Acetone, a relatively unreactive carbonyl compound in the atmosphere, has been detected in concentrations of ≃500 parts per trillion in surface air over the central Atlantic and several hundred ppt (volume) near the tropopause. This represents a substantially higher mixing ratio than anticipated from previous descriptions of its sources which are primarily much more reactive hydrocarbon emissions and skinks. We assimilate new information about the tropospheric sources and sinks of this substance into a numerical model of its chemistry, cloud transport, and removal. Simulations relevant to the tropical and the mid‐latitude atmosphere are presented. Photolysis provides a rather slow sink, giving a first‐order decay time scale of 40 days (near the surface) to 10 days (at 200 mbar). Attack by HO radicals is of comparable importance, with a time scale of about 20 days (near the surface) to a hundred days (at 200 mbar). A significant amount of carbon, approximately 70 Tg (C) yr−1, appears to be cycled as acetone. Many of the sources of acetone involve the oxidation of higher molecular weight hydrocarbons, but the relative contribution of primary, nonanthropogenic, sources, e.g., biogenic and oceanic, may be significant. Acetone, like CO, is a useful tracer recording previous photochemical activity in an air parcel, a tracer that might be used to describe the photochemical production of compounds like ozone. The mid‐latitude simulations required some adaptation of the Chatfield (1982) transport model. The activity of vertical transport was constrained to reproduce available 222Rn data. Simulations of these data suggest that mid‐latitude vertical transport is less active than tropical, but rather more active than Kz parameterizations suggest.

A large discrepancy exists among recent estimates of the primary quantum efficiencies (ϕ1) of nitrogen dioxide photodecomposition, NO2 + hν → NO + O(3P), in the wavelength range from 374 to 396 nm. To resolve this problem, quantum yields of formation of NO, O2, and NO2 loss have been measured for NO2 vapor at low pressures (0.13–0.30 torr) irradiated at selected wavelengths (334.1–404.3 nm) and temperatures (273–370 K). From these data, estimates of ϕ1 were derived which confirm the previous findings of Jones and Bayes (1973b): ϕ1 increases rapidly from near zero at 424 nm to near unity for excitation at λ < 394 nm. The temperature and wavelength dependences of ϕ1 appear to be in qualitative accord with the simple theory of Pitts et al. (1964) that the energy deficiency for photodissociation of NO2 excited at λ > 397.9 nm (λdiss) is made up in large part from the rotational and vibrational energy of the NO2 molecules. Recommended values for ϕ1 based upon a review of these data and estimates made over a period of 58 years, are given as a function of wavelength.

The photodecomposition of acrolein in dilute mixtures of synthetic air (24-760 Torr) has been studied with excitation at 313 or 334 nm. The decomposition of excited acrolein is very inefficient at high air pressures (0d =* 6.5 X 10~3 at 1 atm, 313 nm) but increases with decreasing pressures (0d =* 8.1 X 10~2 at 26 Torr). The quantum yields of acrolein loss and the observed products, C2H4, CO, C02, CH20, (HCO)2, and CH3OH, are elucidated by the primary processes I-V: CH2=CHCHO(S, or T,) -C2H4 + CO (I); -CH2=CH + HCO (II); -CH3CH(S) + CO (III); -CH3CH(T) + CO (IV); -CH2=CHCO + H (V). New evidence is given for the mechanism of the reactions of the vinyl radical with 02: CH2=CH + 02 -(CH2=CH02) -OCH2CHO; OCH2CHO -* CH20 + HCO (4); OCH2CHO + 02 -* (HCO)2 + H02 (5); the data suggest ks/k4 ^6 x 10'19 cm3 molecule'1 11. From computer simulations of the sequence of reactions describing acrolein decay, it is estimated that, at low air pressures (26 Torr), m + 0IV > 4>i> 0v -nl in 1 atm of air, 0v -> 0ii > 0m + 0iv > 0i-J values for acrolein photodecomposition in the troposphere are derived from the data.

No abstract

Photolyses of methylamine have been carried out at 184.9,147.0 and 123.6 nm. Quantum yields of hydrogen, hydrocarbons, and nitrogen have been measured. Evidence has also been obtained for the formation of HCN and CN in primary processes at the shorter wavelengths. Photolyses of CD3NH2 in the presence of oxygen and NO provide evidence that H and D atom production dominates at 184.9 nm while H2 and D2 elimination occurs to an important extent at 147.0 and 123.6 nm. The primary process of molecular hydrogen elimination is partly terminal and partly nonterminal.