PM2.5 Associated PAHs and Inorganic Elements from Combustion of Biomass, Cable Wrapping, Domestic Waste, and Garbage for Power Generation

Received: October 3, 2019
Revised: October 21, 2019
Accepted: October 22, 2019
Download Citation: RIS|BibTeX|https://doi.org/10.4209/aaqr.2019.10.0495  

Cite this article:

Li, Z., Wang, Y., Guo, S., Li, Z., Xing, Y., Liu, G., Fang, R., Hu, Y., Zhu, H. and Yan, Y. (2019). PM2.5 Associated PAHs and Inorganic Elements from Combustion of Biomass, Cable Wrapping, Domestic Waste, and Garbage for Power Generation. Aerosol Air Qual. Res. 19: 2502-2517. https://doi.org/10.4209/aaqr.2019.10.0495

Highlights

• Cutting oil waste was recovered and transformed into refuse derived fuel(RDF).
• RDF was reclaimed in a steam generation plant.
• Substitution of RDF for wood dust improves the combustion condition.

Keywords: PAHs; Heavy metals; Cable wrapping; Garbage-fired power plant; Domestic garbage.

INTRODUCTION

Serious atmospheric quality deterioration has occurred in rapidly developing China (Fang et al., 2009; Pascal et al., 2013; Li et al., 2017a; Hu et al., 2019; Li et al., 2019). Zhai et al. (2019) reported that PM_2.5 should responsible for 1.1 × 10⁶ excess deaths in 2015 in China. The major sources of PM_2.5 in urban areas are incomplete combustion or gas-particle conversion, and PM_2.5 contain a varied mix of chemicals inorganic ions, organic carbon and elemental carbon (Bond et al., 2004; Saikawa et al., 2009). The PM_2.5 associated organic pollutants including PAHs, PCBs, PCDD/Fs, and heavy metals were often linked with damage of health and environment (Li et al., 2011, 2016, 2017a, b, 2018a, b; Mario et al., 2018). Exposure to high loading of atmospheric PAHs originate from both natural and anthropogenic sources is often linked with adverse health effects, especially for urban residents (Hamra et al., 2014; Elzein et al., 2019).

The PM_2.5 bounded micro-pollutants including PAHs, and heavy metals (HMs), and As could threaten the human health and endanger the ecosystem safety since their toxicity even at low levels, which had been paid more attention (Mario et al., 2018). China is the world’s largest PAHs producer and contributes 20.1% to global emissions in 2007 (Zhang et al., 2019). The main emission sources of these pollutants contained incomplete combustion of fossil fuels, agricultural burning, and industrial and agricultural activities (Li et al., 2011, 2017a, b; Abbas et al., 2018; Chen et al., 2018; Dat et al., 2018; Li et al., 2018a, b; Mario et al., 2018). The anthropogenic burning was identified as a predominant contributor to PAHs and HMs compared with the natural sources including volcanic eruptions and forest fires (Xu et al., 2006; Abbas et al, 2018).

The domestic garbage burning for volume reduction and power generation is a more concerned emission source of PAHs, dioxin, HMs, NO_x, and SO_x due to its consequent health and environmental risks (Xiang et al., 2013; Zhang et al., 2013; Liu et al., 2016). As an important measure to dispose the garbage and ease the energy crisis, the garbage-fired power plant (GFPP) was widely used in the world (Murphy and McKeogh, 2004). Various air pollutants including PAHs, PCBs, and HMs originated from GFPP were more concerned and the flue gas purification technology has become a research hotspot in recent years (Consonni et al., 2005; Liu, 2015). Along with the increasing pressure of environment protection, especially the atmospheric environment, the family workshops located at rural areas of China aiming to metal reclamation through the burning of electric cables were closed. However, some workshops were still secretly producing at night driven by the enormous economic benefits. Huang et al. (1992) reported the soil adjacent to these workshops was heavily contaminated by PCBs, PAHs, PCDDs, and PCDFs. Li et al. (2011) also indicated that high levels of PCBs were found in soil near these workshops. China is a big agricultural country with a lots of straw produced every year, which is often directly burned in field to facilitate the cultivation in the next year (Hou et al., 2017). The regional air pollution events derived from biomass burning were frequently observed in recent years regardless of the strict emission limitation from Chinese government (Zhang et al., 2015a; Hou et al., 2017).

To our knowledge, few systematic studies were conducted on the PM_2.5 associated PAHs, HMs, and As originated from the combustion of biomass, garbage for power generation, cable wrapping for metal reclamation, domestic garbage for volume reduction. In this study, a sum of 42 target compounds including 16 PAHs and 26 inorganic elements were analyzed for 18 PM_2.5 samples, which were collected using re-suspension of the ashes from these related emission sources. The 16 target PAHs including naphthalene (NaP), acenaphthylene (Acy), acenaphthene (Ace), fluorine (Fl), benzo(g,h,i)perylene (BghiP), phenanthrene (Phe), anthracene (Ant), fluoranthene (FA), pyrene (Pyr), benzo(a)anthracene (BaA), chrysene (Chr), benzo(b)fluoranthene (BbF), benzo(k)fluoranthene (BkF), benzo(a)pyrene (BaP), indeno(1,2,3-cd)pyrene (Ind), and dibenzo(a,h)anthracene (DBA) were analyzed using the HP6890 GC/5973i MS with selected ion mode (SIM) in this study. The 26 inorganic elements (IEs) such as Li, Be, Na, P, K, Sc, V, Cr, Mn, Co,Ni, Cu, Zn, As, Cd, Sn, Sb, Tl, Pb, Si, Al, Ca, Mg, Fe, Ba, and Sr were analyzed by ICP-MS and ICP-OES. All of this work was to achieve the following objects: 1) the contents of PAHs and IEs in PM_2.5; 2) the PAH and IE composition profile similarity between each pair of emission sources; 3) toxicity evaluation of PAHs; and 4) the carcinogenic and non-carcinogenic exposure risks of HMs and As.

METHODOLOGY

Sample Collection and Re-suspended for Obtainment of PM_2.5

The ash samples were collected from 4 GFPPs in Hebei- and Shaanxi-Province, 4 WSs and 4 PSs in Hebei Province, 3 WCCs in Heibei Province and Tianjin city, and 3 DGs in Hebei Province during January to May, 2015. The samples were collected from the electrostatic precipitators (ESP) of garbage-fired power plants, ash outlet at bottom of the stoves, and outdoor combustion residue using a stainless shovel. Finally, 1 kg of ash was obtained and dried using a vacuum freeze dryer, and subsequently stored in brown glass bottles before re-suspended for PM_2.5 fractions.

The PM_2.5 fractions within particulate matters were collected from fly ashes using re-suspension method described in detail by Chow et al. (1994), Kong et al. (2011), and Li et al. (2017b). Fig. 1 was a diagrammatic sketch for sample collection using the re-suspension chamber equipped with 4 sampling channels. For each PM_2.5 sample, 10 g of ash was blow out from dust feeding bottle using a pump and entered into the re-suspension chamber and then through 4 particle size cutters (2.5 µm) and intercepted by subsequent quartz filters (QFs) (diameter 47 mm; Pall Co. USA) and Teflon membranes (diameter 47 mm; Pall Co. USA). The ash derived from one kind of fuel experienced two repeated re-suspension and finally the ash intercepted in 8 QFs was used for the analysis of 16 PAHs to ensure the detection of as many PAH isomers as possible. The ash intercepted in 2 Teflon filter was used to analyze the 26 inorganic within PM_2.5. Suspension time was set as 10 min for each sample, and then stopped running for 180 min to eliminate any interference derived from the latest experiment for sample re-suspension and collection. Prior to sampling, all the quartz filters were baked at 600℃ for 2 h to eliminate the organic compounds, then equilibrated for 48 h in a room with constant temperature and humidity before and after sampling, and then weighed. The weight difference between pre- and post-sampling was used as collected PM_2.5 mass. The PM_2.5 sampling amounts were in the range of 208–1160 mg. The significant fluctuations existed among PM_2.5 yields of ashes derived from combustion of different fuels, and the maximum was possessed by peanut straw, while the minimum occurred at domestic waste.

Fig. 1. Re-suspension sampling system. (1) Sampling pump, (2) Total flow meter, (3) Filter, (4) Sampling gas path, (5) Support shelf, (6) Brantch flow meter, (7) Vacuum gauge, (8) Sampling platform, (9) Stainless steel particle size cutter, (10) Re-suspension chamber, (11) Gas path of dust injection, (12) Stainless steel bottle holder, (13) Dust feeding bottle, (14) Dust feeding pump.

PAHs Extraction and Analysis

The sample consists of 8 re-suspended QFs was Soxhlet extracted with actone: n-hexane (v:v = 1:1) for 24 h, concentrated by rotary vacuum evaporation, purified by a florisil column and re-concentrated to 0.5 mL under a gentle N₂ stream before analysis. The HP6890 GC/5973i MS system equipped with a DB-5 capillary column (30 m × 0.25 mm × 0.25 µm) was used to PAHs and PAEs analysis using selected ion mode (SIM). A total of 18 PAHs and 6 PAEs (DMP, DEP, DBP, DEHP, DOP, and BBP) were analyzed in this study and the corresponding PAH and PAE standards were all purchased from Dr. Ehrenstorfer GmbH (Augsburg, Germany). The 4-bromo-2-fluorobiphenyl and deuterated terphenyl (Accustandard, USA) were used as surrogate standards. D₈-NaP, D₁₀-Ace D₁₀-Phe, D₁₂-Chr, and D₁₂-Pyr supplied by Dr. Ehrenstorfer GmbH (Augsburg, Germany) were used as internal standards in analysis process. The splitless injection was adopted and Helium was used as the carrier gas with the constant flow keeping at 1.1 mL min^–1. The oven temperature program was as follows: 70°C held for 4 min, ramped to 300°C at 10°C min^–1 and held for 2 min, and then ramped to 340°C at 5°C min^–1.

The limited m/z values were 129, 127 for NaP, 153, 152 for Acy, 151, 153 for Ace, 165, 167 for Fl, 179, 176 for Phe and Ant, 101, 203 for Flu and Pyr, 229, 226 for BaA, 226, 229 for Chr, 256, 126 for BbF and BkF, 253, 126 for BaP and BeP, 138, 227 for Ind, 139, 279 for DBA, 138, 227 for BghiP, 150, 301 for COR, respectively. The m/z values as 128, 154, 152, 166, 178, 178, 202, 202, 228, 228, 252, 252, 252, 252, 276, 278, 276, 300 were selected to quantify the NaP, Acy, Ace, Fl, Phe, Ant, Flu, Pyr, BaA, Chr, BbF, BkF, BaP, BeP, Ind, DBA, BghiP and COR, respectively. The correspondingly limited m/z values and quantitative m/z values (in bracket) were 68 and137 (136) for D₈-NaP, 162 and 165 (164) for D₁₀-Ace, 94 and 189 (188) for D₁₀-Phe, 120 and 241 (240) for D₁₂-Chr, and 260 and 265 (264) for D₁₂-Pyr, respectively. The surrogate standards 4-bromo-2-fluorobiphenyl and deuterated terphenyl possessed the corresponding values as 170 and 151 (250), and 122 (244).

The PAH pre-treatment and analysis were entirely conducted according to the strict quality assurance (QA) and quality control (QC) procedures. The experiments associated with sample and procedure blank, sample duplication, and recoveries of matrix spiked and internal standards were carried on schedule every 6 samples. Sample and procedure blank experiments showed that no target chemicals were detected. The recovery rates ranged from 72.5 to 121% for 4-bromo-2-fluorobiphenyl and from 75.3 to 122% for deuterated terphenyl, and the mean values were 96.3% and 90.9%, respectively. The 18 PAHs in 3 matrix injected samples possessed the recovery fluctuated from 72.0% for Acy to 118% for Pyr with the mean value as 93.5 ± 9.96% and the mass concentrations of 16 PAHs corresponding to triple signal to noise ratio were identified as the method detection limits (MDLs) of them. The corresponding MDLs for 16 PAHs (in ng g^–1) ranged from 0.06 for BaP to 0.44 for Fl with a mean as 0.18 ± 0.11.

Analysis of PM_2.5 Associated Inorganic Elements

Among the 26 inorganic elements (IEs), 19 IEs including Li, Be, Na, P, K, Sc, V, Cr, Mn, Co, Ni, Cu, Zn, As, Cd, Sn, Sb, Tl, and Pb were analyzed by ICP-MS (Agilent 7500a, Agilent Co. USA), and the rest 9 IEs such as Si, Al, Ca, Mg, Fe, Ba, and Sr were analyzed by ICP-OES system (Agilent 5100, Agilent Co. USA). The same analyzed method was detailed documented by Li et al. (2017a, b) and simplified as follows. For ICP-MS, the 1 Teflon filter was heated with 5 mL of aqua regia (V:V = 1:5) and 1 drop of HF acid (V:V = 1:1) at 120℃ for 2 h, and then raised to 130℃ until dryness. Then the HCl acid was added and heated for 20 mins, and the extracts were transferred into a plastic comparison tubes before analysis. For ICP-OES analysis, the other 1 Teflon filter was cut into pieces and heated at 300℃ for 40 mins, and gradually raised to 530–550℃ for completely ashing. Several drops of absolute ethanol and 0.1–0.2 g of NaOH were added and heated at 500℃ for 10 mins, and the water was added and boiled. Finally the extract combined with 2 mL HCl and diluted to 10 mL before analysis.

The method detection limits (MDLs) for 26 IEs were ranged from 0.10 ng of Be to 1810 ng of Fe. The relative standard deviation (RSD) values for 26 elements were 2.12%, 2.51%, 2.57%, 3.12%, 1.77%, 2.58%, 3.01%, 4.55%, 5.80%, 2.43%, 2.23%, 3.22%, 4.28%, 1.72%, 2.58%, 3.02%, 4.28%, 5.15%, 3.58%, 2.12%, 2.57%, 3.25%, 4.43%, 4.59%, 6.15%, and 2.77%, respectively. The standard soil materials GBW07446-GBW07457 (Center for National Standard Matter, China) were used to determine the accuracy of ICP-MS and ICP-OES.

RESULTS AND DISCUSSION

Contents of PM_2.5 Associated PAHs

The PM_2.5 fractions of the ashes originated from five combustion sources differed significantly from each other in their Σ₁₆PAHs concentrations. GFPP possessed the highest Σ₁₆PAHs as 66500 ± 9800 ng g^–1, while the minimum occurred at DG and it was 2480 ng g^–1. The Σ₁₆PAHs values (in ng g^–1) of rest 3 combustion sources conformed to the order WS (13600 ± 1980) > WCC (9890 ± 1280) > PS (3570 ± 852). All the five combustion sources possessed much higher levels of PAHs than those of 12 wood pellets burning (0.064 – 0.90 mg kg^–1) (Orecchio et al., 2016). For biomass fuels, the Σ₁₆PAHs values for PM_2.5 from PS- and WS-burning were significantly higher than 65.0 – 1310 ng g^–1 for the five-sized parts of bottom ashes from 8 kinds of bio-fuels in Beijing-Tianjin-Hebei region, China (Li et al., 2018a). Compared with the other similar studies, the PM_2.5 bounded PAHs from the GFPP in this study were much higher than 21.9 ng g^–1 of fly ashes from the 16 Chinese coal-fired power plants with the individual unit as 600 MW (Li et al., 2016), 29.8–63.8 ng g^–1 for bottom ashes and 3860–148991 ng g^–1 for fly ashes from 48 biomass-fired power plants in Czech Republic (Zdeněk et al., 2016), 2830–5320 ng g^–1 for fly ashes and 930–2080 ng g^–1 for fly ashes from a Chinese coal-fired plant with the individual unit as 300 MW (Wang et al., 2013), 1200–5900 ng g^–1 for bottom ashes from the municipal solid waste (MSW) burning (Peng et al., 2016a), 1970–3710 ng g^–1 for bottom ashes from the hydrothermally treated MSW burning (Peng et al., 2016b), 13540 ng g^–1 for fly ash formed during the burning process of pressurized coal (Zhou et al., 2009), and 31–2160 ng g^–1 for fly ashes from 4 bio-fueled power plants in India (Masto et al., 2015). For biomass combustion process, the WS and PS possessed higher values than those of the bio-fueled power plants in Czech Republic (Zdeněk et al., 2016) and India (Masto et al., 2015), which possibly owning to the more complete combustion of biomass in the industrial boilers (Li et al., 2016, 2017b). The bottom ashes possessed higher TOC values compared than those of fly ashes, so the PAHs in bottom ashes were higher than those of fly ashes. Li et al. (2017b) reported that the sum of 18 PAHs in bottom ashes from burning of 8 types of bio-fuels were in the range of 65–1310 ng g^–1. The PAHs preferred to enrich in finer particles during the combustion process (Kong et al., 2011; Masto et al., 2015; Li et al., 2016, 2017b). In addition to the combustion completeness and ash particle size, fuel type was another important influence factor on PAH emissions (Kong et al., 2011; Masto et al., 2015; Li et al., 2016, 2017b).

Fig. 2 showed the PAH ring-size distributions in PM_2.5 fractions within 5 kinds of ashes. The PAHs were divided to three classes as low molecular weight (LMW) PAHs, medium molecular weight (MMW) PAHs, and high molecular weight (HMW) PAHs based on their molecular weight (Kong et al., 2011; Li et al., 2016, 2018a). The 2- and 3-ring compounds were identified as LMW-PAHs, 4-ring compounds were contained in medium molecular weight (MMW) PAHs, and 5-, 6-, 7-ring compounds were included in HMW PAHs. The contributions of differently sized PAHs varied significantly among the 5 combustion sources. The LMW PAHs dominated in PM_2.5 from burning of the biomass including PS and WS and contributed 82.5 ± 8.56% and 91.6 ± 9.78% to the total PAHs in them. The similar results were found for the bottom ashes from combustion of 6 types of straws in Beijing-Tianjin-Hebei region, and the LMW-PAHs accounted for 78.3%, 80.2%, 92.5%, 69.3%, 96.4%, and 96.7% to the total PAH concentrations in bottom ashes from burning of the straws of soybean, millet, walnut, cotton, corn, and corn cob, respectively (Li et al., 2018a). Also the LMW-PAHs contributed 86.9% to the total PAHs in fly ashes from 4 Indian biomass-fired power plants (Masto et al., 2015), while relatively low contributions of 32.7% and 44.5% were reported for the phytomass- and dendromass-fueled power plants in Czech Republic (Košnář et al., 2016). However, HWM-PAHs dominated in PM_2.5 from GFPP and contributed 52.6 ± 5.11% to the total PAHs, and the LMW-PAHs only contributed 15.3 ± 2.86% to the total PAHs. For DG and WCC, the LMW- and MMW-PAHs occupied a considerable proportion in the total PAHs.

Fig. 2. Ring size distribution and concentration of the PM_2.5 bounded PAHs from 5 combustion sources.

PAH Composition Profiles in PM_2.5 from Five Combustion Sources

Fig. 3 listed the contents of individual PAH congener in PM_2.5 for all the 5 burning sources. The 16 PAH congeners were detectable among all the PM_2.5 samples in this study. The mean mass concentrations (in ng g^–1) of the top 6 PAHs followed the order of BaP (10600) > Chr (7640) > BkF (5970) > Ind (5740) > BghiP (5730)> Phe (5480) for GFPP, while they were NaP (2200) > Phe (2070) > Flu (920) > Pyr (705) > BbF (634) > BaP (492) for WCC, Phe (500) > NaP (275) > BbF (266) > BaP (250) > InD (189) > Flu (175) for DG, NaP (2140) > Phe (657) > Flu (157) > BbF (120) > Pyr (90.4) > BaP (57.8) for PS, and NaP (9810) > Phe (2280) > Flu (487) > Pyr (262) > Ant (187) > Chr (127) for WS, respectively.

Fig. 3. Composition profiles of PAHs for five combustion sources.

Different PAH profiles among different combustion sources were documented elsewhere. The indoor burning of biomass in India caused the emissions of top PAHs such as Ant, Flu, BaA, and Chr (Singh et al., 2013). Masto et al. (2005) indicated that the predominant PAHs NaP, Phe, BbF, and Flu were found in the bottom ashes from 38 Indian biomass-fired power plants, which was similar to those of PS and WS in this study. Orecchio et al. (2016) reported the top PAHs for wood pellet burning for heating in Italy, which were NaP, Phe, and Ant for burning of fir, NaP, Phe, and Flu for burning of the mixture of fir and beech, and NaP, Phe, and Acy for conifers, respectively. In this study, the burned cable wrapping in the metal reclamation workshops contained polyvinyl chloride (PVC), polyethylene (PE), polyperfluoroethylene propylene (PPFEP), polyolefin (PO), chlorinated polyethylene (CPE), chloroprene rubber (CPR), silicon rubber (SR), and chlorosulfonated polyethylene (CSPE). The dominated PAHs derived from burning of the aforementioned mixed materials were NaP, Phe, Flu, Pyr, BbF, and BaP. The predominant PAHs in 12 target PAHs as NaP, BkF, and Ant for polystyrene, NaP, Ant, and Chr for PVC, NaP, Fl, and Phe for low density PE, NaP, Phe, and Fl for high density PE, and NaP, Ace, and Fl for polypropylene and polyethyleneterephthalate were reported by Valavanidis et al. (2008). The different plastic materials, target PAHs, and ash sizes between two studies brought about the difference of PAH profiles.

Fig. 4. Coefficients of divergence for the ten pairs of pollution sources.

A widely known parameter as coefficient of divergence (CD) documented elsewhere was applied to distinguish the similarity of PAH profiles between two emission sources (Kong et al., 2011; Li et al., 2016, 2017b, 2018b). CD value was calculated as follows:

where p is the number of the target compounds detected for each source, x_ij and x_ik are the normalized concentration of i^th PAH in source j and k. The CD value approaches zero indicated the profile similarity between two sources, while the CD value of one suggested the difference of PAH profiles (Wongphatarakul et al., 1998; Kong et al., 2011; Li et al., 2016, 2017b, 2018b). The critical CD value of 0.3 used in identification of source profile similarity was documented frequently in recent studies (Kong et al., 2011; Li et al., 2016, 2017b, 2018a, b).

Except for the lower CD value of WCC vs. DG (0.20 ± 0.05), the CD values for any two sources were higher than 0.3 and they were 0.70 ± 0.15, 0.78 ± 0.11, 0.55 ± 0.09, 0.51 ± 0.08, 0.48 ± 0.06, 0.46 ± 0.06, 0.53 ± 0.10, 0.71 ± 0.21, and 0.73 ± 0.19 for GFPP vs. PS, GFPP vs. WS, GFPP vs. WCC, GFPP vs. DG, PS vs. WS, PS vs. WCC, PS vs. DG, WS vs. WCC, and WCC vs. DG, respectively (Fig .4). The results showed that the PAH profiles for both WCC and DG were similar, while the rest 9 pairs of combustion sources were all different each other (Kong et al., 2011; Li et al., 2016, 2017b, 2018b). The similarity of PAH profiles between WCC and DG was possibly resulted from the similar common principle component of plastics between two fuels.

The difference of mass fractions for NaP and Phe in the total PAHs among the 5 kinds of sources was a common cause for the difference of PAH profiles for all the 9 pairs of sources (Fig. 5). Special cases existed between each pairs of sources, BbF and Pyr were the perpetrators for PAH profile difference between WS and PS, BkF, BghiP, and BaA were the contributors to the profile difference between GFPP and WS, and the profile difference between DG and WS were mainly attributed to BbF, BaP, and Ind.

Fig. 5. Values of coefficient of divergence of PAHs for four pairs of sources.

Potential Health Risk Assessment for PM_2.5 Bounded PAHs

Widely known parameters such as toxicity equivalent concentration (TEQ) based on 2,3,7,8-tetrachlorodibenzodioxin (TCDD), BaP-based equivalent contribution (BaPeq), total carcinogenic PAHs (C-PAHs), and BaP-based equivalent carcinogenic power (BaPE) (Kong et al., 2011; Cheruiyot et al., 2015; Li et al., 2016; Tiwari et al., 2017; Li et al., 2018a). Total toxicity equivalency (TEQ) based on TCDD for the 16 PAHs were calculated as TEQ = ƩPAH_i × TEF_i and the toxicity equivalent factors (TEFs) were same as reported values by Nisbet and LaGoy (1992), and they were 0.001, 0.001, 0.001, 0.001, 0.001, 0.01, 0.001, 0.001, 0.1, 0.01, 0.1, 0.1, 1, 0.1, 1, and 0.01 for NaP, Acy, Ace, Fl, Phe, Ant, Flu, Pyr, BaA, Chr, BbF, BkF, BaP, Ind, DBA, and BgP, respectively. The BaPeq was calculated by Eq. (2) and the relative BaP based TEF values were found in Liu et al. (2009).

where PAH_i,TEF is the BaP based TEF for the i^th individual PAH, PAH_i is the concentration for individual PAH in PM_2.5, PAH_i,TEF is the BaP based TEF for the i^th PAH and it is referred to Liu et al. (2009), PAH_BaP,TEF is equal to 1, which is the BaP TEF, and PAH_BaP is the BaP concentration of PM_2.5 associated BaP. BaPE (reported in ng g^–1) is calculated as BaA × 0.06 + B[b,k]F × 0.07 + BaP + DBA × 0.6 + Ind × 0.08 (Liu et al., 2009; Li et al., 2016, 2018a). C-PAHs is the sum of mass concentrations for all the 7 carcinogenic PAHs including BaA, Chr, BbF, BkF, BaP, Ind, and DBA.

The results related to BaPE, TEQ, BaPeq, and C-PAHs for PM_2.5 from 5 types of ashes were listed in Table 1. BaPeq contributions of each PAH were highly volatile among the five kinds of sources depend on the significant fluctuation of their concentrations and BaP-based toxicity equivalent factors, and the similar status was found in Li et al. (2016, 2018a). For GFPP, the top contributors of BaPeq were DBA, Ind, BbF, and BkF, while they were Flu, DBA, BbF, and Chr for WS, BbF, DBA, Flu, and Ind for PS, BbF, DBA, Ind, and Flu for DG, and DBA, BbF, Flu, and Ind for WCC. GFPP possessed the highest C-PAHs concentration of 41.4 µg g^–1, while the PS had the minimum value as 0.35 µg g^–1, and the C-PAHs values (in µg g^–1) of 2.58, 0.97, and 0.38 were attributed to WCC, DG, and WS, respectively. The BaP concentration as high as 10600 ng g^–1 for GFPP should be paid more attentions. The similar results were found between the TEQ concentrations based on TCDD associated TEFs and BaPeq values for the five sources. TEQ values (in ng g^–1) for the 5 sources followed the order GFPP (15500) > WCC (761) > DG (332) > WS (109) > PS (94.3), and they were GFPP (15400) > WCC (778) > DG (336) > WS (119) > PS (97.2) for BaPeq concentrations (in ng g^–1) of five sources. The GFPP possessed the highest values of TEQ and BaPeq due to its highest BaP concentrations combined with the high TEF values, while biomass burning including PS and WS resulted in the lowest ones. In regard to the BaPE (in ng g^–1), the same trend with both TEQ and BaPeq was found, which was GFPP (13600) > WCC (656) > DG (303) > WS (82.9) > PS (79.4).

Concentrations and Exposure Risks of Inorganic Elements

The total contents of 26 IEs (Σ₂₆IEs) fluctuated from 154000 µg g^–1 for DG to 483000 µg g^–1 for GFPP, and they were 409000 for PS, 367000 for WS, and 320000 for WCC, respectively. Fig. 6 listed the contents for each of 26 IEs for five sources. K dominated in PM_2.5 derived from PS- and WS-burning, which was similar to the reported result by Li et al. (2017b). K was always used as a marker for biomass burning (Duan et al., 2004). K contributed 8.02% and 14.7% to the PM_2.5 mass from PS and WS, and the obvious fluctuation between PS and WS was attributed to the difference of plant physiological characteristics and local soil characteristics (Duan et al., 2004; Wang et al., 2016).

Fig. 6. Mean content of the individual element for five combustion sources.

The IEs were often divided into three classes based on their contents: 1) IEs with high contents including Al, Si, Ca, K, Mg, Fe, P, and Ti; 2) IEs with medium contents such as Mn, Na, Zn, Cu, Sr, Zr, Rb, Pb, and Mo; and 3) IEs with lower contents containing Sc, Li, V, Cr, Ni, Co, Sb, As, Y, Cs, Bi, Tl, Th, Sn, Cd, La, Ce, Sm, W, and U (Li et al., 2017b). The top 9 IEs with higher contents were Si > K > Ca > Mg > Al > Na > Fe > P for PS, Si > K > Ca > Al > Fe > Na > P > Mg > Mn for WS, Ca > Si > Al > Cu > Fe > Mg > Ba > Pb > Zn for WCC, Ca > Si > Al > Fe > Na > K > Mg > P > Zn > Ba for GFPP, and Si > Ca > Al > Na > Fe > K > Mg > P > Ba for DG, respectively. PS and WS possessed higher K contents (80.2 and 147 mg g^–1) than the corresponding 2.98 mg g^–1 for WCC, 15.4 mg g^–1 for GFPP, and 7.14 mg g^–1 for DG, K namely as an important nutrient element of plants would be the explanation. The highest Cu of 15.8 mg g^–1 was found for WCC compared with the 0.05 mg g^–1 for PS, 0.03 mg g^–1 for WS, 0.89 mg g^–1 for GFPP, and 0.36 mg g^–1 for DG. Cu was the main component in wire, which caused the high level of Cu in PM_2.5 from WCC.

HMs and As were more concerned than the other elements in their seriously adverse health effects to people (Li et al., 2017a, b; Hu et al., 2019). In this study, the 10 HMs (e.g., Zn, Mn, Cu, Pb, Co, Ni, V, Cr, Cd, Sn) and As most concerned were all detected. The sum of these 11 inorganic elements (∑₁₁IEs) dominated in WCC due to its high contents of Cu, Zn, and Pb, which was 23.4 ± 2.88 mg g^–1. The ∑₁₁IEs (reported in mg g^–1) followed the order as GFPP (5.44 ± 0.65) > DG (1.16 ± 0.06) > WS (0.68 ± 0.09) > PS (0.63 ± 0.05). The individual HM and As varied significantly among different sources, they were Mn > Zn > Cu >V > Cr > Pb > Ni > Co > As > Sn > Cd for PS, Mn > Zn > As > Cu > Pb > Ni > Co > V > Cr > Sn > Cd for WS, Cu > Pb > Zn > Sn > Mn > As > Ni > Cr > V > Cd > Co for WCC, Zn > Cu > Mn > Pb > Cr > Sn > Ni > Co > V > Cd >As for GFPP, and Cu > Zn > Mn > Pb > Sn > Cr > V > Ni > Cd > Co > As for DG, respectively. The highest levels of Cu, Zn, Sn, and Pb were found in PM_2.5 from WCC.

Also the similarity of HM vs. As profiles between each pair of combustion sources were evaluated using the CD values. Except the lower CD of 0.21 than 0.3 was found for GFPP vs. DG, the other pairs of sources possessed the higher CD values than 0.3, which were 0.36 for PS vs. WS, 0.72 for PS vs. WCC, 0.55 for PS vs. GFPP, 0.51 for PS vs. DG, 0.76 for WS vs. WCC, 0.60 for WS vs. GFPP, 0.58 for WS vs. DG, 0.68 for WCC vs. GFPP, and 0.67 for WCC vs. DG. Aforementioned CD values indicated the HMs and As profiles for all the pairs of sources except for GFPP vs. DG differed from each other. Fig. 7 listed the CD values of HMs vs. As profiles for four pairs of combustion sources. In generally, the difference of mass fractions of Zn, Mn, Pb, As, and Cu weaken the similarity of HM vs. As profiles between two sources.

Fig. 7. Values of coefficient of divergence of HMs and As for four pairs of sources.

Enrichment Factors and Igeo Values of Heavy Metals and As

A widely known parameter namely as enrichment factor (EF) was used to identify the IE was derived from human activities or natural processes, and further access the extent of anthropogenic impact (Li et al., 2017a, b). The IE possessed EF close to 1 indicated its crustal origin, while EF higher than 10 implied the significantly anthropogenic influence. In this study, EFs were used to identify the influence of PM_2.5 associated IEs on local soil and air. EF was obtained by the following equation:

where C_n is the measured content of the target IE in PM_2.5, C_ref is the measured content of reference IE in PM_2.5, B_n is the background content of the target IE in local soil, and B_ref is the background content of reference IE in local soil. Al was always selected as reference element in related studies of China, and Al was also used as reference element in this study (Li et al., 2017a, b). The EF values for the individual IE varied among different sources or the same source (Table 2). K was more enriched in biomass fuel such as PS and WS, Sb, Cu, and Pb were more inclined to enrich in WCC, Sn was preferential enrichment in WCC and GFPP, and Cd intended to enrich in GFPP and DG. Sb and Cu in WCC possess higher EFs as 160 and 114, which suggested the significant influence of human activities.

Geo-accumulation index (Igeo) had been widely used in IEs related studies (Kong et al., 2011; Li et al., 2013; Li et al, 2017a, b), which was calculated as follows:

where C_n and B_n are measured content and soil background content of the target element, and the constant of 1.5 is used to evaluate the natural fluctuation of the target element and assess the slightly anthropogenic impact (Wei et al., 2009). The Igeo values of ≤ 0, 0–1, 1–2, 2–3, 3–4, 4–5, and > 5 indicate uncontaminated, uncontaminated to moderately contaminated, moderately contaminated, moderately to heavily contaminated, heavily contaminated, heavily to extremely contaminated, and extremely contaminated, respectively. Table 3 listed the Igeo values for 24 IEs in PM_2.5 generated from five combustion sources.

In regard to the more concerned HMs and As, Sb, Cu, Pb, Sn, and Cd in PM_2.5 from WCC owned the Igeo values as high as 9.66, 9.18, 7.07, 6.63, and 6.01, which indicated that these HMs were extremely contaminated. Zn and As from WCC also had high Igeo values as 4.98 and 3.00. For GFPP, Cd, Sn, Sb, and Cu possessed higher Igeo values than 5, which were 8.46, 6.57, 5.12, and 5.05 indicating extreme contamination for them. Zn and Pb from GFPP had high Igeo values as 4.96 and 4.17 implying they were heavy to extreme contamination. Cd from DG also was extreme contamination based on its high Igeo value of 6.12.

Exposure Risks of Heavy Metals and As for Five Combustion Sources

EPA Human Health Risk Assessment Guidance (2009) provided information and methods to assess the impact on human health of toxic substances released into environment. The same assumptions were adopted as Mejido in 2017 and briefly stated as follows: 1) The exposure population was restricted to local residents and divided into children (≤ 6 years) and adults (≥ 21 years); 2) Three possible exposure pathways were involved including direct inhalation, dermal adsorption, and ingestion through food and drink containing deposited particles; 3) The intake rates approach those developed for soil (Mejido et al., 2017). The daily intake by ingestion, dermal adsorption dose, and exposure concentration by inhalation labeled as CDI_ing (mg kg^–1 day^–1), DAD_derm (mg^–1 kg^–1 day^–1), and EC_inh (µg m^–3) are calculated as following expressions:

where,

C: the upper bound confidence limit (UCL) for the arithmetic average concentration of metals and As was used as C to obtain the estimate of “reasonable maximum exposure” (Zhang et al., 2015b); IngR: provided by U.S. EPA (1989) for the risk assessment of ingestion of soil/dust; BW: average body weight (kg); EF: exposure rate; ED: exposure duration (6 years for children and 24 years for adults) (U.S. EPA, 2014); AT: averaging time (70 years × 365 days year^–1 for carcinogens and ED × 365 days year^–1 for non-carcinogens); CF: conversation factor (10^–6 kg mg^–1); SA: surface area contacted by airborne particles; AF: adherence factor of skin; ABS: dermal adsorption factor; ET: exposure time (24 h day^–1); and AT_n is average time.

Risk posed by pollutants was classified into carcinogenic and non-carcinogenic effects (Hu et al., 2012).

The cancer risk (CR) is defined as “the incremental probability of an individual developing cancer over a lifetime as a result of exposure to the potential carcinogen” (U.S. EPA, 1989). CR posed by As and heavy metals via inhalation, dermal adsorption, and ingestion was calculated as Eqs. (8)–(10).

The non-carcinogenic risk (NCR) posed by As and heavy metals was calculated using following expressions.

where, SFo is oral slope factor ((mg kg·day^–1)^–1); GIABS is gastrointestinal adsorption factor; IUR is inhalation unit risk ((µg m^–3)^–1); RfDo is oral reference dose ((mg kg·day^–1)^–1); and RfCi is inhalation reference concentration (mg m^–3).

The integrated carcinogenic and non-carcinogenic exposure risks posed by PM_2.5 associated heavy metals and As were listed in Fig. 8. Owning to the lack of the data related to air PM_2.5 concentrations, only two exposure pathways including ingestion (ING) and dermal adsorption (DERM) were involved in this study.

Fig. 8. Integrated carcinogenic and non-carcinogenic exposure risks of PM_2.5 associated HMs and As.

For the carcinogenic risks (CRs) to adults posed by the ING of HMs and As, they were 2.58 × 10^–5 for WCC > 9.86× 10^–6 for GFPP > 6.81 × 10^–6 for WS > 2.58 × 10^–6 for PS > 2.34 × 10^–6 for DG. All the CRs for adults posed by the ING of HMs and As were lower than the acceptable level of 1 × 10^–4 designated by US EPA (U.S. EPA, 2011). The CRs for children posed by ING for 4 out of 5 combustion sources were all higher than the acceptable level of 1 × 10^–4, which were 1.05 × 10^–3 for WCC > 4.05 × 10^–4 for GFPP > 2.80 × 10^–4 for WS > 1.06 × 10^–4 for PS. The CRs to adults posed by DERM for five sources were all higher than those for children. The CRs for adults posed by DERM of PM_2.5 associated HMs and As from PS, WCC, and GFPP were all higher than the accepted level of 1 × 10^–4. WCC possessed the highest CRs for children and adults posed by ING, while the highest CRs posed by DERM were owned by GFPP.

The non-carcinogenic risks (NCRs) for children posed by both ING and DERM for all the five sources were much higher than those for adults. The NCRs for adults posed by both ING and DERM were all lower than safe level of 1, the similar trend occurred at NCRs for children through DERM, while the NCRs for children posed by ING were much higher than 1. WCC possessed the highest NCR (63.8) for children derived from ING, which were much higher than those found for GFPP (10.9), WS (4.11), DG (3.58), and PS (2.19).

Table 4 listed the elements with high exposure CRs. The element Cr from WCC and GFPP resulted in higher CRs than 1 × 10^–4 for children through ING and for adults through DERM. As from WS and WCC, and Pb from WCC contributed to higher CRs than 1 × 10^–4 for children through ING, while Cr from PS could lead to higher CR for adults through DERM.

Table 5 listed the elements with high exposure NCRs. All the NCRs for adults posed by individual IE exposure were lower than the safe level of 1. However, the NCRs for children posed by Co from GFPP, Cu from WCC, As from WS and WCC, Sb from WCC, GFPP, and DG, Tl from WCC and GFPP exceeded the threshold value of 1. Among the aforementioned elements, the maximum value of NCRs was owned by As from WCC (6.96), while the minimum value was attributed to Tl from WCC (1.05). The NCRs for children of the rest of them followed the order Cu from WCC (6.07) > Sb from WCC (4.73) > Sb from GFPP (2.13) > As from WS (2.13) > Tl from GFPP (1.17) > Sb from DG (1.08).

CONCLUSIONS

The burning residues were sampled from the biomass fuels produced in Hebei Province including the peanut straw (PS, n = 4) and wheat stalk (WS, n = 4), workshops of cable combustion (WCC) for metal reclamation in Hebei Province and Tianjin city (n = 3), domestic garbage for its reduction in Hebei Province (DG, n = 3), and garbage-fired power plants (GFPP, n = 4) in Hebei- and Shaanxi- Province during January to May, 2015.

GFPP possessed the highest ∑₁₆PAHs as 66500 ± 9800 ng g^–1, while the minimum occurred at DG and it is 2480 ng g^–1. BaP dominated in GFPP (10600 ng g^–1), which should be paid more attention by local government. TEQ values (in ng g^–1) for the 5 sources based on TCDD and BaP associated TEFs were GFPP (15500) > WCC (761) > DG (332) > WS (109) > PS (94.3) and GFPP (15400) > WCC (778) > DG (336) > WS (119) > PS (97.2). High contents of BaP in GFPP and DG combined with high TEF values might be the explanations. In regard to the BaPE (in ng g^–1), the same trend with both TEQ and BaPeq was found, which was GFPP (13600) > WCC (656) > DG (303) > WS (82.9) > PS (79.4).

The sum of 10 HMs and As (∑₁₁IEs) dominated in WCC resulted from its high contents of Cu, Zn, and Pb (23.4 ± 2.88 mg g^–1). The Σ₁₁IEs followed the order as GFPP (5.44 ± 0.65 mg g^–1) > DG (1.16 ± 0.06 mg g^–1) > WS (0.68 ± 0.09 mg g^–1) > PS (0.63 ± 0.05 mg g^–1). Only one lower CD value than 0.3 was found for GFPP vs. DG, the rest pairs of sources possessed higher CD values than 0.3 and they were 0.36 for PS vs. WS, 0.72 for PS vs. WCC, 0.55 for PS vs. GFPP, 0.51 for PS vs. DG, 0.76 for WS vs. WCC, 0.60 for WS vs. GFPP, 0.58 for WS vs. DG, 0.68 for WCC vs. GFPP, and 0.67 for WCC vs. DG.

According to the enrichment factors of IEs, K was more enriched in biomass fuel such as PS and WS, Sb, Cu, and Pb were more inclined to enrich in WCC, Sn was preferential enrichment in WCC and GFPP, and Cd intended to enrich in GFPP and DG. Sb and Cu in WCC possess higher EF values as 160 and 114, which suggested the significant influence of human activities. Sb, Cu, Pb, Sn, and Cd in PM_2.5 from WCC owned the Igeo values as high as 9.66, 9.18, 7.07, 6.63, and 6.01, Cd, Sn, Sb, and Cu from GFPP possessed Igeo values as 8.46, 6.57, 5.12, and 5.05, and Cd from DG also possessed high Igeo value as 6.12, which indicated these HMs were extremely contamination.

The carcinogenic risks (CRs) for adults posed by dermal adsorption (DERM) were higher than those for children, while the reverse trend was found for CRs posed by ingestion (ING). The CRs for adults posed by ING of HMs and As from all the 5 sources were lower than the safety limit of 1 × 10^–4, which were 2.58 × 10^–5 for WCC > 9.86 × 10^–6 for GFPP > 6.81 × 10^–6 for WS > 2.58 × 10^–6 for PS > 2.34 × 10^–6 for DG. However, the CRs for children posed by ING of HMs and As from 4 sources exceeded the acceptable level, and they were 1.05 × 10^–3 for WCC, 4.05 × 10^–4 for GFPP, 2.08 × 10^–4 for WS, and 1.06 × 10^–4 for PS, respectively. The non-carcinogenic risks (NCRs) for children posed by both ING and DERM for all the five sources were much higher than those for adults. Both NCRs for adults posed by ING and DERM, and NCRs for children through DERM were lower than the acceptable level of 1, while the NCRs for children by ING were much higher than 1. WCC possessed the highest NCR (63.8) for children derived from ING, which were much higher than those found for GFPP (10.9), WS (4.11), DG (3.58), and PS (2.19).

ACKNOWLEDGEMENTS

This study was supported by the Fundamental Research Funds for the Central Universities (2017MS142), the National Natural Science Foundation of China (21407048), and Prime Minister Fund (DQGG-05-13).