Tian Li, Zhu Yongfa, Song Hongwei, Yang Minghui
State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan 430071, China.
Phys Chem Chem Phys. 2019 Jun 7;21(21):11385-11394. doi: 10.1039/c9cp02113b. Epub 2019 May 21.
The highly exothermic hydrogen abstraction reaction of the F atom with NH is investigated using the quasi-classical trajectory method on a newly developed potential energy surface (PES) for the ground electronic state. The full-dimensional PES is constructed by fitting 41 282 ab initio energy points at the level of UCCSD(T)-F12/aug-cc-pVTZ. The flexible fundamental invariant-neural network method is applied in the fitting, resulting in a total root mean square error of 0.13 kcal mol. On one hand, the calculated differential cross sections agree reasonably well with the experimental results and indicate that the reaction is dominated by the direct abstraction and stripping mechanisms while a considerable amount of reaction takes place by the indirect "yo-yo" mechanism. The product energy partition also reproduces well the experimental result, which can be understood according to the geometry change along the minimum energy path. On the other hand, the obtained vibrational state distribution of the product HF follows P≈P > P > P, less consistent with the scattered experimental results. In addition, the calculated thermal rate coefficients have practically no temperature dependence within the statistical errors.
使用准经典轨迹方法,在新开发的基态电子态势能面(PES)上研究了F原子与NH的高放热氢提取反应。通过在UCCSD(T)-F12/aug-cc-pVTZ水平上拟合41282个从头算能量点构建了全维PES。在拟合中应用了灵活的基本不变量神经网络方法,总均方根误差为0.13 kcal/mol。一方面,计算得到的微分截面与实验结果相当吻合,表明该反应以直接提取和剥离机制为主,同时有相当数量的反应通过间接“溜溜球”机制发生。产物能量分配也很好地再现了实验结果,这可以根据沿最小能量路径的几何变化来理解。另一方面,得到的产物HF的振动态分布遵循P≈P > P > P,与分散的实验结果不太一致。此外,在统计误差范围内,计算得到的热速率系数实际上与温度无关。