Biomimetic and Gel‐Mediated Tissue Adhesive Structures for Skin Conformal and Multifunctional Bioelectronics

ABSTRACT

Epidermal electrodes are essential for wearable bioelectronics in health monitoring and rehabilitation. However, current options are limited: adhesives irritate skin, hydrogels dry out, ultrathin films cannot deliver high currents, and sharp microstructures concentrate current, risking burns. Achieving repeatable gel‐mediated adhesion with uniform current distribution on curved, moving skin remains a major challenge. In this study, we present biomimetic electrodes using carbon nanotube‐silicone elastomer (CNT‐elastomer) composites patterned with octopus‐inspired suction units and polyethylenimine (PEIE)‐induced microwrinkles. With an interfacing electrolyte gel, these electrodes achieve strong, repeatable skin adhesion of up to 77 ± 5 J m⁻² in the normal direction and maintain performance over 100 reapplication cycles. Charge‐transfer resistance across the skin is reduced by 20‐fold, establishing low‐impedance interfaces. The electrodes record electrophysiological signals with clinical‐grade fidelity and deliver surface functional electrical stimulation with 50% higher efficiency in large muscle groups. Manufactured through facile fabrication methods of mold casting and leaving no skin residue, the electrodes exhibit skin‐matching mechanical compliance and efficient charge transfer, advancing comfortable and durable bioelectronic interfaces for rehabilitation, assistive neurotechnology, and next‐generation healthcare devices.

Inspired by octopus suction structures, a biomimetic epidermal electrode is developed as a durable, skin‐conformal platform for next‐generation wearable and therapeutic bioelectronics. The electrodes exhibit skin‐matching mechanical properties and achieve strong, reusable adhesion on wet skin, outperforming commercial electrodes by maintaining adhesion over 100 cycles and improving the efficiency of delivered muscle stimulation by up to 50%.

1. Introduction

Epidermal electrodes are a key component of wearable bioelectronics, providing a noninvasive interface between devices and the human body. These systems enable critical healthcare applications ranging from monitoring electrophysiological signals for personalized diagnostics, early disease detection, and prosthetic control [1, 2, 3] to delivering electrical stimulation for wound healing, muscle rehabilitation, and modulation of physiological functions [4, 5, 6]. Because these applications require frequent and long‐term use, establishing repeatable and conformal skin contact is essential for high‐quality signal acquisition and reliable, efficient stimulation [7]. Specifically, the curved geometry of the skin and dynamic body motion necessitates interfacing materials that are intrinsically soft, stretchable, and capable of self‐adhesion.

Recent advances in materials science have introduced a range of soft and flexible solutions for bioelectronic interfaces, including conductive polymers, hydrogels, elastomers, composites, carbon‐based structures, conductive textiles, and hybrid systems [8, 9, 10, 11, 12, 13]. Elastomeric composites such as low‐modulus silicones [14, 15, 16], polydimethylsiloxane (PDMS) [17, 18, 19], and modified polyesters [20, 21, 22] have been widely adopted owing to their mechanical robustness, repeated‐use durability, and elastic modulus comparable to that of the epidermis.

In addition to softness, several skin‐adhesion strategies have been explored. Conventional devices rely on external fixtures, glues, or applied tension that compromises user comfort [23, 24]. Other approaches employ chemical adhesives, including the clinical standard silver–silver chloride (Ag/AgCl) electrodes laminated with acrylamide‐based adhesives [25], covalent or non‐covalent bonding chemistries [26], mussel‐inspired catechol groups [27], or stimuli‐triggered tissue adhesives [28]. While effective, these methods are limited to single use, may leave residues, and can cause skin irritation. As an alternative, elastomer‐based adhesive architectures that rely on physical mechanisms have emerged, enabling residue‐free, repeatable, and long‐term conformal interfacing. For example, tattoo electrodes achieve conformal contact through water‐activated hydrogen bonding or van der Waals interactions [29]. Although suitable for sensing applications and low‐current haptic stimulation (10¹ mA) [30], tattoo electrodes cannot deliver the higher current amplitudes (up to 10² mA) required for rehabilitation of large muscle groups due to their ultrathin, high‐perimeter metallic patterns [31].

In parallel, bioinspired adhesive structures have been developed by mimicking multiscale architectures found in gecko feet, endoparasite proboscises, beetle pads, tree‐frog pads, slug footpads, and octopus suction cups [32]. These biomimetic designs utilize high‐aspect ratio geometries to establish skin adhesion via van der Waals or capillary forces. However, structures such as mushroom‐shaped micropillars introduce sharp corners and microscale features that concentrate electric current, limiting their use for muscle stimulations where localized current hotspots can cause pain and tissue damage [33, 34, 35, 36]. Thus, an interfacing gel layer is required to mediate bidirectional electrical signals across the skin. The presence of mobile ions in electrode gels redistributes current, reduces local current density, and remains the clinical standard for electrical muscle stimulation with metal electrodes [35, 37].

To achieve repeatable adhesion in liquid environments of water and oil, several biomimetic elastomers have been reported. Octopus cup‐inspired structures have been adapted as adhesive patches for hairy skin [38], for underwater and oil‐immersed substrates using polyurethane‐acrylate (sPUA) polymers [39], and for organ adhesion using PDMS [40]. Additional strategies include treefrog pad‐inspired microstructures [41], nanoscale non‐close‐packed geometries [42], and self‐healing chemistry using 4,4’‐methylenebis(phenyl urea) (MPU) and isophorone bisurea (IU) [43]. Elastomeric composites incorporating cellulose nanofibers (CNFs) and PDMS have been developed for drug delivery [44, 45] and strain sensing [46], while conductive composites with carbon black (CB) nanoparticles [46], reduced graphene oxide (rGO) [41, 47], and Ag/AgCl embedded in low‐modulus silicones and PDMS [48] have been developed for electrophysiological applications such as surface electrocardiogram (sECG) and surface electromyography (sEMG) sensing. These biomimetic elastomers demonstrate robust and repeatable adhesion under flowing water or full immersion, highlighting the potential of combining biomimetic architectures with an electrolyte gel layer to achieve both repeatable adhesion and uniform current redistribution. However, as these systems are primarily designed for tissue adhesion, drug delivery, or strain and electrophysiological sensing in air or water environments, there remains a need to adopt biomimetic elastomers for gel‐mediated bioelectronic interfaces.

Here, we present a multi‐walled carbon nanotube (MWCNT)‐embedded silicone elastomer electrode that establishes a biocompatible, skin‐conformal interface, enabling repeatable adhesion in electrolyte gel environments to ensure safe and effective current transfer during both sEMG recording and electrical muscle stimulation. By integrating octopus‐inspired suction geometries with a surface coating of microwrinkles, we develop scalable, elastomeric, adhesive, low‐impedance (SEAL) electrodes to be applied with electrolyte gel for the bioelectronic interface. The SEAL electrodes exhibit skin‐matching properties of stretchability and bending softness, demonstrate a normal adhesive energy of 78.96 J m⁻² in gel environments comparable to clinical standards, and maintain skin conformity after 100 repeated cycles. When applied to skin, the SEAL electrodes achieve low interfacial impedance, improving stimulation efficiency by 50% compared to conventional stainless‐steel fabric‐coated hydrogel electrodes, enabling comfortable and conformal interfaces for effective muscle stimulation demonstrated on the bicep, forearm, and face. Fabricated through a facile molding and casting process using 3D‐printed templates, the SEAL electrodes form conformal, reliable, and non‐irritant skin contact without drying out, losing adhesion, or leaving residues, making them advantageous for long‐term and repeatable use in rehabilitation applications.

2. Results and Discussion

2.1. Design of a Skin‐Conformal Biomimetic Structure for Bidirectional Bioelectronic Communications

To support high stimulation currents up to 10² mA, SEAL electrodes integrate octopus‐inspired suction geometries with a microwrinkled surface on MWCNT‐elastomer composites, as illustrated in Figure 1a. When applied on skin with an interfacing electrolyte gel, the suction‐based adhesion proceeds in three stages: first, placement of the SEAL electrode with water or gel as the interfacing medium; second, application of a preload that compresses the structures, decreases inner cavity volume, and expels the interfacing liquid; third, release of the preload, during which partial elastomer recovery is balanced by the pressure difference within the cavity. Sustained adhesion requires that a sufficient fraction of the suction units remain sealed, as individual units function independently and local failure does not immediately compromise the overall skin adhesion. On soft substrates such as skin, the pressure difference induces a cupping effect that sustains adhesion [44, 45]. Because individual suction units operate independently, local failure does not immediately compromise overall attachment, enabling a stable electrical interface for both electrophysiological sensing and electrical stimulation.

FIGURE 1.

(a) Schematic illustration of SEAL electrodes and their scalable, facile fabrication, elastomeric skin‐matching softness, adhesion mechanism in wet or gel environments, and low‐impedance interface for sensing and stimulation. The electrodes are fabricated by casting MWCNT‐embedded silicone in 3D‐printed molds, forming a skin‐compliant conductor. Under preload, the structure deforms to expel liquid and reduce inner chamber volume; upon release, a pressure drop sustains suction‐based, repeatable adhesion. (b) Schematic rendering of SEAL electrode unit cells with octopus‐inspired suction cup geometries arranged in hexagonal packing. The structure is defined by orifice diameters (D_out , D_in ) and protuberance radius (r_pro ), parametrized by cross‐sectional variables: orifice wall angle (θ _ori ), protuberance tip height to substrate (h_pro ), and aspect ratio scale of the elliptical axis of protuberance (s_pro ). (c) Optical microscope image of a cast SEAL electrode, showing feature uniformity after demolding. (d) Finite‐element analysis of volumetric changes as a function of h_pro , and s_pro , with the optimum design region (dotted white line). (e) Demonstration of SEAL electrodes adhered on skin with lead wires attached during limb motion.

As the proposed adhesion mechanism is based on resultant pressure differential within structural cavities, the adhesive performance will be governed by the change in cavity volume during preload and release, with the extent of volume reduction and subsequent recovery being determined by the geometry of each suction unit. In the presence of an interfacing liquid, adhesion to the substrate is generated by each suction unit, consisting of an orifice and a central protuberance (Figure 1b,c). The orifice, defined by outer and inner diameters prior to preload application (D_out and D_in ), encloses the cavity, while the protuberance, characterized by radius before deformation r_pro , occupies the central chamber as a half‐dome. During preload, outward deformation of the orifice walls and increased contact at the protuberance tip with the substrate both reduce the effective cavity volume. Upon preload release, elastic recovery of the material attempts to restore the original geometry, but sealing at the orifice in the presence of a liquid medium preserves the reduced internal pressure, therefore generating adhesion. Geometry further constrains seal stability through two effects: increasing orifice diameters lengthens the sealing perimeter, introducing a longer pathway for potential leakage, while increasing protuberance radius reduces the initial cavity volume, which limits the achieved pressure differential under preload, resulting in a weakened suction susceptible to failure during elastic recovery. The total adhesion stress in liquid environments is described by Equation (1): [44]

(1)

where P ₀ is the pressure difference established in each unit upon preload release, γ is the surface tension of the interfacing liquid, θ₁ and θ₂ are the contact angles of the liquid on CNT‐elastomer and skin, respectively, h is the liquid gap thickness, and n is the number of suction units. , , and r′′ denote the equilibrium inner orifice diameter, outer orifice diameter, and radius of the contact area of protuberance with substrate after deformation and preload removal. The first term corresponds to suction stress proportional to and P ₀; the second term represents Laplace pressure from capillary stress within the chamber; the third term accounts for stress due to surface tension along the cavity perimeter. These parameters govern deformation‐dependent cavity volume and sealing conditions, which determine the resulting pressure differential and thus the overall adhesive strength.

Qualitative analysis of Equation (1) highlights the sensitivity of adhesion to geometric parameters. Because all terms scale linearly with n, efficient unit packing enhances adhesion, such as hexagonal arrays with center‐to‐center unit spacing of 2.5 D_out (Figure 1b). The capillary terms in Equation (1) scale proportionally to , though gains diminish as sealing becomes limited by surface roughness and the number of units n decreases as D_out increases. Increasing strengthens the suction contribution but reduces the capillary contribution, with theoretical maxima achieved at infinitesimal orifice wall thickness (Equation S1), however, thin orifices are prone to air ingress and seal failure. Larger protuberances improve adhesion by increasing r′′ in the second term of Equation (1) but are limited by the spatial constraint . These interdependent effects complicate the decoupling of individual geometric contributions for analytical optimization; therefore, computational modeling is used to inspect how coupled parameters (, , and r′′) govern adhesion.

Finite element modeling (FEM) was used to identify optimal values of , , and r′′ by parameterizing the structure through cross‐sectional variables θ _ori , h_pro , and s_pro (Figure 1b and Figure S1). θ _ori defines the orifice wall angle and governs diameters under deformation and , h_pro defines the initial protuberance‐substrate gap and controls cavity height and volume change during preload, and s_pro defines the vertical scaling of the protuberance, determining the resulting contact radius after deformation r′′. By expressing the geometry in terms of θ _ori , h_pro , and s_pro , the structural contributions to the adhesive performance can be isolated from that of material‐dependent effects.

Simulations of chamber volume changes under identical preload (Figure 1d and Figure S2) indicate that orifice angles. θ _ori between 0° and 40° maximize suction, protuberance‐substrate gaps h_pro of 0.2–0.8 mm enhance adhesion, and vertically shortened protuberances (s_pro =  0.4) outperform hemispherical or elongated half‐domes. Moreover, uniformly scaling the structures improves electric field distribution and reduces current concentration, as larger suction structures provide greater skin contact and longer current travel paths to minimize density hotspots (Figure S3). Within these optimized ranges, it is observed that SEAL electrodes demonstrate stable self‐adhesion under motion, maintaining attachment during wrist flexion and arm rotation while supporting both its own weight and lead connections (Figure 1a,e).

2.2. Geometric Optimization for Enhanced Skin Adhesion in Gel Environment

Based on the geometry‐dependent parameters θ _ori , h_pro , and s_pro identified by FEM, further validations are conducted in a gel environment to account for material effects and interfacial behavior in the intended environment during surface functional electrical stimulation (sFES) applications, where elastomer deformation, liquid retention, and surface interactions all influence adhesion. To further enhance on‐skin adhesion, suction structures are fabricated and tested for ultimate failure strength and interfacial energy under three modalities: peel, shear, and normal adhesion (Figure S4). For application as epidermal electrodes, these three modes are all crucial for maintaining conformal contact on curved and deforming skin, where bending, sliding, and lifting occur simultaneously during human body motion. Preload forces are calibrated to approximate human finger pressure (Video S1 and Figure S5). Adhesion performance is evaluated by varying parameters (θ _ori , h_pro , and s_pro ) while fixing other geometries of the electrode, applied on porcine skin strips with interfacing electrolyte gel.

Based on multimodal adhesion measurements across varying geometric parameters, θ _ori and s_pro are the dominant factors governing adhesion performance. Widening the orifice produced distinct failure mechanisms across modalities (Figure 2a–c). Furthermore, as θ _ori increases, peel and shear adhesion show a twofold improvement of ultimate strength from 0.6 ± 0.1 to 1.1 ± 0.1 kPa, and 0.6 ± 0.2 to 1.2 ± 0.2 kPa, respectively, while normal adhesion remains comparatively less sensitive due to more uniform load distribution across suction units. This improvement arises because θ _ori controls deformation mode during preload. At θ _ori  < 20°, the orifice walls are nearly perpendicular to the substrate, thus prone to buckling or wrinkling under compression instead of sliding outward, limiting cavity volume reduction and leading to leakage or rapid elastic recovery with minimal sustained suction. At larger angles, outward deformation is stabilized, and the cavity volume change increases, further raising the pressure differential and the resulting adhesion, although the lengthening seal perimeter also increases susceptibility to local lifts. Peel and shear adhesion continue to increase with θ _ori ; however, the optimal range is confined by fabrication: as θ _ori increases, the orifice walls form increasingly undercut profiles that interlock with the mold, causing the demolding yield to fall from 100% at θ _ori =  40° to around half at θ _ori =  50°. Thus, orifice angles (θ _ori ) of 30–40° are adopted as the range that provides the best performance while preserving structural fidelity, enabling stable sliding deformation under preload, minimizing leakage pathways, and maintaining sufficient pressure difference upon preload release, which are critical for strong adhesion and reliable electrode attachments during sFES therapy.

FIGURE 2.

(a–i) Adhesion performance of SEAL electrodes on porcine skin with electrolyte gel interfacing layer. Blue indicates ultimate adhesive strength in kPa, and green indicates work of adhesion in J m⁻². n = 3 electrodes; data presented as mean ± STD. (a–c) Peel, shear, and normal adhesion as a function of orifice wall angle (θ _ori ). (d–f) Peel, shear, and normal adhesion as a function of protuberance tip height to skin (h_pro ). (g–i) Peel, shear, and normal adhesion as a function of scale of protuberance elliptical axis (s_pro ). (j) Representative peel strength‐extension curves of optimized feature at 100% and 150% scale. (k) Work of adhesion across three directional modalities for optimized feature at 100% and 150% scale subjected to identical preload. n = 3 electrodes; data presented as mean ± STD. (l) Optical microscope image of the cross‐section of the optimized suction structure with θ _ori =  40°, h_pro =  0.3 mm, s_pro =  1, and uniform scaling by 150%.

Following optimization of orifice geometry, the effect of protuberance‐substrate gap on adhesion performance is evaluated. As shown in Figure 2d–f, the protuberance‐to‐substrate gap has a moderate influence on adhesion compared to the other geometric parameters. Adhesion energy decreases by 15%–17% across all modalities for h_pro  > 0.6 mm, attributed to the resultant larger separation distance reducing the volumetric change during preload release, and decreasing the effective contact radius r′′ of the protuberance under suction. This lowers the pressure difference and increases deformation stress on the orifice walls, as structures become more prone to venting at the seal edges. On the other hand, although smaller h_pro increases r′′, it results in excessive elastic energy stored during rebound of the protuberance structure, which promotes self‐peeling, exhibited by the lowering of normal adhesive energy (Figure 2f). Thus, an intermediate h_pro value of 0.2–0.4 mm balances the volume change for inducing the pressure difference and elastic rebound for the protuberance material to yield optimal skin‐adhesion and prevent premature detachment due to body deformation in sFES sessions.

Furthermore, adhesion performance is strongly influenced by protuberance aspect ratio across all modalities (Figure 2g–i). Increasing s_pro consistently enhances adhesion across all modalities, with maximum adhesion achieved at s_pro  = 1, corresponding to nearly spherical protuberances and adhesive strengths of 1.1 ± 0.1, 1.4 ± 0.2, and 7.1 ± 0.8 kPa for peel, shear, and normal modes, respectively. This improvement is attributed to more favorable volumetric deformation and increased effective contact radius r′′ under suction, both of which strengthen the suction and capillary contributions to skin adhesion. In contrast, below the threshold of s_pro  = 0.4, protuberances are vertically shortened with geometries that exhibit poor adhesion, partly due to fabrication‐induced surface roughness from the limited z‐axis resolution during additive manufacturing of the structure mold, which disrupts curvature smoothness and sealing capacity. Thus, spherical or near‐spherical protuberances enable optimal adhesion performance by maximizing pressure generation and maintaining smooth, stable contact interfaces with the skin.

In addition to optimizing each protuberance and orifice parameter in isolation, the impact of scaling suction units is examined, as larger units were shown to improve current uniformity (Figure S3). Increasing the unit geometry by 150% significantly enhances both adhesive strength and energy across modalities. Specifically, the work of adhesion increases by 148 ± 27% for peel, 32 ± 13% for shear, and 64 ± 12% for normal adhesion under identical preload (Figure 2k). This improvement is attributed to the increased contact area and greater cavity volume change, both increasing the resultant pressure difference after the preload is removed. However, failure behavior becomes more discrete, reflected by the sharp peaks in Figure 2j, as larger units detach individually rather than through a continuous separation of the surfaces. Despite this difference in behavior of adhesive failure mode, scaling up the structures provides a direct and effective method to enhance adhesion energy substantially without compromising attachment stability.

Based on the combined optimization of structural parameters and scaling effects, the optimal unit geometry for gel‐on‐skin systems is identified as θ _ori =  40°, h_pro =  0.3 mm, and s_pro =  1, corresponding to a spherical protuberance and microstructure scaled by 150% (Figure 2l). This configuration maximizes adhesion strength and energy while maintaining seal stability across all loading modalities, establishing a stable interface with skin in a gel environment of 37 ± 4 J m⁻² peel energy, 45 ± 4 J m⁻² for shear adhesion energy, and 77 ± 5 J m⁻² for normal adhesion energy. Achieving a high and multimodal adhesion performance enables stable, conformal skin interfaces under dynamic motion, which can sustain loading and support reliable operation of bioelectronic devices in practical wearable applications.

2.3. Interfacial Texturing via PEIE for Multi‐substrate Adhesion

To further enhance interfacial contact between the electrode and skin in gel‐mediated environments, localized surface wrinkles are introduced to promote capillary bridge formation and strengthen suction‐based seal at the perimeter. Surface texturing is achieved by partially inhibiting the cross‐linking of silicone elastomers (Figure 3a) [49, 50, 51]. SEAL electrodes were fabricated from a platinum (Pt)‐cured room‐temperature vulcanizing silicone (Si), consisting of vinyl‐functional groups in part A and hydrosiloxane groups with Pt catalyst in part B [52, 53, 54]. In the absence of additives, vinyl groups undergo hydrosilylation with the hydrosiloxane cross‐linkers, forming a network of Si‐based elastomers. When polyethylenimine (PEIE) is added, localized inhibition of cross‐linking occurs through two mechanisms: coordination of amine lone pairs with platinum, which forms amine‐Pt complexes that reduce catalyst availability for hydrosilylation, and consumption of Si─H bonds by hydroxyl and amine groups through dehydrogenative silylation, which reduces available cross‐linking sites.

FIGURE 3.

(a) Schematic of the chemical mechanism and synthesis route for generating surface wrinkles on platinum‐cured silicone elastomers by incorporating PEIE to partially inhibit curing. (b) Optical microscope images of the orifice surface before and after spin‐coating PEIE‐MWCNT‐elastomer to introduce microwrinkles. (c–e) Peel, shear, and normal adhesion on skin substrates. Yellow: smooth MWCNT‐elastomer (no suction structures, no wrinkles); light blue: MWCNT‐elastomer with suction structures (no microwrinkles); dark blue: MWCNT‐elastomer with suction structures and microwrinkles; pink: commercial electrodes with acrylamide‐based hydrogel adhesives. n = 3 electrodes; data presented as mean ± STD. (f, g) Normal and peel adhesion energy with various interfacing liquid viscosity on compliant skin and rigid glass substrates. Yellow: air environment; light blue: water environment; blue: interfacing liquid environment with electrolyte gel diluted 1:1 by weight with water; dark blue: gel environment. n = 3 electrodes; data presented as mean ± STD.

This additive‐based approach provides a simple and scalable method to introduce interfacial microtexturing for promoting the formation of capillary bridges to enhance skin adhesion. The PEIE concentration was selected within an effective window bounded by curing behavior, as concentrations above 2 wt.% excessively delayed cross‐linking and induced MWCNT agglomeration (Figure S6), while reducing the concentration to 1 wt.% produced no discernible change in wrinkle morphology or adhesion.

Following PEIE modification, the previously smooth elastomer surfaces develop irregular microscale ridge‐and‐valley wrinkles on the skin‐contacting regions (Figure 3b). SEM imaging shows the flake‐like fine structure of these ridge topologies (Figure S7). These textured surfaces improve skin adhesion when interfaced with water (Figure 3c–e; Table 1). This improvement arises from the heterogeneous cross‐linking process, which creates mechanically softened regions that form microwrinkles that increase the effective surface area and promote capillary bridge formation at the interface. The resulting enhancement in capillary interactions strengthens suction sealing by improving liquid retention and interfacial conformity. As a result, these microwrinkle‐induced interfaces enable more robust and reliable skin adhesion in non‐dry environments to support consistent device attachment in applications where liquid‐mediated contact is required for bioelectronic signal deliveries.

TABLE 1.

Adhesion of SEAL Electrodes and Commercial Electrodes for Muscle Stimulation.

Electrode Type	Skin‐Interfacing Medium	Normal Ultimate Strength [kPa]	Normal Work of Adhesion [J m−2]	Tensile Ultimate Strength [kPa]	Tensile Work of Adhesion [J m−2]	Peel Ultimate Strength [kPa]	Peel Work of Adhesion [J m−2]
SEAL	Dry	1.0 ± 0.9	8 ± 2	0.5 ± 0.1	4 ± 2	0.08 ± 0.06	1.1 ± 0.2
Commercial	Dry	10 ± 2	21 ± 3	7.0 ± 0.8	41 ± 3	0.3 ± 0.1	1.4 ± 0.2
SEAL	Gel	11 ± 2	79 ± 6	7 ± 2	43 ± 3	4.4 ± 0.7	39 ± 3
Commercial	Gel	0.3 ± 0.1	1.8 ± 0.9	0.3 ± 0.2	3 ± 1	0.4 ± 0.2	1.5 ± 0.5

For normal adhesion in both water and gel environments, wrinkled surfaces show slight improvements, indicating that adhesive strength in this mode is primarily governed by the suction‐based mechanism. In contrast, under shear and peel loading on gelled skin, PEIE‐induced wrinkles resulted in more pronounced improvements in adhesion performance. Samples lacking either suction structures or surface wrinkles show negligible adhesion under aqueous and gel conditions, confirming that suction is the primary contributor while interfacial texturing provides a secondary enhancement. This improvement under shear and peel arises because microwrinkles stabilize the negative pressure within the suction chamber by reducing water leakage. The increased surface area promotes capillary bridge formation, preserving thin liquid pockets and suppressing interfacial slip. Thus, while suction is the mechanism for baseline adhesion, microwrinkles selectively reinforce performance in loading modes most sensitive to interfacial failure.

Across all three adhesion modes, wrinkled SEAL electrodes in gel environments achieve adhesion performance comparable to that of commercial acrylamide‐based adhesives for sFES applications on dry skin (Figure 3c–e; Table 1, and Figure S8). This indicates that combining the suction and capillary mechanism will effectively improve the typically reduced adhesion in hydrated conditions. However, the relative contribution of microwrinkles depends on the properties of the interfacing liquid. In viscous gel environments, adhesion is dominated by the sealing effect of the electrolyte gel, limiting additional benefits to introducing microwrinkles. Rheological measurements (Figure S9) confirm that higher‐viscosity interfacing liquids sustain stronger adhesion by slowing leakage and maintaining negative pressure after preload removal, whereas lower‐viscosity liquids, such as water, leak and evaporate readily, dissipating the pressure differential within the suction chambers of the SEAL structure (Figure 3f,g). These results highlight that optimal adhesion arises from the combined tuning of surface structure and liquid properties, rather than either factor alone.

Lastly, it is shown that substrate compliance further impacts the adhesive performance across all three modes in liquid environments of water, diluted gel, electrolyte gel, and dry interfaces. Adhesion on skin is consistently higher than on rigid glass surfaces (Figure 3f,g and Figure S10), due to the ability of compliant substrates, such as the skin, to deform under negative pressure and maintain conformal contact, thereby reducing stress concentrations that can initiate seal failure. In contrast, glass substrates restrict deformation, allowing gaps to persist and leading to a premature loss of suction. This trend persists across both gel‐assisted interface and dry conditions, although the magnitude of improvement varies depending on the availability of liquid‐mediated sealing. Therefore, substrate mechanics is a non‐negligible factor in achieving gelled skin‐adhesion.

In conclusion, across multiple liquid environments and substrate conditions, PEIE‐induced surface wrinkles provide modest but consistent improvements in adhesion. When combined with suction‐based geometries, this dual strategy enabled SEAL electrodes to achieve adhesion that matches or outperforms commercial single‐use hydrogel electrodes commonly used in sFES for muscle activation. The capability supports reliable, conformal attachment in hydrated and dynamically loaded conditions, essential for stable bioelectronic interfacing and sustained stimulation performance.

2.4. Durability and Mechanical Properties of the SEAL Electrodes

The mechanical durability and adhesion repeatability of SEAL electrodes with optimized geometry and interfacial design are evaluated under benchtop, skin‐relevant application conditions. Across 100 application cycles in gel environments, SEAL electrodes retain high adhesive strength for the first 30 cycles, with a 20%–40% decrease after 100 cycles across peel, shear, and normal modes (Figure 4a–c). In contrast, the adhesive strength of standard commercial hydrogel electrodes on dry skin shows rapid degradation, losing up to 90% of adhesion after 10–15 cycles (Figure S11). This sustained performance of SEAL electrodes is attributed to the suction‐based adhesion mechanism, which is less sensitive to surface contamination such as debris or skin oils in practical sFES applications, where repeatable adhesion without reliance on consumable adhesive layers is advantageous.

FIGURE 4.

(a–f) Cyclic peel, shear, and normal adhesive performance of electrodes on porcine skin after application of identical preloads. Green: SEAL electrodes (MWCNT‐elastomer with suction structures and PEIE‐induced surface wrinkles) with interfacing electrolyte gel on skin, with applied electrolyte gel completely rinsed off from the electrode and skin in between cycles; pink: commercial electrodes with acrylamide‐based adhesives on dry skin. (g) Optical microscope images of the suction structure before and after cyclic testing, demonstrating localized damage of the protuberance wall. (h) Uniaxial stress–strain curves of the SEAL electrodes. (i) Flexural stiffness of the SEAL electrodes measured in bending; data presented as mean with shaded STD.

Adhesive energy dissipation shows similar trends, with SEAL electrodes maintaining stable performance compared to commercial electrodes with rapid loss of skin adhesion (Figure 4d–f). In peel mode (Figure 4d), SEAL electrodes show only a 10.6% decrease after 30 cycles and a 30.0% decrease after 100 cycles, whereas commercial hydrogel electrodes exhibit negligible performance, reaching a maximum peeling energy of 1.38 J m⁻² (3.5% of SEAL electrodes) and decreasing by 54.9% after only 6 cycles. This stability is attributed to the structural origin of energy dissipation in SEAL electrodes, where energy is absorbed through reversible orifice deformation and localized structural damage (Figure 4g), rather than irreversible degradation of a chemical adhesive layer such as acrylamide. Notably, even after partial damage to the orifices, SEAL electrodes maintain shear adhesion and energy (Figure 4b,e), as gel‐mediated sealing preserves negative pressure within the cavities during sliding. Ultimately, performance degradation of SEAL electrodes is primarily attributed to edge defects introduced during repeated gel cleansing, which compromised seal retention (Figure 4g). The optimized structure of SEAL electrodes maintains functional adhesion despite minor mechanical wear, with failure mode governed by peripheral defects rather than bulk degradation.

The intrinsic mechanical properties of SEAL electrodes further support their suitability for skin‐conformal applications (Figure 4h,i). Compared to smooth CNT‐elastomer films, structured SEAL electrodes exhibit higher Young's modulus and localized failure events due to geometric heterogeneity, while incorporation of 2 wt.% CNT further enhances elastic modulus and ultimate stress. Although CNT incorporation reduces strain at break from 471% to 370% (Figure 4h), this remains well above the failure strain of human skin (∼75%) [55], ensuring conformability. The Young's modulus of SEAL electrodes is 103 ± 8 kPa, substantially lower than that of human skin (4.6–20 MPa) [55], while bending stiffness 18.4 ± 0.8 N m⁻¹ also remains lower than that of skin (47.3–128.3 N m⁻¹) [56], enabling compliant deformation (Figure 4i). These properties are attributed to the combination of a soft elastomeric matrix and engineered microstructures to provide mechanical robustness without sacrificing flexibility. Thus, the optimized material‐structure of SEAL electrodes achieves a balance between durability and compliance required for skin interfacing.

In summary, SEAL electrodes exhibit robust, repeatable adhesion in gel‐mediated environments over 100 use cycles, while maintaining mechanical properties closely matched to human skin. Their suction‐based adhesion mechanism ensures durability across loading modalities without reliance on single‐use adhesives, and their intrinsic softness ensures conformal contact during human body motion.

2.5. Bioelectronic Signal Acquisition and Stimulation Performance of the SEAL Electrodes

After establishing stable adhesion and mechanical compliance to skin, the performance of SEAL electrodes in functional bioelectronic applications is next evaluated, where efficient electrical signal transmission across the human body is required. The performance of SEAL electrodes on the human arm is compared against the clinical standard electrode for sFES of stainless‐steel fabric electrodes coated with acrylamide‐based hydrogel. Electrochemical impedance spectroscopy shows that both systems exhibit similar overall impedance (10⁴–10⁶ Ω) (Figure 5a,b), but the underlying mechanisms for electrical current conduction differ substantially. At low frequencies (< 10¹ Hz), interfacial polarization dominates, while at mid to high frequencies, charge transfer and ion transport through the gel and tissue become the primary pathways. Thus, while bulk impedance appears comparable, differences in interfacial processes govern how effectively each electrode supports bioelectronic signal transmission.

FIGURE 5.

(a) Bode plots of impedance magnitude (circles; mean ± STD, n = 3 electrodes) and phase angle (dashed lines) at the electrode‐arm interface, showing that SEAL electrodes applied with electrolyte gel exhibit lower impedance than commercial hydrogel electrodes. (b) Nyquist impedance plots of SEAL electrodes with electrolyte gel and commercial electrodes (hollow circles: experimental data; solid curves: equivalent circuit fits). (c) Equivalent circuit model of the electrode–skin system. Variables: CPE denotes constant phase element, R denotes resistance, and Z_W denotes Warburg element; subscripts: CT denotes charge transfer, SS denotes stainless‐steel fabric of the commercial electrode, SE denotes SEAL electrodes, and SC denotes stratum corneum. Skin tissue equivalent circuit referenced from [57]. (d) ECG signals recorded using commercial and SEAL electrodes of identical surface area, displayed as frequency‐amplitude spectrograms. (e) Measurement configuration for sEMG recording on the forearm using SEAL electrodes. (f) sEMG signals of repeated fist motions measured with commercial electrodes and SEAL electrodes with electrolyte gel of identical surface area. (g) Signal‐to‐noise ratios of recorded sEMG signals (mean ± STD, n = 3 trials). (h) Torque output during electrical stimulation measured by a dynamometer (data displayed as mean with shaded regions showing STD, n = 3 trials), demonstrating robust muscle contraction induced by SEAL electrodes. (i) Resistance and facial expression intensity evaluated by pixel changes output from facial detection algorithms during facial stimulation, plotted as a function of applied sFES current.

To resolve these differences despite biological variability in tissue properties, equivalent circuit modelling and curve fitting are adopted to decouple interfacial and tissue contributions (Figure 5c; Table S1, and Figure S12). For hydrogel‐coated commercial sFES electrodes, the interface is characterized by a large charge‐transfer resistance of 240 kΩ between the stainless‐steel fabric and the hydrogel, producing a phase angle peak associated with capacitive polarization. In contrast, SEAL electrodes exhibit a 20‐fold reduction in charge‐transfer resistance. This improvement is attributed to the intrinsically larger effective surface area of the wrinkled CNT‐elastomer composite and the presence of a percolated CNT network, both of which increase the density of conductive pathways and shorten ionic transport distances at the electrode‐gel boundary. Both electrode systems share a biological‐tissue branch with a moderate resistance of ∼40 kΩ and broad dispersive behavior (constant phase element, or CPE n = 0.39) due to ionic conduction in dermis and hypodermis layers. By isolating these components, the equivalent circuit model reveals that SEAL electrodes achieve more efficient interfacial charge transfer, critical for reducing sEMG signal loss and improving stimulation efficiency during sFES.

These mechanistic differences are reflected in the phase response (Figure 5a). Commercial electrodes display a sharp phase angle peak at 10³ Hz, consistent with the near‐ideal capacitive interface (CPE n = 0.97) in series with high hydrogel resistance. In comparison, SEAL electrodes exhibit phase angles gradually declining from 10³–10⁴ Hz. Curve‐fitted models support a more dispersive double‐layer interface (CPE n = 0.73) combined with a diffusion‐controlled contribution captured by a Warburg‐type element. This behavior is attributed to ion transport dynamics across the CNT‐elastomer composite and the interfacing electrolyte gel. Notably, the gel used with SEAL electrodes exhibits a resistance two orders of magnitude lower than that of skin‐adhesive acrylamide hydrogels, facilitating more efficient ionic conduction and reducing polarization effects. Additionally, dielectric relaxation intrinsic to the CNT‐elastomer composite may further contribute to the phase decrease near 10³ Hz, where both conductivity and permittivity undergo transition (Figure S13). Together, these charge transport characteristics enable SEAL electrodes to maintain a stable, low‐polarization interface, supporting consistent sEMG signal acquisition and efficient electrical stimulation in practical applications.

Before delivering external stimulation signals in the milliamps range, SEAL electrodes are first evaluated for capturing reliable electrophysiological signals, which impose lower current amplitudes but are highly sensitive to interfacial stability. sECG and sEMG recordings confirm that SEAL electrodes with an electrolyte gel interface achieve signal quality comparable to clinical standards (Figure 5d–g), capturing high‐fidelity waveforms (Figure 5d), with an elevated baseline noise attributable to the presence of electrolyte gel (Figure 5g). Despite this, signal‐to‐noise (SNR) ratios remain sufficient for robust muscle activity detection, demonstrating the potential of SEAL electrodes in supporting continuous monitoring and wearable diagnostics.

Lastly, the performance is evaluated under active electrical stimulation, where higher currents and sustained contact are required at the bioelectronic interface. Muscle stimulation (sFES) delivered by SEAL electrodes on the bicep (Figure S14), forearm (Video S2), and face demonstrates strong and controllable actuation. As shown in Figure 5h, sFES delivered via SEAL electrodes induces a higher torque at equivalent applied current amplitudes, achieving up to 50% greater stimulation efficiency in large muscle groups (Figure 5h). This improvement is attributed to the reduced interfacial impedance and more uniform current distribution enabled by the optimized suction geometry and conductive composite interface. The gentle suction‐based adhesion with electrolyte gel ensures safe and comfortable stimulation, with transient redness due to cupping of the skin that resolves within 5 min after 120 min of continuous sFES delivery (Video S3). Beyond inducing motor activation of large muscle groups, SEAL electrodes enable facile application and precise control of facial expression (Figure 5i and Videos S4), where targeted stimulation of discrete facial muscle groups reproduced contractions of action units associated with the Ekman‐defined facial expression for disgust [58], including brow, nasal, and lip movements. These results demonstrate the ability of SEAL electrodes in providing both the milliampere‐stimulation current required for functional limb activation and the precision necessary for delicate facial modulation, highlighting their versatility for advanced bioelectronic therapies and human–machine interfacing.

3. Conclusion

In summary, we introduce a novel biomimetic conductive elastomer that achieves robust and repeatable gelled‐skin adhesion while maintaining skin‐like mechanical compliance. Guided by octopus‐inspired structures, we identified optimal unit geometries of the orifice and protuberance that maximized adhesion of gelled skin across modalities, reaching peel, shear, and normal adhesion energies of 37 ± 4, 45 ± 4, and 77 ± 5 J m⁻², respectively. PEIE‐induced microwrinkles further enhance interfacial performance by promoting capillary bridge formation and stabilizing fluid pockets, providing consistent improvements across hydrated conditions. Compared with standard commercial hydrogel sFES electrodes, which lose up to 95% of their adhesion within 15 cycles, SEAL electrodes retain strong adhesive performance over 100 use cycles, substantially improving durability and reusability.

Mechanically, SEAL electrodes exhibit intrinsic softness (Young's modulus 103 ± 8 kPa; stiffness 18.4 ± 0.8 N m⁻¹) well‐matched to skin, enabling conformal contact without compromising structural integrity throughout repeated use. When applied to the human body, SEAL electrodes deliver electrical stimulation comfortably and efficiently, achieving up to 50% higher muscle activation efficiency and enabling precise facial muscle control without irritation.

By integrating gelled‐skin adhesion, capillary‐enhanced interfacial texturing, and conductive elastomer design within a scalable platform, SEAL electrodes establish a fundamentally different adhesion paradigm for sFES and sEMG applications. The resultant interface overcomes key limitations in poor durability, reliance on consumable adhesives, and instability under hydrated conditions, while enabling high‐performance, reusable interfaces for high‐current stimulation. The advantages of SEAL electrodes demonstrate the potential for this approach to durable, skin‐conformal, and multifunctional bioelectronic interfaces, with broad implications for wearable stimulation, rehabilitation, and next‐generation human–machine interfaces.

4. Experimental Methods

4.1. Materials

Multiwalled carbon nanotubes (MWCNT, NC7000) were obtained from Nanocyl, silicone elastomers (Ecoflex 00–30) were obtained from Smooth‐On Inc., and silicone oil (viscosity 100 cSt) and 2‐Propanol (IPA) were obtained from Sigma Aldrich.

4.2. Fabrication of SEAL Electrodes

In a glass beaker, MWCNTs are added to IPA at a weight ratio of 1:100 (1 g of MWCNT per 100 g of IPA) and dispersed by microtip probe sonication at 50% amplitude (∼50 W) in a pulsed cycle of 4s on and 2s off for 60 min. Silicone oil (1.5 g per 1 g of MWCNT) is then added to the solution, followed by a second sonication under the same settings for 60 min. Part B of the silicone elastomer is added (25 g of Part B per 1 g of CNT), and the mixture is mechanically stirred with an overhead stirrer on a hot plate at 50°C until the IPA fully evaporates. After IPA removal, PEIE is incorporated at 2 wt.% relative to the combined weight of Part A and Part B, added into the Part B mixture, and mechanically stirred for 3 min. Part A is then added at a 1:1 ratio with Part B and mixed under a high‐shear overhead mixer for 3 min. The precursor mixture is poured into 3D‐printed molds and degassed in vacuum for 10 min. Electrical leads and wires are placed onto the mold surface, and the SEAL electrodes are formed after 4 h of cross‐linking at room temperature.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: smll74839‐sup‐0001‐SuppMat.docx.

Supporting File 2: smll74839‐sup‐0002‐VideoS1.mp4.

Supporting File 3: smll74839‐sup‐0003‐VideoS2.mp4.

Supporting File 4: smll74839‐sup‐0004‐VideoS3.mp4.

Supporting File 5: smll74839‐sup‐0005‐VideoS4.mp4.

Data Availability Statement

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