Casimir Effect on a Chip!

One of the strangest effects to arise from the quantum nature of the universe is the Casimir force. This pushes two parallel conducting plates together when they are just a few dozen nanometres apart.

At these kinds of scales, the Casimir force can dominate and engineers are well aware of its unwanted effects. One reason why microelectromechanical machines have never reached their original promise is the stiction that Casimir forces can generate.

On the other hand, many engineers hope to exploit the Casimir force. Various theoretical models predict that the force should be repulsive between objects of certain shapes, a phenomenon that could prevent stiction.

But there is a problem: Casimir force experiments are extremely hard to do. One headache is that nobody has perfected the technology to position different objects accurately with a nanometre scale gap. Another is that microscopic objects tend to warp and bend; any corrugations on a flat surface can dramatically change the amount of Casimir force between them and even its direction. That makes experimental results hard to interpret.

Today, Jie Zou at the University of Florida and a few buddies take a big step towards changing this. These guys have carved a single device out of silicon that is capable of measuring the Casimir force between a pair of parallel silicon beams, the first on-chip device capable of doing this.

The device consists of one fixed beam and another moveable one attached to an electromechanical actuator. The team starts by measuring the separation between them using a scanning electron microscope. They then apply a voltage to the actuator, which pushes the movable beam towards the fixed beam.

The beams oscillate at a natural frequency, which Zou and co can easily measure. However, this frequency depends on the forces on the beams. So as the beams move closer together and the Casimir forces changes, so too does the oscillation frequency. This is how Zou and co measure the force.

Of course, there are other forces at play here too, such as residual electrostatic forces. When Zou and co take these into account, their results more or less exactly match theoretical predictions for the Casimir force that beams of this shape should generate.