Abstract:Artificial synapse networks capable of massively parallel computing and mimicking biological neural networks can potentially improve the processing efficiency of existing information technologies. Semiconductor devices functioning as excitatory and inhibitory synapses are crucial for developing intelligence systems, such as traffic control systems. However, achieving reconfigurability between two working modes (inhibitory and excitatory) and bilingual synaptic behavior in a single transistor remains challenging. This study successfully mimics a bilingual synaptic response using an artificial synapse based on an ambipolar floating gate memory comprising tungsten selenide (WSe2)/hexagonal boron nitride (h-BN)/ molybdenum telluride (MoTe2). In this WSe2/h-BN/MoTe2 structure, ambipolar semiconductors WSe2 and MoTe2 are inserted as channel and floating gates, respectively, and h-BN serves as the tunneling barrier layer. Using either positive or negative pulse amplitude modulations at the control gate, this device with bipolar channel conduction produced eight distinct resistance states. Based on this, we experimentally projected that we could achieve 490 memory states (210 hole-resistance states + 280 electron-resistance states). Using the bipolar charge transport and multistorage states of WSe2/h-BN/MoTe2 floating gate memory, we mimicked reconfigurable excitatory and inhibitory synaptic plasticity in a single device. Furthermore, the convolution neural network formed by these synaptic devices can recognize handwritten digits with an accuracy of >92%. This study identifies the unique properties of heterostructure devices based on two-dimensional materials as well as predicts their applicability in advanced recognition of neuromorphic computing.

A memristive deep belief neural network based on silicon synapses

Reconfigurable Artificial Synapse Based on Ambipolar Floating Gate Memory

Memristive brain-like computing

Flexible Artificial Memristive Synapse Constructed from Solution‐Processed MgO–Graphene Oxide Quantum Dot Hybrid Films

A sub-1-volt analog metal oxide memristive-based synaptic device for energy-efficient spike-based computing systems

A Flexible Artificial Synapse for Neuromorphic System

Nanoscale Memristor Device as Synapse in Neuromorphic Systems

Artificial synapses with a sponge-like double-layer porous oxide memristor

Artificial Synapses Based on an Optical/Electrical Biomemristor

Diverse long-term potentiation and depression based on multilevel LiSiOx memristor for neuromorphic computing

Energy-Efficient Artificial Synapses Based on Oxide Tunnel Junctions

Memristor based on α-In2Se3 for emulating biological synaptic plasticity and learning behavior

CMOS Back-End Compatible Memristors for in Situ Digital and Neuromorphic Computing Applications.

A Memristor Neural Network Based on Simple Logarithmic-Sigmoidal Transfer Function with MOS Transistors

Flexible Low‐Voltage MXene Floating‐Gate Synaptic Transistor for Neuromorphic Computing and Cognitive Learning

Single Neuromorphic Memristor closely Emulates Multiple Synaptic Mechanisms for Energy Efficient Neural Networks

Artificial Neurons Based on Ag/V2C/W Threshold Switching Memristors

Diverse Long-Term Potentiation and Depression Based on Multilevel LiSiOxmemristor for Neuromorphic Computing.

A CMOS spiking neuron for dense memristor-synapse connectivity for brain-inspired computing

Memristors with Diffusive Dynamics As Synaptic Emulators for Neuromorphic Computing

Neuron-inspired flexible memristive device on silicon (100)