Contributions of biogenic and anthropogenic hydrocarbons to photochemical smog formation

Photochemical oxidation of biogenic (Isoprene) and anthropogenic (1-octene) hydrocarbons are examined. Experiments studied the individual daylight reactions of both isoprene and 1-octene, including those of OH, O3, and O([superscript 3]P). Results from both the smog chamber experiments and computer kinetic modeling were then used to develop photochemical oxidation mechanisms for each hydrocarbon. Aerosols formed by isoprene and another biogenic, beta-pinene, are characterized. The OH reaction with isoprene is studied. Methyl nitrite photolysis experiments were carried out in an outdoor smog chamber in an attempt to identify as completely as possible OH-isoprene product spectrum. Emphasis was placed on identification and quantification of oxygenated products. The design of a Tenax-based cryo-trap thermal desorber used to trap, concentrate, and dry chamber samples for identification on a GC/MS is described. Analysis of the products revealed that O([superscript 3]P) can form in reaction systems designed to study OH reactions that include high concentrations of NO, and consequently NO2, hence this reaction is also examined. The yields of methacrolein and methyl vinyl ketone are determined as 25±3 and 35.5±4%, respectively, with an additional 5.1±3% as 3-methyl furan, totaling 66±3%. These results, combined with those of previous studies-allow 80% of isoprene's products to be explicitly identified, and the general structure of the remaining products to be ascertained. The O([superscript 3]P) reaction produces 84±8% epoxides, and 8±3% species which result in production of HO2, and subsequently OH. A heretofore unidentified product of the O([superscript 3]P) reaction, 2,2 methyl butenal, is identified. The rate constant of the NO2-isoprene reaction is measured. A series of experiments have been carried out to study the ozone-isoprene reaction in a smog chamber using externally produced O3, added to the hydrocarbon in the dark. A chemical tracer, methyl cyclohexane, was added to probe the OH formation in the system. O([superscript 3]P) formation was also examined using the known distribution of products that are unique to the O([superscript 3]P)-isoprene reaction (part 1). The results provide clear evidence that both OH and O([superscript 3]P) are produced from the O3-isoprene reaction directly in large quantities; about 0.68±0.15 and 0.45±0.15 per O3 -isoprene reaction, respectively. These additional radicals severely complicate the analysis of the O3 reaction, hence computer kinetic modeling was necessary to ascertain the products of the O3 reaction itself. The product spectrum, which differs dramatically from that published previously, is: 67±9% methacrolein, 26±6% methyl vinyl ketone, and 7±3% propene, accounting for 100±10% of the reacted isoprene. Applicability of these results to the gas-phase O3 reaction with other unsaturated hydrocarbons is briefly discussed. The photooxidation chemistry of 1-octene is examined in detail. Formation of OH from the O3 reaction was examined with the use of a tracer/absorber, methyl cyclohexane. The O3 - 1-octene reaction is found to produce, apparently directly, significant quantities of OH, 0.55+0.2 on a per molecule reacted 1-octene basis. Almost 100% of the reacted 1-octene could be accounted for as 80±10% heptanal, 11±6% thermally stabilized Criegee biradical, and about 1% hexane. The OH - 1-octene reaction was found to produce only 15±5% heptanal. The remainder is assumed to result in the formation of alkyl nitrates (32%), and isomerization and eventual formation of multisubstituted products (52%). A separate experiment examining the O([superscript 3]P)-1-octene reaction, showed that 1-octyl oxide accounted for about 80% of the reacted 1-octene. A photochemical model was developed for 1-octene oxidation, and is compared with smog chamber results from NO/NO2-octene experiments. The most crucial factor in the performance of the model is the quantity of assumed alkyl nitrate formation. A mechanism for the oxidation of isoprene is developed and includes the recent developments on each of isoprene's atmospherically important reactions: O3, OH, O([superscript 3]P), and NO3. The mechanism is tested against chamber data that includes a range of mixtures of these reactions. While it performs reasonably well under conditions where the OH and O([superscript 3]P) reactions dominate, it tends to over predict O3 formation, as well as the speed of development of O3 under conditions where the O3 and NO3 reactions are important. The NO3 reaction is the most uncertain aspect of the isoprene mechanism, and may be responsible for a large part of this discrepancy. The discrepancy may also arise from the difficulty in extrapolating the results of O3 experimental results, necessarily carried out in the absence of NOx, to conditions that include significant concentrations of NOx. An extensive set of outdoor smog chamber experiments was carried out to study aerosol formation by two representative biogenic hydrocarbons: isoprene and beta-pinene. The hydrocarbons, at concentrations ranging from a few ppb to a few ppm, were photooxidized in the presence of NOx. Isoprene was found to produce negligible aerosol at ambient conditions, whereas beta-pinene aerosol carbon yields were as high as 8%, depending strongly on the hydrocarbon to NOx ratio. Aerosol samples subjected to infrared absorption spectroscopy revealed that the dominant aerosol products for both isoprene and beta-pinene are organic nitrates, organic acids, as well as other carbonyls and hydroxy compounds. GCMS of the neutral fraction of the beta-pinene aerosol revealed nopinone and several other compounds with molecular weights ranging from 138-200 amu, indicating mainly mono- and dioxygenated products. The average vapor pressure of the 13-pinene aerosol was estimated to be 37 ± 24 ppt at 31 C. Scanning electron micrographs showed that the particles consist of both liquid droplets and agglomerates of small (40-60 nm) solid particles