Emulating synaptic plasticity in ionic liquid-gated zinc tin oxide neuromorphic transistor

Real-time interpretation and modification of information can impart revolutionary advancements in the evolution of the internet of things. Synaptic transistors belong to a class of devices that mimic a biological synapse's fundamental characteristic functions. Moreover, the adaptive nature makes synaptic transistors the fundamental architectural element towards realizing the artificial brain. An imidazolium-based ionic liquid (IL)-gated zinc tin oxide (ZTO) synaptic transistor on a rigid substrate has been demonstrated here. The ZTO/IL interface was studied using electrochemical impedance spectroscopy, and it was observed that an electrical double-layer capacitance was formed below 1 kHz. Voltage dependent capacitance measurement shows a reasonably stable capacitance in the operating voltage of the synaptic transistor. The estimated capacitance was 0.58 mu F at 20 Hz, resulting in high saturation mobility of 8.73 cm(2) V-1 s(-1) for the fabricated device. The transistor successfully mimics the fundamental biological synaptic functions such as short-term plasticity, long-term plasticity, and paired pulse facilitation. These devices consume very little power of the order of a few nJ/spike. Short-term plasticity is associated with the ion migration dynamics in IL and the IL gate bias induced conductivity change in ZTO channel layer is responsible for long-term plasticity. Device trained with 60 training pulses obtained a synaptic weight change of 768%.