«UNIVERSITY OF CALIFORNIA Santa Barbara Design and Characterization of Fibrillar Adhesives A Dissertation submitted in partial satisfaction of the ...»
7 with (a) showing a front view of the vertical ﬂaps and (b) showing the side view of the angled ﬂaps. The important geometric properties of the vertical and angled ﬂaps are shown in Table 3.1. It should be noted that the vertical ﬂaps are taller by 4.3 µm, thicker by 0.4 µm, have twice the density of the angled ﬂaps, and have a smooth top face. When comparing force values between the two types of ﬂaps, the differences in these geometric properties should be kept in mind. The diﬀerences between the two sets of rectangular ﬂaps are a result of the diﬀerent fabrication processes used to create the adhesive.
Chapter 3. Friction and Adhesion Tester 3.
3.2 Experimental Methods Microfabrication To create the negative mold for vertical ﬂaps, a 0.5 µm layer of silicon dioxide was ﬁrst deposited onto a single crystal silicon wafer using a plasma-enhanced chemical vapor deposition (PECVD) system (Plasma-Therm, St. Petersburg, FL).
An image reversal photoresist (AZ Electronic Materials, Branchburg, NJ) with a thickness of approximately 1 µm was spin coated over the wafer. Using an i-line stepper (GCA, Refurbished by RTS Technologies, Doylestown, PA) a rectangle with a length of 4 µm and a width of 10 µm was photolithographically deﬁned in the resist. After developing, the rectangular voids allowed the silicon dioxide to be etched away with a CHF3 plasma using an inductively coupled plasma reactive ion etching (ICP-RIE) system (Panasonic Factory Solutions, Japan). Once the silicon dioxide has been completely etched away, the photoresist was then stripped away using a stripping solution (Shipley Company, Marlborough, MA). To increase the number of ﬂaps, the following steps can be repeated multiple times: the application of photoresist, the exposure of photoresist in the stepper, the etching of the silicon dioxide, and the stripping of the photoresist. For the angled ﬂaps, only one additional set of ﬂaps was created. The new set of rectangles was oﬀset to ﬁll the gaps left by the ﬁrst generation. The exposed silicon was then vertically Chapter 3. Friction and Adhesion Tester etched 15 µm to create rectangular slots using a Bosch deep reactive-ion etching (DRIE) system (Plasma-therm, St. Petersburg, FL).
To facilitate removal of the polymer from the negative silicon mold, a layer of silane was vapor deposited on the silicon surface. PDMS (Sylgard 184) (Dow Corning, Midland, MI), mixed in a 1:10 ratio of curing agent to base elastomer, was then poured onto the negative mold. The polymer was degassed until no air bubbles were visible and then cured at 100◦ C for 10 minutes. The polymer and silicon were separated by hand, leaving cm-sized adhesive patches and a reusable negative mold.
The procedure used for the creation of vertical ﬂaps could not be reused for angled structures due to the vertical etch and physical characteristics of the Bosch etcher used. A new fabrication scheme was therefore designed. An angled lithography process was implemented for its ability to create angled structures with similar dimensions to the vertical ﬂaps over cm-scale areas. Since a silicon wafer is a reﬂective, glass was chosen as a suitable replacement for the substrate material. The glass wafer was coated with a photoresist bilayer consisting of a 10 µm thick layer of polymethylglutarimide (PMGI) positive tone photoresist (MicroChem Corp., Newton, MA) topped with a 1.4 µm thick layer of image reversal photoresist (AZ Electronic Materials, Branchburg, NJ). Since the materials are sensitive to diﬀerent wavelengths, the photoresist was used to deﬁne the shape Chapter 3. Friction and Adhesion Tester of the features using the i-line stepper and after development, the pattern was then be transferred into PMGI layer using a deep ultraviolet (DUV) ﬂood exposure system (A B Manufacturing, San Jose, CA). The glass sample was mounted at an angle given by Snell’s Law for refraction at the PMGI and air interface with nair =1.00, nP M GI =1.54, and the desired angle in the PMGI. Areas of PMGI without photoresist blocking them are then removed using a developing solution (MicroChem Corp., Newton, MA).
To prepare the photoresist sample for molding, it is ﬁrst exposed to an oxygen plasma for two minutes and then a thin silane layer was deposited to facilitate separation. PDMS molding was performed in the same manner as the vertical ﬂaps with the exception that the two materials were separated by peeling with the angle of the tilt to avoid breaking the ﬂaps. Original angled ﬂaps had poor shape with curvature on all faces of the ﬂap and a small height-to-thickness aspect ratio. Reﬁnement to the shape of the ﬂaps, due to less unwanted removal of the PMGI, was achieved by splitting the exposure and development times of the PMGI into eight equal parts.
Adhesion and Friction Testing Testing was performed using the Bio-F home-built friction and adhesion tester with a four-millimeter diameter ﬂat glass puck as the opposing surface (Section Chapter 3. Friction and Adhesion Tester 3.
1). Movement of the sample was generated by two motorized linear stages (Newport Corporation, Irvine, CA) with incremental movements of 100 nm in the vertical direction and 500 nm in the horizontal direction. Force sensing was provided by a load cell (ATI Industrial, Apex, NC) with a eﬀective force resolution of approximately 3 mN. Position, force, and torque data was recorded during the adhesive characterization.
During testing, the sample was moved towards the glass surface at a vertical speed of 1 µm/s until a given compressive load, the preload, was reached. Vertical retraction speeds remained the same during separation of the adhesive from the glass puck. Any in-plane movement, such as during the drag/shear stage of testing, was performed at a horizontal speed of 3 µm/s in a direction perpendicular to the large face of the ﬂap. The +Y-direction generated movement with the angle of the ﬂap while the −Y-direction was against the angle of tilt.
3.3.3 Results and Discussion Adhesion and friction testing on the structures revealed remarkably diﬀerent behavior between the two designs. As can be seen in Figure 3.8, the vertical ﬂaps were able to create roughly 60 mN of adhesive force above preloads of 85 mN.
Between 25 and 60 mN preloads, maximum contact area had yet to be achieved across the entire puck, as can be seen by the increasing adhesion values. Below Chapter 3. Friction and Adhesion Tester Figure 3.
8: The angled ﬂaps had a non-sticky default state, an important geckoinspired adhesive property, for preloads under 175 mN. Above 175 mN preloads, the pure adhesion forces gradually increased but never reached the level of the vertical ﬂaps. The vertical ﬂaps had large pure adhesion forces for all but the smallest preloads. The roughness on the top of the angled ﬂaps was believed to be responsible for the lack of pure adhesion forces.
25 mN preloads, only small areas of contact have been created which resulted in negligible adhesion values under the minimum detectable force of the load cell.
After the adhesion values plateaued due to full contact across the test area, the values started to fall at preloads above 193 mN. This slight drop with increasing preloads can be explained as a result of buckling the ﬂaps which produced a reduction in contact area over a portion of the sample.
Chapter 3. Friction and Adhesion Tester Figure 3.
9: Long lifetimes are an important characteristic of synthetic adhesives which will be used in commercial and industrial applications. The pure adhesion force was not observed to signiﬁcantly decrease when load-pull endurance tests were repeated 50 times on the vertical ﬂaps.
The angled ﬂaps did not respond in the same manner to load-pull tests.
Preloads under 200 mN did not produce signiﬁcant amounts of adhesion. Only when the force was above this value were small values observed. Surface roughness on the top of the ﬂap, possibly in combination with tilt, was likely responsible for the small adhesion values seen during load-pull tests. While no adhesion may appear to be a poor result, it is in fact one of the key properties identiﬁed by Autumn  for a gecko-like adhesive. This non-sticky default state at small preloads allowed for the possibility to control the adhesive and only stick when desired.
Chapter 3. Friction and Adhesion Tester The adhesive’s lifetime is an important property for synthetic adhesives, requiring force retention and repeatability throughout its use.
Some adhesives degrade quite quickly with less than 20 cycles signiﬁcantly reducing the forces supported . As an initial test, the vertical ﬂaps were tested for durability by repeating the load-pull test 50 times. The maximum vertical preload tested (80 mN) was used to ensure full contact over the entire test area. The results from the durability test can be seen in Figure 3.9. Pure adhesion values were very stable across all tests with an average value of 48.5 mN. The maximum and minimum pure adhesion values only varied by only 2.1 mN across all tests, further showing the adhesive’s high repeatability.
To discern the eﬀect of shearing, load-drag-pull tests were performed on both the vertical ﬂap and angled ﬂaps. The tests aimed to show that shearing would cause a contact area increase with the glass, from the ﬂap tops to the large side or angled faces of the ﬂaps. The vertical ﬂaps were tested at preloads representative of partial and full contact, 45 and 80 mN, and the shear adhesion values can be seen in Figure 3.10. At zero shear length, the test was exactly the same as the load-pull test and adhesion values were very similar. As the shear lengths increase, the ﬂaps were bent until at a shear length around 12.5 µm, only the edge between the top and side face of the ﬂap was in contact. In this orientation, only small loads could be supported. Contact then increased on the large face Chapter 3. Friction and Adhesion Tester Figure 3.
10: Shear adhesion tests were performed on the vertical ﬂaps at preloads of 45 mN, representing partial contact over the glass puck area, and 80 mN, representing full contact over the glass puck area. Shearing the ﬂaps in either direction increased the shear adhesion force due to higher contact area on the side face of the rectangular ﬂap. Misalignment of the adhesive or glass puck are responsible for the lack of symmetry about zero shear length.
Chapter 3. Friction and Adhesion Tester until a maximum contact area was reached at a shear length of approximately 90 µm.
As shear lengths increased from 12.5 to 90 µm, random stick-slip events occur with individual ﬂaps since the 10–15 µm tall ﬂaps cannot extend over these large distances. The maximum shear adhesion value represents the switch from when larger shear lengths no longer caused suﬃcient increases in contact area to counteract the loss of forces due to the stick-slip events. The maximum shear adhesion values of 86 and 53 mN for the 80 and 45 mN preloads, respectively, show that higher forces than those obtained with the top face were possible. In the positive direction, the highest value is 1.4 and 3.3 times the value created without any shear for 80 and 45 mN preloads, respectively. Conﬁrmation of higher contact areas were found in optical images obtained during shearing. After the force maximum, the adhesion forces supported slowly fall to a steady state value as ﬂaps gained and lost contact with the glass surface. It should be noted that the results for positive and negative shear lengths should be symmetric and small misalignment of the sample and/or glass puck was likely responsible for the data skew.
The same load-drag-pull test was then performed on ﬂaps with a 20◦ angle from the vertical to discover the inﬂuence of tilt addition on shear adhesion and shear forces. The results for shear adhesion can be seen in Figure 3.11 and clearly display strong anisotropic shear adhesion properties. When sheared against the Chapter 3. Friction and Adhesion Tester Figure 3.
11: Shear adhesion tests were performed on the angled ﬂaps at preloads of 40, 60, and 100 mN. The shear adhesion forces were highly anisotropic. When shearing in the positive direction (with the direction of tilt), high shear adhesion was seen due to contact with the upward facing side face. When shearing in the negative direction (against the direction of tilt), no shear adhesion was observed.
The angled rectangular ﬂaps demonstrated how adhesion can be controlled based on shear direction and shear length.
Chapter 3. Friction and Adhesion Tester angle of tilt, small areas of contact were created as the trailing edge of the ﬂap was dragged across the glass puck.
When sheared with the angle of tilt, increased contact area on the leading face of the ﬂap results in signiﬁcant shear adhesion forces. The maximum shear adhesion values obtained, 46 mN, 42 mN and 28 mN for preloads of 100 mN, 60 mN, and 40 mN, may appear to be small based on the angled ﬂaps values of 86 mN and 53 mN for preloads of 80 mN and 45 mN, but the diﬀerences in density contribute to the discrepancy. The vertical ﬂaps had twice the density due to diﬀerences in fabrication and a larger face available for contact which ultimately resulted in higher force values.
The shear force was also measured during the load-drag-pull tests and the results, shown in Figure 3.12, display additional advantages of tilted structures.
At positive shear lengths, i.e. the gripping direction, for the distances shown, the angled ﬂaps were able to support higher shear forces despite having half the density of ﬂaps and a smaller area on the side face of the ﬂap. In the negative direction, the angled ﬂaps still had higher forces than the vertical ones at shear lengths less than 30 µm.
The addition of tilt to the rectangular ﬂaps when combined with shearing in the tilt direction, engaged high adhesion and high shear forces, while shearing against the tilt direction resulted in low adhesion and friction. This anisotropic behavior of the tilted ﬂaps is qualitatively similar to the anisotropic attachment Chapter 3. Friction and Adhesion Tester Figure 3.