Application of SQMFF Vibrational Calculations to Transition States: DFT and ab Initio Study of the Kinetics of Methyl Azide and Ethyl Azide Thermolysis Juan F. Arenas,* Juan C. Otero, Adelaida Sa ´ nchez-Ga ´ lvez, and Juan Soto Department of Physical Chemistry, Faculty of Sciences, UniVersity of Ma ´ laga, E-29071-Ma ´ laga, Spain Pedro Viruela Departament of Physical Chemistry, UniVersitat de Vale ` ncia, Dr. Moliner 50, Burjassot, E-46100 Vale ` ncia, Spain ReceiVed: July 29, 1997; In Final Form: NoVember 12, 1997 DFT including nonlocal corrections and ab initio calculations at MP2 and MP4 levels of theory have been performed in order to provide information concerning the mechanism of the rate limiting step of the thermal decomposition of methyl azide and ethyl azide. The chemically interesting points of the ground-state potential energy surface have been fully optimized, and a detailed normal-mode analysis for the reagents and the transition states is presented. The well-established scaled quantum mechanical force field method has been used to obtain reliable vibrational frequencies for these molecular structures. The force fields of transition states have been modified by using the scale factors computed for the force fields of the azides in their ground state. Finally, the activation energies and the Arrhenius preexponential factors for the rate constant have been computed according to transition state theory. The best values for the activation energies are provided by B3-LYP/6-311+G**. For the preexponential factor, the agreement with experiment seems to be independent of the level of theory used. 1. Introduction The chemistry of organic azides has been widely developed in the last years. They constitute a versatile class of compounds used as important reagents in heterocycle syntheses 1 and in several biological methods. 2 For purpose of energy storage, these molecules are substantially higher in energy than their decomposition products, but with activation barriers sufficient for safe handling. Such characteristics have motivated their study as energetic additives for advanced solid propellants, 3-5 which is considered to be one of the practical ways to improve their technical performances. Energetic azides make a signifi- cant energy contribution to propellants and can also minimize the amount of flame and smoke in gases generated during the propulsion phase of solid propellants. In the past two decades, a large number of organic azides that find application in propellants have been synthesized and examined. 3-9 However, fundamental questions regarding the mechanism of the thermal decomposition of such molecules are not completely answered. It is well-known from experimental results that the first step in the thermolysis of azides involves nitrogen elimination to yield the respective imine as the unique product. 10-14 The strong exothermic character of this step is responsible for the applica- tion of azides as propellants and for the chemical activation of the imine molecule, which consequently reacts to yield the final products: H 2 , nitriles, hydrocarbons, etc. 13,14 The first-order overall kinetics of the thermolysis is controlled by the elimina- tion of N 2 and two possible reaction pathways have been traditionally proposed for this process: 14 (a) A synchronous channel where both the elimination of nitrogen and the transposition of one substituent occur simultaneously as a concerted process, (b) An asynchronous channel, which occurs in two steps, in such a way that the azide first decomposes into N 2 and a singlet nitrene radical, that then transforms into an imine by transposition of one of the substituents. For years, it was generally accepted that the intermediate singlet nitrene existed, characterized by an extremely short lifetime, that rapidly suffered a Curtius rearrangement to yield the imine. When the C s symmetry is considered in ab initio calculations, 15-17 the open-shell 1 A′′ of methylnitrene, which is a component of the Jahn-Teller splitting from 1 E state, is a local minimum with a high activation energy for rearrangement. Nevertheless, this process is irrelevant 16 in comparison with the rearrangement of the other Jahn-Teller component, singlet 1 A′ methylnitrene, that is predicted to be a barrier-free process by several ab initio results. 15-19 Therefore, it has been proposed that singlet methylnitrene is not a genuine minimum on its potential energy surface, and this could explain why this species has never been detected in experiments neither of direct photolysis nor pyrolysis of methyl azide. Nevertheless, this radical has been experi- mentally observed by photodetachment spectroscopy of the CH 3 N - anion. 20 A singlet potential energy surface for the elimination of N 2 in methyl azide was obtained by Bock and Dammel 14 at MNDO level, and their results show that both the synchronous and the asynchronous channels compete since the activation energies at those level of theory are similar. However, these authors justify a preference for the synchronous path because the temperature of pyrolysis presents a substituent dependence. Recently, the geometry of the transition state involved in such a process has been localized at MP2/6-31G* level of theory by Nguyen et al. 21 and this channel has also been proposed for the direct photolysis of methyl azide, 22 although no definite experimental evidence exists up to date indicating which channel constitutes the reaction path. 1146 J. Phys. Chem. A 1998, 102, 1146-1151 S1089-5639(97)02500-0 CCC: $15.00 © 1998 American Chemical Society Published on Web 01/27/1998