Low-Power and Multimodal Organic Photoelectric Synaptic Transistors Modulated by Photoisomerization for UV Damage Perception and Artificial Visual Recognition

Low-power and efficiently parallel neuromorphic computing is expected to break the bottleneck of the von Neumann architecture. Due to the direct responses to optical signals, photonic synaptic devices can work as core components of artificial visual systems, accelerating the development of neural computing. Furthermore, the community is looking for effective coupling of photonic and electronic synaptic behaviors within an individual organic device to achieve further functional integration. Photoisomeric molecules with photo-regulatable properties are expected to facilitate this process. Herein, organic photoelectric synaptic transistors (OPSTs) are constructed by introducing poly(2-(3′,3′-dimethyl-6-nitrospiro[chromene-2,2′-indolin]-1′-yl) ethyl methacrylate) (PSPMA) with photoisomeric groups, which effectively improves the photo-synaptic response. Due to the polarization induction and light-assisted charge trapping of PSPMA, the OPSTs simulate typical photo-synaptic behaviors and achieve significant conductance modulation at low voltage with the assistance of UV light. The power consumption is as low as 84 aJ per event. Moreover, the OPSTs mimic UV nociceptors, recognize handwritten digits with 93.33% accuracy, and decode encrypted optical information, demonstrating the potential of applications in UV damage perception and artificial visual recognition. These findings will expand the application of photoisomeric molecules in photonic synaptic devices, and open up new possibilities for hardware architectures with coupling photonic and electronic synapses.