High Amplitude Spike Generator in Au Nanodot-Incorporated NbO_x Mott Memristor

Abstract

NbO x -based Mott memristors exhibit fast threshold switching behaviors, making them suitable for spike generators in neuromorphic computing and stochastic clock generators in security devices. In these applications, a high output spike amplitude is necessary for threshold level control and accurate signal detection. Here, we propose a materialwise solution to obtain the high amplitude spikes by inserting Au nanodots into the NbO x device. The Au nanodots enable increasing the threshold voltage by modulating the oxygen contents at the electrode-oxide interface, providing a higher ON current compared to nanodot-free NbO x devices. Also, the reduction of the local switching region volume decreases the thermal capacitance of the system, allowing the maximum spike amplitude generation. Consequently, the Au nanodot incorporation increases the spike amplitude of the NbO x device by 6 times, without any additional external circuit elements. The results are systematically supported by both a numerical model and a finite-element-method-based multiphysics model.

A non-invasive approach to the resistive switching physical model of ultra-thin organic–inorganic dielectric-based ReRAMs

“…In filament-based ReRAM devices, filament switching instability has been addressed through the incorporation of Au nanodots at the electrode/oxide interface or the integration of a nanoporous-defective bottom layer to promote filament confinement. 14,15 In the new ultra-thin hybrid-based ReRAMs, the reset process is responsible for most reliability issues. When a reset DC bias is applied, multiple consecutive reset → set events occur, leading to the formation of intermediate resistance states (IRS) as depicted in Fig.…”

Harnessing Metal/Oxide Interlayer to Engineer the Memristive Response and Oscillation Dynamics of Two‐Terminal Memristors

“…Many of these applications rely on the ability of the devices to generate reliable current or voltage spikes. [ 15,26,27 ] While the ability to tune oscillation frequency by an external factor (electric field or a circuit component, such as resistor or capacitor) or an internal device variable is desirable for flexibility in device operation, a sufficiently large amplitude is required for accurate data sensing and noise‐tolerant signal transmission. [ 8,27–31 ]…”

“…A major challenge for the practical use of these devices is that achieving reliable device characteristics with controllable parameters is difficult, since device response depends on many different factors, including device structure, oxide stoichiometry, interfacial effects (Schottky barriers and interface reactions), pre‐ or post‐fabrication treatment (e.g., air exposure, annealing in different environments), and operating condition (the magnitude of electrical stimuli). [ 27,32–38 ] Moreover, memristive switching is generally observed after an initial electroforming step, leading to either a local compositional change, or interfacial ion exchange, or modification of interface barrier which greatly affects subsequent switching response and parameters. [ 39–42 ] Therefore, future development of memristor‐based technology relies on understanding the factors governing switching modes and parameters.…”

Enhancement of NbO2-based oscillator neuron device performance via cryogenic operation

“…However, the voltage division between the barrier layer and the on-state NbO 2 can cause an increase in the hold voltage, leading to a decrease in the oscillation amplitude [15]. Consequently, the reduced amplitude poses a challenge in accurately classifying the frequency of the weighted sum [16,17].…”