Performance optimization of TiO2-xnanowires-based optoelectronic memristive synapse
Metal oxide-based optoelectronic memristor has been considered as significant candidate of developing neuromorphic vision system because of its ability of realization the collection,storage and processing of optical information in single device.However,the rapid recombination of photocon-ductive electrons and holes in metal oxide materials would cause the low linearity of device conduct-ance changes,which limits the development of artificial neural networks for high-precision image rec-ognition.In this work,a TiO2-x nanowires-based optoelectronic memristor was prepared by hydro-thermal method.By adopting a plasma treatment,the optoelectronic performance of the device could be improved,especially the linearity of conductance.Moreover,the short-time plasticity and longtime plasticity optoelectronic synaptic behaviors and high linearity of the conductivity change behavior were obtained in the optoelectronic memristor,and the high precision image recognition was realized as well.The experiment demonstrated that the plasma treatment could introduce oxygen vacancy defects in TiO2-x nanowires and enhanced the persistent photoconductivity effect.
万方数据
