New architecture for modulization of membranelessand single-chambered microbial fuel cell usinga bipolar plate-electrode assembly (BEA)

Junyeong An a,b, Bongkyu Kim a, Jae Kyung Jang c, Hyung-Sool Lee b, In Seop Chang a,na School of Environmental Science and Engineering, Gwangju Institute of Science and Technology (GIST), 261 Cheomdan-gwagiro, Buk-gu,Gwangju 500-712, Republic of Koreab Department of Civil and Environmental Engineering, University of Waterloo, 200 University Avenue West Waterloo, Ontario, Canada N2L 3G1c Energy and Environment Engineering Division, National Institute of Agricultural Science, Rural Development Administration, Suwon 441-707,Republic of Korea

a r t i c l e i n f o

Article history:Received 15 November 2013Received in revised form21 February 2014Accepted 25 February 2014Available online 16 March 2014

Keywords:Microbial fuel cellMicrobial electrochemical cellMFC stackSeries connectionStacked MFC

a b s t r a c t

A new architecture for a membraneless and single-chambered microbial fuel cell (MFC) which has aunique bipolar plate-electrode assembly (BEA) design was demonstrated. The maximum power of MFCunits connected in series (denoted as a stacked MFC) was up to 22.870.13 mW/m2 for 0.94670.003 Vworking voltage, which is 2.5 times higher than the averaged maximum power density of the non-stacked MFC units. The power density in the stacked MFC using BEA was comparable to the stacked MFCusing electric wire. These results demonstrate that BEAs having air-exposed cathodes can potentially beused in the stacking of membraneless single-chambered MFCs. In addition, we confirmed that thecurrent in the stacked mode flowed faster than the non-stacked mode due to voltage increase by seriesconnection, and the poorest of the stacked units quickly faced current depletion at higher externalresistance than the non-stacked mode, leading to voltage reversal. These results imply that stacked MFCunits require a relatively large current capacity in order to prevent high voltage reversal at high currentregion. To increase total current capacity and prevent voltage reversal of stacked MFC units, wesuggested series/parallel-integrated MFC module system for scaling-up. This new concept could likelyallow the application of MFC technology to be extended to various wastewater treatment processes orplants.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Microbial fuel cells (MFCs) are promising technologies forproducing electricity from waste organic matter (Kim et al., 1999;Chang et al., 2006). For this reason, researchers have been strivingto use these devises for energy-recovering wastewater treatment.The maximum theoretical voltage from a single individual MFC is!1.1 V at pH 7 at open circuit mode, when acetate is used as theelectron donor and O2 as the electron acceptor (Logan et al., 2006).However, an achievable working voltage in a single MFC rangesfrom !0.2 to !0.5 V that is not sufficient for operating electronics(Kim et al., 2011, 2012), which typically require an input voltage ofover 1.55 V. To satisfy such voltages, MFCs could be connected inseries for voltage-up; indeed, many researchers have been trying toincrease voltage by connecting MFCs in series. Aelterman et al.(2006) demonstrated the possibility of boosting the voltage up to

2 V through a serial connection of six MFCs. Oh and Logan (2007)reported a power increase and several electrochemical phenomenaoccurring in a series connection, further promoting the practicaluse of electrical energy harvested from MFCs. Dekker et al. (2009)produced a working voltage of 3.1 V by connecting 4 MFCs in seriesat an external resistor of 1 k. In addition, Ieropoulos et al. (2013)showed a successful use of serially-connected MFCs for poweringmobile phone. They obtained an open circuit voltage of !7.2 V fora maximum working voltage of !2.15 V (current: 3 mA) in 24MFCs. Despite that these previous studies showed that MFCs couldproduce a voltage for applications by series connection, scale-up ofMFCs in series is being hindered by chambered architecture andcomponents. Most MFCs require either a closed chamber or twochambers separated by an ion exchange membrane in order toprevent organics or oxidants crossing over to a counter chamber.However, chambered configurations having a membrane showlimitations for scaling-up MFCs; for instance, the membranerequires regular cleaning when biofouling or contamination onmembrane surface occurs (Xu et al., 2012). In addition, closed two-chambered MFCs should be designed to ensure that the membrane

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/bios

Biosensors and Bioelectronics

http://dx.doi.org/10.1016/j.bios.2014.02.0630956-5663/& 2014 Elsevier B.V. All rights reserved.

n Corresponding author. Tel.: 82 2 715 3278; fax: 82 2 715 2434.E-mail address: ischang@gist.ac.kr (I.S. Chang).

Biosensors and Bioelectronics 59 (2014) 2834

Gerald

Gerald

Gerald

Gerald

Gerald

Gerald

Gerald

Gerald

www.sciencedirect.com/science/journal/09565663www.elsevier.com/locate/bioshttp://dx.doi.org/10.1016/j.bios.2014.02.063http://dx.doi.org/10.1016/j.bios.2014.02.063http://dx.doi.org/10.1016/j.bios.2014.02.063http://crossmark.crossref.org/dialog/?doi=10.1016/j.bios.2014.02.063&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.bios.2014.02.063&domain=pdfhttp://crossmark.crossref.org/dialog/?doi=10.1016/j.bios.2014.02.063&domain=pdfmailto:ischang@gist.ac.krhttp://dx.doi.org/10.1016/j.bios.2014.02.063

is not physically damaged by the high hydraulic pressure occurringin large-scale systems; even single-chambered MFCs using air-cathodes or membrane electrode assemblies (MEAs) are not freefrom these challenges. Another concern for fabricating large-scaleMFCs and their operation could be the availability of systemoptimization to minimize the internal resistance of the entiresystem. The use of one large electrode can cause MFCs to lose cellvoltage (i.e., voltage drop) because of the internal resistance andmixed potential formation (Dewan et al., 2008; Harnisch et al.,2009).

For MFCs to be applied in the energy-recovering wastewatertreatment process, both the entire configuration and unit cellsshould be scalable, modulable, and stackable in series, along withhaving a low operating cost and be easily serviceable. In this study,we employed a bipolar-plate electrode assembly to demonstrate apragmatic approach for scaling-up in order to connect MFCs inseries, along with passive air provision to the cathode; we theninterpret the overvoltages occurring in the developed economic,stacked MFC configuration.

2. Materials and methods

2.1. Microbial fuel cell construction with a bipolar-plate electrodeassembly

Fig. 1(a) and (b) presents the unique architectures used in thisstudy. In brief, the MFC system employs: (a) a bipolar-plateelectrode assembly (BEA) consisting of an anode (graphite felt)on the bottom of individual MFCs, interfaced plate (rigid graphiteplate) for bipolar function, and air-exposed cathode (graphitefelt) on the top of the MFCs to compactify the architecture andthe structure for a serial connection, and (b) an open compart-ment with no membrane in order to resolve membrane main-tenance issues and a closed single- or two-chambered configura-tion. Fig. 1(c) and (d) illustrates dimensions of the MFC unit andits components, and the series-connected units (denoted as astacked MFC). An individual MFC unit was composed of acylindrical acrylic body (5 cm inner diameter, 9 cm height), ananode (5 cm diameter, 2.54 cm thickness), a cathode (5 cmdiameter, 2.54 cm thickness), and a rigid waterproof graphiteplate (thickness 0.2 cm, diameter 5 cm) on the bottom of the MFCto inhibit water leakage and air influx to the anode. A part of thecathode (2.04 cm) was exposed to air, while the other part of thecathode (0.5 cm) was immersed in medium. By directly connect-ing three MFC units with no external wires, we built a stackedMFC (Fig. 1(d)). A basic description of the stacked MFC config-uration is as follows: the anode at the top of the waterproofgraphite plate in the MFC unit (e.g., upper MFC unit (Fig. 1(c)) isstacked on the cathode in another MFC unit (e.g., lower MFC unit(Fig. 1(e)), which is connected in series in a BEA structure(Fig. 1(a)). The structure of the BEA component is a key compo-nent in this study, which allows for the vertical modularization ofmembraneless air-cathode MECs in series without disturbinganode respiring bacteria (ARB) (Lee et al., 2010) metabolism,due to the presence of a rigid waterproof graphite plate betweenthe anode and cathode.

2.2. Microbial fuel cell inoculation and operation

We inoculated each MFC unit with 157 ml of anaerobic diges-tion sludge, obtained from a brewery wastewater treatment plant(Gwangju, Korea), and fed an acetate (10 mM) medium (An et al.,2011) into the system at a flow rate of 0.07270.8 ml/min(hydraulic retention time (HRT): 36.3573.27 h) using a peristalticpump (Cole-Parmer Instrument Co., East Bunker Court Vernon

Hills, IL, USA). Each unit was operated under closed circuit mode ata 10 external load until the voltage became stable (in !4 days)in order to establish a viable microbial community in the anodecompartment.

2.3. Electrochemical analyses

After confirming the steady voltage output in the MFC, wechanged MFC operation to open circuit mode and conducteddischarging tests for each individual MFC unit (denoted as non-stacked units 1, 2, and 3). We varied the external loads towardlower external resistance from 100 k to 1 k (e.g., 100 k,40 k, 30 k, 20 k, 15 k, 10 k, 5 k, and 1 k) in order.The discharging time for each load was 15 min, and the voltageswere monitored every 10 s using a multimeter (Keithley 2700,Keithley Co., Cleveland, OH, USA). During the tests the voltage inthe non-stacked units was stabilized in 10 min at each externalload. When stabilized the voltage was collected at 30 points for5 min to compute the current density and the power density usingjV/(RA) and P IV/A, respectively, where j is the current density(mA/m2), V is the voltage (V) in individual units, R is the externalload applied to the units, A is the surface area (m2) of the anodeelectrode, P is the power density (mW/m2) in the units, and I is thecurrent (mA) in the units. We averaged 30 points with standarddeviation. In the same manner, we then fully recharged the non-stacked units, converted the units to the stacked MFC mode, andcarried out discharging tests for the stacked MFC. To assess theBEA performance in the stacked MFC, for comparison, dischargingtests were conducted by connecting non-stacked units in seriesusing electrical wires. We then extracted voltagecurrent density(Vj) curves, and power densitycurrent density (Pj) curves wereobtained from the discharging experiments. The internal resis-tances of the non-stacked units (i.e., non-stacked mode) and thestacked MFC (i.e., stacked mode) were subsequently analyzed toassess factors affecting overvoltages in the stacked MFC. Next, theinternal resistances were estimated from the Nyquist plotsobtained via electrochemical impedance spectroscopy (EIS) spec-tra using AutoLab (Eco Chemie, Utrecht, The Netherlands), whichwere combined with an impedance analyzer module (PGSTAT 30with FRA-ADC, Eco Chemie, Utrecht, The Netherlands). EIS spectrawere measured over a frequency range of 100 kHz0.01 Hz underalternating voltage conditions at the cathode in open circuitpotential, at a 710 mV amplitude in a two electrode configuration(An et al., 2011). Then, EIS data were fitted to an equivalent Randlecircuit model to estimate the charge transfer resistance (Rct) andsolution resistance (Rs) for the anode and the cathode of theindividual units.

3. Results and discussion

3.1. Performance of stacked MFCs with BEA

Fig. 2 shows Vj and Pj curves obtained from the non-stackedunits and the stacked MFC. The OCVs of the non-stacked unitswere 0.53 V, 0.51 V, and 0.53 V; the OCV of the stacked MFC tripleat 1.58 V. The maximum power density obtained from the stackedMFC was 22.870.13 mW/m2 for 0.94670.003 V working voltage,which is 2.5 times higher than the averaged maximum powerdensity of the individual MFC units (Fig. 2(A) and (B)). From theseresults, we can see that the BEA structure is applicable tomembraneless MFCs for connection in series.

As the external resistance was lowered, the voltage of thestacked MFC linearly decreased between the OCV and 15 kcondition. The magnitude of the voltage drop in the stacked MFCwas 0.77470.004 V between the OCV and 15 k, which was

J. An et al. / Biosensors and Bioelectronics 59 (2014) 2834 29

Gerald

Gerald

Gerald

Gerald

Gerald

Gerald

much larger than for the non-stacked MFCs (see Fig. 2(A)). Thisovervoltage increase in the stacked MFC mode could probably bedue to an increase of Rct and Rs by stacking the MFCs (An et al.,2011)the total internal resistance of a stacked MFC theoreticallyincreases as much as the sum of internal resistances in stackedunits (Pain, 1996).

To identify whether the overvoltage of 0.77470.004 Vbetween the OCV and 15 k was caused by the increase of Rctand Rs, Nyquist plots were obtained via EIS measurements(Fig. 3). As expected, the anode Rct, cathode Rct, and Rs of thestacked MFC were almost the same as the sum of the anode Rct,cathode Rct, and Rs of the three non-stacked units (Table 1).Among these values, the anode Rct in the stacked MFC was muchsmaller than Rs and the cathode Rct; from these results, it wasidentified that Rs and the cathode Rct in the stacked MFC werethe most critical resistancesaccounting for 96.8% (Rs: 47.9%,cathode Rct: 48.9%) of the total internal resistance. Note thatRs includes an electronic resistance (RE) and ionic resistance(RI) (Rismani-Yazdi et al., 2008); hence, there could be twopossibilities for such a high Rs in the stacked MFC: (1) high RE

occurring in the BEA, or (2) high RI occurring in the solutionbetween the anode and cathode of each stacked unit. It isposited here that if conduction is insufficient in the BEA,resulting from the incomplete contact of rigid graphite to theelectrodes, Rs could thus be increased. In order to confirm whichresistance dominantly affected the Rs of the stacked MFC, the REof the BEAs was analyzed using Nyquist plots. The RE was 1.3between units 1 and 2 and 1.6 between units 2 and 3 of thestacked MFC, corresponding to !1% ([(1.31.6)/303.46]$1000.95%) of the total Rs (EIS data not shown).

In addition to EIS measurements, we determined the perfor-mance of another type of a stacked MFC that connects an anodewith a cathode using Ptcopper wires (Fig. 4). As a result, wefound that the power density of the stacked MFC using BEA wascomparable to the stacked MFC using these wires, indicating thatBEA architecture could be a potent alternative for modulization ofmembraneless and single-chambered MFCs (Fig. 4). Based onthese findings, we thus confirmed that BEAs having air-exposedcathodes can potentially be used in the stacking of membranelessMFCs. In addition, it was confirmed that RI and the cathode Rct are

Fig. 1. Dimensions and architecture of membraneless MFC units; the series connection of the units (stacked MFC) using a bipolar plate-electrode assembly (BEA).

J. An et al. / Biosensors and Bioelectronics 59 (2014) 283430

Gerald

a main source of the overvoltage that occurs between the OCV and15 k external resistance conditions. It should be noted that theovervoltage (i.e., voltage-drop) by RI and cathode Rct is a part of theentire overvoltage incurred throughout the 1001 k range in thestacked MFC.

3.2. Implication of voltage drop at high external resistancein stacked MFC

Another significant overvoltage of 0.73570.003 V in the stackedMFC also occurred between the external resistors of 15 k and 1 k,through which the voltage was gradually decreased and becamesimilar to that in the non-stacked unit 2 (Fig. 2(A)). In addition, thelinearity of the voltage drop of the stacked MFC between 10 k and1 kwas steeper than for the non-stacked units (Fig. 2(A)). Hence, itcould be thought that this overvoltage was caused by a mass transferlimitation. However, it is recognized that this interpretation may besomewhat controversial because in this region the overvoltages innon-stacked units were not really significant compared to thestacked MFC, even at 1 k (Fig. 2(A)). In contrast, it was distinctlyobserved that a significant overvoltage of 0.73570.003 V in thestacked MFC occurred between 10 k and 1 k, through which afaradaic current flows. At 1 k, the voltage of the stacked MFC was0.07370.001 V, similar to that of unit 2 (0.07870.004 V); thecurrent densities in the stacked MFC and unit 2 were 37.170.5 mA/m2 and 39.570.4 mA/m2, respectively. These results indicatethat the performance of the stacked MFC was even worse than thatof unit 2, which showed the worst performance of the threeindividual MFC units. The reason for the lower performance in thestacked MFC could be explained by a voltage reversal phenomenon(An and Lee, in press) and a circuit theory in which the amount ofcurrent is proportional to the voltage applied when electrochemicalcells are connected in series (Pain, 1996). Fig. 5 presents voltageevolution in the stacked MFC and each unit in the stacked MFC(denoted as stacked units 1, 2, and 3) between 100 k and 1 k. Asseen in the figure, the voltage of the stacked unit 2 reached !0 V at10 kwhere units 1 and 2 maintain positive voltages of !0.223 and0.321 V, respectively. At the next external resistance of 5 k, thevoltage in the stacked unit 1 dropped to %0.066 V, implying theoccurrence of voltage reversal. Voltage reversal is a well-knownphenomenon in the research field of chemical fuel cells and MFCs,and elucidates lowering power density in stacked MFCs (Oh andLogan, 2007; An and Lee, 2014). In Fig. 2(A) it is obvious that thecurrent and the voltage in the stacked MFC are larger than thosein the non-stacked units at the external resistances below 10 k(see arrows in the figure), indicating that the current in the stacked

Fig. 2. (A) Vj curves obtained from non-stacked units 1, 2, 3, and the stacked MFCusing 100 k, 40 k, 30 k, 20 k, 15 k, 10 k, 5 k, and 1 k external loads and(B): Pj curves obtained from non-stacked units 1, 2, 3, and the stacked MFCwith BEA.

Fig. 3. (A) Nyquist plots obtained from the non-stacked units 1, 2, 3, and (B) thestacked MFC.

Table 1Internal resistance analyzed from Nyquist plots.

Internal resistance () Non-stacked unit 1 Non-stacked unit 2 Non-stacked unit 3 Non-stacked unit 123 Stacked MFC

Anode Rct 5.2170.11 10.9670.68 6.3970.05 22.4770.84 20.4670.26Cathode Rct 109.3373.66 114.4875.27 102.8679.40 326.67718.33 310.00711.78Rs 83.6470.00 139.0970.01 86.0170.01 308.7470.02 303.4670.01Sum 198.1873.76 264.5575.94 195.2779.45 658.00719.15 633.93712.05

Fig. 4. Comparison of stacked MFC performances using BEA and Ptcopper wires.

J. An et al. / Biosensors and Bioelectronics 59 (2014) 2834 31

MFCs traveled much faster than in the non-stacked units at theexternal resistances below 10 k due to the voltage increasebecause of the series stacking, and that the stacked unit 2 quicklyfaced current depletion resulting in voltage reversal at 5 k (seeFigs. 2(A) and 5). Note that the VI curves of both the stacked andnon-stacked MFC units were plotted in order to obtain the directionof low to high current (i.e., external resistance was high to low).Apparently, the stacked MFC units worked in terms of current andvolt boosting when the external resistance was higher than 5 k,indicating that the resistance of 5 k is a threshold at which thestacked MFC cannot boost power and can cause voltage reversalwhen connected in series (see Figs. 2 and 5). However, the thresholdresistance may be changed by the number of stacked units instacking mode or their current capacity. For instance, if the numberof stacked units is increased, there will be a larger current flow in thestacked units due to the voltage increase. As a result, the threshold

Fig. 5. Evolutions of voltage to external loads in stacked units and stacked MFCobtained from discharging tests.

Fig. 6. Series/parallel-integrated MFC module system for scaling-up.

J. An et al. / Biosensors and Bioelectronics 59 (2014) 283432

resistance could be raised to over 5 k. In contrast, if the stackedunits have a relatively larger current capacity, the threshold externalresistance could become smaller and thus the stacked MFC wouldboost the voltage and current even at an external resistance smallerthan 5 k. Hence, the number of stacking units and their currentcapacity should be considered critical parameters for designing astacked MFC system.

3.3. Conceptual structure of series/parallel-integrated MFC moduleand system upgrade

In this paper, we developed a new MFC stacking configurationusing a BEA structure and then assessed the overvoltage occurringin the stacked MFC system. We found that the current in thestacked MFCs traveled much faster than in the non-stacked unitsuntil 10 k due to the voltage increase resulted from the seriesstacking, and that the stacked unit 2 quickly faced currentdepletion leading to voltage reversal at 5 k (see Figs. 2(A)and 5). Recently An and Lee (2014) interpreted that voltagereversal is a result of electrical energy transition from a powerfulMFC to a weak MFC of stacked MFC units, leading to significantanode potential increase over its cathode potential or vice versa. Itis well known that anode potential regulates ARB activity andcommunity (Aelterman et al., 2008; Torres et al., 2009). Torreset al. (2009) performed the comparison study for ARB character-istic at different anode potentials using aerobic activity sludge asinoculum, and they demonstrated that ARB grown at anodepotentials of %0.15 V and %0.09 V produced approximately 17times higher current compared to that imposed at an higherpotential of 0.37 V. They also showed that the anode microbialcommunity imposed at %0.15 V was 99% similar to Geobactersulfurreducens (G. sulfurreducens), but at the higher anode potentialof 0.37 V the anode has diverse microbial community comprisedof Bacteriodetes, actinoacteria, - and -proteoacteria (Torres et al.,2009). Accordingly increase in anode potential due to voltagereversal could result in change of anode ARB activity and commu-nity, which might cause substantial energy loss. Hence, the preven-tion of voltage reversal in our BEA-equipped MFC system is mostcritical for long-term operation and high voltage output. Impor-tantly, with this BEA structure, our configuration is also designed fora parallel connection which can prevent voltage reversal (Ledezmaet al., 2013; Ieropoulos et al., 2013). The terminal graphite plateillustrated in Fig. 6(a) functions as a bridge to connect the anodeends of stacked MFCs; the cathodes of the top MFC units of thestacked MFCs can be connected using a rigid graphite ring bridge(or metal ring bridge) (Fig. 6(b)). It is posited here that a parallelconnection may not only help in increasing the current capacity butalso prevent voltage reversal by balancing voltages among seriallyconnected BEA modules (Ieropoulos et al., 2013).

Despite the successful current and voltage boosting obtainedusing a series connection in this study, it is not yet sufficient foroperating electric devices or machines, thus indicating that the MFCsystem should be upgraded. There are several options for the MFCupgrade; as one, the performance of stacked MFC systems can beimproved by employing cathode catalysts that lowers the activationenergy of the oxygen reduction reaction (ORR). Note that our MFCunits did not use any metal catalysts for the cathodes, such as Pt/Cand carbon nanotubes, which are commonly used in ORRs (Qiaoet al., 2010; Cheng et al., 2006); furthermore, there was no chemicalpretreatment for the anode or cathode. As such, we expect that thecurrent and power density can be substantially improved in ourstacked MFC configuration by employing metal catalysts and anelectrode pretreatment.

BEA structure is designed to vertically stack MFCs, thus the useof a membrane might cause trapping of biogas such as CH4 or H2which can interferes with ion mass transport (An et al., 2009).

To eliminate the possibility of biogas trapping in the BEA structure,we did not employ a membrane in our system. However, becauseoxygen invasion from the cathode to the anode was an issue in ourmembraneless BEA architecture, we kept the distance between theanode and cathode at !9 cm such that aerobic heterotrophicbacteria can capture the oxygen diffusing to the anode, althoughthe increased distance led to a significant increase in solutionresistance (An et al., 2009, 2010). Despite of the concern of oxygeninvasion from the cathode to the anode in the membraneless BEAsystem, DO concentration of the catholyte in contact with thebottom side of the cathode (1.27 cm depth from solution surface)was almost 0 mg/L, indicating that the electrodes can be posi-tioned as close as possible (41 cm) to minimize solution resis-tance. However, DO concentration in a solution was dependent onorganic strength (An et al., 2010). If COD at the cathode was keptless than 40 mg/L, oxygen invasion from the cathode to the anodecould become significant and could inactivate ARB (An et al., 2010).On the other hand, if the COD was increased over 40 mg/L, the BEAcathode in direct contact with an organic rich solution (medium)with no membranecould suffer biofouling with aerobic hetero-trophic bacteria. This can lower the cathodic kinetic activity due toexcessive oxygen consumption on the cathode and mixed poten-tial formation (An et al., 2009, 2010, 2011). In this case, coating thebottom side of the cathode that is in contact with the solution withsilver-nanoparticles could be an effective solution to increase theDO concentration depleted by aerobic heterotrophic bacteria onthe cathode (An et al., 2011). Since silver-nanoparticles would theninhibit growth of aerobic heterotrophic bacteria on the bottomside of the cathode, a cathode immersed in this solution couldefficiently react to oxygen in the solution (An et al., 2011).

The concept suggested in this paper is relatively simple andcost effective because using an air-exposed cathode (passiveprovision of oxygen to the cathode) with no membrane enablesthe development of a compact MFC system to be modularized in aseries connection without requiring external wires. As such, thisnew concept might allow the application of MFC technology to beextended to large-scale wastewater treatment processes.

4. Conclusions

Our stacked MFC with BEA produced a maximum powerdensity of 22.870.13 mW/m2 for 0.94670.003 V working voltage,which is 2.5 times higher than the averaged maximum powerdensity of the non-stacked MFC units. The power density of thestacked MFC using BEA was comparable to the stacked MFC usingthe wires. Hence, BEA architecture could be a potent alternativefor stacking membraneless single-chambered MFCs vertically.

The current in the stacked mode flowed much faster than thenon-stacked mode until 10 k external load due to voltageincrease by series connection. For this reason, the poorest of thestacked units quickly faced a significant overvoltage at higherexternal resistor, resulting in voltage reversal. Hence, to preventthe high overvoltage and resulted voltage reversal, stacked MFCunits should have a relatively high current capacity. This could becrucial and informative to modeling stacked MFCs.

Acknowledgments

This work was supported by National Research Foundation(NRF) (NRF-2013R1A2A2A03014551), the Pioneer Research Centerfor Nano-Morphic Biology Energy Conversion and Storage (2009-0082812), and the Research Program for Agricultural Science &Technology Development (Project nos. PJ008517 and PJ008517032013) of the Korean government.

J. An et al. / Biosensors and Bioelectronics 59 (2014) 2834 33

Gerald

Gerald

References

Aelterman, P., Rabaey, K., Pham, H.T., Boon, N., Verstraete, W., 2006. Environ. Sci.Technol. 40, 33883394.

Aelterman, P., Freguia, S., Keller, J., Verstraete, W., Rabaey, K., 2008. Appl. Microbiol.Biotechnol. 78, 409418.

An, J., Moon, H., Chang, I.S., 2010. Environ. Sci. Technol. 44, 71457150.An, J., Jeon, H., Lee, J., Chang, I.S., 2011. Environ. Sci. Technol. 45, 54415446.An, J., Kim, D., Chun, Y., Lee, S.J., Ng, H.Y., Chang, I.S., 2009. Environ. Sci. Technol. 43,

16421647.An, J., Lee, H.S., 2014. ChemSusChem http://dx.doi.org/10.1002/cssc.201300949 in

press.Chang, I.S., Moon, H., Bretschger, O., Jang, J.K., Park, H.I., Nealson, K.H., Kim, B.H.,

2006. J. Microbiol. Biotechnol. 16, 163177.Cheng, S., Liu, H., Logan, B.E., 2006. Environ. Sci. Technol. 40, 364369.Dekker, A., Ter Heijne, A., Saakes, M., Hamelers, H.V.M., Buisman, C.J.N., 2009.

Environ. Sci. Technol. 43, 90389042.Dewan, A., Beyenal, H., Lewandowski, Z., 2008. Environ. Sci. Technol. 42,

76437648.Harnisch, F., Wirth, S., Schroder, U., 2009. Electrochem. Commun. 11, 22532256.Ieropoulos, I.A., Ledezma, P., Stinchcombe, A., Papaharalabos, G., Melhuish, C.,

Greenman, J., 2013. Phys. Chem. Chem. Phys. 15, 1531215316.

Kim, B.H., Kim, H.J., Hyun, M.S., Park, D.H., 1999. J. Microbiol. Biotechnol. 9, 127131.Kim, D., An, J., Kim, B., Jang, J.K., Kim, B.H., Chang, I.S., 2012. ChemSusChem 6,

10861091.Kim, Y., Hatzell, M.C., Hutchinson, A.J., Logan, B.E., 2011. Energy Environ. Sci. 4,

46624667.Ledezma, P., Stinchcombe, A., Greenman, J., Ieropoulos, I., 2013. Phys. Chem. Chem.

Phys. 15, 22782281.Lee, H.S., Vermaas, W.F.J., Rittmann, B.E., 2010. Trends Biotechnol. 28, 262271.Logan, B.E., Hamelers, B., Rozendal, R., Schrder, U., Keller, J., Freguia, S., Rabaey, K.,

2006. Environ. Sci. Technol. 40, 51815192.Oh, S.E., Logan, B.E., 2007. J. Power Sources 167, 1117.Pain, R., 1996. Practical Electronic Fault Finding and Troubleshooting. Oxford;

Boston: NewnesQiao, Y., Bao, S.J., Li, C.M., 2010. Energy Environ. Sci. 3, 544553.Rismani-Yazdi, H., Carver, S.M., Christy, A.D., Tuovinen, I.H., 2008. J. Power Sources

180, 683694.Torres, C.I., Krajmalnik-Brown, R., Parameswaran, P., Marcus, A.K., Wanger, G.,

Gorby, Y.A., Rittmann, B.E., 2009. Environ. Sci. Technol. 43, 95199524.Xu, J., Sheng, G.P., Luo, H.W., Li, W.W., Wang, L.F., Yu, H.Q., 2012. Water Res. 46,

18171824.

J. An et al. / Biosensors and Bioelectronics 59 (2014) 283434

http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref1http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref1http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref2http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref2http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref3http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref4http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref5http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref5http://dx.doi.org/10.1002/cssc.201300949http://dx.doi.org/10.1002/cssc.201300949http://dx.doi.org/10.1002/cssc.201300949http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref7http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref7http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref8http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref9http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref9http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref10http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref10http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref12http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref13http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref13http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref14http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref15http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref15http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref16http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref16http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref18http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref18http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref19http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref20http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref20http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref22http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref23http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref23http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref24http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref25http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref25http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref26http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref26http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref27http://refhub.elsevier.com/S0956-5663(14)00151-1/sbref27

New architecture for modulization of membraneless and single-chambered microbial fuel cell using a bipolar...IntroductionMaterials and methodsMicrobial fuel cell construction with a bipolar-plate electrode assemblyMicrobial fuel cell inoculation and operationElectrochemical analyses

Results and discussionPerformance of stacked MFCs with BEAImplication of voltage drop at high external resistance in stacked MFCConceptual structure of series/parallel-integrated MFC module and system upgrade

ConclusionsAcknowledgmentsReferences