skip to main content

Title: Scalable 3D Ta:SiO x Memristive Devices
A highly reliable memristive device based on tantalum‐doped silicon oxide is reported, which exhibits high uniformity, robust endurance (≈1 × 109 cycles), fast switching speed, long retention, and analog conductance modulation. Devices with junction areas ranging from microscale to as small as 60 × 15 nm2 are fabricated and electrically characterized. ON‐/OFF‐ conductance and reset current show weak area dependence when the device is relatively large, and they become proportional to the device area when further scaled down. Two‐layer devices with repeatable switching behavior are achieved. The current study shows the potentials of Ta:SiO2‐based 3D vertical devices for memory and computing applications. It also suggests that doping of the switching layer is an efficient approach to engineer the performance of memristive devices.
; ; ; ; ;
Award ID(s):
Publication Date:
Journal Name:
Advanced Electronic Materials
Page Range or eLocation-ID:
Sponsoring Org:
National Science Foundation
More Like this
  1. The stabilization of the threshold switching characteristics of memristive [Formula: see text] is examined as a function of sample growth and device characteristics. Sub-stoichiometric [Formula: see text] was deposited via magnetron sputtering and patterned in nanoscale ([Formula: see text]–[Formula: see text]) W/Ir/[Formula: see text]/TiN devices and microscale ([Formula: see text]–[Formula: see text]) crossbar Au/Ru/[Formula: see text]/Pt devices. Annealing the nanoscale devices at 700 [Formula: see text]C removed the need for electroforming the devices. The smallest nanoscale devices showed a large asymmetry in the IV curves for positive and negative bias that switched to symmetric behavior for the larger and microscale devices. Electroforming the microscale crossbar devices created conducting [Formula: see text] filaments with symmetric IV curves whose behavior did not change as the device area increased. The smallest devices showed the largest threshold voltages and most stable threshold switching. As the nanoscale device area increased, the resistance of the devices scaled with the area as [Formula: see text], indicating a crystallized bulk [Formula: see text] device. When the nanoscale device size was comparable to the size of the filaments, the annealed nanoscale devices showed similar electrical responses as the electroformed microscale crossbar devices, indicating filament-like behavior in even annealed devices withoutmore »electroforming. Finally, the addition of up to 1.8% Ti dopant into the films did not improve or stabilize the threshold switching in the microscale crossbar devices.« less
  2. MoS 2 has been reported to exhibit a resistive switching phenomenon in a vertical metal–insulator–metal (MIM) structure and has attracted much attention due to its ultra-thin active layer thickness. Here, the resistance evolutions in the high resistance state (HRS) and low resistance state (LRS) are investigated under constant voltage stress (CVS) or constant current stress (CCS) on MoS 2 resistive switching devices. Interestingly, compared with bulk transition metal oxides (TMO), MoS 2 exhibits an opposite characteristic in the fresh or pre-RESET device in the “HRS” wherein the resistance will increase to an even higher resistance after applying CVS, a unique phenomenon only accessible in 2D-based resistive switching devices. It is inferred that instead of in the highest resistance state, the fresh or pre-RESET devices are in an intermediate state with a small amount of Au embedded in the MoS 2 film. Inspired by the capability of both bipolar and unipolar operation, positive and negative CVS measurements are performed and show similar characteristics. In addition, it is observed that the resistance state transition is faster when using higher electric stress. Numerical simulations have been performed to study the temperature effect with small-area integration capability. These results can be explained by amore »modified conductive-bridge-like model based on Au migration, uncovering the switching mechanisms in the ultrathin 2D materials and inspiring future studies in this area.« less
  3. Abstract

    The superior density of passive analog-grade memristive crossbar circuits enables storing large neural network models directly on specialized neuromorphic chips to avoid costly off-chip communication. To ensure efficient use of such circuits in neuromorphic systems, memristor variations must be substantially lower than those of active memory devices. Here we report a 64 × 64 passive crossbar circuit with ~99% functional nonvolatile metal-oxide memristors. The fabrication technology is based on a foundry-compatible process with etch-down patterning and a low-temperature budget. The achieved <26% coefficient of variance in memristor switching voltages is sufficient for programming a 4K-pixel gray-scale pattern with a <4% relative tuning error on average. Analog properties are also successfully verified via experimental demonstration of a 64 × 10 vector-by-matrix multiplication with an average 1% relative conductance import accuracy to model the MNIST image classification by ex-situ trained single-layer perceptron, and modeling of a large-scale multilayer perceptron classifier based on more advanced conductance tuning algorithm.

  4. Abstract Spiking neural network (SNN) in future neuromorphic architectures requires hardware devices to be not only capable of emulating fundamental functionalities of biological synapse such as spike-timing dependent plasticity (STDP) and spike-rate dependent plasticity (SRDP), but also biodegradable to address current ecological challenges of electronic waste. Among different device technologies and materials, memristive synaptic devices based on natural organic materials have emerged as the favourable candidate to meet these demands. The metal–insulator-metal structure is analogous to biological synapse with low power consumption, fast switching speed and simulation of synaptic plasticity, while natural organic materials are water soluble, renewable and environmental friendly. In this study, the potential of a natural organic material—honey-based memristor for SNNs was demonstrated. The device exhibited forming-free bipolar resistive switching, a high switching speed of 100 ns set time and 500 ns reset time, STDP and SRDP learning behaviours, and dissolving in water. The intuitive conduction models for STDP and SRDP were proposed. These results testified that honey-based memristive synaptic devices are promising for SNN implementation in green electronics and biodegradable neuromorphic systems.
  5. The switching performance of unpackaged vertical geometry NiO/ β -Ga 2 O 3 rectifiers with a reverse breakdown voltage of 1.76 kV (0.1 cm diameter, 7.85 × 10 −3 cm 2 area) and an absolute forward current of 1.9 A fabricated on 20 μ m thick epitaxial β -Ga 2 O 3 drift layers and a double layer of NiO to optimize breakdown and contact resistance was measured with an inductive load test circuit. The Baliga figure-of-merit of the devices was 261 −2 , with differential on-state resistance of 11.86 mΩ.cm 2 . The recovery characteristics for these rectifiers switching from forward current of 1 A to reverse off-state voltage of −550 V showed a measurement-parasitic-limited recovery time (t rr ) of 101 ns, with a peak current value of 1.4 A for switching from 640 V. The reverse recovery time was limited by extrinsic parasitic and thus does not represent the intrinsic device characteristics. There was no significant dependence of t rr on switching voltage or forward current.