On-Line Analysis of Reactive VOCs from Urban Lawn Mowing

Methanol and other VOC fluxes from a Danish beech forest during late springtime. Gunnar W. Schade , Sheena J. Solomon , Ebba Dellwik , Kim Pilegaard ...
0 downloads 0 Views 254KB Size
Download PDF
Environ. Sci. Technol. 2001, 35, 2926-2931

On-Line Analysis of Reactive VOCs from Urban Lawn Mowing T H O M A S K A R L , † R A Y F A L L , * ,‡ ALFONS JORDAN,§ AND WERNER LINDINGER§ Atmospheric Chemistry Division, National Center for Atmospheric Research, Boulder, Colorado 80307, Institut fu ¨r Ionenphysik, Universita¨t Innsbruck, Technikerstrasse 25, 6020 Innsbruck, Austria, and Department of Chemistry and Biochemistry and Cooperative Institute for Research in Environmental Sciences, University of Colorado, Boulder, Colorado 80309-0215

We measured the release of volatile organic compounds (VOCs) resulting from lawn mowing during continuous ambient air measurements in July and August 1998 in the outskirts of Innsbruck, Austria. These measurements were made with a proton-transfer-reaction mass spectrometry system, which allowed simultaneous, on-line monitoring of VOCs in the pptv range. We observed the emission of C6 wound compounds, including (Z)-3-hexenal, (E)-2-hexenal, hexenol plus hexanal, and acetaldehyde immediately following lawn mowing, and a rise in background levels of C6 wound compounds that lasted for several hours. Peak levels of biogenic VOCs following mowing were in the same concentration range (20-60 ppbv) as those originating from combustion engines of lawn mowers, and integrated biogenic emissions were much greater in the drying grass clippings. Additional emissions of acetone and other VOCs resulted from rainfall on these clippings. Since the estimated atmospheric chemical reactivity of VOCs resulting from lawn mowing is of the same order of magnitude as unburned hydrocarbons released during the mowing by gasoline-powered lawn mowers, these biogenic VOCs should be considered in urban air-quality control strategies.

Introduction The role of natural and man-made volatile organic compounds (VOCs) in shaping the chemistry of the atmosphere is an important area of research (1-4). The impact of VOCs on the formation of atmospheric oxidants, such as ozone, or secondary organic aerosols is not limited to the Earth’s remote forested areas. For example, forests surrounding metropolitan areas may release more reactive VOCs than man-made sources (5). Have we identified all of the significant natural sources of urban VOCs? Anyone who has mowed a lawn can attest that grass cutting releases VOCs, yet these compounds are not explicitly considered in urban air-quality models (69). It is well-known that wounding plants releases a mixture of C6 hexyl and hexenyl compounds that are derived from major leaf fatty acids; these VOCs are primarily responsible for the odor of cut grass (10). Kirstine et al. (11) used gas * Corresponding author phone: (303)492-7914; fax: (303)492-1149; e-mail: [email protected]. † National Center for Atmospheric Research. ‡ University of Colorado. § Universita ¨ t Innsbruck. 2926

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001

chromatography-mass spectrometry (GC-MS) methods to show that in addition to C6 VOCs, cutting grass or clover releases significant amounts of other oxygenated VOCs (such as methanol, ethanol, acetaldehyde and acetone) plus a variety of minor VOCs. Recently, de Gouw et al. (12) and Fall et al. (13) used proton-transfer-reaction mass spectrometry (PTR-MS) to measure VOC release from wounded plants, including grass. The advantage of the PTR-MS method for these types of experiments is that emissions of most biogenic VOCs can be monitored simultaneously and on-line down to the 10 pptv range without the need for preconcentration or chromatography (14). Here, we wanted to use this method to monitor the levels of VOCs in ambient air before, during, and after lawn mowing. It is pertinent to briefly review how PTR-MS apparatus can be used to monitor VOCs in air; further details are presented elsewhere (14). The apparatus consists of a conventional drift tube equipped with a hollow cathode ion source that produces H3O+ ions with only traces of impurity ions (mainly O2+), so that no mass spectrometer is needed to preselect H3O+ before injection into the drift tube. The air to be analyzed itself acts as the buffer gas in the drift tube (the pressure is typically 10-1 Torr); this is possible because H3O+ does not react with any of the main components of air as they all have lower proton affinities than H2O (15). On the other hand, H3O+ performs proton transfer in nondissociative reactions: k

H3O+ + R 98 RH+ + H2O

(1)

with all VOCs, R, with a higher proton affinity than H2O. The proton-transfer rate constants (k) are large, corresponding to the collisional limiting values (1.0 × 10-9 cm3 s-1 < k < 3 × 10-9 cm3 s-1). The concentration of a VOC is calculated from the ion count rates in counts per second (cps), cps(RH+) and cps(H3O+)0, obtained in the downstream ion detection system using

cps(RH+) ) cps(H3O+)0(1 - e-k[R]t) ≈

cps(H3O+)0[R]kt (2)

where t is the transit time of the H3O+ ions through the drift tube. The value of t is calculated from known mobility values of H3O+ in air. The reaction rate coefficients (k) related to the components studied in the present investigation lie in the range of 1.7-3 × 10-9 cm3 s-1 (15). During PTR-MS, all the variables in eq 2, except [R], are known, so it is easy to convert raw data for each cps(RH+) to the concentrations of individual VOCs. Periodic calibrations of the instrument with known VOC standards are used to verify these determinations. An additional significant feature of the PTR-MS instrument used in these experiments is that it can be operated for extended periods of time; for example, ambient air outside the laboratory in Innsbruck has been sampled for days, weeks, and even months at a time with little or no interruption except for calibrations (16). The stability of the instrument during these measurements, the large data set of ambient VOCs previously measured at this site, and the proximity of the inlet used to a grass lawn on the campus of Innsbruck University prompted us to undertake VOC analysis before, during, and after lawn mowing episodes. We anticipated that we might be able to detect and quantify the VOCs released during lawn mowing activities. The results of these experiments suggest that common lawn mowing releases sub10.1021/es010637y CCC: $20.00

 2001 American Chemical Society Published on Web 06/02/2001

stantial amounts of reactive VOCs and should be considered in urban air-quality control strategies.

Experimental Section PTR-MS equipment used for these measurements has been described in detail elsewhere (14). For these measurements a 10-m 1/8-in. PTFA inlet line, shielded from rain at the distal end, was placed outside the laboratory at Innsbruck University, and a continuous flow [700 standard cm3 min-1 (sccm)] of air was maintained with a diaphragm pump. The PTR-MS was connected to the line in a bypass at a flow rate of 20 sccm. The inlet system was surrounded on three sides by a lawn composed primarily of pasture grasses (orchard grass, Dactylis glomerata, and Kentucky bluegrass, Poa pratensis), maintained by normal watering and fertilization practices by groundskeepers. On two occasions (July and August 1998), we sampled air before, during, and after mowing of this lawn. In each case, three different mowing machines, described below, were used to cut the lawn near the sampling inlet over a period of about 1.5 h. In the first measurements, selected positive ions corresponding to m/z 45 (acetaldehyde), 59 (acetone), 81 ((Z)3-hexenal), and 83 (hexanal plus hexenols) were monitored to follow the formation of leaf wound compounds. Details of the correlation of detected positive ions with leaf wound compounds is presented in ref 13. In the second experiment, full mass scans were performed every 5 min during mowing and could be compared to the selected ion scans of the first experiment. Cuvette measurements of cut bluegrass and orchard grass were performed to calculate typical total VOC emissions per gram dry weight (g dw) of these pasture grasses; the grasses were sampled immediately after collecting them outside the laboratory. A constant flow rate of air (135 sccm) was maintained through a cuvette constructed from a Teflon ring sealed with glass plates on the top and bottom (350 mL volume), and the release of VOCs and water vapor (as the water dimer, H2O‚H3O+, mass 37) was monitored over the whole period of drying. As with the ambient air measurements, the PTR-MS was connected to the exit line of the cuvette in a bypass at a flow rate of 20 sccm, and the resulting peaks of released VOCs were integrated to obtain the total amount emitted. Several drying measurements for bluegrasss were performed at 25, 30, and 35 °C; orchard grass drying was measured at 25 and 35 °C. The first series of cuvette measurements was done in late August 1998; the grass samples showed a high content of leaf tissue, and florets represented the smaller fraction of the total dry weight. A second series of cuvette measurements was performed in May and June 1999 when the lawns were cut for the first time and had reached their full length of 50-60 cm; in this case, the total dry weight of florets was not negligible as compared to the rest of the plant.

Results and Discussion As illustrated in Figure 1, when outside air was continuously monitored by PTR-MS, before, during, and after mowing of lawn on the campus of Innsbruck University, the levels of (Z)-3-hexenal and the sum of hexenols and hexanal could be seen to rise after each pass of the mowers used. These C6 wound compounds were detected by selective ion monitoring; we have previously established that ions 81 and 83 are relatively unique markers of leaf wound C6 compounds arising by dehydration of the protonated parent aldehydes or alcohols in the drift region (13). At the beginning of the experiment shown, background concentrations of (Z)-3hexenal were below 100 pptv; a small signal at mass 81 in the morning was shown in other experiments not to correlate with mass 99 (protonated (Z)-3-hexenal) but was due to a mass 81 fragment arising from monoterpenes released from

a nearby conifer forest. Then, with the first pass of a mower, (Z)-3-hexenal rose within a few minutes to a peak of about 6 ppbv and then declined. In subsequent passes of the mower, even higher levels of (Z)-3-hexenal were detected; 16-20 ppbv was seen after passes 2 and 5. Three different mowers were used to cut the grass: a small power mower helped to cut the grass at edges, a typical two-stroke engine lawnmower cut the grass around trees and parts where the tractor used for the rest of the lawn could not pass, and 90% of the lawn was mowed with a tractor that also subsequently picked up most of the grass clippings. Some grass clippings (about 10%) stayed on the ground. Each pass of the three different mowers is indicated in Figure 1. The higher peaks of (Z)-3-hexenal release, peaks 2 and 5, resulted after one of the portable mowing machines passed within a few meters of the inlet. The smaller peaks 1, 3, 4, 6, and 7 are related to the tractor that collected the cut grass. Most passes of the lawn mowers also resulted in an increase in mass 83, indicative of the release of hexenols and/or hexanal, although the ambient concentrations remained lower than (Z)-3-hexenal, similar to the pattern seen in laboratory leaf wounding experiments (13). We did not see detectable traces of mass 143, which is characteristic of hexenyl acetates, the major hexenyl product from some wounded leaves (13). In addition, hexenyl acetates would also show fragments lying on mass 83. This suggests that the major part of mass 83 comes from hexenols and/or hexanal. Although the major fraction of the cut grass was collected just after cutting, the rest that stayed on the ground produced a small but distinct background of (Z)-3-hexenal and hexenols/hexanal that persisted over several hours. This background may be due to continued release of wound compounds from drying vegetation as discussed below, although on this day the low air temperature (about 18 °C) did not promote rapid drying. At 15:45, a light rainfall started and lasted over several hours with short interruptions of no precipitation. It is clearly seen that concentrations of C6 wound compounds increased from ∼200 to ∼600 pptv at the beginning of the rainfall and then decreased to the 200 pptv range within the next few hours. This decrease is probably caused by a combination of washout and a decrease in the production rate of these compounds. In this experiment, we also monitored a few other potential VOCs that might arise from lawn mowing, including acetaldehyde and acetone which Kirstine et al. (11) found to be significant VOCs in cut pasture grass. As shown in Figure 2, releases of acetaldehyde were also seen to correlate with each pass of the mowers; peak acetaldehyde levels at the inlet reached up to 23 ppbv at mower pass 7. On the other hand, lawn mowing seemed to have a smaller influence on the release of acetone, but a big increase with peaks between 100 and 200 ppbv was seen after the rainfall started. This effect on acetone release from cut and partially dry grass is consistent with a production mechanism proposed by Warneke et al. (17): that a great fraction of biogenic acetone from leaf litter is produced by nonenzymatic Maillard reactions during the drying and decaying phase and is released after wetting. A second lawn mowing experiment was conducted in August 1998. In this case, the PTR-MS was operated in a mass scan mode where positive ions from 20 to 160 amu were scanned continuously every 5 min. This large data set can be summarized briefly. As with the first experiment, with each pass of the mowers spikes of (Z)-3-hexenal and hexenols plus hexanal were seen. Additionally, other ions indicative of C6 wound compounds were seen to increase with mowing: mass 57, (E)-2-hexenal, and mass 85, hexanol. Also seen were releases of methanol (mass 33), which may be released from leaf pools by wounding (18), ethanol or formic VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2927

FIGURE 1. Ambient levels of (Z)-3-hexenal and hexanal plus hexenols before, during, and after lawn mowing. The experiment shown was done on July 16, 1998, with wind speeds varying between 0.2 (morning) and 3.3 m/s (late afternoon), temperature ranging from 13 to 18 °C, relative humidity varying from 60% (between 8:00 and 14:00) up to 100% at 15:30. Rainfall started at 15:40 and continued for several hours with some interruptions. The PTR-MS instrument was operated in selective ion mode so that measurements of ions m/z 81 and 83 were obtained every 42 s. acid (mass 47), acetic acid (masses 61 and 43), butanol (mass 57), butanone (mass 73), pentenols plus 2- and 3-methylbutanal (mass 69), and several other unidentified VOCs. Several of these compounds were previously seen to be released by cut pasture grass (11). In addition to biogenic compounds seen in experiment 2, the PTR-MS signal could be used to monitor compounds coming from exhaust plumes of the lawn mowing machines. For example, benzene and toluene produce distinctive positive ions, m/z 79 and 93, respectively (14); we detected peak concentrations of these hydrocarbons between 20 and 60 ppbv at each pass of a lawn mower. In this experiment, we saw that emissions of VOCs from cut grass are in the same peak concentration range as those originating from combustion engines of lawn mowers. To verify that the primary grasses in the lawn adjacent to the sample inlet, Kentucky bluegrass and orchard grass, give 2928

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001

rise to the major leaf wound compounds seen in the mowing experiments, we placed cut samples of each in a leaf cuvette and analyzed VOCs released during drying by PTR-MS. It has been previously shown that VOC release from cut leaves, including grass, is greatest during the drying phase, presumably because leaf drying leads to the collapse of cellular structures (12, 13). As shown in Figure 3, following cutting of Kentucky bluegrass an initial transient release of (Z)-3hexenal (mass 81) was seen. Then, as the grass clippings dried, a large extended second release of (Z)-3-hexenal occurred over a period of hours. The water release from the drying grass was monitored continuously by mass 37, a protonated water cluster, and as shown in Figure 3, drying was accelerated by raising the temperature of the cuvette. Release of (Z)-3-hexenal decreased at the highest temperature (35 °C), probably due to decreased enzymatic activity associated with wound compound formation. We did not

FIGURE 2. Ambient levels of acetaldehyde and acetone before, during, and after lawn mowing. The experiment is the same as described in the legend of Figure 1. see this trend with acetaldehyde, acetone, and other VOCs known to have pools inside a plant. In general, the release pattern of all VOCs was similar to that of hexenyl compounds showing the highest concentrations at about the same time. Concentrations of acetaldehyde usually decreased slower than those of wound compounds. Table 1 summarizes the total amounts of (Z)-3-hexenal, acetaldehyde, and acetone released from drying grasses; these data were obtained by integrating curves such as those shown in Figure 3. For the two grasses studied, (Z)-3-hexenal release at 25 °C ranged from 100 to 240 µg g dw-1 for bluegrass and was about 70 µg g dw-1 for orchard grass. The range seen for bluegrass represents determinations made in August 1998 (high values) and May/June 1999 (low values); the low values seen in the spring were due in part to the fact that florets and stems contributed significantly to the total dry weight but less to the wound compound emissions. For both grasses, smaller amounts of hexenol plus hexanal, acetaldehyde, and acetone were released (Table 1); again, for bluegrass, higher levels of VOC release on a gram per dry weight basis were seen in samples tested in spring than in summer, possibly reflecting differences in the physiological state of the grass in different seasons. The total emission of hexenyl compounds from drying grasses ranged from 84 to 340 µg g dw-1; these values are of the same magnitude as those reported for hexenyl compound emissions from slashed pasture grass (140 µg of C g dw-1 h-1; 11).

What is the potential significance of these findings? There is continuing concern about the formation of air pollutants in urban areas and increasing scrutiny of reactive VOC sources, including minor sources, that give rise to ozone formation. For example, in California, it is estimated that in 1995 use of gas-powered lawn and garden equipment resulted in the state-wide average releases of 75 t day-1 of reactive organic gas (ROG) out of a total release of 3500 t of ROG/day from all sources (19). Even though emissions from lawn and garden equipment are relatively small ROG sources in California, current efforts and legislation are aimed at decreasing them. We suggest that the ROG emissions from the cut grass should also be considered. This is based on the following line of reasoning. Emission factors for lawn tractors are in the range of 120-180 g kW-1 h-1 (http://www.arb.ca.gov/regact); with a typical engine size between 3 and 18 kW, this equates to about 360-3240 g of ROG h-1 of mowing. If we assume that emission factors are proportional to the observed peak concentrations, we get a rough estimate for the biogenic VOC release per hour from cut grass: 12-20% of 360-3240 g h-1 yields 40-650 g h-1. With an average biomass density from lawn clippings of about 40 g m-2 and total emissions from drying lawn clippings obtained in enclosures, the total amount of ROG during the drying phase yields approximately 24-45 g of VOC for a lawn that could be mowed in about 1 h (0.5 acre ) 2022 m2). Thus, the total biogenic ROG emission from a cut lawn can roughly be VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2929

FIGURE 3. (Z)-3-Hexenal released by drying cut samples of Kentucky bluegrass at different temperatures. (Z)-3-Hexenal was detected by PTR-MS at mass 81; water vapor released during drying was measured by mass 37 (H2O‚H3O+).

TABLE 1. Total Emission of Hexenal, Hexenol/Hexanal, Acetaldehyde, and Acetone from Drying Grassesa drying temp (°C) 25 30 35 a

(Z)-3-hexenal plus (E)-2-hexenal (µg/g dw) bluegrass orchard grass 100-240 100-240 30-100

72 nd 31

hexenol plus hexanal (µg/g dw) bluegrass orchard grass 30-60 30-60 5-30

12 nd 7

acetaldehyde (µg/g dw) bluegrass orchard grass 20-80 20-80 20-80

30 nd 20

acetone (µg/g dw) bluegrass orchard grass 20-40 20-40 20-40

16 nd 25

Abbreviations: bluegrass, Poa pratensis; orchard grass, Dactylis glomerata; nd, not determined.

estimated to be in the order of 64-700 g h-1; this is about 20% of the lawn mower exhaust ROG. However, HO reactivities of biogenic, unsaturated aldehydes are in the range of 5e-11 to 1e-10 cm3 s-1 (20, 21); this is much higher that the potential for oxidant formation by mower exhaust (alkanes, alkenes, and arenes). Only a small percentage of the total ROG emission from combustion engines can be assigned to reactive compounds such as alkenes/alkynes (ethene, acetylene, propene, butene, butadiene) with reaction rates between 9e-12 and 3e-11 cm3 s-1 and aromatic compounds with reaction rates between 1e12 cm3 s-1 (benzene) and 1.8e-11 cm3 s-1 (xylenes) (20). If we compare average reactivities based on propylene equivalents as described by Chameides et al. (22), we get a factor of 2-3 for C5 wound VOCs, 1-2 for C6 wound VOCs, and 0.5 for acetaldhyde. Reactivities for alkenes and arenes on the other hand are typically lower yielding 1 for propene, 0.8 for butenes, 0.04 for benzene, 0.23 for toluene, and 0.6 for xylenes. Multiplying these factors with ambient concentrations of each compound gives ∼120 ppbv for wound VOCs and ∼130 ppbv for ROG released by the lawn mower, which may be a more realistic comparison. If so, the reactivity of wound VOCs released from cut grass as a result of lawn mowing is of the same order of magnitude as of unburned hydrocarbons from gasoline-powered lawn mowers, and it is possible that this represents an unrecognized source of urban ROG. 2930

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 14, 2001

Since the major VOCs we detected are reactive aldehydes, including acetaldehyde, (Z)-3-hexenal, and (E)-2-hexenal, it is likely that their release in environments with abundant NOx will contribute to oxidant formation by well-known mechanisms (4). Acetaldehyde is a precursor to ozone and peroxyacetylnitrate formation, and studies of the gas-phase reactivity of (Z)-3-hexenol have shown it to have an atmospheric lifetime of a few hours and to be a precursor of propanal and 3-hydroxypropanal, which in turn give rise to organic nitrates and ozone formation (20, 23). (Z)-3-Hexanal is also a likely precursor of organic nitrates and ozone (R. Atkinson, personal communication, 1998). It may also be significant that cut grass is a source of acetone as revealed by release of the ketone from the wetting of grass clippings during a rain event (Figure 2) and during direct drying of cut grass in a leaf cuvette (Table 1). There is current interest in biogenic sources of acetone, since acetone photolysis is thought to be an important source of HOx radical formation in the upper troposphere (24, 25). Thus, it seems likely that urban lawn mowing will have small impacts on photochemical reactions in the atmospheric boundary layer and also in higher levels of the atmosphere. It remains to be established how significant these impacts are. This will require knowledge of urban lawn cover, the frequency and extent of lawn mowing, the types of grasses used, the quantification of VOCs released during mowing

and subsequent drying of uncollected grass clippings, the modeling of the results with a suitable urban airshed model, etc. The advent of portable PTR-MS instrumentation that has been operated in the field (26) may assist in these experiments. The new measurements presented here show that lawn mowing releases small but significant amounts of reactive VOCs into the air. These biogenic VOCs are known precursors of atmospheric oxidants. Since the fraction of urban grass landcover is of the order of 10-30% (27; A. Guenther, personal communication) and urban lawns are mowed frequently during the growing season, impacts of lawn mowing may need to be included in urban air chemistry models. The rapid releases of (Z)-3-hexenal, (E)-2-hexenal, hexenols plus hexanal, and hexanol after lawn mowing are consistent with known leaf wounding mechanisms, and their continued release after mowing suggest that drying of cut grass is a significant source of additional wound compounds. Release of other VOCs, such as acetaldehyde after cutting and acetone after wetting of cut grass, are also significant. More work on the quantification of VOC release from lawn mowing is needed to evaluate our observations.

Acknowledgments Funding was provided by Fonds zur Fo¨rderung der Wisenschaftlichen Forschung (Project P 12022) and NSF Grant ATM-9805191 (to R.F.). We thank A. Comrie, P. Crutzen, I. Galbally, and A. Guenther for helpful comments. This paper is dedicated to the memory of W.L.

Literature Cited (1) Fehsenfeld, F.; Calvert, J.; Fall, R.; Goldan, P.; Guenther, A.; Hewitt, C.; Lamb, B.; Liu, S.; Trainer, M.; Westberg, H.; Zimmerman, P. Global Biogeochem. Cycles 1992, 6, 389-430. (2) Guenther, A.; Hewitt, C. N.; Erickson, D.; Fall, R.; Geron, C.; Graedel, T.; Harley, P.; Klinger, L.; Lerdau, M.; McKay, W. A.; Pierce, T.; Scholes, B.; Steinbrecher, R.; Tallamraju, R.; Taylor, J.; Zimmerman, P. J. Geophys. Res. 1995, 100, 8873-8892. (3) Singh, H. B.; Kanakidou, M.; Crutzen, P. J.; Jacob, D. J. Nature 1995, 378, 50-54. (4) Atkinson, R.; Arey, J. Acc. Chem. Res. 1998, 31, 574-583. (5) Chameides, W. L.; Lindsay, R. W.; Richardson, J.; Kiang, C. S. Science 1988, 241, 1473-1475. (6) Cardelino, C. A.; Chameides, W. L. J. Geophys. Res. 1990, 95, 13971-13979. (7) McKeen, S. A.; Hsie, E. V.; Trainer, M.; Tallamraju, R.; Liu, S. C. J. Geophys. Res. 1991, 96, 10809-10845.

(8) Friedrich, R.; Schwarz, U.-B. In Urban Air Pollution- European Aspects; Fenger, J., Hertel, O., Palmgren, F., Eds.; Kluwer Academic: Dordrecht, 1998. (9) Ciccioli, P.; Brancaleoni, E.; Frattoni, M. In Reactive Hydrocarbons in the Atmosphere; Hewitt, C. N., Ed.; Academic Press: San Diego, 1999. (10) Hatanaka, A. Phytochemistry 1993, 34, 1201-1218. (11) Kirstine, W.; Galbally, I.; Ye, Y.; Hooper, M. J. Geophys. Res. 1998, 103, 10605-10619. (12) de Gouw, J. A.; Howard, C. J.; Custer, T. G.; Fall, R. Geophys. Res. Lett. 1999, 26, 811-814. (13) Fall, R.; Karl, T.; Hansel, A.; Jordan, A.; Lindinger, W. J. Geophys. Res. 1999, 104, 15963-15974. (14) Lindinger, W.; Hansel, A.; Jordan, A. Int. J. Mass Spectrom. Ion Proc. 1998, 173, 191-241. (15) Lias, S. G.; Bartness, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard, W. G. J. Phys. Chem. Ref. Data 1988, 17 (Suppl. 1), 1-861. (16) Warnecke, C. Ph.D. Thesis, Innsbruck, 1998. (17) Warnecke, C.; Karl, T.; Judmaier, H.; Hansel, A.; Jordan, A.; Lindinger, W., Crutzen, P. Global Biogeochem. Cycles 1999, 13, 9-17. (18) Fall, R.; Benson, A. A. Trends Plant Sci. 1996, 1, 296-301. (19) Alexis, A.; Gaffney, P.; Garcia, C.; Nystrom, M.; Rood, R. The 1999 California air quality and emissions almanac. 1999 (http:// www.arb.ca.gov/aqd/almanac/almanac99.htm). (20) Atkinson, R.; Arey, J.; Aschmann, S. M.; Corchnoy, S. B.; Shu, Y. Int. J. Chem. Kinet. 1995, 27, 941-955. (21) Orlando, J. J.; Tyndall, G. S.; Ceazan, N. J. Phys. Chem. A 2001, 105, 3564-3569. (22) Chameides, W. L.; Fehsenfeld, F.; Rodgers, M. O.; Cardelino, C.; Martinez, J.; Parrish, D.; Lonneman, W.; Lawson, D. R.; Rasmussen, R. A.; Zimmerman, P.; Greenberg, J.; Middleton, P.; Wang, T. J. Geophys. Res. 1992, 97, 6037-6055. (23) Aschmann, S. A.; Shu, Y.; Arey, J.; Atkinson, R. Atmos. Environ. 1997, 31, 3551-3560. (24) Singh, H. B.; Kanakidou, M.; Crutzen, P. J.; Jacob, D. J. Nature 1995, 378, 50-54. (25) McKeen, S. A.; Gierczak, T.; Burkholder, J. B.; Wennberg, P. O.; Hanisco, T. F.; Keim, E. R.; Gao, R.-S.; Liu, S. C.; Ravishankara, A. R.; Fahey, D. W. Geophys. Res. Lett. 1997, 24, 3177-3180. (26) Hansel, A.; Jordan, A.; Warneke, A.; Holzinger, R.; Wisthaler, A.; Lindinger, W. Plasma Sources Sci. Technol. 1999, 8, 332-336. (27) Comrie, A. C.; Diem, J. E. Smogmap. System for management, observation and GIS modeling of air pollution; University of Arizona: 1998.

Received for review February 14, 2001. Accepted April 3, 200120012001. ES010637Y

VOL. 35, NO. 14, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2931