On-Chip UV/VIS Optical Spectrometer

Optical spectroscopy is one of the most widely used characterization techniques in science and engineering. Miniaturized spectrometer can enable a portable and handheld system for Internet of Things (IoT) and lab-on-a-chip applications. A spectrometer is a highly complex system consisting of optical, mechanical, and image processing units. Miniaturizing such a system is a nontrivial task involving the ability to integrate multiple material platforms and careful planning of performance tradeoffs. In this dissertation, a low-profile reconstructive spectrometer with the thickness of the semiconductors and an operating range spanning the visible wavelength spectrum is reported. In our design, the spectral encoders are based on wide-bandgap gallium nitride (GaN) semiconductors. As GaN semiconductors are an excellent light emitting material, monolithic integration of a reference or excitation light source on the spectrometer chip is possible and thus demonstrated. The GaN-based spectral encoders are programmed using local strain engineering and have a weak dependence on the incident angle of light, eliminating the need to condition the optical signal’s angle distribution with collimation optics or an aperture. Computationally efficient machine learning methods are applied to further reduce the number of encoders needed and the chip area of the spectrometer without greatly sacrificing the spectral reconstruction performance.

Chair: Professor Pei-Cheng Ku