A biomimetic afferent nervous system based on the flexible artificial synapse
Graphical Abstract
Herein, a flexible biomimetic afferent nervous system (ANS) is constructed by fusing the biological ion migration in tactile organ with the MXene artificial synapse’ working mechanism. It is exploited to mimic the famous operant conditioning flex (OCF) behaviors, for the first time. Moreover, not only the common joint flexion behaviors and weak pulse at radial artery can be detected, but also the plastic neural response can be simulated.
Introduction
Afferent nervous system (ANS) is an important part of the biological reflex system, through which the receptor and nerve center are the main components in ANS [1], [2], [3]. The receptors in the biological ANS, entailing the epidermis, dermis, and subcutaneous layers (Fig. 1a), are used for feeling and transmitting the external stimuli (pressure, temperature, etc.) to the nerve center, and to realize the organism’s complex perceptual and self-protect functions [4], [5]. It is accepted in neuroscience that an artificial ANS can provide new methods to study and treat a variety of diseases, like epilepsy, chronic pain, depression and Parkinson’s disease, so great enthusiasms have been ignited by the idea of building a biomimetic ANS [6], [7]. However, in most of the biomimetic ANS studies, they were still limited by using electric pulses to mimic the external stimulation and study the object’s response to it. Making the real outside stimuli (heat or touch) directly generate impulses is still a challenge in the field of biomimetic ANS. The emerging artificial synapses provide a great opportunity for answering this challenge. Synapse is a key part in nervous system, it can make bridges between receptors and the nerve center. Artificial synapses have been extensively studied, due to their powerful capabilities in signal processing [8], [9], integrability [10], [11], and low power consumption [12]. If a synapse-based biomimetic ANS can be built, it can not only contribute a pertinent technique breakthrough to the traditional biomimetic ANS method, but also provide a new tool for building a high-efficient and low-power consumption human-computer interface in the upcoming bionic intelligent robot [3], [13], [14].
Mechanoreceptors are the conventional focuses in the early-stage artificial ANS, and considerable progresses have been made in sensitivity [15] and signal response time [16], [17]. Recent researches in biological receptors indicated the ion migration in the porous scaffolding protein of Merkel cells in tactile organ, named as Merkel cell-neurite complexes (MCNCs), was the main procedure to generate the nerve impulse in creature, it was deduced to be triggered by MCNCs’ elastic behavior [18]. So, an ion-elastic polymer mediated porous background may be a good solution for constructing efficient mechanoreceptor [19], nevertheless, the ion migration is so sensitive that it can be disturbed by external elements (like electric field or defects) [20], how to manipulate it is still a challenge in fabricating artificial ANS. Meanwhile, inspired by the function of biological ANS, a wealth of bionic tactile system was developed by incorporating mechanoreceptors with signal-processing chips, in order to endow them with intelligence. For example, an ANS was proposed by combining a piezoresistive sensor with organic ring oscillators and a synaptic transistor, to fulfill the monosynaptic reflex arc [21]. And the concept of neuromorphic tactile processing system includes a pressure sensor, soft ionic cable and synaptic transistor, was put forward to mimic the sensory neuron [22], such as the recently reported artificial optoelectronic spiking ANS which fulfilled the functions of tactile, neural coding, perceptual learning and memory by combining pressure sensor, ADC-LED and optoelectronic synapse [23]. In those assembling tactics, the tactile sensing and signal transferring still depended on the separate devices, which made them complicated. Therefore, simplifying the architecture of biomimetic ANS is imperative. Recently, the unified tactile perception and transmission were accomplished by using the ferroelectric nanocomposite dielectric gated organic field effect transistor [19]. In that work, the triboelectric-capacitive coupling effect triggered post-synaptic currents could be modulated by touching, and the sensory memory behavior was successfully simulated, which paved a new way for simplifying bionic ANS. However, the above-mentioned researches about ANS do not involve the operant conditioning flex (OCF) behavior, which is an important part of associative learning process in neuroscience. Moreover, to the best of our knowledge, the simulation of real irritation (heat or touch) triggered OCF behavior is still a challenge in the field of neuromorphic devices.
Herein, a chip-like biomimetic neuromorphic ANS in this work is proposed and shown in Fig. 1b, which is obviously distinguished from the current reported artificial ANS. It is a disassembled structure of the proposed ANS chip, there are 4 parts from top to down briefly, i.e., the top electrode layer, ionic conductive elastomer (ICE) layer, the MXene layer, and the bottom electrode layer. The layers of MXene, top and bottom electrodes construct a MXene-based artificial synapse, which not only can transmit the external stimulations (pressure or temperature) felt by ICE, but also the mimic of biological neuro response can be fulfilled. ICE is a porous complex of thermoplastic polyurethane elastomer (TPU) and 1-Ethyl-3-methylimidazolium Tetrafluoroborate (EMIMBF4), as illustrated by the inset in Fig. 1b. The positive (EMIM+) and negative (BF4-) charges of EMIMBF4 can be moved by the external stimuli, the ionic transportation through polymer chains can closely resemble the ion migration in tactile organ [24]. Furthermore, the externally triggered electronic variation in ICE layer can be directly coupled to “Artificial synapse” by the trap-assisted tunneling (TAT) effect, so that the stimuli felt by ICE can be transformed to neuro-electric signal. Briefly, in this layer-by-layer fabricated artificial ANS, the porous ICE is used to mimic MCNCs’ biological ionic migration, simultaneously, few-layer MXene based artificial synapse beneath it can transform the excited electronic changes to synaptic currents. In addition, the applications of the proposed device are also exploited as biomimetic ANS, as well as to detect the wide range joint motions and subtle heart rate. In particular, an interesting “operant conditioned reflex behavior” by using the typical “trial and error learning” under the real irritation (heat or touch) are also successfully simulated, for the first time. We wish this work can open new possibilities for developing high-efficient neuromorphic interface, then a total nervous system could be expected in the upcoming intelligent robot.
Section snippets
Results and discussion
MXene-Ti3C2Tx, a promising two-dimensional electronics material [25], [26], has been widely used for artificial synapse devices owing to the low power consumption [27], small size of device structure[28], fast modulation time [29] and good flexibility [30]. We think, few layer MXene based artificial synapse as depicted in Fig. 1b should be a promising candidate to simulate nerve impulse in biological ANS, so first of all, flexible substrate (polyethylene terephthalate, PET) supported synapse
Conclusion
In summary, a chip-like flexible ANS device by fusing ICE with the MXene artificial synapse is constructed and demonstrated. The MXene layer in ANS device can help to reduce the ion migration dispersity in ICE layer, so as to improve current’s signal-to-noise ratio and generating stable external stimulus modulated synaptic current. Therefore, the proposed ANS device can not only successfully to generate nervous response to real irritation (heat or touch), but also accomplish the detections of
Synthesis of the MXene Ti3C2Tx nanosheets
Firstly, 0.333 g of HF, 2.5 mL 12 M hydrochloric acid, and 0.5 g Ti3AlC2 powder are mixed; then 2.5 mL deionized water are added in the mixture and purged with N2 for 20 min to replace oxygen. After stirring for 24 h in silicon oil bath (40 °C), the resulting solution is centrifuged (4000 rpm / min) in 40 mL DI water for 5 min. After five times centrifugation, the obtained material is dried in the constant temperature drying oven at 80 °C for 24 h. The multi-layer Ti3C2Tx nanosheets are
CRediT authorship contribution statement
Kaiyang Wang: Investigation, Methodology, Data curation, Writing – original draft. Yunfang Jia: Conceptualization, Resources, Validation, Supervision, Project administration, Funding acquisition, Writing – review & editing. Xiaobing Yan: Conceptualization, Supervision, Formal analysis, Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This work is supported by the National Natural Science Foundation of China (Grant no. 61771260).
Kai-Yang Wang is a Ph.D. candidate from the College of Electronic Information and Optical Engineering, Nankai University, Tianjin, China. He obtained his M.S. degree from Hebei University, Baoding, China. His current research interests include preparation and integration of neuromorphic sensory device and wearable sensors.
References (50)
- et al.
An artificial spiking afferent nerve based on Mott memristors for neurorobotics
Nat. Commun.
(2020) - et al.
Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots and peripheral nerves: Basic principles and procedures for routine clinical and research application. An updated report from an I.F.C.N. Committee
Clin. Neurophysiol.
(2015) - et al.
Analogue signal and image processing with large memristor crossbars
Nat. Electron.
(2017) - et al.
Merkel cells transduce and encode tactile stimuli to drive Abeta-afferent impulses
Cell
(2014) - et al.
Ionic elastomers for electric actuators and sensors, engineering
(2021) - et al.
Tunable synaptic behavior realized in C3N composite based memristor
Nano Energy
(2019) Proprioception and locomotor disorders
Nat. Rev. Neurosci.
(2002)- et al.
The microbiome extends to subepidermal compartments of normal skin
Nat. Commun.
(2013) - et al.
Bioinspired microspines for a high-performance spray Ti3C2Tx MXene-based piezoresistive sensor
ACS Nano
(2020) - et al.
Pursuing prosthetic electronic skin
Nat. Mater.
(2016)
tACS motor system effects can be caused by transcutaneous stimulation of peripheral nerves
Nat. Commun.
Current status and prospects of memristors based on novel 2D materials
Mater. Horiz.
A fully integrated reprogrammable memristor–CMOS system for efficient multiply–accumulate operations
Nat. Electron.
Fully hardware-implemented memristor convolutional neural network
Nature
Self-assembled networked PbS distribution quantum dots for resistive switching and artificial synapse performance boost of memristors
Adv. Mater.
Piezotronic graphene artificial sensory synapse
Adv. Funct. Mater.
Vertical organic synapse expandable to 3D crossbar array
Nat. Commun.
Skin-like pressure and strain sensors based on transparent elastic films of carbon nanotubes
Nat. Nanotechnol.
Epidermis microstructure inspired graphene pressure sensor with random distributed spinosum for high sensitivity and large linearity
ACS Nano
A self-powered sensor mimicking slow and fast-adapting cutaneous mechanoreceptors
Adv. Mater.
A flexible artificial intrinsic-synaptic tactile sensory organ
Nat. Commun.
Ion migration in organometal trihalide perovskite and its impact on photovoltaic efficiency and stability
Acc. Chem. Res.
A bioinspired flexible organic artificial afferent nerve
Science
An artificial sensory neuron with tactile perceptual learning
Adv. Mater.
Tactile sensory coding and learning with bio-inspired optoelectronic spiking afferent nerves
Nat. Commun.
Cited by (17)
Novel 2D MXene-based materials in memristors: Fundamentals, resistive switching properties and applications
2024, Surfaces and InterfacesRecent advances in halide perovskite memristors: From materials to applications
2024, Frontiers of PhysicsAn Ultrasensitive Biomimetic Optic Afferent Nervous System with Circadian Learnability
2024, Advanced ScienceFlexible Zn-TCPP Nanosheet-Based Memristor for Ultralow-Power Biomimetic Sensing System and High-Precision Gesture Recognition
2024, Advanced Functional Materials
Kai-Yang Wang is a Ph.D. candidate from the College of Electronic Information and Optical Engineering, Nankai University, Tianjin, China. He obtained his M.S. degree from Hebei University, Baoding, China. His current research interests include preparation and integration of neuromorphic sensory device and wearable sensors.
Prof. Yun-Fang Jia received the Ph.D. degree in Nankai University in 2004. She is currently a professor with College of Electronic Information and Optical Engineering, Nankai University, Tianjin, China. Her current research interests are novel electronic sensitive devices and their applications for biochemistry.
Prof. Xiao-Bing Yan received the Ph.D. degree in Nanjing University in 2011. He is currently a professor with the Key Laboratory of Brain-Like Neuromorphic Devices and Systems of Hebei Province, College of Electron and Information Engineering, Hebei University, China. His current research interests include atomic layer deposition of ferroelectric materials, the fabrication and integration of neuromorphic device, resistive switching device, memristor, and so on.