Diurnal Variations in Fine Particles at Sites Downwind from Tokyo

ARTICLE pubs.acs.org/est

Radiocarbon (14C) Diurnal Variations in Fine Particles at Sites Downwind from Tokyo, Japan in Summer Akihiro Fushimi,*,† Rota Wagai,‡ Masao Uchida,† Shuichi Hasegawa,†,§ Katsuyuki Takahashi,|| Miyuki Kondo,† Motohiro Hirabayashi,^ Yu Morino,† Yasuyuki Shibata,† Toshimasa Ohara,† Shinji Kobayashi,† and Kiyoshi Tanabe† †

National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba 305-8506, Japan. National Institute for Agro-Environmental Sciences, 3-1-3 Kan-nondai, Tsukuba 305-8604, Japan. § Center for Environmental Science in Saitama, 914 Kamitanadare, Kazo, Saitama 347-0115, Japan. Japan Environmental Sanitation Center, 10-6 Yotsuyakami-cho, Kawasaki, Kawasaki 210-0828, Japan. ^ National Institute of Polar Research, 10-3 Midoricho, Tachikawa, Tokyo 190-8518, Japan.

)

‡

bS Supporting Information ABSTRACT: The radiocarbon (14C) of total carbon (TC) in atmospheric fine particles was measured at 6 h or 12 h intervals at two sites, 50 and 100 km downwind from Tokyo, Japan (Kisai and Maebashi) in summer 2007. The percent modern carbon (pMC) showed clear diurnal variations with minimums in the daytime. The mean pMC values at Maebashi were 28 ( 7 in the daytime and 45 ( 16 at night (37 ( 15 for the overall period). Those at Kisai were 26 ( 9 in the daytime and 44 ( 8 at night (37 ( 12 for the overall period). This data indicates that fossil sources were major contributors to the daytime TC, while fossil and modern sources had comparable contributions to nighttime TC in the suburban areas. At both sites, the concentration of fossil carbon as well as O3 and the estimated secondary organic carbon increased in the daytime. These results suggest that fossil sources around Tokyo contributed significantly to the high daytime concentration of secondary organic aerosols (SOA) at the two suburban sites. A comparison of pMC and the ratio of elemental carbon/TC from our particulate samples with those from three end-member sources corroborates the dominant role of fossil SOA in the daytime.

1. INTRODUCTION Fine particles such as PM2.5 (50% cutoff diameter at 2.5 μm) in the atmosphere pose a health risk to humans.1 PM2.5 originates from various sources, and mainly consists of carbonaceous compounds, ions, and elements.2,3 Organic matter contributes significantly (20
90%) to the PM2.5 mass.4 Organic aerosols (OA) comprise primary OA (POA; particles emitted directly in the particulate phase) and secondary OA (SOA; particles produced in the atmosphere by conversion from the gas phase), and SOA is estimated to account for the dominant fraction (60
90%) of OA in various environments.5 However, the source and behavior of OA, especially SOA, are still controversial.4,6 Spatial and temporal distributions of POA and SOA can be estimated using chemical transport models (CTMs), which numerically simulate transport, chemical, and aerosol processes. However, it is widely known that CTMs largely underestimate OA concentrations in many cases,7
10 even though concentrations of O3 and inorganic aerosols are reproduced reasonably accurately.7,8,11
13 It has been shown that a CTM reproduced biogenic SOA relatively well.6 However, SOA concentration was remarkably underestimated in urban areas, suggesting that fossil r 2011 American Chemical Society

SOA is largely underestimated.9,10 Hence, observation-based estimation of the source contributions of POA and SOA, particularly in urban areas, is important both for understanding the atmospheric behaviors and sources of OA and for improving the reproducibility of modeling of OA. Observation of organic tracers, often together with chemical mass balance (CMB) or positive matrix factorization, can attribute POA and SOA to some source types.5,10,14
17 SOA concentrations, formed from each precursor such as toluene and isoprene, were estimated based on the measurement of SOA tracers from each precursor.16 However, this method assumes a fixed ratio of a tracer to organic mass under variable conditions, and the accuracy of the method and appropriate selection of tracers are still being examined.18
20 Radiocarbon (14C) measurement can directly discriminate carbon originating from fossil fuel and carbon from nonfossil Received: April 24, 2011 Accepted: July 7, 2011 Revised: July 1, 2011 Published: July 22, 2011 6784

dx.doi.org/10.1021/es201400p | Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology (modern) sources such as biomass burning and biogenic SOA.6 14 C measurements of total carbon (TC) showed that the contributions of nonfossil carbon to TC are not small (26
59%) even in urban areas, and are large (63
100%) in rural areas.6 14C measurement by carbon fractions indicated that elemental carbon (EC) originated nearly exclusively from fossil fuel, whereas biogenic sources (mostly biogenic SOA) were dominant (60%) for organic carbon (OC) during summer in Z€urich, Switzerland.21 14C measurement of water-soluble OC (WSOC) indicated a relatively high fossil contribution to SOA (≈40%) for aerosols at an urban site in summer, which was of regional origin.22 Compound-class-specific 14C analysis indicated that biomass burning contributed 27
39% of the polycyclic aromatic hydrocarbon burden in PM1.1 composite samples in summer and winter in Tokyo.23 In addition to such 14C analyses of carbon fractions,21
29 time-resolved measurement of 14C would be useful to elucidate the origin and behavior of carbonaceous aerosol. A few studies have measured 14C at 6 h or 12 h intervals 30
32 at urban and suburban areas in summer, and found that biogenic sources dominated OC and TC.30,31 Szidat et al.30 also indicated that biogenic OC depends on the activity of plants and atmospheric oxidants, and anthropogenic OC correlates with black carbon. In contrast, other measurements at urban and suburban areas in summer showed that fossil sources contribute substantially to TC.32 These results suggest that the origin of OA varies a great deal depending on the region and meteorological conditions. Such time-series 14C data, along with CMB and CTM, would enable more detailed analyses of the sources and behavior of OA, especially of SOA. To reveal the origin and mechanism of high OA concentrations at suburban areas in summer, we performed 14C measurements of TC and other chemical analyses at 6 h (12 h at night) intervals and subsequent analyses by CMB and CTM.33,34 We measured 14C simultaneously at two suburban sites, 50 and 100 km downwind from an urban center, to investigate the transport of primary pollutants from the urban center and the formation of fossil SOA involved. The 14C measurements, with a relatively short interval during days of typical weather conditions in summer, first showed a statistically significant increase of fossil carbon in the daytime at the suburban areas and suggested the influence of fossil SOA. In this paper, we examined diurnal variations of fossil carbon and modern carbon with particular focus on the origin and behavior of SOA.

2. MATERIALS AND METHODS 2.1. Sampling. In Japan, a new environmental standard (15 μg m
3 for annual average, 35 μg m
3 for daily average) for PM2.5 was introduced in September 2009 in addition to the standard for suspended particulate matter (100% cutoff diameter at 10 μm). However, annual average PM2.5 concentrations exceeded the environmental standard in urban areas (19.5 μg m
3, n = 14) and suburban areas in 2006 (17.5 μg m
3, n = 2).35 In the Kanto area, almost equivalent to the Tokyo Metropolitan Area, as well as other areas in Japan and foreign countries, severe air pollution used to occur in winter.36 In recent years (2003
2006), however, PM2.5 concentrations in summer have been equivalent to or often higher than those in winter.35 In suburban Tokyo such as Maebashi (Supporting Information (SI) Figure S1), oxidants often reach high concentrations in the daytime in summer.37 In addition, concentrations of WSOC also increase in summer in suburban Tokyo, and the estimated SOA account for 71% of

ARTICLE

PM2.1 mass as a yearly average.38 These phenomena are considered to be the result of transport and photochemical reaction of pollutants emitted around metropolitan Tokyo.37 However, the origin (fossil or modern) and formation mechanism of secondary particles have not been clarified. We conducted a study entitled “Fine Aerosol Measurement and Modeling in Kanto Area” (FAMIKA) in summer 2007 to identify the sources and mechanism of the high daytime concentration of fine particles, and to improve the CTM and emission inventory in cooperation with several institutes. In FAMIKA, chemical analyses based on filter samplings and various online measurements including gaseous compounds and meteorological conditions were conducted from July 30 to August 16 in 2007 at four sites in the Kanto area (SI Figure S1). Fine particulate samples (PM2.0) for 14C analysis were collected at Kisai and Maebashi located 50 and 100 km north
northwest of Tokyo, respectively (SI Figure S1). The 14C analysis was performed during two high-priority weeks (Phase I: July 31
August 3 and phase II: August 6
11) when concentrations of PM2.5 and O3 were remarkably high. During the two phases, a southerly wind (from Tokyo toward Kisai and Maebashi, indicated in SI Figure S1) was dominant in the daytime. The PM2.0 samples were collected at 6 h intervals in the daytime (09:00
15:00) and evening (15:00
21:00) and at 12 h intervals at night (21:00
09:00) using a high-volume Andersen air sampler (AH-600; Tokyo Dylec, Tokyo, Japan mounted on an HVC-1000A, Sibata, Tokyo, Japan) at a rate of 570 L min
1. Photochemical reaction was expected to be the most active in the daytime, and the emissions of primary air pollutants such as nitrogen oxides and carbon monoxide are also the largest during the day.39 Primary emissions are the second largest in the evening,39 and secondary particles would still be at relatively high levels. At night, both primary and secondary emissions would be smaller than during the daytime and evening, and the mountainous forests to the north of the Maebashi site might have had an effect because a gentle mountain breeze (northerly wind) was often observed at Maebashi from 18:00 (SI Figure S1). The Andersen sampler originally consisted of a four-stage impactor (50% cutoff diameters at 7.0, 3.3, 2.0, and 1.1 μm) and a backup (final) filter (203  254 mm quartz fiber filters, deposit area: 39 900 mm2, Pallflex 2500QAT-UP, Pall, East Hills, NY). We replaced the lowest separation plate (diameter: 1.1 μm) with a spacer so as to collect PM2.0 samples on backup filters. Quartz fiber filters were also set on the impaction plates. All quartz fiber filters were baked for 30 min at 600 C in air before sampling. Four blank filters (203  254 mm), two for Maebashi and the other two for Kisai, were transported and stored in the same manner as the samples. 2.2. Gravimetric and Carbon Analysis. PM2.0 mass concentration was determined from the difference in the weight of the backup quartz fiber filters before and after sampling. Weighing was performed using an analytical balance (readability 0.1 mg, LA130S
F, Sartorius AG, Goettingen, Germany) in a chamber (CHAM-1000, Horiba Ltd., Kyoto, Japan) where the temperature and relative humidity were controlled at 25 C and 50%, respectively. The quartz fiber filters were conditioned in the chamber for >24 h before weighing. Each sample was weighed three times and the results were averaged. The weighing error was 1.7
22% (average 3.2%) in case of the Maebashi samples calculated based on the error of repeated weighing (0.2 mg). Quartz fiber filters are fragile and easily damaged during sample handling, thus the measured PM mass might be 6785

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology

ARTICLE

Table 1. Average Measured Values of pMC, O3, PM2.0, Carbonaceous Compounds, And Other Parameters O3

PM2.0

EC

OC

TC

OC/

temp

NOx

(C)

(ppb) (ppb) (μg m
3) (μg m
3) (μg m
3) (μg m
3)

TC

POC

SOC

(μg m
3) (μg m
3)

SOC/

pMC

pMC

OC

(raw) a

(bk-subtracted) b

28.0 ( 7.4

Maebashi day

29.6

12.5

61.2

24.2

2.4

7.1

9.5

0.75

0.53

6.6

0.93

37.3 ( 6.1

evening

29.5

15.3

78.9

27.5

2.6

6.7

9.3

0.72

0.58

6.1

0.91

40.7 ( 5.3

30.6 ( 9.8

night

24.4

13.6

28.4

23.3

2.2

4.5

6.7

0.67

0.49

4.0

0.88

51.3 ( 9.3

45.3 ( 15.6

overall

27.0

13.8

49.2

24.6

2.4

5.7

8.1

0.70

0.53

5.6

0.91

45.1 ( 9.8

37.3 ( 14.6

day

29.6

12.8

53.6

21.3

1.8

6.3

8.0

0.80

0.39

5.9

0.94

37.5 ( 6.2

26.2 ( 9.3

evening night

29.3 25.1

12.6 24.9

50.6 9.9

21.0 15.1

1.8 2.4

4.6 4.0

6.3 6.4

0.70 0.62

0.39 0.53

4.2 3.5

0.91 0.86

43.0 ( 6.9 47.0 ( 8.7

31.0 ( 12.5 44.4 ( 8.1

overall

27.2

18.8

31.0

18.1

2.1

4.7

6.8

0.69

0.44

4.5

0.90

43.6 ( 8.4

36.5 ( 12.3

Kisai

a

Raw data. SD among samples are shown. b Blank-subtracted values. SD among samples are shown. The negative values were not included in the averaging process. For pMC, the average values and standard deviation (SD) are shown. The overall averages are the time-weighted averages in consideration of the longer sampling duration in nighttime compared to daytime and evening.

undervalued. However, other types of filters cannot be used for 14 C analysis, and other field measurements of PM (e.g., PM2.5 by TEOM 1400, Thermo Electron, Waltham, MA) in FAMIKA cannot be substituted either, since their cutoff diameters were not the same as for the 14C analysis. Therefore, we used the PM2.0 mass concentration obtained from weighing the quartz fiber filters. Since the cutoff diameter varies for each measurement, it is reasonable that the PM2.0 values were slightly smaller than the PM2.5 measured by TEOM with a diffusion dryer and the measurement region was kept at 35 C. The ratio of PM2.0/PM2.5 was 0.96 ( 0.31 at Maebashi and 0.83 ( 0.23 at Kisai, although PM2.0 data was not obtained and three of the PM2.0/PM2.5 ratios were small (0.25
0.40). Using a portion (diameter: 8 mm) of the PM2.0 filter samples, EC and OC were analyzed by a thermal/optical carbon analyzer (DRI Model 2001, Desert Research Institute, Las Vegas, NV) based on the IMPROVE protocol.40 Carbon amounts in the samples were determined by carbon fractions (OC1, OC2, OC3, OC4, EC1, EC2, and EC3) with subtraction of the averages of the blank filters (n = 2), which were analyzed on the same day as the samples. Detection limits were calculated as 3σ of the measured values for the blank filters (n = 2) for each carbon fraction. Pyrolyzed OC (OCpy) was corrected by laser reflectance. OC was obtained as the sum of OC1
OC4 and OCpy, EC was obtained by subtracting OCpy from the value of EC1
EC3, and TC was obtained as the sum of EC and OC. There are alternative thermal evolution protocols for carbon analysis such as the NIOSH protocol,41 which has been widely used, and other protocols.42,43 The IMPROVE and NIOSH protocols are equivalent for total carbon for ambient and source samples.40 In contrast, EC determined by the NIOSH protocol was typically less than half of that by the IMPROVE protocol.40 The primary difference is the allocation of carbon evolving in the NIOSH at 850 C in a helium atmosphere to the OC fraction rather than to the EC fraction, and this fraction should be classified as EC.40 Therefore, we considered the IMPROVE protocol to be more reasonable. 2.3. 14C analysis. Sample preparation and 14C analysis of TC were performed in the same manner as that described elsewhere44 except for using a microscale graphitization vacuum line in this

study because the sample amount was smaller (≈300 μgC) than in the previous study (≈1 mgC). Briefly, at first, sample filters were cut and added to a quartz tube (outside diameter: 9 and 30 mm, whole length: 360 mm, length of wider part: 80 mm) with copper oxide (2.5 g), elemental silver wires (0.05 g), and elemental copper (1.25 g). Here, the carbon amount in each filter sample was 210
460 μgC based on the results of carbon analysis. CO2 and water in the quartz tube were removed by gradually evacuating the tube contents. After 6 min in vacuum, the quartz tube was welded. Based on the pilot experiments, loss of OC was less than 2% during this vacuum process. For CO2 production, the samples were combusted at 850 C for 2 h. For the 14C analysis, the CO2 sample was reduced to make a graphite target by using the specially designed automated microscale graphitization vacuum line at the National Institute for Environmental Studies.45 Radiocarbon content was measured at the Accelerator Mass Spectrometer (AMS) facility in the National Institute for Environmental Studies (NIES-TERRA 45,46). All radiocarbon measurements are expressed as percent modern carbon (pMC), as used in related studies (e.g., refs 33,34, 44,47
49) and other studies (e.g., refs 31,50
52), with δ13C correction for samples. The pMC values of the samples are defined as normalized to the atmospheric 14C/12C ratio in 1950, and were calculated using the following equation: pMC ¼

ð14 C=12 CÞsample ð14 C=12 CÞHOxII =134:07

 100

ð1Þ

where HOx II is the standard reference material 4990C (oxalic acid, National Institute of Standards and Technology) with a known 14C/12C ratio (pMC = 134.07). So by definition, 14C analysis apportions the carbon fraction of atmospheric aerosols into the fossil (pMC = 0) and modern (pMC = 100) components. From the repeated signals during the AMS measurement of the FAMIKA samples, the measurement errors were 0.36
1.0% (average 0.52%). For IAEA C6 sucrose small standards of 10
50 μgC, the pMC values obtained by our method were within 0.7% of the consensus value.45 Therefore, a total measurement error of 0.7%, or an AMS measurement error if larger than 0.7%, was indicated with the measured raw pMC values in this paper. 6786

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology

ARTICLE

Figure 1. Diurnal variations at Maebashi and Kisai. (A, B) O3, EC and OC. (C, D) Raw pMC and blank-subtracted pMC. a EC was not detected. b pMC was not obtained. c Mean value of blank-subtracted pMC was negative. In (A) and (B), the plots indicate the mean values of EC and OC. The error bars show the SD of EC and OC calculated from the relative standard deviations (RSD) of the repeated analysis (n = 4) of a sample at Maebashi. In (C) and (D), the plots indicate the mean values of the raw pMC and the blank-subtracted pMC. The error bars of the raw pMC are hidden by the plots since these measurement errors are smaller than 1.0%. The error bars of the blank-subtracted pMC show the SD obtained by Monte Carlo simulation considering every SD for the TC and pMC of the sample and blank filters. The SD of the blank-subtracted pMC strongly depended on the RSD of blank TC (49%) since the RSD of the sample TC, sample pMC, and blank pMC were relatively small (3.9, 0.7, and 2.0%, respectively).

We anticipated that pMC of blank filters may not be negligible because the carbon amounts were small in our filter samples that were taken at relatively high temporal resolution. The 14C analysis of trace amounts of carbon in blank filters is challenging and there are few reports in the literature.31 In this study, we succeeded in obtaining the pMC for two blank filters. We thus subtracted the blank pMC from the sample pMC.

3. RESULTS AND DISCUSSION 3.1. Diurnal Variations in O3, PM, EC, OC, and pMC. In the Kanto area, heavy rain fell on July 30 in 2007, and the rainy season ended on August 1. During August 1
3, Typhoon No. 5 approached western Japan, but the Kanto area received little rain with a continuous southerly wind. Pacific anticyclones then covered the Kanto area, and the fine weather continued until the end of the measurement period except for rainfall during 17:00
21:00 on August 6 at Maebashi. During the whole measurement period, a sea breeze (southerly wind) blew from 08:00
09:00 until 17:00
18:00 in general, and calmed or weak land wind (northerly wind) blew at night in the northern Kanto area such as Maebashi. Average measured values of major components are listed in Table 1. During phases I and II, hourly O3 concentration showed clear diurnal variations with the maximum in the daytime. Daytime maximum O3 concentrations were generally higher and their appearance times were generally a few hours later at the northern downwind sites (Maebashi and Kisai) than at the urban site (Komae). The peak concentrations of O3 were generally

observed at 12:00
15:00 at Kisai and at 16:00
18:00 at Maebashi. These observations indicate that photochemical reactions occurred concurrently with the transport of the air mass from Tokyo to the suburban areas. The daytime maximum hourly O3 concentrations at Maebashi and Kisai exceeded 80 ppb during phase I and 100 ppb during phase II. Most of the 6 h and 12 h average PM2.0 concentrations at both sites exceeded the annual average environmental standard of PM2.5 (Table 1). On average, TC comprised 33 and 37% of PM2.0 mass at Maebashi and Kisai, respectively (Table 1). Figure 1 (A, B) shows the diurnal variations of EC, OC, and O3 for a 6 h or 12 h average at Maebashi and Kisai. Typical and distinct diurnal variations were observed, especially in phase II. The concentration of O3, TC, and OC increased in the daytime. For O3, the diurnal variation and the shift in the peak time, seen in the hourly data, also appeared even for the 6 h average. The pMC of the blank filters was 77.5 ( 1.5. For the subtraction of blank pMC, the average TC concentration (1.00 ( 0.49 μg cm
2) of four blank filters was used according to the size of filter analyzed. TC concentrations in the sample filters at Maebashi and Kisai were 5.8 ( 2.4 μg cm
2 (n = 54). Thus, the blank contributions to the sample TC were 21 ( 13%. Although TC concentrations in blank filters are smaller just after baking, OC adsorption is inevitable because the blank filters were treated as the travel blank. The diurnal variations of the raw and the blank-subtracted pMC at Maebashi and Kisai are shown in Figure 1 (C, D). According to the blank subtraction, the sample pMC decreased by 1.5
29 but the variation pattern never changed. The raw pMC 6787

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology

ARTICLE

Figure 2. Diurnal variations at Maebashi and Kisai. (A, B) Modern carbon and fossil carbon. (C, D) Estimated POC and SOC. a pMC was not obtained. b Concentration of modern carbon was assumed to be zero since the mean pMC value was negative. c EC was not detected. In (A) and (B), the plots indicate the mean values of modern carbon and fossil carbon. The error bars show the SD of modern carbon and fossil carbon obtained by the SD of TC and pMC. In (C) and (D), the plots indicate the maximum likelihood estimates of POC and SOC. The error bars show the range of the lower and higher estimates as detailed in the text.

and blank-subtracted pMC showed distinct diurnal variations with the minimum in the daytime, especially at Maebashi. The errors of blank-subtracted pMC were generally larger in the daytime and evening. This is due to the lower pMC in the daytime and the evening and to the smaller carbon amount per unit filter area because of the shorter sampling duration than at night. Since the decrease rates of pMC by the blank subtraction and the uncertainty of the pMC were large and the blank-subtracted pMC dropped below zero even for the mean values on August 2
3, the pMC behavior in this period cannot be fully discussed. On these days, a typhoon struck the western part of Japan, and strong winds (southerly and easterly) blew from the evening of August 2 to the next evening, and northerly winds never blew at night on August 2, unlike every other day. The strong wind probably resulted in the low concentrations of PM (5.1
16 μg m
3) and TC (0.6
6.7 μg m
3). The blank-subtracted pMC and the relevant examinations are described below. At Maebashi, the daytime average pMC (28.0) was 17.3 lower than the nighttime average pMC (45.3). At Kisai, the daytime average pMC (26.2) was 17.8 lower than the nighttime average pMC (44.0). The daytime
nighttime difference of the average pMC was statistically significant (p < 0.05) at both sites. This data indicates that fossil sources were major contributors (≈70%) to the daytime TC. The 6 h or 12 h pMC values during phases I and II were 9.1
60.6 (average 37.3) at Maebashi and ND
56.3 (average 36.5) at Kisai. The extremely low pMC values (8
11) in the daytime on July 31, which were smaller than the values observed at the roadside (13
1649), may be the result of detecting freshly produced particles in the daytime during fine weather after pre-existing particles were removed by heavy rain throughout the Kanto area on the previous day.

The pMC of the major sources of modern carbon differs greatly among sources as shown in SI Table S1. For example, the pMC of biogenic SOA, biomass burning, and soil is 105, 105, and 56, respectively. Therefore, the contribution from biomass would be smaller than the measured pMC if biogenic sources with pMC > 100 contributed significantly to atmospheric particles. 3.2. Diurnal Variations in Modern Carbon, Fossil Carbon, POC, and SOC. Figure 2 (A, B) shows the diurnal variations of modern carbon and fossil carbon at Maebashi and Kisai. Here, modern carbon and fossil carbon were calculated from pMC  TC and (100
pMC)  TC, respectively. At Maebashi, the concentration of fossil carbon showed quite a distinct diurnal variation with the maximum in the daytime. The concentrations of fossil carbon were significantly higher (p < 0.05) in the daytime than in the evening, and also significantly higher in the evening than at night. In contrast, the modern carbon concentration was sometimes slightly higher at night. The higher concentration of modern carbon at night at Maebashi may have been due to biogenic organic aerosols transported by the mountain breeze. At Kisai, both fossil carbon and modern carbon generally increased in the daytime. The concentration of primary organic carbon (POC) was estimated from the EC concentration using a POC/EC ratio, to grasp the trends of the contribution and diurnal variation of secondary organic carbon (SOC) (Figure 2 (C, D)). SOC was estimated by OC
POC. The POC/EC ratio varies widely among POC sources (0.12
38, SI Table S1). The POC/EC ratio of automobiles, which was the dominant contributor to the TC concentrations (SI Table S1), was 0.075
0.47 (average 0.22 50,53). The POC/EC ratio may change diurnally corresponding to variations of the relative contribution of each source; 6788

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology

Figure 3. Comparison of pMC and EC/TC ratio at Maebashi and Kisai with major sources and other ambient data. a This study. b Bennet et al.50 DEP from a 2.4-L off-highway diesel engine meeting US Tier 2 emission guidelines. c Lemire et al.51 for pMC. Han et al.53 for EC/TC ratio. d Offenberg et al.55 SOAfossil (SOA derived from fossil fuel) was formed from toluene by irradiation in air in the presence of NOx. e The pMC of biogenic SOA (SOAbio) and open burning was assumed to be equal to that of atmospheric CO2 (see the text in the Supporting Information). Kleeman et al.54 for EC/TC ratio for rice straw burning. f Hildemann et al.52,56 for a distillate oil-fired industrial boiler. g Hildemann et al.52,56 for a used radial tire. h Hildemann et al.52,56 for resuspended brake dust from rear drum brakes of a light-duty truck. i Hirabayashi et al.49 Roadside and Tsukuba data are the averages for daily samples collected at the Ikegami-Shincho crossing in Kawasaki City, Japan on February 20
26 in 2002 and collected at Tsukuba (suburban Tokyo) on April 24
26 in 2002, respectively. j Takahashi et al.47 Summer and winter data are the averages for the samples collected from August 2 to October 11 in 2004 and from October 12 in 2004 to January 18 in 2005, respectively, in the center of Tokyo (Kudan). k Gelencser et al.57 Average for May
June 2003 at a high-Alpine site (Sonnblick).

however, a constant POC/EC ratio was used because it is difficult to include the diurnal variations in the analysis and to interpret the results. For the maximum likelihood estimate, we used the average POC/EC ratio of automobiles (0.22). The POC/EC ratio (0.42), for the top seven POC sources in SI Table S1 except SOA when weighted by the contribution by CMB, was in the range for automobiles. We therefore investigated the variation of the POC and SOC estimates using the range of POC/EC ratios of automobiles. Ideally, the same carbon analysis protocol (IMPROVE) as used for the ambient samples should also be used for the source samples when deducing the POC/EC ratios. Although the NIOSH protocol is used for some source samples (e.g., refs 50 and 54), the range of the POC/EC ratios for automobiles was determined by using the IMPROVE protocol,53 so the difference of the protocol should not greatly affect the estimates. At both sites, the estimated SOC was significantly higher (p < 0.05) than the POC in all of the time segments, and SOC increased in the daytime in most cases, even in the higher POC case (Figure 2 (C, D)). The SOC dominance in summer at Maebashi is consistent with previous measurements.38 Although the POC concentration at Maebashi was equivalent to that at Kisai, the SOC concentration at Maebashi was higher on average. The difference between daytime and evening concentration of SOC was larger at Kisai than at Maebashi, suggesting that the time of peak SOC was slightly later at Maebashi, similarly to the O3 case. Actually, O3 had a much higher correlation with SOC (R2 = 0.52
0.57) than with POC (R2 = 0.03
0.11).

ARTICLE

Additionally, the diurnal variation of SOC was similar to that of fossil carbon (R2 = 0.68
0.83) (Figure 2). In contrast, the correlation between SOC and modern carbon was weaker (R2 = 0.55
0.62). As described above, the concentration of O3 and SOC was higher at the northern sites, the time of peak O3 and SOC was delayed at the northern sites, and fossil carbon was highly correlated with O3 and SOC in the daytime with a southerly wind. These results suggest that fossil-derived gaseous organics, emitted around Tokyo, resulted in the daytime increase of SOA in the suburban areas. The daytime increase of fossil SOA was clearly shown in both estimates by CMB and CTM.33,34 Besides the diurnal variation of wind direction, changes in the mixing layer height (MLH) may affect the carbon and pMC levels. For example, reduced MLH at night could increase concentrations of air pollutants. However, this effect was not recognized, at least for our main target components (OC and SOC), which increased in the daytime. The effect of diurnal variation of MLH would be complicated if the contributions of industrial boilers and waste burning are large, since their effective stack heights can be higher than the MLH. However, this effect should also be small since the contributions of these sources are small (totals were below 10% for EC and POC, respectively 33,34). 3.3. Comparison with Source and Other Ambient Data. The pMC and EC/TC (or OC/EC) ratios are important indicators and greatly differ among emission sources (SI Table S1). We therefore attempted to characterize the source
receptor relationship by comparing the relationship between pMC and EC/TC ratio of the FAMIKA samples to the relationship for major sources and other ambient samples reported elsewhere (Figure 3). In Figure 3, all of the source and ambient data lie in the region surrounded by the three end-member sources, that is, fossil-fuel burning (diesel exhaust particles (DEP) and industrial boilers), biogenic sources (biogenic SOA and open burning), and fossil SOA. The data for both Maebashi and Kisai lies between the data for the roadside atmosphere and the suburban site (Tsukuba) in Kanto area. The pMC values at Maebashi and Kisai are lower than the wintertime level at Kudan (Tokyo) and close to the summertime level. It seems reasonable that the EC/ TC ratios at Maebashi and Kisai are smaller than those in Tokyo since combustion sources having large EC/TC ratios such as automobiles, industrial boilers, and other transport modes (SI Table S1) are concentrated around Tokyo. For both Maebashi and Kisai, the nighttime data is close to Line 1, and the daytime points have shifted closer to Line 2. Since fossil SOA could be the driving force toward the origin in the figure, these results suggest that fossil SOA made a greater contribution in the daytime than at night. Fossil fuel and oil itself could produce primary particles, for example, engine-oil derived nanoparticles.58,59 Therefore, they are other possible sources with their data lying around the origin in Figure 3, but their contribution would be very small due to their negligible mass. Thus, this analysis corroborates the dominant role of fossil SOA in the daytime in the suburban areas. Szidat et al.30 and Lewis and Stiles31 suggested that biogenic sources, mostly biogenic SOA, dominated the concentration of OC and TC in ambient aerosols. They reported that the fraction of modern carbon of OC in PM10 was 52
80% (average 69%) at an urban site (Z€urich, Switzerland 30), and the pMC of TC in PM2.5 was 55
95 (average 75) at a suburban site (Tampa, Florida, USA 31). In contrast, Klinedinst and Currie32 reported that the fraction of modern carbon of TC in PM2.5 was 9
70% (average 44%) at an urban site in summer (Welby, Denver, CO). 6789

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology In our study, lower pMC (ND
61, average 37) was obtained at Maebashi and Kisai. These mean pMC values were close to that in Tokyo (≈40 44,47,48). Here, the TC concentration (0.6
13 μg m
3) at our two sites was higher than that in Z€urich (2
9 μg m
3) and Tampa (0.9
7.3 μg m
3), and similar to the level in Welby (2
12 μg m
3). In the Kanto area, especially in Tokyo, there are high densities of humans and automobiles. The difference in density and area of human activity among these cities may have resulted in lower pMC and higher TC and fossil SOA concentration at our sites. As shown above, 6 h or 12 h interval 14C measurements can provide new insights on the origin of fine particles. Further 14C measurements with better time resolution or by carbon fractions are desired. For these measurements, a larger amount of sample is necessary but a larger filter size would make sample preparation more difficult. Therefore, a sampling method which can collect a larger sample amount per filter area needs to be developed.

’ ASSOCIATED CONTENT

bS

Supporting Information. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: +81-29-850-2752; fax: +81-29-850-2574; e-mail: [email protected].

’ ACKNOWLEDGMENT We thank the co-workers of FAMIKA, especially Natsumi Umezawa, Dr. Shinichi Yonemochi (Center for Environmental Science in Saitama), Dr. Akihiro Iijima and Dr. Kimiyo Kumagai (Gunma Prefectural Institute of Public Health and Environmental Sciences) for their help with the field sampling. We also thank Dr. Akiko Kida (National Institute for Environmental Studies) for providing the fly ash samples. We are grateful to Toshiyuki Kobayashi (National Institute for Environmental Studies) for the AMS operation. ’ REFERENCES (1) Dockery, D. W.; Arden Pope, C.; Xu, X.; Spengler, J. D.; Ware, J. H.; Fay, M. E.; Ferris, B. G., Jr.; Speizer, F. E. An association between air pollution and mortality in six U.S. cities. N. Engl. J. Med. 1993, 329, 1753–1759. (2) Harrison, R. M.; Jones, A. M.; Lawrence, R. G. Major component composition of PM10 and PM2.5 from roadside and urban background sites. Atmos. Environ. 2004, 38, 4531–4538. (3) Takahashi, K.; Minoura, H.; Sakamoto, K. Chemical composition of atmospheric aerosols in the general environment and around a trunk road in the Tokyo metropolitan area. Atmos. Environ. 2008, 42, 113–125. (4) Kanakidou, M.; Seinfeld, J. H.; Pandis, S. N.; Barnes, I.; Dentener, F. J.; Facchini, M. C.; Dingenen, R. V.; Ervens, B.; Nenes, A.; Nielsen, C. J.; Swietlicki, E.; Putaud, J. P.; Balkanski, Y.; Fuzzi, S.; Horth, J.; Moortgat, G. K.; Winterhalter, R.; Myhre, C. E. L.; Tsigaridis, K.; Vignati, E.; Stephanou, E. G.; Wilson, J. Organic aerosol and global climate modelling: A review. Atmos. Chem. Phys. 2005, 5, 1053–1123. (5) Zhang, Q.; Jimenez, J. L.; Canagaratna, M. R.; Allan, J. D.; Coe, H.; Ulbrich, I.; Alfarra, M. R.; Takami, A.; Middlebrook, A. M.; Sun, Y. L.; Dzepina, K.; Dunlea, E.; Docherty, K.; DeCarlo, P. F.; Salcedo, D.; Onasch, T.; Jayne, J. T.; Miyoshi, T.; Shimono, A.; Hatakeyama, S.; Takegawa, N.; Kondo, Y.; Schneider, J.; Drewnick, F.; Borrmann, S.;

ARTICLE

Weimer, S.; Demerjian, K.; Williams, P.; Bower, K.; Bahreini, R.; Cottrell, L.; Griffin, R. J.; Rautiainen, J.; Sun, J. Y.; Zhang, Y. M.; Worsnop, D. R. Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophys. Res. Lett. 2007, 34, L13801. (6) Hodzic, A.; Jimenez, J. L.; Prevot, A. S. H.; Szidat, S.; Fast, J. D.; Madronich, S. Can 3-D models explain the observed fractions of fossil and non-fossil carbon in and near Mexico City? Atmos Chem. Phys. 2010, 10, 10997–11016. (7) Morino, Y.; Chatani, S.; Hayami, H.; Sasaki, K.; Mori, Y.; Morikawa, T.; Ohara, T.; Hasegawa, S.; Kobayashi, S. Evaluation of ensemble approach for O3 and PM2.5 simulation. Asian J. Atmos. Environ. 2010, 4, 150–156. (8) Matsui, H.; Koike, M.; Takegawa, N.; Kondo, Y.; Griffin, R. J.; Miyazaki, Y.; Yokouchi, Y.; Ohara, T. Secondary organic aerosol formation in urban air: Temporal variations and possible contributions from unidentified hydrocarbons. J. Geophys. Res. D 2009, 114, D04201. (9) Volkamer, R.; Jimenez, J. L.; Martini, F. S.; Dzepina, K.; Zhang, Q.; Salcedo, D.; Molina, L. T.; Worsnop, D. R.; Molina, M. J. Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than expected. Geophys. Res. Lett. 2006, 33, L17811. (10) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.; Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.; Goldstein, A. H.; Hamilton, J. F.; Herrmann, H.; Hoffmann, T.; Iinuma, Y.; Jang, M.; Jenkin, M. E.; Jimenez, J. L.; Kiendler-Scharr, A.; Maenhaut, W.; McFiggans, G.; Mentel, T. F.; Monod, A.; Prevot, A. S. H.; Seinfeld, J. H.; Surratt, J. D.; Szmigielski, R.; Wildt, J. The formation, properties and impact of secondary organic aerosol: Current and emerging issues. Atmos. Chem. Phys. 2009, 9, 5155–5236. (11) Civerolo, K.; Hogrefe, C.; Zalewsky, E.; Hao, W.; Sistla, G.; Lynn, B.; Rosenzweig, C.; Kinney, P. L. Evaluation of an 18-year CMAQ simulation: Seasonal variations and long-term temporal changes in sulfate and nitrate. Atmos. Environ. 2010, 44, 3745–3752. (12) Carmichael, G. R.; Sakurai, T.; Streets, D.; Hozumi, Y.; Ueda, H.; Park, S. U.; Fung, C.; Han, Z.; Kajino, M.; Engardt, M.; Bennet, C.; Hayami, H.; Sartelet, K.; Holloway, T.; Wang, Z.; Kannari, A.; Fu, J.; Matsuda, K.; Thongboonchoo, N.; Amann, M. MICS-Asia II: The model intercomparison study for Asia Phase II methodology and overview of findings. Atmos. Environ. 2008, 42, 3468–3490. (13) Hayami, H.; Sakurai, T.; Han, Z.; Ueda, H.; Carmichael, G. R.; Streets, D.; Holloway, T.; Wang, Z.; Thongboonchoo, N.; Engardt, M.; Bennet, C.; Fung, C.; Chang, A.; Park, S. U.; Kajino, M.; Sartelet, K.; Matsuda, K.; Amann, M. MICS-Asia II: Model intercomparison and evaluation of particulate sulfate, nitrate and ammonium. Atmos. Environ. 2008, 42, 3510–3527. (14) Lanz, V. A.; Alfarra, M. R.; Baltensperger, U.; Buchmann, B.; Hueglin, C.; Prevot, A. S. H. Source apportionment of submicron organic aerosols at an urban site by factor analytical modelling of aerosol mass spectra. Atmos. Chem. Phys. 2007, 7, 1503–1522. (15) Ding, X.; Zheng, M.; Yu, L.; Zhang, X.; Weber, R. J.; Yan, B.; Russell, A. G.; Edgerton, E. S.; Wang, X. Spatial and seasonal trends in biogenic secondary organic aerosol tracers and water-soluble organic carbon in the Southeastern United States. Environ. Sci. Technol. 2008, 42, 5171–5176. (16) Kleindienst, T. E.; Jaoui, M.; Lewandowski, M.; Offenberg, J. H.; Lewis, C. W.; Bhave, P. V.; Edney, E. O. Estimates of the contributions of biogenic and anthropogenic hydrocarbons to secondary organic aerosol at a southeastern US location. Atmos. Environ. 2007, 41, 8288–8300. (17) Fu, P.; Kawamura, K.; Kanaya, Y.; Wang, Z. Contributions of biogenic volatile organic compounds to the formation of secondary organic aerosols over Mt. Tai, Central East China. Atmos. Environ. 2010, 44, 4817–4826. (18) Lambe, A. T.; Miracolo, M. A.; Hennigan, C. J.; Robinson, A. L.; Donahue, N. M. Effective rate constants and uptake coefficients for the reactions of organic molecular markers (n-alkanes, hopanes, and steranes) in motor oil and diesel primary organic aerosols with hydroxyl radicals. Environ. Sci. Technol. 2009, 43, 8794–8800. 6790

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology (19) Hennigan, C. J.; Sullivan, A. P.; Collett, J. L., Jr.; Robinson, A. L. Levoglucosan stability in biomass burning particles exposed to hydroxyl radicals. Geophys. Res. Lett. 2010, 37, L09806. (20) Kessler, S. H.; Smith, J. D.; Che, D. L.; Worsnop, D. R.; Wilson, K. R.; Kroll, J. H. Chemical sinks of organic aerosol: Kinetics and products of the heterogeneous oxidation of erythritol and levoglucosan. Environ. Sci. Technol. 2010, 44, 7005–7010. (21) Szidat, S.; Jenk, T. M.; Synal, H.-A.; Kalberer, M.; Wacker, L.; Hajdas, I.; Kasper-Giebl, A.; Baltensperger, U. Contributions of fossil fuel, biomass-burning, and biogenic emissions to carbonaceous aerosols in Zurich as traced by 14C. J. Geophys. Res. 2006, 111, D07206. (22) Szidat, S.; Ruff, M.; Perron, N.; Wacker, L.; Synal, H.-A.; Hallquist, M.; Shannigrahi, A. S.; Yttri, K. E.; Dye, C.; Simpson, D. Fossil and non-fossil sources of organic carbon (OC) and elemental carbon (EC) in G€oteborg, Sweden. Atmos. Chem. Phys. 2009, 9, 1521–1535. (23) Kumata, H.; Uchida, M.; Sakuma, E.; Uchida, T.; Fujiwara, K.; Tsuzuki, M.; Yoneda, M.; Shibata, Y. Compound class specific 14C analysis of polycyclic aromatic hydrocarbons associated with PM10 and PM1.1 aerosols from residential areas of suburban Tokyo. Environ. Sci. Technol. 2006, 40, 3474–3480. (24) Weber, R. J.; Sullivan, A. P.; Peltier, R. E.; Russell, A.; Yan, B.; Zheng, M.; de Gouw, J.; Warneke, C.; Brock, C.; Holloway, J. S.; Atlas, E. L.; Edgerton, E. A study of secondary organic aerosol formation in the anthropogenic-influenced southeastern United States. J. Geophys. Res. 2007, 112, D13302. (25) May, B.; Wagenbach, D.; Hammer, S.; Steier, P.; Puxbaum, H.; Pio, C. The anthropogenic influence on carbonaceous aerosol in the European background. Tellus 2009, 61B, 464–472. € Quantification and (26) Zencak, Z.; Elmquist, M.; Gustafsson, O. radiocarbon source apportionment of black carbon in atmospheric aerosols using the CTO-375 method. Atmos. Environ. 2007, 41, 7895– 7906.  € Krusa, (27) Gustafsson, O.; M.; Zencak, Z.; Sheesley, R. J.; Granat, L.; Engstr€om, E.; Praveen, P. S.; Rao, P. S. P.; Leck, C.; Rodhe, H. Brown clouds over south Asia: Biomass or fossil fuel combustion? Science 2009, 323, 495–498  (28) Sheesley, R. J.; Krusa, M.; Krecl, P.; Johansson, C.; Gustafsson, € O. Source apportionment of elevated wintertime PAHs by compoundspecific radiocarbon analysis. Atmos. Chem. Phys. 2009, 9, 3347–3356. (29) Uchida, M.; Kumata, H.; Koike, Y.; Tsuzuki, M.; Uchida, T.; Fujiwara, K.; Shibata, Y. Radiocarbon-based source apportionment of black carbon (BC) in PM10 aerosols from residential area of suburban Tokyo. Nucl. Instrum. Methods Phys. Res., Sect. B 2010, 268, 1120–1124. (30) Szidat, S.; Jenk, T. M.; G€aggeler, H. W.; Synal, H.-A.; Fisseha, R.; Baltensperger, U.; Kalberer, M.; Samburova, V.; Reimann, S.; KasperGiebl, A.; Hajdas, I. Radiocarbon (14C)-deduced biogenic and anthropogenic contributions to organic carbon (OC) of urban aerosols from Z€urich, Switzerland. Atmos. Environ. 2004, 38, 4035–4044. (31) Lewis, C. W.; Stiles, D. C. Radiocarbon content of PM2.5 ambient aerosol in Tampa, FL. Aerosol Sci. Technol. 2006, 40, 189–196. (32) Klinedinst, D. B.; Currie, L. A. Direct quantification of PM2.5 fossil and biomass carbon within the northern front range air quality study’s domain. Environ. Sci. Technol. 1999, 33, 4146–4154. (33) Takahashi, K.; Fushimi, A.; Morino, Y.; Iijima, A.; Yonemochi, S.; Hayami, H., Hasegawa, S.; Tanabe, K.; Kobayashi, S. Source apportionment of ambient fine particles using a receptor model combined with radio carbon content in Northern Kanto area. Jpn. Soc. Atmos. Environ. 2011, 46, 156
163 (in Japanese). (34) Morino, Y.; Takahashi, K.; Fushimi, A.; Tanabe, K.; Ohara, T.; Hasegawa, S.; Uchida, M.; Takami, A.; Yokouchi, Y.; Kobayashi, S. Contrasting diurnal variations in fossil and nonfossil secondary organic aerosol in urban outflow, Japan. Environ. Sci. Technol. 2010, 44, 8581–8586. (35) Report on Exposure and Health Effect of Fine Particulate Matter; Ministry of the Environment of Japan, 2007. (36) Mizuno, T.; Kondo, H. Generation of a local front and high levels of air pollution on the Kanto plain in early winter. Atmos. Environ. A 1992, 26, 137–143.

ARTICLE

(37) Wakamatsu, S.; Uno, I.; Ohara, T.; Schere, K. L. A study of the relationship between photochemical ozone and its precursor emissions of nitrogen oxides and hydrocarbons in Tokyo and surrounding areas. Atmos. Environ. 1999, 33, 3097–3108. (38) Kumagai, K.; Iijima, A.; Tago, H.; Tomioka, A.; Kozawa, K.; Sakamoto, K. Seasonal characteristics of water-soluble organic carbon in atmospheric particles in the inland Kanto plain, Japan. Atmos. Environ. 2009, 43, 3345–3351. (39) Kannari, A.; Tonooka, Y.; Baba, T.; Murano, K. Development of multiple-species 1 km  1 km resolution hourly basis emissions inventory for Japan. Atmos. Environ. 2007, 41, 3428–3439. (40) Chow, J. C.; Watson, J. G.; Crow, D.; Lowenthal, D. H.; Merrifield, T. Comparison of IMPROVE and NIOSH carbon measurements. Aerosol Sci. Technol. 2001, 34, 23–34. (41) Method 5040 Issue 3 (Interim): Elemental Carbon (Diesel Particulate). NIOSH manual of analytical methods; NIOSH, 1999. (42) Cachier, H.; Bremond, M.-P.; Buat-Menard, P. Determination of atmospheric soot carbon with a simple thermal method. Tellus 1989, 41B, 379–390. (43) Cavalli, F.; Viana, M.; Yttri, K. E.; Genberg, J.; Putaud, J.-P. Toward a standardised thermal-optical protocol for measuring atmospheric organic and elemental carbon: the EUSAAR protocol. Atmos. Meas. Tech. 2010, 3, 79–89. (44) Endo, M.; Yamamoto, N.; Yoshinaga, J.; Yanagisawa, Y.; Endo, O.; Goto, S.; Yoneda, M.; Shibata, Y.; Morita, M. 14C measurement for size-fractionated airborne particulate matters. Atmos. Environ. 2004, 38, 6263–6267. (45) Uchida, M.; Shibata, Y.; Yoneda, M.; Kobayashi, T.; Morita, M. Technical progress in AMS microscale radiocarbon analysis. Nucl. Instrum. Methods Phys. Res., Sect. B 2004, 223
224, 313–317. (46) Kume, H.; Shibata, Y.; Tanaka, A.; Yoneda, M.; Kumamoto, Y.; Uehiro, T.; Morita, M. The AMS facility at the National Institute for Environmental Studies (NIES), Japan. Nucl. Instrum. Methods Phys. Res., Sect. B 1997, 123, 31–33. (47) Takahashi, K.; Hirabayashi, M.; Tanabe, K.; Shibata, Y.; Nishikawa, M.; Sakamoto, K. Radiocarbon content in urban atmospheric aerosols. Water, Air, Soil Pollut. 2007, 185, 305–310. (48) Yamamoto, N.; Muramoto, A.; Yoshinaga, J.; Shibata, K.; Endo, M.; Endo, O.; Hirabayashi, M.; Tanabe, K.; Goto, S.; Yoneda, M.; Shibata, Y. Comparison of carbonaceous aerosols in Tokyo before and after implementation of diesel exhaust restrictions. Environ. Sci. Technol. 2007, 41, 6357–6362. (49) Hirabayashi, M.; Shibata, K.; Muramoto, A.; Yoshinaga, J.; Kobayashi, S.; Yoneda, M.; Tanabe, K.; Shibata, Y.; Matsuo, M. Investigation of source apportionment for carbon component in airborne particulate based on 14C measurements of organic and elemental carbons. 10th Int. Conf. Acc. Mass Spectrom. (AMS-10), 2005. (50) Bennett, M.; Volckens, J.; Stanglmaier, R.; McNichol, A. P.; Ellenson, W. D.; Lewis, C. W. Biodiesel effects on particulate radiocarbon (14C) emissions from a diesel engine. J. Aerosol Sci. 2008, 39, 667–678. (51) Lemire, K. R.; Allen, D. T.; Klouda, G. A.; Lewis, C. W. Fine particulate matter source attribution for Southeast Texas using 14C/13C ratios. J. Geophys. Res. 2002, 107 (D22), 4613. (52) Hildemann, L. M.; Klinedinst, D. B.; Klouda, G. A.; Currie, L. A.; Cass, G. R. Sources of urban contemporary carbon aerosol. Environ. Sci. Technol. 1994, 28, 1565–1576. (53) Han, Y.; Cao, J.; Chow, J. C.; Watson, J. G.; An, Z.; Jin, Z.; Fung, K.; Liu, S. Evaluation of the thermal/optical reflectance method for discrimination between char- and soot-EC. Chemosphere 2007, 69, 569–574. (54) Kleeman, M. J.; Robert, M. A.; Riddle, S. G.; Fine, P. M.; Hays, M. D.; Schauer, J. J.; Hannigan, M. P. Size distribution of trace organic species emitted from biomass combustion and meat charbroiling. Atmos. Environ. 2008, 42, 3059–3075. (55) Offenberg, J.; Lewis, C. W.; Lewandowski, M.; Jaoui, M.; Kleindienst, T. E.; Edney, E. O. Contributions of toluene and R-pinene to SOA formed in an irradiated toluene/R-pinene/NOx/air mixture: 6791

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792

Environmental Science & Technology

ARTICLE

Comparison of results using 14C content and SOA organic tracer methods. Environ. Sci. Technol. 2007, 41, 3972–3976. (56) Hildemann, L. M.; Markowski, G. R.; Cas, G. R. Chemical composition of emissions from urban sources of fine organic aerosol. Environ. Sci. Technol. 1991, 25, 744–759. (57) Gelencser, A.; May, B.; Simpson, D.; Sanchez-Ochoa, A.; Kasper-Giebl, A.; Puxbaum, H.; Caseiro, A.; Pio, C.; Legrand, M. Source apportionment of PM2.5 organic aerosol over Europe: Primary/secondary, natural/anthropogenic, and fossil/biogenic origin. J. Geophys. Res. 2007, 112, D23S04. (58) Sakurai, H.; Tobias, H. J.; Park, K.; Zarling, D.; Docherty, K. S.; Kittelson, D. B.; McMurry, P. H.; Ziemann, P. J. On-line measurements of diesel nanoparticle composition and volatility. Atmos. Environ. 2003, 37, 1199–1210. (59) Fushimi, A.; Hasegawa, S.; Takahashi, K.; Fujitani, Y.; Tanabe, K.; Kobayashi, S. Atmospheric fate of nuclei-mode particles estimated from the number concentrations and chemical composition of particles measured at roadside and background sites. Atmos. Environ. 2008, 42, 948–958.

6792

dx.doi.org/10.1021/es201400p |Environ. Sci. Technol. 2011, 45, 6784–6792