Photoelectric Multilevel Memory Device based on Covalent Organic Polymer Film with Keto–Enol Tautomerism for Harsh Environments Applications

Covalent organic polymers (COPs) memristors with multilevel memory behavior in harsh environments and photoelectric regulation are crucial for high-density storage and high-efficiency neuromorphic computing. Here, a donor–acceptor (D–A)-type COP film (Py-COP-3), which is initiated by keto–enol tautomerism, is proposed for high-performance memristors. Satisfactorily, the indium tin oxide (ITO)/Py-COP-3/Ag device demonstrates multilevel memory performance, even in high temperatures, acid-base corrosion, and various organic solvents. Moreover, the performance can be modulated by the photoelectric effect to maintain a great switching behavior. By contrast, Py-COP-0, with similar structure and chemical composition to Py-COP-3 but without keto–enol tautomerism, exhibits binary storage performance. Further studies unravel that both the formation of conductive filaments and charge transfer within D-A Py-COP-3 film contribute to the resistive switching behavior of memory devices.     

Self‐Limited Switching in Ta2O5/TaOx Memristors Exhibiting Uniform Multilevel Changes in Resistance

To facilitate the development of memristive devices, it is essential to resolve the problem of non‐uniformity in switching, which is caused by the random nature of the filamentary switching mechanism in many resistance switching memories based on transition metal oxide. In addition, device parameters such as low‐ and high‐state resistance should be regulated as desired. These issues can be overcome if memristive devices have switching limits for both the low‐ and high‐resistance states and if their resistance values are highly controllable. In this study, a method termed self‐limited switching for uniformly regulating the values of both the low‐ and high‐resistance states is suggested, and the circuit configuration required for the self‐limited switching is established in a Ta₂O₅/TaO_x memristive structure. A method of improving the uniformity of multi‐level resistance states in this memristive system is also proposed.     

Multifunctional neurosynaptic devices for human perception systems

The traditional Von Neumann architecture for processing information is difficult to meet the needs of the big data era, while low-power, small-sized neurosynaptic devices can operate and store information, so that they have received extensive attention. Due to the development of artificial intelligence and robotics, neurosynaptic devices have been given high expectations and requirements. The trend of functionalization, intelligence, and integration of computing and storage is obvious. In this review, the basic principles and types of neurosynaptic devices are summarized, the achievements of neurosynaptic devices for human perception systems are discussed and a prospect on the development trend is also given.     

Ternary Conductance Switching Realized by a Pillar[5]arene-Functionalized Two-Dimensional Imine Polymer Film

Nowadays, most manufacturing memory devices are based on materials with electrical bistability (i. e., “0” and “1”) in response to an applied electric field. Memory devices with multilevel states are highly desired so as to produce high-density and efficient memory devices. Herein, we report the first multichannel strategy to realize a ternary-state memristor. We make use of the intrinsic sub-nanometer channel of pillar[5]arene and nanometer channel of a two-dimensional imine polymer to construct an active layer with multilevel channels for ternary memory devices. Low threshold voltage, long retention time, clearly distinguishable resistance states, high ON/OFF ratio (OFF/ON1/ON2=1 : 10 : 10³), and high ternary yield (75 %) were obtained. In addition, the flexible memory device based on 2DP_TPAZ+TAPB can maintain its stable ternary memory performance after being bent 500 times. The device also exhibits excellent thermal stability and can tolerate a temperature as high as 300 °C. It is envisioned that the results of this work will open up possibilities for multistate, flexible resistive memories with good thermal stability and low energy consumption, and broaden the application of pillar[n]arene.     

Robust DNA-Bridged Memristor for Textile Chips

Electronic textiles may revolutionize many fields, such as communication, health care and artificial intelligence. To date, unfortunately, computing with them is not yet possible. Memristors are compatible with the interwoven structure and manufacturing process in textiles because of its two-terminal crossbar configuration. However, it remains a challenge to realize textile memristors owing to the difficulties in designing advanced memristive materials and achieving high-quality active layers on fiber electrodes. Herein we report a robust textile memristor based on an electrophoretic-deposited active layer of deoxyribonucleic acid (DNA) on fiber electrodes. The unique architecture and orientation of DNA molecules with the incorporation of Ag nanoparticles offer the best-in-class performances, e.g., both ultra-low operation voltage of 0.3 V and power consumption of 100 pW and high switching speed of 20 ns. Fundamental logic calculations such as implication and NAND are demonstrated as functions of textile chips, and it has been thus integrated with power-supplying and light emitting modules to demonstrate an all-fabric information processing system.     

Mean Field Theory of Self-Organizing Memristive Connectomes

Biological neuronal networks are characterized by nonlinear interactions and complex connectivity. Given the growing impetus to build neuromorphic computers, understanding physical devices that exhibit structures and functionalities similar to biological neural networks is an important step toward this goal. Self-organizing circuits of nanodevices are at the forefront of the research in neuromorphic computing, as their behavior mimics synaptic plasticity features of biological neuronal circuits. However, an effective theory to describe their behavior is lacking. This study provides for the first time an effective mean field theory for the emergent voltage-induced polymorphism of circuits of a nanowire connectome, showing that the behavior of these circuits can be explained by a low-dimensional dynamical equation. The equation can be derived from the microscopic dynamics of a single memristive junction in analytical form. The effective model is tested on experiments of nanowire networks and show that it fits both the potentiation and depression of these synapse-mimicking circuits. It is shown that this theory applies beyond the case of nanowire networks by formulating a general mean-field theory of conductance transitions in self-organizing memristive connectomes.     

Neuronal-like Irregular Spiking Dynamics in Highly Volatile Memristive Intermediate-scale AgPt-Nanoparticle Assemblies

Neuromorphic computing seeks functional materials capable of emulating brain-like dynamics to solve computational problems with time and energy efficiency, outclassing current transistor-based hardware architectures. Major efforts are focused on integrating memristive devices into highly regular circuits (i.e., crossbar arrays), where the information representation in individual memristive devices is closely oriented toward the behavior of artificial neurons. However, artificial neurons are rather rigid mathematical concepts than realistic projections of complex neuronal dynamics. Neuroscience suggests that highly efficient information representation on the level of individual neurons relies on dynamical features such as excitatory and inhibitory contributions, irregularity of firing patterns, and temporal correlations. Here, a conductive atomic force microscopy approach is applied to probe the memristive dynamics of nanoscale assemblies of AgPt-nanoparticles at the stability border of the conducting state, where physical forces causing the formation and decay of filamentary structures appear to be balanced. This unveils a dynamic regime, where the memristive response is governed by irregular firing patterns. The significance of such a dynamical regime is motivated by close similarities to excitation and inhibition-governed behavior in biological neuronal systems, which is crucial to tune biological neuronal systems into a state most suitable for information representation and computation.