Enhancing near-infrared sensitivity with GaAs-based hybrid CMOS image sensors

Complementary metal-oxide-semiconductor (CMOS) image sensors are widely used in applications such as time-of-flight systems, machine vision, mobile devices, and medical imaging due to their versatility and compact structure. However, enhancing their sensitivity in the near-infrared (NIR) region remains a significant challenge, largely due to insufficient light absorption within this spectral range. We propose an innovative solution to this limitation by integrating GaAs semiconductor material with silicon in a hybrid CMOS image sensor. The design incorporates a 2 μm×2 μm pixel architecture combined with a deep trench isolation structure, which significantly enhances light absorption and efficiency in the NIR region. This approach achieves more than a twofold increase in NIR sensitivity in the 800- to 1000-nm wavelength range compared with conventional backside-illuminated CMOS image sensors. Simulation results highlight the substantial potential of the GaAs-based hybrid structure in advancing the performance of CMOS image sensors for applications requiring improved sensitivity in the NIR region. To the best of our knowledge, the proposed structure is the first of its kind reported in the literature.