Vertical distribution of polycyclic aromatic hydrocarbons in atmospheric boundary layer of Beijing in winter

Vertical distribution of polycyclic aromatic hydrocarbons in atmospheric boundary layer of Beijing in winter

ARTICLE IN PRESS Atmospheric Environment 41 (2007) 9594–9602 www.elsevier.com/locate/atmosenv Vertical distribution of polycyclic aromatic hydrocarb...

245KB Sizes 0 Downloads 13 Views

ARTICLE IN PRESS

Atmospheric Environment 41 (2007) 9594–9602 www.elsevier.com/locate/atmosenv

Vertical distribution of polycyclic aromatic hydrocarbons in atmospheric boundary layer of Beijing in winter Shu Taoa,, Yi Wanga, Shiming Wua, Shuzheng Liua, Han Doua, Yanan Liua, Chang Langa, Fei Hub, Baoshan Xingc a

Laboratory for Earth Surface Processes, College of Environmental Sciences, Peking University, Beijing 100871, P.R. China b Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, P.R. China c Department of Plant, Soil and Insect Sciences, University of Massachusetts, Stockbridge Hall, Amherst, MA 01003-7245, USA Received 6 May 2007; received in revised form 17 August 2007; accepted 17 August 2007

Abstract Air samples were collected using active samplers at various heights of 8, 15, 32, 47, 65, 80, 102, 120, 140, 160, 180, 200, 240, 280 and 320 m on a meteorological tower in an urban area of Beijing in two campaigns in winter 2006. Altitudinal distributions of polycyclic aromatic hydrocarbons (PAHs) in atmospheric boundary layer of Beijing in winter season were investigated. Meteorological conditions during the studied period were characterized by online measurements of four meteorological parameters as well as trajectory calculation. The mean total concentrations of 15 PAHs except naphthalene of gaseous and particulate phase were 6677450 and 3317144 ng m3 in January and 61719 and 2976 ng m3 in March, respectively. Domestic coal combustion and vehicle emission were the dominant PAH sources in winter. Although the composition profiles derived from the two campaigns were similar, the concentrations were different by one order of magnitude. The higher concentrations in January were partly caused by higher emission due to colder weather than March. Moreover, weak wind, passing through the city center before the sampling site, picked up more contaminants on the way and provided unfavorable dispersion condition in January. For both campaigns, PAH concentrations decreased with heights because of ground-level emission and unfavorable dispersion conditions in winter. The concentration ratio of PAHs in gas versus solid phases was temperature dependent and negatively correlated to their octanol–air partition coefficients. r 2007 Elsevier Ltd. All rights reserved. Keywords: Atmosphere; PAHs; Vertical distribution; Trajectory; Beijing

1. Introduction Air pollution in urban environment has become a major concern to policy makers and general public. This is particularly true for Beijing, which together with several other large cities in China, ranked top Corresponding author. Tel./fax: +86 10 62751938.

E-mail address: [email protected] (S. Tao). 1352-2310/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2007.08.026

in the most polluted cities in the world (Watts, 2005). Of various atmospheric contaminants, a number of polycyclic aromatic hydrocarbons (PAHs) were demonstrated to have carcinogenic effect (Xue and Warshawsky, 2005). The concentrations of PAHs in air samples collected in Beijing were much higher than those in many other cities (Liu et al., 2003; Zhou et al., 2005). Therefore, exposures to high level of various atmospheric

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

contaminants, including PAHs, are blameful for health risk in Beijing and other cities in northern China (Tao et al., 2006). Information on altitudinal distributions of PAHs in atmosphere is scarce in spite of its importance for understanding and modeling environmental fate and behavior of the chemicals. In fact, uniform distribution of PAHs in atmospheric compartment was often assumed practically in compartmental modeling. Such a simplifying assumption often leads to high uncertainties in the estimated advective air fluxes and the predicted air concentrations because convection airflow and air-to-surface exchange are among the most influential processes in the model (Wang et al., 2002, 2004). For most atmospheric dispersion models, on the other hand, the predicted results could only be validated two dimensionally without the information on vertical distribution (Tao et al., 2006). Efforts have been made to fill this information gap recently. Qi et al. (2000) measured the concentrations of PM2.5 and PAHs at three heights in Nanhai, southern China, and found that there was a clear decline trend with height for both PM2.5 and PAHs and vehicle emission at ground level was the primary pollutant source. A similar conclusion was reached from an investigation conducted in Guangzhou (Li et al., 2005). In a study carried out in Tianjin, a northern Chinese city, air samples were collected at three heights of 20, 40 and 60 m on a meteorological tower and the highest concentrations were observed at 40 m (Wu et al., 2006). Recently, Farrar et al. (2005) deployed thin film passive air samplers at six heights up to 360 m on the CN Tower in Toronto, Canada, to investigate the vertical distribution of several persistent organic pollutants including PAHs in the urban atmospheric boundary layer. They observed a decline trend with height for dominant PAHs in most times. Although information on PAH vertical distribution in atmosphere is limited, many studies were conducted on altitudinal distributions of aerosol which are certainly relevant to PAHs because a large fraction of PAHs occur in particulate phase. More often than not, sharp vertical variation of aerosol concentration was observed while the pattern of the distribution largely depended on emission source categories and meteorological conditions (McKendry et al., 2004; Kalberer et al., 2004; Farrar et al., 2005; Wu et al., 2006). The vertical distribution of PAHs partly depends on the position and strength of their sources. The

9595

major emission sources of PAHs in Chinese urban environments including domestic coal combustion and motor vehicle exhaust which are all at or close to ground level (Qi et al., 2000; Xu et al., 2006). In addition, the vertical profile is also strongly affected by the extent of the atmospheric boundary layer and meteorological conditions. Pollutants emitted at ground level may transfer vertically through advection flow and/or mixing, resulting in different profile patterns (Chan et al., 2000; Farrar et al., 2005). The purpose of this study was to determine the vertical distribution of PAHs in the atmospheric boundary layer of urban area of Beijing in early and late winter by using an active sampling technique. The vertical distribution pattern was interpreted based on both emission and the meteorological conditions with the assistance of trajectory calculation. The results of such observations would be useful not only for a better understanding of source and fate of, as well as human exposure risk to, PAHs in the region, but also for improvement of algorithm and parameterization of multimedia fate modeling and dispersion modeling. 2. Methodology 2.1. Sampling Samples were collected on a meteorological tower in urban area of Beijing (391580 2700 N, 1161220 1900 E). The tower was constructed in northern suburb of Beijing decades ago but totally surrounded by buildings now because of the sprawling of the city. Still, most buildings in the neighborhood of several kilometers are no more than several floors tall and there are only a few buildings within the distance of 500 m from the tower. In winter season, in addition to vehicle exhaust from heavy traffic in the area, coal combustion for indoor heating is also a major emission source of PAHs in Beijing (Zhang et al., 2007). Under the influence of continental monsoon climate and local topography, shallow and stable boundary layer with weak vertical mixing and strong ground-based inversion often occur in winter (Xia, 2006). Integrated 24-h air samples were collected in two campaigns in 15–16 January 2006 and 7–8 March 2006 using TMP-1500 active air samplers (Jiangsu Eltong Electric, Jingtan, China) at a flow rate of 0.8–0.9 m3 min1. The flow rates were double calibrated immediately before and after the sampling. Gaseous and particulate phase PAHs were collected using polyurethane foam plugs (PUF,

ARTICLE IN PRESS 9596

S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

22 mm O.D., 7.6 cm) in glass holders and glass fiber filters (GFF, 22 mm O.D.) in metalline screens (Supelco, PA, USA), respectively. No breakthrough was found for all PAHs studied in a preliminary experiment using three holders connected in series. The PUF plugs were previously Soxhlet-extracted using a mixture of n-hexane and cyclohexane (1:1) for 4 h and the GFFs were heated in a furnace at 500 1C for 24 h. Fifteen samplers were deployed simultaneously at 15 platforms of various heights (8, 15, 32, 47, 65, 80, 102, 120, 140, 160, 180, 200, 240, 280 and 320 m) on the tower. Not all data for the 15 heights were available due to the failure of a few samplers over the course of the sampling. After harvesting, all PUF plugs were stored at 18 1C. All GFFs were equilibrated in a desiccator (25 1C) for 24 h and weighed both before and after the sampling. All samples were analyzed within 48 h. 2.2. Extraction and analysis PUF plugs and GFFs were extracted by Soxhlet with a mixture of n-hexane and cyclohexane (1:1) for 4 and 10 h, respectively. The extracts were concentrated by rotary evaporation to about 0.5 ml and 2-floruobiphenyl and p-terphenyl-d14 were then added as internal standard. For both PUF and GFF, the amount of co-extracts was relatively small and sample cleanup was not necessary. The samples were stored in dark at 18 1C prior to analysis on a gas chromatograph (Agilent 6890) coupled with a mass selective detector (MSD, Agilent 5973) and an automatic sampler (HP-7673A, Agilent). A 30 m  0.25 mm I.D.  0.25 mm film thickness HP5MS capillary column was used. Helium was employed as the carrier gas at a flow rate of 1 ml min1. The column was programmed from 60 to 300 1C at 5 1C min1, and then held isothermal for 15 min. The injection (1.0 ml) was operated at a splitless mode with the head pressure of 0.003 MPa and injector temperature of 250 1C. The MSD was operated in selected ion monitoring mode at 70 eV and the ion source temperature was set at 200 1C. Sixteen PAHs were quantified and they were naphthalene (NAP), acenaphthylene (ACY), acenaphthene (ACE), fluorene (FLO), phenanthrene (PHE), anthracene (ANT), fluoranthene (FLA), pyrene (PYR), benz(a)anthracene (BaA), chrysene (CHR), benzo(b)fluoranthene (BbF), benzo(k)fluoranthene (BkF), benzo(a)pyrene (BaP), dibenz(a,h) anthracene (DahA), indeno(l,2,3-cd)pyrene (IcdP) and benzo(g,h,i)perylene (BghiP).

2.3. Analytical quality control Routine quality assessment procedures were followed. Due to limited availability of sampling equipment, only a single sampler was deployed at each of the 15 heights each time. The reproducibility of the procedure was checked in a previous study using the same equipment and procedure and it was demonstrated that average coefficients of variation of the duplicate sampling and analysis were between 22% and 34% for various PAH compounds in gaseous and particulate phases (Tao et al., 2007). Detection limits ranged between 0.85 (NAP) and 6.8 (BghiP) ng ml1 for the extracts. Two procedure blanks were included for every batch of 16 samples and mean of the two were used for procedure blank correction. All the results were laboratory procedure blank corrected using mean blank values (two for every 16 samples) obtained from the extraction to the analysis. The results were also corrected using method recoveries which were determined by spiking the sampling media with a working standard and extracting the media in the same way as the samples. The standard mixture of PAHs (PPH10JM, Chem Service Inc., FL, USA) was diluted with n-hexane. For the 16 spiked PAHs, the recoveries were from 66% to 114% for the PUF plug and from 60% to 115% for the GFF. Several samples were also spiked with a range of deuterated PAHs (NAP-d8, ACE-d10, ANT-d10, CHR-d12 and perelyne-d12, 4 mg l1; J&K Chemical (Beijing), Beijing, China) to monitor the extraction and cleanup procedures. The recoveries of the surrogates ranged from 59% to 78%. Quantification was performed by the internal standard method using 2-fluoro-1,10 -biphenyl and p-terphenyl-d14 (2.0 mg ml1; J&K Chemical (Beijing), Beijing, China). All solvents used were analytical grade (Beijing Chemical Reagent, Beijing, China) and purified by distillation prior to use. All glassware was cleaned using an ultrasonic cleaner (Kunshan KQ-500B) and heated to 400 1C for 6 h. 2.4. Meteorological conditions Meteorological parameters including temperature, humidity, wind direction and wind speed were retrieved every 15 min online at all heights during the sampling period. Air parcel trajectories were calculated using the hybrid single-particle Lagrangian integrated trajectory (HYSPLIT) model and

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

global NOAA-NCEP/NCAR pressure level reanalysis data archive. Both forward (12 h) and backward (72 h) trajectories were calculated at all sampling heights with location of the sampling tower as the starting point every 6 h during the sampling period (Draxler and Rolph, 2003). 3. Results and discussion 3.1. PAH concentrations in ambient air The measured concentrations of several representative PAHs and the total concentrations of 15 PAHs (PAH15) are summarized in Fig. 1 in log scale. NAP was not included in PAH15 because of its relatively low recovery. The statistics were height-averaged for the two campaigns and two phases, respectively. The detailed data for all PAHs are provided in the Supplementary material. With a few exceptions, most data are log-normally distributed, suggested by the symmetric Box–Whisker plots. The mean PAH15 were 3547439 and 1547178 ng m3 for gaseous and particulate phases, respectively. Zhou et al. (2005) measured PAHs in five-stage size segregated aerosol in Beijing in 2003 and reported an annual mean concentration of 116 ng m3 with significant higher winter concentration than the other seasons. During a period 10–26 January 2005, a survey on police exposure was conducted and the PAH15 measured on two sites on Peking University campus ranged from 130 to 1100 ng m3 and from 21 to 1050 ng m3 for gaseous and particulate phase PAHs, respectively (Liu et al., 2007). It appears that the concentrations derived in this study were broadly in line with other measurements in urban areas of northern China except for a few extremely high concentrations reported at

specific locations close to strong sources, e.g. coking factory (Liu et al., 2003). The measured PAH concentrations in air in southern Chinese cities were usually lower than those in north where space heating is a common practice in winter. For example, it was reported that the annual average concentration of PAH15 in PM10 sample collected in urban area of Nanjing was 86.0 ng m3 (Wang et al., 2006). In Guangzhou, the observed average total concentrations of 16 PAHs (PAH16) were 313 and 23.7 ng m3 for gaseous and particulate phases, respectively (Li et al., 2006). Still, all these measurements are generally higher than those reported in other countries in the world (Goldstone et al., 1992; Fraser et al., 1998; Park et al., 2002). 3.2. Profile composition and sources of PAHs It was found that the lighter PAHs, particularly NAP, predominated in the gaseous phase, while the particulate phase PAHs were primarily median to higher molecular weight compounds including PHE, FLA, PYR and CHR. Such a pattern was commonly reported in literature (Harner and Bidleman, 1998; Kaupp and McLachlan, 1999; Liu et al., 2007). The difference between the two campaigns was not as profound as that between the two phases. Such similarities and dissimilarities were further addressed by a principle component analysis, in which, the profiles of major emission sources in this area were also included (Zhang et al., 2007). Table 1 presents the factor loadings of the principal component analysis and those with loading values 40.70 are underlined and marked bold. The factor loadings of the first two principal components (F1 and F2) accounted for 55% and 35% of the total

4 Conc., log(ng/m3)

9597

Jan p99

2

Jan

p95

Jan

p75

Mar 0

p50 mean

Mar

p25 p5

-2

p1

part gas PHE

part gas PYR

part gas CHR

part gas PAH15

Fig. 1. Means and percentiles (pi) of NAP, PHE, PYR, CHR and PAH15 in gaseous and particulate phases measured in the two campaigns in January 2006 and March 2006. The concentrations were log-scaled. Gaseous CHR in March was below the detection limit.

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

9598

Table 1 Factor loadings derived from the principal component analysis PAHs

F1 F2

NAP

ACE

ACY

FLO

PHE

FLA

PYR

BaA

CHR

BbF

BkF

BaP

DahA

BghiP

Variance (%)

0.37 0.89

0.15 0.96

0.06 0.95

0.12 0.98

0.19 0.97

0.93 0.07

0.89 0.02

0.95 0.16

0.93 0.22

0.95 0.03

0.91 0.05

0.94 0.09

0.72 0.53

0.95 0.09

0.55 0.35

2.5 Coal Vehicle Jan-p

F2

March-g

0.0 Jan-g

-2.5 -1.5

March-p

0.5 F1

coke production (Xu et al., 2006). In the case of urban area of Beijing, similar to many other major metropolitan areas, biofuel burning and coking industry are not major sources while motor vehicles contribute a significant portion of the total emission (Xue and Warshawsky, 2005; Zhou et al., 2005; Wang et al., 2006). As shown in Fig. 2, both emission profiles of domestic coal combustion and vehicle emission are similar to the particulate phase PAHs observed in this study.

2.5

Fig. 2. Factor score plot of F1 against F2 based on PAH profile compositions of the measured PAHs as well as the major sources including domestic coal combustion and transportation petroleum consumption. Gaseous and particulate phases are presented as g and p, respectively.

variance (Var.), respectively. F1 was mainly an association of median to higher molecular weight PAHs from FLA to BghiP, while F2 represents principally the linear combination of lower molecular weight PAHs from NAP to PHE. It appears that the composition patterns of gaseous and particulate phases are well distinguished. The factor score of F1 is plotted against that of F2 in Fig. 2 to illustrate the similarity or dissimilarity of individual samples. The gaseous and particular phase PAHs are well distinguished in the plot. The gaseous phase PAH profiles from the two sampling periods have relatively lower F1 values than those of the particulate phase ones, corresponding to relatively abundant heavier species in the latter. A few exceptions among the gaseous phase samples with higher F1 values happened to be those collected at lower heights with much higher total concentrations than the others. It appears that the samples from the two campaigns can be distinguished by neither F1 nor F2. It was reported that the major emission sources of PAHs in China are biofuel burning, domestic coal combustion and

3.3. Difference between the two campaigns Although the composition profiles of the samples collected in two campaigns were similar to each other, significant difference in concentration was observed. The mean concentrations (all heights) of gaseous and particulate phase PAH15 were 6677450 and 3317144 ng m3 in January and 61719 and 2976 ng m3 in March, respectively. Although both sampling periods were within the heating season which lasted from 15 November to 15 March each winter, the difference in PAH15 was around one order of magnitude. It was likely that both emission and meteorological condition were the primary reason causing such a difference. In winter, coal burning for space heating is a dominant emission source of PAHs in China (Xu et al., 2006). Although the space heating was on at both occasions, the different intensities of the activity were reasonably expected while the daily mean temperatures of the two campaigns were 0.4 and 7.3 1C, respectively. Li (1999) investigated the dependence of coal consumption on the ambient temperature in Beijing and derived a quantitative equation. By using Li’s equation for heating coal consumption, together with the non-heating coal consumption provided in literature (Beijing Statistics Bureau, 2006), it was estimated that the coal consumption during the two campaigns were 52,000 and 39,600 t d1, equivalent to PAH15 emission of 17.4 and 12.6 t d1, respectively.

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

To illustrate the possible source of PAHs, a number of trajectories (72 h) for several sampling heights were presented in Fig. 3. According to a county-resolution emission inventory, the emission densities in the regions along the pathways of these trajectories were all significantly lower than that of Beijing urban area and the detected PAHs in this study were primarily from local source (Zhang et al., 2007). In January, not only the movement of air parcels was slower than that in March, they reached the sampling site from southeast direction passing through the city center. On the other hand, the air parcels came from northeast in March with faster speed and less chance of picking up the pollutants on the way and diluting the local pollution. The calculated differences were wellsupported by the on-site measured wind speeds (1.4 m s1 versus 0.9 m s1) and directions. The difference in dispersion status between the two campaigns occurred not only horizontally, but

9599

also vertically. Fig. 4 compares the 5-h forward and 5-h backward trajectories with initial times of 18:00 h. In 7–8 March, strong lifting air from northeast passed the sampling site created a condition relatively favorable for dispersion of pollutants at ground level. In contrast, the air was more static on 15 January, the wind from southeast tended to circle around which was favorable for the accumulation of pollutants at ground level. It is likely that the difference in dispersion conditions was a more important factor controlling PAH concentrations than the difference in emission. 3.4. Vertical distribution of PAHs Among the few studies on vertical distribution of PAHs in atmospheric boundary layer, different concentration profiles were observed. Farrar et al. (2005) found a clear decline of dominant PAHs with height in most cases in the urban atmospheric

56 Russia

Russia

N, °

50 Mongolia

Mongolia

44 China Jan. 15

38

China March 7

Beijing

80

100 E, °

120

80

Beijing

100 E, °

120

Fig. 3. Seventy-two hour back-trajectories at various elevations of 8, 47, 102, 160 and 240 m with initial times of 6:00 h of 15 January (left) and 8 March (right).

480

480

March 7

January 14 Sampling tower

360 Height, m

240

240

120

116.5 E, °

116.0

115.5

40.8 40.3 39.8 ° 39.3 N,

0 39.3

39.8

117.0 116.5 116.0 40.8 115.5 °

120

0 117.0

Sampling tower

N, °

40.3

E,

Height, m

360

Fig. 4. Movement of air mass passing the sampling site at various heights in 14–15 January (left) and 7–8 March (right) derived from forward and backward trajectory calculations.

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

9600 330

gaseous Jan.

gaseous March

particulate Jan.

330

particulate March

Jan.

220 H, m

H, m

220

110

110

March

0

0 0

900 P15, ng/m3

1800 0

100 0

50 P15, ng/m

3

250

500 0

20

P15, ng/m3

40

0.0

P15, ng/m3

2.5

5.0

V, m/s

Fig. 5. Vertical distributions of gaseous and particulate phase PAH15 and daily mean wind velocities in 16–17 January 2006 and 7–8 March 2006.

0 DahA

logRg/p

boundary layer of Toronto, Canada. In Tianjin, the city immediately next to Beijing, Wu et al. (2006) found the highest concentrations of PAHs at a height of 40 m compared to the other two sampling heights of 20 and 60 m, presumably due to the heights of chimneys in the surrounding neighborhood. As illustrated in Fig. 5, clear decreasing trends with height are demonstrated for both campaigns in this study. This is similar to the observed vertical distributions of aerosol in atmosphere of Beijing and other cities (Li et al., 2005; Yang et al., 2005). Farrar et al. (2005) proposed three major scenarios of vertical concentration profiles of atmospheric pollutants due to differences in emission sources, advection and vertical mixing conditions. (1) Even vertical distribution with weak fresh emission, dominant advection and well-mixed condition; (2) decrease with height with ground source domination and stable atmospheric boundary layer conditions and (3) increase with height with upper boundary layer emission sources. The results of this study fit the second scenario well. As discussed previously, the major PAH emission sources in Beijing were motor vehicle emission and domestic coal combustion, which were practically at ground level. Yang et al. (2005) also confirmed that this specific site was under strong influences of emission from transportation and domestic combustion of coal in adjacent neighborhood. Though the weathers were different between the two campaigns, the meteorological conditions prevailing in this region in winter facilitated the accumulation of most atmospheric pollutants at lower boundary layer, because the study area generally features in shallow boundary layer, weak vertical mixing and frequent temperature inversion in winter. Hu et al. (2004) indicated that the average extents of the boundary layer are 510.0 and 1360.0 m in winter and summer, respectively. Under the influence of still current,

-4

r = 0.905

-8 4

10 logKoa

16

Fig. 6. Relationship between gaseous/particulate ratio (Rg/p) and octanol–air partitioning coefficient (KOA), both of which are logtransformed. Standard deviations of Rg/p (vertical bars) were derived from the measurements at various heights and standard deviations of KOA (horizontal bars) were due to changes in temperature at various heights.

local topography, mountain-and-valley wind and cloudy weather, temperature inversion occurred for 230 d each year and for 22 d in January in Beijing. Occasionally, the inversion can even last continuously for several days in winter (Xia, 2006). During the sampling periods, the ground wind speeds were only 0.9 and 1.4 m s1 and ground-based temperature inversion was observed in both cases, particularly in 15–16 January. It was unfavorable for pollutants emitted at ground level to travel upwards, leading to sharp vertical decrease in PAH concentrations both in gaseous and particulate phases. This is especially true in January. 3.5. Partitioning of PAHs between gaseous and particulate phases As a class of semi-volatile chemicals, the gasparticle partitioning of various PAHs are very different. Lighter PAHs with lower vapor pressure are predominantly found in gaseous phase while heavier ones with five or more rings are almost

ARTICLE IN PRESS S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

-2.5

-2.0

9601

330

March

March

r = 0.639

-2.7

Jan.

H, m

logRg/p

logRg/p

220 -2.9

110 Jan. -3.4 3.64 3.65

3.66

3.67

r = 0.578 -3.3 3.55 3.57

1 /T, 10-3 k

0 3.59

1 /T, 10-3 k

-1

4

9

T, °C

Fig. 7. Relationship between log Rg/p and ambient temperature during the two campaigns (left and middle) and vertical variations of temperatures in January and March.

exclusively in the particulate phase (Harner and Bidleman, 1998; Kaupp and McLachlan, 1999). The octanol–air partition coefficient (KOA) was defined as the equilibrium concentration ratio of chemicals in air versus in octanol and often used to describe the partitioning of semi-volatile chemicals between gaseous and particulate phases in atmosphere (Finizio et al., 1997). In this study, the weights of particles were measured for all samples and the PAHs in particulate phase can also be presented as weight concentration (ng g1). The concentration ratios of PAHs in gas (ng m3) versus solid (ng g1) phases (Rg/p) were derived, which show a significant correlation (Fig. 6) with the literature reported KOA after log-transformation (Tolls and McLachlan, 1994). In Fig. 6, data from the two campaigns are pooled together with the vertical and horizontal bars representing standard deviations of the measured Rg/p at various heights and the influence of temperature on KOA, respectively. Relatively high variability of Rg/p for several heavier PAHs can be explained by the fact that the measured concentrations of these compounds in gaseous phase were close or sometimes below to the detection limits. This was also the reason that DahA deviates from the linear correlation. It should also be noted that the active sampling technique has its limitation. It was possible that a re-partitioning of PAHs between the solid and gas phases occurred during the sampling when all gaseous PAHs passed through the layer of collected particles on the filter prior to be trapped on PUF. The temperature dependence of the phase partitioning of PAHs is illustrated in Fig. 7, where log Rg/p is plotted against temperature (1/T) for the two campaigns. The vertical temperature profiles

are also presented for reference (right). Significant correlation (po0.01) can be seen with Pearson correlation coefficient of 0.639 and 0.578 for the two campaigns, respectively, demonstrating that cool ambient temperature, or high altitude was favored particulate phase partitioning. Acknowledgments Funding for this study was provided by National Basic Research Program (2007CB407301) and National Scientific Foundation of China (40332015, 40428005). We are also grateful to the Peking University President Undergraduate Student Research Program for financial support. Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at doi:10.1016/ j.atmosenv.2007.08.026. References Beijing Statistics Bureau, 2006. Statistics Year Book of Beijing. Chinese Statistics Press, Beijing. Chan, L.Y., Kwok, W.S., Chan, C.Y., 2000. Human exposure to respirable suspended particulate and airborne lead in different roadside microenvironments. Chemosphere 41, 93–99. Draxler, R.R., Rolph, G.D., 2003. HYSPLIT (HYbrid SingleParticle Lagrangian Integrated Trajectory) model access via NOAA ARL READY website /http://www.arl.noaa.gov/ ready/hysplit4.htmlS. NOAA Air Resources Laboratory, Silver Spring, MD. Fraser, M.P., Cass, G.R., Simoneit, B.R.T., Rasmussen, R.A., 1998. Air quality model evaluation data for organics: 5. C6-C22 nonpolar and semipolar aromatic compounds. Environmental Science and Technology 32, 1760–1770.

ARTICLE IN PRESS 9602

S. Tao et al. / Atmospheric Environment 41 (2007) 9594–9602

Farrar, N.J., Harner, T., Shoeib, M., Sweetman, A., Jones, K.C., 2005. Field deployment of thin film passive air samplers for persistent organic pollutants: a study in the urban atmospheric boundary layer. Environmental Science and Technology 39, 28–42. Finizio, A., Mackay, D., Bidleman, T.F., Harner, T., 1997. Octanol–air partition coefficient as a predictor of partitioning of semi-volatile organic chemicals to aerosols. Atmospheric Environment 15, 2289–2296. Goldstone, S.O., Kirk, M.E., Lester, P.W.W., Lester, J.N., Perry, R., 1992. Concentration of particulate and gaseous polycyclic aromatic hydrocarbons in London air following a reduction in the lead content of petrol in the United Kingdom. Science of the Total Environment 111, 169–199. Harner, T., Bidleman, T.F., 1998. Measurement of octanol–air partition coefficient for polycyclic aromatic hydrocarbons and polychlorinated naphthalenes. Journal of Chemical Engineering Data 43, 40–46. Hu, H.L., Wu, Y.H., Xie, C.B., Yan, F.Q., Wong, N.Q., Fan, A.Y., Xu, Y.D., Ji, Y.F., Yu, T., Ren, Z.H., 2004. Aerosol pollutant boundary layer measured by Lidar at Beijing. Research of Environmental Science 17, 59–73. Kalberer, M., Henne, S., Prevot, A.S.H., Steinbacher, M., 2004. Vertical transport and degradation of polycyclic aromatic hydrocarbons in an Alpine Valley. Atmospheric Environment 38, 6447–6456. Kaupp, H., McLachlan, M.S., 1999. Gas/particle partitioning of PCDD/Fs, PCBs, PCNs and PAHs. Chemosphere 38, 3411–3421. Li, Q.J., 1999. Temperature dependence of coal consumption loading in design of space heating. Urban Coal Gas 2, 22. Li, C.L., Fu, J.M., Sheng, G.Y., Bi, X.H., Hai, Y.M., Wang, X.M., Mai, B.X., 2005. Vertical distribution of PAHs in the indoor and outdoor PM2.5 in Guangzhou, China. Building and Environment 40, 327–339. Li, J., Zhang, G., Li, X.D., Qi, S.H., Liu, G.Q., Peng, X.Z., 2006. Source seasonality of polycyclic aromatic hydrocarbons (PAHs) in a subtropical city, Guangzhou, South China. Science of the Total Environment 355, 145–155. Liu, D.M., Li, Y.Y., Jiang, B.K., Wang, K., 2003. Preliminary study of organic pollutants from atmospheric particulates in Shougang District, Beijing. Journal of China University of Geosciences 28, 327–332. Liu, Y.N., Tao, S., Yang, Y.F., Dou, H., Yang, Y., Coveney, R.M., 2007. Inhalation exposure of traffic police officers to polycyclic aromatic hydrocarbons (PAHs) during the winter in Beijing, China. Science of the Total Environment 383, 98–105. McKendry, I.G., Sturman, A.P., Vergeiner, J., 2004. Vertical profiles of particulate matter size distributions during winter domestic burning in Christchurch, New Zealand. Atmospheric Environment 38, 4805–4813. Park, S.S., Kim, Y.J., Kang, C.H., 2002. Atmospheric polycyclic aromatic hydrocarbons in Seoul, Korea. Atmospheric Environment 36, 2917–2924.

Qi, S.H., Sheng, G.Y., Fu, J.M., Min, Y.S., Wu, J.X., Jian, Y.T., Ye, Z.X., 2000. Study on distributions of polycyclic aromatic hydrocarbons (PAHs) in aerosols at different levels. Acta Scientiae Circumstantiae 20, 308–311. Tao, S., Li, X.R., Yang, Y., Coveney, R.M., Lu, X.X., Chen, H.T., Shen, W.R., 2006. Dispersion modeling of polycyclic aromatic hydrocarbons from combustion of biomass and fossil fuels and production of coke in Tianjin, China. Environmental Science and Technology 40, 4586–4591. Tao, S., Liu, Y.N., Xu, W., Lang, C., Liu, S.Z., Dou, H., Liu, W.X., 2007. Calibration of a passive sampler for both gaseous and particulate phase polycyclic aromatic hydrocarbons. Environmental Science and Technology 41, 568–573. Tolls, J., McLachlan, M.S., 1994. Partitioning of semivolatile organic compounds between air and Lolium multiflorum (welsh ray grass). Environmental Science and Technology 28, 159–166. Wang, X.L., Tao, S., Xu, F.L., Dawson, R.W., Cao, J., Li, B.G., Fang, J.Y., 2002. Modeling the fate of benzo[a]pyrene in the wastewater-irrigated areas of Tianjin with a fugacity model. Journal of Environment Quality 31, 896–903. Wang, X.L., Tao, S., Dawson, R.W., Wang, X.J., 2004. Uncertainty analysis of parameters for modeling the transfer and fate of benzo(a)pyrene in Tianjin wastewater irrigated areas. Chemosphere 55, 525–531. Wang, G., Huang, L., Zhao, X., Niu, H., Dai, Z., 2006. Aliphatic and polycyclic aromatic hydrocarbons of atmospheric aerosols in five locations of Nanjing urban area, China. Atmospheric Research 81, 54–66. Watts, J., 2005. China: the air pollution capital of the world. Lancet 366 (9499), 1761–1762. Wu, S.P., Tao, S., Liu, W.X., 2006. Particle size distributions of polycyclic aromatic hydrocarbons in rural and urban atmosphere of Tianjin, China. Chemosphere 62, 357–367. Xia, H.X., 2006. The preliminary study of introducing the superhigh chimney to the plain area of Beijing. Muncipal Administration and Technology 8, 70–72. Xu, S.S., Liu, W.X., Tao, S., 2006. Emission of polycyclic aromatic hydrocarbons in China. Environmental Science and Technology 40, 702–708. Xue, W.L., Warshawsky, D., 2005. Metabolic activation of polycyclic and heterocyclic aromatic hydrocarbons and DNA damage: a review. Toxicology and Applied Pharmacology 206, 73–93. Yang, L., He, K.B., Zhang, Q., Wang, Q.D., 2005. Vertical distributive characters of PM2.5 at the ground layer in autumn and winter in Beijing. Research of Environmental Science 18, 23–28. Zhang, Y.X., Tao, S., Cao, J., Coveney, R.M., 2007. Emission of polycyclic aromatic hydrocarbons in China by county. Environmental Science and Technology 41, 683–687. Zhou, J., Wang, T., Huang, Y., Mao, T., Zhong, N., 2005. Size distribution of polycyclic aromatic hydrocarbons in urban and suburban sites of Beijing, China. Chemosphere 61, 792–799.