High-performance near-Infrared computational spectrometer enabled by finely-tuned PbS quantum dots

The bulky footprint of near-infrared (NIR) spectrometers has been limiting their applications in portable and movable systems for probing molecular compositions and structures. Quantum dot (QD) computational spectrometers are a promising strategy for miniaturized NIR spectrometers, whose performance is limited by the poor spectral encoding matrix and, ultimately, the poor quality of PbS QDs. Here, we show that the monodispersity and finely controlled absorption peak of PbS QDs are critical parameters affecting the spectral resolution and noise resistance. Thus, a facile synthesis of a series of monodisperse PbS QDs from a single batch is developed using cation exchange synthesis in a seeded-growth manner. All the as-synthesized PbS QDs have narrow size distributions of below 4%, and the peak intervals can be controlled to within 3 nm. Furthermore, stable PbS QD inks are prepared by considering the compatibility between QD ligands, solvents, and polymers. The PbS QD filter array is fabricated using a contact printing method, exhibiting supreme transmittance curves and a spectral encoding matrix. The filter array is coupled with an InGaAs image sensor to form the QD NIR computational spectrometer. Thanks to the high-quality PbS QDs, the QD spectrometer shows a high spectral resolution of 1.5 nm in a broad wavelength range of 900-1700 nm and excellent spectral reconstruction of narrow and broad spectra with fidelities of above 0.987. Additionally, the QD spectrometer is applied to distinguish materials and accurately measure the alcohol content of white wines, demonstrating the great potential for practical applications of QD NIR spectrometers.