Bubble removal by electric and acoustic actuation for heat transfer enhancement

A wide range of devices essential to modern industry, including integrated circuits, nuclear reactors, and rocket motors, are required to control the heat generated during operation quickly.^1,2 To that end, liquids with higher rates of heat transfer than gases are used as heat transfer media, and boiling heat transfer is used to maximize the rate of heat transfer.³ For decades, boiling heat transfer has been hailed as an effective mechanism for removing large quantities of thermal energy from surfaces.⁴ In this boiling state in which boiling heat transfer takes place, the surface that guarantees the most efficient heat transfer is an area where air bubbles form, grow, and separate periodically.⁵ However, when excessive evaporation of the liquid occurs on this surface, the air bubbles building up on the heating surface form a gas layer, resulting in declining rates of heat transfer.⁶ To tackle these problems, many research groups have conducted research on shortening the cycle of bubble formation and removal and research on removing bubbles from surfaces by keeping them under control efficiently.⁷ Conventionally, research on removing air bubbles is divided into passive and active methods. In the former case, the wettability of a surface is changed by modifying the properties of that surface; in the latter, an external force is applied. Passive methods include surface mechanical machining, surface coating, chemical processing, and micro/nano-electro mechanical system (MEMS/NEMS) techniques.⁸ Jones and co-workers, for instance, conducted research to change the cycle of bubble formation and removal by varying surface roughness.⁹ Zhang and co-workers, on the other hand, created porous surfaces and applied hydrophobic coatings to enhance heat transfer.¹⁰ Rahman and co-workers used the surfaces of two materials with different conductivities to eliminate air bubbles.¹¹ Meanwhile, Choi and co-workers fabricated a downhill capillary wicking mechanism to remove air bubbles quickly and efficiently, in addition to using two materials with different surface conductivities.¹² However, these passive techniques have limitations in changing surface properties.¹³ For example, structural surfaces run the risk of damaging fine structures and are vulnerable to contaminants and clogging.¹⁴ If contaminants in liquids, which serve as heat transfer media, or particles are injected to enhance heat transfer clog cavities in structures, the structural surfaces cannot remain structurally intact enough to enhance boiling heat transfer. Surfaces coated with thin films of 1 µm or thinner to enhance boiling heat transfer are likely to suffer damage or degradation during chemical reactions.^1,11 To resolve these problems with passive techniques, many research groups have conducted research on active techniques that employ external forces. Acoustic waves and electric fields are among the most commonly used active techniques.¹⁵ The acoustic wave-induced bubble removal utilizes the oscillation of the acoustically excited bubble, which strongly occurs at the natural frequency of the bubble determined by the volume of the bubble. Heffington and Glezer first applied ultrasonic waves to air bubbles directly to enhance boiling heat transfer through a gaseous bubble removal.¹⁶ In further investigation, Douglas and co-workers verified the behavior of the gaseous bubble under the acoustic pressure field and facilitated the bubble removal with a light-weight, low-power acoustic actuator at a few kilohertz levels.¹⁷ The acoustic wave-based method, however, can effectively remove only a certain volume of bubbles whose natural frequency is matched with the applied acoustic frequency.

Other research groups have carried out studies on changing the geometries of air bubbles on electric fields [electrohydrodynamic (EHD) phenomenon] to enhance boiling heat transfer.¹⁸ In 1998, Kweon and co-workers applied high voltages from 0 to 13 kV to separate single bubbles from surfaces.¹⁹ In 2007, Chen and co-workers went further by applying voltages from 0 to 3 kV to separate air bubbles from surfaces.²⁰ However, the use of EHD in removing air bubbles comes with disadvantages, such as requiring high voltages.²¹ To resolve such problems, Wang and co-workers utilized electrowetting that operates at lower voltages than EHD, which could eliminate various bubbles forming within liquid films.²² This group also applied electrowetting to an experiment where air bubbles were removed by interfacial waves on the surface of a liquid film and an experiment where bubbles were removed from the liquid film by regulating applied voltages.^23,24 These studies on electrowetting-based bubble removal, however, can only be used in liquid film settings.

This study proposes a new bubble removal technology utilizing the dielectrowetting and acoustic actuation of the bubbles. Figure 1 shows a schematic image of the proposed bubble removal technology. As shown in Fig. 1(b2), when a certain voltage is applied to a surface composed of electrode and hydrophobic dielectric layers, the air bubble gets separated from the surface by the mechanism of dielectrowetting. Then, the bubble gets removed from the surface when it is acoustically excited by acoustic waves generated from a piezoactuator attached on a liquid tank, as shown in Fig. 1(b3). Through the combination of the two different types of actuations, the proposed bubble removal technology is capable of not only removing various volumes of bubbles at once but also operating with a significantly lower voltage than the conventional EHD-based bubble removal method. It can also be used in both liquid film and medium settings. The proposed technology uses the external force induced by acoustic actuation to eliminate air bubbles separated from a surface by dielectrowetting, ensuring faster and more efficient removal. Such advantages will guarantee that the bubble removal technology proposed in this study will find a wide variety of applications.

There are two methods for changing the wettability of liquid droplets on an open substrate using electric fields: electrowetting and dielectrowetting.²⁵ In electrowetting, the movements of free charges in conducting liquid droplets on the substrate are used to change the wettability of liquid droplets with electric fields.²⁶ On the other hand, in dielectrowetting, the polarization of dipoles is used within a non-conducting liquid.²⁷ When electric fields are used to change the wettability of liquid droplets on an open substrate, the change largely exceeds any limit to the wettability of liquid droplets posed in electrowetting. That is why the dielectrowetting technology has recently gained attraction as a future technology.^27,28 Dielectrophoresis refers to an electromechanical force occurring in non-uniform electric fields due to the polarization of neutral matter.²⁹ In 1985, Pellat and co-workers used a chip composed of two electrodes adequately spaced to analyze the movements of droplets using dielectrophoresis or liquid dielectrophoresis (L-DEP). The effect of the L-DEP force was subsequently demonstrated in an experiment where a voltage was applied to two electrodes to control the height of non-conducting liquid droplets laid on a chip.^27,30 In 2011, McHale and co-workers fabricated patterned electrodes through microfabrication, applied a hydrophobic layer to the surface, and used the droplets of propylene glycol to experimentally prove the L-DEP effect on the solid–liquid surface and the relationship between voltages and contact angles in dielectrophoresis. This group was also the first to use the word “dielectrowetting,”³¹ 

where θ_e is the contact angle of a liquid droplet when the applied voltage is V, θ₀ is the contact angle of the droplet when the applied voltage is 0 V, V is the voltage applied across the dielectric liquid, ɛ₀ is the absolute permittivity of vacuum, ɛ_L is the permittivity of the liquid, ɛ_V is the permittivity of the gas surrounding the liquid, γ_LV is the surface tension occurring at the liquid–vapor interface, and δ is the penetration depth of the electric potential.

Equation (1) refers to the relationship between the applied voltage and the contact angle of the droplet during dielectrowetting. Instead of the ratio of the thickness of the dielectric layer (t) to permittivity (ɛ₀ɛ_s) in the Lippmann–Young equation for electrowetting [$cosθe=cosθ0+ε0εs2γLVtV2$⁠, where V is the voltage applied across the dielectric layer, ɛ_s is the permittivity of the dielectric layer, and t is the thickness of the dielectric layer^32,33], this equation shows the ratio of the permittivity difference between the liquid and vapor to the penetration depth of the electric potential.

As with existing L-DEP experiments, dielectrowetting can control not only dielectric droplets on substrates in a gaseous environment but also air bubbles in a non-conductive liquid medium in a diametrically opposite environment.³⁴ Once the electric potential difference occurs between the non-conductive liquid surrounded by air bubbles and the electrode featuring air bubbles, the contact area between the bubbles on the electrode, which consists of hydrophobic and dielectric layers, and the substrate surface can be reduced by dielectrowetting as a result of declining contact angles. Consequently, when a voltage greater than the threshold voltage is applied to the electrode, the contact area between bubbles and the substrate will be significantly reduced, causing the bubbles to separate from the surface.³⁵ 

To separate air bubbles from the substrate surface, a dielectrowetting chip was fabricated through the standard MEMS fabrication processes. Figure 2 shows a schematic view of this MEMS fabrication process. Prior to the fabrication of the chip, sliding glass was submerged in the heated diluted piranha solution for about 10 min, blow-dried using nitrogen gas, and placed on top of a hot plate for about 30 min so that the cleaned glass can be prepared as shown in Fig. 2(a).

To create an electrode layer on the sliding glass as shown in Fig. 2(b), the layer was deposed sequentially with a chrome coating as thick as 200 Å and a gold coating of 1000 Å using sputtering equipment. Then, as shown in Figs. 2(c) and 2(d), photoresist (AZ 7220) was spin-coated on top of the electrode layer. UV photolithography was then performed for patterning. Then, as shown in Fig. 2(e), the patterned chrome and gold were wet-etched to form interdigitated electrodes (200 µm width and 50 µm gap) on a flat surface for dielectrowetting. As shown in Figs. 2(f) and 2(g), acetone was used to remove the remainder of the photoresist. Then, a parylene coating system was used to depose a parylene-C dielectric layer (with a thickness of 2.5 ± 0.1 µm). Finally, as shown in Fig. 2(h), Teflon AF1600 (DuPont, a thickness of ∼100 nm) was spin-coated on top of the dielectric layer to depose a hydrophobic layer.

Figure 3 shows a schematic diagram of experimental setups. To perform dielectrowetting, a certain electrical signal was generated using the signal generator (33210A, Agilent Co.) and amplified 100 times more using the voltage amplifier (PZD700, Trek Co.) before it was applied to the patterned electrodes on the dielectrowetting chip. To perform acoustic actuation, an electrical signal at a certain frequency was generated using the signal generator (33210A, Agilent Co.) and applied to a disc-shaped piezoactuator (KPR3020LC-450, Daeyoung electric Co.) mounted on the side of a liquid tank. All the experimental results were filmed using a CCD (charge-coupled device) camera (EO-1312C, Edmund Optics) featuring a zoom lens (VZMTM 450i eo, Edmund Optics) and a high-speed camera (Phantom Miro eX4, Vision Research, Inc.), and the data were stored on a personal computer.

To understand the behavior of air bubbles affected by the mechanism of dielectrowetting, the substrate was positioned inside a liquid tank [70(L) × 50(W) × 30(H) mm³] filled with 98% propylene glycol liquid, and a micropipette (HHT-D10, HTL high tech Lab) was used to generate air bubbles of 1, 3, and 5 µl at the bottom of the substrate. Then, the bubbles were positioned at the bottom of the substrate, as shown in Fig. 4. Next, a signal generator and an amplifier were used to apply a voltage of 23 kHz to the patterned electrodes. Due to dielectrowetting, the contact angles of the bubbles decreased in proportion to the declining contact area between the bubbles and the substrate as the applied voltage increased. Consequently, Fig. 4 demonstrates that the bubble (5 µl) is completely separated from the substrate surface at the applied voltage of 260 V. It also presents a change in the geometry of air bubbles caused by the mechanism of dielectrowetting, showing that the proposed mechanism of dielectrowetting can separate air bubbles from a substrate surface.

Figure 5 shows the contact angle and separation distance according to the different voltage amplitudes at 23 kHz. When the applied voltage increased from 0 to 240 V in increments of 20 V, the contact angles of 1, 3, and 5 µl bubbles decreased similarly, but the contact angle decreased more significantly at voltages greater than 125 V. Furthermore, the bubbles began to separate from the substrate at 260 V, and the separation distance of the bubbles was proportional to the applied voltage. In addition, for smaller 1 µl bubbles, the separation distance was also smaller than the distances of 3 and 5 µl bubbles and was found to be ∼0.15 mm at 340 V.

Figure 6 presents the variation in the contact angles of air bubbles according to the different frequencies at the constant applied voltage of 220 V. When the frequency of the applied voltage increased from 1 to 29 kHz in increments of 2 kHz, the contact angles of 1, 3, and 5 µl bubbles tended to increase, while the contact angles decreased significantly at 25 kHz or higher. At frequencies around 23 kHz, 1, 3, and 5 µl bubbles displayed the largest changes in the contact angle. The result demonstrates that an electrical signal at 23 kHz can be used to separate air bubbles of various volumes from a substrate through the mechanism of dielectrowetting.

To analyze the movement and removal of air bubbles separated from a substrate by dielectrowetting using acoustic excitation, the piezoactuator (diameter: 30 mm and height: 3 mm, KPR3020LC-450, Daeyoung electric Co.) was mounted on the side of the tank in the same position as the substrate, as shown in Fig. 3. Figure 7 presents the experimental result of the proposed bubble removal method using a single bubble. The dielectrowetting chip was positioned inside the tank, and the micropipette was used to generate bubbles of 5 µl at the bottom of the substrate and position them, as shown in Fig. 7(a). Figure 7(b) shows the result of separating a single bubble of 5 µl from the substrate surface by the mechanism of dielectrowetting by applying 350 V (at 23 kHz) to the patterned electrodes. Figures 7(c) and 7(d) present the results of transporting and removing the bubble excited by acoustic waves that were generated by applying a voltage at 2.4 kHz to the piezoactuator. As the bubble separated from the surface by dielectrowetting was excited by acoustic waves generated by the piezoactuator mounted on the side of the tank, the oscillation and microstreaming that occurred at the same time caused the bubble to move and get removed from the substrate. This result confirms the effectiveness of the bubble removal method proposed in this study. Note that the reference for the oscillating behavior and microstreaming of an acoustically excited bubble can be found.³⁶ 

Figure 8 shows the velocity of bubbles of 1, 3, and 5 µl according to the frequencies of the voltages applied to the piezoactuator to identify the effect of acoustic excitation on bubbles. The highest velocity of the moving bubbles occurred at the resonant frequency of the bubbles [$fv=(2π)−13γp0/ρR2$⁠, where γ is the ratio of the specific heats of air, R is the radius of the bubble, p₀is the hydrostatic liquid pressure, ρ is the liquid density^37–39], and the resonant frequencies of 1, 3, and 5 µl bubbles were 3.6, 3.2, and 2.4 kHz, respectively. The result demonstrated that the resonant frequencies should be used to remove bubbles separated from a surface by acoustic excitation as quickly as possible.

Finally, the results of the experiments conducted earlier prove that air bubbles of various volumes could be removed simultaneously. As shown in Fig. 9(b1), a micropipette was used to create air bubbles of various volumes at the bottom of the substrate, which obstruct the transfer of heat across the substrate surface. To separate the bubbles of various volumes from the substrate surface by the mechanism of dielectrowetting, a voltage (350 V at 23 kHz) was applied to the patterned electrodes, while at the same time, the piezoactuator mounted on the tank was used to transport and remove bubbles by acoustic excitation in Fig. 9(b2). Consequently, as shown in Fig. 9(b3), all the bubbles that obstructed the transfer of heat on the substrate surface were removed in a matter of a few seconds. The results of this experiment show that the proposed method can quickly remove bubbles of various volumes from the substrate.

This paper proposed a new bubble removal method aimed at enhancing heat transfer by applying the two mechanisms of dielectrowetting and acoustic excitation. Before bubbles were removed from the substrate surface, electrodes had first been patterned through the process of microfabrication, so the mechanism of dielectrowetting could be applied. Then, a dielectrowetting chip composed of dielectric and hydrophobic layers was used in the experiment. To investigate the behavior of the bubbles affected by the mechanism of dielectrowetting, quantitative experiments were conducted at various voltages and frequencies using the dielectrowetting chip. This study demonstrated that the contact angle becomes smaller as the voltage becomes higher; that when a voltage of 260 V was applied, the bubbles were separated from the surface; and that the separation angle varies in proportion to the applied voltage. To investigate the movements of the separated bubbles through acoustic excitation, quantitative experiments were conducted on the moving velocity of the bubbles, which varies depending on the frequency applied to the piezoactuator on the side of the tank. The experiment confirmed that the bubbles moved at the highest velocity at resonant frequencies. The experiment for removing bubbles of various volumes on the substrate surface was successfully carried out. Based on the experiment, the proposed method can be used in a variety of industrial applications that require the transfer of large quantities of heat.

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea Government (MSIT) (Grant No. 2020R1F1A1074888) and the Korea Institute for Advancement of Technology (KIAT) grant funded by the Korea Government (MOTIE) (Grant No. P0008458, The Competency Development Program for Specialist).

Y. Hyun and K. Y. Lee contributed equally to this work.

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