Maxwell's demon can use quantum information to generate work

December 18, 2013 by Lisa Zyga, feature
In the quantum information heat engine, the system is attached to the reservoir, and is measured and controlled by the memory consisting of a diatomic molecule with two entangled states. Credit: Park, et al. ©2013 American Physical Society

( —In theory, Maxwell's demon can decrease the entropy of a system by opening and closing a door at appropriate times to separate hot and cold gas molecules. But as physicist Leó Szilárd pointed out in 1929, entropy does not decrease in such a situation because the demon's measurement process requires information, which is a form of entropy. Szilárd's so-called information heat engine, now called the Szilárd engine (SZE), demonstrates how work can be generated by using information.

In the SZE and other heat engines devised since then, the information that is used to generate work has always been classical information. Now for the first time, physicists have theoretically demonstrated that a heat engine can generate work using purely quantum mechanical information.

The researchers, Jung Jun Park and Sang Wook Kim from Pusan National University in Busan, Korea; Kang-Hwan Kim from the Korea Advanced Institute of Science and Technology (KAIST) in Daejeon, Korea; and Takahiro Sagawa from Kyoto University in Kyoto, Japan, have published their paper on their work in a recent issue of Physical Review Letters.

"It is known that classical information can be used to extract work, which is important because this saves the second law of thermodynamics!" Sang Wook Kim told "The mathematical expression of such work is given as the mutual information between the system and the measurement device multiplied by kT. Now for the first time we show that quantum information can also be used to extract work, and its mathematical expression is discord."

As the physicists explain, in many ways the new quantum heat engine is similar to the SZE. Both engines begin in thermodynamic equilibrium, and both have some kind of wall that separates the reservoir of molecules into two spaces. While the wall in the SZE is "perfect," in the quantum heat engine it is semi-permeable.

The demon and the type of molecules used are also different in the quantum heat engine. Here, the demon consists of not one but two memories, which the scientists physically demonstrate using a molecule containing two atoms. Each atom has two internal states that are physically equivalent to each atom's two spin states.

Before measurements are performed, the two atoms are prepared in a maximally entangled quantum state, and are also classically correlated. But as measurements are performed to separate the molecules, the quantum entanglement between the two atoms decreases while the classical correlation does not. It is this , expressed here as quantum discord, that is used to generate work. The quantum discord measures the amount of quantum mechanical correlation compared to the total correlation (both quantum and classical).

Like the SZE, the quantum heat engine does not violate the second law of thermodynamics because entropy, in the form of information, never decreases in these systems. The physicists also note that the quantum heat engine is not cyclic, so the memory does not return to its initial maximally entangled state. In other words, the quantum information is not free, and work must be done to recover the initial state.

The researchers plan to further investigate the potential of generating work from quantum information in the future.

"First we are considering how to realize our engine in experiments," Sang Wook Kim said. "The example that we presented in our paper is just a gedanken experiment, implying that it is extremely hard to directly realize it. Second, we are studying the far from equilibrium. This is related to constructing a quantum fluctuation theorem with feedback control."

Explore further: How losing information can benefit quantum computing

More information: Jung Jun Park, et al. "Heat Engine Driven by Purely Quantum Information." PRL 111, 230402 (2013). DOI: 10.1103/PhysRevLett.111.230402

Related Stories

How losing information can benefit quantum computing

November 24, 2013

Suggesting that quantum computers might benefit from losing some data, physicists at the National Institute of Standards and Technology (NIST) have entangled—linked the quantum properties of—two ions by leaking judiciously ...

Quantum engines must break down

June 26, 2013

Our present understanding of thermodynamics is fundamentally incorrect if applied to small systems and needs to be modified, according to new research from University College London (UCL) and the University of Gdańsk. The ...

Playing quantum tricks with measurements

February 15, 2013

A team of physicists at the University of Innsbruck, Austria, performed an experiment that seems to contradict the foundations of quantum theory—at first glance. The team led by Rainer Blatt reversed a quantum measurement ...

Recommended for you

Walking crystals may lead to new field of crystal robotics

February 23, 2018

Researchers have demonstrated that tiny micrometer-sized crystals—just barely visible to the human eye—can "walk" inchworm-style across the slide of a microscope. Other crystals are capable of different modes of locomotion ...

Researchers turn light upside down

February 23, 2018

Researchers from CIC nanoGUNE (San Sebastian, Spain) and collaborators have reported in Science the development of a so-called hyperbolic metasurface on which light propagates with completely reshaped wafefronts. This scientific ...

Seeing nanoscale details in mammalian cells

February 23, 2018

In 2014, W. E. Moerner, the Harry S. Mosher Professor of Chemistry at Stanford University, won the Nobel Prize in chemistry for co-developing a way of imaging shapes inside cells at very high resolution, called super-resolution ...

Recurrences in an isolated quantum many-body system

February 23, 2018

It is one of the most astonishing results of physics—when a complex system is left alone, it will return to its initial state with almost perfect precision. Gas particles, for example, chaotically swirling around in a container, ...

Hauling antiprotons around in a van

February 22, 2018

A team of researchers working on the antiProton Unstable Matter Annihilation (PUMA) project near CERN's particle laboratory, according to a report in Nature, plans to capture a billion antiprotons, put them in a shipping ...


Adjust slider to filter visible comments by rank

Display comments: newest first

5 / 5 (2) Dec 18, 2013
This seem sto put us on a path where we can be able to compute the maximum possible energy contained in an entangled system. Entangled state batteries anyone? I wonder if this will come out to a potentially low or extremely high energy density. Gut feeling would indicate the latter.
5 / 5 (2) Dec 19, 2013
Who says scientists don't dabble in fiction eh? Maxwell's 'finite being' thought experiement certainly caused ripple...but of course way back then they weren't aware of 'entatngled particles'.
@antialias_physorg Entangled state batteries anyone?' One of several possibilities I think for such a system and if memory serves, laser and photon cooling systems experiments were being constructed last decade.
Okay, so the second law at the end is safe but look what reasearches have found during the 'journey'. Amazing...great stuff
1 / 5 (1) Dec 19, 2013
I cheer whenever a loose stone is found in the foundations of physics. Will this bring down the entire edifice? I certainly hope so.
"Extraordinary claims require extraordinary evidence" is a load of codswallop.
Just one example is sufficient, especially at the venerable foundations that are touted as being unassailable.
The problem is that science advances one death at a time but we do not have time on our side.
The beginning is near.
3 / 5 (2) Dec 19, 2013
Will this bring down the entire edifice?

Probably not - as it isn't a lose stone (thermodynamics isn't violated).
It does shine an intruguing new light on things.

Scientif theories that have withstood the test of time rarely 'fall down'. They get replaced by a fresh look on things (e.g. Newtonian physics isn't outright false. It wasn't 'brought down' by Einstein. It was just shown to be incomplete)

The problem is that science advances one death at a time

That might have been the case 100 years ago. But that time has long since passed. Since the early 1900's when people started to have access to good information dissemination methods (frequent journals and then the internet) and the ability to recreate experiments much more easily the "science by authority" approach has all but vanished.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.