The present state of knowledge of organic, or carbon-based, peroxy radicals (RO2) is reviewed. Data on the chemical and physical properties of peroxy radicals in the gas-phase is considered, as well as the role of peroxy radicals in tropospheric chemistry and measurements of their concentrations in the atmosphere. Where appropriate, peroxy radicals are grouped together by type (alkyl, acyl, oxygen-substituted, halogen-substituted and aromatic radicals) to facilitate comparison. Data on the hydroperoxy radical (HO2) is included where it is directly relevant to measurements on organic peroxy radicals, eg. absorption cross-sections used in measurements of RO2 + HO2 rate constants. The literature data is critically reviewed and recommendations for absorption cross-sections, rate constants and branching ratios are made where considered appropriate. The laboratory experimental techniques which have been used for the generation and detection of peroxy radicals and the products of their reactions are discussed. The structure, spectroscopy and thermochemistry of the radicals are examined. Although the majority of spectroscopic data concerns the u.v. spectra much used for kinetic studies, near-infrared, infrared and electron spin resonance spectra are also considered. In many cases, peroxy radical u.v. spectra are well-fitted by a Gaussian distribution function, enabling the cross-sections to be easily calculated at any wavelength. For the purpose of this review, the chemical reactions of peroxy radicals am divided into reactions with organic peroxy radicals with HO2, with NO and NO2, and finally with other species. Peroxy radical abstraction and addition reactions with closed-shell species are sufficiently slow to be of negligible importance at temperatures pertinent to the atmosphere and are consequently not covered. Data on both the kinetics and mechanisms of peroxy radical reactions are considered. The role of peroxy radicals as intermediates in the atmospheric degradation of volatile organic compounds and in the production of ozone in the troposphere under both low and high [NOx] conditions is discussed. The involvement of peroxy radicals in night-time oxidation chemistry and the oxidation of halocarbons is also indicated. The techniques used for the difficult measurement of peroxy radical concentrations in the atmosphere are described, together with the results to date. Finally, some tentative suggestions as to further avenues of research am made, based on the data reviewed here and with particular reference to the solution of outstanding problems in atmospheric chemistry. Although a great deal of progress has been made in recent years, it is clear that additional work is needed in most areas covered by this review. New, sensitive and selective laboratory techniques are required for studies of peroxy radical kinetics and high level ab initio calculations would help design laser-based detection techniques. Further product studies of photooxidation systems are needed, particularly as a function of temperature. Recent work has shown that the rate constants for RO2 + HO2 reactions used in modelling studies may be too low; if so, these reactions will be correspondingly more important than previously believed in tropospheric oxidation. Recent kinetic studies of the potentially important reactions of methylperoxy radicals with ClO and NO3 need to be confirmed and mechanistic work is necessary. Although substantial progress has been made towards the monitoring of peroxy radical concentrations in the atmosphere, more work is needed, both on measurements and the development of new techniques.
