Meng Qingyong, Chen Jun, Zhang Dong H
State Key Laboratory of Molecular Reaction Dynamics and Center for Theoretical and Computational Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Zhongshan Road 457, 116023 Dalian, China.
J Chem Phys. 2015 Sep 14;143(10):101102. doi: 10.1063/1.4930860.
The ring polymer molecular dynamics (RPMD) calculations are performed to calculate rate constants for the title reaction on the recently constructed potential energy surface based on permutation invariant polynomial (PIP) neural-network (NN) fitting [J. Li et al., J. Chem. Phys. 142, 204302 (2015)]. By inspecting convergence, 16 beads are used in computing free-energy barriers at 300 K ≤ T ≤ 1000 K, while different numbers of beads are used for transmission coefficients. The present RPMD rates are in excellent agreement with quantum rates computed on the same potential energy surface, as well as with the experimental measurements, demonstrating further that the RPMD is capable of producing accurate rates for polyatomic chemical reactions even at rather low temperatures.
基于置换不变多项式(PIP)神经网络(NN)拟合[J. Li等人,《化学物理杂志》142, 204302 (2015)],在最近构建的势能面上进行了环聚合物分子动力学(RPMD)计算,以计算标题反应的速率常数。通过检查收敛性,在300 K ≤ T ≤ 1000 K计算自由能垒时使用了16个珠子,而在计算传输系数时使用了不同数量的珠子。目前的RPMD速率与在同一势能面上计算的量子速率以及实验测量结果非常吻合,进一步证明了RPMD即使在相当低的温度下也能够为多原子化学反应产生准确的速率。