Optical Metasurfaces Saving Energy in Optical Computing and Photovoltaics

I will present our recent collaborative work on the design, fabrication and operation of silicon-based optical metasurfaces that perform mathematical operations in an analog way, using light fields as input and output signals. We show how the interplay between scattering components in silicon metasurfaces creates a mathematical derivative on an input image. We then design and fabricate a silicon metasurface that solves an integral equation by using visible light in which we use grating orders as input and output ports on a periodic metasurface with a specially tailored unit cell. The new analog optical computing concepts operate with very low energy consumption, at the speed of light, and can form the basis of more complex geometries solving multiple equations and can be applied in optical neural networks, control systems, and more.<br/>We then apply optical metasurfaces to control and enhance the interaction of free electrons and highly-localized optical near fields in scanning electron microscopy (SEM). Using angle-resolved cathodoluminesence spectroscopy we analyse the spatial modulation of Smith-Purcell radiation that results from the coherent interplay of free-electron driven optical excitations in a chirped metagrating. We embed cylindrically-shaped metagrating patterns onto the input facet of a multi-mode optical fiber to experimentally demonstrate the coupling of free electrons and guided fiber-optic modes via the Smith-Purcell effect.<br/>In the last part of the talk I will introduce an integrated near field/far-field multiple scattering formalism to control the absorption of light in solar cells. We design and fabricate a metallodielectric metasurface back contact for an ultra-high efficiency InGaAs/InGaAsP/Si multi-junction solar cell and enhance the light trapping inside the silicon bottom cell by multiple scattering, creating a record photovoltaic energy conversion efficiency for silicon-based tandem solar cells of 36.1%.