Facile synthesis of intra-nanogap enhanced Raman tags with different shapes

Abstract

Hot spot engineering in plasmonic nanostructures plays a significant role in surface enhanced Raman scattering for bioanalysis and cell imaging. However, creating stable, reproducible, and strong SERS signals remains challenging due to the potential interference from surrounding chemicals and locating SERS-active analytes into hot-spot regions. Herein, we developed a straightforward approach to synthesize intra-gap nanoparticles encapsulating 4-nitrobenzenethiol (4-NBT) as a reporter molecule within these gaps to avoid outside interference. We made three kinds of intra-gap nanoparticles using nanorods, bipyramids, and nanospheres as cores, in which the nanorod based intra-gap nanoparticles exhibit the highest SERS activity. The advantage of our method is the ease of preparation of high-yield and stable intra-gap nanoparticles characterized by a short incubation time (10 mins) with 4-NBT and quick synthesis without requiring an additional step to centrifuge for the purification of core nanoparticles. The intense localized field in the synthesized hot spots of these plasmonic gap nanostructures holds great promise as a SERS substrate for a broad range of quantitative optical applications.

Graphical Abstract

A straightforward approach to synthesize intra-nanogap enhanced Raman tags using nanorod/bipyramid/sphere as cores was developed, aiming to achieve stable, reproducible, and strong SERS signals without potential interference from surrounding environment.

TABLE OF CONTENTS (TOC)

Authors are required to submit a graphic entry for the Table of Contents (TOC) in conjunction with the manuscript title. This graphic should capture the readers’ attention and give readers a visual impression of the essence of the paper. Labels, formulae, or numbers within the graphic must be legible at publication size. Tables or spectra are not acceptable. Color graphics are highly encouraged. The resolution of the figure should be at least 600 dpi. The size should be at least 50 mm × 80 mm with a rectangular shape (ideally, the ratio of height to width should be less than 1 and larger than 5/8). One to two sentences should be written below the figure to summarize the paper. To create the TOC, please insert your image in the template box below. Fonts, size, and spaces should not be changed.

1. Introduction

Surface-enhanced Raman scattering (SERS) has grown to become an attractive, powerful, and ultra-sensitive analytical tool, since its discovery in the 1970s [1]. Due to the advantages of real-time monitoring and providing specific “fingerprint” information of a wide range of analytes in a non-destructive and non-invasive way, SERS has been widely used in electrochemistry, biological analysis, biomedical applications, environment monitoring, and single molecule detection [2–8]. The localized surface plasmon resonance (LSPR) arises from the collective oscillation of free electrons that generate enhanced local electromagnetic fields on the surfaces of noble metal nanoparticles. The resulting LSPR is confined in a very small volume around subwavelength-scale plasmonic nanostructure when excited by resonant incident light [9]. These extremely intense and highly localized electromagnetic fields induced by LSPR can amplify the Raman scattering signal by several orders of magnitude especially at nanoscale “hot spots” [10]. The LSPR strongly depends on the shape, size, and composition of the nanoparticles (NPs) [11]. Noble metals (typically Au and Ag) are able to exhibit LSPR properties in the visible range of spectrum due to the energy levels of d-d band transitions. In order to achieve a larger Raman enhancement factor (EF), NPs with intentional hot spots formed by tips, gaps, or pores of varying shapes including nanospheres, nanorods, bipyramids, nanostars, nanocubes, nanoflowers, porous or nanogap structures, etc. have been prepared [12].

Among these NPs, it is believed that the SERS signal of nanogap nanostructures primarily arises from the molecules adsorbed in the hot-spot area located in the intra- or inter-gap regions [13]. In general, metal nanogaps are capable of enhancing the local electromagnetic field intensity by 2–5 orders of magnitude [14], and Raman signal intensity variation of molecules trapped at the gap agrees with the theoretical studies, showing higher enhancement factors at shorter distances [15]. Several approaches have been reported for the preparation of plasmonic substrates with nanogaps, and they can be mainly divided into two groups: top-down methods and bottom-up methods. For top-down methods, nanostructured materials are often obtained through lithographic methods by means of structural decomposition from a bulk solid [16]. The dependence of SERS intensities on the gap topography and distance of gold gap-nanorods with different segment lengths made by focused ion beam and on-wire lithography has been elucidated, which indicates that smooth gaps lead to a bigger SERS intensities and EFs than rough gaps [17]. Compared to the top-down methods, bottom-up methods have been more extensively adopted to systematically synthesize nanogap structures due to low manufacturing cost and ease of preparation. Typically, there are three kinds of nanogap structures: 1) Inter-gap NPs formed from NPs [18–26], whose field strengths in the coupling region between adjacent nanoparticles are several orders of magnitude larger than the fields at the surface of single nanoparticles; 2) Intra-gap NPs [27–29], which have 1–2 nm gap inside and can generate highly uniform and reproducible SERS signals; and 3) Nanogaps between nanoparticles and a metal (Au, Ag, Cu) or semiconductive surface or nanoparticles that have been separated by an ultrathin GO film [30–35]. In contrast to the other types, the advantages of intra-gap NPs include: (i) Raman tags or analytes protected in the gaps have no spectral interference from additional Raman bands from unwanted external environment molecules; (ii) no spatial occupation competition with analyte molecules; (iii) excellent photochemical and photothermal stability; and (iv) prolonged shelf life, which significantly improves the accuracy of quantitative SERS determination and thus enables them for multiplexed monitoring of various analyte molecules in complicated biological and chemical systems [36].

Since Nam’s group reported the intra-gap NPs in 2011, the preparation process has involved the design and manipulation of nanogap spacers such as thiolated DNA [37], SiO₂ [38–39], polymers [40–41], and other big molecules [42] with the integration of embedded Raman dyes. Nevertheless, Raman dye-labelled thiolated DNA can be expensive, and the process of implementing it is time-consuming and laborious. Petal-like gap-enhanced Raman tags (P-GERTs) for high-speed Raman imaging are prepared using 4-NBT molecules [43–44], and a few synthetic papers have been reported using Raman-active molecules including 4-mercaptopyridine [45–46], 1,4-benzenedithiol and 4-methylbenzenethiol [47–48]. However, the key advantages of our method over them are the ease of preparing intra-gap nanoparticles through shorter incubation time (4-NBT, 10 mins) compared to other methods (normally more than 30 mins), and not requiring an additional purification step. Herein, we developed a straightforward synthesis of intra-gap NPs without using nanogap spacers but instead encapsulating 4-NBT Raman reporter molecules inside the gaps, in which strong plasmon coupling results in amplified electromagnetic fields and thus significantly enhances the Raman signals of 4-NBT. Specifically, we used three different shaped nanoparticles (nanosphere, nanorod, and nano-bipyramid) as cores immobilized with the Raman reporter 4-NBT. Then silver atoms were reduced onto the cores to form a shell, followed by replacement using HAuCl₄. As a result, small nanogaps with high-density hot spots were formed, encapsulating 4-NBT molecules in the gaps and significantly enhancing the Raman intensity. We found that the nanorod based intra-gap NPs exhibited the highest, uniform, and reproducible Raman intensity. With the advantages of a simplified synthesis and the quantitative diagnostic applications of intra-gap NPs across disciplines, this versatile method can extend to other core shapes and expand the applications of SERS-based bioanalytical sensing.

2. Results and discussion

Seed-mediated procedures were used to make three kinds of gold cores (nanorod, bipyramid, sphere) [49–51]. As shown in Figure. 1(a–c), Au nanorods with average aspect ratio of 3.33 (length ~100 nm, width ~30 nm), Au bipyramids with average aspect ratio of 3.77 (length ~177 nm, width ~47 nm), gold nanospheres (~11 nm diameter) were synthesized, and the size distribution of the Au cores based on the TEM data are shown in Figure S1 in the Electronic Supplementary Material (ESM). The UV-Vis extinction spectra of each particle shape are shown in Figure. 1(d–f). The UV-Vis extinction spectrum of the bipyramid particles indicates that there are small impurities, such as smaller bipyramids or spheres, in the prepared bipyramid core solution. This is indicated in the extinction spectrum by the shoulders on the plasmon resonances and validated in the TEM image in Figure 1(b). Without requiring the step of centrifuging for the removal of surfactant, gold cores in the original solution were directly immobilized with 4-NBT molecules for only 10 mins. Then a Ag layer was adsorbed from the reduction of Ag⁺ ions by ascorbic acid in the solution also containing ammonium hydroxide. Intra-gap nanoparticles with different cores were then prepared by using galvanic replacement reactions (GRR) to substitute Ag with Au, resulting in particles that consist of cores, gaps, and Ag-Au alloy shells with stable and reproducible compositions. The resultant structures were then characterized by transmission electron microscopy (TEM) and energy-dispersive X-ray spectroscopy (EDS) for conformation of gap formation between the Au cores and Au-Ag shells and for elemental analysis to reveal the metal composition, respectively.

Figure 1.

TEM images of Au cores. (a) Au nanorod; (b) Au bipyramid; (c) Au nanosphere, and their corresponding UV-Vis extinction spectra for the rods (d), bipyramids (e), and spheres (f) in the right panel. Scale bars are all 200 nm.

The structures of nanorod based intra-gap NPs with average aspect ratio of 2.2 (length ~130 nm, width ~60 nm), bipyramid based intra-gap NPs with average aspect ratio of 2.7 (length ~200 nm, width ~73 nm), and nanosphere based intra-gap NPs nanosphere (~31 nm diameter) were visualized by TEM images (Figure 2(a–c)). Their respective size distribution curves are shown in Figure S2 in the ESM. The decreased aspect ratio and the increased diameter of intra-gap nanoparticles indicated the coating of Au-Ag shell onto the Au cores, which is further verified by their UV-Vis extinction spectra (Figure 2(d–f)). Specifically, there is a blue shift of longitudinal surface plasmon resonance during the shell growth of gold nanorods and bipyramids with the change from 811 nm to 656 nm and from 1027 nm to 894 nm, respectively. On the contrary, due to the structure change from nanosphere (LSPR 519 nm) to flower-like shape, we can see two new red-shifted LSPR peaks (544 nm and 656 nm) of nanosphere based intra-gap NPs. Due to the heterogeneity of bipyramid cores in the original solution, some impurity-based (smaller bipyramids or spheres) intra-gap nanoparticles were also formed in solution as shown in Figure S3 in the ESM. Interestingly, the nanosphere based intra-gap nanoparticles (Figure 2(c)) have irregular outer shells, that are different from the sphere-like gap-enhanced Raman tags (s-GERTs) with a smooth external shell [43].

Figure 2.

TEM images of intra-gap nanoparticles with different Au cores: (a) Au nanorod; (b) Au bipyramid; (c) Au nanosphere, and their corresponding UV-Vis extinction spectra in the lower panel. Scale bars are all 100 nm.

Next, we explored the factors influencing the structure of the intra-gap nanoparticles. First, to assess if the same method can make nanorod based intra-gap NPs with different sized nanorod cores, we prepared nanorod cores with a smaller size, whose UV-Vis extinction spectrum is shown in Figure S4 in the ESM. It should be noted that the ratio of Au³⁺/Ag⁺ and their concentrations in the smaller nanorod core preparation solution are different from those in the larger nanorod solution in order to study the feasibility of this method. The result in Figure 3(c) shows that the nanorod based intra-gap NPs were successfully prepared using this method. The surface appears rough, which is different from the nanorod based intra gap NPs (Figure 2(a)) that had much smoother external shells. The change in roughness corresponds to the difference of the Au³⁺/Ag⁺ concentrations in the reaction solution. Additionally, the nanorod based intra-gap NPs can only be prepared when the added HAuCl₄/AgNO₃ mixture has a molar ratio of 2:1 (Figure 3(a–c)). Figure S5 in the ESM shows that a few small pores can be seen in the shuttle-like nanoparticles that are made when the ratio of Au³⁺/Ag⁺ is 1:1. We also investigated the shapes of bipyramid based intra-gap NPs by keeping the same concentration of AgNO₃ but varying the amount of HAuCl₄. As shown in Figure 3(d–f), the gaps gradually disappeared as the molar ratio of HAuCl₄/AgNO₃ was increased to 3:2, as the Au atoms progressively occupy the vacancy. In addition, changing the surfactant from CTAB to CTAC also allows the generation of intra-gap NPs, but with a much lower yield (~40% in Figure S6 in the ESM), which is ascribed to a change in the rate of gold ion reduction [52]. The success of intra-gap NPs synthesis based on nanorod and bipyramid cores is highly dependent on both the amount and ratio of HAuCl₄ and AgNO₃. Longitudinal surface plasmon resonance peaks (Figure S7(c–d) in the ESM) have decreased intensity and blue shift in wavelength compared with the spectra in Figure 1, indicating the formation of nanorod or bipyramid based intra-gap NPs. Different from nanorod/bipyramid based intra-gap NPs, we can easily make nanosphere based intra-gap NPs (Figure 3(g–j)) using various amounts and ratios of added HAuCl₄/AgNO₃ mixtures, in which the shells of Figure 3(j) with more petals are much thicker than Figure 3(g) and Figure 3(h) due to the higher amounts of HAuCl₄ and/or less AgNO₃ in the reaction solution. Evidence of successful preparation of sphere based intra-gap NPs here is primarily due to the appearance of the bimodal spectra in Figure S7(g–j) in the ESM.

Figure 3.

TEM images of nanoparticles with different cores after adding various amounts of AgNO₃ (10 mM) and HAuCl₄ (10 mM). Smaller nanorod cores: (a) 0 μL AgNO₃ and 200 μL HAuCl₄; (b) 50 μL AgNO₃ and 200 μL HAuCl₄; (c) 100 μL AgNO₃ and 200 μL HAuCl₄. Bipyramid cores: (d) 200 μL AgNO₃ and 100 μL HAuCl₄; (e) 200 μL AgNO₃ and 200 μL HAuCl₄; (f) 200 μL AgNO₃ and 300 μL HAuCl₄. Nanosphere cores: (g) 300 μL AgNO₃ and 300 μL HAuCl₄; (h) 500 μL AgNO₃ and 500 μL HAuCl₄; (j) 300 μL AgNO₃ and 500 μL HAuCl₄. Scale bars are all 100 nm, and their UV-Vis extinction spectra are shown in Figure S7 in the ESM.

The images in Figure 4(a–c) reveal the core-shell intra-gap structures of particles made with various cores. The corresponding EDS elemental maps in Figure 4(d–f) confirm that the cores primarily consist of Au while the shells are Au and Ag alloy. The replacement of sacrificial metal (Ag) by the higher reduction potential metal (Au) was observed due to the occurrence of galvanic replacement reactions, resulting in the formation of gaps inside the nanoparticles. As shown in Figure S8 in the ESM, while varying in places, these three intra-gap NPs have average gap sizes of ~2.45 nm (nanorod cores), ~2.57 nm (bipyramid cores), and ~2.21 nm (nanosphere cores), respectively.

Figure 4.

Counts per second images (left panel) and EDS elemental maps (right panel) of intra-gap nanoparticles with nanorod cores (a, d), bipyramid cores (b, e), and nanosphere cores (c, f). The green color corresponds to Au and red to Ag, respectively.

We further examined the distribution of the electromagnetic (EM) field located within the intra-gaps of these NPs using the Finite-Difference Time-Domain (FDTD) method. The EM field generated in the gaps is important for amplifying the Raman scattering signal. The EM field enhancement distributions excited by 633 nm (Figure 5(a–c)) of intra-gap nanoparticles with uniform nanogap sizes were simulated using the commercially available software FDTD Solution by Lumerical Inc. Detailed 3D models of these three gapping nanoparticles referenced for these simulations are illustrated in Figure S9 in the ESM. The SERS EF of 2.5 × 10⁶ (rod-based gapping NPs), 1.1 × 10⁴ (bipyramid-based gapping NPs), and 1.2 × 10⁶ (nanosphere-based gapping NPs) were also obtained from 3D-FDTD simulations. The wavelength range of the optical cross section monitor for each type of intra-gap NPs was set corresponding to the relevant LSPRs from their UV-Vis extinction spectra. Also, 400 sampling points were distributed uniformly in this range (Figure 5(d–f)), which accounts for the difference of simulated nanoparticle sizes and spectra from the experimental results.

Figure 5.

FDTD simulation of the electromagnetic field of intra-gap nanoparticles (a–c) and their corresponding extinction cross-sections (d–f). (a,d) Nanorod based intra-gap NPs; (b,e) Bipyramid based intra-gap NPs; (c,f) nanosphere based intra-gap NPs.

Understanding the enhanced Raman signal from these intra-gap nanoparticles is important for their application as a SERS sensor. Figure 6(a–c) shows the Raman spectra of intra-gap nanoparticles with different cores in solution recorded using a custom Snowy Range IM-52 spectrometer with a 638 nm laser. The particle concentrations were determined by nanoparticle tracking analysis (NTA) (Table S1). By comparing the intensity of the υ_NO2 stretching vibration of 4-NBT at 1334 cm⁻¹, the nanorod based intra-gap NPs exhibit the highest Raman signal in comparison with the other two NPs in Figure S10 in the ESM. The average intensity per particle for the nanorod intra-gap particles is considerably larger than other particles (e.g. – nanostars and gap-enhanced Raman tags) measured under comparable conditions and that have been used for cell imaging. [53] This trend is consistent with the nanorod-based particles having the best alignment between the 638 nm laser and their localized surface plasmon resonance. We also investigated the stability of these intra-gap NPs after three months storage shown in Figure 6(d–f). The TEM images show the intra-gaps still exist in the nanorod/bipyramid-based nanoparticles; however, the structure of the nanosphere based intra-gap nanoparticles shows significant changes in which the gaps are gradually shrinking or disappear. This is consistent with nanosphere based intra-gap nanoparticles of smaller sizes having a more negative surface free energy and decreased standard reduction potential [54]. Additionally, as shown in Figure S11 in the ESM, the stability of nanorod based intra-gap nanoparticles has been confirmed by the stable Raman intensity of characteristic peak (1334 cm⁻¹) of 4-NBT with a relative standard deviation (RSD) of 6.1% after three months storage at 5 °C.

Figure 6.

Raman spectra of intra-gap nanoparticles with different cores, and their respective TEM images after three months storage. (a,d) nanorod; (b,e) bipyramid; (c,f) nanosphere. Scale bars are all 100 nm.

Finally, The SERS EF for the nanorod based intra-gap nanoparticles is calculated to be 2.52 × 10⁶ approximately using the 1334 cm⁻¹ peak of 4-NBT (see supporting information for details). This experimental EF agrees with the EF in the 3D-FDTD simulation (2.509 × 10⁶). This enhanced is sufficient for detection in cellular imaging experiments. Figure S12 shows darkfield and Raman images from an individual cell incubated in the presence of the nanorod-based intra-gap particles. The NBT signal is clearly evident further supporting that the reporter molecule is not affected by environmental molecules. This proof-of-concept results supports further investigation with cells, such as monitoring cell secretion events, cell sorting and imaging, cancer diagnosis and treatment, etc.

3. Conclusion

We developed a straightforward approach to synthesize nanorod, bipyramid, and nanosphere based intra-gap nanoparticles by encapsulating 4-nitrobenzenethiol (4-NBT) as probe molecules in the gaps. This simplified method provides high-yield, stable intra-gap nanoparticles with a short incubation time (10 mins) and quick synthesis without requiring additional centrifugation to purify the core nanoparticles. The successful preparation of intra-gap NPs with nanorod and bipyramid cores is highly dependent on the amount and ratios of HAuCl₄ and AgNO₃ added to the shell growth solution. However, the preparation of intra-gap NPs with nanosphere cores depends on the relative amounts of HAuCl₄ and AgNO₃, which affects the shell thickness and shape. The nanorod/bipyramid intra-gap NPs also exhibit longer stability than the nanosphere intra-gap NPs. Notably, the nanorod based intra-gap NPs shows the highest Raman intensity with excellent stability, providing a competitive Raman tag for a wide range of applications in sensors. This method could be expanded to make a variety of intra-gap NPs with a wide range of cores such as stars, plates, or cubes.