| The rapid development of artificial intelligence has changed the way humans interact with the world.Electronic sensors,mimicking biological sensing systems,play a key role in artificial intelligence technology.Among all feelings,haptics undertakes most of the information feeling for learning and memory.It is an important pathway for people to know and understand the world.However,currently developed electronic haptic sensors with poor resolution,low sensitivity,and rigid form,cannot meet the requirements of intelligent information pre-processing and wearability comfort.In addition,an external power supply is needed to drive the haptic sensors,which increases the system’s energy consumption.Moreover,the spatial separation of haptic sensors and neuromorphic devices limits the highly integrated application of intelligent sensing networks.Therefore,to address the above research difficulties,this dissertation presents an artificial tactile system by integrating self-powered tactile sensors and organic synapstic transistors in a flexible fashion.Such neurophic systems are capable of sensing,processing,encoding,and recognizing tactile information effectively.The main research content includes the following three aspects:(1)Study of flexible skin-integrated triboelectric nanogenerators(TENG)as self-powered tactile sensors.To solve the research difficulties of traditional TENG(such as its rigid state,low array density,and difficulity to prepare on a large scale),this work fabricated a flexible stretchable TENG by optimizing the functional material and preparation processes.Polydimethylsiloxane(PDMS)with a 3D interconnected structure was used as the friction layer,and silver nanowires were used as the stretchable electrodes with the aid of spraying and blade coating.Results demonstrated that an open-circuit voltage of 78.7 V,an output power density of 32.6 m W/m~2,a pressure sensitivity of 5.56V/N,and 1.56μA/N were obtained.Furthermore,TENG sensors maintained stable operation under various deformations,such as stretching,bending,twisting,and folding.Furthermore,improved output performances were obtained by introducing the 3D interconnected structure of PDMS as the friction layer.Various TENG arrays were prepared for the applications of electronic shoe insoles and gloves.Sensitive and crosstalk-free haptic information mapping was obtained.It provides the theoretical and design foundation for using such TENG tactile sensors in self-powered neuromorphic tactile sensory systems.(2)Study of flexible,skin-integrated organic electrochemical transistors(OECTs)as synaptic transistors.To solve the research difficulties of traditional OECT(such as rigid state,low array density,and poor biocompatibility),this study prepared a biocompatible100-channel OECT synaptic transistor via multilayer photolithography processes.An ultrathin thickness of 15μm,a light weight of 99.3 mg,and a minimum bending radius of less than 5 mm were achieved.The great mechanical properties enable such an OECT array to seamlessly laminate onto curved surfaces.In this design,poly(lactic acid-hydroxyacetic acid)copolymer,gold,and poly(3,4-ethylenedioxythiophene)polystyrene sulfonate(PEDOT:PSS)were utilized as substrates,source-drain electrodes,and active channels,respectively.An average transconductance of 9.8 m S was obtained,with 100%array yield.In demonstrated examples,these OECT arrays could accurately record neural spike information,including anesthesia,seizures,and electrical stimulation.It provides the theoretical and design foundation for using such OECT synaptic transistors in self-powered neuromorphic tactile sensory systems.(3)Study of a flexible,stretchable neuromorphic tactile sensory system based on the integration of TENG tactile sensors and OECT synaptic transistors.To solve the problems of requiring other power supplies,liquid electrolyte,low integration,and poor flexibility),inspired by biological neural networks,the TENG tactile sensor simulates biological mechanoreceptors,which convert the captured tactile information into electrical signals.Then OECT synaptic transistors simulate biological synapses,which convert the TENG output signals(i.e.,presynaptic voltage signals)into postsynaptic currents.The power consumption is about 6.15μJ/spike and 5.8μJ/spike in the depression and facilitation behavior modes,respectively.Advanced synaptic plasticity behaviors were achieved,including postsynaptic currents,paired-pulse facilitation/depression,and short-and long-term synaptic plasticity.Meanwhile,the integration scheme of stretchable PDMS,liquid metal electrodes,hydrogel electrolytes,and PEDOT:PSS semiconductor materials enables the TENG tactile sensors and OECT synapses to own skin-conformability,great wearability,and high stretchability up to 100%.In the demonstrated examples of a Morse code reader and handwritten information human-machine interface with a size of 5×5,advanced intelligent functions of sensing,processing,feature extraction,and encoding are proposed.It is expected to pave the way for promising next-generation intelligent robots and advanced alternative prostheses.In summary,this dissertation presents a self-powered,flexible,neuromorphic tactile sensory system with neural encoding,perceptual learning,and memory abilities.Inspired by biological nervous systems,flexible TENG and OECT are developed to mimic the biological mechanoreceptors and synapses,and tactile perception and information processing functions are achieved.This dissertation is expected to provide a solid theoretical and application foundation for neuromorphic electrochemistry in artificial intelligence. |
