Abstract:Optoelectronic synapses integrating synaptic and optical-sensing functions exhibit large advantages in neuromorphic computing for visual information processing and complex learning, recognition, and memory in an energy-efficient way. However, electric stimulation is still essential for existing optoelectronic synapses to realize bidirectional weight-updating, restricting the processing speed, bandwidth, and integration density of the devices. Herein, a two-terminal optical synapse based on a wafer-scale pyrenyl graphdiyne/graphene/PbS quantum dot heterostructure is proposed that can emulate both the excitatory and inhibitory synaptic behaviors in an optical pathway. The simple device architecture and low-dimensional features of the heterostructure endow the optical synapse with robust flexibility for wearable electronics. This optical synapse features a linear and symmetric conductance-update trajectory with numerous conductance states and low noise, which facilitates the demonstration of accurate and effective pattern recognition with a strong fault-tolerant capability even at bending states. A series of logic functions and associative learning capabilities have been demonstrated by the optical synapses in optical pathways, significantly enhancing the information processing capability for neuromorphic computing. Moreover, an integrated visible information sensing memory processing system based on the optical synapse array is constructed to perform real-time detection, <i>in situ</i> image memorization, and distinction tasks. This work is an important step toward the development of optogenetics-inspired neuromorphic computing and adaptive parallel processing networks for wearable electronics.The Supporting Information is available free of charge at <a class="ext-link" href="/doi/10.1021/acsnano.0c08921?goto=supporting-info">https://pubs.acs.org/doi/10.1021/acsnano.0c08921</a>.Experimental details including preparation of PbS QDs, 1,3,6,8-TEP monomer and device fabrication; Fermi level and energy band alignment of the Pyr-GDY/Gr/PbS-QDs heterosturcture; characterization, electrical/optoelectrical measurements, and flexibility of the heterostructure; calculation of the nonlinearity, asymmetricity, and write noise; process flow of the integrated visible information sensing memory processing system (<a class="ext-link" href="/doi/suppl/10.1021/acsnano.0c08921/suppl_file/nn0c08921_si_001.pdf">PDF</a>)This article has not yet been cited by other publications.

Programmable Superconducting Optoelectronic Single-Photon Synapses with Integrated Multi-State Memory

Demonstration of Superconducting Optoelectronic Single-Photon Synapses

Optoelectronic synapses based on hot-electron-induced chemical processes.

AlGaN/GaN-Based Optoelectronic Synaptic Devices for Neuromorphic Computing

Optically Stimulated Synaptic Devices Based on the Hybrid Structure of Silicon Nanomembrane and Perovskite.

Circuit designs for superconducting optoelectronic loop neurons

An optoelectronic synapse based on α-In 2 Se 3 with controllable temporal dynamics for multimode and multiscale reservoir computing

Considerations for neuromorphic supercomputing in semiconducting and superconducting optoelectronic hardware

A High-Speed 2D Optoelectronic In-Memory Computing Device with 6-Bit Storage and Pattern Recognition Capabilities

Multifunctional 2D MoS2 Optoelectronic Artificial Synapse with Integrated Arithmetic and Reconfigurable Logic Operations for In‐Memory Neuromorphic Computing Applications

Scalable Synaptic Transistor Memory from Solution‐Processed Carbon Nanotubes for High‐Speed Neuromorphic Data Processing

Artificial Synapse with Mnemonic Functionality using GSST-based Photonic Integrated Memory

All-Optical Synapse with Directional Coupler Structure Based on Phase Change Material

Two‐Terminal Perovskite Optoelectronic Synapse for Rapid Trained Neuromorphic Computation with High Accuracy

Combined optical and electrical control of a low-power consuming ( ∼ fJ) two-terminal organic artificial synapse for associative learning and neuromorphic applications

Vertically Integrated Dual-memtransistor Enabled Reconfigurable Heterosynaptic Sensorimotor Networks and In-memory Neuromorphic Computing

Multifunctional Two-Terminal Optoelectronic Synapse Based on Zinc Oxide/Poly(3-hexylthiophene) Heterojunction for Neuromorphic Computing

A Perovskite-Based Artificial Photonic Synapse with Visible Light Modulation and Ultralow Current for Neuromorphic Computing

Optoelectronic Perovskite Synapses for Neuromorphic Computing

Large-Scale and Flexible Optical Synapses for Neuromorphic Computing and Integrated Visible Information Sensing Memory Processing

A 45nm CMOS Neuromorphic Chip with a Scalable Architecture for Learning in Networks of Spiking Neurons