Wood-Derived Nanopaper Dielectrics for Organic Synaptic Transistors

Abstract

The use of biocompatible and biodegradable materials in electronic devices can be an important trend in the development of the next-generation green electronics. In addition, by integrating the advantages of low power consumption, low-cost processing, and flexibility, organic synaptic devices will be promising elements for the construction of brain-inspired computers. However, previously reported electrolyte-gated synaptic transistors are mainly made of non-biocompatible and non-biodegradable electrolytes. Woods are widely considered as one kind of sustainable and renewable materials. We found that wood-derived cellulose nanopapers have ionic conductivity and, therefore, can be used as dielectric materials for organic synaptic transistors. The fabricated wood-derived cellulose nanopapers exhibit decent ionic conductivity of 7.3 × 10–4 S m–1 and a high lateral coupling effective capacitance of 18.65 nF cm–2 at 30 Hz. The laterally coupled organic synaptic transistors using wood-derived cellulose nanopapers as the dielectric layer present excellent transistor performances at the operating voltage below 1.5 V. More significantly, some important synaptic behaviors, such as excitatory postsynaptic current, signal-filtering characteristics, and dendritic integration are successfully simulated in our synaptic transistors. Because the development of electronic devices with biocompatible and biodegradable materials is essential, this work may inspire new directions for the development of “green” neuromorphic electronics.

The Design of 3D‐Interface Architecture in an Ultralow‐Power, Electrospun Single‐Fiber Synaptic Transistor for Neuromorphic Computing

“…Nowadays, various types of electronic devices have been developed to simulate synaptic properties, such as phase‐change memories,13 synaptic transistors,14–16 and memristors 17–19. Synaptic behaviors have been successfully mimicked by these devices.…”

“…[9] With the technology based on complementary metal-oxide semiconductor (CMOS), at least 10 transistors are needed to implement one synapse, which is too large for an artificial neuromorphic system. [10][11][12] Therefore, developing a single device to achieve synaptic behavior is an important improvement in the basic hardware for the artificial neuromorphic system.Nowadays, various types of electronic devices have been developed to simulate synaptic properties, such as phase-change memories, [13] synaptic transistors, [14][15][16] and memristors. [17][18][19] Synaptic behaviors have been successfully mimicked by these devices.…”

Ionotronic Neuromorphic Devices for Bionic Neural Network Applications

“…Specifically, the release and retreat of neurotransmitters lead to the short‐term depolarization or hyperpolarization of the postsynaptic membrane, i.e., short‐term synaptic plasticity . Thus, based on the EDL mechanism, short‐term synaptic plasticity can be mimicked on EDLTs . Meantime, electrochemical doping/de‐doping can also cause ionic species to trap in the channel and lead to the long‐term changes in channel conductances.…”

“…[50,53] Thus, based on the EDL mechanism, short-term synaptic plasticity can be mimicked on EDLTs. [54][55][56][57][58] Meantime, electrochemical doping/de-doping can also cause ionic species to trap in the channel and lead to the long-term changes in channel conductances. Thus, the electrochemical doping/dedoping mechanism in ECTs is similar to the long-term changes in the synaptic weights.…”

Recent Advances in Transistor‐Based Artificial Synapses

“…Therefore, transistors may be more suitable for simulating synaptic functions than other types of devices, especially for simulating concurrent learning and dendrites integration that require multiterminal operation. Recent years have witnessed the increasing research interests in developing transistor‐based artificial synapses, however, this field is still in its infancy.…”