Abstract. Global atmospheric models play a key role in international assessments of the human impact on global climate and air pollution. To increase the accuracy and facilitate comparison of results from such models, it is essential they contain up-to-date chemical mechanisms. To this end, we present an evaluation of the atmospheric chemistry of the four most abundant organic peroxy radicals: CH302, C2H502, CH3C(O)O2, and CH3C(O)CH202. The literature data for the atmospheric reactions of these radicals are evaluated. In addition, the ultraviolet absorption cross sections for the above radicals and for HO2 have been evaluated. The absorption spectra were fitted to an analytical formula, which enabled published spectra to be screened objectively. Published kinetic and product data were reinterpreted, or in some case reanalyzed, using the new cross sections, leading to a self-consistent set of kinetic, mechanistic, and spectroscopic data. Product studies were also evaluated. A set of peroxy radical reaction rate coefficients and products are recommended for use in atmospheric modeling. A three-dimensional global chemical transport model (the Intermediate Model for the Global Evolution of Species, IMAGES) was run using both previously recommended rate coefficients and the current set to highlight the sensitivity of key atmospheric trace species to the peroxy radical chemistry used in the model.

The reaction of the hydroxycyclohexadienyl radical (HO-C 6 H 6 ) (the adduct from the benzene + OH reaction) with O 2 has been investigated using laser flash photolysis with UV-absorption spectroscopic detection, and DFT and ab initio quantum mechanical calculations. An absolute absorption spectrum was measured for the benzene-OH adduct, and its reaction with O 2 , giving a peroxy radical species, was seen to be equilibrated around room-temperature. An equilibrium constant of 1.15 AE 0.6 Â 10 À19 cm 3 molecule À1 was determined at 295 K from an analysis of transient absorption signals using a detailed reaction mechanism. Equilibrium constants were obtained in this way at six different temperatures between 265 and 345 K. The temperaturedependence of these data indicates that the DH 0 298 and DS 0 298 for the title reaction are À10.5 AE 1.3 kcal mol À1 and À33.9 AE 1.4 cal K À1 mol À1 respectively (second-law analysis of the data, 2s errors). A third-law analysis of the data (using a value for DS 0 298 of À38.3 cal K À1 mol À1 , derived from DFT and ab initio calculations) yields a value for DH 0 298 of À11.7 AE 0.2 kcal mol À1 , which compares with an ab initio calculated value of À12.2 kcal mol À1 . Absorption signals at 260-275 nm, in the presence of high concentrations of O 2 , were observed that are consistent with the presence of the benzene-OH peroxy radical, and with stable products of its chemistry. Equilibrium constants obtained from these data agree well with our other determinations. The effective lifetime of the equilibrium system-adduct + O 2 Ð adduct À O 2 -is dictated either by an additional, irreversible reaction of the benzene-OH adduct with O 2 or by a unimolecular transformation of the peroxy species. Assuming the former case, a bimolecular rate constant of around 5.5 AE 3.0 Â 10 À16 cm 3 molecule À1 s À1 was estimated from a kinetic simulation of our decay signals. This rate constant does not appear to vary significantly between 265 and 320 K, but it must be emphasised that it was estimated with a fairly high uncertainty.

Subsequent oxidation yields the radical cation which is unstable in this case and is transformed into Rum(P)(CN')2. Such processes, which involve the transfer of an electron from the metal to the porphyrin ring, have been observed before, but no kinetic data were provided. We were able to measure the rate of the above process as k ~5 X 103 s"1 and to conclude that its rate-determining step is the removal of the carbonyl ligand. Radiolysis of the Ru11 porphyrin in dichloromethane saturated with KOH led to formation of the µ-dimer [(0H)RuIVP]20. Gradual exchange of the two OH groups by Cl' was also followed by radiolysis and the spectrum of the mixed µ-dimer (OH)RuIV-P-0-RuIVP(Cl") is reported.Acknowledgment. This research was supported by the Office of Basic Energy Sciences of the U.S. Department of Energy.Registry No. Run(OEP)(CO), 41636-35-5; Run(TPP)(CO), 32073-84-0; Run(OEP'+)(CO)(Cl), 118513-65-8; Run(TPP1+)(CO)(Cl), 118494-31-8; Run(OEP)(CO)(CN'), 77340-22-8; Rum(OEP)(CN)2, 77340-18-2; (HO)RuIV(OEP), 77089-60-2; Ru"(OEP)(CO)(OH'),

No abstract