Scaling-up of microbial electrosynthesis with multiple electrodes for in situ production of hydrogen peroxide

Summary

Microbial electrosynthesis system (MES) has recently been shown to be a promising alternative way for realizing in situ and energy-saving synthesis of hydrogen peroxide (H₂O₂). Although promising, the scaling-up feasibility of such a process is rarely reported. In this study, a 20-L up-scaled two-chamber MES reactor was developed and investigated for in situ and efficient H₂O₂ electrosynthesis. Maximum H₂O₂ production rate of 10.82 mg L⁻¹ h⁻¹ and cumulative H₂O₂ concentration of 454.44 mg L⁻¹ within 42 h were obtained with an input voltage of 0.6 V, cathodic aeration velocity of 0.045 mL min⁻¹ mL⁻¹, 50 mM Na₂SO₄, and initial pH 3. The electrical energy consumption regarding direct input voltage was only 0.239 kWh kg⁻¹ H₂O₂, which was further much lower compared with laboratory-scale systems. The obtained results suggested that the future industrialization of MES technology for in situ synthesis of H₂O₂ and further application in environmental remediation have broad prospects.

Graphical Abstract

Highlights

• •Up-scaled microbial electrosynthesis with multiple electrodes to synthesize H₂O₂
• •The H₂O₂ yield was higher than that of laboratory-scale systems using graphite cathode
• •Energy consumption was lower than that of laboratory-scale (bio)electrochemical systems
• •Systematic evaluation of the influence of operating parameters on H₂O₂ production

Electrochemistry; biotechnology; engineering; materials science

Introduction

Hydrogen peroxide (H₂O₂) is a green and environment-friendly oxidant that has been widely used in industrial and environmental applications (Kelly et al., 2019; Sheng et al., 2020; Zhao et al., 2020). According to statistics, the world's H₂O₂ production capacity exceeded 5 million tons in 2015 and maintained steady growth, most of which were synthesized based on the anthraquinone method, which has the disadvantage of high energy consumption and high-risk coefficient (Sheng et al., 2020; Zhao et al., 2020). In addition, due to the unstable chemical properties of H₂O₂, this also provides challenges for transportation and storage and correspondingly increases costs (Chen et al., 2018). For many applications, such as advanced oxidation processes for water treatment, the required concentration of H₂O₂ was usually below 1,000 mg L⁻¹. In this context, it is urgent and important to develop alternative methods for the in situ synthesis of H₂O₂. In the past years, among others, electrochemical methods have received extensive attention due to their relatively high H₂O₂ synthesis efficiency and easy operation (only need electrical energy as input) (Campos-Martin et al., 2006; Siahrostami et al., 2013). Despite its promise, the electrochemical methods are still suffering from the disadvantage of large energy costs (Perry et al., 2019).

Recently, the microbial electrosynthesis system (MES), in which electrochemically active microorganisms can obtain energy from the oxidation of organics in sewage, have been demonstrated as an efficient and cost-effective means for the production of H₂O₂ (Hassan et al., 2019; Li et al., 2018a, 2018b,bib_Li_et_al_2018a; Zhao et al., 2020). Specifically, in a typical two-chamber MES, the anodic electrochemically active microorganisms can oxidize organic matter in wastewater and generate electrons, protons, and carbon dioxide via Equation (1) (Hassan et al., 2019). Among them, the generated electrons and protons are further transferred to the cathode chamber through wires and cation exchange membrane (CEM), respectively, and are further combined with oxygen to generate H₂O₂ through two-electron reduction via Equation (2) (Hassan et al., 2019).

C_aH_bO_c + (2a-c)H₂O → aCO₂ + (b+4a-2c)H⁺ + (b+4a-2c)e⁻ (Equation 1)

O₂ + 2H⁺ + 2e⁻ → H₂O₂ (Equation 2)

To this end, the microbial fuel cell (MFC) and its derived microbial electrolysis cell (MEC)-based MES have been proposed for H₂O₂ production, and the range of produced H₂O₂ was 79–1,300 mg L⁻¹ (Li et al., 2017). This production level would be able to satisfy the application in the water treatment process. Nonetheless, it was worth noting that most of the MES studies on the in situ synthesis of H₂O₂ including MFC or MEC were conducted with laboratory-scale reactors ranging from milliliters to hundreds of milliliters. Scaling-up is of importance to prove the feasibility and economic profits of such a process for large-scale application. However, a systematic evaluation of the scaling-up of MES for H₂O₂ synthesis is rarely reported. So far, the only scaling-up study was reported by Sim et al. who developed a 110-L dual-chamber up-scaled MEC reactor (10 L of cathode chamber and 1.6–2.0 V voltage input) to synthesize H₂O₂. Nevertheless, the H₂O₂ titer was less than 100 mg L⁻¹ after 20 days of operation, and the corresponding H₂O₂ conversion efficiency was only about 7.2% in a form of short communication (Sim et al., 2018). Thus, more efforts are required toward the scaling-up of the MES-based H₂O₂ production. Among others, the electrode material is one of the important factors for a successful scaling-up, which is directly related to the current efficiency (CE), two-electron oxygen reduction reaction (ORR) selectivity (Equation (2)), and energy consumption during the in situ electrosynthesis of H₂O₂ in the cathode. In traditional MFC for bioelectricity production, noble metals (e.g., titanium and platinum) are usually used as electrode materials or catalysts for favoring four-electron oxygen reduction to produce H₂O via Equation (3) (Chung et al., 2020; Hassan et al., 2019).

O₂ + 4H⁺ + 4e⁻ → 2H₂O (Equation 3)

In contrast, carbon-based electrodes without containing noble metals, especially the commonly used graphite-based electrodes such as graphite plate and graphite felt, are often applied in the electrosynthesis of H₂O₂ given their merits of strong electrical conductivity, high two-electron ORR selectivity, inexpensive, long lifetime, weak H₂O₂ decomposition effect, and ease of scaling up (An et al., 2019; Chung et al., 2020; Li et al., 2016, 2017). In addition, it should be noted that although most of the current laboratory-scale studies related to bioelectrochemical synthesis of H₂O₂ used gas diffusion electrodes (GDE) as the cathode to obtain a high H₂O₂ yield, their mechanical property makes it difficult to be scaled up (Li et al., 2016). Moreover, the effect of acidic pH (2–3) on the stability of the GDE and the formation of possible refractory by-products are two main challenges for in situ wastewater treatment (Wang et al., 2020a). In addition, almost all these laboratory-scale MES reactors were only equipped with a single electrode including a bioanode and an abiotic cathode in each chamber, which tended to cause system instability.

This study successfully scaled up an MES with multiple graphite electrodes to 20 L (dual-chamber and 10 L of each chamber) and comprehensively investigated the influence of operating parameters including input voltage, cathodic aeration velocity, catholyte pH, and electrolyte nature and concentration on the in situ synthesis of H₂O₂. To the best of our knowledge, this is the first scaled-up MES reactor equipped with highly stable graphite plate electrodes for in situ H₂O₂ synthesis. The present study not only verified the feasibility of such scaled-up MES for H₂O₂ synthesis but also accumulated valuable experience for subsequent industrial applications on the treatment of wastewater.

Results and discussion

The performance of 20-L scaled-up MES reactor for H₂O₂ synthesis

A series of experiments was carried out to demonstrate the feasibility of the 20-L scaled-up MES reactor for H₂O₂ synthesis. Figure 1 shows the change of H₂O₂ concentration with operating time at the cathode in the 20-L scaled-up MES reactor. The concentration of H₂O₂ increased monotonically with operating time and reached about 454.44 mg L^-1 at 42 h. Specifically, the synthesis rate of H₂O₂ in the first 16 h was significantly higher than that in the next 26 h; especially after 36 h, the increase of H₂O₂ concentration was relatively slow. Moreover, the cathodic pH varied during the process (Figure S1). A rapid rise in pH was observed as the reaction progressed. The cathodic pH reached 8.58 in the first 16 h, and then gradually increased and finally reached 10.07 at 42 h. The fast increase of pH observed in the first 16 h was due to the fast reduction of two electrons of oxygen to H₂O₂ via Equation (2). After that, when the cathodic pH turned to alkaline, the main reaction occurring on the cathode surface was the reduction of oxygen to HO₂^- via Equation (4) (Brillas et al., 1997; Moreira et al., 2017).

O₂ + H₂O + 2e^- → HO₂^- + OH^- (Equation 4)

Figure 1.

Feasibility verification of this 20-L scaled-up MES reactor regarding H₂O₂ production

Operating conditions: input voltage of 0.6 V, cathode aeration velocity of 0.045 mL min⁻¹ mL⁻¹, initial catholyte pH of 3, and electrolyte nature and concentration of 50 mM Na₂SO₄, respectively. Control 1: without cathodic aeration. Control 2: open circuit. Control 3: without input voltage.

In addition, the in situ-generated H₂O₂ would be further reduced to H₂O and OH⁻ under acidic or alkaline conditions via Equations (5) and (6), respectively, and the decomposition of H₂O₂ would also occur simultaneously via Equations (7) and (8), thus more H₂O₂ would be decomposed in a longer operating time (Nidheesh and Gandhimathi, 2012; Qiang et al., 2002; Xia et al., 2015). In short, the above-mentioned results also proved the feasibility of the graphite plate as the cathode of this scaled-up reactor for two-electron ORR to generate H₂O₂ via Equation (2).

H₂O₂ + 2H⁺ + 2e⁻ → 2H₂O (Equation 5)

HO₂^- + H₂O + 2e⁻ → 3OH⁻ (Equation 6)

2H₂O₂ → O₂ + 2H₂O (Equation 7)

H₂O₂ + OH⁻ → HO₂^- + H₂O (Equation 8)

Besides, the control experiments including the MES reactor without cathodic aeration (control 1), open circuit (control 2), and without input voltage (control 3) were also conducted and the results were also exhibited in Figure 1. As expected, no H₂O₂ production was detected when the system was in an open circuit. According to Equation (2), as no electrons were flowing from the anode to the cathode in the open-circuit state, there was no H₂O₂ produced. Similarly, the reactor without cathodic aeration or input voltage (corresponding to operating at MFC mode) only produced a low concentration of H₂O₂ (below 20 and 70 mg L⁻¹, respectively). The low H₂O₂ production could be attributed to insufficient oxygen or electrons for the synthesis of H₂O₂ based on Equation (2). Similar results were also found in the previous studies by using a laboratory-scale MES and electrochemical system (Li et al., 2017; Yu et al., 2015a). Therefore, the obtained results highlighted that the reactor operating in MEC mode (requiring a small amount of voltage input) was effective and superior in H₂O₂ productivity compared with operating in MFC mode. However, it is worth emphasizing that around 70 mg L⁻¹ of H₂O₂ was still produced when this scaled-up reactor was operated in the MFC mode, which further proved that the scaled-up reactor can generate enough H₂O₂ to meet the water and wastewater treatment requirements (e.g., 5–50 mg L⁻¹ of H₂O₂ were needed for water disinfection and micro-pollutant removal) (Chung et al., 2020).

In addition to studying the H₂O₂ synthesis at the cathode, we also monitored the current and cathode potential of the system under different conditions. As depicted in Figure S1, the stable and relatively high current density output (around 2.86 A m⁻²) together with a stable cathode potential (around −0.6 V) were observed during the entire operation, respectively. It proved that the biological anode of this scaled-up system was stable during the test. Furthermore, as the performance of the in situ synthesis of H₂O₂ via MES has been shown to depend on the transfer of effective microbial electrons to the anode (Logan and Regan, 2006), the observed stable current output demonstrated the feasibility of further application of large-scale MES in situ synthesis of H₂O₂ in the future. In contrast, the current density outputs obtained in the control conditions (around 1.25, 0, and 0.56 A m⁻²) were significantly lower than the normal operating condition.

Overall, the above-mentioned results proved that the scaled-up MES reactor with sufficient cathodic aeration and electron flow can provide stable and efficient in situ production of H₂O₂.

Effect of operating parameters

Input voltage

The H₂O₂ synthesis was generally slow in MFC mode due to the low circuit current (Li et al., 2017). As an alternative, H₂O₂ production can be significantly increased by applying a small amount of applied voltage (normally below 0.8 V) in MEC mode. In general, a higher input voltage would lead to a higher current, thus the enhanced electron flow would increase the rate of H₂O₂ synthesis. However, it would not be the case, because the higher input could also promote the occurrence of side reactions, resulting in a decrease in current efficiency, and produce similar or even lower H₂O₂ yields compared with lower input voltage (Moreira et al., 2017; Oturan et al., 2018). Therefore, seven groups of different input voltages (0.1, 0.2, 0.4, 0.5, 0.6, 0.7, and 0.8 V) were selected to investigate their influence on H₂O₂ synthesis. As shown in Figure 2A, the cumulative H₂O₂ production in 42 h was 125.36, 236.36, 314.93, 353.40, 454.44, 408.70, and 385.71 mg L⁻¹, respectively, at the input voltage ranging from 0.1 to 0.8 V. The results indicated that the optimum input voltage was 0.6 V and further increasing the input voltage (0.7 and 0.8 V) led to a decrease in H₂O₂ production. Likewise, the variation trend of current efficiency presented in Figure 2B was also in line with the yield of H₂O₂, and the corresponding values were 1.63%, 2.42%, 2.80%, 3.41%, 2.66%, and 2.37% at input voltage ranging from 0.1 to 0.8 V, respectively. The system current and cathode potential were shown in Figure S2, respectively. It was observed that both system current density (from 1.63 to 3.44 A m⁻²) and cathode potential (in terms of absolute value, from −0.45 to −0.66 V) increased continuously with the increase of the applied voltage from 0.1 to 0.8 V. Thus, the aforementioned results can be explained from two aspects. On the one hand, the increase of current density led to accelerated electron transfer on the surface of the cathode electrode, thereby promoting oxygen reduced to the H₂O₂ via Equation (1) (Yu et al., 2015a, 2015b). On the other hand, higher currents may also enhance the occurrence of side reactions, such as accelerating the conversion of the produced H₂O₂ into H₂O via Equations (5), (7), and (8); even reduction of oxygen to H₂O via Equation (3) (Chen et al., 2015a; Nadais et al., 2018; Wang et al., 2020b); or hydrogen evolution via Equation (9) (Chen et al., 2015a).

2H⁺ + 2e⁻ → H₂ (Equation 9)

Figure 2.

Effect of input voltage

(A) H₂O₂ production and (B) current efficiency in the scaled-up reactor. Operating conditions: cathode aeration velocity of 0.045 mL min⁻¹ mL⁻¹, initial catholyte pH of 3, and electrolyte nature and concentration of 50 mM Na₂SO₄, respectively.

Similar results were also observed in the previous studies regarding abiotic electro or MES for H₂O₂ production (Chen et al., 2015a; Li et al., 2016; Yu et al., 2015a). Notably, the highest H₂O₂ yield (at 0.6 V) obtained in this study was around 10.82 mg L⁻¹ h⁻¹, which was about 3.3 times higher than that observed in the highest value from laboratory-scale (cathodic working volume of 14 mL) MEC reactor (3.92 mg L⁻¹ h⁻¹) (Li et al., 2016). Moreover, it was also found that the production rate of H₂O₂ was slightly higher than that of the laboratory-scale (cathodic working volume of 100 mL) MES reactor is driven by salinity gradient (10.80 mg L⁻¹ h⁻¹) (Li et al., 2017). Taking into account the H₂O₂ production rate, the current efficiency, and the energy consumption, an input voltage of 0.6 V was the optimal value among the investigated voltages.

Cathodic aeration rate

As observed in the previous part, the cathode aeration was essential to H₂O₂ production. However, it was not that the faster the aeration rate, the higher the concentration of H₂O₂ produced at the cathode. Excess aeration could reduce H₂O₂ production and also increase energy consumption. Thus, set an optimal aeration rate could not only maximize the H₂O₂ production at the cathode but could also save energy consumption. In this part, the aeration rates ranging from 4 to 500 mL min⁻¹, corresponding to the cathode aeration velocities of 0.00045–0.056 mL min⁻¹ mL⁻¹, were adopted to study its effect on the H₂O₂ synthesis at the cathode. As shown in Figure 3, the H₂O₂ concentration accumulated at the cathode increased monotonically with the increase of the aeration rate in 42 h, which indicated that the selected aeration rate range has no negative effect on the H₂O₂ production at the cathode. When the aeration velocity was increased from 0.045 to 0.056 mL min⁻¹ mL⁻¹, the accumulated H₂O₂ concentration only increased slightly (from 454.44 to 459.30 mg L⁻¹). Additionally, the concentration of cathodic DO also increased with the increase of aeration velocity (Figure 3). Among them, it can be seen that when the aeration velocity was lower than 0.045 mL min⁻¹ mL⁻¹, the catholyte was unsaturated, whereas the catholyte was in a saturated or supersaturated state when the aeration rate reached 0.045 mL min⁻¹ mL⁻¹or above. These results were in agreement with previous studies using photocatalytic or photo-assisted electrocatalytic processes for H₂O₂ production (Salmerón et al., 2019; Yu et al., 2018). The increased aeration rate increased the catholyte DO and promoted the mass transfer rate of oxygen in the catholyte, which was beneficial to the production of H₂O₂ (Luo et al., 2015a; Yu et al., 2015a). When the aeration rate was increased from 400 to 500 mL min⁻¹, the insignificant increase in the accumulated H₂O₂ concentration could be due to the following reasons. First, the increase in the number of bubbles generated by aeration would increase the internal resistance of the system (Figure S3) (Freakley et al., 2013; Salmerón et al., 2019). Second, the further increase in the size of bubbles generated was adverse to their adsorption to the cathode (Nadais et al., 2018). Therefore, it can be inferred that further increasing the aeration rate over 400 mL min⁻¹ may make the H₂O₂ production at the cathode adverse. The aeration rate of 400 mL min⁻¹ was selected as the optimized aeration rate for subsequent experiments.

Figure 3.

Effect of cathodic aeration velocity

(A) H₂O₂ production and (B) current efficiency in the scaled-up MES reactor. Operating conditions: input voltage of 0.6 V, initial catholyte pH of 3, and electrolyte nature and concentration of 50 mM Na₂SO₄, respectively.

Initial catholyte pH

Apart from the input voltage and cathodic aeration velocity, initial catholyte pH was another key factor because the electrosynthesis of H₂O₂ required the participation of protons according to Equation(2). Accordingly, the impact of initial catholyte pH (3, 5, 7, and 9) on the cathodic electrosynthesis of H₂O₂ was studied. The results showed that the change of initial pH significantly affected the cathodic synthesis of H₂O₂, and cumulative concentration reached approximately 454.44, 412.71, 395.71, and 316.36 mg L⁻¹, respectively, after 42-h operation (Figure 4A). Similarly, the current efficiency and H₂O₂ rates also showed the same variation trend, resulting in final values of 3.41%, 3.10%, 2.97%, and 2.37% and 10.82, 9.83, 9.42, and 7.53 mg L⁻¹ h⁻¹, respectively (Figures 4B and S4). In addition, the change of catholyte pH value over time was shown in Figure 4C. The rapid rise of pH indicated that the consumption rate of protons at the cathode for the synthesis of H₂O₂ was faster than the generation rate at the anode (latterly entered the cathode through the CEM). As mentioned previously, under alkaline conditions, especially at pH > 9, H₂O₂ mainly existed in the form of HO₂^-, which could catalyze the decomposition of H₂O₂ via Equation (10) (Luo et al., 2015b; Sheng et al., 2011).

H₂O₂ + HO₂^- → ^·OH + ^·O₂^- + H₂O (Equation 10)

Figure 4.

Effect of initial catholyte pH

(A) H₂O₂ production, (B) current efficiency in the scaled-up MES reactor, and (C) pH variation during the process. Operating conditions: input voltage of 0.6 V, cathodic aeration velocity of 0.045 mL min⁻¹ mL⁻¹, and electrolyte nature and concentration of 50 mM Na₂SO₄, respectively.

This also explained why the cumulative H₂O₂ concentration was lower with the initial pH 9 than with other pH values. The same experimental results were found in the electrochemical synthesis of H₂O₂ using unmodified graphite electrodes (Yu et al., 2015a). Therefore, when the initial pH was set to 3, more H₂O₂ can be synthesized at the cathode, and this value conformed to the optimal pH range (2–4) of the traditional electro-Fenton and bio-electro-Fenton processes (Li et al., 2018a, 2018b; Moreira et al., 2017). It should be noted that the cumulative H₂O₂ concentration under the neutral and alkaline conditions observed in this study still meet the needs of water treatment, which may broaden the applicability of the scaled-up MES in other fields such as combining other technologies (e.g., UV-based advanced oxidation processes) or catalysts for water treatment. In this way, it may also reduce the chemical cost, as there was no need to use acid and alkali to adjust the pH before water treatment.

Electrolyte nature and concentration

Generally, research processes related to electrochemical systems required high conductivity in the electrolyte to enhance the flow of electrons. In this study, we have selected three widely used electrolytes including Na₂SO₄, NaCl, and Na₂CO₃ with an initial concentration of 50 mM to explore their effects on the synthesis of H₂O₂ at the cathode. As presented in Figures 5A and 5B, when Na₂SO₄ was used as the electrolyte, the accumulated H₂O₂ concentration in 42 h was 454.44 mg L⁻¹, followed by Na₂CO₃ and NaCl (358.50 and 262.17 mg L⁻¹, respectively). The reasons why Na₂SO₄ as an electrolyte was superior to Na₂CO₃ and NaCl can be explained as follows: (1) chloride ion in NaCl could react with H₂O₂ to generate higher reactive HClO and/or chlorine derivatives via (Equation 11), (Equation 12), (Equation 13), (Equation 14), (Equation 15), (Equation 16), (Equation 17), (Equation 18) (De Laat et al., 2004; Lai et al., 2020; Moreira et al., 2017); (2) CO₃²⁻ itself and its hydrolysis product HCO₃^- (main role) could cause the decomposition of H₂O₂ via (Equation 19), (Equation 20), (Equation 21), (Equation 22) (Attiogbe and Francis, 2011; Fabian, 1995; Xia et al., 2020);

H₂O₂ + Cl⁻ → ClO⁻ + H₂O (Equation 11)

H₂O₂ + ClO⁻ → Cl⁻ + O₂ + H₂O (Equation 12)

H₂O₂ + e⁻ → ^⋅OH + OH⁻ (Equation 13)

^⋅OH + Cl⁻ → ^⋅ClOH⁻ (Equation 14)

^⋅ClOH⁻ + Cl⁻ → ^⋅Cl₂^- + OH⁻ (Equation 15)

^⋅Cl₂^- → ^⋅Cl + Cl⁻ (Equation 16)

^⋅Cl₂^- + H₂O₂ → ^⋅HO₂ + H⁺ + 2Cl⁻ (Equation 17)

^⋅Cl + H₂O₂ → ^⋅HO₂ + H⁺ + Cl⁻ (Equation 18)

CO₃²⁻ + H₂O ↔ HCO₃^- + OH⁻ (Equation 19)

CO₃²⁻ + ^⋅OH → ^⋅CO₃^- + OH⁻ (Equation 20)

HCO₃^- + H₂O₂ → HCO₄^- + H₂O (Equation 21)

^⋅CO₃^- + H₂O₂ → HCO₃^- + ^⋅HO₂ (Equation 22)

Figure 5.

Effect of electrolyte nature and concentration

(A and C) H₂O₂ production and (B and D) current efficiency in the scaled-up MES reactor. Operating conditions: input voltage of 0.6 V, cathodic aeration velocity of 0.045 mL min⁻¹ mL⁻¹, and electrolyte nature (NaCl, Na₂SO₄ and Na₂CO₃), and initial Na₂SO₄ concentration of 10, 25, 50, and 100 mM, respectively.

(3) compared with NaCl and Na₂CO₃, Na₂SO₄ has higher conductivity (9.02 ms cm⁻¹ vs 8.02 ms cm⁻¹ of Na₂CO₃ and 5.66 ms cm⁻¹ of NaCl) and correspondingly higher current (see Figure S5) when the initial concentration was 50 mM, and thus, it has the highest production of H₂O₂ (Zhou et al., 2007).

Although the side reactions listed earlier that can affect the synthesis of H₂O₂ will also produce some free radicals with strong oxidizing ability (e.g., ^⋅Cl and ^⋅OH), the amount of those strong oxidizing agents produced is very small. Notably, the successful application of bioelectrochemical synthesis of H₂O₂ for the disinfection of gray water and wetland effluent has been recently reported (Arends et al., 2014; Murawski, 2018). However, the dose of H₂O₂ needed to reach the target fecal coliform level in a real municipal effluent from the wastewater treatment plant ranged from 106 to 285 mg L⁻¹, which made the direct application of H₂O₂ as a disinfectant very inappropriate (Wagner et al., 2002). Furthermore, our previous study has also shown that the bioelectrochemical synthesis of H₂O₂ has a poor killing effect on E. coli with an initial concentration of 10⁷ CFU mL⁻¹, whereas the combination of H₂O₂-producing MEC with Fenton process, named bio-electro-Fenton process, exhibited a significant improvement on the E. coli inactivation (Zhou et al., 2018). Besides, a recent study exhibited that the strong alkaline catholyte (pH > 13) produced by the MFC-based electro-osmosis process can effectively inactivate pathogens, but considering its pH also limits its subsequent application to the disinfection of real wastewater (Gajda et al., 2020). In general, although beneficial attempts have been made to directly apply bioelectrochemical systems for water disinfection, considering the treatment efficiency, integration of bioelectrochemical synthesis of H₂O₂ with other advanced oxidation processes (e.g., Fenton process) may be a more promising way for simultaneous water disinfection and micro-pollutants removal. It was also worth noting that although the application of Na₂CO₃ and NaCl also yielded higher H₂O₂ production, the presence of chloride ions (only when solution pH below 7.2) and carbonates may affect the subsequent application of the synthesized H₂O₂ for wastewater treatment (e.g., they have a scavenging effect on the free radicals [e.g., ^⋅OH] in Fenton reaction, Chen et al., 2019; Kläning and Wolff, 1985; Zhang et al., 2018). Based on the above, Na₂SO₄ was used as an electrolyte in the subsequent experiments.

Besides the electrolyte nature, the electrolyte concentration, which was related to the conductivity, was also crucial for the electrosynthesis of H₂O₂. Thus, different concentrations of Na₂SO₄ (10, 25, 50, and 100 mM) corresponding to the conductivities of 2.44, 4.95, 9.02, and 16.36 ms cm⁻¹, respectively, were tested. As shown in Figures 5C and 5D, the increase of Na₂SO₄ from 10 to 100 mM increased the continuous improvement of cathodic H₂O₂ accumulation. When the Na₂SO₄ concentration was further increased from 50 to 100 mM, the increase in the H₂O₂ synthesis was getting slower. The correlation between electrolyte concentration and H₂O₂ synthesis can be explained as follows. First, the conductivity was higher with higher electrolyte concentration applied, which further increased the current (see Figure S6), thereby promoting the production of H₂O₂ (Zhou et al., 2007). Second, to a certain extent (e.g., below 200 mM), increasing the Na₂SO₄ concentration could increase the mass transfer, thereby increasing the H₂O₂ synthesis (Jin et al., 2014). Third, excessive Na₂SO₄ may be adsorbed on the surface of the electrode, resulting in a decrease in the active site that can bind to oxygen molecules on the electrode, which in turn reduces to produce H₂O₂ (Chen et al., 2015b). Last, excessive Na₂SO₄ could also consume the generated H₂O₂ via Equations 23 and 24 (Jin et al., 2014).

SO₄²⁻ + ^⋅OH → ^⋅SO₄^- + OH ^- (Equation 23)

^⋅SO₄^- + H₂O₂ → SO₄²⁻ + H⁺ + ^⋅HO₂ (Equation 24)

Although the addition of 100 mM Na₂SO₄ resulted in the highest concentration of H₂O₂, it was considered that there was only a small increase compared with that at 50 mM. In addition, considering the cost of chemicals and the subsequent cost of further desalination, 50 mM Na₂SO₄ was selected for the subsequent test.

Energy consumption and performance comparison

To comprehensively understand the feasibility of the scaled-up MES for the in situ electrosynthesis of H₂O₂, the energy consumption required for the entire process under the selected optimal conditions was further calculated. The energy consumption calculation can be divided into two aspects, one was from the input voltage required by the reactor itself, and another was from cathodic aeration using a pump. For the first aspect, the calculated energy consumption was 0.298 kWh kg⁻¹ H₂O₂ (Supplemental information). Moreover, for the second aspect, the energy consumption by the pump within 42 h was recorded as 0.42 kWh, equal to 99.291 kWh kg⁻¹ H₂O₂. Therefore, the total energy consumption was 99.587 kWh kg⁻¹ H₂O₂, of which approximately 99.70% was consumed by the pump for the cathodic aeration. It was mainly due to extremely low oxygen utilization efficiency (calculated value was 1.43% based on the formula reported in the previous literature) (Yu et al., 2015b). However, low oxygen utilization efficiency was a common phenomenon in which graphite plate was used as electrodes (Yu et al., 2015b). Recently, Zhang et al. reported a superhydrophobic natural air diffusion electrode, which had an oxygen utilization efficiency of up to 65% when used in the synthesis of H₂O₂ (Zhang et al., 2020). This provided a solution for the subsequent application of the scaled-up MES system to efficiently synthesize H₂O₂ and further reduce the energy consumption caused by aeration. In addition, it should be pointed out that the energy consumption for H₂O₂ synthesis based on electrochemical systems or bioelectrochemical systems reported in the literature almost did not take into account the energy consumption including pumping for aeration, and most of the aeration rates adopted in the literature were similar to or slightly higher than that adopted in this study.

Additionally, Table 1 compared the energy consumption of this scaled-up MES with previously published studies. First, although the traditional electrochemical systems (e.g., electro-Fenton) in the laboratory scale using graphite as the cathode have slightly higher H₂O₂ production rate compared with the scaled-up MES in this study, the energy consumption was much higher in the traditional electrochemical systems. Second, compared with laboratory-scale MFC systems that did not require additional voltage input, the scaled-up MES exhibited a faster H₂O₂ production rate with little energy consumption (0.298 kWh kg⁻¹ H₂O₂). Last, compared with the laboratory-scale MEC systems, the scaled-up MES exhibited not only lower energy consumption but also faster H₂O₂ production rate. Besides, Table 1 also showed that the H₂O₂ production rate (88.20 vs 10.82 mg L⁻¹) of some MEC systems was higher than that of this study because modified GDE was used in these studies. The reason can be attributed to the fact that the pumped oxygen can be absorbed by the porous structure of the GDE instead of being dissolved in the catholyte, which overcomes the limitation of low H₂O₂ production usually caused by low oxygen solubility (Yu et al., 2015b). However, using GDE may increase the capital cost in the practical application and may cause water leakage when applied to the scaled-up systems (Wang et al., 2020a). There was only one literature in the form of short communication that reported an MEC system with equivalent cathodic volume as in this study, but the reported H₂O₂ production rate was much lower and the energy consumption was expected to be higher because of the higher aeration rate and input voltage adopted. Finally, it should be pointed out that the cathode potential measured during the whole experiment is higher than −0.86V, which indicated that the corrosion of the graphite electrode can be effectively avoided in this scaled-up system, thereby achieving a long lifespan (Qiao et al., 2018).

Table 1.

Previous studies on H₂O₂ production through (bio)electrochemical systems

System Types	Cathode Materials	Operating Conditions	Reactor Volume	H2O2 Production Rate	Energy Consumption	References
Electro-Fenton	Graphite	pH of 2, 50 mM NaClO4, cathodic aeration rate of 8,747 mL min−1, and applied cathode potential of −0.5 versus SCE	7.65 L (3.15 L of cathode)	40.6 mg L−1 h−1	7.8 kWh kg−1 H2O2	(Qiang et al., 2002)
MEC	Vulcan carbon-coated GDE	pH of 7, 200 mM NaCl, cathodic aeration rate of 20 mL min−1, and applied voltage of 0.31 V	218 mL (18 mL of cathode)	8.8 mg L−1 h−1	1.1 kWh kg−1 H2O2	(Young et al., 2016)
MEC	GDE	pH of 7, 50 mM NaCl, and applied voltage of 0.9 V	18.8 mL (9.4 mL of cathode)	4.2 mg L−1 h−1	1.8 kWh kg−1 H2O2	(Modin and Fukushi, 2012)
MEC	Electrochemically modified graphite particle	pH of 7, 50 mM NA2SO4, and applied voltage of 0.4 V	96 mL (64 mL of cathode)	88.2 mg L−1 h−1	0.66 kWh kg−1 H2O2	(Chen et al., 2015a, 2015b)
MEC	Carbon black and graphite hybrid air cathode	pH of 7, 50 mM NaCl, cathodic aeration rate of 1,500 mL min−1, and applied voltage of 0.6 V	42 mL (14 mL of cathode)	3.3 mg L−1 h−1	56 kWh kg−1 H2O2	(Li et al., 2016)
MEC	Carbon GDE	Tap water, cathodic aeration rate of 2,000 mL min−1, and applied voltage of 1.0–1.6 V (CEM) or 1.6–2.0 V(AEM)	110 L (10 L of cathode)	0.019 mg L−1 h−1 (AEM) and 0.274 mg L−1 h−1 (CEM)	–	(Sim et al., 2018)
MFC	3D graphite cathode	pH of 7 and 50 mM NA2SO4	82 mL (50 mL of cathode)	1.5 mg L−1 h−1	–	(Li et al., 2018a, 2018b)
MFC	Graphite	pH of 7 and 100 mM Na2SO4 and cathodic aeration rate of 191 mL min−1	160 mL (80 mL of cathode)	6.6 mg L−1 h−1	–	(Fu et al., 2010)
MFC	Electrochemically modified graphite/carbon black/active carbon particle	pH of 7 and 50 mM NA2SO4	204.5 mL (120 mL of cathode)	6.8/7.1/8.1 mg L−1 h−1	–	(Chen et al., 2014)
MREC	Graphite	pH of 7, 35 g L−1 NaCl, cathodic aeration rate of 12 mL min−1, and salt and fresh water flow rate of 0.5 mL min−1	140 mL (40 mL of cathode)	10.80 mg L−1 h−1	0.45 kWh kg−1 H2O2	(Li et al., 2017)
20-L scaled-up MES reactor	Graphite	pH of 3, 50 mM NA2SO4, cathodic aeration rate of 400 mL min−1, and applied voltage of 0.6 V	20 L (10 L of cathode)	10.82 mg L−1 h−1	0.298 kWh kg−1 H2O2	In this study

AEM, anion exchange membrane; MREC, microbial reverse-electrodialysis electrolysis cell

Limitation of the study

Although promising, there are still several bottlenecks that need to be further addressed. First, in this study, the electrolyte (e.g., 50 mM Na₂SO₄) was added to increase the system current and therefore the production of H₂O₂, which would lead to an increase in operating cost. Future large-scale applications should be concentrated on high-conductivity wastewater (e.g., textile wastewater) or combined with a membrane technology to treat concentrates (e.g., reverse osmosis). Moreover, the cathodic oxygen utilization rate was low, which was manifested as the cathodic aeration provided by the pump occupying around 99.70% of the total energy consumption. This challenge could be solved in the future through optimization of the reactor configuration and the development of new electrode materials with higher oxygen mass transfer efficiency.

Resource availability

Lead contact

Further information and requests for resources and materials should be directed to and will be fulfilled by the Lead Contact, Yifeng Zhang (yifz@env.dtu.dk, yifzmfc@gmail.com).

Materials availability

This study did not yield new unique reagents.

Data and code availability

There is no dataset or code associated with this work.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Published: February 19, 2021