Metasurfaces for light shaping: a look into the past to better appreciate the present and future
Abstract
In integrated circuits, an important prerequisite for success is that every transistor can be driven or read independently, and nanotechnology is used to push the integration limit subject to that essential requirement. In dielectric metasurfaces for wavefront shaping, one aims to imprint spatially varying phases (or amplitudes, polarizations) on an incoming optical beam, and nanotechnology is used to spatially control the phase down to the subwavelength scale, as required for high numerical aperture focusing or imaging, beam steering over large angles, and computer-generated holography. Just like for integrated circuits, it is important that the phase can independently be controlled by adjacent nanostructures (or meta-atoms).
This demanding requirement has marked the history of diffractive optics. I will review this history, evidence some fundamental limitations and introduce recent prospects in my group in relation to disordered metasurfaces.