3D‐Printed Electrohydrodynamic Pump and Development of Anti‐Swelling Organohydrogel for Soft Robotics

Abstract

This study introduces advancements in electrohydrodynamic (EHD) pumps and the development of a 3D‐printable anti‐swelling organohydrogel for soft robotics. Using digital light processing (DLP)technology, precise components with less than 1% size variation are fabricated, enabling a unique manifold pump array. This design achieves an output pressure of 90.2 kPa—18 times higher than traditional configurations—and a flow rate of 800 mL min⁻¹, surpassing previous EHD pumps. To address swelling issues in dielectric liquids, a novel organohydrogel is developed with Young's modulus of 0.33 MPa, 300% stretchability, and a swelling ratio under 10%. Its low swelling is attributed to the shield effect and edge length confinement effect. This durable material ensures consistent pump performance under mechanical stresses like bending and twisting, crucial for dynamic soft robotic environments. These innovations significantly improve EHD pump efficiency and reliability, expanding their potential applications in soft robotics, bioengineering, and vertical farming.

This study advances electrohydrodynamic (EHD) pumps by introducing a 3D‐printable anti‐swelling organohydrogel for soft robotics. Using digital light processing (DLP) technology, precise components are fabricated, resulting in a pump with 90.2 kPa pressure and 800 mL min⁻¹ flow rate. A novel organohydrogel with low swelling and excellent mechanical properties enhances pump performance, supporting diverse soft robotics applications.

1. Introduction

Soft robots have been an exciting area of research and development in robotics.^{[ 1} , ² , ³ , ^{4 ]} They have the potential to revolutionize many areas of industry and healthcare due to their ability to operate more safely and effectively in environments where traditional robots are limited. To drive soft robotics, different types of actuators were developed, including pneumatic/hydraulic actuators,^{[ 5} , ⁶ , ^{7 ]} tendon‐driven actuators,^{[ 8} , ^{9 ]} ionic–metal‐polymer composites (IMPC) based actuators,^{[ 10} , ¹¹ , ^{12 ]} humidity stimulated actuators,^{[ 13} , ^{14 ]} solvent/salt stimulated actuators,^{[ 15} , ^{16 ]} magnetic composite‐based actuators,^{[ 17} , ^{18 ]} thermal stimulated actuators,^{[ 19 ]} and dielectric elastomer actuators (DEA).^{[ 20 ]} Compared to the other types of actuators, pneumatic/hydraulic actuators typically offer high blocking force and rapid response. Nevertheless, the performance of pneumatic/hydraulic actuators is governed by a conventional external pump, which can be challenging to integrate into soft robots due to their weight, rigidity, and size. To expand the utilization of pneumatic/hydraulic actuators, it is essential to develop small, light, soft, high‐pressure, and high‐flow rate pumps. Peristaltic pumps of various designs are created to pump the liquid and create a continuous flow.^{[ 20} , ²¹ , ²² , ^{23 ]} However, a peristaltic pump has a complicated structure, and it requires an expensive, bulky multi‐channel power supply. The diaphragm pump, which is based on the alternate volume changes stimulated by the diaphragm, is another type of developed soft pump.^{[ 24 ]} The diaphragm pump is not an ideal candidate for powering soft robotics due to its low output pressure, which is below 3 kPa. On the other hand, a magnetohydrodynamic soft pump has been developed based on electromagnetic force to push and pull the permanent magnet to create a liquid flow.^{[ 25 ]} The maximum pressure of this pump is only 8 kPa, which is still insufficient for certain soft robotic applications such as grippers and crawling robotics.

Electrohydrodynamic (EHD) pumps show advantages in being soft, light, fast responding, and having a simple structure. Because there are no moving parts, the pumps have a low failure rate and make no noise. However, it must confront two major challenges: inadequate electro‐fluid dynamic response and swelling of the polymeric pump shell caused by the dielectric liquid.

The first drawback of soft EHD pumps is their inferior electro‐fluidic performance, particularly concerning output pressure and flow rate. The first reported soft EHD pump was based on the printed planar interdigitated electrodes with carbon and silver conductive ink.^{[ 26 ]} Due to the low efficiency of planar electrodes, the maximum pressure that could be reached was 15 kPa, and the flow rate was only 6 µL min⁻¹. An EHD pump made of two parallel copper wires wound around the interior of a thermoplastic polymer tube has been reported.^{[ 27 ]} It has the typical limitation of the low flow rate of a planar electrodes‐based EHD pump (50 mL min⁻¹). To increase the flow rate, a soft EHD pump based on 3D electrodes was created.^{[ 28 ]} Although the flow rate could increase to 500 mL min⁻¹, the maximum pressure was only 5 kPa and by attempting to elevate the output pressure through the serial connection of four pumps, the maximum achieved output pressure was still only 6 kPa due to inadequate consistency.

To improve the electro‐fluidic performance of EHD pumps, especially for use in soft robotics, we developed a new 3D electrodes‐based EHD pump with serially connected units. The series connection of pump units is essential for incrementally boosting both pressure and flow rate, thereby significantly enhancing the electro‐fluid performance of EHD. For the serially connection of pump units, the fundamental prerequisite lies in the high uniformity electro‐fluidic performance across all constituent units. The electro‐fluidic performance is contingent upon two influential factors: electrode dimensional precision and assembly accuracy. We utilized high‐precision 3D printing technique, specifically digital light processing (DLP), to fabricate electrode substrates in needle and ring configuration, as well as a concentrator that aids in assembly precision. Subsequently, the electrode substrates are coated with a thin copper film to ensure conductivity. By employing DLP printing, the standard deviation of the sizes of all components has been upheld to within ≈1%. Coupled with the meticulous assembly precision of the concentrator, it has facilitated the attainment of uniformly high electro‐fluidic performance (the output pressure variations are ≈3%) in the EHD pump units, thereby enabling series connection of multiple units. With sixteen serially connected pump units, the soft pump shows a 90.2 kPa output pressure (18 times higher than that of needle/ring configuration EHD pumps) and 800 mL min⁻¹ flow rate.

Second, upon immersing a soft polymer in an organic dielectric liquid, the increased polymer's chain mobility leads to volumetric expansion and resulting in alterations in its physical characteristics, including dimensional expansion, shape modification, and changes in mechanical properties.^{[ 29} , ^{30 ]} The swelling issue in the soft EHD pump poses two significant drawbacks: first, it alters the shape, dimensions, and positions of the electrodes, consequently impairing their electro‐fluidic performance; second, it generates substantial internal stresses within the structure, potentially leading to overall deformation or even structural damage to the pump.

Due to its high polarization from water content, the hydrogel is an excellent candidate for preventing the swelling of dielectric liquids, which usually have low polarization, in EHD pumps. An additional advantage of hydrogel is its UV‐curability and suitability for fabrication into intricate, high‐resolution structures using DLP technology.^{[ 31} , ^{32 ]} However, hydrogels tend to exhibit weak stability and poor performance in dry environments, as their water content is essential for maintaining their structure and functionality. By combining a hydrogel solution with an organogel solution, an organohydrogel is synthesized, which demonstrates enhanced polarization due to the presence of the hydrogel component. This hybrid structure effectively mitigates swelling induced by dielectric liquids and retains water through the organic phase, thereby improving stability and functionality. Based on the above concept, we synthesized a stable UV‐curable organogel–hydrogel that can be DLP printed into a soft shell for the EHD pump. The interlinked structures of the organogel and hydrogel are bound via covalent bonds. Notably, the dielectric liquid employed is insoluble in water, thereby restricting the ingress of solvents into the material and diminishing swelling within the hydrogel networks. Our most promising sample demonstrates an 8% swelling ratio by weight and exhibits promising mechanical properties, with Young's modulus of 0.33 MPa and ≈300% stretchability. Using this organohydrogel to 3D print the shell enhances the softness of the manifold EHD pump, allowing it to easily withstand bending and twisting while maintaining consistent output pressure.

Coupled with the lab‐made control system, we also demonstrate their potential use in vertical farming for vegetable picking. For soft bionic robot application, a soft robotic worm powered by this pump was also created, which was able to climb on a designed substrate and a trunk surface. Combined with a customized minimized programmable power supply, an untethered bionic soft robotic squid was created and performed swimming in water. The successful demonstrations of our soft pump indicate its potential to impact a broad range of applications for electrohydrodynamic pumps, especially in the field of soft robotics.

2. Results and Discussion

2.1. Strategy for Enhancing the EHD Pump's Fluidic Performance and Optimization of 3D Printable Low‐Swelling Organohydrogels

To apply the EHD pump in soft robotics, it is necessary to enhance its electro‐fluidic performance, originally designed for heat and mass transfer applications.^{[ 33} , ^{34 ]} EHD pumps are divided into three types: induction, conduction, and ion‐drag.^{[ 35} , ³⁶ , ³⁷ , ^{38 ]} Our pump functions according to the ion‐drag principle: charges are injected directly into or from the fluid at the interface between the electrode and the fluid via electron tunneling. Subsequently, these injected charges traverse the electric field lines from the emitter to the collector electrode, exerting a drag force on neutral molecules and inducing fluid motion, as shown in Figure 1a. This injection process typically occurs under high applied electric fields generated by a DC voltage. Generally, regardless of whether the needle/ring is connected in the ± or −/+ configuration, the output pressure does not change, as shown in Figure S1 (Supporting Information).

Figure 1.

In this diagram, by utilizing DLP's high precision and the concentrator's design, multiple pump units can be serially connected to enhance the pump's electro‐fluidic performance. A soft, low‐swelling organohydrogel synthesized for DLP 3D printing is employed in fabricating EHD pump shells. a) demonstrates the ion‐drag effect: under high‐voltage direct current, dielectric liquids transform into anions at the needle electrode, propelled toward the positive electrode by the electric field, carrying neutral molecules with them, and finally reverting to neutral molecules at the anode. b,c) depict the high‐precision components of the pump unit in DLP 3D printing with high Young's modulus flexible polyurethane (FPU), along with the concentrator ensuring installation accuracy. d) presents scanning electron microscope (SEM) images showcasing the needle (left) and ring (right) electrode substrates fabricated through DLP printing, thereby evidencing remarkable printing precision. e) depicts the complete architecture of the pump: internally, it comprises interconnected pump units with uniform performance, while externally, it is enveloped by a 3D‐printed shell. f) provides a comparative analysis of the performance of several recently developed soft EHD pumps. Our pump exhibits the highest output pressure and flow rate, thus facilitating its potential deployment in the domain of soft robotics. g) Components of the organogel solution: aliphatic polyurethane prepolymer (AUD); photoinitiator (TPO); linalyl acetate (LA). h) The composition of the hydrogel includes acrylamide, poly(ethylene glycol) diacrylate (PEGDA), a water‐soluble photoinitiator (PI), as well as a surfactant and water. i) The schematic diagram illustrates the DLP 3D printing process and the structure of the organohydrogel. j) This illustration portrays the entirety of the DLP printing process, alongside the crosslinking mechanism inherent to the organohydrogel. k) The hydrogel networks impede the ingress of LA into the interior of the organohydrogel, consequently diminishing the swelling coefficient. l) The molecular chains of the hydrogel, cross‐linked with the organogel at the interface, remain unaffected by LA and subsequent swelling. Consequently, this phenomenon constrains the swelling rate of the organogel at the interface, a phenomenon termed the 'edge length confinement effect. m) A comparative analysis of Young's modulus and the corresponding swelling ratios across various materials reveal a trend wherein softer materials exhibit heightened susceptibility to swelling, while harder materials demonstrate decreased susceptibility. Nevertheless, our organohydrogel achieves the dual function of maintaining a low Young's modulus alongside a reduced swelling ratio.

Connecting more pump units in a series poses challenges due to the inconsistent performance among the different units. To streamline control and power multiple pump units from a single power source, consistent electro‐fluidic performance is crucial. Inconsistent performance may cause some units to fail to generate sufficient pressure or experience breakdowns, resulting in gas production and potential field breakdown in adjacent units, ultimately risking catastrophic failure of the entire pump system. The consistency of pump units is improved by enhancing electrode dimensional precision and assembly accuracy. DLP printing ensures the electrode's dimensional precision. Furthermore, we have devised a concentrator to ensure accuracy in electrode alignment, focusing on both inter‐electrode spacing and electrode concentricity, as shown in Figure 1b,c. Figure 1d demonstrates the regularity of the 3D‐printed electrode substrate, highlighting the advantages of DLP in 3D electrode fabrication. Figure 1e depicts the structural configuration of our EHD pump, wherein pump units are housed within a shell fabricated through 3D printing. This design facilitates effortless modification of the number of series‐connected pump units to accommodate varying operational demands. In Figure 1f, a comparative analysis of several recently developed EHD pumps is presented. Our pump has the highest output pressure and flow rate, making it suitable for powering a wide range of soft robotic systems.

The swelling behavior of soft polymers in organic solvents poses a disadvantage for devices requiring structural stability.^{[ 39} , ^{40 ]} For EHD pumps, swelling presents a considerable challenge that warrants attention. We present a strategy aimed at achieving a low‐swelling, 3D‐printable material while simultaneously reducing its Young's modulus and enhancing its stretchability.

This strategy entails incorporating the dielectric fluid as a plasticizer into the AUD monomer prior to polymerization into polyurethane (PU), forming the organogel as shown in Figure 1g. The primary constituents of the acrylamide‐based hydrogel are predominantly showcased in Figure 1h. The surfactant sodium dodecyl benzene sulfonate (SDBS) aids in the formation of stable oil–water mixtures, with its addition ratio being dependent on the proportion of organogel to hydrogel solutions. By incorporating a photocurable hydrogel solution into the organogel solution, due to the immiscibility between the dielectric fluid and water, a stable emulsion is achieved. This process aids in reducing the swelling upon solidification. Figures S2,S3 (Supporting Information) illustrates the influence of different concentrations of SDBS on the stability of the mixed solution. Figure 1i depicts a schematic illustration of the microstructure of the emulsion alongside the underlying principle of DLP 3D printing. With the aid of SDBS, droplets of the hydrogel solution are dispersed randomly within the organogel solution. Following layer‐by‐layer exposure by the DLP printer, the organogel–hydrogel emulsion undergoes solidification, yielding the formation of organohydrogel. Figure 1j illustrates the solidified microstructure of the organohydrogel solution. Then we shows a photograph of the EHD pump shell fabricated through 3D printing using this solution in Figure S4a (Supporting Information). The solidification process comprises three distinct phases: the solidification of the organogel, the solidification of the hydrogel, and the solidification of the interface between the organogel and hydrogel, as shown in Figure S4b (Supporting Information). Figure S4c,d (Supporting Information) depict the chemical processes of crosslinking and solidification for the PU‐based organogel and acrylamide‐PEGDA‐based hydrogel, respectively. It is noteworthy that the interface between the organogel and hydrogel is also covalently bonded, due to the abundant presence of acrylate groups in both the organogel solution and the hydrogel solution, as shown in Figure S4e (Supporting Information).^{[ 41 ]} To further validate this mechanism, a specimen consisting of half organogel and half hydrogel is prepared. Tensile testing confirmed that the organogel and hydrogel are indeed strongly bonded since the fracture does not occur at the interface, as shown in Figure S5 (Supporting Information). The organohydrogel demonstrates two mechanisms that contribute to lowering its swelling rate: the shield effect and the edge length confinement effect, as shown in Figure 1k,l. Since water and LA are immiscible, it is theoretically expected that hydrogel particles would not swell in an LA solution.

To verify the swelling behavior of the hydrogel in LA, a photocured hydrogel was prepared and immersed in LA for 24 h. Interestingly, no change in the mass of the hydrogel was observed, as illustrated in Figure S6 (Supporting Information). Various shapes of hydrogel particles are dispersed randomly within the organohydrogel, akin to small shields, impeding the ingress of LA.^{[ 42 ]} During the DLP printing process, larger hydrogel droplets are squeezed from spherical shapes into disk shapes, forming a local network of hydrogel and further enhancing its resistance to swelling. Another mechanism aimed at reducing swelling is referred to as the “edge length confinement effect”, as shown in Figure 1l. By optimizing the proportions of each component, we achieved an organohydrogel with a swelling rate below 10% (measured by mass alteration). The material's Young's modulus undergoes a notable reduction from 68 to 0.33 MPa, thereby exhibiting increased softness. Figure 1m illustrates that our organohydrogel displays Young's modulus similar to commonly used soft materials and, more importantly, exhibits a low swelling rate comparable to materials with high degrees of cross‐linking and high Young's modulus. This feature makes it highly suitable for application in soft EHD pumps.

2.2. The Structure and Electro‐Fluidic Performance of the EHD Pump

Figure 2a delineates the fundamental structure of the EHD pump and elucidates the factors influencing its electro‐fluidic performance. The needle diameter, ring diameter, and needle length are pertinent to the component dimensional accuracy, while the concentrator thickness and its printing precision are crucial aspects of assembly precision. Figure 2b illustrates the substrate of needle and ring electrodes fabricated via DLP printing with flexible polyurethane (FPU). Both scanning electron microscopy (SEM) and optical microscopy images depict a close alignment between the printed contours and the intended design. Following thermal evaporation, a layer of ≈40 µm copper (shown in Figure S7, Supporting Information SEM image) is deposited onto the DLP‐printed substrate, thereby forming the copper electrodes. The final step involves inserting a pair of electrodes into the concentrator and securely fixing them in place using UV‐curable adhesive. The grooves of the concentrator match the dimensions of the electrodes, ensuring concentric alignment between the two electrodes, as shown in Figure 2c. Furthermore, the thickness of the concentrator ensures the appropriate distance between the needle tip and the ring surface. Figure 2d presents the quantified values and corresponding variances of all factors influencing EHD electro‐fluidic performance that have been measured in our study. We have successfully constrained the standard deviation for all factors to ≈1%. Figure 2e illustrates a pump configuration featuring four pump units arranged in series, showcasing the flexibility to adjust the series configuration effortlessly through the utilization of diverse 3D‐printed shells. In Figure S8 (Supporting Information), we show the 3D‐printed FPU shell and its flexibility, as well as pumps containing different numbers of pump units.

Figure 2.

The structure and performance of the soft pump. a) The fundamental structure of the EHD pump unit and the two key factors influencing its electrohydrodynamic performance are: precise structural components and assembly accuracy. b) The 3D‐printed flexible electrode substrate and the copper‐coated electrode, demonstrating the precision achieved through the 3D printing process. The 3D‐printed concentrator ensures high assembly accuracy. c) The concentrator maintains the inter‐electrode distance through its thickness, while its framework ensures the alignment between the needle and ring electrodes. d) shows the specific values and error margins of the main factors influencing the consistency of the pump's electrohydrodynamic performance, with all error margins within 1%. e) This is an EHD pump comprised of four pump units arranged in series, with the capacity for expansion as required. f) The incorporation of a concentrator tot only enhances the consistency of electro‐fluidic performance across different pumps but also amplifies the output pressure under equivalent voltage conditions. g) The voltage–pressure curves of five distinct pump units demonstrate a remarkable degree of consistency and robust performance uniformity. h) The voltage–pressure characteristics of pump units arranged in series demonstrate a discernible linear correlation, indicating that pressure elevates proportionally with the incremental pump units at a consistent voltage level. i) The flow rate increases with the number of serially connected pump units. j) The power density and energy efficiency of the pump with two pump units increase with the applied voltage. k) The pump's continuous testing reveals a stable flow rate of 167 mL min⁻¹. l) The output pressure measured at different voltages following various intervals of continuous testing reveals a stable performance. m). A response time of 0.6 s is displayed for the pump with six pump units.

To assess the functionality of the concentrator, we conducted comparative experiments, the results of which are depicted in Figure 2f. The setups for measuring the out pressure and flow rate are shown in Figure S9 (Supporting Information). The concentrator demonstrates dual effects. First, it augments the stability of the voltage–pressure correlation. Under maximum pressure output, the concentrator yields a standard deviation of ≈0.38, whereas the absence of a concentrator results in a standard deviation of 1.44 for the pump's maximum pressure output. The pressure of the pump with the concentrator at 12 kV is ≈50% higher than that of the pump without the concentrator. Second, under equivalent input voltages, the output pressure is amplified. This effect arises from the channel constraints of the concentrator, which limit the divergence flow between the electrodes. We subsequently conducted trials on five pump units, with the results depicted in Figure 2g. The closely aligned pressure–voltage curves demonstrate commendable uniformity among the pump units, thereby facilitating the potential enhancement of performance through the serial connection of multiple units.

We initiated endeavors to augment the number of pump units in series and evaluate their electro‐fluidic performance. Figure 2h shows the output pressure of the pumps, which are composed of different pump units that are serially connected, showing that the pressure is linear with the increased number of pump units. When the number of pump units is increased to eight, the soft pump's peak pressure can reach ≈46 kPa, which is sufficient to drive most soft hydraulic actuators. To assess the maximum output pressure, we interconnected two soft pumps, each equipped with eight pump units, arranged in series. The pressure rises to 90.2 kPa, which is ≈18 times higher than that of a similar needle/ring configuration EHD pump.^{[ 28 ]} Theoretically, the flow rate would also increase as the number of pump units increased. We tested the flow rate of the pumps, and the results are shown in Figure 2i. The outcomes support the prediction, demonstrating a favorable correlation between the flow rate and the quantity of serially connected pump units. With eight pump units, the soft pump's maximum flow rate was ≈600 mL min⁻¹. When the pump units were raised to sixteen, the maximum flow rate was increased to ≈800 mL min⁻¹. Nevertheless, the rate of increase in the flow rate diminishes as the number of pump units is continuously raised, owing to the heightened flow resistance at a high flow rate. Two more crucial features of a well‐performing pump are high power density and energy efficiency. We created a pump with two pump units to assess the power density and energy efficiency of the soft pump. With increasing the applied voltage, both power density and energy efficiency are improved, as shown in Figure 2j. The weight of the tested pump is 8.9 g, and the peak power is 1.2 W related to a power density of 135 W kg⁻¹. The energy efficiency rises with the applied voltage, reaching a maximum of just 0.9% at 12 kV. Compared to commercial solenoid pumps, such as the HAWE HR080, which have an energy efficiency of 30% to 50%, the developed soft pumps all have much lower energy efficiencies. For the DEA‐based peristaltic pump, its energy efficiency is between 0.09% and 0.13%.^{[ 20 ]} Among the developed soft pumps, the magnetohydrodynamic‐based pump has a relatively higher energy efficiency of ≈12%.^{[ 25 ]} The relatively low energy efficiency of charge injection electrohydrodynamic pumps^{[ 26} , ²⁷ , ^{28 ]} is commonly associated with the high electric resistance of dielectric liquid between the pump unit electrodes. We conducted a 10‐hour long continuous pumping test, and the results are shown in Figure 2k. In this lengthy continuous pumping test, the volume of dielectric liquid pumped by the soft pump is linear with the working time, indicating a stable flow rate. Every three hours, we turned off the pump and examined the pressure–voltage curves (Figure 2l). It is apparent from all the tested pressure–voltage curves that there is no degradation of pumping performance during this lengthy continuous test. We tested the response time of the soft pump with six pump units (Figure 2m). The time from the voltage being applied to reaching the stable pressure stage (i.e., the response time) is ≈0.6 s. This duration measured includes the time required for our power source to reach the target voltage, therefore, the effective response time should be smaller than 0.6 s. We conduct a comprehensive comparison of the response times for recently developed EHD pumps, as presented in Table S1 (Supporting Information). The response time varies between 0.2 and 1.4 s.

2.3. The Swelling and Mechanical Behaviors of the Organohydrogel

We conducted assessments of the swelling characteristics of several common soft materials immersed in the LA, including 3D printable soft PU, polydimethylsiloxane (PDMS), and Ecoflex; the results are shown in Figure 3a. Within a span of 5 h, all materials surpassed a swelling ratio of 100%, which makes them unsuitable for the pump construction. The comparison of swelling after 24 h can be found in Figure S10 (Supporting Information). The FPU is anticipated to have a low swelling rate due to its high degree of cross‐linking and low stretchable molecular chains. We tested its swelling performance and found that it reached a swelling limit of ≈11% within 48 h, as shown in Figure 3b. The FPU exhibited a remarkable Young's modulus of 68 MPa, far surpassing the Young's modulus of commonly used soft materials (<2 MPa), as shown in Figure 3c. We added LA to FPU and obtained organogels with different properties by adjusting their proportions. While other organic solvents are capable of dissolving FPU to create an organogel, it's noteworthy that LA exhibits compatibility with a wide range of organic solvents. We conducted mechanical testing on organogels formed with varying LA concentrations, as depicted in Figure 3d. The results demonstrate that higher LA content notably diminishes the Young's modulus of the organogel. An organogel formed with 40wt% LA by weight exhibits Young's modulus of ≈0.48 MPa, comparable to that of commonly used soft materials such as PDMS, Ecoflex, and thermoplastic PU (TPU).

Figure 3.

The swelling and mechanical properties of the organohydrogel. a) The swelling tests (weight changes) of several commonly used flexible materials reveals their pronounced susceptibility to swelling in LA. b,c) showcase a 3D printable FPU material with a high Young's modulus and low swelling rate, which can serve as the matrix for the organohydrogel. d) illustrate the mechanical performance of FPU/LA blends at different ratios. The findings suggest that an increased LA content leads to a reduction in the material's Young's modulus, albeit with a corresponding decrease in tensile strength. e) demonstrates that varying proportions of FPU and LA result in an increased swelling ratio with higher LA content. f) illustrates the swelling performance of different ratios of organogel to hydrogel in LA indicating that a higher proportion of hydrogel can significantly reduce the swelling rate. g) illustrates the influence of sample thickness on swelling rate, with thicker samples showing reduced susceptibility to swelling. h) demonstrates that varying ratios of organohydrogel show altered mechanical properties, with increasing hydrogel showing enhanced tensile strength. i) Illustrates the cross‐sectional morphology using an optical microscope of the samples printed via DLP 3D printing at various hydrogel concentrations. With an increase in hydrogel content, numerous hydrogel droplets are compressed and solidified into disk‐like shapes. The scale bar in the figure represents 250 µm. j) demonstrates the shell printed with organohydrogel, capable of serially connecting different numbers of pump units, showcasing remarkable softness.

Then we tested the swelling performance of organogels with different ratios of FPU in LA, and the results are shown in Figure 3e. The results indicate that as the LA content increases from 30 to 70 wt%, the swelling ratio correspondingly increases from 60% to 100%. This phenomenon is attributed to the reduction in FPU content with the increase in LA content, which subsequently decreases the cross‐linking density of the entire organogel, thereby leading to an elevated swelling ratio. We selected a mixture with an FPU to LA ratio of 60:40 as the organogel to form organohydrogel with hydrogel in varying proportions. We evaluated the swelling properties of organohydrogels with varying component ratios in LA, and the results are presented in Figure 3f. The results demonstrate a discernible trend: with an elevation in the hydrogel ratio, there is a corresponding reduction in swelling rate. At a ratio of 100:10 between organogel and hydrogel, the swelling rate ≈60%. Yet, with a shift to a ratio of 100:50, the swelling rate diminishes to ≈25%. The thickness of the sample is also recognized as a pivotal determinant of the swelling rate. The aforementioned data sets are derived from measurements obtained from samples measuring 2 mm in thickness and 30 mm in diameter. The sample thickness was augmented to 6 and 10 mm, and the corresponding swelling test data has been depicted in Figure 3g. Consequently, the swelling ratio or the 6 mm sample was ≈17%, while for the 10 mm sample, it diminished to ≈8%. Following these observations, we undertook an investigation into the tensile properties of organohydrogels with varying levels of hydrogel content, encompassing both tensile modulus and stretchability, as shown in Figure 3h and Figure S11 (Supporting Information). The Young's modulus decreased from 0.48 MPa for the pure organogel to 0.3 MPa for the organohydrogel, with an organogel to hydrogel ratio of 100:50. Furthermore, the stretchability increased from ≈150% to 300%, which can be attributed to the enhanced toughness imparted by the hydrogel component. To evaluate the post‐dynamic loading stability of the organohydrogel (with an organogel‐to‐hydrogel ratio of 100:50), systematic mechanical testing was conducted following 100 cycles of tensile loading, 200 cycles of bending, and 200 cycles of compression, as shown in Figure S12 (Supporting Information). The stress–strain curves demonstrate no significant alterations after dynamic loading, suggesting exceptional durability. We further evaluated the Shore hardness of organohydrogels with varying hydrogel content as an additional metric for assessing softness, with the results depicted in Figure S13 (Supporting Information). The Shore hardness of the pure organogel is ≈33 A, whereas the Shore hardness of the organohydrogel with an organogel to hydrogel ratio of 100:50 is ≈15 A, which falls within the range indicating softness based on Shore hardness.

We investigated the morphological images of cross‐sections under various ratios of organogel to hydrogel. To enhance contrast, we added 0.5% water‐soluble green dye to the hydrogel. As illustrated in Figure 3i, the green portion represents the hydrogel, while the yellow portion represents the organogel. When the content of hydrogel is relatively low, it is uniformly distributed in small droplet form within the organohydrogel. Nevertheless, with the elevation of hydrogel content, the coalescence of numerous small droplets into larger ones becomes more apparent. Under DLP printing, these larger droplets are compressed into disc‐shaped structures, as depicted by the elongated green blocks in the cross‐sectional view. The maximum thickness of the disc‐shaped hydrogel corresponds to the printing thickness per layer. Within the same layer and between adjacent layers, the hydrogel forms a network, thereby enhancing its resistance to swelling. The original images of the cross‐sectional morphology as well as the microscopic views of the top layer are both presented in Figures S14,S15 (Supporting Information). Additionally, cross‐sectional SEM images captured at high magnification depict both disk‐like and droplet‐shaped hydrogel formations, as shown in Figure S16 (Supporting Information). Furthermore, we tested the transparency of samples with different hydrogel contents in the visible light spectrum, and the results are displayed in Figure S17 (Supporting Information). The findings suggest a gradual reduction in transparency with higher hydrogel content. This phenomenon is attributed to the disparity in refractive indices between the hydrogel and organogel, leading to partial light reflection at the interface. Additionally, the presence of numerous interfaces between the hydrogel and organogel dissipates incident light, contributing to the observed decrease in transparency. Finally, contact angle measurements were performed on organohydrogels with different hydrogel contents, as shown in Figure S18 (Supporting Information). At an organogel‐to‐hydrogel ratio of 100:50, the contact angle measured less than 20°, indicating excellent hydrophilicity. Generally, hydrophilicity and lipophilicity are opposite properties; high hydrophilicity implies high lipophobicity This characteristic serves as a significant contributing factor to its low swelling ratio.

We employed an organohydrogel solution with a ratio of organogel to hydrogel set at 100:50 for 3D printing soft shells capable of serially linking various pump units, as demonstrated in Figure 3j. To evaluate the precision of DLP printing with this organohydrogel, we selected several geometric parameters of the shell for four units, and the results are presented in Figure S19 (Supporting Information). The resolution of the DLP printer is typically in the range of several tens of micrometers, which contributes to the high precision of organohydrogel printing. We examined the stability of large‐sized parts by printing a 10 mm thick part, which exhibited excellent uniformity in its cross‐section, as depicted in Figure S20 (Supporting Information). The organohydrogel shell fabricated via 3D printing demonstrates exceptional softness, enabling effortless bending of up to 180° and substantial torsional flexibility. We compared the swelling and deformation of soft shells (without FPU) made from different materials after 24 h of immersion in LA. Only our organohydrogel‐based shell showed no deformation, while shells made from other flexible materials exhibited significant deformation, as depicted in Figure S21 (Supporting Information). We tested the output pressure of single pump units with organohydrogel shells, as well as the output pressure and flow rate of two pump units connected in series, with results shown in Figure S22 (Supporting Information). The results indicate that both the output pressure and flow rate are consistent with pumps using FPU as the shell. We also tested the influence of bending and twisting on the output performance of the pump, as shown in Figure S23 and Video S1 (Supporting Information). When four pump units are serially integrated within the organohydrogel shell, the output pressure measures 22 kPa under static conditions. Moreover, upon subjecting the assembly to bending and twisting, no discernible variation in the output pressure is observed. This phenomenon stems from the elevated modulus of FPU, compounded by the formation of a stable structure through the concentrator and electrodes. The soft shell primarily absorbs the majority of strain, leaving minimal deformation that could impact the pump units' output. We also fabricated pumps with varying numbers of pump units (Figure S24, Supporting Information) and demonstrated the excellent softness of a pump containing eight pump units in Figure S25 (Supporting Information).

Finally, we conducted a 14‐day long‐term stability experiment on two samples of organohydrogel with a thickness of 10 mm and a ratio of organogel to hydrogel set at 100:50, under laboratory air and in a controlled environment of LA, as shown in Figure S26 (Supporting Information). For the samples in lab air, their weight decreased from 8.454 to 8.155 g, reflecting a 3.5% reduction in mass. This decrease is attributed to the evaporation of LA in the ambient air. In LA, the sample's weight increased from 8.661 g to 9.614 g, reflecting an 11% increase due to swelling. Although there were weight fluctuations, neither sample displayed noticeable changes in volume or deformation, underscoring the robust stability of our organohydrogel. To further investigate the long‐term stability of the organohydrogel, we conducted a 10‐day evaluation under varying pH, humidity, and temperature conditions. The results demonstrate that the organohydrogel maintains stability across a broad range of environmental conditions. Further details are provided in Figure S27 (Supporting Information).

2.4. A Soft EHD Pump is Used to Drive Pneumatic/Hydraulic Grippers

As shown in Figure 4a, we demonstrate the soft pump powering a pneumatic/hydraulic actuator‐based soft gripper. Four pump units are arranged in a forward alignment in the soft pump to create positive pressure and move the dielectric liquid from the right reservoir to the left reservoir. The air will be forced into the actuators to close the gripper (Figure S28, Supporting Information) as the liquid level rises. It performs the role of a pressure converter by converting liquid pressure to air pressure. The secondary purpose of the pressure converter is to prevent the dielectric liquid from infiltrating the soft gripper, which may otherwise be damaged due to the combined effects of pressure and swelling, thus potentially contaminating the objects to be gripped. To create negative pressure under which the gripper is open, two additional pump units are positioned backward to transfer liquid from the left to the right reservoirs. As depicted in Figure 4b, when operated at 12 kV, four forward‐placed pump units can generate a positive pressure of ≈20 kPa. The maximum bending angle of the 3D printable actuator under the stimulation of four forward‐arranged pump units reached 177°, as shown in Figure 4c. Under 20 kPa positive pressure, the printed actuators’ blocking force can reach 0.85 N, as shown in Figure 4d.

Figure 4.

A soft gripper powered by the soft EHD pump. a) The schematic of the soft pump transfering the dielectric liquid between two liquid reservoirs to generate positive/negative pressure used to open/close the gripper. b) The positive pressure produced by the four pump units connected in series is used to close the gripper. c,d) show the bending angle and blocking force of the pneumatic/hydraulic actuator composing the gripper under the activating of forward‐aligned four pump units. e–h) show the gripper grasp of different objects with different weights. i–n) The potential for use in vertical farming is demonstrated by the gripper mounted on a soft robotic arm.

We used various objects to test the gripper system's grasping capacity. As shown in Figure 4e,f, a 37 g green vegetable can be tightly grasped by the gripper in 3 s upon pump activation. Additionally, a 226 g copper coil was tightly grasped in less than 4 seconds after being placed in the gripper, as shown in Figure 4g and Video S2 (Supporting Information). A 500 g metal cylinder was selected as the target object to test the maximum grasping capability of this gripper system, as shown in Figure 4h and Video S2 (Supporting Information). Since it takes more time to achieve maximum pressure when compressing the air in the pressure converter and gripper's chamber, the response time for grabbing the 500 g metal cylinder increased to ≈20 s.

Furthermore, we demonstrated the gripper system combined with a robotic arm to harvest vegetables to mimic its application in vertical farming. To close the gripper, four forward‐positioned pump units were used to generate positive pressure and to open the gripper, two backward‐positioned pump units were used to generate negative pressure. In Figure S29 (Supporting Information). the control circuit diagram and power system are displayed. At the start, the pump units are not in operation and the soft gripper is in its initial state (Figure 4i). As the two backward‐positioned pump units were activated (Figure 4j), the soft pump starts generating negative pressure to open the gripper.

The robotic arm begins to move toward the targeted vegetable once the gripper is open (Figure 4k). To release negative pressure while generating positive pressure when grabbing a vegetable, the two backward‐positioned pump units were turned off, while the four forward‐positioned pump units were turned on (Figure 4l). The robotic arm securely held onto the vegetable before smoothly transferring it to its intended position (Figure 4m). Once in the desired position, the forward‐positioned pump units were turned off and the backward‐positioned pump units were turned on to create negative pressure that opens the gripper and releases the vegetable, as shown in Figure 4n. The whole process mentioned above is shown in Video S3 (Supporting Information).

2.5. The Soft EHD Pump Used to Power a Bioinspired Robotic Worm

The soft pump's high pressure and flow rate enabled the creating soft robotics that operate in the air, where the high gravity and corresponding friction are the challenges to be addressed. To demonstrate this capacity, we implanted the soft pump and a linear hydraulic actuator in a one‐body soft robotic worm that can climb over a rough patterned substrate (Figure 5 ). Figure 5a shows the main structure of the soft robotic worm integrated with a soft pump that can transfer liquid between the liquid reservoir located above the pump and the linear hydraulic actuator to achieve the climbing function. This pump consists of six pump units. Three of these units are positioned in a forward manner, allowing them to transfer liquid from the liquid reservoir to the actuator, and the rest three units are positioned in a reverse manner, enabling them to move liquid from the actuator back to the reservoir. The linear hydraulic actuator is extended when it is filled with the liquid, as shown in Figure 5b. On the contrary, when the liquid is evacuated from the linear hydraulic actuator, it will contract (Video S4, Supporting Information). Figure 5c shows the soft robotic worm climbing on a rough‐patterned substrate. The climbing process consists of two substages: the head is pushed forward as the actuator elongates, and the tail moves to the front as the actuator contracts. The patterned substrate and the structural claw at the bottom, enable an asymmetrical distribution of frictional force. This results in a small frictional force in the forward direction and a significant frictional force in the backward direction. When the actuator is in contraction, the tail is pulled toward the front because the head exerts a greater frictional force compared to the tail. As the actuator extends, the stronger frictional force exerted by the tail pushes the head toward the front. Figure 5d illustrates the pump's output pressure and the actuator's pushing/dragging force. At a maximum output pressure of ≈15 kPa, the linear hydraulic actuator can produce a force of ≈4 N, which is sufficient to drive a 120‐gram soft robotic warm. Compared to a bio‐inspired worm‐type soft robot for in‐pipe locomotion, our linear actuator has the same output force with a smaller diameter of 13 to 20 mm.^{[ 43 ]} We created a programmable high‐voltage power source utilized to power the pump since the forward‐ and backward‐arranged pump units needed to operate alternately.

Figure 5.

A soft robotic worm powered by a soft EHD pump. a) The design of the soft robotic worm that a six‐units pump connected with a linear hydraulic actuator to achieve elongation and contraction. b) The operational principle of the soft robotic worm. Three forward pump units are arranged to transfer dielectric liquid from the liquid reservoir to a linear hydraulic actuator for elongation. Another three backward‐arranged pump units are used for contraction. c) show the design of the soft robotic worm climb on a patterned substrate. d) shows the output pressure and output force increase when the forward and reverse pumps are activated, respectively. e) The linear hydraulic actuator shows elongating and contracting, respectively. The soft robotic worm climbs on the patterned substrate. f) The second‐generation soft worm robot has achieved locomotion on rough surfaces, such as tree trunks. The scale bar in the figure depicts a length of 4 cm.

In Figure S30 (Supporting Information), more information about this programmable high‐voltage power supply is presented. Figure 5e shows the contracted and elongated linear hydraulic actuator, powered by the soft pump. The actuator's length is ≈2 cm during contraction and 5 cm during extension. The soft robotic worm was able to perform the intended climbing, as illustrated in Figure 5e and Video S5 (Supporting Information), when used in conjunction with the head, tail, and patterned substrate that were 3D printed.

To enhance the functionality of the soft robotic worm, we have developed one‐directional paws that enable it to climb rough surfaces like tree trunks. More details are provided in Figure S31 (Supporting Information). The same actuator is then expanded to move the head of the soft robotic worm toward the front as well. We conducted a test to evaluate the soft robotic worm's ability to crawl continuously on the tree trunk. The outcomes of this test can be seen in Figure 5f and Video S6 (Supporting Information). After taking 4 steps, the soft worm moves forward ≈8 cm. Compared to a similar worm‐inspired soft robot, it has a stride length of only 0.7 cm under 15 kPa.^{[ 44 ]}

2.6. An Untethered Soft Robotic Squid Powered by the EHD Pump

By creating a soft robotic squid, we demonstrate the potential of the soft pump used in untethered soft robotics. The squid head contains the control/battery module and the soft pump with four pump units, as shown in Figure 6a. The squid bottom is where the three bending hydraulic actuators function as the legs, and a liquid reservoir is connected. The control/battery module operates in two states: it switches on for a duration of 4 s and then shuts down for 4 s in a cyclical manner. More details about of this control/battery module are shown in Figure S32 (Supporting Information). When the control/battery module is in the opening status, the three bending hydraulic actuators are activated and bend as the dielectric liquid is transferred from the liquid reservoir to them through the soft pump. Conversely, when the control/battery module is in the closing status, the soft pump is deactivated, and the dielectric liquid is forced from the actuators back into the liquid reservoir. This action causes the previously bent actuators to straighten. The soft robotic squid generates propulsion by repeatedly bending and straightening its actuators, pushing the water to create a driving force. Figure 6b displays the soft robotic squid filled with dielectric liquid, while Figure 6c illustrates its components. The swimming ability of the soft robotic squid was tested in a glass tank, as shown in Figure 6d and Video S7 (Supporting Information).

Figure 6.

An untethered soft robotic squid powered by the EHD pump. a) The design schematic of the soft robotic squid involves a mechanism where a soft EHD pump transfers liquid from a reservoir to three bending hydraulic actuators to generate forward‐moving force. b) The photo of the soft robotic squid that is integrated with the control/battery module. c) A picture of the entire soft robotic squid, including the control/battery module, the liquid reservoir, the three bending hydraulic actuators, the three soft tentacles, the soft pump, the leg support, and the soft head. d) The pictures of the squid swimming in the water from left to right. The unit of the ruler is cm.

3. Conclusion

This study introduces notable advancements in the development of EHD pumps and the synthesis of an anti‐swelling organohydrogel, which shows promising potential for soft robotics applications. By utilizing high‐precision DLP technology, we fabricated a manifold pump array with exceptional uniformity and electro‐fluidic performance, achieving an output pressure of 90.2 kPa and a flow rate of 800 mL min⁻¹. Furthermore, the novel organohydrogel developed for the pump shell effectively mitigates swelling issues, with a swelling ratio of under 10%, and enhances durability, offering Young's modulus of 0.33 MPa and 300% stretchability. The pump's stability under various mechanical stresses, such as bending and twisting, makes it ideal for dynamic soft robotic environments. These innovations open new possibilities for soft robotics, bioengineering, and vertical farming, where compact, high‐performance pumps are essential. Future research can focus on optimizing the scalability and control of this technology, potentially enabling its integration into more complex and autonomous soft robotic systems. Expanding the design to include more adaptable materials and configurations may further enhance the versatility of these pumps, reinforcing their role in advancing the field of soft robotics and beyond.

4. Experimental Section

Fabrication of the Pump Units

1. The three components of the pump units were the needle‐electrode, hole‐electrode, and concentrator. These components were printed using a transparent soft UV curable resin (F39T, Dongguan Godsaid Technology Co., Ltd.). Autodesk Fusion 360 was used to create the 3D model of the individual parts. The 3D model was then digitally cut into slices using the slicing program HALOT BOX, with a slicing thickness set to 30 µm. Afterward, the sliced 3D model was transferred into the 3D printer (HALOT‐SKY, Creality 3D Technology Co. Ltd). To ensure a strong adhesion between the printing platform and the printed object, the first layer's exposure time was set to 17 s, and the subsequent layer's exposure time was set to 5 s.

2. To coat a thin layer of copper on the surface, the needle‐ and hole‐substrates were adhered to a thermal evaporator plate. A thermal evaporator (FL400, Syskey Technology Co., Ltd.) was used for the coating. The copper pellets (Kurt J. Lesker Co., Ltd.) have a purity of 99.99%. The resulting copper layer was ≈40 µm thick to ensure good conductivity.

3. The needle‐electrodes and hole‐electrodes were arranged on both sides of the concentrator through the designed framework of the concentrator. UV glue was applied to the gaps between the electrodes and the concentrator, followed by five seconds of UV irradiation for curing to ensure the stability of the structure.

Fabrication of the Soft EHD Pump with Commercial Flexible Resin

1. The shell of the soft EHD pump, which was printed with a soft polymer (F39T, Dongguan Godsaid Technology Co., Ltd.), was designed with two parts: the bottom and the top parts. The printing parameters of the shell were the same as those for the pump units.

2. To ensure that the liquid flowed in the same direction, the pump units were installed into the bottom shell in a head‐end manner. Following that, the top shell was assembled with the bottom shell. Following the steps, the shell was wrapped around the pump units, allowing only a portion of the electrodes to penetrate on the left and right sides of the soft pump. A copper wire connects the electrodes’ penetrated parts on the same side. Finally, UV glue was used to seal the gaps between the shell and the penetrated electrodes.

Fabrication of the Soft EHD Pump with Low Swelling Ratio Organohydrogel

This organohydrogel comprises two components: one part was organogel, and the other part was hydrogel. The organogel solution was relatively simple, consisting of a mixture of 100 g of F39T (FPU) and LA in different ratios, which was then stirred in a high‐speed shearing machine at 2500 revolutions per minute for 10 minutes. While the hydrogel component was more complex, including acrylamide, PEGDA, water‐soluble TPO,^{[ 45 ]} glycerol, water, and SDBS in a ratio of 30:0.5:0.5:19:30:20. Like the organogel solution, this mixture was also blended in a high‐speed shearing machine at 2500 revolutions per minute for 5 min. Then, the organogel solution and hydrogel solution were mixed in different ratios and further blended in the high‐speed shearing machine to obtain a stable organohydrogel suspension, which was finally used in DLP printing to form the soft shell of the EHD pump.

Performance Tests

The soft pump was powered by a high‐voltage power supply (HB‐Z153‐1AC, HengBo Power Supply Co., Ltd.) with a maximum 15 kV DC voltage output. The power supply could display the output voltage and current, and the output voltage was programmable. A soft pump was used to link two liquid reservoirs (each has an internal volume of 64 µL), and one reservoir had a pressure meter (DP380, SanLiang) connected to it to measure the output pressure. The soft pump then freely pumped the dielectric liquid between the two reservoirs once the pressure meter was removed in order to evaluate the flow rate. The input voltage and input current were compounded to determine the soft pump's input power. The tests were in two scenarios to determine the soft pump's energy efficiency. In scenario 1, to ensure that all of the kinetic energy of the liquid created by the soft pump was converted to the internal energy of the system, the two reservoirs were positioned at the same level. It was timed when the soft pump moved all of the liquid from one reservoir that was full to the other that was empty. In scenario 2, the soft pump should first be closed in order to measure the energy that was transferred to the liquid's kinetic energy. Then, the height difference between the two reservoirs was adjusted so that the time it takes for all liquid to transfer from one to the other was equal to the time previously recorded. Due to with the same flow rate in scenario 1 and scenario 2, the work done by electricity was equal to the work done by gravity. The weight of the dielectric liquid multiplied by the height difference between two liquid reservoirs and divided by the transfer time could be used to calculate the soft pump's output power. Finally, the energy efficiency was determined by dividing the output power by the input power (applied voltage multiplied by current).

Soft Gripper System

The bending actuators for the soft gripper were fabricated by 3D printing with a 1:1 ratio of Ebecryl 8413 and Ebecryl 113, and 5% photoinitiator, diphenyl (2 4 6‐trimethylbenzoyl) phosphine oxide. The gripper was made up of four bending actuators that were fixed to a 3D‐printed basement. The basement was printed with a rigid UV‐curable ink (3D Printer UV Sensitive (white) Resin, CREALITY 3D), and designed with inner flow channels connected with the tube to provide pressure to the four printed bending actuators.

A power system was used for this demonstration. Three DC‐DC converters and a 12 V battery make up the demonstration's power supply. Two DC–DC converters were used to power the four forward‐arranged pump units to generate positive pressure. The other one was used to power the pump units that were arranged backward in order to create negative pressure. The DC‐DC converter was composed of three components: DC–AC, AC–AC, and AC–DC. The DC–AC module transformed the battery's DC voltage into AC voltage. The input AC voltage was then raised to ≈1000 kV AC voltage by the AC–AC module. The AC–DC module raised the output AC voltage to a maximum of 16 kV. Through various cable connections, the output voltage could be chosen. The authors settled on a 12 kV output voltage.

Bioinspired Robotic Worm

A DLP printer was used to 3D print every component of the worm. The shell of the worm, linear actuator, and liquid reservoir were printed with a transparent polymer (F39T, Dongguan Godsaid Technology Co., Ltd.). The dielectric liquid was stored in the liquid reservoir and the actuator. Transferring the dielectric liquid between the actuator and the liquid reservoir allowed the actuator to expand and shrink. The structured claw and the patterned substrate were printed with transparent rigid green ink (3D Printable UV Sensitive Resin (green), Creality 3D Technology Co. Ltd). The components of the paws were printed with transparent rigid red ink (3D Printable UV Sensitive Resin (red), Creality 3D Technology Co. Ltd).

The power supply was programmable to control the actuator's expansion and shrinkage in a designed way. Figure S24a,b (Supporting Information) depicted the power supply's schematic and graphic representation. Battery, DC–DC converter, and control center were its three component parts. A single‐chip microcomputer (Arduino UNO, Arduino. cc) was used to open and close micro relays (G6K‐2F‐5 V, Omron Electronics Company) serially connected with DC–DC converters. The code is shown in Figure S24c (Supporting Information). To power the pump, two DC–DC converters were utilized. One drives the three pump units that were arranged forward, while the other drives the three pump units arranged backward. In this power supply, there were two 9 V batteries: one powered the Arduino UNO, and the other powered the DC–DC converter. The power supply was capable of cyclically activating and deactivating the two DC–DC converters.

Untethered Bioinspired Soft Robotic Squid

Like the soft robotic worm, the head of the soft squid, bending actuator, legs, and the liquid reservoir were printed with the transparent polymer (F39T, Dongguan Godsaid Technology Co., Ltd.). The part that connected the legs, and the head was printed with a transparent rigid ink (3D Printable UV Sensitive Resin, Creality 3D Technology Co. Ltd).

The power supply could be programmed to control the actuators, allowing them to bend and straighten alternately. The structure and circuit diagram the of the power supply is shown in Figure S26 (Supporting Information). To reduce the size of the power supply, the Arduino UNO single‐chip microcomputer was replaced with the Arduino Nano and the battery was switched to a smaller lithium battery. Two optimized DC–DC converters were used to power the soft pump, which consisted of four pump units.

Conflict of Interest

The authors declare no conflict of interest.

Supporting information

Supporting Information

Supplemental Movie 1

Supplemental Movie 2

Supplemental Movie 3

Supplemental Movie 4

Supplemental Movie 5

Supplemental Movie 6

Supplemental Movie 7

Xin Y., Zhou X., Tan M. R. J., Chen S., Huang P., Jiang Y., Wu W., Gao D., Lv J., Magdassi S., Lee P. S., 3D‐Printed Electrohydrodynamic Pump and Development of Anti‐Swelling Organohydrogel for Soft Robotics. Adv. Mater. 2025, 37, 2415210. 10.1002/adma.202415210

Data Availability Statement

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