Seed-Mediated Growth of Ag@Au Nanodisks with Improved Chemical Stability and Surface-Enhanced Raman Scattering.

Bimetallic Ag@Au nanoparticles (NPs) have received significant research interest because of their unique optical properties and molecular sensing ability through surface-enhanced Raman scattering (SERS). However, the synthesis of Ag@Au core-shell plasmonic nanostructures with precisely controlled size and shape remained a great challenge. Here, we report a simple approach for the synthesis of bimetallic Ag@Au nanodisks of about 13.5 nm thickness and different diameters through a seed-mediated growth process. The synthesis involves the conformal deposition of Au atoms at the corner sites of Ag nanoplate (AgNPL) seeds coupled with site-selective oxidative etching of AgNPL edges to generate Ag@Au nanodisks. The resultant Ag@Au nanodisks manifest significantly improved chemical stability and tunable localized surface plasmon resonance from the visible to the near-infrared spectral range. Moreover, in comparison to AgNPLs, the Ag@Au nanodisks showed greatly enhanced SERS performance with an enhancement factor up to 0.47 × 105, which is nearly 3-fold higher than that of the original AgNPLs (0.18 × 105). Furthermore, the Ag@Au nanodisks show a high sensitivity for detecting probe molecules such as crystal violet of concentration as low as 10-9 M and excellent reproducibility, with the SERS intensity fluctuation less than 12.5%. The synthesis route adapted for the controlled fabrication of Ag@Au nanodisks can be a potential platform for maneuvering other bimetallic plasmonic nanostructures useful for plasmonics and sensing applications.

Introduction

Noble metal nanostructures have generated a huge research interest owing to their unique tunable localized surface plasmon resonance (LSPR) properties,[1] which render their widespread application in various fields such as photocatalysis,[2] plasmonics,[3,4] surface-enhanced Raman scattering (SERS),[5] biosensing,[6] and biomedicine.[7] Since its discovery, SERS has been regarded as a fascinating analytical technique for detecting analyte molecules with ultrahigh sensitivity.[8−10] Noble metal nanoparticles (NPs) such as Ag[3,11,12] and Au[13−16] with various shapes have been widely investigated for SERS applications because of their morphology-dependent high anisotropy and tunable LSPR, which can produce strong electromagnetic (EM) fields localized around the NP surface, resulting in the generation of near-field enhancement effects.[17] In particular, triangular Ag nanoplates (NPLs) with relatively larger {111} facets have been found to exhibit a strong coupling of LSPR with the analyte molecules because of the strong EM field accumulation at their edges,[18] manifesting a highly enhanced SERS signal for a variety of analyte molecules.[16] However, the Ag NPLs are thermodynamically unstable, and the poor chemical stability toward oxidation in physiological conditions is the principal limitation for their use in practical applications.[19] On the other hand, Ag@Au core–shell-type bimetallic NPs have received considerable attention because of their improved chemical stability and the possibility of tuning the LSPR peak positions from the visible to the near-infrared (NIR) region by tailoring their structure and composition.[20−26] Intensive research efforts have been made on controlling the LSPR properties and the enhancement of chemical stability of such core–shell-type bimetallic nanostructures.[11,27−30] Specifically, the controlled growth of an ultrathin Au shell around Ag seeds to obtain bimetallic Ag@Au core–shell structures is largely considered to be a promising strategy to tune the LSPR and enhance the chemical stability of the plasmonic nanostructure.[27,29] As an example, Yang et al.[28] showed that the conformal deposition of an Au shell of a few atomic-layer thickness (6 atomic layers of Au) over Ag nanocubes can significantly improve their chemical stability and SERS activity.

Recently, seed-mediated solution-phase growth process has become the most attractive technique for the preparation of nanocrystals of bimetallic/multimetallic systems with well-controlled size and shape.[31,32] The controlled synthesis of Ag@Au NPs of diverse shapes with a finely tuned LSPR peak position from the visible to the IR region has been extensively reported in the literature.[29,30,32−34] However, the fabrication of bimetallic Ag@Au core–shell NPs with spatial control over the shell thicknesses remained a great challenge. This challenge primarily arises because of the inherent galvanic exchange reaction that occurred between the Ag NP seeds and Au3+, making the uniform deposition of Au over Ag NPs extremely difficult, which usually leads to the formation of hollow nanoframes.[35−37] Considerable efforts have been made to overcome this obstacle by preventing the galvanic reaction, achieving an Au shell of varying thickness onto the Ag NP seeds.[38−40] Yang et al.[28] developed a strategy for the synthesis of Ag@Au core–shell nanocubes with 3–6 atomic layers of Au-shell thickness by the parallel reduction of Au3+ atoms onto the cubic Ag seeds using ascorbic acid (AA), where the reduction of Au3+ ion dominated by AA resulted in the suppression of the galvanic reaction between Ag and Au3+.[28] Recently, Cathcart et al.[30] demonstrated an effective deposition of Au onto the Ag NP seeds assisted by poly(styrenesulfonate) to obtain Ag@Au core–shell NPs with tunable LSPR from 470 to 800 nm, with a greatly improved chemical stability. On the other hand, oxidative etching of metal NPs has been extensively reported as a novel strategy to control the shape of the metal NPs.[41] For instance, Hong et al.[42] described a process of oxidative etching of Au NPLs by the chlorine ions of the Au precursor, capable of producing Au NPs of different shapes. Lin et al.[43] reported a facile synthesis process for the fabrication of uniform spherical Ag NPs via the seed-mediated growth coupled with oxidative etching of Ag nanocubes using Cu2+ ions. Despite all these efforts, the process of controlled synthesis of Ag@Au core–shell nanostructures with desired sizes and shapes remained unclear.

Herein, we present a facile, seed-mediated strategy for the fabrication of Ag@Au nanodisks through a galvanic reaction-free deposition of Au atoms onto Ag NPLs followed by tip-selective oxidative etching reactions. The galvanic etching-free deposition of Au atoms and oxidative etching occur simultaneously after the injection of HAuCl4 at a relatively higher reaction pH, which leads to a shape evolution from AgNPLs to Ag@Au nanodisks. The resultant Ag@Au nanodisks exhibit excellent chemical stability, and their LSPR peak position can be tuned from 656 to 780 nm just by varying their diameter. Furthermore, owing to the unique disk-like structure of the Ag@Au nanodisks and their LSPR position matching with the laser excitation wavelength (785 nm), the nanostructures manifest significantly improved SERS performance for the detection of crystal violet (CV) molecules with a concentration as low as 10–9 M.

Results and Discussion

Synthesis and Characterization of Ag@Au Nanodisks

The triangular Ag NPLs of 20–35 nm edge lengths (Figure S1) were synthesized following the procedure reported previously.[44] The bimetallic Ag@Au core–shell nanodisks were obtained by the deposition of reduced Au atoms onto the triangular Ag NPL seeds and further controlled oxidative dissolutions at the tips using polyvinylpyrrolidone (PVP) as the surfactant and AA as the reducing agent. The pH value of the reaction solution was adjusted to 11.2 by adding NaOH (0.2 M) solution, with a subsequent addition of the HAuCl4 precursor solution at a relatively slow injection rate. Recent studies have shown that the introduction of HAuCl4 solution at a higher pH (pH = 11.2) reduces the reduction potential of gold ions significantly, which enables the conformal deposition of uniform metallic shells over the Ag seeds.[45] Specifically, during the slow titration in the presence of NaOH, after the quick neutralization of HAuCl4 by OH–, the produced [AuCl4]− undergo a ligand exchange process with OH– to generate AuCl(OH)3– and Au(OH)4–, which have a lower reduction potential than the [AuCl4]−.[46] By controlling the degree of the neutralization process, the smoothness of Au shells and the morphology of the resultant Ag@Au NPs could be effectively tailored by avoiding the galvanic reaction.[47] On the other hand, the presence of Cl– ions generated from the dissociation of AuCl4 favors the selective oxidative etching of the edges of the Ag NPLs.[48,49]

The growth of bimetallic Ag@Au nanodisks reported here is distinct from the previous studies as we combine both the galvanic reaction-free deposition of Au and oxidative dissolution reaction of the edges of Ag NPLs to obtain Ag@Au core–shell nanodisks. The suppression of galvanic replacement at room temperature was demonstrated by Yang et al. for Ag@Au nanocubes,[28] and the strategy has been extended by several other groups for the fabrication of other bimetallic systems.[29,50] On the other hand, the oxidative dissolution reaction has been exploited to other NP systems for controlled transformation of their shapes into other morphologies.[12,41,42,51,52] As an example, O’Brien et al.[52] have demonstrated the transformation of hexagonal Au NPLs to uniform Au circular disks through selective oxidative etching in the presence of cetyltrimethylammonium bromide (CTAB). Uniform circular Au nanodisks were obtained because of the strong binding of HAuCl4 with CTAB, which enabled the reaction to proceed with the lowest metal coordination number in a self-limiting fashion, followed by selective oxidation at the tip of the Au NPLs by Au3+ ions. In the present synthesis, as the selective oxidation of the tip edges of triangular Ag NPLs was performed without a strong reducing agent, the galvanic replacement reaction is unavoidable, which usually leads to the formation of bimetallic nanoframes.[36] The key premise of this synthesis is that the slow addition of HAuCl4 at a higher pH (pH = 11.2) by introducing aqueous NaOH solution decreases the reduction potential of [AuCl4]−. Consequently, a conformal deposition of reduced Au atoms occurs onto Ag NPLs by avoiding the galvanic reaction, and further a conproportionation reaction occurs selectively at the corner sites of the Ag NPLs rather than the other faces. By increasing the amount of HAuCl4, the Ag NPLs could be converted into Ag@Au core–shell nanodisks through the oxidative dissolution reaction.

A schematic illustration of the steps involved in the controlled transformation from the triangular Ag NPLs into Ag@Au nanodisks is presented in Figure A. To understand the synthetic steps better, we carried out the transmission electron microscopy (TEM) analysis of the samples prepared at different Au precursor concentrations (Figure B). The TEM results in Figure B(a–d), confirm the reduced Au atoms are selectively bonded at the tips of the Ag NPLs instead of other triangular faces. For Au precursor concentrations higher than the optimum, the reduced Au atoms diffuse to the triangular faces, and oxidative dissolution takes place at the edges of the Ag NPLs to gradually evolve into circular disk-like structures. It is evident that the galvanic replacement-free conformal deposition of Au atoms proceeds selectively at the tips of the Ag NPLs, which serve as initial seeding sites for controlled oxidation at the edges and subsequent diffusion of Au atoms to other faces.[52]

Figure 1

(A) Schematic representation of the formation of Ag@Au nanodisks from Ag NPL seeds as a function of HAuCl4 concentration; (B) corresponding TEM images showing (a) Ag NPLs and (b–d) evolution of Ag@Au nanodisks at different growth stages upon increasing the amount of HAuCl4 confirming the tip-selective oxidative dissolution at the tips of Ag NPLs and conformal deposition.

Figure a–d shows the low- and high-magnification TEM images of the triangular Ag NPL seeds and the obtained bimetallic Ag@Au nanodisks. The formation of triangular Ag NPLs with sharp corners along with a few hexagonal particles is evident in the TEM images presented in Figure a,b. Figure c,d displays the TEM images of Ag@Au nanodisks after their transformation from Ag NPLs. The formation of uniform disk-like structures with an average size of ca. 45 nm is very clear from these micrographs. The distribution of the elements (Au and Ag) in the disk-like final structures can be seen in the energy-dispersive spectrometry (EDS) elemental mapping images presented in Figures e–g. The elemental mapping images clearly demonstrate that Ag (green) is located at the core and Au (red) remains around the Ag core in the Ag@Au nanodisks. Furthermore, we carried out atomic force microscopy (AFM) imaging of the Ag NPLs and Ag@Au nanodisks to estimate their thicknesses. Typical topographic images of Ag NPLs and Ag@Au nanodisks and their height profiles are displayed in Figure . The AFM images confirm the formation of Ag NPLs (Figure a,b) and Ag@Au nanodisks (Figure d,e). The surface height profile plots (Figure c,f) further reveled the smooth noncorrugated surface of the bimetallic nanodisks. The thicknesses of the Ag NPLs and Ag@Au nanodisks were determined to be 10 and 13.5 nm, respectively.

Figure 2

Typical TEM images of (a,b) Ag NPLs and (c,d) Ag@Au nanodisks. (e–g) EDS elemental mapping of Ag@Au nanodisks showing (e) Ag, (f) Au, and (f) overlay of Ag and Au elements.

Figure 3

Typical AFM images of Ag NPLs (a,b) and Ag@Au nanodisks (d,e) supported over Si substrates. (c,f) Corresponding surface height profiles along the yellow dotted lines in (b,e). The increase of height for the Ag@Au nanodisks with respect to bare Ag NPLs indicates the formation of Au layer over them. The flat profile of the Ag@Au nanodisks (f) indicates their smooth/noncorrugated surface.

To study the effect of the reaction solution pH on the galvanic reaction between Au3+ ions and Ag NPLs, we carried out a set of control experiments by changing the reaction conditions. (i) The synthesis was carried out at pH = 4, without introducing NaOH, keeping all other reaction conditions fixed. Surprisingly, the obtained products were bimetallic Ag@Au nanorings with well-defined interior gaps (Figure S2). EDS mapping and the line scanning analysis of the nanorings reveal their alloy (Ag@Au) nature (Figure S2d,e). As the reduction power of AA is low, at this condition, the rate of galvanic reaction becomes high (Rred < Rgal), resulting in the formation of hollow Ag@Au nanorings with well-defined pore diameter.[35] (ii) We increased the concentration of the HAuCl4 precursor solution (2-fold higher than the concentration used for the fabrication of nanodisks), resulting in the formation of well-defined flower-like porous Ag@Au nanostructures (Figure S3). The formation of flower-like structures occurred because of the partial galvanic reaction between the Ag NPL seeds and Au3+ as well as the overgrowth of Au shells on the Ag NPLs. These results unambiguously confirm that the reaction pH and concentration of HAuCl4 play crucial roles in controlling the reduction kinetics, which directly influence the shape of the final products.

In addition, previous studies have shown that the surfactant PVP can also favor in directing anisotropic growth.[53] In specific, PVP can prevent the oxidative etching events,[54] and the hydroxyl −OH groups present in the reaction mixture can act as mild reducing agents.[55] To validate the role played by PVP, we carried out two control experiments. First, the synthesis was carried out without the introduction of PVP, keeping all other conditions unchanged. As can be observed in Figure S4, disk-shaped nanostructures of different sizes were formed in the absence of PVP. The formation of few small Au NPs unevenly attached to disk-like nanostructures could also be observed (Figure S4). Second, PVP of the reaction mixture was replaced with another stabilizing agent CTAB of the same molar concentration. As can be seen in Figure S5, the addition of CTAB as the stabilizing agent leads to the formation of hexagonal Ag@Au NPs instead of Ag@Au nanodisks. The change in the morphology of the nanostructures can be ascribed to the selective binding of CTAB onto the (100) facet of Ag NPLs, which permits the oxidative etching to occur only at the corner sites of the Ag NPL seeds upon the introduction of Au3+ ion solution,[41] resulting in the formation of hexagonal Ag@Au nanostructures.

As the fabricated Ag@Au nanostructures contain silver (Ag), which is chemically unstable toward many chemical reactions, especially oxidation reactions,[56] we proceed to evaluate the chemical stability of the obtained Ag@Au nanodisks, monitoring their reaction with H2O2, a strong oxidant, through UV–vis spectroscopy. As can be seen in Figure a, the LSPR peak intensity of the original triangular Ag NPLs diminished within 2 min of the addition of about 0.2 mL aqueous H2O2 solution (2.3%). By contrast, the Ag@Au nanodisks showed only a slight decrement in the SPR peak intensity even after 12 h of the addition of H2O2 solution (Figure b). The results clearly suggest a remarkable enhancement in the chemical stability of the Ag NPLs after their transformation to Ag@Au nanodisks. The improved chemical stability of the nanostructures is due to the formation of Au shell over the Ag NP surface.[29,50]

Figure 4

UV–vis absorption spectra of (a) bare Ag NPLs and (b) Ag@Au nanodisks before and after the addition of 2.3% aqueous H2O2. Insets show the corresponding photographic images before (left) and after (right) the addition of H2O2.

It is well-known that the bimetallic Ag@Au nanostructures exhibit size-dependent LSPR properties, which can be precisely adjusted from the visible to the NIR spectral range.[52] To monitor how a change in the diameter of Ag@Au nanodisks affects their LSPR, we fabricated Ag@Au nanodisks of different diameters by varying the amount of Ag NPL seeds in the reaction mixture. For this, we used 2, 5, 10, and 15 mL of Ag NPL seeds in the synthesis solution that yielded Ag@Au nanodisks of diameters 27, 33, 38, and 45 nm, respectively. The utilization of Ag NPLs in higher concentrations in the reaction mixture induced the formation of fewer Au sites at their edges. Consequently, there occurred a less oxidative etching reaction at the edges of the Ag NPLs and a higher overgrowth (of Au layer) onto their other faces, producing Ag@Au nanodisks of higher diameters. In fact, utilizing this procedure, Ag@Au nanodisks of different diameters could be fabricated with great control. The UV–vis absorption spectra of the Ag@Au nanostructures (Figure a) revealed that the characteristic LSPR peak at 656 nm for 27 nm nanodisks, which suffered red shifts to 689, 726, and 780 nm on increasing the diameter to 33, 38, and 45 nm, respectively. The results are in accordance with the expected line, considering the variation of the effective dielectric constant of the bimetallic structures with the increase of the Au shell width.[57,58]

Figure 5

(A) Normalized UV–vis absorption spectra of the aqueous suspension of Ag@Au nanodisks of different diameters (27–45 nm). Insets show the TEM images of individual Au@Au nanodisks of increasing diameter. (B) Comparison of SERS spectra obtained from the aqueous solution of CV (10 μL of 10–6 M)-functionalized Ag NPLs, Ag@Au nanodisks, and Ag@Au nanorings, respectively. (C,D) SERS spectra of CV adsorbed on the Au@Au nanodisks of different diameters and corresponding EFs values.

SERS Properties of Ag@Au Nanodisks

The SERS performance of Ag@Au nanodisks was studied using the aqueous solution-based SERS detection by incubating the plasmonic nanostructures in CV solutions and compared with the SERS performance of Ag@Au nanorings and Ag NPLs under the same conditions. The utilization of CV as a probe molecule in the present study is specifically because of its distinct Raman spectral features and the natural adsorption characteristic over the Au NP surface without chemical modifications.[59]Figure b compares the solution-phase SERS spectra of CV-adsorbed triangular Ag NPLs, Ag@Au nanorings, and Ag@Au nanodisks, respectively. As can be seen from Figure b, the obtained bimetallic Ag@Au nanodisks exhibit a significantly higher SERS signal intensity compared to the bare Ag NPLs and bimetallic Ag@Au nanorings.

The quantification of the SERS signal was performed by the estimation of SERS enhancement factor (EF) for the most prominent Raman peak at 1620 cm–1, using the expression (eq )where ISERS and INor are the SERS signal intensities of the sample and the normal Raman peak of CV at ∼1620 cm–1, and NSERS and NNor are the corresponding values of the analyte probe molecules (CV) at the incident laser spot. We consider the adsorbed CV probe molecules distributed uniformly in aqueous solution; therefore, the values of NSERS and NNor can be used with the CV concentrations such as 1.0 × 10–6 M (CV incubated with NPs) and 1.0 × 10–3 M (CV solution), respectively.

The signal EFs were estimated considering the intensity of the Raman peak appeared at 1620 cm–1 (Figure S6). The calculated EF for the Ag@Au nanodisks was 5.97 × 103, which is approximately 3- and 2.3-fold higher than the same for the Ag NPLs (2.0 × 103) and Ag@Au nanorings (2.55 × 103), respectively. In Figure c,d, we compare the SERS spectra of Ag@Au nanodisks of different diameters and their estimated signal EF values as a function of disk diameter. The results presented in Figure d demonstrate that the EF value increases with the diameter of the Ag@Au nanodisks. The Ag@Au nanodisks of higher diameters (45 nm) exhibit a higher signal EF compared to the Ag@Au nanodisks of lower diameters (Figure d). The observed results are in good agreement with the previous reports.[60,61]

It is largely believed that solid SERS substrates or films produce intense “hot spots”, which improve significantly their SERS intensities.[62,63] Thus, we evaluated the substrate-based SERS performance by depositing the same amount of Ag NPL and Ag@Au nanodisk (45 nm diameter) samples over silicon (Si) substrates after drying the probe CV molecules (solution) over the samples prior to the SERS analysis (Figure a). As can be seen from Figure a, the substrate prepared with Ag@Au nanodisks exhibits much enhanced Raman signals in comparison to the Raman signal intensities obtained from their solution-casted counterpart. Apart from the effect of dispersion media (air in the first case and water in the second case),[64] the substrate-based assembly of plasmonic nanostructures seems to be of help in producing uniform and more intense “hot spots” which are responsible for the enhancement of Raman signal of the analytes. The Ag@Au nanodisks revealed highly intense Raman signals in comparison to the Ag NPLs. The EF values were calculated to be 0.18 × 105 and 0.47 × 105 for Ag NPLs and Ag@Au nanodisks, respectively. In comparison to bare Ag NPLs, the Ag@Au nanodisks exhibited almost 2.5-fold higher EF values. The estimated SERS EF for Ag@Au nanodisk was slightly less compared to the previous reports of Ag@Au nanostructures, as shown in Table S1. The low EF estimated for our Ag@Au nanodisks might be due to their nonuniform distribution/assembly over the Si substrate, as the SERS substrates were prepared simply by drop-casing their colloidal solutions. On the other hand, as has been shown in the AFM results in Figure f, the obtained Ag@Au nanodisks have flat, noncorrugated surfaces, which generate plasmonic hot spots of relatively lower intensities.

Figure 6

(a) Comparison of the SERS signals of bare AgNPLs and Ag@Au nanodisks deposited over Si substrates (CV, 1 × 10–6 M). (b) SERS spectra of Ag@Au nanodisks for different CV concentrations (from 10–6 to 10–9 M). (c) Reproducibility analysis of the SERS spectra of Ag@Au nanodisks (1 × 10–6 M CV) obtained from 12 randomly selected sites of the same substrate. (d) Corresponding peak intensity distribution at 1620 cm–1 for 12 sites of the SERS spectra.

To evaluate the SERS sensitivity of the Ag@Au nanodisks assembled over the Si substrate, we acquired the SERS spectra with different CV concentrations ranging from 1 × 10–6 to 1 × 10–9 M (Figure b). The SERS spectra for 1 × 10–6 M of CV revealed highly intense Raman peaks. However, their intensity decreased gradually on the reduction of CV concentration. To evaluate the SERS response of the fabricated Ag@Au nanodisks, the intensity of the 1620 cm–1 Raman dispersion peak of CV was recorded for different concentrations of the CV solution (Figure b). As can be observed in Figure S7 (Supporting Information), the intensity of this most intense Raman signal of CV varied linearly for the 10–6 to 10–9 M concentration range. In addition, as can be seen from Figure b, SERS signals are detectable even for the CV concentration of 10–9 M, suggesting that the resultant Ag@Au nanodisks are capable to detect CV molecules down to the 1 × 10–9 M concentration range. Notably, the observed sensitivity is comparable to the previous reports for bimetallic nanostructures (Table S1). The enhanced SERS efficiency of the Ag@Au nanodisks can be ascribed to two main plausible factors: (i) the LSPR peak of the Ag@Au disks at 780 nm is very near to the laser excitation wavelength used for recording the SERS spectra, which can be strongly coupled with the incident laser source, producing the resonance Raman effect.[65,66] (ii) The effective adsorption of CV molecules over the Ag@Au nanodisks and enhancement in the EM field between the nanogaps, which can serve as “hot spots” for a further increase in the SERS signal.[21] To demonstrate the reproducibility of the SERS substrates fabricated with Ag@Au nanodisks, the SERS spectra were measured at 12 different randomly selected sites of a substrate, for the CV concentration of 10–6 M. As can be observed in Figure c, the intensity of the SERS signal fluctuates by only about one order from place to place. The SERS reproducibility of the fabricated substrates was assessed by calculating the relative standard deviation (RSD) of the most dominant peak intensity at 1620 cm–1 (Figure d). The RSD value was obtained using the equation: RSD = SD/Im, where SD is the standard deviation of the peak intensity and Im is the average Raman peak intensities of the most dominant peaks.[67] The RSD of the 1620 cm–1 peak for 12 different sites was estimated to be 12.5%, indicating that the SERS substrates fabricated using Ag@Au nanodisks exhibit excellent reproducibility.

Conclusions

In summary, we present a simple, seed-mediated route for the fabrication of chemically stable bimetallic Ag@Au nanodisks with tunable LSPR response extending from the visible to the NIR spectral region. The high reaction pH and the tip-selective oxidative etching process utilized were found to be crucial for the suppression of the galvanic reaction, ultimately converting the triangular Ag NPLs into Ag@Au nanodisks. The obtained Ag@Au nanodisks not only show enhanced chemical stability, but also excellent optical properties. By simply changing the amount of Ag NPL seed particles, Ag@Au nanodisks of diameters 27–45 nm, with tunable LSPR peak position from the visible to the NIR region could be fabricated. The as-prepared Ag@Au nanodisks exhibit highly enhanced SERS activity with the CV signal EF as high as 0.47 × 105. The SERS substrates prepared using the fabricated nanodisks can detect CV concentration down to 10–9 M. The synthesis approach utilized to prepare Ag@Au bimetallic nanostructures of high chemical stability and tunable plasmonic and SERS properties indicates that it is a promising technique for fabricating other type of core–shell bimetallic nanostructures for applications in plasmonics and biosensing.

Experimental Details

Chemicals and Materials

Silver nitrate (AgNO3, 99%), trisodium citrate (Na3C6H5O7, 99%), chloroauric acid (HAuCl4·4H2O, 99.9%), sodium borohydride (NaBH4, >98%), l-ascorbic acid (AA, 99%), PVP (MW of 55 K), CTAB (98%), hydrogen peroxide (H2O2, 30 wt % in H2O), sodium hydroxide (NaOH, 98%), and CV (90%) were purchased from Sigma-Aldrich. All the chemicals were utilized as received and all the glasswares were washed thoroughly using deionized (DI) water prior to their use in experiments.

Synthesis of Ag NPL Seed NPs

The used process for the synthesis of Ag NPLs of ca. 25–35 nm edge lengths was similar to the process we reported earlier.[44] Briefly, 0.4 mL of 0.01 M aqueous AgNO3 solution and 0.6 mL of 0.1 M trisodium citrate were mixed together, followed by the addition of DI water to make the final volume of the reaction solution of 40 mL. The mixture solution was magnetically stirred for 10 min, and then 0.4 mL of freshly prepared NaBH4 solution (100 mM) was added. Subsequently, 0.1 mL of H2O2 was quickly injected to the previous solution. The color of the reaction mixture turned light yellow, yellow, red, green, and finally blue within 3–4 min. The final products (Ag NPLs) were collected by centrifugation at 12 000 rpm for about 30 min, followed by repeated washing in DI water.

Preparation of Bimetallic Ag@Au Nanodisks

The Ag@Au core–shell nanodisks were prepared by the seed-mediated growth process at room temperature using Ag NPLs as the template particles, PVP as the surfactant, and AA as the reducing agent. In a typical synthesis, 100 mg of AA and 66.6 mg of PVP were dissolved in 15 mL of DI water in a glass vial under magnetic stirring, after which 5 mL of the presynthesized aqueous suspension of Ag NPL seeds and 0.5 mL of 0.2 M aqueous NaOH solution were added into the reaction mixture sequentially. NaOH solution was added to increase the pH of the reaction mixture to 11.2. Then, the conformal deposition of Au and subsequent oxidative etching were performed by slow titration, adding 100 μL of 0.1 M aqueous HAuCl4 precursor solution slowly (10 μL/min) by a syringe. The color of the solution mixture changed from dark blue to pink, indicating the formation of Ag@Au nanodisks. The product was collected by centrifugation at 12 000 rpm for 15 min and subsequent washing in DI water. The obtained product was redispersed in DI water. The formation of free AuNPs was avoided by optimizing the concentration of the Au precursor solution, injection rate, and the concentration of AA in the reaction mixture. The diameter of the Ag@Au nanodisks was controlled by simply varying the amount of Ag NPL seeds (5, 10, 15, and 20 mL) in the reaction mixture, keeping all other conditions fixed.

Materials Characterization

The UV–vis spectra of the samples were collected using an Agilent 8453 UV–vis spectrophotometer. The TEM images of the samples were collected in a JEOL JEM-1010 microscope operating at 80 kV. The EDS elemental mapping of the samples was carried out in a Hitachi SU 8020 microscope operating in STEM mode at an accelerating voltage of 20 kV. The samples for TEM and EDS mapping analysis were prepared by dispersing the colloidal nanostructures over carbon-coated Cu grids and subsequent drying at room temperature. AFM of the samples was performed in a Bruker (ScanAsyst-Air model) microscope operating in tapping mode. For AFM analysis, the colloidal nanostructures were deposited onto a Si substrate and dried overnight at ambient conditions. Room-temperature Raman and SERS spectra of the nanostructures were recorded in a Bruker (SENTERRA) Raman spectrometer equipped with an Ar-ion laser of 785 nm excitation wavelength (λex).

Aqueous Solution-Based SERS Detection

The SERS spectra were obtained by using the same concentration of samples incubated in 25 μL of an aqueous solution of CV (10–6 M) for 2 h. The CV-adsorbed colloidal suspensions were then transferred to well-cleaned glass capillary tubes prior to SERS analysis. The Raman spectra of the solutions were collected using a 785 nm laser excitation source of 4.7 mW power with 4.0 μm beam diameter projected through a 50× objective lens.

Silicon Substrate-Based SERS Detection

The samples for substrate-based SERS were prepared by drying the same concentrations of the as-prepared AgNPL and Ag@Au nanodisk samples over clean silicon substrates (8 mm × 8 mm). In a typical procedure, 50 μL of each of the NP-containing suspensions (1 mg/mL) was deposited on Si substrates and then allowed to dry at room temperature. After this, 25 μL of an aqueous CV solution (10–6 M) was deposited over the sample and dried at room temperature. The Raman spectra were recorded using a 785 nm laser with the excitation source at room temperature. For all samples, the laser spot was fixed to 4 μM under a 50× objective lens; power density was 4.7 mW; and the signal acquisition time was set to be 3 s.