Research data supporting "Boosting Optical Nanocavity Coupling by Retardation Matching to Dark Modes"

Data necessary to reproduce the graphs in the associated publication. For the data in Figures 1c,d / 2b,c,d / 4a, / 5a Full-wave 3D simulations are performed using Lumerical FDTD solutions. The AuNP is modelled as a truncated sphere (with a facet width of 20 nm) of radius 40 nm on top of an infinite dielectric sheet of the refractive index of ng = 1.45 and a gap size of 1.3 nm matching the BPT thickness. (29) The thickness of the Au slab placed below the BPT layer is infinite to the perfectly matching layer, and the AuNP is embedded into a dielectric film of different heights and refractive index (nd) as mentioned in the text. The NPoM geometry is illuminated with a plane wave with polarization either perpendicular or parallel to the metal surface to access different sets of modes. For estimating field enhancements, a 2D near-field monitor is placed at the center of the nanogap. To extract the respective field strengths for each different mode, the near-field spectrum at the field maximum is extracted with multipeak fitting for that resonating mode wavelength. Data for Figures 3, 4b,c,d were collected using an in-house built Raman/darkfield microscope. In short an Olympus upright microscope is fibre coupled to an Ocean Insight broad range spectrometer for dark field scattering spectroscopy, free space coupled to a Horiba spectrometer paired with an Andor EMCCD. For K-space imaging and energy momentum spectroscopy a Bertrand lens is introduced into the collection path to project the back focal plane in the CCD instead. All experimental data were collected using a 100x DF Olympus objective with a 0.9NA. Peak fitting was performed using either the Gaussian fit function or multi-gaussian fit function as implemented in Wavemetrics' Igor programme. For darkfield spectra a z-scan is performed (collecting spectra over a range of different focus positions), and the highest intensity for each wavelength position is extracted using a gaussian fit, and combined reconstructing a chromatic aberration corrected darkfield spectrum. For the average and variance depicted in Fig. 3 a standard deviation and mean intensities are calculated from chromatic aberration corrected darkfield spectra taken from 1550, 313, 438, 2235 Nanoparticle-on-mirror geometries respectively.