Synaptic Learning and Memory Functions Achieved Using Oxygen Ion Migration/Diffusion in an Amorphous InGaZnO Memristor

A single synaptic device with inherent learning and memory functions is demonstrated based on an amorphous InGaZnO (α‐IGZO) memristor; several essential synaptic functions are simultaneously achieved in such a single device, including nonlinear transmission characteristics, spike‐rate‐dependent and spike‐timing‐dependent plasticity, long‐term/short‐term plasticity (LSP and STP) and “learning‐experience” behavior. These characteristics bear striking resemblances to certain learning and memory functions of biological systems. Especially, a “learning‐experience” function is obtained for the first time, which is thought to be related to the metastable local structures in α‐IGZO. These functions are interrelated: frequent stimulation can cause an enhancement of LTP, both spike‐rate‐dependent and spike‐timing‐dependent plasticity is the same on this point; and, the STP‐to‐LTP transition can occur through repeated “stimulation” training. The physical mechanism of device operation, which does not strictly follow the memristor model, is attributed to oxygen ion migration/diffusion. A correlation between short‐term memory and ion diffusion is established by studying the temperature dependence of the relaxation processes of STP and ion diffusion. The realization of important synaptic functions and the establishment of a dynamic model would promote more accurate modeling of the synapse for artificial neural network.     

Memristive Logic‐in‐Memory Integrated Circuits for Energy‐Efficient Flexible Electronics

A memristive nonvolatile logic‐in‐memory circuit can provide a novel energy‐efficient computing architecture for battery‐powered flexible electronics. However, the cell‐to‐cell interference existing in the memristor crossbar array impedes both the reading process and parallel computing. Here, it is demonstrated that integration of an amorphous In‐Zn‐Sn‐O (a‐IZTO) semiconductor‐based selector (1S) device and a poly(1,3,5‐trivinyl‐1,3,5‐trimethyl cyclotrisiloxane) (pV3D3)‐based memristor (1M) on a flexible substrate can overcome these problems. The developed a‐IZTO‐based selector device, having a Pd/a‐IZTO/Pd structure, exhibits nonlinear current–voltage (I–V) characteristics with outstanding stability against electrical and mechanical stresses. Its underlying conduction mechanism is systematically determined via the temperature‐dependent I–V characteristics. The flexible one‐selector?one‐memristor (1S–1M) array exhibits reliable electrical characteristics and significant leakage current suppression. Furthermore, single‐instruction multiple‐data (SIMD), the foundation of parallel computing, is successfully implemented by performing NOT and NOR gates over multiple rows within the 1S–1M array. The results presented here will pave the way for development of a flexible nonvolatile logic‐in‐memory circuit for energy‐efficient flexible electronics.     

Bienenstock,Cooper, and Munro Learning Rules Realized in Second‐Order Memristors with Tunable Forgetting Rate

Memristors with synaptic functions are very promising for developing artificial neural networks. Compared with the extensively reported spike‐timing‐dependent plasticity (STDP), Bienenstock, Cooper, and Munro (BCM) learning rules, the most accurate model of the synaptic plasticity to date, are more compatible with the neural computing system; however, the progress in the realization of the BCM rules has been quite limited. The realized BCM rules so far mostly performs just the spike‐rate‐dependent plasticity (SRDP), however, without a tunable sliding frequency threshold, because the memristors used to realize the BCM rules do not have tunable forgetting rates. In this work, the BCM rules with a tunable sliding frequency threshold are biorealistically implemented in SrTiO₃‐based second‐order memristors; the forgetting rate of the memristors is tuned by engineering the electrode/oxide interface through controlling the electrode composition. The approach of this work is precise and efficient, and the biorealistic implementation of the BCM rules in memristors improves the efficiency of the neural network for the artificial intelligent system.     

Adaptive Crystallite Kinetics in Homogenous Bilayer Oxide Memristor for Emulating Diverse Synaptic Plasticity

A critical routine for memristors applied to neuromorphic computing is to approximate synaptic dynamic behaviors as closely as possible. A type of homogenous bilayer memristor with a structure of W/HfO_y/HfO_x/Pt is designed and constructed in this paper. The memristor replicates the structure and oxygen vacancy (V_O) distribution of a complete synapse and its Ca²⁺ distribution, respectively, after the forming process. The detailed characterizations of its atomic structure and phase transformation in and near the conductive channel demonstrate that the crystallite kinetics are adaptively coupled with the V_O migration prompted by directional external bias. The extrusion (injection) of the V_Os and the subsequent crystallite coalescence (separation), phase transformation, and alignment (misalignment) resemble closely the Ca²⁺ flux and neurotransmitter dynamics in chemical synapses. Such adaptation and similarity allow the memristor to emulate diverse synaptic plasticity. This study supplies a kinetic process of conductive channel theory for bilayer memristors. In addition, our memristor has very low energy consumption (5–7.5 fJ per switching for a 0.5 µm diameter device, compatible with a synaptic event) and is therefore suitable for large‐scale integration used in neuromorphic networks.     

Artificial Synapses Based on Multiterminal Memtransistors for Neuromorphic Application

Neuromorphic computing, which emulates the biological neural systems could overcome the high‐power consumption issue of conventional von‐Neumann computing. State‐of‐the‐art artificial synapses made of two‐terminal memristors, however, show variability in filament formation and limited capacity due to their inherent single presynaptic input design. Here, a memtransistor‐based arti?cial synapse is realized by integrating a memristor and selector transistor into a multiterminal device using monolayer polycrys‐talline‐MoS₂ grown by a scalable chemical vapor deposition (CVD) process. Notably, the memtransistor offers both drain‐ and gate‐tunable nonvolatile memory functions, which efficiently emulates the long‐term potentiation/depression, spike‐amplitude, and spike‐timing‐dependent plasticity of biological synapses. Moreover, the gate tunability function that is not achievable in two‐terminal memristors, enables significant bipolar resistive states switching up to four orders‐of‐magnitude and high cycling endurance. First‐principles calculations reveal a new resistive switching mechanism driven by the diffusion of double sulfur vacancy perpendicular to the MoS₂ grain boundary, leading to a conducting switching path without the need for a filament forming process. The seamless integration of multiterminal memtransistors may offer another degree‐of‐freedom to tune the synaptic plasticity by a third gate terminal for enabling complex neuromorphic learning.     

Photoelectric Synaptic Plasticity Realized by 2D Perovskite

Recently, several light‐stimulated artificial synaptic devices have been proposed to mimic photonic synaptic plasticity for neuromorphic computing. Here, the photoelectric synaptic plasticity based on 2D lead‐free perovskite ((PEA)₂SnI₄) is demonstrated. The devices show a photocurrent activation in response to a light stimulus in a neuron‐like way and exhibit several essential synaptic functions such as short‐term plasticity (STP) and long‐term plasticity (LTP) as well as their transmission based on spike frequency control. The strength of synaptic connectivity can be effectively modulated by the duration, irradiance, and wavelength of light spikes. The ternary structure of (PEA)₂SnI₄ causes it to possess varied photoelectric properties by composition control, which enhances the complexity and freedoms required by neuromorphic computing. The physical mechanisms of the memory effect are attributed to two distinct lifetimes of photogenerated carrier trapping/detrapping processes modulated by controlling the proportion of Sn vacancies. This work demonstrates the great potential of (PEA)₂SnI₄ as a platform to develop future multifunctional artificial neuromorphic systems.     

HfO2-Based Memristor as an Artificial Synapse for Neuromorphic Computing with Tri-Layer HfO2/BiFeO3/HfO2 Design

Neuromorphic devices are among the most emerging electronic components to realize artificial neural systems and replace traditional complementary metal–oxide semiconductor devices in recent times. In this work, tri-layer HfO₂/BiFeO₃(BFO)/HfO₂ memristors are designed by inserting traditional ferroelectric BFO layers measuring ≈4 nm after thickness optimization. The novel designed memristor shows excellent resistive switching (RS) performance such as a storage window of 10⁴ and multi-level storage ability. Remarkably, essential synaptic functions can be successfully realized on the basis of the linearity of conductance modulation. The pattern recognition simulation based on neuromorphic network is conducted with 91.2% high recognition accuracy. To explore the RS performance enhancement and artificial synaptic behaviors, conductive filaments (CFs) composed of Hafnium (Hf) single crystal with a hexaganal lattice structure are observed using high-resolution transmission electron microscopy. It is reasonable to believe that the sufficient oxygen vacancies in the inserting BFO thin film play a crucial role in adjusting the morphology and growth of Hf CFs, which leads to the promising synaptic and enhanced RS behavior, thus demonstrating the potential of this memristor for use in neuromorphic computing.     

Direct Observation of Oxygen Ion Dynamics in a WO3-x based Second-Order Memristor with Dendritic Integration Functions

Direct observation of oxygen dynamics in an oxide-based second-order memristor can provide the valid evidence to clarify the memristive mechanism, however, which is still limited for now. In this study, the migration and diffusion of oxygen ions in the region of Pt/WO_3-x Schottky interface are observed in the WO_3-x second-order memristor by using the technique of in situ transmission electron microscopy (TEM) and the electron energy loss spectroscopy. Interestingly, the coexistence of memristive and memcapacitive switching can be implemented in this memristor. Combined with the analysis of depth-profile X-ray photoelectron spectroscopy (XPS), an interface-barrier-modulation second-order memristive model is proposed based on the above results. Notably, temporally correlative oxygen dynamics in the memristor offers the platform to integrate signals from multiple inputs, enabling the realization of the dendritic functions of synchronous and asynchronous integration for the application of logic operations with fault-tolerance capability and associative learning. These findings provide the experimental evidence to in-depth understanding of oxygen dynamics and switching mechanism in second-order memristor, which can support the optimization of memristive performance and the achievement of biorealistic synaptic functions.     

Electrolyte‐Gated Synaptic Transistor with Oxygen Ions

Artificial synaptic devices are the essential hardware of neuromorphic computing systems, which can simultaneously perform signal processing and information storage between two neighboring artificial neurons. Emerging electrolyte‐gated transistors have attracted much attention for efficient synaptic emulation by using an addition gate terminal. Here, an electrolyte‐gated synaptic device based on the SrCoO_x (SCO) films is proposed. It is demonstrated that the reversible modulation of SCO phase transforms the brownmillerite SrCoO_2.5 and perovskite SrCoO_3?δ , through controlling the insertion and extraction of oxygen ions with electrolyte gating. Nonvolatile multilevel conduction states can be realized in the SCO films following this route. The synaptic functions such as the long‐term potentiation and depression of synaptic weight, spike‐timing‐dependent plasticity, as well as spiking logic operations in the device are successfully mimicked. These results provide an alternative avenue for future neuromorphic devices via electrolyte‐gated transistors with oxygen ions.     

Surface Modification of a Titanium Carbide MXene Memristor to Enhance Memory Window and Low-Power Operation

With the demand for low-power-operating artificial intelligence systems, bio-inspired memristor devices exhibit potential in terms of high-density memory functions and the emulation of the synaptic dynamics of the human brain. The 2D material MXene attracts considerable interest for use in resistive-switching memory and artificial synapse devices owing to its excellent physicochemical properties in memristor devices. However, few memristive and synaptic MXene devices that display increased switching performances are reported, with no significant results. Herein, the conductivity of MXene (Ti₃C₂T_x) is engineered via etching and oxidation to enhance the switching performance of the device. The exceptional properties of partially oxidized MXene memristors include large memory windows and low threshold biases, and the complex spike-timing-dependent plasticity synaptic rules are also emulated. The low threshold potential distribution, reliable retention time (10⁴ s), and distinct resistance states with a high ON–OFF ratio (>10⁴) are the main memory-related features of this device. The experimentally determined switching potentials of the optimized device are also uniformly distributed, according to a statistical probability-based approach. This investigation may promote the essential material properties for use in high-density non-volatile memory storage and artificial synapse systems in the field of innovative nanoelectronic devices.     

Bioinspired Artificial Sensory Nerve Based on Nafion Memristor

Bioinspired artificial haptic neuron system has received much attention in the booming artificial intelligence industry for its broad range of high‐impact applications such as personal healthcare monitoring, electronic skins, and human–machine interfaces. An artificial haptic neuron system is designed by integrating a piezoresistive sensor and a Nafion‐based memristor for the first time in this paper. The piezoresistive sensor serves as a sensory receptor to transform mechanical stimuli into electric signals, and the Nafion‐based memristor serves as the synapse to further process the information. The pyramid‐structured sensor exhibits excellent sensitivity (6.7 × 10⁷ kPa^?1 in 1–5 kPa and 3.8 × 10⁵ kPa^?1 in 5–50 kPa) and durability (>7000 cycles), while the memristor realizes fundamental synaptic functions under low power consumption (10–200 pJ) and remains stable for over 10⁴ consecutive tests. The integrated system can detect tactile stimuli encoded with temporal information, such as the count, frequency, duration and speed of the external force. As a proof‐of‐concept, English characters recognition with high accuracy can be achieved on the system under a supervised learning method. This work shows promising potential in bioinspired sensing systems owing to the high performance, excellent durability, and simple fabrication procedure.     

MXene-ZnO Memristor for Multimodal In-Sensor Computing

Recently, in-sensor computing with individual sensors or multiple connected sensors directly processing information has been proposed to improve energy, area, and time efficiency of artificial intelligence systems. Current investigations mainly focus on a single sensory processing such as auditory, visual, tactile, olfactory, and so on. However, a human perception system can sense and process different types of information with a complex environment and small perceptive field simultaneously. For example, the recognition accuracy of human eyes is highly affected by the environment such as extremely low or high relative humidity (RH). Here, a multi-modal MXene-ZnO memristor that combines visual data sensing, RH sensing, and pre-processing functions to emulate the unique environmental adaptive behavior of the human eye is designed and constructed. The multi-field controlled resistive switching of the MXene-ZnO memristor is originated from the photon-/protons-regulated formation of oxygen vacancies filaments. Finally, in-sensor computing with a MXene-ZnO memristor functioning as both filter to preprocess the information and synapse to implement a weight updating process with different humidity adaptability has been demonstrated. Multimodal in-sensor computing provides the potential to reduce the underlying circuitry complexity of the traditional neuromorphic visual system and contributes to the development of intelligence in device-level implementations.     

Graphene Oxide Quantum Dots Based Memristors with Progressive Conduction Tuning for Artificial Synaptic Learning

Memristors as electronic artificial synapses have attracted increasing attention in neuromorphic computing. Emulation of both “learning” and “forgetting” processes requires a bidirectional progressive adjustment of memristor conductance, which is a challenge for cutting‐edge artificial intelligence. In this work, a memristor device with a structure of Ag/Zr_0.5Hf_0.5O₂:graphene oxide quantum dots/Ag is presented with the feature of bidirectional progressive conductance tuning. The conductance of proposed memristor is adjusted through voltage pulse number, amplitude, and width. A series of voltage pulses with an amplitude of 0.6 V and a width of 30 ns is enough to modulate conductance. The impacts of pulses with different parameters on conductance modulation are investigated, and the potential relationship between pulse amplitude and energy is revealed. Furthermore, it is proved that the pulse with low energy can realize the almost linear conductance regulation, which is beneficial to improve the accuracy of pattern recognition. The bidirectional progressive conduction modulation mimics various plastic synapses, such as spike‐timing‐dependent plasticity and paired‐pulse facilitation. This progressive conduction tuning mechanism might be attributed to the coexistence of tunneling effect and extrinsic electrochemical metallization effect. This work provides one way for memristor to attain attractive features such as bidirectional tuning, low‐power consumption, and fast speed switching that is in urgent demand for further evolution of neuromorphic chips.     

Modeling the Metal–Insulator Phase Transition in LixCoO2 for Energy and Information Storage

An electro‐chemomechanical phase‐field model is developed to capture the metal–insulator phase transformation along with the structural and chemical changes that occur in Li_xCoO₂ in the regular operating range of 0.5 < x < 1. Under equilibrium, in the regime of phase coexistence, it is found that transport limitations lead to kinetically arrested states that are not determined by strain‐energy minimization. Further, lithiation profiles are obtained for different discharging rates and the experimentally observed voltage plateau is observed. Finally, a simple model is developed to account for the conductivity changes for a polycrystalline Li_xCoO₂ thin film as it transforms from the metallic phase to the insulating phase and a strategy is outlined for memristor design. The theory can therefore be used for modeling Li_xCoO₂‐electrode batteries as well as low voltage nonvolatile redox transistors for neuromorphic computing architectures.     

Memristor with Ag‐Cluster‐Doped TiO2 Films as Artificial Synapse for Neuroinspired Computing

Memristor, based on the principle of biological synapse, is recognized as one of the key devices in confronting the bottleneck of classical von Neumann computers. However, conventional memristors are difficult to continuously adjust the conduction and dutifully mimic the biosynapse function. Here, TiO₂ films with self‐assembled Ag nanoclusters implemented by gradient Ag dopant are employed to achieve enhanced memristor performance. The memristors exhibit gradual both potentiating and depressing conduction under positive and negative pulse trains, which can fully emulate excitation and inhibition of biosynapse. Moreover, comprehensive biosynaptic functions and plasticity, including the transition from short‐term memory to long‐term memory, long‐term potentiation and depression, spike‐timing‐dependent plasticity, and paired‐pulse facilitation, are implemented with the fabricated memristors in this work. The applied pulses with a width of hundreds of nanoseconds timescale are beneficial to realize fast learning and computing. High‐resolution transmission electron microscopy observations clearly demonstrate that Ag clusters redistribute to form Ag conductive filaments between Ag and Pt electrode under electrical field at ON‐state device. The experimental data confirm that the oxides doped with Ag clusters have the potential for mimicking biosynaptic behavior, which is essential for the further creation of artificial neural systems.     

Modelling of nanostructured TiO2-based memristors

The fourth fundamental circuit element memristor completes the missing link between charge and magnetic flux. It consists of the function of the resistor as well as memory in nonlinear fashion. The property of the memristor depends on the magnitude and direction of applied potential. This unique property makes it the primitive building block for many applications such as resistive memories, soft computing, neuromorphic systems and chaotic circuits etc. In this paper we report TiO2-based nanostructured memristor modelling. The present memristor model is constructed in MATLAB environment with consideration of the linear drift model of memristor. The result obtained from the linear drift model is well matched with earlier reported results by other research groups.     

Functional Demonstration of a Memristive Arithmetic Logic Unit (MemALU) for In‐Memory Computing

The development of in‐memory computing has opened up possibilities to build next‐generation non‐von‐Neumann computing architecture. Implementation of logic functions within the memristors can significantly improve the energy efficiency and alleviate the bandwidth congestion issue. In this work, the demonstration of arithmetic logic unit functions is presented in a memristive crossbar with implemented non‐volatile Boolean logic and arithmetic computing. For logic implementation, a standard operating voltage mode is proposed for executing reconfigurable stateful IMP, destructive OR, NOR, and non‐destructive OR logic on both the word and bit lines. No additional voltages are needed beyond “V_P” and its negative component. With these basic logic functions, other Boolean functions are constructed within five devices in at most five steps. For arithmetic computing, the fundamental functions including an n‐bit full adder with high parallelism as well as efficient increment, decrement, and shift operations are demonstrated. Other arithmetic blocks, such as subtraction, multiplication, and division are further designed. This work provides solid evidence that memristors can be used as the building block for in‐memory computing, targeting various low‐power edge computing applications.     

Optoelectronic Perovskite Synapses for Neuromorphic Computing

Simulating the human brain for neuromorphic computing has attractive prospects in the field of artificial intelligence. Optoelectronic synapses have been considered to be important cornerstones of neuromorphic computing due to their ability to process optoelectronic input signals intelligently. In this work, optoelectronic synapses based on all‐inorganic perovskite nanoplates are fabricated, and the electronic and photonic synaptic plasticity is investigated. Versatile synaptic functions of the nervous system, including paired‐pulse facilitation, short‐term plasticity, long‐term plasticity, transition from short‐ to long‐term memory, and learning‐experience behavior, are successfully emulated. Furthermore, the synapses exhibit a unique memory backtracking function that can extract historical optoelectronic information. This work could be conducive to the development of artificial intelligence and inspire more research on optoelectronic synapses.     

One Transistor One Electrolyte-Gated Transistor Based Spiking Neural Network for Power-Efficient Neuromorphic Computing System

Neuromorphic computing powered by spiking neural networks (SNN) provides a powerful and efficient information processing paradigm. To harvest the advantage of SNNs, compact and low-power synapses that can reliably practice local learning rules are required, posing significant challenges to the conventional silicon-based platform in terms of area- and energy-efficiency, as well as computing throughput. Here, electrolyte-gated transistors (EGTs) paired with transistors are employed to implement power-efficient neuromorphic computing systems. The one-transistor-one-EGT (1T1E) synapse not only alleviates the self-discharging of EGT but also provides a flexible and efficient way to practice the important spike-timing-dependent plasticity learning rule. Based on that, an SNN with a temporal coding scheme is implemented for associative memory that can learn and recover images of handwritten digits with high robustness. Thanks to the temporal coding scheme and low operation current of EGTs, the energy-efficiency of 1T1E-based SNN is ≈ 30 × lower than that of the prevalent rate coding scheme, and the peak performance is estimated to be 2 pJ/SOP (picojoule per synaptic operation) at the training phase and 80 TOPs⁻¹ W⁻¹ (tera operations per second per watt) at inference phase, respectively. These results pave the way for power-efficient neuromorphic computing systems with wide applications for edge computing.     

A Natural Silk Fibroin Protein‐Based Transparent Bio‐Memristor

The recent discovery of nanoelectronics memristor devices has opened up a new wave of enthusiasm and optimism in revolutionizing electronic circuit design, marking the beginning of new era for the advancement of neuromorphic, high‐density logic and memory applications. Here a highly non‐linear dynamic response of a bio‐memristor is demonstrated using natural silk cocoon fibroin protein of silkworm, Bombyx mori. A film that is transparent across most of the visible spectrum is obtained with the electronic‐grade silk fibroin aqueous solution of ca. 2% (wt/v). Bipolar memristive switching is demonstrated; the switching mechanism is confirmed to be the filamentary switching as observed by probing local conduction behavior at nanoscale using scanning tunneling microscopy. The memristive transition is elucidated by a physical model based on the carrier trapping or detrapping in silk fibroin films and this appears to be due to oxidation and reduction procedures, as evidenced from cyclic voltammetry measurements. Hence, silk fibroin protein could be used as a biomaterial for bio‐memristor devices for applications in advanced bio‐inspired very large scale integration circuit design as well as in biologically inspired synapse links for energy‐efficient neuromorphic computing.