Quantum Reaction Dynamics Study Of Cl+CH₄and HH+H₃⁺in Gas Phase

Quantum Reaction Dynamics is a molecular level mechanism of elementary chemicaland physical processes. During the past few decades, quantum dynamics methods welldeveloped for small reaction systems. Use it, We can exactly calculate some dynamicsproperties: initial-state-selected reaction probability, state-to-state reaction probability, rateconstant and cross sections. Quantum reaction dynamics can help us explain the microscopicnature of chemical phenomena to predict experimental results and the experimentalphenomenon, and then to guide the industrial production according to the relevant theoreticalresults. Currently, quantum reaction dynamics has been widely applied in the field of newmaterials, new energy, environmental management, chemical production andpharmacological analysis.This paper presents the quantum reaction dynamics study of Cl+CH4and HD+H3+ingas phase. They are all carried out by the method of time-dependent wave packet propagationwith reduced dimensional method.Firstly, for the quantum reactive dynamics, six-degrees-of-freedom, time-dependentwavepacket propagation method is applied to study the Cl+CH4â†'HCl+CH3reaction onthe newly published potential energy surface(PES) by Czakó and Bowman. This is the firsttime to carry out the reaction dynamics study on this new PES. We confirm not only theexperimental speculation of the reactive resonance by observing a prominent resonance peakon the ground state reaction probability, but also the experimental and quasi-classicaltrajectory finding that at lower total scattering energy the translational energy drives thereactivity more than the vibrational energy for this late barrier reaction. The vibrationalmotions of CH4enhance the reactivity, and the C-H stretching motion has the biggest impacton the reactivity. The vibrational energy overall plays a more efficient role in the reactivitythan the translational energy except at the lower scattering energy. So Polanyi rule can beapplied to this system. The Energy Shifting approximation is employed to obtain an approximate full-dimensional cumulative reaction probability based on the six dimensionalcalculation. The calculated thermal rate coefficients agree very well with experimentalmeasurements after using experimental vibrational frequencies and zero point energy tocorrect the reactant vibrational partition function and to convert the energy for the fulldimensional cumulative reaction probability.Secondly, time-dependent, quantum reaction dynamics wavepacket approach is alsoemployed to investigate the impacts of the translational, vibrational, and rotational motion onthe HD+H3+â†'H2D++H2reaction using the Xie-Braams-Bowman potential energy surface.We treat this five atom reaction with a seven-degree-of-freedom model by fixing one Jacobiand one torsion angle related to H3+at the lowest saddle point geometry of the potentialenergy surface. The initial state selected reaction probabilities show that the rotationalexcitations of H+-H2greatly enhance the reactivity with the reaction probabilities increaseddouble at high rotational states compared to the ground state. However, the vibrationalexcitations of H3+hinder the reactivity. So Polanyi rule can not apply to this system. Theground state reaction probability shows no reaction threshold for this exoergic reaction, andas the translational energy increases, the reaction probability decreases. Furthermore, reactiveresonances and zero point energy play very important roles on the reaction dynamics. Theobtained integral cross section has the character of an exoergic reaction without a threshold: itdecreases with the translational energy increasing. The calculated thermal rate constants usingthis seven-degree-of-freedom model are in agreement with a later experiment measurement.Because the current calculations have used the reduced dimensional method, and welook forward to the full dimensions of the quantum reaction dynamics computing. Ourpresent theoretical calculations can be used as a reference for the subsequent research. Inaddition, we also tried to find a new potential energy surface fitting method. Analogous to theartificial neural network, we tried to use a random decision tree approach method to fit thepotential energy surfaces. But this fitting method in the one-dimensional andtwo-dimensional fitting tasks’ performances are poor. We hope that the new amendmentapproach proposed, which could make the random decision tree method to be able to moreaccurately facilitate the work of the fitting potential energy surfaces.