Context: Guiding light trough or around inhomogeneous media is one of the great challenges of optics. The most widely studied strategy to reach this challenging goal is to work with metamaterials – carefully engineered structures with exotic properties. The theoretical framework supporting such approaches is transformation optics, a conceptually very appealing guide for optical design based on geometric transformation of the underlying coordinate space. Such techniques do, however, come with several limitations like the requirement for very specific metamaterial properties such as near zero dielectric constants and negative refractive indices, making them challenging to implement.
Hypothesis: Our project aims to explore a complementary approach to eliminating scattering in inhomogeneous dielectric media, by adding suitably arranged gain and loss distributions to them. We hypothesize that such techniques can lead to design schemes for media through which broadband undistorted propagation is possible, making these non-Hermitian materials unidirectionally invisible to a distant observer. In addition to these theoretical concepts, we will explore possibilities of their experimental implementation with light in time-multiplexed fiber loop structures and atomic vapours with laser-tailored optical susceptibilites.
Methods: Instead of transforming the coordinate space, an approach used in transformation optics, we will study transformations of the electric field solutions themselves, and analyze the properties of the corresponding complex dielectric functions. Many such schemes can be explored within the framework of classical non-Hermitian optics. The experimental systems will be modeled on the level of classical optics in discrete (time-multiplexed) setups and semiclassical light-matter interaction in continuous media like atomic vapors. Our analytical predictions will be tested by numerically solving Maxwell’s equations using finite element methods.
Innovation: Our approach has the potential to deeply affect the standard practices of photonic material design as it enables an original and conceptually simple framework for tailoring optical media. The great advantage of our approach is that it can be applied to dielectrics with gain and/or elements being flexibly added to a given structure only after its fabrication. Another attractive feature of the proposal is the possibility to connect the results in non-Hermitian optics to the mature field of transformation optics. Such a connection has the potential to bridge the community working on non-Hermitian physics and PT-symmetry with its peers dedicated to cloaking and transformation optics - a highly interesting prospect which improving our understanding of complex optical phenomena.
