Evaluation of Inhalation Exposure to Carcinogenic PM10-Bound PAHs of People at Night Markets of an Urban Area in a Metropolis in Eastern China

Inhalable particle bounded carcinogenic polycyclic aromatic hydrocarbons (PAHs) pose high health risks to both food service workers and consumers cooking or purchasing late night snacks, respectively, in night markets because PAHs are generated in abundance during thermal cooking processes such as barbecue grilling. In the present study, sixteen carcinogenic PAHs in PM10 collected during open hours from eight night markets in an urban area of a metropolis in eastern China were determined. The total concentration of PAHs (ΣPAH) ranged from 145 to 1340 ng m, with an arithmetic mean of 828 ± 360 ng m. Fiveand six-ringed PAHs were predominant in the composition, implying a combination of sources including cooking and traffic. Low coefficient of divergence values for individual PAH homologous among all sampling sites indicated similar sources. The occurrence of PM10-bound PAHs appeared to depend on the gasparticle partitioning processes and emission sources according to the results of principal component analysis (PCA). The daily inhalation rate of particle and predicted gaseous benzo[a]pyrene (BaP)-equivalent adjusted PAHs (BaPeq) for the workers due to occupational exposure was 0.451–3.43 μg day. Correspondingly, the occupational exposure risk for workers with one year of service was less than the acceptable risk level (10). However, the risk for people who had worked for more than 3 years exceeded 10 but was less than the priority risk level (10). For consumers, the maximum consumption time (tmax) for each time under the acceptable risk level would increase with an increase in age, and the exposure risk for infants, toddlers and children is high when PM10-bound PAHs are inhaled. We recommend that tmax should be less than 1 hour for consumers under 25 years old. Therefore, our results indicate that both workers and consumers in the night markets have high cancer risks due to inhalation of PM10-bound PAHs.


INTRODUCTION
Cooking and food processing are known to generate inhalable particle and various toxicants (Jägerstad and Skog, 2005).Among many mutagenic and carcinogenic compounds, polycyclic aromatic hydrocarbons (PAHs) are gaining wide public concern because these multi-ringed organic compounds are one of the first verified airborne carcinogens; some PAH congeners have been classified as "carcinogenic to humans" (Group 1) or "probably carcinogenic to humans" (Group 2A and 2B) (International Agency for Research on Cancer (IARC), 2012).For example, benzo[a]pyrene (BaP) and dibenz [a,h]anthracene (DahA) are listed in Group 1 and Group 2A, respectively.Benz[a]anthracene (BaA), chrysene (Chr), benzo [b]fluoranthene (BbF), and benzo[k]fluoranthene (BkF) are the members of Group 2B.Moreover, PAHs are heavily produced from various types of cooking sources via the incomplete combustion of organic materials (See et al., 2006;Farhadian et al., 2010), and cooking processes usually generate PAHs with high carcinogenic potency.Li et al. (2003) estimated that the total emission inventory for 21 PAHs from cooking sources (8,973 kg year -1 ) was significantly less than that from traffic sources (13,500 kg year -1 ) in a city of southern Taiwan.However, the toxic equivalency factor-adjusted emission rates (BaP was treated as a reference compound) from cooking sources (675 kg year -1 ) were more than those from traffic origins (61.4 kg year -1 ) (Li et al., 2003).Therefore, PAHs from incomplete combustion of organic materials during thermal cooking pose inevitable stressors to both occupational and public health (Kuo et al., 2005;Shen et al., 2011).
Chinese cooking methods and barbecue grilling with liquefied petroleum gas and charcoal as fuels generate large quantities of PAHs (Li et al., 2003;Chen et al., 2012).On one hand, these toxicants are contained the cooked foods, and consumers increase their cancer risk by dietary intake (Alomirah et al., 2011;Kao et al., 2014).On the other hand, PAHs can be released into air within gaseous and particulate phases, and workers and consumers increase their cancer risk by inhalation (Kuo et al., 2006;Chen et al., 2012).Carcinogenic PAHs with high molecular weights apparently tend to bind to particles due to their low volatilities (Shen et al., 2014).PAHs associated with particles having aerodynamic diameters below 10 µm (regarded as PM 10 ) can deposit into the human respiratory system because PM 10 is inhalable.
Nevertheless, most previous works have demonstrated indoor air pollution caused by PAHs from cooking processes (Bhargava et al., 2004;Ding et al., 2012;Shen et al., 2014).Limited studies regarding PAHs from outdoor cooking, e.g., barbecue stalls providing late night snacks in the night markets, have been published (Zhao et al., 2011).Currently, late night snacks, primarily grilled foods, are extremely popular in urban areas around China, and many well-known snack streets or squares in almost every city are composed of several barbecue stalls.Therefore, PM 10 -bound PAHs from barbecue grilling at night markets should be of great concern to protect the workers' and consumers' health.
The aims of the present study were as follows: (1) to investigate the occurrence of PM 10 -associated carcinogenic PAHs at the night markets in a metropolis in eastern China, (2) to estimate the daily inhalation dosage of target PAHs and to evaluate the potential cancer risk by inhalation for food service workers in the night markets, and (3) to provide the maximum consumption time for consumers.

Sampling Sites
Eight sampling sites (Fig. 1) total were selected in the urban area of Hefei, a metropolis in eastern China with a total population of approximately 2.2 million.Except for the site located in the northern gate of the western campus of USTC where many barbecue stalls were closed because of summer vacation for university students, seven sites were night market areas with dense food stalls where barbecue grilling was the primary cooking method (Table 1).Liquefied petroleum gas was commonly used for fuel.Approximately three workers were in each stall, and dining tables for consumers were behind the stalls.Furthermore, the night markets were near the traffic roads, and no factories were near the eight sampling sites.Additionally, a control site in a rural area in southwestern Hefei, which was far from any night market, was selected to collect PM 10 samples for comparison with the results from the night markets.

Air Sampling and Metrological Conditions
According to the questionnaire before our sampling, cooking in the market is performed primarily from 20:00 h to 2:00 h.Thus, the sampling was conducted at night from 20:00 h to 2:00 h.PM 10 samples were collected by air samplers (ZC-Q0102, Zhejiang Hengda Instrument & Meter Co., Ltd., Zhejiang, China) with a Whatman glass fiber filter (GFF,102 mm i.d.).All of filters were pre-heated at 450°C for 5 hours, then wrapped in pre-baked aluminum foil packages, sealed in airtight polyethylene bags and reserved in a desiccator.Before the sampling, the filters were weighed.The samplers were set at a height of approximately 1.5 m above the ground within the breathing zone.The airflow was set to approximately 90 L min -1 .After the sampling, the GFFs were put into refrigerator at -4°C immediately and were ready for the following treatment.
Information regarding the sampling is presented in Table 1.Briefly, the air temperature was approximately 24-29°C during the summer sampling period from July 16 to 23, 2012, and the wind strength was under 3-4 Beaufort wind force scale.

Extraction and Clean up
Before Soxhlet extraction, each sample was spiked with the surrogate standards.The GFFs were extracted with a 1:1 mixture of hexane and dichloromethane for 48 hours.The extracts were concentrated to approximately 1 mL using a rotary evaporator (ZX98-1, Shanghai Yukang), and this step was repeated after the solvent was exchanged to 10 mL hexane.Then, the samples were cleaned using a chromatographic column packed with pre-extracted cotton, 12 cm silica (180°C baked) and 2 cm anhydrous sodium sulfate from the bottom to the top.The fraction containing  PAHs was eluted by 15 mL of hexane:dichloromethane mixture (7:3 in volume).The elute was concentrated to 1 mL using a rotary evaporator, and then the solvent was exchanged to hexane and was further reduced to 0.5 mL with a soft stream of purified nitrogen.The final samples were spiked with the internal standards and stored at -4°C until they were analyzed.

Instrumental Analysis
A gas chromatograph-mass spectrometer (GCMS-QP2010 Plus supplied by Shimadzu Scientific Instruments, Kyoto, Japan) was employed to quantify the individual concentrations of 16 PAH congeners.A DB-5MS (length = 60 m; inner diameter = 0.25 mm; film thickness = 0.25 m) column was used for chromatographic separation.The initial oven temperature was set at 60°C and programmed at a rate of 10 °C minute -1 to 200°C, at a rate of 2 °C minute -1 to 250°C, and finally at a rate of 20°C minute -1 to 290°C and then kept at 290°C for 5 minutes.The injection port temperature was programmed from 100°C to 280°C at a rate of 200 °C minute -1 and kept at 280°C for 10 minutes.Each sample (1 µL) was injected using an automatic AOC-20 Series sampling system in the splitless mode.One minute after injection, the split mode was switched on with a split ratio of 100:1, and five minutes later, the ratio was changed to 20:1 to save the carrier gas.The ion source and interface temperatures were both maintained at 250°C.

Quality Assurance
Three field blanks were applied to monitor the background contamination and were processed sequentially with field samples.Three method blanks were also analyzed.These contaminants did not interfere with our targets.Surrogate standards were added into each sample as the recovery indicators.The average recovery percentages of these substances were 113 ± 12%, 117 ± 10%, and 99 ± 13% (mean ± relative standard deviation, n = 14).

Data Analysis
All measured concentrations of particulate PAHs were not adjusted by the recovery efficiencies.The total concentration of the 16 PAHs was defined as ΣPAH.The analysis of the correlation between the concentrations of ΣPAH and PM 10 was performed using two-sided Spearman's correlation at a significance level of 0.05, and principal component analysis of the mass concentrations of individual PAH compounds was performed using SPSS 16.0 software package (SPSS Inc., Chicago, IL, USA).
(1) Coefficient of divergence (CD) To evaluate the spread of the data points for two datasets, the coefficient of divergence (CD) was employed as described in Eq. (1): where i and j are the two profiles of the sampling sites, p is the number of components; and C ij and C ik are the concentrations of component i at sampling sites j and k, respectively. (

2) Estimation of gaseous concentrations of individual PAH congeners
To estimate the gaseous concentrations of individual PAH congeners, the particle-gas partition coefficient (K p, with a unit of m 3 µg -1 ) was applied as described elsewhere (Cincinelli et al., 2014): 10 ( ) ( ) ( ) where C P (i), and C g (i) are the concentrations of PAH congener i in PM 10 and in gaseous phase (ng m -3 ), respectively, and PM 10 is is the concentration of PM 10 (ng m -3 ).K P of each compound generally is a function of its subcooled liquid vapor pressure (P L ) at 25°C, and log K P can be linearly correlated to the logP L as shown in Eq. ( 3) as proposed by Simcik et al. (1998).P L values ranged from 9.51 × 10 -8 Pa for DahA to 0.115 Pa for Phe (Paasivirta et al., 1999) (3) Inhalation exposure and health risk assessment The inhalation exposure of the PAH mixture was adjusted by the toxicity equivalent factor (TEF) to BaP toxicity equivalent (BaP eq ) of each compound (Nisbet and LaGoy, 1992).
The daily inhalation exposure level (IEL) of PAHs for the occupational staff was calculated by the BaP eq concentration, inhalation rate (IR, m 3 h -1 ) and daily working time (t) in the following equation: Considering that the vendors in the night market are adults (usually ages ranging from 20-50 years old), the IR was selected at 1.6 m 3 h -1 (Duan, 2012), and the daily work time was set at 6 hours (from 20:00 h to 2:00 h).
The incremental lifetime cancer risk (ILCR) of vendors by inhalation exposure to ambient air PAHs was computed by Eq. ( 6): where SF is the cancer slope factor for BaP inhalation exposure (a geometric mean of 3.14 kg day mg -1 ± a geometric standard deviation of 1.80) (Chen and Liao, 2006), EF is the exposure frequency (day year -1 , with 250 d year -1 selected in the present study), T is the exposure duration (year, with 1, 3, 5, 10 and 20 years selected to evaluate the risk of the vendors with 1, 3, 5, 10, and 20 years of seniority, respectively), CF is the conversion factor (10 -6 mg ng -1 ), BW is the body weight of the vendors (67.1 ± 11.4 kg) (Xia et al., 2013), and AT represents the average lifespan for carcinogens (25550 days) (Xia et al., 2013).
(4) Estimation of maximum consumption time (t max ) To estimate t max in the barbecue night market for consumers under acceptable ILCR (ILCR acceptable ) at 10 -6 , we hypothesized consumption frequency in the barbecue night market at once per week before 55 years old with the lifespan for carcinogens of 70 years old (25550 days).The maximum consumption time for each time in the night market could be calculated by Eq. ( 7):

PM 10 and PAH Concentrations
The level of PM 10 in the present study ranged from 168 to 426 µg m -3 , with an arithmetic mean ± deviation of 267 ± 100 µg m -3 , while the concentration of PM 10 at the control site was 55.1 µg m -3 .The mean concentration of PM 10 was 2-fold higher than that observed at a night market for barbecue (Kuo et al., 2005).Compared with other occupational circumstances, field traffic police personnel (Majumder et al., 2012) and poultry farming (Lawniczek-Walczyk et al., 2013) have a greater risk of exposure to PM 10 ; however, the employees of public transportation (Gerber et  al., 2014), lignite-fired power plant (Manousakas et al., 2013), and Sports facilitie (Alves et al., 2013) had lower exposure to PM 10 .All values obtained in the present study exceeded the primary national ambient air quality standard of China for PM 10 (150 µg m -3 ) (Ministry of Environmental Protection of the People's Republic of China, 2012).Therefore, PM 10related airborne particulate matter exposure is considered an important environmental stressor for vendors of the night markets because numerous studies have confirmed the negative effects of chronic and acute exposure to PM 10 on human health (Pope et al., 1991;Giri et al., 2007).
The mean concentration of ∑PAH was 828 ± 360 ng m -3 , with a range from 145 to 1340 ng m -3 at night markets and 101 ng m -3 at control site.The maximum concentration of ΣPAH was observed at Datong Square, which has more than 20 barbecue stalls and which is the most popular barbecue night market in Hefei City.The lowest concentration of ΣPAH was investigated at the northern gate of the western campus of the University of Science and Technology of China.During the sampling period, most barbecue stalls were closed because of summer vacation for university students.Furthermore, the concentrations of PAHs in the present study were significantly lower than those investigated in the barbecue night markets that used charcoal as fuel, where the concentration of 22 PAH congeners ranged from 1690 to 31000 ng m -3 .However, our results were comparable with the values of PM 10 -bound PAH concentrations found in the barbecue night markets that used electricity as fuel (Kuo et al., 2005) and with the values of PM 2.5 -bound PAH concentrations found in the barbecue night markets that used liquefied petroleum gas as fuel (Zhao et al., 2011).The concentrations of ΣPAH in the present study were also lower than those observed in suburban areas (Jamhari et  al., 2014) and in high-density traffic areas (Chen et al., 2013).The low concentrations of PM 10 -bound PAHs found in the barbecue night markets in Hefei City were most likely due to the prohibition of charcoal as fuel for barbecue (Wu et al., 2012).PAHs could be formed both during cooking and food processes (Onyango et al., 2012) and fuel combustion (Hou et al., 2008).Previous studies have demonstrated that the concentrations of PAHs in grilled foods depend on the thermal processes and grilling methods (Farhadian et al., 2010;Viegas et al., 2012).Generally, the concentrations of PAHs in charcoal-grilled foods are significantly higher those in gas-grilled foods and electric oven-grilled foods (Farhadian et al., 2010).Therefore, fewer PAH emissions are expected when grilling foods with liquefied petroleum gas as fuel compared with charcoal as fuel.Furthermore, liquefied petroleum gas combustion during cooking causes lower concentrations of ambient PAHs than does charcoal consumption (Titcombe and Simcik, 2011).However, the PM 10 -bound PAH concentrations in the barbecue night markets in Hefei City were 1-2 orders of magnitude higher than those detected in urban areas (Mantis et al., 2005;Hong et al., 2007;Vu et al., 2011;Limu et al., 2013;Wu et al., 2014a), indicating that staff and consumers in barbecue night markets are significantly exposed to carcinogenic PAHs.
Two-tailed Spearman's correlation analysis indicated that the concentrations of ∑PAH did not correlate to the content of PM 10 (R 2 = 0.31 and p = 0.45), which is partly attributable to the effects of particle sizes because the distribution of airborne PAHs usually depends on the particle sizes (Kawanaka et al., 2009;Zhang et al., 2012) and because these semi-volatile organic pollutants with low vapor pressures are generally associated with ultrafine particles (Offenberg and Baker, 1999).A previous study demonstrated that most PAHs from grilled foods (corn, trout, beef, prawns, and pork) were observed in particles less than 0.43 µm in size (Saito et al., 2014); however, PAHs were predominantly emitted within the 1.1-2.1 µm particles when grilling fishes (Tanaka et al., 2012).A study regarding the PAH emissions from crop residues burned for cooking also indicated that particles with an aerodynamic diameter between 1.1-2.1 µm were capable of concentrating carcinogenic aromatic hydrocarbons (Shen et al., 2011).Additionally, temperature is expected to affect the redistribution of PAHs between airborne particle matter and gases (Tsapakis and Stephanou, 2005), and wind speed also affects the concentration of particle-bound PAHs by the resuspension of street dust (Martuzevicius et  al., 2011).Although all samples were collected during the same time (between 20:00 h to 2:00 h), the air temperature, wind speed and wind direction slightly varied during the sampling period (Table 1) in the present study.Consequently, particle-bound PAHs were not distinctly dependent on the PM 10 content.

Compositional Characteristics
In the present study, the most abundant PAH was BbF (15.0 ± 2.47%), followed by BghiP (14.4 ± 2.49%), BeP (10.8 ± 1.92%), IcdP (10.3 ± 1.89%), and Flu (9.03 ± 3.16) (Fig. 2).The samples contained low concentrations of methylated Phe and Per, with the relative abundance of individual PAH compounds less than 4%.Low CD values (0.15-0.42) between each two sampling sites were investigated, reflecting all sampling sites with similar composition distributions and further implying similar sources of PM 10 -associated PAHs for all sampling sites because a CD value close to zero generally suggests similarities in PAH sources among sites, while a CD value approaching 1 implies different PAH sources (Shi et al., 2009;Kong et al., 2010;Kong et al., 2013).
A previous study indicated that BbF, IcdP, BeP were the most abundant PM 2.5 -associated PAHs in Chinese cooking stalls (See et al., 2006), whereas particle samples collected in the kitchens that used liquefied petroleum gas as fuel (See et al., 2006) and in the night market (Zhao et al., 2011) contained the highest observed concentrations of BghiP, IcdP, Flu, Pyr and BaA.Therefore, the PM 10 -bound PAHs in the night markets of Hefei City were predominantly derived from cooking.Flu and Pyr are usually considered markers of traffic-related sources (Ho et al., 2009).In the present study, high concentrations of Flu and Pyr also suggest vehicle-associated mobile sources because many gasolineand liquefied petroleum gas-fueled vehicles were present around the night markets for consumer transport.Therefore, airborne PAHs in the night markets of Hefei City originated from combined cooking and vehicle emissions.
The gaseous concentrations of individual PAH congeners were estimated to range from 3970 ng m -3 for Phe to < 1 ng m -3 for heavy molecular weight PAHs (BbF, BkF, BaP, Per, IcdP, DahA and BghiP (Fig. 2).High concentrations of light molecular weight PAHs were predicted in the gaseous phase because of their volatilization.Previous studies have indicated that light molecular weight PAHs predominantly occur in the gaseous phase, while heavy molecular weight PAHs tend to accumulate in the particle phase (Tasdemir and Esen, 2007;Vasilakos et al., 2007).However, heavy molecular weight PAHs are likely to be underestimated in the gaseous phase because PAHs from cooking can occur in oil fumes, which are usually considered in gaseous phase.Therefore, high concentrations of heavy molecular weight PAHs were observed in gaseous phase at barbecue night market (Zhao et al., 2011) and in kitchens (Chen et al., 2012).It should be noted that the estimated gaseous concentrations of ΣPAH were 2.09-21.1 times higher than the measured PM 10 -bound PAH concentrations, but BaP eq concentrations in gaseous phase were significantly lower than those found in PM 10 samples.The primary reason is due to the fact that the predicted gaseous concentrations of ΣPAH were largely dominated by Phe and Ant (accounting for 84.4 ± 5.3%), which are characterized by high volatilities and low toxicity equivalent factors when comparing with those high molecular weight PAHs.

Principal Component Analysis
In the present study, principal component analysis based on the particle-bound PAH data potentially classified PAH compounds into two groups (Fig. 3(a)), and the top two components explained 87.6% of the total variance, with component 1 (PC 1) = 60.0%and component 2 (PC 2) = 27.6%.PC 1 was associated primarily with the heavy weight molecular PAHs, including BbF, BkF, BeP, BaP, Per, IcdP, DahA, and BghiP, as well as Ant, with negative loading.This component was possibly affected by the volatilities of PAHs and gas-particle partitioning processes because the PC 1 loading of the rotated component matrix clearly correlated with the subcooled liquid vapor pressures (P L o ) of PAH isomers (Fig. 3  contaminants between gas and particle phases are generally explained by physical adsorption onto the surface of particle matter or adsorption into the organic matter, both of which are associated with the P L o (Simcik et al., 1998).Consequently, numerous studies have demonstrated that the particle-gas partition coefficients of semi-organic contaminants correlate well with their P L o values on a logarithm scale (Zhang et  al., 2009; Tseng et al., 2014; Wu et al., 2014b).Although PAH isomers are emitted primarily in gaseous phase during high-temperature pyrolysis processes, gaseous PAHs are quickly partitioned into particle matter by condensation and adsorption, and the equilibrium between gas and particle phases is reached within a short time (Liang and Pankow, 1996;Spezzano et al., 2009).PC 2 loading of the rotated component matrix was potentially distinguished as cooking-and traffic emission-related PAHs.Briefly, mass concentrations of BghiP, BbF, IcdP, BeP, BaA, and two alkylated PAHs negatively correlated with their PC 2 loading values; however, the concentrations of Ant, BaP, DahA, BkF, Flu, Chr, and Phe positively correlated to their PC 2 loadings (Fig. 3(c)).As discussed above, BghiP, IcdP and BeP would be expected from cooking activities (See et al., 2006), and Flu and Pyr would be considered the markers of vehicleassociated mobile sources.Therefore, PC 2 represented the emission sources of PAHs.

Occupational Inhalation Exposure
The IEL of the night market staff to the BaP-adjusted PAHs was 1.75 ± 0.878 µg day -1 , with a range of 0.451-3.43µg day -1 , indicating high potential exposure risk due to the occupation.The highest IEL value was investigated at Datong Square, and the lowest value of IEL was found at the northern gate of the western campus of the University of Science and Technology of China, where the highest and the lowest concentrations of ΣPAH were obtained, respectively.The daily inhalation dose in the present study is comparable to that for the population in the night market in Taiwan, which was significantly affected by grilling practices (Kuo et al., 2005).A previous study also showed temple workers in Taiwan with a comparable daily inhalation dose (0.076 µg kg day -1 at the 50 th percentile) (Chiang et al., 2009).Notably, only inhalation exposure was considered in the present study; ingestion of and dermal contact with particlebound PAHs would potentially be expected to highly contribute to total human exposure to airborne PAHs (Chiang et al., 2009;Chen et al., 2012).
In the present study, we estimated the potential ILCR by the occupational exposure to PAHs of 1, 3, 5, 10 and 20 years (Fig. 4).Our results suggest that the occupational inhalation risk of the workers with 1 year of service time was primarily less than the acceptable risk level of 10 -6 (1.3 × 10 -7 -8.4 × 10 -7 ) for all sampling sites, while the risk was greater than 10 -6 in five sites for the people who had worked for 3 years (Fig. 4(a)).However, all ILCR values were higher than 10 -6 but lower than the priority risk level (10 -4 ) for the staff with 20 years of work experience (2.5 × 10 -6 -1.7 × 10 -5 ).However, the people at control site would be expected to have a low inhalation exposure risk because the ILCR values were less than 10 -6 , except for 20-year inhalation exposure, which was 1.3 × 10 -6 .These results indicate the potential carcinogenic risk for the workers in the night markets of Hefei City, a typical metropolis in China.The highest and lowest ILCR values were found at Datong Square and at the northern gate of the western campus of the University of Science and Technology of China, where the highest and lowest concentrations of ΣPAH were obtained, respectively.The probability of worker exposure to PM 10bound PAHs in night markets was also evaluated by combining the probit mode with the estimated ILCR values, indicating that these people will be exposed to acceptable risk levels of airborne PAHs, with probabilities of 0.26% for 1 year, 66.0% for 3 years, 85.3% for 5 years, 93.7% for 10 years and 96.1% for 20 years (Fig. 4(b)).
Combining the acceptable risk level, t max was estimated for the customers with ages from 1 to 55 years old (Fig. 5).The acceptable t max was 2.5-7.1 minutes for one-year-old infants, 3.9-11 minutes for 3-year-old toddlers, 5.5-29 minutes for 5-10-year-old children, 20-57 minutes for 15year-old adolescents, 0.48-1.34hours for 20-year-adults, and 1.99-5.60hours for 50-year-old adults.Therefore, the estimated consumption time under the acceptable risk level increased with an increase in age, and infants, toddlers and children are vulnerable to a high exposure risk by inhalation of PM 10 -bound PAHs.Specifically, the maximum exposure time in the night markets for the customers under 25 years old should be less than 1 hour.

Fig. 1 .
Fig. 1.Sampling sites and spatial distribution of PM 10 -bound ΣPAH concentrations.The sizes of the symbols are proportional to the concentration values.

Fig. 3 .
Fig. 3. Rotated component matrix of 16 PAH homologs in PM 10 samples by principal component analysis (a); the loading of component 1 (PC 1) and 2 (PC 2) was plotted against the subcooled liquid vapor pressures (P L o ) of PAHs (b) and their mass concentrations (c).

Fig. 5 .
Fig. 5. Predicted maximum consumption time for consumers of different ages.
(b)).The partition processes of semi-organic Average concentrations for particulate (measured) and gaseous (predicted) phase PAHs in the night markets, Hefei City, eastern China.