Holy sheet. The asymmetric pores in a silicon wafer induce tiny spheres to separate.

Shaken and Stirred

"Shake it to the left. Shake it to the right. Separate those spheres, 'cause theory works just right." OK, maybe it won't be a hit on the pop charts, but physicists are singing the praises of a new scheme that, as theory predicted, can separate objects of different sizes with nothing more than a few good vibrations. The method should hop from the lab to the real world, where it might help analysts with tasks such as sensing bioweapons.

It's not easy to organize things when they're mixed up. It takes energy to sort a bunch of spheres or other objects into groups depending on their size, for instance. A set of sieves can do the trick, but sieves can't operate indefinitely--the spheres have to be unloaded periodically or the whole operation will clog up.

Now physicists Sven Matthias and Frank Müller of the Max Planck Institute of Microstructure Physics in Halle, Germany, have demonstrated a clever scheme for continuously sorting objects by size. They submerged a silicon wafer with thousands of asymmetric holes in a bath of ultrapure water. When they placed an assortment of tiny spheres in the bath and vibrated the whole assembly so that water sloshed back and forth through the wafer, they saw the spheres spontaneously separate by size, they report in the 3 July issue of Nature. The combination of frictional forces, random motion of the spheres, and the vibration-induced fluid flow along the edges and in the middle of the pores shunt bigger spheres to one side of the wafer and smaller spheres to the other--precisely as predicted by a group of theoreticians in 2000. What's more, by changing the frequency of the vibrations, the researchers can change which side of the wafer the spheres end up on.

Unlike other sorting schemes based upon the same physics, the team's wafer can lead to a "massive" sorter, says physicist R. Dean Astumian of the University of Maine, Orono. "It can run on a whole bunch of particles at once, and it's so simple." Although he says that the research is in its early stages, Astumian thinks that the idea will see applications soon. Müller agrees. "It should be possible to use it for sensing in biological systems--if you need to trap small particles, for instance," he says. The idea looks ready to roll--and shake and rattle, for that matter.