Metasurface Photodetectors for Computational Imaging

点击翻译按钮获取中文摘要

Traditional image sensors used in photography and microscopy can only visualize the intensity distribution of the incident light, whereas all information related to the local direction of light propagation and phase profile is lost in the image acquisition process. To address this limitation, the proposed research will develop a new class of image sensors based on nanophotonics technology, where only light incident along a predetermined, geometrically tunable set of directions can be detected. When combined with computational imaging techniques, this distinctive behavior can be exploited to enable a wide range of advanced imaging functionalities. Specific examples include the ability to produce focused images of three-dimensional scenes in a single shot, to image otherwise invisible transparent phase objects, and to automatically detect the edges of an object. These capabilities are ideally suited to a multitude of challenging and technologically significant applications – for example label-free imaging of transparent biological cells, surveillance and navigation under degraded conditions (e.g., rain, fog, and underwater), and image recognition for computer vision. The proposed activities will also promote multidisciplinary education through the training of graduate and undergraduate students in relevant areas at the intersection of optical physics, device engineering, and data science, and through related curriculum development efforts. The proposed devices consist of standard photodetectors coated with specially designed plasmonic metasurfaces (i.e., ordered two-dimensional arrays of different metallic nanoparticles) that can selectively transmit or reflect incident light depending on its direction of propagation. The work plan includes the electromagnetic design, nanofabrication, and testing of prototype devices, and the demonstration of their imaging capabilities through computational studies based on the measured device characteristics and through proof-of-principle imaging experiments. Metasurfaces of increasing complexity will be developed with angular response tailored to meet specific applications and capable of providing polarization-engineered and broadband achromatic operation. Advanced computational imaging techniques will be employed to demonstrate the proposed imaging functionalities, from lightfield acquisition to phase contrast imaging and optical-domain spatial filtering. This research will advance the science and technology of metasurfaces in a novel fashion, by exploring their direct integration within active optoelectronic devices. Furthermore, it will highlight and exploit the great potential of combining the extreme design flexibility and size miniaturization of metasurface flat optics with the enhanced information processing capabilities of computational imaging. The resulting devices are promising for a significant technological and societal impact through their many applications in areas of high relevance and timeliness, ranging from the life sciences to computer vision and autonomous navigation.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.