Synthesis of noble metal nanoarrays via agglomeration and metallurgy for acidic water electrolysis

Tao, Jiawei; Gao, Ruiqin; Lin, Geyu; Chu, Chaoyang; Sun, Yan; Yu, Chunyang; Ma, Yanhang; Qiu, Huibin

doi:10.1038/s41467-025-60419-8

Download PDF

Article
Open access
Published: 29 May 2025

Synthesis of noble metal nanoarrays via agglomeration and metallurgy for acidic water electrolysis

Nature Communications volume 16, Article number: 4996 (2025) Cite this article

7841 Accesses
17 Citations
Metrics details

Subjects

Abstract

Noble metal electrocatalysts remain the mainstay for proton exchange membrane water electrolysis, majorly due to their exceptional activity and durability in acidic media. However, conventional powder and particle catalysts intensively suffer from aggregation, shedding and poor electron conductivity in practical applications. Here, we develop a micellar brush-guided method to agglomerate and smelt metal nanoparticles into erect nanoarrays with designable constitutions on various substrates. While the nanoarrays of stacked nanoparticles show poor stability in the acidic media, the smelting treatment substantially enhances the electron conductivity by more than four order of magnitude and reinforces the nanoarray architectures. This allows the tailored fabrication of self-supported acid-durable metallic and alloy nanoarray catalysts with outstanding hydrogen evolution activity, and metal oxide nanoarray with extraordinary oxygen evolution activity. The integration of metallic Ru-nanoarray and RuO_x-nanoarray in a proton exchange membrane electrolyzer further enables a long-term stable water electrolysis process for more than 500 h at 1 A cm⁻².

Torsion strained iridium oxide for efficient acidic water oxidation in proton exchange membrane electrolyzers

Article 25 October 2021

Construction of Zn-doped RuO₂ nanowires for efficient and stable water oxidation in acidic media

Article Open access 02 May 2023

Turing structuring with multiple nanotwins to engineer efficient and stable catalysts for hydrogen evolution reaction

Article Open access 04 September 2023

Introduction

Hydrogen (H₂) production from electrolysis of water using renewable electricity is widely recognized as a sustainable alternative to polluting fossil fuels^1,2,3. The proton exchange membrane (PEM) water electrolysis technology, featured with scalability, high current density and superior power regulation compared to the dominant alkaline water electrolysis technology^4,5,6,7, is well-suited to the volatility of renewable electricity from wind, light, water wave, etc. However, the PEM water electrolysis is performed in an acidic environment, which remarkably limits the choice of electrocatalysts^8,9,10. The noble metal electrocatalysts remain the major choice for the anode (e.g. IrO₂ and RuO₂ powder) and the cathode (e.g. Pt/C powder) catalysts in acidic water electrolysis mainly due to their remarkable durability^11,12,13. Nevertheless, these powder and particle catalysts are normally bonded onto the membrane or diffusion electrodes (e.g. carbon cloth/paper, Ti fiber felt) for PEM devices through adhesives, which inevitably leads to problems such as active site burial, catalyst shedding and poor electron conductivity, especially at a high current density¹⁴.

Metal nanoarrays with appealing features such as free-standing architecture^15,16,17, open space, fast charge transfer, and low interfacial barriers^18,19,20,21 are highly useful for catalysis²², sensors²³ and optoelectronics²⁴. For transition metals (e.g. Fe, Co, Ni, Cu, Ti), they can normally be grown into corresponding hydroxide nanoarrays on electrodes by hydro/solvothermal processes²⁵. The nanoarray electrodes are more conducive to active site exposure, mass transfer and electrolyte wetting, which significantly improve the activity and stability of the electrocatalysts compared to those in powder form^14,18,26,27. However, limited by the intrinsic chemical properties and engineering difficulties, the growth of self-supported noble metal nanoarrays on electrodes has rarely been reported. Methods of casting or doping noble metals onto supported nanoarrays (e.g. TiO₂ nanoarrays) have been proposed to allow noble metals to take advantage of nanoarray architectures^28,29. Nevertheless, these frequently suffer from low efficiency, low noble metal density, as well as the limited choice and moderate electron conductivity of supported nanoarrays. It remains a critical challenge to precisely and rapidly agglomerate metal species on desired substrates to form self-supported vertically aligned architectures with sufficient catalytic interfaces.

Here, we develop a micellar brush-guided method to regulate the agglomeration of metal NPs to form erect nanoarrays on desired substrates. The highly solvated corona of the soft template also favors a simultaneous penetration of two or more metal NPs, leading to the formation of hybrid nanoarrays. These densely stacked metal NPs are further smelted at high temperatures to produce continuous metallic, alloy and metal oxide nanoarrays, exhibiting enhanced the electron conductivity by more than four order of magnitude and reinforced structures. When applied to the cathode and anode of a PEM water electrolysis device, these self-supported noble metal nanoarray electrocatalysts considerably boost the hydrogen and oxygen evolution reactions, and show sustainably durability at an industrial-level current density for more than 500 h at 1 A cm⁻².

Results and discussion

Agglomeration of metal NPs into nanowires

It has been previously found that a diverse array of metal NPs tend to associate with the poly(2-vinylpydidine) (P2VP) corona of the micellar brushes formed by PFS-b-P2VP [PFS = poly(ferrocenyldimethylsilane)] through coordination interactions^30,31. However, the metal NPs were deposited on the micellar brushes majorly in a rather sparse manner, which was far from the self-supported nanoarrays. In consideration of the factor that the average distance between the P2VP corona chains in the cylindrical micelles of PFS-b-P2VP is ca. 3.5 nm³², ideally, metal NPs with matched size would be more preferrable to fill the coronal gaps, aiming at a higher loading density. To this end, we prepared a series of metal (Ru, Ir, Rh, Pt) NPs with precisely controlled diameters ranging from 1.2 to 3.4 nm, which were stabilized by adsorbed anions (acetate) in ethylene glycol (Supplementary Fig. 1).

Ru NPs with an average diameter of 3.3 nm were first employed to interact with the cylindrical micelles formed by PFS₂₉-b-P2VP₂₃₀ (the subscripts refer to the degree of polymerization of each block, Supplementary Fig. 2) in ethylene glycol. After being mixed for 10 min, the Ru NPs were predominantly concentrated in the periphery of the P2VP corona (Fig. 1A). Subsequently, the Ru NPs gradually penetrated into the inner corona and eventually occupied the entire corona layer. The average spacing between the adjacent Ru NPs decreased from 5.1 ± 1.9 to 0.5 ± 0.08 nm during the agglomeration process (Supplementary Fig. 3). This led to the formation of regular nanowires with smooth contours and tightly stacked Ru NPs (Fig. 1B). X-ray photoelectron spectroscopy (XPS) showed that the Ru NPs were bonded to the pyridine groups in the P2VP corona mainly through coordination interactions, and the Ru-N association increased with the agglomeration period (Fig. 1C). It seemed that the coordination between the metal NPs and the pyridine groups in the P2VP corona was relatively reversible and dynamic. During the agglomeration process, the adjacent P2VP chains with a gap of 3.5 nm were constrained by the pre-loaded Ru NPs, and other Ru NPs were temporarily prevented from penetrating into the inner corona layer (Fig. 1D left). However, the reversible and dynamic coordination promoted the sequential migration of the Ru NPs (Fig. 1D middle), which ultimately led to a dense and linear stacking of metal NPs in the gaps of P2VP chains (Fig. 1D right). Thus, in some regions, the Ru NPs were found to stack in a chain-like fashion along the corona chains (Supplementary Fig. 4). Notably, the agglomeration of Ru NPs was highly dependent on the concentration of the acetate (c_acetate, Supplementary Fig. 5). Ru NPs with smaller diameters of 2.1 and 1.2 nm were also employed to interact with the cylindrical micelles, which revealed less tight stacking and relatively random distribution in the resulting nanowires (Supplementary Fig. 6). The cylindrical micelles also successfully templated the dense agglomeration of Ir, Rh and Pt, NPs with average diameters of ca. 3.0 nm to form regularly shaped nanowires (Supplementary Fig. 7).

**Fig. 1: Agglomeration of Ru NPs (3.3 nm) into nanowires.**

To gain more insights into the agglomeration of metal NPs, we successively and simultaneously added Ru and Pt NPs with comparable sizes into the solution of PFS₂₉-b-P2VP₂₃₀ cylindrical micelles. In terms of the relatively larger Ru (3.3 nm) and Pt NPs (3.4 nm), the first added Ru NPs that originally located in the outer layers of the P2VP corona migrated to the inner space after the sequential addition of Pt NPs (Supplementary Fig. 8A). The pre-loaded Ru NPs prevented the further added Pt NPs from penetrating into the inner corona layer, and thus the latter eventually majorly distributed in the periphery of the resulting nanowire. When the Ru and Pt NPs were simultaneously added, they both uniformly deposited in the corona of the micelles, forming hybrid nanowires (Supplementary Fig. 8B). Regarding the smaller Ru (1.2 nm) and Pt NPs (1.6 nm), the sequential and simultaneous addition of the two metal NPs solely led to the formation of hybrid nanowires with randomly distributed Ru and Pt NPs (Supplementary Fig. 9). Apparently, the pre-deposited small Ru NPs were unable to hinder the diffusion and immobilization of the later added Pt NPs. Preliminary molecular dynamics calculation showed that the association between the metal NPs and the P2VP corona is thermodynamically favored, and the metal NPs with comparable size with the gap of corona chains reveal high affinity with the P2VP corona (Supplementary Fig. 10).

Agglomeration of metal NPs into erect nanoarrays

Micellar brushes were subsequently prepared on desired substrates (e.g. carbon cloth) via living crystallization-driven self-assembly initiated by the pre-immobilized micelle seeds (Supplementary Fig. 11) to direct the agglomeration of metal NPs (Fig. 2A). Upon placing the micellar brush-decorated carbon cloth, the solution of Ru NPs (3.3 nm, c_acetate = 0.015 mol L⁻¹) quickly turned from dark brown to light brownish, indicating an efficient agglomeration process (Fig. 2B). The loading of Ru NPs on the carbon cloth increased with the immersion cycles and eventually reached ca. 0.22 mg cm⁻². Scanning electron microscopy (SEM) images showed that all of the exposed surfaces were completely covered by densely and vertically aligned nanoarrays (Fig. 2C, D, Supplementary Fig. 12), which were solely comprised by tightly stacked Ru NPs (Supplementary Fig. 13). The height of such Ru NPs-nanoarrays can be precisely adjusted in a range of ca. 0.5~1.6 μm by the amount of unimers added during the surface-initiated living growth of the micellar brush templates (Supplementary Fig. 14). Nanoarrays of binary metal NPs (RuPt) or quaternary metal NPs (RuIrRhPt) were also fabricated following a similar agglomeration process (Supplementary Fig. 15).

**Fig. 2: Agglomeration of Ru NPs into erect nanoarrays on desired substrates.**

The electrical conductivity of the metal NPs-nanoarrays on conductive titanium sheets was then evaluated by conductive atomic force microscopy, where the conductive probe was placed on the top of an individual nanowire (Fig. 2E), allowing the detection of electron conductivity along the longitudinal direction. While the PFS₂₉-b-P2VP₂₃₀ micellar brushes on the titanium sheet were absolutely insulative, the nanoarrays with higher loading of Ru NPs that obtained by decreased concentration of acetate showed ascending ohmic responses (Fig. 2F). The corresponding nonlinear current-voltage (I–V) curves illustrated a featured electron tunneling between the nanoparticle islands³³. The closer stacking of Ru NPs allowed a shorter and less hindered electron conduction path, which substantially boosted the electron conductivity. The optimal electron conductivity of Ru NPs-nanoarray was calculated to be 2.9 S m⁻¹ at a high bias voltage region (2.5 ≤ V ≤ 4.0), which was in line with conventional semiconducting materials.

Fabrication of continuous metallic and alloy nanoarrays by smelting

Insulating barriers including organic templates and voids are inevitable in assemblies of metal NPs, and critically limit the electron conductivity. Thus, we aimed to eliminate the insulative templates and the interparticle voids by smelting the metal NPs-nanoarray at high temperatures (Fig. 3A). This was favored by the lower melting temperature of metal NPs than that of the bulk counterpart³⁴. To this end, we smelted the Ru NPs-nanoarray in an inert atmosphere (N₂) at high temperatures (T). While a high temperature (800 °C) destroyed the nanoarray architectures (Supplementary Fig. 16), the Ru NPs-nanoarrays treated at relatively lower temperatures (400~600 °C) integrally maintained the nanoarray architecture (Fig. 3B, Supplementary Fig. 17). The interparticle voids were hardly observed upon increasing the smelting temperature to 600 °C and the stacked Ru NPs were entirely converted into fused continuous phases without losing the nanowire shapes (Fig. 3C). A similar evolution process was also observed under in situ TEM (Supplementary Fig. 18). The corresponding selected area electron diffraction images transformed from a diffuse ring to a clear dot matrix pattern, suggesting the transition from nanoparticle agglomerates to a highly crystalline structure.

**Fig. 3: Metallurgy of Ru NPs-nanoarrays into continuous metallic nanoarrays.**

High-resolution TEM (HRTEM) images further revealed that the stacked Ru NPs fused into interconnected polycrystalline grains (Supplementary Fig. 19), and eventually formed a continuous crystalline morphology (Fig. 3D). Atomic-scale HRTEM image further revealed the presence of abundant atomic vacancies that linearly distributed along the grain boundaries (GBs) in Ru-nanoarray-600 (Supplementary Fig. 20A). The lattice space along the GB was measured to be 0.46 nm, nearly twice the value of other regions (0.24 nm), confirming the loss of atomic layer (Supplementary Fig. 20B). X-ray diffraction (XRD) analysis of metallic Ru-nanoarray-600 presented sharp diffraction peaks, corresponding to a typical hexagonally close-packed (hcp) structure for Ru crystals (Fig. 3E). Extended X-ray absorption fine structure (EXAFS) spectra of Ru-nanoarray-600 showed that the bond length of the first shell Ru-Ru was red-shifted by 0.02 Å compared to that of Ru foil, probably due to the vacancies (Supplementary Fig. 20C). Such lengthened Ru-Ru bonds suggested that the atomic vacancies in metallic Ru-nanoarrays would be able to tune the covalency and electronic structures.

The resulting continuous metallic nanoarrays revealed fairly linear I-V curves throughout the full bias voltage region (−0.4 ≤ V ≤ 0.4) along the longitudinal direction (Fig. 3F). A relatively low bias voltage of 0.4 V gave a current of microampere level, while the nanoarray of stacked Ru NPs failed to be activated (Fig. 2F). Apparently, the elimination of the insulative templates and the interparticle voids, and the subsequent creation of continuous metallic conductive paths substantially enhanced the electron conductivity. Remarkably, the electron conductivity of the metallic Ru-nanoarray-600 reached 3.1 × 10⁴ S m⁻¹ in the longitudinal direction, which was four orders of magnitude higher than that of the Ru NPs-nanoarray. The generation of a continuous monocrystalline phase was undoubtably beneficial to the electron conduction.

Other metal (Ir, Rh, Pt) NPs-nanoarrays were also facilely converted into continuous metallic nanoarrays by smelting (Supplementary Figs. 21–23). Moreover, binary alloy RuPt-nanoarrays (Fig. 4A–D, Supplementary Fig. 24) and quaternary alloy RuIrRhPt-nanoarrays (Fig. 4E, F, Supplementary Fig. 25) were fabricated as well by smelting the corresponding hybrid metal NPs-nanoarray at 600 °C. STEM-EDS mapping images demonstrated a uniform distribution of the constituent elements. High-resolution HAADF-STEM images and XRD analyses further revealed that the alloy phases were highly crystalline and adopted a face-centered cubic (fcc) structure. Obviously, the smelting allowed the coalescence of various metal NPs and the further co-crystallization into an alloy phase. Noteworthy, when smelting at 400 °C, the dual metal (Ru, Pt) NPs-nanoarray was converted to a heterogeneous hybrid, where Pt nanocrystals were randomly embedded on the metallic Ru-nanoarray (Supplementary Fig. 26).

**Fig. 4: Metallurgy of hybrid metal NPs-nanoarrays into alloy nanoarrays.**

Fabrication of continuous metal oxide nanoarrays by smelting

To fabricate the oxide nanoarrays, initially, the Ru NPs nanoarray was directly smelted in air at high temperatures. Despite holding the nanoarray architectures, the Ru NPs were oxidized to stacked RuO_x NPs (labeled as RuO_x NPs-nanoarray) with the presence of a large number of dislodged species (Supplementary Fig. 27). It seemed that the oxidation of the Ru NPs prevented the further fusion³⁵. Subsequently, we smelted the Ru NPs-nanoarray in a nitrogen/oxygen mixture (N₂/O₂ = 99/1, v/v) at high temperatures. The decelerated oxidization of the stacked Ru NPs eventually enabled the formation of erect nanoarray of tightly fused RuO_x crystal grains with a characteristic lattice space of 0.32 nm, corresponding to the (110) plane of RuO₂ (Fig. 4B, Supplementary Fig. 28). The low oxygen atmosphere not only favored the simultaneous fusion and oxidization of Ru NPs, but also provides a feasible access to oxygen vacancies (Supplementary Fig. 29).

Electrocatalytic performance of smelted nanoarrays

In consideration of the highly open space and the prominent charge transporting features, as well as the direct decoration on current collectors, we further evaluated the potential utilization of the smelted nanoarrays in electrocatalysis. As a proof of concept, we chose the metallic and alloy nanoarrays as the catalysts for hydrogen evolution reaction (HER) and the metal oxide nanoarrays for oxygen evolution reaction (OER).

In HER, compared to the carbon cloth decorated with commercial 20 wt% Pt/C or Ru NPs-aggregation (Supplementary Fig. 30), the nanoarray catalysts could achieve a high current density of 1 A cm⁻² more readily (Fig. 5A, B). The metallic Ru-nanoarray-600 revealed a considerably low overpotential of 82.1 mV at 1 A cm⁻² and fast reaction kinetics of 18.6 mV dec⁻¹, much lower than that of Ru NPs-nanoarray (302 mV, 85.8 mV dec⁻¹). The alloy RuPt-nanoarray showed a lower overpotential of 73.1 mV at 1 A cm⁻² than metallic Ru-nanoarray-600, probably due to the more diverse electronic structures and synergistic active centers. The electrochemical double-layer capacitance (C_dl) and the calculated electrochemically active surface area (ECSA) of the nanoarray catalysts were significantly higher than those of the Ru NPs (Supplementary Fig. 31). In particular, the metallic Ru-nanoarray-600 possessed the highest C_dl and ECSA, more than five folds than that of Ru NPs-aggregation (Supplementary Table 1). The Ru-nanoarray-600 and RuPt-nanoarray demonstrate notable mass activity of 2088 A g_Ru⁻¹ and 4267 A g_RuPt⁻¹ respectively at −70 mV (vs. RHE), much larger than that of commercial 20 wt% Pt/C (530 A g_Pt⁻¹) (Supplementary Fig. 32). Considering that potentiostatic electrochemical impedance spectroscopy (EIS) may be not scientifically accurate as the charge transfer is inversely proportional to the current density, we conducted EIS at a fixed current density of 1 A cm⁻². The charge transfer resistance of RuPt-nanoarray and metallic Ru-nanoarray-600 were 1.5 and 2.9 ohm cm², respectively, remarkably smaller than that of the Ru NPs-nanoarray (12.3 ohm cm²) (Supplementary Fig. 33).

**Fig. 5: Customized nanoarrays as electrochemical catalysts.**

In OER, the overpotential for RuO_x-nanoarrays-400 was 330 mV at 1 A cm⁻² (Fig. 5C, D), much lower than RuO_x-nanoarrays-500 (472 mV) and RuO_x-nanoarrays-600 (532 mV). The RuO_x-nanoarray-400 demonstrates a notable mass activity of 4295 A g_Ru⁻¹ at 1.55 V (vs. RHE), representing 123 folds enhancement over the commercial RuO₂ catalyst (35 A g_Ru⁻¹) (Supplementary Fig. 34). The enhanced activity may arise from the better intrinsic activity of the relatively amorphous structures, which probably possessed more abundant defects and oxygen vacancies³⁶ (Supplementary Fig. 29). The RuO_x NPs-nanoarrays showed dramatic fluctuations when the current density reached 200 mA cm⁻², possibly due to the shedding of RuO_x NPs. Compared to the previously reported catalysts in acidic media, the HER and OER performance of the above nanoarray catalysts was undoubtedly comparable at a high current density level (Supplementary Fig. 35, Supplementary Tables 2 and 3).

In stability tests, both the metallic Ru-nanoarray-600 and RuO_x-nanoarrays-400 presented negligible increase in overpotential after running the reaction at 1 A cm⁻² for 200 h (Fig. 5E red line, Supplementary Fig. 36), and the nanoarray architecture and crystalline structure were integrally retained (Fig. 5F right, Supplementary Figs. 37 and 38). At the same current density, the Ru NPs-nanoarray displayed an apparent increase in overpotential after 50 h (Fig. 5E green line), and a large number of Ru NPs-nanoarrays were detached from the carbon cloth fibers under the constant impact of generated bubbles (Fig. 5F left, Supplementary Fig. 39). We also tracked the mass change of the Ru NPs-nanoarray during the test and found that the mass was decreasing as the test went on. After 50 hours of testing, the catalyst was reduced by 45%. (Supplementary Fig. 40). Thus, the extraordinary durability in acid medium of the metallic Ru-nanoarray-600 probably originated from the continuous crystalline structure and the fully metallic nanoarray. Contact angle analyses further demonstrated that the surface of metallic Ru-nanoarray-600 was hydrophilic toward the electrolyte but superaerophobic against the H₂ bubble (Supplementary Fig. 41), which further favored the release of the generated gas at a high current density³⁷.

Considering the higher price of Pt, we investigated the practical application potential of cheaper Ru as catalyst in water electrolysis. A proton exchange membrane (PEM) water electrolyzer was constructed by using RuO_x-nanoarrays-400 as the anode and the metallic Ru-nanoarray-600 as the cathode (Fig. 6A, B). I–V curves clearly showed that RuO_x-nanoarrays-400/Nafion 117/Ru-nanoarray-600 electrolyzer required only 1.88 V to reach a current density of 2 A cm⁻² at 80 °C (without iR-corrected), which was considerably better than the commercial IrO₂/Nafion 117/Pt/C electrolyzer and other reported systems (Fig. 6C, Supplementary Table. 4). The voltage of the RuO_x-nanoarrays-400/Nafion 117/Ru-nanoarray-600 electrolyzer showed negligible increase (degradation rate: 20 μV h⁻¹) and the nanoarray architectures remained undamaged after operating at 1 A cm⁻² for more than 500 h (Fig. 6D, Supplementary Fig. 42), indicating an outstanding durability against the high current density, acidic media and bubbles impact. In contrast, the commercial RuO₂/Nafion 117/Pt/C electrolyzer decay rapidly within 20 h. We monitored Ru dissolution from RuOx-nanoarrays-400/Nafion 117/Ru-nanoarray-600 electrolyzer by inductively coupled plasma mass spectrometry (ICP-MS) during 500 h stability test (Supplementary Fig. 43). The Ru concentration exhibited an initial surge during the early operational phase, reflecting dissolution of unstable Ru species during the activation period. Dissolution rates stabilized progressively during the first 200 h interval, ultimately maintaining a relatively constant level of approximately 55 ppb throughout the subsequent stability assessment phase.

**Fig. 6: PEM water electrolysis device performance.**

Theoretical calculations

Density functional theory (DFT) calculations were applied to analyze the adsorption energy of *H intermediates on the catalyst surface. The Sabatier principle states that a catalyst should bind an H atom neither too weakly nor too strongly to achieve a better HER performance³⁸. First of all, we constructed a model of perfect Ru (100) plane and calculated the Gibbs free energy (∆G_H) of six different adsorption sites, including the top sites (t₁, t₂), hollow sites (h₁, h₂) and bridge sites (b₁, b₂) (Supplementary Fig. 44A). The dominant adsorption sites on perfect Ru (100) are the bridge and hollow sites (b₁ & h₂), which reveal highest binding energy to *H, leading to a poor theoretical HER activity (Supplementary Fig. 44B). The NPs fusion-induced GBs in the metallic Ru-nanoarrays produced abundant atomic vacancies on the surface, whereby we constructed a defective Ru crystal by removing some atoms on the surface (Supplementary Fig. 44C). This created new catalytic sites in the vacancies (v₁, v₂). Compared to the perfect Ru, the ∆G_H of v₂ and b₁ is significantly reduced to −0.04 and −0.31 eV (Supplementary Fig. 44D), respectively, suggesting an enhanced intrinsic activity. Projected density of states (PDOS) demonstrated that the atomic vacancies in the metallic Ru-nanoarray would increase the density of states of the Ru sites near the Fermi level (Supplementary Fig. 44E). The calculated charge density of defective Ru showed distinct charge delocalization around the GBs, and the weakening of charge density at the vacancies reduces the *H adsorption strength (Supplementary Fig. 44F). It appears that more free electrons could transfer from the catalyst to the reactant molecules with lower energy barriers driven by the electric potential, facilitating the recombination of adsorbed *H into H₂.

In summary, we have revealed an intriguing spontaneous agglomeration process of a variety of metal (Ru, Ir, Rh, Pt) NPs into erect nanoarrays with precisely controlled height and components on various substrates. By in situ smelting, one or more metal NPs fused together and recrystallized into continuous metallic, alloy or oxide nanoarray, which prominently promoted the electrolysis of water in PEM cells at industrial-level current density especially in terms of stability, as a result of the highly open space, improved electron conductivity and fully metallic nanoarrays. In principle, this approach is applicable for a wider variety of nanoparticles such as transition metals, quantum dots, clusters and perovskites. The introduction of more types of metal NPs and the cooperative smelting also provides a feasible route to the construction of high-entropy alloys with predefined nanostructures. Future work could explore synergistic integration with membrane impregnation techniques to optimize interfacial contact and mitigate mass transport limitations. The advances in the fabrication of these distinctive nanoarray materials on desired substrates are anticipated to promote upgrade applications in catalysis, thermoelectricity, solar cells, superconductivity, flexible sensors, bioelectrodes, etc.

Methods

General information

Ethylene glycol (≥99.5%, AR), isopropanol (99.8%, HPLC), tetrahydrofuran (99.8%, free of stabilizer, HPLC), K₂PtCl₆ (99.95%, RG) were purchased from Adamas. Ethanol (≥99.5%, GR) and sodium acetate (≥99.0%, AR) were purchased from Greagent. RuCl₃ (97%) and IrCl₃ (99.9%) were purchased from Bidepharm. RhCl₃ (99%) was purchased from Meryer. PFS₂₉-b-P2VP₂₃₀ block copolymer was synthesized via living anionic ROP according to previous reports³⁹. The carbon clothes were purchased from Toray.

Preparation of PFS₂₉-b-P2VP₂₃₀ micelle seeds

In a typical process, 5 mg of PFS₂₉-b-P2VP₂₃₀ polymer powder was added to 10 mL of isopropanol and stirred at 80 °C until the mixture became transparent. The resulting solution was aged at room temperature overnight to allow the formation of long polydisperse cylindrical micelles. Micelles were subsequently fragmentized into shorter and uniform seeds by sonication using a probe ultrasonic processor in an ice-water bath for 80 min.

Preparation of long PFS₂₉-b-P2VP₂₃₀ cylindrical micelles

Typically, 50 μL of a solution of unimers (10 mg mL⁻¹ in tetrahydrofuran) was added to 1 mL of the solution of micelle seeds (0.05 mg mL⁻¹ in isopropanol) to trigger the living growth at the termini of the micelle seeds. The solution was shaken for 30 min and aged for 2 h, followed by solvent exchange to ethylene glycol by evaporation.

Living growth of PFS₂₉-b-P2VP₂₃₀ micellar brushes on desired substrates

A piece of carbon cloth (1 cm × 1 cm) was selected, and firstly sonicated alternatively in acetone and water to remove the surface-adsorbed components. Afterwards, it was treated by plasma to activate the surface and then soaked in a solution of the PFS₂₉-b-P2VP₂₃₀ micelle seeds (500 μL, 0.5 mg mL⁻¹ in isopropanol). After 30 min, the carbon cloth was taken out and blow dried under a gentle stream of nitrogen. The resulting carbon cloth was rinsed with isopropanol to remove the excess and poorly immobilized micelle seeds. Then, the micelle seed-coated carbon cloth was placed in 1 mL of isopropanol, followed by adding 4, 8, 16 and 32 μL of a solution of the PFS₂₉-b-P2VP₂₃₀ unimers (10 mg mL⁻¹ in tetrahydrofuran) and the mixture was vibrated for 30 min using a shaker. Subsequently, the resulting carbon cloth was rinsed with isopropanol several times to remove the excess unimers. The as-prepared micellar brush-decorated carbon cloth was immersed in ethylene glycol for further use.

When using flat titanium (Ti) sheets or quartz glass slides as the substrate, 20 μL of a solution of the micelle seeds (0.5 mg/mL in isopropanol) was cast on the substrate using a spin coater (3000 rpm, 1 min). The subsequent living growth process was the same as described above.

Fabrication of monodisperse metal (Ru, Ir, Rh, Pt) nanoparticles with controlled sizes

Metal nanoparticles (NPs) with a narrow size distribution were synthesized by chemical reduction of their corresponding metal species in ethylene glycol⁴⁰. For small metal NPs, a certain amount of the metal compound (3.2 × 10⁻³ mol L⁻¹) and varied amount of sodium acetate (0.01~0.12 mol L⁻¹) were dissolved in 50 mL of ethylene glycol at room temperature (60 °C for RhCl₃ and IrCl₃) under magnetic stirring. After bubbling with nitrogen for 30 min, the solutions were heated at 160 °C for 90 min. The resulting metal NPs were approximately 1.2 nm in diameter.

Metal NPs with larger sizes were further fabricated by seed-mediated growth. For instance, Ru NPs with diameter of 2.1 nm were synthesized using Ru NPs (1.2 nm) as the seeds. Typically, 20 mL of a solution of the as-prepared Ru NPs (1.2 nm) was mixed with 30 mL of ethylene glycol. Subsequently, RuCl₃ (final concentration 3.2 × 10⁻³ mol L⁻¹) and sodium acetate (final concentration 0.01~0.12 mol L⁻¹) were added at room temperature (60 °C for RhCl₃ and IrCl₃) under magnetic stirring. After bubbling with nitrogen for 30 min, the solutions were heated at 160 °C for 90 min. Similarly, Ru NPs with diameter of 3.3 nm were synthesized using Ru NPs (2.1 nm) as the seeds.

Preparation of nanowires of stacked metal NPs in solution

Nanowires of stacked metal NPs were prepared by mixing 2 mL of a solution of corresponding metal NPs (0.1 mg mL⁻¹ in ethylene glycol) and 1 mL of a solution of PFS₂₉-b-P2VP₂₃₀ cylindrical micelles (0.05 mg mL⁻¹ in ethylene glycol). The mixture was shaken for 10 s and allowed to age at room temperature for 2 h.

For the preparation of nanowires of multiple simultaneously stacked metal NPs, the corresponding metal NPs in ethylene glycol were firstly mixed in a 1:1 mass ratio to obtain a hybrid solution (total concentration 0.1 mg mL⁻¹). Subsequently, 2 mL of the hybrid solution was mixed with 1 mL of a solution of PFS₂₉-b-P2VP₂₃₀ cylindrical micelle (0.05 mg mL⁻¹ in ethylene glycol). The mixture was shaken for 10 s and allowed to age at room temperature for 2 h.

Preparation of nanowires of sequentially agglomerated Ru and Pt NPs in solution

Typically, 1 mL of a solution of Ru NPs (3.3 nm, 0.1 mg mL⁻¹ in ethylene glycol) was firstly mixed with 1 mL of a solution of PFS₂₉-b-P2VP₂₃₀ cylindrical micelles (0.05 mg mL⁻¹ in ethylene glycol). After 10 min, 1 mL of a solution of Pt NPs (3.4 nm, 0.1 mg mL⁻¹ in ethylene glycol) was added, and the mixture was shaken for 10 s and allowed to age at room temperature for 2 h.

Preparation of nanoarrays of stacked metal NPs (M NPs-nanoarrays)

Typically, a piece of micellar brush-decorated carbon cloth was immersed in 2 mL of a solution of metal NPs for 20 min, followed by rinsing with ethylene glycol to remove the excess metal NPs. This operation was repeated up to 5 times. Subsequently, the resulting carbon cloth was rinsed with ethanol to remove the remaining ethylene glycol. Supercritical carbon dioxide drying was applied to eliminate the capillary effect caused by the surface tension, leading to the formation of erect M NPs-nanoarrays in the dry state.

Preparation of continuous metallic, alloy and oxide nanoarrays

In a typical process, individual or hybrid metal NPs-nanoarrays were heated at a high temperature (T = 400, 500, 600 °C) with a speed of 10 °C min⁻¹ in specified gas atmosphere (N₂ for metallic and alloy nanoarrays, air or N₂/O₂ for oxide nanoarrays) and allowed to age for 4 h before cooling to room temperature.

Characterizations

Scanning electron microscopy (SEM)

SEM images were obtained on a RISE-MAGNA microscope operating at 5 keV under a high-resolution mode with a work distance of 5 mm. Energy dispersive spectrometric (EDS) mapping was performed using an Oxford X-Max Extreme operating at 15 keV under analysis mode with a work distance of 8 mm.

Transmission electron microscopy (TEM)

TEM imaging was performed using two configurations: (1) a Talos L120C G2 system equipped with a high-sensitivity CCD detector (120 kV), and (2) a Talos F200X G2 instrument with STEM capabilities (200 kV). For nanoparticle dispersions, samples were deposited onto carbon-coated copper grids via drop-casting followed by ambient drying. Bulk materials were subjected to brief ethanol ultrasonication (5–10 s) prior to grid deposition. Image processing utilized Velox 3.6 and FIJI/ImageJ platforms.

In situ heating experiments employed a JEM-F200 TEM (200 kV Schottky emitter) with a MEMS-based heating holder (Protochips Fusion Select). Specimens were dispersed in ethanol, loaded onto holey carbon MEMS chips (18 nm thick), and dried under infrared illumination⁴¹. Temperature ramping proceeded at 10°C min⁻¹ during time-resolved imaging.

Contact angle

Contact angles were measured by a Lauda Scientific LSA100 system equipped with a high-speed camera at room temperature. Two microliters of electrolyte or bubble was introduced via the sessile drop method and the angle was measured by the circle fitting approach. For each sample, at least three locations were tested in order to minimize the errors.

X-ray photoelectron spectroscopy (XPS)

XPS measurements were carried out using Thermo Scientific K-Alpha. The obtained data were analyzed by Avantage software.

X-ray diffraction (XRD)

XRD patterns were obtained on Rigaku Smartlab 9KW with a grazing incidence of 0.3°. Samples for XRD measurements were placed on quartz glass slides.

Inductively coupled plasma-mass spectrometry (ICP-MS)

ICP-MS data were collected using the NexION2000 Flexar20 HPLC system.

Electron conduction measurements

Electron conductivity analyses were performed at ambient conditions using a Bruker Dimension FastScan Bio atomic force microscopy with an electrical & magnetic module and a Multi75E-G probe, utilizing peak force tuna feedback control. Samples of metal nanoarrays were fabricated on a conductive titanium sheet before the test. The titanium sheet and the test platform were bonded together using conductive copper glue. The electron conductivity σ can be calculated as

$$\sigma=\frac{\nabla I}{\nabla V}\times \frac{H}{A}$$

(1)

where $\nabla I$ and $\nabla V$ were the current and bias voltage calculated from the I–V curves, H and A represented the height and cross-section area of the nanowire, respectively.

Electrocatalytic measurements

Electrocatalytic performance was assessed using a three-electrode electrochemical cell (CH Instrument 760E). The as-prepared samples of the Ru NPs-aggregate on carbon cloth, metal NPs-nanoarrays and metallic nanoarray on carbon cloth (1 cm × 1 cm, containing ca. 0.22 mg of catalyst) were directly used as the working electrodes. The 20 wt% Pt/C electrodes were fabricated by depositing 100 μL of catalysts ink (5.5 mg catalysts + 20 μL of Nafion solution in 2.5 mL of a mixture of 1:1 (v/v) water and isopropanol) onto carbon cloth, followed by drying at 60 °C. The test area for all samples is 1 cm², unless otherwise noted. A saturated calomel electrode (SCE) and a platinum flake were used as the reference and counter electrodes, respectively. An aqueous solution of H₂SO₄ (0.5 M, pH = 0.01 ± 0.01) was used as the electrolyte, which was configured by dilution with sulfuric acid and stored at room temperature in a sealed glass bottle. Before testing, the SCE was calibrated following established protocols⁴². Typically, Pt flake was selected as both the working electrode and counter electrode. The electrolyte (0.5 M H₂SO₄) was saturated with H₂ for at least 30 min. CV was carried out at a scan rate of 1 mV s⁻¹, and the average of the two interconversion point values was taken as the thermodynamic potential. The experimentally value of E⁰(SCE) is 0.243 V in 0.5 M H₂SO₄. The deviation was <5 mV compared to the theoretically value (0.241 V). All the mentioned potential was referenced to the reversible hydrogen electrode (RHE) according to the following equation: E(RHE) = E (vs. SCE) + E⁰(SCE) + 0.059 pH. Linear sweep voltammetry (LSV) measurements were conducted at 5 mV s⁻¹ and results were corrected by iR compensation (E_iR-corrected = E_original − I × 0.85Rs). Tests were performed three times to minimize accidents and errors. The stability was tested at a current density of 1 A cm⁻² with iR correction for 200 h by chronopotentiometry. Electrochemical impedance spectra (EIS) were acquired at 1 A cm⁻² in a frequency range of 100,000–0.01 Hz. The electrochemically active surface area (ECSA) was determined by: ECSA = C_dl/C_s, (C_dl: double-layer capacitance, C_s: specific capacitance)⁴³. In this study, a general specific capacitance of C_s = 0.035 mF cm⁻² was used based on the typical reported values⁴⁴. C_dl was determined by the equation C_dl = i_c/ν, (i_c: charging current, ν: scan rate). A plot of i_c as a function of ν yields a straight line with a slope equal to C_dl.

PEM water electrolyzer tests

A self-made cell was used as the PEM water electrolyzer device. The PEM cell consists of an iridium-coated titanium plate at the anode and a titanium plate at the cathode with serpentine flow channels of 1.6 cm² area on both sides. The Nafion 117 membranes were boiled sequentially for 1 h each in 3 wt.% H₂O₂, pure water, 1.0 M H₂SO₄ and pure water to remove possible contaminants and ensure the membrane was completely protonated. The RuO_x-nanoarray-400 on a titanium fiber felt, and the metallic Ru-nanoarray-600 on a carbon cloth were used as an OER anode and a HER cathode, respectively. Both of them were coated with 10 wt.% Nafion ionomers before being assembled into the membrane electrode assembly, followed by hot-pressing on a Nafion membranes at 135 °C at a mold clamping force of 2 MPa for 3 min⁴⁵. The electrolyzer underwent exposure to water at 80 °C for no less than 5 h. Prior to testing, the electrolyzer underwent sequential testing for galvanostatic (at 0.1 A cm⁻²) and potentiostatic (at 1.6 V) for 0.5 h each step to achieve a steady-state. The cell was then circulated with pure water at 100 mL min⁻¹. I–V curves were measured at 80 °C.

Molecular dynamics calculation

In order to explore the interaction between PFS₂₉-b-P2VP₂₃₀ cylindrical micelles and Ru NPs, two simplified all-atom models were constructed. First, for computational convenience, a nanotube structure was used to represent the PFS core, in consideration of the factor that the Ru NPs predominantly interact with the P2VP corona chains. On this basis, the P2VP chains were grafted onto the surface of the nanotubes with an equal spacing of 3.5 nm. At the same time, for the convenience of calculation, the repeating unit of P2VP is set as 20. In addition, three types of Ru NPs with variable diameters were constructed to investigate the effect of nanoparticle size. Universal force field was used to describe the cylindrical micelles and Ru NPs. At the beginning of the simulation, six NPs were placed in the gaps of the P2VP chains. Then, smart algorithm was used to minimize the energy of the system. At last, 1 ns constant volume-temperature run was performed at 298 K to obtain the equilibrium structure. Nose-Hoover thermostat was employed to control the temperature.

Theoretical methods for HER

All density functional theory (DFT) calculations in this study are performed using Vienna ab initio simulation package (VASP) code^46,47. The generalized gradient approximation (GGA) with the Perdew–Burke–Ernzerhof (PBE) functional is used to describe the exchange-correlation energy⁴⁸. The project augmented wave (PAW) method⁴⁹ is employed and the cutoff energy was set to 450 eV. The convergence limitation for force and energy are set as 0.05 eV Å⁻¹ and 10⁻⁴ eV, respectively. The Brillouin zones are sampled by Monkhorst–Pack scheme⁵⁰, which is carried out automatically by using VASPKIT package⁵¹. The k-points separation length is set as 0.04 2π Å⁻¹ for geometric optimization. A finer separation length, 0.02 2π Å⁻¹ is used in density of state (DOS) calculations. For all surface calculations, the long-range dispersion energy interaction is corrected using the Grimme’s DFT-D3 method^52,53. The dipolar correction is considered and the symmetrization is switched off.

For the hcp Ru (1000) model, a (4 × 4) supercell was constructed by cleaving the bulk structure along the (1000) direction, including 4 metallic atom layers. Then, two sets of Ru atoms located in the first layer are removed out from the Ru (1000) model to simulate the vacancies observed in the TEM image. A 15 Å vacuum layer was added to the slab models to prevent the periodic interactions. During the whole surface calculation process, the bottom two layers were kept frozen, while the upper two layers were fully relaxed.

The Gibbs free-energy of adsorbed H on the surface (ΔG_H*) was evaluated by using the following equation⁵⁴:

$$\Delta {G}_{{H}^{*}}={E}_{{H}^{*}}-{E}^{*}-1/2{E}_{{H}_{2}}+(\Delta {ZPE}-T\Delta S)$$

(2)

in which ${E}_{{H}^{*}}$, ${E}^{*}$ and ${E}_{{H}_{2}}$ were the DFT energies for the adsorbed surface, the initial surface and a single H₂ molecule, respectively. The $\Delta {ZPE}$ was the zero-point energy change $(\Delta {ZPE}={{ZPE}}_{{H}^{*}}-1/2{{ZPE}}_{{H}_{2}})$. The $T\Delta S$ was entropy change ($T\Delta S=T{S}_{{H}^{*}}-1/2{{TS}}_{{H}_{2}}$). Considering the negligible vibrational entropy for the adsorbed H* ($T{S}_{{H}^{*}}=0$), thus, at the condition of 1 atm and 300 K, $T\Delta S=-1/2{{TS}}_{{H}_{2}}=-0.205{eV}$.

Computation of d-band center

The d-band center (${\varepsilon }_{d}$) was calculated from the following equation⁵⁵

$${\varepsilon }_{d}=\frac{{\int }_{-\infty }^{\infty }{n}_{d}(\varepsilon )\varepsilon d\varepsilon }{{\int }_{-\infty }^{\infty }{n}_{d}(\varepsilon )d\varepsilon }$$

(3)

where $\varepsilon$ was energy referring to E_{Fermi level}, ${n}_{d}(\varepsilon )$ was the density of states data plotted versus energy.

Data availability

All data supporting the findings of this study are available within the article and its Supplementary Information/Source Date file, or from the authors upon reasonable request. The atomic structures for DFT calculations can be found at https://doi.org/10.6084/m9.figshare.29117933.v1. Source data are provided with this paper.

References

Turner, J. A. Sustainable hydrogen production. Science 305, 972–974 (2004).
Article ADS CAS PubMed Google Scholar
Chu, S. & Majumdar, A. Opportunities and challenges for a sustainable energy future. Nature 488, 294–303 (2012).
Article ADS CAS PubMed Google Scholar
Seh, Z. W. et al. Combining theory and experiment in electrocatalysis: insights into materials design. Science 355, eaad4998 (2017).
Article PubMed Google Scholar
Klose, C. et al. All‐hydrocarbon MEA for PEM water electrolysis combining low hydrogen crossover and high efficiency. Adv. Energy Mater. 10, 1903995 (2020).
Article CAS Google Scholar
Wang, Y. et al. PEM Fuel cell and electrolysis cell technologies and hydrogen infrastructure development–a review. Energy Environ. Sci. 15, 2288–2328 (2022).
Article CAS Google Scholar
King, L. A. et al. A non-precious metal hydrogen catalyst in a commercial polymer electrolyte membrane electrolyser. Nat. Nanotechnol. 14, 1071–1074 (2019).
Article ADS CAS PubMed Google Scholar
Khan, M. A. et al. Seawater electrolysis for hydrogen production: a solution looking for a problem. Energy Environ. Sci. 14, 4831–4839 (2021).
Article CAS Google Scholar
Pham, C. V. et al. Essentials of high performance water electrolyzers–from catalyst layer materials to electrode engineering. Adv. Energy Mater. 11, 2101998 (2021).
Article CAS Google Scholar
Kong, S. et al. Acid-stable manganese oxides for proton exchange membrane water electrolysis. Nat. Catal. 7, 1–10 (2024).
Lagadec, M. F. & Grimaud, A. Water electrolysers with closed and open electrochemical systems. Nat Mater. 19, 1140–1150 (2020).
Article ADS CAS PubMed Google Scholar
Hao, S. et al. Torsion strained iridium oxide for efficient acidic water oxidation in proton exchange membrane electrolyzers. Nat. Nanotechnol. 16, 1371–1377 (2021).
Article ADS CAS PubMed Google Scholar
Shi, Z. et al. Phase-dependent growth of Pt on MoS₂ for highly efficient H₂ evolution. Nature 621, 300–305 (2023).
Article ADS CAS PubMed Google Scholar
Wu, Z. Y. et al. Non-iridium-based electrocatalyst for durable acidic oxygen evolution reaction in proton exchange membrane water electrolysis. Nat. Mater. 22, 100–108 (2023).
Article ADS CAS PubMed Google Scholar
Luo, Y., Zhang, Z., Chhowalla, M. & Liu, B. Recent advances in design of electrocatalysts for high‐current‐density water splitting. Adv. Mater. 34, 2108133 (2022).
Article CAS Google Scholar
Hou, J. et al. Rational design of nanoarray architectures for electrocatalytic water splitting. Adv. Funct. Mater. 29, 1808367 (2019).
Article Google Scholar
Tang, C. et al. Energy‐saving electrolytic hydrogen generation: Ni₂P nanoarray as a high‐performance non‐noble‐metal Electrocatalyst. Angew. Chem. Int. Ed. 56, 842–846 (2017).
Article CAS Google Scholar
He, S. et al. The Secret of Nanoarrays toward efficient electrochemical water splitting: a vision of self‐dynamic electrolyte. Adv. Mater. 35, 2307017 (2023).
Article CAS Google Scholar
Zhu, S., Sheng, J., Ni, J. & Li, Y. 3D vertical arrays of nanomaterials for microscaled energy storage devices. Acc. Mater. Res. 2, 1215–1226 (2021).
Article CAS Google Scholar
Mor, G. K. et al. A review on highly ordered, vertically oriented TiO₂ nanotube arrays: Fabrication, material properties, and solar energy applications. Sol. Energy Mater. Sol. Cells 90, 2011–2075 (2006).
Article CAS Google Scholar
Elnathan, R., Kwiat, M., Patolsky, F. & Voelcker, N. H. Engineering vertically aligned semiconductor nanowire arrays for applications in the life sciences. Nano Today 9, 172–196 (2014).
Article CAS Google Scholar
Park, W. I. & Yi, G.-C. Electroluminescence in n-ZnO nanorod arrays vertically grown on p-GaN. Adv. Mater. 16, 87–90 (2004).
Article CAS Google Scholar
Ren, Z., Guo, Y. & Gao, P. X. Nano-array based monolithic catalysts: concept, rational materials design and tunable catalytic performance. Catal. Today 258, 441–453 (2015).
Article CAS Google Scholar
Mishra, R. K. et al. Continuous opioid monitoring along with nerve agents on a wearable microneedle sensor array. J. Am. Chem. Soc. 142, 5991–5995 (2020).
Article CAS PubMed Google Scholar
Zhu, Q. B. et al. A flexible ultrasensitive optoelectronic sensor array for neuromorphic vision systems. Nat. Commun. 12, 1798 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Li, H. et. al. Systematic design of superaerophobic nanotube-array electrode comprised of transition-metal sulfides for overall water splitting. Nat. Commun. 9, 2452 (2018).
Article ADS PubMed PubMed Central Google Scholar
Zhou, D. et. al. Layered double hydroxide-based electrocatalysts for the oxygen evolution reaction: identification and tailoring of active sites, and superaerophobic nanoarray electrode assembly. Chem. Soc. Rev. 50, 8790–8817 (2021).
Article CAS PubMed Google Scholar
Lv, X. et al. Atomic‐level surface engineering of nickel phosphide nanoarrays for efficient electrocatalytic water splitting at large current density. Adv. Funct. Mater. 33, 2205161 (2023).
Article CAS Google Scholar
Cheng, J. et al. Impact of Ir-valence control and surface nanostructure on oxygen evolution reaction over a highly efficient Ir–TiO₂ nanorod catalyst. ACS Catal. 9, 6974–6986 (2019).
Article CAS Google Scholar
Zhou, L. et al. Stabilizing non-iridium active sites by non-stoichiometric oxide for acidic water oxidation at high current density. Nat. Commun. 14, 7644 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Cai, J. D. et al. Tailored multifunctional micellar brushes via crystallization-driven growth from a surface. Science 366, 1095–1098 (2019).
Article ADS CAS PubMed Google Scholar
Lin, G. Y. et al. Capillary-bound dense micelle brush supports for continuous flow catalysis. Angew. Chem. Int. Ed. 60, 24637–24643 (2021).
Article CAS Google Scholar
Tian, J. et al. High-resolution cryo-electron microscopy structure of block copolymer nanofibres with a crystalline core. Nat. Mater. 22, 786–792 (2023).
Article ADS CAS PubMed Google Scholar
Wen, Y. et al. A coherent nanomechanical oscillator driven by single-electron tunnelling. Nat. Phys. 16, 75–82 (2020).
Article CAS PubMed Google Scholar
Warren, S. C. et al. Generalized route to metal nanoparticles with liquid behavior. J. Am. Chem. Soc. 128, 12074–12075 (2006).
Article CAS PubMed Google Scholar
Yu, J. et al. Ultra-high thermal stability of sputtering reconstructed Cu-based catalysts. Nat. Commun. 12, 7209 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Deka, N. et al. On the operando structure of ruthenium oxides during the oxygen evolution reaction in acidic media. ACS Catal. 13, 7488–7498 (2023).
Article CAS PubMed PubMed Central Google Scholar
Yu, X. X. et al. Superaerophobic” nickel phosphide nanoarray catalyst for efficient hydrogen evolution at ultrahigh current densities. J. Am. Chem. Soc. 141, 7537–7543 (2019).
Article CAS PubMed Google Scholar
Sun, F. et al. Theoretical advances in understanding and designing the active sites for hydrogen evolution reaction. ACS Catal. 12, 8404–8433 (2022).
Article CAS Google Scholar
Wang, H. et al. Synthesis and self-assembly of poly(ferrocenyldimethylsilane-b-2-vinylpyridine) diblock copolymers. Macromolecules 40, 3784–3789 (2007).
Article ADS CAS Google Scholar
Wang, Y. et al. Preparation of tractable platinum, rhodium, and ruthenium nanoclusters with small particle size in organic media. Chem. Mater. 12, 1622–1627 (2000).
Article Google Scholar
Ling, Y. et al. Atomic-level structural responsiveness to environmental conditions from 3D electron diffraction. Nat. Commun 13, 6625 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Niu, S. et al. How to reliably report the overpotential of an electrocatalyst. ACS Energy Lett. 5, 1083–1087 (2020).
Article CAS Google Scholar
Zhao, X. et al. Electrolyte-free electrochemical oxygen generator for providing sterile and medical-grade oxygen in household applications. Device 2, 100360 (2024).
Article Google Scholar
McCrory, C. C. L. et al. Benchmarking Heterogeneous Electrocatalysts for the Oxygen Evolution Reaction. J. Am. Chem. Soc. 135, 16977–16987 (2013).
Article CAS PubMed Google Scholar
Li, A. et al. Atomically dispersed hexavalent iridium oxide from MnO₂ reduction for oxygen evolution catalysis. Science 384, 666–670 (2024).
Article ADS CAS PubMed Google Scholar
Perdew, J. P. et al. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996).
Article ADS CAS PubMed Google Scholar
Kresse, G. et al. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6, 15–50 (1996).
Article CAS Google Scholar
Kresse, G. et al. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996).
Article ADS CAS Google Scholar
Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953–17979 (1994).
Article ADS Google Scholar
Monkhorst, H. J. & Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B 13, 5188–5192 (1976).
Article ADS MathSciNet Google Scholar
Wang, V. et al. VASPKIT: A user-friendly interface facilitating high-throughput computing and analysis using VASP code. Comput. Phys. Commun. 267, 108033 (2021).
Article CAS Google Scholar
Grimme, S. et al. A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 132, 154104 (2010).
Article ADS PubMed Google Scholar
Grimme, S., Ehrlich, S. & Goerigk, L. Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 32, 1456–1465 (2011).
Article CAS PubMed Google Scholar
Nørskov, J. K. et al. Trends in the exchange current for hydrogen evolution. J. Electrochem. Soc. 152, J23 (2005).
Article Google Scholar
Vojvodic, A., Nørskov, J. K. & Abild-Pedersen, F. Electronic structure effects in transition metal surface chemistry. Top Catal. 57, 25–32 (2014).
Article CAS Google Scholar

Download references

Acknowledgements

We thank the Shanghai Synchrotron Radiation Facility for SAXS and EXAFS characterization; the Analytical Instrumentation Center, ShanghaiTech University for in situ TEM; the Instrumental Analysis Center, Shanghai Jiao Tong University for other characterizations. We thank the financial support from the National Key R&D Program of China (2020YFA0908100, received by H.Q.), the National Natural Science Foundation of China (22425203, 22075180, 22401249, received by H.Q.), the Innovation Program of Shanghai Municipal Education Commission (202101070002E00084, received by H.Q.), the Fundamental Research Funds for the Central Universities (YG2023ZD29, YG2023ZD07, received by H.Q.).

Author information

Authors and Affiliations

State Key Laboratory of Synergistic Chem-Bio Synthesis, School of Chemistry and Chemical Engineering, Frontiers Science Center for Transformative Molecules, Zhangjiang Institute for Advanced Study, State Key Laboratory of Micro-Nano Engineering Science, Shanghai Jiao Tong University, Shanghai, China
Jiawei Tao, Geyu Lin, Yan Sun, Chunyang Yu & Huibin Qiu
School of Biological and Chemical Engineering, NingboTech University, Ningbo, China
Ruiqin Gao
School of Physical Science and Technology & Shanghai Key Laboratory of High-resolution Electron Microscopy, ShanghaiTech University, Shanghai, China
Chaoyang Chu & Yanhang Ma

Authors

Jiawei Tao
View author publications
Search author on:PubMed Google Scholar
Ruiqin Gao
View author publications
Search author on:PubMed Google Scholar
Geyu Lin
View author publications
Search author on:PubMed Google Scholar
Chaoyang Chu
View author publications
Search author on:PubMed Google Scholar
Yan Sun
View author publications
Search author on:PubMed Google Scholar
Chunyang Yu
View author publications
Search author on:PubMed Google Scholar
Yanhang Ma
View author publications
Search author on:PubMed Google Scholar
Huibin Qiu
View author publications
Search author on:PubMed Google Scholar

Contributions

J.T. and H.Q. conceived the project; J.T. performed the synthetic and assembly experiments; J.T., R.G. and Y.S. performed the electrochemical tests; C.C. and Y.M. performed the in situ TEM tests; G.L. prepared the polymer samples; C.Y. performed the molecular dynamics simulations; J.T. and H.Q. analyzed the data and prepared the manuscript.

Corresponding author

Correspondence to Huibin Qiu.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks Bosi Peng, Hao Bin Wu and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Transparent Peer Review file (download PDF )

Source data

Source Data (download XLSX )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Tao, J., Gao, R., Lin, G. et al. Synthesis of noble metal nanoarrays via agglomeration and metallurgy for acidic water electrolysis. Nat Commun 16, 4996 (2025). https://doi.org/10.1038/s41467-025-60419-8

Download citation

Received: 17 December 2024
Accepted: 22 May 2025
Published: 29 May 2025
Version of record: 29 May 2025
DOI: https://doi.org/10.1038/s41467-025-60419-8