Article pubs.acs.org/Langmuir
Mechanism of Delayed Frost Growth on Superhydrophobic Surfaces with Jumping Condensates: More Than Interdrop Freezing Quanyong Hao,† Yichuan Pang,† Ying Zhao,† Jing Zhang,† Jie Feng,*,† and Shuhuai Yao*,‡ †
College of Materials Science & Engineering, Zhejiang University of Technology, Hangzhou 310014, China Department of Mechanical Engineering, The Hong Kong University of Science and Technology, Hong Kong 999077, China
‡
S Supporting Information *
ABSTRACT: Delayed frost growth on superhydrophobic surfaces (SHSs) with jumping condensates has been found by many researchers recently. However, the mechanism of this phenomenon has not been elucidated clearly. In this study, copper SHSs with or without jumping condensates were selected as the substrates for observing condensation icing at a relative humidity (RH) of 60%. The results showed that only SHS with jumping condensates showed delayed condensation icing. Moreover, when such SHSs were placed upward and the surface temperature was held at −10 °C, some discrete frozen drops first appeared on the SHSs. The following icing mainly occurred on these discrete global crystals and then expanded around them until covering the entire surface. Little macroscopic interdrop freezing phenomenon was found. The growth of the frost front is mainly dominated by jumping freezing (the condensed droplets jumped onto the ice crystals and were frozen) or direct vapor−ice deposition. Using microscopy, we found interdrop freezing occurred, in addition to the two mechanisms mentioned above. By placing the SHS downward at −10 °C and intentionally introducing or eliminating tiny dusts, we confirmed that there were no superhydrophobic defects on our SHSs. The discrete frozen drops first appearing on the SHSs were triggered by tiny dusts falling on the surface before or during condensation icing. The key approach in delaying or resisting frost growth on SHSs with jumping condensates is to retard initial ice crystal formation, e.g., eliminating the edge effect and keeping the SHSs clean. Varanasi et al.,15 Wang et al.,17 and Wang et al.19 all demonstrated that frost nucleation (caused by in situcondensed water droplets) occurred on superhydrophobic microtextures without any particular spatial preference, thus resulting in increased ice adhesion over smooth surfaces, Song et al.21 confirmed that the superhydrophobic isotactic polypropylene (i-PP) surface can retard frost formation (delaying ∼300 s) compared with the hydrophobic PP surface. In another work,22 they showed that the jumping condensates delayed the ice formation for 1 h at −15 °C even at a supersaturation of 6.97. In fact, either with the sessile subcooled water or with the in situ-condensed water, the anti-icing ability of SHSs depends not only on their superhydrophobicity but also on the detailed surface morphology. If the structure gaps are too wide, e.g., scale of tens or several tens of micrometers, or the impacting strength is too large,23−26 the sessile subcooled water droplets may penetrate into the microgaps and thus be frozen instantly.27 On the other hand, if the structure gaps are narrow enough, e.g., scale of submicro- or nanometers, the water condensed within these nanogaps could automatically ascend to the top of the structures and form Cassie state droplets.28−30 As
1. INTRODUCTION Anti-icing, retarding icing, and reducing ice adhesion on solid surfaces are persistent problems that need to be resolved because of their extreme importance in human production activities and living condition improvements.1−4 Different types of deicing technologies, including active2,3 and passive5−7 ones, have been developed over the past 60 yeas. However, an effective, economical, and innovative technique still remains to be found. Recently, superhydrophobic surfaces (SHSs) have attracted a great deal of attention and had been expected to be novel candidates for overcoming surface icing problems.8−19 However, while many researchers have confirmed that SHSs can effectively prevent ice formation or reduce ice adhesion,8−13 many other researchers14−19 concluded that SHSs are not always ice-repellent. Through detailed analysis of these different conclusions, we find that studies supporting SHSs possessing anti-icing ability mostly applied supercooled water (simulating “freezing rain”), while studies opposing such a conclusion mostly used in situ-condensed water (simulating “frosting”). With the impact of the supercooled water, Gao et al.20 found that ice did not form on the superhydrophobic coats (−20 °C) prepared using 20 and 50 nm particles. However, when the particle diameter was larger than 50 nm, the icing probability increased remarkably, though the coats were still superhydrophobic. On the other hand, although Kulinich et al.,14 © XXXX American Chemical Society
Received: October 21, 2014 Revised: December 2, 2014
A
dx.doi.org/10.1021/la504166x | Langmuir XXXX, XXX, XXX−XXX
Langmuir
Article
2.3. Observation of Condensation and the Frosting Process. Condensation experiments were performed in a closed room with an area of 25 m2 and a height of 3 m. The ambient temperature was controlled at 20 °C, and the RH was adjusted at 60%. The copper foils (3 cm × 3 cm except where stated) with double SHSs were horizontally placed on the Peltier cooling stage with a surface temperature of 0−1 °C. The spontaneous motion of condensate droplets was observed and visualized by a top-down optical microscope (Nikon LV150) with a 10× objective and a chargecoupled device (CCD) camera at 25 fps. The spontaneous motion phenomenon was quantified by calculating changes in the location of visible droplets within the field of view in the 1 s video (here named “spontaneous motion frequency”).28,29 Condensation icing was performed just as described for the condensation experiments (e.g., in a larger room at 20 °C and 60% RH) but controlling the Peltier stage temperature at −5, −10, and −30 °C. The condensation and succedent frosting process were observed by eye and imaged by a digital camera every 5 min. The microscopic icing procedure was followed with an optical microscope (Nikon LV150) with a 10× objective and a charge-coupled device (CCD) camera at 25 fps. In addition to being deposited upward, the SHSs were also thermally bonded onto the Peltier stage using a thermal paste, and then the stage was reversed, e.g., moving the SHSs downward. Attention was paid on the frosting delay time and the manner of frost covering the entire SHSs with jumping condensation, especially on the effect of spontaneous motion of condensate droplets on the delayed frosting mechanism.
a result, delayed icing may occur because of little heat transfer and coalescence-induced jumping compared with the Wenzel condensates on the SHS with larger gaps. How do the subcooled condensate drops begin to freeze? How does the icing spread over the whole surface? Recently, Collier et al.31 reported that subcooled condensates on a chilled SHS were able to continuously jump off the surface before heterogeneous ice nucleation occurred. However, frost still formed on the SHS because of ice nucleation at neighboring edge defects, which eventually spread over the entire surface via an interdrop frost wave. The growth of the frost front was shown to be up to 3 times slower on the SHS than on a control hydrophobic surface, because of the jumping-drop effect dynamically minimizing the average drop size and surface coverage of the condensate. They attributed the first icing nucleation to the edge defects and demonstrated the icing spread arising from the interdrop freezing; e.g., after a frozen condensate drop contacted the neighboring liquid drops, a chain reaction was activated, which resulted in a propagating frost front. In this paper, we demonstrate that when the surface temperature was not too low (e.g., −5 °C), the anti-icing phenomenon was observed on the SHSs with jumping condensates. However, when the surface temperature was much lower (e.g., below −30 °C), a thin layer of ice was soon observed on the SHSs. When the surface temperature is moderate (−10 °C), the delayed condensation icing phenomenon was observed, and it seemed that icing occurred not only by interdrop freezing but also by direct vapor−ice deposition and jumping freezing. Moreover, when the SHSs were cleaned thoroughly before condensation icing or placed downward, the discrete frozen crystals appeared more slowly and sparsely. This demonstrated that the discrete ice crystals that first appeared on the SHSs were triggered by tiny dusts but not by defects of SHS itself. Therefore, for the SHS with jumping condensates, always keeping the surface clean is important for delaying or restraining initial ice crystal formation because it is the initial ice crystal that finally frosts the whole surface.
3. RESULTS AND DISCUSSION 3.1. Structures and Wettability of As-Prepared Surfaces. In our earlier studies, we confirmed that SHSs composed of nanostructures with a sufficiently narrow space, a higher perpendicularity, and a lower surface energy can result in Cassie condensates or obvious spontaneous motion phenomena.28,29 Figure 1 shows the top view morphologies of surfaces
2. EXPERIMENTAL SECTION 2.1. Preparation of SHSs with Different Condensation Modes. The copper SHSs were created by the procedures described in our earlier studies.28,29 Briefly, 10 cm × 10 cm × 0.5 mm clean copper foils (purity of 99.99%, Aldrich) were first immersed in an aqueous 4 M HCl solution for 5 s to remove surface oxide and then rinsed with a large volume of deionized water. Then, the copper foils were incubated in an aqueous solution of 2.5 M NaOH and 0.1 M (NH4)2S2O8 at 4−5 °C or room temperature (∼20 °C) for 1 h. Then they were thoroughly washed with deionized water and dried at 180 °C for 2 h to convert Cu(OH)2 into stable CuO by completing the dehydration reaction. Afterward, the black copper foils were incubated in a 0.5 wt % hexane solution of 1H,1H,2H,2H-perfluorodecyltriethoxysilane (FAS17, Sigma) at room temperature for 1 h, followed by drying at 120 °C for 1 h. 2.2. Characterization of Morphology and Wettability. The morphology of the resulting surfaces was characterized by field emission scanning electron microscopy (FE-SEM) (S4700, Hitachi). The water contact angles (CAs) and slide angles (SAs) of the asprepared copper surfaces were measured by using a Dataphysics OCA35 contact-angle system with a temperature/humidity control chamber (10 cm × 10 cm × 3 cm). The Peltier stage within this chamber can precisely maintain the temperature of SHS from −30 to 160 °C. The relative humidity (RH) of air within the chamber is controlled at 5−95% by adjusting the feeding rate of dry N2 and water vapor. The static CAs were measured six times and averaged.
Figure 1. SEM images of copper surfaces after their immersion in 2.5 M NaOH and 0.1 M (NH3)2S2O8 for 1 h at 4−5 °C (a and b) and room temperature (20 °C) (c and d), respectively. The insets show the profiles of sessile water droplets (2 μL) with CAs of (a) 159.0 ± 0.7° and (c) 158.2 ± 5.1°.
fabricated by wet chemical etching at various temperatures. One can see that when the surface was oxidized at 4−5 °C, only nanoribbons with good perpendicularity and diameters of ∼100 nm appeared on the surface (Figure 1a,b). However, when the oxidation was performed at room temperature, not only nanohairs with diameters of
