A comparison study on airborne particles during haze days and non-haze days in Beijing

Science of the Total Environment 456–457 (2013) 1–8

Contents lists available at SciVerse ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

A comparison study on airborne particles during haze days and non-haze days in Beijing Zhenquan Sun a, b, Yujing Mu a,⁎, Yanju Liu b, Longyi Shao c a b c

Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China Beijing Centre for Physical and Chemical Analysis, Beijing 100089, China State Key Laboratory of Coal Resources and Safe Mining & College of Geoscience and Surveying Engineering, China University of Mining and Technology, Beijing 100083, China

H I G H L I G H T S • Water soluble inorganic ions in different size airborne particles during haze and non-haze days in Beijing were analyzed. • NH4+, K+, NO3− and SO42− were concentrated in fine particles around 0.56–1.0μm during haze days, while 0.32–0.56μm during non-haze days. • NOx from the vehicle exhaust is playing an important role on fine particle formation.

a r t i c l e

i n f o

Article history: Received 26 November 2012 Received in revised form 1 March 2013 Accepted 2 March 2013 Available online 9 April 2013 Keywords: Haze Water soluble inorganic ions Mass concentration Size distribution

a b s t r a c t Airborne particles in Beijing during haze days and non-haze days were collected by an eleven-stage cascade impactor (MOUDI 110, MSP, USA), and the mass concentrations and water soluble inorganic ions of the size segregated airborne particles were quantitatively analyzed. PM10 concentrations during haze days ranged from 250.5 to 519.4 μg m −3 which were about 3–8 times greater than those (ranged from 67.6 to 94.0 μg m −3) during non-haze days, and PM1.8 concentrations during haze periods were in the range of 117.6–378.6 μg m −3 which were 3–14 times higher than those (27.0 to 36.8 μg m−3) during non-haze days. In comparison with non-haze days, all water soluble inorganic ions investigated in the airborne particles greatly enhanced during haze days. NH4+, NO3− and SO42− were found to be the dominant water soluble inorganic ions, accounting for 91–95% of the total inorganic ions in PM1.8 during haze days, and 73–81% during non-haze days. The size distributions of SO42−, NO3−, Cl −, K+ and Na+ exhibited bimodal types, while single mode was found for NH4+, Ca2+ and Mg2+. Only with exception of Ca2+ and Mg2+, all ions were concentrated in fine particles around 0.56–1.0 μm of “droplet mode” during haze days, while 0.32–0.56 μm of “condensation mode” during non-haze days. The extremely high mole ratio (> 2) of [NH4+]/[SO42−] during haze days implied that the main form of ammonium in PM1.8 might be (NH4)2SO4 and NH4NO3. The mass ratio of NO3−/SO42− was > 1 in PM1.8 during haze days and ~1 during non-haze days, indicating that NOx from the vehicle exhaust in Beijing is playing more and more important role on fine particle formation. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Airborne particles have important adverse influence on ecosystem, especially on human health and visibility (IPCC, 2001; Okada et al., 2001; NRC, 1998; Watson, 2002), and can provide surface for atmospheric heterogeneous (or multiphase) reactions to influence on atmospheric chemistry (Ravishankara, 1997). Human activities have greatly enhanced the concentration of airborne particles, especially in urban areas. Serious pollution of airborne particles in most of Chinese cities has aroused great attention to public and government in recent years, which has pushed stricter air quality standard to be legislated. Although the atmospheric concentrations of primary pollutants (such as SO2, ⁎ Corresponding author. Tel.: +86 10 62849125; fax: +86 10 62849117. E-mail address: [email protected] (Y. Mu). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.03.006

PM10, NOx) have evidently decreased under a series air clean actions during the last decade, more frequent haze days are being observed in most Chinese cities along east coast (Sun et al., 2006b; Wang et al., 2006; Fu et al., 2008; Tan et al., 2009; Li et al., 2010; Du et al., 2011). Haze formation has been well recognized to be dependent on atmospheric relative humidity and the concentrations of airborne particles, especially for the water soluble components in them (Brown et al., 2002; Jacobson, 2001; Chen et al., 2003; Kang et al., 2004). Therefore, study on various water soluble components in the airborne particles can provide useful information for recognizing the key reasons of haze formation in the urban areas. Water soluble inorganic ions (WSIIs), such as NO3−, SO42− and NH4+, are considered as important contributors to visibility impairment (Brown et al., 2002; Jacobson, 2001; Kang et al., 2004). Previous studies showed that WSIIs accounted for one third or more of the fine

2

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

particles in Chinese cities (He et al., 2001; Hu et al., 2002; Wang et al., 2002). The investigation of Wang et al. (2006) revealed that the formation of sulfate and nitrate was accelerated during haze days in Beijing. Li et al. (2010) found that K-rich particles, regard as the tracer of biomass burning, existed in the particles in Beijing during haze days. Du et al. (2011) also found that the maximum K + during haze days was about 19 times higher than the average value of clear days in Shanghai. Schichtel et al. (2001) presented the patterns and trends of haze over the United States, and found that the haze decline was consistent with the reduction of PM2.5 and sulfur emissions. Senaratne and Shooter (2004) found that the accumulation of diesel emissions contributed the most to the appearance of brown haze in Auckland. Beijing, the capital of China, is among the cities with frequent haze episodes in recent years (Che et al., 2009). The number of haze days in Beijing increased from about 50 days in 1980 to 72 days in 2008 (Zhao et al., 2012). The most frequent haze episodes in Beijing usually occur during winter season and in October (Li et al., 2012a). The extra pollutants' emissions from heating were suspected to be responsible to the high frequency of haze days in winter, and the frequent haze days in October are coincident with the prevailing biomass burning (maize leftover) in the countries around Beijing(Zheng et al., 2005). Although many studies have focused on the variation, size distribution, and formation mechanisms of the airborne particles in Beijing (Yao et al., 2003; Wang et al., 2006; Sun et al., 2006b; Mao et al., 2011), the studies on water soluble inorganic ions of different diameter airborne particles during haze days are still very sparse. In this study, airborne particles with eleven aerodynamic equivalent diameters (ranged from 0.056 μm to 18 μm) were collected by a cascade impactor, and the size distributions of water soluble inorganic ions during haze days and non-haze days were comparably investigated to recognize the relative contribution of various water soluble inorganic ions to haze formation. 2. Experimental 2.1. Filter sampling Airborne particles during haze days and non-haze days were collected on polycarbonate filters (47 mm diameter, Millipore, UK, pore size 0.6 μm) by an eleven-stage cascade impactor with flow rate of 30 L min−1(MOUDI 110, MSP, USA) on the roof (~15 m high) of a building at Beijing Academy of Science and Technology. The site is located nearby West 3rd Ring traffic road about 120 m. The MOUDI 110 sampler segregates particles in the following size stage: >18, 10–18, 5.6–10, 3.2–5.6, 1.8–3.2, 1.0–1.8, 0.56–1.0, 0.32–0.56, 0.18–0.32, 0.1–0.18 and 0.056–0.1 μm. The filters were kept in vacuum desiccators for 24 h to

remove any moisture content before mounting them on the MOUDI sampler. After sampling, filters were immediately transferred to the vacuum desiccators again to de-moisturize in the same manner. All of those filters were weighed before and after sampling with an analytical balance (Sartorius CP225D, 0.01 mg, Germany). Totally 7 sets of samples were collected by the MOUDI sampler during haze days and nonhaze days (Table 1). Blank filters (7 pieces of filters) were also kept during each sampling day. In this study, haze days referred to typical days with atmospheric visibility less than 10 km and relative humid (RH) below 90% (Sun et al., 2006b; Du et al., 2011), and non-haze days represented clear days which visibility is more than 10 km. To recognize the influence of biomass burning on haze formation, this study mainly chose the days in October for sampling, and air samples on one haze day and one non-haze day in April were also collected for comparison. Wind speed and direction, relative humidity, barometric pressure, and ambient temperature were automatically recorded by a Kestral 4500 Pocket Weather Tracker (Nielsen-Kellerman Inc., USA). Visibility was monitored by Visibility Sensor, Belfort Model 6000 (USA). 2.2. Chemical analysis Half of each sample filter and blank filter was cut and extracted by 5 mL deionized water (Millipore,18.2 MΩ) respectively, using an ultrasonic bath for 20 min at room temperature, the extraction liquid was filtered using 0.22 μm filters, and then quickly analyzed by Ion Chromatograph (DIONEX, ICS-900). Anions were analyzed by Dionex AS14 Column, ASRS-I suppressor, the eluent was 8 mmol L −1 Na2CO3/1.0 mmol L −1 NaHCO3 with a flow rate of 1.0 mL min −1. Cations were analyzed by CS-12A Column, CSRS-I suppressor, the eluent was 20 mmol L −1MSA with a flow rate of 1.0 mL min −1. Sample volume of 250 μL was used for both anion and cation analyses, respectively. The detection limits (S/N = 3) of the methods were less than 0.01 mg L −1 for anions and 0.005 mg L −1 for cations, and both anion and cation in all samples were above the detection limits. Standard solutions of both anion and cation used for calibrations were purchased from the National Research Center of Certified Reference Materials, China. 3. Results and discussion 3.1. Particle mass size distributions Since MOUDI 110 does not have a 2.5 μm cut-point, 1.8 μm is defined as the boundary between fine and coarse particles, and PM1.8 and PM1.8–10 are designated as fine and coarse particle fractions (Zhuang et al., 1999; Yao et al., 2003), respectively. Fig. 1 shows the

Table 1 Information of collected airborne particle samples. Sample

SD/h

2010–10–07

18

2010–10–08

36

2010–10–30

45

2010–11–01

48

2011–04–13

19

2011–04–14

25.5

2011–10–31

8

Start time (local time)

End time (local time)

2010/10/07 15:00 2010/10/08 09:00 2010/10/30 12:00 2010/11/01 09:00 2011/04/13 17:00 2011/04/14 15:00 2011/10/31 09:30

2010/10/08 09:00 2010/10/09 21:00 2010/11/01 09:00 2010/11/03 09:00 2011/04/14 12:00 2011/04/15 16:30 2011/10/31 17:30

WS/ms−1

T/°C

Climate

Min

Ave

Max

Min

Ave

Max

0.5

1.1

1.8

15

17.9

23.3

Haze

0.4

1.3

2.1

14

19.1

23.9

Haze

0.3

1.8

4.2

7.1

13.7

21.9

Non-haze

0.3

2.1

5.7

3.8

9.4

16.1

Non-haze

1.5

3.4

5.6

13.2

20.3

31.2

Haze

2.1

3.6

6.4

19.3

24

31.8

Non-haze

0.3

1.5

5.3

12

14

15

Haze

SD: sampling duration, WS: wind speed, T: temperature, Min: minimum, Ave: average, Max: maximum.

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

3

35

100

Vis

RH

90

30

70

20

60 50

15

40

RH(%)

Vis(Km)

80 25

30

10

20 5

10

Mass concentration(µgm-3)

0 600

0 Haze

500

Haze

400

Haze

300

Haze

200 Non-Haze

100 0

2010-10-7

2010-10-8

5.6 - 10um 0.32 - 0.56um

2010-10-30

3.2 - 5.6um 0.18 - 0.32um

Non-Haze

Non-Haze

2010-11-1

1.8 - 3.2um 0.10 - 0.18um

2011-4-13

2011-4-14 2011-10-31

1.0 - 1.8um 0.056 - 0.10um

0.56 - 1.0um

Fig. 1. The visibility (Vis), relative humidity (RH), and mass concentrations of size-segregated particles during the sampling days.

visibility (Vis), relative humidity (RH), and the mass concentrations of the size-segregated particles for each sampling day. PM10 concentrations that ranged from 250.5 to 519.4 μg m −3 during haze days were much greater than Chinese daily standard value of 150 μg m − 3 . PM1.8 concentrations that ranged from 117.6 to 378.6 μg m −3 during haze periods greatly exceeded the National Ambient Air Quality Standard (NAAQS) daily average value of 35 μg m −3 for PM2.5 (US EPA, 2006). PM10 concentrations during haze days were 3–8 times higher than those (ranged from 67.6 to 94.0 μg m −3) during nonhaze days, and PM1.8 concentrations during haze days were 3–14 times higher than those (27.0 to 36.8 μg m −3) during non-haze days. It is evident that the extremely high concentrations of the airborne particles during haze days in Beijing are badly threatening the health of citizens with large population. The ratios of PM1.8/PM10 (Fig. 2) during haze days were greater than 0.47, but were less than 0.42 during non-haze days. In addition, the linear correlation (correlation coefficient R 2 = 0.9889) between PM1.8 and PM10 during haze days was more remarkable than that (R2 = 0.5546) during non-haze days. Atmospheric coarse particles are usually

considered from primary sources (such as crustal sources), while atmospheric fine particles are from both primary sources and secondary formation in the atmosphere. The higher ratios of PM1.8/PM10 during haze days indicated evident new formation of fine particles which has been observed by many field measurements (Tan et al., 2009; Du et al., 2011; Yin et al., 2012). The relatively stable meteorological conditions and low boundary layer during haze days favor the accumulation of atmospheric pollutants including airborne particles (Quan et al., 2013; Li and Shao, 2010; Sun et al., 2006b), resulting in the remarkable linear correlation between PM1.8 and PM10. During non-haze days, new fine particles were indeed being formed, but the faster diffusion rates of the fine particles than the coarse particles due to the relatively higher boundary layer would result in relatively low ratios of PM1.8/PM10 and less significant correlation between PM1.8 and PM10. The mass size distributions for the samples are shown in Fig. 3. Bimodal distributions of the particle mass size were found during both haze days and non-haze days, but were more pronounced during haze days than during non-haze days. The peaks of coarse mode at 3.2–5.6 μm were consistent with the results of other field measurement

0.8

Ratios of PM1.8/PM10

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 2010-10-7 Haze

2010-10-8 Haze

2011-4-13 Haze

2011-10-31 Haze

2010-10-30 Non-Haze

2010-11-1 Non-Haze

2011-4-14 Non-Haze

Fig. 2. Ratios of PM1.8/PM10 for the samples collected on haze days and non-haze days.

4

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

900

2010-10-7 Haze

dC/dlgDp

800

2010-10-8 Haze

700

2011-4-13 Haze

600

2011-10-31 Haze

500

2010-11-1 Non-Haze

400

2010-10-30 Non-Haze 2011-4-14 Non-Haze

300 200 100 0 0.01

0.1

1

10

Dp(µm) Fig. 3. Particle mass size distributions during haze days and non-haze days.

(Hu et al., 2005). The peaks of the fine mode were at 0.56–1.0 μm during haze days whereas at 0.32–0.56 μm during non-haze days. Because the concentrations of fine particles were significantly higher during haze days than those during non-haze days, the higher frequency of fine particles coagulation during haze days favored larger fine particle formation (Kulmala et al., 2004). In contrast to the haze days in October, the peak value of coarse mode at 3.2–5.6 μm on the haze day in April was greater than that of the fine mode at 0.56–1.0 μm. The relatively high wind speed on the day in April (Table 1) facilitated re-suspending road dust, resulting in the relatively high value of the coarse mode. 3.2. Concentrations of water-soluble ionic species Water-soluble inorganic ions (WSIIs) of NO3−, SO42−, NH4+, Cl −, K +, Ca 2 +, Mg 2+ and Na + are important species in atmospheric particles. The mass concentrations of WSIIs in PM1.8 and PM10 during haze days were significantly higher than those during non-haze days (Table 2). The total concentrations of the inorganic ions accounted for 28.39–54.95% (average 42.79 ± 12.3%) of the total PM1.8 mass concentrations during the haze days, whereas only 10.71–27.91% (average 21.86 ± 9.67%) during non-haze days. The results obtained by this study were comparable to the previous observation in Beijing (53% of PM2.5 during haze days and 32% during clear days, Wang et al., 2006) and those in Guangzhou (44.9 ± 8.7% and 24.7 ± 4.9% of PM2.5 during haze and non-haze days, respectively, Tan et al., 2009). The major ions of NH4+, NO3− and SO42− contributed 91–95% of the total inorganic ions in PM1.8 during haze days, and 73–81% during non-haze days. The average concentration of NH4+, NO3− and SO42− during haze days was 16 times more than that during non-haze days, suggesting that NH4+, NO3− and SO42− play an important role in the formation of haze.

As listed in Table 2, the concentration of SO42− was the highest WSII in PM1.8 during non-haze days, while NO3− was the highest WSII during haze days, which were consistent with the results observed during haze days and non-haze days in Shanghai by Du et al. (2011) and in Guangzhou by Tan et al. (2009), but were different from the earlier investigation during 2003–2004 in Beijing (Wang et al., 2006). Tan et al. (2009) pointed out that the highest NO3− concentration observed during haze days was ascribed to the high concentration of NOx which greatly surpassed that of SO2 during haze days. On the other hand, atmospheric concentration of H2O2 could be greatly reduced (Poppe et al., 1993) under high NOx condition, and further decreased the possibility of SO42− formation via its dominant formation channel of multiphase reactions. In addition, long range transport from power plant has become the dominant source of SO2 in Beijing, and the relatively stable meteorological conditions during haze days may greatly reduce SO2 diffusion from the upper atmosphere to the earth's surface (Sun et al., 2006a; Li et al., 2012b). 3.3. Mass size distributions of water-soluble ionic species The size distributions of ionic species are shown in Fig. 4. Similar to the mass size distributions, SO42−, NO3−, Cl−, K + and Na+ all exhibited bimodal types, which was in agreement with other investigations (Yao et al., 2003; Wang et al., 2006). The ionic species in the coarse particle mode peaked at 3.2–5.6 μm, and in the fine mode mainly peaked at 0.56–1.0 μm during haze days and at 0.32–0.56 μm during non-haze days. NH4+ was found to be a single mode, mainly peaking at 0.56– 1.0 μm during haze days and at 0.32–0.56 μm during non-haze days, which was in good agreement with previous studies in Hong Kong (Zhuang et al., 1999) and in Jinan City (Yu et al., 2011). The single coarse mode for Ca 2+ and Mg 2+ peaking at 3.2–5.6 μm during both haze and non-haze days was in good agreement with previous studies (Sun et al., 2006b; Wang et al., 2006). The peak values of sulfate and nitrate in the fine mode at 0.32– 0.56 μm during non-haze days were within the “condensation mode” which is usually attributed to the transformation of SO2 and NOx via heterogeneous reactions (Yao et al., 2003), and at 0.56– 1.0 μm during haze days was “droplet mode” which is likely formed in-cloud or through aqueous-phase chemical reactions (Meng and Seinfeld, 1994; Kerminen and Wexler, 1995). The relative high humidity during haze days facilitates the formation of “droplet mode” particles. 3.4. Source analysis of WSIIs Based on the data listed in Table 2, the calculated ratios of K +, Cl − and NH4+ on the two haze days on 7 and 8 October to those on the

Table 2 Concentrations of water soluble ions in PM1.8 and PM10 during haze and non-haze. Species (μg m−3)

Mass SO42− NO3− NH4+ Ca2+ Mg2+ K+ Cl− Na+ TWSI TWSI/PM (%) NO3−/SO42−

2010–10–7 Haze

2010–10–8 Haze

2011–4–13 Haze

2011–10–31 Haze

2010–10–30 Non-haze

2010–11–1 Non-haze

2011–4–14 Non-haze

PM1.8

PM10

PM1.8

PM10

PM1.8

PM10

PM1.8

PM10

PM1.8

PM10

PM1.8

PM10

PM1.8

PM10

378.62 24.51 45.60 27.97 1.02 0.14 3.85 3.32 1.09 107.50 28.39 1.86

519.40 26.93 57.87 28.95 8.48 1.04 4.27 4.69 1.43 133.66 25.73

294.23 25.14 53.03 24.41 0.70 0.08 2.47 2.25 0.99 109.07 37.07 2.11

454.04 30.35 72.72 26.10 7.98 0.88 2.98 3.93 1.58 146.52 32.27

117.56 17.28 32.91 11.33 0.80 0.11 1.12 1.05 – 64.60 54.95 1.90

250.49 20.09 45.22 11.68 8.65 0.77 1.33 2.44 – 90.18 36.00

184.39 28.11 42.46 18.32 0.61 0.05 1.23 2.82 – 93.60 50.76 1.51

341.30 37.44 53.32 22.26 5.09 0.62 1.44 3.86 – 124.03 36.34

36.76 3.15 2.94 2.15 0.55 0.14 0.57 0.45 0.31 10.26 27.91 0.93

93.97 4.12 5.44 2.23 4.51 0.45 0.67 0.67 0.44 18.53 19.72

28.62 2.69 2.63 0.92 0.26 0.06 0.36 0.67 0.13 7.72 26.97 0.98

67.62 3.13 2.96 0.93 2.57 0.23 0.45 0.80 0.20 11.27 16.67

35.11 1.66 0.86 0.51 0.46 0.09 0.13 0.05 – 3.76 10.71 0.52

106.59 3.10 1.51 0.52 2.37 0.39 0.17 0.18 – 8.24 7.73

TWSI: total water soluble ions, PM: particles mass.

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

60

40 30

2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

100

dC/dlgDp

dC/dlgDp

120

2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

50

5

20 10

80 60 40 20

0 0.1

1.0

10.0

0 0.1

100.0

1.0

10.0

Size distribution of NO3-

Size distribution of SO4270

16.0 2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

dC/dlgDp

50 40 30

12.0 10.0

20

8.0 6.0 4.0

10

2.0

0 0.1

0.0 1.0

10.0

100.0

0.1

1

10

Dp (µm)

Dp (µm)

Size distribution of NH4+

Size distribution of Ca2+

1.80

100

9.0 2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

1.40 1.20 1.00 0.80

7.0

0.60

6.0 5.0 4.0 3.0

0.40

2.0

0.20

1.0

0.00 0.1

2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

8.0

dC/dlgDp

1.60

dC/dlgDp

2010-10-7 Haze 2010-10-8 Haze 2011-4-13 Haze 2011-10-31 Haze 2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

14.0

dC/dlgDp

60

1.0

10.0

0.0

100.0

0.1

1.0

Dp (µm)

10.0

100.0

Dp (µm)

Size distribution of Mg2+

Size distribution of K+

7.0

3.0 2010-10-7 Haze

6.0

2010-10-8 Haze

5.0

2011-4-13 Haze

2010-10-7 Haze

2.5 2010-10-8 Haze

2.0

2010-10-30 Non-Haze

4.0

2010-11-1 Non-Haze

3.0

2011-4-14 Non-Haze

dC/dlgDp

dC/dlgDp

100.0

Dp (µm)

Dp (µm)

2.0

2010-10-30 Non-Haze 2010-11-1 Non-Haze

1.5 1.0 0.5

1.0 0.0

0.0 0.1

1.0

10.0

100.0

0.1

1

10

Dp (µm)

Dp (µm)

Size distribution of ClFig. 4. The mass size distributions of inorganic ionic species.

Size distribution of Na+

100

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

1.8 y = 0.630 x + 0.238 R² = 0.835

1.6

Cl-(µgm-3)

1.4 1.2 1.0

y = 0.780 x + 0.196 R² = 0.638

y = 1.939x + 0.141 R² = 0.511

2010-10-7Haze 2010-10-8Haze 2011-4-13Haze 2011-10-31Haze 2010-10-7Haze 2010-10-8Haze 2011-4-13Haze 2011-10-31Haze

0.8 0.6 y = 0.584 x + 0.196 R² = 0.156

0.4 0.2 0.0 0.0

0.5

1.0

1.5

2.0

2.5

K+(µgm-3) 0.30

Cl-(µgm-3)

0.25 0.20

y = 2.073 x - 0.017 R² = 0.945

y = 0.752 x + 0.019 R² = 0.716

0.15

2010-10-30Non-haze 2010-11-1Non-haze 2011-4-14Non-haze 2010-10-30Non-haze 2010-11-1Non-haze 2011-4-14Non-haze

0.10 0.05 0.00 0.00

y = -0.306 x + 0.024 R² = 0.092

0.05

0.10

0.15

0.20

0.25

0.30

K+(µgm-3) Fig. 5. Correlation between K+and Cl− concentrations during haze days and non-haze days.

0.45 0.40 0.35

Mg2+(µgm-3)

haze day in April were remarkably greater than the ratios of other WSIIs. Because K + and Cl − are typical components from biomass burning (Reid et al., 2005; Adachi and Buseck, 2008; Engling et al., 2009) which usually occurs at the beginning days in October just after harvest of Maize in the countries around Beijing, biomass burning was suspected to make great contribution to atmospheric particles in October. As shown in Fig. 5, the significant positive correlation between K + and Cl − during both haze days and nonhaze days in October confirmed the contribution of biomass burning to atmospheric particles. However, no significant correlations between K + and Cl − were found in April during both haze days and non-haze days. Abundant KCl particles had been detected in fresh smoke plumes of biomass burning (Engling et al., 2009; Li et al., 2003; P'osfai et al., 2003; Adachi and Buseck, 2008). Du et al. (2011) also found that KCl was present in particles during haze days in Shanghai. The K + in airborne particles is so soluble in water that their associated particles are expected to be active as cloud condensation nuclei (CCN) (Du et al., 2011), and may therefore significantly contribute to the reduction of radiative forcing and visibility (Liu et al., 2000; Chand et al., 2006). Atmospheric Mg 2+ and Ca 2+ have been recognized to be mainly from crustal sources, such as re-suspended road dust, soil dust, and construction dust. As expected, significant positive correlations during both haze days and non-haze days were found (Fig. 6). Because the stagnation meteorological condition during haze days favors the accumulation of Ca 2+ and Mg 2+ originated from the crustal sources, the concentrations of Mg 2+ and Ca 2+ observed during haze days were about 2–3 times higher than those during non-haze days in PM1.8 and PM10, which is consistent with the results in Guangzhou (Tan et al., 2009). SO42−, NO3−, and NH4+ in the airborne particles are usually considered as secondary pollutants from transformation of their precursors of SO2, NOx and NH3 in the atmosphere (Wang et al., 2005). During haze days, the average concentrations of SO42−, NO3−, and NH4+ (23.76, 43.5 and 20.51 μg m −3 in PM1.8, and 28.70, 57.28 and 22.25 μg m −3 in PM10, respectively) were 10–20 times higher than those during

2010-10-7Haze 2011-4-13Haze 2010-10-7Haze 2011-4-13Haze

2010-10-8Haze 2011-10-31Haze 2010-10-8Haze 2011-10-31Haze

0.30 0.25

y = 0.109 x + 0.018 R² = 0.973 y = 0.095 x + 0.010 R² = 0.870

y = 0.125 x -0.001 R² = 0.992

0.20

y = 0.072 x + 0.020 R² = 0.887

0.15 0.10 0.05 0.00 0.0

1.0

2.0

3.0

4.0

Ca2+(µgm-3) 0.14 y = 0.056 x + 0.025 R² = 0.752

0.12 y = 0.163 x + 0.003 R² = 0.979

Mg2+(µgm-3)

6

0.10

y = 0.050 x + 0.016 R² = 0.843

0.08

2010-10-30Non-haze

0.06

2010-11-1Non-haze 2011-4-14Non-haze

0.04

2010-10-30Non-haze

0.02 0.00 0.0

2010-11-1Non-haze 2011-4-14Non-haze

0.5

1.0

1.5

2.0

Ca2+(µgm-3) Fig. 6. Correlations between Ca2+and Mg2+concentrations during haze days and non-haze days.

non-haze days. In comparison with other ions, the greater enhancements of SO42−, NO3−, and NH4+ during haze days were probably ascribed to fast conversion of their precursors via multiphase reactions under the relatively high humidity (Khoder, 2002; Xiu et al., 2004; Sun et al., 2006b). Because ammonium nitrate has much higher vapor pressure than ammonium sulfate, only when abundant of NH3 and high humidity can be favorable to form NH4NO3 (Calvert et al., 1978; Stelson and Seinfeld, 1982). The mole ratio of [NH4+]/[SO42−] were more than 2 in the airborne particles (aerodynamics diameter ≤ 1.8 μm) during haze days (Fig. 7), suggesting that the main forms of ammonium in PM1.8 might be (NH4)2SO4 and NH4NO3 during haze days. The mass ratio of NO3−/SO42− in airborne particles has been used as an indicator of the relative importance of stationary sources versus mobile sources (Arimoto et al., 1996; Yao et al., 2002). Only with the exception on the non-haze day on 14, April 2011, the mass ratios of NO3−/SO42− were all >0.9 in PM1.8 during both haze days and non-haze days (Table 2), which was much greater than those reported in 1990s (Zhou et al., 1998; Yao et al., 2002). Both the fast increasing car numbers and the strict air clean actions for the reduction of atmospheric SO2 in recent years are responsible for the change of the NO3−/SO42− ratio. The car number in Beijing has increased from about 1 million to 5 million in the recent ten years, making significant contribution to atmospheric NOx which is contributing to NO3− via atmospheric photochemical conversions. On the other hand, the major sources for atmospheric SO2 in Beijing have been greatly reduced by a series of air clean actions, e.g., most boilers with coal as fuel have been replaced by natural gas or oil or electricity. The ratios of NO3−/SO42− during haze days (> 1) and non-haze days (0.52–~ 1) investigated in this study were consistent with those measured in Guangzhou, a value of 1.02 during haze days and of 0.55 during non-haze days (Tan et al., 2009), but was inconsistent with previous study in Beijing by Wang et al. (2006) who reported the ratio of 0.89 in haze days and 0.96 in normal days (0.96). The minimal NO3−/SO42− ratio of 0.52 was measured on the non-haze day of 14, April 2011 when the average wind speed was the maximum among

[NH4+](µmolm-3)

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

Conflict of interest

2010-10-7 Haze 2010-10-8 Haze

There is no conflict of interest.

2011-4-13 Haze 2011-10-31 Haze

Acknowledgments

y = 2x

0

0.05

0.1

0.15

[SO42-](µmolm-3)

2010-10-30 Non-Haze 2010-11-1 Non-Haze 2011-4-14 Non-Haze

0.05 0.04

y = 2x

0.03 0.02 0.01 0

This research was financially supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB05010100), the Chinese National Natural Science Foundation (41075094 and 41175104), and the National Basic Research and the Development Program 973 (No. 2010CB732304). References

0.06

[NH4+](µmolm-3)

7

0

0.01

0.02

0.03

[SO42-](µmolm-3) Fig. 7. The mole ratio of [NH4+]/[SO 42 −] in airborne particles (aerodynamics diameter ≤ 1.8 μm) during haze days and non-haze days.

the investigating days (Table 1). High wind speed not only favors transportation of SO2 and sulfate into Beijing, but also promotes the dilution of NOx local emitted from vehicles, resulting in relatively low ratio of NO3−/SO42− under the windy day. It should be mentioned that the concentrations of NO3− as well as NH4+ measured by this study only represented the lowest limits due to loss of NH4NO3 via heterogeneous process (Solomon et al., 1992) during sampling, and thus the actual NO3−/SO42− ratios in the airborne particles must be greater than measured.

4. Conclusion The mass concentrations of airborne particles observed during haze days greatly exceeded the standard of air quality, indicating that the pollution of airborne particles is badly threatening the health of citizens with large population in the megacity. Because the hygroscopic character of water soluble inorganic ions in the airborne particles can strongly affect particle size and reflection of sunlight etc., the great enhancement of water soluble inorganic ions observed during the haze days implied that they made great contribution to the reduction of visibility. The size distribution of NH4+, NO3− and SO42− peaked at 0.32–0.56 μm during non-haze days was within the “condensation mode”, and at 0.56– 1.0 μm during haze days was a typical “droplet mode”. The great enhancement of K+ in fine airborne particles during haze days in October indicated that the contribution of biomass burning to the reduction of visibility in Beijing couldn't be neglected due to its high water solubility. In comparison with other ions, the greater enhancements of SO42−, NO3−, and NH4+ during haze days were probably ascribed to fast conversion of their precursors via multiphase reactions under the relatively high humidity. The high mole ratios of [NH4+]/[SO42−] in the fine particles observed during haze days indicated that the main form of ammonium in PM1.8 was (NH4)2SO4 and NH4NO3. The size distribution variations of Mg2+ and Ca 2+ were the same during both haze days and nonhaze days, and mainly focused on coarse particle mode, confirming that they mainly came from the crustal sources.

Adachi K, Buseck PR. Internally mixed soot, sulfates, and organic matter in aerosol particles from Mexico City. Atmos Chem Phys 2008;8(21):6469–81. Arimoto R, Duce RA, Savoie DL, Prospero JM, Talbot R, Cullen JD, et al. Relationships among aerosol constituents from Asia and the North Pacific during PEM-West. J Geophys Res 1996;101:2011–23. Brown SG, Herckes P, Ashbaugh L, Hannigan MP, Kreidenweis SM, Collett Jr JL. Characterization of organic aerosol in Big Bend National Park, Texas. Atmos Environ 2002;36(38):5807–18. Calvert JG, Su F, Bottenheim JW. Mechanism of the homogeneous oxidation of sulfur dioxide in the troposphere. Atmos Environ 1978;12:199–210. Chand D, Guyon P, Artaxo P, Schmid O, Frank GP, Rizzo LV, et al. Optical and physical properties of aerosols in the boundary layer and free troposphere over the Amazon basin during the biomass burning season. Atmos Chem Phys 2006;6:2911–25. Che HZ, Zhang XY, Li Y, Zhou ZJ, Qu JJ, Hao XJ. Haze trends over the capital cities of 31 provinces in China, 1981–2005. Theor Appl Climatol 2009;97:235–42. Chen LWA, Chow JC, Doddridge BG, Dickerson RR, Ryan WF, Mueller PK. Analysis of summertime PM2.5 and haze episode in the Mid-Atlantic region. J Air Waste Manag Assoc 2003;53(8):946–56. Du HH, Kong LD, Cheng TT, Chen JM, Du JF, Li L, et al. Insights into summertime haze pollution events over Shanghai based on online water-soluble ionic composition of aerosols. Atmos Environ 2011;45:5131–7. Engling G, Lee JJ, Tsai YW, Lung SC, Chou CK, Chan CY. Size-resolved anhydrosugar composition in smoke aerosol from controlled field burning of rice straw. Aerosol Sci Technol 2009;43(7):662–72. Fu QY, Zhuang GS, Wang J, Xu C, Huang K, Li J, et al. Mechanism of formation of the heaviest pollution episode ever recorded in the Yangtze River Delta, China. Atmos Environ 2008;42:2023–36. He KB, Yang FM, Ma YL, Zhang Q, Yao XH, Chan CK, et al. The characteristics of PM2.5 in Beijing, China. Atmos Environ 2001;35:4959–70. Hu M, He L, Zhang Y, Wang M, Kim YP, Moon KC. Seasonal variation of ionic species in fine particles at Qingdao, China. Atmos Environ 2002;36:5853–9. Hu M, Zhao YL, He LY, Huang XF, Tang XY, Yao XH, et al. Mass size distribution of Beijing particulate matters and its inorganic water-soluble ions in winter and summer. J Environ Sci 2005;4(26):1–6. [in Chinese]. IPCC (Intergovernmental Panel on Climate Change). Climate change 2001. New York: Cambridge University Press; 2001. Jacobson MZ. Strong radiative heating due to the mixing state of black carbon in atmospheric aerosols. Nature 2001;409:695–7. Kang CM, Lee HS, Kang BW, Lee SK, Sunwoo Y. Chemical characteristics of acidic gas pollutants and PM2.5 species during hazy episodes in Seoul, South Korea. Atmos Environ 2004;38:4749–60. Kerminen VM, Wexler AS. Growth laws for atmospheric aerosol particles: an examination of the bimodality of the accumulation mode. Atmos Environ 1995;29(22):3263–75. Khoder MI. Atmospheric conversion of sulfur dioxide to particulate sulfate and nitrogen nitric acid in an urban area. Chemosphere 2002;49:675–84. Kulmala M, Laakso L, Lehtinen KEJ, Riipinen I, Dal Maso M, Anttila T, et al. Initial steps of aerosol growth. Atmos Chem Phys 2004;4:2553–60. Li WJ, Shao LY. Mixing and water-soluble characteristics of particulate organic compounds in individual urban aerosol particles. J Geophys Res 2010;115:D02301. http://dx.doi.org/10.1029/2009JD012575. Li J, Pósfai M, Hobbs PV, Buseck PR. Individual aerosol particles from biomass burning in southern Africa: 2. compositions and aging of inorganic particles. J Geophys Res 2003;108(D13):8484. Li WJ, Shao LY, Buseck PR. Haze types in Beijing and the influence of agricultural biomass burning. Atmos Chem Phys 2010;10:8119–30. Li LJ, Wang Y, Li JX, Xin LZ, Jin J. The analysis of heavy air pollution in Beijing during 2000 2010. J Environ Sci (China) 2012a;2(1):23–30. [in Chinese]. Li WJ, Shi ZB, Zhang DZ, Zhang XY, Li PR, Feng QJ, et al. Haze particles over a coal-burning region in the China Loess Plateau in winter: three flight missions in December 2010. J Geophys Res 2012b;117:D12306. http://dx.doi.org/10.1029/2012JD017720. Liu XD, Espen PV, Adams F, Cafmeyer J, Maenhaut W. Biomass burning in southern Africa: individual particle characterization of atmospheric aerosols and savanna fire samples. J Atmos Chem 2000;36:135–55. Mao HY, Tian G, Huang YH, Li G, Song GW. Mass size distributions and existing forms of sulfate and nitrate at atmospheric environment in Beijing. J Environ Sci 2011;5(23): 1237–41. Meng Z, Seinfeld JH. On the source of the submicrometer droplet mode of urban and regional aerosols. Aerosol Sci Technol 1994;20(3):253–65.

8

Z. Sun et al. / Science of the Total Environment 456–457 (2013) 1–8

NRC. Research priorities for airborne particulate matter. Washington DC: National Academy of Sciences/National Research Council; 1998. Okada K, Ikegami M, Zaizen Y, Makino Y, Jensen JB, Gras JL. The mixture state of individual aerosol particles in the 1997 Indonesian haze episode. J Aerosol Sci 2001;32(11):1269–79. P'osfai M, Simonics R, Li J, Hobbs PV, Buseck PR. Individual aerosol particles from biomass burning in southern Africa: 1. Compositions and size distributions of carbonaceous particles. J Geophys Res 2003;108(D13):8483. Poppe D, Wallasch M, Zimmermann J. The dependence of the concentration of OH on its precursors under moderately polluted conditions: a model study. J Atmos Chem 1993;16:61–78. Quan JN, Gao Y, Zhang Q, Tie XX, Cao JJ, Han SQ, et al. Evolution of planetary boundary layer under different weather conditions, and its impact on aerosol concentrations. Particuology 2013;11:34–40. Ravishankara AR. Heterogeneous and multiphase chemistry in the troposphere. Science 1997;276:1058–65. Reid JS, Koppmann R, Eck TF, Eleuterio DP. A review of biomass burning emissions part II: intensive physical properties of biomass burning particles. Atmos Chem Phys 2005;5:799–825. Schichtel BA, Husar RB, Falke SR, Wilson WE. Haze trends over the United States, 1980–1995. Atmos Environ 2001;35:5205–10. Senaratne I, Shooter D. Elemental composition in source identification of brown haze in Auckland, New Zealand. Atmos Environ 2004;38:3049–59. Solomon PA, Salmon LG, Fall T, Cass GR. Spatial and temporal distribution of atmospheric nitric acid and particulate nitrate concentrations in the Los Angeles area. Environ Sci Technol 1992;26:1594–601. Stelson AW, Seinfeld JH. Relative humidity and temperature dependence of the ammonium nitrate dissociation constant. Atmos Environ 1982;16:983–92. Sun Y, Wang YS, Liu GR ANJL, Ma ZQ, Shi LQ, Xu HH. Analysis for vertical profile of atmospheric SO2 during air seriously polluted days in Beijing. J Environ Sci 2006a;27(3):408–14. [in Chinese]. Sun YL, Zhuang GS, Tang AH, Wang Y, An ZS. Chemical characteristics of PM2.5 and PM10 in haze–fog episodes in Beijing. Environ Sci Technol 2006b;40:3148–55. Tan JH, Duan JC, He KB, Ma YL, Duan FK, Chen Y, et al. Chemical characteristics of PM2.5 during a typical haze episode in Guangzhou. J Environ Sci 2009;21(6):774–81. US EPA (United States Environmental Protection Agency). National Ambient Air Quality Standards (NAAQS), 2006.

Wang GH, Huang LM, Gao SX, Gao ST, Wang LC. Characterization of water-soluble species of PM10 and PM2.5 aerosols in urban area in Nanjing, China. Atmos Environ 2002;36(8):1299–307. Wang Y, Zhuang GS, Tang AH, Yuan H, Sun YL, Chen S, et al. The ion chemistry and the source of PM 2.5 aerosol in Beijing. Atmos Environ 2005;39:3771–84. Wang Y, Zhuang GS, Sun YL, An ZS. The variation of characteristics and formation mechanisms of aerosols in dust, haze, and clear days in Beijing. Atmos Environ 2006;40:6579–91. Watson JG. Visibility: science and regulation. J Air Waste Manag Assoc 2002;52(6): 628–713. Xiu GL, Zhang DN, Chen JZ, Huang XJ, Chen ZX, Guo HL, et al. Characterization of major water-soluble inorganic ions in size-fractionated particulate matters in Shanghai campus ambient air. Atmos Environ 2004;38:227–36. Yao XH, Chan CK, Fang M, Cadle S, Chan T, Mulawa P, et al. The water-soluble ionic composition of PM2.5 in Shanghai and Beijing, China. Atmos Environ 2002;36: 4223–34. Yao XH, Lau AP, Fang M, Chan CK, Hu M. Size distributions and formation of ionic species in atmospheric particulate pollutants in Beijing, China: 1 — inorganic ions. Atmos Environ 2003;37:2991–3000. Yin LQ, Niu ZC, Chen XQ, Chen JS, Xu LL, Zhang FW. Chemical compositions of PM2.5 aerosol during haze periods in the mountainous city of Yong'an, China. J Environ Sci 2012;24(7):1225–33. Yu CY, Dong C, Wang XF, Yang LX, Wang WX. Size distributions of water-soluble inorganic ions of atmospheric aerosol particles in autumn in Jinan. J Environ Sci (China) 2011;31(4):561–7. [in Chinese]. Zhao PS, Xu XF, Meng W, Dong F, He D, Shi QF, et al. Characteristics of hazy days in the region of Beijing, Tianjin, and Hebei. J Environ Sci (China) 2012;2(1):31–6. [in Chinese]. Zheng XY, Liu XD, Zhao FH, Duan FK, Yu T, Cachier H. Seasonal characteristics of biomass burning contribution to Beijing aerosol. Sci China Chem 2005;5:481–8. Zhou FM, Sun QR, Wang MR, Shao KS. Measurement of aerosol strong acidity in Zhouguancun, Beijing. Environ Sci 1998;19:6-11. [in Chinese]. Zhuang H, Chan CK, Fang M, Wexler AS. Size distributions of particulate sulfate, nitrate, and ammonium at a coastal site in Hong Kong. Atmos Environ 1999;33:843–53.