1. Introduction
Seed-mediated growth has been widely used to synthesize core–shell nanoparticles because the size, morphology, composition, and structure of the nanoparticles, which affect their properties, can be easily controlled [
1,
2,
3]. In particular, it has been employed to form noble metal core–shell nanocatalysts with uniform thin shells in aqueous solutions [
4,
5,
6]. These nanocatalysts have been employed in many applications; however, industrial-scale applications are difficult because of the complex and low-efficiency preparation process and their low yield [
7,
8].
To obtain core–shell nanoparticles with thin (<2 nm) and uniform shells, the seed-mediated growth method typically involves two steps: (i) first, the core material seed is synthesized, which requires a washing process to remove unreacted species (e.g., metal precursor, reducing agent, capping agent, and surfactant); (ii) second, after re-dispersing the core material in the reacting solution, the shell growth process is carried out. The washing process between the first and second steps is particularly important because the unreacted species may adversely affect the shell formation. Moreover, this process usually increases the cost of manufacturing of nanomaterials. To overcome these issues, we recently developed a direct seed-mediated growth method that allows for a precise shell control in aqueous solutions [
9,
10,
11]. In this method, the washing process after the synthesis of core material is not necessary, leading to a simple and cost-effective manufacturing procedure. For example, we were able to synthesize Pd@Pt core–shell nanocubes with a finely controlled Pt shell and showed that, compared with commercial Pt/C, the catalytic activity of the nanocubes with extremely low Pt content (0.4 at%) was 2.2 times higher than that in the electrochemical formic acid oxidation [
9]. However, the morphology of the synthesized Pd@Pt core–shell nanocubes was concave cubic, where the Pt shell did not fully cover the surface of the Pd core. Having a full Pt shell may be necessary for the catalysis of certain reactions; therefore, it is necessary to develop a technique to control the degree of Pt coverage on the Pd core.
In the anisotropic growth of nanomaterials, monomer concentration is an important factor that can lead to kinetically controlled growth. Compared with thermodynamically controlled growth, kinetic control can be achieved at low monomer concentrations because the binding energy difference of each crystal facet is maximized under such conditions [
12]. For example, in a previous study, anisotropic Pt concave nanocubes were synthesized by slowly adding aqueous NaBH
4 and Pt precursor solutions [
13]. The slow addition of these solutions promoted kinetically controlled growth rather than thermodynamically favored growth, thus leading to the formation of concave structures. When the solutions were rapidly added into the reaction mixture, truncated spherical nanoparticles were obtained. This indicates that the morphology of the Pd@Pt core–shell nanocubes could be controlled by adjusting the Pt monomer concentration.
In this study, we report on the control of the coverage degree of the Pt shell in Pd@Pt core–shell nanocubes using a direct seed-mediated growth method. Full and partial Pt shells were obtained by varying the type of reducing agent used to form the shell. A weak reducing agent (citric acid) led to a partial Pt shell, while a strong reducing agent (NaBH
4) resulted in the formation of a full Pt shell on the surface of the Pd nanocubes. The catalytic properties of the core–shell nanocubes may be different depending on whether Pd and Pt coexist on the surface (partial Pt shell) or if only Pt is present (full shell). The electrocatalytic properties of the nanocubes were evaluated in the methanol oxidation reaction (MOR), for which Pd@Pt core–shell nanocubes have a significant effect [
14,
15]. The MOR is an important reaction at the anode of a direct methanol fuel cell (DMFC), and the activity and stability of the MOR catalyst have a great influence on the performance of a DMFC. We found that in MOR, the activity and stability of the two types of nanocubes were clearly different.
2. Results
The Pd@Pt core–shell nanocubes were synthesized by a direct seed-mediated growth method, which does not require a washing process between the synthesis of the Pd core and the growth of the Pt shell. The powder X-ray diffraction (XRD) patterns of the synthesized core–shell nanocubes indicate the presence of fcc Pd (Joint Committee on Powder Diffraction Standards (JCPDS) file no. 88-2335), possibly due to the low Pt atomic ratio and/or thin Pt shell (
Figure S1). The morphology of the Pt shell on the Pd@Pt core–shell nanocubes was modulated by varying the reducing agent used for the formation of the Pt shell. Nanocubes with a partial Pt shell (P-PdPt) were synthesized using citric acid (CA) to reduce the Pt cation, as previously reported [
9,
10].
Figure 1a–c and
Figure S2a–c show transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images of P-PdPt nanocubes obtained using different amounts of Pt. It can be seen that the morphology changed from simple cubic to concave cubic as the amount of Pt increased (from 0.6 at% to 13 at%). The atomic ratio of Pd and Pt was measured by an inductively coupled plasma (ICP) spectrometer (
Table S1). Energy-dispersive X-ray spectroscopy (EDX) mapping images for these nanocubes indicate that the Pt shell grew mainly on the corners of the Pd cube, leading to the concave cubic morphology with Pd exposed on the plane of the cube (
Figure 2a,
Figures S1 and S3). When NaBH
4 was used as the reducing agent instead of CA, the nanocubes maintained their cubic shape, and their size increased from 10 to 11 nm as the amount of Pt used in the reaction increased (
Figure 1d–f and
Figure S2). EDX mapping of the nanocubes with 13 at% Pt clearly shows a full Pt shell covering the Pd surface (
Figure 2b and
Figure S3). The Pd@Pt core–shell nanocubes with a full Pt shell are referred to as F-PdPt. When the amount of Pt decreased to 0.6 at% in the nanocubes synthesized with NaBH
4, Pt was found in all the planes of the cube, indicating that the shell grew on the {100} facets of the cube (
Figure S1). TEM and EDX analyses did not show isolated Pt nanoparticles, confirming that all of the Pt precursor was used to form the Pt shell.
Previous studies on the direct seed-mediated growth of Pd@Pt core–shell nanocubes show that CA induces the growth of Pt enclosed by {110} and {111} facets on the Pd nanocube, thus leading to the formation of a concave structure [
9,
10]. In addition, the Pd nanocubes used in this work were covered by bromide ions (Br
−), and it is known that the selective chemisorption of Br
− on the {100} facets of the Pd nanocubes decreases the growth rate of the Pt shell along the <100> direction. To achieve an even growth of the Pt shell on all facets of the Pd nanocubes, we evaluated reaction conditions that involved a high monomer concentration and NaBH
4, which is a stronger reducing agent than CA. With high monomer concentrations, Pt atoms cannot selectively attach to specific facets of the Pd nanocubes; thus, Pt would evenly deposit on all facets (
Figure 3). To further understand the relation between reduction rate and growth behavior, we conducted the reaction using L-ascorbic acid (AA) as the reducing agent in the shell formation step (
Figure 4a). Concave nanocubes were obtained, indicating that a weak reducing agent cannot form fully covered Pd@Pt core–shell nanocubes. However, a mixture of isolated Pt nanoparticles and irregularly shaped Pd@Pt core–shell nanocubes with a bumped shell was obtained with high Pt and NaBH
4 concentrations. This demonstrates that a too-high monomer concentration may cause homogeneous nucleation of Pt atoms (
Figure 4b). The results indicate that monomer concentration is crucial to control the coverage degree of the Pt shell on the Pd core.
The electrochemical catalytic properties of F-PdPt and P-PdPt were investigated in the MOR.
Figure 5 shows the cyclic voltammetry curves of F-PdPt and P-PdPt in an aqueous electrolyte containing 1 M CH
3OH and 1 M KOH. Pd exhibited low MOR activity in alkaline media (
Figure S4). The oxidation current density associated with MOR was similar for both F-PdPt and P-PdPt and increased with increasing Pt content. This is because surface Pt serves as the active site of the reaction. It is worth mentioning that F-PdPt
13 and P-PdPt
13 had high MOR activity close to that of commercial Pt/C (20 wt% Pt), even with small Pt content (3.8 wt% Pt). On the basis of the Pt loading, both F-PdPt and P-PdPt showed significantly higher MOR activity than commercial Pt/C (
Figure S3). In binary PdPt MOR catalysts, the electronic effect and bifunctional mechanism enhance the catalytic properties. The former refers to changes in the electronic and/or geometric structure of Pt and Pd that favor MOR due to the interaction of adjacent Pd and Pt [
15,
16]. This effect mainly occurs in core–shell structures. In addition, during MOR, Pd is more favorable for the dissociative adsorption of H
2O than Pt because Pd–OH is formed. This is known to promote the oxidation of CO species, an intermediate adsorbed on Pt (bifunctional mechanism) [
15,
17]. The Pd@Pt core–shell structure of F-PdPt likely boosted MOR by inducing electronic effects, altering the structure of the Pt shell. In contrast, in P-PdPt, where both Pd and Pt are exposed to the electrolyte, the bifunctional mechanism would have promoted MOR catalysis. Moreover, P-PdPt has a preferential Pt(110) orientation, which favors OH adsorption and may also improve MOR catalysis in alkaline media [
18].
The chemical stability of F-PdPt and P-PdPt during MOR was tested by a potentiostatic method.
Figure 6 shows chronoamperometry curves measured at −0.2 V. A decrease in the current density was observed in all samples; at high Pt content, a high current density was maintained. However, F-PdPt had a higher current density than P-PdPt at the end of the test. This indicates that the former is more stable than the latter. The decrease in the current density occurs mainly due to catalyst poisoning by strongly adsorbed species formed during MOR [
19]. Pd@Pt core–shell nanocubes are known to possess high resistance against catalyst poisoning by reducing the binding energy of the adsorbed species [
14,
15]; this effect was also observed in F-PdPt. However, Pt(110), which is the facet present in P-PdPt, has been reported to be the weakest against catalyst poisoning among all Pt facets [
20]. This resulted in the rapid current density decrease seen in
Figure 6.
3. Materials and Methods
3.1. Materials
Polyvinylpyrrolidone (PVP, MW = 55,000), L-AA (99%), sodium tetrachloropalladate (Na2PdCl4, 98%), CA (99.5%), potassium tetrachloroplatinate (K2PtCl4, 99.99%), sodium borohydride (NaBH4, 98%), and potassium bromide (KBr, 99%) were purchased from Sigma Aldrich (St. Louis, MO, USA) and utilized without further purification.
3.2. Synthesis of Pd@Pt Core–Shell Nanocubes
Pd@Pt core–shell nanocubes were synthesized using a direct seed-mediated growth method [
9,
10]. For the synthesis of the nanocubes with a partial Pt shell (P-PdPt), 36.93 mg of PVP, 50 mg of AA, and 300 mg of KBr were dissolved in 8 mL of deionized water and heated to 80 °C for 20 min under magnetic stirring at 800 rpm. Then, 3 mL of an aqueous solution containing 57 mg of Na
2PdCl
4 was slowly added to the reaction mixture. After 3 h of heating at 80 °C, 0.1 mL of an aqueous CA solution (800 mg/mL) was added. Afterward, 1 mL of aqueous K
2PtCl
4 with different concentrations (0.88, 4.39, and 6.16 mg/mL) was rapidly injected. The solution was then heated to 80 °C for 5 h. The products were collected by centrifugation and washed three times with a deionized (DI) water and acetone mixture. The same experimental procedure was followed to synthesize Pd@Pt core–shell nanocubes with a full Pt shell (F-PdPt) except for the reducing agent; 0.1 mL of an NaBH
4 aqueous solution (1000 mg/mL) was used instead of CA.
3.3. Electrochemical Measurements
The electrochemical experiments were performed using a CHI 760E potentiostat (CH Instruments, Austin, TX, USA) at room temperature. A leak-free Ag/AgCl electrode was used as reference electrode and a Pt mesh (1 × 1 cm−2) as counter electrode. The working electrode was a glassy carbon (GC) rotating disk electrode (RDE, 5 mm diameter; Pine Research Instrumentation, Durham, NC, USA). The disk was mechanically polished before each measurement. Ethanol dispersions of F-PdPt/C (Vulcan XC-72) and P-PdPt/C were sonicated for 15 min, and then the catalyst ink was made (1 mg metal catalyst and 45 μL nafion in ethanol solution). In this solution, the concentration of the catalyst was 1.67 mg/mL. The catalyst ink was dispersed by sonication for 15 min and 4.7 μL was dropped onto GC RDE (0.196 cm−2) and then dried at room temperature for 15 min prior to the MOR evaluation. The total metal loading was 40 μg cm−2 for all samples. The voltammetric curves of the MOR were obtained at 50 mV/s in an N2-saturated aqueous solution containing CH3OH (1 M) and KOH (1 M) solution. Before the MOR measurements, the electrode surface was cleaned by cycling between −0.9 and 0.3 V in 1 M aqueous KOH (20–40 cycles). To test the durability of the catalysts, the I–t amperometric curve was acquired at −0.2 V for 3600 s.
3.4. Characterization of Pd@Pt Core–Shell Nanocubes
TEM, high-resolution TEM (HRTEM), and energy-dispersive X-ray spectroscopy (EDX) images were captured using a JEM-2100F microscope (JEOL, Akishima, Tokyo, Japan) operated at 200 kV. Powder XRD patterns were obtained with a Rigaku D-MAX/A diffractometer (Rigaku, Akishima, Tokyo, Japan) at 35 kV and 35 mA. The elemental composition of the Pd@Pt core–shell nanocubes was measured using a direct reading echelle inductively coupled plasma (ICP) spectrometer (Leeman, Hudson, NH, USA).