The future of electrical computing is inherently hindered by physical limitations to conduction speeds and energy losses. To overcome this obstacle, researchers are working in several different directions.
For the short term, they're trying to develop better conductors like graphene that will show lower losses. For the long term they're working towards developing room temperature superconductors. Also, quantum computing is being sought as a solution to certain kinds of problems.
However, the ultimate future of computing as we know it may be to replace electricity with light. Researchers have been able to trap light, switch light, and more, raising the possibility of processing circuits and memory powered entirely by light. However, in order to develop optical computers, a reliable light source on a microscopic level is essential.
The only light source capable of creating the intensely focused beam needed is a laser. Typically, a laser builds its beam by bouncing light off walls of a cavity. However, a linear cavity approach is too bulky and infeasible for a microscopic scale. In 1992, though, researchers devised using a circular disk as a sort of race track to amplify light into laser beams, which are called "whispering gallery modes" based on their similarity to sound waves bouncing off curved walls.
However, the downside to the mini-disk lasers was that the beams were emitted in random directions, essentially making them useless to optical computing. Martina Hentschel of the Max Planck Institute for the Physics of Complex Systems states, "The experimentalists have a holy grail of unidirectional emission in microlasers."
In recent years, a new type has arisen called spiral microlasers which have a slight notch on the side, making the disk resemble a snail shell. Light is drawn to the notch and tends to emit from it more frequently. However, while the character improved slightly, many tests still showed light to be escaping in varying directions at times.
Now Martina Hentscheland her colleague, Tae-Yoon Kwon, have performed and in depth analysis on spiral lasers, with new theory, which yields substantial improvements and brings the devices closer to computing readiness.
Before a "billiard" model was used to explain spiral lasers, similar to that used with fiber optics. This theory that beams hitting at sharper angles escape, while those that barely graze the surface state inside the circle. However, this theory alone is insufficient to characterize spiral lasers.
The pair of researchers ditched this model in favor of a new electromagnetic wave and laser equations model. This model showed light emitted in two modes clockwise and counterclockwise, and allowed them to understand where to excite the semiconductor to emit light for optimal efficiency.
They found that the notch needs to be twice the wavelength of the laser beam to be emitted. Moreover, the semiconductor excitation to produce light, known as pumping, needs to be confined to the outer 10 percent of the disk.
The result is greatly improved unidirectional emission character, and better efficiencies. Using the model should also allow the researchers to build upon the work and achieve even higher efficiencies. States Ms. Hentscheland, "The optimal geometry and boundary pumping is very useful to know for an experimentalist."
The work is reported in the Optical Society (OSA) journal Optics Letters.