Development and Field Application of a Diffusive Gradients in Thin-Films Passive Sampler for Monitoring Three Polycyclic Aromatic Hydrocarbon Derivatives and One Polycyclic Aromatic Hydrocarbon in Waters

By inergency On Feb 27, 2024

1. Introduction

Polycyclic aromatic hydrocarbons (PAHs) are a group of persistent organic compounds that are widely found in various types of environmental matrixes and have teratogenic, carcinogenic and mutagenic properties [1]. PAHs can be chemically and microbially transformed into PAH derivatives, including nitro polycyclic aromatic hydrocarbons (NPAHs), oxygenated polycyclic aromatic hydrocarbons (OPAHs), chlorinated polycyclic aromatic hydrocarbons (ClPAHs), etc. [2,3]. Certain PAH derivatives are more biotoxic than their parent PAHs [4,5]. NPAHs and OPAHs have a higher water solubility, and they are theoretically more mobile in water than their parent PAHs [6]. Compared to pyrene, the NPAH 1-nitropyrene exhibits a higher carcinogenicity and mutagenicity/genotoxicity in in vitro and in vivo tests [7,8]. A study in Brazil showed the mutagenicity of OPAHs in particulate matter samples increased approximately fourfold compared to their corresponding PAHs [9,10]. ClPAHs, which are structurally similar to dioxins and polychlorinated biphenyls (PCBs), are a type of highly toxic organochlorine compound [11,12,13].

Accurate and stable determination of compounds in an aqueous environment can help us better understand their transport and transformation processes to assess their biological toxicity and impact on the ecosystem and human health [14]. At present, in water, PAH derivatives and PAHs are mainly detected via spot grab sampling, demanding a significant amount of resources and expertise. Moreover, some trace-level pollutants are often undetectable, or their measurements are susceptible to detection errors, which is not conducive to understanding the current pollution status and pollution trends in water [15,16]. Spot grab sampling provides instantaneous concentrations [17], which may not be representative of the pollution over a period of time or describe the temporal variations in the water quality [18,19].

Passive sampling techniques, requiring no collection of water samples or their transport and preservation, can save considerable human, material and financial resources compared with grab sampling. Furthermore, the time-weighted average (TWA) concentration measured by passive sampling is more meaningful, determining the overall water pollution during the assessed period [20]. Some studies have utilized passive samplers, such as semipermeable membrane devices (SPMDs) and polyethylene (PE) passive samplers, to quantify PAHs in water [21,22]. The SPMD procedure is well established and widely used. However, the presence of glyceryl trioleate in the SPMD makes the postanalytical purification process cumbersome [23]. The small capacity of PE passive samplers makes them unsuitable for long-term water monitoring [24,25]. The diffusive gradients in thin films (DGT) technique was first proposed by Davison and Zhang and was originally used to detect trace metals in freshwater systems [26,27]. Compared with other passive samplers, DGT passive sampling is less affected by environmental conditions, and laboratory-measured diffusion coefficients could be used to accurately predict in situ sampling rates, which would be highly convenient for field sampling. DGT techniques have also been widely used for trace organic contaminant detection in water [16,28]. However, to date, there have been limited studies on DGT or other passive sampling techniques specifically designed for sampling PAH derivatives in water.

In this study, we developed a new DGT passive sampler for monitoring PAH derivatives and PAHs in waters. Binding gels based on the XAD18 resin can rapidly adsorb a variety of organic pollutants in water and are simple and inexpensive to produce [29]. Thus, we used XAD18 binding gels as the adsorption phase and agarose gels as the diffusion layer material, which are resistant to temperature, pH and ionic strength changes [17]. The aims of this study are (i) to develop a DGT passive sampler for sampling PAH derivatives and PAHs in waters and study their adsorption kinetics, adsorption capacity, diffusion coefficients D and environmental effects in laboratory tests; (ii) to evaluate the DGT technique in field tests; and (ⅲ) to apply the DGT technique for in situ monitoring of PAH derivatives and PAHs in rivers and lakes.

2. Materials and Methods

2.1. Chemicals and Analysis

In this test, we selected three PAH derivatives—9-fluorenone (9-FL), 9-nitroanthracene (9-NA) and 1-chloroanthraquinone (1-ClAQ)—which are representative OPAHs, NPAHs and ClPAHs, respectively, and one typical PAH, phenanthrene (Phe), as the target compounds. Their physical and chemical properties are shown in Table S1 in the Supporting Information (SI). The composition and sources of the other reagents and materials are shown in SI Table S2.

2.2. DGT Preparation

A standard DGT device consists of a filter membrane, an agarose-based diffusive gel and a binding gel. A detailed description of DGT device construction and its principles is presented in the SI. In this experiment, 0.8 mm diffusion agarose diffusion gels were prepared according to reference [30]. In order to select DGT-appropriate materials that have a low influence on the adsorption of the target compounds, agarose diffusive gel discs and four types of filter membrane (polyethersulfone (PES, 0.14 mm thick), polytetrafluoroethylene (PTFE, 0.18 mm thick), polyvinylidene fluoride (PVDF, 0.11 mm thick) and nylon (0.11 mm thick) (all with 2.5 cm diameter)) were separately immersed in a 100 mL solution containing the four target compounds (9-FL; 1-CLAQ; Phe: 100 μg L⁻¹; 9-NA: 20 μg L⁻¹). At the same time, DGT moldings were immersed in 200 mL of a solution of the same composition. All of the above solutions were shaken for 12 h, and the adsorption of the four compounds on each material was calculated based on the concentration difference from the beginning to the end of the experiment [30]. The reason for the different concentrations of the target compounds is the low solubility of 9-NA and the fact that the 9-NA concentration is usually lower than that of the other three compounds in ambient water. The selected concentrations of each compound in the following experiments are the same as in this section due to the same reason.

The binding gel preparation procedure was as follows: The XAD18 resin was dried and ground into powder form, fully activated with methanol and then rinsed with ultrapure water. Then, 6 g (wet weight after pretreatment) of resin was added to 30 mL of 2% agarose solution, and then the solution was heated to boiling. This agarose solution was pipetted between two preheated glass plates (>80 °C) separated by a 0.5 mm thick PVC plastic sheet and cooled to room temperature. Then, the gel was cut into discs 25 mm in diameter. Diffusive gels and binding gels were placed in ultrapure water, sealed and stored at 4 °C.

2.3. Characteristics of the Binding Gel

2.3.1. Adsorption Kinetics

The ability of the XAD18 binding gel to rapidly adsorb compounds is key to its ability as a binding layer in DGT devices, and thus its adsorption kinetics need to be examined. The XAD18 binding gel was immersed in 100 mL of a solution containing the four target compounds (9-FL; 1-CLAQ; Phe, 100 μg L⁻¹; 9-NA, 20 μg L⁻¹), and shaken for 30 h. Samples were taken from the solution at different time intervals ranging from 5 min to 30 h for analysis. The mass of the target compounds adsorbed by the binding gel at different time intervals was calculated.

2.3.2. Elution Efficiency

The binding gels were immersed separately in 50 mL solutions containing the four target compounds at concentrations of 10 μg L⁻¹, 20 μg L⁻¹ and 100 μg L⁻¹ and then shaken for 4 h. After the adsorption was complete, the binding gels were placed into 10 mL of methanol and sonicated for 90 min; these solutions were then filtered through an organic filter membrane and analyzed via HPLC. The elution efficiency was the ratio of the measured mass of the target compound in the eluate to the adsorbed mass calculated by the difference in concentrations of the solution before and after adsorption.

2.3.3. Adsorption Isotherm

Adsorption isotherm tests were conducted to assess the uptake capacity of the binding gels. Different concentration solutions were prepared; the initial concentrations of 9-FL, 1-CLAQ and Phe were 100 μg L⁻¹, 500 μg L⁻¹, 1 mg L⁻¹ and 2 mg L⁻¹, and the initial 9-NA concentrations were 50 μg L⁻¹, 100 μg L⁻¹, 200 μg L⁻¹ and 300 μg L⁻¹. Next, the ¼ XAD18 binding gels were immersed into the solution of each component separately. The resulting concentrations of the solutions and the mass of the four target compounds loaded on the binding gels were measured after shaking the solutions for 24 h.

2.4. Diffusion Coefficient Determination

The concentrations determined via the DGT technique were calculated using Equation (1) [31].

$C_{D G T} = \frac{M Δ g}{D A t},$

(1)

where M is the mass accumulated on the XAD18 binding gels, Δg is the thickness of the diffusive layer (diffusive gel and filter membrane), D is the diffusion coefficient of the target analyte in the diffusive layer, t is the exposure time, and A is the window area of the DGT device (3.14 cm²).

In the present study, the diffusion coefficients of the target compounds were experimentally determined, and DGT-based compound concentrations in the laboratory and field tests were determined using Equation (1). DGT devices containing XAD18 binding gel, 0.80 mm thick agarose diffusive gel and PTFE filter membranes were placed into 2.5 L of well-stirred solutions containing the four target compounds (9-FL; 1-CLAQ; Phe, 100 μg L⁻¹; 9-NA, 20 μg L⁻¹) and 0.01 M NaCl at 25 °C for 12 h. After adsorption was complete, the binding gel was removed from the DGT device and eluted; then, the mass of the compound adsorbed on the binding gel was calculated.

The diffusion coefficients of the four target compounds were also investigated at a lower concentration due to the low levels of PAHs and their derivatives in ambient water. The DGT devices were deployed in 2.5 L of solution containing the target compounds at 3 μg L⁻¹, along with 0.01 M NaCl, at a temperature of 25 °C. After deployment for 12 h, the compounds’ diffusion coefficients were measured.

The diffusion coefficients of the four target compounds were also investigated at lower concentrations due to the low levels of PAHs and their derivatives in ambient waters. The DGT devices were deployed in 2.5 L of solution containing the target compounds at 3 μg L⁻¹, along with 0.01 M NaCl, at a temperature of 25 °C. After deployment for 12 h, the diffusion coefficients of the compounds were measured.

2.5. DGT Performance Tests under Different Conditions

2.5.1. Effects of pH and IS

To test the effects of IS and pH on DGT performance, the DGT devices were placed for 12 h in two different well-stirred solutions: (a) a 2.5 L solution containing 100 μg L⁻¹ 9-FL, 1-CLAQ and Phe and 20 μg L⁻¹ 9-NA (IS = 0.01 M NaCl) at different pH values (5–8, adjusted with 0.1 M HCl or 0.1 M NaOH); (b) a 2.5 L solution containing 100 μg L⁻¹ 9-FL, 1-CLAQ and Phe and 20 μg L⁻¹ 9-NA (pH = 6) with NaCl concentrations ranging from 0.005 to 0.5 M.

2.5.2. Influences of Diffusion Film Thickness and Deployment Time

To explore how the mass taken up by the DGT device was affected by the diffusive gel thickness, DGT devices containing agarose diffusive gels of various thicknesses (0.68 mm–2.18 mm) were immersed in 2.5 L of solution containing 100 μg L⁻¹ of 9-FL, 1-CLAQ and Phe, 20 μg L⁻¹ of 9-NA and 0.01 M NaCl for 12 h.

To investigate the impact of the deployment time, the DGT devices were immersed in well-stirred solutions (4 L) containing 3 μg L⁻¹ of the target compound and 0.01 M NaCl for a period of 7 days (168 h), with three devices being recovered every other day. To ensure accurate results and minimize any potential effects from compound evaporation, the solutions were replaced once daily.

2.5.3. Competition Effects and Aging Effects

In order to investigate the adsorption competition among the compounds, the DGT devices were deployed for 12 h in four groups of solutions (2.5 L). The compound concentrations in each group were as follows: (a) 9-FL, 1-CLAQ, Phe: 100 μg L⁻¹, 9-NA: 20 μg L⁻¹, IS: 0.01 M NaCl; (b) 9-NA, 1-CLAQ, Phe: 100 μg L⁻¹, 9-FL: 20 μg L⁻¹, IS: 0.01 M NaCl; (c) 9-NA, 9-FL, Phe: 100 μg L⁻¹, 1-CLAQ: 20 μg L⁻¹, IS: 0.01 M NaCl; (d) 9-FL, 9-NA, 1-CLAQ: 100 μg L⁻¹, Phe: 20 μg L⁻¹, IS: 0.01 M NaCl.

The XAD18 binding gels were immediately stored in ultrapure water after preparation; the overall storage duration is called the aging time. The aging times of XAD18 binding gels were set to 25, 45 and 90 days. The DGT devices assembled with binding gels with different aging times were placed into 4 L solutions containing 100 μg L⁻¹ 9-FL, 1-CLAQ and Phe, 20 μg L⁻¹ 9-NA and 0.01 M NaCl for 12 h.

In each performance test described above, the mass of compounds present in the DGT binding gel was measured after device retrieval. Subsequently, the C_DGT was calculated using Equation (1) and compared to the compound concentrations in the actual water samples.

2.6. Field Trials

The field deployments included two phases, one for evaluation and the other for water quality monitoring. First, we deployed the DGT devices in Goose Pond, a landscape water body at Anhui University campus (117°27′12.46″. 31°37′53.82″) in Heifei City, China, and recorded the water temperature every 15 min utilizing a temperature recorder. Three DGT devices were removed after 3, 6, 10 and 15 d, and the concentrations (C_DGT) of 9-FL, 1-CLAQ, 9-NA, and Phe were calculated according to Equation (1). Meanwhile, active sampling, i.e., traditional grab sampling, was also conducted at the pond on Day 3, 10 and 15 of DGT deployment to measure the concentrations (C_active) of the same compounds. Then, the two results were compared. After validating the feasibility of using the XAD18-DGT technique to assess the quantity of four target compounds in the campus pond, we deployed the DGT devices in several rivers flowing into Chaohu Lake, the fifth largest freshwater lake in China, located in Hefei City, during June to September 2023. A map of the sampling sites is shown in Figure 1. The DGT units were retrieved after 7 days at every site. The pretreatment of water samples is presented in the SI.

4. Conclusions

In this work, a feasibility study for using the XAD18-DGT technique for the detection of PAH derivatives and PAHs in waters was conducted, proving that it is a reliable and cost-effective method for PAH derivative and PAH detection. Although only four typical target compounds were studied, this class of organic compounds have similar physicochemical properties, so the DGT technique could be applied to monitor other PAH derivatives and PAHs. The approach was stable and reliable for measuring PAH derivatives and PAHs, and it was not limited by the pH, ionic strength or deployment time within a certain range in natural waters. Thus, DGT devices could be used in a variety of aqueous environments, such as highly polluted fresh water and seawater. The resulting concentrations measured using the DGT technique are time-weighted concentrations, which better reflect the true levels of organic pollutants in water bodies. By employing the DGT technique, the potential risks associated with sampling in adverse weather conditions are mitigated. Compared with other passive samplers, the DGT method’s flux independence and ease of use without field calibration could make it a suitable technique for in situ monitoring of PAH derivatives and PAHs in a variety of aquatic systems.