Growth of pentatwinned gold nanorods into truncated decahedra

Abstract

The growth mechanism from pentatwinned (PTW) gold nanorods into truncated quasi-decahedral particles when a gold salt (HAuCl₄) is reduced by N,N-dimethylformamide (DMF) in the presence of poly(vinylpyrrolidone), was elucidated through a combination of different techniques, including transmission and scanning electron microscopy, high resolution TEM and selected area electron diffraction. Particles with intermediate shapes between the original pentatwinned Au nanorods, used as seeds, and the final quasi-decahedral particles were obtained by simply tuning the [HAuCl₄] to [seeds] ratio. From the thorough structural analysis of all the intermediate morphologies obtained, it was concluded that gradual morphology changes are related to the preferential growth of higher energy crystallographic facets. As a result of the particle growth and concomitant decreased anisotropy, a progressive blue-shift of the surface plasmon resonance bands of the nanoparticles was registered by vis-NIR extinction spectroscopy.

---

Introduction

The synthesis of metal nanoparticles with accurate morphology (shape and size) control is a key stage in realizing the fabrication of functional materials at the nanoscale. This is due to the strong correlation between the morphology and the chemical, physical, optical, electronic and catalytic properties of the nanoparticles.^1–3 In particular, gold and silver are widely studied metals due to potential multiple applications arising from their unique optical properties, in fields such as surface-enhanced Raman scattering (SERS), plasmonics and biosensing.^4–7 The important role of shape and size on the properties of metal nanoparticles makes it essential to understand the processes and mechanisms involved in their growth so as to design synthetic strategies that allow shape tuning. A number of synthesis methods have been developed to prepare gold nanoparticles with different morphologies, including spheres,^8,9 rods,¹⁰ cubes,¹¹ octahedrons,^12–14 decahedrons,^15,16etc. Many of such methods are based on the growth of preformed seeds to avoid uncontrolled nucleation and thereby direct particle growth, the seeds often also act as catalysis to favour reduction of the metal salt on their surface.^9,8,10,14,15 Therefore, understanding the growth mechanism over a particular type of seed and subsequent shape-guiding process is crucial to fabricate particles with a specific shape and crystalline structure. Anisotropic particle growth is directly related to the differences in growth rate of different crystallographic facets while in general, metals tend to grow into thermodynamically stable particles.¹⁷ Interestingly, it has been reported that the presence of certain capping agents or additives can alter the relative surface energies of the facets (in a face-centred cubic gold lattice: γ(110) > γ(100) > γ(111))^17,18 and, thereby, modify the particle shape and even induce the formation of thermodynamically non-preferred nanoparticle shapes.^19–23 For example, Seo et al.²¹ observed that the overgrowth of preformed cubooctahedral particles produced either octahedra or cubes depending on whether the process took place in the absence or in the presence of silver ions. The authors proposed that silver ions were preferentially deposited onto the {100} facets, forming a silver layer and at the same time inhibiting the epitaxial growth of additional gold atoms. Niu et al.²³ reported that cetylpyridinium chloride (CPC) used as capping agent during the growth of single-crystalline gold seeds alters the surface energies of the gold facets in the order {100} > {110} > {111}, inducing the formation of rhombic dodecahedral nanoparticles. Furthermore, they observed that cubic gold particles could also be formed by introducing bromide ions which replace chloride ions in CPC and contribute to stabilize the {100} facets.

Elucidation of the nanoparticles growth mechanism requires the careful morphological and crystallographic analysis not only of the preformed seeds and the final particles, but also of the different intermediate structures (sometimes containing relatively high energy facets). Such study is not always straightforward and the combination of different techniques is often required. Among the most common techniques, transmission and scanning electron microscopy (TEM and SEM), high resolution TEM (HRTEM) and selective area electron diffraction (SAED) deserve special attention.

We have recently described the growth mechanism of single crystal Au nanorods into single crystal octahedrons using N,N-dimethylformamide (DMF) as solvent and reducing agent, in the presence of poly(vinylpyrrolidone) (PVP).¹⁹ The rod to octahedron transition was found to involve intermediate steps where strongly faceted rods with sharp tips were formed. Through detailed morphological and crystallographic analysis, it was revealed that PVP might play an important role in the shape-guiding process by altering the relative surface energies of the fcc faces and therefore their relative growth rate (which followed the order {111} < {110} < {100}). Furthermore, it was also concluded that the crystallographic structure of the seeds determined the final particle morphology.²⁴ Keeping in mind our previous results, we decided to go further and analyze the growth process of pentatwinned (PTW) gold nanorods under similar experimental conditions, that is, using DMF as solvent and reducing agent and PVP as capping agent. In this case, besides the original gold nanorods and the final quasi-decahedra (see below), different intermediate particle morphologies obtained in the course of the process were also morphologically and structurally characterized by means of transmission and scanning electron microscopy (TEM and SEM), high resolution TEM (HRTEM) and selective area electron diffraction (SAED), in order to propose a growth mechanism.

Experimental

Chemicals

Tetrachloroauric acid (HAuCl₄ × 3H₂O), sodium borohydride (NaBH₄), ascorbic acid, cetyltrimethyl ammonium bromide (CTAB), citric acid (trisodium salt dihydrate) and N,N-dimethylformamide (DMF) were purchased from Aldrich. Poly(vinylpyrrolidone) (PVP, Mw 40,000) was supplied by Fluka. All chemicals were used as received. Milli-Q deionized water (resistivity higher than 18 MΩ cm⁻¹) was used for all preparations.

Synthesis of pentatwinned gold nanorods

The initial rods were prepared following a previously reported seeded growth method.²⁵ In a first step, 3.5 nm citrate stabilized Au nanoparticle seeds were prepared as follows: 20 mL of an aqueous solution containing 1.25 × 10⁻⁴ M HAuCl₄ and 2.5 × 10⁻⁴ M trisodium citrate was prepared in a conical flask at room temperature. Then, 0.3 mL of ice-cold, freshly prepared 0.01 M NaBH₄ solution was added to the solution under vigorous stirring. After 30 s, stirring was slowed down and the colloidal dispersion was kept between 40 and 45 °C for 15 min to ensure removal of excess NaBH₄. In a second step the seeds were grown to ca. 5.5 nm as follows: 5 mL of growth solution consisting of 1.25 × 10⁻⁴ M HAuCl₄ and 0.04 M CTAB at 25–30 °C was mixed under stirring with 0.0125 mL of 0.10M ascorbic acid. Subsequently, 1.67 mL of 3.5 nm Au-citrate seed solution was quickly added while stirring. As a result, CTAB capped Au nanoparticles with an average diameter of ca. 5.5 nm were obtained. In the final step, pentatwinned gold nanorods were grown. Briefly, to 250 mL of a growth solution ([HAuCl₄] = 0.125 mM and [CTAB] = 0.008 M) at 20 °C, 0.625 mL of 0.1 M ascorbic acid was added. After homogenization, 750 μL of 5.5 nm Au-CTAB seed solution was added and allowed to react for several hours. As a result, a mixture of spheres, plates and rods was obtained. To separate the gold nanorods, the method reported by Jana²⁶ was employed by centrifuging a total volume of 250 mL (25 tubes, 10 mL each) for 10 min at 6500 rpm. The supernatant was discarded and the precipitate redispersed in CTAB 0.1 M. The samples were concentrated again by centrifugation (10 min at 6500 rpm), and the concentrated sample (4 mL) was first heated at 50 °C for 5 min and then cooled down. After cooling, the precipitate was collected and redispersed in 9 mL of water. The gold (atom) concentration at this stage was approximately 1.96 mM. The purification (the precipitate contained mainly pentatwinned gold nanorods – spheres and plates were removed) was monitored by registering extinction spectra in the vis-NIR range. Whereas in the original sample two intense bands were registered at 530 and 990 nm, corresponding to the surface plasmon band (SPB) of the spheres and the longitudinal SPB of the rods, respectively; after purification only the band at 990 nm was maintained, whereas a weak band at 511 nm, corresponding to the rods transverse SPB, could also be identified in the spectrum (see Figure S1 in the Electronic Supplementary Information (ESI)).†

PVP coating and ethanol transfer

A previously reported procedure²⁷ was used for coating the rods with PVP. 9 mL of Au-CTAB nanorods (1.96 mM) was mixed with an aqueous PVP solution (2.35 mM, 9 mL) and stirred overnight. Then, the mixture was centrifuged at 4500 rpm for 80 min, the clear supernatant discarded and the precipitate redispersed in ethanol (2 mL) under vigorous stirring. The gold concentration at this stage was approximately 4.5 mM.

Growth of pentatwinned gold nanorods

Before adding the seed solution, a certain amount of HAuCl₄ aqueous solution (50 mM) was added to a solution of PVP in DMF (2.5 mM, 7.5 mL) in a 20 mL beaker. The final Au³⁺ concentration ranged from 0.08 to 0.58 mM. The mixture was sonicated until reduction of Au³⁺ into Au⁺ was complete (disappearance of the Au³⁺ CTTS absorption band at 325 nm). Then, 25 μL of the Au rod (4.5 mM) solution in ethanol was added and the mixture was irradiated with ultrasound until the process was completed. Ultrasonic irradiation was performed with a Bandelin Sonopuls HD2200 ultrasonic homogenizer operating at a frequency of 20 kHz and at 12% of the maximum power (200 W) to maintain the solution temperature at 75 °C during the process. It is well-known that rods can undergo thermal reshaping at moderate temperature.¹⁹ In order to avoid this effect during the growth, 75 °C was selected as the optimum temperature to carry out the growth experiments and therefore, all the observed morphological changes during growth can be attributed to the deposition of gold atoms on the nanorods surface.

Characterization

Vis-NIR extinction spectra were recorded using a Cary 5000 UV-vis-NIR spectrophotometer. Measurements were made every 10 min after the reaction was initiated, by stopping the process and withdrawing aliquots. A JEOL JEM 1010 transmission electron microscope operating at an acceleration voltage of 100 kV was used for low magnification imaging. HRTEM (high resolution transmission electron microscopy) images and SAED (selected area electron diffraction) patterns were obtained with a JEOL JEM 2010 FEG TEM operating at an acceleration voltage of 200 kV. The samples for TEM were centrifuged three-fold at 3500 rpm and redispersed in ethanol to avoid the dissolution of the polymer layer on the TEM grid by DMF as well as to decrease the PVP concentration in the sample. The specimens were prepared by depositing a droplet on FORMVAR-carbon-coated (for low magnification imaging) or carbon-coated grids (for SAED and HRTEM) and evaporating the solvent in air at room temperature.

Results and discussion

Pentatwinned gold nanorods were synthesized using the seed-mediated growth method through reduction of HAuCl₄ with ascorbic acid, in the presence of preformed small seeds and CTAB.²⁵ This strategy yields a mixture of nanospheres and nanorods (ca. 50% each), which requires sphere removal prior to subsequent handling the gold nanorods (see Experimental Section and Figure S1 in ESI).† Since the growth was to be carried out in DMF and CTAB capped nanoparticles are not stable in DMF, an additional step was required to exchange CTAB by PVP, prior to solvent transfer. These procedures have been reported to work for single crystal nanorods without aggregation,¹⁹ and successful transfer was also obtained in the present case, as reflected in the corresponding visible-NIR extinction spectra and the TEM images measured for the colloid in ethanol, as exemplified in Fig. 1. From the analysis of the TEM images, an average aspect ratio of 5.2 ± 0.5 (124.4 ± 5.6 nm × 23.9 ± 1.7 nm) was obtained.

Fig. 1 a) Visible-NIR extinction spectrum of the pentatwinned gold nanorods in ethanol. b) Representative TEM micrograph of the same sample.

The pentatwinned Au nanorods were grown through reduction of HAuCl₄ by DMF under ultrasound irradiation, in the presence of PVP, as reported earlier (see Experimental Section for details).^15,19 Since the reduction of the Au ions takes place on the particle surface (which acts as a catalyst), it leads to gradual changes in particle size and morphology, as evidenced by the measured optical changes (Fig. 2B). Regardless of the amount of reduced Au salt, the trend in the spectral evolution involves a blue shift of the longitudinal SPB into a single band, which eventually broadens and splits again into a new second band. Examples of spectral evolution during the growth of gold nanorods with different amounts of HAuCl₄ are displayed in Figure S2 (ESI).† As expected, larger relative amounts of gold salt lead to more pronounced optical changes; but, as previously reported for single crystalline nanorods, the size and morphology of the particles can be tuned through the initial ratio between HAuCl₄ concentration and amount of Au seed (R). As shown in Fig. 2A (see also Figure S3 in ESI),† varying R from 0 to 38.7, particles with different intermediate morphologies between the initial pentagonal cross section rods and the final truncated quasi-decahedra (see below) can be obtained. The blue shift of the longitudinal SPB (Fig. 2B), initially centred at 1025 nm, can be easily related to the decrease in the aspect ratio of the particles. Indeed, the TEM images of Fig. 2A indicate that the growth of the particles does not lead to an increase in their length but they rather get thicker (see also Table S1 in ESI).† When a sufficient amount of gold was grown on the rods (R > 6), the longitudinal SPB was observed to merge with the transverse SPB into a single band, which indicates that the initial anisotropy of the particles has been lost through growth. Further growth leads to an increase of intensity, a subsequent red-shift and broadening of the band, as expected for the growth of quasi-isotropic particles. Finally, for R > 14, a new band appears at 610 nm, which can be assigned to a quadrupolar surface plasmon resonance mode, due to the increased particle size.²⁸

Fig. 2 (A) TEM micrographs of the original Au nanorods (a) and final particles resulting from reduction of different amounts of HAuCl₄ (b–h). (B) Visible-NIR extinction spectra of the corresponding colloids shown in (A). The [HAuCl₄] : [Au_seed] ratios are: (a) 0, (b) 5.3, (c) 8.5, (d) 9.7, (e) 14.0, (f) 19.6, (g) 23.6 and (h) 38.7.

In order to disclose the mechanism involved in the growth of the pentatwinned rods, the morphology and crystalline structure of the final particles obtained for different R values were analysed using TEM, high resolution TEM (HRTEM), selected area electron diffraction (SAED) and scanning electron microscopy (SEM). Although a similar study was previously reported for single-crystal gold nanorods,¹⁹ the difficulty is significantly higher in the present case for several reasons. On one hand, these particles are pentatwinned crystals and therefore polycrystalline, exhibiting five crystalline domains with different orientations along the common [110] twinning axis, and on the other hand the lateral facets of the particle are all of the same type ({100}, see below) but in different relative orientations. This different orientation is a limiting factor for the analysis, which can however be overcome if we take into account previous works related to the crystalline characterization of pentatwinned gold nanorods.^29,30 It has been reported that two main different orientations with respect to electron beam are possible for pentatwinned nanorods.²⁹ In both orientations, the particles show the overlap of two different zone axes: <110> and <111> or <112> and <100> (see Fig. 3b–c). An example of this characterization is shown in Fig. 3. The diffraction patterns in Fig. 3b2–c2 were obtained from the same particle and are related to each other through a clockwise rotation of 18° around the common [110] axis, in agreement with the pentagonal cross-section. In our present study we selected the orientation corresponding to a combination of <110> and <111> zone axes (Fig. 3b1), since it is the only one that allows us to follow the growth of the {100} lateral facets of the particles. As shown in Fig. 3b1, in the <110> and <111> zone axis combination the electron beam is nearly parallel to the side facet on the right, while on the opposite side one of the twin boundaries of the rod is perpendicular to the electron beam. Consequently, the projections of the particles in the TEM images will be asymmetric along the [110] twinning axis of the original rod.

Fig. 3 a) Scheme of a pentatwinned gold nanorod, the common twinning axis is [110]. b1) and c1) Schemes of the pentatwinned nanorod cross section oriented in the <110> and <111>, and in the <112> and <100> zone axes, respectively. The corresponding diffraction patterns are shown in b2 and c2, respectively. Note that only in b1 the electron beam is parallel to a side facet. The diffraction patterns were obtained on the very same nanoparticle and are related to each other by a rotation of 18° around the common [110] axis.

Thus, in practice, all analyzed particles (R = 5.3, 23.6 and 38.7), including the original nanorods, were rotated in the TEM grid until a SAED pattern corresponding to the “<110> and <111> orientation” was registered (maintaining the common [110] twinning axis perpendicular to the electron beam), and the corresponding TEM projection image was then obtained, as well as the HRTEM images of the particle edges. Although the low resolution TEM images together with the corresponding SAED pattern provide the information we need to follow the changes occurring during growth, there is still a significant challenge in identifying which side of the projection corresponds to the edge and which one to the facet (see Fig. 3b1). This issue was resolved by determining the Fourier Transform (FT) from a selected area (in the HRTEM image) of each side of the particle (see Figure S4, ESI).† Shown in Fig. 4 are TEM images of the analyzed particles, oriented in the <111> and <110> zone axes (a–d), as well as their corresponding SAED patterns upon compensation of electron diffraction rotation (e–h).

Fig. 4 (Top row) Representative TEM images of the initial rods, R = 0 (a), and particles obtained at R = 5.3 (b), R = 23.6 (c) and R = 38.7 (d). (Bottom row) Corresponding electron diffraction patterns of the displayed particles (e–h). Electron diffraction rotations were compensated.

Although the nature of the lateral facets is a controversial issue in the literature,^29–31 our evidences suggest that the particles present five {100} lateral facets and ten {111} tip facets. Therefore, we decided to perform the study of the growth mechanism following the evolution of the {100} lateral facet and the <100> edge. As all the particles in Fig. 4 are oriented in the same zone axis and it appears that they grow wider, with no significant changes in particle length (see Table S1, ESI),† the deposition of new gold atoms is assumed to take place always on the lateral facets, which is consistent with their expected relative energies. All SAED patterns measured from the grown particles are consistent with a pentatwinned face-centred cubic structure, which confirms that the growth process does not affect the pentatwinned crystalline structure of the original nanorods (see also Fig. 5a), which also points toward deposition of gold atoms in an epitaxial fashion. Thus, for example, the growth of the original nanorods into the particles shown in Fig. 4b (R = 5.3) can be visualized as the deposition of gold atoms on the {100} side facets, therefore maintaining their pentagonal cross-section. The structural analysis of the particle in Fig. 4b shows that it is delimited by two {113} facets at the right side and by two edges at the left. The morphological changes observed for particles with R = 23.6 and 38.7 (Fig. 4c and 4d) are even more drastic, because the deposition of additional gold atoms on the {113} facets gives rise to asymmetric particles along the common [110] five-fold axis, with two new long {110} facets at the left and two edges and three small facets (one {100} and two {111}), at the right. This particle also has {110} facets in perpendicular direction to the [110] zone axis, at both the upper and lower ends. The biggest particles obtained, R = 38.7, are found quite similar to those obtained with R = 23.6, since deposition of new gold atoms occurs on {110} facets, which leads to observation of only two new edges in the projection images (Fig. 4d).

Fig. 5 (a) TEM image of a particle obtained with R = 23.6, oriented in the [110] zone axis. Note that the twinning interfaces end at the middle of each edge (red arrow). The corresponding SAED pattern demonstrating the pentatwinned structure is shown in the inset (see Figure S5 in ESI for the indexation).† (b–d) SEM images of particles obtained with R = 23.6 (b) and R = 38.7 (c,d), where the truncated decahedral morphology is evident. For clarity, the indexes of the main facets, the edges (dashed lines), and the twin interfaces (red lines) are depicted.

Fig. 5a shows a TEM image of the top view of a particle similar to that shown in Fig. 4c (R = 23.6), that is, oriented in the common [110] zone axis. This image clearly evidences the pentagonal cross-section, which, combined with Fig. 4c, indicates a truncated decahedron morphology. The inset shows the diffraction pattern obtained in the same orientation, which again confirms the [110] orientation (a complete indexation of this SAED pattern is provided in Figure S5, ESI).† The main difference between this truncated-decahedron morphology and a perfect decahedron is that each twinning interface ends at the middle point of an edge of the pentagonal projection, rather than at a tip (see comparison in Figure S6, ESI).† The actual geometry of these particles is more clearly observed through SEM characterization (Fig. 5b). In the image, the position of a particular twin was indicated by red lines. Despite some small differences, the resemblance between particles grown with R = 23.6 and those grown with R = 38.7 is obvious in the SEM images (Fig. 5b and 5c). Both particles present a new {110} facet at the plateau perpendicular to the longitudinal axis of the initial rod, and five small {111} facets that form the beautiful pentagonal star at the top. Thus, the growth from the particle with R = 23.6 through deposition of gold atoms on the {110} lateral facets does not give rise to new facets, while the growth of {111} facets leads to the appearance of new edges, as seen in Fig. 5c.

To clarify the growth process, we propose a representative atomic model of all these structures, on the basis of all the information obtained from SEM, TEM and SAED (Fig. 6). This atomic model explains graphically the mechanism involved in the transition from pentatwinned gold nanorods into truncated decahedrons. Since all five twins of the particles are equivalent, just one twin was depicted in the atomic model. However, the whole particle model is also displayed for more intuitive understanding. The initial rods present a prismatic morphology with a pentagonal cross-section, enclosed by five {100} lateral facets and ten {111} tip facets (Fig. 6, structure 1). During the growth, preferential addition of gold atoms on the {100} facets occurs, leading to consumption of these facets and creation of two new {113} facets in each twin (structure 2), so that the resulting particles maintain the original pentagonal cross section. The ten {111} tip facets remain almost unaltered in structure 3, but two small new {110} facets appear at the top and the bottom of the particles, while four {110} facets form in the lateral sides and join at the apexes forming a small {100} facet (Fig. 6, structure 3). The growth from this geometry to the final one is based on the preferential addition of gold atoms on {110} facets, which leads to bigger {111} facets and smaller {110} facets (Fig. 6, structure 4). Interestingly, all the growth processes described here are consistent with the faster growth of facets with higher energy, which in this case, rather than simplifying (as previously reported for single crystal rods¹⁹) leads to more complicated morphologies. However, such new morphologies are again closer to the spherical shape, which explains the measured blue-shift in the surface plasmon band of the nanoparticle colloids.

Fig. 6 Scheme of the proposed growth mechanism. 1) Initial PTW gold nanorods. 2) Particles grown with R = 5.3. 3) Particles grown with R = 23.6. 4) Particles grown with R = 38.7. Red spheres represent the original rod structure. Atomic models (A1–A4) and solid models (B1–B4) showing the evolution of one of the particle twins. (C1–C4) Solid models showing the evolution of the whole particle. (D1–D4) TEM projection images, which were rotated 180° as compared to Fig. 4 for a better comparison with their respective models.

Conclusion

Careful analysis of the crystallographic and morphological structure of the different particles obtained during the controlled growth of pentatwinned gold nanorods in DMF, in the presence of PVP, allowed us to elucidate the mechanism and morphological transformations involved in the process. Thus, it was observed that the growth at the tips is inhibited and there is a preferential growth of the lateral, more energetic {100} facets, so that the starting nanorods grow wider rather than longer, eventually giving rise to truncated decahedrons. Additionally, the crystallographic analysis also revealed that all the intermediate particles retain a pentatwinned cross-section, thereby confirming the key role of the seeds in determining the crystalline structure of the grown nanoparticles, in accordance with previous studies.

Acknowledgements

E.C.A. acknowledges an FPU scholarship from the Spanish Ministerio de Ciencia e Innovación (MiCInn/FEDER). This work has been funded by MiCInn (MAT2007-62696) and Xunta de Galicia (09TMT011314PR).

References

1. K. M. Bratlie, H. Lee, K. Komvopoulos, P. Yang and G. A. Somorjai, Nano Lett., 2007, 7, 3097–3101.
2. L. M. Liz-Marzán, Langmuir, 2006, 22, 32–41.
3. G. D. Scholes, Adv. Funct. Mater., 2008, 18, 1157–1172.
4. L. Rodríguez-Lorenzo, R. Alvarez-Puebla, I. Pastoriza-Santos, S. Mazzucco, O. Stéphan, M. Kociak, L. M. Liz-Marzán and F. J. García de Abajo, J. Am. Chem. Soc., 2009, 131, 4616–4618.
5. B. Sepúlveda, P. C. Angelomé, L. M. Lechuga and L. M. Liz-Marzán, Nano Today, 2009, 4, 244–251.
6. R. Wilson, Chem. Soc. Rev., 2008, 37, 2028–2045.
7. Special Issue in Chemical and bioanalytical applications of surface enhanced Raman scattering spectroscopy, Chem. Soc. Rev., 2008, 37.
8. N. R. Jana, L. Gearheart and C. J. Murphy, Chem. Mater., 2001, 13, 2313–2322.
9. J. Rodríguez-Fernández, J. Pérez-Juste, F. J. García de Abajo and L. M. Liz-Marzán, Langmuir, 2006, 22, 7007–7010.
10. J. Pérez-Juste, I. Pastoriza-Santos, L. M. Liz-Marzán and P. Mulvaney, Coord. Chem. Rev., 2005, 249, 1870–1901.
11. Y. Sun and Y. Xia, Science, 2002, 298, 2176–2179.
12. J. Zhang, Y. Gao, R. A. Alvarez-Puebla, J. M. Buriak and H. Fenniri, Adv. Mater., 2006, 18, 3233–3237.
13. C. Li, K. L. Shuford, M. Chen, E. J. Lee and S. O. Cho, ACS Nano, 2008, 2, 1760–1769.
14. A. Sánchez-Iglesias, E. Carbó-Argibay, A. Glaria, B. Rodríguez-González, J. Pérez-Juste, I. Pastoriza-Santos and L. M. Liz-Marzán, Chem. Eur. J., 2010, 16, 5558–5563.
15. A. Sánchez-Iglesias, I. Pastoriza-Santos, J. Pérez-Juste, F. J. García de Abajo and L. M. Liz-Marzán, Adv. Mater., 2006, 18, 2529–2534.
16. D. Seo, C. I. Yoo, I. S. Chung, S. M. Park, S. Ryu and H. Song, J. Phys. Chem. C, 2008, 112, 2469–2475.
17. L. D. Marks, Rep. Prog. Phys., 1994, 57, 603–649.
18. L. Vitos, A. V. Ruban, H. L. Skriver and J. Kollár, Surf. Sci., 1998, 411, 186–202.
19. E. Carbó-Argibay, B. Rodríguez-González, J. Pacifico, I. Pastoriza-Santos, J. Pérez-Juste and L. M. Liz-Marzán, Angew. Chem., Int. Ed., 2007, 46, 8983–8987.
20. F. Kim, S. Connor, H. Song, T. Kuykendall and P. Yang, Angew. Chem., Int. Ed., 2004, 43, 3673–3677.
21. D. Seo, C. I. Yoo, J. C. Park, S. M. Park, S. Ryu and H. Song, Angew. Chem., Int. Ed., 2008, 47, 763–767.
22. B. J. Wiley, Y. Xiong, Z.-Y. Li, Y. Yin and Y. Xia, Nano Lett., 2006, 6, 765–768.
23. W. Niu, S. Zheng, D. Wang, X. Liu, H. Li, S. Han, J. Chen, Z. Tang and G. Xu, J. Am. Chem. Soc., 2009, 131, 697–703.
24. I. Pastoriza-Santos and L. M. Liz-Marzán, Adv. Funct. Mater., 2009, 19, 679–688.
25. J. Pérez-Juste, L. M. Liz-Marzán, S. Carnie, D. Y. C. Chan and P. Mulvaney, Adv. Funct. Mater., 2004, 14, 571–579.
26. N. R. Jana, Chem. Commun., 2003, 1950–1951.
27. C. Graf, D. L. J. Vossen, A. Imhof and A. van Blaaderen, Langmuir, 2003, 19, 6693–6700.
28. I. Pastoriza-Santos, A. Sánchez-Iglesias, F. J. García de Abajo and L. M. Liz-Marzán, Adv. Funct. Mater., 2007, 17, 1443–1450.
29. C. J. Johnson, E. Dujardin, S. A. Davis, C. J. Murphy and S. Mann, J. Mater. Chem., 2002, 12, 1765–1770.
30. J. Reyes-Gasga, J. L. Elechiguerra, C. Liu, A. Camacho-Bragado, J. M. Montejano-Carrizales and M. J. Yacaman, J. Cryst. Growth, 2006, 286, 162–172.
31. P. L. Gai and M. A. Harmer, Nano Lett., 2002, 2, 771–774.

---

Footnote

† Electronic supplementary information (ESI) available: Table S1 and Fig. S1–S6. See DOI: 10.1039/c0nr00239a

---