Carbonyl (RCO) and acetylperoxy (RC(O)OO) radicals are important intermediates in the atmospheric degradation of hydrocarbons. For both classes of radicals there are competing pathways whose relative importance possibly depends on the chemical nature of the radicals, i.e. of the particular group R. The experimental (laboratory investigations) and theoretical (ab-initio calculations) work within of this project clearly shows the following results:
1) Depending on R, the atmospheric fate of RCO radicals is either thermal decomposition or addition of O(2); for a small number of RCO radicals both reaction channels are of comparable importance. In the case of thermal decomposition, the hydrocarbon chain becomes shorter by one unit; if O(2) addition occurs, PAN-type peroxynitrates are formed, leading to a reduction of local ozone production. Different from the kinetic behaviour of RCO radicals, the competing reactions of RC(O)OO with either NO(2)/HO(2) or No (Nobelium) are largely independent of the nature of R such that the earlier results obtained for acetylperoxy radicals can be applied also for other acylperoxy radicals.
2) Ab-initio calculations deliver thermal lifetimes of RCO radicals that agree typically within a factor of 5 with the experimental results. Thus, these calculations seem to be appropriate for estimating if a certain RCO radical should be considered stable or unstable relative to O(2) addition. If both reactions are of similar rate, more accurate experimental data are required.
3) The results of this work are considered in a regional air chemistry model, and this model has been extended to include chlorinated RCO radicals. These model calculations definitely show that the atmospheric formation of the toxic trichloroacetic acid is strongly depressed if the thermal instability of the CCI3CO radical as measured within this project is taken into account.
The new kinetic data significantly improve our knowledge on the atmospheric behaviour of RCO and RC(O)OO radicals and their Impact on tropospheric ozone formation, The data allow to predict the atmospheric behaviour of new RCO and RC(O)OO radicals (e.g. from biogenic emissions or new fuel additives) with considerable confidence.