«UNIVERSITY OF CALIFORNIA Santa Barbara A Micro/Nano-Fabricated Gecko-Inspired Reversible Adhesive A Dissertation submitted in partial satisfaction of ...»
However, with the millimeter scale of the aluminum sphere used in testing only one structure is likely being tested and the compliance of the platform fingers are likely the dominant mechanism of conformal enhancement. As the aluminum sphere contacts the platform, the radial extending fingers are able to bend up and contact the aluminum surface. This produces a near linear increase in adhesion with normal load. At some point, however, the finger (or fingers) is completely in contact and can provide no further adhesion. Adhesion then becomes solely dependent on squashing the organorods in the center of the structure (Fig. 8), and little increase in adhesion is seen thereafter.
To better mimic the structure and properties of the gecko tarsus a conformal hydrophobic fluorocarbon coating was deposited on the organorods. With the addition of the coating contact angle measurements show a change from hydrophilic,
42.5° ± 2, to highly-hydrophobic, 145° ± 2 (compared to 160.9° ± 1.4 for the gecko (7)), behavior with the fluorocarbon coating (Fig. 4). This change from hydrophilic to hydrophobic behavior is significant not only in better mimicking the gecko adhesive structure, but may also shed new light on the purpose of the hydrophobic property. By testing the artificial adhesive before and after the addition of the hydrophobic coating, a comparison between hydrophilic and hydrophobic surface adhesion can be made. While with a gecko, it would be non-trivial to change the hydrophobic nature to perform a similar experiment. Surprisingly, adding a hydrophobic coating to the organorod surface improved adhesion (Fig. 8). As will be discussed later, this result may be associated with other factors than just the change in surface composition, e.g. increased organorod size. Adhesion was further increased when the hydrophobic organorods were combined with the compliant platform structures. Improved adhesion with the platforms may be a result of two mechanisms; one that the fingers are improving meso-scale surface conformation and the other because the fingers are able to bend out of plane changing the adhesion vector to include a transverse component Although the adhesion mechanism between the rough aluminum sphere and the organorod coated surfaces is quite complex – and many basic theoretical aspects of adhesion not fully understood – a simplified approach was taken to explain the experimental adhesion measurements. Two simple models based on JohnsonKendall-Roberts (JKR) contact adhesion theory(63) and van der Waals interaction forces(64) were used to capture the basic trend in adhesion seen. JKR theory for a sphere contacting a flat surface predicts a pull-off force of F= (3/2) RW, where R is the radius of the sphere and W is the work of adhesion, taken to be √γAl γpolymer, where γAl = 500 mJ/m2 and γpolymer = 35 mJ/m2(65, 66). For the organorod surface the radius of curvature of the rods is considered to be very small compared to the sphere (making the organorods the spherical contact surface and the aluminum sphere a flat surface) and the adhesion force estimated to be: FJKR= n(3/2)RoW, where Ro is the diameter of the organorods (determined using scanning electron microscopy), and n is the number of organorods in contact with the surface – equal to the surface density of organorods multiplied by the contact area (determined using JKR contact mechanics for a sphere contacting a flat surface(63) and verified using scanning electron microscopy).
Alternately, considering only the van der Waals contribution, the force is taken to be FvdW = HRo/6do2, where H is the Hamaker constant (estimated to be H =
2.1x10-21 W = 2.8x10-19 J)(64), Ro is the diameter of the organorods and do is a characteristic cut-off distance equal to 0.165 nm(64). As can be seen in Figure 6, the JKR method tends toward the upper limit of values obtained in the rough aluminum sphere adhesion test (overestimate of actual number of organorods contributing to adhesion), while the van der Waals method tends toward the lower limit (not accounting for other interaction forces, e.g. capillary forces).
Increased adhesion with the compliant platforms (Fig. 6) can be attributed to greater surface contact as a result of enhanced surface conformation, primarily due to the flexible platform fingers. The additional contact area of the fingers was assumed to increase linearly with applied load and included in the van der Waals model from above. The new model, for the case of 2 and 4 fingers adhering, was compared with the experimental data for the organorod covered flexible platforms (Fig. 6). Although there is a considerable amount of scatter, due in part to the nature of this “real life” rough sphere test, there is good agreement between the predicted adhesion for all four fingers in contact and the maximum measured adhesion strengths. This also helps to understand the origin of much of the scatter. On every successive indent it is possible that not all the platform fingers are making contact – especially considering the roughness of the indenter tip. While this introduces a variance in the data and makes modeling tricky, from an application standpoint larger arrays will produce a conglomerate adhesive strength.
Addition of the hydrophobic coating to the organorod surface increased adhesion. It was expected that the decrease in surface energy with the hydrophobic coating would reduce the adhesion. The hydrophobic coating on the organorods increased the diameter of the organorods and changed the surface composition.
Considering these two factors the van der Waals model was modified, with Ro = 350 nm and γpolymer = 19 mJ/m2(65, 66), and plotted against the experimental data (Fig.
7). While the model predicts a slight increase in adhesion over the smaller organorods, it greatly underestimates the adhesion of the hydrophobic organorods.
Further investigation using scanning electron microscopy revealed that the hydrophobic organorods tended to change conformation under applied pressure effectively changing their radius of curvature. Both JKR and van der Waals models were modified to account for the increased size of the organorods (Fig. 7, lines b and c).
While the JKR model offered a better prediction, both models failed to capture the experimental data trend. Upon examination the van der Waals equation has a strong dependence on do, the estimated cut-off distance – basically determining the range of interaction of van der Waals forces. Reducing the cut-off distance, do, from 0.165 nm to 0.09 nm the van der Waals model better represents the experimental data (Fig. 7 line d). This suggests that the hydrophobic coating may enhance the van der Waals adhesive force by decreasing the surface to surface interaction distance.
A composite model for the hydrophobic organorods on the flexible platforms was made by accounting for the compliant fingers in the van der Waals hydrophobic organorod model, with do=0.09 nm. Comparing this model with the experimental data, excellent agreement is seen (Fig. 7, line e). The slight lag in the experimental curve can be attributed to the delay in finger attachment (the sphere must press into the center of the platform before it becomes sufficiently close to interact with the extended fingers). Note that the highest adhesion attainable experimentally, 462 μN, corresponds to 1 complete finger attachment (the end of line e in figure 7) and a predicted adhesion of 485 μN.
Some simple models to explain a complex adhesion phenomenon have been presented. Although simple, the models do show reasonable agreement with the empirical data, although the testing was initially designed to produce only relative adhesion strengths between the biomimetic adhesives and a rough aluminum sphere.
The rough sphere testing introduces a significant amount of scatter to the data making analytical comparisons tenuous. However, the relative improvement in adhesion with the hydrophobic surface over the hydrophilic surface is an intriguing result. To ascertain the mechanism of this adhesion improvement, and to support the validity of the testing, the nanoindenter testing should be repeated with a smooth and more ideal indenter surface – e.g. a polished flat punch tip. This would certainly help to eliminate much of the scatter and remove the adhesion strength dependence on applied load, as is seen with the spherical tip.
Arrays of flexible silicon dioxide platforms supported by single high-aspect-ratio silicon pillars have been fabricated. These platforms, when coated with polymeric organorods, show a significant increase in adhesion over solid organorod covered substrates. Further improvement in adhesion is measured with the addition of a highly-hydrophobic organorod surface – possibly enhancing van der Waals interactions. This indicates that the super-hydrophobic nature of the gecko pad may serve to improve adhesion, instead of just aiding in self cleaning and wear characteristics. Future work on the synthetic adhesive will focus on optimization of the structure, more ideal testing for theoretical purposes, and scaling up fabrication for macro-scale testing and device integration.
Acknowledgements We thank Marco Aimi, Masa Rao and Brian Thibeault for their valuable processing conversations; Laurent Pelletier and Michael Requa for their constructive comments on previous drafts of this document. Funding was provided by AFOSR Contract Number FA9550-05-1-0045 and UCSB. Fabrication was carried out at Nanotech Nanofabrication Facility at UC Santa Barbara, part of the National Nanotechnology Infrastructure Network.
Correspondence should be addressed to M.T.N. (e-mail:
Figure V-1 Electron micrographs of silicon dioxide platforms supported by single slender pillars and coated with polymeric organorods. The structures are batch fabricated using standard bulk micro fabrication techniques requiring only a single lithography step. a, Array of four-fingered platform structures.
Scale bar, 200 µm. b, Profile view of a organorod coated silicon dioxide platform supported by a single crystal silicon pillar. Scale bar, 20 µm. c, Magnified view of the multi-scale system. Scale bar, 20 µm. The integration of the ~200 nm diameter organorods with the 150 µm platform structures provides enhanced surface conformation. d, High resolution electron micrograph of the edge of an organorod coated finger. Scale bar, 5 µm.
Figure V-2 Silicon dioxide platform supported by a single high-aspect-ratio pillar etched out of single crystal silicon using a modified micro fabrication technique. Scale bar 50 μm.
Figure V-3 Organic looking polymeric nanorods, “organorods”, fabricated utilizing the instability of a dielectric in a large electric field. The ~100 nm features were achieved using a bilayer to reduce the interfacial energy, thus reducing the instability wavelength and corresponding feature size. Scale bars, 10 and 2 μm, left and right respectively.
Figure V-4 Comparison of hydrophilic and hydrophobic surfaces. a, (top) Water droplet partially wetting a hydrophilic organorod surface with a contact angle of 42.5° ± 2 (top). (bottom) Electron micrograph of ~120 nm diameter organic looking nanorods, “organorods”. Scale bar, 5 µm. b, (top) Water droplet resting on top of the highly-hydrophobic organorod surface with a contact angle of 145° ± 2. (bottom) Electron micrograph shows the increased diameter of the organorods to ~350 nm from the fluorocarbon hydrophobic coating applied through plasma deposition. Scale bar, 5 µm.
Figure V-5 Typical nanoindenter adhesion test results for a organorod covered structure indented with a semi-rough 3.175 diameter spherical aluminum tip. The plot represents a typical load vs. displacement curve illustrating loading, unloading, and pull-off adhesion.
Figure V-6 Nanoindenter adhesion testing results and theoretical models for hydrophilic polymeric surfaces. Test surfaces were indented with a semirough 3.175 diameter spherical aluminum tip. Increased normal loads created more surface contact between the spherical tip and surfaces increasing adhesion. The organorod morphology enhanced surface adhesion over the solid photoresist surface. Adhesion was further enhanced by integration of organorods with flexible silicon dioxide platforms, large data scatter was a biproduct of testing meso-scale adhesion with a rough surface. a, van der Waals and b, Johnson-Kendal-Roberts (JKR) adhesion models predicting the collective adhesion of organorods over a given contact area, predicted by JKR contact mechanics for a sphere and flat surface. c and d, Modified van der Waals adhesion accounting for increased contact area attributed to conformation of platform “fingers” to the test sphere; c, two fingers. d, four fingers. The saturation point corresponds to complete finger attachment, in the case of four fingers (d) this value nicely bounds the experimental data.
Figure V-7 Nanoindenter adhesion testing results and theoretical models for hydrophobic polymeric surfaces. Surface adhesion between the sphere and test surfaces is enhanced by a hydrophobic coating on the organorods. The coating increases the size from ~120 nm to ~350nm and reduces the surface energy. a, van der Waals interaction energy model predicts an increase in adhesion with the larger organorods, but fails to capture the trend of the hydrophobic organorods. b, van der Waals and c, JKR models compensating for an increased radius of curvature- determined by scanning electron micrscopy- of organorods squashed during indentation show better agreement with experimental data. d, Reduction of the interaction distance, do, from 0.165 nm to 0.09 nm in the van der Waals model. e, Combining the modified van der Waals models for hydrophobic organorods (do=0.09 nm) and the compliant platform model with 1 finger, excellent agreement with the meso-scale adhesion test is seen. Note the terminus of dotted line represents 1 complete finger adhesion.
Figure V-8 Scanning electron micrographs showing squashed organorods in the central portion of a structure. Scale bars, 20 and 10 μm, left and right respectively.
VI. Batch Fabrication and Characterization of Nanostructures for Enhanced Adhesion