This work performs a time-dependent wavepacket study of the H2 + C2H → H + C2H2 reaction on a new ab initio potential energy surface (PES). The PES is constructed using neural network method based on 68 478 geometries with energies calculated at UCCSD(T)-F12a/aug-cc-pVTZ level and covers H2 + C2H↔H + C2H2, H + C2H2 → HCCH2, and HCCH2 radial isomerization reaction regions. The reaction dynamics of H2 + C2H → H + C2H2 are investigated using full-dimensional quantum dynamics method. The initial-state selected reaction probabilities are calculated for reactants in eight vibrational states. The calculated results showed that the H2 vibrational excitation predominantly enhances the reactivity while the excitation of bending mode of C2H slightly inhibits the reaction. The excitations of two stretching modes of C2H molecule have negligible effect on the reactivity. The integral cross section is calculated with J-shift approximation and the mode selectivity in this reaction is discussed. The rate constants over 200-2000 K are calculated and agree well with the experimental measured values.
The potential energy surface plays an important role in studying molecular reaction dynamics. In this work, a new method, namely the "multi-center partition" method, is proposed to construct the potential energy surface of H 3 . The optimized function is first determined by comparing the London-Eyring-Polanyi-Sato (LEPS) potential, the many-body expansion potential, and the permutation-invariant polynomial potential. This comparison shows that the permutation-invariant polynomial fitting proposed by Bowman is the most efficient method for describing the topology of the H 3 system. The quasi-classical trajectory method is used to analyze the rationality of those potential energy surfaces. By combining the multi-center partition method with the permutation-invariant polynomial method, the accuracy of the H 3 molecular potential energy surface is greatly improved and could possibly be used in the fitting of potential energy surfaces in other systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.