Enhancing extracellular electron transfer through selective enrichment of Geobacter with Fe@CN-modified carbon-based anode in microbial fuel cells

Microbial fuel cells (MFCs) have been demonstrated as a renewable energy strategy to efficiently recover chemical energy stored in wastewater into clean electricity, yet the limited power density limits their practical application. Here, Fe-doped carbon and nitrogen (Fe@CN) nanoparticles were synthesized by a direct pyrolysis process, which was further decorated to fabricate Fe@CN carbon paper anode. The modified Fe@CN anode with a higher electrochemically active surface area was not only benefit for the adhesion of electrochemically active microorganisms (EAMs) and extracellular electron transfer (EET) between the anode and EAMs but also selectively enriched Geobacter, a typical EAMs species. Accordingly, the MFCs with Fe@CN anode successfully achieved a highest voltage output of 792.76 mV and a prolonged stable voltage output of 300 h based on the mixed culture feeding with acetate. Most importantly, the electroactive biofilms on Fe@CN anode achieved more content ratio of proteins to polysaccharides (1.40) in extracellular polymeric substances for the balance between EET and cell protection under a harsh environment. This work demonstrated the feasibility of development on anode catalysts for the elaboration of the catalytic principle about interface modification, which may contribute to the practical application of MFC in energy generation and wastewater treatment.


Introduction
In the past decades, over-exploitation and over-consumption of fossil fuels have created serious threats for human being, such as global warming, environment pollution, and energy shortage. Under this circumstance, bioelectrochemical systems (BESs) have shed light on the research and development of renewable energy sources due to their versatility for the degradation of organic pollutant and production of green-energy simultaneous by the metabolism of exoelectrogens (such as Geobacter and Shewanella) (Chai et al. 2020;Jung 2012;Son et al. 2021a). Some various types of BESs (e.g., microbial electrolysis cells, microbial electrolysis cell, and microbial fuel cells) have given the potentially widespread application (Amrut Pawar et al. 2020;Kang et al. 2017b;Karthic et al. 2020;Son et al. 2021b;Zahid et al. 2022). As typical BESs, however, previous researchers have demonstrated that the high internal resistance and low power density were the obstacles for distinguishing the performance difference of microbial fuel cells (MFCs) (Gurung et al. 2012;Jung and Regan 2007;Koo et al. 2019;Song et al. 2016). It is also well known that the performance of MFCs relies heavily on extracellular electron transfer (EET) between the exoelectrogens and electron acceptors. When the electron acceptors (such as abiotic solid electrode) cannot enter the interior of the cells through cell membrane, Responsible Editor: George Z. Kyzas the exoelectrogens transfer the generated electrons to the conductive proteins embedded in outer-membrane through respiratory chain and finally to electrodes through the direct (external c-type cytochromes or flagella) and/or indirect pathways (electron mediators) (Hu et al. 2019;Liu et al. 2020a).
Commonly, exoelectrogens in the anode chambers tend to aggregate on anodes and evolve into electroactive biofilms (EABs), which are the core component of the power performance of MFCs (Pandit et al. , 2021Savla et al. 2020;Wu et al. 2020;Yong et al. 2014).
Undoubtedly, the formation process of EABs and the EET rate between exoelectrogens and anode was proved to be affected by the anode properties, such as the interfacial area (Sreelekshmy et al. 2020), hydrophobicity , and biocompatibility (Li et al. 2019b). Thus, a great deal of studies was conducted to explore the advanced anode materials for enhancing the performance of MFCs. For example, different porous biomass materials (corn stems, commercial bread, and loofah sponges) were used as precursors to prepare three-dimensional (3D) electrodes with good conductivity (Mei et al. 2020). As a result, the macroporous structure in the carbonized electrodes was preferable for improving the electrochemically active surface area (EASA) and forming EABs compared with that of plain carbon felt electrode. Furthermore, a 3D microporous nitrogen-enriched graphitic carbon electrode was fabricated through facile pyrolysis and embedded in MFCs, which achieved a highest power density of 750 mW/m 2 (2.07-fold higher than that of control group) (You et al. 2017). These works demonstrated that the types and properties of anodes heavily determined the electron-producing capacity of EABs and EET rate accordingly, thus further affecting the bioelectrochemical performance of MFCs.
To date, considering the cost of all BESs and experiment processes, the main types of anodes used in MFCs are carbon-based materials, such as carbon felt (Li et al. 2021a), carbon cloth (Li et al. 2021b), carbon fiber brusher (Hari et al. 2016;Kang et al. 2017a;Nam et al. 2020), and graphite plate (Yang et al. 2019). Nevertheless, the abovementioned anodes does not show extra breaks in the performances of MFCs, particularly since their hydrophobicity and low surface area are unsuitable for bacteria adhesion and high internal resistance unfavorable for EET .
It should be noted that the performance of anode could be enhanced by using better catalyst, optimizing structure, adjusting configuration, and selecting the appropriate counter electrode. Among these, the modification and upgradation of electrodes using better catalyst is an attractive and effective strategy to strengthen anodic biocompatibility, enhance EET rate, and improve performance of MFCs (Jung et al. 2018;Koo and Jung 2021;Li et al. 2020a;Nam et al. 2017;Yang et al. 2020b). Recently, conductive substrates , functional groups , and transition metal carbides and nitrides (Yang et al. 2020b) were consecutively employed to modify the anodes by surface modification. Researchers have found that transition metal-doped carbon and/or nitrogen can significantly facilitate the EET process of exoelectrogens. Song et al. synthesized a 3D anode through combination of carbon nanotubes and Fe 3 O 4 nanospheres into microporous graphene foams to fabricate the MFC system with a higher power output of 782 A/m 3 (11-fold higher than that of control group) and a stronger stability (Song et al. 2016). The high affinity of Fe 3 O 4 nanospheres toward exoelectrogens and the smoother EET pathway paved by microporous graphene foams were mainly responsible for the high-performance of MFCs compared with that of graphite rod anode. Wang et al. also reported that the FeS 2 nanoparticles decorated graphene anode not only enhanced the adhesion of exoelectrogens but also shorten the start-up of MFCs by enhancing the EET process . Moreover, Wu et al. reported that the N-doped carbon nanotubes/reduced graphene oxide (rGO) promoted the formation of EABs on anode and further achieved a maximum power density of 1137 mW/m 2 (Wu et al. 2018). In a word, the enhanced performances by the modification of anodes could be attributed to the following reasons: (1) the 3D structure and porosity were beneficial for exoelectrogen habitation and EAB formation (Song et al. 2016;You et al. 2017); (2) the enrichment of exoelectrogens produced more exoelectrons and achieve higher voltage output (Li et al. 2021c); and (3) the lower charge transfer resistance avoided the loss of electron in EET and improved the coulomb efficiency of MFCs (Yang et al. 2021;Yu et al. 2018).
Although most works about electrode modification with metal or metal oxide integrated with some conductive materials were reported (Wang et al. 2022), few studies explored the effect of co-decoration of anodes with metal (Fe, Mn, Co, and so on) and non-metal (C, S, and N) on the power performance of BESs. In this context, the present study highlights the possibility of a Fe@CN nanosphere catalyst coating on the carbon paper anode for achieving a high bioelectrochemical performance of MFCs, and the reasons for the enhanced performance were also elaborated from both biological and electrochemical perspectives. As shown in Scheme 1, Fe@CN nanospheres were synthesized through one-pot pyrolysis and further coated on the surface of carbon paper. The obtained anode had porous structure and achieved a maximum voltage output of 792.76 mV, which could be attributed to the enrichment of exoelectrogens (Geobacter) and the variation of EET pathway. This work shed new insights on fabricating efficient anodes for selectively enriching exoelectrogens and efficiently enhancing the performance of MFCs.

One-pot method for the synthesis of Fe@CN nanospheres
The Fe@CN nanospheres were synthesized by one-pot pyrolysis of Fe (III) and carbon-nitrogen precursors. In detail, 0.5 g of FeCl 3 and 3 g of urea were dispersed in ultrapure water (10 mL), which was stirred for 20 min and then sonicated under 30 °C for 60 min to obtain a homogenous solution. Next, the suspension was placed into a muffle furnace and annealed at 550 °C for 2 h with a heating rate of 10 °C min −1 . The samples were denoted as Fe@CN1. Accordingly, Fe@CN0 (pure FeCl 3 without urea) was prepared in a similar synthesis process.

Characterization of catalyst
The microstructure detection of Fe@CN nanospheres was conducted by scanning electron microscope (SEM, Regulus-8100) coupled with an energy dispersive spectrometer (EDS). X-ray diffraction (XRD, D8 Advance) patterns of the catalyst were obtained with 2θ range of 5-90°. The Raman images were obtained using Raman spectroscopy (RAM ΙΙ). The X-ray photoelectron spectroscopy (XPS, AXIS) was employed to measure the chemical composition and atomic valence of the synthesized catalysts.

Electrode fabrication and MFC setup
Carbon paper was used as plain template for coating the prepared Fe@CN catalyst. In detail, 5 mg of powdered catalyst was dispersed in 20 μL ultrapure water, which was then mixed with 10 μL Nafion solution to prepare catalyst ink. The surface of a carbon paper (2.0 × 1.0 cm) was coated with the abovementioned catalyst ink and dried under 60 °C for 2 h. The modified electrode was assembled into MFCs as the anode in the following studies.
The typical double-chamber MFC device with 50 mL of each identical volume, separated by a proton exchange membrane (DuPont, USA), was constructed as our previous reports . The anodes were the control carbon paper electrode or the above prepared Fe@CN electrode. The cathode utilized in this work was carbon paper electrode. The anode and cathode were connected with 1,000 Ω external resistance to form a closed circuit. The distance between the two electrodes was about 8 cm. The anolyte was consisted of wastewater (sampled from Nanjing wastewater treatment plant, Nanjing, China) and phosphate-buffered solution (v/v: 1/4): NaH 2 PO 4 , 2.54 g/L; Na 2 HPO 4 , 4.09 g/L; NH 4 Cl, 0.31 g/L; and KCl, 0.13 g/L, along with mineral and vitamin solutions (Li et al. 2019a). Sodium acetate (1.28 g/L) was added into anode chamber as electron donors, and the catholyte was 50 mM potassium ferricyanide (K 3 [Fe(CN) 6 ]) with phosphate buffer solution. During the running process, MFCs were accessed to the data collectors to record the voltage output, and the temperature in all process was maintained at 30 °C. All experiments were set up in triplicate.

Characterization and analytical measurements
The biomass density of EABs was measured using BCA Protein Assay Kit (Li et al. 2021a). For the micromorphology detection of EABs, electrodes with EABs were extracted from MFCs and subsequently cut into small pieces. After being fixed in 2.5% glutaraldehyde for 2 h, the biofilm pieces were dehydrated using a graded ethanol series (25, 50, 70, 80, 90, and 100%). Finally, the samples were dried at room temperature for 12 h and analyzed through scanning electron microscopy (SEM) (Li et al. 2021a). Extracellular polymeric substance (EPS) fractions from the EABs were extracted following the previous report (Li et al. 2019a). Polysaccharide and protein from the EPS were determined by the sulfuric acid-phenol method and BCA Protein Assay Kit, respectively .

Scheme 1 Pyrolytic synthesis of Fe@CN nanospheres
The electrochemically active surface area (EASA) was obtained through cyclic voltammograms (CV) from -0.4 to 0.7 V with a scan rate of 10 mV/s in the electrolyte with 0.1 M Na 2 SO 4 and 10 mM K 3 [Fe(CN) 6 ] ). The electrochemical active surface area (EASA) of prepared electrode was calculated using Matsuda' equation: where Ip was the peak current; n = 1 was the number of the transferred electrons; F = 96,487 C/mol was the Faraday' constant; R = 8.314 J/(mol·K) was the gas constant; T = 303 K was the temperature; A represented the EASA of the prepared electrode; D = 5.79*10 −6 cm 2 /s was the diffusion coefficient of the K 3 [Fe(CN] 6 solution; C 0 = 10 mM was the concentration of the K 3 [Fe(CN) 6 ; and v = 10 mV/s was the scan rate.
The polarization curves were conducted when the MFCs were in steady state with a scan rate of 10 mV/s from -0.8 to 0 V. The cyclic voltammograms (CV) and electrochemical impedance spectroscopy (EIS) experiments were measured on a CHI 660 electrochemical workstation with a three-electrode configuration, including a reference electrode (Ag/AgCl electrode), a working electrode (the EAB anodes), and a counter electrode (carbon paper electrode). The CV curves were recorded from − 0.8 to 0.8 V at a scan rate of 10 mV/s. The EIS experiments were carried out from 10 mHz to 100 kHz at OCP in the MFCs.

Characterization of Fe@CN catalysts
The dispersed iron anchored C-N supporter was prepared by one-pot pyrolysis of FeCl 3 and urea (Scheme 1), which was named as Fe@CN catalysts.
SEM images ( Fig. 1) revealed that these Fe@CN samples obtained at 550℃ from the precursor with different mass ratio of FeCl 3 /urea showed significant different surface morphology. It was noteworthy that Fe@CN0 exhibited an irregular geometric configuration consisting of a smooth sphere and blocky structure, which resulted from the interaction of iron and oxygen during pyrolysis. The corresponding elemental mapping images demonstrated that C, N, O, and Fe elements relatively uniformly distributed Fe@CN1 sample ( Fig. 1c-f), which implied that Fe was successfully doped with C and N element.
XRD patterns of the prepared Fe@CNx catalysts are shown in Fig  . The characteristic diffraction peaks in Fe@CN0 disappeared when Fe doped into urea to form Fe@CN1, which suggested the effective recombination of carbon and nitrogen structure. Moreover, Raman spectra were recorded to further reveal the defect sites and disordered structure of Fe@CN catalysts. The D-band was induced by the vacancies and heteroatom doping while the G-band was related to the in-plane stretching vibration of sp2 hybrid atoms (Yang et al. 2020a). As shown in Fig. S2b, the ratio of I D /I G in Fe@CN1 catalyst was 1.23, indicating that the existence of more defect sites .
The surface chemical composition and valance states of these catalysts were also scrutinized by X-ray photoelectron spectroscopy (XPS). The C1s spectra of Fe@CN1 revealed the presence of C-C (284.78 eV), C-N (286.48 eV), and N-C = N (289.18 eV) functional groups on the surface of Fe@CN1 (Fig. 2a). Moreover, the deconvoluted N1s spectrum of Fe@CN1 exhibited two peaks at 400.38 and 403.38 eV, which were attributed to pyrrolic N and graphitic  (Fig. 2b). A previous study reported that graphitic N could enhance the material biocompatibility and reduced the start-up time of MFCs Zhao et al. 2019). The pyrrolic N was responsible for improving the power density of MFC systems (Liu et al. 2014. The deconvoluted Fe2p spectra of Fe@CN1 is presented in Fig. 2c, which was assigned to Fe 2+/3+ 2p 1/2 , Fe 2+/3+ 2p 3/2 , and their satellites. In details, the peaks at 711.87, 714.52, and 718.90 eV in the Fe2p spectrum were assigned to Fe 2+ 2p 3/2 , Fe 3+ 2p 3/2 , and their satellites, while Fe 2+ 2p 1/2 , Fe 3+ 2p 1/2 , and their satellite peaks appeared at 725.42, 729.38, and 732.90 eV, respectively. Thus, the transition of solid-state redox couples Fe 3+ /Fe 2+ was expected to be beneficial for electron transfer (Li et al. 2015;Rethinasabapathy et al. 2021).

Electrochemical active surface area of different anodes
To investigate the electrochemical performance of the modified electrode, the EASA test was conducted in a three-electrode system with prepared Fe@CN-modified electrodes, titanium wire, and Ag/AgCl electrode as working electrode, counter electrode, and reference electrode, respectively. Previous studies proved that EASA represented the actual electroactive area of electrode and a higher EASA facilitated the adhesion of electrochemical active bacteria and benefited the bioelectrochemical performance of MFCs ultimately (Chen et al. 2020;Xu et al. 2022).
Additionally, similar current curves with a pair of reversible redox peak are displayed in Fig. S3, which indicated that the chemical reduction reaction was not changed by the different anodes. The catalytic curve of the blank electrode (carbon paper without modification) delivered an Ip value of 0.98 mA with a corresponding EASA of 1.53 cm 2 . To be more specific, the highest current value was increased to 1.14 mA with lager EASA of 1.78 cm 2 (the anode modified with Fe@CN0) and further increased to 2.35 cm 2 (the anode modified with Fe@CN1). In short, these results also indicated that Fe@CNx catalysts were successfully coated onto the surface of carbon paper, which provided more electrochemical active sites for interfacial electron transfer between anodes and exoelectrogens (Chen et al. 2020).

Bioelectrochemical performances of MFCs with Fe@ CN-modified anode
The MFCs equipped with the different anodes were inoculated with wastewater as bacteria inoculum to launch the voltage output. The power performance of all the launched MFCs showed a similar trend in the initial 4 running cycles, which increased steadily after being replaced with fresh anolyte (Fig. 3). In particular, the voltage output of MFC-CP (unmodified carbon paper anode) increased and achieved the maximum of 616.39 mV at about 100 h, which indicated that the exoelectrogens evolved to form matured EABs and started to produce extracellular electrons. In contrast, MFC-Fe-CK (anode modified with Fe@CN0) and MFC-Fe@CN (anode modified with Fe@ CN1) achieved higher voltage outputs of 711.89 mV and 792.76 mV without the long start-up periods, respectively (especially in MFC-Fe@CN). Furthermore, the voltage output of MFC-CP declined quickly after reaching the maximum voltage. The voltage output process represented the formation of EABs process. When the voltage output increased, the more electrons were recovered from electron donors through exoelectrogens and transferred into anodes. What is more, exoelectrogens aggregated on the anodes and evolved into EABs in the first cycle. This process was involved in the apoptosis and regeneration of cells, which was a time-cost and dynamic process. We found that the MFC with Fe@CN anode achieved high voltage output at the initial 2 h, which was greatly shorter than that of CK. In addition, the voltage outputs of all MFCs, including the maximum voltage output and the voltage-maintained time, were not influenced by the refreshed anolyte, which indicated that the matured EABs were formed in the first cycle.
The maximum voltage output of MFC-CK with the Fe@CN2 modified anode, however, was not enhanced (573.48 mV), and the start-up time (about 100 h) was also not accelerated compared with that of MFC-CP, which indicated that the increased urea proportion was not the primary cause for the enhanced electrochemical performance of anodes (Fig. S5). Compared with other MFCs, the maximum voltage output of MFC-Fe@CN could be resulted from its large EASA, which was also reported in previous studies. You et al. prepared a 3D microporous nitrogen-enriched graphic carbon electrode for efficient bioelectricity generation in MFCs (You et al. 2017), which showed a higher EASA (109.01 cm 2 ) with higher voltage output (508.00 mV) compared with that of carbon cloth anodes (EASA of 31.23 cm 2 with voltage output of 395.00 mV).
Interestedly, on the other hand, MFC-Fe@CN possessed a stable voltage output during each period. In details, the voltage outputs of MFC-Fe@CN were 706.79, 718.14, 722.42, and 682.54 mV at the end of the 1st to 4th cycle, respectively, while the voltage outputs of MFC-CP decreased to about 0 mV in each cycle (0.11, 5.82, and 26.19, 4.87 mV for the end of the 1st to 4th cycle, respectively). The long stable voltage output was the precondition for the real application of MFCs, which potentially indicated the superiority of the prepared anodes in the present study.
To further investigate the electrochemical activity of EABs on different modified anodes, CV measurements were employed at the scan rates from 10 to 200 mV/s. The CV spectrum (Fig. 4) showed that various redox peaks appeared in MFCs with different anodes, which suggested the EET pathways of EABs could be altered by the modified anodes. Particularly, a pair of redox peaks was found in MFC-CP, with the oxidation peak potential at − 0.31 V and a peak current of 0.23 mA. The CV spectra of the EABs in MFC-Fe-CK and MFC-Fe@CN showed the similar trend, respectively, which implied that the EET pathways could be same in these two MFCs. In addition, the peak at − 0.12 V almost disappeared in MFC-Fe-CK, which, in contrast, possessed 0.45 mA of catalytic current in MFC-Fe@CN. Moreover, the other peaks appeared similar redox potential (0.32 V in MFC-Fe-CK, 0.30 V in MFC-Fe@CN) with the different catalytic current (0.58 mA in MFC-Fe-CK, 0.97 mA in MFC-Fe@CN) indicated that the EABs in these two MFCs had significant different catalytic activities. The abovementioned results demonstrated that the Fe@CNmodified anodes increased electrochemical activity of EABs, which were in good accordance with the voltage outputs (Fig. 3). When the scan rates increased from 10 to 200 mV/s, the peak current also increased consequently. Interestingly, the peak currents of all MFCs exhibited a good line relation with the scan rates, which suggested that the electrochemical reactions on all anodes were the typical surface-controlled processes (Li et al. 2020b).
The polarization and power density curves of the MFCs with different modified anodes were also tested with LSV. The maximum power density of MFC-Fe@CN was 0.434 W/m 2 , which was 13.91%, and 183.66% higher than that of MFC-Fe-CK (0.381 W/m 2 ) and MFC-CP (0.153 W/ m 2 ), respectively (Fig. 5b). The similar open-circuit voltages were exhibited in MFC-Fe-CK (668.18 mV) and MFC-Fe@CN (655.98 mV), which were also 39.49% and 34.55% higher than that of MFC-CP (487.52 mV), respectively. These results indicated that the Fe@CN-modified carbon paper anodes were more suitable for the attachment of exoelectrogens to form robust EABs for power generation .
EIS experiment was also performed on all MFCs to identify the electron transfer resistance between exoelectrogens and anode (Fig. 5b)  . The similar solution resistance (Rs) was obtained (Table. S1), which indicated that the electrodes modified with Fe@ CN had no impact on Rs. Apparently, the values of charge transfer resistance (Rct) in MFC-Fe-CK (29.59 Ω) and MFC-Fe@CN (26.54 Ω) were greatly lower than those of Rct in MFC-CP (36.24 Ω), suggesting that the modified anode decreased the Rct of EABs. As discussed in our recent report, lower Rct could result in a faster extracellular electron transfer rate . These results implied that Fe@CN-modified anodes in MFC-Fe-CK and MFC-Fe@CN possessed the higher conductivity, which was beneficial for more efficient EET between the anodes and EABs.

The structure and morphology of EABs on the anodes modified with Fe@CN
As previously reported, the performance of MFCs was related closely to the architecture of EABs (V. H. Tran et al. 2022). After over 700 h of operation, stable voltage outputs were achieved in all MFCs, indicating that exoelectrogens successfully attached on the surface of modified anodes to form measurable EABs. As shown in Fig. 6, only a small part of anode was coated with exoelectrogens in MFC-CP, on which the cells' adhesion was limited and loose due to the lower specific surface area of the smooth surface of carbon paper. On the other hand, porous and sponge-like structure was obtained after modifying with Fe@CN catalysts, which was more favorable for attracting exoelectrogens from wastewater in the first cycle of voltage output. Therefore, the shorter start-up time was presented in MFC-Fe@CN with the faster maturation of EABs on anodes. In addition, the three-dimensional structure of anodes in MFC-Fe-CK and MFC-Fe@CN could capture and transfer electrons from the outer-membrane of exoelectrogens in multiple directions. The SEM results were further supported by the biomass analysis in Fig. S4a, which showed that the content of biomass in MFC-CP (4.08 mg/cm 2 ) increased to 5.95 mg/cm 2 (MFC-Fe-CK) and 7.27 mg/cm 2 (MFC-Fe@CN), respectively.

The modified anodes modulated the EPS production of EABs
As an integral surrounding component of cells and/or biofilms, EPS spread widely on the surface of microorganism. It is well known that polysaccharides and proteins (the two main components) and other substrates (e.g., c-type cytochromes, nanowires, and humics) were responsible for the heterogeneity of EPS . Polysaccharides were proved to protect cells against unfavorable environment, and proteins were responsible for EET through the thick layer of EPS Li et al. 2021a). During the colonization and maturation of EABs on anodes, the mixed-culture trended to produce some secretions and further to form EPS to package the biofilm from outer environment. In this study, we found that the content of proteins (0.97, 1.31, and 1.57 mg/cm 2 for MFC-CP, MFC-Fe-CK, and MFC-Fe@CN, respectively) in EPS of EABs was apparently higher than that of polysaccharides (0.85, 1.03, and 1.12 mg/cm 2 for MFC-CP, MFC-Fe-CK, and MFC-Fe@CN, respectively) among all MFCs (Fig. S4b). A previous study showed that intracellular electrons generated from the exoelectrogens were hard to be transferred to the electrode due to the thicker EABs (over tens of microns) . Nevertheless, proteins inlaid in EPS were considered to contain multiple redox molecules to overcome the insulativity of EABs (mainly was polysaccharides), which significantly accelerated EET together with other redox-active substrates (e.g., c-type cytochromes) (Li et al. 2021a;. In this case, a higher protein content in EPS of the EABs on Fe@CN-modified anodes was responsible for the enhanced EET efficiency, which also benefited the growth of exoelectrogens in MFC systems with insoluble electron acceptors. In this work, to be more specific, the maximum content of protein in the EPS was 1.57 mg/cm 2 in MFC-Fe@ CN compared with that of MFC-Fe-CK (1.31 mg/cm 2 ) and MFC-CP (0.97 mg/cm 2 ). On the other hand, polysaccharides played an important role to form the carbon skeleton of biofilms, although it was reported to be nonconductive in some studies (Yang et al. 2019). The similar content of polysaccharides was found in MFC-Fe@CN (1.12 mg/cm 2 ) and MFC-Fe-CK (1.03 mg/cm 2 ), which were significantly higher than that of MFC-CP (0.85 mg/cm 2 ). Moreover, it was reported that the content of extracellular polysaccharides was often improved with the increasing environmental stress. Here, the Fe@CN nanomaterials might induce slight stress to the mixed-culture and active the production of polysaccharides. Apparently, the higher content ratio of proteins/ polysaccharides occurred in MFC-Fe-CK (1.27) and MFC-Fe@CN (1.40) compared with that of MFC-CP (1.14). The high content of proteins in EPS compensated the possible electron transfer resistance caused by the poor conductivity of polysaccharides. The Fe@CN-modified anodes improved the electrochemical activity of EPS by adjusting the content of EPS and the ratio between different components and thus enhanced the EET efficiency of EABs. Overall, the production of EPS on different anodes would be finely regulated by cells to keep balance between EET efficiency and selfprotection against environmental stress (Li et al. 2021a).

Microbial community analysis of EABs on Fe@ CN-modified anodes
When achieving a stable voltage output, high-throughput sequencing was performed to investigate the microbial community structure of EABs on different Fe@CN-modified anodes. The dominant phyla of three EABs were all Proteobacteria, Bacteroidetes, and Firmicutes (Fig. 7). Proteobacteria were mainly Gram-negative and considered to be the most well-known exoelectrogens, Bacteroidetes were responsible for the electron generation, and Firmicutes were related to the fermentation and/or degradation of complex organic substrates (Li et al. 2019a(Li et al. , 2021b. It was noted that Proteobacteria had the highest proportion with the abundances of 53.80%, 80.10%, and 55.16% for MFC-CP, MFC-Fe-CK, and MFC-Fe@CN, respectively. The relative abundances of Bacteroidetes decreased from 38.26% (MFC-CP) to 5.60% (MFC-Fe-CK) and 24.37% (MFC-Fe@CN), and Firmicutes increased from 3.65% (MFC-CP) to 11.40% (MFC-Fe-CK) and 11.48% (MFC-Fe@CN), respectively.
These results implied that the microbial community of EABs changed tremendously after the modification with Fe@CN nanocatalysts on carbon paper. Additionally, at the genus level, an obvious difference on the microbial composition of all EABs was obtained, which suggested that the enrichment of different functional microbes was regulated by external environment . Detailedly, in MFC-CP, the microbial population was dominated by Dysgonomonas and Comamonas, which together accounted for 53.58% of the total microbial composition. These bacteria were considered to be the efficient exoelectrogens with higher power generation capacities in previous reports (Kong et al. 2022;You et al. 2021). However, the proportion of these bacteria decreased dramatically in MFC-Fe-CK (Dysgonomonas: 33.68% to 2.48%; Comamonas: 19.90% to 1.35%) and MFC-Fe@-CN (Dysgonomonas: 33.68% to 3.00%; Comamonas:19.90% to 4.40%). At the same time, Geobacter was enriched to the proportion of 10.49% and 15.41% in MFC-Fe-CK and MFC-Fe@CN, respectively. The reasons for the selective enrichment of Geobacter species in MFC-Fe-CK and MFC-Fe@ CN were in multiple fronts. On the one hand, Geobacter was well-known as one of the most typical exoelectrogens in most studies, which had high capability of direct EET pathway via nanowire and c-type cytochromes (Ueki 2021). The two EET pathways strongly depended on the physical connection between the surface of both cells/nanowire and anode. Interestingly, the sponge-like and porous structure of the modified anodes was beneficial for the contact with Geobacter, which could be one prerequisite for the accelerated start-up time of MFC-Fe@CN. On the other hand, a higher EASA of the modified anodes provided more electroactive sites for exoelectrogens. Figure 7b shows that Comamonas and Dysgonomonas enriched in MFC-CP with more abundance while Geobacter reached a low proportion (2.54%), which demonstrated the selective enrichment of exoelectrogens on anode was largely changed by Fe@CN nanospheres. In addition, the shift EET pathway in MFC-CP, MFC-Fe-CK, and MFC-Fe@CN implied the varied main exoelectrogens responsible for electron generation, which was consistent with the dynamics of microbial community. Zhu et al. demonstrated that the relative abundance of Geobacter increased from 10.00% to 33.96% when carbon cloth anode was modified with heteroatom-doped (N, P, S, and Co) porous carbon nanoparticles, while Acinetobacter (also reported as an exoelectrogens) decreased from 28.24% to 9.00% (Zhu et al. 2022). And thus, the higher power density (1.72 W/m 2 ) and current density (4.52 A/m 2 ) were achieved with the modified anode compared with that of plain carbon cloth anode (0.95 W/m 2 of power density and 3.12 A/ m 2 of current density). The above results also revealed that the evolution of EABs in the composition and proportion of exoelectrogens varied greatly on different anodes with different structures and properties, which may be attributed to the variable responses of different microbial species to the Fe@CN-modified anodes.

Conclusions
In summary, the novel Fe-doped carbon and nitrogen (Fe@ CN) catalyst had been successfully synthesized by one-pot pyrolysis and decorated on carbon paper as MFC anode. The Fe@CN nanospheres significantly improved the electrochemically active surface area of carbon paper, which was favorable for the adhesion of electrochemical active bacteria. MFC-Fe@CN achieved a maximum voltage output of 792.76 mV, which was 28.61% higher than that of MFC-CP (616.39 mV). Moreover, MFC-Fe@CN presented a longer stable voltage output, while that of MFC-CP decreased to about 0 mV in each cycle. Such extraordinary performance of MFC-Fe@CN was ascribed to three major aspects: (a) the Fe@CN-modified anode enhanced both of the electrochemically active surface area (EASA) and electrochemical activity of EABs for more active interfacial interaction between exoelectrogens and electrode. (b) The higher content of biomass (7.27 mg/cm 2 ) and EPS (2.69 mg/cm 2 ) with a higher proteins/polysaccharides ratio in MFC-Fe@CN was finely regulated by cells to keep a balance between EET efficiency and self-protection against environmental stress. (c) Geobacter species, the well-known exoelectrogens, were selectively enriched in abundance in MFC-Fe@CN, which was particularly beneficial for efficient electrons generation. The present study provided a new insight into the use of Fe@CN nanoparticles for enhancing the bioelectrochemical performance of MFCs, which may contribute to the practical application of MFCs in energy generation and wastewater treatment.