The thermal decomposition of di-tert-butyl peroxide has been studied in the presence of carbon dioxide at total pressures from 0.05 to 15 at111 and temperatures from 90-130 "C. The first-order rate .constant for the decomposition is independent of total pressure in this range, with Arrhenius parameters E = 37.8 f 0.3 kcal/mole and log A (s-') = 15.8 + 0.2. A reevaluation of previous data on this reaction leads us to recommend E = 37.78k0.06 kcal/mole and log A (s-') = 15.80 F 0.03 over the temperature range 90-350 "C; extension of this range to higher temperatures using a shock tube would be worthwhile.Canadian Journal of Chemistry, 46, 2721 (1968) It has been postulated by Birss (1) and by Bose and Hinshelwood (2) that in the thermal decomposition of di-tert-butyl peroxide (DTBP), a second channel exists between reactants and products which they called induced decomposition. The evidence upon which these arguments are based is in our view not very strong, and we believe that the postulate is unnecessary. It was found that the rate of decomposition at 160 "C increased with pressure between 50 and 600 mm. However, since the explosion limit of the peroxide decomposition is estimated (3) to be around 600 mm at 150 "C, the probable cause of the rise in rate with pressure becomes apparent. It was also found that added gases, namely SF,, CF,, CO,, CCl,, SiF,, and N,O caused an acceleration of the reaction, whereas acetone, 17-hexane, and N2 were without effect. It was recognized later by Archer and Hinshelwood (4) that in the case of CCI,, the acceleration may have been partly a chemical effect involving chloriile atoms. More recently, Flowers, Batt, and Bellson (5) showed that the peroxide decomposition was strongly catalyzed by HC1 by a mechanism involving chlorine atoms, and further evidence for the existence of chlorine atoms in the peroxide/CCl, reaction is available (6). Also, Batt and Cruickshank (7) have shown that CH,F is formed in the peroxide/SF, decomposition. We suggest that all these observed acceleratioils in the presence of CF,, SiF,, CO,, and N,O are due either to chemical reaction of some kind or to the presence of trace quantities of chlorinated impurities; it is reasonable to suppose that the N,, acetone, and hexane used in the above experiments were free from chlorinated impurities. Our own experiments described below show that carefully purified CO, has no effect on the rate of peroxide decomposition up to a pressure of 15 atm, and we believe that the idea of induced decom~osition should not be included in the current theoretical description of unimolecular reactions.
Materials and Experimental TechniqueCommercial di-tert-butyl peroxide contains about 0.2% of minor impurities including 0.04% acetone (8); these were removed by preparative scale gas chromatography. Carbon dioxide was made from commercial dry-ice: the gas was condensed onto phosphorus pentoxide it1 vacuo, degassed, and then sublimed into a storage vessel; mass-spectrometer analysis failed to reveal any impurity.Each experiment wa...