The rate constant for tert-butyl radical recombination has been measured near 700°K by the very-low-pressure pyrolysis (VLPP) technique and was found to be 108.8*0.3 M-1.sec-1 with neglibible temperature dependence. The thermochemical parameters for tert-butyl radicals were varied within reasonable limits to bring into agreement the data for the decomposition of 2,2,3,3-tetramethyl butane and the recombination of terf-butyl radicals. The revised thermochemistry also makes the gas-phase results and liquid-phase results compatible.

The pyrolysis of cyclobutane has been studied by the VLPP technique between 969 and 1280 K. It was found that the Arrhenius parameters which give best agreement between the present results and those of previous workers are log {k~,, sec-1) = 16.5 -65.5/9. Estimates consistent with the above parameters have been made of the Arrhenius parameters for reactions 2, -2, and 3, c-C4H8 ;=± C4H8• -* 2CZH4 (2, forward; -2, reverse). Studies on the pressure dependence of the rate constant in the 22,400-collision reactor indicate a gas-collision efficiency of about 0.1 relative to the wall, while the good agreement expressed above indicates approximately unit efficiency. By lowering the high-pressure Arrhenius parameters, we can reconcile such a low collision efficiency

The four species in the title were decomposed under VLPP conditions at temperatures in the vicinity of 1100°K. Three model transition states were constructed that fit the low‐pressure data thus obtained and that also yield (1) E298 = ΔE298; (2) E1100 = ΔE1100; (3) log A1100 = 16.4 per C–C bond broken. The predictions of these models as to values of the high‐pressure rate constants for bond scission and the reverse rate constants (radical combination) are compared with existing data.

Absolute values of the rate constant for isopropyl radical combination have been determined over the temperature range 683-808°K. The value log kr (M~1 sec-1) = 9.5 ± 0.2 fits all the data, leading to the conclusion that the temperature dependence, if any, is slight. The compatibility of this value with other values determined either directly or from the reverse process via thermochemical data depends on the particular model chosen to represent the transition state since the temperatures of the various determinations differ. It has been found that