Search

We demonstrate that an analog synaptic weight update can be achieved using identical pulses via gradual resistance switching of memristors based on sputter-deposited InGaZnO (IGZO) films. When a positive voltage is applied to the IGZO memristor, oxygen vacancies (V o) are ionized and electrons that serve as dopants are generated. This lowers the Schottky barrier formed at the interface and decreases the resistance. However, a negative voltage provides electrons from the Pd electrode, annihilating the ionized vacancies. As the barrier is reconstructed, the IGZO memristor exhibits a high-resistance state (HRS). As V o plays a crucial role in switching, the effect of the radio frequency (RF) power on the memristor during the IGZO deposition is examined. A physical analysis reveals that the V o concentration in the IGZO is proportional to the RF power, inducing a leaky HRS. However, at the expense of a reduced memory window in the IGZO deposited at a higher power of 200 W, the resultant low-resistance state is reliable over time. The strengthened retention allows the synaptic weights to be linearly decreased even by consecutive negative pulses. Hence, when the IGZO memristors are used for synaptic elements to construct neural networks, a higher recognition accuracy for the MNIST dataset is achieved. [ABSTRACT FROM AUTHOR]

Memristive devices are promising candidates for analog computing applications such as neuromorphic computation. Larger dynamic ranges and more sufficient multilevel states can enable the significant development of memristor‐based utilizations. Herein, a method to improve the analog switching performance of memristors through a hybrid tuning (coarse and fine tuning) of two sub‐filaments is demonstrated. The creation of sub‐filaments inside the dielectric switching layer is realized by deploying Pt metal islands in the switching layer. Given the different material stack configurations of the two sub‐filaments, they exhibit different switching properties to play the roles of coarse and fine tuning respectively in the memristor. Based on the above mechanism, a Pt/Ta/Al2O3/Pt island/Al2O3‐x/TiOy/Al2O3‐x/Pt memristor is proposed and fabricated. Through the hybrid tuning of two sub‐filaments, a combined dynamic range of 600 Ω to 50 kΩ is achieved. Compared to the reference Pt/Ta/Al2O3/Pt memristors (dynamic range: 600 Ω to 8 kΩ), both dynamic range and multilevel resistance states are increased significantly. Meanwhile, the energy efficiency is improved because the resistance of tunable states can be set to larger values. Furthermore, this mechanism can be incorporated into various existing memristors to improve their dynamic range and multilevel states, which extensively enriches the applications of memristors. [ABSTRACT FROM AUTHOR]

This study proposes a high-performance organic–inorganic hybrid memristor for the development of neuromorphic devices in the memristor-based artificial synapse. The memristor consists of a solid polymer electrolyte (SPE) chitosan layer and a titanium oxide (TiOx) layer grown with a low-thermal-budget, microwave-assisted oxidation. The fabricated Ti/SPE–chitosan/TiOx/Pt-structured memristor exhibited steady bipolar resistive switching (BRS) characteristics and demonstrated excellent endurance in 100-cycle repetition tests. Compared to SPE–chitosan memristors without a TiOx layer, the proposed organic–inorganic hybrid memristor demonstrated a higher dynamic range and a higher response to pre-synaptic stimuli such as short-term plasticity via paired-pulse facilitation. The effect of adding the TiOx layer on the BRS properties was examined, and the results showed that the TiOx layer improved the chemical and electrical superiority of the proposed memristor synaptic device. The proposed SPE–chitosan organic–inorganic hybrid memristor also exhibited a stable spike-timing-dependent plasticity, which closely mimics long-term plasticity. The potentiation and depression behaviors that modulate synaptic weights operated stably via repeated spike cycle tests. Therefore, the proposed SPE–chitosan organic–inorganic hybrid memristor is a promising candidate for the development of neuromorphic devices in memristor-based artificial synapses owing to its excellent stability, high dynamic range, and superior response to pre-synaptic stimuli. [ABSTRACT FROM AUTHOR]

Oxide-based memristors have emerged as a promising electronic device for high-density memory and neuromorphic applications. In our study, we explored the tunable analog switching and biological synaptic functions of a Pt/ZnO/Al 2 O 3 /TaN memristive device. Using transmission electron microscopy (TEM) and x-ray photoelectron spectroscopy (XPS), we confirmed the presence of a TaO x N y interface layer at the anode contact, believed to play a critical role in resistance transitions. The memristive device showed excellent performance, including a stable and reproducible analog switching memory with a low operating voltage (μ = ̶ 2.0 / + 1.7 V), good cycling endurance (2 × 102), a high on/off ratio (>103), and retention up to 104 s at 85 °C. Additionally, multi-state resistances were achieved by varying the reset voltage, enabling the creation of neuromorphic synapses and high-density memories. Direct-current mode set and reset transitions showed multi-state resistance changes similar to potentiation and depression behaviors in biological synapses. Further simulations, including long-term potentiation (LTP) and long-term depression (LTD), paired pulse facilitation (PPF), and convolutional neural network (CNN) simulations for handwritten digits, showed an accuracy of 86.5%. These results indicate that the memristive device is highly suitable for use in high-density memory and brain-inspired computer systems. [ABSTRACT FROM AUTHOR]

The two-terminal memristor is a promising neuromorphic artificial electronic device, mirroring biological synapses in structure and replicating various synaptic functions. Despite its advantages, challenges in achieving high reliability, gradual switching, and low energy consumption hinder progress in neuromorphic devices. This study explores electronic synapses and simulates analog switching in a Pt/TiN/ZnO/SnO 2 /ZnO/Pt multilayer (ML) configuration, featuring a ∼3 nm SnO 2 layer between ZnO layers. Results show enhanced cycling endurance (more than 250 cycles), resistance window (102), tunable synaptic plasticity, and multilevel switching. ML memristors exhibit low coefficient of variation (4.5 %) in set voltage, low energy consumption (set = ∼0.12 nj, reset = ∼0.1 nj), and fast switching speeds (set = 300 ns, reset = 200 ns), suitable for high-density memory and neuromorphic systems. They successfully emulate synaptic functions, including paired-pulse facilitation (PPF), spike voltage-dependent plasticity (SVDP), spike width-dependent plasticity (SWDP), spike frequency-dependent plasticity (SFDP), and post-tetanic potentiation (PTP). Modulating pulse amplitude and width achieves multilevel conductance in long-term potentiation (LTP) and long-term depression (LTD). Using nonlinear conductance data, a 96.5 % image pattern recognition accuracy is achieved in a deconvolution neural network (DNN) simulation. These results highlight the ML memristor's potential in efficient neuromorphic computing systems. • Multilayer memristor demonstrates improved cycling (∼300 cycles) and a wide resistance range (∼10^2), enhancing reliability for neuromorphic system. • Memristor device shows low energy consumption (set = ∼0.12 nJ, reset = ∼0.1 nJ) and fast switching (set = 300 ns, reset = 200 ns). • Our study demonstrates tunable synaptic plasticity, enabling the replication of key synaptic functions like PPF and SVDP, essential for neuromorphic computing. • Pattern recognition accuracy of up to 96.5 % is achieved using DNN simulation, revealing the memristor potential for efficient neuromorphic computing. [ABSTRACT FROM AUTHOR]

Mobile species and matrix materials in ion motion‐mediated memristors predominantly determine the switching characteristics and device performance. As a result of exploring a new type of mobile species, a Ru ion‐mediated electrochemical metallization‐like memristor with an amorphous oxide matrix is recently suggested to achieve a low switching current, voltage, and good retention simultaneously. Although the ion migration of Ru in the oxide matrix is previously confirmed, no in‐depth study on how the crystallinity of the oxide matrix influences the Ru ion motion and switching characteristics has not been reported. Therefore, in this study, the crystallinity‐dependent resistive switching behavior of the Pt/HfO2/Ru structure device is investigated. With the crystallized HfO2 layer, the preferred Ru ion migration through the grain boundaries occurs owing to the enhanced ion mobility, resulting in a high switching current (≈100 µA) with continuous metallic Ru conducting filaments. The discontinuous conducting filaments with amorphous HfO2 exhibit a low switching current. In addition, highly linear and symmetric conductance modulation properties are achieved, and over 91.5% accuracy in the Mixed National Institute of Standards and Technology (MNIST) pattern recognition test is demonstrated. [ABSTRACT FROM AUTHOR]

This study proposes a high-performance organic–inorganic hybrid memristor for the development of neuromorphic devices in the memristor-based artificial synapse. The memristor consists of a solid polymer electrolyte (SPE) chitosan layer and a titanium oxide (TiOx) layer grown with a low-thermal-budget, microwave-assisted oxidation. The fabricated Ti/SPE–chitosan/TiOx/Pt-structured memristor exhibited steady bipolar resistive switching (BRS) characteristics and demonstrated excellent endurance in 100-cycle repetition tests. Compared to SPE–chitosan memristors without a TiOx layer, the proposed organic–inorganic hybrid memristor demonstrated a higher dynamic range and a higher response to pre-synaptic stimuli such as short-term plasticity via paired-pulse facilitation. The effect of adding the TiOx layer on the BRS properties was examined, and the results showed that the TiOx layer improved the chemical and electrical superiority of the proposed memristor synaptic device. The proposed SPE–chitosan organic–inorganic hybrid memristor also exhibited a stable spike-timing-dependent plasticity, which closely mimics long-term plasticity. The potentiation and depression behaviors that modulate synaptic weights operated stably via repeated spike cycle tests. Therefore, the proposed SPE–chitosan organic–inorganic hybrid memristor is a promising candidate for the development of neuromorphic devices in memristor-based artificial synapses owing to its excellent stability, high dynamic range, and superior response to pre-synaptic stimuli.

Leaky integrate and fire (LIF) neurons are critical units for constructing a spiking neural network, in which neurons communicate with each other using spikes via synapses. Memristors, due to its specific nonlinear characteristics, are frequently introduced to emulate partial functions of LIF neurons for simplifying the circuit complexity, either the integration process or the fire action. Usually, a relatively complicated peripheral circuit needs to be engineered to assist the memristive device for complete emulation for biological neurons, which certainly would hinder the integration potential. Herein, we fabricated a stacked memristive device possessing both analog and threshold switching behaviors for constructing an artificial LIF neuron. Thus, the integration and fire functions were both accomplished within this single nanoscale device. In addition, the key neuronic functional of a biological neuron, including all-or-nothing spiking, threshold spiking, a refractory period, and strength-modulated frequency response were all successfully mimicked. The results demonstrate that the fabricated stacked memristor-based LIF neurons have great potential to construct high-density spiking neural network for neuromorphic computing. [ABSTRACT FROM AUTHOR]

Memristor is considered as one of the key components of neuromorphic hardware aiming to overcome the limitations of classical von Neumann computers, but the continuous control of the device conduction remains a challenge. Random and stochastic dynamics of weak conducting filaments (CF) hinder the precise control of the resistance states, particularly in the conductive bridge random access memory (CBRAM). The problem of random and stochastic resistive switching of the Cu‐based CBRAM is largely mitigated by inserting the Cu‐cone as a cation source into the memory cell (Cu‐cone device). The electric field concentrated on the Cu‐cone induces stronger CF with a limited number near the tip region, making the device feature distinctive from the planar Cu CBRAM. The Cu‐cone device exhibits enhanced resistance controllability attributed to the well‐controlled CF, improved switching reliability improvement, and multiple intermediate resistance states. Such features are used to explore its usefulness as an artificial synapse, where fluent potentiation and depression can be achieved using voltage pulses. Further research is still required to emulate the artificial synapses fully, but the feasibility is presented in this work by exploiting the similarities between the Cu cations in the CBRAM device and the Ca2+ dynamics in a biological synapse. [ABSTRACT FROM AUTHOR]

The adjustable conductance of a two‐terminal memristor in a crossbar array can facilitate vector‐matrix multiplication in one step, making the memristor a promising synapse for efficiently implementing neuromorphic computing. To achieve controllable and gradual switching of multi‐level conductance, important for neuromorphic computing, a theoretical design of a superlattice‐like (SLL) structure switching layer for the multi‐level memristor is proposed and validated, refining the growth of conductive filaments (CFs) and preventing CFs from the abrupt formation and rupture. Ti/(HfOx/AlOy)SLL/TiN memristors are shown with transmission electron microscopy , X‐ray photoelectron spectroscopy , and ab initio calculation findings corroborate the SLL structure of HfOx/AlOy film. The optimized SLL memristor achieves outstanding conductance modulation performance with linearly synaptic weight update (nonlinear factor α = 1.06), and the convolutional neural network based on the SLL memristive synapse improves the handwritten digit recognition accuracy to 94.95%. Meanwhile, this improved synaptic device has a fast operating speed (30 ns), a long data retention time (≥ 104 s at 85 ℃), scalability, and CMOS process compatibility. Finally, its physical nature is explored and the CF evolution process is characterized using nudged elastic band calculations and the conduction mechanism fitting. In this work, as an example the HfOx/AlOy SLL memristor provides a design viewpoint and optimization strategy for neuromorphic computing. [ABSTRACT FROM AUTHOR]

Abstract The adjustable conductance of a two‐terminal memristor in a crossbar array can facilitate vector‐matrix multiplication in one step, making the memristor a promising synapse for efficiently implementing neuromorphic computing. To achieve controllable and gradual switching of multi‐level conductance, important for neuromorphic computing, a theoretical design of a superlattice‐like (SLL) structure switching layer for the multi‐level memristor is proposed and validated, refining the growth of conductive filaments (CFs) and preventing CFs from the abrupt formation and rupture. Ti/(HfOx/AlOy)SLL/TiN memristors are shown with transmission electron microscopy , X‐ray photoelectron spectroscopy , and ab initio calculation findings corroborate the SLL structure of HfOx/AlOy film. The optimized SLL memristor achieves outstanding conductance modulation performance with linearly synaptic weight update (nonlinear factor α = 1.06), and the convolutional neural network based on the SLL memristive synapse improves the handwritten digit recognition accuracy to 94.95%. Meanwhile, this improved synaptic device has a fast operating speed (30 ns), a long data retention time (≥ 104 s at 85 ℃), scalability, and CMOS process compatibility. Finally, its physical nature is explored and the CF evolution process is characterized using nudged elastic band calculations and the conduction mechanism fitting. In this work, as an example the HfOx/AlOy SLL memristor provides a design viewpoint and optimization strategy for neuromorphic computing.

The antiferromagnet (AFM) and ferromagnet (FM) interface is a unique branch of magnetics of broad scientific interest. AFMs play an important role in spin‐orbit torque devices based on their ability to generate spin‐polarized current and exchange bias when combined with FMs. In this study, an interesting spin‐orbit torque (SOT) ratchet involving the exchange spring effect in an IrMn/CoFeB bilayer device with perpendicular anisotropy and exchange bias is developed. The combined use of electrical and spectroscopic analysis reveals that the exchange spring in IrMn/CoFeB bilayer yields unidirectional anisotropy, resulting in a collinear/orthogonal AFM/FM spin configuration at the interface upon switching CoFeB magnetization upward/downward. The ratcheting characteristics resulting from unidirectional anisotropy manifest in SOT switching. In this process, magnetization against the exchange spring features digital‐like switching with a sharp transition, whereas the reverse function is characteristic of analog switching with a gradual transition tail. The dual digital‐analog characteristics of the IrMn/CoFeB bilayer may be of benefit in neuromorphic and memory applications. [ABSTRACT FROM AUTHOR]

Abstract In this study, by inserting a buffer layer of TiO x between the SiO x :Ag layer and the bottom electrode, we have developed a memristor device with a simple structure of Ag/SiO x :Ag/TiO x /p++-Si by a physical vapor deposition process, in which the filament growth and rupture can be efficiently controlled during analog switching. The synaptic characteristics of the memristor device with a wide range of resistance change for weight modulation by implementing positive or negative pulse trains have been investigated extensively. Several learning and memory functions have been achieved simultaneously, including potentiation/depression, paired-pulse-facilitation (PPF), short-term plasticity (STP), and STP-to-LTP (long-term plasticity) transition controlled by repeating pulses more than a rehearsal operation, and spike-time-dependent-plasticity (STDP) as well. Based on the analysis of logarithmic I-V characteristics, it has been found that the controlled evolution/dissolution of conductive Ag-filaments across the dielectric layers can improve the performance of the testing memristor device.

In this study, we demonstrate both of digital and analog memory operations in InGaZnO (IGZO) memristor devices by controlling the electrode materials for neuromorphic application. The switching properties of the devices are determined by the initial energy barrier characteristics between the metal electrodes and the IGZO switching layer. Digital switching characteristics are obtained after the forming process when Schottky junction occurs at both of top and bottom electrodes. On the other hands, analog resistive switching is achieved when Schottky and Ohmic junctions exist at each side because the applied voltage modulates the Schottky barrier height through the Ohmic contact. In addition, the weight-update properties of the devices are verified depending on identical and incremental pulse schemes. The incremental pulse trains improve the linearity and variation of weight modulation, leading to the stable learning characteristics of neuromorphic system in terms of pattern recognition with MNIST hand-written digit images.

Thin‐film transistors (TFTs) based on amorphous indium‐gallium‐zinc‐oxide (a‐IGZO) have attracted vast attention for use in organic light‐emitting diode (AMOLED) displays due to their high electron mobility and large current on–off ratio. Although amorphous oxide semiconductors show considerably less threshold voltage (Vth) variation than poly‐silicon, large‐area processing and degradation effects can impede the characteristic parameters of a‐IGZO TFTs, which manifests in an uneven brightness distribution across the display panel. Such Vth variations are usually reduced by additional compensation circuits consisting of TFTs and capacitors. Herein, a new approach to compensate such variabilities is demonstrated: the integration of a programmable ferroelectric (FE) film in the gate stack of the TFT. This simplifies the complexity of the pixel cell and potentially minimizes the need for compensation circuits, which is crucial for transparent displays. To test this new approach, fully integrated FE‐TFTs (i.e., with vias contacting a structured bottom gate electrode from the top) based on a‐IGZO and FE hafnium‐zirconium oxide (HZO) are developed. A single low‐temperature post‐fabrication treatment at 350 °C for 1 h in air is used to simultaneously crystallize the HZO film in the FE phase and to reduce the number of defects in the a‐IGZO channel. The structural and electrical characterizations provide comprehensive guidance for the design of effective FE‐TFT gate stacks and device geometries. An accurate control of the polarization state and linear switching between multiple intermediate states is shown by using programming pulses of various amplitudes and widths. Furthermore, a direct correlation between the channel length and the applied pulse width for programming is observed. [ABSTRACT FROM AUTHOR]

In this study, by inserting a buffer layer of TiOx between the SiOx:Ag layer and the bottom electrode, we have developed a memristor device with a simple structure of Ag/SiOx:Ag/TiOx/p++-Si by a physical vapor deposition process, in which the filament growth and rupture can be efficiently controlled during analog switching. The synaptic characteristics of the memristor device with a wide range of resistance change for weight modulation by implementing positive or negative pulse trains have been investigated extensively. Several learning and memory functions have been achieved simultaneously, including potentiation/depression, paired-pulse-facilitation (PPF), short-term plasticity (STP), and STP-to-LTP (long-term plasticity) transition controlled by repeating pulses more than a rehearsal operation, and spike-time-dependent-plasticity (STDP) as well. Based on the analysis of logarithmic I-V characteristics, it has been found that the controlled evolution/dissolution of conductive Ag-filaments across the dielectric layers can improve the performance of the testing memristor device. [ABSTRACT FROM AUTHOR]

In this study, we propose high-performance chitosan-based flexible memristors with embedded single-walled carbon nanotubes (SWCNTs) for neuromorphic electronics. These flexible transparent memristors were applied to a polyethylene naphthalate (PEN) substrate using low-temperature solution processing. The chitosan-based flexible memristors have a bipolar resistive switching (BRS) behavior due to the cation-based electrochemical reaction between a polymeric chitosan electrolyte and mobile ions. The effect of SWCNT addition on the BRS characteristics was analyzed. It was observed that the embedded SWCNTs absorb more metal ions and trigger the conductive filament in the chitosan electrolyte, resulting in a more stable and wider BRS window compared to the device with no SWCNTs. The memory window of the chitosan nanocomposite memristors with SWCNTs was 14.98, which was approximately double that of devices without SWCNTs (6.39). Furthermore, the proposed SWCNT-embedded chitosan-based memristors had memristive properties, such as short-term and long-term plasticity via paired-pulse facilitation and spike-timing-dependent plasticity, respectively. In addition, the conductivity modulation was evaluated with 300 synaptic pulses. These findings suggest that memristors featuring SWCNT-embedded chitosan are a promising building block for future artificial synaptic electronics applications.

Recently, in the field of artificial intelligence, neuromorphic devices have received significant attention for building artificial neural networks. Nano-scale resistive random access memory (RRAM) is one of the emerging electrical components for neuromorphic computing. In the present study, the synaptic characteristics are investigated in the piezoelectric B i F e 1 − x L u x O 3 (BFLO) and B i 1 − x L u x F e O 3 (BLFO) (x = 0.05) based RRAM devices in a vertical configuration. Both devices show bipolar resistive switching behavior with good endurance characteristics. However, analog switching with potentiation and depression response is observed in Ag/BFLO/FTO device. The change in resistance and leakage current flow across the device is explained by the space charge limiting current conduction mechanism with regard to thermally generated and injected charge carriers. Both films possess strong electro-mechanical response with the ferroelectric and ferro-elastic domain walls oriented along 7 1 0 , 10 9 0 , and 18 0 0 directions. The electro-mechanical response and resistive switching properties make the materials viable for sensing, energy harvesting, and non-volatile memory technology. Moreover, the BFLO-based synaptic device has potential application in neuromorphic computation. [Display omitted] • Rare-earth doped single-phase multifunctional BiFe 1-x Lu x O 3 (BFLO) and Bi 1-x Lu x FeO 3 (BLFO) (x = 0.05) films are prepared by sol–gel method. • Voltage controlled ferroelectric domains in both the films by PFM technique. • Both Ag/BFLO(BLFO)/FTO devices show forming-free bipolar resistive switching. • The Ag/BFLO/FTO device exhibits analog resistive switching with a gradual SET and RESET process. • Long-term potentiation (LTD) and long-term depression (LTD) synaptic simulations in Ag/BFLO/FTO artificial synapse. [ABSTRACT FROM AUTHOR]

[Display omitted] • Effect of the Ti buffer layer on synaptic plasticity and analog switching were investigated. • Different proportions of oxygen vacancies in multilayer films were confirmed by XPS. • Multilayer devices exhibit high a ON/OFF ratio (>104), and high switching stability. • Several forms of short-term plasticity were successfully mimicked in this multilayer device. • Oxygen vacancy-based filamentary model was proposed to explain switching mechanism. In this work, electronic synaptic plasticity and analog bipolar switching behavior by using a Pt/TiO x /AlO x /AlTaON/TaN multilayer resistive random-access memory (RRAM) device were examined. By using high-resolution transmission electron microscopy (HRTEM), the oxidation–reduction reaction occurs between the Ti buffer layer and the high-k Al 2 O 3 layer at the top interface, and the high-k Al 2 O 3 layer and the TaN bottom electrode at the bottom interface, resulting in the formation of the Pt/TiO x /AlO x /AlTaON/TaN multilayer RRAM structure. According to X-ray photoelectron spectroscopy (XPS) studies, different proportions of oxygen vacancies in TiO x /AlO x /AlTaON multilayer film play a key role in increasing RS properties uniformity. This multilayer structure-based RRAM device exhibits a stable RS behavior with a very large on/off resistance window (∼104), excellent durability (500 DC endurance cycles), and 104 s retention period with virtuous uniformity during the cycling operation. Furthermore, the device mean value (μ) set-and reset-voltage is −2.82 V, and +1.5 V, resulting in a low-operating voltage that is beneficial for neuromorphic systems. Besides, by adjusting the reset voltage, the device achieved a progressive transition from the low resistance state (LRS) to the high resistance state (HRS). To develop the artificial neural networks, basic biological properties such as potentiation, depression, and paired-pulse facilitation were efficiently mimicked. Ohmic conduction in the LRS and Schottky emission in the HRS are the two types of current transport conduction mechanisms. The analog bipolar switching behavior is governed by the redox reaction, which promotes the formation and rupture of oxygen vacancy-based conical shape conductive filaments. These findings show that multilayer RRAM devices have the potential to be used for future information storage and neuromorphic computing of artificial synapses. [ABSTRACT FROM AUTHOR]

The adjustable conductance of a two-terminal memristor in a crossbar array can facilitate vector-matrix multiplication in one step, making the memristor a promising synapse for efficiently implementing neuromorphic computing. To achieve controllable and gradual switching of multi-level conductance, important for neuromorphic computing, a theoretical design of a superlattice-like (SLL) structure switching layer for the multi-level memristor is proposed and validated, refining the growth of conductive filaments (CFs) and preventing CFs from the abrupt formation and rupture. Ti/(HfO x /AlO y ) SLL /TiN memristors are shown with transmission electron microscopy , X-ray photoelectron spectroscopy , and ab initio calculation findings corroborate the SLL structure of HfO x /AlO y film. The optimized SLL memristor achieves outstanding conductance modulation performance with linearly synaptic weight update (nonlinear factor α = 1.06), and the convolutional neural network based on the SLL memristive synapse improves the handwritten digit recognition accuracy to 94.95%. Meanwhile, this improved synaptic device has a fast operating speed (30 ns), a long data retention time (≥ 10 4 s at 85 ℃), scalability, and CMOS process compatibility. Finally, its physical nature is explored and the CF evolution process is characterized using nudged elastic band calculations and the conduction mechanism fitting. In this work, as an example the HfO x /AlO y SLL memristor provides a design viewpoint and optimization strategy for neuromorphic computing., (© 2022 The Authors. Advanced Science published by Wiley-VCH GmbH.)

This paper describes design and implementation of a single axis de attraction type electromagnetic suspension system where an electromagnet of 2.6 kg mass is levitated over a large gap under a fixed ferromagnetic guide-way. The electromagnet exhibits nonlinear force-current-distance characteristics, and if controllers are to be designed by using linear analysis, the air-gap is restricted to a small region around the chosen nominal operating point. In this work, an attempt has been made to increase the operating range of an electromagnetic suspension system by using the concept of piecewise linear control where the nonlinear force-current-airgap relationships of the magnetic suspension system have been successively linearized at several operating points with a suitable controller designed for each operating point. A novel analog switching scheme has been designed and implemented to automatically switch to the relevant controller depending on the actual air-gap. [ABSTRACT FROM AUTHOR]

In this study, we propose high-performance chitosan-based flexible memristors with embedded single-walled carbon nanotubes (SWCNTs) for neuromorphic electronics. These flexible transparent memristors were applied to a polyethylene naphthalate (PEN) substrate using low-temperature solution processing. The chitosan-based flexible memristors have a bipolar resistive switching (BRS) behavior due to the cation-based electrochemical reaction between a polymeric chitosan electrolyte and mobile ions. The effect of SWCNT addition on the BRS characteristics was analyzed. It was observed that the embedded SWCNTs absorb more metal ions and trigger the conductive filament in the chitosan electrolyte, resulting in a more stable and wider BRS window compared to the device with no SWCNTs. The memory window of the chitosan nanocomposite memristors with SWCNTs was 14.98, which was approximately double that of devices without SWCNTs (6.39). Furthermore, the proposed SWCNT-embedded chitosan-based memristors had memristive properties, such as short-term and long-term plasticity via paired-pulse facilitation and spike-timing-dependent plasticity, respectively. In addition, the conductivity modulation was evaluated with 300 synaptic pulses. These findings suggest that memristors featuring SWCNT-embedded chitosan are a promising building block for future artificial synaptic electronics applications. [ABSTRACT FROM AUTHOR]

A new type of two-dimensional (2D) SnO 2 semiconductor-based gate-tunable memristor, that is, a memtransistor, an integrated device of a memristor and a transistor, was demonstrated to advance next-generation neuromorphic computing technology. The polycrystalline 2D-SnO 2 memristors derived from a low-temperature and vacuum-free liquid metal process offer several interesting resistive switching properties such as excellent digital/analog resistive switching, multistate storage, and gate-tunability function of resistance switching states. Significantly, the gate tunability function that is not achievable in conventional two-terminal memristors provides the capability to implement heterosynaptic analog switching by regulating gate bias for enabling complex neuromorphic learning. We successfully demonstrated that the gate-tunable synaptic device dynamically modulated the analog switching behavior with good linearity and an improved conductance change ratio for high recognition accuracy learning. The presented gate-tunable 2D-oxide memtransistor will advance neuromorphic device technology and open up new opportunities to design learning schemes with an extra degree of freedom.

Memristors with the outstanding advantages are beneficial for neuromorphic computing and next-generation storage. Realizing various resistive switching behaviors in monolayer memristors is essential for understanding the device physics and fabricating fully memristive devices. In this paper, a simple and feasible method was proposed to achieve the digital and analog resistive switching in Cu/AlOx/Ag memristors by using ozone and water precursors in atomic layer deposition. According to the characterization results of surface topography, Raman spectrum and electrical measurement, the transition between the abrupt and gradual resistive switching was ascribed to the migration and diffusion of active electrode metal ions in the sparser, rougher and more amorphous AlOx dielectric films. The key features of biological synapses including long-term potentiation/depression, paired-pulse facilitation and learning-experience behaviors were emulated in the analog monolayer memristors. This study makes an important step towards the development of the sophisticated, multi-functional, and large-scale integrated neuromorphic devices and systems., (© 2021 IOP Publishing Ltd.)

Spiking neural networks (SNNs) sharing large similarity with biological nervous systems are promising to process spatiotemporal information and can provide highly time- and energy-efficient computational paradigms for the Internet-of-Things and edge computing. Nonvolatile electrolyte-gated transistors (EGTs) provide prominent analog switching performance, the most critical feature of synaptic element, and have been recently demonstrated as a promising synaptic device. However, high performance, large-scale EGT arrays, and EGT application for spatiotemporal information processing in an SNN are yet to be demonstrated. Here, an oxide-based EGT employing amorphous Nb 2 O 5 and Li x SiO 2 is introduced as the channel and electrolyte gate materials, respectively, and integrated into a 32 × 32 EGT array. The engineered EGTs show a quasi-linear update, good endurance (10 6 ) and retention, a high switching speed of 100 ns, ultralow readout conductance (<100 nS), and ultralow areal switching energy density (20 fJ µm -2 ). The prominent analog switching performance is leveraged for hardware implementation of an SNN with the capability of spatiotemporal information processing, where spike sequences with different timings are able to be efficiently learned and recognized by the EGT array. Finally, this EGT-based spatiotemporal information processing is deployed to detect moving orientation in a tactile sensing system. These results provide an insight into oxide-based EGT devices for energy-efficient neuromorphic computing to support edge application., (© 2020 Wiley-VCH GmbH.)

The switching parameters and device performance of memristors are predominately determined by their mobile species and matrix materials. Devices with oxygen or oxygen vacancies as the mobile species usually exhibit a great retention but also need a relatively high switching current (e.g., >30 µA), while devices with Ag or Cu as cation mobile species do not require a high switching current but usually show a poor retention. Here, Ru is studied as a new type of mobile species for memristors to achieve low switching current, fast speed, good reliability, scalability, and analog switching property simultaneously. An electrochemical metallization-like memristor with a stack of Pt/Ta 2 O 5 /Ru is developed. Migration of Ru ions is revealed by energy-dispersive X-ray spectroscopy mapping and in situ transmission electron microscopy within a sub-10 nm active device area before and after switching. The results open up a new avenue to engineer memristors for desired properties., (© 2020 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

In this study, by inserting a buffer layer of TiO x between the SiO x :Ag layer and the bottom electrode, we have developed a memristor device with a simple structure of Ag/SiO x :Ag/TiO x /p ++ -Si by a physical vapor deposition process, in which the filament growth and rupture can be efficiently controlled during analog switching. The synaptic characteristics of the memristor device with a wide range of resistance change for weight modulation by implementing positive or negative pulse trains have been investigated extensively. Several learning and memory functions have been achieved simultaneously, including potentiation/depression, paired-pulse-facilitation (PPF), short-term plasticity (STP), and STP-to-LTP (long-term plasticity) transition controlled by repeating pulses more than a rehearsal operation, and spike-time-dependent-plasticity (STDP) as well. Based on the analysis of logarithmic I-V characteristics, it has been found that the controlled evolution/dissolution of conductive Ag-filaments across the dielectric layers can improve the performance of the testing memristor device.

Memristor-based neuromorphic networks have been actively studied as a promising candidate to overcome the von-Neumann bottleneck in future computing applications. Several recent studies have demonstrated memristor network's capability to perform supervised as well as unsupervised learning, where features inherent in the input are identified and analyzed by comparing with features stored in the memristor network. However, even though in some cases the stored feature vectors can be normalized so that the winning neurons can be directly found by the (input) vector-(stored) vector dot-products, in many other cases, normalization of the feature vectors is not trivial or practically feasible, and calculation of the actual Euclidean distance between the input vector and the stored vector is required. Here we report experimental implementation of memristor crossbar hardware systems that can allow direct comparison of the Euclidean distances without normalizing the weights. The experimental system enables unsupervised K-means clustering algorithm through online learning, and produces high classification accuracy (93.3%) for the standard IRIS data set. The approaches and devices can be used in other unsupervised learning systems, and significantly broaden the range of problems a memristor-based network can solve.

Recent electronic applications require an efficient computing system that can perform data processing with limited energy consumption. Inspired by the massive parallelism of the human brain, a neuromorphic system (hardware neural network) may provide an efficient computing unit to perform such tasks as classification and recognition. However, the implementation of synaptic devices (i.e., the essential building blocks for emulating the functions of biological synapses) remains challenging due to their uncontrollable weight update protocol and corresponding uncertain effects on the operation of the system, which can lead to a bottleneck in the continuous design and optimization. Here, we demonstrate a synaptic transistor based on highly purified, preseparated 99% semiconducting carbon nanotubes, which can provide adjustable weight update linearity and variation margin. The pattern recognition efficacy is validated using a device-to-system level simulation framework. The enlarged margin rather than the linear weight update can enhance the fault tolerance of the recognition system, which improves the recognition accuracy.

Inspired by the human brain, a neuromorphic system combining complementary metal-oxide semiconductor (CMOS) and adjustable synaptic devices may offer new computing paradigms by enabling massive neural-network parallelism. In particular, synaptic devices, which are capable of emulating the functions of biological synapses, are used as the essential building blocks for an information storage and processing system. However, previous synaptic devices based on two-terminal resistive devices remain challenging because of their variability and specific physical mechanisms of resistance change, which lead to a bottleneck in the implementation of a high-density synaptic device network. Here we report that a three-terminal synaptic transistor based on carbon nanotubes can provide reliable synaptic functions that encode relative timing and regulate weight change. In addition, using system-level simulations, the developed synaptic transistor network associated with CMOS circuits can perform unsupervised learning for pattern recognition using a simplified spike-timing-dependent plasticity scheme.