Atmospheric reactions of PAHs fall into two broad categories: (1) heterogeneous processes involving particle-associated compounds (including the semivolatile 4-ring and 5- and 6-ring PAHs), e.g., photolysis/photo-oxidation and gas-particle interactions, and (2) homogeneous gas-phase reactions of volatile 2- and 3-ring and semivolatile 4-ring PAHs initiated by OH (daytime) and N03 radicals (nighttime) and ozone (24 h).
As we have seen, a great deal is known about emission sources and strengths, ambient levels, and mutagenic/carcinogenic properties of the particle-phase PAHs in airborne POM. However, because of the tremendous physical and chemical complexity of the aerosol surfaces on which photolysis, photooxida-tions, and gas-particle interactions take place in "real" polluted ambient air, much less is known about the structures, yields, and absolute rates and mechanisms of formation of PAH and PAC reaction products, especially for the more polar PACs. This is one area in which there exists a major gap in our knowledge of their atmospheric chemistry and toxicology.
In 1954, Kotin and co-workers reported the "carcinogenicity of atmospheric extracts" of Los Angeles air. Subsequently, in 1956, they reported the carcinogenic activity of oxidation products of aliphatic hydrocarbons and, in 1958, of ozonized gasoline. Concurrently, Falk and co-workers (1956) published results of their laboratory studies of photooxidations of PAHs. A decade later, Tebbens et al. (1966) and Thomas and co-workers (1968) reported that major decreases in the concentrations of BaP and perylene occurred in smoke that was irradiated while passing through flow chambers (e.g., a 60% decrease in the BaP content of soot).
Based in part upon this pioneering research, a National Academy of Sciences Committee suggested in their f972 document Particulate Polycyclic Organic Matter that photolysis, photooxidations, and gas-particle interactions might lead to significant degradation of PAHs in ambient particles and formation of products more polar than the parent PAHs (NRC, 1972).
Shortly thereafter, laboratory studies of the photoox-idation and ozonolysis of PAHs deposited on several substrates (e.g., glass surfaces and thin-layer chromatography plates) and exposed to sunlight and/or various levels of ozone in air found significant differences in the reactivities of individual PAHs. Half-lives of BaP and benz[a]anthracene were short (~1 h or less), benzo[e]pyrene (BeP) and pyrene intermediate, and benzofluoranthenes long (Lane and Katz, 1977; Katz et al., 1979). Among the reaction products were three BaP diones that had previously been identified in ambient aerosols collected in Toronto, Canada (Pierce and Katz, 1976), and subsequently at Duisburg, West Germany (König et al., 1983), and Elverum, Norway (Ramdahl, 1983b).
During the same time frame, extracts of ambient aerosols collected throughout the world were found to be directly mutagenic in the Ames bacterial assay. Concurrently, BaP deposited on various substrates (e.g., glass fiber filters) and exposed to ppm and sub-ppm levels of ozone was shown to react readily to form direct-acting frameshift mutagens (Pitts et al., 1978). Furthermore, exposures of particle-associated 4- and 5-ring PAHs to similar levels of N02 (and a required "trace of gaseous HN03") in air resulted in significant yields of nitro-PAHs; these products, which reflected a range of reactivities of the parent compounds, were also direct-acting frameshift mutagens (Jäger, 1978 Jäger and Hanüs, 1980; Pitts et al., 1978; Pitts, 1979 Hughes et al., 1980; Tokiwa et al., 1981; Nielsen et al. 1983; vide infra). For example, an 8-h exposure of BaP deposited on a glass fiber filter to 250 ppb of N02 (the California air quality standard) and traces of HN03 gave an ~20% yield of products, predominantly 6-N02-BaP with lesser amounts of the 1- and 3-isomers. Soon thereafter, mutagenicity studies by Wang and co-workers (f980) implied the presence of nitroarenes in ambient air. Subsequently, several nitroarenes were isolated and positively identified by Ramdahl and coworkers (1982b). We note, however, that these initial researchers cautioned that their rate, mechanism, product, and mutagenic data from laboratory exposures of PAHs deposited in several ways on a variety of substrates (evaporation of a BaP solution on a Hi-Vol filter, vapor deposition of PAHs on synthetic and real particle substrates, etc.) and exposed to relatively high levels of simulated sunlight and 03 and N02 in clean air may not be truly representative of "real world" heterogeneous reaction systems.
We discuss in this section four key aspects of heterogeneous reactions: (1) theoretical and experimental structure and reactivity relationships; (2) field measurements of relative and absolute PAH decay rates in near-source ambient air and during downwind transport; (3) laboratory studies of the photolysis/photo-oxidation and gas-particle interactions with 03 and N02 of key 4- and 6-ring PAHs adsorbed on model substrates or ambient aerosols; and (4) environmental chamber studies of the reactions of such PAHs associated with several physically and chemically different kinds of combustion-generated aerosols (e.g., diesel soot, wood smoke, and coal fly ash). Where such data are available, we also briefly consider some toxicologi-cal ramifications of these reactions.
Space considerations preclude extensive discussions of these broad topics; for details, see reviews and critiques by Nielsen and co-workers (1983), Van Cauwenberghe and Van Vaeck (1983), Pitts (1983), Van Cauwenberghe (1985), Bj0rseth and Olufsen (1983), Valerio et al. (1984), Nikolaou et al. (1984), Finlayson-Pitts and Pitts (1986), Pitts (1987, 1993a, 1993b), Baek et al. (1991), Atkinson and Arey (1994) (gas-phase reactions only), and Arey (1998a).
2. Theoretical and Experimental Structure - Reactivity Relationships
In their thoughtful 1983 review, Nielsen and coworkers noted that particles of diesel soot or wood smoke can absorb significant amounts of water. Thus, they suggested that the most plausible mechanism(s) for nitration (and possibly other electrophilic reactions) of particle-associated PAHs in ambient air may involve reactions both in a liquid film and on solid surfaces and that fundamental laboratory studies of the rates, products, and mechanisms of PAHs in polar solvents would be atmospherically relevant for reactions in the liquid films. Based on this, they proposed a classification scheme for the reactivities of key PAHs in electrophilic reactions, which was subsequently described in detail (Nielsen, 1984).
Given the complexities of the system involved, it is interesting that this reactivity scale has proven to be a useful predictor not only for overall relative decay rates of PAHs associated with aerosols in ambient air but also for the relative rates of their electrophilic heterogeneous reactions in simulated atmospheres.
The original suggestion of Nielsen and co-workers (1983) that these heterogeneous processes involved particle-phase PAHs interacting with other compounds, organic and inorganic, present in "liquid" (or quasiliquid) layers on the surfaces of carbonaceous particles (e.g., wood smoke) is consistent with results from subsequent studies. For example, Strommen and Kamens (1997) cite as evidence for this phenomenon the fact that a viscous, oily slick, with little particle integrity, results when particles of diesel or wood soot impact a metal foil (McDow et al., 1994). Furthermore, Kamens, McDow, and co-workers have demonstrated the strong influence of aerosol water content (humidity), solvent effects, and dissolved aerosol constituents on such processes as the photodegradation of benz[a]anthracene (vide infra; also see Odum et al., 1994a; McDow et al., 1994, 1995, 1996; Jang and McDow, 1997; and Jang and Kamens, 1998). The diffusion model of Odum and co-workers (f994b), which describes the gas-particle partitioning of semivolatile organics (including PAHs), incorporates the idea of diffusion of the SOCs into viscous, liquid-like, organic layers on particles of diesel soot. Subsequently, Strommen and Kamens (1997) extended this concept with a dual-impedance model that incorporated rapid diffusion into an outer layer of liquid-like organic material and slower diffusion into an inner core of discrete solids. We discuss in the following the implications of such liquid-like organic layers to the rates and mechanisms of the photochemical and gas-particle reactions of particle-associated PAHs and nitro-PAHs (e.g., see Fan et al., 1996a, 1996b; and Feilberg and Nielsen, 1999b).
In developing the reactivity scale, Nielsen first investigated the transformation rates of 25 PAHs and four derivatives of anthracene in water-methanol-dioxane solutions, taken as a model of "wet particles," and containing small amounts of dinitrogen tetroxide and nitric and nitrous acids. The measured half-lives and relative rates are shown in Table 10.29. The range of reactivities in solution for PAHs of different structures is remarkable, from 100,000 (arbitrarily set) for an-thanthrene (XXXII) to <0.2 for the least reactive compounds:
Was this article helpful?