Neuromorphic Devices Research Articles

Dual biological-clock controllable low-power fibrous synapse array based on heterojunction switched conductive filaments

Central and peripheral biological clocks are essential to regulate creatures’ circadian rhythms, mimicking the biological clock controlled neural and/or immunological responses are research hotspots in the emerging neuromorphic or brain-like systems. However, the current neuromorphic devices still depend on programmable frequency dividing and doubling machine clocks, the high system cost is a challenge to conduct sophisticated brain-like neuromorphic computing. Herein, a dual biological-clocks (including ambient light and endocrine) controllable fibrous synapse crossbar array is proposed. It is demonstrated and confirmed in this work, the Ag conductive filaments, which are critical in the conventional MXene synapses, can be successfully switched by differently biased MoS2/MXene heterojunction. The current and power of the unit fibrous synapse in positive/negative switching processes can reach as low as 460 pA/-170 pA and 330 pW/121 pW, respectively. Then, four kinds of logical calculations to the input light signals are accomplished by using the as-prepared fibrous synapse based logical gate. Meanwhile, the light and serotonin combined modulation on the synaptic currents are manifested and applied to simulate the immune-neural responses (memory and the memory exhausted) of T cells in the skin clock. More interestingly, two biological immune compensation mechanisms controlled by the central clock and peripheral clock, namely balanced and reset immune compensation, are also accomplished. This work may pave a new way for promoting the development of dual biological-clock regulated computations in neuromorphic systems.

Advanced Neuromorphic Applications Enabled by Synaptic Ion-Gating Vertical Transistors.

Bioinspired synaptic devices have shown great potential in artificial intelligence and neuromorphic electronics. Low energy consumption, multi-modal sensing and recording, and multifunctional integration are critical aspects limiting their applications. Recently, a new synaptic device architecture, the ion-gating vertical transistor (IGVT), has been successfully realized and timely applied to perform brain-like perception, such as artificial vision, touch, taste, and hearing. In this short time, IGVTs have already achieved faster data processing speeds and more promising memory capabilities than many conventional neuromorphic devices, even while operating at lower voltages and consuming less power. This work focuses on the cutting-edge progress of IGVT technology, from outstanding fabrication strategies to the design and realization of low-voltage multi-sensing IGVTs for artificial-synapse applications. The fundamental concepts of artificial synaptic IGVTs, such as signal processing, transduction, plasticity, and multi-stimulus perception are discussed comprehensively. The contribution draws special attention to the development and optimization of multi-modal flexible sensor technologies and presents a roadmap for future high-end theoretical and experimental advancements in neuromorphic research that are mostly achievable by the synaptic IGVTs.

MoS2/ZnO-heterostructured optoelectronic synapse for multiwavelength optical information-based sensing, memory, and processing

The rapid development of artificial intelligence and humanoid robots requires hardware-level revolution of synapses-based bionic vision system. However, current optoelectronic synapses featuring multiwavelength sensing and multifunctional processing necessitate intricate integrations of multi-materials and floating gate configurations, which induces complex fabrication process and limited integration density. Here, we propose a novel two-terminal optoelectronic synaptic device based on MoS2/ZnO heterostructure which integrates sense and memory functions and realizes multifunctional optical information processing while circumvents the need of complex configurations. By simultaneously exploiting the photo-adsorption oxygen dissociation property of ZnO and the persistent photoconductivity effect of MoS2, the device demonstrates synaptic capability of four-color perception with a single synaptic event consuming a mere 2.55×10−9 J. Different from the traditional power-density sensitive nociceptor, for the first time photon energy-sensitive nociceptors is developed, which can mimic the process of UV irradiation to the retina leading to fear of visible light. Diverse information processing ability, such as photonic information encryption, associative learning, and logic calculation, was demonstrated in a single device, indicating a practical strategy for facile design for future high integration intensity neuromorphic devices.

Self‐Healing Magnetic Field‐Assisted Threshold Switching Device Utilizing Dual Field‐Driven Filamentary Physics

AbstractAdvanced filamentary devices are crucial for developing low‐power devices to implement high‐speed logic and neuromorphic devices. Among these, HfO2‐based filamentary devices have attracted attention as viable options due to their threshold‐switching characteristics and compatibility with complementary metal‐oxide‐semiconductor (CMOS) technology. However, the unpredictability of conventional filament formation/rupture driven by an electric field challenges consistency and reliability. A paradigm shift from conventional stochastic electric field‐driven ion migration to controllable ion‐based transportation is essential to devise functional low‐power devices capable of controlling the filament process. This work introduces a magnetic field‐assisted threshold switching (MA‐TS) device, which integrates a neodymium magnet and a nickel (Ni) barrier layer to enable controlled dual field‐driven ion transportation. The dual field‐driven process combining the conventional vertical electric field‐driven ion migration with lateral magnetic field‐driven ion transportation, reveals a distinctive aspect of ion movement. The MA‐TS device achieves superior performances characterized by an ultra‐low threshold voltage (≈0 V), minimized leakage current in the off‐state, a variation‐immune hysteresis‐free characteristic, enhanced yield, and revival‐ability (i.e., self‐healing) after a failed TS operation. By overcoming the limitations of conventional filamentary devices, the MA‐TS device opens up a promising avenue for efficient and stable low‐power applications.

Setting Plasma Immersion Ion Implantation of Ar+ Parameters towards Electroforming-Free and Self-Compliance HfO2-Based Memristive Structures.

Memristive structures are among the most promising options to be components of neuromorphic devices. However, the formation of HfO2-based devices in crossbar arrays requires considerable time since electroforming is a single stochastic operation. In this study, we investigate how Ar+ plasma immersion ion implantation (PI) affects the Pt/HfO2 (4 nm)/HfOXNY (3 nm)/TaN electroforming voltage. The advantage of PI is the simultaneous and uniform processing of the entire wafer. It is thought that Ar+ implantation causes defects to the oxide matrix, with the majority of the oxygen anions being shifted in the direction of the TaN electrode. We demonstrate that it is feasible to reduce the electroforming voltages from 7.1 V to values less than 3 V by carefully selecting the implantation energy. A considerable decrease in the electroforming voltage was achievable at an implantation energy that provided the dispersion of recoils over the whole thickness of the oxide without significantly affecting the HfOXNY/TaN interface. At the same time, Ar+ PI at higher and lower energies did not produce the same significant decrease in the electroforming voltage. It is also possible to obtain self-compliance of current in the structure during electroforming after PI with energy less than 2 keV.

Oxide Based Pentachromatic‐Vision Inspired Optoelectronic Synaptic Transistor with Large Conduction States Over 512

AbstractOptoelectronic neuromorphic devices based on oxide semiconductors have been potentially investigated to mimic the functions of human visual synapses. However, the challenge comes from the wide bandgap characteristics of numerous oxide semiconductors, which restricts the response range of the device under ultra‐violet (UV) region. Strategies for widening the response range are mostly focused on artificially generating the defect states, however, most of them results in mimicking the tetrachromatic visual system from UV to visible light range. To be used for industries such as robotics, or autonomous vehicles, mimicking the tetrachromatic vision system should be overcome up to near‐infrared (NIR) region. Here, a facile solution processed indium‐gallium‐zinc‐oxide and silver oxide structured optoelectronic synaptic transistor is fabricated not only to mimic the function of human synapses, but to overcome the tetrachromatic human visual system up to the NIR region. The device not only showed photoresponse characteristics under the entire 405 to 830 nm wavelength region, but also showed significant synaptic behaviors with over 512 conduction states under a reasonable incident light power density of 4.5 mW cm−2. The results will offer a useful facile method for fabricating optoelectronic synaptic transistors that can overcome the tetrachromatic vision systems.

Multi-Level Switching of Spin-Torque Ferromagnetic Resonance in 2D Magnetite.

2D magnetic materials hold substantial promise in information storage and neuromorphic device applications. However, achieving a 2D material with high Curie temperature (TC), environmental stability, and multi-level magnetic states remains a challenge. This is particularly relevant for spintronic devices, which require multi-level resistance states to enhance memory density and fulfil low power consumption and multi-functionality. Here, the synthesis of 2D non-layered triangular and hexagonal magnetite (Fe3O4) nanosheets are proposed with high TC and environmental stability, and demonstrate that the ultrathin triangular nanosheets show broad antiphase boundaries (bAPBs) and sharp antiphase boundaries (sAPBs), which induce multiple spin precession modes and multi-level resistance. Conversely, the hexagonal nanosheets display slip bands with sAPBs associated with pinning effects, resulting in magnetic-field-driven spin texture reversal reminiscent of "0" and "1" switching signals. In support of the micromagnetic simulation, direct explanation is offer to the variation in multi-level resistance under a microwave field, which is ascribed to the multi-spin texture magnetization structure and the randomly distributed APBs within the material. These novel 2D magnetite nanosheets with unique spin textures and spin dynamics provide an exciting platform for constructing real multi-level storage devices catering to emerging information storage and neuromorphic computing requirements.

Concealing Organic Neuromorphic Devices with Neuronal-Inspired Supported Lipid Bilayers.

Neurohybrid systems have gained large attention for their potential as in vitro and in vivo platform to interrogate and modulate the activity of cells and tissue within nervous system. In this scenario organic neuromorphic devices have been engineered as bioelectronic platforms to resemble characteristic neuronal functions. However, aiming to a functional communication with neuronal cells, material synthesis, and surface engineering can yet be exploited for optimizing bio-recognition processes at the neuromorphic-neuronal hybrid interface. In this work, artificial neuronal-inspired lipid bilayers have been assembled on an electrochemical neuromorphic organic device (ENODe) to resemble post-synaptic structural and functional features of living synapses. Here, synaptic conditioning has been achieved by introducing two neurotransmitter-mediated biochemical signals, to induce an irreversible change in the device conductance thus achieving Pavlovian associative learning. This new class of in vitro devices can be further exploited for assembling hybrid neuronal networks and potentially for in vivo integration within living neuronal tissues.

Realization of Multiple Synapse Plasticity by Coexistence of Volatile and Nonvolatile Characteristics of Interface Type Memristor.

Studies on neuromorphic computing systems are becoming increasingly important in the big-data-processing era as these systems are capable of energy-efficient parallel data processing and can overcome the present limitations owing to the von Neumann bottleneck. The Pt/WOx/ITO resistive random-access memory device can be used to implement versatile synapse functions because it possesses both volatile and nonvolatile characteristics. The gradual increase and decrease in the current of the Pt/WOx/ITO device with its uniform resistance state for endurance and retention enables additional synaptic applications that can be controlled using electric pulses. If the volatile and nonvolatile device properties are set through rehearsal and forgetting processes, the device can emulate various synaptic behaviors, such as potentiation and depression, paired-pulse facilitation, post-tetanic potentiation, image training, Hebbian learning rules, excitatory postsynaptic current, and Pavlov's test. Furthermore, reservoir computing can be implemented for applications such as pattern generation and recognition. This emphasizes the various applications of future neuromorphic devices, demonstrating the various favorable characteristics of pulse-enhanced Pt/WOx/ITO devices.

Predicting Blood Glucose Levels with Organic Neuromorphic Micro-Networks.

Accurate glucose prediction is vital for diabetes management. Artificial intelligence and artificial neural networks (ANNs) are showing promising results for reliable glucose predictions, offering timely warnings for glucose fluctuations. The translation of these software-based ANNs into dedicated computing hardware opens a route toward automated insulin delivery systems ultimately enhancing the quality of life for diabetic patients. ANNs are transforming this field, potentially leading to implantable smart prediction devices and ultimately to a fully artificial pancreas. However, this transition presents several challenges, including the need for specialized, compact, lightweight, and low-power hardware. Organic polymer-based electronics are a promising solution as they have the ability to implement the behavior of neural networks, operate at low voltage, and possess key attributes like flexibility, stretchability, and biocompatibility. Here, the study focuses on implementing software-based neural networks for glucose prediction into hardware systems. How to minimize network requirements, downscale the architecture, and integrate the neural network with electrochemical neuromorphic organic devices, meeting the strict demands of smart implants for in-body computation of glucose prediction is investigated.

Magnetization switching by electric field in ZnFe2O4/ZnO heterostructure

We have fabricated the ZnFe2O4/ZnO (ZFO/ZnO) heterostructure on a silicon substrate by pulsed laser deposition technique and studied the magnetization switching by the electric field in the ZFO/ZnO heterostructure using an indigenously developed optical cantilever beam magnetometer setup. The magnetization (M) vs electric field (E) curve reveals that the magnetization of the ZFO film has been switched by an electric field applied along the thickness of the ZnO film. The saturation magnetization is found to be 28.77 MA/m from the M–E curve. The emergence of electric field-driven magnetization switching in the ZFO/ZnO heterostructure is attributed to the strain-mediated magnetoelectric coupling between the electric polarization of the ZnO film and the magnetization of the ZFO film as evidenced by the butterfly-type hysteresis behavior of magnetization with the applied electric field. However, the realization of electric field-controlled magnetization switching in the ZFO/ZnO heterostructure is regarded as a potential aspect for the fabrication of energy-efficient spintronic devices such as magnetoelectric random access memory cells, highly sensitive magnetic field sensors, magneto-logic devices, and neuromorphic devices.

A synaptic transistor based on van der Waals heterojunction HfS2/HfOx/SnS2 with optical modulation properties

Neuromorphic devices, which break the traditional von Neumann architecture, have attracted much attention for their ability to mimic the perception, learning, and memory functions of the human brain. Two-dimensional (2D) materials are excellent candidates for building neuromorphic devices due to their atomic-level thickness and excellent optoelectronic properties. In this work, we designed a nonvolatile floating-gate synaptic device based on an HfS2/HfOx/SnS2 van der Waals heterostructure. This device exhibits a large memory window of 60 V, based on which synaptic properties, such as an excitatory postsynaptic current, and short-term and long-term plasticity, were simulated. In addition, the long-term potentiation/depression (LTP/D) characteristics of the device can be optically modulated. The device has a low nonlinearity of 0.22 for LTP, and the ratio of the number of effective conductance states was 93.3% under 532 nm illumination; this is an improvement on the levels reported using 2D-material floating-gate devices in recent years. This work offers the possibility of future applications of optoelectronic synaptic devices.

Artificial synaptic simulating pain-perceptual nociceptor and brain-inspired computing based on Au/Bi3.2La0.8Ti3O12/ITO memristor

Recently, memristors have garnered widespread attention as neuromorphic devices that can simulate synaptic behavior, holding promise for future commercial applications in neuromorphic computing. In this paper, we present a memristor with an Au/Bi3.2La0.8Ti3O12 (BLTO)/ITO structure, demonstrating a switching ratio of nearly 103 over a duration of 104 s. It successfully simulates a range of synaptic behaviors, including long-term potentiation and depression, paired-pulse facilitation, spike-timing-dependent plasticity, spike-rate-dependent plasticity etc. Interestingly, we also employ it to simulate pain threshold, sensitization, and desensitization behaviors of pain-perceptual nociceptor (PPN). Lastly, by introducing memristor differential pairs (1T1R-1T1R), we train a neural network, effectively simplifying the learning process, reducing training time, and achieving a handwriting digit recognition accuracy of up to 97.19 %. Overall, the proposed device holds immense potential in the field of neuromorphic computing, offering possibilities for the next generation of high-performance neuromorphic computing chips.

Ultralow-power optoelectronic synaptic transistors based on polyzwitterion dielectrics for in-sensor reservoir computing.

Bio-inspired transistor synapses use solid electrolytes to achieve low-power operation and rich synaptic behaviors via ion diffusion and trapping. While these neuromorphic devices hold great promise, they still suffer from challenges such as high leakage currents and power consumption, electrolysis risk, and irreversible conductance changes due to long-range ion migrations and permanent ion trapping. In addition, their response to light is generally limited because of "exciton-polaron quenching", which restricts their potential in in-sensor neuromorphic visions. To address these issues, we propose replacing solid electrolytes with polyzwitterions, where the cation and anion are covalently concatenated via a flexible alkyl chain, thus preventing long-range ion migrations while inducing good photoresponses to the transistors via interfacial charge trapping. Our detailed studies reveal that polyzwitterion-based transistors exhibit optoelectronic synaptic behavior with ultralow-power consumption (~250 aJ per spike) and enable high-performance in-sensor reservoir computing, achieving 95.56% accuracy in perceiving the trajectory of moving basketballs.

Robust compression and detection of epileptiform patterns in ECoG using a real-time spiking neural network hardware framework

Interictal Epileptiform Discharges (IED) and High Frequency Oscillations (HFO) in intraoperative electrocorticography (ECoG) may guide the surgeon by delineating the epileptogenic zone. We designed a modular spiking neural network (SNN) in a mixed-signal neuromorphic device to process the ECoG in real-time. We exploit the variability of the inhomogeneous silicon neurons to achieve efficient sparse and decorrelated temporal signal encoding. We interface the full-custom SNN device to the BCI2000 real-time framework and configure the setup to detect HFO and IED co-occurring with HFO (IED-HFO). We validate the setup on pre-recorded data and obtain HFO rates that are concordant with a previously validated offline algorithm (Spearman’s ρ = 0.75, p = 1e-4), achieving the same postsurgical seizure freedom predictions for all patients. In a remote on-line analysis, intraoperative ECoG recorded in Utrecht was compressed and transferred to Zurich for SNN processing and successful IED-HFO detection in real-time. These results further demonstrate how automated remote real-time detection may enable the use of HFO in clinical practice.

SiC@NiO Core-Shell Nanowire Networks-Based Optoelectronic Synapses for Neuromorphic Computing and Visual Systems at High Temperature.

1D nanowire networks, sharing similarities of structure, information transfer, and computation with biological neural networks, have emerged as a promising platform for neuromorphic systems. Based on brain-like structures of 1D nanowire networks, neuromorphic synaptic devices can overcome the von Neumann bottleneck, achieving intelligent high-efficient sensing and computing function with high information processing rates and low power consumption. Here, high-temperature neuromorphic synaptic devices based on SiC@NiO core-shell nanowire networks optoelectronic memristors (NNOMs) are developed. Experimental results demonstrate that NNOMs attain synaptic short/long-term plasticity and modulation plasticity under both electrical and optical stimulation, and exhibit advanced functions such as short/long-term memory and "learning-forgetting-relearning" under optical stimulation at both room temperature and 200°C. Based on the advanced functions under light stimulus, the constructed 5×3 optoelectronic synaptic array devices exhibit a stable visual memory function up to 200°C, which can be utilized to develop artificial visual systems. Additionally, when exposed to multiple electronic or optical stimuli, the NNOMs effectively replicate the principles of Pavlovian classical conditioning, achieving visual heterologous synaptic functionality and refining neural networks. Overall, with abundant synaptic characteristics and high-temperature thermal stability, these neuromorphic synaptic devices offer a promising route for advancing neuromorphic computing and visual systems.

An organic artificial soma for spatio-temporal pattern recognition via dendritic integration

A novel organic neuromorphic device performing pattern classification is presented and demonstrated. It features an artificial soma capable of dendritic integration from three pre-synaptic neurons. The time-response of the interface between electrolytic solutions and organic mixed ionic-electronic conductors is proposed as the sole computational feature for pattern recognition, and it is easily tuned in the organic dendritic integrator by simply controlling electrolyte ionic strength. The classifier is benchmarked in speech-recognition experiments, with a sample of 14 words, encoded either from audio tracks or from kinematic data, showing excellent discrimination performances in a planar, miniaturizable, fully passive device, designed to be promptly integrated in more complex architectures where on-board pattern classification is required.

Strain-insensitive viscoelastic perovskite film for intrinsically stretchable neuromorphic vision-adaptive transistors

Stretchable neuromorphic optoelectronics present tantalizing opportunities for intelligent vision applications that necessitate high spatial resolution and multimodal interaction. Existing neuromorphic devices are either stretchable but not reconcilable with multifunctionality, or discrete but with low-end neurological function and limited flexibility. Herein, we propose a defect-tunable viscoelastic perovskite film that is assembled into strain-insensitive quasi-continuous microsphere morphologies for intrinsically stretchable neuromorphic vision-adaptive transistors. The resulting device achieves trichromatic photoadaptation and a rapid adaptive speed (<150 s) beyond human eyes (3 ~ 30 min) even under 100% mechanical strain. When acted as an artificial synapse, the device can operate at an ultra-low energy consumption (15 aJ) (far below the human brain of 1 ~ 10 fJ) with a high paired-pulse facilitation index of 270% (one of the best figures of merit in stretchable synaptic phototransistors). Furthermore, adaptive optical imaging is achieved by the strain-insensitive perovskite films, accelerating the implementation of next-generation neuromorphic vision systems.

Ferroelectricity-Defects Synergistic Artificial Synapses for High Recognition Accuracy Neuromorphic Computing.

The ability of ferroelectric memristors to modulate conductance and offer multilevel storage has garnered significant attention in the realm of artificial synapses. On one hand, the resistance change of ferroelectric memristors mainly depends on the polarization reversal. On the other hand, the defects such as oxygen vacancies, which are inevitable presence during high-temperature processes, can undergo diffusion drift with the polarization reversal, thereby change the interface potential barrier. Thus, it is both desirable and necessary to investigate the synergistic effect of ferroelectricity and defects. Here, we prepare BaTiO3 ferroelectric memristor by pulse laser deposition and achieve resistance switching through the synergistic effect of ferroelectricity and oxygen vacancies. The memristor shows excellent switching characteristics with a large switching ratio (104) and good stability (103 s). It effectively emulates the features of artificial synapses and accomplishes decimal logical neural computing. In the neuromorphic system crafted with the memristor, the recognition accuracy of the 28 × 28 pixel image reaches 94.9%. These findings strongly support the research of ferroelectric memristors in neuromorphic devices.

Piezo strain-controlled phase transition in single-crystalline Mott switches for threshold-manipulated leaky integrate-and-fire neurons.

Electrical manipulation of the metal-insulator transition (MIT) in quantum materials has attracted considerable attention toward the development of ultracompact neuromorphic devices because of their stimuli-triggered transformations. VO2 is expected to undergo abrupt electronic phase transition by piezo strain near room temperature; however, the unrestricted integration of defect-free VO2 films on piezoelectric substrates is required to fully exploit this emerging phenomenon in oxide heterostructures. Here, we demonstrate the integration of single-crystalline VO2 films on highly lattice-mismatched PMN-PT piezoelectric substrates using a single-crystal TiO2-nanomembrane (NM) template. Using our strategy on heterogeneous integration, single-crystal-like steep transition was observed in the defect-free VO2 films on TiO2-NM-PMN-PT. Unprecedented TMI modulation (5.2 kelvin) and isothermal resistance of VO2 [ΔR/R (Eg) ≈ 18,000% at 315 kelvin] were achieved by the efficient strain transfer-induced MIT, which cannot be achieved using directly grown VO2/PMN-PT substrates. Our results provide a fundamental strategy to realize a single-crystalline artificial heterojunction for promoting the application of artificial neurons using emergent materials.