ZnCdSe/ZnS quantum dot hybrid film organic optoelectronic synaptic transistor for high-efficiency neuromorphic computing

Abstract

Optoelectronic synaptic transistors have great potential for application in neuromorphic computing, and by integrating sensing and computing capabilities onto a single optoelectronic synaptic transistor, it is expected to overcome the inherent limitations of computing based on the von Neumann architecture. This paper reports an electrolyte gate-controlled organic optoelectronic synaptic transistor. The device uses polyacrylic acid with an electric double layer effect as the gate dielectric layer and a solution method ZnCdSe/ZnS quantum dot and poly[2,5-(2-octyldodecyl)-3,6-dione pyrrole-pyrrole-5,5-(2,5-di(thien-2-yl)thienyl)[3,2-b]thienyl] hybrid film as the semiconductor layer. By taking advantage of the excellent light absorption properties and good charge transport performance of the hybrid film, the separation efficiency of photogenerated carriers can be significantly improved. This results in efficient photoelectric conversion. Taking advantage of this unique feature, the device successfully simulated typical synaptic functions as well as long-term characteristics, enabling the switching of multiple logical functions, supporting Pavlovian conditioned reflex simulation and visual object recognition simulation, with handwritten digit recognition accuracy reaching 91.6%. This work offers broad prospects for photoelectric pulse modulation based on organic synaptic devices and shows great potential in the development of artificial intelligence.