Microbial Electrosynthesis Using 3D Bioprinting of Sporomusa ovata on Copper, Stainless-Steel, and Titanium Cathodes for CO 2 Reduction

: Acetate can be produced from carbon dioxide (CO 2 ) and electricity using bacteria at the cathode of microbial electrosynthesis (MES). This process relies on electrolytically-produced hydrogen (H 2 ). However, the low solubility of H 2 can limit the process. Using metal cathodes to generate H 2 at a high rate can improve MES. Immobilizing bacteria on the metal cathode can further proliferate the H 2 availability to the bacteria. In this study, we investigated the performances of 3D bioprinting of Sporomusa ovata on three metal meshes—copper (Cu), stainless steel (SS), and titanium (Ti), when used individually as a cathode in MES. Bacterial cells were immobilized on the metal using a 3D bioprinter with alginate hydrogel ink. The bioprinted Ti mesh exhibited higher acetate production (53 ± 19 g/m 2 /d) at − 0.8 V vs. Ag/AgCl as compared to other metal cathodes. More than 9 g/L of acetate was achieved with bioprinted Ti, and the least amount was obtained with bioprinted Cu. Although all three metals are known for catalyzing H 2 evolution, the lower biocompatibility and chemical stability of Cu hampered its performance. Stable and biocompatible Ti supported the bioprinted S. ovata effectively. Bioprinting of synthetic bioﬁlm on H 2 -evolving metal cathodes can provide high-performing and robust biocathodes for further application of MES.


Introduction
There have been continuous global attempts to reduce carbon dioxide (CO 2 ) emissions.CO 2 can be captured and transformed chemically or biologically into useful compounds.Microorganisms are being utilized for the electrocatalytic conversion of CO 2 to compounds like methane (CH 4 ), acetate, and ethanol at the cathode with the input of electric power.Bacterial genera such as Sporomusa and Clostridium have been shown to use electric current to turn CO 2 into acetate and other multi-carbon compounds at the cathode in a system called microbial electrosynthesis (MES) [1,2].MES has gained high interest in recent years for capturing and utilizing carbon in valuable chemicals or fuels [3].As such, it shows the potential to reduce our reliance on fossil fuels and combat climate change [4].However, current MES technology is not yet economically viable.Only bulk products such as acetate and/or short-chain carboxylates are reported as being produced in MES [5].Improving the performance of MES systems will be crucial in making these technologies more cost-effective.
The productivity of MES reactors has been successfully increased using a variety of methods, such as S. ovata adaptation to methanol-containing growth medium [6] and structural optimization of the microbial and cathode interface [7][8][9].Studies have focused on the utilization of mixed cultures derived from natural anaerobic sources for MES [10,11], and long-term enrichment of the microbiome and biofilm has led to significant performance gains [12][13][14].Studies obtaining the highest productivity have focused on modifying the cathode in various ways.There have been two main approaches.The first is to increase the biomass concentration on the cathode [13,[15][16][17][18].The second approach is to modify the electrochemical properties of the cathode, either through the surface modification of carbonaceous electrodes or through the addition of metal particles or using metallic electrodes [12,[19][20][21].One of the difficulties in designing ideal MES cathodes is that they need to meet several requirements simultaneously, namely high electrical conductivity, good biocompatibility, and a high surface area, while having a high durability and low production cost [22].Some of these properties tend to be difficult to meet; for example, copper foam electrodes fulfill all the requirements, except for biocompatibility, since copper can be toxic to several microorganisms [23].But investigations on suitable and cost-effective metal cathodes for MES are still lacking.S. ovata has often been used as a model organism in MES, and is well known for its use in microbial electrosynthesis reactors, producing mainly acetate [1,24].Mixed cultures might be more feasible for large-scale operations, but the large variability of microbial structure and function in mixed cultures would add uncertainties in the performance of cathode materials [25].S. ovata is, however, not capable of forming the dense biofilms as observed in other electroactive bacteria such as Geobacter sulfurreducens [26,27].That being said, increasing the amount of microbial cells directly on the electrode has shown significant increase in production; for example, the use of silicon nanowire arrays to entrap the cells [18], electrode modification to make it more biocompatible [28], and also the use of 3D-bioprinted biofilms [16] are the ways to increase cell density in cathode biofilm.For 3D-bioprinting biofilms, a concentrated cell suspension is mixed with a cross-linkable hydrogel and printed directly on the cathode, forming a dense biofilm in approximately a third of the normal time required (less than two days, compared to a week typically required) [16,29].Bioprinted biofilms, for example, allow for significant biofilm growth, even when using microbes that do not naturally form thick biofilms, such as S. ovata, and has been shown to greatly increase acetate production in MES with an average acetate production of 47 ± 5.1 g/m 2 /d (0.31 ± 0.55 g/L/d) [16].
Electron uptake from the cathode in MES systems occurs in two ways, direct electron uptake by the biocatalyst or indirect electron uptake via H 2 production [30].Although the core mechanics of electron uptake are still uncertain, it has become widely accepted that indirect electron uptake is the main driving force for MES [24].MES reactors have been shown to enhance the total production of hydrogen, even when merely using spent media, and it has been proposed to occur via the aid of secreted hydrogenases or catalytically active precipitates on the electrodes [31].This is most likely the reason for the increased MES productivity while enhancing the biofilm thickness.The gas-liquid mass-transfer is also seen as the limiting factor in the uptake of H 2 and CO 2 , with most of the gas escaping the reactor, especially when considering the low solubility of H 2 .Several reactor designs have been tried to improve the gas to liquid mass-transfer [13,20,21].This makes the combination of H 2 evolving electrodes and well-established biofilms especially productive, since the H 2 can be consumed before it can escape the reactor.Metallic cathodes are generally considered the most effective for H 2 evolution via electrochemical water splitting, and non-precious HER catalysts such as Ti, Cu, and stainless steel (SS) can be used for MES reactors [32].It was evident from Bajracharya et al. [21] and Krige et al. [16] that hydrogen is required for the CO 2 reduction, and the rates were high when hydrogen-evolving Ti mesh was introduced as a second cathode in MES [21].Higher rates of acetate production were reported by using Sporomusa ovata in 3D bioprinting [16].Based on the results from these studies, we are proposing to use bioprinting of bacterial cells directly on the Cu, SS, and Ti metal cathodes for achieving a high rate of hydrogen evolution.However, there are some concerns that biofilm formation can be hindered by antimicrobial metal ion formation through the oxidation of such cathodes during the long-term operation of MES [33,34].Few studies have used metallic cathodes for MES, but to our knowledge no studies have been reported using 3D-bioprinted biofilms with metallic electrodes.By combining synthetic 3D-printed S. ovata biofilms and H 2 -evolving metallic cathodes, in this study we are investigating the MES performances of different metal cathodes to minimize the anti-microbial effects, while still maintaining a thick biofilm on the cathode.In addition, we also recirculated the headspace gas to the catholyte in all reactors, to better reutilize the escaping CO 2 and H 2 gases.

Cultivation and Enrichment of S. ovata
Sporomusa ovata cultures (DSM 2662) were purchased from Deutsche Sammlung Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany, and cultivated anaerobically in standard DSM 311 media without casitone.The subsequent cultures were carried out under a mixture of H 2 , CO 2 , and N 2 (60:20:20) in the same media without casitone and betaine.To stimulate the cells for autotrophic growth, only CO 2 and HCO 3 − were used as a carbon source and H 2 was used as an electron donor.A small amount of yeast extract (0.5 g/L) was added to the media to help with initial biomass growth.

MES Reactor Construction and Operation
All experiments were carried out in standard H-type, dual-chambered glass reactors (270 mL per chamber), separated by a Nafion 117 ion exchange membrane (Ion Power GmbH, München-Flughafen, Germany).DSMZ 311 media without organic substrates were used as electrolytes in MES.The reactors were stirred using a magnetic stirrer and kept in a temperature-controlled plexiglass box (35 • C).Graphite rods (Schunk Carbon Technology AB, Lenhovda, Sweden; diameter: 20 mm, height: 70 mm) were used as an anode.The cathode chamber was equipped with an Ag/AgCl reference electrode (eDAQ, Denistone, East NSW, Australia).The cathodes consisted of 25 mm × 25 mm sized square-shaped metal meshes of copper and stainless steel.The titanium mesh was of size 28 mm × 70 mm.Final productivity is therefore standardized with regards to area and reported as area-based productivity to account for differences in electrode area.Each metal mesh cathode had a wire of the same material as a current collector, which also held the cathode by being pierced through the butyl stopper.The copper electrode was first annealed at 300 • C for 8 h to form a CuO x layer [35].Annealed electrodes are efficient in catalyzing CO 2 reduction and are less sensitive to the deactivation phenomena that destroy bulk metal electrodes [23,35].
A gas bag (~3L volume) filled with 100% CO 2 was added to the cathode chamber, and the gas was recirculated to the catholyte using a peristaltic pump.The gas bag had two opening ports, one for collecting the cathode headspace gas from the reactor and another one for pumping the gas back to the catholyte.The recirculation of gas from the gas bag to the catholyte was carried out at 6-10 mL/min using a peristaltic pump.With the operation of the MES, hydrogen gas was produced and accumulated in the gas bag, thus, H 2 and CO 2 gas mixture was being collected in the gas bag and recirculated.The reactors were connected to a multi-channel potentiostat (MultiEmStat+, Palmsens, Houten, The Netherlands).The metal cathodes with 3D-bioprinted S. ovata were used as working electrodes which were poised at −0.8 vs. Ag/AgCl or more negative potentials to ensure cathodic H 2 evolution.At the start-up, the cathode chamber was first sparged with 100% CO 2 for 5 min.The composition of gas in the gas bag changed throughout the MES run, since excess H 2 was produced, and the CO 2 was converted to acetate by bacteria.In addition, the volume of gas in the bag also decreased during operation of MES and the bags were subsequently refilled with 100% CO 2 every 3-4 days to maintain sufficient CO 2 availability.The anode side was not sparged with any gas, but was in continuous stirring using a stirring magnet.During the MES operation, O 2 evolution was occurring at the anode.There was a Nafion 117 proton exchange membrane (PEM) separating the anode and cathode to prevent the passing of O 2 to the cathode side.

Bio-Printing of Synthetic Biofilms
A synthetic biofilm directly onto the cathode was created on the metal-mesh cathodes by 3D-bioprinting an alginate-based hydrogel, containing cellulose nanofibrils (Cellink bioink, Cellink, Sweden), inoculated with S. ovata.The hydrogel preparation and printing were performed in an anaerobic hood, to maintain anaerobic conditions as previously described [16].A concentrated solution of S. ovata was prepared by centrifugation of the entire volume of a 100 mL inoculum culture down to a pellet and resuspending the pellet in 0.3 mL media.The 0.3 mL of cell suspension was then mixed with 3 mL of gel.The gel was then used to print several electrodes, to maintain similar starting cell concentrations between the replicates.The cathodes were held in place by a sterilized 3D-printed fixture, and a single layer of hydrogel (15 mm × 15 mm, ~0.5 mm thick) was printed onto the cathode using the 3D printer.An ionic crosslinking agent (CELLINK, Gothenburg, Sweden) was then applied over the printed layer to cross-link the gel before fixing the cathode in the reactor, containing 270 mL of DSMZ 311 media as catholyte.A photograph of the bioprint layer on the Cu cathode is shown in Figure 1.

Bio-printing of synthetic biofilms
A synthetic biofilm directly onto the cathode was created on the metal-mesh cathodes by 3D-bioprinting an alginate-based hydrogel, containing cellulose nanofibrils (Cellink bioink, Cellink, Sweden), inoculated with S. ovata.The hydrogel preparation and printing were performed in an anaerobic hood, to maintain anaerobic conditions as previously described [16].A concentrated solution of S. ovata was prepared by centrifugation of the entire volume of a 100 mL inoculum culture down to a pellet and resuspending the pellet in 0.3 mL media.The 0.3 mL of cell suspension was then mixed with 3 mL of gel.The gel was then used to print several electrodes, to maintain similar starting cell concentrations between the replicates.The cathodes were held in place by a sterilized 3D-printed fixture, and a single layer of hydrogel (15 mm × 15 mm, ~0.5 mm thick) was printed onto the cathode using the 3D printer.An ionic crosslinking agent (CELLINK, Gothenburg, Sweden) was then applied over the printed layer to cross-link the gel before fixing the cathode in the reactor, containing 270 mL of DSMZ 311 media as catholyte.A photograph of the bioprint layer on the Cu cathode is shown in Figure 1.

Chemical analyses
Samples of catholyte (1~2 mL) were collected from the cathode chambers and filtered through 0.2 µm syringe filters.Acetate and/or other volatile fatty acids were analyzed using a high-performance liquid chromatograph (HPLC) (PerkinElmer Inc., Waltham, MA, USA), equipped with an Aminex HPX-87H column (Bio-Rad Laboratories Inc., Hercules, CA, USA), and a refractive index detector.A solution of 5 mM H2SO4 (0.6 mL/min) was used as the mobile phase and the column was kept at 65 °C.

Calculations
The performance of the MES was evaluated in terms of rate of production of the value-added products and cathodic columbic efficiency (CE).In this study, only acetate production was considered for calculations.The rate of production of any reduced compound (P) in g/L/d in a batch process is given by: Pi,t = (Ci, t − Ci, t-1)/Δt (1) where subscripts i refer to any product from CO2 reduction and subscripts t and t-1 refer to subsequent samples.C is concentration (g/L) of product (here acetate), Δt is the time difference (days) between the two samples.Number of moles of acetate produced was calculated as:

Chemical Analyses
Samples of catholyte (1~2 mL) were collected from the cathode chambers and filtered through 0.2 µm syringe filters.Acetate and/or other volatile fatty acids were analyzed using a high-performance liquid chromatograph (HPLC) (PerkinElmer Inc., Waltham, MA, USA), equipped with an Aminex HPX-87H column (Bio-Rad Laboratories Inc., Hercules, CA, USA), and a refractive index detector.A solution of 5 mM H 2 SO 4 (0.6 mL/min) was used as the mobile phase and the column was kept at 65 • C.

Calculations
The performance of the MES was evaluated in terms of rate of production of the valueadded products and cathodic columbic efficiency (CE).In this study, only acetate production was considered for calculations.The rate of production of any reduced compound (P) in g/L/d in a batch process is given by: where subscripts i refer to any product from CO 2 reduction and subscripts t and t-1 refer to subsequent samples.C is concentration (g/L) of product (here acetate), ∆t is the time difference (days) between the two samples.Number of moles of acetate produced was calculated as: where n is the number of moles of a product, Vcat is the total volume of the catholyte (L), and M is its molecular weight (g/mol).Electric equivalents of acetate (in moles) were calculated from the electric current as: where I is the current (in ampere), 8 is the moles of electrons needed to produce one mole of acetate, F is Faraday's constant F = 96,485 C/mol e−.The current in equivalent acetate was later converted to g/L to match with the measured acetate concentration profile.Coulombic efficiency (CE) accounts specifically for the fraction of electrons that ended up as acetate produced from CO 2 reduction.CE for CO 2 reduction is the efficiency of capturing the electron from the electric currents to the product(s).CE = n acetate /Current equivalent acetate × 100% (4)

Scanning Electron Microscopy (SEM)
Small pieces of metal cathodes, with the bioprint, were cut off, and the biofilm was then fixed by submerging it in 2.5% w/v glutaraldehyde overnight and then freeze-dried to prepare for SEM imaging.Biofilms on the metal cathodes were visualized in FEI Magellan 400 SEM, to study the structure and morphologies.

Results
After centrifugation, the cells of S. ovata were bioprinted on various metal cathodes instead of carbon-based electrodes and tested for their performance to reduce CO 2 .Artificial biofilms were printed successfully on the copper, stainless-steel, and titanium meshes by entrapping the bacterial cells on the alginate-based hydrogel.S. ovata has been shown to reduce CO 2 to organic compounds using either electrons supplied by the cathode or hydrogen produced at the cathode.

MES Performance of Bioprinted Cu Mesh
Two MES reactors with S. ovata bioprinted copper cathodes, namely CuBP1 and CuBP2, were operated as batch reactors at −0.8 V vs. Ag/AgCl.High reduction currents ranging from −11 mA to −28 mA were recorded at the starting phase of both copper bioprinted (CuBP) cathodes (especially in the first 5 days of CuBP1) (Figure 2A).This phenomenon of high current at the starting phase was also mentioned by Chatzipanagiotou et al. [35,36] while using Cu and Ni as cathodes.Electrochemical reduction of metal oxides could be associated with the initial high currents [23,35].Annealing of Cu produced a uniform layer of CuO x (Figure 3A,B) on the electrode surface that was expected to resist the possible corrosion of Cu and to some extent prevented the anti-bacterial effect of Cu ions.In CuBP1 MES, the current started to decrease after 5-10 days and then stabilized around −5 to −8 mA (Figure 2A), whereas the stabilized current in CuBP2 was around −2 mA at −0.8 V (Supplementary Materials Figure S1).
The outer layer of bioprint layer started to fall off the cathode over time.The MES operation was continued, as a thin layer of bioprint, along with the encapsulated bacteria, remained attached.The shredded bioprint bacteria may also have contributed to CO 2 reduction, being in suspended form.Immobilized cells on the surface of the alginate bioprint are visible in SEM images of the CuBP1 cathode (Figure 3C-E).In bioprinting, the hydrogel containing alginate and methacrylate was crosslinked using low-intensity UV radiation after printing on the carbon cloth.After crosslinking, the bioprint layer becomes mechanically robust and chemically resistant.Furthermore, the microbial biodegradation of gelatin is rare and has not been reported for S. ovata.It is therefore unlikely for the bioprint to contribute as a carbon source while operating as biocathode.However, the breakage of bioprint from the cathode was observed, which was likely related to the adhesion on the metal, rather than microbial degradation.Although the bioprinted biofilms were sufficiently stable for the duration of the experiments, the alginate hydrogels were not ideal for long-term operation on cathodes.MES cathodes typically experience a localized increase in pH at the biofilm-cathode interface, resulting in pH values far above the bulk pH [37].This increased pH might result in faster alginate detachment, since alginate is known to degrade at pH values above 10 [38].This might have led to an increase in the breaking rate of the bioprinted biofilm at the cathode-hydrogel interface, resulting in the biofilm sloughing off over time.Although the alginate hydrogels are not ideally suitable for long-term cathodic biofilms, we believe that the results would still be indicative of bioprinted MES performance, since the gels remained attached for a sufficient duration.Further experiments are suggested using more stable, chemically crosslinked hydrogels, such as UV-crosslinked or gelatin-based hydrogels.The outer layer of bioprint layer started to fall off the cathode over time.The MES operation was continued, as a thin layer of bioprint, along with the encapsulated bacteria, remained attached.The shredded bioprint bacteria may also have contributed to CO2 reduction, being in suspended form.Immobilized cells on the surface of the alginate bioprint are visible in SEM images of the CuBP1 cathode (Figure 3C-E).In bioprinting, the hydrogel containing alginate and methacrylate was crosslinked using low-intensity UV radiation after printing on the carbon cloth.After crosslinking, the bioprint layer becomes mechanically robust and chemically resistant.Furthermore, the microbial biodegradation of gelatin is rare and has not been reported for S. ovata.It is therefore unlikely for the bioprint to contribute as a carbon source while operating as biocathode.However, the breakage of bioprint from the cathode was observed, which was likely related to the adhesion on the metal, rather than microbial degradation.Although the bioprinted biofilms were sufficiently stable for the duration of the experiments, the alginate hydrogels were not ideal for long-term operation on cathodes.MES cathodes typically experience a localized increase in pH at the biofilm-cathode interface, resulting in pH values far above the bulk pH [37].This increased pH might result in faster alginate detachment, since alginate is known to degrade at pH values above 10 [38].This might have led to an increase in the The concentration of acetate increased steadily in both CuBP1 and CuBP2 over the first 8 days, reaching around 0.3 to 0.5 g/L, but in CuBP2 the concentration did not increase after 0.3 g/L, but rather started to decline, although high cathodic currents were recorded (Figure 2B).Due to this low production, we are not discussing CuBP2 further.The combined effect of copper leaching and mass-transfer limitation of substrates to the bacteria within the bioprint could be the main reason that triggered the inactivation of bacteria and thus resulted a drop in current in CuBP2 after eight days of operation.
The highest acetate production rate at this initial period of operation in CuBP1 was 0.07 ± 0.01 g/L/d (30.2 ± 5.4 g/m 2 /d) with a coulombic efficiency of 26%.A small amount of formate production (up to 640 mg/L) was also observed in CuBP reactors, but it did not show a steady accumulation, hence it is not presented or discussed here further.To allow more hydrogen evolution at the cathode, the potentials were switched to −0.9 and −1 V in both CuBP1 and CuBP2 after 10-12 days of operation at −0.8 V.The more negative cathode potential resulted in the production of approximately 1 g/L acetate in 18 days in CuBP1.Acetate production in CuBP1 continued at a slower rate, 0.04 ± 0.02 g/L/day (16 ± 10 g/m 2 /d), reaching 1.74 g/L after 72 days of operation, even though a large reduction current was observed.From the acetate production profile in Figure 2B, it appears that the acetate accumulated in the CuBP1 MES only corresponded to a minority of the reduction current recorded in the MES, since the acetate equivalent calculated from the current increased much faster, theoretically supporting an acetate concentration up to 9 g/L within 40 days.Overall, the bioprinted Cu cathode produced on average 0.05 ± 0.02 g/L/day (23 ± 7 g/m 2 /d) acetate production.However, the MES performances of Cu cathodes remained inconsistent, as there were variations in performance with the batch operation of each individual reactor and also between the two reactors.Due to the complexities of cathode reactions, current fluctuations were observed.Indeed, the bioprinting on the metal cathode added complexities at the cathode, due to the combined effects of the electrochemical activity of the metals, mass-transfer limitation within the bioprint, and the bacterial electrochemistry.The combined effects from these multiple interactions created unknown variables at the cathode biofilm.In addition, from practical experience on electrochemical systems, reference electrodes may lose their stability during long-term operation.The changes in current in the reactors can also be attributed to the fluctuations in individual reference electrodes.So, the duplicate reactors showed different currents when the status of individual reference electrodes changes.
The copper oxides deposited on the cathodes are known for their electrochemical catalysis of CO 2 reduction, and the combination of CuOx and microbial catalyst was reported to enhance CO 2 reduction [23,35].Layers of copper oxide formed after annealing at 300 • C were visible on the SEM image of Figure 3B.In our study, the bioprinted Cu cathodes did not show a high rate of CO 2 reduction, most likely due to microbial inhibition caused by Cu ions.Transient metal electrode leaching cannot be avoided while the electrode is submerged in a strong electrolyte without applying reduction potential.Chatzipanagiotou et al. [23,35] showed an increase in the Cu ion concentration of the catholyte at the beginning of batch operation when using Cu cathodes in MES.Over time, while maintaining the cathodic electrical potentials, the concentration of Cu ions slowly declined because the metal ions were reduced and deposited on the cathode [23,35].Aryal et al. [33] also showed that S. ovata did not thrive on a copper-coated graphene cathode, whereas high performance was observed when graphene was coated on top of copper foam.These studies suggest that direct contact of S. ovata with the surface of less corrosion-resistant metals such as copper could be inhibitory to the bacteria under the electrochemical reactor conditions.A comparatively low performance of S. ovata observed in CuBP reactors might be due to the similar conditions in this study.Corrosion-like features were visible on the SEM image (Figure 3F) of the CuBP1 cathode.However, further analyses regarding the Cu ion concentration and the impact of solid Cu on S. ovata growth were not carried out in this study.

MES Performance of Bioprinted SS Mesh
Bioprinting of S. ovata was carried out on a stainless-steel (SS) mesh named SSBP, without the annealing as mentioned for Cu mesh.The acetate concentration and reduction currents from SSBP are presented in Figure 4A,B.Stainless steels are known for chemical stability due to passive layer of oxide formation [39], and SSs are commonly used in MES reactors as a current collector [40].SS cathodes are known low-cost non-noble electrocatalysts for hydrogen evolution [41,42].As soon as the bioprinted SS cathodes were polarized at −0.8 V vs. Ag/AgCl to start MES operation, a high current of −4 mA was observed (Figure 4A).The high current resulted in a high rate of H 2 evolution through water electrolysis.In addition, the reduction of the oxide layer formed on the SS could be the reason for the high initial current, as discussed previously.Within a day, the current decreased to −1 mA, most likely because the available metal oxides might have been reduced.The reduction current in the SSBP MES reactors increased over several days and remained between −5 to −10 mA (Figure 4A).As described earlier for CuBP MES, a portion of the bioprint layer on the SS mesh broke off and a thin layer of bioprinted gel remained on the SS cathodes.SEM images in Figure 4C,D show that the SS cathode was covered with a considerably thick biofilm.Acetate levels in the catholyte started to rise from the beginning when the cathode was polarized at −0.8V vs. Ag/AgCl.The production continued steadily and reached 0.74 g/L of acetate in 8 days (Figure 4B).This acetate level corresponds to a highest production rate of 38.62 ± 1.4 g/m 2 /d and SSBP showed continuous accumulation of acetate afterward, maintaining an average rate of 0.1 ± 0.01 g/L/d (24.6 ± 8.8 g/m 2 /d).The sizes of SS and Cu cathodes used in this study were relatively small (2.5 × 2.5 cm) for the reactor volume (270 mL), thus the volumetric acetate production rates were relatively low in these MES reactors, as compared to the rates obtained with bioprinted carbon cloth in Krige et al. [16].
stability due to passive layer of oxide formation [39], and SSs are commonly used in MES reactors as a current collector [40].SS cathodes are known low-cost non-noble electrocatalysts for hydrogen evolution [41,42].As soon as the bioprinted SS cathodes were polarized at −0.8 V vs. Ag/AgCl to start MES operation, a high current of −4 mA was observed (Figure 4A).The high current resulted in a high rate of H2 evolution through water electrolysis.In addition, the reduction of the oxide layer formed on the SS could be the reason for the high initial current, as discussed previously.Within a day, the current decreased to −1 mA, most likely because the available metal oxides might have been reduced.The reduction current in the SSBP MES reactors increased over several days and remained between −5 to −10 mA (Figure 4A).As described earlier for CuBP MES, a portion of the bioprint layer on the SS mesh broke off and a thin layer of bioprinted gel remained on the SS cathodes.SEM images in Figure 4C,D show that the SS cathode was covered with a considerably thick biofilm.Acetate levels in the catholyte started to rise from the beginning when the cathode was polarized at −0.8V vs. Ag/AgCl.The production continued steadily and reached 0.74 g/L of acetate in 8 days (Figure 4B).This acetate level corresponds to a highest production rate of 38.62 ± 1.4 g/m 2 /d and SSBP showed continuous accumulation of acetate afterward, maintaining an average rate of 0.1 ± 0.01 g/L/d (24.6 ± 8.8 g/m 2 /d).The sizes of SS and Cu cathodes used in this study were relatively small (2.5 × 2.5 cm) for the reactor volume (270 mL), thus the volumetric acetate production rates were relatively low in these MES reactors, as compared to the rates obtained with bioprinted carbon cloth in Krige et al. [16].The reduction currents recorded in the SSBP reactor were relatively low up to 30 days of operation, and thus the acetate equivalents calculated from the currents were lower than the measured acetate (Figure 4A,B).Higher acetate measurements than electrical equivalents indicated the presence of additional electron sources other than electricity during that operation period; thus, the coulombic efficiency exceeded 100% for that period.Although the additional electron sources were not identified, the contribution from cell decay is most probable.A small amount (0.5 g/L) of yeast extract was also available for bacterial growth, which may partly contribute as a carbon source for acetate production.Nevertheless, the contribution remained minimal, since the acetate accumulation in the electrolyte reached 2.6 g/L (significantly higher than the 0.5 g/L yeast extract) over the course of MES operation, confirming that CO 2 was fixed to produce acetate.After 30 days of operation, higher currents were observed and acetate equivalents from electricity exceeded the measured acetate level.The accumulation of slightly more than 2.5 g/L of acetate corresponds to 51% of CE over 70 days of operation of SSBP MES.

MES Performance of Bioprinted Ti Mesh
After bioprinting of S. ovata on Ti mesh, the MES, named TiBP, was operated at −0.8 V vs. Ag/AgCl cathode potential.A high current of −25 to −30 mA was recorded in TiBP MES accompanying a high rate of H 2 evolution through water electrolysis and acetate production (Figure 5A,B).An additional Ti cathode was previously used in an MES to support high-rate CO 2 reduction [21].The reduction of oxides formed on the Ti also contributed to the high reduction current.Ti is a known HER catalyst, with high resistance to corrosion due to the passive oxide layer developed on its outer surface [43].The acetate concentration in the catholyte quickly increased at the early stage of operation, reaching the highest instantaneous production rate of 0.84 g/L/d (130.5 g/m 2 /d).The acetate production stabilized after 4 days, and remained around 0.34 g/L/d (53 g/m 2 /d) on average for 18 days, reaching a concentration of 4 g/L acetate.After day 18, the acetate production continued steadily at an average rate of 0.1 g/L/d for a long time, up to 87-90 days, reaching 9.4 g/L at the end of the run (Figure 5A).The CE of acetate production in the TiBP reactor ranged from 75 to 35% over a long operation of 90 days.What is noteworthy in TiBP is that the bioprint gel layer did not break and remained attached to the Ti mesh, despite the fact that high currents and hydrogen evolution were observed.SEM images of the Ti cathode in Figure 6A,B show a full coverage of the Ti surface with the biofilm.So far, it appeared that the bonding of alginate gel with Ti or with its oxide was stronger than bonding with other two metals.In addition, the pH gradient developed at the electrode-bioprint interface was not severe to the bioprint layer What is noteworthy in TiBP is that the bioprint gel layer did not break and remained attached to the Ti mesh, despite the fact that high currents and hydrogen evolution were observed.SEM images of the Ti cathode in Figure 6A,B show a full coverage of the Ti surface with the biofilm.So far, it appeared that the bonding of alginate gel with Ti or with its oxide was stronger than bonding with other two metals.In addition, the pH gradient developed at the electrode-bioprint interface was not severe to the bioprint layer when using Ti.Ti is resistant to corrosion, catalytic to HER, and also known for high biocompatibility in comparison to SS and Cu.This property of Ti might have allowed the high stability and high performance of the bioprinted S. ovata.However, detailed analyses in this direction remained out of the scope of this study.

Bioprinting of S. ovata relatively increased acetate production from CO2 reduction
S. ovata has been used as a biocatalyst to reduce CO2 to acetate using several metal cathodes in a microbial electrosynthesis (MES) system.A comparative overview of performance of S. ovata in MES with different cathode materials is presented in Table 1.Previous studies have reported that the acetate production rate using S. ovata in MES system with graphene-functionalized carbon cloth at −0.93 V was almost 14 g/m 2 /day [28], and on Si nanowire array cathodes it was 44.3 g/m 2 /d at −1.4 V [18].An acetate production rate of 1.85 g/m 2 /d was reported at −0.6 V with Ni hollow-fiber cathodes with carbon nanotube coating [20].In a similar study of metal cathodes for methane production from CO2, Wang and coworkers employed stainless steel, copper, and nickel mesh biocathodes and reported increased methane generation with nickel, likely due to the higher electrocatalytic activity of this electrode for the synthesis of H2 [44].S. ovata has been used as a biocatalyst to reduce CO 2 to acetate using several metal cathodes in a microbial electrosynthesis (MES) system.A comparative overview of performance of S. ovata in MES with different cathode materials is presented in Table 1.Previous studies have reported that the acetate production rate using S. ovata in MES system with graphene-functionalized carbon cloth at −0.93 V was almost 14 g/m 2 /day [28], and on Si nanowire array cathodes it was 44.3 g/m 2 /d at −1.4 V [18].An acetate production rate of 1.85 g/m 2 /d was reported at −0.6 V with Ni hollow-fiber cathodes with carbon nanotube coating [20].In a similar study of metal cathodes for methane production from CO 2 , Wang and coworkers employed stainless steel, copper, and nickel mesh biocathodes and reported increased methane generation with nickel, likely due to the higher electrocatalytic activity of this electrode for the synthesis of H 2 [44].
The bioprinting technique of S. ovata has been shown to effectively localize a high density of bacteria on the cathode [16].The results from previous publications, Krige et al. [16] and Bajracharya et al. [21], on 3D bioprinting were used as a control study for the present metal bioprinting experiments.In Krige et al. [16] plain carbon cloth was used, whereas in this study the carbon cloth electrodes were replaced with the metal meshes and the same procedure for bioprinting was followed.The comparison of results of this study to our previous studies are presented in Figure 6B and discussed here.While bioprinting on plain carbon cloth, Krige et.al. [16] achieved a maximum of 52 g/m 2 /d, whereas in this present study, the acetate production rate was increased to 72 g/m 2 /d (maximum) with the Ti cathodes and bioprinting.With Cu and SS cathodes, we observed less than half the acetate production rate than Ti with bioprinted S. ovata in this study, and the rates from Cu and SS were lower than the rates from bioprinting of S. ovata on carbon cloth as reported in Krige et al. [16].An important experimental consideration to note here in this study is that the size of the Ti cathode was three times bigger than CuBP and SSBP.The total current recorded in TiBP was relatively higher than CuBP and SSBP due to the higher area of Ti.To compare the currents with different bioprinted cathodes, the results of linear sweep voltammetry (LSV) carried out on the different cathodes at 5 mV per second scan rate are shown in Figure 7. CuBP showed the lowest onset potential (~−0.5 V vs. Ag/AgCl) for cathodic reaction in Figure 7. MES with Ti reached similar current densities as Cu, but the onset potential for cathodic reaction was between −0.6 to −0.7 V vs. Ag/AgCl.Current density from Ti was higher than that of Cu after −0.9 V vs. Ag/AgCl.The HER performances of CuBP and TiBP were higher than that of SSBP and non-bioprinted C-cloth, as the current densities for SSBP and C-cloth are considerably lower in Figure 7.The current densities are normalized to the projected area of the cathode.At more negative potentials than −0.8 V vs. Ag/AgCl, the current densities from TiBP are highest, then those from CuBP and the lowest from SSBP.For visual comparison purposes, the LSV from non-bioprinted C cloth obtained from Krige et al. [16] is also shown.In previous study [21], also it was also reported that a larger area Ti cathode with S. ovata had shown lower currents than that from the graphite rod cathode when highly active S. ovata biofilm was established on it.Thus, the high currents with the Ti cathode were not only due to the area but also due to the higher activity of bacteria on the Ti cathode as evident from Figure 7.In fact, electrochemical catalysis is actually dependent on the active surface area of the electrode rather than the physical area of the electrode [45].Acetate production rates are also presented by normalizing to the electrode area (per m 2 ) for the comparison of performances of different metal cathodes in Table 1.The electrode area-based production rates allow for the comparison of different cathodes' performances although they differ in sizes, and is a standard method for reporting MES productivity.The area-based acetate production rates were more than doubled in TiBP compared to the rates from CuBP and SSBP (Table 1).The use of metal cathodes and the bioprinting of microbes combine the benefit of direct metal electrocatalysis for H 2 production and the subsequent reduction of CO 2 by microorganisms, using H 2 as a substrate for biological reactions, improving the overall production rate.Krige et al. [16] introduced 3D bioprinting of S. ovata on carbon cloth for MES.As a continuation, in this study we used 3D bioprinting on several metal mesh cathodes.Bioprinting allows the bacteria to be localized on the cathode rather than being in suspension and slowly forming cathode biofilm at their own pace, if at all; thus, bioprinting increased the possibilities of bacterial interaction with the cathode [16].Fast start-up and bacterial localization on the cathode for immediate access to the reducing equivalents are the main advantages of bioprinting on the cathode.These advantages improved the CO 2 reduction rates in our MES studies.Additionally, the encapsulated and localized bacterial cells in the cathode surface within the hydrogel of bioprinting also appeared helpful in avoiding the inhibitory effect of minor levels of oxygen penetration from the anode side.Hence, the bioprinting technique is useful to operate MES which has an O 2 -evolving anode for long-term operation, as diffusion limitation is created with the bioprinting technique.However, the bioprinting technique may need further improvement and adaptation for the enhancement of biocathode for CO 2 reduction in MES, as there was low reproducibility in the performances of similarly bioprinted cathodes.This study is one of the pioneering works on the use of bioprinting to form biocathodes for CO 2 reduction; optimized procedures and bioink composition are still to be attained in future study.The use of metal cathodes and the bioprinting of microbes combine the benefit of direct metal electrocatalysis for H2 production and the subsequent reduction of CO2 by microorganisms, using H2 as a substrate for biological reactions, improving the overall production rate.Krige et al. [16] introduced 3D bioprinting of S. ovata on carbon cloth for MES.As a continuation, in this study we used 3D bioprinting on several metal mesh cathodes.Bioprinting allows the bacteria to be localized on the cathode rather than being in suspension and slowly forming cathode biofilm at their own pace, if at all; thus, bioprinting increased the possibilities of bacterial interaction with the cathode [16].Fast start-up and bacterial localization on the cathode for immediate access to the reducing equivalents are the main advantages of bioprinting on the cathode.These advantages improved the CO2 reduction rates in our MES studies.Additionally, the encapsulated and localized bacterial cells in the cathode surface within the hydrogel of bioprinting also appeared helpful in avoiding the inhibitory effect of minor levels of oxygen penetration from the anode side.Hence, the bioprinting technique is useful to operate MES which has an O2-evolving anode for long-term operation, as diffusion limitation is created with the bioprinting technique.However, the bioprinting technique may need further improvement and adaptation for the enhancement of biocathode for CO2 reduction in MES, as there was low reproducibility in the performances of similarly bioprinted cathodes.This study is one of the pioneering works on the use of bioprinting to form biocathodes for CO2 reduction; optimized procedures and bioink composition are still to be attained in future study.

Metal degradation in MES and its effect on MES performance
Non-noble metals on exposure to air form varying thickness of oxides layers having semiconducting properties [39,46].These passivating oxide layers are also likely to play a

Metal Degradation in MES and Its Effect on MES Performance
Non-noble metals on exposure to air form varying thickness of oxides layers having semiconducting properties [39,46].These passivating oxide layers are also likely to play a major role in corrosion resistance and in electrochemical interaction between microorganisms and metals.High chloride and sulfide ion concentrations, increased temperature, and acidic pH can induce significant passive layer disintegration on metal producing metal ions [47].Adsorption of aggressive anions has been shown to react with metal ions and produce chemical dissolution; hence, more aggressive anions trigger the disintegration of a passive layer and commence pitting.Low pH of the solution instigates constant corrosion.Metallic cathodes in MES are likely to be corroded under biotic conditions [33].In our study, SEM images of some spots on a CuBP cathode show morphologies similar to pitting corrosion on the surface (Figure 3F).Stainless-steel mesh in SSBP did not show such morphologies, but SS mesh contains Ni which can be corroded in the electrolyte environment.Baudler et al. [48] have also indicated the incompatible interactions of microorganism with nickel-containing electrodes, showing the lower current density obtained with stainless steel (containing 8% to 10.5% w/w nickel) in comparison to graphite bioanodes.
Bacterial production of catalase enhances the microbial corrosion of metal surfaces due to an increased oxygen reduction current, oxidizing ferric into ferric oxide, resulting in a red color in the growth system [49].Despite the fact that the cathodic redox potential should have prevented the leaching of copper during normal operation, studies have suggested that copper might be harmful to biofilms [50].The risk of air/oxygen infiltration in the cathode chamber and probable risk of forming corrosive species can also not be eliminated.An example from Tashiro et al. [51] showed that the electrochemical conditions for water electrolysis induced the inhibitory effects of reduction products such as H 2 O 2 formed with slight exposure to oxygen which can react with the electrode and also prolonged the lag phase for bacterial growth.However, as discussed in the previous subsection, bioprinting can minimize such an inhibitory effect arising due to the oxygen penetration from anode side to some extent.Moreover, many factors may also impair the microbematerial interaction in MES.Further study can be suggested to characterize the factors that may negatively affect the performance of bioprinted metal cathodes.

Conclusions
This research demonstrates the potential of using 3D bioprinting on metals to create synthetic biofilms of S. ovata for microbial electrosynthesis.When using less corrosionresistant and less biocompatible metals such as Cu, stainless steel, the bioprinting did not enhance the microbial electrosynthesis, presumably due to the instability of the bioprint and the inhibitory effects of metal ions on the bacteria.When using bioprinted S. ovata on a corrosion-resistant and biocompatible Ti mesh cathode, a high acetate production rate of 53 ± 19 g/m 2 /d at −0.8 V vs. Ag/AgCl cathode potential was obtained.Bioprinting artificial biofilms on chemically stable metals can help to develop more efficient biocathodes with locally immobilized biofilm for large-scale microbial electrosynthesis systems.A synthetic biofilm can also improve stability, allowing continuous operation without the risk of cell washing-out.

Figure 1 .
Figure 1.Bioprint layer of S. ovata on Cu cathode (encircled part) placed inside an MES reactor.

Figure 1 .
Figure 1.Bioprint layer of S. ovata on Cu cathode (encircled part) placed inside an MES reactor.

Figure 2 .
Figure 2. (A) Profiles of electric current and cathode potentials in CuBP1 MES.(B) Profiles of acetate concentration measured in CuBP1 and CuBP2 MES, and the corresponding electricity-based acetateequivalents calculated from the recorded electric current in CuBP1.

Figure 2 .
Figure 2. (A) Profiles of electric current and cathode potentials in CuBP1 MES.(B) Profiles of acetate concentration measured in CuBP1 and CuBP2 MES, and the corresponding electricity-based acetate-equivalents calculated from the recorded electric current in CuBP1.

Figure 2 .Figure 3 .
Figure 2. (A) Profiles of electric current and cathode potentials in CuBP1 MES.(B) Profiles of acetate concentration measured in CuBP1 and CuBP2 MES, and the corresponding electricity-based acetateequivalents calculated from the recorded electric current in CuBP1.

Figure 3 .
Figure 3. (A) Photograph of annealed copper meshes.(B) SEM showing the formation of copper oxides (CuOx) layers on the annealed Cu mesh.(C) Photograph of a dry bioprinted copper cathode CuBP1 after the MES operation.(D,E) SEMs showing biofilm on CuBP1.(F) Pit formation-like observation on the CuBP1 cathode after operation of MES.

Figure 4 .
Figure 4. (A) Profiles of electric current and cathode potentials in bioprinted stainless-steel cathode MES (SSBP).(B) Profiles of measured acetate concentration and calculated acetate equivalents from the electric current in SSBP during the long-term operation.(C,D) SEMs showing the biofilm on the SSBP cathode at different magnifications.

Figure 4 .
Figure 4. (A) Profiles of electric current and cathode potentials in bioprinted stainless-steel cathode MES (SSBP).(B) Profiles of measured acetate concentration and calculated acetate equivalents from the electric current in SSBP during the long-term operation.(C,D) SEMs showing the biofilm on the SSBP cathode at different magnifications.

Figure 5 .
Figure 5. (A) Profiles of electric current and cathode potentials in TiBP MES.(B) Long-term profiles of measured acetate concentration and calculated acetate equivalents from the electric current in TiBP MES.For a comparison of performance, acetate production levels of S. ovata with carbon cloth cathodes in non-bioprinted and bioprinted MES reactors from Krige et al. [16] are also plotted, and an enlarged form of the first 10 days' acetate production is shown on the left side.

Figure 5 .
Figure 5. (A) Profiles of electric current and cathode potentials in TiBP MES.(B) Long-term profiles of measured acetate concentration and calculated acetate equivalents from the electric current in TiBP MES.For a comparison of performance, acetate production levels of S. ovata with carbon cloth cathodes in non-bioprinted and bioprinted MES reactors from Krige et al. [16] are also plotted, and an enlarged form of the first 10 days' acetate production is shown on the left side.

Figure 6 .
Figure 6.(A,B) SEMs showing the biofilm on the TiBP at different magnifications.

Figure 6 .
Figure 6.(A,B) SEMs showing the biofilm on the TiBP at different magnifications.

1 .
Bioprinting of S. ovata Relatively Increased Acetate Production from CO 2 Reduction

Figure 7 .
Figure 7. Linear sweep voltammetry (LSV) carried out at 5 mV per second scan rate on CuBP, SSBP, and TiBP with bioprinted S. ovata, showing a comparison of current densities from different cathodes with normalization to projected area.LSV plot from non-bioprinted carbon cloth cathode is provided for a visual impression of higher current densities in metal cathodes.

Figure 7 .
Figure 7. Linear sweep voltammetry (LSV) carried out at 5 mV per second scan rate on CuBP, SSBP, and TiBP with bioprinted S. ovata, showing a comparison of current densities from different cathodes with normalization to projected area.LSV plot from non-bioprinted carbon cloth cathode is provided for a visual impression of higher current densities in metal cathodes.

Table 1 .
A comparative overview of performances of S. ovata in MES with different cathode materials.

Table 1 .
A comparative overview of performances of S. ovata in MES with different cathode materials.