Levels and congener pattern of PAHs
The concentrations of 16 individual PAHs, n-rings PAHs, 7 carcinogenic PAHs (Σ7-carPAHs) and total 16 PAHs (Σ16PAHs) in soils from Jizhou district and Ninghe district in 2001 and 2018 are summarized in Table S1 and Table S2. For Jizhou district in 2001, the Σ16PAHs concentrations ranged from 241 to 777 ng g-1, with a mean value of 447 ng g-1. The concentrations of Σ7-carPAHs were in the range of 58.4 to 265 ng g-1, making up 38.1% of the Σ16PAHs concentrations. Among, 2- and 5-ring PAHs were the predominated congeners with contributions of 37.0% and 28.6%, respectively, followed by 3- (13.2%), 4- (12.0%), and 6-ring PAHs (9.2%). LMW PAHs contributed about 50% to the total PAHs. Among the detected PAHs, NAP, BbF, BkF, and BaP were the major compounds in soils, comprising 37.0%, 9.3%, 8.9%, and 8.9% of the total PAHs, respectively. In 2018, the concentrations of Σ16PAHs in soils from Jizhou district varied from 113 to 1064 ng g-1, with a mean of 711 ng g-1. The Σ7-carPAHs comprised 24.5% of the total PAHs, varying from 25.3 to 364 ng g-1. The relative composition of PAHs ranked from high to low was as follows: 3-ring (68.7%) > 5-ring (20.8%) > 4-ring (5.7%) > 6-ring (3.6%) > 2-ring (1.3%). The ratio of LMW PAHs increased appreciably from 50% to 70%. ANT and DahA were the dominated compounds, accounting for 59.1% and 16.5% of the total PAHs.
As to Ninghe district in 2001, for Σ16PAHs and Σ7-carPAHs in soils, the concentration ranges were 240 to 2410 ng g-1 and 94.1 to 749 ng g-1, and their mean values were 624 ng g-1 and 219 ng g-1, respectively. The Σ7-carPAHs accounted for 36.1% of the Σ16PAHs. The relative composition of PAHs decreased in the following order: 2-ring (36.7%) > 5-ring (25.1%) > 4-ring (14.4%) > 3-ring (14.0%) > 6-ring (9.7%), and the percentage of LMW PAHs was about 50%. In terms of the individual PAHs, NAP, BbF, PHE, BkF, and BaP were regarded as the predominated compounds, with the contributions of 36.7%, 8.1%, 8.0%, 7.8% and 7.8%, respectively. In 2018, the Σ16PAHs concentrations ranged from 202 to 1964 ng g-1 (mean 718 ng g-1), and the Σ7-carPAHs concentrations were in range of 36.4 to 1496 ng g-1 and contributed to 28.8% of the total PAHs. 2-, 3-, 4-, 5- and 6-ring PAHs accounted for 2.5%, 62.7%, 5.8%, 25.0% and 4.0%, respectively. The ratio of LMW PAHs increased to 65%. In addition, ANT and DahA were the major compounds, accounting for 47.2% and 15.9% of the total PAHs.
In summary, the temporal changes in concentrations and compositions of 16 PAHs for Jizhou district and Ninghe district between 2001 and 2018 were obvious and similar. Over this period, the concentrations of Σ16PAHs of both districts increased, while the Σ7-carPAHs concentrations decreased or increased slightly. Moreover, the percentage of 2-ring PAHs fell from its peak to its trough, and the percentage of 4-ring PAHs also decreased, but the percentage of 3-ring PAHs increased significantly. The ratio of LMW PAHs increased appreciably. ANT and DahA, instead of NAP, BbF, BkF, and BaP, became the dominated compounds. The concentrations of ANT increased dramatically, while NAP and the HMW PAHs, except for DahA and BaP, decreased substantially. Although PAHs are characterized persistently, their removal from soils attributed to photolysis and chemical degradation, biodegradation and volatilization cannot be neglected (Haritash and Kaushik, 2009). Thus, previous studies found that the concentrations of PAHs decreased substantially after burying in soils for a long time (Cao et al., 2017; Choi, 2014). Moreover, PAHs with different rings present different environmental behaviors in soils (Wild and Jones, 1995). HMW PAHs tend to bury in soils for a longer period of time (Wang et al., 2011), and the degradation rate decreases with increasing molecular weight (Sun et al., 2014). LMW PAHs are more readily to be photodegraded and biodegraded (Jones et al., 1989; Wilcke et al., 2005; Wild and Jones, 1995). Therefore, the increase in Σ16PAHs concentration and LMW PAHs ratio in soils after nearly 20 years indicate continuing discharges of PAHs in both Jizhou district and Ninghe district. Besides, as shown in Fig. S1, in Jizhou district and Ninghe district, the percentages of soil samples contaminated and heavily contaminated by PAHs were notably higher (> 60%) in 2018 than those (< 25%) in 2001. The percentages of heavily contaminated soil samples in Jizhou district and Ninghe district increased from 0% to 14% and from 7% to 17%, respectively. This dramatically increased trend suggests that soil PAH pollution in the studied areas need to be concerned.
In addition, it is worth noting that, from 2001 to 2018, the mean Σ16PAHs concentration in soils from Jizhou district increased by 60 %, while that from Ninghe district only rose by 15 %. The major reason may be the different geographical locations. Ninghe district is adjacent to Tianjin coastal new region, where there has assembled lots of industrial activities such as coking plants, coal-fired power plants, steel mills, oil exploiting and refinery, and petrochemical in the past few decades. Tianjin coastal new region could be a center of PAH emissions because industrial activities are considered to be the major PAH sources (Manariotis et al., 2011; Wolska et al., 2012; Zhang et al., 2019). Transfer from air to soil is the most important interface process for PAH flux in Tianjin, resulting in very high flux around the industrial area (Tao et al., 2003, 2004). Additionally, more than 90% of the PAHs entering into surface soils attribute to the dry and wet deposition (Wild and Jones, 1995). Therefore, the constant PAH contamination from Tianjin coastal new region may be the main reason for keeping soil PAHs in Ninghe district at a relatively high level. However, Jizhou district is located in the outer suburbs of Tianjin, far away from the areas with severe contamination of PAHs. Many studies in recent years have argued that various and scattered sources, such as industrial facility, traffic, cooking, and space heating in a city, contribute to the regional baseline deposition of PAHs, which is also influenced by the level of economic development and the degree of urbanization and industrialization (Peng et al., 2011, 2013, 2016a; Wang et al., 2015, 2018; Zhang and Chen, 2017). From 2004 to 2016, the number of Jizhou GDP increased by 500%, and the gross product of construction industry in this district increased from 1.5 billion yuan in 2008 to 9.1 billion yuan in 2016. Thus, the relatively larger growth of PAH level in soils from Jizhou district was in part attributed to the local increasingly rapid economic development and urbanization.
Compared with the results reported in worldwide other researches, the concentrations of total PAHs in soils from Jizhou district and Ninghe district in 2018 were at middle level. As shown in Table S3, the Σ16PAHs concentrations were far lower than those in urban and rural soils from Chengdu, China (3110 ng g-1) (Zheng et al., 2018), Bergen, Norway (6780 ng g-1) (Haugland et al., 2008), and Glasgow, UK (11900 ng g-1) (Morillo et al., 2007), and comparable to those in urban soils from Changshu, China (641 ng g-1) (Cao et al., 2017) and Torino, Italy (857 ng g-1) (Morillo et al., 2007), as well as agricultural soils from Delhi, India (830 ng g-1) (Agarwal et al., 2009) and Czech Republic (847 ng g-1) (Holoubek et al., 2009). However, they were also substantially higher than those in urban, rural and agricultural soils collected globally, such as Hong Kong (54.6 ng g-1) (Zhang et al., 2006), Shanghai (360 ng g-1) (Jia et al., 2017) and Xinzhou in China (202 ng g-1) (Zhao et al., 2014), Korea (236 ng g-1) (Nam et al., 2003), and Poland (252 ng g-1) (Maliszewska-Kordybach et al., 2009).
Source apportionment using molecular diagnostic ratios
PAH molecular diagnostic ratios are used to qualitatively distinguish emission sources of PAHs in many studies (Chen et al., 2018; Qi et al., 2019; Tobiszewski and Namieśnik, 2012; Wang et al., 2010; Xing et al., 2020; Yunker et al., 2002). Four diagnostic ratios were used to perform PAH source apportionment in this study (Fig. 2). In 2001, for all of samples from Jizhou district, the values of FLA/(PYR + FLA) were greater than 0.5, for 63.2% of samples, ANT/(ANT + PHE) values were greater than 0.1, for 57.9% of samples, IcdP/(IcdP + BghiP) values were greater than 0.5, and for 68.4% of samples, BaA/(BaA + CHR) values were in the range of 0.2-0.35, indicating that coal and biomass combustion is the primary source of PAHs. Besides, 31.6% and 36.8% of samples having BaA/(BaA + CHR) and ANT/(ANT + PHE) ratio less than 0.2 and 0.1 imply that petroleum is, to a certain degree, an important contributed source. As for Ninghe district, the values of FLA/(PYR + FLA) were greater than 0.5 for 100% of samples, IcdP/(IcdP + BghiP) values were greater than 0.5 for 71.4% of samples, and ANT/(ANT + PHE) values were greater than 0.1 for 42.9% of samples, corresponding to the important contribution of coal and biomass combustion to PAH pollutions. High percent of samples having BaA/(BaA + CHR) and ANT/(ANT + PHE) ratios less than 0.2 and 0.1 suggests that petroleum is another emission source which cannot be ignored.
For samples in Jizhou district collected in 2018, 82.1% of FLA/(PYR + FLA) values were greater than 0.5, 100% of ANT/(ANT + PHE) values were greater than 0.1, 78.6% of IcdP/(IcdP + BghiP) values were greater than 0.5, and 53.6% of BaA/(BaA + CHR) values were greater than 0.35, indicating that coal and biomass combustion is a major PAH source. As for Ninghe district, 82.8% of FLA/(PYR + FLA) values were greater than 0.5, 96.6% of ANT/(ANT + PHE) values were greater than 0.1, 75.9% of IcdP/(IcdP + BghiP) values were greater than 0.5, and 50% of BaA/(BaA + CHR) values were greater than 0.35. These results suggest that biomass and coal combustion account for the main source of PAHs.
To sum up, the diagnostic ratios preliminarily reveal a change of the major PAH sources in these two districts during the urbanization and industrialization process from 2001 to 2018. Over this period, PAHs emissions from petrogenic origins decreased, while those from pyrogenic sources increased substantially.
Source identification by positive matrix factorization (PMF) analysis
Because of possible significant changes of molecular diagnostic ratios during the migration from sources to receptors due to different physicochemical properties of the paired species (Lang et al., 2008), some uncertainties could not be excluded only based on these ratios (Galarneau, 2008; Katsoyiannis and Breivik, 2014; Zhang et al., 2005). Therefore, USEPA PMF model 5.0 (USEPA, 2014) was introduced to further identify sources of soil PAHs and quantify their contributions in this study.
The quality of the data for each of PAHs used in the model is based on the signal-to-noise ratio and the percentage of samples with concentrations above MDLs (USEPA, 2014). Firstly, the number of soil samples in either Jizhou district or Ninghe district in 2001 was not sufficient for PMF to derive a convincing result. Secondly, with more than 50% of samples in the two districts having concentrations lower than MDLs in 2001, DahA and BghiP were classified as bad species and were excluded from PMF. Finally, soil PAHs in Jizhou district and Ninghe district showed a similar congener pattern in both 2001 and 2018. Hence, a dataset of 33 samples and 14 species and a dataset of 57 samples and 16 species were respectively introduced into the PMF model to identify soil PAH sources in 2001 and 2018. The model was set to run from 3 to 6 factors, and each run was initialized with different starting points. The PAH profiles in factors for 2001 and 2018 are shown in Fig. 3 and Fig. 4, respectively.
In 2001, as shown in Fig. 3, four appropriate factors were identified by the PMF analysis. Factor 1, explaining 32.1% of the total PAHs, was mainly characterized by NAP, ACE and FLO. NAP is a significant fraction of crude oil and petroleum products (Chen et al., 2013; Dahle et al., 2003; Saha et al., 2009). In addition, the dominance of 2- and 3-ring PAHs is related to petrogenic sources (Hu et al., 2013a; Olajire et al., 2005; Wang et al., 2013b). Therefore, factor 1 indicates petroleum sources.
Factor 2 accounted for 22.1% of the total PAHs and was loaded dominantly by IcdP which usually associates with vehicular emissions (Fraser et al., 1997; Harrison et al., 1996; Motelay-Massei et al., 2005), corresponding to traffic sources. Factor 3 was dominated by ANT and responsible for 12.9% of the total PAHs. Given the fact that ANT is a marker of wood combustion source (Harrison et al., 1996), this factor is considered as biomass combustion.
Factor 4 contributed 32.9% of the total PAHs and had high loadings of PHE, FLA, PYR, CHR, and BaA. Coal combustion mainly produces PAHs dominated by 3- and 4-ring congeners (Li et al., 2012; Qu et al., 2020). FLA, PYR, BaA, CHR are typical markers for coal combustion (Duval and Friedlander, 1981; Harrison et al., 1996; Larsen and Baker, 2003; Simcik et al., 1999). Therefore, factor 4 indicates coal combustion.
As shown in Fig. 4, four appropriate factors were also identified in 2018 by the PMF analysis. Factor 1 was responsible for 18.7% of the total PAHs. This factor was predominated by ACY as well as FLA, and moderately weighted by DahA, BghiP and IcdP. ACY and FLA are predominated PAH compounds emitted by grass or wood combustion (Jenkins et al., 1996; Khalili et al., 1995; Ou et al., 2010). DahA, BghiP and IcdP have been identified as typical tracers of vehicular sources (Fraser et al., 1997; Harrison et al., 1996; Larsen and Baker, 2003; Motelay-Massei et al., 2005; Simcik et al., 1999). Thus, factor 1 is considered as petroleum and biomass combustion. The second factor accounted for 10.1% of the total PAHs and was dominated by FLO, PHE and CHR. FLO and PHE are sometimes used as indicators for coke oven sources (Duval and Friedlander, 1981; Simcik et al., 1999). Therefore, factor 2 can be closely related to coke production.
Factor 3, contributing 63.7% of the total PAHs, was loaded predominantly by NAP, ANT, BaP and DahA, and moderately weighted by ACE, FLO, ACY and BghiP. DahA, NAP, FLO, and ANT are the main tracers of diesel engines (Fraser et al., 1997; Khalili et al., 1995; Motelay-Massei et al., 2005), and ACE, ACY, FLO, BaP as well as BghiP are the main tracers of gasoline engines (Harrison et al., 1996; Khalili et al., 1995; Ravindra et al., 2008). Therefore, factor 3 indicates traffic sources.
Factor 4 contributed 7.6% of the total PAHs and had high loadings of PYR, BaA, BbF and BkF. BkF is a typical marker for industrial coal combustion (Brown and Brown, 2012). Moreover, coal combustion mainly produces PAHs dominated by 3- and 4-ring congeners (Li et al., 2012; Qu et al., 2020). Therefore, factor 4 is considered as coal combustion.
The changes for soil PAHs sources in Jizhou district and Ninghe district that happened within the past 20 years were identified according to the results calculated by PMF model. From 2001 to 2018, the contribution of coal combustion had decreased by three quarters. This is inconsistent with that obtained by molecular diagnostic ratios, which actually supports that the latter has some limitations on the source apportionment of PAHs (Galarneau, 2008; Katsoyiannis and Breivik, 2014; Zhang et al., 2005). The substantial reduction in the contribution of coal combustion may be related to the ban of coal burning for heating in winter. Furthermore, coal was the major source in the form of industrial energy in these two districts in the 2000s. However, clean energy sources, such as electricity or natural gas, have been widely expanded in China in recent years. Hence, changes in the energy structure during the industrialization process may be also responsible for this decrease. In contrast, a significant increase in the contribution of traffic sources (from 22.1% to 63.7%) led to much higher concentrations of ANT as well as DahA and their predominance, and a corresponding rise in soil Σ16PAHs concentrations of Jizhou district and Ninghe district in 2018. The dominance of traffic sources was likely attributed to a huge surge in car use resulted from the rapid development and expansion of urbanization as well as the sustained economic growth, such as the burgeoning local vigorous tourism in Jizhou district.
Health risk assessment
The incremental lifetime cancer risk (ILCR) model was introduced to assess age- and pathway-specific potential cancer risks of soil PAHs in Jizhou district and Ninghe district. A detailed description of this model is provided in the Supplementary Content. As shown in Table 1 and Table 2, the orders of magnitude of ILCRsinhalation for different ages varied in a range of 10−12 to 10−10, about 104 to 105 times lower than ILCRsingestion and ILCRsdermal. Hence, residents in Jizhou district and Ninghe district possibly exposed to soil PAHs via ingestion and dermal contact, while the exposure via inhalation was negligible.
In 2001, for children, adolescents and adults in Jizhou district, ILCRsingestion and ILCRsdermal ranged from 5.91 × 10-8 to 5.19 × 10-7 and from 1.42 × 10-7 to 8.84 × 10-7, while in Ninghe district, ILCRsingestion and ILCRsdermal for these three age groups were from 6.64 × 10-8 to 1.15 × 10-6 and from 1.59 × 10-7 to 1.95 × 10-6, except one sample with ILCR values greater than 10−6. ILCRsingestion and ILCRsdermal in Ninghe district were slightly higher than those in Jizhou district. Totally, the vast majority of samples in these two districts had ILCRs lower than the safe value of 1 × 10-6, indicating health risks are generally at a safe level.
In 2018, the ILCRs of all exposure pathways for residents were relatively higher than those in 2001. For residents in Jizhou district, ILCRsingestion and ILCRsdermal ranged from 5.90 × 10-8 to 1.72 × 10-6 and from 1.41 × 10-7 to 2.92 × 10-6, respectively. At 11 of the 28 samples, ILCRsdermal were higher than 10−6 for children, adolescents and adults, and ILCRsingestion were greater than 10−6 for children and adults. For Ninghe district, ILCRsingestion and ILCRsdermal were from 7.54 × 10-8 to 1.18 × 10-6 and from 1.80 × 10-7 to 2.01 × 10-6, respectively. For children and adolescents, ILCRsdermal were higher than 10−6 at 6 of the 29 samples, but for adults, ILCRsdermal were higher than 10−6 at 13 samples. It should be noted that there was only one sample with ILCRsingestion greater than 10−6 for children and adults. These results indicate higher cancer risks to residents in these two districts in 2018, with dermal contact as the chief exposure pathway. The increase in cancer risks from 2001 to 2018 occurred mainly because the alterations of PAH sources led to the predominance of DahA which has a highest TEF and therefore a strong carcinogenic potency equal to BaP. In view of the fact that DahA is considered as an indicator for the vehicular emission (Fraser et al., 1997; Motelay-Massei et al., 2005), a substitution of diesel and gasoline by clean energies in vehicles and an option of public transit over private cars should be advocated to reduce the health risks to local residents.