Experimental and theoretical study of atmospheric free radical reactions.

Free radicals play essential roles in atmospheric chemical and radiative processes, thus impacting composition, dynamics and energy budgets of atmosphere. In this work, we have carried out experiments and calculation on three important atmospheric free radical reactions. First, we have applied the laser flash photolysis long-path absorption method to measure the key aqueous halogen free radical reaction rate constants and equilibrium constants: X + X- ↔ X2 - (X = Br, I). This may be part of a route to halogen activation in which halide from sea salt aerosols is converted to photochemically reactive species. Simple reaction systems were designed to eliminate sources of error. The rate constants exhibit strong temperature dependence over the temperature range from 10 °C to 50 °C, whereas, the equilibrium constants do not. The slow reverse reactions strongly depend on ionic strength, but the fast forward reactions do not, probably because they are controlled by diffusion. A detailed error analysis and a comparison with literature data were carried out. The present results have significantly improved error limits and range of validity. Second, we have used quasiclassical trajectory calculations (QTC) with an analytic potential energy surface based on ab initio calculations to investigate the dynamics of the gaseous OH + NO2 reaction. This reaction to form HONO2 has received much attention, but extrapolation of falloff curves towards the high pressure limiting rate constant kinfinity is still uncertain in both magnitude and temperature dependence. Due to the difficulty in measuring kinfinity directly, it has been suggested that the vibrational deactivation rate constant kv can serve as a proxy for kinfinity when vibrational deactivation occurs via formation of a strongly coupled complex. A method was developed for monitoring intramolecular vibrational energy redistribution (IVR) in the trajectories. The calculated IVR time constant is in excellent agreement with experiment. The calculated kv and kinfinity agree with each other, supporting the use of the proxy method, although the energy distributions are not statistical. Studies using a weaker O-N bond show that the proxy method may fail for intermediates with weaker bond energies, thus the proxy method should be tested on a case-by-case basis.Ph.D.Physical chemistryPhysics, Atmospheric SciencePure SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/125151/2/3186695.pd