Optimizing plasmon-enhanced fluorescence with nonlocal metallic nanospheres

The fluorescence signal η of molecules coupled to plasmonic nanoparticles (NPs) is optimized through extended simulations, taking the metal nonlocal optical response fully into account. Solid Au and Ag nanospheres, as well as SiO2/Au(Ag) core/shell NPs (of total radius R), are engineered to maximize the ratio of radiative to nonradiative losses and match the emitter emission wavelength, λem. For a molecule modeled as an electric dipole p, oriented parallel to the incident field E, the optimal emitter-NP distance is then identified within the local response approximation (LRA). Introducing nonlocal corrections, the plasmon blueshift predicted by the hydrodynamic Drude model [1] leads to a small reduction of η. If however the plasmonic mode does not coincide exactly with λem, this blueshift can tune the mode to increase η. Nevertheless, when size-dependent losses are also taken into account, through the recent Generalized Nonlocal Optical Response (GNOR) theory, the concurrent contribution of modal shifts and nonradiative losses, together with a reduced emitter excitation rate due to the decreased field intensity, lead always to a strong reduction of fluorescence (see Fig. 1). Finally, we identify situations where the common, intuitive recipe of tuning the NP modes to match λem can in fact lead to strong fluorescence quenching, instead of the anticipated enhancement. Our results highlight the necessity for careful modeling and design of plasmon-field-enhancement based applications.