Core–Shell Droplet Generation Device Using a Flexural Bolt-Clamped Langevin-Type Ultrasonic Transducer

Droplets with a core–shell structure formed from two immiscible liquids are used in various industrial field owing to their useful physical and chemical characteristics. Efficient generation of uniform core–shell droplets plays an important role in terms of productivity. In this study, monodisperse core-shell droplets were efficiently generated using a flexural bolt-clamped Langevin-type transducer and two micropore plates. Water and silicone oil were used as core and shell phases, respectively, to form core–shell droplets in air. When the applied pressure of the core phase, the applied pressure of the shell phase, and the vibration velocity in the micropore were 200 kPa, 150 kPa, and 8.2 mm/s, respectively, the average diameter and coefficient of variation of the droplets were 207.7 μm and 1.6%, respectively. A production rate of 29,000 core–shell droplets per second was achieved. This result shows that the developed device is effective for generating monodisperse core–shell droplets.


Introduction
Droplets with core-shell structures formed from two immiscible liquids have useful physical and chemical characteristics, such as protection of the core phase by the shell phase, simultaneous use of two substances, and improved shell phase reactivity using the core phase as a catalyst. Therefore, core-shell droplets are commonly used in industrial fields such as medicine, cosmetics, and food. Many studies using core-shell droplets have been reported [1][2][3][4][5], wherein it is necessary to manage each droplet that is generated. Thus, the efficient generation of uniform core-shell droplets has an important role in terms of productivity, and is a topic of considerable research interest [6][7][8][9][10][11][12][13][14][15][16][17].
A large number of core-shell droplets can be simultaneously generated by membrane emulsification [6]. This method is categorized bulk method. However, this method cannot control the generation of each droplet, and two-step emulsification is required.
A microfluidic device composed of glass capillaries is a typical device used for generating core-shell drop-lets [7][8][9][10][11]. The droplets are generated individually in the microchannel, achieving a production rate of thousands of droplets per second.
Core-shell droplets have a complex structure and are more difficult to generate than single droplets. Thus, an efficient core-shell droplet generation method is required. Monodisperse single droplets have been efficiently generated using vibration-based methods [18][19][20][21][22]. It is possible to generate tens of thousands of droplets per second, which is the production rate of droplets required for core-shell droplet generation with high efficiency. Among the flow methods that generate droplets one by one, the vibration-based method has high generation efficiency.
In the present study, monodisperse core-shell droplets were generated by a vibration method. A core-shell droplet generation device using a bolt-clamped Langevin-type flexural transducer and two micropore plates was designed.

Droplets Generation Principle
The principle behind the droplet generation in this research is surface tension. Fluid is converted into droplets by surface tension. This method used relies on electric power to turn fluid into droplets with controlled size. In addition, production rate of droplets is controlled using vibration [18][19][20][21][22].
First, the supplied fluid is ejected from the micropore to air. The ejected jet flow breaks under the effect of surface tension and turns into droplets. When vibration is not applied, uniform droplets are not generated. Conversely, when vibration is applied, uniform droplets are generated. Vibration is applied vertically to the flow direction of the jet flow. Uniform waves with the same period as the vibration are generated on the surface of the jet flow. The driven frequency is equal to the number of droplets.
A double-structure jet flow is necessary for core-shell droplet generation. This doublestructure jet is a coaxial flow in which the core phase is covered by a shell phase, thereby generating core-shell droplets as they exit the device. Figures 1 and 2 show the schematic and photograph of the core-shell droplet generation device, respectively. The length of the device is 42.7 mm and the diameter is 18 mm. A bolt-clamped Langevin-type ultrasonic transducer was used to obtain flexural vibration. Piezoelectric elements (PZT) polarized in the thickness direction are sandwiched between metal blocks.

Droplets Generation Principle
The principle behind the droplet generation in this research is surface tension. Fluid is converted into droplets by surface tension. This method used relies on electric power to turn fluid into droplets with controlled size. In addition, production rate of droplets is controlled using vibration [18][19][20][21][22].
First, the supplied fluid is ejected from the micropore to air. The ejected jet flow breaks under the effect of surface tension and turns into droplets. When vibration is not applied, uniform droplets are not generated. Conversely, when vibration is applied, uniform droplets are generated. Vibration is applied vertically to the flow direction of the jet flow. Uniform waves with the same period as the vibration are generated on the surface of the jet flow. The driven frequency is equal to the number of droplets.
A double-structure jet flow is necessary for core-shell droplet generation. This double-structure jet is a coaxial flow in which the core phase is covered by a shell phase, thereby generating core-shell droplets as they exit the device. Figures 1 and 2 show the schematic and photograph of the core-shell droplet generation device, respectively. The length of the device is 42.7 mm and the diameter is 18 mm. A bolt-clamped Langevin-type ultrasonic transducer was used to obtain flexural vibration. Piezoelectric elements (PZT) polarized in the thickness direction are sandwiched between metal blocks.

Structure
To discharge a double-structure jet flow, the device has a channel part at the tip of the device.
As shown in Figure 3, the channel part is assembled to three plates: an upper plate, an intermediate plate, and a lower plate. The intermediate plate is sandwiched between the upper and lower plates to provide a shell phase flow path. The upper and lower micropore diameters are 50 μm and 100 μm, respectively. The schematic of the upper micropore plate for the core-shell droplet generation device is shown in Figure 4. The micropore plate is 6 mm in diameter and 0.5 mm thick. A schematic for the lower micropore plate for the core-shell droplet generation device is shown in Figure 5. The micropore plate is 12 mm in diameter and 0.5 mm thick. Figure 6 shows the cross section of the device. The core phase supplied from the top of the device is injected into the shell phase supplied from the side to generate the core-shell two phase flow.    To discharge a double-structure jet flow, the device has a channel part at the tip of the device.
As shown in Figure 3, the channel part is assembled to three plates: an upper plate, an intermediate plate, and a lower plate. The intermediate plate is sandwiched between the upper and lower plates to provide a shell phase flow path. The upper and lower micropore diameters are 50 µm and 100 µm, respectively. The schematic of the upper micropore plate for the core-shell droplet generation device is shown in Figure 4. The micropore plate is 6 mm in diameter and 0.5 mm thick. A schematic for the lower micropore plate for the core-shell droplet generation device is shown in Figure 5. The micropore plate is 12 mm in diameter and 0.5 mm thick. Figure 6 shows the cross section of the device. The core phase supplied from the top of the device is injected into the shell phase supplied from the side to generate the core-shell two phase flow.

Vibration Characteristics
This device was designed using finite element method analysis, as shown in Figure  7. The device was subjected to a first-order flexural vibration mode. The fixed part and the vibration node are at the same position; this design prevents vibration damping.
For alignment of the center of the upper and lower micropore plates, the micropores were placed at the center of each micropore plate. The flexural vibration is effective for oscillating the center of the micropore plate. The relationship between frequency, admittance, and phase is shown in Figure 8. This device was designed to provide first-order flexural vibration at 30 kHz. The resonance frequency of the fabricated device was observed using a laser Doppler vibrometer at a frequency of 29 kHz. Figure 9 shows the relationship between distance from the tip of the core-shell droplet generation device and vibration velocity when the driving frequency and the voltage were 29 kHz and 100 Vp-p, respectively. Vibration nodes were observed at 9 and 33 mm; a vibrational antinode was observed at 21 mm, which corresponds to the center of the device. The results show that the device oscillated under a first-order flexural vibration mode. Figure 10 shows the relationship between the applied voltage and the vibration velocity at the tip of the core-shell droplet generation device when the driving frequency was 29 kHz.

Vibration Characteristics
This device was designed using finite element method analysis, as shown in Figure 7. The device was subjected to a first-order flexural vibration mode. The fixed part and the vibration node are at the same position; this design prevents vibration damping.
Actuators 2021, 10, x FOR PEER REVIEW 4 of 10 Figure 6. Fluid supply route for the core-shell droplet generation device under vibration mode.

Vibration Characteristics
This device was designed using finite element method analysis, as shown in Figure  7. The device was subjected to a first-order flexural vibration mode. The fixed part and the vibration node are at the same position; this design prevents vibration damping.
For alignment of the center of the upper and lower micropore plates, the micropores were placed at the center of each micropore plate. The flexural vibration is effective for oscillating the center of the micropore plate. The relationship between frequency, admittance, and phase is shown in Figure 8. This device was designed to provide first-order flexural vibration at 30 kHz. The resonance frequency of the fabricated device was observed using a laser Doppler vibrometer at a frequency of 29 kHz. Figure 9 shows the relationship between distance from the tip of the core-shell droplet generation device and vibration velocity when the driving frequency and the voltage were 29 kHz and 100 Vp-p, respectively. Vibration nodes were observed at 9 and 33 mm; a vibrational antinode was observed at 21 mm, which corresponds to the center of the device. The results show that the device oscillated under a first-order flexural vibration mode. Figure 10 shows the relationship between the applied voltage and the vibration velocity at the tip of the core-shell droplet generation device when the driving frequency was 29 kHz. For alignment of the center of the upper and lower micropore plates, the micropores were placed at the center of each micropore plate. The flexural vibration is effective for oscillating the center of the micropore plate.
The relationship between frequency, admittance, and phase is shown in Figure 8. This device was designed to provide first-order flexural vibration at 30 kHz. The resonance frequency of the fabricated device was observed using a laser Doppler vibrometer at a frequency of 29 kHz.
Actuators 2021, 10, x FOR PEER REVIEW Figure 8. Relationship between frequency, admittance, and phase of the core-shell droplet tion device.  Figure 9 shows the relationship between distance from the tip of the core-shell droplet generation device and vibration velocity when the driving frequency and the voltage were 29 kHz and 100 V p-p , respectively. Vibration nodes were observed at 9 and 33 mm; a vibrational antinode was observed at 21 mm, which corresponds to the center of the device. The results show that the device oscillated under a first-order flexural vibration mode.    Figure 11 shows the core-shell droplet generation experiment system. Core-shell droplets were generated in air when water and silicone oil (1 mm 2 /s) were used as the core and shell phases, respectively. Core and shell phase fluids were supplied separately by applying pressure with a compressor and regulators. The vibration speed was controlled by the voltage and drive frequency applied to the piezoelectric element using a function generator and a high-speed bipolar power supply. The generated droplets were observed with a high-speed camera. The droplet diameter was measured using the image analysis      Figure 11 shows the core-shell droplet generation experiment system. Co droplets were generated in air when water and silicone oil (1 mm 2 /s) were used as and shell phases, respectively. Core and shell phase fluids were supplied separ applying pressure with a compressor and regulators. The vibration speed was co by the voltage and drive frequency applied to the piezoelectric element using a f generator and a high-speed bipolar power supply. The generated droplets were o with a high-speed camera. The droplet diameter was measured using the image  Figure 11 shows the core-shell droplet generation experiment system. Core-shell droplets were generated in air when water and silicone oil (1 mm 2 /s) were used as the core and shell phases, respectively. Core and shell phase fluids were supplied separately by applying pressure with a compressor and regulators. The vibration speed was controlled by the voltage and drive frequency applied to the piezoelectric element using a function generator and a high-speed bipolar power supply. The generated droplets were observed with a high-speed camera. The droplet diameter was measured using the image analysis software "WinROOF" (MITANI Corporation, Japan). One hundred droplets were measured for each type of experimental data. software "WinROOF" (MITANI Corporation, Japan). One hundred droplets wer ured for each type of experimental data.

System Configuration
In this research, monodispersity was assessed using the coefficient of v When the coefficient of variation was 5% or less, it was determined that a uniform was generated. Figure 11. Core-shell droplet generation system.

Experiments with Varying Voltage
Core-shell droplets were generated when the applied voltage was changed. plied pressures of the core and shell phases were 200 kPa and 150 kPa, respectiv driving frequency was 29 kHz. Figure 12 shows photographs of the generated core-shell droplets in the air w applied voltages were 0, 6, 7, and 100 Vp-p. The vibration velocities of the micropo 0, 0.12, 0.13, and 8.2 mm/s. Figure 13 shows the relationship between vibration and the diameter of droplet. Table 1 shows the average droplet diameter when the voltage was changed. Figure 13 and Table 1 shows that monodisperse droplets w erated when the applied voltage was larger than 7 Vp-p.  In this research, monodispersity was assessed using the coefficient of variation. When the coefficient of variation was 5% or less, it was determined that a uniform droplet was generated.

Experiments with Varying Voltage
Core-shell droplets were generated when the applied voltage was changed. The applied pressures of the core and shell phases were 200 kPa and 150 kPa, respectively. The driving frequency was 29 kHz. Figure 12 shows photographs of the generated core-shell droplets in the air when the applied voltages were 0, 6, 7, and 100 V p-p . The vibration velocities of the micropore were 0, 0.12, 0.13, and 8.2 mm/s. Figure 13 shows the relationship between vibration velocity and the diameter of droplet. Table 1 shows the average droplet diameter when the applied voltage was changed. Figure 13 and Table 1 shows that monodisperse droplets were generated when the applied voltage was larger than 7 V p-p .
software "WinROOF" (MITANI Corporation, Japan). One hundred droplets wer ured for each type of experimental data.
In this research, monodispersity was assessed using the coefficient of v When the coefficient of variation was 5% or less, it was determined that a uniform was generated. Figure 11. Core-shell droplet generation system.

Experiments with Varying Voltage
Core-shell droplets were generated when the applied voltage was changed. plied pressures of the core and shell phases were 200 kPa and 150 kPa, respectiv driving frequency was 29 kHz. Figure 12 shows photographs of the generated core-shell droplets in the air w applied voltages were 0, 6, 7, and 100 Vp-p. The vibration velocities of the micropo 0, 0.12, 0.13, and 8.2 mm/s. Figure 13 shows the relationship between vibration and the diameter of droplet. Table 1 shows the average droplet diameter when the voltage was changed. Figure 13 and Table 1 shows that monodisperse droplets w erated when the applied voltage was larger than 7 Vp-p.

Experiments with Varying Pressure
Core-shell droplets were generated by changing the applied pressure. The a voltage was 100 Vp-p and the drive frequency was 29 kHz. Figure 14 shows photographs of the generated core-shell droplets in air wh applied pressure for the core phase was changed. Figure 15 shows the relationsh tween vibration velocity and droplet diameter. Table 2 shows average droplet dia when the applied pressure for the core phase was changed. According to Table 2 m disperse droplets were generated when the applied pressure was between 170 kP 320 kPa. Figure 14. Photographs of generated core-shell droplets when the applied pressure was cha

Experiments with Varying Pressure
Core-shell droplets were generated by changing the applied pressure. The applied voltage was 100 V p-p and the drive frequency was 29 kHz. Figure 14 shows photographs of the generated core-shell droplets in air when the applied pressure for the core phase was changed. Figure 15 shows the relationship between vibration velocity and droplet diameter. Table 2 shows average droplet diameter when the applied pressure for the core phase was changed. According to Table 2 monodisperse droplets were generated when the applied pressure was between 170 kPa and 320 kPa. Figure 13. Relationship between vibration speed and droplet diameter when the driv was 29 kHz and the applied pressure of core and shell phases were 200 kPa and 150

Experiments with Varying Pressure
Core-shell droplets were generated by changing the applied pressure voltage was 100 Vp-p and the drive frequency was 29 kHz. Figure 14 shows photographs of the generated core-shell droplets in applied pressure for the core phase was changed. Figure 15 shows the rel tween vibration velocity and droplet diameter. Table 2 shows average drop when the applied pressure for the core phase was changed. According to T disperse droplets were generated when the applied pressure was between 320 kPa.

Discussion
As shown in the experimental results, monodisperse core-shell droplets wer ated by adjusting the applied voltage and the applied pressure. The voltage has a old value at which core-shell droplets can be generated. Because the applied vo almost proportional to the vibration velocity, the droplet generation requires a v velocity higher than the specified value.
There is a range of applied pressures at which core-shell droplets can be gen The lower limit of the applied pressure is determined by the difference between plied pressure of the core and shell phases ∆P. It is expressed by Equation (1) where γ is the interfacial tension between the core and shell phases, Rc is the sp radius of the discharged core phase, and Ro is the radius of the lower micropore. W is smaller than Ro, monodisperse droplets can be generated. In these experiments 41.6 mN/m and Ro was 0.05 mm. Therefore, ∆P was 16.6 kPa. When the applied p of the core phase was 160 kPa (∆P = 10 kPa), generation of monodisperse dropl impossible. When the applied pressure of the core phase was 170 kPa (∆P = 20 kPa odisperse droplets could be generated.
The upper limit of the applied pressure is determined by the core-shell pha velocity vcs and the core phase flow velocity vc. The core-shell phase flow veloci flow velocity of the coaxial flow composed of the core and shell phases discharg the lower micropore. The core phase flow velocity is the flow velocity of the cor discharged from the upper micropore. When vcs > vc, droplet generation is pos Figure 16, the graph shows the relationship between the applied pressure of t phase and the flow velocity. When the applied pressure of the core phase was 320 was 15.3 m/s, and vc was 14.4 m/s, monodisperse droplets could be generated. W applied pressure of the core phase was 330 kPa, vcs was 15.3 m/s, and vc was 1 Figure 15. Relationship between applied pressure of the core phase and droplet diameter when the driving frequency was 29 kHz and the vibration velocity was 8.2 mm/s.

Discussion
As shown in the experimental results, monodisperse core-shell droplets were generated by adjusting the applied voltage and the applied pressure. The voltage has a threshold value at which core-shell droplets can be generated. Because the applied voltage is almost proportional to the vibration velocity, the droplet generation requires a vibration velocity higher than the specified value.
There is a range of applied pressures at which core-shell droplets can be generated. The lower limit of the applied pressure is determined by the difference between the applied pressure of the core and shell phases ∆P. It is expressed by Equation (1) where γ is the interfacial tension between the core and shell phases, R c is the spherical radius of the discharged core phase, and R o is the radius of the lower micropore. When R c is smaller than R o, monodisperse droplets can be generated. In these experiments, γ was 41.6 mN/m and R o was 0.05 mm. Therefore, ∆P was 16.6 kPa. When the applied pressure of the core phase was 160 kPa (∆P = 10 kPa), generation of monodisperse droplets was impossible. When the applied pressure of the core phase was 170 kPa (∆P = 20 kPa), monodisperse droplets could be generated. The upper limit of the applied pressure is determined by the core-shell phase flow velocity v cs and the core phase flow velocity v c . The core-shell phase flow velocity is the flow velocity of the coaxial flow composed of the core and shell phases discharged from the lower micropore. The core phase flow velocity is the flow velocity of the core phase discharged from the upper micropore. When v cs > v c , droplet generation is possible. In Figure 16, the graph shows the relationship between the applied pressure of the core phase and the flow velocity. When the applied pressure of the core phase was 320 kPa, v cs was 15.3 m/s, and v c was 14.4 m/s, monodisperse droplets could be generated. When the applied pressure of the core phase was 330 kPa, v cs was 15.3 m/s, and v c was 15.3 m/s, generation of monodisperse droplets was impossible.
Actuators 2021, 10, x FOR PEER REVIEW Figure 16. Relationship between applied pressure for core phase and flow velocity when th plied pressure for shell phase was 150 kPa.
Core-shell droplets were collected by ejecting water into a beaker. Figure 17 photographs of the collected core-shell droplets observed with an optical microsco water used in the core phase was colored blue. The water in the beaker was colo ange. Compared to the total number of generated droplets, fewer droplets were ob because many of them collapsed under the high flow velocity when they hit th surface. Figure 17. Relationship between applied pressure for core phase and flow velocity when th plied pressure for shell phase was 150 kPa.

Conclusions
Core-shell droplets were generated using flexural vibration. A core-shell generation device was designed using a bolt-clamped Langevin-type flexural tran and two micropore plates. The droplets were generated in air using water and silic A generation rate of 29,000 core-shell droplets per second was achieved. Monod core-shell droplets were generated by adjusting the applied voltage and the applie sure. The conditions of core-shell droplet generation were evaluated. We have suc in generating highly efficient and monodisperse core-shell droplets by using th sonic transducer. Core-shell droplets were collected by ejecting water into a beaker. Figure 17 shows photographs of the collected core-shell droplets observed with an optical microscope. The water used in the core phase was colored blue. The water in the beaker was colored orange. Compared to the total number of generated droplets, fewer droplets were observed because many of them collapsed under the high flow velocity when they hit the water surface. Figure 16. Relationship between applied pressure for core phase and flow velocity plied pressure for shell phase was 150 kPa.
Core-shell droplets were collected by ejecting water into a beaker. Fi photographs of the collected core-shell droplets observed with an optical m water used in the core phase was colored blue. The water in the beaker w ange. Compared to the total number of generated droplets, fewer droplets because many of them collapsed under the high flow velocity when they surface. Figure 17. Relationship between applied pressure for core phase and flow velocity plied pressure for shell phase was 150 kPa.

Conclusions
Core-shell droplets were generated using flexural vibration. A core generation device was designed using a bolt-clamped Langevin-type flexu and two micropore plates. The droplets were generated in air using water a A generation rate of 29,000 core-shell droplets per second was achieved. core-shell droplets were generated by adjusting the applied voltage and th sure. The conditions of core-shell droplet generation were evaluated. We h in generating highly efficient and monodisperse core-shell droplets by u sonic transducer.

Conclusions
Core-shell droplets were generated using flexural vibration. A core-shell droplet generation device was designed using a bolt-clamped Langevin-type flexural transducer and two micropore plates. The droplets were generated in air using water and silicone oil. A generation rate of 29,000 core-shell droplets per second was achieved. Monodisperse core-shell droplets were generated by adjusting the applied voltage and the applied pressure. The conditions of core-shell droplet generation were evaluated. We have succeeded in generating highly efficient and monodisperse core-shell droplets by using the ultrasonic transducer. Funding: This research was partially supported by a grant for the Promotion of Science and Technology in Okayama Prefecture by NEXT.

Conflicts of Interest:
The authors declare no conflict of interest.