A mirror-based experiment could finally answer one of the oldest open questions in theoretical physics: is gravity a quantum force? Researchers from Caltech, Huazhong University of Science and Technology, and Tokyo Institute of Technology have published a proposal in Physical Review D describing how to test the Schrödinger-Newton equation, a 44-year-old prediction about how quantum systems interact with their own gravitational field. Phys.org reported on the paper this week. If the experiment works and the signal appears, it would immediately rule out several competing theories of quantum gravity. If the signal does not appear, it would rule out others. Either way, the theoretical landscape of fundamental physics gets narrower.
The Schrödinger-Newton equation emerged from two separate theoretical threads that converged in the early 1980s. Roger Penrose proposed in 1975 that a massive quantum system in superposition should generate a gravitational field that is itself in superposition, and that this self-gravity would cause the quantum state to collapse. Ludwig Mantica and others formalized this into what became known as the Schrödinger-Newton limit: a description of quantum matter under the influence of its own gravitational field that sits uneasily between full quantum mechanics and classical general relativity. For four decades, nobody could design a tabletop experiment to distinguish it from competing theories. The new paper changes that.
The mechanism exploits an effect that is usually a nuisance in precision interferometry. When two suspended mirrors move independently, their motions are not perfectly correlated. In gravitational wave detectors like LIGO, this differential motion is noise. The researchers propose to use it as the signal. By positioning two mirrors asymmetrically within a Michelson-type interferometer, they create cross-talk between the common motion of both mirrors and their differential motion relative to each other. This cross-talk imprints itself as a binary yes-or-no signature in the correlation of output light. Classical gravity predicts one outcome. Quantum gravity predicts another.
The authors claim the signature is detectable with current technology, provided two conditions are met. The experiment requires input light with quantum noise reduced by 10 decibels using squeezed states, a technique well-established in gravitational wave detection. It also requires operation at roughly one Kelvin, warm by quantum standards but achievable with commercial cryostats. Under these conditions, three hours of aggregated data should yield sufficient signal-to-noise ratio to distinguish the two predictions. In a real laboratory, decoherence from seismic vibrations, cosmic rays, and electromagnetic interference could extend the required integration time or introduce systematic errors that mimic one signal or the other.
The binary readout is both the proposal's strength and its vulnerability. A clean yes-or-no answer is much easier to interpret than a quantitative measurement requiring precise theoretical input. But three hours is the figure under ideal isolation. The paper models this in simulation; the experiment has not yet been built.
If the result holds, the implications cascade across theoretical physics. A confirmed Schrödinger-Newton signal would falsify semi-classical gravity as a complete description of quantum matter, and simultaneously eliminate every quantum gravity approach that predicts different low-energy signatures. A null result would be equally significant, showing that gravity remains classical at scales where quantum effects should appear and demanding new theoretical frameworks that most current approaches do not anticipate.
Whether either outcome actually arrives depends on which quantum gravity frameworks survive contact with the result. String theory makes few falsifiable predictions at the Schrödinger-Newton scale. Loop quantum gravity has not yet generated consensus on its low-energy phenomenology. Whether either approach has a stake in this particular experiment depends on theoretical work that has not yet been published. What the paper demonstrates is that the question is no longer permanently beyond experimental reach. Forty-four years after the Schrödinger-Newton equation first appeared, the experimental community has a concrete protocol with specific hardware requirements and a defined readout time. The gap between theory and experiment has narrowed. What happens next depends on who decides the question is worth the engineering cost of answering.