Alternate Charging and Discharging of Capacitor to Enhance the

Jun 28, 2011 - A bioelectrochemical system (BES) can be operated in both “microbial fuel cell” (MFC) and “microbial electrolysis cell” (MEC) m...
0 downloads 16 Views 2MB Size
ARTICLE pubs.acs.org/est

Alternate Charging and Discharging of Capacitor to Enhance the Electron Production of Bioelectrochemical Systems Peng Liang, Wenlong Wu, Jincheng Wei, Lulu Yuan, Xue Xia, and Xia Huang* State Key Joint Laboratory of Environment Simulation and Pollution Control, School of Environment, Tsinghua University, Beijing, 100084, P.R. China

bS Supporting Information ABSTRACT: A bioelectrochemical system (BES) can be operated in both “microbial fuel cell” (MFC) and “microbial electrolysis cell” (MEC) modes, in which power is delivered and invested respectively. To enhance the electric current production, a BES was operated in MFC mode first and a capacitor was used to collect power from the system. Then the charged capacitor discharged electrons to the system itself, switching into MEC mode. This alternate charging and discharging (ACD) mode helped the system produce 2232% higher average current compared to an intermittent charging (IC) mode, in which the capacitor was first charged from an MFC and then discharged to a resistor, at 21.6 Ω external resistance, 3.3 F capacitance and 300 mV charging voltage. The effects of external resistance, capacitance and charging voltage on average current were studied. The average current reduced as the external resistance and charging voltage increased and was slightly affected by the capacitance. Acquisition of higher average current in the ACD mode was attributed to the shorter discharging time compared to the charging time, as well as a higher anode potential caused by discharging the capacitor. Results from circuit analysis and quantitatively calculation were consistent with the experimental observations.

’ INTRODUCTION A bioelectrochemical system (BES), which consists of an anode, a cathode and a separation membrane, uses microorganisms as catalysts for reactions occurring at electrodes.1 BESs could be operated in the “microbial fuel cell” (MFC) mode, in which they deliver power, and in the “microbial electrolysis cell” (MEC) mode, in which power is invested to increase the kinetics of the reaction and/or to drive thermodynamically unfavorable reactions. Not only energy, but also a plethora of other applications in bioremediation and the production of fuels and chemicals (such as biological denitrification,2,3 desalination,46 hydrogen,7 methane,8 and caustic productions9) could be obtained by means of BES. Electric current is one of the key parameters influencing the performance of BES. For biocathode denitrification, Virdis et al. demonstrated that the MFC with higher current output could reach a higher nitrogen removal rate.10 In microbial desalination cell (MDC), the external resistance was set as low as 1 Ω in order to get a higher current, and thus a higher desalination rate was obtained.6 The BES generating a current up to 1.05 A achieved a high caustic production efficiency.9 There are several methods to increase the current density, including improvements on the internal BES as well as on the external circuit (loads). As for the former, much attention has been paid to decrease the BES’s internal resistance, including electrochemically active bacteria, applying different electrode materials, separating material and configurations.1116 With respect to the external circuit, most researchers employed a similar external circuit setup made up of resistors to consume r 2011 American Chemical Society

electrical energy.17 As research findings suggested that the current production of MFC could not be increased by only modifying the external resistance,18 Alim Dewan et al. introduced a capacitor, rather than resistors, into the external circuit and successfully collected higher electrical energy intermittently instead of harvesting energy continuously with electrical current passing through a resistor.1921 As MFCs could produce electricity during the organic matter oxidation, they are paid more attentions. Since the power generated from the MFC is limited, the expectation of in situ utilization of the electrical energy stored in the capacitor on MFC itself to enhance the current production has been taken into account. The process when the capacitor was charged from BES is similar to MFC mode, and the process when the capacitor was discharged to BES is similar to the MEC mode. In this research an electrical circuit was designed to achieve a new operation mode, in which a capacitor was first charged from a BES and then discharged directly through the BES (alternative charging and discharging mode, ACD mode). Since the charging phase of the capacitor can improve current production of an MFC, another circuit practicing intermittent charging (IC) mode was also designed as a comparison of the ACD mode.17 In this mode, a capacitor was charged from an BES and then discharged to a resistor. We compared the average current in the two modes, Received: March 5, 2011 Accepted: June 28, 2011 Revised: June 19, 2011 Published: June 28, 2011 6647

dx.doi.org/10.1021/es200759v | Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology

Figure 1. Schematic circuit diagrams of (A) alternate charging and discharging (ACD) and (B) intermittent charging (IC) modes. (Rct, the charge transfer resistance; RΩ, the electrolyte resistance; Ci, the internal capacitance; Vs, constant voltage source; Re, external resistor; Red1, Red2, external discharging resistor; Ce, Ce1, Ce2, external capacitor ; S1, S2, S3, S4, relay switches).

as well as the power density, the COD removal rate, and Coulombic efficiency. Factors affecting the performance were studied. Based on the circuit analysis and quantitative calculation, the advantages of the ACD mode were proved.

’ MATERIALS AND METHODS Reactors and Operation. Two-bottle “H” type MFCs were constructed with the bottles separated by a cation membrane (CMI7000, Membranes International Inc., U.S.). Both anode and cathode chambers were initially sealed with rubber stoppers, and stirred slowly using magnetic stir bars at 60 rpm. Four pieces of graphite felt (Sanye, China) (2  2  0.5 cm) were used as the anode and cathode, and were then connected to the circuit using graphite rods. A saturated calomel electrode (SCE, 0.242 V vs standard hydrogen electrode (SHE), Leici, China) was fitted through the rubber stopper of the anode chamber and used as a reference electrode. The feeding medium of the anode chamber contains (per liter) 1.64 g of NaAc, 1.5 g of NH4Cl, 0.6 g of KH2PO4, 0.1 g of MgCl2 3 6H2O, 0.1 g of CaCl2 3 2H2O, 0.1 g of KCl, 10 mL of trace

ARTICLE

mineral mix, pH 7.0.14 The cathode chamber was filled with a ferricyanide and phosphate buffer catholyte, containing (per liter in deionized water) 16.5 g K3Fe(CN)6, 4.4 g KH2PO4, 3.4 g K2HPO4.22 These MFCs were inoculated with mixed bacterial cultures collected from an acetate-fed MFC operated for more than six months in our laboratory. Before experiments, the MFCs were operated with a 500 Ω external resistance until a peak voltage of 530 mV was repeatedly produced. ACD and IC Operation Modes. In the ACD mode (Figure 1A), the ultracapacitor (Beijing HCC Energy Tech. Co. China, Ce) and external resistor (Re) were in series combined, and then they were connected to the anode and cathode, respectively. Two relays (MY2N-J, Omron Corporation, Japan, S1, S2) controlled by a timer (H3BA-N8H, Omron Corporation, Japan) were used to switch the capacitor from charging (S1, S2 switched to the black points) to discharging (S1, S2 switched to the red points) mode and back and forth. In the equivalent circuit, an MFC can be described as a direct current (DC) power source like a battery (the voltage Vs is a constant), an internal capacitor (Ci, represents the capacitance between the electrode and its surrounding electrolyte) and internal resistors (Rct, represents the charge transfer resistance; RΩ, the electrolyte resistance induced by cation membrane, anolyte, and catholyte).23 In the IC mode, two ultracapacitors (Ce1 and Ce2), two discharging external resistors (Red1 and Red2) and four relays (S1, S2, S3, and S4) were connected to an external resistor (Re) and an MFC as shown in Figure 1B. When Ce1 was charged by the MFC, Ce2 was discharged through R ed2 (S1, S2, S3, S4 switched to the black points). When the voltage of Ce1 was charged to the setting value, four switches controlled by timers switched to the white points automatically. Then the situation reversed: Ce1 was discharged through R ed1 and Ce2 was charged from the MFC. To explore the influence of external resistance, capacitance, and charging voltage on current production, three experiments were performed. In the first experiment, the MFCs were operated with a fixed capacitance (3.3 F) and charging voltage (300 mV) but different external resistances (the nominal resistance were 3.3 Ω, 7.3 Ω, 21 Ω, 54 Ω, 100 Ω, and 200 Ω; the actual values were 3.4 Ω, 7.3 Ω, 21.4 Ω, 54 Ω, 102 Ω, and 201 Ω), which were tested with potentiometry (galvanostatic) methods on an electrochemical workstation (Autolab PGSTAT 128N). In the second experiment, the MFCs were operated with a fixed charging voltage (300 mV) and external resistance (21.4 Ω) but different capacitances (1 F, 3.3 F, 5 F, 6.6 F, 8.3 F, and 10 F). In the third experiment, the MFCs were operated with a fixed capacitance (3.3 F) and external resistance (21.4 Ω) but different charging voltages (50 mV, 100 mV, 200 mV, 300 mV, and 500 mV). The values of Re1 and Re2 in IC mode equaled to Re plus total internal resistance. All experiments were repeated three times, and more than 20 cycles were repeated each time. The medium was replaced at every experiment. Analyses and Calculation. The voltages drop across the ultracapacitor and Re were measured and recorded using a data acquisition system (DAQ2213, ADLINK, China). The current I through the circuit was determined as I ¼ UR =Re

ð1Þ

Where Re is the external resistance (Ω) and UR is the voltage drop across the external resistor (mV). 6648

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology

ARTICLE

Figure 2. Changes in capacitor’s current (A) and voltage (B) over time in two modes (3.3 F capacitance, 300 mV charging voltage, 21.4 Ω external resistance).

The quantity of charge harvested from the MFC (Q) was calculated from the capacitor’s charging voltage: Q ¼ CU

ð2Þ

Where C and U represent the capacitance (F) and charging voltage (V), respectively. The average current (Ia) in a charging/discharging cycle was calculated with the following equation: Ia ¼ Q =T

ð3Þ

Where Q and T represent the charge harvested by MFC (C) and duration time (s), respectively. Based on the step response of a resistorcapacitor circuit (RC circuit), the voltage on the capacitor can be described with eq 4.24 As shown in Figure 1, the capacitor was charged under constant voltage resource conditions. Ut ¼ U∞ þ ðU0  U∞ Þ  et=τ

ð4Þ

Where Ut is the voltage on the capacitor at time t, U∞ is the steady-state voltage on the capacitor when the circuit becomes stabilized, U0 is the initial voltage on the capacitor and τ is the time constant of the circuit (τ is characteristic of the circuit and determines how rapidly the capacitor will discharge.).23 In the IC mode, the initial voltage on the capacitor was 0 and eq 4 could be rewritten as Ut ¼ U∞  ð1  et=τ Þ

ð5Þ

In the ACD mode, the initial voltage on the capacitor was U (see Figure 2B) and eq 4 could be rewritten as Ut ¼ U∞ þ ð U  U∞ Þ  et=τ

ð6Þ

In the IC mode, the energy generated by the MFC was dispensed to the external resistor (PIC-R) and capacitors (PIC-C), whereas in the ACD mode, the energy was only dispensed to the external resistor (PACD-R). The average power densities in the two modes were calculated with eqs 7 and 8. The cycle time (TC: duration of a charging cycle in IC mode, and TD: duration of

Figure 3. The average current in a charging/discharging cycle with (A) different external resistances (3.3 F capacitance, 300 mV charging voltage), (B) capacitances (300 mV charging voltage, 21.4 Ω external resistance), and (C) charging voltages (3.3 F capacitance, 21.4 Ω external resistance).

a charging-discharging cycle in ACD mode) could be calculated by eqs 5 and 6 when the capacitors were charged up to the charging voltage (Ut = U). Z TC I 2 Re dt Ce U 2 þ ð7Þ PIC ¼ PIC-R þ PIC-C ¼ 0 TC 2TC Z PACD ¼ PACD-R ¼ 6649

TD

I 2 Re dt

0

TD

ð8Þ

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology

ARTICLE

’ RESULTS AND DISCUSSION Charging and Discharging Process of the Capacitor Operated in ACD and IC Modes. The MFCs were started up as

described previously. After the maximum voltage reached 530 mV at an external resistance of 500 Ω and was reproducible over five cycles (each cycle lasted three days and the internal resistance was 145 Ω), they were operated with the ultracapacitor. Figure 2 showed the changes in the current and voltage on the capacitor valued 3.3 F in charging and discharging cycles when operating in the ACD and IC modes. In both modes, the charging voltages were fixed at 300 mV, the capacitors were 3.3 F, and the external resistor was 21.4 Ω. The two MFCs were operated for 24 h, and several representative cycles were shown in Figure 2. In ACD mode, the current changed from 5.44 mA to 2.06 mA, while the current in IC mode changed from 3.17 to 1.90 mA. Based on eq 3, the average current in the ACD mode was 3.34 ( 0.01 mA, substantially higher than that in the IC mode (2.55 ( 0.01 mA). Figure 2A showed a cyclic variation of the capacitor’s current in both modes. In the IC mode, each cycle time was 5.94 ( 0.06 min (from time 0 to b, in Figure 2B), containing only a charging process. This charging time was almost equal to that in the ACD mode (5.89 ( 0.08 min, from time a to c, in Figure 2B). However, there was a discharging phase in the ACD mode with a shorter duration of about 3.28 min. Considering the total electric quantity, either the capacitor generated in charging phase or released in discharging phase were the same (equaled to the charging voltage multiplied by the capacitance), a larger average current could be obtained in the ACD mode because of the shorter discharging time. Influence of External Resistance, Capacitance, and Charging Voltage on Current Production and Power Density. Based in Figure 2, the average current and power density in the charging/discharging cycle can be obtained using eqs 3, 7, and 8. Changes in average current and power density with external resistance, capacitance, and charging voltage were shown in Figure 3AC respectively. Figure 3A indicates that the average current gradually decreased with the increase of external resistance in both ACD and IC modes, and the currents obtained when operating in the ACD mode were 2232% higher than those when operating in the IC mode with selected resistances. The average power densities in the two modes increased with the external resistance first and then decreased. Figure 3B showed that the current in the ACD mode were about 814% higher than those in the IC mode with different capacitances; however, power densities in the ACD mode were lower than those in the IC mode. In addition, the currents and the power densities were slightly influenced by variation of capacitances. Figure 3C showed that the current decreased gradually when the charging voltage increased from 50 mV to 500 mV, and the currents in the ACD mode were also higher than those obtained in the IC mode. The power density in the ACD mode decreased as the charging voltage increased, while the power density in the IC mode increased initially and then decreased as the charging voltage increased. COD Removal Rates and Coulombic Efficiencies of the MFCs Operated in the ACD and IC Modes. Three MFCs with internal resistances ranging from 145 Ω to 188 Ω were tested in the experiments. The external capacitance and external resistance were 3.3 F and 21.4 Ω. The voltages on the capacitor over time were shown in Figure 4. The charging voltage was 100 mV in two modes. After operating for 24 h, the cycles charging stabilities of

Figure 4. Changes of voltage on the capacitor over the time in two modes.

Figure 5. The COD removal rates and Coulombic efficiencies of MFCs operated in ACD and IC mode with different internal resistances (the internal resistances R1, R2 and R3 were 145, 175, and 188 Ω, the open circuit voltages were 690, 755, and 760 mV, respectively, each MFC was operated in IC and ACD modes alternatively and repeated three times.).

Figure 6. Temporal change in anode potential and capacitor’s voltage ACD mode (A) and IC mode (B) (3.3 F capacitance, 300 mV charging voltage, 21.4 Ω external resistance).

the two modes were held in the same level (The charge voltage remained at 100 ( 1.2 mV). The COD removal rates and Coulombic efficiencies were shown in Figure 5. The COD removal rates in the ACD mode were about 1825% higher 6650

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology

ARTICLE

than those in the IC mode and the Coulombic efficiencies in the ACD mode were 816% higher than those in the IC mode. Change in Anode Potential of MFCs. The capacitor plate, which was wired to the anode directly in both modes, had the same potential as the anode. As a result, the anode potential varied with the capacitor plate’s potential. As shown in Figure 6, the anode potential in the two modes changed with time. It is noteworthy that the anode potential in the ACD mode varied in the range of 185 mV∼158 mV, whereas that in the IC mode varied only between 188 mV and 175 mV. Performance in the ACD mode produced a higher transient anode potential than in the IC mode. In order to explain the difference in anode potential variation in ACD and IC mode, the potential variation of all parts in an MFC during a complete charging and discharging process were shown in Figure S1 (see Supporting Information). In the anode chamber of the MFCs, electrons were liberated from substrate and passed down an electron transfer chain. Then they were transferred to the anode from a component (e.g., outermembrane protein) on the electron transfer chain.25 The potential of the “electron donor” component should be lower than the anode potential such that a current can be formed on the anode. The higher the anode potential was, the larger potential difference would exist between the anode and that terminal electron transfer component.26 Thus the ACD mode would be beneficial for electron transfer from biofilms to anode. This may be one of the reasons why MFCs in the ACD mode produced higher transient current than in the IC mode. Circuit Analysis. In the circuits of IC and ACD modes, the resistors (including internal resistor and external resistor) were wired in series while the capacitors (including internal capacitor and external capacitor) were in parallel connection, so the time constant denoted by τ could be calculated by eq 9. τ ¼ RC ¼ ðRi þ Re ÞðCi þ Ce Þ

ð9Þ

Where Ri, Re, Ci, and Ce represent the internal resistance, external resistance, internal capacitance, and external capacitance, respectively. Ci was monitored by transient response analysis to be 0.007 ( 0.001F,21 negligible compared to the external capacitor. Thus the eq 9 could be simplified as τ ¼ RC ¼ ðRi þ Re ÞCe

ð10Þ

At the same time, the time constant in the IC and ACD modes could also be obtained by fitting the charging curves of capacitor by eqs 5 and 6. We compared the time constant determined from these two approaches in both modes (detailed data were shown in Figure S2 of Supporting Information.) The average deviation of τ was lower than 4%. Theoretical Explanation to Experimental Results of Current Production and Power Density. The duration of a cycle (TC and TD) could be calculated by eqs 5 and 6. Based on the circuit analysis and the τ calculation (eq 10), the average currents in the two modes could be given by

IaIC ¼

Q ¼ TC

Ce U U ¼ U∞ U∞ τ  ln ðRi þ Re Þln U∞  U U∞  U ð11Þ

IaACD ¼

Q 2Ce U 2U ¼ ¼ U∞ þ U U∞ þ U TD τ  ln ðRi þ Re Þln U∞  U U∞  U ð12Þ

Equations 11 and 12 indicate that the current decreased with the increase of external resistance in both ACD and IC modes. In addition, the current was not influenced by external capacitors. The current decreased with the charging voltage, which is consistent with our observations. The duration of cycle (TC and TD) and τ were calculated with eqs 5 and 6, the current I could be obtained by I = Ce((dUt)/dt)) where, Ut was decided by eqs 5 and 6. The average power densities in the two modes could be calculated with eqs 13 and 14, applying TC, TD, and I into eqs 7 and 8. PIC ¼ PIC-R þ PIC-C ¼

Re Uð2U∞  UÞ   U∞ 2ðRe þ Ri Þ2 ln U∞ -U þ

U2   U∞ 2ðRe þ Ri Þln U∞  U

PACD ¼ PACD-R ¼

2Re UU∞   U∞ þ U 2 ðRe þ Ri Þ ln U∞  U

ð13Þ

ð14Þ

In the IC mode, the electricity generated from MFC was consumed by the external resistor and the capacitor (charging). When the external resistance increased, the PIC-R increased while the PIC-C decreased. This indicated why average power density in the IC mode increased at first and then decreased. When the charging voltage increased, the PIC-R decreased while the PIC-C increased. Accordingly, the PIC increased initially and then decreased as charging voltage increased. The higher external resistance or lower charging voltage could increase the percentage of PIC-R in PIC. When the external resistance increased, the PACD-R increased initially and then decreased. When the charging voltage increased, the PACD-R decreased. Equations 13 and 14 indicate that the external capacitance Ce had no effect on the average power density. Advantage and Disadvantage of ACD and IC Mode. When a capacitor is added into the external circuit of MFC, the current changes as the capacitor gets charged and discharged. In the ACD and IC modes, the capacitor’s charging or discharging processes were used to achieve the transient state. When the current is drawn at a high and constant rate, it is limited by the mass transfer of the electroactive species to the electrode, which is known in electrochemistry as the limiting current.9 The transient state could put the boundary layer proximity to the electrode in the dynamic changes, and the mass transfer limitation could be reduced and the electron production would be enhanced. As the time of discharging process is shorter than that of charging process, the boundary layer charged more frequently. This is one reason why the average current in the ACD mode was higher than that in the IC mode. In addition, the discharging process in the ACD mode could provide higher anode potential 6651

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology than that in IC mode, and thus promote electrochemically active bacteria to produce more electrons. This also contributes to a larger average current in the ACD mode. As illustrated previously, a BES could be operated in the MFC mode or in the MEC mode.1 The BES performed as the MFC mode when operating in the IC mode. In the ACD mode, the process performed as the MFC mode when the capacitor was charged, and was switched to the MEC mode when the capacitor was discharged. To obtain more electric energy from MFC, the IC mode is better than ACD mode. To obtain more current or other application, the ACD mode is better. The capacitor’s discharging phase also could be used in other applications, such as enhancing the performance of MDC and biocathode BESs.

’ ASSOCIATED CONTENT

bS

Supporting Information. Calculation methods of the quantity of electricity; the mechanism of anode potential on the electrochemically active bacteria; time constants in the two modes. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Phone: (+86-10) 62772324; fax: (+86-10) 62771472; e-mail: [email protected].

’ ACKNOWLEDGMENT This work was supported by the National High Technology Research, Development Program (863 Program) (No. 2009AA06Z306) and National Science Foundation China (NSFC No. 50908129). ’ NOMENCLATURE ACD alternate charging and discharging C capacitance (F) External capacitance in ACD mode (F) Ce ultracapacitors in IC mode (F) Ce1, Ce2 internal capacitance (F) Ci I current through the circuit (A) average current in a charging/discharging cycle (A) Ia average power density in ACD mode (W/m3) PACD power density dispensed to external resistor in PACD-R ACD mode (W/m3) average power density in IC mode (W/m3) PIC power density dispensed to capacitor in IC PIC-C mode (W/m3) power density dispensed to external resistor in IC PIC-R mode (W/m3) Q quantity of charge harvested by the MFC (C) charge transfer resistance (Ω) Rct external resistance (Ω) Re external discharging resistors in IC mode (Ω) Red1, Red2 internal resistance (Ω) Ri electrolyte resistance induced by cation membrane, RΩ anolyte and catholyte (Ω) RC circuit resistor-capacitor circuit S1, S2, S3, S4 relay switches T duration time of a charging/discharging cycle (s) duration of a charging cycle in IC mode TC

ARTICLE

TD U U0 UR Ut U∞ Vs τ

duration of a charging-discharging cycle in ACD mode (min) charging voltage on capacitor (V) initial voltage on capacitor (V) voltage drop across the external resistor (mV) voltage on the capacitor at time t (V) steady-state voltage on capacitor when the circuit becomes stabilized (V) electromotive force of MFC (V) time constant of the circuit (min)

’ REFERENCES (1) Rabaey, K.; Rozendal, R. A. Microbial electrosynthesis—Revisiting the electrical route for microbial production. Nat. Rev. Microbiol. 2010, 8, 706–716. (2) Virdis, B.; Rabaey, K.; Yuan, Z. G.; Rozendal, R. A.; Keller, J. Electron fluxes in a microbial fuel cell performing carbon and nitrogen removal. Environ. Sci. Technol. 2009, 43 (13), 5144–5149. (3) Xie, S.; Liang, P.; Chen, Y.; Xia, X.; Huang, X. Simultaneous carbon and nitrogen removal using an oxic/anoxic-biocathode microbial fuel cells coupled system. Bioresour. Technol. 2011, 102 (1), 348–354. (4) Cao, X. X.; Huang, X.; Liang, P.; Xiao, K.; Zhou, Y. J.; Zhang, X. Y.; Logan, B. E. A new method for water desalination using microbial desalination cells. Environ. Sci. Technol. 2009, 43 (18), 7148–7152. (5) Chen, X.; Xia, X.; Liang, P.; Cao, X. X.; Sun, H. T.; Huang, X. Stacked microbial desalination cells to enhance water desalination efficiency. Environ. Sci. Technol. 2011, 45 (6), 2465–2470. (6) Jacobson, K. S.; Drew, D. M.; He, Z. Efficient salt removal in a continuously operated upflow microbial desalination cell with an air cathode. Bioresour. Technol. 2011, 102 (1), 376–380. (7) Liu, H.; Grot, S.; Logan, B. E. Electrochemically assisted microbial production of hydrogen from acetate. Environ. Sci. Technol. 2005, 39 (11), 4317–4320. (8) Selembo, P. A.; Perez, J. M.; Lloyd, W. A.; Logan, B. E. High hydrogen production from glycerol or glucose by electrohydrogenesis using microbial electrolysis cells. Int. J. Hydrogen Energy 2009, 34 (13), 5373–5381. (9) Rabaey, K.; Butzer, S.; Brown, S.; Keller, J.; Rozendal, R. A. High current generation coupled to caustic production using a lamellar bioelectrochemical system. Environ. Sci. Technol. 2010, 44 (11), 4315–4321. (10) Virdis, B.; Rabaey, K.; Rozendal, R. A.; Yuan, Z. G.; Keller, J. Simultaneous nitrification, denitrification and carbon removal in microbial fuel cells. Water Res. 2010, 44 (9), 2970–2980. (11) Bond, D. R.; Lovley, D. R. Electricity production by Geobacter sulfurreducens attached to electrodes. Appl. Environ. Microbiol. 2003, 69 (3), 1548–1555. (12) Kim, N.; Choi, Y.; Jung, S.; Kim, S. Effect of initial carbon sources on the performance of microbial fuel cells containing Proteus vulgaris. Biotechnol. Bioeng. 2000, 70 (1), 109–114. (13) Logan, B. E.; Cheng, S.; Watson, V.; Estadt, G. Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environ. Sci. Technol. 2007, 41 (9), 3341–3346. (14) Scott, K.; Rimbu, G. A.; Katuri, K. P.; Prasad, K. K.; Head, I. M. Application of modified carbon anodes in microbial fuel cells. Process Saf. Environ. Prot. 2007, 85 (5), 481–488. (15) Zhang, X. Y.; Cheng, S.; Wang, X.; Huang, X.; Logan, B. E. Separator characteristics for increasing performance of microbial fuel cells. Environ. Sci. Technol. 2009, 43 (21), 8456–8461. (16) Rabaey, K.; Claowaert, P.; Aelterman, P.; Verstraete, W. Tubular microbial fuel cells for efficient electricity generation. Environ. Sci. Technol. 2005, 39 (20), 8077–8082. (17) Rismani-Yazdi, H.; Christy, A. D.; Carver, S. M.; Yu, Z.; Dehority, B. A.; Tuovinen, O. H. Effect of external resistance on bacterial diversity and metabolism in cellulose-fed microbial fuel cells. Bioresour. Technol. 2011, 102 (1), 278–283. 6652

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653

Environmental Science & Technology

ARTICLE

(18) Lyon, D. Y.; Buret, F.; Vogel, T. M.; Monier, J. M. Is resistance futile? Changing external resistance does not improve microbial fuel cell performance. Bioelectrochemistry. 2010, 78 (1), 2–7. (19) Dewan, A.; Beyenal, H.; Lewandowski, Z. Intermittent energy harvesting improves the performance of microbial fuel cells. Environ. Sci. Technol. 2009, 43 (12), 4600–4605. (20) Donovan, C.; Dewan, A.; Heo, D.; Beyenal, H. Batteryless, wireless sensor powered by a sediment microbial fuel cell. Environ. Sci. Technol. 2008, 42 (22), 8591–8596. (21) Donovan, C.; Dewan, A.; Huan, P.; Deukhyoun, H.; Beyenal, H. Power management system for a 2.5W remote sensor powered by a sediment microbial fuel cell. J. Power Sources. 2011, 196 (3), 1171–1177. (22) Clauwaert, P.; Rabaey, K.; Aelterman, P.; De Schamphelaire, L.; Ham, T. H.; Boeckx, P.; Boon, N.; Verstraete, W. Biological denitrification in microbial fuel cells. Environ. Sci. Technol. 2007, 41 (9), 3354–3360. (23) Ha, P. T.; Moon, H.; Kim, B. H.; Ng, H. Y.; Chang, I. S. Determination of charge transfer resistance and capacitance of microbial fuel cell through a transient response analysis of cell voltage. Biosen. Bioelectron. 2010, 25 (7), 1629–1634. (24) Alexander, C.; Sadiku, M. Fundamentals of Electric Circuits 3e; McGraw-Hill Companies, Inc.: New York, 2005. (25) Schroder, U. Anodic electron transfer mechanisms in microbial fuel cells and their energy efficiency. Phys. Chem. Chem. Phys. 2007, 9 (21), 2619–2629. (26) Finkelstein, D. A.; Tender, L. M.; Zeikus, J. G. Effect of electrode potential on electrode-reducing microbiota. Environ. Sci. Technol. 2006, 40 (22), 6990–6995.

6653

dx.doi.org/10.1021/es200759v |Environ. Sci. Technol. 2011, 45, 6647–6653