Fabrication of functionalized 3D graphene with controllable micro/meso-pores as a superior electrocatalyst for enhanced oxygen reduction in both acidic and alkaline solutions

Abstract

We have synthesized hierarchically 3D porous Fe₅C₂/N-doped graphene (3D Fe₅C₂/N-PG) with numerous micro/meso-pores by a controllable and facile strategy. The density of the micro/meso-pores can be accurately tailored by adjusting the reaction conditions. It is found that the ORR performance of the electrocatalyst is strongly dependent on the pore structures under the same conditions. The electrocatalyst exhibits comparable activity in an acidic solution together with superior activity in an alkaline solution to the state-of-the-art Pt/C catalyst in terms of onset potential, half-wave potential, diffusion-limiting current density and kinetic current density. The excellent electrocatalytic activity is attributed to the balance of the surface area, hierarchical porosity, chemical composition (pyridinic N, graphitic N and Fe₅C₂ nanoparticles) and conductivity of the 3D interconnected graphene. This strategy for the tunable synthesis of hierarchically 3D porous graphene materials offers a platform for developing catalysts with superior activity in energy devices.

---

Introduction

To address the environmental issues of severe energy shortages and pollution, it is necessary to develop sustainable energy technologies, such as fuel cells or lithium-ion batteries. Fuel cells are eco-friendly and high-efficiency energy conversion systems that are attracting increasing interest. The cathodic oxygen reduction reaction (ORR) is the key to determine the function of fuel cells; however, it suffers from inactive kinetics, thus hampering the applications of fuel cells.^1,2 Currently, Pt-based catalysts are commercially used to catalyze the ORR with high activity.^3,4 Nevertheless, the high price, instability and low tolerance to methanol of these catalysts are still great obstacles for the worldwide development of fuel cells.⁵ Therefore, the synthesis of low-cost electrocatalysts with efficient catalytic activity, enhanced stability and excellent methanol tolerance is extremely desirable to provide substitutes for Pt-based catalysts.^6,7

Graphene is promising as an advantageous electrocatalyst because of its high surface area, excellent electrical conductivity and tailorable characteristics by introducing foreign atoms.^8–12 Unfortunately, the chemical/thermal reduction of graphite oxide (GO) often results in aggregation or restacking of graphene sheets due to the attraction between the graphene sheets.¹³ Furthermore, the stacking of graphene sheets will lower their specific surface area and hinder mass transfer, thus reducing their electrocatalytic performance. In addition, remarkably, graphene chemically doped with foreign atoms has emerged as a potential material which displays excellent activity for ORR in alkaline media; however, the activity of most graphene-based catalysts is still very poor in acidic media,^14,15 which prevents their application in proton exchange membrane fuel cells (PEMFCs). Accordingly, it is extraordinary challenging but increasingly necessary to rationally design and tunably synthesize graphene-based catalysts with porous structures and high surface areas which, importantly, have high ORR activity in acidic media.

It is believed that macro-porous and meso-porous structures are unable to satisfy the requirements of porous graphene in the electrocatalytic field. A hierarchically porous structure of graphene, especially with high micro-porosity, is essential to promote its catalytic performance.¹⁶ Although some studies have aimed to obtain graphene materials with high surface areas, and 3D graphene materials have been developed as ORR electrocatalysts,^17–20 there is little focus on observing the effect of hierarchical porosity on the catalytic performance of 3D graphene materials. In addition, the precisely controlled synthesis of 3D graphene nanostructures with abundant micro-pores and meso-pores remains a challenging task.

Generally, the ORR performance of a catalyst is affected by its chemical structure as well as the quantity of catalytic active sites, specific surface area, mass transport and conductivity in the catalyst layer. The two crucial factors are as follows: (1) the chemical structure; (2) the surface area and appropriate pore structure. In order to improve the catalytic activity in both acidic and alkaline media, herein, we provide a protocol to construct novel Fe₅C₂/N-doped 3D graphene with controllable micro/meso-pores by acid oxidation, hard-template and pyrolysis methods. Moreover, this opens new avenues for accurately tailoring the density of micro/meso-pores of 3D graphene. It is observed that the porosity structure of 3D graphene definitely plays an important role in its electrocatalytic performance for ORR. The functionalized 3D graphene (3D Fe₅C₂/N-PG) electrocatalysts exhibit superior activity, higher stability and better methanol tolerance for ORR not only in alkaline medium, but also in acidic medium.

Experimental section

Chemicals

Natural flake graphite powder was purchased from Beijing Chemical Company (China). Nitric acid (HNO₃) was bought from Beijing Chemical Works. Iron(III) nitrate nonahydrate (Fe(NO₃)₃·9H₂O) was provided by Tianjin Guangfu Fine Chemical Research Institute. Melamine (C₃N₃(NH₂)₃) was provided by Tianjin Huadong Technology Development. SiO₂ spheres (d = 30 nm) were bought from the Aladdin Industrial Corporation. Ammonium hydroxide (NH₃·H₂O) was purchased from Tianjin Guangfu Technology Development Co. Ltd (China). Hydrofluoric acid (HF) was provided by Beijing Chemical Works. Pt/C (20 wt% Pt on Vulcan XC-72) was obtained from Alfa Aesar.

Synthesis of porous graphene (PG)

First, graphene oxide (GO) was prepared from natural graphite flakes using a modified Hummers method.²¹ The as-prepared GO product was readily dispersed in deionized (DI) water to obtain a stable aqueous solution at a concentration of 2 mg mL⁻¹ for use. Then, porous GO (PG) sheets were prepared using a previously reported procedure with some modifications.²² Briefly, 40 mL of the above GO aqueous solution (2 mg mL⁻¹) was mixed with 40 mL of concentrated nitric acid (70 wt%) under stirring. The mixture was subsequently subjected to ultrasonic vibration for 0.5 h, 1 h and 2 h, respectively. After completing sonication, the mixture was settled at RT for 2 h and then poured into 300 mL of DI water, centrifuged and washed with DI water several times to remove the acid. Then, it was readily dispersed in 40 mL deionized (DI) water to obtain a stable aqueous solution (2 mg mL⁻¹), which was identified as PG-h (h indicates the ultrasonic time).

Synthesis of 3D Fe₅C₂/N-PG

Briefly, first, 40 mL of the above PG aqueous solution was mixed with 20 mL of ammonium hydroxide (25 wt%) and SiO₂ spheres (d = 30 nm) under stirring for 24 h. Then, the mixture was evaporated to remove ammonia and water; next, 20 mL DI water, 80 mg melamine and 80 mg Fe(NO₃)₃·9H₂O were added under ultrasonic vibration for 1 h. Consequently, the mixture was evaporated to remove water, and the resulting product was dried under vacuum for further use. Lastly, the solid products were heated to 700 °C, 800 °C and 900 °C at 10 °C min⁻¹ under an argon atmosphere in a tube furnace, then maintained at the peak temperature for 2 h and allowed to cool to room temperature; they were then etched by HF (10 wt%) solution. The final products were described as 3D Fe₅C₂/N-PG-h-T (T represents the pyrolyzing temperature). The preparation procedure is schematically illustrated in Fig. 1.

Fig. 1 Schematic of the preparation of 3D Fe₅C₂/N-PG-h-T.

Synthesis of 3D Fe₅C₂/N-RGO-800

As a control experiment, 3D Fe₅C₂/N-RGO-800 was also prepared by a similar procedure as 3D Fe₅C₂/N-PG-800, using GO to replace PG.

Characterization

Scanning electron microscopy (SEM) images were obtained with a JEOL JSM-6701F electron microscope operating at 5 kV. The morphologies and structures of the samples were observed using transmission electron microscopy (TEM, JEOL-2010 transmission electron microscope operating at 200 kV). XPS measurements were recorded on an ESCLAB 250 spectrometer with a monochromatized Al Kα X-ray source to characterize the surface chemical composition and the valence state. The nitrogen sorption experiments were performed at 77 K on a Micromeritics ASAP-2020 sorption analyser. Raman spectral analysis was carried out on a Lab RAM HR Evolution Raman spectrometer at 532 nm. X-ray diffraction (XRD) patterns were collected by a powder X-ray diffractometer at 40 kV and 15 mA using Co-Kα radiation (RIGAK, D/MAX2550 VB/PC) at room temperature.

Electrochemical measurements

First, to prepare the working electrode, 1 mg catalyst was dispersed in 1 mL ethanol by sonication treatment. The samples were bath sonicated for 1.5 h to afford a relatively homogenous dispersion. The surface of a glassy carbon electrode (GCE) was polished with 0.3 and 0.05 mm α-Al₂O₃ powder slurry successively and ultrasonically rinsed with absolute ethanol and distilled water for a short time. Then, the as-prepared dispersion was immediately drop-cast on a GC electrode with a catalyst loading of about 280 μg cm⁻² for all catalysts, followed by drying in air. After that, 1.5 μL of a diluted Nafion solution (5 wt% in ethanol) was further cast over the catalyst layer to prevent loss of the catalyst during electrochemical testing.

A rotating disk electrode (RDE) with a GC disk diameter of 3 mm was used to carry out cyclic voltammetry (CV) or linear scan voltammetry (LSV) in O₂-saturated 0.1 M KOH solution or 0.5 M H₂SO₄ solution. Current–time chronoamperometric response (i–t) and rotating ring disk electrode (RRDE, ddisk = 5.61 mm) measurements were carried out with a Pine Instrument Company AF-MSRCE modulator rate rotator on a CHI 760E electrochemical workstation (Shanghai CHENHUA Company) using a standard three-electrode system consisting of a modified GCE as the working electrode, a platinum wire counter electrode and a saturated calomel reference electrode (SCE) in an O₂-saturated 0.1 M KOH solution or 0.5 M H₂SO₄. The SCEs were converted to RHE scale. In O₂-saturated 0.5 M H₂SO₄, E(RHE) = E(SCE) + 0.273; in O₂-saturated 0.1 M KOH, E(RHE) = E(SCE) + 0.998. In the methanol-tolerance experiments, the working electrode was immersed in an O₂-saturated 0.1 M KOH solution or 0.5 M H₂SO₄ with 3 M CH₃OH. The ORR kinetics parameters of the catalysts can be obtained using the following Koutecky–Levich (K–L) eqn (1) and (2):²³

1/J = 1/J_L + 1/J_K = 1/(Bω^1/2) + 1/J_K (1)

B = 0.62nFC₀(D₀)^2/3ν^−1/6 (2)

where J is the measured current density, J_L and J_K are the diffusion limiting and kinetic limiting current densities, respectively, ω is the angular velocity of the disk (ω = 2πN, N is the electrode rotating speed), B is the Levich slope, n is the transferred electron number, F is the Faraday constant (96 485 C mol⁻¹), C₀ is the bulk concentration of O₂ (1.2 × 10⁻³ mol cm⁻³ in 0.1 M KOH, 1.26 × 10⁻³ mol cm⁻³ in 0.5 M H₂SO₄). D₀ is the diffusion coefficient of O₂ in electrolyte (1.9 × 10⁻⁵ cm² s⁻¹ in 0.1 M KOH, 1.93 × 10⁻⁵ cm² s⁻¹ in 0.5 M H₂SO₄), and ν is the kinetic viscosity (0.01 cm² s⁻¹ in 0.1 M KOH, 0.01009 cm² s⁻¹ in 0.5 M H₂SO₄). Moreover, the diffusion-limiting current (J_L) is obtained from the corresponding current divided by the geometric area of the GCE (0.19625 cm⁻²).

The transferred electron number (n) and H₂O₂ yield during ORR can be calculated using the following equations:²⁴

n = 4I_d/(I_d + I_r/N) (3)

H₂O₂% = 200I_r/NI_d + I_r (4)

where I_r is the ring current, I_d is the disk current, and N is the current collection efficiency of the Pt ring.

Results and discussion

Characterization of structure and morphology

The nitrogen physisorption of 3D Fe₅C₂/N-PG-0.5-800, 3D Fe₅C₂/N-PG-1.0-800 and 3D Fe₅C₂/N-PG-2.0-800 was first measured to investigate the porous structures. The adsorption–desorption isotherms of all samples in Fig. 2a show a IV type, suggesting the existence of meso-pores. Furthermore, the micro-, meso- and macro-pores are observed in the corresponding pore size distribution in Fig. 2b. The inset in Fig. 2b depicts the pore size distribution in the range of 0 to 35 nm. It is worth noting that the micro-pore and meso-pore size distribution can be tailored through controlling the synthesis time of PG. The Brunauer–Emmett–Teller surface area, pore volume and average pore size of 3D Fe₅C₂/N-PG-0.5-800, 3D Fe₅C₂/N-PG-1.0-800 and 3D Fe₅C₂/N-PG-2.0-800 are summarized in Table 1. It is noted that the BET surface area and pore volume of 3D Fe₅C₂/N-PG-1.0-800 are higher than those of the other two samples, which is due to the fact that the pore size and density of Fe₅C₂/N-PG-h-800 initially increase with the treatment time of GO in HNO₃ when the treatment time is less than 1 h. Meanwhile, when the treatment time reaches 2 h, the pore size and volume decrease because some GO sheets may break into small pieces, as previously reported.²⁵ In order to obtain a higher BET surface area, pore volume and appropriate pore size distribution, which are favourable for the exposure of active sites and for fast mass transfer, the sonication time of GO in HNO₃ was optimized to 1 h.

Fig. 2 (a) Nitrogen adsorption–desorption isotherms of 3D Fe₅C₂/N-PG-0.5-800, 3D Fe₅C₂/N-PG-1.0-800 and 3D Fe₅C₂/N-PG-2.0-800. (b) Pore size distribution curves of 3D Fe₅C₂/N-PG-0.5-800, 3D Fe₅C₂/N-PG-1.0-800 and 3D Fe₅C₂/N-PG-2.0-800.

Table 1 Porous parameters of 3D Fe₅C₂/N-PG-0.5-800, 3D Fe₅C₂/N-PG-1.0-800 and 3D Fe₅C₂/N-PG-2.0-800

Material	SBET (m^2 g^−1)	Pore volume (cm^3 g^−1)	Average pore size (nm)
3D Fe5C2/N-PG-0.5-800	366	1.06	11.52
3D Fe5C2/N-PG-1.0-800	474	1.83	15.45
3D Fe5C2/N-PG-2.0-800	404	1.48	14.69

---

SEM and TEM technology were employed to observe the morphology of 3D Fe₅C₂/N-PG-1.0-800. Fig. 3a displays a 3D porous graphene structure, which suggests the PG successfully forms 3D structures by the hard-template and pyrolysis strategies. Furthermore, the TEM image of 3D Fe₅C₂/N-PG-1.0-800 in Fig. 3b clearly exhibits a wrinkled morphology with numerous pores, further proving that the PG sheets aggregate to assemble an interconnected 3D porous architecture, which facilitates the diffusion of electrolytes and reactants. Moreover, a few iron nanoparticles are observed on the 3D graphene sheets. To confirm the elaborate structure of the nanoparticles, the corresponding HRTEM image in Fig. 3c undoubtedly indicates that a small iron nanocrystal particle is embedded into a few graphite carbon layers, confirming the formation of iron carbide nanoparticles, which are considered to be superior active sites for ORR in both acidic and alkaline solutions,^26–28 thereby promoting the ORR activity.

Fig. 3 (a) SEM image of 3D Fe₅C₂/N-PG-1.0-800. (b) TEM image of 3D Fe₅C₂/N-PG-1.0-800. (c) HRTEM image of 3D Fe₅C₂/N-PG-1.0-800.

X-ray photoelectron spectroscopy (XPS) was performed to evaluate the chemical composition of 3D Fe₅C₂/N-PG-1.0-T. In Fig. 4a, the C 1s, N 1s and O 1s signals on all samples are obviously present in the XPS survey spectra, proving that nitrogen is successfully doped into carbons after pyrolyzing. On the other hand, the Fe peak is not clearly found; this is attributed to the encasement of Fe in the graphite carbon layers. The high-resolution N 1s spectrum in Fig. 4b of the 3D Fe₅C₂/N-PG-1.0-800 catalyst shows four peaks at 398.7 eV, 399.8 eV, 401.1 eV and 402.4 eV, revealing the presence of pyridinic N, pyrrolic N, graphitic N and N-oxide, respectively. The high-resolution N 1s spectra of the 3D Fe₅C₂/N-PG-1.0-700 and 3D Fe₅C₂/N-PG-1.0-900 catalysts are shown in Fig. S1.† The X-ray diffraction (XRD) patterns of 3D Fe₅C₂/N-PG-1.0-T are exhibited in Fig. 4c. The peak located at 26.0° corresponds to the graphite carbon; also, the peaks at 44.0° and 44.6° are related to Fe₅C₂ (PDF card no. 20-0508). The crystal plane angles of 44.0° and 44.6° are attributed to the (021) and (−402) planes of Fe₅C₂ phase, demonstrating that the iron carbide nanoparticle in Fig. 3c can be identified as Fe₅C₂ species. As shown in Fig. 4d, the Raman spectra of 3D Fe₅C₂/N-PG-1.0-T show two prominent peaks at 1345 and 1590 cm⁻¹, belonging to the D band arising from the disordered carbon atoms and the G band related to sp²-hybridized graphitic carbon atoms, respectively. The intensity ratio of the D and G band (I_D/I_G) is a defect indicator of graphitic materials.^29–31 The I_D/I_G ratios of 3D Fe₅C₂/N-PG-1.0-T decrease with the pyrolysis temperature because a higher pyrolysis temperature results in a higher graphitization degree of graphene.

Fig. 4 (a) XPS survey spectra of 3D Fe₅C₂/N-PG-1.0-T. (b) High-resolution N 1s spectrum of 3D Fe₅C₂/N-PG-1.0-800. (c) XRD curves of 3D Fe₅C₂/N-PG-1.0-T and standard XRD pattern of Fe₅C₂. (d) Raman spectra of 3D Fe₅C₂/N-PG-1.0-T.

Based on the above characterizations, it can be seen that 3D Fe₅C₂/N-PG-1.0-800 is composed of pyridinic N, pyrrolic N, graphitic N, and N-oxide along with Fe₅C₂ nanoparticles. It has been confirmed that graphitic N can enhance the limiting-current density and pyridinic N can improve the onset potential, while pyrrolic N has little effect on ORR.³² In addition, iron carbide nanoparticles embedded in graphite carbon layers are well known to be superior active sites for ORR in both acidic and alkaline solutions.^26–28

The electrocatalytic evaluation of ORR in acidic medium

In Fig. 5a, the CV curves show no significant peak in the N₂-saturated solution. Contrastingly, a characteristic ORR peak at about 0.50 V vs. RHE is observed in the presence of oxygen, indicating the electrocatalytic activity of 3D Fe₅C₂/N-PG-1.0-800 towards ORR. To further observe the ORR electrocatalytic activity of the catalysts, linear sweep voltammetry (LSV) measurements on a rotating disk electrode (RDE) were performed in O₂ saturated 0.5 M H₂SO₄ solution. From Fig. 5b, it is noted that the onset potential (E_onset) and half-wave potential (E_1/2) of 3D Fe₅C₂/N-PG-1.0-800 are approximately 0.88 and 0.71 V vs. RHE, respectively, vs. RHE; these values are more positive than those of the other Fe₅C₂/N-PG-h-800 samples and Fe₅C₂/N-RGO-800 and are even equal to and 50 mV lower than those of Pt/C, proving that the electrocatalytic activity is strongly dependent on the porosity structure as well as the surface area. Additionally, these values are similar or larger than those of other reported catalysts, as shown in Table S2,† indicating the superior electrocatalytic activity of 3D Fe₅C₂/N-PG-1.0-800 for ORR. Furthermore, the diffusion-limiting current density on Fe₅C₂/N-PG-1.0-800 is larger than on the other samples. The E_onset, E_1/2, and J_L for ORR on the catalyst are given in Table 2. In comparison to Fe₅C₂/N-PG-1.0-800, all the Fe₅C₂/N-PG-h-800 samples present lower E_onset and E_1/2 values. Thus, Fe₅C₂/N-PG-1.0-800 was used in the subsequent research. As exhibited in Fig. 5c, the influence of the pyrolyzing temperature on the electrocatalytic activity is also observed. The E_onset, E_1/2 and J_L of 3D Fe₅C₂/N-PG-1.0-800 are more positive and larger than other 3D Fe₅C₂/N-PG-1.0 samples pyrolyzed at different temperatures, probably owing not only to the appropriate porosity structure and high surface area, which benefit mass transfer and the exposure of active sites, but also the higher content of effectively active sites, including the doped pyridinic N and graphitic N, which can improve the onset potential and boost the limiting-current density, as well as the iron carbide nanoparticles encased in graphite carbon layers, which are well known to be superior active sites for ORR (Table S1†). The reaction kinetics for Fe₅C₂/N-PG-1.0-800 and Pt/C were studied by rotating disk voltammetry at various rotation rates. Fig. 5d and e describe the LSVs of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C in O₂-saturated 0.5 M H₂SO₄; the current density improved with increasing rotation rate from 400 to 2500 rpm, ascribed to the shortened diffusion distance for O₂. The insets in Fig. 5d and e depict the corresponding K–L plots based on eqn (1) and (2) at various potentials. The curves exhibit good linearity and constant slopes at different potentials, indicating first-order reaction kinetics for ORR with regard to the concentration of O₂ dissolved in solution, coupled with a similar electron transfer number. Also, the kinetic current densities of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C at selected potentials were further determined and are shown in Fig. 5f. The kinetic current density of Fe₅C₂/N-PG-1.0-800 is higher than that of Pt/C, indicating better activity on Fe₅C₂/N-PG-1.0-800. The Tafel slopes for 3D Fe₅C₂/N-PG-1.0-800 and Pt/C, shown in Fig. 5g and h, are −107.3 and −131.9 mV dec⁻¹, respectively, suggesting that the ORR activity on 3D Fe₅C₂/N-PG-1.0-800 is better than that of Pt/C catalyst in acidic medium. Moreover, the mass activity (MA) of 3D Fe₅C₂/N-RGO-800, 3D Fe₅C₂/N-PG-h-800 and the commercial Pt/C catalyst at 0.6 V vs. RHE in O₂-saturated 0.5 M H₂SO₄ are given in Fig. S2.† The MA of Fe₅C₂/N-PG-1.0-800 is 69.93 mA mg⁻¹, which is significantly higher than those of Fe₅C₂/N-RGO-800 (23.21 mA mg⁻¹), 3D Fe₅C₂/N-PG-0.5-800 (38.86 mA mg⁻¹), and 3D Fe₅C₂/N-PG-2.0-800 (20.93 mA mg⁻¹) and is close to that of the Pt/C catalyst (80.64 mA mg⁻¹), suggesting better ORR activity on the Fe₅C₂/N-PG-1.0-800 catalyst.

Fig. 5 CV curves of 3D Fe₅C₂/N-PG-1.0-800 at a scan rate of 10 mV s⁻¹ in 0.5 M H₂SO₄ solution saturated with N₂ and O₂ (a), LSV curves for the 3D Fe₅C₂/N-PG-h-800, Fe₅C₂/N-RGO-800 and Pt/C samples at 1600 rpm (b), 3D Fe₅C₂/N-PG-1.0-T at 1600 rpm (c), 3D Fe₅C₂/N-PG-1.0-800 at different rotation rates (d) and Pt/C at different rotation rates (e) in O₂-saturated 0.5 M H₂SO₄ at 10 mV s⁻¹. The insets in (d) and (e) show the corresponding K–L plots. The kinetic current densities of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C (f) at different potentials in O₂-saturated 0.5 M H₂SO₄. Tafel plots of 3D Fe₅C₂/N-PG-1.0-800 (g) and Pt/C (h) in O₂-saturated 0.5 M H₂SO₄ solution.

Table 2 Electrochemical parameters for ORR estimated from LSV in O₂-saturated 0.5 M H₂SO₄ at a scan rate of 10 mV s⁻¹ with a rotation speed of 1600 rpm

Material	Eonset (V)	E1/2 (V)	JL (mA cm^−2)	Electrolyte solution	Reference electrode	Reference
3D Fe5C2/N-PG-0.5-800	0.85	0.68	5.68	0.5 M H2SO4	RHE	In this work
3D Fe5C2/N-PG-1.0-800	0.88	0.71	6.62	0.5 M H2SO4	RHE	In this work
3D Fe5C2/N-PG-2.0-800	0.86	0.67	5.57	0.5 M H2SO4	RHE	In this work
3D Fe5C2/N-RGO	0.82	0.66	5.75	0.5 M H2SO4	RHE	In this work
Pt/C	0.88	0.76	5.03	0.5 M H2SO4	RHE	In this work

---

To further examine the ORR activity on the samples, RRDE measurements were employed to evaluate the ORR process in depth. Fig. 6a shows the disk and ring current density on 3D Fe₅C₂/N-PG-1.0-800 and Pt/C. Calculated from eqn (3) and (4), we can acquire the H₂O₂ yield (%) and the electron transfer number. In Fig. 6b, H₂O₂ is indistinctively detected with a yield below 0.3%, close to that of the Pt/C catalyst. The electron transfer number is about 3.97 to 4.00 at 0.10 to 0.40 V vs. RHE in Fig. 6c, suggesting a 4e⁻ process for ORR. The above-discussed results based on RDE and RRDE measurements significantly demonstrate that 3D Fe₅C₂/N-PG-1.0-800 possesses excellent activity for ORR, similar to that of the commercial Pt/C catalyst.

Fig. 6 (a) Rotating ring disk electrode (RRDE) measurements of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C in O₂-saturated 0.5 M H₂SO₄ at a scan rate of 10 mV s⁻¹ with a rotation speed of 1600 rpm. (b) Hydrogen peroxide percentages at fixed potentials of 0.15 to 0.4 V vs. RHE. (c) Electron transfer numbers at fixed potentials of 0.10 to 0.40 V vs. RHE.

Remarkably, resistance to methanol crossover and stability are significant issues involved in estimating the performance of the catalysts. Firstly, it is shown in Fig. 7a that the i–t curves of Pt/C and 3D Fe₅C₂/N-PG-1.0-800 maintain the original downward trend after the addition of methanol, indicating that the impact of the methanol crossover effect on Pt/C and 3D Fe₅C₂/N-PG-1.0-800 is negligible in acid solution. Both electrocatalysts show similar methanol tolerance. Moreover, the stability characterization of the 3D Fe₅C₂/N-PG-1.0-800 and Pt/C catalysts is shown in Fig. 7b. As can be seen, after the 10 000 s test, the 3D Fe₅C₂/N-PG-1.0-800 catalyst activity loss is 16%, whereas the Pt/C catalyst only retains 50% of its initial current, confirming the much better stability of the 3D Fe₅C₂/N-PG-1.0-800 catalyst compared to that of commercial Pt/C.

Fig. 7 (a) Current–time (i–t) chronoamperometric responses for the 3D Fe₅C₂/N-PG-1.0-800 and Pt/C electrodes upon introduction of 3 M methanol after about 300 s at 0.52 V vs. RHE in O₂-saturated 0.5 M H₂SO₄ solution at a rotation rate of 1600 rpm. (b) i–t chronoamperometric responses at 0.52 V vs. RHE for 3D Fe₅C₂/N-PG-1.0-800 and Pt/C over 10 000 s in O₂-saturated 0.5 M H₂SO₄ solution at a rotation rate of 1600 rpm.

The electrocatalytic evaluation for ORR in alkaline medium

Fig. 8a shows the CV curves of 3D Fe₅C₂/N-PG-1.0-800 in 0.1 M KOH solution saturated with N₂ and O₂ at a scan rate of 10 mV s⁻¹. A prominent reduction peak is detected in the O₂ saturated solution but not in the N₂ saturated solution, demonstrating the ORR activity on 3D Fe₅C₂/N-PG-1.0-800. In Fig. 8b and c, 3D Fe₅C₂/N-PG-1.0-800 displays superior activity to all the other samples, even including commercial Pt/C catalyst, in terms of onset potential and half-wave potential, and the diffusion-limiting current density is almost unchanged on 3D Fe₅C₂/N-PG-h-800. In addition, the electrochemical parameters for ORR are displayed in Table 3. The LSVs and K–L curves of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C, shown in Fig. 8d and e, confirm the first-order reaction kinetics for ORR with regard to the concentration of dissolved O₂ in solution as well as the similar electron transfer numbers, respectively. Moreover, as can be seen in Fig. 8f–h, the higher kinetic current density and the similar Tafel slope of Fe₅C₂/N-PG-1.0-800 suggest better activity than the Pt/C catalyst in alkaline conditions, which agrees well with the LSV test results. The disk and ring current density on 3D Fe₅C₂/N-PG-1.0-800 and Pt/C are exhibited in Fig. 9a. The H₂O₂ (%) yield and the electron transfer number for 3D Fe₅C₂/N-PG-1.0-800, shown in Fig. 9b and c, are close to 5.5% and 4.0, respectively. Moreover, i–t measurements, as shown in Fig. 9d and e, demonstrate the excellent methanol tolerance and durability of 3D Fe₅C₂/N-PG-1.0-800 in alkaline solution. The MA of 3D Fe₅C₂/N-RGO-800, 3D Fe₅C₂/N-PG-h-800 and the commercial Pt/C catalyst at 0.8 vs. RHE in O₂-saturated 0.1 M KOH are given in Fig. S3.† The MA of Fe₅C₂/N-PG-1.0-800 is 47.04 mA mg⁻¹, which is significantly higher than those of Fe₅C₂/N-RGO-800 (30.18 mA mg⁻¹), 3D Fe₅C₂/N-PG-0.5-800 (20.32 mA mg⁻¹), 3D Fe₅C₂/N-PG-2.0-800 (28.86 mA mg⁻¹) and Pt/C catalyst (23.14 mA mg⁻¹), suggesting that Fe₅C₂/N-PG-1.0-800 demonstrated better performance than commercial Pt/C.

Fig. 8 CV curves of 3D Fe₅C₂/N-PG-1.0-800 at a scan rate of 10 mV s⁻¹ in 0.1 M KOH solution saturated with N₂ and O₂ (a). LSV curves for the 3D Fe₅C₂/N-PG-h-800, Fe₅C₂/N-RGO-800 and Pt/C samples at 1600 rpm (b), 3D Fe₅C₂/N-PG-1.0-T at 1600 rpm (c), 3D Fe₅C₂/N-PG-1.0-800 at different rotation rates (d) and Pt/C at different rotation rates (e) in O₂-saturated 0.1 M KOH at 10 mV s⁻¹. The insets in (d) and (e) show the corresponding K–L plots. The kinetic current densities of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C (f) at different potentials in O₂-saturated 0.1 M KOH. Tafel plots of 3D Fe₅C₂/N-PG-1.0-800 (g) and Pt/C (h) in O₂-saturated 0.1 M KOH solution.

Table 3 Electrochemical parameters for ORR estimated from LSV in O₂-saturated 0.1 M KOH at a scan rate of 10 mV s⁻¹ with a rotation speed of 1600 rpm

Material	Eonset (V)	E1/2 (V)	JL (mA cm^−2)	Electrolyte solution	Reference electrode	Reference
3D Fe5C2/N-PG-0.5-800	1.00	0.84	6.23	0.1 M KOH	RHE	In this work
3D Fe5C2/N-PG-1.0-800	1.01	0.87	6.12	0.1 M KOH	RHE	In this work
3D Fe5C2/N-PG-2.0-800	0.99	0.82	6.29	0.1 M KOH	RHE	In this work
3D Fe5C2/N-RGO	1.00	0.85	6.10	0.1 M KOH	RHE	In this work
Pt/C	1.01	0.82	5.89	0.1 M KOH	RHE	In this work

---

Fig. 9 (a) Rotating ring disk electrode (RRDE) measurements of 3D Fe₅C₂/N-PG-1.0-800 and Pt/C in O₂-saturated 0.1 M KOH at a scan rate of 10 mV s⁻¹ with a rotation speed of 1600 rpm. (b) Hydrogen peroxide percentages at fixed potentials of 0.2 to 0.6 V vs. RHE. (c) Electron transfer numbers at fixed potentials of 0.2 to 0.6 V vs. RHE. (d) i–t chronoamperometric responses for 3D Fe₅C₂/N-PG-1.0-800 and Pt/C upon introduction of 3 M methanol after about 300 s at 0.72 V vs. RHE in O₂-saturated 0.1 M KOH solution at a rotation rate of 1600 rpm. (e) i–t chronoamperometric responses at 0.72 V vs. RHE for 3D Fe₅C₂/N-PG-1.0-800 and Pt/C over 10 000 s in O₂-saturated 0.1 M KOH solution at a rotation rate of 1600 rpm.

As discussed in the above sections, firstly, in both acidic and alkaline environments, the 3D Fe₅C₂/N-PG-1.0-800 catalyst possesses superior activity to the other 3D Fe₅C₂/N-PG-h-800 catalysts, which is attributed to the 3D hierarchically interconnected porosity, especially with abundant micro-pores and meso-pores, as well as the high surface area, which contributes to the diffusion of the electrolyte and reactants together with exposure of the active sites. The electrochemical test results confirm that the pore structures and surface area actually play vital roles in the electrocatalytic performance; furthermore, the 3D Fe₅C₂/N-PG-1.0-800 catalyst exhibits better activity than the other 3D Fe₅C₂/N-PG-1.0-T catalysts due to the balance of surface area, pore structure, defects and chemical composition (pyridinic N, graphitic N and Fe₅C₂ nanoparticles). Lastly, Fe₅C₂/N-PG-1.0-800 displays superior stability in both acidic and alkaline environments due to the excellent stability of graphene along with the encapsulation of iron carbide nanoparticles by graphitic carbon layers, which can protect the inner metal from corrosion in harsh conditions.

Conclusions

In summary, we have developed an effective fabrication strategy for synthesizing a 3D Fe₅C₂/N-PG-h-800 catalyst with tunable pore size distribution by acid oxidation, hard-template and pyrolysis methods. Based on the adsorption and desorption isotherms of 3D Fe₅C₂/N-PG-h-800, the density of micro-pores and meso-pores can be purposely tailored by adjusting the reaction time of PG. The electrocatalytic measurement results clearly confirm that, as expected, the electrocatalytic activity is highly related to the porosity structures, surface area and chemical structures. Benefitting from the 3D hierarchically interconnected porosity structure, high surface area and Fe₅C₂ coupled with N functionalities as catalytic active sites, the 3D Fe₅C₂/N-PG-1.0-800 catalyst exhibits superior activity, better methanol tolerance and excellent stability in both acidic and alkaline media, especially compared to Pt/C. This work provides a method to construct novel 3D graphene structures containing a higher density of micro/meso-pores, unlike the graphene aerogel materials (GAs), which mainly contain macro-pore structures. It also confirms that the controllable synthesis of hierarchically porous graphene materials with high surface area is a feasible strategy to promote ORR activity.

Acknowledgements

The work has been supported by the Natural Science Foundation of China (No. 21603017 and 21273024) and Natural Science Foundation of Jilin Province, China (No. 20160101298JC).

References

1. G. Wu, K. L. More, C. M. Johnston and P. Zelenay, Science, 2011, 332, 443–447.
2. B. C. H. Steele and A. Heinzel, Nature, 2001, 414, 345–352.
3. S. Y. Wang, S. P. Jiang, T. J. White, J. Guo and X. Wang, J. Phys. Chem. C, 2009, 113, 18935–18945.
4. B. Lim, M. Jiang, P. H. C. Camargo, E. C. Cho, J. Tao, X. Lu, Y. Zhu and Y. Xia, Science, 2009, 324, 1302–1305.
5. R. L. Liu, C. V. Malotki, L. Arnold, N. Koshino, H. Higashimura, M. Baumgarten and K. Müllen, J. Am. Chem. Soc., 2011, 133, 10372–10375.
6. J. Sanabria-Chinchilla, K. Asazawa, T. Sakamoto, K. Yamada, H. Tanaka and P. Strasser, J. Am. Chem. Soc., 2011, 133, 5425–5431.
7. X. Xu, L. Yang, S. Jiang, Z. Hu and S. Liu, Chem. Commun., 2011, 47, 7137–7139.
8. D. Yu, N. Enoch, N. Rajesh and L. Dai, Angew. Chem., Int. Ed., 2011, 50, 6575–6578.
9. D. Yu, Y. Yang, M. Durstock, J. B. Baek and L. Dai, ACS Nano, 2010, 4, 5633–5640.
10. X. W. Yang, J. W. Zhu, L. Qiu and D. Li, Adv. Mater., 2011, 23, 2833–2838.
11. Y. W. Zhu, S. Murali, W. W. Cai, X. S. Li, J. W. Suk, J. R. Potts and R. S. Ruoff, Adv. Mater., 2010, 22, 3906–3924.
12. H. Bai, C. Li and G. Q. Shi, Adv. Mater., 2011, 23, 1089–1115.
13. S. Stankovich, D. A. Dikin, G. H. B. Dommett, K. M. Kohlhaas, E. J. Zimney, E. A. Stach, R. D. Piner, S. T. Nguyen and R. S. Ruoff, Nature, 2006, 442, 282–286.
14. K. P. Gong, F. Du, Z. H. Xia, M. Durstock and L. M. Dai, Science, 2009, 323, 760–764.
15. X. W. Wang, G. Z. Sun, P. Routh, D.-H. Kim, W. Huang and P. Chen, Chem. Soc. Rev., 2014, 43, 7067–7098.
16. M. Jiang, X. P. Cao, D. D. Zhu, Y. X. Duan and J. M. Zhang, Electrochim. Acta, 2016, 196, 699–707.
17. Z. S. Wu, S. B. Yang, Y. Sun, K. Parvez, X. L. Feng and K. Müllen, J. Am. Chem. Soc., 2012, 134, 9082–9085.
18. X. Q. Yang, A. R. Liu, Y. W. Zhao, H. J. Lu, Y. J. Zhang, W. Wei, Y. Li and S. Q. Liu, ACS Appl. Mater. Interfaces, 2015, 7, 23731–23740.
19. R. W. Chen, J. Yan, Y. Liu and J. H. Li, J. Phys. Chem. C, 2015, 119, 8032–8037.
20. W. Lv, C. Zhang, Z. J. Li and Q.-H. Yang, J. Phys. Chem. Lett., 2015, 6, 658–668.
21. S. B. Liu, T. H. Zeng, M. Hofmann, E. Burcombe, J. Wei, R. R. Jiang, J. Kong and Y. Chen, ACS Nano, 2011, 5, 6971–6980.
22. X. Zhao, C. M. Hayner, M. C. Kung and H. H. Kung, ACS Nano, 2011, 5, 8739–8749.
23. D. H. Long, W. Li, W. M. Qiao, J. Miyawaki, S. H. Yoon, I. Mochida and L. C. Ling, Chem. Commun., 2011, 47, 9429–9431.
24. Y. Liu, Y. Y. Wu, G. J. Lv, T. Pu, X. Q. He and L. L. Cui, Electrochim. Acta, 2013, 112, 269–278.
25. D. S. Yu, L. Wei, W. C Jiang, H. Wang, B. Sun, Q. Zhang, K. L. Goh, R. M. Si and Y. Chen, Nanoscale, 2013, 5, 3457–3464.
26. G. Y. Zhong, H. J. Wang, H. Yu and F. Peng, J. Power Sources, 2015, 286, 495–503.
27. Y. Hu, J. O. Jensen, W. Zhang, S. Martin, R. Chenitz, C. Pan, W. Xing, N. J. Bjerruma and Q. F. Li, J. Mater. Chem. A, 2015, 3, 1752–1760.
28. M. L. Xiao, J. B. Zhu, L. G. Feng, C. P. Liu and W. Xing, Adv. Mater., 2015, 27, 2521–2527.
29. Y. J. Zhang, K. Fugane, T. Mori, L. Niu and J. H. Ye, J. Mater. Chem., 2012, 22, 6575–6580.
30. D. S. Yu, Y. H. Xue and L. M. Dai, J. Phys. Chem. Lett., 2012, 3, 2863–2870.
31. Y. Zhou, Q. L. Bao, L. A. L. Tang, Y. L. Zhong and K. P. Loh, Chem. Mater., 2009, 21, 2950–2956.
32. L. F. Lai, J. R. Potts, D. Zhan, L. Wang, C. K. Poh, C. Tang, H. Gong, Z. X. Shen, J. Y. Linand and R. S. Ruoff, Energy Environ. Sci., 2012, 5, 7936–7942.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16530c

---