Alloying and Dealloying Processes Involved in the Preparation of

Department of Chemistry, UniVersity of Washington, Seattle, Washington ... interiors and highly crystalline walls (see, for example, Sun, Y.; Mayers, ...
0 downloads 0 Views 197KB Size
NANO LETTERS

Alloying and Dealloying Processes Involved in the Preparation of Metal Nanoshells through a Galvanic Replacement Reaction

2003 Vol. 3, No. 11 1569-1572

Yugang Sun and Younan Xia* Department of Chemistry, UniVersity of Washington, Seattle, Washington 98195-1700 Received September 11, 2003; Revised Manuscript Received September 19, 2003

ABSTRACT The galvanic replacement reaction between Ag nanoparticles and an aqueous HAuCl4 solution has recently been demonstrated as a simple and convenient route to metal nanostructures with hollow interiors and highly crystalline walls (see, for example, Sun, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 481. Sun, Y.; Xia, Y. Science, 2002, 298, 2176). However, the details of morphological, compositional, structural, and spectral changes involved in the entire process of this template-engaged reaction is yet to be elucidated. The experimental results described in this letter indicate that the templating process proceeded through two distinctive steps: (i) formation of pinhole-free nanoshells with homogeneous, uniform walls of Au/Ag alloys via a combination of the replacement reaction and alloying; and (ii) formation of porous nanoshells (nanocages) through a dealloying process, in which Ag was selectively dissolved from the walls made of Au/Ag alloys. As alloying and dealloying proceeded, the surface plasmon resonance peaks of resultant metal nanostructures could be continuously tuned from ∼425 to ∼1030 nm.

Nanostructures made of metals have received much attention in the past several decades because of their intriguing applications in optics, electronics, optoelectronics, catalysis, sensing, clinical diagnostics, surface-enhanced Raman scattering (SERS), information storage, and energy conversion/ storage.1-5 In particular, the performance of these nanostructures could be further improved by processing them as hollow entities. For instance, hollow nanostructures are intrinsically lower in density and higher in surface area relative to their solid counterparts. As a result, the quantity of coinage metals (such as Au, Ag, and Cu) could be significantly reduced in making conductive composites by using metallic fillers in the form of hollow nanostructures.6 In another example, nanoshells made of gold have been demonstrated by Halas et al. as a new approach to greatly expand the spectral range of surface plasmon resonance (SPR) features associated with Au nanostructures.7 Hollow metal nanostructures are often prepared by templating against existing entities, such as silica beads or polymer latexes. In a typical procedure, a thin shell of the desired metal (or a precursor to this metal) is deposited on the surface of a template using various methods, and the template can be subsequently removed through wet chemical etching (or calcination) to generate a hollow metal structure.8 Although this approach can be experimentally simple and straightforward, the resultant nanoshells are often characterized by problems such as incomplete coverage, rough surface, * Corresponding author. E-mail: [email protected]. 10.1021/nl034765r CCC: $25.00 Published on Web 10/08/2003

© 2003 American Chemical Society

poor crystallinity, nonuniformity in shell thickness, poorly controlled composition, and structural fragility. We have recently demonstrated a one-step approach based on the galvanic replacement reaction between Ag templates and a metal salt solution to the large-scale synthesis of hollow nanostructures made of various noble metals (e.g., Au, Pt, and Pd).9 However, the mechanisms responsible for the morphological, compositional, and spectral changes involved in this template-engaged process are still elusive. For example, there is still no direct evidence to demonstrate how the Ag templates are completely removed to form hollow nanostructures characterized by pinhole-free walls. The chemical compositions and stabilities of the hollow nanostructures also remain to be evaluated systematically. The major goal of this letter is to address these issues by analyzing the structures corresponding to different stages of the galvanic replacement reaction with scanning electron microscopy (SEM), transmission electron microscopy (TEM), atomic emission spectroscopy, and UV-visible-near-infrared (NIR) absorption spectroscopy.10 Figure 1 shows the SEM images of Ag templates before and after they had reacted with different volumes of 1 mM HAuCl4 solution. Figure 1A gives an SEM image of the templates-Ag nanoparticles that existed in various shapes with a mean diameter of ∼75 nm. After they had reacted with 0.25 mL of the HAuCl4 solution, shells containing pinholes (as indicated by arrows in Figure 1B) were formed

Figure 1. SEM images obtained from silver nanoparticles (A) before and (B-F) after they had reacted with different volumes of 1 mM HAuCl4 solution: (B) 0.25, (C) 0.60, (D) 1.00, (E) 1.20, and (F) 1.50 mL. The insets in (B) and (C) are the corresponding TEM images, indicating the formation of hollow interiors for the nanostructures.

on their surfaces. The appearance of pinholes at specific spots on each template indicated that the replacement reaction was initiated locally rather than over the entire surface. The Au atoms resulting from the galvanic replacement reaction tended to be epitaxially deposited on the surfaces of Ag templates because of their close match in crystal structure (both silver and gold have a face-centered cubic lattice) and lattice parameters (4.0786 Å and 4.0862 Å for gold and silver, respectively). The deposition would lead to the formation of an incomplete Au shell on each individual Ag template, and this shell could prevent the underneath Ag surface from reacting with HAuCl4. As a result, only the pinholes in the newly formed Au shells could serve as active sites for further reaction. The inset of Figure 1B shows a TEM image of the same sample, suggesting that the Ag template had been partially dissolved to generate a void inside each particle. This image also confirms the existence of openings (pinholes) on the surfaces of resultant particles, which allowed all the species involved in the reaction to continuously diffuse in and out of the holes until the Ag template had been completely digested. Note that continuous etching of silver cores would then lead to the formation of hollow nanostructures (in this case, nanoshells) with uniform walls. Figure 1C shows an SEM image of such nanoshells that were prepared by reacting the same amount of Ag nanoparticles with 0.60 mL of HAuCl4 solution. The TEM image (shown as inset) clearly indicates the increase of void size and the uniformity in wall thickness for each shell. At 1570

this stage, essentially all pinholes had disappeared and each shell was characterized by a smooth, hole-free surface. The elimination of pinholes could be attributed to mass-transport processes such as Ostwald ripening.11 Along with the dissolution of Ag templates and the formation of Au shells, alloying also occurred between these two elements because the homogeneous Au/Ag alloy is more thermodynamically stable than either pure Au or Ag.12 It is a combination of galvanic replacement reaction (between Ag and HAuCl4) and alloying (between Au and Ag) that led to the formation of hollow nanoshells with smooth surfaces and homogeneous, uniform, and highly crystalline walls. When the volume of HAuCl4 solution added to a dispersion of Ag templates was further increased, the AuCl4- ions started to etch the Au/Ag nanoshells by selectively removing the silver atoms. In this so-called dealloying process,13 numerous lattice vacancies would be formed because only one Au atom could be produced at the expense of three Ag atoms (as calculated from the stoichiometric relationship discussed in ref 9a). These defects would result in the formation of negative curvature into the solid walls and thus cause an increase in the total surface energy.14 To avoid this energy penalty, the vacancies would coalesce to form pinholes via an Ostwald ripening process to reduce the total surface energy. Figure 1D shows the SEM image of a typical sample where the volume of HAuCl4 solution was increased to 1.00 mL. This image clearly shows that the surface of each nanoshell was decorated with pinholes of 5-15 nm in lateral dimensions. Pinholes of smaller dimensions could still be clearly resolved using SEM when the volume of HAuCl4 solution was between 0.60 and 1.00 mL. On the other hand, the pinholes could be enlarged to form highly porous nanocages as dealloying further proceeded. As shown in Figure 1E, the nanocages with pore sizes larger than 15 nm were formed when 1.20 mL HAuCl4 solution was added to the dispersion of Ag nanoparticles. The broken cages (as indicated by arrows) also imply that each pore penetrated both interior and exterior surfaces of the wall. It is also interesting to note that complete dealloying would force the nanocages to collapse and form small fragments of gold (a typical example is shown in Figure 1F). In general, porous nanoshells with tunable porosities could be conveniently prepared by adding different volumes of HAuCl4 solution to control the dealloying process. We have also studied the compositions of complete nanoshells (purified by removing all AgCl solid) using energy-dispersive X-ray (EDX) spectroscopy. Figure 2A shows a spectrum that was obtained from a sample where 0.6 mL of HAuCl4 solution was used. Both Au and Ag were observed with strong intensities, while no signal for Cl was detected. Combined with the single crystalline structure revealed by high-resolution TEM studies (see, for example, Figure 2 of ref 9b), it is conclusive that these nanoshells were composed of a homogeneous Au/Ag alloy rather than a heterogeneous (or mosaic) microstructure. The exact concentrations of Au and Ag in the hollow nanostructures were analyzed using an atomic emission spectrometer equipped with an inductively coupled plasma (ICP) system. Nano Lett., Vol. 3, No. 11, 2003

Figure 2. (A) EDX spectrum taken from a random assembly of pinhole-free Au/Ag nanoshells (as shown in Figure 1C). The signal of Si originated from the silicon substrate, on which the nanoshells were deposited. (B) Relationship between the molar fraction of Au in the product (as determined using atomic emission spectroscopy) and the volume of 1 mM HAuCl4 solution that was added to react with the Ag nanoparticles.

The hollow nanostructures suspended in water could be directly atomized because the flame of ICP could reach temperatures as high as 9000 K.15 Figure 2B shows the molar fractions of Au in products that were prepared by reacting the same amount of Ag nanoparticles with different volumes of 1 mM HAuCl4 solution. The plot indicates that the complete nanoshells shown in Figure 1C were actually made of a Au/Ag alloy with the molar fraction of Au being ∼0.32. This result implies that pinhole-free nanoshells could be formed when only ∼60% of the Ag atoms participated in the replacement reaction. The molar fraction of Au could reach 0.96 when the volume of HAuCl4 solution was increased to 1.5 mL, indicating that the small fragments shown in Figure 1F (formed through the dealloying process) were composed of essentially pure gold. Because nanostructures made of gold and/or silver are well-known to exhibit distinctive SPR features that are strongly dependent on the composition, shape, and structure (solid versus hollow), the alloying and dealloying processes could also be conveniently followed using the UV-visibleNIR spectroscopic method. Figure 3 shows the extinction spectra taken from a set of samples (dispersed in water, with the AgCl solid being removed) that were prepared by adding different volumes of 1 mM HAuCl4 solution to the same amount of Ag nanoparticles. More specifically, Figure 3A shows the spectral changes involved in the formation of Au/ Ag alloyed nanoshells (similar to Figure 2 of ref 9a). Note Nano Lett., Vol. 3, No. 11, 2003

Figure 3. UV-visible-NIR extinction spectra of an aqueous dispersion of silver nanoparticles before and after reacting with different volumes (as indicated on each panel) of 1 mM HAuCl4 aqueous solution. The entire process can be divided into three different stages: (A) the formation of complete nanoshells through galvanic replacement reaction between Ag templates and HAuCl4 solution and alloying between Au and Ag; (B) the formation of porous nanocages through dealloying and selective dissolution of Ag; and (C) the fragmentation of gold nanocages into nanoparticles. The spectra in (B) have been normalized against the peak intensities.

that the extinction peak corresponding to Ag nanoparticles (at 425 nm) disappeared when 0.60 mL of HAuCl4 solution was introduced into the reaction system. This change indicated the complete consumption of templates made of pure Ag and the formation of nanoshells made of a homogeneous Au/Ag alloy. Accompanying the formation of Au/Ag nanoshells, an extinction peak with increasing intensity appeared at longer wavelengths, whose position was continuously red-shifted toward 750 nm. This peak shift 1571

might be caused by a slight increase in void size for the alloyed nanoshells.16 Figures 3B and 3C show the spectral signatures associated with the dealloying of Au/Ag nanoshells. As the volume of HAuCl4 solution was increased from 0.60 to 1.20 mL, the extinction peak was further red-shifted from 750 to 1030 nm. Such a shift in SPR peak from visible to NIR regions was mainly attributed to the formation of pinholes with increasingly larger dimensions in the walls. Furthermore, thinning of shell walls might also be involved in the dealloying process, which could cause additional red-shift for the SPR peak.16 Since irradiations in the NIR region (e.g., 800-1030 nm) can penetrate soft tissues without significant attenuation, the Au/Ag alloyed nanoshells and Au nanocages described here are expected to find use as photothermal converters in biomedical applications such as drug release and cancer therapy.17 After 1.50 mL of HAuCl4 solution was added, the peak in the NIR region disappeared, indicating that the nanoshells had collapsed to form small fragments of solid gold nanoparticles, whose SPR peak has been documented to be located at ∼520 nm (the blue curve in Figure 3C). All these spectroscopic observations are consistent with the results obtained by SEM and TEM studies. In summary, the metal nanostructures that were formed at the two distinctive steps of the replacement reaction between Ag nanoparticles and aqueous HAuCl4 solutions were intensively studied using SEM, TEM, EDX, atomic emission spectroscopy, and absorption spectroscopy. In the first step, seamless nanoshells with Au/Ag walls were resulted from a combination of replacement reaction (between Ag and HAuCl4) and alloying (between Au and Ag). In the second step, porous nanocages were formed due to the destruction of dealloying process, in which HAuCl4 selectively removed silver atoms from the Au/Ag walls. Complete dealloying led the nanocages to collapse into gold fragments. Acknowledgment. This work has been supported in part by a DARPA-DURINT subcontract from Harvard University, a Career Award from NSF (DMR 9983893) and a research fellowship from the David and Lucile Packard Foundation. Y.X. is a Camille Dreyfus Teacher Scholar and an Alfred P. Sloan Research Fellow. We thank Prof. D. Gamelin for allowing us to use the UV-vis-NIR spectrometer in his research group. References (1) Reviews: (a) Halperin, W. P. ReV. Mod. Phys. 1986, 58, 533. (b) Templeton, A. C.; Wuelfing, W. P.; Murray, R. W. Acc. Chem. Res. 2000, 33, 27. (c) Lam, D. M.-K.; Rossiter, B. W. Sci. Am. 1991, 265(5), 80. (d) Lewis, L. N. Chem. ReV. 1993, 93, 2693. (e) Murray, C. B.; Sun, S.; Doyle, H.; Betley, T.; Mater. Res. Soc. Bull. 2001, 26, 985. (2) Optics and catalysis: (a) Novak, J. P.; Brousseau, L. C.; Vance, F. W.; Johnson, R. C.; Lemon, B. I.; Hupp, J. T.; Feldheim, D. L. J. Am. Chem. Soc. 2000, 122, 12029. (b) Murphy, C. J.; Jana, N. R. AdV. Mater. 2002, 14, 80. (c) Kim, F.; Song, J. H.; Yang, P. J. Am. Chem. Soc. 2002, 124, 14316. (d) Nicewarner-Pen˜a, S. R.; Freeman, R. G.; Reiss, B. D.; He, L.; Pen˜a, D. J.; Walton, I. D.; Cromer, R.; Keating, C. D.; Natan, M. J. Science 2001, 294, 137. (e) Ahmadi, T. S.; Wang, Z. L.; Green, T. C.; Henglein, A.; El-Sayed, M. A. Science 1996, 272, 1924. (e) Teng, X.; Black, D.; Watkins, N. J.; Gao, Y.; Yang, H. Nano Lett. 2003, 3, 261. 1572

(3) Electronics and optoelectronics: (a) Chen, S.; Yang, Y. J. Am. Chem. Soc. 2002, 124, 5280. (b) El-Sayed, M. A. Acc. Chem. Res. 2001, 34, 257. (c) Maier, S. A.; Brongersma, M. L.; Kik, P. G.; Meltzer, S.; Requicha, A. A. G.; Atwater, H. A. AdV. Mater. 2001, 13, 1501. (d) Kamat, P. V. J. Phys. Chem. B 2002, 106, 7729. (4) Sensing and clinical diagnostics: (a) Tkachenko, A. G.; Xie, H.; Coleman, D.; Glomm, W.; Ryan, J.; Anderson, M. F.; Franzen, S.; Feldheim, D. L. J. Am. Chem. Soc. 2003, 125, 4700. (b) Taton, T. A.; Mirkin, C. A.; Letsinger, R. L. Science 2000, 289, 1757. (c) Thanh, N. T. K.; Rosenzweig, Z. Anal. Chem. 2002, 74, 1624. (d) Kim, Y.; Johnson, R. C.; Hupp, J. T. Nano Lett. 2001, 1, 165. (e) Nath, N.; Chilkoti, A. Anal. Chem. 2002, 74, 504. (f) Roll, D.; Malicka, J.; Gryczynski, I.; Cryczynski, Z.; Lakowicz, J. R. Anal. Chem. 2003, 75, 3440. (5) SERS: (a) Tessier, P. M.; Velev, O. D.; Kalambur, A. T.; Rabolt, J. F.; Lenhoff, A. M.; Kaler, E. W. J. Am. Chem. Soc. 2000, 122, 9554. (b) Nie, S.; Emory, S. R. Science 1997, 275, 1102. (c) Dick, L. A.; McFarland, A. D.; Haynes, C. L.; Van Duyne, R. P. J. Phys. Chem. B 2002, 106, 853. (d) Jackson, J. B.; Westcott, S. L.; Hirsch, L. R.; West, J. L.; Halas, N. J. Appl. Phys. Lett. 2003, 82, 257. (6) Ohmori, M.; Matijevic, E. J. Coll. Interface Sci. 1992, 150, 594. (7) Oldenburg, S. J.; Averitt, R. D.; Westcott, S. L.; Halas, N. J. Chem. Phys. Lett. 1998, 288, 243. (8) (a) Kim, S.-W.; Kim, M.; Lee, W. Y.; Hyeon, T. J. Am. Chem. Soc. 2002, 124, 7642. (b) Caruso, F.; Spasova, M.; Salgueirin˜o-Maceira, V.; Liz-Marza´n, L. M. AdV. Mater. 2001, 13, 1090. (9) (a) Sun, Y.; Mayers, B. T.; Xia, Y. Nano Lett. 2002, 2, 481. (b) Sun, Y.; Xia, Y. Anal. Chem. 2002, 74, 5297. (c) Sun, Y.; Xia, Y. Science 2002, 278, 2176. (d) Sun, Y.; Mayers, B.; Xia, Y. AdV. Mater. 2003, 15, 641. (10) The Ag nanoparticles were synthesized using the polyol process. In a typical synthesis, 0.025 g AgNO3 (99+%, Aldrich) and 0.10 g poly(vinyl pyrrolidone) (PVP, Mw ≈ 55 000, Aldrich) were dissolved in 10 mL anhydrous ethylene glycol (99.8%, Aldrich). The mixture was heated at 160 °C (in oil bath) for 1.5 h while it was vigorous stirred. In a typical replacement reaction, 250 µL of the as-obtained dispersion of Ag nanoparticles was added to 5 mL of deionized water (purified with cartridges from Millipore, E-pure, Dubuque, IA). This aqueous suspension was refluxed for 15 min, and then a specific volume of 1 mM HAuCl4 (99.9%, Aldrich) aqueous solution was added dropwise. The reaction mixture was continuously refluxed for another 20 min until the color became stable. Vigorous magnetic stirring was maintained throughout the synthesis. As the solution was cooled to room temperature, white solid (AgCl precipitate) settled to the bottom of the reaction container. The AgCl solid could be removed by dissolving it with a saturated solution of NaCl (99.9%, Fisher). In this case, NaCl powders were added to the aqueous dispersion of product until the solution was saturated with NaCl. The solution was then transferred into a centrifugation tube and centrifuged at 10 000 rpm for 15 min. The supernatant containing the dissolved AgCl could be easily removed using a pipet. The solid was rinsed with water and centrifuged for six times. The final product could be dispersed in water for further characterization. The SEM images were obtained using a FEI field-emission microscope (Siron XL) operated at 15 kV. EDX measurements were carried out with the same microscope. The TEM images were taken using a JEOL microscope (1200EX II) operated at 80 kV. The concentrations of gold and silver in the nanostructures were analyzed using an atomic emission spectrophotometer (Thermo Jarrell Ash Corporation, Franklin, MA) equipped with a Jarrell Ash 955 inductively coupled plasma system. All the UV-vis-NIR extinction spectra were recorded at room temperature on a Cary 5E (Varian) spectrophotometer using the quartz cuvette with a 1-cm optical path. (11) Roosen, A. R.; Carter, W. C. Physica A 1998, 261, 232. (12) Shi, H.; Zhang, L.; Cai, W. J. Appl. Phys. 2000, 87, 1572. (13) Erlebacher, J.; Aziz, M. J.; Karma, A.; Dimitrov, N.; Sieradzki, K. Nature 2001, 410, 450. (14) Sieradzki, K. J. Electrochem. Soc. 1993, 140, 2868. (15) InductiVely Coupled Plasma Emission Spectroscopy, Boumans, P. W. J. M., Ed.; John Wiley & Sons: New York, 1987; Vol. 90, part 1, pp 69-99. (16) Westcott, S. L.; Oldenburg, S. J.; Lee, T. R.; Halas, N. J. Chem. Phys. Lett. 1999, 300, 651. (17) (a) Sershen, S. R.; Westcott, S. L.; Halas, N. J.; West, J. L. J. Biomed. Mater. Res. 2000, 51, 293. (b) West, J. L.; Halas, N. J. Curr. Opin. Biotech. 2000, 11, 215.

NL034765R Nano Lett., Vol. 3, No. 11, 2003