Langmuir 2009, 25, 1659-1666
1659
Lotus-Like Biomimetic Hierarchical Structures Developed by the Self-Assembly of Tubular Plant Waxes Bharat Bhushan,*,† Yong Chae Jung,† Adrian Niemietz,‡ and Kerstin Koch†,‡ Nanoprobe Laboratory for Bio- & Nanotechnology and Biomimetics (NLB2), The Ohio State UniVersity, 201 West 19th AVenue, Columbus, Ohio 43210-1142, and Nees Institute for BiodiVersity of Plants, Rheinische Friedrich-Wilhelms UniVersity of Bonn, Meckenheimer Allee 170, 53115 Bonn, Germany ReceiVed August 1, 2008. ReVised Manuscript ReceiVed September 12, 2008 Hierarchical roughness is beneficial for superhydrophobic and self-cleaning surfaces. Biomimetic hierarchical surfaces were fabricated by replication of a micropatterned master surface and self-assembly of two kinds of tubular wax crystals, which naturally occur on the superhydrophobic leaves of Tropaeolum majus (L.) and Leymus arenarius (L.). These tubule forming waxes are multicomponent waxes, composed of a mixture of long chain hydrocarbons. Thermal evaporation of wax was used to cover artificial surfaces with a homogeneous wax layer and tubule formation was initiated by temperature and a solvent vapor phase. Based on this technique, various nanostructures produced by three-dimensional tubular waxes have been fabricated by changing the wax mass. Fabricated structures and surface chemistry mimic the hierarchical surfaces of superhydrophobic and self-cleaning plant surfaces. The influence of structures on superhydrophobicity at different length scales is demonstrated by investigation of contact angle, contact angle hysteresis, droplet evaporation and propensity of air pocket formation as well as adhesive forces. The optimal structural parameters for superhydrophobicity and low static contact angle hysteresis, superior to natural plant leaves including Lotus, have been identified and provide a useful guide for development of biomimtetic superhydrophobic surfaces.
1. Introduction Superhydrophobic surfaces with high contact angle and low contact angle hysteresis (the difference between the advancing and receding contact angles) have low drag for fluid flow and provide water repellence, self-cleaning and can also lead to energy conversion such as the microscale capillary engine and energy conservation ability from friction and energy dissipation during the sliding contact of solid surfaces.1-3 These superhydrophobic surfaces are of interest for various applications, including selfcleaning windows, windshields, exterior paints for buildings and navigation of ships, utensils, roof tiles, textiles, and applications requiring a reduction of drag in fluid flow, e.g., in micro/ nanochannels. In addition, condensation of water vapor from the environment and/or process liquid film can form menisci leading to high adhesion in devices requiring relative motion.4-7 Superhydrophobic surfaces are needed to minimize adhesion. A model surface for superhydrophobicity and self-cleaning is provided by the leaves of the Lotus plant (Nelumbo nucifera), which have a hierarchical structure, built by convex cell papillae and randomly oriented hydrophobic wax tubules superimposed.8-10 Water on such a surface forms a spherical droplet, and both the contact area and adhesion to the surface are * Corresponding author. E-mail:
[email protected]. † The Ohio State University. ‡ Rheinische Friedrich-Wilhelms University of Bonn.
(1) Bhushan, B.; Jung, Y. C. J. Phys.: Condens. Matter 2008, 20, 225010. (2) Nosonovsky, M.; Bhushan, B. J. Phys.: Condens. Matter 2008, 20, 225009. (3) Nosonovsky, M.; Bhushan, B. Multiscale DissipatiVe Mechanisms and Hierarchical Surfaces: Friction, Superhydrophobicity, and Biomimetics; SpringerVerlag, Heidelberg,Germany, 2008. (4) Bhushan, B. Introduction to Tribology; Wiley, NY, 2002. (5) Bhushan, B. J. Vac. Sci. Technol. B 2003, 21, 2262. (6) Bhushan, B. Springer Handbook of Nanotechnology, 2nd ed.; SpringerVerlag, Heidelberg,Germany, 2007. (7) Bhushan, B. Nanotribology and Nanomechanics - An Introduction, 2nd ed.; Springer-Verlag, Heidelberg, Germany, 2008. (8) Barthlott, W.; Neinhuis, C. Planta 1997, 202, 1. (9) Koch, K.; Bhushan, B.; Barthlott, W. Soft Matter 2008, 4, 1943. (10) Koch, K.; Bhushan, B.; Barthlott, W. Prog. Mater. Sci. 2009, 54, 137.
dramatically reduced.1-3,11,12 The measured static contact angle and contact angle hysteresis are about 164° and 3°, respectively. Waxes are the hydrophobic component of the plant cuticle, which covers nearly all primary surfaces of land plants. The multifunctional properties of the cuticle and its integrated (intracuticular) and superimposed (epicuticular) waxes are summarized by Bargel et al.13 and Koch et al.10 The epicuticular waxes can be divided into three-dimensional structures of varying morphologies and an underlying twodimensional wax film. Three-dimensional epicuticular wax structures usually occur in sizes from 0.5 to 100 µm, whereas two-dimensional wax films range from a few molecular layers up to 0.5 µm. The three-dimensional waxes and their different morphologies can be characterized by scanning electron microscopy (SEM), whereas on leaf surfaces the thin wax film is rarely visible in SEM.14 Epicuticular waxes occur in different morphologies on superhydrophobic leaves.15 The most common ones are tubules, platelets, rodlets and films. Wax morphologies are strongly dependent upon the wax chemistry, which is generally a mixture of long chain hydrocarbons and, in some waxes, cyclic hydrocarbons. Wax tubules are hollow structures, which can be distinguished chemically as well as morphologically. The first type, called nonacosanol-tubules, contains high amounts of asymmetrical secondary alcohols, predominantly nonacosan-10-ol and its homologues and to a certain degree asymmetrical diols.16,17 The nonacosanol tubules are usually 0.3 to 1.1 µm in length and 0.1 (11) Nosonovsky, M.; Bhushan, B. Mater. Sci. Eng. R 2007, 58, 162. (12) Nosonovsky, M.; Bhushan, B. AdV. Funct. Mater. 2008, 18, 843. (13) Bargel, H.; Koch, K.; Cerman, Z.; Neinhuis, C. Funct. Pl. Biol. EV. ReV. 2006, 3, 893. (14) Koch, K.; Ensikat, H. J. Micron. 2008, 39, 759. (15) Neinhuis, C.; Barthlott, W. Ann. Bot. 1997, 79, 667. (16) Holloway, P. J.; Jeffree, C. E.; Baker, E. A. Phytochem. 1976, 15, 1768. (17) Jetter, R.; Riederer, M. Bot. Acta 1995, 108, 111.
10.1021/la802491k CCC: $40.75 2009 American Chemical Society Published on Web 01/08/2009
1660 Langmuir, Vol. 25, No. 3, 2009
Figure 1. SEM micrographs of tubular waxes on Lotus (Nelumbo nucifera), Nasturtium (T. majus) and sand ryegrass (L. arenarius) leaf surfaces. All tubules are hollow and are randomly orientated on the leaf surfaces, thus the terminal openings are only visible in few tubules.
to 0.2 µm in diameter.18 These tubules can be found on Lotus (Nelumbo nucifera) and Nasturtium (Tropaeolum majus) leaves, as shown in Figure 1, and are the characteristic wax type of the Ranunculaceae, and Papaveraceae.19 The second type of tubules contains high amounts of β-diketones, such as hentriacontan14,16-dione.18,20 This particular kind of wax tubule can be found in some Eucalyptus species, as for example on the leaves of Eucalyptus globulus, and within the Poaceae, for example on leaves of Sand Ryegrass (Leymus arenarius). Their length typically varies between 2 and 5 µm, and their diameter varies between 0.2 and 0.3 µm. Both types of tubules are shown in Figure 1. The different wax morphologies arise by self-assembly of the wax molecules, which has been shown by the recrystallization of waxes, which were isolated from plant surfaces.17,18,21-23 Most of the investigated waxes recrystallized in the same morphology as found on the plant surface. In all of these recrystallization studies, the wax has been dissolved in organic solvents, such as chloroform, and recrystallized after evaporation (18) Jeffree, C. E. The Fine Structure of the Plant Cuticle, in Biology of the Plant Cuticle, Eds.;M. Riederer and C. Mu¨ller, Blackwell, Oxford, pp. 11-125, 2006. (19) Barthlott, W.; Theisen, I.; Borsch, T.; Neinhuis, C. Epicuticular waxes and Vascular plant systematics: Integrating micromorphological and chemical data, in:Stuessy T. F.; Mayer V.; Ho¨randl E. (eds) Deep Morphology: Toward a Renaissance of Morphology in Plant Systematics, Reg. Veg. Gantner Verlag Ruggell/Liechtenstein, pp.189-206, 2003. (20) Meusel, I.; Neinhuis, C.; Markstadter, C.; Barthlott, W. Plant Bio. 2000, 2, 462. (21) Jetter, R.; Riederer, M. Planta 1994, 195, 257. (22) Koch, K.; Dommisse, A.; Barthlott, W. Crys. Grow. Design 2006, 6, 2571. (23) Koch, K.; Barthlott, W.; Koch, S.; Hommes, A.; Wandelt, K.; Mamdouh, W.; De-Feyter, S.; Broekmann, P. Planta 2006, 223, 258.
Bhushan et al.
of the solvent. In nonacosanol tubules, the secondary alcohols, and in particular nonacosan-10-ol, as well as the corresponding diols, are responsible for the formation of tubules. This was demonstrated in recrystallization experiments with total waxes as well as with isolated compounds.17,21 However, chemical analysis of these waxes showed that approximately 40-50% of the wax components are not related to the tubule formation. β-diketones and hydroxy-b-diketones were proved to recrystallize as tubules in artificial systems.24 Correlations between the molecular structure of β-diketones and the formation of tubules have been studied in detail by Meusel et al.20 For the wax of L. arenarius, which has been used here, it was shown that the wax mixture and isolated hentriacontane-14,16-dione crystallized into tubules.20 However, recrystallization of waxes from a solution is correlated to a very inhomogeneous distribution of the wax mass on the surface.22 One of the ways to increase the hydrophobic properties of the surface is to increase surface roughness; so roughness-induced hydrophobicity has become a subject of extensive investigations. Wenzel25 suggested a simple model predicting that the contact angle of a liquid with a rough surface is different from that with a smooth surface. Cassie and Baxter26 showed that gaseous phase including water vapor, commonly referred to as “air” in the literature, may be trapped in the cavities of a rough surface, resulting in a composite solid-liquid-air interface, as opposed to the homogeneous solid-liquid interface. These two models describe two possible wetting regimes or states: the homogeneous (Wenzel) and the composite (Cassie-Baxter) regimes. A number of artificial roughness-induced hydrophobic surfaces have been fabricated with hierarchical structures using electrodeposition, nanolithography, colloidal systems and photolithography.27-30 Molding is a low cost and reliable way of surface structure replication and can provide a precision on the order of 10 nm.31,32 Nanostructuring using self-assembly over a molded microstructure can provide tremendous flexibility in the fabrication of a variety of hierarchical structures. In this paper, objective is to mimic hierarchical structures with wax tubules such as in the Lotus plant (Nelumbo nucifera) and to study the effect of wax tubules on the wetting properties and adhesion of surfaces. We fabricated hierarchical structures using replication of a micropatterned master surface and self-assembly of two hydrophobic tubules forming waxes of T. majus and L. arenarius leaves. Traditional recrystallization of these waxes from an organic solvent would lead to inhomogeneous wax mass distribution. To overcome this, thermal evaporation of the wax was used and wax tubule formation was induced by increasing temperature and by providing a solvent vapor phase for molecular mobility. Micro-, nanostructures and hierarchical structure were also used to study their effect at different length scales on superhydrophobicitiy.
2. Material and Fabrication 2.1. Plant Waxes. The plant waxes used in this study are the nonacosanol-waxes which were isolated from mature leaves of several (24) Jeffree, C. E.; Baker, E. A.; Holloway, P. J. Origins of the fine structure of plant epicuticular waxes. Microbiology of Aerial Plant Surfaces. Dickinson, C. H. and Preece, T. F., (eds.),Academic Press, London, NewYork, San Francisco, pp. 119-158, 1976. (25) Wenzel, R. N. Indust. Eng. Chem. 1936, 28, 988. (26) Cassie, A.; Baxter, S. Trans. Faraday Soc. 1944, 40, 546. (27) Shirtcliffe, N. J.; McHale, G.; Newton, M. I.; Chabrol, G.; Perry, C. C. AdV. Mater. 2004, 16, 1929. (28) Ming, W.; Wu, D.; van Benthem, R.; de With, G. Nano Lett. 2005, 5, 2298. (29) Chong, M. A. S.; Zheng, Y. B.; Gao, H.; Tan, L. K. Appl. Phys. Lett. 2006, 89, 233104. (30) Del Campo, A.; Greiner, C. J. Micromech. Microeng. 2007, 17, R81.
Lotus-Like Biomimetic Hierarchical Structures individuals of Tropaeolum majus (L.) The plants were cultivated in the Botanical Gardens of the University of Bonn, Germany (for annual plants no accession number exist). The second type of waxes used is β-diketone waxes, isolated from mature leaves of several individuals of Leymus arenarius (L.) Hochst. The plants were cultivated in the Botanical Gardens of the University of Bonn, Germany, accession number 2961. Data referring to the chemical composition of these waxes are given by Meusel et al.,20 which used the waxes of the same plants from the Botanical Garden Bonn. For wax extraction, single leaves of each species were immersed for 2 s in Tropaeolum majus (L.) and twice for 30 s in Leymus arenarius (L.) in 300-500 mL chloroform (purity, 99.8%; Merck, Darmstadt, Germany). The resulting wax solutions were filtered and evaporated to dryness. 2.2. Fabrication of Flat and Microstructure. A two-step molding process was used to fabricate flat and microstructure, in which at first a negative is generated and then a positive.33,34 As a master template for flat surface, a flat Si surface was used. As a master template for microstructure surface, a Si surface with pillars of 14 µm diameter and 30 µm height with 23 µm pitch, fabricated by photolithography, was used.35 A polyvinylsiloxane dental wax (President Light Body Gel, ISO 4823, Polyvinylsiloxan (PLB), Coltene Whaledent, Hamburg, Germany) was applied via a dispenser on the surface and immediately pressed down with the cap of the Petri dish or with a glass plate. After complete hardening of the molding mass (at room temperature for approximately 5 min), the silicon master surface and the mold (negative) were separated. After a relaxation time of 30 min for the molding material, the negative replicas were filled up with a liquid epoxy resin (Epoxydharz L, No. 236349, Conrad Electronics, Hirschau, Germany) with hardener (Harter S, Nr 236365, Conrad Electronics, Hirschau, Germany). The liquid epoxy resin should be added near the edge of the negative replica to prevent trapped air. Specimens were immediately transferred to a vacuum chamber at 750 mTorr (100 Pa) pressure for 10 s to remove trapped air and to increase the resin infiltration through the structures. After hardening at room temperature (24 h at 22 °C, or 3 h at 50 °C), the positive replica was separated from the negative replica. The second step can be repeated to generate a number of replicas. 2.3. Fabrication of Nanostructure. Nanostructure was created by self-assembly of the T. majus and L. arenarius waxes deposited by thermal evaporation. The specimens mounted on the specimen holder using double-side tape were placed in a vacuum chamber at 30 mTorr (4 kPa pressure), 2 cm above a heating plate loaded with 500, 1000, 1500 and 2000 µg waxes. The waxes were evaporated by heating it up to 120 °C. Evaporation from the point source to the substrate occurs in straight line, thus the amount of condensed material in equal in a hemispherical region over the point of source.36 In order to estimate the amount of condensed mass, the surface area of the half-sphere is first calculated using the formula 2πr2, whereby the radius (r) represents the distance between the specimen to be covered and the heating metal with the substance to be evaporated. Next, the amount of sublimated mass per surface area can be calculated by an amount of alkane loaded on a heating plate divided by surface area. The calculated amount of material condensed on the specimen surfaces was 0.2, 0.4, 0.6 and 0.8 µg/mm2, respectively. It has been reported by Niemietz et al.37 that for the development of nanostructures by tubule formation, specimens need to be stored (31) Madou, M. Fundamentals of Microfabrication; Boca Raton: CRC Press, 1997. (32) Varadan, V. K.; Jiang, X.; Varadan,V. V. Microstereolithography and Other Fabrication Techniques for 3D MEMS; Wiley, New York, 2001. (33) Koch, K.; Dommisse, A.; Barthlott, W.; Gorb, S. Acta Biomat. 2007, 3, 905. (34) Koch, K.; Schulte, A. J.; Fischer, A.; Gorb, S. N.; Barthlott, W. Bioinsp. Biomim. 2008, 3, 046002. (35) Barbieri, L.; Wagner, E.; Hoffmann, P. Langmuir 2007, 23, 1723. (36) Bunshah, R. F., Handbook of Deposition Technologies for Films and Coatings: Science, Technology and Applications, William Andrew, Applied Science Publishers, NJ, 1994. (37) Niemietz, A.; Barthlott, W; Wandelt, K.; Koch, K. Acta Biomaterialia 2008, submitted.
Langmuir, Vol. 25, No. 3, 2009 1661
Figure 2. Schematic of a glass recrystallization chamber used for tubules formation. The filter paper placed at the bottom of the chamber was wetted with 20 mL of the solvent, and slow evaporation of the solvent was provided by placing a thin filter paper between the glass body and the cap placed above. The total volume of the chamber is about 200 cm3.
at 50 °C. The increase of temperature from 21° (room temperature) to 50 °C had a positive effect on the mobilization and diffusion of wax molecules, required for separation of the tubule forming molecules. It is also known that chemical ambient is known to have an influence on the propensity of tubule formation. In this study, the surfaces covered with evaporated waxes were stored for three days at 50 °C in a crystallization chamber as described by Jetter and Riederer21 and Meusel et al.20 and schematically shown in Figure 2. During crystallization, specimens were placed on metal posts and exposed to a solvent in vapor phase. A filter paper (Whatman grade No. 1; Pore size: 11 µm) placed below the specimens was wetted with 20 mL of the solvent, and slow diffusive loss of the solvents in the chamber was provided by placing a thin filter paper between the glass body and the lid. Solvents used were chloroform (purity, 99.8%; Mallinckrodt Inc.) and ethanol (purity, 70%) which are known to dissolve waxes.9 After evaporation of the solvent, specimens were left in the oven at 50 °C in total for seven days. 2.4. Fabrication of Hierarchical Structure. Surfaces with hierarchical structure were made by using the micropatterned epoxy replicas. These surfaces were covered with T. majus and L. arenarius waxes, in the same way as described for the development of nanostructure. 2.5. Fabrication of Flat Surfaces. In order to produce specimens with unstructured coverage of T. majus and L. arenarius waxes, flat and microstructured surfaces were covered with the wax material deposited by thermal evaporation and stored for seven days at 21 °C in a desiccator.
3. Results and Discussion 3.1. Surface Structuring by Wax Tubules. For the development of nanostructures by tubule formation of T. majus wax, specimens were stored at 50 °C with and without ethanol vapor. Figure 3 shows the scanning electron microscope (SEM) micrographs of the nanostructure and hierarchical structure fabricated with two different masses (0.6 and 0.8 µg/mm2) of T. majus. SEM micrographs in Figure 3(a) show an increase in tubule amount on flat and microstructure surfaces after deposition of higher masses of wax. The tubules of T. majus wax grown in an ethanol atmosphere are comparable to the wax morphology found on the leaves of T. majus, as shown in Figure 1. Surfaces show a homogeneous distribution of the wax mass on the specimen surfaces, and tubules provide the desired nanostructure of threedimensional tubules on flat and microstructure surfaced surfaces. The tubule morphology of T. majus wax is shown in detail in Figure 4. The tubular crystals are hollow structures, randomly orientated on the surface and embedded into an amorphous wax layer. They are randomly distributed on the surface, and their shapes and sizes show some variations. As shown in Figure 4, and based on additional specimens, the tubular diameter varied
1662 Langmuir, Vol. 25, No. 3, 2009
Bhushan et al.
Figure 3. SEM micrographs taken at 45° tilt angle (shown using two magnifications) of the nanostructure and hierarchical structure fabricated with two different mass (0.6 and 0.8 µg/mm2) of T. majus wax after storage at 50 °C (a) with and (b) without ethanol vapor.
Figure 4. SEM micrograph taken at 45° tilt angle of three-dimensional tubules forming nanostructures on the surface fabricated with 0.8 µg/ mm2 mass of T. majus wax after storage at 50 °C with ethanol vapor.
between 100 and 300 nm, their length varied between 300 and 1200 nm. However, as shown in Figure 3(b), few crystals were found on the flat and microstructure surfaces, covered with two different masses (0.6 and 0.8 µg/mm2) of T. majus wax after storage at 50 °C, without ethanol vapor. The lower mass of wax formed a granular layer, whereas higher mass of wax led to few formations of single tubules. This result shows that the formation of tubules requires the mobility of wax molecules on the surface and is primarily provided by the solvent in which the waxes were dissolved. For L. arenarius wax, no formation of tubules was found on surfaces after storage at 50 °C with ethanol vapor. The experiments were repeated with chloroform vapor, and tubule formation was found. Figure 5 shows the SEM micrographs of the nanostructure and hierarchical structure fabricated with two different masses (0.6 and 0.8 µg/mm2) of L. arenarius wax after storage at 50 °C
with chloroform vapor. An increase in tubule length, an increase in the amount of tubules, and a more upright orientation of the tubules was found after deposition of higher wax mass. The created nanostructures of L. arenarius wax with chloroform vapor are comparable to the wax crystal morphology found on the superhydrophobic leaves, as shown in Figure 1. A detail of the wax tubule morphology is shown in Figure 6. The tubular crystals are randomly orientated on the surface and embedded into an amorphous wax layer. As shown in Figure 6, and based on additional specimens, the tubular diameter varied between 200 and 300 nm, their length varied between 1500 and 4000 nm; thus, tubules of L. arenarius wax are two to five times longer than the tubules of T. majus wax. The SEM data show an influence of solvent and amount of applied wax mass on the tubule formation. In both kinds of tubules, the increase of mass led to higher amounts of tubules. Recrystallization experiments of Meusel et al.20 showed that for L. arenarius, the tubule forming component is assumed to be the β-diketone (hentriacontan-14,16-dione). It was also shown that the wax also contains alkanes (4.0%), alkyl esters (7.0%), benzyl acyl esters (1.4%), aldehydes (1.2%), primary alcohols (5.1%), fatty acids (3.2%) and other not-identified (14%) components. For T. majus wax, the secondary alcohol nonacosan-10-ol and at least 2% of diols form the tubules.38 The remaining approximately 30% of wax mass are long chain hydrocarbons (alkanes, primary alcohols, few amounts of aliphatic acids, esters, aldehydes and some not identified components) which in plants and in artificial surfaces form the basal wax layer.22 The tubule (38) Dommisse, A. Self-assembly and pattern formation of epicuticular waxes on plant surfaces; Dissertation, Rheinische Friedrich-Wilhelms Universita¨t Bonn, 2007.
Lotus-Like Biomimetic Hierarchical Structures
Langmuir, Vol. 25, No. 3, 2009 1663
Figure 5. SEM micrographs taken at 45° tilt angle (shown using two magnifications) of the nanostructure and hierarchical structure fabricated with two different mass (0.6 and 0.8 µg/mm2) of L. arenarius wax after storage at 50 °C with chloroform vapor.
Figure 6. SEM micrograph taken at 45° tilt angle of three-dimensional tubules forming nanostructures on the surface fabricated with 0.8 µg/ mm2 mass of L. arenarius wax after storage at 50 °C with chloroform vapor.
forming compounds of L. arenarius and T. majus, and their molecular formulas are shown in Table 1. The L. arenarius wax formed no tubules when the polar solvent ethanol was used, but formed tubules when nonpolar chloroform was used. If we compare the polarity of the tubules forming compounds hentriacontane-14,16-dione (L. arenarius) and nonacosan-ol (T. majus), we find twice as many oxygen atoms in L. arenarius. The interpretation of molecular interactions between the solvent and the wax molecules is complex, as many different interactions can be involved, for example hydrogen bonding, π-interactions or van der Waals forces. Thus further studies, in which solvent interactions are quantified should be performed to explain this molecular interactions with the solvents in detail. 3.2. Contact Angle and Adhesion Forces. To study the effect of nanostructure and hierarchical structures with different crystal densities on superhydrophobicity, static contact angle and contact angle hysteresis were measured. For static contact angle, droplets of about 5 µL in volume (with the diameter of a spherical droplet about 2.1 mm) were gently deposited on the surface using a microsyringe. For contact angle hysteresis, the advancing and receding contact angles were measured at the front and back of the droplet moving along the tilted surface, respectively. All measurements were repeated five times. Figure 7 shows a plot of the mean values of static contact angle and contact angle hysteresis as a function of mass of the T. majus and L. arenarius waxes deposited. As the mass of wax increased, the static contact angle and contact angle hysteresis of nanostructures with T. majus and L. arenarius waxes gradually increased and decreased, respectively. The highest static contact angle and lowest contact angle hysteresis for nanostructure are about 161° and 6° at a
mass of 0.8 µg/mm2. As nanostructures with different masses of both waxes were fabricated on microstructures, the static contact angle and contact angle hysteresis showed similar trends as those of nanostructures. The highest static contact angle and lowest contact angle hysteresis for hierarchical structure (L. arenarius) are about 169° and 2° at mass of 0.8 µg/mm2. Atomic force microscopy (AFM) was used to characterize the nanostructure fabricated using T. majus and L. arenarius waxes of 0.8 µg/mm2 after storage at 50 °C with ethanol and chloroform vapor. Statistical parameters of nanostructure (root-mean-square (rms) height, peak to valley height, and summit density (η)) were calculated and are presented in Table 2.4,39 A summit is defined as a point whose height is greater than of its four nearest neighboring points where the height difference is greater than a threshold value of 10% of rms height to avoid measurement errors. The measurement results were reproducible within ( 5%. To study the effect of structures with various length scales on superhydrophobicity, static contact angle, contact angle hysteresis and tilt angle, and adhesive forces of four structures produced using T. majus wax were measured. Nanostructures formed on flat and microstructured surfaces were fabricated using T. majus wax of 0.8 µg/mm2 after storage at 50 °C with ethanol vapor. The data are shown in Figure 8(a). Reproducibility of the data is shown by the error bars in the figure. The static contact angle of a flat surface coated with a film of T. majus wax was 112°, and increased to 164° when T. majus wax formed a nanostructure of tubules on it. On the flat specimen with a microstructure on it, the static contact angle was 154°, but increased to 171° for the hierarchical surface structure. Contact angle hysteresis and tilt angle for flat, micro- and nanostructured surfaces show similar trends. Flat surface showed a contact angle hysteresis of 61° and a tilt angle of 86°. The microstructured surface shows a reduction of contact angle hysteresis and tilt angle, but a water droplet still needs a tilt angle of 31° before sliding. As tubules are formed on the flat and microstructured surfaces, the nanostructured and hierarchical structure surfaces have low contact angle hysteresis of 5° and 3°, respectively. Adhesive force measured using a 15 µm radius borosilicate tip in an AFM also show a similar trend as the wetting properties. Adhesion force of the hierarchical surface structure was lower than either that of micro- and nanostructured surfaces because the contact between the tip and surface was lower as a result of contact area being reduced.4,39 (39) Bhushan, B. Handbook of Micro/Nanotribology, 2nd ed.; Boca Raton, FL: CRC Press, 1999.
1664 Langmuir, Vol. 25, No. 3, 2009
Bhushan et al.
Table 1. Chemical structure of the major wax components used in the study
Table 2. Roughness Statistics for Nanostructure Surfaces Measured Using an AFM (Scan Size 10 µm × 10 µm)a
nanostructure (Tropaeolum majus wax) nanostructure (Leymus arenarius wax) a
rms height (nm)b
peak-to-valley height (nm)
η (µm-2)c
180
1570
0.570
379
5590
1.10
2
Nanostructures were fabricated with a 0.8 µg/mm mass of T. majus and L. arenarius waxes after storage at 50°C with ethanol and chloroform vapors, respectively. b rms, root mean square. c η, summit density.
Table 3. Summary of the Roughness Factor (Rf), Fractional Liquid-Air Interface (fLA), Static Contact Angles, and Contact Angle Hysteresis Measured and Calculated on the Various Surfaces.a The Variation Represents ( 1 Standard Deviation Rf Tropaeolum majus Flat Nanostructure Microstructure Hierarchical structure Leymus arenarius Flat Nanostructure Microstructure Hierarchical structure
fLA
Static contact Contact angle angle (deg) hysteresis (deg)
11 0.86 3.5 0.71 14.5 0.96
112 ( 2.1 164 ( 2.4 154 ( 1.9 171 ( 1.3
61/ ( 2.1 5 ( 1.1 27 ( 1.5 3 ( 0.7
16 0.87 3.5 0.71 19.5 0.96
108 ( 1.6 159 ( 1.5 153 ( 2.1 169 ( 1.1
55// ( 2.1 6 ( 1.1 30 ( 1.9 2 ( 0.7
/ Advancing and receding contact angles are 132° and 71°, respectively. Advancing and receding contact angles are 123° and 68°, respectively. a Nanostructures and hierarchical structures were fabricated with 0.8 µg/ mm2 of T. majus and L. arenarius waxes after storage at 50°C with ethanol and chloroform vapors, respectively. //
Figure 7. Static contact angle and contact angle hysteresis as a function of mass of T. majus and L. arenarius waxes deposited for nanostructures and hierarchical structures.
Next, four structures produced using L. arenarius were measured. Nanostructures formed on flat and microstructured surfaces were fabricated using L. arenarius wax of 0.8 µg/mm2 after storage at 50 °C with chloroform vapor. The data are shown in Figure 8(b). Reproducibility of the data is shown by the error bars in the figure. The data show similar trends with those of T. majus wax. The static contact angle of flat surface coated with a film of L. arenarius wax was 108°, and increased to 159° when
L. arenarius wax formed a nanostructure of tubules on it. On the flat specimen with a microstructure on it, the static contact angle was 153°, but increased to 169° for the hierarchical surface structure. Both nanostructured and hierarchical structure surfaces show low contact angle hysteresis of 6° and 2°, respectively. Adhesive forces of various structures with L. arenarius wax measured using a 15 µm radius borosilicate tip in an AFM also show similar trend as those with T. majus wax. The measured static contact angle and contact angle hysteresis values of Lotus leaf are 164° and 3°, respectively. Hierarchical structure has contact angle properties superior to Lotus leaf. Further optimization can improve these values. As indicated earlier, hierarchical structure can lead to air pocket formation which allows desirable contact angles. In order to identify propensity of air pocket formation for the various surfaces, roughness factor (Rf) and fractional liquid- air interface (fLA) are needed. The Rf for the nanostructure was calculated using the AFM map.40 The calculated results were reproducible within ( 5%. The Rf for the microstructures was calculated for the geometry of flat-top, cylindrical pillars of diameter D, height H, and pitch P distributed in a regular square array. For this case, roughness factor for the microstructures, (Rf)micro ) (1+πDH/P2). The roughness factor for the hierarchical structure is the sum of (Rf)micro and (Rf)nano. The values calculated for various surfaces are summarized in Table 3. For the calculation of fLA, we make the following assumptions. For the microstructure, we consider that a droplet much larger than the pitch value P is in contact with only the flat top of the (40) Burton, Z.; Bhushan, B. Ultramicroscopy 2006, 106, 709.
Lotus-Like Biomimetic Hierarchical Structures
Langmuir, Vol. 25, No. 3, 2009 1665
Figure 8. Bar chart showing the measured static contact angle, contact angle hysteresis and tilt angle on various structures fabricated with 0.8 µm/mm2 mass of (a) T. majus wax after storage at 50 °C with ethanol vapor and (b) L. arenarius wax after storage at 50 °C with chloroform vapor. The bar chart also shows adhesive forces for various structures, measured using a 15 µm radius borosilicate tip. The error bars represent ( 1 standard deviation.
pillars in the composite interface and the cavities are filled with air. For the nanostructure, only the higher crystals are assumed to come into contact with a water droplet. For the microstructure, there are fractional flat geometrical areas of the solid-liquid and liquid-air interfaces under the droplet, (fLA)micro ) 1 - (πD2/ 4P2).1 The fractional geometrical area of the top surface for the nanostructure was calculated from an SEM micrograph with a top view (0° tilt angle). The SEM image was converted to a high-contrast black-and-white image using Adobe Photoshop. The increase in contrast in the SEM image eliminates the smaller tubule structures, which were visible in the original SEM image. The higher crystals led to white signals in the SEM image. The fractional geometrical areas of the top surface with T. majus and L. arenarius waxes were found to be 0.14 and 0.13, leading to fLA ) 0.86 and 0.87, respectively. The calculated results were reproducible to within (5%. For the hierarchical structure, the
fractional flat geometrical area of the liquid-air interface, (fLA)hierarchical ) [1 - (πD2/4P2)][1 - (fLA)nano]. The values calculated for various surfaces are summarized in Table 3. The roughness factor and fractional liquid-air interface of hierarchical structure are higher than those of nano- and microstructures. These results show that air pocket formation in hierarchical structured surfaces occurs, which further decreases the solid-liquid contact and thereby reduces the contact angle hysteresis and tilt angle. To verify further the effect of hierarchical structure on the propensity of air pocket formation, the evaporation experiments with a droplet on microstructured and hierarchical structure fabricated with a 0.8 µm/mm2 mass of T. majus wax with ethanol vapor at 50 °C were performed.41 Figure 9 shows successive (41) Jung, Y. C.; Bhushan, B. J. Microsc. 2008, 229, 127.
1666 Langmuir, Vol. 25, No. 3, 2009
Bhushan et al.
Figure 9. Evaporation of a droplet on microstructured and hierarchically structured surfaces fabricated with a 0.8 µm/mm2 mass of T. majus wax after storage at 50 °C with ethanol vapor. The initial radius of the droplet was about 950 µm. The time interval between the first two photographs was 180 s, and the interval between the latter two was 60 s. As the radius of the droplet reached 425 µm (foot print ) 836 µm) on the microstructured surface, the transition from the Cassie-Baxter regime to the Wenzel regime occurred, as indicated by the arrow. On the hierarchically structured surface, air pockets, visible at the bottom of the droplet, existed until the droplet evaporated completely.
photographs of a droplet evaporating on two structured surfaces. On the microstructured surface, the light passes below the droplet and air pockets can be seen, so the droplet is initially in the Cassie-Baxter regime. When the radius of the droplet decreases to 425 µm, the air pockets are not visible anymore and the droplet is in the Wenzel regime. This transition results from the impalement of the droplet in the patterned surface, characterized by a smaller contact angle. For the hierarchical structure, an air pocket was clearly visible at the bottom of the droplet throughout, and the droplet was in a hydrophobic state until the droplet evaporated completely. This suggests that a hierarchical structure with nanostructures prevents liquid from filling the gaps between the pillars.
4. Conclusions Biomimetic hierarchical structures were produced by the replication of a micropatterned silicon surface using an epoxy resin and by the recrystallization of tubular waxes of T. majus and L. arenarius leaves, which self-assemble on the substrates into 3D crystals. The technique introduced here involves a lowcost molding process to replicate microstructures via the selfassembly of wax to create different tubular nanostructures. This two-step process provides flexibility in the fabrication of a variety of hierarchical structures. For the optimization of crystal density
for high static contact angle and low contact angle hysteresis, various nanostructures produced by 3D tubular crystals were studied by changing the wax mass. The influence of these structures on the static contact angle, contact angle hysteresis, tilt angle, and air pocket formation as well as adhesive force was studied. The static contact angle and contact angle hysteresis of flat surfaces are on the order of 110 and 60°, respectively, whereas tubular crystals on nanostructures and hierarchical structure led to static contact angles on the order of 160 and 170° and contact angle hysteresis on the order of 5 and 2°, respectively. The optimal structural parameters for superhydrophobicity and low static contact angle hysteresis, superior to those of natural plant leaves including lotus leaves, have been identified and provide a useful guide for the development of biomimetic superhydrophobic surfaces. The flexible, low-cost technique advanced in this article demonstrates that surfaces with lotus-like wettability can be produced in the laboratory for further investigation of hierarchical structure properties. Note Added after ASAP Publication. This article posted ASAP on January 8, 2009. In the Introduction section, paragraph 2, sentence 3; reference 8 has been deleted. The correct version posted on January 27, 2009. LA802491K
