Looking through the ground glass

Abstract

We initiate, and to an extent, motivate our discussion of wave propagation through a random medium asking whether we can view or image an object through a light scattering medium. We answer in the affirmative by arguing that the image-bearing ballistic component of light can be time-resolved with respect! To the image-blurring diffusive component that has to traverse relatively much longer distance. This leads jus to the question of diffusion of light or its absence (localization) in disordered media. We discuss some essential differences between photon localization vis-a-vis electron localization. One of these that makes photon localization much harder to realize experimentally is that the photon energy multiplies the dielectric disorder in the Maxwell equation, as a result of which localization is missed in the limit of bath the long wavelength (Rayleigh scattering) and the short wavelength (geometrical optics). The narrow 'window of localization' requires drastic enhancement of effective scattering which is possible by the coincidence of the Mie-resonant scattering and the Bragg-reflection (umklapp) conditions. Photon localization at microwave frequencies (as also the complete photon band gap) has already been achieved by several workers. Localization at visible wavelengths is awaited. We also discuss some fundamental QED effects of photon localization, such as the suppression of spontaneous emission from an excited atom embedded in a random dielectric. We end the discussion with some speculations on photon localization in an active (amplifying) medium and the possibility of a 'mobility-edge' laser.