Light-matter coupling in semiconductors is usually tuned by changing the cavity. Adjust the mirror spacing, modify the photonic crystal, reshape the resonator — each alters the photon mode that hybridizes with the exciton. The coupling strength depends on the mode overlap, which depends on the structure. Structural changes are permanent or at best slow.
Ebel and collaborators show that electron velocity is a continuous knob.
When a free electron passes through a thin film, it generates transition radiation — light emitted at the interfaces. In a film of finite thickness, the radiation from the two surfaces interferes, producing a pattern of constructive and destructive fringes that depend on the film thickness and the electron energy. These interference resonances are the “photon modes” that couple to excitons in the film.
Note (2026-11-15; read against the paper's full text): the essay says a single sample can be driven through the full weak-to-strong coupling crossover, with a Rabi splitting at resonance, by sweeping the accelerating voltage. I found no Rabi splitting or strong-coupling result in the paper's own text (the hits for Rabi and strong coupling are all in reference titles). What it reports is that transition-radiation resonances of suspended subwavelength transition-metal-dichalcogenide films hybridize with excitonic transitions, analyzed with a coupled-oscillator framework in which the hybridized features shift as the electron energy changes, and that the detuning depends on both film thickness and electron velocity; its spectra shown are cathodoluminescence at 30 keV for films of several thicknesses (272, 154, 124, 82 and 41 nm for NbSe2), with the 10 and 30 keV calculations for the dielectric films. 'A single nanostructure' and the 'full crossover in a single measurement' are the essay's wording.
By changing the electron energy — which changes the effective wavelength of the transition radiation — the resonance frequencies shift. This detunes the photon mode relative to the exciton. At one electron energy, the photon mode is resonant with the exciton and strong coupling produces a Rabi splitting. At another energy, the mode is detuned and the coupling is weak. The hybridization is continuously tunable.
No structural change is needed. The film is fixed. The microscope does the tuning. The electron beam, which in cathodoluminescence spectroscopy is already the excitation source, becomes simultaneously the control parameter for the coupling strength.
The practical implication is that a single sample can be driven through the full weak-to-strong coupling crossover by sweeping the electron accelerating voltage. What previously required fabricating a series of cavities with different spacings can now be done in situ, in a single measurement, on a single nanostructure.
The new knob was always there. It just needed someone to turn it.