Journal of Colloid and Interface Science, Journal Year: 2024, Volume and Issue: 678, P. 325 - 335
Published: Sept. 5, 2024
Language: Английский
Advanced Functional Materials, Journal Year: 2024, Volume and Issue: unknown
Published: Sept. 2, 2024
Abstract To achieve cost‐effectiveness, researchers are exploring various memristors for their adaptation in neuromorphic computing. Recent studies have focused on developing versatile functioning singular memristors, such as those involved on‐receptor computing, which integrates sensory functions into current computing paradigms. Additionally, adaptations like reservoir being investigated systems. In this study, a memristor composed of stack Ti/NbO x /Pt layers is fabricated to explore multifunctional behaviors within single memristor. By applying bias toward the top Ti electrode, gradual changes with volatile features demonstrated, revealing an ion‐migration‐based nonfilamentary switching Leveraging functionality, artificial nociceptor first implemented, demonstrating key biological nociceptors including thresholding, relaxation, no‐adaptation, and sensitization. Subsequently, synapse emulation akin brain achieved through easy conductance potentiation depression diverse functions, enabling mimic learning activities spike firing. Lastly, computational applications explored by adapting edge multi‐bit expanding memristor's across fields behaviors.
Language: Английский
Citations
11
Advanced Functional Materials, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 19, 2025
Abstract In this study, a highly rectifying memristor composed of Pt/TaO x /TiN stack, incorporating complementary metal‐oxide semiconductor‐friendly metal oxide switching layer, is fabricated to assess its performance in diverse range applications. The exhibits characteristics due the Schottky barrier formed by work function difference between Pt and TiN electrodes. For compliance current 1 mA, displays volatile memory properties, attributed migration oxygen ions within TaO layer. Leveraging behavior, synaptic functions—where changes plasticity occur response incoming spikes—are emulated. Additionally, complete functions biological nociceptor are demonstrated, including threshold, relaxation, no‐adaptation, sensitization, recovery. These dynamic then utilized mimic neuronal firing with array, Morse code implementation enabling data generation, computing through cost‐effective reservoir computing. simplicity fabrication process broad implemented single make device promising candidate for future
Language: Английский
Citations
0
Advanced Functional Materials, Journal Year: 2025, Volume and Issue: unknown
Published: March 12, 2025
Abstract Confronted with the surge in information volume, ensuring security and efficiency of transmission is crucial. Therefore, a selective encryption strategy necessary. It involves pre‐identifying image deciding whether to encrypt it based on its importance. Given low manual identification, there an urgent need develop fast efficient recognition module ensure effective processing decrypted data. However, previous studies, encryption, decryption, are implemented by separate modules, which require additional circuit connections limit system flexibility. This study proposes photoelectric memristor MXene quantum dots (MQDs) that exhibits stable volatile nonvolatile functions under electrical control. Using this multifunctional device, integrated into single array, achieving desired through programmed control light signals, thereby increasing device utilization 100%. Through technology, volume data needs be encrypted reduced, when reduced 30% original size, speed tripled. neuromorphic provides solution for energy‐saving system.
Language: Английский
Citations
0
ACS Nano, Journal Year: 2025, Volume and Issue: unknown
Published: March 26, 2025
Two-dimensional (2D) ferroelectric field-effect transistors (Fe-FETs) based on p–n junctions are the basic units of future neuromorphic hardware. The In2Se3 semiconductor with ferroelectric, photoelectric, and phase transition properties possesses great application potential for in-sensor computing, but its junction (FePNJ) is not well investigated. Here, we present an optoelectronic synapse made uniformly full-coverage α-In2Se3/WSe2 FePNJ, achieving ultralow-power classification recognition multiscale signal processing. Using chemical vapor deposition (CVD), can obtain β′-In2Se3/WSe2 subferroelectric by direct growth SiO2/Si substrate FePNJ transition. Modulated synergistic effect polarization electric field built-in field, exhibits significantly enhanced highly tunable synaptic effects (memory retention >2500 s >8 multilevel current states under single optical/electrical pulses), along power consumption down to atto-joule levels. Utilizing these photoelectric properties, constructed all-ferroelectric reservoir computing system, comprising both readout networks, handwritten digit recognition. We also created a system through gate-voltage-modulated relaxation time scale which efficiently detect motions in 1 100 km h–1 speed range.
Language: Английский
Citations
0
Organic Electronics, Journal Year: 2025, Volume and Issue: unknown, P. 107242 - 107242
Published: March 1, 2025
Language: Английский
Citations
0
Advanced Science, Journal Year: 2025, Volume and Issue: unknown
Published: May 20, 2025
Abstract An asymmetric dual‐gate heterointerface‐regulated artificial synapse (HRAS) is developed, utilizing a main gate with distinct ion concentrations and lateral to receive synaptic pulses, through dielectric coupling ionic effects, formed indium tin zinc oxide (ITZO) dual‐interface channels that allow precise control over channel charge, thereby simulating multi‐level coordinated actions of dual‐neurotransmitters. The modulation the significantly regulates achieving intricate interplay among inhibition/enhancement short‐/long‐term plasticity at scale for first time. This enables HRAS device simulate frequency‐dependent image filtering spike number‐dependent dynamic visual persistence. By combining temporal inputs modulation, harnesses spatiotemporal properties bio‐inspired cryptographic applications, offering versatile device‐level platform secure information processing. Furthermore, novel input neural network architecture based on has been proposed, which aids in weight update demonstrates enhanced recognition capabilities tasks, highlighting its role computing.
Language: Английский
Citations
0
Nanomaterials, Journal Year: 2024, Volume and Issue: 14(19), P. 1573 - 1573
Published: Sept. 29, 2024
The traditional computer with von Neumann architecture has the characteristics of separate storage and computing units, which leads to sizeable time energy consumption in process data transmission, is also famous “von wall” problem. Inspired by neural synapses, neuromorphic emerged as a promising solution address problem due its excellent adaptive learning parallel capabilities. Notably, 2016, researchers integrated light into computing, inspired extensive exploration optoelectronic all-optical synaptic devices. These optical devices offer obvious advantages over all-electric devices, including wider bandwidth lower latency. This review provides an overview research background on discusses their implementation principles different scenarios, presents application concludes prospects for future developments.
Language: Английский
Citations
1
Chemical Engineering Journal, Journal Year: 2024, Volume and Issue: 501, P. 157503 - 157503
Published: Nov. 8, 2024
Language: Английский
Citations
1
Journal of Colloid and Interface Science, Journal Year: 2024, Volume and Issue: 678, P. 325 - 335
Published: Sept. 5, 2024
Language: Английский
Citations
0