Thermal decomposition mechanism of melamine via ReaxFF molecular dynamics simulation

L Song and Z Mei and J Ye and TL Ren and X Ma and T Fang and DQ Wang and XH Ju, JOURNAL OF MOLECULAR MODELING, 31, 258 (2025).

DOI: 10.1007/s00894-025-06477-7

ContextMelamine, widely employed as a high-efficiency flame retardant, exhibits an intricate high-temperature degradation mechanism that remains poorly understood. Comprehensive insight into its pyrolysis behavior is critical for advancing flame-retardant material design. This study employs ReaxFF molecular dynamics simulations to investigate melamine's thermal decomposition, elucidating initial reaction pathways, intermediate species formation, and final product distribution. Results revealed that melamine undergoes three temperature-dependent reactions: dimerization, NH2 elimination, and ring-opening reactions. At 2500 K, the initial decomposition pathways of melamine involve (i) NH2 removal yielding C3N5H4 radicals, (ii) direct cleavage forming C2N4H4 and CN2H2, and (iii) NH2-assisted dehydrogenation generating NH3. The primary final products comprise NH3, CN2H2, H2, and HCN. Moreover, melamine undergoes a transition to an intermediate with an N-bridge structure, ultimately leading to the formation of a melem structure. This study enhances our understanding at the atomic level of the thermal decomposition mechanism of melamine. Future studies should focus on investigating melamine-based composite materials for the development of high-performance and environmentally friendly flame retardants.MethodsBased on the open source software LAMMPS, this study verified the applicability of the C/H/N ReaxFF force field in the melamine system and studied the thermal decomposition behavior of melamine through reactive molecular dynamics (RMD) simulation. The simulation was performed under the canonical ensemble (NVT) with a damping time constant of 100.0 fs. The integration of the atomic equations of motion was implemented using the velocity- Verlet algorithm, and the total simulation time was 1.0 ns. The RMD simulation trajectory was post-processed using the ChemTrayzer program, and the bond order cutoff was set to 0.3 for molecular identification, thereby supporting species distribution analysis and reaction pathway identification.

Return to Publications page