Polarization‐Enhanced Narrow‐Band GeS2 2‐D SWIR Spectral Phototransistor

Integrated computational spectrometers with gate‐tunable nano heterostructures and reconstruction algorithms are attractive for on‐chip gas‐sensing spectrometers and have enabled versatile spectrum detectors. However, they require the selective and optical filtering capabilities of wavelengths, restricting their efficient implementation in narrow‐band photodetection. In this study, a printable spectral phototransistor is developed with high dynamic detectivity (1012 Jones and 105 Hz at −3 db bandwidth) modulated by a GeS2 nanosheet heterostructure at short‐wave infrared (SWIR) regime. Using the transport mode switching of carriers in a heterostructure and the polarization‐sensitivity of the GeS2 two‐dimension (2‐D) nanosheet, this SWIR spectral phototransistor demonstrates an accurate narrow‐band selective (96.7% accuracy) spectrum detector and performed a deep‐learning analysis of an artificial neural network (ANN). Furthermore, this GeS2 2‐D based spectral phototransistor, characterized by its high in‐plane anisotropy and electrically reconfigurable properties, extends the applicability of narrow‐band photodetection with 15 nm Full Width at Half Maximum (FWHM) to the recognition of trace‐gases at the parts per billion (ppb) level. A polarization‐sensitive GeS2/PbSe two‐dimensional (2‐D short‐wave infrared (SWIR) spectral phototransistor, integrated with an artificial neural network (ANN) for deep‐learning, reconstructs infrared spectra (900–1700 nm) with tunable VG and δ. This detector achieves 96.7% accuracy in narrow‐band selection, supports at least 8 multi‐wavelengths’ channels, and narrow FWHM (15 nm), which is crucial for future hyperspectral detecting applications.