Weak Light‐Stimulated Synaptic Transistors Based on MoS₂/Organic Semiconductor Heterojunction for Neuromorphic Computing

Abstract

Photonic synapses are expected to play an important role in implementing brain‐like computing owing to the wide bandwidth and low mutual interference of optical signals. Herein, photonic synaptic transistors based on inorganic semiconductor molybdenum disulfide (MoS2) and organic semiconductor heterojunction with adjustable short‐term/long‐term plasticity are proposed. Benefitting from the outstanding photosensitive characteristics originating from the heterojunction, the devices have the capability of weak light detection and can exhibit distinct synaptic responses even under an ultraweak light intensity of 40 nW cm−2, which is considerably lower than that of most previously reported photonic synaptic transistors. Furthermore, a low energy consumption of 0.4 fJ per synaptic event can be obtained at a low operating voltage of 0.001 V, lower than that required for a single synaptic event observed in human brains. In addition, the devices can implement high‐pass filtering function and illustrate the potential in image sharpening processing. More significantly, logic functions as well as associative learning behavior are dramatically simulated through all‐optical stimulation. This work demonstrates a feasible approach to developing multifunctional photonic synaptic transistors based on inorganic/organic semiconductor heterojunction for neuromorphic computing.

Thermally and Mechanically Stable Perovskite Artificial Synapse as Tuned by Phase Engineering for Efferent Neuromuscular Control

“…As the number of pulses increases gradually, the growth increment of the current spikes begins to flatten (Figure S10); this trend is similar to plasticity changes observed with varying pulse widths (Figure S11). Conversely, the gain increases with the frequency of pulses (Figure S12), mirroring the frequency response observed in electrolyte/semiconductor devices. − A larger synaptic gain can be employed in high-pass filtering to preprocess the image to sharpen the edge of the image (Figure e). − Here, the contours of the image are deepened by using the cutoff frequency of 4.15 Hz. Higher frequency gain and lower cutoff frequency (<10 Hz) are good for edge sharpening.…”

Recent progress of organic photonic synaptic transistors for artificial visual systems: structure configuration and innovative applications

“…At an operating voltage of 0.001 V, the energy consumption per synapse event is as low as 0.4 fJ. 85 The successful implementation of this work provided a feasible approach for the development of inorganic/organic heterojunction OPSTs in the field of low energy consumption and ultra-sensitive multifunctional neuromorphic computing. This layer-by-layer stacking method for preparing organic–inorganic hybrid heterojunctions can cause problems such as the introduction of impurities during the preparation process, complex preparation techniques, and low transfer success rates.…”

Transistor-Based Synaptic Devices for Neuromorphic Computing

“…Various synaptic functions, such as PPF and LTP, could be mimicked in this structure. Many synaptic devices have been investigated substantially and shown to mimic synaptic functions based on similar mechanisms [65][66][67][68]. It should be mentioned that charge-trap-flash-based synaptic transistors were successfully fabricated based on a silicon nitride film as the charge-trapping layer and a silicon film as the channel layer in transistor devices [69,70].…”