Three-dimensional, highly porous N-doped carbon foam as microorganism propitious, efficient anode for high performance microbial fuel cell

Abstract

Three-dimensional (3D) N-doped open-porous carbon foam was fabricated using the simple procedure of calcining a melamine sponge. The properties of the fabricated carbon foam and its performance in microbial fuel cells (MFCs) using Shewanella oneidensis MR-1 (S. oneidensis) were compared with those of commercial graphite felt. The MFC with the carbon foam anode produced approximately 2 times higher power density than the commercial graphite felt. The superior performance of the as-prepared carbon foam in MFC was attributed to the higher surface area (687.19 m² g⁻¹) and open-porous scaffold structure. Moreover, the appearance of the hydrophilic functional groups such as C=N–C, N–C=O on the surface of the as-prepared carbon foam facilitated extracellular electron transfer, resulting in a decrease in charge transfer resistance and an increase in biocompatibility. Owing to the excellent biocompatibility, a large amount of microbial biomass colonized both the surface and inside the carbon foam, which helped enhance the performance of the MFC.

---

Introduction

The generation of electricity along with wastewater treatment using microbial fuel cells (MFCs) has attracted considerable attention over the last few decades. On the other hand, the widespread practical application of MFCs is still limited by the low power generation and high cost of the anode materials, separator and catalyst used in the cathode.¹ MFC power generation depends on many factors, such as biocatalyst, reactor configurations, material electrodes, and separator.² Among these factors, the anode, as an electron acceptor, plays the most important role in power generation³ and in the cost^1,4 of MFCs. To achieve good performance in MFCs, the anode requires a high surface area with controlled porosity, high electrical conductivity, good biocompatibility, good stability and durability, and cost-effectiveness.⁵

The MFC anode materials are quite diverse, but are generally divided into carbonaceous and non-carbonaceous materials. Carbonaceous materials are used widely because of their high roughness, enhanced conductivity, eco-friendliness, and cost-effectiveness.⁵ A large number of porous and non-porous carbon-based materials have been used, including graphite rods, carbon paper, carbon cloth, graphite felt, and reticulated vitrified carbon.⁶ Compared to a conventional flat anode (graphite rod, carbon paper), anodes with a 3D open-porous structure have been reported to enhance the MFC performance significantly because of the large bio-accessible area, which allows internal bacteria colonization and the efficient substrate transfer simultaneously.⁷ Although 3D materials show promise, their applications are limited by their inherent drawbacks, such as low biocompatibility (polyaniline-hybridized graphene foam⁸) and high cost (carbon nanotube-coated sponges⁹).

In order to improve the biocompatibility of anode, the anode surface treatment procedures, such as ammonia,¹⁰ heat,¹¹ acid treatment¹² and electrochemical oxidation¹³ were exploited. The basic of these treatment procedures is to increases the adhesion of microorganism onto the anode surface by enhancing the positive charge of the electrode surface or adding hydrophilic functional groups.⁵ In addition, the hydrophilicity and biocompatibility of anode can be improved by 3D open-celled carbon scaffold with the hydrophilic functional –C=N group prepared by foaming polyacrylonitrile using supercritical CO₂.¹⁴

Carbon foam with 3D open-celled structure has very high electrical and thermal conductivity, as well as a tunable cell size, making it an interesting candidate for a range of applications, such as electrode storage devices and electrochemical capacitors.¹⁵ Carbon foam has been commercialized under the trade name of reticulated vitreous carbon, and has been applied to MFCs.^16,17 On the other hand, it is expensive and has poor electrical conductivity,¹⁵ resulting in inferior performance compared to MFCs made using carbon foam prepared from sponge-like structures.⁶ Therefore, it is important to develop a low-cost and effective 3D carbon foam that helps increase the MFC performance and reduce the capital cost.

Recently, the direct carbonization of melamine sponge, which is available low cost and environmental friendly, a novel nitrogen-doped carbon foam highly electrically conductive 3D elastic interconnected network could be prepared for supercapacitor application.^18,19 Attributed to these properties, that carbon foam could be a promising candidate to 3D conductive framework for MFCs applications. This paper reports a simple method to synthesize a 3D N-doped open-porous carbon foam electrode using melamine sponge template and resorcinol–formaldehyde polymer. The structure, physical and chemical properties as well as the performance of the as-prepared carbon foam were characterized using a range of analytical and physico-chemical techniques. In addition, the performance of the MFCs with the commercial graphite felt and the as-prepared carbon foam electrode was also evaluated for comparison.

Experimental

Preparation of 3D N-doped carbon foam

Carbon foam was prepared using a melamine sponge (Dae Han Co. Ltd, Korea) as the template (Fig. 1). The melamine sponge was first impregnated with solution containing 50 mL H₂O, 1 g of resorcinol, 1.47 g of formaldehyde, and 2.41 g of sodium carbonate solution (2 wt%) followed by aging at 80 °C for 72 h under closed conditions. The obtained dark red color gel was cut into 10 cm × 3 cm × 1.5 cm pieces, and dried at room temperature for 5 days. The products were then calcined in a tube muffle furnace (GSL-1100X, MIT Corp., Korea) at 900 °C for 1 h under a N₂ atmosphere. The as-prepared carbon foam was 1.5 cm × 3.4 cm × 0.5 cm. Graphite felt was purchased from the Shanghai Xinxing Carbon Corp., China.

Fig. 1 Schematics of 3D N-doped open-porous carbon foam preparation and MFC set-up.

MFC construction and operation

Two H-type MFCs, one with graphite felt and the other with the as-prepared carbon foam as the anode electrode were constructed. As described previously,²⁰ the H-type MFCs were prepared from two 300 mL glass bottles that were, separated physically by a Nafion 117 membrane. Carbon paper coated with a 0.5 mg cm⁻² Pt catalyst loading on one side (Fuel Cell Earth LLC, USA) was used as the cathode electrode. The cathode chamber was filled with phosphate-buffered saline, pH 7 (NH₄Cl 0.31 g L⁻¹, KCl 0.13 g L⁻¹, NaH₂PO₄·2H₂O 3.32 g L⁻¹, Na₂HPO₄·12H₂O 10.36 g L⁻¹) and purged continuously with air at 150 cm³ air per minute. The anode chamber contained 250 mL of Luria broth (LB) medium inoculated with overnight cultures of S. oneidensis at a ratio of 1 : 100. The medium used to inoculate the bacteria and fill the anode chamber was purged with nitrogen gas for 10 min to remove dissolved oxygen. To avoid contamination, the MFCs were autoclaved at 121 °C for 30 min before set-up and all experiments were maintained under sterile conditions. The electrodes were hung by titanium wire and connected to an external resistance box with a fixed load of 1000 Ω. The voltage produced was monitored continuously using a digital multi-meter (Agilent 34405A, Agilent Technologies, Inc., USA) connected to a personal computer. The power density was calculated as P = V²/(AR_ext), where V is the measured voltage, A is the surface area of the anode electrode and R_ext is the external resistance. A polarization curve was obtained by varying the external resistance from 6000 Ω to 100 Ω with each resistor stabilized for 30 min. To obtain reliable and producible data, the experiments were repeated three times. Each set of experiments consisted of two MFCs, one with the commercial graphite felt and the other with the as-prepared carbon foam as the anode. One set of representative data is shown.

Analysis methods

Physical and electrochemical analyses. The chemical state and surface composition was analyzed by X-ray photoelectron spectroscopy (XPS) (ESCALAB 250 XPS system, Thermo Fisher Scientific Inc., UK) using a monochromatized Kα X-ray source (hν = 1486.6 eV). Fourier Transform Infrared Spectroscopy (FT-IR) was carried out on a PerkinElmer Frontier FTIR spectrophotometer (USA) over the wave number range of 400–4000 cm⁻¹. The textural properties of the electrodes were determined by N₂ sorption using a volumetric gas adsorption apparatus (ASAP 2020, Micromeritics Instrument Corp, USA) at 77 K and mercury intrusion porosimetry (AutoPore IV 9500, Micromeritics Instrument Corp., USA). The static water contact angles on anode surface were measured using a video-based an optical contact angle measuring system (OCA 40 Micro, DataPhysics Instruments GmbH, Germany).

The electrochemical studies, such as cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), were carried out using a potentiostat (Versa STAT 3, Princeton Research, USA). Both CV (scan rate of 10 mV s⁻¹) and EIS (0.01 Hz to 100 kHz, amplitude 10 mV) were conducted in a three-electrode single chamber reactor with a platinum plate as the counter electrode, Ag/AgCl saturated KCl as the reference electrode and bare carbon foam/graphite felt as the working electrode. All experiments were conducted in LB media as an electrolyte at room temperature.

Scanning electron microscopy (SEM). The surface morphologies of the anodes before and after use in the MFCs were examined by scanning electron microscopy (SEM) (S-4100, Hitachi Ltd., Japan). The morphology of the bacteria attached to the anode electrodes was examined after 1 day operation in the MFCs using the protocol described elsewhere.²¹ Briefly, the electrode with the biofilm was removed from the MFCs and washed with phosphate buffer (pH = 7) to remove the loosely adhering cells, then cut into 0.5 cm × 0.5 cm pieces. The specimens were incubated at overnight 4 °C with glutaraldehyde (2.5% final) and formaldehyde (2% final). On the next day, the specimens were washed with 0.2 M sodium phosphate buffer before being fixed for 90 min with an osmium solution (containing 1.5 mL of sodium phosphate buffer 0.2 M, 3 mL of 2% OsO₄ and 3 mL deionized water). Subsequently, specimens were dehydrated in a graded series of aqueous ethanol solutions (50%, 70%, 80%, 90%, 95%, and 100%). After dehydration, the specimens were incubated in amyl acetate for 20 min and dried using a critical-point dryer (HCP-2, Hitachi Ltd., Japan). Prior to the SEM observations, the specimens were coated with platinum for 200 seconds by ion sputtering (E-1030, Hitachi Ltd., Japan).

Biocompatibility test. The biocompatibility test was conducted by counting the colony-forming units (CFUs) on electrodes.²² The graphite felt and carbon foam electrodes were suspended in the anode chamber filled with the bacterial suspension. The electrodes were removed from the MFCs after 1 day operation and washed gently to remove the loosely adhering cells. They were then cut into 1 cm × 1 cm pieces and the bacterial cells on that piece of the electrode were dispersed in 15 mL sterile phosphate buffer by ultrasonication for 10 min. The diluted bacteria suspension was then spread over an agar plate and incubated overnight at 37 °C. The CFU was determined by counting the colonies on the agar plate. The analysis was performed in triplicate.

Results and discussion

Surface morphology and characteristics of the electrode materials

SEM observation. Fig. 2 shows the surface morphology of the as-prepared carbon foam and the commercial graphite felt. The as-prepared carbon foam exhibited a 3D open-porous scaffold structure with a pore diameter of up to 100 μm (Fig. 2a) and a wrinkle surface cell wall (Fig. 2b). The commercial graphite felt had a fibrous structure composed of randomly dispersed carbon fibers ∼10 μm in diameter (Fig. 2c). The commercial graphite fiber surface was relatively smooth, which can hinder bacteria colonization inside the material. On the other hand, the as-prepared carbon foam possessed an open-porous scaffold morphology with a rougher surface and higher surface area, which could provide enhanced attachment sites for biofilm formation without clogging.

Fig. 2 SEM images of the as-prepared carbon foam (a and b) and the commercial graphite felt (c and d).

Pore size and specific surface area. The N₂ sorption analysis at 77 K showed that the as-prepared carbon foam has a very high BET surface area, 687.19 m² g⁻¹. As shown in Fig. 3, the BET isotherm is most compatible with the Type-I + Type-IV isotherm, exhibiting microporous and mesoporous textural properties. Moreover, the Type H2 hysteresis loop, which defines the interconnected networks of pores of different sizes and shapes, is present as distinguished by the IUPAC 1985 classification.²³ On the other hand, N₂ sorption analysis failed to measure the surface area of the graphite felt, even though the measurement had been repeated many times, which confirms the absence of internal porosity inside the fibers. The surface area of the as-prepared carbon foam was quite high, and even much more robust than the commercial vitreous carbon foam (0.075 m² g⁻¹) presented elsewhere.¹⁷ This property of graphite felt has been well-characterized.^24,25 Thus, the filamentary analog method was used to calculate the specific surface area. As a result, the graphite felt had a surface area of 0.13 m² g⁻¹,²⁵ which was 5471 times smaller than that of the as-prepared carbon foam. The small surface area of the graphite felt was consistent with other studies.^14,24

Fig. 3 N₂ sorption isotherm of the as-prepared carbon foam at 77 K.

Because the microorganisms in the MFCs anode are a few micrometers in size, the micro-porous surface area determined using N₂ sorption is not accessible to microorganisms and not helpful for predicting the performance of materials in MFCs. In this case, mercury intrusion porosimetry has greater relevance for the inner surface because it can determine the wide pore diameter range (3 nm to 200 μm), whereas N₂ sorption can determine the pore diameter range from 0.3 to 300 nm only.²⁶ Therefore, in this study, mercury intrusion porosimetry was used to characterize the structural properties (porosity, bulk density, mean pore diameter, and total pore area) of the as-prepared carbon foam. As shown in Table 1, significant differences in mean pore diameter and total pore area were observed between the as-prepared carbon foam and commercial graphite felt. The mean pore diameter of the as-prepared carbon foam and the commercial graphite felt were 16 μm and 159.16 μm, respectively. From a structural viewpoint, the commercial graphite felt has a different geometrical structure from the as-prepared carbon foam. The large pore diameter of the commercial graphite felt was caused by the void space between the carbon fibers (so that a higher porosity was observed). Therefore, the pore diameter of the commercial graphite felt has no meaning from a geometrical view point.²⁴ The slightly higher total intrusion volume and bulk density of the commercial graphite felt compared to the as-prepared carbon foam were attributed to the higher pore size of the commercial graphite felt. The as-prepared carbon foam has a mean pore diameter of 16 μm, which is large enough to allow S. oneidensis, 0.2 μm in diameter and 1–2 μm in length to enter and colonize easily. Moreover, the open-porous scaffold structure of the as-prepared carbon foam had a larger pore area of 0.79 m² g⁻¹, which is approximately 7 times higher than that of the commercial graphite felt. Therefore, the as-prepared carbon foam is expected to provide a larger active space inside the pores to allow the internal colonization of S. oneidensis bacteria and good mass transfer between the solution and biofilm. Therefore, extracellular electron transfer from the bacteria to the anode was facilitated, thereby promoting power generation in the MFCs. In summary, a noticeable increase in both the specific surface area and total pore area was observed, as in the case of the as-prepared carbon foam anodes, which is expected to impart high performance to MFCs.

Table 1 Comparisons of the structural properties of the commercial graphite felt and the as-prepared carbon foam

	Unit	Graphite felt	Carbon foam
Total intrusion volume	mL g^−1	4.28	3.1414
Total pore area	m^2 g^−1	0.11	0.79
Mean pore diameter	μm	159.14	16.00
Bulk density	g mL^−1	0.21	0.26
Porosity	%	88.03	81.76

---

Elemental distribution and surface functional groups. The surface functional groups of the as-synthesized carbon foam and the graphite felt were identified qualitatively by FT-IR. As shown in Fig. S1,† the FT-IR absorption bands of the as-synthesized carbon foam and the graphite felt are mainly centered at about 3430, 2920, 1630, 1380, and 1090 cm⁻¹, which are quite similar with that of N-doped carbon foam presented so far.²⁷ The characteristic peak at 3430 cm⁻¹ is attributed to the O–H stretching vibration of surface hydroxyl and acid functional group.²⁸ The peak positioned at 2920 cm⁻¹ corresponds to the stretching vibration of C–H bond.²⁹ The strong and sharp peak centered at 1630 cm⁻¹ is identified as C=X (X = C, N, or O) stretching modes, while the peaks at 1090 cm⁻¹ and 1380 cm⁻¹ assigned to existence of the C–N bonding.²⁷ According to the FT-IR results, it is clear that nitrogen/oxygen-containing functional groups have been introduced in the as-synthesized carbon foam. Since FT-IR technique identifies the surface functional groups qualitatively, FT-IR spectra cannot probe the difference in elemental distribution and surface functional groups between the as-synthesized carbon foam and graphite felt. In order to do that, XPS technique is required.

Fig. 4a shows the survey scan of XPS for the primary C, N and O composition at 284.57, 399.74 and 532.28 eV, respectively.³⁰ The most significant difference in the composition of the as-prepared carbon foam and the commercial graphite felt was the appearance of a N peak in the as-prepared carbon foam. This confirms that the as-prepared carbon foam is N-doped.³¹ Table 2 listed the detailed atomic C, N and O content of the graphite felt and the as-prepared carbon foam. The as-prepared carbon foam had a N/C ratio of 11.13, which was approximately 10 times higher than that of the commercial graphite felt. The increase in N content of the as-prepared carbon foam anode can be a key factor for improving the power production of MFCs, which is consistent with previous studies.^11,32

Fig. 4 Survey XPS scan of the commercial graphite felt and carbon foam electrode (a). C1s spectrum of the graphite felt (b) and the as-prepared carbon foam (c). N1s spectrum of the as-prepared carbon foam (d).

Table 2 Content of C, N, and O on the commercial graphite felt and the as-prepared carbon foam surfaces

Electrodes	C1s (%)	O1s (%)	N1s (%)	N/C (%)
Graphite felt	95.54	3.25	1.21	1.27
Carbon foam	77.99	13.33	8.68	11.13

---

The surface functional groups on the commercial graphite felt and the as-prepared carbon foam were also investigated. As shown in Fig. 4b, the C1s spectrum of the commercial graphite felt was fitted to the main band at 284.6 eV, which was attributed to C–C bonds, identified as graphitic carbon.^33,34 As shown in Fig. 4c, in addition the C–C bonds, the as-prepared carbon foam showed two new peaks at 285.88 eV and 288.55 eV, which were assigned to sp² C–N³⁴ and N–C=O bonds,³⁵ respectively. The more detailed information on chemical interaction peaks was obtained from the fitted N1s XP spectra of the as-prepared carbon foam. As shown in Fig. 4d, the N1s spectrum of the as-prepared carbon foam was deconvoluted into three peaks at 399.48 eV, 400.28 eV and 401.08 eV. The peaks at approximately 399.48 eV and 400.28 eV were assigned to the C=N–C^31,35 and C=N chemical interaction peaks,³⁶ respectively. The feature at approximately 401.08 eV was assigned to N–C=O bond.^37,38 The fitted XP spectra suggested that more N- and O-containing active functional groups were generated on the as-prepared carbon foam surface which are in good agreement with FT-IR measurement. It is well known that the N- and O-containing functional group determines the hydrophilicity of the electrodes.^14,39 Therefore, the larger number of N- and O-containing active functional groups generated on the as-prepared carbon foam surface enhances the hydrophilicity of the carbon foam electrode and hence enhances biofilm formation, which will be analyzed in next section. Furthermore, the presence of N-containing functional groups at the surface could facilitate electron transfer between the biofilm and electrodes,^10,40 which may result in an improvement in power generation.⁴¹

Hydrophilicity of electrode materials. The hydrophilic properties of the anode surface were determined by the static water contact angles. As shown in Fig. 5a, the contact angle of water on the commercial graphite felt is 95°, which demonstrates the hydrophobic nature of the commercial graphite felt.^14,24 It is noticed that the water droplet immediately penetrated the foam resulted in a contact angle lower than 10° (Fig. 5b). This result indicates the hydrophilic nature of the as-prepared carbon foam. The hydrophilic nature of the as-prepared carbon foam is attributed to the presence of hydrophilic functional groups (as shown in Fig. 4d). Compared to the hydrophobic surface of the graphite felt, the hydrophilic surface of the as-prepared carbon foam is more approachable for bacterial colonization and hence increases biocompatibility of anode, which is confirmed by the biocompatibility results presented in next section.

Fig. 5 Water contact angle (CA) of the commercial graphite felt (a) and the as-prepared carbon foam (b).

Biocompatibility

The biocompatibility of the anode electrode with a biocatalyst is a critical factor that determines the MFC power generation.⁴¹ In this study, the biocompatibility was tested by counting the viable bacteria attached to electrodes and SEM observations. As shown in Fig. 6, the number of viable bacteria attached to the as-prepared carbon foam was 334 ± 47 CFU cm⁻², which was 2.6 times higher than that attached to the commercial graphite felt (127 ± 50 CFU cm⁻²). Therefore, the as-prepared carbon foam exhibited significantly higher biofilm formation, which is in accordance with the SEM observations.

Fig. 6 Colony-forming unit (CFU) of S. oneidensis on the commercial graphite felt and the as-prepared carbon foam surface.

The as-prepared carbon foam anode exhibited a more intense biofilm on the surface and walls of the carbon foam (Fig. 7a–d) compared to the commercial graphite felt (Fig. 7e and f). S. oneidensis adhered and spread over the surface, cell walls and even inside the pore window of the carbon foam. Bacteria can occupy not only the inside of the large pore window (∼100 μm, as shown in Fig. 7a), but also the small cracks on the surface (∼3 μm, as shown in Fig. 7d). This suggests that a 3D biofilm had formed and the internal colonization of S. oneidensis inside pore had been maintained. In contrast, only a few S. oneidensis bacteria were observed on the surface of the commercial graphite felt anode (Fig. 7e and f). Most of the bacteria concentrated on the polar areas of the graphite fibers (as marked by the arrow in Fig. 7f), where the surface was much rougher than the smooth wall. These results show that the biocompatibility of the as-prepared carbon foam was superior to that of the commercial graphite felt. Wang et al.¹⁴ also reported the poor biocompatibility and hydrophobic nature of the graphite felt. The excellent biocompatibility of the as-prepared carbon foam was attributed to the rougher surface morphology of the carbon foam, which enables a larger active surface area for the formation of bacteria, thereby facilitating more active biomass attachment.⁴¹ In addition, the more hydrophilic functional groups on the as-prepared carbon foam surface can also cause an increase in hydrophilicity of the electrodes and enhance biofilm formation and development.³⁹

Fig. 7 Morphology of S. oneidensis attached to the surface of the as-prepared carbon foam (a–d) and the commercial graphite felt anodes (e and f).

Electrochemical behavior of electrode materials

The electrochemical behavior of the electrode was examined by cyclic voltammetry (CV) at a scan rate of 10 mV s⁻¹ over the voltage range from −1.0 to 0.8 V. As shown in Fig. 8a, at the same geometric area of the electrode, the as-synthesized carbon foam exhibited much higher capacitance currents than the graphite felt, indicating that the carbon foam has higher electroactive surface area.⁴² Therefore, as-synthesized carbon foam can provide a larger active surface area for bacterial attachment and facilitate conduction of electrons produced in the anode chamber, which expect to provide the MFCs with superior performance.

Fig. 8 CVs of the commercial graphite felt and the as-prepared carbon foam at a scan rate of 10 mV s⁻¹ (a). Nyquist plot of EIS results of the graphite felt and the as-prepared carbon foam electrodes. Frequency 0.01 Hz to 100 kHz, amplitude 10 mV (b). The insert shows an enlarged plot at high frequency.

In addition to the surface area, the internal resistance of the anode electrode is one of the major factors affecting the power density of MFCs. EIS was carried out to determine the electron transfer mechanism in the bare carbon foam and graphite felt. In the Nyquist plot, the semicircular diameter feature observed in the medium frequency is related to the charge-transfer resistance of the electrode.⁴³ The Nyquist impedance plots in Fig. 8b showed that the charge transfer resistance through the electrode/electrolyte interface of the carbon foam was significantly lower than that of the graphite felt. The lower charge transfer resistance of the carbon foam resulted in a higher electron transfer rate,⁷ thereby improving the MFC performance.

Microbial fuel cell performance

Fig. 9 shows the performance parameters, including power density generation, open circuit potential (OCV), polarization, and power density curves of two MFCs with the carbon foam and graphite felt anodes. As shown in Fig. 9a, the MFC with the carbon foam anode produced a maximum power density of 96 mW m⁻³, which was approximately double that with graphite felt anode (47 mW m⁻³). Compared to the MFC equipped with reticulated carbon foam,¹⁷ the MFC equipped with the as-synthesized carbon foam produced 2.4 times higher power density, which highlighted the superior of as-synthesized carbon foam. On the other hand, the polarization curve showed that both MFCs have a similar OCV of 0.8 V, which indicates that the two systems have similar activation resistance (Fig. 9b). At a higher current, the voltage of the MFC with the carbon foam anode felt more slowly, representing the lower ohmic losses and mass transport loss compared to the MFC with the commercial graphite felt anode.² In other words, the results suggest that the MFC with the as-prepared carbon foam electrode could reduce the internal resistance. Indeed, the internal resistance calculated by the power density peak method⁴⁴ showed a decrease in internal resistance from 1064.6 to 777 Ω when the commercial graphite felt anode was replaced with the as-prepared carbon foam anode. This result is in good accordance with the EIS results.

Fig. 9 Power curve (a) and polarization curve (b) of MFCs equipped with the graphite felt and the as-prepared carbon foam anodes.

These results strongly suggest that the as-prepared carbon foam anode enables superior performance of the MFC, which can be explained as follows. First, the enhanced electricity generation might be due to the 3D N-doped open-porous carbon foam with a high surface area and porosity, as observed by N₂ sorption, mercury porosimetry and SEM. An anode with a high microorganism accessible surface area is essential to the performance of MFC. This is because it enables the attachment of a larger amount of microbial biomass on the anode surface, which typically increases the current generation in the MFC. Second, the much better performance of the MFC with the as-prepared carbon foam anode was attributed to the much lower charge transfer resistance compared to the commercial graphite felt anode. Third, hydrophilic functional groups such as C=N–C and N–C=O facilitates extracellular electron transfer and enhances the biocompatibility of the as-prepared carbon foam electrode.

In addition to the superior performance in the MFC, the 3D N-doped open-porous carbon foam preparation was quite simple to prepare and was effective. The raw materials, such as melamine sponge and other basic chemicals, are inexpensive. Therefore, 3D N-doped open-porous carbon foam presents a low cost method that can improve power generation. As mentioned earlier, the electrode cost is one of the limitations for the scale up of MFCs, and may contribute up to 20–50% of the overall cost of MFCs.⁴ Therefore, the as-prepared carbon foam is a promising electrode for the large-scale applications of MFCs.

Conclusions

Cost-effective, 3D N-doped open-porous carbon foam was prepared using a simple method. Compared to the commercial graphite felt electrode, the MFC with the carbon foam electrode can produce two times higher power density. The superior performance of the carbon foam anode in the MFCs was attributed to the synergistic effects of the high surface area, good biocompatibility, excellent electron transfer, and small internal resistance which was proven by SEM, XPS, biocompatibility test, surface area measurements, and electrochemical tests. The results showed that the as-prepared 3D N-doped open-porous carbon foam is a good candidate for the electrode of a MFC.

Acknowledgements

This study was supported by Priority Research Centers Program (grant no.: 2014R1A6A1031189), and by Basic Science Research Program (grant no.: 2015R1D1A3A03018029) through the National Research Foundation of Korea (NRF) funded by the Ministry of Education.

Notes and references

1. B. E. Logan, M. J. Wallack, K. Y. Kim, W. He, Y. Feng and P. E. Saikaly, Environ. Sci. Technol. Lett., 2015, 2, 206.
2. B. E. Logan, B. Hamelers, R. Rozendal, U. Schröder, J. Keller, S. Freguia, P. Aelterman, W. Verstraete and K. Rabaey, Environ. Sci. Technol., 2006, 40, 5181.
3. H. Yuan and Z. He, Nanoscale, 2015, 7, 7022.
4. K. Rabaey, S. Bützer, S. Brown, J. Keller and R. A. Rozendal, Environ. Sci. Technol., 2010, 44, 4315.
5. G. G. Kumar, V. G. S. Sarathi and K. S. Nahm, Biosens. Bioelectron., 2013, 43, 461.
6. S. Chen, Q. Liu, G. He, Y. Zhou, M. Hanif, X. Peng, S. Wang and H. Hou, J. Mater. Chem., 2012, 22, 18609.
7. M. Liu, M. Zhou, H. Yang, Y. Zhao and Y. Hu, RSC Adv., 2015, 5, 84269.
8. Y. Yong, X. Dong, M. Chan-Park, H. Song and P. Chen, ACS Nano, 2012, 6, 2394.
9. X. Xie, M. Ye, L. Hu, N. Liu, J. McDonough, W. Chen, H. C. Alshareef, S. Criddle and Y. Cui, Energy Environ. Sci., 2012, 5, 5265.
10. S. Cheng and B. E. Logan, Electrochem. Commun., 2007, 9, 492.
11. Y. Feng, Q. Yang, X. Wang and B. E. Logan, J. Power Sources, 2010, 195, 1841.
12. K. Scott, G. Rimbu, K. P. Katuri, K. K. Prasad and I. M. Head, Process Saf. Environ. Prot., 2007, 85, 481.
13. X. Tang, K. Guo, H. Li, Z. Du and J. Tian, Bioresour. Technol., 2011, 102, 3558.
14. Y. Q. Wang, H. X. Huang, B. Li and W. S. Li, J. Mater. Chem. A, 2015, 3, 5110.
15. N. Amini, K. F. Aguey-Zinsou and Z. X. Guo, Carbon, 2011, 49, 3857.
16. S. K. Chaudhuri and D. R. Lovley, Nat. Biotechnol., 2003, 21, 1229.
17. G. Lepage, F. O. Albernaz, G. Perrier and G. Merlin, Bioresour. Technol., 2012, 124, 199.
18. S. He and W. Chen, J. Power Sources, 2014, 262, 391.
19. J. Wang, L. Shen, P. Nie, X. Yun, Y. Xu, H. Dou and X. Zhang, J. Mater. Chem. A, 2015, 3, 2853.
20. T. H. Han, M. H. Cho and J. Lee, Biotechnol. Bioprocess Eng., 2014, 19, 126.
21. T. H. Han, J. H. Lee, M. H. Cho and J. Lee, Res. Microbiol., 2011, 162, 108.
22. J. Jayapriya, J. Gopal, V. Ramamurthy, U. Kamachi Mudali and B. Raj, Composites, Part B, 2012, 43, 1329.
23. F. Rouquerol, J. Rouquerol and K. Sing, Adsorption by powders and porous solids, Academic Press, London, 1999.
24. J. González-García, P. Bonete, E. Expósito, V. Montiel, A. Aldaz and R. Torregrosa-Maciá, J. Mater. Chem., 1999, 9, 419.
25. H. Zhou, H. Zhang, P. Zhao and B. Yi, Electrochim. Acta, 2006, 51, 6304.
26. J. Rouquerol, D. Avnir, C. W. Fairbridge, D. H. Everett, J. H. Haynes, N. Pernicone, J. D. F. Ramsay, K. S. W. Sing and K. K. Unger, Pure Appl. Chem., 1994, 66, 1739.
27. Z. Du, Y. Peng, Z. Ma, C. Li, J. Yang, X. Qin and G. Shao, RSC Adv., 2015, 5, 10296.
28. S. Y. Sawant, R. S. Somani and M. H. Cho, RSC Adv., 2015, 5, 46589.
29. S. Y. Sawant, R. S. Somani, S. S. Sharma and H. C. Bajaj, Carbon, 2014, 68, 210.
30. A. P. Dementjev, A. de Graaf, M. C. M. van de Sanden, K. I. Maslakov, A. V. Naumkina and A. A. Serov, Diamond Relat. Mater., 2000, 9, 1904.
31. J. Zhou, Z. Zhang, W. Xing, J. Yu, G. Han, W. Si and S. Zhuo, Electrochim. Acta, 2015, 153, 68.
32. T. Saito, M. Mehanna, X. Wang, R. D. Cusick, Y. Feng, M. A. Hickner and B. E. Logan, Bioresour. Technol., 2011, 102, 395.
33. G. V. Ouedraogo, D. Benlian and L. Porte, J. Chem. Phys., 1980, 73, 642.
34. K. J. Boyd, D. Marton, S. S. Todorov, A. H. Al-Bayati, J. Kulik, R. A. Zuhr and J. W. Rabalais, J. Vac. Sci. Technol., A, 1995, 13, 2110.
35. J. Peeling, F. Hruska, D. M. McKinnon, M. S. Chauhan and N. S. Mclntyre, Can. J. Chem., 1978, 56, 2405.
36. F. Rossi, B. Andre, A. van Veen, P. E. Mijnarends, H. Schut, F. Labohm, H. Dunlop, M. P. Delplancke and K. Hubbard, J. Mater. Res., 1994, 9, 2440.
37. B. Lindberg, A. Berndtsson, R. Nilsson, R. Nyholm and O. Exner, Acta Chem. Scand., Ser. A, 1978, 32, 353.
38. K. B. Yatsimirskii, V. V. Nemoshalenko, V. G. Aleshin, Y. I. BratushkoI and E. P. Moiseenko, Chem. Phys. Lett., 1977, 52, 481.
39. J. A. Cornejo, C. Lopez, S. Babanova, C. Santoro, K. Artyushkova, L. Ista, A. J. Schuler and P. Atanassova, J. Electrochem. Soc., 2015, 162, H597.
40. K. P. Gong, F. Du, Z. H. Xia, M. Durstock and L. M. Dai, Science, 2009, 323, 760.
41. J. Zhang, J. Li, D. Ye, X. Zhu, Q. Liao and B. Zhang, J. Power Sources, 2014, 272, 277.
42. L. Yang, S. Wang, S. Peng, H. Jiang, Y. Zhang, W. Deng, Y. Tan, M. Ma and Q. Xie, Chem.–Eur. J., 2015, 21, 10634.
43. Z. S. Wu, W. Ren, L. Xu, F. Li and H. M. Cheng, ACS Nano, 2011, 5, 5463.
44. B. E. Logan, Microbial fuel cells, Wiley, New Jersey, 2007.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra01842d

‡ Thi Hiep Han and Sandesh Y. Sawant contributed equally to this work.

---