The Quantum Refrigerator
A tiny cooling device could help rewrite the thermodynamic rulebook for quantum machines.
by Nicola Jones
February 9, 2021
A small fridge could have big consequences for physics.
Over the past decade, physicists have been pondering how the laws of thermodynamics—which describe heat and energy transfer in macroscopic systems—will apply (or need to be modified) as machines shrink down to the atomic level. At these scales, quantum physics, which describes the behavior of sub-atomic particles, becomes important. These quantum effects could be exploited to make machines, including cooling devices, that operate in a new and quantum way.
"People talk about thermal machines, and there has been lots of insight, but no one has actually realized a quantum machine in this sense," says
Jens Eisert, a physicist at the Free University in Berlin, Germany.
Now, with the aid of an
FQXi grant of over US$1.3 million, Eisert, along with fellow physicists
Jörg Schmiedmayer, at the Atomic Institute in Vienna, Austria, and
Marcus Huber, from the Institute for Quantum Optics and Quantum Information, also in Vienna, are attempting to build a quantum refrigerator. The physicists hope their tiny fridge will be the first genuine quantum-thermodynamic machine built with many parts.
Investigating how their fridge functions should help the team to fill in the details of the thermodynamic rule book for quantum systems—a 21st-century update to 19th-century steam-engine (or refrigerator) science. "In quantum thermodynamics we still face big questions," says
Alexia Auffèves, a quantum thermodynamic theorist at the Institute Néel in Grenoble, France, who is not involved with the project. "What is heat, what is work at the quantum scale? That’s still not understood and is still under debate." Experiments like these help to settle those debates, she says.
Real working quantum refrigerators are not just theoretically interesting but also potentially useful. For example, they could one day help to cool a quantum computer down to the near-zero temperatures needed for best operation. "If you want really to unleash the power of quantum computing you need some serious cooling," says Auffèves. Quantum enhancement of refrigeration devices might do just that.
The effort is part of a larger field of study into how quantum weirdness may be harnessed to power machines, or, to put it another way, how information can be utilized as fuel.
In the late 1800s, Scottish physicist James Clerk Maxwell conjured up a thought experiment that kicked off this line of thinking. He imagined a container of gas with a gated barrier in the middle, and some tiny little "neat fingered being" who could observe the speed of molecules approaching the barrier and choose to open the gate whenever, say, fast ones were heading his way. In this way the little demon could separate hot from cold, creating order from disorder. In effect, the system converts the demon’s knowledge about the molecules into energy.
Demonic ExperimentUsing knowledge of the speed of particles, it may be possible to create order from
disorder and power a machine.Credit: Htkym, Wikicommons Many versions of such devices have been proposed by modern physicists, often built from just one or two atoms, or qubits—the entities used to encode information in quantum computers. And
some quantum-scale Maxwell’s demons have been brought to life in recent years.
Michele Campisi, a physicist at the Institute for Nano Science in Pisa, Italy, who is not part of the German-Austrian team, and colleagues proposed their own idea for a two-qubit quantum fridge in 2018. It uses the peculiar fact that measurements can disturb quantum systems to its advantage—as a fuel to power refrigeration (
Buffoni, L. et al. Phys. Rev. Lett. 122, 070603 (2019)). Campisi and colleagues’ quantum-measurement-cooling technique doesn’t need to use or even keep the information gained from measurement; it’s the measurement itself that provides energy. "You can throw the information away," laughs Campisi.
Tiny devices like these, and others that have been realized, may ultimately prove useful for quantum computing cooling—but so far they constitute simple proofs-of-principle rather than true machines that can probe thermodynamic operation, says Eisert: it’s "cheating," he jokes. Eisert’s argument is that tiny experiments made with just a few parts are too small and quantifiable to be truly thermodynamic; the researchers just lay a thermodynamic story or interpretation onto a controlled system.
In real thermodynamic systems there is too much going on to have perfect knowledge about it—thermodynamics instead creates a statistical description of the overall effect of what’s going on. For instance, classical thermodynamics explains that it is the average speed of the atoms in a gas that results in the gas’s temperature. The faster the particles, on average, the hotter the gas. But crucially, you do not need to know the actual speed or properties of every individual particle to know how a steam engine would work. It would take more supercomputers than exist in the world to have total knowledge of all the atoms in any system beyond just a handful.
We hope to witness anomalous heat flow, where heat flows from cold to hot—consuming these quantum correlations as fuel.
- Marcus Huber
To build a fridge like that found in your kitchen, Huber explains, "I don’t need to know about every atom, or how every part works. I only need to know its statistical overall behaviour." On the quantum scale, for devices like Campisi’s proposed device, however, "you have perfect control over the system; I don’t need thermodynamics to describe it," Huber adds.
The team wants find the middle ground between these two extremes. "Somewhere in between there’s a sweet spot where you can actually investigate quantum thermodynamics," says Huber. In other words, Eisert and Schmiedmayer’s new machine should bridge the gap between big classical thermodynamic systems and tiny controlled quantum systems.
Taking a different tack to Campisi’s proposed strategy, they aim to do the measurement in the most non-invasive way possible, extracting the information and using it to power their refrigerator. "They are doing something complementary," says Campisi.
To set up a normal, macroscopic refrigerator like the one in your kitchen, you typically need two pools of fluid. When the fluid in one is forced to contract it will heat up; when the fluid in the other is forced to rapidly expand, it cools. By connecting the pools and operating these processes periodically you can transfer heat between them, cooling one pool relative to the other.
Jens EisertFree University in Berlin, Germany. The team’s tiny quantum fridge effectively uses that same standard physics. "Basically, we’re replacing the fluid from your fridge with a quantum system," says Schmiedmayer.
Their system will consist of a few thousand rubidium atoms, trapped by a system of lasers and cooled into a quantum state known as a Bose Einstein condensate—which behaves like a fluid whose particles move in lockstep, like one single atom. The idea is to use those lasers to control two pools of the fluid, forcing one to rapidly expand and the other to contract, creating a tiny heat pump.
Although the general principle and mechanisms are borrowed from classical physics, the system is small enough that quantum effects should prevail. One key feature that will be exploited is entanglement—a quantum phenomenon that should link the two pools, so that operations on one instantly affect its partner. "We’re in the goldilocks zone, where the system is large enough to be thermodynamic, but we have a realistic hope of retaining these correlations," says Huber.
The question the team aims to investigate is how those quantum effects will manifest: how they’ll impact the fridge. "We even hope to witness anomalous heat flow, where heat flows from cold to hot—consuming these quantum correlations as fuel," Huber says.
So far, the team has a working system for creating these pools and manipulating them with lasers. But all the details of how to connect the pools, measure the heat flow, and get the whole thing working as a coherent, tweakable machine, will take years yet. "We are basically building the nuts and bolts right now: how to build the boiler and connect it to the piston and things like this," says Eisert.
"We hope for it tomorrow. Plan for three years. Realistically it will take five," laughs Schmiedmayer.
Campisi thinks the project is doable, given the team’s technology. However, there’s always a worry that they might end up with a tool that compresses and expands gas without doing any useful work like cooling, he says, adding: "If they do it, if they make it work, it will be very exciting."