Kinetics of Soot Oxidation by Molecular Oxygen in a Premixed Flame

In this study, a two-stage burner was used to determine kinetic parameters of soot oxidation by molecular oxygen. The two-stage burner technique produced soot particles from different fuels in the first burner and then oxidized the particles under either fuel lean or rich conditions in the secondary burner. Methyl decanoate/n-dodecane (surrogate for biodiesel/diesel) and n-butanol/n-dodecane (alcohol/diesel) were studied and compared with previous results using pure m-xylene, pure n-dodecane, and m-xylene/n-dodecane mixtures (surrogate for conventional jet fuel (JP-8)). Particle size distributions, determined using samples taken from the center line of the secondary burner and characterized by a scanning mobility particle sizer, were used to determine the experimental oxidation rates. The evolution of major gas-phase species was measured experimentally by an online GC, and kinetic modeling was used to predict the concentration of OH radicals. The results revealed two regions in the flame: (i) a region close to the burner surface with a high O₂ concentration and (ii) a region where OH was formed and the O₂ concentration dropped. A power-law form of the kinetic rate was fitted to the experimental oxidation rates in the first region of flames where oxidation via O₂ dominated close to the burner surface and OH was not yet formed. The fit yielded an activation energy between 120.3 and 149.3 kJ mol^-1 with an average of 134.8 kJ mol^-1. The average of the pre-exponential factor, A, was found to be 1600 with an order in molecular oxygen of 0.76, resulting in a kinetic rate equation of W_O2 (g cm^-2 s^-1) = 1600 × T^-0.5exp[(-134.8 ± 14.5)/RT](P_O2)^0.76. The method was applied to calculate the oxidation rate of soot over the effective experimental temperature range 1350-1750 K and O₂ partial pressures from 0.065 to 0.123 atm. A comparison of the proposed model with the rates obtained for soot samples derived from the same fuels but oxidized in a thermographimetric analyzer showed that the proposed model can be extended to temperatures as low as 750 K. The reaction rate depended on the fuel source and flame condition, and some rates were over 1 order of magnitude higher than rates obtained from Nagle and Strickland-Constable.