Introduction

Gold nanoparticles (AuNPs) with various shapes, size, and organization show unique scattering spectra depending on their geometries (Hao et al. 2004; Xia et al. 2003; Creighton and Eadon 1991; Candreva et al. 2020). Spectroscopy and TEM are precious techniques to identify, characterize, and study nanoparticle features. Anisotropic nanoparticles show more than one plasmonic peak located in the visible and near-infrared region of the electromagnetic spectrum, associated with the various orientations of the particle axes relative to the electric field of light (Hao et al. 2004; Eustis and El-Sayed 2006; Burda et al. 2005; Teranishi et al. 2000; Yamada and Niidome 2006). In particular, for gold nanorods (AuNRs), the transverse oscillation is peaked at around 513 nm (Wu et al. 2016; Murphy et al. 2011), while the longitudinal resonance gives rise to a band whose position, ranging from 700 to 850 nm, depends on the aspect ratio and other factors, such as the covering agent (Pérez-Juste et al. 2005; Chen et al. 2013; Wang et al. 2013a; Wang et al. 2020; Wu and Tracy 2015; Mayer and Hafner 2011). Organized superstructures of gold nanorods show an optical appearance different from the isolated nanorods (Zhang et al. 2015; Vial et al. 2007; Lunn et al. 2015), and this behavior has been attributed to the plasmon coupling (Jain et al. 2006).

To spatially organize anisotropic gold nanoparticles, two strategies should in principle be followed: one based on the interparticle forces due to a proper functionalization of the AuNPs (direct assembly) (Ni et al. 2010; Kumar et al. 2013; Kawamura et al. 2007; Ahijado-Guzmán et al. 2016); the other based on the presence of an orienting medium (indirect assembly) (Coursault et al. 2012; Rožič et al. 2017). In any case, plasmonic coupling was found to be highly dependent on interparticle spacing and the effectiveness of the NP coating in modulating the plasmonic field. Coverage can modulate the plasmonic field in several ways: it can guide the assembly of nanoparticles in a precise structural order, (Jain et al. 2006) or avoid plasmon interaction (Comenge et al. 2016) by increasing the distance between nanoparticles. The effectiveness of plasmonic coupling increases as the distance between the nanorods decreases and the number of interacting NPs increases. (Kumar et al. 2013; Sahu and Raj n.d.).

In this perspective, by using a versatile coating agent such as silica, it is possible to adjust the maximum distance between the nanoparticles necessary for the plasmon interaction. The thickness of the silica coating that covers AuNPs can be adjusted with the consequence of modulating the distance between the assembled gold nanorods and controlling their plasmonic coupling properties. With this in mind, the silica coating strategy (Wu and Tracy 2015; Jiang et al. 2016; Gorelikov and Matsuura 2008; Liz-Marzán et al. 1996; Carrasco et al. 2016; Guerrero-Martínez et al. 2010; Abadeer et al. 2014; Pellas et al. 2020) consists of a pH-controlled condensation of tetraethylorthosilicate (TEOS) precursor onto CTAB-capped gold nanorods. Hydrolysis and condensation of TEOS require alkaline conditions, generally reached by adding to the sample ammonia or strong bases such as NaOH. CTAB surfactant deposited onto AuNRs is a structure-directing agent (Ming et al. 2009) that facilitates the formation of mesostructured silica shell around nanorods. The silica shell grows anisotropically on AuNRs; it is deposited first selectively at the two ends of the nanorods, due to the greater curvature, then grows over the entire surface of the nanoparticles (Adelt et al. 2021; Wang et al. 2013b).

The synthesis of silica-coated AuNRs requires fine-tuning the value of some parameters and the quantities of the reagents: here the individual contributions are discussed, estimating the quantity of each to obtain a homogeneous silica shell. Considering the reaction time and the ratio between the concentrations of the CTAB-coated AuNRs and TEOS, we adjusted the thickness of the silica around NRs. Using this cover agent as spacer between AuNRs, we found that the coupling of the plasmonic fields is effective when the distance between nanoparticles is not higher than 10 nm. When the distance is greater, no nanorod interact each other.

Materials and methods

Chemicals

Hexadecyltrimethylammonium bromide (CTAB, ≥96%), 5-bromosalicylic acid (technical grade, 90%), hydrogen tetrachloroaurate trihydrate (HAuCl4•H2O, ≥99.9%), silver nitrate (AgNO3, ≥99.0%), L-ascorbic acid (≥99%), sodium iodide 99.99%), trisodium citrate (99%), polyvinilpolypyrrolidone (PVP K15, Mn 10000), and sodium borohydride (NaBH4, 99%) were purchased from Aldrich and used as received. Milli-Q water (resistivity 18.2 MΩ•cm at 25 °C) was used in all experiments. All glasswares were washed with aqua regia, rinsed with water, sonicated threefold for 3 min with Milli-Q water, and dried before use. TEOS (Alfa Aesar, 99.9%), NaOH (Sigma Aldrich, 98%), and MeOH (Sigma Aldrich) were used for the SiO2 over coating.

Instruments and methods

A Perkin Elmer Lambda 900 spectrophotometer was employed to obtain the extinction spectra (Cretu et al. 2018; Pf et al. 2020; Ionescu et al. 2019). The size and morphology of the gold nanoparticles were measured using a transmission electron microscope (Jeol JEM-1400 Plus 120 kV). The samples for transmission electron microscopy (TEM) were prepared by depositing a drop of a diluted solution on 300 mesh copper grids. After evaporation of the solvent in air at room temperature, the particles were observed at an operating voltage of 80 kV.

Synthesis

An aqueous solution of CTAB (until a final concentration of 1.0E-3 M) has been added to 2 mL of an aqueous dispersion of AuNR@CTAB (0.3 E-3 M, A.R. 4) (Cretu et al. 2018; Pf et al. 2020). After few seconds, 20 μL of a 0.1 M NaOH aqueous solution were added under vigorous stirring, reaching a pH value of 8.5, followed by three additions, in intervals of 1 h from each other, each of 12 μL, of TEOS 20% v/v in methanol under gentle stirring, at room temperature. After 14 h, the mixture was centrifuged in MilliQ water twice at 6000 rpm for 10 min and the sample was dispersed in 2 mL of MilliQ water (Wu and Tracy 2015; Jiang et al. 2016). Undesired core-free silica particles can be eliminated by the appropriate centrifugation step, but it is not unusual that a small fraction remains in the final sample.

This sample made up of AuNRs coated by silica, AuNR@SiO2, (hereinafter referred to as the Starting Sample) and the protocol followed to obtain it, are the point of reference with which to compare the obtained results.

Results and discussion

Thickness control of silica shell and photophysical measurements of isolated gold nanorods

TEM image of the nanorods of the Starting Sample (Fig. 1a) shows the silica shell around nanorods with a thickness of 24 nm × 12 nm, long axis × short axis, respectively. Silica shell modifies the spectral position of the AuNRs longitudinal LSPR peak (Fig. 1b): the peak at 836 nm of AuNR@SiO2 solution is red-shifted compared to that of the AuNR@CTAB solution, positioned at 783 nm. This is due to the fact that the metal-dielectric interface in the case of AuNR@SiO2 is made up of silica which has a higher refractive index (nd20=1.45) than the AuNR@CTAB interface, which is actually made up of the water (nd20 =1.33) embedding CTAB. As shown below, as the thickness of the silica shell increases, the longitudinal band shifts towards red accordingly.

Fig. 1
figure 1

Starting sample. a TEM image of AuNR@SiO2. b Extinction spectra of AuNR@CTAB (red line; λLSPR = 783 nm) and AuNR@SiO2 (black line; λLSPR = 836 nm) in water solution

Full size image

Experimental parameters controlled during the silica coverage of gold nanorods include pH, concentration of the added CTAB solution, TEOS amount, concentration of the AuNR@CTAB solution, and the reaction time. Firstly, we have optimized the pH value and CTAB concentration, to obtain a detectable and homogeneous silica shell, and then we have studied the correlation between the [TEOS]/[AuNR@SiO2] ratio and the silica coverage.

Hydrolysis and condensation: pH value

The quantity of NaOH necessary to obtain the Starting Sample corresponds to a pH value of 8.5. By changing the amount of NaOH to measure a pH value of around 10, the gold nanorods precipitate to the bottom of the vials in seconds. By lowering the amount of NaOH to reach a final pH value of 8, no precipitation was recorded, but the TEM image (Fig. 1 in SI) showed the absence of an appreciable silica coating.

In the first case (pH=10), the alkalinity of the reaction environment causes a rapid TEOS polymerization creating a sort of cloud that incorporates a number of gold nanorods, triggering their precipitation. In the second case (pH=8), the alkalinity of the reaction environment is insufficient to prompt a good TEOS polymerization, so the silica coverage of nanorods is poor. In conclusion, pH=8.5 can be considered an optimal value.

Homogeneity of the silica shell: CTAB concentration

As previously reported, the first step of the silica shell growth is the addition of an aqueous solution of CTAB to an aqueous dispersion of AuNR@CTAB. In the Starting Sample (see experimental part), we used a solution of CTAB 1.0E-3 M, but we ascertained that using a CTAB concentration of 5.0E-4 M, the silica coverage appears to be unchanged (Fig. 2 in SI). By halving the CTAB concentration again down to 2.5E-4 M, we obtained a disappointing result as shown in the TEM image of Fig. 3 in SI, while reducing again the CTAB concentration down to 2.5E-5 M, after few hours from the latest TEOS addition, a burgundy cloud appeared in the solution which in a few seconds precipitated on the bottom of the vials.

Because CTAB molecules adsorbed on AuNRs act as templates for the growth of the silica shell, it is necessary to maintain this coverage during TEOS condensation; nevertheless, the addition of NaOH can dislodge CTAB molecules from the gold surface (on which they are held by electrostatic interactions) due to the presence of the Na+ hard ions.

Successively, we increased CTAB concentration up to 4.0E-3 M. TEM images (Fig. 4 in SI) show the expected AuNR@SiO2 nanoparticles, but the silica coverage is not uniform on the gold surface, and the size of AuNR@SiO2 appears inhomogeneous. These features can be attributed to the excessive CTAB concentration over the CMC (Scarabelli et al. 2015) that hampers the formation of the templating micelles, leading to a disordered silica growth.

Thickness control: TEOS concentration

The silica shell thickness can be modulated by varying the amount of TEOS. On this basis, we carry out various syntheses keeping constant the concentration of AuNR@CTAB (0.3E-3 M) and the other parameters (as reported in the paragraph Synthesis), and varying only the quantity of TEOS.

We tested four protocols with different amounts of TEOS, obtaining AuNR@SiO2 samples with different thicknesses of silica determined by TEM (Table 1): Figs. 2, 3, 4, and 5 report TEM images, Fig. 6 the extinction spectra of the samples.

Table 1 Thickness control: TEOS concentration
Full size table
Fig. 2
figure 2

TEM image from sample 1

Full size image
Fig. 3
figure 3

TEM image from sample 2

Full size image
Fig. 4
figure 4

TEM image from sample 3

Full size image
Fig. 5
figure 5

TEM image from sample 4

Full size image
Fig. 6
figure 6

Extinction spectra of the samples reported in Table 1: longitudinal LSPR peak of Starting Sample, 1, 2, 3 fall at the same wavelength (red line; λLSPR = 836 nm) because the silica thickness does not change; extinction spectrum of sample 4 (black line; λLSPR = 830 nm)

Full size image

The silica thickness is reduced by decreasing the quantity of TEOS, although a considerable reduction of TEOS does not correspond to a notable thickness reduction. The extinction spectra show the typical red-shift due to the presence silica cover. It is interesting to note that the magnitude of the redshift is reduced by decreasing the amount of TEOS; this is to be attributed to a greater porosity of the shell, which incorporates a greater quantity of water.

Thickness control: AuNR@CTAB concentration

To test the effect of the AuNR@CTAB concentration on silica thickness, we increased this concentration up to 3.0E-3 M (Starting Sample’ in Table 2). We estimated the AuNR concentration, according to Scarabelli et al., considering that all Au3+ has been reduced to Au0 (Scarabelli et al. 2015). By increasing the concentration of AuNR@CTAB, we correspondingly doubled the amount of NaOH to have a pH = 8.5. As expected, the thickness of the silica was significantly reduced compared to the products of the Starting Sample (12 nm × 7 nm vs 24 nm × 12 nm, respectively), because the amount of TEOS available for each AuNR is reduced.

Table 2 Thickness control: AuNR@CTAB concentration
Full size table

Taking into account that TEOS amount used to obtain sample 3 (Table 1) allows to attain the thinner silica thickness, we fixed a TEOS amount of 0.1E-4 mol (12 μL) (while Starting Sample sets a 0.3E-4 mol value). Therefore, we prepared two samples, 3′ and 3′′, with the same TEOS amount (0.1E-4 mol, 12 μL) but different AuNR@CTAB concentration (Table 2), observing a different silica thickness: Fig. 7 a, b, c report TEM images of the samples, Fig. 7d the extinction spectra. In particular, spectra show the usual red-shift of the longitudinal NR band substituting CTB coverage with silica one, which is reduced according to the thinning of silica thickness.

Fig. 7
figure 7

a TEM image from Starting Sample′. b TEM image from sample 3′. c TEM image from sample 3′′. d Extinction spectra of Starting Sample′ (red line; λLSPR = 824 nm), 3′ (black line; λLSPR = 805 nm), 3′′ (blue line; λLSPR = 800 nm)

Full size image

Thickness control: reaction time

Silica shell growth is rapid during the first hour, but slows down considerably during the second hour of reaction. In the Starting Sample, the reaction time (i.e., the time interval between the last addition of TEOS and the purification step) was set at 14 h. To study the effect of the reaction time variation on the silica thickness, we synthesized AuNR@SiO2 samples starting from the three recipes already illustrated (Starting Sample, 3 and 3′), and reducing the reaction time to 1h.

The obtained results are reported in Table 3 and TEM images in Figures 8, 9, and 10.

Table 3 Thickness control: reaction time
Full size table
Fig. 8
figure 8

TEM image of AuNR@SiO2 prepared from the Starting Sample by reducing the reaction time. Unreacted silica present even after centrifugation is visible

Full size image
Fig. 9
figure 9

TEM image of AuNR@SiO2 prepared according to the recipe used to prepare sample 3

Full size image
Fig. 10
figure 10

TEM image of AuNR@SiO2 prepared according to the recipe used to prepare sample 3′

Full size image

Photophysical measurement of assembled gold nanorods covered with silica

To explore the occurrence of the coupling among AuNR@SiO2, taking into account that the reciprocal distance between nanorods can be tuned by controlling the thickness of their silica shells, we compared the Starting Sample′, 3′,3′′. In particular, we compared extinction spectra from the water solution and from the film obtained by drop-casting of the solutions on quartz windows, thus, taking into account that in the solution there are no assemblies and therefore no coupling (Scarabelli et al. 2013), any difference should be attributable to the interaction of gold nanorods.

Observing the TEM image in Fig. 7a, we noticed the tendency of the nanoparticles to remain closed as much as possible, but no spectral variations were noted: the superimposition between the spectrum in solution and the spectrum of the film (Fig. 11) accounts for the absence of coupling. The small blue-shift of the longitudinal band of film sample compared to that in solution (hardly observable also in the previous sample), is attributed to the high sensitivity of this band to the refractive index variation on passing from water (nd20 =1.33) to air (nd20 = 1.00).

Fig. 11
figure 11

From left to right: extinction spectrum of Starting Sample′ from water solution (blue line), and from film sample (red line); extinction spectrum of 3′ from water solution (blue line, and from film sample (red line); extinction spectrum of 3′′ from water solution (blue line) and from film sample (red line)

Full size image

Sample 3′ 10 nm × 8 nm

Observing the TEM images of Fig. 7b, we noticed that, due to the reduced thickness of the silica, the distance between the gold cores appears reduced. In this case, the spectrum of the solution and the spectrum of the film (Fig. 11) are different; in particular, the longitudinal plasmonic bands of the film spectrum are red-shifted and broadened with respect to those of the spectrum in solution. This behavior is attributed to the coupling of the plasmonic fields, allowed by the reduced distance between the gold cores.

Sample 3′′ 5 nm × 3.5 nm

Looking at the TEM images in Fig. 7c, the thickness of the silica shell is so thin that it is barely visible. The longitudinal plasmon band of the film spectrum (Fig. 11), with respect to the spectrum in solution, is red-shifted, but the band broadening in the film spectrum is impressive. This behavior is attributed to the coupling of the plasmonic fields, allowed by the reduced distance between the gold cores of the AuNRs due to a very thin silica shell of 5 nm × 3.5 nm.

Conclusion

Various parameters control the growth and thickness of silica shell of gold nanorods, but the obtained results show that the impact of the parameters is different, and small variations of someone can produce very important effects. For example, pH and CTAB concentration are difficult to modulate, while TEOS concentration is not very sensitive, because, as reported in Table 1, important variations of this parameter do not cause relevant thickness variations; on the contrary, the amount of AuNR@CTAB seems to be more decisive for determining the thickness of the silica (Table 2). What appears to be decisive in controlling silica thickness is the [TEOS]/[AuNR@CTAB] ratio (Fig. 12), while to monitor the actual growth of silica on AuNRs, the most suitable observable is the volume of silica shell, calculated from TEM images.

Fig. 12
figure 12

Correlation between the silica shell volume of AuNR@SiO2 with [TEOS]/[AuNR@CTAB] ratio

Full size image

The thickness of the silica determines the position of the plasmon band of single and assembled gold nanorods. The effectiveness of the coupling depends on the mutual distance of the gold cores, which in turn depends on the thickness of the silica. In this perspective, using the silica shell as a spacer, it is possible to measure the edge-to-edge distance between the gold cores. By measuring the edge-to-edge distance between the gold cores of the examined samples, we found that it shortens by decreasing the silica thickness, thus allowing for better coupling. In particular, the coupling of the plasmonic fields is effective when the distance between the nanoparticles is not greater than 10 nm, that is, in samples 3′ and 3′′, while if the distance is greater, no nanorods interact with each other and no difference is visible between solution and film spectra. The use of silica as a covering agent is an interesting way to regulate the plasmonic properties of gold nanorods. The longitudinal band moves towards the near-infrared as a function of the thickness of the shell, which determines the interaction between the gold cores and consequently the occurrence of plasmonic coupling. Consequently, gold-core/silica-shell nanorods are of considerable interest for a wide range of possible fields of application.