ULTRAVIOLET-ABSORPTION SPECTRUM AND SELF-REACTION OF CYCLOPENTYLPEROXY RADICALS

The kinetics and mechanism of the self-reaction of cyclopentylperoxy radicals: 2 c-C5H9O2 --> 2 c-C5H9O + O2 (1a) --> c-C5H9OH + c-C5H8O + O2 (1b) have been studied using both time-resolved and end-product-analysis techniques. Determination of the product yields from the photolysis of Cl2-c-C5H10-O2-N2 mixtures using FTIR spectroscopy demonstrates that ring-opening of the cyclopentoxy radical formed in channel (1a): c-C5H9O + M --> CH2(CH2)3CHO + M (3) dominates over reaction with oxygen: c-C5H9O + O2 --> c-C5H8O + HO2 (2) under atmospheric conditions. Flash photolysis-UV absorption experiments were used to obtain the UV spectrum of the cyclopentylperoxy radical and the kinetics of reaction (1). The spectrum of c-C5H9O2 is similar to those of other alkylperoxy radicals, with a maximum cross-section of (5.22 +/- 0.20) x 10(-18) cm2 molecule-1 at 250 nm, measured relative to a value of 4.55 x 10(-18) cm2 molecule-1 for CH3O2 at 240 nm. The observed second-order rate constant, k(obs) (-d[c-C5H9O2]/dt = 2k(obs)[c-C5H9O2]-2), for removal of cyclopentylperoxy radicals was dependent on the oxygen partial pressure. Experiments as a function of temperature from 243 to 373 K gave limiting minimum and maximum values of k(obs) at low (< 1 Torr) and high (> 50 Torr) oxygen partial pressures, respectively: k(min)/cm3 molecule-1 s-1 = (1.3 +/- 0.4) x 10(-14) exp[(188 +/- 83)K/T] and k(max)/cm3 molecule-1 s-1 = (2.9 +/- 0.8) x 10(-13) exp[-(555 +/- 77)K/T]. At low oxygen partial pressures, the only effective removal channel for cyclopentylperoxy radicals is the molecular channel (1b) and k(min) can be equated to k1b. Simulations suggest that k(max) represents an upper limit on k1 and is at most 25% greater. In light of the present results on the cyclopentylperoxy radical, further experiments were performed on the cyclohexylperoxy radical self-reaction: 2 c-C6H11O2 --> 2 c-C6H11O + O2 (16a) --> c-C6H11OH + c-C6H10O + O2 (16b) at low oxygen partial pressures, giving k16b/cm3 molecule-1 s-1 = (1.3 +/- 0.3) x 10(-14) exp[(185 +/- 15)K/T] and an estimated k16/cm3 molecule-1 s-1 = 7.7 x 10(-14) exp(-184 K/T). The above errors are 1-sigma and represent experimental uncertainty only.