Pt‐SnO_x Clusters With Weak Interfacial Interaction Boost Hydrogen Evolution

ABSTRACT

Downsizing noble metal catalysts (e.g., Pt) is critical for enhancing atomic efficiency in sustainable energy applications. Herein, we report a Pt–SnO_x nanocluster (NC) system featuring weak Pt/SnO_x interfacial interaction, which enables SnO_x to effectively modulate the reaction pathway without compromising the intrinsic Pt catalytic properties. Notably, characterization results reveal negligible charge transfer at the interface. The Pt–SnO_x NCs exhibit an overpotential of 22 mV at 10 mA cm⁻² and a Tafel slope of 24 mV dec⁻¹, outperforming Pt NCs, Pt nanoparticles, and Pt–SnO₂ structures with strong interfacial interactions. Theoretical calculations suggest that the Pt–SnO_x architecture facilitates hydrogen spillover, which may promote a shift in the dominant hydrogen evolution reaction (HER) mechanism from Volmer–Heyrovsky to a more efficient Volmer–Tafel route. As the cathode in an ethanol‐assisted proton exchange membrane paired‐electrolyzer, the Pt–SnO_x NC catalyst delivers 1.0 A cm⁻² at below 1.5 V under 80°C with stable operation over 70 h.

A weak interfacial interaction is observed in the Pt‐SnO_x nanoclusters, accompanied by negligible charge transfer. This unique feature enables SnO_x to effectively steer the reaction pathway toward a more efficient Volmer–Tafel route without compromising the intrinsic hydrogen evolution reaction (HER) activity of Pt. The Pt‐SnO_x nanocluster catalysts can be effectively utilized in a proton exchange membrane electrolyzer.

1. Introduction

Hydrogen, with its high energy density (120 MJ kg⁻¹) and zero greenhouse gas emissions, is widely regarded as a promising alternative to fossil fuels and plays a crucial role in advancing renewable and sustainable energy development [1, 2]. However, its current production predominantly relies on steam reforming of hydrocarbons, a process that generates significant CO₂ emissions, contributing to the greenhouse effect, and produces CO, which can poison low‐temperature fuel cells. Alternatively, electrochemical water splitting is viewed as a green, clean, and highly efficient technology for large‐scale production of high‐purity hydrogen via the hydrogen evolution reaction (HER), a two‐electron transfer process, 2H⁺ + 2e⁻ → H₂ [3, 4]. Balancing activity and stability, platinum (Pt) is currently the most effective catalyst for HER [5, 6]. Nørskov and co‐workers employed Gibbs free energy calculations for hydrogen adsorption (ΔG _H*) to predict a volcano‐shaped curve for HER catalysts, identifying Pt group metals near the apex of the plot as the most efficient catalysts.

Numerous studies have focused on optimizing Pt‐based catalysts for the HER, driven primarily by the scarcity and high cost of Pt [7, 8, 9, 10, 11, 12]. Currently, one mainstream approach to reduce Pt usage involves constructing Pt‐based alloys or core/shell structures, both of which inherently induce strong electronic coupling and interfacial interactions between Pt and the introduced foreign components [13, 14]. However, these strategies are not without drawbacks. For instance, Pt–M alloys (where M denotes foreign metals) often exhibit relatively poor stability in acidic media due to the leaching of less noble components, a critical limitation for HER under pH‐universal conditions, particularly in proton exchange membrane (PEM) electrolayers [15]. Additionally, the strong electronic coupling in Pt–M alloys may excessively alter the ΔG _H* of Pt sites, shifting it away from the optimal apex of the HER volcano plot and thus degrading intrinsic catalytic activity. This dilemma is not exclusive to alloys but also persists in Pt‐based core‐shell structures. While these core–shell designs aim to introduce value‐added functionalities (e.g., corrosion resistance) at the Pt/M interface, strong interfacial interactions with the foreign component still trigger undesirable perturbations of Pt's electronic structure, leading to similar ΔG _H ^* deviations from the ideal value.

Another mainstream design, downsizing Pt nanoparticles (NPs) to nanoclusters (NCs) or even single atoms (SAs), has emerged as an alternative critical strategy [16, 17]. Yet this approach also carries an inherent trade‐off: unlike large Pt NPs with bulk‐like electronic properties, Pt species scaled down to the NC or SA regime exhibit high sensitivity of their electronic structures to quantum confinement effects and coordination environments. This manifests in significantly altered electronic configurations (e.g., modified d‐band centers, enhanced orbital hybridization, or distorted local coordination geometries), which inherently drive ΔG _H ^* away from the thermodynamically optimal value. As recently highlighted by Whang et al., such structural and electronic perturbations in low‐loading Pt catalysts disrupt their intrinsically favorable electronic characteristics, thereby complicating the HER reaction mechanism and hindering the attainment of theoretical performance limits [18].

Therefore, developing an effective strategy to further optimize low‐loading Pt HER catalysts that incorporates an additional functional secondary species while preserving the intrinsically superior electrochemical performance of Pt, represents a promising avenue. However, such studies remain relatively scarce to date. Several Pt/SnO₂‐ and Pt/F‐SnO₂‐based catalysts have been reported for acidic HER, in which the improved activity is generally attributed to Pt–SnO₂ interfacial electronic modulation, strong metal–support interaction, F‐doping‐induced stabilization, or optimized H* adsorption/desorption behavior [19, 20]. These studies have demonstrated the importance of SnO₂‐containing interfaces in regulating Pt‐based HER catalysis. Nevertheless, most reported systems rely on relatively strong Pt–support interactions or heteroatom‐doped oxide matrices to modify the electronic structure of Pt. In such cases, the intrinsic Pt catalytic properties and active‐site accessibility may be partially perturbed, especially when Pt is used at low loading. Therefore, a distinct design strategy that introduces a functional oxide species while maintaining a weak Pt–oxide interaction remains highly desirable but has been less explored.

In this context, amorphous oxides provide a rational platform for constructing weak metal–oxide interactions. Unlike crystalline oxides with long‐range ordered lattices and well‐defined interfacial bonding, amorphous oxides possess disordered atomic arrangements, abundant unsaturated sites, lattice distortions, and locally diverse coordination environments [21, 22]. Such structural disorder can weaken both homonuclear and heteronuclear atomic interactions, thereby leading to relatively weak contact at metal–oxide interfaces [23]. This concept has been explicitly demonstrated in Ru/HfO₂ heterostructures, where the disordered arrangement in amorphous materials was reported to generate weak interactions, resulting in weak Ru–HfO₂ contact. Moreover, amorphous Sn‐containing oxides have been shown to contain multiple dangling bonds and disordered internal lattice arrangements, while maintaining catalytic functionality and stability during electrochemical reactions [21, 22], suggesting that amorphous SnO_x can serve as a suitable secondary oxide phase for interfacial regulation without necessarily imposing strong electronic perturbation on the metal component.

Herein, we report a Pt–SnO_x NC structure fabricated via a conventional deposition‐precipitation method, which effectively enhances the performance of low‐loading Pt‐based HER catalysts. Experimental characterizations reveal a fairly weak interaction at the interface between Pt and SnO_x with almost negligible charge transfer occurring. This enables the preservation of Pt intrinsic catalytic properties while leveraging the functional benefits of the introduced SnO_x species. Compared with the pristine Pt NCs and Pt NPs, the Pt–SnO_x NCs exhibit improved HER activity and stability, achieving an overpotential of 22 mV at 10 mA cm⁻² and a Tafel slope of 24 mV dec⁻¹. Structural control experiments further confirm that the HER performance is compromised when strong interfacial interactions are present in traditional Pt–SnO₂ heterostructures. Combining experimental observations and theoretical calculations, we reveal that the enhanced activity of Pt–SnO_x NCs originates from a hydrogen spillover mechanism promoted by SnO_x sites. This mechanism may shift the HER pathway from the Volmer–Heyrovsky route to the more efficient Volmer–Tafel route and reduces the energy barrier for H₂ recombination. More importantly, we verify the applicability of this low‐Pt‐loading catalyst in PEM electrolyzers. When employed as the cathode with ethanol oxidation at the anode (an ethanol‐assisted proton exchange membrane paired‐electrolyzer), the Pt–SnO_x NC catalyst achieves 1.0 A cm⁻² at < 1.5 V at 80°C and maintains stable operation for over 70 h. This work highlights that a key to enhancing the HER performance of low‐Pt‐loading catalysts lies in the design of weak interfacial interactions, which enables the introduction of a secondary species with additional functionalities while preserving Pt's favorable ΔG _H*.

2. Results and Discussion

Carbon‐supported Pt NCs (2.9 wt%) and Pt NPs (19.8 wt%) were synthesized via the deposition‐precipitation method, followed by H₂ reduction at 470 K for 1 h. Carbon‐supported Pt–SnO_x NCs (2.7 wt% Pt, Pt/Sn = 59/41) were prepared using a similar procedure (Table S1). Figure 1 presents the characterization result of the as‐prepared Pt‐based catalysts. First, aberration‐corrected high‐angle annular dark‐field scanning transmission electron microscopy (HAADF‐STEM) was employed to elucidate the atomic‐scale structures. Figure 1a reveals the nanoparticle morphology of the Pt/C catalyst. In contrast, Figure 1b displays highly dispersed Pt NCs with an average diameter of approximately 1.2–1.5 nm. Remarkably, Figure 1c provides direct visualization of a distinct heterostructure, where high‐contrast PtNCs are intimately decorated with low‐contrast SnO_x species, highlighting the marked difference in atomic number (Z‐contrast) between the two components. This structural configuration is further corroborated by the corresponding energy‐dispersive X‐ray spectroscopy (EDS) elemental maps, which clearly confirm the spatial distribution of Pt and Sn. These results collectively demonstrate the successful synthesis of ultrafine Pt–SnO_x heterostructures with overall dimensions below 2.0 nm. More examples are provided in Figures S1 and S2.

FIGURE 1.

Structure characterization of Pt‐based HER catalysts. (a–c) HAADF‐STEM image of carbon‐supported Pt NPs, Pt NCs, and Pt–SnO_x NCs. Insets in (c) show the EDS elemental maps of Pt–SnO_x NCs. (d, e) XRD patterns (d) and XPS spectra (e) of carbon‐supported Pt NPs, Pt NCs, and Pt–SnO_x NCs.

Figure 1d displays the X‐ray diffraction (XRD) patterns of the three Pt‐based catalysts. The broad diffraction peak observed at approximately 25° is attributed to the carbon support [24]. The diffraction pattern of the Pt NPs exhibits well‐defined reflections that are in agreement with the standard diffraction data for Pt. As the Pt size decreases to the NC region, the characteristic Pt diffraction peaks become attenuated and broadened, reflecting the reduced crystallite size [19]. Notably, no discernible SnO₂ diffraction peaks are observed for the Pt–SnO_x NCs, suggesting that the SnO_x species are predominantly amorphous and highly dispersed at the Pt interface.

Figure 1e presents the Pt 4f core‐level X‐ray photoelectron spectroscopy (XPS) spectra, providing insights into the Pt surface chemical states. The deconvoluted spectrum of the Pt NPs reveals a metallic Pt⁰ to oxidized Pt²⁺/Pt⁴⁺ species ratio of approximately 70/30, which corresponds to Pt oxidation. Upon decreasing the Pt particle size to form clusters, the proportion of metallic to oxidized Pt species shows a little bit lower (Pt⁰/Ptⁿ⁺ = 67/33), indicating a tendency toward Pt cluster oxidation. Intriguingly, the introduction of SnO_x species does not induce further shift in the Pt binding energies, instead, it restores the Pt valence state to a Pt⁰/Ptⁿ⁺ ratio of 69/31. This observation suggests that, within these ultrafine heterostructures, there is negligible electronic charge transfer between the Pt and the SnO_x clusters, indicating a rather weak electronic interaction at the interface. The chemical identity of SnO_x was further verified by Sn 3d and O 1s XPS analysis, as shown in Figure S3. The Sn 3d spectrum exhibits characteristic Sn⁴⁺ signals, while the O 1s spectrum contains a lattice oxygen component associated with Sn–O bonding. These results confirm that Sn exists predominantly in an oxidized SnO_x state rather than as metallic Sn. Therefore, the absence of SnO₂ diffraction peaks in XRD should be interpreted as evidence for the lack of long‐range crystalline SnO₂ domains, while the XPS results directly confirm the presence of oxidized Sn–O species. Overall, the combined XRD and XPS analyses support that SnO_x in Pt–SnO_x NCs is highly dispersed and/or amorphous at the Pt interface.

Figure 2a compares the X‐ray absorption near‐edge structure (XANES) spectra of the prepared catalysts (Pt NCs, Pt–SnO_x NCs, and Pt NPs) and the control sample (PtO₂). In an XANES spectrum, the intensity (H _A) and width (W _A) of the absorption peak correspond to the extent of electron transition from 2p_3/2 to 5d_5/2 orbitals (i.e., the amount of empty states in 5d_5/2 orbital) and the width of the 5d band, respectively. Accordingly, the oxidation state of Pt atoms increases with H _A, following the trend of PtO₂>> Pt NCs > Pt NPs ∼ Pt–SnO_x NCs (inset in Figure 2a), consistent with the XPS results shown in Figure 1e. Moreover, the broadened W _A indicates an expansion of d‐band due to significant hybridization between Pt and O atoms. The position of the inflection point (denoted by arrow X in Figure S4) suggests the threshold energy for the 2p to 5d transition, which is directly proportional to the extent of oxidation in Pt atoms.

FIGURE 2.

XAS spectra of carbon supported Pt NCs, Pt–SnO_x NCs, Pt NPs and PtO₂ reference at Pt L₃ edge. (a) XANES, (b) Fourier‐transformed EXAFS regions, and (c) wavelet transformation plots for the k³‐weighted EXAFS signals.

Figure 2b compares the corresponding Fourier‐transformed extended X‐ray absorption fine structure spectra (FT‐EXAFS) of these Pt‐based samples. The radial peaks between 1.2 Å to 2.1 Å (peak A) and 2.1 Å to 3.2 Å (peak B) are the typical features of X‐ray interference with the Pt–O and Pt–Pt bond pairs in Pt oxide and metallic Pt crystal, respectively [25]. The structural parameters of the samples are summarized in Table 1. For Pt NCs and Pt–SnO_x NCs, compared to Pt NPs, the intensity of peak B is substantially decreased by 35%, indicating the formation of downsized Pt NCs. Notably, the slightly suppressed intensity of peak B in Pt–SnO_x NCs compared to Pt NCs denotes the increased local disorder around Pt atoms, attributing to the decorated SnO_x. Overall, the atomic environments of Pt NCs and Pt–SnO_x NCs are quite similar, suggesting the SnO_x modification does not introduce strong interfacial interaction.

TABLE 1.

XAS model analysis determined structural parameters of the Pt NCs, Pt–SnO_x NCs, Pt NPs catalysts, and control sample (PtO₂ NPs) at the Pt L₃ edge.

	Neighboring atom CN a				R (Å) b
	Pt	Sn	O	CNtotal c	Pt	Sn	O
Pt NCs	4.02	—	0.80	4.82	2.75	0	1.99
Pt–SnOx NCs	3.91	0.3	0.45	4.66	2.75	2.79	1.98
Pt NPs	8.22	—	0.55	8.77	2.75	0	2.03
PtO2	0	—	4.72	4.72	0	0	2.03

^a CN: coordination number.

^b R: atomic distance.

^c CN_total: total CN around Pt atom; σ² is 0.004, ΔE ₀ is 7–10, R factor is 0.003–0.008 for all fitting models.

Table 1 summarizes the model analysis of the local atomic structures of these Pt‐based samples. The obtained quantitative structural parameters confirm the dimensions of the Pt NCs and the effects of SnO_x decoration. The coordination number (CN) of the Pt‐Pt bond in Pt NCs is 4.02, corresponding to a highly undercoordinated Pt environment that is consistent with ultrasmall Pt NCs of approximately 1.5 nm [26], in good agreement with the STEM results. In the case of Pt–SnO_x NCs, the CN for the Pt–Pt bond is slightly reduced compared to Pt NCs, indicating local disordering around the Pt atoms due to the SnO_x modification. Furthermore, the wavelet transform (WT) of EXAFS was applied to reveal the atomic dispersion in the samples. In the WT transformation plots, bright colors represent regions with a maximum contribution from scattering atoms located at a distance R from the absorber, characterized by a specific value of the photoelectron wavenumber k [27]. As shown in Figure 2c, for Pt NCs, Pt–SnO_x NCs, and Pt NPs, the intense peak detected at R ≈2.55 Å and k ≈11 Å corresponds to the Pt–Pt scattering path. The WT‐EXAFS of Pt–SnO_x NCs is similar to those of Pt NPs and Pt NCs, implying similar Pt local structures in these samples and confirming the weak Pt–SnO_x interaction.

To further verify this weak interfacial interaction, STEM‐based electron energy loss spectroscopy (EELS) was conducted to probe the local electronic structure of Pt–SnO_x NCs with high spatial resolution. As shown in Figure S5, two representative regions were selected for EELS analysis: one located away from the Pt/SnO_x interface and the other near the Pt/SnO_x interface. The corresponding Pt N‐edge spectra exhibit comparable peak positions and spectral profiles at these two locations, suggesting that the electronic structure of Pt is not significantly altered by its distance from the SnO_x species. This spatially resolved evidence suggests the absence of pronounced charge redistribution at the Pt/SnO_x interface, further supporting the relatively weak electronic interaction between Pt and SnO_x.

Figure S6 presents the Cyclic Voltammetry (CV) curves of the catalysts obtained in N₂‐saturated 0.5 m H₂SO₄ solution. These curves display the oxidation and reduction processes of oxides and hydroxides on the catalyst surfaces. During the anodic scan, the peak centered at 0.7 V corresponds to the oxidation of Sn oxides [28, 29]. In the cathodic scan, three characteristic peaks are observed at around 0.5, 0.6, and 0.75 V, attributed to the reduction of Pt–Sn oxides, SnO₂, and Pt oxide, respectively [30]. Compared to the control samples shown in Figure S6b, the anodic and cathodic peaks of the NCs shift negatively, indicating that downsizing Pt/Sn alters their redox behavior. Specifically, Sn in Pt–SnO_x NCs is more easily oxidized, while Pt oxide in Pt–SnO_x NCs is more resistant to reduction than in PtSn NPs.

To evaluate HER activity, Linear sweep voltammetry (LSV) scans were recorded at 2 mV s⁻¹ in N₂‐purged 0.5 m H₂SO₄. Figure 3a illustrates the polarization curves of catalysts. Generally, nanocatalysts such as SAs and NCs, with smaller particle sizes, larger surface areas, and more Pt active sites, exhibit superior activities due to their ability to adsorb H and convert H_ad to H₂ molecules [31, 32]. The HER performance measured by the overpotential at 10 mA cm⁻² (η ₁₀) of Pt NCs surpasses that of Pt NPs due to their uniform dispersion and high Pt utilization. Notably, incorporating SnO_x into Pt NCs significantly boosts HER performance compared to pristine Pt NCs. Pt–SnO_x NCs exhibit the lowest overpotential and highest current density across the entire potential range, with an overpotential of 22 mV at 10 mA cm⁻², as shown in Table S2. The excellent HER performance of Pt–SnO_x NCs can be attributed to their cluster morphology and the interfacial incorporation of SnO_x, which may favor a Volmer–Tafel pathway and accelerate the HER catalytic process.

FIGURE 3.

HER performance of Pt NPs, Pt NCs, and Pt–SnO_x NCs. (a) Polarization curves. (b) Tafel plots. (c) Comparison of overpotentials. (d) Mass activities. (e) LSV curves before and after stability tests. (f) Schematic illustration of the PEM reactor. (g) Stability tests of the ethanol‐assisted PEM paired‐electrolyzers at 1 A cm⁻².

Further Tafel analysis, presented in Figure 3b, provides insight into the reaction kinetics of these Pt‐based catalysts. All Tafel slopes were derived from the iR‐corrected polarization curves (90%) within the low‐overpotential kinetic region to minimize mass‐transport interference. Both Pt NCs and Pt NPs exhibit Tafel slopes close to 35 mV dec⁻¹, indicating a Volmer–Heyrovsky reaction mechanism [33]. In contrast, Pt–SnO_x NCs show a Tafel slope of about 24 mV dec⁻¹, obtained from the linear fitting range of approximately 4–14 mA cm⁻² with a high fitting reliability. This value is close to the typical slope associated with a Tafel recombination‐favorable HER process [34], implying that SnO_x decoration facilitates interfacial hydrogen transfer/recombination and modifies the HER kinetics. To further examine whether hydrogen‐transfer behavior is involved in the enhanced HER activity, acidic HER measurements were conducted in a D₂O‐containing electrolyte. Both Pt NCs and Pt–SnO_x NCs exhibited decreased HER activity in D₂O, which can be attributed to the kinetic isotope effect associated with the slower D⁺/D₂O transfer kinetics compared with H⁺/H₂O. Notably, Pt–SnO_x NCs showed a more pronounced activity loss than Pt NCs, as shown in Figure S7. The stronger isotope‐dependent degradation of Pt–SnO_x NCs suggests that its HER kinetics are more sensitive to hydrogen/deuterium transfer processes. This result provides additional experimental evidence consistent with the involvement of interfacial H‐transfer or hydrogen‐spillover behavior at the Pt–SnO_x interface.

The advantage of NCs is highlighted in the mass activity (MA) comparison in Figure 3d, where downsizing Pt atoms improves the utilization efficiency, and SnO_x decoration on Pt NCs further enhances the MA, although Pt NCs and Pt–SnO_x NCs possess comparable particle sizes. Replicate measurements were also performed, and the standard deviations of key activity metrics, including overpotential and MA, are provided in Figure 3c,d, respectively. To clarify the origin of this enhancement, the electrochemically active surface areas (ECSAs) of Pt NCs and Pt–SnO_x NCs were estimated from Figure S6 and compared in Table S3. The ECSA of Pt–SnO_x NCs is slightly higher than that of Pt NCs, increasing from 95 to 110 m² g_Pt ^{− 1}, indicating a modest increase in accessible Pt surface area after SnO_x modification. Nevertheless, the ECSA‐normalized specific activity (SA) increases from 7.8 to 18.1 A m⁻², demonstrating that SnO_x modification improves the intrinsic activity of Pt sites. More importantly, the turnover frequency (TOF) at an overpotential of 20 mV increases markedly from 0.76 s⁻¹ for Pt NCs to 2.03 s⁻¹ for Pt–SnOx NCs, suggesting that the enhanced HER activity is not solely governed by the number of exposed Pt sites but is mainly associated with improved site‐level catalytic kinetics. To evaluate whether the introduction of SnO_x causes an unfavorable conductivity penalty, electrochemical impedance spectroscopy (EIS) was conducted for Pt NCs and Pt–SnO_x NCs. As shown in Figure S8, Pt–SnO_x NCs exhibit a smaller semicircle in the Nyquist plot than Pt NCs, indicating a lower charge‐transfer resistance and faster interfacial electron‐transfer kinetics (Table S3). These results indicate that the highly dispersed SnO_x species do not severely impede charge transport in the present catalyst configuration. Instead, the Pt–SnO_x interface improves the electrochemical kinetics at the catalyst/electrolyte interface, thereby supporting the enhanced HER activity of Pt–SnO_x NCs.

Stability evaluations, shown in Figure 3c–e, reveal that after chronoamperometry (CA) testing, the overpotential and Tafel slope of Pt NPs increase dramatically (Table S2) while their mass activity decreases, implying the inevitable deactivation of NPs during durability tests in an acidic solution. In contrast, Pt–SnO_x NCs, display more excellent and stable performance after 6 h, with an overpotential of 24 mV at 10 mA cm⁻², confirming that the usage of Pt can be significantly enhanced compared to Pt NPs. The SnO_x modification effectively anchors the Pt NCs, contributing to their superior stability and performance.

To evaluate the potential of Pt–SnO_x NCs as a cathode material under industrially relevant conditions, we assembled a PEM electrolyzer. A commercial IrO₂ catalyst served as the anode. The PEM electrolyzer featured a sandwich configuration with an active area of 4 cm², comprising a porous transport layer (PTL), cathode (anode) catalyst layer, and PEM, as illustrated in Figure 3f. First, DI water without ethanol (conventional PEM electrolyzer) was used as the feed electrolyte. As shown in Figure S9, the IrO₂||Pt–SnO_x NCs cell maintained stable operation at 1 A cm⁻² and 80°C for 20 h, with only a slight increase in cell voltage. Compared with the IrO₂||Pt/C cell tested under the same conditions, the Pt–SnO_x NCs‐based cell exhibited a comparable voltage despite the much lower Pt loading of the cathode, approximately 0.05 mg_Pt cm⁻². This result indicates that Pt–SnO_x NCs possess good PEM electrolyzer durability and improved Pt utilization under conventional OER conditions. In addition to the standard (oxygen evolution reaction)OER‐based PEM electrolyzer test, ethanol was introduced into the anode electrolyte to examine the feasibility of EOR‐assisted hydrogen production (ethanol‐assisted PEM paired‐electrolyzer). Because EOR is kinetically more favorable than OER, the partial replacement of OER by EOR can effectively lower the overall cell voltage. Under this configuration, the Pt–SnO_x NCs‐based ethanol‐assisted PEM paired‐electrolyzer delivered a current density of 1 A cm⁻² at < 1.5 V (80°C) for over 70 h, as shown in Figure 3g. During EOR‐assisted PEM electrolysis, the H₂ Faradaic efficiency remained close to 98%, confirming efficient cathodic hydrogen production under this paired‐electrolysis configuration. Analysis of the anodic electrolyte (Figure S10) showed that acetic acid was the dominant product during EOR. These results further confirm that the reduced cell voltage originates from the coupling of cathodic HER with anodic EOR rather than conventional OER alone.

To further evaluate the durability of Pt–SnO_x NCs after long‐term operation, post‐stability HAADF‐STEM/EDS and XPS analyses were performed after the 70 h PEM electrolyzer test. As shown in Figure S11, the Pt–SnO_x NCs retain their clustered morphology with distinguishable Pt and amorphous SnO_x regions, as supported by the corresponding EDS elemental maps and line profiles. The Pt clusters remain below 5 nm without obvious aggregation, indicating that SnO_x modification effectively stabilizes the Pt nanoclusters during HER operation. Post‐stability XPS (Figure S12) further reveals weakened Pt and Sn signals, likely due to surface interference from membrane/ionomer‐derived species. Nevertheless, the Pt 4f spectrum shows that only approximately 20% of Pt species are oxidized, while Sn remains mainly as SnO₂. The O 1s spectrum shows a decrease in adsorbed H₂O‐related species and an increase in oxygen‐vacancy‐related components, suggesting surface reconstruction during electrolysis. These results confirm that the Pt–SnO_x interfacial structure is largely retained after long‐term ethanol‐assisted PEM paired‐electrolyzer operation.

To further benchmark the performance of Pt–SnO_x NCs, recently reported low‐Pt‐loading HER catalysts, including Pt single‐atoms and Pt‐based nanoclusters, were summarized in Table S4. Key parameters such as Pt loading, overpotential, Tafel slope, stability, and testing configuration were compared to clarify the positioning of this work. As evidenced by the literature comparison summarized in Table S4, the present PEM electrolyzer test result represents a practically relevant and competitive evaluation condition for low‐Pt‐loading HER catalysts. These results highlight the high Pt utilization efficiency and industrially relevant hydrogen‐production potential of Pt–SnO_x NCs, demonstrating their promise as efficient low‐Pt cathodes for both conventional PEM water electrolysis and paired electrolysis systems with reduced anodic energy demand.

To verify whether the superior HER performance of Pt–SnO_x NCs is associated with the weak interfacial interaction between Pt and amorphous SnO_x, an additional Pt–Sn catalyst with stronger metal–oxide interaction, denoted as Pt–SnO₂, was intentionally prepared as a structural control. It should be noted that Pt–SnO₂ was not designed as a loading‐matched counterpart to Pt–SnO_x NCs, but rather as a contrasting sample to establish a strongly coupled Pt/SnO₂ interface. In the Pt–SnO_x NCs, the weak interaction originates from the low Pt loading and the amorphous nature of SnO_x, which prevent the formation of crystalline Pt/SnO₂ heterojunctions and help preserve the intrinsic electronic structure of Pt.

For the preparation of Pt–SnO₂, the Pt loading was increased to 6 wt% (Pt/Sn = 3/2), followed by an additional air treatment at 470 K for 1 h after deposition. These conditions promote the crystallization of SnO_x into SnO₂ and the formation of well‐defined crystalline Pt NPs. As a result, Pt and SnO₂ no longer exist as disordered cluster‐like species, but instead form crystalline Pt/SnO₂ heterogeneous junctions with stronger interfacial coupling. Therefore, the comparison between Pt–SnO_x NCs and Pt–SnO₂ is used to highlight the structural consequence of weakly coupled amorphous Pt–SnO_x interfaces versus strongly coupled crystalline Pt/SnO₂ heterojunctions, rather than to evaluate the effect of Pt loading alone. Atomic‐resolution HAADF‐STEM images (Figure S13) clearly reveal the lattice fringes of both Pt and SnO₂, confirming the formation of intimately coupled hetero‐NPs. Consistently, XRD patterns (Figure S14a) exhibit a pronounced SnO₂ diffraction peak for Pt–SnO₂ compared to Pt–SnO_x NCs, indicating the presence of crystalline SnO₂ and Pt domains. XPS analysis (Figure S14b) reveals stronger electronic coupling in Pt–SnO₂, as evidenced by a more noticeable Pt 4f binding energy shift (0.2 eV) compared with Pt NCs and Pt–SnO_x. XAS measurements (Figure S14c–e and Table S5) provide additional support: EXAFS fitting reveals an increased Pt–Sn coordination number (CN_Pt–Sn ≈ 1.1), while WT‐EXAFS spectra show additional Pt–Sn scattering features, collectively indicating stronger Pt–SnO₂ interactions. Such strong metal–oxide interactions inevitably perturb the intrinsic electronic structure of Pt and modify the surface properties of Pt sites, thereby influencing the adsorption energetics of hydrogen intermediates during the HER process.

Corresponding electrochemical measurements (Figure S15) show that Pt–SnO₂ exhibits an overpotential of 48 mV at 10 mA cm⁻² and a Tafel slope of 35 mV dec⁻¹, indicating that the intrinsic HER kinetics of Pt remain largely unchanged despite the strong interaction. However, the strong coupling partially blocks the Pt active sites, limiting their accessibility and reducing catalytic efficiency, as reflected by the relatively low MA (≈15 A g_Pt ⁻¹) (Table S6). In contrast, Pt–SnO_x preserves accessible Pt surfaces while introducing SnO_x species that improve Pt efficiency without significantly perturbing its intrinsic electronic structure. This balanced configuration ultimately leads to superior HER performance and a markedly enhanced MA (≈340 A g_Pt ⁻¹).

Then, the detailed mechanistic insights were analyzed using density functional theory (DFT) by considering three elementary processes: the electrochemical Volmer adsorption step ((H⁺ + e⁻) + * → H*) followed by either the electrochemical Heyrovsky desorption step ((H⁺ + e⁻) + H* → H₂(g) + *) or the thermochemical Tafel recombination step (2H* → H₂(g) + 2*). Front and top views of the catalyst surfaces for Pt NCs and Pt–SnO_x NCs are shown in Figure S16. As depicted in Figure 4a, the HER was first studied over Pt NCs. The Volmer step was performed over Pt NCs with the activation barriers of 0.37 and 0.25 eV at 0 V and ‐0.16 V vs RHE, respectively, forming H* on Pt sites. The subsequent Heyrovsky step becomes the rate‐determining step, with a higher energy barrier (≈0.55 and 0.40 eV, respectively, at 0 and ‐0.16 V), indicating that hydrogen desorption is kinetically limiting. The alternative Tafel pathway shows an even higher barrier (≈0.70 eV), less favorable on pure Pt NCs. Overall, these energetics indicate that HER on Pt NCs is controlled by electrochemical hydrogen desorption rather than H–H recombination.

FIGURE 4.

Free energy diagrams display reaction mechanisms for HER over (a) Pt NCs, and (b) Pt–SnO_x NCs. The black and blue lines show the energy change at 0 and ‐0.16 V vs RHE, respectively. The solid lines show the more favored reaction pathways in the relevant reaction conditions. The bending curves show the activation barriers for Volmer (V), Heyrovsky (H), and Tafel (T) steps, respectively. The insets show the H* adsorption models used for the DFT calculations, with all elements clearly labeled, where gray, blue, red, and white spheres represent C, Pt, Sn, and H atoms, respectively.

In contrast, the Pt–SnO_x structure introduces a distinct reaction pathway. The Volmer step occurs on Pt sites (Pt–Pt) with a similar energy barrier (≈0.33 eV), after which the adsorbed hydrogen can migrate from Pt to the Pt–Sn interfacial sites, forming H* (Pt–Sn). This site‐transfer step (≈0.32 eV) represents a key intermediate process enabled by the metal–oxide interface [35]. After hydrogen migration, the recombination of two hydrogen species proceeds through a Tafel pathway with a significantly reduced barrier (≈0.18 eV), much lower than that observed on Pt NCs.

The theoretical analysis was performed to describe the representative Pt–SnO_x NC interface and to clarify the kinetic role of SnO_x in the spillover‐assisted Volmer–Tafel pathway, rather than to suggest that SnO_x enhances HER activity simply by optimizing the intrinsic H adsorption strength. As summarized in Table S7, the Pt site exhibits a ΔG _H* value of −0.25 eV, indicating favorable H* adsorption, whereas the Sn site shows a much weaker H* binding with a ΔG _H* value of 0.07 eV. This comparison suggests that Pt serves as the primary site for H* formation, while Sn/SnO_x does not preferentially stabilize H* through direct adsorption. Moreover, the comparable H* adsorption behavior on Pt sites before and after SnO_x incorporation indicates that SnO_x does not substantially alter the intrinsic H* binding strength of Pt. Instead, SnO_x mainly functions as an adjacent interfacial mediator for H* transfer and recombination. The direct formation of H* on Sn through the Volmer step is energetically unfavorable due to its high reaction barrier. Therefore, a kinetically accessible site‐transfer pathway from Pt to Sn is proposed for the adsorbed H*, effectively bypassing the direct Volmer step on Sn sites (Figure 4b). These results indicate that the Pt–SnO_x interface enables a hydrogen spillover mechanism [36], where hydrogen generated on Pt migrates to adjacent interfacial sites prior to recombination. This interfacial pathway is suggested to lower the energy barrier for H₂ formation and facilitates the Tafel recombination step, and may shift the dominant HER mechanism from Volmer–Heyrovsky (Pt NCs) to Volmer–Tafel (Pt–SnO_x NCs). Consequently, the cooperative interaction between Pt and SnO_x promotes hydrogen formation on Pt, interfacial H* transfer, and subsequent recombination, leading to enhanced HER activity. Notably, the theoretically predicted activity trend shows excellent agreement with the experimental observations.

3. Conclusions

To summarize, we have developed a Pt–SnO_x NC catalyst that maximizes Pt utilization while preserving its intrinsic catalytic activity via controlled SnO_x decoration. This architecture establishes a weak Pt/SnO_x interfacial interaction, which minimizes the electronic perturbation of Pt sites while introducing additional functionality of SnO_x. The Pt–SnO_x NCs exhibit superior HER performance relative to pristine Pt NCs, Pt NPs, and strongly coupled Pt–SnO₂ catalysts, delivering an overpotential of 22 mV at 10 mA cm⁻² and a Tafel slope of 24 mV dec⁻¹. DFT calculations suggest that Pt–SnO_x structure may facilitate a hydrogen spillover mechanism, which is associated with a reduced energy barrier for hydrogen formation and may promote a shift in the dominant HER pathway from the Volmer–Heyrovsky route to the more efficient Volmer–Tafel route. When employed as a cathode in an ethanol‐assisted PEM paired‐electrolyzer, the Pt–SnO_x catalyst achieves a current density of 1 A cm⁻² at a cell voltage below 1.5 V at 80°C and maintains stable operation for over 70 h. This work demonstrates that rational design of weak interfacial interactions represents a key strategy for boosting the HER performance of low‐Pt‐loading catalysts. Such a design allows the incorporation of a functional secondary species without disrupting Pt's intrinsically favorable hydrogen adsorption free energy.

Author Contributions

S.D. and K.‐W.W. conceived and supervised the research project. Z.J., T.‐H.H., F.‐Y.C., and T.‐Y. C. fabricated the catalysts and performed the most experiment. J.‐H.W., S.D., and K.‐W.W. performed the characterization experiment and analysis the data. J.‐H.W., S.‐T.C., and T.‐Y.C. performed the theoretical calculation. S.D. and K.‐W.W. write and edit the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: smll74788‐sup‐0001‐SuppMat.docx.

Data Availability Statement

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