An analysis of the atmospheric observations of OZ('Ag) in the dayglow and twilight confirms the crude experimental assessments of Brown [ 1 ] and the conclusions that can be made from recent work of Westenberg, Roscoe, and DeHaas [2] that the reaction H -t Oz('Ag) --+ OH + 0 is relatively slow, much slower than was expected or can be explained easily in theoretical terms. Using a value for the concentration of atomic hydrogen at 85 km, an upper limit rate of 3 ~1 0 -l ~ cm3 molecule-' sec-I would be compatible with current atmospheric models. An evaluation of the available data for the rates of several reactions involving ground state or electronically excited species, for which the values are reasonably well established, is included to better analyze the general effects of an electronically excited reactant. This further illustrates the unusual slowness of the H + O,(lAg) reaction.
Brown [ l ] has inferred an upper limit of 6 . 5 ~1 0 -' ~ cm3 molecule-' sec-' at room temperature for the reaction (1) H + 02('Ag) = OH + 0 from a crude assessment of some earlier discharge flow experiments which studied the effects of water on the production of atomic oxygen by microwave discharge of 02; H and 02('Ag) were detected by electron-paramagnetic resonance. The measured concentrations showed little if any evidence for reaction (1) at these conditions. This appears to be further supported by recent observations of Westenberg, Roscoe, and DeHaas [2]. These workers mixed a stream of H atoms with a stream of discharged 0 2 free of 0 atoms. With undischarged 0 2 they observed H decay from the reaction H + 0 2 + M + HO2 + M followed by reactions of H with HO2. When the 0 2 was discharged, they found the H decay reduced by just that amount that the 0 2 was decreased by conversion to 02('Ag). They state that this result does not prove that reaction (1) is slow because, were it fast, the reaction O H + 0 + H + O2 would rapidly regenerate H. However, they fail to point out that the Ozf'Ag) would also be efficiently quenched so that, rather than observing a decrease in H, the undischarged and discharged cases would appear the same. Currently the estimate of Brown is the only avail-