Article pubs.acs.org/Langmuir
Evaporation Kinetics of Sessile Water Droplets on Micropillared Superhydrophobic Surfaces Wei Xu,† Rajesh Leeladhar,† Yong Tae Kang,‡ and Chang-Hwan Choi*,†,‡ †
Department of Mechanical Engineering, Stevens Institute of Technology, Hoboken, New Jersey 07030, United States Department of Mechanical Engineering, Kyung Hee University, Yong In, Gyeong-Gi, 446-701, Korea
‡
ABSTRACT: Evaporation modes and kinetics of sessile droplets of water on micropillared superhydrophobic surfaces are experimentally investigated. The results show that a constant contact radius (CCR) mode and a constant contact angle (CCA) mode are two dominating evaporation modes during droplet evaporation on the superhydrophobic surfaces. With the decrease in the solid fraction of the superhydrophobic surfaces, the duration of a CCR mode is reduced and that of a CCA mode is increased. Compared to Rowan’s kinetic model, which is based on the vapor diffusion across the droplet boundary, the change in a contact angle in a CCR (pinned) mode shows a remarkable deviation, decreasing at a slower rate on the superhydrophobic surfaces with less-solid fractions. In a CCA (receding) mode, the change in a contact radius agrees well with the theoretical expectation, and the receding speed is slower on the superhydrophobic surfaces with lower solid fractions. The discrepancy between experimental results and Rowan’s model is attributed to the initial large contact angle of a droplet on superhydrophobic surfaces. The droplet geometry with a large contact angle results in a narrow wedge region of air along the contact boundary, where the liquid−vapor diffusion is significantly restricted. Such an effect becomes minor as the evaporation proceeds with the decrease in a contact angle. In both the CCR and CCA modes, the evaporative mass transfer shows the linear relationship between mass2/3 and evaporation time. However, the evaporation rate is slower on the superhydrophobic surfaces, which is more significant on the surfaces with lower solid fractions. As a result, the superhydrophobic surfaces slow down the drying process of a sessile droplet on them.
■
INTRODUCTION Superhydrophobic surfaces1−4 have attracted great interest because of their extreme water-repellent surface property for many potential applications including self-cleaning,5−7 hydrodynamic friction reduction,8−10 anti-icing,11−13 anticorrosion,14−16 biotechnology,17−19 thermal systems,20−24 and micro- and nanodevices.25−27 In many applications as such, they mostly deal with droplets that are open to the atmosphere and evaporative. Thus, the study of the evaporation kinetics and wetting behaviors of liquid droplets on superhydrophobic surfaces is critical to the design and applications of such surfaces for proper operation. To date, the droplet evaporation kinetics on hydrophilic or hydrophobic surfaces has been studied experimentally in many works, and theoretical kinetics models have also been developed.28−41 For example, Rowan et al. developed an evaporation kinetic model for a sessile droplet placed on a substrate based on the diffusion of the vapor across the boundary using Fick’s law.30 Deegan et al. considered that the vapor diffusing in the evaporation should quickly approach a steady-state concentration profile, which would obey the steady-state diffusion equation.31 The vapor concentration distribution above an evaporating droplet, which would be mathematically equivalent to that of a charged conductor, was examined to predict the droplet characteristics, such as the contact angle, contact radius, and evaporation rate.31,33,38 © XXXX American Chemical Society
Recently, Nguyen et al. also developed a model for the evaporation kinetics in constant contact radius and constant contact angle modes.41 Although such models have been extensively studied on hydrophilic and hydrophobic surfaces with experimental verifications, the study of superhydrophobic surfaces and comparison with the theoretical models has still been limited,19,21,42−49 demanding more extensive and systematic studies for a clearer understanding. It has typically been observed that three distinct evaporation modes appear sequentially during sessile droplet evaporation on a superhydrophobic surface. As illustrated in Figure 1, they include a constant contact radius (CCR) mode (or a pinning mode) with a gradual decrease of the contact angle (Figure 1a), a constant contact angle (CCA) mode (or a receding mode) with a gradual decrease in the contact radius (Figure 1b), and a mixed mode with simultaneous decreases in both the contact angle and the contact radius (Figure 1c). When the droplet shrinks significantly during evaporation and the internal Laplace pressure of the droplet increases more than the critical capillary pressure in sustaining the liquid−gas meniscus, a wetting transition from a dewetting (Cassie−Baxter) state50 to a Received: February 4, 2013 Revised: March 25, 2013
A
dx.doi.org/10.1021/la400452e | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
methanol, and deionized (DI) water, dried with flowing nitrogen gas, and dehydrated at 110 °C on a hot plate. After adhesion promoter hexamethyldisilazane (HMDS) was applied to the wafer, photoresist (SPR 3012, Shipley L.L.C. Rohm and Haas Electronic Materials, Marlborough, MA, USA) was spin-coated at 2000 rpm for 1 min to produce a film thickness of ∼2 μm. The substrate was then exposed using a mask aligner (MA-6, SUSS MicroTec, Garching, Germany) in soft-contact mode after soft baking at 95 °C for 1 min and developed in developer solution (MF-319, Shipley L.L.C. Rohm and Haas Electronic Materials, Marlborough, MA, USA) for 1 min. After the microdot patterns of photoresist were obtained on the substrate to serve as an etch mask, a deep reactive-ion etching (DRIE) process at a cryogenic temperature of −100 °C (Oxford PlasmaLab 100 ICP Si etcher, Oxford Plasma Technology Inc.) was employed to create the high-aspect-ratio micropillar structures of silicon with different pattern periodicities. After the DRIE process, the photoresist layer was removed with acetone. The silicon substrate was then rinsed with methanol and DI water, dried with nitrogen gas, and dehydrated at 110 °C on a hot plate. The micropillar patterns were made to have the same diameter (5 μm) and height (25 μm) but varying spaces between pillars (5, 10, 20, and 50 μm) appearing in a square array to have the area fractions of the solid−liquid interface (called a solid fraction) in a dewetting (Cassie−Baxter) state from 0.20 to 0.09 to 0.03 to 0.01, denoted as Φ0.20, Φ0.09, Φ0.03, and Φ0.01, respectively (Figure 2).
Figure 1. Schematics of typical evaporation modes occurring sequentially during the evaporation of a sessile droplet of water on a superhydrophobic surface: (a) a constant contact radius (CCR) mode, (b) a constant contact angle (CCA) mode, and (c) a mixed mode. As the droplet size is reduced during the evaporation process, a wetting transition (d) can also take place when the internal Laplace pressure of a droplet exceeds the capillary pressure of the hydrophobic surface structures. The changes in the contact angle (θ) and the contact radius (r) of a droplet are indicated with red dotted arrows.
wetting (Wenzel) state51 can also take place (Figure 1d), which typically results in a substantial geometric change in the droplet (e.g., a sudden decrease in the contact angle with an increase in the contact radius).52 There has been an attempt to analyze the change in such a droplet profile (e.g., contact angle and contact radius) and the droplet mass based on the kinetics models.28−30,32,34,36,40,41 However, the previous models were mainly developed for a sessile droplet on a planar hydrophilic or hydrophobic surface. Thus, it is necessary to validate, especially by experimental investigation, whether the previous models can simply be extended to predict the kinetic behaviors of a sessile droplet on patterned superhydrophobic surfaces properly. Moreover, the influence of the structural morphology of superhydrophobic surfaces on the evaporation kinetics of sessile droplets has not yet been clearly understood. In this work, we have experimentally investigated the evaporation kinetics of sessile droplets of water on superhydrophobic surfaces with systematically varied micropillar patterns and compared the experimental results to Rowan’s kinetic model. The impacts of the structural morphology of the superhydrophobic surface patterns on the various aspects of evaporation kinetic behaviors have also been analyzed, including the duration of evaporation modes, the evolution of the contact angle in the CCR mode, the evolution of the contact radius in the CCA mode, and the evaporative mass transfer rate in each mode.
■
Figure 2. Scanning electron microscope images (45° tilted view) of the micropillar-patterned superhydrophobic surfaces. The square array of the micropillar structures on the surfaces was designed to have the solid fraction (Φ) in the dewetting (Cassie−Baxter) state changing from (a) 0.20 to (b) 0.09 to (c) 0.03 to (d) 0.01. After the microstructures were fabricated, the surfaces were spincoated with Teflon solution for hydrophobic surface treatment. The Teflon solution was prepared by dissolving 0.2 wt % of amorphous fluoropolymer Teflon AF1600 powder (DuPont, Wilmington, DE, USA) in perfluoro compound FC-75 (Acros Organics). After agitation for 4 days and filtering with a Millipore filter (0.22 μm), the Teflon solution was spin-coated onto the surface and then baked at 112, 165, and 330 °C for 10, 5, and 15 min, respectively. The coating thickness of the Teflon film was controlled by varying the solution concentration and spinning speed. Spin-coating at 1000 rpm for 1 min was used in this work, which resulted in an ∼10 nm film thickness of the Teflon over the surface. The coating thickness was estimated by comparing the scanning electron microscope images of the micropillar structures before and after coating. A polished silicon substrate with a smooth Teflon coating was also prepared as a control, denoted as Φ1.00. The coating thickness on the nonpatterned planar surface was also measured by ellipsometry, resulting in the same thickness of ∼10 nm. For the study of evaporation kinetics, a sessile droplet (droplet volume ≈ 4 μL) of pure DI water was evaporated on each sample under atmospheric conditions (23 ± 1 °C, 32 ± 1% humidity, and atmospheric pressure). During the evaporation, the evolution of the
MATERIALS AND METHODS
For superhydrophobic surface models, micropillar-patterned superhydrophobic surfaces were studied in this work. Micropillar patterns of varying dimensions were fabricated from silicon substrates using a standard photolithography technique followed by a deep reactive-ion etching process at a cryogenic temperature (−100 °C). The detailed fabrication process is summarized in the following section. A 4 in. polished silicon wafer was first cleaned successively with acetone, B
dx.doi.org/10.1021/la400452e | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
Figure 3. Evolution of the contact angle (blue circles) and contact radius (red triangles) of a sessile droplet of water on (a) a planar hydrophobic surface and (b−e) micropillar-patterned superhydrophobic surfaces during evaporation. The distinct evaporation modes (CCR, constant contact radius mode; CCA, constant contact angle mode; Mix, mixed mode; Wetting, wetting state) are divided by vertical dotted lines.
to a wetting (Wenzel) state (CCA mode → wetting mode) takes place near the end of the evaporation process, as indicated by a small but sudden increase in the contact radius accompanying the simultaneous decrease in the contact angle. On the superhydrophobic surface of Φ0.01 with the lowest solid fraction (Figure 3e), no clear CCR mode with pinning appears. Instead, the droplet begins to recede, resembling a CCA mode, despite a slight decrease in the contact angle. A wetting transition also takes place at the end of the evaporation with a more dramatic increase in the contact radius and a decrease in the contact angle. Figure 3 also shows that the durations of the CCA and CCR modes significantly depend on the surface morphology (i.e., solid fraction (Φ)) of the hydrophobic surfaces. Figure 4 compares the duration of each evaporation mode on each surface type in terms of a percentile for the whole lifetime of the evaporation. It shows that after a short period for a CCR
contact angle, contact radius, and droplet volume was measured every 20 s with a goniometer system (model 500, Ramé-Hart Instrument Co., Netcong, NJ, USA) for data analysis until the droplet volume was smaller than the instrumental resolution (
