Surfactant-free Synthesis of Spiky Hollow Ag-Au Nanostars with Chemically Exposed Surfaces for Enhanced Catalysis and Single-Particle SERS.

Spiky/hollow metal nanoparticles have applications across a broad range of fields. However, the current bottom-up methods for producing spiky/hollow metal nanoparticles rely heavily on the use of strongly adsorbing surfactant molecules, which is undesirable because these passivate the product particles' surfaces. Here we report a high-yield surfactant-free synthesis of spiky hollow Au-Ag nanostars (SHAANs). Each SHAAN is composed of >50 spikes attached to a hollow ca. 150 nm diameter cubic core, which makes SHAANs highly plasmonically and catalytically active. Moreover, the surfaces of SHAANs are chemically exposed, which gives them significantly enhanced functionality compared with their surfactant-capped counterparts, as demonstrated in surface-enhanced Raman spectroscopy (SERS) and catalysis. The chemical accessibility of the pristine SHAANs also allows the use of hydroxyethyl cellulose as a weakly bound stabilizing agent. This produces colloidal SHAANs that remain stable for >1 month while retaining the functionalities of the pristine particles and allows even single-particle SERS to be realized.

Introduction

Ag and Au nanoparticles have been extensively studied in the past few decades due to their unique catalytic and plasmonic properties.[1−6] Up to now, the most popular route to produce Ag and Au nanomaterials has been through bottom-up colloidal synthesis, which typically employs a strongly adsorbing surfactant to direct particle growth and provide colloidal stability.[7−11] A variety of anisotropic metal nanoparticles with distinct properties can now be synthesized with the aid of surfactants, for instance, the polyvinylpyrrolidone (PVP)-assisted synthesis of nanocubes,[12−14] the cetyltrimethylammonium bromide (CTAB)-assisted synthesis of nanorods,[15−17] the Triton-X-assisted synthesis of nanostars, and so on.[18−20] Whereas the use of surfactant capping molecules offers easy manipulation of the morphology of the synthesized nanoparticles, their existence is also problematic for nearly every type of application, for example, in plasmonic sensing and catalysis, where the surface-bound surfactant molecules act as both physical and chemical barriers that restrict the free access of analytes/reactants to the nanosurface,[9,21−23] or in bioapplications where the free surfactants induce cytotoxicity.[24,25] As a result, methods for the postsynthesis removal of surfactant molecules or surfactant-free synthesis have become important areas of research.[26−30] For example, Gao et al. showed that diethylamine can be used as a general and highly effective intermediate ligand to facilitate the replacement of strongly adsorbed capping agents with weakly adsorbed capping agents.[29] Odom et al. demonstrated that it was possible to use weakly adsorbing and biocompatible Good’s buffers, such as 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, to synthesize Au nanostars.[30]

Among the anisotropic metal nanoparticles reported to date, spiky or hollow nanoparticles have drawn particular interest as catalysts due to their large surface-to-volume ratio and the presence of high-index crystal planes that are composed of highly active low-coordinated atoms.[31,32] If they are composed of plasmonic materials, such as Au and Ag, then spiky or hollow nanoparticles have also been shown to possess significantly enhanced plasmonic properties compared with their isotropic counterparts.[33−36] To take full advantage of the plasmonic and catalytic properties of hollow and spiky nanoparticles, a few research groups have reported methods for synthesizing Au/Ag nanostars that are both spiky and hollow.[20,37−41] However, in general, bottom-up methods that allow the production of spiky Au/Ag nanostars that contain a hollow interior remain extremely rare. More importantly, the current methods for producing Ag/Au nanostars with hollow interiors are either complex or require the extensive use of surfactants, which has significantly limited the application of these particles, despite their promising properties. For example, Rodríguez-Fernández et al. showed that spiky and hollow Au nanoparticles with tailored near-infrared plasmonic properties could be produced using Triton X-100 as the growth-directing agent.[20] Evans et al. showed that spiky and hollow Au nanoparticles with excellent catalytic and surface-enhanced Raman spectroscopy (SERS) activity could be produced using methyl-orange-FeCl3 templates.[38]

In this work, we demonstrate a mild and surfactant-free synthesis, which can be completed at room temperature within minutes to produce spiky hollow Au–Ag nanoparticles (SHAANs) with ca. 100% morphological yield. The unique morphology of the SHAANs combined with their highly accessible surfaces allows them to exhibit outstanding catalytic and single-particle SERS activities, which were significantly enhanced compared with those of their conventional surfactant-capped counterparts. Moreover, the high accessibility of the SHAAN surface also meant that they could be stabilized in the colloidal state using extremely weakly bound capping ligands, such as hydroxyethyl cellulose (HEC). This allowed the product colloid to remain stable for >1 month while fully retaining the surface accessibility of the particles. This is significant because good colloidal stability is crucial for a variety of important applications, such as catalysis, kinetic monitoring, and slow drug release,[1,42,43] but has typically been overlooked in previous studies regarding the surfactant-free synthesis of colloidal nanoparticles. More generally, the ability to access enhanced nanoproperties with anisotropic nanoparticles prepared using an easy and reliable room-temperature process not only offers a powerful tool for fundamental research but also paves the way for a broad range of important real-life applications, such as trace molecule detection, biomedical diagnostics, or the construction of functional devices.

Results and Discussion

The synthesis of the SHAANs is a simple and rapid one-step process. In brief, NaCl and AgNO3 solutions are first added in sequence to a HAuCl4 solution under constant stirring. The mixture is then incubated, after which l-ascorbic acid solution is quickly injected. (See details in the Materials and Methods section of the Supporting Information.) This causes the color of the solution to change immediately (typically within ca. 15s), indicating the formation of the product nanoparticles. The critical step in the process is the incubation of the mixed HAuCl4, NaCl, and AgNO3 solution before the addition of l-ascorbic acid. As shown in Figure a–e, the morphology of the nanoparticles obtained changed dramatically with different incubation times. The addition of ascorbic acid without incubation produced small spiky nanoparticles with an average diameter of ca. 40 nm. The resulting colloid had a light-blue color with λmax measured via UV–vis at ca. 782 nm (Figure f,g). Further scanning transmission electron microscopy (STEM) characterization and energy-dispersive X-ray (EDX) spectroscopy of this colloid sample revealed that some of the nanoparticles that appeared to have a hollow interior when observed with scanning electron microscopy (SEM) actually had a Ag/AgCl core (Figure S1). As shown in Figure b, when the incubation time was extended to 5 min, approximately half of the nanoparticles became larger SHAANs with a particle diameter of ca. 150 nm. These SHAANs had a larger number of spikes on the surface compared with smaller nanoparticles obtained with less incubation, and the spikes were notably narrower and longer. Additionally, the interiors of particles were found to be hollow, which will be discussed in detail in the following section. The change in morphology with incubation was also reflected in the color of the colloids and their UV–vis extinction spectra. As shown in Figure f,g, the color of the colloids with incubation was clearly different, and their extinction peaks were broadened and red-shifted significantly compared with the nonincubated sample. The optimal incubation time, which led to the highest yield (ca. 100% morphological yield) of SHAANs, was found to be 10 min, as shown in Figure c. The resulting colloid was gray, with its UV–vis extinction peak centered around 805 nm (Figure f,g). Extending the incubation time beyond 10 min led to the population of SHAANs decreasing and eventually nearly completely disappearing. As shown in Figure d, the population of the product colloid with 15 min of incubation was similar to those obtained with 5 min of incubation, which is reflected by their identical colors and UV–vis extinctions (Figure f,g). If the incubation time was increased to 60 min, then very few SHAANs were formed, and the product colloid mainly consisted of small (ca. 30 nm) spheres (Figure e). Unsurprisingly, this colloid was purple, with a UV–vis extinction centered at ca. 527 nm (Figure f,g), which is characteristic of small spherical Au nanoparticles.[44]

Figure 1

(a–e) Scanning electron microscopy (SEM) images of Au/Ag alloy nanoparticles synthesized by incubating the mixture solution of HAuCl4, NaCl, and AgNO3 for (a) 0, (b) 5, (c) 10, (d) 15, and (e) 60 min and then adding ascorbic acid. Scale bars in all images correspond to 500 nm. (f) Corresponding optical images of colloids prepared with different incubation times. (g) UV–vis spectra of colloids prepared with different incubation times.

Further characterization of the nanoparticles was performed by scanning transmission electron microscopy–high-angle annular dark field (STEM-HAADF) imaging. This was used in combination with energy-dispersive X-ray spectroscopy (EDX) to investigate the internal morphology, local elemental distribution, and composition of the SHAANs. Images from the 10 min incubated specimen revealed that the homogeneous population of nanoparticles has cubic hollow cores in addition to the spiky surface morphology observed by SEM (Figure a). The average diameter of the cubic core was found to be 150 ± 16 nm, as measured from 100 individual SHAANs. Figure b shows the high-magnification STEM-HAADF image of a typical SHAAN. The hollow cavity of the SHAAN was measured to be ca. 100 nm in diameter, and the length of spikes ranged between 30 and 45 nm. Elemental mapping revealed a fairly evenly dispersed mixture of Ag and Au, which resembled bimetallic Au–Ag alloys, with a mean composition of 90 at % Au and 10 at % Ag (Figure c,d). Interestingly, as shown in Figure e,f, the high-magnification elemental line scan also revealed that the at % ratio between Ag and Au was clearly higher at the surface of the SHAANs, which suggested the existence of a ∼1 nm thick Ag-rich surface layer that could have been formed due to the lower surface energy of Ag compared with Au.[45] STEM-HAADF analysis for a nonincubated specimen (see Figure S2) also revealed some hollow particles but with a smaller and less regular size and shorter spikes, consistent with the SEM data. These particles also had a composition of 90 at % Au and 10 at % Ag and a few nanometers thick Ag-rich surface layer (Figures S3 and S4).

Figure 2

(a) Low- and (b) high-magnification scanning transmission electron microscopy–high-angle annular dark field (STEM-HAADF) images of SHAANs. Energy-dispersive X-ray (EDX) elemental mapping of SHAANs: (c) Au mapping image and (d) Ag mapping image. (e) EDX elemental mapping of a tip on the SHAAN shown in panel b. (f) Plot showing the material composition obtained with a high-magnification elemental line scan across the line labeled in panel e. Scale bars in panel a, panels b–d, and panel e correspond to 200, 60, and 10 nm, respectively.

The proposed mechanism for the formation of SHAANs is shown in Scheme . The synthesis is initiated by the addition of AgNO3 to a mixed NaCl/HAuCl4 solution, which leads to the formation of small AgCl seed nanoparticles. This is consistent with the observation that the mixture solution gradually changes from transparent to slightly cloudy after the addition of AgNO3. Subsequent incubation leads to the growth of the AgCl seeds to form larger cubic AgCl nanoparticles; however, this process can be interrupted by the addition of ascorbic acid, which triggers a series of redox reactions (Table S1) that ultimately lead to the formation of SHAANs if the optimized incubation time of 10 min is used. More specifically, the addition of ascorbic acid to the incubated solution leads to the reduction of the AgCl nanoparticles to Ag0 nanoparticles, as has been previously observed.[46−48] However, these are not stable in the presence of Au3+ cations, which can galvanically reduce them, leading to the replacement of the Ag0 by Au0 in the nanocrystal structure, a process that would be expected to yield cubic hollow Au shells.[46,47] Simultaneously, the presence of ascorbic acid and Au3+ cations along with residual Ag+ and Cl– leads to growth of spikes on the surface of the particles.[34]

Scheme 1

Schematic Illustration of the Formation of SHAANs with 10 min of Incubation Time Compared with When a 0 min of Incubation Is Used

Without incubation, the reaction goes through three simultaneous pathways, which leads to the formation of a mixture of nanoparticles including solid Au stars, Au hollow spheres, hollow Ag–Au stars, and solid Au stars with spherical/cubic AgCl cores. With 10 min of incubation, the reaction goes through a single pathway, leading to the formation of SHAANs.

This mechanism accounts for the importance of the duration of the incubation step and the nature of the products that are formed when it is changed from the optimum value. If the initial mixture was over-incubated (60 min), then the majority of Ag+ and Cl– ions were consumed in the growth of AgCl nanoparticles, and not enough were left to assist the formation of Au spikes. Therefore, the nanoparticles formed by the subsequent addition of ascorbic acid were mostly quasi-spherical (Figure e). In contrast, if the initial mixture was under-incubated (0 min), then there were not sufficient AgCl seeds to template the growth of SHAANs, and the AgCl seeds that were available were under-grown, which resulted in the formation of polydisperse spiky/hollow nanoparticles (Figure a and Figure S2). Although it is important to use the correct incubation time to achieve an appropriate balance between AgCl seeds, Ag+ and Cl– necessary for the formation of SHAANs (Figure c), it is useful to stress that provided that this time is used, the process is very robust, and it reliably generates SHAANs, as shown by SEM and SERS measurements in Figures S5 and S6. Although AgCl nanoparticles act as intermediates during the synthesis, a significant amount of residual AgCl is not expected to be present in the product SHAANs because it is reduced by the excess ascorbic acid in the solution. This was confirmed by the observation that washing SHAANs with ammonia–water, which would dissolve any residual AgCl, gave no detectable change in the morphology of the SHAANs (Figure S7).

X-ray photoelectron spectroscopy (XPS) was used to further study the surface composition of the SHAANs, in particular, to identify the oxidation state of the metals and the number of species present. Surprisingly, the XPS revealed that Ag was present in just one oxidation state, Ag+, indicated by the photoelectron peak at 367.5 eV (Ag 3d5/2) (Figure a).[49,50] Because, as previously discussed, it is unlikely that a significant amount of AgCl is present in the product SHAANs, the Ag+ observed by XPS is most likely in the form of Ag2O, which is formed postsynthesis due to the high reactivity of colloidal Ag toward oxygen.[51] As shown in Figure b, the XPS also showed a photoelectron peak at 83.4 eV, corresponding to metallic Au 4f7/2. Interestingly, the binding energy observed for Au was lower than the conventional values reported in literature (84.0 eV). This could be due to charge-transfer effects that arise from the formation of Ag–Au alloys in the SHAANs.[52,53]Table S2 and Figure S8 show the XPS data and at % of the relevant elements in the SHAAN sample from XPS. Importantly, as a surface-specific technique, the XPS results were consistent with the Ag surface enrichment observed in the STEM-elemental mapping and showed that the at % ratio between Ag and Au (1:1.6) on the surface of the SHAANs was much higher than the overall ratio between Ag and Au in the whole particle (1:9).

Figure 3

XPS analysis of SHAANs showing the characteristic peaks of (a) Ag and (b) Au.

The plasmonic properties of SHAANs were investigated using a finite element method, with the particle structure and material composition modeled according to the microscopy measurements. (See the Materials and Methods section of the Supporting Information.) The volume-averaged near-field enhancement factor was calculated aswhere the volume V was made by a 2 nm thick shell covering the nanoparticle surface (Figure a). Figure b compares the calculated electromagnetic field distribution of individual SHAANs (x corresponds to the approximated length of the spikes within a SHAAN) and a hollow cube core excited by 785 nm laser. Regardless of the exact spike length, the near-field enhancement factor maps showed that the distribution of the excited surface plasmon resonance was mainly localized around the sharp tips rather than within the hollow cavity. Essentially, the electromagnetic-field distribution around the SHAANs was much more similar to that of solid nanostars rather than hollow/porous nanoparticles or nanocubes.[54−56] The profile and strength of the volume-averaged near-field enhancement were found to be highly dependent on the length of the spikes. For example, Figure c shows a plot comparing the volume-averaged near-field enhancement of individual hollow cubes and SHAANs at different wavelengths, which shows that the optimal excitation wavelength shifted by >200 nm when the spike length increased from 20 to 40 nm. Figure d compares the maximum and average near-field enhancement of hollow nanocubes and SHAANs at the same excitation wavelength as our Raman measurements (see as follows). As shown in the plot, the SHAANs were significantly more plasmonically active compared with hollow nanocubes. Interestingly, the simulations also revealed that the maximum near-field enhancement increased with the increasing spike length of the SHAANs, whereas the volume-averaged enhancement decreased.

Figure 4

Finite element method simulation of the electrodynamic of SHAANs and hollow cube nanostructures. (a) 3D renderings of a SHAAN, in red (30 nm spike length), and of the volume, in light blue, which define the volume-averaged near-field enhancement factor. (b) Near-field enhancement factor maps for SHAANs with three different spike lengths x (x = 20, 30, or 40 nm) and the hollow nanocube core. (c) Volume-averaged near-field enhancement factors for the structures in panel b. (d) Comparison between the volume-averaged near-field enhancement factors and maximum punctual enhancement factor of the structures in panel b calculated at 785 nm excitation wavelength.

The SERS performance of SHAANs was first investigated using 4-mercaptobenzoic acid (MBA) as the model analyte. Strong SERS signals of MBA were observed using freshly prepared, well-dispersed pristine SHAANs colloids as the enhancing substrate, and the limit of detection was measured to be ca. 10–8 M, as shown in Figure S9. The analytical enhancement factor (AEF), which is calculated differently from the enhancement factor (EF), is widely regarded as the gold standard for quantifying the enhancing ability of colloidal SERS substrates because the complex geometry makes the direct calculation of EF difficult.[57] Using MBA as the probe analyte, the AEF of SHAANs was determined to be ca. 4.9 × 105 under nonresonant conditions for the pristine SHAAN colloid (Figure S9). This value is notably stronger than the enhancement predicted in Figure , which is likely due to the presence of additional far-field electromagnetic enhancement and chemical enhancement that are not considered in the simulations. It is worth noting that this experimentally recorded value is comparable to the AEF value reported for state-of-the-art plasmonic particles, such as Au nanotriangles (ca. 1.2 × 105), and to aggregated silver nanoparticles (ca. 5 × 105), which have been used for single-molecule SERS.[57,58] To clearly demonstrate the advantages of the SHAANs, we compared the SERS of MBA on SHAANs with solid nanostars synthesized following a well-recognized surfactant-free approach from the literature and with surfactant-free hollow Ag–Au nanocubes produced via an in-house procedure derived from literature.[56,59] As shown in Figure S10, the hollow nanocubes were ca. 250 nm in diameter, which was similar to the diameter of the cubic hollow core of the SHAANs, whereas the solid nanostars were ca. 50 nm in diameter. As shown in Figure S11, the SERS signals of MBA were clearly the strongest on SHAANs. More specifically, under the same unaggregated experimental conditions, no signals of MBA were observed from the hollow nanocubes, whereas the SERS signals of MBA obtained from the conventional solid nanostars were more than two times weaker than the signals of MBA obtained from SHAANs.

It is important to note that the strong AEF of SHAANs arises not only from their excellent plasmonic properties but also from their chemically exposed surfaces. More specifically, the surfactant-free synthesis means that the only chemical species adsorbed on the surface of the SHAANs were Cl– ions originating from HAuCl4 and NaCl. It is known that charged ligands, such as Cl–, occupy only ca. 20% of the surface of Ag/Au nanoparticles.[60] As a result, the blank SERS signals of the pristine SHAANs colloids showed only a single weak Au–Cl vibration band (Figure S12). This means the surface of SHAANs is much more accessible to analyte molecules than conventional spiky nanostars, which are covered in strongly adsorbed ligands, such as Triton-X, PVP, or poly(ethylene glycol) methyl ether thiol. The lack of capping agents in surfactant-free syntheses does lead to issues in the long-term stability of the product nanoparticle colloids because there are no capping ligands to protect the nanoparticles from aggregation. Indeed, it was found that the pristine SHAANs colloids were very unstable against aggregation and typically precipitated as aggregates from solution within 1 h. The challenge is to be able to stabilize the SHAAN colloid without compromising the surface accessibility of the particles. This was not a problem because another advantage of the surfactant-free environment is that it allowed us to use weakly adsorbed polymeric molecules, such as HEC polymer, as stabilizing agents.[61−63] This allowed the SHAANs to remain highly stable for >1 month, as characterized by UV–vis spectroscopy (Figure S13), while retaining their surface accessibility for further applications. Interestingly, SERS monitoring of the same batch of SHAANs revealed that their SERS activity slightly decreased, by approximately two-fold, during the first day of storage (Figure S14). This is most likely due to oxidation of the thin Ag surface layer rather than aggregation or Ostwald ripening of the SHAANs because the extinction spectrum of the colloid remained unchanged.[17,64] More importantly, the long-term SERS kinetics showed that the slight decrease in SERS activity was observed only during the first day of storage, after which the SHAANs remained highly plasmonically stable for >1 month. The high chemical stability of SHAANs against further oxidation is likely due to SHAANs being mainly composed of a Ag–Au alloy, which is consistent with previous plasmonic studies comparing Ag and Ag–Au nanocubes.[64]

Figure a compares the SERS activity of three batches of pristine, HEC-stabilized, and PVP-stabilized SHAAN colloids for the detection of a model SERS analyte, crystal violet (CV). As shown in spectra sets i and ii, the intensity of the SERS signals of CV was fully retained on the HEC-SHAANs. In contrast, because the Cl–, which must adsorb onto the enhancing surface to facilitate the adsorption of the cationic CV dye, cannot displace PVP, a significant drop in SERS intensity, by approximately ten-fold, was observed when PVP-SHAANs (Figure a, spectrum iii) were used as the enhancing substrate. It is worth noting that CV is not a special case. The exposed surface of SHAANs also provides significant advantages in the SERS detection of important real-life targets. For example, Figure b,c compares the SERS signals of two anticancer drugs, Niraparib and Doxorubicin, obtained using pristine SHAANS, HEC-SHAANs, and PVP-SHAANs at a final drug concentration of 10–5 M. For both anticancer drugs, the SERS signals were intense on pristine SHAANS and HEC-SHAANs but barely visible on PVP-SHAANs, which emphasizes the importance of surface accessibility in the SERS monitoring of real-life analytes.

Figure 5

(a) SERS spectra of 10–6 M crystal violet (CV), (b) 10–5 M doxorubicin (DOX), and (c) 10–5 M niraparib (NI) recorded on pristine SHAAN colloid (i), hydroxyethyl-cellulose (HEC)-stabilized SHAAN colloid (ii), and PVP-stabilized SHAAN colloid (iii). (d) Time-dependent SERS spectra showing single SHAANs moving into and out of the probed laser volume. Spectra with detectable SERS signals of the MBA analyte at 1045 cm–1 are highlighted in red. The plot below shows the change in the SERS signal for an extended period of 20 s, with red dots representing the appearance of the MBA signal and black dots representing the disappearance of the MBA signal. (e,f) Plots showing the decrease in the concentration of 4-nitrophenol with time in the NaBH4 reduction of 4-nitrophenol catalyzed by (e) HEC- and (f) PVP-stabilized SHAANs. Insets show the rates of the reactions.

The outstanding AEF of SHAANs also suggests that they might be suitable for single-particle SERS measurements, which is an important property in both fundamental and applied studies, such as SERS sensing and imaging in biological samples.[65,66] The HEC-SHAAN colloid was diluted, labeled with MBA, which is commonly used as a Raman tag molecule for biological studies, and Raman-probed using a confocal Raman microscope with a diffraction-limited probe diameter of 1.04 μm to test this. Figure d shows a subset of the SERS spectra from a total of 20 spectra series continuously recorded at 1 s intervals. Even though MBA is a nonresonant analyte and the accumulation time was short, the signal of MBA was still intense and was measured to be 38 ± 11 counts using the characteristic peak at 1045 cm–1. More importantly, the MBA signals showed an on/off pattern, which is associated with the movement of the enhancing particle–substrate moving into and out of the probed volume due to Brownian motion. Because within this diluted SHAAN colloid there is only a single particle every 2000 μm3 (see the Supporting Information for details), which is more than 2000 times larger than the probe volume (0.7 μm3) of the confocal microscope, this meant that the consistently observed on/off signals of MBA most likely arose from a single SHAAN in the probe laser acting as a single-particle SERS probe. To validate that the SHAANs were active as single particles, we studied the effect of colloidal aggregation on the plasmonic properties of SHAANs by performing contrast SERS kinetic experiments using pristine and HEC-SHAANs. As shown in Figure S15, the SERS signals obtained from the pristine SHAANs decreased almost linearly as they aggregated naturally without any visible precipitation, whereas the SERS signals of the unaggregated HEC-SHAANs remained constant over the same period of time. This shows that aggregation has a negative impact on the SERS activity of the SHAANs, which is consistent with previous reports regarding star-type Au particles,[67,68] and more importantly, argues against attributing the intermittent signals in Figure d to aggregates. The SERS properties of the SHAANs functionalized with MBA Raman tags were also studied in artificial serum, which contained 3.3 wt % albumin, to mimic a real-life environment. SERS analysis in the presence of albumin is challenging because albumin is known to compete with analyte molecules for the enhancing surface and prevent particle aggregation, which is often crucial for the formation of plasmonic hot spots.[69,70] As shown in Figure S16, because the MBA Raman tag molecules were strongly adsorbed on the single-particle SERS-active SHAANs, this allowed the strong SERS signals of MBA from the SHAANs to be fully retained, even in the presence of albumin, which demonstrates their potential in biochemical applications, for example, as pH reporters.

The HEC-SHAANs also showed superior catalytic properties to PVP-SHAANs. As shown in Figure e, the reduction of 4-nitrophenol by NaBH4 occurred immediately when catalyzed by HEC-stabilized SHAANs, as measured by UV–vis spectroscopy (Figure S17). In contrast, for PVP-stabilized SHAANs, there was a delay of ca. 90 s before any reaction could be detected (Figure f and Figure S18). In addition, the average rate constant for the reaction catalyzed by the HEC-SHAANs (k = 0.0406 ± 0.0015 s–1) was more than 3.8 times faster than that for their PVP-stabilized counterparts (k = 0.0105 ± 0.0015 s–1), as shown in the insets of Figure e,f. This can be attributed to the need for surface reconstruction of PVP-SHAANs when acting as catalysts due to the strongly adsorbed PVP capping on the particles’ surfaces.[71]

The high yield and excellent properties of the SHAANs also makes them promising candidates for the construction of multidimensional nanomaterials with advanced functionalities. For illustration, we show that HEC-SHAANs can be readily used as the building block for the construction of surface-exposed nanoparticle sheets (SENSs) via an interfacial self-assembly technique previously reported by our group.[72]Figure shows typical SEM and SERS results obtained from SHAAN SENSs. As shown in Figure a,b, the SHAANs in SENSs are held as a densely packed 2D monolayer anchored on the surface of the polystyrene. Importantly, the surface of the HEC-SHAANs in SENSs is both chemically and physically exposed for interactions in applications. Figure c shows a 5 μm × 5 μm Raman map of the HEC-SHAAN SENS with MBA adsorbed on the particle surface. As shown in Figure d, the assembled SHAANs remained highly plasmonically active, and the relative standard deviation of the SERS intensity of the MBA peak at 1045 cm–1 across 100 randomly selected points within this mapped area was very low at just 10.8%, which demonstrates the potential of HEC-SHAANs as a plasmonic building block for constructing uniform and active plasmonic devices.

Figure 6

(a) SEM image of a SHAAN SENS showing a monolayer of densely packed SHAANs on the surface. (b) Atomic force microscopy (AFM) and (c) Raman mapping images across a 5 μm × 5 μm area on the SHAAN SENS labeled with MBA. (d) Raman intensity of the MBA peak at 1045 cm–1 obtained from 100 random spots on the Raman mapping image.

Conclusions

We have demonstrated a rapid, mild, and surfactant-free one-pot synthesis to prepare spiky Ag–Au nanoparticles (SHAANs) with near 100% morphological yield. The as-synthesized nanoparticles have a hollow interior with a dense covering of narrow spikes irradiating from the surface, which gave rise to distinct catalytic and plasmonic properties. Moreover, the fact that the synthesis was surfactant-free meant that the surfaces of the SHAANs were chemically exposed, which allowed the use of HEC as a weakly bound stabilizing agent to ensure colloidal stability while retaining surface accessibility for further applications. As a result, the HEC-SHAANs not only were highly stable but also exhibited greatly enhanced functionalities as both SERS substrates and catalysts when compared against their surfactant-capped counterparts. The combination of easy preparation and outstanding functionality means that SHAANs have the potential to be exploited in a broad range of applications as both functional materials in their own right and as building blocks for functional materials with advanced properties.