Electroactive interface–enhanced dielectric elastomer with ultrahigh electroadhesion and multifunctional droplet actuation

Abstract

Dielectric elastomers (DEs) with high energy conversion density are highly desirable for flexible actuation and sensing applications. However, current DEs only perform well under electrical actuation, incapable of regulating interfacial interactions between materials and their environment, such as active control of electroadhesion and electrowetting-on-dielectric (EWOD). Herein, we report a method using polar small-molecule additives to enable electroactive interfacial regulation in DEs. For electroadhesion, the electroactive interface–enhanced dielectric elastomer (EIEDE) with a metal mesh electrode attains adhesion strength of 31.75 kilopascals at 14 megavolts per meter, which is 488 times greater than that before modification. For EWOD performance, the EIEDE induces droplet contact angle to decrease sharply from 83.15° to 9.92°, showing the best electric-field response among DEs and enabling functions like droplet transport and shape modulation. The EIEDE simultaneously integrates large-strain deformation with electroactive interfacial regulation, achieving breakthrough performance in superhigh electroadhesion and excellent EWOD performance.

An EIEDE integrates ultrahigh electroadhesion and multifunctional droplet actuation capabilities.

INTRODUCTION

Dielectric elastomers (DEs) are electroactive polymer (EAP) materials that exhibit electrical activity that is primarily manifested from an energetic perspective in their capability for direct conversion between electrical field energy and mechanical energy. With this electromechanical mechanism, conventional dielectric elastomer actuators (DEAs) demonstrate numerous advantages, including high electromechanical conversion efficiency, high actuation strain, rapid response speeds, and broad frequency bandwidths (1–3), giving them considerable application potential in fields including artificial muscle actuators (4–7) and soft robotics (8–12). The increasing demand for high-performance DEs emphasizes their energy density and electromechanical conversion efficiency as critical characteristics for DEAs (13). When polar functional groups are grafted onto their polymer side chains, organic fillers with high dielectric constants are incorporated, viscoelastic losses in their molecular chains are reduced, and their electrode structures are optimized, DEs have excellent electric-field energy storage and electromechanical conversion (14–18). Although substantial progress has been made with the energy storage and electromechanical conversion capacities of these materials, existing studies have overlooked an important electroactive feature of DEs, i.e., the conversion and regulation mechanism between the electric-field energy and the surface energy.

Therefore, on the basis of a modification approach that uses polar small molecules within a silicone rubber network, we synthesized an electroactive interface–enhanced dielectric elastomer (EIEDE) that combined a versatile range of capabilities, including large strain actuation, high electroadhesion, and multifunctional droplet actuation. The overall DE material design strategy comprises two key aspects. For the electric field–to–mechanical energy conversion, additives were introduced to weaken the intermolecular physical interactions, thereby increasing the dielectric constant while reducing the elastic modulus and mechanical losses to enhance the DE’s actuation performance. For electric field–to–interfacial energy conversion, the reversible charge injection, field-driven oriented migration, and surface aggregation processes of the polar small-molecule additives were used to achieve pronounced modulation of the DE’s surface energy. As Fig. 1A shows, the fabrication process involved stirring and mixing Wacker P7670 component A with additives, followed by adding component B with continued stirring. The mixture was then vacuum degassed and lastly cast on a heated coater. After cross-linking, the additives could migrate freely within the network mesh. Using these characteristics, we fabricated an original EIEDE with large strain actuation, high electroadhesion, and multifunctional droplet actuation properties.

Fig. 1. Polymer design.

(A) Schematic of the preparation process for modified DEAs. (B) Schematic of the actuation mechanism and chemical structures of our EIEDE. (C) Schematic of the electroadhesion principle for metal plate electrodes and metal mesh electrodes. (D) Schematic of EWOD. (E) Schematic of droplet repulsion and directional motion. Software credit: [(A) and (B)] Created with Adobe Photoshop; [(C) to (E)] Created with Microsoft PowerPoint.

For their electromechanical conversion mechanisms, conventional DE deformation involves two primary processes: electric field–induced polarization and electromechanical energy conversion. Under an applied electric field, molecular chain orientation polarization occurs within DEs. Furthermore, under an applied electric field, thickness contraction induced by Maxwell stress leads to areal expansion (19). As Fig. 1B shows, the electroinduced deformation mechanism of the EIEDE developed here exhibited similarities to that of conventional DEs. The elastic modulus was reduced substantially by the additives, and their co-migration with negative charges toward the anode under an applied electric field enhanced the polarization effect. This mechanism realized a major actuation performance improvement when compared with conventional DEs.

For the electrical-surface energy conversion, our work used polar small-molecule additives to enable electroactive interfacial regulation of the DE. Concerning the electrical-to-surface energy conversion process at solid-solid interfaces, the mechanism is primarily manifested through electroadhesion, which has garnered increasing interest over the past decade. When an electric field is applied to a pair of electrodes separated by a dielectric material, electroadhesion on the nonconductive surfaces is achieved through induced charges (20, 21). Conventional DEs rely on dipole moment alterations in the polymer chains, which yield negligible interfacial effects because of the limited polar side-chain reorientation. In contrast, the EIEDE in this work demonstrated charge injection–induced migration and surface aggregation of the additives at the anode, substantially raising the surface energy. By applying an electric field, we can controllably tune the surface energy of the EIEDE in different states. Using this mechanism, we have successfully achieved superhigh electroadhesion at solid-solid interfaces. This electroactive mechanism enabled strong adhesion at solid-solid interfaces, providing an enhancement when compared with conventional DEs (e.g., Wacker P7670). Previous studies indicate that electroadhesion strength on rough surfaces may be substantially lower than that on smooth surfaces because of air gaps between the dielectric material and the substrate. Research on electroadhesive behavior in the presence of air gaps demonstrates that even minute air gaps substantially diminish adhesion strength. Therefore, a challenge in developing electroadhesive devices lies in achieving high adhesion strength on rough surfaces at relatively low operating voltages (20, 21). Our work demonstrates that the EIEDE’s adhesion capability can also be extended to irregular surfaces, which was previously unattainable for traditional DEAs (Fig. 1C).

The electrowetting-on-dielectric (EWOD) behavior of the EIEDE at solid-liquid interfaces was studied systematically. Our developed EIEDE overcomes the limitations of low surface energy and inherent hydrophobicity of conventional DEs coupled with their limited hydrophilic/hydrophobic tunability under applied voltage. When a droplet contacts the anode, charge injection–induced migration of the additive’s molecules toward the anode and their subsequent surface aggregation substantially enhances the surface energy. This converts the EIEDE surface into a hydrophilic state, thereby substantially reducing the contact angle (Fig. 1D) and expanding the tunable range of droplet contact angles. Leveraging this mechanism enables effective droplet manipulation. Conversely, cathode-connected droplets show enhanced hydrophobicity caused by localized migration of the additives and anions. Enhanced hydrophobicity propels the droplet toward the relatively hydrophilic surface, that is, from a region of lower surface energy toward a region of higher surface energy. Confining the electrode geometry causes droplet repulsion and directional motion (Fig. 1E). These findings have enabled advanced droplet manipulation techniques, including droplet transport, deformation, and splitting, considerably optimized electrowetting performance and led to innovative microfluidic control strategies. This work demonstrates the broad applicability of a multifunctional EIEDE in flexible actuation, nonplanar surface adhesion, and microfluidic manipulation, thus establishing a creative paradigm for developing high-performance electroactive materials.

RESULTS

Material properties

In this study, on the basis of the common P7670 elastomer, four polar small-molecule additives [dibutyl adipate (DBA), di(2-ethylhexyl) phthalate (DOP), dioctyl terephthalate (DOTP), and diisooctyl sebacate (DOS)] were used to synthesize EIEDE. The copolymers with DBA are designated as PBAX [where X denotes the DBA fractions (ω) of relative to PBA; tables S1 to S4]. Analogous nomenclature is applied to the other modified systems. In DEs, the elastic modulus and the dielectric constant were identified as critical parameters that govern their electroactive responses. The large dielectric constant (ε_r) plays a critical role in DE actuation. This parameter directly determines the achievable actuation strain as described by the equation S_z = −ε_rε₀E²/Y, where ε₀ denotes the vacuum permittivity, E represents the electric field strength, and Y is the Young’s modulus. As Fig. 2A shows, the dielectric constant of the DBA-modified DEs increased with increasing additives concentration across the frequency range 4 to 10⁶ Hz. This is attributed to interfacial polarization resulting from the orientation of polarized additives molecules (toward the anode) as they migrate within the chain matrix of the DE, indicating substantial additive-induced dielectric modulation. Elevated dielectric loss is a critical consideration for DEs because high dissipation can induce thermal or electrical breakdown via increased temperature and conductivity. Although our approach increased the dielectric constant, subsequent experiments confirmed that the associated rise in dielectric loss remained negligible (see fig. S1, A to G, for the dielectric properties of the other additives). Incorporating additives within DEs disrupts intermolecular interactions between polymer chains. Consequently, the mechanical properties of EIEDE depend strongly on both the physical properties and content of the additives. Mechanical testing revealed a reduced elastic modulus in the PBA systems when compared with P7670 (Fig. 2B), which was consistent with the modulus-reducing mechanism of the additives (additional additives are shown in fig. S2, A to C). The electromechanical sensitivity (β), defined as the ratio of the dielectric constant to Young’s modulus, is an important DE actuation metric. The experimental results showed substantial β enhancements after additive incorporation, particularly for DBA (Fig. 2C), which confirmed improved electric-field responsiveness. Following comprehensive mechanical and dielectric evaluations, DBA was selected as the optimal additive for subsequent testing.

Fig. 2. Material properties.

(A) Dielectric constant characteristic of P7670 with various DBA weight ratios over the range from 4 to 10⁶ Hz. (B) Stress-strain curves of P7670 with various DBA weight ratios. (C) Comparison of actuation electromechanical sensitivity (β) for the different additives. (D) Cyclic stress-strain curves of P7670, PBA5, PBA10, and PBA15 when subjected to strain at 150%. (E) Actuation area strains of P7670 and PBA15 under various electric fields with 150% prestretch. (F) Photographs of the actuation area strains of 200-μm P7670 and PBA15 membranes.

Conventional DEs inherently exhibit pronounced viscoelastic effects, which exacerbates mechanical losses. This study demonstrates that incorporating additives molecules into the EIEDE’s matrix effectively suppresses such viscoelastic behaviors. Specifically, introducing low-molecular-weight additives induces swelling within the polymer network. This swelling increases intermolecular spacing and reduces frictional resistance between polymer chains, thereby diminishing viscous dissipation. Tensile-recovery tests verified the superior resilience and reduced viscoelastic losses of PBA elastomers. PBA15 showed lower hysteresis and residual strain than P7670 across various stretch ratios (Fig. 2D and fig. S2, D and E). The EIEDE exhibits charge (electron) injection from the cathode followed by migration toward the anode, which is unlike the deformation mechanism in conventional DEs. Moreover, adding additives reduces the elastic modulus, whereas DBA migration induces interfacial polarization that enhances the dielectric constant. Such synergistic effects enable EIEDE’s materials to achieve greater actuation strain than conventional systems. On the basis of experimental findings, actuation performance under a 150% biaxial prestretch showed enhanced areal strain with increasing DBA content (fig. S3A), with PBA15 exhibiting 2.5 times greater strain than P7670 at 55 megavolts (MV) m⁻¹ (Fig. 2, E and F). Increasing DBA content progressively reduces the breakdown field strength of the DE, which can be primarily attributed to DBA migration within the material leading to a progressive leakage current and a reduced breakdown field with excessive DBA content. Notably, however, higher DBA concentrations yield enhanced actuation under identical electric field strengths (fig. S3, B to G). For the various prestretch levels, the actuation strain initially increased and then decreased (fig. S3H); this behavior was attributed to the higher prestretch levels stiffening the DE and thereby reducing its performance. We also compared the actuation strains of DE membranes of various thicknesses under a 150% biaxial prestretch: As the membrane thickness increased, the actuation strain under the same electric field decreased as a result of DBA migration, which is affected by the thick membrane network (fig. S3I). Frequency response tests showed that over the 1- to 40-Hz range, PBA15 outperformed P7670 consistently in terms of strain (fig. S3J). The amplitude actuation strain under square-wave excitation exceeded that under sine wave excitation (fig. S3K), and the material demonstrated excellent cycling stability, operating stably for 2500 cycles at 1 Hz and 12% area strain (fig. S3L). These results collectively demonstrated that the additives modulated the EIEDE’s electromechanical performance effectively.

Electroadhesion properties

When the electric field was applied, the DBA molecules in the PBA underwent a massive migration toward the anode, forming strong electrostatic adhesion to the anode. Images and movie recordings indicate that as the voltage increases, DBA molecules undergo vigorous motion beneath the material surface, causing the membrane to creep and adhere to the anode surface (Fig. 3A and movie S1). We used a Raman microscopy system to quantitatively monitor the dynamic variation in plasticizer content within the EIEDE during electrical poling. We first measured Raman spectra of pristine P7670, PBA5, PBA15, PBA25, PBA35, and pure DBA (Fig. 3B). After executing the peak-processing algorithm, we extracted characteristic peak areas corresponding to DBA (C═O, 1676 to 1800 cm⁻¹) and silicone (Si─O, 400 to 750 cm⁻¹). The ratio $\mathrm{\eta }=\frac{S_{\mathrm{C}═\mathrm{O}}}{S_{\text{Si}─\mathrm{O}}}$ was used as a quantitative descriptor of the local DBA concentration. Using the Raman spectra of samples with known plasticizer content, we established a calibration curve linking η to the actual DBA mass fraction ω (Fig. 3C). The relationship follows

$$\mathrm{\eta }=A_1{e}^{-\mathrm{\omega }/{\mathrm{\tau }}_1}+B_1$$ (1)

with fitting parameters $A_1=0.00583$, $B_1=-0.1639$, and ${\mathrm{\tau }}_1=-7.61881 (R^2=0.99)$. Using this calibration, we monitored the DBA distribution in situ before and after electrical stimulation (fig. S4A). The results (Fig. 3D and fig. S4B) show a clear increase in the inferred DBA content near the anode after poling for PBA25. This provides direct chemical evidence that the electric field induces directional migration and interfacial enrichment of DBA molecules within the elastomer matrix. The observed increase in local DBA concentration confirms the proposed mechanism underlying the enhanced surface energy and electroadhesion.

Fig. 3. Electroadhesion properties.

(A) Illustration of electrostatic adhesion between PBA35 and the metal mesh. (B) The ratio η was determined from the Raman spectra of P7670, PBA5, PBA15, PBA25, PBA35, and DBA. a.u., arbitrary units. (C) Established mathematical relationship between Raman spectral ratio η and plasticizer content ω. (D) Changes in plasticizer content after the electrical stimulation of PBA25. (E) Adhesive strengths of P7670 and PBA35 with metal plates under various electric fields. Inset: Demonstration of electroadhesion to a 10-kg object using a metal flat-plate electrode. (F) Adhesive strengths of P7670 and PBA35 with metal meshes under various electric fields. Inset: Demonstration of electroadhesion to a 3-kg object using a metal mesh-plate electrode. (G) Reported adhesive strengths of similar adhesion structures versus that in this work (22, 23). (H) Images of adsorption, handling, and rapid detachment of deformed metal sheets, metal wire meshes, and metal wires.

To investigate this electroadhesion, comparison studies were conducted using a flat-plate electrode and a 100-mesh metal grid electrode and showed that the adhesion force increased with increasing DBA content (fig. S5, A and B). In the flat-plate electrode tests, PBA35 generated adhesion strength of 77.8 kPa at 8 MV m⁻¹, which was 12.5 times greater than the 6.25 kPa strength for P7670. With increasing electric field strength, the adhesive force is rapidly enhanced because of the field-dependent DBA within the DE. Beyond 2 MV m⁻¹, PBA migration approaches saturation, whereupon adhesive force plateaus with diminishing growth rate. We demonstrated the ability of PBA35 to lift a 10-kg weight, suspend it, and then detach rapidly (Fig. 3E and movie S2). When the metal grid electrode was used to eliminate vacuum interference, PBA35 achieved an adhesion strength of 31.75 kPa at 14 MV m⁻¹, which was 488 times higher than P7670 (0.065 kPa). Adhesive force is amplified with increasing electric field strength. The reduced effective contact area of mesh electrodes relative to planar electrodes, combined with DBA migration–induced creep in PBA35, delays the saturation of DBA migration. We also showcased PBA35’s capacity to suspend and release a 3-kg load swiftly (Fig. 3F and movie S3). In addition, we evaluated the decay of electroadhesion. The electroadhesive force was normalized to quantify its degradation. Under a square-wave field of 6 MV m⁻¹ with cyclic loading at 1, 10, and 100 Hz for 1000 to 5000 s, the electroadhesive force showed a maximum reduction of 16% (fig. S6).

Compared with the electrostatic adsorption structures of conventional DEs, PBA demonstrated greatly enhanced electroadhesion across both smooth and rough surfaces (Fig. 3G). PBA also showed high adhesion to irregular engineering surfaces. Contact between the adhered substrate and PBA35 triggers DBA migration toward the interface. This induces localized creep and adhesive force at the PBA35-substrate contact interface. Therefore, we developed an adsorption device that grasped uneven metal sheets, meshes, and wires stably, and enabled rapid detachment (Fig. 3H and movies S4 to S6). PBA35 also achieved a higher electroadhesion strength than P7670 on conductive surfaces. This substantial adhesive capability was unattainable with previous silicone systems, and PBA35 also conforms to rough surfaces to achieve adhesion under applied electric fields. This electroadhesion technology overcomes existing constraints while expanding applicability, delivering innovative solutions for next-generation adhesion systems in robotics, industrial manufacturing, and related sectors.

EWOD properties

Next, we investigated the EWOD effect in the EIEDE. As Fig. 4A shows, upon application of a voltage, the droplet’s contact angle changed markedly, indicating that the material’s surface wettability had been modulated extensively by the voltage. The contact angle variation is attributed to DBA migration within the EIEDE, which altered the surface’s energy and hydrophilicity/hydrophobicity (movie S7). We measured contact angle variations in P7670, PBA5, PBA15, PBA25, and PBA35 under various voltages (Fig. 4B and fig. S7A). As the voltage increased, DBA migrated toward the anode, elevating interfacial energy at the membrane surface through increased DBA concentration. This cascade transformed the originally hydrophobic surface into a hydrophilic state. All samples exhibited notable contact angle reductions, with PBA35 showing the most pronounced change, which confirmed its superior electrowetting. Figure 4C summarizes these advantages via comparison with prior works. As the field increased from 0 to 1.5 kV, the contact angles for all DE systems declined in a graded manner, with PBA35 showing the steepest drop (from 83.15° to 9.92° at 1.5 kV, a 73.23° reduction) and achieving a superhydrophilic state (Fig. 4D and fig. S7B). Compared with conventional DEs, PBA demonstrates substantially enhanced saturation contact angles and a markedly expanded dynamically tunable contact angle range. Compared with other reported systems, the proposed EIEDE has a lower saturation contact angle, thus underscoring its high application potential. We also examined the effects of prestretching and thickness on the EIEDE’s EWOD performance, with the thicker membrane without prestretching exhibiting smaller droplet contact angles under the same applied voltage because of its higher DBA content per unit area (figs. S8 and S9). The decay in EWOD performance was assessed. When subjected to cyclic loading (1 to 100 Hz, 1000 to 5000 s) under a 1000-V square wave, the contact angle (normalized) revealed a maximum reduction of 16.4% (fig. S10). We additionally measured the rolling-angle variations of 50-, 100-, and 200-μl droplets under different applied voltages. All three droplet volumes exhibited increased rolling angles, confirming the excellent anchoring of our material (fig. S11). To assess whether substantial DBA precipitation might contaminate the droplets during voltage application, we conducted a quantitative analysis of DBA leaching. The maximum detected amount was 0.5% (fig. S12), which we consider negligible.

Fig. 4. EWOD behavior of the anode.

(A) Electrowetting effect demonstration of PBA35 under a 2000-V field. (B) Contact angle variation as a function of the electric-field intensity for electrowetting characterization. (C) Reported EWOD (24–36) characteristics versus this work. (D) Comparative contact angle analysis between P7670 and DBA35 under identical field conditions. (E) Adsorption and transport capabilities for metallic objects via droplet manipulation. (F) Droplet translocation and deformation behavior.

Using the same electrowetting principle, we also developed a droplet-adsorption device. As depicted in Fig. 4E and movie S8, PBA35’s strong hydrophilicity caused droplets to adhere firmly to the membrane surface under an applied voltage. Continuous voltage facilitates droplet transport, whereas subsequent voltage termination triggers surface reversion to hydrophobicity. This induces droplet release, enabling stable operations including adsorption, transport, and release. Using PBA35’s outstanding electrowetting, we designed patterned electrodes to control both droplet morphology and translation (Fig. 4F and movie S9). Sequential energization of cathodes with varied shapes (e.g., triangular and square) induced localized DBA migration, creating corresponding hydrophilic regions. Droplets migrated from hydrophobic areas to these newly hydrophilic regions because of surface tension gradients. Consequently, manipulating the surface tension gradient near a droplet can drive its motion. Moreover, sufficiently strong gradients can induce droplet deformation, inducing a contour and size matching the activated electrode’s geometry. Furthermore, to investigate whether the droplet can fully cover the entire electrode area, we analyzed the spreading resolution of droplets on electrodes with three geometries—circular, triangular, and square. The results show that spreading deteriorates in regions where two edges form an angle, with a maximum error of up to 10%. In contrast, spreading along straight boundaries exhibits much smaller deviations, remaining below 5% (fig. S13). This behavior arises from the intrinsic surface tension of the droplet. Among the three geometries, circular electrodes yield the most uniform spreading, followed by triangular and square electrodes.

The experimental results show distinct electrowetting in PBA35 that depended on the electric-field polarity. Droplets showed strong electrowetting when connected to the anode but displayed an antielectrowetting effect when contacting the cathode. First, non–contact-mode operations maintained synchronized motion between the moving electrode and droplet through sustained electrical activation while preserving critical spacing (Fig. 5A and movie S10). The underlying mechanism involves an induced voltage dominating under noncontact conditions between the cathode, droplet, and PBA35. Driven by the voltage, DBA molecules migrate toward the cathode periphery, inducing localized hydrophobicity in the thin film. The resultant surface tension gradient propels droplets toward hydrophilic zones, exhibiting directional cathode avoidance. Critically, weak orientational polarization strength restricts displacement, necessitating continuous electrode advancement to sustain motion. When the cathode contacted both PBA35 and the droplet simultaneously, through an optimized anode electrode geometry design (fig. S14A) used to constrain the droplet trajectories, a characteristic repulsion process involving spreading-contraction-detachment was achieved under square-wave excitation (Fig. 5B and movie S11). This mechanism occurs when the cathode, droplet, and PBA35 establish direct contact. Upon voltage application, direct electroinjection of negative charges into DBA molecules within the material triggers intensified migration, inducing localized hydrophobicity. This provokes pronounced hydrophilic-hydrophobic contrast near the droplet, generating substantially enhanced motion driven by surface-tension gradients. Such a mechanism enables directional spreading of cathode avoidance, wherein strong surface tension induces rapid droplet elongation. This configuration achieves substantially longer displacement trajectories than the non–contact mode. Subsequent voltage removal restores surface energy to baseline hydrophobicity/hydrophilicity states, enabling droplet reconfiguration and accomplishing net displacement away from the cathode.

Fig. 5. EWOD behavior of the cathode.

(A) Droplet actuation demonstration through electrode contact combined with power cycling. (B) Droplet motion visualization during intermittent electrical activation via electrode contact. (C) Displacement characteristics of variable-volume droplets under an electric-field intensity of 4 MV m⁻¹. (D) Electric field–dependent displacement characteristics of 100-μl droplets (droplet bifurcation was observed postdeactivation at excessive field strengths). (E) Schematic illustration and (F) experimental demonstration of programmable droplet division. (G) Demonstration of multiprobe detection with anchoring and movement of multiple droplets.

Systematic investigations showed positive correlations between the forward displacement and the droplet volume under constant voltage (Fig. 5, C and D), which were attributed to the enhanced spreading length and subsequent recovery distance of larger droplets. Two critical thresholds were identified: a lower critical voltage, below which the droplet exhibits a small spreading area that prevents droplet motion, and an upper threshold, above which the droplet splits into residual and mobile portions. Using the antielectrowetting effects, we developed a programmable droplet-splitting technique. Cathodic electron injection induced directional DBA migration toward the anode, creating enhanced hydrophobicity at the cathode-contact regions that propelled the droplet away from the cathode. Through synergistic optimization of the anode geometries (fig. S14) and using custom cutting equipment configurations, we performed precise double-splitting, triple-splitting, and quarter-splitting of droplets (Fig. 5, E and F, and movies S12 to S14). The results indicate that splitting is most effective when dividing the droplet into two parts. This is because, during multidroplet splitting, the droplet’s intrinsic surface tension adversely affects the splitting performance. By leveraging EWOD-induced droplet deformation, transport, splitting, and anchoring, we designed a demonstration of multiprobe detection with anchoring and movement of multiple droplets (Fig. 5G and movie S15). After splitting, the first droplet was driven to the right, whereas the second droplet was directed to the left. Subsequently, the four resulting smaller droplets were simultaneously deformed and anchored onto probes 1 to 4. This technique achieves multidroplet, multiprobe detection and increases the effective sensing area of each probe, offering an innovative concept for future microfluidic sensing platforms. These unique EIEDE-based EWOD characteristics enable preset-path droplet division through controlled electric-field gradient manipulation, providing strategies for advanced microfluidic control and applications. Our work demonstrates polarity-switchable EWOD responses to opposite electrode polarities, paving the way for innovative microfluidics design paradigms.

DISCUSSION

We have proposed an EIEDE and investigated its actuation, electroadhesion, EWOD, and droplet actuation properties. Specifically, the EIEDE demonstrated enhanced intrinsic electromechanical conversion and actuation capabilities, with superior electroactive intelligent interfacial properties. In terms of electromechanical conversion and actuation, it achieved a large area strain of 70% at 55 MV m⁻¹, which is 2.5 times greater than the 28% for conventional P7670. Experiments comparing the electroadhesions of P7670 and the EIEDE on flat metal electrodes and metal meshes showed that the EIEDE attained an adhesion strength of 31.75 kPa on a metal mesh, which was 488 times higher than P7670 (0.065 kPa), and demonstrated effective electroadhesion to irregular conductive materials, including metal sheets, meshes, and wires. At the solid-liquid electroactive interface, the EIEDE’s EWOD produced a marked contact angle change from 83.15° to 9.92°, with the saturation angle and dynamic range notably exceeding those of existing systems. In addition, it exhibited distinct electrowetting behavior under positive and negative polarities, enabling droplet translation, shape modulation, capture-release, and cutting functions and providing a foundation for droplet manipulation applications. Overall, via a polar additive’s strategy, this work integrates large-strain actuation, superhigh electroadhesion, and multifunctional droplet actuation within a single DE material, with promising applications in complex-shaped metal adhesion, wall climbing robotics, and microfluidic droplet control. This study redefines fundamental understanding in DE research, establishing a creative paradigm for future investigations.

MATERIALS AND METHODS

Materials

Wacker Elastosil P7670 was purchased from Wacker Chemie AG (Munich, Germany). Ecoflex 0030 was purchased from Smooth-On Inc. (Pennsylvania, USA). DBA (CAS: 105-99-7, purity ≥ 99%), DOP (CAS: 117-81-7, purity ≥ 97%), DOTP (CAS: 6422-86-2, purity ≥ 98%), and DOS (CAS: 122-62-3, purity ≥ 97%) were all purchased from Macklin Inc. (Shanghai, China). All commercial reagents were used without further purification. Carbon paste (NyoGel 756G, Nye Lubricants) was used to form flexible electrodes. Wacker Elastosil P7670 served as the DE for the comparison studies.

Preparation of the DE

The DBA and P7670A components were mechanically homogenized at 30°C for 30 min on a heating stage. After introduction of P7670B, the mixture underwent additional mechanical blending at 30°C for 10 min. The solution was subsequently subjected to vacuum degassing for 10 min to eliminate any entrapped bubbles. The solution was then cast using a membrane coater on polyethylene terephthalate (PET) substrates at 30°C, followed by ambient curing for 12 hours. Other additive-modified membranes were fabricated by following the same fabrication protocol that was used for the PBA group (fig. S9A).

Dielectric property characterization

To measure their dielectric constants, the membrane samples were cut into circular shapes with a diameter of 42 mm. The tests were conducted using the Hioki LCR Meter IM3536 (Hioki E.E. Corporation, Japan) with an Agilent dielectric test fixture (16451B, Agilent Technologies Co. Ltd., USA) under a low electric field of 1 V/mm. We also performed relative dielectric constant measurements across a frequency range from 4 to 10⁶ Hz (fig. S9B).

Mechanical property characterization

The tensile stress-strain characteristics of the samples were measured using an electronic universal testing machine (CMT 4304, Senstest Co. Ltd., Shenzhen, China). Samples with dimensions of 25 mm by 100 mm were clamped and extended uniaxially at a rate of 5 mm min⁻¹ until failure occurred at each sample’s midpoint. Each sample was tested three times. The Young’s modulus was then calculated on the basis of the initial slope of the stress-strain curve at 5% strain (fig. S9C).

Fabrication of DEAs

To form the DEAs, 200-μm-thick pristine DE membranes were biaxially prestretched to produce the desired multiples using a membrane stretcher (fig. S9D). The stretched membrane was then secured between two annular rigid polymethyl methacrylate (PMMA) frames (inner diameter: 70 mm; outer diameter: 110 mm) to maintain the prestrain. After the membrane was clamped between acrylic plate, a release paper mask was applied to pattern 10-mm-diameter carbon paste electrodes on both surfaces of the membrane through the mask openings. Electrical leads were subsequently connected to the electrodes (fig. S9E).

Static actuation performance measurement

Application of high voltages induced both lateral expansion and thickness reduction in the DE membranes. The nominal electric field was calculated by dividing the applied voltage by the membrane’s initial thickness before actuation. The DEA’s operation was powered using a high-voltage system that comprised a signal generator (UTG1005A, Rigol Technologies Co. Ltd., Suzhou, China) coupled with a high-voltage amplifier (Model 615-10, Trek). The actuation dynamics were recorded using a digital camera (M50, Canon). The area strain quantification was performed through ImageJ analysis of captured images and was calculated as $\frac{a_1-a_0}{a_0}\times 100\%$, where $a_1$ and $a_0$ represent the actuated and initial areas, respectively. A stepwise electric-field escalation protocol was implemented, with each nominal field intensity being maintained for 30 s until membrane dielectric breakdown occurred (fig. S9E).

Dynamic actuation performance measurement

Dynamic actuation performance characterization was performed using prestretched DEAs. For the frequency response analysis, a fixed nominal electric field of 50 MV m⁻¹ was applied whereas the excitation frequencies were varied systematically from 1 to 40 Hz. For the cyclic durability testing, an identical field intensity (50 MV m⁻¹) was maintained under a constant 1-Hz frequency operation.

In situ Raman spectroscopy

The Raman spectroscopy system (Alpha300R, WItec) was acquired using a 532-nm excitation laser, a spectral window of 200 to 2000 cm⁻¹, and an integration time of 60 s per scan.

Subsequently, we constructed an in situ electrical-poling platform by placing a copper foil cathode beneath the film, a metallic mesh anode on top, and a 1-mm glass coverslip to prevent air breakdown (fig. S4A). Raman spectra were then collected at a fixed length of 25 μm beside the anode, both before and after applying the electric field.

Electroadhesion performance measurement

The electroadhesion characterization process used a vacuum chuck system with an effective adsorption area of 8 cm by 8 cm to perform membrane fixation. The metallic anode configuration used either a 4 cm–by–4 cm plate or a 100-mesh metallic grid. The membrane was powered using a high-voltage system that integrated a signal generator (UTG1005A) and a high-voltage amplifier (Model 615-10, Trek). Adhesion force quantification was conducted using a dynamometer to measure the detachment force required to separate the metal anode from the membrane surface (fig. S9F).

The movie presented in Fig. 3A was acquired using a Keyence VHX-6000 digital microscope (Keyence Corp.). In this setup, a copper foil cathode was placed beneath the film, whereas a metallic mesh anode was superimposed on top, enabling us to monitor the surface variations of the film under different electric field strengths.

The decay performance testing of electroadhesion force was conducted according to the following procedure. First, we measured the initial electroadhesion strength of the samples. The system was then actuated with a square-wave field of 6 MV m⁻¹ at 1, 10, and 100 Hz while applying a constant tensile displacement. The adhesion force was recorded after continuous stimulation for 1000, 2000, 3000, 4000, and 5000 s. Each test condition was repeated three times, and the averaged results were normalized for comparative analysis.

Electroadhesion

The electroadhesion system was operated using a dc power supply of 4 MV m⁻¹ that was integrated with a three-axis translational stage. The metallic components used included red copper sheets (surface area: 16 cm²), 100-mesh red copper grids (16 cm²), and 4-cm-long red copper wires. All test objects were subjected to artificial wrinkling to simulate nonplanar metallic surfaces and also ensure adhesion capability verification under geometrically imperfect conditions. Demonstration movies to document the adsorption processes were recorded using a digital camera (M50, Canon).

EWOD measurement

The cathode configuration used indium tin oxide (ITO) glass substrates (resistance: 70 to 100 Ω) obtained from Delta Technologies. The peripheral regions of the glass were electrically isolated using adhesive tape. The test materials were positioned uniformly on the ITO surface. Wettability characterization was conducted using a Krüss DSA100E drop shape analyzer with 50-μl test droplets. Electrical interfacing was established by immersing a wire electrode in the droplet. The electrical stimulation system comprised a signal generator (UTG1005A) paired with a Trek Model 615-10 high-voltage amplifier (fig. S9G).

The EWOD performance decay test was performed as a droplet of controlled volume was dispensed using a micropipette, and its initial contact angle was recorded. The sample was then subjected to continuous electrical conditioning by applying a 1000-V square-wave at frequencies of 1, 10, and 100 Hz for durations ranging from 1000 to 5000 s. Following this conditioning period, a dc voltage was applied, and the resulting steady-state contact angles were measured. This entire procedure was repeated three times for each combination of conditioning frequency and duration; the results were averaged and normalized for analysis.

To precisely quantify the alignment between the droplet and the underlying electrode pattern, we developed an image analysis protocol. Images of the electrode and the corresponding ink droplet were captured at the same location. For circular electrodes, pixel-level image recognition was used to identify the center coordinates and radius. The electrode boundary was divided into 360 equal segments, and from each point, a radial intensity gradient analysis was performed to detect the ink boundary and its offset relative to the electrode edge.

To assess potential additive leaching, the initial masses of the substrate, the cathode, and the EIEDE film were accurately weighed and recorded. A droplet of fixed volume was dispensed onto the film, and the system was actuated with 1000-V square-wave at 1, 10, and 100 Hz. Following stimulation for durations ranging from 100 to 5000 s and subsequent complete evaporation of the droplet under ambient conditions, the total mass of the substrate-cathode-film assembly was remeasured.

The camera was positioned directly above the substrate to capture top-view images of the droplets postcutting. The area of each segmented droplet was quantified using ImageJ software. The relative volume of each droplet was then determined by calculating the ratio of its individual area to the total area of all segments. We performed extensive experiments, repeating the double-splitting, triple-splitting, and quarter-splitting processes 100 times each to gather statistically significant data.

Supplementary Materials

The PDF file includes:

Figs. S1 to S15

Tables S1 to S4

Legends for movies S1 to S15

Other Supplementary Material for this manuscript includes the following:

Movies S1 to S15