Reconfigurable 2D WSe₂‐Based Memtransistor for Mimicking Homosynaptic and Heterosynaptic Plasticity

Abstract

The mimicking of both homosynaptic and heterosynaptic plasticity using a high‐performance synaptic device is important for developing human‐brain–like neuromorphic computing systems to overcome the ever‐increasing challenges caused by the conventional von Neumann architecture. However, the commonly used synaptic devices (e.g., memristors and transistors) require an extra modulate terminal to mimic heterosynaptic plasticity, and their capability of synaptic plasticity simulation is limited by the low weight adjustability. In this study, a WSe2‐based memtransistor for mimicking both homosynaptic and heterosynaptic plasticity is fabricated. By applying spikes on either the drain or gate terminal, the memtransistor can mimic common homosynaptic plasticity, including spiking rate dependent plasticity, paired pulse facilitation/depression, synaptic potentiation/depression, and filtering. Benefitting from the multi‐terminal input and high adjustability, the resistance state number and linearity of the memtransistor can be improved by optimizing the conditions of the two inputs. Moreover, the device can successfully mimic heterosynaptic plasticity without introducing an extra terminal and can simultaneously offer versatile reconfigurability of excitatory and inhibitory plasticity. These highly adjustable and reconfigurable characteristics offer memtransistors more freedom of choice for tuning synaptic weight, optimizing circuit design, and building artificial neuromorphic computing systems.

Neuromorphic Synapses with High Switching Uniformity and Multilevel Memory Storage Enabled through a Hf-Al-O Alloy for Artificial Intelligence

“…Memristor-based neuromorphic systems challenge conventional computers because of their basic combination of storage and computation. , In terms of structure and behavior, memristive devices have properties similar to biological neuromorphic synapses, making them promising contenders for simulating the switching synaptic weights in biological approaches, as their conductivity progressively adjusts with the switch in charge or flow. − The concentration of neurotransmitters at biological synapses weakens or strengthens the connection between neurons. Correspondingly, the movement of cations or vacancies linked with the medium and electrode materials causes potentiation and depression plasticity in memristors. − To put it another way, the electrode media and materials have a big impact on memristor performance features including changeable conductivity and retention. , As a way to develop electrical synapses, the development of memristive devices with progressively modifying conductivity and a great lifespan is considered to develop electrical synapses. Synapses in the brain have established a wide range of memory lifetimes, ranging from a few seconds to decades.…”

Bio‐Inspired 3D Artificial Neuromorphic Circuits

“…[ 234–242 ] In additional to single gate configuration and MOSs as the functional material, multiple‐gate synaptic transistors based on 2D materials have drawn tremendous research efforts as well. [ 243–249 ] As illustrated in Figure 9b, Chen et al. recently combined a double‐gate configuration with intrinsic polarity of 2D materials to perform logic operations in a single synaptic device, which realizes XNOR gate using ambipolar tungsten diselenide (WSe 2 ), NOR gate with black phosphorus, as well as OR and AND gates via molybdenum disulfide (MoS 2 ).…”

“…[234][235][236][237][238][239][240][241][242] In additional to single gate configuration and MOSs as the functional material, multiple-gate synaptic transistors based on 2D materials have drawn tremendous research efforts as well. [243][244][245][246][247][248][249] As illustrated in Figure 9b, Chen et al recently combined a double-gate configuration with intrinsic polarity of 2D materials to perform logic operations in a single synaptic device, which realizes XNOR gate using ambipolar tungsten diselenide (WSe 2 ), NOR gate with black phosphorus, as well as OR and AND gates via molybdenum disulfide (MoS 2 ). [250] Yao et al integrated graphdiyne as a lithium ion trapping layer to prevent back-diffusion of accumulated ions into the LiClO 4 electrolyte, thereby enhancing the retention characteristic of the nonvolatile electrolyte-gated synaptic transistors, as demonstrated in Figure 9c.…”

Advances in Emerging Photonic Memristive and Memristive‐Like Devices

“…The development of 2D materials in the field of memristors has a long history due to the characteristics that are beneficial to the resistance switching behaviors, such as large specific surface area, adjustable band‐gap, and high mobility. [ 165 , 166 , 167 ] A large number of researchers have found the resistive switching behavior in 2D materials, such as graphene and its derivatives, [ 128 , 168 , 169 ] molybdenum disulfide, [ 170 , 171 ] tungsten disulfide, [ 172 , 173 ] and hexagonal boron nitride, [ 174 , 175 ] while the 2D materials are not performed in photonic memristive devices until its concept is proposed in recent years. Even though there are many methods to prepare 2D materials, such as mechanical peel‐off, chemical vapor deposition and sol–gel process, the performances of the 2D materials prepared by various methods are different in memristor due to the limitations of the preparation conditions, which means that the optimization potential of the same 2D material in the photonic memristive and memristive‐like devices is enormous.…”