Effect Of Humidity And Pressure On The Triboelectric Nanogenerator Rapid Communication

Transcript

Nano Energy (2013) 2, 604–608 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/nanoenergy RAPID COMMUNICATION Effect of humidity and pressure on the triboelectric nanogenerator Vu Nguyen, Rusen Yangn Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN 55414, USA Received 7 July 2013; accepted 28 July 2013 Available online 8 August 2013 KEYWORDS Abstract Triboelectric nanogenerator; Energy harvesting; Relative humidity; Pressure; Aluminum; PDMS Triboelectric nanogenerator (TENG) as a novel energy harvester has raised wide research interests recently, although the environmental effect on its performance is still lacking. This paper attempted to experimentally quantify the effect of the relative humidity (RH) as well as the pressure on the charge generated from a TENG. It was found that under the same mechanical excitation, the generated charge of the TENG increased more than 20% when RH decreased from 90% to 10% at the ambient pressure. However, the generated charge decreased as the air pressure decreased from atmospheric pressure to 50 Torr at RH close to 0%, while the optimal pressure shifted to a lower level when the relative humidity was higher. This study can contribute not only to the design and packaging of triboelectric devices to operate in varying environments and extreme conditions, but also to the fundamental understanding of the mechanism of triboelectric effect. & 2013 Elsevier Ltd. All rights reserved. Introduction Recently, the triboelectric effect has been demonstrated for a variety of applications such as energy harvesting, pressure sensor or mercury ion sensor [1–7]. These devices are tested in ambient conditions to convert common mechanical energy, such as human movements, to electrical energy, to instantaneously power electronic devices [4–6], or to serve as selfpowered sensors [2,7]. The above mentioned devices have used the mechanism in which two surfaces are repeatedly pressed together under applied force and separated by some n Corresponding author. Tel.: +1 612 626 4318. E-mail address: [email protected] (R. Yang). 2211-2855/$ - see front matter & 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.nanoen.2013.07.012 types of spring [3–6]. Among the surface materials reported for the triboelectric effect, aluminum and poly(dimethylsiloxane) (PDMS) are of special interests due to their highly opposite positions in the triboelectric series [8]. Their low-cost and good manufacturability at micro/nano scale also make them very suitable for triboelectric effect studies [4–7], although other metals, such as gold and steel, and polymers, such as polystyrene, polytetrafuoroethylene (PTFE), polycarbonate, and polyethylene terephthalate (PET), have also been investigated [9–13]. The nature of triboelectric effect has been widely studied and a comprehensive review is provided in [9]. Triboelectric effect usually involves more than one of the following mechanisms: electron transfer, ion transfer and material transfer. Electron transfer is believed to dominate in a Triboelectric nanogenerator metal-polymer contact because experiments show that the charge transfer depends on the work function of the metal [10]. “Water bridge” layer on polymers can also contribute to the ion transfer and the transferred ion can be either native ions in the polymer or the hydroxide groups [11–13]. In other experiments [15], contact charge is still observed between metals and polymers at 1 μTorr [14], and between insulators at 5 Torr [15], indicating a wide application range of the triboelectric effect at different pressures. In this work, we report the effects of both the relative humidity and the pressure on the triboelectric nanogenerator (TENG) using the aluminum and PDMS as contact surface materials. A functional TENG was fabricated and the surfaces were modified with micro-patterns to enhance the charge generation. The device was operated in a commercial desiccator in which both the relative humidity and the pressure were monitored while the surfaces were repeatedly pressed and separated. This study benefits not only the design and package of triboelectric devices for varying environments, such as in different seasons, or for extreme conditions, such as in space, but also the fundamental understanding of the mechanism of the triboelectric effect. Materials and method Materials As shown in Figure 1a and b, the TENG in this study consisted of aluminum and PDMS electrodes, a spacer, a flexure, and a base where the two electrodes were mounted. The out-ofplane flexure structure was designed and fabricated to enable the complete contact and separation between the aluminum and PDMS surfaces. The flexure was made from a commercial Delrin (Polyoxymethylene) thin sheet (0.8 mm thick), and the spacer and the base were made of acrylic. These plastic parts was fabricated by a laser cutter. The length, width, and thickness of the flexural beam need to be well designed for a suitable stiffness which is high enough for the device to easily spring back after load and low enough to efficiently respond to the mechanical stimulus. The main dimensions of the flexure structure are provided in Figure 1c, which yield about 21 N/cm total stiffness, but the optimization in terms of the efficiency was not done in this work. The metal electrode used in this study was a 20 mm  20 mm patterned silicon substrate sputtering deposited with a 150 nm-thick aluminum layer. The pattern on the silicon surface, shown in Figure 1d, was designed with sharp tips to facilitate protrusion into the PDMS surface for increased contact area. The silicon wafer was first dry-etched in O2 and SF6 to obtain vertical-wall columns about 15 μm tall. The wafer continued to be wet-etched in potassium hydroxide (KOH) at 70 1C for about 4 min to obtain the pyramid structure with sharp tips. The PDMS electrode, which was of the same size as the aluminum electrode, was patterned with pyramids by molding PDMS in the KOH-etched silicon mold, a process similar to the one used in [4]. The Al-coated silicon and PDMS pieces were attached to the spring and the base. The thickness of the spacer was selected so that the distance between the two electrodes at rest was about 605 1 mm. The fabricated TENG was tested with a cyclic pressure force approximately 250 N by a human hand, and a maximum output voltage of 140 V was achieved. The observed maximum voltage was under ambient relative humidity of about 43% and room temperature of 27 1C. The generated charge was estimated to be 50 nC. Method Contradicting humidity dependence of triboelectric effect has been reported. In the experiment described in [11], rolling steel beads with polystyrene powder generates charges that increase monotonically with RH. Another experiment in [16] shows that a water-free environment yields lower charge than humid (about 40%) air for contact between PDMS with various polymers. Considering the significance of the humidity in the environmental control and in the performance of the triboelectric nanogenerator, the humidity dependence for triboelectric charge generation was systematically studied. We also investigated the effect of the pressure on the charge generation, because the pressure can affect the relative humidity and it is also of special interests for devices working at, for example, different sea levels or even in space. An apparatus in Figure 2a was employed to examine the effect of the environment RH and pressure on the charge amplitude. TENG was operated in the desiccator under controlled RH and pressure. The humid air from a water reservoir and the dry air from a dry air cylinder (Matheson, water content o6 ppm) were fed into the desiccator through a 3-way valve. By switching the valve between the humid/dry air sources, the humidity inside the chamber could be either increased or decreased. The desiccator was connected to a vacuum pump and the pressure inside the desiccator was controlled through adjusting the pumping rate. The humidity sensor, HIH-5031 from Honeywell, can measure RH with an accuracy of 73% from 11% to 89% and 77% otherwise, with a specified response time of 5 s. The pressure sensor was the 626B Baratrons Absolute Capacitance Manometer (0.1–1000 Torr). Humidity dependence experiment According to the study in [4], the generated charge on the PDMS surface will not be stabilized and maximized until after a few hundred cycles. To minimize the effect of this “warm-up” process, the linear motor was kept running throughout the whole experiment, pressing and releasing the TENG at a frequency of 2.5 Hz. This frequency was chosen to acquire data in a reasonable time and to maintain the mechanical stability during the experiment. The humidity in the desiccator was increased or decreased in steps of approximately 10% between 10% and 90% by feeding different amounts of the humid air and the dry air into the desiccator at about 5 L/min, with negligible pressure variation. After the desired level of RH was reached, the air flow was stopped and the data was recorded, as shown in Figure 2. This procedure was repeated continuously for different RH levels when RH decreased from 90% to 10% and increased from 10% back to 90%. Many cycles have been tested and the reproducibility was confirmed. The charges from the electrodes were 606 V. Nguyen, R. Yang Figure 1 Structure of fabricated TENG. (a) Schematic diagram of the TENG, (b) Optical image of a real device, (c) Structure of the flexure with dimensions, (d) SEM images of patterns on Al-coated Si surface, and (e) SEM image of patterns on PDMS surface. Figure 2 Experiment setup and measurement. (a) Experiment setup for humidity and pressure control, and (b) triboelectric charge generation when the device was cyclically pressed. recorded by an electrometer (Keithley 6517B). The measured charge was smaller than the maximum charge obtained by the hand pressure due to the smaller force from the linear motor and the larger capacitance from longer leading wires. Since the linear motor was kept running all the time, the charge vs. time plot was periodic as shown in Figure 2b. All the maximum and minimum charge values in a 10 s window were captured, and the peak-to-peak (p–p) values were calculated and averaged. The average p–p charge was plotted against different humidity levels in Figure 3a for three cycles of decreasing and increasing Triboelectric nanogenerator 607 Figure 3 Humidity and pressure dependence of charge generation (a) p–p charge at various RH with fitted trend line, (b) p–p charge at various pressure with fitted trend line, (c) The response of p–p charge with time as the RH is continuously reduced from 95% to 10%, and (d) The response of p–p charge with time as the RH is continuously increased from 10% to 82% (d). All results are at room temperature of 27 1C. RH. Additionally, to examine the transient response of p–p charge to RH, the RH was decreased or increased all the way to 10% or 90% by feeding only the dry air or the humid air respectively, and the p–p amplitude of generated charges with time was recorded and shown in Figure 3c and d. Pressure dependence experiment In this experiment, the chamber was pumped down from the atmospheric pressure to targeted pressures at 650, 550, 450, 350, 250, 150 and 50 Torr with an error of 78 Torr. The RH was maintained close to 0% in the experiment to minimize the effect of the pressure on the RH change. Similar to the humidity experiment, the linear motor was also kept running throughout the experiment at 2.5 Hz. The pressure test was repeated three times and the results were averaged and summarized in Figure 3b with the error bar showing one standard deviation of the data. should be thin, approximately below 2 nm, for the ion transfer. A thick water layer increases the surface conductivity, and therefore, discharges the surfaces. Measurements of the average thickness of the adsorbed water layer have also been performed for materials with various compositions, showing different values for different compositions [11]. Thus, this experiment suggests that at ambient condition, the water adsorbed on the surfaces from the high RH dissipates the generated charge and is not in favor of TENG. The transient result is shown in Figure 3c and d. It should be noted that the sharp increase at the early stage in Figure 3c is due to the fact that reducing RH is faster than increasing RH in this experiment setup. The results show that the p–p charge responds instantaneously to the change of RH, and its value at each RH agrees well with the values obtained in Figure 3a when the RH is varied step by step. The finite response time of the RH sensor does not appear to significantly affect these transient data, possibly because the change in RH happens in an extended time. Results and discussions Humidity dependence experiment Pressure dependence experiment As shown in Figure 3a, the p–p charge increases about 25% when RH decreases from 90% to 10%, and the p–p charge decreases to its original level when the RH increases back to 90% from 10%. RH reading produces a larger error outside 10–90% range, but similar trend has been observed, i.e. generated charges increase when the RH decreases. The trend found in this experiment suggests that in ambient condition, high humidity is not favorable for triboelectric effect between PDMS and aluminum. The effect of the humidity on other materials needs further investigation. The adsorbed water layer plays an important role in the charge generation [13]. Although the water layer is necessary to distribute and separate the charge, the required layer According to Figure 3b, the p–p charge decreases with an increasing rate as the pressure decreases from ambient pressure to 50 Torr. Because lowering the chamber pressure will cause the adsorbed water layer to evaporate, the result suggests that decreasing adsorbed water no longer favors the triboelectric charging when the pressure decreases. Combining with the results from humidity experiment, an optimal amount of adsorbed water appears to be at about atmospheric pressure and 0% RH. Since triboelectric effect usually involves different mechanisms, the charge will probably not go to zero as the pressure is further decreased. It is worth questioning how pressure would affect the charge if RH is at a higher level. Although it is difficult to control RH 608 at high values due to the pressure-induced RH change, it is observed that the charge starts to drop with the pressure at a lower pressure than ambient pressure- at 500 Torr and 55% RH, for example. The charge actually increases slightly as the pressure is decreased from 1 atmospheric pressure to 500 Torr at about 55% RH. In the other words, the optimal working condition of the TENG shifts to a lower pressure than the atmospheric pressure when the humidity is high, and the mechanism is still under further investigation. The described experiments have confirmed the ability of TENG to function at a low pressure environment. They also suggest that the pressure inside the package of the TENG can also be utilized to maximize its performance. Conclusions This paper has experimentally studied the effect of the relative humidity and the pressure on the performance of TENG. TENG can work at different pressure from 50 Torr to an atmospheric pressure and at different relative humidity from 10% to 90%. This study also suggested that the ambient pressure at RH close to 0% maximizes the triboelectric charge generation, and the optimal pressure drops to a pressure lower than the atmospheric pressure when the humidity is higher. This work explains how the surrounding conditions affect the triboelectric nanogenerator and may guide the fabrication and package of the device for the optimal performance. Acknowledgments The authors are truly grateful for the financial support from the Department of Mechanical Engineering and the College of Science and Engineering of the University of Minnesota. Parts of this work were carried out in the Minnesota Nano Center (MNC), University of Minnesota, which receives partial support from NSF through the MRSEC program. References [1] Feng-Ru Fan, Zhong-Gun Tian, Zhong Lin Wang, Nano Energy 1 (2) (2012) 328–334. [2] Feng-Ru Fan, Long Lin, Guang Zhu, Wenzhou Wu, Rui Zhang, Zhonglin Wang, Nano Letters 12 (2012) 3109–3114. [3] Guang Zhu, Caofeng Pan, Wenxi Guo, Chih-Yen Chen, Yusheng Zhou, Ruomeng Yu, Zhong Lin Wang, Nano Letters 12 (2012) 4960–4965. [4] Sihong Wang, Long Lin, Zhong Lin Wang, Nano Letters 12 (2012) 6339–6346. V. Nguyen, R. Yang [5] Guang Zhu, Zong-Hong Lin, Qingshen Jing, Peng Bai, Caofeng Pan, Ya Yang, Yusheng Zhou, Zhong Lin Wang, Nano Letters 13 (2013) 847–853. [6] Xiao-Sheng Zhang, Meng-Di Han, Ren-Xin Wang, Fu-Yun Zhu, Zhi-Hong Li, WeiWang, Hai-Xia Zhang, Nano Letters 13 (2013) 1168–1172. [7] Zong-Hong Lin, Guang Zhu, Yu Sheng Zhou, Ya Yang, Peng Bai, Jun Chen, Zhong Lin Wang, Angewandte Chemie International Edition 52 (2013) 1–6. [8] A.F. Diaz, R.M. Felix-Navarro, Journal of Electrostatics 62 (2004) 277–290. [9] Meurig W. Williams, AIP Advances 2 (2012) 010701. [10] J. Lowell, Journal of Physics D: Applied Physics 9 (1976) 1571–1585. [11] S. Pence, V.J. Novotny, A.F. Diaz, Langmuir 10 (1994) 592–596. [12] Jason A. Wiles, Marcin Fialkowski, Michal R. Radowski, George M. Whitesides, Bartosz A. Grzybowski, Journal of Physical Chemistry 108 (2004) 20296–20302. [13] Logan S. McCarty, George M. Whitesides, Angewandte Chemie International Edition, 47, 2188–2207. [14] D.K. Davies, Journal of Applied Physics 2 (1969) 1533–1537. [15] M.D. Hogue, C.R. Buhler, C.I. Caller, T. Matsuyama, W. Luo, E.E. Groop, Insulator–Insulator Contact Charging and Its Relationship to Atmospheric Pressure. [16] H.Tarik Baytekin, Bilge Baytekin, SiowlingSoh, Bartosz A. Grzybowski, Angewandte Chemie International Edition, 50, 6766–6770. Vu Nguyen received his B.S. degree in Mechanical Engineering from Worcester Polytechnic Institute, Worcester, Massachusetts in 2012. He is currently pursuing Ph.D. degree at the University of Minnesota, Minneapolis, Minnesota. His research interests are energy harvesting and self-power systems at micro/nano scale. Rusen Yang received his Ph.D. degree in Materials Science and Engineering from Georgia Institute of Technology in 2007, where he continued as Post-Doctoral Associate. He joined Mechanical Engineering at the University of Minnesota-Twin Cities as an assistant professor in 2010. He has discovered novel nanostructures, such as ZnO, SnO2, Zn3P2, and investigated their application potentials. His most recent work on energy harvester based on piezoelectric nanomaterials made significant contribution in the field of renewable energy.