Environ. Sci. Technol. 2007, 41, 6990-6995
Impact of Propene on Secondary Organic Aerosol Formation from m-Xylene C H E N S O N G , †,‡,§ K W A N G S A M N A , † B E T H A N Y W A R R E N , †,‡ Q U E N T I N M A L L O Y , †,‡ A N D D A V I D R . C O C K E R , I I I * ,†,‡ Bourns College of EngineeringsCenter for Environmental Research and Technology (CE-CERT), and Department of Chemical and Environmental Engineering, Bourns College of Engineering, University of California, Riverside, California 92521
Propene is widely used in smog chamber experiments to increase the hydroxyl radical (OH) level based on the assumption that the formation of secondary organic aerosol (SOA) from parent hydrocarbon is unaffected. A series of m-xylene/NOx photooxidation experiments were conducted in the presence of propene in the University of California CECERT atmospheric chamber facility. The experimental data are compared with previous m-xylene/NOx photooxidation work performed in the same chamber facility in the absence of propene (Song et al. Environ. Sci. Technol. 2005, 39, 3143-3149). The result shows that, for similar initial conditions, experiments with propene have lower reaction rates of m-xylene than those without propene, which indicates that propene reduces OH in the system. Furthermore, experiments with propene showed more than 15% reduction in SOA yield compared to experiments in the absence of propene. Additional experiments of m-xylene/ NOx with CO showed similar trends of suppressing OH and SOA formation. These results indicate that SOA from m-xylene/NOx photooxidation is strongly dependent on the OH level present, which provides evidence for the critical role of OH in SOA formation from aromatic hydrocarbons.
Introduction Aromatic hydrocarbons are an important class of volatile organic compounds (VOCs) emitted into the atmosphere, and they are well-known to oxidize in the atmosphere to produce a large variety of pollutants including ozone and secondary organic aerosol (SOA). Anthropogenic emissions, such as vehicular and industrial emissions, are major atmospheric sources of aromatic hydrocarbons. Six of the 25 most abundant nonmethane organic hydrocarbons in Los Angeles were reported to be aromatic species (1). It was reported that aromatic hydrocarbons contribute 19% of atmospheric VOCs in Los Angeles, CA, and 30% of the VOCs * Corresponding author phone: 909-781-5695; fax: 909-781-5790; e-mail:
[email protected]. † Bourns College of EngineeringsCenter for Environmental Research and Technology. ‡ Department of Chemical and Environmental Engineering, Bourns College of Engineering. § Current address: Atmospheric Science and Global Change Division Pacific Northwest National Laboratory. phone: 509-3724985; fax: 509-372-6168. 6990
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 20, 2007
in Boston, MA (2). Smog chamber experiments have confirmed significant secondary organic aerosol (SOA) formation from the photooxidation of alkyl-substituted aromatic hydrocarbons (3-6) in the presence of NOx. It is therefore expected that aromatic hydrocarbons may significantly contribute to urban fine particle loadings in areas impacted by photochemical smog (7-9). The main atmospheric oxidation pathway for aromatic hydrocarbons is reaction with hydroxyl radical (OH) (10). Reaction with ozone (O3) and nitrate radical (NO3) can be significant for aromatic hydrocarbons that contain external alkene groups. OH reacts with aromatic hydrocarbons by abstracting hydrogen atoms from the alkyl group or by adding to the aromatic ring (hydrogen atom abstraction from the aromatic ring is negligible except for benzene). It has been shown that, under atmospheric conditions, addition of OH to the aromatic ring accounts for nearly 90% of the reaction yield, resulting in the formation of a variety of oxidation products from ring-retaining and ring-opening pathways (11-14). A large number of these products contain one or more unsaturated bonds and/or multiple functional groups that are believed to undergo further oxidation to generate SOA. Owing to the complexity of the oxidation products and the limitation of the current measurement techniques, our knowledge about the atmospheric degradation mechanisms of aromatic hydrocarbon is still limited, resulting in difficulty to predict the extent of SOA formation. Generally, less than 50% of aromatic photooxidation products are identified (2). Forstner et al. identified only 15-30% of the mass of SOA collected from several aromatic hydrocarbons (11). Because of its key role in the oxidation of aromatic hydrocarbons, OH needs to be maintained at high enough concentrations to elevate sufficient hydrocarbon oxidation during smog chamber experiments. In SOA experiments carried out at Caltech (4-6, 11, 15), as well as other smog chamber experiments (12), propene (C3H6) was added to be a photochemical initiator and to raise OH levels. In most cases, only parent VOCs with at least six carbon atom numbers are needed to generate oxidized products of sufficiently low vapor pressure to form aerosol (5). Since propene has only three carbon atoms, it is not considered to be an SOA precursor. In the troposphere, propene predominantly reacts with OH, with minor losses through reactions with O3 and NO3. The reaction products, which include formaldehyde, acetaldehyde, and small amounts of organic hydroxynitrates, have vapor pressure too high to exist in the aerosol phase. A recent work from Paulsen et al. (16) shows that propene did not enhance OH levels in the 1,3,5-trimethylbenzene/ NOx photooxidation experiments, nor did it change the particle formation. Song et al. (3) showed that, for the m-xylene/NOx photooxidation system, SOA formation is only slightly dependent on or unaffected by O3 and NO3 levels, with the dominant SOA formation process through reaction of OH with intermediate products. Therefore, if propene alters the OH concentration in the chamber, it may indirectly impact the SOA production from m-xylene. In this present work, we report on a series of smog chamber experiments carried out in the UCR indoor chamber facility (17) in which SOA formation from the black light irradiation of m-xylene, NOx, and propene mixtures was monitored. The results are compared with previous work (3) that examined the NOx effect on SOA formation from m-xylene photooxidation in the absence of propene. Additional black light irradiation experiments of m-xylene, NOx, and CO mixtures are presented to provide additional insight into propene’s effect on SOA formation. 10.1021/es062279a CCC: $37.00
2007 American Chemical Society Published on Web 09/22/2007
TABLE 1. Experimental Conditions experiment
compound
HCo (µg/m3)
∆HC (µg/m3)
Mo (µg/m3)
NOx (ppb)
290B 296B 303A 385A 385B 419A 419B 424A 425A 425B 428A 428B 429B 442A 442B 447B 493A 495B 496B 516B
m-xylene + C3H6 m-xylene + C3H6 m-xylene + CO m-xylene + CO m-xylene + CO m-xylene + CO m-xylene + C3H6 m-xylene + CO m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + C3H6 m-xylene + CO m-xylene + C3H6 m-xylene + CO m-xylene + CO
255.4 1068.4 447.0 212.8 212.8 1106.7 1136.5 625.7 1506.9 1498.3 212.8 221.3 655.5 685.3 702.4 1379.2 395.9 515.1 848.4 685.3
178.8 685.3 276.7 132 174.5 1004.6 1047.1 570.4 1076.9 1349.4 102.2 212.8 473.8 205.6 417.2 659.8 374.6 434.2 741.1 489.5
5.4 39.5 6.4 1.0 0.5 16.0 36.8 11.2 115.9 62.0 2.0 1.4 29.1 3.5 21.6 76.2 6.0 5.5 18.4 27.0
25.6 46.5 43.3 23.2 72.4 499.1 502.4 217.7 146.1 674.4 11.2 113.2 54.5 12.5 58.0 130.5 222.5 224.5 380.0 74.2
Experimental Section The experimental methodology and systems are briefly described in the Supporting Information. Briefly, the CECERT chamber has twin 90 m3 Teflon reactors (2 mil FEP Teflon film) inside a purified-air-flushed and temperaturecontrolled enclosure. In this study, all experiments were conducted using black lights. The Teflon bags were collapsed slowly during the experiment by a rigid motor-drive frame to maintain a slight positive differential pressure between the reactors and enclosure. Prior to each experiment, the reactors have no detectable reactive hydrocarbons (1 ppbC detection limit), NOx (500.0b 408.2 135.5 153.2 266.9 175.6 301.6 >500.0b >500.0b 480.1 280.4 >500.0b 64.8 254.3 152.6 53.7 160.8 201.5 479.6 448.9 489.8 282.5
a Experiments 104A to 249B are from our previous study (3), in which the experiments of m-xylene/NOx photooxidations were performed without propene. b Ozone concentration is out of the range of the ozone analyzer.
as our previous study (experiment numbers 5.5 ppbC:ppb). Experiments in Figure 4 were separated into two groups according to the HC:NOx split point of 5.5 identified in the previous work (HCo:NOx < 5.5 (solid triangles) and others (open circles)). Each group was fitted empirically with a two-product partitioning model (5), providing a convenient way to compare the data with previous researchers. Also displayed in Figure 4 are yield curves from our previous study (3) obtained from photooxidation of m-xylene/NOx without propene. The presence of propene does not alter the effect of NOx levels on the SOA formation from the photooxdiation of m-xylene. Consistent with our previous research, experiments with lower NOx levels (higher HCo:NOx) have considerably higher SOA yield than the group of experiments with higher NOx levels (smaller HCo:NOx). Within the same HCo:NOx regime, the SOA yield from the current study is obviously lower than the yield curves from our previous research, indicating that propene reduces SOA yield. Effect of CO on OH and Secondary Organic Aerosol from m-Xylene/NOx System. Eight m-xylene/NOx photooxidation experiments with CO were carried out in addition to the experiments with propene. CO can simply reduce OH levels without affecting m-xylene oxidation chemistry:
OH + CO(+O2) f CO2 + HO2
(R1)
Figure 5 and Figure 6 display comparisons of OH concentrations and SOA yields between experiments with CO and those with and without propene, respectively. As expected, CO reduces OH concentrations in all of the experiments compared to experiments without CO. In most experiments with CO, OH concentrations are comparable to those from experiments with propene. Similar to the propene effect, SOA yield obtained from experiments with CO (Figure 6) was lower than the yield curves without the presence of CO
FIGURE 5. OH radical concentrations as a function of hydrocarbon to NOx ratio for experiments in the presence and absence of propene and in the presence of CO.
FIGURE 6. Secondary organic aerosol yields for m-xylene/NOx photooxidation in the presence of CO as a function of organic aerosol mass concentration. Data are shown as solid triangles and open circles for experiments in the presence of CO in the low NOx regime and high NOx regime, respectively. Dashed lines and solid lines are secondary organic aerosol yield curves for experiments of m-xylene/ NOx photooxidation in the presence and in the absence of propene, respectively. and agreed fairly well with the fitted yield curves from experiments with propene. A clear trend that SOA formation is a function of OH levels is noted for experiments which include propene and CO as well as from simple m-xylene/NOx photooxidation. A similar trend is observed from one of our recent studies (23). In that study, we examined the SOA formation from the photooxidation of m-xylene in the absence of NOx and found that SOA formation increased with the increasing H2O2. Although both studies utilized different methods to manipulate the OH levels, similar trends are still observed, indicating that propene and CO are likely to change the SOA formation by altering OH levels. During aromatic-NOx photooxidation, a large variety of intermediate products, including glyoxal, methylglyoxal, unsaturated dicarbonyl, and furanone coproducts from the ring-opening pathways and aromatic aldehydes, aromatic acids and phenols from the ringretaining pathways (2, 11, 12, 14) are generated. Some of these intermediate products are expected to be oxidized further by O3, OH, and NO3 to form SOA species. Our previous study (3) showed that O3 and NO3 have only a slight or no effect on SOA formation by comparing SOA formation from photooxidation of m-xylene with varied NOx concentrations, which is also supported by Martin-Reviejo and Wirtz (24). Changing OH level impacts the efficiency of the reaction between intermediate products and OH and subsequently alters SOA formation. However, it should be noted that other reactions, such as aerosol-phase heterogeneous reaction, might also participate in the production of SOA and be VOL. 41, NO. 20, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6993
affected by the presence of propene, CO, and their oxidation products. For instance, propene and CO may change the peroxy radical to HO2 radical ratio, thereby perturbing the formation of organic peroxides, which has been speculated as an important precursor of SOA compounds formed in the aerosol phase.
Comparison with Previous Studies Currently, the available SOA yield data from m-xylene/NOx photooxidation are mostly from Odum et al. (5, 6) and Cocker et al. (4). Odum et al.’s experiments were carried out in the Caltech outdoor smog chamber while Cocker et al. obtained their data from the Caltech indoor smog chamber. In both studies, propene was added to initiate and facilitate hydroxyl formation. Our previous study (3) pointed out that Cocker et al.’s SOA yields were smaller than our study, even though the Cocker et al.’s data were measured at a lower temperature. It was previously explained in that study that the elevated aerosol formation at lower temperatures was offset by the decline of aerosol production at high NOx levels. Observations within this study show that the effect of propene should be taken into account as well. During Odum et al.’s experiments, temperatures were much higher than those of the present study and Cocker et al.’s study. Thus, it is not appropriate to directly compare Odum et al.’s data set to examine the propene effect. In the same study, Odum first suggested that SOA yields derived from individual hydrocarbons could be used to predict contribution of a single hydrocarbon to total SOA formed from a mixture, which was supported by experimental evidence from photooxidation of m-xylene/ R-pinene and m-xylene/1,3,5-trimethylbenzene mixtures. This can be true only if the SOA-related chemical mechanisms of individual hydrocarbons are unaffected by each other in these mixtures. However, as shown in this study, even simple non-SOA-forming compounds such as CO and propene could change SOA formation from m-xylene by altering OH levels. In addition, recent studies (25-28) have demonstrated in alkene-O3 reactions that OH scavengers (such as 2-butanol, CO, propanol, and cyclohexane) have significant impact on SOA yields by obviously varying the RO2/HO2 ratios. The photochemical mechanisms of most major ambient VOCs are still far from complete, and therefore caution should be taken when estimating aerosol formation from mixtures as a simple superposition of single compounds. In the m-xylene/ R-pinene system, for instance, R-pinene may act like propene to reduce OH concentration, which could reduce SOA yield from m-xylene. On the other hand, RO2/HO2 ratio may also be different in the mixture system compared to that of photooxidation of R-pinene alone, which could lead to changes of SOA yield from R-pinene (29).
Implication We report herein a series of experiments showing that propene reduced OH concentrations and SOA yields during the atmospheric photooxidation of m-xylene. This is the first attempt to investigate the effect of non-SOA-forming hydrocarbon on SOA formation from aromatic hydrocarbons. Our experimental data indicate that varied OH concentrations may change the relative competition for SOA-related precursors (second or higher order generation intermediate products) among sub-branches of the gas-phase photooxidation chemistries, finally leading to different SOA production. The currently most used SOA model (5) is based on absorptive partitioning theory and does not take into account the OH impact, which could introduce significant uncertainty in SOA prediction. Previous measurements (30, 31) have shown diurnal variation of OH with the maximum concentration in the afternoon and minimum in the early morning and late night. OH concentrations also vary significantly with different 6994
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 20, 2007
locations, seasons, and sunlight radiations. SOA yields of major aromatic hydrocarbons obtained in the smog chamber while propene is present generally represent scenarios with fairly low OH levels. Using parameters generated from such data sets may underpredict SOA when the OH level is high and overpredict SOA when the OH level is very low. Since OH concentration fluctuates with hydrocarbon and NOx levels in classic smog chamber experiments, it is nearly impossible to use this kind of experiment to quantify OH effect on SOA production. A recent study (24) of exploring SOA production from benzene may provide a possible way to evaluate SOA production potential under different OH levels by keeping NOx concentration constant during the photooxidation period.
Acknowledgments The authors wish to acknowledge the National Science Foundation for financial support (ATM-0234111 and ATM0449778). The authors would also like to thank Kurt Bumiller, Irina Malkina, Lindsay Yee, Christina Zapata, and Jean Wang for experimental help with setup and measurement and William P. L. Carter for useful discussions.
Supporting Information Available Experimental section, additional information, and figures. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Lurmann, F. W.; Main, H. H. Analysis of the ambient VOC data collected in the Southern California Air Quality Study. Final Report. ARB Contract No. A832-130. California Air Resource Board: Sacramento, CA, 1992. (2) Calvert, J. G.; Atkinson, R.; Becker, K. H.; Kammens, R. M.; Seinfeld, J. H.; Wallington, T. J.; Yarwood G. The Mechanisms of Atmospheric Oxidation of Aromatic Hydrocarbons; Oxford University Press: London, 2002. (3) Song, C.; Na, K.; Cocker, D. R., III. Impact of the hydrocarbon to NOx ratio on secondary organic aerosol formation. Enivron. Sci. Technol. 2005, 39, 3143-3149. (4) Cocker, D. R., III; Mader, B. T.; Kalberer, M.; Flagan, R. C.; Seinfeld, J. H. The effect of water on gas-particle partitioning of secondary organic aerosol: II. m-xylene and 1,3,5-trimethylbenzene photooxidation systems. Atmos. Environ. 2001, 35, 6073-6085. (5) Odum, J. R.; Hoffmann, T.; Bowman, F.; Collins, D.; Flagan, R. C.; Seinfeld, J. H. Gas/particle partitioning and secondary organic aerosol yields. Environ. Sci. Technol. 1996, 30, 25802585. (6) Odum, J. R.; Jungkamp, T. P. W.; Griffin, R. J.; Forstner, H. J. L.; Flagan, R. C.; Seinfeld, J. H. Aromatics, reformulated gasoline, and atmospheric organic aerosol formation. Environ. Sci. Technol. 1997, 31, 1890-1897. (7) Na, K.; Swant, A. A.; Song, C.; Cocker, D. R., III. Primary and secondary carbonaceous species in the atmosphere of Western Riverside County, California. Atmos. Environ. 2004, 38, 13451355. (8) Turpin, B. J.; Huntzicker, J. J. Secondary formation of organic aerosol in the Los-Angeles basinsa descriptive analysis of organic and elemental carbon concentrations. Atmos. Environ. 1991, 25A (No. 2), 207-215. (9) Turpin, B. J.; Huntzicker, J. J. Identification of secondary organic aerosol episodes and quantitation of primary and secondary organic aerosol concentrations during SCAQ. Atmos. Environ. 1995, 29, 3527-3544. (10) Atkinson, R. J. Atmospheric chemistry of VOCs and NOx. Atmos. Environ. 2000, 34, 2063-2101. (11) Forstner, H. J. L.; Flagan, R. C.; Seinfeld, J. H. Secondary organic aerosol from the photooxidation of aromatic hydrocarbons: molecular composition. Environ. Sci. Technol. 1997, 31, 13451358. (12) Jang, M.; Kamens, R. M. Characterization of secondary aerosol from the photooxidation of toluene in the presence of NOx and 1-propene. Environ. Sci. Technol. 2001, 35, 3626-3639.
(13) Kleindienst, T. E.; Conver, T. S.; McIver, C. D.; Edney, E. O. Determination of secondary organic aerosol products from the photooxidation of toluene and their implications in ambient PM2.5. J. Atmos. Chem. 2004, 47, 79-100. (14) Smith, D. F.; Kleindienst, T. E.; McIver, C. D. Primary product distributions from the reaction of OH with m-, p-xylene, 1,2,4and 1,3,5-trimethylbenzene. J. Atmos. Chem. 1999, 34, 339364. (15) Griffin, R. J.; Cocker, D. R., III; Flagan, R. C.; Seinfeld, J. H. Organic aerosol formation from the oxidation of biogenic hydrocarbons. J. Geophys. Res. 1999, 104, 3555-3567. (16) Paulsen, D.; Dommen, J.; Kalberer, M.; Prevot, A. S. H.; Richter, R.; Sax, M.; Steinbacher, M.; Weingartner, E.; Baltensperger, U. Secondary organic aerosol formation by irradiation of 1,3,5trimethylbenzene-NOx-H2O in a new reaction chamber for atmospheric chemistry and physics. Enivron. Sci. Technol. 2005, 39, 2668-2678. (17) Carter, W. P. L.; Cocker, D. R., III; Fitz, D. R.; Malkina, I. L.; Bumiller, K.; Sauer, C. G.; Pisano, J. T.; Bufalino, C.; Song, C. A new environmental chamber for evaluation of gas-phase chemical mechanisms and secondary aerosol formation. Atmos. Environ. 2005, 39, 7768-7788. (18) Rickard, A. R.; Johnson, D.; McGill, C. D.; Marston, G. OH yields in the gas-phase reactions of ozone with alkenes. J. Phys. Chem. A 1999, 103, 7656-7664. (19) Atkinson, R.; Aschmann, S. M. Hydroxyl radical production from the gas-phase reactions of ozone with a series of alkenes under atmospheric conditions. Environ. Sci. Technol. 1993, 27, 13571363. (20) Spivakovsky, C. M.; Yevich, R.; Logan, J. A.; Wofsy, S. C.; McElroy, M. B.; Prather, M. J. Tropospheric OH in a 3-dimensional chemical tracer modelsan assessment based on observations of CH3CCl3. J. Geophys. Res. 1990, 95, 18441-18471. (21) Poppe, D.; Wallasch, M.; Zimmermann, J. The dependence of the Concentration of OH on its precursors under moderately polluted conditions: A model study. J. Atmos. Chem. 1993, 16, 61-78. (22) Jeffries, H. E.; Tonnesen, S. A Comparison of two photochemical reaction mechanisms using mass balance and process analysis. Atmos. Environ. 1994, 28, 2991-3003.
(23) Song, C.; Na, K.; Warren, B.; Malloy, Q.; Cocker, D. R. Secondary organic aerosol formation from m-xylene in the absence of NOx. Environ. Sci. Technol. 2007, in press. (24) Martin-Reviejo, M.; Wirtz, K. Is benzene a precursor for secondary organic aerosol? Environ. Sci. Technol. 2005, 39, 1045-1054. (25) Docherty, K. S.; Ziemann, P. J. Effects of stabilized criegee intermediate and OH radical scavengers on aerosol formation from reactions of β-pinene with O3. J. Aerosol Sci. Technol. 2003, 37, 877-891. (26) Docherty, K. S.; Ziemann, P. J. Contributions of organic peroxides to secondary aerosol formed from reactions of monoterpenes with O3. Environ. Sci. Technol. 2005, 39, 4049-4059. (27) Keywood, M. D.; Kroll, J. H.; Varutbangkul, V.; Bahreini, R.; Flagan, R. C.; Seinfeld, J. H. Secondary organic aerosol formation from cyclohexene ozonolysis: effect of OH scavenger and the role of radical chemistry. Enivron. Sci. Technol. 2004, 38, 33433350. (28) Ziemann, P. J. Formation of alkoxyhydroperoxy aldehydes and cyclic peroxyhemiacetals from reactions of cyclic alkenes with O3 in the presence of alcohols. J. Phys. Chem. A 2003, 107, 20482060. (29) Kamens, R.; Jang, M.; Chien, C. J.; Leach, K. Aerosol formation from the reaction of alpha-pinene and ozone using a gas-phase kinetics aerosol partitioning model. Enivron. Sci. Technol. 1999, 33, 1430-1438. (30) Ren, X. R.; Harder, H.; Martinez, M.; Lesher, R. L.; Oliger, A.; Shirley, T.; Adams, J.; Simpas, J. B.; Brune, W. H. HOx concentrations and OH reactivity observations in New York City during PMTACS-NY2001. Atmos. Environ. 2003, 37, 36273637. (31) George, L. A.; Hard, T. M.; O’Brien, R. J. Measurement of free radicals OH and O2 in Los Angeles smog. J. Geophys. Res. 1999, 104, 11643-11655.
Received for review September 23, 2006. Revised manuscript received July 2, 2007. Accepted July 11, 2007. ES062279A
VOL. 41, NO. 20, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6995
