Experimental rate parameters for thirty-six reactions between 17 alkenes and five peroxyl radicals are compiled in table 1 (the quoted error bounds are standard errors). All except that of Arsentiev and Mantashyan,[17] have been determined by a competitive technique, using a previously determined rate constant for the reference reaction. The epoxidation rate constants for methylperoxyl and t-butylperoxyl radicals have been reevaluated to take into account more recent critical evaluations of their reference reactions, the self reactions of the peroxyl radicals.[21]
There have been no further evaluations of the reference reaction for epoxidation by acetylperoxyl radicals, the abstraction of hydrogen from acetaldehyde by acetylperoxyl radicals, since that of McDowell and Sharples[22]. However, the standard errors originally quoted for the epoxidation by acetylperoxyl represent the error in the ratio of kepoxidation/kreference.[6,8,9] The values given here include the significant errors for the reference reaction, giving noticeably larger errors for the Arrhenius parameters for epoxidation than previously quoted.
Baldwin et al.[14] have reevaluated their rates for epoxidation by hydroperoxyl radicals using a recent recommendation for the reference reaction, the self reaction of the radical[23]. These rates are quoted here. Of their two measurements of the activation energy for the epoxidation of ethene,[14,16] only the more recent is used in the analysis in this paper.
Rate constants for 14 epoxidation reactions by acetyl and alkyl peroxyl radicals were determined at only one temperature. To allow these results to be included in figures 1, 2, 3, 4, 5 ) , activation energies were calculated using an assumed A factor of log10(A/dm3 mol-1 s-1) = 8.1±0.5, which is the average of the 13 A factors that have been determined (the quoted error is the standard deviation of the measured A factors, and is also equal to the average of the measured A factor standard errors).
The A factors for epoxidation by hydroperoxyl radicals are statistically significantly higher than for epoxidation by alkyl or acylperoxyl radicals, with an average of log10(A/dm3 mol-1 s-1) = 8.83±0.36 (the average of the measured standard errors of the A factors is ± 0.34). Stothard and Walker have noted that the A factors for epoxidation by hydroperoxyl radicals tend to increase with increasing alkene ionisation energy,[13] with a recent evaluation giving log10(A/dm3 mol-1 s-1) = 0.763 Ialkene(eV) + 1.290.[24] However, the variation is due predominantly to the relatively high value for ethene of log10(A/dm3 mol-1 s-1) = 9.35±0.35, which is x8 higher than the average for the other hydroperoxyl epoxidation reactions.
The rate constants for the addition of hydroperoxyl radicals to ethene and 2-methylpropene from the discharge flow experiments of Avramenko et al.[25] have not been included in this table 1 . In their experiments, HO2 was produced via H atom (from discharged H2) addition to oxygen and the review of Lloyd[26] suggests that there was significant contamination by oxygen atoms and hydroxyl radicals, both of which would react with the alkenes to give the main observed product, formaldehyde. Lloyd's review also discounts other early work on hydroperoxyl addition to ethene as unreliable due to the rate data being derived from a poorly understood, complex reaction mechanism describing the co-oxidation of methane and ethene.[27]
Arsentiev et al.[17,28] monitored the total peroxyl radical concentration in the gas phase by ESR during the autoxidation of ethene and propene. The rates of production of the epoxides (determined by GC analysis) were found to correlate well with the product of the peroxyl radical and alkene concentrations and were used to derive (effectively, species averaged) rate constants for the epoxidation of the alkene by the peroxyl radicals present. For ethene autoxidation at the temperatures used (637-688 K), the dominant peroxyl radical present is likely to be the hydroperoxyl radical, so the rate constant of Arsentiev et al.[17] is included in table 1 as a value for epoxidation of ethene by hydroperoxyl radicals. The epoxide formed during the autoxidation of propene at 633K[28] is likely to be due to a variety of radicals (eg. methylperoxyl, 1-hydroxy-2-propylperoxyl, hydroxymethylperoxyl and allylperoxyl,3) so this value is not included.