Nano-scale Fiber Optics Connect Computer Chips - dummies

Nano-scale Fiber Optics Connect Computer Chips

By Earl Boysen, Nancy C. Muir, Desiree Dudley, Christine Peterson

Nanotechnology researchers are planning to use nanoscale components to adapt fiber optics for carrying data within computers. The idea is to use light to carry data between microprocessor cores within a computer chip and between separate chips within a computer, just as fiber optic cable carries data as light between major telecommunication hubs today.

Microprocessors have one or more cores. Multiple core processors allow several mathematical or logic calculations to run at the same time. Within the microprocessor cores are connections between components, such as transistors.

Nano could replace the current technology, which sends data through metal lines, with metallic carbon nanotubes, which conduct electricity better than metal. When information is sent from one core to another, the outgoing electrical signal would be converted to light and travel through a waveguide to another core, where a detector would change the data back to electrical signals.

This method might also lower power consumption. That savings occurs because all metal wires have a resistance to the movement of electrons through them, so some of the voltage used to drive the electrons is converted to heat.

Researchers have developed techniques for transmitting light that customize the nanostructure of crystalline material to form waveguides. These waveguides allow light of a particular wavelength to travel through the material with almost no loss of energy.

Researchers are developing nanoscale light sources, electrically driven optical switches (also called modulators), waveguides, optical routers, and detectors to convert electrical data into optical data, route it to a microprocessor core, and convert the optical data back to electrical data so that the microprocessor core can then process it.

One nanoparticle-based method of generating light, developed at Cornell University, is called a spaser (surface plasmon amplification by stimulated emission of radiation). A spaser is similar to a laser. The difference between a spaser and a laser is that a laser has a cavity in which light bounces back and forth to amplify the light intensity in a process similar to resonance.

That method won’t work very well with a nano-sized light source, whose size is a fraction of the wavelength of the light you’re trying to generate. A spaser is much smaller than the wavelength of light; in fact, the spaser made at Cornell is made of a 44-nm diameter particle, and generates light with a wavelength of 531 nm.

A nanoparticle that works like a laser.
A nanoparticle that works like a laser.

When you excite dye molecules in the outer shell of the spaser, the dye molecules add electrons to the gold core. These electrons, along with electrons in the conduction band of the gold core, form an electron cloud (called a plasmon).

This cloud oscillates on the surface of the gold core at the same frequency as the wavelength of light that you want to generate. These oscillating electrons generate an electric field that is strengthened by the resonance oscillation and additional electrons supplied by the dye molecules until the spaser generates a pulse of light.

In another approach, IBM is developing a carbon nanotube–based laser as a light source. Carbon nanotubes generate light; the wavelength of that light depends on the diameter of the nanotube. Either an electrical signal or a light signal can be used to get a nanotube to initiate light generation.

The nanotubes are located between two mirrored surfaces; the distance between them is half the wavelength of the light being used. These mirrored surfaces act as the resonance cavity of the laser, which amplifies the light generated by the nanotubes.