Rapid Seedless Synthesis of Gold Nanoplates with Microscaled Edge Length in a High Yield and Their Application in SERS

We report a facile and reproducible approach toward rapid seedless synthesis of single crystalline gold nanoplates with edge length on the order of microns. The reaction is carried out by reducing gold ions with ascorbic acid in the presence of cetyltrimethylammonium bromide (CTAB). Reaction temperature and molar ratio of CTAB/Au are critical for the formation of gold nanoplates in a high yield, which are, respectively, optimized to be 85 °C and 6. The highest yield that can be achieved is 60 % at the optimized condition. The synthesis to achieve the microscaled gold nanoplates can be finished in less than 1 h under proper reaction conditions. Therefore, the reported synthesis approach is a time- and cost-effective one. The gold nanoplates were further employed as the surface-enhanced Raman scattering substrates and investigated individually. Interestingly, only those adsorbed with gold nanoparticles exhibit pronounced Raman signals of probe molecules, where a maximum enhancement factor of 1.7 × 107 was obtained. The obtained Raman enhancement can be ascribed to the plasmon coupling between the gold nanoplate and the nanoparticle adsorbed onto it. Electronic supplementary material The online version of this article (doi:10.1007/s40820-016-0092-6) contains supplementary material, which is available to authorized users.

The growth of gold nanoplates can be directed by either templates or capping agents. By providing constrictions in a 2D space or dimension, planar substrates [14] and interfaces in lamellar bilayer membranes [21] have been used as effective templates for growth of gold nanoplates. Alternatively, capping agents can preferentially adsorb on a specific surface of gold so that the adsorption of gold ions to this surface is blocked, and the growth is restricted on a planar direction. These agents should be surfactants [22], polymers [23][24][25][26], biomolecules [27], and halide ions [28,29]. Among them, cetyltrimethylammonium bromide (CTAB) is one of the most frequently used surfactants for the growth of gold nanoplates, which can be easily adsorbed onto the surface of gold through complexing with Electronic supplementary material The online version of this article (doi:10.1007/s40820-016-0092-6) contains supplementary material, which is available to authorized users. halide ions. For example, Mirkin group developed the seed-mediated growth of small gold nanoplates using CTAB as the capping agent [30]. Although the seed-mediated synthesis process effectively prohibits secondary nucleation and easily controls the size and shape of the final product, it involves multistep growth of seeds. To solve this problem, Huang group developed a seedless approach to synthesize gold nanoplates in the presence of CTAB via thermal reduction, where reaction solutions were preheated before they were mixed together to ensure the control of the size distribution [22,31]. However, long preheating time will result in higher time consumption for the process and prove to be cost ineffective. Recently, high-yield synthesis of gold nanoplates with submicron edge length was reported where iodide ions were used as both the capping and etching agents [28]. However, rapid synthesis of gold nanoplates in microscaled edge length with high yield, simplicity, and low-cost still remains challenging.
Herein, we report a facile and reproducible approach of rapid, seedless synthesis of single crystalline gold nanoplates with edge lengths in micron orders of magnitude. The reaction was carried out by reducing gold ions with ascorbic acid in the presence of CTAB. The reaction temperature and molar ratio of CTAB/Au on the products were examined in detail. The SERS properties of the assynthesized gold nanoplates were also investigated.

Synthesis of Gold Nanoplates
A typical synthesis procedure is as follows: first, 100 lL of 0.1 M HAuCl 4 was added into 3 mL of 0.02 M CTAB aqueous solution in a plastic tube, and the mixture was left undisturbed for several minutes. Then, 100 lL of 0.1 M AA was added to the mixture, followed by rapid inversion for 10 s. The resultant solution was immediately placed in a water bath of 85°C and kept undisturbed for about 1 h. The products were washed by centrifugation at 4000 rpm for 10 min and finally dispersed in deionized water.

Preparation of SERS Substrates
The SERS substrate was prepared as follows: gold nanoplate solution was drop-casted onto a clean silicon substrate. The substrate was rinsed and blown dry by nitrogen gas. Afterward, it was immersed into a solution of gold nanoparticles for several minutes, allowing deposition of gold nanoparticles on the gold nanoplates. After thoroughly rinsing with water for several times, it was immersed into a solution of 0.01 M Mph for 3 h. The substrate was carefully rinsed and blown dry by nitrogen gas before the SERS measurement. A cross-bar was finally marked on the substrate for locating the gold nanoplate and investigating them individually under the optical microscope and SEM.

Characterizations
The extinction spectra of the gold nanoplates were recorded using Agilent Cary 60 UV-Vis spectrophotometer using a cuvette having 0.5-cm path length. The morphology of the gold nanoplates was characterized by Hitachi S-4800 field emission and FEI Quanta 250 FEG SEMs. Powder X-ray diffraction (XRD) patterns were recorded on a Bruker D8 Advance powder X-ray diffractometer at a scanning rate of 4°min -1 , using Cu-Ka radiation (k = 1.54056 Å ). Raman scattering spectra were measured on a micro-Raman system (HR evolution 260, Horiba). The sample was excited at 633 nm and 4 mW in the Raman measurement. The Raman scattered light was collected using an Olympus objective (100 X, N.A. = 0.9, W.D. = 1 mm). Raman spectra were recorded using a grating of 600 lines per mm with an integration time of 15 s.

Results and Discussion
As shown in Fig. 1, the synthesis of the gold nanoplates is very straightforward. To be specific, aqueous solutions of HAuCl 4 and CTAB were first mixed and left undisturbed at room temperature for several minutes. The color of the mixture solution slowly changed to red brown owing to the formation of AuBr À 4 complex ions [25,32,33]. Afterward, a reducing reagent of AA was injected, which quickly makes the mixture to be colorless. The reaction solution was kept undisturbed in a water bath of 85°C for 1 h, and some precipitation of the resultant products was found at the bottom of the glass vial. Figure 2a shows the extinction spectra of the as-prepared gold nanoplates in an aqueous solution, and the inset shows the comparison of the colors of the solution before and after the synthesis reaction. The gold nanoplates exhibit a broad absorption band starting from 500 nm, which can be ascribed to the dipole and quadrupole plasmon resonances [25,34,35]. The morphologies of resultant products were examined by SEM (see Fig. 2b) and statistically analyzed, and the results are shown in Fig. S1 of Supporting Information (SI). Although a small amount of spherical nanoparticles were observed as well, the resultant plate-like products were mainly composed of 8.5 % triangular and 91.5 % equilateral hexagonal nanoplates with average edge lengths of 3.5 lm, and thicknesses of around 114 nm (obtained from the tilted SEM image shown in the inset), which strongly confirms the relatively monodisperse and uniform nanoplates.
The XRD pattern of gold nanoplates was recorded using a quartz substrate (Fig. 3). The diffraction peaks are assigned to (111), (200), (220), and (311) planes of facecentered-cubic structure of Au (PDF No. 04-0784). Note that the intensity of Au (111) peak is much stronger than those of (200), (220), and (311). The diffraction intensity ratio of (200)/(111) is extremely lower than the standard file (0.0051 versus 0.52). This XRD result clearly indicates that the as-synthesized gold naonoplates are single crystalline and possess a preferred plane of (111).
To achieve a high yield, the synthesis conditions that affect the growth of gold nanoplates were finely optimized.
It is known that the reaction temperature plays an important role in the formation of anisotropic nanostructures [36,37]. Therefore, in our experiment, the reaction temperature was varied from 25 to 95°C, as illustrated in Fig. 4. Figure 4a shows the morphologies of resultant products synthesized at room temperature (25°C) where most of them are identified as plate-and sphere-like particles. Both the yields and the edge lengths of the nanoplates start to increase when the temperature is increased. Figure 4b, c shows the products synthesized at 45 and 65°C, respectively. At 85°C, a large amount of hexagonal nanoplates with edge length of about *3.5 lm can be found where the highest yield of 60 % is obtained (Fig. 2b). The extinction spectrum suggests that increasing the reaction temperature further to 95°C leads to both inhomogeneous size distribution of the gold nanoplates and the decrease in the yield to 53 % (Fig. 4d). These results indicate that the most preferable reaction temperature for the high-yield synthesis of the gold nanoplates is 85°C. Extinction spectra of the Au products grown at different temperature are shown in Fig. S1, Supporting Information (SI). In contrast to the turbid appearance of the solutions obtained at 25 and  45°C, those at 65, 85, and 95°C exhibit features of plasmon resonances belonging to the gold nanoplates, which is consistent with the SEM images. The gold nanoplate solution obtained at 85°C shows the highest extinction intensity among the these three, suggesting the highest yield.
Besides the reaction temperature, the molar ratio of Br -/ Au has greatly significant influence on the formation of the gold nanoplates. Figure 5 shows the gold nanoplates synthesized at various Br -/Au ratios with the addition of different volumes of 0.2 M NaBr to the reaction solution.
The synthesis was carried out in a water bath of 85°C for 1 h. As shown in Fig. 5a, when molar ratio of Br -/Au is 1:1, the final products are almost nanoparticles, and no plate-like nanocrystals were found. Nanoplates start to appear in the resultant products when molar ratio of Br -/ Au is increased to 3:1 (Fig. 5b). When molar ratio of Br -/ Au is tuned to 6:1, nanoplates dominate in the resultant products (Fig. 2b). The amount of nanoplates start to decrease when molar ratio of Br -/Au is further increased (Fig. 5c, d). These results indicate that the Br -/Au molar ratio of 6:1 is the most preferable for the formation of nanoplates. Also, CTA ? as the capping agent is necessary in the synthesis, in which a proper concentration of 20 mM was used. The results were also confirmed by the extinction spectra (Fig. S1b). A prominent plasmon peak at around 550 nm can be found for the solutions obtained at Br -/Au ratios of 1:1 and 3:1, indicating the presence of considerable amounts of spherical gold nanoparticles in the solutions. This peak becomes less observable at Br -/Au ratios of 6:1, and the spectrum in the near infrared increases, suggesting high yield of nanoplates. The extinction intensity drops when the ratio is further increased to 12:1 and 30:1.
It is generally believed that the formation pathway of nanoplates is ''kinetically controlled'' along with the surfactant as capping agent or template-like CTAB and PVP [16,33]. On the basis of the experimental evidence, the reaction temperature of 85°C and the Br -/Au molar ratio of 6:1 are optimal conditions for obtaining high-yield gold nanoplates. To investigate this reaction route, a series of experimental investigations on the formation process were carried out through sampling gold nanoplates with the increasing reaction times. Figure 6 shows the sizes and shapes of gold nanoplates under different increasing reaction times. It can be found that small plate-like nuclei are formed in less than 5 min (Fig. 6a). They grow bigger into nanoplates with micrometer edge length within 30 min (Fig. 6b), and a high yield is obtained after 60 min (Fig. 6c). The experiment was performed at the preferable reaction temperature and concentration of Brand CTA ? . This result indicates that under the preferable reaction condition, a minimum 1-h reaction time is essential for the synthesis of gold nanoplates in a high yield.
Recently, many endeavors have been devoted to the fabrication of efficient SERS substrates. For example, the nanoscaled gaps provided by sandwiched structures can greatly augment the Raman signals [38][39][40][41][42]. Individual gold nanoplates, however, are rarely used as SERS substrates alone due to their flat surface. Here, we demonstrate that the gold nanoplates with adsorbed gold nanoparticles can be used as efficient SERS substrates for achieving high enhancement factors (EFs). The Raman measurements were performed under excitation of 633 nm where Mph was used as the probing molecule. The same objective was used for the excitation and collection of the Raman scattering light. Figure 7a, b shows a typical gold nanoplate with adsorbed gold nanoparticle. Under the optical microscope, the same nanoplate was located and positioned in the center of the optical view (Fig. 7c). The focal point of the excitation laser at 633 nm was thereafter positioned exactly on the nanoparticle (Fig. 7d). Raman responses from gold nanoplates with and without adsorbed nanoparticles are measured and compared (Fig. 8a). Raman spectra from the individual corresponding gold nanoplates are shown in Fig. 8b. The Raman signals from nanoplates without nanoparticles adsorbed are so weak that no peaks can be identified on the spectrum (cases 1 and 2). It is also the case when a nanoparticle is adsorbed onto the edge of the nanoplate (case 3). This result is related to the lack of hot spots no matter whatever the shapes of these nanoplates be. Interestingly, strong Raman signals are observable when some nanoparticles are adsorbed on the upper surface of the nanoplates (cases 4, 5, and 6), and the excitation laser beam is focused on the nanoparticles. In such cases, hot spots are formed in the gap between nanoparticles and nanoplates, which greatly enhances the Raman response because of the interparticle electromagnetic coupling [43][44][45]. Characteristic Raman peaks of Mph can be found on the measured spectra at 830, 1014, 1083, 1175, and are, respectively, assigned to C-H wagging, C-C bending, and C-S stretching, and 1175 and 1600 cm -1 are belong to C-H bending modes [43,46,47]. Note that the Raman spectra from the SERS substrates are slightly different from the bulk (Fig. 8b).

Conclusions
In summary, we successfully developed a simple but effective route to the synthesis of single crystalline gold nanoplates with edge length on the order of microns. Optimized reaction temperature and molar ratio of CTAB/ Au are found to be, respectively, 85°C and 6:1 for the formation of gold nanoplates in a high yield of 60 %. The synthesis to achieve the microscaled gold nanoplates can be finished in less than 1 h under proper reaction conditions. Therefore, the reported synthesis is time-and cost effective. The gold nanoplates were further employed as the SERS substrates and investigated individually. Interestingly, only those adsorbed with gold nanoparticles exhibit pronounced Raman signals of probe molecules, where a maximum enhancement factor of 1.7 9 10 7 was obtained. Our work demonstrated that a designed nanostructure consisting of a nanoplate adsorbed with a nanoparticle on its upper surface can be used as an efficient SERS substrate for reproducible enhancement.