Use of modeling to design experiments. Doping radicals into complex combustion systems

Numerical modeling can be used to suggest what key experiments should be performed to elucidate complex kinetic systems. In addition, the modeling can be quite useful in the design of such experiments. As an example of such an experimental design we chose a complex combustion system. To elucidate the source of chemiluminescent (A ¹A₂ → X ¹A₁) formaldehyde in the gas phase oxidative decomposition of di-tert-butyl peroxide an incremental concentration of selected precursor radicals can be added into the reactive system. Numerical modeling has suggested the design of experiments to add the radicals in the desired concentrations at the desired location of the reaction zone. Modeling of this system has indicated that elaborate doping probes with their concomittant problems are unnecessary. Further, premixing dimethyl peroxide, azomethane, and methyl hydroperoxide yields suitable sources of methoxy, methylperoxy, and hydroxy radicals, respectively. The inability to directly scale model-calculations is indicated. In addition, the relative (doped to undoped) nature of the calculation ameliorates any uncertainty in the rate constants.