Introduction

Proton exchange membrane fuel cells are significant and high-effective environment-friendly energy conversion devices, capable of direct conversion of chemical energy to electric power by supporting chemical anodic (such as H2 or alcohols oxidation) and cathodic (e.g., O2 reduction) reactions [1, 2]. However, slow kinetics and high overpotentials throughout oxygen reduction reactions (ORRs) on the cathodes dramatically hinder their usages [3]. Currently, carbon-supported Pt-based electrocatalysts process most efficient and practical ORR ability. Nevertheless, expensiveness, poor reliability, besides cross-effects limit their widespread industrial applications [4]. Thus, developing strongly electro-active catalysts containing less Pt with high durability need to be addressed to resolve above drawbacks [5].

The optimization of active site reactivity and their full utilization are critical factors for fabricating efficient Pt-contained ORR catalysts [6]. Thus, catalytic performances can be enhanced by regulating electronic construction and optimization of atomic arrangement on the Pt surface [7, 8]. One approach is to combine Pt particles with a catalytic support, possessing strong resistance to corrosion and sturdy interactions with Pt [9]. Carbon-based materials possess a high electrical conductivity, particularly, it can accommodate high Pt nanoparticles (NPs) density. However, C-supported Pt catalysts would be often corroded in a harsh environment of the membrane fuel cells due to feeble interactions between Pt NPs and support carbon, which leads to catalyst particles aggregation, low catalytic ability, and declining stability [10]. The combination of metal oxides (e.g., TiO2, NbO2, SnO2, and CeO2) into Pt-containing catalysts can solve this problem [11,12,13]. Strong interactions of these metal oxides with Pt NPs can vary their electronic structure and enhance the overall catalytic activity [14].

CeO2 attracted high attention as a catalytic support because of its special oxygen storage and release capacity, which derived from two different valence states (Ce4+, Ce3+) with a unique reversible redox property, and plentiful oxygen vacancies [15]. Thus, CeO2 is employed for encouraging the catalytic performance of Pt-contained catalysts. Enhanced oxygen electrode activities of the Pt-CeO2 catalysts are a result of the strong electron Pt-CeO2 interactions and the presence of the Ce4+/Ce3+ redox couple [16, 17]. Ce3+ cations on the CeO2 surface act as defects and stabilize the catalytic performance, triggering the overall material more durable due to their ability of oxygen vacancies compensation. Besides, it’s worth noting that CeO2 can effectively promote the degradation of H2O2, so that more effective active sites will be retained to further increase stability of the catalyst [18]. However, CeO2 possesses poor electronic conductivity, which impedes applications of Pt-CeO2 catalysts for ORRs. The introduction of C into the Pt-CeO2 materials could raise the electronic conductivity without compromising sturdy interaction between Pt and CeO2 [19, 20].

One of the most efficient ORR catalysts is N-doped C nanomaterials. N-doped C materials can significantly promote the adsorption of oxygen molecular because of N atoms acting as positively charged sites. Besides, N atoms doping results in widely distributed carbon defect sites, serving as catalytic active sites in ORR [21]. Aromatic polyaniline (PANI) is often used to produce N-doped C as it provides N and controls the desired material morphology when heated at a certain temperature. PANI also acts as a protective film to inhibit C corrosion [22, 23]. When N is introduced into the C lattice, the number of the positively charged sites increases, which is beneficial for oxygen molecules absorption and faster ORR [24,25,26]. Usage of CeO2 combined with an N-doped C as composite support for Pt-based catalysts offers an excellent potential for improving its stability and activity.

Up to now, Pt-based ternary electrocatalysts supported by carbon modified metal oxide have been reported to be a near-ideal approach to advance the ORR activity [20, 27]. Synthesis of the hybrid catalyst by combination of as-prepared individuals sequentially [28, 29], and introduction Pt into the as-prepared composite supports including carbon-encapsulated metal oxide [29, 30] were often adopt to fabricate Pt-based ternary electrocatalysts. The above strategies may face problems of intricate-steps and inefficacy in former, and reduce Pt direct contact with metal oxide, leading to poor interaction and stability in latter. Further, employment of special equipment, such as electron-deposition technology and area-selective atomic layer deposition, to achieve these specific Pt-based ternary catalysts, increase the cost of preparation and difficulty of process [31, 32]. Thus, there is thereby an urgent need but it is still a significant challenge to develop an economical strategy, with the advantages of convenience in operation and high efficiency, to develop highly active Pt-based ternary electrocatalysts with strong interaction and uniform distribution of three components.

Thus, we prepared a ternary Ptx/CeO2/N–C composite catalyst by a facile two-step strategy with economic benefits and ease of synthesis to enhance the catalytic property and stability. Ptx/CeO2/PANI composites with K2PtCl6 as doped anions for neutralizing the PANI system was synthesized, followed by a further pyrolysis treatment. The catalysts including nitrogen doping of carbon and Pt NPs anchored CeO2 are achieved during a typical pyrolysis treatment. The well-designed catalyst exhibits surprising catalytic performance and highly long-term durability during ORR assessment.

Results

Fabrication and characterization

Scheme 1 shows the stepwise synthesis procedures. The first step is to involve aniline polymerization on the CeO2 surface to form PANI-covered CeO2. Then PtCl62−, functionalizing for charge compensation of the oxidized PANI backbone, is incorporated into the PANI matrix. The second step is PANI decomposition during the high-temperature annealing, and N-doped C forms during this stage, meanwhile Pt NPs also are fabricated (as a result of PtCl62− decomposition) during the pyrolysis step and loading on N–C/CeO2, forming a ternary Ptx/CeO2/N–C catalyst.

Scheme 1
scheme 1

Preparation steps of the Ptx/CeO2/N–C catalysts synthesis

Full size image

Formation of high purified fluorite-structures CeO2 (JCPDS-43-1002) was first confirmed by XRD diffraction patterns in Fig. 1, without the presence of other Ce-containing phases diffraction. Average particle size of as-prepared CeO2 is approximately 6.0 nm in light of Scherrer equation. XRD of N–C/CeO2 demonstrates a sharp peak at 26.2°, which is consonant with graphitic carbon (002) reflection and the other diffraction peaks at 28.8°, 47.8°, and 56.7° are allocated severally to CeO2(111), (220) and (311) planes, illustrating the formation of N-doped graphitic carbon/CeO2 composite material. A comparison of the XRD patterns of CeO2/PANI-PtCl62−0.1 and Pt0.1/CeO2/N–C indicates that a broad diffraction peak at 25.5° attributes to the PANI matrix [33] (see Fig. 1a, b). Additionally, there is no new peak appeared in the material after K2PtCl6 is introduced into the reaction mixture, confirming that PtCl62− remains stable and does not decompose to Pt NPs. Several new peaks appear in the XRD pattern of Pt0.1/CeO2/N–C at 39.8°, 46.3°, 67.5°, and 81.3° (see Fig. 1b), which correspond to (111), (200), (220), and (311) planes of Pt according to the PDF card number 01-087-0646. Among the planes of (111) and (311) XRD peaks of CeO2 are detected at 28.5° and 47.3°. Thus, Pt NPs and CeO2 coexist in this composite material. The intensity of the Pt XRD peaks in the Ptx/CeO2/N–C spectra (x = 0.05–0.2) in Fig. S1 increases as x increasing. The average diameter of the Pt NPs enlarges from 5.7 to 27.4 nm as the amount of added K2PtCl6 increases from 0.05 to 0.2 g, respectively. ICP-OES exhibits that the actual Pt contents in the Pt0.2/CeO2/N–C, Pt0.1/CeO2/N–C, and Pt0.05/CeO2/N–C catalysts are 6.8%, 5.6%, and 5.1%, respectively. These amounts are far below commercial Pt/C catalysts containing 20 wt% of Pt.

Figure 1
figure 1

XRD patterns of CeO2 (top purple curve in both a, b), a CeO2/PANI composite (material before the pyrolysis, bottom blue curve) and N–C/CeO2 (catalyst formed after the pyrolysis, middle green curve) and of b CeO2/PANI-PtCl62− (materials the before pyrolysis, middle, red curve) and Pt0.1/CeO2/N–C (yellow bottom curve, catalyst formed after the pyrolysis). Raman figures of c the CeO2, CeO2/PANI, and CeO2/PANI- PtCl62− before the pyrolysis and d N–C/CeO2 and Pt0.1/CeO2/N–C materials obtained after the pyrolysis

Full size image

Raman spectra CeO2/PANI-PtCl62− and PANI/CeO2 demonstrate vibrational bands at 410 cm−1 and 434 cm−1of CeO2, and compared to the band position of the pure ceria at 458 cm−1 slightly peaks shift was observed [34, 35] because of the packaging of PANI on the surface of CeO2 NPs [36]. PANI coating on CeO2 NPs result in the appearance of bands at 1150 cm−1 and 1310 cm−1 due to stretching vibrations of C=N and C–N, and the bands at1490 cm−1 and 1602 cm−1 associated to bending vibrations of C=C in PANI [37], as shown in Fig. 1c. These Raman results confirm the presence of PANI on the CeO2 NP surfaces.

Samples subjected to the pyrolysis, N–C/CeO2, and Pt0.1/CeO2/N–C, exhibited two Raman peaks at 1369 and 1598 cm−1 and 1350 and 1602 cm−1, respectively, (see Fig. 1d), which are the D and G bands of carbon. These separately corresponds to disordered graphitic and sp2-hybridized C [38]. The intensity relation between the D and G bands (ID/IG) of Pt0.1/CeO2/N–C specimen (which was equal to 1.08) is higher than for the N–C/CeO2 catalyst (which is equal to 0.99). Thus, the Pt0.1/CeO2/N–C sample possesses more defective sites in the C-support.

As shown in XPS spectra of Pt0.1/CeO2/N–C and CeO2/PANI-PtCl62−, peaks belonged to C and N, conforming PANI presence in the samples (see Fig. 2a). Pt, Cl, O, and S originates from the chloroplatinate and sulfate anions, respectively. A weak Ce XPS signal at ~ 940 eV is detected in the samples prepared before and after the pyrolysis. The samples subjected to the pyrolysis show weak or no Cl peaks. The disappearance of the Cl peak in the Pt0.1/CeO2/N–C spectrum confirmed PtCl62− decomposition during the pyrolysis step. As shown in Fig. 2b, Pt 4f spectrum of the CeO2/PANI-PtCl62− sample, respectively, demonstrates two peaks at 75.9 eV and 72.7 eV, interpreting as Pt 4f5/2 and Pt 4f7/2, which remain typical for compounds containing Pt4+[30]. The Pt 4f peaks in the pyrolyzed catalysts shifted to ~ 74.5 eV and 71.2 eV, respectively. The last peak binding energy is almost identical to the binding energy of pure Pt (which is equal to 71.15 eV), which confirms the formation of Pt NPs. The residual two peaks at 72.4 eV and 75.8 eV are accredited to Pt oxides (see Fig. 2d) [39].

Figure 2
figure 2

a XPS wide spectra and b Pt 4f of the Pt0.1/CeO2/N–C and CeO2/PANI-PtCl62− samples. c Ce 3d and d Pt 4f spectra of the Pt0.1/CeO2/N–C. e N 1s and f C 1s core-level spectra of the Pt0.1/CeO2/PANI

Full size image

Ce 3d XPS spectra for the Pt0.1/CeO2/N–C andCeO2/N–C samples are illustrated in Fig. 2c and Fig. S3b. Owing to the CeO2 lattice defects, Ce in CeO2 is often presented as a mixture of Ce3+ and Ce4+ [40, 41]. The peaks resolved at 885. 15 eV and 903.88 eV are signed to Ce3+, while other peaks belong to Ce4+. Pt0.1/CeO2/N–C sample shows slightly higher Ce3+ content (equal to 22.59%) than N–C/CeO2 sample (which contained 21.3% of Ce3+, see Fig. S3b), very likely because of the partial Ce4+ reduction under the Pt NP presence. This observation might imply that Ce3+ in CeO2 can act as a reducing agent facilitating Pt NP nucleation on the CeO2 surface [42].

In Fig. 3e, four different categories of N were deconvoluted in the N1s XPS spectrum of the Pt0.1/CeO2/N–C sample (see Fig. 2e). Peaks at 401.2 eV and 403.3 eV 398.3 eV are graphitic and oxidized N, peaks at 398.3 eV and 399.8 eV are ascribed to pyridinic and pyrrolic nitrogen [3, 43. Percentages of these species in the total N content are shown in Fig. 3d. Graphitic and pyridinic nitrogen are typically active during ORR. Pyridine N has an unshared pair of electrons, and the electrons that are involved in the conjugated p-bond [44]. Graphitic N is energetic during electron transfer from C to O orbitals. Effect of Pt presence on the percentages of these N species is not apparent (see Table S1). As shown in Fig. 3f, four peaks sited at 284.6, 285.2, 286.2 and 289.12 eV were resolved in the C1s spectrum of the Pt0.1/CeO2/N–C sample, consistent with the graphitic, aromatic or aliphatic C=C bonds, C–C bonds at 285.2 eV, C–N bonds and 4) C=O bonds, correspondingly.

Figure 3
figure 3

a CeO2, b CeO2/PANI-PtCl62− and c, d Pt0.1/N–C/CeO2 TEM and HRTEM micrographs

Full size image

TEM images of the Pt0.1/CeO2/N–C catalyst demonstrate agglomerated sphere-like CeO2 NPs, 5–10 nm in size (see Fig. 3a), which agrees with the XRD results. HRTEM showed lattice fringes 0.31 nm apart, which matches to the (111) plane of CeO2 [45]. Figure 3b shows CeO2 NPs homogeneously disperse in the PANI matrix. EDS of the CeO2/PANI-PtCl62− catalyst shows homogeneous distribution of C, N, Pt, and Cl (see Fig. S3), which confirms PtCl62− presence in the PANI matrix.

Numerous small dark dots with average diameter 5 nm appear in the C substrate after pyrolysis step (see Fig. 3c and Fig. S4, Fig. S5). Lattice fringes by HRTEM analysis of these dots in Fig. 3d display of 0.227 nm and 0.312 nm apart, agreeing with Pt (111) plane and CeO2(111) planes, respectively [10, 39]. Pt and CeO2 NPs overlapped, as shown in Fig. 3d and Fig. S4h). EDS elemental mappings confirm the uniform distribution of Pt NPs in the N–C matrix (see Fig. S4e). Partial Pt NPs distribution on the surface of CeO2 indicates that the Pt NPs nucleate at CeO2 surfaces during the pyrolysis and PtCl62− decomposition. The abundance of oxygen vacancies in CeO2 provides the nucleation sites facilizing the formation of Pt NPs [46, 47]. Thus, Pt NPs nucleate on the defective CeO2 surface, and large Pt-CeO2 interface is achieved. Such Pt NPs arrangement and tight interfaces could be very beneficial for the electronic interactions and transferring, which, in turn, is excellent for improving the conductivity, catalytic activity, and enduring ORR performance of catalyst. The TEM and HR-TEM images illustrate that the average diameter of Pt NPs obvious increases in the catalysts without the presence of CeO2 (Fig. S6).

Electrocatalytic ORR performance of the Pt0.1/CeO2/N–C material

The ORR activity of the Pt0.1/CeO2/N–C is analyzed using CV and associated to performance of commercial 20% Pt/C (see Fig. 4). A very strong ORR activity is observed for our catalyst judging by two strong cathodic peaks and a significantly larger area under the CV curve relative to the commercial 20%Pt/C material. Additionally, Pt0.1/CeO2/N–C catalyst behavior is typical for hydrogen adsorption and desorption reactions in the 0–0.3 V range (see Fig. 4a). However, it does not reveal the peaks by the CV of Pt0.1/N–C, further emphasizing the critical role of CeO2 in the global act of catalyst. Moreover, Pt surface oxidation is more evident for the Pt0.1/N–C sample. Thus, the CeO2 role aims to inhibit Pt oxide formation. Previous studies showed that even minor Ce3+ contents could suppress Pt oxide formation, especially at high voltages, by absorbing excessive oxygen from the system and converting to Ce4+ [42]. Relative to the N–C/CeO2 sample performance during the CV tests, the ORR onset potential of the Pt0.1/CeO2/N–C catalyst significantly shifts positively, and the oxygen reduction current peak significantly enhances because of the Pt NP presence. The ternary Pt0.1/CeO2/N–C catalyst demonstrates dramatically higher overall ORR activity. Furthermore, compared to 20wt%Pt/C, the prepared ternary catalyst displays an expanded CV curve with dramatically higher current densities.

Figure 4
figure 4

a CV and b LSV curves for the Pt0.1/CeO2/N–C, N–C/CeO2, and Pt0.1/N–C, 20%Pt/C catalysts recorded at 10 and 50 mV s−1 sweep rates. c Onset, half-wave potentials and limited current density of the Pt0.1/CeO2/N–C, N–C/CeO2, Pt0.1/N–C, and Pt/C. d LSV curves and e Kouteckye-Levich (K-L) plots, besides n for the Pt0.1/CeO2/N–C catalyst

Full size image

ORR performance of Pt0.1/CeO2/N–C is evaluated by LSV at 1600 rpm (presented in Fig. 4b), which shows a half-wave potential (E1/2) at 0.19 V, and 90 mV more positive than E1/2 of the Pt0.1/N–C and N–C/CeO2, separately (see Fig. 4c and Table S1). The limiting diffusion current density (jl) for the Pt0.1/CeO2/N–C (which is equal to 5.88 mA cm−2) is advanced than Pt0.1/N–C and N–C/CeO2 catalytic agent (which are equal to 4.47 and 3.38 mA cm−2, respectively). And the mass activity of Pt0.1/CeO2/N–C material at 0.9 V was 238.46 mA mgPt−1, which was 4.1 times that of 20% Pt/C catalyst (which is 58.8 mA mgPt−1). Such an excellent behavior attributes to simultaneous attendance of Pt, Ce3+, N in the catalyst, and catalytic synergy provided by the Pt and CeO2 NPs and N-doped C support.

ORR kinetic parameters for the Pt0.1/CeO2/N–C obtained by LSV are recorded at varied rotating speeds. The jl in Fig. 6d increases as the LSV rotating speed rising. The corresponding K-L results in Fig. 6e derived from LSV information express the number of the transferred electrons at 0.4–0.6 V is 3.89–4.02 (see Fig. 4e), which indicates a 4e− reduction path for the Pt0.1/CeO2/N–C electrocatalyst during its participation in the ORR.

The durability of the Pt0.1/CeO2/N–C electrocatalyst and 20%Pt/C is assessed by comparing CV and LSV data before and after 2000 cycles in alkaline environment (see Fig. S7). The E1/2 values decrease slightly after 2000 cycles during the prolonged ADP tests. Insignificant accumulation and increasing dimensions of Pt NPs during ADP tests would result in the stability degradation [30, 50,51,48,49,50,]. However, this degradation of the Pt0.1/CeO2/N–C catalyst is still very acceptable, which could attribute to its novel ternary structural design. The Pt NPs formed in-situ during the carbonization process, are wrapped in the carbon support. The N-doped C support improved Pt NP homogeneous dispersion and adhesion because of facilitated charge transfer from Pt to N. Additionally, forceful mutual effect between CeO2 and Pt NPs enables formation of a stable complex, which can restrain both movement and accumulation of Pt NPs in the N-doped carbon support [51].

Discussion

Formation mechanism of the ternary heterostructure Pt0.1/CeO2/N–C catalyst

During the aniline polymerization to PANI, the [PtCl6]2− as-doped anions are distributed homogenously in the PANI matrix to ensure consequent uniform Pt NP nucleation upon further carbonization. As the PANI matrix is pyrolyzed into an N-doped C support, [PtCl6]2− decomposed into Pt NPs. As stated by the TG-MS results in Fig. 5, the thermal decomposition of the PANI/CeO2/[PtCl6]2− could be divided into three stages. The weight loss below 100 °C is attributed to the physically absorbed water removal judging by the fact that MS detected H2O, CO2, and O2 release at 80 °C (see Fig. 5a). At 200–400 °C, 15 wt% of the catalyst weight loss occurred. The low molecular weight oligomer and doped APS and [PtCl6]2− anions undergo thermal decomposition, leading to the HCl, Cl·, and SO2 escaping from the system according to MS. Pt NP nucleation very likely starts at this temperature range as well. Above 400 °C, the PANI backbone is prone to decompose continuously [45]. C2H2 and CHN fragment release could occur above 500 °C.

Figure 5
figure 5

a TG–DTA curves of CeO2/PANI-PtCl62−. MS results showing the evolution of gaseous species at 80 (b), 280 (c), and 600 °C (d). e–h MS results showing gaseous species evolution from the samples prepared using different amounts of K2PtCl6 as a function of the temperature

Full size image

Subsequently, we analyzed intermediate PANI/CeO2/[PtCl6]2− pyrolysis products obtained at 300 °C, 400 °C, and 600 °C by XRD and TEM (see Fig. S8 and Fig. S9, respectively). XRD result from the catalyst heated to 300 °C shows a hump at 39.8°, belonging to Pt (111) plane. Based on this data, it concludes that Pt NPs nucleation occur at ~ 300 °C. Meanwhile, the decomposition of the PANI matrix result in the absence of the diffraction peaks at 20.1° (110) and 25.5°(100) [33]. The abundant Ce3+ in CeO2 proving the nucleation sites facilities the formation of Pt NPs. Pt NPs are trapped at the site of the vacancies of oxygen in CeO2 [42, 47]. From XPS evaluations in Fig. 2 and Fig. S2, Ce3+ presence with an increment from 21.3 to 22.59% provides the catalyst with abundant oxygen vacancies for Pt NPs to nucleate and grow. When the temperature rises to 300 °C, thermochemical reduction in the doped [PtCl6]2− occurs, conversing the trapped Pt ions into the reduced metal state. As the pyrolyzed temperature increasing, more Pt NPs nucleate, the movement and coalescence of Pt NPs trigger growth of Pt NPs in diameter. Thus, the diffraction intensity of (111) Pt peak located at 39.8° increases significantly because of intensified Pt NP nucleation and growth, which is verified by XRD and TEM (see Fig. S8 and Fig. S9).

Electrocatalytic ORR performance of the Pt0.1/CeO2/N–C catalysts

According to results of electrocatalytic measurements in Fig. 4, the ternary hetero-structured Pt/CeO2/N–C catalysts display superior ORR action in alkaline solution. Particular, we would like to emphasize that the Pt content in these materials is significantly inferior than in the commercial Pt-based catalytic agent. Content of platinum of resulted Pt/CeO2/N–C catalysts is only 5 wt%; however, the mass activity of the Pt0.1/CeO2/N–C material at 0.9 V is 238.46 mA mgPt−1, which is 4.1 times that of 20%Pt/C material (58.8 mA mgPt−1). Additionally, our catalysts demonstrate an outstanding electrocatalytic stability. Table S1 compared ORR ability of lately reported Pt/CeO2/C ternary catalyst with our catalysts. It can be seen that the ternary catalytic material reported in this paper shows the highest limited currents and mass activity.

Schematic diagram of the ternary catalyst is proposed in Scheme 1 to illustrate the excellent electrocatalyst performance. A ternary synergetic effect of individual components leads to the enhancing ORR movement. Firstly, attendance of Pt NPs dramatically shifts the onset and half-wave potential in positive direction by comparison with CeO2/N–C. Both the results of XRD and TEM indicate Pt NPs with plentifully exposed (111) facet, in situ anchored on the CeO2 surface and nitrogen doped carbon support, strongly increase the overall active catalyst active sites [48,51,52,53,48,49,50,51,52,]. In return, the presence of CeO2 in the ternary catalyst, not only play an important role in facilizing Pt NPs nucleation, but also the numerous oxygen vacancies on the CeO2 surface and large Pt-CeO2 interface result in sturdy Pt-CeO2 interactions, which can improve its ORR activity and durability. Furthermore, the Pyridinic and graphitic nitrogen atoms with their sp2 electronic structures are very active ORR sites [3, 26, 55]. They enhance the material electronic conductivity and increase C corrosion resistance, which, again, can effectually advance the movement and constancy of the catalysts.

Effect of the added K2PtCl6 amount on the performance of the Pt0.1/CeO2/N–C catalysts

CV and LSV data of the ternary heterostructure Pt/CeO2/N–C catalysts prepared with various K2PtCl6 amounts are shown in Fig. 6a, b. All CV curves displays the typical hydrogen adsorption and desorption behavior in the 0–0.3 V range. However, their catalytic performance is strongly dependent on the amount of the K2PtCl6 used during the catalyst preparation. The addition of 0.1 g of K2PtCl6 leads to the best ORR catalytic act. The weaker presentation of the other catalysts prepared with higher K2PtCl6 amounts could be enlightened by formation of more Pt NPs, which is highly confirmed by both XRD and TEM measurements. The presence of excessively large Pt NPs could slow down the electrochemical activity of the ternary catalysts, since too large Pt NPs do not provide sufficient surface area. Additionally, an excessive number of Pt NPs can block the micropores of the N-doped carbon support, thus impeding oxygen diffusion. Figure 6c, d demonstrates the nitrogen adsorption/ desorption isotherms and pore size distribution of the as-prepared samples equipped with diverse K2PtCl6 amounts.

Figure 6
figure 6

Pt0.05/CeO2/N–C, Pt0.1/CeO2/N–C Pt0.15/CeO2/N–C, and Pt0.2/CeO2/N–C a CV and b LSV results recorded in O2-saturated 0.1 M KOH solution. c N2 adsorption–desorption isotherms and d pore size distribution results for Pt0.1/CeO2/N–C, Pt0.05/CeO2/N–C and Pt0.2/CeO2/N–C

Full size image

The specific surface area was significantly affected through amount of Pt loaded. Pt0.2/CeO2/N–C has an isotherm which exhibits the II-type curve, and its BET surface area remains 68.3 m2 g−1, as shown in Table S2, both of which indicate a nonporous structure. The surface areas of Pt0.1/CeO2/N–C and Pt0.05/CeO2/N–C samples were significantly higher, 637.0 and 603.7 m2 g−1, respectively. Hierarchical porous structure is noticed with average pore diameter of 4.13 nm in Pt0.1/CeO2/N–C. The hierarchical porous structure grows in specific surface area which provides plentiful energetic spots for catalytic reaction, meanwhile porosity facilitates the transfer of reactants, promotes the occurrence of catalytic response, and thus improves catalytic ability.

Conclusion

This study demonstrates an innovative approach for fabricating low platinum content Pt0.1/CeO2/N–C ternary catalyst with Pt highly uniformly dispersion. Accompanied by the thermal decomposition of polyaniline/CeO2 into N-doped carbon/CeO2 support, numerous Pt NPs nucleate and grow in-situ from the doped PtCl62− anions in polymeric precursor during carbonization, resulting in the porous heterostructure Pt0.1/CeO2/N–C with extraordinary surface area. An optimal amount of K2PtCl6 addition was obtained during synthesis of PANI/CeO2 catalyst precursor for further fabricating the Pt0.1/CeO2/N–C ternary catalyst with the largest specific surface area (equal to 637.0 m2 g−1) and porosity (0.66 cm3g−1). The resulting ternary catalysts exhibit excellent electrochemical properties in 0.1 M KOH solutions. Comparison with 20% Pt/C materials (with activity equal to 58.8 mA mg Pt), the activity of our Pt0.1/CeO2/N–C catalyst (containing only 5.6 wt% of Pt) is four times higher, equal to 238.46 mA mg Pt. The Pt0.1/CeO2/N–C catalyst processes positive half-wave potential at 0.88 V, limit current density valued at 5.88 mA/cm2 (vs. 0. 87 V and 5.50 mA/cm2 of 20% Pt/C, respectively) and enhanced electrochemical permanency. The improved ORR act is ascribed to the unique Pt0.1/CeO2/N–C ternary interconnected structure, in which collaboration of Pt and CeO2 modifies electronic environment of Pt surface, and N-doped carbon scaffold improves catalyst stability and conductivity. These findings reveal an important new strategy for creating affordable and high-efficient Pt-founded ternary electrocatalyst, which should thus have great potential for applications as commercial catalytic materials in fuel cell.

Supporting information

Experimental details for all as-prepared catalysts, including initial materials and chemicals, synthesis approaches, as well as characterization and electrochemical measurements methods details and other supporting information