Multiple Resistance States Research Articles

Van der Waals Multiferroic Tunnel Junctions.

Multiferroic tunnel junctions (MFTJs) have aroused significant interest due to their functional properties useful for nonvolatile memory devices. So far, however, all of the existing MFTJs have been based on perovskite-oxide heterostructures limited by a relatively high resistance-area (RA) product unfavorable for practical applications. Here, using first-principles calculations, we explore spin-dependent transport properties of van der Waals (vdW) MFTJs which consist of two-dimensional (2D) ferromagnetic FenGeTe2 (n = 3, 4, 5) electrodes and 2D ferroelectric In2Se3 barrier layers. We demonstrate that such FemGeTe2/In2Se3/FenGeTe2 (m, n = 3, 4, 5; m ≠ n) MFTJs exhibit multiple nonvolatile resistance states associated with different polarization orientation of the ferroelectric In2Se3 layer and magnetization alignment of the two ferromagnetic FenGeTe2 layers. We find a remarkably low RA product (less than 1 Ω·μm2) which makes the proposed vdW MFTJs superior to the conventional MFTJs in terms of their promise for nonvolatile memory applications.

High-throughput phase-field simulations and machine learning of resistive switching in resistive random-access memory

Metal oxide-based Resistive Random-Access Memory (RRAM) exhibits multiple resistance states, arising from the activation/deactivation of a conductive filament (CF) inside a switching layer. Understanding CF formation kinetics is critical to achieving optimal functionality of RRAM. Here a phase-field model is developed, based on materials properties determined by ab initio calculations, to investigate the role of electrical bias, heat transport and defect-induced Vegard strain in the resistive switching behavior, using MO2−x systems such as HfO2−x as a prototypical model system. It successfully captures the CF formation and resultant bipolar resistive switching characteristics. High-throughput simulations are performed for RRAMs with different material parameters to establish a dataset, based on which a compressed-sensing machine learning is conducted to derive interpretable analytical models for device performance (current on/off ratio and switching time) metrics in terms of key material parameters (electrical and thermal conductivities, Vegard strain coefficients). These analytical models reveal that optimal performance (i.e., high current on/off ratio and low switching time) can be achieved in materials with a low Lorenz number, a fundamental material constant. This work provides a fundamental understanding to the resistive switching in RRAM and demonstrates a computational data-driven methodology of materials selection for improved RRAM performance, which can also be applied to other electro-thermo-mechanical systems.

Optically triggered multilevel resistive switching characteristics of Cu/MoS2/AlN/ITO bilayer memory structure

In this work, the tunable resistive switching (RS) functionality of a Cu/MoS2/AlN/ITO nanostructured device is systematically investigated in dark and white light illumination. The device exhibits bi-state RS behavior in the dark ambient, whereas light illumination induces an extra intermediate resistance state and provides controllable tri-state RS characteristics. A conceptual model is proposed and discussed to elucidate the origin of the switching behavior of two resistance states and multiple resistance states of the device. Under the dark ambient condition, the high resistance state and the low resistance state in the device could be ascribed to the formation/rupture of a Cu metallic filamentary path between the electrodes. However, the formation of an additional ionic filament via trapping/detrapping of electrons in nitride-sulfide-related vacancies along with the Cu metallic filament is responsible for the tri-state switching under the light illumination. Interestingly, the variation of SET voltage with applied light intensity has also been demonstrated. The calculated value of the temperature coefficient and temperature dependency of resistance in various resistance states confirms the existence of the proposed model. The device performed a good undispersed endurance up to 1.5 × 103 cycles and stable retention over 103 s at room temperature. This optical activity dependent functionality of the device provides a possibility to extend resistive switching-based nonvolatile random access memory applications to the optical domain such as imaging sensors, photodetectors, and optoelectronic switches.

Multibit optoelectronic memory using graphene/diamond (carbon sp2-sp3) heterojunctions and its arithmetic functions

This work demonstrates that graphene/diamond (carbon sp2-sp3) heterojunctions can be used as multibit optoelectronic memory, where light information is stored as multilevel resistance in a nonvolatile manner. The carbon heterojunctions exhibit a large memory switching ratio of ∼104 and a retention time of >100 min, which allows for multilevel and nonvolatile data storage of optical information. The carbon heterojunctions also exhibit an apparent response to optical pulses, and the output current (conductivity of the junctions) increased linearly in response to the total number of optical pulses. Simple optical arithmetic operations such as accumulation, subtraction, and counting can be performed by using the multiple resistance states of the heterojunctions. The carbon heterojunctions have light sensing, memory, and arithmetic functions in a single device, and they are expected to pave the way for the production of innovative optical computing devices with multifunctional integration of sensing, memory, and calculation functions.

Reaction-Drift Model for Switching Transients in Pr₀.₇Ca₀.₃MnO₃-Based Resistive RAM

Pr$_{0.7}$Ca$_{0.3}$MnO$_3$ (PCMO) based RRAM shows promising memory properties like non-volatility, low variability, multiple resistance states and scalability. From a modeling perspective, the charge carrier DC current modeling of PCMO RRAM by drift diffusion (DD) in the presence of fixed oxygen ion vacancy traps and self-heating (SH) in Technology Computer Aided Design (TCAD) (but without oxygen ionic transport) was able to explain the experimentally observed space charge limited conduction (SCLC) characteristics, prior to resistive switching. Further, transient analysis using DD+SH model was able to reproduce the experimentally observed fast current increase at ~100 ns timescale, prior to resistive switching. However, a complete quantitative transient current transport plus resistive switching model requires the inclusion of ionic transport. We propose a Reaction-Drift (RD) model for oxygen ion vacancy related trap density variation, which is combined with the DD+SH model. Earlier we have shown that the Set transient consists of 3 stages and Reset transient consists of 4 stages experimentally. In this work, the DD+SH+RD model is able to reproduce the entire transient behavior over 10 ns - 1 s range in timescale for both the Set and Reset operations for different applied biases and ambient temperatures. Remarkably, a universal Reset experimental behavior, log(I) is proportional to (m X log(t)) where m~-1/10 is reproduced in simulations. This model is the first model for PCMO RRAMs to significantly reproduce transient Set/Reset behavior. This model establishes the presence of self-heating and ionic-drift limited resistive switching as primary physical phenomena in these RRAMs.

Understanding of multiple resistance states by current sweeping in MoS2-based non-volatile memory devices

Recently, various two-dimensional materials have been reported to exhibit non-volatile resistance switching phenomenon. The atomristors, featuring memristor effect in atomically thin nanomaterials such as monolayer transition metal dichalcogenides and hexagonal boron nitride, have drawn much attention due to the extremely thin active layer thickness with the advantages of forming-free characteristic, large on/off resistance ratio and fast switching speed. To investigate the switching mechanisms in the 2D monolayers, we introduced an electrical characterization method by current sweeping to illustrate the detailed information hidden in the commonly used voltage-sweep curves. Multiple transition steps have been observed in the SET process of MoS2-based resistance switching devices. The different behaviors of transition steps were attributed to the number of defects or vacancies associated with the switching phenomenon, which is consistent with the previously reported conductive-bridge-like model for 2D atomristors. This work provides an approach using current sweeping to precisely characterize the resistance switching effect and inspires further research to optimize the defect distribution in 2D materials for the applications in multi-bit non-volatile memory and neuromorphic computing.

Non-volatile resistance switching in LaNiO3 films on PMN-PT substrates

Using perovskite oxide films of LaNiO3 as a research object, we demonstrate that the lattice strain has a huge effect on the electron transport properties of the film. The LaNiO3 film was deposited on the PMN-PT substrate, a large change in resistance is generated under the electric field. The substrate will be polarized under an electric field and induced large lattice strain due to the converse piezoelectric effect, the resistivity of the LaNiO3 film is modulated up to three orders of magnitude. Moreover, multiple stable resistance states with endurance properties and good retention can be obtained at room temperature. The modulation on resistive states of LaNiO3 films under an electric field can be used for other oxide films and provides a simple and low energy consumption way to construct multistate resistive memory devices.

Designing Multi‐Level Resistance States in Graphene Ferroelectric Transistors

AbstractConventional memory elements code information in the Boolean “0” and “1” form. Devices that exceed bistability in their resistance are useful as memory for future data storage due to their enhanced memory capacity, and are also a necessity for contemporary applications such as neuromorphic computing. Here, with the aid of an experimentally validated device model, design rules are outlined and more than two stable resistance states in a graphene ferroelectric field‐effect transistor are experimentally demonstrated. The design methodology can be extrapolated for on‐demand introduction of multiple resistance states in ferroelectric transistors for applications both in data storage and neuromorphic computing.

Electronic mechanism for resistive switching in metal/insulator/metal nanodevices

Passing current at given threshold voltages through a metal/insulator/metal sandwich structure device may change its resistive state. Such switching has been rationalized by ion drift models, or changes in electronic states, but the underlying physical mechanism is poorly understood. We propose a new model based on electrostatics to explain multiple resistive states in memristors that contain large defect densities. The different resistive states are due to spontaneously charged states of the insulator ‘storage medium’, characterized by different ‘band bending’ solutions of Poisson’s equation. For an insulator with mainly donor type defects, the low-resistivity state is characterized by a negatively charged insulator due to convex band bending, and the high-resistivity state by a positively charged insulator due to concave band bending; vice versa for insulators with mainly acceptor type defects. We show that these multiple solutions coexist only for nanoscale devices and for bias voltages limited by the switching threshold values, where the system charge spontaneously changes and the system switches to another resistive state. We outline the general principles how this functionality depends on material properties and defect abundance of the insulator ‘storage medium’.

Process Dependent Optimization of Dielectric and Metal Stacks for Multilevel Resistive Random-Access Memory

A HfO2/Al2O3 bilayer structure for a two-terminal ReRAM device with an intention of having multiple resistance states as a function of compliance current (CC) after forming was evaluated. A reduced power consumption was observed when the Al2O3 buffer layer was placed between the top electrode and the HfO2 layer as compared to when it is embedded between the HfO2 layer and the bottom electrode. Gradual resistance change capability was observed with varying CC. The switching power requirement increases even if the Al2O3 buffer layer thickness was decreased when the buffer layer was near the bottom electrode. It was demonstrated that by modifying the deposition process of the top metal layer the switching energy requirement can be altered.

Low power high speed 3-bit multilevel resistive switching in TiO2 thin film using oxidisable electrode

Multilevel cell (MLC) capability of a memory device is an attractive alternative to realise high integration density of memory arrays. In addition, power efficient memory cell along with MLC capability has huge potential in neuromorphic computing and data storage applications. In this study low power, high speed and 3-bit multilevel bipolar resistance switching operation is reported in TiO2 thin films where copper is used as counter electrode. On account of using an oxidisable electrode, low programming voltages (∼+0.2 V, −0.1 V) along with low reset current (∼10 µA) are observed, thereby ensuing low power (∼µW) memory operation. Additionally, observation of fast switching (∼250 ns) translated to ultra-low energy (∼pJ) consumption. By varying compliance current (IC) from 25 µA to 1.6 mA during the set process, eight distinct non-volatile resistance states are reliably obtained, which showed statistically significant and non-overlapping resistance distribution. The multilevel resistive switching in Cu/TiO2/Pt device is explained by considering filamentary the widening phenomenon in the electrochemical metallization memory cell. Numerical calculations revealed that increase in the diameter of the metallic conducting filament from ∼3 nm to 70 nm with increasing IC is accountable for multiple low resistance states. The results of this study are promising towards realisation of emerging memory devices with low operating power, fast switching operation and high spatial density.

Configurable Resistive Response in BaTiO3 Ferroelectric Memristors via Electron Beam Radiation.

Ferroelectric oxide memristors are currently in the highlights of a thriving area of research aiming at the development of nonvolatile, adaptive memories for applications in neuromorphic computing. However, to date a precise control of synapse-like functionalities by adjusting the interplay between ferroelectric polarization and resistive switching processes is still an ongoing challenge. Here, it is shown that by means of controlled electron beam radiation, a prototypical ferroelectric film of BaTiO3 can be turned into a memristor with multiple configurable resistance states. Ex situ and in situ analyses of current/voltage characteristics upon electron beam exposure confirm the quasi-continuous variation of BaTiO3 resistance up to two orders of magnitude under the typical experimental conditions employed in electron beam patterning and characterization techniques. These results demonstrate an unprecedented effective route to locally and scalably engineering multilevel ferroelectric memristors via application of moderate electron beam radiation.

Low-Power Switching through Disorder and Carrier Localization in Bismuth-Doped Germanium Telluride Phase Change Memory Nanowires.

One of the major problems with phase change memory (PCM) is the high current density required for the crystal-amorphous transformation via a melt-quench process. However, alternative low-energy pathways of amorphization via a defect-assisted process have also been proposed. Here, a defect-assisted amorphization pathway in Bi-doped GeTe nanowires is utilized to establish that carrier localization effects can significantly decrease the energy costs of amorphization. We demonstrate a strategy of doping GeTe nanowires with bismuth to engineer carrier localization effects via Fermi level/mobility edge tuning and increased atomic disorder. Enhanced carrier localization increases the carrier-lattice coupling, and therefore, the energy supplied to carriers via electrical pulses can be more efficiently extracted by the lattice to induce the critical bond distortions required for amorphization without an intermediate melting process. RESET (crystal to amorphous transition) current densities as low as ∼0.3 MA cm-2 are achieved for 8% Bi-doped GeTe nanowires, which is nearly a 3-fold reduction compared to undoped GeTe nanowires and is significantly less than GeTe thin film devices (∼50 MA cm-2). We demonstrate good reversibility of switching in the Bi-doped GeTe nanowires and also demonstrate the existence of intermediate resistance states which can be accessed by controlled electrical pulsing. The combination of low-power switching in conjunction with multiple resistance states indicates that doping strategies in PCM nanowires are beneficial for non-volatile memory and neuromorphic computing applications.

Neuromorphic Applications of a Multivalued [SnI4{(C6H5)2SO}2] Memristor Incorporated in the Echo State Machine

The operation of an FTO/[SnI4{(C6H5)2SO}2]/Cu memristor is based on the Schottky barrier modulation due to electron trapping/detrapping at the interface states. The presented memristive bipolar device has an asymmetric current–voltage characteristic and multiple resistance states, which can be achieved by the application of impulses with different amplitudes and durations. STDP measurement performed with symmetric sawtooth voltage pulses results in the asymmetric Hebbian-like learning pattern. The incorporation of the device in a particular type of the reservoir system—a single node echo state machine—allowed observation of signal processing in a feedback-loop equipped system: classification according to the initial pulse amplitude and generation of pulse sequences of a random length.

Dielectrics and Metal Stack Engineering for Multilevel Resistive Random-Access Memory

In this work, we have fabricated and evaluated a HfO2/Al2O3 bilayer structure for a two-terminal ReRAM device to have multiple resistance states as a function of compliance current (CC). Reduced power consumption was observed when the Al2O3 buffer layer was placed between the top electrode and the HfO2 layer as compared to when it is embedded between the HfO2 layer and the bottom electrode. Gradual resistance change capability was observed with varying CC. It was demonstrated that the presence of oxygen vacancies closer to the top electrode reduces the switching energy. Decreasing the thickness of the Al2O3 buffer layer, near the bottom electrode, increases the switching power requirement. It was also observed that the switching energy requirement could be altered by modifying the deposition process of the top metal layer.

Graphene-Boron Nitride-Graphene Cross-Point Memristors with Three Stable Resistive States.

Two-dimensional (2D) material-based memristors have shown several properties that are not shown by traditional ones, such as high transparency, robust mechanical strength and flexibility, superb chemical stability, enhanced thermal heat dissipation, ultralow power consumption, coexistence of bipolar and threshold resistive switching, and ultrastable relaxation when used as electronic synapse (among others). However, several electrical performances often required in memristive applications, such as the generation of multiple stable resistive states for high-density information storage, still have never been demonstrated. Here, we present the first 2D material-based memristors that exhibit three stable and well-distinguishable resistive states. By using a multilayer hexagonal boron nitride (h-BN) stack sandwiched by multilayer graphene (G) electrodes, we fabricate 5 μm × 5 μm cross-point Au/Ti/G/h-BN/G/Au memristors that can switch between each two or three resistive states, depending on the current limitation (CL) and reset voltage used. The use of graphene electrodes plus a small cross-point structure are key elements to observe the tristate operation, which has not been observed in larger (100 μm × 100 μm) devices with an identical Au/Ti/G/h-BN/G/Au structure nor in similar small (5 μm × 5 μm) devices without graphene interfacial layers (i.e., Au/Ti/h-BN/Au). Basically, we generate an intermediate state between the high resistive state and the low resistive state (LRS), named soft-LRS (S-LRS), which may be related to the formation of a narrower conductive nanofilament across the h-BN because of the ability of graphene to limit metal penetration (at low CLs). All the 2D materials have been fabricated using the scalable chemical vapor deposition approach, which is an immediate advantage compared to other works using mechanical exfoliated 2D materials.

Phase-Field Simulation of Resistive Switching Behavior in Bipolar Metal-Oxide Based Resistive Random Access Memory

Metal-oxide based Resistive Random-Access-Memory (RRAM) exhibits multiple resistance states, due to the growth and retraction of a conductive filament between the electrodes of the device. The performance of RRAM relies on the fast switching speed, large current on/off ratio, and longer endurance reliability, which are critically dependent on the nonlinear resistive switching under electrical bias, temperature, interfacial chemical reaction and mechanical stress effects. In this work we developed a comprehensive phase-field model to simulate the resistive switching. The developed model enables us to separately study the aforementioned effects and understand their relative roles in the switching process. By choosing TaOx as an example, we find that voltage-driven ion transport is responsible for the resistance change between low and high states. It also induces a local stress that influences the endurance of the device. Our work helps shed light on understanding the microstructure dependent ion transport behaviors in RRAMs.

Synaptic element for neuromorphic computing using a magnetic domain wall device with synthetic pinning sites

The ability to make devices that mimic the human brain has been a subject of great interest in scientific research in recent years. Current artificial intelligence algorithms are primarily executed on the von Neumann hardware. This causes a bottleneck in processing speeds and is not energy efficient. In this work, we have demonstrated a synaptic element based on a magnetic domain wall device. The domain wall motion was controlled with the use of synthetic pinning sites, which were introduced by boron (B+) ion-implantation for local modification of the magnetic properties. The magnetization switching process of a Co/Pd multilayer structure with perpendicular magnetic anisotropy was observed by using MagVision Kerr microscopy system. The B+ implantation depth was controlled by varying the thickness of a Ta overcoat layer. The Kerr microscopy results correlate with the electrical measurements of the wire which show multiple resistive states. The control of the domain wall motion with the synthetic pinning sites is demonstrated to be a reliable technique for neuromorphic applications.

Resistance switching characteristics and mechanisms of MXene/SiO2 structure-based memristor

Silicon dioxide memristors possess multiple resistance states and can be used as a key component of memory devices and neuromorphic systems. However, their conductive mechanisms are incompletely understood, and their resistance switching (RS) variability is a major challenge for commercialization of memristors. In this work, by combining the desirable properties of silicon dioxide with those of a two-dimensional MXene material (Ti3C2), a memristor based on an MXene/SiO2 structure is fabricated. The Cu/MXene/SiO2/W memristive devices exhibit excellent switching performance compared with traditional Cu/SiO2/W devices under the same conditions. Furthermore, the role of the MXene/SiO2 structure in the SiO2-based memristors is revealed by the physical characterization of the MXene and first-principles calculation of the MXene/SiO2 structure. The results indicate that the conductive filaments (CFs) are more likely to grow along the locations of MXene nanostructures, which reduces the randomness of CFs in the Cu/MXene/SiO2/W memristors and further improves the device performance. Meanwhile, the MXene/SiO2 structure appears to greatly reduce the mobility of Cu ions in the entire RS region, as well as improve the performance of the SiO2-based memristors while maintaining the operating voltages low.

Current assisted memory effect in superconductor–ferromagnet bilayers: a potential candidate for memristors

AbstrsctSuperconductivity and ferromagnetism are two very useful phenomena, hoewever they rarely coexist in bulk materials. Bringing them together in an artificial hybrid bilayer produces some unusual results. We designed and studied a system of superconductor-ferromagnet bilayer with a thin insulating buffer layer between them. Such a superconductor-ferromagnet bilayer with magnetostatic coupling is proposed for use as a multibit superconductor memory device and a potential candidate as a memristor. Numerical simulations were performed by using Ginzburg–Landau and Landau–Lifshitz-Gilbert models for superconductor and ferromagnet materials, which highlighted some interesting resistive memory effects in the superconducting layer in the bilayer system. A vortex pattern in the superconductor was observed to be strongly coupled with the ferromagnet domain structure, while their dynamics were controlled by the current flowing through the superconductor. Carrier concentration, energy components and magnetization in the superconducting layer were studied as a function of applied current pulses in the superconductor layer, indicating the information storage of the current pulses. Multiple resistive states were observed, pointing towards the possibility that such a device could be used as a multibit data storage device.