A near-infrared colloidal quantum dot imager with monolithically integrated readout circuitry

Imagers that operate in the near-infrared region (wavelengths of 0.7–1.4 µm) are of use in applications such as material sorting, machine vision and autonomous driving. However, such imagers typically use the flip-chip method to connect infrared photodiodes with silicon-based readout integrated circuits, as the need for high-temperature processing and single-crystalline substrates prevents direct integration. This increases processing complexity and cost. Here we report high-resolution imagers that monolithically integrate near-infrared colloidal quantum dot photodiodes with complementary metal–oxide–semiconductor readout integrated circuits. The colloidal quantum dot photodetector is designed with a structure compatible with complementary metal–oxide–semiconductors and exhibits a spectral range of 400–1,300 nm, room-temperature detectivity of 2.1 × 1012 Jones, −3 dB bandwidth of 140 kHz and linear dynamic range of over 100 dB. With this approach, we create a large (640 × 512 pixels) imager that exhibits a spatial resolution of 40 line pairs per millimetre at a modulation transfer function of 50%, and we show that it can be used for vein imaging and matter identification. Colloidal quantum dot photodetectors can be integrated with complementary metal–oxide–semiconductor readout integrated circuits, resulting in low-cost, high-resolution infrared imagers.