An atom fell along two paths — and its quantum phase followed Einstein
An interferometer using about 20,000 rubidium atoms measured the phase of a freely falling matter wave and found the predicted time dependence; the test does not quantize gravity.

How can a fall seen from a laboratory be compared with the same fall seen by someone who is falling? An international team turned Einstein’s question into an atom interferometer. Along one path, the wave associated with each rubidium-87 atom remained nearly stationary relative to Earth; along the other, it received a brief upward kick and then followed a parabola under gravity. When the two paths were reunited, the researchers measured the accumulated phase difference. The result, published September 2 in Science Advances, followed the time dependence predicted by the equivalence principle in the regime tested.
The experiment began with a Bose–Einstein condensate of approximately 20,000 atoms, collimated to an effective temperature of 3.3 nanokelvins. The cloud sat about 113 micrometers below an atom chip. Microwave pulses placed every atom in a superposition of two internal states. Because those states respond differently to magnetic fields, microscopic wires on the chip could support one wave packet against gravity while launching the second onto a ballistic path. In the example described in the paper, the maximum separation reached 7.5 micrometers, roughly seven times the packet size.
The decisive quantity was not the fall distance but phase—the position within a quantum wave’s cycle. When the packets again overlapped in position and velocity, a pulse recombined them and produced interference. The fraction of atoms detected in each state revealed the relative phase. As the flight time changed, the team observed the cubic signature: phase grew with the cube of time, as predicted for the transformation between the laboratory frame and a freely falling frame. The data covered nearly 80 radians, about 13 interference oscillations, in measurement sessions lasting three to six hours.
That cubic growth relies on equating inertial mass, which measures resistance to acceleration, with gravitational mass, which determines the response to Earth’s field. Their equality is at the heart of the equivalence principle. In a frame falling with the ballistic packet, gravity locally disappears; in the laboratory, the same evolution appears as an added phase. The new device, called the Quantum Galileo Interferometer, directly compares those two frames without using light pulses as mirrors for the atomic paths. Magnetic fields manipulate the packets but need not remain resonant with an atomic transition during the fall.
An idealized analytical model and a simulation incorporating three-dimensional motion, imperfect recombination, packet expansion and atom–atom interactions reproduced the data. Residuals between the simulation and the measurements were about 2.5%. The leading systematic uncertainty came from the ratio of the currents used to kick and support the packets: a small mismatch adds a phase that grows linearly with time. The agreement therefore confirms the predicted phase for this apparatus and range of parameters; it does not survey every possible violation of equivalence with the precision of the best classical tests.
The result also does not show that gravity is quantized. Earth’s gravitational field entered the experiment as a classical background; it was neither placed in superposition nor measurably changed by the atoms. The novelty is narrower and concrete: a matter wave in superposition retained the phase relation required by the equivalence of gravity and acceleration. The measurement connects two descriptions usually studied separately, but by itself supplies no common theory of general relativity and quantum mechanics.
The architecture opens an experimental route to harder questions. The authors propose manipulating waves that act as quantum clocks and eventually replacing atoms with far more massive nanoparticles. Such systems could make mass itself a consequential participant in the superposition and probe proposals for gravity-induced collapse. Before that, the Quantum Galileo Interferometer already provides a measurable benchmark: when an atom occupies a supported path and a freely falling path at once, the phase between them evolves as Einstein and quantum mechanics jointly predict.
Key points
- About 20,000 rubidium-87 atoms were split between a supported path and a freely falling path.
- The relative phase grew with the cube of time and reached nearly 80 radians; model residuals were close to 2.5%.
- The test confirms the equivalence prediction in this setup, but it neither quantizes gravity nor unifies the two theories.

Comments
No comments have been published yet.
Sign in with a subscription to comment.