Facile Synthesis of Quaternary Structurally Ordered L12-Pt(Fe, Co, Ni)3 Nanoparticles with Low Content of Platinum as Efficient Oxygen Reduction Reaction Electrocatalysts.

Synthesis of electrocatalysts for oxygen reduction reaction (ORR) with not only prominent electrocatalytic performance but also a low amount of Pt is the urgent challenge in the popularization of fuel cells. In this work, through a facile synthetic strategy of spray dehydration on a solid surface and annealing process, we demonstrate the first manufacture of quaternary structurally ordered PtM3 (M = transition metal) intermetallic nanoparticles (NPs), Pt(Fe, Co, Ni)3, in order to lower the content of Pt. The atomic contents of Pt, Fe, Co, and Ni are equal and the chemical structure of Pt(Fe, Co, Ni)3 is a cubic L12-ordered structure. L12-Pt(Fe, Co, Ni)3/C electrocatalysts exhibit enhanced electrocatalytic performance toward ORR with mass activity (MA) 6.6 times higher than the commercial Pt/C and a minimal loss of 17% in MA and 1.5% loss in specific activity (SA) after 10 000 potential cycles at 0.9 V. Furthermore, the stability behavior is confirmed to be attributed to the coaction of particle sizes and the ordering effect. Compared with traditional Pt-based electrocatalysts in the stoichiometric forms of Pt3M and PtM, L12-Pt(Fe, Co, Ni)3 intermetallic NPs exhibit excellent performance and higher cost effectiveness. Moreover, this work also proposes a facile and effective synthetic strategy for manufacturing multicomponent Pt-based electrocatalysts for ORR.

Introduction

Proton-exchange membrane fuel cells (PEMFCs) have been considered to be one of the most satisfactory power supply for electric automobiles and personal devices where chemical energy of the fuel (for instance, hydrogen or methanol) is converted into electric energy.[1−3] In recent decades, a huge number of research studies have been investigated on exploring and designing high-efficiency catalysts with superb performance to expedite the kinetics of oxygen reduction reaction (ORR) at the cathode because of its dominant role in PEMFC.[4−6] On account of the particular electronic structure of Pt for fast reaction kinetics, the current electrocatalysts appropriate for ORR primarily contain the component Pt.[7−9] However, there appears a hindrance for the large-scale application of fuel cells because of the high cost arising from scarcity. In view of this, extensive efforts have been made on boosting the electrocatalytic performance and decreasing the content of Pt simultaneously. Recently, Pt-based alloy electrocatalysts, especially the alloys of Pt and 3d-transition metals such as Fe, Co, Ni, and Cu have attracted tremendous attention and interest because of their excellent properties.[10−13] The enhancement of electrocatalytic performance is attributed to electronic structure modification of Pt and the lower d-band center position resulting from the coalescence of transition metals.[14,15] Some popular strategies for designing active electrocatalysts for ORR primarily focus on the regulation upon morphology and nanocrystalline growth such as Pt surface-enriched or core–shell nanoparticles (NPs),[16−18] hollow structures,[19,20] and specific polyhedral structures to expose high-index facets.[21−23]

However, conventional Pt-based nanocatalysts where the distribution of Pt and transition-metal atoms is in a random configuration are usually incompetent for commercial application because of their unsatisfactory stability arising from particle growth, dissolution of transition metals, and the corrosion of the carbon support under extreme environments.[24,25] Therefore, structurally ordered intermetallic NPs have stood out for their unique geometrical structure and feature,[26] which can be regulated through control over temperature and annealing atmosphere.[27,28] Structurally ordered Pt-based alloy electrocatalysts combine the excellent activity and ideal chemical stability together, serving as perfect elctrocatalysts for the next-generation PEMFCs.

Because of the ligand and strain effect and the accompanying optimum electronic structure of the Pt surface for absorption of Oad and OHad, Fe, Co, and Ni are usually employed in Pt-based catalysts, and hence, Pt–Fe, Pt–Co, and Pt–Ni nanoalloys are identified as the perfect catalytic system for ORR with excellent activities.[29,30] Whereas, Cu is considered not applicable for the practical application of PEMFC on account of the poisoning effect because leached Cu ions can transfer from the cathode and accumulate on the anode.[31] Until now, Pt-based electrocatalysts with excellent performance are basically binary or ternary intermetallic in the stoichiometric form of Pt3M or PtM with a relatively high content of Pt.[18,27,28,32−35] Synthesizing cubic L12-PtM3 NPs with a molar ratio of Pt merely one-third of transition metals can be an effective tactic to prepare active low-Pt catalysts for ORR. However, structurally ordered PtNi3 cannot be obtained even after being annealed from 400 to 800 °C because of its low temperature for disorder-order transition.[36] Hence, synergistically employing Fe, Co, and Ni can be a valid strategy to realize the structurally ordered PtM3 alloy phase. Besides, the widely employed chemosynthesis method for Pt-based electrocatalysts is relatively complicated and low-producing to some degree. As a result, synthesis of intermetallic electrocatalysts in the ordered chemical structure with a low proportion of Pt through a facile and versatile strategy is the key to the popularization of fuel cells.

Inspired by this, we demonstrate the first synthesis of quaternary structurally ordered PtM3 type NP nanocatalysts L12-Pt(Fe, Co, Ni)3/C simply through spray dehydration on the solid surface followed by an annealing process. The coalescence of Fe, Co, and Ni gives rise to the atomic re-establishment and the ordered chemical structure facilitates the anticorrosive ability against acid solution. The quaternary PtM3 type NP eletrocatalysts exhibit outstanding activities and superior stability toward ORR with a considerably low amount of noble metal Pt, serving as a puissant substitute for the commercially popular catalysts in the form of Pt3M and PtM. The novel L12-Pt(Fe, Co, Ni)3/C eletrocatalysts and the facile synthetic strategy reported in this work point out a new direction for exploring promising electrocatalysts suitable for fuel cells.

Results and Discussion

The structurally ordered Pt(Fe, Co, Ni)3 NP catalysts were synthesized through an extremely simple strategy compared with typical chemosynthesis, which generally takes considerable time and effort. Figure shows the schematic illustration of the synthetic route, which can be summarized into a two-step process. The well-dispersed suspension containing specific amounts of chemicals was prepared in advance after an ultrasonic processing for 40 min. Then, the spray dehydration method was adopted to obtain the mixed precursor at molecular level by promptly drying the suspension on a quartz plate at a temperature of 300 °C. Compared with the typical spray drying method where atomized materials are dried out in hot air, spray dehydration on the solid surface has a faster evaporation rate and leads to more uniform precipitation of precursor concurrently. Finally, the leftover was scraped off the quartz plate and then annealed at 600 and 700 °C for 1 h under the H2 atmosphere [marked as Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700]. In order to verify the reliability of this synthetic strategy, quaternary NPs containing Pt, Au, Cu, and Ni were firstly fabricated and investigated because of the weaker magnetism, which makes them more convenient for scanning transmission electron microscopes–energy dispersive X-ray (STEM–EDX) characterization.

Figure 1

Schematic illustration of the synthetic strategy for L12-Pt(Fe, Co, Ni)3 NPs. (a) Suspension comprising specific amounts of chemicals. (b) Sketch of the spray dehydration method on the surface of a quartz plate. (c) Precursors at the molecular level obtained on the quartz plate after spray dehydration. (d) Structurally ordered NPs acquired after being annealed at high temperature under a H2 atmosphere.

The morphological information of PtAuCuNi/C NPs synthesized by the spray dehydration method was first investigated using a transmission electron microscope (TEM). As exhibited in Figure a, the high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) image of PtAuCuNi/C NPs shows a narrow size distribution with an average size of 5.77 nm. Figure b shows the STEM image captured from a representative area and Figure c–f shows the STEM–EDX elemental mappings of all the elements (Pt, Au, Cu, and Ni) contained in the NPs. It is well-marked that Pt, Au, Cu, and Ni are distinctly detected and present a uniform distribution among the NPs. The EDX drift-corrected spectrum images in Figure S1 also confirm the existence of Pt, Au, Cu, and Ni elements. The homogeneously distributed elemental mappings convincingly certify the synthetic strategy of spray dehydration on the solid surface followed by annealing process, making it a convenient and effective strategy for synthesizing multicomponent NP catalysts for energy conversion applications.

Figure 2

(a) HAADF-STEM morphological image of synthesized PtAuCuNi/C NPs. (b) STEM image of PtAuCuNi/C NPs in a particular area. (c–f) EDX elemental mapping images corresponding to Pt, Au, Cu, and Ni, respectively.

The crystal structures of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 NPs were then investigated by using an X-ray diffractometer. XRD patterns in Figure a show a typical simple cubic intermetallic structure. The peak around 25° is considered to be because of the amorphous carbon support. The other four obvious diffraction peaks are corresponding to the lattice planes of (111), (200), (220), and (311) in sequence. It can be observed that the diffraction peak positions are in good match with that of ordered L12-PtFe3 (PDF # 71-8365), revealing a slight shift to higher angles because the atomic radii of Co and Ni are a little bit shorter than that of Fe (shown in Figure S2). XRD patterns indicate that the four kinds of atoms are incorporated to form a cubic L12-type ordered structure with the space group of Pm3̅m. The atomic ratio among Pt, Fe, Co, and Ni is further confirmed to be equal by inductively coupled plasma atomic emission spectroscopy (ICP-AES), which is self-consistent with the XRD patterns and the proportion of metal salt precursors. It should be noticed that there appear extra ordering peaks around 33°, 54°, and 61°, which correspond to the (110), (210), and (211) planes, respectively, suggesting that structurally ordered L12-Pt(Fe, Co, Ni)3 NPs are successfully synthesized with Pt, Fe, Co, and Ni atoms in an ordered spatial arrangement. On the basis of XRD patterns and the Scherrer equation, the average crystallite sizes of Pt(Fe, Co, Ni)3-600 and Pt(Fe, Co, Ni)3-700 NPs are determined to be 5.0 and 6.1 nm, respectively, as details shown in Table S1. In particular, Pt(Fe, Co, Ni)3-700 NPs reveal a higher degree of ordering seen from the more distinct (110) ordered peak compared with that of Pt(Fe, Co, Ni)3-600 NPs. Figure b reveals the magnetic hysteresis loops of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 NPs catalysts carried out from 0 to 2 T at 300 K. The saturation magnetization of Pt(Fe, Co, Ni)3/C-700 NPs is higher than that of Pt(Fe, Co, Ni)3/C-600 because NPs with a larger particle size exhibit higher saturation magnetization at the macro level.[37] According to the thermal decomposition behavior of polyvinyl pyrrolidone (PVP), the percentage of decomposed PVP at high temperature is about 95%, while 5% of the residue still remains.[38,39] In view of this, the alloy loading of Pt(Fe, Co, Ni)3 on the carbon support is confirmed to be 20 wt % through thermogravimetric analysis under the air environment. Supposing that all transition metals (M) are oxidized into the form of M2O3, the theoretical remaining weight of the Pt(Fe, Co, Ni)3/C catalyst is calculated to be 27%. The residue weight is quite close to the result of 27.7% in Figure S3, which testifies the 20% weight loading of the Pt(Fe, Co, Ni)3 alloy on the carbon support (calculation details in Supporting Information). Therefore, the content of Pt in Pt(Fe, Co, Ni)3/C catalysts is extremely low compared with traditional Pt3M and PtM type catalysts.

Figure 3

(a) XRD patterns of synthesized Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 NPs. The green vertical curves are the standard XRD characteristic lines of ordered L12-PtFe3 (PDF # 71-8365). (b) Magnetic hysteresis loops of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 measured at 300 K.

The morphological information and microstructure of Pt(Fe, Co, Ni)3 NPs were further investigated by using a TEM and a high-resolution transmission electron microscope (HR-TEM). Figure a shows the low-resolution TEM image of Pt(Fe, Co, Ni)3-600 NPs distributed on the carbon support. It is observed that the NPs are uniformly distributed on the support with a good sphericity (also shown in Figure S4). Figure b reveals the high-resolution TEM image of Pt(Fe, Co, Ni)3/C-600 NP, indicating that the quaternary NPs are well crystallized. The Fourier transform pattern of Figure b is shown in Figure c, exhibiting three pairs of diffraction spots corresponding to the facets of (200), (111), and (1 – 1 – 1). The pattern is further confirmed to be projected on the zone axis of [0 – 11], and the lattice spacings of these three facets are calculated to be 0.185, 0.213, and 0.213 nm, which are in good agreement with XRD patterns. Similarly, low-resolution TEM images of Pt(Fe, Co, Ni)3-700 NPs are exhibited in Figures d and S5, while the HR-TEM image is displayed in Figure e. Figure f reveals the histograms of the particle size distribution of Pt(Fe, Co, Ni)3 NPs corresponding to the TEM images in Figure a,d. The histograms both exhibit a narrow size distribution, and the average particle sizes of these two NPs are determined to be 5.09 and 6.86 nm severally, which are consistent with the particle sizes obtained from XRD patterns.

Figure 4

(a) Low-resolution TEM image of Pt(Fe, Co, Ni)3-600 NPs supported on carbon. (b) Atomic-scale HR-TEM image of Pt(Fe, Co, Ni)3/C-600 NP. (c) Fourier transform image of the NP in (b). (d) Low-resolution TEM image of Pt(Fe, Co, Ni)3-700 NPs on the carbon support. (e) Atomic-scale HR-TEM image of Pt(Fe, Co, Ni)3-700 NP. (f) Size distribution histograms of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700.

In order to assess the electrocatalytic activity of Pt(Fe, Co, Ni)3/C NPs, electrochemical measurements were investigated in depth with the rotating disk electrode (RDE) method, where cyclic voltammetry (CV) profiles and linear sweep voltammetry (LSV) curves were both carried out. The ORR polarization curves of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 NPs at different rotation speeds in O2-saturated 0.1 M HClO4 solution are displayed in Figure a,b, severally. According to the Koutecky–Levich (K–L) equation, there is a linear relationship between J–1 and ω–1/2 and the electron transfer number of NPs under different potentials can be obtained through the straight slope of K–L plots.[18] Based on the straight slopes from Figure a, the electron transfer numbers of Pt(Fe, Co, Ni)3/C-600 at 0.775, 0.835, 0.895, and 0.905 V are calculated to be 3.95, 4.00, 4.00, and 3.86, respectively, indicating that the Pt(Fe, Co, Ni)3/C-600 NPs are efficient for ORR, and the reaction undergoes a complete four-electron transfer process where O2 is directly reduced to H2O without the intermediate hydrogen peroxide. Similarly, the electron transfer numbers of Pt(Fe, Co, Ni)3/C-700 NPs are also determined to be 3.60, 3.60, 3.73, and 3.69, severally, (summarized in Table S2). As a reference, K–L plots and electron transfer numbers of commercial Pt/C are revealed in Figure S6, which is obviously insufficient compared with Pt(Fe, Co, Ni)3/C NPs catalysts under lower overpotentials. The electron transfer number approaching four indicates that the Pt(Fe, Co, Ni)3/C electrocatalysts exhibit high-efficiency and excellent performance toward ORR even under low overpotentials. Figure c displays the CV profiles measured in N2-saturated 0.1 M HClO4 solution at a rate of 50 mV s–1 in the range of 0.05–1.1 V. Apparently, the cathodic peak positions of Pt(Fe, Co, Ni)3/C-600 (0.759 V) and Pt(Fe, Co, Ni)3/C-700 (0.760 V) catalysts show a positive shift about 30 mV compared to that of commercial Pt/C (0.729 V), declaring a signally reduced absorption free energy of oxygenic species and a faster reduction rate on the surface of catalysts.[16] Beyond that, the electrochemical active specific surface areas (ECSAs) of Pt(Fe, Co, Ni)3/C-600, Pt(Fe, Co, Ni)3/C-700, and Pt/C are determined to be 81.4, 53.7, and 64.0 m2 g–1 through normalizing the charge integral of CV curves from 0.05 to 0.4 V to the total mass of Pt.

Figure 5

Koutecky–Levich plots of (a) Pt(Fe, Co, Ni)3/C-600 and (b) Pt(Fe, Co, Ni)3/C-700 at different potentials acquired from the ORR polarization curves. (c) CV profiles of synthesized Pt(Fe, Co, Ni)3/C catalysts and commercial Pt/C. (d) LSV curves of Pt(Fe, Co, Ni)3/C catalysts and commercial Pt/C. (e) Tafel plots of catalysts normalized to the total mass of Pt. (f) Comparison of mass activities of catalysts calculated at 0.85 and 0.9 V vs RHE.

The ORR polarization curves with deduction of the background current are exhibited in Figure d after pretreatment by repeating the potentials in the range of −0.2–0.8 V for 20 cycles in order to eliminate surface contamination. The ORR polarization curves were executed in the positive direction from 0.3 to 1.1 V at a sweep rate of 20 mV s–1. Basically, the polarization curves of Pt(Fe, Co, Ni)3/C catalysts are at the right side of the commercial Pt/C catalyst, revealing a positive shift in both onset potential and half-wave potential. To be precise, the onset potentials of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 catalysts are determined to be 1.031 and 1.019 V, respectively, surpassing that of Pt/C (1.003 V) by 28 and 16 mV, respectively. The half-wave potentials of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 catalysts were also confirmed to be 0.929 and 0.913 V, showing 48 and 32 mV higher than the commercial Pt/C catalyst (0.881 V). The LSV curves exhibited in Figure d demonstrate that the alloying of Pt, Fe, Co, and Ni in ordered configuration can expedite the kinetics of ORR and meanwhile lower the overpotential needed for the reaction, making the reduction of oxygen easier to occur and follow a more rapid dynamic procedure.

To better understand the utilization effects of Pt in the catalysts, Tafel plots normalized to the total amount of Pt are presented in Figure e with the potential between 0.83 and 1.02 V. It is clear that at an arbitrary potential, the specific kinetic current of the Pt(Fe, Co, Ni)3/C-600 catalyst is the highest and that of the commercial Pt/C catalyst is the minimum, signifying that Pt(Fe, Co, Ni)3/C catalysts utilize the noble metal Pt to a maximum extent with excellent property but at low amounts. In addition to this, the Tafel plot of the commercial Pt/C catalyst shows a higher slope at an arbitrary current, declaring that Pt(Fe, Co, Ni)3/C catalysts can significantly improve the reaction kinetics and lower the overpotential. Based on Tafel plots, the mass activities (MAs) of different catalysts at 0.85 and 0.9 V are calculated. As shown in Figure f, the Pt(Fe, Co, Ni)3/C-600 catalyst possesses the highest mass activity (MA) (3.189 mA μgPt–1 at 0.85 V, 0.605 mA μgPt–1 at 0.9 V), enhancing the activity by 11.8 and 6.6-fold over that of commercial Pt/C (0.269 mA μgPt–1 at 0.85 V, 0.091 mA μgPt–1 at 0.9 V) at 0.85 and 0.9 V versus reversible hydrogen electrode (RHE) . Similarly, the MA of Pt(Fe, Co, Ni)3/C-700 catalysts (1.696 mA μgPt–1 at 0.85 V, 0.322 mA μgPt–1at 0.9 V) also surpasses that of commercial Pt/C catalysts by 6.3 and 3.5 times, revealing excellent electrocatalytic performance toward ORR. It is worth mentioning that the calculated ECSA of commercial Pt/C (64.0 m2 g–1) is comparable to or in excess of that in the previous work,[12,32] and so does the MA (0.091 mA μgPt–1 at 0.9 V),[18,27,40,41] which demonstrates the reliability of the electrochemical measurements.

In order to explore the influence of particle sizes and the ordering effect on the holistic long-term durability of catalysts under acidic conditions, the stability performance was evaluated through accelerated durability tests (ADTs) between 0.6 and 1.0 V carried out in O2-saturated 0.1 M HClO4 solution at a sweep rate of 50 mV s–1. As shown in Figure a,b, there is an inconspicuous difference between the CV curves and ORR polarization curves of Pt(Fe, Co, Ni)3/C catalysts before and after ADTs, while the polarization curves of commercial Pt/C exhibit a distinct separation. More accurately, the degradation in half-wave potentials of different catalysts are determined to be 20, 5, and 41 mV for Pt(Fe, Co, Ni)3/C-600, Pt(Fe, Co, Ni)3/C-700, and commercial Pt/C, respectively, after 10 000 potential cycles. The Pt(Fe, Co, Ni)3/C NPs exhibit the degradation in half-wave potentials far below that of Pt/C, proving that the structurally ordered Pt(Fe, Co, Ni)3 NPs possess superior durability toward ORR under acidic conditions. On this basis, the MAs of three different electrocatalysts before and after 10 000 potential cycles are calculated, as shown in Figure c. To be exact, the remaining MAs of Pt(Fe, Co, Ni)3/C-600, Pt(Fe, Co, Ni)3/C-700, and commercial Pt/C catalysts are determined to be 58, 83, and 44%, respectively, after 10 000 potential cycles at 0.9 V (Table S5). Figure d shows the specific activities (SAs) of electrocatalysts before and after ADTs at 0.9 V, where the Pt(Fe, Co, Ni)3/C-700 catalyst stands out for its distinctive stability performance. The decays in SA of Pt(Fe, Co, Ni)3/C-600, Pt(Fe, Co, Ni)3/C-700, and Pt/C catalysts are determined to be 21.5, 1.5, and 41.5%, severally (Table S6). The mass and specific activities of catalysts at 0.85 V after ADTs are also revealed in Figure S7, where Pt(Fe, Co, Ni)3/C-700 still exhibits the optimal durability. The Pt(Fe, Co, Ni)3/C-700 catalyst reveals the minimal decay in half-wave potential, MA, and SA, while the decays of Pt/C reach the maximum, which further verifies the superior stability of structurally ordered Pt(Fe, Co, Ni)3/C NPs toward ORR than Pt/C catalysts.

Figure 6

(a) CV curves and (b) LSV polarization curves of Pt(Fe, Co, Ni)3/C catalysts and commercial Pt/C before and after ADTs of 10 000 cycles. Contrast diagrams of (c) mass activities and (d) specific activities of catalysts before and after ADTs at 0.9 V.

On the whole, the enhanced electrocatalytic performance of Pt(Fe, Co, Ni)3 NPs toward ORR can be interpreted as the following aspects: electronic structural modification and the ordering effect. Specifically, the integration of transition metals Fe, Co, and Ni into Pt alters the electronic structures of Pt and generates a strain effect, which remarkably enhances ORR activity.[42] Furthermore, under high-temperature and reducing atmosphere environments, the disordered alloy transforms into cubic L12-ordered structure spontaneously, which powerfully restrains the dissolution of transition metals in acid solution and makes the catalysts chemically stable.[26] Because of smaller particle sizes, Pt(Fe, Co, Ni)3/C-600 NPs have a higher proportion of surface atoms. The ECSAs data shown in Table S3 also declare that Pt(Fe, Co, Ni)3/C-600 exhibits more electrocatalytic active sites, which eventually generates the supreme electrocatalytic activity toward ORR. However, in the nanometer scale, NPs with smaller particle sizes tend to show a decreased surface percentage of (111) sites and an increased percentage of edge sites, which may reduce the SA as a consequence.[43] Meanwhile, NPs with smaller particle sizes have a higher surface energy, which provides impetus for particle aggregation and growth. Figure S8 exhibits the TEM images of Pt(Fe, Co, Ni)3/C-600 and Pt(Fe, Co, Ni)3/C-700 catalysts after ADTs of 10 000 cycles. Apparently, Pt(Fe, Co, Ni)3-600 NPs exhibit a conspicuous aggregation behavior, and distribution of Pt(Fe, Co, Ni)3-600 NPs on the carbon support becomes nonuniform during the ADT period. In contrast, Pt(Fe, Co, Ni)3-700 NPs remain its particle sizes and show a more homogeneous distribution on the carbon support. The aggregation and growth of Pt(Fe, Co, Ni)3-600 NPs is responsible for the relatively pronounced decay in activity after ADTs. Moreover, Pt(Fe, Co, Ni)3-700 NPs exhibit a higher degree of ordering seen from the obvious (110) ordered peak in XRD patterns, which dramatically restrains the dissolution of transition metals in acid solution. As a result, Pt(Fe, Co, Ni)3/C-700 NPs reveal the most superior stability after ADTs because of the balance between the ordered chemical structure and particle sizes.

Conclusions

In conclusion, this work demonstrates a facile and effective method for synthesis of quaternary structurally ordered L12-Pt(Fe, Co, Ni)3 NPs supported on carbon through the spray dehydration method on the solid surface accompanied by an annealing process. The orderly spatial configuration is investigated and confirmed, declaring the successful synthesis of novel structurally ordered alloy phase in the type of PtM3. On account of the chemically ordered structure together with electronic structural modification of Pt arising from the coalescence of transition metals, Pt(Fe, Co, Ni)3/C catalysts show boosted ORR activity (maximal MA 6.6 times higher than Pt/C at 0.9 V) and excellent stability performance (minimal loss of 17% in MA and 1.5% in SA). The stability behavior of catalysts is systematically explored to be tightly dependent on the coaction of particle sizes and ordering degree of NPs. The Pt(Fe, Co, Ni)3/C nanocatalysts reported in this work exploit Pt to the maximum extent to guarantee the remarkable electrocatalytic properties, serving as highly competitive substitutes for traditional catalysts in the stoichiometric forms of Pt3M and PtM. Beyond that, compared with the extensively adopted chemosynthesis method at present, the facile and effective synthetic strategy in this work can provide a fresh perspective on synthesizing late-model cost-effective Pt-based nanocatalysts for PEMFCs.

Experimental Section

Chemicals

Chloroplatinic acid (H2PtCl6·6H2O), ferric nitrate (Fe(NO3)3·9H2O), cobalt nitrate (Co(NO3)2·6H2O), nickel nitrate (Ni(NO3)2·6H2O), and PVP were purchased from Aladdin Reagent Co., Ltd. Vulcan XC-72R carbon black was purchased from Cabot Corporation. Nafion (5%) was purchased from DuPont and commercial Pt/C (20% Pt loading supported on XC-72R carbon black with an average size of 3 nm) was obtained from Johnson Matthey.

Synthesis of Pt(Fe, Co, Ni)3 NPs

Before the synthesis, the amount of all the chemicals and materials was calculated. Then, 300 mg of H2PtCl6·6H2O and a predetermined amounts, 234.0 mg of Fe(NO3)3·9H2O, 168.6 mg of Co(NO3)2·6H2O, and 168.4 mg of Ni(NO3)2·6H2O were mixed and dissolved in 40 mL of deionized water to ensure the proportions of Pt, Fe, Co, and Ni atoms in the solution are the same. For better dispersion of NPs on the carbon support, carbon black was added in advance. A specific amount of Vulcan XC-72R carbon black (853.92 mg) was added to achieve a 20 wt % alloy loading on the carbon support. PVP of the same mass as carbon black (853.92 mg) was dissolved in the suspension as well to prevent conglomeration of NPs and the carbon support; then, the suspension went through an ultrasonic processing for 40 min. After all these, the spray dehydration method was carried out by squirting the well-dispersed suspension onto a quartz plate at a temperature of 300 °C. As the solvent evaporated fairly soon on the surface of the quartz plate, the mixed precursor at the molecular level was acquired on the quartz plate. The remaining mixture was collected and then gradually heated in the tube furnace while pumping all the time to remove vestigial PVP as PVP will volatilize under high-temperature conditions and vacuum environment. In the end, the desired NPs were obtained by annealing the precursor at 600 and 700 °C for 1 h under a H2 atmosphere.

Synthesis of PtAuCuNi NPs

The same synthetic strategy was applied to prepare PtAuCuNi/C NPs where H2PtCl6·6H2O, HAuCl4·3H2O, Cu(NO3)2·6H2O, and Ni(NO3)2·6H2O were employed and then they underwent the spray dehydration method and an annealing process.

Morphological, Structural, and Elemental Characterization

The structural analysis of NPs was carried out using an XRD (Rigaka Ultima IV multipurpose XRD) with Cu Κα radiation at the rate of 5° min–1 within the range of 20°–90°. The magnetic properties of NPs were carried out by using a Superconducting Quantum Interface Device (SQUID; Quantum Design) with the magnetic field from 0 to 2 T at 300 K. The morphological information was obtained by transmission electron microscopy (TEM) and HR-TEM (JEM-2100HR JEOL) operated at 200 kV. The STEM–EDX elemental mapping images were acquired by using an FEI Tecnai F-20 microscope (FEI TF-20; FEI) operated at 200 kV. The quantitative analysis of the element content for Pt, Fe, Co, and Ni was detected by ICP-AES (ICP-AES; Thermo Jarrell Ash).

Electrochemical Measurements

The working electrodes were prepared by dispersing 4 mg of catalysts into 2 mL of specially formulated solution consisting of deionized water, isopropanol, and Nafion (5%) with a volume ratio of 3:1:0.05. Then the suspension went through an ultrasonic processing for 40 min to obtain the well-dispersed catalyst ink. The catalyst ink (25 μL) was afterward loaded on the glassy carbon electrode with a diameter of 5 mm. Finally, the uniform catalyst film over the entire glassy carbon surface was acquired by rotating the glassy carbon electrode with a speed of 300 rpm until the ink dried out completely. The electrochemical measurements were carried out using an electrochemical workstation (CHI 660D; CH Instruments) together with a RDE (Pine Research Instrumentation). A platinum wire was used as the counter electrode, while an Ag/AgCl electrode filled with 3 M KCl solution was employed as the reference electrode. The CV curves were carried out between 0.05 and 1.1 V versus RHE at 50 mV s–1 in N2-saturated 0.1 M HClO4 solution. The ORR activities were then measured in O2-saturated 0.1 M HClO4 solution at a rate of 20 mV s–1. The kinetic current density JK can be achieved by the K–L equationJ is the actually measured current density and JL is the diffusion limited current density. Meanwhile, JL can be expressed by the Levich equationwhere n is the electron transfer number; F is Faraday’s constant (96 485 C mol–1); CO is the concentration of molecular oxygen in 0.1 M HClO4 solution (1.2 × 10–6 mol cm–3); D is the diffusion coefficient of O2 in 0.1 M HClO4 solution (1.93 × 10–5 cm2 s–1); ν is the viscosity of the electrolyte (1.01 × 10–2 cm2 s–1); and ω is the angular frequency of rotation.[18,44]

The ECSAs are determined by normalizing the charge integral of CV curves from 0.05 to 0.4 V versus RHE to the mass of Pt. For the ADTs, 10 000 potential cycles of consecutive CV scanning were conducted in the range of 0.6–1.0 V at a rate of 50 mV s–1 in O2-saturated 0.1 M HClO4 solution. All of the electrochemical measurements were accomplished at room temperature, and all the polarization curves were converted to be relative to RHE after taking into consideration the electrode potentials and pH of the solution.