Robust Omniphobic Surfaces
(Ian Burgess, Spring 2012)
1. A. Tuteja, W. Choi, J.M. Mabry, G.H. McKinley, R.E. Cohen, "Robust Omniphobic Surfaces", PNAS, 2008, 105, 18200-18205.
2. T.S. Wong, S.H. Kang, S.K.Y. Tang, E.J. Smythe, B.D. Hatton, A. Grinthal & J. Aizenberg, "Bioinspired self-repairing slippery surfaces with pressure-stable omniphobicity", Nature, 2011, 447. 443-447.
Designing liquid-repellant surfaces is a technological challenge that is important to many industries, cars and airplanes, buildings, to clothing etc. Inspired by the lotus leaf, one popular approach to designing liquid-repellant surfaces is to engineer very rough or porous surfaces having low surface energy. Under certain conditions a drop of liquid may be able to rest on top of the porous structure, while preserving a trapped air region within most of the porosity. This regime of minimal contact area, often referred to as the Cassie-Baxter regime, is typically characterized by very large (>150 degrees) apparent contact-angles and minimal droplet adhesion. While the lotus-leaf approach can be very effective for designing water-repellant surfaces, it is difficult to design surfaces that are repellant to low-surface-tension liquids (e.g. ethanol, hexane, etc.) using this approach. This is because low surface-tension liquids are typically wetting (i.e. the intrinsic contact angle is less than 90 degrees) to even the lowest energy surfaces. Thus it is generally energetically favorable for the liquid to penetrate the roughness rather than preserve the trapped air layer.
To overcome this challenge, Tuteja et al. exploit pinning to create surfaces that support Cassie states even for very low-surface tension liquids. The design strategy is outlined schematically in the figure below.
The geometries in the figure above share the characteristic of reentrant curvature, i.e. that the air voids widen as they descend. This forces a liquid front penetrating the structure to increase the liquid-air surface area as it penetrates, creating an energy tradeoff between the favorable wetting of the solid surface (since the intrinsic contact angle is less than 90 degrees) and the energetically non-favorable creation of liquid-air interface. This tradeoff allows metastable, non-wetting equilibrium conditions to exist, as shown in the figure above.
Thus on surfaces with this type of reentrant geometry, a drop of even low-surface-tension liquid (i.e. one that makes a small intrinsic contact angle with the surface chemistry) can be suspended in a Cassie-like state, displaying a high apparent contact angle and low effective contact area. However, since these states are metastable (i.e. not the global free-energy minimum), a certain amount of pressure applied to these drops will cause them to spread into the porosity. This is why the paper analyzes the robustness of the surfaces' omniphobicity as a function of the geometry. The authors quantify the robustness based on two parameters, H, which is the ratio of the pressure required to sag the pinned liquid contact line until the point where it touches a feature below (i.e. the base of the posts or the next row of fibers) and a reference pressure, defined by the surface tension and capillary length, which estimates the characteristic pressure-difference across the solid-liquid-air interface for typical (~mm scale) droplets. They introduce a second robustness parameter, T, which expresses the critical pressure (divided by the reference pressure) in terms of a critical sagging angle rather than a critical height.
The authors then design two different types of robust omniphobic surfaces. The first consists of electrospun fibers made of a polyhedral oligomeric silsesquioxane (POSS) in which the silsesquioxane core is surrounded by perfluorodecyl groups, which give the fibers a low surface energy and maximize the intrinsic contact angle with respect to most liquids. These fibers, when electrospun, also form a network that is characterized by reentrant geometry as shown in Fig. 1A. Below, a lotus leaf covered with electrospun fibers is shown supporting a high apparent contact angle for drops of water methanol, methylene iodide and octane.
The second geometry the authors design is what they call "micro hoodoos". This structure is shown below, along with a plot of the measured apparent advancing and receding contact angles for liquids with a wide range of surface tensions. These surfaces, fabricated in silicon and then functionalized with perfluoroalkane surface groups, also display robust high apparent contact angles for both high and low-surface-tension liquids.
This paper illustrates the importance of metastable states in the analysis of wetting behavior. If one considers the geometry of the electrospun fiber-coated surfaces, they are not atypical and in fact are similar to what the local geometry looks like in most fabrics. In fact, metastable superhydrophobicity can be observed in many garments on a rainy day. Small droplets will sometimes bead up on a coat that is not waterproof (i.e. not hydrophobic). However, the metastability of this beading can be seen by pressing the drops with your hands and observing them wick entirely into the structure rather than beading up again.
A second point that is worth making is that it is worth thinking about what parameter describing hydrophobicity (or oleophobicity) is actually most important for effective function in most applications. Superhydrophobic surfaces are sought after for many different applications and the figure of merit may be different for each application. One common figure of merit, however, is rolloff angle or the ability of a surface to shed droplets (and the pressure-stability of this property). It is worth noting that this figure of merit is more affected by high contact-angle hystereis than low apparent contact angle. This relationship was shown very well by Wong et al. in their paper bioinspired self-repairing slippery surfaces with pressure-stable omniphobicity, where they create surfaces that display low contact angles with low-surface-tension-liquids, but also exceptionally low contact-angle hysteresis and thus very low rolloff angles. While the surfaces described by Tuteja et al. have low contact angle hysteresis for higher surface-tension-liquids, this breaks down for the lowest surface tension liquids, as shown for example for micro hoodoos in the figure above. This is one disadvantage of exploiting locally pinned contact lines to produce the superomniphobic effect: the more the contact line is locally pinned, the more globally pinned the droplet will be.