Air-borne particulate matter plays an important role in climate and atmospheric chemistry, as it contains various harmful species that may affect human health (Talifu et al. 2015). In particular, the particulate matter with aerodynamic diameter less than 2.5 µm (PM2.5), attracts public attention around the world, because it is deposited throughout the human respiratory tract, causing heart and lung diseases (Pui et al. 2014). Numerous publications have revealed that polycyclic aromatic hydrocarbons (PAHs) can be highly enriched in PM2.5 (Masih et al. 2012), due to their low solubility, low reactivity and moderate vapor pressure (Chang et al. 2006). Particulate-phase PAHs are considered to be a significant hazard to human health (Hong et al. 2007).

PAHs are compounds with two or more aromatic rings, which are mainly derived from the incomplete combustion of carbon enriched materials (Ma et al. 2010a). Although some natural sources (volcanoes and forest fires) could contribute to the PAH burden, anthropogenic sources, such as combustion of fossil fuels, residential heating, and coke production are by far the major sources of PAHs (Li et al. 2014).

Hefei is the capital city of Anhui province, China. In recent decades, Hefei has faced accelerated development of economy, industrial facilities, urban construction and public transportation. In this study, PM2.5 samples were collected systematically from Hefei city and their PAHs were measured. The main aims of this study are to: (1) investigate the spatial and seasonal variations of PAHs, (2) understand the controls of meteorological parameters on the enrichment and distribution of PAHs, and (3) identify the possible sources of PAHs. We aimed to provide a case to understand the organic pollution in particulate matter at developing cities.

Materials and Methods

Hefei City (E117°27′, N 31°86′) is the economic and cultural center of Anhui province. It is located in the middle of the province, covering an urban area of 1, 100 km2. It had 4.86 million residents and more than 1.2 million vehicles in 2015 (Deng et al. 2016). Hefei has a subtropical humid monsoon climate, with four seasons defined as follows: spring (March–April), summer (May–August), fall (September–November) and winter (December–February). PM2.5 was sampled simultaneously at two locations of Hefei: Site A (MS) located at a middle school, northeast of First Ring Rd, Yaohai district, adjacent to the residential area; Site B (ME) located at west section of Fanhua Avenue, Shushan district, adjacent to a road with a very high traffic volume (Fig. 1).

Fig. 1
figure 1

Sketch map showing the PM2.5 sampling locations (ME, MS), Hefei, Anhui, China

Full size image

A middle volume air sampler (LaoYing 2030, Laoshan Electronic Instrument, Qingdao, CN) was used to collect the PM2.5 on a quartz fiber filter (Whatman EPM2000, UK, diameter = 90 mm), at a flow rate of 100–120 L min−1 for approximately 20 h. In total, samples were collected at the time intervals of 20–28 May 2014 in summer, and 16–23 January 2015 in winter. Before sampling, all filters were baked in a muffle furnace at 500°C for 4 h to eliminate the background organic matter. Filters were weighed on a microbalance to ±0.001 mg (Sartorius MSA 3.6P-000-DM, Gottingen, DE) on a clean bench before and after sampling to calculate the mass of collected PM2.5 samples. Upon collection, the filters were stored in clean containers kept at 4°C.

Each filter was cut into small pieces (1cm × 1 cm) using stainless steel scissors, and then was extracted for PAHs by dichloromethane (DCM) and acetone (chromatographically pure grade) in an accelerated solvent extraction apparatus (Dionex-ASE 350, Sunnyvale, CA, USA). The extracts were concentrated to 1 mL by automatic parallel concentrator (MultiVap-8, LabTech, Hopkinton, MA, USA) before analysis with a gas chromatograph (GC, Agilent Technologies 7890, Wilmington, DE, USA) equipped with a gas chromatograph Agilent HP-5MS, non-polar 30 m × 0.25 mm capillary column in tandem with a mass spectrometer (MS, 5975 N).

The helium carrier gas was set at a flow rate of 1.0 mL min−1. Each sample run was programmed as follows: starting temperature of 100°C for 2 min, ramping to 200°C at a rate of 20°C min−1, ramping to 300°C at a rate of 4°C min−1 and keeping for 10 min.

Sixteen priority PAH compounds specified by the Environmental Protection Agency, USA (USEPA), were analyzed, including: NAP—naphthalene; ANY—acenaphthylene; ANA—acenaphthene; FL—fluorene; PHE—phenanthrene; An—anthracen; FLU—Fluoranthene; PYR—pyrene; BaA—benz[a]anthracene; CHR—chrysene; BbF—benzo[b]fluoranthene; BkF—benzo[k]fluoranthene; BaP—benzo[a]pyrene; IcdP—indeno[1,2,3-cd]pyrene; DBA—dibenzo[a,h]anthracene; and BghiP—benzo[ghi]perylene.

Field blank, method blanks (solvent) and spiked blanks (standards spiked into solvent) were analyzed by the same procedure. There were no significant background interferences. In addition, surrogate standards were added to all the samples to monitor procedural performance and matrix effects consisting of phenanthrene-d10, chrysene-d12, and perylene-d12 (Aldrich Chemical, Gillingham, Dorset, UK). The mean recoveries (%) of the three surrogates in field samples were within the acceptable range (75%–128%). The calibration curves were linearly fitted with correlation coefficients (R2) from 0.996 to 1for all of the PAHs when conducting the external calibration. PAH concentrations were corrected for recovery efficiency during extraction.

Pearson correlation and principal component analysis (PCA) and correlation analysis were performed using SPSS version 17.0 software packages (IBM, Somers, NY, USA). Meteorological data were obtained from the Weather Bureau of Anhui Province, and are listed in Table 1.

Table 1 Meteorological parameters during PM2.5 sampling periods
Full size table

Results and Discussion

At ME and MS sampling sites, the daily concentrations of PM2.5 reached 121.75 ± 26.96 and 104 ± 21.75 µg m−3 in winter, whereas in summer, they were 100.44 ± 44.37 and 64.89 ± 30.78 µg m−3, respectively. The average daily value of PM2.5 concentration was 96.88 ± 32.46 µg m−3, higher than the National Ambient Air Quality standard for PM2.5 (NAAQS-2012 24-h) of 35 µg m−3 (Chen et al. 2015; You 2014). It was also higher than the China Ambient Air Quality Standard (GB3095-2012 24-h grade II) of 75 µg m−3 (Brunner 1985; China 2012).

Mean concentrations of PM2.5-bound individual PAHs are shown in Table 2. During sampling times, the total mean ∑PAH concentrations ranged from 4.92 to 71.00 ng m−3 in Hefei, with a mean of 21.34 ng m−3. This mean is lower than those in other large Chinese cities such as Beijing (112.7 ng m−3 in 2003), Qingdao (87.5 ng m−3) and Guangzhou (117.4 ng m−3), but is obviously higher than those in the developed countries/regions such as London (17.23 ng m−3), Hong Kong (7.42 ng m−3), and Los Angeles (2.46 ng m−3)(Baek et al. 1992; Fraser et al. 1998; He et al. 2006; Zheng and Fang 2000). Benzo(a)pyrene has been categorized as a carcinogenic PAH, and is often used as an indicator of human PAH exposure. In this study, the mean BaP concentration was 0.46 to 2.31 ng m−3 (mean = 1.15 ng m−3), which was lower than the maximum allowable annual average BaP concentration of 2.5 ng m−3, according to the China Ambient Air Quality Standard (GB3095-2012) (China 2012). Our BaP concentration was comparable to the published results for other Chinese cities, and was lower than value for Beijing (17.8 ng m−3), Guangzhou (2.98 ng m−3) and Harbin (2.5 ng m−3) (Liu et al. 2015; Ma et al. 2010b; Wang et al. 2008).

Table 2 Mean concentrations (ng m− 3) of individual PAHs in PM2.5
Full size table

The spatial and seasonal variations in PAH concentrations for the two sampling sites are shown in Table 2. The overall mean ∑PAHs in the PM2.5 at ME site (27.23 ng m−3) was higher than that at MS site (18.20 ng m−3). This was largely due to the 2.2-fold higher winter values at site ME. This might relate to the locations of the sampling sites. The ME site is located in the economic development zone, the major industrial area of Hefei. The sampling site was adjacent to a road with a high volume of traffic. Miguel and Pereira (1989) noted that PAHs were the major components of vehicle exhausts, especially gasoline- and diesel-powered vehicles. On the other hand, the MS site was located at a middle school, where there was a mixture of residential, commercial and industrial areas, with less traffic volume and industrial activity than at site ME.

The seasonal variations of PAHs in PM2.5 of Hefei are presented in Table 2. At both sites combined, the mean concentration for ∑PAHs was higher in winter (31.38 ng m−3) than in summer (14.05 ng m−3). Similar seasonal contrasts in PM2.5-bound PAHs have been reported in other areas (Guo et al. 2003; Wang et al. 2015). For example, Guo et al. (2003) found that the ratio of PAHs in winter to summer was in the range of 7.5–8.6. We speculate that this variation might relate to meteorological conditions and human activities. In summer, the high temperature and intense sunlight could enhance the degradation of PAHs in PM2.5, while in winter, the increased domestic heating and consumption of fuels could release PAHs to the atmosphere (He et al. 2006; Tang et al. 2005).

Correlations between meteorological parameters and the concentrations of individual PAHs were calculated to investigate possible relationships (Table 3), in order to investigate the influence of meteorological conditions on the distribution of PAHs. We found a negative correlation between the concentrations of PAHs and ambient temperature. This was largely attributed to the influence of increased emissions during the winter season, with particulate-bound PAH concentrations increasing due to the consumption of fuels associated with domestic heating (Tham et al. 2008). This also can be ascribed to the degradation rate of PAHs commonly being faster under higher temperatures and stronger solar radiation. In this study, the concentrations of PAHs have no significance correlations with the wind speed and relative humidity.

Table 3 Correlations between concentrations of PAHs (ng m−3) and meteorological parameters
Full size table

The PAHs of PM2.5 could derive from various emission sources. The possible sources of PAHs in air can be identified by diagnostic concentrations and PCA (principal component analysis) of some marker PAH compounds.

The concentrations of a group of PAHs, or specific PAHs compounds, have been used to identify the emission sources (Yuan et al. 2015), such as BaA and CHR, An, FLU, PYR, for coal combustion; An, BaP and BghiP for coke production(Cai et al. 2014); FLU and PYR for combustion of wood (Tekasakul et al. 2008); and FLU for oil burning (Khalili et al. 1995). However, in different types of emission source profiles, some overlaps in similarity have been noted for these maker PAHs. (Hu et al. 2012).

Thus, we have used diagnostic ratios, such as (FLU/ [FLU + PYR], BaA/[BaA + CHR], An/[An + PHE], BaP/BghiP and IcdP/[IcdP + BghiP]) to appoint the sources of PAHs in our case study (Table 4 ). Different values of each diagnostic ratio correspond to characterized emission sources. These diagnostic ratios are all indicators of human activities, with coal combustion and motor vehicle emissions being the primary sources of PAHs in PM2.5 of Hefei.

Table 4 Diagnostic ratios of PAHs in PM2.5 of Hefei
Full size table

Principal component analysis was used to aid in the interpretation of PAH emission sources by reducing the large number of variables to a smaller number of factors. Each of the factor from the PCA is associated with source characterization by its most representative PAH compound(s) (Saxena et al. 2014). Source groupings were determined by PCA with varimax rotation, and the retention of principal components having eigen values higher than 1 of the complete data set of PAHs concentrations. In addition, loadings determined the most representative PAH compounds in each factor, with values higher than 0.5 being selected (Hu et al. 2012).

Three factors accounted for 84.03% of the total variance of the data (Table 5), with factor 1 explaining 63.26% of the variance. This factor showed high loading for ANY, FL, PYR, PHE, CHR, FLU, An and IcdP. Factor 2 explained 11.82% of the variance, and showed a high loading for BghiP. As summarized by Harrison et al. (1996), FLU, PYR and CHR are considered to be the tracers of coal combustion; and IcdP, PHE and BghiP to be the tracers of vehicle exhaust(Harrison et al. 1996). The lower molecular weight PAHs, such as FL and An, might be formed in processes during the pyrolysis of fossil fuels (Dachs et al. 2002; Jung et al. 2015). Therefore, this suggested that Factor 1 was representative of pyrogenic sources, including coal burning (power plants and space heating) and vehicle emissions, and Factor 2 was representative of vehicle exhaust. The appointed sources of PCA analysis are consistent with diagnostic ratios. Factor 3 only explained 8.95% the variance. It showed a high loading for DbA, a tracer for coke oven sources (Simcik et al. 1999).

Table 5 Factorial weights matrix for the source analysis of PAHs in Hefei
Full size table

In conclusion, concentrations of 16 PAHs were measured in the PM2.5 of Hefei during the periods of 20–28 May, 2014, and 16–20 January, 2015. The concentration of PAHs ranged from 12.49 to 23.75 ng m−3 at site MS and from 4.92 to 71.00 ng m−3 at site ME. A strong seasonal trend in PAH concentration was observed during the sampling period, with a higher level in winter than in summer. Meteorological conditions influenced the PAH concentrations, with air temperature and PAH concentration being inversely related. According to the analysis of emission source of PAHs based on diagnostic ratios and PCA, it was concluded that combustion of coal and vehicle exhaust were the predominant sources of PAHs in PM2.5 of Hefei.