Single-Chamber Microbial Fuel Cell with Multiple Plates of Bamboo Charcoal Anode: Performance Evaluation

In this study, three single-chamber microbial fuel cells (MFCs), each having Pt-coated carbon cloth as a cathode and four bamboo charcoal (BC) plates as an anode, were run in a fed-batch mode, individually and in series. Simulated potato-processing wastewater was used as a substrate for supporting the growth of a mixed bacterial culture. The maximum power output increased from 0.386 mW with one MFC to 1.047 mW with three MFCs connected in series. The maximum power density, however, decreased from 576 mW/m2 (normalized to the cathode area) with one MFC to 520 mW/m2 with three MFCs in series. The experimental results showed that power can be increased by connecting the MFCs in series; however, choosing low resistance BC is crucial for increasing power density.

MFCs are fundamentally classified into dual-chamber (or two-chamber) and singlechamber MFCs. The dual-chamber MFC consists of an anode chamber and cathode chamber. The anode chamber contains an anode electrode, microorganisms (biocatalysts), and anolyte (substrate, fuel source), while the cathode chamber houses a cathode electrode and catholyte. The two chambers are interfaced by a cation-specific or proton exchange membrane (PEM), and connected with an external circuit (conductive wire). In the singlechamber MFC design, the cathode chamber, catholyte, and PEM are removed, and one side of the cathode is directly exposed to oxidants, usually to the atmosphere (air cathode), and the other side to the anolyte. The air cathode allows oxygen (O 2 ) in the air to contact with protons (H + ) to form water (H 2 O) at the cathode surface. The present study used the single-chamber air-cathode MFC, as it has several advantages over the dual-chamber MFC, including (i) reduced total MFC volume, (ii) reduced internal resistance, (iii) simpler operation and maintenance, (iv) lower construction costs, and (v) an enhanced O 2 reduction rate on the cathode owing to freely available O 2 in the air [12][13][14][15]. Although considerable progress and advancement has been made in recent years, the MFC performance is still insufficient for large-scale industrial applications.

MFC Design and Construction
Three air-cathode single-chamber MFCs, namely MFC 1, MFC 2, and MFC 3, were constructed and operated in a fed-batch cyclic mode. Specifically, the body of each MFC was constructed using acrylic sheets (0.6 cm thickness), which were cut and fastened together ( Figure 1a). The outside dimension of each fuel cell is 13 cm long, 9 cm wide, and 11 cm in height. The top and cathode side covers can be detached from the main body so that the MFC can be cleaned and the cathode can be replaced when needed. Two holes were tapped on the top of the MFC to allow for the removal of a spent substrate solution and the feeding of a fresh solution. These holes were obturated during the operation to maintain an anaerobic condition. In order to expose the cathode to the atmosphere, a circular opening (3 cm diameter) was drilled into the acrylic sheet that holds the cathode. The cathode electrode was made of carbon cloth with platinum loadings of 0.3 mg/cm 2 (Fuel Cell Earth, Woburn, MA, USA). The apparent surface area (one side) of the cathode was 6.7 cm 2 .
The anode electrode was made of bamboo charcoal (Mt Meru Pte, Singapore). Four plates of bamboo charcoal, namely BC1, BC2, BC3, and BC4, were placed as the anode inside the MFC body ( Figure 1b). The BC plates were spaced 2 cm from each other, and In order to expose the cathode to the atmosphere, a circular opening (3 cm diameter) was drilled into the acrylic sheet that holds the cathode. The cathode electrode was made Processes 2021, 9, 2194 3 of 12 of carbon cloth with platinum loadings of 0.3 mg/cm 2 (Fuel Cell Earth, Woburn, MA, USA). The apparent surface area (one side) of the cathode was 6.7 cm 2 .
The anode electrode was made of bamboo charcoal (Mt Meru Pte, Singapore). Four plates of bamboo charcoal, namely BC1, BC2, BC3, and BC4, were placed as the anode inside the MFC body ( Figure 1b). The BC plates were spaced 2 cm from each other, and the distance between the cathode and the nearest BC plate was 3 cm. Specifications of the BC anode plates are presented in Table S1, as supporting information. These BC plates were selected among the BC plates of different resistances on the basis of low resistance. The BC surface was visualized using a scanning electron microscope (SEM) (Quanta 200 FEG, FEI Co., Czech Republic). The apparent (geometric) surface area of each BC anode plate was typically 75 cm 2 ranging from 64 to 81 cm 2 , and the lengthwise and widthwise resistance of the BC anode plates was 30 Ω on average in a range from 16 to 54 Ω. The average resistance of the BC electrodes of MFC 1, MFC 2, and MFC 3 were 26, 24, and 40 Ω, respectively. A stainless-steel wire lead was attached to each electrode using conductive epoxy adhesive (MG Chemicals, Ontario, Canada), and its surface was covered by nonconductive epoxy putty (J-B Weld, Coopersburg, PA, USA) for reinforcement. For each BC anode plate, a hole was drilled in the bottom of the MFC to allow the wire leads to pass through the bottom sheet to connect the electrodes externally. The anode and the cathode were connected through an external resistor to close the circuit. After the installation of the BC anode plates, the working volumes (or anolyte volumes) of MFC-1, MFC-2, and MFC-3 were 530, 530, and 500 mL, respectively.

Experimental Conditions and Analyses
Simulated potato-processing wastewater was used as a substrate (anolyte). To prepare the wastewater, concentrated potato extract obtained from a local food processing plant in Idaho Falls, Idaho, USA, was diluted to the desired concentration (approximately 1000 mg/L) of chemical oxygen demand (COD) using a pH-7 phosphate buffer solution. The potato extract was analyzed by ICP-MS (X-Series II, Thermo Scientific, Germany), and results are presented in Table S2, as supporting information. The wastewater was autoclaved at 121 • C for 20 min to assure that the substrate was free from active microbes, and stored in a refrigerator at 4 • C until used.
The bacterial culture used in this study originated from an anaerobic digester at the Wastewater Treatment Plant in the City of Pocatello, Idaho. The anaerobic mixed bacterial community was acclimated in sterilized synthetic potato-processing wastewater for more than 10 years by the feeding aforementioned sterilized substrate every two weeks to a month, and thus the inoculum was considered to be fully acclimated. The previous study revealed that microbial species present in the anode chamber were predominantly within the phyla of Proteobacteria, Firmicutes, and Bacteroidetes [77].
The organic strength of the wastewater (substrate solution) was measured as COD at the beginning and end of the run cycle. The loss of anolyte (wastewater) in the chamber was compensated by filling deionized water before taking samples. The COD measurement was carried out using the USEPA Reactor Digestion Method (HACH, USA). Wastewater samples were filtered using a 0.2 µm Millex-VV filter (Millipore, France) before the COD analysis. The removal efficiency of individual MFC was approximately 80% in one cycle run.

MFC Operation and Monitoring
The MFCs were run with five different settings: (1) MFC 1 only; (2) MFC 2 only; (3) MFC 3 only; (4) MFC 1 and MFC 2 in-series (noted as MFC 1-2), and (5) MFC 1, MFC 2, and MFC 3 in-series (noted as MFC 1-2-3). The MFCs were operated in a fed-batch mode at 30 ± 2 • C. The MFCs were considered to have reached a quasi-steady state when a stable voltage cycle was achieved, i.e., the MFCs produced similar voltage profiles and repeated at least three cycles. When the system reached a stable condition and voltage reached a high plateau region, the external resistance in the circuit was varied in a range from 20 to 800 Ω to obtain the maximum power output. The voltage produced by the MFCs was measured using a data acquisition system (DAS) and continuously recorded every 15 min. The DAS consisted of a computer equipped with a data acquisition board (PCl-6024E, National Instrument, Austin, TX, USA) and an external connection block (SCB-6024E, National Instrument). The DAS was controlled by LabView-2014 (National Instruments, Austin, TX, USA). The potential difference between the cathode (+) and anode (−) is voltage (V) across a resistor (R). Current (I) was calculated from Ohm's law: I = V/R. Power (P) was calculated by P = I·V or P = V 2 /R. To analyze the performance of MFC, power density and polarization curves were developed. Current and power densities were normalized to the cathode area, anode area, and effective chamber (anolyte) volume. The vales of internal resistance were determined according to the maximum power transfer theorem [78].

Bamboo Charcoal Anode
As was described previously, a large variety of carbon materials have been studied for anode electrodes to improve MFC performance. The present study used bamboo charcoal (BC) for the anode. In the MFC design, electrodes are desired to have the following characteristics: (i) high electrical conductivity and low resistance; (ii) high mechanical strength; (iii) strong biocompatibility, (iv) high chemical stability and corrosion resistance, (v) large surface area, (vi) low cost, and vii) environmental friendliness [1]. Bamboo charcoal meets most of the aforementioned characteristics [66,67,76,[79][80][81][82][83][84][85][86][87][88]. Figure 2 shows SEM images of the external surface and cross-section of the BC anode. The electron micrographs show that the BC used in this study has a rough fiber-like external surface ( Figure 2a) and is comprised of dense round to hexagonal tubes from 20 to 500 µm in diameter, arranged in a honeycomb-like pattern (Figure 2b). The interconnected porous architecture can provide a large surface area for microbial cell attachment and biofilm growth, and facilitates the diffusion of nutrients and protons. These characteristics are essential to the anode for high performance. The major drawback is that being a natural plant, the specifications of the BC such as the thickness and resistance can vary (Table S1). It should be noted that BC plates used in MFC 3 had noticeably larger resistances compared to those of MFC 1 and MFC 2. In a study with BC and carbon fiber anodes, Moqsud et al. [67] found that bamboo charcoal was an effective anode material. A comparison of tubular BC with tubular graphite showed that BC was a better anode material in terms of surface roughness, biocompatibility, electron transfer, and total internal resistance [89]. Additional advantages of using bamboo charcoal are that bamboo is a fast-growing plant [90] and thus it sequesters CO2 rapidly from the atmosphere, stabilizing carbon in a solid form [91]. Bamboo charcoal can be easily manufactured at low cost and with minimal carbon footprint, reused, and disposed of safely (e.g., as biochar) as it is a natural material. In a study with BC and carbon fiber anodes, Moqsud et al. [67] found that bamboo charcoal was an effective anode material. A comparison of tubular BC with tubular graphite showed that BC was a better anode material in terms of surface roughness, biocompatibility, electron transfer, and total internal resistance [89]. Additional advantages of using bamboo Processes 2021, 9, 2194 5 of 12 charcoal are that bamboo is a fast-growing plant [90] and thus it sequesters CO 2 rapidly from the atmosphere, stabilizing carbon in a solid form [91]. Bamboo charcoal can be easily manufactured at low cost and with minimal carbon footprint, reused, and disposed of safely (e.g., as biochar) as it is a natural material.

MFC Performance
The performance of MFCs is generally expressed in either power output or power density normalized to the cathode area, anode area, or effective anode chamber (or anolyte) volume. Preliminary examinations of the MFC 1, MFC 2, and MFC 3 runs in the present study showed that these four expressions are comparable (see Figure S1, supporting information). Hence, in this paper, the power density and current density were normalized to the projected cathode area. Note that all the MFCs used in this study had the same cathode area (6.7 × 10 −4 m 2 ).
The experimental results are analyzed in two stages. Firstly, the performance of MFC 1, MFC 2, and MFC 3 are analyzed individually. Figure 3a shows plots of power (mW) against corresponding current density (mA/m 2 ) for MFC1 (circles), MFC 2 (squares), and MFC 3 (triangles). Lines are third-order polynomial fits to guide the eyes. As is seen, MFC 1 and MFC 2 produced similar power curves, while MFC 3 exhibited a considerably smaller power curve. The maximum power achieved by MFC 1 and MFC 2 is 0.386 and 0.383 mW at a current density of 3.27 × 10 3 and 2.67 × 10 3 mA/m 2 , respectively. The maximum power yielded by MFC 3 is 0.284 mW at a current density of 2.09 × 10 3 mA/m 2 . The trends of polarization for MFC 1, MFC 2, and MFC 3 are shown in Figure 3b. The polarization curves produced by MFC 1 and MFC 2 are nearly identical, whereas the voltage produced by MFC 3 is noticeably smaller, consistent with the power curves. This smaller voltage produced by MFC 3 is most likely attributed to the larger internal resistance of MFC 3.   Figure 4 shows the correlation between current (at the maximum power) and internal resistance. As is seen, current and internal resistance are inversely related. In a comparison of MFC 1 and MFC 3, a 51% increase in internal resistance resulted in a 28% decrease in current.  Figure 4 shows the correlation between current (at the maximum power) and internal resistance. As is seen, current and internal resistance are inversely related. In a comparison of MFC 1 and MFC 3, a 51% increase in internal resistance resulted in a 28% decrease in current.   Figure 4 shows the correlation between current (at the maximum power) and internal resistance. As is seen, current and internal resistance are inversely related. In a comparison of MFC 1 and MFC 3, a 51% increase in internal resistance resulted in a 28% decrease in current. Secondly, the performance of the MFCs connected in series is examined. Figure 5a shows the power curves for MFC 1 (circle), MFC 1-2 (square), and MFC 1-2-3 (triangle). As the number of the MFCs was increased, the power output increased, and the maximum power shifted to the lower current density and larger voltage (Figure 5b). The maximum power output increased from 0.386 mW (at 3.27 × 10 3 mA/m 2 ), 0.875 mW (at 1.75 × 10 3 mA/m 2 ) to 1.047 mW (at 1.14 × 10 3 mA/m 2 ) with the MFC 1, MFC 1-2, and MFC 1-2-3 settings, respectively. The increase of the power output from the MFC 1 to MFC 1-2 setting was 120%, whereas the power increase from the MFC 1-2 setting to the MFC 1-2-3 setting was only 20%. This smaller power increase in the MFC 1-2-3 setting is most likely attributed to the larger internal resistance of MFC 3. Theoretically, if two MFCs are connected in series, the system can increase the power output by increasing the voltage without increasing the current (92). However, MFC 1-2 produced approximately 15% less voltage than the theoretically available voltage, whereas it produced about 35% more current than the theoretical current. Overall, MFC 1-2 produced more than 20% greater power than the theoretical additive power. The decrease in the voltage is likely due to voltage reversal (VR), as this is a common phenomenon in stacked MFCs and MFCs in series [92]. The potential causes of VR are (a) kinetic imbalances between the MFCs in series [93], and (b) substrate limitation from microbial starving conditions [23,94]. Several VR control methods have been discussed and proposed by Kim et al. [92]. The increase in the current could be attributed to the properties of BC or the arrangement of the BC plates. A slightly Secondly, the performance of the MFCs connected in series is examined. Figure 5a shows the power curves for MFC 1 (circle), MFC 1-2 (square), and MFC 1-2-3 (triangle). As the number of the MFCs was increased, the power output increased, and the maximum power shifted to the lower current density and larger voltage (Figure 5b). The maximum power output increased from 0.386 mW (at 3.27 × 10 3 mA/m 2 ), 0.875 mW (at 1.75 × 10 3 mA/m 2 ) to 1.047 mW (at 1.14 × 10 3 mA/m 2 ) with the MFC 1, MFC 1-2, and MFC 1-2-3 settings, respectively. The increase of the power output from the MFC 1 to MFC 1-2 setting was 120%, whereas the power increase from the MFC 1-2 setting to the MFC 1-2-3 setting was only 20%. This smaller power increase in the MFC 1-2-3 setting is most likely attributed to the larger internal resistance of MFC 3. Theoretically, if two MFCs are connected in series, the system can increase the power output by increasing the voltage without increasing the current (92). However, MFC 1-2 produced approximately 15% less voltage than the theoretically available voltage, whereas it produced about 35% more current than the theoretical current. Overall, MFC 1-2 produced more than 20% greater power than the theoretical additive power. The decrease in the voltage is likely due to voltage reversal (VR), as this is a common phenomenon in stacked MFCs and MFCs in series [92]. The potential causes of VR are (a) kinetic imbalances between the MFCs in series [93], and (b) substrate limitation from microbial starving conditions [23,94]. Several VR control methods have been discussed and proposed by Kim et al. [92]. The increase in the current could be attributed to the properties of BC or the arrangement of the BC plates. A slightly super-linear increase in power output by stacking individual MFC units has been reported [95]. Nevertheless, the effects of the BC material and multiple BC plates on the MFC performance require further investigation.
Processes 2021, 9, x FOR PEER REVIEW 7 of 12 super-linear increase in power output by stacking individual MFC units has been reported [95]. Nevertheless, the effects of the BC material and multiple BC plates on the MFC performance require further investigation.  Figure 6 shows the power density curves (mW/m 2 ) for the MFC 1, MFC 1-2, and MFC 1-2-3 settings. As is seen, the power density increased from the MFC 1 to the MFC 1-2 setting but decreased in the MFC 1-2-3 setting. Although the power output was increased by adding MFCs in series, the increase in power by MFC 3 was not large enough to offset  Figure 6 shows the power density curves (mW/m 2 ) for the MFC 1, MFC 1-2, and MFC 1-2-3 settings. As is seen, the power density increased from the MFC 1 to the MFC 1-2 setting but decreased in the MFC 1-2-3 setting. Although the power output was increased by adding MFCs in series, the increase in power by MFC 3 was not large enough to offset the increase in the cathode area. The cathode area increased from 6.7 × 10 −4 m 2 to 1.34 × 10 −3 m 2 to 2.01 × 10 −3 m 2 in the MFC 1, MFC 1-2, and MFC 1-2-3 runs, respectively. Consequently, MFC 1-2-3 exhibited the lowest power density (520 mW/m 2 cathode at 1.14 × 10 3 mA/m 2 ) as compared to MFC 1 (576 mW/m 2 at 3.27 × 10 3 mA/m 2 ) and MFC 1-2 (653 mW/m 2 at 1.75 × 10 3 mA/m 2 ). The results indicate that the power output increases with an increasing number of MFCs in-series; however, the MFC that behaves poorly affects negatively the overall performance of the MFC system.  Figure 6 shows the power density curves (mW/m 2 ) for the MFC 1, MFC 1-2, and MFC 1-2-3 settings. As is seen, the power density increased from the MFC 1 to the MFC 1-2 setting but decreased in the MFC 1-2-3 setting. Although the power output was increased by adding MFCs in series, the increase in power by MFC 3 was not large enough to offset the increase in the cathode area. The cathode area increased from 6.7 × 10 −4 m 2 to 1.34 × 10 −3 m 2 to 2.01 × 10 −3 m 2 in the MFC 1, MFC 1-2, and MFC 1-2-3 runs, respectively. Consequently, MFC 1-2-3 exhibited the lowest power density (520 mW/m 2 cathode at 1.14 × 10 3 mA/m 2 ) as compared to MFC 1 (576 mW/m 2 at 3.27 × 10 3 mA/m 2 ) and MFC 1-2 (653 mW/m 2 at 1.75 × 10 3 mA/m 2 ). The results indicate that the power output increases with an increasing number of MFCs in-series; however, the MFC that behaves poorly affects negatively the overall performance of the MFC system. One of the factors that limit power output by MFCs is their internal resistance, which is comprised of activation (or charge transfer) resistance, ohmic (or solution) resistance, and concentration (or diffusion) resistance [96]. The ohmic resistance represents the resistance of solution, electrode materials, and membrane. In a single-chamber air-cathode MFC, the sources of ohmic resistance are mainly the electrodes and anolyte. Figure 7 shows the relation between the internal resistance and BC plate resistance. Note that the BC anode resistances were measured before they were placed inside the MFC chamber. As is seen, the internal resistance was smaller than the BC anode resistance, indicating One of the factors that limit power output by MFCs is their internal resistance, which is comprised of activation (or charge transfer) resistance, ohmic (or solution) resistance, and concentration (or diffusion) resistance [96]. The ohmic resistance represents the resistance of solution, electrode materials, and membrane. In a single-chamber air-cathode MFC, the sources of ohmic resistance are mainly the electrodes and anolyte. Figure 7 shows the relation between the internal resistance and BC plate resistance. Note that the BC anode resistances were measured before they were placed inside the MFC chamber. As is seen, the internal resistance was smaller than the BC anode resistance, indicating that the BC plates became more conductive, likely due to the establishment of EAB biofilm. that the BC plates became more conductive, likely due to the establishment of EAB biofilm.

Conclusions
This study focused on the single-chamber air-cathode MFCs with the bamboo charcoal (BC) anode and Pt-coated carbon cloth cathode. Three MFCs were run in a fed-batch mode, individually and in series using synthetic potato-processing wastewater as an anolyte. The maximum power output increased from 0.386 mW (at 3.27 × 10 3 mA/m 2 ) with one MFC to 1.047 mW (at 1.14 × 10 3 mA/m 2 ) with three MFCs connected in series. The maximum power density, however, decreased from 576 mW/m 2 (at 3.27 × 10 3 mA/m 2 ) with one MFC to 520 mW/m 2 cathode (at 1.14 × 10 3 mA/m 2 ) with three MFCs in series. The

Conclusions
This study focused on the single-chamber air-cathode MFCs with the bamboo charcoal (BC) anode and Pt-coated carbon cloth cathode. Three MFCs were run in a fed-batch mode, Processes 2021, 9, 2194 8 of 12 individually and in series using synthetic potato-processing wastewater as an anolyte. The maximum power output increased from 0.386 mW (at 3.27 × 10 3 mA/m 2 ) with one MFC to 1.047 mW (at 1.14 × 10 3 mA/m 2 ) with three MFCs connected in series. The maximum power density, however, decreased from 576 mW/m 2 (at 3.27 × 10 3 mA/m 2 ) with one MFC to 520 mW/m 2 cathode (at 1.14 × 10 3 mA/m 2 ) with three MFCs in series. The experimental results showed that power can be increased by connecting MFCs in series; however, choosing high-quality (low resistance) BC charcoal is crucial for increasing power density.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/pr9122194/s1: Table S1: Dimensions and resistance of bamboo charcoal analysis; Table S2: Concentrations of inorganic constituents in potato extracts (ICP analysis results); Figure S1

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to potential proprietary information.