Some Experimental and Theoretical Aspects of Combustion Chemistry
Doctoral thesis, 1995
An experimental apparatus is presented and discussed in terms of the range of combustion systems that can be studied using this apparatus. Based on this characterisation the effect of flame stoichiometry on the production of dioxin precursors is discussed. Results suggest control over flame stoichiometry could facilitate control of dioxin production.
The results of laminar flame experiments are often interpreted using species profiles produced using detailed combustion models. Due to the lack of detailed kinetic data for key reactions these mechanisms contain a certain amount of uncertainty. The full experimental characterisation of any given reaction rate coefficient is difficult over the range of temperatures of interest in combustion modelling (300--2500 K). Therefore a theoretical approach to the full characterisation of an elementary reaction is presented.
Ab initio results are presented mapping the Renner-Teller split, ground state potentials associated with the reaction of atomic carbon and nitric oxide (C+NO). A many-body expansion type fit is presented and results of quasiclassical trajectories (QCT) run on this surface are discussed. The results suggest that product branching should favour the thermodynamically most stable products (CO+N, 60%) over the range of temperatures considered. Both product branching and absolute rate coefficients are in as good agreement with experiments (CO+N, 60%). Statistical calculations based on the Rice-Ramsperger-Kassel-Marcus (RRKM) formalism were not, however, in good agreement with these results (CO+N, 73%) and the RRKM complex lifetimes in particular were found to be much shorter than the QCT equivalent. We suggest that this may largely be due to the suppression of the internal equilibria in the RRKM calculations.
Application of modified Arrhenius type fits of the rate coefficients to the modelling of two laminar flames where the importance of the reaction of atomic carbon and NO has previously been established, is presented. The importance of the temperature dependence of rate coefficients in combustion modelling is discussed.
Finally, RRKM results are presented for the reaction of imidogen and nitric oxide (NH+NO). This reaction has recently been characterised by S. Walch and G.C. Schatz in an analogous manner to our characterisation of the C+NO reaction. Statistical calculations suggest that the thermodynamically least stable products (N2O+H) are dominant (85%) over the range of energies studied. The mechanism for this is essentially entropic as the saddle point associated with the decomposition of the intermediate complex (cis-HNNO) into the thermodynamically most stable products is very narrow. A discussion of the effect of the inclusion of anharmonicity on the complex lifetime is also presented.