Hyperbolic phonon polaritons -- hybrid light-vibration excitations in polar crystals -- propagate with extreme directionality determined by the crystal's anisotropy. Jackering, Saldarelli, Moos, and colleagues demonstrate that these polariton pathways can be rapidly programmed by optically writing nanostructures into a phase-change material substrate. By placing alpha-MoO3 flakes on In3SbTe2, they use focused laser pulses to switch the substrate's phase locally, creating launching and confining structures for the polaritons without any lithographic fabrication.
The critical technical requirement is alignment. Hyperbolic phonon polaritons in alpha-MoO3 propagate preferentially along the crystal's [001] axis, and any launching structure must be precisely oriented relative to this axis to direct the polaritons effectively. Traditional lithographic approaches require careful crystallographic characterization before fabrication. Optical programming eliminates this constraint: the writing laser can be aligned to the crystal axis in real time using the polariton signal itself as feedback. The researchers demonstrate optically programmed disk launchers that perform comparably to gold disk launchers fabricated by electron-beam lithography, and they show that a single disk can be reconfigured into a double-disk cavity by modifying the phase-change pattern, tuning the confinement without starting over.
Note (2026-11-10): I checked the abstract and searched the full text of arXiv 2603.10900. The abstract says that programming after flake deposition enables alignment of the launching stripes to the [001] axis of the flake, that an optically programmed disk shows similar tuning ranges and confinement to gold disks, and that a single disk was reconfigured into a double-disk cavity. The essay adds two things I did not find: that the writing laser is aligned in real time using the polariton signal as feedback, and that optically written circuits can be erased and rewritten. The paper describes phase-change materials in general as reversibly switchable between amorphous and crystalline phases, but the reconfiguration it demonstrates is further crystallization to add or change structures; I did not find a demonstration of erasing.
The reconfigurability is the structural advance. Lithographic polariton circuits are fixed at fabrication; optically written circuits can be erased and rewritten. This transforms phonon polariton optics from a design-once discipline into something closer to a programmable platform, where the same physical crystal can host different optical circuits at different times. The phase-change substrate functions as rewritable memory for nanophotonic architecture.
(arXiv:2603.10900)