Facile Synthesis of the Amorphous Carbon Coated Fe-N-C Nanocatalyst with Efficient Activity for Oxygen Reduction Reaction in Acidic and Alkaline Media

With the assistance of surfactant, Fe nanoparticles are supported on g-C3N4 nanosheets by a simple one-step calcination strategy. Meanwhile, a layer of amorphous carbon is coated on the surface of Fe nanoparticles during calcination. Transmission electron microscopy (TEM), scanning electron microscope (SEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and inductively coupled plasma (ICP) were used to characterize the morphology, structure, and composition of the catalysts. By electrochemical evaluate methods, such as linear sweep voltammetry (LSV) and cyclic voltammetry (CV), it can be found that Fe25-N-C-800 (calcinated in 800 °C, Fe loading content is 5.35 wt.%) exhibits excellent oxygen reduction reaction (ORR) activity and selectivity. In 0.1 M KOH (potassium hydroxide solution), compared with the 20 wt.% Pt/C, Fe25-N-C-800 performs larger onset potential (0.925 V versus the reversible hydrogen electrode (RHE)) and half-wave potential (0.864 V vs. RHE) and limits current density (2.90 mA cm−2, at 400 rpm). In 0.1 M HClO4, it also exhibits comparable activity. Furthermore, the Fe25-N-C-800 displays more excellent stability and methanol tolerance than Pt/C. Therefore, due to convenience synthesis strategy and excellent catalytic activity, the Fe25-N-C-800 will adapt to a suitable candidate for non-noble metal ORR catalyst in fuel cells.


Introduction
As a significant cathode reaction, oxygen reduction reaction (ORR) has received extensive attention in many sustainable energy storage and conversion fields [1][2][3]. Compared with the anode reaction, composed of the hydrogen oxidation reaction, the cathode reaction, composed of the oxygen reduction reaction, has slow reaction rate. Therefore, the development of an oxygen reduction electrode catalyst with high catalytic activity has promising prospects in scientific research and actual production applications. Among all oxygen reduction catalysts, platinum-based catalysts are generally regarded as the best due to their low overpotential, high current density, and the four-electron transfer process in the reaction. Currently, 20 wt.% Pt/C catalysts are used in various commercial fuel cells. However, platinum on earth is limited and expensive. In addition, platinum catalysts also have poor methanol tolerance and insufficient stability, which are the biggest barriers to the commercial application of fuel cells [4].
Fe 25 -N-C-800 was taken as a typical example of Fe-N-C catalyst synthesis. First, 300 mg prepared g-C 3 N 4 and 600 mg P-123 were dissolved in 100 mL deionized water and sonicated for 2 h to form a homogenous dispersion. The well dispersed solution was then stirred vigorously by a magnetic stirrer for 2 h. Subsequently, 25 mg Fe(acac) 3 was sonicated continuously for 2 h after being added into the dispersed solution. After continuous stirring for 12 h, the mixture was dried in an oven at 80 • C for 10 h, and then the dried solids were collected and put into a vacuum tube high-temperature sintering furnace. Under the protection of nitrogen (60 mL·min −1 ), the dried solids were calcined at 550 • C for 2 h and 800 • C for another 2 h. The product of pyrolysis was soaked in pre-prepared 2 M HCl for 24 h, washed with water, and dried to obtain Fe 25 -N-C-800. X in Fe X -N-C-Y catalyst represents the amount of iron precursor added (X = 0, 5, 25, 50 mg). Y represents the calcination temperature during the preparation process (Y = 700, 800, 900 • C). All heating rates in this work were 5 • C·min −1 .

Materials Characterization
The characterization of the catalyst mainly included crystal structure, microscopic morphology, element composition, and pore size distribution. The X-ray diffractometer model DX-2700 (Dandong Fangyuan Instrument Co. LTD, Dandong, China) was used for testing. The X-ray tube was copper palladium (λ = 1.5417 Å), the scanning step width was 0.02 • , and the scanning angle was 10 •~8 0 • . The tube voltage and tube current were set to 40 kV and 30 mA, respectively. A scanning electron microscope (SEM, Zeiss SIGMA 300 field-emission, Oberkochen, Germany) characterized the microscopic morphology of the catalyst. Transmission electron microscopy (TEM) (FEI Tecnai G2 F20, FEI, Eindhoven, The Netherlands) collected information at a voltage of 200 kV. An inductively coupled plasma optical emission spectrometer (ICP-OES) measurement was performed with Agilent 720ES (Santa Clara, CA, USA), and the actual proportion of iron in the catalyst with different iron precursor content was determined. The K-Alpha type of Thermo Scientific (Waltham, MA, USA) was used to test the catalyst to obtain the valence state and surface energy state distribution of the catalyst by analyzing the X-ray photoelectron spectroscopy (XPS) (Thermo Scientific, Waltham, MA, USA).

Preparation of the Working Electrodes
All electrochemical characterization in this work was measured by a rotating disk electrode (RDE) with an electrode diameter of 5 mm. Before the electrochemical test, the glassy carbon electrode was Materials 2020, 13, 4551 4 of 16 polished with 0.5 µm and 0.05 µm alumina slurry, respectively. A 3 mg Fe-N-C catalyst was added into a mixed solution of 300 µL ethanol, 300 µL deionized water, and 300 µL 1 wt.% Nafion solution. The mixed solution was sonicated (KQ5200DE, 40 kHz, Kun Shan Ultrasonic Instruments Co., Ltd, Kunshan, China) for 0.5 h and then used as catalyst ink. Then, 7.95 µL catalyst ink was drip-evenly applied to the electrode surface of the RDE. The catalyst load was maintained at 0.135 mg cm −2 .

Electrocatalytic Measurements
Electrochemical characterization was mainly carried out by CV, LSV, and i-t. Electrochemical characterization of the catalysts were measured by the CHI electrochemical station (CHI 660E, Shanghai Chenhua Instrument Co., Ltd., Shanghai, China) in a standard three-electrode electrochemical cell at room temperature. During the test, a platinum wire electrode was used as the counter electrode, Ag/AgCl electrode as the reference electrode, and glassy carbon electrode loaded with catalyst as the working electrode. The potential was converted to a potential versus RHE according to E: (vs. RHE) = E (vs. Ag/AgCl) + 0.197 + 0.0591 pH.
The ORR catalytic activity of the sample was researched in both alkaline (0.1 M KOH) and acidic (0.1 M HClO 4 ) electrolyte. After an experimental device was set up, O 2 was introduced until the O 2 in the electrolyte was saturated. All electrochemical measurements were conducted in the O 2 -saturated media. For cyclic voltammetry (CV), 20 CV tests were scanned in the voltage range of 0 to 1.20 V (vs. RHE) at a sweep rate of 100 mV s −1 . For linear sweep voltammetry (LSV), ORR, active at the as-prepared catalyst, was surveyed by sweeping the catalysts between 1.20 to 0 V (vs. RHE) at 5 mV s −1 under 0, 100, 400, 900, and 1600 rpm (increasing from low speed to high speed). In this work, the ORR performance of each catalyst was tested more than four times, and the repeatable data were selected in our results analysis.
The electron transfer number of oxygen molecules on the electrode during the ORR process was evaluated via the rotating disk electrode and following Koutecky-Levich equations.
J, J L , and J k are the measured current density, diffusion-limiting current density, and kinetic-limiting current density at a specific potential, respectively. n is the overall number of electrons transferred during the reaction process. F is the Faraday constant (96485 C/mol). C 0 is the solubility of oxygen in this electrolyte, in 0.1 M KOH, C 0 = 1.2 × 10 −6 mol/cm 3 ; in 0.1 M HClO 4 , C 0 = 1.26 × 10 −6 mol/cm 3 . v is the kinematic viscosity of the solution, in 0.1 M KOH, ν = 0.01 cm −2 /s; in 0.1 M HClO 4 ,

Preparation and Physical Characterization of Fe-N-C
A brief synthesis method of the as-prepared catalyst in this work is illustrated schematically in Figure 1. First, the three precursors, g-C 3 N 4 , P-123, and Fe(acac) 3 ), were mixed together (details of the experiment are shown in Section 2.3). Subsequently, the thoroughly mixed and dried mixture was pyrolyzed at 800 • C (under nitrogen protection). The microscopic morphology of the Fe25-N-C-800 was examined by SEM (Figure 2a,b). Combining Figure 2a,b, a mixture of porous fold amorphous carbon and iron nanoparticles is displayed. In Figure 2b, it can be clearly observed that many nano-particle balls are evenly distributed on the carbon folds (bright white dots in Figure 2b), combined with XRD analysis, and it can be seen that the nanoparticles are formed by the agglomeration of iron elemental and Fe3C. Subsequently, the Fe25-N-C-800 catalyst was subjected to TEM testing (Figure 2c,d), which further confirmed that the catalyst was made of aggregated nanoparticles. The measurement found that the size of the nanoparticle ranged from 5 to 70 nm, and the average size was calculated to be 18.53 nm (Figure 2c, inset). Figure 2d is a high-resolution TEM image of the catalyst. It can be clearly seen that the iron nanoparticles were coated with amorphous carbon. Previous studies have shown that wrapping transition metal compounds in amorphous carbon enables them to interact strongly with each other, leading to synergistic effects that effectively enhance their electrochemical properties [23]. The lattice fringe spacing in Figure 2d was measured to be about 0.204 nm, which was consistent with the (110) crystal plane of elemental iron and the (220) crystal plane of Fe3C. This shows that the Fe25-N-C-800 catalyst formed elemental iron and Fe3C. For further research on the catalytic property of iron element in ORR, the elemental composition of the as-prepared catalyst in this work was evaluated by a highangle annular dark-field scanning transmission electron microscope (HAADF-STEM). It is shown in Figure 2e,h that many uniformly scattered bright spots were on the atomic scale. We can see that these bright spots were assigned to iron atoms. In this case, they can prove the existence of the iron element [6]. The microscopic morphology of the Fe 25 -N-C-800 was examined by SEM (Figure 2a,b). Combining Figure 2a,b, a mixture of porous fold amorphous carbon and iron nanoparticles is displayed. In Figure 2b, it can be clearly observed that many nano-particle balls are evenly distributed on the carbon folds (bright white dots in Figure 2b), combined with XRD analysis, and it can be seen that the nanoparticles are formed by the agglomeration of iron elemental and Fe 3 C. Subsequently, the Fe 25 -N-C-800 catalyst was subjected to TEM testing (Figure 2c,d), which further confirmed that the catalyst was made of aggregated nanoparticles. The measurement found that the size of the nanoparticle ranged from 5 to 70 nm, and the average size was calculated to be 18.53 nm (Figure 2c, inset). Figure 2d is a high-resolution TEM image of the catalyst. It can be clearly seen that the iron nanoparticles were coated with amorphous carbon. Previous studies have shown that wrapping transition metal compounds in amorphous carbon enables them to interact strongly with each other, leading to synergistic effects that effectively enhance their electrochemical properties [23]. The lattice fringe spacing in Figure 2d was measured to be about 0.204 nm, which was consistent with the (110) crystal plane of elemental iron and the (220) crystal plane of Fe 3 C. This shows that the Fe 25 -N-C-800 catalyst formed elemental iron and Fe 3 C. For further research on the catalytic property of iron element in ORR, the elemental composition of the as-prepared catalyst in this work was evaluated by a high-angle annular dark-field scanning transmission electron microscope (HAADF-STEM). It is shown in Figure 2e,h that many uniformly scattered bright spots were on the atomic scale. We can see that these bright spots were assigned to iron atoms. In this case, they can prove the existence of the iron element [6]. Figure 3a shows the X-ray diffraction pattern of catalysts with a different iron content. The spectra of different samples were roughly similar and there was no significant difference. It can be clearly seen that an amorphous peak of C appeared at 26 • . Each sample (except Fe 0 -N-C-800) had a set of apparent peaks at 44.67 • and 65.02 • . These peaks were consistent with the (110) Figure 3a, it can be seen that, with the difference of iron precursor content, the characteristic peaks of Fe and Fe 3 C were gradually obvious, from the inconspicuous "broad and relatively weak" peak to the clear "high and narrow" peak. Fe 0 -N-C-800 sample preparation process did not add iron precursor, so it does not contain iron element or iron characteristic peak. Combined with the XRD test data, according to the Scherer formula (D = kλ βcosθ ), we calculated that the size of the crystal grain of Fe 25 -N-C-800 was 17 nm, which was almost consistent with the data obtained by TEM. Meanwhile, ICP was used to determine the actual proportion of iron in the Fe X -N-C-800 catalyst under different iron precursor contents. It is shown in the results indicated in Table 1 that the content of iron was positively correlated with the amount of iron precursor added during the preparation process. That is, the content of iron in the sample increased with the addition of the precursor. Materials 2020, 13, x FOR PEER REVIEW 6 of 16  Figure 3a shows the X-ray diffraction pattern of catalysts with a different iron content. The spectra of different samples were roughly similar and there was no significant difference. It can be clearly seen that an amorphous peak of C appeared at 26°. Each sample (except Fe0-N-C-800) had a set of apparent peaks at 44.67° and 65.02°. These peaks were consistent with the (110) Figure 3a, it can be seen that, with the difference of iron precursor content, the characteristic peaks of Fe and Fe3C were gradually obvious, from the inconspicuous "broad and relatively weak" peak to the clear "high and narrow" peak. Fe0-N-C-800 sample preparation process did not add iron precursor, so it does not contain iron element or iron characteristic peak. Combined with the XRD test data, according to the Scherer formula ( = ), we calculated that the size of the crystal grain of Fe25-N-C-800 was 17 nm, which was almost consistent with the data obtained by TEM. Meanwhile, ICP was used to determine the actual proportion of iron in the FeX-N-C-800 catalyst under different iron precursor contents. It is shown in the results indicated in Table 1 that the content of iron was positively correlated with the amount of iron precursor added during the preparation process. That is, the content of iron in the sample increased with the addition of the precursor.   An XPS test was carried out on the Fe 25 -N-C-Y catalyst. The spectrum was a rough sweep of the full spectrum, so the signal of some elements was not obvious. Figure 3b shows the measured spectra of the catalyst compounded at various pyrolysis temperatures. It is shown in Figure 3b that the Fe-N-C samples consisted of carbon, nitrogen, oxygen, and iron. As shown in Table 2, with the increase of calcining temperature, the nitrogen content decreased from 5.22% to 2.48%. Through comparison, it can be clearly seen that the signal intensity of the N element of the sample calcined at 700 • C was more obvious. The possible cause for this phenomenon was that the combination of Fe(acac) 3 and g-C 3 N 4 ha poor high-temperature resistance under the participation of the surfactant P-123, with which the N element in g-C 3 N 4 was pyrolyzed and volatilized with the rise of pyrolysis temperature. According to previous studies, N-Fe sites are beneficial to increase the ORR catalytic activity of samples [6]. Therefore, it can be reasonably speculated that high pyrolysis temperature is conducive to forming N-Fe sites in the sample, but as the temperature rises, the nitrogen-containing sites will be decomposed at excessively Materials 2020, 13, 4551 7 of 16 high pyrolysis temperature. In consequence, the Fe 25 -N-C-800 catalyst manifests more excellent ORR performance than catalysts synthesised at other pyrolysis temperatures. more obvious. The possible cause for this phenomenon was that the combination of Fe(acac)3 and g-C3N4 ha poor high-temperature resistance under the participation of the surfactant P-123, with which the N element in g-C3N4 was pyrolyzed and volatilized with the rise of pyrolysis temperature. According to previous studies, N-Fe sites are beneficial to increase the ORR catalytic activity of samples [6]. Therefore, it can be reasonably speculated that high pyrolysis temperature is conducive to forming N-Fe sites in the sample, but as the temperature rises, the nitrogen-containing sites will be decomposed at excessively high pyrolysis temperature. In consequence, the Fe25-N-C-800 catalyst manifests more excellent ORR performance than catalysts synthesised at other pyrolysis temperatures.  To obtain further information about the valence state of atom and surface energy distribution of the catalyst, XPS tests were carried out on the Fe 25 -N-C-800 catalyst (Figure 3c-e). In this experiment, the XPS spectrum data of all samples were corrected using the carbon-carbon double bond peak at C 1s 284.8 eV. It can be clearly seen in the C 1s XPS binding energy region of the typical sample that the signal was deconvoluted into five peaks (Figure 3c). The peaks at 283.3, 284.2, 285.5, 286.9, and 289.1 eV can correspond to Fe 3 C, C=C, C=N, C-N, and O-C=O, respectively [24,25]. Among them, Fe 3 C had a positive effect on ORR. Figure 3d reveales the N 1s binding energy region of the Fe 25 -N-C-800 catalyst. The signal is deconvolved into three peaks at 397.8, 399.5, and 400.4 eV, belonging to pyridinic N, pyrrolic N/Fe-N X , and graphitic N, respectively [5,[26][27][28]. At the same time, we calculated the relative composition of the pyridinic N, pyrrolic N/ Fe-N X , and graphitic N species (atomic factions) in the samples (Figure 3f). The content of pyridinic N, pyrrolic N/Fe-N X , and graphitic N in Fe 25 -N-C-800 catalyst were 38.8%, 36.8%, and 24.4%, respectively. According to previous reports, ORR active sites in M-N-C catalysts may be located around the carbon phase [29]. Meanwhile, previous studies have demonstrated that pyridinic N and pyrrolic N are always located on the graphitic edge [30]. Therefore, pyridinic N and pyrrolic N are significant factors to enhance ORR catalytic reaction. Pyridinic N is a six-member heterocyclic compound, containing one nitrogen heterocyclic atom. Each N atom is bound to two carbon atoms and provides one p-electron to the aromatic π system [31]. Pyrrolic N is a five-member heterocyclic compound containing one azo atom, which binds to two carbon atoms and provides two p-electrons to the π-conjugated system [31]. Previous studies have demonstrated high proportions of pyridinic N and pyrrolic N boost O 2 reduction by enhancing the current density, spin density, and π state density of C atoms near the Fermi level [32]. Meanwhile, the two-dimensional planar structure of pyridinic N and pyrrolic N can maintain the original planar large π bond structure on the surface of carbon materials and have good electrical conductivity, so they have excellent ORR catalytic activity. However, due to the uneven three-dimensional structure of graphite nitrogen, the original π-conjugated large bond on the carbon surface is destroyed and the conductivity of carbon materials is reduced. Therefore, the catalytic activity of ORR is relatively low. Since nitrogen doping can cause disorder of the structure, a greater number of active sites are provided by the type of N-active site, such as pyridinic N, pyrrolic N, and Fe-N X , thereby promoting the reaction process of ORR [31][32][33][34]. Consequently, the Fe 25 -N-C-800 catalyst with the total content of pyridinic N and pyrrolic N/Fe-N X up to 76% showed excellent ORR catalytic performance.
In addition, the signal of the Fe 2p shown in Figure 3e was deconvolved into two main peaks, attributing to Fe 2p 3/2 and Fe 2p 1/2 , respectively. It can be found that the Fe 2p 3/2 binding energy region showed various peaks (710.4, 714.0, 719.3 eV), indicating that there were many chemical states of iron species in Fe 25 -N-C-800. Based on previous studies, it can be seen that the peak of Fe 2p 3/2 , at about 710.4 eV, is attributed to Fe-N X , which can accelerate O 2 adsorption and 4e-reduction and has electrocatalytic activity for ORR [11,[35][36][37][38]. Meanwhile, the 706.7 eV peak can be attributed to zero-valent iron and Fe 3 C [39]. Consistent with previous reports, the remaining peak at about 723.9 eV is attributed to Fe 2p 1/2 [6].
Iron is commonly believed to act as a crucial part in enhancing the ORR performance of electrocatalysts. It can improve the catalytic activity of electrocatalysts by promoting the formation of active sites or directly participating in catalyzing ORR [40]. The existence of Fe-N X was confirmed by the XPS spectra of the samples prepared in this work. Based on extensive literature reviews, it was found that Fe-N 4 is the most effective free active site of metal in oxygen reduction reaction among cathode catalysts of fuel cells [41]. Therefore, in combination with the ORR activity assessment results of the catalyst prepared in this work, and considering that the single iron atom in the metallic state is very active and unstable under electrochemical conditions, we reasonably speculated that the iron atom in Fe 25 -N-C-800 coordinated with the N to form the Fe-N 4 active site. Meanwhile, because Fe 3 C nanoparticles can enhance the active site in ORR, a sufficient number of Fe 3 C nanoparticles also have a certain impact on the activity of ORR. Considering all the above characteristics, we expect that Fe 25 -N-C-800 catalyst embracing Fe-N 4 and Fe 3 C can reveal excellent ORR activity.

ORR Catalytic Activity Evaluation
As a series of previous studies revealed, pyrolysis temperature greatly affects the ORR catalytic performance of Fe-N-C catalyst [42]. To explore the impression of pyrolysis temperature and obtain the suitable temperature, the ORR performance of the Fe 25 -N-C catalyst at various temperatures (in the range of 700-900 • C) was investigated. ORR performance was tested in 0.1 M KOH, and the consequences proved that Fe 25 -N-C-800 exhibited the excellent ORR performance with larger onset potential and half-wave potential (Figure 4a). This may be due to the fact that the conductivity, porosity, and active site density of the catalyst prepared at the pyrolysis temperature of 800 • C reached an optimal balance. put forward diverse assumptions for active sites in Fe−N-C catalysts that can enhance ORR properties. In general, there are two main types: one is doped with carbon-based N atoms [11,43,44] and the other is the active site formed by iron species, including Fe-NX, Fe3C, etc. [45][46][47]. Figure 4b provides the LSV curves of the sample with various iron precursor content. It can be found that the onset potential of the sample with the iron precursor (0.887-0.925 V vs. RHE) was significantly better than the sample without the iron precursor (0.794 V vs. RHE), which also facilitated faster reaction kinetics and a higher electron transfer number. Therefore, we can confirm that the active site of iron speciation serves as a significant part in the process of promoting ORR electrocatalytic properties.  Pt/C catalyst is recognized as an effective catalyst for commercial fuel cell cathode reaction (ORR). Therefore, researchers compared ORR catalytic activity evaluation with platinum-based catalysts [4]. The electrocatalytic activity of the as-prepared Fe-N-C catalyst for ORR was evaluated using RDE voltammetry and confronted with the Pt/C catalyst in this work. Previous studies have put forward diverse assumptions for active sites in Fe−N-C catalysts that can enhance ORR properties. In general, there are two main types: one is doped with carbon-based N atoms [11,43,44] and the other is the active site formed by iron species, including Fe-N X , Fe 3 C, etc. [45][46][47]. Figure 4b provides the LSV curves of the sample with various iron precursor content. It can be found that the onset potential of the sample with the iron precursor (0.887-0.925 V vs. RHE) was significantly better than the sample without the iron precursor (0.794 V vs. RHE), which also facilitated faster reaction kinetics and a higher electron transfer number. Therefore, we can confirm that the active site of iron speciation serves as a significant part in the process of promoting ORR electrocatalytic properties.
It is shown in Figure 5a that the onset potential of Fe 25 -N-C-800 (0.925 V vs. RHE) was mildly more negative than Pt/C (0.934 V vs. RHE), while its half-wave potential (0.864 V vs. RHE) was slightly more positive than Pt/C (0.858 V vs. RHE). Figure 5b displays the polarization curves of the Fe 25 -N-C-800 catalyst at various rotation rates (0~1600 rpm). With the enhancement of the rotating speed, the diffusion rate of O 2 in the electrolyte gradually accelerated, which lead to the gradual increase of the limiting of current density of the catalyst.
According to the RDE dates of Fe 25 -N-C-800 at various rotating speeds, combined with the K-L equations, the overall number of electrons transferred during the oxygen reduction process can be calculated. The K-L curves of Fe 25 -N-C-800 reveal an obvious linear relationship, indicating that the reaction process is a first-order kinetic process [48]. After calculation, the average electron transfer number (n) of the oxygen molecules of Fe 25 -N-C-800, over the potential range from 0.2-0.7 V, was 3.86 (Figure 5c), which was approximate to 20 wt.% Pt/C (n = 3.93). Furthermore, as shown in Figure 5d, the Tafel slope is considered to be another crucial parameter for investigating ORR catalytic activity, which mainly investigates the reaction rate of the reaction. In 0.1 M KOH, Fe 25 -N-C-800 had the lowest Tafel slope (29 mV per decade) and lower overpotential compared with other iron precursor content catalysts and Pt/C catalysts. Figure 5a that the onset potential of Fe25-N-C-800 (0.925 V vs. RHE) was mildly more negative than Pt/C (0.934 V vs. RHE), while its half-wave potential (0.864 V vs. RHE) was slightly more positive than Pt/C (0.858 V vs. RHE). Figure 5b displays the polarization curves of the Fe25-N-C-800 catalyst at various rotation rates (0~1600 rpm). With the enhancement of the rotating speed, the diffusion rate of O2 in the electrolyte gradually accelerated, which lead to the gradual increase of the limiting of current density of the catalyst. According to the RDE dates of Fe25-N-C-800 at various rotating speeds, combined with the K-L equations, the overall number of electrons transferred during the oxygen reduction process can be calculated. The K-L curves of Fe25-N-C-800 reveal an obvious linear relationship, indicating that the reaction process is a first-order kinetic process [48]. After calculation, the average electron transfer number (n) of the oxygen molecules of Fe25-N-C-800, over the potential range from 0.2-0.7 V, was 3.86 (Figure 5c), which was approximate to 20 wt.% Pt/C (n = 3.93). Furthermore, as shown in Figure  5d, the Tafel slope is considered to be another crucial parameter for investigating ORR catalytic activity, which mainly investigates the reaction rate of the reaction. In 0.1 M KOH, Fe25-N-C-800 had the lowest Tafel slope (29 mV per decade) and lower overpotential compared with other iron precursor content catalysts and Pt/C catalysts.

It is shown in
Previous researchers have found that the Fe-N-C catalyst has a stable structure but unstable electrochemical performance when exposed to H2O2 (ORR by-product) in acidic electrolyte [49]. The catalytic active site of the Fe-N-C catalyst was not be affected, but its conversion frequency was Previous researchers have found that the Fe-N-C catalyst has a stable structure but unstable electrochemical performance when exposed to H 2 O 2 (ORR by-product) in acidic electrolyte [49]. The catalytic active site of the Fe-N-C catalyst was not be affected, but its conversion frequency was reduced through the oxidation of the carbon surface, resulting in the weakening of the binding ability of the Fe-N-C catalyst to O 2 . Therefore, compared with the alkaline electrolyte, the ORR performance and durability of the Fe-based catalyst in the acidic electrolyte was significantly reduced. It is noteworthy that the Fe 25 -N-C-800 catalyst also had a certain performance for ORR in the harsh acidic medium. Compared with commercial 20 wt.% Pt/C, although the Fe 25 -N-C-800 catalyst's onset potential and half-wave potential showed a certain degree of negative shift, its limiting current density was much bigger in 0.1 M HClO 4 (Figure 6a). Figure 6b displays the LSV curves of the Fe 25 -N-C-800 catalyst at various rates (0~1600 rpm). We can see that, as the rotation speed increased, the O 2 diffusion rate increased and the limiting current density of the catalyst gradually increased, which was the same as when the electrolyte was alkaline. Meanwhile, the K-L plots of the Fe 25 -N-C-800 catalyst presented an obvious linear relationship. Over the potential range from 0.1-0.4 V, the average number of electrons transferred from the oxygen molecules of Fe 25 -N-C-800 during the reaction process was 3.79 (Figure 6c), which was close to that of Pt/C (n = 3.85). This indicates that, even in harsh acidic solutions, there were nearly four electron transfer ORR pathways. It is shown in Figure 6d that the Tafel slopes of the The results demonstrated that, in the low overpotential region, Fe 25 -N-C-800 also had a similar oxygen reduction reaction mechanism to Pt/C. O2 diffusion rate increased and the limiting current density of the catalyst gradually increased, which was the same as when the electrolyte was alkaline. Meanwhile, the K-L plots of the Fe25-N-C-800 catalyst presented an obvious linear relationship. Over the potential range from 0.1-0.4 V, the average number of electrons transferred from the oxygen molecules of Fe25-N-C-800 during the reaction process was 3.79 (Figure 6c), which was close to that of Pt/C (n = 3.85). This indicates that, even in harsh acidic solutions, there were nearly four electron transfer ORR pathways. It is shown in Figure  6d that the Tafel slopes of the Fe25-N-C-800 catalyst and 20 Wt. % Pt/C were approximately 34 and 45 mV per decade, respectively. The results demonstrated that, in the low overpotential region, Fe25-N-C-800 also had a similar oxygen reduction reaction mechanism to Pt/C.

ORR Durability Characterization
In practical applications, catalyst stability and methanol tolerance are also important parameters for direct methanol fuel cells. As shown in Figure 7a,c, either in a 0.1 M KOH or 0.1 M HClO 4 media, after adding CH 3 OH (3 M), the current response of Pt/C dropped sharply, while the current response of Fe 25 -N-C-800 did not change significantly. The consequences prove that Fe 25 -N-C-800 takes on prominent resistance to the methanol crossover effect. Figure 7b,d illustrates the electrocatalytic durability of the Fe 25 -N-C-800 catalyst and Pt/C catalyst. After 28,800 s chronocurrent electrolysis in 0.1 M KOH media, the current density of the Fe 25 -N-C-800 catalyst dropped to 91% of the initial current, whereas the current density of the Pt/C catalyst was only 82%. In 0.1 M HClO 4 media, the current density of the Fe 25 -N-C-800 catalyst left about 83% of the initial current, while that of Pt/C sharply dropped, leaving only 71% of the initial current. According to previous studies, this is because Pt may dissolve in the electrolyte, aggregate into larger particles, and separate from the carrier, resulting in poor durability in acidic media [50]. was only 82%. In 0.1 M HClO4 media, the current density of the Fe25-N-C-800 catalyst left about 83% of the initial current, while that of Pt/C sharply dropped, leaving only 71% of the initial current. According to previous studies, this is because Pt may dissolve in the electrolyte, aggregate into larger particles, and separate from the carrier, resulting in poor durability in acidic media [50].  Table 3 summarizes the comparison of ORR performance between Fe25-N-C-800 prepared in this work and the recently reported Fe-N-C catalyst in alkaline and acidic electrolyte. The electron transfer number of Fe25-N-C-800 at different potentials was close to 4. Therefore, the oxygen reduction reaction on the Fe25-N-C-800 catalyst followed the efficient four-electron path, that is, the oxygen was completely reduced to water, which proves that the catalyst of our vegetation had effective ORR catalytic activity. At the same time, Table 3 confirms that the durability of the catalyst prepared in this work was superior to other types of Fe-N-C to a certain extent, no matter whether it was under alkaline or acidic electrolyte. Table 3. Electrochemical performance of different electrocatalysts for ORR.

Samples
Alkaline Electrolyte Acidic Electrolyte Reference  Table 3 summarizes the comparison of ORR performance between Fe 25 -N-C-800 prepared in this work and the recently reported Fe-N-C catalyst in alkaline and acidic electrolyte. The electron transfer number of Fe 25 -N-C-800 at different potentials was close to 4. Therefore, the oxygen reduction reaction on the Fe 25 -N-C-800 catalyst followed the efficient four-electron path, that is, the oxygen was completely reduced to water, which proves that the catalyst of our vegetation had effective ORR catalytic activity. At the same time, Table 3 confirms that the durability of the catalyst prepared in this work was superior to other types of Fe-N-C to a certain extent, no matter whether it was under alkaline or acidic electrolyte. Fe-N/C 800 3.78-3.89 75.64%/7000 s -- [55] The excellent stability and durability of the Fe 25 -N-C-800 catalyst perhaps contributed to the fact that Fe nanoparticles were wrapped by amorphous carbon to protect metals from being dissolved. A similar result was observed in previous reports, that is, the encapsulated iron nanoparticles cannot be dissolved in acid [56]. A series of results show that the Fe 25 -N-C-800 catalyst possesses a certain prospect in the application of fuel cells, according to its satisfactory methanol resistance and durability in both alkaline and acidic media.

Conclusions
To sum up, we used surfactants to support Fe nanoparticles on g-C 3 N 4 nanosheets through one-step calcination. Meanwhile, a layer of amorphous carbon was formed on the surface of Fe nanoparticles, thereby preparing an iron-based nanocomposite catalyst. According to a series of physical characteristics, such as SEM, TEM, XRD, and XPS, we proposed that the iron atoms, Fe 3 C nanoparticles, and Fe-N X active sites produced during the pyrolysis process raised the ORR catalytic activity to some extent. The optimal pyrolysis temperature (800 • C) and iron precursor content (25 mg) were determined by preparing various catalysts. The Fe 25 -N-C-800 catalyst exhibited comparable catalytic activity to commercial 20 wt.% Pt/C in alkaline electrolyte. Via a series of electrochemical evaluations, it was detected that the initial potential of the Fe 25 -N-C-800 was 0.925 V vs. RHE, the half-wave potential was 0.864 V vs. RHE, and the number of electron transfers was 3.86. At the same time, its excellent stability and methanol resistance were manifested because of the amorphous carbon protection. In acidic electrolyte, the Fe 25 -N-C-800 catalyst also had a four-electron transfer ORR pathway and, at the same time, exhibits more excellent methanol tolerance and stability than Pt/C. Therefore, the catalyst prepared for this work can be used as a fuel cell cathode catalyst.