1. Introduction
Gecko-inspired dry adhesives exhibit considerable potential for applications in grippers [
1], product-line transport [
2], [
3], [
4], skin patch [
5], [
6], [
7], [
8], [
9], and climbing robots [
10], [
11], [
12] due to their innovative adhesion mechanisms [
13], [
14], [
15], [
16] and ingenious prototype design [
17], [
18]. In adhesion-based operations, close parallel contact is essential for effective gripping and transport. In particular, laboratory testing requires precise control of parallelism. However, non-parallel contact is unavoidable in engineering applications and significantly affects the stability and efficiency of the task. For example, non-parallel contact may lead to unstable grasping of fragile materials (liquid crystal display panels, Si wafers, etc.) in integrated circuit (IC) production lines, resulting in the risk of objects dropping and breaking.
Misalignment angles under non-parallel contact with a magnitude of just 0.1° can trigger peel-like detachment, reducing load sharing and decreasing adhesion strength to a flat surface by 50% [
19], [
20], [
21]. Hence, artificial dry adhesives with high theoretical adhesion values tend to exhibit low adhesion during application even on smooth surfaces [
22], [
23], [
24]. Thus, flat-flat adhesion tests require additional complex experimental efforts, such as charge-coupled device (CCD) observation systems and mechanical alignment devices, to ensure effective alignment of the object surface. Similarly, high-precision real-time detection systems are necessary to compensate for the influence of misalignments on industrial manipulation processes. Such methods, which rely on mechanical mechanisms and vision systems to achieve parallel contact, require real-time determination of the positional attitude of the adhesives and the target interface, and can hardly meet the demands of high-speed, fast-paced application scenarios, such as space operations and production line transport. Achieving stable attachment to non-parallel contacts using a simple process remains a significant challenge for gecko-inspired adhesive applications.
An insufficient contact area is a primary cause of reduced adhesion strength in misalignment situations. Hence, various adhesive methods have been explored to improve contact adaptation by reducing the elastic modulus of the adhesives [
19], [
25], [
26], such as material optimization (reducing the elastic modulus of the raw materials used) [
8], [
27], [
28] and structure design (optimization of aspect ratios, multilevel structures, and variable-size gradient structures) [
5], [
29], [
30], [
31], [
32], [
33], [
34]. While these methods improve conformal contact, soft structures with single modulus modulation often exhibit poor equal load sharing, leading to cracks and limited effectiveness [
35], [
36]. Tunable-stiffness adhesion—soft contact and hard separation—effectively optimizes the interfacial stress distribution by adjusting the stiffness of materials [
1], [
37], [
38], [
39], [
40], [
41]. However, external field regulation complicates tunable-stiffness adhesives, causing slow response times unsuitable for rapid adaptation. While effective on rough surfaces, these strategies face challenges with large angle errors in non-parallel contact.
Core-shell structures, combining a soft shell and a rigid core, effectively enhance contact adaptation by soft parts while inhibiting interfacial crack peeling by the rigid part [
42], [
43], [
44], [
45], [
46]. In a previous study, we proposed a mushroom-like dry adhesive structure based on a soft shell and rigid core and verified its excellent adhesion performance on different scales of rough surfaces (ground glass, sandpaper, etc.) [
47]. However, the current core-shell structure can only achieve enhanced adhesion to smooth/rough surfaces under parallel contact and cannot address non-parallel contact at larger scales. This is attributed to the “dead” rigid core, that is, immobilized inside the soft shell, and relies solely on the top thin layer of soft shell to achieve conformal contact at the microscale. Interestingly, geckos can climb flexibly under microscopic morphology and macroscopic angle errors, which also rely on the core-shell structure (soft muscle and rigid bone). Unlike the conventional “dead core”-based core-shell structure, the gecko's rigid core (bones) inside the soft shell (muscle) can rotate flexibly at the joints, rather than remaining fixed (Fig. S1 in Appendix A). This “living core” enables the foot to adapt to macroscopic angular errors through rotation, while regulating stress distribution at the interface due to its high rigidity during parallel contact, ensuring the adaptability and stability of the adhesion interface.
Inspired by this, this study proposes a “live core”-based core-shell dry adhesive by embedding a rigid, thin piece into a soft, thick elastomer. The outer soft shell is divided into the top micro-nano structures, providing high-strength adhesion, and the bottom soft layer for sufficient compliance under large angle errors. The inner rigid core, that is, the rigid piece, can flexibly rotate, enabling conformal contact and equal load sharing at the interface (
Fig. 1(a)). This synergistic design ensures excellent adhesive performance under angle errors, a feat never achieved by adhesive structures alone. Additionally, the core-shell structure effectively offsets energy accumulation from torsional behavior in the adhesion state, providing resistance to overturning when grasping objects.
2. Materials and Methods
2.1. Materials
Unless otherwise stated, solvents and chemicals were obtained commercially and used without further purification. Polydimethylsiloxane (PDMS) prepolymer and crosslinker (Sylgard 184) were purchased from Dow Corning (USA). Silicone rubber (AB-600) was purchased from Shenzhen Hongyejie Technology Co., Ltd., China. A large-aperture wedge lens was purchased from Shenzhen Meidian Acrylic Products Co., Ltd. (China).
2.2. Fabrication of mushroom-shaped adhesive structure
A mushroom-shaped adhesive structure was fabricated using PDMS through a standard molding process. This study prepared the mold using a double-sided exposure process, proposed by our team, enabling the fabrication of mushroom-shaped structures with controllable structures and good uniformity. The body and curing agent of the PDMS were mixed and stirred uniformly at a ratio of 10:1, poured into the mold, and evacuated for 10 min. Subsequently, spin coating was performed, and the curing was completed in an oven at 80 °C in 1 h. Finally, the structure was demolded. The thickness of the backing layer was controlled by varying the casting parameters. The mushroom-shaped micro-structures were approximately 14 μm high, with a 4 μm spacing. The diameter of contact tip and pillar were 18 and 14 μm, respectively.
2.3. Fabrication of core-shell adhesives
The fabrication of self-adaptive adhesives was primarily based on molding. First, a double-sided exposure process was used to prepare a 15 mm × 15 mm mold of the adhesive structure. Subsequently, PDMS was poured on the surface of the mold, placed in a vacuum chamber for 10 min, and then spin-coated at 2000 r∙min−1 for 40 s. Subsequently, a glass piece was placed on the uncured PDMS surface and heated in a 90 °C oven for 1 h for PDMS solidification. Uncured silicone rubber (AB-600, elastic modulus (E) = 400 kPa) was then poured onto the glass surface. To control the thickness of the silicone rubber layer to the millimeter level, a box with dimensions of 15 mm × 15 mm × 10 mm was placed around the sample. Silicone rubber layers with different thicknesses were obtained by pouring silicone rubber prepolymers of different qualities into a box. Finally, the sample was placed in a 90 °C oven for 1 h, removed from the box, and demolded to achieve self-adaptive adhesives.
2.4. Fabrication of soft adhesives
A 15 mm × 15 mm mold of the adhesive structure was prepared using the double-sided exposure technique. Subsequently, PDMS was poured on the surface of the mold, placed in a vacuum chamber for 10 min, and then spin-coated at 2000 r∙min−1 for 40 s. Subsequently, a silicone rubber block (15 mm × 15 mm × 4 mm) was placed on the uncured PDMS surface and heated in a 90 °C oven for 1 h for PDMS solidification. Finally, it was demolded to obtain a homogeneous soft adhesive.
2.5. Fabrication of rigid adhesives
A 15 mm × 15 mm mold of the adhesive structure was prepared using the double-sided exposure technique, and PDMS was poured on the surface of the mold, placed in a vacuum chamber for 10 min, and then spin-coated at 2000 r∙min−1 for 40 s. Subsequently, a block of glass with 15 mm × 15 mm × 4 mm was placed on the uncured PDMS surface and subsequently heated in a 90 °C oven for 1 h for PDMS solidification. Finally, the rigid adhesive was demolded.
2.6. Normal adhesion performance testing
The adhesion of the adhesive structures was measured using the load-pull mode. The probe was first contacted with the sample to generate a specific contact area, and then the reverse movement was performed until complete separation. The maximum tensile force generated before the separation was defined as the maximum adhesion force. The flat probe measured 2 cm × 2 cm. The adhesive structures were attached to the base and adjusted to be parallel/non-parallel to the probe surface. The testing surface was moved down at 2 mm∙min−1 to contact the sample and achieve a predefined preload for 2 s. It was subsequently moved up at the same speed until the testing surface was completely separated from the adhesive structures. The angular displacement platform can adjust the angle between the probe and the adhesive. A CCD camera was used to observe the real-time contact state between the adhesives and the object surface. All tests were repeated thrice, and standard deviations were calculated.
2.7. Material characterization
The microstructure of the adhesive material was observed by scanning electron microscopy (SEM; SU8010, Hitachi, Japan). The adhesion ability of the materials was characterized using a computer servo pull-pressure test machine (PT-1176, Baoda, China). The time-varying contact state between the adhesive material and the probe was observed using a digital microscope (RS-500C, Kone, China). The mechanical properties of the samples were tested using a computer servo pull-pressure test machine (PT-1176, Baoda).
3. Results and Discussion
3.1. Design of the “live core”-based core-shell dry adhesive
The fabrication process used in the realization of core-shell adhesives is presented in Fig. S2 in Appendix A, and the sectional morphology was investigated via SEM (
Fig. 1(b)). The material used to fabricate the core-shell structure was primarily PDMS (Fig. S3 in Appendix A), which has good tensile properties, thermal stability, and optical transparency, as well as low cost and easy molding, and has been widely used in microfluidics, flexible electronics, bioinspired materials, and soft robots. The adhesion region on the top layer of the material was composed of a mushroom-shaped fibrillar array and a 100 μm-thick film. The mushroom-shaped structure was distributed in a square shape, the column diameter, cap diameter, and spacing were 14, 18, and 20 μm, respectively. The thickness of the rigid piece, which was completely embedded in the elastomer, was 800 μm. The mushroom-shaped microstructure at the top of the core-shell adhesive had a modulus of 2 MPa, the soft backing (soft shell) had a modulus of 400 kPa, and the rigid layer inside (rigid core) had a modulus of 55 GPa. The surface roughness of the prepared adherent structures was characterized using atomic force microscopy, as illustrated in Fig. S4 in Appendix A. The test procedure was performed by randomly scanning three different areas on the surface of a single mushroom-shaped pillar with a scanning area of 5 μm × 5 μm, and the images were processed to obtain the RMS roughness of the microstructure surface. The results demonstrate that the surface of the adhesive structure was relatively smooth, with a roughness (
Rq) of only 23 nm, owing to the advantages of the molding process.
Conventional core-shell structures are only partially adapted to micromorphological surfaces because the rigid core is fixed inside the soft shell, ignoring the mechanism of the backing layer in the macroscopic-scale contact state. The self-adaptive core-shell adhesives proposed here optimize adhesion performance by extending the soft shell from the microscale to the macroscale, emphasizing elastic communication between the backing and the microstructure. When contacting a misaligned surface, the rigid core (rigid piece), similar to a joint, ensures the system's adaptability to the angle error through its deformation, increasing the actual contact area, impossible with conventional rigid adhesives owing to their high modulus. When the interface is pulled up, the lateral shrinkage of the elastomer backing is significantly weakened due to the insertion of the rigid core, which makes the angle (
θ) between the interface and the pulling force (
Fpull) closer to 90°, which effectively inhibits peeling and increases the stiffness uniformity of the adhesive force (
Fad). When a conventional soft adhesive is pulled up, significant lateral shrinkage at the edges is likely to occur, reducing the angle (
θ) between the pulling force (
Fpull) and the interface. This can easily cause a peeling effect and lead to local stress concentration (
Fig. 1(c)).
With ideal contact and optimal equal load sharing, the adhesion capacity in misalignment situations is effectively improved (
Fig. 1(d)). Compared with conventional soft/rigid adhesives, the adhesion strength of core-shell adhesives under misalignment (2°) increases by approximately 100 times. Moreover, inserting a rigid piece into a soft elastomer can generate good adhesion strength in alignment and misalignment situations. Such a solution cannot be obtained by manipulating the stiffness of a single homogeneous material. The adaptability of the core-shell adhesives for use on an inclined plane (slope angle = 5°) is shown in
Fig. 1(e). The size characterization of the target is shown in Fig. S5 in Appendix A. The compression of the elastic body causes the rigid piece to rotate until it reaches a state in which it is nearly parallel to the target surface, thus completing the adaptation of the core-shell adhesives. This adaptive capacity can be enhanced by optimizing the elastomer’s elastic modulus. Furthermore, the anti-overturning capacity of the core-shell adhesives was demonstrated in typical grasping situations, such as when the adhesion position is offset from the objective’s center of mass (
Fig. 1(f)). In this situation, the instability caused by the turning moment (
Mturning) on the gripping is greatly reduced.
3.2. Mechanical analysis
To reveal the adhesion mechanism of the macroscopic soft shell and rigid core configuration, a finite element analysis (FEA) model of soft, rigid, and core-shell adhesives was established using the cohesive zone surface method of ABAQUS software (Dassault Systèmes, France). Because the contact areas of the different adhesive structures are the same in the case of alignment, we focus only on the separation stage, that is, the moment when the maximum adhesive force is reached, as shown in
Fig. 2(a). At this moment, the tensile displacement of the elastomer was the largest for the soft adhesives. Additionally, the stress distribution map (
Fig. 2(b)) shows that a more pronounced stress concentration occurred at the contact edges of the soft adhesives, generating cracks and developing a peeling effect. From the dynamic detachment process of the soft adhesives, separation began at the edges of the contact interface and then extended to the entire interface (Movie S1 in Appendix A). In contrast, rigid and core-shell adhesives effectively suppressed interfacial stress concentration, with the latter showing greater inhibition, leading to complete separation. Therefore, the core-shell adhesives demonstrated the best adhesion performance under alignment. The evolution of the interfacial normal stress from the tension stage to the separation stage can be utilized to gain a deeper understanding of the adhesion mechanism from the perspective of detachment (
Fig. 2(c)). The soft adhesives showed the highest normal stress under the same tensile load, and the stress quickly reached the damage threshold value (0.02 MPa), triggering separation of the contact surface and resulting in the corresponding adhesive force (
x-axis: 3 N). In contrast, the normal stress of the core-shell adhesives increased linearly at low stress levels, indicating limited separation. Although the material was primarily subjected to normal tension and compression, shear stress also arose due to the transverse deformation of the material. The contact stress of the core-shell adhesives in the tangential direction also enhanced adhesion (Fig. S6 in Appendix A).
The adhesion enhancement mechanism under misalignment was investigated using an FEA model based on the cohesive zone surface method.
Figs. 2(d) and
(e) illustrates the preloading and separation stages of the different adhesive structures, respectively. Under the same preload, rigid adhesives cannot achieve ideal contact, and a large stress concentration occurs at the edge of the non-contact area, easily leading to crack initiation and propagation. Even if the contact area could be further improved and complete contact could be achieved by increasing the preload, the resulting extremely high interface contact stress would damage the adhesive tip and target surface. In contrast, both soft and core-shell adhesives were capable of achieving a complete fit, and their entire interface remained in contact to reach the maximum adhesion force. Embedding the rigid piece did not appear to affect the adaptability of the structure to possible misalignments, which essentially originated from the rotation of the rigid piece under the mechanical action of the top and bottom elastomers. Specifically, when the adhesive structure is in contact with the target surface under misalignment, the rigid piece is subjected to torque. With a further increase in the loading displacement, the rotation angle of the rigid piece increased gradually until the entire interface was in complete contact with the target surface (Movie S2 in Appendix A). Hence, the rotation of the rigid piece during the loading process effectively improves the adaptability of the system to possible misalignments.
During the tension process, the top and bottom elastomer layers gradually release the strain energy stored inside, and the rigid piece continues to rotate until the maximum adhesion is reached owing to the difference in the thickness and compression state of the two layers. Here, the parameter
α, the angle between the target surface and the upper surface of the rigid piece, is introduced;
Fig. 2(f) illustrates the variations in
α with time. As contact occurs,
α decreases gradually until the maximum preload is reached. Subsequently, the structure begins to be pulled reversely, and
α decreases further. When the maximum adhesion force is reached,
α becomes 0.25°, 75% lower than the initial value. It can be concluded that the rotation of the rigid piece effectively reduces the actual misalignment angle. According to the expression for calculating the adhesion strength considering possible misalignments proposed by Kroner et al. [
48], for the same contact area, a reduced misalignment angle effectively increases the adhesion strength.
In addition, the rotation of the rigid piece during the preloading and separation stages improved the stress distribution at the interface, as illustrated in
Figs. 2(g) and
(h). For the same preloading and maximum adhesion moment, the interface stress of the core-shell adhesives was effectively reduced, and its distribution became more uniform. This flattening of the stress distribution resulted in the maximum normal contact stress (
Fig. 2(i)), and the maximum shear contact stress (Fig. S7 in Appendix A) at the interface remained low during the pulling-up process, thereby hindering interface separation and promoting a higher adhesion strength. Fig. S8 in Appendix A shows the work of attachment and detachment of the adhesive structures on non-parallel surfaces. Owing to the state transition from attachment to detachment, there is an evident turning point in the curve of the external work; that is, the value of the work sequentially increases from small to large, from large to small, and then to larger. The core-shell structure exhibited more detachment than the soft and rigid structures, indicating the strongest adhesion interface.
3.3. Adhesion performance
Based on a theoretical analysis of the coupling effect of the core-shell structure in the adhesion process, the adhesion performance of the adhesive developed in this study was quantitatively characterized, including the normal adhesive strength in situations with alignment and misalignment (
Fig. 3(a)). Because the interfacial adhesion mechanism of the bioinspired adhesive structure primarily originates from van der Waals forces, when the ambient humidity increases, a thin liquid film is formed at the adhesion interface, and the capillary force effect is introduced. In this case, the van der Waals forces are weakened [
49], [
50], [
51], reducing the adhesion strength (Fig. S9 in Appendix A). Therefore, the ambient relative humidity (RH) was maintained at 50% during the experiments. The testing process is detailed in Fig. S10 in Appendix A. The mechanical properties of the three adhesives, which determine the adhesion performance of the structures, were first tested, as illustrated in
Fig. 3(b). The embedding of the rigid material within the structure reduces the thickness of the actual interface adhesion layer, thus increasing the interface stiffness compared with the soft material alone. Under good alignment state, the maximum adhesion strength of the core-shell adhesives was approximately thrice greater than that obtained using soft adhesives, which is attributed to the increased equivalent stiffness (
Fig. 3(c)). For a given contact area, according to the equilibrium theory of adhesion between elastic solids, the adhesive force is proportional to the elastic modulus of the material [
52]. However, the rigid adhesives (which have the highest elastic modulus) do not exhibit the largest adhesion here; they are lower than those of the core-shell adhesives. This may be because absolute 0° may not be reached owing to the errors in the flatness of the adhesion structure and the limitations of the angle observation method (CCD-based observation system). To better illustrate the adhesion performance of different adhesion structures in the parallel contact state, adhesion strength was tested within the range −0.3°-0.3°. The rigid structure demonstrated the highest adhesion performance near 0°; however, it was extremely sensitive to angular changes, with even a 0.1° deviation significantly reducing the adhesion performance (Fig. S11 in Appendix A). Moreover, the adhesion strength of various samples demonstrated minimal dependence on the preload, with adhesion reaching a plateau at relatively low preload values (
Fig. 3(d)). Among these, the core-shell adhesives exhibited superior adhesion performance under different preload values, with the adhesion strength thrice greater than that of the soft adhesives. This enhancement is not solely attributed to the core-shell configuration; the mushroom-shaped microstructures also significantly contribute to enhanced adhesion by optimizing interfacial stresses and controlling internal crack nucleation [
53], [
54] (Fig. S12 in Appendix A).
Fig. 3(e) illustrates the time-force curves of the three structures during the loading-pulling process, with a misalignment of 0.5°. Compared with the results obtained for the tests with good alignment, the existence of alignment errors significantly decreased the adhesion performance of the soft and rigid adhesives. In the case of core-shell adhesives, the curve in the pulling-off process is very steep, and the adhesive exhibits excellent adhesion performance even in the case of misaligned objects. To reveal how misalignment affected the contact-separation states, a CCD camera was used to observe the entire contact-separation process of the three structures and recorded the length of the contact line in real time, as illustrated in
Fig. 3(f). During the loading process, the contact line length of the adhesive plateaued at the required preload. The length of the contact line of the rigid adhesives at this moment was only 9 mm, much smaller than the width of the sample (16 mm). This implies that the adhesive is not in complete contact with the probe, a critical factor in adhesion performance reduction caused by misalignment. In contrast, for the soft adhesives and the core-shell adhesives, the length of the contact line in the plateau stage was 16 mm, indicating that these adhesives were in complete contact with the glass probe. Notably, there is no significant difference in the misalignment adaptability between the core and shell adhesives and soft adhesives, which is attributed to both structures exhibiting the same compression stiffness under misalignment (Fig. S13 in Appendix A).
The adhesion strength as a function of the misalignment angle under different preloads was also investigated, and the maximum adhesion of the core-shell adhesives was over one order of magnitude higher than that of the other two structures (
Fig. 3(g)). The repeatability of the core-shell adhesive structure was demonstrated, with its adhesive force remaining unchanged after 150 cycles (Fig. S14 in Appendix A). This superior bonding performance may be attributed to the interfacial stability between the internal rigid layer and the soft material. In addition, the adhesion properties of the core-shell structure at high temperatures were also tested, as illustrated in Fig. S15 in Appendix A. As the temperature increased, the adhesion properties of the adherent structure remained stable and did not decrease significantly, indicating its potential use at high temperatures. In addition to their high temperature, the materials used in the core-shell structure are also widely used in microfluidic systems because of their excellent mechanical stability. They do not soften in liquid environments and are resistant to reactions with common chemical reagents. The experimental results also demonstrate the good water tolerance of core-shell structures, facilitating their wide application in various complex environments (Fig. S16 in Appendix A).
3.4. Structure optimization
The macroscopic core-shell strategy does not simply insert a rigid piece into the soft material but requires a particular matching design to achieve the desired effect. The rotation of the rigid piece is undoubtedly affected by several factors, such as the stiffness of the soft shell and rigid core, geometric position of the rigid core, and size of the rigid core. Exploring the influence of these factors on the rotation angle will guide further optimization of adhesive structures. Therefore, the effects of the different parameters on the rotation angle of the composite system were investigated (
Fig. 4).
Fig. 4(a) depicts the effect of the elastic modulus of the rigid material on the stiffness of the composite and the rotation angle at maximum adhesion. Here, the mechanical properties of the rigid material were represented by the ratio of the elastic moduli of the core and shell (
Ecore/
Eshell), and the mechanical properties of the soft shell were represented by the ratio of the elastic moduli of the bottom and top layers (
Ebottom/
Etop). The position of the rigid piece is described by the thickness ratio of the upper and lower soft materials (
h2/h1). This study focused on analyzing the variation of
α at maximum adhesion, that is,
α at separation. When the modulus of the rigid material was sufficiently low (< 10 MPa), the rigid piece was subjected to bending under the action of the torque at the moment of maximum adhesion, making it impossible to calculate the rotation angle. Furthermore, a high-stress concentration occurs at the edge of the contact interface owing to bending, which is not conducive to adhesion. When the modulus of the rigid material was high (> 100 MPa), it was observed that the rotation angle decreased gradually with an increasing modulus of the rigid piece. The theoretical data of this model were validated experimentally; different materials (glass, polymethacrylates (PMMA), and high-modulus PDMS) were selected to be embedded in the elastomer, and their normal adhesion strengths were tested (
Fig. 4(b)). The results demonstrate that both PMMA adhesives and glass exhibit excellent performance under alignment and misalignment conditions. In addition, the performances of these two materials in the case of alignment were similar, aligning with the trend of equivalent stiffness (Fig. S17 in Appendix A).
The influence of the mechanical properties of the soft shell on the rotation angle was analyzed using the FEA model, as shown in
Fig. 4(c). During the preloading stage, the mechanical properties of soft materials do not significantly affect the rotation angle. In contrast, during the pulling stage, the rotation angle of the material with a larger
Ebottom/
Etop increases reversely, likey due to the greater torque effect of the bottom material on the rigid piece compared to the top material. The rotation angle at the moment of maximum adhesion affected the adhesion strength in the misalignment state. According to the results, the material with a smaller
Ebottom/
Etop ratio exhibited superior adhesion performance (
Fig. 4(d)). To verify the results of the model, composite adhesive structures with different
Ebottom/
Etop values were fabricated and their maximum adhesion strengths were tested. The experimental data were found to be consistent with the simulation predictions (
Fig. 4(e)). Due to the isolation effect of the rigid piece, the stress distribution of the adhesion interface primarily depends on the mechanical properties of the top material, regardless of the elastic modulus of the bottom layer (
Ebottom). Since these properties do not vary significantly (Fig. S18 in the Appendix A), the adhesion performance remained unchanged.
In addition to the mechanical properties of the soft shell, the position of the rigid piece inside the elastomer is also a critical factor affecting the adhesion strength. When
h2/h1 = 0, soft materials can be considered homogeneous. In this case, the value of the rotation angle was the same as that in the initial state (1°), and the overall equivalent stiffness was minimal (
Fig. 4(f)). When
h2/h1 > 0 (i.e., the rigid piece was embedded in the elastomer), the rotation angle in the maximum adhesion state gradually decreased with increasing
h2/
h1 ratio. This implies that a thinner top layer coupled with a thicker bottom layer improves adhesion in the misalignment state. Considering that the top layer requires a certain degree of compliance to ensure microscale end contact with the target, the
h2/h1 value cannot be infinite. Composite adhesive structures with different top-film thicknesses were fabricated, and the simulation results were verified by testing the adhesion strength in the alignment (0°) and misalignment (0.5°) cases (
Fig. 4(g)). The experimental results demonstrate that a decrease in the thickness of the top layer can improve adhesion performance under misalignment situations, and this principle is also applicable to alignment cases. Under good alignment, although the position of the rigid piece does not change the overall stiffness of the material, a decrease in the thickness of the top layer can still improve adhesion by promoting equal load sharing (Fig. S19 in Appendix A), similar to the results reported by Kim et al. [
35].
A typical feature of a core-shell structure is that the rigid core is encased inside the soft shell; therefore, there is a gap (
l) between the rigid piece and the soft elastomer. It was observed that changes in
l also affected the adhesion strength. Specifically, the adhesion performance can be further improved when the length of the rigid piece is slightly smaller than that of the elastomer. However, the adhesion performance decreases with the length of the rigid piece, even reaching a downward trend (
Fig. 4(h)). This phenomenon is closely related to the interfacial stress distribution of the different structures, as illustrated in
Fig. 4(i). When there is a gap between the rigid piece and the elastic body (
l > 0), the interface shows a pronounced stress concentration in the gap region. At this time, cracks were generated and propagated from the inside rather than from the edge, which reduced the stress concentration at the interface to a certain extent, thereby promoting adhesion. However, as the length of the rigid piece was further reduced, the interfacial stress distribution became uneven owing to the recurring large stress concentration; thus, the adhesion strength started to decrease again. Therefore, an optimized structure was obtained by embedding a rigid piece in the elastomer while leaving a short gap on both sides.
3.5. Technological perspective
Based on the structural optimization findings, the core-shell adhesives exhibited extremely high adhesion performance, especially under misalignment conditions.
Fig. 5(a) illustrates the adhesion strengths of the soft adhesives and the core-shell adhesives for different misalignment angles. As the misalignment angle increased, the adhesion strength of the soft adhesives decreased rapidly, whereas the core-shell adhesives retained relatively good adhesion strength. When the misalignment angle was 5°, the adhesion strength of the soft adhesives dropped to almost zero, whereas that of the core-shell adhesives decreased by 30% compared with the value at 0° (80 kPa); the difference between the two was nearly two orders of magnitude. In addition to adaptability to misalignments, the ability to withstand pull-ups at different angles is another type of misalignment adhesion. The soft and the core-shell adhesives were initially placed in full contact with the glass probe, and the adhesive forces along various pulling directions were tested, as shown in
Fig. 5(b). The core-shell adhesives still exhibited great superiority, further validating the robustness and stability of this adhesive.
Based on their excellent adhesion performance in misalignment situations, the core-shell adhesives have considerable application potential in object manipulation, especially for surfaces with sloping features. The adaptability to sloping features in the contact process and the inhibition of the overturning torque in the pickup process can be achieved simultaneously, which is different from the traditional strategy in which only the stiffness of the homogenous elastomer is tuned (
Fig. 5(c)). This study selected a large-aperture wedge lens as the target object to demonstrate the adhesion ability of the core-shell adhesives in the optical assembly field (Fig. S20 in Appendix A). The adhesive sample adhered to a six-degrees-of-freedom robotic arm, and a manipulation process that included preloading, grasping, and rotation was successfully executed (
Fig. 5(d), Movie S3 in Appendix A). First, the robotic arm gradually approached and contacted wedge lens (
Fig. 5(d-i)). Owing to the rotational characteristics of the rigid core in the composite structure, the adhesive can adaptively fit the sloping surface of the wedge lens (
Fig. 5(d-ii)). Additionally, no large interfacial force can damage the target object because the elastomer substrate bears most of the elastic deformation. The robotic arm then moves in reverse and stably grabs the lens at certain angles (Figs. 5 (d-iii) and (d-iv)). During this process, the contact interface of the wedge lens generates an overturning torque owing to the effects of its own gravity and pulling-up angle. However, the adhesive inhibits the peeling effect, enabling a robust adhesion state. Finally, the robotic arm rotates in different directions to increase the overturning torque at the adhesion interface, and the adhesive maintains a stable grasping state. The proposed dry adhesive, which exhibits considerable adaptability and anti-overturning ability, is a promising candidate for applications in complex optical component assembly.
4. Conclusion
A macroscopic core-shell adhesive was proposed to improve the adhesion under non-parallel conditions, comprising a top adhesion tip with a mushroom-like geometry for interfacial adhesion based on van der Waals forces and a bottom core-shell configuration for interface stress regulation. Unlike traditional core-shell adhesives that emphasize microstructures, this strategy focuses on the macroscopic backing layer. Specifically, the soft shell was separated into top and bottom parts by inserting a rigid piece into a homogeneous elastomer. The top soft layer provided high adhesion, and the bottom soft layer generated large deformations during contact. Crucially, the inner rigid layer was active and rotated during the contact process, enabling adaptive contact with macroscopic interfacial angle errors and optimizing equal load sharing at the interface. Adhesion performance could be flexibly regulated by adjusting the size, position, and stiffness of the composite material. Experimental results demonstrate that the proposed adhesive achieves adhesion strength up to 100 times greater than conventional adhesives under misalignment, without compromising performance in aligned conditions—overcoming the typical trade-off in current adhesive optimization under misalignment.
The adhesive’s excellent adaptability to misalignments and inhibition of interface peeling effects enable stable use in target objects with typical sloping characteristics. The experiments show that the adhesive, combined with a six-degree-of-freedom robot, can flexibly pick up typical optical components with various geometries (e.g., large wedge-shaped lenses) and can achieve stable adhesion in different rotational postures owing to its anti-overturning capability. This innovative and versatile strategy can be applied to various fiber-array adhesive structures, broadening the application range and operational capabilities of adhesive materials and enabling the development of adhesion-based grippers for engineering applications.
CRediT authorship contribution statement
Duorui Wang: Writing - original draft, Methodology, Data curation. Hongmiao Tian: Supervision. Jinyu Zhang: Data curation. Haoran Liu: Software. Xiangming Li: Validation. Chunhui Wang: Methodology. Xiaoliang Chen: Software. Jinyou Shao: Supervision, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was supported by the National Natural Science Foundation (52025055, 52175546, and 52405624), and the Shaanxi University Youth Innovation Team.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2024.12.035.
PDF
(3502KB)
