What if time is not a fundamental feature of the universe, but something that emerges from the relationships between quantum systems? Giovanni Barontini, a professor of physics at the University of Birmingham, has built a tabletop experiment that suggests exactly that.
Using approximately 24,000 ultracold rubidium-87 atoms cooled to billionths of a degree above absolute zero, Barontini created a Bose-Einstein condensate (BEC) and split it in two with a thin sheet of laser light. The two halves, one observed (the “bright sector”), one deliberately left unobserved (the “dark sector”), were hermetically isolated from their surroundings with no external clock available. This was, in effect, a miniature analog of the Wheeler-DeWitt equation that troubles quantum gravity theorists: it had no built-in time parameter.
“Time was speeding up or slowing down, or even stopping, depending on what the system was doing,” Barontini told Live Science.
Entropic time
The key insight was to define time not from an external clock but from the exchange of entropy between the two halves of the system. As atoms crossed the laser barrier between sectors, entropy flowed from the bright sector to the dark sector, or vice versa; Barontini measured this flow every 2 milliseconds for 120 milliseconds and used it to construct what he calls “entropic time” (τ, tau).
When entropy was flowing briskly between the halves, entropic time ran fast. When the exchange slowed, the clock slowed. When the two halves reached equilibrium, with no further entropy to be exchanged, the internal clock stopped entirely.
In a regime with low barrier height, the bright sector cycled through repeated expansions and collapses, mimicking a universe going through alternating Big Bangs and Big Crunches. Crucially, entropic time did not elapse between a crunch and the subsequent bang, because no entropy was exchanged during that transitionless period.
“Both time and the arrow of time, maybe they just are born from ignorance,” Barontini said. “To have time and to observe, you have to give up some degrees of freedom.”
From analog to theory
The experiment, published as a Letter in Physical Review Research (Vol. 8, L022047, June 2026), is explicitly described as an analog Wheeler-DeWitt mini-universe. The Wheeler-DeWitt equation, central to canonical quantum gravity, suggests that at the most fundamental level the universe’s wavefunction does not evolve in time; time itself is absent. Barontini’s system provides a physical test bed for this idea, even if it cannot directly probe quantum gravity.
Using the entropic time variable, Barontini derived a time-dependent Schrödinger equation (iℏ ∂_τ ψ = …) and showed that it accurately reproduced the experimental observations. “This was quite surprising, how well everything came together. Very neatly, in a way. Which is something that doesn’t happen that often in experiments,” he said.
The paper notes several important caveats. The Hamiltonian is structurally analogous to minisuperspace models, but this is a cold-atom simulation, not an actual quantum gravity experiment. The system avoids singularities because the center of the potential barrier prevents the bright sector from ever experiencing the infinite density of a genuine Big Bang or Big Crunch. Some “wiggles” in the entropic time data were caused by coarse sampling in the clock field, meaning the ordering is not perfectly monotonic at all points.
What it means for the nature of time
Barontini’s experiment is a proof of concept, the first demonstration that controlled quantum systems can serve as test beds for fundamental questions about time. The idea that time emerges from quantum correlations, rather than existing as a fundamental background parameter, has been explored theoretically for decades. But this is the first direct experimental test of that principle.
The deep puzzle is this: the fundamental equations of quantum mechanics treat time as a parameter, but the fundamental equations of general relativity (the Wheeler-DeWitt formulation) suggest there is no time at all. Something must bridge this gap. Barontini’s entropic time, defined by information loss between coupled quantum systems, offers a concrete, testable candidate.
“It’s something that you would normally think is impossible to test,” Barontini said. “These are things we can do very simply, using the tools we already have to engineer our systems.”
Sources
1. Live Science, “Time was speeding up, slowing down, or even stopping, physicist demonstrates a key theory of time by building a mini-universe in his lab” (7 July 2026). https://www.livescience.com/physics-mathematics/time-was-speeding-up-slowing-down-or-even-stopping-physicist-demonstrates-a-key-theory-of-time-by-building-a-mini-universe-in-his-lab
2. Barontini, G., “Testing the problem of time with cold atoms”, Physical Review Research 8, L022047 (2026). DOI: 10.1103/1h9j-df4k
3. arXiv preprint: arXiv:2509.07745 (v3, 11 June 2026)