Synthesis and characterization of Pd–Ni core–shell nanocatalysts for alkaline glucose electrooxidation

Abstract

Pd_shellNi_core catalyst decorated carboxylated multi-walled carbon nanotubes (Pd–Ni/C) are synthesized using a two-stage polyol method. The nano-sized Pd–Ni/C catalysts have a metal particle size range of 4.7 to 6.6 nm. The Pd_shellNi_core nanoparticles improve the electrocatalytic activity and durability of glucose oxidation reactions (GORs). X-ray diffraction (XRD), high resolution transmission electron microscopy (HR-TEM), scanning transmission electron microscopy (STEM) and scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM-EDS) are used to characterize the crystalline structure, particle formation, crystalline nature and elemental distribution, respectively. Cyclic voltammetry (CV), Tafel analysis, chronoamperometry (CA) and multi-cycle cyclic voltammetry (mCV) are used to determine the electrochemical properties of the Pd–Ni/C catalysts. The results indicated that Pd–Ni/C (1 : 0.06) exhibits the highest electrochemical surface area (ECSA) of 78.0 m² g⁻¹ which is 4.5 times higher than that of Pd/C and as well as having a 1.5-fold higher GOR current density of 21.2 mA cm⁻². The stability of Pd–Ni/C toward GOR is also significantly enhanced according to the results of the poisoning rate study and 200 cycling CV test. The highest Pd–Ni/C (1 : 0.06) catalyzed GOR current density of 34.2 mA cm⁻² is attained in 0.5 M glucose and 1.0 M NaOH alkaline medium.

---

Introduction

Direct glucose fuel cells (DGFCs) are one of the promising alternative power sources for use in the future because glucose, a six carbon sugar, is renewable, abundant and non-toxic in nature. The theoretical energy density of a complete glucose oxidation reaction (GOR) is 2870 kJ mol⁻¹ and the catalytic process can generate up to 24 electrons for one molecule of glucose.^1–3 DGFCs with high power output have been developed by utilizing noble metals as catalysts and are intended to power electronic devices.^4–6 In recent years, extensive investigations of noble metal catalyzed GORs have been carried out to generate more electricity. It is reported that platinum (Pt) based catalysts are widely used in DGFCs due to their high glucose catalysis activities.^7–10 However, the main obstacles to enhancing GOR performances of Pt-based catalysts are the high cost, low anti-poisoning ability and poor long-term durability. For this reason, there are many studies focusing on finding an ideal catalyst to improve the GOR performance.^5,11–13 Therefore, alternative approaches to developing non-Pt catalysts with high catalytic activities have been widely investigated.

Recently, palladium (Pd) has been studied for methanol and glucose oxidation in alkaline media and provides effective improvement of electrocatalytic performances.^12,14,15 However, the catalytic activities of Pd monometallic catalysts are still not sufficient to reach a notably higher GOR current density. It is known that Pd-based bimetallic catalysts can significantly enhance the activity in alkaline glucose electrooxidation including palladium–gold (Pd–Au) and palladium–rhodium (Pd–Rh).^4,12 The improvement of the glucose catalysis performance is ascribed to the increase of specific active surface sites, the bifunctional mechanism and electronic effects. In this regard, Pd-based bimetallic catalysts are emerging as potential non-Pt catalysts for alkaline GORs.

Moreover, the cost of Pd is also a critical hurdle to the use of Pd-based catalysts in commercial applications. Hence, many researchers have paid more attention to developing Pd-based core–shell catalysts to lower Pd loading.^16–18 Core–shell structures are one of the promising ways to reduce the cost of metal catalysts by replacing the core metal with other cheaper materials. In addition, the electrocatalytic activity and stability of the shell metal can be improved through synergistic effects with the core metal.¹⁶ It has been reported that a great deal of studies have been focused on finding a lower cost metal material for use as the core metal (e.g., nickel, cobalt and silver, etc.).^16,18,19 Among the commonly used core metals, nickel (Ni) shows a remarkable improvement of the catalytic performance for small organic molecules.²⁰ Therefore, Ni would be an ideal candidate for the core metal in Pd-based core–shell electrocatalysts. The prepared catalysts with Pd as the shell and low-cost Ni as the core are synthesized using a facile process with a two-stage polyol method.

It is also known that carbon nanosupports are important for providing nanosized metallic particles with a high catalyst dispersion. In this respect, carbon nanotubes (CNTs) are known as a promising carbon support because of their high specific surface area, electronic conductivity, chemical stability and interaction with metal catalysts.^14,21,22 Recently, it has been demonstrated that functionalized multi-walled carbon nanotubes (MWCNTs) can provide sufficient binding sites and improve the dispersion of decorated metal nanoparticles. According to our earlier study, carboxylated multi-walled carbon nanotubes (cMWCNTs) have more appropriate surface properties for the effective dispersion of Pd nanoparticles.¹⁴ In this work, cMWCNTs are chosen as a carbon support material for Pd_shellNi_core catalysts in the alkaline GOR investigation.

The Pd–Ni/C catalysts with a core–shell structure are synthesized using a two-stage polyol method and the synthesis method is illustrated in Fig. 1.²³ To the best of our knowledge, this is the first study using a carbon supported Pd_shellNi_core anode catalyst for alkaline glucose electrooxidation. The physicochemical properties of the Pd–Ni/C catalysts with different precursor ratios of Pd to Ni are investigated using XRD, HR-TEM, STEM and SEM-EDS. The electrocatalytic performances in terms of the ECSA, GOR current density, Tafel plot, pensioning rate, long-term cycling stability, activation energy and concentration effects are studied using CV and CA methods through the alkaline GOR. Finally, the optimal operation conditions for Pd–Ni/C (1 : 0.06) are determined in this study.

Fig. 1 Illustration for the two-stage polyol synthesis method of the Pd–Ni/C core–shell catalysts.

Experimental

Materials and chemicals

All chemicals used are analytical grade without any further purification. Palladium chloride dihydrate (PdCl₂·2H₂O) and nickel chloride hexahydrate (NiCl₂·6H₂O) were used as the precursors. Ethylene glycol (EG) was used as the reducing agent and stabilizer for the two-stage polyol method. Sodium hydroxide (NaOH) and D-glucose were respectively used as the electrolyte and the fuel in the electrochemical investigations. All of the chemicals were purchased from Sigma-Aldrich. The conventional carboxylated MWCNTs (carboxylation degree ca. 5%) were used as a metal catalyst carbon support and purchased from a local company (Golden Innovation, Taipei, Taiwan). Five percent Nafion® dispersion DE-520 (DuPont, Wilmington, Delaware, USA) was used to prepare the catalyst suspension. All aqueous solutions were prepared using double distilled water (18.2 MΩ cm).

Synthesis of the Pd–Ni/C core–shell catalysts using a two-stage polyol method

cMWCNT supported Pd_shellNi_core catalysts Pd–Ni/C (1 : x) with different precursor weight ratios of Pd to Ni (x = 0.33, 0.14, 0.06, 0.03 and 0.02) were prepared using a two-stage polyol method as follows: a Ni/C dispersion was prepared by adding 100.0 mg of cMWCNTs and the corresponding amount of NiCl₂·6H₂O into 50 mL of ethylene glycol solution. When a homogeneous mixture was obtained, the pH of the mixture was adjusted to 11 by drop-wise addition of 0.5 mol L⁻¹ NaOH in EG. The prepared Ni precursor and cMWNCT mixture was cured at 140 °C with gentle stirring for 2 h and then cooled to room temperature to form cMWCNT supported Ni_core nanoparticles (Ni/C). For the preparation of the Pd_shell coated Ni_core catalysts, the corresponding amount of PdCl₂·2H₂O was mixed with the Ni/C under gentle stirring for 30 min. The prepared mixture was cured at 140 °C with gentle stirring for 2 h and then cooled to room temperature to form cMWCNT supported Pd_shellNi_core nanoparticles (Pd–Ni/C). The resulting slurry was filtered and washed with de-ionized water and ethanol five times. Finally, the Pd–Ni/C catalysts were obtained by drying at 105 °C for 12 h. The theoretical Pd and Ni precursor contents of the Ni–Pd/C catalysts were targeted to be 20.0 mg. For comparison, cMWCNT supported Pd without the Ni_core (Pd/C) and cMWCNT supported Ni (Ni/C) were prepared using the same method.

The preparation of an anode electrode for the electrochemical investigations proceeded as follows: a standard glassy carbon (GC) electrode (d = 3.0 mm, A = 0.07 cm²) was polished with α-Al₂O₃ powder. The polished GC electrode was washed using double distilled water and then dried for 10 min at room temperature before use. A Pd–Ni/C slurry was prepared by dispersing 1.0 mg of catalyst into a 1.0 mL ternary mixture of de-ionized water, ethanol, and 5 wt% Nafion® (10 : 10 : 1, v/v). Afterward, 10 μL of the catalyst slurry was quantitatively drop-coated onto the GC electrode surface using a micropipette and dried under a 50 W halogen lamp for 15 min. The obtained Pd–Ni/C thin film electrodes were prepared for further electrochemical characterization.

Physicochemical characterization of the Pd–Ni/C catalysts

The X-ray diffraction (XRD) patterns were recorded using CuKα radiation (40 kV, 40 mA) and were used to investigate the elemental composition, crystal structure and particle size of the metal nanoparticles, using an X’Pert PRO X-ray diffractometer (PANalytical, Westborough, Massachusetts, USA). The XRD characterization was performed within the angle (2θ) range from 10° to 90°, at a scanning rate of 0.02° s⁻¹ and an angular resolution of 0.02° of the 2θ scan. Field-emission transmission electron microscopy (FE-TEM) and scanning transmission electron microscopy (STEM) with an energy-dispersive detector (EDS) were employed to investigate the dispersion morphology, size of the metal particles, crystalline nature and elemental distribution of the Pd–Ni/C catalyst using a JEM-2100F transmission electron microscope with an accelerating voltage of 200 kV (JEOL, Akishima, Tokyo, Japan). The surface morphology and elemental analysis of the Pd–Ni/C catalysts were investigated using a JSM-5310 scanning electron microscope and an energy dispersive spectrometer (SEM-EDS) (JEOL, Akishima, Tokyo, Japan).

Electrochemical characterization of the Pd–Ni/C catalyst catalyzed GOR

A standard three-electrode cell, connected to an electrochemical analyzer, was used to investigate the electrocatalytic behavior of the Pd–Ni/C catalysts. The catalyst-coated GC electrode was used as a working electrode. The reference and counter electrodes were a Hg/HgO electrode (0.1 mol L⁻¹ NaOH solution) and a platinum wire electrode, respectively. The electrochemical measurements were performed using a programmable potentiostat-galvanostat PGSTAT30 (Metrohm Autolab, Netherlands). The electrochemical surface area (ECSA) and the GOR eletrocatalytic activity of the Pd–Ni/C catalysts were evaluated using cyclic voltammetry (CV) at a scan rate of 50 mV s⁻¹. The Tafel polarization analysis was conducted using linear sweep voltammetry (LSV) at 1 mV s⁻¹ in order to eliminate the mass transfer effect. The chronoamperometry (CA) measurements were performed at a constant potential of +0.1 V for 1000 s and the multi-cycle cyclic voltammetry (mCV) stability was studied using CV at 100 mV s⁻¹ for 500 cycles. All electrochemical measurements were performed in a deoxygenated test solution. Pd–Ni/C (1 : 0.06) was prepared for further study at different operating temperatures to obtain the activation energy value. The effects of the concentrations of glucose and NaOH were evaluated to find optimal operating conditions for attaining a higher GOR current density. The obtained currents were normalized to the geometrical area of the working GC electrode (0.07 cm²).

Results and discussion

XRD patterns of the Pd–Ni/C catalysts

XRD was used to reveal the crystal phase and characterize the atomic structures of the Pd–Ni/C catalysts, as shown in Fig. 2A. The first diffraction peak at 25° is the graphite (0 0 2) facet of the cMWCNTs (JCPDS card no.: 25-0284) for all of the Pd–Ni/C catalysts. The resultant XRD pattern of Pd/C shows that there are four main diffraction peaks, at 38°, 45°, 67° and 79°, which are attributed to the face-centered cubic (fcc) phase of Pd (1 1 1), (2 0 0), (2 2 0) and (3 1 1) (JCPDS card no.: 05-0681), respectively.²⁰ The diffraction peaks of Ni/C at 40° and 47° correspond to the face-centered cubic (fcc) phase of Ni (1 1 1) and Ni (2 0 0) (JCPDS card no.: 04-0850), respectively.¹⁶ In addition, the peaks of Ni (1 1 1) and Ni (2 0 0) have not obviously been found for Pd_shellNi_core, this is possibly because the existing Ni atom amounts are too low and the Ni_core nanoparticles are coated by Pd_shell catalysts.

Fig. 2 (A) XRD patterns of the Pd–Ni/C catalysts: (a) Pd/C, (b) Pd–Ni/C (1 : 0.02), (c) Pd–Ni/C (1 : 0.03), (d) Pd–Ni/C (1 : 0.06), (e) Pd–Ni/C (1 : 0.14), (f) Pd–Ni/C (1 : 0.33) and (g) Ni/C. (B) The enlarged XRD patterns of: (a) Pd/C and (b) Pd–Ni/C (1 : 0.06).

In order to determine whether Ni elements exist in the Pd–Ni/C catalysts, atomic characterization was performed using SEM-EDS measurements. The EDS results show that Ni elements are present in the Pd–Ni/C catalysts, as shown in Fig. S1.† Furthermore, the average crystallite sizes of the Pd_shellNi_core nanoparticles were estimated using the diffraction peak of Pd (2 2 0) at 65° through the Debye-Scherrer equation, as written in eqn (1).

where d is the average particle size (nm), FWHM is the full width at half-maximum of the peak expressed in radian, λ is the corresponding wavelength of the X-rays used (1.54 Å), θ is the angle of the maximum of the peak patterns (rad) and κ is a numerical constant of about 0.9. The estimated Pd crystallite sizes for Pd/C, Pd–Ni/C (1 : 0.02), Pd–Ni/C (1 : 0.03), Pd–Ni/C (1 : 0.06), Pd–Ni/C (1 : 0.14) and Pd–Ni/C (1 : 0.33) are respectively 5.7, 6.0, 6.7, 6.7, 4.9 and 5.3 nm, as shown in Table 1. The diffraction peaks of Pd (2 0 0) and Pd (2 2 0) for Pd–Ni/C (1 : 0.06) both shift to higher 2θ values compared to Pd/C, as shown in Fig. 2B. This means that insertion of Ni atoms into the Pd crystal lattice has occurred and that Pd–Ni might form a bimetallic alloy to some degree.^16,17 According to the HR-TEM investigation (Fig. 3), most of the Pd–Ni/C bimetallic catalysts exist in the core–shell structure rather than a simple alloy composite.

Table 1 Physicochemical characterization of the Pd–Ni/C catalysts from TEM and XRD measurements

Catalyst	Pd/C	Pd–Ni/C (1 : 0.02)	Pd–Ni/C (1 : 0.03)	Pd–Ni/C (1 : 0.06)	Pd–Ni/C (1 : 0.14)	Pd–Ni/C (1 : 0.33)
Crystallite size/nm (XRD)	5.7	6.0	6.7	6.7	4.9	5.3
Particle size/nm (TEM)	4.7 ± 0.25	4.6 ± 0.29	4.2 ± 0.26	4.1 ± 0.21	5.4 ± 0.35	6.6 ± 0.70

---

Fig. 3 (A) HR-TEM morphological characteristics of the Pd–Ni/C catalyst; (B) line profile of the elemental compositions obtained using EDS and the inset is the corresponding HAADF-STEM image. A larger Pd–Ni nanoparticle was chosen to attain high-quality image resolution.

The corresponding particle size, dispersion and surface morphology of the Pd_coreNi_shell nanoparticles were characterized using TEM measurements, as shown in Fig. S2.† As a result of the particle size analysis in Table 1, it was found that the metal particles become smaller as the Pd core metal is replaced by certain amounts of Ni. It can be noted that Pd–Ni/C (1 : 0.06) exhibits the smallest particle size of 4.1 nm. In addition, the Pd_shellNi_core nanocatalysts show a better metal particle dispersion on the cMWCNTs compared to Pd/C. This result means that Ni_core not only increases the utilization of the Pd catalyst but also improves the dispersion of the metal nanoparticles.

In order to identify the structure and understand the elemental distribution of the Pd_shellNi_core, the nanostructure of an individual metal composite of Pd–Ni/C (1 : 0.06) was characterized using HR-TEM and HAADF-STEM with EDS, as shown in Fig. 3A and B, respectively. Fig. 3A shows that the inter-layer distances corresponding to the (1 1 1) lattice plane for Pd_shell and Ni_core of a single crystalline Pd_shellNi_core particle are 0.22 and 0.20 nm, respectively. In Fig. 3B, the EDS cross-sectional compositional line-profile proves that the synthesized Pd–Ni/C has a Ni core and a Pd outside shell.

ECSA estimation and GOR performance investigation for the Pd–Ni/C catalysts

The voltammetric curves of the prepared catalysts, scanned in 0.5 mol L⁻¹ NaOH solution, were used to estimate the ECSA values, as shown in Fig. S3.†^4,14 The ECSA data of the Pd–Ni catalysts are compared, along with other GOR performance factors, in Table 2 and have a range of 25.8 to 78.0 m² g⁻¹. The results indicate that Pd/C has the lowest ECSA value of 17.3 m² g⁻¹ and Pd–Ni/C (1 : 0.06) exhibits the highest ECSA value of 78.0 m² g⁻¹. This can prove that the core–shell structure obviously increases the electroactive surface area of the Pd–Ni/C catalysts.

Table 2 Comparison between the electrocatalytic GOR properties of the Pd–Ni/C catalysts investigated using the cyclic voltammograms

Catalyst	Onset potential^a (V vs. Hg/HgO)	Anodic peak (If)		Cathodic peak (Ib)		ECSA (m^2 g^−1)
		Peak potential (V vs. Hg/HgO)	Peak current (mA cm^−2)	Peak potential (V vs. Hg/HgO)	Peak current (mA cm^−2)
a Onset potential is defined as the potential at which 0.1 mA cm^−2 current density is reached.
Pd/C	−0.53	+0.06	14.2	−0.27	13.2	17.3
Pd–Ni/C (1 : 0.02)	−0.53	+0.08	16.7	−0.08	15.9	25.8
Pd–Ni/C (1 : 0.03)	−0.56	+0.01	18.8	−0.26	17.8	35.4
Pd–Ni/C (1 : 0.06)	−0.63	+0.02	21.2	−0.25	20.6	78.0
Pd–Ni/C (1 : 0.14)	−0.56	+0.06	13.8	−0.25	12.5	70.7
Pd–Ni/C (1 : 0.33)	−0.53	−0.04	3.3	−0.31	2.5	47.5

---

Furthermore, the Pd–Ni/C catalyzed GOR was investigated in 0.5 mol L⁻¹ NaOH solution containing 0.5 mol L⁻¹ glucose using CV measurements with a potential range of −0.8 V to +0.8 V. In Fig. 4A, the forward and backward scans of the Pd/C catalyzed GOR both show a GOR peak and the corresponding peak current densities are respectively expressed as I_f and I_b in Table 2. In the forward scans, the GOR current density is observed at an onset potential of ca. −0.5 V for the Pd–Ni/C catalysts (Fig. 4B). It is known that the onset of the GOR peak current density indicates the chemisorption of glucose molecules to form adsorbed intermediates and to start releasing electrons. Decline of the GOR current density is observed after the peak potential due to the accumulation of the intermediates on the active sites. Consequently, the coverage of the Pd–Ni/C electroactive sites will inhibit the electrooxidation of glucose. The inset of Fig. 4B shows that Pd–Ni/C (1 : 0.06) has the highest current density of 21.2 mA cm², which is 1.5-fold higher than that for Pd/C (14.2 mA cm²). This indicates that the Pd–Ni/C catalysts can enhance the GOR current density due to their higher ECSA.

Fig. 4 Cyclic voltammograms of (A) Pd/C and (B) the forward scan curves for the Pd–Ni/C catalysts (a) Pd/C, (b) Pd–Ni/C (1 : 0.02), (c) Pd–Ni/C (1 : 0.03), (d) Pd–Ni/C (1 : 0.06), (e) Pd–Ni/C (1 : 0.14), (f) Pd–Ni/C (1 : 0.33) and (g) Ni/C in 0.5 mol L⁻¹ NaOH solution containing 0.5 mol L⁻¹ glucose at a scan rate of 50 mV s⁻¹ at room temperature.

Kinetic properties through Tafel analysis of the Pd–Ni/C catalysts

To study the kinetic properties of the Pd–Ni/C catalysts, the Tafel slopes for the alkaline GORs were determined using slow-scan LSV measurements (at 1 mV s⁻¹) and were analyzed as shown in Fig. 5. The rather low scan rate was used in order to eliminate the mass transfer effect. The polarization curves depicted in Fig. 5 show two linear regions that are distinguished by an over potential (η_IR-free) of ca. +0.35 V.

Fig. 5 Tafel plots with linear polarization curves at (A) region I and (B) region II for the Pd–Ni/C catalysts: (a) Pd/C, (b) Pd–Ni/C (1 : 0.02), (c) Pd–Ni/C (1 : 0.03), (d) Pd–Ni/C (1 : 0.06), (e) Pd–Ni/C (1 : 0.14) and (f) Pd–Ni/C (1 : 0.33) for the GOR in 0.5 mol L⁻¹ NaOH solution containing 0.5 mol L⁻¹ glucose at a scan rate of 1 mV s⁻¹ at room temperature.

The Tafel plots at the lower potential (region I) are ascribed to the adsorption of hydroxyl groups and for those at the higher potential (region II) the formation of an oxide layer on the catalyst surface contributes, as shown in Fig. 5A and B, respectively.²⁴ The Tafel slopes for region I which were determined for Pd/C, Pd–Ni/C (1 : 0.02), Pd–Ni/C (1 : 0.03), Pd–Ni/C (1 : 0.06), Pd–Ni/C (1 : 0.14) and Pd–Ni/C (1 : 0.33) are 104, 105, 113, 128, 133 and 116 mV, respectively. In addition, those for the Tafel plots in region II are 126, 127, 145, 192, 153 and 99 mV, respectively. For the Pd_shellNi_core catalysts, the Tafel slopes are associated with the specific values due to the properties of the Pd_shell with the different amounts of underlying Ni_core. It can be noted that the Tafel slopes gradually changed to higher values when the core metal Pd was replaced with Ni metal. The corresponding exchange current densities were also evaluated by extrapolating the Tafel plots until the over potential was equal to zero. The exchange current densities for Pd/C, Pd–Ni/C (1 : 0.02), Pd–Ni/C (1 : 0.03), Pd–Ni/C (1 : 0.06), Pd–Ni/C (1 : 0.14) and Pd–Ni/C (1 : 0.33) are 1.33 × 10⁻⁴, 1.64 × 10⁻⁴, 3.80 × 10⁻⁴, 8.81 × 10⁻⁴, 1.05 × 10⁻³ and 1.47 × 10⁻⁴ mA cm⁻², respectively. The results show that Pd–Ni/C (1 : 0.14) provides the highest exchange current density of 1.05 × 10⁻³ mA cm⁻², which is 7.9 times higher than that of Pd/C. Furthermore, the corresponding activation energy was estimated using an Arrhenius plot and calculated from the GOR peak current density, as shown in Fig. S4.†²⁵ The E_a values for Pd/C and Pd–Ni/C (1 : 0.06) are 23.1 and 16.9 kJ mol⁻¹, respectively, which means that the Pd_shellNi_core also has improved kinetic properties in the alkaline GOR.

Poisoning rate evaluation of the Pd–Ni/C catalysts by CA measurements

To understand the catalyst stability toward GORs, the poisoning rates were studied using a chronoamperometric method. The steady-state current density at a constant exerted potential of +0.1 V was plotted against time for 1000 s at 25 °C and the results are shown in Fig. 6A. The poisoning rate (δ) was calculated by measuring the decay of the current density over a time interval at a fixed voltage of +0.1 V for 500 s and the adopted equation is defined in eqn (2).^12,14

δ = (100/I₀)(dI/dt), t > 500 s (2)

where (dI/dt) is the slope of the linear current decay after 500 s and I₀ is the current density at the start of the polarization back-extrapolated from the linear current decay.

Fig. 6 (A) Current–time curves from chronoamperometric measurements and (B) the change of the poisoning rate and the current density at 1000 s of the Pd–Ni/C catalysts for (a) Pd/C, (b) Pd–Ni/C (1 : 0.02), (c) Pd–Ni/C (1 : 0.03), (d) Pd–Ni/C (1 : 0.06), (e) Pd–Ni/C (1 : 0.14) and (f) Pd–Ni/C (1 : 0.33) in 0.5 mol L⁻¹ NaOH solution containing 0.5 mol L⁻¹ glucose at +0.1 V at room temperature.

There is an initial rapid decrease in the current density with time which is observed for all catalysts. This phenomenon is ascribed to the accumulation of adsorbed species on the catalyst surfaces, as shown in Fig. 6A. Afterward, the GOR current density declines slowly and reaches a steady state after 500 s. It can be found from the results that Pd–Ni/C (1 : 0.14) has the lowest poisoning rate of 1.0 × 10⁻² % s⁻¹, as shown in Fig. 6B. The poisoning rates were also calculated as 1.7 × 10⁻², 1.6 × 10⁻², 1.2 × 10⁻², 1.1 × 10⁻² and 1.3 × 10⁻² % s⁻¹ for Pd/C, Pd–Ni/C (1 : 0.02), Pd–Ni/C (1 : 0.03), Pd–Ni/C (1 : 0.06) and Pd–Ni/C (1 : 0.33), respectively. In addition, Pd–Ni/C (1 : 0.06) shows the highest steady-state current density of 3.95 mA cm⁻² at 1000 s. The corresponding steady-state current densities of the Pd–Ni/C catalysts have the following order: Pd–Ni/C (1 : 0.06) > Pd–Ni/C (1 : 0.03) > Pd–Ni/C (1 : 0.14) > Pd–Ni/C (1 : 0.02) > Pd/C > Pd–Ni/C (1 : 0.33). Among all the prepared catalysts, Pd–Ni/C (1 : 0.06) has a lower poisoning rate and the highest steady-state current density according to the chronoamperometry measurements. To further study the long-term durability of Pd–Ni/C (1 : 0.06), a mCV study was performed in alkaline glucose medium. The percentage of GOR catalytic activity retained for Pd/C and Pd–Ni/C (1 : 0.06) after 500 CV cycles is 28 and 38%, respectively, as shown in Fig. S5.† The results indicate that Pd–Ni/C (1 : 0.06) has a better poison tolerance, as well as long-term durability for the alkaline GOR.

The optimal operating conditions for Pd–Ni/C (1 : 0.06)

To find the optimal conditions in terms of electrolyte and glucose concentration for reaching a higher GOR current density, Pd–Ni/C (1 : 0.06) catalyzed GORs with different glucose and NaOH concentrations are evaluated here.^14,22 The glucose concentrations are in the range between 0 mol L⁻¹ and 1.0 mol L⁻¹ with a fixed NaOH concentration of 0.5 mol L⁻¹, as shown in Fig. 7A. It can be seen that the GOR current density monotonously increases to 24.1 mA cm⁻² while the glucose concentration increases to 0.5 mol L⁻¹, as shown in the inset of Fig. 7A. This can be explained by the dominant glucose molecules becoming more concentrated near the surface of Pd–Ni/C (1 : 0.06) by the increasing glucose concentration. However, the GOR current densities will inversely decrease when the glucose concentrations are higher than 0.5 mol L⁻¹. This is because too high a glucose concentration (>0.5 mol L⁻¹) inhibits the adsorption of hydroxyl ions on the surface of Pd_shellNi_core and GOR intermediates occupy the active sites to lower the GOR performance. Hence there is a maximum GOR current density of 24.1 mA cm⁻² at an optimal glucose concentration of 0.5 mol L⁻¹.

Fig. 7 Concentration effects on Pd–Ni/C (1 : 0.06) at (A) different glucose concentrations (0 mol L⁻¹–1 mol L⁻¹) in 0.5 mol L⁻¹ NaOH solution and (B) different NaOH solution concentrations (0 mol L⁻¹–1.5 mmol L⁻¹) with 0.5 mol L⁻¹ glucose at a scan rate of 50 mV s⁻¹ at room temperature.

The electrocatalytic performances of Pd–Ni/C (1 : 0.06) with different NaOH concentrations were also evaluated at a fixed glucose concentration of 0.5 mol L⁻¹, as shown in Fig. 7B. The GOR current density monotonously increases when the NaOH concentration increases from 0.25 mol L⁻¹ to 1.0 mol L⁻¹, as shown in the inset of Fig. 7B. When the NaOH concentrations are higher than 1.0 mol L⁻¹, the current densities will inversely decline. It is noted that an appropriate NaOH concentration can give faster kinetics for the GOR but that a too high NaOH concentration will lead to decrease the GOR current density. It can be found that there is an optimal NaOH concentration (1.0 mol L⁻¹) to obtain the highest Pd–Ni/C (1 : 0.06) catalyzed GOR current density of 34.2 mA cm⁻². Moreover, the over potential values gradually change to being more negative when the NaOH concentrations increase. This phenomenon proves that hydroxyl ions effectively enhance the catalytic activity. As a result of the effects of the glucose and NaOH electrolyte concentrations, the highest GOR current density of 34.2 mA cm⁻² is obtained in the 1.0 mol L⁻¹ NaOH solution with 0.5 mol L⁻¹ glucose.

Conclusions

Pd–Ni/C nanocatalysts with a core–shell structure have been successfully synthesized using a two-stage polyol method and are extensively studied in this work. The electrocatalytic activities and kinetics are investigated in alkaline GORs to seek an ideal Pd_shellNi_core catalyst with the appropriate Pd-to-Ni ratio. Among all the prepared catalysts, Pd–Ni/C (1 : 0.06) has the highest ECSA (78.0 m² g⁻¹) and GOR current density (21.2 mA cm⁻²). The obvious improvements may possibly be ascribed to the increase of active surface sites on the Pd_shellNi_core catalyst and the bifunctional effect. Pd–Ni/C (1 : 0.06) also has a lower poisoning rate (1.1 × 10⁻²% s⁻¹), higher long-term durability (88% activity remained) and lower activation energy value of 16.9 kJ mol⁻¹ compared to Pd/C. Furthermore, the highest GOR current density of 34.2 mA cm⁻² was obtained in 1.0 mol L⁻¹ NaOH solution containing 0.5 mol L⁻¹ glucose at room temperature. To sum up, the Pd–Ni/C catalyst with a core–shell structure gives a remarkably improved GOR activity and durability, which indicates that Pd–Ni/C is a promising non-Pt catalyst candidate for DGFC applications.

Acknowledgements

The authors gratefully acknowledge the research grant from the Ministry of Science and Technology (103-2923-E-002-008-MY3). The authors are grateful to Professor Kevin C.-W. Wu for technical aid with the XRD analysis. Thanks to Ms. C.-Y. Chien of the Ministry of Science and Technology (National Taiwan University) for assistance with the TEM experiments.

Notes and references

1. R. A. Bullen, T. C. Arnot, J. B. Lakeman and F. C. Walsh, Biosens. Bioelectron., 2006, 21, 2015–2045.
2. S. Kerzenmacher, J. Ducree, R. Zengerle and F. von Stetten, J. Power Sources, 2008, 182, 66–75.
3. N. Fujiwara, S. Yamazaki, Z. Siroma, T. Ioroi, H. Senoh and K. Yasuda, Electrochem. Commun., 2009, 11, 390–392.
4. L. L. Yan, A. Brouzgou, Y. Z. Meng, M. Xiao, P. Tsiakaras and S. Q. Song, Appl. Catal., B, 2014, 150, 268–274.
5. L. Li, K. Scott and E. H. Yu, J. Power Sources, 2013, 221, 1–5.
6. S. Kerzenmacher, J. Ducree, R. Zengerle and F. von Stetten, J. Power Sources, 2008, 182, 1–17.
7. D. Basu, S. Sood and S. Basu, Chem. Eng. J., 2013, 228, 867–870.
8. D. Basu and S. Basu, Int. J. Hydrogen Energy, 2012, 37, 4678–4684.
9. H. F. Cui, J. S. Ye, X. Liu, W. D. Zhang and F. S. Sheu, Nanotechnology, 2006, 17, 2334–2339.
10. L. Hussein, Y. J. Feng, N. Alonso-Vante, G. Urban and M. Krüger, Electrochim. Acta, 2011, 56, 7659–7665.
11. M. Pasta, F. La Mantia, R. Ruffo, F. Peri, C. Della Pina and C. M. Mari, J. Power Sources, 2011, 196, 1273–1278.
12. S. M. EI-Refaei, M. I. Awad, B. E. El-Anadouli and M. M. Saleh, Electrochim. Acta, 2013, 92, 460–467.
13. A. Brouzgou, L. L. Yan, S. Q. Song and P. Tsiakaras, Appl. Catal., B, 2014, 147, 481–489.
14. T. R. Ling, C. S. Li, J. J. Jow and J. F. Lee, Electrochim. Acta, 2011, 56, 1043–1050.
15. C. C. Chen, C. L. Lin and L. C. Chen, Electrochim. Acta, 2015, 152, 408–416.
16. T. Jiang, L. Yan, Y. Meng, M. Xiao, Z. Wu, P. Tsiakaras and S. Song, Appl. Catal., B, 2015, 162, 275–281.
17. M. Zhang, Z. Yan and J. Xie, Electrochim. Acta, 2012, 77, 237–243.
18. N. Kakati, J. Maiti, S. H. Lee and Y. S. Yoon, Int. J. Hydrogen Energy, 2012, 37, 19055–19064.
19. J. H. Jang, C. Pak and Y. U. Kwon, J. Power Sources, 2012, 201, 179–183.
20. M. Rashid, T. S. Jun, Y. Jung and Y. S. Kim, Sens. Actuators, B, 2015, 208, 7–13.
21. S. Y. Shen, T. S. Zhao, J. B. Xu and Y. S. Li, J. Power Sources, 2010, 195, 1001–1006.
22. C. M. Yu, M. J. Yen and L. C. Chen, Biosens. Bioelectron., 2010, 25, 2515–2521.
23. Z. X. Cai, C. C. Liu, G. H. Wu, X. M. Chen and X. Chen, Electrochim. Acta, 2013, 112, 756–762.
24. S. Alayoglu, A. U. Nilekar, M. Mavrikakis and B. Eichhorn, Nat. Mater., 2008, 7, 333–338.
25. Z. X. Liang, T. S. Zhao, J. B. Xu and L. D. Zhu, Electrochim. Acta, 2009, 54, 2203–2208.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra06331k

---