Every frequency parks at its own address, on silicon and in aluminium

What the two devices do

Start with the part that has to work before anything else can. Vibrations travel through a solid as phonons, and left alone they spread out in a disordered, undirected way. Both groups patterned their material with microscopic features — tiny triangular pillars among them — chosen so that a phonon crossing the pattern has its speed and direction changed, as though an electromagnetic field were nudging it. The idea is imported: it is how the motion of electrons in a thin material is constrained by immersing them in a field. Applied to phonons, it groups them by frequency and then holds each group at a different place, which is exactly the sorting an optical rainbow performs on light.^[1]

The two builds are not the same machine. Riyi Zheng and colleagues at the South China University of Technology trapped their rainbow on a silicon chip and then used it: an incoming elastic wave was routed through a single frequency patch, the way you would guide light through one band of colour. Yafeng Chen and colleagues at Tongji University built theirs in an aluminium device, made the rainbow out of ultrasound above human hearing, and imaged it by shining a laser on the surface and recording the vibrations that came back. One result is a routing function, the other is a picture. Both papers went through peer review at Physical Review Letters.^[1]

What the second team buys you

A single device that does something surprising leaves an obvious question open: how much of the behaviour belongs to the physics and how much to that particular slab of material and the run that produced it. Two groups, on silicon and on aluminium, at different scales and with different excitations, close a good deal of that gap. What is left is a recipe another laboratory can follow, and that is a different kind of object from a result. The honest caveat is that both teams took the same borrowed idea from electron physics, so their agreement partly reflects a shared model rather than two independent probes of the world; the difference in material and in scale narrows that concern without removing it. The test that would settle it is a third build that departs from the electron analogy and still produces the rainbow.^[1]

The two quantities that are still missing

Henning Schomerus at Lancaster University prices the remaining distance, and the price comes in two items: the devices would have to be made smaller, and the vibrational energy held inside the rainbow would have to go up. That is a useful kind of criticism, because both items are numbers rather than doubts. Chen says his team is already working in that direction and points at ultra-compact on-chip acoustic devices as the target. So the signal worth watching is narrow: a following paper that reports an elastic-rainbow device together with its dimensions and the fraction of incoming vibrational energy the rainbow actually holds. If those two figures appear within the next year, vibration filters and ultrasonic routers become an engineering schedule. If the papers keep arriving with rainbows and without either figure, what exists is an effect two laboratories have now produced, which is already more than there was last week.^[1]