On-Chip UV/VIS Optical Spectrometer

Abstract

Optical spectroscopy is one of the most widely used analytical techniques in science and engineering. Miniaturizing an optical spectrometer can allow for a portable and handheld system and lead to new opportunities for the Internet of Things (IoT) and lab-on-a-chip applications. Visible-light spectrometers have broad applications in UV-VIS, fluorescence, and chemi-/electro-luminescence spectroscopy. 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, such as spectral resolution, sensitivity, system size, and cost. Among various approaches, spectrometers based on reconstructive algorithms shift the complexity of processing spectral information from physical components to software computations. With the steady growth of computational power per watt-dollar, this approach has become increasingly promising in constructing an extremely compact spectroscopic system. In this work, a low-profile reconstructive spectrometer with only the thickness of the semiconductors and an operating range spanning the visible wavelength spectrum is reported. This spectrometer concept is based on wavelength-selective semiconductor photodiodes monolithically integrated on a chip area of 0.16 mm2. The absorption properties of individual photodiodes were tuned via local strain engineering in compressively strained Indium Gallium Nitride/Gallium Nitride (InGaN/GaN) multiple quantum well heterostructures. By varying the diameters of individual nanopillars, the cutoff wavelengths of absorption were varied across the chip. The intrinsic wavelength selectivity is insensitive to the incident angle of light. The built-in GaN pn junction enabled a direct photocurrent measurement. In this dissertation, we first proposed and demonstrated a proof-of-concept spectrometer based on 14 photodiodes, without any external optics or spectral filtering components, in the wavelength range of 450 – 590 nm. Using a non-negative least square (NNLS) algorithm enhanced by orthogonal matching pursuit (OMP), the spectrum of a test light source was reconstructed. Secondly, we monolithically integrated a GaN-based LED with the spectroscopic chip. An optical blocking structure was used to suppress the LED-photodetector interference and was shown to be essential for spectroscopic functionality. A proof of concept using a reflection spectroscopy configuration was experimentally conducted to validate the feasibility of simultaneously operating the LED excitation light source and the photodiodes. Spectral reconstruction using an NNLS algorithm enhanced with OMP was shown to reconstruct the signal from the reflection spectroscopy. No external optics, such as collimation optics and apertures were used. Thirdly, we discussed the spectrometer design conditions enabling an ultrathin form factor to increase the light-harvesting efficiency. We proposed and demonstrated a simple strategy, utilizing the well-established sapphire substrate patterning process to greatly enhance the absorption efficiency while maintaining a large acceptance angle for the incident light. We also demonstrate that spectroscopic performance can be significantly enhanced with this strategy. Finally, we demonstrated a reconstructive spectrometer consisting of 16 spectral encoders to deliver a decent spectral reconstruction performance in the wavelength range of 400 – 650 nm. The accuracies of spectral peak positions and intensity ratios between peaks were found to be 0.97% and 10.4%, respectively. No external optics such as collimation optics and apertures were used, enabled by angle-insensitive light-harvesting structures, including an array of cone-shaped back-reflector fabricated on the underside of the sapphire substrate. The small chip area and a computationally efficient spectral reconstruction algorithm make the proposed spectrometer especially suitable for wearable applications.