Optimization on Electrical Performance of Advanced Self-Powered Sensing Systems

Xie, Xinkai (2023) Optimization on Electrical Performance of Advanced Self-Powered Sensing Systems. PhD thesis, University of Liverpool.

Text 201470033_Xinkai Xie_2023.pdf - Author Accepted Manuscript Access to this file is embargoed until 1 August 2024. Download (7MB)

Abstract

As the essential core component for the precise sensing of external signals, sensors are the key nodes for the exchange of information between electronic devices and control terminals. Recently, the intelligent Internet of Things (IoT) has developed rapidly with its increasing requirements for large-scale, multi-functional sensor networks. Multi-channel integrated and independent sensors could connect the user entity to the Internet by establishing a sensory information network, which could be applied to smart home, human-machine interfaces, environmental monitoring, and healthcare. However, most of present sensors use external power supplies or traditional batteries. Large and complex power supply systems are becoming a bottleneck limiting its development. The birth of triboelectric nanogenerator (TENG) provides an effective solution to distributed energy supply and self-powered sensing. The electrical output signals of TENG, such as frequency and amplitude, directly reflecting external environmental stimuli with high sensitivity, could allow a variety of active self-powered physical or chemical sensing by monitoring the output parameters of voltage, current or transferred charge. However, the electrical performance of TENG based self-powered sensing still needs to be improved with limitations of external monitoring circuits, integration with artificial intelligence systems, and optimization of its stability. Therefore, based on the intrinsic capacitor model of TENG, establishing an impedance matching effect induced completely self-powered sensing system without external monitoring circuits, which could be utilized for quantitative sensing demonstration with linear response, is the first key point of this thesis. To broaden the applications of self-powered triboelectric sensors in artificial intelligence of things (AIoT), integration with synaptic transistor-based neuromorphic computation to analyze and identify complex output signals for apply human-machine interfaces is the second key point of this thesis. Considering the kinetics of surface charge generation, accumulation and transfer in different layers of TENG, it is essential to maintain the surface charges for a prolonged duration for enhanced stability of triboelectric sensors. Therefore, an intermediate layer-based TENG structure is constructed with in-depth working mechanism investigation through the energy band theory to extend the surface charge storage time and enhance the stability of TENG, which is the third key point of this thesis. Based on the above study points, the main research results of this thesis are as follows. In this work, to construct a fully self-powered sensing system without external monitoring circuits, we prepared a self-powered gyroscope angle sensor based on the resistive impedance matching effect of TENG. It contains a disc shaped free-standing mode TENG, a resistive rotary potentiometer (RRP), and an LED alert display reflecting the rotation angle. It displays high linearity response when rotation angle varies from 0° to 260°, ultra-high sensitivity of 67.3 mV deg-1, and fast response within 20 ms, which could be utilized as an attitude indicator in the unmanned aerial vehicle (UAV) flight control system. In addition, the corresponding LED pattern can be illuminated when the relative rotation angle equals 0o, 90o and 180o, enabling a quantitative sensing alarm function (Chapter 2). Secondly, to broaden the applications of self-powered triboelectric sensors in AIoT, a liquid metal-based dual-mode triboelectric-capacitive coupled tactile sensor (TCTS) array was prepared and integrated with the machine-learning assisted human-machine interaction device. Each sensing unit exhibits capacitance changing from 5.4 pF to 19.1 pF (0-80 kPa) and achieves high triboelectric output response sensitivity of 7.88 kPa-1 in the biological tactile sensing range (0–8.78 kPa). For applications, the flexible 4*4 pixels array implements a visual mapping output for static pressure distribution and the synaptic transistor-based neuromorphic computing mode realizes a 100% recognition accuracy in training 10 classes of handwriting numbers within 80 epochs. Through the incorporation with a signal-processing circuit and wireless interactive display, the perceived tactile information is successfully projected into the mixed reality space, allowing real-time somatosensory translation for Traditional Chinese Medicine physiotherapy perception, which provides a more realistic and immersive experience in converged real and virtual environments (Chapter 3). Finally, to investigate the surface charge modulation mechanism and optimize the stability of TENG, a solution-processed LaZrO high permittivity electron blocking layer (HPEBL) was embedded between the bottom electrode ITO and the triboelectric layer PDMS in a contact-separation mode TENG to create an energy barrier (ΔE > 1.3 eV). The KPFM characterization results shows a 3.1-fold extension of the surface half-charge decay time. The open circuit voltage, short circuit current and transferred charge of the TENG based on La0.1Zr0.9Ox HPEBL reach 215 V, 96.3 mA m-2 and 243.3 μC m-2, respectively. When the load resistance is 100 MΩ, the average mass/volume power density and energy conversion efficiency can be calculated as 59.34 μW g-1, 152.5 μW cm-3 and 39.2%. In addition, metal-insulator-metal (MIM) devices were constructed to study the dielectric behavior of doped LaZrO films with different concentrations. According to the simulation results, Poole-Frenkel (PF) emission was found to dominate the leakage mechanism, where high doping concentrations could result in an increment in the oxygen vacancy concentration and leakage current. Therefore, increasing the relative permittivity and decreasing the oxygen vacancy density to optimize HPEBL provides a general route to construct the stable and high-performance TENG (Chapter 4).

Item Type:	Thesis (PhD)
Uncontrolled Keywords:	triboelectric nanogenerator, self-powered sensing, neuromorphic computation, artificial intelligence application, stability optimization
Divisions:	Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User:	Symplectic Admin
Date Deposited:	25 Aug 2023 15:20
Last Modified:	25 Aug 2023 15:20
DOI:	10.17638/03170947
Supervisors:	Zhao, Chun Yang, Li Tu, Xin
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3170947