Jeffrey S. GatTney Gerald E. Streit W.Dale SpaU John H.Hall Los Alamos National Laboratory Los Alomos, N.M. 87545
Everywhere in the popular and scientific literature one finds the term “acid rain.” It is increasingly apparent that the deposition of acidic sulfate and nitrate is responsible for the acidification of poorly buffered lakes in portions of Scandinavia and North America (1.2). In addition, it may well be the cause of forest decline in West Germany and other industrialized nations (3-5). Because of its regional (and therefore international and political) nature (5, 6) and because of the large costs involved with controlling acid deposition (5, 3, this phenomenon has received a great deal of attention within the environmental research community during the past few years. Numerous programs have been initiated to study the effects of acid rain on crops, lakes, and forest ecosystems. Programs have ranged from laboratory and field studies using simulated acid rain to theoretical investigations and extensive modeling calculations aimed at determining the doses of acidic species to regions of concern (1). To determine the effects on ecosystems and on specific species of plants and animals, almost all of these studies
have focused principaJly on sulfuric acid and pH and to a lesser extent on nimic acid. Because ~ h l r a lsystems also can lead to the deposition of organic and inorganic acids in remote areas, scientists studying acid rain have encountered difficulties in sorting out the man-made effects from those of natural systems. False leads crop up constantly (@, and research must address them if we are to make an accurate assessment of the sources and consequences of polluted precipitation. Although environmental researchers often treat the two phenomena separately, photochemical oxidant formation (smog chemistry) is strongly coupled to the production of acidic species in the atmosphere. Inorganic and organic trace gases and aerosols from natural and man-made sources, as well as their interactions, are important in both regional oxidant formation and in acid rain formation (9, 10). The troposphere (the lower 1C15 km of the atmosphere) is an oxidizing medium; reduced species released into it are likely to be oxidized to form oxidants and acids. In addition to inorganic molecules, these reduced species include a wide variety of organic molecules that typically increase in polarity and aqueous solubility as they are oxidized. The basic physical and chemical processes involved in the washout of soluble gases and aerosol species are not limited to sulfur dioxide, nitric acid, and sulfuric acid. Many atmospheric trace gases are likely to be
~)13g36W87~0921.05191.50/0@ 1987American Chemical Society
highly soluble in precipitation, and they too can undergo further chemical reactions in clouds, fog, and rain (9, 10).
More than acid in acid rain When one considers the variety of soluble oxidants and trace organic compounds found in polluted air masses, it is easy to see that there is much more than hydrogen ion, sulfate, nitrate, and trace inorganic elements in contaminated precipitation. Measurements made since the 1850s have shown that precipitation contains hundreds of organic compounds (11). It also is well documented that a number of stable organic compounds, which are likely to be incorporated into precipitation, are found in the atmosphere in both the gas and the aerosol phases (12). We must begin to reassess those oversimplificationsof research that use simulated acid rain that contains only stable inorganic species, particularly because many other reactive and organic gases and aerosol species are highly soluble in aqueous systems and may be concentrated in snow and rain, as strong acids are. When incorporated into aqueous systems, many species can react with other dissolved gases to produce large numbers of intermediates as well as acidic and oxidized products. Solubility, reactivity, and toxicity in acidic solutions are distinct molecular properties; thus, the solubility, reactivity, and toxicity of a precursor molecule does not define the resultant solubility, reactivity, or toxicity of a product. TWO Environ. Sci. Technol., Vol. 21, No. 6. 1987 510
of the best documented soluble gases are hydrogen peroxide and formaldehyde. These species, which have both natural and anthropogenic origins, are typical examples of reactive species in polluted precipitation. Hydrogen peroxide, formaldehyde Because of the primarily gaseous nature of the troposphere, most acid rain studies have focused on vapor-phase free-radical transformations of primary and secondary pollutants to organic acids. It has become increasingly apparent that gaseous interactions with aqueous solutions in the form of wet aerosols, fogs, clouds, or precipitation are likely to be important. These gasliquid interactions are controlled by gaseous transport to the aqueous surface; transport into the solution; establishment of a gas-liquid equilibrium; or chemical reaction, such as hydrolysis, oxidation, or solvation in the droplet. As Schwartz and Freiberg point out, in many cases transport into the solution, rather than gas-phase transport to the aqueous surface, is the rate-determining step for aqueous reactions (13). Henry’s law is used to describe the equilibrium solubilities of gases with dissolved species in the aqueous phase (14). This is given by
[XI = H3, (1) in which [XI is the equilibrium concentration of X in solution, P, is the gasphase pressure, and H, is the Henry’s law constant in moles per liter-atmosphere. Highly soluble gases, those that have Henry’s law constants >1OOO, can be critical in gas-to-aqueous-phase interactions because they can be concentrated in the aqueous phase and because they reach concentrations appreciablv hieher than thev would if thev remaked-in the gas phase. Hydrogen peroxide and formaldehyde are two such highly soluble species. The role of hydrogen peroxide (H2@) in forming acid rain has received considerable attention recently (9, 10, 13-18) because it reacts quite readily with sulfur dioxide (SO,) in the aqueous phase to form acidic sulfate: H202 + H20(1) HzSOd (2) This rapid liquid-phase reaction is acid catalyzed and is believed to be responsible for a large fraction of the conversion of sulfur dioxide to sulfuric acid during atmospheric transport (9, 1318). This is because sulfur dioxide is reasonably soluble in water and because the gas-phase reaction of sulfur dioxide with the hydroxyl radical (OH) is relatively slow. In contrast, nitric acid formation is dominated by the homogeneous gas-
+ Sa
-
520 Environ. Sci. Technol.. Vol. 21, No. 6. 1987
phase reaction of nitrogen dioxide ( N 4 ) with the hydroxyl radical. The rate constant for hydroxyl radical reaction with nitrogen dioxide (6.6 X IO9 L mol-I s-I) is approximately 10 times greater than that for sulfur dioxide (5 x lo8 L mol-l s-I) at rcmm temperature and atmospheric pressure (9.14). The comparatively low aqueous solubility of nitrogen dioxide, as compared with sulfur dioxide, limits the aqueous oxidation of nitrogen dioxide such that the vapor-phase reaction with the hydroxyl radical is the major route to nitric acid formation. The most likely source of hydrogen peroxide is gas-phase production from hydroperoxyl radical disproportionation (Y, 14):
-
- ___
2H0, H202 + 9 (3) Although aqueous-phase production also has been proposed, gas-phase concentrations of hydrogen peroxide via Equation 3 is expected to reach parts per billion by volume (ppbv) levels in tropospheric air. These concentrations have been postulated by computer models (9) and confirmed by direct Because of its measurement (9,15,13. extreme solubility in water (Figure 1). hydrogen peroxide should partition into ide and formaldehyde also have been the aqueous phase. Analytical techniques similar to those observed to vary strongly with the seaused to make gas-phase measurements sons; higher levels are seen during the reveal that hydrogen peroxide is found summer, when the atmosphere is more in rain and cloud water. These mea- photochemically active. The precursors sured aqueous concentrations represent of these soluble gases are both natural excess hydrogen peroxide that remains and anthropogenic, and they are proafter the available dissolved sulfur di- duced synergistically through interacoxide has been titrated (Equation 2). tions between primary and secondary Acid dissociation at lower pH values pollutants. leads to higher effective solubilities for the stronger acidic species. The data in- Other soluble species Other organic toxins are expected to dicate that rainwater samples taken durinr the summer wpicallv contain excess be observed in precipitation; some exhgrogen peroGae a i concentrations amples, together with their expected > 10 pmol, sometimes reaching 1W gas-aqueous equilibrium solubilities in cloud water, are listed in Table 1. The 150 pmol in acidic solutions. Formaldehyde is another molecule examples presented include organic for which researchers have obtained acids and oxidants as well as some phoboth gas-phase and aqueous-phase data tochemically produced aromatic ni(10,11, 18-20). Although formalde- trates and phenols. This list is not allhyde is less soluble than is hydrogen inclusive, but it does indicate the peroxide (Table 1 and Figure I), its variety of potentially soluble materials gas-phase concentrations have been cal- and their ranges of solubilities. It was culated and observed to be in the range obtained largely from gas chromatoof 1 4 p p b v in the gas phase (19-21). graphic partitioning data and available This leads to expected values of less solubility information (10). The observation of stable organic than 10 pmol to more than 50 pmol in precipitation. Formaldehyde has been compounds in precipitation is not new measured in this concentration range in (11). But a number of these species, fogs and rain at a number of sites especially oxalate, phenols, nitrated worldwide (11). phenols, and formaldehyde-sulfur diFormaldehyde, sulfate, nitrate, and oxide adducts have been determined rehydrogen peroxide have been reported cently in precipitation and have been in precipitation samples taken in the air found in concentrations ranging from a over Long Island, N.Y. (18);levels as few nanomoles per liter to several mihigh as 6Opmol have been observed. cromoles per liter (11,22-23. Some of The concentrations of hydrogen perox- these compounds probably are derived
I
IAULt 1
Gas-aqueous solubilitiesfor selected organic compounds mol I
ForrnWhyde
I
Acetaldehyde Acrolein
Crobonaldehyde Formic acW Oxalc aciff
AceUc a c e Methanol Elhanoi Phenol orthdxesol Peracetc acd Melhylhydroperoxde para-Cresol Peroxyacei-yl ntale Hydrogen wand@ Methyl cyantde N,Imrnethane Nslroethane Acetone Dinmlhyt sulfide BiacetyP
Benzahlehvde >Hydroxy&nzaldehy& 3-Nitrooheno1
yacetyl nitrate) have been studied much iess ixtensive~y. Some preliminary studirs examining the effects of hydrogen peroxide on Norway spruce and other species of young trees have been reponed by Slanina and co-workers ( 2 8 29). Aithough no visible damane was observed initially, microscopic &dies clearly indicate that dissolved hydrogen peroxide in acidic solutions can cause cellular damage. Very little, however, is known of hydrogen peroxide's toxicity to plants. A background of hydrogen peroxide is produced naturally in the atmosphere from the photooxidation of isoprene, methane, and other naturally emitted organic compounds. Because of this background and that of other soluble oxidants and toxins, the interaction of both pH and pollutant concentration on biota must be determined. Because plants themselves produce hydrogen peroxide biochemically and use peroxidase to control its levels in vivo, the potential biochemical synergisms in the total plant and within its subsystems need to be understood when the potential effects of hydrogen peroxide and other codeposited gas or aqueous pollutants are examined. Although formaldehyde is toxic and is a suspected human carcinogen, little information is available regarding its potential biochemical danger to plants and microorganisms at currently observed levels. In the late 1950s and early 1960s studies of plants exposed to gaseous formaldehyde showed little if any visible plant damage. This was not the case for plants exposed to peroxyacetyl nitrate (PAN), ozone, or sulfur dioxide (28-32). Aqueous exposures were not examined, however, and no microscopic or biochemical studies were performed following gas-phase exposures to determine whether plant damage or weakness had occurred that was similar to the hydrogen-peroxideinduced effects observed in Norway spruce. As with hydrogen peroxide, we need to know the potential toxicity and biological effects of formaldehyde as a function of pH. We also should learn which biochemical pathways are affected to assess how other sources of stress, such as frost, drought, and disease-in conjunction with soluble oxidants and toxins in acidic media-will affect forest and agricultural ecosystems. A number of researchers have shown that hydrogen peroxide and formaldehyde have natural backgrounds in the atmosphere because of the oxidation of methane and other naturally occurring hydrocarbons (10, 16, 20,33, 34). It is important IO undentand this if we are to assess the relat~velmportilnce of human activities in contributing 10 tht quantity
4
5
from gas-phase reactions (for example, nitrated phenols and phenols are formed by the photooxidation of aromatic compounds [251). Others, such as organosulfinic acids, are more likely to be formed by aqueous reactions of soluble gases (23). Aqueous hydrogen peroxide-sulfur dioxide reactions are not the only ones that are possible for atmospheric pollutants. We have only begun to touch on the possible importance of reactions that will lead to formation of other soluble species in polluted precipitation.
Potential toxicity Although the observed concentrations of many organic species in polluted precipitation are much lower than the concentrations of sulfate and other major inorganic species, their relative toxicities may be much greater than are those of the inorganic species. Furthermore, although the potentially high toxicity of trace components is now beginning to be understood, most of the work to date has focused on sulfate, hydrogen ion, and nitrate in precipitation and on gas-phase ozone. Some of the other constituents in precipitation (such as hydrogen peroxide, formaldehyde, and trace organic toxins) and gasphase organic toxins (such as perox-
of these toxins. Researchers also must not ignore the synergism between these na~rallyderived pollutants and manmade oxidanh, acids, and inorganic and organic toxins if they propose to evaluate the effectivenus5 of any control strategy. Modelinn work and field data strongly suggest that sulfur dioxide control will decrease sulfuric acid production to some extent, depending on the effective oxidant distribution. Equation 2, however, shows that a decrease in sulfur dioxide with no concurrent reduction in nitrogen oxides may lead to increased concentrations of hydrogen peroxide in precipitation. What are the consequencesof such a control strategy on affected ecosystems? Would lake acidification be lessened while forest damage increased? Dose-response relationships for these ecosystems must be determined if such questions are to be answered. A number of other atmospheric materials are likely to be of concern in any attempt to evaluate ecosystem effects. Ozone, which has long been known to be a potent plant and microbe toxin, has received considerable attention from environmental researchers (5, 14, 35). Organic oxidants, such as PAN and peracetic acid (both of which are gasphase plant toxins [28-32]), have received considerably less attention. Although they have low to moderate solubilities in water (Table l), their action in aqueous solution has not been well characterized. Peroxyacetyl nitrate and peracetic acid chemistries are coupled (9,10,16, 34), and recent data indicate that PAN has a long lifetime in the cold upper troposphere (3639)and that it is transported worldwide (4&46). In recent studies, PAN has been observed to act as a gas-phase plant mutagen (10)and as a potent Ames-active mutagen (47). Peracids have been observed to be strong phytotoxins and bactericides (10,30, 47). Despite these indications of the genotoxic nature of these compounds, and despite their likely solubility (and therefore their mobility) in the lipid-rich waxy cuticles that protect plants (10,36, 3 3 , we know little a b u t the toxicity of these oxidants at levels expected in the environment (nanomolar to low micromolar concentrations in acidic precipitation). Phenols, oxalates, and nitrophenols also are known organic toxins that have long been used in large quantities as fungicidesand bactericides. The effects of these compounds and of the peracids at the levels found in polluted acidic precipitation (11)have not been characierizd. Clearly, the potential for these materials to cause acute and chronic effects in plants and animals must be Enwon. Sco Tecnnol , Voi 21 ho 6 1987 521
evaluated if we are to understand the nature of polluted precipitation or take action to control it. It has been pointed out that all of these toxic organic species are highly soluble in water (Table 1) (10, 11, 22-27). Here again, we do not know the consequences of the presence of these potential toxins in acidic rainfall. These toxins (which can act as potent bactericides) may well have broad regional and global effects on microbial populations, thus effecting their emissions of trace gases. For example, if regional deposition of toxins affects aerobic bacteria selectively, anaerobic populations will increase. Anaerobic bacteria are a major source of atmospheric methane, the concentration of which is currently increasing at a rate of - 2 % annually in the Northern Hemisphere (48).Is the regional deposition of these materials being reflected in the global increase of this greenhouse gas? We must examine these interactions between pollutant deposition and natural trace gas sources if we are to supply answers to this and other questions.
Natural reactive hydrocarbons Another example of the potential for interactions between natural and anthropogenic pollutants involves the emission of natural hemiterpenes and monoterpenes from deciduous and coniferous forests. It is possible that increased levels of anthropogenically generated nitrogen oxides are raising the photooxidation rates of natural isoprene emissions from oak, maple, and other deciduous forests. As these transformations occur during regional transport, we may see either increased levels of isoprene photooxidation products or a change in the location at which such products are observed. It has been shown that isoprene photooxidation causes increases in formaldehyde concentrations in the air above remote deciduous forests (10, 16, 33). This also leads to the production of increased levels of hydrogen peroxide (Equation 3) because the photooxidation of formaldehyde leads directly to the formation of the hydroperoxyl radical. Moreover, higher levels of organic oxidants, such as peracetic acid, are expected in areas that are distant from sources of nitric oxide because peracetic acid is formed by reaction of the peroxyacetyl radical with the hydroperoxyl radical: CH3CO-02 HO2 CH3CO-OOH 0 2 (4) In typical urban environments,which contain local sources of nitric oxide, the HOz, ROz, and RC03 radicals will react to form OH, RO, and RCO;?radi-
+
+
-.
522 Environ. Sci. Technol., Vol. 21, No. 6, 1987
+
cals (and subsequently R02 COz) and N02. These chain-carrying radicals lead to the oxidation of nitric oxide to nitrogen dioxide, which in turn photolyzes to increase ozone levels. Away from the sources of nitric oxide, the peroxyradical termination processes that lead to the formation of hydrogen peroxide, organic peracids, and organic peroxides become more important (Equations 3 and 4). In remote regions where nitric oxide levels are very low, the termination of peroxyradicals becomes important to the production of oxidants other than ozone. Using a simple box model computer simulation, we have examined the potential production of these species first without isoprene and then with increasing emission rates of isoprene (Figure 2). In all cases, the addition of isoprene affects the concentration levels of the species depicted. This is especially true for formaldehyde, hydrogen peroxide, and peracetic acid, all of which are likely to be involved in acid precipitation reactions. The model results deal only with the gas phase and would be modified by aqueous solubilities and reactions. The implication is that isoprene emissions from deciduous forests are likely to have direct consequences for a number of soluble oxidants and organic toxins as they are oxidized in the troposphere in the presence of catalytic oxides of nitrogen and man-made organic pollutants. The production rate trends in this simulation are presented to demonstrate the potential importance of natural hydrocarbons (in this case isoprene) as sources of formaldehyde and other potentially noxious species in remote atmospheres. The key to the production rates from these natural sources lies in the interactions of the anthropogenically derived catalytic photooxidizers, such as ozone and hydroxide, which are expected to increase with increasing levels of nitrogen dioxide. Increased levels of hydrogen peroxide expected to occur over deciduous forests as a result of isoprene oxidation will lead to increased sulfuric acid aerosol over these regions when sulfur dioxide reacts with hydrogen peroxide in wet aerosols, fogs, and cloud droplets (Equation 2). Analysis of the aerosol over such areas indicates that the haze that blankets the Blue Ridge and the Great Smoky mountains (which are predominantly deciduous forests) is mainly sulfate aerosol (20, 49) and not organic aerosol, as has been suggested by Went (50). Although the blue haze of the Smokies predates heavy industrialization, and hence increased levels of sulfur attributable to human activities, the current haze probably is ascribable both to isoprene oxidation and to long-
range transport. Thus, isoprene oxidation leads to formation of hydrogen peroxide, which can efficiently convert any available sulfur dioxide to acidic sulfate in wet aerosols, fogs, and clouds. Field work at Oak Ridge National Laboratory has indicated that formaldehyde levels can be appreciable in deciduous forests. Gas-phase formaldehyde concentrations approaching 50 ppbv were measured in studies carried out there (10, 20). These data are consistent with the expected interaction between anthropogenic nitrogen oxide emissions and reactive natural hydrocarbons that leads to increased production of hydrogen peroxide, formaldehyde, and organic oxidants such as PAN and peracetic acid.
Research focus must widen Polluted precipitation contains much more than hydrogen ion, sulfate, and nitrate. Computer modeling and theoretical work indicate that a number of potentially reactive oxidants and organic toxins are expected to be precipitated along with the acid in acid rain (9, IO). Available field data support this hypothesis, particularly for the cases of hydrogen peroxide and formaldehyde. Many other organic oxidants and toxins are also expected to be present in the gas and aqueous phases. Their toxicities in low nanomolar to micromolar concentrations in acidic solutions have not, however, been established. If we are to continue to study acid rain, particularly by simulation, we cannot neglect the possibility that soluble organic oxidants and toxins may be playing important roles in observed or potential damage to forests and crops. Computer simulations that have addressed acid rain transport and transformation questions use simplified chemical mechanisms aimed at modeling inorganic acid production and deposition. This has been required by the need to conserve computer time and by the narrow focusing of research on the problem as it is defined by the agencies that support the research. Most current models do not simulate the mix of organic species that may be important oxidants or phytotoxins. Although the natural sources of sulfur species and nitrogen oxides are much smaller than those derived from human activities, this is not necessarily the case for reactive hydrocarbons. Many of these materials have natural backgrounds derived from reactive hydrocarbons emitted from trees and other biota, and we need to assess how the increased nitrogen oxide photochemistry caused by the combustion of fossil fuels, for example, affects their levels and reaction products. Many of these
products will be oxidants, aldehydes, exposures should be studied to assess and other potentially toxic species in the potential for a variety of organic the gas and aqueous phases. If the spe- species-alone and in combination with cies are highly soluble, as is the case oxidants and acids-to cause plant with hydrogen peroxide, they will be . stress and other ecological damage. If effectively concentrated in rain and adequate and effective control strakfogs from the gas phase. gies are to be invoked we must underIt is clearly possible that together stand the problem as a whole and not with organic toxins, a number of oxi- just as it is popularly perceived. Finally, we need to learn from the dants other than ozone are being deposited by acid rain onto forest and agri- past. Atmospheric pollution can occur cultural ecosystems. We need to assess on small and large scales, and we are the potential effects by directing b i o becoming more aware that urban polluchemical and plant research into these tion is not the only atmospheric probareas. Gas-phase and aqueous-phase lem we will face. Relatively inert mate-
rials may not seem to be cause for concern in studies of toxicity or determinations of what produces oxidants or acid rain, but we must remember that their infrared spectral properties and upper atmospheric chemistries may be important. For example, when chlorofluorocarbons are emitted into the troposphere they are relatively inert chemically. But once they are transported into the more energetic stratosphere, there is evidence that they affect the levels of upper atmospheric ozone, which acts as an important ultraviolet radiation Environ. Sci. Technal.. Vol. 21, No. 6, 1987 523
screen. Also, although chemically inert in the troposphere, these same substances have infrared absorption prop erties that make them extremely effective as greenhouse gases in the lower atmosphere while they await transport across the tropopause (50). Many trace gases emitted into the atmosphere can have multiple effects; as acids, oxidants, and precursors to toxins and as greenhouse gases and sources o f reactants for upper atmospheric chemistry. Urban emissions lead to the regional and global distribution and deposition of pollutants. These scales are not isolated from one another, and solutions to problems in one may lead to new prob lems for another. For example, the use of tall stacks to disperse pollutants may abate local air pollution, but it can cause the regional and global distribution and deposition of primary pollutants and their reaction products. The atmospheric transport and transformation of air pollutants is a complex process involving organic and inorganic compounds, oxidants, and acids, all of which can interact in the gas and aqueous phases. As we look beyond acid rain to regional atmospheric deposition and its effects, we must continue to develop a better fundamental understanding of how the system works and of how it interacts with a biosphere that is anything but passive. Acknowledgment We thank John Seinfeld and Jack Calvert for their encouragement. T h e support of David Ballantine and David Slade of the Department of Energy's Office of Health and Environmental Research is also appreciated. Helpful discussions with Gunnar Senum, Thomas Kelly, and Yin-Nan Lee of Brookhaven National Laboratory also are acknowledged. This work was supported by the US. Department of Energy and carried out at Los Alamos National Laboratory. T h e opinions expressed in this article a r e those of the authors and d o not reflect the views of the Department of Energy o r Los Alamos National Laboratory. Before publication, this article was reviewed for suitability as an ES&T feature by Jack G . Calvert, National Center for Atmospheric Research, Boulder, Colo. 80307; Steven H. Cadle, General Motors, Warren. Mich. 48090: and James R. Kramer. McMaster University. Hamilion. Ont L8S 4M I , Canada
(8) Havas. M.; Hutchinson. T.C.; Likens, G . E. Environ. Sei. Technol. 1984, 18. 17685A. (9) Calvert, J. G. etal. Notum 1985,317, 2735. (IO) Gaffney, 1. S . ; Senum. 0 .1. In Cos-Liquid Chemistry of Norural Woters; Newman. L.. Ed.; NTlS TIC-4500. UC-II; BNL 51757: Brookhaven National Laboratory: Upton, N.Y., 1984; pp. 5-1 105-7. (11) Mazurek. M. A.; Simoneit, B.R.T. CRC Crit. Rev. Environ. Conrrol 1986, /@I), I 140. (12) Graedel. T E. Chcmicol Compounds in the Almosphere: Academic Press: New Vnrk~ 19711~ ..... . . .. (13) Schwvartr. S. E.; Freiberg, 1. E. Atmos. Environ. 1981,lS. 1145-54. (14) Finlayson-Pills, B. 1.; Pitts, J. N.,Jr. In
.
Atmospheric Chemisrry: Fundomentdr and Experimenrol Techniques; Wiley-Inter-
science: New York, 1986; pp. 273-75. (15) Lind. J. A,; Kok. G. L. J. Geophys. Res. 1986.91, 7889-95. (16) Streit. G.E.;Gaffney. J.S. Paper 79. Presented at the Symposium on the Chemistry of the Atmosphere. 8th Rocky Mountain Regional Meeting of the American Chemical Society. Denver. 1986. (17) Daum, I? H. et al. Atmos. Environ. 1984, 12, 2671-84. (18) Lee, Y.-N.; Shen. 1.; Klotz, P.J. Water Air Soil Pollur. 1986.30, 143-52. (19) Tanner, R. L.; Meng, Z. Environ. Sci. Technol. 1984. 18. 723-26. (20) Tanner, R.L.; Gaffney. J.S. Paper 8416.2. Presented at 77th Annual Air Pollution Control Association meeting, San Francisco, June 1984. (21) Harris. C. W.; Schiff, H. 1.; MacKay. G. 1. EOS Trons. Am. Geophys. Union 1985, 66.817; Abslracl A121-06. (22) Curran. C. M . ; Norton. R. B. EOS Trans. Am. Geophys. Union 1985, 66, 823; Abstract A22A-09. (23) Munger. J. W.; Tiller. C.; Hoffman. M. R. Science 1986,231. 247-49. (24) Leuenberger, C.; Ligocki. M. P.; Pankow, J. F. Environ. Sci. Technol. 1985, 19, 1053-58. (25) Grosjean. D. Environ. Sci. Technol. 1985. 19, 968-74. (26) Tremp. 1. et al. Absrroers of Papers, 192nd National Meeting of the American Chemical Society, Anaheim. Calif.; American Chemical Society: Washington. D.C.. 1986 ENVR4. (27) Kawamura, K.; Kaplan, 1. R. Atmos. Environ. 1986.20, 115-24. (28) Masuch, K. et SI. J. And. Environ. Chem. 1986,27(3). 180-213. (29) Masuch. K. el al. Presented at the International Workshop on the Physics and Biochemistry of Stressed Plants; Neuherberg, West Germany. May 1985. (30) Stephens, E. R. et al. Int. J. Air Water Pollut. 1961.4. 79-100. (31) Taylor. 0. C. J. Air Pollut. Conrrol Assoc. 1969, 19(5). 347-51. (32) Rerponse of Plants to Air Pollution; Mudd, I . B.; Kozolowski. T T.. Eds.; Academic Press: New York, 1975. (33) Brewer. D. A.; Levine, 1. S . In Ozone Symposium; Zerefos. C. S . ; Ghazi, A,. Eds.; Reidel: Hingham. Mass., 1985; pp.
(39) Senum, 0 . 1.; Lee. Y.-N.; Gaffney. J . S. J. Phys. Chem. 1984,88. 1269-70. (40) Lanneman. W. A,; Bufalini. J. J.; Seila. R. L. Environ. Sri. Technol. 1976, IO. 37480. (41) Singh. H . B.; Hans, P L. Gmphys. Res. Lzll. 1981.8, 941-44. (42) Spicer. H. B.; Holdren. M. W.; Keigley. G . W. Amos. Environ. 1983, 17. 1055-58. (43) Nielsen. T. et al. Nature 1981,293. 553. (44) Penkett, S . A,; Sandalls, F. J.: Lovelock. J. E. Almos. Environ. 1975.9. 139. (45) Temple, P I . ; Taylor. 0. C. Atmos. Em;TO". 1983, 17. 1583-87. (46) Hov. 0. J. Amos. Chem. 1984, 1. 187202. (47) Shepson. P. B. et al. Environ. Sei. Techno/.1986.20. 1008-13. (48) Senum. G. 1.; Gaffney. 1. S . In The Corban Cycle and Atmospheric C 0 2 : Norural Yariorims Archman Io Present. Geophysical Monograph 32; American Geophysical Union: Washington, D.C.. 1985; pp. 61-69. (49) Stevens, R . K. el al. Environ. Sci. Techno/. 1980, 14. 1491-97. (50) Went. F. W. Namrc 1960. 187. 641.
Jeffrey S. Caffney (1.) i . ~