- Email: [email protected]
Microfiltration membrane performance in two-chamber microbial fuel cells
Biochemical Engineering Journal 52 (2010) 194–198
Contents lists available at ScienceDirect
Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej
Microfiltration membrane performance in two-chamber microbial fuel cells Xinhua Tang a,b , Kun Guo a,b , Haoran Li a,∗ , Zhuwei Du a , Jinglei Tian c a
State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China Graduate University of Chinese Academy of Sciences, Beijing 100049, China c School of Civil and Environmental Engineering, University of Science and Technology Beijing, Beijing 100083, China b
a r t i c l e
i n f o
Article history: Received 15 July 2010 Received in revised form 10 August 2010 Accepted 15 August 2010
Keywords: Microfiltration membrane Microbial fuel cell COD Coulombic efficiency Power density Proton exchange membranes
a b s t r a c t Proton exchange membranes (PEMs) are typically used in two-chamber microbial fuel cells (MFCs) to separate the anode and cathode chambers while allowing protons to pass between the chambers. However, PEMs such as Nafion are not cost-effective. To reduce the cost of MFCs, we examined the performances of cellulose acetate microfiltration membranes in a two-chamber microbial fuel cell using acetate. The internal resistance, the maximum power density and the coulombic efficiency (CE) of the microfiltration membrane MFC (MMMFC) were 263 , 0.831 ± 0.016 W/m2 and 38.5 ± 3.5%, respectively, in a fed-batch mode, while the corresponding values of the MFC using a PEM were 267 , 0.872 ± 0.021 W/m2 and 74.7 ± 4.6%, respectively. We further used the MMMFC for poultry wastewater treatment. The maximum power density of 0.746 ± 0.024 W/m2 and CE of 35.3 ± 3.2% were achieved when the poultry wastewater containing 566 mg/L COD was used, removing 81.6 ± 6.6% of the COD. These results demonstrate microfiltration membranes, compared with PEMs, have a similar internal resistance and reduce pH gradient across the membrane. They parallel PEMs in maximum power density, while CE is much lower due to the oxygen and substrate diffusion. The MMMFC was effective for poultry wastewater treatment with high COD removal. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Microbial fuel cells (MFCs) use bacteria to produce electricity from the degradation of organic matter [1–3]. Electrons released by bacteria are transferred from the anode to the cathode through an external circuit and typically combine with oxygen and protons to form water. Two-chamber MFCs consist of an anode and a cathode chamber separated by a membrane, which can prevent the bacteria transfer from the anode to the cathode chamber and reduce the oxygen diffusion to the anode. Most of the reported two-chamber MFCs used PEMs (typically Nafion); a few used anion exchange membranes [4–9]. However, these membranes are expensive and approximately account for 38% of the capital costs in MFCs [10]. Therefore, new materials should be explored to make a step forward towards practical implementation of MFCs. Ultrafiltration membranes, J-cloth and Zirfon membranes have been studied in MFCs, and these separators could reduce oxygen diffusion and significantly improve coulombic efficiency [7,11,12]. In this study, we used microfiltration membrane because it has not been previously examined, although it can effectively separate the bacteria from the cathode and reduce the oxygen diffusion to the anode while permitting ion transport. In addition, it can remarkably reduce the capital
∗ Corresponding author. Tel.: +86 10 82627064; fax: +86 10 82627064. E-mail address: [email protected] (H. Li). 1369-703X/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2010.08.007
costs of MFC construction. So, we compared the performance of microfiltration membrane with Nafion membrane when acetate was used in a two-chamber MFC. Furthermore, since most MFCs used pure compounds such as glucose, acetate, sucrose and an amino acid and it has not been studied previously whether MFCs could be used for chicken farm wastewater treatment, we therefore investigated the effectiveness of the two-chamber MFC with a microfiltration membrane for chicken farm wastewater treatment [13–16]. The performance of the MFC was evaluated in terms of internal resistance, maximum power density, coulombic efficiency and COD removal rate at a constant temperature of 20 ◦ C.
2. Materials and methods 2.1. Poultry wastewater and acetate solution Wastewater was collected from Beijing Deqingyuan Agricultural Technology farm (Beijing, China), and stored at 4 ◦ C for 1 week before being used. The wastewater has a COD of 500–600 mg/L and a pH of 6.5. In the tests, 0.1 g KCl, 0.2 g NH4 Cl, 0.6 g NaH2 PO4 , 2.9 g NaCl, 2.5 g NaHCO3 , 5 mL vitamin and 10 mL trace mineral solution [17] were added per liter wastewater, deionized water and 9 mM acetate solution to obtain the amended wastewater, cathodic electrolyte and acetate medium (COD was 560 mg/L) respectively.
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
195
2.2. MFC construction MFCs were constructed by joining two Plexiglas cylindrical chambers 18 cm high by 3.8 cm in diameter (empty volume of 200 mL) with a glass tube containing a proton exchange membrane (PEM) (Nafion 117, thickness 0.19 mm, Dupont Co., DE), or a microfiltration membrane (pore diameter 0.45 m, film thickness 0.13 mm, Q/IEFJ01, Shanghai Xingya Co., Shanghai), clamped between the flattened ends of the glass tubes (inner diameter 2.0 cm). The PEM was sequentially boiled in H2 O2 (30%), deionized water, 0.5 M H2 SO4 , and deionized water (each time for 1 h). Microfiltration membrane was stored in deionized water and gently rinsed prior to use. The anode and the cathode (total electrode spacing of 10 cm) were graphite rod (8 cm high and 2.5 cm in diameter). The cathode was coated with 0.25 mg/cm2 Pt. Titanium wire was used for the connection of the external circuit to the electrodes. 2.3. MFC tests Five MFCs (MFC0-MFC4) were operated in fed-batch mode with the external resistance of 250 (except as indicated) at a fixed temperature of 20 ◦ C. The MFCs were inoculated with a mixed bacterial culture from the anode of a two-chamber MFC, which was originally inoculated with domestic wastewater (Gaobeidian Wastewater Treatment Plant, Beijing) and has been operated for more than 2 years. One hundred and fifty milliliter acetate medium was pumped into the anode of MFC1 with PEM, MFC2 with a microfiltration membrane and MFC0 without a membrane; MFC3 and MFC4 with a microfiltration membrane were filled with 150 mL amended wastewater while MFC4 was operated in open circuit mode. All the cathode chambers were filled with cathodic electrolyte. For each batch cycle, the anode chambers were sparged using ultra high purity nitrogen for 20 min before the operation. During operation, the anode chambers were maintained under aseptic anaerobic condition, while the cathode chambers were sparged with air at 120 mL/min to provide oxidant. The system was considered to be stable when the maximum voltage output was reproducible after refilling the reactor with medium more than two times. The medium in the reactors was refilled when the output voltage dropped below 20 mV. 2.4. Calculations and analysis Microbial growth on the anodic surface of MFCs was investigated by scanning electron microscopy (JSM 6700F, JEOL Ltd.). Voltage was measured using a data acquisition system (AD8201H, Ribohua Co.,China) every 30 s and converted to power density, P (W/m2 ), according to P = IV/A, where I (A) is the current, V (V) is the voltage and A (m2 ) is the cross-sectional area of the membrane. Polarization curve was obtained by varying the circuit resistance. The internal resistance of the cell Rint , was calculated from the slope of V and I using: V = Ecell − IRint
(1)
Coulombic efficiency was calculated as EC = Cp /CTj × 100%. Cp (C) is the total Coulombs calculated by integrating the current over time, calculated as:
Cp =
V dt R
(2)
Fig. 1. Scanning electron microscopy image of bacteria on the anodic surface of MFCs.
where F is Faraday’s constant (96,500 C/mol of electrons), bj is the number of mole of electrons produced per mole of substrate (bw = 4, ba = 8), cj (mol/L) is the substrate concentration, and v (L) is the liquid volume (150 mL). COD was measured according to potassium dichromate method [18]. Acetate concentrations were analyzed using a gas chromatograph (Agilent, 6890) equipped with a flame ionization detector and a 30 m × 0.32 mm × 0.5 m DB-FFAP fused silica capillary column followed the same procedure described by Liu and Logan [19]. pH of the anode and cathode chambers were measured by pH meter (PHS-25 pH meter, Shanghai Precision & Scientific Instrument Co., Ltd.) when the system was stable. Dissolved oxygen analyzer (Model JPSJ-605 D.O. Analyzer, Shanghai Precision & Scientific Instrument Co,. Ltd) was placed in the anode chamber. The mass transfer coefficient of oxygen in the membrane, ko , was determined by monitoring the DO concentration over time and using the equation by Kim and co-workers [7]: ko = −
V ln At
c − c 0 1 c0
(4)
where V is the liquid volume in the anode chamber, A is the membrane cross-sectional area, c0 is the saturated oxygen concentration in the cathode chamber and c1 is the DO in the anode chamber at time t. The diffusion coefficient Do was calculated as Do = Ko L, where L is the membrane thickness. 3. Results 3.1. Microbial enrichment Fig. 1 shows a biofilm was formed on the electrode surface by bacteria in the anode chamber when MFCs were stable. The image clearly demonstrated that bacteria could easily attach to the graphite surface and form a multilayer biofilm, which was considered to play a significant role in electron transfer from bacteria to electrode [3,9].
where R is the external resistance. CTj (C) is the theoretical amount of Coulombs that can be produced from either wastewater (j = w) or acetate (j = a), calculated as:
3.2. pH
CTj = Fbj cj v
The pH in the anode and cathode chambers and DO in the anode chambers were measured when the system was stable. The values
(3)
196
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
Table 1 Oxygen mass transfer and diffusion coefficient. Membrane property
Nafion
Microfiltration membrane
Thickness (mm) ko (10−4 cm/s) Do (10−6 cm2 /s)
0.19 1.4 2.7
0.13 5.9 7.7
were taken at the geometrically center of each chamber for 3 times at an interval of 30 min when the MFCs was stable. For MFC0, the pH in the two chambers was nearly the same (7.0 in the anode chamber and 7.2 in the cathode chamber). However, the pH of the anode and cathode chambers in MFC1 was 6.4 and 9.5, respectively, a total balance of 3.1 pH units; while the corresponding values in MFC2 were 6.9 and 8.0. The results demonstrate microfiltration membrane was effective in decreasing the anode and cathode pH gradient. 3.3. Oxygen mass transfer and diffusion coefficient The oxygen mass transfer coefficient for microfiltration membrane of 5.9 × 10−4 cm/s was larger than that of Nafion 117, which was 1.4 × 10−4 cm/s (Table 1). A high value of ko indicated increased diffusion of oxygen from the cathode to the anode chamber. These were consistent with the result that CE in MFC1 (74.7 ± 4.6%) was larger than (CE 38.5 ± 3.5%) in MFC2. 3.4. Internal resistances Voltage generation cycles of MFCs using acetate were reproducible after 3 feeding cycles with fresh media. By changing the circuit external resistances, polarization curves were obtained. The internal resistance of MFC2 was 263 , while it was 267 in MFC1 (Fig. 2). These internal resistances were much higher than many previously reported MFCs, due to the considerably long electrode spacing [20]. However, the results suggested that microfiltration membrane could maintain a similar internal resistance as PEM. 3.5. Power generation and CE Voltages increased remarkably and reached the maximum after about 4 h both in MFC1 and MFC2 (Fig. 3A). The voltage output in MFC1 ranged from 250 to 260 mV in the next 180 h. The maximum voltage in MFC2 was similar to that of MFC1. However,
Fig. 3. Time–voltage curves (A) and power generation (B) of MFC1 and MFC2 (10 mM acetate) (error bars SD based on the voltages in experiments run in triplicate).
the maximum power output only maintained for approximately 100 h. As a result, the CE in MFC1 was 74.7 ± 4.6% while the CE in MFC2 was 38.5 ± 3.5%. To obtain the polarization curve, the current density was calculated and plotted against power density at different resistances. The maximum power densities were 0.872 ± 0.021 W/m2 normalized by the membrane area in MFC1 and 0.831 ± 0.016 W/m2 in MFC2 (Fig. 3B). These results indicated that microfiltration membranes were as effective as PEMs in power output, but a much lower CE was observed due to the oxygen flux to the anode chamber from the microfiltration membrane.
3.6. Power generation, coulombic efficiency and COD removal for wastewater treatment
Fig. 2. Polarization curves of MFC1 and MFC2: R indicates the internal resistance.
The voltage across the circuit containing a 250 resistor increased to 240–250 mV for poultry wastewater containing up to 566 mg /L COD, lasting for about 70 h and then dropped to 20 mV slowly (Fig. 4A). The maximum power density of MFC3 for wastewater treatment was 0.746 ± 0.024 W/m2 , slightly lower than that of MFC2 (0.831 ± 0.016 W/m2 ) using acetate (Fig. 4B). The CE was 35.3 ± 3.2% and a COD removal rate of up to 81.6 ± 6.6% was achieved in MFC3. MFC4 with an open circuit was used as a control system and the COD removal was 46.7 ± 4.2%.
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
Fig. 4. Time–voltage curve (A) and voltage and power (B) generated in MFC3 (566 mg COD/L) (error bars SD based on the voltages in experiments run in triplicate).
4. Discussion Microbial fuel cell is a novel and promising future technology for the production of renewable energy from organic material in wastewater. Substantial progress has been made in the field in the past few years. For example, power production with MFCs using oxygen at the cathode has increased logarithmically, from less than 100 mW/m2 before 2001 to 5000 mW/m2 in 2009 (normalized to the cathode surface area) [21]. However, implementation of the technology for wastewater treatment is not straightforward because certain challenges need to be solved. For instance, metabolic diversity has to be studied to understand the high diversity of microbial species, maximum power production need to be exploited and electrochemical losses has to be minimized. Among all of the problems, economical challenge is the greatest. It is estimated that the capital cost is 8D /kg COD currently using MFCs for wastewater treatment, which is much higher than conventional methods such as activated sludge and anaerobic digestion [10]. Nonetheless, new materials might bring down the cost in the future and electricity production from MFCs can offset part of the capital cost, turning MFCs to be a cost-effective alternative [11,22]. In some MFCs, membranes were not used in order to reduce the cost of construction. The internal resistance was reduced and the maximum power density increased when membrane was not used. However, an evident drawback of membrane-less operation of MFCs was that it could lead to a reduced electron recovery. Membrane-less operation of MFCs could cause an increased contact between oxygen from the cathode chamber and organic material in the anode chamber, which would lead to direct aerobic conversion of the organic materials [19]. Therefore, membranes need be included in a two-chamber MFC to maintain a high coulom-
197
bic efficiency. Among the various membranes used in MFCs, Nafion membrane is most commonly used in two-chamber MFCs, probably due to its excellent proton conductivity and thermal and mechanical stability in fuel cells [7,8,10]. However, the expensive Nafion does not show great advantage over other membranes in MFC application because of the low proton concentration, moderate reaction condition and potential for biofouling in MFCs [5,11]. The main functions of membranes in MFCs are to prevent the spread of bacteria from the anode to the cathode chamber, reduce or block the oxygen diffusion from the cathode to anode while permitting ion transport. Cellulose acetate microfiltration membrane is an alternative: the 0.45 m pore can effectively cut-off the migration of exoelectrogens (bacteria able to transfer electrons outside the cell) from the anode to the cathode because bacteria typically range in size from 0.5 to 2 m in width or diameter, and 1–10 m in length; they allow ion transport and reduce oxygen diffusion to the anode. Furthermore, cellulose acetate microfiltration membrane can greatly lower the cost of MFC construction, since Nafion 117 is approximately 400D /m2 while the cellulose acetate microfiltration membrane is about 40D /m2 [11]. Due to the reasonable properties and its cheaper cost, cellulose acetate microfiltration membrane should be considered in MFCs. Although a Nafion membrane can be effective in blocking oxygen, it can also inhibit the transport of protons, which was demonstrated by the 3.1 unit difference of pH between the anode and cathode chambers. The anode reactions in MFCs were proton producing and the cathode reactions were proton consuming, which caused the pH decrease in the anode while pH increase in the cathode. From the Nernst equation, the resulting membrane pH gradient causes a potential loss of 0.059 V per pH unit [23]. Therefore, the 3.1 pH differences in MFC1 had a potential loss of about 0.183 V. However, the pH gradient in MFC2 was 1.1, indicating that microfiltration membrane could reduce the potential loss. Kim et al. demonstrated that the oxygen transfer coefficient in ultrafiltration membranes increased with the membrane pore size [7]. Microfiltration membrane pore was much larger than ultrafiltration membrane pore, and the oxygen transfer coefficient of microfiltration (5.9 × 10−4 cm/s) in this study was much larger than that of ultrafiltration membranes (0.19 × 10−4 , 0.41 × 10−4 , 0.42 × 10−4 cm/s) in the study of Kim et al. [7]. As a result, more oxygen could be diffused to the anode chamber within the same time, which led to direct aerobic conversion of the materials. On the other hand, it was reasonable that the acetate transfer coefficient of the microfiltration membrane we used could be greater than that of ultrafiltration membrane due to the much larger pore diameter. Consequently, increased acetate would be transported into the cathode chamber of MFC2 and lead to direct aerobic conversion as well. The diffusion of oxygen to the anode combined with the acetate diffusion to the cathode resulted the decreasing of CE: the CE in MFC2 was only 38.5 ± 3.5% while it was 74.7 ± 4.6% in MFC1. In a future study, we are planning to use a thicker microfiltration membrane in the hope of reducing the oxygen and acetate diffusion and improving CE to a higher level. Though microfiltration membrane had a larger oxygen transfer coefficient than that of ultrafiltration membrane, it had almost the same internal resistance with Nafion 117. However, the internal resistance increased from 91 to 1814 when Nafion 117 was replaced by an ultrafiltration membrane (0.5 kDa), which significantly reduced the maximum power density [7]. In this study, the maximum power densities were 0.872 ± 0.021 W/m2 in MFC1 and 0.831 ± 0.016 W/m2 in MFC2, suggesting that microfiltration membranes were as effective as PEMs in power output. High power density could be achieved by J-colth in MFCs [12], however, Jcolth generally have large pore that it cannot prevent bacterial migration to the cathode, which would cause potential loss and CE decrease.
198
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
When poultry wastewater containing 566 mg/L COD was used in MFC3, the maximum power density was 0.746 ± 0.024 W/m2 , slightly lower than that of MFC2 containing 560 mg/L COD (0.831 ± 0.016 W/m2 ). It was consistent with the previous research that MFCs using pure compounds produced higher power density than using actual wastewater [24]. The power density is known to be limited by high resistance resulting from large distances between the electrodes and correspondingly MFCs tends to have short electrodes spacing [5,7,14,16]. When Liu et al. decreased the electrode spacing from 4 to 2 cm in a single-chamber fedbatch MFC, the maximum power density increased from 720 to 1210 mW/m2 [25]. The CE was 35.3 ± 3.2% and the COD removal was 81.6 ± 6.6% in MFC3. The COD removal of MFC4 with an open circuit was 46.7 ± 4.2%. The CE of 35.3 ± 3.2% indicated considerable amount of organic matters were consumed aerobically and anaerobically. It was reported that a combination of 23.7% and 16.9% of the substrate were lost to aerobic biomass and anaerobic biomass and diffusion in two-bottle MFCs using a Nafion and ultrafiltration membrane (molecular cut-off of 3 kDa) respectively [7]. Because oxygen transfer coefficient and acetate transfer coefficient were larger in microfiltration membranes than that of Nafion and ultrafiltration membranes, a much higher percentage (46.7 ± 4.2%) of the substrate were removed in MFC4 with an open circuit compared with the results of Kim et al. [7]. More efforts are needed to tailor the thickness and pore size of the microfiltration membrane to improve CE. However, the high COD removal demonstrated microfiltration membrane was effective in actual wastewater treatment. Microfiltration membrane shows promising performances in MFCs: reducing pH gradient across the membrane, keeping a similar internal resistance and power density as PEM, and having a high percentage of COD removal for wastewater treatment. Furthermore, it considerably reduces the cost of MFC construction and therefore opens a new avenue for MFC reactor design to meet the practical implementation of wastewater treatment. Acknowledgements The research was supported by the National Science Foundation of China (No. 20876160) and the National High Technology Research and Development Program of China (863 Program) (No. 2007AA05Z158). References [1] S. Suzuki, I. Karube, T. Matsunaga, Application of a biochemical fuel cell to wastewaters, Biotechnol. Bioeng. Symp. 8 (1978) 501–511. [2] L.B. Wingard Jr., C.H. Shaw, J.F. Castner, Bioelectrochemical fuel cells, Enzyme Microb. Technol. 4 (1982) 137–142. [3] X. Tang, Z. Du, H. Li, Anodic electron shuttle mechanism based on 1-hydroxy-4aminoanthraquinone in microbial fuel cells, Electrochem. Commun. 12 (2010) 1140–1143.
[4] R.A. Rozendal, H.V.M. Hamelers, C.J.N. Buisman, Effects of membrane cation transport on pH and microbial fuel cell performance, Environ. Sci. Technol. 40 (2006) 5206–5211. [5] F. Zhao, F. Harnisch, U. Schröder, F. Scholz, P. Bogdanoff, I. Herrmann, Challenges and constraints of using oxygen cathodes in microbial fuel cells, Environ. Sci. Technol. 40 (2006) 5193–5199. [6] G.C. Gil, I. Chang, B.H. Kim, M. Kim, J. Jang, H.S. Park, H.J. Kim, Operational parameters affecting the performance of a mediator-less microbial fuel cell, Biosens. Bioelectron. 18 (2003) 327–334. [7] J. Kim, S. Cheng, S. Eunoh, B. Logan, Power generation using different cation, anion, and ultrafiltration membranes in microbial fuel cells, Environ. Sci. Technol. 38 (2004) 4040–4046. [8] S.E. Oh, B.E. Logan, Proton exchange membrane and electrode surface areas as factors that affect power generation in microbial fuel cells, Appl. Microbiol. Biotechnol. 70 (2006) 162–169. [9] Z. Liu, H. Li, Effects of bio- and abio-factors on electricity production in a mediatorless microbial fuel cell, Biochem. Eng. J. 36 (2007) 209– 214. [10] R.A. Rozendal, H.V.M. Hamelers, K. Rabaey, J. Keller, C.J.N. Buisman, Towards practical implementation of bioelectrochemical wastewater treatment, Trends Biotechnol. 26 (2008) 450–459. [11] D. Pant, G. Van Bogaert, M. De Smet, L. Diels, K. Vanbroekhoven, Use of novel permeable membrane and air cathodes in acetate microbial fuel cell, Electrochim. Acta. doi:10.1016/j.electacta.2009.11.086. (in press). [12] Y. Fan, H. Hu, H. Liu, Enhanced coulombic efficiency and power density of air–cathode microbial fuel cells with an improved cell configuration, J. Power Sources 171 (2007) 348–354. [13] K. Rabaey, G. Lissens, S.D. Siciliano, W. Verstraete, A microbial fuel cell capable of converting glucose to electricity at high rate and efficiency, Biotechnol. Lett. 25 (2003) 1531–1535. [14] H. Liu, S. Cheng, B.E. Logan, Production of electricity from acetate or butyrate in a single chamber microbial fuel cell, Environ. Sci. Technol. 39 (2005) 658– 662. [15] Z. He, N. Wagner, S.D. Minteer, L.T. Angenent, An upflow microbial fuel cell with an internal resistance by impedance spectroscopy, Environ. Sci. Technol. 40 (2006) 5212–5217. [16] B.E. Logan, C. Murano, K. Scott, N.D. Gray, I.M. Head, Electricity generation from cysteine in a microbial fuel cell, Water Res. 39 (2005) 942–952. [17] D.R. Lovley, E.J.P. Phillips, Novel mode of microbial energy metabolism: organism carbon oxidation coupled to dissimilatory reduction of iron and manganese, Appl. Environ. Microbiol. 54 (6) (1988) 1472–1480. [18] C.P.L. Grady, G.T. Daigger, H.C. Lim, Biological Wastewater Treatment, 2nd ed., Marcel Dekker, New York, 1999. [19] H. Liu, B.E. Logan, Electricity generation using an air–cathode single chamber microbial fuel cell in the presence and absence of a proton exchange membrane, Environ. Sci. Technol. 38 (2004) 4040–4046. [20] S. Cheng, H. Liu, B.E. Logan, Increased power generation in a continuous flow MFC with adjective flow through the porous anode and reduced electrode spacing, Environ. Sci. Technol. 40 (2006) 2426–2432. [21] B. Erable, L. Etcheverry, A. Bergel, Increased power from a two-chamber microbial fuel cell with a low-pH air–cathode compartment, Electrochem. Commun. 11 (2009) 619–622. [22] F. Zhang, S. Cheng, D. Pant, G. Van Bogaert, B.E. Logan, Power generation using an activated carbon and metal mesh cathode in a microbial fuel cell, Electrochem. Commun. 11 (2009) 2177–2179. [23] R.A. Rozendal, H.V. Hamelers, R.J. Molenkamp, C.J. Buisman, Performance of single chamber biocatalyzed electrolysis with different types of ion exchange membranes, Water Res. 41 (2007) 1984–1994. [24] D. Pant, G. Van Bogaert, L. Diels, K. Vanbroekhoven, A review of the substrates used in microbial fuel cells (MFCs) for sustainable energy production, Bioresour. Technol. 101 (2010) 1533–1543. [25] H. Liu, S. Cheng, B.E. Logan, Power generation in fedbatch microbial fuel cells as a function of ionic strength, temperature and reactor configuration, Environ. Sci. Technol. 39 (2005) 5488–5493.
Contents lists available at ScienceDirect
Biochemical Engineering Journal journal homepage: www.elsevier.com/locate/bej
Microfiltration membrane performance in two-chamber microbial fuel cells Xinhua Tang a,b , Kun Guo a,b , Haoran Li a,∗ , Zhuwei Du a , Jinglei Tian c a
State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China Graduate University of Chinese Academy of Sciences, Beijing 100049, China c School of Civil and Environmental Engineering, University of Science and Technology Beijing, Beijing 100083, China b
a r t i c l e
i n f o
Article history: Received 15 July 2010 Received in revised form 10 August 2010 Accepted 15 August 2010
Keywords: Microfiltration membrane Microbial fuel cell COD Coulombic efficiency Power density Proton exchange membranes
a b s t r a c t Proton exchange membranes (PEMs) are typically used in two-chamber microbial fuel cells (MFCs) to separate the anode and cathode chambers while allowing protons to pass between the chambers. However, PEMs such as Nafion are not cost-effective. To reduce the cost of MFCs, we examined the performances of cellulose acetate microfiltration membranes in a two-chamber microbial fuel cell using acetate. The internal resistance, the maximum power density and the coulombic efficiency (CE) of the microfiltration membrane MFC (MMMFC) were 263 , 0.831 ± 0.016 W/m2 and 38.5 ± 3.5%, respectively, in a fed-batch mode, while the corresponding values of the MFC using a PEM were 267 , 0.872 ± 0.021 W/m2 and 74.7 ± 4.6%, respectively. We further used the MMMFC for poultry wastewater treatment. The maximum power density of 0.746 ± 0.024 W/m2 and CE of 35.3 ± 3.2% were achieved when the poultry wastewater containing 566 mg/L COD was used, removing 81.6 ± 6.6% of the COD. These results demonstrate microfiltration membranes, compared with PEMs, have a similar internal resistance and reduce pH gradient across the membrane. They parallel PEMs in maximum power density, while CE is much lower due to the oxygen and substrate diffusion. The MMMFC was effective for poultry wastewater treatment with high COD removal. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Microbial fuel cells (MFCs) use bacteria to produce electricity from the degradation of organic matter [1–3]. Electrons released by bacteria are transferred from the anode to the cathode through an external circuit and typically combine with oxygen and protons to form water. Two-chamber MFCs consist of an anode and a cathode chamber separated by a membrane, which can prevent the bacteria transfer from the anode to the cathode chamber and reduce the oxygen diffusion to the anode. Most of the reported two-chamber MFCs used PEMs (typically Nafion); a few used anion exchange membranes [4–9]. However, these membranes are expensive and approximately account for 38% of the capital costs in MFCs [10]. Therefore, new materials should be explored to make a step forward towards practical implementation of MFCs. Ultrafiltration membranes, J-cloth and Zirfon membranes have been studied in MFCs, and these separators could reduce oxygen diffusion and significantly improve coulombic efficiency [7,11,12]. In this study, we used microfiltration membrane because it has not been previously examined, although it can effectively separate the bacteria from the cathode and reduce the oxygen diffusion to the anode while permitting ion transport. In addition, it can remarkably reduce the capital
∗ Corresponding author. Tel.: +86 10 82627064; fax: +86 10 82627064. E-mail address: [email protected] (H. Li). 1369-703X/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.bej.2010.08.007
costs of MFC construction. So, we compared the performance of microfiltration membrane with Nafion membrane when acetate was used in a two-chamber MFC. Furthermore, since most MFCs used pure compounds such as glucose, acetate, sucrose and an amino acid and it has not been studied previously whether MFCs could be used for chicken farm wastewater treatment, we therefore investigated the effectiveness of the two-chamber MFC with a microfiltration membrane for chicken farm wastewater treatment [13–16]. The performance of the MFC was evaluated in terms of internal resistance, maximum power density, coulombic efficiency and COD removal rate at a constant temperature of 20 ◦ C.
2. Materials and methods 2.1. Poultry wastewater and acetate solution Wastewater was collected from Beijing Deqingyuan Agricultural Technology farm (Beijing, China), and stored at 4 ◦ C for 1 week before being used. The wastewater has a COD of 500–600 mg/L and a pH of 6.5. In the tests, 0.1 g KCl, 0.2 g NH4 Cl, 0.6 g NaH2 PO4 , 2.9 g NaCl, 2.5 g NaHCO3 , 5 mL vitamin and 10 mL trace mineral solution [17] were added per liter wastewater, deionized water and 9 mM acetate solution to obtain the amended wastewater, cathodic electrolyte and acetate medium (COD was 560 mg/L) respectively.
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
195
2.2. MFC construction MFCs were constructed by joining two Plexiglas cylindrical chambers 18 cm high by 3.8 cm in diameter (empty volume of 200 mL) with a glass tube containing a proton exchange membrane (PEM) (Nafion 117, thickness 0.19 mm, Dupont Co., DE), or a microfiltration membrane (pore diameter 0.45 m, film thickness 0.13 mm, Q/IEFJ01, Shanghai Xingya Co., Shanghai), clamped between the flattened ends of the glass tubes (inner diameter 2.0 cm). The PEM was sequentially boiled in H2 O2 (30%), deionized water, 0.5 M H2 SO4 , and deionized water (each time for 1 h). Microfiltration membrane was stored in deionized water and gently rinsed prior to use. The anode and the cathode (total electrode spacing of 10 cm) were graphite rod (8 cm high and 2.5 cm in diameter). The cathode was coated with 0.25 mg/cm2 Pt. Titanium wire was used for the connection of the external circuit to the electrodes. 2.3. MFC tests Five MFCs (MFC0-MFC4) were operated in fed-batch mode with the external resistance of 250 (except as indicated) at a fixed temperature of 20 ◦ C. The MFCs were inoculated with a mixed bacterial culture from the anode of a two-chamber MFC, which was originally inoculated with domestic wastewater (Gaobeidian Wastewater Treatment Plant, Beijing) and has been operated for more than 2 years. One hundred and fifty milliliter acetate medium was pumped into the anode of MFC1 with PEM, MFC2 with a microfiltration membrane and MFC0 without a membrane; MFC3 and MFC4 with a microfiltration membrane were filled with 150 mL amended wastewater while MFC4 was operated in open circuit mode. All the cathode chambers were filled with cathodic electrolyte. For each batch cycle, the anode chambers were sparged using ultra high purity nitrogen for 20 min before the operation. During operation, the anode chambers were maintained under aseptic anaerobic condition, while the cathode chambers were sparged with air at 120 mL/min to provide oxidant. The system was considered to be stable when the maximum voltage output was reproducible after refilling the reactor with medium more than two times. The medium in the reactors was refilled when the output voltage dropped below 20 mV. 2.4. Calculations and analysis Microbial growth on the anodic surface of MFCs was investigated by scanning electron microscopy (JSM 6700F, JEOL Ltd.). Voltage was measured using a data acquisition system (AD8201H, Ribohua Co.,China) every 30 s and converted to power density, P (W/m2 ), according to P = IV/A, where I (A) is the current, V (V) is the voltage and A (m2 ) is the cross-sectional area of the membrane. Polarization curve was obtained by varying the circuit resistance. The internal resistance of the cell Rint , was calculated from the slope of V and I using: V = Ecell − IRint
(1)
Coulombic efficiency was calculated as EC = Cp /CTj × 100%. Cp (C) is the total Coulombs calculated by integrating the current over time, calculated as:
Cp =
V dt R
(2)
Fig. 1. Scanning electron microscopy image of bacteria on the anodic surface of MFCs.
where F is Faraday’s constant (96,500 C/mol of electrons), bj is the number of mole of electrons produced per mole of substrate (bw = 4, ba = 8), cj (mol/L) is the substrate concentration, and v (L) is the liquid volume (150 mL). COD was measured according to potassium dichromate method [18]. Acetate concentrations were analyzed using a gas chromatograph (Agilent, 6890) equipped with a flame ionization detector and a 30 m × 0.32 mm × 0.5 m DB-FFAP fused silica capillary column followed the same procedure described by Liu and Logan [19]. pH of the anode and cathode chambers were measured by pH meter (PHS-25 pH meter, Shanghai Precision & Scientific Instrument Co., Ltd.) when the system was stable. Dissolved oxygen analyzer (Model JPSJ-605 D.O. Analyzer, Shanghai Precision & Scientific Instrument Co,. Ltd) was placed in the anode chamber. The mass transfer coefficient of oxygen in the membrane, ko , was determined by monitoring the DO concentration over time and using the equation by Kim and co-workers [7]: ko = −
V ln At
c − c 0 1 c0
(4)
where V is the liquid volume in the anode chamber, A is the membrane cross-sectional area, c0 is the saturated oxygen concentration in the cathode chamber and c1 is the DO in the anode chamber at time t. The diffusion coefficient Do was calculated as Do = Ko L, where L is the membrane thickness. 3. Results 3.1. Microbial enrichment Fig. 1 shows a biofilm was formed on the electrode surface by bacteria in the anode chamber when MFCs were stable. The image clearly demonstrated that bacteria could easily attach to the graphite surface and form a multilayer biofilm, which was considered to play a significant role in electron transfer from bacteria to electrode [3,9].
where R is the external resistance. CTj (C) is the theoretical amount of Coulombs that can be produced from either wastewater (j = w) or acetate (j = a), calculated as:
3.2. pH
CTj = Fbj cj v
The pH in the anode and cathode chambers and DO in the anode chambers were measured when the system was stable. The values
(3)
196
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
Table 1 Oxygen mass transfer and diffusion coefficient. Membrane property
Nafion
Microfiltration membrane
Thickness (mm) ko (10−4 cm/s) Do (10−6 cm2 /s)
0.19 1.4 2.7
0.13 5.9 7.7
were taken at the geometrically center of each chamber for 3 times at an interval of 30 min when the MFCs was stable. For MFC0, the pH in the two chambers was nearly the same (7.0 in the anode chamber and 7.2 in the cathode chamber). However, the pH of the anode and cathode chambers in MFC1 was 6.4 and 9.5, respectively, a total balance of 3.1 pH units; while the corresponding values in MFC2 were 6.9 and 8.0. The results demonstrate microfiltration membrane was effective in decreasing the anode and cathode pH gradient. 3.3. Oxygen mass transfer and diffusion coefficient The oxygen mass transfer coefficient for microfiltration membrane of 5.9 × 10−4 cm/s was larger than that of Nafion 117, which was 1.4 × 10−4 cm/s (Table 1). A high value of ko indicated increased diffusion of oxygen from the cathode to the anode chamber. These were consistent with the result that CE in MFC1 (74.7 ± 4.6%) was larger than (CE 38.5 ± 3.5%) in MFC2. 3.4. Internal resistances Voltage generation cycles of MFCs using acetate were reproducible after 3 feeding cycles with fresh media. By changing the circuit external resistances, polarization curves were obtained. The internal resistance of MFC2 was 263 , while it was 267 in MFC1 (Fig. 2). These internal resistances were much higher than many previously reported MFCs, due to the considerably long electrode spacing [20]. However, the results suggested that microfiltration membrane could maintain a similar internal resistance as PEM. 3.5. Power generation and CE Voltages increased remarkably and reached the maximum after about 4 h both in MFC1 and MFC2 (Fig. 3A). The voltage output in MFC1 ranged from 250 to 260 mV in the next 180 h. The maximum voltage in MFC2 was similar to that of MFC1. However,
Fig. 3. Time–voltage curves (A) and power generation (B) of MFC1 and MFC2 (10 mM acetate) (error bars SD based on the voltages in experiments run in triplicate).
the maximum power output only maintained for approximately 100 h. As a result, the CE in MFC1 was 74.7 ± 4.6% while the CE in MFC2 was 38.5 ± 3.5%. To obtain the polarization curve, the current density was calculated and plotted against power density at different resistances. The maximum power densities were 0.872 ± 0.021 W/m2 normalized by the membrane area in MFC1 and 0.831 ± 0.016 W/m2 in MFC2 (Fig. 3B). These results indicated that microfiltration membranes were as effective as PEMs in power output, but a much lower CE was observed due to the oxygen flux to the anode chamber from the microfiltration membrane.
3.6. Power generation, coulombic efficiency and COD removal for wastewater treatment
Fig. 2. Polarization curves of MFC1 and MFC2: R indicates the internal resistance.
The voltage across the circuit containing a 250 resistor increased to 240–250 mV for poultry wastewater containing up to 566 mg /L COD, lasting for about 70 h and then dropped to 20 mV slowly (Fig. 4A). The maximum power density of MFC3 for wastewater treatment was 0.746 ± 0.024 W/m2 , slightly lower than that of MFC2 (0.831 ± 0.016 W/m2 ) using acetate (Fig. 4B). The CE was 35.3 ± 3.2% and a COD removal rate of up to 81.6 ± 6.6% was achieved in MFC3. MFC4 with an open circuit was used as a control system and the COD removal was 46.7 ± 4.2%.
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
Fig. 4. Time–voltage curve (A) and voltage and power (B) generated in MFC3 (566 mg COD/L) (error bars SD based on the voltages in experiments run in triplicate).
4. Discussion Microbial fuel cell is a novel and promising future technology for the production of renewable energy from organic material in wastewater. Substantial progress has been made in the field in the past few years. For example, power production with MFCs using oxygen at the cathode has increased logarithmically, from less than 100 mW/m2 before 2001 to 5000 mW/m2 in 2009 (normalized to the cathode surface area) [21]. However, implementation of the technology for wastewater treatment is not straightforward because certain challenges need to be solved. For instance, metabolic diversity has to be studied to understand the high diversity of microbial species, maximum power production need to be exploited and electrochemical losses has to be minimized. Among all of the problems, economical challenge is the greatest. It is estimated that the capital cost is 8D /kg COD currently using MFCs for wastewater treatment, which is much higher than conventional methods such as activated sludge and anaerobic digestion [10]. Nonetheless, new materials might bring down the cost in the future and electricity production from MFCs can offset part of the capital cost, turning MFCs to be a cost-effective alternative [11,22]. In some MFCs, membranes were not used in order to reduce the cost of construction. The internal resistance was reduced and the maximum power density increased when membrane was not used. However, an evident drawback of membrane-less operation of MFCs was that it could lead to a reduced electron recovery. Membrane-less operation of MFCs could cause an increased contact between oxygen from the cathode chamber and organic material in the anode chamber, which would lead to direct aerobic conversion of the organic materials [19]. Therefore, membranes need be included in a two-chamber MFC to maintain a high coulom-
197
bic efficiency. Among the various membranes used in MFCs, Nafion membrane is most commonly used in two-chamber MFCs, probably due to its excellent proton conductivity and thermal and mechanical stability in fuel cells [7,8,10]. However, the expensive Nafion does not show great advantage over other membranes in MFC application because of the low proton concentration, moderate reaction condition and potential for biofouling in MFCs [5,11]. The main functions of membranes in MFCs are to prevent the spread of bacteria from the anode to the cathode chamber, reduce or block the oxygen diffusion from the cathode to anode while permitting ion transport. Cellulose acetate microfiltration membrane is an alternative: the 0.45 m pore can effectively cut-off the migration of exoelectrogens (bacteria able to transfer electrons outside the cell) from the anode to the cathode because bacteria typically range in size from 0.5 to 2 m in width or diameter, and 1–10 m in length; they allow ion transport and reduce oxygen diffusion to the anode. Furthermore, cellulose acetate microfiltration membrane can greatly lower the cost of MFC construction, since Nafion 117 is approximately 400D /m2 while the cellulose acetate microfiltration membrane is about 40D /m2 [11]. Due to the reasonable properties and its cheaper cost, cellulose acetate microfiltration membrane should be considered in MFCs. Although a Nafion membrane can be effective in blocking oxygen, it can also inhibit the transport of protons, which was demonstrated by the 3.1 unit difference of pH between the anode and cathode chambers. The anode reactions in MFCs were proton producing and the cathode reactions were proton consuming, which caused the pH decrease in the anode while pH increase in the cathode. From the Nernst equation, the resulting membrane pH gradient causes a potential loss of 0.059 V per pH unit [23]. Therefore, the 3.1 pH differences in MFC1 had a potential loss of about 0.183 V. However, the pH gradient in MFC2 was 1.1, indicating that microfiltration membrane could reduce the potential loss. Kim et al. demonstrated that the oxygen transfer coefficient in ultrafiltration membranes increased with the membrane pore size [7]. Microfiltration membrane pore was much larger than ultrafiltration membrane pore, and the oxygen transfer coefficient of microfiltration (5.9 × 10−4 cm/s) in this study was much larger than that of ultrafiltration membranes (0.19 × 10−4 , 0.41 × 10−4 , 0.42 × 10−4 cm/s) in the study of Kim et al. [7]. As a result, more oxygen could be diffused to the anode chamber within the same time, which led to direct aerobic conversion of the materials. On the other hand, it was reasonable that the acetate transfer coefficient of the microfiltration membrane we used could be greater than that of ultrafiltration membrane due to the much larger pore diameter. Consequently, increased acetate would be transported into the cathode chamber of MFC2 and lead to direct aerobic conversion as well. The diffusion of oxygen to the anode combined with the acetate diffusion to the cathode resulted the decreasing of CE: the CE in MFC2 was only 38.5 ± 3.5% while it was 74.7 ± 4.6% in MFC1. In a future study, we are planning to use a thicker microfiltration membrane in the hope of reducing the oxygen and acetate diffusion and improving CE to a higher level. Though microfiltration membrane had a larger oxygen transfer coefficient than that of ultrafiltration membrane, it had almost the same internal resistance with Nafion 117. However, the internal resistance increased from 91 to 1814 when Nafion 117 was replaced by an ultrafiltration membrane (0.5 kDa), which significantly reduced the maximum power density [7]. In this study, the maximum power densities were 0.872 ± 0.021 W/m2 in MFC1 and 0.831 ± 0.016 W/m2 in MFC2, suggesting that microfiltration membranes were as effective as PEMs in power output. High power density could be achieved by J-colth in MFCs [12], however, Jcolth generally have large pore that it cannot prevent bacterial migration to the cathode, which would cause potential loss and CE decrease.
198
X. Tang et al. / Biochemical Engineering Journal 52 (2010) 194–198
When poultry wastewater containing 566 mg/L COD was used in MFC3, the maximum power density was 0.746 ± 0.024 W/m2 , slightly lower than that of MFC2 containing 560 mg/L COD (0.831 ± 0.016 W/m2 ). It was consistent with the previous research that MFCs using pure compounds produced higher power density than using actual wastewater [24]. The power density is known to be limited by high resistance resulting from large distances between the electrodes and correspondingly MFCs tends to have short electrodes spacing [5,7,14,16]. When Liu et al. decreased the electrode spacing from 4 to 2 cm in a single-chamber fedbatch MFC, the maximum power density increased from 720 to 1210 mW/m2 [25]. The CE was 35.3 ± 3.2% and the COD removal was 81.6 ± 6.6% in MFC3. The COD removal of MFC4 with an open circuit was 46.7 ± 4.2%. The CE of 35.3 ± 3.2% indicated considerable amount of organic matters were consumed aerobically and anaerobically. It was reported that a combination of 23.7% and 16.9% of the substrate were lost to aerobic biomass and anaerobic biomass and diffusion in two-bottle MFCs using a Nafion and ultrafiltration membrane (molecular cut-off of 3 kDa) respectively [7]. Because oxygen transfer coefficient and acetate transfer coefficient were larger in microfiltration membranes than that of Nafion and ultrafiltration membranes, a much higher percentage (46.7 ± 4.2%) of the substrate were removed in MFC4 with an open circuit compared with the results of Kim et al. [7]. More efforts are needed to tailor the thickness and pore size of the microfiltration membrane to improve CE. However, the high COD removal demonstrated microfiltration membrane was effective in actual wastewater treatment. Microfiltration membrane shows promising performances in MFCs: reducing pH gradient across the membrane, keeping a similar internal resistance and power density as PEM, and having a high percentage of COD removal for wastewater treatment. Furthermore, it considerably reduces the cost of MFC construction and therefore opens a new avenue for MFC reactor design to meet the practical implementation of wastewater treatment. Acknowledgements The research was supported by the National Science Foundation of China (No. 20876160) and the National High Technology Research and Development Program of China (863 Program) (No. 2007AA05Z158). References [1] S. Suzuki, I. Karube, T. Matsunaga, Application of a biochemical fuel cell to wastewaters, Biotechnol. Bioeng. Symp. 8 (1978) 501–511. [2] L.B. Wingard Jr., C.H. Shaw, J.F. Castner, Bioelectrochemical fuel cells, Enzyme Microb. Technol. 4 (1982) 137–142. [3] X. Tang, Z. Du, H. Li, Anodic electron shuttle mechanism based on 1-hydroxy-4aminoanthraquinone in microbial fuel cells, Electrochem. Commun. 12 (2010) 1140–1143.
[4] R.A. Rozendal, H.V.M. Hamelers, C.J.N. Buisman, Effects of membrane cation transport on pH and microbial fuel cell performance, Environ. Sci. Technol. 40 (2006) 5206–5211. [5] F. Zhao, F. Harnisch, U. Schröder, F. Scholz, P. Bogdanoff, I. Herrmann, Challenges and constraints of using oxygen cathodes in microbial fuel cells, Environ. Sci. Technol. 40 (2006) 5193–5199. [6] G.C. Gil, I. Chang, B.H. Kim, M. Kim, J. Jang, H.S. Park, H.J. Kim, Operational parameters affecting the performance of a mediator-less microbial fuel cell, Biosens. Bioelectron. 18 (2003) 327–334. [7] J. Kim, S. Cheng, S. Eunoh, B. Logan, Power generation using different cation, anion, and ultrafiltration membranes in microbial fuel cells, Environ. Sci. Technol. 38 (2004) 4040–4046. [8] S.E. Oh, B.E. Logan, Proton exchange membrane and electrode surface areas as factors that affect power generation in microbial fuel cells, Appl. Microbiol. Biotechnol. 70 (2006) 162–169. [9] Z. Liu, H. Li, Effects of bio- and abio-factors on electricity production in a mediatorless microbial fuel cell, Biochem. Eng. J. 36 (2007) 209– 214. [10] R.A. Rozendal, H.V.M. Hamelers, K. Rabaey, J. Keller, C.J.N. Buisman, Towards practical implementation of bioelectrochemical wastewater treatment, Trends Biotechnol. 26 (2008) 450–459. [11] D. Pant, G. Van Bogaert, M. De Smet, L. Diels, K. Vanbroekhoven, Use of novel permeable membrane and air cathodes in acetate microbial fuel cell, Electrochim. Acta. doi:10.1016/j.electacta.2009.11.086. (in press). [12] Y. Fan, H. Hu, H. Liu, Enhanced coulombic efficiency and power density of air–cathode microbial fuel cells with an improved cell configuration, J. Power Sources 171 (2007) 348–354. [13] K. Rabaey, G. Lissens, S.D. Siciliano, W. Verstraete, A microbial fuel cell capable of converting glucose to electricity at high rate and efficiency, Biotechnol. Lett. 25 (2003) 1531–1535. [14] H. Liu, S. Cheng, B.E. Logan, Production of electricity from acetate or butyrate in a single chamber microbial fuel cell, Environ. Sci. Technol. 39 (2005) 658– 662. [15] Z. He, N. Wagner, S.D. Minteer, L.T. Angenent, An upflow microbial fuel cell with an internal resistance by impedance spectroscopy, Environ. Sci. Technol. 40 (2006) 5212–5217. [16] B.E. Logan, C. Murano, K. Scott, N.D. Gray, I.M. Head, Electricity generation from cysteine in a microbial fuel cell, Water Res. 39 (2005) 942–952. [17] D.R. Lovley, E.J.P. Phillips, Novel mode of microbial energy metabolism: organism carbon oxidation coupled to dissimilatory reduction of iron and manganese, Appl. Environ. Microbiol. 54 (6) (1988) 1472–1480. [18] C.P.L. Grady, G.T. Daigger, H.C. Lim, Biological Wastewater Treatment, 2nd ed., Marcel Dekker, New York, 1999. [19] H. Liu, B.E. Logan, Electricity generation using an air–cathode single chamber microbial fuel cell in the presence and absence of a proton exchange membrane, Environ. Sci. Technol. 38 (2004) 4040–4046. [20] S. Cheng, H. Liu, B.E. Logan, Increased power generation in a continuous flow MFC with adjective flow through the porous anode and reduced electrode spacing, Environ. Sci. Technol. 40 (2006) 2426–2432. [21] B. Erable, L. Etcheverry, A. Bergel, Increased power from a two-chamber microbial fuel cell with a low-pH air–cathode compartment, Electrochem. Commun. 11 (2009) 619–622. [22] F. Zhang, S. Cheng, D. Pant, G. Van Bogaert, B.E. Logan, Power generation using an activated carbon and metal mesh cathode in a microbial fuel cell, Electrochem. Commun. 11 (2009) 2177–2179. [23] R.A. Rozendal, H.V. Hamelers, R.J. Molenkamp, C.J. Buisman, Performance of single chamber biocatalyzed electrolysis with different types of ion exchange membranes, Water Res. 41 (2007) 1984–1994. [24] D. Pant, G. Van Bogaert, L. Diels, K. Vanbroekhoven, A review of the substrates used in microbial fuel cells (MFCs) for sustainable energy production, Bioresour. Technol. 101 (2010) 1533–1543. [25] H. Liu, S. Cheng, B.E. Logan, Power generation in fedbatch microbial fuel cells as a function of ionic strength, temperature and reactor configuration, Environ. Sci. Technol. 39 (2005) 5488–5493.