Search

Transition-metal dichalcogenides with their exciton-dominated optical behavior emerge as promising materials for realizing strong light-matter interactions in the visible range and at ambient conditions. When these materials are combined with metals, the energy confining ability of plasmon polaritons in metals below the diffraction limit, allows for further enhancing and tailoring the light-matter interaction, due to the formation of plexcitons in hybrid metal-TMDC structures at the interface. Herein, we demonstrate that the coupling between quasi-propagating plasmons in plasmonic crystals and excitons in WSe2, provides a multi-oscillator playground for tailoring the band structure of plasmonic crystal structures and results in emerging flat bands. The cathodoluminescence spectroscopy and angle-resolved measurements combined with the numerically calculated photonic band structure confirm a strong exciton-plasmon coupling, leading to significant changes in the band diagram of the hybrid lattice and the ability to tailor the band diagram via strong coupling. The hybrid plexcitonic crystal structures investigated here sustain optical waves with remarkably low group velocities. These results could be used for designing tunable slow-light structures based on the strong-coupling effect and pave the way toward plexcitonic topological photonic structures.

In this paper we describe the technical implementation of pump-probe cathodoluminescence microscopy (PP-CL), a novel technique for studying ultrafast dynamics in materials using combined excitation of electron and laser pulses. The PP-CL setup is based on a commercial scanning electron microscope combined with a femtosecond laser. Here we discuss the design, technical details, and characterization of the PP-CL setup. The luminescence emitted by the sample (either photoluminescence, cathodoluminescence or a combination of both) can be analyzed in terms of spectrum, temporal statistics, and frequency-modulated intensity (lock-in detection). The laser spot on the sample is characterized by analyzing the changes in the secondary electron signal on a GaN substrate, from which we extract a laser spot size in the $\mu m$ range. While this method ensures spatial overlap of laser and electron pulses on the sample, temporal alignment is achieved using time-correlated measurements of photoluminescence and cathodoluminescence. We also discuss some considerations to take into account when designing a PP-CL experiment, regarding temporal and spatial resolution. This paper provides a detailed reference for the development of any PP-CL microscope.

Photon bunching in incoherent cathodoluminescence (CL) spectroscopy originates from the fact that a single high-energy electron can generate multiple photons when interacting with a material, thus revealing key properties of electron-matter excitation. Contrary to previous works based on Monte-Carlo modelling, here we present a fully analytical model describing the amplitude and shape of the second order autocorrelation function ($g^{(2)}(\tau)$) for continuous and pulsed electron beams. Moreover, we extend the analysis of photon bunching to ultrashort electron pulses, in which up to 500 electrons per pulse excite the sample within a few picoseconds. We obtain a simple equation relating the bunching strength ($g^{(2)}(0)$ to the electron excitation efficiency ($\gamma$), electron beam current, emitter decay lifetime and pulse duration, in the case of pulsed electron beams. The analytical model shows good agreement with experimental data obtained on InGaN/GaN quantum wells using continuous, ns-pulsed (using beam blanker) and ultrashort ps-pulsed (using photoemission) electron beams. We extract excitation efficiencies of 0.13 and 0.05 for 10 and 8 keV electron beams, respectively, and we observe that non-linear effects play no compelling role even after excitation with ultrashort and dense electron cascades in the quantum wells., Comment: 29 pages (including supplementary information), 14 figures (including supplementary figures)

Nitrogen-vacancy (NV) centers in diamond are reliable single-photon emitters, with applications in quantum technologies and metrology. Two charge states are known for NV centers: NV$^0$ and NV$^-$, with the latter being mostly studied due to its long electron spin coherence time. Therefore, control over the charge state of the NV centers is essential. However, an understanding of the dynamics between the different states still remains challenging. Here, conversion from NV$^-$ to NV$^0$ due to electron-induced carrier generation is shown. Ultrafast pump-probe cathodoluminescence spectroscopy is presented for the first time, with electron pulses as pump, and laser pulses as probe, to prepare and read out the NV states. The experimental data is explained with a model considering carrier dynamics (0.8 ns), NV$^0$ spontaneous emission (20 ns) and NV$^0$ $\rightarrow$NV$^-$ back transfer (500 ms). The results provide new insights into the NV$^-$$\leftrightarrow$NV$^0$ conversion dynamics, and into the use of pump-probe cathodoluminescence as a nanoscale NV characterization tool.

Cathodoluminescence spectroscopy (CL) allows characterizing light emission in bulk and nanostructured materials and is a key tool in fields ranging from materials science to nanophotonics. Previously, CL measurements focused on the spectral content and angular distribution of emission, while the polarization was not fully determined. Here we demonstrate a technique to access the full polarization state of the cathodoluminescence emission, that is the Stokes parameters as a function of the emission angle. Using this technique, we measure the emission of metallic bullseye nanostructures and show that the handedness of the structure as well as nanoscale changes in excitation position induce large changes in polarization ellipticity and helicity. Furthermore, by exploiting the ability of polarimetry to distinguish polarized from unpolarized light, we quantify the contributions of different types of coherent and incoherent radiation to the emission of a gold surface, silicon and gallium arsenide bulk semiconductors. This technique paves the way for in-depth analysis of the emission mechanisms of nanostructured devices as well as macroscopic media., Comment: 8 figures. Includes supplementary information

We use spatially and angle-resolved cathodoluminescence imaging spectroscopy to study, with deep subwavelength resolution, the radiation mechanism of single plasmonic ridge antennas with lengths ranging from 100 to 2000 nm. We measure the antenna's standing wave resonances up to the fifth order and measure the dispersion of the strongly confined guided plasmon mode. By directly detecting the emitted antenna radiation with a 2D CCD camera we are able to measure the angular emission patterns associated with each individual antenna resonance. We demonstrate that the shortest ridges can be modeled as a single point dipole emitter oriented either upward (m=0) or in-plane (m=1). The far-field emission pattern for longer antennas (m>2) is well described by two interfering in-plane point dipoles at the end facets giving rise to an angular fringe pattern, where the number of fringes increases as the antenna becomes longer. Taking advantage of the deep subwavelength excitation resolution of the cathodoluminescence technique, we are able to determine the antenna radiation pattern as function of excitation position. By including the phase of the radiating dipoles into our simple dipole model we completely reproduce this effect. This work demonstrates how angle-resolved cathodoluminescence spectroscopy can be used to fully determine the emission properties of subwavelength ridge antennas, which ultimately can be used for the design of more complex and efficient antenna structures., Comment: 25 pages, 7 figures

We present measurements on plasmonic meta-molecules under local excitation using cathodoluminescence which show a spatial redistribution of the local density of optical states (LDOS) at the same frequency where a sharp spectral Fano-feature in the extinction cross section has been observed. Our analytical model shows that both near- and far-field effects arise due to interference of the same two eigenmodes of the system. We present quantitative insights both in a bare state, and in a dressed state picture that describe plasmonic Fano interference either as near-field amplitude transfer between three coupled bare states, or as interference of two uncoupled eigenmodes in the far field. We identify the same eigenmode causing a dip in extinction to strongly enhance the radiative LDOS, making it a promising candidate for spontaneous emission control.