Separator Characteristics for Increasing Performance of Microbial Fuel

Environ. Sci. Technol. 2009, 43, 8456–8461

Separator Characteristics for Increasing Performance of Microbial Fuel Cells XIAOYUAN ZHANG,† SHAOAN CHENG,‡ XIN WANG,† XIA HUANG,† AND B R U C E E . L O G A N * ,‡ State Key Joint Laboratory of Environment Simulation and Pollution Control, Department of Environmental Science & Engineering, Tsinghua University, Beijing 100084, P.R. China, Department of Civil & Environmental Engineering, 231Q Sackett Building, Pennsylvania State University, University Park, Pennsylvania 16802, and State Key Laboratory of Urban Water Resource and Environment, No. 73 Huanghe Road, Nangang District, Harbin 150090, China

Received June 3, 2009. Revised manuscript received August 25, 2009. Accepted August 25, 2009.

Two challenges for improving the performance of air cathode, single-chamber microbial fuel cells (MFCs) include increasing Coulombic efficiency (CE) and decreasing internal resistance. Nonbiodegradable glass fiber separators between the two electrodes were shown to increase power and CE, compared to cloth separators (J-cloth) that were degraded over time. MFC tests were conducted using glass fiber mats with thicknesses of 1.0 mm (GF1) or 0.4 mm (GF0.4), a cation exchange membrane (CEM), and a J-cloth (JC), using reactors with different configurations. Higher power densities were obtained with either GF1 (46 ( 4 W/m3) or JC (46 ( 1 W/m3) in MFCs with a 2 cm electrode spacing, when the separator was placed against the cathode (S-configuration), rather than MFCs with GF0.4 (36 ( 1 W/m3) or CEM (14 ( 1 W/m3). Power was increased to 70 ( 2 W/m3 by placing the electrodes on either side of the GF1 separator (single separator electrode assembly, SSEA) and further to 150 ( 6 W/m3 using two sets of electrodes spaced 2 cm apart (double separator electrode assembly, DSEA). Reducing the DSEA electrode spacing to 0.3 cm increased power to 696 ( 26 W/m3 as a result of a decrease in the ohmic resistance from 5.9 to 2.2 Ω. The main advantages of a GF1 separator compared to JC were an improvement in the CE from 40% to 81% (S-configuration), compared to only 20-40% for JC under similar conditions, and the fact that GF1 was not biodegradable. The high CE for the GF1 separator was attributed to a low oxygen mass transfer coefficient (kO ) 5.0 × 10-5 cm/s). The GF1 and JC materials differed in the amount of biomass that accumulated on the separator and its biodegradability, which affected long-term power production and oxygen transport. These results show that materials and mass transfer properties of separators are important factors for improving power densities, CE, and long-term performance of MFCs.

* Corresponding author phone: (1)814-863-7908; e-mail: blogan@ psu.edu. † Tsinghua University. ‡ Pennsylvania State University. † State Key Laboratory of Urban Water Resource and Environment. 8456

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 21, 2009

Introduction Microbial fuel cells (MFCs), which are devices that can generate electricity from biomass using bacteria (1), have drawn increasing attention as a promising technology for wastewater treatment (2-4). Single-chamber, air cathode MFCs have the greatest potential for practical applications because of their simple design and the fact that no chemical regeneration is needed when using air (5). The main challenges for constructing practical MFCs include an increase in power and recovery of electrons from the substrate (Coulombic efficiency, CE), and reducing the cost of materials. One method to increase power is to reduce the electrode spacing as this decreases internal resistance (6). When the electrodes get too close, however, power output can be reduced as a result of oxygen utilization by bacteria on the anode (7). To overcome this adverse effect of oxygen, Fan et al. (5) placed a cloth separator (J-cloth, JC) between the electrodes and avoided substantial decreases in power per surface area due to oxygen intrusion. Placing a JC separator next to the cathode initially reduced the volumetric power density from 80 W/m3 to 71 W/m3 compared to an MFC lacking a JC separator, but power was increased by placing the electrodes on either side of the separator (single separator electrode assembly, SSEA) and by using two closely placed electrodes in a reactor with a small volume (double separator electrode assembly, DSEA). Power densities were substantially improved using a cloth separator; however, the CE remained low (24% under 1k Ω) with a single cloth separator (5). CE was increased further by using multiple layers of cloth, presumably due to reduction in oxygen transfer to the anode. A main limitation of the JC material, however, is that it is biodegradable, so over time the cloth is completely degraded in the reactor, effectively removing the separator (Figure 1S, Supporting Information). Thus, nonbiodegradable separators are needed to replace JC in MFCs. Studies by Fan et al. (5) and others have shown that separator characteristics are very important to the design of an MFC, but so far there has been little systematic examination of the effects of separator characteristics on power production or CE. An ideal separator material has a high proton transfer coefficient to ensure that the material does not inhibit protons from reaching the cathode and a low oxygen transfer coefficient to improve CE, and it must be relatively nonbiodegradable. Cation exchange membranes (CEMs) such as Nafion (8), anion exchange membranes (AEM) (9), and ultrafiltration membranes (9) have been used in several different types of MFCs. While these membranes improve the CE by inhibiting oxygen transfer, CEM and AEM membranes substantially reduce power production as a result of pH gradients that develop across the membrane (10) and by an increase in internal resistance, and UF membranes substantially increase internal resistance by inhibiting proton transfer. Several other separator materials have been used, including nylon, cellulose, and polycarbonate filters (11), but their effects on power generation and oxygen transfer have not been well-examined. Another factor that can affect power generation is the growth of bacteria on the cathode. As the biofilm develops on the cathode, power production decreases over time as the biofilm gets thicker and increasingly blocks proton transport to the cathode (12). Bacteria on the cathode consume oxygen and substrate and lower the CE, although over time a thicker biofilm can also hinder oxygen diffusion 10.1021/es901631p CCC: $40.75

 2009 American Chemical Society

Published on Web 09/10/2009

FIGURE 1. Schematic of MFCs in different configurations: NS, “control” without separator and 2 cm electrode spacing; S, separator against cathode and a 2 cm electrode spacing; SSEA, single separator electrode assembly; DSEA2, double SEA with anodes spaced 2 cm apart; and DSEA0.3, double SEA with anodes spaced 0.3 cm apart. into the anode chamber and improve the CE. The effect of biofilm growth on (and in) separators has not been previously considered. Glass fiber mats are commonly used in lead acid batteries as separators and have the additional advantage of being nonbiodegradable. In this study, we examined the use of glass fiber separators having thicknesses of 1 mm (GF1) or 0.4 mm (GF0.4) in MFCs with several different architectures. Performance in terms of volumetric power density and CE with the GF separators was compared to that of J-cloth (JC) and a cation exchange membrane (CEM) in order to identify the specific factors affecting performance such as oxygen transfer.

Materials and Methods MFC Reactors. Anodes were ammonia gas-treated (13) carbon cloth (E-Tek, type A, nonwet proofing, BASF Fuel Cell, Inc., NJ) with 7 cm2 projected area. Cathodes were made of carbon cloth (E-Tek, type B, 30% wet proofing, BASF Fuel Cell, Inc., NJ) with 0.5 mg/cm2 platinum (7 cm2 projected area) and 4 PTFE diffusion layers to prevent water loss (14). Single-chamber, air cathode, cubic-shaped MFC reactors were constructed on the basis five previously described configurations (5, 7, 15-17) (Figure 1). The “control” MFC lacked a separator, with the anode and cathode spaced 2 cm apart (NS). This design was modified by including a separator placed adjacent to the cathode, while maintaining the 2 cm electrode spacing (S). The third type of MFC used a single separator electrode assembly (SSEA), where the separator was sandwiched between the anode and cathode placed on the same side of the reactor. The liquid volumes of the NS, S, and SSEA reactors were 12 mL. The final two reactors were double-separator electrode assembly MFCs with two SEAs (one on each side), with either a 2 cm spacing between the electrodes (DSEA2, 9 mL liquid volume) or 0.3 cm spacing (DSEA0.3, 1.8 mL liquid volume). All anodes were enriched in NS-type MFCs using suspended bacteria from an MFC (originally inoculated with primary clarifier overflow) that had been operated in fedbatch mode for over one year. The reactors were fed 1 g/L of acetate in a 50 mM PBS medium (18) containing mineral (12.5 mL/L) and vitamin (5 mL/L) solutions (19). Following operation of the NS MFCs for more than 20 cycles, where they all exhibited stable and parallel performance, the anodes were then used in the different MFC designs described above. All MFCs were operated under ambient temperature conditions in the laboratory (23 ( 3 °C) with a 1000 Ω resistor except as noted. Separator Selection and Analysis. Glass fiber separators had a 1.0 mm (GF1; type DC1.0, Jiafu Co., China) or 0.4 mm thickness (GF0.4; DC0.4, Jiafu Co., China). The performance

of these separators was compared to a cation exchange membrane (CEM; CMI-7000, Membrane International, Inc., NJ) or J-cloth (JC; Associated Brands LP, Canada). Oxygen and proton mass transfer and coefficients of the separators were measured in uninoculated twochamber bottle reactors that were mixed with a magnetic stir bar (350 rpm). Before measurement of the oxygen mass transfer coefficients (kO, cm/s), the reactors with separators were assembled in an anaerobic glovebox to keep anode and cathode chambers under anoxic conditions and filled with an oxygen-free 50 mM phosphate buffer solution (PBS) containing (g/L in deionized water, pH 7): NH4Cl (0.31), KCl (0.13), NaH2PO4 · H2O (2.69), and Na2HPO4 (4.33) (9). The assembled reactor was removed from the glovebox, and the cathode camber was continuously sparged with air to maintain saturated dissolved oxygen (DO) conditions. The DO in the anode chamber was monitored by using a nonconsumptive DO probe (Foxy-21G, Ocean Optics, Inc., FL), with kO calculated (9) using kO ) -

[

c1,0 - c2 v ln At c1,0

]

(1)

where v is the liquid volume in the anode chamber (200 mL), A is the membrane cross-sectional area (1.33 cm2), c1,0 is the saturated DO concentration in the cathode chamber, and c2 is the DO in the anode chamber at time t. The diffusion coefficient (DO, cm2/s) was calculated from the thickness of the separator (LD), as DO ) KOLD. The proton mass transfer coefficients (kH, cm/s) was calculated from similar tests using the same reactors filled with DI water (pH, ∼7, recorded for c1.0 calculation). NaOH solution was then added into the anode chamber to adjust the pH of anolyte to ∼8.5, and the pH in anode chamber was continuously monitored using a pH probe. The mass transfer coefficient was then calculated as kH ) -

[

c1,0 + c2,0 - 2c2 v ln 2At c1,0

]

(2)

where c2.0 is the initial proton concentration in the anode chamber (Supporting Information). Ohmic resistances for the separators (RS) were determined in uninoculated two-chamber cubic MFCs with 50 mM PBS by electrochemical impedance spectroscopy (EIS) using a potentiostat (PC 4/750, Gamry Instrument, Inc., PA) (18, 20). Impedance measurements were conducted at the open circuit voltage (OCV) over a frequency range of 10000 to 0.1 Hz with a sinusoidal perturbation of 10 mV amplitude. The ohmic resistances of reactors with or without separators (RW or RB) was determined using Nyquist plots of the impedance spectra from the real impedance Zre, where it intersects the X axis (imaginary impedance Zim ) 0). Thus, RS ) RW - RB. Chemical Measurements and Analyses. Voltage (V) across the external resistor in the MFC circuit was measured at 20 min intervals using a data acquisition system (2700, Keithley Instrument, OH) connected to a personal computer. Current (I ) V/R), power (P ) IV), and Coulombic efficiency (CE) were calculated as previously described (8), with the current density and power density normalized by the projected surface area of the cathode and the volumetric power density normalized by the liquid volume. The polarization curves were obtained by varying external resistance from 1000 to 20 Ω in decreasing order, with a single resistor used per fedbatch cycle. Biofilm growth was evaluated on the basis of protein concentrations measured using the bicinchoninic acid method (21). The separator was placed in a vial with 3 mL of 0.2 N NaOH and then shaken to disperse the biofilm into solution. The separator was then rinsed with an additional VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8457

FIGURE 2. Volumetric power density (filled symbols, primary axis; open symbols, secondary axis) as a function of current density obtained by varying the circuit resistance for MFCs using glass fiber and JC (two layers) materials in various reactor configurations. 3 mL of deionized water. The liquids were combined, yielding a 6 mL sample that was then analyzed for protein (all reagents from Sigma Chemical Co., St. Louis, MO).

Results Power Generation Using Separators Directly between the Electrodes (SSEA and DSEA). When the electrodes in the MFC were placed close together (SSEA and DSEA configurations), separators increased performance on the basis of overall volumetric power density compared to other configurations examined here. The GF1 separator consistently produced the best performance among the different separators tested, with a maximum volumetric power density of 70 ( 2 W/m3 (1195 ( 30 mW/m2) when using a single electrode assembly (SEA) (Figure 2). This was slightly larger than that achieved with a two-layer JC (64 ( 1 W/m3, and 1099 ( 8 mW/m2) and much larger than power densities with GF0.4 or CEM separators (Table 1). The improved performance of the SEA configuration was a result of the low internal resistance of 10.1 ( 3.4 Ω, compared to 38.1 ( 0.1 Ω when the electrodes were spaced 2 cm apart (see below). When the MFCs were operated with two complete sets of electrodes, power production was further increased. A maximum volumetric power density of 150 ( 6 W/m3 was obtained with DSEA2 (2 cm anode spacing), which was 115% more power than produced with the single SEA (Figure 2 and Table 1). Decreasing the distance between the two anodes from 2 to 0.3 cm produced a decrease in the ohmic resistance from 5.9 ( 0.1 to 2.2 ( 0.1 Ω, and the power density increased to 696 ( 26 W/m3 (3.57 A/m2, Figure 2 and Table 1). The power density based on cathode area, however, decreased from 963 ( 39 mW/m2 (2 cm spacing) to 895 ( 33 mW/m2 (0.3 cm spacing). This decrease based on surface area could be due to oxygen intrusion as noted by others (5). Thus, while the separator did slightly reduce the performance of the cathode on a projected area basis, the substantial reduction in internal resistance resulted in an overall improvement in volumetric power production. Power Generation Using Separators in MFCs with 2 cm Electrode Spacing. When the electrodes were spaced 2 cm apart, separators reduced power generation compared to the MFC lacking a separator (Table 1). With a 2 cm electrode spacing between the anode and cathode, the power densities using GF1 (46 ( 4 W/m3, 791 ( 69 mW/m2) or JC (46 ( 1 W/m3, 786 ( 23 mW/m2) separators were similar, but they were lower with GF0.4 (36 ( 1 W/m3 and 623 ( 4 mW/m2) or a CEM separator (14 ( 1 W/m3 and 267 ( 22 mW/m2). In comparison, the maximum power density for the MFC without a separator (NS) was 52 ( 2 W/m3 (896 ( 49 mW/ m2) (Figure 3A). This effect of a separator on decreasing power generation when electrodes are spaced relatively far apart is 8458

9


consistent with that observed by others (15). The placement of the separator next to the cathode decreases the cathode potential (Figure 3B) and increases ohmic internal resistance (Table 1). The anode potential is not appreciably affected by the separator. Coulombic Efficiencies. The main advantage of the GF separators was to increase Coulombic efficiencies (CE). The MFC lacking a separator produced the lowest CEs, with values that increased from 19% to 34% with the current (Figure 4). The MFC with a JC separator similarly had low CEs of 20-40%. The JC material is highly porous and water permeable and has a very large oxygen mass transfer coefficient of kO ) 290 × 10-5 cm/s; therefore, it did not result in a CE much different from that of the control MFC lacking a separator. However, the GF1 separator produced the highest CE of 81%, at a current density of 3.85 A/m2. The reason for this high CE is the low oxygen mass transport coefficient of the GF1 separator of kO ) 5.0 × 10-5 cm/s (see below). When the GF1 separator was used in MFCs in the SSEA and DSEA configurations, CE performance further improved over each fed-batch cycle. The CE increased from 43% to 80%, with an increase in current density of 0.4 to 2.0 A/m2, and then reached a range of 83-91% above a current density of 2.0 A/m2 for the GF1 separator. A CE of 85% was obtained at a current density of 3.4 A/m2 in SEA and double SEA MFCs with glass fiber 1.0, which was 150% higher than that of the MFC in absence of a separator (34%) at the same current density. Separator Properties. The porous nature of the JC material resulted in a low ohmic resistance of RS ) 0.21 ( 0.08 Ω, but this open structure also produced a very high oxygen mass transfer coefficient of kO ) 290 × 10-5 cm/s when it was not covered with a biofilm (Figure 1S of the Supporting Information). The JC and glass fiber (GF1 and GF0.4) all had similar proton diffusion coefficients, which ranged from DH ) 9.08 to 9.40 × 10-5 cm2/s (Table 2). The GF1 separator had the lowest oxygen mass transfer coefficient of kO ) 5.0 × 10-5 cm/s. The transport coefficients for GF0.4 and CEM were kO ) 7.5 × 10-5 cm/s and 9.4× 10-5 cm/s, respectively. The CEM produced a relatively high internal resistance (RS ) 3.78 ( 0.37 Ω), with lower values for GF1 (RS )2.26 ( 0.13 Ω) and GF0.4 (RS ) 2.39 ( 0.30 Ω). Effects on Cathodic Biofilm. Although the MFC without separator had the lowest CE after 20 cycles, the CE significantly increased over time (Figure 5A). To further study this effect of the biofilm, we inoculated and acclimated (external resistance of 1000 Ω) three new MFCs. Power output was stable after 10 cycles, but after 30 cycles the power started to slowly decrease, and the CE increased (Figure 5A). Visual inspection of the cathodes showed the presence of a thick biofilm. Following the 40th cycle, the biofilm was scraped off the cathode. This resulted in an increase in the voltage and a decrease in CE to values similar to those originally obtained with the acclimated reactor (Figure 5B). The increase in voltage resulted in a higher current density and thus a shorter cycle time. Note that the anode potentials were not affected by the biofilm removal (39th and 40th cycle) (Figure 5C), only the cathode potentials. Thus, we concluded that the cathode biofilm functioned in a manner similar to that of the separator, which was to reduce oxygen diffusion and proton transfer, resulting in an increase in CE and a decrease in power. The protein content of the cathodic biofilm was 1245 µg/ cm2 on the basis of projected cathode area after 39 cycles in MFC without a separator. The protein content of the biofilm grown in the JC material when it was placed against a cathode (S-type MFC) measured after 39 cycles was 1218 µg/cm2, a value comparable to that of the biofilm from the cathode lacking a separator. The effect of the growth of the biofilm was an increase in the CE from less than 20% (1st cycle) to

TABLE 1. Power Production and Internal Resistance of MFCs maximum power density configuration

separator

ohmic resistance (Ω)

(W/m3)

(mW/m2)

NSa Sb S S S SSEAc SSEA DSEA2d DSEA0.3e

glass fiber 1.0 (GF1) glass fiber 0.4 (GF0.4) one-layer J-cloth (JC) cation exchange membrane (CEM) glass fiber 1.0 (GF1) two-layer J-cloth (JC) glass fiber 1.0 (GF1) glass fiber 1.0 (GF1)

37.6 ( 1.4 38.1 ( 0.1 40.1 ( 0.4 38.1 ( 0.1 131.7 ( 8.4 10.1 ( 3.4 11.3 ( 0.3 5.9 ( 0.1 2.2 ( 0.1

52 ( 2 46 ( 4 36 ( 1 46 ( 1 14 ( 1 70 ( 2 64 ( 1 150 ( 6 696 ( 26

896 ( 49 791 ( 69 623 ( 4 786 ( 23 267 ( 22 1195 ( 30 1099 ( 8 963 ( 39 895 ( 33

a NS, without separator, 2 cm space between cathode and anode. b S, with separator against cathode, 2 cm space between cathode and anode. c SSEA, single separator electrode assembly. d DSEA2, double SEA, with 2 cm anodes space. e DSEA0.3, double SEA, with 0.3 cm anodes space.

TABLE 2. Mass Transfer Coefficients and Diffusivities and Oxygen and Proton and Ohmic Resistance Measured for Each Separator separator glass fiber 1.0 glass fiber 0.4 property L (cm) kO (× 10-5 cm/s) DO (× 10-6 cm2/s) kH (× 10-4 cm/s) DH (× 10-5 cm2/s) RS (Ω) a

CEMa

J-cloth

0.1 5.0

0.04 7.5

0.03 290

0.046 9.4

5.0

3.0

86.9

4.3

9.40

23.18

30.27

-

9.40

9.27

9.08

-

2.26 ( 0.13

2.39 ( 0.30 0.21 ( 0.08 3.78 ( 0.37

From Kim et al. (9).

penetrate past the cathode and biofilm. In contrast, GF1 had a much lower oxygen mass transfer coefficient, and it was found that this material resulted in much less biofilm growth as shown by a protein measurement of only 409 µg/cm2 after 39 cycles.

Discussion FIGURE 3. (A) Power density and (B) electrode potentials (cathode, filled symbols; anode, open symbols) versus Ag/AgCl reference electrode (0.195 V versus NHE) as a function of current density obtained by varying the external circuit resistance for MFCs in various configurations.

FIGURE 4. Coulombic efficiencies as a function of current density for MFCs with various configurations. 25% (30th cycle). The presence of the biofilm was therefore important for a successful increase in power density for the JC separator. Without a biofilm, the JC had a very high oxygen mass transfer coefficient, but with a biofilm any oxygen transfer into the MFC would be removed before it could

The separator is a critical component of an MFC as it allows the electrodes to be closely spaced, reducing ohmic resistance, and allowing for increased power generation. However, oxygen diffusion past the cathode and into the anode chamber must be blocked or eliminated for power to increase as oxygen utilization by the anode bacteria reduces electron flow into the circuit. A reduction in oxygen transfer into the MFC can be accomplished using a separator in one of two ways: by reducing oxygen transfer due to blockage by the separator material or by enhancing the growth of a biofilm in the separator that consumes oxygen before it can reach the anode. It was shown here that a glass fiber separator GF1 has excellent separator characteristics for MFCs as it has a high proton transfer coefficient, low oxygen transfer coefficient, and low ohmic resistance, and it is nonbiodegradable. The GF1 material reduced oxygen transfer and thus increased CE to values higher than other separator materials. While J-cloth (JC) also increased power (5), it has an oxygen mass transfer coefficient that is 57 times larger than that of GF1 (290 × 10-5 cm/s), and therefore it produces very low CEs. Growth of the biofilm in the JC separator is an important factor for its success in increasing power. The JC material supported a biofilm (on the basis of a protein content of 1218 µg/cm2) that was similar to that produced on a cathode in the absence of the separator. Oxygen transfer through the cathode was reduced by biofilm growth on the cathode (or VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8459

transfer reduction (biofilm or separator properties) affected the CE and the overall MFC performance. The maximum volumetric power densities were obtained in MFCs that had the lowest internal resistances, which resulted from close electrode spacing (which also decreased the reactor volume) and effective blockage of oxygen transfer. The ohmic internal resistance was decreased from 37.6 Ω (NS-configuration) to 10.1 Ω by moving the electrodes together and using a GF1 separator (SSEA). In this SSEA configuration, the anode surface area per volume of reactor was 58 m2/m3. By increasing the specific surface area in the reactor to 778 m2/m3 by adding a second set of electrodes spaced only 0.3 cm apart, we found the internal resistance was reduced to 2.2 Ω and power density was further increased to 696 W/m3. CEMs and other membranes that have been used in MFCs can be used to reduce oxygen transfer, but so far they have not produced performance comparable to MFCs using GF1 or JC materials as a result of high ohmic resistance due to pH gradients across the membrane (10). These results show that separators that produce high power and CE must inhibit oxygen transfer and should minimize bacteria growth within and on the separator, and they must be nonbiodegradable. While glass fiber materials have been shown to function well as separators, other materials should continue to be explored as separators to further increase power. The GF1 and JC materials reduce power on a surface area basis, indicating that they adversely affect power generation likely through the inhibition of charge transfer to the cathode. The development of inexpensive materials and coatings on the cathode that can achieve oxygen transfer reduction and good charge transfer to the cathode will be useful for improving performance and reducing the cost of manufacturing MFCs.

Acknowledgments

FIGURE 5. (A) Power density (P) and Coulombic efficiency (CE) as a function of cycle number for MFC without a separator (1000 Ω). At the 40th cycle, the cathodic biofilm was removed. Error bars indicate results based on triplicate reactors. (B) Voltage output and (C) electrode potentials (vs Ag/AgCl reference electrode, 0.195 V vs NHE) as a function of time for the 39th and 40th cycles of the MFC without a separator before and after cathodic biofilm removal. in the separator) so that power was not affected by the presence of dissolved oxygen. However, removal of the oxygen by the biofilm requires substrate, and thus CE is low as this substrate is not available for current generation. This biofilm also reduced power on a surface area basis, presumably as a result of physical blockage of charge transfer (i.e., transfer of charged ionic species) near the cathode. If the biofilm on the cathode was removed, power increased and CE decreased. For example, in the MFC in the S-configuration, the CE increased from 20% to 25% over 30 cycles, but the power decreased from 437 to 403 mW/m2. When the cathodic biofilm was then removed, the power was restored to the original power density of 425 mW/m2. These CE values are comparable to a value of 24% reported by Fan et al. (5) for an MFC in a similar configuration. In contrast, the GF1 in the S-configuration produced a much higher CE of 70% (current density of 3.4 A/m2). These values are all larger than those produced in the same reactor lacking a separator (NS configuration) of CE ) 14%. Thus, these separators all decreased oxygen transfer, but the mechanism for oxygen 8460

9


This research was supported by Award KUS-I1-003-13 from the King Abdullah University of Science and Technology (KAUST), the U.S. National Science Foundation (CBET0730359), the International Program of MOST (2006DFA91120) in China, and a scholarship from the China Scholarship Council (CSC).

Supporting Information Available Photograph of degradable J-cloth and calculation of diffusion coefficients of protons on a separator. This material is available free of charge via the Internet at http://pubs.acs. org/.

Literature Cited (1) Logan, B. E. Microbial Fuel Cells. John Wiley & Sons, Inc.: Hoboken, NJ, 2008. (2) Lovley, D. R. The microbe electric: Conversion of organic matter to electricity. Curr. Opin. Biotechnol. 2008, 19 (6), 564–571. (3) Logan, B. E. Exoelectrogenic bacteria that power microbial fuel cells. Nat. Rev. Microbiol. 2009, 7 (5), 375–381. (4) Rabaey, K.; Verstraete, W. Microbial fuel cells: Novel biotechnology for energy generation. Trends Biotechnol. 2005, 23 (6), 291–298. (5) Fan, Y. Z.; Hu, H. Q.; Liu, H. Enhanced Coulombic efficiency and power density of air-cathode microbial fuel cells with an improved cell configuration. J. Power Sources 2007, 171 (2), 348–354. (6) Liu, H.; Cheng, S. A.; Logan, B. E. Power generation in fed-batch microbial fuel cells as a function of ionic strength, temperature, and reactor configuration. Environ. Sci. Technol. 2005, 39 (14), 5488–5493. (7) Cheng, S.; Liu, H.; Logan, B. E. Increased power generation in a continuous flow MFC with advective flow through the porous anode and reduced electrode spacing. Environ. Sci. Technol. 2006, 40 (7), 2426–2432.

(8) Logan, B. E.; Hamelers, B.; Rozendal, R.; Schrorder, U.; Keller, J.; Freguia, S.; Aelterman, P.; Verstraete, W.; Rabaey, K. Microbial fuel cells: Methodology and technology. Environ. Sci. Technol. 200640 (17), 5181–5192. (9) Kim, J. R.; Cheng, S.; Oh, S. E.; Logan, B. E. Power generation using different cation, anion, and ultrafiltration membranes in microbial fuel cells. Environ. Sci. Technol. 2007, 41 (3), 1004– 1009. (10) Rozendal, R. A.; Hamelers, H. V. M.; Molenkamp, R. J.; Buisman, C. J. N. Performance of single chamber biocatalyzed electrolysis with different types of ion exchange membranes. Water Res. 2007, 41 (9), 1984–1994. (11) Biffinger, J. C.; Ray, R.; Little, B.; Ringeisen, B. R. Diversifying biological fuel cell designs by use of nanoporous filters. Environ. Sci. Technol. 2007, 41 (4), 1444–1449. (12) Cheng, S.; Liu, H.; Logan, B. E. Power densities using different cathode catalysts (Pt and CoTMPP) and polymer binders (Nafion and PTFE) in single chamber microbial fuel cells. Environ. Sci. Technol. 2006, 40 (1), 364–369. (13) Cheng, S. A.; Logan, B. E. Ammonia treatment of carbon cloth anodes to enhance power generation of microbial fuel cells. Electrochem. Commun. 2007, 9 (3), 492–496. (14) Cheng, S.; Liu, H.; Logan, B. E. Increased performance of single-chamber microbial fuel cells using an improved cathode structure. Electrochem. Commun. 2006, 8 (3), 489– 494.

(15) Liu, H.; Logan, B. E. Electricity generation using an air-cathode single chamber microbial fuel cell in the presence and absence of a proton exchange membrane. Environ. Sci. Technol. 2004, 38 (14), 4040–4046. (16) Sammes, N. M. Fuel Cell Technology: Reaching Towards Commercialization. Springer: London, 2006. (17) Liang, P.; Huang, X.; Fan, M. Z.; Cao, X. X.; Wang, C. Composition and distribution of internal resistance in three types of microbial fuel cells. Appl. Microbiol. Biotechnol. 2007, 77 (3), 551–558. (18) 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. (19) Lovley, D. R.; Phillips, E. J. P. Novel mode of microbial energy metabolism: Organic carbon oxidation coupled to dissimilatory reduction of iron or manganese. Appl. Environ. Microbiol. 1988, 54 (6), 1472–1480. (20) He, Z.; Wagner, N.; Minteer, S. D.; Angenent, L. T. An upflow microbial fuel cell with an interior cathode: Assessment of the internal resistance by impedance spectroscopy. Environ. Sci. Technol. 2006, 40 (17), 5212–5217. (21) Bond, D. R.; Lovley, D. R. Electricity production by Geobacter sulfurreducens attached to electrodes. Appl. Environ. Microbiol. 2003, 69 (3), 1548–1555.

ES901631P

VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8461

Separator Characteristics for Increasing Performance of Microbial Fuel

Recommend Documents