Generalized wavefront-engineered metalens integrated with colloidal quantum dot detectors for on-chip near-infrared imager

High-performance near-infrared (NIR) imaging is essential for non-destructive industrial inspection and advanced sensing. While chip-scale integration of metalenses with colloidal quantum dot (CQD) detectors provides a cost-effective and miniaturized pathway, performance is often limited by the high sensitivity of metalenses to fabrication imperfections. Here, we propose a generalized wavefront-engineering methodology (GWEM) to enhance fabrication robustness and efficiency by optimizing the spatial distribution of meta-atoms via equiphase surface translation. This approach preserves near-diffraction-limited focusing while maximizing the density of optimal-sized nanostructures to ensure high fabrication tolerance. As a proof of concept, we fabricated a GWEM with a numerical aperture of 0.38, achieving 85% transmission and 78% focusing efficiency at λ = 1700 nm. By further integrating this metalens with a CQD image sensor, we successfully demonstrated a functional on-chip NIR imager that achieved high-fidelity imaging through silicon wafers. This direct stacking integration approach provides a scalable and low-cost solution that paves the way for high-performance compact imaging systems in diverse fields such as industrial quality control, environmental monitoring, and portable intelligent sensing.