Temporal Trends in Motor Vehicle and Secondary Organic Tracers

Nov 24, 2010 - Science Department, Baylor University, Waco, Texas, and. Wisconsin State ... lecular marker source apportionment of motor vehicle exhau...
1 downloads 0 Views 440KB Size
Download PDF
Environ. Sci. Technol. 2010, 44, 9398–9404

Temporal Trends in Motor Vehicle and Secondary Organic Tracers Using in Situ Methylation Thermal Desorption GCMS R E B E C C A J . S H E E S L E Y , * ,†,‡ JEFFREY T. DEMINTER,§ MARK MEIRITZ,§ DAVID C. SNYDER,† AND JAMES J. SCHAUER† Environmental Chemistry and Technology Program, University of Wisconsin-Madison, Madison, Wisconsin, Environmental Science Department, Baylor University, Waco, Texas, and Wisconsin State Laboratory of Hygiene, University of Wisconsin-Madison, Madison, Wisconsin

Received July 9, 2010. Revised manuscript received October 19, 2010. Accepted November 11, 2010.

Organic aerosol measurements with high temporal resolution can differentiate primary organic carbon (POC) from secondary organic carbon (SOC) and can be used to distinguish morning rush hour traffic emissions and subsequent photo-oxidation. In the current study, five hour filter samples were collected during the Summer Study for Organic Aerosols at Riverside (SOAR-1 in CA, USA) for analysis of organic molecular markers. To achieve the low detection limits required for the high temporal resolution data, a laboratory-based in situ methylation thermal desorption gas chromatography-mass spectrometry method was developed. This enabled the measurement of potential markers of SOC, including phthalic acid, along with markers for traffic emissions, including norhopane. The aromatic acids correlated well with unapportioned OC from a molecular marker chemical mass balance model (SOC-cmb; r2 ) 0.46-0.70) and SOC from the elemental carbon tracer method (SOC-ec; r2 ) 0.40-0.56). The aromatic acid/norhopane ratio increased substantially over the course of each day. The average midday phthalic acid ratio compared to previously published roadway emissions was a factor of 4 times higher, while the average 1,2,3benzenetricarboxylic acid ratio was a factor of 40 times higher than roadway emissions. Using correlation plots of SOCcmb and phthalic acid, it was estimated that 2.9 ( 0.6 µg m-3 SOC was associated with mid-day aromatic acid production in Riverside.

Introduction Real-time and short-duration atmospheric measurements can provide new approaches for source apportionment and source allocation and can more effectively track photochemical processing in the atmosphere than traditional 24 h filter-based analysis. There are many real-time and semicontinuous carbon measurements which offer a range of * Corresponding author phone: (254)710-3158; e-mail: [email protected]. † Environmental Chemistry and Technology Program, University of Wisconsin-Madison. ‡ Baylor University. § Wisconsin State Laboratory of Hygiene, University of Wisconsin-Madison. 9398

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010

chemical specificity. On one end, bulk carbon measurements can be obtained including black carbon by aetholometer or PSAP (particle soot absorption photometer), water-soluble carbon (WSOC) by PILS (particle-into-liquid sampler) (1), and semicontinuous organic and elemental carbon analysis (OCEC), at one minute, six minute, and one hour time resolution, respectively. All of these techniques give valuable information about general carbon trends. Other real-time instrumentation provides more detailed characterization, like the aerosol mass spectrometer (AMS), which provides information on the organic carbon backbone that can be used to assess the overall level of oxidation (2, 3) and the ATOFMS (atmospheric time-of-flight mass spectrometer) which gives very detailed mass spectral information on single particles (4). The mass spectrometry based carbon analyses, which operate at high time resolution, often have less sourcespecificity than filter based mass spectrometry methods due to lack of chromatographic separation and the higher ionization energies which heavily fragments organic molecules. However, it is difficult to achieve sufficient aerosol mass for filter-based methods that use mass spectrometry for detailed organic compound speciation at time resolutions that are less than 24 h. Organic tracers have been shown to be effective at apportioning organic carbon sources and have been reported for a wide variety of emission sources. The development of thermal desorption gas chromatography-mass spectrometry (TD-GCMS) methods have reduced the mass requirement for nonpolar molecular marker analysis of polycyclic aromatic hydrocarbons (PAHs), hopanes, steranes, and alkanes (5-7). This advance in molecular marker analysis has enabled molecular marker source apportionment of motor vehicle exhaust in personal exposure samples (8) and has been able to differentiate between smoking and high load diesel exhaust contributions. However, to fully support source apportionment efforts and to study SOC in the atmosphere, TD-GCMS methods need to be expanded beyond measurement of nonpolar organic tracers. Measurement of polar organic acids been recently reported for thermal desorption methods of GCMS (9) but is not common due to the necessity of a derivatization step prior to analysis. In the Beiner et al. study, online derivatization by tetramethylammonium acetate was combined with a Curie point pyrolyzer coupled to a gas chromatograph-mass spectrometer (CPP-GC-MS) (9); the reported pyrolysis method was offered as an alternative to traditional thermal desorption (maximum temperature 360 °C), although the CPP-GC-MS heats the sample to 510 °C before transfer to the GC-MS. The higher initial temperature may result in biased results as larger organic molecules could experience thermal break down and yield some of the low molecular weight target analytes. It has been postulated that certain organic di- and tricarboxylic acids may be used as proxies for aged organic aerosol (10, 11) and would be a valuable addition to the set of nonpolar source tracers. The aromatic acids have been previously reported as potential secondary species from aromatic precursor gases (12). Although the organic aliphatic and aromatic diacids are all reported in primary emissions as well, the relative increase in aged aerosol was determined in the current study. A recent study has demonstrated the quick production of phthalic acid (1,2-benzenedicarboxylic acid) under high and low NOx (NO+NO2) conditions (13) 10.1021/es102301t

 2010 American Chemical Society

Published on Web 11/24/2010

with naphthalene as the precursor. However, aliphatic diacids were not measured during the course of the same chamber study (13). A previous study of SOC during the SOAR-1 project focused on the month-long average secondary contribution and an intercomparison of different methods for predicting that average contribution (2). In the current analysis, the driving hypothesis is that aromatic acids can be used as markers for the production of a fraction of SOC in Riverside. The ambient profile of aromatic and aliphatic diacids has been used previously as a qualitative indicator of different SOC precursors (11). Results of this study suggest that aromatic acids can be used to semiquantitatively represent the motor vehicle-associated fraction of the daily SOC in the LA basin. SOC cannot be directly measured but was estimated by two different methods in the current study: the OC remainder after chemical mass balance modeling (SOC-cmb) and the OC remainder from the EC tracer method (SOC-ec) (2). To investigate the utility of the aromatic acids for tracking fractions of SOC, the relationships of aromatic acids with SOC-cmb, SOC-ec, and motor vehicle exhaust markers were examined. In addition, validation parameters were presented for the in situ methylation TD-GCMS method, which added analysis of aliphatic diacids and aromatic acids to the analysis of nonpolar organic tracers.

Experimental Section Ambient Sampling and Chemical Analysis. The Summer SOAR project was conducted from July-August 2005 using medium volume PM2.5 samplers (90 L min-1, URG Corp). The current analysis focuses on the period of July 25-31, 2005. The carbon fraction data have been previously reported for EC, OC, primary organic carbon (POC), and SOC (2). The OCEC was measured thermal-optically (14) and was also used to calculate POC-ec and SOC-ec using the EC tracer method (15) with the noncombustion POCec (NCPOC-ec) included with the POC-ec and not the SOCec, unlike previously reported (2). Briefly, POC-ec and SOCec are calculated as follows with (OC/EC)p as the primary OC/EC ratio POC-ec ) [(OC/EC)p × EC] + NCPOC-ec

(1)

SOC-ec ) OC - POC-ec

(2)

Details of the calculation of (OC/EC)p and NCPOC-ec used in this study are included in ref 2. Briefly, the value used for (OC/EC)p was 1.4, while the value for NCPOC-ec was 1.1. The (OC/EC)p was calculated using emission inventories for the South Coast Air Basin and emission factors from recently published California tunnel studies (16-18), as described in ref 2. The OCEC was reported from samples collected on a four-a-day schedule (Morning: 5-10:00; Midday: 10-15:00; Evening: 15-20:00; Night: 20-5:00). Each four-a-day filter from July 25-31 was also analyzed by in situ methylation TD-GCMS for organic tracers. A Markes International Thermal Desorption Unit (Model M-10140) (Foster City, CA, USA) coupled with an Agilent Technologies 5973 GC-MS was used for the organic tracer analysis. The system is optimized for loadings ranging from 15-40 µg of OC with two punches (1.45 cm2) from a 47 mm quartz fiber filter used in the analysis. The filter sample is first spiked with isotopically labeled internal standard at the same mass present in the quantification standards. The internal standard contains several polycylic aromatic hydrocarbons (PAH), one sterane, several n-alkanes, phthalic acid, and suberic acid (octanedioic acid) which are used as surrogates for compounds of similar structure and molecular weight. The solvent is allowed to evaporate

before the filter is spiked with diazomethane to methylate all acid groups. The wetted filter is then inserted into a glass desorption tube placed into the autosampler. The sample tube is ramped to 360 °C over 20 min to desorb the compounds of interest from the filter. A glass bead focusing trap (0 °C) is used to focus the sample before the temperature is ramped to 360 °C again and desorbed onto the GC column; this focusing step concentrates the analytes desorbed from the filter into a smaller volume of vapor which improves detection in the GCMS. The transfer line between the TD and GC is kept at 210 °C. The GCMS parameters are similar to those previously published for solvent extraction GCMS (19). A 3 point calibration curve is run at the start of each sample set. The quantification standard includes 17 PAH, 7 hopanes and steranes, 12 alkanes, 13 alkanoic acids, 8 aromatic acids, and 7 aliphatic diacids which are used to quantify the corresponding compounds within the filter samples as well as compounds of similar structure and molecular weight. The quantification standards are spiked onto a blank filter punch, allowed to evaporate, and then spiked with diazomethane; this is done to more closely parallel the particulate matter samples. After every 5 samples, the middle calibration standard is rerun and quantified to verify the continued accuracy of the calibration curve; this is referred to as the check standard. A duplicate filter sample or matrix spike (quantification standard spike directly onto a sample filter) is alternated every 10 samples to assess reproducibility and matrix effects. A full table of measured species has been included in Table S1 in the Supporting Information. Molecular Marker-Chemical Mass Balance Modeling (MM-CMB). MM-CMB calculations were performed using EPA CMB 8.2 and molecular marker profiles from the literature to apportion primary source contribution to OC for the study. The same profiles and a similar set of key markers were employed as in the previously reported SOAR-1 study (2). Briefly, EC, norhopane (17R(H)-21β(H)29-norhopane), hopane (17R(H)-21β(H)-hopane), indeno[cd]pyrene (IcdP), and benzo[ghi]perylene (BghiP) were used to apportion spark ignition (SI) motor vehicle exhaust, lubricating oil-impacted motor vehicle exhaust and high load diesel exhaust (20). For this study the benzo[a]pyrene, benzo[b]fluoranthene, and benzo[k]fluoranthene were not included for apportionment of natural gas combustion due to inconsistent detection. Inclusion of these compounds during test model runs where all were above detection did not change the apportionment to the motor vehicle exhaust sources, so they were eliminated from all reported runs. In the few samples where IcdP and BghiP were below detection, only two motor vehicle sources were apportioned - lubricating oil-impacted motor vehicle exhaust and high load diesel. The current study does not include levoglucosan and therefore has no apportionment of biomass burning. However, previous reports on SOAR-1 with levoglucosan measurements have demonstrated little to no contribution from biomass burning (2).

Results and Discussion TD-GCMS Method Validation. Previous work with thermal desorption techniques for the analysis of filter-based atmospheric aerosols has focused on nonpolar compounds. The current work adds a derivatization step and quantification of n-alkanoic acids, aliphatic diacids, and aromatic acids. For the aliphatic diacids, the method is limited to measurement of C6-C10 diacids (hexanedioic to decanedioic acid). The method validation had two objectives: 1) verify that the added acid moieties are successfully quantified and (2) confirm that the added derivatization step has not negatively impacted recovery of nonpolar VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9399

FIGURE 1. 5-9 h average atmospheric concentrations of a) primary carbon fractions including EC (elemental carbon), POC (primary organic carbon by the EC tracer method), and POC-cmb (primary organic carbon apportioned by chemical mass balance modeling) and b) primary organic markers for motor vehicle exhaust. compounds. The tools for the method validation include assessment of check standards, matrix spike, and standard reference material recoveries. Briefly, the validation tools demonstrated the stability of the method for both polar and nonpolar compounds with check standards (almost all within (20% of full recovery, Figure S1), matrix spike recoveries (all within 60-120%, Figure S2), and analysis of National Institute of Standards and Measures urban dust standard reference material (NIST SRM 1649a; almost all PAH within 75-120%, Figure S3) all showing reasonable recoveries. The matrix spike and check standard use the same concentration of standard mixture. Since this is the first reported results for in situ methylation TD-GCMS, the limit of detection (LOD) for the acid compound classes will be reported here: for alkanoic acids (nC25-30) the LOD is 4.7 ng; LOD for aromatic acids is 1.1 ng; LOD for aliphatic diacids (C6-10) is 3.9 ng. For this method, field blank values are the limiting factor and determine method detection limits. The LOD was calculated for each compound class by adding the average field blank to the standard deviation of the field blanks. The complete discussion of the method validation is included in the Supporting Information (text plus Figures S1-3). Time of Day Trends. The apportioned OC (POC-cmb) and the unapportioned OC (SOC-cmb) from the MM-CMB calculations have been included with the calculated POC-ec and SOC-ec from the EC tracer model in Figures 1 and 2. The ambient concentrations of these carbon fractions in Figures 1 and 2 show distinct maxima every day and minima during the night time. There is not a distinct multiday accumulation of ambient particulate matter and no strong day-of-the-week 9400

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010

trend for this week (Monday July 25 to Sunday July 31). In Figure 1, the primary carbon fractions and organic markers have been plotted as a function of time. EC and POC-ec are highly correlated, which is a result of the EC tracer method. Figure 2 shows the secondary-associated fractions and species. EC, POC-ec, and POC-cmb generally have daily maxima in the morning (Figure 1), while OC, SOC-cmb, and SOC-ec generally have daily maxima that are shifted later to the midday (Figure 2). The molecular marker concentrations are shown in Figures 1b and 2b. Norhopane, hopane, and IcdP generally peak in the morning to midday time slot, with a bit more variability than the bulk carbon fractions. The aromatic acids shift from midday to evening daily maxima for methylphthalic versus phthalic and 1,2,3-tricarboxylic acids. The aromatic acids discussed in this study have been offered as potential indicators of secondary organic aerosol in Los Angeles (12, 21), and some have been reported by Kautzman et al. for a chamber study of the oxidation of naphthalene (13). The di- and tricarboxylic acids have also been reported in motor vehicle exhaust and roadway emissions but are enriched relative to motor vehicle exhaust markers in remote areas (11). To illustrate this enrichment during the SOAR study with higher time resolution, the ratios of the three aromatic acids to norhopane have been plotted in Figure 3 divided by the ratios reported by Fraser et al. for these compounds in a Los Angeles roadway tunnel (12). The road way measurements should reflect primary emission ratios. It is quickly apparent that the acid/norhopane ambient to emitted ratios generally increase over the course of the day;

FIGURE 2. 5-9 h average atmospheric concentrations of a) secondary carbon fractions including SOC-ec (secondary organic carbon by the EC tracer method) and SOC-cmb (secondary organic carbon from the unapportioned OC from chemical mass balance modeling) and b) proposed secondary organic markers. however, the difference in axis maximum should be noted for Figure 3a and b. This increase is quite substantial for 1,2,3-benzenetricarboxylic acid with maximum ambient loadings a factor of 100 over reported emission ratios for July 25, 30, and 31 and a factor of 10 over the emitted ratio. Methylphthalic (4-methylbenzene-1,2-dicarboxylic acid) and phthalic acids have a more modest, but still substantial, daily average increases of four times the emitted ratio and maxima of six and seven times, respectively. On each day except July 25, the ratio of phthalic and methylphthalic acids to norhopane in the night sample is roughly equivalent to the emitted ratio and then generally increases in a stepwise fashion over the course of the day. It appears that the rate of reaction during the day is quicker than can be captured with these five hour increments, as the morning sample (which incorporates the rush hour traffic pulse) is already elevated above the emitted ratio. This is substantiated by Kautzman et al., who shows high phthalic acid concentrations after 350 min reaction time and distinct increases over background as early as 80 min. Correlation of Bulk Carbon Fractions and Organic Tracers. Because there are so many intrarelated carbon fractions and species presented here, a complete table of regression coefficients has been included in the Supporting Information Table S2. In the previous section, the daily trends for carbon fractions and molecular markers were visualized to illustrate the increase in aromatic acids relative to hopanes (Figure 3) over the course of each day. In the section below, regression statistics will be used to determine which markers correlate with the different carbon fractions and whether

the motor vehicle markers correlate with the aromatic acids. To delimit the correlation trends, cut points were assigned for “correlated” r2 > 0.40 and “minor correlation” 0.4 > r2 > 0.20 in Table S2 and in this discussion. POC-ec did not appear to be representative of the organic fraction of motor vehicle exhaust, the major primary emission source for the SOAR-1 study period (2). This was demonstrated by poor to minor correlation with hopane and IcdP (r2 ) 0.11 and 0.3, respectively in Figure S4), organic markers for gasoline-powered motor vehicle exhaust, and lubricating oil-impacted motor vehicle exhaust. However, hopane was well correlated with POC-cmb (r2 ) 0.81). IcdP had a minor correlation for the total POC-cmb (r2 ) 0.31), but slightly higher correlation specifically with the spark ignition motor vehicle CMB source contribution (r2 ) 0.37). The correlation plots for the primary OC have been included in the Supporting Information (Figure S4). Intriguingly the hopanes are correlated to POC-cmb (0.80), secondary components of OC (SOC-cmb, 0.50 and SOC-ec, 0.40), and the three key aromatic acids (0.22-0.59). Norhopane’s correlation with the aromatic acids has a decreasing trend: r2 ) 0.59, 0.43, and 0.22 for methylphthalic, phthalic, and 1,2,3-benzenetricarboxylic acid, respectively. This may indicate increasing number of reaction steps to produce these compounds in the atmosphere. These aromatic acids may also have multiple aromatic precursor sources, with the trend in correlation due to differences in precursor source contributions. Without Figure 3 it might be assumed that the correlation exists because the acids are also primary emissions, but in this case it more likely indicates that the acids VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9401

FIGURE 3. Daily increase in aromatic acids over roadway emission profile. Ratios of aromatic acids to norhopane for ambient ratio divided by the emitted ratio from a Los Angeles roadway tunnel study by Fraser et al. (12): a) 1,2,3-benzenetricarboxylic acid and (b) phthalic acid and methylphthalic acid. are being formed from precursors emitted by motor vehicles. As an example of this relationship between phthalic acid and motor vehicle emissions, naphthalene oxidizes to form phthalic acid (13), and emission inventories indicate that over 50% of naphthalene is emitted by motor vehicles in the LA basin (22). The three aromatic acids have been regressed against SOC-cmb and SOC-ec in Figure 4. SOC-ec and SOC-cmb have the highest correlation with methylphthalic and the lowest correlation with 1,2,3- benzenetricarboxylic acid, which mimics the relationship seen with the hopane correlations. The regression lines for SOC-cmb and SOCec reveal important information about these two OC fractions. For SOC-cmb the y-intercept for all acid comparisons is roughly 3.4-4.4 µg m-3, and it is roughly 2-3 µg m-3 for SOC-ec. The several microgram y-intercept for both SOC-ec and SOC-cmb indicates either the presence of SOC that does not have an aromatic precursor or a potential underestimation of the primary OC contributions (both SOC-ec and SOC-cmb are calculated by subtracting the respective POC-ec or POC-cmb from measured OC). The slopes for phthalic and 1,2,3-benzenetricarboxylic are very similar for SOC-ec (0.35 ( 0.08 and 0.29 ( 0.07, respectively) and SOC-cmb (0.30 ( 0.06 and 0.24 ( 0.05, respectively); this indicates that a relatively equivalent mass of SOC is associated with these two compounds. Using the slopes from Figure 4, a mid-day average 2.9 ( 0.6 µg m-3 SOC-cmb was associated with phthalic acid production and 2.3 ( 0.5 µg m-3 SOC-cmb was associated with 1,2,3benzenetricarboxylic acid production. This is a range of 9402

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010

34-43% of total SOC-cmb for that time period associated with oxidation of aromatic precursors. These two aromatic acid-associated SOC values are likely not independent of one another but are two related estimates of total, daily scale production of SOC from aromatic precursors. A sensitivity test was conducted to assess the impact of EC tracer method parameters (2, 15); however, it was shown that altering the parameters had minimal impact on the final calculated aromatic acid-associated SOC (Supporting Information Table S3). For comparison, the ratio of phthalic acid to OC from a naphthalene oxidation chamber study was also used to estimate SOC from naphthalene oxidation during the current SOAR study (13). The resulting SOC from naphthalene oxidation was estimated at 0.12 µg m-3 SOC during mid-day, which is 2% of SOC-cmb. This is appreciably lower than the phthalic acid-associated SOC (2.9 µg m-3). The discrepancy may be due to the minor contribution to atmospheric SOC from naphthalene or differences between the oxidizing conditions during the current SOAR campaign versus the chamber study in ref 13. The SOAR-1 study provided an excellent backdrop for investigating the development of SOC. Since the primary particulate emission sources in the LA air basin system have been well-characterized and there were limited multiday SOA events during the SOAR-1 study period, the daily SOC production can be studied in detail. This is advantageous for statistical analysis as similar conditions are repeated each day allowing more robust correlations to be demonstrated under ambient conditions. Three key aromatic acids

FIGURE 4. Correlation plots for a,c,e) secondary organic tracers vs SOC-cmb and b,d,f) secondary organic markers vs SOC-ec. (methylphthalic, phthalic, and 1,2,3-benzenetricarboxylic acid) correlate well with SOC-cmb and moderately well with SOC-ec by the EC tracer method. The aromatic acids also display an increasing ratio vs hopane over the course of each day. These compounds seem to indicate the oxidation of SOA precursors from motor vehicle and other aromatic precursor sources as evidenced by correlation trends with hopane and the minor correlation of hopane with SOC-cmb. Recent chamber studies of naphthalene, the highest emitted volatile PAH, support the idea that these aromatic acids and phthalic acid in particular are formed by oxidation of aromatic precursors in the atmosphere (13). The higher time resolution

of the filter sampling (5-9 h) with measurement by in situ methylation TD-GCMS was key to determining the daily trends in these molecular markers and distinguishing the primary versus secondary nature of the marker species.

Acknowledgments We acknowledge funding for method development provided by the U.S. Environmental Protection Agency (EPA) STAR grant RD-83108801. We also acknowledge the funding for the general SOAR-1 site: EPA STAR grants R831080 and RD832161010 and a grant from the National Science Foundation (NSF grant ATM-0449815). We gratefully acknowledge VOL. 44, NO. 24, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

9403

the assistance of Jose Jimenez and Ken Docherty from Colorado UniversitysBoulder and Paul Ziemann from the University of CaliforniasRiverside for assistance with site preparation and support.

Supporting Information Available Discussion and validation plots for the in situ methylation TD-GCMS method (S1-S3), correlation plots for primary carbon fractions and molecular markers (S4), and tables of ambient concentrations (S1), statistical data (S2), and a sensitivity test (S3). This material is available free of charge via the Internet at http://pubs.acs.org.

(10) (11) (12)

(13)

Literature Cited (1) Peltier, R. E.; Weber, R. J.; Sullivan, A. P. Investigating a liquidbased method for online organic carbon detection in atmospheric particles. Aerosol Sci. Technol. 2007, 41 (12), 1117–1127, 10.1080/02786820701777465. (2) Docherty, K. S.; Stone, E. A.; Ulbrich, I. M.; DeCarlo, P. F.; Snyder, D. C.; Schauer, J. J.; Peltier, R. E.; Weber, R. J. Apportionment of primary and secondary organic aerosols in southern california during the 2005 study of organic aerosols in riverside (soar-1). Environ. Sci. Technol. 2008, 42 (20), 7655–7662, 10.1021/ es8008166. (3) Jimenez, J. L.; Canagaratna, M. R.; Donahue, N. M.; Prevot, A. S. H.; Zhang, Q.; Kroll, J. H.; DeCarlo, P. F.; Allan, J. D. Evolution of organic aerosols in the atmosphere. Science 2009, 326 (5959), 1525–1529, 10.1126/science.1180353. (4) Holecek, J. C.; Spencer, M. T.; Prather, K. A. Analysis of rainwater samples: Comparison of single particle residues with ambient particle chemistry from the northeast pacific and indian oceans. J. Geophys. Res., [Atmos.] 2007, 112 (D22), XXX. (5) Hays, M. D.; Smith, N. D.; Kinsey, J.; Dong, Y. J.; Kariher, P. Polycyclic aromatic hydrocarbon size distributions in aerosols from appliances of residential wood combustion as determined by direct thermal desorption - gc/ms. J. Aerosol. Sci. 2003, 34 (8), 1061–1084. (6) Hays, M. D.; Smith, N. D.; Dong, Y. J. Nature of unresolved complex mixture in size-distributed emissions from residential wood combustion as measured by thermal desorption-gas chromatography-mass spectrometry. J. Geophys. Res., [Atmos.] 2004, 109 (D16), XXX. (7) Sheesley, R. J.; Schauer, J. J.; Meiritz, M.; DeMinter, J. T.; Bae, M. S.; Turner, J. R. Daily variation in particle-phase source tracers in an urban atmosphere. Aerosol Sci. Technol. 2007, 41 (11), 981–993. (8) Sheesley, R. J.; Schauer, J. J.; Garshick, E.; Laden, F.; Smith, T. J.; Blicharz, A. P.; Deminter, J. T. Tracking personal exposure to particulate diesel exhaust in a diesel freight terminal using organic tracer analysis. J. Exposure Sci. Environ. Epidemiol. 2009, 19 (2), 172–186, 10.1038/jes.2008.11. (9) Beiner, K.; Plewka, A.; Haferkorn, S.; Iinuma, Y.; Engewald, W.; Herrmann, H. Quantification of organic acids in particulate matter by coupling of thermally assisted hydrolysis and methylation with thermodesorption-gas chromatography-mass

9404

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 24, 2010

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21)

(22)

spectrometry. J. Chromatogr., A 2009, 1216 (38), 6642–6650, 10.1016/j.chroma.2009.07.054. Ray, J.; McDow, S. R. Dicarboxylic acid concentration trends and sampling artifacts. Atmos. Environ. 2005, 39, 7906–7919, doi:10.1016/j.atmosenv.2005.09.024. Sheesley, R. J.; Schauer, J. J.; Kenski, D. Trends in secondary organic aerosol at a remote site in michigan’s upper peninsula. Environ. Sci. Technol. 2004, 38 (24), 6491–6500. Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T. Gas-phase and particle-phase organic compounds emitted from motor vehicle traffic in a los angeles roadway tunnel. Environ. Sci. Technol. 1998, 32 (14), 2051–2060. Kautzman, K. E.; Surratt, J. D.; Chan, M. N.; Chan, A. W. H.; Hersey, S. P.; Chhabra, P. S.; Dalleska, N. F.; Wennberg, P. O. Chemical composition of gas- and aerosol-phase products from the photooxidation of naphthalene. J. Phys. Chem. A 2010, 114 (2), 913. 10.1021/jp908530s. Schauer, J. J.; Mader, B. T.; Deminter, J. T.; Heidemann, G.; Bae, M. S.; Seinfeld, J. H.; Flagan, R. C.; Cary, R. A. Ace-asia intercomparison of a thermal-optical method for the determination of particle-phase organic and elemental carbon. Environ. Sci. Technol. 2003, 37 (5), 993–1001. Turpin, B. J.; Huntzicker, J. J. Identification of secondary organic aerosol episodes and quantitation of primary and secondary organic aerosol concentrations during scaqs. Atmos. Environ. 1995, 29 (23), 3527–3544. Allen, J. O.; Mayo, P. R.; Hughes, L. S.; Salmon, L. G.; Cass, G. R. Emissions of size-segregated aerosols from on-road vehicles in the caldecott tunnel. Environ. Sci. Technol. 2001, 35 (21), 4189– 4197. Ban-Weiss, G. A.; McLaughlin, J. P.; Harley, R. A.; Lunden, M. M.; Kirchstetter, T. W.; Kean, A. J.; Strawa, A. W.; Stevenson, E. D. Long-term changes in emissions of nitrogen oxides and particulate matter from on-road gasoline and diesel vehicles. Atmos. Environ. 2008, 42 (2), 220–232, 10.1016/j.atmosenv. 2007.09.049. Kirchstetter, T. W.; Harley, R. A.; Kreisberg, N. M.; Stolzenburg, M. R.; Hering, S. V. On-road measurement of fine particle and nitrogen oxide emissions from light- and heavy-duty motor vehicles. Atmos. Environ. 1999, 33 (18), 2955–2968. Sheesley, R. J.; Schauer, J. J.; Smith, N. D.; Hays, M. D. Development of a standardized method for the analysis of organic compounds present in pm2.5. In Proceedings of the AWMA Annual Meeting 2000. 2000. Salt Lake City, UT. Lough, G. C.; Schauer, J. J. Sensitivity of source apportionment of urban particulate matter to uncertainty in motor vehicle emissions profiles. J. Air Waste Manage. Assoc. 2007, 57 (10), 1200–1213. Schauer, J. J.; Fraser, M. P.; Cass, G. R.; Simoneit, B. R. T. Source reconciliation of atmospheric gas-phase and particle-phase pollutants during a severe photochemical smog episode. Environ. Sci. Technol. 2002, 36 (17), 3806–3814. Lu, R.; Wu, J.; Turco, R. P.; Winer, A. M.; Atkinson, R.; Arey, J.; Paulson, S. E.; Lurmann, F. W. Naphthalene distributions and human exposure in Southern California. Atmos. Environ. 2005, 39 (3), 489. 10.1016/j.atmosenv.2004.09.045.

ES102301T