Figure 1

Calculated radiative decay rate for a dipole emitting at $\lambda =615\mathrm{n}\mathrm{m}$ inside a dielectric sphere ($n=1.59$) placed in vacuum. The rates have been averaged over different dipole orientations and normalized to the radiative decay rate in the bulk. The horizontal axes display the diameter $2R$ of the sphere and the emitter position in it. Note that some of the sharp resonances have been truncated.Reuse & Permissions

Figure 2

(a) A scanning electron microscope (SEM) image of the colloids on a glass substrate. (b) The schematic arrangement of the AFM and confocal microscope with a scanning stage. The tip is withdrawn by a few $\mu \mathrm{m}$ when the optical image is recorded. (c) A typical fluorescence confocal image of a sample containing 100 nm colloids. (d) The AFM image of the same part of the sample resolves aggregates not identified in (c). Note that the slightly extended shape of the spots in (d) is due to the convolution with an imperfect tip.Reuse & Permissions

Figure 3

(a) Experimental raw data for the fluorescence decay rates as a function of the sphere diameter. Colors represent different surrounding media (green: air; blue: water; red: immersion oil). The inset shows the difference between the fluorescence decay signal of a single 100 nm sphere and a normalized single exponential fit function. (b) The radiative decay rates extracted from the data in (a) are normalized to the bulk rate and fitted by the theoretical curves. For larger particles in air the theoretical curve has been truncated. The inset nicely reveals that a slight spread in the diameter results in a large variation of the rates due to Mie resonances for larger spheres.Reuse & Permissions