For decades, physicists believed that even the strangest quantum objects had a hard limit on how strongly their behaviour could be linked across time.
No matter how quantum a system was, there was thought to be a ceiling to how much its present could be correlated with its past and future.
However, now, a team of researchers from India and Poland has shown that this limit can be dramatically broken using a qubit, one of the most basic building blocks of quantum technology.
By letting a qubit evolve under a superposition (the property that enables a quantum object to exist in multiple states at the same time) of different time evolutions, they have pushed quantum correlations (the link between quantum objects or systems) beyond what was long considered an unbreakable bound.
This result not only advances our understanding of what quantumness really means over time, but also points to new ways of making quantum computers and sensors work reliably for much longer than before.
The difference between classical and quantum objects
At the heart of this research is a famous idea from 1985, proposed by physicists Anthony Leggett and Anupam Garg. They introduced a mathematical test, now called the Leggett–Garg inequality, that checks whether an object behaves like a classical object or a quantum one when measured at different moments in time.
Classical objects obey this inequality, while quantum objects can violate it because their properties at different times are unusually strongly connected. However, theorists believed that even quantum systems could only violate this inequality up to a strict maximum, known as the temporal Tsirelson’s bound (TTB).
Breaking this bound was considered impossible under ordinary quantum evolution frameworks, but the new study shows that this assumption wasn’t always correct.
The researchers suggest that instead of forcing a quantum system to follow just one time evolution, let it follow two incompatible evolutions at once. In everyday terms, this is like telling a spinning object to rotate clockwise and counterclockwise at the same time.
This is something that makes no sense in the classical world, but is allowed in quantum mechanics through superposition.
Going against the Leggett–Garg inequality
The study authors focused on qubits, which are the quantum version of classical bits and the basic units of quantum computers. In their experiment, they used a molecule that effectively hosted three qubits and studied it using nuclear magnetic resonance (NMR) techniques.
Each qubit had a specific role. The first qubit acted as a controller, prepared in a quantum superposition state. This controller decided how the second qubit (target qubit) would evolve. The third qubit was used to read out information about the target qubit at different moments.
Normally, when a qubit evolves in time under a single, well-defined rule (called a unitary evolution), the amount by which it can violate the Leggett–Garg inequality is limited by the TTB.
However, in this experiment, the controller qubit placed the target qubit in a superposition of two different evolutions at once. As a result, the target qubit was following two different histories simultaneously.
When the researchers measured correlations in the target qubit’s behaviour at different times, they found that the violation of the Leggett–Garg inequality went far beyond the TTB. More importantly, the enhancement was not random.
The researchers showed that the more strongly the time evolutions were put into superposition, the larger the violation became. This means the effect could be carefully tuned and controlled. Even more surprising things happened when noise and environmental disturbances, known as decoherence, were introduced.
Decoherence normally destroys quantum behaviour quickly, limiting how long qubits can store or process information. Here, however, the superposition of time evolutions made the quantum correlations last much longer. In fact, in the experiment, the target qubit retained measurable quantum temporal correlations for about five times longer than usual.
“We experimentally realize superposition of unitaries in NMR systems and demonstrate this enhanced violation. In the presence of noise, such a superposition of unitaries remarkably extends the time of LGI violation, showcasing improved robustness against decoherence,” the study authors note.
The next challenge is scaling
At a fundamental level, this study changes our understanding of how quantum systems behave over time. Violating the TTB means that the correlations between a quantum system’s past and future can be far stronger than previously thought.
This pushes the boundary of where the quantum world ends and classical behaviour begins. From a practical perspective, the work offers a powerful new tool for quantum technologies. Long-lasting and controllable quantum behaviour is essential for quantum computing, where fragile qubits must be manipulated with extreme precision.
The same ideas could also benefit quantum metrology, such as ultra-sensitive measurements of magnetic or electromagnetic fields, where stronger and longer-lived quantum correlations can directly improve performance.
However, the current demonstration was carried out in a carefully controlled laboratory setting using NMR systems, and scaling these ideas to larger, more complex quantum devices will be challenging.
Future work will focus on testing similar superpositions of time evolution in other platforms and understanding how these extreme temporal correlations behave in more general measurement scenarios.
The study is published in the journal Physical Review Letters.