Joule heating to grain-boundary-rich RuP₂ for efficient electrocatalytic hydrogen evolution in a wide pH range

INTRODUCTION

The increasing global energy demand, along with the consequential climate changes and environmental issues, has prompted both science and industry to explore sustainable and environmentally friendly alternative sources of energy as a substitute for depletable fossil fuels. Due to its high calorific value, superior energy density, and complete absence of carbon emissions, hydrogen is widely recognized as an optimal energy carrier^[1-3]. The production of green hydrogen via a water electrolyzer, facilitated by renewable energy sources such as solar, wind, and hydropower, represents a promising avenue toward achieving sustainable energy solutions. Platinum (Pt)-based compounds remain the most effective electrocatalysts for hydrogen evolution reaction (HER: 2H⁺ + 2e^- → H₂)^[4-6]. Nevertheless, the commercialization of Pt as an electrocatalyst for widespread H₂ production is hindered by its exorbitant cost and limited availability^[7]. Therefore, it is crucial to identify cost-effective, high-performance, and durable alternative catalysts to efficiently catalyze the hydrogen evolution reaction (HER).

In the past decades, transition-metal phosphides (TMPs) have attracted considerable attention due to their outstanding physicochemical properties, including good conductivity and high catalytic activities^[8-12]. However, developing Pt-group metal (PGM)-free TMPs that offer catalytic performance comparable to or exceeding that of Pt remains a significant challenge. Therefore, PGM-based TMPs are often considered the optimal choice. Among these, ruthenium (Ru) emerges as the most promising noble metal for this purpose. Recent studies have reported various Ru-based phosphides, such as RuP₂, RuP, and Ru₂P, exhibiting HER performance that is comparable to or even surpasses that of commercial Pt/C catalysts^[13,14]. For example, Mu’s group first demonstrated that N,P dual-doped carbon-encapsulated RuP₂ catalyst exhibits Pt-like activity towards HER under a wide pH range^[15]. Additionally, Ru₂P/C was synthesized by Ishikawa et al. using a temperature-programmed reduction method with H₂ as the reductant^[16]. Notably, most TMPs were synthesized through high temperature and/or high-pressure pyrolysis of the metal and phosphorus precursors over extended reaction times (≥ 2 h)^[17-19]. In general, these conventional methods for preparing TMPs are not only energy-intensive, but also result in larger grain sizes and lower surface areas due to prolonged pyrolysis, leading to a decreased electrochemical performance^[20-22]. Therefore, it is highly desirable to explore alternative synthetic approaches that enable the production of TMPs with smaller nanoparticle sizes and increased surface areas through lower pyrolysis temperatures or shorter pyrolysis durations.

Recently, due to its ultrafast heating process in milliseconds to ~3,000 °C, Joule heating has received widespread utilization in the synthesis of various nanomaterials with ultrafine nanostructures, such as materials-based single atoms, nanocrystals composed of transition metal carbide and intermetallic nanoparticles with a size distribution of approximately ~6 nm^[23-26]. The primary advantage of Joule heating lies in its ability to achieve sintering in seconds, thereby preventing the occurrence of sintering and agglomeration of active components in prolonged heat treatment. Undoubtedly, this ensures rapid and effective dispersion of active atoms, facilitating precise regulation of particle size, and structure in the synthesized material. On the other hand, several studies have demonstrated the superior catalytic efficiency of P-rich TMPs compared to their P-deficient counterparts in HER^[10,27,28]. Furthermore, when it comes to noble metal-based catalysts, doping with phosphorus elements has been proven effective in reducing the reliance on noble metals while enhancing their catalytic activity. However, the synthesis of P-rich TMPs is challenging due to prolonged pyrolysis resulting in the decomposition of the P-rich TMPs to P-deficient TMPs^[29].

Drawing inspiration from the aforementioned descriptions, this study demonstrates that Joule heating for just nine seconds enables the synthesis of grain-boundary-enriched, P-rich RuP₂ (RuP₂ JH). Remarkably, the obtained RuP₂ JH exhibits platinum-like HER activity with overpotentials of 22, 22 and 270 millivolts at a current density of 10 milliamperes per square centimeter in 0.5 M H₂SO₄, 1.0 M KOH and 0.1 M phosphoric acid buffer solution, respectively. These high HER activities not only surpass the majority of previously reported electrocatalysts based on precious-metal-based phosphides but also rival the activity of state-of-the-art commercial Pt/C materials. Therefore, this study presents a pioneering approach to fabricating TMPs in seconds by Joule heating, resulting in enhanced performance in HER and other energy conversion and storage devices.

EXPERIMENTAL

RESULTS AND DISCUSSION

The detailed synthesis process for the RuP₂ nanoparticles through the Joule heating method is illustrated in Figure 1A. Initially, a homogeneous mixture of ammonium chlororuthenite (IV) and red P was prepared. Next, the obtained precursor mixture underwent Joule heating under an argon flow. After cooling and rinsing with deionized water to eliminate the by-product chloride ions, etc., the final product of RuP₂ JH was obtained.

Figure 1. (A) Schematic diagram for the preparation of RuP₂ nanoparticles by JH method; (B-D) Low- and high-magnification TEM images of RuP₂ JH; (E) HRTEM, and (F-I) HAADF images and corresponding EDS elemental mapping of Ru and P of RuP₂ JH. TEM: Transmission electron microscopy; HRTEM: high-resolution TEM; HAADF: high-angle annular dark-field; EDS: energy dispersive spectroscopy.

The morphology of RuP₂ JH is initially characterized using high-angle annular dark-field scanning TEM (HAADF-STEM). The low-magnification TEM image in Figure 1B reveals the presence of numerous nanoparticles within RuP₂ JH. The high-magnification TEM image further reveals that the RuP₂ JH exhibits a particle diameter ranging from 15 nm to 60 nm, as shown in Figure 1C and Supplementary Figure 1. This suggests that the resulting material displays a relatively good uniformity. However, further control experiments should be conducted in future research to improve the homogeneity of the nanoparticles. In contrast, the bulk RuP₂ obtained through conventional pyrolysis at 800 °C for 2 h under an inert atmosphere shows a particle size exceeding 200 nm [Supplementary Figure 2]. Therefore, the small nanoparticle size of RuP₂ obtained can be attributed to the rapid JH method, which effectively prevents sintering and agglomeration typically induced by prolonged heat treatment. Consequently, it guarantees prompt and efficient dispersion of active atoms, while simultaneously controlling the micro-morphology, composition, and structure of synthetic materials. Crucially, the presence of grain boundaries is pervasive within the RuP₂ JH nanoparticles [Figure 1D], which can be attributed to the ultrafast heating and cooling process during sample processing. Additionally, the high-resolution TEM (HRTEM) image in Figure 1E reveals a well-defined lattice fringe measuring 0.386 nm, which closely corresponds to the observed diffraction pattern associated with the (110) crystal plane of orthorhombic phase RuP₂^[30]. The energy dispersive spectroscopy (EDS) analysis reveals the presence of both P and Ru signals in the RuP₂ JH samples [Supplementary Figure 3]. Moreover, the RuP₂ JH nanoparticles exhibit a uniform distribution of both P and Ru elements, as evidenced by HAADF-STEM and EDS elemental mapping images [Figure 1F-I].

The powder XRD was used to investigate the crystallographic structure of RuP₂ JH. As illustrated in Figure 2A, the obtained sample exhibited a well-matched powder XRD pattern with the orthorhombic RuP₂ phase as indicated by the standard PDF card (PDF No. 34-0333). The space group was determined as Pnnm(58) with cell parameters a₀ = 5.12 Å, b₀ = 5.89 Å, c₀ = 2.87 Å; Supplementary Figure 4 provides a visual representation^[31]. Additionally, XPS was employed to investigate the intricate elemental composition and surface chemical state of RuP₂ JH. The obtained results reveal that the surface of RuP₂ JH is predominantly composed of Ru, P, C, and O [Supplementary Figure 5]. The presence of elemental C and O can be attributed to the absorbed CO₂, H₂O, and O₂, or to slight surface oxidation of RuP₂ JH due to potential exposure to air. The XPS spectrum of the Ru 3d and C 1s regions [Figure 2B] reveals peaks at binding energies (BEs) of 284.5 eV and 280.4 eV, which correspond to the Ru 3d_3/2 and Ru 3d_5/2 states of RuP₂ JH, respectively. Concerning the Ru 3p spectrum, the peaks detected at BEs of 461.9 eV and 484.3 eV can be ascribed to the Ru 3p_3/2 and Ru 3p_1/2 states [Figure 2C], respectively. It is noteworthy that the peak of Ru 3p_3/2 is blue-shifted by 0.4 eV from the metal Ru (461.5 eV)^[32]. The blue shift primarily arises from the bonding structures involving Ru-P. Furthermore, Figure 2D illustrates the P 2p spectrum, which exhibits three distinct subpeaks at 133.7 eV, 130.3 eV and 129.6 eV, respectively. The peaks observed at BEs of 130.3 eV and 129.6 eV can be attributed to the presence of the Ru-P bond, while the peak detected at BE of 133.7 eV suggests the existence of oxidized phosphorus species (PO₄^3-). It is noteworthy that the P 2p_3/2 peaks in RuP₂ JH exhibit a negative shift compared to elemental phosphorus (130.2 eV)^[33]. These observations suggest the presence of a partial negative charge (δ^-) on P and a partial positive charge (δ⁺) on Ru in RuP₂ JH, indicating a weak electron transfer from Ru to P. The negatively charged P atom tends to attract a positively charged proton, similar to a base, thereby modulating the ability of active sites H atoms to attach and detach during the subsequent HER process. Therefore, the formation of the RuP₂ phase by Joule heating is unequivocally confirmed through comprehensive characterizations using TEM, XRD, and XPS techniques. Notably, a slight surface passivation effect is observed in RuP₂ JH.

Figure 2. (A) The XRD pattern for RuP₂ JH. XPS spectra in the (B) Ru 3d and C 1s; (C) Ru 3p and (D) P 2p regions for RuP₂ JH. XRD: X-ray powder diffraction; XPS: X-ray photoelectron spectroscopic.

The HER activity of the RuP₂ JH was examined in 0.5 M H₂SO₄ electrolytes, with all potentials calibrated using an RHE as the reference scale. Before conducting the HER tests, the RuP₂ JH material was activated by performing a series of linear sweep voltammetry (LSV) scans ranging from 0 V to -0.5 V vs. SCE. In order to facilitate comparison, we also assessed the HER catalytic activities of bulk RuP₂ (bulk RuP₂, obtained by a traditional method as reported in our previous work)^[30], and commercial Pt/C under identical conditions. Typically, Figure 3A illustrates the HER polarization curves over RuP₂ JH, bulk RuP₂ and commercial Pt/C. The polarization curves of all HER measurements were iR-corrected, with the internal resistance determined through EIS analysis [Supplementary Figure 6]. The RuP₂ JH shows remarkable HER catalytic activities, as anticipated, with an onset potential close to 0 mV. It is worth noting that Pt/C, RuP₂ JH and bulk RuP₂ electrodes were tested three times for each sample. Specifically, as illustrated in Figure 3B, in 0.5 M H₂SO₄ solutions, the overpotentials needed for RuP₂ JH to achieve cathodic current densities of 10 mA cm^-2,50 mA cm^-2, and 100 mA cm^-2 are 22±1.5 mV, 67±3 mV and 95±2.5 mV, respectively. Notably, such outstanding HER performance of RuP₂ JH is comparable to the state-of-the-art commercial Pt/C catalysts (20 mV, 48 mV, and 92 mV). However, the bulk RuP₂ synthesized by conventional pyrolysis techniques exhibits a relatively modest HER activity of 130 mV at 10 mA cm^-2. Furthermore, the intrinsic performance of RuP₂ JH is further demonstrated by normalizing the HER activity with respect to the mass of Ru [Supplementary Figure 7]. More importantly, the catalytic activity of RuP₂ JH surpasses that of recently reported Ru-based TMPs, including RuP₂NPC (38 mV@10 mA cm^-2, Tafel slope 38 mV dec^-1)^[15], RuP₂ double-walled nanotubes synthesized through anion/cation exchange reactions (46 mV@10 mA cm^-2)^[34], and pure RuP synthesized by 24 h thermal treatment (36 mV@10 mA cm^-2, Tafel slope 39.8 mV dec^-1)^[35]. Similarly, the Tafel slope obtained from the linear region of the Tafel plots for RuP₂ JH is 49 mV dec^-1 [Figure 3C], representing a notable improvement compared to that observed for bulk RuP₂(114 mV dec^-1). This observation suggests the superiority of Joule heating over conventional methods in preparing RuP₂. Furthermore, to the best of our knowledge, the remarkable performance of RuP₂ JH is favorably compared to that of most TMPs synthesized using traditional strategies and other PGM-based materials for HER reported in acidic electrolytes [Figure 3D and Supplementary Table 1].

Figure 3. (A and C) Polarization curves and corresponding Tafel plots of commercial Pt/C, RuP₂ JH and bulk RuP₂ in 0.5 M H₂SO₄ solution; (B) Overpotentials at j = 10, 50 and 100 mA cm^-2 for Pt/C, RuP₂ JH and bulk RuP₂, the error bars represent the standard deviations of three samples, with three measurements conducted for each sample; (D) Overpotentials comparison of RuP₂ JH at j = 10 mA cm^-2 with recently reported HER electrocatalysts; (E) Polarization curves were recorded before and after 1000, 2000, 3000 and 4000 CV potential cycles for RuP₂ JH; (F) Chronoamperometric test for RuP₂ JH in 0.5 M sulfuric acid electrolyte at constant potential for 45 h (without iR correction). HER: Hydrogen evolution reaction.

The long-term stability of RuP₂ JH was investigated using cyclic voltammetry (CV). Figure 3E presents the polarization curves obtained before and after 1000, 2000, 3000 and 4000 CVs at 5 mV s^-1. As depicted in Figure 3E, negligible degradation was observed in the LSV curves of RuP₂ JH even after undergoing 4000 CV potential cycles. Furthermore, the long-term durability of the RuP₂ JH electrocatalyst is investigated through continuous electrolysis at a fixed overpotential of approximately 50 mV. As shown in Figure 3F, the current density of RuP₂ JH exhibits only minimal degradation under acidic conditions during a 45 h electrolysis period, thereby indicating the excellent durability of the as-synthesized RuP₂ JH. Additionally, the XRD analysis of the fresh and post-HER RuP₂ JH materials [Supplementary Figure 8] confirms the similarity in their crystal structures, providing further evidence for the preservation of RuP₂ JH composition and its exceptional durability towards HER under acidic media. Therefore, these comprehensive analyses unequivocally establish RuP₂ JH as an excellent and enduring catalyst material for the HER.

The investigation into the FE of hydrogen evolution by the RuP₂ JH catalyst is further extended. Specifically, the water drainage strategy has been used to quantify the amount of the generated H₂. The presence of generated H₂ bubbles on the working electrode surface is easily discernible [Supplementary Figure 9]. In addition, Supplementary Video 1 captures the dynamic process of HER and subsequent accumulation of hydrogen gas. As illustrated in Supplementary Figure 10, upon collecting the evolved hydrogen at a specific time, the quantity closely corresponds to the theoretical value predicted by Faraday’s law, indicating an FE exceeding 96% for RuP₂ JH [Supplementary Figure 11]. The results unequivocally demonstrate that RuP₂ JH serves as an exceptional and enduring catalyst for HER.

Furthermore, the HER performance of RuP₂ JH was investigated under both alkaline (1.0 M KOH) and neutral (0.1 M PBS) electrolytes. Specifically, as depicted in Figure 4A-C, RuP₂ JH exhibits remarkable catalytic performance in alkaline electrolytes with a low onset potential of approximately 0 mV and a modest Tafel slope of around 57 mV dec^-1. Moreover, the RuP₂ JH exhibits overpotentials of 22 mV and 92 mV in alkaline media to achieve current densities of 10 mA cm^-2 and 50 mA cm^-2, respectively. Notably, the HER catalytic activity provided by RuP₂ JH surpasses that of commercial Pt/C even at a current density exceeding 25 mA cm^-2. Importantly, the exceptional HER activity of RuP₂ JH surpasses that of the majority of recently reported PGM-based TMPs in alkaline solutions [Figure 4D and Supplementary Table 2]. Similarly, as shown in Figure 4E-G, RuP₂ JH exhibits exceptional HER activity with an overpotential of approximately 65 mV at j = 2 mA cm^-2 and a low Tafel slope of 86 mV dec^-1 under 0.1 M PBS solution. Additionally, the chronopotentiometric analysis presented in Supplementary Figure 12 demonstrates that the RuP₂ JH catalyst exhibits good stability in HER for over 20 h under alkaline and neutral solutions at a current density of 10 mA cm^-2. Importantly, the superior HER activity of RuP₂ JH surpasses that of the majority of recently reported materials in neutral solutions [Supplementary Figure 13 and Supplementary Table 3]. These findings underscore the outstanding HER performance of RuP₂ JH over a wide pH range.

Figure 4. (A and E) LSV curves and (C and G) corresponding Tafel slopes for commercial Pt/C and RuP₂ JH in (A and C) alkaline and (E and G) neutral solutions; (B and F) Overpotentials at j = 2, 10 and 50 mA cm^-2 for Pt/C and RuP₂ JH in (B) alkaline and (F) neutral solutions; (D) Overpotentials comparison of RuP₂ JH at j = 10 mA cm^-2 with recently fabricated HER electrocatalysts in alkaline media. HER: Hydrogen evolution reaction; LSV: linear sweep voltammetry.

Based on the aforementioned analytical findings, the high HER performance of RuP₂ JH can be explained as follows. On the one hand, the ultrafast synthesis process of RuP₂ results in the formation of relatively small nanoparticles, which in turn enhances the electrochemically active surface area. On the other hand, the improved catalytic performance of RuP₂ JH may originate from the grain boundaries, which facilitate the formation of a high density of defects. These defects create a unique coordination environment around the atoms near the grain boundaries, leading to lattice strain effects that effectively modulate the electronic structure of the active site^[36,37]. More importantly, the presence of grain boundaries not only contributes to the high catalytic activity for HER, but also improves the stability of the catalyst^[38].

CONCLUSIONS

In summary, a rapid Joule heating process has been developed for the ultrafast synthesis of RuP₂ nanomaterials within nine seconds. Extensive physical characterizations have revealed that the RuP₂ JH nanomaterials possess abundant grain-boundary-rich structures, with an average nanoparticle size of approximately 40 nm. Experimental data suggest that RuP₂ JH, rich in grain boundaries, exhibits low overpotentials of 22 mV, 22 mV and 270 mV at j = 10 mA cm^-2 under acid, alkaline, and neutral electrolytes, respectively. This exceptional HER activity is comparable to that of the most extensively reported TMPs based on PGMs, which have traditionally been fabricated using solid-state or solution-phase strategies. Therefore, the Joule heating synthesis approach for RuP₂ at the seconds level holds significant promise for advancing the development of a range of TMP-based electrocatalysts in energy-related applications.