Rapid Synthesis of Highly Monodisperse AuxAg1−x Alloy

Apr 4, 2011 - For thousands of years, ancient Egyptians carefully mummified the bodies of their dead in readiness for... BUSINESS CONCENTRATES ...
0 downloads 0 Views 5MB Size
ARTICLE pubs.acs.org/Langmuir

Rapid Synthesis of Highly Monodisperse AuxAg1x Alloy Nanoparticles via a Half-Seeding Approach Ting Ting Chng,†,‡ Lakshminarayana Polavarapu,‡ Qing-Hua Xu,‡ Wei Ji,§ and Hua Chun Zeng*,† †

Department of Chemical and Biomolecular Engineering and KAUST-NUS GCR Program, Faculty of Engineering, Department of Chemistry, and §Department of Physics, Faculty of Science, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260



bS Supporting Information ABSTRACT: Goldsilver alloy AuxAg1x is an important class of functional materials promising new applications across a wide array of technological fields. In this paper, we report a fast and facile synthetic protocol for preparation of highly monodisperse AuxAg1x alloy nanoparticles in the size range of 36 nm. The precursors employed in this work are M(I)alkanethiolates (M = Au and Ag), which can be easily prepared by mixing common chemicals such as HAuCl4 or AgNO3 with alkanethiols at room temperature. In this half-seeding approach, one of the M(I)alkanethiolates is first heated and reduced in oleylamine solvent, and freshly formed metal clusters will then act as premature seeds on which both the first and second metals (from M(I)alkanethiolates, M = Au and Ag) can grow accordingly without additional nucleation and thus achieve high monodispersity for product alloy nanoparticles. Unlike in other prevailing methods, both Au and Ag elements present in these solid precursors are in the same monovalent state and have identical supramolecular structures, which may lead to a more homogeneous reduction and complete interdiffusion at elevated reaction temperatures. When the M(I)alkanethiolates are reduced to metallic forms, the detached alkanethiolate ligands will serve as capping agent to control the growth. More importantly, composition, particle size, and optical properties of AuxAg1x alloy nanoparticles can be conveniently tuned with this approach. The optical limiting properties of the prepared particles have also been investigated at 532 and 1064 nm using 7 ns laser pulses, which reveals that the as-prepared alloy nanoparticles exhibit outstanding broadband optical limiting properties with low thresholds.

1. INTRODUCTION Bimetallic nanoparticles belong to a new class of versatile functional materials whose chemical and physical properties can be tuned by simply changing the composition and particle shape and size. For example, these nanoparticles have displayed attractive catalytic activities due to the synergistic effect from individual components.14 Apart from the application in the field of electrocatalysis,5,6 bimetallic nanoparticles have been found to exhibit optical limiting properties and applicability for use in biodiagnostics.7,8 With the found versatility and enhanced properties of bimetallic nanoparticles, a myriad of opportunities exist to tap into their various potential applications. In general, bimetallic nanoparticles can be broadly divided into two main categories: alloy nanoparticles, where the two kinds of metals are mixed homogeneously, and coreshell nanoparticles, where the core and shell metals differ. Because Au and Ag are among the most important noble metals, bimetallic nanoparticles made from them have attracted significant interest. In particular, several methods have been employed to synthesize AuxAg1x (or AuAg as a simplified short form) alloy nanoparticles, and the most popular ones are the co-reduction of their common metal precursors (e.g., HAuCl4 and AgNO3) in the presence of reducing agents, such as hydrazine and citrate in an aqueous solution,913 and the photochemical method.14 Nevertheless, the direct use of r 2011 American Chemical Society

combination of HAuCl4 with AgNO3 in this type of alloy preparation has the disadvantage of having to take precaution against the formation of silver chloride, which can often lead to failure in the synthesis. The precipitation of silver chloride has been overcome by using methods such as laser ablation of bulk alloys, which also result in the formation of alloy nanoparticles.15 The particles synthesized are, however, not monodisperse. Recently, synthesis of AuAg alloy nanoparticles has also been achieved by a replacement reaction between silver nanoparticles and the gold metal complex.16,17 Heating silver and gold precursors in oleylamine also yields AuAg alloy nanoparticles.18 However, the silver precursor has to be added in great excess in order to compensate for its slower reduction, and this results in an unpredictable composition of the final nanoparticles and possible waste of this precious metal. Digestive ripening of Au@Ag coreshell nanoparticles in 4-tertbutyltoluene for 8 h can also lead to the formation of AuAg alloys,19 which is similar to a previous work of annealing Ag@Au coreshell nanoparticles in oleylamine at 100 °C to obtain the same alloy nanoparticles,20 in which oleylamine was used as a Received: February 9, 2011 Revised: March 8, 2011 Published: April 04, 2011 5633

dx.doi.org/10.1021/la2005216 | Langmuir 2011, 27, 5633–5643

Langmuir

ARTICLE

solvent, reductant, and capping agent. Small AuAg alloy nanoparticles less than 5 nm protected with alkanethiolates and dendrimers have also been synthesized, but the reaction time for these procedures takes several hours.21,22 It should be pointed out that, in contrast to the formation of AuAg alloy nanoparticles, the procedures for synthesis of Ag@Au or Au@Ag coreshell nanoparticles involve the deposition of the shell metal onto the preformed core metal particles.20,2326 Metal nanoparticle systems have emerged as promising materials for nonlinear optical (NLO) applications.2731 Optical limiting materials display a high transmittance at low input fluence and low transmittance at high input fluence, and they are thus very useful to protect the eyes and sensors from laserinduced damage. In this connection, optical limiting properties of gold and silver nanoparticles have been extensively studied, and large optical limiting effects were observed from silverdendrimer nanocomposites,27 gold nanoparticle/polylysine hybrid material,28 and gold nanoparticle aggregates.29 The size effect on the optical limiting performance has also been studied, and it is found that increasing the size of gold nanoparticles enhances the performance due to the nonlinear scattering effect.30 Although there are many investigations on the optical limiting properties of pure gold or pure silver nanoparticles, very few studies have been devoted to the NLO properties of AuAg alloy nanoparticles. For instance, Au and Ag nanoparticles have been found to exhibit similar optical limiting efficiency, but the AuAg alloy nanoparticles are less efficient with picosecond laser pulses of wavelength 532 nm.7,31 Therefore, it is desirable to study the broadband optical limiting properties of AuAg alloy nanoparticles in order to optimize their composition for better optical limiting performance. In this paper, we present our recent research for a fast and facile half-seeding approach to synthesize highly monodisperse alkanethiolate-protected AuAg alloy nanoparticles in a tunable range of 36 nm using supramolecular M(I)alkanethiolates (M = Au or Ag) as solid precursors. The total time for reaction takes less than 30 min, and the conventionally combined use of HAuCl4 and AgNO3 can be avoided, which frees us from the limitation of having to control the concentration of the two metal precursors. The proposed approach also greatly simplifies the synthetic procedure and enables the mixing of the precursors in any mole ratio of Ag:Au, such that tunable optical properties of the metal nanoparticles can be obtained. Our present alloy synthesis is motivated by a recent study that the monodispersity of pure Au nanoparticles could be achieved from heating Au(I)dodecanethiolate nanotubes in hexadecylamine.32 The bilayered solid precursor Au(I) dodecanethiolate,33 which is isostructural to its counterpart Ag(I)dodecanethiolate,3436 can be simply formed from dodecanethiol and HAuCl4 at room temperature. In general, the M(I)alkanethiolates can be synthesized by mixing an excess of alkanethiols with their respective metal precursors; the following equation is for the case of gold.33,37 HAuCl4 þ 3RSH f AuðIÞ  SR þ RS  SR þ 4HCl

ð1Þ

The formation of M(I)alkanethiolates (M = Au or Ag) can be completed within a very short time scale of only 5 min. These supramolecular solids are stable and are ideal to be used as precursors of their metallic phases. Because they are bonded directly to Au(I) or Ag(I), alkanethiolates in these precursors can further function as a capping surfactant for the metal products as

Figure 1. Illustration of synthetic schemes developed in this work (OA = oleylamine and ddt = dodecanethiolate): (a) Ag(I)ddt precursor is added at two different times, (b) all Au(I)ddt precursor is added in the beginning, (c) Au(I)ddt and Ag(I)ddt precursors are added respectively at two different times, (d) a reverse process of scheme c, and (e) Au(I)ddt and Ag(I)ddt precursors are added together in the beginning.

well. As described in Figure 1, the precursor M(I)alkanethiolates are added sequentially into oleylamine solvent, which also acts as a reducing agent for the redox reactions involved in the metal formation. This approach, though somewhat similar to the sequential synthesis of coreshell materials, results in rapid formation of AuAg alloy nanoparticles due to the high reaction temperature, which also facilitates the diffusion and intermixing of metal atoms. Importantly, when the second M(I)alkanethiolate is added, the formation of the first metal nanoparticles has not yet been complete, as indicated by the faintness of the color of the precursor suspension, noting that darker purple/brown color would have been observed if the formation of the gold/silver nanoparticles had reached completion. We also note that seeding methods have been used to separate the nucleation from growth in order to improve particle uniformity.3840 In comparison, because it does not strictly involve the deposition of a shell on a preformed core, our synthetic protocol can be considered as a “half-seeding” approach, where the first formed clusters can act 5634

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir as “premature seeds” into which both the first metal and the second metal can be continuously integrated. Due to the presence of clusters of the first metal, any secondary nucleation could have been inhibited upon the sudden injection of the second M(I)alkanethiolate. In addition to the synthetic experiments, optical limiting properties of the above-prepared monometallic and bimetallic alloy nanoparticles have also been investigated by using 7 ns laser pulses at two wavelengths of 532 and 1064 nm. The transmittance and scattering of light through the metal nanoparticle suspensions have been measured as a function of the laser fluence, and the mechanism for the observed optical limiting behavior will also be discussed.

2. EXPERIMENTAL SECTION 2.1. Chemicals. Gold(III) chloride trihydrate (HAuCl4 3 3H2O, 99.9þ%, Aldrich), 1-dodecanethiol (C12H25SH, DDT, 98þ%, Aldrich), silver nitrate (AgNO3, Mallinckrodt Chemical), oleylamine [CH3(CH2)7 CHdCH(CH2)7CH2NH2, Acros Organics, 8090%], toluene (C7H8, J. T. Baker), and ethanol (C2H5OH, Merck) were used in the synthesis without further purification. 2.2. Preparation of Au(I)Alkanethiolate (Au(I)ddt). HAuCl4 3 3H2O was dissolved in ethanol solvent to a concentration of 0.100 M, and DDT was also dissolved in ethanol to a concentration of 0.100 M. In the initial precursor solution, the mole ratio of HAuCl4 3 3 H2O to DDT was kept at 1:5 (i.e., Au3þ:DDT = 1:5). Details of a typical experimental procedure were as such: at room temperature, 1.0 mL of DDT in ethanol (0.100 M) was added to 2.4 mL of ethanol, following which 0.2 mL of HAuCl4 3 3H2O in ethanol (0.100 M) was added and the mixture stirred for 5 min (total volume 3.6 mL). The resultant light brown precipitate was washed twice with ethanol and then dispersed in ethanol to a concentration of ∼0.005 M.

2.3. Preparation of Ag(I)Alkanethiolate (Ag(I)ddt). AgNO3 salt was dissolved in ethanol solvent to a concentration of 0.010 M, and DDT was also dissolved in ethanol to a concentration of 0.100 M. In the initial precursor solution, the mole ratio of AgNO3 to DDT was kept at 1:5 (i.e., Agþ:DDT = 1:5). Details of a typical experimental procedure were as such: at room temperature, 1.0 mL of DDT in ethanol (0.100 M) was added to 0.6 mL of ethanol, following which 2.0 mL of AgNO3 in ethanol (0.010 M) was added and the mixture stirred for 5 min (total volume 3.6 mL). The resultant white precipitate was washed twice with ethanol and then dispersed in ethanol to a concentration of ∼0.005 M. 2.4. Characterization of the M(I)Alkanethiolates. The above prepared M(I)alkanethiolates (Au(I)ddt and Ag(I)ddt) were characterized by powder X-ray diffraction technique [XRD; Shimadzu X-ray diffractometer, model 6000 with Cu KR radiation (λ = 1.5406 Å) from 0.5° to 35° (2θ angle) at a scanning speed of 2° min1] and ultravioletvisible spectrophotometry (UVvisible; Shimadzu UV-2450). 2.5. Preparation of AuAg Alloy Nanoparticles. A certain volume (0.3, 0.5, and 0.7 mL, respectively) of the ethanolic suspension of the above Au(I)ddt (∼0.005 M) was centrifuged to remove the bulk of the ethanol. A 10.0 mL portion of oleylamine was then added to the semidried solid and the mixture heated to 200 °C in a silicon oil bath. Immediately upon detecting a slight color change, a certain volume (0.7, 0.5, and 0.3 mL, respectively) of Ag(I)ddt (0.005 M, in ethanol) was added. The Au/(Ag þ Au) mole fractions investigated were 0.3, 0.5, and 0.7, which corresponds to 0.3, 0.5, and 0.7 mL of the ethanolic suspension of the Au(I)ddt, respectively. The total number of moles of both metals was kept a constant for all experiments. The heating was continued under reflux for a total of 1215 min. When the reaction was complete,

ARTICLE

aggregation or fusion among the nanoparticles was avoided by quenching the hot solidliquid mixture with the addition of it to a large volume of ethanol (typically 80 mL). The product nanoparticles were then washed with ethanol and redispersed in toluene. To facilitate our discussion in the next section, the above alloy nanoparticle samples are represented by AuAg(X), where X is the metal mole fraction of Au in the starting precursors. For instance, AuAg(0.3) represents a binary alloy sample prepared with an Au(I)ddt:Ag(I)ddt mole ratio of 3:7. Besides the above syntheses, the addition sequence of the M(I)alkanethiolate precursors was also reversed. For example, the heating of the Ag(I)ddt in oleylamine, followed by the addition of Au(I)ddt in ethanol, had also been performed for the Au/(Ag þ Au) mole fraction of 0.5. 2.6. Preparation of Ag Nanoparticles. A 0.5 mL portion of the ethanolic suspension of the Ag(I)ddt (0.005 M) was centrifuged to remove the bulk of the ethanol. Ten milliliters of oleylamine was then added to the semidried solid and the mixture heated to 200 °C in a silicon oil bath. Immediately upon detecting a slight color change, another 0.5 mL of Ag(I)ddt in ethanol was added. The heating was continued under reflux for a total of about 15 min. After this time, the hot solidliquid mixture was quenched by adding it to a large volume of ethanol (80 mL). The Ag nanoparticles were then washed with ethanol and redispersed in toluene. 2.7. Preparation of Au Nanoparticles. A 1.0 mL portion of the ethanolic suspension of the Au(I)ddt (0.005 M) was centrifuged to remove the bulk of the ethanol. Ten milliliters of oleylamine was then added to the semidried solid and the mixture heated to 200 °C in a silicon oil bath. After about 10 min, the hot solidliquid mixture was quenched by adding it to a large volume of ethanol (80 mL). The Au nanoparticles were then washed with ethanol and redispersed in toluene. 2.8. Characterization of the Metal Nanoparticles. The optical properties of the as-prepared gold, silver, and their alloy nanoparticles were characterized with ultravioletvisible spectrophotometry (UVvisible; Shimadzu UV-2450). The particle size and morphology were characterized by transmission electron microscopy (TEM; JEM-2010) and high-resolution TEM (HRTEM; JEM 2100F). The presence of surfactants in the nanoparticle products was also characterized by Fourier-transform infrared spectroscopy (FTIR; Bio-Rad FTS135) using a KBr pellet technique. The composition of the product alloy nanoparticles was determined by inductively coupled plasma mass spectrometry (ICP-MS; Agilent Technologies 7500 Series). 2.9. Optical Limiting Measurement. The broadband optical limiting properties of the metal nanoparticles dispersed in toluene solvent were studied by input fluence-dependent transmittance measurements using the nanosecond laser at wavelengths of 532 and 1064 nm. Nanosecond laser pulses of pulse width 7 ns were generated from a Q-switched neodymium-doped yttriumaluminumgarnet (Nd:YAG) laser (Spectra Physics DCR3) with a repetition rate of 10 Hz. The original output wavelength of the laser was 1064 nm, which was frequency-doubled to obtain the 532 nm laser output. Each sample of nanoparticles dispersed in toluene was placed in a 1 cm cuvette and the laser beam focused on the metal nanoparticles suspension with a spot size of 165 and 330 μm for the experiments using the 532 and 1064 nm laser sources, respectively. The transmitted light was collected in the propagation direction of the laser beam and the scattered light at an angle of 30° from the propagation direction of the laser beam. The concentration of the nanoparticle suspensions in the toluene solvent was adjusted, such that the linear transmittance was ∼70% for experiments employing the 532 nm laser and ∼80% for experiments employing the 1064 nm laser.

3. RESULTS AND DISCUSSION 3.1. Synthesis of the M(I)Alkanethiolates (M = Au or Ag). In this work, the supramolecular solid precursors M(I)alkanethiolates 5635

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

ARTICLE

Figure 3. (a) Normalized UVvis spectra of the Ag (scheme a, Figure 1), AuAg (scheme c, Figure 1), and Au (scheme b, Figure 1) nanoparticles dispersed in toluene solvent and (b) plot of wavelength at UVvis peak maxima (refer to part a) versus the metal mole fraction of Au in the reactant feed.

Figure 2. (a) Normalized XRD patterns of Au(I)ddt precursor formed in 5 min and 24 h, (b) normalized XRD patterns of Ag(I)ddt precursor formed in 5 min and 24 h, and (c) normalized UVvis absorbance spectra of Au(I)ddt and Ag(I)ddt precursors dispersed in ethanol solvent.

(M = Au or Ag) can be prepared rapidly within 5 min at room temperature, which greatly shortens the time required for the entire synthetic procedure of the alloy nanoparticles. In Figure 2a,b, the XRD patterns of the Au(I)ddt and Ag(I)ddt show typical characteristics of layered structures of both inorganicorganic hybrids. Interestingly, on the basis of diffraction peak intensities and widths, the ordering in these samples is almost identical, despite a great difference in reaction time (as short as 5 min versus as long as 24 h). The pattern recorded for the Au(I)ddt is similar to what had been reported earlier in the literature.41 The periodic peaks appearing at the small 2θ angle region are associated with the layered structure, and the interlayer distance in this solid determined by the XRD technique is d001 = 35.6 ( 0.1 Å. Similarly, the layered hybrid Ag(I)ddt also shows pronounced periodic diffraction peaks, and its interlayer distance determined by the XRD technique is d001 = 34.5 ( 0.1 Å, in good agreement with literature data.35,36 Using the Scherrer equation, the average thicknesses (along the [001] direction) of these layered hybrids are about 13.0 and 17.3 nm for Au(I)ddt and Ag(I)ddt, respectively. The UVvisible absorbance spectra of the two M(I)alkanethiolates prepared in 5 min are shown in Figure 2c. For the Au(I)ddt sample, it is observed that the absorption peak due to the ligand-to-metal charge transfer at 320 nm in the AuCl4 complex ion is no longer present,42,43 indicating that all the initial Au(III) has been transformed into monovalent

oxidation state Au(I). Instead, two broad peaks appear at 347 and 387 nm of the spectrum, which are fingerprint absorptions of the Au(I)ddt;32 the peak at 347 nm could be attributed to “ligand to Au” charge transfer while that at 387 nm to “ligand to Au 3 3 3 Au” charge transfer, considering the aurophilicity effect of Au(I) in Au(I)ddt.35 As for the solid Ag(I)ddt, a large absorption peak appears at 340 nm (characteristic absorption of two-coordinate Ag(I) alkanethiolate compounds due to “ligand to Ag” charge transfer), similar to what had been reported previously.35 These absorption peaks will be further discussed, as they could be used as an indication of the extent of reaction in our later AuAg alloy nanoparticle synthesis. 3.2. Preparation of AuAg Alloy Nanoparticles. It has been well-established that bimetallic alloy nanoparticles can be differentiated from their coreshell counterparts by characterizing their optical properties. For example, a mixture of separate Au and Ag nanoparticles would give rise to two different absorption peaks [e.g., due to different surface plasmon resonances (SPR)], whereas the AuAg alloy nanoparticles only show a single absorption band in their UVvisible spectra, which red-shifts when the percentage of gold increases.911,16,21,44 Even for the AuAg alloy particles with a size smaller than 2 nm, a single absorption peak was obtained at a λmax value intermediate of that for pure Ag or Au nanoparticles.22 On the other hand, Au@Ag coreshell particles with the size smaller than 3 nm readily exhibit two absorption peaks, one for the Au cores and the other for Ag shells. Only when there is a large excess of Ag shells will the SPR peaks of Au cores disappear such that only a single peak at the position of Ag absorption is present. For larger Au@Ag coreshell nanoparticles, two SPR peaks are normally observed near the energies of the absorption bands of their respective parent metals (Au and Ag).14,2426 In our synthesis, a series of color changes in the suspensions of AuAg alloy nanoparticles was observed, which ranges from orange-brown with a higher percentage of silver to orange-red 5636

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

ARTICLE

Figure 5. HRTEM images of the as-prepared AuAg alloy nanoparticles (scheme c, Figure 1): (a) AuAg(0.3), (b) AuAg(0.5), and (c) AuAg(0.7).

Figure 4. TEM images of the as-prepared Ag (scheme a, Figure 1) and AuAg alloy nanoparticles (scheme c, Figure 1): (a, b) Ag, (c, d) AuAg(0.3), (e, f) AuAg(0.5), and (g, h) AuAg(0.7). TEM images of Au nanoparticles (schemes a and b, Figure 1) are displayed in Supporting Information SI-2.

with a higher content of gold. The corresponding UVvisible absorbance spectra of the end product suspensions in toluene are reported in Figure 3a. In particular, we do not observe the double-hump feature (refer to Figure 2c) or surface plasmon resonance bands at the position of their parent metals (Au and Ag), but we only observe one singlet peak between those of pure Ag and Au nanoparticles. It is known that the width of the SPR band increases with decreasing particle size in the intrinsic size region (mean diameter smaller than 25 nm).44 In the present case, nevertheless, the Ag sample exhibits much broader bandwidth than the other alloys, which indicates that the SPR bandwidth depends primarily on sample composition in addition to the size effect, noting that the SPR bandwidth of Ag nanoparticles is twice that of the AuAg(0.5), but the particle size of Ag is only 37% smaller than that of the AuAg(0.5). The observed singlet peaks also red-shifts with increasing content of gold, strongly suggesting that AuAg alloys are formed. In Figure 3b, a correlation plot of the wavelength of maximum absorbance with the Au mole fraction in the precursor feed reveals a near-linear

trend, consistent with previous reports for the AuAg alloy nanoparticles.10,11 The absence of the featuring absorption peaks (see Figure 2c) of the M(I)alkanethiolate precursors in these UVvisible absorbance spectra confirms once again the complete reduction of the two organicinorganic hybrids. The TEM images of the nanoparticles synthesized in this work are presented in Figure 4. This TEM investigation reveals that high monodispersity can be facilely achieved for nanoparticle products of pure Ag and all the alloys using the present synthetic protocol (schemes a and c, Figure 1). A measurement with a sampling size of 30 random particles of each sample shows a gradually increasing diameter of the nanoparticles as the Au content increases, owing to the ease of Au(I)ddt reduction. The particle sizes measured in the samples of Figure 4 are 3.5 ( 0.7 nm (Ag), 3.7 ( 0.4 nm (AuAg(0.3)), 4.8 ( 0.5 nm (AuAg(0.5)), and 5.6 ( 0.7 nm (AuAg(0.7)), respectively; also refer to Supporting Information SI-1. Apparently, in this series of samples, the particle size increases monotonically with the amount of initial Au(I)ddt used in synthesis. Therefore, the composition and particle size of AuAg alloys can be controlled facilely with the amount of initial reactants. For the pure Au nanoparticles, high monodispersity has not been achieved (scheme b, Figure 1; section 2.7; and Supporting Information SI-2, which will be further discussed in the final part of this paper). In fact, we have also tried the half-seeding method to synthesize pure Au nanoparticles, similar to that used in the synthesis of pure Ag nanoparticles (scheme a, Figure 1; section 2.6). With this half-seeding method, the monodispersity of Au nanoparticles was improved to some degree (Supporting Information SI-2). However, the Au nanoparticles formed agglomerated and settled to the bottom of the container, disallowing further characterization by UVvisible absorption spectrometry. 5637

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

ARTICLE

Figure 6. (a, b) TEM images of the AuAg(0.5) alloy nanoparticles synthesized by scheme d of Figure 1, (c) normalized UVvis spectra of the AuAg(0.5) nanoparticles formed by two different orders of addition of the metalalkanethiolates according to scheme c (black curve) and d (red curve) of Figure 1, and (d) FTIR spectra of the as-prepared Ag nanoparticles (scheme a, Figure 1) and AuAg(0.5) alloy nanoparticles (scheme c, Figure 1); transmittance of FTIR spectra is in arbitrary units.

Compositions of AuAg alloy nanoparticles determined by ICP-MS technique are essentially the same as designated Au: Ag molar ratios in input precursors (Supporting Information SI-3). The HRTEM images of three types of representative AuAg alloy nanoparticles are displayed in Figure 5. The asprepared alloy nanoparticles are pseudo (polyhedral) spheres. Many of the particles exhibit a 5-fold symmetry (multitwinned structure with the {111} facets). A common interplanar distance of 0.24 nm, corresponding to the lattice fringe of (111) planes of gold or silver, can be clearly resolved in these images. No obvious image contrast or phase boundary relatable to coreshell-like structures among these nanoparticles was observed, affirming that these polyhedral products are not of the coreshell type. For the AuAg(0.5) sample, furthermore, we have also tried reversing the order of adding the M(I) alkanethiolate precursors (scheme d, Figure 1), that is, by heating the Ag(I)ddt in oleylamine first, followed by adding the ethanolic suspension of Au(I)ddt immediately upon detecting a color change (refer to section 2.5). The AuAg alloy nanoparticles resulting from this process are also highly monodisperse, as reported in Figure 6a,b. The particle size of this sample is 3.9 ( 0.7 nm, noting that that synthesized via the normal procedure (scheme c, Figure 1 and Figure 4e,f) is 4.8 ( 0.5 nm. The size variation observed here can be attributed to different starting seed clusters nucleated from different precursor solids, noting that the Au(I)ddt is relatively easy to be reduced under our current reaction conditions. In Figure 6c, two SPR bands in the UVvis absorbance spectra measured from the two samples have almost the same bandwidth, despite being of different particle sizes. Consistent with the results shown in Figure 3a, we believe that indeed the composition of the particles plays a more dominant role in influencing the bandwidth of the SPR band, as compared to the particle size. The particle sizes of the two samples are about 23% different, and this difference might be insignificant in causing a change in the bandwidth.

Figure 7. (a) Optical limiting response, (b) output fluence, and (c) nonlinear scattering results of the AuAg alloy nanoparticles, Au nanoparticles (scheme b), Ag nanoparticles (refer to the samples of Figures 3 and 4), and toluene studied by 532 nm laser.

In Figure 6d, FTIR spectra of the as-synthesized metal nanoparticles show the absence of NH stretching absorption at 33003500 cm1, NH bending at ∼1583 cm1, and dCH bending at ∼1070 cm1 of oleylamine molecules, indicating that the oleylamine does not function as the surface-capping agent in the synthesis. However, the presence of IR absorption peaks which are associated with CH vibrational modes,45,46 reveals the presence of hydrocarbon surfactants. This is rather expected, since the sulfur of dodecanethiolate has a stronger binding affinity to Au and Ag, compared to nitrogen of the oleylamine, although the latter was in great excess. It has been proposed that imines and nitriles could be generated from the reaction of silver ions and oleylamine.47 These resultant species could also passivate the surface of the metal particles through the lone pair of electrons on nitrogen or the π electrons of the double or triple bond. We however do not observe the vibrations assignable to the CdN or CtN bonds in their characteristic regions. For neat dodecanethiol, the CH2 stretching bands appear at 2853 and 2924 cm1.48 In the AuAg(0.5) alloy nanoparticles, the respective peaks appear at 2849 and 2919 cm1. The observed shift to lower wavenumber values is indicative of an ordered all-trans zigzag conformation of the alkyl chains on the nanoparticle surface.4850 For the Ag nanoparticles, however, the two peaks appear at 2853 and 2922 cm1, closer to those of the neat dodecanethiol. Understandably, this IR observation implies that the interaction between the alkyl chains and the pure Ag nanoparticles is less intense, and the alkyl chains on the silver surface are less ordered. Such a phenomenon can be attributed to different morphologies 5638

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

Figure 8. (a) Optical limiting response, (b) output fluence, and (c) nonlinear scattering results of the AuAg alloy nanoparticles, Au nanoparticles (scheme b), Ag nanoparticles (refer to the samples of Figures 3 and 4), and toluene studied by 1064 nm laser.

of the AuAg alloy and pure Ag nanoparticles, where the former takes on a faceted morphology while the latter a spherical shape (due to absence of all-trans zigzag conformation of the alkyl chains).48 Unlike other investigations in which oleylamine functions as both a reducing agent and a capping agent,18,47 the oleylamine in our synthetic protocol serves primarily as a reductant, as evident from the above FTIR results. Instead, the dodecanethiolate, which is adjacent to the Au(I) or Ag(I) cations in the precursor solid of Au(I)ddt or Ag(I)ddt, serves as a surface capping agent. It should be mentioned that there was an abundant supply of this capping agent in our reaction environment, as the mole ratio of the metal cation to dodecanethiolate was 1:1 in either Au(I)ddt or Ag(I)ddt. Finally, it should be mentioned that the presence of the tiny peaks at nearby 2500 cm1, which is probably assignable to SH stretching vibration, is most likely due to the presence of trapped ddt molecules. 3.3. Optical Limiting Properties of AuAg Alloy Nanoparticles. Figure 7a shows the optical limiting response of the nanoparticles measured with laser pulses at 532 nm. The transmittance of all the samples was almost constant at low input fluence and the transmittance decreases at high input fluence. From the figure, it can be seen that all the prepared nanoparticles exhibit the optical limiting effect, with some being better optical limiters than others. The limiting threshold is defined as the incident fluence at which the transmittance falls to 50% of the normalized linear transmittance.51 The observed limiting threshold values are 0.68, 0.48, 0.46, 3.15, and 0.90 J/cm2 for the Ag, AuAg(0.3), AuAg(0.5), AuAg(0.7), and Au nanoparticles,

ARTICLE

respectively. The best optical limiters are the AuAg(0.3) and AuAg(0.5) nanoparticles, and their thresholds are even lower than the optical limiting threshold of carbon nanotubes (1.0 J/cm2), which is a commonly used benchmark optical limiting material.52 This finding is different from that of an earlier report in which AuAg alloys were found to be less efficient optical limiters than their parent pure metals for picosecond laser pulses at a wavelength of 532 nm.31 We have also tested the optical limiting efficiency of toluene and confirmed that this solvent also shows optical limiting effect at high input fluence for the 532 nm laser input, similar to what has been reported.52 Figure 7b further exemplifies the optical limiting effect of these samples, whereby a larger deviation from linearity, which represents low output fluence at high input fluence, is evidence of an optical limiting effect. Scattering measurements of the nanoparticles were also performed to understand the mechanism of the optical limiting effect. Analysis of the results in Figure 7c shows that at low input fluence, the scattering signals are quite linearly related to the input fluence. But at higher input fluence, they deviated from linearity. Apparently, at high input fluence, the nonlinear scattering can cause low transmission of the light, and thus nonlinear scattering is the main reason for the observed optical limiting phenomenon. Similar results were obtained for the same samples studied using the 1064 nm laser source, as shown in Figure 8a,b, with limiting threshold values of 5.7, 7.0, 2.8, 17.7, and 4.9 J/cm2 for the Ag, AuAg(0.3), AuAg(0.5), AuAg(0.7), and Au nanoparticle sample, respectively. These optical limiting threshold values are much better than that of carbon nanotubes with a threshold of 10.0 J/ cm2.52 In Figure 8c, nonlinear scattering of these samples is also found to be the mechanism underlying the optical limiting effect. The nonlinear scattering effect is likely due to the formation of scattering centers owing to the expansion of the nanoparticles, which have been heated and excited by the laser beam. Furthermore, energy transfer to the solvent will result in a formation of microplasmas, which can act as secondary scattering centers and lead to further reduction in the transmittance with increasing input laser fluence.52 This is highly likely in our experiments because toluene is a solvent with low thermal conductivity, which results in the confinement of the heat energy at the particlesolvent interface, enhancing the evaporation of the solvent into bubbles.52 It had been previously proposed that the excited state absorption was responsible for the observed optical limiting performance.31 Recently, we have experimentally observed that strong nonlinear scattering was responsible for the observed broadband optical limiting properties of oleylamine-capped gold nanoparticles.52 Our results show that, for the alloy samples we synthesized, the AuAg(0.3) and AuAg(0.5) nanoparticles give rise to superior optical limiting properties that are greater than that of the pure metals. The AuAg(0.7) alloy sample is, however, not as effective as the rest of the samples in terms of the optical limiting effect. Upon excitation of the surface plasmon by the laser, nanoparticles are known to get heated up and fragmented.53 However, we did not notice any significant change in the UVvisible spectra or the TEM images of the AuAg(0.7) nanoparticles after laser irradiation, ruling out the possibility of fragmentation of the particles due to their instability. Furthermore, it has been recently observed that interband excitation of Au nanoparticles leads to a greater heating efficiency compared to the intraband transition of conduction electrons, that is, the heating efficiency is dependent on the excitation wavelength.54 This new finding is in contrast to the common belief that excitation of the surface plasmon 5639

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir resonance is the most efficient way of heating. In the present study, we used nanoparticles having different compositions and SPRs, and hence we expect these particles to have varying band structures. When a single laser wavelength of 532 or 1064 nm was used for photoexcitation, there will be different degrees of interand intraband transitions in the particles (e.g., in the case of gold nanoparticles, intraband transition is dominant at the excitation wavelength of 532 nm54). As such, the heating efficiency of the particles will be different. The subsequent process of electron relaxation and formation of primary and secondary scattering centers will differ, leading to a varied optical limiting effect. This is because when the hot electron relaxes by electronphonon coupling, energy is transferred to the lattice, causing it to expand and function as the primary scattering center. On the other hand, phononphonon coupling also occurs whereby energy exchange takes place between the lattice and the solvent, forming the secondary scattering centers. These scattering centers are essential for the observed optical limiting phenomenon. Since the nanoparticles of AuAg(0.5) exhibit the best optical limiting properties, the primary and secondary scattering centers that originated from the particles should be the most efficient. The reverse can be said of the nanoparticles of AuAg(0.7). We believe that the difference in the optical limiting performance is therefore related to the difference in the electron thermalization of the particles, though at the present stage the explicit details of why the AuAg(0.7) nanoparticles give rise to the weakest thermal efficiency is still unclear. In an earlier report, monometallic clusters were found to be better optical limiters than alloy clusters.31 It was mentioned that electron ejection and photofragmentation occurred more readily in the alloy clusters. Due to the different procedures used in our preparation of alloy nanoparticles from the reported work,31 the density of surfactants on the particles should also differ. Since these surfactants can affect the stability and optical limiting performance of nanoparticles, the optical limiting properties will depend on the preparation procedures as well. This explains why our observations are different. However, it had been observed that among the alloy nanoclusters, those with a higher Ag content are better optical limiters,31 which our results agree with. The above experiments show that the resultant optical properties of the alloy nanoparticles can be tuned by the composition of gold and silver, which can be readily selected in any mole ratio of Au:Ag via the synthetic protocol we have devised herein. 3.4. Further Explorations of the Synthetic Method. In Figure 6a,b, we have shown that the order of adding M(I)alkanethiolates is not so crucial to the synthesis, as high monodispersity in the alloy nanoparticles was still achieved whether the Ag(I)ddt was added to the Au in oleylamine or vice versa (i.e., scheme c or d, Figure 1). To further examine these synthetic parameters, we have also investigated the simultaneous heating of both M(I)alkanethiolates in the oleylamine solvent, as depicted in scheme e of Figure 1. It is found that the AuAg alloy nanoparticles could still be obtained, but the nanoparticles formed are not so monodisperse. TEM images of Figure 9a,b show a product of AuAg(0.5) nanoparticles prepared with scheme e in Figure 1. Although both the M(I)alkanethiolates were heated from room temperature to 200 °C (N.B.: We have tried the same experiment at 180 °C, but the product particles are polydisperse.), the rate of heating and reduction of each sample are different, which could result in additional nucleation during the growth. For example, newly formed metal atoms of the M(I)alkanethiolate, which was reduced at a slower pace, might

ARTICLE

Figure 9. (a and b) TEM images of the AuAg(0.5) alloy nanoparticles synthesized according to scheme e of Figure 1, (c) normalized UVvis spectra of the AuAg(0.5) nanoparticles formed by scheme c (black curve) and scheme e (red curve) of Figure 1, and (d) TEM image of the AuAg(0.5) alloy nanoparticles synthesized according to scheme c of Figure 1, but Ag(I)ddt was added to the suspension with a 1 min delay upon detecting a color change.

not grow on the nanoparticle seeds that had already formed, resulting in separate nucleation and hence the observed lower monodispersity. Ultimately, though, the diffusion and intermixing of the metal atoms at the high temperature employed would still lead to alloy nanoparticles being formed. From the UV visible spectra shown in Figure 9c, the presence of a single peak confirms the formation of an AuAg alloy. Interestingly, the peak is blue-shifted relative to the other, and this suggests different degrees of segregation and mixing in the nanoparticles prepared via the two different routes,55 although the composition should be the same. Indeed, the particle size of the simultaneous heating method is 3.4 ( 1.1 nm (Figure 9a,b), suggesting that monodispersity is not quite achieved with this scheme. Hence, we believe that the monodispersity achieved in our experiments can be attributed to the staggered addition of the second metal alkanethiolate. One reason that staggering the addition of the M(I)alkanethiolate could lead to higher monodispersity is because when the color of the reaction mixture just changed, it indicates that part of the first metal element has been clustered into “half-seeds” and they are just about to grow. At this stage, these premature seed clusters are still very small in size but possess high surface energy, which the system seeks to minimize, as depicted in Figure 10. Once the second M(I)alkanethiolate is added, it can be reduced readily due to the high temperature of the oleylamine. Moreover, together with the remaining first metal precursor, the second M(I)alkanethiolate would be reduced and grown on the existing small clusters of the first metal in order to minimize the energy of the reaction system, which prevents further nucleation. The metal atoms formed after this half-seeding process can then diffuse and mix to form an alloy. However, the timing of addition must be precisely controlled. On the basis of the present work, the optimum point of addition of the second M(I)alkanethiolate can be started when 5640

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

Figure 10. A schematic illustration of the half-seeding process developed in this work using scheme c of Figure 1 as an example. Small yellow spheres in the Au(I)ddt structure represent Au(I) ions and small green spheres in the Ag(I)ddt represent Ag(I) ions in the supramolecular solids.

the reaction mixture just changes its color. If the wait is too long, enlargement and agglomeration of the metal particles could result. Shown in Figure 9d, the AuAg(0.5) nanoparticles were formed by adding Ag(I)ddt to the reacting suspension only 1 min after a color change was detected. The particles are agglomerated together rather than being discrete. They are no longer dispersible in toluene, settling to the bottom of the container after some time, hence suggesting that they are in the agglomerated state with minimal ddt derivatization. The average size of these particles is 9.4 ( 1.4 nm, which is substantially larger than the ones prepared without any delay [4.8 ( 0.5 nm (AuAg(0.5)), Figure 4]. Due to the similar lattice constants of gold and silver (2.36 Å for Ag and 2.35 Å for Au),15 XRD could not be used effectively to differentiate the alloys from the pure metals, as their diffraction peaks would appear at the same positions. In fact, it has been reported that there has been no obvious contrast change or diffraction peak shift in the HRTEM images or XRD patterns upon changing from coreshell Ag@Au particles to the well mixed AuAg alloy nanoparticles.20 However, because of this similarity in lattice constants, the two metal elements are highly miscible in all proportions, facilitating the formation of their alloys. Of course, this miscibility is also aided by the similar atomic sizes of Au and Ag. At the elevated reaction temperatures, the M(I)alkanethiolates can partially decompose and oleylamine, being a solvent and a reducing agent, can bring about the reduction of the Au(I) and Ag(I) to their metallic forms. Due to the high-temperature environment, the resultant gold and silver atoms acquire sufficient thermal energy, interdiffusing into alloys under a set of reaction conditions. The diffusivity (D) of an element is dependent on temperature (T), and if the dependence is a BoltzmannArrhenius type, it can be expressed by the following equation56   ΔHd D ¼ D0 exp  kT where D0 is the pre-exponential factor of the element, ΔHd is the activation enthalpy of diffusion, and k is the Boltzmann constant. Apparently, diffusivity of the studied metal elements increases

ARTICLE

with temperature. It has been recently illustrated that coreshell nanoparticles of Ag@Au are prepared at 50 °C, but AuAg alloy nanoparticles can be easily obtained at 100 °C.20 Hence, we believe that at 200 °C, the temperature employed in our synthesis, AuAg alloy nanoparticles are definitely formed instead of coreshell particles. Furthermore, it has been previously reported that formation of AuAg alloy from Au@Ag coreshell particles is due to the presence of vacancy defects at the interface, which enhances the mixing of the metal atoms.57 The presence of a stabilizer at the interface could have caused the defects. This may also be applicable to our case, whereby the presence of defects at the surface of the newly formed seed particles aids the depositing and mixing of the second metal. Despite highly monodisperse pure Ag nanoparticles being formed via this method, such a quality characteristic has not been achieved for pure Au nanoparticles under the similar conditions (Supporting Information SI-2), which is somehow contrary to the previous report where larger monodisperse Au nanoparticles could be obtained by heating the Au(I)alkanethiolate nanotubes in hexadecylamine [CH3(CH2)15NH2] in that work.32 Because the oleylamine [CH3(CH2)7CHdCH(CH2)8NH2] used in this work has a stronger reducing power (due to the presence of a double bond) for Au(I) ions in comparison with hexadecylamine, the formation rate of Au0 may be a bit too fast to produce monodisperse Au nanoparticles. Therefore, the composition and structure of amines used in the reduction of the Au(I)alkanethiolates to Au0 are also crucial to the formation of final nanoparticles. A systematic investigation on the reducing ability of different kinds of amines will be carried out in future synthesis of AuAg alloy nanoparticles using this synthetic approach.

4. CONCLUSION In summary, we have developed a fast and facile approach for preparation of highly monodisperse AuAg alloy nanoparticles using supramolecular solids M(I)alkanethiolates (M = Au and Ag) as metal sources. These starting solids can be prepared easily by mixing an excess of alkanethiols with their respective metal precursors (e.g., HAuCl4 or AgNO3) at room temperature. In this half-seeding approach, the first and second M(I) alkanethiolates are added sequentially to oleylamine solvent at different reaction stages. Because of the time difference, the originally formed metal clusters can then act as premature seeds into which both the first and the second metals (from M(I) alkanethiolates, M = Au and Ag) can be continuously reduced and integrated without introducing unwanted spontaneous nucleation. Interestingly, after the metal ions in the precursors have been reduced, the detached alkanethiolates can further function as a capping surfactant on the metal products, while the oleylamine used in the synthesis acts as a reducing agent for the transformation of M(I). Unlike in other prevailing methods, both the Au and Ag elements in our solid precursors are in the same monovalent state and have an identical supramolecular structure, which may lead to a more homogeneous reduction and complete interdiffusion at elevated reaction temperatures. The resultant AuAg alloy nanoparticles are mainly faceted with {111} crystal planes with diameters in the range of 36 nm. Under identical reaction conditions, the average diameter of nanoparticles gradually increases with the rise of gold content in the feed precursors, owing to the relative ease of reduction of Au(I)alkanethiolates. The selection of a suitable solvent for synthesis is also crucial. In this regard, the oleylamine solvent 5641

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir employed in this work has been proven to be very efficient for formation of AuAg alloy nanoparticles. The evidence of alloys being formed is obtained by transmission electron microscopic and UVvisible spectroscopic techniques, where we only observe a single metal phase and a single surface plasma resonance absorption band that red-shifts as the mole ratio of Au:Ag increases. The thus-obtained alloy nanoparticles have optical properties that can be conveniently tuned, as a linear relationship is found for the wavelength of surface plasmon resonance peak and the Au content in the feed. These alloy nanoparticles have also exhibited excellent broadband optical limiting properties that are much better than those of carbon nanotubes (a commonly used benchmark optical limiter). Furthermore, our scattering measurements suggest that nonlinear scattering plays an important role in the observed optical limiting performances. Therefore, the AuAg alloy nanoparticles fabricated with this simple protocol have the potential to be used as broadband optical limiting materials.

’ ASSOCIATED CONTENT

bS

Supporting Information. Particle histograms, TEM images, and ICP-MS results of some studied nanoparticles (PDF). This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT The authors gratefully acknowledge the Economic Development Board, Singapore, and King Abdullah University of Science and Technology, Saudi Arabia, for support of this research. T.T.C. would also like to thank the National University of Singapore for providing her a postgraduate scholarship. ’ REFERENCES (1) Liu, X. Y.; Wang, A. Q.; Yang, X. F.; Zhang, T.; Mou, C. Y.; Su, D. S.; Li, J. Chem. Mater. 2009, 21, 410–418. (2) Liu, X. Y.; Wang, A. Q.; Wang, X. D.; Mou, C. Y.; Zhang, T. Chem. Commun. 2008, 3187–3189. (3) Yen, C. W.; Lin, M. L.; Wang, A. Q.; Chen, S. A.; Chen, J. M.; Mou, C. Y. J. Phys. Chem. C 2009, 113, 17831–17839. (4) Chiment~ao, R. J.; Medina, F.; Fierro, J. L. G.; Llorca, J.; Sueiras, J. E.; Cesteros, Y.; Salagre, P. J. Mol. Catal. A: Chem. 2007, 274, 159–168. (5) Lou, Y. B.; Maye, M. M.; Han, L.; Luo, J.; Zhong, C. J. Chem. Commun. 2001, 473–474. (6) Yuan, Q.; Zhou, Z. Y.; Zhuang, J.; Wang, X. Chem. Commun. 2010, 46, 1491–1493. (7) Nair, A. S.; Suryanarayanan, V.; Pradeep, T.; Thomas, J.; Anija, M.; Philip, R. Mater. Sci. Eng., B 2005, 117, 173–182. (8) Cao, Y. W.; Jin, R.; Mirkin, C. A. J. Am. Chem. Soc. 2001, 123, 7961–7962. (9) Chen, D. H.; Chen, C. J. J. Mater. Chem. 2002, 12, 1557–1562. (10) Sun, Y. G.; Xia, Y. N. Analyst 2003, 128, 686–691. (11) Link, S.; Wang, Z. L.; El-Sayed, M. A. J. Phys. Chem. B 1999, 103, 3529–3533. (12) Mallin, M. P.; Murphy, C. J. Nano Lett. 2002, 2, 1235–1237. (13) Rodríguez-Gonzalez, B.; Sanchez-Iglesias, A.; Giersig, M.; Liz-Marzan, L. M. Faraday Discuss. 2004, 125, 133–144.

ARTICLE

(14) Gonzalez, C. M.; Liu, Y.; Scaiano, J. C. J. Phys. Chem. C 2009, 113, 11861–11867. (15) Lee, I.; Han, S. W.; Kim, K. Chem. Commun. 2001, 1782–1783. (16) Zhang, Q. B.; Xie, J. P.; Liang, J.; Lee, J. Y. Adv. Funct. Mater. 2009, 19, 1387–1398. (17) Zhang, Q. B.; Lee, J. Y.; Yang, J.; Boothroyd, C.; Zhang, J. X. Nanotechnology 2007, 18, 245605. (18) Wang, C.; Yin, H. G.; Chan, R.; Peng, S.; Dai, S.; Sun, S. H. Chem. Mater. 2009, 21, 433–435. (19) Yang, Y.; Gong, X. Z.; Zeng, H. M.; Zhang, L. J.; Zhang, X. H.; Zou, C.; Huang, S. M. J. Phys. Chem. C 2010, 114, 256–264. (20) Wang, C.; Peng, S.; Chan, R.; Sun, S. H. Small 2009, 5, 567–570. (21) Kariuki, N. N.; Luo, J.; Maye, M. M.; Hassan, S. A.; Menard, T.; Naslund, H. R.; Lin, Y. H.; Wang, C. M.; Engelhard, M. H.; Zhong, C. J. Langmuir 2004, 20, 11240–11246. (22) Wilson, O. M.; Scott, R. W. J.; Garcia-Martinez, J. C.; Crooks, R. M. J. Am. Chem. Soc. 2005, 127, 1015–1024. (23) Mandal, S.; Selvakannan, P. R.; Pasricha, R.; Sastry, M. J. Am. Chem. Soc. 2003, 125, 8440–8441. (24) Mallik, K.; Mandal, M.; Pradhan, N.; Pal, T. Nano Lett. 2001, 1, 319–322. (25) Wang, X.; Zhang, Z. Y.; Hartland, G. V. J. Phys. Chem. B 2005, 109, 20324–20330. (26) Kim, Y. J.; Johnson, R. C.; Li, J. G.; Hupp, J. T.; Schatz, G. C. Chem. Phys. Lett. 2002, 352, 421–428. (27) Ispasoiu, R. G.; Balogh, L.; Varnavski, O. P.; Tomalia, D. A.; Goodson, T. J. Am. Chem. Soc. 2000, 122, 11005–11006. (28) Sun, W. F.; Dai, Q.; Worden, J. G.; Huo, Q. J. Phys. Chem. B 2005, 109, 20854–20857. (29) Wang, G.; Sun, W. F. J. Phys. Chem. B 2006, 110, 20901–20905. (30) Francois, L.; Mostafavi, M.; Belloni, J.; Delouis, J. F.; Delaire, J.; Feneyrou, P. J. Phys. Chem. B 2000, 104, 6133–6137. (31) Philip, R.; Kumar, G. R.; Sandhyarani, N.; Pradeep, T. Phys. Rev. B 2000, 62, 13160–13166. (32) Zhang, Y. X.; Zeng, H. C. Adv. Mater. 2009, 21, 4962–4965. (33) Cha, S. H.; Kim, J. U.; Kim, K. H.; Lee, J. C. Chem. Mater. 2007, 19, 6297–6303. (34) Bardeau, J. F.; Parikh, A. N.; Beers, J. D.; Swanson, B. I. J. Phys. Chem. B 2000, 104, 627–635. (35) (a) Bensebaa, F.; Ellis, T. H.; Kruus, E.; Voicu, R.; Zhou, Y. Langmuir 1998, 14, 6579–6587. (b) Schmidbaur, H. Gold Bull. 1990, 23, 11–21. (c) Li, C. K.; Lu, X. X.; Wong, K. M. C.; Chan, C. L.; Zhu, N. Y.; Yam, V. W. W. Inorg. Chem. 2004, 43, 7421–7430. (d) Fijolek, H. G.; GonzalezDuarte, P.; Park, S. H.; Suib, S. L.; Natan, M. J. Inorg. Chem. 1997, 36, 5299–5305. (36) Levchenko, A. A.; Yee, C. K.; Parikh, A. N.; Navrotsky, A. Chem. Mater. 2005, 17, 5428–5438. (37) Schaaff, T. G.; Shafigullin, M. N.; Khoury, J. T.; Vezmar, I.; Whetten, R. L. J. Phys. Chem. B 2001, 105, 8785–8796. (38) Jana, N. R.; Gearheart, L.; Murphy, C. J. Langmuir 2001, 17, 6782–6786. (39) Brown, K. R.; Natan, M. J. Langmuir 1998, 14, 726–728. (40) Henglein, A. Langmuir 1999, 15, 6738–6744. (41) Cha, S.-H.; Kim, K.-H.; Kim, J.-U.; Lee, W.-K.; Lee, J.-C. J. Phys. Chem. C 2008, 112, 13862–13868. (42) Harada, M.; Einaga, H. Langmuir 2007, 23, 6536–6543. (43) Tanahashi, I.; Mitsuyu, T. J. Non-Cryst. Solids 1995, 181 77–82. (44) (a) Hostetler, M. J.; Zhong, C. J.; Yen, B. K. H.; Anderegg, J.; Gross, S. M.; Evans, N. D.; Porter, M.; Murray, R. W. J. Am. Chem. Soc. 1998, 120, 9396–9397. (b) Daniel, M. C.; Astruc, D. Chem. Rev. 2004, 104, 293–346. (45) Zhang, Y. X.; Zeng, H. C. J. Phys. Chem. B 2006, 110, 16812–16815. (46) Zhang, Y. X.; Zeng, H. C. Langmuir 2008, 24, 3740–3746. (47) Chen, M.; Feng, Y. G.; Wang, X.; Li, T. C.; Zhang, J. Y.; Qian, D. J. Langmuir 2007, 23, 5296–5304. 5642

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643

Langmuir

ARTICLE

(48) Han, S. W.; Kim, Y.; Kim, K. J. Colloid Interface Sci. 1998, 208, 272–278. (49) Ang, T. P.; Chin, W. S. J. Phys. Chem. B 2005, 109, 22228–22236. (50) Sandhyarani, N.; Antony, M. P.; Selvam, G. P.; Pradeep, T. J. Chem. Phys. 2000, 113, 9794–9803. (51) Chen, P.; Wu, X.; Sun, X.; Lin, J.; Ji, W.; Tan, K. L. Phys. Rev. Lett. 1999, 82, 2548–2551. (52) Polavarapu, L.; Venkatram, N.; Ji, W.; Xu, Q. H. ACS Appl. Mater. Interfaces 2009, 1, 2298–2303. (53) Mafune, F.; Kohno, J.; Takeda, Y.; Kondow, T. J. Phys. Chem. B 2001, 105, 9050–9056. (54) Werner, D.; Hashimoto, S.; Uwada, T. Langmuir 2010, 26, 9956–9963. (55) Ferrando, R.; Jellinek, J.; Johnston, R. L. Chem. Rev. 2008, 108, 845–910. (56) Tu, K. N.; Mayer, J. W.; Feldman, L. C. in Electronic Thin Film Science; Macmillan Publishing Co.: New York, 1992. (57) Shibata, T.; Bunker, B. A.; Zhang, Z. Y.; Meisel, D.; Vardeman, C. F.; Gezelter, J. D. J. Am. Chem. Soc. 2002, 124, 11989–11996.

5643

dx.doi.org/10.1021/la2005216 |Langmuir 2011, 27, 5633–5643