Superionic rubidium silver iodide gated low voltage synaptic transistor

Abstract

Nonvolatile resistive switching based memristor and memtransistor devices have emerged as a leading platform in neuromorphic computing. In this work, we have fabricated a multifunctional synaptic transistor (ST) using a conjugated polymer P3HT channel and a superionic rubidium silver iodide (RbAg4I5) thin film coated over a polyethylene oxide (PEO) layer as the gate dielectric. Large hysteresis in the transfer curve represents the memristive behavior with at least 105 current On/Off ratio. Enormously large specific capacitance induced by the electrical double layers at the interfaces of PEO/RbAg4I5 dielectric induces polaron (P3HT+) generation in the channel through bound states formation by the electrons with Ag+ ions and consequent movement of iodine (I−) counter ions toward the P3HT channel under a negative gate bias stress. This is strongly supported by the blue shift of the Raman peak from 1444.2 to 1447.9 cm−1 and the appearance of a new peak at 1464.6 cm−1. Interestingly, the proposed ST device exhibits various synaptic actions, which include an excitatory postsynaptic current, paired-pulse facilitation, and short-term potentiation to long-term potentiation after repeated rehearsal on top of standard nonvolatile data storage capability. Our ST also depicts an enhanced retention to 103 s and more than 103 discrete On- and Off-states during potentiation and depression function modulation, respectively, just by consuming a very low energy of about 2.0 pJ per synaptic event. These results are very significant to make this organic synaptic transistor as a potential candidate in terms of the desired metrics for neuromorphic computation at low cost and improved accuracy in the future.

Large-Area Growth of Two-Dimensional Rhenium Disulfide Depicting Robust Artificial Photoreceptor-cum-Optic Nerve Synaptic Functionality

“…Interestingly, the relaxation times extracted from our atomically thick ReS 2 photoreceptor and optic-neural synaptic device are comparable to the actual optical nerves . The smaller interval between two pulses and the increased relaxation time indicate the intensive memory and low forgetting rate, which also confirm the transition of synaptic weight from short-term to long-term potentiation (STP to LTP) …”

“…45 The smaller interval between two pulses and the increased relaxation time indicate the intensive memory and low forgetting rate, which also confirm the transition of synaptic weight from short-term to long-term potentiation (STP to LTP). 46 The effects of light pulse width and intensity are also investigated by recording the change of EPSC (ΔEPSC) responses under a constant bias voltage of 10 V, as shown in Figure 6a,b. When the light pulse width varies from 1 to 50 s, the change in EPSC response enhances gradually.…”

A Singlet-Diradical Co(III)-Dimer as a Nonvolatile Resistive Switching Device: Synthesis, Redox-Induced Interconversion, and Current–Voltage Characteristics

“…This observed conductivity jump from a low (OFF) to a high (ON) state is an exciting feature for molecular memristor devices . Our present work represents a remarkable achievement toward the requirement of desired metrics for nonvolatile RS-based molecular memristor devices . The robustness of our molecular memristor device is further emphasized by testing for resistive random access memory (ReRAM) application in the form of endurance and retention.…”

“…Accordingly, the memristor technology has emerged as an attractive alternative for highly efficient computing platforms such as in-memory and neuromorphic computing. Various materials, including organic semiconductors (OSCs), polymers, , metal oxides, − quantum dots (QDs), and two-dimensional (2D) transition metal dichalcogenides (TMDs), − have been explored to develop high-performance memristors. Among these, the oxide-based devices produce the most promising results; however, their applications are restricted because of the high voltage/current and large set/reset voltage.…”

A self-tuning PID controller based on analog–digital hybrid computing with a double-gate SnS2 memtransistor

“…[7][8][9] In addition, nonlinear and temporal response of memristors to external stimuli enables cognitive sensing/leaning capability in robotic systems, which eventually implements high-speed and low-powered robotic operation over traditional digital computing platforms. [10][11][12][13][14] Particularly, memtransistors (memristors with a three-terminal transistor structure [15][16][17][18][19] ) have garnered significant attention due to their distinct advantages over conventional memristors. One particular noteworthy advantage of memtransistors is their improved reliability.…”