Adhesion Circle: A New Approach To Better ... - ACS Publications

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

The adhesion circle: A new approach to better characterize directional gecko-inspired dry adhesives Yue Wang, Samuel Lehmann, Jinyou Shao, and Dan Sameoto ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b11708 • Publication Date (Web): 30 Dec 2016 Downloaded from http://pubs.acs.org on January 1, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

The adhesion circle: A new approach to better characterize directional gecko-inspired dry adhesives Yue Wang†‡, Samuel Lehmann†, Jinyou Shao‡ and Dan Sameoto*† †

Department of Mechanical Engineering, University of Alberta

10-203 Donadeo Innovation Centre for Engineering, 9211 116 St. NW, Edmonton, AB, Canada, T6G 1H9 ‡

Micro- and Nano-technology Research Center, State Key Laboratory for Manufacturing Systems

Engineering, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China *E-mail: [email protected]

Abstract: The number of different designs of directional gecko inspired adhesives has proliferated over the past 15 years, but some basic characterization tools are still non-standardized, which can make direct comparisons of different adhesives in the literature difficult. By far the most common type of test for directional adhesives, the Load-Drag-Pull (LDP) test is useful but can miss substantial information on the exact behavior of gecko-inspired adhesives in a variety of loading conditions. Other test techniques, including angled approaches and pull-offs have been employed by a few groups but they are not as widely adopted; peel tests can be employed but require a larger amount of adhesive material to use in the test, which is not always practical given some current manufacturing constraints. Very few tests have looked at the effect of off-main axis loads on the performance of directional adhesives however, and this quality of performance may be very important in applications where direct control over displacements or angle of pull-off in pitch and yaw of the peeling interface may not be practical or possible. To address this overlooked area of characterization, we introduce a new test concept for anisotropic adhesives, the adhesion circle, and also compare how the radial normal adhesion performance is altered depending on whether the pull-off comes after a displacement drag or when

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

pulled at a constant angle from vertical after a pre-load. Testing directional adhesive designs made with different geometries shows that unexpected behaviors at pull-off angles not in the direction of the strong-weak axis can sometimes be seen. The complete adhesion circle tests should help better design directional adhesives for scaled up performance, and can be completed with relatively simple hardware that is typically used in most current directional adhesive tests.

Keywords: Gecko, anisotropic fiber, angular, dry adhesion, radial test

1. Introduction

Since the discovery of van der Waals interactions as the primary mechanisms behind the remarkable adhesive capabilities of geckos1, researchers have attempted to manufacture materials that provide the same properties. In particular, Keller Autumn2 identified seven key properties of gecko inspired adhesives which should be duplicated for true gecko-inspired functionality, including anisotropic behavior, high adhesion coefficient (ratio of pre-load to pull-off force), low detachment force (when needed), material independent adhesion, self-cleaning ability, anti-self-adhesion, and non-sticky in their default state. The anisotropic behavior has been the greatest research focus to date, with development of angled fibers3-7, asymmetric fibers8, 9, Janus fibers, offset caps10-13 , and wedges 14-16 all introduced as means to reproduce the direction sensitive behavior of the gecko foot hairs. The number of tests run for these anisotropic adhesives is nearly as large as the types of manufacturing processes used to produce them, which can make direct comparisons of device performance challenging. For example one of the earliest tests was the simple load-drag-pull (LDP) test, introduced first in actual gecko seta measurements17, 18, and then adapted for a variety of synthetic mimics7, 9, 14, 19-23. The attraction of this test is that only small areas of adhesives are needed to test for directionality, unlike peel tests11 or macroscale shear tests5. There are several variations on the LDP (trials reported in literature depending on the equipment used; several LDP trials use a hemispherical indenter, which eliminates measurement


Page 2 of 21

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


errors but introduces complexity into the actual contact area achieved7, 19, 22, 23. Others use carefully aligned flat plates, which can have various surface roughness associated with them, but are more frequently very smooth9, 14, 20, 21. Some tests include both shear and normal force measurements simultaneously, while others report normal adhesion force vs. displacement as a primary metric. The difficulty of measuring multiple axes simultaneously is that it requires multiple force probes (sometimes custom made), or costly multi-axis force probes which are not often capable of very fine force resolution (< 1mN). However, one thing that nearly all reported LDP type tests have failed to report on is the sensitivity to off-axis displacements/trajectories compared to the primary strong-weak direction. While some LDP papers focus on the importance of incoming or out-going angles15, 21, only one report to our knowledge has reported a full radial angle dependence of adhesion in flat-flat contact24 and used a large sample (3.9 cm2) on a flat surface to complete tests. For this work, the anisotropic adhesive was being used in a climbing robot and had negligible adhesion in the absence of shear. Unfortunately, fewer details on the full three dimensional performances were listed with respect to sensitivity with angles, but rather a full limit surface was described with respect to maximum tangential and lateral forces that could be supported by a compressive normal load as well as adhesion. Anisotropic adhesives that are based on mushroom or spatula type geometries can support a significant normal adhesive force even in the absences of shear loads12, 22, 25, and what is desirable is the means to have a preferentially weak direction when loaded in a specific way to remove the material, but otherwise a strong passive adhesive performance. As an alternative method for testing, a simpler version of large-scale limit surface tests is completed with a hemispherical indenter and provides high quality information on the directional sensitivity of anisotropic mushroom shaped fibers produced in a variety of processes and compares for the first time the performance of a load-drag-pull and angled load drag pull in an adhesion circle. Different results than expected are seen for these fibers when tested in a complete adhesion circle, which indicates that for anisotropic adhesives, it is not just the on-off ratio of adhesion in the strong-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

weak axis that is important, but the tolerance to off-axis loading, and therefore should be considered in the test procedures. This adhesion circle approach can be a very useful tool to characterize anisotropy without the full complexity of earlier test techniques.

2. Fabrication and Designs: 2.1 Sample preparation Two typical types of anisotropic fibers were manufactured in different processes for the adhesion circle test. One has a square post with an offset cap. The fabrication of this fiber is based on a bi-layer photoresist mold consisting of polymethylglutarimide (PMGI) and AZ4620, where cap and fiber dimensions can be defined directly with two photolithographic masks in a negative mold template and then replicated in Sylgard 184 polydimethylsiloxane (PDMS)11, 22. The other fiber design introduced a defect on an overhanging cap with circular post. The fabrication process for this fiber is based on a deliberate defect method, in which a master template made from SU-8 and acrylic is formed with indented shapes on the overhanging caps12, 26. The negative silicone rubber molds (TC-5030 from BJB Enterprises) are produced as described previously and are then surface treated using CF4 Plasma27 to permit demolding of both silicone rubbers and other elastomers from the molds. Next, Sylgard 184 or Kraton G1657 styrene-ethylene-butylene-styrene (SEBS) is molded from the negative template to form fibers. These designs can be compared with the exact same adhesion trial types to determine quantitative and qualitative adhesion performance differences. Details on the fabrication protocols are reported in the Supporting Information. 2.2 Sample Designs The designs tested are shown in Figure 1 below. The bi-layer photoresist process produced rectangular fibers ~20 µm tall with caps 1.5 µm thick and a center to center spacing of 20 µm, while the acrylic


Page 4 of 21

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


molding defective caps process produced circular fibers that were approximately 25 µm tall with a cap 3 µm thick and a center to center spacing of 50 µm. Some important geometry dimensions and more detailed description of the microscope images are shown in the Supporting Information.

Figure 1. SEM images of two distinct fiber types manufactured in this process: (a) square fibers with an offset rectangular cap, (b) circular fibers with a deliberate defect etched in an otherwise symmetric mushroom shaped fiber. The left corner insets are microscope images to show the cap offset direction and the defect position. 3. Adhesion Test System

3.1 Equipment A 3-axis system of linear stages (Newport MFA-CC) aligned perpendicular to each other controlled by a Newport ESP-301 Motor controller is used in this work, in order to allow motion over all three axes (as shown in Figure 2a). Attached to one of the linear stages is a single axis Transducer Techniques GSO-25 load cell which uses a 6 mm diameter sapphire lens (Edmund Optics NT49-556) to contact adhesive samples, and the load cell is connected to a National Instruments USB-6289 data acquisition module. Both the ESP-301 motor controller and the National Instruments DAQ card are connected to a computer, which is then interfaced with through the use of a custom-built open-source application. This application is written in C# and allows the operation of various tests performed in our work, and the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

noise resolution of the final collected adhesion data by the computer is approximately 0.02mN and measurement error within 0.2 mN from the sampling and longer term variance. All force measurements are of adhesion normal to the surface of the sample.

Figure 2. (a) Adhesion setup and schematic for the adhesion circle tests. Three linear stages are used to move a single axis load cell with a hemispherical sapphire indenter. The sample is placed on a goniometer to minimize in-plane misalignments when completing drag displacements. (b) Schematic for two kinds of test: Pull-angle and Load-Drag-Pull test and definition for the pull-angle θ and angle α in X-Y plane. 3.2 Radial test A radial test consists of a series of outwards radiating strokes performed from a central point at varying angles (θ) above the horizontal plane, or drag displacements (DL) in the horizontal plane. Each of these strokes is separated by a desired angle (α) from the Y-axis in the horizontal plane (X-Y plane) until a full revolution of 360 degrees is reached, performed in a random order to help prevent the Mullins effect from influencing the adhesion data. Therefore, a single stroke is separated by an angle α from the x-axis in the X-Y plane and an angle θ in the vertical plane (normal to the X-Y plane), as shown in Figure 2b. Each stroke consists of several stages: a descent and preload stage, an ascent stage, and a reversion stage. At the beginning of each stroke, the probe descends vertically at a constant velocity until a specified preload is reached. Once this preload is reached, the probe remains stationary for a set


Page 6 of 21

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


amount of time. After this time has expired, there are two kinds of test used in our experiments. One test method is that the probe is dragged parallel to the surface at a velocity of 5 μm/s for displacement (DL) at angle α from the x-axis, and then the probe is withdrawn vertically (θ=90°). Another test method is that the probe is withdrawn directly without any drag at an angle of θ to characterize the adhesion property with different pull-angles (θ). Once the probe has lost contact with the sample, it returns back to its original position and is ready to begin another stroke at an increment of angle α =5° for these tests. This process then repeats until all of the strokes are completed, after a full 360 degree circle has been tested. To obtain a consistent adhesion test pattern, the direction of the offset-cap is placed aligned with the negative direction of the Y-axis (α =180°), and the defect-cap direction is aligned with the positive direction of Y-axis (α =0°) during all the radial tests. While it is possible to load the fibers at a pull oﬀ angle θ ≠ 90° aXer a drag displacement, this was not done here to minimize the total experiment design space. The software used to run this program collects data at a rate of ten samples per second, which is both displayed as a plot in real-time and collected for further analysis. Furthermore, the maximum and minimum value obtained by the force probe over the descent, preload, and ascent stages in each stroke is collected and can be imported into excel or other spreadsheet software for analysis. This maximum and minimum value is labeled as a stroke’s preload force and adhesion force respectively. For all trials, the preload was selected to be a nominal value of 5 mN, which had some slight variation due to lag between sensing and reaction of the linear stages, but was usually consistent within approximately 0.5 mN. For this preload, the expected contact diameter was at least 330 µm based on separate optical imaging. For the majority of fibers tested, the anisotropic adhesion was much less than the maximum that could be achieved with isotropic designs, so the maximum strains were low enough to not cause significant changes with repeated trials, but randomization of the order of α was maintained for all the experiments completed. The individual points on each adhesion circle represent singular tests in this



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

work, although once preferred loading conditions/angles are selected, multiple repeated trials can be completed for more averaged results. General trends and behavior are the focus of these trials to provide basic comparisons.

4. Results and Discussion 4.1 Adhesion Circle with Pull-angle

Different pull-angles (θ) were investigated to characterize the adhesion circle of these fibers, and no drag on the surface is first applied. θ varies from 30° to 90° in increments of 10°. At θ =90°, the measuring probe just moves down and up vertically and hence there is nearly no adhesion difference with α as shown in the plots, but these tests help to determine if there is any substantial change in adhesion with the number of cycles, and act as a baseline adhesion value. As shown in Figure 3a, with the θ increasing from 30° to 50°, the adhesion circle profile shifts towards the side where the overhanging caps stick out, and its area increases. At θ=50°, the adhesion force reaches a maximum (αmax) of approximately 25mN and a minimum of just below 1mN at αmax - 180°. The ratio of adhesion can be used to characterize the directional or anisotropic property of the adhesives and is over 25 for this fiber. Further increasing the pull-angle, the adhesion circle gradually returns back to the balanced position while the circle area also becomes smaller, which means that both the adhesion force and adhesion directionality becomes weak. In addition to providing the degree of the directionality, the adhesion circle can also present the information about adhesion azimuth. Taking the θ=50° pull test as an example, the adhesion force is larger than 15mN from 125° to 235°. Therefore, this kind of fiber can provide a strong adhesion (>15mN) in a loading angle of 110°. Also, the adhesion force at these angles is basically uniform and no large fluctuations occur, consequently leading to a stable adhesion area. More interestingly, the adhesion force decreases sharply once outside this region. The transition angle is


Page 8 of 21

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


merely 30° for the adhesion force to drop from over 20mN to below 5mN in the case of θ = 50°. This is very useful for practical applications involving a quick switch from an adhesion state to a release state.

Figure 3. Comparison of adhesion circles for different pull-angles (θ) by using the offset cap rectangular design (a) and defect cap circular design (b). Compared with the offset-cap designs, the line-defect cap shows similar qualitative adhesion properties but lower overall values of adhesion. This type of fiber also demonstrates good anisotropic adhesion property but not as high a ratio because the maximum adhesion values are lower. The adhesion force reaches a maximum value of approximately 17mN at one side and below 0.5mN on the other side when θ=50°. Also, taking the 50° pull test as an example, the strong adhesion (>10mN) azimuth is from 110° to 210°, which means this adhesive can provide more than 10mN force in a 100° angle which is a little less compared with Figure 3a. In general, using these two typical kinds of anisotropic designs, the adhesion circle patterns are employed as a new characterization tool to demonstrate the anisotropic adhesion property. In this case, with offset fiber caps, the geometries can lead to directional adhesion strength, but it is seen to be



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

substantially influenced by loads and displacements not in line with the nominal strong direction. Also, from the adhesion circle pattern, it can be concluded that even a small line-defect is able to produce substantial directional adhesion and the maximum force reaches nearly seven times that of the balanced position at the specific pull direction (α) and pull angle (θ).” These anisotropic behaviors are mainly caused by the asymmetric geometry designs of the fiber caps. The mushroom shaped cap functions to reduce stress concentrations present in a flat punch design, and results in high adhesion strength. The exact amount of cap overhang is strongly influential in the total adhesion strength and so if a portion of the cap is missing or defective overall adhesion will be lowed. If the cap is sheared upon loading, the defective side can have its local stress in tension lowered or increased, substantially changing the normal strength of the adhesive11,12. 4.2 Adhesion Circle with Load-Drag-Pull Test The LDP test is an adhesion measurement that includes three main steps: preload, drag and pull. The drag displacement is the major variable of interest in the LDP test process, and previous versions of this test mainly focused on one axis performance in the strong-weak direction. Here, by introducing the adhesion circle, the adhesion force in all drag directions around the adhesive fibers is presented. Figure 4a shows the test results of the offset type fibers when operated in LDP type tests. The drag displacement varies from 10 to 40 μm in an increment of 10 μm. With increasing drag displacement, the adhesion circle shifts to the strong side which is similar to the pull-angle test, and the adhesion circle pattern reaches its maximum values at 20 μm drag displacement. For the 30 μm displacement however, the adhesion force starts to dramatically drop in the “strong” adhesion axis because the caps begin detaching from the hemisphere indenter at this displacement. However, the adhesion force is still very large and little change in adhesion strength is seen when displaced towards the two corners of the opposite cap-offset side. The exact mechanism of this behavior is unclear, but it could be a combination


Page 10 of 21

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of factors including the slightly greater overhang at the corners of the square shaped fibers, slightly larger undercuts due to uncollimated light exposure of the defective cap fibers in the square array. Therefore, it needs a larger dragging displacement to cause the cap detachment from the contact

surface. Further increasing the drag displacement to 40 µm, the cap starts to separate from the indenter even at the corner position, leading to a generally decreased adhesion force, but with much more variations between, and less consistent performance at, similar angles. Usually, the one-axis LDP test is oriented with the strong-weak axis, which will miss substantial adhesion information in other orientations. Here, the adhesion force shows a peculiar heart shape when plotted against α when the drag displacement is larger than 20 µm.

Figure 4. Comparison of adhesion circles for different drag displacement (DL) by using the offset cap rectangular design (a) and defect cap circular design (b). For the defect-cap design, the drag displacement varies from 20 to 80 µm with an increment of 20 µm, which is roughly equivalent because the fiber height and width is approximately double that of the offset cap designs. Although it needs a larger drag displacement to reach the maximum adhesion, the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

results show the same general trends as the offset cap fibers. At 40 µm drag displacement, the adhesion circle reaches a maximum. As the in-plane displacement is increased to 60 µm, the adhesion is reduced in the nominal “strong” direction, yet still shows high strength at an α value approximately 30° off of the strong axis. After the drag displacement increases to 80 µm, the adhesion force drops further at nearly all strong adhesion side directions. The maximum adhesion strength and range of α is substantially lower than the rectangular offset cap design when pulled off at different θ, but has greater adhesion when run as a LDP type test. Similar offset maximum adhesion values are again seen for the LDP type test compared to the designed maximum α direction. In general, both these designs show very strong directional adhesion changes is pull-angle and drag displacements controls. Similar to the pull-angle test, the drag step can also contribute a lot to the anisotropic adhesion and the adhesion circle also provides more detailed information about the changing adhesion force with drag direction. However, the pull-angle test can provide a more stable adhesion measurement at the strong adhesion side because full cap contact can be lost during preload at large shear displacements and will strongly influence total adhesion strength. 4.3 Adhesion Circle with different material In order to demonstrate the effect of the fiber structural material on the adhesion circle measurement, the fibers with line-defect caps were fabricated using a thermoplastic elastomer (Kraton G1657) and compared with Sylgard 184 fibers made from the same master mold. Here θ=20°, 50° and 90° are chosen as the comparisons. This is because 20° and 90° are the smallest and largest pull-angle in the test respectively, and the adhesion circle is largest at 50° for Sylgard 184 fibers. The SEBS fibers show nearly two times larger adhesion force than that of Sylgard fibers for a purely vertical pull-off (θ=90° or drag displacement = 0). After the pull-angle decreases to 50°, the adhesion circle of SEBS fibers also shifts to one side, as shown in Figure 5a. However, the magnitude of the shift is much smaller than that of


Page 12 of 21

Page 13 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Sylgard 184 fibers, which appears to make this particular material less inherently directional. Here, the ratio of the maximum to minimum adhesion force is approximately 3. Meanwhile, although the SEBS elastomer fibers’ adhesion circle area is bigger than that of the Sylgard 184, the maximum force is nearly the same at this pull-angle. Therefore, for applications that have precise control over the loading conditions of the fibers, Sylgard 184 would likely be a better choice to achieve more anisotropic behavior with all other properties being equal. The one case where the SEBS is outperforming the Sylgard 184 adhesive is that using very low pull-off angles (20°). In this case, the adhesion is quite strong in all directions with the exception of a narrow range of α in the weak direction. This could be desirable in select applications where shear loads may be high and not always well controlled and adhesion needs to be maintained. For the LDP test, 10 and 40 µm drag displacements are chosen as the comparison in Figure 5b. The SEBS fibers also show anisotropic adhesion with these values of drag displacement (10 and 40 µm). However, both the adhesion magnitude and the degree of the directionality are smaller than that of the Sylgard 184 fibers. When the drag displacement is 10 µm, the adhesion circle of SEBS fibers shifts only a little bit in the strong direction, and therefore the degree of the adhesion anisotropy is not as good as the Sylgard 184 fibers. Further increasing the drag displacement to 40 µm, the adhesion force drops dramatically at all values of α and the adhesion circle becomes very small and irregular. Meanwhile, at this condition, the adhesion circle shows similar, if less dramatic, maxima at angles offset to the presumed strong direction as well as local maxima offset to the weak direction which is not seen in the Sylgard 184 fibers because their adhesion force is too low (nearly zero).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 5. Direct comparison of Sylgard 184 and Kraton G1657 SEBS elastomers with the same defective fiber design. For both angled pull-offs test (a) and LDP type test (b), the Sylgard 184 outperforms in anisotropy ratios, and in the case of drag displacement, maximum adhesion strength. As shown in Figure 5b, for low drag displacements, the SEBS adhesive has a much wider range of α which is close to the maximum value, (~130°) compared with Sylgard 184 (~55°). This wide range of maximum adhesion values comes at a cost of the lowest adhesion values being nearly the same as those of Sylgard 184 when not loaded in any preferential direction, while the Sylgard 184 adhesive has a negligible adhesion strength at any angle in the negative orientation. More significantly, with larger shear displacements, the Sylgard 184 adhesive increases in strength beyond that of the SEBS and does so at a wider range of α than before (~75°). The complete data indicates that Sylgard may have a higher maximum adhesion strength, but is far more sensitive to the introduction of a defect than the SEBS material. In summary, both Sylgard 184 and SEBS with line-defect caps show anisotropic adhesion with the pull-angle or the drag displacement as a control variable. The adhesion circle can not only provide the adhesion magnitude but also give more information about the adhesion directionality and azimuth


Page 14 of 21

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in a whole plane. Therefore, the adhesion circle approach can provide more comprehensive and useful data for practical applications such as climbing robots, mechanical transportation arm, etc. 4.4 Improving performance What is seen for these adhesion circle tests is that adhesion performance and directionality is strongly influenced by the exact means of pull-off trajectory, and adhesive structural material when the geometric design is the same. Through comparing different structural materials, manufacturing processes and geometries of adhesives through this adhesion circle test, more details are revealed about performance than single axis anisotropic tests can provide. One of the biggest surprises was the particular shape of the adhesion circle profile of the fibers with LDP type tests, where the maximum adhesion was not in the direction of the presumed strongest axis, but had two maxima at angles corresponding to the corners of the caps. While previous work had indicated that the size of the overhang of the cap makes a large difference in resistance of mushroom shaped fibers to off-axis loads22, this particular profile wasn’t expected because other work loading a mushroom shaped cap to a sharp corner showed a lower adhesion value than the same loads to a wider side8. In the case of the circular type fibers, the particular fabrication process using an uncollimated exposure of acrylic tends to undercut the fiber more in the direction of the larger gaps between fibers, resulting in an effective overhang that is larger in those directions. As a result, we suspect that for the LDP tests of those materials a similar effect is occurring where the fibers can simply withstand a higher shear force/moment while being preloaded so as to maintain full cap contact longer during these trials. An interesting consequence of seeing the anisotropic performance presented in this manner is that it raises the question about what would the “ideal” dry adhesive look like when tested in a similar manner. There are two possibilities (as shown in Figure 6a): one would be an adhesive that reaches its maximum adhesion force at any value of α between 180° and 360° and zero adhesion at all other values of α. Such



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

an adhesive would be a classical shear-activated design with maximum robustness to off-axis loads, and the adhesion circle profile would look like a “D”. Another possibility is that it may be feasible to enhance the adhesion performance seen in a few trials (Figure 4a, Figure 5a) in which the profile broadly looks like a “heart”. In such a case, the adhesion strength in the absence of shear can be relatively high, and enhanced if loaded in the strong side of the adhesive design, but only a very narrow range of α displacements will permit an easy release. This is in fact opposite to the gecko, where the default state of its foot hairs is non-sticky and they need to be loaded in a specific manner to adhere. Regardless, this particular type of anisotropic adhesive may be useful in a variety of applications where easy removal is not as desirable as a high strength default state. It is important to note that these particular fiber designs were not optimized for this particular behavior in mind, but future modifications can be made to the shape, size and defect placement with the goal to maximize performance for either of the two types of anisotropic behavior that is desired. To compare the anisotropic performance to a practical maximum for the specific geometry and structural material of a dry adhesive, we also tested the circular fibers from an area of the mold that did not have the defect introduced, as shown in Figure 6b. The results were surprising based on our earlier findings28 that SEBS fibers show higher adhesion than Sylgard 184, but in this specific case, the maximum adhesion force of the Sylgard 184 fibers (~80mN) was higher than that of the SEBS (~40mN), despite the overall lower surface energy and viscoelasticity of the structural material. In fact, this particular fiber design was one of the strongest we have tested to date reaching over 80mN against a 6mm hemisphere indenter at a 5mN preload, and suspect that the primary improvement was due to better surface roughness of the cap via the original mold, or subsequent replication processes. What the data shows is that the maximum possible adhesion of the Sylgard 184 fibers could be much higher than that of the SEBS, but at the same time, it is much more vulnerable to any imperfections in the contact area. As a result, to achieve best directionality with this material, we would have to introduce a much smaller defect, if all other properties/geometries


Page 16 of 21

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


remained the same. Such information will be very useful in the design and optimization of future anisotropic dry adhesives using Sylgard 184 or alternative structural materials.

Figure 6. (a) The adhesion circle with an “Ideal” pattern; (b) Comparison of the adhesion force between intact-cap and defect-cap for two materials at a 90° pull-angle test without any drag displacements. 5. Conclusion We demonstrate the usefulness of an adhesion circle approach to testing the performance of a variety of directionally sensitive fiber designs. If these materials are compared in only a single strong-weak axis, substantial performance differences may be missed and a more complete picture emerges when samples are examined for their behavior in radial orientations. For example, these fibers show a different maximum adhesion orientation at large shear displacements than the designed strong axis, but their performance reverts back to more expected behaviors at lower shear displacements. This off-axis maximum adhesion strength is not seen however if the samples are tested by simply pulling at an angle off of the normal direction rather than applying shear displacements first. This influences the types of systems that could use these particular anisotropic adhesives, because if very precise control over trajectory is possible, then substantial adhesive performance can be maintained while still permitting a low force removal of the adhesive. Other behavior or improved performance may yet be seen by the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

introduction of a shear displacement followed by an angled pull-off, but these tests will be saved for a future work. The current test system is designed to only need very small adhesive samples, and can function with a single axis high precision load cell, making it more accessible to other researchers in the field. The code as written can take inputs from a two-axis load cell if needed and could provide data in 3 axes at once with minor modification so it is feasible to introduce new functionality in future work and help optimize adhesive not just for maximum adhesion strength but minimum shear loads to engage the fibers. Because significant insight can be gained into the adhesive performance by adding off-axis shear displacements of anisotropic adhesive designs, the adhesion circle can be a valuable tool to guide future research into directionally sensitive gecko-inspired adhesives. Supporting Information The Supporting Information is available free of charge on the ACS Publications website. A detailed description of the fabrication of microfibers with offset-caps and the fabrication of microfibers with defect-caps; Microscope images of the array of fibers with offset-cap and defect-cap and some important geometry dimensions.

Acknowledgement The authors would like to thank NSERC, CMC Microsystems and the China Scholarship Council for financial support of the work presented.

References (1) Autumn, K.; Sitti, M.; Liang, Y. A.; Peattie, A. M.; Hansen, W. R.; Sponberg, S.; Kenny, T. W.; Fearing, R.; Israelachvili, J. N.; Full, R. J. Evidence for van der Waals Adhesion in Gecko Setae. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 12252-12256. (2) Autumn, K. In Properties, Principles, and Parameters of the Gecko Adhesive System. Biological adhesives; Springer: 2006; pp 225-256.


Page 18 of 21

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(3) Aksak, B.; Murphy, M. P.; Sitti, M. Adhesion of Biologically Inspired Vertical and Angled Polymer Microfiber Arrays. Langmuir 2007, 23, 3322-3332. (4) Lee, J.; Fearing, R. S.; Komvopoulos, K. Directional Adhesion of Gecko-Inspired Angled Microfiber Arrays. Appl. Phys. Lett. 2008, 93, 191910. (5) Jeong, H. E.; Lee, J. K.; Kim, H. N.; Moon, S. H.; Suh, K. Y. A Nontransferring Dry Adhesive with Hierarchical Polymer Nanohairs. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 5639-5644. (6) Kim, T.; Jeong, H. E.; Suh, K. Y.; Lee, H. H. Stooped Nanohairs: Geometry-Controllable, Unidirectional, Reversible, and Robust Gecko-like Dry Adhesive. Adv Mater 2009, 21, 2276-2281. (7) Murphy, M. P.; Aksak, B.; Sitti, M. Gecko-Inspired Directional and Controllable Adhesion. Small 2009, 5, 170-175. (8) Kwak, M. K.; Jeong, H. E.; Bae, W. G.; Jung, H.; Suh, K. Y. Anisotropic Adhesion Properties of Triangular-Tip-Shaped Micropillars. Small 2011, 7, 2296-2300. (9) Tamelier, J.; Chary, S.; Turner, K. L. Vertical Anisotropic Microfibers for a Gecko-Inspired Adhesive. Langmuir 2012, 28, 8746-8752. (10) Del Campo, A.; Greiner, C.; Arzt, E. Contact Shape Controls Adhesion of Bioinspired Fibrillar Surfaces. Langmuir 2007, 23, 10235-10243. (11) Sameoto, D.; Menon, C. Direct Molding of Dry Adhesives with Anisotropic Peel Strength Using an Offset Lift-off Photoresist Mold. J. Micromech. Microeng. 2009, 19, 115026. (12) Khaled, W. B.; Sameoto, D. Anisotropic Dry Adhesive via Cap Defects. Bioinspiration Biomimetics 2013, 8, 044002. (13) Xue, L.; Pham, J. T.; Iturri, J.; del Campo, A. Stick–Slip Friction of PDMS Surfaces for Bioinspired Adhesives. Langmuir 2016, 32, 2428-2435. (14) Parness, A.; Soto, D.; Esparza, N.; Gravish, N.; Wilkinson, M.; Autumn, K.; Cutkosky, M. A Microfabricated Wedge-Shaped Adhesive Array Displaying Gecko-Like Dynamic Adhesion, Directionality and Long Lifetime. J. R. Soc. Interface 2009, 6, 1223-1232. (15) Day, P.; Eason, E. V.; Esparza, N.; Christensen, D.; Cutkosky, M. Microwedge Machining for the Manufacture of Directional Dry Adhesives. Journal of Micro and Nano-Manufacturing 2013, 1, 011001. (16) Parness, A.; Hilgendorf, T.; Daniel, P.; Frost, M.; White, V.; Kennedy, B. Controllable on-off Adhesion for Earth Orbit Grappling Applications. Aerospace Conference, 2013 IEEE. 2013, 1-11. (17) Autumn, K.; Majidi, C.; Groff, R. E.; Dittmore, A.; Fearing, R. Effective Elastic Modulus of Isolated Gecko Setal Arrays. J. Exp. Biol. 2006, 209, 3558-3568.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(18) Gravish, N.; Wilkinson, M.; Autumn, K. Frictional and Elastic Energy in Gecko Adhesive Detachment. J. R. Soc. Interface 2008, 5, 339-348. (19) Schubert, B.; Lee, J.; Majidi, C.; Fearing, R. S. Sliding-Induced Adhesion of Stiff Polymer Microfibre Arrays. II. Microscale behaviour. J. R. Soc. Interface 2008, 5, 845-853. (20) Chary, S.; Tamelier, J.; Turner, K. A Microfabricated Gecko-Inspired Controllable and Reusable Dry Adhesive. Smart Mater. Struct. 2013, 22, 025013. (21) Tamelier, J.; Chary, S.; Turner, K. L. Importance of Loading and Unloading Procedures for GeckoInspired Controllable Adhesives. Langmuir 2013, 29, 10881-10890. (22) Sameoto, D.; Sharif, H.; Díaz Téllez, J. P.; Ferguson, B.; Menon, C. Nonangled Anisotropic Elastomeric Dry Adhesives with Tailorable Normal Adhesion Strength and High Directionality. J. Adhes. Sci. Technol. 2014, 28, 354-366. (23) Das, S.; Cadirov, N.; Chary, S.; Kaufman, Y.; Hogan, J.; Turner, K. L.; Israelachvili, J. N. Stick-slip Friction of Gecko-Mimetic Flaps on Smooth and Rough Surfaces. J. R. Soc. Interface 2015, 12, 20141346. (24) Santos, D.; Spenko, M.; Parness, A.; Kim, S.; Cutkosky, M. Directional Adhesion for Climbing: Theoretical and Practical Considerations. J. Adhes. Sci. Technol. 2007, 21, 1317-1341. (25) Wang, Y.; Tian, H.; Shao, J.; Sameoto, D.; Li, X.; Wang, L.; Hu, H.; Ding, Y.; Lu, B. Switchable Dry Adhesion with Step-Like Micropillars and Controllable Interfacial Contact. ACS Appl. Mater. Interfaces 2016, 8, 10029-10037. (26) Khaled, W. B.; Sameoto, D. Fabrication and Characterization of Thermoplastic Elastomer Dry Adhesives with High Strength and Low Contamination. ACS Appl. Mater. Interfaces 2014, 6, 6806-6815. (27) Manca, M.; Cortese, B.; Viola, I.; Aricò, A. S.; Cingolani, R.; Gigli, G. Influence of Chemistry and Topology Effects on Superhydrophobic CF4-Plasma-Treated Poly (dimethylsiloxane)(PDMS). Langmuir 2008, 24, 1833-1843. (28) Soltannia, B.; Sameoto, D. Strong, Reversible Underwater Adhesion via Gecko-Inspired Hydrophobic Fibers. ACS Appl. Mater. Interfaces 2014, 6, 21995-22003.


Page 20 of 21

Page 21 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC


Adhesion Circle: A New Approach To Better ... - ACS Publications

Recommend Documents