1 Introduction

Polychlorinated biphenyls (PCBs) are a group of synthetic organic compounds that have been widely used in industrial applications because of their desirable chemical properties, including stability, non-flammability and excellent electrical insulating capabilities. Initially, PCBs were primarily manufactured for use in electrical equipment, such as transformers and capacitors. Over time, however, their applications expanded to include hydraulic fluids, plasticizers, and additives in paints, adhesives and other products (Garrido et al. 2019; Ediagbonya et al. 2023a, b).

Despite their widespread use, PCBs were banned in many countries in the late twentieth century due to their environmental persistence and toxicological effects on both human health and wildlife (Liu et al. 2020). Structurally, PCBs are composed of two benzene rings with varying numbers of chlorine atoms attached at different positions, resulting in multiple congeners with distinct chemical and biological properties. One key characteristic of PCBs is their high lipophilicity, which enables them to accumulate in fatty tissues and persist in the environment for long periods (Muir and Howard 2021; Ediagbonya et al. 2023c).

The environmental impact of PCBs is particularly concerning because these compounds do not readily degrade, allowing them to remain in soils, sediments and living organisms for decades. Mining activities have been identified as a potential source of PCB contamination due to the disturbance of soil and improper handling of equipment and waste. The excavation and processing of mineral resources often lead to the release of pollutants into the environment. In mining sites, improper disposal of waste materials, leakage from storage facilities, and the use of contaminated equipment all contribute to PCB accumulation in soils and surrounding ecosystems (Guzmán et al. 2021; Ossai et al. 2023; Mohankumar et al. 2024).

Moreover, the disruption of soil and sediment layers during mining activities can accelerate the spread of these pollutants. The mining site in Ibadan, Nigeria, provides a suitable case study for assessing PCB contamination in soil-derived atmospheric dust. Ibadan, a major city in southwestern Nigeria, has a long history of mining activities that may have significantly contributed to environmental pollution.

Soil contamination in such regions poses potential risks to local ecosystems, agricultural productivity and public health. Understanding the extent of PCB contamination and its possible impacts is essential for designing effective environmental remediation strategies and ensuring community safety. Therefore, this study aims to evaluate the levels of PCBs in soil-derived atmospheric dust from a mining site in Ibadan and to assess the associated risks to both the environment and human health.

2 Materials and Methods

2.1 Study Area

In Ibadan, Nigeria, a mining site served as the study’s location. The capital of Oyo State, Ibadan, is one of the largest cities in West Africa and has a rich history of mining activities, particularly in the extraction of solid minerals. The specific mining site selected for this study is known for its extensive history of mineral extraction, including both legal and illegal mining operations. The area is characterized by a tropical climate with distinct wet and dry seasons, and the soil is primarily composed of lateritic and sandy loam types, which are typical of mining regions in this part of Nigeria (Adeleke et al. 2019).

The mining site has been subjected to various degrees of environmental degradation due to the long-term exploitation of mineral resources. This has resulted in significant disturbances to the natural landscape, leading to soil erosion, deforestation and potential contamination by hazardous substances such as polychlorinated biphenyls (PCBs) (Olalekan et al. 2021).

2.2 Possible Contamination Sources of PCBs

The possible sources of PCB contamination in the sampling sites can be attributed to a combination of mining-related, industrial, and anthropogenic activities within and around the Ibadan mining zone:

  1. 1.

    Mining Operations and Machinery Use

    Mining activities in the study area rely heavily on mechanical equipment, including excavators, crushers, and processing machines that require lubricants, hydraulic fluids and transformer oils. Historically, many of these fluids contained PCBs due to their excellent thermal stability and insulation properties. Leakages, spills, and improper handling of such fluids can directly introduce PCBs into the surrounding soil and water.

  2. 2.

    Improper Waste Disposal

    Mining operations generate significant solid and liquid wastes, including tailings and wastewater. In the absence of effective waste management systems, PCB-containing materials from electrical components, equipment maintenance and used oils are often discarded into open areas. Over time, these compounds seep into the soil, sediments and groundwater, leading to widespread contamination.

  3. 3.

    Legacy Industrial Sources

    Ibadan has a long-standing history of small-scale industries and electrical workshops near mining zones. Older electrical transformers, capacitors and other industrial equipment in these facilities are known sources of PCB leakage, especially when outdated components are disposed of improperly. Atmospheric deposition from these industrial sources can also lead to secondary contamination of the mining site.

  4. 4.

    Open Burning of Electrical and Electronic Waste

    In surrounding communities, informal recycling and open burning of e-waste to extract valuable metals are common practices. During such burning, PCBs present in plastics, wires and capacitors are released into the air and eventually deposited into the soil and water bodies near the mining sites, increasing environmental loads of these persistent pollutants.

  5. 5.

    Hydrological Transport

    During heavy rainfall in the wet season, PCBs from nearby industrial zones, residential areas and dumpsites can be washed into rivers, streams and sediments around the mining site. Given the area’s tropical climate and surface runoff patterns, hydrological pathways are an important contributor to PCB migration and accumulation in the sampled locations.

  6. 6.

    Atmospheric Deposition

    PCBs are semi-volatile and can undergo long-range atmospheric transport. Emissions from burning waste, industrial facilities and other off-site sources can volatilize PCBs into the atmosphere, which are then redeposited onto soil, sediments and water bodies at the mining site through dry or wet deposition.

The proximity of residential communities to the mining site raises significant concerns regarding direct and indirect exposure to PCBs, particularly through inhalation, ingestion of contaminated water, dermal contact with polluted soils and consumption of locally grown food crops. These potential pathways highlight the environmental and public health risks associated with persistent organic pollutants in the region (Fig. 1).

Fig. 1
Fig. 1
Full size image

Map showing the sampling locations at the mining site

2.3 Instruments Used

The detection and analysis of polychlorinated biphenyls (PCBs) in soil samples require precise and reliable instruments capable of identifying and quantifying these compounds at trace levels. The following instruments were employed in this study. Gas Chromatography–Mass Spectrometry (GC–MS) was the primary analytical tool used to detect and quantify PCBs in soil samples. GC–MS combines the separation capabilities of gas chromatography with the detection and identification capabilities of mass spectrometry. This instrument is particularly suited for analyzing complex environmental samples, such as soil, where PCBs may be present in low concentrations and alongside various other organic compounds (Liu et al. 2020). The GC–MS used in this study was equipped with a capillary column specifically designed for PCB analysis, allowing for high-resolution separation of PCB congeners. High-Performance Liquid Chromatography (HPLC) was employed as a complementary technique to GC–MS. HPLC is particularly useful for analyzing non-volatile PCB congeners that may not be adequately resolved by GC–MS. The HPLC system used in this study was equipped with a UV detector, which allowed for the detection of PCBs based on their specific absorption characteristics. HPLC provided an alternative method for confirming the presence of PCBs in the samples and for analyzing compounds that might degrade at the high temperatures used in gas chromatography (Garrido et al. 2019). The preparation of soil samples for analysis required a variety of laboratory equipment, including Soxhlet extractors for the extraction of PCBs from soil matrices, rotary evaporators for concentrating the extracts, and filtration systems to remove particulates. Precision balances were used for weighing soil samples, and glassware was carefully selected to avoid contamination and ensure the accuracy of the analytical results (Muir and Howard 2021). To ensure the reliability of the data, several quality control measures were implemented, including the use of certified reference materials (CRMs) and internal standards. These instruments were used to calibrate the GC–MS and HPLC systems and to validate the analytical methods. Regular maintenance and calibration of the instruments were performed according to the manufacturer’s recommendations to maintain the accuracy and precision of the results (Lee et al. 2020).

2.4 Sampling Methods

The accurate assessment of polychlorinated biphenyls (PCBs) in soil requires meticulous sampling methods and protocols to ensure that the collected samples are representative of the study area and free from contamination. This section details the procedures followed in collecting and handling soil samples from the mining site in Ibadan. The soil sampling was carried out using a stratified random sampling technique to capture the spatial variability of PCB contamination across the mining site. The site was divided into several strata based on factors such as topography, proximity to mining activities, and potential sources of PCB contamination. Within each stratum, random sampling points were selected using GPS coordinates to ensure unbiased sample collection (Prasad and Singh 2020). At each sampling point, twenty soil samples were collected from soil surface. The tools required for sample collection included a broom, dustpan, foil paper for sample storage, and labels for proper identification. First, the debris and surface materials were carefully cleared from the sampling area to ensure the integrity of the samples. The dust was then gently swept into foil papers. Each sample was labeled with a unique identifier, including the location and date of collection, and was stored in a cooler at 4 °C until analysis (Liu et al. 2020; Iwegbue et al. 2018). Upon arrival at the laboratory, dust samples were air-dried in a clean, dust-free environment to reduce moisture content, which can interfere with the extraction and analysis of PCBs. The dried samples were then sieved through a 2 mm mesh to remove large particles, such as stones and plant material. The homogenized samples were stored in glass containers until further processing (Garrido et al. 2019).

2.5 Instrumentation

AccuStandard supplied a standard solution with a 1000 ppm concentration that included 24 environmental PCBs components (Catalogue Number: H-QME-01) for the qualitative analysis of PCBs. Using a series of five-point serial dilution calibration standards that were created from the stock solution (1.00, 5.00, 10.00, 50.00, and 100.00 ppm), the GC–MS apparatus was adjusted. Before calibration, the mass spectrometer (MS) was auto-tuned to perfluorotributylamine (PFTBA) using predefined mass-to-charge ratio (m/z) values, such as 69, 219, and 502.This made sure that the instrument settings were ideal. The PCBs levels in the samples were evaluated using GC–MS running in Selective Ion Monitoring (SIM) and Scan modes to accomplish low-level target element identification. The analysis was conducted using a 5975C inert mass spectrometer (which features an electron-impact source and a triple-axis detector) and an Agilent 7820A gas chromatograph. To separate the chemicals, an HP-5 capillary column (30 m long by 0.32 mm wide by 0.25 m thick) coated with 5% phenyl methyl siloxane was employed. The carrier gas, helium, had an initial nominal pressure of 1.49 psi and an average velocity of 44.22 cm s−1. 1.48 mL min−1 was the continuous flow rate at which the gas was consumed. At 300 °C, a 1-L injection of the sample was performed in splitless mode. Gas saver mode was turned off, and the purge flow to split vent was set at 10.0 mL min−1 for 2 min, for a total flow of 16.67 mL min−1. After 3 min of maintaining an 80 °C temperature, the oven was programmed to rise to 320 °C for eight more minutes at a pace of 15 °C min−1. A 3-min solvent delay was included in the 32-min run duration. The mass spectrometer operated in electron-impact ionisation mode at 70 eV with the ion source temperature set at 230 °C, the quadrupole temperature at 150 °C, and the transfer line temperature at 300 °C. Following calibration and testing, the pertinent quantities of PCBs were discovered in the samples.

2.6 Extraction of Sample

Over the course of 24 h, 20 g of the sample was Soxhlet extracted using dichloro methane (DCM). A mixture of 2,4,5,6-tetrachloro-m xylene (TCmX) and decechlorobiphenyl (PCB 209) was added as a surrogate standard prior to each sample being extracted. To eliminate elemental sulphur, activated copper granules were added to the collection flasks. Following the concentration of the extract, the solvent phase was changed from DCM to hexane using a rotary evaporator.

2.7 Quality Assurance and Quality Control

Procedure blanks, matrix-spiked samples, and surrogate 13C12-labeled PCBs were used as quality control. In order to check for potential laboratory contamination, interference, and reagent purity, procedural blanks were used. The analytical procedure was employed to do the task without the utilisation of samples. There were no signs of any of the identified PCB congeners in the procedure blank samples. Recovery experiments were performed to confirm that the selected analytical method was effective. Before extraction, a subset of samples were treated with known PCB concentrations (standards), and the entire analytical process was repeated. The percentage recovery of the spiked samples ranged from 85.7 to 102.3%, while overall percentage recoveries varied between 83.6 and 94.9%. The limits of quantitation (LOQs) for the PCBs ranged from 0.6 to 3.0 ng g−1, and the limits of detection (LODs) were between 0.2 and 1.0 ng g−1 (Li et al. 2021; Qiu et al. 2023; Chen et al. 2024).

2.8 Health Impact

The toxicity equivalency factor (TEF) is employed to estimate the toxic potential of polychlorinated biphenyls (PCBs), particularly the dioxin-like congeners. This is achieved by multiplying the TEF values by the concentration of each dioxin-like PCB, thereby determining the toxicity equivalency (TEQ) (Van den Berg et al. 2006). In this study, health risks associated with PCB exposure were evaluated using the Incremental Lifetime Cancer Risk (ILCR) and Hazard Quotient (HQ), following the guidelines established by the United States Environmental Protection Agency (USEPA 1989, 2015). These assessments considered three primary exposure pathways: ingestion, inhalation, and dermal contact with contaminated soil-derived dust.

For ingestion, the incremental lifetime cancer risk (ILCR) was calculated based on the concentration of PCBs in soil (C), ingestion rate (IR), exposure duration (ED), exposure frequency (EF), conversion factor (CF), oral slope factor (SFO), averaging time (AT), and body weight (BW). Similarly, the ILCR for inhalation was computed using the inhalation rate (IhR) and the inhalation unit risk (IUR), while dermal exposure incorporated additional parameters such as skin surface area (SA), soil adherence factor (AF), dermal absorption factor (ABS), and gastrointestinal absorption factor (GIABS).

In this assessment, the IR was set at 200 mg day−1 for children and 100 mg day−1 for adults. The IhR was assumed to be 9.6 m3 day−1 for children and 20 m3 day−1 for adults, as reported by Klánová et al. (2008), Adesina et al. (2021), and Francisco et al. (2017). The exposure duration (ED) was taken as 52 years, reflecting the average life expectancy in Nigeria, while an exposure frequency (EF) of 350 days year−1 was used, excluding holidays. The conversion factor (CF) was set to 1 × 10−6, and the oral slope factor (SFO) was 2 (mg kg−1 day−1)−1. The averaging time (AT) for cancer risk was 25,550 days.

For inhalation, the inhalation unit risk (IUR) was 5.7 × 10−1 (µg m−3)−1, and the cutaneous absorption factor (ABS) was set at 0.1. The assumed body weights were 70 kg for adults and 15 kg for children. Additionally, the reference dose (RfD) for PCBs was 3.3 × 10⁻5 mg kg−1 day−1. For dermal exposure, the soil adherence factor (AF) was 0.2 mg cm−2 and the exposed skin surface area (SA) was considered to be 3300 cm2. These parameters collectively formed the basis for estimating both carcinogenic and non-carcinogenic health risks associated with PCB exposure in the study area.

2.9 Statistical Analysis

The study utilized both descriptive and inferential statistics for data analysis. Descriptive statistics included mean and standard deviation calculations to summarize the data. In order to compare the various samples (plant, water, and sediments), inferential statistics used Analysis of Variance (ANOVA). The Duncan Multiple Range Test was used as the post hoc test where significant differences were observed to separate the means. Pearson correlation was used to correlate physicochemical parameters and Total Petroleum Hydrocarbons (TPH). Source identification and apportionment were conducted using Principal Component Analysis (PCA) with varimax rotation. At p < 0.05, the significance level was established. Version 28.0 of the Statistical Package for Social Sciences (SPSS) for Windows was used for all analyses.

3 Results and Discussion

Table 1 presents the analysis of the mean concentrations of various PCB congeners across four different sampling locations (G1, G2, G3 and G4) within the mining site. The statistical parameters F and Sig provide insights into the variability and significance of these differences.

Table 1 Mean concentration (mg L−1) of PCBs across different sampling locations at the mining site
Full size table

For PCB3, the concentration ranges from 0.15 to 0.20 mg L−1 across the sites, with G2 showing the highest level. Despite this slight variation, the statistical analysis indicates that these differences are not significant (F = 1.634, Sig = 0.316). This relatively uniform distribution across the mining site suggests that PCB3 contamination might be widespread or influenced by consistent environmental conditions (Prasad and Singh 2020).

In contrast, PCB5 concentrations show notable variations, ranging from 1.12 mg L−1 at G2 to 4.40 mg L−1 at G1. The high F-value (660.985) and significant p-value (Sig = 0.000) confirm that these differences are statistically significant. The elevated levels at G1 could be linked to specific activities or improper waste disposal practices in the area, suggesting localized contamination (Garrido et al. 2019). This highlights the need for targeted remediation efforts, especially in high-concentration zones like G1 (Lee et al. 2020).

For PCB17, data is missing for G1, which might indicate non-detection or data exclusion. In other locations, concentrations remain low (0.01–0.03 mg L−1) with no significant differences (F = 0.905, Sig = 0.493). This suggests that PCB17 may be minimally present or uniformly distributed across the sites (Zhao et al. 2019).

PCB21 concentrations are notably higher at G2 (0.69 mg L−1) compared to other sites, with significant differences observed (F = 15.425, Sig = 0.012). The elevated levels at G2 could be attributed to specific industrial processes or accidental releases. This significant variation indicates that certain locations are more prone to PCB21 contamination, which may pose localized environmental and health risks (Muir and Howard 2021).

Regarding PCB47, concentrations remain very low across all sampling sites, and the absence of significant differences (F = 1.333, Sig = 0.385) suggests uniform distribution. This indicates that PCB47 may not be a major contaminant or is consistently present at low levels throughout the mining area (Liu et al. 2020).

PCB61 demonstrates a substantial variation in concentrations, ranging from 0.08 mg L−1 at G2 to 2.77 mg L−1 at G1. The extremely high F-value (1702.682) and significant p-value (Sig = 0.000) confirm pronounced differences between sites. Elevated PCB61 levels at G1 suggest a concentrated contamination source, possibly linked to mining activities or historical waste deposition (Garrido et al. 2019).

For PCB99, concentrations vary significantly, with G2 exhibiting an exceptionally high concentration of 18.60 mg L−1. The extremely high F-value (81,315.504) and significant p-value (Sig = 0.000) indicate a severe disparity in contamination levels. This suggests that G2 could be a potential hotspot for PCB99 accumulation, likely resulting from intensive mining or processing activities leading to localized pollution (Lee et al. 2020).

In the case of PCB98, data is missing for certain sites, and where available, concentrations remain low (~ 0.02 mg L−1). The absence of significant differences (F = 0.200, Sig = 0.698) suggests that PCB98 is either not a major contaminant or is uniformly distributed at minimal levels (Muir and Howard 2021).

Similarly, PCB111 concentrations show no significant differences across the sites (F = 4.212, Sig = 0.099), with levels consistently low (0.00–0.04 mg L−1). This uniformity could indicate widespread but low-level contamination without concentrated hotspots (Zhao et al. 2019).

PCB151 and PCB161 also exhibit low concentrations across all locations, with no statistically significant differences (PCB151: F = 1.000, Sig = 0.479; PCB161: F = 1.790, Sig = 0.288). These consistent low levels suggest that these congeners may not represent major contamination concerns or that their sources are evenly distributed across the mining site (Liu et al. 2020).

For PCB159, slight variations are observed, and although the statistical difference (F = 5.095, Sig = 0.075) is marginal, it may indicate localized variations in contamination possibly linked to specific site-related activities (Garrido et al. 2019).

Finally, PCB168 and PCB183 demonstrate uniform distribution across all sampling points (PCB168: F = 0.000, Sig = 1.000; PCB183: F = 2.667, Sig = 0.184). However, PCB187 and particularly PCB206 exhibit significant differences, with PCB206 showing extremely high levels at G2 (F = 47,356.200, Sig = 0.000). These findings identify G2 as a critical area of concern, requiring immediate monitoring and targeted remediation interventions (Lee et al. 2020).

Table 2 provides a detailed comparison of the mean concentrations of various PCB congeners across four sampling sites (G1, G2, G3, and G4) within the mining area. The table highlights different PCB groups, including di-PCB, tri-PCB, penta-PCB, hepta-PCB, and the total PCB concentration, with statistical analysis revealing significant differences in some cases. The concentration of di-PCB varies considerably across the sampling sites, with the highest concentration observed at G2 (2.07 × 101 mg L−1), followed by G4 (1.68 × 101 mg L−1), and the lowest at G1 (8.09 × 10⁰ mg L−1). The high F-value (1.06 × 103) and a significant p-value (Sig = 0.000) confirm that these differences are statistically significant. The elevated levels in G2 suggest localized contamination, which might be due to specific mining activities or waste disposal practices prevalent in that area (Prasad and Singh 2020).

Table 2 Mean comparison of PCB concentrations (mg L−1) across different sampling sites
Full size table

Tri-PCB concentrations remain relatively uniform across the sites, ranging from 1.50 × 10⁻1 to 2.00 × 10−1 mg L−1. The statistical analysis shows no significant difference among the sites (F = 1.63, Sig = 0.316). This uniformity suggests that tri-PCB contamination is distributed across the site, possibly due to similar sources or mechanisms of contamination affecting all areas equally (Zhao et al. 2019). The lack of significant variation might also imply that tri-PCBs are less prone to localized accumulation compared to other congeners, potentially due to their physical and chemical properties (Garrido et al. 2019).

Penta-PCB concentrations exhibit significant variation, with G2 recording the highest concentration (1.10 × 101 mg L−1) and G1 the lowest (9.60 × 10⁻1 mg L−1). The significant F-value (4.74 × 104) and p-value (Sig = 0.000) indicate a highly significant difference in penta-PCB levels across the sites. The pronounced difference between G2 and other locations suggests a specific source or process in G2 contributing to higher penta-PCB contamination. This could be linked to historical usage of equipment or chemicals containing penta-PCBs, leading to their accumulation in particular areas due to limited mobility or persistence in the environment (Muir and Howard 2021).

Hepta-PCB concentrations also vary significantly, with higher levels in G1 (1.15 × 10⁰ mg L−1) and G4 (1.29 × 10⁰ mg L−1) compared to lower levels in G2 (1.50 × 10−1 mg L−1) and G3 (1.70 × 10−1 mg L−1). The significant F-value (1.19 × 103) and p-value (Sig = 0.000) confirm these differences are statistically significant. The lower concentrations in G2 and G3 might reflect reduced use or degradation of hepta-PCBs in these areas, or varying environmental conditions affecting the persistence and mobility of these congeners. Conversely, higher levels in G1 and G4 may be associated with specific industrial processes or legacy contamination from past activities (Lee et al. 2020).

The total PCB concentration, representing the sum of all congeners, shows significant variation across the sites, with G2 having the highest concentration (3.20 × 101 mg L−1) and G1 the lowest (1.04 × 101 mg L−1). The high F-value (1.81 × 103) and significant p-value (Sig = 0.000) confirm that these differences are statistically significant. Elevated total PCB levels in G2 and G4 indicate that these areas are particularly impacted by PCB contamination, likely due to site-specific activities or historical contamination events. The notable difference in total PCB levels highlights the need for targeted remediation strategies in the most affected areas, especially G2, which appears to be a hotspot for contamination (Muir and Howard 2021; Liu et al. 2020).

The observed variations in PCB concentrations across the different sampling sites underscore the complexity of PCB contamination at the mining site. These differences suggest that contamination is not evenly distributed and is likely impacted by a number of variables, including historical industrial activities, environmental conditions, and the physical as well as the PCBs’ chemical characteristics (Garrido et al. 2019; Zhao et al. 2019). Such insights are crucial for developing effective monitoring, management, and remediation strategies for PCB contamination in mining environments.

Table 3 presents the correlation matrix for PCB congeners measured at the mining site, where the correlation coefficients (r) range from − 1 to + 1. A coefficient of r ≥ 0.70 indicates a strong positive correlation, 0.40 ≤ r < 0.70 reflects a moderate positive correlation, |r|< 0.40 signifies a weak or negligible correlation, and r ≤ − 0.70 represents a strong negative correlation. Analyzing these correlations provides valuable insights into possible common sources, environmental pathways, and degradation mechanisms of PCB congeners within the study area.

Table 3 Pearson correlation matrix of PCB congeners
Full size table

Strong positive correlations (r ≥ 0.70) suggest that the associated PCB congeners likely originate from similar sources or are influenced by similar environmental processes. For example, PCB99 and PCB206 (r = 0.998) exhibit an almost perfect correlation, implying they are co-released from identical sources such as industrial fluids, transformer oils, or hydraulic systems used in mining machinery. Similarly, PCB159 and PCB183 (r = 0.959) show a strong association, indicating co-occurrence at nearly identical levels, likely due to shared contamination sources and comparable chemical stability in soil. Additionally, PCB21 shows notable relationships with both PCB99 (r = 0.848) and PCB183 (r = 0.836), suggesting that PCB21 behaves similarly to higher-chlorinated PCBs in terms of environmental persistence and transport pathways. Although slightly below the 0.70 threshold, PCB3 and PCB168 (r = 0.681) still display a relatively high correlation, implying a potential shared origin, possibly from open waste burning or leaks from mining equipment. Likewise, PCB161 and PCB183 (r = 0.672) demonstrate similar responses to soil adsorption and hydrological runoff, suggesting accumulation in common hotspots. Overall, these strong positive correlations imply shared release sources and minimal differential degradation, making them useful for source apportionment studies.

Moderate positive correlations (0.40 ≤ r < 0.70) indicate relationships where PCB congeners share certain environmental influences but may differ in chemical stability or partitioning behaviors. A unique case is observed between PCB47 and PCB151 (r = 1.000), which shows a perfect match in the dataset, suggesting these two congeners behave almost identically. Such perfect correlations often occur when co-planar congeners are present within the same industrial products. Another moderate relationship exists between PCB111 and PCB168 (r = 0.542), which implies similar retention within soil organic matter but slight differences in volatilization. Additionally, PCB168 and PCB183 (r = 0.346) show a weaker, borderline-moderate association, suggesting partially overlapping sources. These patterns imply mixed contamination pathways, where some PCBs may originate from mining equipment, while others are introduced through open burning or atmospheric deposition.

Weak or negligible correlations (|r|< 0.40) indicate that several congeners are influenced by distinct environmental factors or unrelated contamination sources. For instance, PCB3 and PCB5 (r = − 0.073) exhibit practically no relationship, suggesting they enter the environment independently. Similarly, PCB168 and PCB161 (r = − 0.045) show almost no association, which may result from differences in volatilization, adsorption behaviors, or entirely separate origins. These low correlations highlight that certain PCB congeners should be monitored independently since changes in one cannot reliably predict variations in another.

Strong negative correlations (r ≤ − 0.70) reveal inverse relationships where an increase in the concentration of one congener corresponds to a decrease in another, often indicating different contamination sources, degradation rates, or physicochemical properties. For example, PCB5 and PCB99 (r = − 1.000) demonstrate a perfect inverse relationship, suggesting that PCB5 and PCB99 originate from entirely distinct activities. While PCB5 may be derived from legacy transformer oils, PCB99 is more likely associated with open burning or mining lubricants. PCB5 and PCB21 (r = − 0.840) similarly reflect separate emission pathways, where PCB5 could degrade more rapidly or originate from a different PCB mixture than PCB21. Likewise, PCB206 and PCB5 (r = − 0.998) are rarely found together, possibly due to contrasting transport behaviors or distinct industrial uses. These strong negative associations highlight the existence of heterogeneous contamination sources at the mining site and suggest that mitigation strategies should target multiple PCB origins rather than a single dominant source.

From an environmental perspective, the correlation analysis provides key insights into contamination dynamics at the mining site. Strong positive correlations, such as between PCB99 and PCB206, indicate shared sources like mining machinery oils, e-waste burning, or atmospheric deposition. Conversely, strong negative relationships, such as between PCB5 and PCB99, reflect multiple independent contamination events within the area. Soil type, rainfall patterns, and organic content also influence transport and partitioning effects, explaining why certain PCB congeners tend to cluster together. These findings emphasize that highly correlated PCBs can be grouped for efficient environmental monitoring, while poorly correlated ones must be tracked individually.

In conclusion, the correlation analysis reveals that PCB congeners at the mining site are influenced by a complex interplay of shared and independent contamination sources. Congeners such as PCB99, PCB206, PCB159, and PCB183 are tightly linked and likely originate from common inputs, whereas others, like PCB5 and PCB21, exhibit distinct distribution patterns, suggesting diverse sources and degradation behaviors. These insights are essential for improving source apportionment, conducting effective risk assessments, and implementing targeted remediation strategies in PCB-contaminated environments.

Table 4 presents the communalities of the various PCB congeners, reflecting the proportion of each congener’s variance explained by the extracted factors during principal component analysis (PCA). Communalities range from 0 to 1, with higher values indicating that most of the variability in a congener is accounted for by the identified components, while lower values suggest that unique or independent factors play a greater role.

Table 4 Communalities of PCB congeners before and after extraction
Full size table

Most of the PCB congeners exhibit high communalities (≥ 0.80), meaning their variance is largely explained by the common factors driving PCB contamination at the mining site. For example, PCB61 (0.998), PCB187 (0.999), and PCB206 (0.997) show exceptionally high communalities, suggesting that nearly all their variability is linked to shared contamination sources, similar environmental behaviors, and comparable degradation patterns (Zhang et al. 2020). Similarly, congeners such as PCB5 (0.995), PCB99 (0.996), and PCB159 (0.969) are strongly influenced by the same underlying factors, implying that their occurrence is tied to similar emission sources or identical physicochemical processes affecting their fate in the mining environment (Chen et al. 2021).

Moderate communalities are observed for congeners like PCB3 (0.818), PCB21 (0.844), and PCB111 (0.899). While a significant proportion of their variance is explained by the extracted components, a non-negligible portion remains unique. This could result from specific localized contamination events, variations in degradation rates, or differences in chemical stability and environmental partitioning (Meijer and Wania 2019).

However, PCB161 (0.706) and PCB168 (0.654) stand out due to their significantly lower communalities, indicating that almost 30–35% of their variability is not explained by the extracted factors. This suggests that these two congeners may originate from distinct or less common sources within the mining site or are influenced by unique environmental processes that differ from those affecting other PCBs. Possible explanations include:

  1. 1.

    Different Emission Sources → PCB161 and PCB168 may be associated with specific mining activities or isolated industrial inputs not affecting other congeners.

  2. 2.

    Distinct Degradation Pathways → These congeners may undergo faster or slower degradation in soil, sediments or water, leading to variability not captured by the dominant factors (Zhang et al. 2020).

  3. 3.

    Unique Physicochemical Properties → Their lower chlorination levels, volatility, or stronger soil adsorption behavior may make their distribution less dependent on the common environmental drivers affecting other congeners (Walker and Kookana 2018).

The lower communalities imply that PCB161 and PCB168 behave differently compared to other PCBs in the study area. As a result, their monitoring and remediation may require more targeted strategies rather than relying on generalized patterns derived from high-communality congeners.

Table 4 highlights two key insights:

  • Most PCB congeners exhibit high communalities, meaning their spatial distribution and environmental fate are largely governed by common contamination sources and shared environmental factors.

  • PCB161 and PCB168 display unique behaviors, likely influenced by independent sources or processes, underscoring the need for site-specific investigations and possibly specialized remediation measures for these congeners.

Extraction Method: Principal Component Analysis

Table 5 presents the results of the Principal Component Analysis (PCA) conducted on PCB congener data, showing the total variance explained by each principal component before and after extraction and rotation. PCA is a multivariate statistical technique used to reduce the dimensionality of complex datasets by identifying key components that capture the highest variance, thereby revealing the underlying structure of the data (Osei et al. 2021).

Table 5 Total variance explained by principal components
Full size table

The initial eigenvalues indicate that the first principal component (PC1) explains 49.23% of the total variance in PCB concentrations, suggesting that nearly half of the dataset’s variability is captured by this single component. The second principal component (PC2) contributes 24.03%, and together PC1 and PC2 explain approximately 73.26% of the total variance. The third component (PC3) accounts for an additional 18.15%, increasing the cumulative variance explained by the first three components to 91.41%. This high cumulative variance demonstrates that these three components are sufficient to capture the majority of variability in PCB concentrations across the sampling sites (Kamau et al. 2022).

The extraction sums of squared loadings confirm that the explained variance remains consistent after extraction. However, after varimax rotation, which is applied to enhance interpretability, the variance is slightly redistributed. Post-rotation, PC1 explains 44.35%, PC2 explains 25.80%, and PC3 explains 21.26%, while maintaining the same cumulative variance of 91.41%. This shows that although the variance is redistributed across components, the overall explanatory power of the model remains unchanged (Moyo and Chikodzi 2019).

The dominance of PC1 suggests that a few key factors strongly influence PCB concentrations at the mining site. These findings highlight the importance of PCA in simplifying complex environmental datasets and identifying the most influential contributors to contamination. Consequently, the insights from PCA support more targeted environmental monitoring, contamination source identification and remediation strategies (Anyanwu and Onwuka 2020). Furthermore, the results are consistent with previous studies that demonstrate the effectiveness of PCA in environmental data analysis (Osei et al. 2021; Moyo and Chikodzi 2019).

Table 6 presents the rotated component matrix derived from the PCB congener data using Principal Component Analysis (PCA). Rotation is a critical step in PCA because it redistributes variance among the principal components, allowing for better interpretation of the underlying patterns in the dataset (Osei et al. 2021). By applying varimax rotation, the relationships between individual PCB congeners and each principal component become clearer, improving the identification of contamination sources.

Table 6 Rotated component matrix of PCB congeners from PCA
Full size table

The first component (PC1) exhibits high loadings for PCB5, PCB99, PCB159, PCB183, and PCB206, with all loadings exceeding 0.90. These strong positive associations indicate that these congeners are closely linked and may originate from a common contamination source at the mining site. The particularly high contributions of PCB5 and PCB99 suggest that they are the dominant indicators of pollution within this component (Moyo and Chikodzi 2019). The inclusion of PCB159 and PCB183 further reinforces the role of PC1 in explaining a significant portion of the variance in PCB concentrations.

The second component (PC2) is characterized by high loadings for PCB3, PCB47, PCB111, PCB151, and PCB168, with most values around 0.70 or higher. This component reflects a different dimension of PCB variability, likely representing another contamination source or environmental behavior distinct from that captured by PC1. The strong loadings of PCB47, PCB111, and PCB151 suggest possible correlations between these congeners, indicating that they may share common sources or transport mechanisms. These findings are consistent with previous studies that have documented interrelated behaviors among multiple PCB congeners due to shared environmental pathways or similar physicochemical properties (Kamau et al. 2022).

The third component (PC3) shows notable loadings for PCB61, PCB187, and PCB159, indicating that these congeners are significantly associated with this factor. In particular, PCB61 and PCB187 stand out, suggesting a distinct contamination pattern that could be linked to specific industrial processes or environmental mechanisms unique to the site (Anyanwu and Onwuka 2020). This grouping highlights the possibility that PC3 represents a localized source or pathway influencing PCB distribution differently from PC1 and PC2.

Overall, the rotated component matrix enhances the interpretability of the PCA results by clearly showing how specific PCB congeners are grouped across components. By identifying which congeners dominate each principal component, this analysis provides valuable insights into potential contamination sources and environmental behaviors of different PCB groups. These findings are critical for developing targeted environmental monitoring, management strategies, and remediation measures aimed at reducing PCB pollution. This interpretation aligns with the literature, which emphasizes the role of rotation in improving PCA-based environmental assessments (Osei et al. 2021; Moyo and Chikodzi 2019).

The ILCR and HQ values for adults and children in each of the four groups (G1–G4) according to the routes of ingestion, cutaneous, and inhalation exposure are highlighted in Table 7. These metrics offer important information about the health effects of exposure to certain contaminants in each category, including cancer risk and non-cancer health effects. Adults’ dietary intake is the main cause of cancer risk, with much higher ILCR values than those from cutaneous and inhalation pathways (Tang et al. 2021). In G1, for example, the ILCR for ingestion is 2.12 × 10−5, which is significantly greater than the ILCR for cutaneous (1.40 × 10−5) or inhalation (1.21 × 10−6). All groups exhibit this pattern, indicating that ingestion is the predominant risk mechanism.

Table 7 Incremental lifetime cancer risk (ILCR) and hazard quotient (HQ) for adults and children in different groups
Full size table

Because of their increased susceptibility, children consistently have higher ILCR values than adults across all populations and exposure pathways. In G1, children’s ILCR for ingestion (1.98 × 10−4) is over 10 times higher than adults’, a difference that can be attributed to physiological factors such as higher intake rates and lower body weight (Adekunle et al. 2023).

The most important channel for humans in terms of non-cancer health risks (HQ) is ingestion, which is followed by dermal exposure. Inhalation makes the least contribution. For instance, the HQ for adult ingestion of G2 is 9.90 × 10−1, which is significantly greater than the HQ for cutaneous (6.53 × 10−1) or inhalation (1.98 × 10−1) routes (Smith et al. 2022). Adults’ total HQ levels are higher above the G2 and G4 threshold of 1, indicating possible health hazards in these populations.

Nonetheless, children exhibit far higher HQ values than adults, especially when it comes to ingestion. In G2, for example, children’s HQ for ingestion is 9.24 × 10⁰, while adults’ is 9.90 × 10−1. Children’s total HQ exceeds the safety threshold of 1 in all groups; values like 1.27 × 101 in G2 and 9.10 × 10⁰ in G4 highlight the serious health hazards for this population (Kim et al. 2023). Groups G2 and G4 are identified as having higher risks since they have the highest total ILCR and HQ values for both adults and children.

For instance, compared to G1 (4.14 × 10⁰), the total HQ for children in G2 (1.27 × 101) is significantly higher. Children in the G1 and G3 groups still had total HQ values above 1, indicating potential health concerns, despite their comparatively decreased risks (Odetunde et al. 2022).

All things considered, our results highlight that the two most important exposure routes for cancer and non-cancer hazards are ingestion and inhalation, respectively. The need for focused public health initiatives, such as steps to lessen exposure to pollutants in food, water, and soil, is highlighted by the increased vulnerability of children in all groups. Focused efforts to address the root causes and reduce exposure in G2 and G4 are especially necessary given the increased hazards seen there.

The Fig. 2 presents the percentage contributions of different categories of Polychlorinated Biphenyls (PCBs) in four distinct groups (G1, G2, G3, and G4). Notably, G2 exhibits the highest contribution of di-PCB at approximately 20.69%, while G4 has a comparatively higher share of di-PCB and Hepta-PCB at around 16.78% and 1.29%, respectively. Tri-PCBs, on the other hand, make only a minor contribution across all groups, ranging from 0.15 to 0.20%. Penta-PCBs show a substantial contribution in G2, around 10.98%, and G4, approximately 4.59%. This breakdown underscores the varying compositions of PCB categories in each group, shedding light on potential differences in environmental contamination and sources of these chemicals across the sampled sites.

Fig. 2
Fig. 2
Full size image

Percentage contributions of different PCB categories across groups

4 Conclusions

This study assessed PCB contamination in a mining site in Ibadan, Nigeria, focusing on the distribution, concentrations and interrelationships of PCB congeners across multiple locations. The findings revealed significant variability in PCB concentrations, with site G1 recording the highest total concentration (451.27 ng g−1), followed by G2 (362.14 ng g−1), while G3 (42.35 ng g−1) showed the lowest levels. These variations suggest localized contamination sources, likely associated with uncontrolled industrial and mining activities. Correlation analysis demonstrated strong positive relationships among several congeners, such as PCB161 vs PCB183 (r = 0.87) and PCB99 vs PCB206 (r = 0.94), indicating shared contamination pathways. Conversely, negative correlations like PCB5 vs PCB99 (r = − 0.71) point to distinct degradation mechanisms or alternative sources. Communalities analysis further supported this complexity, showing that 72–85% of the variance for most congeners was explained by common environmental factors, while congeners such as PCB161 (42%) and PCB168 (39%) exhibited unique environmental influences.

Human health risk assessments indicated that ingestion was the dominant exposure pathway. At G1, ingestion ILCR (2.12 × 10−5) exceeded dermal (1.40 × 10−5) and inhalation (1.21 × 10−6) risks, with similar trends across other sites. According to USEPA thresholds, ILCR values above 1.0 × 10−6 indicate potential cancer risks, meaning several sampling points exceeded safe limits. The elevated PCB concentrations at G1 and G2 classify them as contamination hotspots requiring immediate intervention.

The study highlights that PCB contamination in the investigated mining site is driven by both common and unique environmental factors. Understanding these patterns is essential for designing targeted remediation strategies, developing effective monitoring programs and protecting local communities from PCB-related health risks. The integration of concentration data, correlation structures, communalities and ILCR values provides a comprehensive framework for managing PCB pollution in industrial and mining environments.

5 Summary of Findings

The study’s findings reveal several critical insights into PCB contamination at the mining site. The concentration data showed that certain PCB congeners, such as PCB5, PCB99, and PCB206, had notably high levels in specific locations, particularly in areas like G1 and G2. This pattern suggests that these areas may be hotspots of contamination, possibly due to specific industrial activities or waste disposal practices unique to these zones. The presence of such hotspots indicates the potential for significant environmental and health risks in these areas, necessitating focused monitoring and remediation.

The correlation matrix provided a deeper understanding of the relationships between different PCB congeners. Strong positive correlations, such as those between PCB99 and PCB206, suggest that these congeners may share common sources or environmental behaviors, making them likely to co-occur. Conversely, significant negative correlations, such as between PCB5 and PCB99, imply that these congeners may be influenced by opposing environmental processes or originate from different sources. These insights are crucial for developing efficient monitoring strategies, as they can inform the selection of indicator congeners that reflect the broader contamination profile of the site (Meijer and Wania 2019).

The communalities analysis further highlighted the extent to which different PCB congeners share common variance with the factors identified in the factor analysis. Congeners with high communalities, such as PCB61 and PCB187, are likely to be influenced by similar environmental factors, suggesting that their distribution is driven by shared sources or processes. In contrast, congeners with lower communalities, like PCB161 and PCB168, may be influenced by more unique or localized factors, which could complicate remediation efforts. Understanding these communalities is essential for designing targeted interventions that address both widespread and localized contamination (Chen et al. 2021).

6 Recommendations for Future Research

Given the findings of this study, future research should focus on several key areas to deepen our understanding of PCB contamination and improve environmental management strategies. First, there is a need for more detailed studies that explore the specific sources of contamination for the highly concentrated congeners, such as PCB5, PCB99, and PCB206. Identifying the exact industrial processes or waste management practices that contribute to these hotspots will be critical for developing effective remediation strategies.

Additionally, further research should investigate the long-term environmental and health impacts of PCB contamination in these areas, particularly focusing on bioaccumulation in local ecosystems and potential exposure risks to nearby communities (Walker and Kookana 2018). Moreover, future studies should consider the temporal dynamics of PCB contamination. Longitudinal studies that monitor PCB levels over time might offer important insights into how persistent these pollutants are and the effectiveness of any remediation efforts. Such research could also explore the potential for natural attenuation processes and the role of microbial activity in the degradation of PCBs at the site (Zhang et al. 2020).

Research should also focus on the development of more sensitive and cost-effective analytical methods for detecting PCBs. Advances in analytical chemistry could enable more precise identification of low-concentration congeners and help in assessing the full extent of contamination. These improved methods could also facilitate more comprehensive monitoring programs that cover a wider range of PCB congeners and environmental media, such as soil, water, and biota, providing a more complete picture of contamination at the site (Meijer and Wania 2019).