Bioresource Technology 169 (2014) 532–536
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Highly durable anodes of microbial fuel cells using a reduced graphene oxide/carbon nanotube-coated scaffold Hung-Tao Chou a, Hui-Ju Lee b, Chi-Young Lee a, Nyan-Hwa Tai a,⇑, Hwan-You Chang b a b
Department of Materials Science and Engineering, Hsin Chu 300, Taiwan Department of Medical Science, National Tsing Hua University, Hsin Chu 300, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
A porous, conductive carbon coated
sponge is used as anode. High specific area and
biocompatibility anode promotes bacteria immobilisation. ONPG assay is used to analyse the activity of bacteria immobilised on anodes. Well transportation proliferate the bacteria, prolonging the MFC durability. Sponge anodes yield high durability and current density.
a r t i c l e
i n f o
Article history: Received 20 April 2014 Received in revised form 28 June 2014 Accepted 4 July 2014 Available online 16 July 2014 Keywords: Microbial fuel cell Graphene Macroporous anodes Bioenergy Biocatalyst
a b s t r a c t Melamine sponges coated with reduced graphene oxide/carbon nanotube (rGO–CNT sponges) through a dip-coating method were fabricated that provide a large electrical conductive surface for Escherichia coli growth and electron transfer in microbial fuel cell. Four rGO–CNT sponges with different thicknesses and arrangements were tested as an anode in this study. The thinnest one (with a thickness of 1.5 mm) exhibited the best performance, providing a maximum current density of 335 A m 3 and a remarkably durable life time of 20 days at 37 °C. Analyses of bacterial colonisation on the rGO–CNT sponges using FE-SEM and the bacterial metabolic activity using the b-galactosidase assay indicates that the rGO–CNT sponges provide excellent biocompatibility for E. coli proliferation and could help to maintain high bacterial metabolic activity, presumably due to the high mass transfer rate of the porous scaffold. In this regard, the rGO–CNT sponges showed higher durability and performed better electrochemical properties than traditional carbon-based and metal-based anodes. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Fossil fuels are essential for our daily energy needs. However, the exhaustion of fossil fuel reserves and the negative impacts by fossil fuel consumption on our environment, including air
⇑ Corresponding author. Tel.: +886 3 5742568. E-mail addresses: [email protected] (N.-H. Tai), [email protected] (H.-Y. Chang). http://dx.doi.org/10.1016/j.biortech.2014.07.027 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.
pollution and global warming, have prompted scientists to develop renewable energy technologies. One promising low-cost alternative energy technology is microbial fuel cells (MFCs), which are devices that convert bacterial metabolic chemical energy into electrical energy (Mohan et al., 2008; Rabaey and Verstraete, 2005). In MFCs, electrons are generated at the anode when bacteria oxidising carbon sources to produce protons and carbon dioxide. The electrons travel through an external circuit while the protons move by diffusion to the cathode, where they react with electron acceptors, most frequently oxygen, to complete the electricity
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
generation process. Numerous publications have demonstrated that microbial fuel cells show great potential for power generation, waste water treatment, and biosensor applications (Li and Sheng, 2012; Kalathil et al., 2012). An MFC typically comprises two compartments: an aerobic chamber containing the cathode and an anoxic chamber that houses the anode and the bacteria which act as catalysts. Because of its advantageous properties including rapid growth rate and avirulence, Escherichia coli is one of the most frequently used bacterial model in laboratory for electrochemical oxidation of carbon sources in MFCs although several other bacteria species such as Pseudomonas aeruginosa and Shewanella spp. that are more suitable for the use in wastewater have also been tested (HerreroHernandez et al., 2013). The desired properties of an anode in an MFC should include excellent electric conductivity, large surface area, and high biocompatibility for bacteria colonisation (Cheng and Logan, 2007; Qiao et al., 2007; Wang et al., 2009). Carbon materials, such as carbon cloth, carbon paper, and activated carbon meet these requirements. However, surface fouling of these carbon materials by microorganism secretions during MFC operation is common. The poor mass transfer of metabolic wastes and nutrients and decreases of surface area for bacterial colonisation can lead to lose of electrode functionality, exhibited by electric current decay and life-span reduction (Gutierrez et al., 2007; Katuri et al., 2011). To enhance the durability of MFC and reduced the cost, significant effort has been exerted on providing a large specific surface area of the anode (Gutierrez et al., 2007; Katuri et al., 2011; VazquezLarios et al., 2010). Graphene is an atomic layer of carbons arranged in a hexagonal manner. Because of its high specific surface area and high electrical conductivity, graphene has been widely used in lithium ion batteries, solar cells, and super capacitors. Additionally, the material can support catalysts and is a promising material in biofuel applications (Xiao et al., 2012; Yuan et al., 2012; Liu et al., 2012; Zhang et al., 2011). Several published articles have demonstrated that graphene is capable of promoting microorganism to secrete signalling molecules, which could accelerate the growth of bacteria and act as an electrical mediator to enhance the power transfer efficiency (Liu et al., 2012; Zhang et al., 2011; Jain et al., 2012). Studies on the applications of graphene in MFCs have been focused on the promotion of the oxygen-reducing capability of the cathode. However, few articles have addressed utilising the biocompatibility and high specific surface area features of graphene to improve anode performance (Wen et al., 2012). This work reports the fabrication of a conductive 3-dimensional structure for the purpose of improving anode performance of MFCs. The device is generated by coating reduced graphene oxide (rGO) and multi-walled carbon nanotubes (MWCNTs) on commercial melamine sponge through a simple dip-coating process. The fabricated structure exhibits high porosity and specific surface area, which are beneficial for bacteria immobilisation to provide higher current output. An enhancement in current generation stability and electrode durability are also reported herein. 2. Experimental Suppliers of raw materials and chemicals are listed in Table S1. 2.1. Nanomaterial preparation and characterisation Graphene oxide was prepared by a modified Hummer’s method and was further treated with NaBH4 at 120 °C for 7 days to form rGO (Hummers and Offeman, 1958). The resulting rGO was subjected to functionalisation to improve its dispersibility in solvent.
533
In short, 1.0 g of rGO suspended in 50 mL 10% KMnO4 (w/v) was acidified using 200 mL of 3 M H2SO4. After under sonication for 6 h, the rGO was mixed with 50 mL 10 M HCl solution and the mixture was stirring at 75 °C for 30 min. Finally, the rGO was washed to neutral pH with double distilled water. The product after each processing step was collected by filtration through a 0.22-lm cellulose filter (ADVANTEC) examined by X-ray diffraction using a Shimadzu XRD6000 instrument. The well-dispersed functionalised MWCNTs (f-MWCNTs) were obtained as described previously (Chou et al., 2013). Both the f-MWCNTs and the functionalised rGO (f-rGO) were analysed by Fourier Transform InfraRed spectroscopy (Spectrum RX, Perkin-Elmer) to examine the presence of carboxyl function groups. The lateral size of f-rGO was determined using Zeta potential measurement by the Zeta-sizer (Malvern). The purity of the synthesized f-MWCNTs and the f-rGO were estimated based on residual weights obtained from thermogravimetric analysis (TGA, Pyris Diamond TG/DTA, Perkin-Elmer).
2.2. Anode preparation and characterisation Functionalised rGO and f-MWCNTs were mixed with poly(3, 4-ethylenedioxyth-iophene) poly(styrenesulphonate) (PEDOT:PSS) ink, each at a final concentration of 0.5%. The ink was used to enhance conductivity and adherence of carbon materials on the sponge scaffold. The mixture is designated as rGO–CNT ink. Two melamine sponges, each has a thickness of 1.5 mm and 3.0 mm, respectively and a cross sectional area of 2.5 cm 2.5 cm were prepared and dipped into rGO–CNT ink for 10 min followed by a drying process. The dipping–drying process was repeated three times to increase the carbon load and to enhance the uniformity of the coating layer on the scaffold. The surface topology of the rGO–CNT-coated sponge was examined using a field-emission scanning microscope (JSM-6500F FE-SEM, JEOL, Japan). The as-prepared rGO–CNT sponge was adhered to a 6 cm 6 cm stainless steel slice (SSS) using carbon paste. The stainless steel surface was coated with polydimethylsiloxane (PDMS) to create an insulation layer and prevent microorganism from immobilisation. Graphite slice (GS), SSS, and SSS coated with rGO–CNT ink (C-SSS) were then tested under the same experimental condition using the rGO–CNT sponge as a comparison. The rGO–CNT coated on the C-SSS surface has the same composition as those coated on the sponge.
2.3. Microbial fuel cell construction and operation The MFC, as schematically depicted in Fig. S1, was assembled according to the procedure described in published articles (Rabaey and Verstraete, 2005; Wen et al., 2012). The anolyte consists of sterilized Luria-Bertani (LB) broth, phosphate buffered saline (PBS), methylene blue (an electron transfer mediator, 0.03 mM), glucose, and bacteria solution (OD = 0.3). The catholyte contained 0.1 M K3[Fe(CN)6] in PBS buffer. After the MFC was assembled, the electrolyte was pour into the chamber at an equal volume (400 mL), and the whole device was placed in an incubator set at 37 °C. The current generation of the MFC was recorded when 10 mL E. coli JM109 (OD = 0.3) were injected into the anodic chamber and the electric potential was zero. Additional glucose was added every 60 h. The fuel cell was monitored using a digital multi-metre, and the circuit was operated under a fixed external resistance of 1000 X, which was determined according to the polarisation curve.
534
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
2.4. Evaluation of bacteria colonisation on anode To examine the degree of bacterial colonisation on anode, the scaffold was detached from the SSS and immersed in 2% of glutaraldehyde for 1 h followed by sequential ethanol dehydration (20%, 40%, 50%, 70%, 90%, and 100%). The samples were air dried and coated with Pt prior to inspection using an FE-SEM. The bacteria metabolic activity of different anodes was estimated by measuring the activity of b-galactosidase, a widely used reporter for bacterial gene expression study using the o-nitrophenyl-b-D-galactoside (ONPG) assay. After incubating in the MFC for different periods of time, the sponges were detached from the SSS, washed gently in a 37 °C double distilled water bath to remove unattached bacteria, and transferred to a sterile 15 mL centrifuge tube containing physiological saline for b-galactosidase activity measurement. A few drops of toluene were added to the tube containing the sponges and incubated for 5 min to dissolve the bacterial cell membrane. Then a 0.0133 M 7 mL ONPG solution was added into the tube, and formation of yellow chromophore o-nitrophenol (ONP) from colourless ONPG catalysed by b-galactosidase was quantified spectrophotometrically at 420 nm (Griffith and Wolf, 2002). A calibration curve was established and used to estimate the number of bacteria on the sponges. 3. Results and discussion 3.1. Nano carbon material fabrication Graphene oxide was prepared from graphite flakes though the modified Hummer’s method, through the process hydroxyl and carboxylic groups were grafted on its surface. The X-ray diffraction pattern (Fig. S2a) shows a narrow diffraction peak at approximately 2h = 26.4°, which is typical of a graphite structure with a d-spacing of 0.34 nm. The broad diffraction peak of GO appears at approximately 2h = 10.9° corresponding to a d-spacing of 0.88 nm. The increase in d-spacing is due to the presence of the oxygen-containing functional groups of GO and water molecules in the interlayer. A broad and low intensity diffraction peak at approximately 2h = 24.5°, a signature of rGO, indicates that the restacking structure of rGO is different from that of graphite. This may be attributed to the remaining H atoms and OH groups bonded to the aromatic rings after NaBH4 reduction. To enhance the uniform coating of rGO–CNT on the sponge, both carbon materials were functionalised to improve their dispersibility. The FTIR spectra of the MWCNTs and rGO with and without functionalisation are shown in Fig. S2b. In contrast to the pristine MWCNTs and the rGO, f-MWCNTs and f-rGO displayed strong peaks at 1718 and 1747 cm 1, respectively indicating the presence of a C@O bond and also demonstrating the presence of carboxyl groups on their surface. The TGA plots in Fig. S2c show that only 1.4% and 1.6% impurity are in the f-rGO and f-MWCNTs preparations, respectively reflecting their high quality. Microstructure characterisations were performed using an FE-SEM and the images are shown in Fig. S3.
is commercially available, inexpensive, and with good tensile strength, as a scaffold for coating the electrical conductive carbon nanomaterials. To prevent the sponge pores from obstructing by the carbon nanomaterials during the dip-coating process, rGO of smaller lateral sizes was used in this study. It is also documented that defects and gaps between f-rGOs could disrupt electron transfer. To improve the electron transfer efficiency, f-MWCNTs, a kind of one-dimensional carbon nanomaterial, were introduced into the f-rGO preparation as fillers. Our rationale is that, the f-rGO serves as the major coating material for the electrode and f-MWCNTs bridge rGOs to provide more routes for electron transfer when fabricated in a three-dimensional connected network. To test how stabile the rGO–CNT coated on melamine sponge is, the rGO–CNT sponge fabricated through our standard procedure was immersed in E. coli culture for 20 days, and the culture medium was collected and analysed using Raman microscopy. No characteristic GO peak could be detected in the bacterial culture medium indicating that the rGO–CNT coated on the sponge was very stable and did not shed off under standard operation condition. This result implies that the anode is durable and could be used for extended periods. To form a complete MFC, carbon paste was used to fix rGO–CNT sponges to the SSS, which was connected to the electric circuit with an alligator clip. The apparatus was designed to ensure a consistent resistance is present from sample to sample. 3.3. Current outputs of the microbial fuel cells using rGO–CNT anode Current densities generated from SSS, GS, and C-SSS are shown in Fig. 1. The results show that MFC with the SSS anode produced the lowest current level presumably because the smooth surface of the SSS electrode provided the lowest surface area for bacteria immobilisation and electron transfer. Moreover, electrodes made of certain metals are not friendly for bacteria to attach and can also lower the current density of the MFC (Cheng and Logan, 2007). This notion was verified by FE-SEM and the result shows that little bacteria are adhered on the SSS even after dipping in solution for 20 days (Fig. S5a). The current density generated from the C-SSS was greater than that of the GS, which was presumably due to the higher specific surface area of the carbon material coated stainless steel (SEM image not shown). These results indicate that specific surface area is an important factor affecting current output. In this regard, sponge anodes were designed to enhance the efficiency of the MFC. Fig. 2 shows the comparison of output currents generated from sponge 1.5 (one rGO–CNT sponge with a thickness of 1.5 mm), sponge 1.5 * 2 (two rGO–CNT sponges with a thickness of 1.5 mm), sponge 3.0 (one rGO–CNT sponge with a thickness of
3.2. Carbon materials coated on the sponge and microbial fuel cell assembly The degree of porosity of the MFC electrode scaffold is closely associated with nutrient and waste transfer efficiency, and thus plays an important role in microorganism proliferation on the electrode. Furthermore, a higher porosity on the electrode can provide a higher surface area for bacteria to colonise and potentially to produce more current. To fabricate an electrode with high porosity, this study utilised melamine sponge, a highly porous polymer that
Fig. 1. Current density output of MFC using different anodes: SSS (green line), GS (blue line), and C-SSS (red line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
Fig. 2. Current density output of MFC using different anodes: sponge 1.5 (red-dash line), sponge 1.5 * 2 (red line), sponge 3.0 (blue-dash line), and sponge 3.0 array (blue line) anodes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3.0 mm), and a sponge 3.0 array (four rGO–CNT sponges with the dimensions of 1.25 cm 1.25 cm and a thickness of 3.0 mm; the distance between each sponges is nearly 0.25 cm), all the aforementioned sponges were adhered on a piece of SSS. Sponge anodes with different specific surface areas and microstructures may influence bacteria proliferation and nutrient consumption. For the anodes of sponge 1.5 * 2, sponge 3.0, and sponge 3.0 array, a glucose concentration which was fourfold greater than that used in the experiments for SSS, GS, C-SSS, and sponge 1.5 was applied to prevent unstable current output attributed to insufficient nutrient from occurring, as was reported by Herrero-Hernandez et al. (2013). Higher current densities (over 0.4 Am 2; data shown in Fig. S6) were achieved in sponge anodes compared with plane anodes (C-SSS anode), which had maximum current densities of 0.2 Am 2. In this work, the current output generated by the sponge anode was calculated based on the volume of the sponge to illustrate the advantages of a porous anode. Current density generated from sponge 3.0 was less than that of sponge 1.5, and its value decreased with time after 80 h of operation. In Fig. S5c, the micrograph of sponge 1.5 shows that bacteria firmly adhered on the skeleton of the scaffold. In Fig. S5d, the formation of biofilm can be observed on the scaffold of sponge 3.0, demonstrating that the biofilm partially blocked the channel in the porous anode of the MFC. Partial channel blockage negatively influences the transportation of waste and nutrient, and as a result, affects E. coli colonisation and its metabolic activity. In this regard, sponge 1.5 * 2 and sponge 3.0 array, with the same volumes as sponge 3.0, were used to verify how transportation obstruction affects the stability of current output from the MFC. Sponge 1.5 * 2 and the sponge 3.0 array showed similar current densities and had higher current densities and better stabilities than sponge 3.0 under 200 h of operation. However, sponge 1.5 * 2 exhibited higher current density and stability than the sponge 3.0 array after operating for 260 h. The difference may be due to the formation of biofilm, which not only partially blocked the sponge channels but also reduced electron transfer due to the intrinsic insulation properties of the biofilm. Although sponge 1.5 * 2 had the same volume as the sponge 3.0 array, it was thinner, but had twofold greater facial surface area compared with sponge 3.0. These properties of sponge 1.5 * 2 provided advantages in the transportation of waste and nutrient, which resulted in greater stability and higher current output. Fig. S7a shows that the porous structure of sponge 1.5 * 2 was maintained even after E. coli immobilisation for 20 days. However, the porous structure was partially blocked for the sponge 3.0 array, as shown in Fig. S7b. Sponge thickness is the primary parameter affecting the transportation of wastes and nutrients inside the sponge anode. In the
535
sponge 1.5 anode, biofilm formation was not detected, indicating that a sponge as thin as sponge 1.5 is suitable for bacteria proliferation, even at point of contacts with the SSS. Fig. S8 shows bacteria colonised at different depths in sponge 1.5. The transportation of nutrient and waste plays an important role in maintaining a stable current output. An insufficient nutrient supply negatively affects microorganism growth. Under these conditions, bacteria colonies trend towards dormancy to reduce nutrient consumption and their growth falls to levels inadequate to act as a catalyst for electrochemical oxidation; further discussion is provided in Section 3.4. The build-up of wastes resulting from secretions and metabolism resulted in the over-fouling of the scaffold, which reduced the living space and retarded electron transfer. 3.4. Quantity evaluation of bacteria colonised on anode Bacterial quantity and metabolic activity are the two primary factors affecting the current density of an MFC. To examine how different types of anode affect the overall bacterial metabolic activity, an ONPG assay was performed. No bacteria metabolic activity was detected on SSS, suggesting a poor bacterial colonisation on the electrode (Fig. 3). The overall bacterial metabolic activity varied with time and anodes. Starting from day 5, the bacterial metabolic activity gradually increased and reached saturation for all anodes. Higher levels of the b-galactosidase activity were observed in both sponge 3.0 and the sponge 3.0 array than on the sponge 1.5 anode because sponge 3.0 and the sponge 3.0 array possessed greater surface area for bacterial colonisation. However, the levels of bacteria b-galactosidase activity on sponge 3.0 and the sponge 3.0 array decreased after day 5. At day 7, the bacterial enzyme activity of sponge 3.0 decreased to 63% of that in day 5. The reduction in the enzyme activity may have been resulted from a decrease in bacteria propagation and conversion into a dormant state like biofilm. This phenomenon could be easily observed when GS was used as the anode. Reductions in bacteria activity at day 5 and day 13 for sponge 3.0 and the sponge 3.0 array were observed. The results could be attributed to the different transportation capabilities in different sponges prepared in this study. In contrast with the sponge 3.0 array, sponge 3.0 had lower nutrient and waste transporting capabilities. Bacterial secretions and wastes built up on its surface resulting in decreased space for bacteria colonisation and promoted the formation of biofilm, which caused the decrease in the b-galactosidase assay values. The number of glucose-oxidising bacteria on the sponge 3.0 array was maintained at a stable level because the sponge provided unblocked channels for the transportation of nutrients and wastes. However, the level of the bacterial
Fig. 3. Amount of bacteria colonies on anodes over time for GS (black), sponge 1.5 (blue), sponge 3.0 (red), and sponge 3.0 array (green) anodes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
536
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
b-galactosidase was reduced after 13 days of operation, likely due to the similar reason with that for the sponge 3.0 anode. In contrast with the sponge 3.0 and the sponge 3.0 array anodes, the thinner sponge 1.5 anode had improved exchange of nutrients and wastes, which enhanced the durability and stability of the MFC anode. After reaching saturation, variations between bacteria colonies on the sponge 1.5 anode were not as obvious as those on the sponge 3.0 and the sponge 3.0 array anodes even though the level of bacteria b-galactosidase and bacterial numbers on the sponge 1.5 anode was smaller than that of the sponge 3.0 and the sponge 3.0 array anodes. The b-galactosidase, used to analyse the overall bacterial number and activity on the anode, is related to the durability and the stability of the MFC, and correlates well with the current density-time curve. The enzyme activity also correlates well with E. coli colonisation on different anodes, as shown in Figs. 1 and 2 and S5. Among the three sponge anodes, sponge 1.5 had the highest stability and current density even though it performed at lower current values than the sponge 3.0 and the sponge 3.0 array anodes. 4. Conclusions This study demonstrates the rGO–CNT sponges, fabricated by a dip-coating method, could be used for MFC application. The sponge anodes provide high surface areas for E. coli immobilisation and high porosity for efficient mass transport supporting growth of the biocatalyst. Such an arrangement can produce a high current density of 335 A m 3 and last a long operation time of 20 days. FE-SEM images and ONPG assays were used to analyse the proliferation and the activity of bacteria, revealing that formation of biofilm, composed of less active bacteria, significantly affected the current output, as a result, reduced the stability of MFC. Acknowledgement The authors thank National Tsing-Hua University for financial support under the contract No. 101N7049E1. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2014. 07.027.
References Mohan, Y., Kumar, S.M.M., Das, D., 2008. Electricity generation using microbial fuel cells. Int. J. Hydrogen Energy 33, 423–426. Rabaey, K., Verstraete, W., 2005. Microbial fuel cells: novel biotechnology for energy generation. Trends Biotechnol. 23, 291–298. Li, W.W., Sheng, G.P., 2012. Microbial fuel cells in power generation and extended applications. Adv. Biochem. Eng. Biotechnol. 128, 165–197. Kalathil, S., Lee, J., Cho, M.H., 2012. Efficient decolorization of real dye wastewater and bioelectricity generation using a novel single chamber biocathodemicrobial fuel cell. Bioresour. Technol. 119, 22–27. Herrero-Hernandez, E., Smith, T.J., Akid, R., 2013. Electricity generation from wastewaters with starch as carbon source using a mediatorless microbial fuel cell. Biosens. Bioelectron. 39, 194–198. Cheng, S.A., Logan, B.E., 2007. Ammonia treatment of carbon cloth anodes to enhance power generation of microbial fuel cells. Electrochem. Commun. 9, 492–496. Qiao, Y., Li, C.M., Bao, S.J., Bao, Q.L., 2007. Carbon nanotube/polyaniline composite as anode material for microbial fuel cells. J. Power Sources 170, 79–84. Wang, X., Cheng, S.A., Feng, Y.J., Merrill, M.D., Saito, T., Logan, B.E., 2009. Use of carbon mesh anodes and the effect of different pretreatment methods on power production in microbial fuel cells. Environ. Sci. Technol. 43, 6870–6874. Gutierrez, M.C., Garcia-Carvajal, Z.Y., Hortiguela, M.J., Yuste, L., Rojo, F., Ferrer, M.L., del Monte, F., 2007. Biocompatible MWCNT scaffolds for immobilization and proliferation of E. coli. J. Mater. Chem. 17, 2992–2995. Katuri, K., Ferrer, M.L., Gutierrez, M.C., Jimenez, R., del Monte, F., Leech, D., 2011. Three-dimensional microchanelled electrodes in flow-through configuration for bioanode formation and current generation. Energy Environ. Sci. 4, 4201–4210. Vazquez-Larios, A.L., Solorza-Feria, O., Vazquez-Huerta, G., Esparza-Garcia, F., RiosLeal, E., Rinderknecht-Seijas, N., Poggi-Varaldo, H.M., 2010. A new design improves performance of a single chamber microbial fuel cell. J. New Mat. Electrochem. Syst. 13, 219–226. Xiao, L., Damien, J., Luo, J.Y., Jang, H.D., Huang, J.X., He, Z., 2012. Crumpled graphene particles for microbial fuel cell electrodes. J. Power Sources 208, 187–192. Yuan, Y., Zhou, S.G., Zhao, B., Zhuang, L., Wang, Y.Q., 2012. Microbially-reduced graphene scaffolds to facilitate extracellular electron transfer in microbial fuel cells. Bioresour. Technol. 116, 453–458. Liu, J., Qiao, Y., Guo, C.X., Lim, S., Song, H., Li, C.M., 2012. Graphene/carbon cloth anode for high-performance mediatorless microbial fuel cells. Bioresour. Technol. 114, 275–280. Zhang, Y.Z., Mo, G.Q., Li, X.W., Zhang, W.D., Zhang, J.Q., Ye, J.S., Huang, X.D., Yu, C.Z., 2011. A graphene modified anode to improve the performance of microbial fuel cells. J. Power Sources 196, 5402–5407. Jain, A., Zhang, X.M., Pastorella, G., Connolly, J.O., Barry, N., Woolley, R., Krishnamurthy, S., Marsili, E., 2012. Electron transfer mechanism in Shewanella loihica PV-4 biofilms formed at graphite electrode. Bioelectrochemistry 87, 28–32. Wen, Q., Wang, S.Y., Yan, J., Cong, L.J., Pan, Z.C., Ren, Y.M., Fan, Z.J., 2012. MnO2graphene hybrid as an alternative cathodic catalyst to platinum in microbial fuel cells. J. Power Sources 216, 187–191. Hummers, W.S., Offeman, R.E., 1958. Preparation of graphitic oxide. J. Am. Chem. Soc. 80, 1339. Chou, H.T., Wang, T.P., Lee, C.Y., Tai, N.H., Chang, H.Y., 2013. Photothermal effects of multi-walled carbon nanotubes on the viability of BT-474 cancer cells. Mater. Sci. Eng. C-Mater. Biol. Appl. 33, 989–995. Griffith, K.L., Wolf, R.E., 2002. Measuring beta-galactosidase activity in bacteria: cell growth, permeabilization, and enzyme assays in 96-well arrays. Biochem. Biophys. Res. Commun. 290, 397–402.
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Highly durable anodes of microbial fuel cells using a reduced graphene oxide/carbon nanotube-coated scaffold Hung-Tao Chou a, Hui-Ju Lee b, Chi-Young Lee a, Nyan-Hwa Tai a,⇑, Hwan-You Chang b a b
Department of Materials Science and Engineering, Hsin Chu 300, Taiwan Department of Medical Science, National Tsing Hua University, Hsin Chu 300, Taiwan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
A porous, conductive carbon coated
sponge is used as anode. High specific area and
biocompatibility anode promotes bacteria immobilisation. ONPG assay is used to analyse the activity of bacteria immobilised on anodes. Well transportation proliferate the bacteria, prolonging the MFC durability. Sponge anodes yield high durability and current density.
a r t i c l e
i n f o
Article history: Received 20 April 2014 Received in revised form 28 June 2014 Accepted 4 July 2014 Available online 16 July 2014 Keywords: Microbial fuel cell Graphene Macroporous anodes Bioenergy Biocatalyst
a b s t r a c t Melamine sponges coated with reduced graphene oxide/carbon nanotube (rGO–CNT sponges) through a dip-coating method were fabricated that provide a large electrical conductive surface for Escherichia coli growth and electron transfer in microbial fuel cell. Four rGO–CNT sponges with different thicknesses and arrangements were tested as an anode in this study. The thinnest one (with a thickness of 1.5 mm) exhibited the best performance, providing a maximum current density of 335 A m 3 and a remarkably durable life time of 20 days at 37 °C. Analyses of bacterial colonisation on the rGO–CNT sponges using FE-SEM and the bacterial metabolic activity using the b-galactosidase assay indicates that the rGO–CNT sponges provide excellent biocompatibility for E. coli proliferation and could help to maintain high bacterial metabolic activity, presumably due to the high mass transfer rate of the porous scaffold. In this regard, the rGO–CNT sponges showed higher durability and performed better electrochemical properties than traditional carbon-based and metal-based anodes. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction Fossil fuels are essential for our daily energy needs. However, the exhaustion of fossil fuel reserves and the negative impacts by fossil fuel consumption on our environment, including air
⇑ Corresponding author. Tel.: +886 3 5742568. E-mail addresses: [email protected] (N.-H. Tai), [email protected] (H.-Y. Chang). http://dx.doi.org/10.1016/j.biortech.2014.07.027 0960-8524/Ó 2014 Elsevier Ltd. All rights reserved.
pollution and global warming, have prompted scientists to develop renewable energy technologies. One promising low-cost alternative energy technology is microbial fuel cells (MFCs), which are devices that convert bacterial metabolic chemical energy into electrical energy (Mohan et al., 2008; Rabaey and Verstraete, 2005). In MFCs, electrons are generated at the anode when bacteria oxidising carbon sources to produce protons and carbon dioxide. The electrons travel through an external circuit while the protons move by diffusion to the cathode, where they react with electron acceptors, most frequently oxygen, to complete the electricity
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
generation process. Numerous publications have demonstrated that microbial fuel cells show great potential for power generation, waste water treatment, and biosensor applications (Li and Sheng, 2012; Kalathil et al., 2012). An MFC typically comprises two compartments: an aerobic chamber containing the cathode and an anoxic chamber that houses the anode and the bacteria which act as catalysts. Because of its advantageous properties including rapid growth rate and avirulence, Escherichia coli is one of the most frequently used bacterial model in laboratory for electrochemical oxidation of carbon sources in MFCs although several other bacteria species such as Pseudomonas aeruginosa and Shewanella spp. that are more suitable for the use in wastewater have also been tested (HerreroHernandez et al., 2013). The desired properties of an anode in an MFC should include excellent electric conductivity, large surface area, and high biocompatibility for bacteria colonisation (Cheng and Logan, 2007; Qiao et al., 2007; Wang et al., 2009). Carbon materials, such as carbon cloth, carbon paper, and activated carbon meet these requirements. However, surface fouling of these carbon materials by microorganism secretions during MFC operation is common. The poor mass transfer of metabolic wastes and nutrients and decreases of surface area for bacterial colonisation can lead to lose of electrode functionality, exhibited by electric current decay and life-span reduction (Gutierrez et al., 2007; Katuri et al., 2011). To enhance the durability of MFC and reduced the cost, significant effort has been exerted on providing a large specific surface area of the anode (Gutierrez et al., 2007; Katuri et al., 2011; VazquezLarios et al., 2010). Graphene is an atomic layer of carbons arranged in a hexagonal manner. Because of its high specific surface area and high electrical conductivity, graphene has been widely used in lithium ion batteries, solar cells, and super capacitors. Additionally, the material can support catalysts and is a promising material in biofuel applications (Xiao et al., 2012; Yuan et al., 2012; Liu et al., 2012; Zhang et al., 2011). Several published articles have demonstrated that graphene is capable of promoting microorganism to secrete signalling molecules, which could accelerate the growth of bacteria and act as an electrical mediator to enhance the power transfer efficiency (Liu et al., 2012; Zhang et al., 2011; Jain et al., 2012). Studies on the applications of graphene in MFCs have been focused on the promotion of the oxygen-reducing capability of the cathode. However, few articles have addressed utilising the biocompatibility and high specific surface area features of graphene to improve anode performance (Wen et al., 2012). This work reports the fabrication of a conductive 3-dimensional structure for the purpose of improving anode performance of MFCs. The device is generated by coating reduced graphene oxide (rGO) and multi-walled carbon nanotubes (MWCNTs) on commercial melamine sponge through a simple dip-coating process. The fabricated structure exhibits high porosity and specific surface area, which are beneficial for bacteria immobilisation to provide higher current output. An enhancement in current generation stability and electrode durability are also reported herein. 2. Experimental Suppliers of raw materials and chemicals are listed in Table S1. 2.1. Nanomaterial preparation and characterisation Graphene oxide was prepared by a modified Hummer’s method and was further treated with NaBH4 at 120 °C for 7 days to form rGO (Hummers and Offeman, 1958). The resulting rGO was subjected to functionalisation to improve its dispersibility in solvent.
533
In short, 1.0 g of rGO suspended in 50 mL 10% KMnO4 (w/v) was acidified using 200 mL of 3 M H2SO4. After under sonication for 6 h, the rGO was mixed with 50 mL 10 M HCl solution and the mixture was stirring at 75 °C for 30 min. Finally, the rGO was washed to neutral pH with double distilled water. The product after each processing step was collected by filtration through a 0.22-lm cellulose filter (ADVANTEC) examined by X-ray diffraction using a Shimadzu XRD6000 instrument. The well-dispersed functionalised MWCNTs (f-MWCNTs) were obtained as described previously (Chou et al., 2013). Both the f-MWCNTs and the functionalised rGO (f-rGO) were analysed by Fourier Transform InfraRed spectroscopy (Spectrum RX, Perkin-Elmer) to examine the presence of carboxyl function groups. The lateral size of f-rGO was determined using Zeta potential measurement by the Zeta-sizer (Malvern). The purity of the synthesized f-MWCNTs and the f-rGO were estimated based on residual weights obtained from thermogravimetric analysis (TGA, Pyris Diamond TG/DTA, Perkin-Elmer).
2.2. Anode preparation and characterisation Functionalised rGO and f-MWCNTs were mixed with poly(3, 4-ethylenedioxyth-iophene) poly(styrenesulphonate) (PEDOT:PSS) ink, each at a final concentration of 0.5%. The ink was used to enhance conductivity and adherence of carbon materials on the sponge scaffold. The mixture is designated as rGO–CNT ink. Two melamine sponges, each has a thickness of 1.5 mm and 3.0 mm, respectively and a cross sectional area of 2.5 cm 2.5 cm were prepared and dipped into rGO–CNT ink for 10 min followed by a drying process. The dipping–drying process was repeated three times to increase the carbon load and to enhance the uniformity of the coating layer on the scaffold. The surface topology of the rGO–CNT-coated sponge was examined using a field-emission scanning microscope (JSM-6500F FE-SEM, JEOL, Japan). The as-prepared rGO–CNT sponge was adhered to a 6 cm 6 cm stainless steel slice (SSS) using carbon paste. The stainless steel surface was coated with polydimethylsiloxane (PDMS) to create an insulation layer and prevent microorganism from immobilisation. Graphite slice (GS), SSS, and SSS coated with rGO–CNT ink (C-SSS) were then tested under the same experimental condition using the rGO–CNT sponge as a comparison. The rGO–CNT coated on the C-SSS surface has the same composition as those coated on the sponge.
2.3. Microbial fuel cell construction and operation The MFC, as schematically depicted in Fig. S1, was assembled according to the procedure described in published articles (Rabaey and Verstraete, 2005; Wen et al., 2012). The anolyte consists of sterilized Luria-Bertani (LB) broth, phosphate buffered saline (PBS), methylene blue (an electron transfer mediator, 0.03 mM), glucose, and bacteria solution (OD = 0.3). The catholyte contained 0.1 M K3[Fe(CN)6] in PBS buffer. After the MFC was assembled, the electrolyte was pour into the chamber at an equal volume (400 mL), and the whole device was placed in an incubator set at 37 °C. The current generation of the MFC was recorded when 10 mL E. coli JM109 (OD = 0.3) were injected into the anodic chamber and the electric potential was zero. Additional glucose was added every 60 h. The fuel cell was monitored using a digital multi-metre, and the circuit was operated under a fixed external resistance of 1000 X, which was determined according to the polarisation curve.
534
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
2.4. Evaluation of bacteria colonisation on anode To examine the degree of bacterial colonisation on anode, the scaffold was detached from the SSS and immersed in 2% of glutaraldehyde for 1 h followed by sequential ethanol dehydration (20%, 40%, 50%, 70%, 90%, and 100%). The samples were air dried and coated with Pt prior to inspection using an FE-SEM. The bacteria metabolic activity of different anodes was estimated by measuring the activity of b-galactosidase, a widely used reporter for bacterial gene expression study using the o-nitrophenyl-b-D-galactoside (ONPG) assay. After incubating in the MFC for different periods of time, the sponges were detached from the SSS, washed gently in a 37 °C double distilled water bath to remove unattached bacteria, and transferred to a sterile 15 mL centrifuge tube containing physiological saline for b-galactosidase activity measurement. A few drops of toluene were added to the tube containing the sponges and incubated for 5 min to dissolve the bacterial cell membrane. Then a 0.0133 M 7 mL ONPG solution was added into the tube, and formation of yellow chromophore o-nitrophenol (ONP) from colourless ONPG catalysed by b-galactosidase was quantified spectrophotometrically at 420 nm (Griffith and Wolf, 2002). A calibration curve was established and used to estimate the number of bacteria on the sponges. 3. Results and discussion 3.1. Nano carbon material fabrication Graphene oxide was prepared from graphite flakes though the modified Hummer’s method, through the process hydroxyl and carboxylic groups were grafted on its surface. The X-ray diffraction pattern (Fig. S2a) shows a narrow diffraction peak at approximately 2h = 26.4°, which is typical of a graphite structure with a d-spacing of 0.34 nm. The broad diffraction peak of GO appears at approximately 2h = 10.9° corresponding to a d-spacing of 0.88 nm. The increase in d-spacing is due to the presence of the oxygen-containing functional groups of GO and water molecules in the interlayer. A broad and low intensity diffraction peak at approximately 2h = 24.5°, a signature of rGO, indicates that the restacking structure of rGO is different from that of graphite. This may be attributed to the remaining H atoms and OH groups bonded to the aromatic rings after NaBH4 reduction. To enhance the uniform coating of rGO–CNT on the sponge, both carbon materials were functionalised to improve their dispersibility. The FTIR spectra of the MWCNTs and rGO with and without functionalisation are shown in Fig. S2b. In contrast to the pristine MWCNTs and the rGO, f-MWCNTs and f-rGO displayed strong peaks at 1718 and 1747 cm 1, respectively indicating the presence of a C@O bond and also demonstrating the presence of carboxyl groups on their surface. The TGA plots in Fig. S2c show that only 1.4% and 1.6% impurity are in the f-rGO and f-MWCNTs preparations, respectively reflecting their high quality. Microstructure characterisations were performed using an FE-SEM and the images are shown in Fig. S3.
is commercially available, inexpensive, and with good tensile strength, as a scaffold for coating the electrical conductive carbon nanomaterials. To prevent the sponge pores from obstructing by the carbon nanomaterials during the dip-coating process, rGO of smaller lateral sizes was used in this study. It is also documented that defects and gaps between f-rGOs could disrupt electron transfer. To improve the electron transfer efficiency, f-MWCNTs, a kind of one-dimensional carbon nanomaterial, were introduced into the f-rGO preparation as fillers. Our rationale is that, the f-rGO serves as the major coating material for the electrode and f-MWCNTs bridge rGOs to provide more routes for electron transfer when fabricated in a three-dimensional connected network. To test how stabile the rGO–CNT coated on melamine sponge is, the rGO–CNT sponge fabricated through our standard procedure was immersed in E. coli culture for 20 days, and the culture medium was collected and analysed using Raman microscopy. No characteristic GO peak could be detected in the bacterial culture medium indicating that the rGO–CNT coated on the sponge was very stable and did not shed off under standard operation condition. This result implies that the anode is durable and could be used for extended periods. To form a complete MFC, carbon paste was used to fix rGO–CNT sponges to the SSS, which was connected to the electric circuit with an alligator clip. The apparatus was designed to ensure a consistent resistance is present from sample to sample. 3.3. Current outputs of the microbial fuel cells using rGO–CNT anode Current densities generated from SSS, GS, and C-SSS are shown in Fig. 1. The results show that MFC with the SSS anode produced the lowest current level presumably because the smooth surface of the SSS electrode provided the lowest surface area for bacteria immobilisation and electron transfer. Moreover, electrodes made of certain metals are not friendly for bacteria to attach and can also lower the current density of the MFC (Cheng and Logan, 2007). This notion was verified by FE-SEM and the result shows that little bacteria are adhered on the SSS even after dipping in solution for 20 days (Fig. S5a). The current density generated from the C-SSS was greater than that of the GS, which was presumably due to the higher specific surface area of the carbon material coated stainless steel (SEM image not shown). These results indicate that specific surface area is an important factor affecting current output. In this regard, sponge anodes were designed to enhance the efficiency of the MFC. Fig. 2 shows the comparison of output currents generated from sponge 1.5 (one rGO–CNT sponge with a thickness of 1.5 mm), sponge 1.5 * 2 (two rGO–CNT sponges with a thickness of 1.5 mm), sponge 3.0 (one rGO–CNT sponge with a thickness of
3.2. Carbon materials coated on the sponge and microbial fuel cell assembly The degree of porosity of the MFC electrode scaffold is closely associated with nutrient and waste transfer efficiency, and thus plays an important role in microorganism proliferation on the electrode. Furthermore, a higher porosity on the electrode can provide a higher surface area for bacteria to colonise and potentially to produce more current. To fabricate an electrode with high porosity, this study utilised melamine sponge, a highly porous polymer that
Fig. 1. Current density output of MFC using different anodes: SSS (green line), GS (blue line), and C-SSS (red line). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
Fig. 2. Current density output of MFC using different anodes: sponge 1.5 (red-dash line), sponge 1.5 * 2 (red line), sponge 3.0 (blue-dash line), and sponge 3.0 array (blue line) anodes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
3.0 mm), and a sponge 3.0 array (four rGO–CNT sponges with the dimensions of 1.25 cm 1.25 cm and a thickness of 3.0 mm; the distance between each sponges is nearly 0.25 cm), all the aforementioned sponges were adhered on a piece of SSS. Sponge anodes with different specific surface areas and microstructures may influence bacteria proliferation and nutrient consumption. For the anodes of sponge 1.5 * 2, sponge 3.0, and sponge 3.0 array, a glucose concentration which was fourfold greater than that used in the experiments for SSS, GS, C-SSS, and sponge 1.5 was applied to prevent unstable current output attributed to insufficient nutrient from occurring, as was reported by Herrero-Hernandez et al. (2013). Higher current densities (over 0.4 Am 2; data shown in Fig. S6) were achieved in sponge anodes compared with plane anodes (C-SSS anode), which had maximum current densities of 0.2 Am 2. In this work, the current output generated by the sponge anode was calculated based on the volume of the sponge to illustrate the advantages of a porous anode. Current density generated from sponge 3.0 was less than that of sponge 1.5, and its value decreased with time after 80 h of operation. In Fig. S5c, the micrograph of sponge 1.5 shows that bacteria firmly adhered on the skeleton of the scaffold. In Fig. S5d, the formation of biofilm can be observed on the scaffold of sponge 3.0, demonstrating that the biofilm partially blocked the channel in the porous anode of the MFC. Partial channel blockage negatively influences the transportation of waste and nutrient, and as a result, affects E. coli colonisation and its metabolic activity. In this regard, sponge 1.5 * 2 and sponge 3.0 array, with the same volumes as sponge 3.0, were used to verify how transportation obstruction affects the stability of current output from the MFC. Sponge 1.5 * 2 and the sponge 3.0 array showed similar current densities and had higher current densities and better stabilities than sponge 3.0 under 200 h of operation. However, sponge 1.5 * 2 exhibited higher current density and stability than the sponge 3.0 array after operating for 260 h. The difference may be due to the formation of biofilm, which not only partially blocked the sponge channels but also reduced electron transfer due to the intrinsic insulation properties of the biofilm. Although sponge 1.5 * 2 had the same volume as the sponge 3.0 array, it was thinner, but had twofold greater facial surface area compared with sponge 3.0. These properties of sponge 1.5 * 2 provided advantages in the transportation of waste and nutrient, which resulted in greater stability and higher current output. Fig. S7a shows that the porous structure of sponge 1.5 * 2 was maintained even after E. coli immobilisation for 20 days. However, the porous structure was partially blocked for the sponge 3.0 array, as shown in Fig. S7b. Sponge thickness is the primary parameter affecting the transportation of wastes and nutrients inside the sponge anode. In the
535
sponge 1.5 anode, biofilm formation was not detected, indicating that a sponge as thin as sponge 1.5 is suitable for bacteria proliferation, even at point of contacts with the SSS. Fig. S8 shows bacteria colonised at different depths in sponge 1.5. The transportation of nutrient and waste plays an important role in maintaining a stable current output. An insufficient nutrient supply negatively affects microorganism growth. Under these conditions, bacteria colonies trend towards dormancy to reduce nutrient consumption and their growth falls to levels inadequate to act as a catalyst for electrochemical oxidation; further discussion is provided in Section 3.4. The build-up of wastes resulting from secretions and metabolism resulted in the over-fouling of the scaffold, which reduced the living space and retarded electron transfer. 3.4. Quantity evaluation of bacteria colonised on anode Bacterial quantity and metabolic activity are the two primary factors affecting the current density of an MFC. To examine how different types of anode affect the overall bacterial metabolic activity, an ONPG assay was performed. No bacteria metabolic activity was detected on SSS, suggesting a poor bacterial colonisation on the electrode (Fig. 3). The overall bacterial metabolic activity varied with time and anodes. Starting from day 5, the bacterial metabolic activity gradually increased and reached saturation for all anodes. Higher levels of the b-galactosidase activity were observed in both sponge 3.0 and the sponge 3.0 array than on the sponge 1.5 anode because sponge 3.0 and the sponge 3.0 array possessed greater surface area for bacterial colonisation. However, the levels of bacteria b-galactosidase activity on sponge 3.0 and the sponge 3.0 array decreased after day 5. At day 7, the bacterial enzyme activity of sponge 3.0 decreased to 63% of that in day 5. The reduction in the enzyme activity may have been resulted from a decrease in bacteria propagation and conversion into a dormant state like biofilm. This phenomenon could be easily observed when GS was used as the anode. Reductions in bacteria activity at day 5 and day 13 for sponge 3.0 and the sponge 3.0 array were observed. The results could be attributed to the different transportation capabilities in different sponges prepared in this study. In contrast with the sponge 3.0 array, sponge 3.0 had lower nutrient and waste transporting capabilities. Bacterial secretions and wastes built up on its surface resulting in decreased space for bacteria colonisation and promoted the formation of biofilm, which caused the decrease in the b-galactosidase assay values. The number of glucose-oxidising bacteria on the sponge 3.0 array was maintained at a stable level because the sponge provided unblocked channels for the transportation of nutrients and wastes. However, the level of the bacterial
Fig. 3. Amount of bacteria colonies on anodes over time for GS (black), sponge 1.5 (blue), sponge 3.0 (red), and sponge 3.0 array (green) anodes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
536
H.-T. Chou et al. / Bioresource Technology 169 (2014) 532–536
b-galactosidase was reduced after 13 days of operation, likely due to the similar reason with that for the sponge 3.0 anode. In contrast with the sponge 3.0 and the sponge 3.0 array anodes, the thinner sponge 1.5 anode had improved exchange of nutrients and wastes, which enhanced the durability and stability of the MFC anode. After reaching saturation, variations between bacteria colonies on the sponge 1.5 anode were not as obvious as those on the sponge 3.0 and the sponge 3.0 array anodes even though the level of bacteria b-galactosidase and bacterial numbers on the sponge 1.5 anode was smaller than that of the sponge 3.0 and the sponge 3.0 array anodes. The b-galactosidase, used to analyse the overall bacterial number and activity on the anode, is related to the durability and the stability of the MFC, and correlates well with the current density-time curve. The enzyme activity also correlates well with E. coli colonisation on different anodes, as shown in Figs. 1 and 2 and S5. Among the three sponge anodes, sponge 1.5 had the highest stability and current density even though it performed at lower current values than the sponge 3.0 and the sponge 3.0 array anodes. 4. Conclusions This study demonstrates the rGO–CNT sponges, fabricated by a dip-coating method, could be used for MFC application. The sponge anodes provide high surface areas for E. coli immobilisation and high porosity for efficient mass transport supporting growth of the biocatalyst. Such an arrangement can produce a high current density of 335 A m 3 and last a long operation time of 20 days. FE-SEM images and ONPG assays were used to analyse the proliferation and the activity of bacteria, revealing that formation of biofilm, composed of less active bacteria, significantly affected the current output, as a result, reduced the stability of MFC. Acknowledgement The authors thank National Tsing-Hua University for financial support under the contract No. 101N7049E1. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2014. 07.027.
References Mohan, Y., Kumar, S.M.M., Das, D., 2008. Electricity generation using microbial fuel cells. Int. J. Hydrogen Energy 33, 423–426. Rabaey, K., Verstraete, W., 2005. Microbial fuel cells: novel biotechnology for energy generation. Trends Biotechnol. 23, 291–298. Li, W.W., Sheng, G.P., 2012. Microbial fuel cells in power generation and extended applications. Adv. Biochem. Eng. Biotechnol. 128, 165–197. Kalathil, S., Lee, J., Cho, M.H., 2012. Efficient decolorization of real dye wastewater and bioelectricity generation using a novel single chamber biocathodemicrobial fuel cell. Bioresour. Technol. 119, 22–27. Herrero-Hernandez, E., Smith, T.J., Akid, R., 2013. Electricity generation from wastewaters with starch as carbon source using a mediatorless microbial fuel cell. Biosens. Bioelectron. 39, 194–198. Cheng, S.A., Logan, B.E., 2007. Ammonia treatment of carbon cloth anodes to enhance power generation of microbial fuel cells. Electrochem. Commun. 9, 492–496. Qiao, Y., Li, C.M., Bao, S.J., Bao, Q.L., 2007. Carbon nanotube/polyaniline composite as anode material for microbial fuel cells. J. Power Sources 170, 79–84. Wang, X., Cheng, S.A., Feng, Y.J., Merrill, M.D., Saito, T., Logan, B.E., 2009. Use of carbon mesh anodes and the effect of different pretreatment methods on power production in microbial fuel cells. Environ. Sci. Technol. 43, 6870–6874. Gutierrez, M.C., Garcia-Carvajal, Z.Y., Hortiguela, M.J., Yuste, L., Rojo, F., Ferrer, M.L., del Monte, F., 2007. Biocompatible MWCNT scaffolds for immobilization and proliferation of E. coli. J. Mater. Chem. 17, 2992–2995. Katuri, K., Ferrer, M.L., Gutierrez, M.C., Jimenez, R., del Monte, F., Leech, D., 2011. Three-dimensional microchanelled electrodes in flow-through configuration for bioanode formation and current generation. Energy Environ. Sci. 4, 4201–4210. Vazquez-Larios, A.L., Solorza-Feria, O., Vazquez-Huerta, G., Esparza-Garcia, F., RiosLeal, E., Rinderknecht-Seijas, N., Poggi-Varaldo, H.M., 2010. A new design improves performance of a single chamber microbial fuel cell. J. New Mat. Electrochem. Syst. 13, 219–226. Xiao, L., Damien, J., Luo, J.Y., Jang, H.D., Huang, J.X., He, Z., 2012. Crumpled graphene particles for microbial fuel cell electrodes. J. Power Sources 208, 187–192. Yuan, Y., Zhou, S.G., Zhao, B., Zhuang, L., Wang, Y.Q., 2012. Microbially-reduced graphene scaffolds to facilitate extracellular electron transfer in microbial fuel cells. Bioresour. Technol. 116, 453–458. Liu, J., Qiao, Y., Guo, C.X., Lim, S., Song, H., Li, C.M., 2012. Graphene/carbon cloth anode for high-performance mediatorless microbial fuel cells. Bioresour. Technol. 114, 275–280. Zhang, Y.Z., Mo, G.Q., Li, X.W., Zhang, W.D., Zhang, J.Q., Ye, J.S., Huang, X.D., Yu, C.Z., 2011. A graphene modified anode to improve the performance of microbial fuel cells. J. Power Sources 196, 5402–5407. Jain, A., Zhang, X.M., Pastorella, G., Connolly, J.O., Barry, N., Woolley, R., Krishnamurthy, S., Marsili, E., 2012. Electron transfer mechanism in Shewanella loihica PV-4 biofilms formed at graphite electrode. Bioelectrochemistry 87, 28–32. Wen, Q., Wang, S.Y., Yan, J., Cong, L.J., Pan, Z.C., Ren, Y.M., Fan, Z.J., 2012. MnO2graphene hybrid as an alternative cathodic catalyst to platinum in microbial fuel cells. J. Power Sources 216, 187–191. Hummers, W.S., Offeman, R.E., 1958. Preparation of graphitic oxide. J. Am. Chem. Soc. 80, 1339. Chou, H.T., Wang, T.P., Lee, C.Y., Tai, N.H., Chang, H.Y., 2013. Photothermal effects of multi-walled carbon nanotubes on the viability of BT-474 cancer cells. Mater. Sci. Eng. C-Mater. Biol. Appl. 33, 989–995. Griffith, K.L., Wolf, R.E., 2002. Measuring beta-galactosidase activity in bacteria: cell growth, permeabilization, and enzyme assays in 96-well arrays. Biochem. Biophys. Res. Commun. 290, 397–402.
