Formation of Oligomers in Secondary Organic Aerosol - American

Delaware, Newark, Delaware 19716, and Department of. Environmental Sciences and ... North Carolina, Chapel Hill, North Carolina 27599. The formation o...
2 downloads 0 Views 130KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 1428-1434

Formation of Oligomers in Secondary Organic Aerosol MICHAEL P. TOLOCKA,† MYOSEON JANG,‡ JOY M. GINTER,† FREDERICK J. COX,† RICHARD M. KAMENS,‡ AND M U R R A Y V . J O H N S T O N * ,† Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716, and Department of Environmental Sciences and Engineering, University of North Carolina, Chapel Hill, North Carolina 27599

The formation of oligomeric molecules, an important step in secondary organic aerosol production, is reported. Aerosols were produced by the reaction of R-pinene and ozone in the presence of acid seed aerosol and characterized by exact mass measurements and tandem mass spectrometry. Oligomeric products between 200 and 900 u were detected with both electrospray ionization and matrix-assisted laser desorption ionization. The exact masses and dissociation products of these ions were consistent with various combinations of the known primary products of this reaction (“monomers”) with and/or without the expected acid-catalyzed decomposition products of the monomers. Oligomers as large as tetramers were detected. Both aldol condensations and gem-diol reactions are suggested as possible pathways for oligomer formation. Exact mass measurements also revealed reaction products that cannot be explained by simple oligomerization of monomers and monomer decomposition products, suggesting the existence of complex reaction channels. Chemical reactions leading to oligomer formation provide a reasonable answer to a difficult problem associated with secondary organic aerosol production in the atmosphere. It is unlikely that monomers alone play an important role in the formation and growth of nuclei in the atmosphere as their Kelvin vapor pressures are too high for them to significantly partition into the particle phase. Polymerization provides a mechanism by which partitioning to the particle phase becomes favored.

type of activity will change during the chemical transformations that take place through heterogeneous and homogeneous reactions along the aerosol trajectory (9, 10). For the most part, these reactions are not well understood because the chemistry is complex and potentially hundreds of individual organic compounds may be involved. Many volatile organic compounds are released into the atmosphere by plants including terpenes, alcohols, and ketones (11, 12). Of these, terpenes receive the most attention because of their high emission rates and reactivity. SOA is formed when biogenic terpenes or aromatic species from anthropogenic sources react with atmospheric oxidants such as hydroxyl radicals, ozone, and nitrogen oxides (NOx). The emission rates of the terpenoid compounds from plants are estimated to be 10 times greater than nonmethane hydrocarbons from anthropogenic sources (13, 14). Tropospheric ozone levels are thought to be greatly influenced by the reactions of biogenic alkenes (3), and SOA formation may indirectly influence cloud processing (15). For these reasons, much effort has been devoted to understanding the kinetics and mechanism of terpene oxidation and its role in SOA formation (3, 12, 16-26). It is well-known that terpene oxidation leads to the formation of carbonyls and carboxylic acids that become major components of SOA (27, 28). However, as pointed out by Calogirou and co-workers (27), many uncertainties remain. While some primary products of gas-phase oxidation have sufficiently low vapor pressures to partition to the particle phase (23, 24, 29-33), most primary products do not, even though they are found in SOA (33). If this is the case, then how is SOA formed? Recent work has implicated the production of high molecular weight compounds by heterogeneous and multiphase reactions as a driving force for SOA formation and growth (34). For example, formaldehyde may polymerize through gem-diol reactions, acetals may form through intermediates, and ketones may undergo aldol condensations in the presence of acid (35-37). Reactions such as these are consistent with the observation that sulfuric acid aerosol catalyzes the formation of SOA (33, 38, 39). However, the expected oligomeric products have defied detection and characterization probably because the methods typically used to study SOA induce decomposition back to the primary monomers. We present here direct evidence for oligomer formation in SOA produced by the reaction of R-pinene and ozone and show that these products are consistent with an acid-catalyzed polymerization formation mechanism.

Experimental Section Introduction Carbon is one of the main components of fine particulate matter (PM2.5), comprising from 10 to 70% of the fine particle mass (1, 2). The physicochemical properties of organic compounds involved in secondary organic aerosol (SOA) formation directly influence estimates of the impact of SOA on visibility, global climate change, and health effects (3-5). Fine particulate matter has been linked to adverse health effects including premature death and an increased probability of respiratory and circulatory diseases (5, 6). Moreover, many organic compounds in particulate matter exhibit direct mutagenic and carcinogenic activity (7, 8). The amount and * Corresponding author phone: (302)831-8014; fax: (302)8316335; e-mail: [email protected]. † University of Delaware. ‡ University of North Carolina, Chapel Hill. 1428

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

The reaction of R-pinene with ozone was performed in a 0.5 m3 Teflon film bag (40). The experiment was begun by filling the bag with dry clean air from the Pure Air Generator (AADCO 737, Rockville, MD). Ozone was generated photolytically with a Xonics ozone generator. The initial ozone concentration was measured with a photometric ozone detector (Thermo Environmental Instruments, Model 49, Hopkinton, MA), and then acidic inorganic seed aerosol was injected into the bag. Seed aerosol was generated using a commercially available large volume nebulizer (TSI Model 3076 Constant Output Atomizer, St. Paul, MN) to aspirate aqueous salt solution of 0.0035 M (NH4)2SO4 and 0.005 M sulfuric acid into the bag. (Experiments performed in the absence of acid seed contained ammonium sulfate but not sulfuric acid.) Next, R-pinene was injected by volatilizing the liquid in a gently heated manifold with a flowing stream of dry air. Temperature and relative humidity measurements were taken with an 10.1021/es035030r CCC: $27.50

 2004 American Chemical Society Published on Web 01/29/2004

FIGURE 1. ESI-QTOF mass spectrum of an extract of the secondary organic aerosol produced by reaction of r-pinene with ozone in the presence of acid seed aerosol. electronic thermohygrometer (Hanna Instruments, Padova, Italy). The relative humidity was 28-30% after injecting aqueous seed droplets into the bags. The initial R-pinene concentration was 3.4 ppm, the ozone concentration was 0.8 ppm, and the temperature was maintained at 294-295 K. After 1 h of reaction, the aerosol was collected on a 47 mm Teflon impregnated glass fiber filter (type T60A20, Pallflex Products Corp., Putnam, CT). The filter was promptly sealed, packed in dry ice, and delivered for analysis by overnight express. After receipt, the filter was split in two. Each half was extracted separately in 1.5 mL of methanol or acetonitrile by vortexing/agitation for 15 min, centrifuged under vacuum to dryness, reconstituted with 0.5 mL of solvent, and analyzed by matrix-assisted laser desorption ionization (MALDI), electrospray ionization (ESI), and desorption ammonia chemical ionization (CI) mass spectrometry. Similar oligomer distributions were observed with MALDI (Bruker Biflex III, Billerica, MA; matrix, 2,5-DHB) and ESI (Micromass API-US QTOF, Beverly, MA) using both solvents. The mass spectra did not change with the length of time that the filter or solvent extract was stored. Desorption CI (VG Autospec, Beverly, MA) using ammonia as the reagent gas was found to induce extensive decomposition to monomeric species. Exact mass measurements were made by the National High Magnetic Field Laboratory at Florida State University with a 9.4 T Fourier transform ion cyclotron resonance (FTICR) mass spectrometer using ESI (41). Ions in the m/z range of 250-2000 were excited and detected. The broadband spectrum was first externally calibrated to assign at least the two peaks of interest within about 2 ppm error. Internal calibration was then performed with these peaks, giving an uncertainty of 0.2 ppm or less, which was sufficient to assign

a unique molecular formula. For this work, the solvent extract from a filter was sealed, packed in dry ice, and delivered for analysis by overnight express. The ESI-QTOF spectrum obtained before shipping matched the ESI-FTICR spectrum obtained after shipping. Six separate experiments were performed, three with acid seed and three without acid seed. A sample blank was obtained for each experiment. The filter containing aerosol from each experiment was divided in two, and each half was extracted and analyzed separately. Similar mass spectra were obtained from different extracts of the same filter and from different experiments of the same type (e.g. with/without acid seed).

Results Figure 1 shows the ESI-QTOF mass spectrum of an extract of the aerosol formed from the reaction of R-pinene and ozone in the presence of sulfuric acid seed aerosol. Ions are observed across the mass range of 200 to greater than 900 m/z. Adjacent ions are separated by 1.0 u, suggesting that no multiply charged ions are present. Most, if not all, of the peaks observed are identified as MNa+ (molecule plus a sodium ion), as ion exchange with potassium (performed by adding KCl to the extract solution to give MK+ in the mass spectrum) shifts each nominal mass by 16 u, the difference between Na+ and K+. These assignments are also confirmed by exact measurements with ESI-FTICR (see data below). It is not possible to observe MH+ ions under the conditions used in this experiment, as even trace levels of sodium contamination are sufficient for strong binding to the analyte molecules. The broad groupings of singly charged ions in Figure 1 are characteristic of a low-molecular weight polydisperse copolymer system (42-46). It is known that the VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1429

primary products from the reaction of R-pinene and ozone are pinonaldehyde (MW ) 168), norpinonic acid (MW ) 170), 10-hydroxylpinonaldehyde (MW ) 184), and norpinonaldehyde (MW)154), along with carboxylic acids such as pinonic acid (MW ) 184), pinic acid (MW ) 186), and norpinic acid (MW ) 172) (27, 28, 47-49). These products are the “monomers” from which oligomeric molecules may form. The broad groupings are separated by about 180 u, which roughly corresponds to the median mass of a monomer unit. Within each grouping are smaller clusters of peaks. The clusters are separated by 14-18 u, and the prominent peaks within each cluster are separated by 2 u. These groupings and clusters are consistent with numerous combinations of monomer units with multiple sites of oxidation (alcohols, ethers, carbonyls, and acids). While the median mass of each grouping does correspond to a whole number of monomers, monomer decomposition may accompany oligomer formation as evidenced by the broad distribution of oligomers. Note that the ion intensities in Figure 1 may not be directly related to the concentration of the products. Oligomers were detected in aerosols produced without acid seed, but the signal intensity was 10 times lower than in the acid-catalyzed experiments. Sample blanks were also analyzed for each experiment; in each case, no oligomer peaks were observed. Scheme 1 illustrates reactions that could lead to the oligomers observed in Figure 1. Parts A and B of Scheme 1 show representative building blocks of the oligomers, which may include both monomers (the primary products of ozonolysis, some of which are given in Scheme 1A) and acidcatalyzed ring cleavage products (50-52) of the monomers (examples shown in Scheme 1B). Parts C and D of Scheme 1 show polymerization reactions induced by acid-catalyzed aldol condensation and gem-diol formation respectively using the dimerization of pinonaldehyde as an example. Note that the gem-diol reaction in Scheme 1D includes a dehydration step, which yields a product having the same molecular formula as that in Scheme 1C. If dehydration does not occur, then the hemiacetal product will have a mass 18 u higher than the aldol condensation product. Many other combinations of monomers and monomer ring cleavage products are possible, leading to dimers, trimers, and so forth. Tandem mass spectrometry (performed with the ESIQTOF) provides evidence for reactions such as those in Scheme 1C,D. In tandem mass spectrometry, a specific precursor ion from the spectrum in Figure 1 is selected, and the product ions obtained by collision-induced dissociation of that precursor are recorded. In general, the product ion signal intensities from these precursors were weak and high collision energies were required to achieve detectable signals, which is not surprising for sodiated precursors (MNa+). Figure 2 shows the product ion spectrum of the 359 m/z precursor ion, whose m/z corresponds to MNa+ of the pinonaldehyde dimer, structure 9. The ESI-FTICR spectrum of this precursor m/z shows two peaks of approximately equal intensity. The two peaks have different exact masses that correspond to distinct molecular formulas. The first, C20H32NaO4 (359.219 83 expected, 359.219 86 measured), corresponds to the exact mass of structure 9. While structure 9 is the only combination of two monomers that gives a product at this exact mass, five other combinations that involve three total species, one monomer and two decomposition products, can lead to a product at this exact mass. The second peak in the ESI-FTICR spectrum corresponds to a formula of C19H28NaO5 (359.183 44 expected, 359.183 49 measured), which cannot be formed by direct coupling of the common monomers and monomer decomposition products and suggests a more complex reaction sequence. Since it is not possible to obtain a product ion spectrum of these peaks individually even with ESI FTICR, 1430

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

the product ion spectrum is a convolution of precursor ions having both exact masses. The product ion spectrum of 359 m/z, shown in Figure 2, contains many ions that would be expected from structure 9. Because of structural similarities, most of these ions would also be expected from the other combinations of monomers and decomposition products at this exact mass. Product ions consistent with structure 9 include 191 m/z (dissociation to the pinonaldehyde monomer); 341 m/z (dehydration of the molecular ion, most likely from the alcohol group); 289 and 261 m/z (rupture of the cyclobutane moieties); and 316, 225, 207, and 177 m/z (R cleavage to a carbonyl group). Note that the bond cleavages shown in the inset of Figure 2 do not include hydrogen rearrangements and/or dehydrogenation reactions that accompany many of these fragmentations. The possibility of dimerization through the gem-diol route (structure 11, Scheme 1D) should be considered, even though this structure may be unstable in acidic media such as that found in the aerosol. Structure 11 is very similar to structure 9 and equally consistent with the product ion spectrum in Figure 2. The consistency of Figure 2 with a pinonaldehyde dimer was checked by reacting pinonaldehyde with sodium hydroxide in tetrahydrofuran (THF) as described previously (53). (Dimerization is catalyzed by addition of either acid or base; base catalysis is more robust for organic synthesis in the laboratory.) The ESI-QTOF mass spectrum of this sample included numerous ions between 300 and 700 m/z. The product ion spectrum of the 359 m/z precursor from this sample contained several of the ions found in Figure 2, most notably 316, 313, and 301 m/z plus a series of ion groupings between 180 and 300 m/z separated by 14-16 m/z (spectrum provided as Supporting Information, Figure S1). It should be noted that the composition (product distribution, absolute concentration) of this sample solution and the solution used to obtain Figure 2 were not identical, which could easily influence the internal energy of the precursor ions and hence the relative intensities of ions in the product spectra. Figure 3 shows the product ion spectrum of the 347 m/z precursor. The ESI-FTICR spectrum of this precursor m/z shows five separate peaks, each corresponding to a unique molecular formula. However, one of these dominates the rest, having a signal intensity greater than 25 times the intensities of the remaining four peaks: C18H28NaO5 (347.183 44 expected, 347.183 39 measured). This formula corresponds to a dimer, structure 12 in Figure 3, which is formed by coupling monomers 1 and 3. While structure 12 is the only combination of two monomers that gives a product at this exact mass, seven other combinations that involve three total species, one monomer and two decomposition products, can lead to a product at this exact mass. Product ions in Figure 3 that are consistent with structure 12 include 329 m/z (dehydration of the alcohol); 303 and 207 m/z (R-cleavage to the carbonyl); 301 m/z (COOH loss); 289 m/z (both dehydration and R-cleavage); 277, 275, and 249 m/z (rupture of a cyclobutane moiety); and 193 and 177 m/z (dissociation to the monomers). Figure 4 shows the product ion spectrum of the 361 m/z precursor, which corresponds to structures 13 and 14, dimers of 2 + 3 and 1 + 4, respectively. The ESI-FTICR spectrum of this precursor m/z shows two peaks. The first corresponds to C19H30NaO5 (361.199 09 expected, 361.198 97 measured), which is the chemical formula of structures 13 and 14. It should be noted two other structures, both resulting from the combination of one monomer and two decomposition products, also have this formula. The second species is four times less intense than the first and corresponds to C20H34NaO4 (361.235 48 expected, 361.235 41 measured). This species cannot be formed directly from monomers and decomposition products, although it does correspond to the

SCHEME 1 a

a (A) Representative monomer products of the reaction of R-pinene with ozone. (B) Possible decomposition products of the monomers using pinonaldehyde as an example. (C) Dimerization of pinonaldehyde via aldol condensation. (D) Dimerization of pinonaldehyde by gem-diol formation with subsequent dehydration.

alcohol rather than ketone form of the product 9. Product ions in Figure 4 that are consistent with structures 13 and 14 include 343 m/z (dehydration of the alcohol); 329 m/z (methanol loss from 14); 317 and 315 m/z (loss of both ketone

and carboxylic acid from 13); 301 m/z (R-cleavage to the carbonyl on the hydroxyl side of 14); 247 m/z (rupture of a cyclobutane moiety from 14); and 207, 193, 191, and 177 (dissociation to the monomers). VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1431

FIGURE 2. Product ion spectrum of the 359 m/z precursor. Hydrogen rearrangements and/or dehydrogenation reactions that accompany many of these fragmentations are not shown in the inset.

FIGURE 3. Product ion spectrum of the 347 m/z precursor. Hydrogen rearrangements and/or dehydrogenation reactions that accompany many of these fragmentations are not shown in the inset.

FIGURE 5. Product ion spectrum of the 375 m/z precursor. Hydrogen rearrangements and/or dehydrogenation reactions that accompany many of these fragmentations are not shown in the inset. 375.214 69 measured), which is consistent with structure 15. While structure 15 is the only combination of two monomers that gives a product at this exact mass, nine other combinations that involve three total species, one monomer and two decomposition products, can lead to a product at this exact mass. The second peak corresponds to C19H28NaO6 (375.178 36 expected, 375.178 35 measured), which does not represent a simple combination of monomers and/or decomposition products. Product ions in Figure 5 that are consistent with structure 15 include 357 m/z (dehydration), 343 m/z (loss of methanol), 261 m/z (rupture of the cyclobutane moiety), 223 (R-cleavage to a carbonyl), and 207 and 191 m/z (dissociation to the monomer). It should be noted that the ammonia chemical ionization mass spectrum did not resemble Figure 1. Almost all of the m/z values observed in this spectrum corresponded to the monomers and their decomposition products. It is likely that oligomers observed with ESI or MALDI decompose with conventional electron and chemical ionization, explaining why they have not been previously reported.

Discussion

FIGURE 4. Product ion spectrum of the 361 m/z precursor. Hydrogen rearrangements and/or dehydrogenation reactions that accompany many of these fragmentations are not shown in the inset. Figure 5 shows the product ion spectrum of the 375 m/z precursor. This precursor m/z corresponds to structure 15, a dimer of 2 and 4. Two peaks with equal intensities were observed in the ESI-FTICR spectrum of this precursor m/z. The first corresponds to C20H32NaO5 (375.214 74 expected, 1432

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

Figures 2-5 are typical of the product ion spectra of the precursors in Figure 1. The product ion spectra of different precursors are quite similar, containing many ions of the same m/z. These results clearly show that secondary organic aerosol produced by R-pinene ozonolysis in the presence of an acid catalyst is strongly influenced by oligomerization reactions of the primary ozonolysis products, most likely by aldol condensation and/or gem-diol formation (Scheme 1C,D). Exact mass measurements of the precursors in each case confirm the presence of an ion having the expected molecular formula for direct oligomerization of the monomers (primary products). The product ion spectra confirm the presence of monomer building blocks in two ways: product ions corresponding to the specific monomers involved in oligomer formation and product ions corresponding to rupture of the cyclobutane moiety within the monomers involved in oligomer formation. While this study has focused on dimers, Figure 1 shows that reactions of this type give products at least to the tetramer length. It is also clear that the chemistry of secondary organic aerosol formation is not restricted to simple oligomerization of the primary ozonolysis products. Calculation of the combinations of either one or two monomers (primary products) with either two or one monomer decomposition products (cyclobutane rupture, Scheme 1B) gives an m/z distribution that closely matches the experimental distribu-

tion in Figure 1 (distribution is given in Supporting Information, Figure S2). The m/z values corresponding to the direct coupling of monomers are generally more intense in Figure 1 than those which can only be formed by a combination of monomers and monomer decomposition products. In a similar manner, the expected m/z values for various combinations of monomers to yield trimers and monomers plus monomer decomposition products to yield tetramers closely match the ions observed in Figure 1 between 520 and 650 m/z. Exact mass measurements reveal even greater complexity in secondary aerosol formation, since they indicate the presence of species that cannot be rationalized on the basis of the direct oligomerization of monomers and/or monomer decomposition products. Therefore, chemical routes in addition to Scheme 1C,D must be operative. Oligomerization may be prevalent in bag experiments such as those studied here simply because of the high aerosol loading inherent to the experiment. Are reactions such as these likely to occur in ambient air? To investigate this possibility, samples were taken over a 24 h period in Research Triangle Park, NC with a high-volume sampler (54). The aerosol in Research Triangle Park is expected to be strongly influenced by secondary organic aerosol (1). The samples were analyzed in the same fashion as the bag samples described above. The ESI mass spectra showed a complex series of ions over the m/z range of interest (spectrum given in Supporting Information, Figure S3). Some of the ions in Figure 1 are observed in the ambient spectrum (such as 359 m/z), but others are not (such as 347, 361, and 375 m/z) perhaps because other monomer products of R-pinene oxidation are less abundant than pinonaldehyde. The product ion spectrum of the 359 m/z precursor revealed fragments similar to those in Figure 2 (spectrum given in Supporting Information, Figure S4). The implications of these observations are significant when viewed in the context of other recent investigations of ambient aerosol (55-58). Once formed, particles containing oligomeric molecules will likely have absorptive and scattering properties that affect the global radiation balance and have yet to be included in current modeling efforts. The hygroscopicity of these compounds are unknown; the ability of particles containing these oligomers to act as cloud condensation nuclei is likely to be modified (55, 56). Oligomers may also represent a significant portion of the unaccounted organic mass in ambient aerosol measurements (57). Moreover, biological systems exposed to particles containing oligomeric molecules may have unforeseen negative consequences (58). For these reasons, additional experiments are needed to unravel the details of oligomer formation in secondary organic aerosols including intermediates, products, and reaction kinetics.

Acknowledgments John Dykins provided much of the ESI-QTOF data. Geoff Klein, Ryan Rodgers, Mark Emmett, and Alan Marshall at the Florida State University National High Magnetic Field Laboratory obtained and analyzed the ESI-FTICR data. John Volckens and David Olsen at the U.S. Environmental Protection Agency provided ambient aerosol samples from Research Triangle Park, NC. Matthew Dreyfus performed the pinonaldehyde oligomerization experiment. Terrence Schull and Douglass Taber provided insight into organic reaction mechanisms. This material is based on work supported the National Science Foundation under Grants CHE-0098831 (M.V.J.), ATM-0314128 (R.M.K.), and CHE-9909502 (NHMFL).

Supporting Information Available Four figures. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Tolocka, M. P.; Solomon, P. A.; Mitchell, W.; Gemmill, D.; Wiener, R. W.; Homolya, J.; Natarajan, S.; Vanderpool, R. W. Aerosol Sci. Technol. 2001, 34, 88-96. (2) Watson, J. G.; Zhu, T.; Chow, J. C.; Engelbrecht, J.; Fujita, E. M.; Wilson, W. E. Chemosphere 2002, 49, 1093-1136. (3) Chameides, W. L.; Lindsay, R. W.; Richardson, J.; Kiang, C. S. Science 1988, 241, 1473-1475. (4) Hansen, J. E.; Sato, M. K. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 14778-14783. (5) Laden, F.; Neas, L. M.; Dockery, D. W.; Schwartz, J. Environ. Health Perspect. 2000, 108, 941-947. (6) Pope, C. A.; Burnett, R. T.; Thun, M. J.; Calle, E. E.; Krewski, D.; Ito, K.; Thurston, G. D. JAMA, J. Am. Med. Assoc. 2002, 287, 1132-1141. (7) Lewis, C. W.; Baumgarder, R. E.; Stevens, R. K.; Claxton, L.; Lewtas, J. Environ. Sci. Technol. 1988, 22, 968-971. (8) Hannigan, M. P.; Cass, G. R.; Penman, B. W.; Crespi, C. L.; Lafleur, A. L.; W. F. Busby, J.; Thilly, W. G. Environ. Sci. Technol. 1997, 31, 438-447. (9) Kamens, R. M.; Bell, D.; Dietrich, A.; Perry, J.; Goodman, R.; Claxton, L.; Tejada, S. Environ. Sci. Technol. 1985, 19, 63-69. (10) Sasaki, J.; Arey, J.; Harger, W. P. Environ. Sci. Technol. 1995, 29, 1324-1335. (11) Arey, J.; Crowley, D. E.; Resketo, M.; Lester, J. Atmos. Environ. 1995, 29, 297-2988. (12) Konig, G.; Brunda, M.; Puxbaum, H.; Hewitt, C. N.; Duckham, S. C.; Rudolf, J. Atmos. Environ. 1995, 29, 861-874. (13) Fehsenfeld, F.; et al. Global Biogeochem. Cycles 1992, 6, 389430. (14) Lamb, B.; Gay, D.; Westberg, H.; Pierce, T. Atmos. Environ. 1993, 27A, 1673-1690. (15) Cruz, C. N.; Pandis, S. N. Aerosol Sci. Technol. 1999, 31, 392407. (16) Pandis, S. N.; Wexler, A. S.; Seinfeld, J. H. Atmos. Environ. 1993, 27A, 2403-2416. (17) Stephanou, E. G.; Stratigakis, N. Environ. Sci. Technol. 1993, 27, 1403-1407. (18) Calogirou, A.; Jensen, N. R.; Nielson, C. J.; Kotzias, D.; Hjorth, J. Environ. Sci. Technol. 1999, 33, 453-460. (19) Odum, J. R.; Hoffmann, M. R.; Bowman, F. M.; Collins, D.; Flagan, R. C.; Seinfeld, J. H. Environ. Sci. Technol. 1996, 30, 2580-2585. (20) Hallquist, M.; Wangberg, I.; Ljungstrom, E. Environ. Sci. Technol. 1997, 31, 3166-3172. (21) Wangberg, I.; Barnes, I.; Becker, K. H. Environ. Sci. Technol. 1997, 31, 2130-2135. (22) Alvarado, A.; Tuazon, E. C.; Aschmann, S. M.; Atkinson, R.; Arey, J. J. Geophys. Res. 1998, 103, 25541-25552. (23) Blando, J. D.; Porcia, R. J.; Li, T.-H.; Bowman, D.; Lioy, P. J.; Turpin, B. J. Environ. Sci. Technol. 1998, 32, 604-613. (24) Kavouras, I. G.; Mihalopoulos, N.; Stephanou, E. G. Environ. Sci. Technol. 1999, 33, 1028-1037. (25) Kleindienst, T. E.; Smith, D. F.; Li, W.; Edney, E. O.; Driscoll, D. J.; Speer, R. E.; Weathers, W. S. Atmos. Environ. 1999, 33, 36693681. (26) Vereecken, L.; Peeters, J. J. Phys. Chem. A 2000, 104, 1114011146. (27) Calogirou, A.; Larsen, B. R.; Kotzias, D. Atmos. Environ. 1999, 33, 1423-1439. (28) Jang, M.; Kamens, R. M. Atmos. Environ. 1999, 33, 459-474. (29) Hatakeyama, S.; Izumi, K.; Fukuyama, T.; Akimoto, H.; Washida, N. J. Geophys. Res. 1991, 96, 947-958. (30) Pandis, S. N.; Paulson, S. E.; Seinfeld, J. H.; Flagan, R. C. Atmos. Environ. 1991, 25A, 997-1008. (31) Kamens, R.; Jang, M.; Chien, C.-J.; Leach, K. Environ. Sci. Technol. 1999, 33, 1430-1438. (32) Kamens, R. M.; Jaoui, M. Environ. Sci. Technol. 2001, 35, 13941405. (33) Zhang, K. M.; Wexler, A. S. J. Geophys. Res. 2002, 107, art. no. 4577. (34) Jang, M.; Czoschke, N. M.; Lee, S.; Kamens, R. M. Science 2002, 298, 814-817. (35) Iraci, L. T.; Essin, A. M.; Golden, D. M. J. Phys. Chem. A 2002, 106, 4054-4060. (36) Noziere, B.; Riemer, D. D. Atmos. Environ. 2003, 37, 841-851. (37) Ziemann, P. J. J. Phys. Chem. A 2003, 107, 2048-2060. (38) Jang, M.; Carroll, B.; Chandramouli, B.; Kamens, R. M. Environ. Sci. Technol. 2003, 37, 3828-3837. (39) Jang, M.; Lee, S.; Kamens, R. M. Atmos. Environ. 2003, 37, 21252138. VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1433

(40) Czoschke, N. M.; Jang, M.; Kamens, R. M. Atmos. Environ. 2003, 37, 4287-4299. (41) Hughley, C. A.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G. Energy Fuels 2001, 15, 1186-1193. (42) Hanton, S. D. Chem. Rev. 2001, 101, 527-569. (43) Rader, H. J.; W., S. Acta Polym. 1998, 49, 272-293. (44) Nielen, M. W. F. Mass Spectrom. Rev. 1999, 18, 309-344. (45) Zoller, D. L.; Johnston, M. V. Macromolecules 2000, 33, 16641670. (46) Cox, F. J.; Dasgupta, A.; Johnston, M. V. J. Am. Soc. Mass Spectrom., in press. (47) Hatakeyama, S.; Ohno, M.; Weng, J.-H.; Takagi, H.; Akimoto, H. Environ. Sci. Technol. 1987, 21, 52-57. (48) Hatakeyama, S.; Izumi, K.; Fukuyama, T.; Akimoto, H. J. Geophys. Res. 1989, 94, 13013-13024. (49) Yu, J.; Flagan, R. C.; Seinfeld, J. H. Environ. Sci. Technol. 1998, 32, 2357-2370. (50) Bunnelle, W. H.; Shangraw, W. R. Tetrahedron 1987, 43, 20052011. (51) Grieco, P. A.; Parker, D. T.; Fobare, W. F.; Ruckle, R. J. Am. Chem. Soc. 1987, 109, 5859-5861.

1434

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

(52) Wijnen, J. W.; Engberts, J. B. F. N. J. Org. Chem. 1997, 62, 20392044. (53) Glasius, M.; Calogirou, A.; Jensen, N. R.; Hjorth, J.; Nielson, C. J. Int. J. Chem. Kinet. 1997, 29, 527-533. (54) Peters, A. J.; Lane, D. A.; Gundel, L. A.; Northcott, G. L.; Jones, K. C. Environ. Sci. Technol. 2000, 34, 5001-5006. (55) Cruz, C. N.; Pandis, S. N. J. Geophys. Res. 1998, 103, 1311113123. (56) Andracchio, A.; Cavicchi, C.; Tonelli, D.; Zappoli, S. Atmos. Environ. 2002, 36, 5097-5107. (57) Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.; Simoneit, B. R. T. Atmos. Environ. 1993, 27A, 1309-1330. (58) Oberdorster, G.; Sharp, Z.; Atudorei, V.; Elder, A.; Gelein, R.; Lunts, A.; Kreyling, W.; Cox, C. J. Toxicol. Environ. Health 2002, 65, 1531-1543.

Received for review September 17, 2003. Revised manuscript received November 26, 2003. Accepted December 16, 2003. ES035030R