Is CH3NC isomerization an intrinsic non-RRKM unimolecular reaction?

Abstract

Direct dynamics simulations, using B3LYP/6-311++G(2d,2p) theory, were used to study the unimolecular and intramolecular dynamics of vibrationally excited CH3NC. Microcanonical ensembles of CH3NC, excited with 150, 120, and 100 kcal/mol of vibrational energy, isomerized to CH3CN nonexponentially, indicative of intrinsic non-Rice-Ramsperger-Kassel-Marcus (RRKM) dynamics. The distribution of surviving CH3NC molecules vs time, i.e., N(t)/N(0), was described by two separate functions, valid above and below a time limit, a single exponential for the former and a biexponential for the latter. The dynamics for the short-time component are consistent with a separable phase space model. The importance of this component decreases with vibrational energy and may be unimportant for energies relevant to experimental studies of CH3NC isomerization. Classical power spectra calculated for vibrationally excited CH3NC, at the experimental average energy of isomerizing molecules, show that the intramolecular dynamics of CH3NC are not chaotic and the C—N≡C and CH3 units are weakly coupled. The biexponential N(t)/N(0) at 100 kcal/mol is used as a model to study CH3NC → CH3CN isomerization with biexponential dynamics. The Hinshelwood-Lindemann rate constant kuni(ω,E) found from the biexponential N(t)/N(0) agrees with the Hinshelwood-Lindemann-RRKM kuni(ω,E) at the high and low pressure limits, but is lower at intermediate pressures. As found from previous work [S. Malpathak and W. L. Hase, J. Phys. Chem. A 123, 1923 (2019)], the two kuni(ω,E) curves may be brought into agreement by scaling ω in the Hinshelwood-Lindemann-RRKM kuni(ω,E) by a collisional energy transfer efficiency factor βc. The interplay between the value of βc, for the actual intermolecular energy transfer, and the ways the treatment of the rotational quantum number K and nonexponential unimolecular dynamics affect βc suggests that the ability to fit an experimental kuni(ω,T) with Hinshelwood-Lindemann-RRKM theory does not identify a unimolecular reactant as an intrinsic RRKM molecule.

Comparison of Exponential and Biexponential Models of the Unimolecular Decomposition Probability for the Hinshelwood–Lindemann Mechanism

“…In summary, the work presented here, and that presented previously, , indicates that fitting an experimental thermal unimolecular rate constant versus temperature and pressure with the Hinshelwood–Lindemann–RRKM model does not unambiguously identify the unimolecular reactant as a RRKM molecule.…”

Pragmatic Solution for a Fully E,J-Resolved Master Equation

“…This is in contrast to results obtained from classical trajectory calculations done by Bunker and Hase, , who reported nonstatistical behavior (e.g., nonexponential decay) for this reaction. Very recently, Hase and co-workers have restudied this reaction using trajectory calculations based on the density functional theory (DFT) based potential energy surface (PES) and again drawn the same conclusion. This reaction has also recently been investigated using the high accuracy extrapolated ab initio thermochemistry (HEAT) method for the potential energy surface (PES) in combination with Miller’s semiclassical transition state theory (SCTST) and the fixed- J 2DME method alluded to the previous section .…”

“…2500 cm −1 with the fixed-J 2DME approach. 33 The latter value seems to be unusually large, suggesting that its unphysical value is somehow absorbing into the parameter space the effects of the nonstatistical behavior noted by Hase et al 74 At the highpressure limit where the Boltzmann thermal energy distribution is established, rate constants calculated with the HEAT/ SCTST approach from first-principles (without any adjusted parameters) agree well with those of experiments within 20%. 33 Here, we investigate the effects of angular momentum transfer by collisions on the calculated rate constants in the falloff regime through a full solution of the 2DME.…”