Robust Ag/ZrO₂/WS₂/Pt Memristor for Neuromorphic Computing

Abstract

The development of the information age has made resistive random access memory (RRAM) a critical nanoscale memristor device (MD). However, due to the randomness of the area formed by the conductive filaments (CFs), the RRAM MD still suffers from a problem of insufficient reliability. In this study, the memristor of Ag/ ZrO 2 /WS 2 /Pt structure is proposed for the first time, and a layer of two-dimensional (2D) WS 2 nanosheets was inserted into the MD to form 2D material and oxide double-layer MD (2DOMD) to improve the reliability of single-layer devices. The results indicate that the electrochemical metallization memory cell exhibits a highly stable memristive switching and concentrated ON-and OFF-state voltage distribution, high speed (∼10 ns), and robust endurance (>10 9 cycles). This result is superior to MDs with a single-layer ZrO 2 or WS 2 film because two layers have different ion transport rates, thereby limiting the rupture/rejuvenation of CFs to the bilayer interface region, which can greatly reduce the randomness of CFs in MDs. Moreover, we used the handwritten recognition dataset (i.e., the Modified National Institute of Standards and Technology (MNIST) database) for neuromorphic simulations. Furthermore, biosynaptic functions and plasticity, including spike-timing-dependent plasticity and paired-pulse facilitation, have been successfully achieved. By incorporating 2D materials and oxides into a doublelayer MD, the practical application of RRAM MD can be significantly enhanced to facilitate the development of artificial synapses for brain-enhanced computing systems in the future.

A Pure 2H‐MoS2 Nanosheet‐Based Memristor with Low Power Consumption and Linear Multilevel Storage for Artificial Synapse Emulator

“…The Hebbian learning rule, an essential neural theory, reflects the effect of pre‐synaptic and post‐synaptic activity 40,41. One of the most important biological features underlying this theory is the spike‐time‐dependent plasticity (STDP) 42,43. Adjusting the weights of synapses according to the time (Δ t ) is an important feature of STDP.…”

Emerging Memristive Artificial Synapses and Neurons for Energy‐Efficient Neuromorphic Computing

“…The electrochemical reaction is responsible for the switching behavior in this type of synapse, [36,54] which generally utilizes electrochemically active (e.g., Ag and Cu) and inert (e.g., Pt and Au) metals as the two-terminal electrodes. [39,[64][65][66][67][68][69][70][89][90][91][92][93][94] Furthermore, diverse solid electrolyte materials, such as Ag 2 S, [39,67] Cu 2 S, [89] α-Si, [90] SiO 2 , [92] and Ta 2 O 5 , [93,94] have been used as the active layer. For example, when a negative bias is applied to the inert electrode, the active Ag (or Cu) electrode at the interface becomes oxidized by the charge transfer reaction, and thus resulting in the formation of Ag + cations.…”

“…Hence, a wide range of electrochemical metallization memristive synapses has recently been suggested and developed. [39,[64][65][66][67][68][69][70][89][90][91][92][93][94] For example, Shi et al reported amorphous ZrTe alloy/Al 2 O 3 /Ta memristive synapses. [64] In this structure, the Te-based conductive filament grew by the migration and nucleation of Te + cations at the interface that can be controlled by the applied electric field.…”

Memristors Based on 2D Materials as an Artificial Synapse for Neuromorphic Electronics

“…Finally, heterojunction-structure devices, where the active switching layer consists of 2D-2D or 2D-3D hybrid heterojunctions instead of a single material, operate based on the filament formation, charge trapping, and vacancy migration mechanisms unlike the previously suggested structures. [62][63][64] While phase transition, [65][66][67] vacancy migration, [68][69][70] filament formation, [71][72][73] and charge trapping [74][75][76] are reported commonly in both bulk and 2D materials, migration of specific defects, phase transition induced by cation migration, and SE to DT transition in a atomically thin active layer are uniquely observed in the 2D platform. Table 1 categorizes the mechanisms of the memristive switching according to the structures and materials and further summarizes synaptic functions with characteristic features such as SET voltage (V SET ), operation current, power consumption, and stability.…”

“…d) STDP characteristics with 0.8 V pulse amplitude and 100 ns pulse width. a-d) Reproduced with permission [62]. Copyright 2019, American Chemical Society.poly(ethylene naphthalate) foil substrates with inkjet-printed bottom Ag electrodes, were oxidized in air at 200 °C to form an ultrathin (<3 nm) oxide layer as a switching layer.…”