Magnetic lifting triboelectric nanogenerators (ml-TENG) for energy harvesting and active sensing

Multifunctional devices are of significance to the rapid development of smart cities in recent years, especially for the functionalities of energy harvesting and active sensing.¹ Due to the complexity of application scenarios, however, sensing devices typically need to be flexible with good applicability at the multiscale, which is ideally able to generate electrical power under the environment.² Nanogenerators have been developed to harvest mechanical energy, which is one of the most extensive energy sources in the environment, and provide the revolutionary technique for active sensing as well.^3–6 As a competitive blue energy solution, triboelectric nanogenerators (TENG) have been reported with the advantages of low cost, high environmental compatibility, and good applicability,^7,8 which have therefore been applied in various applications, such as heat transferring,⁹ active sensing,^10,11 and position and gesture monitoring.^12,13 In particular, TENG have been applied to detect the velocity of moving objects in transportation since TENG-enabled sensors can identify the electrical signals resulting from mechanical movements.^14–16 Given the variety of transportation situations, however, it is necessary to effectively program the TENG sensors to adapt to the vehicle signals while increasing the energy harvesting efficiency (i.e., effectively trigger TENG under various excitations).^17–19 The magnetic field has been used to enhance and maneuver the mechanical excitations for TENG.^20–22 The macroscopic self-assembly network has been developed using the encapsulated TENG for energy harvesting in water waves.²³ The network has demonstrated the capabilities of self-assembly and self-healing using the self-adaptive magnetic joints of the encapsulated TENG, which provides a reliable route toward large-scale utilization of water wave energy for self-powered devices in the ocean.

Here, we present magnetic lifting triboelectric nanogenerators (ml-TENG) for energy harvesting and self-powered active sensing. The ml-TENG are designed with magnetic mover systems to trigger the dielectric capsules to slide for generating electrical voltage/power. The magnetic capsules are particularly designed in three modes with respect to the assembly of the copper, aluminum, and Kapton layers. Experiments are conducted to investigate the energy harvesting response (i.e., voltage and electrical power) of the copper and aluminum ml-TENG in the three modes. The comparison indicates that the ml-TENG design with copper in mode 2 provides the highest energy harvesting efficiency, and the maximum open-circuit voltage of 4 V and output power of 340 µW per copper capsule are obtained. Numerical simulations are developed to validate the energy harvesting performance of the copper ml-TENG in mode 2, and satisfactory agreement is obtained. In the end, we develop a self-powered velocity active sensing system using the ml-TENG, and a field test is conducted to detect the velocity of an electrical bicycle with different velocities and to light the LED board. The maximum open-circuit voltage has been obtained as 7.2 V at the velocity of 15 km/h. The reported ml-TENG have been used to develop active sensing systems for real-world applications, such as velocity detection in transportation. This study takes advantage of magnets as active excitations to more effectively trigger the TENG capsules subjected to arbitrary vehicular motions. The magnetic system enables the return of the capsules to their original state in the unloading phase. In addition, the magnetic system in the ml-TENG is able to reduce the sliding friction and possible fatigue that typically exist in the existing lever-arm or spring systems. As a consequence, this study obtains improved electrical performance and durability over a longer time.

The reported ml-TENG consists of dielectric capsules and a magnetic frame to create the magnetic field that can be triggered by a magnetic mover subjected to axial displacement. Figure 1 demonstrates the design principle of the ml-TENG. Figure 1(a) shows the main design principle of the dielectric capsules in the aluminum and copper modes. The components of the copper ml-TENG include a copper capsule and a magnetic frame with central magnets. These components were fabricated by 3D printing using PLA materials. End magnets are installed on the end magnetic mover M1, and the central magnets that have the same magnetic pole as the end magnets are placed on the frame M2. The magnetic mover is overall hexagonal in shape, which is installed with rectangular magnets on each side M3. Note that the magnetic poles of these rectangular magnets are the same as the magnetic pole of the capsule magnet M5 such that the dielectric capsules can be pushed when the magnetic mover is descended. The central magnets of the frame M2 push the magnetic mover to return to the original position. Figure 1(b) presents the different positions of the magnetic capsules in the copper ml-TENG. Voltage and electrical power are generated from the motion of the copper capsules in the magnetic frame. The oscillation response and energy harvesting performance of the copper ml-TENG are demonstrated in Supporting Video 1. The material and geometric properties of the ml-TENG in the experiments are provided in the supplementary material, Sec. 1.

The ml-TENG were fabricated using copper and aluminum in three modes. Figure 2(a) shows the experimental setup of the copper and aluminum dielectric capsules of ml-TENG. The loading machine was applied to create the cyclic loading with different speeds (0.9, 1, and 1.1 Hz) during the time period of 12 s, and the digital oscilloscope was used to collect the voltage signal. Figure 2(b) presents the dimensions of the copper and aluminum layers used in the ml-TENG. The gap between each layer is fixed as 1 mm. Figure 2(c) displays the design of the copper and aluminum capsulates in three modes. In mode 1, the dielectric capsules are covered with one layer of copper or aluminum and the dielectric capsules are covered with one Kapton layer. In mode 2, the magnetic frame consists of two separate layers of copper or aluminum (i.e., the electrodes). The distance between the two electrode layers was 1 mm. In mode 3, the dielectric capsule was only covered with one Kapton layer. To collect the output voltage in modes 1 and 3, the oscilloscope was connected to the two electrode layers of the magnetic frame. For mode 2, on the other hand, one end of the oscilloscope wire was connected to the copper or aluminum layer of the magnetic frame, and the other end of the oscilloscope wire was connected to the dielectric capsule. Figure 2(d) shows the fabrication of the copper and aluminum capsules in the ml-TENG. The structures of the ml-TENG and the magnets were maintained the same when the material was changed from copper to aluminum.

Figure 3 illustrates the energy harvesting performance of the copper and aluminum ml-TENG consisting of one capsule in three modes. The loading time is fixed as 2 s for each case, and the electrical resistance in the closed circuit is varied from 10 kΩ to 5 GΩ. Figure 3(a) shows the output voltage of the copper ml-TENG in mode 1 in the open circuit and closed circuit with different electrical resistances. The maximum voltage is obtained in the open circuit as V_max = 2.96 V, and the minimum voltage output is V_min = 1.68 V with the resistance of 10 kΩ. Figure 3(b) shows the output voltages of the aluminum ml-TENG in mode 1 in the open and closed circuits. It can be seen that the same voltage dispersion trend is obtained with respect to the electrical resistance. The maximum voltage in the open circuit is V_max = 2.6 V and the minimum voltage output is V_min = 1.6 V with the resistance of 10 kΩ. Figures 3(c) and 3(d) compare the output voltages of the copper and aluminum ml-TENG in mode 2 in the open and closed circuits. In the same manner, the maximum voltages are gained in the open circuit and the minimum voltages are in the closed circuit with the electrical resistance of 10 kΩ. Figures 3(e) and 3(f) compare the voltages of the copper and aluminum ml-TENG in mode 3 in the open and closed circuits. In general, the output voltage of mode 1 is higher than those of modes 2 and 3 due to the design difference of the magnetic capsules [see Fig. 2(c)]. In addition, the voltage output of the aluminum specimens is slightly smaller than that of the copper ones because of the materials in the copper and aluminum ml-TENG.

Figure 4 shows the voltage and electrical power of the open-circuit ml-TENG subjected to the loading frequencies of 0.9, 1, and 1.1 Hz. Figure 4(a) presents the output voltage of the copper ml-TENG in mode 1 in the open circuit, and the maximum voltage is obtained as V_max = 2.96 V under the loading frequency of 1.1 Hz. Figure 4(b) shows the voltage of the aluminum ml-TENG in mode 1. A similar voltage distribution trend is obtained, and the maximum voltage is obtained as V_max = 2.6 V under the loading frequency of 1.1 Hz. Figures 4(c) and 4(d) compare the voltages of the copper and aluminum ml-TENG in mode 2. The maximum voltages are obtained as, respectively, 4 and 3.96 V under the frequency of 1.1 Hz. Figures 4(e) and 4(f) compare the voltages of the copper and aluminum ml-TENG in mode 3, and similar voltage distribution trends are illustrated for modes 1 and 2. Figure 4(g) shows the voltage of the copper and aluminum ml-TENG in the three modes with respect to the electrical resistance and the voltage distribution of the ml-TENG in mode 2 with respect to the electrical resistance. Note that the voltage of copper and aluminum ml-TENG in modes 1 and 3 has a little difference. Figure 4(h) shows the power of the ml-TENG with the electrical resistance. The peak power of ∼340 µW is obtained during the electrical resistance of 0–10 kΩ.

The ml-TENG consist of two concepts of triboelectric nanogenerators: freestanding (FS) and lateral sliding (LS) modes. Figure 5(a) shows the FS modes, where L₁ is the dielectric length (Kapton) and L₂ is the electrode length (copper) for electrode 1 and electrode 2; g is the distance between the two electrodes used in the FS modes, which is 1 mm; d is the distance between the electrode and the dielectric, which is near zero; and d₀ is the thickness of the dielectric. Furthermore, Fig. 5(b) illustrates the LS modes, where L₁ is the dielectric length (Kapton) and electrode length (copper) for electrode 2. In addition, 2 L₂ + 1 mm is the electrode length (copper) for electrode 1. In ml-TENG, FS modes used for modes 1 and 3 and LS modes are used for mode 2. Increasing the surface area of the copper electrode in mode 2 (LS) provides additional electron transfer; thus, more voltage is generated in mode 2 relative to modes 1 and 3 at the same frequency.

Figure 6 demonstrates the applications of the copper ml-TENG in mode 2 for energy harvesting and active sensing. The reported ml-TENG can be used as active sensors for measuring vehicle speed since voltage is generated when vehicles pass the ml-TENG. Figure 6(a) shows the field test setup of the ml-TENG speed sensors, including the copper ml-TENG, wireless power supply, monitoring data system, digital oscilloscope, circuit for connection, and electronic bike. Figure 6(b) presents the output voltages of the ml-TENG under the velocities of 15, 10, and 5 km/h. Since different velocities lead to different levels of voltage signals, the peak voltages are observed as V_max = 4.2, 5, and 7.2 V, respectively (see Supporting Video 2). Figure 6(c) illustrates the application of the ml-TENG embedded in roads for active sensing of vehicle speed. Figure 6(d) displays the application of the ml-TENG in energy harvesting to light the LED board. 12 LEDs are lighted by the ml-TENG with a rectifying circuit under the loading frequency of 1.1 Hz. The rectifying circuit in the ml-TENG is shown in Supporting Video 2.

Numerical simulations are conducted to validate the energy harvesting performance of the ml-TENG. The numerical study was developed in accordance with the method applied in Ref. 24. Numerical models are developed using the same geometric and material properties as the magnetic capsules of the copper ml-TENG in mode 2 in the experiments. Figure 7 shows the numerical modeling and results obtained in COMSOL Multiphysics. Figures 7(a) and 7(b) show the numerical setup, mesh, and electrical potential contour of the open-circuit copper capsules in mode 2. Figure 7(c) compares the experimental and numerical voltages of the copper ml-TENG in three modes. The maximum difference is obtained as Diff_max = 5.7% in mode 2. Figures 7(d) and 7(e) present the electrical potential and output voltages of the copper ml-TENG with the capsules designed in modes 1, 2, and 3, respectively. Good agreement is obtained between the experiments and numerical simulations.

In this study, we developed the magnetic lifting triboelectric nanogenerators (ml-TENG) for energy harvesting and active sensing. The ml-TENG were designed with the magnetic capsules with copper or aluminum layers, the magnetic mover to trigger the capsules to move, and the magnetic frame to create the magnetic field. Experiments and field testing were carried out to investigate the electrical performance of the ml-TENG. The findings indicated that the energy harvesting performance was significantly affected by the structural design (i.e., modes 1, 2, and 3) of the magnetic capsules in the ml-TENG, which was slightly affected by the triboelectric materials and loading speed. The copper ml-TENG in mode 2 was found as the most effective design with the maximum open-circuit voltage of 4.0 V and the peak power of ∼340 µW. Numerical simulations were conducted to validate the output voltage of the copper ml-TENG in mode 2, and satisfactory agreement was observed.

The dielectric capsules, magnetic frame, and magnetic mover in the ml-TENG were 3D printed with the material of PLA (Polymaker-Tough) using the 3D printer (Ultimaker-S3, Ultimaker, Inc.) with the maximum printing size of 223 × 223 × 205 mm³ and the maximum printing velocity of 50 mm/s. The capsules were manufactured using copper, aluminum, and Kapton. The driving and driven magnets were tightly fixed to the magnetic movers and frame, respectively, using super glue. The copper or aluminum layers were coupled with PLA capsules and frame, and the light coating of Kapton was spread on the surfaces of the capsules. Care was taken to ensure no air bubbles existed between the bonding surfaces.

In the experiments, the cyclic loading of the ml-TENG was provided by the loading machine. The output voltage was measured using the digital oscilloscope (RIGOL DS2102A-EDU Tech., Inc.). The high-speed camera FR-500s-25G with the maximum speed of 1550FPS at a full resolution model was used to record the loading and speed of the magnetic capsules in the ml-TENG.

The 3D finite element (FE) models were developed using COMSOL. The energy harvesting performance of the copper ml-TENG in mode 2 was investigated. The linear quadrilateral shell elements (S4R) were considered in the numerical study. The simulations were carried out using the same geometries, boundary conditions, angular displacement, and material properties as the experiments. The frequency and modal dynamic analyses were particularly carried out using COMSOL Multiphysics.

The supplementary material introduces details on the material and geometric properties of the capsulate TENG, experimental setup, and mesh independence in the numerical simulations.

This study was supported, in part, by the Key Research and Development Program of Zhejiang, China (Grant Nos. 2021C03180 and 2021C03181), the Fundamental Research Funds for the Central Universities, China (Grant No. 2020-KYY-529112-0002), the Chinese Government Scholarship (Grant No. 2019ZFY011294), and the China Scholarship Council (Grant No. 2019S0A023394). P.J. acknowledges the Startup Fund of the Hundred Talents Program at Zhejiang University, China.

The authors declare that they have no known competing financial interest or personal relationships that could have appeared to influence their conclusions.

A.M.N. contributed to conceptualization, investigation, data curation, and writing of the original draft; K.-J.I.E. contributed to investigation and data curation; P.J. contributed to conceptualization, resources, supervision, and writing (review and editing); Y.W. contributed to investigation; and Y.Y. contributed to writing of the original draft.

The data that support the findings of this study are available from the corresponding author upon reasonable request.