📄 Showing the "Simple" version — "Popular" is not ready yet. Add it to favorites to help prioritize it. Ask for the full read

Light is trapped in a tiny cavity and circles around like a ball in a groove. If you slightly change the material inside, the light notices immediately, and gives a signal. Usually, such sensors work like spring scales: the harder you press, the more noticeable the shift. But there's another, trickier approach: make even a weak perturbation produce a disproportionately large response.

Previously, this was done using exceptional points in non-Hermitian systems — where two states merge into one. There, sensitivity grows as the square root of the perturbation, not linearly. But these devices have a problem: noise amplifies along with the signal, and the so-called Petermann factor ruins everything. It diverges at the exceptional point, and the advantage melts away.

The new work proposes a way around this barrier. The authors took a simple chiral cavity — a regular resonator where light travels in only one direction, like a one-way street. No phase transitions, no exotic physics. But when the refractive index changes, the response is the same as at an exceptional point: the signal grows as the square root of the perturbation.

Why does this work? In a chiral cavity, the eigenstates remain orthogonal, like two perpendicular streets that don't interfere with each other. So noise doesn't amplify catastrophically, and the Petermann factor doesn't blow up. The sensor gains sensitivity without losing stability.

This approach breaks the previous limitation: it seemed that ultra-sensitivity in non-Hermitian sensors would always hit a noise wall. Now there's a path to sensors that are sharper than both those at exceptional points and those at ordinary points. Light trapped in a simple cavity turns out to be the best watchdog.