Promoting the Electrocatalytic Ethanol Oxidation Activity of Pt by Alloying with Cu

Abstract

The development and commercialization of direct ethanol fuel cells requires active and durable electro-catalysts towards the ethanol oxidation reactions (EOR). Rational composition and morphology control of Pt-based alloy nanocrystals can not only enhance their EOR reactivity but also reduce the consumption of precious Pt. Herein, PtCu nanocubes (NCs)/CB enclosed by well-defined (100) facets were prepared by solution synthesis, exhibiting much higher mass activity (4.96 A mg_Pt⁻¹) than PtCu nanoparticles (NPs)/CB with irregular shapes (3.26 A mg_Pt⁻¹) and commercial Pt/C (1.67 A mg_Pt⁻¹). CO stripping and in situ Fourier transform infrared spectroscopy (FTIR) experiments indicate that the alloying of Cu enhanced the adsorption of ethanol, accelerated the subsequent oxidation of intermediate species, and increased the resistance to CO poisoning of PtCu NCs/CB, as compared with commercial Pt/C. Therefore, alloying Pt with earth-abundant Cu under rational composition and surface control can optimize its surface and electronic structures and represents a promising strategy to promote the performance of electro-catalysts while reduce the use of precious metals.

1. Introduction

Fuel cells represent an efficient route for the direct conversion of chemical energy to electricity with potential applications in sustainable energy storage, electricity generation, and transportation [1,2]. Among all kinds of fuel cells, direct ethanol fuel cells (DEFCs) can generate electricity under mild conditions by utilizing nontoxic ethanol with high energy density [3,4,5]. In particular, when coupled with the developing technology of sustainable photocatalytic/electrocatalytic CO₂ reduction reaction (CO₂RR), the CO₂RR-ethanol-DEFCs route would hold great potential for generating green electricity with a neutral carbon cycle [6,7]. To achieve this goal, one of the most essential challenges is to accelerate the sluggish kinetics for the ethanol oxidation reaction (EOR) at the anode of DEFCs [8,9,10]. To date, the precious noble metal Pt is generally considered the most efficient catalyst for EOR, along with other noble metals (e.g., Rh, Pd, etc.) and their alloys [11,12,13,14]. However, the overall efficiency of DEFCs is still limited by the inevitable CO poisoning of Pt-based EOR catalysts and the low selectivity to CO₂ as the final oxidation products [15]. In addition, the high costs of such noble metals impede the large-scale commercialization of DEFCs. One promising strategy is to modify the surface geometric and electronic structures of Pt-based catalysts though morphology and composition-controllable synthesis of the alloy nanocrystals of Pt and non-precious metals (Pt–M nanocrystals) [16,17,18,19,20,21]. For instance, various PtCu nanocrystals have been prepared and applied in an electrocatalytic formic acid oxidation reaction and a methanol oxidation reaction, displaying the composition and morphology effects on electrocatalysis reactions [17,22,23]. However, the application and comprehensive understanding of PtCu alloys in EOR are still elusive. For EOR applications, on one hand, rationally modifying the surface structures of Pt–M nanocrystals could tune the adsorption behavior of CO* and other key intermediates for EOR and thus prevent the CO poisoning, promote EOR efficiency, and enhance catalyst stability [20,21]. On the other hand, alloying with a non-precious metal could reduce the use of high-cost Pt to cut the catalyst costs for commercialization of DEFCs. Despite the enormous efforts devoted by researchers, there is still plenty of room for promoting the EOR catalysts’ activity and extending their stability, while reducing their costs.

Herein, we report that alloying with non-precious Cu can not only modify the surface electronic structures of Pt-based EOR catalysts to boost their activity and stability with enhanced CO resistance and CO₂ selectivity, but also reduce the catalyst cost by minimizing the use of precious Pt. Monodispersed PtCu NCs and PtCu NPs could be selectively prepared through the co-reduction of platinum acetylacetonate [Pt(acac)₂] and copper sulfate (CuSO₄) species in oleylamine (OAm) solutions with the addition of appropriate amounts of cetyltrimethylammonium bromide (CTAB) (Figure 1). X-ray photoelectron spectroscopy (XPS) and CO stripping results indicated that alloying with non-precious Cu tuned the electronic structures of Pt through charge transfer and/or compressing the lattice distances of Pt. The removal of CO* intermediates from the Pt surface was accelerated to increase the catalysts’ resistance to CO poisoning during EOR. The electrochemical in situ Fourier transform infrared spectroscopy (FTIR) study illuminated that the low-coordinated (100) facets of PtCu NCs/CB could drive the C−C bond cleavage and CO* intermediate oxidation, showing higher selectivity for CO₂, the final product for complete EOR. As a result, the PtCu NCs/CB showed much higher EOR mass activity (4.96 mA mg_Pt⁻¹) than PtCu NPs/CB (3.26 mA mg_Pt⁻¹) and commercial Pt/C (1.67 mA mg_Pt⁻¹) at the anodic potential of 0.75 V vs. RHE, with improved durability in both the chronoamperometry and accelerated cycling measurements.

2. Results and Discussion

2.1. Catalyst Synthesis and Characterization

We synthesized the PtCu NCs by a hot injection method. CuSO₄ and Pt(acac)₂ were used as the Cu and Pt sources, respectively. OAm was used as the solvent, and CTAB was added as the morphology modification surfactant. Pt–Cu alloy nanocrystals with varying compositions were also prepared by tuning the molar ratios of the Pt and Cu precursors. The compositions of all Pt–Cu nanocrystals were confirmed by inductively coupled plasma optical emission spectroscopy (ICP-OES), which were in good agreement with the molar ratios of the Pt and Cu precursors, suggesting that the Pt and Cu precursors were completely reduced during the synthesis. The X-ray diffraction (XRD) patterns of the different PtCu NPs showed that the peaks shifted from the Pt standard peaks (JCPDS #04–0802) to Cu standard ones (JCPDS #04–0836) when more Cu was added, demonstrating that the compositions of the PtCu alloy nanocrystals could be widely tuned to form the alloy of Pt and Cu (Figure S1) [22,23].

Monodispersed PtCu NCs enclosed by well-defined (100) facets with an average edge length of (14.08 ± 1.35) nm were obtained with high selectivity, as shown in the transmission electron microscopy (TEM) images (Figure 2a and Figure S2a). High-resolution TEM (HRTEM) images showed the high crystalline nature of the PtCu NC, where the lattice fringe spacing along the edge of the nanocube was 0.185 nm, located between the (100) lattice distances of face-centered cubic (fcc) Pt (0.196 nm) and Cu (0.181 nm) (Figure 2b,c). High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) imaging and the corresponding elemental mapping images of PtCu NCs suggested that Pt and Cu elements were evenly distributed throughout the nanocubes (Figure 2d). Energy-dispersive spectrometer (EDS) line-scan profile also suggested the homogeneous distribution of Pt and Cu elements in the PtCu NCs (Figure 2e and Figure S3). In addition, the Pt/Cu atomic ratio determined by EDS (Pt/Cu = 32.97/67.03, Figure S4a) was very close to that measured by ICP-OES (32.77/67.23), also suggesting the uniform distribution of Pt and Cu in the as-obtained PtCu NCs.

PtCu NPs were obtained by decreasing the amount of CTAB in the synthesis process. TEM images exhibited uniform PtCu NPs with sphere-like morphologies with a narrow size distribution of 14.04 ± 1.31 nm (Figure 2f and Figure S2b). The PtCu NPs showed lattice fringe spacings of 0.213 nm, which could be assigned to the corresponding (111) facets of fcc Pt–Cu alloy (Figure 2g). HAADF-STEM and corresponding elemental mapping images suggested that Pt and Cu were also homogeneously distributed, which was also confirmed by the line-scanning profile of a PtCu NP (Figure 2h,i). The atomic percentages of Pt and Cu were determined to be 33.62% and 66.38% by EDS (Figure S4b), which was in good agreement with ICP-AES (32.87% and 67.13%). The XRD patterns also confirmed that the as-obtained PtCu NPs were of the fcc phase Pt–Cu alloy (Figure S5).

The CTAB used in the synthesis was essential for the formation of the nanocubes [24,25]. A mixture of nanocubes and irregularly shaped nanoparticles was obtained with excessive CTAB (Figure S6a). Aggregations of irregularly shaped nanoparticles were obtained in the absence of CTAB (Figure S6d). It was worth noting that aggregations of irregularly shaped nanoparticles were obtained by substituting CTAB with potassium bromide (KBr) (Figure S7), suggesting that the combination of bromide ions and the long carbon chains of CTAB facilitated the formation and dispersion of the uniform PtCu NCs. In addition, the atomic ratios of Pt and Cu were critical for forming nanocubes. The Pt₄₉Cu₅₁ NPs showed irregular spherical profiles and the Pt₂₃Cu₇₇ NPs presented a mixture of nanocubes and nanotripods (Figure S8).

2.2. Surface Structures

XPS analyses suggested that the electronic states of Pt could be tuned by alloying with Cu. The high-resolution Pt 4f core-level spectrum of PtCu NCs/CB, PtCu NPs/CB, and commercial Pt/C are shown in Figure 3a. Compared with commercial Pt/C, these peaks for PtCu NCs/CB and PtCu NPs/CB shifted toward lower binding energy, indicating that strong interactions between Pt and Cu were accompanied by the electron transfer from Cu to Pt [21,26]. Systematic shift of the binding energy of Pt 4f _5/2 and Pt 4f _7/2 peaks was also observed in the high-resolution Pt 4f core-level spectrum of the Pt₄₉Cu₅₁/CB, PtCu NCs (Pt₃₃Cu₆₇)/CB, and Pt₂₃Cu₇₇/CB (Figure 3b), confirming the electron transfer from Cu to Pt. Additionally, two characteristic peaks of Cu 2p for the PtCu alloy nanocrystals were assigned to Cu 2p_1/2 and Cu 2p_3/2, respectively (Figure S9). The absence of observable satellite peaks for Cu²⁺ suggested that most Cu species in the Pt–Cu alloy were in the metallic state [21]. The electron transfer from Cu to Pt, as well as the compression of Pt lattices by alloying with small Cu atoms, would lower the d-band center of Pt [21], further promoting the adsorption of OH* species on Pt atoms to improve intrinsic activity and prevent the CO poisoning [18,23].

2.3. CO-Stripping Experiments

Preventing the CO poisoning for Pt active sites was critical for EOR promotion [16,17]. As shown in the CO-stripping experiments (Figure 3c), the PtCu NCs/CB showed lower onset and peak potentials (0.336 V and 0.631 V vs. RHE) for CO oxidation than for PtCu NPs/CB (0.343 V and 0.643 V vs. RHE) and Pt/C (0.436 V and 0.682 V vs. RHE), indicating a higher CO oxidation ability for PtCu NCs (Pt₃₃Cu₆₇)/CB in 1.0 M potassium hydroxide (KOH). The onset and peak potentials shift caused by the alloying of Pt with Cu indicated easier adsorption of OH* species on the surface of PtCu alloy than on pure Pt, which could promote the oxidation of CO* [21]. In addition, as shown in Figure 3d, a systematic shift of the CO oxidation peaks was also observed, with the Pt₂₃Cu₇₇/CB showing the lowest onset and peak potentials (0.331 V and 0.628 V vs. RHE). Such results suggest that alloying Cu could facilitate the oxidation of CO* intermediates on the Pt surface, and thus potentially promote the EOR process on such Pt–Cu alloy nanocrystals.

2.4. In Situ FTIR

Electrochemical in situ FTIR was performed to manifest the dissolved products in the thin layer during EOR on the PtCu NCs/CB and Pt/C. (Figure 3e,f). The peaks at ca. 1045 and 1087 cm⁻¹ were assigned to the C−O stretching vibration of ethanol [16,27,28,29], which could be firstly observed at 0.5 V and 0.7 V vs. RHE on PtCu NCs/CB and Pt/C, respectively. The signature peaks at 1550 cm⁻¹ were attributed to the asymmetric stretching vibration of the O−C−O band (ν_as (OCO)) for the incomplete oxidation products [28,29]. Moreover, the peaks at ca. 1414 cm⁻¹ were attributed to the symmetric stretching vibration of the O−C−O band (ν_s(OCO)) including the contributions from both acetate and carbonate (CO₃²⁻) species [28,29]. For neat acetate, the intensity ratio of the asymmetric and symmetric stretching vibration of the O−C−O band (Iν_as(OCO)/Iν_s(OCO)) should be fixed to ca. 2.6 [29]. However, the corresponding ratio obtained on PtCu NCs/CB samples was approximately 1.16, demonstrating the existence of carbonate, of which the IR band (1400–1490 cm⁻¹) could be covered in the 1414 cm⁻¹ band of acetate [29]. Such CO₃²⁻ species could be produced by reacting CO₂ with KOH in solution. That is, the PtCu NCs/CB could significantly promote the selectivity and activity for CO₂ in the EOR process, possibly due to the high density of Pt (100) surface sites, which favored the cleavage of the C−C bond of the ethanol molecule to produce CO₂ products [30,31].

Thus, all these spectrum characterizations and electrochemical experiments suggested that the composition and morphology effects could effectively enhance the activity, stability, and CO₂ selectivity of PtCu NCs/CB, making them promising candidates for EOR catalysts.

2.5. EOR Performance

The EOR properties of the as-prepared Pt–Cu alloy catalysts (PtCu NCs/CB and PtCu NPs/CB) and commercial Pt/C were tested in a typical three-electrode system with the electrolytes of 1.0 M KOH and 1.0 M C₂H₅OH. The peak current density (j) of the forward-scanning curve at 0.75 V vs. RHE was used to evaluate the EOR performance. The obtained current was normalized to the Pt mass and electrochemically active surface areas (ECSA) of the catalysts to obtain the mass activity and specific activity. As shown in Figure 4a,b, the mass activity of PtCu NCs/CB was 4.96 A mg_Pt⁻¹ for EOR, which was 1.5 and 3 times that of PtCu NPs/CB (3.26 A mg_Pt⁻¹) and commercial Pt/C (1.67 A mg_Pt⁻¹), respectively. The ECSA of each catalyst was examined by the charge transferred in the hydrogen underpotential (H_upd) region of the cyclic voltammogram (CV) curves (Figure S10). The specific activity of PtCu NCs/CB was 6.75 mA cm⁻²_(ECSA), which was 1.1 and 2.9 times that of PtCu NPs/CB (6.22 mA cm⁻²_(ECSA)) and commercial Pt/C (2.34 mA cm⁻²_(ECSA)), respectively (Figure 4b and Figure S11). That is, the PtCu NCs/CB exhibited superior catalytic activity for EOR (Table S1). In the Nyquist plots (Figure S12), PtCu NCs/CB exhibited the smallest diameter for the semicircle in the mixture of 1.0 M KOH and 1.0 M C₂H₅OH, suggesting that the PtCu NCs/CB could promote the electron transport for EOR [18,32].

For the accelerated durability performed from 0.00 V to 1.20 V vs. RHE at a scan rate of 50 mV s⁻¹ at room temperature, the mass activity of the PtCu NCs/CB, PtCu NPs/CB, and commercial Pt/C dropped to 1.69, 0.80, and 0.14 A mg_Pt⁻¹ after 500 cycles, respectively (Figure 4c and Figure S13). Notably, the bimetallic PtCu NCs/CB showed higher durability in cycling measurements toward EOR than PtCu NPs/CB and Pt/C. TEM images revealed no significant change in the morphology of PtCu NCs/CB after 500 testing cycles (Figure S14). In addition, the chronoamperometric curves (at 0.60 V vs. RHE, 1800s) indicated that the current density of PtCu NCs/CB was higher than that of PtCu NPs/CB and the commercial Pt/C for the entire time course (Figure 4d), further confirming the higher durability of PtCu NCs/CB in EOR. Moreover, we examined the composition–activity dependence for Pt-based nanocrystals with different Pt/Cu ratios (Figure 4e,f). The PtCu NCs (Pt₃₃Cu₆₇)/CB exhibited the highest mass activity for EOR among the obtained alloy catalysts, showing the activity order of PtCu NCs (Pt₃₃Cu₆₇)/CB > Pt₂₃Cu₇₇/CB > Pt₄₉Cu₅₁/CB > Pt/C and PtCu NCs (Pt₃₃Cu₆₇)/CB > PtCu NPs (Pt₃₃Cu₆₇)/CB > Pt/C. This trend is in good accordance with the XPS, CO stripping, and in situ FTIR results, and clearly demonstrates that tuning the compositions and morphologies of Pt–Cu alloy nanocrystals could significantly enhance their EOR performances. As a result, the as-prepared PtCu NCs/CB showed higher EOR activity than the most reported Pt-based catalysts (Table S2), indicating that alloying with earth-abundant Cu could not only promote the activity and durability of Pt-based EOR catalysts, but also reduce the cost of the precious Pt metals.

3. Materials and Methods

3.1. Materials

Pt(acac)₂ (98%, Alfa Aesar, Ward Hill, MA, USA), CuSO₄ (99.95%, Aladdin, Shanghai, China), OAm (80–90%, Aladdin, Shanghai, China), CTAB (99%, Aladdin, Shanghai, China), KBr (99.99%, Aladdin, Shanghai, China), and KOH (99.99%, Aladdin, Shanghai, China) were used as received without further purification.

3.2. Preparation of Pt-Based Catalysts

In a typical synthesis of PtCu NCs, 12.6 mg (0.08 mmol) of CuSO₄, 28.8 mg (0.08 mmol) of CTAB, and 10 mL of OAm were loaded into a 50 mL three-necked flask. The system was evacuated for 3 min at room temperature and then purged with N₂, and it was repeated three times. Then the solution was heated to 120 °C, kept for 30 min, and then further heated to 250 °C under N₂ atmosphere. Then Pt stock solution, which was made by dissolving 16.0 mg (0.04 mmol) of Pt(acac)₂ in 2 mL of OAm at 60 °C, was rapidly injected into the hot solution. The reaction mixture was kept at 250 °C for another 30 min and gradually turned black. Finally, the reaction mixture was cooled down to room temperature naturally, centrifuged, and washed by cyclohexane and ethanol three times. The as-obtained PtCu NCs were redispersed in cyclohexane to form a stable colloidal dispersion. The synthesis of PtCu NPs followed similar procedures except that 14.4 mg (0.04 mmol) of CTAB was added. The Pt₄₉Cu₅₁ NPs and Pt₂₃Cu₇₇ NPs were prepared under similar conditions, except that CuSO₄ was added with 6.3 mg (0.04 mmol), and 18.9 mg (0.12 mmol), respectively.

3.3. Loading Catalysts on Carbon Supports

The high surface area carbon black (CB, Vulcan XC-72, Carbot Co., Boston, MA, USA) was dispersed in cyclohexane and then sonicated for 30 min. The as-prepared colloidal dispersion of PtCu NCs was added dropwise to the carbon cyclohexane dispersion. The mixture was further sonicated for 30 min and stirred for another 120 min. The resulting CB-supported catalyst was collected by filtration and then dried in the oven. It was further calcinated in the air at 200 °C for 5 h to remove the surfactant. The mass loading of metals in the CB-supported catalysts was measured by ICP-OES.

3.4. Physicochemical Characterization

The XRD patterns were recorded on a Rigaku D/Max 2500 VB2+/PC X-ray powder diffractometer (Tokyo, Japan) with a Cu K_α radiation (λ = 0.154 nm) source at a scan rate of 5°·min⁻¹. The TEM imaging was conducted with a JEOL JEM-2100 (Akishima, Japan) transmission electron microscope. EDS, HRTEM, HAADF-STEM, and mapping/line scanning were performed on a Talos F200X transmission electron microscope (Waltham, MA, USA). The XPS spectra were recorded on a Thermo Fisher ESCALAB 250 Xi XPS system (Waltham, MA, USA) with a monochromatic Al K_α X-ray source. ICP-OES studies were performed on an Agilent ICP-OES 720 spectrometer (Santa Clara, CA, USA).

In situ FTIR spectroscopy experiments were conducted on a Nicolet iS50 FTIR (Waltham, MA, USA) spectrometer equipped with an MCT detector cooled with liquid nitrogen. A calcium fluoride window and an in situ EC-IR thin cell were applied in the in situ FTIR spectroscopy test. The working electrode was a glassy carbon electrode (10 mm in diameter) with a Pt-based catalyst. SCE and Pt wires were used as the reference electrode and counter electrode, respectively. The working electrode was pressed onto the calcium fluoride window with a gap less than 10 μm when collecting IR spectrum. The electrolyte was a mixture of 1.0 M KOH and 1.0 M ethanol. The spectral resolution was set to 8 cm⁻¹, and 64 interferograms were co-added for each spectrum. A reference spectrum was acquired at 0.1 V vs. RHE. The sample spectra were recorded during the chronopotentiometry tests with the anodic potential increased from 0.3 to 1.1 V vs. RHE in the step of 0.1 V.

3.5. Electrochemical Characterization

Electrochemical measurements were conducted on an electrochemical workstation (CHI 760E, Shanghai, China) with a standard three-electrode system at room temperature. A saturated calomel electrode (SCE) was used as the reference, and a platinum plate (1 × 1 cm²) was used as the counter electrode. The catalyst ink was prepared by dispersing 1 mg of the CB-supported catalyst powder in a mixture of ethanol (700 μL), water (290 μL), and Nafion dispersion (5 wt%, 10 μL). A 10 μL volume of the catalyst ink was dropped onto glassy carbon electrodes (5 mm in diameter, PINE instruments) and used as the working electrode.

4. Conclusions

In summary, this study reports the synthesis of Pt–Cu alloy catalysts with delicate composition and morphology control. Notably, PtCu NCs/CB with appropriate Pt/Cu atomic ratios exhibited boosted activity and stability towards the oxidation of ethanol due to the regulated surface geometric and electronic structures. CO-stripping experiments and in situ FTIR studies revealed that alloying with Cu could prevent the CO poisoning and promote the selectivity for CO₂ as the final product of ethanol oxidation. Therefore, the strategy of introducing earth-abundant Cu as a promoter in Pt-based catalysts could enhance the adsorption of C₂H₅OH, accelerate the subsequent oxidation of key intermediates, prevent the CO-poisoning, enhance catalyst stability, and reduce the catalyst costs simultaneously. Our study may find more applications in promoting the performance of electro-catalysts while reducing their costs.