Huygens metasurface-enabled mid-wave infrared computational spectrometers

The mid-wave infrared (MWIR) spectrum is critical for environmental monitoring, gas detection, and related fields as it enables the identification of molecular fingerprints. Despite recent unprecedented progress, computational spectrometers are still constrained by system performance, including spectral encoding and reconstruction algorithms. Here, we propose a promising mechanism for low-correlation spectral encoding, enabled by Huygens metasurfaces (HMSs) consisting of germanium-antimony-telluride (GST) nanodisk arrays on a calcium fluoride (CaF₂) substrate. Flexible manipulation of electric and magnetic dipole resonances can be realized following Huygens' principle, leading to an average correlation coefficient as low as 0.34 for 16 HMSs. Together with a compressed sensing algorithm, the proposed HMS-enabled spectrometer can achieve a 10 nm spectral resolution over the MWIR spectrum. Narrow spectra and wide-band absorption spectra have also been successfully demonstrated.