Partial Cloaking of a Gold Particle by a Single Molecule

Extinction of light by material particles stems from losses incurred by absorption or scattering. The extinction cross section is usually treated as an additive quantity, leading to the exponential laws that govern the macroscopic attenuation of light. In this Letter, we demonstrate that the extinction cross section of a large gold nanoparticle can be substantially reduced—i.e., the particle becomes more transparent—if a single molecule is placed in its near field. This partial cloaking effect results from a coherent plasmonic interaction between the molecule and the nanoparticle, whereby each of them acts as a nanoantenna to modify the radiative properties of the other.

Figure 1

(a) Schematics of the experimental arrangement. The sample consists of DBT molecules in a pDCB crystal surrounding an array of gold nanoparticles prepared in a nanochannel (see text for details). SIL: solid-immersion lens, GNP: gold nanoparticle. Arrows at the bottom show the translational degrees of freedom. (b) Scanning electron microscope image of the GNP array. Inset: enlargement of a single GNP. (c) Optical transmission image of two GNPs at $\lambda =740\mathrm{nm}$. Upper panel shows a normalized cross section along the dashed line. (d) Fluorescence spectrum of DBT recorded upon excitation via transition from $|g,v=0⟩$ to $|e,v\ne 0⟩$. The strong emission line at $\lambda =740.3\mathrm{nm}$ represents the 00ZPL. Inset: molecular structure and Jablonski diagram of DBT. (e) Plasmon resonance of a GNP measured in transmission (dots) fitted by a Lorentzian profile (black curve). (f) Predicted reduction of the GNP extinction at the resonance of a single molecule that is coupled to it in the near field.

Figure 2

(a) Fluorescence of DBT molecules located within a focal spot of one GNP as a function of the excitation laser frequency. The applied laser power of 79 nW corresponds to an excitation intensity above saturation. Inset: an enlargement of the spectrum of $M0$ recorded at low excitation power. (b) An extinction image of a single GNP recorded in transmission overlaid with the locations of the GNP obtained from its fluorescence image (yellow), $M0$ (magenta), and $M1$ (blue). The arrows depict the in-plane orientations of the molecular dipole moments. (c) 240 spectra recorded at $20\mathrm{GHz}/\mathrm{s}$ over one minute well below saturation (excitation power 0.04 nW). (d) Sum of the aligned spectra shown in (c) (symbols) and a Lorentzian fit (solid curve).

Figure 3

(a) Second-order autocorrelation function, $g^{(2)}(\tau )$ for $M0$ excited via 00ZPL (magenta), $M1$ excited via 00ZPL (blue), and $M1$ excited via a higher vibrational level $|e,v\ne 0⟩$ (green). (b) Fluorescence signal versus excitation power for $M0$ (magenta) and $M1$ (blue) under 00ZPL excitation. (c) 00ZPL linewidth (FWHM) versus excitation power extracted from the spectra used in (b). (d) Same as in (a) but excited above saturation. Solid black lines are fits to the data (see the Supplemental Material [25]).

Figure 4

(a) Transmission spectra recorded about the 00ZPL resonances of $M1$ (blue) and $M0$ (magenta) placed in the focus of the excitation beam, corresponding to the middle dashed cut in (d). (b),(c) Same as in (a) but for molecules placed on two opposite sides of the focus, as indicated by the lower and upper dashed cuts in (d). (d) Series of transmission scans through the 00ZPL resonance of $M1$ at 15 different axial positions of the sample over an estimated total distance of $3.5\mu \mathrm{m}$ (see the Supplemental Material [25]). All data in this figure were recorded in the weak excitation regime [0.04 nW, see Figs. 3 and 3]. (e) Calculated transmission spectra to fit the data in (d). See the Supplemental Material [25] for details.