed results, particularly when there was a mismatch in the reaction distributions between the training and target datasets. Feature engineering, which involves adding computed molecular properties as input features, showed modest gains, particularly when incorporating thermodynamic properties. Our study highlights the trade-offs between accuracy and computational demand in selecting the best approach for enhancing activation energy predictions. These insights provide valuable guidelines for researchers aiming to apply machine learning in chemical reaction engineering, helping to balance accuracy with resource constraints.

Correlated RNN Framework to Quickly Generate Molecules with Desired Properties for Energetic Materials in the Low Data Regime

Deep learning for energetic material detonation performance

Prediction of Energetic Material Properties from Electronic Structure Using 3D Convolutional Neural Networks

Deep learning for synthetic microstructure generation in a materials-by-design framework for heterogeneous energetic materials

Machine Learning Enabled Quickly Predicting of Detonation Properties of N‐Containing Molecules for Discovering New Energetic Materials

Applying machine learning techniques to predict the properties of energetic materials

ReacNetGenerator: an Automatic Reaction Network Generator for Reactive Molecular Dynamics Simulations.

Low-data deep quantum chemical learning for accurate MP2 and coupled-cluster correlations

Molecular descriptor-enhanced graph neural network for energetic molecular property prediction

Interpretable and Physicochemical-Intuitive Deep Learning Approach for the Design of Thermal Resistance of Energetic Compounds

Data-Driven Tabulation for Chemistry Integration Using Recurrent Neural Networks

Design of Organic Electronic Materials With a Goal-Directed Generative Model Powered by Deep Neural Networks and High-Throughput Molecular Simulations

Data Distillation for Neural Network Potentials toward Foundational Dataset

Diffusion-Driven Generative Framework for Molecular Conformation Prediction

High Pressure and Temperature Neural Network Reactive Force Field for Energetic Materials

Force field-inspired transformer network assisted crystal density prediction for energetic materials

PARC: Physics-Aware Recurrent Convolutional Neural Networks to Assimilate Meso-scale Reactive Mechanics of Energetic Materials

IRNet: A General Purpose Deep Residual Regression Framework for Materials Discovery

Deep Spatial Learning with Molecular Vibration

Machine Learning of Energetic Material Properties

Enhancing Activation Energy Predictions under Data Constraints Using Graph Neural Networks