Suppressing the Leidenfrost effect by air discharge assisted electrowetting-on-dielectrics

Evaporation or boiling of droplets can effectively dissipate heat from high-temperature surfaces. In recent years, such phase-change as a paramount component has been widely used in power generation,^1,2 water purification,³ electronic/optoelectronic cooling,^4,5 and even biosciences.⁶ However, this method is only valid as long as the surface temperature is lower than a certain temperature. Once the surface temperature exceeds this value, the bottom of the droplet vaporizes rapidly and forms an insulating vapor layer between the droplet and solid surface, separating the droplet from the over-heated surface with a large contact angle of nearly 180°; this phenomenon is called the Leidenfrost effect (LFE),^7,8 and the critical temperature is called Leidenfrost point (LFP). The effective heat transfer between droplets and solids depends on their contact area, and this extremely large contact angle prevents the droplets from contacting the solid surface, which greatly reduces the heat transfer from the solid surface to the droplets, limiting the evaporation or boiling of droplets for heat dissipation, and can easily cause surface overheating or equipment damage. Therefore, in many practical applications, better wettability is desired to meet the demands of heat transfer.

Conventional strategies for augmenting the surface wettability rely on the introduction of micro-/nanotextures on solid surfaces,^9–13 where the surface microtextures decouple the liquid–solid contact and the insulating vapor. Using this strategy, efficient liquid–solid contact can be achieved, which significantly improves the wettability of the solid surface for facilitating the heat transfer. However, this strategy usually requires changing the original morphology and properties of the solid surface, and these micro-/nanotextures require special design; otherwise, the wettability of the over-heated solid surface will be deteriorated, leading to the premature emergence of the Leidenfrost effect.^14–16 Electric field was ever been used to try to influence the Leidenfrost effect by taking the advantage of electromechanical force to reduce the insulating vapor layer, which is usually referred to as electrowetting-on-dielectrics (EWOD).^17–21 EWOD is a versatile and convenient method to control wettability by electromechanically forcing a droplet to spread on a conductive substrate with dielectric layer in response to applied voltage.^22,23 EWOD relies on electromechanical force to properly compress the vapor film between the droplet and the substrate,^24–26 bringing the droplet into contact with the over-heated surface, thereby elevating the LFP to a certain extent. However, this effectiveness by EWOD was much limited because the electromechanical force was not strong enough to properly eliminate the vapor layer.

Here, we demonstrate a method to control the wettability of over-heated solid surface by the air discharge assisted electrowetting-on-dielectrics (ADA-EWOD), where the electrically induced air discharge around droplet was able to increase the solid surface free energy to cooperatively work with electromechanism force, decreasing the droplet contact angle well below the value obtained by the traditional EWOD strategy. The ADA-EWOD is high temperature tolerant that the contact angle can be decreased from nearly 180° down to less than 10° on an over-heated surface above 350°, well inhabiting the Leidenfrost effect. By taking the advantage of superhydrophilicity at high temperature, we employed the ADA-EWOD in droplet evaporation for heat dissipation, where the droplet lifetime is reduced by more than one order of magnitude, and a heated silicon wafer at 600 °C is cooled down to less than 200 °C within 20 s.

To show the mechanism of ADA-EWOD, we used the common test configuration that was usually used for traditional EWOD studies on a SiO₂/Si substrates, as described in Fig. 1. To maximize the initial contact angle and minimize the contact angle hysteresis, the silicon oxide insulating layer with the thickness of 1000 nm was hydrophobized with a topcoat of an amorphous fluoropolymer (Chemours Teflon® AF 1601X), producing an initial contact angle of 115°–120° on the substrates, as shown in Figs. 1(a) and 1(b). A drop of sodium chloride solution (0.5% mass concentration) was deposited onto the silicon oxide surface, and then, a needle electrode was injected into the droplet as a counter electrode. A sinusoidal AC signal with a frequency of f = 500 Hz was generated by an arbitrary waveform generator (Agilent 33220A) and then amplified by a high-voltage amplifier (TREK 610E), to output a range of voltages onto the droplet. A contact angle goniometer (Dataphysics OCA20) was used to measure the droplet contact angle under various applied voltages.

By gradually increasing the applied voltage to 350 V in the peak value, we realized stable and continuous air discharge near the three-phase contact line of the droplet. All voltages in this paper are peak values. According to previous literature, high-energy particles, such as O⁺, N²⁺, and OH⁻, generated by gas discharge acting on the solid surface to generate active groups, such as hydroxy and carbonyl, will significantly increase the solid surface free energy.²⁷ Therefore, in our ADA-EWOD strategy, there are two factors to enhance the solid wettability, including the electromechanical forces and the increase in solid surface free energy, and the combination of these two wetting factors results in a sharp drop in the water droplet minimum contact angle; as shown in Fig. 1(b), the droplet contact angle decreased from about 120° to 29°. In comparison, by using the conventional EWOD strategy, the contact angle of droplet was only reduced to 76.9° (Fig. S1 in the supplementary material), which is similar to what has been reported in the existing literature.^28,29

We explored the application of ADA-EWOD to improve the over-heated solid surface wettability performance and a silicon oxide wafer without hydrophobic layer as the substrate. It can be seen that driven by double wetting factors, the vapor gap between the droplet and the over-heated surface disappeared, allowing the droplet completed the transition from superhydrophobic state to superhydrophilic state, well inhabiting the Leidenfrost effect [Figs. 1(c) and 1(d)]. As a comparison, the cases without applied voltage and by traditional EWOD were also studied. In the experimental session, the SiO₂/Si substrate surfaces were first brought to 350 °C by a hot plate (Lichen FL-2Y), and then at a fixed voltage (U = 350 V), the droplet with a volume of ∼13 μl was released along the needle electrode at a distance of about 10 mm from the silicon wafer by a pipette. The morphology of droplets in contact with over-heated surfaces under different conditions was captured by a high speed camera (Optronis CP80-3-M-540), operating at 1000 frames per second. Figure 2 and Videos S1 and S2 (Multimedia view) illustrate the different dynamic behaviors of the droplets (volume of ∼13 μl) when they initially contacted the over-heated surface (T = 350 °C). It can be seen that without applied voltage, the droplet could not effectively contact the high-temperature surface due to the existence of the vapor film, and the droplet was in the Leidenfrost state [Fig. 2(a)]. Using traditional EWOD, under the action of electric field force, the vapor gap was squeezed when the droplet impacted the solid surface, which slightly increased the contact area between the droplet and the solid, but the droplet maintained the superhydrophobic state throughout the evaporation process, and more vapor was generated to make the droplet bounce off the solid surface [Fig. 2(b)]. In sharp contrast, using ADA-EWOD, the vapor gap disappeared and the contact area between the droplet and the solid surface increased significantly. Moreover, after the droplet bounced off, there was a liquid film on the solid surface, which means that the droplet completed the transition from superhydrophobic state to superhydrophilic state, and the contact angle of the liquid film was less than 10° [Fig. 2(c)].

The droplet lifetime, critical metrics for droplet cooling, was influenced by the applied voltage and solid surface temperature in the ADA-EWOD. In the droplet lifetime test experiment, the droplet was released along the needle electrode at a distance of about 10 mm from the silicon wafer by a pipette. The entire process of the droplet evaporation on the high-temperature silicon oxide surface was recorded by the CCD camera (Imagesource DFK 33UX265), operating at 30 frames per second. Figure 3 and Video S3 in the supplementary material show the droplets lifetime varies as a function of solid surface temperature and applied voltage by using different strategies. It can be seen that the droplet lifetime decreased with the increase in applied voltage until the minimum value, as observed on a solid surface with a given temperature of 350 °C [Fig. 3(a)]. This decrease in droplet lifetime was directly related to the decrease in contact angle, resulting in the larger contact area between liquid and solid to facilitate the heat transport, as confirmed by the red curve in Fig. 3(a). For instance, the droplet lifetime achieved by ADA-EWOD at the temperature of 350 °C was as short as 3.5 s, at least one order of magnitude shorter than the value of 54.3 s without applied voltage. At a proper applied voltage (such as 350 V) for generation of an adequate air discharge, the droplet lifetime always showed the significantly reduced value compared with that without applying a voltage or by traditional EWOD over a wide range of temperature from 100 °C to at least 450 °C [Fig. 3(b)]. The potential effect of heat generated by air discharge on droplet evaporation was also tested using an infrared thermal imaging camera with a macro lens. It can be seen that the local temperature change near the three-phase contact line of droplet before and after air discharge was less than 1 °C (Fig. S2 in the supplementary material). This magnitude of temperature change has a very limited effect on droplet evaporation, which once again confirms that the reduction in droplet lifetime is caused by the improvement in surface wettability. It was also notable that the droplet lifetime by ADA-EWOD tended to slightly increase with solid surface temperature above 350 °C, which was mainly due to that the droplet became more excited at a higher temperature by the violent insulator vapor, destabilizing the electric contact to the needle electrode (Video S4 in the supplementary material).

By taking advantages of the ability to turn a superrepellent, over-heated surface superhydrophilic, inhabiting the Leidenfrost effect, we employed the ADA-EWOD in droplet evaporation for efficient heat dissipation, as shown in Fig. 4. We measured the substrate temperature to demonstrate the capability of heat dissipation by ADA-EWOD, and the schematic diagram of the cooling performance test platform for droplet ADA-EWOD is shown in Fig. 4(a). The change in temperature during droplet evaporation was measured by a thermocouple sensor (Omega KMTXL-040G-6) located under the silicon wafer, in which the wafer was in direct contact with the sensor, and collected by a data collector (Jinko JK9000-8) with a sampling interval of 100 ms. After the temperature reached the set value, droplets were released by the syringe pump at a height of about 10 mm from the silicon wafer. An infrared camera (FLIR X6520sc) operating at 50 Hz was used to record temperature changes caused by water droplets.

It can be seen from Fig. 4(b) that the temperature reduction by ADA-EWOD was as high as about 40 °C after a 13 μl droplet impacting the 350 °C solid surface from a falling height of about 1 cm, whereas the temperature reduction was only 16 or 4 °C by traditional EWOD or direct impact without applied voltage, respectively, using the same volume of droplet from the same falling height. The decrease in temperature is directly related to the decrease in the contact angle, resulting in a larger contact area between liquid and solid surface, which is beneficial to heat dissipation, as confirmed by the infrared images in Fig. 4(c). The maximum droplet cooling efficiency³⁰ under the action of the ADA-EWOD can be estimated as $ε = ∫ 0 t c ∫ 0 A max q ″ d A d t / m c p Δ T$ ≈ 0.074, see the supplementary material for the details. Here, t_c is the time required for the droplet to reach its maximum extent, A_max the maximum contacted area between the droplets and the over-heated surface, q″ the heat flux from the interface, m and c_p the droplet mass and specific heat, and ΔT the difference between the droplet and over-heated substrate temperatures. In comparison, the droplet cooling efficiencies when using conventional EWOD and direct impact without applied voltage were 0.008 and 0.007, respectively. The heat transfer coefficient was estimated to be about 9.1 kW m⁻² K⁻¹ using Newton's law of cooling,³¹ which is at the same level as other reported data.^7,32 It should be noted that the heat transfer coefficient is a function of the droplet properties and temperature,^15,33,34 which means that the heat transfer coefficient can be further improved by changing the droplet material in the future.

In addition, we tested the continuous heat dissipation performance of ADA-EWOD. In the experiment, we used a syringe pump (Wenhao WH-SP-1) instead of the pipette to provide water droplets. As can be seen from Fig. S3 in the supplementary material, during the test time of nearly 300 min, there was no observable attenuation of the temperature drop caused by the ADA-EWOD, and the dielectric layer on the top of the substrate did not break down, which demonstrated that ADA-EWOD has good stability in heat dissipation from over-heated surfaces.

In engineering applications, heat dissipation can be enhanced by increasing the fluid supply. When droplets continuously impacting a 450 °C surface at a flow rate of 1 ml min⁻¹, the temperature was reduced down to a steady value of about 248 °C by ADA-EWOD [Fig. 4(d)], whereas the steady temperature was about 380 or 405 °C for that by traditional EWOD or direct drop without applied voltage, respectively, using the same flow rate of 1 ml min⁻¹ [Fig. 4(d)]. Moreover, by adjusting the flow rate, we found that the cooling performance by ADA-EWOD was controllable, as shown in Fig. 4(e). At a flow rate of 5 ml min⁻¹, the temperature dropped to below 200 °C within 18 s and eventually stabilized at about 170 °C, demonstrating the capability to effectively and stably inhibited Leidenfrost effect.

Considering what has been mentioned in the past literature,^13,15 the introduction of microtextures on the over-heated surface can effectively accelerate the heat dissipation because it can decouple the generation of water vapor from the liquid–solid contact; we further explored the heat dissipation performance of the combination of surface microstructuring and ADA-EWOD strategies. We employed a stainless steel mesh with a mesh umber of 50 as electrode laying on the solid surface, where the mesh worked as microstructures and in place of a single needle electrode, as shown in Fig. 5(a). The mesh openings allow for transport of both liquid and vapor, and the interconnected metal wires worked as electrode. Compared with the droplet lifetime of 12 s at 600 °C by simply relying on the surface microstructuring without applying voltage, the droplet lifetime of a 13 μl droplet on 600 °C surface was reduced down to 3.7 s at 500 V by ADA-EWOD [Fig. 5(a) and Video S5 in the supplementary material], indicating the efficiently improved heat dissipation. Moreover, at a droplet flow rate of 3 ml min⁻¹, the substrate temperature was reduced down to a steady value of less than 200 °C by ADA-EWOD, whereas the steady temperature was above 500 °C only by surface microstructuring, still in the Leidenfrost state [Fig. 5(b)]. The infrared thermogram of the solid surface visually showed that this significant decrease in temperature was owing to the spread and contact of water over a large area substrate by ADA-EWOD [Fig. 5(c)]. These results suggested that ADA-EWOD was a highly effective, practically viable technique to elevate the Leidenfrost point, up to at least 600 °C, well improving the droplet evaporation for heat dissipation.

In conclusion, the strategy of ADA-EWOD we demonstrated here provides a powerful solution to suppressing the Leidenfrost effect. This inhibition of Leidenfrost effect by ADA-EOWD was mainly due to maximized wetting property on a flat solid surface, fundamentally different from the commonly used method of solid surface microtexturing on bulk metal body. Therefore, our cooling method provides a possible strategy for improving the cooling of high-power chips for laser, highly integrated electronic/optoelectronic devices, LEDs, and even radars, where limited space is available for designing the millimeter-scale surface textures. Moreover, since various conductive liquids besides water—such as nano inks, liquid metals, and electrolytes—can generate air discharge on dielectric layers, we believe the ADA-EWOD may be useful in a wide range of droplet manipulations besides droplet evaporation demonstrated here, including the printing of flexible electronics,³⁵ electrochemical energy storage,³⁶ and ultrafast water transport,³⁷ where an extraordinary wetting property plays a critical role.