Emerging Contaminants in Groundwater and Surface Water at a Mine Reclamation Site: Occurrence, Fate and Remediation

Abstract

Mine operations can generate large volumes of sulfide-rich waste rock and tailings. When exposed to atmospheric O2 and water, sulfide minerals can oxidize, producing low-quality drainage that contains high concentrations of dissolved metal(loid)s, sulfate, and acid. To limit the release of these toxic substances and promote landscape revegetation, municipal biosolids are increasingly utilized in mine rehabilitation programs. A multilayer cover system, consisting of a 0.5 m biosolids amended organic carbon (OC) layer and a 2 m desulfurized tailings (DST) layer, was applied on the high-sulfide tailings area in northern Ontario to retain moisture in the underlying tailings at high degrees of saturation, and promote revegetation of the landscape. Water samples from the vadose and saturated zones in the covered tailings system were collected versus depth using discrete sampling, whereas adjacent surface water was monitored using both discrete sampling and two passive samplers (diffusive gradients in thin films (DGT) and polar organic chemical integrative samplers (POCIS)) to assess spatial and temporal variations in major ions, trace elements, pharmaceutical and personal care products (PPCPs), artificial sweeteners, and per- and polyfluoroalkyl substances (PFASs) over a two-year period. Field investigations showed improved geochemical conditions below the cover system. In the vadose zone, the composite cover maintained circumneutral porewater pH through O2 consumption by OC, the enhanced neutralization capacity provided by added CaO, and the restriction of O2 infiltration in the high-moisture content DST layer, which effectively limited sulfide mineral oxidation compared to the acidic conditions (pH 3–4) observed in uncovered areas. While elevated concentrations of SO₄²⁻ and trace metals (e.g., Ni, Co, Mn, Cu, Zn) were observed in the subsurface groundwater due to historical oxidation, lower Fe concentrations beneath the biosolids layer are attributed to precipitation of goethite favoured under the circumneutral pH conditions. Concentrations of major ions and trace elements peaked adjacent to the tailings area (Ca: 152–298 mg L-1; SO42-: 347–745 mg L-1, Fe: 122–715 µg L-1; Ni: 810–1960 µg L-1) then decreased downstream within the alkaline treatment system decreasing critical trace element concentrations to well below regulatory limits (e.g., 6.0–426 µg L⁻¹ for Fe; 15.5–663 µg L⁻¹ for Ni). Porewater and groundwater monitoring indicated leaching of five pharmaceutical and personal care products (PPCPs) (<LOQ–1460 ng L-1) and two artificial sweeteners (<LOQ–820 μg L⁻¹). These compounds were transported to uncovered areas and to DST only covered areas of the tailings. The environmental fate of carbamazepine (CBZ) was primarily controlled by hydrophobic sorption. (Bio)degradation under different redox conditions influenced the fate of caffeine (CAF), ibuprofen (IBU) and naproxen (NAP). The fate of CAF was also affected by H-bonding processes. The fate of gemfibrozil (GEM), acesulfame-K (ACE-K), and sucralose (SCL) was mainly governed by physical transport of groundwater. Correlation analysis with the conservative chloride (Cl⁻) ion indicated that select PPCPs (IBU, GEM, and SCL) also showed conservative transport, indicating potential use as co-tracers for biosolids leachate. Eight target per- and polyfluoroalkyl substances (PFASs), predominantly perfluorocarboxylic acids (PFCAs), were observed in the underlying vadose and saturated zones, with average total PFAS concentrations reaching 710 ng L⁻¹ in 2021 and 785 ng L⁻¹ in 2022. Subsurface transport was primarily controlled by compound hydrophobicity and air-water interfacial adsorption, limiting long-chained PFCAs to the upper cover while short-chained species were transported into deeper groundwater. For instance, the highest concentrations of PFOA observed in the vadose zone of the DST layer below the biosolids ranging from 8.8 to 278 ng L-1 and decreased in the deepest sampling locations (5 to 6 m) ranging from 3.7 to 37.5 ng L-1. In contrast, PFHxA showed increasing concentrations with depth, ranging from 36.5 to 362.4 ng L-1 in the vadose zone of the DST layer, and from 239.3 to 1684.0 ng L-1 in the deepest sampling locations (5 to 6 m). Concentrations of the target emerging contaminants (ECs) in the surrounding surface waters were substantially lower than those observed in groundwater under the cover system. Caffeine (CAF) showed the highest concentrations adjacent to the tailings area (up to 15.4 ng L−1), followed by IBU (up to 3.5 ng L−1) and NAP (up to 2.0 ng L−1). Concentrations in effluent from the final polishing pond were lower: CAF (2.1–6.3 ng L−1), IBU (1.0–2.2 ng L−1), and NAP (<LOQ–0.2 ng L⁻¹). The highest concentrations of PFOA (up to 8.5 ng L⁻¹) and PFOS (up to 7.2 ng L−1) were adjacent to the tailings area, with lower concentrations observed in the final effluent (PFOA up to 4.9 ng L−1; PFOS up to 3.7 ng L−1). The overall low concentrations of PPCPs and PFASs may indicate that groundwater containing these compounds may not have yet reached the surface water system. Alternatively, dilution, sorption, photodegradation and/or biodegradation processes may have lowered concentrations of the target compounds. Good agreement between observed concentrations obtained using discrete sampling and time-weighted passive samplers (DGT and POCIS) indicates the reliability and efficacy of passive sampling techniques for the detection and monitoring of trace elements, PPCPs, and PFASs in dynamic mine water settings, particularly at trace concentration levels near LOQs. DGT and POCIS can be complementary tools for contaminant monitoring in mine settings. To remove PFASs from aqueous solutions, a novel biochar-supported bimetallic nanocomposite (nZnNi-BC) was synthesized via the pyrolysis of oak-derived biochar impregnated with Zn and Ni sulfate for reductive degradation. Batch experiments conducted at 50°C achieved rapid, near-complete removal of both PFOA and PFOS (>95% within 480 hours), which followed a double first-order in parallel (DFOP) kinetic mechanism. The defluorination process approached 44.7±0.5% at 40 minutes and was accompanied by the formation of various shorter-chained PFASs and possible generation of organic acids (e.g., formic and acetic acid), indicating concurrent decarboxylation and defluorination pathways. Fluoride recovery, as the sum of the total inorganic and organic fluorine, reached a maximum of 67.4±1.5% at 40 min and decreased to 37.6±1.1% at 240 h. Given the near-complete removal of both PFOA and PFOS and low masses of generated short-chained PFASs (C4-C6), it is likely that PFOA and PFOS were transformed into PFBA, PFPeS and/or ultrashort-chain PFASs (C<4) that were not analyzed in the study. Alternatively, geochemical analysis indicated a potential fluoride sink via precipitation of fluorapatite associated with Ca and phosphate release from biochar. These outcomes demonstrate the feasibility of nZnNi-BC as an effective material for PFAS degradation and highlights the importance of geochemical sinks in controlling fluoride mass balance.