Storage of light: A useful concept?

Abstract

We show both analytically and numerically that photons from a probe pulse are not stored in several recent experiments. Rather, they are absorbed to produce a two-photon excitation. More importantly, when an identical coupling pulse is re-injected into the medium, we show that the regenerated optical field has a pulse width that is very different from the original probe field. It is therefore, not a faithful copy of the original probe pulse. Recent experiments [1,2] on the extremely slow propagation of a pair of optical pulses in a resonant medium and the subsequent coherent regeneration of the “probe ” pulse have generated much excitement in the fields of optical computing and quantum optics. The physical explanation offered to these extraordinary phenomena rests on the adiabatic theory developed a decade ago by Eberly and coworkers[3]. To a large degree, the “dark-state polariton” theory [4] and all other recent works formulated around it are basically a field theoretical reformulation of the adiabatic approximation of pulse pair propagation in resonant media originally studied reasonably complete in the early 90’s. These new treatments have provided some further understanding of the process, especially from the view point of atomic spin wave excitation and atomic coherence. Two main results of the these works, however, do not, in our view, accurately describe the physical picture of such a highly nonlinear excitation process. In this letter, we show, based on the adiabatic theory, that the claim of the “revival of the probe pulse ” in recent experiments [1,2,5] is not accurate.