Bio‐Electrocatalytic Application of Microorganisms for Carbon Dioxide Reduction to Methane

Abstract We present a study on a microbial electrolysis cell with methanogenic microorganisms adapted to reduce CO2 to CH4 with the direct injection of electrons and without the artificial addition of H2 or an additional carbon source except gaseous CO2. This is a new approach in comparison to previous work in which both bicarbonate and gaseous CO2 served as the carbon source. The methanogens used are known to perform well in anaerobic reactors and metabolize H2 and CO2 to CH4 and water. This study shows the biofilm formation of those microorganisms on a carbon felt electrode and the long‐term performance for CO2 reduction to CH4 using direct electrochemical reduction. CO2 reduction is performed simply by electron uptake with gaseous CO2 as the sole carbon source in a defined medium. This “electrometabolism” in such microbial electrolysis cells depends strongly on the potential applied as well as on the environmental conditions. We investigated the performance using different adaption mechanisms and a constant potential of −700 mV vs. Ag/AgCl for CH4 generation at 30–35 °C. The experiments were performed by using two‐compartment electrochemical cells. Production rates with Faradaic efficiencies of around 22 % were observed.


Introduction
CO 2 reduction has gained high interest in the last decadeb ecause of research in the field of carbon capture and utilization (CCU), which is av iable strategy for cyclic carbon use. From ac hemical point of view,C O 2 is av aluablef eedstock for many reactions and productss uch as acids, alcohols, and gases,t hat is, acetic acid, methanol, and methane. CO 2 is av ery stable molecule, therefore, its reduction reactions require ah igh energy input because of overpotentials and multielectronr eductionsteps.
According to the standard redox potentials shown in Equations (1)- (6), the reduction of CO 2 can be tuned toward different products. Indeed, the given potentials are much more negative in reality because of overpotentials. To lower these energy barriers, several electrochemical and biological systems have been investigated to catalyze the reduction of CO 2 .P articularly,t he biological pathway of CO 2 reduction, whichu ses microorganisms or enzymes, offers ab iocompatiblea nd sustainable energy storage approach. This is particularly attractive for industry because of the use of mild reactionc onditions such as ambient temperature and pressure. Furthermore,b iocatalysts are capable of self-regeneration anda re, therefore, highly suitable for long-term performance systems without the loss of catalyst. [1][2][3][4] In particular,r eductionr eactions towarda lcohols, aldehydes, and other hydrocarbons are of high interesta sm ost of those substances can be appliedd irectly as fuels. If we consider biocatalysts for the reduction of CO 2 ,d ehydrogenase enzymes are prime candidates. In 1976, Ruschig et al. reported the application of formate dehydrogenase for the reduction of CO 2 to formate with the aid of the coenzymeN ADH. [5] Further investigations on the carboxylation of CO 2 using dehydrogenase enzymes have been presented by Aresta and Dibenedetto. [6] Sri-We present as tudy on am icrobial electrolysis cell with methanogenic microorganisms adapted to reduce CO 2 to CH 4 with the direct injection of electrons and without the artificial addition of H 2 or an additional carbon source except gaseous CO 2 . This is an ew approach in comparison to previous work in which both bicarbonate and gaseous CO 2 served as the carbon source.T he methanogensu sed are knownt op erformw ell in anaerobicr eactors and metabolize H 2 and CO 2 to CH 4 and water.T his study shows the biofilm formation of those microorganisms on ac arbon felt electrode and the long-term performance for CO 2 reduction to CH 4 using directelectrochemical reduction.C O 2 reduction is performed simply by electron uptake with gaseous CO 2 as the sole carbon source in adefined medium. This "electrometabolism" in such microbial electrolysis cells depends strongly on the potential applied as well as on the environmental conditions. We investigated the performance using different adaptionm echanisms and ac onstant potentialo fÀ700 mV vs. Ag/AgClf or CH 4 generation at 30-35 8C. The experiments were performed by using two-compartment electrochemical cells. Production rates with Faradaic efficiencies of around 22 %w ere observed.
kanth andc o-workersu sed ab io-electrochemical system (BES) with formate dehydrogenase and NADH as ac ofactor for the reduction of CO 2 to formate and further presented ad etailed study on the electrochemical application of enzymes for the generation of, for example, fuels and chemicals. [7,8] Reda et al. showedabio-electrocatalytic approachw ith the electrochemical reduction of CO 2 to formate assisted by af ormated ehydrogenase( F ate DH) enzyme without as acrificial coenzyme. [9] Recently,o ur group has shown that electrodes with such enzymes immobilized onto it can be used efficiently for the electrocatalytic generation of highera lcohols such as butanol and the reduction of CO 2 to methanol. [10,11] These nonliving biocatalysts, however,a re each single molecules that are isolated from correspondingmicroorganism strains.
In ad ifferent approach, the direct application of the living biocatalysts or microorganisms, respectively,w as investigated. In particular, living microorganisms have gained interesti n comparison to enzymes as their productsa re tunable by the choice of microorganism strains and environmental parameters. Desired products can be generated with high yield and selectivity after the adaptation is complete.
Severals tudies have been presented for the application of microorganisms in the fieldo fC O 2 conversion,b iosynthesis, and the production of biofuels with H 2 equivalents added artificially.I nt he 1980s, Kerby and Zeikus, and Sharak Genthner and Bryantp resented the growth of differentm icroorganisms utilizing CO 2 as the carbon source amongo thers. [12,13] Tanner et al. showed the generation of acetate by growing Clostridium lungdahlii with, for example, CO, H 2 ,o rC O 2 . [14] Sakai et al. presented work on ethanol and acetate generation from Moorella sp.and furtherinvestigated the influence of the pH and the activities of the corresponding dehydrogenase enzymes. [15,16] Liou et al. presented work on Clostridium strains for the generation of, for example, acetate, ethanol, and higher alcohols such as butanol from CO, CO 2 ,a nd others, such as sugars, as the growth support and carbon source. [17] Logan further discussed the application of microorganisms in an electrochemical system to establish microbialf uel cells. [18] Logan and coworkers screeneds tudies with af ocus on microbial electrosynthesis. [19] Additionally,K undiyana et al. investigated Clostridium ragsdalei for ethanolp roduction and the influence of different parameters such as pH and temperature on their performance during fermentation. [20] Tracy et al. depicted detailed pathways with all of the corresponding enzymatic steps in reactions of Clostridia. They showedt he large variety of possible chemical products and biofuels that can be obtained by using thesebiocatalysts startingf rom CO 2 . [21] Indeed all of these studies showed the properties and potential to generate valuablec arbon-based products using microorganisms. However,most of the approaches that aimed towards CO 2 conversion required fermentation processes and basically focusedo nt he metabolism of the microorganisms to generatef uels and chemicals. In our work we wanted to investigate the possibility of direct electrochemical reduction of CO 2 using such living microorganisms grown on an electrode as biocatalyst. This offers the possibility to tune the metabolisms of the microorganisms toward ac ertain product, according to the potential applied, and therefore, to increase the selectivity. Furthermore, this approach depictsamethod that opens ways for renewable energy storage as solaro rw ind energy could serve as electrical sources. Therefore, the approach presented here introduces direct electron injection into microorganisms and ac harge transfer mechanism for CO 2 reduction.
In contrast to molecular catalysts, for example, metalorganic compounds or enzymes, in which charge transfer occurs through conjugatedb onds and metal ions, charge transfer in living systems, such as microorganisms, may occur outside of the cell of the living biological systems through the outer membrane.F or microorganisms, it is proposed that bioelectrochemical reactions occur mainly because of so-called extracellulare lectron transfersw ith or without the aid of an electron shuttle:R osenbaum et al. [22] suggested three different cathodic extracellular electron transfer mechanisms for biocathodic microorganisms. In addition to ad irect electron transfer that involves c-type cytochrome electron transfer chains,t hey propose am ediated electron transfer to ap eriplasmic hydrogenaseo rad irect electron transfer that involves cytochromehydrogenase partnerships. Furthermore, Villano et al. discussed the influence of abiotic hydrogen generation on indirect extracellular electron transfer,w hich is also considered as ap ossible pathway for microbialc athodic reactions. [23] Ajo-Franklin et al. investigated charge transfer between living and nonliving organismsa nd tried to apply nanostructures for improved charge transport through cell membranes. [24,25] Bio-electrocatalytic species such as microorganisms, which are capable of direct charge transfer,t herefore, have gained interest for applications in electrochemical CO 2 reduction. This work focuseso n the conversion of CO 2 to CH 4 using hydrogenotrophicm ethanogens in am icrobial electrolysis cell. The proposed mechanism for methanogenic mixed cultures is correlated to the well-known mechanisms of the conversion of CO 2 and H 2 to CH 4 and water in anaerobic digesters. Deppenmeier et al., Shima et al., and Ferry have discussed not only the detailed enzymological pathways, which include oxidationa nd reduction reactions fore lectron transfer within the metabolism, but also the role of metabolic groups for CH 4 production from biomass with such methanogenic mixed cultures. [26][27][28][29] The metabolic pathways of these methanogensf or the conversiono f CO 2 to CH 4 can be summarized in the followingo verall reaction equation [Eq. (7)].
However,H 2 added artificially,w hich is generated before CH 4 synthesis, makes these processes unfavorable because of the high cost and energy loss from the storage and transport of H 2 .A sadifferent approach, the direct electrochemical reduction of CO 2 using microorganisms as biocatalysts withoutt he need for H 2 added artificially is desired. Thisw ould offer direct CO 2 reduction on abiocathode withoutthe need for any mediator or supplementary process such as water splitting for H 2 generation and enablet he storageo fr enewable energies in the form of fuels and chemicals.T he first study on CH 4 synthesized bio-electrochemically from CO 2 without any electron ChemSusChem 2017, 10,226 -233 www.chemsuschem.org shuttle or mediators was reported by Chenge tal. in 2009. [30] VanEerten-Jansen et al. [31] studied the performance of aC H 4producing microbial electrolysis cell (MEC)f or 188 days. The maximum energy efficiency obtained in this study was 51.3 % in ay ield test. Villano et al. presented high CH 4 production rates by using am icrobial biocathode based on ah ydrogenophilic methanogenic culture. [23] In addition, they showedt he possibility to establish biofilm reactors and the use of aC H 4producing MEC for wastewater treatment. Ab ioanode able to oxidize acetate and aC H 4 -producing biocathode were used in these studies. High acetate removalf rom the influent and efficient conversion to CH 4 was shown, and 75 %o ft he energy was captured in the resulting CH 4 gas. [32,33] Moreover,S ato and co-workersd iscussed the possible implementation of the bioelectrochemical conversion of CO 2 to CH 4 for geological storage reservoirs. Electromethanogenic CO 2 reduction can be achieved by using biocathodes based on subsurface methanogens. [34][35][36][37] Furthermore, Li et al. presented the utilizationo f electromicrobial systems for CO 2 reduction and showed the conversion of CO 2 to higher alcohols by using genetically modified Ralstonia eutropha H16. [38] Recently,J iang et al. showedab io-electrochemical approach that used am ethanogenic mixed culture for the simultaneous productiono fC H 4 and CH 3 COOH from CO 2 .T he CO 3 2À -rich mediuma nd gaseous CO 2 acted as carbon-based nutrients. They adapted microorganismst oacarbon-nutrient-only metabolism by reducing the amount of H 2 added stepwise in four cycles of 10 days each. [39] Am ore recent study on microbial electrosynthesis has been presented by Bajracharya et al. who used pure and mixed cultures for CO 2 reduction. In their investigations, they applied an assembly of graphite felt and stainless steel as the cathode for the generation of acetate and CH 4 from the conversion of bicarbonate. [40] However,a ll of those approaches werep erformed using carbonate salts dissolvedi nt he mediuma st he carbon source in place of or in addition to gaseous CO 2 .I n contrast, Bajracharya et al. also presented as tudy on ag as diffusion biocathode to provide CO 2 directly. [41] In this work we were interested in using CO 2 in its gaseous form for reduction to CH 4 .T og et an idea of the bio-electrocatalytic process, we also investigated the system in as tate without any CO 2 or bicarbonate but under inert conditions. Here we show as imilara pproacht ot he work of Jiang et al. [39] on direct electron injection into methanogenic mixed cultures and the reduction of CO 2 to CH 4 [Eq. (8)].
However,i nc ontrast to previous studies, we did not add CO 3 2À and used gaseous CO 2 only as the carbon source. This was done to preserve the possibility for investigations without any CO 2 in the system but under N 2 -saturated and, therefore, inert conditions. Furthermore,t his provides ac ontrolled supply of the carbon source and, therefore, the exact determination of the efficiency and electrochemical characterization of the system for CO 2 reduction. These investigations provedt hat CH 4 was only generated if CO 2 was added.T he approachp resented here shows the conversion of CO 2 ,a dded by purging the gas directly through the system,t oC H 4 without any other additives required. This is favorable as H 2 ,a ne xplosive gas obtained from energy-costly processes such as the steam reforming of fossil fuels or water electrolysis,can be avoided.

Results and Discussion
After the inoculation of the microorganism suspension (Figure 1a), ac onstant potentialo fÀ700 mV vs. Ag/AgCl was applied, and the cathode compartment was purged fora pproxi-mately5hper day with CO 2 and H 2 .The negative potential applied was necessary for the growth of ab iofilm as the utilized microorganisms are exoelectrogenic and can immobilize on the carbon-based electrode because of their ability to take up electrons.H owever,t he application of ac onstant negative potential is required for continuousC O 2 reduction.T he reduction potentialw as set at À700 mV vs. Ag/AgCla ccordingt ot he theoretical reduction potentialo fC O 2 to CH 4 [Eqs.
(1)- (6)]. For this, the target was to use as low an overpotential as possible for the reduction to CH 4 and to avoid competing reduction reactions such hydrogen evolution, which occurs at even lower potentials. Biofilm formation was observed after 24 ho ft he application of ac onstant negative potential ( Figure 1b). After one week, the biofilm had multiplied distinctly,a nd the nourishing medium was exchanged (Figure 1c).
During the biofilm formation,C H 4 production was monitored. Gas chromatograms during differents tates of the growing process are depicted in Figure 2. The increasing CH 4 concentration in the headspacei ndicates ac ontinuousa nd advancedproduction of CH 4 .
To characterize the biocathode, cyclic voltammetry was performed before and after the adaption was completed to investigate the redox processes associated with the microorganisms.
In the abiotic( noninoculated) MEC with ap ristine carbon felt electrode, redox peaks were not detected either with N 2 or with CO 2 /H 2 purging (Figure 3, gray line with triangles and blue line with stars). In the biotic (inoculated) nonadapted MEC ad istinct increase in the reductive current from an offset of À200 mV vs. Ag/AgCla nd ap eak value at À700 mV vs. Ag/ AgCl waso bserved. This reductive peak correlates with the reductiono fC O 2 to CH 4 at at heoretical potential of À0.446 Vv s. Ag/AgCl (À0.24 Vv s. normal hydrogen electrode (NHE); , whichi sa ssumed to be because of insufficient flushing with N 2 before the measurement and the removal of CO 2 in the cathodec ompartment, respectively.H owever,t he reduction current density is much higher in the case of CO 2 /H 2 purging. This behavior supports the expectation of electrochemically activem icroorganisms established on the carbon felt electrode.
The first adaption process was performed using the technique of the successive decrease of the amount of H 2 purged throught he system in three cycles. Each cycle consisted of 5days of purging with ac ertain ratio of CO 2 /H 2 and 2days of no purging. During the adaption process, the CH 4 production of the microorganisms on the cathode was investigated continuously by measuring headspace gas samples by using GC, and the chromatograms of each cycle are shown in Figure 4.
Even though the amounto fH 2 added was reduced continuously,C H 4 generation was not affected distinctly,a sc an be seen from the rather uniform peak and, therefore, comparable concentrationo fC H 4 during all three cycles.
The long-term performance of the microbial electrosynthesis of CH 4 was investigated for the adapted MEC. The only carbon sourcet ob er educed to CH 4 was gaseous CO 2 bubbled through the cathode compartment regularly.T he CH 4 production from long-term microbial electrolysis over 22 weeks is shown in Figure 5. The CH 4 concentration in the headspace was rather constant.
In addition to CH 4 ,H 2 generation was observed, which is expected to be produced by the microorganisms themselves. As there was as teady potential applied to thes ystem,w ew ould expect water splitting-independent of the microorganisms on the electrode-with ac onstant generation and constant    . Gas chromatograms that show long-term performance after completeda daption with CO 2 and H 2 (3 weeks). In general, the CH 4 production decreased compared to the performance immediately after adaption.H owever,C H 4 generation was ratherconstant over 22 weeks of performance despite its low concentration, as seen for the CH 4 peaksa t0 .48 min by using GC. H owever,H 2 was not generated with ap ristine carbon felt electrode at the same potential, and it is clear from the resultsp resented in Figure 5s teady H 2 evolution was not observed. We propose that the CO 2 reduction to CH 4 using the bioelectrodeo ccurs either throughadirect electrochemical reductionbecause of electron uptake or through indirect electrochemicalr eduction because of intermediate H 2 generation by the microorganisms and subsequent conversion with CO 2 to CH 4 .
Althought he performance of the microbial electrolysis cell was rather constant over several weeks according to CH 4 generation from CO 2 reduction,e fficiencies were low.T herefore, we tried improve the process. For this asecond adaption, adifferent approachu sing glucose and an enhancemento ft he biofilm on the carbon felt cathode was examined.
To improveb iofilm formation, more microorganism suspension was added to the cathode compartment.
For the second adaption process, glucose (0.1 mL saturated aqueous solution) was added to the cathodec ompartmenti nstead of reducing the H 2 purged through the cathode compartment continuously ( Figure 4). Additionally,t he cathode compartmentw as purged regularly with CO 2 .T he adaption processu sing glucose within three cycles was monitored ( Figure 6). As glucose and CO 2 served as carbon-based nutrients but the amount of glucosei nt he electrolyte medium was depleted during the three cycles,g aseous CO 2 was the remaining and only carbons ource after adaption and, therefore, the source for the reduction to CH 4 .I nc ontrastt oa daption with H 2 ,t he CH 4 generation first increased and then stabilized at as lightly lower amount. The addition of glucosep rovides as imple wayf or adaption withoutt he need for H 2 addition. This is favorable as H 2 can be avoided.
Cyclic voltammograms (CVs) were recordedf or the characterization of the final biocathode after the completion of the second adaptionw ith glucose ( Figure 7). There is an increase in the reductive currentf rom À300 mV vs. Ag/AgCl observed for the CO 2 -saturated system (dotted red line). Saturation with N 2 and experimentsw ithout biofilm did not deliver an increase in reductive current,w hich indicates that the predominant re-action of the microbiale lectrolysis is the reduction of CO 2 to CH 4 .I nc omparison to values obtained for CVs recordedf or the nonadapted state (Figure 3), current densities are lower by approximately an order of magnitude. However,t he CVs displayed in Figure 7c annotbec omparedd irectly as not only the CV conditions were changed (only CO 2 purging instead of CO 2 / H 2 )b ut also the constitution of the mixed culture and the metabolism of the microorganisms was modified after reinoculation, second adaption, and furtherw eeks of long-term performance. Nevertheless,t he electrochemical characterization of the biocathode showedt hat reduction reactions that take place are only observed if CO 2 and microorganisms were both present.
The long-term performance of the final adapted biocathode was monitored by investigating the constitution of the headspace gas samples over severalweeks. The gas chromatograms measured over 25 weeks after the adaptionw ith glucose are presented in Figure 8.
As observed previously after adaptionw ithC O 2 /H 2 ,t he CH 4 generation was rather constant during long-term performance after the glucose adaption of the biocathode.F or H 2 generation, the same effects in terms of nonconstant concentrations were observed as for the first long-term investigation ( Figure 5). It is expected that H 2 is not only produced by the microorganism but also partly used by the microorganism for its metabolism and the reduction of CO 2 to CH 4 .T he direct electrochemical reduction of CO 2 and the indirect electrochemical reduction of CO 2 using intermediate hydrogen is, therefore, not distinguishable.
The correlation of the detected amounts of H 2 and CH 4 from headspace gas samples of the cathode compartment of the MEC over that time shows that fluctuations of H 2 and CH 4 concentration were mainly parallel (Figure 9). The peak values of concentrations for both CH 4 and H 2 were reached at approximately the same time during the long-termp erformance.
Potentiostatic electrolysis for 4h at ap otentialo fÀ700 mV vs. Ag/AgCla fter biofilm improvement and completed adap-  These results show an efficient, biological approach of CO 2 reduction to av aluablef uel at rather high efficienciesa nd with gaseous CO 2 as the only carbon source.M icrobial electrosynthesis has been shown to be applicable in al ong-term and continuousr un with ar ather constantp roduction rate of CH 4 . Furthermore, the efficiency can be tuned by enhancing the biofilm formation and increasing the amount of microorganisms immobilized on the cathode. It is also expectedt hat mixed cultures could be modified with regard to their constitution and, therefore, their ability to metabolize CO 2 sufficiently to CH 4 by tuning the adaption method for aC O 2 -only process. These observations depict av ery interesting approachf or CO 2 recycling. Additionally,s uch electrochemical processes could be driven by renewable energy sources and, therefore, repre-sent an attractive, sustainable method forr enewable energy storage.

Conclusions
We showedt he utilization of methanogenic microorganisms, from digestate, for the reduction of CO 2 to CH 4 in am icrobial electrocatalytic synthesis approach. Microbial electrolysis cells offer great potential for sustainable, highly efficient, and selective CO 2 reduction.C O 2 was reduced in ap rocess in which microorganisms were grown on ac arbon-basedc athode for heterogeneous electrocatalysis. CH 4 was generated by direct electron injection into biocatalysts through ad irect or indirect electrochemical reduction pathway with Faradaic efficiencies of approximately 22 %. For this neither H 2 ,c arbonate, nor any other carbon source except gaseous CO 2 was added artificially. In addition, no mediatoro rs upplementary cofactors were required. This is new and different to previous studies, for example, that of Jiang et al.,i nw hich bicarbonate was used as ac arbon source in the electrolyte solution. [39] Therefore, they could not fully comprehend if the CH 4 generation was from bicarbonate or gaseous CO 2 flushed through the system or investigate the properties in as ystem withoutt he addition of any carbon source. Moreover,w es howed an extraordinary long-term performance of one year of continuous CH 4 and H 2 generation,w hiche ven extended earlier long-term investigations of, for example, van Eerten-Jansen et al. [31] If H 2 was produced during these electrochemical processes, it could act as as upply for the metabolism of the microorganisms. However, the direct electrochemical reduction of CO 2 and the indirect electrochemical reduction of CO 2 using intermediate H 2 are not distinguishable.
The application of living organismsi sa dvantageous because of their self-regeneration and adaptability to certain conditions. Here, we present as imple and efficient process that is suitable for al ong-term performance of several months with continuous and stable production rates.
Additionally,a ni mprovement of the biocathode was obtained and as econd more favorable adaption technique for aC O 2 -only process was investigated. Here, fort he first time, two different adaption techniques were appliedd uring ac ontinuously runninge xperiment. [12,30] We found ac onvenienti mprovement of an existing technology that will gain interest in the topics of renewable and sustainable energy,C O 2 reduction, and energy storagewith biocatalysts.

Experimental Section
Setup All experiments were performed by using at wo-compartment cell with separate anode and cathode compartments to avoid the diffusion of oxygen generated anodically and the reoxidation of products generated cathodically.B oth compartments were sealed with silicon septa to enable gas purging, sample withdrawal, and to guarantee anaerobic conditions for the microorganisms. Carbon felt (CF) with asize of 2.5 6 0.6 cm (the active area dipped in the electrolyte solution was approximately one third of the electrode, Figure 8. Gas chromatography records for the long-term performanceof MEC1 after repeated adaption with glucose. The production of CH 4 from CO 2 reduction showsaremarkable increase in comparison to results from long-termperformance before adaption with glucose. , with aP tw ire as the electrical contact served as working electrode. Pt foil was applied as the counter electrode. The potentials were applied versus aA g/AgCl reference electrode mounted in the cathode compartment. The anode compartment contained phosphate buffer of pH 7a st he electrolyte solution and was purged with N 2 regularly to prevent oxygen diffusion. For the cathode compartment, at emperature of 30-35 8Cw as chosen, and an utrient medium that consisted of phosphate buffer,v itamins, and trace elements with pH 7w as used as the electrolyte solution. The medium contained the following ingredients (per liter): KH 2 PO 4 (3 g), K 2 HPO 4 ·H 2 O( 2.5 g), NH 4 Cl (310 mg), NaCl (130 mg), trace element solution (12.5 mL), and vitamin solution (5 mL 00 mg), pyridoxine hydrochloride (10.00 mg), thiamine hydrochloride dihydrate (5.00 mg), riboflavin (5.00 mg), nicotinic acid (5.00 mg), d-calcium pantothenate (5.00 mg), Vitamin B12 (0.10 mg), p-aminobenzoic acid (5.00 mg), and lipoic acid (5.00 mg).

Enrichment of microorganisms
The source for the microbial inoculum was digestate collected from aw astewater treatment plant Asten (Austria). The digestate was centrifuged at 4000 rpm for 10 min. The supernatant was cultivated in headspace vials in an utrient medium, prepared as described above, in aH 2 /CO 2 atmosphere (4:1) at 37 8Cu nder orbital shaking.

Biofilm formationo nt he cathode
For the inoculation of the microorganisms, the system was purged with N 2 gas to achieve appropriate anaerobic conditions for the methanogenic mixed culture. The enriched microorganism suspension (9 mL, 10 vol %) was inoculated into the cell. The cathode chamber was purged with CO 2 and H 2 of ar atio of approximately 1:1, and ap otential of À700 mV vs. Ag/AgCl was applied. After the clarification of the electrolyte solution caused by the attachment of the microorganisms onto the electrode as ab iofilm, the medium was refreshed in the cathode compartment ( Figure 1).

Adaption of the biocathode
The inoculated microorganisms were first adapted by nourishing with CO 2 /H 2 and reducing the H 2 amount continuously over three cycles (1:1, 2:1, and 3:1). For every adaption step, the ratio was kept for 5days and for 5h every day followed by 2days of no purging. The medium of the compartment was refreshed every 2-3weeks to provide an appropriate nutrient solution for the microorganism. Furthermore, ap otential of À700 mV vs. Ag/AgCl was applied constantly.P roduct generation was investigated daily by measuring the gas composition of the headspace by injecting 2mLo fh eadspace gas into ag as chromatograph (Thermo Scientific, Trace GC Ultra) with ag as-tight syringe. After the long-term performance of 28 weeks, microorganism suspension was inoculated into the cell to enhance the biofilm on the cathode and to improve performance of the cell. Adaption to aCO 2 -only performance (without any H 2 purging required) was then undertaken in ad iffer-ent approach. For this, 0.1 mL of saturated glucose solution was added, and the cathode compartment was purged with CO 2 for approximately 5h per day.T he amount of glucose in the system decreased automatically because of the microorganisms themselves as glucose was metabolized to CH 4 and, moreover,b yr efreshing the medium in the cathode compartment after 2weeks to provide appropriate conditions (vitamins, trace elements) for the microorganisms.

Electrochemical characterization
The biofilm electrode was characterized electrochemically in its nonadapted state after the first inoculation, after the biocathode was improved, and after the second adaption with glucose was completed to ap rocess with CO 2 as the only carbon source. Electrochemical characterization was performed by using CV.C Vs were recorded after purging the cell with N 2 and CO 2 /H 2 (1:1) or CO 2 ,r espectively,f or comparison. Furthermore, also ab lank CF electrode was characterized by using CV under the same conditions. CVs were recorded by using aJ aissle Potentiostat-Galvanostat IMP 88 PC-R or 1030 PC.T., and electrolysis measurements were realized by using aJ aissle Potentiostat P-M 100. For CV,p otentials were swept between 0a nd À700 mV vs. Ag/AgCl with as can rate of 1mVs À1 .

Electrolysis for efficiencydetermination
Potentiostatic electrolysis at À700 mV vs. Ag/AgCl was conducted for 24 ha fter the first adaption with CO 2 and H 2 and for 4h after completed glucose adaption of the microorganisms to determine Faradaic efficiencies. CO 2 gas was the sole carbon source to be reduced sufficiently to CH 4 for both electrolysis experiments. For potentiostatic electrolysis, the cathode chamber was purged with N 2 for 1h to ensure inert conditions and, subsequently,w ith CO 2 for 2h to saturate the electrolyte solution. Headspace samples were analyzed before and after electrolysis by using GC. The charge [C] consumed during potentiostatic electrolysis was calculated from the current over time curve and correlated, by means of the eightelectron reduction of CO 2 shown in Equations (1)- (6), to the amount of CH 4 produced for the calculation of the Faradaic efficiency.