Journal of Colloid and Interface Science 363 (2011) 187–192

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Nanoscale biomimetics studies of Salvinia molesta for micropattern fabrication

James Hunt, Bharat Bhushan ⇑Nanoprobe Laboratory for Bio- & Nanotechnology and Biomimetics, The Ohio State University, Columbus, OH 43210, United States

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 May 2011Accepted 30 June 2011Available online 12 July 2011

Keywords:BiomimeticsSalvinia molestaMicropatternsMicrofabricationAtomic force microscopyContact angle

0021-9797/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.jcis.2011.06.084

⇑ Corresponding author.E-mail address: Bhushan.2@osu.edu (B. Bhushan).

The emerging field of biomimetics allows one to take inspiration from nature and mimic it in order tocreate various products, devices and structures. There are a large number of objects, including bacteria,plants, land and aquatic animals and seashells, with properties of commercial interest. The subject ofinterest for this research is the water fern Salvinia molesta because of its ability to trap air. Air-retainingsurfaces are of technological interest due to their ability to reduce drag when used for fluid transport, shipcoatings and other submersible industrial products in which drag is a concern. The purpose of thisresearch is to mimic the air trapping ability of S. molesta in order to prove that a structure can be createdin the lab that can mimic the behavior of the fern as well as demonstrate microfabrication techniquesthat can be utilized in industry to produce such materials. In this work, a novel methodology for the fab-rication of microstructures that mimic the water-pinning and air-trapping ability of S. molesta is intro-duced. Water contact angle, water roll angle and adhesive force of the new microstructure and waterfern are investigated.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

Throughout millions of years, nature has evolved to form com-plex optimized surfaces that can be mimicked for the developmentof new materials. The emerging field of biomimetics allows one totake inspiration from nature and mimic it in order to create variousproducts, devices and structures. Surfaces with superhydrophobic-ity, self-cleaning, drag reduction in fluid flow, energy conversionand conservation, high adhesion, thermal insulation, and self-healing mechanisms are some of the examples found in nature thatare of commercial interest [3]. Various features found in nature’sobjects are on the nanoscale. The major emphasis on nanoscienceand nanotechnology since the early 1990s has provided a significantimpetus in mimicking nature using nanofabrication techniques forcommercial applications [4].

There are a large number of objects, including bacteria, plants,land and aquatic animals and seashells, with properties of commer-cial interest. A classic example of biomimetics used for commercialapplications is the self cleaning ability of Nelumbo nucifera (lotus)which is known to be superhydrophobic and displays self cleaningdue to a hierarchical surface roughness and the presence of a hydro-phobic coating [7,15,16,18–20]. For a water droplet on the surfaceof a lotus leaf, the droplet is in the so called Cassie state and sitson top of the microstructure of the lotus leaf and forms air pocketsbetween the leaf and water [9]. Self cleaning surfaces are of interestfor various applications such as self-cleaning windows, windshields

ll rights reserved.

and exterior paints. Additionally, surfaces such as the lotus leafhave significantly less surface friction drag when submerged inwater due to the air pockets formed between the microstructuresof the lotus leaf.

Another subject of interest is the floating water ferns of genusSalvinia, because of their ability to trap air. Specifically, Salvinia mo-lesta has been an area of research for its complex hierarchical sur-face structures which are able to retain air and has been reportedby Koch et al. [16]. The floating water fern, S. molesta is an aquaticfern commonly known as giant salvinia and is native to south-east-ern Brazil. S. molesta is a free-floating plant that does not requiresoil and consists of leaves that are roughly 0.5–4 cm wide and long.The air-retaining properties of biological surfaces have also beenexplored by Barthlott et al. [2] where a mechanism for long-termair-retention for the floating water fern S. molesta is explored. Ithas been shown that the hierarchical nature of the S. molesta leafis predominantly composed of tiny eggbeater-shaped hairs, shownin Fig. 1, which are almost completely hydrophobic due to a coat-ing of nanoscopic wax crystals except for the terminal cells of eachhair which lack the crystals thus making them hydrophilic. Thesehydrophilic patches are located at the top of each hair where theindividual follicles forming the egg beater shape join together.Due to the hydrophilic patches at the tip of each hair, S. molestaexhibits a pinning effect of water against the top of the hairs whichenables the formation of air pockets between each hair as shown inFig. 2. The combination of hydrophilic patches coupled with an in-ner hydrophobic coating of the S. molesta hairs, and the subsequentability of S. molesta to pin water and retain air when submergedunderwater is referred to as ‘‘Salvinia Effect.’’ Air-retaining surfaces

Fig. 1. Optical micrograph of S. molesta leaf.

Fig. 2. Water droplet suspended by S. molesta hair at horizontal and verticalorientations demonstrating air pocket formation and water pinning at thehydrophilic tips where the terminal cells of each S. molesta hair is located.

188 J. Hunt, B. Bhushan / Journal of Colloid and Interface Science 363 (2011) 187–192

are of technological interest due to their ability to reduce dragwhen used for fluid transport, ship coatings and other submersibleindustrial products in which drag is a concern. Superhydrophobicsurfaces have been utilized to obtain the desired air film for thepreviously stated applications, however the effect called ‘‘giant li-quid slip’’ has been shown to deteriorate the presence of such airfilms in a matter of minutes [1,10,11,17]. Therefore, S. molesta’s un-ique ability to pin water and trap air is of importance in order to

increase the durability and lifetime of air pocket formation for dragreduction in industrial use.

The purpose of this research is to mimic the air trapping abilityof S. molesta in order to prove that a structure can be created in thelab that can mimic the behavior of the fern as well as demonstratemicrofabrication techniques that can be utilized in industry to pro-duce such materials. In this study a novel approach to fabricating amicrostructure which mimics the behavior of S. molesta is given. Toaccomplish this, a micropattern is created in the lab with compara-ble dimensions to the S. molesta hairs. The micropattern is thentreated with a hydrophobic coating which is then stripped awayto produce a new microstructure which is hydrophobic every-where except for the tips of the micropattern in the same manneras the S. molesta hairs themselves. The new micropattern is thenstudied to determine air trapping ability as well as its ability topin water in the same fashion as S. molesta.

2. Experimental

2.1. Characterization, fabrication and coating

Samples of S. molesta were provided for this study by the OhioState University Biological Sciences Greenhouse. To store the waterfern throughout testing, a 10-gallon aquarium, water filter, and UVlamp was purchased. To characterize the fern hair, an opticalmicroscope was used to image S. molesta as well as determinethe spacing of the fern hairs (Fig. 1). Using the microscope images,an average hair spacing of 490 lm with a range of 250–750 lmwas observed. The apparent surface area of the tip of the S. molestahair where the four eggbeater-shaped hairs come to a point wasobserved to be on the order of 20 lm. The height of the S. molestahairs was about 2 mm. To observe the effect that water pressurehad on the ability of S. molesta to maintain an air layer when sub-merged in water, the leaf was subjected to a hydrostatic pressuretest at a depth of 0.3 m under a column of water.

In order to create the micropatterns used in this study, a twostep molding technique was employed. Using the methods re-ported by Bhushan and Jung [8] two master Si microstructureswere fabricated by photolithography. The technique was used tocreate two microstructures of 14 lm diameter (mimicking the sur-face area of the S. molesta hair tips) and 30 lm height with pitchesof 210 and 26 lm. The 30 lm pillar height, while much smallerthan the observed 2 mm height of the S. molesta hairs was selectedbased on the tallest pattern size readily available for the givenpitch and diameter dimensions. The 210 lm pitch was selectedin order to recreate the minimum hair spacing observed on S. mo-lesta as closely as possible. The 26 lm pitch was selected in orderto observe the effects that pitch had on the hydrophobic behaviorof the microstructure as well as its ability to trap air. Additionally,the 26 lm pitch was selected to compensate for the low height ofthe micropillars as compared to the observed height of the S. mo-lesta hairs in order to reproduce a closer ratio between heightand pitch of the fern hairs. Negative molds were created fromthe master Si microstructures and were then filled with a liquidepoxy resin [8]. The specimens were then transferred to a vacuumchamber at 500 mTorr (66.7 Pa) for one minute to remove trappedair and to enhance the saturation of the epoxy resin into the neg-ative mold. After trapped air bubbles ceased to release from themold, the mixture was removed from the vacuum and placed atroom temperature (24 h at 22 �C). After hardening, the positivereplica of the original Si master was removed from the mold.

Once the epoxy replicas were created, a coating of (tridecaflu-oro-1,1,2,2-tetrahydrooctyl) trichlorosilane (Gelest) was then ap-plied by a syringe to the surface of the micropatterns to imparthydrophobicity. After coating, the micropatterns were placed in a

Fig. 3. Micropillars created in lab (14 lm diameter, 30 lm height, 26 lm pitch)coated with (tridecafluoro-1,1,2,2-tetrahydrooctyl) trichlorosilane (Gelest) andstripped away.

J. Hunt, B. Bhushan / Journal of Colloid and Interface Science 363 (2011) 187–192 189

vacuum at 500 mTorr for one minute to remove air bubbles and toassist with saturation. Finally, the hydrophobic coating was re-moved using double sided tape to strip away the trichlorosilanefrom the top of the microstructures leaving the hydrophobic coat-ing in between the micropillars. The tape was attached to a flat sur-face and then lowered onto the top of the micropattern using amotor-driven arm until the surface of the micropattern came intocontact with the double sided tape which ensured that the tapestayed flat and did not bend while making contact with the tipsof the micropattern. The hydrophobic coating was stripped fromthe tips of the micropillars, creating a microstructure which ishydrophobic with small patches at the tips of the micropillarswhich are hydrophilic. This was done in order to mimic the S. mo-lesta hairs in which the surface of the leaf is almost completelyhydrophobic except for tiny hydrophilic patches at the tips of thefern hairs. Optical microscope imaging of the 26 lm pitch micro-structure is shown in Fig. 3 for the coated and strippedmicropillars.

2.2. Contact angle and nanoscale adhesion measurements

Water droplets of roughly 1.5 mm radius (15 lL volume) wereapplied to the uncoated micropatterns using a microsyringe forboth pitch values in order to observe a difference in contact angle.Contact angle was measured using proprietary software developedfor the ImageJ software program [21]. The uncoated micropatternswere then tilted to investigate the angle at which water begins to

roll off of the surface of the micropattern. The contact angle and tilttests were repeated for both the coated 210 and 26 lm pitches. Fi-nally, the trichlorosilane was stripped away from the top of themicropillars as previously described and contact angle measure-ment and tilt testing was repeated.

Adhesive force measurements were taken for the S. molesta hairtips as well as the coated and stripped micropatterns. A commer-cial AFM (Nanoscope IIIa, Veeco, Santa Barbara, CA, USA) was usedfor this study [4,5]. A silicon tip with an 8 lm diameter borosilicateparticle (Duke Scientific Corporation, 9008) attached to the end of arectangular cantilever beam (spring constant of 3 N/m) were usedfor adhesion measurements. Adhesive force was calculated usingthe force distance curve technique [4,5]. The experiments wereperformed at room temperature (22 �C) and 45–55% relativehumidity. In order to collect adhesion data for the micropillarsan initial scan size of 30 lm � 30 lm was selected at a 1 Hz scanrate, a single pillar was identified and the scan size was decreasedincrementally until just the top of one micropillar was scanned. Asingle force distance curve was generated for the center of one pil-lar, with 128 data points recorded for each of the extending andretracting deflections of the AFM tip. This was repeated a mini-mum of 30 times for both the coated and stripped micropatternson various pillars at different locations on the microstructure.

3. Results and discussion

3.1. Cassie–Baxter and Wenzel transition criteria

In order to understand the results of the contact angle experi-ments it is necessary to examine the importance of the Cassie–Bax-ter and Wenzel transition criteria. The effect of water dropletcurvature on the Cassie–Baxter and Wenzel regime transition hasbeen investigated by Bhushan and Jung [6,7] and Jung and Bhushan[12–14]. A small water droplet suspended on a superhydrophobicsurface consisting of a regular array of circular pillars with diame-ter D, height H, and pitch P is considered. The local deformation forsmall droplets is governed by surface effects rather than gravity,and the curvature of a droplet is governed by the Laplace equation,which relates the pressure inside the water droplet to its curvature.Therefore, the curvature is the same at the top and bottom of thewater droplet.

For a micropatterned surface, the maximum droop of the drop-let occurs in the center of the square formed by four adjacentmicropillars. The maximum distance between two pillars is wherethe water droplet will droop the most and is the diagonal of thesquare formed by the four pillars. This diagonal distance betweentwo pillars is given by ð

ffiffiffi

2p

P � DÞ2=ð8RÞ where P is the microstruc-ture pitch, D is the micropillar diameter, and R is the water dropletradius. If the droop of the water droplet is greater than the pillarheight, the droplet will just come into contact with the bottomof the cavity between two pillars and transition from the Cassie–Baxter to Wenzel regime occurs. The equation used to determinewhich regime the droplet and micropattern interface resides is gi-ven in Eq. (1) [8].

ðffiffiffi

2p

P � DÞ2

RP H ð1Þ

3.2. Air pocket formation, contact angle and roll angle

When submerged in water, a silvery layer is visible on the sur-face of the S. molesta leaf indicating a layer of air trapped againstthe surface. When submerged to a depth of 0.3 m in a water col-umn, the fern is able to retain an air layer under a hydrostatic pres-sure of 2.99 kPa. With this result in mind, it is necessary to explore

190 J. Hunt, B. Bhushan / Journal of Colloid and Interface Science 363 (2011) 187–192

the interaction between water droplets and the micropatterns cre-ated in this study to observe the micropatterns’ ability to create anair layer and pin water to its surface. This is accomplished bystudying the contact angle and angle at which water will remainpinned to the surface of the micropatterns.

Results from the water contact angle test are shown in Fig. 4.Also shown in Fig. 4 are the results of the tilt test used to checkfor water pinning at vertical orientation. To begin, the uncoated210 lm pitch micropillars are completely saturated by the waterdroplet. This is physically observed and is also an expected result

Fig. 4. Contact angle measurement for coated, uncoated and stripped micropillars (210 apinning occurs.

due to the Cassie–Baxter and Wenzel region transition criteria.Using Eq. (1), it is seen that the interface of the water dropletand 210 lm pitch micropillars falls within the Wenzel regimedue to the 30 lm height of the micropillars. When the values forthe dimensions of the 210 lm pitch micropillars are substitutedinto Eq. (1), the left hand side of the equation will be greater thanthe pillar height, indicating a Wenzel state. The pillars are not highenough to create an air layer between the water droplet and baseof the micropattern. In order to recreate the air layer created bythe S. molesta hairs for a micropillar pitch of 210 lm, the pillar

nd 26 lm pitch) and samples turned to 90� vertical orientation to determine if water

J. Hunt, B. Bhushan / Journal of Colloid and Interface Science 363 (2011) 187–192 191

height would have to be increased by a factor of two to three, giventhat the other microstructure dimensions remain the same, in or-der to satisfy the relationship in Eq. (1). As a result of the hydro-philic nature of the hardened epoxy microstructures, the waterdoes not roll off of the 210 lm micropillars up to an angle of 90�(vertical orientation). The coated 210 lm micropillars show an in-creased contact angle due to the deposition of the hydrophobiccoating. The stripped 210 lm pitch micropattern shows an in-creased contact angle over the uncoated sample and pins the waterdroplet to the surface. For the 210 lm pitch that is coated andstripped, the water remains pinned to the microstructure at verti-cal orientation because the water droplet is in the Wenzel State.

The uncoated 26 lm micropattern shows an increased contactangle of 139�. As predicted by the transition criteria discussed previ-ously, the interface falls within the Cassie–Baxter regime, thereforecreating air pockets between the water droplet and base of themicropattern. However, despite the higher contact angle observedon the 26 lm pitch, the water does not roll off of the epoxy micro-structure due to the hydrophilic nature of the micropillars. Theattraction between the micropillars and water droplet is high en-ough to pin the water to the surface of the microstructure at verticalorientation. The coated 26 lm micropattern displays a contact angleof 147� and is higher than the 139� contact angle of the uncoatedsample.

The most notable difference between the 210 and 26 lmpitches is that water begins to roll off of the 26 lm pitch at an in-cline of 14� whereas the water droplet stays pinned to the coated210 lm microstructure up to vertical orientation. When the26 lm microstructure is stripped, the contact angle is reduced to136� which is on par with the uncoated 26 lm sample and waterremains pinned to the surface due to the hydrophilic nature ofthe now-exposed tips of the micropillars as a result of stripping.Because of the hydrophilic nature on the surface of the microstruc-tures, the water droplet remains pinned to the surface at verticalorientation despite the hydrophobic nature of the coating betweenthe micropillars and air pocket resulting from the interface resid-ing in the Cassie–Baxter regime.

This is the first time a microstructure has been fabricated whichmimics the behavior of S. molesta. The micropattern traps air at theinterface of the water droplet and also has a hydrophobic coatingin between the micropillars while the tips of the pillars arehydrophilic in nature, thus exhibiting the same characteristics of S.molesta.

Fig. 5. Force–distance curves for fern and stripped 26 lm pitch microstructureshowing extending and retracting motion of AFM tip and illustrating adhesive forcecalculation. The vertical deflection of the AFM tip is given by Fad/k where Fad is theadhesive force between AFM tip and the surface and k is the spring stiffness of thecantilever (3 N/m). Fad is calculated by multiplying the tip deflection (read from thegraph) and multiplying by the spring stiffness. Summary of adhesive force for fernhairs and 26 lm pitch micropillars (coated and stripped).

3.3. Adhesion

For the first time, AFM has been utilized to quantify the adhe-sive force of S. molesta hairs. The force–distance curve resultingfrom the adhesion study of S. molesta is shown in Fig. 5. The dipin Fig. 5 shows a vertical tip deflection in the retracting directionof 69 nm which is used to calculate the adhesive force (Fad) by mul-tiplying the vertical deflection of the tip by the cantilever springconstant (k) of 3 N/m. This corresponds with a resulting adhesiveforce of 207 nN. The force–distance curve for the stripped 26 lmmicropattern is also shown in Fig. 5. Similar to the S. molestaforce–distance curve, a tip deflection of 67 nm is shown which cor-responds with an adhesive force of 201 nN.

A comparison of the adhesive force for S. molesta and the 26 lmmicropattern is also shown in Fig. 5 for both the coated andstripped micropatterns. As expected, the trichlorosilane-coatedmicropattern exhibits a much lower average adhesive force of54 nN. The S. molesta hairs show an average adhesive force of165 nN with a standard deviation of 47 nN and the stripped26 lm microstructure shows an average adhesive force of 180 nNwith a standard deviation of 19 nN.

The similarity between the adhesive force of the strippedmicrostructure and the S. molesta hair is an important result be-cause it shows that the stripped microstructure will exhibit similarsurface characteristics as the S. molesta hair. The higher adhesiveforce of the stripped microstructure compared to the coated26 lm microstructure explains why water will roll off of the coatedsample at an incline of 14� whereas water will remain trapped atvertical orientation for the stripped sample. For the first time, a

192 J. Hunt, B. Bhushan / Journal of Colloid and Interface Science 363 (2011) 187–192

microstructure has been created in the lab that exhibits a compa-rable wetting behavior to the S. molesta hair.

4. Conclusions

For the first time, a micropattern that mimics the hydrophobic/hydrophilic nature and air-trapping/water pinning behavior of S.molesta has been created. Using the methods described previously,a 26 lm pitch microstructure composed of micropillars has beencreated such that the surface is predominantly hydrophobic whilestill having hydrophilic tips. The resultant microstructure is shownto create trapped air pockets between a water droplet and micro-structure base due to the water droplet residing in the Cassie–Bax-ter state. Through the use of AFM, the adhesive force of S. molestahas been characterized for the first time. Additionally, the newlycreated microstructure has been shown to exhibit similar adhesionas the S. molesta hair. As a result of the similar adhesive forces,water pinning on the microstructure has been observed up to ver-tical orientation in the same fashion as S. molesta.

The 210 lm pitch microstructure does not exhibit the air trap-ping ability of S. molesta, despite the micropatterns similarity ofpitch with the S. molesta hairs. While water remains pinned tothe 210 lm micropillars in the same fashion as S. molesta, an airlayer is not formed in between a water droplet and the base ofthe microstructure due to the water droplet residing in the Wenzelstate as predicted by Eq. (1). To improve on this design, it is re-quired that microstructures with the same height as S. molestahairs (roughly 2 mm) are created. Water droplets sitting on micro-pillars with this height would reside in the Cassie–Baxter state.This new structure, when coupled with the hydrophobic treat-ments presented in this paper would mimic the air trapping abilityof S. molesta more closely.

It has been shown that the Salvinia Effect of pinning water andcreating air pockets at the water surface interface, as well as the nec-essary production techniques required for such a surface, is viablethrough the production of hydrophobic micropatterns with a prede-fined level of hydrophilic patches. The commercial applications forsuch a technology include industrial uses in which fluid transport,drag reduction, thermal insulation, and increased buoyancy are of

interest, especially in applications in which giant liquid slip has beenshown to deteriorate air films created by superhydrophobicsurfaces.

Acknowledgments

The authors would like to thank Joan Leonard, GreenhouseCoordinator of the Biological Sciences Greenhouse at The OhioState University for providing the S. molesta samples used in thisresearch as well as general horticultural guidance. We would alsolike to thank Dr. Hyungoo Lee for data collection support as wellas scientific discussions.

References

[1] A.K. Balasubramanian, A.C. Miller, O.K. Rediniotis, AIAA J. 42 (2004) 411–414.[2] W. Barthlott, T. Schimmel, S. Wiersch, K. Koch, M. Brede, M. Barczewski, S.

Walheim, A. Weis, A. Kaltenmaier, A. Leder, H.F. Bohn, Adv. Mater. 22 (2010)2325–2328.

[3] B. Bhushan, Philos. Trans. R. Soc. A 367 (2009) 1445–1486.[4] B. Bhushan (Ed.), Springer handbook of nanotechnology, third ed., Springer,

Heidelberg, Germany, 2010.[5] B. Bhushan, Nanotribology and nanomechanics I – measurement techniques

and nanomechanics, II – nanotribology, biomimetics, and industrialapplications, third ed., Springer-Verlag, Heidelberg, Germany, 2011.

[6] B. Bhushan, Y.C. Jung, Ultramicroscopy 107 (2007) 1033–1041.[7] B. Bhushan, Y.C. Jung, J. Phys.: Condens. Matter 20 (2008) 225010.[8] B. Bhushan, Y.C. Jung, Prog. Mater. Sci. 56 (2011) 1–108.[9] A.B.D. Cassie, S. Baxter, Trans. Faraday Soc. 40 (1944) 546–551.

[10] C.-H. Choi, C.-J. Kim, Phys. Rev. Lett. 96 (2006) 66001–66004.[11] C. Henoch, T.N. Krupenkin, P. Kolodner, J.A. Taylor, M.S. Hodes, A.M. Lyons, K.

Breuer, Turbulent drag reduction using superhydrophobic surfaces, in: 3rdAIAA Flow Control Conf., June 5–8, San Fransisco, CA, AIAA-2006-3192, 2006.

[12] Y.C. Jung, B. Bhushan, Scripta Mater 57 (2007) 1057–1060.[13] Y.C. Jung, B. Bhushan, J. Microsc. 229 (2008) 127–140.[14] Y.C. Jung, B. Bhushan, Langmuir 24 (2008) 6262–6269.[15] K. Koch, B. Bhushan, W. Barthlott, Soft Matter 4 (2008) 1943–1963.[16] K. Koch, B. Bhushan, W. Barthlott, Prog. Mater. Sci. 54 (2009) 137–178.[17] C. Lee, C.-J. Kim, Langmuir 25 (2009) 12812.[18] M. Nosonovsky, B. Bhushan, Multiscale dissipative mechanisms and

hierarchical surfaces: friction, superhydrophobicity, and biomimetics,Springer, Heidelberg, Germany, 2008.

[19] M. Nosonovsky, B. Bhushan, J. Phys.: Condens. Matter 20 (2008) 225009,doi:10.1088/0953-8984/20/22/225009.

[20] M. Nosonovsky, B. Bhushan, Adv. Funct. Mater. 18 (2008) 843–855.[21] A.F. Stalder, G. Kulik, D. Sage, L. Barbieri, P. Hoffmann, Colloids Surf., A –

Physicochem. Eng. Aspects 286 (2006) 92–103.