Solvent Effect on the Synthesis of Oleylamine Modified Au Nanoparticles and Their Self-Assembled Film for SERS Substrate

Abstract

The preparation and self-assembling of monodisperse gold nanoparticles (Au NPs) is of great significance for its SERS application. According to the oleylamine-reduced method, oleylamine (OAm) serves as the reducing agent and stabilizing agent, and the effects of different reaction parameters such as solvent and temperature on the size and dispersity of Au NPs have been evaluated. The Au NPs synthesized with toluene as the solvent have the best dispersity and narrowest particle size distribution with adjustable sizes. The particle size gradually increases with the increase in reaction temperature. The highly ordered self-assembly film of Au NPs was employed as surface-enhanced Raman scattering (SERS) substrate for the probing molecule of rhodamine 6G. The Au substrate exhibits excellent spatial uniformity and SERS reproducibility, which indicates its practicability as a substrate. This study provides a simple synthesis strategy of highly ordered monodispersed Au NPs, which can serve as a SERS substrate with excellent spatial uniformity and SERS re-producibility.

1. Introduction

Surface-enhanced Raman scattering (SERS) is a highly sensitive and potential spectral analysis method, which combines traditional Raman scattering and emerging nano-material science and technology [1,2,3,4,5]. It has high sensitivity and a good enhancement effect and can be widely used for single component detection and real-time detection in many fields [6,7,8]. The preparation of a SERS active substrate is a prerequisite for obtaining a higher Raman enhanced signal. The SERS substrate is of great significance to accelerate molecular adsorption and SERS sensitivity through electromagnetic and chemical mechanisms for the SERS substrate [9]. The stability and reproducibility of the SERS substrate directly affect SERS enhanced performance. Currently, it is still a key challenge to fabricate a highly active, stable, and reproducible SERS substrate. Consequently, it is critical to fabricate a SERS substrate with spatial uniformity and good repeatability.

In recent years, Au nanoparticles (Au NPs) as a kind of precious metal nanoparticles have caused a great research boom in the field of photonics, catalysis, bio-sensing and nanomedicine [10,11,12,13,14,15,16,17]. This is due to the strong localized surface plasmon resonance (LSPR) of Au. The source of LSPR is that the wavelength of the incident light source is resonantly coupled to the vibrational frequency of the free electrons on the Au atomic surface. More importantly, Au NPs are also widely applied in SERS substrates due to their LSPR, which makes the detection of an ultra-low concentration target molecules more sensitive [18,19,20,21,22]. These applications are all based on the successful preparation of monodisperse Au NPs and self-assembling. Conventionally, except when using the Faraday method via the sodium citrate reduction of HAuCl₄ and the Brust–Schiffrin method with thiol-ligands protection, a single-phase method is introduced for the better synthesis of monodisperse Au NPs [23,24,25]. It has been confirmed that the single-phase synthetic method can significantly decrease the size distribution. In general, the synthesis of monodisperse Au NPs is easily affected by various synthetic parameters, such as capping agent, temperature, reductant, and final treatment process [26,27,28,29,30,31]. These effects are significant. It is worth mentioning that there are many reports about the influence of capping agent, temperature, reductant, final treatment process and so on. For example, our group reported the influence of ethanol as an antisolvent for the postsynthesis purification of Au NPs synthesized with DDAB and alkylamine [30]. Lee’s group prepared relatively larger Au NPs and Au@SiO₂ core–shell structures to study the size-dependent plasmonic effect [28]. However, there are relatively few reports on the effect of different solvents on reaction results. In addition, amine-capped Au is susceptible to solvent. Oleylamine (OAm) has also been reported to act as a solvent, reducing agent and capping ligand for the synthesis of Au nanomaterials [32,33,34,35]. To the best of our knowledge, in Zheng’s work, OAm-capped monodisperse Au nanoparticles were successfully synthesized in different solvents (chain alkanes and simple aromatics) [36]. Therefore, the investigation of the solvent effect of the fabrication of amine-capped Au NPs is necessary and urgent for the stronger universality of the synthesis method.

The self-assembly of gold nanoparticles as a SERS substrate has a good surface-enhanced Raman scattering effect on rhodamine 6G (R6G) [37,38,39]. For example, Wu’s group prepared AuNP/graphene substrates, which were significantly enhanced with an improvement factor of 400% on Raman characteristics of R6G dyes compared to the case without graphene [37]. Traditional methods of nanoparticle self-assembly have poor stability and repeatability, restricting its wide application. Currently, it is a common method to modify the surface of Au nanoparticles with appropriate chemical ligands to mediate self-assembly. When metal is determined, the influence of the SERS enhanced effect is mainly on the particles morphology, size contribution, surface ligand and arrangement mode. Narrow particle size distribution is particularly important for nanoparticle self-assembly [40]. Therefore, it is essential to fabricate the monodisperse Au NPs with size control.

In the current research, we synthesized OAm-capped Au NPs with the single-phase method with different solvents and reaction temperature to explore solvent effects and size effects. In the synthesis, OAm acts as a surfactant and reductant, reducing HAuCl₄·4H₂O in an organic solvent. High-quality monodisperse Au NPs were obtained when toluene acts as a solvent. With the increase in temperature, the synthesis of Au NPs is more homogeneous and larger in size. Meanwhile, the size is not uniform when n-hexane, tetralin and diphenyl ether act as solvent. The results of ultraviolet-visible (UV-Vis) light spectra, transmission electron microscopy (TEM) and the X-ray diffraction (XRD) showed the consistency of these changes. SERS substrates were prepared by Murray’s method, which is a general method for growing a centimeter-scale uniform film that can be easily transferred to any substrate. The OAm-capped Au NPs also exhibited an outstanding surface-enhanced Raman scattering (SERS) effect on the probe molecule R6G. This study provides a simple synthesis strategy of highly ordered monodispersed Au NPs, which can serve as a SERS substrate with excellent spatial uniformity and SERS reproducibility.

2. Materials and Methods

2.1. Chemicals

All syntheses were carried out using commercially available regents. Chloroauricacid (HAuCl₄·4H₂O), oleylamine (OAm, C18 content: 80–90%, Aladdin), rhodamine 6G (R6G) and solvents including toluene, n-hexane, tetralin, diphenyl ether, acetone, and ethanol were obtained from Sinopharm Chemical Reagent Beijing Co., Ltd. (Shanghai, China) All chemicals were used without further purification.

2.2. Synthesis of Au Nanoparticles

HAuCl₄·4H₂O (50.0 mg 47.8%) and oleylamine (2.5 mL) were dissolved in a three-necked flask by sonication to form a reddish orange-colored solution. Then, 2.5 mL of solvent (toluene, n-hexane, tetralin and diphenyl ether) was injected into the precursor solution. The mixture was heated to 120 °C under magnetic stirring. The solution was kept at this temperature for 1 h and cooled down to room temperature. The color of the suspension changed from reddish orange to pale yellow, to nearly colorless and finally to wine red as time elapsed. The Au NPs were precipitated by adding acetone and isolated by centrifugation. The top supernatant layer with extra surfactant and impurities was then decanted from the vial, and the bottom precipitate was re-dispersed in toluene or hexane. The control experiment only added HAuCl₄·4H₂O (50.0 mg 47.8%) and oleylamine (OAm) (2.5 mL). We used the same method in toluene at 75 °C and 90 °C for 1 h to synthesize Au NPs.

2.3. Self-Assembly Au Nanoparticle Film for SERS Substrate

For the SERS measurements, Au NPs substrates were prepared by the reported procedure [41]. Briefly, after dropping ≈50 μL of Au NPs dispersions over the surface of diethylene glycol (DEG) in a Teflon well, with the volatilization of hexane, a solid film appears on the subphase surface of DEG. We used the silicon pellet as a substrate to transfer Au NPs, maintaining the substrate/subphase angle at 120° and then gently lifting [42]. In this way, centimeter-scale Au NPs can be readily transferred to the substrate after drying. We prepared monolayer films of well-ordered nanoparticles using the above-described technique. Then, 20 μL of different concentrations of ethanol solution of R6G was dropped over the surface of the Au substrate and dried at room temperature.

2.4. Instrumentation

The UV-vis absorption spectra of Au NPs diluted with toluene and recorded by a Perkin Elmer Lambda 40 P UV-vis spectrometer with a variable wavelength from 350 to 750 nm. Transmission electron microscopic (TEM) and high-resolution TEM (HRTEM) images were taken on an FEI Talos F200X field emission transmission electron microscope operating at 200 kV accelerating voltage. The samples for TEM measurements were prepared by dropping Au dispersions on 200 mesh carbon-coated copper grids. The average size and standard deviation of Au NPs were obtained by manually measuring more than 200 particles with the TEM image with Image Tool. Scanning electron microscopy (SEM) was carried out via a ZEISS SIGMA SEM field emission scanning electron microscope. The X-ray diffraction (XRD) test was performed on an X-ray diffractometer (Bruker, D8 Advance) with Cu radiation (Kα = 1.54059Å). The scanning rate is 0.02°/s, ranging from 30° to 90° of 2θ. Samples for SEM and XRD measurements were prepared by dropping Au dispersions on sheet glass and drying under ambient condition. Raman spectra were collected with a 633 nm excitation line using a Thermo DXR2 microscopy Raman microscope. The signal detection was achieved using a sensitive charge-coupled device array detector. The laser power at the sample position was about 20 mW. The data acquisition time was 10 s with two accumulations.

3. Results and Discussion

In this synthesis, monodisperse Au NPs were obtained by reducing HAuCl₄·4H₂O with OAm as the capping agent and reductant in toluene, n-hexane, tetralin and diphenyl ether solvents at 120 °C. The TEM and HRTEM images and the corresponding particle size distributions of OAm-capped Au NPs synthesized at 120 °C in toluene are shown in Figure 1. As shown in Figure 1a, the Au NPs are high-quality monodispersed, uniform, and even self-assembled into two-dimensional (2D) nanocrystal monolayers. The result might be attributed to the fact that Au NPs with a narrow size distribution could self-assemble into two-dimensional or three-dimensional superlattices. They are all nicely aligned in two-dimensional hexagonal closed packed arrays [43]. Therefore, this is another way of confirming the narrow size distribution of Au NPs. This phenomenon indicates that it has the potential to act as an active SERS substrate. In addition, Au NPs were obtained without any solvent. Its particle size distribution is not uniform. This indicates that the presence of solvent promotes the homogeneity and dispersion of nanoparticle morphology. The average size and size distribution of Au NPs was ascertained by manually measuring at least 200 particles with ImageJ software. In Figure 1b, the average size of Au NPs is (9.7 ± 0.6) nm, further indicating the narrow size distribution. It was observed that the lattice distance of the sample is 0.24 nm in Figure 1c, which was very close to the crystalline planes of the (111) families of the face-centered cubic structure of Au [44]. As shown in Figure 1d, a three-dimensional (3D) superlattice structure is formed when the concentration of gold nanoparticles reaches a certain value, and the second layer of particles is in the three-fold sites of the single-layer particles. As a result, the OAm-capped Au NPs are highly monodispersed, homogeneous, and have a narrow size distribution, which has the potential to be applied in a SERS substrate.

The UV-vis absorption spectra of Au NPs obtained with toluene are shown in Figure 2a. A single, narrow, symmetric absorption peak around 525 nm was observed in the Au NPs dispersions, originating from the strong and characteristic plasmon resonance absorbance of Au nanostructures in the visible to near infrared region. When the chemical bond between the NPs and the ligand is relatively weak, changes in the ligand chain length do not affect the position of the SPR band. The size and crystallinity of Au NPs were further characterized by XRD. Figure 2b shows that the four diffraction peaks can belong to the (111), (200), (220) and (311) lattice planes of face-centered cubic (fcc) Au, respectively, which is consistent with the HR-TEM result. The crystallite size calculated by the Scherrer formula (D = kλ/βcosθ, k is the Scherrer constant, λ is the X-ray wavelength, β is the full width at half maximum in radians, and 2θ is the diffraction angle) with a half-peak breadth of the (111) plane is 6.1 nm [45,46]. The grain size estimated from the XRD patterns using Scherrer’s equation is smaller than that measured from the TEM images, indicating that the Au NPs are polycrystalline according to the previous study [47].

To explore the influence of temperature on particle size, Au NPs in toluene at different temperature were further synthesized. Figure 3 shows the TEM and the corresponding particle size distributions of OAm-capped Au NPs synthesized in the toluene at different temperatures. As can be seen in Figure 3a,c, the relatively monodispersed and uniform Au NPs are worse than the sample synthesized at 120 °C (Figure 1). The average size is (6.4 ± 0.5) nm at 75 °C and (7.3 ± 0.8) nm at 90 °C by particle size statistical analysis in Figure 3b,d. Obviously, compared with the particle size of Au NPs at different temperatures, its size increases with the increase in reaction temperature. Their dispersion and uniformity become better. This is because the reduction rate of Au³⁺ and growth process were accelerated with the increase in reaction temperature, making Au NPs more uniform and larger in size. These results directly show that the particle sizes of Au NPs can be controlled by adjusting the reacted temperature.

To study the solvent effect on the synthesis of OAm-capped Au NPs, we further synthesized OAm-capped Au NPs using other different reaction solvents, such as n-hexane, tetralin and diphenyl ether. Using linear hydrocarbons as the solvent, such as n-hexane, as shown in Figure 4a, most of the Au NPs are relatively uniform compared to toluene as the solvent. Whereas n-hexane has a low boiling point, it is difficult to control the particle size through a higher temperature change. However, in Figure 4b,c, when tetralin and diphenyl ether were used as the solvent, synthesized Au NPs have a much broader size dispersity. In addition, non-spherical nanoparticles are present when tetralin was used as the solvent. According to the LaMer model, the formation of a monodisperse and homogeneous system requires a fast nucleation rate and slow growth rate [48]. It is known that HAuCl₄ is highly soluble in polar solvents [49]. Therefore, strong polar solvents such as toluene make the nucleation rate of nanoparticle fast and make the growth rate slow. This results in a uniform morphology and narrow size distribution of nanoparticles, which is in accordance with the TEM images and size contribution. In contrast, weak polar solvents slow down the nucleation rate of nanoparticles due to the poor solubility of gold salts. Higher solvent polarity leads to a sharp decrease in the energy of nanoparticles, which reduces the polydispersity of nanoparticles [50]. The conclusion is consistent with the previous literature reports. When the polarity of the solvent increases, the nucleation rate of nanoparticles becomes faster and the growth rate becomes slower. Therefore, the particle size distribution of the product becomes narrow, and the morphology becomes more uniform. This shows that the reaction solvents can significantly influence the quality of synthesized Au NPs.

To investigate the SERS performance of a probe molecule on the monodispersed Au NPs, Au NPs films were primarily prepared by the reported procedure of Murray’s method [41]. In Figure 5a, the SEM image of the Au NPs film reveals that it is very uniform and has no noticeable cracks and intervals, which can be applied in the SERS substrate. It is not like the superlattice with coffee rings assembled by the solvent evaporation method. As shown in Figure 5b, HR-SEM images of Au NPs films further presented the oriented arrangement and uniformity of Au NPs in a monolayer, which is consistent with the TEM results. The film thickness is continuous and uniform. Obviously, the film is ordered in a compact hexagonal network, which is due to the narrow size contribution of Au NPs. Particle–particle and particle–substrate interactions also play an important role in this oriented arrangement. These results suggested that the Au NPs films prepared by this method can be used as a SERS substrate.

The process of the synthesis and self-assembly of Au NPs is shown in Scheme 1. Firstly, Au NPs were prepared by the one-phase method with HAuCl₄·4H₂O as the metal precursor, OAm as the capping agent and reductant, and toluene as the solvent. The stability of the Au NPs dispersion is a combination of the Van der Waals attraction between the Au cores and the spatial repulsive forces between the ligand OAm shells [51]. OAm as a capping agent is used to mediate the growth of NC and stabilize the NPs in solution. According to the above TEM results, the average particle size of Au NPs is 9.7 nm, whereas the effective nanocrystal diameter should be the sum of the metal diameter and double the surface ligand thickness [52]. For the OAm-modified NPs, the thickness of OAm is assumed to be 1.8 nm [53]. Therefore, the effect of Au NPs is about 13.3 nm. By statistical analysis of TEM images, the distance between Au NPs is less than 3.6 nm (twice the length of OAm). The phenomenon indicates that overlapping and interlacing occurs between ligand and ligand. Then, the Au NPs film was obtained by self-assembling with Murray’s method. The mechanism is drying-mediated self-assembly of Au NPs on an immiscible liquid surface under ambient condition. The hexane dispersion of OAm-modified Au NPs was dropped and spread on the DEG surface in a Teflon wall. The driving force of Au NPs self-assembly is based on the on-covalent interactions, such as van der Waals and dipole coupling, combining with hard-sphere space-filling rules. During self-assembly, the head group is firmly attached to the nanoparticle surface, and the hydrophobic alkane chain is exposed to the air. In this procedure of self-assembly, the proper speed of solvent evaporation and the subphase are important for the formation of uniform Au NPs film. There are several advantages using the self-assembly for preparing Au NPs film. (1) Hexane will evaporate and the self-assembly process will finish in several minutes. (2) The Au NPs film formed on the surface of DEG can be easily transferred to silica or glass substrates for further applications. (3) The large-scale film of long-range ordered Au NPs could provide a repeatability of signals and be used as reliable chemical or biological substrates.

The self-assembled film made of plasmonic NPs is particularly interesting in SERS due to the two-dimensional (2D) hexagonal close-packed structure with several nanometers’ interparticle gaps. These gaps generated numerous uniform “hot spots” for SERS activity, high uniformity, and high stability. As illustrated in the above discussion, the present synthesis and self-assembly approach allowed us to prepare large area assemblies of monodispersed Au NPs, which can be used to investigate the SERS efficiency and uniformity on such substrates. The repeatability of Raman signals in SERS analysis can demonstrate the spatial uniformity of the substrate. Therefore, we selected a spot at 4 µm intervals and measured by the Raman-line mapping mode. The upper half of Figure 6 shows the SERS spectra (excited at 633 nm) of 1 × 10⁻⁵ M R6G on the Au NPs substrate. Compared with the normal Raman signals of R6G, it exhibits stronger SERS ability. The SERS enhancement factor of Au NPs film is 1.8 × 10⁶. It is obvious that there are five characteristic peaks at 1182 cm⁻¹, 1310 cm⁻¹, 1362 cm⁻¹, 1512 cm⁻¹ and 1650 cm⁻¹, which can be assigned to C-C stretching vibration, C-H bending vibration, C-C stretching vibration of oxanthracene ring, C-C stretching vibration of oxanthracene ring and C-C stretching vibration of oxanthracene ring of R6G, respectively [54,55]. The lower half of Figure 6 is the SERS contour from the line mapping of five spots, which was selected at 4 µm intervals. All SERS contours are consistent with the characteristic peaks of R6G. In addition, all SERS spectra have similar intensities on the same characteristic peak, which demonstrates the spatial uniformity of this substrate.

To further explore the repeatability of Raman signals of the Au substrate, we also selected 10 spots on the same silicon chip for testing and calculated the relative standard deviation (RSD) from the intensity of the Raman signal at 1511 cm⁻¹. For the same concentration of R6G, the intensity of Raman signals collected from all random spots on the same substrate was averaged. As shown in Figure 7, the RSD is 9.5, 4.1, 3.5 and 7.6, corresponding to different concentrations of R6G (0.005, 0.05, 0.1 and 0.5 mM), respectively. The repeatability of Raman signals and the results of RSD indicate the good SERS performance of the Au substrate.

In addition, the coefficient of determination (R²) of the calibration curve is also a key factor to evaluate the SERS performance. Figure 8a shows the SERS spectra of different concentrations of R6G as the model molecule using Au NPs film as the SERS substrate under 633 laser lines: 0.005, 0.05, 0.1 and 0.5 mM (a–d), respectively. The SERS intensity of the characteristic band of each of the R6G increased regularly with the R6G concentration. Figure 8b shows the scatter plots of intensity of SERS spectra at 1511 cm⁻¹ versus the concentrations of R6G. The inset shows the calibration curve of the intensity of SERS spectra versus the logarithm of the concentration of R6G, with each data point obtained from 10 randomly chosen spots on the same substrate. The intensity of Raman signals has very consistent results with relatively small error bars. From the calibration curve, the coefficient of determination (R²) was indicated as 0.999, which revealed that the Au NPs substrate can be used as a reliable SERS substrate for real applications. In conclusion, the highly monodispersed Au NPs can be applied in SERS substrate, which is highly ordered and homogeneous. The substrate exhibits outstanding spatial uniformity, repeatability of signals and the narrowest of errors. Therefore, the Au substrate has excellent SERS performance.

4. Conclusions

We successfully synthesized OAm-capped Au NPs by reducing HAuCl₄·4H₂O in different solvents including toluene, n-hexane, tetralin, and diphenyl ether. The reaction temperature and solvents can both affect the size and morphology of Au NPs. When the polarity of the solvent increases, the particle size distribution of the product becomes narrow, and the morphology becomes better. The Au nanoparticles synthesized with toluene have good dispersity and narrow particle size distribution. Its particle size can be controlled by changing the temperature. The Au nanoparticle substrate prepared by Murray’s method has good SERS performance. This study provides a simple strategy to obtain size-controllable high-quality monodispersed OAm-capped Au NPs. It is of great significance to exploit the practical SERS application of Au NPs.