Wood carbon electrode in microbial fuel cell enhances chromium reduction and bioelectricity generation

Microbial fuel cell (MFC) is a recommended treatment to remediate hexavalent chromium (Cr(VI)) in wastewater. In this study, a wood carbon (WC) electrode was introduced in MFC to enhance the Cr(VI) removal efficiency. WC electrode in MFC completely removed Cr(VI) as compared to the carbon cloth (31.12%) and carbon felt (34.83) within 48 h of operation at 20 mg L−1 of Cr(VI) concentration. The maximum power density of WC electrode was 62.59 mW m−2 higher than 0.115 and 3.154 mW m−2 of carbon cloth and felt respectively. The specific surface area of WC increased to 158.47 m−2 g−1 after high-temperature carbonization, and electrochemical tests indicate it has higher electrocatalytic ability. Therefore, WC might be a good electrode material to effectively remove Cr(VI) and generate bioelectricity simultaneously.


Introduction
Microbial fuel cell (MFC) is an effective biological technique for the enhancement of wastewater treatment and heavy metal removal efficiency. Heavy metal ions could serve as electron acceptors to decrease and precipitate with redox potentials (Wang and Ren 2014). Utilization of organic feedstock in anode MFC for bioelectricity harvesting reduces metal ions including Cu 2+ , Cr 6+ , and Ag + (Nancharaiah et al. 2015;Wu et al. 2018). Enhancement of voltage output and total reduction of heavy metal are the key factors of the MFC approach. Hexavalent chromium (Cr(VI)) with high solubility, toxicity, and long-term exposure to the environment caused a serious threat to human health (Pophali et al. 2021). MFC treatment could remove Cr(VI) and had more advantages including high-cost efficiency and eco-friendliness as compared to traditional treatments such as adsorption, photocatalytic reduction, electrocoagulation, and bacteria (Lei et al. 2020). TiO 2 / Fe 2 O 3 photoanode in MFC could remove 90.9% of Cr(VI) at the 50-ppm concentration within 13.5 h of incubation (Ren et al. 2018). Cr(VI) also acts as an electron acceptor, a Hongyuhang Ni, Aman Khan and Zi Yang contributed equally to this work.
Responsible Editor: Ioannis A. Katsoyiannis * Fengjuan Chen chenfj@lzu.edu.cn considerable reduction rate of it can be achieved (Niu et al. 2018). However, the forms of Cr(VI) in the cathode of MFC (CrO 4 2− and Cr 2 O 7 2− ) will lead to a charge repulsion reaction due to the electrons with the same negative charge and thus lower the reduction efficiency and electricity output . However, MFC is restricted because of less removal efficiency of metals and low generation along with its high cost (Choudhury et al. 2017). It is essential to identify the key steps that would optimize the process of MFC and enhance Cr(VI) removal efficiency.
Modification of electrode materials is a vital method to enhance the usability and performance of MFC, such as the increment of power generation and reduction of organic pollutants (Sekar et al. 2019;Yellappa et al. 2019). A wide variety of electrodes have been used in MFC depending upon the applications and feasibility of the system. For example, carbon felt like a support material for immobilization of multi-wall carbon nanotube (MWCNT) enhances nitrate reduction efficiency in dual chamber bio-electrochemical system Safari et al. 2014). Al/Ni nanoparticles were prepared on the surface of carbon nanofiber, and reduction rate of Cr(VI) was increased to 100% (Gupta et al. 2017). Moreover, a FeS@rGO modified graphite felt removes 1.43 mg L −1 h −1 of Cr(VI) with 154 mW m −2 of power density in MFC (Ali et al. 2019). However, the high cost and complex preparation method of modified electrodes still become limiting factors for the application of MFC. Hence, the development of electrode materials with good catalytic ability and a simple preparation process is required.
Recently, natural biomass materials have been receiving attention and used in MFC due to their inherent porous structure, high electrical conductivity, and low-cost properties (Bi et al. 2019;Tang et al. 2017). For example, kenaf, pomelo peel, and king mushroom have been carbonized under high temperatures and enhanced voltage generation and heavy metal removal efficiency MFC (Chen et al. 2012a;Chen et al. 2012b;Karthikeyan et al. 2015). Natural carbon materials have a strong electrocatalytic ability (Gao et al. 2015). Therefore, we hypothesize that carbon materials derived from wood would be appropriate to use as an electrode in cathode MFCs. In the process of carbonization, a wood carbon (WC) electrode was prepared. Specific surface area and basic element contents of the electrode were optimized at high temperature and applied in the cathode MFC. The employment of WC electrode instead of carbon cloth and carbon felt electrodes in MFC was a novel strategy to enhance Cr(VI) removal efficiency and bioelectricity harvesting simultaneously. Moreover, surface morphology, electrochemical tests, and valence change of chromium were also conducted in anticipation of using the electrode for practical application. With the inherent porous structure and excellent biocompatibility with simple chemical modifications, WC can become a new material with more features.

Electrode preparation
To synthesize the wood carbon electrodes, wood was carbonized following (Yang et al. 2021) study protocol. The naturel basswood (purchased from Chenlin Wood Company) was cut into a square (3.0 cm × 3.0 cm × 0.7 cm) with the same geometric surface area (26.4 cm 2 ) and dried at 50°C for 48 h to remove the moisture inside the material. Then, the square was calcined at 800°C under N 2 flow for 3 h with a 5°C/min heating rate. The wood retains its pore-like structure and toughness after carbonization, and its internal channels become laxer with high-temperature modification. Therefore, the titanium wire can be easily inserted directly into the wood to assemble the electrode. The electrodes were named a WC electrode connected to titanium wire. Carbon cloth (3.73 cm × 3.0 cm × 0.3 cm) and carbon felt (3.0 cm × 2.55 cm × 1.0 cm) were control cathodes with a 26.4 cm 2 geometric surface area pretreated accordingly (Bond and Lovley 2003). They were soaked in 1M of NaOH and HCl to remove impurities and linked to an external circuit by a titanium wire.

MFC fabrication and operation
A dual-chamber MFC was assembled with two glass bottles (250 mL of each) and divided by a PEM (9.6 cm 2 ). PEM was pretreated accordingly (Khan et al. 2020) and connected via a 1000-Ω external resistance. The carbon felt (3.0 cm × 2.55 cm × 1.0 cm) was used as the anode electrode and preimmobilized with Shewanella oneidensisMR-1 for 1 month to improve the electrochemical performance referring to our previous work (Shi et al. 2018). WC electrode was tested for various pH such as 5, 6, 7, 8, and 9 for optimization of pH for the further experiment as shown in Fig. S1. The anolyte was modified M 9 medium and consisted of the following: 3.0 g L −1 KH 2 PO 4 , 0.011 g L −1 CaCl 2 , 0.498 g L −1 MgSO 4 ·7H 2 O, 17.105 g L −1 Na 2 HPO 4 ·12H 2 O, 0.5 g L −1 NaCl, 4 g L −1 C 6 H 12 O 6 ·H 2 O, and 1.0 g/L g L −1 NH 4 Cl at pH 7.0. The catholyte was a modified media (3.0 g L −1 KH 2 PO 4 , 0.011 g L −1 CaCl 2 , 0.498 g L −1 MgSO 4 ·7H 2 O, 17.105 g L −1 Na 2 HPO 4 ·12H 2 O, 0.5 g L −1 NaCl, 1.0 g L −1 NH 4 Cl, and 0.1 g L −1 NaHCO 3 at pH 7.0) with a 20 g L −1 initial Cr(VI) concentration for testing the removal effect. To remove the dissolved oxygen, nitrogen was regularly employed in the cathode chamber for 30 min. A constant temperature (30°C) was used for operating batch mode MFCs. Each reactor was run in triplicate.

Analytical techniques and calculations
Data acquisition apparatus, paralleled with a 1000-Ω resistance to record the voltage output with a stable interval (10 min), was connected with a computer. The polarization curve was obtained by a variable resistance box (100,000-100 Ω) for the calculation of power density (Simeon et al. 2020). The power density (mW m −2 ) and current (mA m −2 ) were normalized to the geometric surface area (26.4 cm 2 ) of electrodes. The current (I) was obtained through Ohm's law: I = U/R, where U is the determined voltage, and R is the external resistance (1000 Ω). The electrochemical measurements were performed by using a potentiostat (CHI604E, Shanghai, China) and conclude the Tafel plot (TAFEL), AC impedance (IMP), and cyclic voltammetry (CV) in a 3-electrode arrangement. Ag/AgCl, Pt sheet, and cathode were used as the reference electrode, counter electrode, and working electrode, respectively. CV test was performed in the range of −0.8 to 0.8 V, while the IMP test was done at the open circuit potential in the frequency range from 1 × 10 5 to 0.1 Hz with a 10-mV amplitude sinusoidal perturbation. The electrochemical impedance of cathodes was then analyzed according to the Nyquist plots using the ZView software (El-Hajjaji et al. 2018). The Tafel plots (ln(j/A)~η) were recorded by the sweeping over potential at 1 mVS −2 from −40 to 48 mV. The linear portion of the Tafel plot was presented by semi-empirical Tafel equations as follows: where β a (α a nFE/RT) and β c (α c nFE/RT) were the Tafel slopes that are oxidative and reductive slopes of the linear fitting respectively. E is the applied voltage, I represents current (A), T is the temperature in Kelvin (298), F is Faraday's constant (96,485 C mole -1 ), R is the gas constant (8.314 J mol −1 K), and n is the number of electrons transferred at the rate-limiting step. To analyze the content of initial chromium and residual soluble Cr(VI), the samples of catholyte were filtered through membrane filters (0.45 μm) to remove the insoluble substances. The optical density with 1,5diphenylcarbazide at 540 nm was measured by a UV spectrophotometer (Omer et al. 2019). The removal percentage of soluble Cr(VI) can be determined accordingly: Removal effi- where C 0 and C t are initial concentrations (mg L −1 ) and time t of Cr(VI), respectively. The amount of soluble Cr(III) was calculated with a mass balance, the difference between concentration of residual soluble Cr(VI) and total soluble chromium. Furthermore, to calculate the reduction efficiency with a different electrode material, a pseudo-first-order kinetic model was used to analyze the results: where C is the Cr(VI) concentration at time t and C 0 is the initial Cr(VI) concentration, 20 mg L −1 ; k is the rate constant of the first-order model, h −1 ; t represents time, h. The k value was calculated by the slope of ln (C 0 /C) vs. time pattern, which exhibited the Cr(VI) reduction rates mediated by different electrodes. The oxidation reduction potential (ORP) amount was detected by using an ORP-pH meter (4010-3W Multilab, YSI, USA). Scanning electron microscopy equipped with energy dispersive spectroscopy (SEM-EDS, FEI Apreo, Czech) was applied to observe the morphology of the electrode surface. Samples were coated with Au particles before observation. An SZ85 digital multimeter (Suzhou Telecommunication Factory, China) was used to test the electrical conductivity of the WC electrode according to the standard four-probe technique (Guo et al. 2018). The Fourier transform infrared (FTIR) spectra were applied to analyze the functional groups of wood materials with FTIR spectrometer (NICOLET, NEXUS 670) in a wavenumber range of 4000-400 cm −1 . The specific surface area (SSA) was analyzed from N 2 adsorption-desorption experiment using a micrometric absorber (3 FLEX 3500, USA). Elemental analysis (C, N, H) of wood carbon material was performed on VarioEL Elemental Analyzer (Elementar, Germany). XPS spectra was obtained on an AXIS-ULTRA instrument (Kratos, England), and the results were fitted with software (XPS Peak 41). Elemental compositions, the specific valence of Cr and O on the surface of cathodes were obtained by the X-ray photoelectron spectroscopy (XPS) technique. All graphs and fit curves were made by Graphpad Prism 7 (Graphpad, San Diego, CA, USA).

Physiochemical characterization of developed electrode
The morphological structure of wood carbon was observed under SEM after the development. The result showed that a large number of hole was found in the cross section of the WC electrode (Fig. 1a, b). The specific surface area of WC was characterized by N 2 adsorption-desorption which was 158.47 m 2 g −1 (Fig. 1c). Furthermore, the functional group analysis of WC revealed 3 stretching vibration modes at 1093, 1622, and 3427 cm −1 respectively, after an 800°C modification (Fig.  1d). In addition, the elemental analysis showed that C and N content reaches 85.34% and 0.52% for WC in comparison with unmodified natural wood, whereas, the C and N contents were 48.97% and 0.23% respectively (Table S1).
SEM results suggested that the fold pore structure of the WC electrode increased the effective surface area for making a microenvironment to facilitate the electron transfer mechanism. The performance of wood material exceeds the carbonized chestnut shell (48.12 m 2 g −1 ) that was used in the anode MFCs previously ). Due to the stable porosity of wood by 66%, the WC electrode has a lower resistivity and increased electron transferability (Jia et al. 2017;Wan et al. 2015). Moreover, the modes of FTIR patterns were corresponding to C-O, C=O, and O-H in which the C group becomes a major component after carbonization (Mansur et al. 2008), consistent with the present results (Table S1). The increment of C content can enhance the electrical conductivity of WC, whereas N content can strengthen the electron transfer capacity and promote the electrochemical performance of electrodes (Liu et al. 2014;Wu et al. 2021). In addition, the conductivity of WC was 4.1 S cm −1 stronger than traditional carbon cloth electrode, which was only up to 2.51 S cm −1 (Geng et al. 2010). These results suggested that the modification of carbonization can increase the porous structure of WC and endow it with better electrochemical activity, which is conducive to its function as an electrode.

Cr(VI) reduction and electricity generation in MFC
Cr(VI) removal efficiency was evaluated in MFC under closed-circuit conditions. The results revealed that 100% of Cr(VI) was removed in the WC electrode equipped chamber which was higher than MFC carbon felt (34.83% ± 0.12) or carbon cloth (31.12 ± 0.31%) electrodes within 48 h of incubation in MFC (Fig. 2a). The highest k value in the WC cathode was (0.08345 ± 0.01063 h −1 , R 2 =0.84) calculated according to the pseudo-first-order kinetic equation. This result was almost 10 times higher than MFCs with carbon cloth (0.007583 ± 0.0005731 h −1 , R 2 =0.93) and carbon felt  0.007648 ± 0.0004977 h −1 , R 2 =0.95) electrodes, respectively (Fig. 2b). The amounts of total chromium were declined to 8.44 ± 0.32 mg L −1 significantly in WC cathode chamber compared to carbon felt (18.09 ± 0.03 mg L −1 ) and carbon (18.78 ± 0.01 mg L −1 ), whereas the amounts of Cr(III) were increased to 5.00 ± 0.15 mg L −1 (carbon cloth), 5.05 ± 0.02 mg L −1 (carbon felt), and 8.44 ± 0.24 mg L −1 (WC) respectively (Fig. S2). Furthermore, the influence of different electrodes on the bioelectricity generation of MFCs was also investigated. The result showed that the voltage output decreases from 3 ± 1 to 1 mV with 14 h of incubation by using the carbon cloth cathode, while carbon felt electrode in MFC increased voltage production 136 ± 2 to 227 ± 3 mV within 2 h. Similarly, WC electrode in MFC increased electricity generation to 435 ± 2 mV from 127 ± 3 mV within 8 h (Fig. 2c). The current density also showed a similar trend in WC (166.03 mA m −2 ) followed by carbon felt (85.50 mA m −2 ) and carbon cloth (1.53 mA m −2 ) at various time incubations in MFC operation (Fig. S3). The highest power density (P max ) (62.59 ± 0.27 mW m −2 ) was observed in the MFC with WC cathode higher than carbon cloth (0.115 ± 0.001 mW m −2 ) and carbon felt (3.154 ± 0.035 mW m −2 ) respectively (Fig. 2d). Moreover, the internal resistance of different electrodes was also evaluated by slopes of voltage versus current. The maximum resistance was 611.3 ± 40.1 Ω of carbon cloth, followed by 5.9 ± 0.1 Ω of carbon felt and 0.9 ± 0.01 Ω of WC electrode (Fig. 2e, f).
WC assists MFC by enhancing Cr(VI) removal and current generation under the close-circuit condition simultaneously. As shown in (Fig. 2a), Cr(VI) was completely removed at 20 mg L −1 in WC electrode MFC. In contrast, the carbon cloth electrode in MFC reduces 51.64% of Cr(VI) at 50 mg L −1 (Li and Zhou 2019). Similarly, in the MFC system, 5 mg L −1 of Cr(VI) was completely reduced (Sophia and Sai 2016), while 86% of the reduction was achieved at 8 mg L −1 of Cr (VI) (Sophia and Saikant 2016). The variation of total chromium and Cr(III) indicates that some contents of Cr are dissolved in the buffer after the reduction, while the other is probably deposited on the surface of the electrode and bottom of the reactor in the form of an insoluble precipitate. Therefore, the large specific surface area and porous structure of wood electrodes were favorable for the enhancement of electron transfer and chemical catalytic activity, which is responsible for the high Cr(VI) removal efficiency in MFC with WC cathode rather than traditional carbon electrode. The maximum bioelectricity generation in the present study was 435 mV (Fig.  2c) higher than the previously reported result for the traditional MFC with graphene modified graphite felt cathode (411 ± 12 mV), where the Cr(VI) was used as an electron acceptor . Moreover, the maximum power density in our study was also compared with other previous studies which use different cathode materials. The graphite plate electrode used in MFC produces maximum power density of 55.5 mW m −2 by using Trichococcus pasteurii and Pseudomonas aeruginosa as a biocatalyst (Tandukar et al. 2009). The HNO 3 -NaX modified graphite felt cathode with 28.90 ± 3.18 mW m −2 , the pure graphite block cathode (45.9 mW m −2 ) (Pandit et al. 2011), and liquid crystal polaroid glass cathode (10 mW m −2 ) (Gangadharan and Nambi 2015). These previously reported data were lower than the present maximum power density (62.59 ± 0.27 mW m −2 ) obtained in MFC equipped with WC cathode (Fig. 2d). The electrodes with smaller resistance enhance electron transfer and increase power output. The data indicates that the carbonized natural wood electrode enhances Cr(VI) reduction and bioelectricity generation in the MFC.

Electrochemical characteristics of electrode
In order to understand the electrochemical processes occurring on the electrode surface, CV, Tafel plot, and Nyquist plot tests were used to determine different characterizations mediated by various electrodes (Fig. 3). The CV patterns indicated the peak of reduction at 0.235 V, 0.401 V, and 0.443 V (vs Ag/ AgCl) for carbon cloth, carbon felt, and WC electrode, respectively (Fig. 3a). Tafel plots have exhibited the exchange current densities between the electrodes and electron acceptor (Cr(VI)) (Fig. 3b). The results showed that the values of the reductive Tafel slope were 277.3, 139.8, and 110.5 V dec −1 respectively, for the carbon cloth, carbon felt, and WC electrodes (Fig. 3c). EIS was performed to analyze the impedance (charge transfer resistance (R ct ) & solution resistance (R s )) of three different cathode-based MFC (Fig. 3d). The values of impedance were calculated by using Nyquist plots. All cathodes have similar R s values as follows: 0.72 ± 0.01 Ω (carbon cloth), 0.41 ± 0.01 Ω (carbon felt), and 0.28 ± 0.02 Ω (WC). WC electrode showed significantly low charge transfer resistance (0.42 ± 0.01 Ω) compared with carbon cloth (43.77 ± 7.67 Ω) and carbon felt (201.3 ± 31.9 Ω) electrodes (Fig. 3e).
In the MFC operating system, CV analyzes the electrochemical ability of the electrode surface. The positions of loop peaks can exhibit the redox potential mediated by Cr(VI) reduction, and the size of peaks can illustrate the electrochemical activity of electrode (Yang et al. 2016;Zhao et al. 2018). The oxidationreduction loop areas of these electrodes were also different from each other, which can reveal the electron transfer capacity best at the WC electrode surface. The electro-active ability of fuel cells could be defined by stoichiometry (Tafel plots), where the ability is inversely proportional to slope values (Limaye et al. 2021). The results indicated that WC has a stronger electrocatalytic ability than traditional carbon electrodes. The internal resistance of electrodes has limitations for bioelectricity generation in MFC . The low resistance of the wood electrode indicates better and faster charge transferability on the surface. A significant reduction of R ct was attributed to the porous structure of wood materials, which improved the interfacial interaction between the electron acceptor and electrode surface. Besides, the lower charge transfer resistance is more favorable for the electron's transportation, which is vital for the electrode conductivity (conductivity is inversely related to impedance; σ=(1/R t )(1/K)) (Densakulprasert et al. 2005;He et al. 2020). Therefore, the stronger conductivity for the WC electrode was further confirmed. These results proved that a larger surface area of the WC cathode has an advanced strong absorption ability to reduce Cr(VI). Possible mechanism of Cr(VI) reduction by wood carbon Cr(VI) reduction was examined in the MFC system by using the wood electrode. Therefore, XPS technique was used to analyze the composition of cathode surface and the valence of elements (Fig. 4). No apparent signal of elemental chromium was observed in the carbon cloth cathode, which might be due to the limited adsorption capacity of materials for Cr(VI). However, two characteristic peaks of chromium appeared on the WC cathode at the binding energy of 574.3 and 584.7 eV (Fig. 4c, d), which indicated that chromium was deposited on the surface. One characteristic peak of oxygen appeared on the WC at the binding energy of 529.2 eV (Fig. 4e, f). The ORP for the carbon cloth, carbon felt, and WC was (+) 180 mV, (+) 123 mV, and (+) 31 mV, respectively in the MFC system (Fig.  S4). This indicates that WC allows more electrons to transfer Cr(VI) rapidly and its reduction to Cr(III) products.
The binding energy of chromium peaks was corresponding to Cr(III) (Cr 2p 3/2 and Cr 2p 1/2 ) for WC cathode, which was compared with the referenced Lookup Table for Signals from Elements and Common Chemical Species (Crist and Crisst 2000;Song et al. 2016). This result indicated that the Cr(VI) absorption of wood electrodes was stronger than conventional carbon cloth electrode, and Cr(VI) can be reduced to some Cr(III) precipitated substances (Cr 2 O 3 or Cr(OH) 3 ). In addition, the binding energy peak of oxygen (O1S) corresponded to the O1 peak, which represented the metal-O band (Cheng et al. 2019). The initial ORP is about (+) 240 mV, which is similar to the ORP required for the reduction of Cr(VI) reported in the previous study (Gangadharan and Nambi 2015). The difference between the ORP values for the WC-mediated MFC in whole run was much larger than the results of other groups (Kumari and Dutta 2020), which suggested that WC allows more electrons to transfer to Cr(VI). In addition, the decrease in pH due to the Cr(VI)mediated reduction is also associated with the potential Cr(III) product formation (Matsena et al. 2021). In addition to the formation of Cr(III) precipitate obtained by mass balance extrapolation, the final pH value greater than 6 is fit with the product distribution pattern of Cr(OH) 3 (Xafenias et al. 2014).
Therefore, these results proved that WC electrode absorbs chromium ions, reduces them to metal oxide product Cr 2 O 3 , and then deposits them on the surface of the wood (Eq. (5)).
The cost of the WC electrode was evaluated with previous electrodes used in MFC for metal reduction. The preparation cost of it is about 0.00085 $/piece, far lower than the traditional graphene felt (0.114 $/piece) with the same specific surface area (Table S2) (Guo et al. 2016;Li et al. 2018). This study demonstrated the potential application of WC in MFC, while some practical operational parameters are still lacking. It needs to be further investigated and explored in the future.

Conclusions
MFC with WC electrode completely removed Cr(VI) and produced 62.59 ± 0.27 mW m −2 of power density which was higher than traditional carbon felt cathode. The proposed mechanism of the high Cr(VI) removal efficiency and power output were ascribed to the advanced electrocatalytic effect and microstructure with partially aligned and irregular channels after a high temperature (800°C) modification. It has been recommended that mesoporous, three-dimensional WC electrodes could be utilized for efficient Cr(VI) removal and bioelectricity generation in dual-chamber MFC by treating wastewater.
Funding This research work was sponsored by the National Natural Science Foundation of China funds (21876072) and (31870082).
Data availability All the data obtained in this study are presented, and specific results and element determination data can be obtained from the author.

Declarations
Ethics approval and consent to participate Not applicable.
Consent for publication Not applicable.