Table of Contents
Fetching ...

Velocity-tunable exciton-photon hybridization in cathodoluminescence

Sven Ebel, Martin Nørgaard, Christian Nicolaisen Hansen, N. Asger Mortensen, Sergii Morozov

Abstract

Exciton-photon hybridization is typically realised in geometrically defined optical cavities, where tunability is achieved by modifying either the cavity or the excitonic medium. Here we investigate transition-radiation interferences in suspended subwavelength films resembling a free-electron-defined resonance and explore their interaction with excitons in transition metal dichalcogenides. We demonstrate that these resonances hybridize with excitonic transitions and can be tuned continuously by varying the electron energy. The resulting detuning depends on both film thickness and electron velocity, establishing the latter as an external and continuous knob for exciton-photon coupling. This approach enables tunable hybridization without structural modification and provides a free-electron-driven nanoscale platform for studying exciton-light interactions.

Velocity-tunable exciton-photon hybridization in cathodoluminescence

Abstract

Exciton-photon hybridization is typically realised in geometrically defined optical cavities, where tunability is achieved by modifying either the cavity or the excitonic medium. Here we investigate transition-radiation interferences in suspended subwavelength films resembling a free-electron-defined resonance and explore their interaction with excitons in transition metal dichalcogenides. We demonstrate that these resonances hybridize with excitonic transitions and can be tuned continuously by varying the electron energy. The resulting detuning depends on both film thickness and electron velocity, establishing the latter as an external and continuous knob for exciton-photon coupling. This approach enables tunable hybridization without structural modification and provides a free-electron-driven nanoscale platform for studying exciton-light interactions.
Paper Structure (12 sections, 6 equations, 5 figures)

This paper contains 12 sections, 6 equations, 5 figures.

Figures (5)

  • Figure 1: Transition radiation resonances.(a) Experimental setup showing a parabolic mirror positioned in the electron beam path to collect the generated transition radiation (TR) and direct it to a spectrometer. (b) Schematic illustration of electron-velocity-dependent resonant TR in thin films of thickness $d$. (c) Dielectric function of a model semiconductor with a bandgap of 1.5 eV. (d) Calculated TR intensity for dielectric films described by the dielectric response shown in panel (c), plotted as a function of film thickness $d$ and emission wavelength for electron energies of 10 and 30 keV. (e) Optical image of a suspended 83 nm thick WS$_2$ crystal. The red dashed line marks the suspended region.
  • Figure 2: Transition radiation from NbSe$_2$ films.(a) Calculated TR intensity from NbSe$_2$ films as a function of film thickness $d$ and emission wavelength for a free-electron energy of 30 keV. (b) Experimental in-plane dielectric function of NbSe$_2$ as reported by Munkhbat et al.Munkhbat2022. (c) Experimental CL spectra at 30 keV and corresponding theoretical TR spectra for NbSe$_2$ films with thicknesses of 272, 154, 124, 82, and 41 nm. The selected thicknesses are indicated by triangles in panel (a). (d) Spectrally integrated experimental CL (pink diamonds) and theoretical TR (blue) intensities. In the experiments, the beam current was kept at 1.4 nA and the exposure time was 90 s.
  • Figure 3: Transition radiation from MoS$_2$ films.(a) Calculated TR intensity from MoS$_2$ films as a function of film thickness $d$ and emission wavelength for a free-electron energy of 30 keV. (b) Experimental in-plane dielectric function of MoS$_2$ as reported by Ermolaev et al.Ermolaev2021. (c) Experimental CL spectra at 30 keV and corresponding theoretical TR spectra for MoS$_2$ films with thicknesses of 131, 113, 85, 32, and 6 nm. The selected thicknesses are indicated by triangles in (a). (d) Spectrally integrated experimental CL (pink diamonds) and theoretical TR (blue) intensities. In the experiments, the beam current was kept at 1.4 nA and the exposure time was 90 s.
  • Figure 4: Electron-energy dependent hybridization of TR resonances and excitons in WS$_2$ films. Calculated TR intensity spectra from WS$_2$ films as a function of film thickness $d$ and emission wavelength for free-electron energies of (a) 30 keV and (b) 10 keV. Black dashed lines indicate the spectral positions of the A and B excitons in WS$_2$. White solid lines show coupled-oscillator fits to the calculated TR resonance maxima.
  • Figure 5: Tunable hybridization between TR resonances and excitons in WS$_2$ films.(a,e) Calculated electron-energy-dependent TR spectra for WS$_2$ films with thicknesses of 83 nm (a) and 100 nm (e). (b,f) Measured electron-energy-dependent CL spectra for WS$_2$ films with thicknesses of 83 nm (b) and 100 nm (f). (c,g) Extracted experimental spectral positions of the intensity maxima in the CL spectra (red circles). Black solid lines show solutions of the coupled-oscillator model (COM) illustrating the tunability of the TR–exciton hybridization with electron energy. (d,h) Experimental CL spectra from WS$_2$ films with thicknesses of 83 nm (d) and 100 nm (h) acquired at electron energies of 30, 20, 15, and 10 keV. The spectra are normalized around 600 nm and compared to calculated TR spectra of the same thickness and electron energy. During the measurements, the beam current was maintained at 1.4 nA, with exposure times of 90--120 s.