Strong and switchable adhesives utilizing the rubber-to-glass transition

Abstract

The current invention relates to a gripping member, the member comprising a base portion formed from a rubber-to-glass phase transitioning material; and one or more fibrils extending from the base portion made from a rubber-to-glass phase transitioning material, wherein the rubber-to-glass phase transitioning material can transition from a glass phase to a rubber phase upon application of a stimulus and return to the glass phase in an absence of the stimulus; has an elastic modulus of less than 1 MPa in the rubber phase; and the transition from a glass phase to a rubber phase occurs in the one or more fibrils having a radius of at least 0.1 mm in size.

Description

[0001] STRONG AND SWITCHABLE ADHESIVES UTILIZING THE RUBBER-TO-GLASS TRANSITION

[0002] Field of Invention

[0003] The current invention relates to a gripping member, a gripping apparatus, a gripping device, and a method of gripping using the gripping device.

[0004] Background

[0005] The listing or discussion of a prior-published document in this specification should not necessarily be taken as an acknowledgement that the document is part of the state of the art or is common general knowledge.

[0006] Ever since Neanderthals made tar from dry distillation of birch bark to join stone tools to wooden handles 200,000 years ago, intelligent species like humans have been using adhesives to join different materials together (Journal of archaeological science 2006, 33, 1310-1318; Progress in Materials Science 2021, 120, 100823). Today, adhesives are becoming smarter (Progress in Materials Science 2021, 120, 100823; Advanced Materials

[0007] 2021, 33, 2102983; Advanced Functional Materials 2020, 30, 1901693), with strong adhesion for robust attachment to various surfaces and weak adhesion for on-demand detachment. Smart adhesives, which provide strong adhesion when needed and easy detachment when desired, are widely used in various applications, including in daily life (Adv. Mater. Interfaces, 2019, 6(3), 1801542), agriculture (Sci. Robot., 2020, 5(49), eabd9120), and industry (Progress in Materials Science 2021, 120, 100823). They have also played important roles in daily life and in advanced technologies including additive manufacturing (Mater. Today, 2018, 21(9), 951-959; Nat. Commun., 2021 , 12(1), 1-9), micro-nano assembly (npj Flex. Electron., 2018, 2(1), 1-14; Soft Matter, 2019, 75(1), 30-37; Extreme Meeh. Lett. 2019, 27, 76-82; J. Meeh. Phys. Solids, 2020, 143, 104066; Sci. Adv., 2020, 6(25), eabb2393; Adv. Funct. Mater., 2021 , 31(52), 2105407; Science, 2022, 377(6605), 517-523), robotics and automation (Extreme Meeh. Lett., 2018, 22, 51-59; Small, 2019, 15(4), 1803870; Sci. Robot., 2020, 5(49), eabd9120; Sci. Adv., 2020, 6(7), eaay5120; Sci. Adv., 2022, 8(28), eabq1905; Adv. Mater.,

[0008] 2022, 34(13), 2109325), wearable electronic and biomedical devices (Mater. Horizons, 2019, 6(4), 642-683; Sci. Adv., 2022, 8(23), eabo0537; Natl. Sci. Rev., 2022, 9, nwac016; Science, 2022, 377(6605), 517-523). Stronger adhesion strength and greater adhesion switchability have been two major challenges for smart adhesives (Progress in Materials Science 2021, 720, 100823; Advanced Materials 2016, 28, 7457-7465; Advanced Materials Technologies 2021, 6, 2100335). First, it is very difficult to achieve strong adhesion between macroscopic objects because of the adhesion paradox (Physical Review E 2020, 702, 042803; K. Kendall, "Introduction to Molecular Adhesion and Fracture: The Adhesion Paradox" in Molecular Adhesion and Its Applications: The Sticky Universe. (Springer US, Boston, MA, 2001), 10.1007 / 0-306-48406- 4_1 , pp. 3-23; Proceedings of the National Academy of Sciences 2014, 777, 3298-3303) - the adhesion strength decays dramatically with increasing surface roughness on common objects, despite molecules adhering to one another with considerable forces at the atomic scale. Macroscopic objects usually do not adhere to each other unless the contact surfaces are polished to near atomic scale smoothness, or one of them is soft enough to conform to surface roughness, or special chemical bonds are formed between the two adherent surfaces. An adhesive is said to be sticky when the elastic energy required for it to achieve conformal contact with a rough surface is less than the surface energy gained during microscopic contact formation (Proceedings of the National Academy of Sciences 2014, 777, 3298-3303; Physical review letters 2005, 95, 124301; The journal of adhesion 2021, 97, 509-527; Tribology International 2021, 759, 106971 ; Friction, 2022, 77, 1027; Tribology International 2020, 746, 106031), and soft elastomers are often used for smart adhesives in consideration of this stickiness criterion. However, elastomeric adhesives often store a large amount of strain energy during contact formation, which is released during pull-off and breakage of adhesive bonding. Consequently, the adhesion strength of elastomeric adhesives decreases rapidly with increasing surface roughness (Journal of The Royal Society Interface 2013, 70, 20130620; Acta Biomaterialia 2012, 8, 282-288; Beilstein Journal of Nanotechnology 2019, 70, 1-8; ACS Appl Mater Interfaces 2022, 74, 46239-46251 ; Journal of the Mechanics and Physics of Solids 2021 , 757, 104402; International Journal of Solids and Structures 2016, 87, 244-249). As shown in FIG. 1(A), the adhesion strength of the bulk adhesives made of elastomers (e.g. Silicone rubber (Mathematical and Physical Sciences 1975, 345, 327-342), polydimethylsiloxane (PDMS) (Physical review letters 2005, 95, 124301 ; Acta Biomaterialia 2012, 8, 282-288), polyvinylsiloxane (PVS) (Journal of The Royal Society Interface 2013, 70, 20130620; Beilstein Journal of Nanotechnology 2019, 70, 1-8), Dow soft skin adhesive (SSA) (ACS Appl Mater Interfaces 2022, 74, 46239-46251)) is typically weakened as the surface becomes rougher. Through evolution, gecko and many insects have developed hierarchical hairy adhesives to increase their adaptability on rough surfaces (Nature 2000, 405, 681-685). This design principle has inspired many synthetic counterparts (Progress in Materials Science 2021, 720, 100823). Despite improvements over bulk adhesives, the elastomeric hairy adhesives still lose adhesion on rough surfaces (Acta Biomaterialia 2012, 8, 282-288; Journal of The Royal Society Interface 2013, 10, 20130620), but generally at a slower rate.

[0009] Many animals, such as gecko (Progress in Materials Science 2021, 720, 100823; Nature 2000, 405, 681-685; MRS bulletin 2007, 32, 473-478), have evolved sophisticated hairy adhesive pads to overcome the adhesion paradox. These hairy structures ensure that the effective modulus of the adhesive pad can be as low as 100 kPa (Journal of Experimental Biology 2006, 209, 3558-3568), making them sticky on most rough surfaces. At the same time, the hierarchical structures increase the number of contact points of the adhesive pads. The hairy adhesive structures have inspired many synthetic counterparts in various applications (Progress in Materials Science 2021, 720, 100823). These adhesives, which have been inspired by natural fibrillar adhesive systems (Progress in Materials Science 2021, 720, 100823; Nature 2000, 405, 681-685) have shown potentials in soft grippers (Science advances 2020, 6, eaay5120; Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353) robotics (Proceedings of the National Academy of Sciences 2017, 774, E4344- E4353), advanced manufacturing (npj Flex. Electron., 2018, 2(1), 1-14), and wearables (Progress in Materials Science 2021, 720, 100823).

[0010] Despite improvements over bulk adhesives, elastomeric hairy adhesives still lose adhesion on rough surfaces (Journal of The Royal Society Interface 2013, 70, 20130620; Acta Biomaterialia 2012, 8, 282-288), but generally at a slower rate. Some optimal designs, such as contact splitting (Beilstein Journal of Nanotechnology 2019, 70, 1-8), soft tip (Royal Society Open Science 2017, 4, 161105; ACS Applied Materials & Interfaces 2017, 9, 1036-1044), or terminated film (ACS AppI Mater Interfaces 2022, 74, 46239-46251 ; ACS Applied Materials & Interfaces 2017, 9, 7752-7760) (FIG. 1(B)), have also been proposed to further reduce the rate of adhesive degradation. Also, the existing approaches to achieve switchable adhesion using fibrils, based on mechanical principles such as controlled shearing (Advanced Materials Interfaces 2022, 9, 2101764), buckling (Advanced Functional Materials 2021, 22, 1246-1254) or peeling (Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353), have been limited by the trade-off between adhesion strength and switchability (the ratio of the maximum adhesion load to the lowest detachment load), as well as the scaling limits at both the single fibril and fibrillar array levels (Progress in Materials Science 2021, 720, 100823; Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353). The trade-off between adhesion strength and adhesion switchability, can be defined as the ratio of the maximum adhesion force when the adhesion is turned “ON” and the force during intentional release when the adhesion is turned “OFF” (Progress in Materials Science 2021, 720, 100823). Stronger adhesives are usually more difficult to detach. For bioinspired adhesives made of elastomers, the adhesion switchability decreases as the adhesion strength increases. For example, when the adhesion strength is increased to 1000 kPa, the bioinspired hairy adhesives become almost non-switchable with adhesion switchability down to about one. Typical values of the adhesion strength of geckos range from 10 to 100 kPa and the corresponding values of adhesion switchability range from tens to several hundred (Nature 2000, 405, 681-685). One of the commonly used elastomers for synthetic hairy adhesives is PDMS (Progress in Materials Science 2021, 120, 100823). Other elastomers used include polyurethane (PU) (Advanced Materials 2014, 26, 4901-4906), polyurethane acrylate (PUA) (Advanced Functional Materials 2021, 31, 2107023), liquid crystal elastomer (LOE) (Bioinspired photocontrollable microstructured transport device. Sci. Robot 2 (2017)), carbon nanotubes (ONT) (Science 2008, 322, 238-242) and polyethylene glycol dimethacrylate (PEGDMA) hydrogel (Advanced Functional Materials 2018, 28, 1706498). The adhesion switchability and adhesion strength values of bioinspired hairy adhesives using these elastomers are compiled and shown in FIG. 2. The bioinspired hairy adhesives can achieve a very high adhesion switchability of about 2000, while the adhesion strength is in the order of 10 kPa. However, as the adhesion strength increases, the adhesion switchability decreases dramatically. As the adhesion strength is increased to 1000 kPa by some special designs such as mushroom-shaped tips (Advanced Functional Materials 2021, 31, 2107023), the bioinspired hairy adhesives become almost non-switchable (i.e., with an adhesion switchability of about one).

[0011] Theoretical models have revealed (Proc. Natl. Acad. Sci., 2004, 707(21), 7851-7856; Mechanics of Materials 2005, 37, 275-285; Journal of The Royal Society Interface 2004, 7, 35-48) that the adhesion strength of fibrils is strongly dependent on their size. Small fibrils can achieve attachment in the DMT-like regime (J. Colloid Interface Sci., 1975, 53(2), 314-326; J. Meeh. Phys. Solids, 2020, 739, 103937), reaching the theoretical adhesion strength Oth and leading to size-independent strong adhesion. In nature, adhesive fibrils that approach the DMT-like regime are very small, ranging from a few hundred nanometers to a few micrometers (Progress in Materials Science 2021, 720, 100823; Proc. Natl. Acad. Sci., 2004, 707(21), 7851-7856). On the other hand, large fibrils tend to detach in the JKR-like regime (J. Meeh. Phys. Solids, 2020, 739, 103937; Proceedings of the royal society of London. A. mathematical and physical sciences 1971, 324, 301-313), resulting in weak adhesion due to stress concentration at the edge (Progress in Materials Science 2021, 720, 100823; Small 2012, 8, 2464-2468). As the radius of a single fibril increases from 100 pm to 1 mm, the adhesion strength decreases from -100 kPa to less than 25 kPa, imposing limitations on the size and complexity of fibril geometry required to achieve strong adhesion (Progress in Materials Science 2021 , 120, 100823; Soft Matter 2011, 7, 5545-5552) and leading to various practical challenges such as expensive micro / nano-fabrication processes (Small 2012, 8, 2464-2468), susceptibility to lateral collapse, bundling (Progress in Materials Science 2021, 120, 100823), fracture and abrasion (ACS nano 2018, 72, 1273-1284; Nature 2020, 582, 55-59). Despite various proposed optimizations (e.g., modulus gradients (ACS nano 2018, 72, 1273-1284), mushroom-shaped tips (Proceedings of the National Academy of Sciences 2017, 774, E4344- E4353; Advanced Materials Interfaces 2022, 9, 2101764; Soft Matter 2011, 7, 5545-5552) (Progress in Materials Science 2021, 720, 100823), the adhesion strength of mm-scale fibrils remains significantly lower than the intrinsic level of Van der Waals interactions, and achieving both high adhesion strength and easy detachment remains challenging (npj Flex. Electron., 2018, 2(1), 1-14).

[0012] At the array level, the effective adhesion strength (the adhesion force over the apparent contact area) may not scale up with the number of fibrils due to uneven load sharing (Progress in Materials Science 2021, 720, 100823) among fibrils. This uneven load sharing can be caused by interfibrillar mechanical interactions and deformation of the backing layer (Progress in Materials Science 2021, 720, 100823; Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353). Increasing the backing layer stiffness can improve the adhesion efficiency (defined as the ratio of the effective adhesion strength of the fibril arrays under a certain apparent contact area to the theoretical adhesion strength but reduces adhesive system compliance during contact (Adv. Mater. Interfaces, 2019, 6(3), 1801542; Science advances 2020, 6, eaay5120; Polymer Chemistry 2015, 6, 3046-3053), limiting its adaptability to surface textures. Some designs have utilized a releasable stiffer backing6 to address this issue, but only a limited percentage (<26%) (Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353) of the adhesive fibrils’ full potential can be exploited when scaling up the apparent contact area.

[0013] To date, it is still an open challenge to resolve the adhesion paradox (i.e., how to prevent the adhesion strength from decreasing when the adherend becomes rougher) and the switchability conflict (i.e., how to increase the adhesion strength and switchability simultaneously). Due to these limitations, natural adhesives and their synthetic counterparts made of elastomers are limited in loading capabilities (e g., the heaviest animals using adhesives for locomotion is less than 250 g in weight (Proceedings of the National Academy of Sciences 2016, 773, 1297-1302)) or restricted to a limited range of surfaces (Science 2016, 352, 895-896). Contrary to the hairy structure designs of gecko and many insects, another strategy used by animals like snail (Proceedings of the National Academy of Sciences 2019, 116, 13774-13779) realizes conformal contact with rough surfaces through secreting mucus. When dried, mucus becomes stiff and locks the contact, producing very strong adhesion. When activated again, the mucus becomes soft, and the snail can detach from trees and rocks after dormancy.

[0014] Somewhat similar to snails’ mucus, stimulus responsive materials like shape memory polymers (SMPs) are smart materials with dramatically tunable elastic modulus, temporary shape-locking effect, and permanent shape-memory effect (Polym. Chem., 2015, 6(16), 3046- 3053) when transitioning between different material phases (e.g., glass transition (Science advances 2020, 6, eaay5120; Polymer Chemistry 2015, 6, 3046-3053) or hydrationdehydration transition (Proceedings of the National Academy of Sciences 2019, 116, 13774- 13779; Small 2019, 15, 1803870) upon external stimulations. SMPs have been explored as a new class of smart adhesives (Soft Matter, 2017, 13(31), 5317-5323; Langmuir, 2021 , 37(45), 13420-13429; Chem. Sci., 2021 , 12(19), 6472-6487; Adv. Mater., 2021, 33(6), 2000713) for multiple applications. This includes adhesive hooks or stickers (Int. J. Adhes. Adhes., 2018, 81 , 74-78; Adv. Mater. Interfaces, 2019, 6(3), 1801542; ACS Appl. Mater. Interfaces, 2021, 13(23), 27656-27662), micro-assembly (J. Microelectromech. Syst., 2014, 23(5), 1012-1014; Adv. Mater. Technol., 2016, 1(7), 1600098; Int. J. Extreme Manuf., 2021 , 3(3), 035103; ACS nano, 2022, 16(4), 6134-6144), flexible and wearable electronics (Adv. Funct. Mater., 2019, 29(34), 1902720), and soft grippers and soft robotics (Small, 2019, 15(4), 1803870; Small, 2019, 15(50), 1904248; Sci. Adv., 2020, 6(7), eaay5120; Sci. Adv. 2020, 6(5), eaax1464; Adv. Funct. Mater., 2021, 31 (21), 2010872; Soft Robot., 2022, 00(00), 1-9).

[0015] SMPs are stiff in the glassy state with a modulus value of more than 1 GPa. The SMPs can shift from a soft, rubbery state to a stiff, glassy state with a modulus change of more than 3 orders of magnitude by external stimuli through glass transition (Polym. Chem., 2015, 6(16), 3046-3053; Sci. Adv., 2020, 6(7), eaay5120), phase change (Nat. Mater., 2019, 18(8), 874- 882; Mater. Horiz., 2021 , 8(4), 1230-1241) or hydration / dehydration (Proc. Natl. Acad. Sci., 2019, 116(28), 13774-13779). Upon transitioning to the rubbery state, they become very soft, with modulus reduced by three orders of magnitude to less than 1 MPa. When SMPs undergo a rubber-to-glass (R2G) transition after deformation in the rubbery state, they can temporarily retain their deformed configuration, which is referred to as the shape-locking phenomenon. When the stimulation is reversed to return the SMPs to the rubbery state, they regain their original shape, referred to as the shape-memory effect. These effects have endowed the SMP smart adhesives strong and switchable adhesion (ACS Appl. Mater. Interfaces, 2016, 8(51), 35628-35633; Proc. Natl. Acad. Sci., 2019, 116(28), 13774-13779; Sci. Adv., 2020, 6(7), eaay5120; Int. J. Extreme Manuf., 2021, 3(3), 035103; ACS nano, 2022, 16(4), 6134- 6144), These effects have made SMPs a unique material as switchable adhesives (Advanced Materials 2016, 28, 7457-7465; Advanced Functional Materials 2021, 31, 2010297; Advanced Materials 2016, 28, 5088-5092; Journal of Microelectromechanical Systems 2014, 23, 1012- 1014) with good repeatability and superior surface adaptability (Science advances 2020, 6, eaax1464; Advanced Functional Materials 2019, 29, 1902720, Macromolecular Rapid Communications 2022, 43, 2200012, Sci. Adv., 2020, 6(7), eaay5120; Macromol. Rapid Commun., 2022, 43(8), 2200012).

[0016] In particular, it has been found that the adhesion capability of SMPs is greatly enhanced when attached in the rubbery state and detached in the glassy state (Proc. Natl. Acad. Sci., 2019, 116(28), 13774-13779; Sci. Adv., 2020, 6(7), eaay5120), hereafter referred to as R2G adhesion. Cho et al. (Proceedings of the National Academy of Sciences 2019, 116, 13774- 13779) and Linghu et al. (Sci. Adv., 2020, 6, eaay5120) have shown that the SMP adhesives are characterized by ultra-strong adhesion strength of more than 1 MPa and good switchability of up to 1000 simultaneously. Xue et al. (Journal of the Mechanics and Physics of Solids 2015, 77, 27-42) discussed the mechanics of adhesion switchability of SMP adhesives, and found that the shape-memory effect is the main factor for switchability.

[0017] While Xue et al (J. Meeh. Phys. Solids, 2015, 77, 27-42) have discussed the mechanics of adhesion switchability in SMP adhesives based on an energy method, they did not reveal the underlying mechanisms responsible for the tremendous adhesion enhancement. Till now, fundamental mechanics of R2G adhesion has not been systematically investigated.

[0018] In conventional adhesive contact mechanics, the adhesive strength and adhesion regime can be energy-controlled (Proc. Math. Phys. Eng. Sci., 1971 , 324(1558), 301-313; J. Colloid Interface Sci., 1992, 150(1), 243-269) or strength-limited (J. Colloid Interface Sci., 1975, 53(2), 314-326; J. Colloid Interface Sci., 1992, 150(1), 243-269), depending on the value of the so- called Tabor parameter, which is a function of the effective radius of curvature effective modulus theoretical adhesion strength and work of adhesion The interfacial tensile traction at detachment is uniform in the strength-limited regime, but it becomes concentrated at the contact edge and the indenter detaches from the substrate in the form of crack propagation in the energy-controlled regime. The adhesion capability or pull-off force is limited by the value independent of the indentation depth where a is a constant ranging from 5 / 6 to 2 according to the adhesion regime under fixed-grip (or 1.5 - 2 under fixed-load) condition (J. Colloid Interface Sci., 1992, 150(1), 243-269). However, the R2G adhesion of SMPs cannot be directly described by the conventional adhesive contact models. Other than the extraordinary enhancement in adhesion strength by more than an order of magnitude (Proc. Natl. Acad. Sci., 2019, 776(28), 13774-13779; Sci. Adv., 2020, 6(7), eaay5120), the R2G adhesion strength is also dependent on the indentation depth (Sci. Adv., 2020, 6(7), eaay5120). A thorough understanding of the fundamental mechanics of R2G adhesion will be vital to the design and application of SMP adhesives.

[0019] In fact, SMPs has been utilized as smart adhesives, employing either the tunable modulus effect (Soft Matter 2009, 5, 3689-3693; Int. J. Adhes. Adhes., 2018, 81 , 74-78), or the shapememory effect (Soft Matter, 2017, 73(31), 5317-5323; ACS applied materials & interfaces

[0020] 2013, 5, 7714-7717). Particularly noteworthy is the design introduced by Seok Kim et al. (of the National Academy of Sciences 2010, 107, 17095-17100), which proposed microstructured SMP surfaces capable of concurrently achieving high adhesion strength (~1 MPa) and remarkable switchability (>1000). This achievement was realized through controlled contact area manipulation, facilitated by the temperature-mediated collapsing and recovery of SMP micro-pyramids on the adhesive surface. This design, stemming from their prior work54 on microstructured PDMS surfaces, proved successful in applications such as transfer printing and heterogeneous integration of microdevices (Journal of Microelectromechanical Systems

[0021] 2014, 23, 1012-1014; Advanced Materials Technologies 2016, 1, 1600098; Micromachines 2019, 70, 267; ACS applied materials & interfaces 2016, 8, 35628-35633).

[0022] These results predominantly focus on technological aspects and adhere to conventional fundamental mechanics models for elastic bodies. However, these models fall short of explaining the notable adhesion enhancement and unique indentation depth dependence observed in R2G adhesive systems.

[0023] Once the abovementioned hurdles are overcome, such adhesives that addresses the adhesion paradox and switchability conflict on rough surfaces would greatly promote advancements in other technologies, such as in the field of robotics.

[0024] With the rapid development of robots and their replacement of living beings in various fields (Science Robotics 2022, 7, eabg5561 ; Extreme Mechanics Letters 2018, 22, 51-59), including industry (Science Robotics 2021, 6, eabd9461), agriculture (Robotics and automation for improving agriculture, 2019, 137-160; Nature Food 2021, 2, 306-309), healthcare (Science robotics 2021 , 6, eabi8017), micro / nano fabrication (npj Flex. Electron., 2018, 2(1), 1-14), and even in social aspects such as education and services (Science Robotics 2017, 2, eaam5410), the interactions and information exchange between robots and the external world are becoming increasingly crucial (Science Robotics 2021, 6, eabl7022).

[0025] Electronic skins (E-skins) mirroring the exceptional sensing abilities of natural counterparts are revolutionizing robotics by enhancing their interactions with the environment. Skins, the body’s largest sensory organ, serves as a crucial interface between the internal systems and the external environment. The skin tissues contain a complex network of receptors for various stimuli, facilitating somato sensation, and relaying information to our brains about external factors like pressure, temperature, and humidity. Inspired by natural skins, researchers have developed a myriad of E-skins (Science Robotics 2020, 5, eaaz9239) incorporating various sensors, including strain (Advanced Materials Technologies 2021, 6, 2100038), pressure (Advanced Functional Materials 2020, 30, 1907312), sliding (Science Robotics 2018, 3, eaau6914), temperature (Science advances 2019, 5, eaav9653), and electrophysiology (Science robotics 2022, 7, eabn0495) sensors. Extensive research (Advanced Functional Materials 2021, 31, 2104686) has focused on the materials, sensors, data processing, and fabrication, further advancing this field.

[0026] The natural skins possess intricate design and comprehensive functionality, and researchers have been continuously working to approach this natural ideal. Other than sensors in skin tissues, there exist a lot of skin appendages such as hair, feathers, nails, and scales adjacent to skins. These skin appendages serve particular functions beyond sensation, such as protection and lubrication. However, prior studies on E-skins have predominantly concentrated on replicating sensory functions of skin tissues, neglecting the multifaceted functions of natural skin appendages beyond sensation. Especially, there exists a special kind of appendages, the adhesive skin appendages (Progress in Materials Science 2021, 120, 100823, which provide switchable adhesion to enable the living creatures’ extraordinary capabilities in adhering (Proceedings of the National Academy of Sciences 2019, 116, 13774-13779), parasitism (Science Robotics 2022, 7, eabm6695), capturing (Nature Physics 2016, 12, 931-935), climbing (Proceedings of the National Academy of Sciences 2016, 113, 1297-1302) and so on. The adhesive skin appendages have inspired numerous smart adhesives (Progress in Materials Science 2021, 120, 100823; ACS nano 2018, 12, 1273-1284) with adaptable adhesion with gentle pressure, superior to those friction-based designs (Nature communications 2021 , 12, 7177), facilitating the safe and reliable gripping / release of heavy, large or fragile objects and leading to many new applications in various fields, including heterogeneous assembly (Advanced Materials 2022, 34, 2107748), soft robots (Progress in Materials Science 2021, 120, 100823) and soft grippers (Sci. Adv., 2020, 6(7), eaay5120) et al.. Nevertheless, these biomimetic smart adhesives, while providing adaptable adhesion between the high and low values, have experienced restricted adoption within electronic skin (E-skin) applications for robotics. This underutilization of their potential capabilities calls for further exploration and integration in the field of E-skin technology, promising enhanced robotic interactions with the environment. Furthermore, prior bio-inspired smart adhesives designs mainly rely on soft elastomers (K. Kendall, "Introduction to Molecular Adhesion and Fracture: The Adhesion Paradox" in Molecular Adhesion and Its Applications: The Sticky Universe. (Springer US, Boston, MA, 2001)) to from attachment with the objects and exhibit immediate adhesion upon contact, and the low adhesion strength value increases with the high adhesion strength values. In various scenarios, adhesion can hinder robot-environment interactions. For example, in dexterous manipulation (Science robotics 2021, 6, eabd7935), where precise relative movement between the robot’s fingers and objects is necessary, adhesion can be an obstacle. Similarly, in sensing and detection applications (Proceedings of the National Academy of Sciences 2020, 117, 25352-25359), where the robot's end effector must move smoothly across a target surface, excessive adhesion can disrupt this movement. Additionally, when handling extremely lightweight and small objects (Science Robotics 2021, 6, eabd8803) (e.g., microchips, clothes, and papers), adhesion can cause these objects to stick to the robot’s gripper, complicating the release process. This has presented acritical challenge in smart adhesive design, that is, it cannot be used in carrying heavy load and used in scenarios where adhesion should be avoided simultaneously.

[0027] It is apparent that there exist multiple hurdles to overcome in order to arrive at an adhesive suitable for the abovementioned use. To achive so, firstly a thorough understanding of the fundamental mechanics of R2G adhesion will be vital to the design and application of SMP adhesives. Once the fundamental mechanics of the R2G adhesion with SMPs are addressed, it would be possible to reveal the origin and underlying mechanisms of the R2G adhesion enhancement, and to develop an approximate analytical solution to the R2G pull-off force. Further from there, it would be possible as well for one to elucidate the required materials properties, underlying mechanisms, as well as models to predict the adhesion strength and adhesion switchability of the SMP adhesives in addressing the adhesion paradox and switchability conflict on rough surfaces. Upon successfully demonstrating the capabilities of the R2G fibrillar adhesive, it would finally be possible to fabricate adhesive fibrils with different radii and fibril arrays with varying apparent contact areas for said purposes.

[0028] Summary of Invention Here, we study the R2G adhesion from a fundamental adhesive contact mechanics point of view using SMPs. FIG. 3 illustrates the adhesive contact and detachment processes of a rigid sphere on a flat substrate. FIG. 3(A) shows the conventional adhesive contact between a rigid sphere and an elastic substrate while FIG. 3(B) the corresponding R2G adhesion. During the R2G adhesion process, the rigid sphere is first pushed onto the SMP substrate in the rubbery state to an indentation depth of 6. While holding the indentation depth constant, the SMP substrate is brought to its glassy state (e.g., by cooling or dehydration), where the deformation incurred under the rubbery state is now “frozen” due to the shape locking effect of SMP. Detachment under this condition then gives rise to a large pull-off force far beyond that typically associated with an elastic substrate.

[0029] The current invention also addresses the fundamental mechanics of the R2G adhesion with SMPs, to reveal the origin and underlying mechanisms of the R2G adhesion enhancement, and to develop an approximate analytical solution to the R2G pull-off force.

[0030] The current invention also discloses SMPs that can be used to overcome the adhesion paradox as well as switchability conflict on rough surfaces of various materials with different degrees of roughness. A E44 epoxy SMP (FIG. 4(A)(i)), actuated by a thermally controlled glass transition mechanism, is used in this work. The E44 epoxy SMP, cured by Jeffamine D- 230 (see FIG. 4(A)(i) for the chemical structures for the monomers), with a dramatically tunable modulus through the glass transition temperature exhibits the shape-locking effect when cooled below and shape-memory effect when heated above (as illustrated in FIG. 4(A)(ii)). The E44 SMP can be easily cast and cured into desired geometry for characterization and practical applications, as shown in FIG. 4(A)(iii). FIG. 4(B) shows the working principle of the SMP adhesive on rough adherends. When heated to the rubbery state, the SMP becomes soft and can conform to a rough adherend under a preload (FIG. 4(B)(i) and 4(B)(ii)). Cooling the SMP to the glassy state under the preload fixes the conformal contact, rendering a strong R2G adhesion for robust attachment due to the shape-locking effect (FIG. 4(B)(i I) and 4(B)(iii)). When the SMP is re-heated to the rubbery state, it recovers its original shape and loosens its contact with the adherend due to the shape-memory effect, rendering an on-demand, weak rubbery-state adhesion (FIG. 4(B)(iv)).

[0031] FIG. 4(C) summarizes the SMP R2G adhesion for attachment and its rubbery-state detachment measured on glass surfaces with different hms (the root mean square amplitude of surface roughness) values, together with the adhesion strengths of various elastomers. As shown in FIG. 4(C), the adhesion strength of the SMP adhesive in the rubbery state follows the decreasing trend of those of the elastomeric adhesives as increases. In contrast, the SMP R2G adhesion strength increases whe increases. The increase of the SMP R2G adhesion and the decrease of its rubbery-state adhesion as the surface roughness increases lead to simultaneous enhancement in both adhesion capability and adhesion switchability, as shown in FIG. 4(D) . Results in FIG. 4(C) and 4(D) demonstrate that the SMP adhesives can overcome both adhesion paradox and switchability conflict that have long limited the performance of elastomeric adhesives. Detailed discussions are provided in later sections to elucidate the required materials properties, underlying mechanisms, as well as models to predict the adhesion strength and adhesion switchability of the SMP adhesives in addressing the adhesion paradox and switchability conflict on rough surfaces.

[0032] The current invention also disclose a new design paradigm that leverages the JKR-DMT transition of adhesive fibrils’ adhesion regime to overcome the abovementioned limitations. This design enables the development of strong, switchable, and scalable smart fibrillar adhesives by tuning geometrical, elastic, and adhesive parameters, as depicted in FIG. 5(A). We introduce the Rubber-to-Glass (R2G) phase changing fibrillar adhesives concept, utilizing smart materials with two distinct phases: a soft rubber-phase and a stiff glass-phase (Polym. Chem., 2015, 6(16), 3046-3053; Nat. Mater., 2019, 18(8), 874-882). The adhesive transitions between these two phases to achieve the desired adhesion regimes. During attachment (FIG. 5(A)-(i)), the R2G fibrillar adhesive is in the soft rubber-phase, facilitating enhanced contact and sticky attachment (Physical Mesomechanics 2018, 21, 67-75). Upon transitioning to the glass-phase after contact, the fibrillar adhesive exhibits strong R2G adhesion (contact in the rubber-phase and detach in the glass-phase), with uniform stress distribution at each fibril and equal load sharing among fibrils (FIG. 5(A)-(ii)) During detachment, the fibrillar adhesive returns to the soft rubber-phase, which induces detachment in the JKR-like regime at each fibril and uneven load sharing among fibrils (FIG. 5(A)-(iii)), resulting in weak rubber-phase adhesion for easy detachment.

[0033] To demonstrate the capabilities of the R2G fibrillar adhesive, we fabricated adhesive fibrils with different radii (FIG. 5(B)-(i)) and fibril arrays with varying apparent contact areas (FIG. 5(B)-(ii)) using a thermally-controlled E44 shape-memory polymer (E44-SMP) (Polymer Chemistry 2015, 6, 3046-3053). Our proposed breakthrough adhesion design paradigm, anchored in the disclosed in R2G adhesion, shows promise for unmatched adhesion strength, switchability, and scalability.

[0034] Further, we introduce a versatile robotic adhesive skin owning a wide range of adhesion tunability utilizing the smart phase-changing materials as the adhesive skin appendages, which can be used for all kinds of interaction scenarios, whether the adhesion is required or not. This adhesive skin is fully integrated with a surface layer of adhesive fibrils emulating natural adhesive skin appendages make of phase-changing materials, a flexible heater controlling the materials phases of the skin appendages, and a flexible pressure sensor for sensing and detection application. Utilizing the shape memory polymer (SMP) (Int. J. Adhes. Adhes., 2018, 81, 74-78; Polymer Chemistry 2015, 6, 3046-3053) as the skin appendages’ materials, we validate the large adhesion tunability of the proposed adhesive skin through systematic adhesion measurement experiments and finite element simulations, as well as demonstrating the functionality and versatility by integrating the adhesion skins into a robotic hand and using it in scenarios where adhesion should be avoided (the handling of a cloth and detecting of the surface morphologies) or highly required (the grasping of fragile, large and curved, or heavy objects).

[0035] The current invention moves the research on E-skins forward in replicating the natural ideal counterparts by further incorporating the skin appendages. Plus, it can broaden the substantially broadens the range of interactions achievable by robotics across various environmental scenarios, whether adhesion is required or not, representing a significant contribution to the field and promoting the development of enhanced intelligent robotics.

[0036] Aspects and embodiments of the invention will now be discussed by reference to the following numbered clauses.

[0037] 1. A gripping member, the member comprising: a base portion formed from a rubber-to-glass phase transitioning material; and one or more fibrils extending from the base portion made from a rubber-to-glass phase transitioning material, wherein the rubber-to-glass phase transitioning material: can transition from a glass phase to a rubber phase upon application of a stimulus and return to the glass phase in an absence of the stimulus; has an elastic modulus of less than 1 MPa in the rubber phase; and the transition from a glass phase to a rubber phase occurs in the one or more fibrils having a radius of at least 0.1 mm in size.

[0038] 2. The member according to Clause 1, wherein the rubber-to-glass phase transitioning is selected from one or more of a shape memory polymer, a liquid metal, a liquid-crystalline polymer, and a jamming material. 3. The member according to Clause 2, wherein the rubber-to-glass phase transitioning is a shape memory polymer.

[0039] 4. The member according to Clause 3, wherein the shape memory polymer is an E44 epoxy shape memory polymer.

[0040] 5. The member according to any one of the preceding clauses, wherein the size of one or more fibrils are selected to have: a dimensionless paramete in the glass phase, such that it has adhesion properties under a DMT-like adhesion regimen; and a dimensionless parameter ^ in the rubber phase, such that it has adhesion properties under a JKR-like adhesion regimen, where Xfis obtained from formula (1) where R is the radius of a single fibril being the Young’s modulus and Poisson’s ratio of the fibril material, respectively, being the theoretical strength of adhesion, and being the work of adhesion.

[0041] 6. The member according to any one of the preceding clauses, wherein each of the one or more fibrils has a radius that is greater than a critical radius in a rubber phase and less than a critical radius in the glass phase as calculated according to Formula (2) and Formula (3), respectively: where are the Young’s modulus under the rubber phase and glass phase, respectively, and and are as defined in Clause 5.

[0042] 7. The member according to any one of the preceding clauses, wherein when the rubber- to-glass phase transitioning material is an E44 shape memory polymer, then the radius of each of the one or more fibrils may be from 0.5 mm to 3 mm, such as about 2.5 mm.

[0043] 8. The member according to any one of the preceding clauses, wherein the member displays a switchability of from 20 to 2,000, such as greater than or equal to 1 ,000 to less than or equal to 2,000. (i.e., the ratio between the maximum adhesion force for attachment and the minimum adhesion force for detachment)

[0044] 9. The member according to any one of the preceding clauses, wherein the one or more fibrils are a plurality of fibrils, each fibril having a radius of from 0.5 mm to 3 mm, such as from 1.5 mm to 2.5 mm.

[0045] 10. The member according to Clause 9, wherein the plurality of fibrils are arranged in an array, optionally wherein the effective contact area is up to 1 ,000 mm2.

[0046] 11 . The member according to any one of the preceding clauses, wherein each of the one or more fibrils has an aspect ratio of from 1 to 50, such as from 5 to 10, where the aspect ratio is the height of a fibril divided by its radius.

[0047] 12. The member according to any one of the preceding clauses, wherein the stimulus is selected from one or more of heat, light (e.g. Ultraviolet light), an electrical current, and pressure (e.g. negative pressure relative to standard pressure).

[0048] 13. The member according to any one of the preceding clauses, wherein one or both of the following apply:

[0049] (i) the adhesion efficiency of the member is greater than 50%, such as greater than 55%, such as greater than 57.8%; and

[0050] (ii) the adhesion strength of the member is more than 1 MPa, such as from 1 MPa to 5 MPa, such as from 1.5 MPa to 2 MPa, such as about 1.89 MPa.

[0051] 14. A gripping apparatus comprising: at least one gripping member according to any one of Clauses 1 to 13; and a stimulating means or apparatus, wherein the stimulating means or apparatus is configured to supply a stimulus to transition the gripping member from a glassy phase to a rubber phase.

[0052] 15. The gripping apparatus according to Clause 14, wherein the stimulating means or apparatus is formed integrally with the gripping member.

[0053] 16. The gripping apparatus according to clause 15, wherein the stimulating means or apparatus is formed integrally with the base of the gripping member. 17. A gripping device, comprising: at least one gripping apparatus according to any one of Clauses 14 to 16; and a substrate, wherein the at least one gripping apparatus is attached to the substrate,

[0054] 18. The gripping device according to Clause 17, wherein the substrate is a robotic hand or a glove suitable for use by a human.

[0055] 19. A method of gripping using a gripping device according to Clause 17 or Clause 18, wherein the method comprises:

[0056] (A) providing an object for gripping and a device according to Clause 17 or Clause 18;

[0057] (B) applying a stimulus to the at least one gripping apparatus to transition the one of more fibrils of the rubber-to-glass phase transitioning material into a rubber phase, then bringing the one or more fibrils into contact with a surface of the object to form an adhesion between the surface of the object and the one or more fibrils; and

[0058] (C) removing the stimulus to return the one or more fibrils to a glass phase, thereby gripping the object.

[0059] 20. The method according to Clause 19, wherein the method further comprises the step(s):

[0060] (D) optionally physically manipulating the gripped object; and

[0061] (E) reapplying the stimulus to the at least one gripping apparatus to release the object.

[0062] Drawings

[0063] FIG. 1 depicts the summary of adhesion strength of (A) bulk elastomeric adhesives and (B) bioinspired hairy elastomeric adhesives on rough surfaces. Hexagon for silicone sphere (Mathematical and Physical Sciences 1975, 345, 327-342), pentagon for PDMS sphere (Physical review letters 2005, 95, 124301), squares for PDMS bulk and mushroom (Acta Biomaterialia 2012, 8, 282-288), diamonds for PVS bulk and mushroom (Journal of The Royal Society Interface 2013, 10, 20130620), upper triangles for PU mushroom with a soft tip design (Royal Society Open Science 2017, 4, 161105), lower triangles for PVS microflaps with a contact splitting design (Beilstein Journal of Nanotechnology 2019, 10, 1-8), and circles for SSA surface with a terminated film design (ACS Appl Mater Interfaces 2022, 14, 46239-46251). Insets show the magnifications of the data points when the root-mean-square (rms) roughness values are small.

[0064] FIG. 2 depicts the summary of adhesion switchability versus adhesion strength of elastomeric adhesives: PU (data from (Advanced Materials 2014, 26, 4901-4906)), PUA (data from (Advanced Functional Materials 2021, 31, 2107023)), LCE (data from (Bioinspired photocontrollable microstructured transport device. Sci. Robot 2 (2017))), CNT (data from (47)), PEGDMA hydrogel (data from (Science 2008, 322, 238-242)) and PDMS (data from (Progress in Materials Science 2021, 720, 100823)). The shadowed area shows the typical adhesion switchability versus the adhesion strength range of gecko (Nature 2000, 405, 681- 685; Mechanics of Materials 2005, 37, 275-285).

[0065] FIG. 3 depicts the schematic illustrations of (A) adhesive contact and detachment between a rigid sphere and an elastic substrate and (B) the corresponding R2G adhesion with an SMP substrate.

[0066] FIG. 4 depicts (A-B) the design, mechanism, and (C-D) the performance of the SMP Smart Adhesive. (A) Material and adhesion characterization systems adopted in this work, (i) Chemical structures of the epoxy precursors of the SMP used, (ii) Illustrations of the cured networks and working principles of the SMP. (Hi) Photograph of the clamp-mounted SMP sample held on top of a rough glass surface for the adhesion test. (B) Illustrations of the SMP adhesion on rough surfaces, (i-ii) Full contact is achieved when the SMP is heated to the soft, rubbery state, (ii-iii) Strong R2G adhesion is achieved when the SMP is cooled to the stiff, glassy state, with the full contact configuration from the rubbery state frozen due to the temporary shape-locking effect, (iii-iv) Weak rubbery-state adhesion for on-demand detachment is achieved when the SMP is reheated to the rubbery state, where the contact reverts to a tip contact configuration due to the shape-memory effect of the SMP. (C) Adhesion strengths of SMP adhesives compared to those of adhesives made of typical elastomers on rough surfaces. The stars and asterisks represent the R2G and rubbery-state adhesion of SMP, respectively, measured on glass substrates (preload = 255 kPa). The other dots are adhesion strength values reported in the literature for elastomeric adhesives. Hexagon (A. Mathematical and Physical Sciences 1975, 345, 327-342): silicone rubber, pentagon (Physical review letters 2005, 95, 124301) and square (Acta Biomaterialia 2012, 8, 282-288): PDMS - polydimethylsiloxane, prism (Journal of The Royal Society Interface 2013, 70, 20130620): PVS - poly(vinylsiloxane), upper triangle (Royal Society Open Science 2017, 4, 161105): PU - polyurethane, lower triangle (Beilstein Journal of Nanotechnology 2019, 70, 1-8): PVS - polyvinylsiloxane, circle (ACS Appl Mater Interfaces 2022, 74, 46239-46251): Dow soft skin adhesive. (D) Adhesion switchability versus adhesion strength of the SMP compared to typical elastomers. The spheres represent the adhesion switchability of SMP adhesives and squares the adhesion switchability of elastomeric adhesives reported in the literature (Progress in Materials Science 2021, 720, 100823). The shaded area in the dotted-lined oval shows the range of adhesion switchability and adhesion strength of gecko (Nature 2000, 405, 681-685; Mechanics of Materials 2005, 37, 275-285).

[0067] FIG. 5 depicts the overview of the R2G fibrillar adhesives. (A) Illustrations showing the working principle of the R2G phase changing fibrillar adhesive, (i) On-demand attachment in the rubber-phase. During the contact process, the adhesive fibrils become soft upon transitioning into the rubber-phase under stimulation, which facilitates sticky attachment of the adhesive fibrils on the adherend when Dahlquist criterion (i 1 MPa) is satisfied, (ii) Strong and scalable adhesion in the DMT-like regime. After attachment, the adhesive fibrils become stiff when transitioning into the glass-phase upon reverse stimulation. When loaded in the stiffstate, the adhesive fibrils work under the DMT-like regime with uniform stress distribution at every fibril-adherend contact interface when and equal load sharing among different fibrils, resulting in strong and scalable adhesion to support heavy load, (iii) On-demand detachment in the JKR-like regime. The adhesive fibrils become soft upon transitioning into the rubber-phase under restimulation. When loaded in the soft-state, the adhesive fibrils work under the JKR-like regime with high stress concentration at the periphery of every fibriladherend interface when and uneven load sharing among different fibrils as well as progressive detachment from periphery inwards, resulting in weak adhesion for easy detachment. (B) Photos of (i) single adhesive fibrils with different fibril radii and (ii) fibril arrays with different apparent contact areas made of E44-SMP. (C) Adhesion strength scaling performance of the R2G adhesive fibrils made of E44-SMP at the single fibril level compared to those made of conventional elastomers. The R2G adhesive fibrils made of E44-SMP can sustain the DMT-like strong adhesion with fibril radius scaled all the way up to the millimetersize, breaking through the scaling limit of elastomeric adhesives. (D) Adhesion switchability versus adhesion strength of the R2G adhesive fibrils made of E44-SMP as compared to those made of elastomers actuated by controlled peeling (squares), controlled buckling (upper triangles), and controlled shearing (lower triangles). (E) Adhesion strength scaling performance of the R2G fibrillar adhesives made of E44-SMP at the array level compared to fibrillar adhesives made of elastomers. The SMP fibrillar adhesive showed an improved adhesion efficiency (57.8%) over the elastomeric ones even when the apparent contact area is scaled up to 1000 mm2. (F) Array adhesion switchability of the R2G fibril arrays and bulks made of E44-SMP. The switchability can be increased by two orders of magnitude through controlled buckling.

[0068] FIG. 6 depicts the details of the contact test demonstrating the shape-conformability and temporary shape-locking capabilities of the E44 SMP. (A) Illustrations of the contact test procedure, (i) An SMP block (20 mm x 20 mm x 7.5 mm) is brought into contact with a serrate steel surface in the glassy state (at about 25 °C) with a preload of 200 kPa, and (ii) then heated to 90°C to facilitate conformal contact, (ill) Upon cooling to the glassy state, the deformation is temporarily fixed, which (iv) allows direct measurement of the deformed shape after the SMP sample is separated from the substrate. (B) The results of the measurements showing shapeconformability and shape-locking capabilities. Photos of (i) the SMP block sample (20 mm x 20 mm x 7.5 mm) and (ii) the serrate steel substrate, (ill) SMP sample in contact with the serrated steel surface in the glassy state, (iv) Magnification of the contact interface in the glassy state, (v) SMP sample in contact with the serrated steel surface in the rubbery state, (vi) Magnification of the contact interface in the rubbery state, (vii) Photo of the SMP with the deformation fixed after separated from the serrated surface of the steel substrate. (C) Comparison of the 3D surface morphologies of (i) the serrated steel surface and (ii) the deformed SMP sample. The 3D morphology was measured using a Keyence VK-X1100 Optical Profilometer.

[0069] FIG. 7 depicts the material properties of the thermally controlled E44 epoxy SMP adhesive. (A) Storage modulus of the E44 SMP at different temperatures determined by DMA test (Test frequency 1 Hz, temperature rise rate 3°C / min). (B) Demonstration of the stickiness of SMP in the rubbery state on a highly roughened steel substrate with serrated surface (peak-to- valley amplitude A = 0.5 mm, wavelength Photographs of the serrated steel surface in contact with the SMP block sample (20 mm x 20 mm x 7.5 mm) in the (i) glassy and (ii) rubbery states. The magnified surfaces in (Hi) and (iv) show the corresponding contact interfaces between the serrated steel substrate and the SMP block in the (i) glassy (25°C) and (ii) rubbery (90°C) states; Preload = 200 kPa. (C) Demonstration of the shape-locking effect of SMP. (i) Photograph of the SMP block (20 mm x 20 mm x 7.5 mm) as prepared, (ii) The cross-sectional view of the contact interface between the SMP block and the serrated steel substrate (peak-to-valley amplitude A = 500 wavelength 1 mm) in the rubbery state (90 °C); preload = 200 kPa. (iii) Photograph of the SMP block separated from the serrated steel substrate after cooling to room temperature, (iv) Comparison of surface profiles between the SMP block in (iii) and the serrated steel substrate. (D) Demonstration of the shape-memory effect of SMP. SEM Images of the SMP surface (i) as prepared, (ii) after contact with a rough glass, and (iii-v) restored to the original shape by heating after (iii) one, (iv) ten, and (v) twenty contact restoration cycles, (vi) 2D profiles of SMP surfaces corresponding to the SEM images. Ra denotes the arithmetic average roughness.

[0070] FIG. 8 depicts the finite element simulation setup of the R2G Adhesion with SMP. (A) Schematic illustration of the axisymmetric model for finite element simulations of the R2G adhesion. (B-C) Thermomechanical response of an SMP strip (25 mm x 5 mm x 1 mm) under uniaxial tension. (B) Illustrations and (C) temperature / strain-time curves during a full cycle of deformation, shape locking and shape recovery.

[0071] FIG. 9 depicts the FEA results of the nominal stress-strain curves of an SMP strip (25 mm x 5 mm x 1 mm) under uniaxial tension in the (A) glassy (30°C) and (B) rubbery (90°C) states.

[0072] FIG. 10 depicts the mixed-mode energy release rate near the crack front during detachment.

[0073] FIG. 11 depicts the comparison of contact radius values from FEA simulations and predictions from modified Hertz theory for different indenter radii and indentation depths.

[0074] FIG. 12 depicts (A-B) the dependence of the R2G pull-off force on indentation depth and (C-D) the associated adhesion enhancement results. (A) Experimental data and FEA simulation of the R2G pull-off force in the case of a sphere with radius of 1 mm and comparison to predictions from the conventional adhesive contact models under the same adhesion parameters. (B) Normalized R2G pull-off force vs. normalized indentation depth according to Eq. (9). (C) Comparison of pull-off forces in elastic states (E = 1 MPa and 2 GPa) and R2G adhesion (6 / R = 0.1) under different indenter radii. (D) Comparison of adhesion enhancement ratios due to the modulus effect (the ratio between the pull-off forces for the elastic states E= 2 GPa and E = 1 MPa ), the net R2G effect (the ratio between the R2G pull-off forces with 5 / R - 0.1 and that of the basic elastic state with E - 1 MPa), and the shape locking effect for different indenter radii.

[0075] FIG. 13 depicts the illustration of (A) the normal traction distribution and (B) the integration unit at the contact interface in the R2G adhesion with a rigid sphere.

[0076] FIG. 14 depicts the comparison of the evolution of normal contact stress distribution due to modulus effect vs. shape locking effect in R2G adhesion (R = 1 mm, <5 = 0.01 mm). (A) Typical normal contact force-displacement curve during the contact and detachment process for elastic adhesion with modulus of E -2 GPa. (B) Evolution of the normal contact stress distribution over the contact area during the pull process at the corresponding pulling stages numbered in (a) for the elastic adhesion with modulus of E -2 GPa. (C) Typical normal contact force-displacement curve for the contact and detachment process for the R2G adhesion. (D) Evolution of the normal contact stress distribution over the contact area during the pull process at the corresponding pulling stages numbered in (C) for the R2G adhesion. In (B) and (D), the dashed line, dot dashed line and dotted line indicate the compression-dominant states and the solid lines the tension-dominant states. The contact pressure is normalized by its maximum value during contact and the pulling force is normalized by iwiadR. The compression stress is normalized by its maximum value during contact and tensile stress normalized by the theoretical adhesion strength. Insets in (A) and (C) show the normal stress distribution in the substrate at the pull-off point.

[0077] FIG. 15 depicts the flowchart showing the procedure for calculating the effective radius of curvature for the R2G adhesion between a rigid sphere and an SMP substrate.

[0078] FIG. 16 depicts the effective radius of curvature of the contact surface and adhesion coefficient in the R2G adhesion with an SMP substrate. (A) Effective radius of curvature as a function of normalized indentation depth. (B) Transition of adhesion regime due to increased

[0079] FIG. 17 depicts the comparison of the Kendall solution and DB model of flat punch adhesion with the FEA results of the R2G adhesion between a rigid spherical indenter and an SMP substrate under different indenter radii at increasing normalized contact radius (related to indentation depth).

[0080] FIG. 18 depicts the comparison of stress distribution in the R2G adhesion with a rigid sphere and that in the rigid flat punch / elastic flat substrate adhesive system (E = 2 GPa). (A) Illustration of the adhesive system configuration and the adopted linear traction-separation law. (B) Normal stress (S22) distribution in the substrate and normal traction distribution along the contact interface at the pull-off point.

[0081] FIG. 19 depicts experimental validation of the flat-punch model prediction for the pull-off force in the R2G adhesion between a rigid spherical indenter and an SMP substrate with various radii under different indentation depths.

[0082] FIG. 20 depicts the adhesion map of the R2G adhesion between a rigid sphere and an SMP substrate based on the flat-punch approximation. Simulation data from FEA, with the pink circles denote the strength-limited regime (defined as having adhesion strength within 5% of the theoretical adhesion strength), the black hexagons denote the transitional regime, and the blue rectangles denote the energy-controlled regime (defined as having adhesion strength within 5% of the Kendall solution).

[0083] FIG. 21 depicts the Fabrication process of the SMP adhesive sample. (A) Illustration of the fabrication process. (B) Photographs of the molds and the sample. The SMP sample is dyed for better visualization in (iii). FIG. 22 depicts the illustrations of the test process for (A) SMP R2G adhesion, (8) SMP glassystate adhesion, (C) SMP rubbery-state adhesion and (D) PDMS adhesion.

[0084] FIG. 23 depicts the influence of the preload on SMP (A) rubbery-state and (B) R2G adhesion strength on glass samples with different surface roughness values. The surface roughness increases from samples #0 to #4.

[0085] FIG. 24 depicts the surface morphology of the smooth glass sample (sample #0) used in this study. (A) 3D morphology of the smooth glass sample measured with Cypher ES Environmental AFM (Asylum Research / Oxford Instrument) and (B) its surface roughness power spectrum.

[0086] FIG. 25 depicts the Surface morphology of rough glass samples #1 - #4 used in this study. (A-D) The microscopic optical images and height distributions of the rough glass samples (measured by Keyence VK-X1100 Optical Profilometer with a pixel size resolution of ~ 0.6837 pm). (E) The corresponding power spectra of the surface roughness of the rough glass samples in A-D.

[0087] FIG. 26 depicts the Root mean square amplitude and slopes of different materials. Measured by Keyence VK-X1100 Optical Profilometer with a pixel size resolution of ~ 0.6837 pm.

[0088] FIG. 27 depicts the experimental validation of scaling laws for (A) SMP R2G adhesion, (B) SMP glassy-state adhesion, (C) SMP rubbery-state adhesion compared to (D) PDMS adhesion on glasses with different roughness values. Insets illustrate the adopted mechanics models. The SMP and PDMS adhesive samples are both cylinders with radius R = 2.5 mm and aspect ratio H / R = 4. The results were measured under a preload of 255 kPa. The SMP R2G adhesion is proportional to the surface area, while the SMP glassy-state adhesion, SMP rubbery-state adhesion and PDMS adhesion are all proportional to the reciprocal of (the root mean square slope) of the rough surfaces. (No fitting parameters are included in the theoretical predications)

[0089] FIG. 28 depicts the stickiness of SMP adhesives on rough glasses. The SMP adhesive sample is a cylinder with radius R = 2.5 mm and aspect ratio (A) Normalized adhesion strength of SMP rubbery-state adhesion on rough glasses under different preloads. The normalization standard is the value of SMP rubbery-state adhesion on smooth glass (sample #0) under a preload of 255 kPa. (B) Normalized adhesion strength of SMP R2G adhesion on rough glasses under different preloads. The normalization standard is the value of SMP R2G adhesion strength on smooth glass (sample #0) under a preload of 255 kPa. The vertical orange dashed lines represent the critical condition for the stickiness of elastomeric adhesives, above which indicating a large decay of adhesion strength.

[0090] FIG. 29 depicts the measured (A) adhesion strength and (B) adhesion switchability of SMP adhesive on various adherends. The SMP adhesive sample is a cylinder with radius R = 2.5 mm and aspect ratio H / R - 4. Results are measured under a preload of 255 kPa.

[0091] FIG. 30 depicts the materials of the R2G fibrillar adhesive. (A) Key materials parameters governing the adhesion regime of an adhesive fibril, (i) Illustrations showing the mechanics model and key parameters determining the transition between JKR-like and DMT-like adhesion regimes of an adhesive fibril on a rigid substrate. Summary and comparison of the

[0092] (i) adhesion strength, (ii) work of adhesion and (iii) elastic modulus values of typical materials used for the fibrillar adhesives. (B) Molecular structures of the E44-SMP (i) precursors and (ii) crosslinked networks, and (iii) the illustrations of related working principle. (C-D) Mechanical behaviors of the E44 epoxy SMP. (C) Storage modulus of the E44-SMP under different temperatures obtained from DMA tests. (D) Typical strain-stress curves of the E44-SMP in the glass-phase (30 °C) and rubber-phases (90 °C), at the tension speed 10 mm / min. Inset shows the hyperelastic fitting of the rubber-phase strain-stress curves using Ogden model. (E-F) Adhesive parameters of the E44-SMP. (E) Pull-off forces measured on the E44-SMP samples with different thicknesses utilizing a rigid glass punch (20 mm in diameter). (F) Fitted adhesive parameters (theoretical adhesion strength and work of adhesion) between the E44-SMP and the glass punch from the measured pull-off forces in (E).

[0093] FIG. 31 depicts the details for the measurement of the adhesive parameters of the E44-SMP. (A) Illustration of the pull-off test platform. (B) Surface morphology of the (i) glass punch and

[0094] (ii) the E44-SMP sample, (c) Illustrations of the measurement processes of the (i) R2G and (ii) rubbery adhesion parameters. (D) Representative curves showing the dependence of the pull-off force on the preload under the (i) R2G condition and (ii) rubber-phase (R =10 mm, t = 10 mm).

[0095] FIG. 32 depicts the statistics of the adhesion strength scaling performance of elastomeric adhesive fibrils under different fibril radii for homogeneous flat fibrils (data collected from Bioinspiration & biomimetics 2015, 10, 066002; Advanced Functional Materials 2016, 26, 4687-4694; Applied Physics A 2020, 126, 1-9; Advanced Functional Materials 2021 , 31, 2104636; ACS Applied Materials & Interfaces 2015, 9, 1036-1044; ACS nano 2018, 12, 1273- 1284; Nature materials 2003, 2, 461-463; Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena 2006, 24, 331-335; Nature 2007, 448, 338-341 ; Advanced materials 2007, 19, 3833-3837; Science 2008, 322, 238-242; Advanced Materials 2010, 22, 2125-2137; Faraday discussions 2012, 156, 235-241 ; Materials Letters 2013, 92, 409-412; Soft Matter 2013, 9, 1985-1996; Chinese Science Bulletin 2012, 57, 1343-1349; Applied Physics Letters 2015, 106, 201604; Soft Matter

[0096] 2020, 16, 9534-9542), gradient fibrils (data collected from ACS Applied Materials & Interfaces 2015, 9, 1036-1044; ACS nano 2018, 12, 1273-1284; Applied Physics Letters 2015, 706, 201604; Soft Matter 2020, 16, 9534-9542), mushroom fibrils (data collected from ACS Applied Materials & Interfaces 2015, 7, 24127-24135; Bioinspiration & biomimetics 2015, 70, 066002; Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353; Journal of the Mechanics and Physics of Solids 2019, 123, 20-27; Soft Matter 2019, 15, 5827-5834; ACS Applied Materials & Interfaces 2020, 12, 27717-27725; Small 2021, 17, 2005493; Advanced Materials Interfaces 2022, 9, 2101764; Advanced Materials 2022, 34, 2107748; Advanced Functional Matehals 2022, 32, 2109076; Royal Society Open Science 2017, 4, 161105; ACS applied materials & interfaces 2009, 7, 849-855; Small 2012, 8, 3009-3015; Science Advances

[0097] 2021, 7, eabf5695) and gradient-mushroom (G-M) fibrils (data collected from Small 2021, 17, 2005493; Royal Society Open Science 2017, 4, 161105; Advanced Functional Materials 2021, 31, 2107023; Science Advances 2021, 7, eabf5695).

[0098] FIG. 33 depicts the summary of the adhesion switchability versus adhesion strength of adhesive fibrils made of elastomers under different adhesion actuation methods. Squares - elastomeric adhesive fibrils actuated by controlled peeling (Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353; Soft Matter 2019, 15, 5827-5834; ACS Applied Materials & Interfaces 2020, 72, 27717-27725; Advanced Materials 2022, 34, 2107748; Advanced Functional Materials 2021, 31, 2107023; Science 2008, 322, 238-242; Advanced Materials 2014, 26, 4901-4906; Advanced Materials 2016, 28, 7457-7465; Nanoscale 2013, 5, 11876-11884; ACS applied materials & interfaces 2014, 6, 19247-19253; ACS applied materials & interfaces 2014, 6, 1345-1350; Soft Matter, 2019, 75(1), 30-37; Advanced Functional Materials 2018, 28, 1706498; Song, S., Majidi, C. & Sitti, M. in 2074 IEEE / RSJ International Conference on Intelligent Robots and Systems. 4624-4629 (IEEE); Advanced Materials Interfaces 2019, 6, 1900875; ACS Applied Materials & Interfaces 2022, 74, 31448-31454; Applied Physics Letters 2012, 700, 171909). Upper triangles - elastomeric adhesive fibrils actuated by controlled buckling (ACS Applied Materials & Interfaces 2015, 7, 24127-24135; Journal of the Mechanics and Physics of Solids 2019, 123, 20-27; Advanced Functional Materials 2021, 22, 1246-1254; International Journal of Precision Engineering and Manufacturing 2017, 18, 133-137). Lower triangles - elastomeric adhesive fibrils actuated by controlled shearing (Advanced Materials Interfaces 2022, 9, 2101764; ACS nano 2018, 12, 1273-1284; Advanced materials 2007, 19, 3833-3837; Applied Physics Letters 2015, 106, 201604; Advanced Functional Materials 2021, 22, 1246-1254; ACS Applied Materials & Interfaces 2016, 8, 10029-10037; Bioinspiration & Biomimetics 2020, 15, 056007; Micro and Nano Engineering 2020, 7, 100052; Composites Science and Technology 2020 192, 108115; Advanced Materials 2016, 28, 5088-5092; Smart Materials and Structures 2010, 20, 015014).

[0099] FIG. 34 depicts the photos of (A) the SMP fibril arrays and (B) single fibril adhesives with comparable apparent contact areas. The apparent contact area of the SMP fibril arrays is defined by the total area an array covers.

[0100] FIG. 35 depicts the scaling performance of the load capabilities of E44-SMP R2G fibril arrays compared to those of E44-SMP single fibrils and existing elastomeric fibril arrays. Data for elastomeric fibrils are from literatures: PDMS (Bioinspiration & biomimetics 2015, 10, 066002; Advanced Functional Materials 2016, 26, 4687-4694; Proceedings of the National Academy of Sciences 2017, 114, E4344-E4353; Journal of the Mechanics and Physics of Solids 2019, 123, 20-27; Advanced Functional Materials 2022, 32, 2109076), CNT (Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena 2006, 24, 331-335; Science 2008, 322, 238-242; Advanced materials 2007, 19, 3844-3849), PU (Royal Society Open Science 2017, 4, 161105; ACS applied materials & interfaces 2009, 1, 849-855) and PS (Trends in Biotechnology 2021 , 39, 150-164).

[0101] FIG. 36 depicts the design, validation, and characterization of the R2G fibrillar adhesive at the fibril level using E44-SMP. (A) Theoretical predictions and FEA validations of the adhesion regimes of the SMP adhesive fibril under various radii. With the purple shadowed ‘work window’ indicating the range of the fibril radius within which the R2G and rubber-phase adhesion regimes of a single fibril are DMT-like and JKR-like, respectively. (B) FEA simulation results of the detachment processes of the SMP adhesive fibril in the work window (R = 1 mm) under the (i) R2G condition and (ii) rubber-phase. (C) Normal stress distribution at the SMP adhesive fibril - glass substrate interface with various fibril radii under the (i-iii) R2G condition and (iv-vi) rubber-phase. (D) Measured adhesion strength of an individual SMP adhesive fibril with a glass substrate under different radii. Inset shows the photo of an SMP adhesive fibril (R =1 mm, h = 4 mm). (E) Snapshots of the detachment process of an individual SMP adhesive fibril (R =1 mm, h = 4 mm) under the (i) R2G condition and (ii) rubber-phase. (F) Switchability of an individual SMP adhesive fibril under different radii. (G) Measured adhesion strength of the SMP adhesive fibril during 30 repeated cycles on a glass adherend. (H) Measured adhesion strength of an individual SMP adhesive fibril (R = 1 mm) under the R2G condition and rubber-phase with various aspect ratios. (I) Influence of the aspect ratio on the switchability of the SMP adhesive fibrils (R = 1 mm). Insets show the configurations of the deformed SMP fibril under a large preload.

[0102] FIG. 37 depicts the detailed setup for the Finite Element Analysis (FEA) simulations. (A) Illustration of the FEA simulation setup. (B) Linear traction-separation law used to model the adhesive interactions.

[0103] FIG. 38 depicts the FEA simulation results of SMP fibril. (A) calculated R2G adhesion strength normalized by the R2G theoretical adhesion strength = 1.893 MPa and (B) calculated rubbery adhesion strength normalized by the rubbery theoretical adhesion strength = 0.4202 MPa under various fibril radii. The adhesion regime is regarded as DMT-like when the adhesion strength is larger than 80% of the theoretical value. (C-D) Typical normal adhesion stress distribution at the contact interface under (A) the R2G condition and (B) rubber-phase.

[0104] FIG. 39 depicts the details of the adhesion test of a single SMP fibril. (A-B) Illustrations showing the test process of the adhesion strength under the (A) R2G condition and (B) rubberphase. (C-D) Representative curves showing the dependence of the R2G and rubbery pull-off forces of a single SMP fibril (R =2.7 mm, h = 10 mm) on the preloads.

[0105] FIG. 40 depicts the measured adhesion strength of an individual SMP fibril under different preloads in the rubber-phase under (A) a small (R = 1 mm, h = 2 mm) and (B) large (R = 1 mm, aspect ratio. Insets show photographs of the SMP fibril under different preloads.

[0106] FIG. 41 depicts the FEA simulation and characterization of the E44-SMP R2G fibrillar adhesive at the array level. (A) Comparison of the 2D-FEA results of the normal stress distributions at the contact interface with comparable apparent contact areas in the (i, iii) single fibril adhesive (R = 29 mm) and (ii, iv) fibril arrays (R = 2.5 mm, center distance between fibrils w = 6 mm) (i, ii) under the R2G conditions and (iii, iv) in the rubber-phase. (B) Stability and durability test of the SMP fibril arrays (fibril radius R = 2.5 mm and fibril height h = 10 mm), (i) Photo of the fibril arrays as prepared, (ii) Force-cycle curves of the contact-detach test of fibril arrays in rubbe phase. (iii) Photo of the fibril array after 100 test cycles. FIG. 42 depicts the static and dynamic stability of the SMP fibril arrays with large aspect ratios. (A) Theoretical predictions of the critical aspect ratio values under which lateral collapse and bundling of the SMP fibrils in the rubber-phase do not occur under static conditions. (B) Photos and microscope images showing the SMP fibril arrays as prepared (i - h / R = 10, Hi - h / R = 20, v - h / R = 30) and after 100 repeated test cycles (ii - h / R = 10, iv - h / R = 20, vi - h / R = 30). Fibril radius R = 250 gm and compression displacement = 1 mm. all of the fibrils are compressed to buckling in each test.

[0107] FIG. 43 depicts the demonstrations of utilizing the SMP fibrillar adhesives as soft grippers and detachable superglues. (A) Demonstrations of the potential to use (i) the R2G SMP adhesive fibril (R = 2.5 mm, h = 10 mm) as a soft gripper to grip various objects including (ii) a pottery disk (41 g), (iii) a raw egg (58 g), (iv) a cobble stone (241 g) and (v) a paper tube for potato chips (170 g). (B) Photos of the SMP fibrillar adhesives of various sizes adhered to a glass ceiling and holding heavy objects, (i-ii) A single adhesive fibril can support (i) a wrench (0.3 kg) and (ii) a drill (1 .56 kg), (iii-iv) An array of seven fibrils can support (iii) a backpack (10 kg) and (iv) a tank of water (15 kg), (v-vii) An array of 19 fibrils can support (v) a suitcase (30 kg) and (vi) a wood box (40 kg), (vii) An array of 37 fibrils can support a bag of steel weights (60 kg). Inset shows the photo of an SMP fibrillar adhesive hook, the SMP is dyed blue for better visualization. (C) Illustrations and snapshots showing the on-demand and benign detachment process of the SMP fibrillar adhesive hook by heating the SMP fibrils using a hair dryer. (D) Summary and comparison of the performance of different fibrillar adhesive designs.

[0108] FIG. 44 depicts the demonstrations of the potential to use the R2G SMP adhesive fibril (R = 2.5 mm, h = 10 mm) as a soft gripper to grip heavy and rough objects. (A) Gripping a wood cube (60 mm x 60 mm x 60 mm, 64 g) using the SMP adhesive fibril, (i-iv) Gripping process - (i) heating the SMP adhesive fibril, (ii) contacting with the wood cube (preload = 254.6 kPa), (iii) cooling the SMP adhesive fibril to the glass-phase, (vi) gripping the wood cube, (v-viii) Release process - (v) heating the SMP adhesive fibril, (vi) shape recovery of the SMP adhesive fibril, (vii) releasing the wood cube, (vii) wood cube released. (B) (i) Photo of the SMP adhesive fibril (R = 2.5 mm, h - 10 mm) with a backing plate (D = 25 mm, t - 2 mm). The SMP sample is dyed blue for better visualization, and (ii-v) photos of the SMP adhesive holding glass bottles with different volumes filled with water (ii, 100 ml, 292 g; iii, 250 ml, 535 g; iv, 500 ml, 970 g; vi, 1000 ml, 1660 g).

[0109] FIG. 45 depicts the illustrations of the set-up used to measure the heating process of the SMP: (A) fibril arrays and (B) a single fibril adhesive. FIG. 46 depicts the heating process of a SMP single fibril adhesive and fibril arrays using a hair dryer. (A) Temperature evolution at the center of the central fibril of the SMP fibril arrays in contact with the acrylic adherend as compared to that at the center of the large single fibril with comparable apparent contact area when heated by a hair dryer laterally. (B) Time required to heat the center of the SMP fibril tip in contact with the adherend to 60 °C using a hair dryer. Inset shows the temperature distribution of the SMP fibril arrays (apparent contact area 1194.8 mm2) and the single fibril (apparent contact area 1017.9 mm2) after heated for 3 minutes. The small fibril refers to a fibril with a radius R = 2.5 mm, the fibril arrays are shown in FIG. 34(A) and single fibril adhesives in FIG. 34(B).

[0110] FIG. 47 depicts the fabrication process for adhesive skin patches. (A) 3D printing technology was utilized to create a rigid mold. (B) Liquid Ecoflex precursors were poured into the rigid mold and subsequently cured. (C) The Ecoflex soft mold was acquired after demolding. (D) Liquid adhesive precursors were introduced into the Ecoflex soft mold. (E) The flexible heater was positioned and integrated into the liquid adhesive precursors. (F) Following the curing process and demolding, adhesive skin patches featuring an adhesive fibrillar top layer and embedded flexible heaters were produced. These patches were configured into various geometries to accommodate different regions of the robotic hand.

[0111] FIG. 48 depicts the overview of the SMP versatile robotic adhesive skin. (A) Integrated robotic adhesive skin system featuring embedded flexible heaters and pressure sensors mounted on a robotic hand, (i) Photograph of the robotic adhesive skin system integrated on a robotic hand, (ii) Close-up view of the adhesive skin patches on the fingertips, (iii) Three-dimensional model of the surface fibrillar structures of the adhesive skin, characterized by hexagonal arrays of hexagonal fibrils, (iv) Exploded illustrations displaying individual components of the adhesive skin patch. (B) Operating principle of the on-demand attachment and detachment of the SMP adhesive skin, (i) Key material properties of the SMP as the materials system for the versatile robotic adhesive skin, (ii) Schematic representation showing (1) minimal contact and adhesion when contacting in the glassy phase, (2) on-demand attachment in the rubbery phase, (3) strong DMT-like R2G adhesion, and (4) weak JKR-like rubbery adhesion of the SMP adhesive fibril. (C) Versatile applications of the robotic adhesive skins in various scenarios, whether adhesion is necessary or not. Gentle and reliable gripping of fragile items (e.g., a raw egg), large objects (e.g., a basketball), and heavy objects (e.g., a 1.20 kg heavy aluminum block) is attainable through high adhesion strength in the R2G condition, with on-demand release achieved by heating the SMP to the rubbery phase, resulting in low adhesion. Dexterous manipulation (e.g., surface texture detection, handling of lightweight objects) is feasible due to minimal adhesion when contacting in the glassy phase. FIG. 49 depicts the details of the storage modulus and loss factor of the Epoxy SMP utilized in the current invention.

[0112] FIG. 50 depicts the fabrication of rough SMP adhesive fibrils. (A) Illustration of the fabrication procedure, (i) Preparation of a rigid mold using 3D Printer (Form 3+, FORMLABS INC.), with a sandpaper template affixed to the tip of the fibril, (ii) Ecoflex casting into the rigid mold to generate a flexible negative mold, (ill) Pouring SMP material into the soft Ecoflex mold and heating it to cure, (iv) Resulting SMP fibril with a rough tip. (B) Image of a smooth SMP adhesive fibril created without the use of a sandpaper template, along with the corresponding surface profile acquired using a TalyScan 150 3-D Scanner (Taylor Hobson, Inc.). (C) 3D morphologies of the SMP fibrils replicated from sandpapers with varying grit sizes: (i) 10O-grit, (ii) 120-grit, (iii) 240-grit, and (iv) 800-grit, as measured with the Keyence VK-X1100 Optical Profilometer.

[0113] FIG. 51 depicts the shear adhesion strength testing of an SMP adhesive fibril. (A) Photographic depiction, and (B) schematic representation of the experimental setup for shearing test. (C) Analysis of the measured force components during testing and calibration. (D) Results of the frictional force between the sliding glass and the stationary glass in relation to applied preloads.

[0114] FIG. 52 depicts the adhesion performance of SMP adhesive fibrils. (A) Impact of preload on the adhesion strength under (i) normal and (ii) shear modes (measured on smooth glass, aspect ratio = 1). (B) Effect of fibril aspect ratio on the adhesion strength under (i) normal and (ii) shear modes (measured on smooth glass, preload = 192 kPa). (C) Adhesion strength measured on various substrates under (i) normal and (ii) shear modes (aspect ratio = 1 , preload = 192 kPa). (D) Influence of SMP adhesive fibrils’ surface roughness on adhesion strength under (i) normal and (ii) shear modes (measured on smooth glass, aspect ratio = 1 , preload =192 kPa). (E-F) Comparison of achievable minimum and maximum adhesion strength values of SMP adhesive fibrils under (E) normal and shear (F) modes compared to other smart adhesives made from various intelligent materials. (G-H) Summary and comparison of adhesion switchability versus adhesion strength of SMP adhesive fibrils (the ratio between the R2G adhesion to rubbery adhesion) under (G) normal and (H) shear modes compared to reported results made of both conventional elastomers and intelligent materials.

[0115] FIG. 53 depicts the comparison of adhesion performance of different smart adhesives made of typical intelligent materials under both (A) normal (Data are from literatures: SMPs (Small, 2019, 15(50), 1904248; Macromolecular Rapid Communications 2022, 43, 2200012; Langmuir, 2021 , 37(45), 13420-13429; J. Microelectromech. Syst., 2014, 23(5), 1012-1014; Advanced Materials Interfaces 2019, 6, 1801542), ionogels (Soft Matter 2022, 18, 5934-5938; Advanced Intelligent Systems, 2023, 2300127), hydrogels (Nature Communications 2022, 13, 6964; Science Robotics 2021, 6, eabe1858), MRF / PDMS Composites (Science Advances 2023, 9, eadf4051), LM / PDMS Composites (Extreme Mechanics Letters 2021, 49, 101488; Advanced Materials 2016, 28, 5088-5092)); and (B) shear (Data are from literatures: SMPs (Extreme Mechanics Letters 2016, 9, 207-214), ionogels (Soft Matter 2022, 18, 5934-5938; / ACS Applied Polymer Materials 2023, 5, 3398-3404), hydrogels (Journal of Colloid and Interface Science 2021 , 582, 187-200; ACS nano 2023, 17, 13948-13960; Proceedings of the National Academy of Sciences 2022, 119, e2203074119; ACS Applied Materials & Interfaces 2021, 13, 36574-36586), and copolymers (Polymer Chemistry 2017, 8, 7207-7216; Angewandte Chemie International Edition 2015, 54, 13063-13067; Polymer 2015, 68, 270- 278; Polymer Chemistry 2018, 8, 6300-6308; ACS Macro Letters 2019, 8, 968-972; Nature communications 2016, 7, 12094)) modes.

[0116] FIG. 54 depicts the mechanisms behind the versatile robotic adhesive skin. (A) Understanding on-demand attachment: (i) Illustrations showing non-sticky attachment in the glassy phase and sticky attachment in the rubbery phase of the SMP adhesive fibrils, (ii) Free-contact (i.e., zero-preload) stickiness diagram for SMP adhesive fibrils (a = 1 mm, h = 1 mm) under glassy and rubbery phases, compared to the surface roughness of typical adherends, (iii) Examination of the forced contact between an SMP adhesive fibril (a= 5 mm, h= 10 mm) and a segregated steel substrate. The lines with dots show the FEA results of the true contact area normalized by the projected contact area. Insets show the (1 , 3) FEA simulation results and (2, 4) experimental observation of the contact conditions under the (1 , 2) glassy and (3, 4) rubbery phases. (B) Deciphering on-demand adhesion switching of an SMP adhesive fibril (a =1 mm, h = 1 mm), (i) Illustration showing the FEA simulation set up under normal mode and the contact stress distribution at pull-off point under (ii) R2G and (iii) rubbery conditions, respectively, (iv) Illustration showing the FEA simulation set up under shear mode and the contact stress distribution at fracture point under the (v) R2G and (vi) rubbery conditions, respectively.

[0117] FIG. 55 depicts the surface profiles of typical adherends. (A) Smooth glass (B) casted acrylic (C) polished steel (D) ceramic plate (E) red brick (F) cobble stone (G) wood block and (H) clean cloth FIG. 56 depicts the details of the FEA calculations of the forced contact between an SMP adhesive fibril and a segregated substrate. (A) Illustration showing the contact test configuration with the key geometries. (B) Illustration of the cross-sectional view of the FEA modeling of the contact test.

[0118] FIG. 57 depicts the experimental imaging of the contact area. (A) Photos of the samples used for contact area imaging, including (i) the SMP fibril, (ii) the serrate steel plate and (iii) the fluorescent nanoparticles under the natural light and 365 nm UV light illumination (B) Illustration showing process to prepare a thin layer of fluorescence nanoparticles coating on the serrate substrate. (C-D) Contact and imaging process of the contact condition between the (C) glassy and (D) rubbery SMP with the serrate substrate (preload = 231 kPa).

[0119] FIG. 58 depicts the details of the FEA setup for the simulation of the detachment of an SMP adhesive fibril in the (A-C) normal and (B-D) shear modes. (A, D) Illustration depicting the FEA setup. The fibril geometry resembles a hexagonal pillar with an edge length of 1 mm and a height of 1 mm. During the simulation, the height is adjusted to 0.8 mm since the fibril undergoes a compression of 0.2 mm in the adhesion test. (B, E) Force-displacement curves illustrate the behavior of the SMP adhesive fibril during the loading process under the R2G condition and the rubbery phase. (C, F) Snapshots exhibit the detachment process under the (i) R2G condition and (ii) rubbery phase.

[0120] FIG. 59 depicts the utilizations of the integrated robotic adhesive skin system in non-sticky states. (A) Surface morphology detection. In its glassy state, the adhesion between the SMP adhesive fibrils and the substrate is minimal, enabling the smooth traversal of the robotic hand across surfaces with (i) small, (ii) medium, and (iii) large amplitude textures, (iv) The recorded signals from the pressure sensor. (B-C) Handling ultra-light object. A clean cloth (3.89 g) is used for demonstration. (B) Upon contact in the rubbery state, even the low adhesion of the adhesive skins results in sticking, obstructing the release process. (C) In contrast, while in the glassy state, the frictional force facilitates successful gripping, and the minimal adhesion ensures a seamless release.

[0121] FIG. 60 depicts the utilizations of the integrated robotic adhesive skin system in Switchable Adhesion states. (A) Gripping of various kinds of objects, including quail eggs, plastic ball, wood cube, glass bub and ceramic plate. (B) Gripping and release of large objects, including steel shells with radii ranging from 10 cm to 40 cm, and a flat glass plate. (C-D) Gripping manipulation comparison of a heavy aluminum disk with an acrylic hand in the (C) glassy phase relying on friction, versus (D) that in the R2G condition and rubbery phase relying on switchable adhesion.

[0122] FIG. 61 depicts the smart adhesive gloves integrated with SMP adhesive skins. (A) Photo of an individual wearing a smart adhesive glove. Employing the smart glove (B) to grasp multiple quail eggs and (C) to hold a large steel shell in both normal and shear modes.

[0123] Description

[0124] In a first aspect of the invention, there is provided a gripping member, the member comprising: a base portion formed from a rubber-to-glass phase transitioning material; and one or more fibrils extending from the base portion made from a rubber-to-glass phase transitioning material, wherein the rubber-to-glass phase transitioning material: can transition from a glass phase to a rubber phase upon application of a stimulus and return to the glass phase in an absence of the stimulus; has an elastic modulus of less than 1 MPa in the rubber phase; and the transition from a glass phase to a rubber phase occurs in the one or more fibrils having a radius of at least 0.1 mm in size.

[0125] In embodiments herein, the word “comprising” may be interpreted as requiring the features mentioned, but not limiting the presence of other features. Alternatively, the word “comprising” may also relate to the situation where only the components / features listed are intended to be present (e.g. the word “comprising” may be replaced by the phrases “consists of” or “consists essentially o It is explicitly contemplated that both the broader and narrower interpretations can be applied to all aspects and embodiments of the present invention. In other words, the word “comprising” and synonyms thereof may be replaced by the phrase “consisting of’ or the phrase “consists essentially or synonyms thereof and vice versa.

[0126] The phrase, “consists essentially and its pseudonyms may be interpreted herein to refer to a material where minor impurities may be present. For example, the material may be greater than or equal to 90% pure, such as greater than 95% pure, such as greater than 97% pure, such as greater than 99% pure, such as greater than 99.9% pure, such as greater than 99.99% pure, such as greater than 99.999% pure, such as 100% pure. As used herein, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to “a composition” includes mixtures of two or more such compositions, and the like.

[0127] As will be appreciated, the gripping member may also be considered to be a smart adhesive device. However, for reasons of conciseness and clarity it will be referred to herein as a gripping member only.

[0128] As noted above, the “rubber-to-glass phase transitioning material” refers to a material that is capable of transitioning from a glass phase to a rubber phase upon application of a stimulus, and back to the glass phase upon removal of said stimulus.

[0129] Any suitable material may be used as the rubber-to-glass phase transitioning material. For example, the rubber-to-glass phase transitioning may be selected from one or more of a shape memory polymer, a liquid metal, a liquid-crystalline polymer, and a jamming material.

[0130] The stimulation applied may be any stimulus that can be used to cause a transition of a suitable material. Such stimuli may be selected from, but are not limited to, heating / cooling, light (e.g. UV light), electricity, and negative pressure.

[0131] In particular embodiments that may be mentioned herein, the materials listed in Table 1 below may be used, along with their suitable stimuli. In particular embodiments the rubber-to-glass phase transitioning is a shape memory polymer. For example, the shape memory polymer may be a material may be stimulated by electricity.

[0132] In yet further embodiments of the invention that may be mentioned herein, the shape memory polymer may be an E44 epoxy shape memory polymer.

[0133] An advantage of the current invention is that the radius of the fibrils used herein can be of a larger size (and hence adhesion strength) that is significantly larger than has been previously achieved. The size of the one or more fibrils may be selected to have: a dimensionless paramete of <0.5 in the glass phase, such that it has adhesion properties under a DMT-like adhesion regimen; and a dimensionless parameter of >0.5 in the rubber phase, such that it has adhesion properties under a JKR-like adhesion regimen, where is obtained from formula (1) where R is the radius of a single fibril, E and being the Young’s modulus and Poisson’s ratio of the fibril material, respectively, being the theoretical strength of adhesion, and being the work of adhesion.

[0134] In certain embodiments, the each of the one or more fibrils may have a radius that is greater than a critical radius in a rubbery phase and less than a critical radius in the glassy phase ( as calculated according to Formula (2) and Formula (3), respectively: where and are the Young’s modulus under the rubber phase and glass phase, respectively, and and are as defined in Clause 5.

[0135] In certain embodiments that may be mentioned herein, the rubber-to-glass phase transitioning material may be an E44 shape memory polymer. In such embodiments, the radius of each of the one or more fibrils may be from 0.5 mm to 3 mm, such as about 2.5 mm.

[0136] As an example, an E44 shape memory polymer may be prepared using the precursor materials shown below in (i) (i.e. precursors (poly(propyleneglycol)bis (2-aminopropylether) and E44-epoxy)) to provide the E44 shape memory polymer, which has a crosslinked polymeric network of the kind shown in (ii). The member disclosed herein may display a switchability of from 20 to 2,000, such as greater than or equal to 1 ,000 to less than or equal to 2,000. When used herein, the term “switchability” may refer to the ratio between the maximum adhesion force for attachment and the minimum adhesion force for detachment. In certain embodiments disclosed herein, the one or more fibrils may be a plurality of fibrils, each fibril may have a radius of from 0.5 mm to 3 mm, such as from 1.5 mm to 2.5 mm. The properties mentioned here may be generic or may be specific for a material using E44 shape memory polymer. As will be appreciated and discussed above, other materials may have different suitable radius windows as determined by their particular properties.

[0137] As will be appreciated, when there are a plurality of fibrils, these may be arranged into an array. The array may have any suitable size or shape, depending on the application envisaged for the gripping member. For example, the effective contact area of the gripping member may be any suitable size, such as up to 1 ,000 mm2or even higher.

[0138] Each of the one or more fibrils may have any suitable aspect ratio. For example, each of the one or more fibrils may have an aspect ratio of from 1 to 50, such as from 5 to 10, where the aspect ratio is the height of a fibril divided by its radius.

[0139] The gripping members disclosed herein may have a high adhesion efficiency. For example, the adhesion efficiency of the member is greater than 50%, such as greater than 55%, such as greater than 57.8%. The adhesion efficiency may be measured as described in the examples section hereinbelow.

[0140] The gripping members may have significantly higher adhesion strength than previously reported. For example, the adhesion strength of the member may be more than 1 MPa, such as from 1 MPa to 5 MPa, such as from 1 .5 MPa to 2 MPa, such as about 1 .89 MPa.

[0141] As will be appreciated, the gripping members disclosed herein may be suited for use in an apparatus. Thus, in a further aspect of the invention, there is provided a gripping apparatus comprising: at least one gripping member as described hereinbefore; and a stimulating means or apparatus, wherein the stimulating means or apparatus is configured to supply a stimulus to transition the gripping member from a glassy phase to a rubber phase.

[0142] As will be appreciated, the stimulating means or apparatus is intended to provide the desired stimuli discussed hereinbefore. This may be provided by an ambient environmental change or be a direct change, depending on the type of stimulus to be provided (e.g. the turning on or off of a particular light source, the supply of electricity or heat, etc ). As such, in certain embodiments of the invention, the stimulating means or apparatus may be formed integrally with the gripping member. More particularly, the stimulating means or apparatus may be formed integrally with the base of the gripping member. Such arrangements may, for example be suitable for the supply or electricity and / or heat through an electrical circuit or a resistive member that generates heat. Similarly, the base may be arranged to include a suitable light source light emitting diode, such that a suitable wavelength of light is provided that generates the desired stimulus.

[0143] In a further aspect of the invention, there is provided a gripping device, comprising: at least one gripping apparatus as described hereinbefore; and a substrate, wherein the at least one gripping apparatus is attached to the substrate,

[0144] The gripping device may be any suitable gripping device. Such devices include, but are not limited to a robotic hand or a glove suitable for use by a human. As will be appreciated, the robotic hand may be provided in a form that approximates that of a human or it may be significantly more rudimentary in form.

[0145] The gripping device disclosed herein may be used for gripping and releasing an object, thereby allowing it to be picked up from one position and placed in a new position (or orientation). Thus, in a further aspect of the invention, there is provided a method of gripping using a gripping device as described hereinbefore, wherein the method comprises:

[0146] (A) providing an object for gripping and a device as described hereinbefore;

[0147] (B) applying a stimulus to the at least one gripping apparatus to transition the one of more fibrils of the rubber-to-glass phase transitioning material into a rubber phase, then bringing the one or more fibrils into contact with a surface of the object to form an adhesion between the surface of the object and the one or more fibrils; and

[0148] (C) removing the stimulus to return the one or more fibrils to a glass phase, thereby gripping the object.

[0149] As will be appreciated, the above method allows for an object to be gripped by the gripping device. In order for the device to release the object, the method may further comprise the step(s):

[0150] (D) optionally physically manipulating the gripped object; and

[0151] (E) reapplying the stimulus to the at least one gripping apparatus to release the object.

[0152] As will be appreciated, step (D) is only required if there is a need to change the location and / or orientation of the object.

[0153] In the current invention, we have disclosed a novel concept and successfully engineered versatile robotic adhesive skins by integrating the SMP adhesive fibrils, flexible heaters, and flexible pressure sensors, which emulate the dual functionalities found in natural skin tissues for sensing and skin appendages for functions beyond sensation (e g., adhesion shown here). The results presented herein showcase the groundbreaking potential of the versatile robotic adhesive skin, offering a remarkable blend of on-demand attachment and detachment capabilities under varying environmental conditions. By harnessing the tunable properties of SMPs, this technology demonstrates impressive feats in object manipulation, with applications spanning from smooth detection of the surface textures, dexterous handling of lightweight items to securely gripping fragile, large, and heavy objects. These findings significantly extend the scope of interactions available to robotics within diverse environmental contexts, representing a notable advancement in intelligent robotics.

[0154] One of the primary implications of the current invention lies in its potential to reshape industries that rely on precise and adaptable object manipulation, owing to the broad adhesion spectrum the robotic adhesive skin can cover (from 0.65 kPa to 2 MPa in the normal mode, and from 2.86 kPa to 5.1 MPa in the shear mode) and the large adhesion switchability (>100) under high adhesion strength (> 1 MPa). In the realms of measurement, maintenance, textile, and paper industries, activities that encompass surface detection and dexterous manipulation hold paramount importance. Within these sectors, the capacity to handle objects with minimal adhesion is not merely advantageous, and the robotic adhesive skin technology promises to enhance productivity, safety, and precision. Conversely, in manufacturing, where fragile and intricate components are commonplace, the ability to handle objects with switchable adhesion can significantly enhance product quality and production efficiency. Moreover, applications demanding robust adhesion, such as construction and logistics, can substantially benefit from the technology's capacity to securely grasp heavy loads. Additionally, the versatility of this adhesive skin makes it a compelling candidate for applications in the field of healthcare, enabling delicate and precise operations with minimal risk to patients. Moreover, the successful integration of these adhesive skins into electronic adhesive gloves highlights their potential to become a part of daily life, potentially leading to a revolutionary impact on society.

[0155] Further aspects and embodiments of the invention will now be discussed by reference to the following non-limiting embodiments.

[0156] Examples

[0157] Conventional Theories of Adhesive Contact Mechanics

[0158] Herein we briefly cover the conventional theories of adhesive contact mechanics to facilitate the understanding of the following Examples. Hertz (J. Reine. Angew. Math., 1882, 92(22), 156-171) developed the classical theory of non-adhesive contact. In the case of a rigid sphere of radius R indenting on an elastic substrate, the contact radius is related to the indentation depth in the Hertz theory which assumes an infinitesimally small For finite inden tation depth the Hertz solution is modified as (J. Appl. Meeh., 2020, 87(5), 051003, Eq. (1)) to account for geometrical non-linearity. This modified Hertz model for contact radius given by Eq. (1) is still applicable for the incompressible neo-Hookean materials under finite spherical indentation up to Although adhesion can influence the contact radius, it was found to be negligible for finite indentation depths (Proc. Math. Phys. Eng. Sci., 1971, 324(1558), 301- 313; Phys. Rev. Lett., 2017, 118(23), 238001).

[0159] For general contact problem between two elastic bodies with curved surfaces, while the contact area size is small compared to the radii of curvature of the contacting bodies, Hertz’s theory remains valid if R is replaced with an effective radius of curvature of the contact surface, hereafter referred to simply as effective radius of curvature, defined as where R1 and R2 are the radii of curvature of the two contacting bodies.

[0160] Johnson, Kendall, and Roberts (Proc. Math. Phys. Eng. Sci., 1971 , 324(1558), 301-313) proposed the well-known JKR model of adhesive contact, where the influence of the adhesive force inside of the contact region of Hertz’s model is taken into consideration. The JKR model predicts a pull-off force of under fixed-load condition and under fixed-grip condition (J. Colloid Interface Sci., 1992, 150(1), 243-269), where is the adhesive energy or work of adhesion. Derjaguin, Muller, and Toporov (J. Colloid Interface Sci., 1975, 53(2), 314-326) proposed another model by considering the adhesive interaction outside of the Hertz contact area, and the so-called DMT model predicts a pull-off force of The discrepancy of the JKR and DMT models was a subject of intense debate until calculations based on the Lennard-Jones potential of interfacial interactions (J. Colloid Interface Sci., 1980, 77(1), 91-101) showed that there is a gradual transition from the DMT model, which falls in the strength-limited regime of adhesion, to JKR model, which governs the energy-controlled regime of adhesion, as the Tabor parameter increases, where s the theoretical interfacial adhesion strength and E* is the effective modulus of contact given b being the Young’s modulus and Poisson’s ratio of the two contacting bodies, respectively. The Tabor parameter indicates that the DMT model generally works for stiff bodies with small radii of curvature and weak interfacial adhesion (i.e. , under a small Tabor parameter), while the JKR model is suitable for soft bodies with large radii of curvature and strong interfacial adhesion (i.e., under a large Tabor parameter). Using a rectangular traction-separation law of interfacial interaction, or the so-called Dugdale model, Maugis (J. Colloid Interface Sci., 1992, 150(1), 243-269) derived a close-form solution for the pull-off force as where a is a coefficient that varies from the DMT limit of 2 to the JKT limit of 5 / 6 (under fixed- grip condition) or 3 / 2 (under fixed-load condition). Obviously, such small variations in a cannot explain the large adhesion enhancement by more than an order of magnitude in the R2G adhesion of SMPs. In the following examples, we conduct a systematic theoretical analysis, FEA calculations and experiments to uncover the fundamental mechanisms of R2G adhesion.

[0161] Example 1. Synthesis of the SMP Substrate

[0162] Synthesis of the SMP Substrate

[0163] A thermally-responsive epoxy shape memory polymer (Polym. Chem., 2015, 6(16), 3046- 3053; Sci. Adv., 2020, 6(7), eaay5120) is used in the current experimental investigation. The SMP substrate is prepared by pouring a mixture of the epoxy monomer E44 (China Feicheng Deyuan Chemical Co.) and curing agent JEFFAMINE D-230 (Sigma-Aldrich) with a mass ratio of 81 :46 into a circular container mold. Specifically, an annular silicone rubber spacer with desired inner diameter of 80 mm is first cut from a thick silicone rubber sheet with thickness of 12 mm, which serves as the container mold when put on a clean glass plate. The mold is then filled by the mixture of the epoxy monomer and curing agent, covered by another clean glass plate on the top and then fixed with binder clips. The SMP substrate (D = 80 mm, H = 12 mm) is obtained upon demolding after curing in the oven at 100°C for one hour and postcuring at 130°C for another hour. The SMP substrate surface is replicated from the smooth glass and is thus very smooth, with the arithmetic mean deviation value of surface roughness of Sa = 1.870 nm. More details of the fabrication process can be found in the literature (Sci. Adv., 2020, 6(7), eaay5120).

[0164] Results and Disucssions

[0165] This E44 SMP has a glass transition temperature of around 45°C above which its modulus decreases dramatically. For example, it is very stiff in the glassy state with a modulus of around 2 GPa at 30°C, but very soft in the rubbery state with a modulus of around 1 MPa at 90°C. When cooled below the glass transition temperature, the E44 SMP can “freeze” in a deformed configuration with a shape fixation ratio of around 99%. When heated above the glass transition temperature again, it recovers to the original shape with a shape recovery ratio close to 100%.

[0166] Example 2. Contact and Pull-off Tests of the E44 SMP Substrates

[0167] Contact and Pull-off Tests

[0168] Smooth glass sphere lenses (Sa ~ 1.752 nm) with different radii (R = 1 , 5, 10, and 50 mm) are glued to an acrylic handle using UV glue and serve as spherical indenters to probe the R2G adhesion of the E44 SMP substrate. The indenter is connected to an Instron 5566 universal test machine and brought into contact with the SMP substrate heated to the rubbery state (90°C) in fixed-grip (or displacement-controlled) manner at very low speed (1 - 10 um / s). After a dwelling of 5 min under a fixed displacement for relaxation, the SMP substrate is cooled down to the glassy state (30°C) at certain indentation depth, and the indenter is then pulled up under a fixed speed of 100 um / s. The load cell connected to the indenter record the pulling force. The pull-off force is obtained from the load-displacement curve for given indenter radius and indentation depth.

[0169] Example 3. Materials Properties of the E44 SMP

[0170] Dynamic Mechanical Analysis (DMA)

[0171] Dynamic Mechanical Analysis (DMA) was utilized to assess the storage modulus (£’) of the SMP materials across a range of temperatures, employing a measuring frequency of 0.5 Hz and a temperature ramp rate of 2°C / min. To create the rectangular test specimens, each measuring 30 mm in length, 6 mm in width, and 2 mm in thickness, an initial 2 mm thick SMP membrane was formed within the mold. Subsequently, these specimens were precision-cut into the desired geometry using a laser cutter.

[0172] Contact Tests to Demonstrate the Shape-Locking and Shape-Memory Effects ofE44 SMP Contact tests of a shape-memory polymer (SMP) block (20 mm * 20 mm * 7.5 mm) with a steel substrate with serrated surface (peak-to-valley amplitude A - 0.5 mm, wavelength A = 1 mm) were performed to show the shape-conformability of the SMP when heated to the rubbery state and the temporary shape-locking effect when cooled to the glassy state. In addition, contact tests were conducted on a rough glass substrate to demonstrate the shape-memory effect of the SMP. The detailed procedure for demonstrating the shape-locking effect is shown in FIG. 6. First, the SMP block (20 mm x 20 mm x 7.5 mm) is brought into contact with the serrated steel surface in the glassy state (25 °C) with a preload of 200 kPa (FIG. 6(A)-(i) and 6(B)-(iii)). In the glassy state, the SMP is very stiff, and consequently, the SMP has contact with the steel surface only at the serrated tips (FIG. 6(B)-(iv)). Next, the SMP block is heated to the rubbery state at 90 °C (FIG. 6(A)-(ii) and 6(B)-(v)), where it becomes very soft. This facilitates conformal contact between the SMP and the serrated steel surface (FIG. 6(B)-(vi)). Then the SMP is cooled to the glassy state (FIG. 6(A)-(iii)), with deformation temporarily fixed due to the shape-locking effect, allowing direct measurement of the deformed shape after the SMP sample is separated from the substrate (FIG. 6(A)-(iv)). FIG. 6(B)-(vii) shows the image of the locked deformation of the SMP block after separated from the serrate steel surface and FIG. 6(C) shows the measured morphologies. The SMP perfectly replicates the surface morphology of the serrate steel surface, which confirms the shape conformability of the SMP in the rubbery state due to temporary shape-locking effect.

[0173] The contact test aimed to demonstrate the permanent shape-memory effect is similar to that shown in FIG. 6(A), with the adherend replaced by a rough glass substrate and the SMP heated after separation to recover its original shape. The recovered shape is measured after the sample is fully heated and then cooled.

[0174] Results and Discussions

[0175] FIG. 7(A) shows the dynamic mechanical analysis (DMA) results of the storage modulus and loss factor of the E44 SMP at different temperatures. The E44 SMP has a glass transition temperature corresponding to the peak of the loss factor. Below the E44 SMP is in the glassy state and stiff, with a modulus of around 2 GPa (for example, E’ = 2.3 GPa at 30°C). When heated above the E44 SMP transitions into the rubbery state and becomes soft, with a modulus of around 1 MPa (for example, E = 0.95 MPa at 90°C).

[0176] The dramatically tunable modulus value of E44 SMP, spanning three orders of magnitude, led to very limited contact between the SMP block and the peaks of the serrated surface (peak- to-valley amplitude A - 500 pm and wavelength A - 1 mm) of a steel adherend under a high preload of 200 kPa in the glassy state (FIG. 7(B)-(i) and 7(B)-(iii), but it facilitated conformal contact between the SMP block and a rough surface (i.e., the serrated surface of steel adherend) in the rubbery state, as demonstrated in FIG. 7(B)-(ii) and 7(B)-(iv). When cooled to the glassy state after conformal contact with the serrated steel surface (FIG. 7(C)-(i) to 7(C)-(iii)), the E44 SMP freezes the deformed shape. FIG. 7(C)-(iv) shows the surface profile of the E44 SMP block separated from the serrated steel surface after it was cooled to the glassy state while in contact. The E44 SMP could replicate the serrate steel surface both at a coarse scale up to 100 pm and fine scale down to 1 pm (FIG. 6), showing the capabilities of the E44 SMP to establish conformal contact upon heating and maintaining the deformed shape upon cooling.

[0177] Contact testing on a rough glass surface (FIG. 6) further demonstrated the shape-memory behavior of the SMP, which facilitated the recovery of the smooth surface morphology of the SMP block after repeated contact of up to 20 cycles, as evidenced by the SEM images (FIG. 7(D)-(i) to 7(D)-(v)) and 2D profiles (FIG. 7(D)-(vi)) of the SMP block surface before and after contact. Even though the SMP was highly deformed after contacting with the rough glass surface (FIG. 7(D)-(ii)), it could restore the original shape (FIG. 7(D)-(i)) after repeated contact tests (FIG. 7(D)-(ii) to 7(D)-(v)), which would be crucial to the switchability of the SMP adhesives.

[0178] Example 4. Finite Element Analysis (FEA) Setup

[0179] A thermomechanical constitutive model of SMP (J. Intell. Mater. Syst. Struct., 1997, 8(8), 711- 718; Meeh. Mater., 2001 , 33(10), 545-554; J. Intell. Mater. Syst. Struct., 2020, 31, 2128-2136) was used in the FEA simulations. Detailed descriptions of the thermomechanical constitutive model can be found below.

[0180] Finite Element Analysis (FEA) Model

[0181] FEA simulations of the R2G adhesion between a rigid sphere and an SMP substrate were carried out using ABAQUS Standard (Static, General). FIG. 8(A) shows a schematic of the axisymmetric FEA model of a rigid sphere in adhesive interaction with the SMP substrate. The bottom surface of the SMP substrate was fixed during the simulation. RAX2 (rigid, axisymmetric 2-node) elements and CAX4 (continuum, axisymmetric 4-node) elements were used for the rigid sphere and the SMP substrate, respectively. A linear traction-separation law was used to model adhesive interactions after contact is established and non-slipping adhesive contact was assumed during the contact and pulling process. A quadratic nominal stress criterion is used for damage initiation and a power-law (a = 2) mixed mode behavior (J. Adhes. Sci. Technol., 1995, 9(2), 119-147; J. Compos. Mater. 2003, 37(16), 1415-1438) is used for damage evolution. The adhesive parameters between glass and E44 SMP are measured using a method proposed by Peng et al (Peng et al., 2020) as GH, = 1.893 MPa = 5 MPa 1.61 J / m2and 3.22 J / m2, where and Tlhare the theoretical adhesion strength in the normal and radial shear directions, respectively, and are the critical energy release rate (or work of adhesion) of Mode-I and Mode-ll crack, respectively.

[0182] SMP Thermomechanical Constitutive Model

[0183] A thermomechanical constitutive model (J. Intell. Mater. Syst. Struct., 1997, 8(8), 711-718, Meeh. Mater., 2001 , 33(10), 545-554) adapted from a linear viscoelastic model by adding a friction-slip element is used to model the behaviors of SMP in FEA simulation where and are the elastic modulus, viscosity and retardation time of the SMP, respectively; and denote stress, strain and temperature, respectively; is the creep irrecoverable strain and OIT is the coefficient of thermal expansion; the dot denotes the time derivative. The material parameters are related to temperature as (J. Intell. Mater. Syst. Struct., 1997, 8(8), 711-718) where denotes one of the material parameters, and denotes the value at the glass transition temperature and is the corresponding proportional coefficient.

[0184] According to the experimental observations, the irrecoverable creep strain appears only when the total creep strain ecis smaller than a certain threshold value, otherwise the Once the threshold value is exceeded, the irreversible creep strain es(t,7) increases in linear proportion to the total creep strain as the creep strain increases (i.e., when For a given temperature, the irrecoverable strain esis described by with the proportion coefficient and the threshold value a (7) related to temperature as Eq. (7)

[0185] Extension of this model to 3D conditions can be found in (Comput. Model Eng. Sci., 2013, 90,

[0186] 339-358; Smart Mater. Struct., 2016, 25(11), 115034; Mater. Des., 2020, 196, 109153; J. Intell. Mater. Syst. Struct., 2020, 31, 2128-2136; Int. J. Meeh. Sci., 2021 , 195, 106267). For the finite element analysis of SMP structures under complex stress states, the above SMP constitutive equations can be written in tensor form and integrated into ABAQUS using a user material subroutine (Comput. Model Eng. Sci., 2013, 90, 339-358; Smart Mater. Struct., 2016, 25(11), 115034). The model parameters adopted in our simulations are summarized in Table 2, and more details can be found in (Comput. Model Eng. Sci., 2013, 90, 339-358; Smart Mater. Struct., 2016, 25(11), 115034; Mater. Des., 2020, 196, 109153; J. Intell. Mater. Syst. Struct., 2020, 31 , 2128-2136).

[0187] Results and Discussions

[0188] A benchmark simulation on an SMP strip (25 mm x 5 mm x 1 mm) subject to uniaxial tension in glassy (30 °C) and rubbery states (90 °C) demonstrates that this constitutive model can capture the tunable modulus effect with correct modulus values in the glassy state (2 GPa) and rubbery state (1 MPa), as shown in FIG. 9. A full cycle of deformation, shape locking and shape recovery of the SMP strip was also simulated in four steps (FIG. 8(B)): 1) loading under a displacement of 5 mm in the rubbery state (90 °C); 2) cooling down to the glassy state (30 °C) under fixed displacement; 3) freezing of deformation in the glassy state and removal of external load; and 4) heating to the rubbery state to recover the original shape. FIG. 8(C) shows the strain / temperature-time curves for the whole simulation cycle. These results show that the thermomechanical constitutive model can simulate the temporary shape locking effect with a shape fixation ratio of 99.27% and shape memory effect with a shape recovery ratio of 99.99%. In simulating the R2G adhesion (FIG. 8(A)), a temperature of 90 °C was applied to the SMP substrate at the beginning, and the spherical indenter was moved toward the SMP substrate until the indentation depth reached the target value in step 1. The SMP substrate was then cooled to 30 °C in step 2, and finally a negative displacement was applied on the sphere until it was pulled-off from the SMP substrate in step 3. Example 5. Fracture Mode and Contact Radius

[0189] Fracture Mode

[0190] At finite indentation depth, the fracture mode is no longer pure Mode I. For the adopted traction-separation law, it is found that for all indenter radii (R = 1, 5, 10, 50 mm) and indentation depths discussed herein, the fracture is typically mode-l dominant with and and denote the normal traction, normal separation, shear traction and shear separation at the contact interface, respectively, and the superscript denotes the fracture point. As the indentation depth can become comparable to the radius of the indenter (Sci. Adv., 2020, 6(7), eaay5120), and the SMP substrate can be very soft and undergo large deformation around the indenter (Small, 2019, 75(4), 1803870), substantial shear interaction and even mechanical interlocking at the contact interface can occur. However, we limit our discussion to a relatively small indentation depth range and neglect mixed mode adhesive contact in the theoretical analysis.

[0191] Contact Radius

[0192] The contact radius between the spherical indenter and SMP substrate can be estimated from the modified Hertz model for finite indentation depth as given by Eq. (1), in good agreement with the FEA simulations (FIG. 11), with relative errors smaller than 10%. For all FEA simulations, the thickness of the SMP substrate was set as at least 4 times the contact radius so that the substrate can be approximated as a half-space.

[0193] Example 6. Mechanisms of the R2G Adhesion

[0194] Dependence on the Indentation Depth

[0195] A major difference between the R2G adhesion and the conventional adhesive contact is that the adhesive strength of the former increases with indentation depth, as shown in FIG. 12(A). For the conventional adhesion between an elastic substrate and a rigid sphere, the pull-off force is given by Eq. (3), which is independent of the indentation depth, and the upper limit of the pull-off force is given by the DMT solution of for a given and For R2G adhesion, in contrast, the upper limit of the pull-off force can be obtained by integrating the vertical component of the theoretical interfacial adhesion strength over the contact area (see “Strength-Limited Results of R2G Adhesion" for details) as

[0196] For small indenter radius, the R2G pull-off force is very close to this limit, as validated by FEA simulations and experimental data, as shown in FIG. 12(A) in the case of a sphere with radius of 1 mm and follows the scaling law of Eq. (5), as shown in FIG. 12(B). As the indenter radius increases, the R2G adhesion falls below the strength limit in Eq. (5).

[0197] Strength-Limited Results of R2G Adhesion

[0198] Assume the normal traction at the pull-off point has a distribution of asshown in

[0199] FIG. 13(A), where r is the radial distance to the center of the sphere. Integration of the vertical component of the normal traction over the contact interface gives the net pull force as where according to geometrical relations (FIG. 13(B)), and

[0200] In the strength-limited adhesion regime, the normal traction reaches the theoretical value and which gives the upper limit of the R2G pull-off force with a spherical indenter as compared to the strength-limited value for a flat-punch model as

[0201] Adhesion Enhancement

[0202] For the R2G adhesion, there are two potential effects during the rubbery to glassy transition: the modulus effect and the shape locking effect. For the E44 SMP, the modulus increases from 1 MPa in the rubbery state to 2 GPa in the glassy state. At the same time, the deformation is locked and the compressed shape under the contact region is frozen. It is important to determine which effect dominates the R2G adhesion. The pull-off forces for a rigid sphere in adhesion with an elastic substrate with moduli of E = 1 MPa and E = 2 GPa (i.e., values before and after the glass transition), and that for the net R2G adhesion with = 0.1 under the same adhesive parameters are compared in FIG. 12(C). The adhesion enhancement ratios due to the modulus effect and the net R2G transition effect are obtained by comparing the pull-off forces after the modulus change in the elastic state of E = 2 GPa and after the R2G transition to the reference elastic value under E = 1 MPa. By subtracting the adhesion enhancement due to modulus effect from the net R2G effect, the contribution of the shape locking effect is obtained and shown in FIG. 12(D).

[0203] For elastic substrate with modulus of 1 MPa, the adhesion coefficient is equal to the fixed-grip JKR value of 5 / 6 for all indenter radii under investigation (see “Summary of Adhesion Enhancement’ for details), indicating that the adhesion regime is energy-controlled due to the softness of the substrate. As the modulus of the substrate is shifted from 1 MPa by 3 orders of magnitude to 2 GPa in the glassy state, the Tabor parameter is reduced by two orders of magnitude, which moves the adhesion regime close to the DMT regime with adhesion coefficient increased above 1.90. This shows that the modulus effect only contributes to an adhesion enhancement ratio of 2.3 - 2.4, independent of the indentation depth. In contrast, the enhancement ratio of the R2G adhesion compared to the elastic reference state of E - 1 MPa for is 131 , 558, 1030, and 2936 for indenter radius of 1, 5, 10 and 50 mm, respectively. Obviously, the modulus effect only contributes a very small fraction of the adhesion enhancement (< 2% here), and the shape locking effect accounts for more than 98% of adhesion enhancement (as shown in FIG. 12(D)). In the conventional models of adhesive contact, the compressive stress from elastic deformation within the contact area counteracts the tensile stress due to adhesive interactions, leading to a pull-off force that is independent of the indentation depth. In contrast, the R2G adhesion is greatly enhanced as the indentation depth increases (as shown in FIG. 12(A) and 12(B)). The agreement between experiments and FEA simulations in FIG. 12 also validates the accuracy of the FEA model. In the following sections, more FEM simulations are conducted to probe deeper into the R2G adhesion to reveal the underlying mechanisms.

[0204] Summary of Adhesion Enhancement

[0205] The adhesion enhancement ratios due to the modulus change and the net effect of the R2G transition are summarized and compared here. During the R2G transition, the SMP is shifted from the rubbery state (E = 1 MPa) to the glassy state (E= 2 GPa), and the deformation is locked at the same time. Consequently, two effects - modulus change and shape locking - are involved in the R2G adhesion. Contributions of each effect are summarized and compared here. The pull-off forces and adhesion coefficients for elastic conditions (E = 1 MPa and 2 GPa) are calculated from Eq. (3) according to the MD theory (J. Colloid Interface Sci., 1992, 150(1), 243-269) for adhesion of elastic spheres. The adhesion coefficient for the elastic condition before modulus change (E = 1 MPa) is 5 / 6 for all indenter radii, indicating an energy- controlled adhesion regime under fixed-grip condition. The adhesion coefficient for the elastic condition after modulus change (E = 2 GPa) ranges from 1.91 to 1.99 as the indenter radius decreases, showing that the adhesion regime is very close to the strength-limited regime. The modulus effect shifts the adhesion regime from an energy-controlled one to a strength-limited one, leading to an adhesion enhancement ratio in the range of 2.3 - 2.39. Typical values of the pull-off force in the R2G transition (97R = 0.1) are obtained from FEA calculations, and the adhesion enhancement ratios due to the R2G transition are obtained by comparing the R2G pull-off forces to the basic elastic state pull-off forces prior to transition (Elastic, E = 1 MPa). The contribution from the shape locking effect is obtained by subtracting the contribution of the modulus effect from that of the net R2G effect. The results are summarized in Table 3.

[0206] Normal Stress Evolution during Pull-Off

[0207] FIG. 14 shows the typical curves of the normalized contact force as a function of normalized pulling displacement for elastic adhesion (FIG. 14(A)) and R2G adhesion (FIG. 14(C)), and the corresponding evolution of normalized normal stress distribution in the contact area (R =1 mm, 5 =0.01 mm) for elastic adhesion with modulus of E=2 GPa (FIG. 14(B)) and for R2G adhesion with both modulus effect and shape locking effect (FIG. 14(D)). The number-labelled dots in FIG. 14(A) and 14(C) indicate different loading stages during pulling and the number labelled curves in FIG. 14(B) and 14(D) correspond to the numbered loading stages in FIG. 14(A) and 14(C) , respectively. The compressive contact force (positive) is normalized by its maximum value during the contact process, and the pulling force (negative) by moadR . Similarly, the compressive normal stress (positive) is normalized by its maximum value during the contact process, and the tensile normal stress (negative) by the theoretical adhesion strengt

[0208] In the elastic state (E=2 GPa), the force-displacement curve during contact and pulling process shows little hysteresis and coincides before the indentation depth approaches zero (FIG. 14(A)). Most of the contact and pull-off process is in a compression-dominant state. That is, the resultant force of the compressive stress due to elastic deformation within the contact area exceeds the resultant force of tensile stress due to adhesion, as shown by the dashed line ), dot dashed line and dotted line in FIG. 14(B). Only when the indentation depth approaches zero does the compressive deformation within the contact area become negligible and the tensile stress begin to dominate (i.e., the pulling process is in a tension-dominant state), as shown by the solid lines in FIG. 14(B). The inset in FIG. 14(A) shows the normal stress distribution in the elastic substrate at the pull-off point, where adhesion only remains in a very small residual portion of the contact area. In contrast, with shape locking effect in the R2G adhesion, there is a substantial hysteresis in the forcedisplacement curve during the contact and pulling process (FIG. 14(C)). In the R2G adhesion, the pulling process is dominated by adhesion over a large contact area (as evidenced by the normal stress evolution at the contact interface shown in FIG. 14(D)) since the deformed surface profile is locked. As the pulling proceeds, the stress at the contact interface can increase to the theoretical adhesion strength over most of the contact area, as seen in the inset of FIG. 14(C) and the solid lines in FIG. 14(D), which is responsible for the greatly enhanced pull-off force in R2G adhesion.

[0209] Example 7. Pull-Off Force and Adhesion Map

[0210] Effective Radius of Curvature of the Contact Surface Due to Shape Locking Effect

[0211] In the R2G adhesion, the shape locking effect of the SMP substrate leads to nearly conformal contact with the contacting sphere, thus the local radius of curvature of the SMP substrate becomes nearly equal to, but opposite in sign with that of the contacting sphere, i.e., is close to -R (as illustrated in FIG. 3(A)), leading to a very large effective radius of curvature like that defined in Eq. (2). While it is difficult to get an explicit expression ta we can estimate it iteratively from the pull-off force and the (adhesion coefficient) (Tabor parameter) curve obtained from the Maugis-Dugdale (MD) model (J. Colloid Interface Sci., 1992, 150(1), 243-269) (see FIG. 15 shows Res for different indenter radii under different indentation depths. The calculated Reff increases with indentation depth by more than 2 orders of magnitude. For example, for is as large as 57.14, 295.68, 716.93, and 2935.798 for 5, 10, and 50 mm, respectively. The increase of herewith indentation depth gives rise to a pull- off force that depends on the indentation depth and results in the observed adhesion enhancement. FIG. 16(B) shows the corresponding adhesion coefficient and adhesion regime. As the indentation depth increases, the effective radius of curvature increases. The increased effective radius of curvature leads to increased Tabor parameter and moves the adhesion regime towards the JKR regime.

[0212] Calculation of Effective Radius of Curvature of the Contact Surface and Adhesion Regime According to the Curve

[0213] The adhesion coefficient a and effective radius of curvature Reff are calculated iteratively as shown in FIG. 15. First, an initial value for the effective radius of curvature is calculated as and then a new Tabor parameter is given using the estimated effective radius of curvature. After that, the adhesion coefficient is updated according to the (adhesion coefficient) (Tabor parameter) curve obtained from the MD model solution (J. Colloid Interface Sci., 1992, 150(1), 243-269). Then Res is updated using the new adhesion coefficient until the relative error of between two neighboring iteration is less than 0.1%.

[0214] Flat Punch Model for R2G Adhesion

[0215] The shape locking of the SMP substrate conformal to a spherical indenter and the large effective radius of curvature of the contact surface suggest that it might be possible to use a flat-punch model to approximate the R2G adhesion with a sphere. In the flat-punch model, a rigid flat punch with radius a is detached from a flat elastic substrate.

[0216] The effective adhesion strength of a rigid flat punch on an elastic substrate in the energy- controlled regime was given by Kendall (J. Phys. D, 1971, 4(8), 1186) according to the Griffith (Philos. Trans. R. Soc. London A, 1921 , 221 , 163-198) condition of fracture in the form of (J. Phys. D, 1971 , 4(8), 1186; Proc. Natl. Acad. Sci., 2004, 101(21), 7851-7856)

[0217] Hui et. al (J. R. Soc., Interface, 2004, 1(1), 35-48) used a non-slipping Dugdale-Barenblatt (DB) cohesive model to obtain a solution of flat punch adhesion from the strength-limited to energy-controlled regime as a function of the non-dimensional paramete The effective adhesion strength can be expressed as where r| is the ratio of the cohesive zone size c to the punch radius a determined by

[0218] Fo f|atpunch adhesion falls in the strength-limited regime (with effective adhesion strength within 5% of theoretical adhesion strength). For the flat punch adhesion becomes energy-controlled (with effective adhesion strength within 5% of the Kendall solution).

[0219] FIG. 17 shows a comparison between the FEA predictions of effective adhesion strength (pull-off force over the projected contact area, i.e., of R2G with those of the flat punch model for different indenter radii and indentation depths. It is seen that the R2G adhesion can be reasonably approximated by the flat punch model. When , the FEA results of the R2G adhesion strength are within 15% of the theoretical adhesion strength, very close to the strength-limited regime. When , the results transition into the energy-controlled regime with effective adhesion strength within 5% of the Kendall solution. These results validate the flat punch model as a reasonable approximation of the R2G adhesion. Furthermore, the stress distribution over the contact interface in R2G adhesion is similar to that of flat punch adhesion with the same contact radius (see “Comparison of the Interfacial Stress Distribution at the Pull- Off Point between the R2G Adhesion with a Rigid Sphere and the Flat Punch Model” and FIG. 18). Therefore, the R2G pull-off force can be estimated according to Eq. (14) and Eq. (15), where the pull-off force is dependent on indentation depth 3 through contact radius a.

[0220] Comparison of the Interfacial Stress Distribution at the Pull-Off Point between the R2G Adhesion with a Rigid Sphere and the Flat Punch Model

[0221] FEA simulations are conducted to compare the stress distribution in the R2G adhesion with a rigid sphere and in the rigid flat punch - elastic substrate adhesive system (E = 2 GPa) under the same contact radius, as shown in FIG. 18(A). A linear traction-separation law is used to model the adhesive interaction. The remaining FEA setups are identical to those stated in the other examples. FIG. 17(B) shows the normal stress distribution in the substrate and the normal traction distribution along the contact interface in the strength-limited and energy- controlled regimes of the two systems, which are very similar.

[0222] Pull-Off Force Expression

[0223] It is of practical importance to have an explicit expression for the pull-off force of the R2G adhesion for various indenter radii under different indentation depths. Jiang et al (J. Phys. D, 47(32), 325301) proposed an approximate solution (MD-n solution) for the flat punch adhesion from the Maugis-Dugdale-n (MD-n) model for power-law-shaped bodies in frictionless contact as, Eq. (16) where

[0224] Eq. (17) and Eq. (18) Substituting Eq. (1) into Eq. (16-17) and Eq. (18), we obtain an approximate expression for the pull-off force in the R2G adhesion as

[0225] Eq. (19) where is given by Eq. (17) and Eq. (20)

[0226] We have already shown in Eq. (9) that the upper limit of the R2G pull-off force is given by Using the flat punch model, the strength limit for the pull- force, however, is given by

[0227] Eq. (21) which is slightly larger than the exact R2G strength limit of a sphere with an SMP substrate. FIG. 19 shows a comparison of the flat punch approximation given by Eq. (19) with FEA and experimental results for the R2G adhesion between a sphere and an SMP substrate, as well as the strength limit given by Eq. (9). For small radius (R = 1 mm, FIG. 19(A)) and small indentation depths (FIG. 19(B) to 19(D)), the contact radius between the indenter and SMP substrate is small, in which case the adhesion regime is strength-limited (as also shown in FIG. 17). In such cases, the pull-off force is limited by the exact strength limit of the spherical configuration given by Eq. (9). For large indenter radii and indentation depths (FIG. 19(B) to 19(D)), the contact radius can become sufficiently large that the adhesion regime deviates significantly from the strength limit. In these cases, the flat punch approximation using DB model given by Eq. (19) can fall below the exact strength limit of the spherical configuration. Nevertheless, results in FIG. 19 demonstrate that the flat punch solution agrees well with FEA results in all cases and with experimental results in most conditions. The only exception is the case with indenter radius as large as 50 mm and indentation depths as large as = 2 mm and 3 mm, where the SMP substrate can no longer be approximated as a half space due to the finite thickness of experimental sample (12 mm). Overall, the flat punch approximation given by Eq. (19) gives a reasonable estimation of the pull-off force in the R2G adhesion. The empirical solution of Peng et al. (J. Meeh. Phys. Solids, 2020, 139, 103937) for the flat punch model (assuming frictionless adhesive contact) with a finite thickness ( - 12 mm) is also shown for comparison in FIG. 19(D) (dashdotted line). It is found that the confinement effect due to the finite thickness of the substrate in the flat punch model only leads to a slight increase (smaller than the experimental results) in the pull-off force compared to the FEA simulations (squares) and theoretical predictions (solid line) with infinite substrate thickness. This indicates that the finite thickness effect of the substrate in the flat-punch model of Peng et al. (J. Meeh. Phys. Solids, 2020, 139, 103937) cannot fully account for the finite-thickness effect in the R2G adhesion of the rigid sphere with the SMP substrate. There are clearly other factors that should be considered for the correction of the finite thickness effect in the R2G adhesion in future work.

[0228] Adhesion Map

[0229] Jiang et al (J. Phys. D, 47(32), 325301) also gave a criterion to determine the adhesion regime of a flat punch: the adhesion is in the strength-limited regime when 0.73 (equivalent to < 0.29); in the energy-controlled regime when 2.79 (equivalent to 1.13); and in the transitional regime for intermediate values of Similarly, we can determine the adhesion regime according to the value of s predefined by the geometric, elastic, and adhesive characteristics of the contacting bodies. For example, if we still define the strength-limited regime as having adhesion strength within 5% of the theoretical adhesion strength and the energy-controlled regime as having adhesion strength within 5% of the Kendall solution, we can get the critical values determining the adhesion regime transition of R2G adhesion as and which are at the same order of magnitude as these in flat punch model. Thus, the adhesion regime of R2G can be determined from system characteristics and indentation depth 3 / R as (FIG. 20)

[0230] It is seen from FIG. 20 that the FEA results show good agreement with the theoretical predictions based on flat punch model.

[0231] Example 8. Mechanics of SMP Adhesion on Rough Surfaces

[0232] E44 SMP adhesive samples (cylinders with radius R = 2.5 mm, height H = 4R) were fabricated using molding methods (FIG. 21) to test the adhesion performance (FIG. 22) of the SMP on glass surfaces of different roughness.

[0233] Details for the SMP Adhesive Sample Fabrication (Molding Methods)

[0234] The detailed fabrication process of the SMP adhesive sample for adhesion tests is shown in FIG. 21(A). First, a stiff mold with a cylinder in the center of a circular container was printed using a 3D printer (Form 3, FormLabs), as shown in FIG. 21(A)-(i). Then, the Ecoflex precursors were mixed and poured into the stiff mold. After curing and demolding, the soft Ecoflex mold with a central cylinder hole (FIG. 21 (A)-(ii)) was glued to a glass plate as a mold for the SMP adhesive sample. The SMP precursors are mixed and poured into the mold prepared as shown in FIG. 21 (A)-(iii), degassed in a vacuum oven for 30 minutes, and then cured in the oven at 100 °C for one hour and at 130 °C for another hour. After curing, the SMP adhesive sample is demolded and ready for testing (FIG. 21 (A)-(iv)). FIG. 21 (B)-(i) to 21(B)- (iii) show the photos of the stiff mold, the soft Ecoflex mold and the SMP adhesive sample. A PDMS adhesive sample (mass ratio of base to crosslinker at 10:1) with the same geometry was also prepared and used as a control. The fabrication process of the PDMS adhesive sample is the same as that of the SMP adhesive sample, but with different curing conditions (curing temperature 75 °C, curing time 4 h). Details for the SMP Adhesion Test

[0235] Adhesion Test. The SMP and PDMS adhesive samples (cylinders with radius R 2.5 mm and height H = 4R) were glued to an acrylic handle and clamped onto an Instron universal tensile machine (Instron 5566) for the adhesion test. As shown in FIG. 22, the adhesive samples were brought into contact with the adhered surface at a very low speed of 10 pm / s and then pulled up at a fixed speed of 100 pm / s after dwelling for 5 minutes by the linear stage under displacement control, with the inline load cell measuring the force throughout the process. The rough glass samples were mounted on a hot plate that controlled the test temperature. During the SMP rubber-to-glass (R2G) adhesion test (FIG. 22(A)), the temperature was maintained at 90 °C during the contacting and dwelling stages so that the SMP was in the rubbery state, then adjusted to the room temperature (25 °C) to bring the SMP to the glassy state before pulling. During the SMP glassy-state adhesion test (FIG. 22(B)), the temperature was kept at the room temperature (25 °C) all the time and the SMP stayed in the glassy state during the contact, dwell and pull stages. During the SMP rubbery-state adhesion test (FIG. 22(C)), the temperature was kept above the glass transition temperature at 90 °C all the time and the SMP stayed in the rubbery state during the contact, dwell and pull stages. PDMS adhesion was measured at room temperature (25 °C) as shown in FIG. 22(D).

[0236] Influence of Preload The dependence of adhesion force (or adhesion strength) on preload was studied by changing the preload from 51 kPa to 255 kPa. The adhesion strength increases with preload and reaches a plateau after conformal contact is established. In all cases studied here, the adhesion strength values reached plateau at a preload smaller than 255 kPa. FIG. 23(A) shows the results of relative change in adhesion strength for SMP rubbery-state adhesion on rough glasses, which shows saturations at preload smaller than 204 kPa. For all preload values studied, the adhesion strength decreases monotonically as the surface becomes rougher. FIG. 23(B) shows the results of the relative change in adhesion strength for SMP R2G adhesion on rough glasses, where the adhesion strength values are saturated under preload values smaller than 204 kPa for all glass samples. The adhesion strength is lower on rougher glasses at small preload values but becomes larger at large preload as the surface becomes rougher.

[0237] Morphologies and Roughness Power Spectra of Rough Surfaces

[0238] Rough glass samples are fabricated using abrasive blasting techniques. The surface morphology of a smooth glass (referred to as sample #0) was measured using the Cypher ES Environmental Atomic Force Microscope (AFM) with a pixel size of ~ 0.06 pm. The other samples (rough glass samples #1 - #4 and samples of other materials such as acrylic, polyimide, steel, pottery, brick, stone, and wood) are measured using the Keyence VK-X1100 Optical Profilometer with a pixel size of ~ 0.6837 The power spectra of surface roughness C(q) (or power spectral density) are calculated in MATLAB from the measured 2D surface heights z = h(x) = h(x,y) as defined (see Eq. (30) for more details) where z = h(x) = h(x,y) is the surface height measured from the average surface plane, defined such that h(0)-0. The < - - > stands for ensemble averaging, i.e., averaging over a collection of different surfaces with identical statistical properties.

[0239] FIG. 24(A) shows the surface morphology of the smooth glass surface (sample #0) and FIG. 24(B) shows the corresponding power spectrum of the surface roughness, which has a low roll-off wavenumber on the order of 106 rrr1, corresponding to = 1 pm, reflecting the largest surface roughness features in the smooth glass and indicating a very smooth surface.

[0240] FIG. 25(A) to 25(D) show the surface morphologies of the rough glass surfaces (sample #1 - #4) and FIG. 25(E) \shows the corresponding power spectrum of the surface roughness, where the smallest wavenumber -16828 rrr1corresponds to the lateral dimension of the measured area and the largest wavenumber 6068503 corresponds to the pixel size in measurement.

[0241] The surface area of a randomly rough surface is calculated through the power spectra of the surface roughness according to (Advanced Materials Interfaces 2019, 6, 1801542; Acta Biomaterialia 2012, 8, 282-288) where and is the nominal contact area (or planar projection of true contact area).

[0242] The elastic energy at the contact interface per unit area under conformal contact is given by (Advanced Materials Interfaces 2019, 6, 1801542; Acta Biomaterialia 2012, 8, 282-288) Notice that for rough fractal surfaces, key roughness quantities such as the real contact area given by Eq. (32), and the elastic energy at the contact interface given by Eq. (33), depend on the magnification (Beilstein Journal of Nanotechnology 2019, 10, 1-8) or the high wavenumber cut-off cji (ACS Appt Mater Interfaces 2022, 14, 46239-46251). For small fractal surfaces with D < 2.5 (for most surfaces, D < 2.3 (Journal of the Mechanics and Physics of Solids 2021, 151, 104402)), it was shown that the pull-off force is converged at fine-scale roughness with a certain dimension (International Journal of Solids and Structures 2016, 81, 244-249; Nature 2000, 405, 681-685; MRS bulletin 2007, 32, 473-478). Here we take the high wavenumber cut-off as the value with the corresponding wavelength at the order of minimal value for adhesive interaction distance of the PDMS and SMP. For SMP R2G adhesion, Therefore, is estimated with a smallest cut-off wavelength and the measurement resolution was set at the order of (close to the length scale of the bridging and the coalescence of microcracks with the main crack in an epoxy / aluminum adhesive interface (Beilstein Journal of Nanotechnology 2019, 10, 1-8)), leading to a value of 6068503 at the order of 107 nr1, which is at the same order of magnitude as the resolution for convergence of key roughness qualities given in the literature (MRS bulletin 2007, 32, 473-478). The rms amplitude and slopes of different materials are also measured (shown in FIG. 26) at a pixel resolution of ~ 0.6837 using the Keyence VK- X1100 Optical Profilometer.

[0243] Results and Discussions

[0244] Five randomly rough glass substrates (labeled as sample #0 to sample #4) with ranging from 2.8 nm to 9.2 urn, and h'ms (the root-mean-square surface slope) ranging from 0.07 to 0.67 were prepared using abrasive blasting techniques (see “Morphologies and Roughness Power Spectra of Rough Surfaces” and FIG. 24, FIG. 25, FIG. 26). The adhesion strength of the SMP was measured on rough glass surfaces under different conditions: 1) SMP state adhesion, i.e., contact and detach in the glassy state; 2) SMP R2G adhesion, i.e. , contact in the rubbery state and detach in the glassy state; 3) SMP rubbery-state adhesion, i.e., contact and detach in the rubbery state. As a control, PDMS (polydimethylsiloxane) elastomeric adhesive samples were also fabricated and tested.

[0245] The adhesion test results of SMP R2G adhesion and SMP rubbery-state adhesion versus hms presented in FIG. 4(C) show that the SMP R2G adhesion is orders of magnitude stronger than the rest, and it actually increases with surface roughness, in contrast to the conventional behavior of reduced adhesion strength as the surfaces become rougher for elastomeric adhesives and SMP rubbery-state adhesion. In the following, in-depth mechanics analyses are presented to reveal the underlying mechanisms responsible for the differences between the strengths of the SMP R2G adhesion, SMP rubbery-state adhesion, SMP glassy-state adhesion, and PDMS adhesion on rough surfaces.

[0246] Mechanics models on rough surfaces can be classified into two categories (The journal of adhesion 2021, 97, 509-527) according to the assumptions on the adhesion regimes of the asperities: 1) those in the energy-controlled adhesion regime or JKR-type model (A. Mathematical and Physical Sciences 1975, 345, 327-342; D. Tabor, "Surface forces and surface interactions" in Plenary and invited lectures. (Elsevier, 1977), pp. 3-14); and 2) those in the strength-limited regime, i.e., DMT-type model (Proceedings of the Institution of Mechanical Engineers, Part J: Journal of Engineering Tribology 2019, 233, 1585-1593; Tribology International 2019, 130, 36-42). To calculate the adhesion force on a rough surface, it is necessary to determine whether the adhesion regime is in the energy-controlled regime or strength-limited regime.

[0247] Examples above have showed that the adhesion regime of SMP R2G adhesion with an asperity with an effective radius of curvature (FIG. 27(A)) could be estimated from a flat punch model (inset (iv) in FIG. 27(A)), which depends on the parameter where is the contact radius between the adhesive and an asperity, is the theoretical adhesion strength, is the effective work of adhesion, and the effective contact modulus defined as being the Young’s modulus and Poisson’s ratio of the adhesive and adherend substrate, respectively. Since the modulus of the glass substrate is much larger than that of the adhesives, the effective modulus can be approximated as

[0248] On the other hand, for SMP glassy-state adhesion (FIG. 27(B)), SMP rubbery-state adhesion (FIG. 27(C)) and PDMS adhesion (FIG. 27(D)), the adhesion regime of a single asperity and the adhesive can be estimated using the adhesive contact mechanics of elastic spheres according to the Tabor parameter, as (Journal of Colloid and Interface Science 1980, 77, 91- 101)

[0249] With the elastic and adhesive parameters of the SMP and PDMS (Soft Matter, 2019, 75(1), 30-37; Proceedings of the National Academy of Sciences 2017, 774, E4344-E4353), the critical contact radius AN or effective radius of curvature determining the adhesion regimes between an asperity and the SMP / PDMS adhesives can be estimated according to Eq. (27) and Eq. (28), with results listed in Table 4, where is adhesive interaction distance.

[0250] Table 4. Summary of the elastic / adhesive parameters and adhesion regimes.

[0251] Since the radius of the adhesive sample is 2.5 mm and the maximum size of the surface asperities are typically in the order of serval micrometers or smaller, the adhesion regimes between the SMP / PDMS adhesives and the asperities are close to the strength-limited regimes for all the adhesion states. Under the strength-limited adhesion regime assumption, the apparent adhesion strength on rough surfaces can be obtained through the convolution of the elementary tension-separation law with a distribution of gaps, and for the Maugis potential, simplified as where A is the true contact area and is the nominal contact area (planar projection of true contact area) (The journal of adhesion 2021, 97, 509-527). The pull-off force is obtained by multiplying the apparent adhesion strength by the nominal contact area as

[0252] In other word, the adhesion force is linearly proportional to the true contact area.

[0253] For SMP R2G adhesion, the full contact configuration with a rough surface in the rubbery state under a preload (inset i in FIG. 27(A)) is frozen after cooled to the glassy state (inset ii in FIG. 27(A)) due to shape-locking. Consequently, the SMP R2G pull-off force should be linearly proportional to the true area of the rough surface and follow the scaling law wher are the SMP R2G pull-off force, adhesion strength and the surface area of a reference surface (e.g., smooth glass sample denote the

[0254] SMP R2G pull-off force, apparent adhesion strength and true area of the rough surface. The true area of surfaces with random roughness can be calculated from its power spectra of surface roughness C(q) (or power spectral density) (FIG. 24 and FIG. 25) (Physical review letters 2005, 95, 124301; Physical review letters 2022, 89, 245502) according to where in which qo and qn are the low and high cut-off wave vectors, respectively. Here, the smallest wavevector go -16828 nr1corresponds to the lateral dimension of the measured area and the largest wavevector q-i - 107 nr1corresponds to the minimal adhesive interaction distance of the SMP and PDMS (FIG. 24 and FIG. 25).

[0255] FIG. 27(A) shows the model prediction of Eq. (31) and the experimental measurement of the normalized SMP R2G pull-off force versus the normalized surface area S / So, in which the experimental conformal contact was achieved under a preload of 255 kPa. Clearly the experimental results agree very well with the scaling law given by Eq. (31), validating that the SMP R2G adhesion strength is linearly proportional to the true contact area of the rough surface. This behavior is similar to the toughening mechanism of silicon {Journal of the American Ceramic Society 1995, 78, 3201-3208) and epoxy / aluminum joints (86), where the fracture resistance was shown to be proportional to the roughness number

[0256] For SMP glassy-state adhesion, the adhesive block attaches onto and detaches from the adherend in the same glassy state. However, due to the high stiffness of SMP in the glassy state, even under a relatively large preload, only the tips of the asperities on the rough glass surface can make contact with the SMP (insets i and ii in FIG. 27(B)). Consequently, the glassy-state adhesion of SMP corresponds to a tip contact configuration (insets iii to iv in FIG. 27(B), as also illustrated in FIG. 4(B)-(i)), in contrast to the full contact configuration (as illustrated in FIG. 4(B)-(iii), and inset ii in FIG. 27(A)) in the R2G adhesion. For SMP rubbery- state adhesion and PDMS adhesion, although full contact may be realized under sufficiently large preloads due to the low modulus of the adhesives (inset i in FIG. 27(C) and 27(D)), the full contact configuration cannot be sustained due to the elastic recovery of the adhesive material. The stored strain energy at the contact interface during contact formation will cause the adhesives to spring back, and the contact between the adhesives and the rough surface will become a tip contact configuration prior to pulling-off (inset ii in FIG. 27(C) and 27(D)).

[0257] It was shown that when the contact area is relatively sma the true contact area is inversely proportional to the root-mean-square surface slope where is the contact pressure and K is a constant on the order 1. Combining Eq. (29) and Eq. (34), it follows that the pull-off forces of glassy-state SMP rubbery-state SMP and PDMS on rough surfaces should scale inversely proportional to where are the corresponding values on the reference surface

[0258] (the smooth glass, sample #0 here).

[0259] FIG. 27(B) to 27(D) show the measured pull-off forces (black circles) of the glassy-state SMP, rubbery-state SMP and PDMS on rough glass samples, compared to the theoretical predictions (solid lines) given in Eq. (35). The agreement between the experimental data and theoretical predictions validates the scaling law in Eq. (35). This further shows that the adhesion behavior of the rubbery-state SMP is the same as that of conventional elastomeric adhesives, as shown in FIG. 4(C).

[0260] Example 9. Stickiness of SMP Adhesion on Rough Surfaces

[0261] One index commonly used to evaluate the adhesion performance of adhesives on rough surfaces is the so-called stickiness or tackiness parameter (The journal of adhesion 2021, 97, 509-527), i.e., the possibility to sustain macroscopic tensile pressures or a nonzero contact area without external load. An adhesive sticky (also called tacky) to a surface will yield a considerable pull-off force while the adhesion of a non-sticky (or non-tacky) adhesive to a rough surface is greatly weakened compared to that on smooth surfaces. There are several interpretations of the stickiness criteria (Tribology International 2020, 746, 106031 ; Proceedings of the National Academy of Sciences 2019, 776, 13774-13779). The best known among them is the so-called golden rule used in pressure sensitive adhesives — the Dahlquist criterion (The Journal of Adhesion 1991, 34, 189-200; Advanced Functional Materials 2021, 31, 2104296), which states that pressure sensitive adhesives are sticky to ordinary surfaces with a shear storage modulus MPa or an equivalent tensile storage modulus MPa. Gecko's adhesive pad has an effective modulus on the order of 100 kPa (Journal of Experimental Biology 2006, 209, 3558-3568) due to its hierarchical and hairy structure and would therefore be considered sticky to various adherends. However, the underlying physics of this criterion (Physical review letters 2005, 95, 124301 ; Friction, 2022, 77, 1027; The journal of adhesion 2021, 97, 509-527) is the competition between the gained surface energy y (under a thermal equilibrium condition) and the stored elastic strain energy vel per unit area during contact. If on a surface, the adhesive is sticky and can produce observable adhesion on the surface. Otherwise, the adhesion will be easily diminished. The elastic energy at the contact interface per unit area (Physical review letters 2005, 95, 124301 ; Physical review letters 2022, 89, 245502) is given by

[0262] Contact between solids occurs commonly via van der Waals forces (The journal of adhesion 2021, 97, 509-527; Proc. Natl. Acad. Sci., 2004, 707(21), 7851-7856), and the surface energy The ratio on glass surfaces can be estimated according to Eq. (26) for various surface roughness. We measured the SMP rubbery-state and R2G adhesion under different preloads, normalized by their values on the reference smooth surface (sample #0) under full contact conditions (preload equal to 255 kPa). The results were plotted as a function of in FIG. 28 (see FIG. 23 and “Morphologies and Roughness Power Spectra of Rough Surfaces" for more details). For SMP rubbery-state adhesion (FIG. 28(A)), the adhesion strength was reduced by more than 50% when This is similar to the adhesion behavior of elastomers, where the preload has little effect on the adhesive strength since the enhanced contact at higher preload would be easily reverted by the release of stored elastic energy upon unloading. In contrast, the SMP R2G adhesion shows a strong dependence on the preload since the contact configuration, frozen when the SMP is cooled to the glassy state, is determined by the elastic deformation during preloading. As a result, the SMP R2G adhesion is strongly retained (at least 90%) even when Under large preloads (for example, 255 kPa), full contact is achieved during the contact process and fixed upon R2G transition. Consequently, when vei / y increases as the surface becomes rougher, the R2G adhesion strength also increases. In fact, our experiments showed that, even for a very rough surface with a value as high as 100, the adhesion force was not weakened, rather it increased with surface roughness if the preload was large enough to ensure conformal contact. Consequently, the R2G SMP remains sticky to surfaces that are challenging to most elastic adhesives.

[0263] However, due to the limitations of the material system used here, the preload required to achieve the stickiness for R2G adhesion is up to 200 kPa, especially when there are mesoscopic structures on the object surfaces (as with the serrated steel shown in FIG. 7(B)). This preload might damage relatively fragile target objects in engineering applications. In such cases, similar E44 SMP with a different monomer to crosslinker ratio (Polym. Chem., 2015, 6(16), 3046-3053) or a new material system (Proceedings of the National Academy of Sciences 2019, 116, 13774-13779; Small, 2019, 15(4), 1803870) with a lower modulus in the rubbery state should be used in extending the working principle of R2G adhesion to practical applications that require a smaller preload to achieve stickiness.

[0264] Example 10. Switchability of SMP Adhesion on Rough Surfaces

[0265] As discussed above, the shape conformability during contact through tunable modulus effect of the SMP adhesive in the rubbery state and the shape-locking effect when the SMP adhesive is cooled to the glassy state produce drastically increased adhesion on rough surfaces, with adhesion strength linearly proportional to the true surface area, as given by Eq. (31). These effects make the SMP adhesives sticky to various rough surfaces, resulting in R2G adhesion strength of more than 1 MPa, as shown in FIG. 29(A), with the root mean square amplitude of surface roughness up to about 100 (see FIG. 26). On the other hand, the shape memory effect enables the release of stored elastic energy at the contact interface and recovers the adhesive to tip contact with rough adherends, resulting in decreases in SMP adhesion on rougher surfaces according to Eq. (35). The increased R2G adhesion and the decreased rubbery-state adhesion make the SMP an extraordinarily strong and highly switchable adhesive, as demonstrated in FIG. 29(A) for a wide range of adherend materials.

[0266] According to Eq. (31) and Eq. (35), the switchability of the SMP adhesive SR, defined by the ratio between the strong SMP R2G adhesion for attachment and weak SMP rubbery-state adhesion for detachment is given by

[0267] Since the area increase is relatively small (<0.25), S / So should be close to 1. Therefore, Eq. (36) can be approximated as which gives the scaling law for adhesion switchability, as confirmed by the measured values of adhesion switchability on different materials in FIG. 29(B).

[0268] Although only the adhesion performance on nominally flat surfaces with different roughness values are measured, the working principle of the SMP smart adhesive can be extended to curved macroscopic objects, as shown in (Science advances 2020, 6, eaay5120; Proceedings of the National Academy of Sciences 2019, 116, 13774-13779). However, for the prediction of R2G adhesion strength and switchability on surfaces that contain both mesoscopic waviness and microscopic roughness (e.g., serrated steel shown in FIG. 7(B)), the formulae for calculating the true contact area may need to be modified due to the possible coupling between waviness and roughness, as well as anisotropy in surface structures and corresponding power spectra density (The European Physical Journal E 2009, 29, 275-284), which requires further investigation.

[0269] Example 11. Materials Selection and Characterisation for Adhesive Fibrils

[0270] Working Principle of Materials for R2G Adhesive Fibrils

[0271] The working principle depicted in FIG. 5(A) and FIG. 30(A)-(i) requires that the materials for

[0272] R2G adhesive fibrils must satisfy: 1) Sufficient stickiness (elastic modulus Ef <1 MPa) in the rubber-phase to adhere to rough surfaces according to the Dahlquist criterion (The journal of adhesion 2021, 97, 509-527; Sci. Robot., 2020, 5(49), eabd9120) Transition between JKR- DMT adhesion regimes occurs in fibrils of millimeter size. Theoretical studies (Journal of The Royal Society Interface 2004, 1, 35-48; J. Meeh. Phys. Solids, 2020, 139, 103937) indicate that this transition depends on geometric, elastic, and interfacial properties of the adhesive system. The selection of materials and the design of fibrillar structures are essential for the proposed R2G adhesive system to exhibit distinctive, easily achievable JKR / DMT-like regimes. Elastic modulus Ef, theoretical adhesion strength oth and work of adhesion Mod are key parameters affecting the JKR-DMT transition. The JKR-DMT transition of an elastomeric fibril on a rigid substrate is governed by the dimensionless parameter (Journal of The Royal Society Interface 2004, 1, 35-48) where is the radius of the fibril and Ef and being the Young’s modulus and Poisson’s ratio of the fibril material, respectively. The adhesion regime is DMT- like when and JKR-like for To ensure that the adhesive fibril remains in the DMT-like regime to support high load (FIG. 5(A)-(ii) and FIG. 30(A)-(i)), must be smaller than 0.5; and to detach easily in the JKR-like regime (FIG. 5(A)-(iii) and FIG. 30(A)-(i)), xr must be larger than 0.5 during operation.

[0273] Sample Preparation

[0274] To demonstrate the feasibility of our adhesive paradigm, we selected a thermally tunable epoxy SMP E44-SMP (Polymer Chemistry 2015, 6, 3046-3053) as the adhesive material that exhibits R2G transition (FIG. 30(B)). The E44-SMP was prepared by following the protocols disclosed in the Examples above.

[0275] Modulus Characterization

[0276] By following the protocols disclosed in the Examples above, SMP samples with a rectangle shape (30 mm in length, 6 mm in width and 2 mm in thickness) were prepared using moulding methods. Dynamic Mechanical Analysis (DMA) of the SMP materials were conducted to measure the storage modulus E’ (unless stated otherwise, storage modulus herein refers to tensile storage modulus) under different temperatures.

[0277] FIG. 30(D) shows the typical stress-strain curves, where the SMP shows an elastic-plastic behaviour in the glass-phase (30 °C). In consideration of the small deformation in the stiff glass-phase, the SMP is regarded as a linear elastic material in the glass-phase. In the rubber- phase, the SMP shows a hyperelastic behaviour which can be described by the first-order Ogden-Roxburgh model (Acta Mechanics Sinica 2017, 33, 543-554; Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences 1999, 455, 2861-2877): where H4iais the strain energy density, l / l / devand l / l / voiits deviatoric and volumetric parts, the initial bulk modulus, the initial shear modulus, the Ogden parameter, the hprincipal stretch and the total volume change. The fitted parameters are = 0.1785 MPa, = 2.7265 and = 0.2256 MPa1.

[0278] Pull-Off Tests

[0279] By following the protocols disclosed in the Examples above, pull off tests using a flat punch (R - 10 mm) on E44-SMP samples (diameter in 100 mm) with various thicknesses were conducted to measure the interfacial theoretical strength trtti and the work of adhesion wadof the E44-SMP material under the rubber-to-glass condition and the rubbery phase. A homemade testing platform was constructed, as illustrated in FIG. 31(A). Attached to the Instron 5566, the testing platform was equipped with linear and tip / tilt stages to ensure parallel alignment between the flat punch and the SMP samples. FIG. 31(B) shows surface morphologies of the glass punch (FIG. 31 (B)-(i)) and the E44-SMP sample (FIG. 31 (B)-(ii)), which are both very smooth with an arithmetic mean roughness value of Sa = 1.752 nm and Sa = 1 .870 nm, respectively. The test processes are illustrated in Fig. S4c. For the test of the R2G adhesion (FIG. 31 (C)-(i)), the SMP was heated to the rubber-phase (90 °C) first. Then the glass punch was attached to the rubbery-phase SMP at a constant speed of 10 ftm / s until a certain preload (the preload was increased until a pull-off force plateau is reached, as shown in FIG. 31(D)), dwelled for 5 minutes and cooled down to the glass-phase. Finally, the glass punch was pulled away at a speed of 100 jum / s, with the measured maximum force giving the pull-off force. The adhesion tests of rubbery adhesion (FIG. 31 (C)-(ii)) were the same as those for the R2G adhesion, except that the SMP is kept in the rubber-phase throughout the process.

[0280] The values of can be obtained through non-linear fitting of the theoretical expression for the pull-off force as a function of sample thickness as shown in FIG. 30(E) (Journal of the Mechanics and Physics of Solids 2020, 139, 103937), according to where is the thickness of the E44-SMP substrate, R is the punch radius; E and are the elastic modulus and Poisson’s ratio of the E44-SMP substrate, and are dimensionless functions of and v. Values of ath and wad can be determined by fitting experimental data to Eq. (40) and are given in FIG. 30(F).

[0281] Results and Discussions

[0282] FIG. 30(A)-(ii) to 30(A)-(iv) provide the range of elastic modulus E, adhesion strength aa, and work of adhesion Madof typical materials for fibrillar adhesives on glass substrates. Among them, animal adhesive fibrils (Mechanics of Materials 2005, 37, 275-285) and carbon nanotube (CNT) (Science 2008, 322, 238-242) fibiris are stiff (>1 GPa) with a low (~10 , exhibiting DMT-like regime for small fibril sizes. Other synthetic materials, such as Polydimethylsiloxane (PDMS), are relatively soft, with large (~ 0.1 - 10 J / m2). The adhesion strength of these synthetic fibrils is typically below 1 MPa, and achieving the DMT-like regime is challenging (Progress in Materials Science 2021, 120, 100823) unless the fibril tip is optimized (e.g., using mushroom-shaped tips (Soft Matter 2011, 7, 5545-5552)) at microscales (<10 . In contrast, smart materials such as hydrogels (Advanced Materials 2021, 33, 2102983; Proceedings of the National Academy of Sciences 2019, 116, 13774-13779) and SMPs (Adv. Mater. Interfaces, 2019, 6(3), 1801542; Science advances 2020, 6, eaay5120; Polymer Chemistry 2015, 6, 3046-3053; Adv. Mater., 2021, 33(6), 2000713) offer a wide range of values. However, hydrogels are not suitable for long-term applications due to dehydration. SMPs (Adv. Mater. Interfaces, 2019, 6(3), 1801542; Science advances 2020, 6, eaay5120; Polymer Chemistry 2015, 6, 3046-3053), therefore, are good candidates for smart adhesive fibrils with good material stability, tunable modulus (0.1 kPa- 10 GPa) during R2G transition, and strong adhesion to various materials (~MPa).

[0283] The highly tunable modulus of E44-SMP, ranging from 2 GPa at room temperature (25-30 °C) to 1 MPa at 90 °C, satisfies the Dahlquist criterion for stickiness, as shown in FIG. 30(C). FIG. 30(C) shows the storage modulus of the SMP under different temperatures. At the room temperature (25 - 30 °C), the SMP is in the stiff glassy phase (E’ ~ 2 GPa). When heated above the glass transition temperature (~ 45 °C), the SMP shows a sharp decrease in storage modulus. At around 90 °C, the storage modulus falls below 1 MPa. Tensile tests were further carried out on standard universal tensile test machine (Instron 5566) with a furnace under 30 °C and 90 °C.

[0284] FIG. 30(D) presents typical stress-strain curves of the E44-SMP at different temperatures. In the glass-phase (30 °C), it exhibits a linear elastic behaviour; whereas in the rubber-phase (90 °C), it shows a hyperelastic behaviour that can be characterized by the first-order Ogden- Roxburgh model (Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences 1999, 455, 2861-2877). Oth and are the two other key parameters governing the JKR-DMT transition.

[0285] Pull-off tests were conducted using flat punches on SMP samples of various thicknesses, as illustrated in FIG. 30(E). By fitting the measured results to the theoretical expression (J. Meeh. Phys. Solids, 2020, 139, 103937) relating the pull-off force Pc and the sample thickness, we can determine the values of as shown in FIG. 30(F). The results show that, under R2G conditions (contact at 90 °C, detach at 30 °C), E44-SMP demonstrates a high adhesion strength of 1.893 MPa compared to the rubber-phase (90 °C contact / detach) strength of 0.4202 MPa. However, the work of adhesion in the glass-phase (1.61 J / m2) is lower than that in the rubber-phase (20.27 J / m2). These measured values will be used in designing the R2G smart fibrillar adhesives.

[0286] Example 12. Fabrication of Individual Adhesive Fibrils and Adhesive Fibril Arrays

[0287] To demonstrate the capabilities of the R2G fibrillar adhesive, we fabricated adhesive fibrils with different radii (FIG. 5(B)-(i)) and fibril arrays with varying apparent contact areas (FIG. 5(B)-(ii)) using a thermally-controlled E44 shape-memory polymer (E44-SMP).

[0288] Fabrication of Individual Adhesive Fibrils

[0289] The fabrication was done by following the protocol disclosed in Example 8.

[0290] Fabrication of Adhesive Fibrils Arrays

[0291] The fabrication was done by following the protocol disclosed in Example 8.

[0292] Example 13. Adhesion Scaling of Individual Fibrils

[0293] FIG. 5(C) presents a summary of the adhesion scaling performance of E44-SMP R2G adhesive fibrils (FIG. 5(B)-(i)) compared to those made of other elastomers (FIG. 32). Results and Discussions

[0294] Natural adhesive fibrils are made of stiff materials such as keratin or resilin, with elastic modulus in the range of 1-10 GPa (Biomimetics 2017, 2, 10). They approach the DMT-like regime mainly by reducing fibril size to the micro / nano-scale (Mechanics of Materials 2005, 37, 275-285) and optimizing the tip geometry, such as using a mushroom shape. However, as mentioned in above sections, existing synthetic adhesive fibrils are usually made of some soft to moderately stiff materials, such as Polydimethylsiloxane (PDMS), Polyurethane (PU) and Polyurethane acrylate (PUA), with relatively large work of adhesion values (~ 0.1 - 10 J / m2) in consideration of limitations such as the adaptability to surface roughness and ease of fabrication. The DMT-like regime is usually inaccessible for these materials unless the fibril tip is optimized (e.g., using a mushroom-shaped tip) at a moderately small size (e.g., < 10 jttm). The maximum adhesion strength of the adhesive fibrils reported so far is summarized in FIG. 32. The adhesion strength decreases with increasing fibril radius, due to stress concentration at the contact interface in the JKR-like adhesion regime. The fibril adhesion strength of homogenous flat fibrils scales according to a power law as the fibril radius increases. The fibril adhesion strength due to Van der Waals interactions changes from about 100 kPa to less than 25 kPa when the fibril radius increases from 100 gm to 1 mm. In nature, the adhesive fibrils are usually not homogeneous but graded with a stiffer steam and a softer tip, which has inspired the development of gradient adhesive fibrils (Applied Physics Letters 2015, 106, 201604; Soft Matter2020, 16, 9534-9542) that can alleviate stress concentration to enhance adhesion strength (~75 kPa when the fibril radius is scaled to millimeter size). Further optimization of the shape of the fibril tip by using the mushroom shape may allow for a more uniform distribution of stress at the contact interface, hence increasing the adhesion strength to around 100 kPa when the fibril radius is scaled to millimeter size. The combination of gradient-modulus and mushroom-shape designs can further increase the adhesion strength of adhesive fibrils to around 200 kPa when the fibril radius is scaled to millimeter size. In spite of these efforts, while scaled to the millimeter scale, current designs still fail to reach an adhesion strength in the order of 1 MPa, which is desired for many high payload applications.

[0295] With conventional elastomeric adhesive fibrils, the maximum radius for the DMT-like regime, s ~ 10 ,um. In contrast, E44-SMP R2G adhesive fibrils can maintain the DMT-like regime when its radius is as large as 2.5 mm, as demonstrated by experiments and confirmed by finite element (FEA) simulations (FIG. 5(C)). E44-SMP R2G adhesive fibrils can achieve adhesion strength of 1.893 MPa with radius up to the millimeter-size.

[0296] FIG. 5(D) (and FIG. 33) shows the trade-off between adhesion strength and switchability. For conventional elastomeric adhesive fibrils, increasing the adhesion strength to ~2 MPa significantly reduces the adhesion switchability to ~1, making the adhesion nearly non- switchable. In contrast, E44-SMP R2G adhesive fibrils, under R2G conditions, achieve adhesion switchability of 20 (circles in FIG. 5(D)) due to JKR-like detachment in the rubberphase. By subjecting R2G adhesive fibrils to a large preload in the rubber-phase to induce fibril buckling, the detachment strength can be further reduced by two orders of magnitude, leading to an extraordinary adhesion switchability of approximately 2000 (stars in FIG. 5(D)). Consequently, the R2G adhesive fibrils can achieve unprecedented adhesion strength and switchability.

[0297] Example 14. Adhesion Scaling of Fibril Arrays

[0298] To fully benefit DMT-like adhesion, the radius of an individual E44-SMP R2G adhesive fibril cannot exceed *(FIG. 5(C)). Devices requiring higher load capacity can employ fibril arrays with radius of each fibril smaller than

[0299] Comparison of the Single Fibril and Fibril Arrays

[0300] The SMP fibril arrays and their single fibril counterparts (as shown in FIG. 34) with comparable apparent contact areas were prepared utilizing molding as described (Sci. Adv., 2020, 6(7), eaay5120). The SMP fibril arrays (FIG. 34(A)) are characterized by smaller SMP fibrils (R = 2.5 mm, h = 10 mm) arranged in a hexagonal pattern (center to center distance of 6 mm) on an SMP backing of 5 mm thickness. The SMP single fibril adhesives are cylinders fabricated on the larger SMP backing plates of 2 mm thickness. The backing plates are designed to enhance the contact area between the adhesives and the backing glasses for fixation convenience. The total height of the SMP single fibril adhesive including the backing is 15 mm, equal to the total height of the SMP adhesive fibril arrays. For fibril arrays, the contact area is the sum of the areas of every single fibril tip area (areas in the circles as shown in the illustrations in FIG. 34(A) ) and the apparent contact area is the whole area the array has covered (areas in the hexagons as shown in the illustrations in FIG. 34(A) ). For the SMP single fibril adhesives (see FIG. 34(B)), the apparent contact area is identical to the contact area, both equal to the tip area of the fibril. FIG. 35 shows the measured load capabilities of the SMP R2G fibril arrays, SMP R2G single adhesive fibrils under the R2G condition as compared that in the rubber-phase, with those of the elastomeric fibril arrays as control. As the apparent contact area increases, the R2G pull- off force of the SMP R2G fibril arrays and SMP R2G single adhesive fibril scales as

[0301] In contrast, the rubbery adhesion forces of both single SMP fibrils and SMP fibril arrays are much smaller than their R2G values and similar to those of the elastomeric fibrillar adhesives

[0302] When the apparent contact area is scaled to 1000 mm2, the load capability (i.e., pull-off force) of the SMP R2G fibril arrays is 1.4 folds of the SMP R2G single adhesive fibril and 23.89 folds of the elastomeric fibril arrays.

[0303] Results and Discussions

[0304] The adhesion load capacity of a fibril array device scales with its apparent contact area. Our experimental measurements show that, for a device with the apparent contact area of 1000 mm2, the load capability of R2G fibril arrays is 1.4 times that of the R2G single fibril and 23.89 times that of the elastomeric fibril arrays (FIG. 35).

[0305] FIG. 5(E) shows the measured scaling performance of the effective adhesion strength of fibril arrays compared to the elastomeric ones. The measured rubber-phase array adhesion strength of the SMP fibril arrays follows a scaling law , similar to the scaling law observed for elastomeric ones, where is the apparent contact area.

[0306] In contrast, R2G fibril arrays show a much slower decay rate of resulting in an adhesion efficiency of 57.8% up to the apparent contact area of 1000 mm2. The R2G fibrillar adhesive can thus utilize more than half of its full potential, representing a significant improvement over elastomeric adhesives, even when compared to the state-of-the-art result (26%) achieved through complex controlled-load-sharing designs (Proceedings of the National Academy of Sciences 2017, 114, E4344-E4353). Dividing a single fibril into fibril arrays can further enhance adhesion switchability, as demonstrated by the experimental results in FIG. 5(F). Through switching between the R2G conditions and rubber-phase, single adhesive fibrils have intermediate switchability (~20) that decreases with increasing contact area. After dividing into fibril arrays with each fibril radius of 2.5 mm, the switchability increases slightly with the contact area. Preload-induced buckling of fibril arrays can reduce the rubber-phase detachment strength, leading to a dramatic increase in switchability (>1000).

[0307] Example 15. Viable Parameter Space for JKR-DMT Transition and Validation

[0308] The transition between adhesion regimes of an elastomeric fibril on a rigid substrate is governed by Eq. (38), which helps determine the size of fibril radius where the adhesion regime switches from the DMT-like regime under R2G condition to the JKR-like regime in the rubber-phase. The solid lines and the circular / diamond symbols in FIG. 36(A) represent the theoretical predictions and FEM simulation results of the transition between DMT-like and JKR-like regimes under R2G conditions (solid line) and in the rubber-phase (dotted line). Taking as the critical value at which the transition occurs, the results indicate the fibril radius R must be smaller than 3.2 mm to maintain the DMT regime under R2G conditions, and larger than 475 pm to reach JKR regime in the rubber-phase, suggesting a viable fibril size range of 475 mm for strong adhesion and easy detachment.

[0309] The criterion based on / f assumes infinite fibril height and material incompressibility To assess more accurately the adhesion regime transition with finite fibril height and different Poisson's ratio, we conducted FEA simulations (ABAQUS 2022) with a cohesive zone model.

[0310] FEA simulations of the adhesion regimes of the SMP adhesive fibril

[0311] Finite Element Analysis (FEA) simulations were conducted using ABAQUS / Explicit and a cohesive zone model to capture the adhesion regimes for different fibril radii (as shown FIG. 37). An axisymmetric model with fibril height h equals to 4 times of the fibril radius R is used, as shown in FIG. 37(A). The E44-SMP was regarded as a linear elastic material with an elastic modulus of 2 GPa and Poisson’s ratio of 0.35 in the glass-phase, and a hyperelastic material described by the first-order Ogden model in the rubber-phase 0.1785 MPa 2.7265 D = 0.2256 MPa-1). A thin layer of cohesive elements was used to simulate the adhesive behavior. The cohesive elements were tied to the bottom of the SMP fibril on the top surface and fixed at its bottom surface. A pre-crack with the length of 1 / 200 of the fibril radius was introduced at the outer edge by deleting one cohesive element. A linear traction-separation law was used to model the adhesive interactions, as shown in FIG. 37(B). The stiffness of the cohesive element was set as to ensure a left triangular and a maximum nominal stress criterion = 1 for damage initiation. The SMP was modeled with CAX4R elements, and the minimum element size was set to 1 / 200 of the fibril radius. The cohesive layer was modeled with COHAX4 elements, and the element length set to the minimum element size of the SMP. Detailed parameters of the FEA simulations are given in Table 5.

[0312] Table 5. Summary of the modeling parameters for FEA simulations.

[0313] Results and Discussions

[0314] The results show that, under the R2G condition, the adhesion regime transition occurs at a fibril radius of 3 mm and, in the rubber-phase, it occurs at 0.4 mm.

[0315] FIG. 36(B)-(i) and 36(B)-(ii) show typical detachment processes using FEA simulations in the DMT-like and JKR-like regimes. In the DMT-like regime (FIG. 36(B)-(i), R = 1 mm, R2G condition), the fibril detaches uniformly at the contact interface, while in the JKR-like regime (FIG. 36(B)-(ii), R = 1 mm, rubber-phase), detachment initiates at the periphery and propagates inwards. FIG. 36(C)-(i) to 36(C)-(iii) (R2G condition) and FIG. 36(C)-(iv) to 36(C)- (vi) (rubber-phase) show normal stress distributions at the contact interface for fibril radii of 0.1 , 1.0, and 10 mm, with uniform stress in the DMT-like regime (FIG. 36(C)-(i), to 36(C)-(ii), and 36(C)-(iv)) and highly concentrated stress at the edge in the JKR-like regime (FIG. 36(C)- (iii), to 36(C)-(v), and 36(C)-(vi)), resulting in strong and weak adhesion, respectively. The FEA simulations validate the theoretical prediction of the JKR-DMT transition criterion for the E44-SMP R2G adhesive fibrils, suggesting a ‘work window’ of radius ranging from 0.5 - 3.0 mm to achieve DMT-like adhesion under R2G condition and JKR-like detachment in the rubber-phase.

[0316] FIG. 38(A) and 38(B) show the FEA results of the fibril adhesion strength under various fibril radii and FIG. 38(C) and 38(D) the normal stress distribution under typical fibril radii. Under the R2G condition, the adhesion regime is DMT-like with an adhesion strength larger than 80% of the theoretical value (FIG. 38(A)) and a uniform stress distribution when the fibril radius is smaller than 3 mm (e.g., R = 0.1 and 1 mm in FIG. 38(C)), and JKR-like with stress concentration at the edge part when the fibril radius exceeds 3 mm (e.g., R = 10 mm in 38(C)), resulting in rapid decay of adhesion strength with increasing fibril radius (FIG. 38(A)). In the rubbery-phase, the adhesion regime is DMT-like with an adhesion strength larger than 80% of the theoretical value (as shown in FIG. 38(B)) and a uniform stress distribution when the fibril radius is smaller than 0.4 mm (e.g., R = 0.1 mm in FIG. 38(D)). When the fibril radius is larger than 0.4 mm, the adhesion regime is JKR-like with stress concentration at the edge part (e.g., R = 1 mm and 10 mm in FIG. 38(D)), leading to rapid decay of adhesion strength with increasing fibril radius.

[0317] Example 16. Characterisation of R2G Adhesive Fibrils

[0318] To validate the parameter space for viable R2G adhesive fibrils, we fabricated single fibrils with different radii R and an aspect ratio h / R - 4 (FIG. 5(B)-(i)) using molding methods (as disclosed in above Examples). Pull off tests on E44-SMP fibrils of various radii under R2G conditions were conducted to measure their adhesion strength, as shown in FIG. 36(D).

[0319] Adhesion Test of an Individual SMP Adhesive Fibril

[0320] To further validate the designed ‘work window’ of our SMP adhesive fibrils, SMP fibrils with different radii R (0.2, 0.5, 1, 2.5, 5, 7.5 and 10 mm) and an aspect ratio of h / R = 4 were fabricated using molding 96. Pull off tests were conducted to measure the adhesion strength of SMP fibrils with different radii. Adhesion measurement details of the SMP adhesive fibrils are shown in FIG. 39. During the pull off tests, the SMP fibril was heated to the rubber-phase (90 °C) and brought into contact with the glass substrate at a speed of 10 jum / s until a sufficiently large preload, dwelled for 5 minutes and then pulled away under 90 °C for rubbery adhesion measurement (FIG. 39(A)) and under 30 °C for R2G adhesion measurement (FIG. 39(B)). The adhesion strength is given by the pull-off force over the contact area. The preload was increased until the pull-off force reaches a plateau, as shown in FIG. 39(C) and 39(D).

[0321] Results and Discussions

[0322] The measured adhesion strength matches the theoretical value (1.893 MPa) for radii up to 2.7 mm, indicating a DMT-like adhesion regime. This remarkable scaling performance of the R2G adhesive fibril, as shown in FIG. 5(C), outperforms the elastomeric fibrils (Acta biomaterialia 2005, 1, 5-13) and advanced fibrillar designs such as gradient fibrils (ACS nano 2018, 12, 1273-1284), mushroom fibrils (Proceedings of the National Academy of Sciences 2017, 114, E4344-E4353; Advanced Materials Interfaces 2022, 9, 2101764) and gradient-mushroom fibrils (Science Advances 2021, 7, eabf5695). These other designs experience a significant loss of adhesion strength when scaled up to the millimeter size (FIG. 5(C) and FIG. 32). Our experimental measurements represent the first systematic validation of the size effect of adhesive fibrils, governed by the ideal cohesive strength, across a wide range of fibril radii from micro to mm scale (Progress in Materials Science 2021, 120, 100823; Acta biomaterialia 2005, 7, 5-13).

[0323] The experimental measurements in FIG. 36(D) also reveal that the adhesion strength of E44- SMP fibrils in the rubber-phase is lower than 100 kPa, significantly below its theoretical strength (420.2 kPa), indicating a JKR-like adhesion regime for the range of fibril radii. FIG. 36(E) depicts the detachment process of an E44-SMP fibril of 1 mm radius under R2G and rubber-phase conditions. Under the R2G condition, the fibril detaches uniformly (FIG. 36(E)- (i)), in agreement with the FEA simulation shown in FIG. 36(B)-(i). In the rubber-phase, the fibril detaches gradually from the periphery (FIG. 36(E)-(ii)), in agreement with the FEA simulation shown in FIG. 36(B)-(ii). The adhesion of E44-SMP fibrils with radii within the ‘work window’ can be switched, on-demand, by heating and cooling, between the (strong) DMT-like and (weak) JKR-like regimes. The switchability can exceed 20 by harnessing the adhesion regime switching through R2G transition (FIG. 36(F)).

[0324] To assess the repeatability of the R2G adhesive fibrils, repeated tests were conducted (FIG. 36(G)). Over 30 testing cycles, the adhesion strength varied within ± 4 % of the mean value in the DMT-like regime, and within about ± 2 % in the JKR-like regime, demonstrating excellent repeatability of the R2G adhesive fibrils.

[0325] We conducted pull off tests on SMP adhesive fibrils with various aspect ratios. FIG. 36(H) shows that the R2G adhesion strength is independent of the aspect ratios, whereas the rubber-phase adhesion strength decreases slightly as the aspect ratio increases (squares in FIG. 36(l)). In the rubber-phase, the adhesion strength of the single SMP adhesive fibril can be further reduced by a large compression until it bulges or buckles. However, when compressed to bulging / buckling shapes, the adhesion strength of the fibrils is highly sensitive to the aspect ratio when it is below 4, resulting in a rapid increase of the adhesion switchability to the order of 2000 (circles in FIG. 36(l)) . When the aspect ratio is small than 4 (e g., 2, see FIG. 40(A)-(i)), the fibril bulges under a large compression, as shown in Fig. S8a-ii. Further increase of compression will lead to bulging of the fibril (FIG. 40(A)-(iii)) but no compression- induced detachment is observed. The adhesion strength reduces gradually with increasing compression preload and plateaus in the end (FIG. 40(A)). On the contrary, when the aspect ratio is larger than 4 (e.g., 4, see FIG. 40(B)-(i)), the fibril will be compressed to buckling under a large preload. After buckling, the SMP fibril will tilt and detach, leaving a small residual contact between the SMP adhesive fibril tip and the adhered, as shown in FIG. 40(B)-(ii), leading to a sharp drop of the adhesion strength (FIG. 40(B)). When the aspect ratio is above 4, the switchability increases slowly with increasing aspect ratio. To balance the requirements for fibril stability, heating time, and sufficient switchability in practical applications, an aspect ratio of 4-5 is chosen for the following studies.

[0326] Example 17. Characterisation of R2G Fibril Arrays

[0327] When the fibril radius exceeds the DMT-limi the adhesion regime becomes JKR-like (FIG. 5(C)), with decreasing strength as the fibril radius increases. One strategy to maintain the load capability (FIG. 5(E)) and switchability (FIG. 5(F)) is to use fibril arrays with each fibril below the DMT-limit in radius (FIG. 5(B)-(ii)). To understand the mechanisms governing the fibril dimensions in the array, we compared the 2D FEA simulation results of normal stress distribution at the pull-off point of a single fibril (R - 29 mm) and fibril array (R - 2.5 mm) with the same apparent contact area (FIG. 41(A)).

[0328] Mechanical Stability and Adhesion Repeatability of the SMP Fibrillar Adhesive

[0329] Lateral collapse and bundling problems of adhesive fibrils (Progress in Materials Science 2021 , 120, 100823) have been a big challenge to the mechanical stability and adhesion repeatability of fibrillar adhesives since they were first proposed (Nature materials 2003, 2, 461-463) in 2000s. The adhesive fibrils are susceptible to sticking to each other, especially when the fibril radius is small, aspect ratio large and the elastic modulus low.

[0330] It was found that for a given ratio between the fibril center distance w and fibril radius R, the critical aspect ratio under which lateral collapse does not occur (the so-called static stability) is given by where h is the height f and v the elastic modulus and Poisson’s ratio of the adhesive fibrils, respectively, and is the work of adhesion (on the order of 300 J / m2) between adhesive fibrils. (Nature materials 2003, 2, 461-463; Progress in Materials Science 2021, 120, 100823) Results and Discussions

[0331] FIG. 41 (A)-(i) shows that, under R2G conditions, the normal stress is concentrated at the edge of the single fibril adhesive. When the large single fibril is divided into smaller fibrils of the radius within the ‘work window’ (FIG. 36(A)), the normal stress across the contact interface is more uniform (FIG. 41 (A)-(ii)), resulting in increased R2G load capability (FIG. 5(D)). FIG. 41 (A)-(ii) shows the fibrils, upon loading, detach simultaneously and share the loading equally among themselves, resulting in a slow decay of R2G adhesion strength with increasing apparent contact area (FIG. 5(E)).

[0332] On the other hand, when heated to the rubber-phases, FIG. 41 (A)-(iii) and 41(A)-(iv) show that the normal stress is concentrated at the edge for both the large single fibril and the fibril array, resulting in low adhesion strength and switchability values of ~20 (FIG. 5(F)). Controlled buckling of the fibrils further increases switchability by two orders of magnitude (>1000). Moreover, in the rubber-phase, the SMP is soft, leading to large deformation in the backing and fibrils (FIG. 41(A)-(iv)), and failure by progressive detachment of the fibrils from the periphery inwards. As a result, the rubbery adhesion strength weakens rapidly with increasing apparent contact area (FIG. 5(E)).

[0333] Bundling of adhesive fibrils during operation and adhesion repeatability affect the performance of these adhesive devices (Nature 2000, 405, 681-685). To evaluate the mechanical stability and adhesion durability of E44-SMP fibrillar adhesives, we carried out repeated contact and detachment tests of E44-SMP fibril arrays (FIG. 41 (B)-(i), R = 2.5 mm, h / R = 4) in the rubberphase (at 90 °C). The adhesion force exhibits good repeatability and durability with only slight degradation (<10%, FIG. 41 (B)-(ii)) and no bundling, fracture, or wear in the fibrils after 100 cycles (FIG. 41 (B)-(iii)).

[0334] According to Eq. (45), larger fibrils can have bigger aspect ratios, which has the advantage of larger contact compliance to adapt to surface textures and a smaller preload to actuate buckling for detachment. FIG. 42 shows the critical aspect ratio values for static stability of fibril arrays under a given center distance among fibrils as w / R = 4. When the fibril radius is in the order of microns, the fibril aspect ratio should remain below 10 or lateral collapse will occur. In contrast, the critical aspect ratio can exceed 10 with a fibril radius in the order of tens of microns and 30 with a fibril radius in the order of hundreds of microns.

[0335] Even with aspect ratios of 10 and 20, the SMP fibril arrays demonstrated good stability and repeatability when compressed to buckling during repeated tests. FIG. 41 (C) has already shown the mechanical stability and repeatability of the E44-SMP fibril arrays with a millimeter- sized radius (R = 2.5 mm) and a small aspect ratio (h / R =5). To inspect the static and dynamic stability (here dynamic stability refers to the stability after repeated contact and detach tests) of the E44-SMP fibril arrays with large aspect ratio values compressed to buckling, samples (R = 250 jizm, h / R = 10, 20, 30) were fabricated and they all showed good static stability as fabricated (FIG. 42(B)-(i) , 42(B)-(iii), 42(B)-(v)). Repeated contact and detachment tests with a large compression displacement of 1 mm (under which all the adhesive fibrils are buckled) were conducted in the soft, rubber-phase at a test temperature of 90 °C. After 100 test cycles, the adhesive fibrils did not stick to each other and show little observable fracture or wear for samples with aspect ratios of 10 and 20 (FIG. 42(B)-(ii), 42(B)-(iv)). Under an aspect ratio (h / R = 30) close to the static stability limit, the fibrils stick to each other after 100 cycled tests (FIG. 42(B)-(vi)).

[0336] Example 18. Application of Adhesive Fibrils

[0337] Soft Gripper

[0338] We demonstrate a soft gripper based on a R2G smart adhesive device. A millimeter-sized adhesive fibril (R = 2.5 mm, h / R = 4, FIG. 43(A)-(i)) is fabricated via molding (as described in Examples above), and used for the gripping, moving and placing of various objects, with a heating gun as the heating source, as shown in FIG. 44. These results prove strong adhesion and on-demand switchability of the R2G adhesive fibril.

[0339] As shown in FIG. 44(A), the gripping of a wood cube (60 mm x 60 mm x 60 mm, 64 g) was demonstrated using the SMP adhesive fibril. Gripping process includes: (i) heating the SMP adhesive fibril; (ii) contacting with the wood cube (preload = 254.6 kPa); (iii) cooling the SMP adhesive fibril to the glass-phase; and (iv) gripping the wood cube. Release process includes: (i) heating the SMP adhesive fibril; (ii) shape recovery of the SMP adhesive fibril; (iii) releasing the wood cube: and (iv) wood cube released.

[0340] Reusable Super Glue

[0341] To demonstrate the capabilities of the SMP fibrillar adhesives, we fabricated SMP fibril arrays (R - 2.5 mm, h / R - 4) arranged in a hexagonal structure (see “Comparison of the Single Fibril and Fibril Arrays”), and integrated them on a steel plate with a hook (FIG. 43(B) .

[0342] The adhesive hook utilizing SMP fibrillar adhesive as a superglue can be removed easily with the assistance of only a hair dryer. Due to the space in between the SMP adhesive fibrils, the SMP fibril arrays can be heated quickly as compared to the large single fibril adhesive with a comparable apparent contact area. To demonstrate this, a thermal imaging system (FLUKE, Ti200) is used to monitor the temperature of the SMP fibril arrays and the single fibril adhesive with a comparable apparent area when they are heated by a hair dryer, as illustrated in FIG. 45. Small holes are opened on the acrylic adherend, which facilitates exposure of the adhesive tip adhering to the adherend and direct monitoring of its temperature.

[0343] Results and Discussions

[0344] The demonstrated soft gripper showcases the adhesive fibril’s strong adhesion and on- demand switchability. With R2G adhesion, a single SMP adhesive fibril can hold glass bottles of 100-1000 ml filled with water, with weights up to 1.66 kg (FIG. 44(B)). Especially, the SMP adhesive exhibits excellent adaptability to surface roughness, allowing to firmly hold objects with diverse surface textures (FIG. 43(A)).

[0345] FIG. 43(B) demonstrates the load capacity of the device with a single fibril of 19.6 mm2cross- sectional area compared to that with 37 fibrils of effective contact area of 1091 mm2, supporting loads from 0.3 kg (FIG. 43(B)-(i)) to a box of 40 kg (FIG. 43(B)-(vi) ), and to a weight of 60 kg (-2307 times the weight of the SMP adhesive, FIG. 43(B)-(vii)).

[0346] Moreover, the SMP fibril arrays can be heated fast and detached easily, as depicted in FIG. 43(C) . When heated with a hair dryer, the time required for the temperature of the fibril arrays to reach 60 °C was less than one minute (FIG. 45 and FIG. 46). Consequently, the SMP fibril array adhesive device can be easily detached upon heating with a hair dryer after use, as demonstrated in FIG. 43(C) .

[0347] FIG. 46(A) shows the temperature evolution at the center of the central fibril of the SMP fibril arrays in contact with the acrylic adherend as compared to that at the center of the large single fibril adhesive when heated by a hair dryer laterally. It can be seen that the SMP fibril arrays are heated much faster than the SMP single fibril adhesive with the comparable apparent contact area. For the fibril arrays, it takes less than one minute to heat the fibril tip to 60 °C, even with an apparent contact area of 1194.8 mm2. However, this process takes more than 30 minutes for the single fibril with a comparable apparent contact area of 1017.9 mm2, as shown in FIG. 46(B).

[0348] Example 19. Summary and Comparison of the Performance of Different Fibrillar Adhesive Designs

[0349] The scaling limit at fibril (critical fibril radius when adhesion strength is 1 MPa) and array levels (effective array adhesion when the apparent contact area is scaled up to 1000 mm2), adhesion switchability when the adhesion strength is on the order of 100 kPa, the surface roughness adaptability (RMS roughness value of the adherend surface when adhesion strength is reduced by 50%) and the adhesion durability of the SMP R2G fibrillar adhesive are comprehensively compared with those of the gecko fibrillar adhesives (Biosurface and Biotribology 2015, 7, 42-49; Proc. Natl. Acad. Sci., 2004, 707(21), 7851 ; Biological journal of the Linnean Society 1996, 59, 21-35; Acta biomaterialia 2007, 3, 607-610; Journal of Experimental Biology 2006, 209, 3558-3568; MRS bulletin 2007, 32, 473-478) and bioinspired fibrillar adhesives with flat (ACS nano 2018, 72, 1273-1284; Nature materials 2003, 2, 461- 463; Suthisomboon, T. et al. in 2027 IEEE 4th International Conference on Soft Robotics (RoboSoft). 543-546 (IEEE); Beilstein Journal of Nanotechnology 2019, 70, 1-8), mushroom (Journal of Applied Mechanics 2022, 89, doi: 10.1115 / 1.4054628; Acta Biomaterialia 2012, 8, 282-288; Journal of The Royal Society Interface 2013, 70, 20130620), gradient (Royal Society Open Science 2017, 4, 161105; ACSAppI Mater Interfaces 2022, 74, 46239-46251 ; Advanced materials 2017, 29, 1701353), gradient-mushroom (Royal Society Open Science 2017, 4, 161105; ACS Appl Mater Interfaces 2022, 74, 46239-46251 ; Advanced materials 2017, 29,

[0350] 1701353) designs, as shown in Table 6. From Table 6 we can see that the SMP R2G fibrillar adhesives exhibit superior performance over other designs in all five aspects.

[0351] References cited in Table 6 are:1FIG. 5(C);2FIG. 32;3Proc. Natl. Acad. Sci., 2004, 707(21), 7851-7856;4FIG. 5(E);5Biosurface and Biotribology 2015, 7, 42-49;6FIG. 5(D) and 5(F);7Data presented in Examples above;8Beilstein Journal of Nanotechnology 2019, 70, 1-8;9Acta Biomaterialia 2012, 8, 282-288; Journal of The Royal Society Interface 2013, 70, 20130620;10Royal Society Open Science 2017, 4, 161105; ; ACS Appl Mater Interfaces 2022, 74, 46239-46251 ;11Nature 2000, 405, 681-685; Acta biomaterialia 2007, 3, 607-610; Journal of Experimental Biology 2006, 209, 3558-3568; MRS bulletin 2007, 32, 473-478;12FIG. 41(B);

[0352] 13ACS nano 2018, 72, 1273-1284; Nature materials 2003, 2, 461-463; Suthisomboon, T. et al. in 2027 IEEE 4th International Conference on Soft Robotics (RoboSoft). 543-546 (IEEE);

[0353] 14Journal of Applied Mechanics 2022, 89, doi:10.1115 / 1.4054628;15ACS nano 2018, 72, 1273-1284;16Advanced materials 2017, 29, 1701353;17MRS bulletin 2007, 32, 473-478.

[0354] Example 20. Structure, Materials and Operating Principles of Adhesive Skin System

[0355] Synthesis of the E-44 SMP

[0356] By referring to the protocols disclosed in the above Examples, the E44-SMP disclosed in this Example was prepared through the mixing of the liquid crosslinker, Poly(propyleneglycol)bis(2-aminopropylether), with the liquid E44 monomer at a mass ratio of 81 :48. aminopropylether), with the liquid E44 monomer at a mass ratio of 81 :48. Subsequently, this mixture underwent a 30-minute degassing process in a vacuum chamber. Following degassing, the liquid compound was poured into molds and subjected to a two-hour precuring phase at 50°C. This was succeeded by one hour of curing at 100°C, and a subsequent post curing period lasting an hour at 130°C. Modulus characterization of the E44 SMP

[0357] By referring to the protocols disclosed in the above Examples, Dynamic Mechanical Analysis (DMA) was utilized to assess the storage modulus (E) of the SMP materials disclosed in this Example across a range of temperatures, employing a measuring frequency of 0.5 Hz and a temperature ramp rate of 2°C / min. To create the rectangular test specimens, each measuring 30 mm in length, 6 mm in width, and 2 mm in thickness, an initial 2 mm thick SMP membrane was formed within the mold. Subsequently, these specimens were precision-cut into the desired geometry using a laser cutter.

[0358] Fabrication of the Adhesive Skin Patches

[0359] FIG. 47 illustrates the fabrication process of the adhesive skin patches. Initially, a rigid mold with pillar arrays was produced using clear resin through a Form 3+ SLA 3D printer (FORMUKBS INC.). Next, liquid Ecoflex precursors were prepared by blending Ecoflex 00-30 (Smooth-On, Inc.) Part A and Part B in a 1:1 mass ratio in a beaker, which was then degassed for 3 minutes. The degassed liquid Ecoflex precursors were poured into the rigid mold and degassed for an additional 5 minutes. Following a 2-hour curing period at room temperature, the soft Ecoflex mold containing pillar holes was obtained. Subsequently, the liquid adhesive precursors were prepared and cast into the flexible Ecoflex mold, with the flexible heater positioned on top and submerged into the liquid adhesive precursors. After another degassing step, the adhesive liquid precursors were allowed to fully cure. Finally, the adhesive skin patch was obtained, complete with top fibrillar structures and embedded flexible heaters. The inplane dimensions of the rigid molds and the resulting adhesive skin patches were tailored to accommodate various regions of the robotic hand, including the fingers, palm center, hypothenar, and thenar regions.

[0360] Results and Discussions

[0361] The fully integrated adhesive skin system is shown in FIG. 48(A)-(i), featured by adhesive skin patches (FIG. 48(A)-(ii)) with tailored geometries to accommodate the shape of different regions ona robotic hand. Each adhesive skin patch features a surface adorned with adhesive fibril arrays (FIG. 48(A)-(iii)) mimicking the adhesive skin appendages found in nature (e.g., geckos and tree frogs). These arrays are structured as hexagonal patterns, with the adhesive fibrils resembling hexagonal pillars(edge dimensions of a - 1 mm, fibril height of h - 1 mm, and a center-to-center distance of d = 2.2 mm). Embedded within the adhesive top layer and interlayer are flexible heaters that regulate the phase changes of the adhesive fibrils.

[0362] The skin patch’s fabrication involves molding on micro-structured molds using liquid adhesive precursors, with the flexible heaters embedded into the adhesive precursors and subsequently cured together. For applications related to sensing and detection, sensors like pressure sensors can be positioned on the rear side of the adhesive skin patches (FIG. 48(A)-(iv)) and then affixed to the robotic hand through methods such as gluing or mechanical fixation.

[0363] The key to achieving widely tunable adhesion lies in selecting the appropriate material system for the adhesive. The above examples have demonstrated that the E44 epoxy shape memory polymer (E44-SMP) with dramatically tunable modulus, temporary shape-locking and permeant shape-memory effects, is a promising candidate. However, E44-SMP’s actuation temperature mentioned above was relatively high (~90 °C), which could potentially cause thermal damage to certain objects like fabrics or plasticware. In this example, the monomer- to-curing agent ratio was adjusted to 81 :48 to lower the glass transition temperature, making it more compatible with a wider range of objects.

[0364] FIG. 49 shows the storage modulus and loss factor of the SMP at different temperature levels. The peak observed in the loss factor curves signifies the glass transition temperature of the E-44 SMP material at 36°C. Dynamic Mechanical Analysis (DMA) tests (FIG. 48(B)-(i)) reveal that the modified E44-SMP is stiff (E > 1 GPa) in the glassy phase at room temperatures (T < 30°C) and becomes soft (E < 1 MPa) when heated to the rubbery phase (T ~ 51°C). After the transition from the rubbery phase to the glassy phase, the material retains its deformed shape due to the shape-locking effect. Upon reheating to the soft rubbery phase, the E44- SMP regains its original shape, demonstrating the shape-memory effect.

[0365] The on-demand attachment and detachment capabilities and large adhesion tunability of the adhesive skin replies on the tunable materials properties of the E44-SMP (FIG. 48(B)-(ii)). While contacting with the objects under the glassy phase, the stiff SMP adhesive fibril cannot form good attachment with the adherend, and the adhesion strength is minimal (FIG. 48(B)- (ii)-(1 )). While heated to the soft rubbery phase, the adhesive fibrils are able to form conformal contact with the adherend (FIG. 48(B)-(ii)-(2)). When transitioning into the glassy state, the conformal contact will be locked and the stiffness of the SMP adhesive fibril will lead to a DMT- like adhesion regime (as shown in above Examples) with the interfacial stress reaching the theoretical value uniformly, rendering the so-called strong R2G (rubbery-to-glassy) adhesion (as shown in above Examples) for secure grasping. When transitioning back to the rubbery phase again, the compliance of the SMP adhesive fibril will lead to a JKR-like adhesion regime (as shown in above Examples) with the interfacial stress concentrated at the edge of the contacting interface, causing crack-propagation detachment behaviors and weak rubbery adhesion for easy release. Due to the tunable material properties of E44 SMP, the adhesive skin exhibits minimal adhesion strength in the stiff glassy phase, weak adhesion in the rubbery phase, and high adhesion strength in the R2G conditions as stated above. This broad range of tunable adhesive properties allows the SMP adhesive skins to adapt to various environmental interaction conditions (see FIG. 48(C)). Whether adhesion might hinder the process, such as in cases of dexterous manipulation, surface condition detection, or handling lightweight objects, or whether adhesion would be highly advantageous, such as in on-demand grasping and releasing of fragile, large, or heavy objects, these adhesive skins are expected to effectively accommodate the specific requirements.

[0366] Example 21. Adhesion Characterisation and Optimisation of the Adhesive Skin

[0367] To access the adhesion performance of the versatile robotic adhesive skin, the SMP adhesive fibrils (hexagonal pillars, edge length a = 1 mm with various aspect ratios) were fabricated (see more details in “Fabrication of the Rough SMP Adhesive Fibrils”) and their glassy adhesion (contact and detach in the glassy phase), rubbery adhesion (contact and detach in the rubbery phase) and R2G adhesion (contact in the rubbery phase and detach in the glassy phase)were characterized experimentally through systematic pull off and shearing tests under the normal and shear modes, receptivity (see more details in “Adhesion Test of the SMP Adhesive Fibrils").

[0368] Fabrication of the Rough SMP Adhesive Fibrils

[0369] To further reduce the minimal adhesion in the glassy phase, and the weak adhesion in the rubbery phase, surface roughness is introduced on the top of the SMP adhesive fibril. FIG. 50(A) illustrates the fabrication process of the rough SMP adhesive fibrils using the replica molding method. First, a stiff fibril mold was prepared using 3D printing, and a sandpaper template was glued on top of the fibril mold (FIG. 50(A)-(i)). Then, liquid Ecoflex precursors mixed in the baker and then fully degassed were poured into the stiff mold (FIG. 50(A)-(ii)). After curing at room temperature for 2 hours, the soft Ecoflex mold was obtained. Then, SMP liquid precursors, also mixed in the baker and then fully degassed, were introduced into the soft Ecoflex mold (FIG. 50(A)-(iii)), and then degassed in the vacuum chamber for 30 mins. Finally, an SMP rough fibril replicating the surface texture of the sandpaper template was obtained (FIG. 50(A)-(iv)) after curing in the oven (see more details in previous Examples).

[0370] FIG. 50(B) shows the photo of an SMP adhesive fibril fabricated without using the sandpaper template, and the corresponding surface profile of the SMP adhesive fibril tip. It shows a surface roughness value RMS (Root-mean-square amplitude) of 0.455 urn (measured using TalyScan 150 3-D Scanner). FIG. 50(C) shows the 3D morphologies of the SMP fibrils replicated from sandpapers with varying grit sizes of 10O-grit (FIG. 50(C)-(i)), 120-grit (FIG. 50(C)-(ii)), 240-grit (FIG. 50(C)-(iii)), and 800-grit (FIG. 50(C)-(iv)), as measured using the Keyence VK-X1100 Optical Profilometer. The corresponding RMS values of the fibril tip are 46.96 29.97 16.25 pm, and 10.04 pm, respectively.

[0371] Adhesion Test of the SMP Adhesive Fibrils

[0372] To verify the broad tunability of adhesion in the versatile robotic adhesive skin, the adhesive fibrils (hexagonal pillar, edge length = 1 mm) were fabricated as mentioned above, and the normal and shear adhesion strength under the glassy, rubbery, and R2G conditions were measured experimentally. For the glassy adhesion test, the temperature was kept at the room temperature (25 °C) all the time and the SMP was in the glassy phase during both the contacting and detachment (by pulling / sharing) stages. For the rubbery adhesion test, the temperature was kept above the Tg(60 °C) all the time and the SMP was in the rubbery phase during both the contacting and detachment stages. For the R2G adhesion test, the temperature was 60 °C during the contacting stage so that the SMP was in the rubbery phase, then was adjusted to the room temperature (25 °C) to bring the SMP to the glassy phase before detachment.

[0373] Pull off tests were used to measure the normal adhesion performance. The details of the normal adhesion testing setup and procedures were similar to those outlined in the disclosed Examples above, where the customized adhesion test platform was constructed on a universal tensile machine (Instron 5566), and a hot plate was used to control the temperature. The SMP adhesive fibrils were glued to a glass indenter and clamped onto an Instron Universal tensile machine (Instron 5566) for the adhesion test. The adhesive samples were brought into contact with the glass adherend surface at a very low speed of 10 pm / s and then pulled off at a fixed speed of 10 mm / min by the linear stage in displacement control, with the inline load cell measuring the force throughout the process.

[0374] Shearing tests were used to measure the shear adhesion performance. The experimental setup is given in FIG. 51(A) and 51 (B), with a stationary glass fixed on top of the heating plate. The SMP adhesive fibrils were glued to a glass indenter and connected to the vertical load cell through a clamp. The vertical load cell connected to the clamp was used to bring the SMP adhesive fibrils into contact with the sliding glass placed on the stationary glass at a very low speed of 10 to the predefined preload. After contacting, the horizontal load cell pulled the sliding glass laterally at a speed of 10 mm / min through a connecting string. The measured total force by the horizontal load cell was composed of the shear adhesion force between the SMP adhesive fibril and the sliding glass, and the friction force Function between the sliding glass and stationary glass (FIG. 51 (C)-(i)). To calibrate Ffnction, another shearing test with a dead load from the weight placed on top of the sliding glass was conducted (FIG. 51(C)- (ii)) and the measured force denoted as is shown in FIG. 51(D), indicating a friction coefficient of 0.121 between the sliding glass and stationary glass. Finally, the shear adhesion force can be obtained as

[0375] Results and Discussions

[0376] FIG. 52(A) shows the influence of the preload on the adhesion strength of the SMP adhesive fibril in the normal (FIG. 52(A)-(i)) and shear (FIG. 52(A)-(ii)) modes. Under both the normal and shear modes, both the R2G adhesion and rubbery adhesion get saturated under a very small pressure 38.5 kPa due to the compliance of the SMP in the rubbery phase during the contact process. The R2G adhesion can be as strong as 2.0 MPa in the normal mode and 5.1 MPa in the shear mode, while the rubbery adhesionvaluesare164.4kPain the normal mode and 206.9 kPa in the shear mode. The ratio between the high adhesion strength (i.e., R2G adhesion) and the preload can be up to 90 and 50 folds under a preload of 19.2 and 38.5 kPa, respectively, outperforming most dry adhesives (with a ratio up to 1) and even the state-of- the-art design (Advanced Functional Materials 2021, 31, 2107023) of around 6 folds. This enables the reliable grasping of various objects, including these fragile ones such as eggs and shells, under a gentle pressure. In contrast, the glassy adhesion saturates at a higher preload of around 192 kPa due to the stiffness of the SMP in the glassy phase during contacting. The saturated glassy adhesion is 5.7 kPa in the normal mode and 20.5 kPa in the shear mode. In the shear mode, the shear force increases linearly after the shear adhesion is saturated as the preload further increases due the contribution of the friction. The measured friction coefficient is around 0.3803, much smaller than 1. If frictional method is used to grip an object, it will require a very large preload and cause significant deformation to the object, or sliding will occur due to limited gripping force under a give preload. However, the minimal adhesion strength will be of great importance for scenarios where adhesion is not desired.

[0377] The fibril aspect ratio also has a huge influence on the adhesion performance of the adhesive fibrils. To select a proper aspect ratio for the adhesive fibrils, adhesion tests were conducted, and the measured results are shown in FIG. 52(B). It was shown that under the normal mode (FIG. 52(B)-(i)), the R2G adhesion is not influenced by the fibril aspect ratio, and the rubbery adhesion and glassy adhesion strength reduces as the fibril aspect ratio increases. Under the shear mode (FIG. 52(B)-(ii)), when the aspect ratio is smaller than 1 , the deformation is block shearing dominant and the adhesion strength shows limited dependence on the fibril aspect ratio, while aspect ratio exceeds 1 , the adhesion strength decreases quickly as the aspect ratio increases due to a beam bending deformation. In short, a smaller aspect leading to a stronger R2G adhesion to support heavy load, but at the cost of reduced fibril contacting compliance to adapt to different surface textures. Therefore, an aspect ratio of 1 where the R2G adhesion can be maintained is selected and used for future discussion and applications.

[0378] Pull off and shearing tests were further conducted to exam the R2G, rubbery and glassy adhesion of the SMP adhesive fibril on various adherends other than the glass, including steel, acrylic and Polypropylene (PP). It has been proven in above Examples that the SMP can adapt to various surface materials and conditions due to the shape locking effect. The measured results (FIG. 52(C)) show that on all the adherends examined here, switching the SMP to the rubbery phase leading to a low rubbery adhesion, which is one order of magnitude smaller than R2G adhesion. In contrast, the glassy adhesion can be further reduced by more than one order of magnitude further on the basis of rubbery adhesion.

[0379] To further reduce weak rubbery adhesion and low glassy adhesion, surface roughness can be introduced on the tip of the SMP fibrils. It was shown that the R2G adhesion is not influenced by the surface roughness, thanks to the shape-locking of the conformal contacting during the rubbery-to-glassy transition process (as disclosed in the above Examples). In contrast, the rubbery adhesion and glassy adhesion reduce quickly as the surface roughness increases. When the surface roughness value (root-mean-square amplitude -RMS) of the SMP adhesive fibril increases from 0.455 pm to 46.96 pm, the rubbery adhesion is reduced from 164.4 kPa to 4.3 kPa, and from 206.9 kPa to 41.7 kPa in the normal and shear mode, respectively; and the glassy adhesion is reduced from 5.7 kPa to 0.65kPa, and from 20.5kPa to 2.86kPa in the normal and shear mode, respectively (FIG. 52(D) ). The further reduced rubbery adhesion enables larger adhesion switchability from R2G to rubbery conditions for on- demand gripping and release, and reduced minimal glassy adhesion enables broader utility in non-adhesive conditions.

[0380] In conventional designs of smart adhesives made of soft elastomers, the adhesion forms whenever contact forms. Intelligent materials own tunable modulus can alleviate this problem. However, the minimum adhesion of the smart adhesives made of intelligent materials increases as the maximum adhesion strength increases (as shown in FIG. 52(E) and 52(F), see more details in FIG. 53, Table 7 and Table 8). Table 7. Summary of the adhesion performance of smart adhesives made of typical intelligent materials and conventional elastomers measured on glass substrates under normal loading.

[0381] Table 7 summarizes the reported normal adhesion performance of the smart adhesives made of typical intelligent materials where the adhesion is mediated through controlled switching of the materials’ properties, encompassing SMPs (Small, 2019, 15(50), 1904248; Macromolecular Rapid Communications 2022, 43, 2200012; Langmuir, 2021 , 37(45), 13420- 13429; J. Microelectromech. Syst., 2014, 23(5), 1012-1014; Advanced Materials Interfaces 2019, 6, 1801542), ionogels (Soft Matter 2022, 18, 5934-5938; Advanced Intelligent Systems, 2023, 2300127), hydrogels (Nature Communications 2022, 13, 6964; Science Robotics 2021, 6, eabe1858), Magneto-rheological Fluid (MRF) / Polydimethylsiloxane (PDMS) Composites (Science Advances 2023, 9, eadf4051), Liquid Metal (LM) / PDMS Composites (Extreme Mechanics Letters 2021, 49, 101488; Advanced Materials 2016, 28, 5088-5092)), and the smart fibrillar adhesives inspired by geckos made of conventional elastomers where the adhesion strength is adjusted by controlled mechanical loadings, including the PDMS fibrils controlled by peeling (Proceedings of the National Academy of Sciences 2017, 114, E4344- E4353; ACS Applied Materials & Interfaces 2020, 12, 27717-27725; Nanoscale 2013, 5, 11876-11884; ACS applied materials & interfaces 2014, 6, 19247-19253; ACS applied materials & interfaces 2014, 6, 1345-1350; Advanced Materials Interfaces 2019, 6, 1900875; / ACS Applied Materials & Interfaces 2022, 14, 31448-31454), PDMS fibrils controlled by buckling (Journal of the Mechanics and Physics of Solids 2019, 123, 20-27; ACS Applied Materials & Interfaces 2015, 7, 24127-24135), PDMS fibrils controlled by wrinkling (Langmuir 2010, 26, 2223-2226; International Journal of Precision Engineering and Manufacturing 2017,

[0382] 18, 133-137), PDMS fibrils controlled by shearing (ACS Applied Materials & Interfaces 2016, 8, 10029-10037), carbon nanotubes (CNT) fibrils controlled by shearing (Nano Letters 2019,

[0383] 19, 6756-6764; The Journal of Adhesion 2012, 88, 243-252; ACS applied materials & interfaces 2023, 15, 17029-17044; RSC Advances 2014, 4, 9056-9060; Science 2008, 322, 238-242), and vinylpolysiloxane (VPS) fibrils controlled by buckling (Advanced Functional Materials 2023, 33, 2209905).

[0384] Table 8 summarizes the reported shear adhesion performance of the stimulus-responsive smart adhesives made of typical intelligent materials, encompassing SMPs (Extreme Mechanics Letters 2016, 9, 207-214), ionogels (Soft Matter 2022, 18, 5934-5938; ACS Applied Polymer Materials 2023, 5, 3398-3404), hydrogels (Journal of Colloid and Interface Science 2021, 582, 187-200; ACS nano 2023, 17, 13948-13960; Proceedings of the National Academy of Sciences 2022, 119, e2203074119; ACS Applied Materials & Interfaces 2021, 13, 36574- 36586), and copolymers (Polymer Chemistry 2017, 8, 7207-7216; Angewandte Chemie International Edition 2015, 54, 13063-13067; Polymer 2015, 68, 270-278; Polymer Chemistry 2018, 8, 6300-6308; ACS Macro Letters 2019, 8, 968-972; Nature communications 2016, 7, 12094), and the mechanical-loading-adjustable smart fibrillar adhesives inspired by geckos made of conventional elastomers, including PDMS fibrils controlled by peeling (ACS applied materials & interfaces 2014, 6, 2213-2218; Langmuir 2023, 39, 3784-3791 ; Langmuir 2012, 28, 8746-8752; Journal of the Royal Society Interface 2015, 12, 20150090), PDMS fibrils controlled by wrinkling (Langmuir 2010, 26, 2223-2226), Polyurethane acrylate (PUA) fibrils controlled by peeling (Proceedings of the National Academy of Sciences 2009, 106, 5639- 5644; Advanced Functional Materials 2014, 24, 574-579; Nanotechnology 2015, 26, 415301; T.-i. Kim, H. E. Jeong, K. Y. Suh, H. H. Lee, Stooped nanohairs: Geometry-controllable, unidirectional, reversible, and robust gecko-like dry adhesive. (2009); Nano Today 2009, 4, 385-392; Soft Matter 2014, 10, 3301-3310; Langmuir 2012, 28, 5737-5742; ACS nano 2018, 12, 1273-1284), and CNT fibrils controlled by peeling (RSC Advances 2014, 4, 9056-9060; Science 2008, 322, 238-242; Proceedings of the National Academy of Sciences 2007, 104, 10792-10795; Advanced Materials 2014, 26, 1456-1461).

[0385] Table 8. Summary of the adhesion performance of smart adhesives made of typical intelligent materials and conventional elastomers measured on glass substrates under shear loading. FIG. 53(A) and FIG. 53(B) summarize the maximum versus the minimum adhesion strength of the smart adhesives made of typical intelligent materials under the normal (FIG. 53(A)) and shear (FIG. 53(B)) modes, respectively. Notably, it is evident that as the maximum adhesion strength increases, the minimum adhesion strength also increases. This observation conforms to the following scaling laws:

[0386] That implies that it is difficult to maintain a low minimum adhesion strength value under a large maximum adhesion strength for the smart adhesive design, leading to a critical challenge to use the smart adhesives in carrying heavy loads and in scenarios where adhesion should be avoided simultaneously.

[0387] Therefore, these smart adhesives cannot be used to support heave load and used in nonadhesive conditions at the same time. In contrast, as shown before, the shape-locking effect of the SMP enables high R2G adhesion on various surfaces, and switching the SMP to the rubbery phase leads to an adhesion reduction by 1-2 order of magnitude (hexagons in FIG. 52(E) and 52(F)). While contacting in the glassy phase, the adhesion strength can be further reduced by another 1-2 order of magnitude (stars in FIG. 52(E) and 52(F) ). This enables a huge increase of the range of the adhesion strength a smart adhesive can achieve simultaneously. FIG. 52(G) and 52(H) further compare the adhesion strength vs. the adhesion switchability of the existing smart adhesives made of soft elastomers or intelligent materials (see more details in Table 7 and Table 8 above). It is shown that even the adhesion switching capability of this design from the R2G condition to the rubbery phase can outperform other designs. And the further reduction of the adhesion in the glassy phase makes the proposed design more extraordinary.

[0388] Example 22. Analysis of the On-Demand Attachment

[0389] One extraordinary aspect of the adhesive skin is the unprecedentedly broad adhesion range it can cover (from 0.65 kPa to2 MPa in the normal mode, and from 2.86 kPa to 5.1 MPa in the shear mode) as given in FIG. 52(E) and 52(F), enabled by the on-demand attachment capability of the SMP adhesive fibrils (FIG. 48(B)-(ii)-(1) and 48(B)-(ii)-(2)). Here, more extensive analyses are provided to shed further light on the mechanisms underlying this on- demand attachment capability. To provide strong adhesion, a prerequisiteis to from macroscopic contact between the adhesive and the adhered. However, adhesion occurs once contact forms in most smart adhesive designs, since these designs utilize soft elastomerswith low modulus values. In contrast, the reported adhesive skins utilize SMP as the adhesive material, which has a dramatically tunable modulus spanning 3 orders of magnitude, enabling on-demand attachment capabilities, as shown in FIG. 48(B) and FIG. 54(A)-(i). When contacting in the glassy phase, the glassy SMP lead to non-sticky contact (The journal of adhesion 2021, 97, 509-527; Proceedings of the National Academy of Sciences 2014, 111, 3298-3303), i.e., very small actual contact area, with the adherend due to the stiffness (modulus ~ 2 GPa at room temperature) of the SMP and the ubiquitous surface roughness. This non-sticky contact leads to the minimum adhesion for non-adhesive applications. While switched into the rubbery phase, the SMP become very soft with a modulus of 1 MPa, enabling sticky contact (i.e., large contact area) and facilitating adhesion on typical adherends, according to the Dalquist Criterion.

[0390] Typical Adherend Surfaces Profiles

[0391] Typical adherends were prepared to evaluate the stickiness of the SMP adhesive fibril in both the glassy and rubbery phases. FIG. 55 illustrates the surface profiles of these adherends, as measured using the TalyScan 150 3-D Scanner (Taylor Hobson, Inc.). The root-mean-square amplitude values of these adherends were derived from the measured profiles and ranged from 0.0028

[0392] FEA Simulations of forced contact

[0393] 3D Finite Element Analyses (FEA) were conducted in ABAQUS to study the forced contact between an SMP fibril (hexagonal pillar, edge length = 5 mm, height = 10 mm) and a segregated steel plate (amplitude = 0.5 mm, wavelength = 1 mm) in the rubbery and glassy phases (FIG. 56).

[0394] To simulate the contact, a square die representing the compression tool was positioned at the top of the SMP fibril, and a segregated plate was placed at the bottom of the SMP fibril. The die and plate were assumed to be much stiffer than the SMP fibril, and no deformation of these parts was expected. The SMP fibril in this model was composed of a temperature-dependent shape memory polymer. For the SMP fibril in the rubbery phase, the One-Order Ogden- Roxburgh Model (Acta Mechanica Sinica 2017, 33, 543-554; Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences 1999, 455, 2861- 2877) was adopted to simulate the rubbery behavior during the contact process with modeling parameters as 0.1785 MPa, 2.7265, Di = 0.2256 The SMP in the glassy phase was modeled using an isotropic elastic model with Young’s modulus of 2 GPa and Poisson ratio of 0.35. The density of SMP bulk was set as ton / mm3for both glassy and rubbery phases.

[0395] The contact process was simulated through the enforcement of displacement boundary conditions. The bottom segregated plate was constrained to a fixed location. Both analysis cases were conducted using quasi-static explicit dynamic procedures. The displacement loading velocity was 1 mm / s. A general contact definition with a friction coefficient of 0.3 enforced contact interactions between all dies and the Eulerian materials in the analysis. The SMP fibril contact simulation was conducted in Abaqus / Explicit using coupled Eulerian- Lagrangian (CEL) analysis. In this analysis, SMP fibrils that undergone large deformations were meshed with Eulerian elements, while the die and plate in the model were meshed with Lagrangian elements. In both models, the Eulerian part was a rectangular prism measuring 15 x 15 x 11 mm3meshed with EC3D8R elements. The global element size was 0.1 mm. This mesh did not define the geometry of the SMP fibril, but the domain in which the SMP material could exist. The SMP fibril geometry was defined by assigning SMP material to a portion of this mesh corresponding to the shape of the SMP fibril. The Eulerian mesh needed to be large enough to contain the SMP material completely as it was deformed. The simulation of the SMP was invalid if the material flowed out of the mesh model. The segregated plate and the top die were modeled as discrete rigid with mesh seed of 0.03 and 0.4 respectively. It was worth noting that the Eulerian volume fraction output variable (EVF) was requested as field output to visualize geometric results. Then the images showing the contact pressure were obtained from the FEA calculation results and analyzed using Imaged software to get the contact area between the SMP fibril and the segregated substrate under different pressures.

[0396] Samples for Contact Test

[0397] To validate the on-demand attachment capabilities of the SMP and the FEA simulation results of the forced contact condition when the SMP switches from the glassy phase to the rubbery phase, contact tests were conducted experimentally, with the basic experimental setup illustrated in FIG. 56(A). The samples used in the contact test are shown in FIG. 57(A). The macroscopic SMP fibril (FIG. 57(A)-(i), hexagonal pillar, edge length = 5 mm, height = 10 mm) was fabricated using the molding method, the segregated steel plate (FIG. 57(A)-(ii), amplitude = 0.5 mm, wavelength = 1 mm) was prepared using machining, and the fluorescent nanoparticles (FIG. 57(A)-(iii)), which emit green fluorescence after UV illumination (FIG. 57(A)-(iv)), were used to observe the contact area directly. Coating the Substrate with Fluorescent Nanoparticles

[0398] Before the contact tests, the fluorescent nanoparticles were coated onto the segregated substrate through the following process: First, the segregated steel substrate was coated with a thin layer of PDMS (FIG. 57(B)-(i) and 57(B)-(ii)) to facilitate strong adhesion between the fluorescent nanoparticles and the substrate. Then, the fluorescent nanoparticles were sprinkled onto the substrate (FIG. 57(B)-(iii)) and the extra particles were blown away gently using an ear syringe (FIG. 57(B)-(iv)).

[0399] Contact Tests in Different Phases

[0400] The contact test process in the glassy phase is given in FIG. 57(C). First, the glassy SMP pillar was brought into contact with the segregated substrate coated with fluorescent nanoparticles, causing the nanoparticles to adhere to the SMP within the contact area (FIG. 57(C)-(i) and 57(C)-(ii)) . After contact and separation, the SMP with fluorescent nanoparticles in the contact regions was illuminated under 365 nm UV irradiation (FIG. 57(C)-(iii)). Then, the sample was then moved to a dark room, and a camera was used to capture the contact condition indicated by the fluorescence from the illuminated fluorescent nanoparticles (FIG. 57(C)-(iv)). The contact test in the rubbery phase is shown in FIG. 57(D), where the contact condition was locked due to the shape-locking effect when it contacted the segregated substrate in the rubbery phase and was subsequently cooled to the glassy phase (FIG. 57(D)-(ii)).

[0401] Results and Discussions

[0402] To further unravel the underlying mechanisms behind the on-demand switching from the non- sticky to sticky contact conditions of the SMP adhesive fibrils, the free contact condition (i.e., contact under zero-loading due to adhesive interactions)of between SMP adhesive fibril and common objects are examined using the stickiness parameter The journal of adhesion 2021,

[0403] 97, 509-527 jSjsthe maximum RMS roughness amplitude of the SMP adhesive fibril and the adherend and |Sthe adhesive interaction distance defines a characteristic adhesion length which can be used to quantify the relative strength of adhesion, where with and as the modulus and Poisson’s ratio of SMP adhesive fibril is the largest wavelength in the roughness spectrum with L as the largest lateral characteristic sale. For Van der Waals force (Journal of Adhesion Science and Technology 2014, 28, 226- 239), m and 10 mJ / m2. For the SMP adhesive fibril, 2 mm, leading to the The modulus values of the SMP are 1 MPa and 2 GPa, and Poisson’s ratios are 0.5 and 0.35, in the rubbery and glassy phases, respectively. When an adhesive is non-sticky our an adhered, i.e., macroscopic contact area cannot be formed, and the adhesion force is negligible under a zero-preload. When an adhesive is sticky on an adherend, i.e., reasonable large contact true area can form, facilitating measurable adhesion force when detached.

[0404] FIG. 54(A)-(ii) shows the stickiness parameters of the SMP adhesive fibril under various adherends. It is shown that in the glassy phase, on all the adherends, indicating non- sticky adhesion state. When switched to the rubbery phase, on most adherends with a surface roughness The only exception is the highly roughened wood block surface. However, conformal contact between the rubbery SMP and a highly roughened adherend can also be possible under a preload, as pointed in the above Examples. To demonstrate the on-demand attachment capability of the SMP from the glassy to the rubbery phase, the contact area between a large SMP sample edge length = 5 mm, height = 10 mm) and a highly roughened adherend (a segregated steel plate amplitude = 0.5 mm, wavelength = 1 mm) were further examined using Finite Element Analysis (FEA). FIG. 54(A)-(iii) shows calculated true contact area between the SMP fibril and the segregated adherend, normalized by the nominal or projection area In the glassy phase, only a small part of the SMP is in contact with the adherend, as shown by the simulated contact deformation of the SMP fibril through FEA (Inset 1 in FIG. 54(A)-(iii)) and the experimental observation of the fluoresce spanning very limited part on the bottom of the fibril indicating the contact area (Inset 2 in FIG. 54(A)-(iii)). Even under a high preload of 250 kPa, the true contact area is smaller than 10% of the nominal contact area. In contrast, while switched into the rubbery phase, the true contact area increases monotonically with the increase of the preload. When the preload exceeds 162 kPa, the true contact area becomes larger than the nominal contact area and enable full contact under a preload of 231 kPa, as indicated by the FEA result (Inset 3 in FIG. 54(A)-(iii) ) and the experimental observation of the fluoresce spanning all over the bottom of the fibril (Inset 4 in FIG. 54(A)-(iii).

[0405] In short, the dramatically tunable modulus of the SMP facilitates non-sticky attachment in the stiff glassy phase and sticky attachment in the rubbery phase, enabling on-demand attachment for applications in all adhesion spectrum scenarios, whether non-adhesion or adhesion is required. This is the key mechanisms leading to the broad adhesion range that the SMP adhesive fibril can achieve, as given in FIG. 52(E) and 52(F). Example 23. Deciphering the On-demand Adhesions Switching

[0406] Another outstanding characteristic of the adhesive skin is the large adhesion switchability (~102) under very high adhesion conditions (~ MPa) inaccessible to other smart adhesive designs, as shown in FIG. 52(G) and 52(H). This on-demand adhesion switching under strong adhesion conditions is enabled by the on-demand switching of the SMP from the R2G to rubbery conditions (FIG. 48(B)-(ii)-(3) and 48(B)-(ii)-(4)) after sticky attachment in the rubbery phase. In this section, the physics controlling the adhesion strength switching behind the materials phase changing is further explored.

[0407] Fundamental research disclosed in the above Examples on the adhesion between a spherical indenter and the SMP substrate has demonstrated that the adhesion states of the SMP adhesives can be switched between the strong R2G and weak rubbery conditions under normal loading. Here, comprehensive FEA modeling and simulations of the detachment of the selected adhesive fibril are conducted under the most commonly encountered loading conditions, including both normal and shear modes.

[0408] FEA Simulations of On-Demand Detachment

[0409] 3D FEA simulations (FIG. 58) were conducted in ABAQUS Explicit to study the mechanics of the switchable adhesion of the SMP adhesive fibril under the selected geometry (hexagonal pillar, edge length a = 1 mm, height h = 0.8 mm) after on-demand attachment, under both the (FIG. 58(A) to 58(C)) normal and the shear (FIG. 58(D) to 58(F)) modes.

[0410] In the normal-mode cases, a pulling displacement in the vertical direction was specified on the top surface of the SMP fibril (FIG. 58(A)). In the shear-mode cases, a rigid plate was tied with the upper surface of the SMP adhesive fibril, on which the shear displacements were applied, as shown in FIG. 58(D). A layer of cohesive elements with zero thickness was tied to the bottom surface of the SMP adhesive fibril to model the adhesion. The bottom surfaces of the cohesive elements were fixed in the simulation. C3D8R (8-node linear brick, reduced integration with hourglass control) elements were used for the SMP adhesive fibril and COH3D8 (8-node three-dimensional) cohesive elements were used for the cohesive layer. The mesh size is 50 pm, and the change of the calculated adhesion force is smaller than 6.5% if the mesh size is reduced to 25 which leads to an increase in the number of elements by 3 times.

[0411] The SMP was regarded as a stiff elastic material with an elastic modulus of 2 GPa and Poisson’s ratio of 0.35 in the glassy phase, and a hyperelastic material described by a one- order Ogden model (JJ = 0.1785 MPa, a = 2.7265 D = 0.2256 MPa'1) in the rubbery phase, as mentioned in Examples above. The density of the SMP was 2x10-9ton / mm3. A linear tractionseparation law was used to model adhesive interactions. In the normal-mode cases, a maximum nominal stress criterion was used for damage initiation, and the normal adhesive nominal stress criterion was used for damage initiation and a power-law mixed mode behavior {Journal of adhesion science and technology 1995, 9, 119-147; Journal of composite materials 2003, 37, 1415-1438) was used for damage evolution. The shear adhesive parameters were and theoretical adhesion strength in the normal, first and second shear directions, respectively, and e are the critical energy release rate (or work of adhesion) of Mode-1 and Mode- II cracks, respectively

[0412] Results and Discussions

[0413] FIG. 54(B)-(i) illustrates the detachment behavior of the SMP adhesive fibril under normal mode, and FIG. 54(B)-(ii) and 54(B)-(iii) give the stress distribution at the contact interface at the detachment point.

[0414] Four cases were explored in FEA simulations, including the normal / shear modes in the rubbery / glassy phases. In the normal mode, a normal displacement was applied on the top surface of the fibril, until the detachment of the SMP adhesive fibril occurred (FIG. 58(A)). In the shear mode, a shear displacement was applied on the rigid plate until the detachment of the SMP adhesive fibril occurred (FIG. 58(D)). FIG. 58(B) shows the force-displacement curves of the SMP adhesive fibril under the normal mode and FIG. 58(E) under the shear mode. It can be seen from FIG. 58(B) and 58(E) that the adhesion can be switched from the high, R2G condition to the weak, rubbery phase. FIG. 58(C) further shows the detachment process of SMP fibril under the normal mode. It is shown that the SMP adhesive fibril detaches uniformly under the R2G condition (FIG. 58(C)-(i)) while showing a gradual crack-propagation detachment from the edge in the rubbery phase (FIG. 58(C)-(ii)). Fig. S9F further shows the detachment process of SMP fibril under the shear mode. It is shown that the SMP adhesive fibril detaches uniformly under the R2G condition with a shear-dominant behavior across the contact interface (FIG. 58(F)-(i)) but a gradual crack-opening dominant detachment from the edge in the rubbery phase (FIG. 58(F)-(ii) ).

[0415] It is shown that under the normal mode, the interfacial stress is dominant by the normal stress. Under the R2G condition, the normal stress is evenly distributed at the contact interface (FIG. 54(B)-(ii)), indicating a DMT-like or strength-limited strong adhesion state where detachment occurs when the interfacial normal stress reaches the theoretical limit everywhere. While transitioning into the rubbery phase, the SMP adhesive fibril become soft and the deformation thus the stress is concentrated at the edge part of the contact interface (FIG. 54(B)-(iii)). Therefore, the detachment happens in a crack-propagation form when the energy release rate value reaches the critical valueand the Griffith’s criterion is satisfied, leading to uneven loading sharing at the contact interface and JKR-like weak adhesion. FIG. 54(B)-(iv) depicts the detachment characteristics of the SMP adhesive fibril under shear mode, and FIG. 54(B)-(v) and 54(B)-(vi) show the stress distribution at the contact interface at the detachment point. Under the shear loading, the interfacial stress is shear dominant (FIG. 54(B)-(v)), leading to strong shear adhesion strength. While in the rubbery phase, the interfacial stress is dominant by the normal stress at the crack front or the left edge part due to the large deformation of the SMP adhesive there, leading to an opening-dominant detachment behaviour and a weak adhesion strength. This interesting behavior of opening-dominant failure under a shear loading has been reported elsewhere (Carbon 2021, 173, 410-418 (2021) in the multilayer nanosheets.

[0416] Example 24. Applications in Non-Sticky States

[0417] As pointed in one of the Examples above, adhesion is not desired in many scenarios. Existing designs of smart adhesives, however, are unable to be used in both adhesive and non- adhesive conditions. The proposed smart adhesive skins, in contrast, owns a broad range of tunable adhesion, can adapt to various application needs. While contact in the glassy phase, the SMP adhesive skins are under non-sticky attachment states with various target surfaces, offering minimal adhesion to avoid the adverse influence of adhesion. FIG. 59(A) shows the snap shots of utilizing the integrated electronic adhesive skins (see FIG. 48(A) for details of the structures and components) in surface texture detection and the recorded signals from the pressure sensor integrated in the robotic adhesive skin). When contacted in the glassy state, the adhesion between the robotic finger equipped with the electronic adhesive skin and the segregated substrate is minimal, enables the smooth traverse movement of the robotic finger across surfaces with small-amplitude (FIG. 59(A)-(i)), medium-amplitude (FIG. 59(A)-(ii)) and large amplitude textures (FIG. 59(A)-(iii)), and the recorded relative change of the resistance of the pressure sensor can clearly discern the change of the surface textures, both in wavelength and wave amplitude. FIG. 59(B) and 59(C) further compares the application of the robotic adhesive skin in handling an ultra-light clean cloth weighting only 3.89 g. While contacting in the soft rubbery phase (FIG. 59(B)-(i) and 59(B)-(ii)), even the weak rubbery adhesion will lead to the sticking of the cloth on the adhesive skin and hinder the release process (FIG. 59(B)-(iii) to 59(B)-(v)). Opening the pinch of the robotic fingers or even subsequent continuous swing of the fingers also cannot remove the cloth from the robotic finger due to the existence of residual adhesion. In contrast, while contacting in the glassy phase, the frictional force from the opposite fingers is able to grip the light cloth(FIG. 59(C)-(i) and 59(C)-(ii)), and when the pressure is removed after the robotic hand open the finger, no residual adhesion force will hinder the release process (FIG. 59(B)-(iii) to 59(B)-(iv)). The glassy adhesion can be reduced further when the preload is removed, and therefore the robotic adhesive skins can be utilized in more human-machine interaction scenarios where adhesion is not required.

[0418] Example 25. Applications in Switchable Adhesion States

[0419] After forming sticky contact under the rubbery phase, the robotic adhesive skins can provide switchable adhesion between the high and low adhesion strength values under the R2G and rubbery conditions, respectively. This on-demand switchable adhesion can be used to empowering the robots’ manipulation capabilities. FIG. 60(A) shows the snapshots of utilize the robotic adhesive skin mounted on the robotic finger for a gripping and release of a fragile quail egg weighting 14 g. First, the adhesive skin is heated to the rubbery phase and bring contact with the quail egg (FIG. 60(A)-(i)). Then the adhesive skin is cooled down to the glassy phase in situ (FIG. 60(A)-(ii)). Friction, known for its strong dependence on the preload and a small ratio between the gripping force and the preload (i.e., the coefficient of friction, usually smaller than one), requiring a large preload for gripping and tend to damage and deform the objects during gripping. In contrast, the high R2G adhesion enables gentle gripping of the quail egg with a very small preload with a large ratio (>50) between the grip force and preload. The strong R2G adhesion can ensure that the quail egg adheres strongly on the skin for resisting acceleration and vibration during movement (FIG. 60(A)-(iii) and 60(A)-(iv)). Finally, when heated to the rubbery phase, the adhesion is switched to a low value for easy release (FIG. 60(A)-(v)).

[0420] With the adhesive skins, a robotic hand can also own other superior handling capabilities. For example, parallel gripping of multiple quail eggs can be realized by adhering quail eggs to different adhering sites on the adhesive skins simultaneously, which can greatly enhance the handling efficiency of a robotic hand (FIG. 60(A)-(vi)). Since the adhesion is based on physical interactions (i.e., shape-locking and van de van der Waals force) that do not have strong dependence on the adherend materials and surface conditions, this robotic adhesive skin can grip various kinds of materials, including plastics, wood, glass, and ceramics, as shown in FIG. 60(A)-(vii) to 60(A)-(x).

[0421] Furthermore, with adhesion, a robot can also grip objects with various radius of curvatures, including the smooth steel shells with radii ranging from 10 cm to 40 cm, to the flat glass plate with a radius of curvature of infinity, as shown in FIG. 60(B). These handling capabilities can never be possible for frictional designs. Gripping multiple objects at one time and a large ball (e.g., basketball) is also of great challenge for humans. However, when equipped with an electronic adhesive glove using the robotic adhesive skins (FIG. 61(A)), an ordinary person can own the extraordinary superpower of gripping like a Spiderman, such as gripping quail eggs securely on the tips of every finger (FIG. 61(B)) and firmly gripping a large and smooth steel shell with a radius of 40 cm (FIG. 61(C)).

[0422] The preload that a robotic gripper can apply is always limited, by either the damage limit of the object or the loading capability of the motors. Therefore, the gripping capability using friction is always limited. As shown in FIG. 60(C), the robotic hand fails to grip a heavy aluminum block (1.2 kg) through an acrylic hand under the glassy phase due to limited frictional force and resulting sliding of the gripping. In contrast, if the robotic hand contacts the object in the rubbery phase (FIG. 60(C)-(i) and 60(C)-(ii)) and then cool to the glassy phase (FIG. 60(C)-(iii)), the high R2G adhesion strength enables reliable gripping (FIG. 60(C)-(iv)) and movement (FIG. 60(C)-(v)) of the aluminum block. When heated to the rubbery phase again (FIG. 60(C)-(vi)), the adhesion of the SMP adhesive skin is reduced to low rubbery state, facilitating subsequent release of the block when the robotic hand is opened (FIG. 60(C)-(vii)).

Claims

Claims1. A gripping member, the member comprising: a base portion formed from a rubber-to-glass phase transitioning material; and one or more fibrils extending from the base portion made from a rubber-to-glass phase transitioning material, wherein the rubber-to-glass phase transitioning material: can transition from a glass phase to a rubber phase upon application of a stimulus and return to the glass phase in an absence of the stimulus; has an elastic modulus of less than 1 MPa in the rubber phase; and the transition from a glass phase to a rubber phase occurs in the one or more fibrils having a radius of at least 0.1 mm in size.

2. The member according to Claim 1 , wherein the rubber-to-glass phase transitioning is selected from one or more of a shape memory polymer, a liquid metal, a liquid-crystalline polymer, and a jamming material.

3. The member according to Claim 2, wherein the rubber-to-glass phase transitioning is a shape memory polymer.

4. The member according to Claim 3, wherein the shape memory polymer is an E44 epoxy shape memory polymer.

5. The member according to any one of the preceding claims, wherein the size of one or more fibrils are selected to have: a dimensionless parameterin the glass phase, such that it has adhesion properties under a DMT-like adhesion regimen; and a dimensionless parameterin the rubber phase, such that it has adhesion properties under a JKR-like adhesion regimen, where is obtained from formula (1)where is the radius of a single fibril, and vbeing the Young’s modulus andPoisson’s ratio of the fibril material, respectively being the theoretical strength of adhesion,and being the work of adhesion.

6. The member according to any one of the preceding claims, wherein each of the one or more fibrils has a radius that is greater than a critical radius in a rubber phase andless than a critical radius in the glass phase as calculated according to Formula (2)and Formula (3), respectively:where and are the Young’s modulus under the rubber phase and glass phase,respectively, and and are as defined in Claim 5.

7. The member according to any one of the preceding claims, wherein when the rubber- to-glass phase transitioning material is an E44 shape memory polymer, then the radius of each of the one or more fibrils may be from 0.5 mm to 3 mm, such as about 2.5 mm.

8. The member according to any one of the preceding claims, wherein the member displays a switchability of from 20 to 2,000, such as greater than or equal to 1 ,000 to less than or equal to 2,000. (i.e., the ratio between the maximum adhesion force for attachment and the minimum adhesion force for detachment)9. The member according to any one of the preceding claims, wherein the one or more fibrils are a plurality of fibrils, each fibril having a radius of from 0.5 mm to 3 mm, such as from 1.5 mm to 2.5 mm.

10. The member according to Claim 9, wherein the plurality of fibrils are arranged in an array, optionally wherein the effective contact area is up to 1 ,000 mm2.

11. The member according to any one of the preceding claims, wherein each of the one or more fibrils has an aspect ratio of from 1 to 50, such as from 5 to 10, where the aspect ratio is the height of a fibril divided by its radius.

12. The member according to any one of the preceding claims, wherein the stimulus is selected from one or more of heat, light (e.g. Ultraviolet light), an electrical current, and pressure (e.g. negative pressure relative to standard pressure).

13. The member according to any one of the preceding claims, wherein one or both of the following apply:(I) the adhesion efficiency of the member is greater than 50%, such as greater than 55%, such as greater than 57.8%; and(ii) the adhesion strength of the member is more than 1 MPa, such as from 1 MPa to 5 MPa, such as from 1.5 MPa to 2 MPa, such as about 1.89 MPa.

14. A gripping apparatus comprising: at least one gripping member according to any one of Claims 1 to 13; and a stimulating means or apparatus, wherein the stimulating means or apparatus is configured to supply a stimulus to transition the gripping member from a glassy phase to a rubber phase.

15. The gripping apparatus according to Claim 14, wherein the stimulating means or apparatus is formed integrally with the gripping member.

16. The gripping apparatus according to claim 15, wherein the stimulating means or apparatus is formed integrally with the base of the gripping member.

17. A gripping device, comprising: at least one gripping apparatus according to any one of Claims 14 to 16; and a substrate, wherein the at least one gripping apparatus is attached to the substrate,18. The gripping device according to Claim 17, wherein the substrate is a robotic hand or a glove suitable for use by a human.

19. A method of gripping using a gripping device according to Claim 17 or Claim 18, wherein the method comprises:(A) providing an object for gripping and a device according to Claim 17 or Claim 18;(B) applying a stimulus to the at least one gripping apparatus to transition the one of more fibrils of the rubber-to-glass phase transitioning material into a rubber phase, then bringing the one or more fibrils into contact with a surface of the object to form an adhesion between the surface of the object and the one or more fibrils; and(C) removing the stimulus to return the one or more fibrils to a glass phase, thereby gripping the object.

20. The method according to Claim 19, wherein the method further comprises the step(s):(D) optionally physically manipulating the gripped object; and(E) reapplying the stimulus to the at least one gripping apparatus to release the object.

Images