Pollution Characteristics and Health Risk Assessment of Summertime Atmospheric Volatile Halogenated Hydrocarbons in a Typical Urban Area of Beijing, China

Twenty-three atmospheric volatile halogenated hydrocarbons (VHHs) were detected in a typical urban area of Beijing, China from 24 August to 4 September, 2012. The mean and range in daily mass concentrations of the 23 VHHs were 30.53 and 13.45–76.33 µg/m3, respectively. Seven of those VHHs were controlled ozone-depleting substances in China, with a mean of 12.95 µg/m3, accounting for 42.43% of the total. Compared with other national and international cities, the concentrations of the selected 11 VHHs in this study were relatively higher. Dichloroethane had the highest mass concentration, followed by difluorochloromethane. Maxima of total VHHs occurred within the period 8:30–9:00 a.m., while minima occurred during 1:30–2:00 p.m. Source apportionment suggested that the main sources of VHHs in the study area were solvents usage and industrial processes, leakage of chlorofluorocarbons banks, refrigerants, and fumigant usage. Among the selected 7 VHHs, trichloromethane, tetrachloromethane, 1,2-dichloroethane, and 1,4-dichlorobenzene posed potential carcinogenic risks to exposed populations, while none of the selected 11 VHHs posed appreciable non-carcinogenic risks to exposed populations. The carcinogenic risks from atmospheric VHHs in Beijing are higher than in other Chinese cities, indicating that it is necessary to implement immediate control measures for atmospheric VHHs in Beijing.

. Location of sampling site and surroundings [21].
Before sampling, all canisters were cleaned using a canister cleaner (3100A, Entech Instruments, Inc., Simi Valley, CA, California, USA) and vacuumed to 50 mtorr. The sampling speed was controlled by a flow-limiting valve. The canisters were placed 1.5 m above the rooftop. Meteorological data at the sampling site, including wind direction, wind speed, temperature, dew point temperature, solar radiation, ultraviolet radiation, and visibility, were monitored hourly using an automatic weather station (Vaisala Inc., Helsinki, Finland) from 28 August to 4 September, 2012.

Sample Analysis
All air samples were analyzed using cryogenic cold trap preconcentration followed by gas chromatography (GC) coupled with mass spectrometry (MS) and flame ionization detection (FID) within 15 days. Each volatile organic compound (VOC) sample was enriched and concentrated by passing it through the Entech 7100A pre-concentration system (Entech Instruments, Inc., Simi Valley, CA, USA), whereby water and CO2 were also removed from the sample. VOCs were rapidly gasified and fed into the GC-MS/FID system (GC, HP-7890A, Agilent Technology, Inc., California, USA; MS, HP-5975C, Agilent Technology, Inc., California, USA) to be separated and quantitatively analyzed [23]. Ninety-seven VOCs, including 25 VHHs, were separated by the DB-624 chromatographic column (60 m × 0.25 mm × 1.8 µm; J&W Scientific, Folsom, CA, USA) and the PLOT chromatographic column (30 m × 0.25 mm × 3.0 µm; J&W Scientific, Folsom, CA, USA). All these species were quantitatively analyzed by MS or FID. For the MS analysis, the temperature and energy of the ion source were 200 °C and 70 eV, respectively; the scanning speed was 4 scan/s; and the scanning scope was 35-350 amu. For the chromatographic analysis, the total run time was 47 min. The initial temperature in the GC oven was kept at 30 °C for 7 min, then increased to 120 °C at a rate of 5 °C/min. It was kept at 120 °C for 5 min, and then increased to 180 °C with a rate of 6 °C/min, and held at 180 °C for 7 min.
Standard compounds and internal standards were used to build multi-point calibration curves for quantitative analysis of the VOCs. The standard gases were multi-component gas, including 56 VOCs (PAMS, Scott Specialty Gases Company, Plumsteadville, PA, USA), as well as the multicomponent standard gas recommended in the TO-15 method (Scott Specialty Gases Company, Plumsteadville, PA, USA). The three internal standard compounds were bromochloromethane, pdifluorobenzene, and 1-bromo-3-fluorobenzene (Scott Specialty Gases Company, Plumsteadville, PA, USA). Samples were collected from 24 August to 4 September 2012 over four 30-min periods: 8:30-9:00 a.m., 1:30-2:00 p.m., 6:00-6:30 p.m., and 10:00-10:30 p.m. However, sampling was postponed from noon on 1 September to the morning of 2 September because of rain. In total, 43 air samples were collected in SUMMA canisters (Entech Instruments, Inc., Simi Valley, CA, USA); thirty canisters had volumes of 3.2 L, while the others had volumes of 3.0 L. The recommended TO-15 sampling method of the US Environmental Protection Agency (US EPA) was used as our sampling method [22].
Before sampling, all canisters were cleaned using a canister cleaner (3100A, Entech Instruments, Inc., Simi Valley, CA, USA) and vacuumed to 50 mtorr. The sampling speed was controlled by a flow-limiting valve. The canisters were placed 1.5 m above the rooftop. Meteorological data at the sampling site, including wind direction, wind speed, temperature, dew point temperature, solar radiation, ultraviolet radiation, and visibility, were monitored hourly using an automatic weather station (Vaisala Inc., Helsinki, Finland) from 28 August to 4 September 2012.

Sample Analysis
All air samples were analyzed using cryogenic cold trap preconcentration followed by gas chromatography (GC) coupled with mass spectrometry (MS) and flame ionization detection (FID) within 15 days. Each volatile organic compound (VOC) sample was enriched and concentrated by passing it through the Entech 7100A pre-concentration system (Entech Instruments, Inc., Simi Valley, CA, USA), whereby water and CO 2 were also removed from the sample. VOCs were rapidly gasified and fed into the GC-MS/FID system (GC, HP-7890A, Agilent Technology, Inc., Santa Clara, California, USA; MS, HP-5975C, Agilent Technology, Inc., Santa Clara, California, USA) to be separated and quantitatively analyzed [23]. Ninety-seven VOCs, including 25 VHHs, were separated by the DB-624 chromatographic column (60 m × 0.25 mm × 1.8 µm; J&W Scientific, Folsom, CA, USA) and the PLOT chromatographic column (30 m × 0.25 mm × 3.0 µm; J&W Scientific, Folsom, CA, USA). All these species were quantitatively analyzed by MS or FID. For the MS analysis, the temperature and energy of the ion source were 200 • C and 70 eV, respectively; the scanning speed was 4 scan/s; and the scanning scope was 35-350 amu. For the chromatographic analysis, the total run time was 47 min. The initial temperature in the GC oven was kept at 30 • C for 7 min, then increased to 120 • C at a rate of 5 • C/min. It was kept at 120 • C for 5 min, and then increased to 180 • C with a rate of 6 • C/min, and held at 180 • C for 7 min.
Standard compounds and internal standards were used to build multi-point calibration curves for quantitative analysis of the VOCs. The standard gases were multi-component gas, including 56 VOCs (PAMS, Scott Specialty Gases Company, Plumsteadville, PA, USA), as well as the multi-component standard gas recommended in the TO-15 method (Scott Specialty Gases Company, Plumsteadville, PA, USA). The three internal standard compounds were bromochloromethane, p-difluorobenzene, and 1-bromo-3-fluorobenzene (Scott Specialty Gases Company, Plumsteadville, PA, USA). Twenty-five VHHs were analyzed, including 4 chlorofluorocarbons (CFCs), 1 hydrochlorofluorocarbons (HCFCs), 10 chloroalkanes, 4 chloroalkenes, 3 chlorinated benzenes, 2 bromohydrocarbons, and 1 bromine chlorine hydrocarbon. However, 1,1,1-trichloroethane and bromine chlorine hydrocarbon were not detected in all samples. Daily average mass concentrations of the other 23 VHHs, as well as their method detection limits and linear correlation coefficients (r 2 ) of their working curves, are given in Table 1.

Quality Assurance and Quality Control
Quality assurance and quality control in this study relied on method detection limits, retention times, detection accuracy, and reference sample analyses. The standard working curves were established using standard gases outlined in TO-15, PAMS, and three internal standard gases. The collection and analysis process for all samples were performed according to the requirements of the US EPA Compendium Method TO-15 [22]. The R 2 of the calibration curves for all samples was >0.998 and the range in deviation between our results and theoretical values was 1 ± 20%.

Data Processing
Levels, temporal variations, sources, and health risk were studied based on the data in this observation. It should be noted that the related conclusions based on these limited data may have some uncertainty. In order to obtain more general research results, long-term and systematic observations of atmospheric VHHs are needed.

Source Apportionment Using Positive Matrix Factorization
The positive matrix factorization (PMF) model is a multivariate factor analysis tool and matrix decomposition method that integrates error estimates in data to the restricted-weight, least-squares linear model. It is widely used to analyze the source of aerosols, wet and dry deposition, and VOCs [24,25]. In this study, PMF5.0 recommended by US EPA was used to analyze the sources of atmospheric VHHs in the study area. The details of species selection and data processing were documented elsewhere [21,26]. In brief, species with no available mixing ratios or species for which >25% samples had mixing ratios lower than the detection limit have been removed. Species with high reactivity or low mixing ratios or those which were not important tracers of pollution also have been removed [26]. Eventually, 19 VHHs were selected to analyze the sources.

Health Risk Assessment
Health risk assessment methods can be used to estimate the adverse effect of pollutants on human health. Such methods have been adopted worldwide and are continually updated. In 1983, the US National Academy of Sciences proposed a four-step health risk assessment, which includes hazard identification, dose-response assessment, exposure assessment, and risk characterization [27].
Given the carcinogenicity of some pollutants, the health risk assessment is divided into carcinogenic risk and non-carcinogenic risk assessments; both assessments should have been carried out for carcinogenic pollutants. The US EPA proposed a health risk assessment method based on respiratory rate and body weight for inhalable pollutants in 1989 [28], and updated it in 2009 [29]. In this latest version, concentrations of the pollutant in air is used as the exposure metric (e.g., mg/m 3 ), rather than its inhalation intake based on human inhalation rate and body weight. This methodology has been widely accepted for human health assessment worldwide [30][31][32][33][34]. We used the method proposed by the US EPA to make the health risk assessment in this study. Here, carcinogenic risk is expressed as Risk (dimensionless), while non-carcinogenic risk is expressed using the hazard index (HI, dimensionless), i.e., a sum of the hazard quotients (HQ, dimensionless) for several pollutants [29]. The details are shown as below: The chronic or subchronic exposure concentration (EC, µg/m 3 ) can be estimated with the following equation: The Risk can be estimated with the following equation: The HQ and HI can be estimated with the following equation: In Equations (1)-(4), CA (µg/m 3 ) is the contaminant concentration in air; ET (24 h/d) is exposure time; EF (365 day/year) is exposure frequency; ED (70 year) is exposure duration; AT (70 year × 365 day × 24 h) is averaging time; IUR (m 3 /µg) is inhalation unit risk [35]; and RfC (mg/m 3 ) is the reference concentration [35].
Atmosphere 2020, 11, x FOR PEER REVIEW 6 of 19 5.55, 17.06, 0.63, 0.70, and 0.09 µg/m 3 , respectively, yielding percentages of 21.32%, 18.17%, 55.87%, 2.05%, 2.28%, and 0.31% of the total mass concentration of the 23 VHHs ( Figure 2). Chloroalkanes, CFCs, and HCFCs were the main VHHs in this study area, accounting for 95% of the 23 VHHs. The daily average mass concentrations of dichloromethane, chlorodifluoromethane, dichlorodifluoromethane, 1,2-dichloroethane, monochloromethane, 1,2-dichloropropane, and trichlorofluoromethane were 6.43, 5.55, 3.46, 3.09, 2.87, 2.38, and 2.17 µg/m 3 , respectively, accounting for 21.07%, 18.17%, 11.35%, 10.13%, 9.40%, 7.79%, and 7.09% of the total mass concentration of the 23 VHHs ( Figure 3). Thus, these seven species were the main VHHs in this study, accounting for 85% of the total mass concentration of the 23 VHHs.  With VHHs getting more widespread in the atmosphere, some VHH species have become a concern to atmospheric scientists. Long-term monitoring of some volatile organic compounds (VOCs), which have important impacts on the global climate or are radiative gases, has been carried out as part of the Advanced Global Atmospheric Gases Experiment (AGAGE) and its subsidiary network. Among these VOCs, 11 VHH species, including trichlorofluoromethane, dichlorodifluoromethane, trichlorotrifluoroethane, dichlorotetrafluoroethane, chlorodifluoromethane, tetrachloromethane, bromomethane, monochloromethane, dichloromethane, trichloromethane, and tetrachloroethene, were observed in this study. Comparisons of mixing ratios of the 11 VHHs in the study area against their global background levels from AGAGE are shown in Table 2. Mixing ratios of the 11 VHHs, including 7 controlled ODS in China, were all higher than their global background levels. In particular, the mixing ratios of dichloromethane, trichloromethane, and tetrachloroethene in this study were 42.34, 14.17, and 25.73 times higher than their global background levels, respectively, Atmosphere 2020, 11, x FOR PEER REVIEW 6 of 19 5.55, 17.06, 0.63, 0.70, and 0.09 µg/m 3 , respectively, yielding percentages of 21.32%, 18.17%, 55.87%, 2.05%, 2.28%, and 0.31% of the total mass concentration of the 23 VHHs ( Figure 2). Chloroalkanes, CFCs, and HCFCs were the main VHHs in this study area, accounting for 95% of the 23 VHHs. The daily average mass concentrations of dichloromethane, chlorodifluoromethane, dichlorodifluoromethane, 1,2-dichloroethane, monochloromethane, 1,2-dichloropropane, and trichlorofluoromethane were 6.43, 5.55, 3.46, 3.09, 2.87, 2.38, and 2.17 µg/m 3 , respectively, accounting for 21.07%, 18.17%, 11.35%, 10.13%, 9.40%, 7.79%, and 7.09% of the total mass concentration of the 23 VHHs ( Figure 3). Thus, these seven species were the main VHHs in this study, accounting for 85% of the total mass concentration of the 23 VHHs.  With VHHs getting more widespread in the atmosphere, some VHH species have become a concern to atmospheric scientists. Long-term monitoring of some volatile organic compounds (VOCs), which have important impacts on the global climate or are radiative gases, has been carried out as part of the Advanced Global Atmospheric Gases Experiment (AGAGE) and its subsidiary network. Among these VOCs, 11 VHH species, including trichlorofluoromethane, dichlorodifluoromethane, trichlorotrifluoroethane, dichlorotetrafluoroethane, chlorodifluoromethane, tetrachloromethane, bromomethane, monochloromethane, dichloromethane, trichloromethane, and tetrachloroethene, were observed in this study. Comparisons of mixing ratios of the 11 VHHs in the study area against their global background levels from AGAGE are shown in Table 2. Mixing ratios of the 11 VHHs, including 7 controlled ODS in China, were all higher than their global background levels. In particular, the mixing ratios of dichloromethane, trichloromethane, and tetrachloroethene in this study were 42.34, 14.17, and 25.73 times higher than their global background levels, respectively, With VHHs getting more widespread in the atmosphere, some VHH species have become a concern to atmospheric scientists. Long-term monitoring of some volatile organic compounds (VOCs), which have important impacts on the global climate or are radiative gases, has been carried out as part of the Advanced Global Atmospheric Gases Experiment (AGAGE) and its subsidiary network. Among these VOCs, 11 VHH species, including trichlorofluoromethane, dichlorodifluoromethane, trichlorotrifluoroethane, dichlorotetrafluoroethane, chlorodifluoromethane, tetrachloromethane, bromomethane, monochloromethane, dichloromethane, trichloromethane, and tetrachloroethene, were observed in this study. Comparisons of mixing ratios of the 11 VHHs in the study area against their global background levels from AGAGE are shown in Table 2. Mixing ratios of the 11 VHHs, including 7 controlled ODS in China, were all higher than their global background levels. In particular, the mixing ratios of dichloromethane, trichloromethane, and tetrachloroethene in this study were 42.34, 14.17, and 25.73 times higher than their global background levels, respectively, suggesting that there were stronger source emissions for these three VHHs in the study area [36]. Although the mixing ratios of monochloromethane and chlorodifluoromethane in this study were as high as 1413.6 and 1589.3 pptV, respectively, they have relatively lower levels compared with their global background levels (540.51 and Atmosphere 2020, 11, 1021 7 of 19 231.16 pptV, respectively). As controlled ODS in China, the mixing ratios of trichlorofluoromethane, dichlorodifluoromethane, dichlorotetrafluoroethane, 1,1,2-trichlorotrifluoroethane, tetrachloromethane, and bromomethane were comparable with their global background levels.
The concentrations of these 11 VHHs in the study area compared with other national and international cities are also shown in Table 2 to provide context to their pollution status in Beijing. The mixing ratio of monochloromethane in the study area was higher than the average mixing ratio of monochloromethane in 45 Chinese cities, as well as Bristol (UK), but lower than in Changsha (China). The mixing ratio of dichloromethane in the study area was higher than the average concentration of dichloromethane in 45 Chinese cities, as well as Changsha (China), Bristol (UK), Philadelphia (USA), and Marseille (France), but lower than in Tianjin (China) and Shenyang (China). Likewise, the mixing ratio of trichloromethane in the study area was higher than the average mixing ratio of trichloromethane in 45 Chinese cities, as well as Shenyang (China), Changsha (China), Bristol (UK), Philadelphia (USA), and Marseille (France), but lower than in Tianjin (China). In contrast, the mixing ratio of tetrachloroethene was lower than the average mixing ratio of tetrachloroethene in 45 Chinese cities, as well as Shenyang (China), Tianjin (China), Philadelphia (USA), and Marseille (France), but higher than in Changsha (China) and Bristol (UK). As controlled ODS in China, the mixing ratios of CFCs in this study, including trichlorofluoromethane, dichlorodifluoromethane, trichlorotrifluoroethane, and dichlorotetrafluoroethane, were slightly higher than in the other Chinese cities and selected international ones, but lower than in industrial cities, such as Shenyang. As the major substitute for dichlorodifluoromethane, the mixing ratio of chlorodifluoromethane in this study was much higher than in the other Chinese cities and selected international ones. It might be because the sampling time was in summer in this study, which could lead to active air-conditioning and refrigerating, and rapid residue discharge from foam products under high temperature [37]. The higher population density in Beijing could generate a greater demand for air conditioning, leading to larger chlorodifluoromethane emissions [37]. The mixing ratios of tetrachloromethane and bromomethane were comparable with their mixing ratios in other national and international cities.

Daily Variations
The daily averages of the total mass concentrations of the 23 VHHs over the study period had two peaks (Figure 4). The first peak occurred on 26 August, with a daily concentration of 35.36 µg/m 3 , while the second peak occurred on 30 August, with a daily concentration of 41.30 µg/m 3 . The occurrence of these peaks indicates that low wind speed, high temperature, high RH, and weak ultraviolet radiation might lead to the accumulation of VHHs in the ambient air ( Figure 5). After 1-2 September, the daily average concentration of VHH species declined markedly, indicating that high wind speed is conducive to VHHs diffusion ( Figure 5). Variations in the daily concentrations of the 23 VHH species are shown in Figure 6. It can be seen that the daily average concentrations of chloroalkanes varied in a similar manner to the total concentration of VHHs, while others species had different trends. ultraviolet radiation might lead to the accumulation of VHHs in the ambient air ( Figure 5). After 1-2 September, the daily average concentration of VHH species declined markedly, indicating that high wind speed is conducive to VHHs diffusion ( Figure 5). Variations in the daily concentrations of the 23 VHH species are shown in Figure 6. It can be seen that the daily average concentrations of chloroalkanes varied in a similar manner to the total concentration of VHHs, while others species had different trends.  ultraviolet radiation might lead to the accumulation of VHHs in the ambient air ( Figure 5). After 1-2 September, the daily average concentration of VHH species declined markedly, indicating that high wind speed is conducive to VHHs diffusion ( Figure 5). Variations in the daily concentrations of the 23 VHH species are shown in Figure 6. It can be seen that the daily average concentrations of chloroalkanes varied in a similar manner to the total concentration of VHHs, while others species had different trends.   The daily average concentrations of the four CFCs, including trichlorofluoromethane, dichlorodifluoromethane, dichlorotetrafluoroethane, and 1,1,2-trichlorotrifluoroethane, did not show obvious temporal variations (Figure 6a). Since these CFCs are controlled ODS under the agreement of the "Montreal Protocol on Substances that Deplete the Ozone Layer", China has completely banned consumption and production of these four CFCs since 1 January, 2010. The The daily average concentrations of the four CFCs, including trichlorofluoromethane, dichlorodifluoromethane, dichlorotetrafluoroethane, and 1,1,2-trichlorotrifluoroethane, did not show obvious temporal variations (Figure 6a). Since these CFCs are controlled ODS under the agreement of the "Montreal Protocol on Substances that Deplete the Ozone Layer", China has completely banned consumption and production of these four CFCs since 1 January, 2010. The detection of these CFCs in the study area indicates that more strict environmental management countermeasures need to be taken to prevent such pollution. Emission sources of these four CFCs appear to be stable in the study area given their small variations in daily average concentrations over the study period.
In contrast, the daily average concentrations of chlorodifluoromethane varied markedly (Figure 6b), which could reflect that leakage of refrigerants is a source of VHHs in the study area because chlorodifluoromethane is used in air conditioners, as well as industrial and commercial refrigerators in China.
The daily average concentrations of dichloromethane varied most among the 23 VHHs. The maximum concentration of dichloromethane occurred on 30 August, and the minimum concentration on 2 September, representing 1.45 and 0.28 times its daily average concentration, respectively ( Figure 6c). As can be seen from the daily variations of dichloromethane, its concentrations were relatively high and unstable, indicating that there are local emission sources close to the sampling site.
The daily average concentrations of tetrachloroethene changed most significantly among the chlorinated hydrocarbons ( Figure 6d); its highest and lowest values were 1.88 and 0.19 times its average value over the study period. Daily average concentrations of 1,4-dichlorobenzene varied most among the chlorinated benzenes (Figure 6e), with highest and lowest values being 2.19 and 0.28 times its average value over the study period. For bromohydrocarbons, the daily average concentrations of tribromomethane varied more than those of bromomethane (Figure 6f).

Diurnal Variations
The diurnal variations of the mass concentration of VHHs in the study area are shown in Figures 7 and 8. The total mass concentration of VHHs had a "V-Shaped" trend, being higher in the morning and at night, and lower at noon. Maxima of the total VHHs concentrations occurred within the period 8:30-9:00 a.m., while minima occurred during 1:30-2:00 p.m., and then increased again slowly (Figure 7). Aside from fluorodichloromethane, tetrachloroethene, 1,4-dichlorobenzene, 1,3-dichlorobenzene, and tribromomethane, the concentrations of most VHH species had similar diurnal variations with the total VHHs concentrations (Figure 8).
Atmosphere 2020, 11, x FOR PEER REVIEW 11 of 19 detection of these CFCs in the study area indicates that more strict environmental management countermeasures need to be taken to prevent such pollution. Emission sources of these four CFCs appear to be stable in the study area given their small variations in daily average concentrations over the study period. In contrast, the daily average concentrations of chlorodifluoromethane varied markedly ( Figure  6b), which could reflect that leakage of refrigerants is a source of VHHs in the study area because chlorodifluoromethane is used in air conditioners, as well as industrial and commercial refrigerators in China.
The daily average concentrations of dichloromethane varied most among the 23 VHHs. The maximum concentration of dichloromethane occurred on 30 August, and the minimum concentration on 2 September, representing 1.45 and 0.28 times its daily average concentration, respectively ( Figure  6c). As can be seen from the daily variations of dichloromethane, its concentrations were relatively high and unstable, indicating that there are local emission sources close to the sampling site.
The daily average concentrations of tetrachloroethene changed most significantly among the chlorinated hydrocarbons ( Figure 6d); its highest and lowest values were 1.88 and 0.19 times its average value over the study period. Daily average concentrations of 1,4-dichlorobenzene varied most among the chlorinated benzenes (Figure 6e), with highest and lowest values being 2.19 and 0.28 times its average value over the study period. For bromohydrocarbons, the daily average concentrations of tribromomethane varied more than those of bromomethane (Figure 6f).
The diurnal variations of VHH concentrations in the ambient air are influenced by many factors, such as emission source characteristics, meteorological conditions, and rates of local atmospheric photochemical reactions [43]. In the morning, the mass concentrations of VHHs were maintained at a high level because of the atmospheric stability, a low boundary-layer height, and anthropogenic emissions. Then, the mass concentrations of VHHs reached the lowest mass concentration between 1:30 and 2:00 p.m. as the temperature, the photochemical reaction, and wind speed increased ( Figure 5), which accelerates the consumption and diffusion of VHHs. Afterwards, as the temperature, atmospheric boundary layer, and wind speed decreased, the mass concentrations of VHHs increased gradually. Figure 9 shows the main sources of volatile halogenated hydrocarbons in ambient air in the study area and the contributions of individual compounds. Seven factors were finally identified using PMF 5.0.

Source Apportionment
Atmosphere 2020, 11, x FOR PEER REVIEW 13 of 19 a high level because of the atmospheric stability, a low boundary-layer height, and anthropogenic emissions. Then, the mass concentrations of VHHs reached the lowest mass concentration between 1:30 and 2:00 p.m. as the temperature, the photochemical reaction, and wind speed increased ( Figure  5), which accelerates the consumption and diffusion of VHHs. Afterwards, as the temperature, atmospheric boundary layer, and wind speed decreased, the mass concentrations of VHHs increased gradually. Figure 9 shows the main sources of volatile halogenated hydrocarbons in ambient air in the study area and the contributions of individual compounds. Seven factors were finally identified using PMF 5.0.  1 was characterized by high loadings of tetrachloroethene (58.89%). Tetrachloroethene is mainly used as industrial cleaning solvents in the electronic industry [25,44]. So, we concluded that factor 1 described emissions from solvents usage in the electronic industry.

Source Apportionment
Factor 2 was characterized by high loadings of chlorodifluoromethane (53.40%). Chlorodifluoromethane is a major refrigerant replacement and widely used in commercial refrigeration and refrigeration transport [43]. So, we concluded that factor 2 described emissions from refrigerants.
Factor 3 was characterized by high loadings of 1,4-dichlorobenzene (71.63%). 1,4dichlorobenzene is used mainly as a fumigant for the control of moths, molds, and mildews, and as  1 was characterized by high loadings of tetrachloroethene (58.89%). Tetrachloroethene is mainly used as industrial cleaning solvents in the electronic industry [25,44]. So, we concluded that factor 1 described emissions from solvents usage in the electronic industry.
Factor 2 was characterized by high loadings of chlorodifluoromethane (53.40%). Chlorodifluoromethane is a major refrigerant replacement and widely used in commercial refrigeration and refrigeration transport [43]. So, we concluded that factor 2 described emissions from refrigerants. Factor 3 was characterized by high loadings of 1,4-dichlorobenzene (71.63%). 1,4-dichlorobenzene is used mainly as a fumigant for the control of moths, molds, and mildews, and as a space deodorant for toilets and refuse containers [35]. So, we concluded that factor 3 described emissions from fumigant usage. Factor 4 was characterized by high loadings of chlorobenzene, monochloroethane, and trichloromethane. Chlorobenzene (93.17%) is used as a solvent for some pesticide formulations, to degrease automobile parts, and as a chemical intermediate to make several other chemicals [45]. Monochloroethane (63.80%) is used in the production of cellulose, dyes, medicinal drugs, and other commercial products, and as a solvent and refrigerant [45]. Trichloromethane (58.49%) is formerly used as an inhaled anesthetic during surgery, a solvent, and in the production of the refrigerant freon [36]. So, we concluded that factor 4 described emissions from solvents usage and industrial processes.
Factor 5 was characterized by high loadings of CFCs and tetrachloromethane. CFCs are mainly used as refrigerants in cooling appliances and air conditioning, whereas tetrachloromethane is mainly used as feedstock for CFC production [44]. Although CFCs and tetrachloromethane were supposed to be entirely phased-out in 2010 in China, these species produced before 2010 are still stored in equipment and can leak into the environment [25]. So, we concluded that factor 5 described emissions from leakage of CFCs banks.
Factor 6 was characterized by high loadings of dichloromethane, dichloroethane, and dichloropropane. Chloroalkanes are widely used as solvents and in the production of other chemicals. Dichloromethane is well known as a feedstock for foam plastic products, metal cleaning, and other solvent uses [44]. Dichloroethane and dichloropropane are also widely used in industry as solvents. So, we concluded that factor 6 described emissions from solvents usage and industrial processes.
Factor 7 was characterized by high loadings of trichloroethene. Trichloroethene is widely used in electronic and textile industries as a solvent and degreaser [44]. So, we concluded that factor 7 described emissions from solvents usage in the electronic and textile industries.
Average contributions of each factor to the total volatile halogenated hydrocarbons are shown in Figure 10. Since factor 1, factor 4, factor 6, and factor 7 are related to solvents usage, their contributions are summed as the contribution of solvents usage and industrial processes to volatile halogenated hydrocarbons. It can be seen that solvents usage and industrial processes (51.73%) is a major contributor in the study area, followed by leakage of CFCs banks (21.92%). Refrigerants and fumigant usage contributed 17.31% and 9.04%, respectively.
Atmosphere 2020, 11, x FOR PEER REVIEW 14 of 19 a space deodorant for toilets and refuse containers [35]. So, we concluded that factor 3 described emissions from fumigant usage. Factor 4 was characterized by high loadings of chlorobenzene, monochloroethane, and trichloromethane. Chlorobenzene (93.17%) is used as a solvent for some pesticide formulations, to degrease automobile parts, and as a chemical intermediate to make several other chemicals [45]. Monochloroethane (63.80%) is used in the production of cellulose, dyes, medicinal drugs, and other commercial products, and as a solvent and refrigerant [45]. Trichloromethane (58.49%) is formerly used as an inhaled anesthetic during surgery, a solvent, and in the production of the refrigerant freon [36]. So, we concluded that factor 4 described emissions from solvents usage and industrial processes.
Factor 5 was characterized by high loadings of CFCs and tetrachloromethane. CFCs are mainly used as refrigerants in cooling appliances and air conditioning, whereas tetrachloromethane is mainly used as feedstock for CFC production [44]. Although CFCs and tetrachloromethane were supposed to be entirely phased-out in 2010 in China, these species produced before 2010 are still stored in equipment and can leak into the environment [25]. So, we concluded that factor 5 described emissions from leakage of CFCs banks.
Factor 6 was characterized by high loadings of dichloromethane, dichloroethane, and dichloropropane. Chloroalkanes are widely used as solvents and in the production of other chemicals. Dichloromethane is well known as a feedstock for foam plastic products, metal cleaning, and other solvent uses [44]. Dichloroethane and dichloropropane are also widely used in industry as solvents. So, we concluded that factor 6 described emissions from solvents usage and industrial processes.
Factor 7 was characterized by high loadings of trichloroethene. Trichloroethene is widely used in electronic and textile industries as a solvent and degreaser [44]. So, we concluded that factor 7 described emissions from solvents usage in the electronic and textile industries.
Average contributions of each factor to the total volatile halogenated hydrocarbons are shown in Figure 10. Since factor 1, factor 4, factor 6, and factor 7 are related to solvents usage, their contributions are summed as the contribution of solvents usage and industrial processes to volatile halogenated hydrocarbons. It can be seen that solvents usage and industrial processes (51.73%) is a major contributor in the study area, followed by leakage of CFCs banks (21.92%). Refrigerants and fumigant usage contributed 17.31% and 9.04%, respectively.

Health Risk Assessment
According to the International Agency for Research on Cancer (IARC), chemicals are divided into five categories (1, 2A, 2B, 3, and 4) based on their carcinogenicity [35,46]. Categories 1, 2A, and 2B are acknowledged to be carcinogenic to humans. Hence, the carcinogenic risk of seven VHH species belonging to these categories was assessed. The non-carcinogenic risk for adults in the study area was assessed for 11 VHH species whose RfC values could be obtained from the Integrated Risk Information System.

Health Risk Assessment
According to the International Agency for Research on Cancer (IARC), chemicals are divided into five categories (1, 2A, 2B, 3, and 4) based on their carcinogenicity [35,46]. Categories 1, 2A, and 2B are acknowledged to be carcinogenic to humans. Hence, the carcinogenic risk of seven VHH species belonging to these categories was assessed. The non-carcinogenic risk for adults in the study area was assessed for 11 VHH species whose RfC values could be obtained from the Integrated Risk Information System. The parameters used for the carcinogenic risk assessment for these VHHs as well as for the non-carcinogenic risk assessment of the dominant 11 VHH species are shown in Tables 3 and 4, respectively. According to the US EPA document "EPA-540-R-070-002", the acceptable value of Risk for a specific pollutant for an ordinary adult is 1 × 10 −6 [47]. The Risk values for the seven carcinogenic VHHs varied from 6.43 × 10 −8 to 8.04 × 10 −5 , with the Risk of trichloromethane, tetrachloromethane, 1,2-dichloroethane, and 1,4-dichlorobenzene being more than 1 × 10 −6 . 1,2-Dichloroethane had the highest Risk, with a Risk of 80.4 times 1 × 10 −6 , followed by trichloromethane with a Risk of 18.6 times 1 × 10 −6 , and 1,4-dichlorobenzene and tetrachloromethane with Risks of five-six times 1 × 10 −6 . Thus, trichloromethane, tetrachloromethane, 1,2-dichloroethane, and 1,4-dichlorobenzene in the ambient air posed relatively high carcinogenic risks to the long-term exposed populations in the study area.  According to the US EPA, if the value of HQ for a specific pollutant is lower than 1, the pollutant has no obvious non-carcinogenic risk to humans. The HQ of the 11 VHH species varied from 1.54 × 10 −5 to 5.94 × 10 −1 , while HI was 0.728. The result indicates that these VHHs had no obvious non-carcinogenic risk to the general population in the study area. Among the 11 VHH species, 1,2-dichloropropane had the highest HQ value (5.94 × 10 −1 ), followed by trichloroethene (9.00 × 10 −2 ).
Huang et al. [48] determined values of Risk for trichloromethane and 1,4-dichlorobenzene in a primary school to be 6.5 × 10 −7 and 1.6 × 10 −7 , respectively. Clearly, the values of Risk for trichloromethane and 1,4-dichlorobenzene in this study were 28 and 36 times higher, posing a far greater risk to residents in the study area. In contrast, the values of Risk for trichloroethene and tetrachloroethene for this study were comparable to those in ambient air in Nanjing [49], suggesting that these species do not pose a threat to local residents in Beijing or Nanjing.
Comparisons of non-carcinogenic risk assessments for nine volatile halogenated hydrocarbons in Liaoning Province [12] and this study are shown in Table 5. Clearly, the HQ values of monochloromethane, tetrachloromethane, and 1,1-dichloroethene in this study were lower than in some cities of Liaoning Province, whereas the HQ values of 1,2-dichloropropane, tetrachloroethene, and bromomethane were mostly higher. No apparent difference between the two studies was observed for dichloromethane, monochloroethane, and trichloroethene. Health risk assessments of VHHs in the ambient air in urban areas of China suggest that non-carcinogenic risks are at a relatively low level, but Risks for trichloromethane, tetrachloromethane, and 1,2-dichloroethane are relatively high. Given that there are few studies on the carcinogenic risk assessment of VHHs in China to date, it is crucial to carry out more health risk assessments of toxic VHHs in the near future.

Conclusions
In this study, VOC samples were collected with SUMMA canisters in a typical urban area of Beijing. VHH species were determined using a cryogenic cold trap pre-concentration prior to GC-MS/FID analysis. Ambient levels, temporal variation, and sources of VHHs were determined for the study area, and health risks of selected toxic and hazardous VHH species were assessed. The results showed that the mean and the range of daily total mass concentrations of the 23 VHHs were 30.53 and 13.45-76.33 µg/m 3 , respectively. Among the 23 VHHs, 7 VHHs were controlled ozone-depleting substances in China. The average and range of the seven VHHs were 12.95 and 7.69-36.17 µg/m 3 , respectively, accounting for 42.43% of the total. Dichloroethane had the highest concentrations, followed by chlorodifluoromethane, making chloroalkanes the main VHH constituent in the study area. Compared with other national and international cities, the concentrations of the selected 11 VHHs in the study area, including 7 controlled ODS in China, were relatively higher. Diurnal variation in the total mass concentrations of the 23 VHHs was marked by higher levels in the morning and evening, and lower levels at noon. Solvents usage and industrial processes, leakage of CFCs banks, refrigerants, and fumigant usage were the main sources of VHHs in the study area. Health risk assessment showed that four VHH species, including trichloromethane, tetrachloromethane, 1,2-dichloroethane, and 1,4-dichlorobenzene, posed potential carcinogenic risks to the exposed populations in the study area. None of the selected 11 VHH species posed appreciable non-carcinogenic risks to exposed populations. Therefore, it is imperative to take effective countermeasures to control VHH emissions in Beijing. For example, it is recommended that the construction and improvement of VHH emission inventories for key sources should be promoted nationwide, and not only the total amount but also the toxic VHHs should be included in the emission inventories, in order to lay the foundation for accurate source apportionment of ambient VHHs, as well as for the identification and refined control of key sources of VHHs.