Hydrogen Peroxide Activation with Sulfidated Zero-Valent Iron for Synchronous Removal of Cr(VI) and BPA

Abstract

In this work, the synchronous removal of Cr(VI) and bisphenol A (BPA) in a heterogeneous Fenton process with sulfidated nanoscale zero-valent iron (S-nZVI) as the reductant and catalyst was systematically evaluated. Compared to other systems including S-nZVI or H₂O₂ alone, a simultaneous BPA degradation and Cr(VI) removal could be achieved in the S-nZVI/H₂O₂ system at an optimum pH of 3. It was, interestingly, found that 7.8% of BPA and 98.2% of Cr(VI) were removed within 60 min in presence of S-nZVI alone, whereas, correspondingly, 98.2% of BPA and 96.9% of Cr(VI) were eliminated in the S-nZVI/H₂O₂ system. Specifically, humic acid (HA) and H₂PO₄⁻ inhibited the deterioration of BPA but posed no significant effect on Cr(VI) removal. NO₃⁻ had a slight lifting effect on the removal of BPA and Cr(VI), while HCO₃⁻ showed a relatively weak prohibition. Experiments with EPR and radical probe tests also provide direct evidence that hydroxyl radicals was monitored in the S-nZVI/H₂O₂ system, which not only degraded BPA but also inhibited the reduction of Cr(VI). It could not be ignored that FeS accelerated Fe⁰ corrosion to release Fe²⁺. In, addition, Fe⁰, Fe²⁺ and S²⁺ could react with Cr(VI) while the most of produced Cr(III) was co-precipitated in the form of Cr_xFe_1−xOOH film. The study confirmed that it was feasible for S-nZVI/H₂O₂ system to remove synchronously organic pollutants and heavy metal.

1. Introduction

The co-contamination of heavy metals and refractory organic compounds in wastewater have become an urgent environmental problem nowadays. One typical example is that chromate (Cr(VI)) is commonly present together with organic pollutants in industrial waste of metal finishing, wood preserving, petroleum refining, and even in contaminated fields. It is well known that Cr(VI) is acutely toxic, carcinogenic and mutagenic to living organisms. In contrast, Cr(III) is less toxic and can be easily removed through precipitation [1]. The effective transformation of Cr(VI) into Cr(III) is a preferred method for reducing contamination. However, the reduction of Cr(VI) and the oxidation of organic pollutants seems theoretically contradictory, making this kind of wastewater difficult to treat [2,3]. Thus, efficient and simultaneous removal of Cr(VI) and organic pollutants is of great significance for water remediation.

Nanoscale zero-valent iron (nZVI) is a versatile material with high reductive capacity (E⁰ = −0.44 V). Due to this strong reductivity, nZVI has shown attractive potential in the ex/in situ transformation of Cr(VI) to the less toxic Cr(III) [4]. Ferrous and ferric ions, as corrosion products of nZVI, would be sacrificed to precipitate with Cr(III) to form insoluble minerals, making nZVI a promising material for Cr(VI)-bearing wastewater treatment. In addition, the Fenton or Fenton-like reactions are attractive for the removal of recalcitrant contaminants from aqueous solutions via the generation of powerful radicals (such as ·OH and SO₄·⁻) from the disintegration of oxidants (such as H₂O₂ and persulfate) with Fe²⁺ oxidized to Fe³⁺, as per Equations (1) and (2) [5]. Typically, iron acts as the catalyst to activate H₂O₂ or persulfate (PS, S₂O₈²⁻) to generate into radicals, which exhibit strong oxidizing capability and are capable of degrading most of the organic pollutants to CO₂, H₂O, or intermediate products such as organic acids with a series of rapid chain reactions [6]. In particular, the dissolved Fe²⁺ is essential for radical production, and nZVI is widely used as a source of in situ Fe²⁺ precursor in heterogeneous Fenton-like systems [7,8]. Compared to the homogeneous Fenton reaction, the nZVI-Fenton reaction can run in a wider pH range because of the fast circulation of Fe²⁺/Fe³⁺ on the Fe⁰ surface [9,10]. With these advantages, the nZVI-based Fenton-like reaction is expected to be an efficient means of wastewater purification.

Fe²⁺ + H₂O₂ → Fe³⁺ + OH⁻ + ·OH

(1)

Fe²⁺ + S₂O₈²⁻ → Fe³⁺ + SO₄²⁻ + SO₄·⁻

(2)

In a typical Fenton-like process with nZVI, the reducing capacity of Fe⁰ was used to activate H₂O₂ or persulfate with the formation of ·OH or SO₄·⁻ for the oxidation of organic pollutants. In this sense, most studies have only focused on the oxidation of organic pollutants, ignoring cases of co-contamination. In fact, nZVI-based Fenton-like systems have the potential to remove both heavy metals and organic pollutants in a single system [11,12]. Yin et al. studied the removal of coexisting Cr(VI) and 4-chlorophenol by reducing Cr(VI) in a first step with subsequent H₂O₂ oxidation. They found that 100% of Cr(VI) could be reduced by nZVI within 20 min, and the subsequent addition of H₂O₂ initiated a Fenton reaction for 4-CP degradation [13]. Yang et al. found that the total dissolved copper (TCu) and methylene blue (MB) could be removed together in the ZVI/H₂O₂ system [14]. In addition to the nZVI/H₂O₂ system, the simultaneous removal of phenol and Cr(VI) could be achieved using a system containing bentonite-supported nZVI (B-nZVI) and persulfate, as reported by Diao et al. [15]. A positive increase in dissolved Fe²⁺ was found throughout persulfate decomposition, and SO₄·⁻ was identified as acting toward phenol degradation, while B-nZVI was responsible for Cr(VI) [16]. However, the previous studies ignored the interaction effect between Cr(VI) and organic pollutants, and the efficiency and mechanism of their simultaneous removal in the nZVI/H₂O₂ Fenton-like system are still unclear.

In this study, the synchronous removal of Cr(VI) and bisphenol A (BPA) was evaluated in both H₂O₂ and persulfate oxidation systems. A sulfidated nanoscale zero-valent iron (S-nZVI) was chosen as the activator of the reaction system, because it has been reported to show significantly better efficiency with respect to Cr(VI) removal [17]. To elucidate the interaction effects between Cr(VI) and BPA, the effect of pH, H₂O₂ and catalyst dosage were investigated. The possible mechanism of pollutant elimination was further investigated by radical analysis and material characterization.

2. Materials and Methods

All chemicals, including ferrous sulfate (FeSO₄·7H₂O), sodium sulfide (Na₂S·9H₂O), sodium borohydride (NaBH₄), absolute ethyl alcohol, sodium sulfate hydrate (Na₂SO₄·10H₂O) potassium dichromate (K₂Cr₂O₇), bisphenol A (BPA), hydrogen peroxide (H₂O₂), and 1,10-phenanthroline, were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Deionized water (≥18.0 MΩ cm⁻¹) was deoxygenated by high-purity nitrogen purging for 1 h before use.

2.1. Preparation nZVI and S-nZVI

nZVI and S-nZVI nanoparticles were prepared according to our previous research [17]. Firstly, 50 mL of NaBH₄ (0.036 mol/L) and 100 mL of FeSO₄·7H₂O (0.018 mol/L) was added dropwise into a three-neck flask under N₂ protection. Then, the resulting nZVI suspensions in the three-neck flask were magnetically separated and rinsed with anaerobic water and ethanol three times.

To obtain S-nZVI, prepared FeSO₄·7H₂O solution was mixed with nZVI to obtain uniform dispersion, and then Na₂S solution was added dropwise into the flask and the mixture was stirred continuously for 20 min. Finally, the zero-valent iron was magnetically separated and rinsed as described above, and vacuum-dried at 50 °C overnight. The whole preparation process was performed under a N₂ atmosphere.

2.2. Experimental Procedures

The synchronous Cr(VI) sequestration and BPA degradation were conducted in a 250 mL beaker with 100 mL solution open to the atmosphere. The initial concentration of Cr(VI) and BPA was the same: 5 mg/L. Unless otherwise specified, an aliquot of stock solution of H₂O₂ (0.2 mmol/L) and S-nZVI (50 mg/L) was sequentially injected as a trigger for the reaction. At specific time intervals, 2 mL of samples were periodically collected and immediately mixed with 2 mL of ethanol to terminate the reaction, then filtered with 0.22 µm PTFE membrane filters for further analysis. The effects of initial pH, oxidant dosage, and catalyst dosage on synchronous removal of BPA and Cr(VI) by S-nZVI/H₂O₂ system were investigated. Specifically, latent co-existing water anions (HCO₃⁻, H₂PO₄⁻, NO₃⁻) and humic acid (HA) were also examined to obtain a better understanding of their adaptability in an aquatic environment. All experiments were conducted in duplicate.

To investigate the free radical species, ethanol (2 mmol/L) was added as a radical scavenger and 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) was used as a spin trap agent for electron paramagnetic resonance (EPR) analysis (EMX-8/2.7C, Bruker, Germany) [18]. Synchronous removal of BPA and Cr(VI) with or without o-phenanthroline (o-phenanthroline: S-nZVI = 10:1) was performed to investigate the effect of Fe(II).

2.3. Analysis

The surface morphologies and element mappings of the S-nZVI were observed by scanning electron microscopy (SEM) (ZEISS Ultra 55., Germany) combined with energy dispersive X-ray spectroscopy (EDX). The mineralogical identity and structures were determined by a D/Max-IIIA Powder X-ray Diffractometer (XRD) (Rigaku Corp., Tokyo, Japan) equipped with Cu–K_α radiation. The surface compositions of the S-nZVI were determined by X-ray photoelectron spectroscopy (XPS) (Thermo-VG Scientific ESCALB-250 with Al–Kα radiation).

The concentration of BPA was measured by high-performance liquid chromatography (HPLC, LBP, 3000, RIGOL, Beijing, China). Cr(VI) was measured using a UV-vis spectrometer for the 1,5-diphenylcarbazide colorimetric method at a wavelength of 544 nm, and the concentration of total chromium was determined by flame atomic adsorption spectrometry (Flame-AAS, 240, Agilent, Santa Clara, CA, USA). Cr(III) was identified by calculating the difference between the total chromium concentration and Cr(VI). Hydrogen peroxide was examined by a UV-vis spectrophotometer at 400 nm after coloration with Ti(SO₄)₂. The concentration of ferrous ions was measured by the o-phenanthroline method at wavelength of 510 nm. The concentration of total iron ions was measured by adding hydroxylamine hydrochloride to reduce ferric ions into ferrous ions. Total organic carbon (TOC) was determined by TOC analyzer (Liqui TOC II).

3. Results and Discussion

3.1. Characterization of S-nZVI

The surface morphology and element distribution of pristine S-nZVI was characterized by SEM-EDX as well as TEM. As shown in Figures S1 and S2, the spherical particles of Fe⁰ are covered with a thin layer of micrometer-sized material, which can be attributed to FeS. In addition, peaks of Fe and S elements were observed in the EDS mapping (Figure S1c). As shown in Figure S3, both S-nZVI and nZVI shows shrill 2θ peaks at 44.79°, which correspond to the [110] directions of α-Fe, indicating the high purity and crystallinity of Fe⁰. Moreover, the diffraction peak of S-nZVI was much lower than that of nZVI, because the Fe⁰ core was encapsulated by a coverage of amorphous FeS. The above characterization indicates that S-nZVI is a core–shell structure with a coverage of FeS.

3.2. Cr(VI) and BPA Removal in Different Systems

The different systems on synchronous removal of BPA and Cr(VI) were investigated. As shown in Figure 1a, only 8.6–11.3% of BPA was removed when H₂O₂ or nZVI or S-nZVI were used individually during the reaction. However, the removal rates of BPA by nZVI/H₂O₂ and S-nZVI/H₂O₂ systems were 97.7% and 98.5% after 20 min, respectively, indicating that the oxidative role of H₂O₂ should be activated to produce reactive species responsible for the degradation of BPA (Equations (3) and (4)) [19]. Noteworthily, the coupling of S-nZVI with H₂O₂ presented the highest efficiency for the degradation of BPA, and sulfur promotes the nZVI corrosion with generation of more hydroxyl radicals for BPA degradation [20]. Upon the detection of TOC variation (Figure S4), it was found that rapid TOC removal was achieved during the initial stage of the reaction. Then, after 10 min reaction, more TOC was gradually removed, and finally, 67.7% of TOC was eliminated in the S-nZVI/H₂O₂/Cr/BPA system. This result suggests that a considerable amount of BPA can be mineralized to carbon dioxide and water in the reaction. Interestingly, it was observed that 34.9% of Cr(VI) was removed by H₂O₂ alone within 60 min (Figure 1b). As a strong oxidant, K₂Cr₂O₇ (E⁰ = 1.33 V) can be reduced by H₂O₂ (E⁰ = 0.678 V) according to Equation (5) [21]. Furthermore, 22.8% and 28.8% of Cr(VI) were removed by nZVI/H₂O₂ and S-nZVI/H₂O₂ systems after 5 min, respectively, values which were lower than those obtained for nZVI (70.9%) and S-nZVI (76.3%) systems. The reaction rate of Fe²⁺/H₂O₂ (2.58~3.14 min⁻¹) is much higher than that of Cr(VI) with Fe²⁺ (0.0763~0.1078 min⁻¹), and the produced ·OH could oxidize Cr(III) back to Cr(VI) [22]. Noteworthily, the removal efficiency of Cr(VI) in H₂O₂/S-nZVI system (79.4%) was much higher than that of the H₂O₂/nZVI system (37.5%) after 10 min, indicating the better performance of S-nZVI compared to nZVI (Equations (6)–(8)). After 20–30 min, both H₂O₂/nZVI and H₂O₂/S-nZVI systems achieved over 95% of Cr(VI) removal, close to that obtained by nZVI (97.8%) and S-nZVI (98.2%). The above results clearly suggest that the synchronous removal of BPA and Cr(VI) can be successfully achieved in nZVI-based Fenton-like systems, and S-nZVI performed better than bare nZVI.

Fe²⁺ + H₂O₂ → Fe³⁺ + ·OH + OH⁻

(3)

Cr(n) +H₂O₂ → Cr(n + 1) + ·OH + OH⁻ (n = IV or V)

(4)

2HCrO₄⁻ + 8H⁺ + 3H₂O₂ ⇌ 2Cr₃⁺ + 3O₂ + 8H₂O

(5)

S²⁻ + Cr₂O₇²⁻ + 2H⁺ → SO₃²⁻ + Cr(OH)₃↓ +H₂O

(6)

Cr₂O₇²⁻ + 6Fe²⁺ + 14H⁺ → 6Fe³⁺ + 2Cr³⁺ + 7H₂O

(7)

Cr₂O₇²⁻ + 3Fe⁰ + 14H⁺ → 3Fe³⁺ + 2Cr³⁺ + 7H₂O

(8)

To further detect the interaction of BPA and Cr(VI) in a S-nZVI/H₂O₂ system, the effect of one contaminant on the other contaminant was investigated via batch experiments. For example, the effects of BPA on Cr(VI) removal were studied with different initial concentrations. Figure 1c shows that the degradation efficiency of BPA decreased when Cr(VI) was added into the S-nZVI/H₂O₂ system, because Cr(VI) consumed the reductive Fe species and also reacted with H₂O₂. Therefore, the degradation of BPA was eventually inhibited with increasing dosage of Cr(VI), inactivating more H₂O₂ and decreasing the yield of hydroxyl radicals (Figure 1e). Conversely, the presence of BPA showed no significant inhibitory effect, but rather a positive promotion of Cr(VI) removal (Figure 1d). However, it was noticed that this promotion was not further enhanced with increasing BPA concentration (Figure 1f). The competitive reaction between BPA and H₂O₂ seemed to protect Cr(VI) from oxidation and facilitate its reduction on the S-nZVI surface.

In the past decade, SO₄·⁻-based advanced oxidation processes using PS as the oxidant have been studied intensively for wastewater treatment. In comparison with the S-nZVI/PS Fenton-like system, as shown in Figure S5, S-nZVI/H₂O₂ exhibited a significantly higher efficiency in BPA oxidation, indicating a higher oxidation capacity of S-nZVI/H₂O₂ than S-nZVI/PS. Conversely, Cr(VI) seemed to be removed more quickly in the S-nZVI/PS system than in the S-nZVI/H₂O₂ system, implying a lower interference of PS activation on Cr(VI). This result is consistent with that reported by Diao et al., where Cr(VI) removal was negligibly influenced by PS in the nZVI/PS system. However, after 20 min reaction, the difference between the two reaction schemes was really minor, and was surprisingly uniform after 30 min. The reduction of Cr(VI) after 30 min was over 95%, and achieved equilibrium in the remaining reaction time. Considering the joint removal performance of BPA and Cr(VI), the S-nZVI/H₂O₂ system seemingly showed a more comprehensive potential than S-nZVI/PDS with respect to the treatment of co-contamination of BPA and Cr(VI).

3.3. Effect of pH, Oxidant Dosage, Catalyst Dosage

It was observed that Cr(VI) reduction and BPA oxidation were dependent on varying pH conditions (Figure S6a). Approximately 90% of BPA was removed after 20 min at pH = 3, whereas there was almost no obvious BPA degradation at pH 7.0 and 9.0, indicating that BPA removal increased with decreasing initial pH value, which might be attributed to the generation of more hydroxyl radicals by S-nZVI/H₂O₂ in acidic conditions [23]. It was observed that 97.8%, 54.6%, 57.6%, 53.1% and 45.5% of Cr(VI) removal were obtained at pH values of 3, 4, 5, 7 and 9, respectively (Figure S6b). This suggests that acidic conditions favor Cr(VI) removal. More indissoluble Fe³⁺ hydroxide was precipitated with Cr(III) to form a Fe-Cr hydroxide layer under neutral or alkaline conditions [17]. These precipitates on nZVI surface occupy active sites, limiting Fe²⁺ release, and retarding electron transfer from S-nZVI to Cr(VI) [24]. In addition, the oxidation potential of ·OH decreased along with the increase in solution pH, from 2.65–2.80 V at pH 3.0 to 1.90 V at pH 7.0, as previously reported [19]. Nevertheless, some researchers have reported that Cr(VI) sequestration by nano-FeS₂ increases notably with an increase in suspension pH from 5.0 to 9.0 [25]. This is because S²⁻ plays a primary role in the reactive surface of nano-FeS₂, having higher nucleophilicity in an alkaline solution, and may enhance the reduction of Cr(VI). In this work, Cr(VI) was completely removed at pH 3, while approximately 50% Cr(VI) removal was obtained under other pH conditions, indicating that acidic conditions (pH = 3) were conducive to Cr(VI) removal.

The influence of oxidant dosage on the synchronous removal of Cr(VI) and BPA in the S-nZVI/H₂O₂ system is depicted in Figure S6c,d. The BPA removal rate increased from 88.5% to 98.0% after 20 min when the amount of oxidant added to the system was increased from 0.2 mmol/L to 0.4 mmol/L (Figure S6c). These results conclusively demonstrate the key role played by H₂O₂ in the generation of ·OH for BPA degradation in this system. Interestingly, despite there being no obvious difference observed in BPA removal when the H₂O₂ dosages were in the range of 0.4 to 0.8mmol/L, the reaction constant decreased from 0.298 min⁻¹ to 0.209 min⁻¹. In fact, excessive H₂O₂ could cause the release of a mass of ·OH, resulting in the self-quenching of hydroxyl radicals, as shown in Equations (9) and (10) [25].

H₂O₂ + OH· → H₂O + ·O₂H

(9)

·O₂H + OH· → H₂O + O₂

(10)

As shown in Figure S6d, when the H₂O₂ dosages exceeded 0.4 mmol/L, Cr(VI) removal rapidly increased during the first 2 min, then the amount of Cr(VI) gradually decreased, but an increase in Cr(VI) removal was observed after 30 min, meaning that Cr(VI) removal decreased with increasing H₂O₂ dosage. This may be due to the generation of hydroxyl radicals with increasing H₂O₂ dosage, causing Cr(III) to be reconverted into Cr(VI) and the competition between excess H₂O₂ and Fe²⁺, which could quickly remove Cr(VI). Therefore, the larger the oxidant dosage, the slower the removal rate of Cr(VI) at H₂O₂ concentrations higher than 0.4 mmol/L. In addition, the dosage of S-nZVI is also a vital control variable for BPA and Cr(VI) removal. As shown in Figure S6e,f, the increment of S-nZVI significantly accelerated the synchronous removal of BPA and Cr(VI) when increasing the S-nZVI dosage from 0.02 to 0.15 g/L. Specifically, the reaction kinetics constant of BPA degradation increased from 0.088 min⁻¹ to 0.276 min⁻¹. Increasing the dosage of S-nZVI provided more active sites for not only Fenton-like catalysis, but also Cr(VI) sequestration. In addition, the ultimate removal rates of BPA and Cr(VI) were similar when the S-nZVI dosage was increased from 0.05 to 0.15 g/L. Hence, considering the cost and practical application of S-nZVI, 0.05 g/L of it was chosen for the subsequent experiments.

3.4. Effect of Water Matrices

Humic acid (HA) and coexisting water anions (including HCO₃⁻, NO₃⁻, H₂PO₄⁻) are widely present in the environment and interfere with the process of environmental restoration. As shown in Figure 2a, the removal of BPA decreased from 97.7% to 63.7% when HA concentration was increased from 0 to 50 mg/L. Since humic acid is a kind of organic substance, it competes with BPA to quench ·OH, resulting in a decrease in BPA degradation rate [26]. Additionally, humic acid might integrate with surface iron (such as Fe²⁺, Fe³⁺ and Fe(0)) to form dissolved Fe–humate complexes, causing a loss of reducing substances as well as decreasing the removal efficiency of BPA [27]. However, when a small amount of HA (5 mg/L) was spiked into the reaction solution, almost no influence on Cr(VI) removal was observed (Figure 2b). In particular, humic acid also acts as an additional adsorbent and electron transfer mediator at tiny concentrations due to its multiple functional groups [28], which ultimately offset the negative impact of highly concentrated HA for Cr(VI) removal.

As presented in Figure 2c,d, the removal rate of both BPA and Cr(VI) by S-nZVI/H₂O₂ was significantly influenced in the presence of NO₃⁻ and H₂PO₄⁻, but slightly suppressed by HCO₃⁻. The slight inhibition of HCO₃⁻ here could be because of its reaction with Fe²⁺ to form FeCO₃ or hydroxyl carbonate, which attached to the surface of S-nZVI, resulting in a decrease in iron ions and reactive sites [28]. However, in the presences of NO₃⁻, the removal efficiency of BPA increased from 98.9% to 99.7%, while the initial removal efficiency of Cr(VI) increased from 95.7% to 98.7% in 30 min. This is probably because NO₃⁻ competes with Cr(VI) for reactive sites in S-nZVI and produces more Fe(II) at the solid–liquid interface (as per Equations (11) and (12)) [20]. Moreover, H₂PO₄⁻ showed a different effect, wherein the rate of BPA removal decreased from 98.2% to 75.9% while the Cr(VI) removal increased from 96.9% to 99.1%. Firstly, this could be because H₂PO₄⁻ can complex with Cr(VI) to form a water-soluble complex [29], preventing the formation of a zero-valent iron surface passivation layer and improving the removal of Cr(VI). Secondly, H₂PO₄⁻ is a typical radical scavenger, thus inhibiting the degradation of BPA [30].

Fe⁰ + NO₃⁻ + 2H⁺ → Fe²⁺ + H₂O + NO₂⁻

(11)

4Fe⁰ + NO₃⁻ + 10H⁺ → 4Fe²⁺ + 3H₂O + NH₄⁺

(12)

3.5. Proposed Reaction Mechanism

Accordingly, it can be deduced that BPA was degraded by activating H₂O₂ with S-nZVI, but the removal of Cr(VI) was achieved by ferrous on S-nZVI surface. To clarify the role of Fe(II) on BPA and Cr(VI) elimination, o-phenanthroline was added into the reaction system as a Fe(II) shielding agent. As shown in Figure 3a,b, the removal of Cr(VI) and BPA was significantly inhibited after adding o-phenanthroline. The BPA removal rate decreased from 97.7% to 29.1%, and the Cr(VI) removal rate decreased from 97.8% to 17.6%, implying that Fe(II) not only participated in the Fenton reaction as an activator of H₂O₂, but also reduced Cr(VI) into Cr(III) as an electron donor [31]. In addition, the H₂O₂ use rates were determined for different systems as shown in Figure 3c. It can be observed that H₂O₂ was completely used in 60 min for the H₂O₂/S-nZVI/BPA/Cr system, while 82.4% and 68.8% were achieved for the H₂O₂/S-nZVI/BPA and H₂O₂/S-nZVI/Cr(VI) systems, respectively. Moreover, the use of H₂O₂ also reached 64.7% when S-nZVI was not added, due to the catalysis of chromium, which is consistent with the results in Figure 1, suggesting that H₂O₂ could be decomposed in presence of Cr(VI) via a redox process (Equations (14) and (15)). Both Fe and Cr species are able to react with H₂O₂, thus consuming more H₂O₂ than systems with a single contaminant [32].

HCrO₄⁻ + 2H₂O₂ + H⁺ ⇌ CrO(O₂)₂ + 3H₂O

(13)

2CrO(O₂)₂ + 6H⁺ ⇌ 2Cr³⁺ + 3O₂ + H₂O₂ + 2H₂O

(14)

The catalytic activation of H₂O₂ is most likely to produce ·OH via a Haber-Weiss-type reaction [33]. As a typical ·OH quencher, ethanol was used to identify the presence of ·OH in the present reaction system, and the removal efficiencies of BPA with or without ethanol were evaluated. As observed in Figure 3d, 14.22% of BPA was removed after 20 min in the presence of ethanol, whereas the removal rate of BPA was 89.92% without ethanol, suggesting an obvious inhibitory effect of ethanol on BPA degradation. In contrast, no significant difference was observed in Cr(VI) removal with scavenger (Figure 3e). The EPR technique with 5,5-dimethyl-1-pyrrolidine-N-oxide (DMPO) as a spin-trapping agent was used to further identify the production of ·OH in the S-nZVI/H₂O₂ system. As shown in Figure S7, there exists a characteristic peak with a peak height ratio of 1:2:2:1, and the signal intensity of DMPO-OH increased in first 2 min but then gradually decreased due to the consumption of H₂O₂. These results suggest that ·OH was the principal species responsible for BPA degradation in the S-nZVI/H₂O₂/Cr(VI)/BPA reaction system.

The transformation of Cr(VI) is important for detoxification, and the species of aqueous Cr varies with S-nZVI reduction and ·OH oxidation. As presented in Figure 3f, the concentration of total Cr in solution decreased along with the elimination of Cr(VI). Aqueous chromium was continuously reduced and immobilized by S-nZVI, which is consistent with previous studies reporting that Fe⁰ corrodes and yields the iron (hydro)oxides to remove heavy metals throguh adsorption and co-precipitation [2]. However, it is worth noting that the concentration of total Cr at every time point is higher than that of Cr(VI). This result is different from that reported in our previous study, which indicated that the total Cr and Cr(VI) displayed almost same removal tendency [17]. This is because a significant portion of Cr(VI) was transformed to Cr(III) through the reaction with H₂O₂, and therefore was not captured by S-nZVI [34]. This phenomenon is evidenced by the variation of Cr species. For example, the concentration of Cr(III) decreased from 0.84 mg/L to 0.66 mg/L in 5 min, which could be due to the reversible conversion of Cr species by ·OH. In addition, Cr(VI) can be reduced by surface Fe²⁺ and S²⁻ from the corrosion of S-nZVI during the reaction process [4]. The newly formed Cr(III) and Fe(III) co-precipitate simultaneously on the S-nZVI to form Cr/Fe (oxy)-hydroxide, as per Equations (15) and (16) [35]. The reusability of S-nZVI was further investigated by separating S-nZVI with a magnet after the reaction, followed by washing in deionized water and subsequent reuse for synchronous removal of Cr(VI) and BPA. As presented in Figure S8, BPA can be fully degraded even when S-nZVI was reused three times, but the removal efficiency of Cr(VI) decreased along the number of times that S-nZVI was reused. In the third run, 69.2% of Cr(VI) was removed in the S-nZVI/H₂O₂/Cr/BPA reaction system. The deposition of Fe-Cr oxides impeded the smooth generation of ferrous species from S-nZVI corrosion, leading to a decline in performance for Cr(VI) sequestration.

xCr³⁺ + (1 − x)Fe³⁺ + 2H₂O → Cr_xFe_1−xOOH↓ + 3H⁺

(15)

xCr³⁺ + (1 − x)Fe³⁺ + 3H₂O → (Cr_xFe_1−x)(OH)₃↓ + 3H⁺

(16)

After the reaction, a thick iron oxide was formed on the S-nZVI surface, as indicated by the XRD spectrum (Figure S9). Compared to fresh S-nZVI, a new peak appeared at a 2θ angle of 35.6°, which attributed to the facet of hematite (110), implying the formation of iron oxides during the reaction. To better understand the mechanism of BPA elimination and Cr(VI) sequestration, the XPS spectra of S-nZVI were detected before and after the reaction. As shown in Figure 4a, the XPS spectra of Fe 2p on the surface of the pristine S-nZVI at 706.96 eV corresponded to Fe 2p_3/2 of Fe⁰ with low content of 2.55% [36]. The binding energies at 712.05 eV and 726.11 eV, coupled with a satellite peak obtained at 719.07 eV and 732.35 eV, respectively, were assigned to Fe 2p_3/2 and Fe 2p_1/2 of Fe (III) in terms of iron (oxy) hydroxide [37]. In addition, the peaks at 710.40 eV and 724.01 eV were attributed to Fe 2p_3/2 and Fe 2p_1/2 of Fe(II), respectively. These results demonstrate that the surface of S-nZVI was slightly oxidized and wrapped by a mixture of Fe(II) and Fe(III), which is consistent with the results of EDX analysis. After the reaction, it can be observed that the peak of Fe⁰ disappeared, and the molar ratio of Fe(II) remained nearly unchanged, while the Fe(III) fraction increased from 23.74% to 27.32%, as depicted in Figure 4a and Table S1. Thus, it can be deduced that Fe⁰ was sacrificed as the electron donor and converted to Fe(II), which then reacted with Cr(VI) and H₂O₂, resulting in a redox transformation from Fe(II) to Fe(III) [35]. Moreover, four distinguishing peaks can be found in the S2p spectrum (Figure 4b). The binding energies at 161.37 eV and 163.07 eV were attributed to iron monosulfide (FeS) and iron disulfide (FeS₂) [38], and the peaks with binding energy of 166.49 eV and 168.42 eV were attributed to SO₃²⁻ and SO₄²⁻, respectively [39]. After the reaction, the peak intensity of S²⁻ decreased from 56.45% to 14.65% (Table S1), but the peak intensities for SO₃²⁻ and SO₄²⁻ increased remarkably to 6.38% and 78.97%, respectively, demonstrating the involvement of S²⁻ during the reductive adsorption of Cr(VI) and the catalytic activation of H₂O₂ for BPA degradation (Figure 4b). It has been verified that iron sulfides are able to act as good semiconductors or metallic conductors for S-nZVI, thus facilitating the electron transfer from the iron core to the adsorbed contaminants and facilitating the cycle of Fe(II)/Fe(III) [40]. The XPS spectrum of Cr was further tested to confirm the conversion of Cr(VI) to low-toxic Cr(III). After the reaction, as shown in Figure 4c, the deconvolution peaks of Cr 2p at 577.16 eV and 586.76 eV corresponded to Cr₂O₃ [25], and the peaks at 577.37 eV and 587.46 eV were assigned to Cr(OH)₃ [41], indicating that Cr(VI) sequestration by S-nZVI was primarily a process of chemisorption, with reduction of Cr(VI) to Cr(III). Additionally, the oxygen peaks with binding energies of 529.72 eV and 531.11eV were assigned to O²⁻ and OH⁻ (Figure 4d), suggesting the presence of iron oxide and hydroxyl species of Fe before use, respectively. After the reaction, the content of O²⁻ decreased from 31.98% to 15.65% while the content of OH⁻ increased from 68.02% to 84.35% (Table S1), indicating the formation of hydroxides on the surface of S-nZVI. These results imply that both Cr(OH)₃ and Cr(III)/Fe(III) (oxy)hydroxide were on the surface of S-nZVI after Cr(VI) sequestration. In addition, the atomic fraction of C-C component at a binding energy of 284.8 increased remarkably from 41.83 to 68.19 after the reaction (Figure S10), which may indicate an adventitious organic carbon contamination on S-nZVI surface.

In summary, a probable mechanism of BPA degradation and Cr(VI) removal by S-nZVI/H₂O₂ system was schematically elucidated, as shown in Figure 5. Under acidic conditions, H₂O₂ was activated by Fe²⁺, Fe³⁺ and Fe⁰ on the S-nZVI surface, producing ·OH for BPA degradation. Interestingly, the co-existence of Cr(VI) negatively slowed the oxidation rate of BPA because of the competitive consumption of H₂O₂ and ·OH. Conversely, the removal of Cr(VI) was positively promoted in the presence of BPA due to its oxidative reaction with ·OH, which inhibited the reoxidation of Cr(III) by ·OH. The stepwise removal of Cr(VI) involved reduction, re-oxidation, adsorption and immobilization. Initially, Cr(VI) was absorbed on the S-nZVI surface and occupied the reactive sites of S-nZVI, and then Cr(VI) was directly reduced to Cr(III) on the S-nZVI surface by Fe⁰, Fe(II) and S²⁻. The dissolved Fe(III) on the surface of S-nZVI could be reduced back to Fe(II) by Fe⁰ and S²⁻, giving rise to more ferrous sites for further Cr(VI) reduction. The produced Cr(III) and Fe(III) were present in the forms of mixed Cr-Fe (oxy) hydroxides, and formed a passivation layer that impeded further reaction. However, it should be highlighted that not all Cr species could be captured by S-nZVI within the reaction time, and a considerable fraction of Cr(VI) was converted directly to Cr(III) through reaction with H₂O₂ under acidic conditions. The formed Cr(III) presented as dissolved ions at acidic pH, but might sediment further in the form of chromium hydroxides.

4. Conclusions

The present study used S-nZVI as a reductant and catalyst for heterogeneously activated H₂O₂ for the synchronous removal of BPA and Cr(VI) via processes including both degradation and reductive adsorption. The removal efficiency reached up to 98.2% for BPA and 96.9% for Cr(VI) within 60 min under optimum conditions, which were pH 3, and concentrations of 0.05 g/L of S-nZVI, and 0.2 mmol/L of H₂O₂, respectively. Compared to nZVI/H₂O₂, the S-nZVI/H₂O₂ Fenton-like system showed a very similar efficiency toward BPA degradation, but a faster rate for Cr(VI) removal. HA and H₂PO₄⁻ inhibited, but NO₃⁻ promoted the degradation of BPA. This was different from BPA degradation in that the effects of water matrix on Cr(VI) removal were inconspicuous and Cr(VI) could be completely converted to Cr(III) after 30 min reaction time. In addition, ·OH was identified as the principal radical species in the S-nZVI/H₂O₂ Fenton-like system for BPA degradation. However, the reaction with Cr(VI) was a complex process, involving not only reduction by both S-nZVI and H₂O₂, but also ·OH-mediated oxidation of Cr(III). Additionally, the removal of Cr(VI) involved adsorption, reduction and immobilization on the S-nZVI surface, as further evidenced by XPS characterization. This study provides a new feasible solution for the treatment of complex polluted water bodies containing refractory organic pollutants and heavy metal pollutants.