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

Area of Science:

• Materials Science
• Optoelectronics
• Neuroscience

Background:

• Tin-lead perovskites possess narrow bandgaps and strong near-infrared absorption, making them promising for neuromorphic optoelectronics.
• High-performance three-terminal artificial synapses using these materials are limited by difficulties in achieving high-quality semiconductor films.

Purpose of the Study:

• To develop a high-performance three-terminal perovskite synaptic field-effect transistor (FET) for efficient near-infrared (NIR) sensing and neuromorphic modulation.
• To address challenges in forming high-quality semiconductor films for perovskite-based artificial synapses.

Main Methods:

• Fabrication of FASn0.8Pb0.2I3 thin films using a molecular additive (1-bromo-4-(methylsulfinyl)benzene, BMSB) to control crystallization and film quality.
• Characterization of the synaptic FETs' photoresponse, synaptic behaviors, and performance in a reservoir-computing framework for NIR facial recognition.

Main Results:

• Achieved uniform, high-crystallinity FASn0.8Pb0.2I3 thin films, enabling efficient 940 nm sensing and neuromorphic modulation.
• Demonstrated exceptional responsivity (231 A W-1) at 940 nm, the first for three-terminal perovskite artificial synapses.
• Exhibited reliable synaptic functions (e.g., excitatory postsynaptic currents, learning-forgetting cycles) due to balanced ion-electron coupling.
• Successfully implemented the synaptic FETs in a reservoir-computing system for accurate NIR facial recognition.

Conclusions:

• A molecular-level strategy effectively harmonizes ionic and electronic processes in Sn-Pb perovskites for advanced neuromorphic transistors.
• The developed perovskite synaptic FETs show significant potential for next-generation intelligent NIR vision systems and in-sensor computing.