Take two flashlights at night and cross them. You know that the light beams pass right through each other, right? As we all know, this is because Maxwell's equations (the equations of light) are linear. Light cannot interact with light -- usually.
But it can under the right circumstances.
Now take a single laser and split it into two coherent beams of photons. If these laser beams cross inside a type of a nonlindar crystal, they mutually influence each others propagation. In some cases, one beam will donate its photons to the other beam. This results in nonreciprocal energy transfer between the beams, and constitutes one of the few instances of optical amplification without stimulated emission.
The control of light by light is the optical analog of the control of electgrons by electrons in transistors. One of the great challenges facing quantum optics today is the search for the appropriate medium in which photons can control photons. The discovery of such a medium would open to way for photonic computers.
The electronic revolution of the 1960's is now giving way to the photonics revolution. Almost all high-data-rate systems in use today use photons to carry information. Even microprocessors in your personal computers are on the verge of using photonics to bus information and to provide masssively parallel interconnections. Hybrid electronic/photonic microprocessors are only about 5 years away from the marketplace.
Still, there are (almost) no good optical analogs of the transistor. All the current photonics technology uses electronics to generate and detect photons, without letting the photons control themselves. But with the appropriate nonlinear optical material, this may just be a step away.
An electronic computer can be divided into two roles, that of memory, which needs to hold information for as long as possible, and that of processing, which needs to use information as fast as possible. Static holograms perform the role of memory in the optical analogy to the electronic computer, while dynamic holograms perform the role of the central processor.
Holographic optical memories are a technology that have been pursued for several decades. On the other hand, dynamic holographic processors -- that have speeds compatible with image processing applications -- have only come about in the past ten years through the discovery of photorefractive semiconductor materials. Semiconductors, like GaAs, can act as dynamic holographic media. They have high carrier mobilities, that make the refresh rate of the holograms fast enough for video applications.
Photorefractive quantum wells represent the culmination of developments in photorefractive semiconductors. They have all the advantages of high mobility and speed for dynamic hologram recording, while relying on quantum-confined enhancements of optical properties to produce the highest-sensitivity dynamic holographic films currently known. While photorefractive quantum wells cannot yet be used as optical analogs to the transistor, they provide new avenues for the control of light by light in wide areas of new applications.
The photorefractive effect is an effective optical nonlinearity in which the coherent interference of the two beams produces a pattern of bright and dark fringes. These fringes cause electrical charge to separate inside the photorefractive crystal, producing space-charge field patterns that mimic the light patterns. The electric fields, in turn, modify the refractive index of the material, creating a diffraction grating. This light-induced diffraction grating diffracts light from the two laser beams, redirecting the photons in the direction of the other beam. When the phase relationship is just right between the transmitted and diffracted beams, then net constructive interference will occur in one transmitted beam, but destructive interference will occur in the opposite beam. The optical amplification is therefore a simple consequence of diffraction and interference -- two aspects of classical optics -- but in a unique combination.
The photorefractive effect is a type of dynamic holography. Holograms that move and change in time in response to changing light images are called dynamic holograms. They are recorded in real-time just as an ordinary hologram is, using two laser beams. One laser beam carries the information from the object, while the other laser beam acts as a reference. The use of two light beams rather than one (in rodinary photography) that makes it possible for a hologram to record phase as well as intensity.
Dynamic holograms are constantly changing, or updating, as the information on the signal beam changes. This means that dynamic holographic films perform an information processing function.