A Monolithic Electrostatic–Hydraulic Coupled Suction Pad

Suction cups are one of the most widely adopted switchable adhesive techniques and play a crucial role in manufacturing and robotics. However, conventional negative pressure sources (e.g., diaphragm vacuum pumps) suffer from poor scalability from meso‐ to microscale, unavoidable vibration and noise, while the performances of existing smart material‐driven suction cups are often restricted by their bulky structures, poor active adhesive stresses, or high power consumptions. To address these limitations, a monolithic suction pad is proposed where its active adhesion is realized by the electrostatic–hydraulic coupling generated negative pressure cavity and the sealing is achieved by the elastomeric–hydraulic coupled outer rim. This suction pad features a peak pull‐off stress of 7.4 kPa and good switch ability in its adhesive stress of over 12 times. Compared with the existing smart material‐driven suction cups, the proposed suction pad demonstrates clear advantages in its significantly lower profile structure, lower preload stress requirement, and a higher adhesive force to power consumption ratio. This suction pad is envisioned to have promises in broad applications such as digital manufacturing, transfer printing, medical robotics, and robotic locomotion.

DOI: 10.1002/aisy.202200425 Suction cups are one of the most widely adopted switchable adhesive techniques and play a crucial role in manufacturing and robotics. However, conventional negative pressure sources (e.g., diaphragm vacuum pumps) suffer from poor scalability from meso-to microscale, unavoidable vibration and noise, while the performances of existing smart material-driven suction cups are often restricted by their bulky structures, poor active adhesive stresses, or high power consumptions. To address these limitations, a monolithic suction pad is proposed where its active adhesion is realized by the electrostatic-hydraulic coupling generated negative pressure cavity and the sealing is achieved by the elastomeric-hydraulic coupled outer rim. This suction pad features a peak pull-off stress of 7.4 kPa and good switch ability in its adhesive stress of over 12 times. Compared with the existing smart material-driven suction cups, the proposed suction pad demonstrates clear advantages in its significantly lower profile structure, lower preload stress requirement, and a higher adhesive force to power consumption ratio. This suction pad is envisioned to have promises in broad applications such as digital manufacturing, transfer printing, medical robotics, and robotic locomotion.
high energy density, and relatively large actuation force. Follador et al. developed a dome DEA-driven octopus-inspired suction cup with a 6 kPa suction stress in water. [23] Although several aforementioned solutions have been developed for air-supply-free suction cups, some limitations still remain. First, current designs rely on a simple stack of a passive suction cup and an actuator on top, the structure of which will inevitably increase the overall height of the system, making them impossible to miniaturize in the height dimension. Second, these designs often require a substantial preload (e.g., >1 N, 5 kPa) on the suction cup when engage with an object to ensure a good sealing and thus successful adhesion. However, this high preload will pose potential damage issues to fragile objects and complicate the system if a feedback control is added to the suction system. In contrast to the aforementioned smart materialdriven suction cups with stacked structures and high preload requirements, in this work we develop a monolithic suction pad which features a low-profile structure (2 mm in height), high active adhesive stress (%7.4 kPa), good adhesive switch ability (>12 times), and very low preload demands (≤0.16 kPa). A prototype of the suction pad is shown in Figure 1a. These attractive features are realized by adopting an electrostatic-hydraulic coupling actuation mechanism. The unique electromechanical design of the suction pad allows the electrostatic-hydraulic coupling actuation in its center to create a negative pressure cavity while the elastomeric-hydraulic coupled outer rim forms good sealing with the object in contact. Extensive experimental work is carried out in this work to characterize its adhesive performance. The influences of several key design and actuation parameters are also investigated in this study.
The rest of this article is organized as follows. In Section 2, design of the proposed suction pad is overviewed and its actuation principle is explained, adhesive forces of the suction pad are characterized, and the influences of key parameters are also investigated. The proposed suction pad is compared with other existing smart material-driven suction mechanisms and its advantages and limitations are discussed in Section 2. Finally, conclusion is drawn in Section 3.

Suction Pad Design
The design of the proposed suction pad is illustrated in Figure 1b. Its bottom layer is a 15 mm radius circular printed circuit board (PCB) with a 7 mm radius circular copper electrode being printed on the top side. A 1 mm diameter through-hole is included close to the copper electrode, which connects the electrode to the bottom side of the PCB substrate for easy wire connection and it also serves as the oil filling channel. On top of the PCB is a thin layer of FR-4 ring shape spacer. Together with the PCB substrate, they contain the dielectric liquid, which is then sealed by a piece of elastomer membrane with stretchable electrode attached. Finally, on top is another layer of FR-4 ring to maintain the tension in the elastomer membrane. The suction pad features a diameter of 30 mm and a height of 2 mm, which demonstrates a significantly lower height-to-diameter ratio than the current smart material-driven suction cups.
The actuation of the proposed suction pad is based on the electrostatic force to deform the elastomer membrane, hence creating a negative pressure cavity. It should be noted that relying on the local contraction in the membrane in its thickness alone (i.e., as in DEAs) can only lead to a very minor deformation, which is incapable of achieving a sufficient negative pressure cavity. To amplify the actuation deformation, we take inspiration from the hydraulically amplified self-healing electrostatic (HASEL) actuators [27] and electro-ribbon actuators [28] and adopt the electrostatic-hydraulic coupling actuation, as illustrated in Figure 1c. When an electric potential is applied across the electrodes, the electrostatic stress forces the membrane to move toward the bottom electrode, which drives the dielectric liquid to move  (c) (d) Figure 1. a) Photo of the proposed suction pad prototype next to a 1 RMB Yuan coin. b) Illustration of the proposed suction pad with its cross-sectional view (left bottom) and expanded view (right). c) Actuation principle of the proposed suction pad. The electrostatic-hydraulic coupling actuation will force the dielectric liquid out from the central region toward the outer edge to achieve negative pressure cavity. d) Laser scanned deformation of the membrane along the center line with and without actuation.
www.advancedsciencenews.com www.advintellsyst.com toward the sides. Due to the incompressibility of the dielectric liquid, the outer rim of the membrane will bulge to form a good sealing with the substrate (Figure 1c bottom). This electrostatichydraulic coupling actuation eventually forms a negative pressure cavity in the central region of the suction pad, which, therefore, generates suction force on the substrate. The fabrication process of the suction pad is illustrated in Figure 2 and is explained in detail in the Experimental Section.
To demonstrate the feasibility of the proposed electrostatichydraulic coupling actuation, the deformations of the membrane along its center line with and without actuation are measured using laser scanning and the results are plotted in Figure 1d. It can be noted that the actuation successfully caused the electrode-covered region of the membrane to deform downward by approximately 0.15 mm, which is the designed depth of the dielectric liquid reservoir, indicating that the electrostatic force pushed all dielectric liquid out from the center. Also note that the outer ends of the membrane bulged upward compared to its passive state, which will act as the sealing of the suction pad. This preliminary test confirmed that the electrostatichydraulic coupling actuation is capable of achieving the desired deformation. The adhesive forces of the suction pad will be characterized in-depth in the later sections.

Maximum Pull-Off Force Characterizations
The maximum pull-off force F max is a key parameter that describes the performance of a suction mechanism. The force characterization experimental setup is described in the Experimental Section and the measured results of the generated adhesive force of the suction pad during a maximum pull-off force test are plotted in Figure 3a. A 4 kV voltage is applied to the suction pad during 2-7 s. It can be noted that the adhesive force applied to the substrate increases approximately linearly with the increasing displacement and drops sharply to zero due to the detachment. The maximum pull-off force is %1.1 N, which leads to a maximum adhesive stress of about 7.14 kPa. To analyze the adhesion switch ability of the suction pad, Figure 3b plots the adhesive force-displacement relationships with the voltage ON and OFF. Note that, when no voltage is applied, the suction pad generated a negligible passive adhesive force of %0.08 N due to the residual tackiness of the suction pad surface. The comparison results in Figure 3b demonstrate that the electrostatic-hydraulic coupling actuation can lead to over 12 times increase in the adhesive force of the suction pad. An analytical model that describes the quasistatic active adhesive force of the proposed suction pad is developed in the Supporting Information. The simulation results in Figure S1a, Supporting Information, agree well with the experimental data reported in this subsection.
To investigate the repeatability and consistency of the suction pad, five samples were fabricated and the maximum pull-off force of each sample was measured five times. The statistic results are plotted in Figure 3c. Good consistency can be found between each sample and between each set of tests. A mean pull-off force is calculated at 1.14 N and the mean passive adhesive force is %0.1 N, which is equivalent to 7.4 and 0.65 kPa, respectively. The deviation in the results could be due to the fabrication tolerance and the slight differences in the contact condition during each set of tests.

Parameter Characterizations
The adhesive performance of the suction pad can be influenced by many parameters. In this subsection, we experimentally investigate the effects of three key parameters, such as the oil mass, actuation voltage amplitude, and pulling velocity, and the results are plotted in Figure 3d-f, respectively.
It can be noted from Figure 3d that the oil mass has a significant impact on the F max of the suction pad. At the minimum oil mass, the suction pad suffers for a very poor adhesive force. As the oil mass increases, the adhesive force reaches its optimal region before drops gradually. The measured force-oil mass relationship could be explained as follows: the low oil mass (volume) does not allow the membrane to deform sufficiently to create the required negative pressure cavity. On the other hand, a much larger oil mass (volume) separates the two electrodes, thus weakens the electrostatic force generated at the same voltage. Figure 3e plots the F max against the DC voltage amplitude. It can be noted that the force increases with the increasing voltage amplitude due to the increased electric field and hence electrostatic force. This is verified in Figure S1a, Supporting Information, using the analytical model. It is also noteworthy that a maximum voltage of 4 kV was tested because a voltage higher than this value was found to significantly increase the electric breakdown risk of the suction pad. Electric breakdown was observed to cause permanent damage on the membrane, hence should be avoided by controlling the voltage in the safe range in practical applications. Another parameter that could affect the F max is the pulling velocity. As shown in Figure 3f, a higher pulling velocity leads to an increase in both passive and active pull-off forces, which could be due to the viscoelasticity www.advancedsciencenews.com www.advintellsyst.com of the elastomer membrane and viscosity of the dielectric liquid. However, it should be noted that a higher pulling velocity does not ensure a better suction performance in practical applications because higher velocity can also lead to higher acceleration and deceleration if not properly controlled, which will affect the stability during the suction process.

Hold Time Characterizations
In some applications (e.g., pick and place tasks in manufacturing), it is usually required the suction pad to lift up and hold the object for a certain period (e.g., during the transition from one place to another in an assembly line). As a result, in this subsection we investigate the adhesive force holding capability of the suction pad. The test duration was chosen as 120 s, which is believed to be sufficient for many applications such as robotic grasping and climbing. Figure 4a plots the measured adhesive forces of the suction pad and the corresponding displacements over time at 65%, 72.5%, 80%, and 90% F max . It can be noted that in all four cases the suction pad is able to maintain its adhesion to the substrate without detachment. However, note that in 90% F max case the actual suction force decreases gradually over time, which could be due to the air leakage of suction pad. Also note that in 72.5%, 80%, and 90% F max cases the adhesion can be removed almost instantly once the voltage is switched off. In 65% F max case a second delay is observed, which could be the slow back flow of the dielectric liquid without the aid of a sufficiently large external pulling force.
In many applications, the energy consumption of the suction pad during operation is also critical. Here, we recorded the  www.advancedsciencenews.com www.advintellsyst.com voltage and current supplied to the suction pad in the hold time tests and the current data are plotted in Figure 4b. Two peaks can be noted at the instants of switching the voltage ON and OFF, which are due to the charges flow into and out of the suction pad, respectively. During the short charging period, it is calculated that approximately 13.8 mJ electric energy is consumed. Once the suction pad is fully charged, a minor current leakage of %7 μA can be detected, which corresponds to power consumption of <30 mW during the holding period.

Demonstration of the Suction Pad
To illustrate the feasibility of utilizing the suction pad in applications, a simple pickup task is developed, as shown in Figure 5.
A 100 g payload is mounted to the acrylic substrate and is placed on top of a piece of foam. A red LED is added to indicate the condition of the actuation voltage where a glowing LED means the voltage is ON. It can be noted from Figure 5 that after the suction pad is activated, it is capable of picking up the 100 g mass and can quickly release the object when the voltage is removed.

Discussions
To better demonstrate the advantages and potential limitations of the proposed suction pad, in Table 1 we compare its performance against a conventional vacuum pump-driven suction cup and the existing smart material-driven suction cups. It can be noted that, at the same mesoscale, conventional vacuum pump-driven suction cup has the highest adhesion stress. However, electric motor scales down poorly from meso-to microscales [29] and the unavoidable vibration and high noise level of the diaphragm pumps makes them unsuitable for many applications. The several smart material-driven suction cups all feature a lower noise level. However, shape-memory alloy (SMA) suffers from substantially high-power consumption while electrohydrodynamics shows the weakest adhesion stress among all. Magnetic fields, DEA, and our electrostatic-hydraulic coupling demonstrate close active adhesive stresses, yet note that the electrostatic-hydraulic coupling proposed in this work features the three advantages: 1) A significantly lower profile structure (an order of magnitude reduction in the height dimension) due to the monolithic configuration. Thanks to its much smaller form factor, the proposed suction pad features the highest adhesive stress per unit volume among these suction techniques. This allows the suction pad to better integrate to robotic manipulators or climbing robots in volume sensitive cases (e.g., medical robotics). 2) A much lower preload stress. In the other suction cup designs, a relatively large preload stress is typically required, which poses potential damage issues to fragile objects. In this design, only a gentle preload force close to the weight of the suction pad is required, which greatly reduces the damage risks and control complexity (e.g., can rely on the suction pad's weight to apply preload force instead of using complex feedback force-position controls). 3) A high adhesive stress to power consumption ratio at an approximately similar scale. Despite that the proposed suction pad has a high power consumption density due to its compact form factor; it also shows the highest adhesive stress per unit power consumption among these suction techniques. This high adhesive stresspower ratio can be advantageous in applications such as untethered climbing robots where a lower power consumption can expand its cursing duration or allow for more electronic devices. It should also be acknowledged that several limitations still exist in this suction pad design, which requires dedicated studies in the future. First, current design relies on a piece of copper tape and a FR-4 ring as the connection between the carbon powder electrode and the power supply. This piece of rigid structure stops the suction pad to form a reliable sealing if the substrate is larger than the suction pad (the FR-4 ring and the copper tape is higher than the sealing rim formed by the activated elastomer membrane). Future work should investigate more advanced flexible electrode fabrication technologies which then eliminate the demand for the copper tape and the FR-4 ring. Second, as the sealing is solely enabled by the bulge in the elastomer membrane, it only works well on smooth substrates. To expand the  Excluding the permanent magnet. b) Data measured in water environment.
www.advancedsciencenews.com www.advintellsyst.com versatility and adaptability of the suction pad, micropatterns can be introduced to the surface of the membrane, which can potentially enhance its performance on rough surface substrates. Third, it should be noted that the current suction pad prototype works mainly with objects with flat and rigid surfaces due to the stiff PCB back plate adopted in the prototype. Future work will develop flexible suction pad arrays which consist of multiple miniature suction pads supported by flexible back plates to improve the adaptability to curved or irregular surfaces. Finally, the current prototype requires a high actuation voltage up to 4 kV, which unavoidably increases the design complexity in circuit insulation and safety protection. The actuation voltage can be reduced by: 1) further reducing the thicknesses of the elastomer and oil layer, 2) selecting elastomer materials and transformer oils with higher permittivity (as demonstrated theoretically in Figure S1b, Supporting Information, using the analytical model in the Supporting Information), or 3) blending transformer oil with highly polarizable fillers to improve its permittivity.

Conclusions
In summary, we presented a novel monolithic suction pad that achieves an excellent active adhesion by exploiting the electrostatic-hydraulic coupling mechanism. The active adhesion is realized by the negative pressure cavity from the electrostatic-hydraulic coupling actuation and the sealing achieved by the elastomeric-hydraulic coupled outer rim. Extensive experimental studies were conducted to characterize its performances and a peak pull-off stress of 7.4 kPa and good switch ability of its adhesive stress of over 12 times were reported. The proposed suction pad demonstrated clear advantages over other smart material-driven suction cups in its much lower profile structure, lower preload stress requirement, and higher adhesive force to power consumption ratio. Our suction pad is envisioned to have promises in broad applications such as industrial manufacturing, medical robotics, climbing robotics, and robotic manipulations.

Experimental Section
Fabrication of the Suction Pad: A piece of VHB (4905, 3M, USA) was adopted as the elastomer membrane for its good electrical properties, relatively low stiffness, and good compatibility with many types of dielectric liquids. The VHB membrane was first prestretched equal biaxially by 3 times before bonded to a rigid support frame.
A paper mask with the designed electrode pattern cut was covered on the membrane and carbon black powder (Ketjenblack EC-300J, Nouryon, The Netherlands) was applied to the exposed area of the membrane via a paint brush.
A 0.2 mm thickness FR-4 ring was cut by the laser cutter (Universal Laser VLS2.30, USA) and then attached to the membrane. A piece of copper tape was attached to the ring for connecting the carbon powder electrode and power supply cable.
A thin layer of Eco-flex 30 (Smooth-on, USA) was spin-coated on top of the membrane at 1500 rpm for 30 s to protect the electrode.
A custom PCB substrate and a piece of FR-4 ring were bonded together by TPU adhesives and by hot-press at 200°C for 20 s.
The base fabricated in step (5) was bonded to the underside of the membrane before the whole device was removed from the support frame.
Transformer oil (#25, SINOPEC, China) was adopted as the dielectric liquid in this study for its low viscosity and compatibility with the VHB material. Approximately 75 mg oil was injected into the suction pad through a filling channel on the PCB substrate and then sealed by tape.
Note that, to minimize any tackiness on suction pad surface, a thin layer of talcum powder was applied to the surface after the prototype was fabricated. The fabricated device features a 30 mm outer diameter, a 2 mm thickness, and weighs 2.8 g with oil filled.
Maximum Pull-Off Force Test: To characterize the maximum pull-off force of the suction pad, it was mounted to a linear rail (X-LSQ150B-E01, ZABER, Canada) with a load cell to measure the generated force (S/N 835827, FUTEK, USA) and a laser displacement sensor (LK-G152 and LKGD500, Keyence, Japan) to measure the displacement of the linear rail. A 20 mm diameter acrylic disk substrate was chosen as the target substrate, which was fixed to a spring platform. The suction pad was first lowered to make contact with the acrylic substrate and the preload force was controlled at 0.05 N by a fine tuning of the linear rail. A direct current (DC) voltage of 4 kV was applied to the suction pad by a high voltage amplifier (10/40A-HS, TREK, USA). The linear rail then drove the suction pad to lift the substrate up at a constant velocity of 1 mm s À1 , which also stretched the spring platform to yield an increasing reaction force until the substrate detached from the suction pad. The force and displacement data were collected by the data acquisition device (USB-6303, National Instruments, USA) at a sampling frequency of 10 kHz and recorded by a custom script using MATLAB (MathWorks, USA).
Hold Time Test: This test aims to characterize the hold capability of the suction pad at different suction force levels. This test followed the same initial steps as in the maximum pull-off force test, where the actuated suction pad lifted the substrate and stretched the springs. The linear rail was stopped at a target position to ensure the suction force was maintained at 65%, 72.5%, 80%, and 90% of the maximum pull-off force. The position was maintained for 120 s and the force data were recorded.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.