Neuromorphic Transistors Research Articles

A Biomimetic Nociceptor Based on a Vertical Multigate, Multichannel Neuromorphic Transistor.

Nociceptors, crucial sensory receptors within biological systems, are essential for survival in diverse and potentially hazardous environments. Efforts to replicate nociceptors through advanced electronic devices, such as memristors and neuromorphic transistors, have achieved limited success, capturing basic nociceptive functions while more advanced characteristics like various forms of central sensitization and analgesic effect remain out of reach. Here, we introduce a vertical multigate, multichannel electrolyte-gated transistor (Vm-EGT), designed to mimic nociceptors. Utilizing the hybrid mechanism combining electric-double-layer (EDL) with ion intercalation/deintercalation in EGTs, our approach successfully replicates peripheral sensitization and desensitization characteristics of nociceptors. The intricate multigate and multichannel design of the Vm-EGT enables the emulation of more advanced nociceptive functionalities, including central sensitization and analgesic effect. Furthermore, we demonstrate that by exploiting the inherent current-voltage relationship, the Vm-EGT can simulate these advanced nociceptive features and seamlessly transition between them. Integrating a Vm-EGT with a thermistor and a heating plate, we have developed an artificial thermal nociceptor that closely mirrors the sensory attributes of its biological counterpart. Our approach significantly advances the emulation of nociceptors, providing a basis for the development of sophisticated artificial sensory systems and intelligent robotics.

IGZO/PVP Composite Nanofiber Neuromorphic Transistors with Optoelectronic Synapse Emulation and Reservoir Computing.

Nanofiber neuromorphic transistors are regarded as promising candidates for mimicking brain-like learning and advancing high-performance computing. Composite nanofibers (CNFs) typically exhibit enhanced optoelectronic and mechanical properties. In this study, indium-gallium-zinc oxide (IGZO)/polyvinylpyrrolidone (PVP) CNFs were synthesized, and the neuromorphic transistor was integrated on both rigid and flexible substrates. The learning behavior, characterized by the transition from short-term plasticity (STP) to long-term plasticity, was achieved through photoelectric stimulation of the rigid neuromorphic transistor. The nonlinear STP was simulated by the flexible neuromorphic transistor through electrical pulses, matching effectively with a reservoir computing (RC) system. Hand gesture recognition with little energy consumption (49 pJ per reservoir state) and a maximum accuracy of 92.86% has been achieved by the RC system, proving the substantial potential of the IGZO/PVP CNF neuromorphic transistor for wearable intelligent processing tasks.

TiO2 Channel Material for Lithium-Ion-Gated Neuromorphic Transistors

Ubiquitous Internet of Things (IoT) platforms have exponentially increased the relay of an abundance of data between memory and processor in current computing architectures, ultimately limiting throughput and advancing towards the von Neumann bottleneck. Brain-inspired computing using neuromorphic devices presents an encouraging alternative framework capable of power-efficient real-time inference, self-learning, and decision-making. This work focuses specifically on ion-gated, three-terminal transistors that rely on electrochemically driven ion intercalation between an electronically insulating electrolyte and a conducting channel. While other nonvolatile memories (NVMs) can be used in neuromorphic devices, electrochemical random-access memory (ECRAM) advantageously enables the programming of multi-states and their retention in the absence of applied power. This talk will present our recent work on anatase TiO2 as a channel material in a lithium-ion-gated, three-terminal transistor. Based on the literature, anatase TiO2 covers a wide channel conductance range and exhibits a linear response, allowing for numerous well-defined synaptic weights [1]. We explore a range of channel conductance values using both potentiostatic and galvanostatic pulsing and correlate material properties to electrical measurements. These results provide a benchmark and roadmap for lithium-ion-based neuromorphic transistors capable of multi-states.[1] Y. Li, E.J. Fuller, S. Asapu, S. Agarwal, T. Kurita, J.J. Yang, and A.A. Talin, ACS Appl. Mater. Interfaces 11, 38982 (2019).

Low-voltage solution-processed Sn-doped CuI neuromorphic transistors with synaptic plasticity and pain mimicked

In this article, SnxCu1−xI thin-film transistors were fabricated on a glass substrate, with CuI doped with varying concentrations of SnI2 serving as the channel and chitosan as the dielectric. When x = 0.06, the device exhibited optimal performance: a current on/off ratio of 2.56 × 105, a subthreshold slope of 31.67 mV/dec, a threshold voltage of 1.33 V, and a saturated field-effect mobility of 21.75 cm2 V−1 s−1. Due to the electric double layer effect of chitosan, the operating voltage of the devices was reduced to below 2 V. Simulations were also conducted on the behavior and functionality of artificial synapses, such as short-term plasticity, long-term plasticity, and paired-pulse facilitation. Building upon the functionalities of artificial synapses, the Sn0.06Cu0.94I neuromorphic transistors simulated the fundamental pain perception function of biological nociceptors. Finally, the effects of bias stress and laser irradiation on the devices were investigated, indicating the excellent stability of the Sn0.06Cu0.94I neuromorphic transistors. Fabricated via the solution process, this low-voltage neuromorphic transistors hold significant implications for applications in bionic sensing systems and neuromorphic chip technology.

Proton gated oxide neuromorphic transistors with temperature-sensitive thermochromic function

Temperature is always one of the most important factors that critically affects the electrical performance of devices. In the present work, the thermochromic function has been realized on oxide neuromorphic transistors gated by a chitosan based electrolyte mixed with temperature-sensitive thermochromic powder. Except for extremely strong proton gating effects, the device exhibits temperature-sensitive thermochromic activities with good electrical performance. It also demonstrates typical synaptic functions at different temperatures, mimicking heat pain sensation activities. Additionally, a high recognition accuracy of ∼93.80% for MNIST handwritten digits is obtained at different temperatures, indicating good stabilities for neuromorphic computing. The present work opens up interesting idea for bionic cognitive electronic skin and bionic thermochromic cognitive platforms.

Proton Competitive Activities in Flexible Hydrogel Gated Oxide Neuromorphic Transistor for Mimicking Bienenstock– Cooper–Munro Learning Rule

Proton Competitive Activities in Flexible Hydrogel Gated Oxide Neuromorphic Transistor for Mimicking Bienenstock– Cooper–Munro Learning Rule

Visualized in-sensor computing

In artificial nervous systems, conductivity changes indicate synaptic weight updates, but they provide limited information compared to living organisms. We present the pioneering design and production of an electrochromic neuromorphic transistor employing color updates to represent synaptic weight for in-sensor computing. Here, we engineer a specialized mechanism for adaptively regulating ion doping through an ion-exchange membrane, enabling precise control over color-coded synaptic weight, an unprecedented achievement. The electrochromic neuromorphic transistor not only enhances electrochromatic capabilities for hardware coding but also establishes a visualized pattern-recognition network. Integrating the electrochromic neuromorphic transistor with an artificial whisker, we simulate a bionic reflex system inspired by the longicorn beetle, achieving real-time visualization of signal flow within the reflex arc in response to environmental stimuli. This research holds promise in extending the biomimetic coding paradigm and advancing the development of bio-hybrid interfaces, particularly in incorporating color-based expressions.

Multi-terminal pectin/chitosan hybrid electrolyte gated oxide neuromorphic transistor with multi-mode cognitive activities

Multi-terminal pectin/chitosan hybrid electrolyte gated oxide neuromorphic transistor with multi-mode cognitive activities

Reconfigurable optoelectronic transistors for multimodal recognition

Biological nervous system outperforms in both dynamic and static information perception due to their capability to integrate the sensing, memory and processing functions. Reconfigurable neuromorphic transistors, which can be used to emulate different types of biological analogues in a single device, are important for creating compact and efficient neuromorphic computing networks, but their design remains challenging due to the need for opposing physical mechanisms to achieve different functions. Here we report a neuromorphic electrolyte-gated transistor that can be reconfigured to perform physical reservoir and synaptic functions. The device exhibits dynamics with tunable time-scales under optical and electrical stimuli. The nonlinear volatile property is suitable for reservoir computing, which can be used for multimodal pre-processing. The nonvolatility and programmability of the device through ion insertion/extraction achieved via electrolyte gating, which are required to realize synaptic functions, are verified. The device’s superior performance in mimicking human perception of dynamic and static multisensory information based on the reconfigurable neuromorphic functions is also demonstrated. The present study provides an exciting paradigm for the realization of multimodal reconfigurable devices and opens an avenue for mimicking biological multisensory fusion.

Harness Background Lighting for Separability Enhancement in Optical Sensing Reservoir for Temporal Signal Processing Utilizing Amorphous IGZO Neuromorphic Transistors

Harness Background Lighting for Separability Enhancement in Optical Sensing Reservoir for Temporal Signal Processing Utilizing Amorphous IGZO Neuromorphic Transistors

Feedforward Photoadaptive Organic Neuromorphic Transistor with Mixed‐Weight Plasticity for Augmenting Perception

AbstractOrganic photoelectric neuromorphic devices that mimic the brain are widely explored for advanced perceptual computing. However, current individual neuromorphic synaptic devices mainly focus on utilizing linear models to process optoelectronic signals, which means that there is a lack of effective response to nonlinear structural information from the real world, severely limiting the computational efficiency and adaptability of networks to static and dynamic information. Here, a feedforward photoadaptive organic neuromorphic transistor with mixed‐weight plasticity is reported. By introducing the potential of the space charge to couple gate potential, photoexcitation, and photoinhibition occur successively in the channel under the interference of constant light intensity, which enables the device to transform from a linear model to a nonlinear model. As a result, the device exhibits a dynamic range of over 100 dB, exceeding the currently reported similar neuromorphic synaptic devices. Further, the device achieves adaptive tone mapping within 5 s for static information and achieves over 90% robustness recognition accuracy for dynamic information. Therefore, this work provides a new strategy for developing advanced neuromorphic devices and has great potential in the fields of intelligent driving and brain‐like computing.

Realization of High Mobility Synaptic Transistor through Control of Cross‐Linking Agent in a Polymer Dielectric Layer for Emerging Electric Double Layer

AbstractThe development of high‐mobility neuromorphic transistors is essential to increase signal transmission speed and solve the von Neumann bottleneck issue. Herein, this work proposes cross‐linked Poly(4‐vinylphenol) (c‐PVP) as a dielectric layer to form an electric double layer (EDL), which plays a key role in synaptic transistors, and measure the ratio‐dependent characteristics of cross‐linking agents in c‐PVP. When the ratio of PVP to poly(melamine‐co‐formaldehyde) methylated (PMF) is 10:1, neuromorphic transistors is found to show the best performance with a memory window of 2.2 V and a mobility of 93.4 cm2 V−1 s−1. Fourier‐transform infrared spectroscopy (FT‐IR) results show that the reduction in the concentration of the cross‐linking agent generates more hydroxyl groups within the c‐PVP film. The 10:1 c‐PVP‐based synaptic device has an ultra‐low energy consumption of 15.8 pJ for a single pulse and a maximum paired pulse facilitation (PPF) index value of 291%. Additionally, synaptic characteristics, such as pulse duration time dependent plasticity, pulse intensity dependent plasticity, pulse rate dependent plasticity (SRDP), high band filtering, and short‐term memory (STM) conversion to long‐term memory (LTM), are also described and discussed. These results suggest that c‐PVP/ZnON‐based neuromorphic devices can be promising artificial synapses for memory and learning capabilities.

Synaptic Plasticity Modulation of Neuromorphic Transistors through Phosphorus Concentration in Phosphosilicate Glass Electrolyte Gate.

This study proposes a phosphosilicate glass (PSG)-based electrolyte gate synaptic transistor with varying phosphorus (P) concentrations. A metal oxide semiconductor capacitor structure device was employed to measure the frequency-dependent (C-f) capacitance curve, demonstrating that the PSG electric double-layer capacitance increased at 103 Hz with rising P concentration. Fourier transform infrared spectroscopy spectra analysis facilitated a theoretical understanding of the C-f curve results, examining peak differences in the P-OH structure based on P concentration. Using the proposed synaptic transistors with different P concentrations, changes in the hysteresis window were investigated by measuring the double-sweep transfer curves. Subsequently, alterations in proton movement within the PSG and charge characteristics at the channel/PSG electrolyte interface were observed through excitatory post-synaptic currents, paired-pulse facilitation, signal-filtering functions, resting current levels, and potentiation and depression characteristics. Finally, we demonstrated the proposed neuromorphic system's feasibility based on P concentration using the Modified National Institute of Standards and Technology learning simulations. The study findings suggest that, by adjusting the PSG film's P concentration for the same electrical stimulus, it is possible to selectively mimic the synaptic signal strength of human synapses. Therefore, this approach can positively contribute to the implementation of various neuromorphic systems.

Polyvinyl alcohol electrolyte-gated oxide transistors with tetanization activities for neuromorphic computing

Polyvinyl alcohol electrolyte-gated oxide neuromorphic transistors with tetanization activities were fabricated, demonstrating an ultra-low energy consumption of 36 aJ and a quality factor of ∼0.0036 fJ V.

Stretchable and neuromorphic transistors for pain perception and sensitization emulation.

Pain perception nociceptors (PPN), an important type of sensory neuron, are capable of sending out alarm signals when the human body is exposed to destructive stimuli. Simulating the human ability to perceive the external environment and spontaneously avoid injury is a critical function of neural sensing of artificial intelligence devices. The demand for developing artificial PPN has subsequently increased. However, due to the application scenarios of bionic electronic devices such as human skin, electronic prostheses, and robot bodies, where a certain degree of surface deformation constantly occurs, the ideal artificial PPN should have the stretchability to adapt to real scenarios. Here, an organic semiconductor nanofiber artificial pain perception nociceptor (NAPPN) based on a pre-stretching strategy is demonstrated to achieve key pain aspects such as threshold, sensitization, and desensitization. Remarkably, while stretching up to 50%, the synaptic behaviors and injury warning ability of NAPPN can be retained. To verify the wearability of the device, NAPPN was attached to a curved human finger joint, on which PPN behaviors were successfully mimicked. This provides a promising strategy for realizing neural sensing function on either deformed or mobile electronic devices.

Polarized Tunneling Transistor for Ultralow-Energy-Consumption Artificial Synapse toward Neuromorphic Computing.

Neural networks based on low-power artificial synapses can significantly reduce energy consumption, which is of great importance in today's era of artificial intelligence. Two-dimensional (2D) material-based floating-gate transistors (FGTs) have emerged as compelling candidates for simulating artificial synapses owing to their multilevel and nonvolatile data storage capabilities. However, the low erasing/programming speed of FGTs renders them unsuitable for low-energy-consumption artificial synapses, thereby limiting their potential in high-energy-efficient neuromorphic computing. Here, we introduce a FGT-inspired MoS2/Trap/PZT heterostructure-based polarized tunneling transistor (PTT) with a simple fabrication process and significantly enhanced erasing/programming speed. Distinct from the FGT, the PTT lacks a tunnel layer, leading to a marked improvement in its erasing/programming speed. The PTT's highest erasing/programming (operation) speed can reach ∼20 ns, which outperforms the performance of most FGTs based on 2D heterostructures. Furthermore, the PTT has been utilized as an artificial synapse, and its weight-update energy consumption can be as low as 0.0002 femtojoule (fJ), which benefits from the PTT's ultrahigh operation speed. Additionally, PTT-based artificial synapses have been employed in constructing artificial neural network simulations, achieving facial-recognition accuracy (95%). This groundbreaking work makes it possible for fabricating future high-energy-efficient neuromorphic transistors utilizing 2D materials.

Triboelectric potential tuned oxide artificial tactile sensory platform with ultra-low power consumption

Sophistiated cognitive platform calls for multifunctional intelligent sensory devices with ultra-low power consumption. Triboelectric nanogenerator (TENG) provides an energy-efficient strategy to gate neuromorphic transistor with multiple sensing applications. To this end, a self-powered artificial tactile sensory neuron is proposed. Fish gelatin (GEL) and carboxylated-chitosan (CCs) are cross-linked to gate indium tin oxide (ITO) neuromorphic transistor. Thanks to the extremely strong protonic/electronic interfacial coupling at GEL-CCs based hydrogel/channel interface, good electrical performances are obtained at low voltage of only ∼0.8 V. The device demonstrates fundamental synaptic behaviors. Moreover, it exhibits ultra-low energy consumption of only ∼31.1 aJ for a single synaptic response with a good sensitivity of ∼4.65 dB. TENG is fabricated using polydimethylsiloxane elastomer and polytetrafluoroethylene, demonstrating good sensitivity of ∼8.0 V/kPa and good stabilities under pressure of 400 Pa. Thus, a tactile sensory platform is built with co-substrate configuration, integrating TENG and ITO neuromorphic transistor. Image edge detection is conceptually demonstrated by adopting Sobel algorithm. Furthermore, a wearable arterial pulse monitor is proposed based on the haptic sensory platform. The present work provides a candidate solution for realizing environment-friendly, energy-saving neuromorphic system and tactile perception system, demonstrating potentials in portable intelligent platform, prosthetics and human-machine interface.

Carbon nanotube neurotransistors with ambipolar memory and learning functions

In recent years, neuromorphic computing has gained attention as a promising approach to enhance computing efficiency. Among existing approaches, neurotransistors have emerged as a particularly promising option as they accurately represent neuron structure, integrating the plasticity of synapses along with that of the neuronal membrane. An ambipolar character could offer designers more flexibility in customizing the charge flow to construct circuits of higher complexity. We propose a novel design for an ambipolar neuromorphic transistor, utilizing carbon nanotubes as the semiconducting channel and an ion-doped sol–gel as the polarizable gate dielectric. Due to its tunability and high dielectric constant, the sol–gel effectively modulates the conductivity of nanotubes, leading to efficient and controllable short-term potentiation and depression. Experimental results indicate that the proposed design achieves reliable and tunable synaptic responses with low power consumption. Our findings suggest that the method can potentially provide an efficient solution for realizing more adaptable cognitive computing systems.Impact statementThe huge amount of data generated by the current society makes it necessary to explore new computing methods with higher efficiency to overcome the bottleneck formed between data storage and processing tasks. Neuromorphic computing aims at emulating the functioning of our brain, which performs both tasks utilizing the same hardware. Here, we propose ambipolar field-effect transistors based on carbon nanotubes with a polarizable gate dielectric, capable of providing memory functions reminiscent of neuronal synapses, at both polarities of the device. The ambipolar characteristic doubles the possibilities of previously demonstrated neurotransistors. The short-term and ambipolar behavior of the device can find its place in novel applications in the future. Machine learning-enabled gas sensing is an excellent example, where real-time processing of large amounts of data is beneficial. In addition, interaction with oxidative and reductive gases will result in dual responses due to the ambipolarity of the transistor, along with the possibility of storing the sensing data.Graphical abstract

Brain‐Inspired Organic Electronics: Merging Neuromorphic Computing and Bioelectronics Using Conductive Polymers

AbstractNeuromorphic computing offers the opportunity to curtail the huge energy demands of modern artificial intelligence (AI) applications by implementing computations into new, brain‐inspired computing architectures. However, the lack of fabrication processes able to integrate several computing units into monolithic systems and the need for new, hardware‐tailored training algorithms still limit the scope of application and performance of neuromorphic hardware. Recent advancements in the field of organic transistors present new opportunities for neuromorphic systems and smart sensing applications, thanks to their unique properties such as neuromorphic behavior, low‐voltage operation, and mixed ionic‐electronic conductivity. Organic neuromorphic transistors push the boundaries of energy efficient brain‐inspired hardware AI, facilitating decentralized on‐chip learning and serving as a foundation for the advancement of closed‐loop intelligent systems in the next generation. The biocompatibility and dual ionic‐electronic conductivity of organic materials introduce new prospects for biointegration and bioelectronics. Their ability to sense and regulate biosystems, as well as their neuro‐inspired functions can be combined with neuromorphic computing to create the next‐generation of bioelectronics. These systems will be able to seamlessly interact with biological systems and locally compute biosignals in a relevant matter.

Biodegradable Oxide Neuromorphic Transistors for Neuromorphic Computing and Anxiety Disorder Emulation.

Brain-inspired neuromorphic computing and portable intelligent electronic products have received increasing attention. In the present work, nanocellulose-gated indium tin oxide neuromorphic transistors are fabricated. The device exhibits good electrical performance. Short-term synaptic plasticities were mimicked, including excitatory postsynaptic current, paired-pulse facilitation, and dynamic high-pass synaptic filtering. Interestingly, an effective linear synaptic weight updating strategy was adopted, resulting in an excellent recognition accuracy of ∼92.93% for the Modified National Institute of Standard and Technology database adopting a two-layer multilayer perceptron neural network. Moreover, with unique interfacial protonic coupling, anxiety disorder behavior was conceptually emulated, exhibiting "neurosensitization", "primary and secondary fear", and "fear-adrenaline secretion-exacerbated fear". Finally, the neuromorphic transistors could be dissolved in water, demonstrating potential in "green" electronics. These findings indicate that the proposed oxide neuromorphic transistors would have potential as implantable chips for nerve health diagnosis, neural prostheses, and brain-machine interfaces.