Element-Specific Analysis of the Growth Mechanism, Local Structure

Article pubs.acs.org/JPCC

Element-Specific Analysis of the Growth Mechanism, Local Structure, and Electronic Properties of Pt Clusters Formed on Ag Nanoparticle Surfaces Paul N. Duchesne and Peng Zhang* Department of Chemistry, Dalhousie University, Halifax, NS B3H 4R2, Canada ABSTRACT: Bimetallic nanoparticles (NPs) consisting of small Pt clusters on the surface of larger metal NPs are potential catalysts that combine the advantages of reduced cost, high surface area, and a strong alloying interaction between Pt and the substrate. Herein we report the preparation of a series of small Pt clusters deposited on the surface of Ag NPs with systematically varied compositions (Ag93Pt7, Ag81Pt19, Ag72Pt28, and Ag65Pt35) via a galvanic replacement reaction. UV−vis spectroscopy, transmission electron microscopy, and inductively coupled plasma optical emission spectroscopy are used to provide evidence of NPs formation and evaluate their gross structural and compositional characteristics. Extended X-ray absorption fine structure (EXAFS) measurements at the Pt L3- and Ag K-edges are then used to elucidate the structural evolution of small surface Pt clusters as the Pt concentration is increased, allowing the formation mechanism of the bimetallic NPs to be deduced. X-ray absorption near-edge structure (XANES) provides further information regarding the electronic properties of the bimetallic NPs from both Ag and Pt perspectives. Finally, the Pt L3-edge XANES spectra are used in conjunction with ab initio calculations to verify the accuracy of structural models based on the EXAFS fitting results and to provide insight into the size-dependent nature of the Ag−Pt bonding interaction.

1. INTRODUCTION Pt-based heterogeneous catalysts currently see widespread study and application in a variety of chemical processes, including fuel cell operation and the production of combustible fuels from biomass.1−3 As Pt is both costly and relatively scarce, the importance of small particle size has been well established, with fine Pt nanoparticles (NPs) resulting in very high surface area catalysts.4 The field of nanomaterials research is well suited to address this need, as it emphasizes the controlled synthesis of structured materials with dimensions on the order of 100 nm or less. It has been observed previously that the formation of Pt-based alloys and composites can result in catalysts whose activity matches or even exceeds that of pure Pt.5,6 This approach presents its own challenges, as the incorporation of some baser elements (e.g., Fe) into Pt materials can be hindered by their rapid oxidation,7 but a simple solution can be found in the use of still relatively inexpensive and unreactive noble metals such as Ag. By using these suitable elements to form “seed” NPs and depositing Pt exclusively onto the surfaces to form a core−shell structure, the amount of exposed Pt can be maximized to further improve cost-efficiency. In particular, galvanic replacement reactions, being limited by the ability of metal ions to access reductive surface sites, are well-suited for the controlled deposition of noble metals onto a less-noble substrate.8 When dealing with core−shell NPs, the thickness of the surface layer is also important. If the layer is too thick, many Pt atoms will lie beneath the exposed surface and be inaccessible and inactive; monolayer, or even submonolayer, coverage of the © XXXX American Chemical Society

particle core by Pt is thus desirable for the optimization of Pt site accessibility; in the latter case, interfacial regions between the two metals can exist at the surface, upon which the catalyzed reaction may proceed differently or even more efficiently than on Pt alone.9 Unfortunately, the characterization of NP surface structure using conventional techniques remains a challenge; however, through the use of more specialized methods, in particular X-ray absorption spectroscopy (XAS), such a challenge can be overcome.9,10 Reported herein are the galvanic replacement synthesis and characterization of a series of AgPt NPs with Pt-enriched surfaces, with emphasis on the potential of using both experimental and ab initio XAS techniques to obtain element-specific information regarding their structural and electronic properties.

2. MATERIALS AND METHODS 2.1. Sample Preparation. Silver nitrate (AgNO3, 99.9+%), polyvinylpyrrolidone (PVP, MW 8000), and dihydrogen hexachloroplatinate(IV) hexahydrate (H2PtCl6·6H2O, 99.9%) were purchased from Alfa Aesar. Sodium borohydride (NaBH4, 99%) and n-butylamine (BNH2, 99.5%) were purchased from Sigma-Aldrich. Potassium hydroxide (KOH, 85+%) and hydrochloric acid (HCl(aq), 36.5−38.0% w/w) were purchased from ACP Chemicals. Nitric acid (HNO3(aq), 67−70% w/w) was purchased from Caledon Laboratory Chemicals. Vulcan Received: July 23, 2014 Revised: August 28, 2014

A

dx.doi.org/10.1021/jp507382z | J. Phys. Chem. C XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry C

Article

flow of 1.50 L/min, and a read time of 10 s with three replicates. Ultraviolet−visible spectroscopy measurements were performed using a Cary 100 Bio UV−visible spectrophotometer. For each AgPt sample, a 100 μL aliquot was taken via micropipet and diluted to a total volume of 3 mL before being analyzed. A DIW blank reference was also measured and used to background-subtract each sample spectrum. Spectra were normalized at 320 and 600 nm to more clearly illustrate the evolution of the Ag SPR absorption peak. Each sample was imaged using TEM; the relatively large amount of the PVP ligand material present in the final products proved problematic. In order to facilitate TEM analysis, a small portion of each AgPt sample was deposited onto carbon powder and washed thoroughly according to standard procedures.7 In brief, several milligrams of lyophilized NP product were dispersed in a 1 mL of n-butylamine suspension of Vulcan XC-72 and stirred for 48 h. The samples were then successively purified (three times) via high-speed centrifugation in absolute ethanol (laboratory grade), with the supernatant removed using a Pasteur pipet. After the final such purification, a small amount of the product dispersed in ethanol was dropcast onto Formvar-coated copper TEM grids. After the samples dried, they were analyzed using bright field imaging on an FEI Tecnai-12 transmission electron microscope (80 kV operating voltage). Mean particle diameters were determined by the digital image processing tool ImageJ.12 2.3. X-ray Spectroscopy and ab Initio Calculations. XAS data acquisition was performed at beamline BM-20 of the Advanced Photon Source at Argonne National Laboratories, Argonne, IL. The lyophilized AgPt powders were analyzed without further modification or dilution. In order to suppress thermal vibrations and enhance EXAFS signal intensity, a cryostat was used to maintain a constant sample temperature of 50 ± 1 K throughout the course of the measurements. A double-crystal Si(111) mirror monochromator was used for wavelength selection, with the incident beam intensity detuned to 80% in order to reject higher harmonics of the selected wavelength. Data were acquired using 32-element solid state Ge X-ray fluorescence detector (AgPt NPs) or gas ionization chambers (metal foil references) as appropriate. XAS data processing was performed using the IFEFFIT software suite.13 Absorption edge (E0) energies for Pt and Ag XANES spectra were calibrated relative to an in-line metal foil reference (Pt or Ag, as appropriate) and normalized using first- and third-order polynomials to fit the pre- and postedge regions of the spectra, respectively. XANES simulations of Pt L3-edge spectra were generated using FEFF software (version 8.2).14 Structural models were generated by manipulating a simple FCC lattice and substituting atoms at the appropriate sites as necessary. For the fitting of the experimental EXAFS data, two metal−metal paths were used, with their Debye−Waller coefficient (σ2) and E0-shift (ΔE0) parameters correlated in order to provide more degrees of freedom. The Pt and Ag spectra were first fitted separately to obtain reasonable starting values for subsequent fits; the spectra were then fitted again simultaneously under the identical constraints, save that the Ag−Pt and Pt−Ag bond lengths were correlated to reflect the fact that they represent a single bonding distance. Uncertainties for the reported parameter values were calculated from the off-diagonal elements of the correlation matrix for the final fit. These values were further weighted by the square root of the χ2 value, taking into account shot noise in the range 15−25 Å, as

XC-72 carbon powder was purchased from Cabot Corporation. Deionized water (DIW, 18.2 MΩ) was obtained using a Barnstead NANOpure DIamond UV ultrapure water system. All chemicals were used as received and without further purification. A single large batch of silver NPs (“seed particles”) was first synthesized using a procedure derived from that published by Zhang et al.11 Fresh stock solutions of PVP(aq) (8 g in 88 mL of DIW), KOH(aq) (28 mg in 25 mL of DIW), and AgNO3(aq) (17 mg in 5 mL of DIW) were prepared. The PVP(aq) was first transferred to a 250 mL round-bottom flask, to which 2 mL of KOH(aq) and 5 mL of the AgNO3(aq) were then added; this mixture was stirred and bubbled with nitrogen for 30 min. A fresh solution of NaBH4 (19 mg in 5 mL of ice-cold DIW) was prepared and added to the reaction mixture, which was then sealed under a N2(g) atmosphere and allowed to react for 1 h. Finally, the mixture was aged for 48 h without stirring so that any excess NaBH4 was allowed to decompose. A fresh stock solution of H2PtCl6·6H2O (13 mg in 25 mL of DIW) was used to prepare the bimetallic AgPt samples. First, 20 mL of the previously synthesized Ag seed NPs was transferred to a 50 mL round-bottom flask, followed by a predetermined volume of H2PtCl6·6H2O(aq) (1, 3, 5 mL, etc.). This mixture was then bubbled with N2(g) and refluxed (ca. 100 °C) for 10 min before being allowed to cool for 24 h while still under a N2(g) atmosphere. After cooling, the mixture was centrifuged at a relative centrifugal force (RCF) of 15 000 to remove any silver chloride or other insoluble byproducts formed during the reaction. In order to purify the final AgPt NPs, each sample (dispersed in DIW) was first treated with 4 vol equiv of acetone (laboratory grade) to induce precipitation of its constituent particles; at this point, the suspensions became visibly cloudy. Low-temperature centrifugation (3 °C, RCF = 15 000) was then performed to isolate the precipitated NPs, which were then redispersed in a minimal quantity of DIW before being precipitated with acetone and centrifuged once more. The final, twice-purified product was again redispersed in a minimal quantity of DIW, before being lyophilized. The resultant yellow or brown powders were light and easily handled for further analyses; however, due to the large PVP content of the final product, further preparation was required to prepare samples for transmission electron microscopy (TEM) measurements (see section 2.2.). In preparation for ICP-OES analysis, 10−20 mg of each AgPt NP sample was added to a 5 mL volumetric flask and dissolved in 1 mL of HNO3(aq); the flask was then stoppered, and the NPs/HNO3(aq) mixture was gently stirred for 24 h. After this treatment the suspension took on a partially opaque, brownish coloration that was visibly darker for samples to which more Pt had been added during synthesis. Next, 3 mL of HCl(aq) was added to the mixture to form an aqua regia solution; the flask was then gently stoppered (to prevent pressurization), and the mixture was stirred for a further 24 h. Following this stage of treatment, the solutions had visibly cleared and taken on a yellow-orange coloration; each solution was then diluted to a volume of 5 mL before being analyzed. 2.2. Standard Characterization Methods. The elemental analysis was conducted on a Varian Vista Pro (radial view) ICP OES instrument. The samples were prepared by dilution into 10% HCl, from 2× to 10× dilution depending on the concentrations of Ag/Pt. The instrument was operated with a power of 1.30 kW, a plasma flow of 15.0 L/min, an auxiliary B



Article

nm for Ag93Pt7, 4.6 ± 1.1 nm for Ag81Pt19, 3.5 ± 0.9 nm for Ag72Pt28, and 4.5 ± 1.2 nm for Ag65Pt35 were determined. Representative images of each NP sample are presented in Figure 1c−g. The AgPt NPs are visible as small (ca. 2−5 nm diameter) dark particles in the bright field images; the larger structures (ca. 20 nm or greater diameter) are particles of the carbon support material used to facilitate TEM data collection and improve sample stability for long-term storage. The unimodality of the particle diameter histograms supports the deposition of Pt onto the Ag seeds, as the nucleation of discrete Pt NPs would likely result in a bimodal distribution. From these images, the successful deposition of the NPs onto the carbon support is also clearly confirmed. To further characterize the properties of these AgPt NPs, UV−vis spectra were acquired for each sample. Background-corrected UV−vis absorption spectra for AgPt NP samples prior to deposition on Vulcan XC-72 are presented in Figure 2. An intense surface plasmon resonance (SPR) band, characteristic of nanosized Ag, is prominent in the spectrum of the Ag NPs sample. With increased amounts of Pt being added to the Ag seed particles, damping of this SPR band was observed. Experimental evidence from similar noble metal alloys reveals that the peak absorption of an SPR band is dependent on the both the dielectric constant of the constituent material and particle size.16 Given that the observed differences in mean particle diameter are not sufficient to significantly affect the SPR band17 and that TEM imaging precludes the possibility of NP aggregation, it is understood that the observed loss of absorption must be due to the incorporation of Pt atoms into the of Ag NP seeds. To further

recommended by Newville et al.15 Note that the Fouriertransformed regions were obtained over the range 2.5−12.5 Å−1 and fitted only within the range 2.2−3.1 Å.

3. RESULTS AND DISCUSSION 3.1. Initial Characterization. ICP-OES elemental analysis of the samples was performed in order to quantify the amount of Pt incorporated into each AgPt sample. Mass concentrations of Ag and Pt and elemental compositions corresponding to each volume of Pt solution added to Ag NP seeds are presented in Table 1. For convenience, samples are hereafter identified by their atomic compositions (i.e., Ag93Pt7, Ag81Pt19, Ag72Pt28, and Ag65Pt35). Table 1. ICP-OES and Calculated Elemental Compositions of Synthesized AgPt NPs Pt added (mL)

Ag (mg L−1)

Pt (mg L−1)

elemental composition

1 3 5 8

9.9 6.5 6.0 15.6

1.3 2.8 4.3 15.5

Ag93Pt7 Ag81Pt19 Ag72Pt28 Ag65Pt35

Preferential deposition of Pt atoms onto the surfaces of the Ag seed NPs was encouraged over homogeneous alloy formation through the use of a galvanic reduction reaction during synthesis (as illustrated in Figure 1a). Data on average particle size and standard deviation are shown in Figure 1b. Mean particle diameters of 3.9 ± 1.2 nm for pure Ag, 3.9 ± 1.4

Figure 1. (a) Schematic synthesis procedure, (b) mean particle diameters, and (c−g) TEM images depicting AgPt NPs deposited onto a Vulcan XC72 carbon powder support. C



Article

Fourier-transformed spectra, the reduced intensity of the peak between 2 and 3 Å arising from the combination of Ag−Ag and Ag−Pt scattering paths is again caused by destructive interference between Ag−Ag and Ag−Pt scattering paths. A similar trend in peak intensity is also observed from Pt L3-edge perspective, corresponding to the Pt−Pt and Pt−Ag scattering paths in the range of 2−3 Å. Experimental Fourier-transformed EXAFS spectra and computed fits from the Ag K-edge and Pt L3-edge perspectives are presented in Figures 4 and 5, respectively. Specific parameter values obtained by fitting these EXAFS spectra using standard methods7,18 are presented in Table 2. Figure 2. Background-corrected UV−vis absorption spectra of AgPt NP samples.

investigate this hypothesis, XAS analysis was employed to measure the resulting electronic and structural changes. 3.2. EXAFS Analysis. In order to characterize the structure of these bimetallic NPs from both Pt and Ag perspectives, we focus on the element-specific EXAFS analysis in this work. High-resolution TEM (HRTEM), a routine atomic structure analysis tool, is not included in the present study due to the different resolutions of the two techniques (i.e., ∼1 Å for HRTEM vs ∼0.01 Å for EXAFS). EXAFS k2-weighted k-space spectra corresponding to the AgPt samples are presented in Figure 3. Both the Ag K-edge and Pt L3-edge EXAFS oscillation patterns of the NPs closely resemble those of their respective metal foil references. The relatively low intensities of the Pt foil reference spectrum arise as a result of being measured at ambient temperature (ca. 300 K), whereas the Ag foil and AgPt samples were measured at low temperatures (50 ± 1 K). Among the AgPt NP samples, those containing greater proportions of platinum were observed to have less-intense EXAFS oscillations due to destructive interference between the Ag−Ag and Ag−Pt scattering paths in these bimetallic NPs. Conversely, analysis of the Pt L3-edge spectra revealed more consistent amplitudes between samples, which could be reflective of their similar local structural environments. In the

Figure 4. Experimental data (empty circles) and fits (solid red lines) of Fourier-transformed Ag K-edge EXAFS spectra.

Figure 3. (left) Raw EXAFS and (right) Fourier-transformed EXAFS spectra from AgPt NP samples acquired at (top) the Ag K-edge and (bottom) Pt L3-edge. D



Article

numbers. The low value of the former, especially in the Ag93Pt7 sample, is due to the relatively small amount of Ag−Pt interaction from the Ag perspective, which occurs only at the NP surface. Furthermore, based on the NP sizes measured using TEM, a large proportion of the Ag atoms must exist in the core, where they have little to no interaction with Pt. By contrast, a greater degree of coordination is observed for the Pt−Ag interaction from the Pt perspective, due again to the relatively lesser concentration of Pt in the particles. Overall, the proportion of Pt atoms which experience direct bonding to Ag is significantly larger, making its influence on Pt much more significant. Taken together, and in conjunction with the SPR band extinction observed in optical spectroscopy, these observations provide strong evidence in support of the formation of small Pt clusters on the surface of the (now significantly etched) Ag NP cores. The preference of Pt to form small clusters rather than a more homogeneous alloy can be attributed to its tendency to form shorter, and therefore stronger, bonds with other Pt atoms than with those of Ag. Although the formation of a continuous Pt monolayer at the particle surface could also satisfy the observed Pt−Pt and Pt− Ag CNs, such an occurrence is unlikely due to the significant surface roughening suggested by the low Ag−Ag CNs, particularly in the samples with the highest Pt contents. The only clear trends in terms of bond distances were observed for the Ag−Ag scattering path, where a gradual decrease was observed as Pt content was increased. This effect is likely due to lattice contraction caused by the electronic effects of having greater quantities of Pt present in these samples. Based on the structural information provided by the EXAFS fitting results, it is possible to construct a picture of the Pt cluster growth mechanism. Island growth mechanisms are common for metal NPs formed on a variety of substrate materials,20−22 and so it can be reasonably assumed that a similar process could be at work in these AgPt NPs. As illustrated in Figure 6, it is proposed that small Pt clusters initially nucleate at the Ag surface, become larger as the concentration of Pt is increased, and finally continue to increase their surface density once a maximum cluster size has been reached. The Pt−Pt CN data in Table 2 can also offer quantitative information on the structural evolution of the Pt

Figure 5. Experimental data (empty circles) and fits (solid red lines) of Fourier-transformed Pt-L3-edge EXAFS spectra.

The obtained fitting parameter values indicated a decrease in the total Ag−Ag CN as Pt was incorporated into the Ag NPs, likely due to etching from the galvanic reduction of Pt. The extent of this etching is extensive due to the fact that (as indicated by the corresponding partial redox reaction: Pt4+(aq) + 4Ag0(s) → Pt0 + 4Ag+(aq)) four atoms of silver must be oxidized in order to reduce a single atom of Pt. In reality, the reaction occurring is likely more complicated, thanks to the reductive action of PVP,19 which could lead to conversion of Pt4+ species to Pt2+ and/or rereduction of Ag+ to Ag0 on the NP surfaces. Investigation of the specific role of played by PVP in this reduction process could prove interesting for future research. In contrast, the Pt−Pt coordination was observed to initially increase between compositions of Ag93Pt7 and Ag81Pt19 and then remain constant despite a continued increase in relative Pt content. While this result may at first appear to be unusual, it can be interpreted as indicating an increase in the density of clusters of Pt atoms rather than a constant growth in their size. The degree of interaction between the Ag and Pt components of the NP are reflected by Ag−Pt and Pt−Ag coordination

Table 2. Parameter Values Obtained from Fitting Ag K-Edge and Pt L3-Edge EXAFS Spectra sample

edge

path

CN

R (Å)

σ2 (10−3 Å2)

ΔE0 (eV)

Ag NPs Ag93Pt7

Ag K Ag K

Ag−Ag Ag−Ag Ag−Pt Pt−Pt Pt−Ag Ag−Ag Ag−Pt Pt−Pt Pt−Ag Ag−Ag Ag−Pt Pt−Pt Pt−Ag Ag−Ag Ag−Pt Pt−Pt Pt−Ag

9.9(8) 10.0(5) ∗∗ 2(1) 6(2) 7(1) 1(1) 5(1) 3.0(7) 7(2) 1(1) 5(1) 4(1) 5(1) 2(1) 6(1) 3.0(6)

2.875(3) 2.872(2) ∗∗ 2.77(5) 2.84(1) 2.847(9) 2.81(2)* 2.77(1) 2.81(2)∗ 2.85(1) 2.83(1)* 2.78(2) 2.83(1)∗ 2.83(1) 2.81(1)∗ 2.77(1) 2.81(1)∗

2.8(4) 3.2(2) ∗∗ 3.1(1)

2.5(7) 2.9(4) ∗∗ 5(2)

4.2(9)

1(1)

2(2)

2(2)

4(1)

1(1)

3(2)

4(2)

5(2)

−1(1)

2(1)

3(2)

Pt L3 Ag81Pt19

Ag K Pt L3

Ag72Pt28

Ag K Pt L3

Ag65Pt35

Ag K Pt L3

E



Article

Figure 6. Proposed growth scheme for Pt clusters at the surface of Ag seed NPs: Pt ions are reduced onto the initial Ag surface (a) to form very small clusters (b), which grow larger when sufficient quantities of Pt ions are available in solution (c).

clusters. The Ag93Pt7 sample has a Pt−Pt CN of 2, suggesting a Pt3 (Pt−Pt CN = 2) cluster structure illustrated in Figure 6. Interestingly, for higher Pt concentrations, the other three samples (i.e., Ag81Pt19, Ag72Pt28, and Ag65Pt35) all show similar values of 5−6 for the Pt−Pt CN, which may roughly correspond to a Pt13 (Pt−Pt CN = 5.5) cluster model shown in Figure 6. For the four samples obtained in the present work, no intermediate structures (such as tetrahedral or planar Pt4 with a Pt−Pt CN = 3) were observed. This implies that certain structures (like Pt3 and Pt13) may be more favorable during the Pt cluster growth process. Additionally, it is interesting to discuss how the Pt:Ag reaction mechanism (1:4 [Pt4+] or 1:2 [Pt2+]) may affect the resulting Ag−Pt NP structure. If a 1:4 Pt:Ag mechanism were dominant during the reaction, the resulting Ag−Pt particles with high Pt concentration (e.g., Ag72Pt28 sample) would be considerably smaller than the pure Ag NPs. This is because the reduction of Pt4+ ions would result in the loss of greater number of Ag atoms from the Ag seeds than would the reduction of Pt2+. However, our TEM results on mean NP diameter (Figure 1b) indicate that there is no noticeable size decrease for samples with higher Pt concentrations, suggesting that the 1:2 Pt:Ag mechanism is more favorable. It also provides evidence to support the reductive potential of PVP under these reaction conditions, as discussed previously. 3.3. XANES Analysis and Simulation. In order to probe the electronic properties of the NPs, XANES spectra for the AgPt NPs were acquired. Ag K-edge spectra (Figure 7a) closely resembled that of the Ag foil reference, indicating their metallic structure and ruling out significant oxidation or contamination with unreduced metal salts. The white line in the Pt L3-edge spectra (Figure 7b) arises from an excitation of core p electrons to vacancies in the valence d level, and so its intensity can be used to monitor the oxidation state of Pt. This white line intensity is sensitive enough to make it a useful tool for probing the structural and electronic properties of Pt sites in nanomaterials.23 The general shape of the AgPt NP XANES spectra was observed to be very similar to that of the reference Pt foil, again indicating their metallic character. Interestingly, a slight decrease and broadening of the Pt white line peak were observed in the Pt L3-edge spectra of AgPt NP samples (Figure 7b) relative to the Pt foil reference; this indicates the possibility of charge transfer from Ag to Pt and is most significant by far for Ag93Pt7. As mentioned previously, the more significantly reduced white line intensity and increased peak breadth of the Ag93Pt7 spectrum stand out as unusual. In order to better understand the origin of these features, FEFF8 was used to perform ab initio simulations of XANES spectra, which were then

Figure 7. (left) Overlain and (right) stacked plots of Ag K-edge (a) and Pt L3-edge (b) XANES spectra from AgPt NP samples.

compared with the experimental data. Before these simulations could be performed, however, it was necessary to generate reasonable structural models to reflect the structure of Pt sites in the Ag93Pt7 NPs. Based on CN and bond length parameter values obtained from EXAFS data fitting, a small (three-atom) Pt cluster on the Ag surface was selected as a representative model for the Ag93Pt7 sample. Experimental Ag93Pt7 and simulated Pt cluster spectra are presented together in Figure 8a. In order to verify the reliability of the simulation, a bulk Pt spectrum was also generated and compared with experimental data from a Pt foil (see Figure 8b). It can be seen by comparison that the simulated spectrum of small Pt clusters on Ag closely resembles that of the Ag93Pt7 NPs. Also with respect to the white line peak shape of Ag93Pt7, the observed differences relative to bulk Pt could be ascribed to some combination of the Pt−Ag bonding interaction and the F



Article

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge financial support from NSERC Canada. PNC/XSD facilities at the Advanced Photon Source (APS) and research at these facilities are supported by the U.S. Department of Energy−Basic Energy Sciences, a Major Resources Support grant from NSERC, the University of Washington, the Canadian Light Source, and the Advanced Photon Source. Use of the APS, an Office of Science User Facility operated for the U.S. Department of Energy (DOE) Office of Science by Argonne National Laboratory, was supported by the U.S. DOE under Contract DE-AC0206CH11357. Beamline scientist Dr. Robert Gordon is also acknowledged for his assistance with technical support and XAS data acquisition.

Figure 8. Experimental (black) and simulated (red) Pt L3-edge XANES spectra for (a) a Pt cluster on an Ag surface (dashed line: “free” Pt cluster) and (b) bulk Pt. The corresponding structural models of the Pt3 cluster (c) and (d) bulk Pt are also shown for reference.

■

small size of the Pt clusters; thus, it was further undertaken to determine which of the two provided a more significant contribution. By examining the simulated spectrum of a “free” Pt cluster with no supporting Ag surface (also shown in Figure 8a), it is clear that virtually no resemblance to the experimental data exists. It can thus be concluded that the interesting electronic behavior of Ag93Pt7 arises primarily as a result of Ag− Pt bonding interactions, with little influence due to the small cluster size alone. Moreover, it can be inferred from these results that the selected Pt cluster model is a reasonable representation of Pt sites in the Ag93Pt7 NP sample, verifying its reliability.

4. CONCLUSIONS To summarize, we have prepared a series of AgPt NPs consisting of small Pt clusters formed on the surfaces of Ag seed particles. As was confirmed by elemental analysis, the successful synthesis of AgPt alloy NPs with varying compositions was achieved through the use of a simple galvanic replacement reaction. Evidence supporting the existence of Ag−Pt bonding in these bimetallic NPs (as opposed to formation of discrete Ag and Pt particles) was provided by multiple characterization techniques, including TEM, UV−vis absorption, EXAFS data fitting, and analysis of experimental and simulated XANES spectra. More importantly, structural models generated using information gleaned from EXAFs data fitting allowed the unusual electronic structure observed in the Ag93Pt7 XANES spectrum to be reproduced in ab initio simulated spectra, illustrating the important synergy between these techniques in the determination of local structure and electronic properties of nanomaterials. Finally, it was demonstrated that small Pt clusters with controlled size and alloyed bonding could be formed on Ag nanoparticle surfaces, which may be useful materials for further catalytic studies. The detailed information presented herein on the local structure, electronic properties, and formation mechanism of AgPt nanoparticles could also be useful for better understanding the structure−property relationship of bimetallic nanoparticles with Pt-enriched surfaces.

■

REFERENCES

(1) Huber, G. W.; Iborra, S.; Corma, A. Synthesis of Transportation Fuels from Biomass: Chemistry, Catalysts, and Engineering. Chem. Rev. 2006, 106, 4044−4098. (2) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Activity Benchmarks and Requirements for Pt, Pt-Alloy, and Non-Pt Oxygen Reduction Catalysts for PEMFCs. Appl. Catal., B 2005, 56, 9−35. (3) Wasmus, S.; Ku, A. Methanol Oxidation and Direct Methanol Fuel Cells: A Selective Review. J. Electroanal. Chem. 1999, 461, 14−31. (4) Aricò, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J.-M.; van Schalkwijk, W. Nanostructured Materials for Advanced Energy Conversion and Storage Devices. Nat. Mater. 2005, 4, 366−377. (5) Luo, J.; Yin, J.; Loukrakpam, R.; Wanjala, B. N.; Fang, B.; Shan, S.; Yang, L.; Nie, M.; Ng, M. S.; Kinzler, J.; et al. Design and Electrochemical Characterization of Ternary Alloy Electrocatalysts for Oxygen Reduction Reaction. J. Electroanal. Chem. 2013, 688, 196−206. (6) Loukrakpam, R.; Luo, J.; He, T.; Chen, Y.; Xu, Z.; Njoki, P. N.; Wanjala, B. N.; Fang, B.; Mott, D.; Yin, J.; et al. Nanoengineered PtCo and PtNi Catalysts for Oxygen Reduction Reaction: An Assessment of the Structural and Electrocatalytic Properties. J. Phys. Chem. C 2011, 115, 1682−1694. (7) Duchesne, P. N.; Chen, G.; Zheng, N.; Zhang, P. Local Structure, Electronic Behavior, and Electrocatalytic Reactivity of CO-Reduced Platinum−Iron Oxide Nanoparticles. J. Phys. Chem. C 2013, 117, 26324−26333. (8) Zhang, Q.; Xie, J.; Liang, J.; Lee, J. Y. Synthesis of Monodisperse Ag-Au Alloy Nanoparticles with Independently Tunable Morphology, Composition, Size, and Surface Chemistry and Their 3-D Superlattices. Adv. Funct. Mater. 2009, 19, 1387−1398. (9) Wanjala, B. N.; Fang, B.; Luo, J.; Chen, Y.; Yin, J.; Engelhard, M. H.; Loukrakpam, R.; Zhong, C.-J. Correlation between Atomic Coordination Structure and Enhanced Electrocatalytic Activity for Trimetallic Alloy Catalysts. J. Am. Chem. Soc. 2011, 133, 12714− 12727. (10) Chen, G.; Zhao, Y.; Fu, G.; Duchesne, P. N.; Gu, L.; Zheng, Y.; Weng, X.; Chen, M.; Zhang, P.; Pao, C.-W.; et al. Interfacial Effects in Iron-Nickel Hydroxide-Platinum Nanoparticles Enhance Catalytic Oxidation. Science 2014, 344, 495−499. (11) Zhang, Q.; Lee, J. Y.; Yang, J.; Boothroyd, C.; Zhang, J. Size and Composition Tunable Ag−Au Alloy Nanoparticles by Replacement Reactions. Nanotechnology 2007, 18, 245605. (12) Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W. NIH Image to ImageJ: 25 Years of Image Analysis. Nat. Methods 2012, 9, 671. (13) Ravel, B.; Newville, M. ATHENA, ARTEMIS, HEPHAESTUS: Data Analysis for X-Ray Absorption Spectroscopy Using IFEFFIT. J. Synchrotron Radiat. 2005, 12, 537−541. (14) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. RealSpace Multiple-Scattering Calculation and Interpretation of X-RayAbsorption Near-Edge Structure. Phys. Rev. B 1998, 58, 7565.

AUTHOR INFORMATION

Corresponding Author

*Tel +1-902-494-3323; e-mail [email protected] (P.Z.). G



Article

(15) Newville, M.; Boyanov, B. I.; Sayers, D. E. Estimation of Uncertainties in XAFS Data. J. Synchrotron Radiat. 1999, 6, 264−265. (16) Verbruggen, S. W.; Keulemans, M.; Martens, J. A.; Lenaerts, S. Predicting the Surface Plasmon Resonance Wavelength of Gold−Silver Alloy Nanoparticles. J. Phys. Chem. C 2013, 117, 19142−19145. (17) Liz-Marzán, L. M. Nanometals: Formation and Color. Mater. Today 2004, 7, 26−31. (18) Duchesne, P. N.; Zhang, P. Local Structure of Fluorescent Platinum Nanoclusters. Nanoscale 2012, 4, 4199. (19) Wu, C. W.; Mosher, B. P.; Lyons, K.; Zeng, T. F. Reducing Ability and Mechanism for Polyvinylpyrrolidone (PVP) in Silver Nanoparticles Synthesis. J. Nanosci. Nanotechnol. 2010, 10, 2342− 2347. (20) Wawro, A.; Sobańska, M.; Petroutchik, A.; Baczewski, L. T.; Pankowski, P. Self-Assembled Growth of Au Islands on a Mo(110) Surface. Nanotechnology 2010, 21, 335606. (21) Finetti, P.; Dhanak, V. R.; Binns, C.; Edmonds, K. W.; Baker, S. H.; D’Addato, S. Structural and Magnetic Properties of Self-Assembled Nanoscale Fe Islands on Cu(100). J. Electron Spectrosc. Relat. Phenom. 2001, 114−116, 251−256. (22) Kong, D.; Wang, G.; Pan, Y.; Hu, S.; Hou, J.; Pan, H.; Campbell, C. T.; Zhu, J. Growth, Structure, and Stability of Ag on CeO 2 (111): Synchrotron Radiation Photoemission Studies. J. Phys. Chem. C 2011, 2, 6715−6725. (23) Bayindir, Z.; Duchesne, P. N.; Cook, S. C.; MacDonald, M. A.; Zhang, P. X-ray Spectroscopy Studies on the Surface Structural Characteristics and Electronic Properties of Platinum Nanoparticles. J. Chem. Phys. 2009, 131, 244716.

H


Element-Specific Analysis of the Growth Mechanism, Local Structure

Recommend Documents