Resistive Switching Characteristics Research Articles

Non-Volatile Memory and Synaptic Characteristics of TiN/CeOx/Pt RRAM Devices.

In this study, we investigate the synaptic characteristics and the non-volatile memory characteristics of TiN/CeOx/Pt RRAM devices for a neuromorphic system. The thickness and chemical properties of the CeOx are confirmed through TEM, EDS, and XPS analysis. A lot of oxygen vacancies (ions) in CeOx film enhance resistive switching. The stable bipolar resistive switching characteristics, endurance cycling (>100 cycles), and non-volatile properties in the retention test (>10,000 s) are assessed through DC sweep. The filamentary switching model and Schottky emission-based conduction model are presented for TiN/CeOx/Pt RRAM devices in the LRS and HRS. The compliance current (1~5 mA) and reset stop voltage (−1.3~−2.2 V) are used in the set and reset processes, respectively, to implement multi-level cell (MLC) in DC sweep mode. Based on neural activity, a neuromorphic system is performed by electrical stimulation. Accordingly, the pulse responses achieve longer endurance cycling (>10,000 cycles), MLC (potentiation and depression), spike-timing dependent plasticity (STDP), and excitatory postsynaptic current (EPSC) to mimic synapse using TiN/CeOx/Pt RRAM devices.

Effects of Bottom Electrode Materials on the Resistive Switching Characteristics of HfO2-Based RRAM Devices

Effects of Bottom Electrode Materials on the Resistive Switching Characteristics of HfO2-Based RRAM Devices

Ferroelectric memristor based on Li-doped BiFeO3 for information processing

As a nanoscale semiconductor memory device, a ferroelectric memristor has promising prospects to break through the von Neumann framework in terms of artificial synaptic function, information processing, and integration. This study presents the fabrication of Li0.09Bi0.91FeO3 as the functional layer for a memristor device based on the Si substrate, enabling the integration of silicon complementary metal oxide semiconductor technology. In addition, it exhibits bipolar resistance switching characteristics in a direct current mode and can rapidly achieve stable conductance tunability at higher frequencies through the applied pulse for biosynapse simulation. More importantly, multiple devices are connected into electrical circuits to realize storage functions with information processing and programmable characteristics. This work paves the way for near-future applications of ferroelectric memristors in information processing.

Ultrahigh tunability of resistive switching in strongly correlated functional oxide

Vanadium dioxide (VO2) is a strongly correlated material that undergoes reversible resistive switching. However, the underlying mechanism of such resistive switching is still under debate. In this work we report an ultrahigh decrease in the VO2 switching threshold-voltage by more than 35 V and a giant 680% change by modifying the material stoichiometry through tungsten (W) doping. This remarkable effect of chemical doping on the resistive switching characteristics of VO2 illuminates on the fundamental mechanism of the insulator-to-metal transition, which reveals that such abrupt resistive switching is attributed to the electro-thermal actuation process. Furthermore, we report ultralow voltage spontaneous electrical oscillation in W-doped VO2 for the first time. Advances achieved here through the design of a new class of transition metal oxide based strongly correlated materials are of paramount importance for emerging electronics.

MoTe2-based low energy consumption artificial synapse for neuromorphic behavior and decimal arithmetic

Two-dimensional material-based memristors have shown attractive research prospects as brain-like devices for neuromorphic computing. Among them, transition metal dichalcogenides–based memristors have proved to be one of the most promising competitors. In this work, a two-dimensional memristor based on MoTe2 nanosheets was fabricated and demonstrated. The experimental results illustrate that the two-terminal synaptic based on the Ag/MoTe2/ITO structure exhibits stable bipolar and non-volatile resistive switching characteristics attributed to the controllable formation and rupturing of silver conductive filaments. The device can be successively modulated by a pulse train with a minimum pulse width of 40 ns. More interestingly, the energy consumption of the device to complete one write event is only 74.2 pJ. In addition, biological synaptic behaviors, such as excitatory postsynaptic current gain properties, long-term potentiation (LTP)/long-term depression, spike-timing-dependent- plasticity, short-term plasticity, long-term potentiation (LTP), paired-pulse facilitation, post-tetanic potentiation, and learning-experimental behaviors were mimicked faithfully. Finally, the decimal arithmetic application was introduced to the device, and it is confirmed that addition and multiplication functions can be performed. Therefore, the artificial synapse based on MoTe2 nanosheets not only exhibits the stable non-volatile resistive switching behavior but also facilitates the development of low-energy consumption neuromorphic computing chips based on transition metal dichalcogenides.

Pseudo-flexible resistive switching characteristics of nano-bowl-like NiO arrays on mica substrates

Nano-bowl-like NiO (nb-NiO) arrays on flexible Pt-coated mica substrates have been fabricated by electro-deposition and magnetron sputtering with the assistance of the monolayer colloidal sphere template. The nano-bowl array makes the polycrystalline NiO film thickness change periodically in the range of 150–450 nm, and the HfO2 buffer layer effectively depresses the local leakage current. A voltage which is lower than ±0.4 V can make the Ag/HfO2/nb-NiO/Pt cell reversibly switch between the high and the low resistance states within 11 μs. This cell exhibits a ratio of the high to the low resistance (RH/RL) of 103 ∼ 104, a data retention time over 104 s and an endurance of >1200 cycles. Driven by the non-uniform electric field, oxygen ions accumulated at the HfO2/NiO interface prefer to occur the Fowler-Nordheim tunneling at the thinnest bowl bottom, which decreases switching voltages. The bent cell almost keeps initial resistive switching properties before the bending radius (r) is below 2.5 cm. Though its properties greatly degrade when the r further reduces below 1.5 cm, it can recover most of initial memory window after returning to the flat state again, showing potential applications in new types of pseudo-flexible memory devices.

Variability-Controlled HfZrO2 Ferroelectric Tunnel Junctions for Reservoir Computing

We propose reservoir computing (RC), a framework for constructing recurrent neural networks (RNNs) with simple training rule, using ferroelectric tunnel junction (FTJ) crossbar array. Not only the weights of connection between reservoir layer and output layer need to be trained, but also random fixed weights within reservoir layer can be realized by the conductance of FTJs. Random weights within reservoir layers have optimal variability, which is dependent not on the type of the distribution but on the reservoir size. Optimal variability can be attained by device-to-device variability in FTJs together with analog converters on the wordlines and the bitlines. Analog resistive switching characteristics in FTJs can also realize the weights connected to output layer, which need to be trained. Furthermore, device-to-device variability in Hf <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$_{{1}- {x}}$ </tex-math></inline-formula> ZrxO2 (HZO) FTJ is extensively investigated. Taking into account the low current operation in FTJs, RC with FTJ crossbar array is a potential candidate for neuromorphic computing system with low power consumption.

Fast Fitting of the Dynamic Memdiode Model to the Conduction Characteristics of RRAM Devices Using Convolutional Neural Networks.

In this paper, the use of Artificial Neural Networks (ANNs) in the form of Convolutional Neural Networks (AlexNET) for the fast and energy-efficient fitting of the Dynamic Memdiode Model (DMM) to the conduction characteristics of bipolar-type resistive switching (RS) devices is investigated. Despite an initial computationally intensive training phase the ANNs allow obtaining a mapping between the experimental Current-Voltage (I-V) curve and the corresponding DMM parameters without incurring a costly iterative process as typically considered in error minimization-based optimization algorithms. In order to demonstrate the fitting capabilities of the proposed approach, a complete set of I-Vs obtained from Y2O3-based RRAM devices, fabricated with different oxidation conditions and measured with different current compliances, is considered. In this way, in addition to the intrinsic RS variability, extrinsic variation is achieved by means of external factors (oxygen content and damage control during the set process). We show that the reported method provides a significant reduction of the fitting time (one order of magnitude), especially in the case of large data sets. This issue is crucial when the extraction of the model parameters and their statistical characterization are required.

Interface barrier-induced conversion of resistive switching mechanism in Mn-doped BiFeO3 memristor

Different from conductive filament (CF)-type counterparts, interface-type devices exhibit continuously gradual conductance changes, making them the potential for artificial synapses. In this paper, Mn-doped BiFeO3 (BFMO) devices with SrRuO3 and TiN bottom electrodes demonstrate the clear CF rather than the interface barrier type resistance-switching feature due to the high Schottky barrier. Considering the measured electron affinity of 3.52 eV and work function of 4.22 eV in the as-synthesized BFMO film (a weak n-type semiconductor, marked as n−), we fabricated a hetero-junction device with the Nb-doped SrTiO3 (NSTO) bottom electrode (a strong n-type semiconductor, marked as n+) exhibiting analog switch characteristics. The n−–n+ hetero-junction between BFMO and NSTO reverses the operation polarity and leads to a barrier transition-dominated conductive behavior in the BFMO-based memristor. The device shows a large ON/OFF ratio over 1200, favorable stability after 104 s, continual multi-value characteristics, symmetrical long-term potentiation and depression, and synaptic plasticity with about 80 ns time constant. The investigation of resistive switching features, band structure, and synapse performance in this work provides a reference for the application of BiFeO3 in the field of the memristor.

Remotely tuned multistate resistive switching in MoS2/NiMnIn thin film heterostructure for highly flexible ReRAM application

In this article, the Cu/MoS2/NiMnIn memory structure was fabricated over stainless steel substrate for flexible electronics applications. The bipolar resistive switching characteristics were observed in two different fabricated devices (i) without sulfur vacancies (D1) and (ii) with sulfur vacancies (D2) in MoS2 thin film. Three different resistance states such as the high resistance state (HRS), intermediate resistance state (IRS) and low resistance state (LRS) have been detected for the D2 memory device. It could be ascribed to the two filamentary models based on the movement of sulfur vacancies and the formation of Cu metallic filament with applied bias voltage. Moreover, the Ohmic and modified space charge limited conduction mechanisms clearly explain the current conduction in different resistance states of device D2. The fabricated MoS2 thin film-based memory structure exhibits stable resistive switching behavior with a high OFF/ON ratio of ∼3.6 × 103, good consistency of ∼3600 endurance cycles and excellent data retention capability up to 3000 s. Moreover, the remote tuning of device D2 was thoroughly investigated and an appreciable change in SET voltage was detected with external temperature and magnetic field. Additionally, the external magnetic field altered the switching states and enhanced the multi-bit data storage capability of the memory device. This can be ascribed to the effect of the Lorentz force on ionic movement in the presence of the external magnetic field. The mechanical flexibility of the memory structure was tested for 1000 bending cycles at various bending angles in both the tensile and compressive bending modes. Hence, the present study opens up new ways for the futuristic flexible device for high data storage and neuromorphic computing applications.

CMOS-Compatible Memristor for Optoelectronic Neuromorphic Computing

Optoelectronic memristor is a promising candidate for future light-controllable high-density storage and neuromorphic computing. In this work, light-tunable resistive switching (RS) characteristics are demonstrated in the CMOS process-compatible ITO/HfO2/TiO2/ITO optoelectronic memristor. The device shows an average of 79.24% transmittance under visible light. After electroforming, stable bipolar analog switching, data retention beyond 104 s, and endurance of 106 cycles are realized. An obvious current increase is observed under 405 nm wavelength light irradiation both in high and in low resistance states. The long-term potentiation of synaptic property can be achieved by both electrical and optical stimulation. Moreover, based on the optical potentiation and electrical depression of conductances, the simulated Hopfield neural network (HNN) is trained for learning the 10 × 10 pixels size image. The HNN can be successfully trained to recognize the input image with a training accuracy of 100% in 13 iterations. These results suggest that this optoelectronic memristor has a high potential for neuromorphic application.

Understanding the Origin of Unreliable Low-Resistance State During Initialization Process

The main factor hindering the massive employment for resistive random access memory (RRAM) can be ascribed to reliability issues. It has been reported that resistive switching (RS) characteristics were unstable during initial <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${I}-{V}$ </tex-math></inline-formula> cycles; however, the reason has not been studied yet. In this article, the mechanism for a significant increase in a low-resistance state (LRS) value during the initial cycles has been investigated. In this article, we suggest that the initialization process is actually the formation process of a stable TaOx interface layer between the RS layer and the top electrode (TE). Furthermore, a co-regulation model of the filament-type and the interface-type switching mechanisms on RS behavior is proposed, which explains the complete working process of devices. By reducing the size of the device and increasing the positive stop voltage, the initialization process of the device has been greatly shortened. This article may promote the development of advanced and reliable RRAM devices.

Enhanced Resistive Switching and Synaptic Characteristics of ALD Deposited AlN-Based RRAM by Positive Soft Breakdown Process.

Nitride film played an essential role as an excellent diffusion barrier in the semiconductor field for several decades. In addition, interest in next-generation memories induced researchers' attention to nitride film as a new storage medium. A Pt/AlN/TaN device was investigated for resistive random-access memory (RRAM) application in this work. Resistive switching properties were examined in the AlN thin film formed by atomic layer deposition (ALD). The unique switching feature conducted under the positive voltage was investigated, while the typical bipolar switching was conducted under the application of negative voltage. Good retention and DC, and pulse endurances were achieved in both conditions and compared to the memory performances. Finally, the electronic behaviors based on the unique switching feature were analyzed through X-ray photoelectron spectroscopy (XPS) and the current-voltage (I-V) linear fitting model.

Effect of oxygen flow rate on long-term and short-term Schottky barrier modulations in Pd/IGZO/SiO2/p+-Si memristors

In this work, we investigate the long- and short-term Schottky barrier modulations in a Pd/IGZO/SiO2/p+-Si memristor under oxygen flow rate (OFR) control. The thickness of the SiO2 layer verified using transmission electron microscopy (TEM) was found to affect resistive-switching characteristics such as the on/off ratio. A high barrier was observed corresponding to a high OFR in the thermionic emission model. In addition, we present an energy band diagram considering traps and oxygen vacancies for oxygen-rich (O-rich) and oxygen-poor (O-poor) devices. Moreover, physical parameters such as a change in barrier height, activation energy, and potentiation/depression were experimentally extracted through various pulse schemes. Experimental results reveal that an O-rich device has high accuracy with a neural network on the MNIST dataset. In this study, we optimized the resistive-switching characteristics by varying OFR as an experimental factor. Further, we provide guidelines for designing hardware-based neuromorphic systems by separating short- and long-term components.

Resistive Switching Characteristics of Alloyed AlSiOx Insulator for Neuromorphic Devices

Charge-based memories, such as NAND flash and dynamic random-access memory (DRAM), have reached scaling limits and various next-generation memories are being studied to overcome their issues. Resistive random-access memory (RRAM) has advantages in structural scalability and long retention characteristics, and thus has been studied as a next-generation memory application and neuromorphic system area. In this paper, AlSiOx, which was used as an alloyed insulator, was used to secure stable switching. We demonstrate synaptic characteristics, as well as the basic resistive switching characteristics with multi-level cells (MLC) by applying the DC sweep and pulses. Conduction mechanism analysis for resistive switching characteristics was conducted to understand the resistive switching properties of the device. MLC, retention, and endurance are evaluated and potentiation/depression curves are mimicked for a neuromorphic device.

Resistive switching of two-dimensional NiAl-layered double hydroxides and memory logical functions

Two-dimensional NiAl layered double hydroxides (LDHs) were synthesized by urea hydrolysis method. Considering its unique layered structure and high-speed carrier transfer channel, NiAl-LDHs were applied to the active layer of memristor through spin coating process. It is found that the Al/NiAl-LDHs/Al memristor exhibits unipolar resistance switching characteristics, which resistive switching mechanism is attributed to formation and fusing of conductive filaments controlled by oxygen vacancies. Based on this Al/NiAl-LDHs/Al memristor, a D-type latch was constructed. A memtransistor with Si/SiO2/graphene oxide (GO)/Al/NiAl-LDHs/Al structure was built by connecting Al/NiAl-LDHs/Al memristor in series on the drain electrode of GO transistor. The memtransistor exhibits excellent gate controllability and resistive switching characteristics. Combined with its characteristics, we used the memtransistor to build a D-trigger and verify its memory logical function. The result shows the great potential of layered bimetallic hydroxide materials in the field of resistance switching memory and memory computing.

Emergence of electric field-induced conducting states in single-crystalline MoTe[formula omitted] nanoflakes and its application in memristive devices

Conductive atomic force microscopy (cAFM) reveals the emergence of conducting states in MoTe2 NFs under applied electric field. In this report, we explore the use of electric field induced conducting states in the single-crystalline MoTe2 nanoflakes (NFs) for nonvolatile resistive memory technologies. The memristive devices are fabricated in vertical metal–insulator (active layer)-metal architecture with aluminum top and fluorine doped tin oxide bottom electrodes. The MoTe2 NFs were embedded in polymethyl methacrylate (PMMA) to form the active layer. The device shows appreciable resistive switching (RS) characteristics while switching from a high resistive state (HRS) to a low resistive state (LRS) and vice versa. The low-frequency conductance noise (LFCN) measurements show that the power spectral density of conductance fluctuation follows the 1/f2 nature in the HRS, while 1/f in LRS. The stochastic analysis of conductance fluctuation data exhibits random telegraphic noise in the HRS, indicating the role of NFs as charge trapping sites. The LFCN measurements and cAFM study confirm that electric field-induced conducting states are pivotal in enhancing the RS performance of the memory device.

Resistance switching characteristics of Ag/ZnO/graphene resistive random access memory

In this work, an Ag/ZnO/Graphene structure resistive random access memory (RRAM) is prepared through RF magnetron sputtering. Compared to the Ag/ZnO/Au structure device, the endurance of the device using a graphene electrode up to 3 × 103 cycles. It has a self-current compliance function of 2 × 10−4 A and 30 times switching resistance ratio and characteristics without electroforming. The density of states (DOS) of ZnO supercells with different space groups are calculated by using the first-principles, and it is analyzed that the contributions of different supercells to the conductivity and contact type of the RRAM. Combined with the I–V curve of the RRAM, it illustrated that the conduction path is created by the space-charge-limited conduction (SCLC) of electrons captured by oxygen vacancy traps. The biocompatibility of the device is tested by a cell toxicity test. It proves that the Ag/ZnO/Graphene RRAM has high biocompatibility and broad applicability in implantable biomedical devices.

Flexible Threshold-Type Switching Devices with Low Threshold and High Stability Based on Silkworm Hemolymph.

In this paper, a floating-gate flexible nonvolatile memory is reported that is composed of natural biological materials, namely, silkworm hemolymph, graphene quantum dots as the floating-gate layer, and polymethyl methacrylate (PMMA) as the insulating layer. The device has a high ON/OFF current ratio (4.76 × 106), a low setting voltage (<−1.75 V), and good durability and retention ability. The device has two storage characteristics, namely, Flash and WORM, which can be effectively and accurately controlled by adjusting the limiting current during device setting. The resistance switching characteristics are the result of the formation and fracture of conductive filaments. The floating-gate flexible bioresistive random access memory prepared in this paper provides a new idea for the development of multifunctional and biocompatible flexible memory.

Protons: Critical Species for Resistive Switching in Interface‐Type Memristors

AbstractInterface‐type (IT) resistive switching (RS) memories are promising for next generation memory and computing technologies owing to the filament‐free switching, high on/off ratio, low power consumption, and low spatial variability. Although the switching mechanisms of memristors have been widely studied in filament‐type devices, they are largely unknown in IT memristors. In this work, using the simple Au/Nb:SrTiO3 (Nb:STO) as a model Schottky system, it is identified that protons from moisture are key element in determining the RS characteristics in IT memristors. The Au/Nb:STO devices show typical Schottky interface controlled current–voltage (I–V) curves with a large on/off ratio under ambient conditions. Surprisingly, in a controlled environment without protons/moisture, the large I–V hysteresis collapses with the disappearance of a high resistance state (HRS) and the Schottky barrier. Once the devices are re‐exposed to a humid environment, the typical large I–V hysteresis can be recovered within hours as the HRS and Schottky interface are restored. The RS mechanism in Au/Nb:STO is attributed to the Schottky barrier modulation by a proton assisted electron trapping and detrapping process. This work highlights the important role of protons/moisture in the RS properties of IT memristors and provides fundamental insight for switching mechanisms in metal oxides‐based memory devices.