Tag Archives: demonology

Thermodynamic demonology: extracting energy from information

Here’s a potential plot device for one of Charlie Stross’ Laundry novels or, rather more seriously (though perhaps not as entertainingly), as the hinge for a Greg Egan story; a team of Tokyo researchers reckon they’ve managed to summon up a very tiny version of what physicists call “Maxwell’s demon”. Ars Technica breaks it down:

Maxwell’s demon has haunted thermodynamics for well over a century, since James Clerk Maxwell first suggested that a small demon might be able to selectively allow only hot atoms through a small gate, gradually extracting heat from a gas without expending much in the way of energy. But there’s no such thing as a free lunch, and the demon feeds on information: it needs to know which atoms are hot. Eventually, it was recognized that information was being exchanged for energy, and an equivalence between the two was calculated based on theoretical considerations. Until now, however, nobody has managed to build a demon that could help see how well real-world behavior matched the theory.


The metaphor the authors use is a spiral staircase. A particle placed on a small staircase will be buffeted by energy, and typically go up or down a stair; on average, it’ll go down more often than up, eventually settling at the bottom of the staircase. The demon stands at the side of the stairs with a barrier. When, by chance, the particle happens to move up to a higher energy state, it inserts the barrier behind it, preventing it from dropping down. Given time, the particle will reach the top of the staircase.

Their real-world implementation involves a bead on a tether that is able to freely rotate around a full 360° axis. Below the bead, the authors set up four electrodes that generated electric fields that were shaped like a sine wave. When the bead was in the trough of the wave, it would be at its lowest energy state. If the bead was jostled anywhere else, it gained potential energy that would eventually be lost again when it fell back down to the trough.

Remember, though, that not even demonologists get a free lunch:

If you ignore the apparatus involved, the authors could directly compare the energy gained against the amount of information required to flip the switches involved. And the results appear to agree very well with the theoretical predications.

As far as thermodynamics is concerned, however, you have to consider the apparatus, since it’s necessary to balance the books in order to avoid thinking that we’re getting energy for free. And, as it turns out, the system for tracking the bead and switching currents is rather elaborate, involving “microscope by constructing a real-time feedback system including video capture, image analysis, potential modulation and data storage.” As an accompanying perspective notes, you’d probably also have to throw in the energetic cost of the grad student who was operating the whole thing, too.

So, no chance of heating my garret with information, then… unless I burn all my books. (Not. Gonna. Happen.)