Turning Information Into Energy
Japanese physicists turn Brownian motion—plus plenty of external energy—into work
16 November 2010—Energy scavenging is all the rage these days. Sensors power themselves with harvested radio-frequency radiation, and piezoelectric roadbeds are in development to turn highway rumblings into electricity for traffic signals and streetlights. Now Japanese physicists claim to have pushed this trend to a thermodynamic extreme by harvesting the ubiquitous random vibrations known as Brownian motion. The key to doing this, they explain in this week’s issue of the journal Nature Physics, is to first gather a few bits of information.
Chuo University physicist Shoichi Toyabe and University of Tokyo physicist Masaki Sano, along with three colleagues, predict that their experimental system could produce smart devices that could power themselves using Brownian motion—the low-quality energy left over by the inexorable flow toward increasing entropy that is enshrined in the second law of thermodynamics. Experts in microelectromechanical systems, or MEMS, say it’s a worthy goal but see the Japanese experiment, at best, as a proof of principle.
In their experimental setup, the scientists use a high-speed feedback scheme to manipulate a linked pair of polystyrene beads 287 nanometers in diameter—about the size of a large virus. The beads are sandwiched between glass plates and chemically bound to the top plate at a single point, around which they exhibit Brownian motion by randomly spinning clockwise and counterclockwise with equal probability. The feedback control detects and blocks swings in one direction, thus causing the nanoparticles to consistently spin in the other direction.
The paper’s primary claim is that the research provides the first experimental confirmation of a theorem named for University of Maryland physicist Christopher Jarzynski, which relates how adding information to a thermodynamic system (in this case its knowledge of the particles’ direction) increases its free energy (the Brownian motion converted to work). "Their motivation was to demonstrate the Jarzynski relation, which was very hard to do, and they succeeded," says Argonne National Laboratory microfluidics expert Igor Aranson.
In practice, however, the Japanese team’s experimental setup hardly looks like a blueprint for practical devices, because it takes much more than a few bits of information to set their nanoparticles spinning. To track the nanoparticles, they rely on a microscope attached to a high-speed camera that posts images every 1.1 milliseconds. Those images feed into a computer running the control algorithm, which predicts the particles’ movement and decides whether or not to block the spin. Blocking, meanwhile, requires an electric field set up by electrodes etched into the bottom plate of glass.
"You’re using your 110-volt outlet to power a camera and a computer, and you have to apply electrostatic forces to block the particle. That means you’re applying more than just information," says Amit Lal, who directs Cornell University’s SonicMEMS Laboratory. If the goal is to rotate the nanoparticles, it would be simpler to just drive them with the electric field.
However, Lal believes that it could be possible to build simple feedback systems into micro- or nanomechanical devices. In that case, such systems could become a ready source of onboard energy and more. Lal says that feedback to counteract Brownian motion could squelch the vibration-induced noise that plagues sensors or efficiently guide nanoparticles through nanoscale pores and channels where Brownian motion would otherwise randomize the flow.
Aranson agrees that it is too early to write off the team’s information-to-energy system. While the present system may look like a clumsy way to spin a pair of particles, the same technology could look smarter in a complex microfluidic system where the information feedback simultaneously manipulates the flow of hundreds or thousands of particles through narrow channels. "If you just focus on one particle, it may not be very useful," says Aranson. "But if you do it in parallel for many particles, then it’s a totally different story."
About the Author
Peter Fairley is a contributing editor at IEEE Spectrum and the Spectrum Online blog Energywise. In the October 2010 issue he reported on conservation voltage reduction as a means to save power in smart grids.