Search for: All records

Chip-scale integrated imaging spectrometers show significant potential for high-performance spectral analysis due to advancements in fabrication and computational techniques. Many practical applications, such as astronomy and molecular spectroscopy, require analyzing light at sub-nanowatt levels, where inherent enhancement in spectrometer signals can reduce the need for expensive photodetectors or long integration time. Previously, we introduced an integrated spectrometer scheme using machine learning to reconstruct spectra from imaging the wavelength-dependent patterns scattered out of a multimode interference (MMI) waveguide. In this work, we report a signal enhancement of 13.6 dB and an increase of device sensitivity and dynamic range by 15 dB by selective roughening of the waveguide surface via plasma etching. By imaging interference patterns at various points along the waveguide, we determine that the best spectrometer performance is achieved by imaging MMI sections with highest pattern variation. We report accurate spectral measurements using convolutional neural network-based spectral reconstruction with 1 nm resolution at input powers as low as 300 pW for the present experimental configuration, and a scattering coefficient of 1.109 cm^-1from the etched section. 

Spectral analysis of light is one of the oldest and most versatile scientific methods and the basis of countless techniques and instruments. Miniaturized spectrometers have recently seen great advances, but challenges remain before they are widely deployed. We report an integrated photonic spectrometer that achieves high performance with minimal component complexity by combining imaging of light propagation patterns in multi-mode interference waveguides with machine learning analysis. We demonstrate broadband operation in the visible and near-infrared, 0.05 nm spectral resolution, and an array of four spectrometers on a single chip. Two canonical applications are implemented: spectral analysis of the solar spectrum with neural network reconstruction and detection of Rayleigh scattering from microbeads on an optofluidic chip using principal component classification. These results illustrate the potential of this approach for high-performance spectroscopy across disciplines.