Thermal Decomposition Simulations of Hydroxylamine Pentazolate With Deep Neural Network Potential

Area of Science:

• Computational Chemistry
• Materials Science
• Chemical Engineering

Background:

• Limited understanding of microscopic reaction mechanisms in pentazole anion (N5-) salts.
• Need for accurate simulation methods for energetic materials.

Purpose of the Study:

• Develop a deep neural network potential (DNNP) model for N5- salts.
• Investigate the thermal decomposition mechanism of hydroxylamine pentazole (NH3OHN5) using molecular dynamics (MD) simulations.
• Elucidate atomic-scale reaction pathways and kinetics.

Main Methods:

• Constructed a high-precision DNNP model using active learning.
• Calibrated the DNNP model with first-principles data (Density Functional Theory - DFT).
• Performed large-scale MD simulations to analyze thermal decomposition.

Main Results:

• DNNP model showed excellent agreement with DFT for energy and atomic forces.
• Thermal decomposition initiated via a hydrogen transfer reaction.
• Protonation of N5- reduced ring-opening energy barrier, facilitating decomposition.
• Predominant decomposition products identified as N2, H2O, and NH3.

Conclusions:

• Elucidated the decomposition pathways and reaction mechanism of NH3OHN5 at the atomic scale.
• Demonstrated the capability of DNNP in simulating reaction dynamics of energetic materials.
• Provided a theoretical foundation for designing high-performance, green energetic materials.