A Directional Electrode Separator Improves Anodic Biofilm Current Density in a Well-Mixed Single-Chamber Bioelectrochemical System

Abstract

In this study, a directional electrode separator (DES) was designed and incorporated into a single-chamber bioelectrochemical system (BES) to reduce migration and reoxidation of hydrogen (H₂). This issue arises when H₂, generated at the cathode, travels to the anode where anodic biofilms use H₂ as an electron acceptor, resulting in low current density. To test the feasibility of our design, a 3D-printed BES reactor equipped with a DES was inoculated with anaerobic digestor granules and operated under fed-batch conditions using fermented corn stover effluent. The DES equipped reactor achieved significantly higher current densities (~53 A/m²) compared to a conventional single-chamber BES without a separator (~16 A/m²), showing a 3.3 times improvement. Control abiotic electrochemical experiments revealed that the DES exhibited significantly higher proton conductivity (456±127 μS/mm) compared to a proton exchange membrane (67±21 μS/mm) with a statistical significance of P=0.03. The DES also effectively reduced H₂ migration to the anode by 21-fold relative to the control. Overall, incorporating a DES in a single-chamber BES enhanced anodic current density by reducing H₂ reoxidation at the anode.

1. INTRODUCTION

A bioelectrochemical system (BES) is a biological system that leverages the ability of microbes to catalyze electrochemical reactions [1]. In a BES, electrochemically active microbes such as exoelectrogenic Geobacter spp. facilitate a directly exchange electrons with an inert electrode without a need for electron shuttles or electron transfer mediators [2–4]. This unique characteristic of electrochemically active microbes enables us to harvest energy in the form of electricity from a microbial fuel cell (MFC) or hydrogen (H₂) from a microbial electrolysis cell (MEC) [5–10]. The process is directly linked to the microbial attachment to the surface of the electrode (anode) and formation of an electrochemically active biofilm (EAB) [11,12] which transfers electrons derived from the oxidation of organic compounds such as acetate to the anode via the following reaction (Equation 1) [2–4].

$$\mathrm{C}{\mathrm{H}}_3\mathrm{C}\mathrm{O}\mathrm{O}\mathrm{H}+2{\mathrm{H}}_2\mathrm{O}\to 2\mathrm{C}{\mathrm{O}}_2+8{\mathrm{H}}^{+}+{8\mathrm{e}}^{-}$$ (1)

The electrons generated by the EAB reaction at the anode are transferred to the cathode via an external circuit or potentiostat for the reduction reaction. In MFCs, the electrons and the protons on the cathode react with dissolved oxygen to generate H₂O [13]. In MECs, which are operated anaerobically, the cathodic reaction drives the direct reduction of protons resulting in H₂ production (Equation 2) [14–18].

$$2{\mathrm{H}}^{+}+2{\mathrm{e}}^{-}\to {\mathrm{H}}_2$$ (2)

In both systems, anodic current depends on the oxidation of organic compounds by the anodic biofilm. The current density is significantly influenced by the metabolic state and the electrogenic properties of the microbial species present. The microbial biofilms on the anode facilitate the oxidation of organic substrates, and their metabolic state affects the efficiency of electron transfer and overall electroactivity. Therefore, improving anodic current density is imperative for enhancing the overall performance of a BES.

In a BES system, multiple factors can affect the anodic current density: pH, ion transfer, mass transfer of nutrients, biofilm composition, H₂ reoxidation at anode, etc. [19–21]. H₂ reoxidation occurs in a single-chamber BES when H₂ generated at the cathode migrates to the anode and the anodic biofilm uses the H₂ as an electron acceptor instead of donating electrons to the anode [20,22]. A methanogenic community can also use H₂, which reduces H₂ yield [23]. H₂ reoxidation can reduce the anodic current, which reduces the performance of BESs, especially MECs (resulting in a low H₂ yield). To prevent H₂ reoxidation, membranes such as proton exchange membranes (PEMs) and anion exchange membranes (AEMs) have been used in BESs to separate the anodic and cathodic chambers [24,25]. However, membranes employed in BES are susceptible to fouling, which is caused by dissolved organic molecules, precipitation of dissolved inorganic components, bacterial adhesion resulting in biofilm growth, etc. [29–32] and can limit the performance of BESs by decreasing ion transport [26–28]. Fouling affects membrane performance through both chemical and physical modifications that result in decreased ion transport. Physically, fouling agents such as organic molecules, and inorganic compounds accumulate on the membrane surface and within its pores, creating a physical barrier that hinders the passage of ions. This accumulation would increase the membrane resistance and reduce ion flux. Chemically, the fouling process can alter the membrane’s surface properties that can lead to decrease ion transport. For instance, chemical adsorption of organic compounds can lead to a decreased surface area for ion transport [30].Therefore, equipping a BES with a separator which moves H₂ gas away from the anode and thus effectively reduces H₂ reoxidation without decreasing ion transport could be beneficial.

To address membrane fouling and improve BES performance, we developed a directional electrode separator with pores angled at 60° to act as a physical barrier between the anode and cathode. This separator, placed inside a 3D-printed single-chamber BES, allows H₂ gas generated at the cathode to be removed to the BES headspace using N₂ gas sparging from the bottom of the reactor. The nitrogen inlet was positioned directly underneath the cathode to ensure that H₂ is removed before reaching the anode. We hypothesized that a directional electrode separator in a well-mixed single-chamber BES would reduce H₂ reoxidation at the anode while maintaining ion transfer and increasing nutrient mass transfer due to agitation. To test this, we conducted an abiotic electrochemical experiment to measure H₂ levels on both sides of the separator, evaluated its proton conductivity, and determined the optimal electrode placement. The feasibility of the directional electrode separator equipped BES was further demonstrated using native organisms and Geobacter anodireducens SD1 under fed-batch operating conditions with fermentation effluent derived from lignocellulosic feedstock.

2. MATERIALS AND METHODS

2.1. Reactor medium, inoculation, and chemicals

Wastewater samples were collected from an anaerobic digestor at the Moscow Water Reclamation and Reuse Facility (Moscow, ID, USA). Anaerobic granules were obtained from Ingredion Incorporated, Richland, WA, USA. Ferric citrate (CAT#F3388), PIPES (CAT#P1851), sodium acetate (CAT#S2889), sodium DL-lactate solution (CAT#L4263), potassium ferricyanide (CAT#244023), sodium phosphate dibasic dehydrate (CAT#101835432) and potassium hexacyano-ferrate trihydrate (CAT#P3289) were purchased from Sigma-Aldrich. Potassium chloride (CAT#3040-01), sodium phosphate monobasic (CAT#3818-05), potassium phosphate monobasic (CAT#3246-05), and ammonium chloride (CAT#0660-05) were purchased from J.T. Baker. Sodium hydroxide (CAT#7708-10) was purchased from Macron Chemicals. Sodium chloride (CAT#S271-3) was purchased from Fisher Scientific.

Deacetylation and mechanical refining (DMR) effluent was received from the Lawrence Berkeley National Laboratory (LBNL) [33]. DMR was processed from corn stover (Hurley County, South Dakota) at the National Renewable Energy Laboratory (NREL), and processed effluent was used in a fermentation reactor at LBNL [33]. Finally, the effluent from LBNL was used for this study. DMR was autoclaved at 121°C for 15 min/L prior to use and stored in sealed serum bottles with nitrogen headspace after purging with N₂. The organic acid content of DMR broth was quantified using a high-performance liquid chromatography (Table S1). Experimental protocols for Geobacter anodireducens SD1 can be found in the supplementary information.

2.2. Electrochemical analysis of the 3D printed single-chamber BES with a directional electrode separator

2.2.1. Cyclic voltammetry and electrochemical impedance spectroscopy for assessing the placement of the electrode in the 3D printed single-chamber BES with a directional electrode separator

Cyclic voltammetry (CV) was performed on the 3D printed single-chamber BES. Four different configurations of the working and counter electrodes were used to assess the effect of the electrode positions on the directional electrode separator (Supplementary Fig. S1). Ten mM of potassium ferri-/ferrocyanide in 100 mM of KCl were used as the electrolyte. The working and counter electrodes were identical and made of 14.51 cm² of carbon fabric (Panex 30 PW-06, Zoltek Companies, Inc.). A 30 American wire gauge (AWG) titanium wire (TEMCo, part number #RW0517) was sewn through the carbon fabric electrode to serve as a connection wire between the electrode and the potentiostat. The configurations for placing the working and counter electrodes are summarized in Supplementary Table S2. Briefly, in configuration ‘a’ the working and counter electrodes were placed inside the 3D printed single-chamber BES without a directional electrode separator. In configuration ‘b’ the working and counter electrodes were placed on the inside and outside of the directional electrode separator, respectively. In configuration ‘c’ the working and counter electrodes were attached to the inside and outside of the directional electrode separator, respectively, with a monofilament fishing line (Part no. PSM8LB100, South Bend, Calcutta Outdoors, USA). In configuration ‘d’ the working and counter electrodes were attached to the inside and outside of the directional electrode separator, respectively, with a monofilament fishing line. In Configuration ‘d’, the directional electrode separator was cut about 25% of the projected surface area of the electrodes, onto which the electrodes were then placed (Supplementary Fig. S2). A scan rate of ten mV s⁻¹ was used to scan from 0.6 to −0.7 V vs. Ag/AgCl using a benchtop potentiostat (Gamry Interface 1010T, Gamry, USA). Potentiostatic electrochemical impedance spectroscopy (EIS) was performed using a two-electrode system with only ten mM of KCl. This was done by removing the Ag/AgCl reference electrode and connecting the reference electrode probe of the potentiostat directly to the counter electrode. EIS was done at the open circuit potential from 10,000 kHz to 10 mHz using a benchtop potentiostat (Gamry G 300^™, Gamry, USA).

2.2.2. Proton conductivity of the directional electrode separator

An H-shaped cell (Microbial H-cell, Fuel Cell Store, USA) was used for measuring the proton conductivity of the directional electrode separator. A 37-mm-diameter circular disk with a thickness of 2 mm was 3D printed with perforations (2-mm diameter) at a steep 60° angle (same characteristic as the directional electrode separator) using a Formlabs Form 3B+ SLA 3D printer. This printer used autoclavable BioMed Clear resin at a layer thickness of 0.050 mm (Supplementary Fig. S3A). Biomed clear resin is USP Class VI certified material which is suitable for various applications including pharmaceutical and drug delivery. This resin is compatible with common sterilization techniques including autoclave. This 3D printed disk (identical to the directional electrode separator) was used to assess the proton conductivity. One hundred mM of KCl was used as the electrolyte. The dimensions of the working and counter electrodes were the same as described in the section above. A proton exchange membrane (Nafion 117, CAT# 50-109-0478, Fisher Scientific) was also used for performance evaluation. The 3D printed directional electrode separator, or the proton exchange membrane (PEM) was placed in the connection channel between the working and counter electrode chambers (Supplementary Fig. S3B). The channel was kept open to measure the background. EIS was done at the open circuit potential from 10,000 kHz to 10 mHz using a benchtop potentiostat (Gamry G 300^™, Gamry, USA). The proton conductivity was calculated using the following equation (Equation 3) [34]:

$$\mathrm{\sigma }=\mathrm{L}/(\mathrm{R}\mathrm{s}\times \mathrm{A})$$ (3)

where $\sigma$ is the proton conductivity (μS mm⁻¹), L is the thickness (0.17 mm for PEM and 2 mm for the 3D directional electrode separator disk), A is the exposed surface area (572.55 mm² for PEM and the equivalent of a sum of small circles, ~125.80 mm², for the 3D printed directional electrode separator), and Rs is the uncompensated resistance, which can be obtained from the model fitting the EIS data.

2.2.3. The migration of H₂ from the cathode side to the anode side in the 3D printed single-chamber BES with and without the directional electrode separator

The prevention of H₂ migration from the cathode side to the anode side in a 3D printed single-chamber BES with and without a directional electrode separator was assessed by measuring H₂ concentration electrochemically using a 2 mm diameter Platinum (Pt) electrode. Briefly, the directional electrode separator equipped with carbon fabric electrodes was placed within the 3D printed single-chamber BES. The electrode that was placed on the outside of the directional electrode separator served as the working electrode (Supplementary Fig. S4). The working electrode was polarized at −1.3 V vs. Ag/AgCl using a benchtop potentiostat (Gamry Interface 1010T, Gamry, USA) to generate H₂ gas on the electrode surface. H₂ was measured using the Pt electrode by polarizing the Pt electrode to 0.6 V vs. Ag/AgCl using a benchtop potentiostat (Gamry Reference 600^™, Gamry, USA). 10× Phosphate-buffered saline (PBS, 10×) was used as the electrolyte. The Pt electrode was calibrated using pure nitrogen gas and 4% hydrogen gas with 96% nitrogen gas (two-point calibration). Three different agitation speeds and two different nitrogen gas flow rates were used. The Pt electrode was placed in two different locations within the reactor to measure in situ H₂: 1) outside the directional electrode separator and 2) inside the directional electrode separator (Supplementary Fig. S4). The prevention of H₂ migration was reported as the percentage of the H₂ migration going to the anode; this is the H₂ measured inside the directional electrode separator (anodic side of the BES) relative to the total H₂ measured (sum of both measurements). As a control, A 3D printed single-chamber BES without a directional electrode separator was used.

2.3. Single-chamber BES without a separator

The single-chamber BES without a separator is a three-electrode system. The working and counter electrodes are identical and made of 14.51 cm² of carbon fabric (Panex 30 PW-06, Zoltek Companies, Inc.) (Supplementary Fig. S5A). A thin line of silicone adhesive was applied to the edges of the electrode to keep the electrode from tearing. Four 8/32 nuts, #8 washers, and an 8/32 0.75” bolt were used to secure 0.025” diameter titanium wire (Malin Co., USA) against the carbon fabric, as shown in Supplementary Fig. S5A. The electrode was bent to a half-circle shape in order to place two electrodes (anode and cathode) inside the reactor without them touching each other (Supplementary Fig. S5A). A custom-made silver/silver-chloride (Ag/AgCl in saturated KCl) electrode was used as a reference electrode [35]. A single compartment of the glass H-cell configuration was used for the reactor. A rubber separator was used to block the connection of the two chambers so that each chamber could be used as a single-chamber reactor (Supplementary Fig. S5B). Rubber tubing was placed inside the liquid in the reactor to bubble nitrogen gas to maintain an anaerobic condition. Supplementary Fig.S6 shows a schematic of the single-chamber BES without a separator used in the experiment.

2.4. Design of 3D printed single-chamber BES with directional electrode separator

A cylindrically shaped 3D printed single-chamber BES was printed with a FormsLabs Form 3B+SLA 3D printer using autoclavable BioMed Clear resin at a layer thickness of 0.050 mm. The 3D printed single-chamber BES has an N₂ gas inlet at the bottom, enabling N₂ gas to flow along the outer edge of the reactor (Fig. 1A). The total working volume of the reactor was 200 mL, and it had a water jacket to control the temperature, which was kept at 30°C. The directional electrode separator was also 3D printed with BioMed Clear resin. The directional electrode separator featured a cylindrical shape with a thickness of 2 mm, a length of 84 mm, and an inner diameter of 39.4 mm. The directional electrode separator had perforations (2 mm in diameter) at a steep 60° angle and were directed downward and toward the center of the reactor (Fig. 1B). The pores of the directional electrode separator are circular in shape and angled at 60° inward. The pore angle determines how effectively the pore aligns with the flow of ions and gas through the separator. A more acute pore angle can decrease the H₂ transport. Additionally, the size of the pore may increase the surface area of ions transport but will also increase the H₂ transport which may lead to reduced current density. Moreover, a larger pore size may also increase the chance of short-circuiting the electrochemical system. The directional electrode separator acted as a physical separator between the working and counter electrodes. Drawing of 3D reactors and relevant data is available from the authors if requested.

Figure 1.

A) Schematic of a 3D printed single-chamber BES equipped with a directional electrode separator. B) Schematic of a directional electrode separator.

2.5. Construction of the 3D printed single-chamber BES with directional electrode separator

The working and counter electrodes of the 3D printed single-chamber BES with a directional electrode separator were identical and made of 14.51 cm² of carbon fabric (Panex 30 PW-06, Zoltek Companies, Inc.). A thin line of silicone adhesive was applied to the edges of the electrode to keep the electrode from tearing. The working electrode was placed on the inside of the directional electrode separator, and the counter electrode was placed on the outside of the directional electrode separator (Fig. 1A). The electrodes were attached to the directional electrode separator with a monofilament fishing line (Part no. PSM8LB100, South Bend, Calcutta Outdoors, USA). Prior to attaching the electrodes to the directional electrode separator, A 30 American wire gauge (AWG) titanium wire (TEMCo, part number #RW0517) was sewn through the carbon fabric electrode to serve as a connection wire between the electrode and the potentiostat. A custom-made Ag/AgCl electrode (Ag/AgCl in saturated KCl) was used as a reference electrode.

2.6. Inoculation and refreshment of the medium of the BES

2.6.1. Single-chamber BES without a separator inoculated with homogenized AD granules

A 125-ml single-chamber BES without a separator was inoculated with homogenized AD granules (10% of the total reactor volume—Ingredion Incorporated, Richland, WA, USA) and enriched with wastewater from an anaerobic digestor (Moscow Water Reclamation and Reuse Facility, Idaho, USA). In a control experiment, a single-chamber BES was inoculated with homogenized AD granules (Ingredion Incorporated, Richland, WA, USA) and enriched solely with sterilized DMR. When needed, the medium was changed by taking out some of the existing medium (90%) with a 60-mL syringe and needle. The syringe and needle were purged with N₂ before the medium was changed. The working electrode was polarized at 0.3 V vs. Ag/AgCl with a benchtop potentiostat (Reference 600^™, Gamry, USA) connected to a multiplexer (ECM8 electrochemical multiplexer, Gamry, USA). The polarization was stopped during the transfer of the medium. Ferric citrate (5 mM) was used as an iron supplement when the medium was refreshed with DMR broth.

2.6.2. A 3D printed single-chamber BES inoculated with homogenized AD granules

A 3D printed single-chamber BES was inoculated with homogenized AD granules (10% of the total reactor volume) and enriched three times with wastewater from an anaerobic digestor. Afterwards, the medium was switched to DMR. The total working volume of the reactor was 200 mL. The temperature of the 3D printed single-chamber BES was held at 30°C by passing temperature-controlled water from a recirculation system (Neslab Merlin M33). A continuous flow of 2 L min⁻¹ of nitrogen was employed to maintain anerobic conditions, and this nitrogen was introduced from the bottom of the reactor. The working electrode of the BES was polarized at 0.3 V vs. Ag/AgCl with a benchtop potentiostat (Squidstat Prime, Admiral Instruments, USA). The polarization was stopped during the refreshment of the medium. Ferric citrate (5 mM) was used as an iron supplement when the medium was refreshed with DMR broth.

2.7. Statistics

EIS data are presented as mean ± standard deviation of mean of three replicates. The H₂ migration data are presented as mean ± standard deviation of mean of ten data points. The mass transfer coefficient data are presented as mean ± standard deviation of mean of five replicates. A student’s t-test was performed to compare the proton conductivity of the 3D printed directional electrode separator and a PEM. MATLAB^® (version R2021b) was used for performing statistical analysis and generating graphs.

3. RESULTS AND DISCUSSION

3.1. Electrodes placed on the directional electrode separator provide optimum conditions for the 3D printed single-chamber BES

We first evaluated the optimum placement of the electrode on the directional electrode separator. In a controlled abiotic environment, we placed electrodes in various locations within the 3D printed BES with a directional electrode separator (Supplementary Fig. S1). We conducted cyclic voltammetry (CV) on four different electrode placement configurations. The electrodes, along with the directional electrode separator, were inserted into the 3D printed single-chamber BES reactor. A 10 mM potassium ferri-/ferrocyanide redox couple (K₃[Fe(CN)₆]/K₄[Fe(CN)₆]) with 100 mM KCl was used as the electrolyte. Among the configurations that were tested, the order of oxidation peak current density was as follows: configuration ‘d’ exhibited the highest at 24.52 A m⁻², followed by ‘c’ at 21.15 A m⁻², ‘a’ at 13.43 A m⁻², and ‘b’ at 11.98 A m⁻² (Supplementary Fig. S7). The oxidation current density suggests an oxidation electrochemical reaction (ferrocyanide to ferricyanide) at the anode. Because the oxidation and reduction reactions of potassium ferri-/ferrocyanide are interconnected, an elevation in both the oxidation and reduction peaks was observed when the electrodes were placed in close proximity, as seen in configurations ‘c’ and ‘d’ (Supplementary Fig. S7). This observation suggests that the protons generated from the biological oxidation reaction in the BES (Equation 1) could efficiently transfer to the cathode in configurations ‘c’ and ‘d’ [36]. However, we employed configuration ‘d’ to investigate the impact of the directional electrode separator. Since configuration ‘d’ compromises the integrity of the directional electrode separator, making this configuration unsuitable for use in the BES, we opted for configuration ‘c’ to position the electrodes on the directional electrode separator within the 3D printed single-chamber BES equipped with a directional electrode separator.

We also conducted EIS to assess the resistance associated with the directional electrode separator for these four different configurations. Supplementary Fig. S8 shows the Nyquist plots for these configurations of the electrode placements. An equivalent circuit consisting of uncompensated resistance (Ru), parallel resistance (Rp) and a constant phase element (Q1) was used to fit the data to the model (Supplementary Fig. S9). Supplementary Table S3 summarizes all values obtained from this process. The value of n, which is an exponent, indicates the capacitive or resistive behavior of the element. For example, n=1 indicates pure capacitive behavior, and n=0 indicates pure resistive behavior of the constant phase element. The uncompensated resistance (Ru) indicates the resistance between the working and counter electrodes. This includes solution resistance, electrode resistance and resistance due to the directional electrode separator. The Ru value was found to be 14.91 ± 1.15 Ω, 17.02 ± 2.85 Ω, 12.15 ± 1.05 Ω and 8.36 ± 0.27 Ω for configurations ‘a’, ‘b’, ‘c’ and ‘d’, respectively (Supplementary Table S3). This indicates that the total uncompensated resistance decreased as the distance between the electrodes (anode and cathode) was reduced. Additionally, the n values show that the constant phase element behaved like a capacitor (Supplementary Table S3) where Rp indicates the charge transfer resistance. This observation suggests that all the electrode configurations have similar charge transfer resistance (Supplementary Table S3). Considering its combination of a low total uncompensated resistance and a high oxidation-reduction peak current density (21.15 A m⁻² for oxidation and −0.02 A m⁻² for reduction) (Supplementary Fig. S7) in a configuration that does not compromise the integrity of the directional electrode separator, configuration ‘c’ was selected for use in the biological experiments.

3.2. The directional electrode separator provides significantly higher proton conductivity than a proton exchange membrane

EIS was used to measure the proton conductivity of the directional electrode separator. A 3D printed directional electrode separator was used for this purpose, and a PEM (Nafion 117) was used as a control to compare the proton conductivity as it is commonly used in BESs. One hundred mM KCl was used as electrolyte. Supplementary Fig. S10 shows the Nyquist plots of the three different conditions that were tested. An equivalent circuit consisting of uncompensated resistance (Ru), parallel resistance (Rp) and a constant phase element (Q1) was used to fit the data to the model (Supplementary Fig. S9). Table 1 summarizes all the values obtained from the model. The proton conductivities of the 3D printed directional electrode separator and the PEM were found to be at 456 ± 127 and 67 ± 21 μS mm⁻¹, respectively, using equation 3. This suggests that the 3D printed directional electrode separator has significantly higher proton conductivity (P = 0.03), approximately seven times higher, than a PEM.

Table 1.

Uncompensated resistance (Ru), parallel resistance (Rp), constant phase element (Q1) and exponent (n) for each condition tested for proton conductivity.

Configuration	Ru (ohm)	Rp (ohm)	Q1(S*sn)	n
Directional electrode separator	137.93 ± 10.36	6.1×109 ± 7.49×108	9.61×10−4 ± 2.03×10−4	0.98 ± 0.00
PEM	106.77 ± 1.00	6.62×108 ± 9.72×108	9.11×10−4 ± 2.23×10−4	0.97 ± 0.00
Open channel for background	101.77 ± 0.95	1.83×109 ± 1.81	1.59×10−3 ± 3.43×10−4	0.98 ± 0.00

3.3. The directional electrode separator reduces H₂ migration to the anode in a 3D printed well-mixed single-chamber BES

We evaluated H₂ migration in a 3D printed well-mixed single-chamber BES by measuring in situ H₂ concentrations at various flow rates and agitation speeds. When a directional electrode separator was incorporated into the 3D printed well-mixed single-chamber BES, at a 2-L min⁻¹ nitrogen sparging rate, H₂ migration to the anode was about 12.57% ± 1.96%, 15.25% ± 1.53%, and 4.14% ± 2.02% for rotational Reynolds numbers (R.R.N.) of 0, 3268 and 6765, respectively (Fig. 2). However, when we removed the directional electrode separator, we found an increase of H₂ migration to the anode. At a 2-L min⁻¹ nitrogen sparging rate, H₂ migration to the anode measured about 32.82% ± 8.41%, 34.14% ± 9.44%, and 30.85% ± 11.53% for R.R.N. of 0, 3268 and 6765 R.R.N, respectively (Fig. 2). We also tested for a lower (1-L min⁻¹) nitrogen sparging rate; when the directional electrode separator was used, we found that H₂ migration to the anode was about 14.58% ± 1.77%, 9.36% ± 3.96%, and 11.65% ± 1.13% for R.R.N. of 0, 3268 and 6765, respectively (Supplementary Fig. S11). When we removed the directional electrode separator, about 28.62% ± 3.15%, 28.79% ± 5.16%, and 38.62% ± 6.41% H₂ migration to anode was measured for 0, 3268 and 6765 R.R.N., respectively (Supplementary Fig. S11). A high deviation (approximate average of ± 7.35 when the directional electrode separator was not incorporated, compared to an average of ± 2.06 when it was) was observed when measurements were conducted without the directional electrode separator. This could be due to heterogeneous nitrogen sparging throughout the system, as the reactor was not designed to operate without the directional electrode separator. It is expected that bubble sizes, bubbling efficiency, viscosity, and the distance between could play a role in anodic current and reactor efficiency and such factors can be a topic for another study. This result indicates that using a directional electrode separator is beneficial for reducing H₂ migration to the anode in 3D printed well-mixed single-chamber BESs. A directional electrode separator reduces H₂ migration to the anode by approximately 21-fold in a 3D printed well-mixed single-chamber BES compared to operating without the directional electrode separator.

Figure 2.

H₂ migration to the anode decreased when a directional electrode separator was equipped, compared to without incorporating the directional electrode separator in the 3D printed well-mixed single-chamber BES at 2 L min⁻¹ nitrogen sparging rate. Agitation speeds were reported as rotational Reynolds numbers (R.R.N.). The data are represented as the means and standard deviations of ten measurements.

3.4. Enrichment with wastewater provides higher anodic current density than enrichment with DMR alone in a single-chamber BES without a separator inoculated with homogenized AD granules

Prior to testing the 3D printed single-chamber BES, we evaluated the effects of initial biofilm enrichment with wastewater on current density in a single-chamber BES without a separator inoculated with homogenized AD granules. A control system was operated with DMR alone. For comparative purposes, we tested the system with an axenic strain (G. anodireducens SD1) (Supplementary Fig. S12). While G. anodireducens SD1 showed excellent performance in a BES and was able to use lactate, formate and acetate, it exhibited low current density (1.4 A m⁻²) with DMR due to its weak tolerance of succinate and intolerance of ethanol and butanol (which are found in DMR, Supplementary Table S1) [37,38]. When homogenized AD granules were used as the inoculum, a notably higher current density was observed (Fig. 3). Thus, homogenized AD granules were selected as the inoculum in all subsequent biological experiments. The single-chamber BES without a separator was initially enriched with wastewater for a duration of ten days (Fig. 3 lines A and B). Subsequently, the medium was refreshed with DMR. Following the medium refreshment, there was a notable increase in the current density (Fig. 3 lines A and B). For the control, a single-chamber BES without a separator was enriched solely with DMR and the medium was refreshed on the 22^nd, 55^th, and 65^th day (Fig. 3 lines C and D). An additional medium refreshment was conducted on the 35^th day (Fig. 3, line D). Fig. 3 shows that enriching anodic biofilm with wastewater followed by switching to DMR in a single-chamber BES without a separator resulted in a higher current density, reaching maxima of 14 A m⁻² and 16 A m⁻² (as shown in Fig. 3 lines A and B), than enriching solely with DMR broth, which resulted in maxima of 6 A m⁻² and 3 A m⁻² (as shown in Fig. 3 lines C and D). In these reactors (Fig. 3), we utilized homogenized AD granules to develop a mixed culture anodic biofilm. This mixed culture anodic biofilm exhibited both a higher current density and a robust nature as reported in the literature [39,40]. The single-chamber BESs without a separator that were initially enriched with wastewater showed ~ 4-fold higher current densities than the single-chamber BESs without a separator enriched solely with DMR.

Figure 3.

Performance of a single-chamber BES without a separator inoculated with homogenized AD granules: Lines A&B) Initial enrichment with wastewater, followed by a switch to DMR. The highest peak current densities were reached after the switch to DMR medium, 14 A m⁻² and 16 A m⁻² (replicates). Lines C&D) Enrichment solely with DMR broth. The highest peak current densities reached were 6 A m⁻² and 3 A m⁻² (replicates). BESs enriched with wastewater initially showed ~ 4-fold higher current densities than BESs enriched with DMR.

3.5. A well-mixed 3D printed single-chamber BES equipped with a directional electrode separator inoculated with homogenized AD granules showed sustainable high current density with three rounds of wastewater enrichment

Finally, we assessed the performance of the single-chamber BES equipped with a directional electrode separator on DMR. We found that high agitation improved the mass transfer but that the mass transfer decreased slightly when a directional electrode separator was incorporated (Supplementary Fig. S13). Moreover, the sequence of initially enriching with wastewater followed by switching to DMR resulted in a higher current density in the BES (Fig. 3). Therefore, we enriched the homogenized AD granules three successive times with wastewater before switching to DMR. The peak current densities recorded during the wastewater enrichment phase were 0.75 A m⁻² (as shown in Fig. 4A), 7.33 A m⁻² (Fig. 4B), and 2.88 A m⁻² (Fig. 4C). After the wastewater enrichment phase, the medium was renewed (on day 18) with DMR. Subsequent to successive medium refreshments with DMR, the current density reached peaks of 37.75 A m⁻² (Fig. 4A), 55.86 A m⁻² (Fig. 4B), and 65.22 A m⁻² (Fig. 4C). Fig. 4 shows that a sustainable current density was achieved with a well-mixed 3D printed single-chamber BES with a directional electrode separator for approximately 8 days. The reduction in current density during this phase signifies the exhaustion of nutrients, and following medium refreshment, there was an increase in current density. However, there was a substantial variation in current density across the BESs that may be attributable to the diverse community of AD granules [41]. This 3D printed well-mixed single-chamber BES equipped with a directional electrode separator can be effectively employed for achieving a sustained, long-term (over a week) high (minimum of ~30 A m⁻² and maximum of ~60 A m⁻²) current density with homogenized AD granules.

Figure 4.

Performance of the well-mixed 3D printed single-chamber BES with AD granules on DMR, following an initial three rounds of initial wastewater enrichment. Peak current densities of (A) 37.75 A m⁻², (B) 55.86 A m⁻², and (C) 65.22 A m⁻² were recorded after the medium was refreshed with DMR.

A single-chamber BES equipped with a directional electrode separator holds a great potential for implementation as an MEC for producing substantial volumes of H₂. The core characteristic of the directional electrode separator, reducing the H₂ migration from the cathode to the anode (Fig. 2), reduces the H₂ concentration at the anode, ultimately reducing the H₂ reoxidation and increasing the current density. Protons produced by the anodic biofilm from the EAB-induced oxidation reaction of the substrate (acetate, glucose, etc.) can easily migrate to the cathode (Fig. 5). The hydrogen evolution reaction occurs at the cathode, and H₂ bubbles formed on the counter electrode rise to the headspace along with the N₂ gas (Fig. 5).

Figure 5.

Directional electrode separator in a well-mixed single-chamber BES, where it acts as a physical electrode separator and reduces H₂ gas migration from the cathode to the anode. Protons generated from the working electrode can transfer to the counter electrode easily.

Furthermore, the high mixing rate of this 3D printed well-mixed BES contributes to a higher mass transfer, addressing the local nutrient deprivation at the anode. We have shown that, with all these conditions combined, a 3D printed well-mixed single-chamber BES equipped with a directional electrode separator can achieve a sustainable high current density (minimum of ~30 A m⁻² and maximum of ~60 A m⁻²) for a long term. This technology offers an opportunity to enhance the efficiency of H₂ production by implementing it in an MEC.

4. CONCLUSIONS

The directional electrode separator effectively reduces the H₂ migration from the cathode to the anode (~21-fold) and enhances proton conductivity (~7-fold) compared to a PEM. Additionally, placing an electrode onto the directional electrode separator showed reduced internal resistance. In biological experiments in a single-chamber BES without a separator inoculated with homogenized AD granules, initial wastewater enrichment showed higher current density (~4-fold) than enrichment with DMR alone. Finally, the directional electrode separator in a 3D printed well-mixed single-chamber BES inoculated with homogenized AD granules and enriched 3 times with wastewater before being switched to DMR provided a sustainable high current density (minimum of ~30 Am⁻² and maximum of ~60 Am⁻²) for 8 days.

Data availability statement:

Data are available from the corresponding author upon request.