The silver particles were synthesized with the terminated carboxylic ligands and displayed good solubility and chemical stability in water. TEM images showed the silver particles were approximately homogeneous in size distribution (inset of ). The average diameter of metal cores was detected to be 20 nm, and the chemical composition was estimated to be ca. (Ag)2.5×105(Tio)5.0×103.
Figure 1 Absorbance spectra of tiopronin-coated silver particles and the metal dimers that were formed by hybridization of double-length oligonucleotide with the bound oligonucleotide on the metal particle with the different metal core sizes. Transmission electron (more ...)
The ligand exchange is known to occur at a mole ratio of 1:1, so (2-mercapto-propionylamino) acetic acid 2,5-dioxopyrrolidin-1-ylester was co-dissolved with the silver particle at a mole ratio of 1/2 in water to ensure loading only one succinimidylated ligand per metal particle.16
At this molar ratio, even though the succinimidyl ligands were completely transferred onto the metal particles, at the most, there were only half of the metal particles that were single-succinimidylated. This strict limitation of a single succinimidyl ligand on the metal particle may guarantee that it is definitely bound by a single aminated oligonucleotide in the following surface reaction and then labeled by the single fluorophore through hybridization with the long-chain oligonucleotide. We did not separate the succinimidylated particles from the unmodified particles after the surface reactions because only the succinimidylated particle could be bound by the aminated DNA and then dimerized via the hybridization with the long-chain labeled DNA. Because both the unmodified particles and DNA are negatively charged on their surfaces, nonspecific interactions between them were supposed to occur and result in the aggregation of metal particles.
Absorbance spectra of silver particles were performed in 10 mM PBS buffer solution. It was shown that the tiopronin-coated silver particles displayed a metal plasmon absorbance at 401 nm (). The surface reactions including the ligand exchange and oligonucleotide binding did not significantly alter the absorbance spectrum of the metal particles, indicating that only a few ligands were involved. Simultaneously, the TEM images of metal particles showed no significant changes after these reactions occurred. As described above, the metal monomers and dimers were synthesized by hybridization of oligonucleotides 2 and 3, respectively, with the bound oligonucleotide 1 on the metal particles in 50 mM PBS buffer solution at pH = 7.2 ().16,22
The hybridization on the metal monomers did not alter the absorbance wavelength, but the hybridization on the metal dimers could result in a slight red-shifting of the metal plasmon absorbance to 403 nm ().27
This wavelength shift was smaller than the reported value, reflecting that only a portion of metal particles was dimerized by the hybridization. The formation of metal dimer was also verified clearly by the TEM images (inset b of ). The inset b of showed up the typical TEM images of formed dimers. Actually, many individual and few aggregated metal particles were also observed. When the TEM images were analyzed by at least 200 counting, it was shown that about 75% of particles exist as the monomers, 23% as the dimers, and less than 2% as the aggregates that contain more than two individuals. Because some dimer or aggregates were regarded to form occasionally when drying the solution on the grid, we thought that there were 30-40% of metal particles existing as the dimers formed by DNA hybridization. This ratio approximately consisted with the value obtained by the emission intensity
Ensemble emission spectra were first used to monitor the binding of fluorophores on the metal particles. Unbound Cy5-labeled oligonucleotide in 10 mM PBS buffer solution displayed an emission maximum at 661 nm upon excitation at 620 nm when measured in ensemble spectroscopy. The emission wavelength was slightly red-shifted to 663 nm when binding the fluorophore on either monomer or dimer metal particles, consistent with previous observations.28
We employed metal monomers to estimate the loading number of fluorophores per metal nanoparticle. Experimentally, a known concentration metal particle in buffer solution was added to a couple drops of 0.1 N NaCN aqueous solution.27
The metal cores were dissolved and the fluorophore-labeled oligonucleotides were released. The released fluorophores displayed an identical emission spectrum to the free oligonucleotide in the absence of metal, so the concentration of fluorophore was estimated by the emission intensity. The loading number was calculated to be ca. 0.3, indicating 30% of silver particles having one fluorophore as we required in this case.
We investigated the single-molecule spectral properties using the SMD technique on the labeled metal particles. The fluorescence was first examined for the fluorophores without the metal particles and on the metal monomers. Typical fluorescence images of a 5 μ
m × 5 μ
m region were recorded ().17
The recorded images provided initial and quantitative information on the brightness of single molecules. All image spots were circle-shaped, but the labeled metal monomers () were much brighter than the free fluorophores in the absence of metal (). Histograms of the intensity were constructed by counting more than 50 molecules. Only spots displaying single-step photobleaching were included in the analysis. The time profiles of traces were collected under the excitation (). Most traces showed clear one-step photobleaching, corresponding to the typical behavior from single fluorophores. The intensity was observed to be fairly constant until dropping abruptly to the background level in a single step. The fluorescence of single fluorophore on the metal monomers was enhanced 7-fold over the intensity of free Cy5-labeled oligonucleotide in the absence of metal. Because the scattering intensity from the unlabeled metal particles was less than 5% of the emission intensity from the free fluorophores, we attributed the emission enhancement to the metal-enhanced fluorescence on the metal monomers.
Figure 2 Respective fluorescence images of single-labeled (a) free Cy5, (b) single fluorophore on metal monomer, and (c) on metal dimer. The 5 μm × 5 μm images are 100 pixels × 100 pixels with in integration time of 0.6 ms per pixel. (more ...)
Respective time-traces of single Cy5-labeled oligonucleotide in the absence of metal, on metal monomer, and on metal dimer.
Next, we studied the fluorescence signal changes on the metal dimers. It was observed that, like the case of the metal monomers, the collected images on the metal dimers were also circle-shaped but became brighter (). The histogram of these labeled metal dimers, collected by an analogous method as the labeled metal monomers, showed a 13-fold intensity increase relative to the free fluorophores. This value was about double than the labeled metal monomers. To the best of our knowledge, this is the first experimental demonstration of the coupling effect between the metal particles that can lead to a more efficient MEF. The time trace profile of the labeled metal dimers showed a higher intensity scale than the metal monomers, and the emission persisted much longer than the free fluorophores in the absence of metal. This was also observed on immobilized silver particles as islands.14
MEF is known to occur via an increase of the intrinsic decay rate for a fluorophore near a metal surface.9a
As a result, the lifetime of the fluorophore should be shortened dramatically. This result is important evidence to confirm the metal-fluorophore near-field interaction. In this case, the lifetimes were derived from the time traces using time-corrected single-photon counting (TCSPC) technique and analyzed in terms of a single-exponential decay (). The lifetime of Cy5-labeled fluorophore was estimated to be 2.3 ns in the absence of metal, shortened to be 0.75 ns on the metal monomer, and furthermore altered to 0.61 ns on the metal dimer. The difference of lifetime value was small between the labeled monomer and dimer, which may be due to the limitations in the time resolution of our instrument. But we still can conclude that the trend observed in the shortening of lifetime was consistent with the intensity enhancement on the metal particles.
Emission decay curves fits for single Cy5-labeled oligonucleotide in the absence of metal, on metal monomer, and on metal dimer.
The lifetime of fluorophores near metal particles can influence their photostability.9b
A shorter lifetime results in a smaller time for photochemistry while in the excited state and thus more excitation-emission cycles prior to photobleaching. In this study, the time traces showed that both labeled metal monomers and dimers had emissions lasting for at least 20 s before bleaching, but the free fluorophore in the absence of metal had emissions lasting for only 2 s (), indicating that the photostability of fluorophore was extended 10-fold on the metal particles.
We know that MEF occurs via a near-field interaction of the fluorophore with the metal nanoparticle. The electric field near the metal particle is an important factor that influences MEF. These fields generated near a silver particle when under illumination from incident light can be calculated by the FDTD technique ().18-21
In this case, the hybridized DNA duplexes are regarded as rigid rods that separate the fluorophores and the surfaces of metal cores. The oligonucleotides on the metal monomer contain 23 base pairs that are about 8 nm long. So the MEF efficiency of the labeled metal monomers can be assumed to depend on the intensity distribution profile of the electric field at distance of 8 nm from the metal core. Our FDTD calculations reveal that the electric field intensity on the metal monomer decays significantly at this distance, so the MEF efficiency cannot be expected to be significant. However, our calculations also show that the electric field intensity in the space between the metal dimers is increased due to the coupling effect of metal particles, which can result in higher MEF efficiencies on the metal dimers. Although it is difficult to directly compare the experimental results of the metal monomer and dimer to the FDTD calculations in a quantitative manner, we still see qualitative agreements between our experimental observations and FDTD calculations. Of course, a complete numerical analysis would require further considerations of the field induced by the fluorophore.
Electric fields near silver (a) monomer and (b) dimer were calculated by FDTD model under an incident light of 635 nm. Note the incident light is propagating along the y-axis and is polarized along the x-axis.
In this paper, we studied MEF at the single-fluorophore level on silver monomers and dimers. The silver particles were fluorescently labeled by Cy5 via DNA hybridization. Single-molecule fluorescence images were recorded using scanning confocal microscopy. The emission image analysis of single fluorophores showed a 7-fold enhancement on the metal monomers and a 13-fold enhancement on the metal dimers when compared to the free fluorophore in the absence of metal. The lifetime results verified the near-field interaction mechanism of fluorophore with the metal particle. The electric fields near the metal monomer and dimer were calculated by the FDTD method and were further employed to explain the coupling effect between the metal particles. In subsequent studies, we will investigate the effect of different core sizes of silver particles and different distances between the coupled particles.