Screening the Oxygen Reduction Reaction Performance of Carbon-Supported Pt-M (M = Ni, Cu, Co) Binary Electrocatalysts via Tuning Metal–Support Interaction

Abstract

Platinum-based catalysts remain the benchmark for the oxygen reduction reaction (ORR) in fuel cells, owing to their exceptional catalytic activity in the harsh chemical environment. However, optimizing Pt utilization and improving performance through support engineering are essential for commercial viability. In this study, we synthesized carbon-supported binary Pt-M (M = Ni, Cu, Co) electrocatalysts to investigate the influence of metal–support interactions on ORR activity. The Pt-M nanoparticles were fabricated on carbon supports, enabling the systematic screening of electronic and structural interactions. Among all compositions, Pt@Co exhibited the highest ORR mass activity, delivering 817 mA mg_Pt⁻¹ at 0.85 V and 464 mA mg_Pt⁻¹ at 0.90 V vs. RHE, surpassing both commercial Pt/C (J.M. 20 wt.%) and its Pt@Ni, Pt@Cu, and Pt@CNT counterparts. Structural and spectroscopic analyses reveal a strong electronic interaction between Pt and Co, leading to localized electron transfer from Co to Pt domains. This electronic modulation facilitates an optimal surface binding energy, enhancing oxygen adsorption–desorption kinetics and ORR activity. These findings highlight the critical role of transition metal–support synergy in the rational design of high-performance Pt-based electrocatalysts for next-generation fuel cell applications.

1. Introduction

Fuel cells are among the most promising clean energy technologies due to their high energy efficiency, low operational temperatures, and near-zero emissions [1]. Fuel cells convert chemical energy directly into electricity through redox reactions involving hydrogen and oxygen. However, the widespread deployment of fuel cells is hampered by the sluggish kinetics of the cathodic oxygen reduction reaction (ORR), which significantly lowers the overall efficiency of the fuel cell [2,3]. To address this kinetic bottleneck, the development of high-performance electrocatalysts is imperative, with platinum (Pt) remaining the benchmark catalyst due to its unparalleled ORR activity and durability towards ORR [4,5]. Nevertheless, the high cost, limited abundance, and susceptibility to degradation of Pt-based catalysts necessitate strategies that maximize their catalytic efficiency while minimizing the Pt content [6].

A widely adopted strategy to enhance Pt utilization and ORR activity is the engineering of metal–support interactions (MSIs) in binary or ternary catalyst systems [7,8]. This approach not only tunes the electronic structure of Pt but also modifies its geometric configuration and surface adsorption properties [9]. Carbon-based materials, such as carbon nanotubes (CNTs), are often employed as supports due to their high electrical conductivity, large surface area, and stability in electrochemical environments [10]. In addition, the incorporation of transition metals (TMs) such as Ni, Cu, and Co into Pt-based systems has shown promise in modulating the d-band center of Pt, thereby enhancing ORR kinetics and catalyst durability [11,12,13]. Notably, these transition metals can serve as either alloying components or oxide domains to interact electronically and structurally with Pt nanoparticles, thereby optimizing the catalytic interface [14,15]. Prior studies have demonstrated that the formation of Pt-M (M = Ni, Cu, Co) alloys can improve catalytic activity via electronic synergism [16]. For example, it has been reported that Pt-Ni octahedral nanoparticles exhibited superior ORR activity owing to lattice strain and electron redistribution effects [17]. Similarly, Co has been shown to induce favorable Pt electronic states, leading to enhanced O₂ adsorption and dissociation kinetics [18]. The significance of MSIs becomes particularly evident when considering the electron transfer dynamics and active site availability on Pt surfaces. Strong interactions between Pt and TM or carbon supports can induce charge redistribution, which influences the binding energy of ORR intermediates such as O*, OH*, and OOH* [19,20]. This modulation plays a crucial role in accelerating reaction kinetics and reducing overpotential. Such interactions not only boost activity but also improve long-term stability by preventing Pt agglomeration and dissolution. Furthermore, the support architecture and chemical environment can stabilize the optimal oxidation state of Pt and facilitate rapid mass and charge transport.

In this context, we investigate a series of binary Pt-M/C catalysts with a fixed Pt loading (10 wt.%) on distinct supports, including carbon nanotubes (CNTs) and transition metals (Ni, Cu, and Co), to systematically assess the effect of MSIs, composition, and configuration on ORR performance. The synthesized catalysts Pt@Co, Pt@Cu, Pt@Ni, and Pt@CNT are characterized structurally and electrochemically to identify the key parameters influencing ORR activity. Among the studied systems, Pt@Co exhibits the highest mass activity, achieving 817 mA mgPt⁻¹ at 0.85 V and 464 mA mgPt⁻¹ at 0.90 V vs. RHE, outperforming commercial J.M. Pt/C and the other prepared catalysts. Physical characterizations using TEM, XAS, and XPS reveal that Pt@Co possesses a unique interfacial environment wherein strong electronic coupling between Pt and Co results in significant electron transfer from Co to Pt domains. This study provides valuable insight into the rational design of Pt-based catalysts through the engineering of metal–support interactions. By systematically comparing Pt-M systems with controlled compositions and support properties, we establish a clear correlation between the nature of the support, the degree of electron transfer, and the resulting catalytic activity. The high performance of Pt@Co, in particular, illustrates the potential of using earth-abundant transition metals to amplify the intrinsic activity of Pt and reduce overall precious metal consumption. These results pave the way for future efforts in catalyst optimization through support engineering and electronic structure modulation, ultimately contributing to the development of more cost-effective and efficient electrocatalysts for sustainable energy applications.

2. Experimental Section

2.1. Preparation of Pt-M/C Binary Electrocatalysts

Pt-M/C binary electrocatalysts were prepared using a wet chemical reduction method, following a carefully controlled sequence of metal ion adsorption and reduction steps (Scheme 1). Carbon nanotube (CNT) support (Sino Applied Technology, Taoyuan, Taiwan) was first activated using potassium hydroxide (KOH) to increase surface defects and introduce surface functional groups that facilitate better metal ion attachment. For this activation process, CNT powder was mixed with KOH in a metal crucible and heated in a nitrogen (N₂) environment first at 350 °C for 30 min, then at 800 °C for 2 h. After activation, the CNTs were washed thoroughly with hydrochloric acid (HCl) and deionized (DI) water until the wash water reached a neutral pH (around 7.0).

To synthesize Pt@CNT, 1.2 g of 5 wt.% CNT solution in DI water (equivalent to 60 mg of dry CNT) was mixed with 1.02 g of 0.1 M hydrogen hexachloroplatinate (IV) hexahydrate (H₂PtCl₆·6H₂O, Sigma-Aldrich, Burlington, MA, USA). This mixture was stirred at 800 rpm for 6 h to allow platinum ions (Pt²⁺) to adsorb onto the CNT surface. The solution contained approximately 0.102 mmol (6 mg) of Pt²⁺, corresponding to a Pt loading of 10 wt.%. In the next step, 5 mL of a freshly prepared aqueous solution containing 0.1 g of sodium borohydride (NaBH₄, Sigma-Aldrich) was added to reduce the Pt²⁺ ions. This reduction was performed under stirring (800 rpm) for 10 s, resulting in the formation of Pt nanoparticles on the CNT surface, i.e., Pt@CNT.

For the Pt-M/C (M = Ni, Cu, Co) binary catalysts, the synthesis started by dispersing 1.2 g of 5 wt.% CNT solution in 3.06 g of 0.1 M aqueous solution of the chosen transition metal (Ni²⁺, Cu²⁺, or Co²⁺). The mixture was stirred for 6 h at 800 rpm, allowing the metal ions to adsorb onto the CNT surface. This step introduced approximately 0.306 mmol (18 mg) of metal ions (M^x+), corresponding to a metal loading of 30 wt.% to CNT. Then, 5 mL of a 0.16 g NaBH₄ solution was added and stirred for 10 s to reduce the metal ions to their respective nanoparticles. Due to their inherent instability, these freshly formed nanoparticles were partially oxidized to metal oxides during the process. In the final step, 1.02 g of 0.1 M hydrogen hexachloroplatinate solution (containing 0.102 mmol of Pt²⁺) was added to the CNT-supported metal oxide solution. The remaining NaBH₄ from the previous step facilitated the in situ reduction of Pt²⁺, depositing Pt nanoparticles onto the metal oxide-modified CNT support. The resulting Pt-M/C electrocatalysts were washed several times with acetone, isopropyl alcohol (IPA), and DI water, followed by centrifugation and drying at 70 °C.

2.2. Structural Characterizations of Binary Electrocatalyst

The structural properties of the Pt-M/C binary catalysts were studied using a combination of microscopy and X-ray-based techniques. High-resolution transmission electron microscopy (HRTEM) was performed at the Electron Microscopy Center, National Tsing Hua University, Taiwan, to observe the particle size, distribution, and morphology of the synthesized catalysts. X-ray absorption spectroscopy (XAS) was conducted at beamlines BL-17C and 01C1 of NSRRC to examine the local atomic environment and electronic structure around the metal atoms. To ensure accuracy, each XAS measurement was repeated at least twice, and the results were averaged for analysis. For extended X-ray absorption fine structure (EXAFS) analysis, the raw spectra were processed by subtracting both pre-edge and post-edge backgrounds and then normalized based on the edge step. The normalized spectra were transformed from energy to k-space, followed by k²-weighting to highlight the contributions of atoms at different distances from the absorbing atom. X-ray photoelectron spectroscopy (XPS) was performed at beamline BL-24A1 of NSRRC to analyze the electron localization between the elements. The binding energies were calibrated using the C 1s peak at 284.6 eV as a reference. X-ray diffraction (XRD) measurements were carried out at beamline BL-01C2 of the National Synchrotron Radiation Research Center (NSRRC), Taiwan, using X-rays with a wavelength of 0.6888 Å (corresponding to 18.0 keV).

2.3. Electrochemical Analysis

All electrochemical tests were performed at room temperature (25 ± 1 °C) using a CHI 600B potentiostat (CH Instruments, Bee Cave, TX, USA) with a standard three-electrode setup. To prepare the catalyst ink for the oxygen reduction reaction (ORR) experiments, 5 mg of the catalyst powder was mixed with 1.0 mL of isopropanol (IPA) and 50 μL of Nafion-117 solution (99%, Sigma-Aldrich) to act as a binder and conductive additive. This mixture was sonicated for 30 min to ensure uniform dispersion before use. For ORR testing, 10 μL of the prepared ink was drop-cast onto a glassy carbon rotating disk electrode (RDE) with an area of 0.196 cm² and left to air-dry, forming the working electrode. A Hg/HgCl₂ electrode immersed in a KCl solution was used as the reference electrode. The reference potential was calibrated to 0.242 V vs. the reversible hydrogen electrode (RHE). A graphite rod served as the counter electrode to eliminate the risk of platinum contamination during testing. Cyclic voltammetry (CV) and linear sweep voltammetry (LSV) were used to evaluate the ORR activity. CV was performed at a scan rate of 0.02 V/s over a potential range of 0.1 to 1.3 V (vs. RHE) under a nitrogen (N₂) atmosphere. LSV measurements were carried out in oxygen (O₂)-saturated 0.1 M KOH solution (pH 13) at a scan rate of 0.001 V/s within the potential window of 0.4 to 1.1 V (vs. RHE). The rotation speed of the RDE varied from 400 to 3600 rpm to study the kinetic behavior of the catalysts under different diffusion conditions.

3. Results and Discussion

3.1. Morphological and Structural Properties

Figure 1 presents HRTEM images of four different electrocatalysts, (a) Pt@CNT, (b) Pt@Ni, (c) Pt@Cu, and (d) Pt@Co, highlighting the structural evolution of Pt nanoparticles when interfaced with various transition metal oxides supported on carbon nanotubes (CNTs). As shown in Figure 1a, for Pt@CNT, highly crystalline Pt nanoparticles are grown on the CNT support with clear lattice fringes and a measured interplanar spacing of ~0.24 nm. The slightly enhanced d-spacing as compared to the standard Pt (111) plane can be attributed to the strain relaxation due to the absence of surface defects [21]. The hexagonal FFT pattern further confirms the highly crystalline nature of Pt, while the inverse Fourier transform (IFT) image and line profile further validate the lattice periodicity. In Figure 1b, the Pt@Ni displays a slightly reduced d-spacing of ~0.208 nm, indicating the formation of a Ni_xPt alloy [22], supported by the contrast difference in the HRTEM image and diminished FFT spots (denoted by yellow circles). Such a scenario can be attributed to the galvanic replacement between Ni⁰ ←→ Pt²⁺ due to the electronegative difference [23]. Meanwhile, the existence of clear lattice fringes (denoted by yellow rectangle) and the co-existence of a d-spacing of 0.201 nm indicate the growth of small atomic clusters on the surface region with strong compressive lattice strain due to the lattice mismatch between Ni and Pt [24,25]. Figure 1c shows the HRTEM image of Pt@Cu, where the Pt lattice retains an expanded spacing of ~0.25 nm, suggesting surface oxidation (due to the adjacent copper oxide domains) induced expansive surface strain. Such results can be confirmed by the fuzzy surface (denoted by the yellow arrows), which is a typical feature of surface oxidation [26]. The FFT and IFT patterns show less symmetry compared to Pt@CNT, reflecting structural heterogeneity. In Figure 1d, Pt@Co reveals the presence of distinct lattice fringes oriented in different directions, suggesting the growth of polycrystalline Pt nanoparticles, which is consistently confirmed by the distorted hexagonal FFT pattern. In the absence of surface oxidation or alloy formation along with the presence of lattice dislocations (denoted by white arrows), the reduced d-spacing of ~0.202 is indicative of the strong compressive lattice strain.

Figure 2a–c show the X-ray absorption near-edge structure (XANES), first-derivative spectra, and the corresponding extended X-ray absorption fine structure (EXAFS) spectra of binary electrocatalysts, respectively. The XANES spectra at the Pt-L₃ edge are compared in Figure 2a. Accordingly, with an identical inflection point X (peak X in the first-derivative curve of Figure 2b) for all samples, the Pt atoms preserve a similar oxidation (valence) state in all the samples. Among all samples, the highest white line intensity (H_A) confirms the highest density of occupied Pt-5d_5/2 orbitals (i.e., highest extent of charge transfer to Pt atoms) in the Pt@Co sample [27]. An even closer inspection of white line intensities reveals that Pt-M/C catalysts exhibit suppressed H_A as compared to Pt@CNT, suggesting the highest density of occupied Pt-5d_5/2 orbitals in Pt-M/C catalysts. Such a scenario confirms the strong metal–support interaction between transition metal oxides and Pt nanoparticles as compared to carbon support. Figure 2c compares the corresponding Fourier transform EXAFS (FT-EXAFS) spectra. In FT-EXAFS spectra, the position and intensity of the radial peaks indicate the distances and number of backscattering (neighboring atoms) around the target atoms, respectively. In the case of Pt@CNT, the radial peak B results from the X-ray interference with the Pt-Pt bond pair at 2.751 Å. In the case of the M@Pt catalysts, the radial peak of the Pt-Pt bond pair splits into two peaks (peaks C and D), which comprise contributions from the backscattering of the heteroatomic bond pairs of Pt-M and Pt-Pt [28]. In general, the suppression of the radial peak intensity corresponds to a decrease in the total coordination number around the target atoms. Consistently, the Pt-M/C catalysts exhibit suppressed radial peaks as compared to Pt@CNT. Moreover, the left shift of peak C (corresponding to the Pt-M bond pairs) for Pt@Ni and Pt@Co suggests a reduced bond length, which is in good agreement with the former HRTEM results. The aforementioned results are in good agreement with the XAS-determined structural parameters (

Table 1. Pt L₃ edge XAS determined structural parameters of experimental samples.

Sample	Pt L3-Edge
	Bond Pair	CN	R
Pt@CNT	Pt-Pt	6.95	2.783
	Pt-O	1.46	2.781
Pt@Ni	Pt-Pt	1.75	2.712
	Pt-O	2.46	2.765
	Pt-Ni	2.35	2.523
Pt@Cu	Pt-Pt	3.76	2.734
	Pt-O	2.79	2.762
	Pt-Cu	1.87	2.536
Pt@Co	Pt-Pt	4.02	2.722
	Pt-O	2.44	2.768
	Pt-Co	1.12	2.522

CN represents the coordination number, and R corresponds to the radial distance.

).

Figure 3 shows the high-resolution X-ray photoelectron spectroscopy (XPS) spectra of Pt@CNT, Pt@Ni, Pt@Cu, and Pt@Co electrocatalysts at Pt-4f orbitals. Each spectrum was deconvoluted into three chemical states of platinum: metallic Pt⁰, Pt²⁺, and Pt⁴⁺. The binding energy peaks corresponding to Pt⁰ appear around 71.2 eV (Pt 4f₇/₂) and 74.5 eV (Pt 4f₅/₂), as indicated by the vertical dashed lines. For Pt@CNT, the dominant peaks are attributed to metallic Pt⁰, indicating that most of the Pt remains in its unoxidized state when supported on carbon nanotubes. In contrast, the Pt@Ni, Pt@Cu, and Pt@Co samples exhibit a relative decrease in Pt⁰ peak intensity and an increase in the contributions from Pt²⁺ and Pt⁴⁺ species. Such a scenario is obvious due to adjacent oxide domains. Moreover, the Pt-M/C catalysts exhibit a systematic shift toward lower binding energies compared to Pt@CNT. This shift reflects a higher electron density on the Pt atoms, which can be attributed to stronger electron donation from the adjacent metal oxide domains to the Pt nanoparticles. Among all the samples, Pt@Co shows the most pronounced shift to lower binding energy, indicating the highest degree of electron relocation. This observation suggests that the Pt-Co interface facilitates the strongest metal–support interaction, resulting in greater charge transfer to Pt and potentially leading to improved catalytic properties due to the modulation of the electronic structure of the active sites.

3.2. Electrochemical Performance in ORR

Inspired by the potential advantages, the ORR performances of as-prepared catalysts were first assessed by cyclic voltammetry (CV). As shown in Figure 4a, the superimposed CV curves exhibit three potential regions, including the hydrogen underpotential deposition region (denoted by green region; H^ads in reverse sweep) and hydrogen desorption region (yellow region; H^des in forward sweep) below 0.40 V vs. RHE (due to hydrogen adsorption and desorption), followed by the double-layer region (denoted by redish region between 0.40 V < E < 0.60 V vs. RHE) due to OH⁻ ion adsorption and oxide formation (O^ads.; forward sweep)/reduction (O^des.; backward sweep) region over 0.60 V vs. RHE [29,30]. The position and intensity of current signals (i.e., peaks) in these regions can be assigned to the adsorption and/or desorption of intermediate species. Notably, all the samples except Pt@Co exhibit profound hydrogen adsorption (H¹/H²)/desorption (H^1*/H^2*) peaks below 0.4 V vs. RHE, confirming the strong hydrogen evolution activity on the surface of these catalysts. On the other hand, the Pt@Co catalyst exhibits a smeared peak profile in this region, confirming that surface reaction sites are free from hydrogen evolution activities. Furthermore, the smallest width (Δh) of the double region suggests the lowest density of OH⁻ on the surface of Pt@Co. These two results integrally confirm that the surface reaction sites of the Pt@Co catalyst are free for ORR. These scenarios are further confirmed by the offset of oxygen reduction peak potentials (O^des.) to the highest potential (Figure 4b), indicating the lowest energy barrier for oxide reduction on the surface of Pt@Co as compared to commercial Pt/C and other counterparts [31].

Figure 5a presents the ORR polarization curves of the synthesized catalysts compared to the commercial Pt/C (Johnson Matthey) catalyst, with current densities normalized to the geometric area of the carbon electrode. Among all samples, Pt@Co shows the highest onset potential (V_OC) and half-wave potential (E_1/2), indicating a lower energy barrier and faster ORR kinetics (Figure 5b). This trend in V_OC and E_1/2 also aligns well with the position of the oxide reduction peak observed in the CV profiles (Figure 4a). To evaluate the ORR activity more quantitatively, the kinetic current densities (J_k) were calculated from the LSV data (Figure 5c). These J_k values were then normalized to the Pt loading to determine the mass activity (MA) at 0.85 V and 0.90 V vs. RHE (Figure 5d), following a previously reported method. As expected, Pt@Co displays the highest mass activity, reaching ~817 mA mg_Pt⁻¹ at 0.85 V and 464 mA mg_Pt⁻¹ at 0.90 V vs. RHE. In contrast, the commercial Pt/C shows significantly lower MA values of only 67 and 24.9 mA mgPt⁻¹ at 0.85 V and 0.90 V, respectively. Notably, all Pt-M/C catalysts outperform the commercial Pt/C catalyst, highlighting the advantage of bimetallic support interaction in enhancing ORR performance. The four-electron transfer pathway of the catalysts was confirmed by the Koutecky–Levich (K.L.) plots (Figure 5e), confirming efficient catalytic behavior with minimal formation of H₂O₂ as an intermediate. Additionally, the Tafel slope analysis (Figure 5f) was conducted to investigate the reaction kinetics. Among the tested samples, the Pt@Co catalyst exhibits the lowest Tafel slope of approximately 52 mV dec⁻¹, reflecting its superior kinetic performance for the ORR process.

4. Conclusions

In conclusion, this study demonstrates that strategic support engineering through the incorporation of transition metals significantly enhances the catalytic performance of Pt-based electrocatalysts for the oxygen reduction reaction (ORR). Among the synthesized binary Pt-M (M = Ni, Cu, Co) systems supported on carbon, Pt@Co exhibits superior ORR mass activity, outperforming both commercial Pt/C and other Pt-M/C counterparts. Detailed structural and spectroscopic investigations confirm strong electronic interactions between Pt and Co, resulting in electron transfer from Co to Pt and favorable modulation of the Pt d-band center. This electronic restructuring enhances oxygen adsorption–desorption kinetics, which directly contributes to improved ORR efficiency. These insights emphasize the importance of metal–support synergy in designing high-performance Pt-based catalysts, advancing their applicability in next-generation fuel cell technologies.