Growth of Silver Nanowires from Solutions: A ... - ACS Publications

Gilles Berhault, Marta Bausach, Laure Bisson, Loïc Becerra, Cécile .... Ivan Moreno , Nuria Navascues , Manuel Arruebo , Silvia Irusta , Jesus Santa...
0 downloads 0 Views 543KB Size
Download PDF
9416

J. Phys. Chem. B 2005, 109, 9416-9421

Growth of Silver Nanowires from Solutions: A Cyclic Penta-twinned-Crystal Growth Mechanism Shu-Hong Zhang, Zhi-Yuan Jiang, Zhao-Xiong Xie,* Xin Xu, Rong-Bin Huang, and Lan-Sun Zheng State Key Laboratory of Physical Chemistry of Solid Surfaces and Department of Chemistry, Xiamen UniVersity, Xiamen 361005, China ReceiVed: December 27, 2004; In Final Form: March 11, 2005

Silver nanowires are synthesized by simple reduction of the silver ions with reductants such as glucose, sodium citrate, and sodium hypophosphite, etc., in the absence of the so-called surfactants or capping reagents at the temperature from 80 to 200 °C. Regardless of the reductants, the nanowires prepared at a given temperature are uniform in diameters, ranging from 30 to 50 nm at 100 °C. Nanoparticles coexist with nanowires in the products with larger diameters (usually larger than 50 nm). We find that all the silver nanowires in the as-prepared products are of cyclic penta-twinned structure, where five crystallites bond by the {111} facets. We propose that the intrinsic factor of the cyclic penta-twinned structure, i.e., the angular mismatch of the five crystallites in forming a gapless rod, controls the size of the nanowires and guides the directional growth of the nanowires with {110} as the active facets. The nanoparticles in the products are aggregates of imperfect penta-twinned crystals, which inhibits them from growing into nanowires and results in larger size. From the structural information of the nanoparticles synthesized at room temperature, we propose that the formation of the cyclic penta-twinned structure is due to the stacking fault and the intrinsic equilibrium structures of the lower energy.

1. Introduction One-dimensional (1-D) nanostructures of noble metals are of considerable interest due to their different or superior electrical, optical, and chemical properties as compared to the bulk counterparts.1-3 Among all the metals, silver is especially attractive as it possesses the highest electrical and thermal conductivities, promising some potential applications in many aspects such as catalysts and optical and electronic nanodevices.8,9 Great effort has been devoted to the controlled growth of various kinds of 1-D nanostructures of noble metals in recent years.4-7 Up to now, a variety of techniques, including hardtemplate, soft-template, and template-free methods, have been developed.10-20 Among these methods, the solution-phase synthesis should be the most promising hard-template-free route to the preparation of silver nanostructures in terms of cost, yield, and simplicity. The growth of nanowires in the hard templates, such as nanopores of porous alumina,10,11 can usually be attributed to the geometrical confinement of the pore structures. The growth mechanism for the 1-D nanostructures without the topographical confinement is still elusive. It is generally accepted that organic additives, such as surfactants or capping reagents, play an important role in directing the 1-D growth.18,20-22 Recently, Xia et al. demonstrated an approach for the largescale synthesis of silver nanowires with uniform diameters based on the polyol process from the solutions. The nanowires thus produced were determined to have a penta-twinned structure with the side surfaces being bounded by the {100} facets and the end surfaces being bounded by the {111} facets. It was suggested that silver nanowires started from the multiply * Corresponding author. Telephone: +86-592-2185667. Fax: +86-5922183047. E-mail: [email protected].

twinned decahedral nanoparticles and that it was the passivation of the more active {100} surfaces by adsorption of poly(vinyl pyrrolidone) (PVP) that led to the anisotropic 1-D growth of silver nanowires by the {111} facets.22 The growth of gold nanorods was claimed to follow the same mechanism.23 In this paper, we show that silver nanowires can be synthesized without the assistance of surfactants or capping reagents. We propose a cyclic penta-twinned-crystal growth mechanism with bare {110} facets as the active surfaces, in which the surface energy (including the high strain energy) in the penta-twinned crystals plays a key role in confinement of the diameters of the nanowires and in directing the 1-D growth of the nanowires. 2. Experimental Section In a typical procedure for the preparation of silver nanocrystals, equal volumes of two aqueous solutions A and B freshly prepared in glassware were mixed with each other thoroughly. Solution A was a 0.1 mM silver-ammonia solution, while B was a 0.1 mM reductant solution containing either glucose or sodium citrate or sodium hypophosphite (NaH2PO2). The final pH of the mixture was 8-9. The mixture was kept steady at certain temperature. The reaction lasted for several minutes to 1 h. The reacting solutions initially became bright yellow, indicating the appearance of Ag nanoparticles, and then turned to suspension with yellowish gray precipitates. Finally, the precipitates were isolated by centrifugal separation, which were then rinsed several times with water, and suspended in alcohol for later use. The morphology and the crystal structure of the products were observed by using transmission electron microscopy (TEM) and high-resolution TEM (HRTEM). The TEM and HRTEM images

10.1021/jp0441036 CCC: $30.25 © 2005 American Chemical Society Published on Web 04/20/2005

Growth of Silver Nanowires from Solutions

J. Phys. Chem. B, Vol. 109, No. 19, 2005 9417

Figure 1. Typical TEM images of silver nanostructures synthesized at 100 °C by simple reduction of the silver ions with glucose (a), sodium citrate (b), and NaH2PO2 (c), respectively; and typical SAED patterns (d, e) of silver nanowires. Most of the diffraction spots in d can be indexed as the diffractions of the [1h12] zone axis (italics) and the [001] zone axis (upright), while those in e, as those of the [11h0] (italics) and the [11h1h] zone axis (upright), suggesting that the nanowires should possess a multiply twinned structure.

were recorded on a JEM-100CXII microscope operating at 100 kV and a Tecnai F30 microscope at 300 kV, respectively. The selective area electron diffraction (SAED) patterns were taken on both JEM-100CXII and Tecnai F30 microscopes. 3. Results and Discussion Parts a-c of Figure 1 show the typical TEM images for the products prepared by the simple reduction of the silver ions with glucose, sodium citrate, and NaH2PO2 at 100 °C, respectively. It is clearly seen that all the products are the mixture of nanowires and nanoparticles. The only difference among Figure 1a-c is that the products contain much more nanowires when using glucose as reductant than they do when using sodium citrate or NaH2PO2 as reductant. Nanowires make up 70% of the total crystal population for glucose, while the nanowire population is around 35% for sodium citrate and 40% for NaH2PO2. We found that three reductants led to nanowires of similar radial sizes. From the TEM images of the products synthesized at 100 °C, it can be seen that the size distribution of the nanowires is very narrow, ranging from 30 to 50 nm. The average diameter of the nanowires is about 40 nm. We also find that the size of the nanoparticles in the products is usually larger than the diameter of the nanowires (usually larger than 50 nm). Parts d and e of Figure 1 show two typical SAED patterns observed for the silver nanowires. By tilting the sample along the longitudinal axis, Figure 1d,e can be interchanged. The diffraction patterns indicate that silver nanowires are crystallized in the face-centered cubic (fcc) structure, but neither of them could be assigned to a single-crystal structure. As shown in Figure 1d,e, every diffraction pattern contains two sets of diffractions from the fcc silver. Most of the diffraction spots in Figure 1d can be indexed as the diffractions of the [1h12] zone axis and the [001] zone axis, while those in Figure 1e, as those of the [11h0] and the [11h1h] zone axis. This fact suggests that the

nanowires should possess a multiply twinned structure. The longitudinal direction of the nanowires is [110], which is the common direction for the twinned crystals. A few additional weak diffraction spots (as marked by asterisks in Figure 1d) that are not attributable to the basic diffraction points of silver can be indexed by the double diffractions.23 These observations are consistent with the results obtained for cyclic penta-twinned nanorods of gold,23 copper,24 and silver.25 As a consequence, a cyclic penta-twinned fcc structure model (Figure 2a) with five {111} twinned boundaries and five {100} side surfaces is suggested for silver nanorods/nanowires synthesized in our experiments. Parts b and c of Figure 2 show the pentagonal cross-section viewed along the 5-fold symmetry axis. We have, in an ideal situation, five similar subunits labeled as T1, T2, T3, T4, and T5. Every subunit is twinned with its two neighbors by two (111) lattice planes. The side surfaces are the {100} facets, and [110] is the common axial direction. When the electron beam is directed perpendicularly to one side surface of the nanowire (Figure 2b), T1, T3, and T4 will generate diffraction spots, which belong to the [001] (for T1) and the [1h12] (for T3 and T4) crystallographic zones, respectively. When rotating the nanowire by 18° along the 5-fold axis, we change the pentagon from the position of Figure 2b to that of Figure 2c such that the electron beam runs parallel with one side surface of the nanowire. In Figure 2c, the T5 subunit will cause the diffraction of the [11h0] zone axis, and T2 and T3 will give the diffraction of the [11h1h] zone axis. Thus the diffraction patterns of models in Figure 2b,c are in good agreement with the corresponding experimental data shown in Figure 1d,e, respectively. These observations compare well with the results obtained for the cyclic penta-twinned nanorods of gold.23 Wet chemical syntheses have been successfully applied to the preparation of silver nanowires, where capping reagents or surfactants have been employed to direct the 1-D growth. The advocated mechanism is that each silver nanowire evolved from

9418 J. Phys. Chem. B, Vol. 109, No. 19, 2005

Zhang et al.

Figure 2. Schematic model of the silver nanowires. (a) Stereoview of cyclic penta-twinned nanostructures with five {111} twinned boundaries and five {100} side surfaces. Here one subunit is explicitly shown. (b-d) the cross-section view along the longitudinal axis (〈110〉, the 5-fold axis). Panel b corresponds to the electron diffraction pattern where the direction of the electron beam is perpendicular to one side surface, while c corresponds to the electron diffraction pattern where the direction of the electron beam is parallel to one side surface. Panel d shows the angular mismatch for five perfect fcc crystallites to form a planar cyclic penta-twinned structure.

a multiply twinned decahedral nanoparticle of silver, that the anisotropic growth was maintained by selectively covering the {100} facets with PVP while leaving the {111} facets largely uncovered by PVP and thus highly reactive, and hence that the reduced silver atoms were deposited on the top {111} surfaces, which then form an elongated decahedron, resulting in the 1-D growth22 (such an elongated decahedron is shown by the dotted lines on the front end of the rod in Figure 2a). As {100} surfaces are more active than {111} surfaces, it is clear, based on this mechanism, that passivation of {100} by surfactants such as PVP is vital for the 1-D growth. This mechanism is conceivable for the 1-D growth of metal nanowires in the presence of surfactants or capping reagents; it, at the same time, poses a question, why 1-D growth can be observed in the absence of surfactants, as shown in the present study as well as other studies.17,26 Especially with this mechanism, it is difficult to explain why Ag nanowires with elongated decahedron shape could be grown from a vapor phase by evaporating silver powders without any carrier gas.26 While the previous mechanism emphasized the role of the twin boundaries, we emphasize the intrinsic factor of the cyclic penta-twinned crystals and believe that bare {110} are the active facets for growth. The five twinned crystals have the common 〈110〉 5-fold axis (Figure 2a), and they are connected with the {111} facets by the stacking fault. It can be calculated that the interfacial angle between two {111} facets of the subunits along the longitudinal axis on the top plane of the nanowire is 360°/5 ) 72° in the cyclic penta-twinned crystals. However, for a perfect fcc crystallite, the interfacial angle between {111} lattice planes is 70.53°. The angle difference is then 1.47°, which indicates that there should be lattice distortion (microstrain) in the crystallites that consist of the cyclic penta-twinned crystals (see Figure 2d). Considering a nanowire with 20 nm in diameter, the angle deviation of 1.47° will lead to a gap of 0.256 nm on the side surface, which corresponds to the size of one silver atom. The bigger the diameter of the nanowire grows, the larger

Figure 3. Typical TEM image of silver nanostructures synthesized at 200 °C by simple reduction of silver ions with glucose.

the gap becomes. To form gapless nanowires, the atoms surrounding the radial direction must overcome the energy cost of microstrains. As a result, the radial size of the nanowire is controlled by such surface-free-energy minimization. In the literature, the sizes of multiply twinned decahedral nanoparticles were also claimed to be controlled in terms of surface-freeenergy minimization.27 On the basis of this mechanism, it would be anticipated that by changing reaction temperature, the free energy of the reduced silver atoms should change, such that a higher reaction temperature will lead to an increased diameter of the nanowires. Indeed, we have successfully prepared silver nanowires from temperature of 80 to 200 °C, which does show an increased diameter of the nanowires accompanied with an increase of the reaction temperature. As compared to Figure 1 for nanowires of average radii of 40 nm synthesized at 100 °C, Figure 3 shows a typical TEM image of silver nanowires synthesized at 200 °C with the average radii of 80 nm. Along this line, it is also anticipated that physical or chemical adsorptions of additives on the nanowires would modify the surface free energies to affect the size of the nanowires. There

Growth of Silver Nanowires from Solutions

Figure 4. (a) Typical TEM image which shows the existence of an on-top plane vertical to the longitudinal axis of a silver nanowire. (b) HRTEM image of the on-top plane, showing the bare {110} surfaces. (c) TEM image of a bottleneck type nanowire. (d) TEM image of a torsion (V) shape nanowire.

are many experimental results, lending support to this point of view.18,20,26 Figure 4a presents direct evidence for the existence of an on-top plane vertical to the longitudinal axis of a silver nanowire as indicated by the planar penta-twinned fcc structure model of the bare {110} facets (Figure 2). By checking carefully the nanorods and the nanowires in the products, we found that some short nanorods in the products did show the shape of an elongated decahedron with both ends of the nanorods consisting of inclined surfaces of the {111} facets. However, many nanowires and nanorods have their ends vertical to the longitudinal axes (Figure 4a). HRTEM images clearly show that these on-top planes of the nanowires correspond to the {110} facets (Figure 4b). These results indicate that silver nanowires synthesized in the present condition may not be envisioned as elongated decahedrons evolved from deposition of silver atoms on the {111} facets. Instead, the bare {110} surfaces are conceivable to be the active surfaces to grow into silver nanowires. For the fcc metals, {111} are the most stable facets, {100} are the next stable facets, and {110} are the least stable ones. The growth rates of the {110} facets should be much faster than those of the other two facets. Hence we infer that it is the free energy cost to fuse the {111} facets that define the radial size of the nanowire and that it is the layer by layer growth of the {110} facets that leads to the preferential growth along [110]. It is known that surfactants change the aspect ratios of the nanowires. This infers that formation of ending {111} surfaces and the layer by layer growth of the {110} facets are in a kinetic balance. In some cases, the stable {111} surfaces are formed at an early stage of growth on both ends of the nanorods and the growth processes may stop. Therefore, some nanorods with the morphology of the elongated decahedrons may not be the growing nanowires but “homunculus” of nanowires. In some special cases, ending and growth processes are in competition, leading to the formation of bottleneck type nanowires, as shown in Figure 4c.

J. Phys. Chem. B, Vol. 109, No. 19, 2005 9419 Figure 4d shows a torsion (“V”) shape nanowire occasionally found in the products. HRTEM images of the turning area show that two parts of the nanowire are bonded with the {111} facets. Although the joint area of nanowires becomes smaller (marked with dotted circle in Figure 4d), the nanowire grows to its normal size shortly after the joint. One may explain it as being a new cyclic penta-twinned structure formed on the terminated {111} facets that results in the V shape nanowire. Another possibility is that two neighboring penta-twinned nuclei bond together by the twinning of {111} facets and further grow into two nanowires independently. Both cases support that the intrinsic factor of the cyclic penta-twinned structure determines the diameters of the nanowires. Our experiments show that nanoparticles coexist with nanowires in the products and that the size of particles is usually larger than the radial size of the nanowires. The detailed HRTEM images show that the nanoparticles are also multiply twinned crystals. Some of them are imperfect penta-twinned decahedral particles. As shown in Figure 5a,b, five crystallites bond together with the {111} lattice planes, but one additional crystallite embeds into the particle, as marked by a dotted parallelogram. Furthermore, packing defects clearly exist in the joint part. Although embedding and defect formation provide a way to release the strain originated from the wedge Volterra disclination (with an angle of about 7.5°), it also lead to the formation of larger nanoparticles instead of nanorods or nanowires. Besides such incomplete penta-twinned particles, most particles consist of several parts of pyramidal shape, each of which, as shown in Figure 5c, is similar to a penta-twinned decahedral particle. Therefore, we considered the nanoparticles as the aggregations of incomplete penta-twinned particles. As the cyclic penta-twinned structures are concluded to play a key role in the growth of nanowires, formation of such incomplete penta-twinned particles or their aggregates will block the 1-D growth. It can be expected that the population of nanowires can be increased if aggregations of the individual penta-twinned particles can be avoided in the first place. Some surfactants may be effective to separate those particles by adsorption on the surfaces of particles during their nucleations. As an example, we reported high-yield syntheses of silver nanowires in the presence of surfactants (sodium dodecylsulfonate20), which were also confirmed to have the same cyclic penta-twinned structure recently. It should be pointed out that, in the present study, the reductants themselves may also play the role of surfactants to some extent, and therefore the percentages of the nanowires in the products are not the same because of their different adsorption abilities. However, the nanowires produced by different reductants give similar radial size, as shown in Figure 1, which supports that the intrinsic factor of the cyclic pentatwinned structure determines the diameters of the nanowires. The formation of the multiply twinned particles has been explained by several mechanisms, such as errors during the growth leading to twins, intrinsic equilibrium structures of the lower energy at smaller size, layer by layer growth around the 5-fold symmetry axes, and a phase transformation to an orthorhombic or rhombic form. To date, most of the evidence directs one to the mechanism of intrinsic equilibrium.27 To explore the detailed formation mechanism of the multiply twinned crystals, the products at different temperatures were examined. At room temperature, no 1-D nanostructures can be found in the products. The products are nanoparticles (Figure 6a) with the size ranging from 40 to 60 nm. However, being different from the nanoparticles in the products of higher temperature, no penta-twinned-crystal-like particles or multiply

9420 J. Phys. Chem. B, Vol. 109, No. 19, 2005

Zhang et al.

Figure 5. Typical TEM images of nanoparticles coexisting with nanowires: (a) TEM image of an imperfect penta-twinned nanoparticle, where one additional crystallite is embedded; (b) HRTEM image with large magnification showing the additional crystallite as marked by an arrow; (c) multiply twinned nanoparticle having several pyramidal shapes as circled by dash lines.

Figure 6. (a) Typical TEM image of silver nanoparticles synthesized at room temperature with no coexistence of silver nanowires. (b) SAED images of the nanoparticles synthesized at room temperature, showing the appearance of some additional diffraction spots (marked by arrows) in comparison with the fcc crystallites. The inset is the TEM image of the corresponding nanoparticle.

pyramidal particles can be found, which prevents the formation of nanorods or nanowires. Figure 6b shows the SAED of particles synthesized at room temperature. The SAED pattern can be indexed as fcc silver as the electron beam was set perpendicular to the (111) plane of the crystallite. However, some additional diffraction spots at 1/3 of 〈422〉 diffraction were observed. Such additional diffraction can be explained as either a flat thin plate or the stacking faults. To understand the additional diffraction spots, an equivalent hexagonal cell is then introduced to express the fcc cell, in which the c axis of the hexagonal cell is perpendicular to the (111) lattice plane of the fcc cell as shown in Figure 7. The Miller indices of the fcc cell are related to those of the corresponding hexagonal cell (hex) by the following formulas.

h ) -4H/3 - 2K/3 + L/3 k ) 2H/3 - 2K/3 + L/3 l ) 2H/3 + 4K/3 + L/3 For example, we have (111)fcc ) (003)hex, (002)fcc ) (012)hex, and (24h2)fcc ) (3h30)hex. For the fcc structure, ABCABC close packing results in the same electron densities on lattice planes I, II, and III as indicated in the model. Therefore, only (-3n 3n 0)hex diffractions can be observed, and others are missing due to the systematic extinction. However, for the flat thin plate with [111]fcc as the surface normal, the electron densities on the lattice planes I, II, and III are not the same when the number of the atomic layers is not a multiple of 3. For example, for a thin plate consisting of 20 layers, the ratio of the electron densities on the lattice planes I, II, and III should be 7:7:6. Such a change of the electron densities breaks the systematic

Figure 7. Schematic model of the equivalent hexagonal cell for the fcc crystals (ABCABC... stacking), where the c axis is perpendicular to the (111)cubic lattice plane of the fcc cell.

extinction, and therefore additional diffractions of (-n n 0)hex and (-2n 2n 0)hex (i.e. 1/3 and 2/3 of {422}fcc) appear. However, the difference of electron densities on the planes I, II, and III is small, and usually only weak diffractions appear. An example of such a case was shown in ref 28, where weak additional diffractions appeared at 1/3 and 2/3 of 〈422〉 for the atomically flat silver nanoprisms. Another possibility for the appearance of such additional diffractions is due to the stacking faults, i.e., an imperfect ABCABC... packing. These kinds of stacking faults would cause the transition from the fcc crystal cell to the hexagonal crystal cell. In the extreme, the hexagonal close packing (hcp, ABAB... close packing) could be considered as an example of the stacking faults for the fcc crystals. For this kind of stacking faults, the additional diffractions can be much stronger as compared to the case of a flat plane. In our cases as shown in Figure 6b, the additional diffractions are not weak points, suggesting that there should be many stacking faults along the [111]fcc for the nanoparticles prepared at room temperature. As has been discussed above, the cyclic penta-twinned particles are bounded by the {111} planes; i.e., the twinned crystal of the mirror symmetry is due to one stacking fault on the {111} plane. Figure 6 shows the existence of many stacking faults along the [111] direction in the room-temperature products, which might be the origin of the penta-twinned particles. With the increase of the temperature, the increased thermal energy of atoms fulfils the energy requirements for the formation of stacking faults, leading to the formation of the cyclic penta-twinned structures. Although the detailed formation mechanism for the penta-twinned structures is currently unknown, we suspect that both the stacking fault (errors during the growth) and intrinsic equilibrium structures of lower energy lead to the twinned structures.

Growth of Silver Nanowires from Solutions 4. Concluding Remarks In this paper, we show that silver nanowires can be synthesized without the assistance of surfactants. TEM images demonstrate that the products are the mixture of nanowires and nanoparticles, whose typical SAED patterns can be correlated to a cyclic penta-twinned structure and an imperfect pentatwinned decahedral structure, respectively. The cyclic pentatwinned structures are concluded to play a key role in the growth of nanowires or formation of incomplete penta-twinned particles, or their aggregates will thus block the 1-D growth. HRTEM images clearly show the existence of on-top {110} planes vertical to the longitudinal axis of a silver nanowire, which suggests that the bare {110} surfaces are conceivably the active surfaces to grow into silver nanowires. Our results show that the diameters of nanowires are controlled by the intrinsic factor of the cyclic penta-twinned structures, where penalty due to the angular mismatch has to be paid in forming nanowires of the penta-twinned-crystal structure. We propose that the cyclic penta-twinned structure is formed due to a combined effect from the stacking fault and the intrinsic equilibrium structures of the lower energy. Acknowledgment. This work is supported by the National Natural Science Foundation of China (Grant Nos. 20021002, 20473069, and 20273052), the Ministry of Science and Technology of China (Grant Nos. 2001CB610506 and 2002CCA01600), NCET from the Ministry of Education of China, and the Fok Ying-Tung Educational Foundation. References and Notes (1) Favier, F.; Walter, E. C.; Zach, M. P.; Benter, T.; Penner, R. M. Science 2001, 293, 2227. (2) Cui, Y.; Wei, Q. Q.; Park, H. Q.; Lieber, C. M. Science 2001, 293, 1289. (3) Chung, S. W.; Yu, J. Y.; Heath, J. R. Appl. Phys. Lett. 2000, 76, 2068. (4) Yu, Y. Y.; Chang, S. S.; Lee, C. L.; Wang, C. R. C. J. Phys. Chem. B 1997, 101, 6661.

J. Phys. Chem. B, Vol. 109, No. 19, 2005 9421 (5) Jana, N. R.; Gearheart, L. A.; Obare, S. O.; Johnson, C. J.; Edler, K. J.; Mann, S.; Murphy, C. J. J. Mater. Chem. 2002, 12, 2909. (6) Han, Y. J.; Kim, J. M.; Stucky, G. D. Chem. Mater. 2000, 12, 2068. (7) Forrer, P.; Schlottig, F.; Siegenthaler, H.; Textor, M. J. Appl. Electron. 2000, 30, 533. (8) Jin, R. C.; Cao, Y. W.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.; Zheng, J. G. Science 2001, 294, 1901. (9) Gould, I. R.; Lenhard, J. R.; Muenter, A. A.; Godleski, S. A.; Farid, S. J. Am. Chem. Soc. 2000, 122, 11934. (10) Zhang, S. H.; Xie, Z. X.; Jiang, Z. Y.; Xu, X.; Xiang, J.; Huang, R. B.; Zheng, L. S. Chem. Commun. 2004, 1106. (11) Sauer, G.; Brehm, G.; Schneider, S.; Nielsch, K.; Wehrspohn, R. B.; Choi, J.; Hofmeister, H.; Gosele, U. J. Appl. Phys. 2002, 91, 3243. (12) Hong, B. H.; Bae, S. C.; Youn, S. J.; Kim, K. S. Science 2002, 296, 611. (13) Hong, B. H.; Bae, S. C.; Lee, C. W.; Jeong, S.; Kim, K. S. Science 2001, 294, 348. (14) Sloan, J.; Wright, D. M.; Woo, H. G.; Bailey, S.; Brown, G.; York, A. P. E.; Coleman, K. S.; Hutchison, J. L.; Green, M. L. H. Chem. Commun. 1999, 699. (15) Sun, Y. G.; Gates, B.; Mayers, B.; Xia, Y. N. Nano Lett. 2002, 2, 165. (16) Jana, N. R.; Gearheart, L.; Murphy, C. J. Chem. Commun. 2001, 617. (17) Caswell, K. K.; Bender, C. M.; Murphy, C. J. Nano Lett. 2003, 3, 667. (18) Sun, Y. G.; Xia, Y. N. AdV. Mater. 2002, 14, 833. (19) Tian, M. L.; Wang, J. U.; Kurtz, J.; Mallouk, T. E.; Chan, M. H. W. Nano Lett. 2003, 3, 919. (20) Hu, J. Q.; Chen, Q.; Xie, Z. X.; Han, G. B.; Wang, R. H.; Ren, B.; Zhang, Y.; Yang, Z. L.; Tian, Z. Q. AdV. Funct. Mater. 2004, 14, 183. (21) Bonet, F.; Tekaia-Elhsissen, K.; Sarathy, K. V. Bull. Mater. Sci. 2000, 23, 165. (22) Sun, Y. G.; Mayers, B.; Herricks, T.; Xia, Y. N. Nano Lett. 2003, 3, 955. (23) Johnson, C. J.; Dujardin, E.; Davis, S. A.; Murphy, C. J.; Mann, S. J. Mater. Chem. 2002, 12, 1765. (24) Lisiecki, I.; Filankembo, A.; Sack-Kongehl, H.; Weiss, K.; Pileni, M. P.; Urban, J. Phys. ReV. B 2000, 61, 4968. (25) Chen, H. Y.; Gao, Y.; Yu, H. C.; Zhang, H. R.; Liu, L. B.; Shi, Y. G.; Tian, H. F.; Xie, S. S.; Li, J. Q. Micron 2004, 35, 469. (26) Bo¨gels, G.; Meekes, H.; Bennema, P.; Bollen, D. J. Phys. Chem. B 1999, 103, 7577. (27) Marks, L. D. Rep. Prog. Phys. 1994, 57, 603. (28) Jin, R.; Cao, Y. W.; Mirkin, C. A.; Kelly, K. L.; Schatz, G. C.; Zheng, J. G. Science, 2001, 294, 1901.