Dynamic Metamaterials for Far-Infrared Imaging and Spectroscopy

As early as 1949, it was predicted that a technological gap would form in the far infrared. This so-called ``terahertz gap" is the result of two limitations. On one side, the atomic phenomena giving rise to laser technologies are difficult to extend below $10$ terahertz (THz), and on the other, microwave technologies are difficult to extend above $0.1$ THz. Even today, while this gap has closed to some extent, the generation and detection of electromagnetic radiation in this bandwidth remains inefficient and impractical, especially when compared to more mature technologies based in optical and microwave frequencies. The terahertz gap thus provides an exciting opportunity for innovation and the development of novel imaging techniques.

Metamaterials are a natural fit for the above problem because their electromagnetic properties are determined by their geometry, so they are fundamentally less limited by the physical properties of the materials of which they are composed. This means that a designed electromagnetic response can be scaled to many different bands--including the terahertz--simply by scaling the geometry accordingly. However, the process of designing and optimizing metamaterials is nontrivial and still very much an area of active research. Chapter 6 in particular will describe some new approaches for metamaterial design based on machine learning methods.