AI-augmented photon-trapping spectrometer-on-a-chip on silicon platform with extended near-infrared sensitivity

We present a compact, noise-resilient reconstructive spectrometer-on-a-chip on silicon that achieves high-resolution hyperspectral imaging across an extended near-infrared (NIR) range up to 1100 nm. The device integrates monolithically fabricated silicon photodiodes enhanced with photon-trapping surface textures, enabling improved responsivity in the low-absorption NIR regime. Leveraging a fully connected neural network, we demonstrate accurate spectral reconstruction from only 16 uniquely engineered detectors, achieving <0.05 root mean squared error and ∼8 nm resolution over a wide spectral range of 640 to 1100 nm. Our system outperforms conventional spectrometers, maintaining a signal-to-noise ratio above 30 dB even with 40 dB of added detector noise—extending functionality to longer wavelengths up to 1100 nm, whereas the traditional spectrometers fail to perform beyond 950 nm due to poor detector efficiency and noise performance. With a footprint of 0.4 mm2, dynamic range of 50 dB, ultrafast time response (57 ps), and high photodiode gain (>7000), this AI-augmented silicon spectrometer is well-suited for portable, real-time, and low-light applications in biomedical imaging, environmental monitoring, and remote sensing. The results establish a pathway toward fully integrated, high-performance hyperspectral sensing in a CMOS-compatible platform.