Article pubs.acs.org/Langmuir
Superhydrophobic and Adhesive Properties of Surfaces: Testing the Quality by an Elaborated Scanning Electron Microscopy Method Hans J. Ensikat,* Matthias Mayser, and Wilhelm Barthlott Nees Institute for Biodiversity of Plants, University of Bonn, Meckenheimer Allee 170, 53115 Bonn, Germany ABSTRACT: In contrast to advancements in the fabrication of new superhydrophobic materials, the characterization of their water repellency and quality is often coarse and unsatisfactory. In view of the problems and inaccuracies, particularly in the measurement of very high contact angles, we developed alternative methods for the characterization of superhydrophobic surfaces. It was found that adhering water remnants after immersion are a useful criterion in determining the repellency quality. In this study, we introduce microscopy methods to detect traces of water-resembling test liquids on superhydrophobic surfaces by scanning electron microscopy (SEM) or fluorescence light microscopy (FLM). Diverse plant surfaces and some artificial superhydrophobic samples were examined. Instead of pure water, we used aqueous solutions containing a detectable stain and glycerol in order to prevent immediate evaporation of the microdroplets. For the SEM examinations, aqueous solutions of lead acetate were used, which could be detected in a frozen state at −90 °C with high sensitivity using a backscattered electron detector. For fluorescence microscopy, aqueous solutions of auramine were used. On different species of superhydrophobic plants, varying patterns of remaining microdroplets were found on their leaves. On some species, drop remnants occurred only on surface defects such as damaged epicuticular waxes. On others, microdroplets regularly decorated the locations of increased adhesion, particularly on hierarchically structured surfaces. Furthermore, it is demonstrated that the method is suitable for testing the limits of repellency under harsh conditions, such as drop impact or long-enduring contact. The supplementation of the visualization method by the measurement of the pull-off force between a water drop and the sample allowed us to determine the adhesive properties of superhydrophobic surfaces quantitatively. The results were in good agreement with former studies of the water repellency and contact angles. In contrast to contact angle measurements, the acqusition of SEM images with high resolution and wide depth of sharpness gives better insight into the wetting behavior and susceptibility of the structural elements of the superhydrophobic surfaces.
■
INTRODUCTION Superhydrophobic surfaces are found on many plant and animal species. Notably, the emerging of their self-cleaning properties, the so-called Lotus effect,1,2 enhanced the activities in elucidating the fundamentals and the development of technical superhydrophobic surfaces.3 The development of robust, durable self-cleaning surfaces is challenging because the superhydrophobicity is based on delicate micro- or nanostructures. Recently developed alternative types of self-cleaning and antiadhesive surfaces, such as photocatalytic surfaces4,5 and liquid hydrophobic coatings,6 may become serious competition. Thus, the improvement of commercial superhydrophobic materials requires reliable diagnostic methods to analyze their quality and alterations. The quality of superhydrophobic surfaces depends on various factors such as the apolar character of the material, the surface topography (nano-, micro-, or hierarchical structure), and the occurrence of surface defects or discontinuities. The most common method of characterization of superhydrophobic surfaces is the measurement of the contact angles of water drops resting on the surface, often in combination with the tilting angle, which causes the drop to roll off. However, it is © 2012 American Chemical Society
well known that the measurement of very high contact angles with standard equipment can be quite inexact, notably, with the sessile-drop method. Slight inaccuracies in the determination of the baseline or the drop shape cause errors in the contact angles, which are often larger than the contact angle differences on diverse superhydrophobic samples.7,8 These inaccuracies increase on rough, uneven surfaces. Furthermore, variations in the water drop size and the gravitation influence complicate the comparability of contact angle measurements. Only methods that exclude the influence of gravitation seem to be capable of measuring extremely high contact angles of >179°, as demonstrated by Gao and McCarthy,9 who observed the affinity between a drop and a surface during attachment and release. Sophisticated analyses of the quality of and imperfections in superhydrophobic surfaces and their causes require precise measurements and high-resolution optical characterization. Received: July 16, 2012 Revised: September 11, 2012 Published: September 14, 2012 14338
dx.doi.org/10.1021/la302856b | Langmuir 2012, 28, 14338−14346
Langmuir
Article
and Tropaeolum majus L.. Except for Alocasia, the top sides of the leaves were examined. Technical Samples. Microstructured silicon wafer samples with a hydrophobic coating were used as a model for technical surfaces. They had a regular structure of pillars with a 10 × 10 μm2 area and a 20 μm height. The hydrophobic coatings were applied by physical vapor deposition (PVD) of either paraffin to obtain a single-scale-structured sample or natural wax from Tropaeolum majus leaves following recrystallization at 50 °C to obtain a dual-scale-structured sample with wax tubules on top of the pillars.16 Test Liquids. Aqueous solutions of heavy-metal salts were used for SEM examinations with a backscattered-electron (BSE) detector, and solutions of auramine were used for SEM examinations with a cathodoluminescence (CL) detector and for FLM. The addition of glycerol inhibited the evaporation of the liquid and suppressed the formation of coarse crystallites during freezing for low-temperature SEM. The so-called Pb solution contained 30% lead acetate [Pb(CH3COO)2·3H2O] and 23% glycerol in aqua dest., and the auramine solution contained 1% auramine and 20% glycerol in aqua dest. Furthermore, a refined test solution with lead acetate (30%) and sugar (sucrose) (23%) instead of glycerol has been tested. Its wetting properties are closer to those of water. The surface tension of the solutions was determined using the bubble pressure method by comparison with water. We measured 68 mN/m for the Pb solution, 62 mN/m for the auramine solution, and 72 mN/m for the lead acetate/sugar solution (water: 73 mN/m). Methods. Equipment. Scanning electron microscopy was carried out using a Cambridge Stereoscan S200 SEM or an LEO 1450 SEM (Cambridge, U.K.), both equipped with custom-made cooling holders for low-temperature examinations. Two different BSE detectors were used for SEM material-contrast imaging: a Link Tetra and a scintillator detector (Robinson type) together with the standard secondary electron (SE) detector. A CL detector was used to visualize auraminestained liquid by SEM. The SEMs operated in the standard highvacuum mode. FLM was performed with a Zeiss Axio Imager.M2m (Carl Zeiss GmbH, Oberkochen, Germany) using filter set 38 (green fluorescence). Contact angles (CA) of water drops on the sample surfaces were measured with a contact angle measurement system (OCA 30-2, Dataphysics Instruments GmbH, Filderstadt, Germany) using drops of 10 μL. The adhesion of water drops on the samples was measured with a self-developed device by recording force−distance curves while the drop was attached to and retracted from the surface at a constant velocity of 40 μm/s. Hemispherical drops with a diameter of 3.0 mm were pressed against the surface after the initial contact for another 0.4 mm, causing a deformation of the drop and slight pressure. Then the pull-off forces during retraction were determined. The force data were acquired using a 12-bit A/D converter and recorded at 50 measurements/s. SEM Preparation and Microscopy. The samples (e.g., pieces of fresh leaves) were mounted in the coolable SEM sample holder. For the standard experiments, a drop of the test liquid was gently pressed onto the sample surface and removed. (Varying applications of the liquid are described in the Results section.) The sample holder was then closed with a cover, cooled with liquid nitrogen to below −100 °C, and inserted into the SEM. (For details of the cooling procedure, see ref 15.) After good vacuum was obtained, the cover of the holder was removed. At temperatures between −90 and −60 °C, the samples were examined without metal coating. In this temperature range, the internal water of the plant samples provides sufficient electrical conductivity to avoid charging effects that are too strong. Images were recorded using the secondary electron (SE) and the backscattered electron (BSE) signal.
Superhydrophobic materials have various features that may be used for characterization: - when submerged in water, they are covered with a layer of air; - after re-emerging from the water, the surface appears to be dry; - water drops on the surface roll off at slight tilting angles because of the weak adhesion; and - sessile water drops have high contact angles. Which one of these features can be best used for the analysis of superhydrophobic surfaces? The measurement of the airlayer thickness and its durability are important for certain applications but do not indicate the absence or presence of surface defects. The contact angles (static and advancing) depend predominantly on the topography and the hydrophobic character of the material but are only slightly affected by small defects that cause drop pinning. The detection of the causes of pinning requires additional measures such as the observation of the receding contact angles or the tilting angles, which depend on the adhesion of the drops. However, these measurements are sometimes not very accurate. The slight adhesion between a drop and a surface can be measured more exactly by recording calibrated force−distance curves.10,11 This is also suitable on rough, uneven surfaces and excludes personal reading errors. Besides these spot measurements, sample areas can be analyzed with moving or scanning force sensors, as shown by Ng and Panduputra,12 or by the observation of the contact line during immersion and withdrawal from water. The most simple but distinct method of checking the quality of superhydrophobic surfaces is to observe whether they are completely dry after dipping them into water. Adhering water drops indicate defects, but are the superhydrophobic surfaces really completely dry when they are lifted out of the water, even on a microscopic scale? Microscopic water droplets will of course evaporate immediately, and then the surface is really dry. Krumpfer et al.13 treated superhydrophobic pillar structures with an organic “ionic liquid” that was characterized by high contact angles and low vapor pressure. They found via SEM remnants of the liquid on top of the pillars. Cohen et al.14 studied in detail the dewetting dynamics of hydrophobic surfaces and discussed the importance of a detailed knowledge of the adhesive properties. The imaging of microscopic traces of aqueous liquid remnants on superhydrophobic surfaces in combination with the measurement of adhesion forces could provide the feasibility to detect defects and the reasons for adhesion at high resolution in a 2D field instead of point measurements. Thanks to our experiences in the scanning electron microscopy of plantsinclusive low-temperature SEM of frozen hydrated samples without metal coating15we were able to examine the remnants of the aqueous test liquids, which have similar wetting properties to water, using SEM as well as fluorescence light microscopy (FLM). The results are in accordance with pull-off force measurements and show the variability of the surface properties in great detail.
■
EXPERIMENTAL SECTION
■
Material. Plant Samples. Eight plant species with superhydrophobic leaves, which have been used for many years as standard samples because of the variations in their surface morphologies, were chosen for this study: Alocasia macrorrhiza L., Brassica oleracea var. gongylodes L., Colocasia esculenta L., Euphorbia myrsinites L., Iris germanica L., Nelumbo nucifera Gaertn., Salvinia molesta D. Mitchell,
RESULTS When a water drop is attached to a flat surface and removed from it vertically, usually a remaining droplet can be found on the surface; its size depends on the wettability of the surface 14339
dx.doi.org/10.1021/la302856b | Langmuir 2012, 28, 14338−14346
Langmuir
Article
greater adhesion, resulting in slightly lower contact angles compared to those of water but similar superhydrophobic behavior. Drop Remnants on Different Superhydrophobic Surfaces. Two types of regularly structured superhydrophobic test samples were used: silicon pillars coated with a paraffin layer formed a single-scale structure and pillars coated with wax tubules formed a dual-scale structure. Both were tested to be superhydrophobic but, as expected, with differences in their contact and tilting angles (Table 1). The two samples displayed differences in their adhesion behavior for the Pb solution as well as for water. After contact with the Pb solution, the singlescale-structured sample carried microdroplets on top of all pillars that had been in contact with the liquid (Figure 3a,b). In the margin of the contact area, the remaining droplets were small (ca. 2 μm diameter), and near the center they reached ca. 5 μm diameter. In contrast, no remnants of the liquid were found on the dual-scale-structured test sample (Figure 3c,d). The difference in superhydrophobicity was also confirmed by measurements of the pull-off force, the adhesion during retracting the sample from the test liquid (water or Pb solution, Figure 4). The force−distance curves show the forces measured during the approach and retraction of the sample from the drop. The negative values (up to −300 μN) indicate a repulsive force due to the deformation of the water drop and are characteristic of superhydrophobic surfaces; the maximal positive value during retraction represents the adhesion or pulloff force that arises before the contact breaks off. A collection of superhydrophobic plant leaves with different surface structures was used for this study. Plant leaves show a wide variability in surface structures, such as various epidermis cell shapes (microstructure), epicuticular wax crystals (nanostructure), varying chemical composition of the wax, and coarse inhomogeneities such as leaf veins. All of these parameters seem to affect the wettability. The tests for the remaining microdroplets demonstrated significant differences for the samples. The surfaces of four species (Brassica, Colocasia, Nelumbo, and Tropaeolum; Figure 5) were largely free of drop remnants, but spots with microdroplets occurred sporadically on most samples (e.g., on leaf veins or other locations where the wax coating was damaged). The other plant species examined (Euphorbia, Iris, and Alocasia; Figure 6) had drop remnants on every epidermal cell that had been in contact with the liquid, even the hierarchically structured surface of Euphorbia with wax platelets on epidermis papillae.
(Figure 1). Even on hydrophobic materials such as paraffin a small droplet remains. On superhydrophobic surfaces, no water
Figure 1. Image sequences from a video recording showing (I) the approach of a water drop, (II) contact, (III) retraction, and (IV) separation of the drop from the samples. After separation, part of the drop remains on the flat surfaces. (a) On the steel sample, a larger drop remains than (b) on the hydrophobic paraffin. (c) On the superhydrophobic Iris germanica leaf, no remaining drop is visible.
remnants are visible with the naked eye, except for surface defects. If the surfaces are placed in contact with the nonevaporating test solution, which contains lead acetate and glycerol, then it is possible to detect drop remnants or microdroplets with high resolution using the SEM, preferably by low-temperature SEM (LT-SEM). Microdroplets of the Pb solution on a superhydrophobic Iris leaf are shown in Figure 2. A good detectability of the heavy-metal solution with the SEM was achieved with a BSE detector because of the strong material contrast. The lead-containing droplets appear bright in the BSE image with good contrast against the organic tissue. The content of glycerol in the solution inhibited the evaporation of the liquid and stabilized the shape of the droplets during freezing. The wetting properties of the test solutions were tested by measurements of the contact angles, tilting angles, and pull-off forces on hydrophobic and superhydrophobic samples in comparison to water (Table 1). The Pb solution had
Figure 2. (a) SE and (b) BSE image of microdroplets on an Iris leaf after contact with the Pb solution. Becauase of the heavy-metal content, the droplets can be reliably recognized in the BSE image as bright dots. The inserted high-magnification BSE image shows very small droplets besides the main droplet. 14340
dx.doi.org/10.1021/la302856b | Langmuir 2012, 28, 14338−14346
Langmuir
Article
Table 1. Comparison of the Wetting Properties of the Test Solutions with Watera water
Pb solution
auramine
sample
CA (deg)
TA (deg)
POF (μN)
CA (deg)
TA (deg)
POF (μN)
paraffin (smooth surface) pillars with paraffin pillars with wax tubules Colocasia esc. leaf
109.5 (±1.3) 160.5 (±3.7) 159.2 (±2.0) ca. 160
20.5 (±2.7) 15.5 (±1.3)
