Arrangement optimization of water-driven triboelectric nanogenerators considering capillary phenomenon between hydrophobic surfaces

The rise in environmental issues has stimulated research on alternative energy. In this regard, triboelectric generation has received much attention as one of several new alternative energy sources. Among the triboelectric generation methods, solid-liquid triboelectric nanogenerators (SLTENGs) have been actively investigated owing to their durability and broad applicability. In this paper, we report on the optimum arrangement of SLTENGs to increase the generation of electrical energy. When hydrophobic SLTENGs are arranged in parallel with a specific intervening gap, the friction area between the water and the surface of the SLTENGs is changed owing to the different penetration distances of water between them. This difference affects the amount of triboelectricity generated; this change in the water contact area is caused by the capillary phenomenon. Therefore, we investigated the effect of the gap on water penetration and formulated an optimum arrangement to achieve optimum electricity generation efficiency when multiple SLTENGs are contained in a limited volume. The proposed optimum arrangement of SLTENGs is expected to have high utilization in energy harvesting from natural environment sources such as wave energy or water flow.

Nowadays, environmental issues such as global warming and climate change have become severe owing to the indiscriminate use of fossil energy. Moreover, the price of fossil energy will increase steadily because it is a finite resource. Therefore, many researchers have endeavoured to develop alternative green energies using sunlight, wind, tide, geothermal heat, and the like [1][2][3][4][5][6][7][8][9][10] . Among alternative energies, triboelectric nanogenerator (TENG) has received much attention in academia. TENG is a new invention that harvests energy using triboelectricity, which is common in our lives. Triboelectrification, also called contact electrification, is a common phenomenon that occurs when two different materials are in contact with each other  . Triboelectrification is commonly classified as solid-solid contact electrification and solid-liquid contact electrification 34,35 . Solid-solid contact electrification generally generates more energy than the solid-liquid type, but it is difficult to obtain sustainable energy from this owing to surface abrasion. In other words, the solid-liquid contact type can generate lower but continuous energy because the contact between solid and liquid does not cause abrasion of the solid surface. Therefore, studies have been carried out to amplify the amount of energy output through modification of the surface structure or stacking multiple TENGs [36][37][38] . Lee et al. reported on the effect of surface roughness on solid-liquid triboelectrification and proposed a method for fabricating high-performance solid-liquid type TENG [39][40][41][42][43][44][45] . When hydrophobic TENGs are stacked in parallel and submerged into a liquid, the penetration distance of the liquid into the gap between the TENGs varies owing to the capillary effect. Consequently, the effect leads to difference in the electricity generation.

Results and Discussion
Water-solid contact electrification of SLTENG. When the sliding contact electrification occurs, the charge and voltage follow the formula: If the immersed depth is defined as x, the total charge (Q) is given by contact where σ is the surface charge density and w is the width of the SLTENG. With regard to the relative interfacial velocity (v r (t)), the output current (I) can be expressed as The output voltage (V) can be expressed as where d and l are the effective dielectric thickness and the total friction length, respectively; ε 0 and ε r are the dielectric constant of vacuum and relative permittivity of the dielectric material, respectively 46,47 .
In triboelectric energy generation, definite contact and separation, and large contact area between liquid and the solid surface are critical factors for high electrical output. Therefore, a hydrophobic nano-hole structured surface of SLTENG was fabricated, as shown in Fig. 1(a). The surface roughness and impurities of bare aluminium were flattened and cleaned by electropolishing ( Fig. 1(b,c)). The removal of surface impurities slightly enhanced the hydrophilicity ( Fig. 1(a-I,a-II)). Through the subsequent anodization process, a superhydrophilic Al 2 O 3 surface with evenly distributed 20-nm nanoholes was fabricated ( Fig. 1(a-III,d)). Both the HDFS and the PTFE self-assembled monolayer (SAM) coatings chemically modify the surface and make it hydrophobic without making any changes to the surface structures. Although the SAM coating of HDFS can solely function as a hydrophobic surface, the chain formation of PTFE, -(CF 2 -CF 2 ) n -, increases durability against abrasion caused by continuous sliding friction of water and the surface.
www.nature.com/scientificreports www.nature.com/scientificreports/ The fabricated SLTENG comprised three layers: electrode, dielectric layer, and electrification layer ( Fig. 1(a-IV)). The dielectric layer was the AAO layer of approximately 20-μm thickness, which provided insulation between liquid and the electrode ( Fig. 1(e)). The electrification layer, which loses or gains electrons on contact, consisted of a PTFE coating. PTFE tends to be negatively charged in a triboelectric series. The SLTENG had a CA of 120°.
The simplified solid-liquid triboelectrification process of SLTENG is shown in Fig. 2. Before the SLTENG contacts with water, the entire system, including the SLTENG and water, is in a state of electrical equilibrium ( Fig. 2(a)). When the SLTENG comes into contact with water, the electrification layer of the SLTENG and water  www.nature.com/scientificreports www.nature.com/scientificreports/ are respectively electrified negatively and positively due to the triboelectric series and current flows from the SLTENG to water to maintain the electrical equilibrium ( Fig. 2(b)). When fully dipped in water, the SLTENG and water are in electrical equilibrium (Fig. 2(c)). When the SLTENG is out of the water, the surface of the SLTENG and water are electrified and current flows reversely ( Fig. 2(d)). When the SLTENG is fully out of water, the state is the same as in Fig. 2(a) and the process is repeated. The electrical output measurements of a single SLTENG are shown in Fig. 2(e,f). The output voltage and current of a single SLTENG are approximately 3 V and 3.2 μA, respectively, at a frequency of 2 Hz and amplitude of 1 cm using the electrodynamic shaker. The power variation by frequency and amplitude is shown in Fig. S1. The amount of electricity generated differs according to the frequency and amplitude, but we applied 2 Hz and 1 cm as the natural wave environment in this experiment. A parallel arrangement with a specific gap was devised to increase the amount of electricity generation.

Gap analysis between SLTENGs considering capillary effect. An experimental model was devised
to analyse the amount of electrical output based on the size of the gap between the SLTENGs arranged in parallel. Increasing in the number of electrodes resulted in increased electricity generation, but the volume of the entire TENG system also increased. In order to reduce the volume, it is essential to apply a narrower gap when multiple SLTENGs are stacked. However, the narrow gap increases the capillary pressure, which resists the penetration of water into the gap between the SLTENGs because of their hydrophobic property (Fig. 3(a)). Thus, the surface contact area with the water decreases and this leads to decrease in the electrical output. Capillary pressure is the difference in the pressures at the interface between two immiscible fluids such as water and air (Eq. (4)).
The capillary pressure Δp is expressed as 48 , where a, γ, and θ are the radius of the capillary, the liquid/vapor interfacial tension, and the contact angle of the material, respectively. As the gap decreases, the capillary pressure increases. In the case of a hydrophobic surface, the pressure is negative, which prevents water from penetrating the capillary. In this experiment, the gap is the sole parameter because the applied surfaces have the same wetting properties. As the gap between the hydrophobic plates varies, the height of the water in the gap varies owing to negative capillary pressure.
To identify the quantity of water that penetrates between the hydrophobic plates, stacked SLTENGs separated by a specific gap were immersed in a circular chalet with 1-cm-deep water. The heights to which water had www.nature.com/scientificreports www.nature.com/scientificreports/ penetrated between the plates were observed while increasing the gap between two surfaces from 1 mm to 5 mm ( Fig. 3(b-f)). As shown in the figure, the equilibrium of force is reached at the lower water level between the plates than outside the plates because of the higher capillary pressure. The penetration height of water between the plates is the lowest at the gap of 1 mm, which means that the capillary pressure is the maximum at 1 mm. As the gap increases gradually, the water level also increases. After 3 mm, it becomes almost equal to the height of the water outside the plates. This is because the influence of the capillary pressure generated between the plates is negligible, irrespective of the gap. Computer Aided Engineering (CAE) was performed to verify the height of water penetration between SLTENGs due to capillary phenomenon. A commercial program, ANSYS Fluent, was used for the analysis, and the result showed the same tendency of water penetration (Fig. S2). Also, the same tendency was observed under the real-time condition (Fig. S3).
Gap optimization between two SLTENGs. Figure 4(a) shows the actual experimental setup and illustration of the gap optimization between two SLTENGs. To investigate the change in electrical generation with gap between two SLTENGs, the sides and back of the fabricated SLTENGs were insulated using a maskant. The insulation was applied to obtain the electricity generated only from the inner gap between the two SLTENGs. The gap between the two SLTENGs was varied from 0 mm to 5 mm and their electrical energy evaluated at a frequency of 2 Hz and amplitude of 1 cm using the electrodynamic shaker.
The results are shown in Fig. 4(b,c). When the gap between the SLTENGs was 0 mm, the electrical output was virtually zero because there was no contact with water. When the gap was 1 mm, the measured generation value was 0.2 V and 0.25 μA. In spite of the 1-mm gap between the SLTENGs, the power generation was very low. Because the capillary pressure, which is inversely proportional to the gap, interfered with the penetration of water, only a small contact area existed between the SLTENG surface and water. In other words, the electrical output increased with the gap size because of the lowering of the capillary pressure. As a result, the power output gradually increased as the gap increased from 0 mm to 3 mm, and the maximum electrical output was 2.6 V and www.nature.com/scientificreports www.nature.com/scientificreports/ 3 μA at a gap of 3 mm. This amount was slightly lower than the energy output of the single SLTENG because the sides and back of the SLTENGs were insulated by the maskant. The measured energies at gaps of 4 mm and 5 mm were almost the same as that at the 3-mm gap. It is considered that the capillary pressure no longer affects water penetration when the gap increases beyond 3 mm, as observed in the previous gap analysis experiment. Therefore, the 3-mm gap between two SLTENGs is the best gap for maximum electrical output.
Areal density optimization in limited volume. The number of electrodes is also a critical factor for arranging electrodes efficiently with a specific gap within a limited volume. Therefore, it is necessary to study the dependence of power generation on the number of electrodes and the gap between the electrodes in a limited volume. Consequently, five electrodes with a 1-mm gap, four electrodes with a 2-mm gap, three electrodes with a 3-mm gap, and two pairs of electrodes with 4-mm and 5-mm gaps were placed within a limited distance of 1 cm (Fig. 5(a)). Other conditions for measurement were the same as in the gap optimization experiment.
The electrical outputs depending on the number of electrodes and the gap are shown in Fig. 5(b,c). Both output voltage and current were measured by a probe with an impedance of 10 MΩ. For reference, high-impedance probe measurement for open-circuit voltage was conducted and only the water penetration height was affected regardless of the contact area (Fig. S4). As shown in Fig. 5, the electrical outputs of the five electrodes with 1-mm gap are the lowest because the contact area of the SLTENGs with water is the smallest, as also shown in the gap optimization test. However, unlike in the previous test, the highest power occurs at the 2-mm gap. The power density can be expressed in watts, the product of voltage and current. Therefore, power density also follows the same trends of voltage and current, showing the highest amount of energy at 2 mm-gap. The highest power density was 8.84 μW/cm 3 , which was over 7 times higher than the same volume of 1.15 μW/cm 3 with 1 mm-gap condition. This result shows that the number of electrodes has greater influence on the power generation than electrode spacing when water in the gap penetrates to a certain height in the limited volume. Therefore, for efficient enhancement of triboelectric power generation, it can be concluded that while increasing the number of SLTENGs, due consideration should be given to the limited volume and the capillary phenomenon. www.nature.com/scientificreports www.nature.com/scientificreports/ conclusions In this paper, we proposed an optimum arrangement for SLTENGs for improving the power generation from solid-liquid triboelectrification. When multiple SLTENGs are arranged to increase the contact area with the liquid, a narrow gap between the SLTENGs needs to be avoided to allow water to penetrate into the gap. On the other hand, the gap size should not be excessively large to ensure high areal density considering the limited volume. Therefore, the optimum gap size for the highest efficiency of the SLTENGs was experimentally investigated. While 3 mm was the optimum gap size when considering solely the capillary phenomenon, the maximum power was generated at 2 mm-gap size considering the limited volume. Furthermore, although aluminium-based TENGs were used in this study, the applications of this study are not limited to the materials or the surface property. Rather, it can be applied to a variety of applications using water flow or wave power to generate a continuous electrical output, which is an advantage of solid-liquid TENG.