Enhancement of Synaptic Properties using Ta₂O₅/ZnO/Ta₂O₅ Tri-layer Device

This paper investigated the electrical characteristics of Ta<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>O<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</inf>-based synaptic devices by designing and fabricating a device incorporating a ZnO layer, based on critical parameters for learning in neural network-based algorithms. A comparative analysis was conducted between synaptic devices based on a single and double layer. The results show that inserting ZnO into the Ta<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>O<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</inf>-based synaptic device decreases the SET/RESET voltages. Furthermore, the triple-layer in synaptic devices improved yield, switching behavior uniformity, and conductance change uniformity in response to pulse inputs. These findings suggested that the triple-layer structure provided advantages in consistently forming oxygen vacancy-based conductive filaments (CFs) and allowed for better-controlled migration of oxygen ions from a structural perspective.