Reverse engineering of a wool fibre to mimic the structural hierarchy of a gecko's foot

The gecko's foot employs a five tier hierarchical system to achieve consistent adhesion to a surface; ranging from mm sized lamellae, seta, branch, spatula to nanoscaled spatula tips [1–5] and figure 1(a). The spatula tips provide the final contact with a surface, such that a non-muscular action is required for adhesion. This was proven in a rather unconventional way, by attaching a dead tokay gecko by a single digit to the ceiling of its glass tank; the digit was able to support its own body for three days where upon it was removed due to the smell becoming unbearable [1]. There have been several attempts to replicate the gecko's multi-levelled foot structure, and various new techniques and novel materials have been studied (because traditional machining methods do not replicate the appropriate macro to micro structures) ranging from microelectromechanical systems to multi walled carbon nanotube arrays [2–6]. Thus, synthetic dry adhesives have been produced consisting of adhesive elements with gecko like properties such as reusability, directionality, self-cleaning ability, rough surface adhesion and high adhesive stress however, scale-up of the systems has proved more of a challenge. Geckos increase adhesion with animal size by increasing toepad area and decreasing compliance of the toepad in the loading direction [7]. Scaling efficiency is provided by the hierarchical structure of the gecko's foot as a whole; with controllable adhesion at the spatula tip (by van der Waals forces), discrete adhesive elements (lamellae) and compliance normal to the adhering surface while remaining relatively stiff in the loading direction (close to parallel to the adhering surface). These key hierarchical features have been exploited to demonstrate a gecko-inspired adhesive system based on polydimethylsiloxane (PDMS) slanted micro-wedge adhesive elements, that was scaled to enable a 70 kg human to climb vertical glass with 140 cm² of adhesive per hand [8]. Further work to enhance compliance control with respect to biomimetic, structured dry-adhesives continues [9].

Zoom In Zoom Out Reset image size

The hair of the gecko's foot belongs to the keratin protein family as do the fibres of wool. Wool fibres are in fact composite materials that have highly complex physical and chemical composition (figure 1(b)), that has evolved over millions of years in order to protect sheep from the extremes of hot and cold. Wool fibres contain more than 170 different proteins [10]; each type is located in a specific region of the fibre. There are two main structures in the wool matrix; the outer cuticle, or scales, and the inner core called the cortex. The cortex is composed of cortical cells, in turn these cortical cells contain keratin based macrofibrils that run along the length of the mature wool fibre. The macrofibrils have been shown to consist of bundles of microfibrils held together by interfibrillar cement to form bundles that are ≈200 nm in diameter [10]. The cuticle scales, which act as armour plating for the wool, are comprised of three layers; epicuticle, exocuticle and the endocuticle, each have a specific role to protect the fibre from moisture penetration and mechanical abrasion [11–13]. Bradbury however showed complete cuticle removal is possible by submersion of wool fibres in boiling formic acid [14]. Furthermore it is well established that the cortex of wool is susceptible to further breakdown by enzymatic digestion, as well as reduction via thiol groups [15]. Recent work has reported how stretching of single wool fibres and bundles of fibres can be used to explore hierarchal structure transformation mechanisms to produce 230% strains in single fibres as the disulphide bonds and peptide chains were taken apart or reconstructed under load by new crosslinking bonds inside the fibres [16].

The physical dimensions of the hierarchical wool fibre components and that of the gecko's foot hair have been summarised and compared in the table in figure 1(c). It is immediately apparent that the diameter of each structural component of the gecko hair is of comparable size to those of a wool fibre, especially the spatula tip and macrofibril which share a diameter of 200 nm. The wool fibre has a further sub division of the macrofibril, the microfibril (which is in the region of 10 nm in diameter). We suggest that if wool is processed to release these subcomponents at a fibre end then the similarity of these protein structures to those of the gecko should make it possible to replicate the compliance and adhesive power of the gecko's foot. This would require complete removal of the cuticle scales of a cross-sectioned wool fibre and then controlled removal of the cortical matrix to fibrillate and release the component macrofibrils. It is clear from the work of Bradbury that a combined chemical and physical attack is required to achieve such fibrillation. If successful, exposed macrofibril ends (themselves composed of many microfibrils) could provide adhesive surfaces (similar to the spatula tip of the gecko), while the amount of retained cortical matrix could provide control over surface compliance (as per the seta to branch ratio of the gecko foot). If undertaken on an array of wool fibre ends presented at the surface of a textile it may be possible to produce an adhesive fabric with each single wool fibre end acting as a discrete adhesive element (as per the lamella of the gecko's foot).

We report here on the release of the hierarchical structure of Merino wool fibres using a combination of formic acid refluxing with agitation and trypsin digestion with ultrasonication and on subsequent adhesion measurements of exposed macrofibrils at the surface of cross-sections of descaled, fibrillated wool fibres.

Removal of cuticle scales from individual merino wool fibres has been previously demonstrated by a process of agitation in formic acid [14]. Transferring this technique to Merino wool yarns required modification to the methodology; including increases in the duration of the reflux time and a subsequent agitation. Confirmation that complete removal of cuticles scales could be achieved is demonstrated by a combination of measured mass loss, FTIR spectroscopy and scanning electron microscopy (SEM) (figure 2). The mass loss data show that after an initial increase a plateau is reached after 60 min of reflux (figure 2(a)). FTIR spectroscopy of untreated and refluxed (for 60 min) fibres showed that there was a shift to higher wavenumber of the amide I absorption band (figure 2(b)). This observation is consistent with prior work that shows a shift in the peak of the amide I absorption band occurs upon the loss of cuticle cells, because the cuticle cells contain more cysteine, proline, serine, and valine residues, which are non-α forming amino acids [17–20]. The final piece of evidence for cuticle removal is that from SEM, which shows near elimination of cuticle with increasing reflux time, (figures 2(c)–(f) show progressive removal of surface scales or cuticle, figure 2(f) shows smooth fibres consistent with complete cuticle removal). Furthermore, SEM suggests that the formic acid is delaminating the cuticle scales i.e. favourably attacking the cell membrane complex that holds the cuticles to the cortex (figure 2(d)), as previously reported [14]. The resultant solution was passed through a 2 μm pore sized filter, upon which the delaminated cuticle scales were retained and subsequently imaged (figure 2(g)).

Zoom In Zoom Out Reset image size

Control of cellular breakdown at the fibres' end was achieved through a combination enzymatic digestion and mechanical attack of resin embedded and diamond knife, cross sectioned, refluxed yarns (as described in the control of digestion section of the methods and shown in figure 3(a)). SEM shows how the combination of trypsin immersion and ultrasonics breakdown the cellular structure of the cortex (figure 3(b)). Optimum conditions to achieve this breakdown were found to be 0.2 g trypsin per 100 ml of disodium hydrogen phosphate solution combined with 180 s of 40 kHz ultrasonics. Higher magnification SEM images show that this treatment successfully releases fibrils at the surface of the cross-sectioned yarns (figure 3(c)). Image analysis of the exposed fibril ends reveals a mean diameter 260 +/− 58 nm s, (table (d) in figure 3).

Zoom In Zoom Out Reset image size

Atomic force microscopy (AFM) analysis on the resin embedded yarn cross-sections shows the effect digestion has on the fibre end (figure 4). The tapping mode images, of the resin mounted and cross-sectioned, non-digested and digested fibre ends clearly show a difference in the surface structure and topography. We can see that there is a clean and smooth surface present in the non digested cross section (figure 4(a)) whereas in the digested cross-section there is clear evidence that the structure has under gone morphological changes (figure 4(b)). Higher magnification of the untreated sample is shown in figure 4(c). We can see at higher magnifications that in the treated sample potential macrofibrils are exposed within the cells (figure 4(d)). The areal density of the macrofibrial ends (260 nm across) exposed within a cortical cell is estimated to 6.9 × 10⁶ per mm². The digested cross section has a mean roughness 6 times greater than that of the as received sample (table in figure 4(e)). An increase in the surface roughness is a key mechanism used by several biological attachments systems, not just geckos [21]. AFM was further utilised here to perform adhesion measurements between the tip and the surface of the digested and non-digested yarn cross-sections. AFM tip retraction or adhesive force was made as a measure of local adhesion on both the non-digested and digested cross-sections (with 100 repetitions for each characteristic region). The non-digested cross-section had a mean adhesive force of 0.18 nN (and a standard deviation, SD of 0.09 nN) whereas the adhesive force in the digested cross sections increased by a factor of 100 to a mean values of 18.2, nN (with a significantly greater range or SD of 8.1 nN) (table in figure 4(e)).

Zoom In Zoom Out Reset image size

This research has developed a processing route to expose, layer by layer, the components in a Merino wool yarn down to the release of nanoscale macrofibrils at fibre ends. An overview of the experimental procedure and the optimum conditions for each processing step is presented in (and discussed below), and includes indication of which aspect of the gecko's hierarchical foot structure the processing is attempting to release and replicate on a wool fibre. Importantly the physical structure and shape of individual keratin-based macrofibrils can be maintained throughout (figure 5).

Zoom In Zoom Out Reset image size

The initial processing reported here expands on Bradbury's work for cuticle removal of wool fibres [13, 14]. Use of boiling formic acid to strip the wool fibre of cuticle was adapted in this instance to completely remove the cuticle from a yarn of wool. Using a combination of analytical techniques, a repeatable procedure for complete cuticle removal has been achieved without causing significant damage to the underlying structure of the cortex. This is an essential step as the cuticle protects the cortex of the wool and we had to be confident that all cuticle was removed before any further reverse engineering. The optimum conditions found for complete cuticle removal of a wool yarn are 80 min of reflux in boiling formic acid to attack the cell membrane complex that holds the cuticle to the cortex with an additional 16 h of laboratory shaking in the acid at room temperature to fully detach the cuticle cells.

Once the cuticle was successfully removed it is then possible to further break down wool into macrofibril subcomponents. Immersion of refluxed yarns in a trypsin solution and application of an ultrasonic environment supply the appropriate driving force for the trypsin to digest the matrix that binds the cortical cells and release sub unit macrofibrils. The optimum condition for digestion was found to be a solution of 0.1 M Na2HPO4 containing 0.2 g of trypsin per 100 ml. The refluxed and partially digested wool was submersed and subjected to 3 min of ultrasonic agitation (at 40 kHz) in this solution. This indicates that replication of gecko adhesion by fibrillation of wool is potentially viable, however the mechanical stability of the resulting macrofibrils within the wool fibres is not optimal; it became apparent on this digestion that the released macrofibrils are unable to self-support in a manner similar to a gecko spatula (data not shown, see Caven PhD thesis 2011 [22]). So resin embedding and cross-sectioning of refluxed yarns was essential here to achieve greater control over the fibrillation process.

Resin embedding the fibrillated yarn is an unwanted step as it means that the transfer from yarn to fabric will be a major obstacle however, it was necessary at this stage in order to demonstrate digestion for fibrillation into a hierarchical structure. Digestion at the surface of cross-sectioned, embedded yarns exposes macrofibrils (figure 3) which are measured to occupy an areal density of 6.9 × 10⁶ per mm², comparing favourably to that of the gecko's foot which has been reported to have a spatula density of 1–14 × 10⁶ per mm² [2–6].

Mechanical testing to analyse the surface structure of the digested wool by AFM revealed differences in the surface structures of the non-digested and digested or fibrillated cross sections of the wool yarns. Topography images show clear cortical cell boundaries; furthermore roughness analysis shows that fibrillation increased the surface roughness of the digested cross sections by a factor of 6 suggesting macrofibril ends have been exposed (figure 4). The key result of the AFM work is the adhesive force measurements which show that fibrillation of refluxed and digested yarn cross-sections increases the retraction force required to detach an AFM cantilever from the wool surface by 100 from a mean of 0.18 up to 18 nN. The combined adhesive force of a gecko's seta, composed of multiple spatula tips, was reported to be 194 μN [1]. We have not be able to measure adhesion of bundles of macrofibril ends of the digested, fibrillated cross sections of the wool yarns because the force analysis technique used here differs to that of prior AFM work done on the gecko's foot [1]. Nonetheless, digestion of the wool fibre surface has shown a very significant and repeatable increase in adhesive force and if we have indeed exposed macrofibril ends, as suggested by the AFM imaging and surface roughness measurements, then a mean adhesive force of 18 nN is comparable to that reported for the gecko's foot spatula tip of 11 nN [1]. We are yet to confirm if the depth profile is optimised to meet the buckling (and therefore compliance) conditions reported for a gecko's foot, however we believe this could be achievable with further digestion analysis e.g. (prolonged ultrasonication in trypsin) [23, 24].

The current work has developed an experimental route to reserve engineer wool back to its sub unit macrofibrils, where each experimental step only targets the desired wool component that needs to be removed, and does not significantly damage the remaining structure. Thus, the format and to some extent the adhesive properties of the gecko's hierarchal foot structure have been successfully replicated in wool yarns in terms of structure and geometric conformation at a resin embedded, cross-sectioned surface of a formic acid refluxed and trypsin digested yarn. Further depth optimisation is required to fully mimic the hierarchical structure of a gecko foot to provide compliance and adhesion however, an optimal release of microfibrils from the exposed cortex of wool could potentially increase surface roughness further and generate a higher density of van der Waals forces on surface contact than that of other biological attachment systems.

5.1. Material

5.2. Cuticle removal

5.3. Cortical cell breakdown

5.4. Control of digestion

5.5. Microscopy

5.6. Atomic force microscopy

The authors gratefully acknowledge the Engineering and Physical Sciences Research Council (UKRI-EPSRC) for funding the research as a Doctoral Training Partnership. We also acknowledge support from the Leeds EPSRC Nanoscience and Nanoequipment User Facility Grant EP/R02863X/1 at the University of Leeds. The study was also partly supported by the Bergrettung Tirol (Austria) and the Comet K-Project 'Sports Textiles' (Grant No. 820494), funded by the Austrian Research Promotion Company (FFG), Standortagentur Tirol (Austria), and region Vorarlberg (Austria).

All data that support the findings of this study are included within the article (and any supplementary files).