Nonvolatile Devices Research Articles

Emerging tunnel FET and spintronics-based hardware-secure circuit design with ultra-low energy consumption

Present complementary metal–oxide–semiconductor (CMOS) technology with scaled channel lengths exhibits higher energy consumption in designing secure electronic circuits against hardware vulnerabilities and breaches. Specifically, CMOS sense amplifier-based secure differential power analysis (DPA) countermeasures at scaled channel lengths show large energy consumption, with increased vulnerability. Additionally, spin-transfer torque magnetic tunnel junction (STT-MTJ) and CMOS-based logic-in-memory (LiM) cells demonstrate high energy consumption due to the large write current requirement of the STT-MTJ and poor MOS device performance at scaled channel lengths. This paper for the first time leverages emerging tunnel field effect transistor (TFET) steep-slope device characteristics and compatible non-volatile STT-MTJ devices for enhanced hardware security with ultra-low energy consumption at lower supply voltages. TFET-based sense amplifier-based logic (SABL) gates are proposed that achieve 3× lower energy consumption than the Si FinFET SABL designs. Further, utilizing TFET SABL gates, a TFET PRIDE S-box is designed that exhibits higher DPA resilience with 3.2× lower energy consumption than the FinFET designs. With the resulting lower static power consumption, TFET SABL-based cryptosystems are thus less vulnerable to static power side-channel attacks. Additionally, the proposed STT-MTJ and TFET LiM gates achieve 4× lower energy consumption than the STT-MTJ and FinFET designs. Lastly, these gates are explored in a logic encryption/locking technique that shows 3.1× lower energy consumption than the STT-MTJ and FinFET-based design.

Anisotropic laser-pulse-induced magnetization dynamics in van der Waals magnet Fe3GeTe2

Femtosecond laser-pulse excitation provides an energy efficient and fast way to control magnetization at the nanoscale, providing great potential for ultrafast next-generation data manipulation and nonvolatile storage devices. Ferromagnetic van der Waals materials have garnered much attention over the past few years due to their low dimensionality, excellent magnetic properties, and large response to external stimuli. Nonetheless, their behaviour upon fs laser-pulse excitation remains largely unexplored. Here, we investigate the ultrafast magnetization dynamics of a thin flake of Fe3GeTe2 (FGT) and extract its intrinsic magnetic properties using a microscopic framework. We find that our data is well described by our modeling, with FGT undergoing a slow two-step demagnetization, and we experimentally extract the spin-relaxation timescale as a function of temperature, magnetic field and excitation fluence. Our observations indicate a large spin-flip probability in agreement with a theoretically expected large spin–orbit coupling, as well as a weak interlayer exchange coupling. The spin-flip probability is found to increase when the magnetization is pulled away from its quantization axis, opening doors to an external control over the spins in this material. Our results provide a deeper understanding of the dynamics van der Waals materials upon fs laser-pulse excitation, paving the way towards two-dimensional materials-based ultrafast spintronics.

Photo‐Enhanced Resistive Switching Effect in High‐Performance MAPbI3 Memristors

AbstractThe rapid development of artificial intelligence (AI) requires processing vast amounts of complex information, which has accelerated the exploration of neuromorphic computing systems. Artificial neuromorphic synapses based on memristors of organic–inorganic halide perovskites (OHPs) potentially exploit a niche area for brain‐inspired neuromorphic computing, which can be operated as biological synapses to realize signal processing. Here, MAPbI3‐based memristors with reliable resistance states triggered by electric fields or photons are reported. A model for resistive switching (RS) originated from conductive filaments (CFs) based on intrinsic defect migrations is proposed. Importantly, the unique photoresponsive characteristic provides the opportunity to enhance the RS through multifunctional photo‐coupling. Enhanced by monochromatic illumination, memristors exhibit RS with remarkable characteristics such as ultralow operating voltage, high ON/OFF ratio (4.3 × 103), small HRS/LRS variation coefficient (29.91%/13.82%), stable endurance (104 cycles), long retention time (105 s), and ultralow power consumption. Moreover, photons can modulate the nonvolatile devices to maintain a great ON/OFF ratio over 9 days under ambient conditions without any encapsulation. The research presents plausible applications of memristors in coupling ions, electrons, and photons, thus contributing to applicability for multifunctional optoelectronics and optogenetics tunable neuromorphic systems.

A Singlet-Diradical Co(III)-Dimer as a Nonvolatile Resistive Switching Device: Synthesis, Redox-Induced Interconversion, and Current-Voltage Characteristics.

Herein we report a ligand-centered redox-controlled strategy for the synthesis of an unusual binuclear diradical cobalt(III) complex, [Co2III(L•3-)2] (1), featuring two three-electron reduced trianionic monoradical 2,9-bis(phenyldiazo)-1,10-phenanthroline ligands (L•3-) and two intermediate-spin cobalt(III) centers having a Co-Co bond. Controlled ligand-centered oxidation of 1 afforded two mononuclear complexes, [CoII(L•-)(L0)]+ ([3])+ and [CoII(L0)2]2+ ([2]2+), which upon further ligand-centered reduction yielded a di-azo-anion diradical complex, [CoII(L•-)2] (4). In complex 1, two three-electron reduced di-azo-anion monoradical ligands (L•3-) bridge two intermediate Co(III) centers at a distance of 2.387(2) Å, while upon oxidation, one of the coordinating azo-arms of L becomes pendent, and in complexes [2]2+, [3]+, and 4, two tetradentate ligands coordinate a single Co(II) center in a tridentate meridional fashion with one uncoordinated azo-arm from each of the ligands. In the presence of reducing agents, the monomers [2]2+, [3]+, and 4 undergo ligand-centered reduction to form azo-anion radicals, and the otherwise pendent azo-arms in the presence of cobalt(II)-salts like Co(ClO4)2 or CoCl2 bind the second Co(II)-ion; further internal electron transfer from the cobalt center to the arylazo backbone produces the binuclear complex 1. Spectroscopic analysis, DFT studies, and control experiments were performed to understand the electronic structures and the ligand-centered redox-controlled interconversion. The application of complex 1 as a molecular memory device (memristor) was also explored. Complex 1 showed encouraging results as a memristor with a current ON/OFF ratio > 104 and is highly promising for resistive RAM/ROM applications.

2D electrene LaH2 monolayer: an ideal ferrovalley direct semiconductor with room-temperature ferromagnetic stability

In developing nonvolatile valleytronic devices, ferromagnetic (FM) ferrovalley semiconductors are critically needed due to the existence of spontaneous valley polarization. At present, however, the known real materials have various drawbacks towards practical applications, including the in-plane FM ground state, low Curie temperature (), small valley polarization, narrow energy window with clean polarized valley, and indirect bandgap. From first-principles calculations, here we predict an ideal ferrovalley semiconductor, honeycomb LaH2 monolayer (ML), whose intrinsic properties can overcome all these shortcomings. We demonstrate that LaH2 ML, having satisfied structural stability, is a FM half-semiconducting electrene () with its magnetic moments localized at the lattice interstitial sites rather than La atoms. At the same time, LaH2 ML holds the following desired attributes: a robust out-of-plane FM ground state with a high (334 K), a sizable valley polarization (166 meV), a wide energy window (137 meV) harboring clean single-valley carriers, and a direct bandgap. These results identify a much needed ideal ferrovalley semiconductor candidate, holding the promising application potential in valleytronics and spintronics devices.

Variation of magnetism in two-dimensional MnS2 thin films

Atomically thin layers with variable magnetism are highly desirable for next-generation data storage and spintronic devices. Variations of magnetism in two-dimensional (2D) materials have inspired 2D device designs that mimic their functions. In this work, the electronic structure and charge transfer of 2D MnS2 thin films consisting of two monolayers can be changed between two different structures. Variation of magnetism between two different antiferromagnetic (AFM) configurations are realized in the MnS2 thin films. The calculated magnetic anisotropy shows that the easy axes of the thin films are perpendicular to the xy-plane. The underlying variation of magnetism in the air-stable, semiconductive, and facile 2D MnS2 system is revealed. Variations in the magnetism of the 2D thin films are determined from charge transfer between the Mn and S as induced by changes in the stacking orders of the nonmagnetic S atoms. Our propose-designed MnS2 thin films enable emerging 2D magnetism and exotic physical phenomena for nonvolatile spintronic devices that operate without external strains or, electric or magnetic fields.

Multiple polarization phases and strong magnetoelectric coupling in the layered transition metal phosphorus chalcogenides TMP2X6(T=Cu,Ag;M=Cr,V;X=S,Se) by controlling the interlayer interaction and dimension

Magnetoelectric multiferroic materials are the potential candidates for the high-density nonvolatile data storage devices. However, up to now, multiferroic materials with strong magnetoelectric coupling are still rare. Here, based on the first-principles calculations and theoretical model, we predict a different class of single phase multiferroic materials, transition metal phosphorus chalcogenides $TM{\mathrm{P}}_{2}{X}_{6}(T=\mathrm{Cu},\phantom{\rule{0.16em}{0ex}}\mathrm{Ag};\phantom{\rule{0.16em}{0ex}}M=\mathrm{Cr},\phantom{\rule{0.16em}{0ex}}\mathrm{V};\phantom{\rule{0.16em}{0ex}}X=\mathrm{S},\phantom{\rule{0.16em}{0ex}}\mathrm{Se})$ with multiple polarization phases and strong magnetoelectric coupling. The ferroelectric polarization originates from the movement of Cu/Ag atoms breaking the symmetry of spatial inversion and the magnetism arises from partially filled $d$ orbitals of the V/Cr atoms. It is predicted that the different ferroelectric phases of $TM{\mathrm{P}}_{2}{X}_{6}$ bulk have different band gaps, providing a way to control electronic and transport properties by the external electric field. Most prominently, for $\mathrm{CuV}{\mathrm{P}}_{2}{\mathrm{S}}_{6}$ bilayer and few layers, one of the ferroelectric phases has ferromagnetic ground state and the other has antiferromagnetic states, realizing the electric-field control of magnetism. We reveal that the physical mechanism of the strong magnetoelectric coupling is from the reduced dimension and symmetry by constructing a theoretical model including the crystal field splitting, electric polarization effect, and exchange interaction. This work not only predicts a different class of magnetoelectric multiferroic materials, but also proposes a strategy to design them by controlling the interlayer interaction in van der Waals layered materials.

Memristive FHN spiking neuron model and brain-inspired threshold logic computing

The FitzHugh-Nagumo (FHN) spiking model has rich dynamics behaviors and can imitate the firing process of a neuron. The memristor is a nonvolatile and resistance tunable device, which gradually becomes a potential candidate for performing dynamic behaviors and implementing neuromorphic computation in the nervous system. In this paper, the memristive FitzHugh-Nagumo (MFHN) spiking model is presented. Firstly, we introduce a memristor to the FHN spiking model to build the MFHN spiking model and analyze the phase plane trajectory of the MFHN model. Secondly, we couple two MFHN models with a memristor and experimentally simulate the dynamic behaviors of two coupled neurons. The synchronous and asynchronous phenomena of two coupled neurons are discussed. Thirdly, the feasibility of the MFHN spiking model is demonstrated by the realization of binary logical operations and the implementation of binary adders. The comparison between the MFHN binary adder and the FHN binary adder is conducted. The simulation results illustrate that the proposed model efficiently performs rich, dynamic behaviors and higher firing frequency. The coupled MFHN models show the effectiveness of a memristor acting as an electric coupling synapse. The threshold logic computation can be completed efficiently by the MFHN spiking model. The MFHN binary adders reproduce the computing functions and behaviors of the biological neuron.

A compute-in-memory chip based on resistive random-access memory

Realizing increasingly complex artificial intelligence (AI) functionalities directly on edge devices calls for unprecedented energy efficiency of edge hardware. Compute-in-memory (CIM) based on resistive random-access memory (RRAM)1 promises to meet such demand by storing AI model weights in dense, analogue and non-volatile RRAM devices, and by performing AI computation directly within RRAM, thus eliminating power-hungry data movement between separate compute and memory2–5. Although recent studies have demonstrated in-memory matrix-vector multiplication on fully integrated RRAM-CIM hardware6–17, it remains a goal for a RRAM-CIM chip to simultaneously deliver high energy efficiency, versatility to support diverse models and software-comparable accuracy. Although efficiency, versatility and accuracy are all indispensable for broad adoption of the technology, the inter-related trade-offs among them cannot be addressed by isolated improvements on any single abstraction level of the design. Here, by co-optimizing across all hierarchies of the design from algorithms and architecture to circuits and devices, we present NeuRRAM—a RRAM-based CIM chip that simultaneously delivers versatility in reconfiguring CIM cores for diverse model architectures, energy efficiency that is two-times better than previous state-of-the-art RRAM-CIM chips across various computational bit-precisions, and inference accuracy comparable to software models quantized to four-bit weights across various AI tasks, including accuracy of 99.0 percent on MNIST18 and 85.7 percent on CIFAR-1019 image classification, 84.7-percent accuracy on Google speech command recognition20, and a 70-percent reduction in image-reconstruction error on a Bayesian image-recovery task.

Terahertz multi-level nonvolatile optically rewritable encryption memory based on chalcogenide phase-change materials

SummaryFast and efficient information processing and encryption, including writing, reading, and encryption memory, is essential for upcoming terahertz (THz) communications and information encryption. Here, we demonstrate a THz multi-level, nonvolatile, optically rewritable memory and encryption memory based on chalcogenide phase-change materials, Ge2Sb2Te5 (GST). By tuning the laser fluence irradiated on GST, we experimentally achieve multiple intermediate states and large-area amorphization with a diameter of centimeter-level in the THz regime. Our memory unit features a high operating speed of up to 4 ns, excellent reproducibility, and long-term stability. Utilizing this approach, hexadecimal coding information memories are implemented, and multiple writing-erasing tests are successfully carried out in the same active area. Finally, terahertz photoprint memory is demonstrated, verifying the feasibility of lithography-free devices. The demonstration suggests a practical way to protect and store information and paves a new avenue toward nonvolatile active THz devices.

Realization of the skyrmionic logic gates and diodes in the same racetrack with enhanced and modified edges

Magnetic skyrmions are topological quasiparticles with nanoscale size and high mobility, which have potential applications in information storage and spintronic devices. Here, we computationally investigate the dynamics of isolated skyrmions in a ferromagnetic racetrack, where magnetic properties of the edges are enhanced and modified, forming a channel with lower magnetic anisotropy for skyrmion motion. It is found that the rectangular notch at the edge can have a pinning effect on the skyrmion and enrich the dynamics of the skyrmion. Based on the racetrack with modified edges and the notch, we design a racetrack that realizes the skyrmionic logic AND, OR, and NOT gates as well as the diode in the same magnetic racetrack. It is found that the driving current density could be much smaller than those used in previous designs of skyrmion-based logic gates. By slightly altering the shape of the racetrack, we also design the NAND and NOR gates. Finally, we study the feasibility of our design at finite temperatures. Our results may contribute to the design of nonvolatile spintronic devices with integrated multiple functions and ultra-low energy consumption.

Sintering characteristics of plate-like bismuth titanate ceramics with preferred c - axis orientation for memory storage devices

Bismuth titanate (Bi4Ti3O12) (BIT) has been an attractive material due to its high Curie temperature, large remnant polarization and relatively low processing temperature, making it suitable for piezoelectric, pyroelectric and non-volatile random access memory storage devices. In this present work, polycrystalline textured bismuth titanate ceramics were synthesised by the cost-effective molten salt synthesis method. The effect of sintering conditions on the structural, morphological and dielectric properties have been carried out in detail. The X-ray diffraction studies reveal the formation of the monophasic structure of BIT ceramics with a preferential orientation along the c-axis. It is also noticed that the orientation along the c-axis increases with an increase in sintering temperature. Interestingly, the parameters like crystallite size and micro-strain calculated from the Williamson - Hall method could be tailored as a function of sintering temperature. The microstructural studies carried out using FESEM technique suggested the presence of plate-like morphology of BIT ceramics and subsequently, the elemental mapping analysis used to confirm the uniform distribution of the elements Bi, Ti and O present in the ceramic sample. The Fourier transform infrared and X-ray photoelectron spectroscopic studies have been done to identify the constituent chemical bonds and oxidation states of BIT ceramics. Interestingly, it is noted that the value of the dielectric constant increases with the increase in sintering temperature of BIT ceramics. Furthermore, the high dielectric constant value of 200 obtained at 100 kHz for 950 °C sintered sample could establish this material for the embedded capacitor in memory storage devices.

Tunable physical properties of Ba doped BiFeO3 multiferroic nanoceramics for capacitor and memory storage devices

Recently, perovskite structured bismuth ferrite (BiFeO3) multiferroic material has been widely studied owing to its potential applications in various devices like spintronics, energy storage, sensors and photovoltaic devices. Interestingly, the presence of two or greater stable polarization states in this material and the ease of polarization switching by way of electric field exhibited high energy storage densities, making it appropriate for information storage devices. However, the large leakage current and weak magnetism associated with the material at 300 K hinders its usage in practical applications. It is noticed that the substitution of Bi by alkaline earth elements (Mg2+, Sr2+, Ca2+) could significantly enhance the physical properties of BiFeO3 ceramics. In this work, Bi1-xBaxFeO3 (x = 0.05–0.2 mol %) nanoceramics were prepared by adopting low temperature molten salt synthesis route using NaCl as flux. The X-ray diffraction studies revealed the formation of distorted perovskite structure of BiFeO3 phase in all the range of compositions studied. The FESEM analysis demonstrated the presence of nanosized grains of BiFeO3 and the average grain size decreases with an increase in Ba doping content. The X-ray photoelectron spectra revealed the existence of different valence states of the elements present in BiFeO3 nanoceramics. It is interesting to observe that the dielectric properties of Ba doped BiFeO3 ceramics are superior to the undoped samples, especially at the higher frequency range due to the ionic size difference of barium and bismuth ions in the crystal lattice of BiFeO3, which would significantly influence the energy storage capacity in the non-volatile random access memory storage devices.

An improved reconfigurable logic in resistive random access memory

Emerging non-volatile devices have shown great potential in computing in-memory (CIM). This work proposes a logic design method based on the complementary resistance switching (CRS) structure, which is connected by two anti-serially memristors. Three logics including AND, OR, and NIMP can be implemented in one step. Moreover, XOR logic can be implemented in two steps. Based on the proposed design, a 1-bit full adder is implemented with the fewer number of memristors and operation steps compared with existing works. The feasibility and correctness of our design are verified by the SPICE simulations with the Voltage Threshold Adaptive Memristor (VTEAM) model. In addition, we evaluate the ISCAS’85 benchmark circuit based on the proposed logic. Compared with existing methods, the performance and area overhead have been improved by the proposed method.

In-Memory Mathematical Operations with Spin-Orbit Torque Devices.

Analog arithmetic operations are the most fundamental mathematical operations used in image and signal processing as well as artificial intelligence (AI). In‐memory computing (IMC) offers a high performance and energy‐efficient computing paradigm. To date, in‐memory analog arithmetic operations with emerging nonvolatile devices are usually implemented using discrete components, which limits the scalability and blocks large scale integration. Here, a prototypical implementation of in‐memory analog arithmetic operations (summation, subtraction and multiplication) is experimentally demonstrated, based on in‐memory electrical current sensing units using spin‐orbit torque (SOT) devices. The proposed structures for analog arithmetic operations are smaller than the state‐of‐the‐art complementary metal oxide semiconductor (CMOS) counterparts by several orders of magnitude. Moreover, data to be processed and computing results can be locally stored, or the analog computing can be done in the nonvolatile SOT devices, which are exploited to experimentally implement the image edge detection and signal amplitude modulation with a simple structure. Furthermore, an artificial neural network (ANN) with SOT devices based synapses is constructed to realize pattern recognition with high accuracy of ≈95%.

Energy‐Efficient Photonic Memory Cell with Spatially Separated Recording/Erasing and Readout Channels

Phase-change material Ge2Sb2Te5 (GST) is a promising candidate for nonvolatile photonic devices due to its phase stability, different absorption coefficients in the amorphous and crystalline phase, and relatively fast response. Here, an energy-efficient design of a photonic nonvolatile memory that stores its logical status as GST phase states is proposed. The device is based on the silicon-on-insulator (SOI) substrate and consists of the fourfold symmetric strip waveguide crossing incorporated with the square GST patch and a multimode interference (MMI) loop reflector. The distinctive feature of the proposed cell is two spatially separated recording/erasing and readout channels. One of the crossed waveguides is connected to the MMI loop reflector and used for recording/erasing. The other waveguide is used for readout. The numerical simulation of the light transmission through the device in question shows that the MMI loop reflector provides 45% and 80% higher absorption values in comparison with waveguide end reflector and straightforward transmission, respectively. This fact allows the more efficient utilization of the guided mode energy and makes the proposed photonic memory cell an attractive node element for photonic integrated circuits.

Resonant multilevel optical switching with phase change material GST

Abstract We demonstrate a multilevel optical memristive switch based on a silicon Fabry–Perot resonator. The resonator is constructed by a pair of waveguide Bragg gratings at the ends of a multimode interferometer (MMI) covered with sub-micrometer-size Ge2Sb2Te5 (GST) thin film on top. The interaction between the optical field and GST is greatly enhanced due to the resonant effect. The GST phase transition is triggered by applying electrical pulses to the doped-silicon microheater. Light is transmitted when GST is amorphous while it is highly absorbed by the crystalline GST at the resonance wavelength, leading to a higher on-off extinction ratio (ER) compared to the non-resonant device. The resonant device achieves a maximum transmission contrast of 10.29 dB and a total of 38 distinct nonvolatile switching levels. Our work provides an effective solution to improving the multilevel switching performance of phase-change devices and paves the way for future nonvolatile silicon photonics devices.

The technique to symmetrize domain switching hysteresis loops in LiNbO3 domain-wall nanodevices with improved polarization retention

Ferroelectric devices have wide applications in nonvolatile random-access memories, sensors, actuators, and transducers. The built-in potential at the interfaces could result in poor polarization retention. Here, we found an effective way to independently adjust two coercive fields of LiNbO3 mesa-like domain wall devices in contact to two side electrodes at the surface. Taking advantage of the electrode shielding effect on the depolarization field across an interfacial layer, the near-zero coercive field increases almost linearly with respect to the extending length of one side electrode covering the tail of the reversed domain, enabling nonvolatile ferroelectric domain wall devices in good polarization retention and fast operation speeds.

Stable Resistive Switching in ZnO/PVA:MoS2 Bilayer Memristor.

Reliability of nonvolatile resistive switching devices is the key point for practical applications of next-generation nonvolatile memories. Nowadays, nanostructured organic/inorganic heterojunction composites have gained wide attention due to their application potential in terms of large scalability and low-cost fabrication technique. In this study, the interaction between polyvinyl alcohol (PVA) and two-dimensional material molybdenum disulfide (MoS2) with different mixing ratios was investigated. The result confirms that the optimal ratio of PVA:MoS2 is 4:1, which presents an excellent resistive switching behavior. Moreover, we propose a resistive switching model of Ag/ZnO/PVA:MoS2/ITO bilayer structure, which inserts the ZnO as the protective layer between the electrode and the composite film. Compared with the device without ZnO layer structure, the resistive switching performance of Ag/ZnO/PVA:MoS2/ITO was improved greatly. Furthermore, a large resistive memory window up to 104 was observed in the Ag/ZnO/PVA:MoS2/ITO device, which enhanced at least three orders of magnitude more than the Ag/PVA:MoS2/ITO device. The proposed nanostructured Ag/ZnO/PVA:MoS2/ITO device has shown great application potential for the nonvolatile multilevel data storage memory.

Realization of long retention properties of quantum conductance through confining the oxygen vacancy diffusion

Quantum conductance, known as Sharvin point contact, has been extensively investigated in many electronic devices, including diodes, transistors, and switches, especially in conductive filaments-based memristors. Quantum conductance with one or multiple atoms point connection can overcome the limitations of scaling and operating speed of nonvolatile multiple memory, logic device, and brain-inspired computing systems. However, because of the instability of the atomic arrangement in the one/multiple atoms connection in a conductive filaments-based memristor, it is a great challenge to maintain quantum conductance states for a long time. Here, we demonstrate that the stable long-time retention of multi-level quantum conductance states can be realized in Mott insulator vanadium dioxide with a highly oriented crystalline texture. According to in situ transmission electron microscope, conductive atomic force microscope, and detailed energy band analysis results, it is proposed that the grain boundaries act as reservoirs for oxygen vacancies and confine the oxygen vacancy diffusion in the narrow grain boundaries due to the higher bulk diffusion barrier. Our approach is extremely crucial for realizing quantum conductance-based electronic devices, such as multi-level and high-density storage and neuromorphic computing.