In the classical Universe, the arrow of time flies one way – but in the quantum realm, where classical physics breaks down, could even time’s direction be ripe for the tweaking?
The ability to manipulate the direction of quantum evolution could offer new ways to process information – with implications especially for quantum computing.
And it may be within reach.
Korean physicists have demonstrated a qubit whose processes can be run backward in time… and the deliciously ironic twist is that they had to go back in time to find it.
In a new preprint, physicist Tai Hyun Yoon of Korea University developed a theory describing what he calls a quantum vacuum-seeded single-photon time qubit – and showed that an experiment he and colleagues conducted in 2018 for an entirely different purpose had inadvertently already shown it in action.
The notion of a time qubit is a relatively recent one, arising in a preprint penned by physicist Yakov Bloch of Tel Aviv University, uploaded to arXiv in November 2025.
It’s not actually a qubit that can travel backward in time; rather, as Bloch explains it, it’s a quantum system that can evolve in two opposite temporal directions at once.
Imagine a clock whose hands can run both clockwise and counterclockwise simultaneously. That’s roughly the idea – except that instead of turning back time itself, the qubit allows physicists to manipulate the direction in which a quantum process unfolds.
He proposed that this could be experimentally tested by splitting the path of a quantum particle in two, with opposing magnetic fields along the two paths, then bringing them back together. The interference pattern of their reunion could reveal how the two opposing evolutions are connected.
As of time of writing, that idea remains solidly in the theoretical realm; but, like a quantum particle split in two, Yoon took the idea and ran with it in another direction.
Instead of using magnetic fields to reverse the evolution of a spinning particle, Yoon turned to light. His approach uses two crystal sources that generate entangled photon pairs; one photon from each pair carries the single-photon qubit. The technique is called entangled nonlinear biphoton source (ENBS) interferometry.
The direction of a particular quantum interaction is then controlled by adjusting the phase of the light used to generate the photons.
Tweaking the phase by half a cycle reverses the quantum interaction – not because time itself is flowing backward, but because the mathematical process governing the interaction runs in the opposite direction.
In quantum mechanics, the evolution of a system over time is governed by a mathematical description called a Hamiltonian. It determines how the system’s quantum state changes from one moment to the next.
Reverse the sign of that Hamiltonian, and the system evolves as though time itself were running backward, retracing the changes it would otherwise undergo.
Yoon’s method doesn’t reverse the entire Hamiltonian, but it does reverse the part governing the interaction he controls. This gives him two opposing directions of quantum evolution – the equivalent of a quantum clock whose hands can turn either way.
But two opposing directions are not in themselves a time qubit. That requires something extra – superposition, where both directions exist simultaneously.
This is where the shared laser light becomes important. Ordinarily, generating a photon pair leaves behind information that could reveal which crystal produced it. That information distinguishes the two possible paths, preventing them from producing the coherent superposition needed for a qubit.
But by supplying both crystals with light from the same coherent laser source, the distinguishing information can be suppressed, allowing the two opposing quantum evolutions to interfere.
The result is a single-photon qubit that encodes not just two different paths, but two opposing directions of quantum evolution.
Yoon didn’t need to design and build an entirely new experiment to show this approach is feasible. Instead, he turned to measurements from experiments he and colleagues had already conducted, including one in 2018.
Those experiments had demonstrated the phase-dependent behavior and interference that his new theoretical framework requires. By reexamining the data he and his colleagues had already collected, Yoon showed that the apparatus could be understood as an implementation of a single-photon time qubit.
Because it uses data from older experiments to probe a new theory, it’s not a perfect demonstration. However, as a proof of concept, it suggests there could be a lot to gain by pursuing it in a dedicated investigation of time qubits.
A more specialized experiment with tighter control over the quantum states could test the full range of operations predicted by the theory.
Or, as Yoon puts it, “Coupling the discrete ENBS time qubit via controlled time-frequency displacements to hardware-native Gottesman-Kitaev-Preskill (TF-GKP) bosonic grid states in chronocyclic phase space establishes a robust pathway toward fault-tolerant temporal quantum error correction, high-dimensional qudit processing, and tabletop simulations of relativistic quantum detector dynamics.”
Quite.
The research, which has yet to be peer-reviewed, is available on arXiv.
This article was fact-checked by Fiona MacDonald and edited by Fiona MacDonald. While we pride ourselves on our process, we are only human. If you spot a mistake, please let us know.
