Abstract

Bistability is at the root of numerous applications, present or envisaged. In the case of optics this is especially relevant from the viewpoint of information processing. Bistability can be of intensity type (the classical optical bistability of nonlinear dispersive cavities) but also of phase type (like that displayed by type I degenerate optical oscillators (DOPO), which can equally likely emit light with either of two opposite phase values, differing by π). However bistability is not a usual feature and only few optical systems display it: Lasers do not. A relevant question is the following: Can we do something to force a laser(-like) system to display phase bistability? [By laser-like we refer to optical systems that, like the laser, exhibit phase invariance, non-degenerate optical parametric oscillators or two-wave mixing photorefractive oscillators (TW-PRO) among them.] The positive answer is given by a new type of forcing that receives the name of rocking [1], which, in optics, consists in the optical injection of an almost resonant, amplitude modulated plane wave (in the simplest case, a bichromatic injection) into the laser(like) cavity. When the rocking parameters (carrier frequency, and modulation frequency and amplitude) are properly tuned laser emission frequency locks to the carrier frequency and its phase displays bistability between two opposite values [1]. In the case of very large Fresnel number nonlinear optical cavities this phenomenon manifests additionally from the viewpoint of pattern formation: Free running lasers spontaneously display vortices owed to phase invariance (a vortex is a black filament parallel to the light beam, around which the light phase rotates by 2π; on the transverse plane it is seen as a black spot), which are transformed into domain walls typical of DOPO, in which two regions (domains) of the transverse plane with equal light intensity but opposite phase values are connected by a dark line (the wall). Optical domain walls could have potential applications in information processing, as they can be controlled and steered as our group has shown. The verification of these predictions, which we communicate in this paper, could open new fields of applications for lasers.

© 2007 IEEE