Demonstration of a feasible energy-water-environment nexus: waste sulfur dioxide for water treatment

Abstract

Sulfur dioxide (SO₂) mitigation and water treatment are two key aspects towards a sustainable environment, and simultaneous achievement of these two goals is extremely attractive. Inspired by the iron ion catalyzed auto-oxidation of aqueous SO₂ (i.e., Fe(II)/sulfite process), that generates intermediary sulfate radical able to oxidize organic compounds, we propose a feasible energy-water-environment nexus by using waste SO₂ to alleviate water contamination. As a demonstration, electrolysis is used to assist Fe(II)/sulfite (i.e., electro/Fe(II)/sulfite) process to enhance contaminant removals. Results showed 91% of 10 μM ibuprofen (a nonsteroidal anti-inflammatory drug) contaminant at neutral pH was removed, due to sulfate radical oxidation. Synergy mechanisms of electro/Fe(II)/sulfite process were revealed. Moreover, the electro/Fe(II)/sulfite process could effectively degrade contaminants in water bodies from fields, indicating its promising practical application. Energetic analysis indicated that electric treatment cost is 8.65 cent/m³, and is affordable by most water treatment plants. Possible procedures to realize the proposed energy-water-environment nexus were also suggested. The strategy proposed in this study adds new value to the waste produced in energy production and will create renewed interest considering its potential use towards environmental treatment.

Graphical Abstract

1. Introduction

Sulfur dioxide (SO₂) is regarded as an atmospheric pollutant, which causes acid rain and human respiratory system diseases after short-term exposure [1,2]. Around 100 million tons of SO₂ are emitted into the atmosphere annually [3], a large fraction of which is produced by the fossil fuel power plants although clean oil and gas production techniques are being developed [4,5]. It is hence imperative that power plants flue gas must be desulfurized before emission into the atmosphere. The most widely adopted technique to remove SO₂ from the emitted flue gas of power plants is called flue gas desulfurization (FGD) process. Conventional FGD process relies on a sprayed limestone slurry to scrub dissolved SO₂ in water droplets (Fig. 1a) [6,7]. During the FGD process, trace metals present in solution such as iron ions are catalytically active to convert SO₂ into sulfuric acid through “bi-cycle” radical chain reactions (Fig. 1b) [8,9]. Briefly, the first cycle of iron species turnover generates SO₃^•− (sulfite radical) (Equations 1-4), which then initiates the second cycle of oxysulfur radical evolution to sequentially derive SO₅^•− (peroxymonosulfate radical) and SO₄^•− (sulfate radical) in the presence of O₂ (Equations 5-7). These oxysulfur radicals, particularly SO₄^•−, are as highly reactive as hydroxyl radical (HO^•), and they can potently degrade the co-existed organic molecules [9]. Inspired by this, we coupled ferrous iron catalyst and aqueous SO₂ (i.e., Fe(II)/sulfite process) to treat various organic contaminants, such as azo dyes [10], endocrine disruptor [11], herbicide [12], and pharmaceutical [13]. Owing to the strong bactericidal potential, SO₄^•− was also used to inactivate water pathogens [14].

Fig. 1.

(a) Conventional FGD process to desulfurize the flue gas of fossil fuel power plants, and (b) “bi-cycle” radical chain reactions of iron catalyzed sulfite auto-oxidation that could degrade organic molecules during FGD process. S(IV) denotes sulfite, including HSO₃⁻/SO₃²⁻.

Cycle of Fe(II/III) turnover:

$${\mathrm{Fe}}^{2+}+{{\mathrm{HSO}}_3}^{-}\to {{\mathrm{FeHSO}}_3}^{+}$$ (1)

$${{\mathrm{FeHSO}}_3}^{+}+¼{\mathrm{O}}_2\to {{\mathrm{FeSO}}_3}^{+}+½{\mathrm{H}}_2\mathrm{O}$$ (2)

$${\mathrm{Fe}}^{3+}+{{\mathrm{HSO}}_3}^{-}\to {{\mathrm{FeSO}}_3}^{+}+{\mathrm{H}}^{+}$$ (3)

$${{\mathrm{FeSO}}_3}^{+}\to {\mathrm{Fe}}^{2+}+{{\mathrm{SO}}_3}^{•-}$$ (4)

Cycle of oxysulfur radicals evolution:

$${{\mathrm{SO}}_3}^{•-}+{\mathrm{O}}_2\to {{\mathrm{SO}}_5}^{•-}$$ (5)

$${{\mathrm{SO}}_5}^{•-}+{{\mathrm{HSO}}_3}^{-}\to {{\mathrm{SO}}_4}^{2-}+{{\mathrm{SO}}_4}^{•-}+{\mathrm{H}}^{+}$$ (6)

$${{\mathrm{HSO}}_3}^{-}+{{\mathrm{SO}}_4}^{•-}\to {{\mathrm{SO}}_4}^{2-}+{\mathrm{H}}^{+}+{{\mathrm{SO}}_3}^{•-}$$ (7)

Two major challenges impede the application of Fe(II)/sulfite process into practical water treatment. First, the Fe(II)/sulfite process generates SO₄^•− effectively at pH below 5, but the capability rapidly diminishes at near-neutral pH due to iron precipitation [10]. Since most water bodies in the fields are neutral [15,16], the effectiveness of the Fe(II)/sulfite process is compromised. Recently, researchers have used chromium [17], copper [12], manganese [14], and metallic composites [13] as catalysts to replace iron ions for sulfite activation. However, these new types of catalysts suffer from significant toxicity to human beings, low catalytic activity, and high preparation cost, respectively [9]. Therefore, Fe(II)/sulfite process is still a dominant sulfite activation platform. Secondly, it can be seen from Equations 1-7 that the SO₄^•− generation by the Fe(II)/sulfite process is strongly dependent on solution O₂ content. However, the evolution of oxylsulfur radicals in the second cycle of Fe(II)/sulfite process causes rapid depletion of dissolved oxygen within tens of seconds, far exceeding the dissolution rate of atmospheric oxygen [9,10]. The rapid depletion of solution O₂ was widely observed in batch reaction of various volumes [14]. The low oxygen content retards the turnover of Fe(II/III) in the first cycle and oxysulfur radical evolution in the second cycle, leading to relatively weak SO₄^•− generation.

To address the above challenges, we here apply electricity to Fe(II)/sulfite process (i.e., electro/Fe(II)/sulfite process). This is because 1) electrolysis is able to create local acid zone around the anode [18], 2) electrolytically generated nascent oxygen [19] can support the “bi-cycle” reactions, and 3) electrolysis of sulfite solution directly produces the oxysulfur radicals (i.e., SO₃^•−, SO₅^•− and SO₄^•−) [20] for possible organic contaminants degradation. To date, this novel energy-water-environment nexus utilizing fossil fuel combustion waste SO₂ for enhanced water treatment has never been reported previously. As a bench-scale demonstration, sulfite represents the captured SO₂ from flue gas and ibuprofen (a nonsteroidal anti-inflammatory drug) is selected as a typical water contaminant [21]. The electro/Fe(II)/sulfite process was used to degrade ibuprofen at neutral pH that is close to natural conditions.

In this strategy, the waste SO₂ from electrical energy generation is instead a useful resource holding unexplored chemical energy that can be used to treat contaminated water. Therefore, the aim of this work is to thoroughly investigate the feasibility of applying chemical energy of SO₂ into water treatment. Techno-economic analysis indicates that the electro/Fe(II)/sulfite process is energetically viable and relatively low-cost. The possible procedures to realize this process are also proposed to facilitate future research. It is anticipated that, this feasible energy-water-environment nexus could simultaneously achieve SO₂ mitigation and sustainable water treatment.

2. Materials and Methods

2.1. Materials

Sodium sulfite (Na₂SO₃), ferrous sulfate (FeSO₄), and sodium sulfate (Na₂SO₄) were purchased from Acros Organics (Thermo Fisher Scientific, Inc.). Ibuprofen (2-(4-isobutylphenyl)propanoic acid, C₁₃H₁₈O₂) used as the target compound was purchased from Alfa Aesar (Haverhill, MA). Ethanol (reactive toward SO₄^•−) and tert-butanol (inert toward SO₄^•−) from Thermo Fisher Scientific, Inc. were used to identify SO₄^•−, because of their selective oxidation property. Details were also included in Section 3.4. Milli-Q water was used throughout this study unless indicated. All other chemicals were from Thermo Fisher Scientific, Inc.

2.2. Ibuprofen degradation experiments

The reactor to carry out experiments is shown in Fig. S1 of supplementary material. Ibuprofen degradation assays were performed under different conditions. Typically, 400 mL solution of 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, and 10 mM Na₂SO₄ as supporting electrolyte was prepared in a beaker, and immediately adjusted to pH 7 with 0.1 M NaOH or H₂SO₄. Solution of pH was recorded with a pH meter (Mettler Torledo). During the reaction, the uncontrolled change of solution pH was constantly monitored. 0.1 mM Fe(II) was diluted from a 100 mM FeSO₄ stock solution that was prepared in advance and stored in 4 °C refrigerator. Two identical Ti-based mixed metal oxide (Ti/MMO) mesh electrodes (5 cm²) with a fixed distance of 5 cm were placed in parallel in the reactor as anode and cathode, and a 200-mA constant current was applied by an Agilent E3612A DC power supply. The solution was constantly stirred by a PTFE-coated magnetic bar. Parameters including sulfite concentration, current intensity, and solution pH are varied if stated to study their effects on ibuprofen removal. Specifically, the effect of sulfite concentration (0-2.5 mM), current intensity (0-200 mA), and solution pH (4-10) were investigated toward ibuprofen degradation efficiency. For carbonate buffering effect, solutions containing 0.5, 1, or 2 mM Na₂CO₃ and indicated amounts of Fe(II) and/or sulfite were used. Solutions were rapidly adjusted to pH 7 prior to initiation of assays.

2.3. Analysis methods

At specific times, 1 mL samples were collected from the solution and the reactions were stopped by adding 1 mL of 100 mM ethanol. Before quantification, all samples were filtered through 0.45 μm nitrocellulose membranes. Concentrations of ibuprofen were then analyzed by a high-performance liquid chromatography (HPLC, Agilent 1200 Infinity Series) equipped with an Agilent Eclipse AAA C18 column (4.6 × 150 mm). 0.5 mL/min methanol/1% phosphoric acid (68/32) was used as the mobile phase, and ibuprofen was detected at 228 nm wavelength using Agilent 1260 diode array detector. Sulfite concentrations were quantified using a modified high throughput method [22], with an injection volume of 100 μL. The detection limit of this modified method is 0.05 mM. Solution pHs were recorded during a brief switch off of current supply to exclude the disturbance of electrodes electrolysis. All assays were conducted for triplicates. The ibuprofen pseudo first-order degradation rate constant (k) was obtained by calculating the slope of ln(C_t/C₀) versus time curve.

2.4. Quantifications of dissolved Fe(III) via colorimetric assays

Concentration of dissolved Fe(III) (i.e., [Fe(III)]_dissolved) was calculated from total dissolved iron concentration (i.e., [total iron]_dissolved) minus the dissolved Fe(II) concentration (i.e., [Fe(II)]_dissolved) (Equation 8). Total dissolved iron concentration was measured via a colorimetric assay. In details, 1 mL sample after filtration with 0.45 μm nitrocellulose membranes was added into solutions containing 4 mL of 1,10-o-phenanthroline (17 mM) and 2 mL of potassium hydrogen phthalate (20 mM). 0.5 mL of 10 mM hydroxylamine was used to reduce Fe(III) into Fe(II). After reaction for 15 min, all Fe(II) complexed with 1,10-o-phenanthroline into a pink product with peak wavelength of 510 nm, and was then quantified with a UV-Vis spectrometer. Dissolved Fe(II) concentration was quantified through the same manner, without addition of hydroxylamine.

$${\left[\mathrm{Fe}(\mathrm{III})\right]}_{\text{dissolved}}={\left[\text{total iron}\right]}_{\text{dissolved}}-{\left[\mathrm{Fe}(\mathrm{II})\right]}_{\text{dissolved}}$$ (8)

2.5. Cyclic voltammetry

The cyclic voltammetry of sulfite solution has been performed with many electrodes [20,23,24], and we attempt to further investigate the sulfur species evolution on Ti/MMO electrode in this study. Both 0-10 mM sulfite and 100 mM sodium sulfate electrolyte were prepared in 200 mL solution. To run cyclic voltammetry for the sulfite solution, a Ti/MMO electrode (2 cm²) was used as the working electrode, a Pt electrode (4 cm²) as the counter electrode, and a saturated Ag/AgCl electrode as the reference electrode. Scans were performed from −0.6 V to 1.7 V (vs SHE, standard hydrogen electrode) at a rate of 50 mV/s with SP-300 electrochemical workstation (BioLogic, France). Stable and closed cyclic voltammetry curves, which represent the steady-state electrocatalytic reactions, were obtained typically after two runs. 0.1 mM EDTA solution was used when stated, to exclude the effect of metal catalysis for sulfite oxidation on Ti/MMO anode. All cyclic voltammetry assays were performed at pH 7.

2.6. Electron spin resonance assay

To further probe the produced intermediate of the electro/sulfite reaction, 50 mM 5,5-dimethyl-1-pyrrolidine-N-oxide (DMPO) solution was used to trap the generated radical. Both the electro/sulfite reaction (1 mM sulfite solution and 200 mA current at pH 7) and the control reactions (1 mM sulfite solution or 200 mA current at pH 7) were carried out at room temperature for 1 min. Samples withdrawn by a capillary tube were inserted into the cavity of the electron spin resonance spectrometer (JES-FA200, Japan Electron Optics Laboratory Co. Ltd., Japan) for analysis under the following parameters: center field, 326 mT; sweep width, 30 mT; microwave frequency, 9.147 GHz; microwave power, 3 mW.

2.7. Radical-scavenging assay

Tert-butanol (inert toward SO₄^•−) and ethanol (reactive toward SO₄^•−) were used to scavenge relevant radicals in the Fe(II)/sulfite (0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, initial pH 7), electro/sulfite (1 mM sulfite, 10 mM Na₂SO₄, 200 mA current, initial pH 7), and electro/Fe(II)/sulfite (0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA current, initial pH 7) processes. 5 mM or 10 mM of tert-butanol and ethanol solutions were individually added into each of the three solutions containing 10 μM ibuprofen, and ibuprofen concentrations were immediately analyzed by HPLC after specific sampling times.

2.8. Removal of ibuprofen from field water matrices

Ibuprofen degradations by Fe(II)/sulfite, electro/sulfite, and electro/Fe(II)/sulfite processes were tested in four water bodies from fields, i.e., two groundwaters extracted from our Tallonal (TAL) and Pozo Mita (MIT) Superfund sites in Puerto Rico, and two surface waters collected from local Jamaica pond (JAM) and Wards pond (WAR) in Boston, Massachusetts. Extracted water matrices were filtered through 0.45 μm nitrocellulose membranes for three times to remove suspended solid particles, and subsequently stored in a 4 °C cold room. During ibuprofen degradation assays, 10 μM ibuprofen were prepared in 400 mL field water matrices, and then treated by Fe(II)/sulfite (0.2 mM Fe(II), 2 mM sulfite, initial pH 7), electro/sulfite (2 mM sulfite, 200 mA, initial pH 7), and electro/Fe(II)/sulfite (0.2 mM Fe(II), 2 mM sulfite, 200 mA, initial pH 7) processes. The intrinsic conductive ions together with added iron and sulfite supported the electrolysis reactions, without additional use of sulfate electrolyte.

3. Results and discussion

3.1. Electro/Fe(II)/sulfite process for water treatment

It is reported that during FGD process, trace iron ions catalyze the auto-oxidation of aqueous SO₂ (sulfite), and three intermediary oxysulfur radicals (i.e., SO₃^•−, SO₅^•−, and SO₄^•−) are generated [8,9]. Among them, SO₄^•− is the most powerful radical and could oxidatively degrade a large variety of organic compounds. Fe(II)/sulfite process was then developed for organic contaminants degradation inspired by this [10]. The most intriguing feature of the Fe(II)/sulfite process for water treatment is that both Fe(II) and Fe(III) are catalytically active toward sulfite oxidation (supplementary material Fig. S2) [9]. Therefore, the persistent turnovers of minimum iron dose can continuously steer sulfite activation to produce immense SO₄^•− radicals for organic contaminant degradation.

As shown in Fig. 2a, Fe(II)/sulfite process produced moderate oxidation capability at neutral pH. A solution of 0.1 mM Fe(II) and 1 mM sulfite mediated 28% removal of 10 μM ibuprofen after 30 min at pH 7. By comparison, solutions with only Fe(II) or only sulfite did not induce any noticeable removal of ibuprofen. It is expected that, the SO₄^•− generated from the Fe(II)/sulfite reaction was primarily responsible for ibuprofen removal [10]. But the Fe(II)/sulfite process was majorly hindered by significant precipitative deactivation of Fe(II) at near-neutral pH, which explained the incomplete ibuprofen removal observed in Fig. 2a.

Fig. 2.

(a) Electrolysis enhanced Fe(II)/sulfite process toward ibuprofen removal at neutral pH, and (b) ibuprofen pseudo first-order degradation rate constant (k) linearly increased versus current intensity. Conditions: 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, initial pH 7.

Application of 200 mA electric current to the Fe(II)/sulfite (i.e., electro/Fe(II)/sulfite) process greatly enhanced ibuprofen removal (Fig. 2a). Up to 91% of 10 μM ibuprofen was removed by the electro/Fe(II)/sulfite process after 30 min, reflecting a 3.6-fold enhancement of the ibuprofen pseudo first-order degradation rate constant (k) compared with Fe(II)/sulfite process. Moreover, the enhancement of ibuprofen degradation kinetics appeared to be linearly correlated with current intensity in the range of 0-200 mA (Fig. 2b). The mechanism of enhanced ibuprofen degradation is investigated and discussed below.

3.2. Synergy mechanism of electro/Fe(II)/sulfite process

3.2.1. Mechanism 1: acid zone around anode supported Fe(II)/sulfite process

During Fe(II)/sulfite process treatment, iron ion catalyst shows higher catalytic activity at acidic pH. Electrolysis of water solution is able to create an acid zone around the anode, due to the intense generation of protons [18]. As a result, the solution pH around the anode was 3.3 while the bulk solution pH was neutral in our experiment. The local acidification due to anode electrolysis significantly improved iron catalytic activity, and consequently enhanced water treatment efficiency of Fe(II)/sulfite process. To further validate the role of acid zone around the anode, ibuprofen degradation was performed by electro/Fe(II)/sulfite process with 50 mM Tris buffer. Significantly less ibuprofen removal was observed (Fig. 3a), because of weakened acidification effect and minimum dissolved iron catalyst (Fig. 3b).

Fig. 3.

(a) Degradation of ibuprofen by electro/Fe(II)/sulfite process in unbuffered or buffered solution, and (b) total dissolved iron concentration and pH at the end of reaction. Conditions: 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, with or without 50 mM Tris buffer, pH 7. Reaction for 30 min.

3.2.2. Mechanism 2: electrolytic oxygenation supported Fe(II)/sulfite process

As shown in Equations 1-7, Fe(II)/sulfite process efficiency is highly dependent on oxygen content, which supports both the turnovers of iron species in the first cycle and evolution of oxysulfur radicals in the second cycle. With sufficient oxygen, repeated turnovers of a minimum dose of iron, regardless of Fe(II) or Fe(III), will continuously drive sulfite auto-oxidation and produce SO₄^•−. However, the low oxygen content (Fig. 4a), mainly because SO₃^•− scavenged O₂ at an approaching-diffusion rate (Equation 5) [25], could adversely reduce SO₄^•− radical yield by inhibiting the conversion of Fe(II) into Fe(III) (first cycle) and also retarding SO₃^•− evolution to SO₄^•− (second cycle).

Fig. 4.

Changes of (a,b) dissolved oxygen, (c,d) dissolved Fe(III), (e,f) sulfite concentration, and (g,h) ibuprofen concentration in Fe(II)/sulfite, oxygenated Fe(II)/sulfite, and electro/Fe(II)/sulfite processes at pH 4 or pH 7. Conditions: Fe(II)/sulfite – 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄; oxygenated Fe(II)/sulfite – 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 1 L/min oxygenation; electro/Fe(II)/sulfite – 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA.

We first showed that purging gaseous O₂ to enhance the oxygen content (Fig. 4a,b) could remarkably accelerate the turnovers of dissolved Fe(II/III) catalyst (represented by conversion of Fe(II) into Fe(III)) (Fig. 4c,d) and oxysulfur radicals evolution (represented by sulfite consumption) (Fig. 4e,f), and thus consequently enhance ibuprofen degradation by Fe(II)/sulfite process particularly at acidic pH (Fig. 4g,h). For instance, k was increased from 0.0961 to 0.2591 min⁻¹ at pH 4, and from 0.0151 to 0.0308 at pH 7, after oxygenation to the Fe(II)/sulfite process (Fig. 4g,h). The diminished catalytic activity of iron at alkaline pH, due to precipitation, caused the low oxidation capability of the Fe(II)/sulfite process, and additional oxygenation did not enhance ibuprofen removal (Fig. 4h).

Electrolysis was then manifested as a mechanism to generate nascent O₂, instead of purging gaseous oxygen, to accelerate the “bi-cycle” of Fe(II)/sulfite process (Fig. 4). This is because, the O₂ generated from water electrolysis tends to be hydrated and more actively engaged in aqueous reactions compared with purged gaseous oxygen [19]. As a result, electrolysis of 400 mL solution with 200 mA current, corresponding to 0.7 mL O₂/min generation under utmost faradaic efficiency, recovered the dissolved oxygen contents of Fe(II)/sulfite reaction solution similarly to those produced by purging 1 L/min gaseous O₂ (Fig. 4a,b). Moreover, the oxidation capability of Fe(II)/sulfite reaction was observed to be positively correlated with the oxygenation rate (supplementary material Fig. S3). As a result, increasing the current intensity led to more nascent O₂ generation, and consequently mediated stronger oxidation capability of the Fe(II)/sulfite process (Fig. 2b). We hence conclude that electrolytic oxygenation played a vital role to enhance the oxidation capability of the Fe(II)/sulfite process.

3.2.3. Mechanism 3: electro/sulfite reaction generated sulfate radical

It was also observed that, electrolysis of the sulfite solution under 200 mA current without Fe(II) catalyst (i.e., electro/sulfite process) mediated a significant fraction of ibuprofen removal (34% removal within 30 min), compared to almost no removal of ibuprofen by either electrolysis or sulfite treatment alone (Fig. 2a). Reactive intermediates generated in the electro/sulfite reaction were likely responsible for ibuprofen removal. Below we investigate the mechanism of radical generation by electro/sulfite reaction.

Previous reports indicated that sulfite could be oxidized on both active (e.g. graphite, glassy carbon) [20,26] and inert anodes (e.g. noble metal, metal oxide) [24,27], in which direct electron transfer is the predominant mechanism. Cyclic voltammetry was at first performed to detect electron transfer from sulfite to Ti/MMO anode (Fig. 5a). A peak of onset potential at 1.1 V (vs SHE) in the anodic scan of the sulfite solution climbed above the background current, corresponding to sulfite oxidation. The peak height of the signal increased while increasing the concentration of sulfite, which was consistent with the literature [23].

Fig. 5.

Electro/sulfite reaction mechanism investigated through (a,b) cyclic voltammetry, and (c) electron spin resonance assay. Conditions: (a) 0-10 mM sulfite, 100 mM Na₂SO₄, pH 7, 50 mV/s scan rate; (b) 10 mM sulfite, 0.1 mM EDTA, 100 mM Na₂SO₄, pH 7, 50 mV/s scan rate; (c) 50 mM DMPO, 1 mM sulfite, 200 mA current, 10 mM Na₂SO₄, pH 7. DMPO/SC₃^•− adduct (a^H_• = 16.0 G, a^N = 14.7 G) showed up in electro-sulfite reaction. Note: DMPO, 5,5-dimethyl-1-pyrrolidine-N-oxide.

It was further revealed that, unlike transition metal mediated sulfite auto-oxidation which requires intermediary metal-sulfite complex [9], anodic sulfite oxidation is free from participation of metal species in Ti/MMO electrode since EDTA complexation did not inhibit the electron transfer from sulfite to anode (Fig. 5b). Anodic sulfite oxidation is a pure single-electron transfer process from the adsorbed sulfite molecules to the immediate conductive surface of the Ti/MMO electrode, consistent with the mechanism of sulfite oxidation on other electrodes [20].

Sulfite oxidation on the anode is accomplished via two successive single-electron transfers involving the formation of SO₃^•− radical [20]. Herein, we attempted to further validate the radical-electron mechanism by directly capturing the intermediary SO₃^•− of the electro/sulfite reaction using electron spin resonance (Fig. 5c). A pronounced signal corresponding to DMPO/SO₃^•− adduct (a^H_• = 16.0 G, a^N = 14.7 G) [28,29] was observed in electro/sulfite reaction, whereas electrolysis or sulfite alone did not produce any such signal. The presence of SO₃^•− radical, as a result of sulfite oxidation by anode under high redox potential, corroborated the mechanism of a single-electron transfer from sulfite to anode. Following the discharge of adsorbed sulfite on the anode surface, the generated SO₃^•− could evolve into SO₄^•− via Equations 9-11 (ads denotes adsorbed). Ibuprofen degradation by SO₄^•− generated from sulfite oxidation on Pt anode was also observed (supplementary material Fig. S4), following the same mechanism. Since these transient oxysulfur radicals are relatively short-lived, the electro/sulfite reaction is hence an in situ sulfite discharge process followed by radical chain reactions (Equations 9-11) on anode surface (Fig. 6).

Fig. 6.

Illustration of the synergy mechanism of electro/Fe(II)/sulfite process for organic contaminant degradation. S(IV) denotes sulfite, including HSO₃⁻/SO₃²⁻.

$${{\mathrm{HSO}}_3}^{-}∕{{\mathrm{SO}}_3}^{2-}\to {{\mathrm{HSO}}_3}^{-}∕{{{\mathrm{SO}}_3}^{2-}}_{\text{ads}}\to {{{\mathrm{SO}}_3}^{•-}}_{\text{ads}}+{\mathrm{e}}^{-}$$ (9)

$${{{\mathrm{SO}}_3}^{•-}}_{\text{ads}}+{\mathrm{O}}_2\to {{{\mathrm{SO}}_5}^{•-}}_{\text{ads}}$$ (10)

$${{{\mathrm{SO}}_5}^{•-}}_{\text{ads}}+{{\mathrm{HSO}}_3}^{-}∕{{{\mathrm{SO}}_3}^{2-}}_{\text{ads}}\to {{\mathrm{SO}}_4}^{2-}+{{{\mathrm{SO}}_4}^{•-}}_{\text{ads}}$$ (11)

3.3. Effect of pH and sulfite dose

3.3.1. Effect of pH

Fe(II)/sulfite process efficiently removed ibuprofen at acidic pH, and its capability rapidly declined at higher pH (Fig. 7a). At pH 4, k was 0.0961 min⁻¹, and it decreased to 0.0151 min⁻¹ at pH 7. At pH higher than 7, negligible ibuprofen removal was observed. This is because iron ion starts to precipitate at pH 5 and catalytic activity quickly decays as a consequence.

Fig. 7.

Effect of (a-c) pH and (d-f) sulfite concentration over ibuprofen pseudo first-order degradation rate constant (k) by the Fe(II)/sulfite, electro-sulfite, and electro-Fe(II)/sulfite processes. Conditions: Fe(II)/sulfite process – 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, initial pH 7; electro/sulfite process – 10 μM ibuprofen, 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, initial pH 7; electro/Fe(II)/sulfite process – 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, initial pH 7. Solution pH and sulfite concentration varied as indicated.

The effect of pH over the ibuprofen degradation by electro/sulfite process was also investigated. The degradation efficiency of ibuprofen by electro/sulfite process was more evident at alkaline pH (Fig. 7b). For instance, k was improved from 0.0144 min⁻¹ at pH 7 to 0.0837 min⁻¹ at pH 10 in the electro/sulfite process. Sulfite (pKa = 7) species variation [30] subjected to solution pH presumably accounted for this, since SO₃²⁻ formed at alkaline pH (SO₃²⁻ → SO₃^•− + e⁻, E⁰ = 0.63 V vs SHE) [30] was easier to be oxidized than its counterpart in acidic pH (HSO₃⁻ → H⁺ + SO₃^•− + e⁻, E⁰ = 0.84 V vs SHE) [25]. Notably, the profile of k value against pH slightly differed from that of sulfite species variation against pH. For instance, the fraction of SO₃²⁻ was plateaued above pH 9 (Fig. 8), but k continued to exponentially increase until pH 10 (Fig. 7b), consistent with the steady-state concentration of SO₄^•− calculated from benzoic acid quantification assays (supplementary material Fig. S5). An acidic electrolyte layer in the vicinity of anode due to proton release, which lowered the pH around anode than in bulk solution [23], likely accounted for this shift in pH effect.

Fig. 8.

Sulfite species distribution against pH in the range of 0-14.

Collectively, the electro/Fe(II)/sulfite process for water treatment was composed of the electrolytically oxygenated Fe(II)/sulfite process and the electro/sulfite process. As a result, the simultaneous roles of electrolytic oxygenation (that remarkably enhanced Fe(II)/sulfite oxidation capability at acidic pH) (Fig. 7a) and anodic sulfite oxidation (that significantly mediated ibuprofen degradation at alkaline pH) (Fig. 7b) together led to an upward parabola pattern in pH effect over ibuprofen degradation by the electro/Fe(II)/sulfite process (Fig. 7c).

3.3.2. Effect of sulfite dose

The role of sulfite concentration was also explored, as sulfite is the precursor of derived SO₄^•− radical. In the Fe(II)/sulfite process, increase of sulfite concentration initially improved the oxidation capability; however, higher concentration (over-dose) of sulfite reduced SO₄^•− generation due to self-scavenging (Fig. 7d) [31]. A typical molar ratio of 1:10 (Fe(II):sulfite) usually led to optimal substrate oxidations [10].

In contrast, ibuprofen oxidation rate by the electro/sulfite process linearly increased against sulfite concentration, and self-scavenging was not observed within the tested range (i.e., 0-2 mM sulfite) (Fig. 7e). Unlike homogeneous Fe(II)/sulfite reaction, only adsorbed sulfite molecules participated in the electron transfer during electro/sulfite process [23] followed by oxysulfur radical evolution to derive SO₄^•−. Consequently, higher sulfite concentration in bulk solution linearly increased sulfite occupancy of active sites on the anode surface, and mediated higher SO₄^•− generation efficiency via above steps.

During water treatment by electro/Fe(II)/sulfite process in neutral solution, the locally induced acidic zone and electrolytic oxygenation near the anode that both favorably promoted the Fe(II)/sulfite process, played a dominant role in the electro/Fe(II)/sulfite process, and consequently mediated a volcano pattern of k versus sulfite dose (Fig. 7f).

3.4. Dominant role of sulfate radical

The group of oxysulfur radicals (i.e., SO₃^•−, SO₅^•−, and SO₄^•−) have been found to react with organic substrates under different conditions. In details, SO₃^•− accounts for substrate removal in anaerobic assays, in which SO₅^•− and SO₄^•− do not exist [32]. Under aerobic conditions, SO₃^•− is completely transformed into SO₅^•− and SO₄^•− (Equations 5-7), which then tend to oxidize distinct species of substrates. SO₅^•− shows high reactivity towards anilines [33], whereas SO₄^•− has much broader selectivity including phenols, alcohols, and azo dyes [9]. Therefore, SO₄^•− most likely accounted for ibuprofen removal with the three aerobic processes (i.e., Fe(II)/sulfite, electro/sulfite, and electro/Fe(II)/sulfite) in this study.

A well-established differential method was adopted to identify SO₄^•−. In short, ethanol scavenges SO₄^•− at a relatively high reaction rate [31], while tert-butanol is quite inert toward SO₄^•− [34]. Utilizing this method, tert-butanol at up to 10 mM (1000-fold of ibuprofen concentration) did not affect the oxidation of ibuprofen, while the addition of ethanol significantly inhibited ibuprofen removal in the three processes (Fig. 9). In particular, with 10 mM ethanol, almost complete inhibition of ibuprofen oxidations was observed in three processes. As expected, SO₄^•− was the predominant radical responsible for the oxidation of ibuprofen in all three processes.

Fig. 9.

Identification of sulfate radical in (a) Fe(II)/sulfite process, (b) electro/sulfite process, and (c) electro/Fe(II)/sulfite process through radical-scavenging assays. Conditions: (a) 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, initial pH 7; (b) 10 μM ibuprofen, 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, initial pH 7; (c) 10 μM ibuprofen, 0.1 mM Fe(II), 1 mM sulfite, 10 mM Na₂SO₄, 200 mA, initial pH 7.

Furthermore, the attacking mode of SO₄^•− has been investigated previously [35]. In details, SO₄^•− abstracts an electron from the organic compound, and forms stable sulfate anion (i.e., SO₄²⁻). The oxidation products of the organic substrate are at first an addition product between the aromatic ring and sulfate group, followed by its hydrolysis into hydroxylated aromatic ring. The intermediate organic radicals will continuously propagate until stable products are formed. To investigate the degradation mechanism, degradation products of ibuprofen by the electro/Fe(II)/sulfite process were scanned (Q1 MS mode) in the range of 50 to 300 Da using HPLC-MS (supplementary material Text S2), and several fingerprint products of ibuprofen transformation by SO₄^•− were identified (supplementary material Fig. S6 and S7). Specifically, SO₄^•− at first oxidized either the butyl group leading to byproduct 2 (4-(1-carboxyethyl)benzoic acid), or carboxyl group leading to byproduct 3 (4-isobutylacetophenone). Both byproducts 2 and 3 could be oxidized by SO₄^•− into byproduct 4 (4-acetylbenzoic acid), followed by further oxidative transformations. The proposed ibuprofen degradation pathway was consistent with previous reports [36-38], validating the contribution of SO₄^•− during the transformation.

3.5. Treatment of other water contaminants

Besides ibuprofen, several other water contaminants were tested (supplementary material Fig. S8). Particularly, 4-chloroaniline and N,N-dimethylaniline are precursors for industrial dyes or insecticides [39]; paracetamol is a drug for fever or headache [40]; and Orange II, Rhodamine B, and Reactive Brilliant Red are common dyes of textile industry [41]. They are considered recalcitrant contaminants since microbial activities are ineffective towards their degradations. Their quantification methods were included in Table S1 of supplementary material. It was shown that electro/Fe(II)/sulfite process could remove 70-97% of these compounds at a concentration of 10 μM at pH 4, significantly higher than Fe(II)/sulfite or electro/sulfite process (supplementary material Fig. S8). The removal efficiencies were slightly lower at pH 7, ranging from 52-82% by electro/Fe(II)/sulfite process (supplementary material Fig. S8). The structure and molecular weight are shown in Table S2 of supplementary material. The slightly lower removal efficiencies of 4-chloroaniline and Reactive Brilliant Red are possibly ascribed to the low electron density of aromatic rings due to the presence of heteroatoms such as nitrogen and chloride. Nevertheless, it is believed that, electro/Fe(II)/sulfite process could non-selectively degrade a wide variety of water contaminants, showing potential applications into practical water treatment.

3.6. Treatment of field water bodies

Our final goal is to apply electro/Fe(II)/sulfite process to treat natural contaminated waters. However, water bodies in the fields often contain strong buffering agent that might compromise the water treatment efficiency of most technologies. Carbonate is the most common buffering agent in field water bodies, in the range of 0.2 – 2 mM [42,43]. It was observed that electro/Fe(II)/sulfite process could efficiently resist carbonate buffering effect (supplementary material Fig. S9), and removed up to 76% ibuprofen in solution containing 2 mM carbonate (supplementary material Fig. S9l). Moreover, treatment of water matrices of various carbonate contents by electro/sulfite process, led to over 80% removal of 10 μM ibuprofen after three consecutive additions of sulfite (supplementary material Fig. S10 and S11), highlighting a promising use of electro/sulfite process to treat water if iron is prohibited.

Ibuprofen removals from water samples collected from the fields were then tested by the Fe(II)/sulfite, electro/sulfite, and electro/Fe(II)/sulfite processes, respectively. Specifically, two groundwater samples with high carbonate alkalinity were extracted from Superfund sites of karst topography in Puerto Rico, and two typical surface waters containing high organic carbon contents were sampled from local Boston ponds (Table 1). Considering the complex conditions of natural groundwater matrices, we doubled the concentrations of Fe(II) and sulfite. Intrinsic ions in the field water bodies, together with the added Fe(II) and sulfite, supported the conductivity of the electrolysis. As a result, over 80% of ibuprofen was removed by the electro/Fe(II)/sulfite process in samples collected from two groundwater wells (referred to as TAL and MIT), compared to lower ibuprofen removal by both Fe(II)/sulfite and electro/sulfite processes (Fig. 10a,b). On the other hand, removals of ibuprofen in the surface water samples (referred to as JAM and WAR) by electro/Fe(II)/sulfite process were slightly lower (Fig. 10c,d), presumably due to the high content of organic carbon present in these samples (Table 1). Organic carbon owns similar reactivity with ibuprofen towards SO₄^•−, and such competition would spare a large fraction of SO₄^•− for organic content oxidation. Besides, the high carbonate alkalinity (Table 1) might also compromise ibuprofen oxidation by affecting solution pH.

Table 1.

Characterizations of used field water matrices for ibuprofen removal assays.

	TAL	MIT	JAM	WAR
pH	7.70	7.51	7.34	7.81
DO (mg/L)	7.94	7.86	8.14	7.94
ORP (mV, vs Ag/AgCl)	315	281	347	329
TOC (mg/L)	79.8	73.37	307	448
Alkalinity (mg/L) a	150	123	48	81

Note: TAL, Tallonal Superfund site; MIT, Pozo Mita Superfund site; JAM, Jamaica pond; WAR, Wards pond. DO, dissolved oxygen; ORP, oxidation/reduction potential; TOC, total organic carbon.

^awater alkalinity was presented in the form of CaCO₃ (ref. 66).

Fig. 10.

Degradation of ibuprofen in field groundwater matrices extracted from (a) TAL and (b) MIT Superfund sites of Puerto Rico, and in natural surface water matrices collected from local (c) JAM and (d) WAR ponds of Boston, Massachusetts. Conditions: Fe(II)/sulfite process – 10 μM ibuprofen, 0.2 mM Fe(II), 2 mM sulfite, initial pH 7; electro/sulfite process – 10 μM ibuprofen, 2 mM sulfite, 200 mA, initial pH 7; electro/Fe(II)/sulfite process – 10 μM ibuprofen, 0.2 mM Fe(II), 2 mM sulfite, 200 mA, initial pH 7. Note: TAL, Tallonal Superfund site; MIT, Pozo Mita Superfund site; JAM, Jamaica pond; WAR, Wards pond.

3.7. Feasibility and energetic analysis

Feasibility and energetic analysis of electro/Fe(II)/sulfite process are performed to gain more insights. Various sophisticated water treatment techniques have been developed to meet different demands. For example, wet oxidation under high air pressure is an energy-intensive process and is used to treat concentrated biomass and organic matters [44,45]. Membrane filtration (such as reverse osmosis and ultrafiltration) is installed at the terminal of a centralized water treatment facility to obtain ultrapure water, and this unit faces the problem of costly membranes [46]. More platforms are discussed in Table S3 of supplementary material. Advanced oxidation process (AOP) using various oxidative radicals such as SO₄^•− is a quite common water pretreatment process [47]. AOP is usually utilized to remove recalcitrant contaminants so as to alleviate the burden of following units. The herein developed electro/Fe(II)/sulfite process belongs to AOP family, which is used to preliminarily eliminate most hazards. In practice, this process can be performed by simply adding iron ions and collected sulfite under electric current.

The energetic analysis is then conducted to evaluate the energy cost of electro/Fe(II)/sulfite process. Over the course of reactions, applied electric current was 200 mA, and cell voltage was around 5 V, and experiments endured for 30 min. Therefore, the energy used to treat 400 mL solution was determined to be 0.5 W·h by Equation 12. Given that the average industrial electricity price is 6.92 cent/(KW·h) in US [48], the electric treatment cost is 8.65 cent/m³ as estimated by Equation 13, where P is the price of electricity and E is the energy cost to treat 1 m³ water.

$$\mathit{Energy}={\int }_0^{t}I\times U\times t$$ (12)

$$\mathit{Cost}=P\times E$$ (13)

The cost of electro/Fe(II)/sulfite process is significantly lower than many other water treatment processes. For example, Fenton reaction (Fe(II)/H₂O₂) generating hydroxyl radical (HO^•) is a classical AOP in water treatment industry [49,50]. However, several aspects increase the treatment cost of Fenton reaction. At first, Fe(II) is much more catalytically active than Fe(lll) to transform H₂O₂ into HO^• [49]. This requires continuous addition of Fe(II) ions during practical water treatment. Besides the increased reagent cost, the treatment of iron hydroxide sludge is another concern [51]. Secondly, the price of liquid H₂O₂ is $750/ton (50%) [52], and the overall cost could be higher by considering its storage and transportation. It is reported that the cost for Fenton process treatment is around $0.6/m³ (after translation of euro into US dollar) [53], and the cost could rise up to $10-30/m³ when the targets are high-strength wastewater [54,55]. In sharp contrast, utilization of sulfite for water treatment is a net gain in terms of cost, because sulfite is otherwise a waste. Besides Fenton process, the cost of electro/Fe(II)/sulfite process is also significantly lower than many other well-established water purification facilities. For instance, a standard ultrafiltration system costs around $0.28/m³ in a large plant [56], and costs of ion exchange and reduction/coagulation for heavy metal removal are $0.29/m³ and $0.38/m³, respectively [57]. The electricity energy cost of electro/Fe(II)/sulfite process (8.65 cent/m³) is significantly lower than these processes. Collectively, our proposed electro/Fe(II)/sulfite process for water treatment is not only sustainable but also rather low-cost.

4. Possible procedures to realize the proposed energy-water-environment nexus

The electro/Fe(II)/sulfite process has been demonstrated to effectively treat water contaminants, with sulfite to represent the pure and aqueous form of SO₂. Other concerns to be considered are the capture and purification process of SO₂, and the electricity source to power this process. Here we propose sequential procedures to realize this energy-water-environment nexus (Fig. 11).

Fig. 11.

Schematic illustration of procedures to realize the proposed energy-water-environment nexus.

(1). Capture of SO₂ from flue gas. SO₂ can be captured and concentrated from the flue gas by alkaline solutions such as ammonia [58-60]. At this stage, most particulate matters are removed. Captured SO₂ then enters the next step.

(2). Purification of captured SO₂. Among all the hazards in flue gas, mercury poses the greatest ecological/health concern due to its toxicity. Well-established mercury removal techniques such as Thief process [61] or other processes [62,63] can be adopted. Other control processes are necessary if additional hazards are identified. At this stage, the captured SO₂ is relatively clean and safe for water treatment.

(3). Solar panels to power electro/Fe(II)/sulfite process. Harnessing solar energy with a high efficiency has been well studied for the past few decades [64,65], and could be used to supply the power of electric equipment.

5. Conclusion

In summary, we manifested a novel energy-water-environment nexus, which is using waste SO₂ from electricity energy production to alleviate water contamination. As a demonstration, an electro/Fe(II)/sulfite process was used to degrade ibuprofen contaminant using both synthetic water and field water matrices. The excellent water treatment efficiency of electro/Fe(II)/sulfite process is due to following synergetic mechanisms: 1) anode-induced acid zone facilitates Fe(II)/sulfite process; 2) electrolytic oxygenation supports the “bi-cycle” of sulfite evolutions; and 3) sulfite oxidation on anode generates oxysulfur radicals for contaminant degradation. The strong potential of electro/Fe(II)/sulfite process was further demonstrated by non-selective degradation of various recalcitrant contaminants and effectiveness in highly carbonated media. Feasibility and energetic analysis were performed to indicate the advantages of this process. It is believed that the tens of millions of tons SO₂ “waste” holds numerous chemical energy for primary water treatment. Implications of this energy-water-environment nexus are significant: SO₂ mitigation and sustainable water treatment are simultaneously achieved.

Highlights:

• Achieved simultaneous sulfur dioxide mitigation and water contaminant treatment

• Synergy mechanism of electro/Fe(II)/sulfite process was investigated

• Feasibility and energetic analysis indicated it is a viable process

• Procedures to realize the energy-water-environment nexus were proposed