Understanding Thermal Stability in Energetic Materials via High-Throughput Quantum Molecular Dynamics Simulations
POSTER
Abstract
Thermal stability is a crucial factor in ensuring the safety of energetic materials (EMs). Current methods to evaluate the stability of materials, such as thermogravimetric analysis and differential scanning calorimetry can produce noisy data and are highly dependent on the experimental conditions. Furthermore, the molecular features that contribute to thermal stability are not well understood, which makes it a challenge to develop novel molecules with improved stability. Therefore, it is highly desirable to study EMs in a controlled simulation environment to better understand the factors controlling thermal stability.
In this work, high-throughput quantum molecular dynamics (QMD) simulations are employed to investigate the onset of decomposition reactions across a diverse set of EMs. From these simulations, effective kinetic parameters are extracted and correlated with experimental decomposition temperatures. Additionally, QMD simulations provide insight into the reactions that occur at the onset of decomposition. Such reactions are analyzed in the context of proposed trigger-linkages to reveal crucial chemistries. The results of these QMD simulations also support building a large database of material properties, which are used to train AI models for inverse design of novel EMs.
In this work, high-throughput quantum molecular dynamics (QMD) simulations are employed to investigate the onset of decomposition reactions across a diverse set of EMs. From these simulations, effective kinetic parameters are extracted and correlated with experimental decomposition temperatures. Additionally, QMD simulations provide insight into the reactions that occur at the onset of decomposition. Such reactions are analyzed in the context of proposed trigger-linkages to reveal crucial chemistries. The results of these QMD simulations also support building a large database of material properties, which are used to train AI models for inverse design of novel EMs.
*Research presented in this presentation was supported by the Laboratory Directed Research and Development program of Los Alamos National Laboratory under project number 20250006DR.
Presenters
-
R. Seaton S Ullberg
- Theoretical Division, Los Alamos National Laboratory