Single-Crystal PZT-Driven Organic Piezo-Phototronic Adaptive Transistors Toward Advanced Spatiotemporal Visual Computing.

Integrating event detection and grayscale sensing in a single pixel/transistor enables compact, intelligent, flexible neuromorphic spatiotemporal visual imaging. Memory phototransistors based on organic phase-change semiconductors (OPCSs) are promising due to the high theoretical photo-sensing-storage capacity, excellent conductance linearity/symmetry, and intrinsic flexibility. However, such systems are constrained by low phase-change efficiency (narrow memory window/capacity) arising from weak and poorly controllable organic molecular interactions, restricting complex feature extraction and increasing energy consumption during information perception and processing. Here, we propose a single-crystal PZT-driven piezo-phototronic organic adaptive memory transistor (OAMT) with optimized stress distribution and multi-field control, significantly enhancing molecular conformation transition efficiency under low-power operation. The device achieves a record memory window capacity factor (γ) of ∼0.87 at a subthreshold swing (SS) of 200 mV/decade, with over 90% recognition accuracy from the OAMT device's actual LTP/LTD synaptic functions in neuromorphic simulations. Furthermore, the device's adaptive multistage phase-transition behavior in response to varying UV pulse densities enables stable current changes-transitioning from molecular conformation 2 to mixed conformations (1+2) in the PCS layer-as well as transient current spikes from conformation 2 to 1. The device simulates real-time flight attitude and dynamic grayscale detection via precise spatio-temporal synchronization, showing great potential for advanced visual technology.