Memristor-based neural networks: a bridge from device to artificial intelligence

Optoelectronic neuron based on transistor combined with volatile threshold switching memristors for neuromorphic computing

The human perception and learning heavily rely on the visual system, where the retina plays a vital role in preprocessing visual information. Developing neuromorphic vision hardware is based on imitating the neurobiological functions of the retina. In this work, an optoelectronic neuron is developed by combining a gate-modulated PDVT-10 channel with a volatile threshold switching memristor, enabling the achievement of optoelectronic performance through a resistance-matching mechanism. The optoelectronic spiking neuron exhibits the ability to alter its spiking behavior in a manner resembling that of a retina. Incorporating electrical and optical modulation, the artificial neuron accurately replicates neuronal signal transmission in a biologically manner. Moreover, it demonstrates inhibition of neuronal firing during darkness and activation upon exposure to light. Finally, the evaluation of a perceptron spiking neural network utilizing these leaky integrate-and-fire neurons is conducted through simulation to assess its capability in classifying image recognition algorithms. This research offers a hopeful direction for the development of easily expandable and hierarchically structured spiking electronics, broadening the range of potential applications in biomimetic vision within the emerging field of neuromorphic hardware.

Artificial synapses based on HfOx/TiOymemristor devices for neuromorphic applications

As a result of enormous progress in nanoscale electronics, interest in artificial intelligence (AI) supported systems has also increased greatly. These systems are typically designed to process computationally intensive data. Parallel processing neural network architectures are particularly noteworthy for their ability to process dense data at high speeds, making them suitable candidates for AI algorithms. Due to their ability to combine processing and memory functions in a single device, memristors offer a significant advantage over other electronic platforms in terms of area scaling efficiency and energy savings. In this study, single-layer and bilayer metal-oxide HfOxand TiOymemristor devices inspired by biological synapses were fabricated by pulsed laser and magnetron sputtering deposition techniques in high vacuum with different oxide thicknesses. The structural and electrical properties of the fabricated devices were analysed using x-ray reflectivity, x-ray photoelectron spectroscopy, and standard two-probe electrical characterization measurements. The stoichiometry and degree of oxidation of the elements in the oxide material for each thin film were determined. Moreover, the switching characteristics of the metal oxide upper layer in bilayer devices indicated its potential as a selective layer for synapse. The devices successfully maintained the previous conductivity values, and the conductivity increased after each pulse and reached its maximum value. Furthermore, the study successfully observed synaptic behaviours with long-term potentiation, long-term depression (LTD), paired-pulse facilitation, and spike-timing-dependent plasticity, showcasing potential of the devices for neuromorphic computing applications.

Recent developments of artificial intelligence in MXene-based devices: from synthesis to applications

Two-dimensional transition metal carbides, nitrides, or carbonitrides (MXenes) have garnered remarkable attention in various energy and environmental applications due to their high electrical conductivity, good thermal properties, large surface area, high mechanical strength, rapid charge transport mechanism, and tunable surface properties. Recently, artificial intelligence has been considered an emerging technology, and has been widely used in materials science, engineering, and biomedical applications due to its high efficiency and precision. In this review, we focus on the role of artificial intelligence-based technology in MXene-based devices and discuss the latest research directions of artificial intelligence in MXene-based devices, especially the use of artificial intelligence-based modeling tools for energy storage devices, sensors, and memristors. In addition, emphasis is given to recent progress made in synthesis methods for various MXenes and their advantages and disadvantages. Finally, the review ends with several recommendations and suggestions regarding the role of artificial intelligence in fabricating MXene-based devices. We anticipate that this review will provide guidelines on future research directions suitable for practical applications.

Exploring Nanoscale Perovskite Materials for Next-Generation Photodetectors: A Comprehensive Review and Future Directions

The rapid advancement of nanotechnology has sparked much interest in applying nanoscale perovskite materials for photodetection applications. These materials are promising candidates for next-generation photodetectors (PDs) due to their unique optoelectronic properties and flexible synthesis routes. This review explores the approaches used in the development and use of optoelectronic devices made of different nanoscale perovskite architectures, including quantum dots, nanosheets, nanorods, nanowires, and nanocrystals. Through a thorough analysis of recent literature, the review also addresses common issues like the mechanisms underlying the degradation of perovskite PDs and offers perspectives on potential solutions to improve stability and scalability that impede widespread implementation. In addition, it highlights that photodetection encompasses the detection of light fields in dimensions other than light intensity and suggests potential avenues for future research to overcome these obstacles and fully realize the potential of nanoscale perovskite materials in state-of-the-art photodetection systems. This review provides a comprehensive overview of nanoscale perovskite PDs and guides future research efforts towards improved performance and wider applicability, making it a valuable resource for researchers.

Multicolor Fully Light-Modulated Artificial Synapse Based on P-MoSe2/PxOy Heterostructured Memristor

Developing brain-inspired neuromorphic paradigms is imperative to breaking through the von Neumann bottleneck. The emulation of synaptic functionality has motivated the exploration of optoelectronic memristive devices as high-performance artificial synapses, yet the realization of such a modulatory terminal capable of full light-modulation, especially near-infrared stimuli, remains a challenge. Here, a fully light-modulated synaptic memristor is reported on a P-MoSe2/PxOy heterostructure formed by a facile one-step selenization process. The results demonstrate successful achievement of multiwavelength (visible 470 nm to near-infrared 808 nm) modulated switching operations (reset in 0.21-0.97 V) and diverse synaptic behaviors, including postsynaptic current, paired-pulse facilitation, short- and long-term memory (STM and LTM), and learning-forgetting. Notably, the device can exhibit a 3.42 μA PSC increase under six identical 655 nm stimuli, a 11.90-46.24 μA PSC modulation by changing 808 nm light intensity from 6 to 14 mW/cm2, and a transition from STM to LTM lasting between 2.47 and 4.27 s by a prolonged 808 nm pulse from 1 to 30 s. A novel possible light-induced switching mechanism in such a heterostructure is proposed. Furthermore, brain-like light-stimulated memory behavior and Pavlov's classical conditioning demonstrate the device's capacity for processing complex inputs. The study presents a design toward a multiwavelength modulated artificial visual system for color recognition.

Understanding the Reversible Transition of Unipolar and Bipolar Resistive Switching Characteristics in Solution-Derived Nanocrystalline Au-Co3O4 Thin-Film Memristors

Nanocrystalline Au-Co3O4 thin films were fabricated through a facile solution-processing method, and voltage-polarity-dependent resistive switching (RS) characteristics including variability of Set/Reset voltages, stability of cycling endurance, and carriers transport mechanism were studied in the Pt/Au-Co3O4/Pt memory devices. The switching voltages of the Set and Reset processes in the bipolar RS memristors exhibited lower variability as compared to the unipolar resistance switching devices. Moreover, the switching performance of cycle-to-cycle endurance in bipolar mode had a smaller fluctuation than those of unipolar switching behavior. Based on the current-voltage curve fitting analysis, it was found that Ohmic-conduction behaviors dominated the carriers transport of low-resistance state regardless of unipolar or bipolar switching behaviors. The carrier transport of high resistance state was governed following Poole-Frenkel emission and Schottky emission mechanisms in the unipolar and bipolar switching modes, respectively. The physical switching mechanism of Pt/Au-Co3O4/Pt memory devices was proposed using the model by means of growth and disruption of conducting filaments, involved in memory effects of thermochemical mechanism and valence change mechanism in the unipolar and bipolar RS, respectively. The proposed model following the finite element method revealed the roles of electrical field distribution and temperature gradient to further clarify the RS mechanism. Our results open a door for understanding and optimizing oxide-based thin-film resistance switching memory devices.

Dynamic memristor for physical reservoir computing

Reservoir computing (RC) has attracted considerable attention for its efficient handling of temporal signals and lower training costs. As a nonlinear dynamic system, RC can map low-dimensional inputs into high-dimensional spaces and implement classification using a simple linear readout layer. The memristor exhibits complex dynamic characteristics due to its internal physical processes, which renders them an ideal choice for the implementation of physical reservoir computing (PRC) systems. This review focuses on PRC systems based on memristors, explaining the resistive switching mechanism at the device level and emphasizing the tunability of their dynamic behavior. The development of memristor-based reservoir computing systems is highlighted, along with discussions on the challenges faced by this field and potential future research directions.

A reversible implantable memristor for health monitoring applications

Conventional implantable electronics based on von Neumann architectures encounter significant limitations in computing and processing vast biological information due to computational bottlenecks. The memristor with integrated memory-computing and low power consumption offer a promising solution to overcome the computational bottleneck and Moore's law limitations of traditional silicon-based implantable devices, making them the most promising candidates for next-generation implantable devices. In this work, a highly stable memristor with an Ag/BaTiO3/MnO2/FTO structure was fabricated, demonstrating retention characteristics exceeding 1200 cycles and endurance above 1000 s. The device successfully exhibited three-stage responses to biological signals after implantation in SD (Sprague-Dawley) rats. Importantly, the memristor perform remarkable reversibility, maintaining over 100 cycles of stable repetition even after extraction from the rat. This study provides a new perspective on the biomedical application of memristors, expanding the potential of implantable memristive devices in intelligent medical fields such as health monitoring and auxiliary diagnostics.

Memristor-Based Bionic Tactile Devices: Opening the Door for Next-Generation Artificial Intelligence

Bioinspired tactile devices can effectively mimic and reproduce the functions of the human tactile system, presenting significant potential in the field of next-generation wearable electronics. In particular, memristor-based bionic tactile devices have attracted considerable attention due to their exceptional characteristics of high flexibility, low power consumption, and adaptability. These devices provide advanced wearability and high-precision tactile sensing capabilities, thus emerging as an important research area within bioinspired electronics. This paper delves into the integration of memristors with other sensing and controlling systems and offers a comprehensive analysis of the recent research advancements in memristor-based bionic tactile devices. These advancements incorporate artificial nociceptors and flexible electronic skin (e-skin) into the category of bio-inspired sensors equipped with capabilities for sensing, processing, and responding to stimuli, which are expected to catalyze revolutionary changes in human-computer interaction. Finally, this review discusses the challenges faced by memristor-based bionic tactile devices in terms of material selection, structural design, and sensor signal processing for the development of artificial intelligence. Additionally, it also outlines future research directions and application prospects of these devices, while proposing feasible solutions to address the identified challenges.

Graphene oxide-DNA/graphene oxide-PDDA sandwiched membranes with neuromorphic function

The behavior of polyelectrolytes in confined spaces has direct relevance to the protein mediated ion transport in living organisms. In this paper, we govern lithium chloride transport by the interface provided by polyelectrolytes, polycation, poly(diallyldimethylammonium chloride) (PDDA) and, polyanion, double stranded deoxyribonucleic acid (dsDNA), in confined graphene oxide (GO) membranes. Polyelectrolyte-GO interfaces demonstrate neuromorphic functions that were successfully applied with nanochannel ion interactions contributed, resulting in ion memory effects. Excitatory and inhibitory post-synaptic currents were tuned continuously as the number of pulses applied increased accordingly, increasing decay times. Furthermore, we demonstrated the short-term memory of a trained vs untrained device in computation. On account of its simple and safe production along with its robustness and stability, we anticipate our device to be a low dimensional building block for arrays to embed artificial neural networks in hardware for neuromorphic computing. Additionally, incorporating such devices with sensing and actuating parts for a complete feedback loop produces robotics with its own ability to learn by modifying actuation based on sensing data.

Tunable Neuromorphic Switching Dynamics via Porosity Control in Mesoporous Silica Diffusive Memristors

In response to the growing need for efficient processing of temporal information, neuromorphic computing systems are placing increased emphasis on the switching dynamics of memristors. While the switching dynamics can be regulated by the properties of input signals, the ability of controlling it via electrolyte properties of a memristor is essential to further enrich the switching states and improve data processing capability. This study presents the synthesis of mesoporous silica (mSiO2) films using a sol-gel process, which enables the creation of films with controllable porosities. These films can serve as electrolyte layers in the diffusive memristors and lead to tunable neuromorphic switching dynamics. The mSiO2 memristors demonstrate short-term plasticity, which is essential for temporal signal processing. As porosity increases, discernible changes in operating currents, facilitation ratios, and relaxation times are observed. The underlying mechanism of such systematic control was investigated and attributed to the modulation of hydrogen-bonded networks within the porous structure of the silica layer, which significantly influences both anodic oxidation and ion migration processes during switching events. The result of this work presents mesoporous silica as a unique platform for precise control of neuromorphic switching dynamics in diffusive memristors.

Unraveling the Neural Circuits: Techniques, Opportunities and Challenges in Epilepsy Research

Epilepsy, a prevalent neurological disorder characterized by high morbidity, frequent recurrence, and potential drug resistance, profoundly affects millions of people globally. Understanding the microscopic mechanisms underlying seizures is crucial for effective epilepsy treatment, and a thorough understanding of the intricate neural circuits underlying epilepsy is vital for the development of targeted therapies and the enhancement of clinical outcomes. This review begins with an exploration of the historical evolution of techniques used in studying neural circuits related to epilepsy. It then provides an extensive overview of diverse techniques employed in this domain, discussing their fundamental principles, strengths, limitations, as well as their application. Additionally, the synthesis of multiple techniques to unveil the complexity of neural circuits is summarized. Finally, this review also presents targeted drug therapies associated with epileptic neural circuits. By providing a critical assessment of methodologies used in the study of epileptic neural circuits, this review seeks to enhance the understanding of these techniques, stimulate innovative approaches for unraveling epilepsy's complexities, and ultimately facilitate improved treatment and clinical translation for epilepsy.

Technology and Integration Roadmap for Optoelectronic Memristor

Optoelectronic memristors (OMs) have emerged as a promising optoelectronic Neuromorphic computing paradigm, opening up new opportunities for neurosynaptic devices and optoelectronic systems. These OMs possess a range of desirable features including minimal crosstalk, high bandwidth, low power consumption, zero latency, and the ability to replicate crucial neurological functions such as vision and optical memory. By incorporating large-scale parallel synaptic structures, OMs are anticipated to greatly enhance high-performance and low-power in-memory computing, effectively overcoming the limitations of the von Neumann bottleneck. However, progress in this field necessitates a comprehensive understanding of suitable structures and techniques for integrating low-dimensional materials into optoelectronic integrated circuit platforms. This review aims to offer a comprehensive overview of the fundamental performance, mechanisms, design of structures, applications, and integration roadmap of optoelectronic synaptic memristors. By establishing connections between materials, multilayer optoelectronic memristor units, and monolithic optoelectronic integrated circuits, this review seeks to provide insights into emerging technologies and future prospects that are expected to drive innovation and widespread adoption in the near future.

Unveiling transient current response in bilayer oxide-based physical reservoirs for time-series data analysis

Physical reservoirs employed to map time-series data and analyze extracted features have attracted interest owing to their low training cost and mitigated interconnection complexity. This study reports a physical reservoir based on a bilayer oxide-based dynamic memristor. The proposed device exhibits a nonlinear current response and short-term memory (STM), satisfying the requirements of reservoir computing (RC). These characteristics are validated using a compact model to account for resistive switching (RS) via the dynamic evolution of the internal state variable and the relocation of oxygen vacancies. Mathematically, the transient current response can be quantitatively described according to a simple set of equations to correlate the theoretical framework with experimental results. Furthermore, the device shows significant reliability and ability to distinguish 4-bit inputs and four diverse neural firing patterns. Therefore, this work shows the feasibility of implementing physical reservoirs in hardware and advances the understanding of the dynamic response.

Interface Engineering Enables Multilevel Resistive Switching in Ultra-Low-Power Chemobrionic Copper Silicate

Memristor is assuming prominence due to its exceptionally low power consumption, adaptable, and parallel signal processing capabilities that address the limitations of the von Neumann architecture to meet the growing demand for advanced technologies such as artificial intelligence, Internet of Things (IoTs), and neuromorphic computation. In this work, we demonstrate resistive switching in copper silicate-based hollow tube-forming self-organized membrane structures belonging to the category of chemobrionics or chemical gardens to demonstrate cost-effective and highly efficient memristor devices. The device architecture is configured as ITO/PEDOT:PSS/active layer (copper silicate)/PMMA/Ag, an arrangement that serves to stabilize current-voltage hysteresis and exhibit a low SET voltage ∼0.2 V with a 0.8 nJ power consumption while manifesting robust data endurance and multilevel resistive switching. The inherent self-rectifying behavior, characterized by a high rectification ratio of 60, underscores the potential utility of these devices across a spectrum of electronic applications. To emulate the functionality of biological synapses, fundamental synaptic characteristics are assessed, including paired-pulse facilitation (PPF) and potentiation and depression (P&D). We validate the potential of copper silicate chemical garden-based memristor devices for applications that require real-time synaptic processing. Importantly, the fabrication of these devices was accomplished through a comprehensive solution-based, low-temperature process conducted under ambient environmental conditions, obviating the need for specialized glovebox facilities.

Memristor-Based Artificial Chips

Memristors, promising nanoelectronic devices with in-memory resistive switching behavior that is assembled with a physically integrated core processing unit (CPU) and memory unit and even possesses highly possible multistate electrical behavior, could avoid the von Neumann bottleneck of traditional computing devices and show a highly efficient ability of parallel computation and high information storage. These advantages position them as potential candidates for future data-centric computing requirements and add remarkable vigor to the research of next-generation artificial intelligence (AI) systems, particularly those that involve brain-like intelligence applications. This work provides an overview of the evolution of memristor-based devices, from their initial use in creating artificial synapses and neural networks to their application in developing advanced AI systems and brain-like chips. It offers a broad perspective of the key device primitives enabling their special applications from the view of materials, nanostructure, and mechanism models. We highlight these demonstrations of memristor-based nanoelectronic devices that have potential for use in the field of brain-like AI, point out the existing challenges of memristor-based nanodevices toward brain-like chips, and propose the guiding principle and promising outlook for future device promotion and system optimization in the biomedical AI field.

Resistive Switching in Bigraphene/Diamane Nanostructures Formed on a La3Ga5SiO14 Substrate Using Electron Beam Irradiation

Memristors, resistive switching memory devices, play a crucial role in the energy-efficient implementation of artificial intelligence. This study investigates resistive switching behavior in a lateral 2D composite structure composed of bilayer graphene and 2D diamond (diamane) nanostructures formed using electron beam irradiation. The resulting bigraphene/diamane structure exhibits nonlinear charge carrier transport behavior and a significant increase in resistance. It is shown that the resistive switching of the nanostructure is well controlled using bias voltage. The impact of an electrical field on the bonding of diamane-stabilizing functional groups is investigated. By subjecting the lateral bigraphene/diamane/bigraphene nanostructure to a sufficiently strong electric field, the migration of hydrogen ions and/or oxygen-related groups located on one or both sides of the nanostructure can occur. This process leads to the disruption of sp3 carbon bonds, restoring the high conductivity of bigraphene.

Effect of Hydrogen Annealing on Performances of BN-Based RRAM

BN-based resistive random-access memory (RRAM) has emerged as a potential candidate for non-volatile memory (NVM) in aerospace applications, offering high thermal conductivity, excellent mechanical, and chemical stability, low power consumption, high density, and reliability. However, the presence of defects and trap states in BN-based RRAM can limit its performance and reliability in aerospace applications. As a result, higher set voltages of 1.4 and 1.23 V were obtained for non-annealed and nitrogen-annealed BN-based RRAM, respectively, but lower set voltages of 1.06 V were obtained for hydrogen-annealed BN-based RRAM. In addition, the hydrogen-annealed BN-based RRAM showed an on/off ratio of 100, which is 10 times higher than the non-annealed BN-based RRAM. We observed that the LRS changed to the HRS state before 10,000 s for both the non-annealed and nitrogen-annealed BN-based RRAMs. In contrast, the hydrogen-annealed BN-based RRAM showed excellent retention characteristics, with data retained up to 10,000 s.