1. Introduction
Polycyclic aromatic hydrocarbons (PAHs) are pervasive, persistent organic pollutants, predominantly originating from the incomplete combustion of fossil fuels and biomass [
1]. Recognized for their teratogenic, carcinogenic, and mutagenic properties, 16 types of PAHs have been prioritized as pollutants by the United States Environmental Protection Agency (USEPA), underscoring their significant threat to human health and attracting considerable attention [
2]. Due to their hydrophobic nature, PAHs tend to adsorb onto organic matter–rich environmental media, such as soil and sediment, upon release into the environment. These pollutants can be re-released when environmental conditions fluctuate, posing ongoing risks to both the ecological environment and human health. Hence, a thorough understanding of the re-release dynamics of PAHs from environmental media is essential for accurately assessing the associated environmental and health risks.
Human exposure to PAHs occurs through various pathways, including ingestion and inhalation, with the latter two being the most prevalent [
3]. Children, in particular, are at an increased risk of absorbing environmental pollutants from soils and dust due to their behavioral patterns, such as hand–to–mouth activity [
4]. In vitro assays have been instrumental in determining the portion of pollutants that desorb from particles upon contact with biological fluids [
5]. Extensive research has demonstrated that considering the total pollutant load in media overestimates health risks. For instance, Zhang et al. [
6] employed various digestion models to extract PAHs from contaminated soil, revealing bioaccessibility rates below 13.4% for individual PAHs. In soils treated with contaminated wastewater, the bioaccessibility of PAHs in the small intestine was higher, peaking at 53% [
7]. Similarly, the bioaccessibility of PAHs in PM2.5 in the respiratory tract reached a maximum of 24.5% [
8]. Xie et al. [
9] utilized Modified Gamble’s Solution (MGS) and Artificial Lysosomal Fluid (ALF) to extract PAHs from e-waste particles, with results indicating the highest release rate nearing 90% and consistently higher bioaccessibility in ALF compared to MGS. Furthermore, the “aging effect” refers to the decline in extractability and bioavailability of soil-bound PAHs over time, reducing their availability to microorganisms and other ecological receptors [
10,
11]. Wei et al. [
10] observed a significant decrease in the available extraction rate of pyrene with increasing contact time, particularly in the rapid desorption fraction. After 150 days of exposure, the desorption rate of phenanthrene dropped from 82% to 65%, with pyrene showing a less pronounced change. This aging effect leads to an increase in the irreversible fraction of PAHs, thereby diminishing ecological risks [
12].
In summary, the re-release of PAHs is influenced by multiple factors, including the number of rings, hydrophobicity, the nature of the media, the properties of the extractant, and the placement in the human body. Current research predominantly focuses on the bioaccessibility of adsorbed PAHs in the digestive and respiratory tracts, with less emphasis on the processes and mechanisms governing these dynamics. Given the significance of inhalation and hand-to-mouth exposure pathways for particle-bound PAHs, further investigation into these processes and mechanisms is warranted. Additionally, environmental conditions are subject to change, which can alter the physical and chemical properties of soil and other media, subsequently affecting the re-release of adsorbed PAHs. For example, freeze-thaw cycles can lead to the rearrangement of soil particles and alterations in soil structure, impacting the re-release processes of adsorbed PAHs. The implications of these changes on the re-release of PAHs and their potential impact on human health risks require further exploration.
Unraveling the re-releasing process of particle-bound PAH in body fluid may provide insight into its fate and related drivers, allowing for a more precise estimation of the potential human health risk of PAH. The objectives of the present study were (i) to compare the discrepancies in bioaccessibilities of selected PAHs before and after freeze-thaw cycles, (ii) to employ the desorption kinetics model to reveal the re-release process of PAHs in simulated body fluids, and (iii) to identify the key drivers of responding to environmental change in the bioaccessible processes. This study hopes to enhance understanding of the geochemical dynamics of PAHs and provide critical information for risk assessors in accurately characterizing and prioritizing contaminated sites.
2. Materials and Methods
2.1. Sorbents Preparation
For this study, two distinct types of surface soils were utilized as geosorbents: ferrallitic soil (FS); and calcareous soil (CS). The ferrallitic soil was obtained from the Maolan National Reserve in Libo, while the calcareous soil was sourced from the Duxi Forest Park in Guiyang, both located in Guizhou Province, Southwest China. The soil samples were carefully collected from the upper 0 to 20 cm layer and air-dried to remove moisture. Subsequently, the soil was sieved through a 4 mm stainless-steel mesh to eliminate any roots and stones, ensuring a uniform particle size.
The sieved soil was then meticulously ground using a mortar and pestle to further homogenize the sample. Finally, the soil was passed through a 150 μm stainless-steel sieve for the bioaccessibility experiments. Although these soil particles might not have perfectly replicated the characteristics of particles that are actually inhaled or ingested, they served as suitable proxies for in vitro assays. These assays were designed to reveal the interactions between soil particles and PAHs in simulated body fluids, allowing us to evaluate the effects of various assay parameters on the behavior of these particles at the interface.
2.2. Simulated Body Fluids Preparation
The background solution used in this study contained 0.005 M CaCl
2 and 0.001 M NaHCO
3. Two artificial body fluids, Gamble’s solution (namely, simulated lung fluid) and simulated saliva, were employed. The detailed compositions of these simulated body fluids are presented in
Table 1. To minimize the risk of precipitation, each component was added in the order presented. All solutions were prepared with Milli-Q water, and sodium azide (NaN
3) with a concentration of 100 mg L
−1 was added to prevent the degradation of PAHs by microorganisms. Organic reagents (guaranteed reagents) were purchased from Sigma-Aldrich Chemical Co. (Shanghai, China).
2.3. Soil Spiking and Aging
Phenanthrene (C
14H
10) and pyrene (C
16H
10), two common probe PAHs, were selected as target compounds (
Figure 1). These PAHs were initially dissolved in methanol (HPLC grade, Merck, Germany) to create stock solutions with concentrations of 1000 μg L
−1 for phenanthrene and 100 μg L
−1 for pyrene. Subsequently, soil particles were spiked with the target PAHs to achieve concentrations of approximately 20 mg kg
−1 for phenanthrene and 8 mg kg
−1 for pyrene at a temperature maintained at 21 ± 1 °C for the bioaccessibility kinetics experiments. Prior to incubation, the concentrations of spiked soil PAHs were confirmed analytically. The determination of solid-phase PAH concentrations was performed in accordance with the method established by Huang and Pignatello [
15]. Briefly, approximately 1 g of soil was placed into a 20 mL glass vial, followed by the addition of a methanol-water solution in a 4:1 volume ratio, leaving approximately 1 mL of headspace within the vial. The vials were then sealed and shaken horizontally in a water bath at a temperature of 85 °C for 8 h. Post-extraction, the vials were centrifuged at 3000×
g for 30 min, after which, the supernatant was carefully collected for further analysis.
The soil samples spiked with PAHs were transferred into 50 mL Teflon centrifugation tubes and maintained under moist conditions. They were then subjected to a freeze-thaw process by being frozen at −18 ± 1 °C for 24 h and subsequently thawed at 21 ± 1 °C in a temperature-controlled chamber in the dark for 24 h. This cycle was repeated 15 times (FT15) and 30 times (FT30) for the respective samples to simulate the process of repeated freezing and melting of the surface soil with varying frequency in the natural environment within a short time. Following the freeze-thaw cycles, the samples were freeze-dried under vacuum conditions for subsequent analysis.
2.4. Desorption Kinetics Experiments
The desorption kinetics experiments were conducted in duplicate. Briefly, 20 mg of PAH-spiked soil sample was placed into a glass vial with Teflon-lined caps. Subsequently, 15 mL of simulated body fluid was added to the vial, ensuring that the headspace was minimized to less than 0.1 mL. The vials were then securely sealed and horizontally placed into a temperature-controlled shaker to incubate at 37 ± 1 °C with a shaking speed of 120 rpm. The incubation periods were systematically varied as follows: 15 min; 30 min; 1 h; 2 h; 8 h; 24 h; 48 h; 1 week; and 2 weeks, with all incubations conducted in the dark. At each predetermined time interval, the vials were centrifuged at 3000 rpm for 20 min. Following centrifugation, approximately 2 mL of the supernatant was sampled and transferred into a 5 mL glass vial containing pre-weighted methanol (approximately 2 mL) to stabilize phenanthrene and pyrene within the vial. These vials were then stored at 4 °C in a dark environment to await subsequent analytical measurements. A control experiment of no sorbent in the vial was conducted simultaneously to quantify the loss of solute during experiments and was used as a reference for data reduction.
2.5. Sample Analysis and Data Calculation
2.5.1. Sample Analysis
The concentrations of phenanthrene and pyrene were determined using ultra-high-performance-liquid-chromatography (1290 UHPLC, Agilent, Santa Clara, CA, USA) equipped with a diode array UV detector (DAD) (G4212A), a fluorescence detector (FLD) (G1321B), and an auto-sampler unit (G4226A), employing an Eclipse XDB-C18 column (2.1 × 150 mm, 3.5 μm) at a column temperature of 37 °C. Phenanthrene and pyrene were analyzed using DAD at wavelengths of 254 nm and 272 nm, respectively, and using FLD with excitation/emission wavelengths of 254 nm/366 nm and 270 nm/392 nm, respectively.
The sorbate concentration in the supernatant or initial solution was calculated by adjusting the UHPLC-measured sorbate concentration with a corresponding dilution factor based on the mass ratio of methanol to water in the mixture.
The total organic carbon (TOC) of each sample was determined using the dry combustion method with an elemental analyzer (Vario El III, Langenselbold, Germany). Prior to analysis, the soil sample underwent treatment with 0.5 mol L−1 HCl (with a volume ratio of soil to HCl solution of approximately 1:30) to remove inorganic carbon. Soil particle-size distributions were determined following the hydrometer method (ASTM D422-63). Two blind replicates were included in the analysis, and good repeatability was obtained (<3%).
2.5.2. Data Calculation
Desorption kinetics from soils or sediments are often characterized by fast processes followed by a plateau of slower progress [
16]:
where S
t (mg kg
−1) is the PAH content in the soil at time t (h) and S
0 (mg kg
−1) at the start of the experiment; F
rap and F
slow (–) are the rapidly and slowly desorbing fractions; k
rap and k
slow (h
−1) are the rate constants of rapid and slow desorption compartments, assuming that k
slow ≪ k
rap. We additionally assumed that the two defined fractions covered the
Parameters in Equations (1) and (2) were determined by direct curve fitting, minimizing the cumulative squared residuals between experimental and calculated values of St/S0 (Origin 2022).
The bioaccessibility (%BA) of PAH was calculated by adding the percentages released in simulated lung fluid and simulated saliva:
The Mann-Whitney U test was used to verify the significance of variations between compounds, soils, and simulated body fluids, while the Jonckheere-Terpstra test was used to determine the significances of variations among groups of freeze-thaw cycles. All statistical analyses were performed using SPSS (v23.0, IBM Corp, Armonk, NY, USA). All the figures were prepared using GraphPad Prism 9.
2.6. Quality Assurance and Quality Control
A method blank, a spiked blank, and a pair of matrix-spiked sample duplicates were processed and analyzed in parallel. A total of 20% of the samples were selected to calculate the recovery, which was (87 ± 12)% for phenanthrene and (95 ± 11)% for pyrene.