Nature study reveals unexpected stability of driven quantum systems

A universal quantum effect, first predicted at Imperial College London, has now been confirmed in a major international experiment.

New research has revealed how large quantum systems can remain stable for unexpectedly long periods even while being driven far from equilibrium, a key challenge for developing future quantum technologies.

Published in Nature, the work confirms predictions developed by Imperial researchers. Their theory predicted that certain forms of “structured randomness” can prevent driven quantum systems from heating up too quickly.  Instead of losing all structure, the system enters a long-lived stable regime known as a prethermal state.

The international study, involving researchers at the Chinese Academy of Sciences, Peking University, the Max Planck Institute, TU Munich and Imperial College London, provided the first experimental validation of this effect using a 78-qubit superconducting quantum processor.

Shaking a quantum system

When left undisturbed, a quantum system settles into a steady, predictable state known as equilibrium. However, modern quantum physics is increasingly concerned with a simple question: what happens when a quantum system is pushed far from this equilibrium?

Many emerging quantum technologies, from quantum simulators to quantum computers, operate in this far-from-equilibrium regime, using controlled time-dependent changes to perform useful tasks. Physicists call this “driving” or “shaking” the system.

This can be highly useful, allowing researchers to create new kinds of behaviour that would never appear in equilibrium. However, it also introduces a serious challenge: driven quantum systems tend to absorb energy from the drive and heat up. If heating is not controlled, the system loses its structure, and all interesting quantum behaviour disappears.

The heating problem is one of the main obstacles to building reliable quantum technologies. Until recently, it was believed that introducing randomness to the driving, often unavoidable in realistic devices, would speed up heating and make it impossible to control.  

Seeing the effect in the lab

To test whether this assumption always holds, the research team carried out experiments on Chuang-tzu 2.0, a state-of-the-art superconducting quantum processor made up of 78 qubits arranged on a two-dimensional grid. The device allows researchers to precisely control how qubits interact, making it possible to simulate complex systems with many interacting particles.

On this platform, the researchers applied structured random driving, in which random elements were arranged in a specific pattern predicted to slow down heating. They then tracked the system over more than a thousand driving cycles.

Instead of heating rapidly, the system entered a predicted prethermal state, remaining stable and highly ordered for far longer than would normally be expected before eventually heating up.

As the experiment progressed, quantum correlations spread across the entire processor, making the system’s behaviour extremely complex – a regime that cannot be realistically simulated on classical computers.  

Remarkably, this behaviour had been predicted in earlier theoretical work carried out at Imperial. During his PhD, Dr Hongzheng Zhao, supervised by Professor Johannes Knolle and Professor Florian Mintert, predicted that stability in driven quantum systems should follow a universal pattern rather than relying on careful fine-tuning.

Professor Johannes Knolle, who along with Zhao, Mintert and Roderich Moessner from the Max Planck Institute for the Physics of Complex Systems were part of this theory-experiment collaboration, said, “Having Hongzheng Zhao as my first PhD student at Imperial was a stroke of luck. With his enthusiasm for research, he discovered that randomly driven quantum systems could remain stable for long periods before heating up. This fundamental prediction of ours has now been realized for the first time on a quantum computer." 

Dr Zhao added, “It is amazing to see one’s own theory realised in the lab, and we hope it can be exploited in future quantum technology.”

As Dr Zhao predicted, the results show that carefully designed randomness can be used to control how quantum systems absorb energy. This could help extend the lifetime of future quantum simulators and computers and highlights the unique role of quantum processors in exploring physics beyond the reach of classical methods.