Artificial Neuromorphic Systems Research Articles

Self-Powered FTO/CdSe/Bi2Se3 Photodetector with Bipolar Photoresponse Characteristics.

Self-powered photodetectors with bipolar photoresponse characteristics are expected to play a critical role in the field of secure optical communication, artificial neuromorphic systems, and intelligent color sensors. In this work, asymmetric heterojunction devices exhibiting wavelength-dependent bipolar photoresponse with a structure of Glass/FTO/CdSe/Bi2Se3/Au were fabricated. Under a short wavelength light irradiation, the top CdSe absorber generates a high carrier concentration; the excited carriers are quickly separated by the built-in electric field induced by the FTO/CdSe diode, resulting in a negative photocurrent. For light with wavelengths beyond the CdSe absorption edge, it is absorbed by the bottom Bi2Se3 absorber, and a positive photocurrent can be observed. Therefore, based on the bandgap difference between the top CdSe absorber and the bottom Bi2Se3 absorber, combined with the photogenerated carriers separated by asymmetric back-to-back diode, a wavelength-dependent bipolar response is realized. In this work, by employing this structure, the responsivities of -33.3 and 0.3 mA/W were achieved under the illumination of 405 and 830 nm, respectively. This work provides important indications in the preparation and performance optimization for wavelength-dependent bipolar photodetectors.

VO2/MoO3 Heterojunctions Artificial Optoelectronic Synapse Devices for Near-Infrared Optical Communication.

Artificial optoelectronic synapses (OES) have attracted extensive attention in brain-inspired information processing and neuromorphic computing. However, OES at near-infrared wavelengths have rarely been reported, seriously limiting the application in modern optical communication. Herein, high-performance near-infrared OES devices based on VO2/MoO3 heterojunctions are presented. The textured MoO3 films are deposited on the sputtered VO2 film by using the glancing-angle deposition technique to form a heterojunction device. Through tuning the oxygen defects in the VO2 film, the fabricated VO2/MoO3 heterojunction exhibits versatile electrical synaptic functions. Benefiting from the highly efficient light harvesting and the unique interface effect, the photonic synaptic characteristics, mainly including the short/long-term plasticity and learning experience behavior are successfully realized at the O (1342nm) and C (1550nm) optical communication wavebands. Moreover, a single OES device can output messages accurately by converting light signals of the Morse code to distinct synaptic currents. More importantly, a 3×3 artificial OES array is constructed to demonstrate the impressive image perceiving and learning capabilities. This work not only indicates the feasibility of defect and interface engineering in modulating the synaptic plasticity of OES devices, but also provides effective strategies to develop advanced artificial neuromorphic visual systems for next-generation optical communication systems.

Remaining useful life prediction towards cycling stability of organic electrochemical transistors

Organic electrochemical transistors (OECTs) show abundant potential in biosensors, artificial neuromorphic systems, brain-machine interfaces, etc With the fast development of novel functional materials and new device structures, OECTs with high transconductance (gm > mS) and good cycling stabilities (> 10,000 cycles) have been developed. While stability characterization is always time-consuming, to accelerate the development and commercialization of OECTs, tools for stability prediction are urgently needed. In this paper, OECTs with good cycling stabilities are realized by minimizing the gate voltage amplitude during cycling, while a remaining useful life (RUL) prediction framework for OECTs is proposed. Specifically, OECTs based on p(g2T-T) show tremendously enhanced stability which exhibits only 46.1% on-current (ION) and 33.2% peak gm decreases after 80,000 cycles (53 min). Then, RUL prediction is proposed based on the run-to-failure (RtF) aging tests (cycling stability test of OECTs). By selecting two aging parameters (ION and peak gm) as health indicators (HI), a novel multi-scale feature fusion (MFF) method for RUL prediction is proposed, which consists of a long short-term memory (LSTM) neural network based multi-scale feature generator (MFG) module for feature extraction and an attention-based feature fusion (AFF) module for feature fusion. Consequently, richer effective information is utilized to improve the prediction performance, where the experimental results show the superiority of the proposed framework on multiple OECTs in RUL prediction tasks. Therefore, by introducing such a powerful framework for the evaluation of the lifetime of OECTs, further optimization of materials, devices, and integrated systems relevant to OECTs will be stimulated. Moreover, this tool can also be extended to other relevant bioelectronics.

Stretchable and stable neuromorphic tactile system

A performance-stable tactile neuron is developed, which integrates a stretch-insensitive triboelectric nanogenerator with an artificial neuron in a single device, and a 64 × 64 neuromorphic tactile matrix is established to process touch signals.

All‐Optically Controlled Artificial Synapse Based on Full Oxides for Low‐Power Visible Neural Network Computing

AbstractArtificial synapse devices are dedicated to overcoming the von Neumann bottleneck. Adopting light signals in visual information processing and computing is vital for developing next‐generation artificial neuromorphic systems. A strategy to construct all‐optically controlled artificial synaptic devices based on full oxides with amorphous ZnAlSnO/SnO heterojunction in a two‐terminal planar configuration is proposed. All synaptic behaviors are operated in the visible optical pathway, with excitatory synapse under red (635 nm) light and inhibitory synapse under green (532 nm) and blue (405 nm) lights. Based on the different inhibitory effects, two modes of long‐term depression (LTD) and RESET processes can be implemented through green and blue lights, respectively. The energy consumption of an event can be as low as 0.75 pJ. A three‐layer perceptron model is designed to classify 28 × 28‐pixel handwritten digital images and performed supervised learning using a backpropagation algorithm, demonstrating the bio‐visually inspired neuromorphic computing with a training accuracy of 92.74%. The all‐optically controlled artificial synapses with write/erasure behaviors in visible RGB region and rational microelectronic process, as presented in this work, are essential in developing future artificial neuromorphic systems and highlight the huge potential of next‐generation computer systems.

Wavelength-Controlled Photoconductance Polarity Switching via Harnessing Defects in Doped PdSe2 for Artificial Synaptic Features.

Optoelectronic synapses are currently drawing significant attention as fundamental building blocks of neuromorphic computing to mimic brain functions. In this study, a two-terminal synaptic device based on a doped PdSe2 flake is proposed to imitate the key neural functions in an optical pathway. Due to the wavelength-dependent desorption of oxygen clusters near the intrinsic selenide vacancy defects, the doped PdSe2 photodetector achieves a high negative photoresponsivity of -7.8 × 103 A W-1 at 473nm and a positive photoresponsivity of 181 A W-1 at 1064nm. This wavelength-selective bi-direction photoresponse endows an all-optical pathway to imitate the fundamental functions of artificial synapses on a device level, such as psychological learning and forgetting capability, as well as dynamic logic functions. The underpinning photoresponse is further demonstrated on a flexible platform, providing a viable technology for neuromorphic computing in wearable electronics. Furthermore, the p-type doping results in an effective increase of the channel's electrical conductivity and a significant reduction in power consumption. Such low-power-consuming optical synapses with simple device architecture and low-dimensional features demonstrate tremendous promise for building multifunctional artificial neuromorphic systems in the future.

Bio‐Inspired Artificial Fast‐Adaptive and Slow‐Adaptive Mechanoreceptors With Synapse‐Like Functions

AbstractDevelopment of artificial mechanoreceptors capable of sensing and pre‐processing external mechanical stimuli is a crucial step toward constructing neuromorphic perception systems that can learn and store information. Here, bio‐inspired artificial fast‐adaptive (FA) and slow‐adaptive (SA) mechanoreceptors with synapse‐like functions are demonstrated for tactile perception. These mechanoreceptors integrate self‐powered piezoelectric pressure sensors with synaptic electrolyte‐gated field‐effect transistors (EGFETs) featuring a reduced graphene oxide channel. The FA pressure sensor is based on a piezoelectric poly(vinylidene fluoride‐trifluoroethylene) (P(VDF‐TrFE)) thin film, while the SA pressure sensor is enabled by a piezoelectric ionogel with the piezoelectric‐ionic coupling effect based on P(VDF‐TrFE) and an ionic liquid. Changes in post‐synaptic current are achieved through the synaptic effect of the EGFET by regulating the amplitude, number, duration, and frequency of tactile stimuli (pre‐synaptic pulses). These devices have great potential to serve as artificial biological mechanoreceptors for future artificial neuromorphic perception systems.

Tribo‐ferro‐optoelectronic neuromorphic transistor of α‐In2Se3

AbstractInspired by biological neural networks, the fabrication of artificial neuromorphic systems with multimodal perception capacity shows promises in overcoming the “von Neumann bottleneck” and takes advantage of the efficient perception and computation of diverse types of signals. Here, we combine a triboelectric nanogenerator with an α‐phase indium selenide (α‐In2Se3) optoelectronic synaptic transistor to construct a tribo‐ferro‐optoelectronic artificial neuromorphic device with multimodal plasticity. Based on the excellent ferroelectric and optoelectronic characteristics of the α‐In2Se3 channel, typical synaptic behaviors (e.g., pair‐pulse facilitation and short‐term/long‐term plasticity) are successfully simulated in response to the synergistic effect of mechanical and optical stimuli. The interaction of mechanical displacement and light illumination enables heterosynaptic plasticity and spatiotemporal dynamic logic. Furthermore, multiple Boolean logical functions and associative learning behaviors are successfully implemented using the paired stimuli of displacement pulses and light pulses. The proposed tribo‐ferro‐optoelectronic artificial neuromorphic devices have great potential for application in interactive neural networks and next‐generation artificial intelligence.

MXene-Based Memristor for Artificial Optoelectronic Neuron

With high efficiency and low energy consumption, bio-inspired artificial neuromorphic systems are regarded as the next generation of computing methods and have attracted tremendous attention in recent years. In bio-inspired artificial neuromorphic systems, multi-functional artificial neurons function as processing units to process complex information with high efficiency. However, the majority of reports on artificial neurons are based on electrical stimulation, whereas light-simulated neurons receive less attention. In this work, an artificial optoelectronic neural device based on an Ag/MXene/SiO2/Si structure is demonstrated. By introducing an oxidized MXene (O-MXene) layer, photoelectric integration can be realized on a single device. Compared to pure electrical stimulation, the synergistic effect of light and electrical stimulation can efficiently accelerate the firing behavior of neurons. In addition, we successfully designed and demonstrated a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$64\times64$ </tex-math></inline-formula> sensing array based on optoelectronic neurons to recognize and sharpen the input signal trajectory. The proposed artificial optoelectronic neural device shows great potential for optoelectronic neuromorphic systems and is expected to promote the development of multifunctional, high-performance neuromorphic systems.

Forming-free plant resistive random access memory based on the Coulomb blockade effect produced by gold nanoparticles.

Resistive random access memory has been widely studied as a powerful candidate for building future memories and realizing high-efficiency artificial neuromorphic systems. In this paper, gold nanoparticles (Au NPs) are doped in a leaf solution of Scindapsus aureus (SA) as the active layer to fabricate Al/SA:Au NPs/ITO/glass resistive random access memory. The device exhibits stable bipolar resistance switching characteristics. More importantly, the device's multilevel storage capabilities and synaptic potentiation and depression behaviour have been proven. Compared with the device without doped Au NPs in the active layer, the device has a higher ON/OFF current ratio, which is attributed to the Coulomb blockade effect caused by the Au NPs. The device plays an important role in realizing high-density memory and efficient artificial neuromorphic systems.

Sensitization of ZnO Photoconductivity in the Visible Range by Colloidal Cesium Lead Halide Nanocrystals

In this work, colloidal perovskite nanocrystals (PNCs) are used to sensitize the photoconductivity of nanocrystalline ZnO films in the visible range. Nanocrystalline ZnO with a crystallite size of 12-16 nm was synthesized by precipitation of a zinc basic carbonate from an aqueous solution, followed by annealing at 300 °C. Perovskite oleic acid- and oleylamine-capped CsPbBr3, CsPb(Cl/Br)3 and CsPb(Br/I)3 PNCs with a size of 6-13 nm were synthesized by a hot injection method at 170 °C in 1-octadecene. Photoconductive nanocomposites were prepared by applying a hexane sol of PNCs to a thick (100 μm) polycrystalline conductive ZnO layer. The spectral dependence of the photoconductivity, the dependence of the photoconductivity on irradiation, and the relaxation of the photoconductivity of the obtained nanocomposites have been studied. Sensitization of ZnO by CsPbBr3 and CsPb(Cl/Br)3 PNCs leads to enhanced photoconductivity in the visible range, the maximum of which is observed at 460 and 500 nm, respectively; close to the absorption maximum of PNCs. Nanocomposites ZnO/CsPb(Br/I)3 turned out to be practically not photosensitive when irradiated with light in the visible range. The data obtained are discussed in terms of the position of the energy levels of ZnO and PNCs and the probable PNCs photodegradation. The structure, morphology, composition, and optical properties of the synthesized nanocrystals have also been studied by XRD, TEM, and XPS. The results can be applied to the creation of artificial neuromorphic systems in the visible optical range.

Ion transport to temperature and gate in organic electrochemical transistors with anti-freezing hydrogel

Organic electrochemical transistors (OECTs) are useful for biochemical sensors, logic circuits, and artificial neuromorphic systems and call for insights into operation mechanisms to allow constructing complicated digital systems of high qualities. Here, we fabricate anti-freezing hydrogel based, dual-channel OECTs to probe the temperature-dependence and distance-dependence of current to investigate the ion transport. In the temperature-dependence measurement down to −30 °C, a threshold gate voltage is observed as V th , E = 0.5 V , above which the electrochemical doping mode dominates. By tuning the gate voltage above V th , E , the activation energy for hole transport is dramatically raised from 0.2 to 245.2 meV. Via the in-situ multi-point measurement, the transient response potential in neighboring channels are found not synchronous but limited by the ion transport when the ion concentration is low. The results are compared with those of OECTs based on aqueous electrolyte and reveal the operational mechanisms of OECT based on anti-freezing hydrogel. • The anti-freezing hydrogel based OECTs were fabricated to temperature-dependence measurement. • The in-situ multi-point measurement of dual-channel OECT was implemented to reveal ion transport. • The operation mode switching of OECTs was determined by the threshold gate voltage ( V th , E ). • Distance-dependence effect of ion transport was exhibited in low ionic concentration.

Tunable Bi‐directional Photoresponse in Hybrid PtSe2−x Thin Films Based on Precisely Controllable Selenization Engineering

AbstractBi‐directional photodetection with positive and negative photoconductance (PPC/NPC) plays a critical role in building next‐generation artificial neuromorphic systems, visual information processing and optoelectronic logic gates of the future. However, the typical device configuration based on heterostructures with strict working principles restricts the diversity of material selection. In this work, a novel strategy is proposed to fabricate bi‐directional photoresponse devices based on PtSe2−x films with precisely controllable selenization engineering. It is found that the PtSe2−x films are composed of Pt and PtSe2 components. The direction of the photocurrent can be reversibly modulated in a broadband range by tuning the ratio of PtSe2 in the film, which is ascribed to the synergistic mechanism of bolometric effect of Pt and photoconductive effect of PtSe2. The polarity switching wavelength from NPC to PPC can be further modulated through tuning the components of the film for specific demand. The tunable bi‐directional photoresponse enables the device to mimic the fundamental functions of artificial synapses all optically, such as aversion learning capability and logical functions, with high paired pulse facilitation index. This work potentially opens an opportunity for the development of advanced artificial neuromorphic systems and optoelectronic logic gates.

Adaptive immunomorphic hardware based on organic semiconductors and oxidized MXene heterostructures for feature information recognition

Neuromorphic computing with brain-like functions has become one of the important strategies for the von Neumann bottleneck. However, current artificial neuromorphic systems are mainly based on volatile or non-volatile synaptic devices, which limit the flexibility and computational efficiency of neuromorphic computing systems. Here, we report an adaptive immunomorphic hardware based on the heterostructure of MXene-TiO 2 complexes and organic semiconductors. The hardware has photon-triggered synaptic plasticity for accurate recognition and electrically triggered non-volatile retention for effective preservation of weight values. As a result, the retraining time and power consumption of the hardware can be reduced by 95% and 96%, respectively. Moreover, the array system expanded to 5 × 5 can extract special information from complex signals within 0.2 s, enabling feature information recognition. This work, therefore, provides a new strategy for improving the efficiency of artificial neuromorphic computation and has significant application prospects in intelligent sensing systems and edge computing. • Photoelectric transistor with memory window and memory ratio exceeding 60 V and 10 5 • Combination non-volatile and volatile retention for feature information recognition • The power consumption and time were reduced by 96% and 95% in retraining Gao et al. report a photoelectric transistor with the heterostructure of MXene-TiO 2 and P2FDIID. The device can achieve non-volatile and volatile retention behavior under electrical and optical pulse, respectively. As a result, the combined operation of the two retention characteristics can significantly reduce power consumption and time in retraining.

Intrinsically stretchable carbon nanotube synaptic transistors with associative learning ability and mechanical deformation response

Stretchable synaptic transistors with the ability to emulate multiplexed neural signal transmission is an important step toward mimicking complex brain activities in bionic memory sensors, intelligent soft robots, and skin electronics. The reports about stretchable artificial synaptic transistors are very limited and further exploration is urgently required. In this work, intrinsically stretchable carbon nanotube synaptic transistors (IS-CNT-STs) with poly (urea-urethane) dielectric have been realized. The synaptic characteristics and the synaptic mechanism have been analysed. Synaptic functions of the biological synapses including excitatory and inhibitory behaviours, short-term plasticity, and filtering characteristic are successfully emulated in the IS-CNT-STs. Moreover, the IS-CNT-STs have exhibited excellent mechanical stretchability even after experiencing large stretching deformation. After stretched by 50% strain, and even after 400 stretching cycles by 25% strain, both the transistor characteristics and the synaptic functions remain correctly. The realized IS-CNT-STs possess low operating voltage, multiple synaptic plasticity, and mechanical deformation response. The demonstrations indicate their capability to work in the stretchable artificial neuromorphic systems.

A binary classifier based on a reconfigurable dense network of metallic nanojunctions

Major efforts to reproduce the brain performances in terms of classification and pattern recognition have been focussed on the development of artificial neuromorphic systems based on top-down lithographic technologies typical of highly integrated components of digital computers. Unconventional computing has been proposed as an alternative exploiting the complexity and collective phenomena originating from various classes of physical substrates. Materials composed of a large number of non-linear nanoscale junctions are of particular interest: these systems, obtained by the self-assembling of nano-objects like nanoparticles and nanowires, results in non-linear conduction properties characterized by spatiotemporal correlation in their electrical activity. This appears particularly useful for classification of complex features: nonlinear projection into a high-dimensional space can make data linearly separable, providing classification solutions that are computationally very expensive with digital computers. Recently we reported that nanostructured Au films fabricated from the assembling of gold clusters by supersonic cluster beam deposition show a complex resistive switching behaviour. Their non-linear electric behaviour is remarkably stable and reproducible allowing the facile training of the devices on precise resistive states. Here we report about the fabrication and characterization of a device that allows the binary classification of Boolean functions by exploiting the properties of cluster-assembled Au films interconnecting a generic pattern of electrodes. This device, that constitutes a generalization of the perceptron, can receive inputs from different electrode configurations and generate a complete set of Boolean functions of n variables for classification tasks. We also show that the non-linear and non-local electrical conduction of cluster-assembled gold films, working at room temperature, allows the classification of non-linearly separable functions without previous training of the device.

Graphene-Based Artificial Synapses with Tunable Plasticity

Design and implementation of artificial neuromorphic systems able to provide brain akin computation and/or bio-compatible interfacing ability are crucial for understanding the human brain’s complex functionality and unleashing brain-inspired computation’s full potential. To this end, the realization of energy-efficient, low-area, and bio-compatible artificial synapses, which sustain the signal transmission between neurons, is of particular interest for any large-scale neuromorphic system. Graphene is a prime candidate material with excellent electronic properties, atomic dimensions, and low-energy envelope perspectives, which was already proven effective for logic gates implementations. Furthermore, distinct from any other materials used in current artificial synapse implementations, graphene is biocompatible, which offers perspectives for neural interfaces. In view of this, we investigate the feasibility of graphene-based synapses to emulate various synaptic plasticity behaviors and look into their potential area and energy consumption for large-scale implementations. In this article, we propose a generic graphene-based synapse structure, which can emulate the fundamental synaptic functionalities, i.e., Spike-Timing-Dependent Plasticity (STDP) and Long-Term Plasticity . Additionally, the graphene synapse is programable by means of back-gate bias voltage and can exhibit both excitatory or inhibitory behavior. We investigate its capability to obtain different potentiation/depression time scale for STDP with identical synaptic weight change amplitude when the input spike duration varies. Our simulation results, for various synaptic plasticities, indicate that a maximum 30% synaptic weight change and potentiation/depression time scale range from [-1.5 ms, 1.1 ms to [-32.2 ms, 24.1 ms] are achievable. We further explore the effect of our proposal at the Spiking Neural Network (SNN) level by performing NEST-based simulations of a small SNN implemented with 5 leaky-integrate-and-fire neurons connected via graphene-based synapses. Our experiments indicate that the number of SNN firing events exhibits a strong connection with the synaptic plasticity type, and monotonously varies with respect to the input spike frequency. Moreover, for graphene-based Hebbian STDP and spike duration of 20ms we obtain an SNN behavior relatively similar with the one provided by the same SNN with biological STDP. The proposed graphene-based synapse requires a small area (max. 30 nm 2 ), operates at low voltage (200 mV), and can emulate various plasticity types, which makes it an outstanding candidate for implementing large-scale brain-inspired computation systems.

Recent Progress of Protein‐Based Data Storage and Neuromorphic Devices

By virtue of energy efficiency, high speed, and parallelism, brain‐inspired neuromorphic computing is a promising technology to overcome the von Neumann bottleneck and capable of processing massive sophisticated tasks in the background of big data. The abilities of perceiving and reacting to events in artificial neuromorphic systems allow us to build the communicative electronic–biological interfaces to get closer to electronic life. Protein materials offer great application potentials in such a system due to their sustainability, low cost, controllable hierarchical structure, intrinsic biocompatibility, and biodegradability. Herein, a timely review of the development of protein‐based memories for data storage and neuromorphic computing is provided. Proteins’ unique mechanical, electronic, optical properties, and their broad applications are discussed. Then, the progress of protein‐based two‐terminal memristor and three‐terminal transistor‐type memory is reviewed, and their applications for data storage, logic circuit, and neuromorphic computing are introduced. Finally, the major challenges and outlook toward the future developing directions of protein‐based computing systems are pointed out.

Optical Monitoring of the Resistive States of a Polyaniline‐Based Memristive Device

AbstractOrganic memristive devices are of great interest nowadays due to their low costs, flexibility, and prospects of use in artificial neuromorphic systems. Polyaniline (PANI) is a conductive and electrochemically active polymer and a promising material for redox‐gated memristive devices. In this work, the electrochromic properties of PANI are utilized for spatially resolved mapping of electrochemical reactions in the active zone of a model PANI‐based device. The observed nonuniform propagation of the reaction front defines the difference in the kinetics of the device while switching from the low‐ to high‐resistance state and vice versa. This study both reveals the mechanism of resistive switching in redox‐gated memristive elements and offers a novel nondestructive method for monitoring the resistive state of PANI‐based memristive devices.

Solution-Processable Anion-doped Conjugated Polymer for Nonvolatile Organic Transistor Memory with Synaptic Behaviors

Brain-inspired synaptic transistors have been considered as a promising device for next-generation electronics. To mimic the behavior of a biological synapse, both data processing and nonvolatile memory capability are simultaneously required for a single electronic device. In this work, a simple approach to realize a synaptic transistor with improved memory characteristics is demonstrated by doping an ionic additive, tetrabutylammonium perchlorate (TBAP), into an active polymer semiconductor without using any extra charge storage layer. TBAP doping is first revealed to improve the memory window of a derived transistor memory device from 19 to 32 V (∼68% enhancement) with an on/off current ratio over 103 at VG = -10 V. Through morphological analysis and theoretical calculations, it is revealed that the association of anion with polymers enhances the charge retention capability of the polymer and facilitates the interchain interactions to result in improved memory characteristics. More critically, the doped device is shown to successfully mimic the synaptic behaviors, such as paired-pulse facilitation (PPF), excitatory and inhibitory postsynaptic currents, and spike-rate dependent plasticity. Notably, the TBAP-doped device is shown to deliver a PPF index of up to 204% in contrast to the negligible value of an undoped device. This study describes a novel approach to prepare a synaptic transistor by doping conjugated polymers, which can promote the future development of artificial neuromorphic systems.