This study used 2-photon three-dimensional lithographic printing and replica molding to fabricate a micro-texture based on the Ocellated Skink. The fabricated surface texture was studied for friction and wear behavior using linear reciprocating tribological tests with a chrome steel ball counterface under various loading conditions and compared with samples fabricated with the same methods without a surface texture. The texture was found to decrease friction at low loads and provide steady friction under all loading conditions. The textured samples also decreased the average wear track width up to 61%. Wear was reduced on surfaces by the texture through the controlled formation of microcracking, which both reduced the amount of debris built up on samples and effectively reduced the worn area.
It has been estimated that about one-fifth of all energy produced annually, 100 million terajoules, is used to overcome friction and wear . This fact alone is all a person needs to understand why controlling friction and wear is so important. Researchers have investigated many ways of controlling friction and wear, although the focus has mostly been on lubrication, using both liquids such as oil [2,3] and solids such as polytetrafluoroethylene (PTFE) [4–6]. Recently, researchers have also been focusing on the effect of micro- and nano-texturing on surface friction and wear properties. In order to do this, some have looked to nature for inspiration .
When it comes to bioinspiration with the goal of controlling wear and friction, studies have investigated all manner of things: lotus texturing for friction control on micro/nano-electromechanical systems devices , toucan beaks, chiton teeth, stomatopod clubs, nacre, fish scales , synovial fluids in mammalian joints, shark skin , and gecko skin [9,10]. Some studies have chosen to concentrate their efforts on reptiles, specifically snakes, lizards, and skinks. This is because reptiles, especially those found in sandy environments, are known for having extremely durable scale-covered surfaces that are also slippery. One skink that has been of particular interest, the Sandfish Lizard, was even found to have lower adhesion forces than PTFE during testing . Multiple studies have focused on making a scale-like surface texture using laser texturing. These do not directly focus on the smaller micro-texture on the surface of the individual reptile scales, but instead, use laser texturing to fabricate 50 µm scale-like features that function as the larger texture would for the animal . The studies showed a reduction of friction up to about 40% when made of steel using pin-on-disc tests , and similar reductions were found when using a variety of other materials . Similar studies have been done investigating scale-like surface textures ranging from 13 to 150 µm diameters .
The Sandfish lizard, or Scincus scincus, is a species of skink native to northern Africa and southwestern Asia. Sandfish is commonly studied because of the method the skinks travel underground through sand, by swimming with an oscillatory motion . The surface of these skinks has also been studied due to the durability of the surface against the sand. However, the studies have mainly focused on the chemistry of the surface and its Beta Keratin layer [15,16]. Although some studies have been performed through direct replication of the surface of the Sandfish, the studies have shown no significant effect from the texture alone on friction and wear properties. However, the studies mostly examined the sliding angle at which the sand dropped on the surface starting to move  and/or limited friction testing of the entirety of the scale, not the microstructure on the scale . The test is not the clearest method for quantifying wear and friction behaviors. This means that the surface texture of Sandfish has not been fully investigated . The surface texture is hierarchical in nature with a wave-like micro-texture that has nanoscale scalloping at the edge of each wave .
From the investigation of the surfaces of various animals with a laser scanning microscope, we found that many other reptiles native to sandy desert biomes also present with a similar texture to Sandfish . One such species is the Chalcides ocellatus or Ocellated Skink. Like Sandfish, it is native to parts of northern Africa and southwestern Asia but can also be found in arid areas in southwestern Europe.
The objective of this study is to isolate the surface texture found on these desert lizards, specifically the Ocellated Skinks, to investigate the effect of the texture on surface wetting and tribological properties. Three-dimensional (3D) nano printing was used as a vehicle to understand the mechanisms of the surface texture on the lizard skin. This was accomplished by fabricating the micro-texture with a 2-photon lithographic 3D printer and then replica molding the printed sample for the quicker fabrication of multiple samples. These samples were tested and compared with samples made the same way without any texturing to analyze differences in friction and wear properties between the samples. The hypothesis was that the textured surface would have more controlled friction and a reduction in wear to the smooth samples.
2 Experimental Methods
2.1 Characterizing Surfaces From Nature.
The surface topographies of various reptiles from desert biomes were investigated using a 3D laser scanning confocal microscope (VK-X260K, Keyence Corporation of America). This type of imaging allows for three-dimensional analysis of the surfaces without damaging the specimens used and can account for the complex surface curvature without needing to remove the skin from the bodies of the animals. The specimens obtained from the University of Arkansas Department of Biology were preserved in formalin, and those from the University of Arkansas Collections Facility were preserved using dry methods. It was determined that neither preservation method would have affected the nature of the surface texture on the scales. Figure 1 shows three images taken at the same magnification from similar areas on an Ocellated Skink, a desert-based Gecko, and a Texas Blind Snake. All of these reptiles have similar wave-like surface textures but with some minor differences: the Ocellated Skink has nano-scalloping on the edges of each wave, the Gecko scales have a simpler texture, but more curvature on the scales as a whole, and the Texas Blind Snake has some pocking evenly spread throughout the linear texture.
The Ocellated Skink shown in Fig. 2 was determined to be the most significant to the study of friction and wear because of the skink’s similarities to the Sandfish Lizard, an animal that also belongs to the family of Scincidae and has been studied extensively before [11,14–19]. The surface texture is only one of the similarities between the two species, and other similarities include body type and native area.
2.2 Sample Fabrication.
A texture was designed in Solidworks based on the surface texture of the Ocellated Skink and was exported as an STL file. To fabricate the designed texture, a lithographic 3D printer (Photonic Professional GT System, Nanoscribe GmbH) was used. The system was chosen for its extremely high resolution. The smallest lattice dimension of 130 nm has been printed using this printer. The laser of the 3D printer operates with a wavelength of 780 nm, a pulse width of 100 fs, and a repetition rate of 80 MHz. This system works through two-photon polymerization of photo-curable resins, creating an ovoid voxel of cured photoresist. When using the 63× lens attachment to the 3D printer, the dimensions of the base ovoid are 0.5 µm in height and 0.3 µm in width. This voxel travels along a path programmed from the STL file as it is sliced in DeScribe, the 3D printer’s import software. The samples were printed with the Nanoscribe company’s proprietary photo resin known as IP-Dip. The substrate used for this 3D nano printing process was a 1 in. square indium tin oxide (ITO) coated glass slide so that the refraction index between the IP-Dip and the glass slide is large enough that the system can record that as the printing interface. A few microliters of IP-Dip photoresist were dropped onto the center of the slide, and the surface texture was printed over an area of 4 mm by 12 mm. For this study, samples with the same surface texture were fabricated multiple times and the textures were measured and found to be consistent across the samples.
After the master sample was printed, it was used to make a polydimethylsiloxane (PDMS) mold for replica molding samples out of the same photoresist for rapid fabrication of multiple samples for this study, Fig. 3 shows this process. The PDMS was cured for 4 h at 60 °C in a fully enclosed mold. Replica molding and curing of the photo resin were carried out under compression for 400 s with a ultraviolet (UV) light source of 280 mW/cm2. The same process was done to make smooth samples by making a PDMS Mold against a plain glass slide instead of one with a 3D printed texture.
2.3 Tribological Testing.
Friction and wear testing were performed with an automatic friction abrasion analyzer (Triboster TS-501, Kyowa Interface Science Co., Ltd.). This tool uses linear reciprocating motions while recording the forces acting on the counterface as it runs across a sample and automatically calculates the coefficient of friction. Tests were completed on smooth and textured surfaces fabricated through replica molding with a 6.3 mm diameter chrome steel ball counterface, as shown in Fig. 4. The linear reciprocating tests were run with 100 repeats. The speed of the tests was 2 mm/s, and the stroke length was 4 mm. Various loads were used for the tests: 20 g, 50 g, and 100 g, which resulted in the maximum Hertzian contact pressure of 35.0 MPa, 47.5 MPa, 59.8 MPa, respectively, calculated from assuming the Young’s modulus of IP-Dip and steel to be 2.6 GPa and 200 GPa, respectively, and the Poisson’s ratio of IP-Dip and steel to be 0.49 and 0.3, respectively.
3 Results and Discussion
3.1 Surface Topography.
Figure 5 shows the laser image of the surface texture of the Ocellated Skink. The average height of the texture in Fig. 5(b) from 25 measurements is about 174 nm with a standard deviation of 27 nm; the average width in between peaks is 3.59 µm with a standard deviation of 0.41 µm. Although there is visible nano-scalloping on the edge of each wave, making it a hierarchical structure, this study focused on mimicking the microscale texturing alone.
The final replica-molded smooth and textured samples had an average polymer thickness of about 130 µm. The average surface roughness, Ra, for the smooth control samples was about 0.13 µm. Figure 6 shows a laser microscope image of a final replica-molded sample texture and a line profile of the texture. The peaks of these waves are about 150 nm tall, with a distance between the peaks of 3 µm. The fabricated texture has more regular patterning than the organic nature of the texture on the surface of the skinks, and it lacks the variation in the dimensions (height and width) of the wave-like texture referenced. The dimensions are also slightly smaller than the average values measured from the skink but are within the range of the realistic values.
There is some tessellation from the process of 3D printing that can be seen in the final sample. The tessellation on the sample is a pattern of 100 µm squares repeating over the course of the sample. This effect exists because a 63× lens was used for the 3D printing in order to have the best printing resolution possible. However, the printing area of the 63× lens was limited to 100 µm2. This requires larger areas needed for tribological testing to be printed one square at a time. The change in focus that occurs between each printed square means that there can be minor differences in height from one printed area to the next. The average value from 40 measurements of the height difference between the adjacent fields of views for the 3D printed surface was about 240 nm, with a standard deviation of 150 nm. The patterning was designed to have each of these height changes aligned with the standard edge of the design so that there was less effect from the height change. The tessellation could also act similarly to the height variations between scales on a reptile’s skin, forming a hierarchical structure that adds to the bioinspiration aspect of the design.
3.2 Surface Wetting Property.
Although dry friction was studied, wettability is still important because, in a humid environment, wettability may affect the adhesion between contacting surfaces through meniscus formation, thus affecting friction and wear. Therefore, water contact angle measurements were taken on the textured and smooth surfaces to determine water’s affinity to the smooth and textured surfaces. Figure 7 shows representative water droplets on the samples. There is not a significant change in water contact angle between the two surfaces. The smooth samples had an average water contact angle of 67.7 deg with a standard deviation of 3.2 deg, and the textured samples had an average water contact angle of 69.1 deg with a standard deviation of 2.6 deg. The printed texture has an extremely small height of 150 nm with a small distance about 5 µm between the textures. The small and sharp nature of the texture resulted in a similar water contact angle to smooth surfaces.
3.3 Frictional Behavior.
Figure 8 shows the friction profiles of the textured and smooth samples during the first ten passes of the 20 g tests. There is clearly more regular and controlled friction in the textured sample data, where the tension is built and released at regular intervals, as compared with the random peaks from the smooth sample. The number of peaks in the COF data is equivalent to the number of squares from tessellation in the worn path. This is because the tessellation behaves in some ways as scales would of a lizard.
The bar chart in Fig. 9 shows average friction coefficients with standard deviation comparing the friction results under various loading conditions for textured and smooth samples. There is a clear reduction of the average value of the coefficient of friction for the 20 and 50 g tests, but not necessarily for the 100 g tests. The plots do show a consistent decrease in the standard deviations and error bars between the smooth and textured samples at each load. The t values (t = 19.5 for the 20 g test, t = 15.9 for the 50-g test, and t = 0.928 for the 100-g test) communicate that, in fact, the change in the mean value for the 20 and 50 g tests are statistically significant but not for the 100 g tests. The reduction of friction for the 20 and 50 g tests but not the 100 g test shows that the texture is more effective at lower loads when the adhesion contribution to friction is more prominent.
To further understand this behavior, the average friction profiles under each loading condition, shown in Fig. 10, can be examined. For the 20 and 50 g loading condition tests, the coefficient of friction is reduced (by 30% and 26%, respectively) for the textured samples when compared with the smooth samples, while this is not the case for the 100 g friction data. For all three loading conditions, the coefficients of friction of the textured samples remain steadier and more consistent than that of the smooth samples throughout the tests. On the smooth sample tested under 20 g load, there is a step change in the coefficient of friction that occurred around the 60th repeat of linear reciprocating motion, after which point damage can be seen on the samples. Similarly, a small step change in the coefficient of friction can also be observed around the 42nd testing cycles on the smooth sample tested under 50 g load. These step changes were from the smoothening of the surface during the testing, generating higher adhesion force between the contacting surfaces. However, throughout the smooth sample tested under 100 g, the coefficient of friction values on the smooth surface behaves erratically. This occurred because of the wear debris generated on the surface. When wear debris is trapped between the counterface and the sample, the friction behavior is no longer based on sliding friction but the rolling friction from the debris. This friction behavior also lowered the average coefficient of friction of the smooth surface shown in Fig. 9.
3.4 Wear Mechanisms.
The effects of texture on the wear of the surfaces after various tests can be seen in Fig. 11. It should be noted that the most significant difference between the wear tracks on the smooth and textured samples is that the average wear track width of the textured samples is significantly narrower than that of the smooth samples. The individual wear track width is defined as follows. First, a 3D optical image of a wear track was obtained. The distance between two points across the wear track was then measured at 20 different locations along the sliding direction. Then the average of these values was calculated and defined as the wear track width. For the 20 g tests, the average wear track of the smooth samples was 176 µm, while that of the textured sample was 69 µm. For the 50 g tests, the average wear track width of the smooth sample was 188 µm, while that of the textured sample was 100 µm. For the 100 g tests, the average wear track width of the smooth sample was 205 µm, while that of the textured sample was 117 µm. While the reduction in wear track width is more prominent at lower loads, it is extremely effective for all loading conditions tested. Although the value of the friction coefficient for the 100 g tests is not significantly lower for the textured samples, the wear reduction is visually apparent and numerically validated as the textured samples have only 57% of the average wear track width of that of the smooth sample. Figure 12 is a bar chart showing these relationships as well as the standard deviations of the data sets, which also shows a reduction in the variation of the data from that of the smooth samples.
Figure 13 contains images from the smooth and textured surfaces at various states of wear and helps to understand the difference in wear formation from both samples. There are two significant differences between the two. First, there is clearly more debris formed, especially in the early stages of wear, on the smooth samples, while on the textured surface, the debris did not begin to form and accumulate until there was already an extreme amount of wear. Second, when looking closely at Fig. 13(e) or Fig. 14, while there is minimal wear on the texture, the formation of many microcracks can be seen in between the waves in the texture. During the formation of these microcracks, stresses built in the surface were released in a controlled manner instead of at random, as it occurred on the smooth samples, as shown in Fig. 13(c). The texture also continues to aid in the control of wear after the crack formation by providing a location where damage can accumulate as larger cracks formed from the smaller ones. This leads to overall cleaner cracks forming on the surface of the textured sample as compared with the smooth samples. It should be noted that the small size of the texture is extremely important for controlling wear in this way as the scale restricts the amount of stress allowed to buildup on the surface of the sample before it needs to release by forming these microcracks.
Optical images of some counterfaces from the various wear tests can be seen in Fig. 15. While the counterfaces from the wear tests on smooth samples are covered in debris, those from the wear tests on the textured samples are clear of debris and show extremely consistent patterns of wear and transfer film buildup near the contact area if any was apparent.
In this study, a general manufacturing method was developed for fabricating bioinspired surface textures using 2-photon lithographic 3D printing and replica molding with PDMS. The method was used to fabricate a surface based on the micro-texture found on a desert-based reptile, the Ocellated Skink. The wettability, friction, and wear behavior of the samples were studied. There was no significant effect on the water contact angle of the textured samples as compared with smooth samples due to the fact that the texture is only 150 nm tall, and this is too small of a feature to significantly affect the wettability of the surface. Although there was no significant reduction of friction under a loading condition of 100 g, tests under 20 and 50 g loading conditions showed that the coefficient of friction was reduced by 30% and 26%, respectively. This behavior suggests that the effect of the texture is the control of the adhesion contribution of friction since only the lower load conditions are significantly affected. The control of wear comes from the formation of microcracks facilitated by the texture of the samples. The average widths of the wear tracks formed were reduced under all loading conditions, with a 61%, 48%, and 44% reduction, respectively, under 20 g, 50 g, and 100 g loading conditions.
This work was supported by the National Science Foundation under Grant CMMI-1463306 and the support from the Center for Advanced Surface Engineering under NSF Grant No. OIA-1457888 through the Arkansas EPSCoR Program, ASSET III. Special thanks to Dr. Jason Ortega from the Biology Department of the University of Arkansas and Dr. Nancy Glover McCartney, the Curator of Zoology at the University of Arkansas Collections Facility for providing reptiles for examination, including the Chalcides ocellatus and Leptotyphlops.
Conflict of Interest
There are no conflicts of interest.
Data Availability Statement
The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request. The authors attest that all data for this study are included in the paper.