Experimental and ab initio study of the HO2.CH3OH complex: thermodynamics and kinetics of formation

Near-infrared spectroscopy was used to monitor HO2 formed by pulsed laser photolysis of Cl2-O2-CH3OH-N2 mixtures. On the microsecond time scale, [HO2] exhibited a time dependence consistent with a mechanism in which [HO2] approached equilibrium via HO2 + HO2.CH3OH (3, -3). The equilibrium constant for reaction 3, K(p), was measured between 231 and 261 K at 50 and 100 Torr, leading to standard reaction enthalpy and entropy values (1 sigma) of delta(r) = -37.4 +/- 4.8 kJ mol(-1) and delta(r) = -100 +/- 19 J mol(-1) K(-1). The effective bimolecular rate constant, k3, for formation of the HO2.CH3OH complex is .10(-15).exp[(1800 +/- 500)/T] cm3 molecule(-1) s(-1) at 100 Torr (1 sigma). Ab initio calculations of the optimized structure and energetics of the HO2.CH3OH complex were performed at the CCSD(T)/6-311++G(3df,3pd)//MP2(full)/6-311++G(2df,2pd) level. The complex was found to have a strong hydrogen bond (D(e) = 43.9 kJ mol(-1)) with the hydrogen in HO2 binding to the oxygen in CH3OH. The calculated enthalpy for association is delta(r) = -36.8 kJ mol(-1). The potentials for the torsion about the O2-H bond and for the hydrogen-bond stretch were computed and 1D vibrational levels determined. After explicitly accounting for these degrees of freedom, the calculated Third Law entropy of association is delta(r) = -106 J mol(-1) K(-1). Both the calculated enthalpy and entropy of association are in reasonably good agreement with experiment. When combined with results from our previous study (Christensen et al. Geophys. Res. Lett. 2002, 29; doi:10.1029/2001GL014525), the rate coefficient for the reaction of HO2 with the complex, HO2 + HO2.CH3OH, is determined to be (2.1 +/- 0.7) x 10(-11) cm3 molecule(-1) s(-1). The results of the present work argue for a reinterpretation of the recent measurement of the HO2 self-reaction rate constant by Stone and Rowley (Phys. Chem. Chem. Phys. 2005, 7, 2156). Significant complex concentrations are present at the high methanol concentrations used in that work and lead to a nonlinear methanol dependence of the apparent rate constant. This nonlinearity introduces substantial uncertainty in the extrapolation to zero methanol.