Effect of NOx on Secondary Organic Aerosol Concentrations

Jul 12, 2008 - The secondary organic aerosol (SOA) module in PMCAMx, a three-dimensional chemical transport model, has been updated to incorporate NOx...
0 downloads 7 Views 6MB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 6022–6027

Effect of NOx on Secondary Organic Aerosol Concentrations TIMOTHY E. LANE,† N E I L M . D O N A H U E , †,‡ A N D S P Y R O S N . P A N D I S * ,†,§ Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, Department of Chemistry, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213, and Department of Chemical Engineering, University of Patras, Patra, 26500, Greece

Received December 22, 2007. Revised manuscript received April 30, 2008. Accepted June 9, 2008.

The secondary organic aerosol (SOA) module in PMCAMx, a three-dimensional chemical transport model, has been updated to incorporate NOx-dependent SOA yields. Under low-NOx conditions, the RO2 radicals react with other peroxy radicals to form a distribution of products with lower volatilities, resulting in higher SOA yields. At high-NOx conditions, the SOA yields are lower because aldehydes, ketones, and nitrates dominate the product distribution. Based on recent laboratory smog chamber experiments, high-NOx SOA parametrizations were created using the volatility basis-set approach. The organic aerosol (OA) concentrations in the Eastern US are simulated for a summer episode, and are compared to the available ambient measurements. Changes in NOx levels result in changes of both the oxidants (ozone, OH radical, etc.) and the SOA yields during the oxidation of the corresponding organic vapors. The NOx dependent SOA parametrization predicts a maximum average SOA concentration of 5.2 µg m-3 and a domain average concentration of 0.6 µg m-3. As the NOx emissions are reduced by 25%, the domain average SOA concentration does not significantly change, but the response is quite variable spatially. However, the predicted average SOA concentrations increase in northern US cities by around 3% but decrease in the rural southeast US by approximately 5%. A decrease of the average biogenic SOA by roughly 0.5 µg m-3 is predicted for the southeast US for a 50% reduction in NOx emissions.

1. Introduction The yields of condensable products from the oxidation of volatile organic compounds (VOC) show substantial variation as the VOC/NOx ratio changes (1–7). One of the most commonly studied VOC oxidation reactions is the ozonolysis of R-pinene (3, 4, 8–14). Ozonolysis is an important atmospheric sink for R-pinene with measured secondary organic aeresol (SOA) yields as high as 60% (4, 8, 13). Measured SOA yields from the ozonolysis of R-pinene are lower in the presence of high-NOx conditions than for low-NOx conditions (2, 5). Recent experiments for the photooxidation of benzene and m-xylene showed that more VOC is needed to react in * Corresponding author phone: (412) 268-3531; fax: (412) 2687139; e-mail: [email protected]. † Department of Chemical Engineering, Carnegie Mellon University. ‡ Department of Chemistry, Carnegie Mellon University. § Department of Chemical Engineering, University of Patras. 6022

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008

order to initiate SOA formation as the NO concentrations increases (15–17). Johnson et al. (18) suggested that the SOA yields from the oxidation of toluene increase as the NOx concentrations decrease because the organic hydroperoxides react with aldehydes in the aerosol phase to form products with lower volatilities. More recently, Ng et al. (7) showed that the SOA yields for toluene, benzene, and m-xylene are dependent on the NOx concentration and that the SOA yields for these VOCs are higher than previously observed. Most chemical transport models do not account for the effect of NOx on the SOA yields during the oxidation of VOCs. In a recent study by Tsigaridis and Kanakidou (19), the NOx emissions in a future scenario are much higher than present day emissions, causing an increase in ozone and SOA production. Tsigaridis and Kanakidou (19) examined the importance of SOA in the future atmosphere scaling the dependence of biogenic SOA yields on the VOC/NOx ratio according to existing isoprene SOA smog chamber results. In this study, NOx dependent SOA yields were included in a regional three-dimensional chemical transport model, PMCAMx (20, 21), extending the model of Lane et al. (22) that assumed NOx-independent SOA yields. The results of the revised model are compared to the daily average observed OA concentrations from the EPA Speciation Trends Network (STN) and the Interagency Monitoring of Protected Visual Environments (IMPROVE) for July 12-28, 2001. The effect of NOx emission reductions on the predicted SOA levels is quantified.

2. Modeling the NOx Effect on SOA Yields in PMCAMx PMCAMx is a three-dimensional chemical transport model which uses the framework of CAMx (20–23). For this study, PMCAMx is applied to a seventeen day period in the eastern and central United States starting on July 12, 2001. The modeling domain covers a 3492 × 3240 × 6 km region with a 36 × 36 km grid resolution. Anthropogenic emissions are derived from the 1999 National Emissions Inventory Version 3, projected to 2001 using updated data from the MOBILE6 model (24, 25). Biogenic emissions are generated using the Biogenic Emissions Inventory System version 3.13, and MM5 is used to generate the meteorological inputs (20, 26, 27). In this study, sesquiterpene emissions have been added to the above biogenic emissions, assuming that they are equal to 0.3 times the monoterpenes emissions (22, 28). The SAPRC99 chemistry mechanism was used as the gas-phase chemistry mechanism within PMCAMx (29, 30). Additional details about PMCAMx and the results of its evaluation for this episode can be found in Gaydos et al. (20) and Karydis et al. (31). The primary organic aerosol is assumed to be nonvolatile and inert. The aerosol yield or aerosol mass fraction (AMF) is defined as the ratio of the produced SOA concentration to the concentration of the reacted precursor (both in µg m-3). If one assumes that the various compounds form a pseudoideal solution in the particulate phase (21, 22, 32–36), then the yield can be calculated as: yield )

∑ R ( 1 + c1/C i

i

/ i

OA

)

(1)

where Riis the mass-based stoichiometric yield for product i and ci/ is the effective saturation concentration in µg m-3 (33). Odum et al. (33) demonstrated that the SOA yields measured in smog chamber experiments agree well with eq 1, if one uses two surrogate species. The stoichiometric yields and saturation concentration of these two species 10.1021/es703225a CCC: $40.75

 2008 American Chemical Society

Published on Web 07/12/2008

(four parameters) can be estimated by fitting the measurements to eq 1. An alternative approach is the splitting of the volatility range of the SOA components into predetermined bins (saturation concentrations ci*) and the definition of the corresponding surrogate species. Different basis-sets of saturation concentrations can be used, for example using four (1, 10, 100, 1000 µg m-3 at 300 K) or seven (0.01, 0.1, 1, 10, 100, 1000, 104 µg m-3 at 300 K) surrogate species, depending on the required accuracy and the type of environment simulated (22). The corresponding stoichiometric yields are then calculated by fitting eq 1 to the experimental data. The basis-set approach allows the model to cover accurately a wider SOA concentration range (5), increases its computational efficiency by lumping species according to volatility (22), and can easily accommodate descriptions of the chemical aging of organic aerosol (32). NO reacts with organo-peroxy radicals (RO2) to form a distribution of oxidation products under high NOx conditions. Under low NOx conditions, a different distribution of products forms from the reaction of the RO2 radicals with other peroxy radicals (RO2 and HO2). Since the distribution of products from each pathway have different product vapor pressures, the amount of NOx present will influence the SOA yields. Since the basis set approach is used in this study for the saturation concentrations, only the stoichiometric mass yields are affected (2, 22, 32). The enthalpy of vaporization is assumed to be independent of the NOx level due to the lack of experimental data at highNOx conditions. The different product distributions formed from the oxidation of a VOC under both low and high-NOx conditions are described by: VOC + oxidant f R1,lowP1 + R2,lowP2 + R3,lowP3 + R4,lowP4 (low NOx)(2) VOC + oxidant f R1,highP1 + R2,highP2 + R3,highP3 + R4,highP4 (high NOx)(3) For the purposes of this study we define the low NOx regime as the conditions for which VOC/NOx > 10 ppbC/ ppb and the high NOx conditions as VOC/NOx< 3 ppbC/ ppb. The same products P1 to P4 are used for both pathways in the volatility basis-set approach. However, the stoichiometric mass yields that determine the product distributions in these two pathways are different. The overall mass yields, Ri, are combinations of the low- and high-NOx mass yields: Ri ) Ri,highB + Ri,low(1 - B)

(4)

where B is the branching ratio, which determines the fraction of organo-peroxy radicals that react with NO (the high-NOx pathway) given by:

B)

(rate of RO2 + NO) (rate of RO2 + NO) + (rate of RO2 + RO2) + (rate of RO2 + HO2) (5)

When the branching ratio is zero (low-NOx), the organoperoxy radicals (RO2) react with other peroxy radicals (RO2 + HO2), whereas the organo-peroxy radicals react with NO at a branching ratio of one. Table 1 lists the parametrizations for the oxidation of each SAPRC99 precursor (details can be found in (22)) under lowNOx and high-NOx conditions, and Figure 1 compares the low and high-NOx fits for the lumped aromatics (ARO1 and ARO2), monoterpenes (TERP), and isoprene (ISOP). ARO1 is composed mainly of toluene, benzene, and ethylbenzene, while ARO2 consists of xylenes, trimethylbenzenes, ethyltoluene, and higher aromatics. Descriptions of each lumped VOC can be found in Lane et al. (22) or Carter (29). The high anthropogenic yields for ARO1 and ARO2 presented in Lane et al. (22) are used as the low-NOx parametrization for these two VOCs. OLE1 contains all the terminal alkenes, OLE2 consists of all the internal alkenes and cyclic alkenes, and SESQ contains all the sesquiterpenes. The low-NOx parameters for all the other VOCs were the same as in (22). The Lane et al. (22) study, in contrast to this work, used NOxindependent SOA yields for both the baseline and all the sensitivity tests. In Table 1, the parametrizations assume a density of 1 g cm-3 (normalized yield). In this work, the predicted SOA concentrations are multiplied by 1.5 to account for an SOA density of 1.5 g cm-3 (37). Similar to Lane et al. (22), all the SOA products are assumed to have an effective enthalpy of vaporization of 30 kJ mol-1 (5). The high-NOx parametrizations were calculated using the same ratio measured by Pathak et al. (5) for the decrease in SOA yields from the ozonolysis of R-pinene under low and high-NOx conditions. Pathak et al. (5) found that the SOA yields for R-pinene under high-NOx conditions are 69%, 45%, and 37% less than the low-NOx yields at 1, 10, and 100 µg m-3 of organic aerosol, respectively. For the lumped alkanes (ALK4 and ALK5), which have one-product fits, the high-NOx parametrization predicts SOA yields that are 50% lower than the low-NOx yields for all organic aerosol concentrations. The sesquiterpenes are the only VOC which does not have NOx-dependent yields in this model due to the high uncertainties of the corresponding emissions and SOA yields. Ng et al. (38) recently reported that the sesquiterpenes longifolene and aromadendrene have higher SOA yields at high NOx conditions, the opposite behavior from the rest of the SOA precursors. The sensitivity of the predicted SOA concentrations to these choices of the NOx-dependence of the corresponding yields is discussed in the next section.

TABLE 1. High- and Low-NOx SOA Mass Yields for Each Lumped VOC Species within the SAPRC Chemistry Mechanism Using a Four-Product Basis Set Fit with Saturation Concentrations Set to 1, 10, 100, and 1000 µg m-3 at 300 K high-NOx parameterizationa

ALK4 ALK5 OLE1 OLE2 ARO1 ARO2 ISOP SESQ TERP a

low-NOx parameterizationa

1

10

100

1000

1

10

100

1000

0.0 0.0 0.0001 0.0004 0.002 0.001 0.0002 0.05 0.008

0.005 0.050 0.001 0.0035 0.11 0.13 0.015 0.1 0.081

0.0 0.0 0.008 0.014 0.2 0.2 0.010 0.5 0.134

0.0 0.0 0.030 0.055 0.29 0.29 0.0 0.0 0.338

0.0 0.0 0.001 0.003 0.05 0.05 0.006 0.05 0.072

0.01 0.10 0.002 0.006 0.15 0.2 0.02 0.1 0.061

0.0 0.0 0.012 0.023 0.25 0.25 0.01 0.5 0.239

0.0 0.0 0.045 0.076 0.35 0.35 0.0 0.6 0.405

The yield that corresponds to these values assumes a density equal to 1 g cm-3 (normalized yield).

VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6023

FIGURE 2. Predicted average total SOA concentrations (µg m-3) using the (a) NOx-dependent SOA parametrization and (b) the low-NOx (B ) 0) parametrization during July 12-28, 2001.

FIGURE 1. Comparison of low-NOx (solid) and high-NOx (dashed) basis set fits for (a) ARO1 (black) and ARO2 (blue) and (b) TERP (blue) and ISOP (black).

3. Results and Discussion Figure 2 shows the average predicted total SOA concentrations across the eastern US using the NOx-dependent and low-NOx (B ) 0) SOA yields during July 12-28, 2001. Most existing chemical transport models (20, 30, 31, 35) use yields measured during experiments where the precursor organic vapor reacted with ozone or hydroxyl radicals without any NOx present. Therefore, they use yields similar to the lowNOx case discussed here. For the low-NOx case, the predicted domain average SOA concentration is 0.95 µg m-3, with a maximum average of 7.2 µg m-3 over Atlanta, GA. The NOxdependent SOA yields decrease the predicted domain average SOA concentration to 0.63 µg m-3, with a maximum average of 5.2 µg m-3. Figure 3 shows the percent decrease in total SOA across the eastern US when simulating NOx-dependent SOA yields compared to just low-NOx parametrizations. In the southeast US, the SOA concentrations decrease by 20-30%, while the SOA concentrations in the northeast decrease by 40-50%. The larger decrease in the northeast US is due to higher NOx concentrations, which decrease the SOA yields. These differences between the predictions of the models using the NOx-dependent and NOx-independent yields can also be viewed as a first approximation of the sensitivity of the SOA concentrations to the new yield parametrizations. Figure 4 compares the daily average OA concentrations predicted from the NOx-dependent model and the measured OA from the STN and IMPROVE sites (the measured OC was converted to OA by multiplying by 1.4). For the mostly rural IMPROVE sites, the results have little bias (fractional bias equal to 0.05) but a fractional error of 0.45. The performance 6024

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008

FIGURE 3. Predicted percent change in total SOA concentrations when comparing the results of the NOx-dependent SOA yields with the low-NOx parametrizations during July 12-28, 2001. of the current model compared to the IMPROVE measurements is better than that of Lane et al. (22) and is comparable to that of Karydis et al. (31). Similar to these two studies, the current model underpredicts the urban OA concentrations. The bias, error, fractional bias, and fractional error for the blank-corrected OA concentrations at STN sites are -1.62 µg m-3, 2.13 µg m-3, -0.33, and 0.54, respectively. If the IMPROVE OC is instead multiplied by a molecular weight to carbon weight ratio of 1.8 (39–42), the bias between the NOx-dependent SOA model and the measurements decreases to -0.57 µg m-3. In that case, the OA in both the urban and rural areas would be underpredicted. As the NOx emissions are decreased, the ozone and OH radical concentrations decrease since most of the modeling domain is in the NOx-limited regime (42). Across the entire domain, the average ozone concentration decreases by 2.2

FIGURE 4. Comparison of the predicted daily average OA concentrations (µg m-3) between the NOx-dependent SOA model and the sites from the IMPROVE and STN monitoring networks. Dashed lines represent an error of (50%. and 5.4 ppb as the NOx emissions are decreased by 25% and 50%, respectively. The largest decreases in ozone occur in the southeastern US, an area where a large fraction of the SOA is predicted to form from the ozonolysis of the monoterpenes. Reducing the NOx emissions by 25% and 50% slightly lowered the domain average SOA concentrations from 0.63 µg m-3 to 0.62 and 0.61 µg m-3, respectively. Figure 5 shows the differences between the average total SOA concentrations predicted with 25% and 50% NOx emission reductions and the corresponding predictions of the NOx-dependent SOA model. As the NOx emissions decrease, the total SOA concentrations decrease by 0.2-0.3 µg m-3 in the southeast with a 25% reduction and by 0.4-0.6 µg m-3 with a 50% reduction of the NOx emissions. The SOA concentrations in

many of the urban areas across the eastern US are predicted to increase by 0.0-0.2 µg m-3 with 50% reductions in NOx emissions. The SOA concentration changes for a reduction of the NOx emissions by 50% are less than 10% and can be either positive or negative. These changes are non-negligible, but they are relatively modest. An increase of SOA by 0.1-0.2 µg m-3 is predicted for the oceanic regions within a few hundred kilometers from the coast (Figure 5a and 5b). The average B values for the NOx-dependent simulations, as well as the change in the average branching ratio value as the NOx emissions are reduced by 25% and 50% respectively are displayed in Figure 6. Across the entire domain, lower NOx emissions decrease the B value, increasing the predicted SOA yields. The domain average B value decreased from 0.58 to 0.55 and 0.51 with NOx emissions reductions of 25% and 50%, respectively. The maximum average B value reduction was 14% and 31% with NOx emissions lowered by 25% and 50%. While the B values are not greatly reduced, the SOA concentrations would still increase if the oxidant concentrations remained the same. As shown in Figure 5, any decrease in the average SOA concentration is caused by a reduction of the oxidant concentrations. Although the SOA yields slightly increase with a NOx reduction, the decrease in oxidant concentration is more significant than the increase in SOA yields in the southeastern US. In the Central Plains and Midwest, the SOA concentrations remain relatively constant or slightly increase because the increases in the SOA yields are offset by the small decreases in oxidant concentrations. The largest increases in average SOA concentrations as the NOx emission are reduced are located in urban areas in the north. The results of the simulations with reduced NOx emissions for specific sites in the domain are shown in Table 2. For the rural areas listed in Table 2, the average percent change in the SOA concentrations are -5.2% and -7.4% for the south and -0.9% and -1.2% for the north, as the NOx emissions are reduced by 25% and 50%, respectively. In the northern

FIGURE 5. The average difference in total SOA concentrations (µg m-3) predicted as the NOx emissions are reduced by: (a) 25% and (b) 50%. A positive value corresponds to an increase of the SOA. Also shown the percent change in total SOA concentrations as the NOx emissions are reduced by (c) 25% and (d) 50%. VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6025

TABLE 2. Predicted Average Total SOA Concentrations at Rural and Urban Locations across the Eastern US SOA concentrations (µg m-3)

location

NOx emissions reduced by 25%

NOx emissions reduced by 50%

5.22 3.84 1.90 2.34 0.31

5.03 3.71 1.82 2.22 0.34

4.69 3.45 1.69 2.06 0.38

2.73 2.71 2.66 2.88 4.62

2.61 2.59 2.52 2.71 4.34

2.45 2.42 2.33 2.50 3.96

1.65 0.96 1.19 1.13 0.73 1.09 0.81

1.66 1.00 1.22 1.20 0.74 1.13 0.82

1.67 1.04 1.24 1.29 0.75 1.17 0.84

0.52 0.75 0.90 0.64 0.73

0.51 0.74 0.89 0.63 0.73

0.50 0.73 0.88 0.63 0.74

base case South

urban Atlanta, GA Birmingham, AL Gulfport, MS Penscola, FL Houston, TX rural Yorkville, GA Centreville, AL Oak Grove, MS OLF#8, FL south-central Arkansas

North urban St. Louis, MO Philadelphia, PA Chicago, IL NYC, NY Detroit, MI Boston, MA Pittsburgh, PA rural Potsdam, NY North-central PA Acadia National Park, ME Kalamazoo, MI Southeast Iowa

FIGURE 6. The average predicted (a) branching ratio (B) during the NOx-dependent SOA yield simulation and the difference between B and the predicted B values when the NOx emission are reduced by (b) 25% and (c) 50%. cities, the SOA concentrations increase on average by 2.8% and 3.2% as the NOx emissions are reduced by 25% and 50%. Over Houston, the average SOA concentration increases by 9.8% and 12.8% with 25% and 50% reductions in the NOx emissions. Most of the urban areas in the modeling domain are VOClimited (that is NOx-saturated). As a result, the ozone and OH radical concentrations increase as the NOx emissions are reduced (43). Since the NOx concentrations in urban areas are high, reduction of the NOx emissions by 25% and 50% does not significantly shift the SOA yields toward the lowNOx regime, and the SOA yields do not change significantly (Figure 6). The urban cities from the Southeastern Aerosol Research and Characterization (SEARCH) network have lower SOA percent decreases compared to their corresponding rural sites as the NOx emissions decrease (AtlantasYorkville, BirminghamsCentreville, GulfportsOak Grove, Penscolas OLF#8). The average percent change in SOA concentrations as the NOx emissions are reduced 25% for the urban and rural SEARCH sites are -4.2% and -5.0%. In the northern urban cities and Houston, both the ozone and OH radical concentrations increased (Tsimpidi et al., 2007). The higher NOx-dependent SOA yields and oxidant concentrations lead to higher predicted SOA concentrations. As NOx emissions are reduced, three different scenarios may occur depending on the current atmospheric conditions. In many urban areas the NOx-dependent SOA yields will either increase or remain relatively constant, but an increase in oxidant concentration will produce an increase in SOA. In 6026

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 16, 2008

rural NOx-limited areas, decreasing the NOx emissions will reduce the SOA concentrations by reducing the amount of oxidants present. Here, the NOx-dependent SOA yields do not significantly increase. In other areas, the SOA concentrations will remain relatively constant as the NOx emissions are reduced because the increase in the NOx-dependent SOA yields are offset by the decrease in oxidant concentrations. The results of the present study are based on the relatively simple parametrization of the SOA yield dependence on NOx levels using the branching ratio B (eq 3) and the available low and high-NOx smog chamber yield measurements for monoterpenes and isoprene. These simplifications together with the uncertainties about the NOx-dependence of the aromatic VOCs and sesquiterpenes introduce significant uncertainties in the PMCAMx predictions discussed here. Using the sensitivity of the model to the parametrization, we estimate that these uncertainties are probably in the 20-50% range. Additional experiments and corresponding theoretical analyses are needed to reduce these uncertainties.

Acknowledgments The authors thank the EPA for providing the emission files used in this study. This research was supported by the EPA STAR program through the National Center for Environmental Research (NCER). This paper has not been subject to EPA’s required peer and policy review and therefore does not necessarily reflect the views of the Agency. No official endorsement should be inferred.

Literature Cited (1) Pandis, S. N.; Paulson, S. E.; Seinfeld, J. H.; Flagan, R. C. Aerosol formation in the photooxidation of isoprene and β-pinene. Atmos. Environ. 1991, 25A, 997–1008. (2) Presto, A. A.; Huff Hartz, K. E.; Donahue, N. M. Secondary organic aerosol production from terpene ozonolysis: 2. Effect of NOx

concentration. Environ. Sci. Technol. 2005, 39, 7046–7054. (3) Presto, A. A.; Donahue, N. M. Investigation of r-Pinene + Ozone Secondary Organic Aerosol Formation at Low Total Aerosol Mass. Environ. Sci. Technol. 2006, 40, 3536–3543. (4) Pathak, R. K.; Stanier, C.; Donahue, N.; Pandis, S. N. Ozonolysis of alpha-pinene at atmospherically relevant concentrations: temperature dependence of aerosol mass fractions (yields). J. Geophys. Res. 2007, 112, D03201, doi:, 10.1029/2006JD007436. (5) Pathak, R. K.; Presto, A. A.; Lane, T. E.; Stanier, C. O.; Donahue, N. M.; Pandis, S. N. Ozonolysis of R-pinene: parameterization of secondary organic aerosol mass fraction. Atmos. Chem. Phys. 2007, 7, 1–11. (6) Kroll, J. H.; Ng, N. L.; Murphy, S. M.; Flagan, R. C.; Seinfeld, J. H. Secondary organic aerosol formation from isoprene photooxidation. Environ. Sci. Technol. 2006, 40, 1869–1877. (7) Ng, N. L.; Kroll, J. H.; Chan, A. W. H.; Chhabra, P. S.; Flagan, R. C.; Seinfeld, J. H. Secondary organic aerosol formation from m-xylene, toluene, and benzene. Atmos. Chem. Phys. Discuss. 2007, 7, 4085–4126. (8) Hoffmann, T.; Odum, J. R.; Bowman, F.; Collins, D.; Klockow, D.; Flagan, R. C.; Seinfeld, J. H. Formation of organic aerosols from the oxidation of biogenic hydrocarbons. J. Atmos. Chem. 1997, 26, 189–222. (9) Yu, J.; D.R.; Cocker, D. R.; Griffin, R. J.; Flagan, R. C.; Seinfeld, J. H. Gas phase ozone oxidation of monoterpenes: gaseous and particulate products. J. Atmos. Chem 1999, 34, 207–258. (10) Griffin, R. J.; Cocker, D. R.; Flagan, R. C.; Seinfeld, J. H. Organic aerosol formation from the oxidation of biogenic hydrocarbons. J. Geophys. Res. 1999, 104, 3555–3567. (11) Winterhalter, R.; Van Dingenen, R.; Larsen, B. R.; Jensen, N. R.; Hjorth, J. LC-MS analysis of aerosol particles from the oxidation of R-pinene by ozone and OH-radicals. Atmos. Chem. Phys. Dis. 2003, 3, 1–39. (12) Presto, A. A.; Huff Hartz, K. E.; Donahue, N. M. Secondary organic aerosol production from terpene ozonolysis: 1. Effect of UV radiation. Environ. Sci. Technol. 2005, 39, 7036–7045. (13) Ng, N. L.; Kroll, J. H.; Keywood, M. D.; Bahreini, R.; Varutbangkul, V.; Flagan, R. C.; Seinfeld, J. H.; Lee, A.; Goldstein, A. H. Contribution of first- versus second-generation products to secondary organic aerosols formed in the oxidation of biogenic hydrocarbons. Environ. Sci. Technol. 2006, 40, 2283–2297. (14) Lee, A.; Goldstein, A. H.; Keywood, M. D.; Gao, S.; Varutbangkul, V.; Bahreini, R.; Ng, N. L.; Flagan, R. C.; Seinfeld, J. H. Gas-phase products and secondary organic aerosol yields from the ozonolysis of ten different terpenes. J. Geophys. Res 2006, 111, D07302, doi:, 10.1029/2005JD006437. (15) Martin-Reviejo, M.; Wirtz, K. Is benzene a precursor for secondary organic aerosol. Environ. Sci. Technol. 2005, 39, 1045– 1054. (16) Song, C.; Na, K.; Warren, B.; Malloy, Q.; Cocker, D. R. Secondary organic aerosol formation from the photooxidation of p- and o-xylene. Environ. Sci. Technol. 2007, 41, 7403–7408. (17) Song, C.; Na, K.; Cocker, D. R. Impact of the hyrdrocarbon to NOx ratio on secondary organic aerosol formation. Environ. Sci. Technol. 2005, 39, 3143–3149. (18) Johnson, D.; Jenkin, M. E.; Wirtz, K.; Martin-Reviejo, M. Simulating the formation of secondary organic aerosol from the photooxidation of toluene. Environ. Chem. 2004, 1, 150– 165. (19) Tsigaridis, K.; Kanakidou, M. Secondary organic aerosol importance in the future atmosphere. Atmos. Environ. 2007, 41, 4682–4692. (20) Gaydos, T. M.; Pinder, R. W.; Koo, B.; Fahey, K. M.; Pandis, S. N. Development and application of a three-dimensional aerosol chemical transport model, PMCAMx. Atmos. Environ. 2007, 41, 2594–2611. (21) Lane, T. E.; Pandis, S. N. Predicted secondary organic aerosol concentrations from the oxidation of isoprene in the Eastern United States. Environ. Sci. Technol. 2007, 41, 3984–3990. (22) Lane, T. E.; Donahue, N. M.; Pandis, S. N. Simulating secondary organic aerosol formation using the volatility basis-set approach in a chemical transport model. Atmos. Environ. 2008, in press. (23) Environ. User’s guide to the comprehensive air quality model with extensions (CAMx). Version 4.10. Report prepared by ENVIRON International Corporation, Novato, CA, 2003. (24) NEIv3 1999: U.S. EPA, 1999 National Emission Inventory Documentation and Data, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC. http://www.epa.gov/ttn/chief/net/1999inventory.html, 2002, accessed December 15, 2007.

(25) MOBILE6: U.S. EPA, User’s Guide to MOBILE6.1 and MOBILE6.2: Mobile Source Emission Factor Model, Office of Transportation and Air Quality, Ann Arbor, MI, October 2002; Report No. EPA420-R-02-028. http://www.epa.gov/otaq/m6.htm, 2002, accessed December 15, 2007. (26) BEISv3.13: Schwede, D.; Pouliot, G.; Pierce, T. 2005: Changes to the Biogenic Emissions Inventory System version 3 (BEIS3). Presented at the 4th Annual CMAS Models-3 Users’ Conference, Sept 26-28, 2005, Friday Center, University of North Carolina, Chapel Hill, NC. Available at http://www.cmascenter.org/html/ 2005_conference/abstracts/2_7.pdf, accessed December 15, 2007. (27) Grell, G. A.; Dudhia, J.; Stauffer, D. R. A Description of the fifthgeneration Penn State/NCAR Mesoscale Model (MM5), 1995, NCAR/TN-398+STR. http://www.mmm.ucar.edu/mm5/documents/mm5-desc-doc.html, accessed December 15, 2007. (28) Helmig, D.; Ortega, J.; Guenther, A.; Herrick, J. D.; Geron, C. Sesquiterpene emissions from loblolly pine and their potential contribution to biogenic aerosol formation in the Southeastern US. Atmos. Environ. 2006, 40, 4150–4157. (29) Carter, W. P. L. Programs and Files Implementing the SAPRC99 Mechanism and its Associates Emissions Processing Procedures for Models-3 and Other Regional Models. January 31, 2000. http://pah.cert.ucr.edu/∼carter/SAPRC99.htm, accessed December 15, 2007. (30) Environ. User’s guide to the comprehensive air quality model with extensions (CAMx). Version 4.30. Report prepared by ENVIRON International Corporation, Novato, CA, 2006. (31) Karydis, V. A.; Tsimpidi, A. P.; Pandis, S. N. Evaluation of a three-dimensional chemical transport model (PMCAMx) in the eastern United States for all four seasons. J. Geophys. Res. 2007, 112, D14211,doi:, 10.1029/2006JD007890. (32) Donahue, N. M.; Robinson, A. L.; Stanier, C. O.; Pandis, S. N. Coupled partitioning, dilution, and chemical aging of semivolatile organics. Environ. Sci. Technol. 2006, 40, 2635–2643. (33) 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, 2580– 2585. (34) Odum, J. R.; Jungkamp, T.; Griffin, R. J.; Forstner, H.; Flagan, R. C.; Seinfeld, J. H. Aromatics, reformulated gasoline, and atmospheric organic aerosol formation. Environ. Sci. Technol. 1997, 31, 1890–1897. (35) Bowman, F.; Odum, J.; Pandis, S. N.; Seinfeld, J. H. Mathematical model for gas-particle partitioning of secondary organic aerosol. Atmos. Environ. 1997, 31, 3921–3931. (36) Strader, R.; Lurmann, F.; Pandis, S. N. Evaluation of secondary organic aerosol formation in winter. Atmos. Environ. 1999, 33, 4849–4863. (37) Kostenidou, E.; Pathak, R. K.; Pandis, S. N. An algorithm for the calculation of secondary organic aerosol density combining AMS and SMPS data. Aerosol Sci. Technol. 2007, 41, 1002–1010. (38) Ng, N. L.; Chhabra, P. S.; Chan, A. W. H.; Surratt, J. D.; Kroll, J. H.; Kwan, A. J.; McCabe, D. C.; Wennberg, P. O.; Sorooshian, A.; Murphy, S. M.; Falleska, N. F.; Flagan, R. C.; Seinfeld, J. H. Effect of NOx level on secondary organic aerosol (SOA) formation from the photooxidation of terpenes. Atmos. Chem. Phys. 2007, 7, 5159–5174. (39) Turpin, B. J.; Lim, H.-J. Species contributions to PM2.5 mass concentrations: revisiting common assumptions for estimating organic mass. Aerosol Sci. Technol. 2001, 35, 602–610. (40) Poirot, R. L.; Husar, R. B. Chemical and physical characteristics of wood smoke in the northeastern US during July 2002: impacts from Quebec forest fires. Presented at the A&WMA Specialty Conference: Regional and Global Perspectives on Haze: Causes, Consequences and Controversies, Asheville, NC, Oct 25-29, 2004; paper #94. (41) Malm, W. C.; Hand, J. L. An examination of physical and optical properties of aerosols collected in the IMPROVE program. Atmos. Environ. 2007, 41, 3407–3427. (42) Malm, W. C.; Day, D. E.; Carrico, C. M.; Kreidenweis, S. M., Jr.; McMeeking, G.; Lee, T.; Carrillo, J.; Schichtel, B. A. Intercomparison and closure calculations using measurements of aerosol species and optical properties during the Yosemite aerosol characterization study. J. Geophys. Res. 2005, 110, D14302. (43) Tsimpidi, A. P., Karydis, V. A., Pandis, S. N. Response of fine particulate matter to emission changes of NOx and anthropogenic VOCs in the eastern US. J. Air Waste Manage. Assoc. 2008, in press.

ES703225A

VOL. 42, NO. 16, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6027