NANO LETTERS
Triangular Nanoframes Made of Gold and Silver
2003 Vol. 3, No. 4 519-522
Gabriella S. Me´traux, Yunwei Charles Cao, Rongchao Jin, and Chad A. Mirkin* Department of Chemistry and Institute for Nanotechnology, Northwestern UniVersity, 2145 Sheridan Road, EVanston, Illinois 60208 Received February 19, 2003
ABSTRACT Silver nanoprisms (∼74-nm edge length) were etched with HAuCl4 to generate gold−silver alloy framework structures. Interestingly, these triangular nanoframes possess no strong surface plasmon bands in the ultraviolet or visible regions of the optical spectrum. By adding a mild reducing agent, L-ascorbic acid, gold and silver ions remaining in solution can be reduced, resulting in metal plating and reformation of nanoprisms. The extent of the back-filling process can be controlled, allowing one to form novel prisms with nanopores. This back-filling process is accompanied by a regeneration of the surface plasmon bands in the UV−vis spectrum.
Metallic nanoparticles have generated significant scientific and technological interest because of their unusual optical properties as well as their novel chemical and catalytic properties. Nonspherical particles, and in particular anisotropic ones, are of major interest because they allow one to investigate how shape affects the physical and chemical properties of such structures. A variety of shapes, including stars,1 cubes,2,3 rods,4-6 disks,7-9 and prisms,10-14 have been synthesized, and their properties have been preliminarily characterized. Hollow nanoparticles are an interesting emerging class of materials that will help one to understand the structure-property relationship in nanoparticles better. Although a significant amount of work has already been done in developing synthetic methods for hollow spheres, cubes, and rods,15-19 little has been done with triangular nanostructures. Herein, we show how one can chemically convert triangular nanoprisms into triangular nanoframes. Importantly, we demonstrate that this can be done in a faceselective manner to generate triangular two-component nanostructures with filled or partially filled cores. We recently reported a novel process in which silver nanospheres were converted, via a photomediated reaction, to larger triangular silver nanoprisms.10 The faces of the triangular nanoprisms have atomically flat faces ([111] crystal face) and faceted sides composed primarily of [110] crystal planes. We report here that a gold salt can be used to faceselectively etch the [111] faces of the silver nanoprisms to generate gold-silver triangular nanoframes (Scheme 1, step A). In a typical experiment, silver nanoprisms with ∼74nm edge lengths (σ ) 13%, N ) 200) and 9-nm thickness (σ ) 27%, N ) 46) were prepared by literature methods.10 * Corresponding author. E-mail:
[email protected]. Phone: (847) 491-2907. Fax (847) 467-5123. 10.1021/nl034097+ CCC: $25.00 Published on Web 03/15/2003
© 2003 American Chemical Society
The silver nanoprism colloid was synthesized from a 0.1 mM AgNO3 solution. For all experiments performed herein, Au/ Ag molar ratios (i.e., Au/Ag ) 1:9) were calculated with the assumption that the silver concentration was 0.1 mM. To generate Au/Ag nanoframes, 10 mL of silver nanoprisms was first diluted with Nanopure water (18.2 MΩ) to 1/5 the starting concentration. This was done to prevent aggregation of the resulting Ag/Au triangular nanoframes. Under ambient conditions, aqueous HAuCl4 (5 mM) was added dropwise to the rapidly stirred colloid. As the gold salt was added, the turquoise-blue color of the colloid gradually changed to purple and finally to blue or gray. Samples with low gold content (Au/Ag ) 1:9) formed pale-blue solutions and exhibited a low-intensity, broad surface plasmon band around 775 nm. In contrast, colloids containing high gold concentrations (Au/Ag ) 1:5, 1:3) were pale gray (essentially colorless) and displayed no strong surface plasmon bands in the UV-vis spectrum (Figure 1A). Transmission electron microscopy (TEM) images after gold addition revealed that the resulting nanostructures were triangular in shape with hollow centers (Figure 1B-D). Both the wall width and thickness were measured (Scheme 1). The wall width of the nanoframes increased slightly with gold content: 7.7 nm (σ ) 11%, N ) 245) for Au/Ag ) 1:9 and 10.3 nm (σ ) 21%, N ) 230) for Au/Ag ) 1:3. The thickness of the gold-silver nanoframes (10 nm, σ ) 20%, N ) 24) was similar to that of the pure silver nanoprisms from which they were derived (9 nm). TEM analysis at high magnification (200 000×) revealed that the center of each nanoframe was indeed hollow; the amorphous carbon film of the TEM support grid could be clearly seen in the underlying area. Tapping-mode AFM analysis (using
Scheme 1.
Scheme of Nanoframe Synthesisa
a In step A, silver nanoprisms are etched with aqueous HAuCl . Subsequent addition of L-ascorbic acid (step B) causes gold and silver 4 ions in solution to crystallize primarily on the inner walls of the nanoframes, causing the central pore to shrink in size. This gold salt/Lascorbic acid cycle (steps A + B) can be repeated to shrink the size of the central pore progressively.
Figure 1. Gold-silver nanoframes. (A) UV-vis spectra of triangular nanoframes with varying Au/Ag ratios. I: Silver nanoprisms; II: Au/Ag ) 1:9; III: Au/Ag ) 1:5; IV: Au/Ag ) 1:3. (B-D) TEM images of gold-silver nanoframes. (B) Au/Ag ) 1:9; (C) Au/Ag ) 1:5; (D) Au/Ag ) 1:3. Scale bars: (B, D) 100 nm; (C) 125 nm.
a Nanoscope IIIa AFM, Digital Instruments) also confirmed that these nanoframes are hollow structures. A variation of this etching strategy has been employed by Xia and co-workers to make hollow forms of cubes and rods,15,16 but to the best of our knowledge, this is the first procedure for making gold-silver frames from silver nanoprisms. This method takes advantage of the large difference in reduction potential of the two metals, the Ag+/Ag pair (0.8 V vs SHE) and the AuCl4-/Au pair (0.99 V vs SHE),15 to oxidize and etch the silver nanoprisms using gold ions in a type of nano-Galvanic cell reaction. In this type of reaction, at least three scenarios are possible. (1) Gold atoms selfnucleate to form gold nanoparticles, (2) gold crystallizes on 520
the [111] faces (top and bottom faces) of the silver nanoprisms, and (3) gold crystallizes on the faceted (mostly [110] planes) edges of the silver nanoprisms. Self-nucleation is a high-energy process that is unlikely to occur under our reaction conditions. Because they are atomically flat, the [111] crystal planes of the nanoprism faces are lower in surface energy than the faceted edges. Thus, gold crystallization on the [111] faces will result in a much larger energy increase of the system than gold plating on the edges. This reasoning accounts for why we retain the triangular shape of the initial silver nanoprisms while the gold salt etches the central silver matrix. Interestingly, our procecure does not yield hollow nanoprisms, as would be expected using Xia’s technique, but rather it generates triangularly shaped frames with solid walls and a hole in the center. This may be due to the fact that our initial particles are thinner (∼9 nm) or that our reaction conditions are milder (Xia uses elevated temperatures).15,16 Such face-selective etching is not observed under comparable conditions with other metal ions such as H2PtCl6 (Supporting Information), possibly because of the relatively large lattice mismatch between Pt and Ag (Pt ) 3.9231 Å, Ag ) 4.0862 Å).20 In addition, chloride ions play an important role in the etching process of silver nanoprisms. Gold precursors without chloride ions (i.e., gold acetate) do not demonstrate the same etching behavior as AuCl4-. This may be due to the fact that the formation of AgCl drives down the reduction potential of Ag+/Ag, favoring the oxidation of silver atoms by AuIII. Importantly, we have also discovered that these structures can be back-filled with gold to generate gold-silver alloy nanoprisms. To investigate the optical properties of the nanoframes more fully, we developed a protocol that would allow us to change the size of the central pore. A mild reducing agent, L-ascorbic acid, was used to reduce gold and silver ions in solution (generated from the first addition of gold salt) onto the triangular nanoframes, causing the walls to thicken and the central pore to shrink (Scheme 1, step B). Subsequent additions of HAuCl4 followed by L-ascorbic Nano Lett., Vol. 3, No. 4, 2003
Figure 2. UV-vis spectra and TEM images monitoring the backfilling process of Au/Ag ) 1:9 nanoframes. (A, D) After the addition of L-ascorbic acid to triangular nanoframes. (B, E) After two cycles of HAuCl4/L-ascorbic acid. (C, F) After three cycles of HAuCl4/L-ascorbic acid. Note how the pore size is becoming gradually smaller with an increasing number of cycles. Scale bars: (D, F) 100 nm; (E) 125 nm.
acid (Scheme 1, steps A + B) were performed to reduce the size of the triangular nanoframe pore progressively. In a typical experiment, an excess of L-ascorbic acid (1 mL, 5mM) was added dropwise to a rapidly stirring colloid of two-component nanoframes (50 mL, Au/Ag ) 1:9 nanoframes). After the addition of the reducing agent, the pale-blue colloid gradually became turquoise, as evidenced by an increase in intensity accompanied by a blue shift in the absorption band from 775 to 650 nm in the UV-vis spectrum of the solution (Figure 2A). The growth of a second band centered at 463 nm was also observed. The observed change in the UV-vis spectrum is consistent with Ag+ ions, generated from the initial etching process with gold, being reduced back onto the nanoframe. A second aliquot of HAuCl4 (22 µL, 5 mM) was added dropwise, followed by more L-ascorbic acid (1 mL, 5 mM). The surface plasmon bands associated with the partially filled triangles at 650 and 463 nm red shifted to 665 and 480 nm, respectively (Figure 2B). The third and final addition of HAuCl4 (22 µL, 5 mM) and L-ascorbic acid (1 mL, 5 mM) caused the most intense band at 665 nm to further red shift to 693 nm (Figure 2C). The shorter-wavelength band also red shifted to 508 nm but became a shoulder on the main surface plasmon band. The red shift observed after the second and third HAuCl4/Lascorbic acid additions is consistent with gold ions being reduced onto the nanoframe. Reduction of gold ions onto the nanoframe walls is responsible for the back-filling of the nanoframes to form alloy nanoprisms. The change in the pore size of the triangular nanoframes as a function of Au deposition was monitored by TEM (Figure 2D-F). Interestingly, as the pore closing occurs, the metal ions (gold and silver) seem to crystallize primarily on the inner walls of the nanoframe. After the first reduction, the pore size decreased from 33 (σ ) 23%, N ) 286) to 14 nm (σ ) 16%, N ) 695). The second addition of gold salt followed by reduction generated triangular nanoframes with an average pore size of 7 nm (σ ) 14%, N ) 744). After Nano Lett., Vol. 3, No. 4, 2003
the third gold/reduction cycle, many of the nanoframes were completely filled, and the remaining particles possessed an average pore size of 4 nm (σ ) 13%, N ) 659). After one cycle of HAuCl4/L-ascorbic acid, the thickness of the nanoframes increased from approximately 10 nm (σ ) 13%) to 12.4 nm (σ ) 11%, N ) 89). After three HAuCl4/Lascorbic acid cycles, the thickness had increased to 15.2 nm (σ ) 10%, N ) 8). The UV-vis spectrum of the filled nanoframes was red shifted and dampened with respect to the pure silver nanoprisms (Figure 2C). This is the same phenomenon observed in spherical Au-Ag alloy nanoparticles: the surface plasmon band of silver nanoparticles is red shifted and damped with increasing amounts of gold.21 Hence, in our case, UV-vis data indicate that Au-Ag alloy nanoprisms are formed. TEM-energy-dispersive X-ray (EDX) analysis confirmed that the back-filled nanoprisms were, in fact, gold-silver alloys (Supporting Information). As observed in the TEM images, the faces and edges of the gold nanoprisms are rough in texture, and their corners are truncated. The average edge length of the gold-silver alloy nanoprisms is approximately 63 nm. The primary reaction observed in the back-filling process is gold crystallization on the inner and outer edges of the Au/Ag nanoframe; the thickness of the nanoframes does not increase substantially. Hence, face-selective back-filling must also be taking place. This can also be explained in terms of surface energy. Because of their roughness, the inner edges of the Au/Ag nanoframes possess higher surface energy than the outer edges and flat faces. To minimize surface energy, gold will crystallize primarily on surfaces that will not greatly increase the energy of the system. For this reason, gold plating occurs predominantly on the inner walls of the Au/Ag nanoframes, leading to back-filling. In addition, there is less total surface area on the interior of the frames as compared to that on the exterior. Therefore, if the absolute amount of plated Au is compared for the exterior and interior surfaces, then backfilling will occur faster than total prism growth. This paper is important for the following reasons. First, it describes a method for forming a new class of nanostructures, bimetallic triangular nanoframes. Second, all of the data are consistent with gold salt-induced, face-selective etching and a novel back-filling process, which allow one to form triangular nanoframes and prisms with different compositions. Finally, these synthetic methods allow one to convert silver nanoprisms to gold-silver alloy nanoprisms, which are otherwise not accessible via thermal and photochemical methods for making monometallic nanoprisms.10-14 Acknowledgment. This work was supported primarily by the Nanoscale Science and Engineering Initiative of the National Science Foundation under NSF award number EEC0118025. C.A.M. also acknowledges ARO and DARPA for support of this research. Supporting Information Available: TEM image depicting the effect of H2PtCl6 addition on silver nanoprisms as well as EDX data of nanoframes after one and three cycles of HAuCl4 and L-ascorbic acid addition. This material is available free of charge via the Internet at http://pubs.acs.org. 521
References (1) Lee, S.-M.; Jun, Y.-W.; Cho, S.-N.; Cheon, J. J. Am. Chem. Soc. 2002, 124, 11244. (2) Petroski, J. M.; Green, T. C.; El-Sayed, M. A. J. Phys. Chem. A 2001, 105, 5542. (3) Ahmadi, T. S.; Wang, Z. L.; Green, T. C.; Henglein, A.; El-Sayed, M. A. Science (Washington, D.C.) 1996, 272, 1924. (4) Jana, N. R.; Gearheart, L.; Murphy, C. J. J. Phys. Chem. B 2001, 105, 4065. (5) Yu, Y.-Y.; Chang, S.-S.; Lee, C.-L.; Wang, C. R. C. J. Phys. Chem. B 1997, 101, 6661. (6) Kim, F.; Song, J. H.; Yang, P. J. Am. Chem. Soc. 2002, 124, 14316. (7) Puntes, V. F.; Krishnan, K. M.; Alivisatos, A. P. Science (Washington, D.C.) 2001, 291, 2115. (8) Maillard, M.; Giorgio, S.; Pileni, M.-P. AdV. Mater. 2002, 14, 1084. (9) Hao, E.; Kelly, K. L.; Hupp, J. T.; Schatz, G. C. J. Am. Chem. Soc. 2002, 124, 15182. (10) Jin, R.; Cao, Y.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.; Zheng, J. G. Science (Washington, D.C.) 2001, 294, 1901.
522
(11) Pinna, N.; Weiss, K.; Urban, J.; Pileni, M.-P. AdV. Mater. 2001, 13, 261. (12) Hulteen, J. C.; Treichel, D. A.; Smith, M. T.; Duval, M. L.; Jensen, T. R.; Van Duyne, R. P. J. Phys. Chem. B 1999, 103, 3854. (13) Bradley, J. S.; Tesche, B.; Busser, W.; Maase, M.; Reetz, M. T. J. Am. Chem. Soc. 2000, 122, 4631. (14) Chen, S.; Carroll, D. L. Nano Lett. 2002, 2, 1003. (15) Sun, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 481. (16) Sun, Y.; Xia, Y. Science (Washington, D.C.) 2002, 298, 2176. (17) Antipov, A. A.; Sukhorukov, G. B.; Fedutik, Y. A.; Hartmann, J.; Giersig, M.; Mo¨hwald, H. Langmuir 2002, 18, 6687. (18) Jackson, J. B.; Halas, N. J. J. Phys. Chem. B 2001, 105, 2743. (19) Oldenburg, S. J.; Jackson, J. B.; Westcott, S. L.; Halas, N. J. Appl. Phys. Lett. 1999, 75, 2897. (20) CRC Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 1999. (21) Link, S.; Wang, Z. L.; El-Sayed, M. A. J. Phys. Chem. B 1999, 103, 3529.
NL034097+
Nano Lett., Vol. 3, No. 4, 2003
