Hyperbolic phonon polaritons, often abbreviated HPhPs, are among the most remarkable optical excitations known to materials science. They arise when infrared electromagnetic fields couple strongly to the natural lattice vibrations, or optical phonons, of certain crystals. The resulting hybrid waves behave in ways that ordinary light cannot: they can squeeze electromagnetic energy into volumes thousands of times smaller than the wavelength of the light itself, guiding it coherently along a material’s surface or through its bulk. This extreme confinement makes them attractive candidates for nanoscale light manipulation, high-sensitivity molecular spectroscopy, and the miniaturization of optical circuits to dimensions that rival modern microelectronics.

Hexagonal boron nitride, or hBN, a layered van der Waals crystal famous for its role as a protective encapsulant in two-dimensional electronics, is a particularly powerful host for these waves. Within specific infrared frequency bands, hBN behaves as a hyperbolic medium, meaning that its dielectric response is positive along one crystal axis and negative along another. In this regime, the polariton wavefronts take on hyperbolic shapes, and the waves can propagate with wavelengths dramatically shorter than free-space light, carrying electromagnetic energy along the crystal with remarkable confinement. But harnessing such waves in practical devices requires more than passive propagation; engineers need the ability to switch, steer, and reshape them on demand, and to do so at speeds fast enough to matter for information processing.

That is precisely where the experimental challenge has always lain. Hyperbolic phonon polaritons are strongly dispersive, which means their wavelength changes substantially depending on the frequency of the light that excites them. Femtosecond laser pulses, the only tools capable of resolving dynamics on the relevant timescales, are inherently broadband: a single pulse contains a wide spread of frequencies. When such a pulse strikes a sample, it excites polaritons of many different wavelengths simultaneously. In a real-space image, the interference fringes produced by all these waves overlap and average together, washing out the very propagation patterns scientists want to see. The obvious remedy, narrowing the spectrum of the pulse before it reaches the sample, comes at a cruel price: by the fundamental rules of Fourier optics, a narrower spectrum demands a longer pulse, which destroys the femtosecond time resolution that motivated the experiment in the first place.

A research team led by Kazuki Kamada of the Institute for Molecular Science (IMS) and Osaka Metropolitan University, together with Jun Nishida, Assistant Professor at IMS, and Takashi Kumagai, Associate Professor at IMS, has now found an elegant way around this trade-off. Their solution was to leave the excitation pulse untouched and instead perform the frequency selection after the fact, during detection. By inserting a diffraction grating into the detection path of an ultrafast infrared near-field optical microscope, the team spectrally separated the broadband infrared light scattered from the sharp metallic tip of an atomic force microscope before it reached the detector. Because each frequency component could then be imaged independently, the overlapping interference patterns were cleanly disentangled without ever lengthening the excitation pulse.

The resulting instrument is a genuine tour de force of measurement engineering. It combines nanoscale spatial resolution, inherited from the near-field interaction between the microscope tip and the sample, with a temporal resolution of approximately 150 femtoseconds and a spectral resolution of roughly 10 inverse centimeters. In practical terms, the technique allows researchers to select a single narrow slice of the infrared spectrum, watch the polariton associated with that frequency propagate across the sample in a snapshot lasting only a few hundred atomic vibrations of time, and then repeat the measurement at successive pump-probe delays to assemble a full movie of the wave’s evolution. No previous near-field imaging scheme had brought all three of these capabilities together for strongly dispersive polaritons.

With the new microscope in hand, the team turned it on van der Waals heterostructures composed of tungsten disulfide (WS2) and hexagonal boron nitride. The experimental geometry was a classic pump-probe arrangement: a pulse of visible light generated photoexcited charge carriers, mobile electrons and holes, within the WS2 layer, while a time-delayed infrared near-field pulse probed how those carriers affected the hyperbolic phonon polaritons propagating in the adjacent hBN. Because the two atomically thin materials lie in intimate contact across the van der Waals interface, any change in the optical properties of one should, in principle, be felt by the waves traveling in the other. The question was whether the effect would be fast enough and strong enough to see directly.

The answer was a clear yes, and the details proved richer than a simple on-off switching picture. In heterostructures containing a thin WS2 layer, the researchers observed a transient modulation dominated by a change in the electric-field amplitude of the polaritons, as if the photoexcited carriers were momentarily absorbing or damping the wave. More strikingly, in structures containing a relatively thick WS2 layer, the team resolved a change in the polariton’s wavelength itself. This is a qualitatively different and more powerful form of control: it demonstrates that optical excitation can modify not merely how strongly the wave is present but how it propagates, effectively rewriting the dispersion relation of the polariton on the fly. Electromagnetic simulations confirmed that both observations could be explained by a transient, carrier-induced change in the dielectric response of WS2, which then modifies the HPhPs propagating in the neighboring hBN through the shared interface.

The physical mechanism is worth appreciating in detail. When visible light excites electron-hole pairs in WS2, the free carriers alter the material’s dielectric function, particularly at infrared frequencies relevant to the hBN polaritons. Since the hyperbolic phonon polariton is a hybrid mode whose electromagnetic field extends across both layers of the heterostructure, this sudden change in the WS2 dielectric environment is imprinted directly onto the wave. The carriers then relax and recombine on their own timescale, and the polariton returns to its unperturbed state. The entire cycle, from optical switching to recovery, unfolds within a few hundred femtoseconds, which is exactly the regime the new imaging technique was designed to capture. In effect, the researchers have demonstrated an all-optical gate for nanoscale light, in which one beam of light controls another through the mediation of photoexcited carriers and a van der Waals interface.

The implications extend well beyond this particular material system. The study, published online in Nano Letters on July 27, 2026, establishes a general platform for observing and controlling nanoscale light propagation on femtosecond timescales, and the same imaging strategy should apply to other strongly dispersive polaritons, including surface plasmon polaritons in metals and phonon polaritons in other hyperbolic crystals. As the authors note, the results open new opportunities for ultrafast nanophotonics, a field in which devices might one day modulate, route, and process signals encoded in confined light fields at speeds approaching the fundamental limits set by the optical oscillations themselves. The work was supported by JSPS KAKENHI, the JST FOREST and CREST programs, the MEXT World Premier International Research Center Initiative, and several other Japanese research foundations, and it involved collaborators from the National Institute for Materials Science, the Fritz Haber Institute in Berlin, and Osaka Metropolitan University. For now, the immediate contribution is conceptual and methodological: a way to see, frame by femtosecond frame, how light confined to the nanoscale can be commanded to change its shape, its speed, and its very wavelength, all at the whim of a flash of visible light.

Subject of Research: Ultrafast nano-imaging and optical control of hyperbolic phonon polaritons in van der Waals heterostructures

Article Title: Capturing fleeting changes in “nanoscale light”—femtosecond nano-imaging reveals ultrafast optical control of phonon polariton

Article References: Capturing fleeting changes in “nanoscale light”—femtosecond nano-imaging reveals ultrafast optical control of phonon polariton. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: phonon polaritons, hexagonal boron nitride, tungsten disulfide, femtosecond imaging, near-field microscopy, nanophotonics, van der Waals heterostructures, ultrafast optics, pump-probe spectroscopy, dielectric response, Nano Letters, light-matter coupling