Taiyuan is in the northern part of China and on the eastern edge of the loess hilly region, and it belongs to the continental climate of the northern temperate zone. It is hot and rainy in the summer and cold and dry in the winter. The regional profile is bat-shaped, with hills surrounding the area in the west and east, while the southern area is a valley plain; the average elevation ranges from 800 to 900 m (Figure 1B). The annual mean temperature and precipitation are 9.5°C and 456 mm, respectively.

According to research on the topography of Taiyuan city (Liang et al., 2000), plains and mountains were classified by an elevation of 900 m with SRTM-DEM (https://engine.piesat.cn/dataset-list) data. Forty-nine topsoil samples (obtained at 0–10-cm depths) from the plain area and 21 topsoil samples (at 0–10-cm depths) from the mountainous area were collected in July 2021 using wooden spades (Figure 1b). All sites were at least 50 m away from pollution sources and main roads, without anthropocentric interference activities. At each site, five topsoil subsamples were collected within a 5 m × 5 m (20 m × 20 m) grid and mixed to yield a mixed soil sample of approximately 500-1000 g. Then, one-quarter of each soil sample was collected, stored in a polyethylene bag, transported back to the laboratory to remove debris such as gravel and root debris, air-dried at room temperature and screened with a 70-mesh sieve (Tao et al., 2011; Duan et al., 2015). The treatment samples were mixed and reserved in clean brown bottles (Duan et al., 2015).

Eight grams of each soil sample were weighed into a perfluoroalkoxy polymer extraction vessel with a Teflon-sealed, lip-tight cap and a polyetheretherketone liner. A microwave-accelerated system (MARS 240/50, CEM, United States) was used to extract PAHs from the soil samples using 20 mL of a hexane/acetone mixture (1:1, v/v) at 1200 W (100%). The extraction temperature was programmed to 110°C for 10 min and then held for an additional 10 min. After extraction, the soil and solvent were separated by filtration, and the solvent was decanted into a pear-shaped flask. The extracts were then concentrated to approximately 1 mL using a rotary vacuum evaporator (EYEL4, Japan) and transferred to a silica/alumina gel column (43 cm in length with 10 mm i.d.) filled (from bottom to top) with 12 cm silica gel, 12 cm alumina, and 1 cm anhydrous sodium sulfate for cleanup (Shen et al., 2012b). The column was first eluted with 20 mL hexane and then with 70 mL hexane/dichloromethane (1:1, v/v). The latter eluate was further concentrated to 1 mL and spiked with Nap-d8, Ant-d10, Ace-d10, Chr-d12, and Perylene-d12 (J&W Chemical Ltd., United States) for the chromatography analysis.

PAHs were measured using a gas chromatograph (GC 6890, Agilent, United States) connected to a mass spectrometer (MS5973, Agilent, United States) in electron ionization mode. An HP-5MS capillary column (30 m × 0.25 mm×0.25 μm) was used, and the oven temperature was held at 50°C for 1 min, increased to 150°C at a rate of 10°C/min, increased again to 240°C at a rate of 3°C/min, and then increased finally to 280°C for another 20 min. Helium was used as the carrier gas. Twenty-one parent PAHs were measured in this study, including naphthalene (NAP), acenaphthylene (ACY), acenaphthene (ACE), fluorene (FLO), phenanthrene (PHE), anthracene (ANT), fluoranthene (FLA), pyrene (PYR), retene (RET), cyclopenta [c,d]pyrene (CPP), benzo(a)anthracene (BaA), chrysene (CHR), benzo(b)fluoranthene (BbF), benzo(k)fluoranthene (BkF), benzo(a)pyrene (BaP), benzo(e)pyrene (BeP), perylene (PER), dibenz (a,h)anthracene (DahA), indeno (l, 2,3-cd)pyrene (IcdP), benzo (g,h,i)perylene (BghiP), and coronene (COR). The PAHs were identified by their retention time and qualifying ions from standards in selected ion-monitoring mode.

All glassware was cleaned in an ultrasonic cleaner and baked at 500°C for at least 10 h. The silica and alumina used were prebaked at 450°C for 4 h and reactivated at 130°C for 16 h prior to use. The solvents were purified by redistillation and checked for blanks. For each extraction, laboratory blanks were analyzed. Two field blanks (samplers transported to and from the field without significant exposure) were analyzed alongside the actual samples collected from different landforms.Procedural blanks were measured, and the PAHs concentration in the blanks were subtracted from the samples. Instrumental detection limits defined as 3 times the signal-to-noise ratio were established for the PAHs and ranged from 0.13 ng/g (ACY) to 0.92 ng/g (BghiP). Preliminary recovery tests showed that the recoveries of spiked PAHs standard ranged from 82% to 114%. The results of the reported PAHs were not corrected by the recoveries. The recoveries of the surrogates (2-fluoro-1,1-biphenyl and p-terphenyl-d14) added into 15 samples randomly selected before the extraction process were 81% ± 12% (64%–103% range) and 84% ± 25% (62%–137%) for 2-fluoro-1,1-biphenyl and p-terphenyl-d14, respectively. The statistical analysis was performed using SPSS with a significance level of 0.05. The spatial distributions of the PAHs contents were mapped with ArcGIS Pro software.

The estimated incremental lifetime cancer risk (ILCR) metric of the United States Environmental Protection Agency (USEPA) was used to assess the potential health risks caused by PAHs in the soils in Taiyuan. Ingestion, dermal, and inhalation ILCR pathways were calculated using the following equations: ILCR ingestion = CS × CSF ingestion × BW / 70 3 × IR ingestion × EF × ED BW × AT × 10 6 (1) ILCR dermal = CS × CSF dermal × BW / 70 3 × SA × AF × ABS × EF × ED BW × AT × 10 6 (2) ILCR inhalation = CS × CSF inhalation × BW / 70 3 × IR inhalation × EF × ED BW × AT × PEF (3) where CS is the ΣBaP_eq concentration (mg/kg) with 21 PAHs; the individual BaP_eq values were calculated by multiplying the measured component concentrations with the corresponding toxicity equivalent factors (TEFs, see Supplementary Table S7). CSF_ingestion, CSF_dermal, and CSF_inhalation are the carcinogenic slope factors of oral intake, skin contact and breath intake, with values of 7.3, 25, and 3.85, respectively (mg/kg/d) (Jia et al., 2021); BW is body weight (kg); AT is an average life (y); EF is the exposure frequency (d/y); ED is the exposure time (y); IR_inhalation is the inhalation rate (m³/d); IR_ingestion is the soil intake rate (mg/d); SA is the skin exposure area (cm²); AF is the skin adhesion factor (mg/(cm²∙h)); and ABS is the skin adsorption fraction. The parameters of the ILCR model are listed in Supplementary Table S6.

The soil sample concentrations of PAHs and the sums of the concentrations of 21 PAHs (Σ₂₁PAHs), including the concentrations of the 16 USEPA priority PAHs (Σ₁₆PAHs) and 7 carcinogenic PAHs (BaA, CHR, BbF, BkF, BaP, IcdP and DahA; Σ₇CarPAHs), are shown in Figure 2, Supplementary Table S1, S2. The Σ₂₁PAHs concentration in the plain soil samples ranged from 133.2 to 6,410.6 ng/g with a mean of 1,444.7 ng/g. The Σ₁₆PAHs concentration were 100.1–5,448.3 ng/g (mean 1,240.0 ng/g), accounting for approximately 85.8% of Σ₂₁PAHs. The Σ₇CarPAHs contributed approximately 43.9% of the Σ₁₆PAHs, with a mean of 544.8 ng/g (range 49.6–2,657.9 ng/g). The Σ₂₁PAHs concentration in the mountain soil samples ranged from 66.5 to 2,250.2 ng/g with a mean of 585.5 ng/g. The Σ₁₆PAHs concentration were 50.6–2065.8 ng/g (mean 500.3 ng/g ng/g), accounting for approximately 85.5% of Σ₂₁PAHs. The Σ₇CarPAHs contributed approximately 37.6% of the Σ₁₆PAHs, with a mean of 188.0 ng/g (range 12.8–803.2 ng/g). A past study indicated that human activities are the main source of PAHs (Abdel-Shafy and Mansour, 2016; Li et al., 2020), and as shown in Figure 1B, the landform of Taiyuan city is a basin, with low elevations in the middle region and high elevations in the east and west. Human activities and heavy pollution enterprises are distributed mainly in the middle plain area at low elevations. This may be the reason why the Σ₂₁PAHs in the plain soils were 2.47 times more concentrated than those in the mountainous soils and why the 7 carcinogenic PAHs in the plain soils were 3 times more concentrated than those in the mountain soils.

The PAHs concentration measured in plain and mountainous soils of other regions of China and other countries are shown in Supplementary Table S3. The concentrations of Σ₁₆PAHs in the plain soils of Taiyuan were higher than those in Nanjing (China) (Yang et al., 2017), Harbin (Ma et al., 2009), Shanghai (Wang et al., 2015), Beijing-Tianjin (China) (Wang et al., 2010), Kurukshedra (India) (Kumar et al., 2013) and Ulsan (South Korea) (Kwon and Choi, 2014). Moreover, the Σ₁₆PAHs concentration in the plain soils of Taiyuan were lower than those measured in Xi’an (China) (Wang et al., 2018), Chengdu (China) (Xing et al., 2011), Lisbon (Portugal) (Cachada et al., 2012) and Orlando (United States) (Liu et al., 2019).In addition, the Σ₁₆PAHs concentration measured in mountainous soils were relatively high in other study regions; those measured in this study were higher than those of Changbai Mountain (China) (Zhao et al., 2015), Wang Swan (China) (Wang et al., 2019), and the eastern edge of the Himalayas (India) (Devi et al., 2016) and lower than those of the Qinling Mountains (China) (Chen et al., 2022) and British-Norwegian woodland (UK) (Nam et al., 2008).Drawing from the information provided, Taiyuan is characterized by its intensive industrial activities and coal combustion, which are recognized as significant sources of PAH emissions (Wu et al., 2023). This contributes to elevated concentrations of PAHs in the soil. Conversely, regions exhibiting lower PAH levels might be benefiting from stringent environmental policies, reduced dependency on heavy industries, or more efficacious pollution control strategies, particularly in large urban centers. Furthermore, the mountainous vicinities encircling Taiyuan, being in close proximity to urban activities, exhibit PAH concentrations surpassing those in areas with purely scenic and less affected by human activities. Nonetheless, these concentrations are marginally lower than those observed in the peripheries of more heavily industrialized mountainous regions.

The concentrations of 21 individual PAHs in soil samples are shown in Supplementary Table S1, S2. In the plain area, FLA was dominant in Σ₂₁PAHs, accounting for 13.0%, followed by BbF (10.3%), ANT (10.2%), PHE (8.3%) and PYR (8.1%). ANT was dominant in the Σ₂₁PAHs of the mountainous soils, accounting for 13.0%, followed by FLA (11.9%), PHE (11.4%), BbF (9.2%), and PYR (7.1%). Therefore, FLA and ANT were the main PAHs, followed by PHE, BbF and PYR. A previous study (Duval and Friedlander, 1981) suggested that FLA, ANT, PHE and PYR are markers of coal combustion, while ANT and PHE are the main compounds of PAHs produced by coking, and the individual source of BbF is diesel-powered vehicles. From this information, the main sources of PAHs in Taiyuan may be coal combustion, followed by coke production and vehicle emissions.

The spatial distributions and triangle distributions by of PAHs composited with different numbers of rings in the plain and mountainous surface soils are shown in Figure 3. Among the soil samples collected in the plain area, the 4-ring and 5-ring PAHs were dominant in Σ₂₁PAHs, accounting for 35.2% and 29.7%, respectively, followed by 3-ring (23.9%) and 6-ring (8.9%) PAHs. In addition, 2-ring and 7-ring PAHs were present at lower concentrations, contributing 1.5% and 0.8% of the Σ₂₁PAHs, respectively. In the mountains, the 3-ring and 4-ring PAHs were dominant in Σ₂₁PAHs, accounting for 32.7% and 31.2%, respectively, followed by 5-ring (25.3%) and 6-ring (7.8%) PAHs. In addition, 2-ring and 7-ring PAHs were present at lower concentrations, with contents of 2% and 0.9% of the Σ₂₁PAHs, respectively.

Investigations have revealed that 2-ring PAHs predominantly originate from the volatilization of petroleum products, whereas 3-ring and 4-ring PAHs are primarily generated through incomplete combustion processes, including those of coal, coke, and biomass, with coal combustion being a significant contributor to the presence of 4-ring PAHs. Conversely, PAHs encompassing more than 5 rings are typically associated with vehicular exhaust emissions (Guo et al., 2015; Li and Duan, 2015). In the context of Taiyuan, the analysis indicates a preponderance of 4-ring PAHs, which constitute over 32% of the Σ21PAHs, suggesting coal combustion as the principal source of pollution. This pattern aligns with findings from other Chinese urban centers such as Changzhi (Liu et al., 2017), Xinzhou (Zhao et al., 2014), and Lanzhou (Wang et al., 2016), where similar sources have been identified.However, this trend diverges in cities like Xi’an (Zhou and Lu, 2017) and Nanjing (Yang et al., 2017), where the dominant PAHs were observed to be in the 2–3 ring and 5–6 ring categories, respectively. Additionally, in the plains surrounding Taiyuan, high molecular weight (HMW) PAHs (5–7 rings) were detected at the second highest concentrations, implicating vehicle emissions as a significant contributor. In contrast, in the mountainous regions, low molecular weight (LMW) PAHs, possibly from biomass combustion (2–3 rings), were found to be the second most prevalent, highlighting the variegated sources of PAH pollution across different topographies and underscoring the complex interplay of urban and natural influences on the distribution of PAHs in the environment.

The average concentration of the 21 PAHs in all soil samples was 1,444.7 ng/g with a standard deviation of 1,451.3 ng/g, indicating significant spatial differences in the distribution of PAHs. The spatial distributions of the Σ₂₁PAHs concentration obtained by the ordinary kriging interpolation method are shown in Figure 4. From Figures 1, 2, 4, the results demonstrated that the plain area was a hot spots for PAHs pollution. According to the (Maliszewska-Kordybach, 1996) pollution classification method that is widely used to estimate soil contamination by PAHs, PAH-contaminated soils were divided into four categories: uncontaminated (< 200 ng/g), weakly contaminated (200–600 ng/g), contaminated (600–1,000 ng/g), and heavily contaminated (> 1,000 ng/g). Accordingly, 96% of the soil in the plain was found to be contaminated by 16 priority PAHs at levels higher than 200 ng/g. Heavy contamination accounted for 40.8%, while weak contamination and contamination accounted for 40.8% and 14.3%, respectively. In the mountain, 38.1% of the sample soils were uncontaminated, and weak contamination, contamination and heavy contamination contributed 42.9%, 4.7% and 14.3%, respectively.

Analysis of Figure 1 revealed that the highest Σ21PAHs concentrations were predominantly detected in the plains within urban territories. Among these, notable pollution hotspots included sites P7 (6,410.6 ng/g), P39 (5,845.3 ng/g), P43 (4,457.1 ng/g), P31 (4,441.6 ng/g), P33 (3,530.7 ng/g), P49 (3,294.1 ng/g), P10 (3,248.5 ng/g), and P26 (3,217.4 ng/g), each recording Σ21PAHs concentrations exceeding 3,000 ng/g. Conversely, site P19 manifested the minimal pollution level at 133.2 ng/g. The spatial distribution of pollution, as depicted in Figures 1, 4, correlated with the clustering of major industrial facilities (such as thermal power stations, steel, and coking plants) and intensified human activities in the plain areas. Specifically, the highly contaminated sites P39 and P43 were situated to the southeast of the Taiyuan Iron and Steel Plant and the Taiyuan Second Thermal Power Plant, established in 1937 and 1956, respectively, signifying them as primary pollution emitters.Supporting data from Yang et al. (2017) and Supplementary Figure S1 indicated Taiyuan’s dominant northerly wind patterns, particularly during winter—a season marked by significant pollution and northwest winds. This meteorological factor, coupled with the sites’ proximity to pollution sources, likely exacerbated contamination levels at P39 and P43. Previous research (Li H. et al., 2016; Li et al., 2020) identified vehicular emissions as a significant contributor to Taiyuan’s PAH burden. This aligned with the elevated PAH concentrations at sites P31, P33, P26, and P49, which were proximal to high-traffic areas including the Taiyuan Railway Station, Liuxiang’s bustling commercial zones, Fenhe Park, and the North University of China, respectively. The dense human and vehicular activity in these locales offered a plausible explanation for the heightened soil PAH levels.The persistence and accumulation of non-volatile, decomposed, and degraded PAHs, particularly high molecular weight (HMW) PAHs, in soil could be attributed to their adsorption onto soil particles, with prolonged exposure leading to increased concentrations. This phenomenon might have accounted for the significant PAH levels at sites P7 and P10, situated near the historical Jinci Park. The long-standing pollution legacy from older industrial sites such as the Taiyuan First Power Plant (1953–2019), Taiyuan Chemical Fertilizer Plant (1960–2011), and Taiyuan Chemical Plant (1960–2011) (Wu and Yang, 1995) further compounded the issue. In contrast, the relatively low PAH levels at site P19 could be ascribed to its location within a newly established park, distanced from traditional pollution hotspots and with limited exposure duration to pollution sources, as construction commenced only in 2010 (Liu et al., 2010).

In the mountainous terrain, concentrations of Σ21PAHs at sites M6, M7, and M8 surpassed 1,000 ng/g, with the apex concentration observed at M8 (2,250.2 ng/g), succeeded by M6 (2,110.9 ng/g) and M7 (1,373.5 ng/g). The nadir was recorded at M2 with 66.5 ng/g, as depicted in Figure 2. According to Figure 1, Taiyuan’s mountainous expanse bifurcates into eastern and western sectors. The eastern sector exhibited lower pollution levels with a mean Σ21PAHs concentration of 303.2 ± 213.5 ng/g, indicating minimal pollution and spatial variance. Conversely, the western sector demonstrated elevated and variable concentrations, with a mean of 961.9 ± 782 ng/g. Notably, all Σ21PAHs concentrations exceeding 1,000 ng/g in the mountainous domain were situated in the northwestern quadrant of the western mountains. This sector has historically been a principal coal mining hub since 1933 (Qiao et al., 2011), which, alongside prolonged industrial and anthropogenic activities, likely contributed to the heightened soil contamination observed.

Spatial interpolation through ordinary kriging, as illustrated in Figure 4, pinpointed the central-northern region of Taiyuan as a primary locus for soil PAH pollution. Proximity to industrial zones or heavily trafficked areas corresponded with increased pollutant levels. Analysis of Figure 3A elucidated that in the plains, the Σ21PAHs profile was predominantly characterized by HMW PAHs, whereas the mountainous regions were chiefly influenced by LMW PAHs. Literature reviews (Lang et al., 2008), highlight that HMW PAHs tend to exhibit growth trends over time due to their low vapor pressure and propensity for accumulation. In contrast, LMW PAHs, primarily present in the gaseous phase, are more prone to photodegradation in soil environments. This may be the reason why HMW PAHs showed significant domination in the plain soil samples. Instead, the PAH concentrations in the mountainous soil samples dominated by LMW PAHs may have been partly caused by the local short-term emission of PAH characteristics in the surrounding countryside.

The Lifelong Incremental Cancer Risk (ILCR) model was used to assess the incremental cancer risks resulting from inhalation, ingestion and dermal exposure to PAHs, and the ILCR values for children and adults corresponding to the soils in the plain and mountainous regions were calculated using Eqs 1–3.

As shown in Figure 6 and Supplementary Table S8, the inhalation ILCR values of children and adults in the mountainous and plain regions were approximately 10⁻¹², much lower than the ingestion and dermal exposure values, which had the same order of magnitude of approximately 10⁻⁷, indicating that the inhalation of heavy particles through the mouth and nose was negligible in terms of the carcinogenic risk posed by soil-induced PAHs; this finding was similar to previously published results (Jia et al., 2021). As shown in Supplementary Table S8, in the plain soils, the ingestion and dermal ILCR values for children were 3.4 × 10⁻⁸–1.8 × 10⁻⁶ (mean 3.4 × 10⁻⁷) and 4.2 × 10⁻⁸–2.2 × 10⁻⁶ (mean 4.3 × 10⁻⁷), respectively. Adults’ ingestion and dermal ILCR values were 4.2 × 10⁻⁸–2.2 × 10⁻⁶ (mean 2.7 × 10⁻⁷) and 4.7 × 10⁻⁸–2.4 × 10⁻⁶ (mean 4.8 × 10⁻⁷), respectively. These results demonstrated that children had a 1.3 times greater ingestion ILCR and 0.9 times greater dermal ILCR than adults. According to previous studies (Zhang et al., 2020; Jia et al., 2021), the main reason for these differences is that children are more likely to ingest soils and dust through hand-matched activities than adults, and adults should surpass children in terms of their skin exposure area. The ILCR differences for the ingestion and dermal contact between children and adults in the mountainous region were similar to those in the plain region. However, in the mountainous region, the cancer risk value was lower than that in the plain area; children’s ingestion and dermal exposure were both 0.32 times greater than those in the plain area, and adults had ingestion and dermal exposure risks 0.32 and 0.31 times greater than those in the plain area, respectively.

According to the USEPA standard, ILCRs below 10⁶ represent a safe levels with negligible carcinogenicity, while ILCRs between 10⁶ and 10⁴ or greater than 10⁴ represent a potential and high carcinogenic risks, respectively. As shown in Supplementary Table S8, the mean total ILCR values were 7.7 × 10⁻⁷ and 7.5 × 10⁻⁷ for children and adults in the plain region, respectively. The values in the mountains were 2.5 × 10⁻⁷ and 2.4 × 10⁻⁷, respectively. These results indicate that the cancer risks of PAHs in Taiyuan were at a safe level, with children more likely to have higher exposure risks than adults, and those in the plain region were more easily exposed to PAH risk than those in the mountains. However, some samples showed potential risk points at all sites except point M8 located in the plain area, such as sites P7, P39, P33, P43, P31, P49, P26, and P10. According to the previous analysis, these sites were mainly located near industrial areas, commercial areas, urban parks and coal mining areas. The frequent human activities near these sites lead to the high PAH concentrations in these areas, especially of carcinogenic PAHs such as BaP and DahA.