Center for Functional Nanomaterials Colloquium

It is traditionally assumed that different photonic devices are needed to measure different properties of light: power meters for intensity, spectrometers for wavelength, and polarimeters for polarization. In this talk, I will present an intelligent photonic sensing approach that challenges this paradigm by combining tunable optoelectronic devices with deep neural networks. I will first discuss our work on twisted double bilayer graphene (TDBG), where moiré-induced symmetry breaking and quantum geometric effects produce a tunable mid-infrared bulk photovoltaic response. Exploiting the strong wavelength and polarization dependence of the photoresponse, together with neural-network-based signal processing, we demonstrate simultaneous full-Stokes polarimetry and wavelength detection using a single subwavelength device. I will then present our recent efforts on long-wavelength infrared sensing using InGaAs/InP heterostructure phototransistor. By leveraging a gain-feedback mechanism within a single transistor architecture, we achieve dramatically enhanced temperature sensitivity for compact thermal detection. Finally, I will discuss how computationally assisted sensing can extend from quantum materials to classical semiconductor systems, enabling compact, integrated, and intelligent optical sensing technologies.