Synthesis of Hollow N,P-Doped Carbon/Co2P2O7 Nanotubular Crystals as an Effective Electrocatalyst for the Oxygen Reduction Reaction.

Herein, N,P-rich carbon/carbon/Co2P2O7 hollow nanotubes with a multilayered wall structure were successfully fabricated for the ORR electrocatalyst. The hollow tube structure catalysts were obtained by carbonizing Co2P2O7/C coated with the phytate-doped PANI. The Co2P2O7/C was obtained by phosphorylating a basic cobalt carbonate with phytic acid (PA). Onset and positive half-wave potentials were measured at 0.90 and 0.84 V, respectively, with a diffusion-limited current density of 4.58 mA/cm2. Effect of the thickness of polyaniline (PANI) in the electrocatalyst precursor was also investigated. The specific surface area as well as the content of graphitic N altered as the time of PANI polymerization increased, resulting in remarkably different catalytic activities. This study of hollow nanotube catalysts exhibits efficient noble-metal-free oxygen reduction reaction electrocatalysts for other chemical systems, which will provide abundant electrochemical active centers and sufficient energy.

Introduction

Scientists are exploring new energy alternatives, such as solar energy and electrochemical energy, to combat environmental pollution and the current energy crisis.[1−3] In various electrochemical energy systems, fuel cells have received a lot of attention due to their high energy conversion efficiency, high specific energy, low environmental impact, and high reliability. Several factors limit the use of fuel cells on a large scale, including the poor reaction kinetics of the oxygen reduction reaction (ORR) on the cathode[4] and the use of noble metal catalysts.[4−6] Thus, to address the scarcity and high cost of noble-metal-based catalysts, nonmetallic electrocatalysts,[7,8] nonprecious metal catalysts,[9−12] and low noble metal catalysts[13,14] have therefore been investigated and inspected.

Among the various alternatives, transition-metal phosphates are regarded as potential alternatives due to their good electrical conductivity, low price, stable properties, and environmental friendliness. The performance of pyrophosphates of divalent metals with the general formula M2P2O7 (M1/4 Co, Ni, etc.) is expected to be a promising material of choice for fuel cells.[15] Heteroatom codoping (for example, boron, nitrogen, phosphorus, sulfur, and others) can considerably improve conductivity and functionality while disrupting the initial conjugated electron coordination environment and even achieve metallicity.[16−20] Taking Co-based materials as examples, most zeolite-imidazole framework (ZIF)-based precursors were used to develop active and nonvolatile bimetallic ZIF-derived catalysts to enhance the performance of oxygen reduction. Wang prepared CS Co@NC-700 and Co@Co4N/MnO-NC catalysts using Zn and Mn bimetallic MOFs as precursors, respectively, which not only enhanced the performance of ORR but also acted on Zn-ion batteries to improve their stability.[21,22]

As one of the Co–P materials, Co2P2O7 has superior performance in supercapacitors, magnetism, microwave absorption, and multiphase catalysis.[23,24] In general, Co2P2O7 is mostly used for testing OER mainly stemming from the fact that the valence electron density within the Co center of Co2P2O7 decreases significantly when the metal undergoes carbonization. However, the performance of Co2P2O7 in ORR catalytic performance was far from expectations. It has been reported that the catalytic performance of catalytic materials can be improved through structural design and composition optimization. Graphene nanocages doped with N,P and Co2P2O7 act as Mott–Schottky heterojunction electrocatalysts to enhance their interfacial[25] and thus drive their intrinsic catalytic activity. The resulting N,P-doped carbon layer having a metallic nature not only regulates the overfilling of the Co center, for example, the orbitals occupied by the Co2P2O7 nanorods (NRs) as cocatalysts, but also ensures continuous and long-term operation.[26] However, the smart structure design and synthesis of heterojunction-based Co2P2O7 and N,P-codoped carbon nanostates for electrocatalysts with enhanced performance are problems that urgently need to be addressed.

Herein, hollow nanotube (HNTs) catalysts of N,P-rich carbon/carbon/Co2P2O7 (Co2P2O7/C@N,P–C) were successfully fabricated by combining the advantages of both heteroatom-doped carbon materials and transition-metal phosphides to realize the multifunctionality of the catalyst by carbonizing the PANI-coated Co2P2O7. The effect of PANI amount varying by the polymerization time in the catalyst precursor on the catalytic performance was also investigated. The as-prepared catalysts exhibited a much higher electrocatalytic performance and stability for ORR.

Experimental Section

The steps needed for electrocatalyst syntheses are depicted in Figure . Co2P2O7/C nanorods were fabricated first by PA-coated chemical Co(CO3)0.5(OH)·0.11H2O carbonization. The final N,P-rich carbon-coated Co2P2O7/C catalysts were produced in an in situ reaction through high-temperature carbonization and PANI polymerization.

Figure 1

Schematic illustration of the formation of Co2P2O7/C@N,P–C.

Synthesis of Co(CO3)0.5(OH)·0.11H2O

In 30 mL of deionized water (DI), dissolution of urea (0.0902 g, 1.5 mmol) and Co(NO3)2·6H2O (0.1455 g, 0.5 mmol) took place. After 1 h of continuous stirring, the formed light-pink solution was transferred to a stainless steel autoclave composed of poly(tetrafluoroethylene) and placed in an electric oven at 120 °C for 12 h. The resulting precipitates were then subjected to centrifugation followed by washing with DI water after cooling, and Co(CO3)0.5(OH)·0.11H2O was collected as a light-pink solid powder after drying in an oven at 60 °C.

Synthesis of Co2P2O7/C NRs

Co2P2O7/C nanorods were synthesized by dissolving 0.1 g of Co(CO3)0.5(OH)·0.11H2O prepared above in 10 mmol/L (50 mL) of PA solution and stirring at 60 °C for 1 h. The precipitate was then collected and dried before being annealed in an Ar atmosphere at 800 °C for 2 h at a heating rate of 5 °C/min from room temperature. The resultant samples were labeled Co2P2O7/C NRs.

Synthesis of N,P-Rich Carbon-Coated Co2P2O7/C Hollow Nanotubes (HNTs)

First, 0.1 g of Co2P2O7/C NRs was placed in a 50 mL PA 10 mM solution in a bilayer flask. The temperature of the flask was cooled to 0–5 °C by a recirculating cooling pump. Then, 0.1 mL of aniline monomer was injected into the solution. After that, 0.1 g of ammonium persulfate (APS) as an oxidant was sonicated in 50 mL of a 10 mM PA solution and dropped into the bilayer flask at a constant rate using a constant pressure drop funnel. The polymerization reaction was carried out for 3–5 h at 5 °C with constant stirring. The PANI-coated Co2P2O7 NR product (PANI-co-Co2P2O7/C) was filtered and rinsed multiple times with DI before drying in a vacuum at 60 °C. Subsequently, the PANI-co-Co2P2O7/C was annealed for 2 h at 800 °C in an Ar atmosphere at a heating rate of 5 °C/min. The resultant N,P-rich carbon-coated Co2P2O7/C samples were labeled Co2P2O7/C@N,P–C HNTs. For comparison, the catalyst of PANI-derived N,P-doped carbon was obtained by paralyzing the phytate-doped PANI free from the presence of Co2P2O7 at 800 °C for 2 h.

Material Characterizations

An X-ray diffraction (XRD) device of the Bruker D8-advance model was used at 40 mA and 40 kV with a 2θ range of 5–90° to analyze the XRD of the catalysts. The X-ray photon spectra (XPS) were collected using a JPS-9200 instrument supplemented with the radiation of Mg Ka. The Bunko–Keiki M30-TP-M setup was used to perform Raman spectroscopy with a polychromator using YVO4 532 nm laser for excitation. Investigation of the thermal stability of specimens was carried out through the Perkin Elmer Diamond TG/DTA Lab system’s thermogravimetry (TG). The specimens were heated up to 1000 °C at the rate of 10 °C/min under a constant N2 flow of 50 mL/min. Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) were performed under the JEOL-2010F and JEOL JSM-6510LA instruments, respectively. Porosity and specific surface area were calculated employing data from an ASAP-2000 device utilizing the Horvath–Kawazoe (HK) and Brunner–Emmet–Teller (BET) methods. The electrical conductivity measurements of the composite particles were performed using an ST2253 digital four-probe resistance tester (Suzhou Jingge Electronics Co., Ltd.) at room temperature.

Electrochemical Analysis

The electrocatalytic performance of the as-prepared samples was evaluated with cyclic voltammetry (CV) and linear sweep voltammetry (LSV) tests on a Gamry workstation connected with a rotating disc setup. Rotating discs and rotation ring-disk electrodes (RRDEs) (PINE) were used to assess the electrochemical catalytic performance. The electrolyte used in this experiment was 0.1 mol/L KOH solution, and the working electrode was a glassy carbon disc with a diameter of about 5 mm equipped with an electrocatalyst, while the reference and counter electrodes were an Ag/AgCl solution using 0.1 mol/L KOH and a platinum wire, respectively. A two-channel Gamry 1010E constant potential meter was used for the rotating disc ring device. For testing, a cleaned reaction cell was used and the electrolyte was passed through oxygen until oxygen saturation; the working electrode was polished and cleaned, and a drop of catalyst ink was applied. The catalyst ink was composed of 2 mg of catalysts, ethanol (400 μL), and Nafion (25 μL, 5 wt %). The working electrode had 0.2 mg/cm2 catalyst loading density. Also, the specific tests can be found in a previous work of our group.[27] The catalytic electrode was first activated by CV performed at 0–1.2 V (vs reversible hydrogen electrode (RHE)) at a 50 mV/s scanning rate in an O2-saturated 0.1 M KOH solution until a steady curve was observed. LSV was collected in the 0–1.2 V range (vs RHE) at 10 mV/s rate in the O2-saturated 0.1 M KOH at 100, 400, 900, 1225, 1600, 2000, 2025, and 2500 rpm. All ORR tests were carried out in the presence of a constant oxygen gas stream.

The way to prepare the RRED electrode is to apply a potential of 0.5 V to the Pt ring electrode. Based on the data obtained by RRED, the number of electrons transferred during ORR, n, and the H2O2 yield are determined by the following equations.where ID and IR are the disk and the ring currents, respectively, and N is the current collection efficiency (equal to 0.37).

All of the potential values were measured relative to the Ag/AgCl potential and were recalculated to the RHE scale using the Nernst equation below

Results and Discussion

Synthesis and Characterization of Co2P2O7/C@N,P–C HNTs

The hollowed Co2P2O7/C@N,P–C NTs were synthesized through carbonizing the PANI-coated Co2P2O7, and the Co2P2O7 was obtained by paralyzing the PA-surrounded Co(CO3)0.5(OH)·0.11H2O NRs, schematically illustrated in Figure . Specifically, the Co(CO3)0.5(OH)·0.11H2O clubs with widths of 100–200 nm were first synthesized from Co(NO3)2·6H2O and urea. Then, PA with a high P-content was allowed to coat on the surface of Co(CO3)0.5(OH)·0.11H2O with the help of electrostatic attraction. Subsequently, the protonated PA will react with Co(CO3)0.5(OH)·0.11H2O and form the coordination of Co2+ and phytate (Co-phytate). Co2P2O7/C was achieved after carbonization of phytate/Co(CO3)0.5(OH)·0.11H2O. Then, PANI was polymerized on the surface of Co2P2O7/C from the PA-containing solution. Finally, hollow-tube-structured Co2P2O7/C@N,P–C catalysts are achieved, after calcination, in which PANI acts as the C,N source, while PA acts as the P source.

Powder XRD was employed to ascertain the sample’s crystal structure and phase purity. The XRD outcomes in Figure a indicated that the products for the samples with the PA cladding pretreatment and PA cladding treatment followed by PANI were composed of Co2P2O7. Other than the peaks of diffraction related to Co2P2O7 (JCPDS No. 49-1091), no impurities could be found in the XRD pattern. The main diffraction and peaks at 29.6, 30.1, 35.3, 43.7, 49.3 and 58.3° could be assigned to the (012), (-3̅02), (130), (032), (-424) and (-1̅34) facets of the Co2P2O7 phase, respectively. Further, as shown in Figure a, the intensity of the Co2P2O7/C@N,P–C diffraction peak is significantly higher than that of Co2P2O7/C after the introduction of the conducting polymer. This indicates that the crystallinity of the nanotubule is improved. However, it is interesting that, when carbonizing the PANI-PA-coated Co-(CO3)0.5(OH)·0.11H2O without the PA cladding pretreatment, the calcination product was composed of Co2P2O7 and Co2P. The peaks around 40.7, 43.3, 52, and 54.1° are related to the (121), (211), (130), and (002) crystallographic planes of Co2P, respectively (JCPDS No. 54-0413). Further, we adopted the solution of bicarbonization to obtain the hollow-structured Co2P2O7/C@N,P–C NT catalysts.

Figure 2

(a) Patterns of XRD for Co2P2O7/C, Co2P2O7/C@N,P–C, and Co2P-Co2P2O7/C@N,P–C. (b) Raman spectra of Co2P2O7/C and Co2P2O7/C@N,P–C.

Figure b shows Raman spectra of Co2P2O7/C and Co2P2O7/C@N,P–C samples. The symmetric stretching vibrations vs PO3 were attributed to the weak peak at 1033 cm–1, while the POP bridge in Co2P2O7 was related to the peak at 674 cm–1,[28] confirming the formation of Co2P2O7. The Raman spectrum of the PANI-coated Co2P2O7/C is provided in Figure S1, in which the Raman shift at 578 cm–1 was allocated to Co2P2O7 and the other peaks were attributed to PANI, confirming the successful coating of PANI on Co2P2O7. After carbonization, as shown in Figure b, there are two distinct Raman peaks: the G-band at 1600 cm–1 and the D-band at 1350 cm–1,[29] exhibiting the coating of carbon layers. The D-band emerges as a result of the vibration of sp3 hybridized C and is thus typical of a graphitic plane that is somewhat disordered or flawed. The appearance of the G-band is due to the E2g vibrations of sp2 hybridized C and hence shows the degree of graphitization.[30,31] The intensity ratio (ID/IG) of these two peaks is often used to assess the degree of disorder or defects in graphitic materials.[32] Defects will change the charge distribution of the neighboring carbon, which may be favorable for electrochemical reactions.[33−35] The ID/IG of Co2P2O7/C @N,P–C at 1.13 is higher than that of the Co2P2O7/C catalyst (which is 1.05), suggesting a higher defect degree in Co2P2O7/C@N,P–C. As a result, the N-doped carbon coated on Co2P2O7/C has a lattice with several defects and disorders.[36]

The surface electronic states and composition of the synthesized catalysts were determined using XPS measurements, and the outcomes are illustrated in Figure . The results showed that Co, P, C, and O elements were present in both the Co2P2O7/C and Co2P2O7/C@N,P–C samples (see Figure a). The phosphorization of the phytate derivatives was successful because of the presence of P elements. The PANI resulted in the presence of N in Co2P2O7/C@N,P–C. Detailed information on the structure and composition of Co2P2O7/C@N,P–C is revealed by the Co 2p, P 2p, N 1s, and C 1s XPS spectra (Figure ). Co 2p3/2 and Co 2p1/2 peaks are situated at 781.3 and 797.4 eV, and 786.1 and 802.7 eV, respectively, in the Co 2p spectra of Co2P2O7/C, as depicted in Figure b. Compared to the Co2P2O7/C NRs, the corresponding peaks of Co2P2O7/C@N,P–C in the low-energy direction of Co 2p3/2 and Co 2p1/2 are shifted slightly (Figure b). This is attributed to the Co electron cloud migration due to the PANI surface-doped nitrogen species’ strong electronegativity, which indicates the covalent coupling between Co2P2O7/C and the PANI support.[13−15] There is a close association between these two components, making the catalysts more conductive and electrochemically active, which aims to enhance ORR activity.[37,38] As shown in Figure c, the C 1s spectrum of Co2P2O7/C@N,P–C contains four distinct peaks, one for aromatic, aliphatic, and graphitic C=C bonds at 284.1 eV, while the peaks at 284.4 and 285.6 eV represent C–C and C–N bonds, respectively, with one disordered peak at 288.1 eV. During the pyrolysis step, we observed carbon bonds (unsaturated and saturated) in samples, which indicates that PANI graphitization had been successful. In particular, compared to Co2P2O7/C, the C–N bond indicates the successful incorporation of N atoms of PANI into the carbon matrix.[39,40] The P 2p peak of Co2P2O7/C@N,P–C (Figure d) demonstrates three peaks at 133.1, 133.9, and 134.9 eV. The first two peaks correspond to the P 2p3/2 and P 2p1/2 phosphate group central phosphorus atom nuclear levels, whereas the latter (134.9 eV) is linked to the phosphorus-like bonding to the carbon lattice structure.[41−44] In the P 2p region of Co2P2O7/C, two distinct P species were introduced (Figure d). The first (133.5 eV) is because of P atoms chemically bound to O atoms, whereas the second (135.2 eV) is due to P–C coordination. The presence of P–O–P and P–C bonding is in agreement with the FTIR study in Figure S2, in which the vibrations of P–O–P and P–C in the infrared spectrum were attributed to the peaks at 765 and 1076 cm–1, respectively.[28]

Figure 3

(a) Survey of XPS, (b) Co 2p spectra of Co2P2O7/C and Co2P2O7/C@N,P–C samples. C 1s (c) and P 2p (d) spectra of Co2P2O7/C@N,P–C.

The morphology and structure of the synthesized catalysts were then observed with SEM and TEM. The results in Figure a–c show that the sample Co2P2O7/C is composed of carbon-coated wrapping nanobars (Co2P2O7) with a length of over 1 μm and a diameter of approximately 100 nm. However, after the pyrolyzation treatment of the PANI-co-Co2P2O7/C precursor, one-dimensional (1D) hollow nanotube catalysts were observed, as shown in Figure d,e. Interestingly, the hollow tube structure is more remarkably pronounced with an inner diameter of around 100 nm. This hollowed structure is very beneficial to the penetration of the electrolyte and the transfer of electrons/protons during electrocatalysis. Meanwhile, the rough surface added by the carbon shell provides the potential for efficient mass transport and ion diffusion.

Figure 4

(a–c) Images of SEM for the Co2P2O7/C nanorods. (d–f) Images of SEM for the Co2P2O7/C@N,P–C nanotubes.

The shape and microstructure of the catalysts, as well as the precursors, were further observed with TEM and high-resolution TEM (HRTEM). The precursor of Co(CO3)0.5(OH)·0.11H2O nanowires (solid) with a diameter of around 100 nm and different lengths of ∼μm can be observed in Figure S3. Then, after the PA was introduced into the solution of Co (CO3)0.5(OH)·0.11H2O, PA induced decomposition of Co (CO3)0.5(OH)·0.11H2O, thus resulting in the coordination of Co2+ with surface phytate (Co-phytate) with the help of electrostatic attraction. Figure S3c,d shows that an organic layer covered the porous substrates with a thickness of approximately 10 nm, indicating that PA was successfully covered on the surface of Co(CO3)0.5(OH)·0.11H2O. Figure S4 shows the TEM images of Co2P2O7/C, which were obtained by carbonization of the PA-treated Co (CO3)0.5(OH)·0.11H2O NRs. Interestingly, the rodlike Co2P2O7/C with a carbon shell can be seen in Figure S4, and further high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and element-mapping images in Figure S4 verify that the C and P elements are distributed around Co2P2O7.

Further, after the carbonization of the PANI-PA-coated Co2P2O7/C, as presented in the low-magnification TEM images in Figure a, the feature with hollowed nanotube walls was confirmed. The Co2P2O7 nanoparticles are wrapped by the hollowed PANI-derived N, P-doped carbon tube. The Co2P2O7 nanoparticles are inserted into the carbon skeleton when we magnify the wall (Figure b). The lattice fringes show that the 0.31 nm interlayer spacing corresponds to the (012) plane of Co2P2O7, which is in agreement with the XRD assessment. The hollow feature is further demonstrated in Figure d by the high-angle annular dark-field (HAADF) image and related EDS maps of Co2P2O7@N,P–C with uniform distributions of Co, P, and O elements in the inside and N, C elements in the exterior. These findings contribute to the development of a heterogeneous electrocatalyst comprising Co2P2O7 nanoparticles embedded in N,P-codoped tubular carbon. According to previous reports on carbon nanostructures wrapping metal or other inorganic nanoparticle hybrids, the introduced carbon layer can not only “armor” the Co2P2O7 cores and prevent them from being destroyed during electrochemical cycling but also enhance electron penetration or tunneling and facilitate electrocatalytic application.

Figure 5

(a, b) Images of TEM for the Co2P2O7/C@N,P–C nanocages; (c) image of HRTEM for Co2P2O7/C@N,P–C; (d) image of HAADF-STEM as well as (e–i) element-mapping images of a magnified branched region in Co2P2O7/C@N,P–C nanotubes.

N2 adsorption–desorption tests were employed to assess the specific surface area and porosity of the prepared materials. The BET specific surface area of Co2P2O7/C@N,P–C was found to be 459.3 m2/g, which was greater than that of Co2P2O7/C (104.7 m2/g, Figure ). The formation of the 1D hollowed tube structure obviously promotes the BET specific surface area. The hierarchical porous structure is noted to have an average pore diameter of 2.63 nm in Co2P2O7/C@N,P–C. Meanwhile, the large pore with the diameter close to 80 nm in the pore distribution curve of Figure b mainly corresponded to the inner diameter of the formed hollow Co2P2O7/C@N,P–C tubes. The porous carbon derived from the carbonized PANI matrix as well as the 1D hollowed tube structure was responsible for the increased specific surface area. The hierarchical porous structure develops a specific surface area, which provides numerous energetic spots for the catalytic reaction; meanwhile, porosity facilitates the transfer of reactants, promotes the occurrence of catalytic response, and thus improves the catalytic ability.

Figure 6

(a)Adsorption–desorption isotherms for N2 for Co2P2O7/C@N,P–C and Co2P2O7/C; (b) pore diameter distribution diagram.

Electrochemical Activity of Co2P2O7/C@N,P–C HNTs

To study the electrochemical properties, a three-electrode system was used in an aqueous 0.1 M KOH. The voltage range was 0–0.5 V (relative to RHE). Figure a illustrates the curves of CV for Co2P2O7/C, Co2P2O7/C@N,P–C, PANI-derived N,P–C, and commercial 20% Pt/C at a scan rate of 50 mV/s. The area around the ring of the CV curve for Co2P2O7/C@N,P–C is larger than that of the Co2P2O7/C electrode, exhibiting better capacitive performance. LSV tests of Co2P2O7/C@N,P–C conducted at 1600 rpm exhibited a diffusion-limited current density of 4.6 mA/cm2 at 0.2 V (Figure c), which is substantially greater than that of Co2P2O7/C at 2.23 mA/cm2 and PANI-derived N,P–C at 3.46 cm–2. A more positive onset potential was observed for Co2P2O7/C @N,P–C (equal to 0.91 V) than that of Co2P2O7/C at 0.6 V and PANI-derived N,P–C at 0.84 V. The half-wave potential of Co2P2O7/C@N,P–C is 0.85 V, and the value is only 50 mV lower than commercial 20 wt % Pt/C catalysts. The ORR pathway was computed, and the OH2– ORR yield was screened using RRDE. During ORR with the Co2P2O7/C@N,P–C catalyst, the OH2– yield was 1.0% (see Figure c), and the value of n (electron transfer number) was observed in the range of 3.97–3.98, which is quite close to the 20 wt % Pt/C catalyst value. However, the n values and OH2– yield attained for the Co2P2O7/C catalyst-assisted ORR were substantially different from those achieved for the reactions assisted by the 20% Pt/C and Co2P2O7/C@N,P–C materials, which demonstrates a four-/two-electron mixed transfer pathway.

Figure 7

(a) Comparison of the CV of different Co2P2O7/C, Co2P2O7/C@N,P–C,Co2P-Co2P2O7/C@N,P–C, and 20% Pt/C. (b) LSV curves observed in 0.1 M KOH (O2-saturated) at 20 and 50 mV/s sweep rates. (c) Curves of RRDE for 20 wt % Pt/C, Co2P2O7/C@N, P–C, Co2P2O7/C, and PANI-derived N,P–C catalysts. Densities of the ring (upper) and disc (bottom) currents observed at 1600 rpm. (d) Calculated electron transfer numbers n for the RRDE data for Co2P2O7/C@N,P–C, Co2P2O7/C, PANI-derived N,P–C, and 20% Pt/C. (e) Half-wave (blue) and onset (green) potentials together with the restricted current density (gray) values for Co2P2O7 /C, Co2P2O7/C@N,P–C, Co2P-Co2P2O7/C@N,P–C, PANI-derived N,P–C, and 20% Pt/C catalysts.

In Table , we compare the ORR ability of the lately reported Co2P2O7-based catalyst with our catalysts. Gratifyingly, the catalyst here reported exhibits a superior performance no matter the limiting current density and E1/2. These findings indicate that the Co2P2O7/C@N,P–C materials exhibit an outstanding ORR catalytic activity, which may result from the special hollow tube nanohybrid structure. A synergetic effect of individual components leads to enhancing the ORR movement. In terms of electrocatalytic performance results (Figure b), the performance of the composite ternary catalyst is much higher than the performance of PANI-derived N,P-doped C and Co2P2O7/C. The Co2P2O7 NRs obtained from the Co(CO3)0.5(OH)·0.11H2O precursors are more likely to serve as templates for the fabrication of N,P–C, instead of taking effect in the ORR process. The promising ORR activities of the Co2P2O7/C@N,P–C catalysts also highlight the advantages of the exposure of more active sites given by the well-defined 1D hollow nanotube structure and the synergetic interactions between Co2P2O7/C and PANI-derived N,P–C in favor of the absorption/desorption of the oxygenated species during the electrocatalytic processes. Moreover, the N,P-doped carbon enhances the specific surface area; thus, the resulting abundant pores and open tunnels permit the facile access of O2 bubbles and electrolytes toward the active sites to accelerate the ORR process. Further, there also is a considerable electronic coupling effect in the Co2P2O7/C and PANI-derived N,P–C components, and electrons can be transferred into the thin C shells from Co2P2O7 cores. The conductivity of the catalyst material obtained by four-probe methods was much improved due to the wrapping of PANI-derived N,P–C, with σCo2P2O7/C = 0.468 S/cm up to σCo2P2O7/C@N,P-C = 1.87 S/cm. The thin C shells, in other words, can improve the interfacial electron or electron penetration transfer. Additionally, the pyridinic and graphitic nitrogen atoms with their sp2 electronic structures are very active ORR sites.[8,45,46] They enhance the material electronic conductivity and increase the catalyst corrosion resistance, which, again, can effectually advance the movement and constancy of the catalysts.

Table 1

ORR Electrocatalytic Activity Recently Reported for Nonprecious Metal-Based Catalyst Containing Co

	catalyst	electrolyte	onset potential (V)	half-wave potential (V)	limiting current density (mA/cm2)	refs
1	N–C@CoP	1 M KOH	0.85 V	0.68 V	4.48 mA/cm2	(47)
2	CoP/NP-HPC	0.1 M KOH	0.95 V	0.83 V	5.2 mA/cm2	(48)
3	NC-CoP	0.1 M KOH	0.82 V	0.69 V	5.2 mA/cm2	(49)
4	CoP-DC	0.1 M KOH	N.A.	0.81 V	N.A.	(50)
5	Co/CoP-HNC	0.1 M KOH	0.93 V	0.83 V	N.A.	(51)
6	CoP@SNC	1 M KOH	0.87 V	0.79 V	4.8 mA/cm2	(52)
7	CoP-PBSCF	0.1 M KOH	N.A.	0.752 V	N.A.	(53)
8	CoP@PNC-DoS	0.1 M KOH	0.94 V	0.803 V	N.A.	(54)
9	NPMCNT-300	0.1 M KOH	0.93 V	N.A.	N.A.	(55)
10	CoP@C	0.1 M KOH	0.91 V	0.87 V	4.2 mA/cm2	(56)
11	CoPi/NPGA	0.1 M KOH	0.91 V	0.80	5.1 mA/cm2	(57)
12	CoP@PNC-DoS	0.1 M KOH	0.94 V	0.803	N.A.	(54)
13	Co2P2O7/N-rGO-800 (1800 rpm)	0.1 M KOH	0.9 V	0.8	4.7 mA/cm2	(58)
14	Co2P2O7/C@N,P–C	0.1 M KOH	0.9 V	0.84 V	4.6 mA/cm2	this work

From the above discussion, it is easy to see that the presence of the pyrolytic N,P-doped carbon layer derived from the PANI layer plays an important role in the ORR catalytic performance. A huge improvement of ORR over the original Co2P2O7/C was observed. Thus, the effects of polymerized PANI amounts controlled by the polymerization time on the catalyst performance were then investigated.

Figure compares the curves of CV and LSV for Co2P2O7/C@N,P-C with different PANI polymerization times during the synthesis process of their carbonized precursors. A distinct cathodic peak can be observed for all of the Co2P2O7C@N,P–C samples, which corresponds to the featureless CV curves in O2-saturated solutions, and the catalytic performance is highly dependent on the quantity of PANI loaded during catalyst preparation. Further, the ORR onset potential of the Co2P2O7/C@N,P–C catalyst significantly shifts negatively, and the oxygen reduction current peak dramatically reduced with the increase in polymerization time of PANI. The Co2P2O7/C@N,P–C-3h catalyst exhibits the highest overall ORR activity, which is close to that of 20 wt % Pt/C in Figure . LSV curves recorded at 1600 rpm show that Co2P2O7/C@N,P–C-3h has a higher onset potential (Eonset) of 0.90 V vs RHE and a half-wave potential (E1/2) of 0.84 V than that of Co2P2O7/C@N,P–C-4h and Co2P2O7/C@N,P–C 5h, separately (Figure b). The polymerization time in the catalyst precursor had a significant influence on the OH2– yield and the electron transfer number n attained for the ORR assisted by the Co2P2O7/C@N,P–C catalyst. During ORR with the Co2P2O7/C@N,P–C-3h catalyst, the OH2– yield was 1.0% (see Figure c), and the electron transfer number n was in the 3.97–3.98 range, which is extremely similar to the value of the 20 wt % Pt/C catalyst. Moreover, increasing the polymerization duration of PANI in the corresponding precursors resulted in OH2– yields and n that were not the same as the ones found in reactions facilitated by 20% Pt/C and Co2P2O7C@N,P–C-3h materials, which suggests a two-electron transfer pathway. The weaker presentation of the other catalysts prepared with a longer polymerization time could be explained by the formation of the thicker N-doped carbon, which is highly confirmed by the following results of XPS and BET measurements.

Figure 8

(a) Typical curves of CV for the Co2P2O7/C @N,P–C pyrolyzed at different polymerization times in a solution of O2-saturated 0.1 M KOH with a scan rate of 5 mV/s. (b) LSV recorded in O2-saturated 0.1 M KOH at 10 and 50 mV/s sweep rates. (c) Curves of RRDE for 20 wt % Pt/C, Co2P2O7/C@N,P–C-3h, Co2P2O7/C@N,P–C-4h, and Co2P2O7/C@N,P–C-5h catalysts. The ring current densities (upper) and disk current densities (bottom) were achieved at 1600 rpm. (d) Evaluated electron transfer numbers n for the RRDE data for Co2P2O7/C@N,P–C-3h, Co2P2O7/C@N,P–C-4h, Co2P2O7/C@N,P–C-5h, and 20% Pt/C.

First, the N1s core-level XPS spectra of the synthesized catalysts were shown in Figure , and four types of N bonds were divided. The composed ratio of the different type N was summarized in Table . Based on the XPS results, it was noticed that the different N species ratios in the catalysts were strongly influenced by the varied polymerization times of PANI in the catalytic precious that leads to the variety of loaded outside N–C thickness after pyrolysis treatment (see Figure and Table ). As shown in Table , as the polymerization time in the precursor of the Co2P2O7/C@N,P-C catalyst enhanced, the total content of graphite N and pyridine N decreased. Co2P2O7/C@N,P–C-3h has the greatest total content of graphite N and pyridine N, which is 68.15%. It has been shown that pyridine nitrogen and graphite nitrogen are involved in ORR as active sites, with pyridine nitrogen possessing a pair of lone electrons that can weaken the O–O bond of O2; thus, it enhanced oxygen adsorption and exhibited better catalytic performance.[8,45,46,59] It can be deduced that the synergistic effects between Co2P2O7/C and PANI-derived N,P–C with the active N species (pyridine N, and graphitic N) can effectively reduce the activation barrier of the adsorbed O2 to promote the capture of the first electron to enhance the cleavage of O–O bonds, thereby improving the ORR activity.[60]

Figure 9

XPS High-resolution N 1s spectra for (a) Co2P2O7/C@N,P–C-3h, (b) Co2P2O7/C@N,P–C-4h, and (c) Co2P2O7/C@N,P–C-5h.

Table 2

Weight Content Prepared at T = 3, 4, and 5 of Pyridinic, Pyrrolic, Oxidized, and Graphitic Nitrogen in the Co2P2O7/C@N,P–C-H Catalysts

sample	graphitic-N/Ntotal(%)	pyrrolic-N/Ntotal(%)	pyridinc-N/Ntotal(%)	oxidized-N/Ntotal(%)	graphitic-N + pyridinic-N contents/Ntotal(%)
Co2P2O7/C@N,P–C-3h	51.64	21.45	16.51	10.4	68.15
Co2P2O7/C@N,P–C-4h	46.88	17.7	5.36	29.06	52.24
Co2P2O7/C@N,P–C-5h	21.08	32.22	10.56	20.8	41.88

Further, we can see that the specific surface area was also significantly affected through the PANI polarization time. All of the N2 adsorption–desorption isotherms exhibit the II-type curve in Figure , and the BET surface area is compiled in Table . The Co2P2O7/C@N,P–C-3h catalyst possesses a surface area of 459.30 m2/g, which is approximately twice that of the Co2P2O7/C@N,P–C-4h catalyst, which possesses a surface area of 237.71 m2/g. The hierarchical porous structure is noticed with an average pore diameter of 4.63 nm in Co2P2O7/C@N,P–C. The hierarchical porous structure develops a specific surface area, which provides numerous energetic spots for the catalytic reaction; meanwhile, porosity facilitates the transfer of reactants, promotes the occurrence of catalytic response, and thus improves the catalytic ability (Figures and 11).

Table 3

BET Surface for Co2P2O7/C@N,P-C HNTs

	BET surface (m2/g)	total pore volume (cm3/g)	average pore diameter (nm)
Co2P2O7/C@N,P–C-3h	459.30	0.65	4.63
Co2P2O7/C@N,P–C-4h	237.71	0.46	4.52
Co2P2O7/C@N,P–C-5h	203.40	0.15	4.58

Figure 10

(a) N2 adsorption–desorption isotherms for Co2P2O7/C@N,P–C-3h, Co2P2O7/C@N,P–C-4h, and Co2P2O7/C@N,P–C-5h and (b) the corresponding pore diameter distribution diagram.

Figure 11

Nyquist plot of Co2P2O7/C@N,P–C-3h, Co2P2O7/C@N,P–C-4h, and Co2P2O7/C@N,P–C-5h catalysts under 0.1 M KOH conditions.

Electrochemical impedance spectroscopy (EIS) measurement Nyquist diagram is one of the effective methods for measuring electrochemically active sites. The Nyquist diagram shows that the slope of the Co2P2O7/C@N,P–C-3h catalyst is larger than that of other samples, indicating faster kinetics of charge transfer and an increase in the ion diffusion rate.

We also compared the long-term electrochemical stability of Co2P2O7/C@N,P–C and 20% Pt/C catalysts. The i–t response of Co2P2O7/C@N,P–C and Pt/C recorded for 8000 s at 500 rpm is shown in Figure . The Co2P2O7/C@N,P–C catalyst exhibited better durability than the commercial Pt/C catalyst: only 9.1% of the activity was lost after 8000 s of a continuous ORR. For comparison, the commercial Pt/C catalyst demonstrated 7.7% lower activity after 8000 s.

Figure 12

Long-term stability tests performed by the cathodic current–time (i–t) method using Co2P2O7/C@N,P–C and 20% Pt/C catalyst performance during ORR.

Conclusions

To summarize, a simple hydrothermal technique and a subsequent pyrolysis strategy were used to synthesize N,P-rich carbon/carbon Co2P2O7 (Co2P2O7/C@N,P–C) tubular composite materials. The obtained Co2P2O7/C@N,P–C demonstrated a high catalytic activity toward the ORR due to the special hollowed structure and synergistic effect of the Co2P2O7 and PANI-derived N,P-doped C. The onset and positive half-wave potentials were recorded at 0.90 and 0.84 V, respectively, while the diffusion-limited current density was 4.6 mA/cm2, values that were similar to those of commercial 20% Pt/C. The Co2P2O7/C@N,P–C catalyst also has excellent stability compared to 20% Pt/C. It was demonstrated that the specific surface area, as well as the content of graphitic N, changed with the time of PANI polymerization, resulting in a significantly different catalytic performance, and an optimal value was achieved with 3 h of PANI polymerization in its precursor. This investigation of hollow tube metal phosphate-based materials reveals significant potential for energy storage regeneration and conversion technologies.