Anion-Modulated Platinum for High-Performance Multifunctional Electrocatalysis toward HER, HOR, and ORR

Summary

Efficient electrocatalyst toward hydrogen evolution/oxidation reactions (HER/HOR) and oxygen reduction reaction (ORR) is desirable for water splitting, fuel cells, etc. Herein, we report an advanced platinum phosphide (PtP₂) material with only 3.5 wt % Pt loading embedded in phosphorus and nitrogen dual-doped carbon (PNC) layer (PtP₂@PNC). The obtained catalyst exhibits robust HER, HOR, and ORR performance. For the HER, a much low overpotential of 8 mV is required to achieve the current density of 10 mA cm⁻² compared with Pt/C (22 mV). For the HOR, its mass activity (MA) at an overpotential of 40 mV is 2.3-fold over that of the Pt/C catalyst. Interestingly, PtP₂@PNC also shows exceptional ORR MA which is 2.6 times higher than that of Pt/C and has robust stability in alkaline solutions. Undoubtedly, this work reveals that PtP₂@PNC can be employed as nanocatalysts with an impressive catalytic activity and stability for broad applications in electrocatalysis.

Graphical Abstract

Highlights

• •PtP₂@PNC is synthesized under ambient pressure and moderate temperatures
• •The formed PtP₂@PNC exhibits outstanding performance toward HER, HOR, and ORR
• •The synergistic effect between PtP₂ and PNC is responsible for the high activity

Chemistry; Electrochemistry; Electrochemical Energy Conversion; Materials Science; Energy Materials

Introduction

The increasing requirement for clean, eco-friendly, and sustainable energy has prompted the development of energy technologies to focus more on renewability, efficiency, and environmental protection (Steele et al., 2001; Zhang et al., 2016a, 2016b; Mallouk et al., 2013). In view of its unique properties, hydrogen perfectly accords with those requirements, thus making it the most promising candidate to replace fossil fuels in the future. Undoubtedly, among those hydrogen generation technologies, electrochemical water splitting is an important method, meanwhile hydrogen evolution reaction (HER) is a vital kinetic process of electrochemical water splitting (Balat et al., 2009; Turner, 2004). Besides, efficient hydrogen oxidation reaction (HOR) and oxygen reduction reaction (ORR) can improve the energy efficiency, which also plays an important role in the sustainable energy devices, especially for fuel cells and metal-air batteries (Lee et al., 2011; Huang et al., 2015). However, on the one hand, the sluggish kinetics of ORR generally requires a large overpotential to drive the reaction. On the other hand, it is challenging but urgent to rationally design a practical and efficient electrocatalyst with better performance and lower cost. In addition, alternative electrocatalysts for HER, HOR, and ORR with various structures and chemical compositions have already been widely explored (such as precious metals, non-noble metals, and even metal-free materials) (Zheng et al., 2014a, 2014b, 2016; Huang et al., 2017; Shan et al., 2019; Liu et al., 2014; Popczun et al., 2013; Morales-Guio et al., 2014; Bu et al., 2017; Ma et al., 2015; Zhu et al., 2017). However, Pt group metal-based (PGM-based) catalysts are still the best materials for these electrocatalytic reactions. Their high cost, scarcity, and poor stability of PGM-based catalysts have driven efforts to reduce Pt usage toward their commercialization in energy devices.

During the past few years, intense investigations have been made in this field and a bunch of optimized strategies has been designed to lower the Pt loading and improve utilization efficiency. Among those strategies, the most widely applied method to reach this significant challenging goal is to alloy Pt together with other low-cost 3d transition metals (Tian et al., 2020a, 2020b; Wu et al., 2013; Porter et al., 2013). Owing to the optimized structure and electronic effects caused by M (M = Fe, Co, Ni, Cu, Cr, Ti, V, Pb, Mn, etc.), the PtM alloys present significant potentials in improving the electrochemical activity for a series of catalytic reactions (such as HER, ORR, methanol/ethanol oxidation reaction [MOR or EOR]) (Huang et al., 2017; Bu et al., 2017; Li et al., 2018; Cui et al., 2014; Ma et al., 2020a, 2020b; Wang et al., 2017; Ma et al., 2020a, 2020b; Fan et al., 2016). Nevertheless, the M in the PtM-based nanoparticles (NPs) is inclined to dissolve during the highly oxidizing and/or acidic environment, which is inevitable to degrade proton exchange membranes in fuel cells or to deteriorate the electrolyte solutions in water splitting during the practical application (Wang et al., 2011; Wang et al., 2015; Wang et al., 2018a, 2018b; Dubau et al., 2011).

It is worth noting that previous researches usually focus on 3d transition metal alloying Pt. However, the long-term durability of alloy catalysts, due to the second metal dissolution and particle growth, still remains a huge challenge (Cui et al., 2015a, 2015b). In addition, for disordered Pt alloying M structure, the atomic positions are occupied randomly by M and Pt. Thus, the Pt alloy has randomly distributed active centers and changing surface composition. Oppositely, if Pt combines anions (such as P, S, Se, Te) and forms Pt compounds that have a fixed crystal structure, provide predictable control of the structure. As an example, non-metal P anion combination in Pt formed platinum phosphide (PtP₂) has been rarely reported but is expected to improve the intrinsic activity of electrocatalysts, because P atoms with more electronegativity can grab electrons from Pt atoms, significantly tuning the compound electronic structure (Shi and Zhang, 2016). More importantly, for transition metal phosphides (TMPs), it is reported that improving the atomic percentage of P may significantly enhance HER activity. Additionally, to further improve the dispersibility and stability of metal compounds, the heteroatoms (N, P, S, etc.)-doped carbon materials are generally used as the support of metal compounds, which can effectively prevent the migration and aggregation of NPs. At the same time, such heteroatom-doped carbon materials are able to synergistically boost the activities of metal compounds-based catalysts.

Inspired by the above-mentioned description, in this work, PtP₂ NPs are fabricated at moderate temperatures and ambient pressure in terms of P, N dual-doped carbon layer (PNC) encapsulation (PtP₂@PNC). As expected, the PtP₂@PNC catalyst with a low Pt mass loading (3.5 wt %) exhibits outstanding HER, HOR, and ORR, performance even better than commercial Pt/C (20 wt %). On the one hand, the high catalytic activity can be attributed to that the negatively charged P atoms can catch electrons from Pt atoms, significantly tuning the compound electronic structure, and therefore enhancing the Pt activity. On the other hand, the unique structural relationship of the P, N dual-doped C mainly provides a protective shield to inhibit PtP₂ migration and agglomeration, as well as facilitate the conductivity and charge/mass transport in the materials during the electrochemical process.

Result and Discussion

As illustrated in Figure 1, PtP₂@PNC was fabricated by a facile pyrolysis method. First, phytic acid (PA) cross-linked platinum complexes (PtPA) were formed based on the strong chelating ability of PA (Tian et al., 2020a, 2020b; Zhou et al., 2019). Then, PtPA-dicyandiamide supermolecular aggregate was produced via a simple cooperative assembly process in water at room temperature (Zhang et al., 2016a, 2016b). Finally, the PtP₂ NPs embedded in PNC layers were obtained by a thermal treatment of the PtPA-dicyandiamide precursor at 900°C under Ar atmosphere. It is worth noting that, without introducing the PA precursor in our experiments, only Pt nanoparticles supported on N-doped carbon (Pt-NC) was obtained (Figure S1). That is, PA plays an important role in the TMPs preparation process and can be used as a green phosphorus source. As a reference, PNC layers also can be prepared by the same synthesis strategy but without adding the H₂PtCl₆ precursor (Figures S2–S4). The formation of PNC layers can be described to the presence of P- and N-containing functional groups in the PA and dicyandiamide precursors during the carbonization process, leading to simultaneous self-doping of P and N into the carbon layers. Furthermore, PNC-supported Pt (Pt-PNC) has also been prepared by using the similar method (Figure S5).

Figure 1.

Schematic Illustration of the Synthesis of PtP₂@PNC Catalysts

The obtained product was first characterized by powder XRD. As displayed in Figure 2A, the 2θ peaks at 27.1°, 31.4°, 44.9°, 53.3°, 55.9°, 72.2°, and 74.4° are assigned to the (111), (200), (220), (311), (222), (331), and (420) crystal planes, respectively, indicating a typical crystalline cubic PtP₂ phase (PDF No. 65-0004, space group: Pa-3, a_o = b_o = c_o = 5.695 Å). Subsequently, XPS was carried out and illustrated in Figures 2B–2F. The survey scan (Figure 2B) shows the presence of all elemental components including Pt, P, C, N, and O. The fitting spectra of Pt4f indicate four peaks, corresponding to oxidized Pt and Pt. The peaks located at 71.6 and 74.9 eV are assigned to Pt4f_7/2 and Pt4f_5/2 of Pt (Figure 2C) (Kalinkin et al., 2010; Yang et al., 2018). Compared with Pt4f_7/2 peak located at 71.1 eV of metallic Pt, the binding energy (BE) in PtP₂@PNC is positive shifted, suggesting Pt in PtP₂@PNC exhibits a slightly positive charge (δ+). The rest of two peaks at 73.2 and 76.5 eV are ascribed to the oxidized Pt. Regarding P2p core-level spectra (Figure 2D), the subpeaks at 133.4 and 132.5 eV are attributed to the presence of P-O and P-C species, as well as two weak subpeaks observed at about 129.3 and 130.2 eV are assigned to the Pt−P bond, suggesting a successful formation of PtP₂@PNC (Zhang et al., 2015). The BE of 129.3 eV for P 2p_3/2 exhibits a negative shift from that of P⁰ (130.2 eV) (Tian et al., 2020a, 2020b), which indicates that P is negative charge (δ−). These results points to weak electron density transfer from Pt to P in the PtP₂@PNC.

Figure 2.

Structure and Microscopy Characterization of PtP₂@PNC

(A) XRD pattern of PtP₂@PNC catalysts.

(B–F) XPS spectra of (B) survey scan, (C) Pt 4f regions, (D) P 2p regions, (E) N 1s regions, and (F) C 1s regions for PtP₂@PNC.

(G–I) (G) TEM, (H) HAADF-STEM, and (I) HRTEM images of PtP₂@PNC (inset: corresponding FFT pattern).

The N 1s spectrum (Figure 2E) can be fitted into three separated peaks with binding energies (BEs) at 398.6. 400.5, and 401.2 eV corresponding to pyridinic-N, pyrrolic-N, and graphitic-N species, respectively. As reported, graphitic-N is considered to improve the diffusion-limited properties, while pyridinic-N, as an active site for ORR, boosts the onset potential, electrical conductivity, and surface wettability (Amiinu et al., 2017; Zhao et al., 2013). In addition, it is demonstrated that pyridinic-N, pyrrolic-N, and graphitic-N species are beneficial for HER electrocatalysis (Zheng et al., 2014a, 2014b). The C 1s XPS spectrum (Figure 2F) demonstrates the existence of C=C (284.6 eV), C–O/C–N/C–P (285.9 eV), C=O/C=N (286.4 eV), and C–C=O (289.3 eV) bonds in PtP₂@PNC. It should be noteworthy that the aforementioned C–P and C–N bonding further indicate the successful doping of P and N into carbon layers. In addition, the carbon component was further characterized by the Raman spectrum. As illustrated in Figure S6, two strong peaks at ∼1,345 and 1,598 cm⁻¹ are corresponded to the D- and G-band, respectively. The intensity ratio of I_D/I_G is found to be 1.04, indicating that the PtP₂@PNC sample contains many defective carbons. Moreover, the content of Pt loaded on PNC is confirmed to be about 3.5 wt % by ICP-AES.

The morphology of PtP₂@PNC was characterized by SEM and TEM. As shown in Figure S7A, numerous carbon nanosheets are observed from the SEM images. The low- and high-magnification TEM images of PtP₂@PNC indicate homogeneous dispersion of PtP₂ NPs in PNC layers (Figures S7B and 2G). High-angle annular dark-field scanning TEM (HAADF-STEM) further clearly demonstrates the presence of numerous small PtP₂ NPs (Figure 2H). As shown in Figures S8 and S9, the average size of the PtP₂ NPs is about 11.7 ± 2.5 nm (∼100 nanoparticles were measured). The lattice fringe with an interplanar distance of 3.288 Å is clearly displayed in the high-resolution TEM (HRTEM) image (Figure 2I), corresponding to the (111) facet of cubic PtP₂ phase. Additionally, the fast Fourier transform (FFT) pattern (Figure S10B) deriving from the PtP₂ (circled in Figure S10A) verifies the existence of (311), (211), and (200) facets. The HRTEM images in Figures 3A and 3B further reveal that PtP₂ NPs are embedded within P and N dual-doped carbon with ∼1–6 layers (PNC layers are indicated by the white arrows). Simultaneously, HAADF-STEM and the corresponding energy dispersive spectrometer (EDS) elemental mappings (Figure 3C) distinctly confirm the homogeneous distributions of C, N, Pt, and P elementals. All of these characterizations (XRD, XPS, and TEM) indicate that PtP₂@PNC was successfully achieved.

Figure 3.

Microscopy Characterization of PtP₂@PNC

(A–C) (A and B) HRTEM images of PtP₂@PNC and PNC are indicated by the white arrows. (C) HAADF-STEM image and corresponding EDX elemental maps of C (yellow), N (cyan), Pt (blue), P (green).

HER tests were evaluated in H₂-saturated acid solutions (0.5 M H₂SO₄) in a two-compartment, three-electrode cell with a scan rate of 5 mV s⁻¹. All the LSVs curves were not iR corrected. Before the tests, the saturated calomel electrode (SCE) was calibrated by continually bubbling the solution (0.5 M H₂SO₄) with ∼1 atm of research-grade H₂ (g) using a clean platinum electrode as the working electrode (Figure S11). As shown in Figure 4A nearly zero onset potential is observed for both PtP₂@PNC and commercial Pt/C. To attain the current density (j) of 10 milliamperes per square centimeter (mA cm⁻²), the catalyst for PtP₂@PNC only needs a much lower overpotential (8, 12 mV without iR correction, Figure S12) compared with commercial Pt/C (22 mV). Furthermore, to the best of our knowledge, the HER performance of PtP₂@PNC is superior to the PNC layers (183 mV @ 10 mA cm⁻²), Pt-PNC (19 mV @ 10 mA cm⁻²), Pt-NC (59 mV @ 10 mA cm⁻²) and all of the noble-metals, non-precious metal-based catalysts as well as non-metal HER catalysts (Figures S13, 14, and 4B and Table S1). Tafel analysis of the PtP₂@PNC nanomaterial in 0.5 M H₂SO₄ indicates a Tafel slope of ∼30 mV dec⁻¹ in the region of η = 5–30 mV (Figure 4C). This value is consistent with the known mechanism of the HER on commercial Pt/C. At higher overpotentials (η = 60–100 mV), the Tafel slope increased to ∼122 mV dec⁻¹. This value does not match the expected Tafel slopes of 29, 38, and 116 mV dec⁻¹, each of which correlates with a different rate-determining step of the HER. Furthermore, Figure S15 displays the electrochemical impedance spectroscopy (EIS) data of different samples. From the plots, we can learn that PtP₂@PNC shows more favorable HER kinetics.

Figure 4.

HER Activities of PtP₂@PNC and Commercial Pt/C

(A–F) (A) HER polarization curves for PtP₂@PNC and commercial Pt/C recorded at 5 mV s⁻¹. (B) The HER activity comparison between PtP₂@PNC with other HER catalyst, references are detailed in Table S1. (C) Corresponding Tafel slops. (D) Time-dependent current density curve of the PtP₂@PNC in 0.5 M H₂SO₄ (without iR correction). Polarization curves for (E) PtP₂@PNC and (F) commercial Pt/C initial and after 1,000 CV scanning between +0.1 and −0.2 V versus RHE.

(G–H) (G) The theoretical model of PtP₂@PNC used in DFT calculations. (H) Calculated free energy diagram of the catalyst samples.

Besides the high HER catalytic activity, stability is also another important parameter to promote the materials' practical application. As shown in Figure 4D, PtP₂@PNC presents good stability with a slight degradation of the current density even after 24 h of the continuous electrochemical scan. Similarly, it also maintains high activity after being subjected to an accelerated degradation test for certain CV cycles (Figure 4E). It presents that PtP₂@PNC can sustain the current density of 10 mA cm⁻² with only a small potential degradation ∼3 mV after 2,000 CV cycles, whereas the commercial Pt/C deteriorates by ≈ 6 mV even after 1,000 CV cycles (Figure 4F), indicating the superior durability for the PtP₂/PNC catalyst. To further evaluate the catalyst durability, we conducted both TEM and XPS analysis of PtP₂@PNC after electrochemical stability tests. As illustrated in Figure S16, the TEM image shows that the PtP₂ NPs still maintain good dispersion without obvious migration and aggregation owing to the robust thin PNC layers entrapment. More importantly, the similarity of high-resolution Pt4f, P2p, N1s, and C1s XPS spectra (Figure S17) of the fresh and post-HER PtP₂@PNC materials further demonstrates the retention of the materials in terms of composition, confirming its outstanding robustness for HER electrocatalysis. However, after the HER stability test, part of P species is detected in the electrolyte by the ICP-AES (Table S2), possibly caused by the detachment of the catalyst from the electrode surface due to the vigorous gas evolution (Andronescu et al., 2017). Furthermore, the HER activity of PtP₂@PNC in neutral and alkaline media was also investigated. As shown in Figure S18, the PtP₂@PNC materials also show nearly commercial Pt/C activity under neutral and alkaline conditions. Specifically, to reach the current density of 10 mA cm⁻², PtP₂@PNC requires overpotentials of 64 and 45 mV in 1.0 M phosphate buffered saline (PBS) and 1.0 M KOH solutions, respectively. All of the above results suggest that PtP₂@PNC possesses Pt-like HER catalytic performance under multi-pH conditions.

Based on the above results, the high HER performance of PtP₂@PNC can be ascribed to the following points. (1) As described in Figures 4G, 4H, and S19, when PtP₂ is incorporated with PNC, the resultant PtP₂@PNC yields an optimal adsorption free energy of H (ΔG_H∗), which is even lower than that of Pt, indicating that the introduction of P to Pt could weaken the binding energy between Pt and hydrogen atoms, which further facilitates the hydrogen evolution during the HER process (Robinson et al., 2017; Luo et al., 2018; Wang et al., 2018a, 2018b). (2) As evidenced by the XPS, the main Pt 4f_7/2 peak in PtP₂@PNC is at 71.6 eV, representing positive shifts in the binding energies of +0.05 eV, compared with the binding energy of the Pt 4f_7/2 peak in pure Pt (71.1 eV). This shows the electronic structure of metal Pt was changed after introducing P. In other words, negatively charged P atoms can capture electrons from Pt atoms and play an important role as Lewis base to work with positively charged protons in the HER process (Shi and Zhang, 2016; Guo et al., 2018; Zhuang et al., 2016). Like metal complex and [NiFe] hydrogenases catalysts, the P and Pt serve as the hydride-acceptor and proton-acceptor center, respectively, improving the HER catalytic activity (Popczun et al., 2013; Kibsgaard et al., 2014; Cui et al., 2015a, 2015b). (3) The PtP₂ NPs uniformly encapsulated in the PNC layers can effectively prevent the aggregation and migration of those PtP₂ active centers, thus endowing the PtP₂@PNC superior catalytic stability. (4) The presence of the PNC layer in the catalyst may further improve the conductivity of the PtP₂@PNC, therefore enhancing the electron transfer during the HER process. As reported P and/or N doping carbon could improve HER activity to some extent in comparison with pure carbon materials (Zhang et al., 2015; Zheng et al., 2014a, 2014b). In other words, the P, N dual-doped carbon materials are able to synergistically improve the activities of PtP₂@PNC catalysts. Moreover, we conducted a band structure analysis to examine the electronic-coupling effect between PNC and PtP₂. As illustrated in Figure S20, pure PtP₂ displays a band gap of 0.3 eV, while coupling PtP₂ with PNC, the band gap of PtP₂@PNC is decreased to 0.087 eV. Therefore, after coupling PtP₂ with PNC, the charge density of the PtP₂@PNC is redistributed in the form of an apparent electron transfer from the conductive PNC to PtP₂, leading to electron enrichment of the PtP₂ layer. Besides, the total density of states (DOS) in PtP₂@PNC is quite different from that of PtP₂. These investigations further demonstrate that the coupling of PNC layers strongly influences the electronic structures of PtP₂ in the PtP₂@PNC by enhancing electron mobility and HER performance. (5) PtP₂@PNC has a high electrochemically active surface area (ECSA: ≈ 510 cm²) (Figure S21), which can favor the effective accessibility of the intrinsic active sites. Such large ECSA benefits from the high BET specific surface area (≈159.3 m² g⁻¹) (Figure S22).

Additionally, it also remains a huge challenge in developing low-Pt HOR catalysts under the acidic condition for the widespread employment of proton-exchange membrane fuel cell (PEMFC). Therefore, the HOR activity for PtP₂@PNC was examined by using a rotating disk electrode (RDE) test with a scan rate of 2 mV s⁻¹ in H₂-saturated 0.1 M HClO₄ solution. For comparison, we also tested the HOR activities of Pt-PNC and bare PNC (Figure S23). As shown in Figure 5A, the HOR mass activity (MA) of PtP₂@PNC is higher than that of commercial Pt/C. Figure 5B displays the HOR polarization curves on PtP₂@PNC at different rotating speeds from 400 to 1,600 rpm. When the current is controlled by both reaction kinetics and H₂ diffusion, Koutecky-Levich equation applies (Hunt et al., 2016):

$$\frac{1}{J}=\frac{1}{J_l}+\frac{1}{J_k}=\frac{1}{J_k}+\frac{1}{B{\omega }^2}$$ (Equation 1)

$$B=0.62n{C}_0{D}_0^{\frac{2}{3}}{v}^{\frac{-1}{6}}$$ (Equation 2)

$$J_k=nFkx{C}_0$$ (Equation 3)

where J, J_l, and J_K are the measured, kinetic, and diffusional current densities, respectively. F is the Faraday constant, D₀ is the diffusivity (4.5 × 10⁻⁵ cm²·s⁻¹) of hydrogen in 0.1 M HClO₄, n is the electron transfer number in HOR, ν is the kinetic viscosity of the electrolyte (0.008 cm²·s⁻¹), C₀ is the solubility of hydrogen in 0.1 M HClO₄ (7.2 × 10⁻⁷ mol·cm⁻³), and ω is the rotating speed. By fitting the data at an overpotential of 0.15 V with Equations (1)–(3), a linear plot of ω^−1/2 with j⁻¹ was obtained (inset of Figure 5B). The slope is 5.49 cm² mA⁻¹∙s^−1/2, which is reasonably closed to the two-electron transfer of HOR. Furthermore, as illustrated in Figure 5C, the corresponding MA of PtP₂@PNC and commercial Pt/C were calculated by the loading amount of Pt metal, respectively. The MA value for PtP₂@PNC is 0.172 A mg⁻¹_Pt, which is nearly 2.3 times greater than that of the commercial Pt/C (0.074 A mg⁻¹_Pt for MA). In fact, ΔG_H∗ is also an important descriptor for HOR (Wang et al., 2019). According to the Sabatier principle, ideal HOR electrocatalysts should have ΔG_H∗ close to zero. The |ΔG_H∗| value is calculated to be 0.09 eV on Pt(111) from density functional theory (DFT) calculations. As for PtP₂@PNC, the |ΔG_H∗| is lower than that of Pt, indicating that the introduction of P to Pt could weaken the hydrogen adsorption. All of these indicate that PtP₂@PNC is a promising HOR catalyst in acidic electrolytes for the widespread employment of PEMFC.

Figure 5.

HOR and ORR Activities of PtP₂@PNC and Commercial Pt/C

(A–C) (A) Polarization curves of PtP₂@PNC and Pt/C for HOR in H₂-saturated 0.1 M HClO₄. (B) HOR polarization curves on PtP₂@PNC catalyst at different rotation rates from 400 to 1,600 rpm. The inset image shows a Koutecky-Levich plot at 0.15 V versus RHE. (C) Mass activity at 40 mV of the PtP₂@PNC and Pt/C.

(D–F) (D) LSV curves of PtP₂@PNC and commercial 20 wt % Pt/C catalyst at 5 mV s⁻¹ at a rotating speed of 1,600 rpm obtained in O₂-saturated 0.1 M KOH solutions. (E) Mass activity of the Pt/C and PtP₂@PNC for ORR at 0.85 and 0.9 V, respectively. (F) Current versus time (i-t) chronoamperometric curves of the PtP₂@PNC and Pt/C at the rotation speed of 1,600 rpm and the constant potential of −0.3 V.

(G-I) (G) CVs of PtP₂@PNC and Pt/C in 0.1 M HClO₄ solutions. (H) LSV curves of PtP₂@PNC and commercial 20 wt % Pt/C catalysts at 5 mV s⁻¹ at a rotating speed of 1,600 rpm obtained in O₂-saturated 0.1 M HClO₄ solutions. (I) Histogram of mass and specific activities of PtP₂@PNC and Pt/C electrocatalysts at 0.9 V versus RHE.

To investigate the multifunctional catalytic activities, we also probed the ORR activity of the PtP₂@PNC samples in basic solutions. LSVs were measured to investigate the kinetics and mechanism of PtP₂@PNC in O₂-saturated basic solutions (0.1 M KOH) at a rotation rate of 1,600 rpm (Figure 5D). Its onset and half-wave potentials for PtP₂@PNC is about 0.97 and 0.82 V, respectively, which exhibits nearly identical ORR activity compared with the 20 wt % commercial Pt/C. However, owing to the limited 3.5 wt % content of Pt in PtP₂@PNC, the MA of PtP₂@PNC is far greater than that of commercial Pt/C. As illustrated in Figure 5E, PtP₂@PNC further exhibits a remarkable MA at 0.85 and 0.9 V, which is nearly 2.6 and 2.0 times higher than that of commercial Pt/C. Additionally, compared with PtP₂@PNC, PNC layers and Pt-PNC show an even worse ORR activity, as confirmed by their more negative onset potential and lower steady-state current density (Figures S24 and S25). To further investigate the ORR kinetics, the LSVs of PtP₂@PNC were scanned at different rotating speeds from 400 to 2,000 rpm (Figure S26A). Koutecky-Levich plots (J⁻¹ versus ω^−1/2) were obtained to better understand the number of electrons transferred (n) during ORR (inset of Figure S26B). And the calculated n value of PtP₂@PNC is approximately equal to 3.96, which confirms the preferred four-electron transfer reaction for PtP₂@PNC during the ORR process (Zhang et al., 2015; Greeley et al., 2019; Wang et al., 2012). In addition, to understand the possible mechanism for the improved ORR performance of PtP₂@PNC, DFT calculations were also carried out. The O₂ adsorption energy (ΔE_O2) on Pt is 3.3 eV (Figure S27), which is larger than that on PtP₂@PNC (1.8 eV). In other words, there is easier desorption of oxygen molecules on PtP₂@PNC than on Pt, further enhancing the O₂ reduction performance.

Additionally, long-term stability is also a critical factor to evaluate good ORR material. So, the amperometric i-t curve tests were further applied to explore the durability of the PtP₂@PNC. The chronoamperometric measurement results of PtP₂@PNC and commercial Pt/C are depicted in Figure 5F. After 24-h i–t test, the current of PtP₂@PNC exhibits a slight decrease (∼8%). Nevertheless, the current of Pt/C shifts negatively by about ∼14% after only 3-h test, suggesting the current loss of PtP₂@PNC is much smaller than the commercial Pt/C. Such excellent stability of PtP₂@PNC can be mainly attributed to the improved phosphorus and nitrogen-doped carbon structure (Pu et al., 2017; Qin et al., 2018, 2019). Additionally, the robust PNC layers could confine the PtP₂ NPs and prevent the PtP₂ active centers from aggregation. The morphology data of PtP₂@PNC catalyst after the stability test were characterized by TEM. As shown in Figures S28 and S29, TEM suggests that negligible change has been observed for the morphology of catalysts after HOR and ORR stability test, which further demonstrates that the catalyst has good durability.

It is noteworthy that Figures 5G–5I and S30–S32 present the ORR polarization curves of Pt/C and PtP₂@PNC materials in O₂-saturated 0.1 M HClO₄ electrolytes. The PtP₂@PNC catalyst has similar ECSA to the commercial Pt/C (Figures 5G and S30). Furthermore, PtP₂@PNC shows an ORR onset potential of 0.87 V and E_1/2 of 0.74 V, which are slightly different from commercial Pt/C (onset potential: 0.94 V; E_1/2 = 0.85 V) (Figures 5H and S31). Additionally, commercial Pt/C exhibits MA and specific activity of 48.2 mA mg_Pt⁻¹ and 0.066 mA cm⁻² at 0.9 V versus RHE, respectively, which are slightly better than that of commercial PtP₂@PNC (17.6 mA mg_Pt⁻¹ and 0.0 38 mA cm⁻² at 0.9 V vs. RHE) (Figure 5I). It is worth noting that the ORR path of PtP₂@PNC is nearly four electron transfer path in the acidic electrolytes based on K-L formula (Figure S32).

More importantly, in order to achieve a comparable or even superior ORR activity of PtP_2@PNC materials to the Pt/C, we increased the PtP₂ content in the compounds by adding the H₂PtCl₆ (∼300 mg) precursor during the synthesis process (the obtained materials named as PtP₂@PNC-3, ∼24 wt % Pt). As illustrated in Figure S33A, PtP₂@PNC-3 shows a higher ORR activity than that of commercial Pt/C in alkaline solutions. In addition, it further exhibits identical ORR activity compared with the 20 wt % commercial Pt/C in acidic solutions (Figure S33B).

For industrial application in fuel cells, besides superior activity and high durability, a good catalyst should be able to afford the possible methanol (CH₃OH) crossover and carbon monoxide (CO) poisoning, which may significantly harm the performance of the full cell (Zitolo et al., 2015; Dai et al., 2015). Thus, we measured the i-t response of the PtP₂@PNC in the presence of CO and methanol, respectively. As shown in Figure S34A, when introducing 3.0 M methanol into the 0.1 M KOH solutions, the current density of commercial Pt/C instantaneously decreases, indicating the occurrence of methanol oxidation. In contrast, the ORR current density of PtP₂@PNC only shows a slight change, suggesting that PtP₂@PNC possesses better methanol resistance than that of Pt/C. Similarly, PtP₂@PNC has no CO poisoning, whereas Pt/C is fast poisoned with a gradually dropped current density (Figure S34B). These results demonstrate that PtP₂@PNC has much higher ORR activity and selectivity than that of commercial Pt/C and is free from the CH₃OH and CO poisoning, promising for practical applications in fuel cells and other energy-related conversion and storage devices.

Conclusion

In summary, we have achieved high-efficiency PtP₂ materials embedded in PNC layer with a low Pt mass loading (only 3.5 wt %) by a facile solid-state pyrolysis approach under ambient pressure and moderate temperatures. As expected, the obtained PtP₂@PNC catalyst possesses excellent HER activity (8 mV @j = 10 mA cm⁻²) and durability in acidic solutions. Such high HER activity is superior to not only commercial Pt/C material but also all the reported HER catalysts. For the HOR, its mass activity (MA) at an overpotential of 40 mV is 2.3-fold over that of the commercial Pt/C catalyst. Furthermore, the PtP₂@PNC catalyst exhibits nearly 2.6 times ORR MA than that of 20 wt % commercial Pt/C, as well as greatly improved ORR stability and fuel tolerance than that of Pt/C. The outstanding HER, HOR, and ORR catalytic performances of our PtP₂@PNC catalyst make PtP₂ a promising multifunctional catalyst for wide applications in energy conversion and storage devices.

Limitations of the Study

This study has demonstrated an advanced multifunctional catalyst for hydrogen evolution/oxidation reactions and oxygen reduction reaction from experimental and theoretical perspectives. The experimental results suggested that the catalysts illustrated high HER, HOR, and ORR performance. DFT calculation was also used to investigate the mechanism. However, an in-depth understanding of the catalytic mechanism is still needed by further in situ characterizations. As a result, we will keep working on development and perfection on related mechanism exploration based on a series of in situ technologies.

Resource Availability

Lead Contact

Further information and requests for resources should be directed to and will be fulfilled by the Lead Contact, Shichun Mu (msc@whut.edu.cn).

Materials Availability

All unique/stable reagents generated in this study are available from the Lead Contact without restriction.

Data and Code Availability

All data used in the study are included in this publication. The present research did not use any new codes.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.

Published: December 18, 2020