Enhanced single-molecule detection on plasmonic nanostructures

High demand for ultrasensitive and fast analysis in biomedical research has stimulated efforts to develop novel detection methods and technologies. Emerging advances in nanotechnology, surface engineering, and optics open up opportunities to use near-field interactions and radically increase optical signal strengths. Resonant interactions of localized surface plasmons in particles with plasmon polaritons on metal surfaces yield significantly enhanced localized fields covering large surface areas. Excitation of fluorophores deposited onto these nanostructures provides unprecedented high fluorescence signals with multiple hotspots that show enhancements by three orders of magnitude. Such a platform can easily and repetitively be fabricated by simply pouring silver-colloid solution onto a metallic surface and drying. Slow drying allows formation of colloidal structures of various shapes and sizes over the conductive metal surface: see Figure 1 (bottom insert).

Figure 1. Experimental configuration for front-face measurements. The insert at bottom left shows the colloidal assembly and the corresponding zoomed-in atomic-force-microscopy image. The top insert shows the emission spectra (in arbitrary units, a.u.) of the deposits on self-assembled colloidal structures (SACS) and glass (green and blue, respectively). Alexa555: AlexaFluor 555 dye.

Here, we present a new, metal-enhanced fluorescence platform composed of self-assembled colloidal structures (SACS) on a thin metallic film that produces more than 1000× fluorescence enhancement, enabling single-molecule (SM) monitoring with laser powers well below 100nW. We used surface-deposited immunoglobulin G antibodies labeled with AlexaFluor 555 dye to test the signal enhancement. We deposited Alexa-labeled antibodies uniformly onto both glass and SACS surfaces with comparable density.¹We used a front-face configuration (see Figure 1) for macroscopic (ensemble) detection with simple optics (a setup that can also be used in sensing devices). For SM studies, we used an MT200 confocal-microscopy system from Picoquant. For excitation we used a 470nm pulsed-laser diode with a repetition rate of 20MHz and average power of up to 2mW.

Both steady-state intensities and fluorescence decay were measured using the front-face configuration and an FT200 fluorometer (Picoquant) equipped with a monochromator and a multichannel photomultiplier detector. The top insert in Figure 1 shows the emission spectra measured of deposits on glass (blue) and SACS (green). The large excitation spot size of ~1mm² ensures that the measured signal is an average that covers many structures. The detected signal does not change significantly when the spot is moved across the glass or SACS surface. The observed enhancement represents an average of five measurements. Figure 2 shows the intensity decays measured on both glass and SACS surfaces. The fluorescence lifetime on glass is 1.8ns while on SACS it drops to below 0.4ns, confirming a significant enhancement of the fluorescence signal due to an increase of the radiative rate of fluorophore.

Figure 2. Fluorescence intensity decays (in counts, Cnts) of AlexaFluor 555 dye measured on glass and SACS on a silver platform, and fit residuals. χ²_R: Reduced χ²value, an often-used goodness-of-fit parameter. τ_av: Average fluorescence lifetime. IRF: Instrumental response function.

With decreasing spot size the signal variation from spot to spot on the SACS surface increases significantly. Figure 3 shows the signal as measured with the diffraction-limited spot size using the confocal-microscopy system. The inserts in Figure 3 show images of 30×30μm² fields. The intensity (count rate) measured along the red line shown in Figure 3(a) is low and does not change much. The corresponding fluorescence lifetime (bottom right) oscillates around 1.5ns. The situation is drastically different for the SACS surface: see Figure 3(b). The signal is much stronger and changes significantly from one point to another. It is very interesting to compare the changes in intensity with those in the measured fluorescence lifetimes. They follow each other closely, but in opposite directions. For strong signals, the corresponding fluorescence lifetimes are shorter. High, thousand-fold intensity increases of multiple spots on SACS (compared to glass) indicate that both radiative-rate and excitation-field enhancements are of significant importance.

Figure 3. (a) Intensity (top) and fluorescence lifetime (bottom) for model assay with Alexa555 on glass. The image represents a 30×30μm²field. (b) Intensity and lifetime for assay deposited onto SACS. P: Laser power. The red arrows represent the positions of the traces shown in the graphs at right.

Figure 4. SM images of AlexaFluor 555 dye deposited onto SACS and glass, and recorded with different laser powers.

Next, we reduced the concentration (density) of deposited molecules to detect SMs. Images of SMs on glass—see Figure 4(a)—were collected with a 3μW excitation beam. For SACS surfaces—see Figure 4(c)—a laser power of 70nW already results in excellent SM spots. For comparison, 70nW applied to glass—see Figure 4(b)—does not show any fluorescent spots. To see comparable SM on glass, an 80× higher laser power is required. Under these conditions, we observed that the time required for bleaching (photodegradation) was significantly longer for SACS surfaces than for glass (see Figure 5), so typical single fluorophores could be observed for over 10×longer without applying any oxygen scavengers.

Figure 5. Photon counts for single-fluorophore-labeled antibodies on glass (top) and SACS surface (bottom).

Silver nanostructures deposited onto a silver layer produce enormous enhancements of fluorescence signals. The overall macroscopic enhancement is many hundreds of times, and at hot spots the SM signal increase can be greater than 1000×. This will enable SM studies with minimal laser power (<100nW) and make single-antibody immunoassay feasible, which is the focus of our future research directions.