Trapping polarization of light in nonlinear optical fibers: An ideal Raman polarizer

The main subject of this contribution is the all-optical control over the state of polarization (SOP) of light, understood as the control over the SOP of a signal beam by the SOP of a pump beam. We will show how the possibility of such control arises naturally from a vectorial study of pump-probe Raman interactions in optical fibers. Most studies on the Raman effect in optical fibers assume a scalar model, which is only valid for high-PMD fibers (here, PMD stands for the polarization-mode dispersion). Modern technology enables manufacturing of low-PMD fibers, the description of which requires a full vectorial model. Within this model we gain full control over the SOP of the signal beam. In particular we show how the signal SOP is pulled towards and trapped by the pump SOP. The isotropic symmetry of the fiber is broken by the presence of the polarized pump. This trapping effect is used in experiments for the design of new nonlinear optical devices named Raman polarizers. Along with the property of improved signal amplification, these devices transform an arbitrary input SOP of the signal beam into one and the same SOP towards the output end. This output SOP is fully controlled by the SOP of the pump beam. We overview the sate-of-the-art of the subject and introduce the notion of an "ideal Raman polarizer".