881

Q IWA Publishing 2008 Water Science & TechnologyWST | 58.4 | 2008

A microbial fuel cell equipped with a biocathode for organic removal and denitrication
O. Lefebvre, A. Al-Mamun and H. Y. Ng

ABSTRACT
Microbial fuel cells (MFCs) are a promising anaerobic technology but they are limited by the high cost of the catalyst used at the cathode (typically platinum). In this study, we designed a novel type of two-chambered MFC wherein an autoheterotrophic denitrifying biolm replaced the costly catalyst on the cathode surface. Micro-organisms performed denitrication by using electrons supplied by bacteria oxidizing domestic wastewater and acetate as substrates in the anode chamber. This two-chambered MFC equipped with a biocathode generated during more than 1.5 month up to 9.4 mW m22 of anode surface or 0.19 W m23 of anode chamber volume, while removing over 65% of COD, 84% of total nitrogen and nearly 30% of suspended solids with domestic wastewater as a substrate, and nearly 95% of acetate in the subsequent experiments.
Key words
O. Lefebvre A. Al-Mamun H. Y. Ng Centre for Water Research, Division of Environmental Science and Engineering, National University of Singapore, 9 Engineering Dr. 1, Singapore 117576, Republic of Singapore E-mail: [email protected]; [email protected]; [email protected]

| anaerobic digestion, biocathode, denitrication, microbial fuel cell, wastewater

INTRODUCTION
A microbial fuel cell (MFC) is fundamentally an anaerobic process where bacteria grow in the absence of oxygen in a chamber containing an anode and form a biolm that covers the anode. These bacteria act as a biocatalyst at the anode and replace the use of costly platinum catalyst at the anode. To generate electricity, bacteria in the anode chamber degrade organic matter (the fuel) and transfer the electrons to the anode. The digestion of organic matters occurs in anaerobic environment so that the electrons are not consumed by electron acceptors such as oxygen. The transfer of these electrons occurs via bacterial respiratory enzymes. Protons, created at the anode to maintain a charge balance, typically migrate through the solution to a cathode where a platinum catalyst causes them to combine with oxygen and the electrons produced at the anode to form water. Hence, a typical MFC comprises an anaerobic anode chamber and an aerobic cathode chamber separated by a proton exchange membrane (PEM) that allows the protons to pass through it from the anode toward the cathode. (Figure 1a).
doi: 10.2166/wst.2008.343

The concept of MFCs for electricity generation using wastewater has already been established. However, there are still a number of issues that have to be resolved through research and development before this technology can be an attractive technology for alternative energy production. Of major concern is the optimization of MFC design in order to maximize power output and reduce the installation and operation costs (Lefebvre et al. 2008). The cost of the platinum catalyst used at the cathode is a major limitation to MFC application and economic viability. For this reason some researchers have recently started working on the concept of biocathodes that would use bacteria instead of platinum as a biocatalyst at the cathode (He & Angenent 2006). Instead of being catalyzed by platinum, the combination of protons, electrons and oxidants at the cathode would be catalyzed by a bacterial reaction. Biocathodes are of two types: (i) aerobic biocathodes use oxygen as the oxidant and microorganisms to assist the oxidation of transition metal compounds, such as Mn(II) or Fe(II), for electron delivery to oxygen; (ii) anaerobic biocathodes use

882

O. Lefebvre et al. | A microbial fuel cell equipped with a biocathode

Water Science & TechnologyWST | 58.4 | 2008

compounds such as nitrate, sulfate, iron, manganese, selenate, arsenate, urinate, fumarate and carbon dioxide as terminal electron acceptors. Out of these compounds, nitrate is of major interest due to the relatively high redox potential
0 of the couple NO2 3 =N2 (E 0.74 V). A MFC operating with

Clauwaert et al. (2007) recently published their ndings on the same technology.

METHODS
Design and operation of batch-type MFC As shown in Figure 2, the present invention consisted of two acrylic cubic chambers of 5 cm width separated with a PEM membrane made of Naon. Both chambers possessed an opening on top allowing the provision of fresh substrate, as well as the removal of the treated one. Hence, the MFC operated in batch mode. On each side of the PEM, 2 electrodes consisting of non wet-proof carbon paper (Gashub, Singapore) were coated: one anode and one cathode, both of equal size (25 cm2). The volume of each chamber was 125 cm3.

a biocathode and performing denitrication (Equation 1) in the cathode chamber would hence be of major interest (Figure 1b). In addition anaerobic biocathodes would offer the advantage of having a MFC where both the anode and cathode chambers would be anaerobic, which limits the risk of oxygen leaking in the anode chamber, thus increasing the efciency of the anodic reaction. Finally, of high interest is also the perspective of using MFCs equipped with biocathodes to achieve simultaneous treatment of carbonaceous substrates in the anode chamber and of nitrogenous compounds in the cathode chamber.
2 2NO2 3 10e 12H ! N2 6H2 O 0:74V

In the anode chamber, sewage was used as a substrate and as an inoculum simultaneously during the start-up of the MFC. The characteristics of domestic wastewater averaged 394 ^ 143 mg L21 of COD, 349 ^ 62 mg L21 of suspended solids (SS), 273 ^ 98 mg L21 of volatile suspended solids (VSS), 1,087 ^ 603 mg L21 of total dissolved solids (TDS), 38 ^ 4 mg L21 of total nitrogen (TN), 39 ^ 34 mg L21 of
2 phosphate (PO3 4 ). The pH was 7.2 ^ 0.1. Later on, the

A group of researchers have already observed biological nitrate reduction by accepting electrons from the cathode electrode in a potentiostat poised half cell (Park et al. 2005). In their study, nitrate was reduced completely to N2, according to equation (1). However, our MFC produced electricity from microbiological processes exclusively. In other words, this is a 100% microbial fuel cell. It could generate electricity at a consistent voltage during more than 1.5 months, using domestic wastewater or sodium acetate as the fuel at the anode and nitrate at the cathode. Independent of the work done by our laboratory, the researchers

anode chamber was fed with a sodium acetate (20 mM) synthetic wastewater prepared in a nutrient solution following Oh et al. (2004). The cathode chamber was seeded with a mixed culture of denitrifying bacteria issued from a lab-scale reactor and was operating with a synthetic

Figure 1

Schematic diagram of two-chambered microbial fuel cells (MFCs). a) MFC operating with a conventional cathode achieving aerobic oxidation of O2 in the cathode chamber; b) MFC operating with a biocathode achieving anaerobic denitrication in the cathode chamber; PEM, proton exchange membrane.

883

O. Lefebvre et al. | A microbial fuel cell equipped with a biocathode

Water Science & TechnologyWST | 58.4 | 2008

(Shimadzu, AOC-20i) equipped with a FIT detector and a 25 m 0.32 mm 0.5mm HP-FFAP column. Samples were ltered through a 0.2mm pore diameter membrane and acidied using formic acid before analysis. Nitrate and nitrite determination used an Ion Chromatograph (Dionex) and again samples were ltered beforehand through a 0.2mm pore diameter membrane. Gas analyses were performed using a gas chromatograph (GC-17A, Shimadzu).

RESULTS AND DISCUSSION

Lab-scale batch-type double-chambered microbial fuel cell equipped with a biocathode operating on domestic wastewater.

Operation of the batch-type MFC with domestic wastewater When the anode chamber was inoculated with domestic wastewater and the cathode chamber was fed with the synthetic nitrate solution (660 mg NO32 L21), stable power generation was obtained after 5 batches (20 d) and an illustration of typical proles of cell potentials obtained with batches of domestic wastewater is shown in Figure 3. The maximum Ecell was 0.095 ^ 0.001 V and the total reaction time averaged 118.5 ^ 0.5 h (external resistance, Rext 1,000V). The chamber was relled when Ecell decreased to less than 50 mV (Rext 1,000V). The COD removal efciency averaged 65.3 ^ 1.4%, with a corresponding ec of 14.5 ^ 0.5. SS, VSS and TN removal also averaged 28.9 ^ 0.1, 45.1 ^ 1.0 and 84.3 ^ 0.1%, respectively. Such performance was achieved under an organic loading rate (OLR) of 0.08 ^ 0.01 kg COD m23 d21. The relatively low value of ec shows that most of the organic matter was removed through other microbial processes, such as methanogenesis.

wastewater containing nitrate. This synthetic wastewater, prepared according to (Park et al. 2005), contained 6.8 g L
21

of K2HPO4, 8.7 g L21 of KH2PO4, 2.0 g L21 of NaHCO3 and 1.0 ml of trace element solution. The trace mineral solution was prepared following Balch et al. (1979), as modied by Lovley et al. (1984). This synthetic wastewater was nally amended with KNO3 to the desired concentration and fed into the cathode chamber.

Calculations Cell potential (Ecell, V) was measured with a multimeter connected to a computer by a data acquisition system (PC1604, TTi, RS, Singapore). Power (P, W), power density (Pan, W m22) and volumetric power (Pv, W m23) were calculated following Logan et al. (2006). The Coulombic efciency (Gc, %), evaluated over a period of time tb (s), was t calculated as: ec 100 M 0b Idt=FbvDCOD, where M 62 g mol21 is the molecular weight of nitrate, F 964,853 C mol21 is the Faradays constant, b 5 is the number of electrons exchanged per mole of nitrate, and ACOD (g m23) is the change in COD over time tb. When the substrate was acetate, the formula was similar except that M 82 g mol21 (molecular weight of sodium acetate), b 8 and ACOD was replaced by AS, the change of acetate concentration.

Analytics COD, SS, VSS, TDS were measured following Standard Methods. Acetate was analyzed using a gas chromatograph
Figure 3

Electricity generation by batch-mode MFCs using domestic wastewater in the anode chamber and nitrate solution in the cathode chamber.

884

O. Lefebvre et al. | A microbial fuel cell equipped with a biocathode

Water Science & TechnologyWST | 58.4 | 2008

A polarization curve was drawn by varying Rext using a resistor box (Figure 4). The open circuit voltage (OCV) of the MFC was 0.63 V. This is signicantly lower than the maximum value of 1.08 V as determined by the NADH (2 0.32 V) and nitrate ( 0.82 V) redox potentials. However, it is in the same range as conventional MFCs. For instance, Liu et al. (2005) reported an OCV of 0.8 V with a single-chambered MFC. Ecell dropped linearly with increasing current for, showing that ohmic lossescaused by the resistance to the ow of electrons through the electrodes and through the electrolytewere dominant. From the slope of the linear section of the polarization curve an internal resistance (Rint) of 5,095 V could be determined. From Figure 4 it also appears that the power output was maximal when Rint Rext 5,000 V i.e. 0.024 mW, which corresponds to a maximum power density of 9.7 mW m22 of anode surface and a maximum volumetric power of 0.19 W m23 of anode chamber volume. constant cell potential throughout the batch and a sharp drop when the fuel was exhausted. From an engineering point of view, HRT and loading rate are essential parameters that can be determined from MFC performance with sodium acetate When the substrate in the anode chamber was changed from domestic wastewater to sodium acetate synthetic solution, electricity was generated immediately, which shows the versatility of the electrophilic bacteria that were able to adapt instantly to a new substrate. The fuel cell was fed with several batches of synthetic wastewater at different acetate and nitrate concentrations and an illustration of typical proles of cell potentials obtained is shown in Figure 5. These proles were characterized by immediate electricity production following fuel addition, fairly the batch duration. Over the experimental period, these results are compiled in Table 1. It can also be seen from Table 1 that the acetate removal efciency remained at a high level in the range of 92.9 95.7%, whatever the acetate concentration tested. This was predictable since the reaction time was adjusted depending on electricity generation to allow all the acetate added to be treated. Consequently, increasing acetate concentration induced increasing reaction time, HRT and loading rate but decreasing ec. The drop in ec shows that when the acetate concentration increased, the loss of substrate due to alternative microbial processes, such as acetoclastic methanogenesis was higher. This is coherent with other studies that showed that Ecell and ec are highly dependent on the concentration of the wastewater (Min et al. 2005). In the cathode chamber, NO2 3 removal also increased with increasing initial acetate concentration. In all cases, around 2 ^ 0.2 mg L21 of NO2 2 were found in the cathode chamber at the end of the batch, which shows
2 that 27 ^ 4% of the NO2 3 removed was turned into NO2 ,
Figure 5

Electricity generation by batch-mode MFCs of sodium acetate synthetic solution in the anode chamber and nitrate solution in the cathode chamber at different concentrations.

on a molar basis according to equation (2). The remaining 73 ^ 4% of the NO2 3 removed was actually converted into N2 according to equation (1). N2 was found to constitute nearly 100% of the gas found in air-space of the cathode
Figure 4

chamber.
Polarization (-B-) and power (-W-) curves of MFC operating on domestic wastewater in the anode chamber and nitrate solution in the cathode chamber.

2 2 NO2 3 2e 2H ! NO2 H2 O 0:43V

885

O. Lefebvre et al. | A microbial fuel cell equipped with a biocathode

Water Science & TechnologyWST | 58.4 | 2008

Table 1

MFC performance using batches of sodium acetate synthetic solution in the anode chamber and nitrate solution in the cathode chamber at different concentrations NO2 3 removal (%)

Initial NO2 3 concentration (mg L21)

Initial acetate concentration (mg L21)

Acetate removal (%)

ec (%)

HRT (h)

Loading rate (kg acetate m23 d21)

596 518 520

61 87 122

2.3 3.4 4.0

92.9 94.0 95.7

24.5 21.3 18.6

54 62 73

0.027 0.034 0.040

As with domestic wastewater, a polarization curve and power density curve were drawn and showed a similar pattern as observed previously with domestic wastewater (data not shown). The OCV was 0.68 V, Rint was 3,683V, the maximum power density was 9.4 mW m surface and the maximum volumetric
22

The denitrication process has to be controlled in the cathode chamber in order to avoid the formation of NO2 2,
0 because the redox potential of NO2 3 =NO2 (E 0.43 V) is 0 inferior to that of NO2 3 =N2 (E 0.74 V).

of anode was

power

0.19 W m23 of anode chamber volume. Even though the power generated here is much lower than that commonly observed with a typical chemical MFC equipped with a platinum cathode (up to several thousand mW m
22

REFERENCES
Balch, W. E., Fox, G. E., Magrum, L. J., Woese, C. R. & Wolfe, R. S. 1979 Methanogens: reevaluation of a unique biological group. Microbiol. Rev. 43(2), 260 296. Clauwaert, P., Rabaey, K., Aelterman, P., De Schamphelaire, L., Ham, T. H., Boeckx, P., Boon, N. & Verstraete, W. 2007 Biological denitrication in microbial fuel cells. Environ. Sci. Technol. 41(9), 3354 3360. Du, Z. W., Li, H. R. & Gu, T. Y. 2007 A state of the art review on microbial fuel cells: a promising technology for wastewater treatment and bioenergy. Biotechnol. Adv. 25(5), 464 482. He, Z. & Angenent, L. T. 2006 Application of bacterial biocathodes in microbial fuel cells. Electroanalysis 18(19-20), 2009 2015. Lefebvre, O., Al-Mamun, A., Ooi, W. K., Ng, H. Y., Tang, Z. & Chua, D. H. C. 2008 An insight into cathode options for microbial fuel cells. Water Sci. Technol. 57, 2031 2037. Liu, H., Cheng, S. A. & Logan, B. E. 2005 Production of electricity from acetate or butyrate using a single-chamber microbial fuel cell. Environ. Sci. Technol. 39(2), 658662. Logan, B. E., Hamelers, B., Rozendal, R., Schrorder, U., Keller, J., Freguia, S., Aelterman, P., Verstraete, W. & Rabaey, K. 2006 Microbial fuel cells: methodology and technology. Environ. Sci. Technol. 40(17), 5181 5192. Lovley, D. R., Greening, R. C. & Ferry, J. G. 1984 Rapidly growing rumen methanogenic organism that synthesizes coenzyme-M and has a high-afnity for formate. Appl. Environ. Microbiol. 48(1), 81 87. Min, B., Kim, J. R., Oh, S. E., Regan, J. M. & Logan, B. E. 2005 Electricity generation from swine wastewater using microbial fuel cells. Water Res. 39(20), 4961 4968. Oh, S., Min, B. & Logan, B. E. 2004 Cathode performance as a factor in electricity generation in microbial fuel cells. Environ. Sci. Technol. 38(18), 4900 4904. Park, H. I., Kim, D. K., Choi, Y. J. & Pak, D. 2005 Nitrate reduction using an electrode as direct electron donor in a biolmelectrode reactor. Process Biochem. 40(10), 3383 3388.

of

anode surface area; see for instance Du et al. (2007) for a review of actual MFC performance), biocathodes are inexpensive since they do not require any metal incorporation. In addition, simple improvements in design aiming at reducing the MFC internal resistance will enhance the power production. For instance, Clauwaert et al. (2007) who conducted a similar experiment independent of the work done by our laboratory have been able to generate up to 4 W m23 of total cathodic compartment by using graphite granules, a material that allows to maximize the electrode surface area, both as the anode and as the cathode material.

CONCLUSION
From these results, MFCs operating with a biocathode and performing denitrifaction in the cathode chamber appear to be a promising technology, avoiding the use of costly platinum catalyst at the cathode. The MFC showcased in this study is far from being optimized, as can be seen from its very high internal resistance. However, it was already able to generate up to 9.4 mW m22 of anode surface and 0.19 W m23 of anode chamber volume during more than 1.5 month, while removing over 65% of COD, 84% of TN and nearly 30% of SS with domestic wastewater as a substrate and nearly 95% of acetate in the subsequent experiments.