Plumbojarosite formation in contaminated soil to mitigate childhood exposure to lead, arsenic and antimony

Abstract

In this study, a novel method for lead (Pb) immobilization was developed in contaminated soils using iron (III) (Fe³⁺) in conjunction with 0.05 M H₂SO₄. During method optimization, a range of microwave treatment times, solid to solution ratios, and Fe₂(SO₄)₃/H₂SO₄ concentrations were assessed using a mining/smelting impacted soil (BHK2, Pb: 3031 mg/kg), followed by treatment of additional Pb contaminated soils (PP, Pb: 1506 mg/kg, G10, Pb: 2454 mg/kg and SoFC-1, Pb: 6340 mg/kg) using the optimized method. Pb bioaccessibility was assessed using USEPA Method 1340, with Pb speciation determined by X-ray Absorption (XAS) spectroscopy. Treatment efficacy was also validated using an in vivo mouse assay, where Pb accumulation in femur, kidney and liver was assessed to confirm in vitro bioaccessibility outcomes. Results showed that Pb bioaccessibility could be reduced by 77.4–97.0% following treatment of soil with Fe₂(SO₄)₃ (0.4–1.0 M), H₂SO₄ (0.05 M) at 150 °C for 60 min in a closed microwave system. Results of bioavailability assessment demonstrated treatment effect ratio of 0.06–0.07 in femur, 0.06–0.27 in kidney and 0.06–0.11 in liver (bioavailability reduction between 73% and 93%). Formation of plumbojarosite in treated soils was confirmed by XAS analysis.

1. Introduction

Lead (Pb) is a ubiquitous developmental neurotoxin (Reuben et al., 2017); the link between early childhood Pb exposure (~3–10 μg Pb/dL of blood) and cognitive development impairment in children is well-established (Chiodo et al., 2004; Lanphear et al., 2005; Maloney et al., 2018; Reuben et al., 2017). Although diet (Pb contaminated food and water) may contribute to Pb exposure (Carrington et al., 2019), the Agency for Toxic Substances and Disease Registry (ATSDR, 2019) considers incidental ingestion of Pb contaminated soil as the primary pathway for childhood Pb exposure. In addition to impairing neurological and cognitive capacity in children, Pb may exert wide ranging systemic toxicity (ATSDR, 2019), highlighting the need to reduce exposure from incidental soil ingestion.

Traditional approaches of Pb exposure mitigation include soil excavation, removal, capping and in situ amendment. Among the notable in situ Pb immobilization strategies are amending soil with organic materials [e.g. compost, manure, biochar, biosolids] (Zhang and Hay, 2020) and inorganic materials [e.g. phosphorus (P) and iron (Fe) containing compounds (Bradham et al., 2018; Mele et al., 2015), phosphate amended cement (Wang et al., 2018), lime (Hamid et al., 2020), industrial waste, such as fly ash (Tang et al., 2020), clay (Gu et al., 2020), quartz sand (Li et al., 2020a), and nanoparticles (Fajardo et al., 2020)].

Although strategies for the remediation of Pb contaminated soil have been developed as detailed above, each strategy represents different limitations for field application. For example, soil excavation/replacement, capping or using heavy mulch to suppress soil resuspension requires supply of clean materials and/or disposal of contaminated material, which may be unsustainable (Henry et al., 2015; Karna et al., 2017; Palansooriya et al., 2020). Similarly, use of biosolid, municipal soil waste and fly ash to immobilize Pb may increase the load of potentially toxic elements.

According to Palansooriya et al. (2020), lime, biochar and phosphate amendments are the most effective for in situ Pb immobilization to date. Liming relies on an increase in soil pH, causing Pb to precipitate. However, soil pH tends to reduce over time, requiring multiple liming applications (Palansooriya et al., 2020). Biochar application to soil may result in increased mobility of toxic co-contaminants, for example, arsenic (As) and antimony (Sb) (Gu et al., 2018; Igalavithana et al., 2017; Li et al., 2020b). Therefore, caution is recommended when biochar is applied to metalloid co-contaminated soils (Palansooriya et al., 2020). Similarly, although phosphate amendments [e.g. phosphoric acid (PA), triple super phosphate (TSP)] have been suggested as a promising approach to mitigate Pb exposure (Bradham et al., 2018; Cao et al., 2009; Juhasz et al., 2014; Kastury et al., 2019b; Li et al., 2017), this strategy has not yet been applied successfully during large scale soil remediation (Karna et al., 2020). Phosphate amendments rely on changing Pb speciation in situ into less soluble forms (e.g. pyromorphite and tertiary Pb phosphate), thereby reducing Pb bioavailability (Pb absorption into the systemic circulation) and/or Pb bioaccessibility (Pb dissolution in simulated gastrointestinal solution, e.g. using USEPA method 1340) (Scheckel et al., 2013; USEPA, 2017; Bradham et al., 2018). In a long term field trial conducted in Joplin, Missouri (USA), PA and TSP amended soil showed a 32–50% reduction in Pb relative bioavailability (Pb bioavailability in soil relative to the bioavailability of Pb acetate, the toxicity reference value material) sixteen years after the treatment was applied (Bradham et al., 2018). However, prior to application, the most effective phosphate type and application rate for each contaminated soil must be determined to counteract the possibility of competition with other elements [e.g. Ca, Fe, nickel (Ni) and Zn] (Cerklewski and Forbes, 1976; Gulson et al., 2019; Kastury et al., 2019a; Noonan et al., 2003). Lowering of soil pH to <4 may also be required to facilitate dissolution of Pb and phosphate to enhance the formation of pyromorphite, necessitating a secondary treatment with lime to restore soil pH (Scheckel and Ryan, 2004). Additionally, to prevent excess phosphate run-off into water bodies, environmental monitoring may be required when utilizing this strategy. Moreover, because of declining global phosphate supplies (Geissler et al., 2020), large scale soil amendment to mitigate Pb exposure using phosphate may not be economically feasible or practical in the future.

An alternative approach for minimizing Pb exposure through bioavailability reduction may be the incorporation of Pb into Fe minerals to form plumbojarosite [PbFe₆(SO₄)₄(OH)₁₂] (Karna et al., 2020). A member of the alunite super group, plumbojarosite has been identified in galena rich oxidized sulphidic ore (Smith et al., 2006), metallurgical waste [e.g. waste from zinc (Zn) industries] (Kerolli-Mustafa et al., 2015) and acid rock or mine drainage system (Hochella et al., 1999). It is sparingly soluble, showing solubility constant of 10^−26.2 (Lindsay, 1979) and may be formed by reacting Pb with ferric sulfate [Fe₂(SO₄)₃] and sulfuric acid [(H₂SO₄)] solution at 95–100 °C (Dutrizac et al., 1980; Karna et al., 2020). Thermodynamic and dissolution studies using synthetic plumbojarosite suggest that it is stable at temperatures <667 °C (Frost et al., 2010) with low solubility at acidic (pH 2) and alkaline (pH 8) conditions (Smith et al., 2006).

A ‘slow addition technique’ for plumbojarosite formation was detailed by Dutrizac et al. (1980), where Pb(NO₃)₂ was added at 1.8 mL/h to 0.1 M Fe₂(SO₄)₃−0.01 M H₂SO₄ solution at 95–100 °C with stirring. Karna et al. (2020) examined a slight modification of the ‘slow addition technique’ using soil-free pure Pb compounds (PbCO₃, PbCl₂, Pb₃(PO₄)₂, PbSO₄), clean topsoils spiked with Pb minerals, as well as naturally contaminated soils. The study conducted by Karna et al. (2020) differed from Dutrizac et al. (1980) only in the sense that Fe₂(SO₄)₃ - H₂SO₄ was being supplied to either Pb compounds or soil to represent the conditions of a field application, where a large solution volume was needed for treatment. Karna et al. (2020) found that >90% of all Pb-species were converted to plumbojarosite within 5 h with a corresponding >90% reduction in Pb bioavailability relative to untreated soils. However, this ‘slow addition technique’ has limited application for the treatment of large quantities of Pb contaminated soil because of high water usage and time requirements (e.g. the treatment of 2.0 g of Pb(NO₃)₂ required 1 L of solution, with the addition of 1.8 mL/h 0.1 M Fe₂(SO₄)₃−0.01 M H₂SO₄ to Pb(NO₃)₂ for 67 h).

This research aimed to develop a Pb immobilization strategy for Pb-contaminated soil via the formation of plumbojarosite, so that the treatment may be scaled up and applied to large quantities of soil in the shortest amount of time, using the smallest volume of water. To demonstrate the effectiveness of treatment strategies, in vitro bioaccessibility assay was conducted, followed by validation using in vivo assessment using a mouse bioassay. Additionally, changes in Pb speciation (i.e. plumbojarosite formation) in treated soil was determined using X-ray absorption spectroscopy (XAS) analysis as a multiple lines of evidence approach for treatment efficacy.

2. Material and methods

2.1. Lead contaminated soil

A Pb contaminated soil from Broken Hill, Australia (BHK2, <250 μm soil particle fraction) was utilized during the optimization of treatment parameters for plumbojarosite formation [e.g. treatment time, solid to solution ratio, H₂SO₄ and Fe₂(SO₄)₃ concentrations]. Subsequently, using the optimized treatment parameters, three additional Pb contaminated soils (<250 μm soil particle fraction) were utilized to assess treatment efficacy [smelter impacted (PP), munition impacted (G10) and a reference material (SoFC-1).

2.2. Assessment of pseudo-total elemental concentration, particle size and carbon content

The pre- and post-treated soil (0.5 g) was digested using aqua-regia (5 mL) in a CEM Mars 6 microwave following USEPA Method 3051 (USEPA, 2007). A standard reference material (SRM) from National Institute of Standards and Technology (NIST 2710a) was used to calculate the accuracy of digestion to assess the suitability of the digestion method for the element of interest (i.e. Pb, more information in Section 2.7. Quality assurance and quality control). After digestion, the solution was made up to 50 mL with MilliQ water (>18 MΩ), syringe filtered (0.45 μm, cellulose acetate) and stored at 4 °C until analyzed by ICP-OES (USEPA, 1998a).

Particle size was determined following dispersion of 5 mg of soil in 50 mL NaCl (0.1 M) overnight (to prevent aggregates) and analysis using Accusizer 780 (NICOMP) (Kastury et al. 2020a, 2020b). Total carbon content in pre- and post-treated soil (0.2 g, n=2) was analyzed using a LECO TruMac CNS. Organic carbon was removed from pre- and post-treated soil (0.2 g, n=2) using dropwise addition of sulfur dioxide (H₂O₃S) solution (≥6%) until bubble evolution ceased for two consecutive additions (Kastury et al., 2020b). Analysis of carbon content was undertaken using LECO TruMac CNS and total organic carbon as the difference between total carbon and total inorganic carbon. Sample pH was determined using m (g) : v(mL) = 1:5 using MilliQ water (>18 MΩ) (Kastury et al. 2019a, 2019b).

2.3. Formation of plumbojarosite in Pb contaminated soil

The ‘slow addition method’ of plumbojarosite formation described first in Dutrizac et al. (1980) and utilized by Karna et al. (2020) was modified in this study to minimize treatment time and solution volume so that treatment can be scaled up during field application. A closed reaction vessel system (CEM Mars 6 microwave) was utilized to apply heat to samples by ramping to 150 °C for 10 min, followed by holding at 150 °C for 30 min at 800 psi (MARS 6 Method Notes, 2020). This treatment method was chosen to achieve quick dissolution of Pb phases in contaminated soil (i.e. 1 h) for rapid assessment of treatment efficacy of multiple combination of experimental parameters.

During treatment optimization, three experimental parameters were varied: treatment time, solid to solution ratio and concentrations of Fe³⁺ [in the form of Fe₂(SO₄)₃] and H₂SO₄. After treatment, soil was washed three times with MilliQ water (>18 MΩ) and dried (40 °C), followed by confirmation of total Pb concentration (see Section 2.2) and assessment of Pb bioaccessibility using USEPA Method 1340 (USEPA, 2017). This post treatment analytical approach was utilized for all subsequent method modification to assess the efficacy of Pb immobilization (2.3.1–2.3.2, 2.4).

2.3.1. Treatment time

BHK2 was treated in a final solution volume of 10 mL of H₂SO₄ (0.01 M) and Fe₂(SO₄)₃ (0.1 M), with a solid to solution ratio=1:10, for 15, 30, 60 and 120 min

2.3.2. Solid to solution ratio

Using the results from the above step (2.3.1), a treatment time of 60 min was chosen to investigate the influence of solid to solution ratio on plumbojarosite formation and Pb bioaccessibility reduction. Solid to solution ratio was varied between 1:10–1:1, while keeping all other treatment parameters constant (i.e. 1.0, 2.5, 5.0, 7.5 and 10.0 g of BHK2 soil was treated in 10 mL of solution containing 0.01 M H₂SO₄+0.1 M Fe₂(SO₄)₃).

2.3.3. Concentrations of H₂SO₄ and Fe₂(SO₄)₃

A treatment time of 60 min and a concentration of 0.05 M H₂SO₄ was used in this experiment, while solid to solution ratio and Fe₂(SO₄)₃ concentrations were varied. Solid to solution ratio was varied by using three different masses of soil (2.5, 5.0 and 10.0 g) in a final solution volume of 10 mL. For each solid to solution ratio, Fe₂(SO₄)₃ concentrations that were tested were 0.10, 0.20, 0.30, 0.40, 0.50 and 1.0 M Fe₂(SO₄)₃.

2.4. Treatment with additional soils

To assess if the treatment is effective for other Pb contaminated soils with different Pb concentration, speciation and physico-chemical properties, three additional Pb contaminated soils (e.g. G10, PP and SoFC-1), were treated at 150 °C for 60 min using 0.05 M H₂SO₄, 0.10–1.00 M Fe₂(SO₄)₃ and solid to solution ratio of 1:1 (i.e. 10 g soil in 10 mL of solution).

2.5. Assessment of Pb bioaccessibility

Lead bioaccessibility in all soils and As and Sb bioaccessibility in SoFC-1 was assessed using USEPA Method 1340 (USEPA, 2017) (gastric phase extraction) with metal(loid) analysis conducted using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS) following USEPA Method 6020A (USEPA, 1998). Lead, As and Sb bioaccessibility was calculated according to Eq. (1).

$$\begin{gathered} Invitrometal(loid)bioaccessibility(\%) \\ =\left(\frac{Invitrometal(loid)}{Pseudo-totalmetal(loid)}\right)x100 \end{gathered}$$ (1)

where:

In vitro metal(loid) = metal(loid) extracted (mg/kg) during the in vitro assay.

Pseudo-total metal(loid) = metal(loid) concentration (mg/kg) in contaminated soil determined using aqua regia digestion.

2.6. Assessment of Pb bioavailability

Female C57BL/6 mice (4–6 week old) were utilized during in vivo studies with experimental and ethical approval provided by the South Australian Health and Medical Research Institute Animal Ethics Committee (application number SAM268). Animal care was compliant with the Standard Operating Procedures of the South Australian Health and Medical Research Institute, and the Guidelines for the Care and Use of Laboratory Animals in the National Research Council (NRC, 2010). Briefly, mice were housed in three metabolic cages or operational units (4 per cage; 12 per treatment) and supplied with AIN93G mouse chow amended with untreated or treated Pb-contaminated soil (1% w/w) for a period of 9 days. Food consumption was monitored daily by determining the weight difference of the food hopper after filling and before replenishment while cumulative food consumption was the sum of the daily food consumption. A control group of mice was also included which consumed unamended AIN93G mouse chow. At the end of the exposure period, mice were maintained on unamended AIN93G mouse chow for an additional 24 h after which they were humanely euthanized. Targeted tissues (femur, kidney, liver) were collected as Pb bioavailability endpoints. Tissues from each operational unit (n=3) were combined, frozen (−20 °C), then freeze-dried (Modulyod Freeze Dryer) prior to digestion. Pooled tissue from each operational unit were pre-digested with 70% HNO₃ overnight. 5 mL of HNO₃ was used for femur (0.1 - 0.3 g) and kidney (0.1 - 0.28 g), while 10 mL HNO₃ was used for liver because of the higher mass of tissue (0.7 - 1.1 g). Digestion was conducted using a CEM Mars 6 Microwave (ramped up to 150 °C for 10 min, held at 150 °C for 30 min at 800 psi). NIST SRM 2976 was used to assess the accuracy of tissue digestion (more information regarding digestion accuracy is given in Section 2.7. Quality assurance and quality control). Following digestion, samples were diluted with Milli-Q water (to 50 mL), syringe filtered (0.45 μm, cellulose acetate) and stored at 4 °C until analyzed by ICP-MS. To determine the impact of treatment on As and Sb exposure, urinary excretion was utilized as the bioavailability endpoint for mice that received pre and post treated SoFC-1. Urine was vortexted thoroughly to mix and 15 mL was digested using 70% HNO₃ (5 mL) overnight, followed by digestion using CEM Mars 6 Microwave (ramped up to 150 °C for 10 min, held at 150 °C for 30 min at 800 psi). Post-digestion processing and analysis of As and Sb was done using ICP-MS as detailed above.

2.7. Assessment of Pb speciation

X-ray absorption spectroscopy (XAS) data were collected at the Materials Research Collaborative Access Team (MRCAT) beamline 10-ID, Sector 10, at the Advanced Photon Source (APS) of the Argonne National Laboratory (ANL), United States of America. Further details of the methodology are given in the Supplementary Information (SI).

2.8. Data and statistical analysis

Arsenic and Sb bioavailability was assessed as Eq. (2):

$$\text{Urinary excretion}factor=\left(\frac{\text{Cumulative urinary}excretion}{\text{Cumulative dietary}intake}\right)$$ (2)

Lead, As and Sb bioaccessibility and bioavailability determined in pre and post treated soils were utilized to calculate treatment effect ratio (TER, Eq. (3)). Values <1 indicated that Fe₂(SO₄)₃/H₂SO₄ treatment was effective at reducing metal(loid) bioaccessibility/bioavailability, with lower values indicating greater efficiency.

$$\text{TER}=\left(\frac{{\text{Bioaccessibility/bioavailability}}_{\text{treated soil}}}{{\text{Bioaccessibility/bioavailability}}_{\text{untreated soil}}}\right)$$ (3)

where:

Bioaccessibility/bioavailability _{treated soil} = bioaccessibility (USEPA 1340) or bioavailability (Pb accumulation in femur, kidney, liver and As/Sb urinary excretion factor) in contaminated soil following Fe₂(SO₄)₃ treatment.

Bioaccessibility/bioavailability _{untreated soil} = bioaccessibility (USEPA 1340) or bioavailability (Pb accumulation in femur, kidney, liver and As/Sb urinary excretion factor) in contaminated soil.

The statistical difference between Pb bioavailability in untreated and remediated soil (accumulation in femur, kidney and liver) was conducted by t-test (α=0.05).

2.9. Quality assurance and quality control

The accuracy of soil and tissue digestion was confirmed using NIST SRM 2710a (certified Pb reference value of 5520 mg/kg) and NIST SRM 2976 (certified Pb reference value of 1190 mg/kg). Quantitative recovery of Pb during soil digestion from SRM 2710a was 106% (n=6) and for tissue digestion from SRM 2976 was 116% (n=6). During analysis of soil digests with ICP-OES, deviation from duplicates (n=12) was 2.6% and recovery from check values (n=24) and spiked samples (n=12) were 91.3–108.71%, and 93.1% respectively. Additionally, during Pb concentration analysis in tissues using ICP-MS, deviations from duplicates (n=5) were 1.16% and recoveries from check values (n=12) and spiked samples (n=6) were 98.9–100.8% and 91.2% respectively.

3. Results and discussion

3.1. Physicochemical characterization of Pb contaminated soils

Table 1 lists major and trace elements in the <250 μm soil particle fraction. The concentration of Pb in the four samples used in this study were 1506 mg/kg in PP (smelter impacted), 2454 mg/kg in G10 (munition impacted), 3124 mg/kg in BHK2 (mining impacted), 6340 mg/kg (reference material for the assessment of bioaccessibility assays). All four soils exhibited similar Fe concentration prior to treatment (31,133–37,531 mg/kg). Major cations which may compete for inclusion within jarosite structures include monovalent elements such as potassium (K), and divalent elements such as calcium (Ca) (Karna et al., 2020). Potassium concentration in the 4 soils was variable ranging from 2475 mg/kg (SoFC-1) to 11,392 mg/kg (PP). Calcium concentrations was lowest in SoFC-1 (2172 mg/kg), while the other three soils exhibited a similar range (17,357–20,906 mg/kg). Average particle size in the <250 μm soil particle fraction ranged from 1.52±0.01 μm (G10) to 2.12±0.17 μm (PP). Soil pH of the <250 μm soil particle fractions were 6.8±0.03 in BHK2, 8.02±0.01 for PP and 7.4±0.01 for G10, while pH of SoFC-1 was not determined due to low supply (Table 1). All samples contained low total organic carbon content (0–0.61%) (Table 2).

Table 1.

Major and minor elemental concentrations in the <250 μm soil particle fraction of Pb-contaminated soils.

Soil		BHK2	SoFC1	G10	PP
Elemental concentration (mg/kg) a
Pb		3124	6340	2454	1506
Fe		31133	37531	35242	32277
Al		22990	9669	28345	38089
As		26.1	726	11.6	55.8
Ba		145	124	338	126
Ca		20906	2172	17357	19959
Cd		9.9	55.8	1.5	6.4
Co		12.3	6.1	21.4	14.5
Cr		27.1	14	61.2	42.6
Cu		67.7	48.5	543	59.4
K		6861	2475	3464	11392
Mg		5591	3289	5618	15932
Mn		3892	2487	524	842
Ni		15.8	8.1	60.4	17.3
Sb		0.1	1030	3.4	21.2
Zn		2443	8569	227	1560
pH		6.8 ± 0.03	No data	7.4 ± 0.01	8.0 ± 0.01
Particle Size (μm)	Mean ± SEM	2.06 ± 0.74	1.68 ± 0.14	1.52 ± 0.01	0.93 ± 0.01
	Median ± SEM	1.10 ± 0.08	1.07 ± 0.03	2.12 ± 0.17	1.15 ± 0.01

^aValues represent the mean of triplicate analyses; the standard deviation of means was <5%.

Table 2.

Total and organic carbon in pre and post treated soils.

Soil	Untreated		Fe2(SO4)3 treateda
	Total carbon (%)	Total organic carbon (%)	Total carbon (%)	Total organic carbon (%)
BHK2	1.87±0.13	0.60±0.13	0.87±0.02	0.03±0.02
SoFC1	1.93±0.04	0.00	1.02±0.03	0.07±0.02
G10	7.22±0.01	0.28±0.01	5.72±0.08	0.26±0.08
PP	2.62±0.06	0.61±0.06	1.52	0.26

^aUsing a solid: solution ratio 1:1, BHK2, SoFC1 and G10 was treated with 1.0 M Fe2(SO4)3–0.05 M H2SO4, while PP was treated with 0.5 M Fe2(SO4) 3–0.05 M H2SO4.

3.2. Optimizing the formation of plumbojarosite in Pb contaminated soil

A mining/smelting impacted soil (BHK2, 3124±332 mg/kg of Pb, Table 1) was utilized to optimize plumbojarosite formation. During initial experiments, treatment time was varied from 15 to 120 min, whilst keeping the solid to solution ratio at 1:10 and concentrations of H₂SO₄ and Fe₂(SO₄)₃ (0.01 and 0.1 M respectively) according to Dutrizac et al. (1980). Fig. 1A shows changes in Pb bioaccessibility (mg/kg) with increasing treatment time, while Fig. 1B shows complementary Pb speciation (weighted %) data for untreated BHK2 soil, and soil treated for 15 and 60 min. When untreated BHK2 was assessed using gastric phase extraction (pH 1.5), Pb bioaccessibility was 91.5% (2857±89.3 mg/kg). However, following 15 min treatment, Pb bioaccessibility was reduced to 6.1% (190±3.7 mg/kg), which further decreased to 5.1% (160±9.58 mg/kg), 4.2% (132±5.55 mg/kg) and 2.8% (87.3±1.82 mg/kg) after treatment of 30, 60 and 120 min respectively.

Fig. 1.

Effect of varying treatment time on Pb bioaccessibility (A) and speciation (B) using BHK2. The red line represents a Pb concentration of 300 mg/kg, corresponding to the Health based Investigation Level (HIL) a, set by the National Environment Protection Measure for the Assessment of Site Contamination (NEPM 2013).

Analysis of Pb speciation using XAS identified significant changes in the weighted % of Pb as a result of H₂SO₄/Fe₂(SO₄)₃ treatment. In untreated soil, the predominant Pb phase in BHK2 was magnetoplumbite [(Pb,Mn)(Fe,Mn)₁₂O₁₉, 65%], with anglesite (PbSO₄) and Pb orthophosphate [Pb₃(PO₄)₂] consisting of 21% and 15% respectively. However, following treatment for 15 min, 84% of Pb was converted to plumbojarosite, with 16% remaining as anglesite. Extending the treatment time to 60 min resulted in only plumbojarosite being identified. Anglesite is a weathering product of galena (PbS) and its formation during the synthesis of plumbojarosite may indicate the incomplete conversion of Pb into plumbojarosite (Dutrizac et al., 1980; Smith et al., 2006). In the context of reducing Pb exposure, anglesite formation is unfavorable due to its propensity to dissolve under acidic conditions found in the stomach (Zhang and Ryan, 1998), which would become potentially available for absorption into the systemic circulation as shown in Kastury et al. (2019b). As a result, outcomes from the time variable experiment suggest that a treatment time of 60 min may be appropriate to reduce anglesite formation, while maximizing plumbojarosite formation in treated soils.

Outcomes from the initial experiment also demonstrated that plumbojarosite formation in BHK soil was efficient at a solid to solution ratio of 1:10. However, smaller solid to solution ratios would be preferred to minimize water usage per unit volume of soil treated and to minimize post remediation water treatment. As a consequence, the efficiency of the treatment process was assessed using varying soil masses (1–10 g), solution volumes [0.1 M Fe₂(SO₄)₃, 0.01 M H₂SO₄] and a treatment time of 60 min Fig. 2A shows that as soil mass was increased, Pb bioaccessibility increased, with a corresponding reduction in the percentage of plumbojarosite formed during the treatment process (Fig. 2B). For example, using a solid to solution ratio of 1:10, Pb bioaccessibility was 4.2% (131±5.5 mg/kg) with only plumbojarosite detected in post treatment residues. However, when a solid to solution ratio of 1:5 was used, Pb bioaccessibility increased to 49.6% (1550±63.0 mg/kg), with 40.7% of Pb being present as plumbojarosite (Fig. 2B). Consistent with this trend, when a solid to solution of 1:1 was used, Pb bioaccessibility increased to 78.5% (2453±80.0 mg/kg) and only 12% of Pb was converted to plumbojarosite. Fig. 2B suggests that anglesite precipitation increased in the post-treated soil at higher solid to solution ratios, for example, 17% and 36% anglesite was observed when 2.5 and 5 g of soil was used respectively. In addition, the majority of Pb precipitated as ferrihydrite sorbed Pb [Fe³⁺₂O₃•0.5(H₂O)] at the highest soil to solution ratio; 60% ferrihydrite sorbed Pb was observed when 10 g of soil was used. The results suggested that at high solid to solution ratio, Pb:Fe ratio was not conducive for plumbojarosite formation.

Fig. 2.

Effect of varying solid (g): solution (mL) ratio on Pb bioaccessibility (A) and speciation (B) using BHK2. The solution volume was kept constant at 10 mL, while the soil mass was varied between 1 and 10 g. The red line represents a Pb concentration of 300 mg/kg, corresponding to the Health based Investigation Level (HIL) a, set by the National Environment Protection Measure for the Assessment of Site Contamination (NEPM 2013).

To overcome potential issues associated with the depletion of Fe³⁺, the impact of varying concentrations of Fe₂(SO₄)₃ (0.1–1.0 M) on plumbojarosite formation was assessed using soil masses of 2.5, 5 and 10 g (solid to solution ratios of 1:4, 1:2 and 1:1 respectively). Additionally, H₂SO₄ concentration was also increased to 0.05 M to ensure that the pH was sufficiently low (~1.5–1.6) for efficient Pb solubilization. This was in line with Dutrizac et al. (1980) who suggested that acid content of up to 0.1 M will increase plumbojarosite content in the residual mass. Fig. 3 shows the results of varying Fe₂(SO₄)₃ concentration and solid to solution ratios for BHK2 (top panel: 2.5 g, middle panel: 5 g and bottom panel: 10 g) on Pb bioaccessibility outcomes. At all three solid to solution ratios, a decrease in Pb bioaccessibility was observed with increasing Fe₂(SO₄)₃ concentration, confirming that an insufficient Fe³⁺ supply was the limiting factor for plumbojarosite formation at high soil masses. For example, to achieve >90% reduction in Pb bioaccessibility with 2.5 g of soil, 0.2 M Fe₂(SO₄)₃ was required (Fig. 3A), while to achieve the same bioaccessibility result with 5 and 10 g of soil, 0.3 M Fe₂(SO₄)₃ and 0.5 M Fe₂(SO₄)₃ was required respectively (Fig. 3C and E).

Fig. 3.

Effect of varying Fe₂(SO₄)₃ concentration (0.1–1.0 M) on Pb bioaccessibility and speciation in BHK2. Solid to solution ratio was kept constant at 1:1, while varying the soil mass between 2.5 g (A), 5 g (B) and 10 g (C). The red line represents a Pb concentration of 300 mg/kg, corresponding to the Health based Investigation Level (HIL) a, set by the National Environment Protection Measure for the Assessment of Site Contamination (NEPM 2013).

Although anglesite was identified at low concentrations in treated soil (e.g. 10.7–16.2% using 2.5 g soil in Fig. 3B and 30.0% using 5 g soil in Fig. 3B), once treatment conditions were optimized, the % of anglesite in treated soils appeared to be independent of Pb bioaccessibility results (i.e. no trend was observed relating the % of anglesite to Fe₂(SO₄)₃ concentration or solid to solution ratio). These results suggest that encapsulation of anglesite by plumbojarosite may have occurred during treatment, which served to protect it from dissolution under simulated gastric phase solutions (Dutrizac et al., 1980). Similarly, although Fig. 3F showed decreasing % of ferrihydrite sorbed Pb with increasing Fe₂(SO₄)₃ concentration, Fig. 3D identified 9% ferrihydrite sorbed Pb when 0.50 M Fe₂(SO₄)₃ was used to treat 5 g of soil without affecting Pb bioaccessibility, which may also be a result of this Pb species being encapsulated with plumbojarosite.

These results suggest that a treatment time of 60 min, a solid to solution ratio of 1:1, using 1.0 M Fe₂(SO₄)₃ and 0.05 M H₂SO₄ may be optimum treatment parameters to maximize conversion of Pb in BHK2 into plumbojarosite and for Pb bioaccessibility reduction. However, depending on soil physicochemical characteristics (e.g. the concentration of Pb), variable concentrations of Fe₂(SO₄)₃ may be required to maximize plumbojarosite formation.

3.3. Application of plumbojarosite formation treatment strategy in other Pb contaminated soils

Following optimization of treatment parameters in BHK2, three additional Pb contaminated soils (<250 μm soil particle fraction) containing variable Pb concentrations and speciation were utilized to assess treatment efficacy for reducing Pb bioaccessibility. Soil designated PP, collected from Port Pirie, South Australia, had a Pb concentration of 1506 mg/kg arising from smelter impacts. Three predominant Pb phases were identified in PP including anglesite (14%), Pb orthophosphate (27%) and magnetoplumbite (59%). Soil designated G10 was collected from a former munition plant in Victoria. Total Pb concentration in G10 was 2454 mg/kg, while ferrihydrite sorbed Pb (90%) and plumbojarosite (11%) were the predominant Pb phases. Soil designated SoFC-1 was collected from Flat Creek Iron Mountain (USA), and contained 6340 mg/kg of Pb, with anglesite (47%) and ferrihydrite sorbed Pb (53%) being the predominant Pb species. Pb bioaccessibility in untreated PP, G10 and SoFC-1 was 1404±1.45 mg/kg (93.2%), 1983±50.0 mg/kg (80.8%), and 4767±20.4 mg/kg (75.2%) respectively. As these soils contained a wide range of Pb concentrations, Fe₂(SO₄)₃ additions during treatment were varied (0.1–1.0 M) in order to determine optimum treatment parameters for each soil.

Fig. 4 shows reduction in Pb bioaccessibility and changes in Pb speciation following treatment (60 min, 150 °C, solid to solution ratio of 1:1, 0.05 M H₂SO₄ 0.1–1.0 M Fe₂(SO₄)₃ concentrations): PP (Fig. 4A, B), G10 (Fig. 4C, D) and SoFC-1 (Fig. 4E, F). Outcomes of treatability studies were similar to those observed for BHK2 whereby lower Pb bioaccessibility was observed in treatments with increasing Fe₂(SO₄)₃ concentration (biphasic response) as a consequence of Pb speciation trending towards plumbojarosite formation. In PP, a minor reduction in Pb bioaccessibility, from 93.2% (1404 mg/kg) to 88.2% (1327 mg/kg), was observed when 0.1 M Fe₂(SO₄)₃ was utilized. However, Fig. 4A shows that when Fe₂(SO₄)₃ was increased to 0.2 and 0.4 M, Pb bioaccessibility was reduced to 37.1% (559 mg/kg) and 13.3% (200 mg/kg) respectively, with a corresponding 100% plumbojarosite formation when 0.4 M Fe₂(SO₄)₃ was utilized (Fig. 4B).

Fig. 4.

Pb bioaccessibility and speciation in PP (A and B), G10 (C and D) and SoFC-1 (E and F) using a treatment time of 60 min, Solid to solution ratio of 1:1 and 0.05 M H₂SO₄, while Fe₂(SO₄)₃ concentration was varied from 0.1 M to 1.0 M. The red line represents a Pb concentration of 300, corresponding to the Health based Investigation Level (HIL) a, set by the National Environment Protection Measure for the Assessment of Site Contamination (NEPM 2013).

Similarly, compared to untreated soil, Pb bioaccessibility in G10 and SoFC-1 was reduced to 75.4% (1851 mg/kg) and 37.8% (259 mg/kg) respectively when 0.1 M Fe₂(SO₄)₃ was used in the treatment process (Fig. 4C and E). However, when Fe₂(SO₄)₃ concentration was increased to 0.2 M, Pb bioaccessibility was reduced significantly in both soils (34.7% in G10 and 28.8% in SoFC-1). Speciation analysis showed that when using 0.2 M Fe₂(SO₄)₃, 78% of Pb in G10 was present as plumbojarosite, with 22% remaining as ferrihydrite sorbed Pb (Fig. 4B). Similarly, 75% of Pb in SoFC-1 was present as plumbojarosite when treated with 0.2 M Fe₂(SO₄)₃, with 9% and 16% present as ferrihydrite and anglesite respectively. When treatment utilized the highest Fe₂(SO₄)₃ concentration (1.0 M), Pb bioaccessibility in SoFC-1 was reduced to 4.1% (259 mg/kg), although both ferrihydrite sorbed Pb (10%) and anglesite (17%) were present in treated soil (plumbojarosite being the major Pb species; 72%). These results are similar to BHK2, where encapsulation of anglesite and ferrihydrite sorbed Pb by plumbojarosite may have limited its dissolution under simulated gastric phase conditions. However, the trend of Pb bioaccessibility reduction in G10 appeared to plateau at Fe₂(SO₄)₃ concentrations >0.4 M, with a Pb bioaccessibility value of 18.3% (449 mg/kg) when treated with 1.0 M Fe₂(SO₄)₃.

Although the reason for the discrepancy in treatment efficacy between G10 and the other three soils is unclear, Rasmussen et al. (2011) suggested that Pb associated with organic matter (e.g. Pb citrate and Pb adsorbed onto humate) may influence oral Pb bioaccessibility using the gastric phase of Solubility Bioaccessibility Research Consortium (SBRC) assay or USEPA Method 1340 assay used in this study. Similarly, a negative correlation (r:−0.7079) between the % of organic matter bound Pb in dust recovered from mining/smelting soil and inhalation bioaccessibility using simulated lung fluid (pH 7.4) has been reported (Kastury et al., 2020a). Organic components, e.g. humic acid, may bind to Pb and reduce its mobility in soil (Bolan et al., 2014); although the effect of organic matter on Pb (im)mobilization during the conditions of plumbojarosite precipitation (i.e. low pH and high temperature) has not been assessed. Therefore, analysis of total and organic matter content in pre and post treated soil was conducted to assess this. Table 2 shows that total carbon content ranged from 1.87±0.13% (BHK2) to 7.22±0.01% (G10) in untreated soil and from 0.87±0.02% (BHK2) to 5.72±0.08% (G10) in treated soil. However, all four soils exhibited low organic matter content (0.0–0.6% pre-treatment and 0.03–0.07% post-treatment with Fe₂(SO₄)₃), with G10 and PP exhibiting similar total organic matter content (0.26%) in treated soil. As a consequence, total organic matter content was unlikely to influence Pb bioaccessibility in soils treated with Fe₂(SO₄)₃, warranting future research into factors influencing the efficacy of plumbojarosite formation using Fe₂(SO₄)₃. Soil particle size is another factor, which has been reported to influence Pb bioaccessibility, with <50 μm showing higher bioaccessibility than <250 μm soil particle fraction (Juhasz et al., 2011). Table 1 shows that the four soils used in this study were similar in the number weighted average particle size (1.51 – 2.12 μm) and number weighted median particle size (0.93–1.15 μm). Therefore, although G10 displayed the smallest sizes among the four soils tested, because of the similarity of the average and median particle size ranges, particle size was not thought to influence Pb immobilization via plumbojarosite formation in this study. However, it is noteworthy that the combination of decreasing soil pH by using H₂SO₄ and high temperature treatment (150°), may deplete organic matter content; future studies should focus on approaches for restoring soil pH and organic matter content of treated soil.

3.4. Reduction in Pb bioavailability in treated soils

Although USEPA Method 1340 has been validated for the prediction of Pb relative bioavailability in mining/smelting impacted soil (Drexler and Brattin, 2007), recent studies have highlighted that uncertainties remain in its predictive capabilities for amended soils [e.g. following phosphate treatment in Kastury et al. (2019a)]. Therefore, untreated soil and treated soil showing the highest reduction in Pb bioaccessibility was administered to mice to assess treatment efficacy in vivo. After 10 days exposure to treated and untreated soil, Pb accumulation (μg/organ) in mice femur, liver and kidney were determined as bioavailability endpoint (Fig. 5). Treatment efficacy was assessed using Pb accumulation TER, which are illustrated as numbers above Pb accumulation bars in Fig. 5. Following administration of untreated soil, Pb accumulation was the greatest in femurs (0.82–1.36 μg of Pb/organ), as bones act as long term Pb deposit (ATSDR, 2019). Pb accumulation using untreated soil was similarly high in kidneys (0.80–1.14 μg of Pb/organ), while liver Pb accumulation was significantly lower (0.21–0.65 μg of Pb/organ, p<0.05) compared to the other endpoints.

Fig. 5.

Comparison of Pb accumulation in femurs, kidneys and livers following administration of pre and post treated soil to an in vivo mouse model. The numbers illustrated above the top of the green bar represents the treatment effect ratio.

Lead accumulations in the untreated BHK2 (1.03 μg in femur, 0.99 μg in kidney and 0.42 μg in liver) and PP (0.09 μg in femur, 0.80 μg in kidney and 0.21 μg in liver) was significantly higher (p<0.05) compared to treated soil. After treatment, 0.07±0.02 μg Pb was observed in femur for both BHK2 and PP (92.8% and 93.6% reduction in Pb accumulation in BHK2 and PP respectively, with a corresponding TER of 0.06 and 0.07). Comparable TERs were also observed when kidney and liver Pb accumulation was used to calculate treatment efficacy in BHK2 and PP (TER: 0.06–0.07).

Lead accumulation following administration of untreated G10 was 0.82±0.04 μg in femurs, 0.84±0.1 μg in kidneys and 0.40±0.02 μg in livers while for untreated SoFC-1 accumulation was 1.36±0.24 μg in femurs, 1.14±0.02 μg in kidneys and 0.65±0.06 μg in livers. Both G10 and SoFC-1 contained high concentrations of ‘bioavailable’ Pb species (89% ferrihydrite sorbed Pb in G10; 47% anglesite and 53% ferrihydrite sorbed Pb in SoFC-1). Lead contaminated soil containing these Pb species have been associated with high Pb relative bioavailability and bioaccessibility. For example, in a soil containing 67% anglesite, Kastury et al. (2019b) determined a Pb relative bioavailability of 47.9±2.6%, while in soil containing ferrihydrite sorbed Pb, Beak et al. (2006) reported 53–88% gastric phase Pb bioaccessibility. After treatment, Pb accumulation in G10 was reduced to 0.10±0.16 μg in femurs (TER: 0.12), 0.23±0.08 μg in kidneys (TER: 0.27) and 0.04±0.01 μg in livers (TER: 0.11), while in SoFC-1, Pb accumulation was reduced to 0.09±0.02 μg in femurs (TER: 0.07), 0.11±0.004 μg in kidneys (TER: 0.11) and 0.05±0.05 μg in livers (TER: 0.08). Although both G10 and SoFC-1 exhibited between 17% and 21% anglesite in treated soil, TERs for Pb accumulation in femurs were comparable to BHK2 and PP, where 100% plumbojarosite was present in treated soil. The similarity in treatment efficacy between soils suggests that encapsulation of bioavailable Pb species by plumbojarosite may reduce dissolution in the gastro-intestinal track and absorption/accumulation in target tissues. It also suggests that although X-ray absorption spectroscopy is considered to be a sensitive measure of Pb speciation (Scheckel and Ryan, 2004), because bulk Pb speciation measurement, X-ray absorption spectroscopy may not be able to distinguish between anglesite that is encapsulated by an insoluble Pb species such as plumbojarosite. Consequently, speciation results arising from X-ray absorption spectroscopy and other bulk speciation approaches should be interpreted in light of results obtained using in vitro bioaccessibility assays, such as USEPA Method 1340, or in vivo bioavailability assays. However, it is noteworthy that the observed reduction in Pb accumulation closely aligned with changes in Pb bioaccessibility, suggesting that USEPA Method 1340 may be utilized to predict Pb exposure reduction via plumbojarosite formation unlike phosphate amended soils (USEPA, 2017).

3.5. Effect of treatment on co-contaminant mobility and implications for remediation

Co-contamination of agricultural, shooting range or mining/smelting impacted Pb contaminated soil by other toxic elements has been frequently reported, for example, As, cadmium (Cd), manganese (Mn), Ni, Sb and Zn (Gil-Díaz et al., 2017; Karna et al., 2018; Kastury et al., 2019a, 2019b; Li et al., 2017; Mele et al., 2015; Okkenhaug et al., 2016; Qian et al., 2009). Among the four soils studied here, As and Sb were present at concentrations that may pose unacceptable risk in SoFc-1 (e.g. As: 726 mg/kg and Sb: 1030 mg/kg, Table 1). Recent studies have demonstrated that several in situ soil amendment strategies that decrease Pb bioavailability may increase the mobility of metalloids in soil. For example, phosphate amendment may result in higher As bioaccessibility or bioavailability (Li et al., 2017; Mele et al., 2015), presumably as a result of competitive sorption processes between arsenate (${\text{AsO}}_4^{3\text{-}}$) and phosphate (${\text{PO}}_4^{3\text{-}}$) anions due to their structural similarity (Arco-Lázaro et al., 2016). Similarly, increased As and Sb mobility in biochar amended soil may occur, owing to the increase in pH, release of dissolved organic carbon and competition between ${\text{HAO}}_4^{2\text{-}}$ and OH⁻ (Gu et al., 2018; Igalavithana et al., 2017). Additionally, a limited number of benchtop studies have suggested that incorporating Fe rich amendments directly into soil may reduce As bioaccessibility when assessed in SBRC assay (using gastric and small intestinal phase assays); however, these amendments were less effective at reducing Pb relative bioavailability (Mele et al., 2015). Increasing co-contaminant bioavailability during remediation of Pb contaminated soil may exacerbate childhood Pb exposure health impacts via synergistic or additive effects (Xue et al., 2017). Therefore, As and Sb bioaccessibility (gastric phase extraction) and bioavailability (urinary excretion) in treated and untreated SoFC-1 was investigated to assess changes in metalloid exposure during the application of plumbojarosite formation technique.

Table 3 shows that As bioaccessibility in SoFC-1 was reduced from 167 ± 0.85 mg/kg in untreated soil to 63.4 ± 0.30 mg/kg in treated soil (TER: 0.38), while Sb bioaccessibility in SoFC-1 was reduced from 136 ± 2.20 mg/kg in the untreated soil to 14.8 ± 0.05 mg/kg in the treated soil (TER: 0.11). When assessed using an in vivo model, As urinary excretion factor was reduced from 0.38 ± 0.12 (TER: 0.70), while Sb urinary excretion factor was reduced from 0.03 ± 0.004 – 0.02 ± 0.005 (TER: 0.66). Smith et al. (2006) and Aguilar-Carrillo et al. (2018) demonstrated that ${\text{AsO}}_4^{3\text{-}}$ may be partially incorporated into the jarosite structure, resulting in Pb-As-jarosites. Additionally, Aguilar-Carrillo et al. (2018) demonstrated that incorporation of Pb into the jarosite structure increased as the concentration of As was increased, until both As and Pb were present at equal amounts. When the concentration of As exceeds Pb, a mixture of Pb-As-jarosite, PbSO₄ and poorly crystalline ferric arsenate results (Aguilar-Carrillo et al., 2018). Because of similarity in chemical properties between As and Sb, it is conceivable that Sb may also be incorporated into the plumbojarosite structure, making this treatment applicable for Pb contaminated soil with moderate As and Sb co-contamination. Furthermore, there is evidence that several other cations with coordination number greater than 9 may also be incorporated into jarosite structures (Smith et al., 2006), e.g., mercury (Hg), cobalt (Co), Ni and silver (Ag) (Dutrizac and Chen, 1981; Dutrizac and Chen, 2004; Dutrizac and Kaiman, 1976). Therefore, this study highlights that the plumbojarosite treatment strategy for Pb immobilization may also be applied to sequester other co-contaminants and reduce their mobility in soil. Further work is necessary to optimize treatment parameters for soils containing additional co-contaminants, to ascertain their mode of incorporation and whether variability in co-contaminant concentration affect Pb immobilization efficacy.

Table 3.

Total, bioaccessible (USEPA 1340) and bioavailable (Urinary excretion factor) As and Sb concentrations in pre and post-treated SoFC-1 (solid: solution ratio=1:1, 1.0 M Fe₂(SO₄)₃−0.05 M H₂SO₄).

	As			Sb
	Total (mg/kg)	Bioaccessible (mg/kg)	Urinary excretion factor (mg/L)	Total (mg/kg)	Bioaccessible (mg/kg)	Urinary excretion factor (mg/L)
Untreated	726	167±0.85	0.55±0.08	1030	136±2.20	0.03±0.004
Treated	559±11.4	63.4±0.30	0.38±0.12	925±17.9	14.8±0.05	0.02±0.005
TER		0.38	0.70		0.11	0.66

4. Conclusions

This study demonstrated that formation of plumbojarosite in Pb contaminated soil using a solid to solution ratio of 1:1, 0.4–1.0 M Fe₂(SO₄)₃, 0.05 M H₂SO₄ is a highly efficient Pb immobilization strategy. The presence of plumbojarosite in treated soil (determined using XAS) and the concurrent reduction in Pb bioaccessibility and bioavailability indicated that plumbojarosite was sparingly soluble under gastrointestinal conditions, thereby reducing absorption following soil exposure. An important finding of this study was that the supply of Fe₂(SO₄)₃ to treat a contaminated soil requires optimization based on initial Pb concentrations. This treatment approach offers an advantage over the ‘slow addition technique’ by minimizing water requirements and treatment times, which represent a significant advancement from previous plumbojarosite formation studies. Furthermore, reduced As and Sb bioaccessibility/bioavailability in treated soil indicates that this technique is well suited for soil treatment scenarios where cocontaminants are present.

Although a closed vessel microwave system was employed in this study to rapidly heat Pb contaminated soils, it is acknowledged that the use of microwaves to treat large soil volumes for field application may not be practical. However, other low-technology thermal options may be utilized to facilitate plumbojarosite formation which would provide a pragmatic strategy for soil treatment at larger scale. Additionally, further studies are required to assess the long-term stability of plumbojarosite and the influence of cations (e.g. Na, K, Ca) and cocontaminants that may compete for inclusion in jarosite structures. Nevertheless, results from this study highlight the potential of Fe₂(SO₄)₃ induced plumbojarosite formation as a soil treatment strategy for reducing exposure to Pb and other co-contaminants.