HgTe Nanocrystal-Gated Infrared Phototransistors With High Sensitivity, Extended Dynamic Range, and In-Sensor Image Processing Capability.

Mercury telluride (HgTe) nanocrystals (NCs) offer adjustable absorption and solution-processable fabrication, making them promising materials for low-cost, high-resolution imaging across a wide infrared (IR) spectrum. However, photodetectors based on HgTe NCs often suffer from high dark current, elevated noise arising from trap states and interface defects, and limited structural tunability, which constrain their sensitivity, dynamic range, and applicability in intelligent vision applications. Here, it is reported a dual-gate carbon nanotubes (CNTs) field-effect transistor incorporating an HgTe NC-based PIN heterojunction as the top gate, which converts incident IR light into a photovoltage that functions as a dynamic optical gate, while an independently addressable local bottom gate adjusts the carrier concentration in the CNT channel. This opto-electrically decoupled yet synergistic architecture enables high responsivity (>10 A/W), excellent room-temperature specific detectivity (10 Jones) under low-power IR illumination, and a wide dynamic range of 170 dB to 1650 nm infrared irradiation when biased in the subthreshold region. Furthermore, by leveraging gate-controllable and self-adaptive photoresponse, it is demonstrated in-sensor convolutional processing and image fusion at the device level. This dual-gate architecture provides a new pathway toward high-performance IR photodetectors with in-sensor computing capabilities, advancing their potential for next-generation machine vision systems.