To date, slow Set operation speed and high Reset operation power remain to be important limitations for substituting dynamic random access memory by phase change memory. Here, we demonstrate phase change memory cell based on Ti0.4Sb2Te3 alloy, showing one order of magnitude faster Set operation speed and as low as one-fifth Reset operation power, compared with Ge2Sb2Te5-based phase change memory cell at the same size. The enhancements may be rooted in the common presence of titanium-centred octahedral motifs in both amorphous and crystalline Ti0.4Sb2Te3 phases. The essentially unchanged local structures around the titanium atoms may be responsible for the significantly improved performance, as these structures could act as nucleation centres to facilitate a swift, low-energy order-disorder transition for the rest of the Sb-centred octahedrons. Our study may provide an alternative to the development of high-speed, low-power dynamic random access memory-like phase change memory technology.

Van der Waals heterostructure superlattices of Sb2 Te1 and GeTe are strain-engineered to promote switchable atomic disordering, which is confined to the GeTe layer. Careful control of the strain in the structures presents a new degree of freedom to design the properties of functional superlattice structures for data storage and photonics applications.

Neuromorphic systems aim to implement large‐scale artificial neural network on hardware to ultimately realize human‐level intelligence. The recent development of nonsilicon nanodevices has opened the huge potential of full memristive neural networks (FMNN), consisting of memristive neurons and synapses, for neuromorphic applications. Unlike the widely reported memristive synapses, the development of artificial neurons on memristive devices has less progress. Sophisticated neural dynamics is the major obstacle behind the lagging. Here a rich dynamics‐driven artificial neuron is demonstrated, which successfully emulates partial essential neural features of neural processing, including leaky integration, automatic threshold‐driven fire, and self‐recovery, in a unified manner. The realization of bioplausible artificial neurons on a single device with ultralow power consumption paves the way for constructing energy‐efficient large‐scale FMNN and may boost the development of neuromorphic systems with high density, low power, and fast speed.

No abstract

and high density. [4][5][6][7] Moreover, memristors with analog switching behaviors can faithfully resemble biological computational elements in both structure and switching dynamics. With the intrinsic biomimetic features, memristors could act as the basic computational element in artificial neural networks and have been demonstrated with the capability of solving cognitive computing tasks with spatiotemporal complexity without complex peripheral circuits. [7] Among various material systems, 2D materials recently demonstrated memristive switching behaviors that possess biologically comparable energy consumption compared with the traditional memristors based on oxide materials. [8][9][10][11] Thanks to their atomically thin layers and planar configurations, 2D material based memristors have provided an intriguing window into the motions of ions and opportunities to achieve outstanding electrical performances. [12][13][14] It has been reported that vertical synapses built in 2D MoS 2 push the switching threshold voltages to an extremely low value of 0.1 V. [14] More recently, multiterminal memtransistor consisting of hybrid memristor and transistor were fabricated using 2D materials to realize gate-tunable heterosynaptic functionality, which could not be achieved with transitional materials. [4,15,16] In addition, the rapid development of chemical vapor deposition (CVD) technology enables wafer scale production of 2D material, paving the way for large scale integration of 2D devices. Therefore, dimensionality reduction from 3D to 2D provides an innovative way for further advancing memristor devices in both scalability and electrical performance.Despite enormous efforts have been devoted in investigating 2D material based memristors, progresses are only made on emulating various synaptic functions. Neuromorphic networks comprise layers of artificial neurons that receive, process and transmit signals, and synapses that connect the neurons and evolve to alter the connection patterns during learning. [17,18] Although artificial neurons based on traditional oxide and phase change materials have been implemented, 2D materials have their distinct advantages. [19,20] For instance, the physical properties of 2D materials can easily be modulated by multi factors, such as doping and interface engineering, 2D material based memristors have exhibited superior performance as artificial synapses for neuromorphic computing. However, 2D artificial neurons as have note been exploited as an indispensable computational element owing to the rich dynamics of neurons, which impede the construction of a 2D neuromorphic network. A methodology is developed by introducing ionic migration dynamics and electrochemical reaction into monolayer MoS 2 single crystal and a 2D artificial neuron is realized. The sophisticated electrophysiology process of leaky integrate-and-fire (LIF) is emulated by the injection and extraction of Ag + ions under an e-field in a monolayer MoS 2 device with fine-tuned channel length. Moreover, the fire frequency and ...

No abstract

No abstract

No abstract