Highly Adaptable and Biocompatible Octopus‐Like Adhesive Patches with Meniscus‐Controlled Unfoldable 3D Microtips for Underwater Surface and Hairy Skin

Abstract Adhesion capabilities of various skin architectures found in nature can generate remarkable physical interactions with their engaged surfaces. Among them, octopus suckers have unique hierarchical structures for reversible adhesion in dry and wet conditions. Here, highly adaptable, biocompatible, and repeatable adhesive patches with unfoldable, 3D microtips in micropillars inspired by the rim and infundibulum of octopus suction cup are presented. The bioinspired synthetic adhesives are fabricated by controlling the meniscus of a liquid precursor in a simple molding process without any hierarchical assemblies or additional surface treatments. Experimental and theoretical studies are investigated upon to increase the effective contact area between unfoldable microtips of devices, and enhance adhesion performances and adaptability on a Si wafer in both dry and underwater conditions (max. 11 N cm−2 in pull‐off strength) as well as on a moist pigskin (max. 14.6 mJ peeling energy). Moreover, the geometry‐controlled microsuckers exhibit high‐repeatability (over 100 cycles) in a pull‐off direction. The adhesive demonstrates stable attachments on a moist, hairy, and rough skin, without any observable chemical residues.


Fabrication of octopus-inspired adhesives having -SCs.
The hole-patterned molds made of silicon (30 m in diameter, AR 1, and SR 1) were prepared by photolithography and subsequent reactive ion etching. Here, AR is the width-to-depth ratio and SR is the distance between each structure (d) divided by the diameter (2r). The molds were treated with a atmosphere. Using cyanoacrylate glue, the as-prepared molds were fixed to the upper side of a custom-built pressing machine as shown in Figure S1a. Afterwards, a liquid precursor (s-PUA:MINS-301RM purchased from Minuta Tech, Korea) was dispensed onto a polyethylene terephthalate film (PET; thickness 50 m for a backbone) (see Figure S1a). The polymeric precursor and hole-patterned mold were then compressed with a pressure of 300 kPa to preciously control the capillarity of the precursor between a mold and a PET film as illustrated in Figure 1b-c. Subsequently, ultraviolet-light located in the pressing unit was irradiated for 2 minutes (see Figure S1b). By peeling the cured S-PUA/PET film off, we obtained octopus-inspired adhesives with -SCs. (see Figure S1c). The PUA replica was further exposed to ultraviolet light for several hours to achieve complete curing. To control the curvature of 3D micro-tips in the octopus-inspired adhesive (See Figure 2c-f and Figure   S4), we used solid molds (s-PUA, silicon, and PDMS) of different surface energies (see Figure S3) to fabricate various -SCs in the same manner. The various solid molds obtained through multiple replications using Si-based molds are shown in Figure S2. For scale studies as shown in Figure 2h, we used s-PUA molds of 3 different diameters for the hole-patterns (30 m, 100 m, and 1000 m with AR 1 and SR 1) and fabricated various -SCs through the same method. By contrast, the cylindrical micro-pillars were fabricated by removing air submitted to bubbles completely by using a vacuum pump as shown in Fig. 2c. Details regarding the synthesis and characterization of the PUA polymers can be found elsewhere [1] .

Fabrication of the softened PDMS-based skin-patches with and without -SCs.
To obtain the skin-patch, we used soft, elastic PDMS (Polydimethylsiloxane: Sylgard, 184, Dow Corning), produced by mixing a curing agent (5 wt%) with a PDMS precursor. The soft PDMS-based patch with -SCs was fabricated through multiple replications using the asprepared adhesive with size-maximized -SCs made using a s-PUA mold (CR~0.93) as shown in Figure S10. The mixture was then cured at 70 o C for 2 hours. On the other hand, PDMS-based patch without microstructures was fabricated with a flat FOTCS-treated Siwafer. Herein, the thickness of the adhesives (~ 600 m) was controlled by using a spincoater.
Detailed mechanisms to control the geometry of -SC. To explain the underlying mechanism for tailoring the geometry of -SCs, a simple model based on the liquid-meniscus in a capillary action was investigated. As shown in Fig. 1c, the particular geometry of -SCs is achieved through stress balance between pressure of pressed air ( ; blue arrow), Laplace pressure by capillarity [2] ( ; green arrow), and the external pressure loaded by a compressor ( ; black arrow): . Here, the Laplace pressure ( ) in trapped air is given by , where is the radius of the hole-pattern of mold, is the interfacial tension, and is the contact angle between a liquid precursor and solid molds (see Figure 1c (i) and Figure S3). According to Young's equation [4] , the 3D curvature of the meniscus of a polymer precursor results from the contact angle ( ) on the side-wall of a cylindrical-hole in a mold as shown in Figure 2a. We used 3 different patterned-molds of various surface energies: silicon (~60 mJ/m 2 ), s-PUA (~40 mJ/m 2 ), and PDMS (~20 mJ/m 2 ) [3] . For convenience, we defined the curvature-ratio (CR) of meniscus as , resulting in the tip geometry of -SCs as shown in Figure 2b and d-f. Here, R is the radius of trapped-air.
In addition, the meniscus-height (h) in Figure 1c may determine the height of -SCs after curing process. Herein, the meniscus-height (h) can be controlled by external load of the pressing machine. Since the volume of a trapped air ( ) is estimated with the energy balance between Laplace pressure ( ) via capillarity an external load ( ), we can obtain the following equation (2) by Boyle's law: Here, is the initial air-pressure at ambient condition (~101 kPa) and is the volume of cylindrical hole ( ), wherein H is the depth of cylindrical hole. Because the curvature of the meniscus is decided by surface properties of solid molds (see Figure 1c (i)), the volume of trapped air ( ) can be expressed as By combining equations (3) and (4), we can determine the height of SC structures as a function of compressor pressure ( ) .

submitted to
Adhesion Measurements. Using a custom-built equipment (Adhesion tester, Neo-Plus, Korea; see Figure S5a), adhesive patches (1×1 cm 2 ) with different geometries were attached with a preload of 3.5 N/cm 2 to the Si wafer and glass until failure occurred in the pull-off direction. To test our device's adhesive performance in different conditions, we measured the normal adhesion forces on the substrate in dry (relative humidity ~50 %) and underwater conditions (see the inset images in Figure S5b). Specifically, the external preload ( ) applied to the contact between the sample and the substrate is assumed to a positive force as shown in . Here, the van der Waals adhesion is negligible for underwater adhesion. [5] As the artificial adhesive is attached, the tailored 3D micro-tips could be unfolded to increase the effective contact-area on the surface, expressed with (see Figure 2j(i) and Figure S8a). Here, the radius of the unfolded tip r' is achieved by the curvature of 3D micro-tips in -SCs ( Figure S8a), given by submitted to To prevent the interference among -SCs during their unfolding behaviours, we designed our adhesive patches ( ; see Figure S7-8). Thus, the -SCs cannot physically contact each other in all experiments when completely unfolded ( Figure S7b). In case of capillary interaction ( ), the inside of suction chambers could widen and reach vacuum state during attachment, resulted from the deformation of elastic pillars by an applied preload (see Figure   2j(i) and Figure S7). Herein, adhesion via capillary-bridges ( ) between engaged parallel surfaces is calculated with the dynamic contact area ( )) of enlarged tips, which may be expressed as [6] ( ) Here, is the surface tension of water (~0.072 J/m 2 ), n is the number of -SCs per unit area Next, the suction stress ( ) is defined as .
As -SC structures are pulled off, void formation occurs inside the structures (see images of Figure 2j(ii) and Figure S7). Here, is the pressure difference between the ambient pressure and the pressure inside the -SC (nearly creating a vacuum state; ~101 kPa) and is the area of the void inside of suction chamber ( ). Since total adhesion is sum of suction stress and capillary interaction, can be obtained with equation (9) and (10) submitted to In underwater condition the assistance of capillary-bridges between widened concave-tips of suckers and an engaged surface could be expected due to the existence of residues of water molecules, preventing air leakages. [7] In particular, the contact area between an unfolded 3D micro-tips and engaged surface, which is the dynamic contact area ( )), would change as the adhesive is pulled off. The suction effect may be estimated by the void area ( ), whereas the capillary interaction ( ) may be expressed as a function of . By monitoring , we can predict the suction stress (green line), capillarity adhesion (blue line), and total adhesion strength (red line) of the bioinspired adhesive with -SCs (n ~ #/cm 2 ) as plotted in Figure 2k. This is well in agreement with our experimental data (black dots).

Preparation and characterization of pigskin. The postmortem pigskins for adhesion
tests were obtained from a local slaughterhouse (Jidong, Korea) and stored frozen, as following well-established methods [7,8] . Pig cadaver grafts cut in ~2 × 2 cm 2 area were immersed in DI water for 2 hours. DI water was then removed from the surface of pigskins with blotting paper to use as the substrate for measuring the adhesion of the PDMS-based SC-adhesive. Micro-scale roughness profile of the pigskin surface was determined using a cross-sectional scanning electron microscope (SEM) image. We collected the vertical distances ( ) from the mean-line of surface step-height corresponding to each point of the horizontal displacements and calculated the root-mean-square (RMS) roughness using atomic force microscopy (AFM), given by √ ∑ as shown in Figure 3c.

Measurement of adhesion forces with pigskin.
With well-established methods, [7,8] postmortem pigskins may be used for various adhesion tests (see Figure S12). This is because pigskin is known to have similar morphological features to human skin. [7] After preparing the pigskin through the above method, flat grafts of pigskin were affixed to all areas underlying the surfaces using cyanoacrylate glue for adhesion tests (see Figure S12). The -SC patch or non-patterned patch was glued onto the opposing jig of the mechanical tester. Using samples of the same size (1 cm x 1 cm), pull-off adhesion tests on pigskin were performed with the custom-built equipment (Adhesion tester, Neo-Plus, Korea) as shown in Figure S5. The test samples were applied to the surface of pigskin with preload 3.5 N/cm 2 for 5 seconds. The adhesion force corresponding to the pull-off force was then obtained by detaching the test samples from the dry or wet pigskin (Figure 3d and S11a). For the peel-off test, the patch with -SCs and non-patterned patch were bonded to a rigid string and attached to the epidermis of pigskins as shown in Figure S12b              submitted to