In this study, we introduce a lithium (Li) ion-based three-terminal (3-T) synapse device using WO x as a channel. Our study reveals a key stoichiometry of WO 2.7 for excellent synaptic characteristics that is related to Li-ion diffusivity. The open-lattice structure formed by oxygen deficiency promoted Li-ion injection and diffusion. The optimized stoichiometry and improved Li-ion diffusivity were confirmed by x-ray photoelectron spectroscopy analysis and cyclic voltammetry, respectively. Furthermore, the transient conductance change that inevitably occurs in ion-based synaptic transistors was resolved by applying a two-step voltage pulse scheme. As a result, we achieved a symmetric and linear weight-update characteristic with reduced program/ erase operation time.

arrays, are being actively studied as synaptic devices for use in neuromorphic computing. However, the need for a new type of synaptic device has emerged, owing to write-disturbance issues [6] and the difficulty of obtaining sufficient synaptic properties. [7,8] Among various synaptic memory devices, ion-based synaptic transistors are being studied as promising next-generation solutions based on nearideal synaptic behaviors, low-power operations, reasonable retention, and excellent endurance characteristics.Ion-based synaptic transistors mainly use protons and Li ions as doping elements, which can be easily moved by external biases, owing to their small mass. [9] The electrical conductivity changes linearly according to the concentration of ions injected into the channel [10] and by using it as a synaptic weight ideal training characteristics can be achieved. In the case of a protonbased synaptic transistors, low-voltage operation is possible owing to the small atomic size (ion radius 0.04 Å), and excellent weight-update linearity and endurance characteristics are obtained. [11][12][13][14][15] However, to provide excellent proton-hopping characteristics, polymer materials, such as poly(3,4-ethylenedioxythiophene) polystyrene sulfonate, are used as channels and electrolytes. Hence, it is difficult to utilize conventional complementary metal-oxide-semiconductor (CMOS) technology. [14,15] The structure and operation mechanism of a Li-based synaptic transistor is very similar to that of a Li-ion battery. Li ions are injected or extracted into a channel that acts as a cathode, and its electrical conductivity (i.e., synaptic weight) is determined by the concentration of Li ions in the channel. [10] Li-ion-based synaptic transistors have excellent synaptic properties owing to their small atomic size (ion radius 0.9Å), [16][17][18][19][20] but there are disadvantages, such as Li dendrite generation and open-circuit potential issues. [16,21] Accordingly, on the basis of various prior studies in the RRAM and solid oxide fuel cell fields, research on oxygen ion-based synaptic transistors (OISTs) that can utilize conventional CMOS technology is being actively conducted. [22][23][24][25][26][27][28] The OIST uses transition metal oxide (TMO) as a channel, which can change its electrical conductivity depending on the extraction/injection of oxygen ions. In a recently published

In this study, we investigate a proton-based three-terminal (3-T) synapse device to realize linear weight-update and I-V linearity characteristics for neuromorphic systems. The conductance states of the 3-T synapse device can be controlled by modulating the proton concentration in the WO x channel. Therefore, we estimate the dynamic change of proton concentration in the channel region, which directly affects synaptic behaviors. Our findings indicate that the supply of an excess number of protons from the SiO 2 -H electrolyte and low proton diffusivity in the WO x channel result in asymmetric and non-linear weight-update characteristics. In addition, though the linear I-V characteristics can be obtained using non-stoichiometric WO x , we observe that significant oxygen deficiency in the channel region increases the operating current levels. Thus, based on this information, we introduce optimized conditions of each component in the 3-T synapse device and shape of the gate voltage pulses. As a result, an excellent classification accuracy is achieved using linear weight-update and I-V linearity characteristics under optimized device and pulse conditions.

engineering advancements are speeding up the development of electrolyte materials with lower activation energy barriers and channel materials with lower redox energy barriers. [1a,4] Proton based ECRAM (H-ECRAM) is potentially advantageous because proton (H + ) is a smaller and more rapidly diffusing ion than Li + and O 2− . However, the memory state-retention times of H-ECRAM far remained relatively short with cycling instability. [5] Due to H-ECRAM poor memory state-retention, inference accuracy degrades over time. The poor retention in H-ECRAM is caused by self-discharge of intercalated ions from the channel layer due to existence of non-zero open circuit potential (OCP), which is mainly occurred in polymer-based H-ECRAM. [6] When the gate circuit opens during the read operation, the electrically conductive electrolyte is unable to halt the backflow of electrons, resulting in non-zero OCP and subsequent self-discharge of H + ions. Fuller et al. connected a selector in series with the H-ECRAM gate terminal to address the OCP issue, forming a one-selector-one-H-ECRAM (1S1E) structure that isolates the device and prevents leakage current. [7] In our opinion, the memory state-retention and cycling stability of H-ECRAM can be improved by developing proton-conducting solid electrolytes with electron-blocking properties to lower the self-discharge issue. However, the unavailability of a CMOS-compatible proton-conducting solid electrolyte is the main obstacle. All H-ECRAM presently relied on electrolytes that either cannot be integrated and scale down, such as polymer, [8] ionic liquid, [9] ionic gel, [10] organic material. [5] Herein, atomically thin single-layer hexagonal boron nitride (hBN) is integrated into H-ECRAM as a proton-conducting solid-state electrolyte. Atomically thin 2D material has not yet been exploited in prior research for the purpose of improving memory state-retention and cycling stability of ECRAM devices. Recent research has proven that a few 2D materials exhibit ion transport properties both experimentally and theoretically. [11] Hexagonal boron nitride (hBN) single-layers have been evaluated as a possible material for developing novel ionic transport layers. [12] The honeycomb structure of 2D h-BN is composed of alternating boron and nitrogen atoms. [13] It exhibits superior chemical and thermal stability, as well as mechanical strength. [14] ProtonThe first report on ion transport through atomic sieves of atomically thin 2D material is provided to solve critical limitations of electrochemical randomaccess memory (ECRAM) devices. Conventional ECRAMs have random and localized ion migration paths; as a result, the analog switching efficiency is inadequate to perform in-memory logic operations. Herein ion transport path scaled down to the one-atom-thick (≈0.33 nm) hexagonal boron nitride (hBN), and the ionic transport area is confined to a small pore (≈0.3 nm 2 ) at the single-hexagonal ring. One-atom-thick hBN has ion-permeable pores at the center of each hexagonal ring due to weakened elect...

Oxygen-based electrochemical random-access memories (O-ECRAMs) are promising synaptic devices for neuromorphic applications because of their near-ideal synaptic characteristics and compatibility with complementary metal–oxide–semiconductor processes. However, the correlation between material parameters and synaptic properties of O-ECRAM devices has not yet been elucidated. Here, we propose the critical design parameters to fabricate an ideal ECRAM device. Based on the experimental data and simulation results, it is revealed that consistent ion supply from the electrolyte and rapid ion diffusion in the channel are critical factors for ideal synaptic characteristics. To optimize these parameters, crystalline WO2.7 exhibiting fast ion diffusivity and ZrO1.7 exhibiting an appropriate ion conduction energy barrier (1.1 eV) are used as a channel and an electrolyte, respectively. As a result, synaptic characteristics with near-ideal weight-update linearity in the nanosiemens conductance range are achieved. Finally, a selector-less O-ECRAM device is integrated into a 2 × 2 array, and high recognition accuracy (94.83%) of the Modified National Institute of Standards and Technology pattern is evaluated.