Exploring Area-Dependent Pr0.7Ca0.3MnO3-Based Memristive Devices as Synapses in Spiking and Artificial Neural Networks

Abstract

Memristive devices are novel electronic devices, which resistance can be tuned by an external voltage in a non-volatile way. Due to their analog resistive switching behavior, they are considered to emulate the behavior of synapses in neuronal networks. In this work, we investigate memristive devices based on the field-driven redox process between the p-conducting Pr0.7Ca0.3MnO3 (PCMO) and different tunnel barriers, namely, Al2O3, Ta2O5, and WO3. In contrast to the more common filamentary-type switching devices, the resistance range of these area-dependent switching devices can be adapted to the requirements of the surrounding circuit. We investigate the impact of the tunnel barrier layer on the switching performance including area scaling of the current and variability. Best performance with respect to the resistance window and the variability is observed for PCMO with a native Al2O3 tunnel oxide. For all different layer stacks, we demonstrate a spike timing dependent plasticity like behavior of the investigated PCMO cells. Furthermore, we can also tune the resistance in an analog fashion by repeated switching the device with voltage pulses of the same amplitude and polarity. Both measurements resemble the plasticity of biological synapses. We investigate in detail the impact of different pulse heights and pulse lengths on the shape of the stepwise SET and RESET curves. We use these measurements as input for the simulation of training and inference in a multilayer perceptron for pattern recognition, to show the use of PCMO-based ReRAM devices as weights in artificial neural networks which are trained by gradient descent methods. Based on this, we identify certain trends for the impact of the applied voltages and pulse length on the resulting shape of the measured curves and on the learning rate and accuracy of the multilayer perceptron.

Flexible Active Crossbar Arrays Using Amorphous Oxide Semiconductor Technology toward Artificial Neural Networks Hardware

“…In Supporting Information S7, retention data of seven states can be observed for 10 3 s. The states remain distinct despite some retention loss, expected from the interfacetype of resistive switching. [53,54] The retention time is, therefore, the remaining challenge especially for these types of resistive switching. A strong nonlinearity is required to tackle the issue of voltage-time dilemma.…”

Epitaxial La0.5Sr0.5MnO3‐δ Bipolar Memristive Devices with Tunable and Stable Multilevel States

“…[3,10] On top of that, VCMs can be used for neuromorphic computing, where the device itself shows the behavior that is typically hardcoded in most AI applications, and hence, VCMs may present larger storage density and lower power consumption in applications such as those involving AI. [11][12][13] In our previous work, we studied pure LaMnO 3+δ , and proved that it was possible to change the resistance state up to two orders of magnitude by oxygen drift and by the concomitant redox reactions taking place at the materials' surface, as experimentally proven by combining conductive atomic force microscopy with X-ray photoemission electron spectroscopy Valence change memories are novel data storage devices in which the resistance is determined by a reversible redox reaction triggered by voltage. The oxygen content and mobility within the active materials of these devices play a crucial role in their performance.…”

“…The memristive properties of LSM50 are studied using titanium as an active electrode. As for other material combinations, [3,11,13,[33][34][35][36][37][38][39] the Ti/LSM50 stack can undergo a reversible redox reaction that originates two additional interlayers: oxidized metal electrode and oxygen-deficient oxide. The resistivity of both interlayers is higher than the parent compound.…”

La2NiO4+δ‐Based Memristive Devices Integrated on Si‐Based Substrates

“…Besides, a strong renewal of interest for non-von Neumann computing architecture has recently led to the development of solid state-based synapses dedicated to neuromorphic computing applications using memristive oxides as active layers. [8][9][10][11][12] Oxide-based valence-change memories (VCMs), in which a field-driven oxygen motion triggers the resistance switching of the device by internal redox reactions (the local change of valence in the cation sublattice), are one of the most promising candidates for memory and artificial synapse applications. The valence change can occur in localized regions (filament-type RS) or over the entire memristor area (interface-type RS).…”

“…Besides, a strong renewal of interest for non‐von Neumann computing architecture has recently led to the development of solid state‐based synapses dedicated to neuromorphic computing applications using memristive oxides as active layers. [ 8–12 ]…”