Resistive Switching Memory Research Articles

Low Conductance State Drift Characterization and Mitigation in Resistive Switching Memories (RRAM) for Artificial Neural Networks

The crossbar structure of Resistive-switching random access memory (RRAM) arrays enabled the In-Memory Computing circuits paradigm, since they imply the native acceleration of a crucial operations in this scenario, namely the Matrix-Vector-Multiplication (MVM). However, RRAM arrays are affected by several issues materializing in conductance variations that might cause severe performance degradation. A critical one is related to the drift of the low conductance states appearing immediately at the end of program and verify algorithms that are mandatory for an accurate multi-level conductance operation. In this work, we analyze the benefits of a new programming algorithm that embodies Set and Reset switching operations to achieve better conductance control and lower variability. Data retention analysis performed with different temperatures for 168 hours evidence its superior performance with respect to standard programming approach. Finally, we explored the benefits of using our methodology at a higher abstraction level, through the simulation of an Artificial Neural Network for image recognition task (MNIST dataset). The accuracy achieved shows higher performance stability over temperature and time.

Electrical conduction and resistive switching in cow milk-based devices prepared using the spin-coat method

Electronic devices made of organic materials are critical for sustainable technology and the reduction of electronic waste. In this paper, we report on the resistive switching (RS) behavior of an organic full cream cow milk-based film placed between two electrodes. The switching properties of the milk film were investigated for use in RS memory (ReRAM), a new type of memory device. Two devices were created: one with electrodes made of silver (Ag) and indium doped tin oxide (ITO) (Ag/cow milk/ITO) and the other with electrodes made of Ag and tungsten (W) (Ag/cow milk/W). The study is unique in two ways: (1) the use of the spin coat method, which allowed us to deposit 15 nm thin milk films, and (2) the use of ITO and W bottom electrodes, which allowed us to compare the results. Our findings show that both ITO- and W-based devices exhibit RS properties with “S-type” bipolar behavior. The ITO-based device exhibited switching at 0.77 V and an ION/IOFF ratio of ∼2. The memory behavior of this device lasted ∼7 write/erase cycles, indicating the possibility of nonvolatile memory application. In contrast, the switching in the W-based device was characterized by low (0.1 V) voltage, and it lasted for only one cycle. Increasing the compliance current to 0.5 A improved the ON/OFF ratio to 102, but the device could not repeat the same hysteresis behavior. The analysis of the conduction and switching mechanisms in both devices revealed a conductive bridge mechanism attributed to the metalization of metal ions naturally found in milk. Overall, our findings show that the thickness of the milk film influences both conduction and switching mechanisms and that the (milk film)/ITO and (milk film)/W interfaces influence the conductance of full cream cow milk-based ReRAM devices.

Atomic Layer Deposited SiOX-Based Resistive Switching Memory for Multi-Level Cell Storage

Herein, stable resistive switching characteristics are demonstrated in an atomic-layer-deposited SiOX-based resistive memory device. The thickness and chemical properties of the Pt/SiOX/TaN stack are verified by transmission electron microscopy (TEM) and X-ray photoemission spectroscopy (XPS). It is demonstrated that much better resistive switching is obtained using a negative set and positive reset compared to the opposite polarity. In addition, multi-level switching is demonstrated by changing the compliance current (CC) and the reset stop voltage, and potentiation and depression are emulated by applying pulses to achieve a synaptic device. Finally, a pulse endurance of 10,000 cycles and a retention time of 5000 s are confirmed by modulating the pulse input and reading voltage, respectively.

Effects of drying temperature on preparation of pectin polysaccharide thin film for resistive switching memory

Effects of drying temperature on preparation of pectin polysaccharide thin film for resistive switching memory

An energy-efficient in-memory computing architecture for survival data analysis based on resistive switching memories.

One of the objectives fostered in medical science is the so-called precision medicine, which requires the analysis of a large amount of survival data from patients to deeply understand treatment options. Tools like machine learning (ML) and deep neural networks are becoming a de-facto standard. Nowadays, computing facilities based on the Von Neumann architecture are devoted to these tasks, yet rapidly hitting a bottleneck in performance and energy efficiency. The in-memory computing (IMC) architecture emerged as a revolutionary approach to overcome that issue. In this work, we propose an IMC architecture based on resistive switching memory (RRAM) crossbar arrays to provide a convenient primitive for matrix-vector multiplication in a single computational step. This opens massive performance improvement in the acceleration of a neural network that is frequently used in survival analysis of biomedical records, namely the DeepSurv. We explored how the synaptic weights mapping strategy and the programming algorithms developed to counter RRAM non-idealities expose a performance/energy trade-off. Finally, we discussed how this application is tailored for the IMC architecture rather than being executed on commodity systems.

Preparation and incorporation of NiSe@MoSe2 nano arrays in PVA matrix for resistive switching memory

In the present work, we have synthesized Nickel selenide@Molybdenum diselenide (NiSe@MoSe2) heteronanostructured arrays by a one-step hydrothermal method. The structural and optical analysis of nanopowder is done through Field emission scanning electron microscopy (FESEM), elemental mapping, Energy-dispersive X-ray spectroscopy (EDX), X-ray diffraction (XRD), UV–vis, and Photoluminescence (PL) spectroscopy. XRD reveals the formation of a hexagonal structured system with a crystallite size of 166 nm. The bandgap of NiSe@MoSe2 nanosheets is found to be 1.71 eV, calculated from UV–vis spectra. FESEM showed the homogenous growth of NiSe@MoSe2 nanosheets while electronic transitions are investigated through PL. Further, the polymer nanocomposites of NiSe@MoSe2/PVA are synthesized by the ex-situ method at six different NiSe@MoSe2 concentrations in the PVA matrix. XRD shows good interaction between NiSe@MoSe2 and PVA polymer. The crystallinity increases as concentrations of NiSe@MoSe2 in PVA increases. FESEM shows the uniform distribution of NiSe@MoSe2 nanosheets in the PVA matrix. The two-terminal Ag/NiSe@MoSe2-PVA/FTO devices have been fabricated and characterized electrically by I-V measurements. I-V shows hysteresis behavior and also, bipolar switching is observed with a maximum Ion/Ioff ratio of ~103 at 2 wt. % NiSe@MoSe2 concentration in the PVA matrix. The good stability and reliability are checked through endurance and write-read-erase-read cyclic sequences. The obtained results open up the possibility of a next-generation non-volatile resistive switching memory device.

Demonstration of enhanced resistance switching performance of HfO2/WOx-based bilayer devices embedded with Ti nano island array by applying a rapid thermal annealing process

In this work, a novel set of methods is prοposed for tuning the oxygen vacancy distribution through the combination of several effects. More specifically, the incorporation of a Ti nano-island (NI), as well as the insertion of a thin layer of HfO 2 with 10 nm thickness and the subsequent enforcement of an annealing step at 400 ℃ are employed to improve the performance of the resistive switching memory devices. The acquired results indicate that the embedded Ti NI array can significantly reduce the switching voltage and the statistical dispersion of the switching characteristics, whereas the increase of both the high resistance state (HRS) and low resistance state (LRS) resistance levels and the reduction of the operating current values are attributed to the existence of the HfO 2 layer. Subsequently, the annealing process under air conditions can effectively reduce the oxygen vacancy content within the device and further improve the R HRS and on/off ratio. The reduction of the oxygen vacancy concentration is caused by the diffusion of oxygen ions in the air to the dielectric layer during annealing. Hence, it is concluded that the simultaneous incorporation of a Ti NI, a functional layer and a rapid thermal annealing step are regarded as a novel promising and practical technology for significantly improving the whole performance of the memristive elements. • Ti NI can significantly improve the whole performance of the memristive elements. • A rapid thermal annealing can improve the performance of the memristive elements. • The annealing can effectively reduce the oxygen vacancy concentration within the device.

Active layer nitrogen doping technique with excellent thermal stability for resistive switching memristor

In this study, we propose a thermally stable memristor with nitrogen-doped hafnium oxide (HfO:N)-based resistive switching (RS) memory. The memristor with HfO:N as an active layer showed only a 7% change in the resistance in the high resistance state (HRS) after post-metal annealing (PMA) at 400 °C for 1 h. In contrast, the HfO2-based memristor exhibited an 83% change in the resistance at HRS after PMA at 400 °C for 1 h and lost RS operating characteristics after PMA over 400 °C. In addition, although the resistance of the HRS decreased by 80% after PMA at 550 °C for 1 h, the HfO:N-based memristor showed that the RS operation was maintained up to 550 °C. Through the nitrogen doping technique, a thermal budget of 550 °C can be achieved, which is one of the highest thermal budgets in RS memory with PMA. Such thermal stability enhancement of the memristor is a result of nitrogen doping, which improves the structural stability of the active layer and suppresses the generation of oxygen vacancies in the active layer. This experimental approach can facilitate the development of advanced memristor devices with a good thermal budget of up to 550 °C.

Multifunctionality in ferromagnetic shape memory alloy-based resistive switching memory for flexible ReRAM application

Multifunctional flexible electronics is the ongoing demand for fabricating wearable data storage and communication devices. The magnetoelectric (ME) heterostructure consisting of piezoelectric (AlN) and ferromagnetic magnetic shape memory alloy [FSMA (Ni–Mn–In)] was fabricated over stainless steel (SS) substrate for resistive random access memory application. The Cu/AlN/FSMA/SS metal–insulator–metal based memory cell displays bipolar resistive switching (RS) behavior. The formation of Cu metallic filament at a particular SET voltage leads the memory cell in a low resistance state (LRS) from its pristine high resistance state (HRS). The LRS and HRS are explained well by Ohmic and space charge limited conduction mechanisms, respectively. The fabricated memory cell displays excellent endurance and data retention capability with a high OFF/ON ratio of ∼1.2 × 103. Furthermore, the multifunctionality of the ME heterostructure-based RAM was investigated by tuning the SET voltage with ambiance temperature and external magnetic field remotely. A significant change in the SET voltage could be ascribed to the temperature and magnetic field-induced strain transferred to the AlN piezoelectric layer from the magnetostrictive FSMA (Ni–Mn–In) bottom electrode. The residual Lorentz force explains the remotely tuned LRS and HRS in the transverse magnetic field for multi-bit data storage applications. Moreover, the RS characteristics remain stable even after 800 bending cycles as well as with bending angle (0°–180°). Hence, the present ME heterostructure integrated with flexible SS substrate can be a better choice for highly flexible, low-cost, and multifunctional futuristic RAM applications.

High On/Off Ratio Organic Resistive Switching Memory Based on Carbazolyl Dicyanobenzene and a Polymer Composite

High On/Off Ratio Organic Resistive Switching Memory Based on Carbazolyl Dicyanobenzene and a Polymer Composite

Aliovalent Ta-Doping-Engineered Oxygen Vacancy Configurations for Ultralow-Voltage Resistive Memory Devices: A DFT-Supported Experimental Study.

Alteration of transport properties of any material, especially metal oxides, by doping suitable impurities is not straightforward as it may introduce multiple defects like oxygen vacancies (Vo) in the system. It plays a decisive role in controlling the resistive switching (RS) performance of metal oxide-based memory devices. Therefore, a judicious choice of dopants and their atomic concentrations is crucial for achieving an optimum Vo configuration. Here, we show that the rational designing of RS memory devices with cationic dopants (Ta), in particular, Au/Ti1-xTaxO2-δ/Pt devices, is promising for the upcoming non-volatile memory technology. Indeed, a current window of ∼104 is realized at an ultralow voltage as low as 0.25 V with significant retention (∼104 s) and endurance (∼105 cycles) of the device by considering 1.11 at % Ta doping. The obtained device parameters are compared with those in the available literature to establish its excellent performance. Furthermore, using detailed experimental analyses and density functional theory (DFT)-based first-principles calculations, we comprehend that the meticulous presence of Vo configurations and the columnar-like dendritic structures is crucial for achieving ultralow-voltage bipolar RS characteristics. In fact, the dopant-mediated Vo interactions are found to be responsible for the enhancement in local current conduction, as evidenced from the DFT-simulated electron localization function plots, and these, in turn, augment the device performance. Overall, the present study on cationic-dopant-controlled defect engineering could pave a neoteric direction for future energy-efficient oxide memristors.

Tuning oxygen vacancies and resistive switching behaviors in amorphous Y2O3 film-based memories

Yttrium oxide (Y2O3) has attracted attentions as the new and promising functional material in resistive switching (RS) memories. However, the oxygen vacancies tuning by various electrodes is lack of understanding for Y2O3-based memories, which plays a crucial role for RS behaviors. Here, non-crystalline Y2O3 films prepared through magnetron sputtering deposition are used to construct RS memory devices with different structure. All the devices exhibit nonvolatile bipolar RS behaviors and the switching phenomena are depended on the formation/rupture of conductive filaments (CFs), which are composed of oxygen vacancies. In special, the Y2O3-based memories with Pt/Y2O3/Pt structure demonstrate competitive switching properties, including the ultra-lower set/reset voltages and high retention characteristic. Furthermore, the influence of different electrodes on oxygen vacancies, as well as morphologies of CFs, are discussed in detail. Besides that, physical models are proposed to further clarify carrier transport mechanisms and switching behaviors for memories with different structure. This study provides an in-depth understanding of the structural design and oxygen vacancies tunning by selecting electrodes in amorphous Y2O3 film-based RS memories.

Interfacial Resistive Switching by Multiphase Polarization in Ion-Intercalation Nanofilms.

Nonvolatile resistive-switching (RS) memories promise to revolutionize hardware architectures with in-memory computing. Recently, ion-interclation materials have attracted increasing attention as potential RS materials for their ion-modulated electronic conductivity. In this Letter, we propose RS by multiphase polarization (MP) of ion-intercalated thin films between ion-blocking electrodes, in which interfacial phase separation triggered by an applied voltage switches the electron-transfer resistance. We develop an electrochemical phase-field model for simulations of coupled ion-electron transport and ion-modulated electron-transfer rates and use it to analyze the MP switching current and time, resistance ratio, and current-voltage response. The model is able to reproduce the complex cyclic voltammograms of lithium titanate (LTO) memristors, which cannot be explained by existing models based on bulk dielectric breakdown. The theory predicts the achievable switching speeds for multiphase ion-intercalation materials and could be used to guide the design of high-performance MP-based RS memories.

Top electrode modulated W/Ag/MgO/Au resistive random access memory for improved electronic synapse performance

Resistive random access memory (ReRAM) is touted to replace silicon-based flash memory due to its low operating voltage, fast access speeds, and the potential to scale down to nm range for ultra-high density storage. In addition, its ability to retain multi-level resistance states makes it suitable for neuromorphic computing application. Here, we develop a cationic ReRAM with a sputtered MgO as the insulating layer. The resistive switching properties of the Ag/MgO/Au ReRAM stack reveal a strong dependence on the sputtering conditions of MgO. Due to the highly stable sputtered MgO, repeatable resistive switching memory is achieved with a low ON voltage of ∼0.7 V and a memory window of ∼1 × 105. Limiting Ag diffusion through a modified top electrode in the W/Ag/MgO/Au stack significantly reduces the abruptness of resistive switching, thereby demonstrating analog switching capability. This phenomenon is evident in the improved linearity and symmetry of potentiation and depression weight modulation pulses, demonstrating ideal Hebbian synaptic learning rules.

Improved resistive switching of RGO and SnO2 based resistive memory device for non-volatile memory application

Resistive switching memory, which combines simple architectures with the ability of high density, high switching speed, and low power consumption, has sparked a lot of attention in the field of non-volatile memory devices. In the present work, the resistive switching performance of reduced graphene oxide (RGO) and tin oxide (SnO2) based nanocomposite have been carried out. RGO, SnO2, and RGO-SnO2 nanocomposite were synthesized via hydrothermal technique and the formation of nanocomposite is confirmed by X-ray diffraction and Raman spectroscopy techniques. We herein present a bipolar resistive switching memory based on SnO2 and RGO-SnO2 fabricated by a simple spin-coating process. The two different memory devices have been fabricated by depositing pure SnO2, and RGO-SnO2 films over the bottom ITO electrode and top Al electrodes deposited through a shadow mask. It was observed that devices with RGO-SnO2 showed improved resistive switching in comparison to pure SnO2 film. This improvement of the composite-based film is recognized in terms of switching parameters, such as a reduction in the device operating voltage from 4.11 V for SnO2 to 1 V for RGO-SnO2 composite-based devices. In addition, the resistance ratio in the composite-based films showed significant enhancement. Further, the retention and endurance behavior of the fabricated composite film-based device was tested up to 1 * 103 s and up to 100 cycles respectively which didn’t exhibit any degradation. The operating mechanism of the memory device was explained by fitting the current-voltage characteristics in the low-resistance and high-resistance states. It may be highlighted that the present metal oxide and RGO-based devices have great potential for future non-volatile memory devices.

Impact of Relaxation on the Performance of GeSe True Random Number Generator Based on Ovonic Threshold Switching

Volatile Ovonic threshold switching (OTS) are promising not only as the selector in crossbar resistive switching memory arrays, but also as true random number generators (TRNG) by utilizing its probabilistic switching characteristics. However, investigation on the reliability of OTS-based TRNG is still lacking, which hinders its practical application. Previously, we found that switching probability is dependent on the pulse amplitude and width. In this work, we report that relaxation which happens during the time interval between pulses can also cause switching probability drift. Optimizing the bit-generation waveform and modulating the pulse conditions could provide a practical solution, in addition to the impact of external bias and temperature. This work provides useful guidance for the practical design and operation of OTS-based TRNGs.

Transparent ZnO resistive switching memory fabricated by neutral oxygen beam treatment

In this work, a Cu/ZnO/ITO resistive random access memory (RRAM) structure in which ZnO films are irradiated with neutral oxygen beams was employed to investigate the effect of neutral oxygen beams as a surface treatment. It was confirmed that the treatment reduced the defect concentration in the sputtered-ZnO film and improved the resistance change characteristics of the device. These results indicate the great potential of neutral oxygen beams in the development of RRAM devices using ZnO films.

Conduction mechanism and impedance analysis of HfO[formula omitted]-based RRAM at different resistive states

Conduction mechanism and impedance analysis of both the low resistance state (LRS) and high resistance state (HRS) are performed on TiN/HfOx/Pt resistive-switching memory devices at various temperatures. The dc conduction mechanism in LRS is dictated by Ohmic conduction, whereas by trap controlled space-charge-limited current conduction in HRS. The XPS analysis results on the surface of the HfOx film indicated the presence of oxygen deficiencies. The equivalent physical model under various resistance states is established by using ac impedance spectroscopy method, that is related to oxygen vacancies in the HfOx film. We further discussed the activation energy of dc and ac in the HRS and found the conduction occurs at the grain boundary. Through analysis of different resistance states of HfOx-based resistive switching memories, the conduction mechanism correlation between time and frequency domain is established.

Highly Flexible and Asymmetric Hexagonal‐Shaped Crystalline Structured Germanium Dioxide‐Based Multistate Resistive Switching Memory Device for Data Storage and Neuromorphic Computing

AbstractWith the increase of big data and artificial intelligence (AI) applications, fast and energy‐efficient computing is critical in future electronics. Fortunately, nonvolatile resistive memory devices can be potential candidates for these issues due to their in‐computing and neuromorphic computational abilities. Hence, the paper proposes a highly flexible and asymmetric hexagonal‐shaped crystalline structured germanium dioxide‐based Ag/GeO2/ITO device for high data storage and neuromorphic computing. The proposed device shows the highly asymmetric memristor behavior at low operating voltage to block backward current. The operational behaviors are observed by modulating the applied amplitude, current compliance, and varying the frequency, which shows excellent stability and repeatability in electrical characterizations. Furthermore, the neuromorphic device exhibits synaptic learning properties such as potentiation‐depression, pulse amplification, and spike time‐dependent plasticity rules (STDP). Here, the weights update of the memristive synaptic device is analyzed using a multilayer perceptron convolutional neural network (CNN) by optimizing the learning rate, training epochs, and algorithm to achieve higher accuracy for pattern recognition using CIFAR‐10 data. Undoubtedly, the demonstrated results suggest that the proposed device is a promising candidate to develop high‐density storage and neuromorphic computing technology for wearable and AI electronics.

MESO-LUT: A design approach of look up tables based on MESO devices

The look-up-table (LUT) is widely used in the field of the field programmable gate arrays (FPGA), in which the high efficiencies of energy consumption, area, and speed are primarily concerned. In this paper, a new interconnected LUT structure constructed by the core element of the magnetoelectric spin–orbit logic (MESO) device is proposed. A SPICE model of a MESO-LUT circuit to achieve a two-input logic function is established by which the writing, reading and standby power consumptions are investigated. Comparing with the state-of-art LUT designs based on spin–orbit torque device, resistive switching memory and clockless spin-based LUT, the number of the auxiliary transistors of the proposed 6-input MESO-LUT respectively decreased ∼ 13% , ∼41% and ∼ 81%,due to the property of representing the information by current direction which is promising in the future low power application scenario.