Ordering‐Driven Biaxial Strain Engineering in PtNi Intermetallic Nanowires for Oxygen Reduction Catalysis

ABSTRACT

Precise regulation of lattice strain in platinum (Pt)‐based intermetallic catalysts is essential for optimizing oxygen reduction reaction (ORR) performance, yet strain evolution during atomic ordering is often simplified as isotropic compression, masking its structural complexity. Herein, we systematically modulate the ordering degree of PtNi nanowires from 4.5% to 65.8% to dynamically track this structural evolution. A key finding that refines the conventional understanding of strain evolution is the emergence of characteristic biaxial strain during the disorder‐to‐order transition, which we quantitatively distinguish from the universally assumed isotropic compression model. Specifically, this transition induces in‐plane lattice expansion coupled with out‐of‐plane lattice contraction. As the ordering degree increases, this biaxial strain progressively relaxes the compressive stress on the catalytically active (111) surface, rather than amplifying it as traditional models predict. Density functional theory calculations confirm this anisotropic strain reduces the energy barrier of the rate‐determining *O protonation step from 0.51 to 0.39 eV. The optimized catalyst exhibits a mass activity of 1.30 A mg_Pt ⁻¹ with 86.0% retention after 60 000 cycles, and a peak power density of 1014.5 mW cm⁻² in membrane electrode assemblies. This work identifies tunable ordering‐induced biaxial strain as a critical structural parameter for designing high‐performance Pt‐based ORR catalysts.

Ordering‐driven biaxial strain in PtNi intermetallic nanowires is revealed via controlled disorder‐to‐order transition, featuring in‐plane expansion and out‐of‐plane contraction. This anisotropic strain progressively relieves compressive stress on the (111) surface, lowering the *O protonation barrier from 0.51 to 0.39 eV. The optimized catalyst achieves a mass activity of 1.30 A mg_Pt ⁻¹ with 86.0% retention after 60 000 cycles.

1. Introduction

Proton exchange membrane fuel cells (PEMFCs) are widely recognized as promising clean energy conversion technologies for electric vehicles and stationary power systems, owing to their high energy efficiency and zero‐emission operation [1, 2, 3, 4, 5]. However, the sluggish kinetics of the oxygen reduction reaction (ORR) at the cathode remains the major performance bottleneck, necessitating the development of high‐activity and durable electrocatalysts [6, 7, 8, 9, 10, 11, 12]. Currently, carbon‐supported platinum nanoparticles (Pt/C) are commercially deployed as ORR catalysts, but their high cost, limited activity, and insufficient durability fall short of the United States of America. Department of Energy (DOE) targets, hindering large‐scale commercialization [13, 14, 15, 16, 17].

Alloying Pt with 3d transition metals (M = Fe, Co, Ni, etc.) has emerged as an effective strategy to enhance ORR performance by modifying the geometric and electronic structure of Pt active sites [18, 19, 20, 21, 22]. Among Pt‐based alloys, ordered intermetallic compounds (IMCs), such as L1₀‐PtM or L1₂‐Pt₃M, have drawn particular attention because their chemically ordered structure provides enhanced thermodynamic stability and resistance to transition metal leaching compared with disordered alloys [23, 24, 25, 26]. The periodic arrangement of Pt and M atoms introduces long‐range atomic ordering, which not only adjusts the local coordination environment of Pt but also generates intrinsic lattice strain that shifts the Pt d‐band center and optimizes oxygen intermediate adsorption [27, 28, 29, 30]. Consequently, IMCs catalysts typically demonstrate improved ORR activity and remarkable durability, making structural ordering a promising strategy for catalyst optimization beyond conventional composition engineering [31, 32, 33, 34].

To date, the strain engineering paradigm for Pt‐based IMCs has been dominated by the isotropic uniform compressive strain model, which attributes lattice strain to atomic size mismatch and overall lattice contraction during alloying [35, 36, 37]. This view has guided the design of intermetallic catalysts for decades, yet it does not account for the symmetry‐breaking lattice deformation inherent to certain ordering transitions, particularly the face‐centered cubic (fcc) to L1₀ transformation [38, 39, 40, 41]. Importantly, the L1₀ ordered structure is tetragonal rather than cubic, meaning that the disorder‐to‐order transition intrinsically generates direction‐dependent lattice deformation, in‐plane expansion along the a/b axes coupled with out‐of‐plane contraction along the c‐axis, resulting in a biaxial strain state fundamentally distinct from isotropic compression [42, 43, 44, 45]. For example, a recent study on PtZn nanoparticles revealed that during the transformation from a disordered fcc phase to an ordered tetragonal phase, the lattice undergoes anisotropic deformation, generating in‐plane tensile and out‐of‐plane compressive strain components [43]. This discovery suggests that lattice ordering can introduce direction‐dependent strain states fundamentally distinct from conventional isotropic strain models. Further evidence of the catalytic significance of biaxial strain was provided by a study on PtPb/Pt core‐shell nanoplates [44]. Owing to the inherently anisotropic hexagonal crystal structure of the PtPb intermetallic core, substantial biaxial tensile strain was imposed on the Pt shell, which led to markedly enhanced ORR activity by optimizing Pt‐O binding energetics. These findings confirm that biaxial strain can act as an independent and powerful descriptor for catalyst design. Nevertheless, it remains unclear whether such biaxial strain can be intrinsically generated and continuously tuned within Pt‐based IMCs catalysts, since systematic studies revealing how atomic ordering produces this strain in the alloys and how the strain evolves during ordering are lacking, severely limiting the rational design of advanced intermetallic catalysts.

In this work, we depart from the conventional isotropic strain paradigm by demonstrating that the disorder‐to‐order transition in L1₀‐PtNi nanowires generates a distinctive biaxial strain state, in‐plane expansion coupled with out‐of‐plane contraction, which progressively alleviates compressive stress on the active surface. This finding not only refines the fundamental understanding of strain evolution in intermetallic catalysts but also establishes ordering‐induced biaxial strain as a previously unrecognized structural descriptor for optimizing ORR performance. Here, we address this gap by establishing a model ordering‐strain evolution system based on PtNi nanowires. By systematically varying heat‐treatment duration, we obtain a series of samples spanning low, intermediate, and high ordering degrees and directly track structural evolution with x‐ray diffraction (XRD) and atomic‐resolution imaging. Our results demonstrate that the disorder‐to‐order transition triggers a characteristic biaxial strain involving simultaneous in‐plane lattice expansion and out‐of‐plane contraction. This anisotropic deformation progressively alleviates the intrinsic compressive stress on the Pt(111) surface, which contrasts with the conventional understanding that ordering primarily induces isotropic compression. This strain evolution strengthens Pt‐Ni orbital hybridization, thereby significantly lowering the energy barrier for the rate‐determining step, specifically the protonation of adsorbed oxygen intermediates, from 0.51 to 0.39 eV. As a result, the highly ordered PtNi nanowires deliver a mass activity of 1.30 A mg_Pt ⁻¹, outstanding stability with 86.0% activity retention after 60 000 cycles, and a peak power density of 1014.5 mW cm⁻² in membrane electrode assembly tests. By identifying ordering‐induced biaxial strain as a key structural parameter, this work offers a comprehensive understanding of the interplay between atomic arrangement, lattice strain, and catalytic activity, providing reliable design principles for engineering high‐performance intermetallic nanostructures for energy conversion technologies.

2. Results and Discussion

In this study, we employed a surface hydrogen adsorption strategy to lower the surface energy of nanowires, enabling the successful synthesis of a series of PtNi nanowires (PtNi NWs) with uniform one‐dimensional morphology, but distinct degrees of atomic ordering [24]. The synthesis began with a colloidal solvothermal approach to prepare initial PtNi alloy nanowires (I‐PtNi NWs), in which cetyltrimethylammonium chloride acted as a structure‐directing agent and glucose served as a mild reducing agent. The as‐prepared I‐PtNi NWs exhibited a relatively uniform diameter (∼20 nm) and a rough surface texture, as confirmed by transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM) images (Figure S1). To enable subsequent thermal annealing, the I‐PtNi NWs were deposited onto Vulcan XC‐72R carbon to form a composite (as I‐PtNi NWs/C), which provided a thermally stable matrix for structural evolution. TEM images of I‐PtNi NWs/C confirmed that the nanowires were uniformly distributed on the carbon substrate without noticeable aggregation (Figure S2a–c). TEM energy‐dispersive x‐ray spectroscopy (EDS) revealed a near‐stoichiometric Pt:Ni atomic ratio of 48.8:51.2 (Figure S2d). Subsequently, the I‐PtNi NWs/C were annealed under a mixed H₂/Ar atmosphere at 550°C. Hydrogen adsorption plays a vital mechanistic role during this high‐temperature annealing process. Specifically, the strong orbital hybridization between the adsorbed H 1s and the transition metal d‐orbitals significantly elevates the migration energy barrier of surface atoms. This thermodynamic stabilization effectively lowers the overall surface energy and profoundly suppresses surface diffusion‐driven Ostwald ripening and coalescence, thereby firmly preserving the one‐dimensional nanowire structure [24]. By precisely controlling the annealing duration, we obtained a series of PtNi NWs with gradually increased ordering degrees, thus establishing an ideal platform for systematically investigating the structure‐ordering‐performance relationship in intermetallic nanowire catalysts.

We systematically investigated the morphological evolution of I‐PtNi NWs/C during annealing at 550°C for different durations. STEM images revealed that after annealing for 1, 5, and even up to 24 h, the nanowires largely preserved their original one‐dimensional morphology without noticeable aggregation, fragmentation, or structural collapse (Figure 1a‐c). This morphological stability is notable and highlights the effectiveness of the hydrogen‐assisted strategy in stabilizing the nanowire structure. To verify the crucial role of hydrogen in this process, we conducted control annealing experiments under an Ar atmosphere and vacuum at −0.1 MPa (gauge). While XRD confirmed that ordering transitions could also occur under these conditions, TEM analysis revealed that the one‐dimensional morphology was completely lost, with severe wire fragmentation and aggregation (Figures S3 and S4). This comparison clearly demonstrates that hydrogen plays a pivotal role in preserving the nanowire morphology during high‐temperature ordering.

FIGURE 1.

Structural evolution of I‐PtNi NWs during ordering transition under hydrogen‐assisted annealing. STEM images of I‐PtNi NWs/C held at 550°C for (a) 1 h, (b) 5 h, and (c) 24 h. (d) High‐magnification TEM images with diameter histograms and (e) corresponding STEM images of representative single nanowires annealed for different durations. (f) XRD patterns of I‐PtNi NWs/C under different holding times. (g) Evolution of ordering degree and nanowire diameter with annealing time.

To gain further insight into the time‐dependent structural and compositional evolution during annealing, we performed time‐resolved TEM analysis. Regardless of annealing duration, the average diameter of the nanowires remained constant at ∼20 ± 2 nm, and no obvious coarsening or shape degradation was observed (Figure 1d,e,g; Figures S5–S9). To qualitatively assess the compositional distribution within individual nanowires, we took advantage of the Z‐contrast in STEM imaging, where the image intensity scales approximately with the square of the atomic number. For the as‐synthesized I‐PtNi NWs/C, pronounced contrast variations were observed, indicating a significant compositional inhomogeneity with Pt‐rich regions located both at the surface and within the nanowire core (Figure S10). After annealing for 1 h, the contrast variations became noticeably weaker; the distinct Pt‐rich core disappeared, although some localized segregation regions persisted, suggesting that substantial elemental diffusion had occurred, but the distribution was not yet fully homogenized (Figure 1e). Upon annealing for 2 h, the contrast difference further decreased, and the elemental distribution became more uniform throughout the nanowire. After 5 h, the contrast appeared largely homogeneous across the entire nanowire, indicating a marked reduction in compositional segregation (Figure 1e). With extended annealing, this homogenization became even more pronounced (Figure 1e). To corroborate these observations, we conducted STEM‐EDS elemental mapping at different annealing stages (Figure S11). The elemental maps were in excellent agreement with the STEM contrast analysis: the I‐PtNi NWs/C initially showed clear Pt enrichment at the surface, which disappeared after 1 h of annealing as Pt and Ni underwent extensive interdiffusion, leading to a relatively uniform distribution within the nanowire interior. Owing to the spatial resolution limits of STEM‐EDS mapping, more subtle diffusion features at longer annealing times were not resolvable. Nevertheless, the combined evidence from Z‐contrast STEM and STEM‐EDS mapping unambiguously indicates pronounced atomic interdiffusion during thermal treatment. These findings collectively demonstrate that during the hydrogen‐assisted ordering process, the one‐dimensional morphology is well preserved, implying limited surface atom migration, while extensive compositional reorganization occurs within the nanowire interior.

To elucidate the structural evolution of I‐PtNi NWs/C during thermal annealing, time‐resolved XRD measurements were conducted (Figure 1f and Figure S12). The as‐synthesized I‐PtNi NWs/C exhibited a disordered fcc structure, with the three major diffraction peaks located between those of pure Pt and Ni, confirming the formation of a PtNi alloy (Figure S12). Notably, the (111) diffraction peak displayed a distinct asymmetry, indicative of phase separation and coexistence of Pt‐rich and Ni‐rich domains rather than a homogeneous alloy phase. This observation is consistent with the compositional inhomogeneity revealed by STEM and STEM‐EDS mapping. After annealing at 550°C for 1 h, the (111) peak became more symmetric and gradually shifted to higher 2θ values, suggesting enhanced alloying and partial atomic rearrangement toward a more homogeneous distribution. Simultaneously, very weak superlattice reflections emerged at 24.8° and 33.1°, corresponding to the (100) and (110) planes of the ordered L1₀ PtNi intermetallic phase (PDF#03‐65‐2797). This indicates that the structural transition from a disordered alloy to an ordered intermetallic phase had already begun. After 2 h of annealing, the intensities of the (100) and (110) superlattice reflections increased, signifying a higher degree of ordering. With further prolongation of annealing, the superlattice peak intensities continued to grow, indicating a progressive enhancement of atomic ordering within the PtNi lattice. The degree of ordering was quantitatively evaluated by comparing the intensity ratios of superlattice to fundamental peaks against standard diffraction data [26]. The calculated ordering degrees increased from 4.5% after 1 h to 19.2% after 2 h and reached 65.8% after 24 h of annealing (Figure 1g), confirming a time‐dependent disorder‐to‐order transformation toward the L1₀ intermetallic structure.

Collectively, these results confirm that the hydrogen adsorption strategy enables precise control over the ordering degree of PtNi NWs while preserving their one‐dimensional morphology during annealing. Importantly, the tunability of atomic ordering achieved here establishes a robust experimental platform for systematically exploring the structure‐ordering‐performance relationship in Pt‐based intermetallic nanocatalysts, especially regarding the influence of ordering degree on their electrocatalytic behavior.

To gain deeper insight into the lattice evolution during the disorder‐order transition, we conducted a detailed examination of the XRD patterns in Figure 1f, with locally magnified reflections shown in Figure 2a–c. The main (111) diffraction peak of I‐PtNi NWs/C exhibits a gradual shift from 42.04° after 1 h to 41.92° after 24 h, indicating a slight increase in the interplanar spacing (Figure 2a). In contrast, the evolution of the (200) and (202) diffraction peaks is more pronounced. The (200) peak shifts markedly from 48.83° to 47.78°, while the (202) peak shifts from 71.72° to 72.24° with prolonged annealing. Since these reflections are dominated by distinct crystallographic directions, their opposite trends imply an anisotropic lattice distortion during ordering. Specifically, the (200) plane is predominantly sensitive to variations along the a‐axis; thus, its shift to lower 2θ reflects lattice expansion along the a‐axis. In contrast, the (202) plane is jointly influenced by both a and c‐axis lattice parameters, and its shift to higher 2theta indicates the presence of lattice contraction along the c‐axis. This asymmetric peak evolution can be rationalized by comparing the crystallographic parameters of the disordered fcc and ordered L1₀ PtNi phases (PDF#97‐010‐5317 and PDF#03‐65‐2797). The lattice of disordered PtNi is cubic (a = b = c = 3.750 Å), while the ordered phase exhibits tetragonal distortion (a = b = 3.823 Å, c = 3.589 Å), confirming that the ordering process introduces biaxial strain, tensile strain along the a and b axes, and compressive strain along the c‐axis. A crystallographic stress analysis of the (111) plane further reveals that expansion along the a and b axes generates tensile stress along the <110> direction, while contraction of the c‐axis induces compressive stress along c‐axis‐related directions such as <101> (Figure S13). Consequently, the competing effects along different crystallographic axes progressively alleviate the compressive strain on the (111) plane as the degree of ordering increases, consistent with the gradual shift of the (111) peak toward lower 2theta (Figure 2a). Crucially, these ex‐situ crystallographic observations are dynamically validated by our in‐situ variable‐temperature XRD measurements (Figure S15). The real‐time patterns explicitly capture the identical direction‐dependent shifting behaviors of these primary reflections during the isothermal phase transition, directly visualizing the simultaneous in‐plane expansion and out‐of‐plane contraction within the heated lattice. This biaxial strain evolution is schematically illustrated in Figure 2d. Notably, similar opposite shifts of the (200) and (202) reflections were also observed for PtNi nanoparticles annealed under identical H₂/Ar conditions, further confirming that the biaxial lattice deformation originates from the intrinsic fcc‐to‐L1₀ crystallographic transition of PtNi (Figure S16). However, the evolution of the (111) reflection in PtNi nanoparticles is much less regular, likely due to pronounced particle growth and aggregation during prolonged annealing, highlighting the advantage of morphology‐stable nanowires for reliably tracking ordering‐induced strain evolution (Figure S16; Figure S17).

FIGURE 2.

Atomic‐scale structural analyses of D‐PtNi NWs/C and H‐PtNi NWs/C. (a–c) Enlarged XRD patterns of I‐PtNi NWs/C after different annealing durations, highlighting the evolution of (111), (200), and (202) planes with increasing ordering degrees. (d) Schematic comparison of lattice deformation along (111) plane in L1₀‐type and fcc‐type PtNi alloys. STEM images of (e) H‐PtNi NWs/C and (j) D‐PtNi NWs/C. AC‐HAADF‐STEM images of selected regions in (e) and (j) for (f) H‐PtNi NWs/C and (k) D‐PtNi NWs/C. (g) Enlarged view of orange‐marked region in (f) and corresponding atomic configuration of L1₀‐PtNi along the [110] zone axis. (h) Z‐contrast intensity profile extracted from the region in (g) and (i) AC‐HAADF‐STEM elemental mappings for H‐PtNi NWs/C. (l) Enlarged view of blue‐marked region in (k) and corresponding atomic configuration of fcc‐PtNi, and (m) Z‐contrast intensity profile extracted from the region in (l) for D‐PtNi NWs/C.

To further investigate the origin of the biaxial strain observed during the ordering of I‐PtNi NWs/C, we selected three representative samples from the time‐resolved experiments for detailed microscopic characterization. The sample annealed at 550°C for 1 h, with an ordering degree of 4.5%, was considered nearly fully disordered and designated D‐PtNi NWs/C. Samples annealed for 5 and 24 h, with 36.2% and 65.8% ordering, were denoted as M‐PtNi NWs/C (medium‐order) and H‐PtNi NWs/C (high‐order), respectively. High‐resolution transmission electron microscopy (HRTEM) combined with fast Fourier transform (FFT) analysis was employed to evaluate the atomic ordering in these samples. For D‐PtNi NWs/C, three random positions along a single nanowire were analyzed. The FFT patterns along [111] and [100] directions matched the standard fcc structure, and no obvious superlattice reflections were observed, confirming the ultra‐low ordering consistent with the XRD‐derived results (Figure S18a). In contrast, M‐PtNi NWs/C exhibited weak but discernible superlattice spots corresponding to the (110) and (001) planes in localized regions, indicating the formation of partially ordered domains (Figure S18b). For H‐PtNi NWs/C, all three examined regions displayed prominent (001) superlattice reflections, reflecting a significant increase in atomic ordering (Figure S18c). These observations are in excellent agreement with the XRD‐based ordering quantification and confirm the time‐dependent evolution from disordered alloy to intermetallic phase.

To precisely quantify the biaxial strain in the three representative samples, aberration‐corrected high‐angle annular dark‐field scanning transmission electron microscopy (AC‐HAADF‐STEM) was performed (Figure 2e–m; Figure S19). For H‐PtNi NWs/C, high‐magnification AC‐HAADF‐STEM images allowed measurement of lattice spacings of the (100) and (111) planes, yielding 0.379 and 0.226 nm, respectively (Figure 2f). In M‐PtNi NWs/C, the corresponding lattice spacings were 0.368 nm for (100) and 0.221 nm for (111), both compressed relative to H‐PtNi NWs/C (Figure S19a). For D‐PtNi NWs/C, the (111) plane spacing was 0.211 nm (Figure 2k), showing the largest compressive strain, again in agreement with XRD results. These findings directly confirm that the biaxial strain, with expansion along the a and b axes and compression along the c‐axis, diminishes the compressive strain on the (111) plane as the ordering degree increases. Further structural insights were obtained by magnifying local regions in high‐resolution AC‐HAADF‐STEM images (Figure 2g,l; Figure S19b–e). In H‐PtNi NWs/C, magnification of the highlighted orange region revealed alternating bright and dark atomic columns, corresponding to Pt and Ni, respectively, consistent with the simulated L1₀‐type ordered PtNi structure along [110] (Figure 2g). The atomic‐resolution elemental mapping of H‐PtNi NWs/C further confirmed the long‐range chemical ordering (Figure 2i). For M‐PtNi NWs/C, the blue region showed disordered atomic columns, whereas the orange region displayed periodic atomic arrangement matching the L1₀‐type structure, indicating coexistence of disordered fcc and partially ordered L1₀ domains (Figure S19b–e). In D‐PtNi NWs/C, all examined regions exhibited random atomic columns, confirming the fully disordered fcc structure. Importantly, the Z‐contrast intensity profiles extracted from the blue region of Figure 2l (corresponding to D‐PtNi NWs/C) and the orange region of Figure 2g (corresponding to H‐PtNi NWs/C) provide a direct visualization of the lattice expansion associated with strain evolution. The (111) lattice spacing increases from 0.211 nm in D‐PtNi NWs/C to 0.226 nm in H‐PtNi NWs/C, consistent with the gradual relaxation of compressive strain revealed by the XRD peak shifts. Collectively, these results demonstrate that the three representative samples possess distinct crystal structures, which induce different degrees of biaxial strain within the nanowires. This biaxial strain is responsible for the gradual reduction of compressive strain in the (111) plane as the PtNi nanowires become increasingly ordered, providing a microscopic explanation for the strain evolution observed in XRD.

Subsequently, x‐ray photoelectron spectroscopy (XPS) and x‐ray absorption spectroscopy (XAS) were employed to elucidate how the ordering degree affects the electronic structure and local coordination environment of Pt. For the Pt 4f XPS spectra (Figure 3a), the peaks can be deconvoluted into contributions from metallic Pt (Pt⁰) and oxidized Pt (Pt²⁺). Among the three samples, H‐PtNi NWs/C exhibits the highest proportion of metallic Pt (88.9%), followed by M‐PtNi NWs/C (83.2%) and D‐PtNi NWs/C (50.3%). This trend indicates that the stability of metallic Pt species increases progressively with the ordering degree, reflecting a reduced susceptibility to surface oxidation in the ordered intermetallic structure. Moreover, the Pt 4f _7/2 binding energy exhibits a systematic negative shift for M‐PtNi NWs/C and H‐PtNi NWs/C compared to D‐PtNi NWs/C (Table S1). In particular, H‐PtNi NWs/C shows a negative shift of 0.11 eV, indicative of enhanced Pt‐Ni orbital hybridization and electron transfer from Ni to Pt in the more ordered structure. This conclusion is corroborated by the Ni 2p spectra, where the Ni 2p3/2 binding energy of H‐PtNi NWs/C shifts positively by 0.13 eV relative to D‐PtNi NWs/C, confirming the directional charge redistribution from Ni to Pt (Figure S20 and Table S1).

FIGURE 3.

Electronic and coordination structure evolution of D‐PtNi NWs/C, M‐PtNi NWs/C and H‐PtNi NWs/C. (a) Pt 4f XPS spectra of D‐PtNi NWs/C, M‐PtNi NWs/C and H‐PtNi NWs/C. (b) Pt L₃‐edge XANES spectra, (c) corresponding FT‐EXAFS spectra, and (d) WT‐EXAFS spectra of Pt foil, D‐PtNi NWs/C, M‐PtNi NWs/C and H‐PtNi NWs/C.

To probe the local atomic environment of Pt, x‐ray absorption near‐edge structure (XANES) and extended x‐ray absorption fine structure (EXAFS) analyses were conducted at the Pt L₃‐edge (Figure 3b–d and Figure S21a–d). In the XANES spectra, the white‐line intensity of H‐PtNi NWs/C and M‐PtNi NWs/C is significantly lower than that of D‐PtNi NWs/C and approaches that of Pt foil, indicating higher d‐band electron occupancy and lower oxidation states for Pt in the more ordered samples (Figure 3b). This observation agrees well with the XPS results and demonstrates the increasing metallic character of Pt upon ordering, resulting from stronger Pt‐Ni hybridization and lattice strain modulation. The EXAFS radial distribution functions reveal that the main scattering peaks of all PtNi samples appear at shorter distances than metallic Pt, consistent with lattice contraction caused by alloying with smaller Ni atoms (Figure 3c). To accurately capture this evolving local coordination environment and the associated anisotropic lattice distortion, we conducted quantitative EXAFS fitting using structural models that account for the transition from the isotropic cubic phase to the symmetry‐breaking tetragonal phase. Specifically, the theoretical fitting for the H‐PtNi NWs/C and M‐PtNi NWs/C strictly employed a tetragonal L1₀‐PtNi model featuring characteristic biaxial parameters (a = b = 3.823 Å, c = 3.589 Å), contrasting with the isotropic cubic fcc‐PtNi reference model (a = b = c = 3.750 Å) utilized for the disordered sample. Quantitative fitting shows that the Pt‐Pt bond lengths increase from 2.66 Å for D‐PtNi NWs/C to 2.68 and 2.69 Å for M‐PtNi NWs/C and H‐PtNi NWs/C, respectively (Table S2). This progressive increase in Pt‐Pt bond length with ordering reflects the lattice expansion along the a–b axes during the disordered‐to‐ordered transition, consistent with the biaxial strain mechanism revealed by XRD and AC‐HAADF‐STEM. To further visualize the evolution of local coordination, wavelet transform (WT) analysis of the k³‐weighted EXAFS signals was performed (Figure 3d). WT contour plots offer simultaneous resolution in k‐ and R‐space, enabling discrimination of different backscattering paths. For all samples, the dominant WT maxima appear at k = 7–8 Å⁻¹ and R = 2.3‐2.7 Å, corresponding to the first‐shell Pt‐Pt and Pt‐Ni coordination. As the ordering degree increases from D‐PtNi NWs/C (4.5%) to H‐PtNi NWs/C (65.8%), the WT maxima shift to higher R values and become sharper and more intense, signifying longer Pt–Pt distances and increased structural order. Notably, the WT features of H‐PtNi NWs/C are more concentrated and better defined in k‐space, reflecting a more uniform and periodic local environment around Pt atoms in the ordered intermetallic phase. These spectroscopic findings are consistent with the XRD and AC‐HAADF‐STEM results. They confirm that the biaxial lattice strain, originating from simultaneous a–b expansion and c‐axis compression, modulates Pt–Pt bond lengths and Pt‐Ni electronic interactions. From an electronic structure perspective, increasing ordering strengthens Pt‐Ni orbital hybridization and promotes charge transfer from Ni to Pt, which in turn alleviates the excessive compressive strain present in disordered and moderately ordered samples. As a result, the highly ordered H‐PtNi NWs/C exhibits a more balanced compressive strain state, leading to an optimized Pt d‐band center and more appropriate oxygen adsorption strength, thereby facilitating ORR kinetics. Overall, XPS and XAS analyses consistently indicate that increasing the ordering degree of I‐PtNi NWs/C induces charge redistribution from Ni to Pt and systematically tunes the local electronic structure and coordination environment of Pt. This structural‐electronic coupling provides the fundamental basis for the enhanced activity and durability of PtNi IMCs catalysts.

To elucidate the correlation between atomic ordering and electrocatalytic behavior, the ORR performances of D‐PtNi NWs/C, M‐PtNi NWs/C and H‐PtNi NWs/C were systematically evaluated using cyclic voltammetry (CV) and linear sweep voltammetry (LSV), with commercial Pt/C as a benchmark (Figure 4a,b). CV measurements were first conducted in N₂‐saturated 0.1 M HClO₄ electrolyte to determine the electrochemical surface area (ECSA) from the hydrogen underpotential deposition (H _upd) region (Figure 4a). Subsequently, LSV measurements were performed in an O₂‐saturated electrolyte to assess the ORR activity of the catalysts (Figure 4b). The calculated ECSAs for D‐PtNi NWs/C, M‐PtNi NWs/C, H‐PtNi NWs/C and commercial Pt/C were 27.5, 28.3, 24.0, and 70.0 m² g_Pt ⁻¹, respectively. Despite the lower ECSA, H‐PtNi NWs/C exhibited the highest ORR activity among all catalysts (Figure 4b). At 0.9 V versus RHE, the mass activity (MA) of H‐PtNi NWs/C reached 1.30 A mg_Pt ⁻¹, outperforming M‐PtNi NWs/C (0.74 A mg_Pt ⁻¹), D‐PtNi NWs/C (0.48 A mg_Pt ⁻¹) and commercial Pt/C (0.15 A mg_Pt ⁻¹) (Figure 4c). Similarly, the specific activity (SA) normalized to ECSA exhibited the same trend, with H‐PtNi NWs/C (5.4 mA cm⁻²) showing superior activity compared with M‐PtNi NWs/C (2.61 mA cm⁻²), D‐PtNi NWs/C (1.73 mA cm⁻²) and commercial Pt/C (0.23 mA cm⁻²), representing an approximately 23‐fold enhancement relative to commercial Pt/C. As a control, L1₀‐PtNi NPs/C obtained by annealing PtNi NPs/C under identical H₂/Ar conditions for 24 h was also evaluated, delivering a MA of 0.68 A mg_Pt ⁻¹ and a SA of 1.4 mA cm⁻², both lower than those of H‐PtNi NWs/C (Figure S22). Tafel slope analysis further provides insight into the reaction kinetics (Figure S23). H‐PtNi NWs/C exhibits the lowest Tafel slope (87.4 mV dec⁻¹), followed by M‐PtNi NWs/C (87.6 mV dec⁻¹), D‐PtNi NWs/C (93.8 mV dec⁻¹) and commercial Pt/C (114.8 mV dec⁻¹). The progressive decrease in Tafel slope with increasing ordering degree suggests that the ordered intermetallic structure effectively lowers the energy barrier of the rate‐determining step, thereby accelerating ORR kinetics. To establish a direct structure‐performance correlation, the compressive strain of the (111) facet was determined by calculating the interplanar spacing from the lattice parameters obtained through quantitative XRD Rietveld refinement (Figure S24 and Table S3). To further enrich this correlation, we additionally evaluated two intermediate samples annealed for 2 h and 12 h (denoted as I‐PtNi NWs/C‐2 and I‐PtNi NWs/C‐12), which exhibited MAs of 0.60 and 0.92 A mg_Pt ⁻¹, respectively (Figure 4b,c). As the ordering degree increases from D‐PtNi NWs/C to H‐PtNi NWs/C, the compressive strain gradually decreases, accompanied by a monotonic rise in MA (Figure 4d). This trend can be rationalized by considering that disordered PtNi alloys experience excessive lattice compression due to random Ni incorporation, which can overly downshift the Pt d‐band center and weaken the adsorption of oxygenated intermediates (*O, *OH) beyond the optimal range. Upon increasing ordering, biaxial lattice relaxation occurs, leading to a more appropriate compressive strain state in H‐PtNi NWs/C. This optimized strain, together with enhanced Pt‐Ni orbital hybridization and charge transfer, fine‐tunes the Pt d‐band center to an energetically favorable position, achieving balanced oxygen adsorption strength and thus superior intrinsic activity.

FIGURE 4.

Electrochemical performance evaluation of H‐PtNi NWs/C and its counterparts. (a) CV curves, (b) LSV curves, and (c) MAs and SAs of Pt/C and I‐PtNi NWs/C with varying ordering degrees. (d) Correlation of ordering degree with MAs and lattice strain. (e–g) Durability tests of H‐PtNi NWs/C, M‐PtNi NWs/C and D‐PtNi NWs/C with insets showing corresponding CV curves before and after cycling. (h) TEM image of D‐PtNi NWs/C after stability testing. (i) Fuel cell polarization curves of H‐PtNi NWs/C, D‐PtNi NWs/C and commercial Pt/C under H₂‐air conditions. Fuel cell polarization curves of (j) H‐PtNi NWs/C and (k) commercial Pt/C before and after AST. (l) TEM image, HRTEM image, and corresponding FFT pattern of H‐PtNi NWs/C after AST.

Catalyst durability is another key criterion for practical applications. The accelerated stress test (AST) was conducted by cycling the potential between 0.6 and 1.0 V (vs. RHE) in O₂‐saturated 0.1 M HClO₄ (Figure 4e–g, Figures S25 and S26). The activity retention was evaluated by comparing CV and ORR polarization curves before and after AST. Remarkably, even after the 60 000 cycles, H‐PtNi NWs/C demonstrated outstanding electrochemical robustness, retaining 98.2% of its initial ECSA and 96.5% of its SA. Furthermore, its MA underwent only an expected slight attenuation with a high retention rate of 86.0% (Figure S23 and Table S4). Meanwhile, the control samples including M‐PtNi NWs/C, D‐PtNi NWs/C, and commercial Pt/C underwent a standard 30 000‐cycle protocol. M‐PtNi NWs/C showed negligible ECSA change, indicating good structural stability, while D‐PtNi NWs/C experienced a pronounced decrease from 27.5 to 22.1 m² g_Pt ⁻¹ (19.6% loss) (Figure S26). In contrast, commercial Pt/C underwent a severe ECSA loss from 73.9 to 36.7 m² g_Pt ⁻¹ (50.3% loss) (Figure S26). Consistent with these trends, M‐PtNi NWs/C and D‐PtNi NWs/C exhibited MA losses of 9.7% and 33.3%, respectively (Figure S26). Notably, all three PtNi nanowire catalysts demonstrated significantly better durability than commercial Pt/C. The origin of enhanced catalyst stability can be further elucidated by examining the morphology and structure of the catalysts after AST. TEM and EDS analyses were performed on the cycled catalysts. TEM, HRTEM, and EDS analyses reveal that H‐PtNi NWs/C perfectly preserved its one‐dimensional nanowire morphology and long‐range crystalline ordering even after 60 000 cycles, with only minimal Ni leaching observed (Figure 4l and Figure S30a,b). To explicitly verify whether the performance‐enhancing biaxial strain is preserved after such an extended operation and slight compositional change, we conducted additional HRTEM analyses on the cycled H‐PtNi NWs/C to precisely measure the interplanar spacings of the (111) crystal plane (Figure S30c–h). Measurements across three distinct surface regions yielded values of 0.225, 0.224, and 0.224 nm. These expanded spacings are remarkably consistent with the initial pre‐cycled value of 0.226 nm and remain significantly larger than the 0.211 nm spacing observed in the completely disordered D‐PtNi NWs/C. This direct microscopic evidence definitively proves that the ordering‐induced biaxial strain is robustly maintained despite minor Ni leaching. It is precisely the preservation of this optimal strain state that fundamentally explains why H‐PtNi NWs/C securely retains 86.0% of its exceptional mass activity even after 60 000 demanding cycles. Similarly, M‐PtNi NWs/C maintained its structural features after a 30 000‐cycle test (Figure S29). In contrast, commercial Pt/C exhibited significant particle coarsening (Figure S27). D‐PtNi NWs/C underwent pronounced structural degradation, including morphological distortion and extensive Ni loss (Figure 4h, Figure S28). The leaching of Ni not only weakens the Pt‐Ni electronic interactions but also compromises the structural integrity of the catalyst surface, ultimately leading to diminished ORR activity of D‐PtNi NWs/C. The enhanced durability of H‐PtNi NWs/C and M‐PtNi NWs/C arises from the synergistic contributions of their unique architecture and atomic structure. Their one‐dimensional interconnected morphology ensures multiple anchoring points and extensive catalyst‐support contact, effectively suppressing nanoparticle detachment and coalescence during cycling. At the same time, the ordered intermetallic framework provides strong resistance against atomic rearrangement and metal dissolution, which mitigates Ostwald ripening and prevents structural degradation. These combined features endow the catalysts with exceptional long‐term stability under harsh electrochemical conditions. Collectively, our results demonstrate that increasing atomic ordering in PtNi nanowires not only enhances initial ORR activity but also significantly improves long‐term durability. The synergistic interplay between atomic‐scale ordering, structural integrity, and compositional stability provides a reliable design principle for next‐generation ORR electrocatalysts in PEMFCs.

To further evaluate their practical applicability, membrane electrode assemblies (MEAs) were fabricated using H‐PtNi NWs/C, D‐PtNi NWs/C and commercial Pt/C as cathode catalysts, and their single‐cell fuel cell performances were systematically examined (Figure 4i–k). The MEAs were operated under H₂‐air conditions at 80°C and 150 kPa_abs with fully humidified gases to simulate realistic PEMFC operating environments. All cathode catalysts had the same Pt loading to ensure a fair comparison of intrinsic catalytic behavior. Among the three catalysts, the MEA based on H‐PtNi NWs/C exhibited the highest peak power density, reaching 1014.5 mW cm⁻², which is significantly higher than those of D‐PtNi NWs/C (817.7 mW cm⁻²) and commercial Pt/C (636.4 mW cm⁻²) (Figure 4i, Figure S31a). At 0.7 V, the current density of the H‐PtNi NWs/C MEA reached 1.04 A cm⁻², outperforming D‐PtNi NWs/C (0.86 A cm⁻²) and commercial Pt/C (0.49 A cm⁻²) (Figure S31a). These superior cell performances can be attributed to the unique combination of one‐dimensional morphology and ordered intermetallic structure in H‐PtNi NWs/C, which ensures efficient charge/mass transport, high intrinsic ORR activity, and robust catalyst‐ionomer interfaces within the catalyst layer. Durability tests were further carried out by subjecting the MEAs to 30 000 square‐wave potential cycles between 0.6 V and 0.9 V to mimic dynamic operating conditions (Figure 4j,k and Figure S32). After durability testing, the current density at 0.7 V of the H‐PtNi NWs/C MEA decreased by only 2.0%, much lower than the degradation observed for D‐PtNi NWs/C (40.7%) and Pt/C (46.9%) (Figure S31b). The slower performance loss of H‐PtNi NWs/C is consistent with its superior electrochemical and structural stability observed in half‐cell measurements, confirming that the ordered intermetallic nanowire architecture effectively suppresses degradation phenomena such as Ni leaching, nanoparticle coalescence, and mass‐transport limitations during prolonged operation. Overall, these single‐cell tests demonstrate that the advantages of H‐PtNi NWs/C observed at the half‐cell level, including high intrinsic activity, structural robustness, and durability, translate effectively into MEA configurations. Furthermore, a comprehensive benchmarking analysis reveals that the overall performance of H‐PtNi NWs/C exhibits significant advantages over various recently reported state‐of‐the‐art Pt‐based catalysts (Table S4), underscoring its strong potential as a next‐generation cathode catalyst for practical PEMFC applications.

The enhanced ORR performance of H‐PtNi NWs/C may originate from its strain‐modulated surface that adjusts the binding strengths of intermediates such as *O₂, *OOH, *O and *OH, thereby facilitating faster turnover of active sites. Among these species, *O adsorption has been widely recognized as an effective descriptor of ORR kinetics on Pt‐based catalysts, because excessively strong *O binding inhibits the formation and removal of subsequent intermediates, whereas overly weak binding limits O‐O activation [43, 46, 47]. The unique combination of tensile strain in the a‐b plane and compressive strain along the c‐axis in H‐PtNi NWs/C is therefore anticipated to fine‐tune the intrinsic *O adsorption energy toward a more optimal range, thereby promoting faster ORR kinetics (Figure 5a). This strain‐mediated surface electronic modulation is consistent with the experimentally observed high intrinsic activity of H‐PtNi NWs/C. To validate this structural hypothesis, density functional theory (DFT) calculations were performed on the (111) facet, which serves as the primary active surface for Pt‐based catalysts and was therefore selected as our benchmark model. For the disordered PtNi alloy with an fcc structure, the incorporation of Ni into the Pt lattice induces a nearly isotropic compressive strain along the <110>, <101>, and <011> directions, as shown in Figures S14 and S33. In contrast, the ordered PtNi alloy with the L1₀ structure experiences biaxial lattice distortion, featuring tensile strain along the a and b axes and compressive strain along the c‐axis, resulting in anisotropic compressive strain on the (111) plane (Figures S13 and S33). To accurately capture these structural variations, we established surface models featuring the adsorbed *O intermediate on Pt(111) under two distinct strain regimes. These regimes encompass isotropic compression through uniform scaling of the a, b, and c axes (Figure S34) and L1₀‐symmetry‐breaking‐induced biaxial strain defined by a coupled in‐plane expansion and out‐of‐plane contraction (Figure S35). By evaluating these specific configurations, we systematically calculated the resulting oxygen adsorption energies (E _O). To establish a clear thermodynamic baseline, the optimal oxygen adsorption energy on Pt(111) was set as 0 eV, and ΔE_O was defined as the deviation from this reference. The results show that while it maintains a linear correlation with compressive strain in both regimes, significant thermodynamic shifts and distinct scaling slopes occur between the two models (Figure 5b). Specifically, the thermodynamic optimum for the isotropic compression model is identified at a strain of 3.1%. In contrast, the realistic biaxial strain model reveals a markedly shifted optimum of 3.8% (Figure 5b). This discovery provides a more nuanced mechanistic understanding of ordered intermetallic catalysts, illustrating that the anisotropic lattice distortion induced during the disorder‐to‐order transition reconfigures the surface electronic states and the Pt d‐band center in a fundamentally distinct manner from conventional isotropic compression. By explicitly differentiating the thermodynamic boundaries of these two strain regimes, we refine the theoretical scaling relationships to demonstrate that symmetry‐breaking deformation imposes a unique and specific regulatory effect on adsorbate energetics. Crucially, the experimental compressive strain of the highly ordered H‐PtNi NWs/C (4.71%, derived from quantitative Rietveld refinement) is significantly closer to the newly identified theoretical biaxial optimum of 3.8% compared to the disordered D‐PtNi NWs/C (5.18%), which suffers from excessive lattice compression. This strain relaxation effect brings the surface oxygen binding energy back toward the peak of the thermodynamic volcano, laying the foundation for enhanced intrinsic activity. To further elucidate the reaction kinetics, full free energy diagrams for the ORR pathway were calculated to directly compare the strain‐free Pt(111) reference facet against the H‐PtNi NWs(111) surface under a 4.71% biaxial strain (Figure 5c and Figure S36). At a potential of U = 0 V, all intermediate transformation steps are exergonic on both surfaces. However, at the equilibrium potential of U = 1.23 V, the thermodynamic bottlenecks become evident. The results confirm that the protonation of *O to form the hydroxyl intermediate serves as the rate‐determining step (RDS) for both systems. Notably, the H‐PtNi NWs(111) surface under 4.71% biaxial strain exhibits a substantially reduced energy barrier of 0.39 eV for this RDS, compared to 0.51 eV for the strain‐free Pt(111) surface. This kinetic reduction suggests that the unique biaxial strain state optimizes the surface energetics by weakening the excessively strong Pt‐O interaction, thereby accelerating the overall turnover frequency and granting the catalyst superior mass activity and kinetic current density. Finally, DFT calculations were conducted to understand the superior structural stability of the ordered L1₀ phase. The vacancy formation energy of Ni was evaluated for both ordered L1₀‐PtNi and disordered fcc‐PtNi alloys by substituting a neighboring Pt atom with a Ni atom to mimic the fcc configuration (Figures S37 and S38). As shown in Figure 5d, the vacancy formation energy of Ni in L1₀‐PtNi is significantly higher than that in fcc‐PtNi, indicating that Ni dissolution is thermodynamically less favorable in the ordered structure. This finding rationalizes the superior durability of H‐PtNi NWs/C under acidic ORR conditions, underscoring the role of atomic ordering in stabilizing the catalytic framework.

FIGURE 5.

Theoretical analyses of activity and stability differences for different catalysts. (a) Schematic illustration of O₂ adsorption and dynamic transformation pathway on fcc‐PtNi and L1₀‐PtNi(111) surfaces. (b) Calculated ΔE _O on Pt (111) as a function of strain, including isotropic compressive strain and biaxial strain, where the optimal E _O is set to zero. Insets show the corresponding strain models. (c) Free energy diagrams for the ORR on Pt(111) and H‐PtNi NWs(111) at U = 0 and 1.23 V. (d) Comparison of vacancy formation energies of Ni in L1₀‐PtNi and fcc‐PtNi.

3. Conclusion

In summary, we have developed a controllable ordering strategy to precisely tune the lattice strain and catalytic behavior of PtNi NWs. By regulating the annealing process, a series of I‐PtNi NWs with varying ordering degrees (4.5% to 65.8%) were synthesized, enabling a systematic quantitative correlation of ordering degree, lattice strain, and catalytic activity. As the ordering degree increases, a distinct biaxial strain emerges, which involves in‐plane lattice expansion and out‐of‐plane contraction and progressively alleviates the overall compressive stress on the Pt(111) surface, a trend opposite to the conventional expectation that higher ordering leads to stronger compression. This strain evolution strengthens the Pt‐Ni orbital hybridization and yields favorable oxygen adsorption energetics for accelerated ORR kinetics by significantly lowering the energy barrier of the RDS from 0.51 to 0.39 eV. Consequently, the H‐PtNi NWs/C exhibits an MA of 1.30 A mg_Pt ⁻¹, along with superior durability, retaining 86.0% of this activity after 60 000 cycles, and excellent fuel‐cell performance with a high peak power density of 1014.5 mW cm⁻². The ordered intermetallic framework further suppresses Ni leaching and structural degradation, ensuring long‐term operational stability. Overall, this work identifies ordering‐induced biaxial strain as a structurally informative parameter for optimizing the activity‐stability balance of Pt‐based ORR catalysts and offers insights for the design of engineering robust intermetallic nanostructures for practical PEMFCclearcharge applications.

Author Contributions

Xing Hu: data curation, formal analysis, visualization, validation, methodology, software, investigation, project administration, writing – original draft. Shize Geng: validation, formal analysis, data curation. Yu Cao: data curation, software, methodology. Kezhu Jiang: conceptualization, funding acquisition, visualization, project administration, resources, writing – original draft, writing – review and editing, supervision, formal analysis. Yibo Liu: methodology, validation, visualization. Ruifan Li: formal analysis, visualization, investigation. Yangyang Zhang: methodology, software, data curation. Shuang Meng: methodology, software, validation. Shan Zhu: methodology, visualization, formal analysis. Lingzheng Bu: writing – review and editing, conceptualization, resources. Cong Chen: formal analysis, validation, visualization, writing – review and editing, resources. Shijian Zheng: writing – review and editing, conceptualization, resources, methodology, supervision.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: anie73655‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.