雷亚克夫
化学
热分解
动力学
分子动力学
放热反应
激进的
化学分解
分解
计算化学
物理化学
有机化学
量子力学
物理
原子间势
作者
Chunxing Ren,Han Liu,Xiaoxia Li,Li Guo
摘要
Understanding the similarities and differences of decomposition mechanisms of CL-20 and its cocrystals is of great interest for practical applications of CL-20 cocrystals. The responses of CL-20 cocrystals to thermal stimulus were investigated using ReaxFF molecular dynamics simulations of two representative cocrystals, CL-20/HMX and CL-20/TNT, under adiabatic conditions and comparing to the baseline system of pure CL-20. The comprehensive chemical details were revealed with the aid of the unique code of VARxMD. The three CL-20-involved reactive systems all exhibit a distinct three-stage character during adiabatic decomposition when using the double peaks of the major intermediate NO2 amount as the boundary. By taking advantage of the three-stage classification, a clear scenario for the similar stimulus-response of the CL-20 cocrystals can be elucidated for the dominant primary decomposition of CL-20 in stage I and the transition of favored chemical mechanisms from the generation of intermediates/radicals in stage II into their consumption to form stable products in stage III. The similar chemical behaviors are rooted in the dominance of CL-20 chemistry in the initial response of its cocrystals to thermal stimulus. The prolonged reaction zone uncovers the slowed decomposition kinetics of CL-20/HMX and CL-20/TNT, which is associated with the altered kinetics of CL-20 decomposition specifically by N-NO2 bond scission and CL-20 skeleton decay. The retarded CL-20 decomposition in its cocrystals consequently results in more moderate self-heating and less violent exothermic reactions that agrees with the experimental observations of improved stability and damaged detonation performance of CL-20 cocrystals, particularly for CL-20/TNT. The results obtained in this work suggest that ReaxFF MD simulations can provide useful insight for the modulated chemical properties of varied CL-20 cocrystals.
科研通智能强力驱动
Strongly Powered by AbleSci AI