To appreciate why the result matters, it helps to recall what a valley is. In many semiconductors, the conduction and valence band extrema do not sit at a single point in momentum space but appear at several symmetry-related locations, or valleys. Because these valleys are distinct pockets in the electronic band structure, an electron occupying one of them carries an extra quantum label beyond its charge and spin. Valleytronics proposes to exploit that label in much the same way conventional electronics exploits charge or spintronics exploits spin. The catch has always been control: to write information into valleys, one must be able to populate some valleys and not others, and to read or preserve that population before scattering events wash it out.

The dominant platform for such experiments over the past decade has been the family of two-dimensional transition metal dichalcogenides, monolayers such as molybdenum disulfide and tungsten diselenide. There, broken inversion symmetry allows circularly polarized light to address one valley or the other selectively, a phenomenon known as valley-selective circular dichroism. But those materials pay a price: valley polarization in monolayer dichalcogenides typically decays rapidly through a combination of electron-hole exchange interactions, intervalley scattering and excitonic effects, and achieving near-unity polarization often requires heroic measures such as enormous magnetic fields or cryogenic temperatures. SnSe offers an entirely different route, one based not on circular polarization but on the crystal’s own pronounced in-plane anisotropy.

Tin selenide belongs to the group-IV monochalcogenides, compounds with a puckered, heavily distorted rock-salt structure that are strongly anisotropic within their layers. That anisotropy means the optical absorption of the crystal depends dramatically on the linear polarization direction of incoming light, a property known as linear dichroism. Crucially, theory has predicted that in these materials, linearly polarized light can selectively excite carriers into distinct, non-degenerate valleys depending on the orientation of the polarization axis. Rotate the polarization of the pump laser by ninety degrees, and you populate a different set of valleys. This is polarization-controlled valley addressing without any need for circular light or magnetic fields, and it works in principle even in bulk, layered crystals rather than fragile monolayers.

Demonstrating that prediction in the ultrafast regime, however, required a formidable experimental apparatus. The team employed time- and angle-resolved extreme-ultraviolet photoemission spectroscopy, a technique that uses femtosecond pump pulses to excite the material and delayed extreme-ultraviolet probe pulses to eject electrons from the sample. By measuring the energy and momentum of those photoelectrons with a momentum microscope, the researchers obtain what amounts to a movie of the electronic band structure, resolved in energy, momentum and time. Because the probe is polarization-tunable, the pump pulse could be set to address either of the two valley channels selectively, and the subsequent evolution of the valley populations could be tracked directly rather than inferred from optical signatures alone.

What the movie revealed is striking. When the pump polarization was set to excite carriers into the valleys corresponding to the global conduction band minima, the resulting valley polarization was essentially perfect, approaching unity, and, remarkably, it remained constant over the entire temporal window of the measurement. The carriers simply stayed where they had been put. This near-unity, time-independent polarization in a channel reached by a simple rotation of light polarization is exactly the kind of robust, addressable valley degree of freedom that valleytronic device concepts demand, and it was achieved without the extreme conditions that have been necessary in other material systems.

The second channel tells a very different and, in some ways, more intriguing story. When the pump polarization was rotated to excite the other valley channel, the initial valley polarization did not merely decay; it collapsed and reversed sign on subpicosecond timescales. The explanation, confirmed by time-dependent Boltzmann equation simulations built on first-principles calculations of the electron-phonon coupling, lies in the extraordinarily strong interaction between the excited electrons and a particular optical phonon mode of the lattice. Carriers deposited in that channel scatter between valleys so efficiently, mediated by this vibrational mode, that the population imbalance between valleys not only vanishes but overshoots, producing a transient reversal of the polarization before the system relaxes toward equilibrium.

This contrast between two coexisting channels is what the authors describe as strongly anisotropic and radically different non-equilibrium valley physics compared with the familiar two-dimensional valleytronics materials. In a monolayer dichalcogenide, valley depolarization is generally a nuisance to be suppressed. In SnSe, the two behaviors are built into the band structure itself, accessible at will through the polarization of light. That opens conceptual possibilities that go beyond simply storing bits in valleys: the ultrafast, phonon-mediated reversal channel could itself be harnessed, for instance in schemes where valley information is deliberately scrambled, converted or routed on femtosecond timescales, or where the coupling to a specific phonon mode acts as an ultrafast valve between valley populations.

The theoretical side of the work deserves emphasis as well. By solving the time-dependent Boltzmann equation with ab initio electron-phonon matrix elements, the team could reproduce the measured dynamics and identify the microscopic culprit behind the subpicosecond depolarization: intervalley scattering driven by strong coupling to an optical phonon. This marriage of momentum-microscopy experiments with nonequilibrium theory is becoming the standard for disentangling the many competing processes, carrier-carrier scattering, carrier-phonon scattering, and radiative recombination, that unfold simultaneously after photoexcitation. It also connects to earlier theoretical work by members of the team showing that thermal effects can enhance electron-phonon coupling in SnSe, hinting that the valley dynamics here may be tunable with temperature in ways that remain to be explored.

For the broader field, the study marks a coming of age for valleytronics beyond the dichalcogenide paradigm. Group-IV monochalcogenides, which include tin sulfide, germanium sulfide and their selenide counterparts, have long been predicted to support ferroelectric-like distortions, chiral phonons and rich valley physics, but direct, time-resolved, momentum-resolved evidence of their ultrafast valley dynamics had been lacking. The SnSe results now provide that evidence and suggest a design principle: strong in-plane anisotropy can be as powerful a resource for valley control as broken inversion symmetry, and it can come packaged with valley channels whose lifetimes differ by orders of magnitude within the same material. Whether robust polarization, ultrafast reversal, or some engineered combination of the two proves most useful, the crystal itself now offers both options, selected by nothing more exotic than the orientation of a laser pulse.

Subject of Research: Ultrafast valley polarization dynamics in the anisotropic semiconductor SnSe

Article Title: Ultrafast anisotropic valleytronics in SnSe

Article References: Pan, Y., Fragkos, S., Descamps, D., Petit, S., Caruso, F., & Beaulieu, S. (2026). Ultrafast anisotropic valleytronics in SnSe. Nature Materials. https://doi.org/10.1038/s41563-026-02765-6

Image Credits: AI Generated

DOI: 10.1038/s41563-026-02765-6

Keywords: valleytronics, SnSe, tin selenide, ultrafast dynamics, photoemission spectroscopy, electron-phonon coupling, intervalley scattering, valley polarization, group-IV monochalcogenides, linear dichroism, Boltzmann equation, Nature Materials