Article pubs.acs.org/Langmuir
Spongelike Nanoporous Pd and Pd/Au Structures: Facile Synthesis and Enhanced Electrocatalytic Activity Jungwoo Son,† Seunggi Cho,† Chongmok Lee,*,§ Youngmi Lee,*,§ and Jun Ho Shim*,†,‡ †
Department of Chemistry, Daegu University, Gyeongsan 712-714, Korea Center for Bio-Nanomaterials and Institute of Basic Science, Daegu University, Gyeongsan 712-714, Korea § Department of Chemistry & Nano Science, Ewha Womans University, Seoul 120-750, Korea ‡
S Supporting Information *
ABSTRACT: This paper reports the facile synthesis and characterization of spongelike nanoporous Pd (snPd) and Pd/Au (snPd/Au) prepared by a tailored galvanic replacement reaction (GRR). Initially, a large amount of Co particles as sacrificial templates was electrodeposited onto the glassy carbon surface using a cyclic voltammetric method. This is the key step to the subsequent fabrication of the snPd/ Au (or snPd) architectures by a surface replacement reaction. Using Co films as sacrificial templates, snPd/Au catalysts were prepared through a two-step GRR technique. In the first step, the Pd metal precursor (at different concentrations), K2PdCl4, reacted spontaneously to the formed Co frames through the GRR, resulting in a snPd series. snPd/Au was then prepared via the second GRR between snPd (prepared with 27.5 mM Pd precursor) and Au precursor (10 mM HAuCl4). The morphology and surface area of the prepared snPd series and snPd/Au were characterized using spectroscopic and electrochemical methods. Rotating disk electrode (RDE) experiments for oxygen reduction in 0.1 M NaOH showed that the snPd/Au has higher catalytic activity than snPd and the commercial Pd-20/C and Pt-20/C catalysts. Rotating ring-disk electrode (RRDE) experiments reconfirmed that four electrons were involved in the electrocatalytic reduction of oxygen at the snPd/Au. Furthermore, RDE voltammetry for the H2O2 oxidation/reduction was used to monitor the catalytic activity of snPd/ Au. The amperometric i−t curves of the snPd/Au catalyst for a H2O2 electrochemical reaction revealed the possibility of applications as a H2O2 oxidation/reduction sensor with high sensitivity (0.98 mA mM−1 cm−2 (r = 0.9997) for H2O2 oxidation and −0.95 mA mM−1 cm−2 (r = 0.9997) for H2O2 reduction), low detection limit (1.0 μM), and a rapid response ( snPd-50 > snPd-10. The decrease in charge density from snPd-27.5 to snPd-50, which suggests a decreased Pd surface area of snPd-50, is possibly caused by the larger particle formation within the surface Pd film at a higher Pd precursor concentration. On the other hand, the introduction of the Au precursor via the second GRR suppresses both the Hads/des peaks to a certain degree, but not completely, compared to the corresponding peaks of snPdE
dx.doi.org/10.1021/la4047947 | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Table 1. Comparison of the Surface Area and the Catalytic Activity for snPd, snPd/Au, and Commercial Catalysts (Data Obtained from CVs and RDE) catalyst
GSAa (cm2)
ESAb (cm2)
onset potentialc (V)
E1/2 (V)
nd
RDE slopes for ORRe (V)
RFf
snPd-27.5 snPd/Au Pd-20/C Pt-20/C
0.080 0.081 0.081 0.094
2.6 2.6 0.3 0.4
−0.055 −0.034 −0.064 −0.041
−0.165 −0.143 −0.163 −0.177
3.87 3.92 3.82 3.83
0.177 0.174 0.215 0.248
31.9 31.5 3.2 4.0
a
Calculated from the slop of Q vs t1/2 obtained from the plots of the chronocoulograms. bSurface area of the exposed metal calculated using the charges corresponding to hydrogen adsorption and desorption with a conversion factor of 210 μC cm−2 for Pt-20/C and monolayer Pd oxide reduction with a conversion factor of 310 μC cm−2 for the remaining catalysts.52,53 cEstimated from the cathodic current exceeding four times of the standard deviation of S/N ratio. dCalculated from the RDE data recorded at a potential of −0.6 V vs SCE. eThe RDE curve slopes in the mixed kinetic-diffusion-controlled region was confirmed from the ΔE values for the middle 90% changes in the limiting current densities. fCalculated from the RSA/GSA.
C (4.0) catalysts, respectively, possibly due to the distinctively different surface morphology. In general, pure Pd-based catalysts show lower ORR activity than Pt or Pt-based ones, and Au is a poor material for ORR. Interestingly, snPd/Au, despite the lower activity of each metal, showed better ORR activity than commercial Pt-20/C (Table 1). This suggests that the specific compositional/structural features are other important factors governing the ORR electrocatalysis, in addition to the effect of the distinctive morphology of snPd/ Au (highly porous in macro- and microlevels). The number of electrons transferred (n) for the ORR was calculated from the RDE voltammetry results obtained at various electrode rotation speeds using the Koutecky−Levich (K−L) equation54
5B showed that the exposed surface areas of Pd in the snPd series was in the sequence, snPd-27.5 > snPd-50 > snPd-10, which is consistent with the trends in the Hads/des charge density and SEM images. Electrocatalytic Activities of snPd/Au toward ORR. The catalytic performance of a series of snPd and snPd/Au catalysts toward the ORR was examined by RDE voltammetry under alkaline conditions. This study examined the effects of the Au precursor’s concentration for a systematic series of snPd/Au catalysts on the performance of ORR, as shown in Figure S3. Among the snPd/Au catalysts studied, those prepared with the Au precursor’s concentration of 10 mM (denoted as snPd/Au) exhibited the best ORR performance. Therefore, the concentration of the Au precursor was fixed to 10 mM. Figure 6A shows the normalized RDE voltammetric curves recorded in the O2-saturated 0.1 M NaOH solution at 1600 rpm with each catalyst-deposited GC electrode (for RDE results at different rpm, see Figure S4). For comparison, the ORR activities of the commercial Pd-20/C and Pt-20/C catalyst-loaded GC electrodes were also examined. The ORR began at a slightly more positive potential for snPd/Au followed very closely by the others, i.e., snPd-27.5, Pt-20/C, and Pd-20/C. The ORR onset potentials were more positive in the following order: snPd/Au > Pt-20/C > snPd-27.5 ≈ Pd-20/C (see Figure 6B). The trend of the half-wave potentials (E1/2) was slightly different to that of the ORR onset potentials. The E1/2 value at snPd/Au (−0.143 V vs SCE) was more positive than the ones, not only at snPd-27.5 but also at both commercial Pt-20/C and Pd-20/C catalysts, indicating rather faster ORR kinetics at snPd/Au. The RDE curve sharpness in the mixed kinetic-diffusion controlled region was confirmed more clearly from the ΔE values for the middle 90% changes in the limiting current densities, i.e., the curve slope increases with decreasing ΔE: the measured ΔE values for snPd-27.5, snPd/Au, Pd-20/C, and Pt-20/C were 0.177, 0.174, 0.215, and 0.248 V, respectively. From the CVs experiments, the electrochemical active surface area (ESA) was measured from the integrated charge amounts passed for hydrogen adsorption on Pt-20/C52 with a conversion factor of 210 μC cm−2 and for the surface Pd oxide reduction on the others,53 i.e., snPd-27.5, snPd/Au, and Pd-20/ C, with a conversion factor of 310 μC cm−2. The effect of the structural difference can be observed more clearly in Figure 6C, showing the RDE voltammetry curves based on jESA (current normalized to the electrode ESA) instead of jGSA. As summarized in Table 1, the estimated roughness factor (i.e., RF = ESA/GSA) was ca. 8−10 times higher for snPd-27.5 (31.9) and snPd/Au (31.5) than for Pd-20/C (3.2) and Pt-20/
1 1 1 = + j jk jd
(3)
where j is the measured limiting current density, jk is the kinetic current density, and jd is the diffusion-limited current density expressed as jd = 0.62nFCO2DO2 2/3V −1/6ω1/2
(4) −6
−3
where F is the Faraday constant, CO2 (1.2 × 10 mol cm ) is the saturated concentration of oxygen, DO2 (1.9 × 10−5 cm2 s−1) is the diffusion coefficient of oxygen, v is the kinematic viscosity of the solution, and ω is the electrode rotation rate.55 As shown in Figure 6D, the K−L plots at −0.6 V (vs SCE) showed good linearity, indicating first-order kinetics with respect to the reactant concentration. The n values for snPd27.5, snPd/Au, Pd-20/C, and Pt-20/C calculated from the slopes in the K−L plots were 3.87, 3.92, 3.82, and 3.83, respectively. This suggests that the oxygen is reduced via a direct four-electron transfer pathway at both snPd-27.5 and snPd/Au in an alkaline solution, whereas snPd/Au showed a relatively higher n value. The n values were a function of the potentials, as shown in Figure S5. The potential dependency was much less significant, particularly for the snPd-27.5, snPd/ Au, and Pt-20/C, whereas Pd-20/C showed much stronger dependency of the n values on the potential. For the snPd/Au, the ORR via four-electron transfer occurred mainly without depending on the potential (
