Abstract
Chip-level spectrometers provide a stable and cost-effective solution for spectral analysis in various applications. Here we present a silicon on-chip digital Fourier transform spectrometer consisting of eight cascaded optical switches connected by delay waveguides. By configuring the states of the optical switches, this chip can realize 127 Mach–Zehnder interferometers with linearly increased optical path differences. A machine-learning regularization method is utilized to reconstruct the spectrum. Experimental results show that our chip can retrieve both sparse and broadband optical spectra with negligible reconstruction errors. The spectral resolution can be further improved by cascading more stages of optical switches. Our method has the advantages of compact size, high scalability, and high signal-to-noise ratio, making it a promising candidate for realizing miniaturized spectrometers.
© 2022 Optical Society of America
Full Article | PDF ArticleMore Like This
Huijie Wang, Zhongjin Lin, Qifeng Li, and Wei Shi
Opt. Lett. 44(11) 2923-2926 (2019)
Hugh Podmore, Alan Scott, Pavel Cheben, Aitor V. Velasco, Jens H. Schmid, Martin Vachon, and Regina Lee
Opt. Lett. 42(7) 1440-1443 (2017)
Huijie Wang, Qifeng Li, and Wei Shi
Opt. Lett. 45(6) 1479-1482 (2020)