Preserving quantum information is key to developing useful quantum computing systems. But interacting quantum systems are chaotic and follow laws of thermodynamics, eventually leading to information loss. 

Physicists have long known of a strange exception, called dynamical freezing, when quantum systems shaken at precisely tuned frequencies evade these laws. But how long can this phenomenon postpone thermodynamics? 

Not forever, but for an astonishingly long time, Cornell physicists have determined, giving the first quantitative answer. Using a new mathematical framework, they demonstrate that the frozen state can be stabilized long enough to be a useful strategy for preserving information in quantum systems. This can be a promising route for maintaining coherence in quantum computers as the numbers of qubits scale up to the millions.

“It’s like asking, how do you evade the laws of physics from eventually taking over?” said Debanjan Chowdhury, associate professor of physics in the College of Arts and Sciences. “Imagine that you had a hot cup of coffee that even without a heater stayed hot. Or a block of ice placed on a heater that never melts. Is that even possible? This has been one of the big open problems in the field of quantum many-body systems.” 

With their analytical calculations, Chowdhury and two members of his research group discovered that while quantum systems can be driven to maintain information for extremely long periods – potentially approaching the age of the universe – the frozen state is not permanent and will inevitably thermalize through extremely rare quantum processes.

Their paper “Floquet Thermalization via Instantons Near Dynamical Freezing,” published in Physical Review X on Feb. 27. The co-first authors are Haoyu Guo, Bethe/KIC postdoctoral fellow with Cornell’s Laboratory of Atomic and Solid State Physics (LAASP) and Rohit Mukherjee, former Fulbright visiting fellow. 

“There is something special about this effect,” Chowdhury said. “It doesn’t last forever, but now we can calculate precisely how long the protection persists. The time scales are exponentially long in the presence of the drive. There is a very, very long time over which the information can be preserved.”

On why the drive needs to persist, Mukherjee said, “the system does not stay naturally frozen on its own. Think of a playground swing: If you give it small, well-timed pushes over and over, you can keep its motion controlled in a particular way. Here, the periodic drive is like those regular pushes.” 

This work shows theoretically how the coherence ultimately fades. 

“Most of the time the system remains stable, but every so often it makes a sudden quantum jump to a different state,” Guo said. “Imagine a ball sitting quietly in a valley that unexpectedly shows up in the next valley over – not by rolling uphill, but by passing through the mountain itself, something only quantum physics allows.”

Chowdhury said the continued periodic drive at the precisely tuned frequencies causes a subtle quantum mechanical cancellation of processes that lead to chaos.

“What we see here is like a noise-canceling headphones for quantum chaos,” he said. 

This work is theoretical, but there are huge experimental implications connecting directly to ongoing efforts across different quantum computing platforms, Chowdhury said. As quantum processors grow larger, preserving coherence becomes dramatically harder; a single unstable qubit can trigger cascading errors across millions of interacting components.

“With a few qubits, control is manageable,” Chowdhury said. “With millions, even small chaotic processes can avalanche. We need strategies that remain effective as systems scale.”

Dynamical freezing isn’t the only strategy for preserving quantum information. But it is a particularly promising one for scaling up from a handful of qubits interacting to millions in a real device in the future. 

“This work shows that dynamical freezing does not manifest as a violation of thermodynamics, but as a finely balanced state, poised between order and chaos, whose lifetime can now be predicted from first principles,” Chowdhury said.

The research was supported by the Alfred P. Sloan foundation, the National Science Foundation and a New Frontier Grant from the College of Arts and Sciences.

Kate Blackwood is a writer for the College of Arts and Sciences.