Chemical composition, structural properties, and

Download PDF
0 downloads 0 Views 1MB Size Report
Comment
apportionment of organic macromolecules in atmospheric. PM10 in a coastal .... our previous study as organic macromolecule fractions (Zhao et al. 2011; Chen ...
Missing:
Environ Sci Pollut Res DOI 10.1007/s11356-016-8314-5

RESEARCH ARTICLE

Chemical composition, structural properties, and source apportionment of organic macromolecules in atmospheric PM10 in a coastal city of Southeast China Yanting Chen 1,2 & Wenjiao Du 1,2 & Jinsheng Chen 1,2 & Youwei Hong 1,2 & Jinping Zhao 3 & Lingling Xu 1,2 & Hang Xiao 1,2

Received: 21 July 2016 / Accepted: 21 December 2016 # Springer-Verlag Berlin Heidelberg 2017

Abstract Particulate matter (PM10) associated with the fractions of organic macromolecules, including humic acid (HA), kerogen + black carbon (KB), and black carbon (BC), was determined during summer and winter at urban and suburban sites in a coastal city of southeast China. The organic macromolecules were characterized by elemental analysis (EA), scanning electron microscopy (SEM), and Fourier transform infrared spectroscopy (FTIR), and their sources were identified by using stable carbon/nitrogen isotope (δ13C/δ15N) and the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) Model. The results showed that HA, kerogen (K), and BC accounted for the range of 3.89 to 4.55 % in PM10, while they were the dominant fractions of total organic carbon (TOC), ranging from 64.70 to 84.99 %. SEM analysis indicated that BC particles were porous/nonporous and consisted of spherical and non-spherical (i.e., cylindrical and elongate) structures. The FTIR spectra of HA, KB, and BC exhibited similar functional groups, but the difference of various sites and seasons was observed. HA in PM10 contained a higher fraction of aliphatic structures, such as long-chain fatty Responsible editor: Constantini Samara * Jinsheng Chen [email protected] * Youwei Hong [email protected]

1

Center for Excellence in Regional Atmospheric Environment, Institute of Urban Environment, Chinese Academy of Sciences, 1799 Jimei Road, Xiamen 361021, China

2

Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China

3

Guangdong Environmental Monitoring Center, Guangzhou 510308, China

and carbohydrates with a carboxylic extremity. The C/N ratio, SEM, and δ13C/δ15N values provided reliable indicators of the sources of HA, K, and BC in PM10. The results suggested that HA and K majorly originated from terrestrial plants, and BC came from the mixture of combustion of terrestrial plants, fossil fuel, and charcoal. The air masses in winter originated from Mongolia (4 %), the northern area of China (48 %), and northern adjacent cities (48 %), suggesting the influence of anthropogenic sources through long-range transport, while the air masses for the summer period came from South China Sea (34 %) and Western Pacific Sea (66 %), representing clean marine air masses with low concentrations of organic macromolecules. Keywords Atmospheric PM10 . Organic macromolecules . Chemical composition . Source apportionment . Coastal city

Introduction The organic macromolecules were involved in the transportation of toxic organic and inorganic substances and nutrient cycling in the environment, due to their large surface area and strong structural binding sites. Aerosol particles containing organic macromolecules would readily bind organic pollutants, such as polycyclic aromatic hydrocarbons (PAHs), hexachlorocyclohexane isomers (HCHs), polychlorinated biphenyls (PCBs), and toxic metals (Xiao et al. 2004; Chefetz and Xing 2009; Janssen et al. 2011; Vargas et al. 2016). As a result of these properties, the fractions of organic macromolecules bound to PM10 might play an important role in climate change and further threaten human health through inhalation exposure. Thereby, it is necessary to investigate chemical composition, structural characterization, and sources of organic macromolecules in PM10. However, current studies for macromolecules

Environ Sci Pollut Res

organic matter predominantly focused on soils, volcanic ashes, and sediments through a variety of analytical methods (Song et al. 2002; Xiao et al. 2004; Zhang et al. 2011). Organic material in coarse particulate matter (PM10) is of increasing interest as it plays an important role in understanding atmospheric chemistry and environmental effects. Many solvent-solubilized individual organics, such as organic acids, amino acids, organic nitrates, organic sulfates, and polysaccharides, in atmospheric particulates have been widely characterized, quantified, and identified (Karthikeyan and Balasubramanian 2006; Barbaro et al. 2011; Staudt et al. 2014; Lopes et al. 2015; Siudek et al. 2015). However, a few studies focused on chemical compositions and structural properties of organic macromolecules, such as fluvic acid (FA), humic acids (HA), kerogen (K), and black carbon (BC) in atmospheric environment (Zappoli et al. 1999; Salma et al. 2010; Zhao et al. 2011; Song et al. 2012; Peng et al. 2013). The latter three fractions were often described in our previous study as organic macromolecule fractions (Zhao et al. 2011; Chen et al. 2013; Peng et al. 2013). With complex biomacromolecular structure, humic acid was widely presented in soil and aqueous systems (Song et al. 2002; Cao et al. 2006; Zhang et al. 2011). They were mostly derived from decomposition products of plant and animal tissues. Kerogen was defined as particulate organic matter in sedimentary rock and was the backbone structure of coal (Vandenbroucke and Largeau 2007; Bushnev and Burdel’naya 2009). In soil and sediment, kerogen was the degradation products of algae, phytoplankton, and terrestrial plants (Vandenbroucke 1980; Song et al. 2002). In air particulate matter, HA and K majorly came from combustion sources, soil dust, vegetative detritus, and biomass burning (Lin et al. 2010a; Lin et al. 2010b; Peng et al. 2013). Simultaneously, they can be partially formed from photochemical oxidation, deoxidization, and polymerization of NOx and volatile organic compounds (VOCs) (Gelencsér et al. 2003; Salma et al. 2007; Zhao et al. 2011). In general, BC in atmospheric aerosol was thought to be the product of incomplete combustion of fossil fuel and biomass burning (Masiello et al. 2002; Masiello 2004; Koch et al. 2009; Lyamani et al. 2011). The fractions of HA, KB, and BC in this study were isolated by chemical treatments with sodium hydroxide, hydrochloric acid/hydrofluoric acid, and potassium dichromate/ sulfuric acid, which was widely used in soils and sediments and also occasionally in atmospheric samples (Zhao et al. 2011; Peng et al. 2013). It should be kept in mind that the same name of organic matter might refer to different materials in atmospheric chemistry and geochemistry. For instance, the term BC used in geochemistry referred to carbonaceous materials produced by incomplete combustion of fossil fuels and vegetation, while BC was chemically defined as a strongly light-absorbing carbon determined by the optical instrument

(e.g., the aethalometer). In addition, HA in PM10 actually referred to humic acid and FA that were both soluble in base solutions, which was different from humic substances in soils and sediments. Humic-like substances (HULIS) were a class of organic compounds identified in atmospheric samples that bear some similarities to humic substances (HS) common in soils and aqueous environments. Therefore, the definitions of organic macromolecule fractions in PM10 were associated to the geochemistry, rather than chemistry. Due to the complex nature of atmospheric environment as well as the limitations of extraction methods, a few studies focused on the fractions of organic macromolecules in the atmosphere. Chemical compositions, structure and source apportionment of BC, and HULIS have been investigated in dust (Zhao et al. 2011; Zhao et al. 2012), total suspended particulate (TSP) (Zhao et al. 2009; Song et al. 2012; Peng et al. 2013), and PM2.5 (Lin et al. 2010a; Lin et al. 2010b; Claeys et al. 2012; Fan et al. 2012; Fan et al. 2013). However, pollution characteristics of HA, KB, and BC in atmospheric PM10 remain unclear. In this study, elemental analyzer (EA), Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and stable carbon/nitrogen isotope (δ13C/δ15N) were selected to study chemical composition, structure, and source of three organic macromolecules in PM10 in a coastal city in China. Moreover, backward trajectories (Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) Model) were performed to characterize the long-range transport sources through analysis of the air mass during the sampling period. This study would seek to provide more details about chemical features and possible sources of organic macromolecules in PM10. The pollution characteristics in the urban and suburban sites during the summer and winter months were also discussed.

Materials and methods Sample collection PM10 sampling was simultaneously carried out at two different functional areas during the summer (June, July, and August, in 2010) and winter (December in 2010 and January and February in 2011) months in Xiamen, China. The Siming (SM, 118.15 E, 24.48 N) represented an urban site, which was located at the rooftop (20 m height) of a teaching building of primary school in Siming District and close to commercial, educational, and residential area. Potential pollution sources include vehicular emission, cement and construction dust, natural emission (i.e., bare soil, trees, and road dust), and cooking emissions. In addition, the suburban site of Jimei (JM, 118.06 E, 24.61 N) was chosen at the Institute of Urban Environment, Chinese Academy of Sciences in Jimei District, a region with rapid urbanization. Samples were

Environ Sci Pollut Res

collected at a height of 5 m on the rooftop of one building, surrounded by a highway, construction land, and Xinglin Bay. The suburban site had a relative lower population and traffic density compared with the urban site. PM10 samples were collected by glass fiber filters (GFFs) through high-volume samplers (Thermo GUV-15H-1, USA) at flow rates of about 1130 L min−1 for 24 h. The sampling period lasted at least 2 months each season to sampling about 5 g weight of PM10, as each 24-h continuous sampling could only collect about 50–200 mg of PM10. In order to obtain sufficient mass of these organic macromolecules, several filters were combined to carry out the extraction of HA, K, and BC. For sampling, GFFs were kept in a furnace at 450 °C for 5 h to remove organic materials or impurities. Then, they were put into a desiccator with constant temperature (25 ± 1 °C) and humidity (52 ± 1 %) for 24 h and stored in baked aluminum foil within sealed polyethylene plastic bags. After sampling, the amount of PM10 was calculated gravimetrically. The PM10loaded filters were stored at about −20 °C until extraction. Extraction of organic macromolecules in PM10 The extraction protocol of three organic macromolecules (including HA, KB, and BC) was based on our previous method (Zhao et al. 2011; Chen et al. 2013). Briefly, the procedures include three major steps: (1) base extraction. Numerous filters for each site and season were combined to get particulate quantities of about 5.0 g. GFFs were cut into small pieces and were immersed into 0.1 mol·L−1 NaOH solution under nitrogen gas flow to extract humic substances (i.e., humic acid and fluvic acid) by ultrasonic bath (Model KQ300DE, China). For HA precipitation, the sample was centrifuged, and the supernatant was acidified to pH = 1.0–2.0 with 6 mol L−1 HCl. Then, the HA fraction was dialyzed, followed by being dried under vacuum. HA extraction with NaOH was repeated for several times to achieve maximum recovery. (2) Acid demineralization. The base-extracted residues of PM10 samples were sequentially treated with HCl (6 mol L−1), HCl/HF mixture at a 1 (6 mol L−1):2 (22 mol L−1) volumetric ratio, and HCl (6 mol L−1) to remove carbonates, silicate mineral, and minerals (i.e., the fluorite formed during the demineralization), respectively. During acid demineralization, ultrasonic process with bath temperature of 65 °C was performed to get the most optimal rate of demineralization. The supernatant was removed by centrifugation at 4500 rev./min for 30 min. Finally, the KB residue was washed again with Milli-Q water until pH 7 was reached and dried in an oven at 60 °C. (3) Dichromate oxidation. Another experiment was carried out for isolation of BC. The KB fractions were placed in a bottle and then treated with the mixture of dichromate and sulfuric acid (0.1 mol L−1 K2Cr2O7 + 2 mol L−1 H2SO4) more than

three times to ensure complete oxidation of the kerogen and other organic matters. The bottle containing the KB fractions was placed in a water bath at 55 ± 1 °C for 60 h. That is the reason why ultrasonic bath (60 ± 1 °C) was applied in conjunction with the solvent/acid wash. After the oxidation step, the bottle was centrifuged at 4500 rev./ min for 30 min. After removal of the supernatant, the residues were rinsed with ultrapure water and dried to obtain BC fractions. Note, KB was a mixture fraction of K and BC, and BC was isolated from KB. Chemical and meteorological data analysis Analysis methods were required notable advantages in sample amount, such as they only need submilligram to milligram samples and they were capable to identify chemical and structural characterization. The total contents of elements C, H, and N in the organic macromolecule fractions were measured using EA (Elementar Vario EL Cube, Hanau, Germany) accompanying with a standard high-temperature combustion procedure. Samples were weighted (∼3.0 mg) and placed in a tin capsule. The morphologies (shapes and sizes) were determined using SEM (Hitachi S-4800, Japan). Samples were coated with Au for 180 s and were analyzed at 5 kV with a distance of about 6–8 mm. The chemical functional groups were determined using FTIR (Nicolet iS10, Thermo, USA). About 1-mg sample was blended with KBr (100 mg) and then was pressed into a tablet form. A spectral region of 5000–400 cm−1 with a resolution of 3 cm−1 was selected. The δ13C/δ15N values were measured by elemental analyzer/isotope ratio mass spectrometry (EA/IRMS). An elemental analyzer (Flash HT 2000, Thermo Fisher Scientific) was connected in continuous flow mode through a gas chromatography (GC) column to an isotope ratio mass spectrometry system (DELTA V Advantage, Thermo Fisher Scientific). About 1–5 mg of organic macromolecules in tin capsules was placed into combustion oven at 960 °C under oxygen-containing condition, converting organic macromolecules to CO2 or N2 for GC IRMS analysis. Trajectory analysis HYSPLIT Model (http://ready.arl.noaa.gov/HYSPLIT. php) developed by National Oceanic and Atmospheric Administration (NOAA) Air Resources Laboratory (ARL) was performed to analyze the air masses during the sampling period. Meteorological data was obtained from National Centers for Environmental Prediction and Global Reanalysis (ftp://arlftp.arlhq.noaa.gov/pub/). Air mass trajectories were calculated for wintertime and summertime with 12 h intervals (at 0:00, 6:00, 12:00, and

Environ Sci Pollut Res

18:00). This approach was allowed to identify potential source input from long-range transport of air masses toward the monitoring site in Xiamen.

Results and discussion Elemental compositions and atomic ratios of organic macromolecules The elemental compositions and atomic ratios of organic macromolecules in PM10 are shown in Fig. 1. The contents of elements C, H, and N in K fraction were estimated by subtracting each content in BC from that in KB. The C contents in HA, K, and BC ranged from 32.41 to 45.62 %, 54.40 to 70.40 %, and 58.28 to 66.69 %, respectively. The C contents in HA were significantly lower than those in K and BC, while the N and H contents showed an opposite order with higher values in HA. The C, N, and H contents in K fraction were generally similar to those in BC. In JM, the C content for HA increased from summer to winter in JM, with the increasing of 11 %. In addition, the C content for BC in winter was

Fig. 1 Elemental compositions and atomic ratios of organic macromolecules in PM10 from the suburban (Jimei, JM) and urban (Siming, SM) site

normally higher than those in summer, regardless of the type of sampling site. The seasonal pattern could partly be attributed to the low atmospheric boundary layer in winter, which caused the accumulation of air pollutants and resulted in high concentrations of organic matter. The H/C ratio ranged from 1.52 to 1.75 for HA, 0.87 to 0.99 for K, and 0.32 to 0.98 for BC (Fig. 1). The H/C ratios for three organic macromolecule fractions were generally higher in JM than those in SM, especially for BC. No significant seasonal difference was observed for H/C ratio in SM. However, the seasonal variations of H/C ratios for different organic macromolecule fractions were not consistent in JM, which presented higher ratio for K and BC in winter but higher ratio for HA in summer. The C/N ratios of organic macromolecule fractions showed a reverse order, namely HA < K < BC. The C/N ratio for BC was lower in JM than those in SM. Compared with HA and K, seasonal variation of C/N ratios for BC was more apparent, which showed the lowest C/N ratios in summer in JM. In this study, the H/C ratios (1.52–1.75) for HA extracted from PM10 were higher than those from soil or river/marine/ mangrove sediments (0.69–1.34) (Barančíková et al. 1997; Song et al. 2002; Zhang et al. 2011) but comparable to those extracted from atmospheric dusts (1.52–2.06) and TSP (1.07– 1.90) (Zhao et al. 2011; Song et al. 2012). The C/N ratios for HA in this study were also comparable to those data reported from TSP sample in Guangzhou (16.7–33.3) (Peng et al. 2013). The HULIS, being extracted from water-soluble organic aerosol fractions in PM2.5 and determined by a solid-phase extraction (SPE) protocol, presented similar H/C and C/N ratios of 1.49 and 22, respectively (Salma et al. 2007). Atmospheric HA and K come from multiple sources, such as photochemical reactions of VOCs from vehicle exhaust (Zhao et al. 2011). Aromatic compounds originated from biomass burning might undergo complex oxidation processes and lead to the formation of HA (Gelencsér et al. 2003). Also, previous studies (Zhao et al. 2011) found that a portion of K in the atmospheric environment could derive from the oxidation and intermolecular combination of HA. This provides a reasonable explanation for the fact that K in PM10 were more mature than HA fractions. HA in PM10 was a product of oxidation of VOCs, showing high H/C ratio, and low C/N ratio compared to K. Further oxidation processes involving HA or intermolecular combinations of HA would cause significant loss of CO2, H2O, and NOx, resulting in the decrease of H/C ratios and increase of C/N ratios. In addition, the majority of BC fractions in PM10 were regarded as primary aerosols originating from incomplete combustion of fuels and biomass. During the process, most of O, N, and H were lost and resulted in the lowest H/C ratios and highest C/N ratios. Therefore, these provided a reasonable explanation for the order of C contents and H/C and C/N ratios of HA, KB, and BC extracted from PM10.

Environ Sci Pollut Res

Relative contents of organic macromolecules and TOC in PM10 The relative contents of HA, KB, and BC in TOC were calculated based on organic carbon contents, using the following Eq. (1): O⋮i ¼

TOCO⋮i  M O⋮i 00 % TOCO⋮o  M O⋮o

ð1Þ

where OM is the organic macromolecules, OMi (wt%) is the relative content of HA, KB, and BC in TOC, MOMi (g) is the mass of the isolated fraction, TOCOMi ( %) is the carbon content of the isolated fraction, MOMo (g) is the mass of PM10 sample, from which the fractions of organic macromolecules were isolated, and TOCOMo ( %) is the TOC contents in PM10. The average value of Mgi was calculated based on six replicates, and the relative standard deviations were less than 6 %. Contents of different organic macromolecule fractions in PM10 and TOC are summarized in Table 1. HA, K, and BC accounted for 3.89 to 4.55 % of PM10, while they were the dominant fractions of TOC, ranging from 64.70 to 84.99 %. The relative contents of HA, K, and BC in TOC ranged from 1.60 to 2.71 %, 29.28 to 55.52 %, and 18.09 to 29.77 %, respectively. We found that more than 50 % of the TOC were comprised by K and BC. It suggested that anthropogenic input was the predominant source of organic matter in PM10. The relative contents of organic macromolecules of TOC in this study were comparable with previous work performed (Zhao et al. 2011), who found that K and BC contained 41.51–60.76 and 25.22–33.32 % of TOC in dust, respectively. Some studies also reported that BC contents of TOC in PM10 were proportionate with those in TSP (10.1–41.3 %) (Salam et al. 2003). The increase of HA/PM10 from various seasons and sites was related with the increased H/C ratios, while the relationships between KB/PM10 and H/C ratios as well as BC/PM10 and H/C ratios were inverse. The positive correlation between HA/PM10 and H/C ratios might indicate that H–C functional group were the predominant component in HA. The contents of organic macromolecule fractions in TOC from atmospheric samples are listed in Table 2. The relative contents of HA (2.22 %) and BC (25.93 %) in Xiamen were similar to those in falling dust (2.43 % for HA and 28.92 % for Table 1 Relative content of organic macromolecules in PM10 and TOC

Samples

JM-summer JM-winter SM-summer SM-winter

BC) in Guangzhou. However, the relative content of HA in PM10 was significantly lower than that in TSP. The HA, K, and BC contents in TOC for atmospheric samples were quite consistent, with the higher contribution for BC and K than for HA (Peng et al. 2013; Zhao et al. 2011). The HA in this study and HULIS in aerosols reported from other studies were different concepts. HULIS were a class of organic compounds identified in atmospheric samples, which lied in water-soluble organic matters. The large gap was observed between HA content in TOC (2.22–5.61 %) and HULIS content in TOC (20– 54.6 %). However, similar H/C and C/N ratios were found in HA in this study and in HULIS from other reports, as previous mentioned, which might suggest the common sources for them, such as biomass burning (Fig. 2). Morphological properties of KB and BC To understand particle morphology and sources of organic macromolecules, SEM photographs of KB and BC extracted from PM10 are investigated (Fig. 3). However, due to high concentration content of free radicals, HA in PM10 were unsuitable for SEM measurements (Zhao et al. 2011). The results showed various morphological properties of KB (plant debris and irregular structures, Fig. 3a, b) and BC (spherical structure, Fig. 3c, d). Morphology of spherical BC particles was influenced by temperature and the degree of combustion, as well as the nature of the materials. In this study, BC particles were spherical (Fig. 3c, d) and non-spherical (cylindrical and elongated structures, Fig. 3e, f). The unique pore structure on the surface of BC particles was formed during the combustion processes. The spherical BC particles with porous/nonporous characteristics might originate from oil or coal, in agreement with previous reports from sediment and dust (Song et al. 2002; Zhao et al. 2011; Chen et al. 2013). In addition, the elongate BC particle with a porous structure (Fig. 3f) was similar to that of a porous combustion-derived carbon particle in wood charcoal (Jonker and Koelmans 2002; Fernandes et al. 2003). FTIR spectroscopic analysis of organic macromolecules The FTIR spectra of HA, KB, and BC in PM10 are shown in Fig. 4. The exhibited peaks of HA, KB, and BC were

Organic macromolecules/PM10 (%)

Organic macromolecules/TOC (%)

HA

KB

BC

K

Sum

HA

KB

BC

K

Sum

0.33 0.13 0.15 0.24

4.21 3.76 3.94 3.76

1.89 1.06 1.81 1.74

2.33 2.70 2.13 2.02

4.55 3.89 4.09 3.99

2.65 1.60 1.91 2.71

56.18 73.61 58.32 60.07

26.90 18.09 28.97 29.77

29.28 55.52 29.35 30.31

58.82 75.20 60.23 62.78

JM Jimei suburban site, SM Siming urban site

Environ Sci Pollut Res Table 2 Percentage of organic macromolecules in TOC from atmospheric particles Samples

HA/(%) KB/(%) BC/(%) K/(%) Reference

PM10 TSP

2.22

62.05

25.93

36.12

This study

5.61

45.68

23.28

22.40

Peng et al. (2013)

Falling dust 2.43

80.56

28.92

51.63

Zhao et al. (2011)

HA humic acid, KB kerogen + black carbon, BC black carbon, K kerogen

similar to each other among all PM10 samples from different monitoring sites and seasons. The results indicated that the structural and functional groups of HA, KB, and BC have some similarity. The assignments of the main bands were based on previous studies (Deniau et al. 2001; Zhao et al. 2011; Ji et al. 2015; Roy et al. 2016). In this study, HA in PM 1 0 showed strong absorption at around 3400 cm−1 (Fig. 4), which corresponded to H-bonded OH groups of alcohols, phenols, and organic acids, as well as H-bonded N–H groups. The stronger hydrogen bonds in HA could be attributed to the carboxylic acid functional group, which created intermolecular linkages. The absorption peaks at 2930 and 2860 cm−1 were affected by symmetric and asymmetric stretching vibrations of C–H bonds in –CH3 and –CH2– of alkyl structures. HA, KB, and BC had absorption bands around 1710 cm−1, which was attributed to the C=O stretching vibration of COOH groups and the C=O stretching vibration of other carbonyl groups, such as ketonic acid. Similar peak locations and intensities were observed at ca. 1600 and 1400 cm−1 for HA, KB, and BC. The former bands were attributed to the C=C aromatic, COO−, and C = O and the later bands assigned to C–H bending vibrations of C–(CH3)2 or C–(CH3)3 groups. The transmittance at around 1100 cm−1 was affected by the –C– O–C of carbohydrates and aromatic ethers. As shown in Fig. 4, the notable peaks at 3400, 2930, and 2860 cm−1 in HA are attributed to aliphatic structures, such as long-chain fatty and carbohydrates with a carboxylic extremity. The results were similar to those reported in dusts (Zhao et al. 2011) and were consistent with the fact that HA in PM10 had the highest H/C ratios. This also could be further proved by high correlation coefficient

Fig. 2 Correlations between HA/PM10, KB/PM10, BC/PM10, and H/C ratio

(r = 0.926; P < 0.05) between HA/PM10 and H/C ratios (Fig. 2). Roy et al. (2016) found that dominant peaks at 1020, 1430–1455, 2360–2336, and 3448 cm−1 in PM10 samples reflected the presence of S–O, CO2, SO2, C≡C, and SiO–H bond stretch. This bond stretching obtained from the spectral analysis indicated mine fire as a possible source of PM 10 . In addition, a remarkable peak at 1390 cm−1 that appeared in PM2.5 samples was attributed to NH4NO3, suggesting the occurrence of secondary processes (Ji et al. 2015). Current studies reported that HA in the atmosphere mainly originated from the combustion sources, soil dust, vegetative detritus, biomass burning, and photochemical degradation of NOx and volatile organic compounds (Gelencsér et al. 2003; Salma et al. 2007; Lin et al. 2010a; Zhao et al. 2011; Peng et al. 2013). Source apportionment of organic macromolecules in PM10 C/N ratios The C/N ratios have been proved as reliable indicators of potential sources of organic matter in terrestrial, estuarine, and coastal environments (Jensen et al. 2005; Selvaraj et al. 2012; Wang et al. 2013). For example, C3 vascular plant material had the C/N ratios of above 12, whereas C4 grasses showed C/N ratios above 30 (Meyers 1994). The difference of C/N values could be attributed to the variations of nitrogen content in plants. Some studies found that bacteria and algae had low C/N ratios of 4–6 and below 10, respectively (Meyers 1994). The C/N ratios for wood charcoal and diesel soot were high (over 50), while the C/N ratios were about 40 for vegetation fire residues, 25 for straw charcoal, and 17 for chimney soot (Fernandes et al. 2003). In this study, the average C/N ratios of HA, K, and BC in PM10 were 20.90, 55.97, and 114.35, respectively, indicating the influence of multiple sources. The C/N ratios of K were significantly higher than previous studies (14–33 of K) (Hassan and Spalding 2001), which could be explained by BC with high C/N ratio in KB samples. In our previous study, at the urban (SM), suburban (JM), and background sites,

Environ Sci Pollut Res

Fig. 3 SEM photographs of KB and BC in PM10 from the suburban (Jimei, JM) and urban (Siming, SM) site. SEM photographs of KB and BC in PM10 from the suburban (Jimei, JM) and urban (Siming, SM) site. a JMSKB, plant debris (×11.0K). b JMSKB, irregular structure with smooth surface (×12.0K). c JMSKB, spherical BC. Particles with nonporous surface (×15.0K). d JMWKB, spherical BC particles with

random pores locating on the surface (×13.0K). e SMSBC, cylindrical structure with smooth, porous surface (×3.00k). f SMSBC, elongate BC particle with porous structure (×5.00k). Naming is site (JM: suburban site; SM: urban site) followed by season (S: summer; W: winter), and the subscript is each organic macromolecule fraction

carbonaceous species, levoglucosan, and 14C in PM 2.5 were measured to study source contributions to total carbon (TC). The results showed that fossil fuel combustion is the main contributor (62.9–72.2 %) to TC at urban and suburban sites. Biogenic emissions have important contribution (winter, 52.98 %; summer, 45.71 %) to TC at background site (Niu et al. 2013a). We also found that 42.1 ± 15.1 % of TC in November 2010 was attributed to biomass burning in the coastal urban agglomeration, such as Xiamen, Quanzhou, and Fuzhou city (Niu et al. 2013b). Therefore, in this study, HA in PM10 was possibly affected by biomass burning, especially emitted from terrestrial C3 vascular plants, while K and BC mainly originated from the contributions from wood charcoal, diesel, and vegetation fire residues.

Carbon/nitrogen isotope values (δ13C/δ15N) The δ13C/δ15N values were generally used to trace the sources of organic compounds isolated from various environment samples (Grice et al. 2003; Hiradate et al. 2004; Cao et al. 2008; Cao et al. 2011; Natali and Bianchini 2015; Hegde et al. 2016). The δ13C/δ15N values of HA, KB, and BC in PM10 are listed in Table 3. The results showed that δ13C values of organic macromolecules in PM10 exhibited obviously seasonal and spatial pattern. The δ13C values of PM10 ranged from −27.19 to −25.36 ‰, with a discrepancy of 1.38 ‰. The variances of δ13C values were changed by 1.67 ‰ from −27.59 to −25.92 ‰ for HA, by 1.27 ‰ from −26.75 to −25.48 ‰ for KB, and by 1.30 ‰ from −27.03 to −25.73 ‰ for BC.

Fig. 4 FTIR spectra of HA, KB and BC in PM10 from the suburban (Jimei, JM) and urban (Siming, SM) site

Environ Sci Pollut Res Table 3 The δ13C/δ15N values of the organic macromolecule fractions in PM10

Sample

δ13C/‰

δ15N/‰

PM10

HA

KB

BC

PM10

HA

KB

BC

JM-summer

−25.36

−25.92

−25.96

−26.34

12.53

3.74

1.71

1.26

JM-winter

−26.05

−27.47

−26.75

−26.88

7.47

9.22

1.86

1.98

SM-summer SM-winter

−26.18 −27.19

−26.52 −27.59

−25.48 −26.50

−25.73 −27.03

7.57 5.47

3.87 6.79

4.14 2.05

5.56 2.99

JM Jimei suburban site, SM Siming urban site

The δ13C values of HA, KB, and BC were close to C3 vegetation (average δ13C of −29.0 ± 1.8 ‰), differing from those of C 4 plants (−13.1 ± 0.5 ‰) (Yu et al. 2010). In this study, sampling sites were located in a subtropical area, where C3 plant (e.g., rice) was the predominant contribution to the fractions of organic macromolecules in PM10. The variations of δ13C values in organic matters from C3 plants had been reported (Kawashima and Haneishi 2012), such as rice of −28.0 ‰, dry leaf of −29.4 ‰, and soybean of −28.8 ‰. In addition, δ13C of cellulose and hemicelluloses in plant materials ranged from −28.5 to −23.7 ‰ (Sang et al. 2012). Current studies suggested that the δ13C values of BC were affected by chemical composition of fuels, combustion type, temperature, and efficiency of emission control devices. It was found that the δ13C values of BC in fresh aerosols from fossil fuel sources were about −27 ‰ (Huang et al. 2006). The coal combustion had the δ13C values of −24.9 to −21.0 ‰, while motor vehicle exhausts were lighter (−28.4 to −26.0 ‰) (Gleason and Kyser 1984; Widory 2006; Bush et al. 2007). In this study, BC extracted from PM10 showed lighter δ13C values of −25.73 to −27.03 ‰, indicating the contributions of fossil fuels and vehicle exhausts. The average δ13C values of BC fractions in PM2.5 collected in Xiamen were −25.68 ‰ for winter and −26.63 ‰ for summer (Cao et al. 2011), which were comparable to those of BC in PM10.

The δ15N values ranged from 3.74 to 9.22 ‰ for HA, from 1.74 to 4.14 ‰ for KB, and from 1.26 to 5.56 ‰ for BC. The δ 15N values of HA showed the influence of terrestrial plants, while KB and BC were affected by coal, gasoline particle, and terrestrial plants (Kelly et al. 2005; Widory 2007; Hegde et al. 2016). The δ15N values for HA in summer (3.74 ‰ in JM and 3.87 ‰ in SM) were lower than that in winter (9.22 ‰ in JM and 6.79 ‰ in SM). The variations of δ15N values in different seasons were supported by the fact that there was a negative relationship between terrestrial plants δ15N with precipitation (Liu et al. 2014). In summer, there was much more precipitation than that in winter, causing lighter δ15N values from terrestrial plant in PM10. Compared to JM, SM presented various δ15N values of KB and BC in both summer and winter, suggesting the difference of emission source. The temporal variations of δ15N values indicated that these organic macromolecules might be controlled by changes in local sources, photochemical transformations, and air mass origins. Transport pathways of organic macromolecules Pollution characteristics of organic macromolecules in PM10 are closely related to local and regional emission sources. In this study, cluster analysis of the 3-day

Fig. 5 Seventy-two-hour backward trajectories for air masses during the sampling period

Environ Sci Pollut Res

backward trajectories was performed through different seasons to visualize the long-range transport of organic macromolecules to the monitoring site (Fig. 5). In summer, the air masses mainly came from the South China Sea (34 %) and the Western Pacific Sea (66 %), representing clean marine air masses with low concentrations of organic macromolecules. In winter, the air masses originated from Mongolia (4 %), the northern area of China (48 %), and northern adjacent cities (48 %), indicating the influence of anthropogenic sources through long-range transport. This might be the reason why the C and H content of organic macromolecules in winter was normally higher than those in summer (Fig. 1). Meanwhile, as shown in Fig. 4, seasonal variations of functional group of HA, KB, and BC in PM10 were also observed. Previous studies found that long-range transport of carbonaceous aerosols could make an important contribution to the elevated BC concentrations in the atmosphere. These results implied that both anthropogenic emissions from local sources and regional long-range transport were required to comprehensively identify their sources of organic macromolecules in particular matter.

Conclusions Elemental analysis, SEM, FTIR, δ13C/δ15N values, and the HYSPLIT Model were used to investigate the characterization and sources of HA, K, and BC in PM10 in a coastal city in China. K and BC were the dominant fractions of TOC, ranging from 56.18 to 73.61 %. The morphology of BC showed that particles in PM10 were usually porous, which were influenced by the combustion processes. HA in PM10 contained a higher fraction of aliphatic structures, such as long-chain fatty and carbohydrates with a carboxylic extremity. Through the C/N ratio, SEM, and δ13C/δ15N values, the results suggested that HA and K majorly originated from terrestrial plants, and BC came from the mixture of combustion of terrestrial plants, fossil fuel, and charcoal. The seasonal variations of the δ15N values in HA were supported by the fact that there was a negative relationship between terrestrial plants δ15N and precipitation. The HYSPLIT trajectory model showed that the air masses in summer mainly came from the South China Sea (34 %) and the Western Pacific Sea (66 %), representing clean marine air masses with low concentrations of organic macromolecules. In winter, the air masses originated from Mongolia (4 %), the northern area of China (48 %), and northern adjacent cities (48 %), indicating the influence of anthropogenic sources through long-range transport. These results implied that both anthropogenic emissions from local sources and regional long-range transport were required to identify their sources of organic macromolecules in atmospheric particles.

Acknowledgements This research was financially supported by the National Natural Science Foundation of China (Nos. U1405235, 41303072, 41575146, 31300435, and 21507127), the Key Research Program of the Chinese Academy of Sciences (No. KJZDEW-TZ-G0602), the Knowledge Innovation Program of the Chinese Academy of Sciences (No. IUEZD201403), National Key Research and Development Program (No. 2016YFC0200501), Chinese Academy of Sciences Interdisciplinary Innovation Team, and the Natural Science Foundation of Fujian Province, China (2016J01201 and 2016J05050).

References Barančíková G, Senesi N, Brunetti G (1997) Chemical and spectroscopic characterization of humic acids isolated from different Slovak soil types. Geoderma 78(3–4):251–266 Barbaro E, Zangrando R, Moret I, Barbante C, Cescon P, Gambaro A (2011) Free amino acids in atmospheric particulate matter of Venice, Italy. Atmos Environ 45(28):5050–5057 Bush SE, Pataki DE, Ehleringer JR (2007) Sources of variation in δ13C of fossil fuel emissions in Salt Lake City, USA. Appl Geochem 22(4): 715–723 Bushnev DA, Burdel’naya NS (2009) Kerogen: chemical structure and formation conditions. Russ Geol Geophys 50(7):638–643 Cao G, Zhang X, Zheng F (2006) Inventory of black carbon and organic carbon emissions from China. Atmos Environ 40(34):6516–6527 Cao J, Chow JC, Tao J, Lee S, Watson JG, Ho K, Wang G, Zhu C, Han Y (2011) Stable carbon isotopes in aerosols from Chinese cities: influence of fossil fuels. Atmos Environ 45(6):1359–1363 Cao J, Zhu C, Chow J, Liu WG, Han YM, Watson JG (2008) Stable carbon and oxygen isotopic composition of carbonate in fugitive dust in the Chinese Loess Plateau. Atmos Environ 42(40): 9118–9122 Chefetz B, Xing B (2009) Relative role of aliphatic and aromatic moieties as sorption domains for organic compounds: a review. Environ Sci Technol 43(6):1680–1688 Chen Y, Zhao J, Yin L, Chen J, Yuan D (2013) Quantification, morphology and source of humic acid, kerogen and black carbon in offshore marine sediments from Xiamen Gulf, China. J Environ Sci 25(2): 287–294 Claeys M, Vermeylen R, Yasmeen F, Gómez-González Y, Chi X, Maenhaut W, Mészáros T, Salma I (2012) Chemical characterisation of humic-like substances from urban, rural and tropical biomass burning environments using liquid chromatography with UV/vis photodiode array detection and electrospray ionisation mass spectrometry. Environ Chem 9(3):273–284 Deniau I, Derenne S, Beaucaire C, Pitsch H, Largeau C (2001) Morphological and chemical features of a kerogen from the underground Mol laboratory (Boom Clay Formation, Oligocene, Belgium): structure, source organisms and formation pathways. Org Geochem 32(11):1343–1356 Fan X, Song J, Peng P (2013) Comparative study for separation of atmospheric humic-like substance (HULIS) by ENVI-18, HLB, XAD-8 and DEAE sorbents: elemental composition, FT-IR, 1H NMR and off-line thermochemolysis with tetramethylammonium hydroxide (TMAH). Chemosphere 93(9):1710–1719 Fan X, Song J, Peng PA (2012) Comparison of isolation and quantification methods to measure humic-like substances (HULIS) in atmospheric particles. Atmos Environ 60:366–374 Fernandes MB, Skjemstad JO, Johnson BB, Wells JD, Brooks P (2003) Characterization of carbonaceous combustion residues. I. Morphological, elemental and spectroscopic features. Chemosphere 51(8):785–795

Environ Sci Pollut Res Gelencsér A, Hoffer A, Kiss G (2003) In-situ formation of light-absorbing organic matter in cloud water. J Atmos Chem 45(1):25–33 Gleason JD, Kyser TK (1984) Stable isotope compositions of gases and vegetation near naturally burning coal. Nature 307(5948):254–257 Grice K, Schouten S, Blokker P, Derenne S, Largeau C, Nissenbaum A, Damsté JSS (2003) Structural and isotopic analysis of kerogens in sediments rich in free sulfurised Botryococcus braunii biomarkers. Org Geochem 34(3):471–482 Hassan KM, Spalding RF (2001) Hydrogen isotope values in lacustrine kerogen. Chem Geol 175(3–4):713–721 Hegde P, Kawamura K, Joshi H, Naja M (2016) Organic and inorganic components of aerosols over the central Himalayas: winter and summer variations in stable carbon and nitrogen isotopic composition. Environ Sci Pollut Res 23(7):6102–6118 Hiradate S, Nakadai T, Shindo H, Yoneyama T (2004) Carbon source of humic substances in some Japanese volcanic ash soils determined by carbon stable isotopic ratio, δ13C. Geoderma 119(1–2):133–141 Huang L, Brook JR, Zhang W, Li SM, Graham L, Ernst D, Chivulescu A, Lu G (2006) Stable isotope measurements of carbon fractions (OC/ EC) in airborne particulate: a new dimension for source characterization and apportionment. Atmos Environ 40(15):2690–2705 Janssen NA, Hoek G, Simic-Lawson M, Fischer P, van Bree L, ten Brink H, Keuken M, Atkinson RW, Anderson HR, Brunekreef B, Cassee FR (2011) Black carbon as an additional indicator of the adverse health effects of airborne particles compared with PM10 and PM2.5. Environ Health Perspect 119(12):1691–1699 Jensen LS, Salo T, Palmason F, Breland TA, Henriksen TM, Stenberg B, Pedersen A, Lundström C, Esala M (2005) Influence of biochemical quality on C and N mineralisation from a broad variety of plant materials in soil. Plant Soil 273(1–2):307–326 Ji Z, Dai R, Zhang Z (2015) Characterization of fine particulate matter in ambient air by combining TEM and multiple spectroscopic techniques—NMR, FTIR and Raman spectroscopy. Environmental Science-Processes & Impacts 17:552–560 Jonker MTO, Koelmans AA (2002) Sorption of polycyclic aromatic hydrocarbons and polychlorinated biphenyls to soot and soot-like materials in the aqueous environment: mechanistic considerations. Environ Sci Technol 36(17):3725–3734 Karthikeyan S, Balasubramanian R (2006) Determination of watersoluble inorganic and organic species in atmospheric fine particulate matter. Microchem J 82(1):49–55 Kawashima H, Haneishi Y (2012) Effects of combustion emissions from the Eurasian continent in winter on seasonal δ13C of elemental carbon in aerosols in Japan. Atmos Environ 46:568–579 Kelly SD, Stein C, Jickells TD (2005) Carbon and nitrogen isotopic analysis of atmospheric organic matter. Atmos Environ 39(32): 6007–6011 Koch D, Schulz M, Kinne S, McNaughton C, Spackman JR, Balkanski Y, Bauer S, Berntsen T, Bond TC, Boucher O, Chin M, Clarke A, De Luca N, Dentener F, Diehl T, Dubovik O, Easter R, Fahey DW, Feichter J, Fillmore D, Freitag S, Ghan S, Ginoux P, Gong S, Horowitz L, Iversen T, Kirkevag A, Klimont Z, Kondo Y, Krol M, Liu X, Miller R, Montanaro V, Moteki N, Myhre G, Penner JE, Perlwitz J, Pitari G, Reddy S, Sahu L, Sakamoto H, Schuster G, Schwarz JP, Seland Ø, Stier P, Takegawa N, Takemura T, Textor C, van Aardenne JA, Zhao Y (2009) Evaluation of black carbon estimations in global aerosol models. Atmos Chem Phys 9(22): 9001–9026 Lin P, Engling G, Yu JZ (2010a) Humic-like substances in fresh emissions of rice straw burning and in ambient aerosols in the Pearl River Delta Region, China. Atmos Chem Phys 10(14):6487–6500 Lin P, Huang X, He L, Yu J (2010b) Abundance and size distribution of HULIS in ambient aerosols at a rural site in South China. J Aerosol Sci 41(1):74–87 Liu X, Zhang Y, Su Q, Tian Y, Wang Q, Quan B (2014) Progress of research on relationships between terrestrial plant nitrogen isotope

composition and climate environment change. Advances in Earth Science 29(2):216–226 (in Chinese) Lopes SP, Matos JTV, Silva AMS, Duarte AC, Duarte RMBO (2015) 1H NMR studies of water- and alkaline-soluble organic matter from fine urban atmospheric aerosols. Atmos Environ 119:374–380 Lyamani H, Olmo FJ, Foyo I, Alados-Arboledas L (2011) Black carbon aerosols over an urban area in south-eastern Spain: changes detected after the 2008 economic crisis. Atmos Environ 45(35):6423–6432 Masiello CA (2004) New directions in black carbon organic geochemistry. Mar Chem 92(1–4):201–213 Masiello CA, Druffel ERM, Currie LA (2002) Radiocarbon measurements of black carbon in aerosols and ocean sediments. Geochim Cosmochim Acta 66(6):1025–1036 Meyers PA (1994) Preservation of elemental and isotopic source identification of sedimentary organic matter. Chem Geol 114(3–4):289–302 Natali C, Bianchini G (2015) Thermally based isotopic speciation of carbon in complex matrices: a tool for environmental investigation. Environ Sci Pollut Res 22(16):12162–12173 Niu ZC, Wang S, Chen JS, Zhang FW, Chen XQ, He C, Lin LF, Yin LQ, Xu LL (2013a) Source contributions to carbonaceous species in PM2.5 and their uncertainty analysis at typical urban, peri-urban and background sites in southeast China. Environ Pollut 181:107–114 Niu ZC, Zhang FW, Chen JS, Yin LQ, Wang S, Xu LL (2013b) Carbonaceous species in PM2.5 in the coastal urban agglomeration in the Western Taiwan Strait Region, China. Atmos Res 122:102–110 Peng PA, Zhao J, Song J, Ma S (2013) Humic acid, kerogen and black carbon isolated from atmospheric total suspended particulate from Guangzhou, China. Aerosol Sci Technol 47:1342–1352 Roy D, Gautam S, Singh P, Singh G, Das BK, Patra AK (2016) Carbonaceous species and physicochemical characteristics of PM10 in coal mine fire area—a case study. Air Quality Atmosphere and Health 9:429–437 Salam A, Bauer H, Kassin K, Mohammad Ullah S, Puxbaum H (2003) Aerosol chemical characteristics of a mega-city in Southeast Asia (Dhaka–Bangladesh). Atmos Environ 37(18):2517–2528 Salma I, Mészáros T, Maenhaut W, Vass E, Majer Z (2010) Chirality and the origin of atmospheric humic-like substances. Atmos Chem Phys 10(3):1315–1327 Salma I, Ocskay R, Chi X, Maenhaut W (2007) Sampling artefacts, concentration and chemical composition of fine water-soluble organic carbon and humic-like substances in a continental urban atmospheric environment. Atmos Environ 41(19):4106–4118 Sang XF, Gensch I, Laumer W, Kammer B, Chan CY, Engling G, Wahner A, Wissel H, Kiendler-Scharr A (2012) Stable carbon isotope ratio analysis of anhydrosugars in biomass burning aerosol particles from source samples. Environ Sci Technol 46(6):3312–3318 Selvaraj K, Wei KY, Liu KK, Kao SJ (2012) Late Holocene monsoon climate of northeastern Taiwan inferred from elemental (C, N) and isotopic (δ13C, δ15N) data in lake sediments. Quaternary Sci Rev 37:48–60 Siudek P, Frankowski M, Siepak J (2015) Seasonal variations of dissolved organic carbon in precipitation over urban and forest sites in central Poland. Environ Sci Pollut Res 22(14):11087–11096 Song J, He L, Peng PA, Zhao J, Ma S (2012) Chemical and isotopic composition of humic-like substances (HULIS) in ambient aerosols in Guangzhou, South China. Aerosol Sci Technol 46(5):533–546 Song J, Peng PA, Huang W (2002) Black carbon and kerogen in soils and sediments. 1. Quantification and characterization. Environ Sci Technol 36(18):3960–3967 Staudt S, Kundu S, Lehmler HJ, He X, Cui T, Lin YH, Kristensen K, Glasius M, Zhang X, Weber RJ, Surratt JD, Stone EA (2014) Aromatic organosulfates in atmospheric aerosols: synthesis, characterization, and abundance. Atmos Environ 94:366–373 Vandenbroucke M (1980) Structure of kerogens as seen by investigations on soluble extracts in kerogen: insoluble organic matter from sedimentary rocks. Edition Technip, France, pp 1415–1443

Environ Sci Pollut Res Vandenbroucke M, Largeau C (2007) Kerogen origin, evolution and structure. Org Geochem 38(5):719–833 Vargas C, Perez-Esteban J, Escolastico C, Masaguer A, Moliner A (2016) Phytoremediation of Cu and Zn by vetiver grass in mine soils amended with humic acids. Environ Sci Pollut Res 23(13):13521–13530 Wang L, Mackay AW, Leng MJ, Rioual P, Panizzo VN, Lu H, Gu Z, Chu G, Han J, Kendrick CP (2013) Influence of the ratio of planktonic to benthic diatoms on lacustrine organic matter δ13C from Erlongwan maar lake, northeast China. Org Geochem 54:62–68 Widory D (2006) Combustibles, fuels and their combustion products: a view through carbon isotopes. Combust Theor Model 10(5):831–841 Widory D (2007) Nitrogen isotopes: tracers of origin and processes affecting PM10 in the atmosphere of Paris. Atmos Environ 41(11): 2382–2390 Xiao B, Yu Z, Huang W, Song J, Peng PA (2004) Black carbon and kerogen in soils and sediments. 2. Their roles in equilibrium sorption of lesspolar organic pollutants. Environ Sci Technol 38(22):5842–5852 Yu F, Zong Y, Lloyd JM, Huang G, Leng MJ, Kendrick C, Lamb AL, Yim WWS (2010) Bulk organic δ13C and C/N as indicators for sediment sources in the Pearl River delta and estuary, southern China. Estua Coast Shelf S 87(4):618–630

Zappoli S, Andracchio A, Fuzzi S, Facchini MC, Gelencsér A, Kiss G, Krivácsy Z, Molnár Á, Mészáros E, Hansson HC, Rosman K, Zebühr Y (1999) Inorganic, organic and macromolecular components of fine aerosol in different areas of Europe in relation to their water solubility. Atmos Environ 33(17):2733–2743 Zhang Y, Du J, Zhang F, Yu Y, Zhang J (2011) Chemical characterization of humic substances isolated from mangrove swamp sediments: the Qinglan area of Hainan Island, China. Estua Coast Shelf S 93(3): 220–227 Zhao J, Peng PA, Song J, Ma S, Sheng G, Fu J (2009) Characterization of organic matter in total suspended particles by thermodesorption and pyrolysis-gas chromatography-mass spectrometry. J Environ Sci 21(12):1658–1666 Zhao J, Peng PA, Song J, Ma S, Sheng G, Fu J (2011) Characterization of macromolecular organic matter in atmospheric dust from Guangzhou, China. Atmos Environ 45(31):5612–5620 Zhao J, Peng PA, Song J, Ma S, Sheng G, Fu J, Yuan D (2012) Characterization of humic acid-like substances extracted from atmospheric falling dust using Py-GC-MS. Aerosol Air Qual Res 12(1):83–92