Magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator for advanced human-machine interfaces

✅ 全文

磁相互作用辅助的混合摩擦电-电磁纳米发电机用于先进人机界面

作者 Long Liu; Qiongfeng Shi; Zhongda Sun; Chengkuo Lee 期刊 Nano Energy 发表日期 2021 DOI 10.1016/j.nanoen.2021.106154 类型 原创研究 (Original Research)

📄 英文摘要 English Abstract

EN

Abstract A magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator (MAHN) is designed and optimized with two magnets in attraction, and a silicone-based cushion with microstructures. The whole device including a fixed part and a moving part has a small size of 35 mm × 35 mm × 8 mm. After various energy harvesting and sensing characterizations, the MAHN has been verified as an efficient energy harvester, as well as a self-powered sensor for advanced human-computer interface (HMI). For energy harvesting, the triboelectric nanogenerator (TENG) delivers a peak power of 6.89 mW at the load of 1.1 MΩ, and the electromagnetic generator (EMG) delivers a peak power of 2.7 mW at the load of 1.1 kΩ. With the hybridized buck circuit, outputs from both the TENG and the EMG can be rectified and integrated efficiently to charge capacitors. A capacitor of 100 μF is charged to 2.03 V within 30 s, which is better than the performances of individual generators and their parallel-connection with only rectifiers. A thermometer and a Bluetooth module are powered with different amounts of the attracted magnet, implying the wide adaptability of the MAHN. Advanced applications of HMI have also been developed with cross-divided electrodes in the MAHN, such as orientation control in a game of Snake, real-time operation of a PowerPoint presentation, and recognition of simple air gestures in contactless control. Moreover, a 3 × 3 array is prepared and signal channels are simplified by traversal method without loss of functions to achieve higher-level control. A virtual football game is demonstrated by a shoe moving on the array, and different kinds of shootings have been mapped to the player in the game.

📄 中文摘要 Chinese Abstract

中文
本研究旨在探索一种面向普通用户的可行触觉设备形态。我们提出HAPmini,一种新型可握持触觉设备,能够增强用户的触摸交互体验。为实现这一增强效果,HAPmini在设计上具有较低的机械复杂度、较少的执行器以及简单的结构,同时仍能为用户提供力反馈和触觉反馈。 早期的触觉设备在几十年前开发,体积庞大且结构复杂,不适合日常使用。已有多种类型的触觉设备被提出,涵盖从带有机械臂的重型固定设备到外骨骼、盲文式触觉显示器,以及利用振动或电粘附的轻量嵌入式设备。这些设备能够再现丰富的触觉感受,例如在触摸物体时的排斥力或摩擦力,或其弹性或质地。然而,商业上可用的设备在机械结构上比文献中的设备简单得多。为了在保持相对简单机械结构的同时再现超越简单嗡嗡声的更逼真的触觉感受,触觉设备已被开发为在多个自由度(DOF)或多个维度上运行;它们需要包含多个执行器和精密的机械结构。因此,降低触觉设备的机械复杂度可能对普通用户有益,使其能够享受更精简、更易获取的触觉体验。

📋 英文结构化总结 English Structured Summary

全文整理

EN

Background:

This study aims to explore a feasible form of a haptic device for common users. We propose HAPmini, a novel graspable haptic device that enhances the user’s touch interaction. To achieve this enhancement, the HAPmini is designed with low mechanical complexity, few actuators, and a simple structure, while still providing force and tactile feedback to users.

Early haptic devices, developed a few decades ago, were extremely large and intricate for daily use. Various types of haptic devices have been proposed, ranging from heavy and grounded devices with robot arms to exoskeletons, braille-type tactile displays, as well as light and embedded devices that use vibrations or electroadhesion. These devices reproduce rich haptic sensations, such as repulsive forces or frictional forces, while touching an object, or its elasticity or texture in a manner. However, commercially available devices have a significantly simpler mechanical structure than those from the literature. To reproduce a more plausible haptic sensation beyond the simple buzzing sensation while maintaining a relatively simple mechanical structure, haptic devices have been developed to operate in multiple degrees of freedom (DOFs) or multiple dimensions; they are required to comprise multiple actuators and sophisticated mechanical structures. Therefore, reducing the mechanical complexity of haptic devices may be beneficial for common users by enabling a more streamlined and accessible haptic experience.

Methods:

HAPmini is a joystick-type input and haptic device that responds to a user’s two-dimensional touch interaction through 1-DOF actuation. HAPmini provides two types of haptic feedback (force and tactile) in response to a user’s two-dimensional manipulation on a touchscreen. It consists of a solenoid-magnet actuator and spiral spring. The actuator consists of a solenoid and a permanent magnet in coaxial alignment, which enables the user to manipulate the HAPmini in a tangential two-dimensional plane. At this point, if the polarity of the voltage applied to the solenoid changes, a pushing or pulling force can be applied in the direction of user operation. This enables instantaneous physically capture of a user’s operation via force feedback and the creation of various tactile sensations via vibrotactile feedback. We implemented a magnetic snap function (hardware magnetic snap function) that can help the user’s pointing task using force feedback and a function that simulates various textures on the touchscreen using this vibrotactile feedback (virtual texture). In this study, five virtual textures (i.e., reproductions of the textures of paper, jean, wood, sandpaper, and cardboard) were designed for HAPmini. Both HAPmini functions were tested in three experiments. First, a comparative experiment was conducted, and it was confirmed that the hardware magnetic snap function could increase the performance of pointing tasks to the same extent as the software magnetic snap function could, which is commonly used in graphical tools. Second, ABX and matching tests were conducted to determine whether HAPmini could generate the five virtual textures, which were designed differently and sufficiently well for the participants to be distinguished from each other.

Results:

The correctness rates of the ABX and the matching tests were 97.3% and 93.3%, respectively. The results confirmed that the participants could distinguish the virtual textures generated using HAPmini. The experiments indicate that HAPmini enhances the usability of touch interaction (hardware magnetic snap function) and also provides additional texture information that was previously unavailable on the touchscreen (virtual texture).

Data Summary:

The correctness rates of the ABX and the matching tests were 97.3% and 93.3%, respectively. In this study, five virtual textures (i.e., reproductions of the textures of paper, jean, wood, sandpaper, and cardboard) were designed for HAPmini. The hardware magnetic snap function could increase the performance of pointing tasks to the same extent as the software magnetic snap function could, which is commonly used in graphical tools.

Conclusions:

The experiments indicate that HAPmini enhances the usability of touch interaction (hardware magnetic snap function) and also provides additional texture information that was previously unavailable on the touchscreen (virtual texture).

Practical Significance:

HAPmini enhances the usability of touch interaction (hardware magnetic snap function) and also provides additional texture information that was previously unavailable on the touchscreen (virtual texture). Reducing the mechanical complexity of haptic devices may be beneficial for common users by enabling a more streamlined and accessible haptic experience.

📋 中文结构化总结 Chinese Structured Summary

中文

背景:

本研究旨在探索一种面向普通用户的可行触觉设备形态。我们提出HAPmini,一种新型可握持触觉设备,能够增强用户的触摸交互体验。为实现这一增强效果,HAPmini在设计上具有较低的机械复杂度、较少的执行器以及简单的结构,同时仍能为用户提供力反馈和触觉反馈。

早期的触觉设备在几十年前开发,体积庞大且结构复杂,不适合日常使用。已有多种类型的触觉设备被提出,涵盖从带有机械臂的重型固定设备到外骨骼、盲文式触觉显示器,以及利用振动或电粘附的轻量嵌入式设备。这些设备能够再现丰富的触觉感受,例如在触摸物体时的排斥力或摩擦力,或其弹性或质地。然而,商业上可用的设备在机械结构上比文献中的设备简单得多。为了在保持相对简单机械结构的同时再现超越简单嗡嗡声的更逼真的触觉感受,触觉设备已被开发为在多个自由度(DOF)或多个维度上运行;它们需要包含多个执行器和精密的机械结构。因此,降低触觉设备的机械复杂度可能对普通用户有益,使其能够享受更精简、更易获取的触觉体验。

方法:

HAPmini是一种操纵杆式输入和触觉设备,通过1自由度执行器响应用户的二维触摸交互。HAPmini在用户于触摸屏上进行二维操作时提供两种类型的触觉反馈(力反馈和触觉反馈)。它由螺线管-磁铁执行器和螺旋弹簧组成。该执行器由同轴排列的螺线管和永磁体组成,使用户能够在切向二维平面上操控HAPmini。此时,如果施加到螺线管上的电压极性发生变化,则可以在用户操作方向上施加推力或拉力。这使得能够通过力反馈即时物理捕捉用户的操作,并通过振动触觉反馈创造各种触觉感受。我们实现了一种磁吸功能(硬件磁吸功能),可利用力反馈辅助用户的指向任务,以及一种利用该振动触觉反馈模拟触摸屏上各种纹理的功能(虚拟纹理)。在本研究中,为HAPmini设计了五种虚拟纹理(即纸张、牛仔布、木材、砂纸和纸板纹理的再现)。HAPmini的两项功能均在三个实验中进行了测试。首先进行了对比实验,确认硬件磁吸功能能够将指向任务的性能提升至与图形工具中常用的软件磁吸功能相同的程度。其次,进行了ABX测试和匹配测试,以确定HAPmini是否能够生成五种设计各异且足够清晰、使参与者能够相互区分的虚拟纹理。

结果:

ABX测试和匹配测试的正确率分别为97.3%和93.3%。结果确认参与者能够区分使用HAPmini生成的虚拟纹理。实验表明,HAPmini增强了触摸交互的可用性(硬件磁吸功能),并提供了触摸屏上此前无法获得的额外纹理信息(虚拟纹理)。

数据摘要:

ABX测试和匹配测试的正确率分别为97.3%和93.3%。在本研究中,为HAPmini设计了五种虚拟纹理(即纸张、牛仔布、木材、砂纸和纸板纹理的再现)。硬件磁吸功能能够将指向任务的性能提升至与图形工具中常用的软件磁吸功能相同的程度。

结论:

实验表明,HAPmini增强了触摸交互的可用性(硬件磁吸功能),并提供了触摸屏上此前无法获得的额外纹理信息(虚拟纹理)。

实际意义:

HAPmini增强了触摸交互的可用性(硬件磁吸功能),并提供了触摸屏上此前无法获得的额外纹理信息(虚拟纹理)。降低触觉设备的机械复杂度可能对普通用户有益,使其能够享受更精简、更易获取的触觉体验。

📖 英文全文 English Full Text

EN

Nano Energy 86 (2021) 106154 Available online 13 May 2021

2211-2855/© 2021 Elsevier Ltd. All rights reserved.

Full paper Magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator for advanced human-machine interfaces

Long Liu a,b,c,d, Qiongfeng Shi a,b,c,d, Zhongda Sun a,b,c,d, Chengkuo Lee a,b,c,d,e,* a Department of Electrical and Computer Engineering, National University of Singapore, 4 Engineering Drive 3, 117576, Singapore b Center for Intelligent Sensors and MEMS, National University of Singapore, Block E6 #05-11, 5 Engineering Drive 1, 117608, Singapore c Hybrid-Integrated Flexible (Stretchable) Electronic Systems Program (HIFES), National University of Singapore, Block E6 #05-3, 5 Engineering Drive 1, 117608,

Singapore d NUS Suzhou Research Institute (NUSRI), Suzhou Industrial Park, Suzhou 215123, PR China e NUS Graduate School - Integrative Sciences and Engineering Programme (ISEP), National University of Singapore, 119077, Singapore

A R T I C L E I N F O Keywords:

Hybridized triboelectric-electromagnetic nano­ generator

Magnetic attraction Internet of things Human-machine interface

A B S T R A C T A magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator (MAHN) is designed and optimized with two magnets in attraction, and a silicone-based cushion with microstructures. The whole device including a fixed part and a moving part has a small size of 35 mm × 35 mm × 8 mm. After various energy harvesting and sensing characterizations, the MAHN has been verified as an efficient energy harvester, as well as a self-powered sensor for advanced human-computer interface (HMI). For energy harvesting, the triboelectric nanogenerator (TENG) delivers a peak power of 6.89 mW at the load of 1.1 MΩ, and the electromagnetic generator (EMG) delivers a peak power of 2.7 mW at the load of 1.1 kΩ. With the hybridized buck circuit, outputs from both the TENG and the EMG can be rectified and integrated efficiently to charge capacitors. A capacitor of

100 μF is charged to 2.03 V within 30 s, which is better than the performances of individual generators and their parallel-connection with only rectifiers. A thermometer and a Bluetooth module are powered with different amounts of the attracted magnet, implying the wide adaptability of the MAHN. Advanced applications of HMI have also been developed with cross-divided electrodes in the MAHN, such as orientation control in a game of

Snake, real-time operation of a PowerPoint presentation, and recognition of simple air gestures in contactless control. Moreover, a 3 × 3 array is prepared and signal channels are simplified by traversal method without loss of functions to achieve higher-level control. A virtual football game is demonstrated by a shoe moving on the array, and different kinds of shootings have been mapped to the player in the game.

1. Introduction With the rapid development of the Internet of Things (IoT), vastly distributed devices are playing important roles in the areas of environ­ mental monitoring, healthcare, and smart home, etc [1,2]. To addresses the power challenges of numerous devices with abundant functional sensors nowadays, various energy harvesting technologies including triboelectric nanogenerator (TENG) [3–7], electromagnetic generator (EMG) [8], and piezoelectric nanogenerator (PENG) [9], are introduced to convert irregular mechanical energy into electric power. Thereinto, the TENG has been extensively studied with different mechanical sources like rotation [10–12], vibration [13–15], human motion [16–18], and water wave [19–21], and proved to be effective power supplies with advantages of high output, easy fabrication and low cost.

However, hybridized mechanism of combining TENG with other energy harvesting technologies like piezoelectric [22–24], electromagnetic [25–37], and solar cells [38–41], is still required in the prototype to provide sufficient power and adapt various kinds of operations.

The hybridized triboelectric-electromagnetic nanogenerator is one of the largest branches among hybridized devices [42–44]. The structure design of hybridized triboelectric-electromagnetic nanogenerator can be roughly divided into two types: contact-separation type based on contact-separation mode TENG [25–28], and sliding/rotating type based on the sliding-mode/freestanding-mode TENG [34–37]. Conven­ tionally, a single unit of the EMG possesses one magnet. Recently, multiple magnets have been applied to achieve dynamic equilibrium of

* Corresponding author at: Department of Electrical and Computer Engineering, National University of Singapore, 4 Engineering Drive 3, Singapore 117576,

Singapore.

E-mail address: elelc@nus.edu.sg (C. Lee).

Contents lists available at ScienceDirect Nano Energy journal homepage: www.elsevier.com/locate/nanoen https://doi.org/10.1016/j.nanoen.2021.106154

Received 20 February 2021; Received in revised form 13 April 2021; Accepted 7 May 2021

Nano Energy 86 (2021) 106154 2 the movable magnet in its balance state in harvesting low-frequency mechanical energy. For example, Seol et al. proposed an oscillating magnet suspended in the tube by dual-directional magnetic repulsive forces, which enabled versatile energy harvesting capability at fre­ quencies in the sub-10 Hz range [45]. Additionally, Shao et al. intro­ duced magnet pairs that produce attractions to achieve contact-separation mode TENGs and rotary freestanding-mode EMGs in harvesting blue energy [46]. Although with multiple magnets struc­ ture, mechanical equilibrium of the movable magnet can be achieved, yet this inevitably increases the overall system complexity. Theoreti­ cally, multiple magnets can produce a higher electromagnetic field which has a great potential to enhance the output performance of the

EMG, but the influence of multiple magnets on the output performance is rarely studied in the literature.

Although the hybridized triboelectric-electromagnetic nano­ generators have been reported a lot, most works are focusing on energy harvesting rather than applying in advanced human-machine interfaces (HMIs) [47]. As for the HMIs, three mainstream applications have been developed by a large amount of wearable/portable devices, such as controlling the robotic arm with wearable patches [48,49], driving the drone with tactile devices [50], and playing VR/AR game with gloves [51]. Reported hybridized triboelectric-electromagnetic nanogenerators as shown in Table S1 have been demonstrated for various sensing ap­ plications, show potential for realizing the above applications about

HMIs. Chen et al. proposed a hybridized triboelectric-electromagnetic nanogenerator based on freestanding magnet moving in-plane arbi­ trarily [52]. However, only triboelectric signals are utilized when the prepared device is applied as a vibration sensor in playing the

“HitHamster” game. Furthermore, Wan et al. developed an array of flexible hybridized electromagnetic-triboelectric nanogenerators for 3D trajectory sensing by a magnetic part attached on a finger crossing over the copper FPCB coil array [53], while only electromagnetic signals obtained from the integrated coil array are utilized for sensing. Most recently, Bhatta et al. prepared a magnetic repulsion-assisted hybrid nanogenerator [54]. A central magnet can drive the side magnet aligned in the same magnetization direction moving, resulting in only electro­ magnetic signals are applied for detecting the motion parameters along with in-plane arbitrary directions. The synergistically integrated hy­ bridized mechanism can provide more flexibility in HMI design, more diverse controllable signals, and higher degrees of freedom in opera­ tions, but few works illustrate using both triboelectric signals and electromagnetic signals in HMI applications.

Herein, a magnetic-interaction assisted hybridized triboelectric- electromagnetic nanogenerator (MAHN) is designed with one moving part and one fixed part, where a pair of attracted magnets (Diameter 30 mm, Thickness 2 mm) are embedded in each part and the whole device is with a small size of 35 mm × 35 mm × 8 mm. So that, the TENG and the

EMG can simultaneously convert the mechanical energy of impacts into electric power as two parts attracting to each other during contact- separation processes. Outputs of the TENG and the EMG have been enhanced by the powerful attraction, and the scaling-down effect of output performance by lower operating frequency has been weakened with magnetic attraction induced by the two-magnet design. The MAHN has been proved as an efficient energy harvester working with hybrid­ ized buck convert circuit, which is better than the performances of in­ dividual and parallel-connection. Attributed to the adaptivity of magnets` amount, the MAHN is applied as an adjustable energy harvester in powering different electronics. By utilizing both the tribo­ electric signals and electromagnetic signals, the MAHN can function as an advanced HMI in entertainment, office, and gaming control, providing higher controlling capability and flexibility in various sce­ narios. Moreover, non-contact operations have been developed by the user waves the vertical moving part above the fixed part, resulting in recognizing simple air gestures for contactless controlling. Finally, the 3

× 3 array of the MAHN is prepared and channels are simplified by traversal method without loss of functions. And a virtual football game is demonstrated with the user moving the shoe on the array, assisting the player in the game in locating at different zones and shooting in different directions. To sum up, the MAHN can be applied as an efficient energy harvester, as well as an advanced human-machine interface in various practical scenarios.

2. Results and discussion 2.1. Design and optimization of the MAHN as a hybridized energy harvester

The MAHN is schematically presented in Fig. 1, in which the moving part and the fixed part are marked with two braces. As shown in Fig. 1a, a pair of magnets are placed inside two parts respectively. The magne­ tization direction is along the thickness direction, and magnet poles are opposite in confronting faces of two magnets. Due to magnetic attrac­ tion, two parts of the MAHN are attached at the rest state, as shown in

Fig. 1b. The device in attraction presents a dimension of 35 mm × 35 mm × 8 mm, which is easily separated by manual operations, and two parts in the discrete state are shown in Fig. 1c. The fixed part has a thickness of 4.5 mm, and multi wires are led out from this part for ap­ plications in energy harvesting and self-powered sensing. The moving part has a thickness of 3.5 mm and is applied as an independent part without leading wires. A promising application scenario is also shown in

Fig. 1a, where the moving part of the MAHN is operated with the user`s hand, and the fixed part is attached on the podium for connecting external load. After operations of contact mode and no-contact mode, the user can put the moving part back and locate it on the fixed part.

Through a COMSOL simulation shown in Fig. 1d, magnetic induction lines are gathered in between two magnets in attraction. Based on this gathering trend, the output of EMG can be enhanced due to more intensive magnetic flux through the coil when compared with the simulation in Fig. S1. To verify the simulation, comparison tests between two magnets in attraction and a single magnet are carried with the force gauge (Mecmesin, MultiTest 2.5-i) to remove most of the uncertainties from manual operation, where the instrument is moving at the largest speed of 900 mm/min. The results shown in Fig. S2 have proved that the output voltage of the coil in fixed part increases by introduced another magnet under the coil, which is in accord with the simulation.

Moreover, the effect of the cushion is also explored with the force gauge, and results are shown in Fig. S3. The cushion in the MAHN is designed by curing Eco-flex rubber on a conductive textile. The braided structure is applied as a mold for fabricating the microstructure on the cushion`s bottom surface. After demolding from the conductive textile, obvious microstructures have been obtained and shown in Fig. S3c, which is similar to the braided structure shown in Fig. S3b. The microstructure helps the cushion recover from the bottom contact sur­ face after impact, rather than stick on the bottom contact surface. As shown in Fig. S3a, increasing output voltage of different TENGs defined as D1 (Single magnet & Without cushion), D2 (Two magnets in attrac­ tion & Without cushion), D3 (Two magnets in attraction & Cushion without microstructure), D4 (Two magnets in attraction & Cushion with microstructure), have proved the positive effect of magnetic attraction for close contact and designed cushion for increasing contact surface area. Through the above discussion, introduced magnetic attraction has both improved performance of the TENG and the EMG in the MAHN.

In Fig. 2, the working principle and output performance have been further discussed. As a lateral view shown in Fig. 2a, The MAHN is based on contact-separation structured TENG and EMG. An intact square electrode is utilized in TENG to apply as a hybridized energy harvester in

Fig. 2a. At rest state, the moving part and the fixed part are attracted to each other, where charge balance is achieved among contact interface.

According to the triboelectric series, the FEP layer generates negative charges to maintain a neutralization with a positively charged Al layer in the moving part. Due to the cushion`s microstructure rubbing with the bottom interface, a small number of negative charges generated in the

L. Liu et al.

Nano Energy 86 (2021) 106154 3 cushion area also contributes to charge balance, which is marked in red and compatible with the comparison result in Fig. S2. On the other hand, there is no current flow in the coil of the EMG under rest state. Next, when the moving part is separating from the fixed part with external force overcoming magnetic attraction, the charge balance is broken, and the magnetic flux crossing the coil decreases. As a result, positive charges flow from the ground to TENG`s electrode, an induced current is generated in the coil. On the contrary, positive charges will flow from

TENG`s electrode to the ground when the moving part is approaching and attracted to the fixed part, as while an induced current in the opposite direction is generated in the coil due to increasing magnetic flux. It should be noted that approaching actions are converted into impacts eventually as the moving part in the magnetic attraction range.

From the above analysis, the TENG and EMG in the MAHN are cooperatively working along with the contact-separation processes. The outputs are measured with a multichannel oscilloscope, where leading wires from the fixed part are connected and tested simultaneously. The distance apart is controlled within 2 cm, and the frequency of the manual impact is changed according to a metronome. As shown in

Fig. 2b, there are positive peaks obtained during impacts and negative peaks during separations. Meanwhile, both voltages and currents of impacts are higher than those of separation, either for the TENG or the

EMG. Through tests under different frequencies, the TENG has presented an average output voltage of 280.7 V at 0.5 Hz, 296.8 V at 1 Hz, 309.4 V at 2 Hz, 339.0 V at 3 Hz, and 369.2 V at 4 Hz. In contrast, the EMG has delivered an average output voltage of 1.04 V at 0.5 Hz, 1.26 V at 1 Hz,

1.44 V at 2 Hz, 1.59 V at 3 Hz, 1.87 V at 4 Hz. As a result, even when the operating frequency decreases from 4 Hz to 0.5 Hz, the output voltage of the TENG can still maintain 76% while that of the EMG can maintain

56%. Output currents have also been tested in different frequencies, where average currents of the TENG reach 6.7 μA at 0.5 Hz, 7.4 μA at

1 Hz, 8.5 μA at 2 Hz, 8.71 μA at 3 Hz, and 9.9 μA at 4 Hz, average cur­ rents of the EMG reach 0.73 mA at 0.5 Hz, 0.79 mA at 1 Hz, 0.89 mA at

2 Hz, 0.81 mA at 3 Hz, 0.95 mA at 4 Hz. The outputs have presented the trend of increasing as the operating frequency is increased. When the operating frequency decreases from 4 Hz to 0.5 Hz, the output current of the TENG can maintain 68% while that of the EMG can maintain 77%. It is worth mentioning that the outputs of the TENG and the EMG are increased when the distance apart of manual impacts are not limited, as shown in Fig. S4.

The above analysis is based on the moving part is unfixed. In contrast, when the moving part is fixed on a linear motor to exclude influences of the magnetic attraction and changed operation distance, the results have demonstrated increased trends in Fig. S5. Both the

TENG and the EMG present increase trends as frequency increases, either voltages or currents. But all the output values are lower than those under manual operations with the unfixed moving part. After calcula­ tion, when the operating frequency decreases from 4 Hz to 0.5 Hz, the output current of the TENG can only maintain 16% while that of the

EMG can maintain 13%. And output current of the TENG can only maintain 13% from 4 Hz to 0.5 Hz, while the output current of the EMG has reduced by 14% from 4 Hz to 0.5 Hz. By comparison, the manual operations of the split design (one moving part and one fixed part) in the

MAHN have weakened the scaling-down effect of output performance by lower operating frequency, due to the magnetic attraction which can help improve output performances. In brief, results have demonstrated the adaptability of the MAHN as an efficient hybridized energy harvester for impact energy under different operating frequencies.

To evaluate the performance of the MAHN as a hybridized energy harvester, output power and capacitor charging have been tested in

Fig. 3. First, a series of resistors have been parallel-connected with the

TENG and the EMG, and corresponding voltages are tested for calcu­ lating output power. As shown in Fig. 3a, the output voltage of the TENG increases from 0.78 V to 179.6 V with the increasing value of the external loading resistance from 10 kΩ to 100 MΩ. Through calculation, maximum peak power has reached 6.89 mW at the load of 1.1 MΩ.

Similar to Fig. 3b, the output voltage of the EMG increases from 0.03 V to 3.25 V by increasing external loading resistance from 10 Ω to 100 kΩ.

And maximum peak power has reached 2.7 mW at the load of 1.1 kΩ.

Take into consideration that the MAHN is small-sized (35 mm × 35 mm

× 8 mm) and light-weight (33.1 g), the TENG has a maximum power density of 208.2 mW/Kg and the EMG has a maximum power density of

81.6 mW/Kg. Moreover, different capacitors are charged with individ­ ual energy harvesting parts and the hybridized device. As shown in

Fig. 3d, capacitors of 1 μF, 4.7 μF, 10 μF, and 47 μF have been respec­ tively charged to 3.15 V, 0.94 V, 0.56 V, and 0.11 V within 30 s with the rectified TENG under manual impacts of 2 Hz. On the other hand, ca­ pacitors of 1 μF, 4.7 μF, 10 μF, 47 μF, and 100 μF have been respectively charged to 3.49 V, 3.24 V, 2.85 V, 2.18 V, and 1.61 V within 30 s with the rectified EMG in the same situation, which is shown in Fig. 3e.

Moreover, the rectifier circuit shown in Fig. 3c is applied to integrate two energy harvesting units with the parallel connection. As result, ca­ pacitors of 1 μF, 4.7 μF, 10 μF, 47 μF, and 100 μF have been respectively charged to 5.48 V, 3.99 V, 3.74 V, 2.63 V, and 1.63 V within 30 s in the

Fig. 1. (a) Schematic diagram of the magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator (MAHN). (b) Photo of the prepared device in the rest state. (c) Photo of the prepared device in the discrete state. (d) Simulated distribution of the magnetic field of two magnets in attraction.

L. Liu et al.

Nano Energy 86 (2021) 106154 4 same situation, which is better than charged with the individual unit.

Although rectifiers in the parallel circuit have decreased the inter­ action effect of the TENG and the EMG, the hybridized buck circuit can further improve charging performance as shown in Fig. 4. Xi et.al, and

Liang et.al have proved that the buck circuit is a universal power management strategy for TENG, in which the inductor is applied to absorb and store a part of the energy as the magnetic field energy [55–57]. Based on enlarged output signals of the TENG and the EMG shown in Fig. S6, the impact process has contributed a large part in outputs. With the induction coil of the EMG applied in the hybridized buck circuit, outputs in impact processes are integrated with rectified

TENG signals, and a detailed design is shown in Fig. 4a. Results of charging performance are shown in Fig. 4b, capacitors of 1 μF, 4.7 μF,

10 μF, 47 μF, and 100 μF have been respectively charged to 7.54 V,

5.53 V, 4.25 V, 2.98 V and 2.03 V within 30 s with the hybridized buck circuit, which is better than performances of individual and parallel connection. With the charging capacitor of 100 μF, a thermometer is successfully powered in Fig. 4d and Movie V1. That capacitor is firstly charged to 1.62 V then connected with the thermometer and displayed temperature on-screen has proved the MAHN as an efficient hybridized energy harvester.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

Owing to the discrete design of the moving part, the MAHN also present adaptability of hybridized energy harvester by changing the amount of magnet. The 3D printed structure of the moving part is also adapted to different requirements by changing dimensions. As shown in

Figs. 4c, 3D printed structures have been adjusted to add more circular magnets, and the charging performances of two magnets and three magnets are tested with a capacitor of 100 μF. Results have shown that

Fig. 2. (a) Working principle of the MAHN. (b) The output voltage of the TENG under contact-separation operations of different frequencies. (c) The output current of the TENG under contact-separation operations of different frequencies. (d) The output voltage of the EMG under contact-separation operations of different fre­ quencies. (e) The output current of the EMG under contact-separation operations of different frequencies.

L. Liu et al.

Nano Energy 86 (2021) 106154 5 capacitor can be charged to 3.51 V by two magnets within 30 s with manual impacts of 2 Hz, and 4.89 V by three magnets in the same sit­ uation. Therefore, a Bluetooth module is successfully powered with an adapted moving part that contains three magnets. As shown in Fig. 4e and Movie V2, the capacitor of 100 μF has been charged to 3.34 V while directly connecting to the Bluetooth module. Thus, the module starts working and transferring signals of humidity and temperature to a ter­ minal of a mobile phone. Along with continuous manual impacts, the

Fig. 3. (a) Resistance dependence of the voltage and peak output power of the TENG. (b) Resistance dependence of the voltage and peak output power of the EMG. (c) Rectified circuit of the TENG and the EMG in parallel connection. (d) Charging different capacitors with the TENG with manual impacts of 2 Hz. (e) Charging different capacitors with the EMG with manual impacts of 2 Hz. (f) Charging different capacitors with the parallel circuit with manual impacts of 2 Hz.

Fig. 4. (a) Hybridized buck circuit designed for the MAHN. (b) Charging different capacitors with the hybridized buck circuit with manual impacts of 2 Hz. (c)

Charging a capacitor of 100 μF with different amounts of magnets in the moving part of the MAHN. (d) Powering a thermometer. (e) Powering a Bluetooth module.

L. Liu et al.

Nano Energy 86 (2021) 106154 6 Bluetooth module can work again as the capacitor is charged to 3.34 V.

Briefly, the MAHN can be an efficient energy harvester for impact en­ ergy and shows the promising potential of applying in scenarios where exists separated parts irregularly impacting each other, such as door`s opening and closing belong to the smart home, and buffer structures in bridges belong to the smart traffic.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

2.2. Design and analysis of the MAHN as a self-powered sensor for advanced human-machine interface

Due to the design of magnetic attraction, the MAHN at the rest state presents a similar feature in the joystick for homing after transfer commands of direction. The intact triboelectric electrode in the fixed part can be crossed cut into four electrodes as shown in Fig. 1. A lateral view is shown in Fig. 5 to illustrate the working principle of the MAHN applied as a self-powered sensor in transferring commands of direction.

At rest state, four electrodes are coved with the Al layer in the moving part, thus four single-electrode-based sliding TENGs have been formed.

In Fig. 5a, two TENGs are displayed in view to explain the working principle, in which the charge balance is achieved referred to Fig. 2.

When the moving part deviates from the rest position and slides to the left side, the negative charges in the FEP layer will lead to a change number of induced charges on two electrodes. As a result, positive charges flow from the left electrode to the ground and flow from the ground to the right electrode. When the moving part returns to the rest position, reverse signals are generated in two electrodes. Based on this working principle, four electrodes in the fixed part are designed and named as A1, A2, A3, and A4 in Fig. 5b. Designed operations are demonstrated in Movie V3. When the moving part of the MAHN moves along the directions of up, down, left, and right, four electrodes will generate signals according to the principle in Fig. 5a. As shown in

Fig. 5c, four electrodes are tested with a multichannel oscilloscope, simultaneously, and directions of up, down, left, and right can be easily determined by observing peaks among four electrodes. Obtained output voltages are within the range of −30–30 V, changes in values of four electrodes are due to external press and release from handlers. When the moving part slides up and away from A3 and A4, there will be detected positive peaks in A1 and A2, while negative peaks in A3 and A4. In contrast, there will be negative peaks in A1 and A2 while positive peaks in A3 and A4 when the moving part slides down. The motions of left and right can be done in the same manner by determining which pair of electrodes are detected positive and which pair of electrodes are detected negative.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

Primary data collected from the oscilloscope have proved to distin­ guish moving directions of the moving part. To demonstrate the MAHN as an advanced human-machine interface, another open-source elec­ tronics platform of Arduino 2650 and match circuit are applied to read voltage signals in four electrodes. Signals read by Arduino board are illustrated in Fig. S7, in which four kinds of operations are identified with different changes. Moreover, a classical game of “Snake” is devel­ oped with Python code in Fig. 5d and Movie V4, in which the user can control directions of the Snake to eat randomly distributed food dis­ played on a computer. In Fig. S8, the interactions of the MAHN applying in the HMI are clarified. The Arduino board can read the first peaks of electrodes, and find out which two electrodes present positive peaks.

Then the Arduino sends command codes of up, down, left, or right, into the Python code, and eventually changes the motions of the Snake. To sum up, the MAHN has been played as a self-powered “Joystick” in this application of entertainment, and the moving part will back to the rest position with help of magnetic attraction after transferring commands of directions.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

In Fig. 6, another application domain in the office is demonstrated by

Fig. 5. (a) Working principle of the MAHN as a human-machine interface (HMI). (b) Operations designed for the HMI. (c) Output voltages of four electrodes named

A1, A2, A3, and A4. (d) Demonstration of “Snake” Game.

L. Liu et al.

Nano Energy 86 (2021) 106154 7 applying the MAHN as an advanced human-machine interface. Unlike achieving direction determination with TENGs` signals, the EMG`s sig­ nals are also introduced to convey motion information. As shown in

Figs. 6a and b, basic operations are no-contact typed, which are along with the progressive trend of human-machine interface. Due to the un­ even distribution of the magnetic field in two orthogonal placed magnets shown in Fig. 6b, positive peaks and negative peaks will be detected when the vertical moving part drifts over the fixed part. The simulated distribution of the magnetic field is also clarified in Fig. S9, where the positions of vertical magnet influence magnetic induction lines, and correspond moving above the fixed magnet present electric signals` variations in the coil. In Fig. S10, the voltage signals have been decreased as increasing the distance between two parts. As shown in

Fig. 6c, the EMG unit in the MAHN has delivered output voltages of

3 mV as the moving part waving with the user’s hand above the fixed part about a distance of 10 cm. In the tests, positive peaks are generated when the user waves the vertical moving part from left to right, and negative peaks are generated when the user wave that from right to left.

Thus, the obvious difference can be applied to detect the user`s simple gesture of waving a hand. Moreover, four electrodes in the TENG part also help to detect when the moving part is off the fixed part. In Fig. S11, output voltages in four electrodes are tested in contact-separation pro­ cesses, where positive peaks are generated as contacting while negative peaks are generated as separating. The designed operations are sum­ marized in Movie V5, the moving part has been lifted, moved from left to right, moved from right to left, and put down to achieve no-contact moded interactions in HMI.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

In the demonstration shown in Fig. 6d and Movie V6, the MAHN has been applied in controlling a PowerPoint document with Arduino 2650, match circuit including operational amplifier circuit, and Python code.

Arduino read data is illustrated in Fig. S12, in which designed operations are identified with different signals from the TENG and the EMG. First, when the user lifts the moving part of the MAHN, the document starts playing with Fullscreen as the Arduino has read out four positive values in four electrodes of the TENGs. Next, operations of the next slide and previous slid of slides have been achieved with the EMG`s signal in no- contact mode. Finally, the user can end the presentation by setting the moving part down on the fixed part. The flow chart of interactions is illustrated in Fig. S13. The signals of TENGs are applied to offset the transient unstable EMG`s signals and play key roles in determining motions of contact and separation. Besides, the EMG`s signals are read through an operational amplifier circuit and applied to distinguish waving directions. Finally, the Arduino sends definite command codes into Python and changes states of presentations. Technically, the MAHN here is played as a commercial laser pointer and realized key functions based on a self-powered sensor. To sum up, the MAHN has shown po­ tential as an advanced human-machine interface both in entertainment and office.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

It is worth mentioning that two parts of the MAHN are both magnetic and easily attracted to iron-based furniture and household appliances. In

Fig. 6e and Movie V7, two parts of the MAHN can be anywhere on an iron cupboard, similar to refrigerator magnets. A virtual book model is created to demonstrate recognizing simple air gestures in the Smart home. The waving gestures are designed to accord with directions of page-turning. As shown in the demonstration, the user can grab down the moving part and wave it in front of the other part of the MAHN, resulting in page-turning as the user waving hand along the same di­ rection. Likewise, the user can put down the moving part anywhere on an iron cupboard after the above no-contact interactions. Briefly, simple air gestures like waving hands have been recognized by the MAHN.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

The array with multiple devices is further developed to illustrated another advanced application in HMI, which is discussed in Fig. 7. As shown in Fig. 7a, a single device of the MAHN needs four channels to be combined into 4 pairs to achieve four kinds of functions. Although four

Fig. 6. (a) Working principle of the MAHN as a non-contact HMI. (b) Simulated distribution of magnetic field with two orthogonal magnets. (c) Output voltages of

EMG unit as the moving part waving above the fixed part. (d) Demonstration of “PPT Controller”. (e) Demonstration of non-contact interactions in the smart home.

L. Liu et al.

Nano Energy 86 (2021) 106154 8 channels (A1, A2, A3, A4) can be combined into 6 pairs including (A1,

A2), (A1, A3), (A1, A4), (A2, A3), (A2, A4), and (A3, A4), these pairs result in only one kind of combination which satisfies four different pairs in one device. Moreover, in Fig. 7b, the traversal method is applied to find out how many channels are needed to build an array of 3 × 3.

Through calculations in Fig. S14, nine channels generate seven combi­ nations satisfying twenty-eight different pairs in seven devices, ten channels generate nine combinations satisfying thirty-six different pairs in nine devices, and eleven channels generate thirteen combinations satisfying fifty-two different pairs in thirteen devices. As shown in

Fig. 7b, ten channels have been labeled in each electrode, but these zones can rotate and switch places. So that, 10 channels are applied to build the array of 3 × 3, instead of 36 channels counted for thirty-six electrodes. To sum up, nine devices are distinct from each other, and every device presents four kinds of functions like the MAHN in the

“Snake” game.

With the channels` optimization, a virtual football game is demon­ strated in Fig. 7c and Movie V8. To meet a practical application sce­ nario, outsoles of shoes are attached with Nitrile films, and Nitrile films play as the Aluminum film in the above discussed moving part. Simi­ larly, outputs in operations of contact, separation, moving along with different directions are tested and shown in Fig. S15. The magnet in the above-discussed moving part also can be embedded into the insole, and the magnetic attraction between two parts is in favor of the user`s locating at specific points. As shown in Fig. 7d and Fig. S16, when the user puts the foot down on the fixed part of the MAHN, the Arduino board will recognize which zone is the user in. In the demonstration, nine zones of the array and nine positions are in the consistent one-to- one match. When the user steps on “zone 2′′, the player in the game locates at “position 2′′. Likewise, “position 6′′ and “position 7′′ are located when the player completes a successful shooting. The four electrodes in every zone generate different signals when the user does actions of kick forward, kick left, and kick right. At “zone 2′′, a successful shooting has been completed with a front kick. At “zone 6′′, the user loses the first try with a left kick, then achieves success with the front kick. Another two attempts at “zone 7′′ are operated with the right kick and front kick. To sum up, a 3 × 3 array of the MAHN is build and corresponding channels are simplified by traversal method without loss of functions.

Supplementary material related to this article can be found online at doi:10.1016/j.nanoen.2021.106154.

3. Conclusion In summary, a magnetic-interaction assisted hybridized triboelectric-electromagnetic nanogenerator is designed and optimized with two magnets in attraction, and a silicone-based cushion with mi­ crostructures. The MAHN is based on a contact-separation structured

TENG and EMG. And the whole device including a fixed part and a moving part is with a small size of 35 mm × 35 mm × 8 mm. At rest state, the moving part and the fixed part are attracted to each other.

Outputs of TENG and EMG have been enhanced by the powerful attraction, and the scaling-down effect of output performance by lower operating frequency has been weakened with magnetic attraction induced by the two-magnet design. The MAHN has demonstrated as an efficient hybridized energy harvester for impact energy, as well as a self- powered sensor for the advanced HMI. The TENG delivers the maximum peak power of 6.89 mW at the load of 1.1 MΩ, and the EMG delivers the maximum peak power of 2.7 mW at the load of 1.1 kΩ. Benefiting from the designed hybridized buck circuit, outputs of the TENG and the EMG are rectified and integrated to charge different capacitors. Thus, a capacitor of 100 μF is charged to 2.03 V within 30 s, which is better than the performances of individual parts and MAHN with a parallel- connected circuit. A thermometer and a Bluetooth module are pow­ ered with different amounts of attracted magnets, implying the high adaptability of the energy harvester. Attributed to the adaptivity of

Fig. 7. (a) Combination of four channels in the MAHN. (b) Optimization channels by traversal method. (c) Demonstration of a “ Football ” game. (d) Real-time game display.

L. Liu et al.

Nano Energy 86 (2021) 106154 9 magnets` amount, the MAHN is applied as an adjustable energy harvester in powering different electronics. Moreover, the intact tribo­ electric electrode has been cross-divided into four electrodes for appli­ cations of the human-machine interface. By utilizing both the triboelectric signals and electromagnetic signals, the MAHN can func­ tion as an advanced HMI in entertainment, office, and gaming control, providing higher controlling capability and flexibility in various sce­ narios. Along with the Arduino platform, the MAHN can transfer moving directions by determining positive peaks in four electrodes, and achieve control like a joystick in a classic game of Snake. Moreover, non-contact operations have been developed by the user waves the vertical moving part above the fixed part, resulting in recognizing simple air gestures for contactless controlling. Combined with contact-separation of the split design, the MAHN helps the user control a PowerPoint document including fullscreen, page down, page up, and end presentation, which realizes most functions of the commercial laser pointer but with a self- powered design. The MAHN also can be easily attached to iron-based furniture and household appliances, similarly to refrigerator magnets, for non-contact interactions in the smart home. Moreover, a 3 × 3 array of the MAHN is prepared and channels are simplified by traversal method without loss of functions to achieve more complex control. Thus, a virtual football game is demonstrated with the Nitrile film attached to the outsole. The user steps on different zones to assist the player in the game in locating at different positions and functions in different zones assist the player in shooting in different directions. Therefore, the hy­ bridized triboelectric-electromagnetic nanogenerator with magnetic interactions is a promising candidate in harvesting distributed me­ chanical energy and shows great innovation in the HMI.

4. Experimental section 4.1. Fabrication of the magnetic attraction-assisted hybridized triboelectric-electromagnetic nanogenerator (MAHN)

First, two rigid square structures with circular holes are prepared by the 3D printer (ANYCUBIC 4Max Pro) with polylactic acid (PLA). A pair of attracted magnets are respectively attached in the hole with double- sided tape so that two rigid surfaces can be attracted to each other closely. Then a Cu coil with 3800 turns, 24 mm diameter and 1 mm thickness, is fixed in a 3D printed rigid frame with Al tape and attached on the rigid surface of one square part. The other square part is covered with Al tapes.

Second, an Exo-flex 0030 based cushion is prepared by smearing and curing on a commercial conductive textile of 35 mm × 35 mm. Then the cushion is torn off and flatted on the former rigid frame, where the surface with microstructure is facing the rigid frame.

Third, a square electrode applied for energy harvester and crossed electrodes for the self-powered sensor is fabricated by attaching Al tape on a sheet of FEP film (50 µm thickness,10 cm width, DUPONT). Then the film is flatted on the cushion and fixed with extra parts attached to the rigid margin.

4.2. Characterization and electrical measurement The outputs data is acquired and saved by Model DSOX3034T

Multichannel Oscilloscope (Keysight), where 1000X probes (TT-HVP- 15HF) are applied to measure voltage signals of the TENGs, and 1X probe (GTL-101) is applied to measure voltage signals of the EMG. The current signals are measured with Electrometer (Keithley Model 6514) for TENG and low noise current preamplifier (SR570) for EMG. The charging performances are measured by Electrometer (Keithley Model

6514). Optical photos are photoed with OLYMPUS BX53M.

CRediT authorship contribution statement Long Liu: Conceptualization, Data curation, Formal analysis,

Investigation, Methodology, Resources, Software, Validation, Visuali­ zation, Writing - original draft, Writing - review & editing. Qiongfeng

Shi: Investigation, Validation, Formal analysis, Visualization, Writing - review & editing. Zhongda Sun: Investigation, Validation, Formal analysis, Visualization, Writing - review & editing. Chengkuo Lee:

Conceptualization, Methodology, Supervision, Project administration,

Funding acquisition, 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.

Acknowledgments This work was partly supported the National Key Research and

Development Program of China (Grant No. 2019YFB2004800, Project

No. R-2020-S-002) at NUSRI, Suzhou, China; and Singapore-Poland

Joint Grant (R-263-000-C91-305) “Chip-Scale MEMS Micro­

Spectrometer for Monitoring Harsh Industrial Gases” by Agency for

Science, Technology and Research (A*STAR), Singapore and NAWA

“Academic International Partnerships of Wroclaw University of Science and Technology” programmed by Polish National Agency for Academic

Exchange Programme.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.nanoen.2021.106154.

References [1] Q. Shi, B. Dong, T. He, Z. Sun, J. Zhu, Z. Zhang, C. Lee, Progress in wearable electronics/photonics—Moving toward the era of artificial intelligence and internet of things, InfoMat 2 (2020) 1131–1162, https://doi.org/10.1002/ inf2.12122. [2] F. Wen, T. He, H. Liu, H.-Y. Chen, T. Zhang, C. Lee, Advances in chemical sensing technology for enabling the next-generation self-sustainable integrated wearable system in the IoT era, Nano Energy 78 (2020), 105155, https://doi.org/10.1016/j. nanoen.2020.105155. [3] J. Zhu, M. Zhu, Q. Shi, F. Wen, L. Liu, B. Dong, A. Haroun, Y. Yang, P. Vachon,

X. Guo, T. He, C. Lee, Progress in TENG technology—A journey from energy harvesting to nanoenergy and nanosystem, EcoMat 2 (2020), https://doi.org/

10.1002/eom2.12058. [4] Q. Shi, T. He, C. Lee, More than energy harvesting – Combining triboelectric nanogenerator and flexible electronics technology for enabling novel micro-/nano- systems, Nano Energy 57 (2019) 851–871, https://doi.org/10.1016/j. nanoen.2019.01.002. [5] B. Chen, Y. Yang, Z.L. Wang, Scavenging wind energy by triboelectric nanogenerators, Adv. Energy Mater. 8 (2018), 1702649, https://doi.org/10.1002/ aenm.201702649. [6] Y. Zou, V. Raveendran, J. Chen, Wearable triboelectric nanogenerators for biomechanical energy harvesting, Nano Energy 77 (2020), 105303, https://doi. org/10.1016/j.nanoen.2020.105303. [7] C. Wu, A.C. Wang, W. Ding, H. Guo, Z.L. Wang, Triboelectric nanogenerator: a foundation of the energy for the new era, Adv. Energy Mater. 9 (2019), 1802906, https://doi.org/10.1002/aenm.201802906. [8] P. Carneiro, M.P. Soares dos Santos, A. Rodrigues, J.A.F. Ferreira, J.A.O. Sim˜oes, A.

T. Marques, A.L. Kholkin, Electromagnetic energy harvesting using magnetic levitation architectures: a review, Appl. Energy 260 (2020), 114191, https://doi. org/10.1016/j.apenergy.2019.114191. [9] H. Liu, J. Zhong, C. Lee, S.-W. Lee, L. Lin, A comprehensive review on piezoelectric energy harvesting technology: materials, mechanisms, and applications, Appl.

Phys. Rev. 5 (2018), 041306, https://doi.org/10.1063/1.5074184. [10] P. Bai, G. Zhu, Y. Liu, J. Chen, Q. Jing, W. Yang, J. Ma, G. Zhang, Z.L. Wang,

Cylindrical rotating triboelectric nanogenerator, ACS Nano 7 (2013) 6361–6366, https://doi.org/10.1021/nn402491y. [11] H. Zhang, Y. Yang, X. Zhong, Y. Su, Y. Zhou, C. Hu, Z.L. Wang, Single-electrode- based rotating triboelectric nanogenerator for harvesting energy from tires, ACS

Nano 8 (2014) 680–689, https://doi.org/10.1021/nn4053292. [12] L. Lin, S. Wang, Y. Xie, Q. Jing, S. Niu, Y. Hu, Z.L. Wang, Segmentally structured disk triboelectric nanogenerator for harvesting rotational mechanical energy, Nano

Lett. 13 (2013) 2916–2923, https://doi.org/10.1021/nl4013002.

L. Liu et al.

Nano Energy 86 (2021) 106154 10 [13] W. Yang, J. Chen, Q. Jing, J. Yang, X. Wen, Y. Su, G. Zhu, P. Bai, Z.L. Wang, 3D stack integrated triboelectric nanogenerator for harvesting vibration energy, Adv.

Funct. Mater. 24 (2014) 4090–4096, https://doi.org/10.1002/adfm.201304211. [14] C. Wu, R. Liu, J. Wang, Y. Zi, L. Lin, Z.L. Wang, A spring-based resonance coupling for hugely enhancing the performance of triboelectric nanogenerators for harvesting low-frequency vibration energy, Nano Energy 32 (2017) 287–293, https://doi.org/10.1016/j.nanoen.2016.12.061. [15] X. Xiao, X. Zhang, S. Wang, H. Ouyang, P. Chen, L. Song, H. Yuan, Y. Ji, P. Wang,

Z. Li, M. Xu, Z.L. Wang, Honeycomb structure inspired triboelectric nanogenerator for highly effective vibration energy harvesting and self-powered engine condition monitoring, Adv. Energy Mater. 9 (2019), 1902460, https://doi.org/10.1002/ aenm.201902460. [16] P. Bai, G. Zhu, Z.H. Lin, Q. Jing, J. Chen, G. Zhang, J. Ma, Z.L. Wang, Integrated multilayered triboelectric nanogenerator for harvesting biomechanical energy from human motions, ACS Nano 7 (2013) 3713–3719, https://doi.org/10.1021/ nn4007708. [17] Z. Zhao, C. Yan, Z. Liu, X. Fu, L.M. Peng, Y. Hu, Z. Zheng, Machine-washable textile triboelectric nanogenerators for effective human respiratory monitoring through loom weaving of metallic yarns, Adv. Mater. 28 (2016) 10267–10274, https://doi. org/10.1002/adma.201603679. [18] S. Wang, Y. Xie, S. Niu, L. Lin, Z.L. Wang, Freestanding triboelectric-layer-based nanogenerators for harvesting energy from a moving object or human motion in contact and non-contact modes, Adv. Mater. 26 (2014) 2818–2824, https://doi. org/10.1002/adma.201305303. [19] L. Xu, T. Jiang, P. Lin, J.J. Shao, C. He, W. Zhong, X.Y. Chen, Z.L. Wang, Coupled triboelectric nanogenerator networks for efficient water wave energy harvesting,

ACS Nano 12 (2018) 1849–1858, https://doi.org/10.1021/acsnano.7b08674. [20] T. Jiang, L.M. Zhang, X. Chen, C.B. Han, W. Tang, C. Zhang, L. Xu, Z.L. Wang,

Structural optimization of triboelectric nanogenerator for harvesting water wave energy, ACS Nano 9 (2015) 12562–12572, https://doi.org/10.1021/ acsnano.5b06372. [21] L. Liu, Q. Shi, J.S. Ho, C. Lee, Study of thin film blue energy harvester based on triboelectric nanogenerator and seashore IoT applications, Nano Energy 66 (2019),

104167, https://doi.org/10.1016/j.nanoen.2019.104167. [22] S. Wang, Z.L. Wang, Y. Yang, A one-structure-based hybridized nanogenerator for scavenging mechanical and thermal energies by triboelectric-piezoelectric- pyroelectric effects, Adv. Mater. 28 (2016) 2881–2887, https://doi.org/10.1002/ adma.201505684. [23] M. Zhu, Q. Shi, T. He, Z. Yi, Y. Ma, B. Yang, T. Chen, C. Lee, Self-powered and self- functional cotton sock using piezoelectric and triboelectric hybrid mechanism for healthcare and sports monitoring, ACS Nano 13 (2019) 1940–1952, https://doi. org/10.1021/acsnano.8b08329. [24] L. Wang, T. He, Z. Zhang, L. Zhao, C. Lee, G. Luo, Q. Mao, P. Yang, Q. Lin, X. Li,

R. Maeda, Z. Jiang, Self-sustained autonomous wireless sensing based on a hybridized TENG and PEG vibration mechanism, Nano Energy 80 (2021), 105555, https://doi.org/10.1016/j.nanoen.2020.105555. [25] K. Zhang, X. Wang, Y. Yang, Z.L. Wang, Hybridized electromagnetic-triboelectric nanogenerator for scavenging biomechanical energy for sustainably powering wearable electronics, ACS Nano 9 (2015) 3521–3529, https://doi.org/10.1021/ nn507455f. [26] L. Liu, W. Tang, C. Deng, B. Chen, K. Han, W. Zhong, Z.L. Wang, Self-powered versatile shoes based on hybrid nanogenerators, Nano Res. 11 (2018) 3972–3978, https://doi.org/10.1007/s12274-018-1978-z. [27] Z. Zhang, J. He, T. Wen, C. Zhai, J. Han, J. Mu, W. Jia, B. Zhang, W. Zhang,

X. Chou, C. Xue, Magnetically levitated-triboelectric nanogenerator as a self- powered vibration monitoring sensor, Nano Energy 33 (2017) 88–97, https://doi. org/10.1016/j.nanoen.2017.01.031. [28] P. Maharjan, T. Bhatta, H. Cho, X. Hui, C. Park, S. Yoon, M. Salauddin, M.

T. Rahman, S. Rana, J.Y. Park, A fully functional universal self-chargeable power module for portable/wearable electronics and self-powered IoT applications, Adv.

Energy Mater. 10 (2020), 2002782, https://doi.org/10.1002/aenm.202002782. [29] L. Liu, Q. Shi, C. Lee, A novel hybridized blue energy harvester aiming at all- weather IoT applications, Nano Energy 76 (2020), 105052, https://doi.org/

10.1016/j.nanoen.2020.105052. [30] Y. Wu, Q. Zeng, Q. Tang, W. Liu, G. Liu, Y. Zhang, J. Wu, C. Hu, X. Wang,

A teeterboard-like hybrid nanogenerator for efficient harvesting of low-frequency ocean wave energy, Nano Energy 67 (2020), 104205, https://doi.org/10.1016/j. nanoen.2019.104205. [31] H. Guo, Z. Wen, Y. Zi, M.-H. Yeh, J. Wang, L. Zhu, C. Hu, Z.L. Wang, A water-proof triboelectric-electromagnetic hybrid generator for energy harvesting in harsh environments, Adv. Energy Mater. 6 (2016), 1501593, https://doi.org/10.1002/ aenm.201501593. [32] Z. Wen, H. Guo, Y. Zi, M.H. Yeh, X. Wang, J. Deng, J. Wang, S. Li, C. Hu, L. Zhu, Z.

L. Wang, Harvesting broad frequency band blue energy by a triboelectric- electromagnetic hybrid nanogenerator, ACS Nano 10 (2016) 6526–6534, https:// doi.org/10.1021/acsnano.6b03293. [33] Y. Zhang, Q. Zeng, Y. Wu, J. Wu, S. Yuan, D. Tan, C. Hu, X. Wang, An ultra-durable windmill-like hybrid nanogenerator for steady and efficient harvesting of low- speed wind energy, Nano Micro Lett. 12 (2020) 175, https://doi.org/10.1007/ s40820-020-00513-2. [34] H. Yang, M. Wang, M. Deng, H. Guo, W. Zhang, H. Yang, Y. Xi, X. Li, C. Hu,

Z. Wang, A full-packaged rolling triboelectric-electromagnetic hybrid nanogenerator for energy harvesting and building up self-powered wireless systems, Nano Energy 56 (2019) 300–306, https://doi.org/10.1016/j. nanoen.2018.11.043. [35] X. Wang, Z. Wen, H. Guo, C. Wu, X. He, L. Lin, X. Cao, Z.L. Wang, Fully packaged blue energy harvester by hybridizing a rolling triboelectric nanogenerator and an electromagnetic generator, ACS Nano 10 (2016) 11369–11376, https://doi.org/

10.1021/acsnano.6b06622. [36] P. Hao, C. Guo, M. Wang, J. Shen, Y. Fu, Two-dimensional triboelectric- electromagnetic hybrid nanogenerator for wave energy harvesting, Nano Energy

58 (2019) 147–157, https://doi.org/10.1016/j.nanoen.2019.01.033. [37] L. Feng, G. Liu, H. Guo, Q. Tang, X. Pu, J. Chen, X. Wang, Y. Xi, C. Hu, Hybridized nanogenerator based on honeycomb-like three electrodes for efficient ocean wave energy harvesting, Nano Energy 47 (2018) 217–223, https://doi.org/10.1016/j. nanoen.2018.02.042. [38] X. Pu, W. Song, M. Liu, C. Sun, C. Du, C. Jiang, X. Huang, D. Zou, W. Hu, Z.L. Wang,

Wearable power-textiles by integrating fabric triboelectric nanogenerators and fiber-shaped dye-sensitized solar cells, Adv. Energy Mater. 6 (2016), 1601048, https://doi.org/10.1002/aenm.201601048. [39] L. Zhao, J. Duan, L. Liu, J. Wang, Y. Duan, L. Vaillant-Roca, X. Yang, Q. Tang,

Boosting power conversion efficiency by hybrid triboelectric nanogenerator/ silicon tandem solar cell toward rain energy harvesting, Nano Energy 82 (2021),

105773, https://doi.org/10.1016/j.nanoen.2021.105773. [40] H. Shao, Z. Wen, P. Cheng, N. Sun, Q. Shen, C. Zhou, M. Peng, Y. Yang, X. Xie,

X. Sun, Multifunctional power unit by hybridizing contact-separate triboelectric nanogenerator, electromagnetic generator and solar cell for harvesting blue energy, Nano Energy 39 (2017) 608–615, https://doi.org/10.1016/j. nanoen.2017.07.045. [41] X. Wei, Z. Wen, Y. Liu, N. Zhai, A. Wei, K. Feng, G. Yuan, J. Zhong, Y. Qiang,

X. Sun, Hybridized mechanical and solar energy-driven self-powered hydrogen production, Nano Micro Lett. 12 (2020) 88, https://doi.org/10.1007/s40820-020- 00422-4. [42] H. Liu, H. Fu, L. Sun, C. Lee, E.M. Yeatman, Hybrid energy harvesting technology: from materials, structural design, system integration to applications, Renew.

Sustain. Energy Rev. 137 (2021), 110473, https://doi.org/10.1016/j. rser.2020.110473. [43] T. Zhang, Z. Wen, Y. Liu, Z. Zhang, Y. Xie, X. Sun, Hybridized nanogenerators for multifunctional self-powered sensing: principles, prototypes, and perspectives, iScience 23 (2020), 101813, https://doi.org/10.1016/j.isci.2020.101813. [44] X. Chen, Z. Ren, M. Han, J. Wan, H. Zhang, Hybrid energy cells based on triboelectric nanogenerator: from principle to system, Nano Energy 75 (2020),

104980, https://doi.org/10.1016/j.nanoen.2020.104980. [45] M.-L. Seol, J.-W. Han, S.-J. Park, S.-B. Jeon, Y.-K. Choi, Hybrid energy harvester with simultaneous triboelectric and electromagnetic generation from an embedded floating oscillator in a single package, Nano Energy 23 (2016) 50–59, https://doi. org/10.1016/j.nanoen.2016.03.004. [46] H. Shao, P. Cheng, R. Chen, L. Xie, N. Sun, Q. Shen, X. Chen, Q. Zhu, Y. Zhang,

Y. Liu, Z. Wen, X. Sun, Triboelectric-electromagnetic hybrid generator for harvesting blue energy, Nanomicro Lett. 10 (2018) 54, https://doi.org/10.1007/ s40820-018-0207-3. [47] M. Zhu, T. He, C. Lee, Technologies toward next generation human machine interfaces: from machine learning enhanced tactile sensing to neuromorphic sensory systems, Appl. Phys. Rev. 7 (2020), 031305, https://doi.org/10.1063/

5.0016485. [48] T. Chen, Q. Shi, M. Zhu, T. He, L. Sun, L. Yang, C. Lee, Triboelectric self-powered wearable flexible patch as 3D motion control interface for robotic manipulator,

ACS Nano 12 (2018) 11561–11571, https://doi.org/10.1021/acsnano.8b06747. [49] X. Pu, H. Guo, Q. Tang, J. Chen, L. Feng, G. Liu, X. Wang, Y. Xi, C. Hu, Z.L. Wang,

Rotation sensing and gesture control of a robot joint via triboelectric quantization sensor, Nano Energy 54 (2018) 453–460, https://doi.org/10.1016/j. nanoen.2018.10.044. [50] W. Choi, I. Yun, J. Jeung, Y.S. Park, S. Cho, D.W. Kim, I.S. Kang, Y. Chung,

U. Jeong, Stretchable triboelectric multimodal tactile interface simultaneously recognizing various dynamic body motions, Nano Energy 56 (2019) 347–356, https://doi.org/10.1016/j.nanoen.2018.11.066. [51] F. Wen, Z. Sun, T. He, Q. Shi, M. Zhu, Z. Zhang, L. Li, T. Zhang, C. Lee, Machine learning glove using self-powered conductive superhydrophobic triboelectric textile for gesture recognition in VR/AR applications, Adv. Sci. (Weinh.) 7 (2020),

2000261, https://doi.org/10.1002/advs.202000261. [52] X. Chen, H. Guo, H. Wu, H. Chen, Y. Song, Z. Su, H. Zhang, Hybrid generator based on freestanding magnet as all-direction in-plane energy harvester and vibration sensor, Nano Energy 49 (2018) 51–58, https://doi.org/10.1016/j. nanoen.2018.04.024. [53] J. Wan, H. Wang, L. Miao, X. Chen, Y. Song, H. Guo, C. Xu, Z. Ren, H. Zhang,

A flexible hybridized electromagnetic-triboelectric nanogenerator and its application for 3D trajectory sensing, Nano Energy 74 (2020), 104878, https://doi. org/10.1016/j.nanoen.2020.104878. [54] T. Bhatta, P. Maharjan, M. Salauddin, M.T. Rahman, S.M.S. Rana, J.Y. Park,

A battery-less arbitrary motion sensing system using magnetic repulsion-based self- powered motion sensors and hybrid nanogenerator, Adv. Funct. Mater. 30 (2020),

2003276, https://doi.org/10.1002/adfm.202003276.

L. Liu et al.

Nano Energy 86 (2021) 106154 11 [55] F. Xi, Y. Pang, W. Li, T. Jiang, L. Zhang, T. Guo, G. Liu, C. Zhang, Z.L. Wang,

Universal power management strategy for triboelectric nanogenerator, Nano

Energy 37 (2017) 168–176, https://doi.org/10.1016/j.nanoen.2017.05.027. [56] X. Liang, T. Jiang, G. Liu, T. Xiao, L. Xu, W. Li, F. Xi, C. Zhang, Z.L. Wang,

Triboelectric nanogenerator networks integrated with power management module for water wave energy harvesting, Adv. Funct. Mater. 29 (2019), 1807241, https:// doi.org/10.1002/adfm.201807241. [57] C. Fang, T. Tong, T. Bu, Y. Cao, S. Xu, Y. Qi, C. Zhang, Overview of power management for triboelectric nanogenerators, Adv. Intell. Syst. 2 (2020), 1900129, https://doi.org/10.1002/aisy.201900129.

Long Liu received his Ph.D. degree from Beijing Institute of

Nano energy and Nanosystems, Chinese Academy of Sciences in 2018. He is currently a Research Fellow in the Department of

Electrical and Computer Engineering, National University of

Singapore. His research interests are focused on energy har­ vesters and self-powered sensors, wearable electronics, IoT applications in the 5G era.

Qiongfeng Shi received his B.Eng. degree from the Depart­ ment of Electronic Engineering and Information Science, Uni­ versity of Science and Technology of China (USTC) in 2012, and received his Ph.D. degree from the Department of Elec­ trical and Computer Engineering, National University of

Singapore (NUS) in 2018. He is currently a Research Fellow in the Department of Electrical and Computer Engineering, Na­ tional University of Singapore. His research interests include energy harvesters, triboelectric nanogenerators, self-powered sensors, and wearable/implantable electronics.

Zhongda Sun Received his B.Eng. degree from the School of

Electronic and Information at Soochow University, Suzhou,

China, in 2018. After that he received his Master degree from the Department of Electrical and Computer Engineering, NUS, in 2019. He is now a PhD student at the Department of Elec­ trical and Computer Engineering, NUS. His research interests include self-powered wearable sensors and triboelectric nanogenerator.

Chengkuo Lee received his Ph.D. degree in Precision Engi­ neering from The University of Tokyo in 1996. Currently, he is the director of Center for Intelligent Sensors and MEMS at

National University of Singapore, Singapore. In 2001, he cofounded Asia Pacific Microsystems, Inc., where he was the

Vice President. From 2006–2009, he was a Senior Member of the Technical Staff at the Institute of Microelectronics, A-STAR,

Singapore. He has contributed to more than 380 peer-reviewed international journal articles. His ORCID is 0000-0002-8886- 3649.

L. Liu et al.

📖 中文全文 Chinese Full Text

中文

Nano Energy 86 (2021) 106154 2021年5月13日在线发表 2211-2855/© 2021 Elsevier Ltd. 保留所有权利。

全文 磁相互作用辅助的混合式摩擦电-电磁纳米发电机及其在人机接口中的先进应用

刘龙 a,b,c,d, 施琼峰 a,b,c,d, 孙忠达 a,b,c,d, 李焯耀 a,b,c,d,e,* a 新加坡国立大学电气与计算机工程系,工程驱动3号4号,117576,新加坡 b 新加坡国立大学智能传感器与MEMS中心,E6栋#05-11,工程驱动1号5号,117608,新加坡 c 新加坡国立大学混合集成柔性(可拉伸)电子系统项目(HIFES),E6栋#05-3,工程驱动1号5号,117608,新加坡 d 新加坡国立大学苏州研究院(NUSRI),苏州工业园区,苏州215123,中华人民共和国 e 新加坡国立大学研究生院—综合科学与工程项目(ISEP),119077,新加坡

**A R T I C L E I N F O** 关键词: 混合式摩擦电-电磁纳米发电机 磁吸引力 物联网 人机接口

**A B S T R A C T** 本文设计并优化了一种磁相互作用辅助的混合式摩擦电-电磁纳米发电机(MAHN),其采用一对相互吸引的磁铁以及带有微结构的硅胶基缓冲垫。整个装置包括固定部分和活动部分,尺寸仅为35 mm × 35 mm × 8 mm。经过多种能量收集与传感特性测试,MAHN被验证为一种高效的能量收集器,同时也是用于先进人机接口(HMI)的自供电传感器。在能量收集方面,摩擦电纳米发电机(TENG)在1.1 MΩ负载下输出功率峰值为6.89 mW,电磁发电机(EMG)在1.1 kΩ负载下输出功率峰值为2.7 mW。通过混合式降压电路,TENG和EMG的输出可被高效整流与集成,用于对电容器充电。在30秒内,一个100 μF的电容器可充电至2.03 V,性能优于单个发电机及仅使用整流器并联连接的情况。通过调节磁铁数量,MAHN可驱动温度计和蓝牙模块等不同电子器件,展现出广泛的适应性。此外,利用MAHN中交叉分割电极,开发了多种先进HMI应用,如贪吃蛇游戏中的方向控制、PowerPoint演示文稿的实时操作,以及非接触控制中简单空中手势的识别。同时,制备了3×3阵列,并通过遍历法简化信号通道而不损失功能,实现了更高级别的控制。通过鞋在阵列上的移动演示了虚拟足球游戏,并将不同类型的射门动作映射到游戏中的球员。

**1. 引言** 随着物联网(IoT)的快速发展,大量分布式设备在环境监测、医疗健康、智能家居等领域发挥着重要作用[1,2]。为应对当前众多功能传感器设备的供电挑战,多种能量收集技术被引入,包括摩擦电纳米发电机(TENG)[3–7]、电磁发电机(EMG)[8]和压电纳米发电机(PENG)[9],用于将不规则机械能转化为电能。其中,TENG已被广泛研究,其机械源涵盖旋转[10–12]、振动[13–15]、人体运动[16–18]和水波[19–21],并因其高输出、易制备和低成本等优势被证明为有效的电源。然而,仍需将TENG与其他能量收集技术(如压电[22–24]、电磁[25–37]和太阳能电池[38–41])结合,以提供充足电力并适应多种操作需求。

混合式摩擦电-电磁纳米发电机是混合式器件中最大的分支之一[42–44]。其结构设计大致分为两类:基于接触分离模式TENG的接触分离型[25–28],以及基于滑动模式/独立模式TENG的滑动/旋转型[34–37]。传统上,单个EMG单元仅含一块磁铁。近年来,多磁铁结构被用于实现可动磁铁在低频机械能收集中的动态平衡。例如,Seol等人提出了一种由双向磁斥力悬浮在管中的振荡磁铁,可在低于10 Hz频率范围内实现多模式能量收集[45]。此外,Shao等人引入磁对吸引力,实现了接触分离模式TENG和旋转独立模式EMG在蓝色能量收集中的应用[46]。尽管多磁铁结构可实现机械平衡,但不可避免地增加了系统复杂性。理论上,多磁铁可产生更强的磁场,有望提升EMG输出性能,但相关研究在文献中仍较少涉及。

尽管混合式摩擦电-电磁纳米发电机已有大量报道,但多数研究侧重于能量收集,而非应用于先进人机接口(HMI)[47]。在HMI领域,可穿戴/便携设备已开发出三大主流应用,如通过可穿戴贴片控制机械臂[48,49]、触觉设备驱动无人机[50],以及手套操控VR/AR游戏[51]。如表S1所示,已有混合式摩擦电-电磁纳米发电机被用于多种传感应用,显示出实现上述HMI应用的潜力。Chen等人提出了一种基于磁铁在平面内任意移动的混合式摩擦电-电磁纳米发电机[52],但在将其作为振动传感器用于“打地鼠”游戏时,仅利用了摩擦电信号。此外,Wan等人开发了一种柔性混合电磁-摩擦电纳米发电机阵列,用于三维轨迹传感,通过手指上的磁铁部分跨越铜FPCB线圈阵列[53],但仅利用集成线圈阵列产生的电磁信号进行传感。最近,Bhatta等人制备了一种磁斥力辅助的混合纳米发电机[54],其中中心磁铁驱动同向磁化的侧磁铁移动,仅利用电磁信号检测平面内任意方向的运动参数。协同集成的混合机制可为HMI设计提供更高灵活性、更多样化的可控信号和更高的操作自由度,但鲜有研究同时利用摩擦电信号和电磁信号实现HMI应用。

本文设计了一种磁相互作用辅助的混合式摩擦电-电磁纳米发电机(MAHN),包含一个活动部分和一个固定部分,每部分嵌入一对相互吸引的磁铁(直径30 mm,厚度2 mm),整体尺寸仅为35 mm × 35 mm × 8 mm。在接触分离过程中,TENG和EMG可将撞击机械能同时转化为电能。TENG和EMG的输出因强磁吸引力而增强,且双磁铁设计产生的磁吸引力减弱了低频运行下输出性能的下降效应。MAHN被证明为一种高效的混合式能量收集器,其性能优于单个发电机及并联连接。得益于磁铁数量的可调性,MAHN可作为可调能量收集器为不同电子器件供电。通过同时利用摩擦电信号和电磁信号,MAHN可在娱乐、办公和游戏控制中作为先进HMI,提供更高的控制能力和场景适应性。此外,通过用户手持活动部分在固定部分上方挥动,实现了非接触操作,可识别简单空中手势用于非接触控制。最后,制备了MAHN的3×3阵列,并通过遍历法简化通道而不损失功能。通过鞋在阵列上的移动演示了虚拟足球游戏,使球员能在不同区域定位并向不同方向射门。综上,MAHN可作为高效能量收集器和先进人机接口,广泛应用于多种实际场景。

**2. 结果与讨论** **2.1. MAHN作为混合式能量收集器的设计与优化** MAHN的示意图如图1所示,其中活动部分和固定部分分别用括号标出。如图1a所示,一对磁铁分别置于两部分中,磁化方向沿厚度方向,且两磁铁相对面磁极为异性。由于磁吸引力,MAHN的两部分在静止状态下相互吸附(图1b)。吸附状态下的装置尺寸为35 mm × 35 mm × 8 mm,易于手动分离,分离状态如图1c所示。固定部分厚度为4.5 mm,引出多根导线用于能量收集和自供电传感应用。活动部分厚度为3.5 mm,作为独立部分无导线引出。图1a还展示了一种有前景的应用场景:用户手持MAHN的活动部分,固定部分置于讲台以连接外部负载。在接触模式和非接触模式操作后,用户可将活动部分放回并定位于固定部分。通过COMSOL仿真(图1d),磁力线在两磁铁间聚集。基于此聚集趋势,EMG的输出可因穿过线圈的磁通量更密集而增强(与图S1仿真对比)。为验证仿真,使用测力计(Mecmesin, MultiTest 2.5-i)对双磁铁吸引与单磁铁进行对比测试,以消除手动操作的不确定性,仪器以最大速度900 mm/min移动。图S2结果表明,引入另一磁铁后,固定部分线圈的输出电压提高,与仿真一致。

此外,还通过测力计探究了缓冲垫的影响,结果如图S3所示。MAHN中的缓冲垫通过在导电织物上固化Eco-flex橡胶制成。编织结构被用作模具,在缓冲垫底面制备微结构。从导电织物脱模后,获得明显微结构(图S3c),与图S3b所示编织结构相似。微结构有助于缓冲垫在撞击后从底面恢复,而非粘附于底面。如图S3a所示,不同TENG(定义为D1:单磁铁&无缓冲垫;D2:双磁铁吸引&无缓冲垫;D3:双磁铁吸引&无微结构缓冲垫;D4:双磁铁吸引&有微结构缓冲垫)的输出电压增加,证明了磁吸引力对紧密接触的积极作用,以及所设计缓冲垫对增加接触表面积的有效性。综上,引入的磁吸引力同时提升了MAHN中TENG和EMG的性能。

图2进一步讨论了其工作原理与输出性能。如图2a侧视图所示,MAHN基于接触分离结构的TENG和EMG。在图2a中,TENG使用完整方形电极作为混合式能量收集器。静止状态下,活动部分与固定部分相互吸引,接触界面达到电荷平衡。根据摩擦电序,FEP层产生负电荷,以与活动部分中带正电的Al层中和。由于缓冲垫微结构与底面摩擦,缓冲垫区域产生的少量负电荷(红色标记)也有助于电荷平衡,与图S2对比结果一致。另一方面,EMG线圈在静止状态下无电流。当活动部分在外力克服磁吸引力作用下与固定部分分离时,电荷平衡被破坏,穿过线圈的磁通量减少。结果,正电荷从地流向TENG电极,线圈中产生感应电流。相反,当活动部分接近并被吸引至固定部分时,正电荷从TENG电极流向地,同时由于磁通量增加,线圈中产生反向感应电流。需注意的是,接近动作最终会转化为撞击,因为活动部分处于磁吸引力范围内。

综上,MAHN中的TENG与EMG在接触分离过程中协同工作。输出通过多通道示波器测量,固定部分引出的导线同时连接测试。分离距离控制在2 cm内,手动撞击频率根据节拍器调节。如图2b所示,撞击时产生正峰值,分离时产生负峰值。同时,撞击时的电压和电流均高于分离时,无论是TENG还是EMG。在不同频率测试下,TENG在0.5 Hz、1 Hz、2 Hz、3 Hz和4 Hz下的平均输出电压分别为280.7 V、296.8 V、309.4 V、339.0 V和369.2 V。而EMG在相同频率下的平均输出电压分别为1.04 V、1.26 V、1.44 V、1.59 V和1.87 V。因此,当工作频率从4 Hz降至0.5 Hz时,TENG输出电压仍可维持76%,EMG可维持56%。输出电流测试显示,TENG在0.5 Hz至4 Hz下的平均电流分别为6.7 μA、7.4 μA、8.5 μA、8.71 μA和9.9 μA;EMG分别为0.73 mA、0.79 mA、0.89 mA、0.81 mA和0.95 mA。输出随工作频率增加而增加。当频率从4 Hz降至0.5 Hz时,TENG输出电流可维持68%,EMG可维持77%。值得注意的是,当手动撞击距离不受限时,TENG和EMG的输出均提高(图S4)。

上述分析基于活动部分未固定。相反,当活动部分固定于线性电机以排除磁吸引力和操作距离变化的影响时,结果(图S5)显示TENG和EMG的输出均随频率增加而增加,但所有输出值均低于手动操作未固定活动部分的情况。计算表明,当工作频率从4 Hz降至0.5 Hz时,TENG输出电流仅能维持16%,EMG仅能维持13%。而TENG输出电流从4 Hz到0.5 Hz仅维持13%,EMG输出电流下降14%。相比之下,MAHN的分体式设计(一个活动部分和一个固定部分)的手动操作减弱了低频运行下输出性能的下降效应,这得益于磁吸引力对输出性能的提升。简言之,结果证明了MAHN作为高效混合式能量收集器在不同工作频率下对撞击能量的适应性。

为评估MAHN作为混合式能量收集器的性能,测试了输出功率和电容器充电(图3)。首先,将一系列电阻与TENG和EMG并联,测量相应电压以计算输出功率。如图3a所示,当外部负载电阻从10 kΩ增至100 MΩ时,TENG输出电压从0.78 V增至179.6 V。计算得最大峰值功率为6.89 mW(负载1.1 MΩ)。类似地,图3b显示,当外部负载电阻从10 Ω增至100 kΩ时,EMG输出电压从0.03 V增至3.25 V,最大峰值功率为2.7 mW(负载1.1 kΩ)。考虑到MAHN尺寸小(35 mm × 35 mm × 8 mm)且重量轻(33.1 g),TENG的最大功率密度为208.2 mW/kg,EMG为81.6 mW/kg。此外,使用单个能量收集单元和混合式器件对不同的电容器充电。如图3d所示,在2 Hz手动撞击下,经整流的TENG在30秒内分别将1 μF、4.7 μF、10 μF和47 μF电容器充电至3.15 V、0.94 V、0.56 V和0.11 V。另一方面,图3e显示,相同条件下,经整流的EMG在30秒内分别将1 μF、4.7 μF、10 μF、47 μF和100 μF电容器充电至3.49 V、3.24 V、2.85 V、2.18 V和1.61 V。此外,图3c所示整流电路用于将两个能量收集单元并联集成。结果,在30秒内,1 μF、4.7 μF、10 μF、47 μF和100 μF电容器分别被充电至5.48 V、3.99 V、3.74 V、2.63 V和1.63 V,性能优于单个单元充电。尽管并联电路中的整流器减弱了TENG与EMG的交互作用,但混合式降压电路可进一步提升充电性能(图4)。Xi等人和Liang人已证明降压电路是TENG的通用电源管理策略,其中电感用于吸收并存储部分能量作为磁场能量[55–57]。基于图S6中TENG和EMG的放大输出信号,撞击过程对输出贡献显著。通过将EMG电感线圈应用于混合式降压电路,撞击过程中的输出与整流后的TENG信号集成,详细设计如图4a所示。充电性能结果如图4b所示,在30秒内,混合式降压电路分别将1 μF、4.7 μF、10 μF、47 μF和100 μF电容器充电至7.54 V、5.53 V、4.25 V、2.98 V和2.03 V,性能优于单个单元和并联连接。使用100 μF电容器充电后,成功驱动了温度计(图4d和电影V1)。该电容器先充电至1.62 V,再连接温度计,屏幕显示温度,证明了MAHN作为高效混合式能量收集器的能力。

由于活动部分的分体式设计,MAHN还通过改变磁铁数量展现出混合式能量收集器的适应性。活动部分的3D打印结构也可通过调整尺寸满足不同需求。如图4c所示,调整3D打印结构以添加更多圆形磁铁,并使用100 μF电容器测试双磁铁与三磁铁的充电性能。结果表明,在2 Hz手动撞击下,双磁铁可在30秒内将电容器充电至3.51 V,三磁铁可充电至4.89 V。因此,使用含三磁铁的改进活动部分成功驱动了蓝牙模块。如图4e和电影V2所示,100 μF电容器充电至3.34 V后直接连接蓝牙模块,模块开始工作并将湿度和温度信号传输至手机终端。随着持续手动撞击,电容器再次充电至3.34 V,蓝牙模块可重新工作。简言之,MAHN可作为撞击能量的高效能量收集器,在存在分离部件不规则撞击的场景中具有应用潜力,如智能家居中的门窗开关,以及智能交通中的桥梁缓冲结构。

**2.2. MAHN作为自供电传感器用于先进人机接口的设计与分析** 由于磁吸引力设计,MAHN在静止状态下呈现类似摇杆的复位特性。固定部分的完整摩擦电电极可被交叉分割为四个电极(图1)。图5展示了MAHN作为自供电传感器用于方向指令传输的工作原理侧视图。静止状态下,四个电极被活动部分的Al层覆盖,形成四个单电极滑动TENG。图5a展示了两个TENG以解释工作原理,其电荷平衡参考图2。当活动部分偏离静止位置并向左滑动时,FEP层中的负电荷会引起两电极上感应电荷数量变化。结果,正电荷从左电极流向地,并从地流向右电极。当活动部分返回静止位置时,两电极产生反向信号。基于此原理,固定部分设计了四个电极,命名为A1、A2、A3和A4(图5b)。设计操作在电影V3中展示。当MAHN活动部分沿上下左右方向移动时,四个电极将根据图5a原理产生信号。如图5c所示,使用多通道示波器同时测试四个电极,通过观察各电极峰值可轻松判断上下左右方向。输出电压在−30至30 V范围内,四电极值变化源于操作者的外部按压与释放。当活动部分向上滑动远离A3和A4时,A1和A2检测到正峰值,A3和A4为负峰值。相反,当活动部分向下滑动时,A1和A2为负峰值,A3和A4为正峰值。左右运动可通过判断哪对电极检测到正峰值、哪对检测到负峰值来识别。

从示波器收集的原始数据已证明可区分活动部分的运动方向。为展示MAHN作为先进人机接口,采用Arduino 2650开源电子平台和匹配电路读取四电极电压信号。Arduino读取的信号如图S7所示,四种操作通过不同变化识别。此外,使用Python代码开发了经典“贪吃蛇”游戏(图5d和电影V4),用户可控制蛇的方向以吃掉屏幕上随机出现的食物。图S8阐明了MAHN在HMI中的交互过程。Arduino读取电极的首个峰值,并识别哪两电极呈现正峰值。随后,Arduino向Python代码发送上下左右指令代码,最终改变蛇的运动。综上,MAHN在此娱乐应用中作为自供电“摇杆”,在传输方向指令后,活动部分在磁吸引力作用下返回静止位置。

图6展示了MAHN作为先进人机接口在办公领域的另一应用。与利用TENG信号判断方向不同,此处引入EMG信号以传输运动信息。如图6a和6b所示,基本操作属于非接触式,符合人机接口的发展趋势。由于两正交放置磁铁的磁场分布不均匀(图6b),当垂直活动部分在固定部分上方漂移时,会检测到正负峰值。磁场分布仿真如图S9所示,垂直磁铁位置影响磁力线,在固定磁铁上方移动时,线圈中电信号相应变化。图S10显示,随着两部分距离增加,电压信号减弱。如图6c所示,当用户手持活动部分在固定部分上方约10 cm处挥动时,MAHN中的EMG单元输出电压为3 mV。测试中,用户从左向右挥动时产生正峰值,从右向左挥动时产生负峰值。因此,该明显差异可用于检测用户简单挥手手势。此外,TENG部分的四个电极也有助于检测活动部分是否脱离固定部分。图S11显示,在接触分离过程中测试四电极输出电压,接触时产生正峰值,分离时产生负峰值。设计操作在电影V5中总结:活动部分被拿起、从左向右移动、从右向左移动并放下,实现HMI中的非接触交互。

在图6d和电影V6的演示中,MAHN被用于控制PowerPoint文档,使用Arduino 2650、包含运算放大器电路的匹配电路和Python代码。Arduino读取数据如图S12所示,设计操作通过TENG和EMG的不同信号识别。首先,当用户拿起MAHN活动部分时,文档开始全屏播放,因Arduino读取到TENG四电极的四个正值。接着,通过EMG信号在非接触模式下实现下一页和上一页操作。最后,用户可将活动部分放回固定部分以结束演示。交互流程图如图S13所示。TENG信号用于抵消EMG的瞬时不稳定信号,并在判断接触与分离动作中起关键作用。EMG信号通过运算放大器电路读取,用于区分挥动方向。最终,Arduino向Python发送明确指令代码,改变演示状态。技术上,MAHN在此作为商用激光笔,基于自供电传感器实现关键功能。综上,MAHN在娱乐和办公领域均展现出作为先进人机接口的潜力。

值得注意的是,MAHN的两部分均具有磁性,易吸附于铁质家具和家用电器。如图6e和电影V7所示,MAHN的两部分可置于铁质橱柜上,类似于冰箱贴。创建了虚拟书模型以展示智能家居中简单空中手势的识别。挥手手势设计为与翻页方向一致。演示中,用户可拿起活动部分并在MAHN另一部分前方挥动,实现沿挥手方向的翻页。同样,用户可在非接触交互后将活动部分放回铁质橱柜任意位置。简言之,MAHN可识别挥手等简单空中手势。

进一步开发了多器件阵列以展示HMI中的另一先进应用(图7)。如图7a所示,单个MAHN器件需四个通道组合成四对以实现四种功能。尽管四个通道(A1、A2、A3、A4)可组合成六对(A1,A2)、(A1,A3)、(A1,A4)、(A2,A3)、(A2,A4)和(A3,A4),但仅有一种组合满足单个器件中四对不同功能。此外,图7b采用遍历法计算构建3×3阵列所需通道数。图S14计算表明,九通道生成七种组合,满足七器件二十八对不同对;十通道生成九种组合,满足九器件三十六对不同对;十一通道生成十三种组合,满足十三器件五十二对不同对。如图7b所示,每个电极标记了十个通道,但这些区域可旋转切换位置。因此,使用十通道构建3×3阵列,而非三十六电极对应的三十六通道。综上,九个器件彼此不同,每个器件呈现四种功能,类似于MAHN在“贪吃蛇”游戏中的应用。

通过通道优化,在图7c和电影V8中演示了虚拟足球游戏。为满足实际应用场景,鞋外底粘贴丁腈薄膜,丁腈薄膜充当上述活动部分中的铝薄膜。类似地,接触、分离及不同方向移动的输出测试如图S15所示。上述活动部分中的磁铁也可嵌入鞋垫,两部分间的磁吸引力有助于用户定位特定点。如图7d和图S16所示,当用户踩踏MAHN固定部分时,Arduino板将识别用户所在区域。演示中,阵列的九个区域与九个位置一一对应。当用户踩踏“区域2”时,游戏中的球员定位于“位置2”。类似地,“位置6”和“位置7”在成功射门时定位。用户在前踢、左踢和右踢时,每个区域的四个电极产生不同信号。在“区域2”通过前踢完成成功射门。在“区域6”左踢失败后前踢成功。在“区域7”尝试右踢和前踢。综上,构建了MAHN的3×3阵列,并通过遍历法简化通道而不损失功能,实现更复杂控制。

**3. 结论** 本文设计并优化了一种磁相互作用辅助的混合式摩擦电-电磁纳米发电机,采用一对相互吸引的磁铁及带有微结构的硅胶基缓冲垫。MAHN基于接触分离结构的TENG和EMG,整体尺寸仅为35 mm × 35 mm × 8 mm。静止状态下,活动部分与固定部分相互吸引。TENG和EMG的输出因强磁吸引力而增强,且双磁铁设计产生的磁吸引力减弱了低频运行下输出性能的下降效应。MAHN被证明为撞击能量的高效混合式能量收集器,同时也是先进HMI的自供电传感器。TENG在1.1 MΩ负载下最大峰值功率为6.89 mW,EMG在1.1 kΩ负载下最大峰值功率为2.7 mW。得益于所设计的混合式降压电路,TENG和EMG的输出被整流与集成,用于对不同电容器充电。在30秒内,100 μF电容器可充电至2.03 V,性能优于单个单元及并联连接电路。通过调节磁铁数量驱动温度计和蓝牙模块,展现了能量收集器的高适应性。此外,完整摩擦电电极被交叉分割为四个电极用于人机接口应用。通过同时利用摩擦电信号和电磁信号,MAHN可在娱乐、办公和游戏控制中作为先进HMI,提供更高的控制能力和场景适应性。结合Arduino平台,MAHN可通过判断四电极正峰值传输方向指令,在经典贪吃蛇游戏中实现摇杆式控制。此外,通过用户手持活动部分在固定部分上方挥动,开发了非接触操作,可识别简单空中手势用于非接触控制。结合分体式设计的接触分离,MAHN帮助用户控制PowerPoint文档,包括全屏、翻页和结束演示,实现了商用激光指针的大部分功能,但采用自供电设计。MAHN还可轻松吸附于铁质家具和家用电器,类似于冰箱贴,用于智能家居中的非接触交互。此外,制备了MAHN的3×3阵列,并通过遍历法简化通道而不损失功能,实现更复杂控制。因此,将丁腈薄膜粘贴于鞋外底,演示了虚拟足球游戏。用户踩踏不同区域以协助球员定位,不同区域的功能协助球员向不同方向射门。因此,具有磁相互作用的混合式摩擦电-电磁纳米发电机在分布式机械能收集中具有广阔前景,并在HMI领域展现出重大创新。

**4. 实验部分** **4.1. 磁吸引力辅助混合式摩擦电-电磁纳米发电机(MAHN)的制备** 首先,使用3D打印机(ANYCUBIC 4Max Pro)制备两个带圆孔的刚性方形结构(聚乳酸PLA)。一对相互吸引的磁铁分别用双面胶带粘贴于孔中,使两刚性表面紧密吸附。然后,将匝数为3800、直径24 mm、厚度1 mm的Cu线圈固定于3D打印刚性框架,并用Al胶带粘贴于其中一个方形部分的刚性表面。另一方形部分覆盖Al胶带。

其次,通过在35 mm × 35 mm商业导电织物上涂覆并固化Exo-flex 0030制备缓冲垫。然后将缓冲垫剥离并平铺于前述刚性框架上,微结构面朝向框架。

第三,通过将Al胶带粘贴于FEP薄膜(50 µm厚,10 cm宽,DUPONT)上制备用于能量收集器的方形电极和用于自供电传感器的交叉电极。然后将薄膜平铺于缓冲垫上,并用附加部分固定于刚性边缘。

**4.2. 表征与电学测量** 输出数据由多通道示波器(Keysight Model DSOX3034T)采集与保存,其中使用1000X探头(TT-HVP-15HF)测量TENG电压信号,1X探头(GTL-101)测量EMG电压信号。TENG电流信号使用静电计(Keithley Model 6514)测量,EMG电流信号使用低噪声电流前置放大器(SR570)测量。充电性能使用静电计(Keithley Model 6514)测量。光学照片使用OLYMPUS BX53M拍摄。

**作者贡献声明** 刘龙:概念化、数据整理、形式分析、研究、方法论、资源、软件、验证、可视化、初稿撰写、审阅与编辑。施琼峰:研究、验证、形式分析、可视化、审阅与编辑。孙忠达:研究、验证、形式分析、可视化、审阅与编辑。李焯耀:概念化、方法论、监督、项目管理、资金获取、审阅与编辑。

**利益冲突声明** 作者声明不存在可能影响本论文报道的已知竞争性财务利益或个人关系。

**致谢** 本研究部分受中国国家重点研发计划(项目编号2019YFB2004800,子课题R-2020-S-002)资助,该计划在新加坡国立大学苏州研究院(苏州)执行;同时受新加坡-波兰联合资助项目(R-263-000-C91-305)“用于监测恶劣工业气体的芯片级MEMS微光谱仪”资助,该项目由新加坡科技研究局(A*STAR)与波兰国家学术交流署(NAWA)“弗罗茨瓦夫科技大学学术国际合作项目”共同支持。

**附录A. 补充信息** 与本文相关的补充数据可在在线版本获取,doi:10.1016/j.nanoen.2021.106154。