Scientists demo optical metamaterial control

Light can move matter. That is old physics. But most optical control at tiny scales works in one basic way — you pull particles or structures toward where the light is brightest. The new result is interesting because it breaks that one-way logic. A team at ETH Zurich built a metasurface system that lets light either pull or push a suspended nanomembrane, on command, by changing the geometry of the structure itself. ### What did they actually build? The device is a metasurface integrated onto a suspended silicon nanomembrane. A metasurface is basically a very thin sheet patterned with tiny engineered features — often called meta-atoms — that reshape how light interacts with the material. Here, those patterns do more than bend or filter light. They change the mechanical force the light exerts on the membrane. (arxiv.org) ### What is the new trick? The team showed deterministic control over the sign of the optical force. That is the key phrase. Usually, the question is how strong the force is. Here, the question is whether the force points one way or the opposite way. Their experiments showed both attractive and repulsive forces in different metasurface designs under the same broader optical setup. (arxiv.org) ### How can geometry flip a force? The short version is interference. Each meta-atom supports electromagnetic modes, and those modes combine with one another. By tailoring the geometry, the researchers tuned the coherent superposition of multipolar modes. That changed the momentum exchange between light and the metasurface, which changed the net force direction. So the force is not coming from a moving part or a feedback controller — it is baked into the optical response of the pattern. (arxiv.org) ### Why is that unusual? Most optical trapping and manipulation systems rely on matter being drawn toward intensity maxima — the bright regions of a field. That is the standard picture behind a lot of optical tweezing and microscale light control. This work shows a controlled departure from that behavior. In other words, bright light does not have to mean “come here.” With the right meta-optical design, it can also mean “move away.” (arxiv.org) ### Why use a standing wave? The metasurface sits in a phase-controlled optical standing wave, where the electromagnetic field has a stable spatial pattern. That gives the experiment a clean way to probe how the structure responds as the optical phase changes. The standing wave is part of what makes the force tunable and measurable rather than just a vague radiation-pressure effect. ### Is this like optical tweezers? (arxiv.org) Kind of — but it is not the same thing. Optical tweezers usually manipulate separate particles suspended in a medium. This platform manipulates a nanostructured mechanical element that is itself engineered to respond in a specific way. A decent analogy is the difference between pushing a loose ping-pong ball with air and designing a sail whose shape decides whether the wind stalls, pulls, or shoves. The structure is doing real work here. ### What could this be good for? The paper frames it as a platform for nanoscale mechanical control. That points toward optomechanical devices, on-chip actuators, precision sensing, and tiny systems where contact is a problem. The broader appeal is simple — if light can both pull and push micro- or nanoscale parts in a programmable way, you get a new control knob for devices too small for conventional mechanical handling. (arxiv.org) ### What is the real takeaway? This is not a finished product story. It is a control story. The advance is that meta-optics can program the direction of optical force itself, not just the path of light. That is a deeper capability — and it could matter anywhere tiny machines need frictionless, precise motion. (arxiv.org)