Metal-Substrate-Mediated Plasmon Hybridization in a Nanoparticle

Metal-Substrate-Mediated Plasmon Hybridization in a Nanoparticle Dimer for Photoluminescence Line-Width Shrinking and Intensity Enhancement Guang-Can Li,† Yong-Liang Zhang,† Jing Jiang,‡ Yu Luo,‡ and Dang Yuan Lei*,† †

Department of Applied Physics, The Hong Kong Polytechnic University, Hong Kong, China School of Electrical & Electronic Engineering, Nanyang Technological University, Nanyang Avenue, 639798, Singapore

‡

S Supporting Information *

ABSTRACT: Metal-film-coupled nanoparticles with subnanometer particle−film gaps possess an ultrasmall mode volume, responsible for a variety of intriguing phenomena in plasmonic nanophotonics. Due to the large radiative loss associated with dipolar coupling, however, the plasmonic-film-coupled nanocavities usually feature a low-quality factor, setting an ultimate limit of the increased light−matter interaction strength. Here, we demonstrate a plasmonic nanocavity composed of a metal-film-coupled nanoparticle dimer, exhibiting a significantly improved quality factor. Compared to a silica-supported dimer, the spectral line width of the nanocavity plasmon resonance is reduced by a factor of ∼4.6 and is even smaller than its monomer counterpart (∼30% reduction). Comprehensive theoretical analyses reveal that this pronounced resonance narrowing effect can be attributed to intense film-mediated plasmon hybridization between the bonding dipolar and quadrupolar gap modes in the dimer. More importantly, the invoking of the dark quadrupole resonance leads to a giant photoluminescence intensity enhancement (∼200 times) and dramatic emission line-width narrowing (∼4.6 times), compared to the silica-supported dimer. The similar spectral characteristics of the measured plasmonic scattering and photoluminescence emission indicate that the radiative decay of the coupled plasmons in the nanocavity is the origin of the observed photoluminescence, consistent with a proposed phenomenological model. Numerical calculations show that the intensity enhancement is mainly contributed by the dimer−film gap rather than the interparticle gap. These findings not only shed more light on the hybridized interaction between plasmon modes but also deepen the understanding of photoluminescence emission in coupled plasmonic nanostructures. KEYWORDS: metal nanoparticle dimer, gap plasmon mode, line-width narrowing, plasmon hybridization, enhanced photoluminescence

M

interparticle gap distance still faces great challenges in planar composite nanostructures. More importantly, precise loading of active materials such as quantum emitters and two-dimensional (2D) materials into these nanoscale gap cavities, which is of crucial importance in cavity quantum electrodynamics14,15 (QED) and ultrasmall nanolaser technology,16 also remains largely hindered due to the in-plane arrangement. In these regards, metal-film-coupled plasmonic nanoparticles with vertical gaps seem to be an excellent candidate that has received intensive study in the past decade.17−22 Benefiting

etallic nanostructures with nanometer-scale gaps have received considerable research interest in the past decade. Due to the existence of strong capacitive electromagnetic interaction, these nanostructures often feature one or more near-field hot spots at their gap regions, making them promising platforms for surface-enhanced Raman scattering,1−4 sensitive biosensing,5−7 surface-enhanced photoemission,8−10 and nonlinear optics.11−13 So far, most of such composite plasmonic nanostructures have been fabricated either with lithographic methods, which limit the smallest achievable gap size to around sub-10 nm, or by self-assembling of metal nanoparticles in a planar arrangement. Although subnanometer gaps can be realized with the latter approach, achieving tunable optical response by readily controlling © XXXX American Chemical Society

Received: January 4, 2017 Accepted: March 14, 2017 Published: March 14, 2017 A

DOI: 10.1021/acsnano.7b00048 ACS Nano XXXX, XXX, XXX−XXX

Article

www.acsnano.org

Article

ACS Nano

Figure 1. (a) Schematic of two CTAB-coated (light blue) gold nanosphere dimers positioned respectively on a thin gold film (yellow) and on a glass substrate (light gray). (b) Dark-field images of individual dimers and monomers on a 70-nm-thick gold film (upper left) and on a glass substrate (upper right). The SEM images of the nanoparticles labeled in the dark-field images (a red circle encloses the dimer and a green circle encloses the monomer) are presented in the lower panel. The scale bar is 3 μm in the dark-field image and 200 nm in the SEM image. (c, d) Measured polarization-resolved scattering spectra of the gold nanosphere (c) monomers and (d) dimers on the gold film (orange) and on silica (light blue). The insets depict schematically the excitation configurations used in the respective measurements. The sold lines are the fits with a Lorentz function. The spectra in (c) and (d) are offset for better comparison of their resonance line-width characteristics.

from the well-established technologies for thin film deposition and spacer fabrication,23,24 the particle−film gap distance can be precisely controlled for realizing tunable plasmonic properties25 and also be scaled down to the sub-nanometer regime for addressing nonlocal effects26 and quantum tunneling.27 Moreover, the ultrasmall particle−film gap nanocavities can largely facilitate the accessibility of various active monolayer molecules or 2D materials,28 forming a versatile platform responsible for a series of breakthroughs in nanophotonic applications such as plasmonic nanolasers at deep subwavelength scale,29 realization of huge spontaneous emission enhancement,30,31 and, more recently, single-molecule-light strong coupling at room temperature32 and single-molecule optomechanics in “picocavities”.33 Up until now, most of the studies on such plasmonic nanocavities have been concentrated on the metal-film-coupled nanoparticle monomer configuration. In this configuration, the dipolar gap plasmon resonance possesses an ultrasmall mode volume (V), which dominantly contributes to the enhanced light−matter interaction in the aforementioned applications.30−32,34 However, similar to other coupled plasmonic systems,35,36 the internal capacitive interaction in the nanocavity results in increased radiative losses and consequently broadens the line width of the gap plasmon resonance. This significantly reduces the quality factor (Q-factor) of the

nanocavity, which sets an ultimate limit of the enhanced light−matter interaction strength such as the spontaneous emission enhancement of an emitter embedded in the cavity because of η ≈ Q/V,31,32 where η is the emitter−cavity coupling efficiency. Toward improving the quality factor of a plasmonic particle− film nanocavity, Sobhani and Halas et al. recently demonstrated that the spectral line width of the plasmon resonance in an aluminum nanoparticle monomer can be narrowed by a factor of 2 through near-field coupling with an underlying aluminum thin film,37 where symmetry breaking induces plasmon hybridization between the dipolar and quadrupolar modes. This experimental realization provides the metal-film-coupled nanoparticle system a fascinating feature: increased Q-factor in addition to its intrinsic ultralow mode volume. They also pointed out that this mechanism appears to be general to metalfilm-coupled nanoparticles but far less effective for gold-based systems due to their lower plasma frequency. Here we experimentally demonstrated that this is not necessarily the case for the gold-film-coupled nanoparticle dimer. While placing a metal nanoparticle close to one on silica forms a nanoparticle dimer with a significantly broadened bonding dipolar plasmon resonance, we have found that replacing the silica substrate with a gold thin film results in pronounced line-width shrinking of B


Article

ACS Nano

Figure 2. (a) Measured scattering intensity profiles for the silica-supported (light blue) and gold-film-coupled (yellow) dimers in the far-field imaging plane. The data points, corresponding to the intensity values along the dashed line in the dark-field images, are extracted from the red channel of the detector that mainly covers the radiation of the plasmon band centered at ∼650 nm. The sold lines are the fits with a Gaussian function. (b, c) Calculated far-field scattering patterns at the plasmon resonance wavelength of 650 nm for the silica-supported (b) and goldfilm-coupled (c) dimers.

colors remain similar to the gold film case. Figure 1c,d show the measured scattering spectra of two monomers and dimers on gold and silica under incident polarization along the dimer axis. This incident polarization can preferentially excite the bonding dipolar plasmon resonance without invoking other plasmon modes in the two dimers. The silica-supported dimer exhibits a scattering peak at around 650 nm (see Figure 1d), which is broader than and red-shifted compared to the monomer counterpart (see Figure 1c), simply due to the strong dipolar plasmon coupling in the gap of the dimer. Upon being positioned on the gold film, the scattering peak of the monomer shows a noticeable line-width narrowing, which is consistent with the previous observations for a similar system by Halas37 and Smith38 and their co-workers. Surprisingly, we observe from Figure 1d that placing the dimer on the gold film leads to more pronounced line-width shrinking than in the monomer case. Note that the scattering spectra for the monomer and dimer on gold (yellow curves in Figure 1c,d) reproduce the same spectral features as observed in the previous study.22 To quantify the resonance line-width narrowing, we determine accurately the full-width at half-maximum (fwhm) of each scattering peak by fitting the peak with a Lorentz function. The fitting results show that, by replacing the silica substrate with the gold film, the fwhm of the 650 nm scattering peak in the dimer decreases from ∼209 nm to 45 nm, narrowed by a factor of ∼4.6 (4.4 for fwhm in eV). Accordingly, this corresponds to an increase of ∼4.9 times in the quality factor for the gold-film-coupled nanosphere dimer (Q-factor: ∼14.4) compared with the silica-supported counterpart (Q-factor: ∼2.9). However, this line-width shrinking effect is less effective for the monomer case, where the fwhm of the 550 nm scattering peak decreases from 126 nm to 65 nm (narrowed by a factor of 1.9). In other words, although the dipolar plasmon resonance of the nanosphere monomer on silica is spectrally broadened upon coupling to a second nanosphere nearby to form the dimer (fwhm increased from 65 nm to 209 nm), the introduction of the metal film can remarkably narrow the spectral line width of the new resonance from 209 nm to 45 nm. An interesting observation here is that, by placing the second gold nanosphere near the film-coupled monomer, the

the dimer plasmon resonance by a factor of ∼4.6, leading to an extremely narrow line width of ∼45 nm, fully consistent with our theoretical analyses based on the multipole expansion model and full-wave electromagnetic simulations, which all reveal the intense plasmon hybridization of the dipolar and quadrupolar modes, being responsible for the significant plasmon line-width narrowing. The improved quality factor, together with the nanoscale mode volume, makes the gold-filmcoupled dimer an excellent plasmonic nanocavity with an ultrahigh photoluminescence (PL) yield (∼200 times enhancement compared to the silica-supported dimer) and a significantly narrowed emission line width inherited from the reduced radiative damping of the hybrid plasmon resonance.

RESULTS AND DISCUSSION Preparation of Gold-Film-Coupled and Silica-Supported Gold Nanosphere Dimers. In our study, gold nanosphere dimers on a thin gold film and a silica substrate, as schematically depicted in Figure 1a, were prepared through a simple drop-casting method. The diameter of the nanospheres is ∼100 nm, and the thickness of the gold film is ∼70 nm as measured by AFM mapping. Due to the presence of a surfactant polymer coatingcetyltrimethylammonium bromide (CTAB)on the gold nanospheres, the gap distance between the gold film and the nanospheres (between the nanospheres in the dimer) is estimated to be about 1 nm (2 nm). More details about sample fabrication can be found in the Methods section. Although we made no additional efforts to enrich the production yield, a sufficient number of dimers assembled on the gold film were still identified with the aid of a home-built dark-field microscope (Figure S1). Metal-Substrate-Induced Plasmon Resonance LineWidth Narrowing. Figure 1b shows the dark-field images of the nanoparticles on the gold film and on the silica substrate (upper panel), along with the corresponding scanning electron microscopy (SEM) images of the individual nanostructures labeled in the dark-field images (lower panel). One can see that on the gold film the nanosphere dimer (monomer) produces bright far-field radiation in orange (green). In sharp contrast, the scattering intensities of both dimer and monomer on the glass substrate diminish significantly, although the radiation C


Article

ACS Nano

Figure 3. (a, b) Simulated scattering spectra of the gold nanosphere monomers (a) and dimers (b) deposited on the gold film (orange dots) and on the silica substrate (light blue dots), respectively. The solid lines are the fits with a Lorentz function. The excitation conditions are the same as the experiment as depicted in Figure 1c. (c) Surface charge distribution profile and deduced charge dipole interaction representation for the silica-supported dimer (left column) and the gold-film-coupled dimer (right column) at their scattering peak wavelengths extracted from (b). Note that all the charge distribution profiles are viewed in the substrate plane (XY plane). The dimers are tilted by an angle of 45° with respect to the substrate normal for better visualization of the charge distribution within the nanosphere−substrate gap. (d) Multipolar expansion analysis of the respective contribution from the electric dipole (ED) and quadrupole (EQ) to the total scattering cross sections of the two systems studied in (b). The insets depict the bonding dipolar mode (red) and the quadrupolar mode (blue).

scattering peak fwhm of the newly formed film-coupled dimer can further be narrowed from 65 nm to 45 nm. This can be largely attributed to the suppressed nonradiative bulk damping associated with the interband optical transitions of gold at the plasmon resonance of the film-coupled dimer (∼650 nm) in comparison with the resonance of the film-coupled monomer (∼550 nm). Considering the dominant role of dipolar radiation damping responsible for the resonance line-width broadening in typical plasmonic systems with relatively large sizes,39−41 however, the remarkable line-width narrowing effect in the gold-film-coupled dimer, in comparison with the silicasupported dimer, implies a significant reduction in radiation losses when brought in close proximity to the metal film. In addition to the resonance line-width shrinking effect as discussed above, the light-scattering intensity of the gold-filmcoupled dimer is significantly increased, in comparison with the silica case, by 3-fold, as shown in Figure 2a. It has been proposed that the field reflected by the metal film can add to the incident field, thereby causing a stronger scattering efficiency for the metal-film-coupled nanoparticle.37 More generally, the effects of metal and dielectric substrates on the far-field radiation distribution of emitters placed nearby cannot be neglected.42,43 Here we numerically simulate the far-field radiation pattern at the scattering peak wavelength of 650 nm for the dimers on silica and gold. The simulation model used in

our calculations is shown in Figure S2 in the Supporting Information. As can be seen from the results in Figure 2b, the silica-supported dimer mainly radiates to the substrate side, leaving only a small amount of the scattered light within the collection angle of the objective lens atop the sample. In contrast, almost all the far-field radiation of the gold-filmcoupled dimer is injected into the superstrate side as shown in Figure 2c, and thus the majority of the scattered light falls into the collection angle of the objective, resulting in the much brighter radiation spots than the former case as compared in the dark-field images of Figure 1b. Metal-Substrate-Mediated Dipolar−Quadrupolar Plasmon Hybridization. To uncover the physical mechanism responsible for the drastic resonance narrowing observed above, here we combine full-wave electromagnetic simulations with an analytical multipole expansion model to calculate the near-field distribution and far-field response of the gold-filmcoupled and silica-supported nanosphere dimers, respectively. In addition, a conformal transformation optics scheme is developed to provide a qualitative interpretation of plasmon hybridization in the metal-film-coupled nanosphere dimer system. As can be seen from Figure 3a,b, the simulated scattering spectra of the monomers and dimers on gold and silica under the same excitation condition as the experiment (depicted in D


Article

ACS Nano

an admixture of highly radiative (or bright) dipole and nonradiative (or dark) higher order modes. The dark mode, when coupled to the bright mode, will acquire part of the energy from the bright mode, thereby reducing the radiation loss of the system.54,55 As mentioned earlier, this concept has recently been verified by Sobhani et al. by realizing significant plasmon resonance line-width narrowing in an aluminum-filmcoupled aluminum nanosphere through dipole−quadrupole mode hybridization.37 Although the semianalytical model used in their study can well explain the observed line-width shrinking effect and predict less effective line-width narrowing for a goldfilm-coupled nanosphere monomer due to lower plasma frequency, it cannot be directly applied to our metal-filmcoupled nanosphere dimer system due to the reduced insubstrate-plane symmetry in the presence of the horizontal dipolar bonding (note that the metal-film-coupled monomer system possesses a rotational or spherical symmetry in the substrate plane). To uncover the mode-hybridization-induced line-width narrowing in the film-coupled dimer, here we adopt a multipolar expansion model with which the respective radiation contribution of the electric dipolar and the multipolar moment to the total far-field scattering intensity can be distinguished. The far-field radiation that originates from the induced displacement current in a nanosphere can be described in terms of a vector potential A(r), which, in the quasi-static approximation, could be expanded into multipoles as56

the inset of Figure 1c) show good agreement with the measured spectra by reproducing the scattering peak positions and line-width shrinking of the plasmon resonance in the presence of the gold film. For simplicity, the CTAB surfactant shell was modeled as air in the simulation, leading to a slight blue-shift in the simulated scattering peaks compared to the measured ones. It is worth noting that the simulated reduction factors of the plasmon resonance line width for both monomers and dimers are smaller than the experimentally determined ones. This slight discrepancy can be attributed to the ideal geometric parameters of the nanostructures used in the simulations (such as gap distance, local structural feature, and the removal of the dielectric shell). Without loss of generality, the simplified particle−film systems can be used as a representative model to explore the mechanisms of the linewidth narrowing effect. Figure 3c presents the charge distribution profiles at the scattering peaks identified from Figure 3b, i.e., 648 nm for the silica-supported dimer (left) and 635 nm for the gold-film-coupled dimer (right). The left column shows that the induced surface charges are mainly concentrated within the lateral sphere−sphere gap in the silicasupported dimer, resembling those of the bonding dipolar mode of a nanosphere dimer in a homogeneous medium, except the negligible induced charges at the vertical sphere− silica gaps. In sharp contrast, the gold-film-coupled dimer presents a distorted charge profile with surface charges accumulated within both lateral sphere−sphere and vertical sphere−film gap junctions, as can be seen from the right column.22 To have an intuitive understanding of the complicated electromagnetic coupling in the dimer systems, the plasmon hybridization model developed by Nordlander et al. is applied to reveal the underlying physics from the perspective of plasmonic dipole−dipole interaction.44−48 On the basis of the simulated charge profiles shown in Figure 3c, the plasmon interaction in both dimers can be schematically represented with charged dipoles of different orientation and strength, as shown in the lower part in each column of Figure 3c. In the silica-supported dimer, the surface charge density is weakly screened by the silica substrate, thereby breaking the symmetry of the charge distribution along the substrate norm and giving rise to a tilted bonding dipolar mode on the dimer. Due to the relatively small electric permittivity of silica, the magnitude of the induced charge in the substrate is so weak that its screening effect on the bonding dipole mode is negligible, leaving the broad scattering peak observed in both simulation and experiment. For the gold-film-coupled dimer, however, the magnitude of the induced image charges can be very strong due to the presence of a sea of conduction electrons in the gold film, resulting in the strongly distorted dipolar alignment on the dimer, which can be decomposed as a horizontally bonding dipolar mode and a quadrupolar mode (formed by two vertically bonding dipolar modes) as illustrated in the insets of Figure 3d. The strength of the two induced vertical dipoles in the quadrupolar mode on the dimer can be comparable to that of the two horizontal bonding dipoles, thereby allowing for intensive plasmon hybridization between the quadrupolar and dipolar modes. Up until now, many studies have demonstrated line-width shaping of plasmon resonances in metal nanostructures through the concept of mode hybridization.49−53 Of particular interest is to relax the requirement of the dipolar coupling selection rules by reducing the symmetry of a plasmonic system, resulting in

A(r ) = −iωμ0 P

iωμ0 eik 0r e ik 0r + Q ·∇ 4πr 2 4πr

(1)

where P and Q are the electric dipolar and quadrupolar moments, respectively, under the e−iωt time conversion. Higher order terms in the multipole expansion can be neglected due to the subwavelength size of the constituent nanospheres studied in our experiment. The contribution from each term is displayed on the same scale in Figure 3d. While the electric dipole dominantly contributes to the total far-field scattering intensity for the gold-film-coupled dimer, the contribution of the quadrupolar moment is considerable due to the energy injection from the dipolar mode through mode hybridization. As a result, the larger quality factor of the quadrupolar mode increases the dephasing time of the hybridized plasmon and thus dramatically reduces the radiation loss of the dipolar mode, leading to the significant line-width shrinking effect. Comparatively, in the silica-supported dimer, the quadrupole contribution is almost zero, leaving the broad dipolar plasmon resonance nearly unaffected and dominant in the total scattering spectrum. Additionally, we also simulate the gold film thickness dependent scattering spectral response for the gold-film-coupled nanosphere dimer. One can see from Figure S3 that the plasmon resonance of the system shows a continuous red-shift in peak position and line-width broadening with decreasing film thickness. A detailed discussion can be found in the Supporting Information. For an intuitive analytical understanding of the observation in the gold-film-coupled dimer system, we can deploy the theory of two-dimensional conformal transformation optics to relate the original system to a simpler one with higher symmetry, following the approaches developed in our earlier works.57−60 The inversion transformation defined by

ξ′ = E

g2 ξ − R0

(2) DOI: 10.1021/acsnano.7b00048 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Figure 4. (a, b) Schematics of the two-dimensional conformal transformation from a nanowire dimer on a substrate to an annular system with an extra nanowire, through performing an inversion transformation. (c, d) Real part of the magnetic field in the (c) physical and (d) transformed frames at a wavelength of 500 nm.

with ξ = x + iy and ξ′ = x′ + iy′, transforms a film-coupled nanowire dimer system in the physical frame into an annular system with an extra nanowire corresponding to the semiinfinite substrate, as shown in Figure 4a,b. Points at infinity are mapped to the inversion point, R0, and the incident fields are converted into a dipole source at the inversion point. When the dimer is placed on a gold substrate, compared with the dimer on a silica substrate, the surface plasmons oscillating at the metal surface create a stronger substrate effect. The plasmonic interaction drags down the charges and confines them in the narrow gap between the dimer and the substrate. The effective dipole moment of the dimer is therefore decreased due to the distraction of charges to the dimer−substrate gap, which is confirmed by the contour plot of the calculated magnetic field (representing the distribution of induced surface charges; see Figure 4c). Consequently, optical radiation into the far-field regime is suppressed and the quality factor of the scattering resonance becomes larger than that of the dimer on silica. Alternatively, if viewed from the transformed frame, the nearfield energy is mainly distributed along the horizontal direction due to the symmetric nature of the annular system. However, after introducing the gold substrate, which can be transformed into a nanowire tangent to the inversion point with radius g2/ 2d, the near-field energy is now concentrated at the intermediate space between the nanowire and the annular system due to significant charge accumulation (see Figure 4d). Thus, the energy at the inversion point, which corresponds to the energy radiated into the far-field regime in the physical frame, is reduced to a considerable extent, resulting in the increase of the quality factor. Plasmon-Hybridization-Enhanced Photoluminescence Spectroscopy. Due to the inhibition of radiation

losses, a dark quadrupolar mode, by virtue of having a vanishing moment, can be more efficient than a bright dipolar mode to store electromagnetic energy in the near-field domain.61−63 Thus, we can expect that exciting the quadrupole−dipole hybridized mode in the gold-film-coupled dimer can significantly prompt optical processes occurring within its nanoscale gaps, in a more effective manner than the silica-supported dimer. In the meanwhile, the narrowed plasmon line width, corresponding to a higher quality factor, which is favorable to many plasmon-enhanced radiation phenomena, makes the filmcoupled dimer structure a promising platform for modulating the radiation properties of an optical emitter positioned inside the gap region. As a demonstration, here we perform the PL spectroscopy on the film-coupled dimer and the silicasupported dimer. The experimental setup for doing photoluminescence spectroscopy at the single-particle level is depicted in Figure 5a. A continuous wave laser of wavelength 633 nm, with the polarization direction controlled by a halfwavelength plate, is focused on a single gold nanosphere dimer by a 100× microscope objective. The PL emission signal from the structure is collected by the same objective and then delivered to a spectrometer for spectral analysis after passing through a long-pass dichroic mirror. A home-built illumination arm at the left side of the sample stage is used as a dark-field illumination module, enabling us to image single nanostructures. More details of the experimental setup are given in the Methods section. To avoid possible photothermal damage to the nanostructures under laser irradiation, the incident laser power was kept at a moderate level, ∼0.5 mW, equivalent to a power density of 1.6 × 104 W/cm2 at the sample plane. The scattering spectrum of the gold-film-coupled dimer measured after PL spectroscopy resembles the initial spectrum, F


Article

ACS Nano

Figure 5. (a) Schematic of the experimental setup for single-particle photoluminescence (PL) spectroscopy. HWP: half-wavelength plate, DM: dichroic mirror, BS: beam splitter. (b, d) Measured PL spectra of (b) the film-coupled gold nanosphere dimer and monomer (red) and (d) the silica-supported dimer (red), both under illumination by a 633 nm laser with the polarization direction along the dimer axis as depicted in the inset. The corresponding scattering spectra measured for both dimers (blue curves), both under white-light illumination with polarization along the dimer axis, are overlaid on their PL spectra for spectral characteristics comparison. (c) Excitation (blue) and emission (red) polarization dependent PL intensity of the film-coupled gold nanosphere dimer (yellow−blue core−shell). The magnitude of each data point is an intensity integral over a wavelength band of 20 nm centered at the emission peak. Similar results for the silica-supported dimer are shown in Figure S5 in the Supporting Information.

confirming that no perceivable photothermal damage has occurred (Figure S4). Note that the PL collection for the silicasupported nanoparticles was performed from the substrate side by overturning the sample because the PL emission of these nanoparticles mainly radiates into the silica substrate side, as we will show in Figure 7 later. Figure 5b,d show respectively the PL spectra of the gold-filmcoupled and silica-supported dimers measured with an incidence polarization along the dimer axis. Note that the cutoff in the PL spectra at wavelengths shorter than 630 nm is due to the cutoff response of the dichroic mirror that is used to filter out the excitation laser band and allow the longwavelength emission to pass through. The PL spectrum of the film-coupled dimer features a substantially enhanced (∼200 times) yet narrowed (∼4.6 times) emission peak compared to that of the silica-supported dimer. In particular, the PL spectral profile of each dimer inherits its scattering spectral characteristics: nearly the same radiation peak position and line width. These spectral similarities between PL and scattering for both dimers imply a plasmon-decay-mediated emission mechanism as elucidated in the following. Figure 6b schematically depicts the band structure of gold near the Fermi level around the X point in the first Brillouin

zone, while Figure 6a,c show the calculated scattering spectra representing the magnitude of the local density of plasmonic states (LDPSs)64 of the film-coupled dimer (Figure 6a) and the silica-supported dimer (Figure 6c), respectively. In analogy to the spontaneous emission process of an emitting dipole, a dipolar-like plasmon mode can decay in radiative and nonradiative channels. The radiative decay rate is proportional to the LDPSs in the same way as the radiative decay, γrad, of an emitter linearly depends on the local density of optical states πω (LDOSs), ρ(ω), following γrad(ω) = 3ℏϵ |p|2 ρ(ω), where p is 0

the transition dipole moment of the emitter. More details on the LDPSs are given in the Methods section. Through singlephoton absorption processes a large number of the d-band electrons in gold can be excited into higher energy levels in the sp-band, simultaneously leaving plenty of energetic holes in the d-band. The accumulated photoexcited electrons will then decay nonradiatively into the energy-matched LSP band of the dimer, with an exponentially decreased population distribution65 that accounts for the slight blue-shift in the PL emission peak with respect to the corresponding scattering peak in both dimers (see Figure 5b and d). The energetic electrons within the LSP band can either recombine directly with the photoexcited holes within the top region of the d-band through G


Article

ACS Nano

Figure 6. Schematic diagram elucidating the plasmon-mediated photoluminescence emission processes in both gold- and silica-supported dimers. (a, c) Simulated scattering spectrum of (a) the film-coupled dimer and (c) the silica-supported dimer, respectively. The red dashed line represents the laser line. (b) Energy diagram of gold band structure near the Fermi level around the X point in the first Brillouin zone. The thick upward arrow represents the transition of a d-band electron to the sp-band by absorbing a 633 nm photon. The photoexcited electron will decay nonradiatively into a localized plasmon resonance (LSP) energy matched band with exponentially decreased population redistribution (dashed red arrows). The majority of the accumulated energetic electrons within the LSP band will decay radiatively, giving rise to the PL emission, while the minority may recombine directly with the photoexcited holes within the top region of the d-band and emit photons (thin downward arrow), similar to the photoemission process in bulk gold.

Figure 7. (a, b) Calculated PL intensity distribution (cross-sectional view) at the emission peak of 650 nm for the nanosphere dimers on silica (b) and metal film (b), respectively. The color bars represent the calculated PL intensity by eq 4. (c, d) Calculated PL emission pattern at 650 nm. The blue dashed sectors correspond to the collection angles defined by the objective NA of 0.9.

interband transitions, emitting depolarized photons,66,67 or decay radiatively via plasmon damping, emitting plasmonmodulated polarized photons. Note that the plasmon-decaymediated photoluminescence emission is mainly determined by the LDPSs that could be very large at the LSP resonance band, thereby leading to an increased radiative decay rate and enhanced emission intensity. In general, the two emission

processes occur simultaneously, and the contribution of each decay channel can be evaluated by the degree of polarization (DoP) of both excitation and emission, defined as (Imax − Imin)/ (Imax + Imin), where Imax and Imin are the maximal and minimal emission intensity in the polarization-resolved PL measurements. The highly linearly polarized excitation and emission responses observed for both the film-coupled (DoP ∼0.88 and H


Article

ACS Nano

Figure 8. (a−d) Relative photoluminescence intensity calculated for different integrating domains: (a) the entire film-coupled dimer (orange) and the silica-supported dimer (light blue), (b, c) the horizontal (red) and vertical (purple) gap domains in (b) the silica-supported dimer and (c) the film-coupled dimer, and (d) the two nanospheres (red) and the underlying film (purple) of the film-coupled dimer. The corresponding integrating domains are indicated in yellow as shown in the insets. The apex angle θ, defining the domain size in (b) and (c), is 90°. The size of each slab in (d) is 12 × 12 × 6 nm3, with more details shown in Figure S5 in the Supporting Information.

radiative recombination at the frequency ω. The following assumptions were made in the calculations. (i) The absorption probability Yabs(ω0), which is determined by the imaginary part of the gold dielectric function due to the interband transitions at the excitation frequency ω0, can be assumed to be constant.18 (ii) The relaxation probability YR(ω), as an indication of the population redistribution of photoexcited electrons in the spband, corresponds to their nonradiative decay to the emission states, i.e., lower energy levels within the sp-band for bulk gold and the LSP band for the gold nanosphere dimer. The frequency dependence of this term is associated with the intrinsic bulk property of gold, and this probability is confirmed to be exponentially decreased with decreasing the frequency below the excitation level, as depicted by the dashed red curve in Figure 6. In a plasmon-decay-dominated emission process, the effect of the emission frequency dependence of YR(ω) is manifested by the slight blue-shift in the emission peak with respect to the scattering peak, as shown in Figure 5b,d.65 In the following calculations, the contribution of YR(ω) to the emission intensity of both dimer systems can therefore be assumed to be equal and excluded from eq 3 for a relative intensity comparison. (iii) Yem(ω), the emission probability of radiative recombination in bulk gold, can be modified to associate with the plasmon radiative decay rate in the dimer, which is determined by its LDPSs, which is proportional to the field intensity at the emission frequency.31,64 For comparison between theory and experiment, the collection efficiency, η, determined by the spatial radiation pattern and the numerical aperture of the microscope objective, has to be taken into consideration as well. Thus, we can evaluate the relative PL

∼0.83 for emission and excitation polarization in Figure 5c, respectively) and the silica-supported dimer (DoP ∼0.71 and ∼0.62 in Figure S5) indicate that the radiative plasmon decay substantially dominates over the interband transitions in their PL emissions, and the larger excitation and emission DoP values observed in the film-coupled dimer imply a more pronounced plasmon-enhanced excitation and emission over its counterpart on the silica substrate. Quantitative Evaluation of Photoluminescence Enhancement and Enhancement Sites. In this section, we provide a quantitative evaluation of the PL intensity and the PL enhancement sites in the two systems. In light of their geometric features, we infer that the PL emission enhancement in each system is mainly contributed either by the vertical sphere−film gap or the horizontal sphere−sphere gap. To evaluate the quantitative contribution of each gap to the total emission intensity, we calculate the relative PL spectra of both dimers using a generalized one-photon luminescence model. According to the general theory of single-photon absorption induced photoemission, the photoluminescence from a volume element dV in a material can be expressed as68 ⇀

IPL(ω) dV = I0(ω0 , r ) Yabs(ω0) YR (ω) Yem(ω) dV

(3)

where ω0 and ω0 are the angular frequency of the excitation and ⇀ emission, respectively. I0(ω0, r ) is the field intensity of the ⇀ excitation at point r . Yabs(ω0) is the absorption probability of a single photon (or single-photon absorption cross section) at ω0 for generating an electron−hole pair. YR(ω) is the relaxation probability of photoexcited carriers from an excited state to an emission state, and Yem(ω) is the emission probability of I


Article

ACS Nano intensity in the dimer by solely considering the field intensity and the collection efficiency as ⎯⇀ ⎯

IPL(ω) ∝ η ⎯⇀ ⎯

⇀

⎯⇀ ⎯

⇀

∫V |E0(ω0 , r )|2 ·|Eem(ω , r )|2 dV ⇀

⎯⇀ ⎯

the entire silica-supported dimer, revealing that the total PL emission mainly comes from the sphere−sphere gap region. In remarkable contrast, Figure 8c shows that the two sphere−film gaps have a dominant contribution to the total PL intensity of the film-coupled dimer, although the incidence polarization is set along the dimer axis, which would preferentially excite the sphere−sphere gap plasmon resonance. This observation is consistent with the above-proposed metal-film-mediated dipolar−quadrupolar plasmon hybridization mechanism, through which the bright dipolar mode effectively harvests the incident light energy, transfers the energy to the dark quadrupolar mode through near-field coupling, and significantly enhances the energy density in the vertical gap regions, leading to efficient photoluminescence. Furthermore, Figure 8d shows that the PL intensity of the underlying gold slab overwhelmingly exceeds that of the nanosphere dimer by a factor of more than 20. This could be attributed to the denser charge accumulation in the flat surface of the gold film than in the curved surface of the nanospheres, leading to higher field intensities in the gold film domain.

(4)

⇀

where (E0 (ω0, r ) and Eem (ω, r ) are the electric fields at the excitation frequency ω0 and the emission frequency ω, respectively, and V is the integration volume. Using the finite-difference time domain method, we can calculate the electric field terms in eq 4 and subsequently calculate the relative PL emission intensity. Note that this calculation model is applicable only to plasmon-radiation-dominated PL emission processes based on the aforementioned assumptions. As shown in Figure 7a,b, the calculated PL intensity distributions at the peak emission wavelength of 650 nm reveal that the PL emission of the silica-supported dimer mainly originates from the particle−particle gap, while that of the goldfilm-coupled dimer comes from both the sphere−sphere gap and the sphere−film ones. We also observe that the PL intensity at the sphere−film gaps is substantially larger than that of the sphere−sphere gap in the film-coupled dimer, and, on the other hand, the latter is much stronger than that of the sphere−sphere gap in the silica-supported dimer. This observation is very counterintuitive, considering the fact that the incidence polarization is aligned along the dimer axis. Since the sources of PL emissions are highly confined within the nanoscale gaps in both dimers, we can evaluate their radiation directionalities by simulating the far-field radiation patterns of the two systems with an emitter−cavity model, with details described in the Methods section. The calculation results shown in Figure 7c,d demonstrate that the emission patterns for both systems are similar to their plasmonic scattering patterns as shown in Figure 2b,c, which again supports the radiative decay of plasmons to be the origin of the observed PL emissions. By integrating the radiation intensity over the collection angle of the objective in the silica substrate side in Figure 7c and in the upper-half space in Figure 7d, an emission efficiency of 84% for the silica-supported dimer and 87% for the gold-film-coupled dimer are obtained, respectively, with which the relative PL intensity spectra can be calculated for comparison with the wavelength-dependent experimental results. Figure 8 presents the calculated relative PL intensity spectra over specific integrating domains in the gold-film-coupled and silica-supported dimers. Note that a modified approach was adopted to distinguish the plasmon-enhanced nonlinear emission contribution from a discrete domain of an entire hybrid plasmonic system.64,69,70 Surprisingly, the calculation results shown in Figure 7a well reproduce the measured PL spectral profile for both dimers. Additionally, the results also give an enhancement factor of ∼253 for the film-coupled dimer compared to the silica-supported one, showing excellent agreement with experiment (∼200 times). The negligible discrepancy could come from the effect of nonradiative plasmon decays that are absent in the above discussions on the mechanism of PL emission8 or the relatively weak bulk photoluminescence beyond consideration. The above comparisons thus demonstrate the metal-film-coupled nanoparticle dimers have a more efficient nanocavity in boosting light− matter interactions for enhanced spectroscopy applications. Figure 8b compares the relative PL intensity spectra calculated for the sphere−sphere gap, sphere−film gap, and

CONCLUSION In conclusion, we have realized in experiment pronounced linewidth narrowing of the bonding dipolar plasmon resonance in a gold nanosphere dimer by replacing a silica substrate with a gold film. Theoretical analyses using electromagnetic simulations, a multipole expansion model, and conformal transformation optics suggest that the intensive plasmon hybridization between the bonding dipolar mode and an induced quadrupolar mode is responsible for shrinking the resonance line width in the film-coupled dimer, thus leading the way in engineering the resonance line width of coupled plasmonic modes in general nanoclusters by metal-substrate-mediated mode hybridization. Moreover, upon near-resonant excitation of the hybridized dipole−quadrupole plasmon resonance, the film-coupled dimer shows a photoluminescence intensity enhancement up to ∼200 times yet with a much narrower emission line width in comparison with the silica-supported dimer. The close similarities between the spectral profile and polarization dependence of the PL emission and the plasmonic scattering in both the silica-supported and gold-film-coupled dimers suggest that the radiative decay of the coupled plasmons in the two systems dominates the photoluminescence emission. In addition, the relative PL intensity distribution calculated over the two dimers reveals that the contribution from the vertical sphere−film gap junctions dominates the total PL emission of the entire system. These findings demonstrate that the filmcoupled metal nanosphere dimer can provide a nanoscale cavity with both ultrasmall mode volume and ultrahigh quality factor, offering a promising platform for various plasmon-enhanced spectroscopy applications such as SERS,71 nonlinear plasmonic sensing,72 and plasmon−molecule strong coupling.32 METHODS Sample Preparation. The gold film was deposited on an ultraclean coverslip (Schott Nexterion, Germany) using a thermal evaporator (Nexdep, Angstrom Engineering, Inc.). The thickness and surface roughness (RMS) of the gold film was determined to be about 70 and 0.5 nm by an atomic force microscope (AFM, Nanoscope V, Veeco). CTAB-enclosed gold nanospheres with a nominal diameter of 100 ± 8 nm were purchased from Nanoseedz, Inc. (Hong Kong, China). The gold nanospheres were rinsed and diluted with deionized water to reach a relatively low concentration (typically 109 particles/ J


Article

ACS Nano mL). A drop of diluted colloidal solution (∼100 μL) was casted on the gold film; upon drying in air, some nanosphere were likely to selfassemble into dimers, trimers, or even large clusters. The dimers can be identified with the aid of a dark-field microscope. Single-Particle Dark-Field Scattering Spectroscopy. Optical scattering spectroscopy at the single-particle level was performed with an Olympus BX-51 upright microscope equipped with a standard darkfield optical module and a 100× dark-field objective (NA 0.8, LMPlan, Olympus). Two sets of white-light illumination configurations, including a standard unpolarized illumination module and a homebuilt polarization-controlled oblique illumination module, were used for unpolarized and polarization-resolved excitations, respectively. Dark-field images and scattering spectra of individual nanostructures were recorded respectively using a color CCD camera (QICAM 12-bit, QImaging, Inc.) and an imaging spectrometer (Acton SP2300, Princeton Instruments) equipped with a gray CCD camera (PIXIS:400BR eXcelon, Princeton Instruments). Single-Particle Photoluminescence Spectroscopy. The photoluminescence measurements were carried out on a Horiba Raman microscope platform (LabRAM HR800, Horiba). A continuous-wave laser beam (wavelength 633 nm) with an average power of ∼0.5 mW was guided to the microscope (BX-30, Olympus) through a 100× objective (NA 0.9, MPlan Olympus) and focused on the sample. The diameter of the focused laser spot in the sample plane was estimated to be ∼1 μm. The photoluminescence emission signal was collected by the same objective and sent through a dichroic mirror that blocks the Rayleigh scattering of the excitation laser and simultaneously allows the long-wavelength emission signal to pass through. After a second filtering with a long-pass filter, the collected emission signal was delivered to the spectrometer for spectral analysis. Note that the PL collection by the objective is from the substrate side for the silicasupported nanoparticles and from the superstrate side for the goldfilm-coupled particles. Electrodynamics Simulations. Distribution profiles of the surface charge density and the near-field intensity for the gold-filmcoupled and silica-supported dimers were numerically calculated using a commercially available FEM package (COMSOL Multiphysics 4.3a with RF module). The scattering field of a gold nanosphere dimer was calculated with a standard two-step method, where the background field is first computed for the same geometry in the absence of the nanospheres under illumination polarized along the dimer axis at an incidence angle of 70° from the surface normal. The surface charge density is calculated from the normal component of the electric field. In all calculations, the permittivity of Au was modeled using the experimental data of Johnson and Christy with linear interpolation,73 and the refractive index of silica was taken as 1.5. For simplicity without loss of generality, we neglected the surfactant layer (CTAB) on the gold nanospheres. In order to simulate the semi-infinite air and substrate region, the computation domain was truncated by perfectly matched layers to reduce reflections. The simulation domain was finely meshed with a mesh size of 0.25 nm in the gap region such that the convergence of the results and the accuracy of the computed fields were ensured. Multipole Expansion Model. The different moments in the multipole expansion (see eq 1) are computed from the numerically calculated distribution of polarization current J = −ωP, where P is the polarization vector inside each gold nanosphere. For a nonmagnetic medium, the electric dipole moment is

=

∫V ε0(εr − 1)[E(r′)r′ + r′E(r′)] dV ′

The relative radiation contribution of each multiple to the total average power is then obtained by integrating the complex Poynting vector in the far field. The closed-form solutions for the radiated powers by the dipole moment and quadrupole moment are respectively

⟨Wrad,P⟩ =

k04 12πη0ε0 2

|P|2

and

⟨Wrad,Q ⟩ =

k 06 40πη0ε0 2

|Q |2

where η0 is the impedance of free space. Finally, the scattering crosssection contribution from each multipole is simply computed by normalizing to the incident power density. The total scattering cross section of the dimer structures can then be obtained by summing the scattering cross-section contributions from each multipole. Local Density of Plasmonic States and Plasmon-DecayGenerated Photoluminescence. The local density of plasmonic states is also called surface plasmonic local density of optical states (SP-LDOSs). In analogy with the LDOSs conventionally used in an emitter−cavity system, the introduction of LDPSs can simplify the theoretical description of plasmon-decay-generated emission in the manner of plasmonic molecule−cavity interactions. Therefore, the related concepts and definitions used in the LDOSs can readily be applied to LDPSs for analysis of radiative plasmon decay. In general, the spontaneous emission rate of a dipolar emitter in a cavity is given by πω 0 γsp(r , ω) = |p|2 ρ(r , ω) + γint 3ℏϵ0 where r is the position, ω is the emission frequency, ϵ0 is the permittivity of free space, p is the transition dipole moment, and γ0int is the internal nonradiative decay rate of the emitter. The ρ(r, ω), termed the LDOSs of the cavity, can be correlated with the dyadic Green’s function G as ρ(r , ω) ∝ np · Im{G(r , r )}· np where np is the orientation of the transition dipole, and

⎡Gxx Gxy Gxz ⎤ ⎢ ⎥ G = ⎢ Gxy Gyy Gyz ⎥ ⎢ ⎥ ⎢⎣Gxz Gyz Gzz ⎥⎦ The total LDOSs can be obtained by averaging over the different orientations, which translates to averaging the three diagonal G components as

ρ ̅ (r , ω) =

1 ⎡ 6ω 6ω Im{Gxx(r , r , ω)} + 2 Im{Gyy(r , r , ω)} ⎢ 3 ⎣ πc 2 πc ⎤ 6ω + 2 Im{Gzz(r , r , ω)}⎥ ⎦ πc

where

1 P=− iω

∫V J(r′) dV ′ = ∫V ε0(εr − 1) E(r′) dV ′

Gxx =

where V is the volume of the particle, E(r′) is the electric field, and ε0 and εr are the permittivity of a vacuum and the relative permittivity of gold, respectively. The second-order term of the expansion, that is, the electric quadrupole, is given by Q=−

1 iω

Gyy =

∫V [J(r′)r′ + r′J(r′)] dV ′

Gzz = K

Ex c 2ε0εr μω2 Eyc 2ε0εr μω2

Ezc 2ε0εr μω2 DOI: 10.1021/acsnano.7b00048 ACS Nano XXXX, XXX, XXX−XXX

Article

ACS Nano

Noble-Metal Nanoparticles. ACS Appl. Mater. Interfaces 2012, 4, 4746−4751. (11) Slablab, a.; Le Xuan, L.; Zielinski, M.; de Wilde, Y.; Jacques, V.; Chauvat, D.; Roch, J.-F. Second-Harmonic Generation from Coupled Plasmon Modes in a Single Dimer of Gold Nanospheres. Opt. Express 2012, 20, 220. (12) Celebrano, M.; Wu, X.; Baselli, M.; Großmann, S.; Biagioni, P.; Locatelli, A.; De Angelis, C.; Cerullo, G.; Osellame, R.; Hecht, B.; Duò, L.; Ciccacci, F.; Finazzi, M. Mode Matching in Multiresonant Plasmonic Nanoantennas for Enhanced Second Harmonic Generation. Nat. Nanotechnol. 2015, 10, 412−417. (13) Guan, Z.; Gao, N.; Jiang, X. F.; Yuan, P.; Han, F.; Xu, Q. H. Huge Enhancement in Two-Photon Photoluminescence of Au Nanoparticle Clusters Revealed by Single-Particle Spectroscopy. J. Am. Chem. Soc. 2013, 135, 7272−7277. (14) Englund, D.; Faraon, A.; Fushman, I.; Stoltz, N.; Petroff, P.; Vučković, J. Controlling Cavity Reflectivity with a Single Quantum Dot. Nature 2007, 450, 857−861. (15) Faraon, A.; Fushman, I.; Englund, D.; Stoltz, N.; Petroff, P.; Vučković, J. Coherent Generation of Non-Classical Light on a Chip via Photon-Induced Tunnelling and Blockade. Nat. Phys. 2008, 4, 859− 863. (16) Nomura, M.; Kumagai, N.; Iwamoto, S.; Ota, Y.; Arakawa, Y. Laser Oscillation in a Strongly Coupled Single-Quantum-Dot− nanocavity System. Nat. Phys. 2010, 6, 279−283. (17) Lévêque, G.; Martin, O. J. F. Optical Interactions in a Plasmonic Particle Coupled to a Metallic Film. Opt. Express 2006, 14, 9971− 9981. (18) Lumdee, C.; Yun, B.; Kik, P. G. Gap-Plasmon Enhanced Gold Nanoparticle Photoluminescence. ACS Photonics 2014, 1, 1224−1230. (19) Schertz, F.; Schmelzeisen, M.; Mohammadi, R.; Kreiter, M.; Elmers, H. J.; Schönhense, G. Near Field of Strongly Coupled Plasmons: Uncovering Dark Modes. Nano Lett. 2012, 12, 1885−1890. (20) Lassiter, J. B.; McGuire, F.; Mock, J. J.; Ciracì, C.; Hill, R. T.; Wiley, B. J.; Chilkoti, A.; Smith, D. R. Plasmonic Waveguide Modes of Film-Coupled Metallic Nanocubes. Nano Lett. 2013, 13, 5866−5872. (21) Lei, D. Y.; Fernández-Domínguez, A. I.; Sonnefraud, Y.; Appavoo, K.; Haglund, R. F.; Pendry, J. B.; Maier, S. A. Revealing Plasmonic Gap Modes in Particle-on-Film Systems Using Dark-Field Spectroscopy. ACS Nano 2012, 6, 1380−1386. (22) Li, G.-C.; Zhang, Y.-L.; Lei, D. Y. Hybrid Plasmonic Gap Modes in Metal Film-Coupled Dimers and Their Physical Origins Revealed by Polarization Resolved Dark Field Spectroscopy. Nanoscale 2016, 8, 7119−7126. (23) Hill, R. T.; Mock, J. J.; Hucknall, A.; Wolter, S. D.; Jokerst, N. M.; Smith, D. R.; Chilkoti, A. Plasmon Ruler with Angstrom Length Resolution. ACS Nano 2012, 6, 9237−9246. (24) Mubeen, S.; Zhang, S.; Kim, N.; Lee, S.; Krämer, S.; Xu, H.; Moskovits, M. Plasmonic Properties of Gold Nanoparticles Separated from a Gold Mirror by an Ultrathin Oxide. Nano Lett. 2012, 12, 2088− 2094. (25) Mock, J. J.; Hill, R. T.; Degiron, A.; Zauscher, S.; Chilkoti, A.; Smith, D. R. Distance-Dependent Plasmon Resonant Coupling between a Gold Nanoparticle and Gold Film. Nano Lett. 2008, 8, 2245−2252. (26) Ciraci, C.; Hill, R. T.; Mock, J. J.; Urzhumov, Y.; FernandezDominguez, a. I.; Maier, S. A.; Pendry, J. B.; Chilkoti, a.; Smith, D. R. Probing the Ultimate Limits of Plasmonic Enhancement. Science 2012, 337, 1072−1074. (27) Hajisalem, G.; Nezami, M. S.; Gordon, R. Probing the Quantum Tunneling Limit of Plasmonic Enhancement by Third Harmonic Generation. Nano Lett. 2014, 14, 6651−6654. (28) Mertens, J.; Eiden, A. L.; Sigle, D. O.; Huang, F.; Lombardo, A.; Sun, Z.; Sundaram, R. S.; Colli, A.; Tserkezis, C.; Aizpurua, J.; Milana, S.; Ferrari, A. C.; Baumberg, J. J. Controlling Subnanometer Gaps in Plasmonic Dimers Using Graphene. Nano Lett. 2013, 13, 5033−5038. (29) Oulton, R. F.; Sorger, V. J.; Zentgraf, T.; Ma, R.-M.; Gladden, C.; Dai, L.; Bartal, G.; Zhang, X. Plasmon Lasers at Deep Subwavelength Scale. Nature 2009, 461, 629−632.

Analogously, the LDPSs of a plasmonic nanocavity are directly correlated with the diagonal G components, and, therefore, the relative plasmon emission intensity can be readily evaluated by the field intensity at emission frequency with eq 4 described in the main text.

ASSOCIATED CONTENT S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.7b00048. Additional experimental and simulation details as well as some supplementary results mentoned in the main text (PDF)

AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. ORCID

Guang-Can Li: 0000-0002-9903-8900 Dang Yuan Lei: 0000-0002-8963-0193 Notes

The authors declare no competing financial interest.

ACKNOWLEDGMENTS We acknowledge the financial support by the Hong Kong Research Grants Council (ECS Grant No. 509513). Y.L. acknowledges the funding support from Singapore Ministry of Education (Grant No. MOE2015-T2-1-145). REFERENCES (1) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Single Molecule Detection Using SurfaceEnhanced Raman Scattering (SERS). Phys. Rev. Lett. 1997, 78, 1667− 1670. (2) Nie, S.; Emory, S. R. Probing Single Molecules and Single Nanoparticles by Surface-Enhanced Raman Scattering. Science 1997, 275, 1102−1106. (3) Lim, D.-K.; Jeon, K.-S.; Kim, H. M.; Nam, J.-M.; Suh, Y. D. Nanogap-Engineerable Raman-Active Nanodumbbells for SingleMolecule Detection. Nat. Mater. 2010, 9, 60−67. (4) Zhang, X.; Zheng, Y.; Liu, X.; Lu, W.; Dai, J.; Lei, D. Y.; MacFarlane, D. R. Hierarchical Porous Plasmonic Metamaterials for Reproducible Ultrasensitive Surface-Enhanced Raman Spectroscopy. Adv. Mater. 2015, 27, 1090−1096. (5) Chen, S.; Svedendahl, M.; Van Duyne, R. P.; Käll, M. PlasmonEnhanced Colorimetric ELISA with Single Molecule Sensitivity. Nano Lett. 2011, 11, 1826−1830. (6) Jun, Y.; Sheikholeslami, S.; Hostetter, D. R.; Tajon, C.; Craik, C. S.; Alivisatos, A. P. Continuous Imaging of Plasmon Rulers in Live Cells Reveals Early-Stage Caspase-3 Activation at the Single-Molecule Level. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 17735−17740. (7) Storhoff, J. J.; Elghanian, R.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L. One-Pot Colorimetric Differentiation of Polynucleotides with Single Base Imperfections Using Gold Nanoparticle Probes. J. Am. Chem. Soc. 1998, 120, 1959−1964. (8) Huang, D.; Byers, C. P.; Wang, L.; Hoggard, A.; Hoener, B.; Dominguez-Medina, S.; Chen, S.; Chang, W.; Landes, C. F.; Link, S. Photoluminescence of a Plasmonic Molecule. ACS Nano 2015, 9, 7072−7079. (9) Sivun, D.; Vidal, C.; Munkhbat, B.; Arnold, N.; Klar, T. A.; Hrelescu, C. Anticorrelation of Photoluminescence from Gold Nanoparticle Dimers with Hot-Spot Intensity. Nano Lett. 2016, 16, 7203−7209. (10) Han, F.; Guan, Z.; Tan, T. S.; Xu, Q.-H. Size-Dependent TwoPhoton Excitation Photoluminescence Enhancement in Coupled L


Article

ACS Nano (30) Russell, K. J.; Liu, T.-L.; Cui, S.; Hu, E. L. Large Spontaneous Emission Enhancement in Plasmonic Nanocavities. Nat. Photon. 2012, 6, 459−462. (31) Akselrod, G. M.; Argyropoulos, C.; Hoang, T. B.; Ciracì, C.; Fang, C.; Huang, J.; Smith, D. R.; Mikkelsen, M. H. Probing the Mechanisms of Large Purcell Enhancement in Plasmonic Nanoantennas. Nat. Photon. 2014, 8, 835−840. (32) Chikkaraddy, R.; de Nijs, B.; Benz, F.; Barrow, S. J.; Scherman, O. A.; Rosta, E.; Demetriadou, A.; Fox, P.; Hess, O.; Baumberg, J. J. Single-Molecule Strong Coupling at Room Temperature in Plasmonic Nanocavities. Nature 2016, 535, 127−130. (33) Benz, F.; Schmidt, M. K.; Dreismann, A.; Chikkaraddy, R.; Zhang, Y.; Demetriadou, A.; Carnegie, C.; Ohadi, H.; de Nijs, B.; Esteban, R.; Aizpurua, J.; Baumberg, J. J. Single-Molecule Optomechanics in “picocavities. Science 2016, 354, 726−729. (34) Chen, H.; Ming, T.; Zhang, S.; Jin, Z.; Yang, B.; Wang, J. Effect of the Dielectric Properties of Substrates on the Scattering Patterns of Gold Nanorods. ACS Nano 2011, 5, 4865−4877. (35) Muskens, O. L.; Giannini, V.; Sánchez-Gil, J. a; Gómez Rivas, J. Optical Scattering Resonances of Single and Coupled Dimer Plasmonic Nanoantennas. Opt. Express 2007, 15, 17736−17746. (36) Wen, F.; Zhang, Y.; Gottheim, S.; King, N. S.; Zhang, Y.; Nordlander, P.; Halas, N. J. Charge Transfer Plasmons: Optical Frequency Conductances and Tunable Infrared Resonances. ACS Nano 2015, 9, 6428−6435. (37) Sobhani, A.; Manjavacas, A.; Cao, Y.; McClain, M. J.; García de Abajo, F. J.; Nordlander, P.; Halas, N. J. Pronounced Linewidth Narrowing of an Aluminum Nanoparticle Plasmon Resonance by Interaction with an Aluminum Metallic Film. Nano Lett. 2015, 15, 6946−6951. (38) Mock, J. J.; Hill, R. T.; Tsai, Y. J.; Chilkoti, A.; Smith, D. R. Probing Dynamically Tunable Localized Surface Plasmon Resonances of Film-Coupled Nanoparticles by Evanescent Wave Excitation. Nano Lett. 2012, 12, 1757−1764. (39) Berciaud, S.; Cognet, L.; Tamarat, P.; Lounis, B. Observation of Intrinsic Size Effects in the Optical Response of Individual Gold Nanoparticles. Nano Lett. 2005, 5, 515−518. (40) Sönnichsen, C.; Franzl, T.; Wilk, T.; von Plessen, G.; Feldmann, J.; Wilson, O.; Mulvaney, P. Drastic Reduction of Plasmon Damping in Gold Nanorods. Phys. Rev. Lett. 2002, 88, 77402. (41) Novo, C.; Gomez, D.; Perez-Juste, J.; Zhang, Z.; Petrova, H.; Reismann, M.; Mulvaney, P.; Hartland, G. V. Contributions from Radiation Damping and Surface Scattering to the Linewidth of the Longitudinal Plasmon Band of Gold Nanorods: A Single Particle Study. Phys. Chem. Chem. Phys. 2006, 8, 3540−3546. (42) Lukosz, W. Light Emission by Magnetic and Electric Dipoles close to a Plane Dielectric Interface. III. Radiation Patterns of Dipoles with Arbitrary Orientation. J. Opt. Soc. Am. 1979, 69, 1495−1503. (43) Wang, H.; Liu, P.; Ke, Y.; Su, Y.; Zhang, L.; Xu, N.; Deng, S.; Chen, H. Janus Magneto−Electric Nanosphere Dimers Exhibiting Unidirectional Visible Light Scattering and Strong Electromagnetic Field Enhancement. ACS Nano 2015, 9, 436−448. (44) Hentschel, M.; Dregely, D.; Vogelgesang, R.; Giessen, H.; Liu, N. Plasmonic Oligomers: The Role of Individual Particles in Collective Behavior. ACS Nano 2011, 5, 2042−2050. (45) Hentschel, M.; Saliba, M.; Vogelgesang, R.; Giessen, H.; Alivisatos, A. P.; Liu, N. Transition from Isolated to Collective Modes in Plasmonic Oligomers. Nano Lett. 2010, 10, 2721−2726. (46) Lassiter, J. B.; Sobhani, H.; Fan, J. A.; Kundu, J.; Capasso, F.; Nordlander, P.; Halas, N. J. Fano Resonances in Plasmonic Nanoclusters: Geometrical and Chemical Tunability. Nano Lett. 2010, 10, 3184−3189. (47) Fan, J. a; Wu, C.; Bao, K.; Bao, J.; Bardhan, R.; Halas, N. J.; Manoharan, V. N.; Nordlander, P.; Shvets, G.; Capasso, F. SelfAssembled Plasmonic Nanoparticle Clusters. Science 2010, 328, 1135− 1138. (48) Stockman, M. I. Nanoscience: Dark-Hot Resonances. Nature 2010, 467, 541−542.

(49) Wang, H.; Wu, Y.; Lassiter, B.; Nehl, C. L.; Hafner, J. H.; Nordlander, P.; Halas, N. J. Symmetry Breaking in Individual Plasmonic Nanoparticles. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 10856−10860. (50) Bachelier, G.; Russier-Antoine, I.; Benichou, E.; Jonin, C.; Del Fatti, N.; Vallée, F.; Brevet, P. F. Fano Profiles Induced by near-Field Coupling in Heterogeneous Dimers of Gold and Silver Nanoparticles. Phys. Rev. Lett. 2008, 101, 197401. (51) Christ, A.; Martin, O. J. F.; Ekinci, Y.; Gippius, N. A.; Tikhodeev, S. G. Symmetry Breaking in a Plasmonic Metamaterial at Optical Wavelength. Nano Lett. 2008, 8, 2171−2175. (52) Christ, A.; Ekinci, Y.; Solak, H. H.; Gippius, N. A.; Tikhodeev, S. G.; Martin, O. J. F. Controlling the Fano Interference in a Plasmonic Lattice. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 76, 201405. (53) Chen, H.; Shao, L.; Ming, T.; Woo, K. C.; Man, Y. C.; Wang, J.; Lin, H.-Q. Observation of the Fano Resonance in Gold Nanorods Supported on High-Dielectric-Constant Substrates. ACS Nano 2011, 5, 6754−6763. (54) Wu, Y.; Nordlander, P. Finite-Difference Time-Domain Modeling of the Optical Properties of Nanoparticles near Dielectric Substrates. J. Phys. Chem. C 2010, 114, 7302−7307. (55) Zhang, S.; Bao, K.; Halas, N. J.; Xu, H.; Nordlander, P. Substrate-Induced Fano Resonances of a Plasmonic Nanocube: A Route to Increased-Sensitivity Localized Surface Plasmon Resonance Sensors Revealed. Nano Lett. 2011, 11, 1657−1663. (56) Shafiei, F.; Monticone, F.; Le, K. Q.; Liu, X.; Hartseld, T.; Alù, A.; Li, X. A Subwavelength Plasmonic Metamolecule Exhibiting Magnetic-Based Optical Fano Resonance. Nat. Nanotechnol. 2013, 8, 95−99. (57) Aubry, A.; Lei, D. Y.; Maier, S. A.; Pendry, J. B. Interaction between Plasmonic Nanoparticles Revisited with Transformation Optics. Phys. Rev. Lett. 2010, 105, 233901. (58) Lei, D. Y.; Aubry, A.; Maier, S. A.; Pendry, J. B. Broadband Nano-Focusing of Light Using Kissing Nanowires. New J. Phys. 2010, 12, 93030. (59) Aubry, A.; Lei, D. Y.; Maier, S. A.; Pendry, J. B. Plasmonic Hybridization between Nanowires and a Metallic Surface: A Transformation Optics Approach. ACS Nano 2011, 5, 3293−3308. (60) Luo, Y.; Lei, D. Y.; Maier, S. A.; Pendry, J. B. TransformationOptics Description of Plasmonic Nanostructures Containing Blunt Edges/Corners: From Symmetric to Asymmetric Edge Rounding. ACS Nano 2012, 6, 6492−6506. (61) Liaw, J.-W.; Liu, C.-L.; Kuo, M.-K. Dual-Band Plasmonic Enhancement of Ag-NS@SiO2 on Gain Medium’s Spontaneous Emission. Plasmonics 2011, 6, 673−680. (62) Hastings, S. P.; Swanglap, P.; Qian, Z.; Fang, Y.; Park, S. J.; Link, S.; Engheta, N.; Fakhraai, Z. Quadrupole-Enhanced Raman Scattering. ACS Nano 2014, 8, 9025−9034. (63) Butet, J.; Russier-Antoine, I.; Jonin, C.; Lascoux, N.; Benichou, E.; Brevet, P. F. Sensing with Multipolar Second Harmonic Generation from Spherical Metallic Nanoparticles. Nano Lett. 2012, 12, 1697− 1701. (64) Viarbitskaya, S.; Teulle, A.; Marty, R.; Sharma, J.; Girard, C.; Arbouet, A.; Dujardin, E. Tailoring and Imaging the Plasmonic Local Density of States in Crystalline Nanoprisms. Nat. Mater. 2013, 12, 426−432. (65) Yin, T.; Dong, Z.; Jiang, L.; Zhang, L.; Hu, H.; Qiu, C.-W.; Yang, J. K. W.; Shen, Z. X. Anomalous Shift Behaviors in the Photoluminescence of Dolmen-Like Plasmonic Nanostructures. ACS Photonics 2016, 3, 979−984. (66) Boyd, G. T.; Yu, Z. H.; Shen, Y. R. Photoinduced Luminescence from the Noble Metals and Its Enhancement on Roughened Surfaces. Phys. Rev. B: Condens. Matter Mater. Phys. 1986, 33, 7923−7936. (67) Hu, H.; Duan, H.; Yang, J. K. W.; Shen, Z. X. PlasmonModulated Photoluminescence of Individual Gold Nanostructures. ACS Nano 2012, 6, 10147−10155. (68) Beversluis, M.; Bouhelier, A.; Novotny, L. Continuum Generation from Single Gold Nanostructures through near-Field M


Article

ACS Nano Mediated Intraband Transitions. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 68, 1−10. (69) Aouani, H.; Rahmani, M.; Navarro-Cía, M.; Maier, S. A. ThirdHarmonic-Upconversion Enhancement from a Single Semiconductor Nanoparticle Coupled to a Plasmonic Antenna. Nat. Nanotechnol. 2014, 9, 290−294. (70) Aouani, H.; Navarro-Cía, M.; Rahmani, M.; Maier, S. A. Unveiling the Origin of Third Harmonic Generation in Hybrid ITOPlasmonic Crystals. Adv. Opt. Mater. 2015, 3, 1059−1065. (71) Wang, X.; Li, M.; Meng, L.; Lin, K.; Feng, J.; Huang, T.; Yang, Z.; Ren, B. Probing the Location of Hot Spots by Surface-Enhanced Raman Spectroscopy: Toward Uniform Substrates. ACS Nano 2014, 8, 528−536. (72) Mesch, M.; Metzger, B.; Hentschel, M.; Giessen, H. Nonlinear Plasmonic Sensing. Nano Lett. 2016, 16, 3155−3159. (73) Johnson, P. B.; Christy, R. W. Optical Constants of the Noble Metals. Phys. Rev. B 1972, 6, 4370−4379.

N


Metal-Substrate-Mediated Plasmon Hybridization in a Nanoparticle

Recommend Documents