Electrochemical Synthesis and Catalytic Property of Sub-10 nm

Mar 18, 2010 - (1-3) For example, both Pt nanotubes and hollow spheres can exhibit ... It is still quite challenging to make shape-controlled metal ho...
2 downloads 0 Views 4MB Size
Download PDF
pubs.acs.org/NanoLett

Electrochemical Synthesis and Catalytic Property of Sub-10 nm Platinum Cubic Nanoboxes Zhenmeng Peng,†,§ Hongjun You,†,‡,§ Jianbo Wu,† and Hong Yang*,† †

Department of Chemical Engineering, University of Rochester, Rochester, New York 14627-0166 and ‡ State Key Laboratory for Mechanical Behavior of Materials, Materials Science and Engineering School, Xi’an Jiaotong University, Xi’an, Shannxi 710049, People’s Republic of China ABSTRACT We report an electrochemical synthesis of ultrafine Pt cubic nanoboxes from Pt-on-Ag heteronanostructures. These cubic nanoboxes have an average edge length of about 6 nm and a wall thickness of 1.5 nm. Several reaction parameters including the profile of applied potentials were examined to develop an optimal procedure for controlling the size, shape, and surface morphology of the nanoboxes. A strong shape-dependent catalytic property is observed for Pt cubic nanoboxes, which is 1.5 times more active than hollow spheres in terms of turn over frequency for catalytic oxidation of methanol. KEYWORDS Platinum, nanobox, nanocube, catalysis, methanol oxidation

M

etal nanostructures with hollow interior can have special structural properties, such as high specific surface, low density, and high void ratio, all of which can be interesting for electronic, optical, catalytic, and biological applications.1-3 For example, both Pt nanotubes and hollow spheres can exhibit higher activity than its solid particles in oxygen reduction reaction (ORR) and methanol oxidation reaction (MOR).4-7 Hollow Pd nanospheres have been used in catalyzing Suzuki cross coupling reaction.8 Gold nanocages have been conjugated with monoclonal antibodies for biological imaging and cancer therapy.9-11 Gold nanocages can also be modified by polymers to enable a controlled release of anticancer drugs and enzymes with near-infrared light.3 Metal hollow nanostructures can be made in several different forms including box, cage, sphere, and tube. Templating method has long been used as the main approach to the synthesis of metal hollow nanostructures.1-4,9,12,13 In general, metal nanoparticles are used as both template and support for depositing targeted metal elements. The bestknown method for metal deposition in such cases is the galvanic replacement, in which metal shells are generated from salt precursors through oxidation of the metal cores. The shape of resulting hollow structures is largely determined by that of the metal core. Currently, the achievable size of those shape-controlled hollow nanostructures are usually in the range of tens of nanometers although smaller hollow structures would be preferred in catalysis and many

other applications. Small hollow nanostructures can have large specific surface areas, which are necessary for obtaining high mass specific activity. It is still quite challenging to make shape-controlled metal hollow structures in the sub-ten nm-sized regime. There is no reliable method available for producing cubic or other crystalline surface-specific metal nanoboxes at this size range. In this paper, we present a new method for producing Pt cubic nanoboxes that have average edge length of 6 nm and wall thickness of 1.5 nm which is equal to the length of about 4 unit cells. These cubic nanoboxes are produced through electrochemically restructuring Pt hollow nanospheres that are made using truncated octahedral and other faceted Ag nanoparticles as the templates. We further demonstrate shape-dependent catalytic property of these nanoboxes in MOR. The enhanced catalytic activity can be attributed to the cubic morphology, in which {100} facets are dominated and preferred for catalytic oxidation of methanol. The Pt nanoboxes were prepared by treating Pt-on-Ag nanoparticles electrochemically using different potential cycling profiles. The Pt-on-Ag nanoparticles were made via a sequential approach (see Supporting Information).14-16 Briefly, Ag nanoparticles were first synthesized by reducing silver trifluoroacetate (AgTFA) in isoamyl ether in the presence of oleylamine (OAm) at 160 °C.13,17-20 These faceted Ag nanoparticles had an average diameter of 9.5 nm and were used as the supports for the subsequent deposition of Pt nanoparticles (Figure S1a). Platinum acetylacetonate (Pt(acac)2) was reduced into Pt which grew on the surface of Ag nanoparticles at 180 °C in a mixture of diphenyl ether (DPE) and OAm to form Pt-on-Ag heteronanostructures (Supporting Information Figure S1b). These Pt-on-Ag nanoparticles were loaded onto carbon black (Vulcan XC-72) before they

* To whom correspondence should be addressed. E-mail: hongyang@ che.rochester.edu. Telephone: (585) 275-2110. Fax: (585) 273-1348. § These authors contributed equally to this work. Received for review: 02/15/2010 Published on Web: 03/18/2010

© 2010 American Chemical Society

1492

DOI: 10.1021/nl100559y | Nano Lett. 2010, 10, 1492–1496

FIGURE 2. (a) HR-TEM image, (b) EDX spectrum, (c) HAADF-STEM image with a Pt-M line scan, and (d) TEM images taken at different rotating angles between the direction of the imaging beam and the plane of the sample stage. All TEM images were taken with the same Pt hollow cube.

FIGURE 1. Representative TEM images of Pt hollow nanocubes at (a) low and (b) high magnifications and (c) individual cubes imaged under various tilting angles with respect to the direction of imaging beam. Inset in (a) shows the corresponding potential cycling profile.

of TEM grid (Figure 1c). The difference in contrast could largely disappear between core and shell regions for those boxes aligned in certain angles with respect to the imaging beam. To further elucidate the structure and chemical composition of nanoboxes, we conducted the HR-TEM and energy dispersive X-ray (EDX) analysis with individual nanoboxes (Figure 2). For the batch of cubic nanoboxes including the one shown in Figure 2a, the EDX analysis shows they were made of pure Pt, and no Ag was detected in the final product (Figure 2a,b). The observed Cu signals were from the copper grid used in our EDX study. HAADF-STEM image and a corresponding EDX line scan show the Pt elemental distribution (Figure 2c). The sharp difference in contrast between the center and edge in the HAADF-STEM image indicates the structure was hollow, an observation that is further supported by the saddle-type shape of the Pt M-line curve with the weakest signal in the center region. By rotating the TEM grid along either x- (R) or y- (β) axis at a step angle of 10°, we could unambiguously confirm that the Pt particle was hollow despite that there existed low contrast difference across the cubes under some imaging conditions, such as at β ) 30° (Figure 2d). Both square and rhombic shapes were observed for the same hollow cube when the tilting angle changed. In this particular case, the cube was projected as a rhombic shape, when the tilting angles were at β ) -10 and -30°, respectively. The nearly perfect cubic shape could be projected as if it had a truncated shape with rounded

were treated with acetic acid to remove the residual surface capping agents.14,21 Twenty potential cycles were carried out between 0.0 and 1.3 V in a 0.1 M perchloric acid (HClO4) aqueous solution to remove the Ag metal cores. The sample was then transferred into a 0.5 M sulfuric acid (H2SO4) aqueous solution for further electrochemical treatment. Figure 1 shows representative TEM images of the nanoboxes made after the treatment. To obtain the Pt nanoboxes shown in Figure 1a, 3000 linear sweeps were carried out in a potential range between 0.6 and 1.0 V and at a scan rate of 100 mV/s. Most of the obtained Pt nanostructures were cube-like and had fairly uniform size. The average edge length of the cubes was about 6 nm. The outside and inside surfaces of walls seemed parallel with the average wall thickness of about 1.5 nm or four unit cell length. Highresolution TEM (HR-TEM) image of an individual nanobox reveals that the d-spacing was 2.01 Å for those lattices normal to the surface, and 2.26 Å for those at 45° angle with respect to the surface (Figure 1b). These fringes can be indexed to {100} and {111} planes of pure Pt metal, indicating that the hollow structures were Pt nanoboxes with both inner- and outer-sides bound by {100} surfaces. As TEM images are two-dimensional (2D) projections of 3D objects, cubic nanoboxes can be imaged as a range of different shapes, ranging from rectangle to hexagon depending on the angles between the imaging beam and the plane © 2010 American Chemical Society

1493

DOI: 10.1021/nl100559y | Nano Lett. 2010, 10, 1492-–1496

FIGURE 4. (a) TEM and (b) HR-TEM images of the Pt hollow nanospheres obtained after the removal of Ag cores from Pt-on-Ag heteronanostructures.

became solid and had even smaller sizes after 1500 scans and eventually turned into small and dense 2 nm Pt nanoparticles after 3000 cycles (Supporting Information Figure S2c). This morphological change can be readily understood, as the electrochemical oxidation of Pt metal occurs at about 1.2 V, which could result in rapid dissolution of Pt species.22,23 Thus, a potential of 1.2 V was too harsh for electrochemical conversion into Pt cubic nanoboxes. The potential scanning rate was examined to determine if the reaction kinetics could play a role in the formation of nanoboxes. We observed that if the scan rate was slowed down from 100 to 20 mV/s, the population of the cubic nanoboxes was reduced and sizes of the Pt nanoparticles became nonuniform (Figure 3c). If the scan rate was too fast, well-defined cubic nanoboxes could not be generated either. The shape of scan profile also affected the final Pt nanostructures dramatically. Previously, square wave potential was used to control the shape of Pt crystal electrochemically,24 so we examined the restructuring of the shape of Pt hollow spheres using such scan profile in the potential range between 0.6 and 1.0 V. Quite a few of the Pt shells collapsed into solid nanoparticles after 3000 cycles (Figure 3d), indicating the square wave was not the optimal cycling profile for the conversion to nanoboxes. This observation suggests that continuous increase and decrease of cycling potential were necessary in maintain the structural integrity of the nanoboxes and abrupt change in cycling potential was detrimental to the formation of hollow structures. The above observations indicate careful control of the reaction kinetics is necessary for obtaining well-defined cubic nanoboxes. As the Ag nanoparticle templates were not cubic, extensive restructuring of Pt shells had to occur during the electrochemical treatment. Figure 4 shows the TEM images of Pt nanostructures produced right after the initial twenty cycles of CV treatment in 0.1 M HClO4 aqueous solution. The TEM image shows the formed hollow structures had pseudospherical shapes, resembling those of the particle templates (Figure 4a). Small nanoparticles could readily be found on the surface of these hollow nanoparticles (Figure 4b). The overall size of these hollow structures was comparable with those of as-made Pt-on-Ag nanoparticles, and no obvious shrinkage was observed at this stage. EDX analyses show no detectable signal of Ag element in the sample

FIGURE 3. Representative TEM images of Pt nanostructures obtained under various potential cycling profiles. (a-c) Triangular wave with different potential ranges and scan rates and (d) square wave. Insets illustrate the details of the corresponding cycling profiles.

corners in 2D images. For instance, at the tilting angle of R ) 30° or β ) 20°, parts or all of the sharp external corners of the square shown in Figure 2a became round. As the Ag nanoparticle templates had truncated octahedral and other noncubic shapes, the formation of Pt cubic nanoboxes had to be different from the Au boxes made through galvanic replacement reaction from Ag cubes.1,9,10,13 Experimentally, we observed that several parameters of the electrochemical process could play important roles in controlling the morphology of Pt hollow nanostructures. The scan rate, number of cycle, range, and profile of the applied potentials are some of the critical ones. Change in one or more of these parameters could result in the formation of ill-defined morphologies other than cubic nanoboxes (Figure 3). If the potential range of CV cycles was changed to between 0.6 and 0.8 V, hollow nanospheres instead of cubic nanoboxes were obtained (Figure 3a). Under this condition, the surface Pt atoms show limited mobility and the hollow shells largely preserved those shapes of the Ag templates. Truncated hollows with rough surfaces could be observed, with no extended restructuring of Pt shells. On the other hand, if the scan range changed to between 0.6 and 1.1 V, Pt shells formed size-reduced hollows or even collapsed into solid nanoparticles. The hollow particles formed under this condition were largely spherical and had smooth surfaces (Supporting Information Figure S2a). The size reduction was accelerated with a further increase of the scan range to between 0.6 and 1.2 V (Figure 3b). Hollow nanospheres with reduced size could be observed after 800 linear scans (Supporting Information Figure S2b). These hollow spheres © 2010 American Chemical Society

1494

DOI: 10.1021/nl100559y | Nano Lett. 2010, 10, 1492-–1496

surface species, which were associated with the potential cycling, should result in the movement of Pt surface species and the formation of stable cubic nanoboxes.24,30,31 To examine the effect of shape on the catalytic activity of these Pt hollow nanostructures, we chose the electrochemical oxidation of methanol, as this reaction is sensitive to the surface electronic structure and atomic arrangement of Pt.32,33 The electrochemical active surface areas (ECSAs) of the Pt cubic nanoboxes and hollow nanospheres, as shown respectively in Figures 1a and 4a, were calculated by measuring the hydrogen adsorption on Pt (Figure 5). To be specific, ECSA value was obtained by integrating the shaded area between 0.05 and 0.4 V (vs RHE) in the cyclic voltammetry (CV) curve. The Pt hollow nanoshperes had an ECSA of 82.4 m2/gPt, close to that of the state-of-the-art commercial Pt/C catalysts that are made of 2-3 nm Pt nanoparticles on carbon.34 The ECSA of Pt cubic nanoboxes was 58.0 m2/gPt, indicating some area loss occurred during the surface restructuring. Interesting, the shape of CV curves was different between the Pt cubic nanoboxes and hollow spheres. This difference could be attributed to the effect of morphology, because hydrogen species adsorbs and desorbs on the lowindex planes of Pt at different potential ranges.27,35,36 The electrocatalytic property of both Pt cubic nanoboxes and hollow nanospheres in MOR was tested and compared with that of a commercial Pt reference catalyst (TKK, 46.7 wt %). There was little difference in the shape and peak potentials of their CV curves among these three catalysts, suggesting the reaction pathway should be similar (Figure 6a). All three show the characteristic peaks for pure Pt in the forward and backward scans. The specific activity for Pt cubic nanoboxes was 1.10 mA/cm2Pt at its peak potential of 0.85 V; this value is much higher than 0.74 mA/cm2Pt for the hollow nanospheres and 0.63 mA/cm2Pt for the Pt reference catalyst. This result indicates that Pt cubic nanoboxes have a better intrinsic activity than the hollow spheres. We further analyzed the turnover frequency (TOF) of these catalysts at their peak potentials (Figure 6b). The TOF is defined as the number of oxidized methanol molecules per Pt surface site per second and estimated using the oxidation current in the

FIGURE 5. The CV curves of Pt cubic nanoboxes and hollow nanoparticles in a 0.5 M sulfuric acid aqueous solution. The inset shows the enlarged region for hydrogen desorption.

(Supporting Information Figure S3). Mixed cubic boxes and hollow spheres were observed after the hollow nanoparticles were treated electrochemically with linear potential scan for 1000 cycles in a 0.5 M H2SO4 solution (Supporting Information Figure S4a). The resulting hollow particles shrank in size after the treatment, suggesting the restructuring process was gradual. Once the cubic nanoboxes were formed, they became stable and showed little variation in shape between those obtained after 3000 and 6000 cycles (Supporting Information Figure S4b). Thus, the formation of cubic nanoboxes was most likely driven by the relative stability of the low-indexed surfaces of Pt metal in H2SO4 solution. This mode of formation could be understood, as the adsorption of hydroxyl group on Pt varies among the different lowindex surfaces and Pt (100) surface is more difficult to restructure than the other low-index surfaces using a linear potential cycle.25,26 There also exists a large difference in chemisorption of (bi)sulfate anions among the three lowindex surfaces of Pt, which could also contribute to the restructuringprocess.27-29 TheformationofPtcubicnanoboxes in electrolytes other than sulfuric acid, perchloric acid, and nitric acid in the experiments, suggests the critical role of hydroxyl groups, while their lower yields suggest a synergistic effect of (bi)sulfate anions (Supporting Information Figure S5). A continuous adsorption-desorption of these

FIGURE 6. (a) Methanol oxidation reaction catalyzed by Pt cubic nanobox, hollow nanosphere, and the reference catalysts, and (b) their turnover frequencies (TOF) at the peak potentials. © 2010 American Chemical Society

1495

DOI: 10.1021/nl100559y | Nano Lett. 2010, 10, 1492-–1496

CV, the ECSA and the number of electrons transferred in the reaction. The Pt hollow nanospheres showed a small enhancement in TOF than the reference Pt catalyst by a factor of 1.2 (0.59 vs 0.48 s-1), while the Pt cubic nanoboxes exhibited a significantly higher TOF number of 0.89 s-1 at 0.85 V, or a 1.9-fold increase in catalytic activity. We attribute this improved activity to the cubic morphology, as Pt (100) surface has been known to be more active than (111) surface toward MOR.32,35,37,38 In conclusion, we have demonstrated a new electrochemical restructuring method for making ultrafine Pt cubic nanoboxes. It is intriguing that the mode of formation of cubic nanoboxes is the result of the most stable structures, which are controlled by the reaction conditions, rather than the shapes of original templates. Thus, both the size and shape uniformity of the obtained hollow cubes are controllable. Since the generation of nanoboxes is based on the difference in standard reduction potentials between two metals, and the final shape of the hollow nanostructures is the stable form under the designed conditions, this approach should be generic for producing cubic nanoboxes and other shape-controlled hollow nanostructures. As both the size (6 nm in edge length) and wall thickness (1.5 nm or 4 unit cell length) are among the smallest cubic nanoboxes with ultrathin walls available, these Pt hollow nanostructures should be of interest in the development of facet-specific catalysts that are often required for achieving the maximum selectivity and high activity.16,39,40

(6)

Acknowledgment. This work was supported by U.S. National Science Foundation (Grant DMR-0449849). It made use of Shared Facilities at University of Rochester River Campus EM Lab supported in part by DOE. Z.M.P. holds a Hooker Fellowship and H.J.Y. is a CSC Scholarship recipient.

(25)

(7) (8) (9) (10) (11)

(12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23)

(24)

(26) (27) (28) (29)

Supporting Information Available. Description of experimental details, TEM images of as-synthesized Ag and Pton-Ag nanoparticles (Figure S1), Pt nanostructures obtained under various synthetic conditions (Figures S2, S4, and S5), and EDX spectrum and STEM image of Pt hollow nanospheres (Figure S3). This material is available free of charge via the Internet at http://pubs.acs.org.

(30) (31) (32) (33)

REFERENCES AND NOTES

(34) (35)

(1) (2)

(36)

(3) (4) (5)

Sun, Y. G.; Mayers, B.; Xia, Y. N. Adv. Mater. 2003, 15, 641–646. Zhang, Q.; Wang, W. S.; Goebl, J.; Yin, Y. D. Nano Today 2009, 4, 494–507. Yavuz, M. S.; Cheng, Y. Y.; Chen, J. Y.; Cobley, C. M.; Zhang, Q.; Rycenga, M.; Xie, J. W.; Kim, C.; Song, K. H.; Schwartz, A. G.; Wang, L. V.; Xia, Y. N. Nat. Mater. 2009, 8, 935–939. Chen, Z. W.; Waje, M.; Li, W. Z.; Yan, Y. S. Angew. Chem., Int. Ed. 2007, 46, 4060–4063. Chen, H. M.; Liu, R. S.; Lo, M. Y.; Chang, S. C.; Tsai, L. D.; Peng, Y. M.; Lee, J. F. J. Phys. Chem. C 2008, 112, 7522–7526.

© 2010 American Chemical Society

(37) (38) (39) (40)

1496

Liang, H. P.; Zhang, H. M.; Hu, J. S.; Guo, Y. G.; Wan, L. J.; Bai, C. L. Angew. Chem., Int. Ed. 2004, 43, 1540–1543. Guo, S. J.; Dong, S. J.; Wang, E. Chem.sEur. J. 2008, 14, 4689– 4695. Kim, S. W.; Kim, M.; Lee, W. Y.; Hyeon, T. J. Am. Chem. Soc. 2002, 124, 7642–7643. Skrabalak, S. E.; Chen, J. Y.; Sun, Y. G.; Lu, X. M.; Au, L.; Cobley, C. M.; Xia, Y. N. Acc. Chem. Res. 2008, 41, 1587–1595. Song, K. H.; Kim, C. H.; Cobley, C. M.; Xia, Y. N.; Wang, L. V. Nano Lett. 2009, 9, 183–188. Turner, M.; Golovko, V. B.; Vaughan, O. P. H.; Abdulkin, P.; Berenguer-Murcia, A.; Tikhov, M. S.; Johnson, B. F. G.; Lambert, R. M. Nature 2008, 454, 981–U31. Mayers, B.; Jiang, X. C.; Sunderland, D.; Cattle, B.; Xia, Y. N. J. Am. Chem. Soc. 2003, 125, 13364–13365. Lu, X. M.; Tuan, H. Y.; Chen, J. Y.; Li, Z. Y.; Korgel, B. A.; Xia, Y. N. J. Am. Chem. Soc. 2007, 129, 1733–1742. Peng, Z. M.; Wu, J. B.; Yang, H. Chem. Mater. 2010, 22, 1098– 1106. Peng, Z. M.; Yang, H. Nano Res. 2009, 2, 406–415. Peng, Z. M.; Yang, H. Nano Today 2009, 4, 143–164. Wang, Y.; Wong, J. F.; Teng, X. W.; Lin, X. Z.; Yang, H. Nano Lett. 2003, 3, 1555–1559. Lin, X. Z.; Teng, X. W.; Yang, H. Langmuir 2003, 19, 10081– 10085. Wang, C.; Hu, Y. J.; Lieber, C. M.; Sun, S. H. J. Am. Chem. Soc. 2008, 130, 8902–8903. Wang, C.; Yin, H. G.; Chan, R.; Peng, S.; Dai, S.; Sun, S. H. Chem. Mater. 2009, 21, 433–435. Mazumder, V.; Sun, S. H. J. Am. Chem. Soc. 2009, 131, 4588– 4589. de Bruijn, F. A.; Dam, V. A. T.; Janssen, G. J. M. Fuel Cells 2008, 8, 3–22. Ferreira, P. J.; la O, G. J.; Shao-Horn, Y.; Morgan, D.; Makharia, R.; Kocha, S.; Gasteiger, H. A. J. Electrochem. Soc. 2005, 152, A2256–A2271. Tian, N.; Zhou, Z. Y.; Sun, S. G.; Ding, Y.; Wang, Z. L. Science 2007, 316, 732–735. Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G. F.; Ross, P. N.; Lucas, C. A.; Markovic, N. M. Science 2007, 315, 493–497. Clavilier, J.; Orts, J. M.; Feliu, J. M. J. Phys. IV 1994, 4, 303–308. Markovic, N. M.; Gasteiger, H. A.; Ross, P. N. J. Phys. Chem. 1995, 99, 3411–3415. Mostany, J.; Herrero, E.; Feliu, J. M.; Lipkowski, J. J. Phys. Chem. B 2002, 106, 12787–12796. Hoshi, N.; Sakurada, A.; Nakamura, S.; Teruya, S.; Koga, O.; Hori, Y. J. Phys. Chem. B 2002, 106, 1985–1990. Furuya, N.; Ichinose, M.; Shibata, M. J. Electroanal. Chem. 1999, 460, 251–253. Furuya, N.; Shibata, M. J. Electroanal. Chem. 1999, 467, 85–91. Hamnett, A. Catal. Today 1997, 38, 445–457. Lee, S. W.; Chen, S. O.; Sheng, W. C.; Yabuuchi, N.; Kim, Y. T.; Mitani, T.; Vescovo, E.; Shao-Horn, Y. J. Am. Chem. Soc. 2009, 131, 15669–15677. Peng, Z. M.; Yang, H. J. Am. Chem. Soc. 2009, 131, 7542–7543. Herrero, E.; Franaszczuk, K.; Wieckowski, A. J. Phys. Chem. 1994, 98, 5074–5083. Formo, E.; Peng, Z. M.; Lee, E.; Lu, X. M.; Yang, H.; Xia, Y. N. J. Phys. Chem. C 2008, 112, 9970–9975. Kita, H.; Gao, Y. Z.; Nakato, T.; Hattori, H. J. Electroanal. Chem. 1994, 373, 177–183. Sun, S. G.; Clavilier, J. J. Electroanal. Chem. 1987, 236, 95–112. Xia, Y.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int. Ed. 2009, 48, 60–103. Somorjai, G. A.; Frei, H.; Park, J. Y. J. Am. Chem. Soc. 2009, 131, 16589–16605.

DOI: 10.1021/nl100559y | Nano Lett. 2010, 10, 1492-–1496