Additive manufacturing of flexible 3D surface electrodes for electrostatic adhesion control and smart robotic gripping

Mechanically flexible surface structures with embedded conductive electrodes are attractive in contact-based devices, such as those used in reversible dry/adhesion and tactile sensing. Geometrical shapes of the surface structures strongly determine the contact behavior and therefore the resulting adhesion and sensing functionalities; however, available features are often restricted by fabrication techniques. Here, we additively manufacture elastomeric structure arrays with diverse angles, shapes, and sizes; this is followed by integration of conductive nanowire electrodes. The fabricated flexible three-dimensional (3D) surface electrodes are mechanically compliant and electrically conductive, providing multifunctional ability to sense touch and to switch adhesion via a combined effect of shear- and electro adhesives. We designed soft, anisotropic flexible structures to mimic the gecko’s reversible adhesion, which is governed by van der Waals forces; we integrated nanowires to further manipulate the localized electric field among the adjacent flexible 3D surface electrodes to provide additional means to digitally tune the electrostatic attraction at the contact interface. In addition, the composite surface can sense the contact force via capacitive sensing. Using our flexible 3D surface electrodes, we demonstrate a complete soft gripper that can grasp diverse convex objects, including metal, ceramic, and plastic products, as well as fresh fruits, and that exhibits 72% greater electroadhesive gripping force when voltage is applied.


Introduction
Small-scale structural shapes on surfaces with mechanical flexibility often provide intriguing functionalities. The most well-known example is the reversible manipulation of surface adhesion inspired by geckos in nature [1][2][3]. Geckos have tiny fibrillar surface features on their feet which are mechanically flexible. While they can climb walls freely, their switchable adhesion is enabled by the deformable fibrillar surface structures regulating the shear-induced contact area [4,5]. Mimicking these gecko-like surface structures, several engineering approaches have been introduced and commercialized [6][7][8][9]. Other intriguing structures are cupped microstructured surfaces inspired by the suction organs of aquatic animals and phalanx surfaces of end effectors with triboelectric nanogenerators (TENG) inspired by human fingers. Elastomeric microstructures fabricated by replica molding can hold objects by the suction force generated by surface deformation and cavity collapse [10,11]. Soft smart actuators integrated with TENG that can enable versatile actuation mechanisms, tactile sensing, and energy harvesting upon bending and contact were also reported [12].
Because they are extremely soft yet highly recoverable, elastomers are commonly used as materials for realizing engineered surfaces with flexible www.Springer.com/journal/40544 | Friction characteristics. Different fabrication technologies have been introduced for structuring elastomer surfaces. The most common manufacturing method is the microreplication process, which involves manufacturing a mold using wafer-based lithography, followed by pouring and polymerizing a resin-based polymer such as polyurethane or polydimethylsiloxane (PDMS) into the negative mold [13][14][15]. Surface microstructures larger than 100 μm can also be fabricated by surface and shape deposition manufacturing (S 2 DM) combined with micro-sculpting, shape deposition manufacturing (SDM), laser cutting, and laser patterning [16]. However, these fabrication approaches are limited in their promotion of design complexity for the surface structures [8,15]. The focused ion beam (FIB) process injects ions onto the surface, enabling precise patterning of three-dimensional features [17], but the extremely low production speed restricts its usage.
Compared to other fabrication technologies, additive manufacturing approaches provide effective routes for fabricating complex structural designs [18,19]. Their recent advances toward real product manufacturing in diverse fields have been facilitated by expansions in both processes and materials [20][21][22]. Therefore, many deformable commercialized elastomeric filaments have been introduced for material extrusion printing, and photo-curable resins in a range of mechanical properties are also available via vat photopolymerization processes, such as stereolithography (SLA) or digital light processing (DLP) [23][24][25][26]. Using stereolithography techniques, direct three-dimensional (3D) printing of elastomeric adhesive has been inspired by the gecko [27]. In addition, an anisotropic surface that mimics shark skin to reduce fluid drag on surfaces has been realized by recent 3D printing methods and this surface has been functionally applied in pipes, teeth, and artificial lotus leaves [28].
In this study, we introduce additively manufactured 3D surface structures comprising soft elastomers integrated with conductive yet compliant metal nanowires; these structures can sense touch and switch their adhesion via a combined effect of shear-and electro adhesives, as in Figs. 1(a) and 1(b). We design and fabricate surface composite structures with increased complexity enabled by the additive manufacturing process, as in Fig. 1(c), and integration of metal wires on top of the 3D structures forming an interdigitated circuit, as in Fig. 1(d). We then explore the synergic effect of both shear-and electro-adhesion according to different structural designs, including design factors such as structural angle, shape, and size. The resulting mechanical shear force and further electrical adhesion tunability with applied voltage are characterized. In addition, we investigate the ability of 3D surface electrodes to sense the touch of objects, perceive contact force, and distinguish material type using capacitance change between adjacent structures in deformation. Using these new 3D surface electrodes, we demonstrate a robotic gripper that can grasp and release diverse objects, such as a tangerine, as in Fig. 1(e).

Fabrication of flexible 3D surface electrodes
We designed the flexible structures to be 3D and anisotropic, comprising straight pillars standing at an angle against the bottom film. Figures S1 and S2 in the Electronic Supplementary Material (ESM) illustrate the fabrication procedure of our flexible 3D surface electrodes. The soft base structures were additively manufactured via stereolithography (SLA) printing (Form3; FormLabs) using an elastomeric resin (Elastic Resin; FormLabs). The use of additive manufacturing allowed us to vary the structural shapes, slant angles, and length-scale. The representative design of the surface structure has a width of 1.5 mm, a length of 8 mm, and an angle of 30°. Each pillar was designed to be tapered to 2.5°, making it thinner, and therefore more compliant, at the top, and thicker, and thereby more durable, at the bottom. After all post-curing processes, the angles of the flexible structures fabricated through SLA 3D printing were measured through optical image analysis using a camera and the image processing program (Image J) confirmed that the angle errors are within ±3°.
To build interdigitated circuits on the 3D flexible structures, we integrated copper wires on both sides of the bottom film, and then deposited AgNW (silver nanowire) ink on the surfaces of both the pillars and the copper wires. The elastomer-AgNW composite pillars were electrically conductive and mechanically compliant, prerequisite characteristics to achieve both the capacitive tactile sensitivity and electromechanical adhesion tunability. The AgNWs were particularly suitable as they can maintain their conductivity even under large mechanical deformation [29]. Finally, the whole surface was coated with an insulating cover layer with the same elastomeric resin as base structure.

Characterization of mechanical stiffness and shear force (friction)
To determine the mechanical stiffness of the elastomeric surface structures, we conducted compression tests using a universal mechanical tester (uta-500N, YEONJIN S-tech) and a hemispherical tip, as in Fig. S3 in the ESM. We first placed the flexible 3D surface electrodes upside down facing a flat metal surface, while an acrylic plate was attached on the other side of the surface to distribute the compressive force uniformly throughout the structures when they are compressed by the hemispherical tip. The stiffness was determined as the force required to produce a vertical displacement of 1.5 mm at a constant strain rate of 3 mm/min. To minimize the effect of friction force, we applied a lubricant between the acrylic plate and the hemispherical tip to enable horizontal sliding. Each test was repeated three times or more. For shear force measurement, we used two different types of equipment, a universal mechanical tester and a tribometer (UMT TriboLab, Bruker). When testing with the universal mechanical tester, we first fixed the flexible 3D surface electrodes in one tensile gripper and an medium density fiberboard (MDF) wood plate in another, as in Fig. S4 in the ESM. The two surfaces were manually controlled to induce physical contact; we then slid each structure over the MDF wood to a distance of 3 mm at a constant speed of 20 mm/min while collecting friction data via the load cell. The average kinetic friction values were used to determine the shear force. In general, we adjusted the initial shear force, without external voltage, to be 0.2 N (±0.02 N). For several of the high load tests, the initial shear forces were adjusted to be 1 N (±0.1 N) and 5 N (±0.5 N). When www.Springer.com/journal/40544 | Friction testing with the tribometer, the MDF wood plate was attached to a load cell (DFM-2.0, Bruker) capable of measuring a load of 0.2-20 N with a resolution of 1 mN while sliding the flexible structures at a sliding speed of 1 mm/s, as in Fig. S5 in the ESM. The tribometer allowed us to change the normal force in the middle of the sliding test. Accordingly, we measured the friction force of the sliding structure at different normal forces of 0.5, 1, and 5 N without extra adjustment. Each measurement was conducted for 20 s and we used the average value of kinetic frictional force to determine the shear force. Between each measurement, we calibrated the normal force for 20 s.

Characterization of tactile sensing
We conducted vertical compression of the flexible 3D surface electrodes and measured the capacitance using a multimeter (CEM, DT6500), which controlled the contact force, displacement, contact area, and contact location. We used the reference size of the flexible 3D surface electrode to measure the capacitance change ratio according to the contact force, and quarter-sized surface electrodes with angle of 30° to investigate the charge dependency on contact location, contact area, and material type. For measurement with the controlled contact force and displacement, we employed the universal mechanical tester (uta-500N, YEONJIN S-tech). For measurement with the controlled contact position and contact area, we placed a dead weight on top of the flexible 3D surface electrodes using a 3D printed polylactic acid (PLA) cuboid. When measuring capacitance according to different materials, we first placed the flexible 3D surface electrodes upside down facing the targeted material at the bottom and put a 500 g weight on the top.

Soft gripper design and gripping force measurement
For demonstration of the flexible 3D surface electrodes with tunable adhesion, we built a gripper prototype, as in Fig. 2, following the design originated by Hawkes et al. [2]. This gripper design comprises two main components: a set of opposed gecko-inspired adhesives enabled by directional surface fibers, and a bi-stable support. In particular, the compliant surface fibers allow the gripper to grasp convex objects, even those with size larger than the gripper, using almost exclusively the shear forces. In this study, we used elastomer-AgNW composite surface electrodes with diverse angles, shapes, and sizes as the adhesives, and investigated the additional electrical tunability of the resulting gripping force. The gripping forces of the soft robotic gripper were determined as the maximum weight that the gripper could hold. Cylindrical objects with diameters of 30, 40, 50, and 60 mm were prepared using material extrusion 3D printing; their surfaces were covered with either paper or polypropylene (PP). Then, we set the soft gripper to grasp the cylindrical object and fixed it with the tensile grips in the universal mechanical tester. While pulling the gripper and object in opposite directions at a constant speed of  | https://mc03.manuscriptcentral.com/friction 50 mm/min, we collected data on tensile force from the load cell and defined the maximum tensile force as the gripping force. All tests were repeated with different external voltages. The gripping force was determined as the maximum force at which slip occurs.

Interfacial adhesion according to slanted angle
Despite the extreme freedom of pillar design and fabrication, we first altered only the slanted angle, as in Fig. 3(a), and investigated its effect on adhesion behavior. We selected five different angles from 15° to 75°, representing surface structures with different levels of vertical compliance, as in Fig. S6(a) in the ESM. When we experimentally measured the friction force of the structure arrays, they were found all to be similar regardless of the slanted angles, as in Fig. 3(b). However, when an external voltage was applied to the surface electrodes, the friction forces varied according to the slanted angle and the difference became greater with higher applied voltages, as in Fig. 3(c). Here, we fixed the reference adhesion at 0 V at ~0.2 N (±0.02 N); the greatest increase of 162% in lateral adhesion with voltage of 3 kV was observed for the smallest slanted angle, whereas the lowest increase of 20% was observed for the largest angle.
The difference additional shear force driven by the external voltage is due to the greater contact area created in the surface structures with smaller slanted angles and more compliant behavior. When the slanted surface electrodes are in touch with an object and slide over the object, the individual structure will deform and create a contact area with the counter surface, as in Figs. 3(d) and 3(e). The pillar structures with low slanted angle will likely have more contact area due to their geometry and low stiffness. Although the contact area differs, which results in different levels of adhesion, the reason why the friction forces remain constant is the change of resistance due to elastic recovery. In steady state, the shear force F s may balance with the springback force and the adhesive force of the directional elastomeric structures, which can be expressed as  where F n is the normal load, *  is the angle of the deformed structure, A is the contact area, and α is the lateral adhesive strength coefficient based on the assumption that the interfacial adhesion between the elastomer and the object is proportional to the contact area. When elastomeric structures with small slanged angle are in sliding contact, the resistance of the deformed pillars is low but the interfacial adhesive strength becomes high due to the large contact area. On the other hand, when pillars with large slanged angle slide over with shear, the surface adhesion is low because the contact area is small; however, the strong elastic recovery dominates the shear resistance. Accordingly, small slanted angles may be beneficial for shear-induced adhesives, yet pillars with angles or levels of stiffness that are too low will easily collapse and lose the functionality they obtained from their structural directionality. When the contact area differs according to the slanted angles, the amount of electrical adhesion induced by the external voltage through the surface electrode structures will change, as in Fig. 3(f). When voltage is applied to an interdigitated structured array, each row of its pillars can be electrically charged to have potential opposed to those of adjacent rows. Then, localized electric fields are generated at the interface, which induces electrical charge on the object surface. Additional adhesion develops at the interface by electrical charge induction. Such electrostatic induction will scale with the contact area; therefore, pillars with smaller angle and lower vertical stiffness exhibit larger electrical adhesion tunability, as in Fig. S6(b) in the ESM.

Lateral adhesion according to electrode shape
We next expanded the design freedom to the structured electrode shape. Using the additive manufacturing process, we were able easily to realize complex structural shapes of individual pillars. We printed diverse shapes of flexible structures, as in Fig. 4(a), and compared the enhancement of lateral adhesion by external voltage, as in Fig. 4(b). The reference adhesion at 0 V was again fixed at ~0.2 N (±0.02 N), and all the pillar angles were 30°. The lateral adhesion tunability of the rectangular shape was measured at 57% at 3 kV. First, we found that the shallow width at the pillar top reduced the electrical adhesion enhancement. Triangular-shaped pillars having small width at the top and large width at the bottom resulted in low electrical adhesion tunability of 25%, because the interfacial adhesion highly depends on the contact area at the tops of the pillars. A tree shape with the same width at the top pillars exhibited approximately 31% greater adhesion tunability than that of the triangular pillars, because the hinge-like bottom reduces the vertical stiffness. Next, we designed inverted triangles and T-shaped pillars that had large width at the pillar top and small width at the bottom, expecting improved electrical tunability. However, the electrical tunability values were 58% and 49%, respectively, which were similar to or less than that of the rectangular shape.
Such results are in accordance with results determined according to slant angles because the electrical tunability during adhesion depends highly on the structure, which likely induces the large contact area. Because the geometrical shape with wide top and narrow bottom allows large contact width and low stiffness, we can expect that the inverted triangle and T-shape will create a contact area larger than that of the triangular and tree shaped pillars. The friction coefficient of the inverted triangle, which is greater than that of the tree, also supports the large contact area of the inverted triangle. According to Eq. (1), resistance by elastic recovery should remain almost constant as long as the slanted angle of deformed structures and normal load are the same; therefore, the difference in coefficient of friction may represent the difference in contact area according to the structural shape. However, the friction coefficient of the T-shaped surface structure showed small shear force and yet large electrical tunability. This was due to the low stiffness of the narrow bottom of the T-shape, which reduced the angle of the deformed structure in the first term of Eq. (1), offsetting the increase in the second term by the large contact area.
To further enhance the electrical adhesion tunability, we utilized the T-shape design, which provides extra space for placing additional pillar rows between the shallow bottoms. One of the design restrictions of 3D Friction 11(11): 1974-1986 (2023) | https://mc03.manuscriptcentral.com/friction interdigitated electrodes is that the electrode structures should not overlap when they are perfectly flat. This is particularly important when the arrays contact convex or arbitrary objects where pillar deformations can be spatially concentrated. Therefore, the rectangular pillar rows should have spacing greater than the pillar length. Using the T-shape, in contrast, we can reduce the spacing between adjacent rows without overlap, as in Fig. 4(c). In addition, the width of the top pillars can remain large while the bottom width is kept small. Such structured arrays can increase the density of electric field between opposed charges and therefore enhance the electroadhesion; they can also provide low vertical stiffness at the bottom structure and large contact width at the top, which can maximize the contact area. Using a zigzag-T shape, we were able to achieve adhesion tunability of ~103% with an applied voltage of 3 kV. We note that such electrical tunability may decrease when the initial mechanical force is high, as in Fig. S7 in the ESM.

Lateral adhesion according to electrode size
We further altered the sizes of the surface structures, an essential process for dry adhesives. Half-scale and quarter-scale flexible structures were designed and their shear forces were compared with that of the reference structure. The half-scale surface electrode comprised structures with height of 4 mm, structure row spacing of 3.4 mm, and 12 rows; the quarter-scale had height of 2 mm, structure row spacing of 1.7 mm, and 24 rows, as in Fig. 5(a). We first measured the shear force of each scale without any external voltage, www.Springer.com/journal/40544 | Friction as in Fig. 5(b). As the size of the structures decreased, the shear force increased and the amount of increase according to the size reduction was greater when smaller normal load was applied. This tendency was more apparent in the coefficient of friction (COF) plot according to the normal load, as in Fig. 5(c). We believe that the pillar density plays an essential role, particularly when material is in slight contact, because the increase in number of contacting pillars enhances the overall contact area. Accordingly, we can expect that surface structures with smaller size may induce greater contact area. We next compared the electrical tunability by measuring the ratio of the shear force at a voltage of 3 kV to that at a voltage of 0 kV, as in Fig. 5(d). In all cases, the electrical tunability increased as the structure size decreased, exhibiting improvements of 108%, 47%, and 10% at fixed reference adhesions of 0.2, 1.0, and 5.0 N respectively. The additional adhesion induced by electroadhesion was proportional to the contact area; therefore, the electrical tunability increased in miniaturized surface electrodes.

Capacitive tactile sensing
Flexible 3D surface electrodes can also function as capacitive tactile sensors. The tactile sensing of flexible 3D surface electrodes is practically useful for detecting, without any change or additional device, normal forces or contact areas applied to the surface. We experimentally investigated the sensitivity of the capacitance change according to different vertical displacements, contact locations, normal forces, contact areas, and material types. First, the capacitance change had approximately linear relationships with the normal force and the vertical displacement, as in Figs. 6(a) and 6(b), which informs us that the capacitance relies strongly on the amount of structural deformation. Accordingly, it was possible to correlate the capacitance change to the contact force. The sensitivity and linearity are significant factors that determine the performance of strain sensors [32]. In

1982
Friction 11 (11): 1974-1986 (2023) | https://mc03.manuscriptcentral.com/friction a comparison of sensitivity, a human hand can feel an indentation of 0.01-0.04 mm [33], and yet the minimum deformation we were able to perceive by capacitance change of a single pillar was 0.18 mm. Although the flexible 3D surface electrodes showed sufficient sensitivity and linearity, a more precise measurement system, including a compressive indenter, should be designed to further investigate the sensitivity limit of such surface structures. Second, contacts with identical normal forces were made at different locations, as in Figs. 6(c) and 6(d). The results verified that the tactile sensitivity was independent of the contact location. With the same contact force, the capacitance change due to the contact area difference was less than 4%, as in Fig. 6(e), which confirmed that the surface can sense the contact force regardless of the contact area. Finally, the flexible 3D surface electrodes detected different capacitance changes for different materials at the same normal force (and the same amount of deformation). In general, capacitance changes of conductive materials (human hand and steel) were greater than those of nonconductive materials (paper and PLA), as in Fig. 6(f). The capacitance increases by contact of PLA plastic and paper were 22% and 31%, respectively, whereas that of steel and the human hand were observed to be 154% and 473%, respectively.

Shear-adhesive gripping using surfaces with flexible 3D surface electrodes
We used the soft 3D surface electrodes in a shearadhesive robotic gripper and explored their utility for object grasping of objects with different sizes and shapes. This gripper design allowed effective grasping of objects having convex shapes without squeezing, utilizing the shear adhesion derived from the flexible structures. The flexible 3D surface electrodes also provided means to induce tactile sense of the object touched by tracking the capacitance change. Furthermore, applying a strong DC voltage can further tune the surface adhesion, because the mechanically compliant anisotropic structures with embedded electrically conductive electrodes can couple mechanical and electrical adhesive behaviors, resulting in enhanced maximum adhesive strength, as in Videos S1 and S2 in the ESM.
We first tested the gripping forces of the soft gripper integrated with 3D surface structures having rectangular and zigzag-T shapes. As the gripper design allowed grasping of any convex shape, including those with sizes greater than the gripper, we tested cylindrical objects with a range of diameters of 30, 40, 50, and 60 mm, as in Fig. 7(a). To minimize the effect of the rough surfaces of the 3D cylinders printed using fused filament fabrication (FFF) printer, the objects were covered with smooth polypropylene (PP) and paper. The bi-stable gripper design [2] allowed shear adhesion induced by object weight and therefore induced mechanical gripping force even without external voltage.
When we applied external voltages to the flexible 3D surface electrodes, we were able to observe enhancement in gripping force that scaled with the voltage, as in Figs. 7(b) and 7(c). The rectangular pillars in general exhibited greater gripping force than that of the zigzag-T structures. These results indicate that the rectangular structures induced more contact area against convex objects and therefore are more efficient than the zigzag-T structures for current gripper design. When external voltage of 3 kV was applied to the 3D structures, however, we found that the gripping force of the zigzag-T structures increased to a level identical to or even greater than that of the rectangles, as in Fig. S8 in the ESM. The gripping force of the zigzag-T was 72% greater at 3 kV, while that of the rectangular device was 45% greater. The higher electrical adhesion tunability of the zigzag-T is in consistent with the results of the previous shear force measurements. Accordingly, while both rectangular and zigzag-T structures provide the same range of overall tunable gripping force, the zigzag-T structure has a greater range of electrical tunability. Although the surface electrodes repetitively deformed during the gripping test, degradation in their functionalities did not appear, which ensured sufficient reliability. We proceeded with at least 96 gripping tests using one gripper, and yet observed neither permanent deformation nor failure. However, the performance may degrade by accumulation of surface contaminants. Therefore, it was important for us to frequently clean the cylindrical object using Isopropyl alcohol (IPA) and air-blowing.
www.Springer.com/journal/40544 | Friction For further demonstration, we provide images in Figs. 7(d)-7(f) of the soft robotic gripper with 3D surface electrodes handling fruits (a tangerine), metals (a tumbler), ceramics (a mug), and plastic (a water bottle). We applied external voltage of 3-4 kV for all cases to pick objects up. Although the gripping forces highly depend on the object surface material and roughness, we showed that soft 3D surface electrodes can be used for almost any material.

Conclusions
In this work, we designed and fabricated flexible three-dimensional (3D) surface electrodes that can digitally sense contact and switch adhesive strength via a combination of shear-and electro-adhesion. We additively manufactured base elastomeric structures using stereolithography and coated them with metal nanowires to enable mechanically-deformable, yet electrically-conductive surface electrodes. The soft, anisotropic flexible structures can mimic the gecko's reversible adhesion, which is governed by van der Waals forces, whereas manipulation of localized electric field among adjacent 3D flexible surface electrodes can provide additional means to digitally tune the electrostatic attraction at the contact interface. We experimentally investigated the friction force and its electrical tunability in flexible structures with diverse standing angles, shapes, and scales. High electrical tunability in surface adhesion can be achieved in structures that can easily create large contact area when deformed. The flexible 3D surface electrodes can also perceive object information from capacitance change. The capacitance was highly correlated with the deformation of flexible 3D surface electrodes and with the material type. Finally, we integrated the surface electrodes in a soft robotic gripper designed to grasp convex objects based on shear induced adhesion, without squeezing. Our gripper can increase the gripping force up to ~72% by applying voltage of 3 kV. Using the gripper, we demonstrated grasping of diverse objects, including fruit, metal, ceramic, and plastic products, by control of both shear-and electro-adhesion. For broader use of our flexible 3D surface electrodes, additive manufacturing can be a powerful tool to further tailor the sensing and adhesion-switching functionalities. and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.