The same phenomenon can be found in other types of wave propagation, including light propagating in e.g. a 3D photonic crystal, although it has proved experimentally challenging to prove optical Anderson localization in 3D. Achieving strong localization in 1D or 2D is significantly more achievable in practice. A 2D system could consist of an optical waveguide with disorder in the transverse refractive index distribution, and being constant along the propagation direction. An example of this was demonstrated recently by making an optical fiber out of two different polymers; the polymers had a refractive index difference of 0.1 and were distributed randomly in the transverse plane of the fiber.
It would be interesting to use Anderson localization in an optical fiber for applications such as spatially multiplexed optical communications or imaging, but this requires designs with small localized beam radius and low sample-to-sample variation in the beam size. It turns out that this can be achieved by raising the refractive index contrast in the random transverse profile, e.g. by using silica (n=1.45) and air (n=1), instead of two different polymers. This has now been tried by Karbasi and colleagues by drawing a rod of highly porous glass into an optical fiber with randomly distributed air-holes. The average air-fill fraction achieved (5.5%) is far from the theoretically optimal 50%, and indeed the researchers found no localization to occur when they launched a beam into the center of the fiber. But, when they launched light into the near boundary region of the fiber, they saw that the beam coming out of the fiber was localized. At first glance this directly contradicts other studies finding that boundaries of a disordered medium are de-localizing. In this case, however, the Authors point out that the degree of disorder in the transverse plane of their fiber is not uniform: the air-fill fraction is lower in the center of the fiber and higher at the boundaries. They therefore believe, and support this by theoretical calculations, that the increased disorder at the boundaries is just enough for localization to occur.
There is currently much interest going into increasing optical communication capacity by utilizing multi-core fibers for spatial multiplexing. The work described here may in the future lead to fibers without a set of a few pre-defined index-guiding cores, but that instead have a distribution of many “cores” in a random refractive index profile. Issues such as limiting cross-talk between these cores will of course have to be solved, which requires high air-fill fraction and high uniformity at the same time. The Authors already have a more recent paper in Opt. Express (Vol. 21, p. 305) in which they investigate the simultaneous propagation of multiple beams in a similar type of optical fiber, and examine how much the beams get displaced when bending the fiber. It will no doubt be interesting to see how this utilization of deliberate disorder continues in the future.
You must log in to add comments.