Hydrogen Oxidation Activity and Co-Tolerance of Pt-Ni Alloy Nanoparticles Prepared on Ni Substrate

Eiji Higuchi, Masahiro Ieguchi and Hiroshi Inoue

Copyright AD-TECH; licensee AZoM.com Pty Ltd.
This is an AZo Open Access Rewards System (AZo-OARS) article distributed under the terms of the AZo-OARS https://www.azom.com/oars.asp which permits unrestricted use provided the original work is properly cited but is limited to non-commercial distribution and reproduction.
AZojomo (ISSN 1833-122X) Volume 6 December 2010

Topics Covered

Abstract

Two kinds of Pt-Ni alloy nanoparticles were prepared by casting a 1 or 2 mM PtCl₆²- solution on a Ni substrate. The lower concentration of the PtCl₆²- solution led to the lower Pt content in the alloy nanoparticles. Hydrodynamic voltammograms (HVs) of the hydrogen oxidation reaction (HOR) at the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes with and without CO exposure clearly demonstrated that both electrodes have much higher tolerance to CO poisoning than the polycrystalline Pt electrode. Pt4f core level peaks in X-ray photoelectron spectra for the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni before the HOR shifted to higher binding energies, suggesting the increased d vacancy in the valence band 5d orbital of Pt. Thus, which chemical adsorption of CO on Pt atoms is weaken and CO coverage decreased. Moreover, irrespective of CO exposure, the HOR activity for the Pt-Ni(1 mM)/Ni electrode at 40°C and 60°C was higher than that of Pt-Ni(1 mM)/Ni at 25°C.

Keywords

Fuel Cells, Catalysis, Electrode Materials, Metals and Alloys

Introduction

Polymer electrolyte fuel cells (PEFCs) are at present attracting considerable interest as a primary power source for zero emission electric vehicles and residential co-generation systems due to their high energy efficiency. In PEFCs, Pt or Pt-based alloys are used as active anode electrocatalysts for the hydrogen oxidation reaction (HOR). Pt is the most active element for HOR, but it is very expensive and has lower CO tolerance. A strategy for overcoming these problems is to alloy Pt with foreign metals. A Pt-Ru alloy is the best electrocatalyst developed to date [1-3]; however, from the viewpoint of low cost, Pt-based alloys with non-precious metals such as Co, Fe, or Ni as the second element are more attractive and also have good CO tolerance (e.g., Pt-Fe, Pt-Co, and Pt-Ni) [4-11].

Spontaneous deposition is a simple and unique method that is based on a redox reaction with the aid of a potential gap as the driving force [12–19]. For example, Pt⁴⁺ ions are reduced by Cu adatoms deposited by underpotential deposition. As a result, the Cu monolayer is exchanged with the Pt monolayer. It has been reported that spontaneous deposition can also be used to prepare not only monolayers but also multilayers, and that the resultant bimetallic catalysts have high activity for reactions including HOR, O₂ reduction, and CO oxidation [18-20]. The deposition method does not require any reductant that would have to be removed from the final product, thus resulting in savings of time and cost. We have succeeded in preparing Pt-Ni alloy nanoparticles on the surface of a metallic Ni substrate with the deposition method and found that the particle size and composition of the resulting alloys could be controlled by varying the concentration of PtCl₆²- used as a precursor complex [21]. Moreover, the Pt-Ni alloy nanoparticles showed higher electrocatalytic activity in methanol oxidation and oxygen reduction reactions than Pt [21, 22]. In the present study, we prepare two kinds of Pt-Ni/Ni electrodes using different concentrations of PtCl₆²- solutions and demonstrate that these electrodes have much higher activity for HOR than the polycrystalline Pt electrode at different temperatures, even when they have been pre-exposed to CO.

Experimental

Preparation and Evaluation of Pt-Ni/Ni

A Ni column with a diameter of 5 mm and height of 5 mm was used as a substrate in the present study. The Pt-Ni alloy nanoparticles were deposited on the Ni column, followed by heat-treating at 500°C for 1 h in an H2 atmosphere according to the previous paper [21]. 1 or 2 mM potassium chloroplatinate, referred to as 1 or 2 mM PtCl₆²- solution hereafter, was used as a precursor. The deposits are referred to as Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni hereafter.

The morphology of the deposits on the Ni substrate was observed using a field-emission scanning electron microscope (SEM). Structural analysis of the deposits was carried out using an X-ray diffractometer equipped with a CuKa source (? = 0.1541 nm, 50 kV, 30 mA). The surface chemical states of the Pt-Ni/Ni electrodes were measured by X-ray photoelectron spectroscopy (XPS). The X-ray source was MgKa (1253.6 eV) operating at 8 kV and 30 mA. The base pressure of the system was 1.33 ×10^-7 Pa. The deconvolution of the measured Pt 4f spectra of each species was carried out according to Refs [23, 24]. Background removal was carried out using the Shirley baseline method [25]. For quantitative evaluation, literature values of atomic sensitivity factors (ASF) for Pt 4f and Ni 2p were used [26]. This simple relationship permits estimations of the composition of Pt and Ni.

Electrochemical Measurements

A rotating disk electrode (RDE) apparatus with a gas-tight Pyrex glass cell was used to examine the HOR activity of the Pt-Ni/Ni electrodes. Each Pt-Ni/Ni electrode was fixed on a Cu rotating rod (diameter: 5 mm) with a shrinkable tube and then used as the rotating working electrode. The side of the Ni substrate was sealed with a methyl methacrylate polymer thin film. On the other hand, Pt plate with a diameter of 0.6 mm (Pt disk electrode, BAS) as a polycrystalline Pt was used for comparison. A Pt plate and a reversible hydrogen electrode (RHE) were used as the counter and reference electrodes, respectively. The 0.05 M H₂SO₄ aqueous solution was prepared from reagent grade chemicals and Milli-Q water.

Prior to the HOR experiments, potential cycling between 0.05 and 1.2 V at 20 mV s^-1 in Ar-saturated 0.05 M H₂SO₄ aqueous solution at 25°C was performed 10 times for each electrode. The electrochemically active surface area (EASA) of Pt was evaluated based on the electric charge for hydrogen desorption in the positive-going potential scan from 0.05 to 1.2 V in cyclic voltammograms (CVs) measured at a sweep rate of 20 mV s^-1, assuming 210 µC cm^-2 for smooth polycrystalline Pt. After bubbling H2 in 0.05 M H₂SO₄ aqueous solution for 30 min, hydrodynamic voltammograms (HVs) for the HOR at the working disk electrode were recorded by sweeping the potential from 0 to 0.1 V at 1 mV s^-1 at 25 – 60°C under rotating rates of 3600 to 400 rpm. In order to investigate the CO tolerance of each Pt-Ni/Ni electrode, each working electrode was immersed into a 0.05 M H₂SO₄ aqueous solution saturated with 99.9% CO at 25°C for 60 min with keeping a potential of 0.05 V vs. RHE before recording HVs in the H-saturated 0.05 M H₂SO₄ aqueous solution.

Results and Discussions

Characterization of Deposit Prepared by Casting 1 or 2 mM PtCl₆^2- Solution on a Ni Substrate

Figure 1 shows SEM images of Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni. The morphology and size of deposits on the Ni substrate seem to depend on the concentration of PtCl₆^2- solutions as a precursor. For the Pt-Ni(1 mM)/Ni, nanoparticles with an average size of ca. 30 nm were distributed uniformly onto the Ni substrate surface. For the Pt-Ni(2 mM)/Ni, leaf-like deposits were observed in addition to nanoparticles deposited over the surface of the Ni substrate.

XRD patterns of Pt-Ni(1 mM)/Ni and Pt-Ni(2mM)/Ni are shown in Fig. 2. Typically, Pt had a face-centered cubic (fcc) structure, and the diffraction peak assigned to the (111) plane of the fcc structure was observed at 2θ = 39.8° (Fig.2 (c). Each diffraction peak in Fig. 2 (a) and (b) was shifted to higher diffraction angles than the Pt(111) peak, suggesting the formation of not pure Pt but alloys of Pt and Ni. The degree of peak shift for the Pt-Ni(2 mM)/Ni was larger than that for the Pt-Ni(1 mM)/Ni. Table 1 lists the lattice constant, alloy composition determined with Vegard’s equation [21], and crystallite size determined with Scherrer’s equation [21] for each sample. Pt content in alloys for the Pt-Ni(2 mM)/Ni was higher than that for the Pt-Ni(1 mM)/Ni, leading to lower lattice constant. The crystallite size of Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni was 8.5 nm and 9.4 nm, respectively.

Figure 1. SEM images of (a) Pt-Ni (1 mM)/Ni and (b) Pt-Ni (2 mM)/Ni.

Table 1. Lattice constant (a), composition, crystallite size (d) and surface content of Pt and Ni for Pt-Ni (1 mM) / Ni and Pt-Ni (2 mM) / Ni

Figure 2. XRD patterns of (a) Pt-Ni (1 mM)/Ni, (b) Pt-Ni (2 mM)/Ni and (c) polycrystalline Pt .

HOR Activity of Bare and CO-Adsorbed Pt-Ni (1 mM) / Ni and Pt-Ni (2 mM) / Ni Electrodes

Figure 3 shows CVs (10th cycle) of the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes in an Ar-saturated 0.05 M H₂SO₄ aqueous solution. Both CVs had a large oxidation peak at ca. 0.5 V in the 1st cycle due to the dissolution and subsequent passivation of Ni, and this peak almost disappeared in the 2nd cycle. In the 10th cycle, as shown in Fig. 3, both CVs were similar to that of a polycrystalline Pt electrode, exhibiting reversible waves assigned to the adsorption/desorption of atomic hydrogen at potentials less positive than 0.4 V vs. RHE. Based on the electric charge required for hydrogen adsorption/desorption in each CV, we evaluated the EASA of each electrode. The EASA and roughness factor were 0.41 cmcm² and 2.1 for the Pt-Ni(1 mM)/Ni electrode and 0.69 cm² and 3.5 for the Pt-Ni(2 mM)/Ni electrode. The EASA was increased with increasing concentration of the PtCl₆²- solution due to the increase in the amount of deposited Pt.

Figure 4 shows the HVs of the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes in an H₂-saturated 0.05 M H₂SO₄ solution at a sweep rate of 1 mV s^-1. In both cases, each HV was similar to that of the polycrystalline Pt electrode although the current density for the Pt-Ni/Ni electrodes at smaller overpotentials was slightly lower than that for the polycrystalline Pt electrode. Moreover, in both cases, the current due to the HOR commences at ca. 0 V vs. RHE irrespective of rotating speed and reaches the diffusion limit at approximately 0.05 V, which is similar to the polycrystalline Pt electrode.

Figure 3. CVs (tenth cycle) of (a) Pt-Ni(1 mM)/Ni and (b) Pt-Ni(2 mM)/Ni electrodes in an Ar-saturated 0.05 M H₂SO₄ aqueous solution at 25°C. Scan rate = 20 mV s^-1.

Figure 4. HVs for HOR at various rotating speeds for (a) Pt-Ni(1 mM)/Ni and (b) Pt-Ni(2 mM)/Ni electrodes in an H₂-saturated 0.05 M H₂SO₄ aqueous solution.

Figure 5 shows Levich-Koutecky plots at various potentials for each Pt-Ni/Ni electrode. The number of electrons (n) involved in the HOR was determined from a slope of each linear Levich-Koutecky plot. In both cases, the n value was approximately 2, which agrees well with the theoretical value for the HOR. We determined the kinetically controlled current density, jK, by using the equation jK = (I_L·I)/[(I_L–I)(1/S_Pt)], where I, I_L and S_Pt are the measured current, limiting diffusion current and EASA, respectively. Plots of E - log (jK), known as Tafel plots, are shown in Fig. 6. For all electrodes, a linear relationship holds in the potential range from ca. 0.02 V to 0.05 V vs. RHE. At any potential, the jK values of the Pt-Ni/Ni electrodes were slightly lower than that of the Pt electrode. Igarashi et al. have reported that under conditions free of CO, the HOR activity of Pt-Ni alloy is slightly lower than that of Pt at 26°C [27]. The Tafel slope of the polycrystalline Pt electrode was ca. 30 mV dec^–1, which is in agreement with the value reported for a pure Pt electrode at 30°C in 0.1 M HClO₄ solution [10]. The following reactions are known to occur in the HOR.

H₂ + 2M  →     2MH                (1)
H₂ + M    →     MH + H⁺ + e^-    (2)
2MH        →     M + H⁺ + e^-       (3)

Figure 5. Levich-Koutecky plots for various potentials in Fig. 4 for (a) Pt-Ni(1 mM)/Ni and (b) Pt-Ni(2 mM)/Ni electrodes

Figure 6. Tafel plots for Pt-Ni(1 mM)/Ni, Pt-Ni(2 mM)/Ni and polycrystalline Pt electrodes in an HH₂-saturated 0.05 M H₂SO₄ aqueous solution.

There are two mechanisms for the HOR, namely, the Tafel-Volmer mechanism (eqs. (1)–(3)) and Heyrovsky-Volmer mechanism (eqs. (2)–(3)). It has been shown that the HOR on Pt proceeds through the Tafel-Volmer mechanism at small overpotentials and through the Heyrovsky-Volmer mechanism at ? > 50 mV [28]. The Tafel slope of 30 mV dec^–1 for the polycrystalline Pt electrode indicates that the dissociative adsorption of H₂ on Pt, the Tafel reaction (eq. (1)), was the rate-determining step (rds). For the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes, the Tafel slope was ca. 30 mV dec^–1 in the potential rage from ca. 0.02 to 0.05 V, suggesting that the mechanism and rds of the HOR at the Pt-Ni/Ni electrodes were the same as those at the polycrystalline Pt electrode.

Figure 7. X-ray photoelectron spectra of Pt4f for (a) Pt-Ni(1 mM)/Ni and (b) Pt-Ni(2 mM)/Ni electrodes after 10 potential cycles

X-ray photoelectron spectra of Pt4f on the surface of the Pt-Ni(1 mM)/Ni, and Pt-Ni(2 mM)/Ni electrodes after the potential sweeps of 10 cycles in an Ar-saturated 0.05 M H₂SO₄ aqueous solution are shown in Fig. 7. Two Pt4f peaks in each spectrum are assigned to those of Pt0 [29], while there are no peaks assigned to Pt oxides, suggesting that the surface Pt atoms were not oxidized. Moreover, the Pt4f peaks are shifted to higher binding energies than those of the polycrystalline Pt, which has also been observed for other Pt-Ni/Ni electrodes before potential cycling. As shown in Fig. 7, the shift of the Pt4f peaks is around 0.2 eV for the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes, which strongly suggests that the electronic structure of the Pt sites on the Pt-Ni alloy nanoparticle surface with higher Ni contents is different of that of inherent Pt. Some research groups have reported that Pt skin and Pt skeleton layers formed on the Pt-based alloys had the modified electronic structures, resulting in chemical shifts to higher binding energies [30, 31]. The chemical shift indicates a lowering of the Fermi level or an increasing of d vacancy in the valence band 5d orbital of the surface Pt, which have been shown to be responsible for the improvement of electrocatalytic activity and stability against methanol oxidation [21, 22, 32].

CO Tolerance of Pt-Ni (1 mM) / Ni and Pt-Ni (2 mM) / Ni Electrodes

In order to investigate the CO tolerance of the Pt-Ni/Ni electrodes, the Pt-Ni/Ni and polycrystalline Pt electrodes were immersed in a CO-saturated 0.05 M H₂SO₄ aqueous solution before the evaluation of HOR activity. Figure 8 shows HVs at 1600 rpm in an H₂-saturated 0.05 M H₂SO₄ aqueous solution for the Pt-Ni/Ni and Pt electrodes with and without CO exposure. It is well known that CO is strongly adsorbed on Pt. As shown in Fig. 8(c), the polycrystalline Pt electrode completely lost HOR activity after CO exposure. On the other hand, the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes clearly exhibited current density due to the HOR after CO exposure, as shown in Fig. 8(a) and Fig. 8(b). These results suggest that the adsorption of CO on the surface Pt sites of the Pt-Ni nanoparticles was weakened, and the coverage with CO was less extensive. Figure 7 clearly shows that the electronic structure of the surface Pt sites of Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni was modified or the d vacancy in the valence band 5d orbital increased, which were likely responsible for the remarkable CO tolerance.

Tafel plots for the Pt-Ni/Ni and polycrystalline Pt electrodes that were pre-exposed to CO are shown in Fig. 9. For all electrodes, a linear relationship holds in the potential range from ca. 0.02 V to 0.05 V vs. RHE, as was the case for electrodes without CO exposure. The Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes had HOR activity that was much more tolerant to CO poisoning than the polycrystalline Pt electrodes, which is likely attributable to the positive Pt4f core level shift of ca. 0.2 eV.

As shown in Fig. 9, the Tafel slope of each Pt-Ni/Ni electrode after CO exposure was ca. 38 mV dec^–1, which was similar to that (30 mV dec-1) of the electrodes without CO exposure, suggesting that the mechanism and rate determining step (rds) of the HOR at the Pt-Ni/Ni electrodes were not affected by CO exposure.

HVs for the HOR at CO-Adsorbed Pt-Ni/Ni Electrode at Different Temperatures

In order to investigate the temperature dependence of CO tolerance for Pt-Ni/Ni electrodes, the Pt-Ni(1 mM)/Ni electrode was immersed in a CO-saturated 0.05 M H₂SO₄ aqueous solution before evaluating HOR activity at 25 – 60 °C. Figure 10 shows HVs at 1600 rpm in H₂-saturated 0.05 M H₂SO₄ aqueous solution at various temperatures for the Pt-Ni(1 mM)/Ni electrode with and without CO exposure. The Pt-Ni(1 mM)/Ni electrode at 40°C and 60°C clearly exhibited current density due to the HOR even after CO exposure, as shown in Fig. 10. Irrespective of CO exposure, the HOR activity for the Pt-Ni(1 mM)/Ni electrode at 40°C and 60°C was higher than that of Pt-Ni(1 mM)/Ni at 25°C. These results indicate that the adsorption of CO on the surface Pt sites of the Pt-Ni nanoparticles was weakened, and the coverage with CO was still less extensive at elevated temperature.

Figure 8. HVs at 1600 rpm for (a) Pt-Ni(1 mM)/Ni, (b) Pt-Ni(2 mM)/Ni and (c) polycrystalline Pt electrodes with and without CO exposure in an H₂-saturated 0.05 M H₂SO₄ aqueous solution.

Tafel plots for the Pt-Ni(1 mM)/Ni electrode with CO exposure are shown in Fig. 11. For all temperature, a linear relationship holds in the potential range from ca. 0.02 V to 0.05 V vs. RHE, as was the case for electrodes without CO exposure. As shown in Fig. 11, the Tafel slope of Pt-Ni/Ni electrode at all the temperatures after CO exposure was ca. 30 mV dec^–1, suggesting that the mechanism and rds of the HOR at the Pt-Ni/Ni electrodes were not affected by CO exposure.

Figure 9. Tafel plots for CO-adsorbed Pt-Ni(1 mM)/Ni, Pt-Ni(2 mM)/Ni and polycrystalline Pt electrodes in an HH₂-saturated 0.05 M H₂SO₄ aqueous solution. Rotating speed = 1600 rpm.

Figure 10. HVs at various temperatures for Pt-Ni(1 mM)/Ni electrode with and without CO exposure in H₂-saturated 0.05 M H₂SO₄ aqueous solution. Rotating speed = 1600 rpm

Figure 11. Tafel plots for the Pt-Ni(1 mM)/Ni electrode with CO exposure at various temperatures. Rotating speed = 1600 rpm

Conclusions

The results obtained in the present study are summarized as follows.

1) Pt-Ni alloy nanoparticles are prepared by casting 1 and 2 mM PtCl₆²- solution on a Ni substrate and the Pt content for the former is higher than that for the latter.

2) XRD spectra of Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni show that their diffraction peaks shift to higher diffraction angles than the (111) peak of pure Pt, suggesting the formation of not pure Pt but alloys of Pt and Ni.

3) X-ray photoelectron spectra of Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni show that Pt4f peaks shift to higher binding energies due to the increased d vacancy in the valence band 5d orbital of Pt, which weakens the chemical adsorption of CO or decreases CO coverage.

4) HVs in an H₂-saturated 0.05 M H₂SO₄ aqueous solution show that even after CO exposure, the Pt-Ni(1 mM)/Ni and Pt-Ni(2 mM)/Ni electrodes have much smaller loss of HOR activity than the polycrystalline Pt electrode, indicating clearly that the former is superior to the latter in terms of CO tolerance.

5) Irrespective of CO exposure, the HOR activity for the Pt-Ni(1 mM)/Ni electrode at 40°C and 60°C is higher than that of Pt-Ni(1 mM)/Ni at 25°C, indicating that the adsorption of CO on the surface Pt sites of the Pt-Ni nanoparticles is weakened, and the coverage with CO is still less extensive at elevated temperature.

Acknowledgment

This work was partly supported by a Grant-in-Aid for Scientific Research (B), No. 17360358, from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References

Contact Details

Eiji Higuchi, Masahiro Ieguchi and Hiroshi Inoue
Department of Applied Chemistry, Graduate School of Engineering,
Osaka Prefecture University, Sakai, Osaka 599-8531, Japan

This paper was also published in print form in "Advances in Technology of Materials and Materials Processing", 12[2] (2010) 35-42.