Electrodeposited Co Crystalline Islands Shelled with Facile Spontaneously Deposited Pt for Improved Oxygen Reduction

Abstract

The cobalt crystalline islands (Co_cryst) were electrochemically deposited onto a glassy carbon (GC) support and then modified by a facile spontaneous deposition of platinum. The electrocatalytic activity of the resulting Co_cryst-Pt core-shell catalyst was evaluated for the oxygen reduction reaction (ORR) in an alkaline medium. The XRD characterization of the Co_cryst-Pt islands revealed that the cobalt core had a hexagonal close-packed (hcp) crystalline structure, and that the platinum shell exhibited a crystalline structure with a preferential (111) orientation. SEM images showed that the average lateral size of the Co_cryst islands was 1.17 μm, which increased to 1.32 μm after adding platinum. The XPS analysis indicated that the outer layer of the bulk metallic Co_cryst islands was fully oxidized. During the spontaneous deposition of platinum, the outer Co(OH)₂ layer was dissolved, leaving the cobalt core in a metallic state, while the platinum shell remained only partially oxidized. The high electrochemically active surface area of the Co_cryst-Pt/GC electrode, along with a suitable crystalline structure of the Co_cryst-Pt islands, contributes to enhancing its ORR activity by providing a greater number of surface active sites for oxygen adsorption and subsequent reduction. The ORR on the Co_cryst-Pt catalyst occurs via a four-electron reaction pathway, with onset and half-wave potentials of 1.07 V and 0.87 V, respectively, which exceed those of polycrystalline platinum and a commercial benchmark Pt/C.

1. Introduction

Developing low-cost catalysts for an efficient oxygen reduction reaction (ORR) is crucial for advancing renewable energy technologies, including fuel cells, batteries, and water splitting [1,2,3]. Platinum-group metals are the most effective cathode catalysts, with platinum exhibiting the highest activity for oxygen reduction [4]. However, due to the high price of platinum, the key challenge for its commercial use is reducing the amount of pure platinum in a catalyst while improving its catalytic performance. In recent years, catalysts containing a small amount of Pt supported by inexpensive, electro-conductive, carbon-based materials have gained significant attention [5,6,7]. Nanostructured catalysts containing non-precious metals, which typically show low intrinsic activity for the oxygen reduction reaction, demonstrate enhanced performance when combined with other metals or appropriate supports [8]. Another promising category of materials for catalysts or supports includes transition metal oxides [9,10]. Among these, CoO_x-based catalysts, such as Co₃O₄ and Co(OH)₂, have been extensively researched for ORR and water splitting, either independently or when supported by carbon-based materials like graphene [11,12].

Among the various methods for preparing nanostructured CoO_x-based catalysts [13], electrochemical deposition is frequently favored due to its simplicity and ease of use. The composition of the depositing solutions and the type of support are important factors [14]. In addition, the CoO_x-species chemistry, quantity, and structure can vary depending on the electrochemical techniques employed. For potentiostatic deposition, key parameters include the applied potential and deposition time [15]. For the deposition via cyclic voltammetry, the parameters are the potential range and the number of cycles [16,17].

Various cobalt oxide (CoO_x)-based catalysts have been designed and tested for their effectiveness in improving oxygen reduction reaction performance [18,19,20,21,22]. The atomic tuning of single-crystal CoO nanorods involved engineering their surface structure to create oxygen vacancies on pyramidal nanofacets, resulting in exceptionally high catalytic activity and excellent durability [18]. The CoO hybrid with nitrogen-doped carbon nanotubes exhibited a higher ORR activity compared to the Co₃O₄/graphene hybrid, demonstrating significant stability in a highly aggressive corrosive environment (10 M NaOH at 80 °C) [19]. Additionally, a nitrogen-doped 3D crumpled graphene-CoO hybrid displayed outstanding catalytic activity comparable to that of the benchmark Pt/C catalyst and durability that surpassed Pt/C in an alkaline medium [20]. Similar excellent ORR performance was also observed in ultra-dispersed cobalt loaded on nitrogen-doped carbon (Co-NC900) [21] and CoO@Co nitrogen-doped carbon (CoO@Co/N–C) catalysts [22]. Numerous studies indicate that combining Pt with the early transition metals (Fe, Ni, Co) can reduce costs while enhancing catalytic activity for oxygen reduction [23,24,25,26,27,28,29]. This combination leads to the formation of bimetallic catalysts, which utilize synergy and strain effects to optimize the adsorption of reactants and increase the availability of active sites on the catalyst’s surface [30].

Among the various non-precious metals, Co (Co(OH)₂ and Co₃O₄), as a low-cost catalyst or support, combined with precious metals, is a promising material for ORR electrocatalysts. The ORR activity and durability of such Co-based catalysts can be improved by tuning their morphology and combining them with Pt nanoparticles [28,29,30,31]. Oxygen reduction was studied on carbon-supported Pt–Co nanoparticles with low Pt loading in an alkaline solution. Despite the low platinum content, significant activity was observed in the Pt-rich shell and Pt-Co core particles obtained through potential cycling [32]. Various platinum-cobalt alloy nanoparticles, prepared using the nanocapsule method, were found to exhibit higher ORR catalytic activity and stability in alkaline electrolytes than in acidic [33]. Bimetallic platinum-cobalt nanoparticles supported on reduced graphene oxide, Pt₃Co/rGO, exhibited superior oxygen reduction activity compared to other tested catalysts, including Pt/C, Pt/rGO, and Pt₃Ni/rGO, in alkaline solutions [34]. In a separate study, PtCo alloyed nanoparticles supported on thiolated graphene (referred to as ER/PtCo-tG) were electrochemically reduced to improve their catalytic activity. The enhanced ORR activity can be attributed to the well-dispersed PtCo nanoparticles, the effective formation of the PtCo alloy, and the presence of Co(OH)₂ [35].

In this work, we investigated glassy carbon (GC)-supported Co-Pt catalysts for oxygen reduction in an alkaline medium. The electrochemically deposited cobalt crystalline catalysts, obtained by applying various constant potentials and durations, were explored as substrates for further modification by a spontaneously deposited platinum layer. Although the spontaneous deposition of Pt on bare GC is nearly hindered [36], it occurs with remarkable ease on GC covered with previously electrodeposited Co crystalline islands. The layer of platinum shells the cobalt crystalline islands, thus preserving their structural integrity and enhancing their catalytic performance during oxygen reduction. The most ORR active GC-supported Co_cryst and Co_cryst-Pt catalysts were characterized using cyclic voltammetry (CV), X-ray diffraction (XRD), scanning electron microscopy (SEM), and X-ray photoelectron spectroscopy (XPS). A rotating disc electrode (RDE) arrangement was explored for investigations of the oxygen reduction reaction on Co_cryst/GC and Pt-Co_cryst-Pt/GC electrodes using electroanalytical techniques, including linear sweep voltammetry (LSV) and chronoamperometry (CA).

2. Results and Discussion

2.1. The Preparation and Characterization of Co/GC and Co-Pt/GC Electrodes

2.1.1. The Preparation of Co/GC and Co-Pt/GC Electrodes

The electrochemical deposition of Co onto a GC substrate was carried out from a deaerated 10⁻⁴ M CoSO₄ aqueous solution at various deposition times and different constant potentials, as described in the Supplementary Materials and illustrated in Figure S1. The cyclic voltammetry characterization of various Co/GC electrodes, conducted in a deaerated 0.1 M NaOH, along with preliminary testing of their ORR activity in an oxygen-saturated 0.1 M NaOH, is detailed in the SI and depicted in Figure S2. Since the Co/GC electrode produced by Co deposition at a constant applied potential of −0.36 V for 30 min demonstrated the best ORR activity, it will be further utilized as a substrate for subsequent spontaneous Pt deposition. The chronoamperometry curve (Figure S3) and the calculation of the Co metal loading achieved during the 30 min deposition (0.3 mg cm⁻²) are presented in the Supplementary Materials.

According to the Nernst equation and the standard potential for Co²⁺/Co of −0.28 V vs. SHE [37], the calculated equilibrium potential for Co electrodeposition from a 10⁻⁴ M Co²⁺ solution at a pH of 6.1 is − 0.34 V vs. RHE. A slightly lower selected potential of −0.36 V (−0.71 V vs. SHE), according to the Pourbaix diagram [38,39], indicates that only crystalline metallic Co is deposited, according to Equation (1). This electrode obtained after 30 min Co electrochemical deposition at −0.36 V will be referred to as Co_cryst/GC or the Co_cryst catalyst.

$${\mathrm{C}\mathrm{o}}^{2+}+2\mathrm{e}\to {\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{r}\mathrm{y}\mathrm{s}\mathrm{t}}^0$$

(1)

In addition, in contact with a close to neutral Co depositing solution, the outer layer of the deposited cobalt islands undergoes oxidation to Co(OH)₂, with a simultaneous hydrogen evolution. The Co_cryst islands will be referred to as Co_core_Co⁰, or Co_core_Co(OH)₂, indicating that the Co atoms in the outer layer are in a zero-oxidation state or oxidized.

$${\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}\_{\mathrm{C}\mathrm{o}}^0\to {\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}\_{\mathrm{C}\mathrm{o}\left(\mathrm{O}\mathrm{H}\right)}_2+{\mathrm{H}}_2$$

(2)

Co-Pt catalysts were prepared by immersing the most ORR-active Co_cryst/GC electrode in a deposition solution consisting of 1 mM H₂PtCl₆ and 0.05 M H₂SO₄ for various times. This is described in the SI and illustrated in the CVs shown in Figure S4a. According to their corresponding ORR polarization curves in Figure S4b, the best ORR activity is achieved for 60 min Pt deposition. This electrode will be referred to as the Co_cryst-Pt/GC or as the Co_cryst-Pt catalyst.

2.1.2. The Electrochemical Characterization of Co_cryst/GC and Co_cryst-Pt/GC Electrodes

The CV of Co_cryst/GC obtained for Co deposition at −0.36 V (Figure 1a) exhibits several distinct properties during the anodic scan characteristic of the bulk cobalt electrode. These include an extended peak at approximately 0.42 V with a shoulder at a lower potential attributed to the stripping of Co(OH)₂ [3] along with several anodic peaks at higher potentials associated with the transition to higher oxidation states of cobalt, including Co(II)/Co(III), and the formation of Co₃O₄ and CoOOH [39,40,41,42]. Additionally, there may also be peaks related to the Co(III)/Co(IV) transition [43].

The first anodic peak (from 0.0 to 0.9 V), the stripping of the Co(OH)₂ outer layer, may be used to determine the electrochemically active surface area (ECSA) of the Co_cryst/GC electrode. The charge obtained by integrating the Co/Co(OH)₂ peak is 17.8 mC cm⁻². By dividing this value with the calculated charge passed during the stripping of the first Co(OH)₂ monolayer of 406 μC/cm² (considering the exchange of 2e per single cobalt atom, a Co(OH)₂ unit cell dimension of 0.32 x 0.32 nm, and the geometric surface area of 0.196 cm²), the resulting ECSA value for Co_cryst/GC is 8.6 cm².

Figure 1b shows changes in the open circuit potential (OCP) during Pt spontaneous deposition. The starting OCP value for the Co_cryst/GC electrode is −0.02 V vs. RHE (−0.08 V vs. SHE), confirming that at this potential, according to the Pourbaix diagram [38,39], the surface of Co_cryst islands is covered with Co(OH)₂. Such islands consisting of a metallic Co core encapsulated by a Co(OH)₂ shell will be referred to as ${\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0\_{\mathrm{C}\mathrm{o}\left(\mathrm{O}\mathrm{H}\right)}_{2,\mathrm{s}\mathrm{h}\mathrm{e}\mathrm{l}\mathrm{l}}$. Analogous core-shell Co_Co(OH)₂ nanoparticles were reported in ref. [39]. This Co(OH)₂ shell undergoes a dissolution in an acid Pt depositing solution (pH 1), leaving the metallic Co⁰ place free to become simultaneously occupied by the adsorbed ${\mathrm{P}\mathrm{t}\mathrm{C}\mathrm{l}}_{6,\mathrm{a}\mathrm{q}}^{2-}$ anions from the depositing solution according to Equations (3) and (4):

$${\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0\_{\mathrm{C}\mathrm{o}\left(\mathrm{O}\mathrm{H}\right)}_{2,\mathrm{s}\mathrm{h}\mathrm{e}\mathrm{l}\mathrm{l}}+2{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0+{\mathrm{C}\mathrm{o}}_{\mathrm{a}\mathrm{q}}^{2+}+{2\mathrm{H}}_2\mathrm{O}$$

(3)

$${\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0+{\mathrm{P}\mathrm{t}\mathrm{C}\mathrm{l}}_{6,\mathrm{a}\mathrm{q}}^{2-}\to {\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0{\_\mathrm{P}\mathrm{t}\mathrm{C}\mathrm{l}}_{6,\mathrm{a}\mathrm{d}\mathrm{s}}^{2-}$$

(4)

Due to the higher Nernst potential for Pt⁴⁺/Pt⁰ (0.07 V vs. SHE, see ref. [36]) than for Co²⁺/Co⁰, the Pt⁴⁺ from the adsorbed ${\mathrm{P}\mathrm{t}\mathrm{C}\mathrm{l}}_{6,\mathrm{a}\mathrm{d}\mathrm{s}}^{2-}$ anion is spontaneously reduced to metallic Pt, while the two Co⁰ atoms from the surface of a Co core are simultaneously oxidized to Co²⁺, thus forming the islands consisting of metallic Co-core and metallic Pt-shell:

$${\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0{\_{2\mathrm{C}\mathrm{o}}^0\_\mathrm{P}\mathrm{t}\mathrm{C}\mathrm{l}}_{6,\mathrm{a}\mathrm{d}\mathrm{s}}^{2-}\to {\mathrm{C}\mathrm{o}}_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{e}}^0\_{\mathrm{P}\mathrm{t}}_{\mathrm{s}\mathrm{h}\mathrm{e}\mathrm{l}\mathrm{l}}^0+2{\mathrm{C}\mathrm{o}}_{\mathrm{a}\mathrm{q}}^{2+}+6{\mathrm{C}\mathrm{l}}_{\mathrm{a}\mathrm{q}}^{-}$$

(5)

A slight increase in the OCP over time indicates that the surface Co(OH)₂ oxide is gradually dissolving and subsequently being replaced by spontaneously deposited metallic Pt. Upon completing this process, there is a further sharp increase in the OCP to 0.92 V, corresponding to the OCP of bare polycrystalline Pt in 0.05 M H₂SO₄ (Figure S5). According to the CV of bare Pt(poly) in 0.05 M H₂SO₄ (insert in Figure S5), at the potential corresponding to this OCP value, Pt is in the oxidized state. With prolonged spontaneous Pt deposition, the OCP remains constant, indicating that Pt is additionally deposited on the existing Pt shell, resulting in an increased thickness of the Pt layer.

Figure 1c,d display the CVs of Co_cryst-Pt/GC, Pt_spont/GC, and Pt(poly), respectively, in a 0.1 M NaOH solution, recorded after holding the potential for 5 min at 0.2 V to reduce Pt oxide to its metallic state, and after the first scan in wider potential limits to remove the traces of chlorides and other impurities. The CVs of Co_cryst-Pt/GC and Pt(poly) exhibit all the typical features of polycrystalline Pt in 0.1 M NaOH, consistent with previous reports [44]. The hydrogen adsorption/desorption process occurs within the potential range of 0.05 V to 0.18 V, featuring two pairs of peaks in both the cathodic and anodic directions. Oxidation begins at approximately 0.60 V with the formation of Pt(OH)₂ and PtO at higher potentials, showing a distinct peak at 0.86 V. The reduction peak is observed at 0.70 V. In contrast, the CV of the Pt_spont/GC electrode, obtained by a spontaneous Pt deposition for 60 min (Figure 1d), indicates that there is little to no Pt on the GC support, which is obvious when compared to the current densities in the CV of bare Pt. In addition, much higher current densities for Co_cryst-Pt/GC compared to bare Pt(poly) suggest much higher electrochemically active surface area (ECSA). The charge obtained by integrating hydrogen desorption peaks is 1.53 mC cm⁻² for Co_cryst-Pt/GC, and 259 μC cm⁻² for the Pt(poly) electrode used in the work. Considering the commonly accepted value for a hydrogen monolayer on a polycrystalline Pt of 210 μC/cm², the ECSAs for the Co_cryst-Pt/GC and Pt(poly) are 1.43 cm² and 0.24 cm², respectively.

2.1.3. Structural, Morphological, and Chemical Characterization of the Co_cryst/GC and Co_cryst-Pt/GC

XRD, SEM, and XPS measurements were performed on the as-prepared electrodes. After the deposition process is finished, the electrodes are transferred through the air before additional measurements. Since both Co_cryst/GC and Co_cryst-Pt/GC electrodes contain oxides formed in contact with the solution, their outer spheres are also susceptible to further oxidation in air.

Figure 2 shows XRD profiles and corresponding SEM images of electrochemically deposited Co_cryst/GC and Co_cryst-Pt/GC, acquired through additional spontaneous Pt deposition. Amorphous peaks of glassy carbon dominate both XRD profiles of Co_cryst/GC (Figure 2a) and Co_cryst-Pt/GC (Figure 2b). For the Co_cryst/GC, the diffraction pattern displays three peaks centered at 41.7⁰ (100), 44.5⁰ (002), and 47.5⁰ (101), the signature peaks for the metallic hexagonal close-packed (hcp) Co phase (PDF 1-1256; hexagonal s.g. P63/mmc no.194) [45,46]. There are no traces of face-centered cubic (fcc) (signature peak at 52⁰) structure [46] or the Co oxide (that would be at 36.3 (111) and 42.3 (200)) [47]. This result indicates that the electrochemically deposited Co_cryst islands are composed of elemental crystalline Co within their volume, with the exposed preferential signature (100) and (002) peaks. This is consistent with the observation for Co nanoparticles electrodeposited at the same applied potential of −0.36 V [48]. Due to the size of the deposited Co crystallites, the exposure of the (110) orientation is likely associated with the preferential alignment of their edges.

For the Co_cryst-Pt/GC, in addition to the peaks of glassy carbon, small but noticeable peaks of metal Pt (PDF 1-1194; cubic s. g. Fm-3m no. 225) appear in the XRD profile, with no traces of Co metal or Co oxide phases, suggesting a very thick Pt shell. The preferred orientation of the Pt shell includes (111) and (110), consistent with the face-centered cubic (fcc) crystal structure of platinum. Moreover, the hcp Co surface of (100) orientation is structurally similar to the fcc-Pt(111), both exhibiting a hexagonal structure of surface atoms, suggesting that the Pt layer is expected to exhibit a preferential (111) orientation.

Figure 2c,d show SEM images of the Co_cryst/GC (left column) and Pt-Co_cryst-Pt/GC surface (right column), showing surface topography, revealing that in both cases, the deposited larger Co_cryst and Co_cryst-Pt hexagonally shaped crystalline islands are randomly distributed over the GC surface. In addition, the image of Co_cryst/GC reveals a layer of the deposited cobalt species between the Co_cryst islands. In contrast, the image of Pt-Co_cryst-Pt/GC reveals only bare GC support between the deposited Co_cryst-Pt islands, meaning that the previously deposited CoO_x layer was dissolved during spontaneous Pt deposition from an acidic deposition solution and that Pt is deposited on the remaining bare metallic Co_cryst islands. A spontaneous deposition of Pt on bare GC support can be regarded as negligible. The average lateral sizes of individual Co_cryst and Co_cryst-Pt islands are 1.17 μm and 1.32 μm, respectively, as shown in the island size distribution diagrams in Figure 2e,f, obtained from the images with a lower magnification as illustrated in the SI, Figure S6.

The chemical composition analysis was performed using energy dispersive X-ray spectroscopy (EDS). This hcp structure of the deposited Co crystalline islands aligns well with the hexagonal shape of the Co_cryst islands in the corresponding SEM image (Figure 3a). The SEM image in Figure 3b reveals that the hexagonal shape of the deposited islands is preserved upon the addition of a Pt shell. EDS analysis (Figure 3c,d) confirms that Co_cryst/GC consists of carbon, oxygen, and cobalt, while Coc_ryst-Pt/GC consists of carbon, oxygen, cobalt, and platinum.

The XPS analyses presented in Figure 4 provide information on the surface composition of the outer layer of Co_cryst islands and of the Pt shell of the Co_cryst-Pt core-shell islands. The XPS survey spectrum of Co_cryst/GC, Figure 4a, provides a broad overview of the electrode surface composition, revealing multiple peaks corresponding to the elements present, including Co, C, and O. No other elements are detected in the spectrum.

The Co 2p spectrum for Co_cryst/GC, Figure 4b, displays prominent peaks at binding energies of 781.6 eV and 797.4 eV, corresponding to the Co 2p_3/2 and Co 2p_1/2 levels, respectively. These peaks are ascribed to the cobalt oxide species present in the Co²⁺ oxidation state [49,50,51,52], most likely Co(OH)₂ [49,50], or as a mixture of CoO and Co(OH)₂ [50]. Additionally, two satellite peaks are observed at 786.3 eV and 803.6 eV, characteristic of Co²⁺ species resulting from multiplet splitting, which arises from the interaction of the 2p core hole with the unfilled 3d states of the Co²⁺ ion. These satellites are commonly observed in cobalt oxides, further confirming the presence of Co²⁺ in the Co(OH)₂ phase [49]. No Co³⁺ [51,52] and metal Co⁰ [49] are observed in the spectrum.

The O 1s spectrum for Co_cryst/GC (Figure 4c) shows two distinct peaks. The peak at 530.3 eV is attributed to the oxygen in Co(OH)₂-CoO (7.4% of O 1s emission), confirming the presence of cobalt oxide on the surface. The second peak, at 532.0 eV, can be attributed to adsorbed oxygen species (O⁻, O²⁻, O₂⁻), accounting for a large amount of oxygen detected in the survey spectrum (92.6% of O 1s emission). In addition, based on a small C 1s signal in the survey spectrum, the same peak at 532.0 eV can be attributed to C-O bonds from the GC support, which are located at a similar binding energy [51,52], although in a much smaller amount.

The XPS survey spectrum of Co_cryst-Pt/GC, Figure 4d, displays multiple peaks corresponding to Pt, C, O, and Cl. In this case, Co cannot be detected, as most Co-crystalline islands are encapsulated by platinum, which is preferentially deposited on the pre-existing cobalt crystals. Since XPS is a surface-sensitive technique, it cannot detect cobalt beneath the platinum layer. The peak corresponding to Cl is attributed to the chlorine from the precursor H₂PtCl₆ chemical used for platinum deposition.

The Pt 4f spectrum of Co_cryst-Pt/GC, Figure 4e, exhibits six peaks, corresponding to three distinct components in the Pt 4f₇/₂ and Pt 4f₅/₂ regions. These peaks are assigned to distinct chemical states of platinum present on the surface. The peaks at 75.2 eV and 71.8 eV (47.7% of Pt 4f emission) are associated with Pt⁰, indicating the presence of platinum in its metallic state [36,53]. The peaks at 75.7 eV and 72.8 eV (48.6% of Pt 4f emission) correspond to PtO_x, which represents platinum in an oxidized state, or platinum species that have undergone partial oxidation during the exposure of the electrode in air, possibly as platinum oxide, platinum hydroxide Pt(OH)₂ [36,53]. A roughly equal amount of platinum is in the metallic and oxidized states. The remaining small peaks at 78.3 eV and 73.8 eV (3.7% of Pt 4f emission) are attributed to PtCl₄, indicating platinum in a chloro-complex form, likely due to residual chlorine from the H₂PtCl₆ precursor chemical used during platinum deposition [36,53].

The spectrum of O 1s for Co_cryst-Pt/GC, Figure 4f, shows two distinct peaks at 531.7 eV and 532.6 eV, corresponding to oxygen bound to Pt in various PtO_x states (67.4% of O 1s emission) in either various adsorbed oxygen species (O⁻, O²⁻, O²⁻) or C-O bonds from the GC substrate (32.6% of O 1s emission), respectively [36,53]. Compared to the O 1s spectrum of Co_cryst/GC, the Co_cryst-Pt/GC spectrum shows significantly less oxygen overall, with the C 1s peak exhibiting much higher intensity. This suggests that the O 1s signal in the Co_cryst-Pt/GC originates mainly from the GC support rather than from adsorbed oxygen species. The increase in the relative intensity of the C 1s peak indicates that the carbon support dominates the surface more. In contrast to the previous case, where the GC surface between Co crystalline islands was also covered by the deposited Co oxide, during spontaneous Pt deposition, these Co oxides are dissolved in acidic depositing solution, leaving partially exposed bare GC substrate among the Co_cryst-Pt islands. This increases the amount of oxygen bound to the carbon from the GC support and reduces the amount of oxygen bound to surface metal oxide species, specifically PtO_x.

The atomic and weight percentages of the constitutive elements for Co_cryst/GC and Co_cryst-Pt/GC are shown in

Table 1. Atomic and weight percentage of the elements for Co_cryst/GC and Co_cryst-Pt/GC.

Line	Cocryst/GC		Cocryst-Pt/GC
	at%	wt%	at%	wt%
C 1s	11.6	6.5	50.5	24.7
O 1s	74.4	55.2	36.2	23.6
Co 2p	14.0	38.3	-	-
Pt 4f7/2	-	-	5.0	39.7
Cl 2p	-	-	8.3	12.0

, as estimated from the relative intensity of their corresponding peaks in the survey spectra. It is important to note that in the case of Co_cryst-Pt/GC, cobalt is not detected by XPS, and its amount cannot be estimated. This indicates that Co_cryst islands are fully covered with a Pt layer. According to the island size distribution (see Figure 2e,f), the average sizes of the Co_cryst and Co_cryst-Pt islands are 1.17 and 1.32 μm, respectively. The difference of 150 nm suggests that the average thickness of the Pt layer is 75 nm. Since XPS enables analysis of samples to a depth of 3–10 nm, the Co core lies beyond this range, allowing only the outer Pt layer to be detected and analyzed.

2.2. Oxygen Reduction Reaction on Co_cryst/GC and Co_cryst-Pt Catalysts

The polarization curves for the oxygen reduction reaction, recorded in an oxygen-saturated 0.1 M NaOH solution and the corresponding Koutecky–Levich (K–L) plots, are shown in Figure 5.

The ORR curves in Figure 5a,b were recorded in the cathodic scan direction for five rotation rates. Generally, the oxygen reduction on Co_crystPt/GC occurs at more positive potentials and with higher current densities than on C_cryst/GC. The onset (E_onset) and half-wave (E_1/2) potentials for ORR on C_cryst/GC are 0.82 and 0.62 V, respectively. On the other hand, for ORR on Co_crystPt/GC, E_onset and E_1/2 are 1.07 and 0.87 V, respectively.

Koutecky–Levich plots, Figure 5b,d, are constructed using data from polarization curves (Figure 5a,c) when the inverse current density (1/j) is plotted as a function of the inverse of the square root of the rotation rate (ω^1/2).

For Co_cryst/GC, Figure 5b illustrates that K–L diagrams are parallel across the entire ORR potential region, indicating that the reaction adheres to first-order kinetics with respect to molecular oxygen. The calculated number of exchanged electrons remains consistently four, irrespective of variations in potential. These results demonstrate that the Co_cryst/GC electrode facilitates ORR through a 4e⁻ reaction pathway. Since in the entire ORR potential region, the surface is covered with Co(OH)₂, it may be assumed that the OH acts as a catalyst for the adsorption of oxygen and its subsequent reduction, like in the case of Au(100) single crystal [54].

The ECSA provided above for Co_cryst/GC, estimated from CV in Figure 1a and corresponding to the stripping of the first Co(OH)₂ layer, reflects the state of the as-prepared surface. However, at the ORR onset potential of 0.82 V, the surface has already transformed from its initial state (as illustrated in Figure S7 by the repeated cycling in the potential region from 0.0 to 0.9 V). This indicates that the ECSA corresponding to the onset of ORR differs from the initial one at around 0.0 V, which is demonstrated by the difference in double-layer capacity (Figure S8a,b). Considering that the C_dl value for bare Co is 40 μF cm⁻² [55], the obtained ECSA for the potential region from −0.02 to 0.08 V (0.03 V midpoint potential) is 9.8 cm², while for the potential region around ORR onset potential from 0.79 to 0.85 V, it is 2.9 cm² (see Figure S9a,b). This implies that, at more positive potentials, the surface is partially passivated due to the formation of higher oxides, resulting in a lower availability of Co(OH)₂ surface sites, which act as catalysts for oxygen adsorption and its subsequent reduction.

For Co_cryst-Pt/GC, Figure 5d, the linear K–L diagrams suggest first-order kinetics of the ORR concerning molecular oxygen. Unlike in the previous case, slight non-parallelism of the K–L plots at higher potentials indicates a change in the total number of electrons exchanged with potential. In these activation and mixed control regions, the ORR occurs on the oxidized Pt layer, suggesting that the presence of PtOH on the surface partially hinders oxygen adsorption [56]. In contrast, in the diffusion control region at lower potentials, the K–L curves become parallel, indicating that the total number of electrons exchanged remains constant and does not vary with potential. The total number of electrons exchanged is four, and the ORR follows a 4e⁻ reaction pathway.

Like in the previous case, considering that the C_dl value for bare Pt is 60 μF cm⁻² [55], the ECSAs for Co_cryst-Pt/GC obtained from CV in Figure 1c and from C_dl around the ORR onset potential (from 1.02 to 1.2 V, midpoint potential 1.07 V, see Figure S9c,d) differ. The ECSA value of 5.2 cm² obtained from C_dl is significantly higher than the 1.43 cm² from the hydrogen desorption peak, contributing to the enhanced ORR activity.

Comparison of the ORR Catalytic Activity and Stability of Co_cryst-Pt/GC with That of Pt_spont/GC, Pt(poly), and Co_cryst/GC

Figure 6 shows LSV curves for oxygen reduction in 0.1 M NaOH recorded in the cathodic direction at a rotation rate of 1600 rpm, along with the corresponding Tafel plots for the Pt_spont/GC, Pt(poly), Co_cryst/GC, and Co_cryst-Pt/GC electrodes.

The oxygen reduction polarization curves (Figure 6a) indicate that the ORR catalytic activity of Pt_spont/GC is relatively low, with an onset potential (E_onset) of 0.72 V and a half-wave potential (E_1/2) of 0.58 V. This low ORR activity, which is only slightly better than that of bare GC (shown in Figure S1b), is attributed to the negligible Pt content. In contrast, the Co_cryst/GC electrode demonstrates much higher catalytic activity, although it still does not reach that of Pt(poly) (E_onset = 0.98 V, E_1/2 = 0.84 V). On the other hand, the Co_cryst-Pt/GC shows a significant enhancement in catalytic activity, surpassing that of Pt(poly).

The corresponding Tafel slopes are shown in Figure 6b. The slope of −100 mV dec⁻¹ for the ORR on Pt_spont/GC over the entire potential range matches the value obtained for bare GC [57], as well as for Co_cryst/GC. The slope of −100 mV dec⁻¹, obtained for the ORR on Co_cryst/GC, is comparable to that of various carbon-based electrocatalysts featuring Co nanoparticles [21,51,58,59]. For the Pt(poly) and Pt-Co_cryst/GC electrodes, two different Tafel slope values were identified at low and high overpotentials. The −60 and −120 mV dec⁻¹ slopes for Pt(poly) align with previous reports [36]. The two slopes at high and low overpotentials reflect the changes in adsorption conditions, transitioning from a Temkin (−60 mV dec⁻¹) to a Langmuir (−120 mV dec⁻¹) model, as the transfer of the first electron to O₂ adsorbed on the surface is the rate-determining step, as already discussed in ref. [36].

The parameters for the GC-supported Co_cryst, Co_cryst-Pt, and Pt_spont catalysts compared to bare Pt(poly), including ECSA, E_onset, E_1/2, and Tafel slope for the ORR in 0.1 M NaOH solution at 1600 rpm, are summarized in SI, Table S1.

Figure 7 illustrates the long-term stability testing of the Co_cryst-Pt catalyst during the ORR in oxygen-saturated 0.1 M NaOH. The testing involved recording a chronoamperometry curve over 5 h at a constant applied potential and 2000 LSV curves during continuous cycling within the ORR potential range at a scan rate of 100 mV s⁻¹. In all cases, the rotation rate was 1600 rpm.

The CA curves (Figure 7a) illustrate the long-term stability performance of the Co_cryst-Pt core-shell islands compared to Co_cryst islands and Pt(poly). At an applied potential of 0.87 V, the ORR exhibits significantly higher current density for the Co_cryst-Pt than for Pt(poly). Initially, the current density decreases, but then stabilizes during prolonged measurements. However, a slight decline in activity is observed for extended durations of the ORR on Co_cryst-Pt. A similar trend is noted for the ORR on Co_cryst alone (recorded at the potential of 0.67 V), although there is a greater decrease in activity during the first hour.

Figure 7b,c show LSV curves for ORR on Co_cryst-Pt catalyst before and after the stability measurements and before and after 2000 cycles (performed at a scan rate of 100 mV s⁻¹), respectively, recorded at a scan rate of 50 mV s⁻¹. In both cases, LSV curves exhibited a slight decline in stability (similar to that of bare Pt(poly), as illustrated in Figure S10a). In contrast, the LSV curves for ORR on Co_cryst islands (Figure S10b) demonstrate a significant ORR activity decline, likely due to the formation of higher Co oxide, which contributes to increased surface passivation. The better stability of the Co_cryst-Pt catalyst suggests the protection of the Co_cryst core by a Pt layer. However, further work is necessary to improve the long-term stability of the Co_cryst-Pt catalyst, primarily focusing on preventing the degradation and dissolution of the outer Pt layer.

Due to the large size of Co_cryst islands at the micrometer scale and the thickness of the platinum layer of a few tenths of nanometers, no electronic effects are expected to play a role in the increased ORR activity of the Co_cryst-Pt core-shell islands compared to Co_cryst alone. Based on the presented results, it is reasonable to conclude that, besides the higher intrinsic activity of platinum compared to cobalt, the primary reason for the enhanced ORR activity is the favorable surface chemical composition and structure of the Co_cryst islands. This Co surface chemistry and structure facilitate the spontaneous deposition of Pt. Moreover, a hexagonal structure of both Co(100) and Pt(111) surface atoms promotes the growth of a crystalline Pt(111) layer and maintains the crystalline integrity of the electrochemically deposited Co_cryst islands. This leads to the significantly increased ECSA of Co_cryst-Pt core-shell islands compared to both Co_cryst islands alone and Pt(poly), meaning a greater number of surface active sites for oxygen adsorption and its subsequent reduction. As a result, the electrochemically active surface area of the Co_cryst-Pt core-shell islands is significantly greater than that of polycrystalline platinum. Such a high ECSA corresponds to many platinum surface active sites available for oxygen adsorption and subsequent reduction, thus enhancing the overall ORR efficiency.

In addition, the results obtained in this work are compared with the literature data for ORR in both alkaline and acid media on various Co-Pt catalysts supported by carbon-based materials, as well as with the benchmark Pt/C catalyst, as summarized in

Table 2. Comparison of the ORR activity of various Pt-Co catalysts.

Catalyst	Solution	Eonset (V)	E1/2 (V)	Tafel Slope (mV dec−1)	Reference
Pt3Co/r-GO	0.1 M KOH	1.13	0.76	-	[34]
ER/PtCo-tG	0.1 M KOH	0.94	0.86	-	[35]
PtCo@NMC	0.1 M KOH	1.04	0.96	67.4	[60]
Pt76Co24	1 M NaOH	0.98	-	59.98	[61]
Co/Pt(111)	0.1 M KOH	-	-	51	[62]
PtCo-3	1 M KOH	-	0.86	39.2	[63]
125-PtCo@C–Co3O4	1 M KOH	0.95	-	38	[64]
Pt3Co/NC	1 M KOH	-	0.91	-	[65]
PtCo-NC-4	0.1 M HClO4	1.03	0.93	53.7	[66]
PtCo@CoNC/NTG	0.1 M HClO4	-	0.94	71	[67]
5-wt% Pt/h-Co-NC	0.1 M HClO4	1.00	0.87	91	[68]
PtCo-DPC	0.1 M HClO4	1.02	0.85	87.5	[69]
Cocryst-Pt	0.1 M NaOH	1.07	0.87	63	This work
Pt/C	0.1 M KOH	1.01	0.89	102.6	[60]
20 wt%Pt/C	0.1 M NaOH	-	0.85	69	[70]

.

The ORR activity of the obtained GC-supported Co_cryst-Pt catalyst is among the best concerning the initial potential of 1.07 V, which is within the range of the best Pt-Co catalysts reported in refs. [34,60], surpassing the activity of Pt(poly) and Pt/C [60,70]. The Tafel slopes of −63 and −120 mV dec⁻¹ for the Pt-Co_cryst/GC are consistent with those previously reported in refs. [60,61,62], although a lower value is reported in refs. [63,64].

3. Materials and Methods

3.1. Catalyst Preparation

The electrochemical deposition of cobalt onto a glassy carbon supporting electrode was conducted in a three-electrode cell using a 10⁻⁴ M CoSO₄∙7H₂O aqueous solution. Different constant deposition potentials and times were applied to create Co/GC electrodes. Additionally, platinum was spontaneously deposited by immersing previously prepared Co/GC electrodes into a 10⁻³ M H₂PtCl₆ + 0.05 M H₂SO₄ solution for various durations.

3.2. Material Characterization

XRD measurements were performed using a Philips PW 1050 X-ray powder diffractometer (Philips, Delft, The Netherlands). The measurements utilized Ni-filtered Cu Kα radiation and employed Bragg–Brentano focusing geometry. Diffraction intensity data were recorded over the 2θ range of 10–80° with a step size of 0.02° and a counting time of 5 s per step.

SEM images and EDS spectra were recorded using a FESEM microscope (FEI Scios 2, Thermo Fisher Scientific, Waltham, MA, USA) at a pressure of 1 × 10⁻⁶ mbar and an electron beam voltage of 10 kV. The duration for spectrum acquisition from the selected micro areas was 30 min.

Survey and high-resolution XPS spectra were recorded using SPECS Systems (SPECS Surface Nano Analysis GmbH, Berlin, Germany) with XP50M X-ray source for Focus 500 and PHOIBOS 100/150, using AlKα source (1486.74 eV) at a 12.5 kV, and 32 mA at a pressure of 9 × 10⁻⁹ mbar. C 1s peak at 284.5 eV was used as a reference for all peak positions.

3.3. Electrochemical Measurements

The PINE potentiostat (PINE Instruments Co., Grove City, PA, USA) and a three-electrode electrochemical cell were employed for electrochemical measurements. Different Co and Co-Pt catalysts supported by GC were prepared and used as the working electrodes. A Pt-wire served as the counter, while an Ag/AgCl, 3 M KCl was utilized as the reference electrode. All potential values were recalculated relative to the reversible hydrogen electrode (RHE).

The electrochemical characterization of the electrodes was conducted using cyclic voltammetry in a deaerated 0.1 M NaOH solution. Their ORR electrocatalytic activity was investigated in an oxygen-saturated 0.1 M NaOH solution through linear sweep voltammetry in a rotating-disc electrode setup.

The Koutecky–Levich (K–L) plot presents the relationship between the inverse current density value (1/j) and the inverse square root of the rotation rate (ω^−1/2).

The Koutecky–Levich equation was used for a detailed ORR analysis:

1/j = 1/j_k + 1/j_l = 1/j_k + 1/Bω^1/2

(6)

where j is the measured current density and j_k and j_l represent the kinetic and diffusion current density, respectively. B is a constant expressed as:

$$\mathrm{B}=0.62\mathrm{nF}\mathrm{C}{\mathrm{o}}_2{\mathrm{D}{\mathrm{o}}_2}^{2/3}{\mathsf{\nu }}^{-1/6}$$

(7)

where n is the total number of exchanged electrons, F is the Faraday constant (96,485 C mol⁻¹), $C{o}_2$ is the oxygen solubility (1.22 × 10⁻³ mol L⁻¹ [36]), $\mathrm{D}{\mathrm{o}}_2$ is the oxygen diffusivity (1.90 × 10⁻⁵ cm² s⁻¹ [36]), and ν is the kinematic viscosity (0.01 cm² s⁻¹ [36]) of the 0.1 M NaOH solution. The value of the constant B determined experimentally from the K–L diagram was used to calculate the total number of electrons exchanged.

3.4. Chemicals

Glassy carbon discs (Alfa Aesar GmbH & Co KG, Karlsruhe, Germany), mechanically polished with alumina powder of 1, 0.3, and 0.05 µm particle sizes, were used as supporting electrodes. A glassy carbon disc with a diameter of 5 mm (geometric area of 0.196 cm²) was mounted in a Teflon holder for electrochemical measurements. Additionally, a 7 mm diameter disc (geometric area of 0.385 cm²) was used for ex situ characterization.

CoSO₄∙7H₂O and H₂PtCl₆ (both from MaTech GmbH, Jülich, Germany), H₂SO₄ (Merck, Germany), and ultrapure Milli-Q water were used to prepare the Co and Pt deposition solutions. NaOH pellets (from Merck, Darmstadt, Germany) and Milli-Q water were used for the working solution for electrochemical measurements. The solutions were either deaerated with nitrogen (N₂, 99.9995%, Messer, Frankfurt, Germany) or saturated with oxygen (O₂, 99.9995%, Messer, Frankfurt, Germany).

4. Conclusions

The results presented in this study demonstrate a straightforward method for creating a Co-Pt core-shell catalyst by electrochemically depositing cobalt onto a glassy carbon support, followed by a spontaneous deposition of platinum. The Co_cryst and Co_cryst-Pt catalysts that have shown the best ORR activity were characterized, and their electrochemical performances were investigated in 0.1 M NaOH in detail.

For Co_cryst/GC, the XRD indicated the crystalline nature of the deposited Co islands with the preferential (100) surface structure, while SEM analyses revealed an average lateral size of 1.17 μm. XPS detected cobalt in only the Co²⁺ oxidation state, suggesting that Co_cryst islands consisted of a metallic Co-core and Co(OH)₂ shell. After adding Pt, XRD indicated that the metallic hexagonal close-packed Co core promotes the formation of a crystalline Pt shell with a preferred (111) orientation. SEM analysis revealed an increase in the average size of the Co_cryst-Pt islands to 1.32 μm. XPS detected only Pt, confirming a full coverage of the Co core with a thick Pt shell.

RDE results on the ORR activity of the Co_cryst-Pt/GC electrode demonstrated that the reaction proceeds through a 4e-reaction mechanism with an onset and half-wave potentials of 1.07 and 0.87 V, respectively, and a Tafel slope of 63 mV dec⁻¹. Due to the high intrinsic activity of the platinum shell and the large electrochemically active surface area, the resulting Co_cryst-Pt/GC catalyst exhibited enhanced catalytic activity for the oxygen reduction reaction, exceeding that of a polycrystalline platinum and a benchmark Pt/C.