J Mater Sci (2013) 48:2142–2150 DOI 10.1007/s10853-012-6989-7
Carbon-supported AuPd bimetallic nanoparticles synthesized by high-energy electron beam irradiation for direct formic acid fuel cell Yuji Ohkubo • Masashi Shibata • Satoru Kageyama • Satoshi Seino • Takashi Nakagawa • Junichiro Kugai • Hiroaki Nitani • Takao A. Yamamoto
Received: 17 August 2012 / Accepted: 24 October 2012 / Published online: 10 November 2012 Ó Springer Science+Business Media New York 2012
Abstract Nanoparticle catalysts of carbon-supported Pd (Pd/C) and carbon-supported AuPd (AuPd/C) for the direct formic acid fuel cell (DFAFC) anode were synthesized by the reduction of precursor ions in an aqueous solution irradiated with a high-energy electron beam. We obtained three kinds of nanoparticle catalysts: (1) Pd/C, (2) AuPd/C of the core–shell structure, and (3) AuPd/C of the alloy structure. The structures of AuPd nanoparticles were controlled by the addition of citric acid as a chelate agent, and sodium hydroxide as a pH controller. The structures of nanoparticle catalysts were characterized using transmission electron microscopy, inductively coupled plasma atomic emission spectrometry, the techniques of X-ray diffraction and X-ray absorption fine structure. The catalytic activity of the formic acid oxidation was evaluated using linear sweep voltammetry. The oxidation current value of AuPd/C was higher than that of Pd/C. This indicated that the addition of Au to Pd/C improved the oxidation activity of the DFAFC anode. In addition, the AuPd/ C of the alloy structure had higher oxidation activity than the AuPd/C of the core–shell structure. The control of the AuPd mixing state was effective in enhancing the formic acid oxidation activity.
Y. Ohkubo (&) M. Shibata S. Kageyama S. Seino T. Nakagawa J. Kugai T. A. Yamamoto Graduate School of Engineering, Osaka University, 2-1 Yamadaoka, Suita, Osaka 565-0871, Japan e-mail:
[email protected] H. Nitani Institute of Materials Structure Science, High Energy Accelerator Research Organization (KEK), 1-1 Oho, Tsukuba, Ibaraki 305-0801, Japan
123
Introduction Study of bimetallic nanoparticles is one of the most interesting topics among material scientists [1], and these particles are expected to be used as materials for various catalysts [2–4]. Of the various bimetallic nanoparticles in existence, AuPd nanoparticles have attracted the attention of many researchers, because, despite being Pt-free, they exhibit high catalytic activity for various reactions such as the synthesis of hydrogen peroxide and the oxidation of alcohols [5–7]. In addition, AuPd nanoparticles can be used as electrode catalysts for direct formic acid fuel cells (DFAFCs). DFAFC is expected to be one of the nextgeneration polymer electrolyte fuel cells because of two reasons: (1) The theoretical electromotive potential of DFAFC, as calculated from the Gibbs free energy, is higher than that of direct methanol fuel cell [8, 9]. (2) The fuel crossover of formic acid is considerably less than that of methanol [9, 10]. The structure control of the bimetallic nanoparticle, especially core–shell and alloy structures, is effective for improving the catalytic activity. However, in the case of a liquid phase method, the preferential synthesis of the alloy structure is generally not easy on the AuPd system, because Au and Pd ions are not reduced simultaneously in the solution on account of the large difference in the redox potentials between them. The relationship between the structure of AuPd nanoparticles and catalytic activity has, therefore, not been established. We proposed a solution to the problem by employing a radiolytic process that uses a high-energy electron beam. We succeeded with the bimetallic nanoparticles of the core–shell, alloy and intermediate structures of PtRu and PtCu systems, and previously reported the effects of the structure control on methanol oxidation activity [11] and preferential oxidation of CO [4, 12], respectively. With regards to the AuPd
J Mater Sci (2013) 48:2142–2150
system, we successfully obtained nanoparticles of the alloy structure by means of a one-pot method together with a combination of radiolytic synthesis and the addition of citric acid [13], although most researchers have prepared AuPd nanoparticles of the only core–shell structure, while some researchers have prepared AuPd nanoparticles of the alloy structure using a multistep process. In this study, we show a relatively simple preparation method for AuPd nanoparticles of both core–shell and alloy structures using the electron beam irradiation reduction method (EBIRM), and clarify the relationship between the preparation condition and the structure of bimetallic nanoparticles. Finally, we discuss the relationship between the AuPd structure and the catalytic activity of the formic acid oxidation.
Experimental Synthesis of AuPd/C and Pd/C by electron beam irradiation The AuPd/C samples were synthesized by EBIRM. The procedure and mechanism of this radiochemical method were described in previous reports [4, 13–16]. HAu Cl44H2O (Wako) and PdCl22NaCl3H2O (Wako) were used as metal precursors. Ultrapure water (18 MXcm) was used as a solvent. An aqueous 400 mL solution containing 0.0625 mM of HAuCl44H2O and 0.25 mM of PdCl2 2NaCl3H2O (atomic ratio Au/Pd = 20/80) was prepared in a polystyrene culture flask of 1900 mL. Ninety one milligrams of carbon black powder (Vulcan XC-72R, Cabot Corp.) was added to the solution as a support material with conductive and corrosion resistance properties. The amount of carbon particles was about five times that of the noble metals by weight (weight ratio AuPd/C = 1/5). The metal loading was adjusted to be approximately 20 wt%. 1 vol.% of 2-propanol (Wako) was added to the solution as a reduction enhancer. Prior to irradiation, the flask was sealed and sonicated to have well-dispersed carbon particles. In order to reduce the noble metal ions, the solution was irradiated with an electron beam of 4.8 MeV, 10 mA. The total dose was controlled to 20 kGy, such that irradiation of each flask took only 6–7 s. The surface dose was measured using a radiochromic dosimeter pasted on the flask wall. Radiation-induced radicals reduced the metal precursors, and the AuPd particles were stabilized on the carbon particles. After the irradiation, the solution was segregated into the powder and the supernatant solution using suction filtration. The powder was then dried at 75 °C overnight. For comparison purposes, the Pd/C samples were synthesized in the same way without the addition of HAuCl44H2O.
2143
Characterization The dispersibility, the particle size, and the size distribution of the metal particles on the carbon particles were investigated by transmission electron microscopy (TEM) using H-8100 (Hitachi) with an acceleration voltage of 200 kV. The samples dispersed in ethanol were dropped on a TEM grid (Nisshin EM) and dried prior to observation. The particle size distribution was obtained by measuring the size of over 100 metal particles. Then, the average diameter of metal particles was obtained as a geometric mean. The amounts of Au and Pd in the obtained samples were analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES) using ICPS7500 (Shimazu) to determine the metallic chemical composition and metal loading ratio to the carbon particles. First, the Au and Pd in the samples were dissolved by aqua regia (volume ratio HCl/HNO3 = 3/1), and the diluted solution containing Au and Pd was then sprayed into a plasma torch through a nebulizer. The molar and atomic ratios of Au and Pd were calculated by a calibration curve method. Metallic phases in the obtained samples were investigated by X-ray diffraction (XRD) measurement using RINT2100-Ultima (Rigaku) at room temperature with Cu ˚ . Diffraction data Ka radiation with a wavelength of 1.54 A were collected from 30° to 50° in 2h. The chemical states of metals were investigated by comparing the X-ray absorption near edge structure (XANES) spectra around the Au-LIII- and Pd-K edge of the obtained samples together with those of the reference materials. Both the XANES spectra around the Au-LIII(11910 eV) and Pd-K edge (24300 eV) were measured at the PF-AR NW10A beam line of the Advanced Ring for Pulse X-rays at High Energy Accelerator Research Organization (KEK), Tsukuba, Japan. At the Au-LIII edge, the intensities of the incident and transmitted X-rays were measured in the ion chambers filled with a mixed gas of N2 (85 %) ? Ar (15 %) and a monoatomic gas of Ar (100 %), respectively. At the Pd-K edge, the intensities of the incident and transmitted X-rays were measured in the ion chambers filled with a monoatomic gas of Ar (100 %) and a mixed gas of Kr (50 %) ? Ar (50 %), respectively. The X-ray beam was focused using a Si(311) crystal monochromator. The obtained samples were mixed with a boron nitride powder and were shaped into pellets having 7 mm diameters. The pellets were then put into the X-ray path. In addition, the chemical coordination from Au was investigated by performing an extended X-ray absorption fine structure (EXAFS) analysis. The EXAFS oscillation was extracted by subtracting the background from experimentally obtained raw X-ray absorption spectrum. The EXAFS analyses for the Au-LIII edge were performed by calculating the theoretical EXAFS function. The curve fits
123
2144
were performed with the Artemis code [17], and backscattering amplitudes, phase shifts, and mean free paths were calculated using the FEFF6 code [18]. Assuming the nearest coordination shell consisting of Au and Pd atoms, the coordination numbers to these atoms from Au, NAu–Au, and NAu–Pd were determined by fitting v(k) to that obtained by the inverse Fourier transform of the experimental space r data using conventional EXAFS parameters and the IFEFFIT program [19]. The analysis procedure was similar to those we reported previously [20–22]. The formic acid oxidation activities of the obtained samples were evaluated by linear sweep voltammetry (LSV) using a three-electrode-type beaker cell. A gold wire with a 0.5 mm diameter and an Ag/AgCl electrode (RE-1B, BAS) were used as the counter and reference electrodes, respectively. The obtained sample sandwiched between two carbon papers (TGP-H-060, Chemix) was used as a working electrode. Potential–current curves were obtained with an electrochemical analyzer (ALS630B, ALS) by varying the potential from -0.1 to 0.6 V versus the standard hydrogen electrode (SHE) at a sweep rate of 10 mV/s. LSV measurements were performed in a solution containing 0.5 mol/L of H2SO4 and 1.2 mol/L of formic acid (Wako) under a nitrogen atmosphere. The solution temperature was maintained at 35 °C using a magnetic hot stirrer (CERAMAG Midi, IKA Works). In addition, two types of commercial Pd black samples with different specific surface areas were used to confirm the effect of specific surface area on the catalytic activity of the formic acid oxidation. One commercial Pd black sample (99.9 wt%, specific surface area 20 m2/g) was purchased from Alfa Aesar, and the other commercial Pd black sample with high surface area (99.8 wt%, specific surface area 40–60 m2/g) was purchased from Sigma-Aldrich.
J Mater Sci (2013) 48:2142–2150
On the other hand, in our EBIRM, the precursor solution contains not only citric acid but also metal ions and carbon supports. Although such composition of the precursor solution does not ensure that the dissociation reaction of citric acid finishes at pH = 6, it would be apparent that a high pH promotes the complexation between metal ions and chelate agents. This behavior would help us to understand the importance of pH control.
ð1Þ
Particle morphology using TEM Figure 2 shows TEM micrographs of Pd/C obtained with pH = 12 (Pd/C_pH12), AuPd/C obtained with pH = 12 (AuPd/C_pH12), and AuPd/C obtained with pH = 10 (AuPd/C_pH10). Pd/C obtained with pH = 10 (Pd/ C_pH10) was also observed by TEM, but is not shown here. The micrograph of Pd/C_pH10 was similar to that of AuPd/C_pH10. The smaller black images with diameters less than 10 nm are metal particles. The average grain sizes of Pd/C_pH12, AuPd/C_pH12, and AuPd/C_pH10 are approximately 7, 6, and 5 nm, respectively. Some metal particles are agglomerated on the surface of the carbon particles in Pd/C_pH12 (Fig. 2a) and AuPd/C_pH12 (Fig. 2b), resulting in their increase of average grain sizes and the uneven distribution. Whereas, many metal particles are relatively well dispersed on the surface of the carbon particles in AuPd/C_pH10 (Fig. 2c). As shown in Fig. 2a 100
Effect of addition of chelate agents with pH control Chemical equation (1) shows the dissociation equilibrium of citric acid. The carboxyl group releases a hydrogen ion. In other words, when COOH changes to COO-, the citric acid can form the coordinate bond with metal ions. As reported by Onodera et al. [23], the ratio of anions with increasing pH was calculated using the dissociation constant of protons (pKa = 2.73). Figure 1 shows the calculated results and illustrates the ratio of anions for each pH in the aqueous solution of citric acid. It is clear that the dissociation reaction of citric acid starts at pH = 2 and finishes at pH = 6. In other words, in theory, all of the hydrogen dissociates from citric acid at pH [ 6, resulting in chelate agents that work effectively.
123
Ratio of anion for citric acid [%]
Results and discussion 80
60
40
20
0
0
2
4
6
8
10
pH
Fig. 1 Ratio of anion for each pH in the aqueous solution of citric acid
J Mater Sci (2013) 48:2142–2150
2145
(a)
30
Average grain size: 7.1 nm
Carbon
Frequency (%)
25 20 15 10
Metal 5 0
0
5
10
15
20
Diameter (nm)
20 nm
(b)
30
Average grain size: 6.0 nm
Frequency (%)
25 20 15 10 5 0
0
5
20 nm
10
15
20
Diameter (nm)
(c) 30
Average grain size: 5.1 nm
Frequency (%)
25 20 15 10 5 0
20 nm
0
5
10
15
20
Diameter (nm)
Fig. 2 TEM micrographs of the Pd/C and AuPd/C samples: a Pd/C_pH12, b AuPd/C_pH12, and c AuPd/C_pH10. b TEM micrograph of AuPd/ C_pH12 is the same as that in our previous study (Ohkubo et al. [13])
and b, the addition of Au does not significantly affect the dispersibility, the particle size, and the size distribution of the metal particles on the carbon particles. As shown in Fig. 2b and c, the change in pH affects the dispersibility of the metal particles supported on the carbon particles,
resulting in a change in the specific surface area of the metal particles. Therefore, it was predicted that the catalytic activity will be reduced for the samples obtained at high pH because some agglomeration occurred in Pd/ C_pH12 and AuPd/C_pH12.
123
2146
J Mater Sci (2013) 48:2142–2150
Rate of reduction and chemical composition using ICP-AES Table 1 shows the target and analyzed atomic ratios of metal loading. In Pd/C_pH12, 70 % of the target Pd loading was supported on the carbon particles. This value means that 70 % of the Pd ions in the precursor solution was reduced and deposited on carbon particles under the irradiation with electron beam, and the rest of the Pd ions remained in the supernatant solution. In summary, this behavior indicates that not all of the Pd ions are reduced and deposited on the carbon particles. In AuPd/C_pH12, 59 % of the target Pd loading (=80 %) was supported on the carbon particles. When the target Pd loading is 100 %, the percentage of reduction and deposition of Pd ions becomes 74 %. This percentage is close to that of Pd/C_pH12, which is 70 %. A similar behavior is also observed between Pd/C_pH10 and AuPd/C_pH10. These results indicate that the addition of Au has little effect on the reduction of Pd ions.
Each pattern has a large peak, which is very broad because of the fine grain size. XRD pattern of Pd/C_pH12 has the peak very close to 40.11° due to Pd(111). In contrast, XRD pattern of AuPd/C_pH12 has the peak indexed to Pd(111), which shifts to lower degrees with the addition of Au. This peak shift is a good, and a direct indicator of the alloying of Au into a Pd lattice. If monometallic Au and Pd nanoparticles are separately formed, both peaks at 38.18° due to Au(111) and at 40.11° due to Pd(111) would appear. However, the peaks are not observed in Fig. 3, indicating that AuPd/C_pH12 does not have monometallic Au and Pd nanoparticles. It is observed that the positions of the peak indexed to Pd(111) are different from those between AuPd/ C_pH10 and AuPd/C_pH12. XRD pattern of AuPd/ C_pH10 shows a greater shift in the peak indexed to Pd(111) compared with those of AuPd/C_pH12 because of pH control. This behavior indicates that the alloying depends on the pH of the precursor solution. Metal and oxide fractions using XANES
Structure of Au and Pd using XRD Figure 3 shows XRD patterns obtained with the three samples: Pd/C_pH12, AuPd/C_pH10, and AuPd/C_pH12. Table 1 Atomic ratios of the target and analyzed metal loading of the Pd/C and AuPd/C samples Sample ID
Target loading
Analyzed loading
Au (at.%)
Pd (at.%)
Au (at.%)
Pd (at.%)
Pd/C_pH10
–
100
–
63
Pd/C_pH12
–
100
–
70
AuPd/C_pH10 AuPd/C_pH12
20 20
80 80
12 8
48 59
Au(111)
Pd(111)
lntensity (arb. unit)
AuPd/C_pH10
AuPd/C_pH12
Pd/C_pH12 36
37
38
39
40
41
42
Diffraction angle 2q (degree)
Fig. 3 XRD patterns of the Pd/C and AuPd/C samples. The vertical lines denote positions of diffraction peaks in pure Au and Pd metal
123
Figure 4a shows the normalized Au-LIII edge XANES spectra of AuPd/C_pH10 and AuPd/C_pH12. The spectral shapes of both AuPd/C_pH10 and AuPd/C_pH12 resemble that of the reference Au-foil as shown in Fig. 4a. These spectra were found to be successfully described by the linear combinations of the reference spectra of the Au-foil and the Au2O3 in the region of -20 to 50 eV from the absorption edge. We used the Athena code, and the associated R factors were all less than 0.00032, and were sufficiently low to support the reliabilities of the measurements and analysis procedures. The calculation resulted in Au-foil/Au2O3 ratios of 100/0 for both AuPd/ C_pH10 and AuPd/C_pH12. It is clear that both samples of AuPd/C_pH10 and AuPd/C_pH12 show all Au-metal fractions, indicating that the Au was not oxidized. Figure 4b shows the normalized Pd-K edge XANES spectra of AuPd/C_pH10 and AuPd/C_pH12. The spectral shape of AuPd/C_pH12 is slightly different from that of AuPd/C_pH10, as shown in Fig. 4b. The spectral shape of AuPd/C_pH10 resembles that of the reference PdO at the Pd-K edge while the spectral shape of AuPd/C_pH12 does not resemble it. The Pd-K edge XANES spectrum of AuPd/ C_pH12 has a second peak at 24388 eV which corresponds to the second peak of the reference Pd black. These spectra were successfully described by the linear combination analyses of the reference spectra of the Pd black and the PdO in the region of -20 to 50 eV from the absorption edge. We used the Athena code, the associated R factors were all less than 0.00006, and were sufficiently low to support the reliabilities of the measurements and analysis procedures. The calculation resulted in Pd black/PdO ratios of 26/74 for AuPd/C_pH10 and 49/51 for AuPd/C_pH12.
2147
(a)
Data
(a)
Window
Au2O3 Au foil AuPd/C_pH12 AuPd/C_pH10
11850
11900
11950
12000
12050
Fit
Fourier Transform (arb. unit)
Adsorption coefficience (arb. unit)
J Mater Sci (2013) 48:2142–2150
Energy (eV)
0
0.1
0.2
(b)
24300
0.3
0.4
0.5
r (nm) PdO Pd black AuPd/C_pH12
(b)
Data Window
AuPd/C_pH10
24350
24400
24450
24500
Energy (eV)
Fourier Transform (arb. unit)
Adosorption coefficience (arb. unit)
Second peak (24388 eV)
Fit
Fig. 4 Normalized XANES spectra of the AuPd/C samples: a at the Au-LIII edge and b at the Pd-K edge
0
It is clear that the Pd-oxide fractions of AuPd/C_pH10 and AuPd/C_pH12 are different. Taking into account the previous reports on bimetallic nanoparticles [11, 24], this difference implies that more Pd is located on the outside of AuPd nanoparticles in AuPd/C_pH10 than in AuPd/ C_pH12. AuPd mixing state using EXAFS Figures 5 and 6 are radial structure functions (RSFs) and Fourier back-transformed EXAFS oscillations of AuPd/ C_pH10 and AuPd/C_pH12 obtained at the Au-LIII edge. The frame lines for RSF represent the Hanning window functions used in the analyses. Structural parameters were optimized by curve fits in the q-space assuming single scattering paths. The curve fits were performed with the Artemis code; and backscattering amplitudes, phase shifts, and mean free paths were calculated using the FEFF6 code. Dots denote experimental values of the back-transformed
0.1
0.2
0.3
0.4
0.5
r (nm)
Fig. 5 RSF of the AuPd/C samples at the Au-LIII edge: a AuPd/ C_pH10 and b AuPd/C_pH12
oscillation and solid lines are drawn based on the optimized parameters. Note that the experimental plots are well reproduced. Four parameters: interatomic distance r, coordination number N, Debye–Waller factor r2, and delta energy DE0, were optimized using the Artemis code with Au-LIII edge EXAFS analyses assuming Au–Au and Au–Pd bonds. Here, we did not assume an Au–O bond because it was clear that Au did not oxidize according to XANES spectra. Table 2 shows r, N, r2, and DE0 obtained by fitting the parameters of the EXAFS spectra. In this table, the interatomic distance and coordination number from A to B are denoted as rA–B and NA–B, respectively. The reliability of the EXAFS analyses was checked by referring to the
123
2148
J Mater Sci (2013) 48:2142–2150
(a)
Data
χ(k)k3 (arb. unit)
Fit
0
50
100
150
200
k (nm-1)
(b) Data
χ(k)k3 (arb. unit)
Fit
The coordination numbers, NAu–Au and NAu–Pd, are plotted in Fig. 7 against pH control. In AuPd/C_pH10, NAu–Pd is considerably smaller than NAu–Au, and the difference between NAu–Au and NAu–Pd is large. These results, which are consistent with the previous report [21], concluded that the bimetallic AuPd nanoparticle had a core– shell structure. In summary, these coordination numbers and the difference in our data also indicate that an Au core– Pd shell structure is formed in AuPd/C_pH10. In AuPd/ C_pH12, the NAu–Pd approaches NAu–Au, and the difference between NAu–Au and NAu–Pd is small. This result indicates that a partially randomized alloy phase is formed in AuPd/ C_pH12. Assessments made from NAu–Au indicate that Au–Au bonds more often occur in AuPd/C_pH10 than in AuPd/ C_pH12. In contrast, assessments made from NAu–Pd indicate that Au–Pd bonds more often occur in AuPd/C_pH12 than in AuPd/C_pH10. These behaviors indicate that AuPd/C_pH12 consists of a mixed bimetal that is better than that in AuPd/C_pH10. In other words, the AuPd mixing is enhanced by pH control. This observation is consistent with the promotion of AuPd alloying that is confirmed by the peak shifts found in the XRD patterns. Formic acid oxidation activity
0
50
100
150
200
-1
k (nm )
Fig. 6 Fourier back-transformed EXAFS oscillations of the AuPd/C samples at the Au-LIII edge: a AuPd/C_pH10 and b AuPd/C_pH12
R factors [17] listed in the last columns. The associated R factors were all smaller than 0.005, which indicates a reliability that is sufficient to allow us to discuss the bimetallic structures.
Figure 8 shows the formic acid oxidation current versus potential curves obtained with the three samples: Pd/ C_pH12, AuPd/C_pH10, and AuPd/C_pH12. The oxidation current was normalized to the total metal loading weight contained in each sample in Fig. 8. The formic acid oxidation current of AuPd/C_pH12 is twice as high as that of Pd/C_pH12 at 0.5 V versus SHE. This result indicates that the addition of Au to Pd/C significantly enhances the formic acid oxidation activity. In addition, this result agrees with other studies [25] which indicate that AuPd/C obtained by the liquid phase method using the NaBH4 has higher formic acid oxidation activity than that of Pd/C. The formic acid oxidation current of AuPd/C_pH12 is higher than that of Pd/C_pH10 at 0.5 V versus SHE. Since the characterization results with XRD, XANES, and EXAFS imply that AuPd/C_pH12 is a better mixed bimetal than AuPd/C_pH10, these LSV results indicate that the AuPd mixing enhances the formic acid oxidation activity.
Table 2 Optimized structural parameters estimated by EXAFS analyses of AuPd/C at the Au-LIII edge Sample ID
Au coordination rAu–Au (nm)
AuPd/C_pH10 AuPd/C_pH12 a
0.283 0.283
Pd coordination NAu–Au a
9.1 ± 0.9
a
6.4 ± 0.3
r2 (10-5 nm2)
DE0
R factor
rAu–Pd (nm)
NAu–Pd
0.275
0.7 ± 0.2a
9.3
1.6
0.0043
0.278
a
8.8
2.5
0.0026
3.2 ± 0.2
Coordination numbers: NAu–Au and NAu–Pd are the same as that in our previous study (Ohkubo et al. [13])
123
J Mater Sci (2013) 48:2142–2150
2149
NAu-Pd
NAu-Au 10
Coordination numbers
8
6
4
2
0
9
10
11
12
13
pH
Fig. 7 Coordination numbers of Au and Pd atoms around an Au atom estimated by EXAFS analyses of the AuPd/C samples at the Au-LIII edge
Fig. 9 Formic acid oxidation current versus potential curves for the commercial catalysts of Pd
consistent with the common knowledge. However, the relationship between the specific surface area and catalytic activity is not established in AuPd/C_pH10 and AuPd/ C_pH12. Note that AuPd/C_pH10 has a smaller average grain size than AuPd/C_pH12 in Fig. 2, implying that AuPd/C_pH10 has a larger specific surface area than AuPd/ C. Therefore, it was predicted that the catalytic activity of AuPd/C_pH10 is higher than that of AuPd/C_pH12. Actually, the catalytic activity of AuPd/C_pH12 is higher than that of AuPd/C_pH10. This behavior is precisely explained by the AuPd mixing, which indicates that the AuPd mixing state is a significant factor for enhancing the catalytic activity of formic acid oxidation.
Conclusion Fig. 8 Formic acid oxidation current versus potential curves for the Pd/C and AuPd/C samples
In comparison, the formic acid oxidation activity of commercial catalysts of Pd were also evaluated by LSV. Figure 9 shows the formic acid oxidation current versus potential curves obtained by commercial catalysts of Pd. These results help us to understand the factors affecting catalytic activity. Note that commercial Pd black (Alfa Aesar) and Pd black (Aldrich) are vastly different in specific surface area. The formic acid oxidation currents of Pd black (Aldrich) are higher than that of Pd black (Aesar). This result indicates that the specific surface area is a significant factor affecting catalytic activity, which is
We have successfully synthesized AuPd/C catalysts for DFAFC using EBIRM. TEM micrographs indicate that AuPd nanoparticles were supported on the carbon particles and the dispersibility was changed by a pH in the precursor solution. XANES spectra indicate that Au atoms exist as a metal, while Pd atoms exist not only as a metal but also as an oxide. The results of XRD and XAFS analyses indicate that the addition of chelate agents with high pH control enhances the AuPd mixing. The relationship between the AuPd mixing state and the catalytic activity of the formic acid oxidation was investigated using XRD and EXAFS analyses and LSV. These investigations confirmed that the AuPd mixing is effective for enhancing the catalytic activity of the formic acid oxidation.
123
2150 Acknowledgements The electron beam irradiation was performed at Japan Electron Beam Irradiation Service, Ltd. We thank the staff, especially K. Ueno, for their assistance with the electron beam irradiation experiments. We also thank Mr. T. Kawaguchi (graduate student) for his dedication to this study.
References 1. Rodriguez JA (1996) Surf Sci Rep 24:223. doi:10.1016/ 0167-5729(96)00004-0 2. Kim IT, Lee HK, Shim J (2008) J Nanosci Nanotechnol 8:5302. doi:10.1166/jnn.2008.1147 3. Ca´rdenas-Lizana F, Go´mez-Quero S, Baddeley CJ, Keane MA (2010) Appl Catal A 387:155. doi:10.1016/j.apcata.2010.08.019 4. Yamamoto TA, Nakagawa T, Seino S, Nitani H (2010) Appl Catal A 387:195. doi:10.1016/j.apcata.2010.08.020 5. Edwards JK, Solsona B, Landon P, Carley AF, Herzing A, Watanabe M, Kiely CJ, Hutchings GJ (2005) J Mater Chem 15:4595. doi:10.1039/b509542e 6. Enache DI, Edwards JK, Landon P, Solsona-Espriu B, Carley AF, Herzing AA, Watanabe M, Kiely CJ, Knight DW, Hutchings GJ (2006) Science 311:362. doi:10.1126/science.1120560 7. Solsona BE, Edwards JK, Landon P, Carley AF, Herzing A, Kiely CJ, Hutchings GJ (2006) Chem Mater 18:2689. doi:10.1021/ cm052633o 8. Rice C, Ha S, Masel RI, Waszczuk P, Wieckowski A, Barnard T (2002) J Power Sources 111:83. doi:10.1016/S0378-7753(02) 00271-9 9. Rice C, Ha S, Masel RI, Wieckowski A (2003) J Power Sources 115:229. doi:10.1016/S0378-7753(03)00026-0 10. Rhee YW, Ha SY, Masel RI (2003) J Power Sources 117:35. doi: 10.1016/S0378-7753(03)00352-5 11. Kageyama S, Seino S, Nakagawa T, Nitani H, Ueno K, Daimon H, Yamamoto TA (2011) J Nanopart Res 13:5275. doi:10.1007/ s11051-011-0513-x
123
J Mater Sci (2013) 48:2142–2150 12. Kugai J, Kitagawa R, Seino S, Nakagawa T, Ohkubo Y, Nitani H, Daimon H, Yamamoto TA (2011) Appl Catal A 406:43. doi: 10.1016/j.apcata.2011.08.006 13. Ohkubo Y, Shibata M, Kageyama S, Seino S, Nakagawa T, Kugai J, Yamamoto TA (2011) Mater Lett 65:2165. doi:10.1016/ j.matlet.2011.04.023 14. Belloni J (2006) Catal Today 113:141. doi:10.1016/j.cattod. 2005.11.082 15. Seino S, Kinoshita T, Nakagawa T, Kojima T, Taniguci R, Okuda S, Yamamoto TA (2008) J Nanopart Res 10:1071. doi:10.1007/ s11051-007-9334-3 16. Yamamoto TA, Kageyama S, Seino S, Nitani H, Nakagawa T, Horioka R, Honda Y, Ueno K, Daimon H (2011) Appl Catal A 396:68. doi:10.1016/j.apcata.2011.01.037 17. Ravel B, Newville M (2005) J Synchrotron Radiat 12:537. doi: 10.1107/S0909049505012719 18. Rehr JJ, Albers RC (2000) Rev Mod Phys 72:621. doi: 10.1103/RevModPhys.72.621 19. Newville M, Ravel B, Haskel D, Rehr JJ, Stern EA, Yacoby Y (2000) Physica B 208&209:154. doi:10.1016/0921-4526(94) 00655-F 20. Nakagawa T, Nitani H, Tanbabe S, Okitsu K, Seino S, Mizukoshi Y, Yamamoto TA (2005) Ultrason Sonochem 12:249. doi: 10.1016/j.ultsonch.2004.02.002 21. Nitani H, Yuya M, Ono T, Nakagawa T, Seino S, Okitsu K, Mizukoshi Y, Emura S, Yamamoto TA (2006) J Nanopart Res 8:951. doi:10.1007/s11051-005-9048-3 22. Nitani H, Nakagawa T, Daimon H, Kurobe Y, Ono T, Honda Y, Koizumi A, Seino S, Yamamoto TA (2007) Appl Catal A 326:194. doi:10.1016/j.apcata.2007.04.018 23. Onodera T, Suzuki S, Takamori Y, Daimon H (2010) Appl Catal A 379:69. doi:10.1016/j.apcata.2010.03.003 24. Kugai J, Moriya T, Seino S, Nakagawa T, Ohkubo Y, Nitani H, Daimon H, Yamamoto TA (2012) Int J Hydrog Energy 37:4787. doi:10.1016/j.ijhydene.2011.12.070 25. Zhou W, Lee JY (2007) Electrochem Commun 9:1725. doi: 10.1016/j.elecom.2007.03.016
