EFFECTS OF SOLVATION ON THE ENTHALPIES OF REACTION OF TERT-BUTOXYL RADICALS WITH PHENOL AND ON THE CALCULATED O-H BOND STRENGTH IN PHENOL

The enthalpy for the reaction of di-tert-butyl peroxide with phenol to give tert-butyl alcohol and phenoxyl radical (i.e. t-BuOOBu-t + 2 PhOH --> 2 t-BuOH + 2 PhO(.)) has been determined in a number of solvents using photoacoustic calorimetry. The effect of the solvent on the thermochemistry of this process is remarkably large with the reaction being ca. 10 kcal mol(-1) more exothermic in acetonitrile or ethyl acetate than in isooctane. The relationship between the observed enthalpy changes and the PhO-H bond energy is discussed in detail. It is shown that in order to extract the bond energy from the experimental (apparent) enthalpy change, it is necessary to account for a number of processes, viz., the volume change for the overall reaction, the solvent effect associated with the conversion of 1 mol of di-tert-butyl peroxide to 2 mol of tert-butyl alcohol, and the differences in solvation of phenol and the phenoxyl radical. The contributions from each of these processes to the observed reaction enthalpy were derived from a separate set of experiments with 1,4-cyclohexadiene instead of phenol or from data available in the literature. These data allow one to determine solution bond energies, i.e., the enthalpy of homolysis for which the standard state is the solvated reactant and products, and to quantify the solvent effect on these values. Thus, PhO-H bond energies in isooctane (88 kcal mol(-1)), benzene (89 kcal mol(-1)), carbon tetrachloride (90 kcal mol(-1)), ethyl acetate (95 kcal mol(-1)), and acetonitrile (95 kcal mol(-1)) have been obtained. Most of the differences between these values can be accounted for from the known hydrogen bonding equilibrium between the solvents and the phenol. A number of purported determinations of the PhO-H ''gas-phase bond energy'' which utilized electrochemical (EC) measurements and, of necessity, highly polar solvents are shown to be seriously in error. Similar errors must be present in many other EC ''gas-phase'' bond energies which also were determined in polar solvents.