Co₃O₄ and CDots nanocrystals on g-C₃N₄ as a synergetic catalyst for oxygen reduction reaction

2.1 Synthesis of free Co₃O₄, CDots, C₃N₄, Co₃O₄-C, Co₃O₄-C₃N₄, C/C₃N₄and Co₃O₄-C/C₃N₄ catalysts

Two graphite rods were inserted parallelly into the ultrapure water (18.4 MΩ/cm) as the anode and cathode, respectively. Static potential of 25–30 V was applied to the two electrodes using a direct current power supply. After 140 h of continuous stirring, the anode was eroded, and a dark yellow solution appeared gradually in the reactor. CDots were obtained by filtering the solution with slow speed quantitative filter paper followed by high speed centrifugation (60,000 rpm). g-C₃N₄ sheets were fabricated by heating urea at 0.5°C/min to 550°C and then holding for 9 h. The as-produced g-C₃N₄ sheets were dispersed uniformly in anhydrous ethanol (EtOH) by stirring continually for 40 min, and the concentration of the final C₃N₄ solution was about 0.3 mg/ml. To prepare the Co₃O₄-C/C₃N₄ hybrid, the first step reaction mixture was prepared by adding 1.2 ml of 0.2 m Co(Ac)₂ aqueous solution to 25 ml of as-prepared C3N4 EtOH suspension, followed by addition 1 ml of H₂O at room temperature. The reaction was kept at 80°C with stirring continually for 2 h. After that, 1 ml condensed CDots aqueous solution was added dropwise to the above mixture whilst stirring, and then the reaction mixture was transferred into a 40 ml autoclave for solvothermal reactions at 150°C for 8 h. The obtained product was collected by high speed centrifugation (25,000 rpm) and washed several times with ethanol and water. The same steps were applied for the preparation of Co₃O₄-C, Co₃O₄-C₃N₄, and C/C₃N₄ and free Co₃O₄ nanopowders without the addition of g-C₃N₄, CDots, Co(Ac) ₂ or g-C₃N₄ and CDots to the reaction mixture, respectively. The schematic illustration of the preparation of Co₃O₄-C/C₃N₄ catalysts is shown in Figure 1.

Figure 1:

Schematic illustration of the preparation of Co₃O₄-C/C₃N₄ sheets catalysts and their catalytic ability for oxygen reduction reaction (ORR).

2.2 Characterizations

Transmission electron microscopy (TEM) and high resolution TEM (HRTEM) images were recorded on an FEI Tecnai G² F20 with an acceleration voltage of 200 kV. X-ray photoelectron spectroscopy (XPS) measurement was performed on a KRATOS Axis ultra-DLD X-ray photoelectron spectrometer using an Mg Kα radiation excitation source (hv=1283.3 eV). The phase component of the synthesized catalysts was characterized using X-ray diffraction (XRD) (PW3040/60) with Cu Kα radiation by comparing the peak positions and the intensities with those from the Joint Committee on Powder Diffraction Standards (JCPDS) databases.

2.3 ORR performance measurement

Electrochemical measurements, including cyclic voltammetry (CV) and linear sweep voltammograms measurements, were performed in a standard three-electrode electrochemical cell using a CHI 760E workstation (CH Instruments, Shanghai, China) equipped with a rotating disk and ring-disk electrode apparatus (RRDE-3A, ALS, Co. Ltd., Japan) at room temperature. A glass carbon electrode loaded with the electrocatalyst, a saturated calomel electrode (SCE) and platinum wire were used as the working electrode, reference electrode and counter electrode, respectively. For the preparation of the working electrode, ~6 μl of catalyst suspension (~4 mg/ml) was loaded on a glassy carbon electrode of 3 mm in diameter, followed by coating with 5 μl of Nafion solution (0.5 mass %) and dried at ambient conditions. CV measurements were performed in Ar- and O₂-saturated 0.1 m KOH solution at a scanning rate of 50 mV/s. Rotating-disk electrode (RDE) measurements were conducted in the O₂-saturated 0.1 m KOH solution at a scanning rate of 5 mV/s by varying the rotating speed from 600 to 2000 rpm. Kinetic parameters were analyzed on the basis of the Koutechy-Levich (K-L) equations.

3.1 Physical characterization of Co₃O₄-C/C₃N₄ catalysts

Figure 2A–D present typical TEM images and XRD analysis of the synthesized catalysts. The Co₃O₄-C/C₃N₄ sample displays typical porous structures (Figure 2A). Also, the Co₃O₄ and CDots nanoparticles with the size <10 nm are dispersed uniformly on the surface of the C₃N₄ sheets shown in the inset of Figure 2A. The synthesized Co₃O₄-C/C₃N₄ sample exhibits the combination of two-dimensional sheets and nanoparticles, making it possess a larger surface area and more active sites (Figure 2B). The inset of Figure 2B displays the diffraction rings of Co₃O₄ marked in the red dash rectangle, confirming the formation of a spinel structure of Co₃O₄ nanocrystals. HRTEM of Co₃O₄-C/C₃N₄ marked in the yellow dash rectangle also shows the presence of the crystalline CDots, identified according to the calculated lattice spacing of 0.321 nm shown in Figure 2C [20]. The XRD pattern of Co₃O₄-C/C₃N₄ sheets (Figure 2D) shows a diffraction peak at 2θ of about 25°, indicating the combination of the C₃N₄ and CDots (compare the XRD of C₃N₄, CDots and Co₃O₄-C/C₃N₄ sheet). All the other diffraction peaks are ascribed to the Co₃O₄ phase (JCPDS no. 42–1467).

Figure 2:

(A–C) Transmission electron microscopy (TEM) images and (D) X-ray diffraction (XRD) analysis of the carbon nanodots (CDots), free Co₃O₄, C₃N₄ and synthesized Co₃O₄-C/C₃N₄ sheets. The inset in (A) is the magnified lamellar structure of the Co₃O₄-C/C₃N₄ sheets, the inset in (B) is the diffraction rings of Co₃O₄ marked in the red dash rectangle in (B), and (C) is the high resolution TEM (HRTEM) image of crystalline CDots marked in the blue dash rectangle in (B).

As the carbon nanomaterials with higher atomic ratios of N and C have been documented to exhibit much better electrocatalytic activity, XPS measurements were performed to probe the element chemical states and the atomic ratios of N and C in the Co₃O₄-C/C₃N₄ sheet (Figure 3). In order to identify more active sites obtained from the implantation of g-C₃N₄, nitrogen-doped CDots were also used to compare the atomic ratio of N and C. It should be noted that the nitrogen-doped CDots catalyst (N/C) was also prepared by the same solvothermal process. Generally, the reported atomic ratio of N and C in nitrogen-doped carbon materials was only around 4% [21, 22]. The survey scan spectrum from XPS analysis for the Co₃O₄-C/C₃N₄ sheet revealed the existence of C 1s, O 1s, N 1s and Co 2p after the hydrothermal treatment (Figure 3A). The atomic ratio of N and C in the synthesized Co₃O₄-C/C₃N₄ sheet was calculated to be about 17.2% from the peak areas of C 1s and N 1s, whereas that of the nitrogen-doped CDots was only 13.4%. Thus, the atomic ratio of N and C in the present Co₃O₄-C/C₃N₄ sheet was significantly enhanced by the implantation of active sites. The high-resolution XPS spectrum of C 1s can be fitted with three different peaks located at 284.6 eV, 285.8 eV and 288.5 eV, corresponding to C-C, C-O and C-N, respectively (Figure 3B) [23]. The existence of C-O bonding could be attributed to the moisture, atmospheric O₂, or CO₂ adsorbed on Co₃O₄-C/C₃N₄ sheets, as well as the residual oxygen-containing groups, such as carboxyl groups existing in Co(Ac) ₂ precursors. It has been reported previously that a high content of oxygen in carbon nanomaterials can lead to a strong ability for O₂ adsorption [24], which might be another advantage for Co₃O₄-C/C₃N₄ sheets when being used as the electrocatalyst for oxygen reduction. The C-N bonding could be regarded as the new combination between C₃N₄ and CDots. Comparing the featureless C-N bonding observed in C 1s of the N/C catalyst (inset in Figure 3B) and that of the Co₃O₄-C/C₃N₄ sheet, it is found that the CDots were partly bonded with the nitrogen atom after the hydrothermal process. The detailed N 1s spectrum of the Co₃O₄-C/C₃N₄ sheet ranging from 390 eV to 410 eV is also shown in Figure 3C. It can be found that there are three different N groups in the XPS spectrum of the Co₃O₄-C/C₃N₄ sheet with binding energies of 401 eV, 399.8 eV and 398.6 eV, which can be attributed to sp³-hybridized nitrogen (N–[C]₃) (graphite N) and amino functional groups with a hydrogen atom (C-NH) (pyrrolic N), sp²-hybridized nitrogen (C-N-C) (pyridinic N), respectively. Further exploration found that the content of pyrrolic N in the Co₃O₄-C/C₃N₄ sheet was improved by the implantation of active sites (compare the high resolution N 1s spectrum of Co₃O₄-C/C₃N₄ sheets and that of N/C shown in the inset of Figure 3C). These analyses revealed the effective nitrogen transition in the sheets by the hydrothermal strategy [8, 25–28]. Apart from the C 1s and N 1s spectrum, the high resolution XPS spectrum of Co 2p is also displayed in Figure 3D, in which three peaks are fitted with bonding energies of 781.35eV, 788.25eV and 797.5eV. These peaks are assigned to Co 2p^3/2, Co 2p^3/2 and Co 2p^1/2, corresponding to the partly doped Co coordinated with N (781.3 eV) and undoped oxidized Co (Co₃O₄) (788.2 and 797.6 eV), respectively.

Figure 3:

(A) Full X-ray photoelectron spectroscopy (XPS) scan of Co₃O₄-C/C₃N₄ and C/N, (B) high resolution of C 1s, (C) N 1s, and (D) Co 2p spectra of the synthesized catalysts.

3.2 ORR performances of Co₃O₄-C/C₃N₄ sheets

3.2.1 CV

To investigate the catalytic activities of the synthesized catalysts toward ORR, CV experiments are performed in an Ar and O₂-saturated 0.1 m KOH solution at a scanning rate of 50 mV/s. All potential values given here are in reference to the SCE. Figure 4 shows the CV curves of synthesized catalysts in the absence (Ar bubbling) and presence of oxygen (O₂bubbling) in 0.1 m KOH solution. As shown in Figure 4, all catalysts exhibit the ORR peaks in O₂-saturated solutions in contrast to the featureless peaks observed in Ar-saturated solutions within the same potential range (compare the dot lines and solid lines in Figure 4), confirming the electrocatalytic activity of the synthesized catalysts toward ORR in alkaline medium. It also notes that the Co₃O₄-C/C₃N₄ sheet exhibits a more positive ORR onset potential (-0.09 V vs. SCE) than the potentials of free Co₃O₄, C₃N₄, CDots, Co₃O₄-C₃N₄, Co₃O₄-C, and C/C₃N₄, indicating more active sites produced from the transition of nitrogen in C₃N₄ and the nanosized Co₃O₄.

Figure 4:

Cyclic voltammetry (CV) curves of free Co₃O₄, carbon nanodots (CDots), C₃N₄, Co₃O₄-C, Co₃O₄-C₃N₄ and Co₃O₄-C/C₃N₄ in (solid line) O₂-saturated or (dot line) Ar-saturated 0.1 m KOH solutions at a scan rate of 50 mV/s.

3.2.2 RDE results (linear sweep voltammograms)

In order to obtain more information into the kinetics of ORR, RDE measurements were performed to investigate the ORR activities of the synthesized samples, including the free Co₃O₄, CDots, C₃N₄, Co₃O₄-C₃N₄, Co₃O₄-CDots, Co₃O₄-C/C₃N₄ sheets and commercial 20 mass % Pt/C, in O₂-satuated 0.1 m KOH electrolyte at a scanning rate of 5 mV/s. As shown in Figure 5A, the free Co₃O₄, C₃N₄ and Co₃O₄-C₃N₄ exhibited the low ORR electrocatalytic activities due to their poor conductivity. Among these prepared catalysts, Co₃O₄-C/C₃N₄ sheets showed an enhanced ORR performance with more positive onset potential and larger current density, which are comparable to those of the commercial Pt/C. The desirable properties could be attributed to the synergistic effects of three materials and the combination of CDots and C₃N₄, effectively increasing electrical conductivity. During the ORR process, oxygen dissolved in the solution diffuses to the surface of the glass carbon electrode and reacts to form water through a four-electron pathway, or to produce hydrogen peroxide through a two-electron pathway. The four-electron pathway is preferred to the two-electron pathway toward ORR because the intermediate H₂O₂ produced in the two-electron pathway may erode the membrane and electrocatalyst which frustrate the efficiency of fuel cells [29]. Therefore, when evaluating the ORR activity of the Co₃O₄-C/C₃N₄ sheets, an RDE at various rotating speeds (600~2000 rpm) was employed to determine its electron-transferred numbers (Figure 5B). As shown in Figure 5B, the oxygen reduction current density increased with increasing rotation rate, suggesting the enhanced mass transfer effect of oxygen dissolved in the solution. The electron-transferred numbers/oxygen molecule could be calculated by the K-L equation:

Figure 5:

(A) Rotating-disk electrode (RDE) curves of samples (free Co₃O₄, carbon nanodots [CDots], C₃N₄, Co₃O₄-C, Co₃O₄-C₃N₄ and Co₃O₄- C/C₃N₄) in O₂-saturated 0.1 m KOH solution with a rotation speed of 1600 rpm and a sweep rate of 5 mV/s. (B) RDE voltammograms of the Co₃O₄-C/C₃N₄ in an O₂-saturated 0.1 m KOH solution at 5 mV/s and various rotation speeds. The inset in (B) is the Koutechy-Levich (K-L) plots (j^-1 vs. ω^-1/2) of Co₃O₄-C/C₃N₄ at potentials of -0.8 V, -0.9 V and -1 V. (C) K-L plots of a series of prepared catalysts at -0.8 V. (D) Electron transfer number given at -0.8 V for samples (a–g). a: Co₃O₄; b: C₃N₄; c: CDots; d: Co₃O₄-C₃N₄; f: Co₃O₄-C; and g: Co₃O₄-C/C₃N₄ sheet.

(1)1/j=1/Bω1/2+1/jk (1)

in which j is the measured current density and ω is the rotation speed expressed with rad/s; B is the Levich slope given by:

(2)B=0.62nFD2/3ν-1/6C (2)

where n is the overall number of transferred electrons in the ORR process, F is the Faraday constant (96,485 C/mol), D is the diffusion coefficient of oxygen, ν is the kinematic viscosity of the electrolyte and C is the saturated oxygen concentration in the electrolyte. The reported values for D, ν and C in 0.1 m KOH solution are 1.9×10^-5 cm²/s, 0.01 cm²/s and 1.2×10^-6mol/cm³, respectively [30]. The corresponding K-L plots (J^-1 vs. ω^-1/2) at various electrode potentials of -0.8, -0.9 and -1 V (vs. SCE) exhibit good linearity shown in the inset of Figure 5B. The linearity and parallelism of the K-L plots suggest first-order reaction kinetics toward the dissolved oxygen and similar electron transfer numbers for ORR at different potentials. Based on the slopes of K-L plots in the inset of Figure 5B, the electron- transferred number (n) is calculated to be about 3.69 at -0.8 V, indicating that Co₃O₄-C/C₃N₄ sheets favored a 4e^- oxygen reduction process. Figure 5C shows the K-L plots of a series of the controlled samples at -0.8 V according to their corresponding RDE curves (ESI, Supplementary Figures S1–S5). The electron-transferred numbers (n) of different catalysts are clearly drawn (Figure 5D) in terms of the results of K-L plots shown in Figure 5C, which are 1.82, 2.67, 3.42, 3.12, 3.48 and 3.69 corresponding to free Co₃O₄, C₃N₄, CDots, Co₃O₄-C₃N₄, Co₃O₄-CDots and Co₃O₄- C/C₃N₄ sheets, respectively.

3.2.3 Methanol-tolerant capability and stability of Co₃O₄-C/C₃N₄

The promising ORR catalysts could exhibit a high catalytic selectivity for cathode reaction against the oxidation of methanol because of the possible crossover effect of methanol through the polymer electrolyte membrane from anode to cathode in direct methanol fuel cells [31]. To examine the tolerance of the Co₃O₄-C/C₃N₄ catalyst to the methanol crossover effects, CV curves were carried out in an O₂-saturated 0.1 m KOH solution with and without the addition of 0.5 m methanol, respectively. It should be noted the volume ratio of KOH/methanol is 5. It was found that the Co₃O₄-C/C₃N₄catalyst exhibited a strong ability to avoid methanol crossover effects (Figure 6A). In contrast, for the Pt/C catalyst, after the addition of 0.5 m methanol into the O₂-saturated 0.1 m KOH electrolyte, the typical methanol oxidation peak at about -0.1 V was observed, while the cathodic peak for ORR disappeared (Figure 6B), indicating the competitive reaction between oxygen reduction and methanol oxidation.

Figure 6:

(A, B) Cyclic voltammograms of Co₃O₄-C/C₃N₄ and Pt/C (20%) in O₂-saturated 0.1 m KOH with or without 0.5 m methanol. (C) Current-time (i-t) curves of Co₃O₄-C/C₃N₄ and 20% Pt/C catalysts at -0.4 V (vs. SCE) in an O₂-saturated 0.1 m KOH solution for 10 h.

Moreover, the long-term durability of the Co₃O₄- C/C₃N₄ sheet is considered to be another important factor to evaluate ORR performance. To evaluate the durability of Co₃O₄-C/C₃N₄ catalysts, chronoamperometric (i-t) measurement was performed in O₂-saturated 0.1 m KOH electrolyte compared to that of the commercial Pt/C catalyst. As shown in Figure 6C, the continuous ORR process caused tardy attenuation (8.8%) of current for the Co₃O₄-C/C₃N₄ catalyst over 10 h, but the commercial Pt/C exhibited significant decline of stability within the same time, which merely kept 62.1% of the original current.

All of the above electrochemical results indicate that Co₃O₄-C/C₃N₄ catalysts display good catalytic ability compared to commercial Pt/C (20%), superior stability and outperformed methanol tolerance in an alkaline solution.