What happens when quantum mechanics and relativity meet?

Experiment put atoms in a superposition of trajectories to find out.

Almost a hundred years ago, physicists theorized out what free fall should do to a quantum wave. If the solution they came up with is wrong, quantum mechanics and Einstein’s theory of gravity flatly contradict each other. But testing it has been impossible because nobody has managed to build an interferometer that could perform the necessary measurement.

Now, a team led by Ron Folman, a physicist at Ben-Gurion University of the Negev, with collaborators in Germany, the UK, and the US, including Nobel laureate Roger Penrose, has done it. They built a new interferometer that gives a single atom two possible paths at once: one that involves a free fall, and another where it is held perfectly still. Both paths end at the same place at the same moment, allowing the team to measure what the fall does to a wave-like property of the atom.

Long time coming

Ever since Galileo, physicists have known how to describe a falling object—where it is, how fast it goes, or how quickly it accelerates. Quantum mechanics, though, insists that every object is also a wave. “Every particle, doesn’t matter if it’s a car or a spaceship or an atom, is a wave,” Folman says. “Everything that is a wave, like sea waves or sound waves, goes up and down. And if you’re up or down, this is measured by something called a phase. A phase just tells you if you are at the top of the wave or at the bottom.”

The wave nature of an atom shows itself only when the atom is barely moving, which only happens when it is cooled down to nearly absolute zero. For many years after theorists had first looked at this problem, this sort of temperature wasn’t an option—cooling atoms down to such temperatures only became possible in the late 1990s. “But this was just the start of the journey of this experiment,” Folman says.

The second obstacle is that a phase cannot be measured on its own—it appears only when you perform a comparison. “It can only be a relative measure done by splitting a single particle into two trajectories, and then bringing the two trajectories together,” Folman says. He told Ars that this is the same logic behind the double-slit experiment, where particles fired through two narrow slits toward a background screen form an interference pattern that reveals their phase.

Measuring the same thing in free fall adds another layer of complexity. “Drop a stone and after one second it is 5 meters below you,” Folman says. The question asked by the team was how to bring a wave packet that has fallen back into contact with the one that stayed behind—the same atom on the other path—so that the two can interfere. Finding the solution took his team years, even though in hindsight it sounds rather obvious.

The cannon and the parachute

Grabbing a falling particle and somehow bringing it back to where it started was out of the question. “Even if you could do it with some tweezer or some trap, this very violent action would create so much noise that we wouldn’t have known what is fake news and what is really the effect that we want to measure,” Folman joked. So the team figured the right way to do it would be to put an atom into a superposition of paths, where one path sees the atom thrown upward and has gravity return it to its original location. The other path isn’t really a path at all; the atom just stays where it started. “You have to shoot one with a cannon, into a ballistic motion like artillery,” Folman says. But that was just the start of the problems.

The challenge, obviously, is designing a cannon that may or may not affect an atom (technically, it only affects one wave packet of the atom without affecting the other one).

Then there is the issue of keeping the system in a quantum state. Interference only occurs when there is no way, even in principle, to tell which path the particle took—every measurement destroys the quantum superposition and forces the particle to behave like a localized classical object. Unfortunately, a particle returning from a ballistic flight is moving fast, while its stationary twin is not. That difference in speed is itself information about its path, so no interference pattern would appear.

Atom interferometers have existed for 30 years, but none could accomplish what Folman and his colleagues wanted to do. So, they built a new one called the Quantum Galileo Interferometer (QGI).

Atoms under a chip

The QGI is rather tiny and relies on microwave and magnetic pulses. It uses about 20,000 rubidium atoms maintained as a Bose-Einstein condensate, held by currents in microscopic wires on a chip that hangs upside-down roughly 113 micrometers above them. A microwave pulse puts every atom into superposition of two states, where one state feels magnetic forces, and the other is blind to them. A magnetic pulse then plays the role of the cannon, kicking any magnetically sensitive atoms upward.

This converts the superposition of states into a superposition of trajectories.

Immediately afterward, a second microwave pulse flips the atom into the opposite state, so the superposition that’s flying upward changes into a magnetically blind one and starts falling under gravity, with no other forces influencing it. At the same time, its stationary twin becomes sensitive to magnetism, but the field is tuned so its upward force exactly cancels the atom’s weight, effectively making it hover motionless with respect to the Earth.

As one of the trajectories falls back, a second magnetic pulse acts as a parachute, killing its speed just as it returns to where it started. In Folman’s experiment, at their farthest apart, the two trajectories were about 7.5 micrometers apart, and the longest flight lasted two-thousandths of a second.

At the end of each run, the atoms that landed in one of the interferometer’s two exits were sorted by the phase accumulated by the falling path. As the researchers lengthened the fall, this phase increased—a behavior in line with the theory.

“When somebody falls in free fall, he doesn’t feel his weight. He has no weight. Sometimes we feel [a bit of this] in an elevator that goes down very fast,” Folman says. In the quantum world, the falling path from the atom’s point of view experiences no gravity or any other force, and its wave should not pick up any extra phase. At the same time, from the laboratory’s point of view, gravity is pulling the same falling phase down, so its wave should pick up a specific amount of phase predicted by the theory. That phase difference did show up in the data.

“We were able to show that the behavior of a quantum system still follows the principle of equivalence,” Folman says. “This has never been shown before—that the equivalence principle can work in a quantum system in a superposition.”

In the end, the measured phase of the quantum wave sits 2.5 percent from the theoretical prediction. The result, the team argues, is a checkpoint on the border between Einstein’s relativity and quantum mechanics—two great theories in physics that stubbornly refuse to work together.

“This [incompatibility] is such a problem for physicists that very smart people say that it hints that at least one of the two theories is fundamentally wrong,” Folman says. Unfortunately, the QGI did not bring us much closer to figuring out which one it might be. But Folman and his lab already have a road map they hope will get us there one day.

The diamond experiment

The place to look for the answers, Folman argues, is in heavier masses. Today, his lab is building a larger version of the QGI experiment that puts nanodiamonds into the same superposition—objects 10 orders of magnitude heavier than atoms, massive enough to warp spacetime around themselves. “If we put the same nanodiamond in two places, we actually have a superposition of two different curvatures of spacetime,” Folman says. And this, according to Penrose’s hypothesis, developed independently by the Hungarian physicist Lajos Diósi, is where quantum mechanics should break down.

“Roger Penrose says that the Universe cannot sustain a superposition of curved spacetime. That this should collapse,” Folman says. Watching that collapse happen, or fail to, he says, would be the most direct test yet of whether gravity is quantum. The team hopes that putting nanodiamonds in superposition will become possible within the next four or five years. “It’s a very, very challenging project,” Folman says.

Science Advances, 2026. DOI: 10.1126/sciadv.aec8045