Heat-Treated Stainless Steel Felt as a New Cathode Material in a Methane-Producing Bioelectrochemical System.

Methane-producing bioelectrochemical systems (BESs) are a promising technology to convert renewable surplus electricity into the form of storable methane. One of the key challenges for this technology is the search for suitable cathode materials with improved biocompatibility and low cost. Here, we study heat-treated stainless steel felt (HSSF) for its performance as biocathode. The HSSF had superior electrocatalytic properties for hydrogen evolution compared to untreated stainless steel felt (SSF) and graphite felt (GF), leading to a faster start-up of the biocathodes. At cathode potentials of -1.3 and -1.1 V, the methane production rates for HSSF biocathodes were higher than the SSF, while its performance was similar to GF biocathodes at -1.1 V and lower than GF at -1.3 V. The HSSF biocathodes had a current-to-methane efficiency of 60.8% and energy efficiency of 21.9% at -1.3 V. HSSF is an alternative cathode material with similar performance compared to graphite felt, suited for application in methane-producing BESs.

Introduction

Renewable energy plays an important role in addressing the global energy crisis and environmental pollution, as it can reduce the demand for fossil fuels, and thus counteract the global warming caused by greenhouse gas (GHG) emissions.[1] Renewable energy supplies, such as wind and solar power, are fluctuating and intermittent.[2] Therefore, energy storage systems for integrating renewable energy into a balancing energy grid are essential.[2] Power-to-gas is an emerging renewable energy storage technology which can convert electrical energy into gas fuel (H2 or CH4).[3,4] This technology balances the grid with high flexibility and stability by creating a connection of electrical and gas networks.[5] A potentially convenient power-to-gas technology is methane-producing bioelectrochemical systems (BESs) which convert electricity and CO2 into methane in one processing step.[6,7] In methane-producing BESs, CO2 serves as the sole carbon source at the biocathode and is reduced to CH4 via direct and/or indirect pathways (via H2) by microorganisms.[8,9]

Hydrogenotrophic methanogens that utilize hydrogen for their growth, such as Methanobacterium and Methanobrevibacter, are known to dominate the cathode electrode.[10−13] To drive CH4 production at reasonable rates, a cathode potential that is more negative than the theoretical hydrogen evolution potential (−0.61 V versus Ag/AgCl, at pH 7, 1 M solute concentration[14]) is usually applied. The additional cathode potential, i.e., the overpotential, reflects the extra energy investment at the cathode to drive the reaction. With the introduction of cathode materials that catalyze hydrogen evolution in methane-producing BESs, methane production could be enhanced, possibly at lower energy input.

Metal-based electrodes are known to catalyze the hydrogen evolution reaction (HER),[15] thereby stimulating methane production rate via hydrogenotrophic methanogenesis. Siegert et al. (2014) showed that a platinum electrode resulted in the highest biotic methane production rate (250 ± 90 nmol cm–3 d–1; d = day), and also abiotic hydrogen production rate (1600 ± 200 nmol cm–3 d–1), compared with several different carbon-based and metal-based cathode materials.[16] However, platinum is an expensive material, which is a foreseeable practical barrier for implementing methane-producing BESs.[16] Inexpensive metal-based electrodes, e.g., stainless steel, are more cost-effective and applicable in methane-producing BESs.

Stainless steel has been found to be a good cathode material for hydrogen-producing biocathodes, with a performance comparable to that of platinum.[17] In addition, the high durability and low cost of stainless steel compared with graphite felt are desirable in practical applications.[18] Recently, several strategies have been applied to enhance the performance of stainless steel as an electrode, e.g., surface modification.[17] Recently, Guo et al. (2015) demonstrated a simple and economical way to obtain 3D nanostructure stainless steel felt by application of heat treatment.[19] The presence of 3D iron oxide nanoparticles increased the biocompatibility of stainless steel materials, which resulted in several-fold enhancement in current density (up to about 1.5 ± 0.13 mA/cm2) for bioanodes. These iron minerals can potentially enhance methane evolution by facilitating electron transfer between electrode and methanogens.[20] Heat-treated stainless steel felt may, therefore, be an attractive cathode material for methane-producing BESs. To our best knowledge, although stainless steel felt has been widely used as a cathode material in methane-producing BESs, there are no studies investigating the potential of heat-treated stainless steel felt as a cathode material in methane-producing BESs.

In this study, we examined the use of heat-treated stainless steel felt (HSSF) in a methane-producing BES and compared the performance of the HSSF with untreated stainless steel felt (SSF) and graphite felt (GF). The performance was investigated in terms of CH4 production rate, current-to-methane efficiency, current-to-hydrogen efficiency, and energy efficiency at different cathode potentials of −1.3, −1.1, and −0.8 V versus Ag/AgCl. Polarization curves were used to determine the catalytic activity of different cathode electrode materials.

Materials and Methods

Electrode Preparation

Three materials were tested as cathodes: heat-treated stainless steel felt (HSSF), stainless steel felt (SSF), and graphite felt (GF). Samples of 0.28 cm thick GF (CGT Carbon GmbH, Germany) and 0.1 cm thick 316 L SSF (Lier Filter Ltd., China) were cut into a circle with a diameter of 5 cm. The projected surface area of each electrode was 20 cm2. To obtain heat-treated stainless steel felt (HSSF), SSFs were treated in the same way as described by Guo et al.: placing SSFs into a muffle furnace at 600 °C for 5 min, then taking SSFs out of the muffle furnace, and finally cooling them down under air to ambient temperature overnight.[21] Platinum foil (5 cm length × 2.5 cm width) with a projected surface area of 12.5 cm2 was used as anode material for each reactor. Titanium wires (0.1 cm in diameter) served as current collector for both the anode and the cathode electrode.

Reactor Setup

Each reactor system consisted of 3 chambers, one anodic chamber in the middle facing two cathodic chambers (Figure ). Each chamber had a cylindrical volume of 25 mL (5 cm diameter × 1.26 cm thickness), and the three chambers were separated by two Nafion 117 cation exchange membranes (Sigma-Aldrich, Sigma, St. Louis, MO) pretreated by boiling in H2O2 (30%), deionized (DI) water, 0.5 M H2SO4, and DI (deionized) water, each solution for 1 h at 80 °C.[22] In each reactor, two separate platinum foils were inserted in the middle anodic chamber to serve as anodes; the same cathode electrode material was used in these two cathodic chambers in order to perform duplicate testing. Each cathode chamber contained a Ag/AgCl reference electrode (3 M KCl, ProSenseQiS, Netherlands). All potentials were measured and reported against the Ag/AgCl reference electrode.

Figure 1

Schematic overview of the methane-producing bioelectrochemical reactor. Each cathode material was tested in duplicate and was connected to one of the anodes in the anode chamber. Gas production was collected in the headspace of the recirculation bottle in combination with the 500 mL of the gas bag.

The catholyte of each cathode chamber was connected via a circulation bottle (total catholyte volume of 500 mL) at a pump speed of 1.0 mL/min. The pH was measured daily and controlled in the range of 7.1 and 7.6 manually. Gas bags (500 mL, Cali-5-Bond, Calibrated Instruments INC) were connected to the headspace (25 mL) of the recirculation bottle. All anodic chambers shared the same anolyte (in total 5 L) that was recirculated at a pump speed of 4.0 mL/min.

Inoculum and Electrolytes

Each cathode chamber was inoculated with 50 mL of anaerobic sludge with a volatile suspended solids (VSS) concentration of 5.7 g/L, from the wastewater treatment plant in Ede, The Netherlands. The procedure of VSS measurement was according to Standard Method 2540-E.[23] The catholyte contained 0.2 g/L NH4Cl, 0.13 g/L KCl, 1 mL/L vitamin and 1 mL/L mineral solution,[24] and 50 mM phosphate buffer solution (4.58 g/L Na2HPO4 and 2.77 g/L NaH2PO4·2H2O). The catholyte was flushed with N2 gas for 30 min before feeding to all cathode chambers of all cells, and afterward, 5 g/L NaHCO3 was added to the catholyte as a carbon source. The anolyte contained the same phosphate buffer (50 mM) as the catholyte solution. The anolyte was flushed with N2 continuously to minimize dissolved oxygen diffusion across the membrane from the anodic to the cathodic chamber.

Reactor Operation

The cell voltage of each reactor was controlled by an FP-AO-210 module (National Instruments Field Point system, Austin, TX). The applied voltage was controlled and adapted to reach a certain cathode potential. The current was measured by the voltage difference over a 10 Ω resistor in the electrical circuit between the anode (counter electrode) and the power source. The cathode potentials (measured versus the Ag/AgCl reference electrode) and voltages over the resistor were recorded every minute using LabVIEW, supported by an FP-AI-110 module (National Instruments Field Point system, Austin, TX).

For a start-up, all biocathodes were controlled at −0.9 ± 0.03 V for 3 weeks to allow methanogenic growth, as −0.9 V is a typical cathode potential for a methane-producing biocathode.[11,12] After the start-up phase, each cathode was controlled for 4 weeks first at −1.1 ± 0.03 V, afterward at −0.8 ± 0.03 V, and finally at −1.3 ± 0.03 V. When the cathode potential deviated more than 30 mV from the desired cathode potential, the cell voltage was adjusted to reach the desired cathode potential. Each reactor was operated in batches with a length of 168 h. Four batches were performed at each cathode potential to achieve stable performance (at least two similar batches). Each batch was started by replacing half of the catholyte with fresh medium to ensure sufficient HCO3–, nutrients, and buffer.

Gas Analysis

Two or four measurements were done per batch to analyze the methane production rates. The gas volume was determined by emptying the gas bags with a syringe. A gas sample of each cathode chamber was taken from the headspace through the butyl rubber stopper. The gas composition in the headspace was identical to that in the gas bag because the headspace was connected with the gas bag. The gas composition produced at the cathode by two types of gas chromatography: the HP 5890A gas chromatograph and the Finsons Instruments GC 8340 gas chromatograph. The HP 5890A gas chromatograph measured H2 by injecting 100 μL of a gas sample into a molecular sieve column with thermal conductivity detection (TCD); the Finsons Instruments GC 8340 gas chromatograph measured CH4, CO2, O2, and N2 by injecting 50 μL of a gas sample into a molecular sieve column with TCD 90 °C.

The cumulative methane yield was calculated as follows:Here, VCH was the cumulative methane production (mL) at sampling time t; VT, was the total gas production collected in the gas bag (mL). Vhs is the headspace volume (mL); CCH represented the methane concentration (%).

The methane production rate was calculated over the entire batch and normalized to the cathode electrode projected surface area:Here, γCH4–A represents methane production rate (L CH4/m2 catproj/d); VCH4,total is the total amount of methane yield over the entire batch (L). Aproj is the projected surface area of the cathode electrode (20 cm2); t is the experimental time between each sample (day).

VFA Analysis

The volatile fatty acids (VFA), herein including formate, acetate, and lactate, were measured by high-performance liquid chromatography (HPLC).[25] Each liquid sample was centrifuged for 10 min at 10 000 RCF, and then 1 mL of the supernatant was directly put into the sample vial. Samples were injected by an autosampler and separated with an Alltech OA-1000 column at 60 °C and 6.0–6.5 MPa.

Dissolved CH4 Analysis

The dissolved CH4 was measured by the following procedures:[26] (1) adding 5.3 g of NaCl into a 50 mL tube sealed by a stopper, (2) extracting 20 mL of air from the tube using a syringe with a needle, (3) slowly injecting 15 mL of the catholyte into the tube, (4) shaking the tube to fully mix the salt and catholyte, (5) waiting for 30 min to make sure CH4 is removed to the gas phase, (6) measuring the pressure by the pressure meter (GMH 3150, Germany), (7) measuring gas composition by the gas chromatography (Finsons Instruments GC 8340), and (8) calculating the amount of dissolved CH4 by the following formula.Here, nCH4–dissolved represents the moles of dissolved methane; P is the pressure of headspace in the sample tube (kPa). C is the methane percentage (%) in the headspace of the tube; V is headspace volume in the tube (0.035 L). R is the gas constant value (8.314 J mol–1 K–1); T is 293 K.

Cathodic Efficiency (ηCE)

The efficiency of capturing electrons from the electric current in products is the sum of current-to-methane efficiency (ηCH4) and current-to-hydrogen efficiency (ηH2). The efficiency of capturing the electrons from the electric current in methane or hydrogen was calculated via:Here, the following abbreviations apply: NCH4 is the total moles of methane produced; NH2 is the total moles of hydrogen produced; F is the Faraday constant (96 485 C/mol e–); I is the current (A), and t is the time (s).

Voltage Efficiency (ηvoltage)

is described as the part of the applied cell voltage that ends up in methane, and was calculated asHere, ΔGCH is the Gibb’s free energy of methane oxidation (−890.4 kJ/mol CH4),[6]Ecell is the applied cell voltage (V), and F is the Faraday constant (96 485 C/mol e–).

Energy Efficiency (ηenergy)

Energy efficiency was calculated by taking the product of ηCH and ηvoltage, which represents the part of the external electrical energy that ends up in methane.[6]

Electrochemical Analysis

Polarization tests were performed every 2 weeks using a potentiostat (Ivium Technologies, Eindhoven, The Netherlands). For the polarization test, the cathode potential was decreased from −0.7 to −1.1 V with steps of 0.1 V. Each step lasted for 10 min while the catalytic current was recorded and plotted according to the literature.[8]

Scanning Electron Microscopy

Surface morphology of the biofilms on different cathode materials was analyzed by a scanning electron microscope (SEM, FEI Magellan 400). All samples were pretreated in the same method according to the standard procedure.[27] Processing of SEM images was performed at Wageningen Electron Microscopy Center (WemC, The Netherlands).

Results

Methane Yields and Methane Production Rates

To achieve stable performance, each biocathode underwent four batches (1 week per batch) at −1.3, −1.1, and −0.8 V. The average methane production rates for the duplicates achieved in the last two stable cycles are shown in Figure . When cathode potentials were controlled at −1.3 V, the methane production rate of HSSF was 7.2 L CH4/m2 catproj/d, which was 1.4 times higher than that of SSF. It was lower than the methane production rate of 8.8 L CH4/m2 catproj/d for GF. At a cathode potential of −1.1 V, HSSF and GF had similar methane production rates of 1.0 L CH4/m2 catproj/d, which was higher than that of SSF with 0.7 L CH4/m2 catproj/d. At a cathode potential of −0.8 V, methane production for all three materials was low. The highest methane production rate was in the SSF of around 0.08 L CH4/m2 catproj/d, followed by the HSSF of 0.02 L CH4/m2 catproj/d, while methane production rates of the GF were below 0.0015 L CH4/m2 catproj/d. As the thickness of the electrodes could affect the availability of substrate, proton diffusion (from catholyte to electrode), and biofilm development on the electrode, the different thickness electrodes (between GF 3 cm and SSF 1 cm) could affect the results of our study. On the other hand, Sleutels et al. has shown that anode electrode (felt) thickness between 1 mm to 3 mm did not affect the current density (normalized to projected surface area) in microbial electrolysis cells.[28] From an engineering perspective, we normalized each methane production rate by the volume of its cathode electrode (GF 5.6 cm3; SSF/HSSF 2 cm3). HSSF showed the superiority over GF; for example, the methane production rate in HSSF at −1.3 V was 2.3 times higher than that in GF.

Figure 2

Methane production rate calculated at cathode potentials of −1.3, −1.1, and −0.8 V by taking an average of cycles 3 and 4. The highest methane production rate was achieved at the most negative potential. Error bars indicate the standard deviation, calculated from duplicate reactors of the last two stable cycles.

Cumulative methane yields over four consecutive batches for all three cathode electrodes at −1.1 V are shown in Figure . Within each batch, a clear increase in cumulative methane yield over time was observed for all cathode materials. Within each batch, the HSSF had a stable methane yield between 11 and 13 mL. The SSH had a stable but lower methane yield of around 9 mL. The GF had the lowest methane yield (4 mL) in the first batch; however, it increased to the same level (12 mL) as HSSF in batch 3 and batch 4.

Figure 3

Cumulative methane yields over four consecutive batches for all of the three cathode materials at the cathode potential of −1.1 V. The dashed lines indicate 50% medium replacement at the end of each batch.

System Efficiency

The total cathodic efficiency (ηCE) represents the part of the electrons that end up in products (CH4 and/or H2) and is shown in Figure . The highest total cathodic efficiency (including CH4 and H2), between 60% and 80%, was found for the biocathodes controlled at −1.3 V, which had the highest current densities. At less negative cathode potentials (−0.8 and −1.1 V), the cathodic efficiency decreased to below 35%. Slight differences were observed between the materials, with no clear relation between material and ηCE. In addition to methane and hydrogen, dissolved methane[26] and volatile fatty acids (VFAs)[29] were analyzed to see if these could explain the low cathodic efficiencies. However, neither dissolved methane nor VFAs were detected in any of the experiments.

Figure 4

Total cathodic efficiency versus current density of the GF, the SSF, and the HSSF at different cathode potentials of −0.8, −1.1, and −1.3 V.

To further analyze the system efficiency, the results of the current-to-methane (ηCH4), current-to-hydrogen (ηH2) and energy efficiency (ηenergy) for the different potentials of these different electrode materials are shown in Table . Highest current-to-methane efficiency was 60.8% for HSSF at −1.3 V, 56.9% for SSF at −1.3 V and 69.4% for GF at −1.3 V. Hydrogen was only detected at −1.3 V for all cathode materials. HSSF and the GF obtained a similar current-to-hydrogen efficiency of around 40% at the beginning of the batch (Day 1–2), which was higher than that of SSF (30%). However, the HSSF reached a current-to-hydrogen efficiency of 23% at the end of the batch (Day 4–7), which was similar as that of SSF but higher than that of GF (16%). In general, the current-to-hydrogen efficiency decreased with increasing current-to-methane efficiency within each batch, for all cathode materials at the cathode potential of −1.3 V.

Table 1

Overview of the Average Current-to-Methane Efficiency, Current-to-Hydrogen Efficiency, and Energy Efficiency of Each Cathode Material at Different Cathode Potentialsa

		–1.3 V			–1.1 V		–0.8 V
material	relative period in batch 3 and 4b	ηCH4 (%)	ηH2 (%)	ηenergy (%)	ηCH4 (%)	ηenergy (%)	ηCH4 (%)	ηenergy (%)
GF	day 0–1				31.8	13.1	2.6	1.4
	day 1–2	33.6	41.4	11.1	20.3	8.4	1.4	0.8
	day 2–4				19.2	7.9	1.1	0.6
	day 4–7	69.4	15.5	22.9	14.5	6.0	0.5	0.3
SSF	day 0–1				22.9	9.4	10.0	5.5
	day 1–2	28.2	30.6	8.1	18.6	7.7	5.2	2.9
	day 2–4				17.7	7.3	4.7	2.6
	day 4–7	56.9	22.9	16.4	12.7	5.2	2.4	1.3
HSSF	day 0–1				32.9	13.6	2.4	1.3
	day 1–2	27.8	43.8	10.0	18.7	7.7	1.1	0.6
	day 2–4				17.7	7.3	0.9	0.5
	day 4–7	60.8	22.8	21.9	13.7	5.6	0.2	0.1

The average and standard deviation (less than 5%, not shown) were calculated on the basis of 4 separate samples, which were taken from 2 weeks of stable performance (batch 3 and batch 4) with duplicate cathode electrodes for each cathode material. The current-to-hydrogen efficiencies for all reactors were zero at −1.1 and −0.8 V, which are not included in this table.

For example, day 0–1 refers to day 14–15 and day 21–22 within 4 consecutive operational batches at each cathode potential.

The applied voltage in methane-producing BESs can be divided into two parts: a reversible potential loss recovered in CH4, and an irreversible potential loss dissipated in the form of electrode overpotential, ionic losses, and pH gradient over the membrane.[30] The irreversible potential is the lost energy and reflects the extra voltage required in addition to the thermodynamical equilibrium voltage of CH4 generation from CO2 and H2O (1.06 V at standard conditions, which is 1 mol 1 bar for all chemicals involved in the reaction, pH 7, and 298 K).[7] Thus, the higher the applied voltage is, the lower the voltage efficiency. In this study, the cell voltages were similar for all three cathode materials at each cathode potential, which were around 2.1 ± 0.2 V, 2.8 ± 0.1 V, and 3.5 ± 0.3 V for cathode potential of −0.8, −1.1, and −1.3 V, respectively. Therefore, the voltage efficiency was similar for all three cathode materials. The voltage efficiency was 33% at −1.3 V, 41% at −1.1 V, and 55% at −0.8 V. The energy efficiency was calculated as the product of voltage efficiency and current efficiency. The highest energy efficiency was found for HSSF: 22% at −1.3 V, decreasing to 14% at −1.1 V, and further decreasing to 1% at −0.8 V. At each cathode potential of −1.3, −1.1, and −0.8 V, the energy efficiency of HSSF was similar to the energy efficiency of GF and higher than that of SSF.

Electrochemical Analysis

Polarization curves of the three different cathode materials were analyzed before inoculation and after the operation at the cathode potential of −0.8 V (Figure ). For the abiotic test, there was a clear difference between the three cathode materials in terms of catalytic behavior for the hydrogen evolution reaction (HER). The onset potential of the HER of HSSF started already at −0.6 V, which was less negative than that of SSF (−0.8 V). The GF, however, showed almost no catalytic current for the HER in the chosen potential range. After biofilm growth, there was hardly any difference in polarization behavior between the three cathode materials. The onset potential of the reaction (hydrogen or methane production) for all the cathode materials was around −0.8 V. The cathodic current of HSSF was similar before and after inoculation. However, after inoculation, the cathodic current of GF and SSF increased to values almost the same as that of HSSF, showing that the biofilm catalyzed either hydrogen or methane production more effectively than the bare material.[31]

Figure 5

Polarization curves of three different cathode materials (a) before inoculum and (b) after growth at the cathode potential of −0.8 V.

At cathode potentials, less negative than −0.8 V, a small positive current was observed for HSSF and SSF in the polarization curves. This indicates that these materials may be prone to corrosion if used under typical anode conditions,[32] although this effect was not observed in our study as biocathode.

Morphology of Biofilm

SEM images of three different biocathodes showed the presence of microorganisms on the surface of the three cathode materials (Figure ). In general, good coverage of biofilm on the GF electrode was observed, whereas for SSF and HSSF, biofilm coverage was less dense.

Figure 6

SEM images of microorganisms’ attachment on the surface of the (a and b) HSSF, (c) SSF, and (d) GF after growth at the cathode potential of −1.3 V for almost one month.

Discussion

After the start-up phase, all reactors were poised at a cathode potential of −1.1 V, which could be sufficient to drive the hydrogen evolution reaction (HER), especially for HSSF and SSF, as shown in the polarization curves (Figure a). The HER in the HSSF and the SSF probably promoted methane production with a faster start-up process, whereas the poor HER of GF resulted in lower methane production yield in the first biotic batch. In the absence of H2, the growth of methanogens may be slower because hydrogenotrophic methanogenesis (indirect via H2) has been suggested as a vital pathway for the methane-producing BESs.[10] The heat treatment process enhanced the performance of the SSF both in abiotic HER (Figure a) and in methane production yield at more negative cathode potentials, i.e., −1.1 and −1.3 V (Figure ). This can be attributed to the formation of 3D iron oxide nanoparticles on the surface of the HSSF. It has been found that nanostructured electrodes could increase the abiotic reaction rate by enlarging the electro-active surface area, and also stimulate the development of an electro-active biofilm by proving additional anchoring points for microbial adhesion.[19] The SEM images showed that microorganisms on the HSSF seemed to form a matrix of extracellular polymeric substances (EPS) which allowed adhesion of cell-to-cell and cell-to-electrode surface (Figure b), whereas the microorganisms on the SSF formed a loose matrix (Figure c). It has also been found that the HSSF bioanode could facilitate robust electro-active biofilm formation and increased current generation in BESs: current densities achieved for bioanodes on HSSF were several-fold higher than for SSF and the carbon-based felt.[33,34] When applied as a biocathode for methane production, however, HSSF had a performance similar to that of GF at the cathode potential of −1.1 V, and its performance was slightly lower than GF at the cathode potential of −1.3 V. The fact that HSSF showed lower improvement compared to GF when used as cathode rather than as anode could be due to different mechanisms for electron transfer between microorganisms and the electrode. At a cathode potential of −0.8 V, the low methane production rates obtained in all the biocathodes were in line with the results from a polarization curve (Figure b), which indicated that cathode potential of −0.8 V was not negative enough to obtain a substantial current in all these biocathodes. It is worth noticing that the methane yield for GF gradually increased, and reached similar and stable performance as HSSF during batch 3 and batch 4 at the cathode potential of −1.1 V (Figure ). Furthermore, polarization curves changed after the biofilm developed on GF (Figure ): (a) current density increased considerably after biofilm formation (biotic GF) compared to abiotic tests at the same potential, and (b) the onset potential for hydrogen evolution was less negative in the biotic case compared to the abiotic experiments. These results suggest that the presence of biofilm on the GF could play a role by catalyzing hydrogen evolution that enhances methane production.[16,35] Although no hydrogen was detected in the headspace, it might have been directly consumed by the methanogens.[7] As GF is a material with good biocompatibility,[18] there was evidence for a dense biofilm formation on the GF in one of the SEM images (Figure d).

Concerning cathodic efficiency, the impact of dissolved methane and volatile fatty acids (VFAs) was eliminated as neither dissolved methane nor VFAs were detected. However, it is confirmed by Van Eerten-Jansen et al. that using water as anolyte can lower cathodic efficiency compared to other anolytes,[6] e.g., hexacyanoferrate(II) or acetate. Oxygen diffusion through the membrane from the anode to the cathode can lead to a lower cathodic efficiency, as oxygen was the most favorable compound to be reduced at the cathode.[6,36] In this study, around 1% of oxygen was found in the headspace of the cathode circulation bottle, which suggested oxygen diffusion, and its reduction was occurring continuously at the cathode. This process can consume electrons and lower cathodic efficiencies. A low cathodic efficiency (below 35% at cathode potentials of both −0.8 and −1.1 V) can also be caused by other factors that vary with studies: inoculum, catholyte, biomass growth, membrane, system configuration, and duration of the experiment.[10,37,38]

Furthermore, we observed that the cathodic efficiency was related to the current densities rather than to the cathode materials, with higher efficiencies found for higher current densities. In contrast, the voltage efficiency decreased with the increased current density because of the higher applied voltages. Therefore, there is a trade-off between the two, with an optimum somewhere at which the highest energy efficiency is achieved. This study is one of the first ones analyzing the energy efficiency of methane-producing BES in more detail, with maximum values of 22% (Table ), values similar to those determined for short-term yield tests in another study utilizing water as the electron donor.[6] The energy efficiency is a crucial factor to determine the performance of the methane-producing BESs and to assess its capability as an energy storage system.[7] A further increase in energy efficiency is required, which needs to be achieved by further improvements in the system’s performance, to bring methane-producing BESs closer to the application.

Conclusions

Heat treatment of stainless steel felt improved methane production rates of SSF in the methane-producing BESs when operated at −1.1 and −1.3 V versus Ag/AgCl, with performance similar to GF. HSSF had a maximum current-to-methane efficiency of 60.8% and energy efficiency of 21.9% at −1.3 V. These values were similar to the ones found for GF, and higher than those for untreated SSF. Moreover, HSSF had a better electrocatalytic property for hydrogen evolution, leading to a fast start-up of the biocathode. HSSF is an alternative cathode material with similar performance compared to graphite felt, suited for application in methane-producing BESs.