Quantification of the Internal Resistance Distribution of Microbial Fuel

Environ. Sci. Technol. 2008, 42, 8101–8107

Quantification of the Internal Resistance Distribution of Microbial Fuel Cells YANZHEN FAN, EVAN SHARBROUGH, AND HONG LIU* Department of Biological and Ecological Engineering, Oregon State University, 116 Gilmore Hall, Corvallis, Oregon 97331

Received August 11, 2007. Revised manuscript received July 30, 2008. Accepted August 11, 2008.

Identifying the limiting factors in a microbial fuel cell (MFC) system requires qualifying the contribution of each component of an MFC to internal resistance. In this study, a new method was developed to calculate the internal resistance distribution of an MFC. Experiments were conducted to identify the limiting factors in single-chamber MFCs by varying the anode surface areas, cathode surface areas, and phosphate buffer concentrations. For the MFCs with equally sized electrodes (7 cm2) and 200 mM phosphate buffer, the anode contributed just 5.4% of the internal resistance, while the cathode and the electrolyte each contributed 47.3%, indicating that the anode was not the limiting factor in power generation. The limitation of the cathode was further revealed by the 780% higher area-specific resistance (284.4 Ω cm2) than the 32.3 Ω cm2 of the anode. The electrolyte limitation was also evidenced by the greatly increased contribution of electrolyte in internal resistance from 47.3 to 78.2% when the concentration of phosphate buffer was decreased from 200 to 50 mM. An anodic power density of 6860 mW/m2 was achieved at a current density of 2.62 mA/ cm2 using the MFCs with an anode/cathode area ratio of 1/14 and 200 mM phosphate buffer. The method was also successfully applied to analyze the internal resistance distribution of the two chamber MFCs from a previously reported study. The comparison of the internal resistances of the two air cathode systems indicates that the much lower resistances, including anode, cathode, and membrane resistances, contributed to the much better performance of the singlechamber MFCs than the two-chamber system.

Introduction Microbial fuel cells (MFCs), which can directly generate electricity from biodegradable substances, have rapidly gained increasing research attention. Although the power density of an MFC is still low compared to a hydrogen fuel cell, the renewable and widely available fuel sources and moderate operational conditions make it very promising in renewable energy generation, wastewater treatment, and as potable power supplies or power sources for remote electronic devices. Significant increases in power density have been achieved in recent studies (1-4). Power densities of 115 W/m3 (1) and 2400 mW/m2 (73 W/m3) (2) using air cathode MFCs were reported. An even higher power density of 500 W/m3 was achieved with a miniature MFC using * Corresponding Author phone: 541-737-6309; fax: 541-737-2082; e-mail: [email protected]. 10.1021/es801229j CCC: $40.75

Published on Web 09/24/2008

 2008 American Chemical Society

ferricyanide as catholyte (3). More recently, Fan et al. increased the power density to 1010 W/m3 using air cathode MFCs with cloth electrode assemblies (CEAs) (4). However, further amplifying the power density remains one of the greatest challenges for realizing the practical applications of MFCs. Identifying the limiting factors in an MFC system is critical for further enhancing the MFC performance. Internal resistance, including anode resistance, cathode resistance, electrolyte resistance, and membrane resistance (if any), limits the power output of an MFC (5). The internal resistance can be reduced by increasing the anode surface area (2, 6), the cathode surface area (7), the surface area of the proton exchange membrane (PEM) (6), the ionic strength of the electrolyte (8), or the pH (1). Identification of the limiting factor of an MFC requires quantification of the contribution of each component of an MFC to internal resistance. Like any electrochemical cell, the internal resistance of an MFC can also be classified as ohmic resistance, charge-transfer resistance, and diffusion resistance (12), which can be analyzed using some alternating current (AC) methods such as electrochemical impedance spectroscopy (EIS) (9). The resistance of each fuel cell component may also be estimated using EIS if an appropriate equivalent circuit can be constructed (9). EIS has been utilized to separate the ohmic resistance from the other resistances of MFCs (2, 4, 10). However, the analysis of charge-transfer resistance and diffusion resistance of an MFC remains a great challenge (2, 4, 10, 11) probably due to the 3-dimensional nature of the MFC anode, especially with a developed biofilm, and the relative larger electrode spacing (9). The internal resistance has to be measured over a wide frequency range in order to cover the diffusion resistance, which is usually found in the lowest frequency region, down to 10 mHz, or even lower for chemical fuel cell (9). The relatively larger electrode spacing in MFCs requires longer time constant thus even lower frequency (9). However, the EIS of MFCs at low frequency region is often unstable and inaccurate. In an up-flow two chamber MFC system, the EIS data were inaccurate beyond the range 9 Hz to 4.4 kHz (11). The lowest frequency (9 Hz) might be too high to analyze the diffusion resistance, which might be the reason for the inconsequential diffusion resistance. Although much wider frequency range (5 mHz to 300 kHz) were employed in single-chamber MFCs (2, 4, 10), the charge transfer and diffusion resistances haven′t been separated. Direct current (DC) analytical methods such as polarization curve can be used to measure the total internal resistance (12, 13), but not the resistance of the individual components. There is substantial need for the development of a novel method to analyze the quantitative contribution of anode, cathode, membrane, and electrolyte resistances to the overall internal resistance. Such a quantitative method can be used to identify the limiting factor of an MFC that governs the internal resistance, and in turn the power output. Recently, some computational models have been developed (14-16), which provide comprehensive insights of the biofilm development and kinetics of the anodic reaction. However, it remains difficult to model the whole microbial fuel cell, including anode, cathode, electrolyte, and membrane. In this paper, a new method was developed to calculate the contribution of the anode, cathode, electrolyte, and membrane to internal resistance. Experiments were conducted to investigate the limiting factors in single-chamber MFCs by varying the anode surface areas, cathode surface areas, and phosphate buffer concentrations. The method was also successfully implemented to calculate the distribuVOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8101

(ohmic) region of the curve (5). For the linear polarization curve, the relationship between the external voltage (E) and current (I) can be described as follows: E ) Eb - RintI

FIGURE 1. The polarization curve of an MFC shows the three regions of polarization overpotentials. tion of internal resistance of two-chamber MFCs using the data reported in a previous study (6).

Materials and Methods MFC Construction and Operation. Single-chamber membrane-free MFCs (12 mL liquid volume, 1.7 cm electrode spacing) were constructed as reported previously (17) and were used to investigate the power density at various anode areas (1-7 cm2; nonwet proofing, type A, E-Tek) using 7 cm2 carbon cloth cathodes. In addition, a double-cathode reactor setup was arranged wherein the 1 cm2 anode was placed equidistant between two 7 cm2 cathodes in order to further investigate the maximum power density generated using this carbon fiber cloth. The cathode has a platinum loading of 0.5 mg/cm2 Pt coated with a PTFE layer to prevent water loss. Details of cathode fabrication can be found in a previous study (18). The MFCs were inoculated with a mixed bacterial culture from the anode of a single-chamber MFC, which was originally inoculated with domestic wastewater (Corvallis Wastewater Treatment Plant, Corvallis, OR) and has been operated for about 1 year using acetate. All MFCs were operated in batch mode using nutrient medium (4) amended with acetate (30 mM). The electrolyte solution was replaced when the voltage decreased to less than 30 mV. The system was considered to be operating under steady conditions when voltage output was reproducible after refilling reactor with medium at least twice. Repeatable power output can normally be obtained in about 50 h (three batches) at a fixed external resistance of 1000 Ω. Resistance was gradually decreased from 1000 to 60 Ω for polarization curve generation. At each resistance, MFCs were run for at least two batches for reproducible voltage output. The 10 data points around the maximum voltage at each batch were averaged and used to make the polarization curve. 2-3 weeks (15-20 batches) of operation was sufficient to finish one polarization curve. Voltage (V) was measured using a multimeter with a data acquisition system (2700, Keithley, U.S.) and used to calculate the power (P) according to P ) IV. Power density was then calculated using the power normalized by the cross sectional area (projected) of the anode. Calculation of Internal Resistance Distribution. A polarization curve is commonly used to study the performance of a fuel cell and the various current related losses and includes three regions (Figure 1). The cell voltage drops rapidly and nonlinearly in (1) the activation region, followed by a slow and near-linear drop in (2) the ohmic region. The continuous current increase may reflect (3) the concentration polarization region or a condition where mass transport of reactants becomes the limiting factor. In this region, the voltage drops rapidly and nonlinearly with the current increase. In a MFC, operation often takes place in the linear 8102

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 21, 2008

(1)

where Rint is the total internal resistance of an MFC, which can be easily obtained from the slope of the linear curve. The intercept of the linear curve with the voltage (E) axis, Eb, is referred to as the linear extrapolation open circuit voltage (LE-OCV) in this study, i.e., the voltage that is determined by the extrapolation of the linear part of the polarization curve. The value of Eb should be close to, but less than, the measured open circuit voltage (OCV). The calculated resistance using this method is referred to as internal resistance rather than ohmic resistance because activation loss and concentration polarization affect the slope of the polarization curve, and the influence has been considered in the calculation of the internal resistance using this method. The total internal resistance Rint of an MFC can be further dissected into anodic resistance Ra, cathodic resistance Rc, membrane resistance Rm (if it exists), and electrolyte resistance Re. Therefore, eq 1 can be rewritten as follows: E ) Eb - (Ra + Rc + Rm + Re)I

(2)

The Ra, and Rc, refer to the resistances caused by bioelectrochemical reactions at anode and oxygen reduction at cathode, respectively. The contact resistance is neglected because it is normally less than 1% of the total internal resistance in a well designed MFC. It can be simply included in eq 2 as a constant if its value is too large to be neglected. The anode resistance Ra depends on the size of the anode and the activity of the electricity-generating bacteria. Both the biofilm on the anode and planktonic bacteria in bulk solution (through self-produced mediators) (19) may contribute to the power generation of MFCs when mixed bacterial cultures are utilized. However, the planktonic bacteria had a negligible contribution to the power generation in the tested MFC system (see the Supporting Information). Therefore, for a plate-type anode (e.g., carbon cloth), Ra can be assumed inversely proportional to the projected area of the anode Sa, i.e., Ra ) ra ⁄ Sa

(3)

where ra is the area-specific resistance (ASR) of the anode, or the resistance normalized by anode area. Similarly, Rc ) rc ⁄ Sc

(4)

Rm ) rm ⁄ Sm

(5)

Re ) re ⁄ Sr

(6)

where rc, rm, and re are the ASRs of the cathode, membrane, and electrolyte, respectively; Sc, Sm, and Sr are the projected areas of the cathode, membrane, and electrolyte (reactor), respectively. It should be noted that the values of Ra, Rc, and Rm might be different from the values measured with traditional electrochemical methods, e.g., EIS using Luggin capillary reference electrode. While it is impossible to totally separate the contribution of electrodes or membranes from that of electrolytes using the traditional methods, the method developed here includes the contribution of electrolyte only in electrolyte resistance Re, which represents all the resistances caused by electrolyte, including ohmic resistance, charge transfer resistance, and diffusion resistance. Assuming Re is proportional to the electrode spacing L, while inversely proportional to the cross-sectional area of the reactor Sr and

the concentration of charge transfer electrolyte Ce (10), the electrolyte resistance can be computed as follows: Re ) aL ⁄ (Sr × Ce)

(7)

pmax ) 0 . 25Eb2 ⁄ Rint

where a is a constant. Therefore, eq 2 can be rewritten as E ) Eb - (ra ⁄ Sa + rc ⁄ Sc + rm ⁄ Sm + aL ⁄ (Sr × Ce))I

(8)

The total internal resistance, Rint, can be written as Rint ) ra ⁄ Sa + rc ⁄ Sc + rm ⁄ Sm + aL ⁄ (Sr × Ce)

(9)

Equation 8 can be further transformed into the equation below in order to calculate the power density directly. p ) Ebi - Rinti2

From eq 10, we can easily derive that the power density (p) of an MFC is the quadratic function of the current density (i), while the maximum power density, or the maximum value of the quadratic function, can be computed as follows:

(10)

where p is the power density and i the current density.

(11)

The values of Eb, ra, rc, rm, a, and pmax can be determined with a best fit of the experimental data with eqs 8, 10 or 11 using the SOLVER function in Microsoft Office Excel. In this study, the sums of the squares of the differences in the observed and calculated values were minimized by varying the initial guess 1000 times over a wide range of possible parameter values with a quasi-Newton search method.

Results and Discussion Single-Chamber MFC System. Power Generation. To explore the power density generated by the mixed culture using carbon cloth electrodes and the relationship between the

FIGURE 2. Comparison of observed (A) voltage and (B) anodic power density (points) and the calculated results (lines) as a function of current density in MFCs with different anode areas (1, 2, 3.6, and 7 cm2), cathode areas (7 and 14 cm2), and phosphate buffer concentrations (50, 100, and 200 mM). Inserts show the correlations between the observed and calculated voltage/power density. The format of the legend is anode area (cm2)scathode area (cm2)sbuffer concentration (mM). VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8103

TABLE 1. Internal Resistance Distribution and Anodic Maximum Power Density of MFCs with Different Electrode Areas and Phosphate Buffer Concentrations MFC b

1 2 3 b 4 b 5 b 6

7 7 1 1 1 1

b b

a

Sa cm2

Sc cm2

Cb mM

7 7 7 7 14 ∞

200 50 50 200 200 ∞

Ra Ω (%)

Rc Ω (%)

Re Ω (%)

RintΩ

pmax,ra mW/m2

4.6 (5.4%) 4.6 (2.2%) 32.3(13.7%) 32.3(28.4%) 32.3(34.6%) 32.3(100%)

40.6(47.3%) 40.6(19.5%) 40.6(17.2%) 40.6(35.8%) 20.3(21.8%) 0

40.7(47.3%) 162.7(78.2%) 162.7(69.0%) 40.7(35.8%) 40.7(43.6%) 0

85.9 208.0 235.7 113.6 93.3 32.3

1040 430 379 787 958 -

Based on the cross-sectional area of the MFC reactor.

b

anode and cathode surface areas, the cross-sectional areas were systematically decreased from 7 to 3.6, 2, and 1 cm2 using single or double cathode arrangements. Decreasing the anode surface area relative to the cathode surface area increased the power density (Figure 2). A maximum power density of 1050 mW/m2 was obtained at a current density of 0.39 mA/cm2 in MFCs using equally sized electrodes (7 cm2) (Figure 2). This result is comparable to a previous study using the same electrode materials and similar solution chemistry (8). Reducing the anode surface area from 7 to 1 cm2 resulted in a 430% increase in anodic power density (5570 mW/m2). A power density of 6860 mW/m2 was achieved when double cathodes (14 cm2) were used, corresponding to an anodic current density of 2.62 mA/cm2. The power density of the reactor was also greatly affected by the concentration of the phosphate buffer. As demonstrated in Figure 2, the measured maximum power density was 2680 mW/m2 for MFCs with 50 mM phosphate buffer. The maximum power density increased by 49% to 4000 mW/ m2 when the buffer concentration increased to 100 mM. Further increasing the buffer concentration to 200 mM resulted in an additional 39% increase in power density. Internal Resistance Distribution. The increase in current density resulted in a uniformly linear decrease in voltage, especially in a voltage range of 0.2-0.45 V. For the air cathode single-chamber MFCs used in this study, there was no membrane and the dimensions of the MFCs (L ) 1.7 cm; Sr ) 7 cm2) were fixed. Therefore, eq 8 can be simplified to the following: E ) Eb - (ra ⁄ Sa + rc ⁄ Sc + ac ⁄ Cb)I

(12)

where E (V) is the external voltage, Eb (V) is the linear extrapolation open circuit voltage, ra (Ω cm2) is the anodic ASR, Sa (cm2) is the surface area of anode, rc (Ω cm2) is the cathodic ASR, Sa (cm2) is the surface area of cathode, ac (Ω mM) is a constant, Cb (mM) is the buffer concentration, and I (A) is the current. Similarly, eq 10 can be simplified as follows: p ) Ebi - (ra ⁄ Sa + rc ⁄ Sc + ac ⁄ Cb)i2

(13)

Both eqs 12 and 13 were used to fit all the experimental data except those with an actual voltage greater than 0.45V which were not in the linear region of the polarization curve. Similar results were obtained using eqs 12 and 13 and both equations fit the experimental data very well (R2 > 0.99, n ) 47) (Figure 2). The calculated LE-OCV of the MFCs (0.5004 V) was much lower than the OCV of the same MFCs, which was in the range of 0.7-0.8 V. The ASR of the cathode (284.4 Ω cm2) was 780% higher than the 32.3 Ω cm2 of the anode, indicating that the cathode rather than the anode was limiting the MFC performance. The distributions of internal resistances for MFCs with different electrode areas and buffer concentrations are shown in Table 1. For all tested MFCs, the electrolyte resistance accounted for 36-78% of the total resistance, while the 8104

9


pmaxb mW/m2 1040 430 2656 5509 6708 19369

Based on the projected area of the anode.

resistance of the anode ranged from 2 to 35%. For the MFCs with equally sized anodes and cathodes (7 cm2) and 200 mM phosphate buffer, the anode only accounted for 5.4% of the internal resistance, while the cathode and the electrolyte each contributed 47.3%. Increasing the phosphate buffer concentration from 50 to 200 mM greatly decreased the total internal resistance from 208 Ω (MFC no. 2) to 85.9 Ω (MFC no. 1), whereas the contribution of electrolyte to the internal resistance decreased from 78.2 to 47.3%. The decrease of anode area from 7 cm2 to 1 cm2 only slightly increased the internal resistance to 235.7 Ω (MFC no. 3) because the electrolyte (50 mM phosphate buffer) was still limiting the performance of the MFC. The MFC (no. 4) with a 1 cm2 anode, 7 cm2 cathode, and 200 mM phosphate buffer resulted in a more balanced internal resistance distribution. Doubling the surface area of the cathode resulted in an 18% decrease in internal resistance from 113.6 Ω to 93.3 Ω (MFC no. 5). The calculated anodic maximum power density was only 430 mW/m2 for MFC no. 2 with equally sized electrodes and 50 mM buffer. Increasing the cathode/anode area ratio and the buffer concentration can greatly increase the anodic power density. A power density of 6860 mW/m2 was achieved with a cathode/anode area ratio of 14 and 200 mM buffer, indicating that future studies on this kind of MFC system should focus on the development of high performance cathode materials, buffer solution, and reactor design to minimize the internal resistances. Although platinum cathode has achieved very high current density in chemical fuel cells, the performance of the cathode in MFCs is greatly reduced due to the extremely low availability of protons at the catalytic sites at near neutral pH condition (10). New cathode design should focus on the structure of cathode to improve the proton availability. The reduction in electrolyte resistance can be achieved through the increase in pH buffer concentration and the reduction in electrode spacing. In a recent study, the electrolyte resistances of MFCs with double cloth electrode assemblies (CEAs) (14 cm2, 100 mM PBS) were reduced to about one tenth of that in MFC no. 1 (7 cm2, 200 mM PBS) due to the much smaller electrode spacing in CEA MFCs (about 1 mm). A maximum anodic power density of 19.4 W/ m2 is predicted assuming the carbon cloth anode, microbial consortia, and biofilm thickness and activity are the same and assuming both cathode resistance and electrolyte resistance are neglectable. Such a power density is about 1 order of magnitude higher than the maximum power density achieved using the state-of-the-art MFCs. Two-Chamber MFC System. Membrance Resistance. Oh and Logan tested the power generation in two-chamber MFCs with various electrodes and PEM areas (6). Equation 8 can be modified to fit the data presented in that study. Figure 5 in Oh and Logan′s paper (6) illustrated the effect of the PEM area on the power generation of MFCs using dissolved oxygen or ferricyanide as electron acceptors. The anodic resistance, cathodic resistance and electrolyte resistance were fixed in this case since the electrode areas (22.5 cm2), reactor

FIGURE 3. Comparison of (A) the observed voltage and (B) anodic power density (points) and the calculated results (lines) of MFCs with different PEM areas (3.5 cm2, 6.2 cm2, and 30.6 cm2). The open marks and dashed lines are the results for air cathode MFCs (denoted air), and solid marks and solid lines are the results for MFCs using ferricyanide as the catholyte (denoted FeCN). Inserts show the correlation between the observed and the calculated voltage/power density. Original data were from Figure 5 of ref (6). configuration, and electrolyte were unchanged. Therefore, eq 8 can be simplified as E ) Eb - (Rs + rm ⁄ Sm)I

(14)

where E (V) is the external voltage; Eb (V) is the linear extrapolation open circuit voltage; Rs (Ω) is the sum of anodic resistance, cathodic resistance, and electrolyte resistance; rm (Ω cm2) is the ASR of the PEM; Sm (cm2) is the surface area of the PEM; and I (A) is the current. Similarly, to calculate the power P (W), eq 10 can be simplified as P ) EbI - (Rs + rm ⁄ Sm)I2

(15)

Both eqs 14 and 15 fit well with the data from Figure 5 of ref 6 with a voltage greater than 0.1 V. Figure 3 illustrates the fitting results using eq 14 for air cathode MFCs and eq 15 for ferricyanide cathode MFCs. The calculated Eb, Rs, Rm, and Pmax are listed in Table 2. The calculated voltages and powers agree very well with the experimental results with correlation coefficients of R2 ) 0.9924 (n ) 59) for power and R2 ) 0.9632 (n ) 59) for voltage. The calculated Eb (0.739 V) of MFCs using ferricyanide was higher than the 0.652 V of air cathode MFCs. The relatively lower internal resistances of MFCs using ferricyanide also contributed to the improved power generation. The major

TABLE 2. Calculated Internal Resistance Distribution and Maximum Power of Two-Chamber MFCs with Different PEM Areas MFCa

Sm cm2

Eb V

Rs Ω (%)

air-3.5 3.5 0.652 145 (16%) air-6.2 6.2 0.652 145 (25%) air-30.6 30.5 0.652 145 (62%) FeCN-3.5 3.5 0.739 102 (14%) FeCN-6.2 6.2 0.739 102 (22%) FeCN-30.6 30.5 0.739 102 (58%)

Rm Ω (%)

Rint Ω

Pmax mW

780 (84%) 440 (75%) 89 (38%) 633 (86%) 357 (78%) 72 (42%)

925 585 234 735 459 174

0.115 0.182 0.454 0.186 0.297 0.783

a Air: MFCs using dissolved oxygen as electron acceptor; FeCN: MFCs using ferricyanide as electron acceptor; The number indicates the area of the PEM (cm2).

(38-86%) contribution of the PEM to the total internal resistance indicates that the PEM may limit the performance of two-chamber MFCs. Therefore, the role of the PEM needs to be carefully evaluated and considered in the design and operation of MFCs. Internal Resistance Distribution. The data in Figure 4 of ref 6 can be used to further calculate the distribution of the Rs. In this figure, the areas of the anode, cathode and PEM were varied, while the reactor dimension and electrolyte were fixed, allowing eq 10 to be rewritten as VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8105

P ) EbI - (ra ⁄ Sa + rc ⁄ Sc + rm ⁄ Sm + Re)I2

(16)

To reduce the number of variables, the values of Eb (0.652 V) and rm (2729 Ω cm2) obtained above were used directly in fitting the experimental results using eq 16. All the data from the figure were used except two points with measured voltages of 0.079 V, which fell out of the linear region of the polarization curve. Figure 4 shows that the calculated power agrees well with the experimental data (R2 ) 0.9861, n ) 23). The calculated internal resistance distribution of the twochamber MFCs is listed in Table 3. Single-Chamber MFCs vs Two-Chamber MFCs. A comparison of the calculated parameters of single-chamber and two-chamber MFCs is shown in Table 3. The calculated electrolyte resistances for the two chamber MFCs (33.6 Ω) were lower than that of single-chamber MFCs (40.7 Ω), possibly due to differences in reactor dimensions and electrolyte chemistry. The Eb value (0.5004 V) of the singlechamber air cathode MFCs was lower than the 0.652 V of two-chamber MFCs which may have resulted from the higher oxygen cross-over in the single-chamber system. However, the much lower internal resistances of single-chamber MFCs make it possible to produce more than a 10-fold increase in power density. The anodic ASR of two chamber MFCs with carbon paper anodes were 24 times higher than the single-chamber MFCs with carbon cloth anodes, especially considering similar performances were seen in single-chamber MFCs using both materials (8). The differences in microbial consortia may be the major reason for such a difference. Besides the difference in the inocula, the high resistances of the membrane, electrolyte and cathode in the two-chamber MFCs, which limited the anodic current density, may also affect the development of biofilms and highly efficient microbial consortia. Although both the single-chamber MFCs and the two-chamber air cathode MFCs used platinum as the catalyst and oxygen as the electron acceptor, the cathodic ASR of the two-chamber system was 8 times higher than that of the single-chamber system, indicating direct air is much better than dissolved oxygen in supplying oxygen to the catalytic sites. The difference in supplying oxygen might be the result from the much higher oxygen concentration and mass transfer rate in air than those in water. Although the microorganisms and exopolymeric substance (EPS) on the cathode of a single-chamber MFC might increase the resistances of the cathode, the cathode resistance of the singlechamber MFC was less than 1/8 of the cathode or membrane resistance of the two-chamber MFC (Table 3), indicating the resistance caused by microbes and EPS was not significant. The high membrane resistance (2729 Ω cm2), which is not present in membraneless single-chamber MFCs, is another important factor that caused the poor performance of the two chamber system. The inclusion of a PEM, such as Nafion 117, needs to be carefully considered due to its high area specific resistance, which could contribute 38-86% of the total internal resistances of two-chamber MFCs. The total ASR of the single-chamber MFC with equal sized electrodes (1.7 cm spacing, 200 mM PBS) was 601 Ω cm2. It is lower than the 3220 Ω cm2 of an upflow MFC with interior cathode (membrane surface area 188 cm2) (11), but higher than the ohmic ASRs of 49-56 Ω cm2 of MFCs with cloth-electrode assemblies (CEAs) (4, 10) and 56 Ω cm2 (based on the cathode projected area of 7 cm2) of a brush-anode MFC using 0.2 M phosphate buffer (2). Implications. The method developed in this study provides a versatile tool to evaluate the limiting factor(s) and to simulate/predict the power generation of an MFC. In addition to the tested single-chamber and two-chamber MFC systems, the method can also be easily adapted to other MFC systems. The calculated parameters, such as ra, rc, rm, and Re 8106

9


FIGURE 4. Comparison of observed (points) and calculated (lines) power of MFCs with different electrode and PEM areas. Original data were from Figure 4 of ref (6).

TABLE 3. Comparison of the Calculated Parameters of Single-Chamber and Two-Chamber Air Cathode MFCs MFC singlechamber twochamber

Eb V

ra Ω cm2

rc Ω cm2

rm Ω cm2

re Ω cm2

Re Ω

0.500

32.3

284

0

285a,b

40.7a

2566

2729

1028c

33.6

0.652

a

820

b

200 mM phosphate buffer. Based on cross-section area of 7 cm2. c Based on cross-section area of 30.6 cm2.

are linked directly to the components of MFCs including the anode, cathode, membrane, and electrolyte. These parameters can then be directly used to evaluate the performance of the MFC components and the whole MFC system. A comparison of these parameters can be utilized to determine the limiting factors of MFCs. In addition, these parameters can also be compared among different types of MFCs, providing valuable information in reactor design, electrode/ membrane material selection and development, and electrolyte selection. It should be noted that this is a simplified method and is only suitable for MFCs with near linear polarization curves, which is the case for the majority of the reported MFC systems (5). Further improvement can be made to simulate the non-linear polarization curves through the incorporation of the Butler-Volmer equation for electron transfer rates, first-order mass transport rates, and ohmic potential drop across the membrane and in solution (20). The power generation of MFCs with different anode surface areas, cathode surface areas, and phosphate buffer concentrations indicated that carbon cloth anodes with well developed biofilms did not limit the power generation of the single-chamber air cathode MFCs. The electrolyte resistance, membrane resistance, and cathode resistance contribute the most to the internal resistance of most current MFC systems. Significant enhancement in the performance of MFCs can be achieved through the development of high efficiency cathodes and the reduction of the internal resistance by reducing the electrode spacing, increasing the concentration of pH buffers, and selecting membranes with low resistance. An anodic power density of 6860 mW/m2 was achieved using single-chamber air cathode MFCs with an anode/cathode area ratio of 1/14 and 200 mM phosphate buffer, indicating the great potential of MFC technology.

Acknowledgments This research was partially supported by Agricultural Research Foundation and Oregon State University General Research

Fund. We acknowledge Sang-Eun Oh and Bruce Logan for providing the data of Figures 4 and 5 in ref 6.

Supporting Information Available The experimental results and discussion on the contribution of biofilms and planktonic bacteria to power generation of single-chamber MFCs. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Cheng, S.; Logan, B. E. Ammonia treatment of carbon cloth anodes to enhance power generation of microbial fuel cells. Electrochem. Commun. 2007, 9, 492–496. (2) Logan, S.; Cheng, V.; Watson, G. Estadt, Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environ. Sci. Technol. 2007, 41, 3341–3346. (3) Ringeisen, B. R.; Henderson, E.; Wu, P. K.; Pietron, J.; Ray, R.; Little, B.; Biffinger, J. C.; Jones-Meehan, J. M. High power denstity from a miniature microbial fuel cell using Shewanella oneidensis DSP 10. Environ. Sci. Technol. 2006, 40, 2629–2634. (4) Fan, Y.; Hu, H.; Liu, H. Enhanced columbic efficiency and power density of air-cathode microbial fuel cells with an improved cell configuration. J. Power Source 2007, 171 (2)), 348–354. (5) Logan, B. E.; Hamelers, B.; Rozendal, R.; Schroder, U.; Keller, J.; Freguia, S.; Aelterman, P.; Verstraete, W.; Rabaey, K. Microbial fuel cells: methodology and technology. Environ. Sci. Technol. 2006, 40, 5181–5192. (6) Oh, S. E.; Logan, B. E. Proton exchange membrane and electrode surface areas as factors that affect power generation in microbial fuel cells. Appl. Microbiol. Biotechnol. 2006, 70, 162–169. (7) Oh, S. E.; Min, B.; Logan, B. E. Cathode perfermance as a factor in electricity generation in microbial fuel cells. Environ. Sci. Technol. 2004, 38, 4900–4904. (8) Liu, H.; Cheng, S.; 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, 5488– 5493. (9) Wagner, N. Characterization of membrane electrode assemblies in polymer electrolyte fuel cells using a.c. impedance spectroscopy. J. Appl. Electrochem. 2002, 32, 859–863.

(10) Fan, Y.; Hu, H.; Liu, H. Sustainable power generation in microbial fuel cells using bicarbonate buffer and proton transfer mechanisms. Environ. Sci. Technol. 2007, 41 (23), 81548158. (11) 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, 5212–5217. (12) Larminie, J.; Dicks, A. Fuel Cell Systems Explained, John Wiley & Sons.: New York, 2000. (13) Aelterman, P.; Rabaey, K.; Pham, H. T. Continuous electricity generation at high voltages and currents using stacked microbial fuel cells. Environ. Sci. Technol. 2006, 40, 3388–3394. (14) Picioreanu, C.; Head, I. M.; Katuri, K. P.; van Loosdrecht, M. C.; Scott, K. A computational model for biofilm-based microbial fuel cells. Water. Res. 2007, 41, 2921–2940. (15) Kato Marcus, A.; Torres, C. I.; Rittmann, B. E. Conduction-based modeling of the biofilm anode of a microbial fuel cell. Biotechnol. Bioeng. 2007, 98 (6)), 1171–82. (16) Torres, C. S.; Kato Marcus, A.; Rittmann, B. E. Kinetics of consumption of fermentation products by anode-respiring bacteria. Environ. Sci. Technol. 2007, 77, 689–697. (17) 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, 4040–4046. (18) Cheng, S.; Liu, H.; Logan, B. E. Increased power and Coulombic efficiency of single-chamber microbial fuel cells through an improved cathode structure. Electrochem. Commun. 2006, 8, 489–494. (19) Rabaey, K.; Boon, N.; Ho¨fte, M.; Verstraete, W. Microbial phenazine production enhances electron transfer in biofuel cells. Environ. Sci. Technol. 2005, 39, 3401–3408. (20) Westall, J. C., Reimers, C. E., Stecher III, H. A., Newman, D. K., Lies, D. Mathematical Models for Optimization of the Performance of Biofuel Cells. Abstracts of Papers, 230th National Meeting of the American Chemical Society; American Chemical Society, August 28 to September 01, Washington, DC, 2005; Vol. 230, p U1748.

ES801229J

VOL. 42, NO. 21, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8107

Quantification of the Internal Resistance Distribution of Microbial Fuel

Recommend Documents