$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
As proof of concept, DSNB was chosen as the analyte molecule and coated onto the surface of AgNPs. The typical SERS spectra of DSNB enhanced by the plasmonic AgNP assembly and dispersed AgNP are shown in Figure 6. Without the trapping laser, the dispersed AgNPs in the sample chamber generated a black spectrum (Figure 6A) upon excitation by the Raman probe laser. A weak and broad SERS signal was observed at approximately 1380-1450 cm-1, the characteristic peak of DSNB from its symmetric NO2 stretch, which is consistent with literature reports35,36. Since the dispersed AgNPs were under Brownian motion, the interparticle junctions were large and unstable, as illustrated in Figure 6C. Thus, the SERS signal amplification of DSNB was low for the dispersed AgNPs.
AgNPs are gathered to form a plasmonic AgNP assembly when the trapping laser is on. Increasing the power and extending the irradiation time of the trapping laser could attract more AgNPs and generate a dark spot, as shown in Figure 6B. Here, we applied a trapping laser power of 700 mM and a 20 s irradiation time to create a plasmonic AgNP assembly in a 0.05 nM DSNB-coated AgNP solution at a designated location and moment. The SERS spectrum of DSNB was obtained in the region of the plasmonic AgNP assembly (Figure 6A, red). The strong Raman band at 930 cm-1 is assigned to the nitro scissoring vibration, and the large bands at 1078 cm-1, 1152 cm-1, and 1191 cm-1 likely correspond to the succinimidyl N-C-O stretch overlapping with the aromatic ring modes of DSNB35,37. The feature bands at 1385 cm-1 and 1444 cm-1 arise from the symmetric nitro stretch of DSNB and are significantly enhanced and slightly shifted due to the reaction with the surface of AgNP35,37. Based on the previously reported SERS fingerprints of DSNB35,36,37, the band at 1579 cm-1 was assigned to the aromatic ring mode of DSNB. The overall intensities of DSNB in the plasmonic AgNP assembly were higher than those of the dispersed AgNP. Considering the intensity of the characteristic peak at 1444 cm-1, the plasmonic AgNP assembly can provide approximately a 50-fold enhancement of the SERS signal of DSNB compared to that of the dispersed AgNP. As shown in Figure 7, SERS spectra of DSNB were recorded repeatedly (20 times) for the AgNP assembly in the experiment, demonstrating identical vibrational features. The intensities of the characteristic peaks of DSNB at 1152 cm−1, 1444 cm−1, and 1579 cm−1 across these 20 SERS spectra were plotted as histograms with relative standard deviations (RSD) of 6.88%, 6.59%, and 5.48%, respectively. This further verified the reproducibility and stability. Hence, this approach is reliable for manipulating plasmonic nanoparticles and SERS detection of analyte molecules in solution.

Figure 1: Schematic representation of the optical tweezer-coupled Raman spectroscopic platform. Please click here to view a larger version of this figure.

Figure 2: Preparation of AgNP for SERS measurement. (A) SEM image of AgNP. (B) Size distribution of AgNP by DLS. Please click here to view a larger version of this figure.

Figure 3: Interaction of AgNP and DSNB. (A) Schematic of the coating of DSNB on the surface of AgNP. (B) UV-visible spectra of AgNP and AgNP-DSNB. Please click here to view a larger version of this figure.

Figure 4: Schematic of sample chamber preparation. (A) Sample chamber preparation process. (B) Prepared sample chamber. Scale bar = 1 cm. Please click here to view a larger version of this figure.

Figure 5: Position overlapping of 532 nm Raman laser and 1064 nm trapping laser. (A) Position of 532 nm Raman laser indicated by white spot. (B) Position of 1064 nm trapping laser indicated by red circle. Please click here to view a larger version of this figure.

Figure 6: Typical SERS spectra of the analyte molecules enhanced by the plasmonic AgNP assembly. (A) SERS spectra of DSNB at the plasmonic AgNP assembly (red) and the dispersed AgNP (black). (B) The plasmonic AgNP assembly when the trapping laser is on shows a dark spot under microscopic visualization. (C) The dispersed AgNP when the trapping laser is off. (D) Illustration of the mechanism of AgNP assembly formation. (E) Concentration-dependent SERS intensity in the absence of the trapping laser. Please click here to view a larger version of this figure.

Figure 7: Reproducibility of SERS signal of DSNB. (A) 20 SERS spectra of DSNB at the plasmonic AgNP assembly recorded repeatly in the experiment. (B) Histograms of the intensities of the characteristic DSNB peaks at 1152 cm-1 (RSD = 6.88%), 1444 cm-1 (RSD = 6.59%), and 1579 cm-1 (RSD = 5.48%). Please click here to view a larger version of this figure.

Figure 8: AgNP assembly generated under different experimental parameters. (A) Different trapping laser power; irradiation time 20 s and AgNP concentration 0.05 nM. (B) Different irradiation time; trapping laser power 700 mW and AgNP concentration 0.05 nM. (C) Different AgNP concentration; irradiation time 20 s and trapping laser power 700 mW. Please click here to view a larger version of this figure.
Supplementary Figure 1: The microscope camera images of AgNP assembly in time series when the trapping laser was turned off. Please click here to download this File.