940 Nm Near-Infrared Photosynapses Based on Sn─Pb Perovskite for Efficient Face Recognition

Tin-lead perovskites offer great potentials for neuromorphic optoelectronics owing to their narrow bandgap and robust near-infrared (NIR) absorption. However, high-performance three-terminal artificial synapses based on these materials remain scarce due to challenges in forming high-quality semiconductor films. Here, we demonstrate a perovskite synaptic field-effect transistor (FET) capable of efficient 940 nm sensing and neuromorphic modulation, enabled by uniform, high-crystallinity FASn_0.8Pb_0.2I₃ thin films. A molecular additive, 1-bromo-4-(methylsulfinyl)benzene (BMSB), precisely regulates crystallization, enlarges grains, and suppresses trap formation, thereby reducing ion migration and enhancing charge transport. The optimized devices achieve high hole mobility and an exceptional responsivity of 231 A W^-1 at 940 nm, marking the first demonstration of efficient 940 nm infrared photoresponse in three-terminal perovskite artificial synapses. Benefiting from balanced ion-electron coupling, the devices exhibit reliable synaptic behaviors, including excitatory postsynaptic currents, paired-pulse facilitation, and learning-forgetting cycles. Integrated into a reservoir-computing framework, the synaptic FETs enable accurate NIR facial recognition, underscoring their potential for in-sensor computing. This work establishes a molecular-level strategy to harmonize ionic and electronic processes in Sn─Pb perovskites, advancing light-programmable neuromorphic transistors for next-generation intelligent NIR vision systems.