According to the illustration shown in Fig. 1, the cheap and accessible candle soot were utilized as a rough surface template to directly deposit onto the substrate. The candle soot was composed of carbon nanoparticles forming by incompletely burned paraffin, which loosely accumulated on the substrate and formed a rough surface. Following, a hard and porous silica shell was deposited on the candle soot template through chemical vapor deposition of SiCl4 like Stöber reaction, which completely replicate the roughness of the template and formed a robust rough surface. Because the thickness of the silica shell was much smaller than the wavelength of visible light (420 ~ 700 nm), it showed good transparency. After removal of the carbon nanoparticles template by calcination in air, the hard and transparent silica shell was reserved on substrate. Due to the abundant hydroxyl on this silica shell, it can be easily modified into superhydrophobicity through grafting trichloro (1H, 1H, 2H, 2H-perfluorooctyl) silane. As a result, a transparent silica coating with robust superhydrophobicity was obtained by rational surface design.
Surface modification and characterization. During preparation of this transparent and robust superhydrophobic coating, the surface wettability has changed significantly. Taking the quartz wafer sample as representative, its surface modifications have been successfully demonstrated (Fig. 2). As shown in Fig. 2a, the bare quartz wafer was washed carefully with dry ethanol, ethanol-deionized water (v/v 1:1) and deionized water in turn, and then characterized by SEM, CA and EDS. The SEM images indicated that the surface morphology of bare quartz wafer was homogeneous and smooth, and the EDS analysis only detected the signals of Si and O, the main elements that constituted quartz wafer, indicating a thoroughly clean surface. Meanwhile, the bare quartz wafer showed a hydrophilic surface with the water droplet contact angle of 58.4°, which was mainly due to the presence of hydroxyl on its surface.
Following, to prepare the candle soot template, the abluent quartz wafer was held above the flame of a candle for 3–5 s until a black soot layer with a few micrometers thickness was deposited. As can be seen in Fig. 2b, the SEM and EDS analysis indicated that the candle soot had formed a loose and porous layer, which was consisted entirely of carbon nanoparticles with the diameter of 30 to 40 nm. Meanwhile, a rough surface formed by accumulating candle soot nanoparticles can be observed on the quartz wafer. As is well recognized, the surface chemical composition and morphology are considered to be the key factors for obtaining superhydrophobicity on solid surface. 28 The surface chemical composition with low surface energy ensures its hydrophobicity, and the rough surface morphology with micro-nano hierarchical structure amplifies hydrophobicity into superhydrophobicity furtherly. 29 For instance, the surface of an abluent plastic sheet is only naturally hydrophobic, but not superhydrophobic (CA above 150°). However, it is interesting to note that a rough surface accumulated by hydrophobic nanoparticles can be amplified into superhydrophobicity. Therefore, the rough surface formed by accumulating candle soot nanoparticles showed a significant superhydrophobicity with water contact angle of 154.4°. Due to the weak physical interaction between the soot nanoparticles, this soot deposition layer was fragile. As water droplets rolled down from it, the soot nanoparticles were easily taken off by the droplets until nearly all of the deposits were removed, causing a wettability transition of the surface. Inspired by the promising morphology and super-wettability of candle soot deposits, a novel method to replicate and enhance the roughness structure was adopted 30. In this method, a silica shell was prepared to cover the candle soot layer by chemical vapor deposition of SiCl4, as shown in Fig. 1. As can be seen from Fig. 2c obviously, the average particle diameter of the candle soot increased slightly after the silica deposition, but its original morphology has been completely replicated. This has demonstrated that the roughness structure of the candle soot layer was not destroyed by chemical vapor deposition of SiCl4. Thus, the silica shell obtained superhydrophilicity (CA about 0°) due to the hydrophilicity of the silica and the roughness structure of candle soot layer. The SEM image and EDS analysis of the silica shell after calcination was shown in Fig. 2d. As can be seen from this figure, the candle soot template composed by carbon nanoparticles has been removed during calcination in air, but its rough and porous surface morphology was completely reserved by the hard silica shell. After surface modification by grafting with trichloro (1H, 1H, 2H, 2H-perfluorooctyl) silane, the silica shell surface showed a static contact angle of 153.0°. Because of the extremely low adhesion interaction between water and this superhydrohobic surface, the water droplets would be difficult to deposit and immediately roll off. In some special application scenarios, such as goggles or touch screens, the superhydrophobic coating would require transparency, heat stability and mechanical robustness. Since the thickness of the hollow silica shell was smaller than the wavelength of visible light, this shell was highly transparent. By calcination, the candle soot carbon nanoparticles disappeared and cavities were formed. The visible light transmittance of the quartz wafers before and after modification with transparent superhydrophobic coating were measured and shown in Fig. 2e top, the transparent superhydrophobic coating modified quartz wafer showed almost the same transparence with the bare quartz wafer. Moreover, this transparent superhydrophobic coating also showed excellent thermal stability, which remained superhydrophobicity even after 1 h calcination under high temperature up to 300ºC (Fig. 2e bottom).
Robustness. When exposed to the natural environment, the superhydrophobic coating should be endowed mechanical robustness against the harsh application scenarios. Therefore, sand abrasion test was performed on this superhydrophobic coating to characterize its mechanical durability. Sand particles with an average diameter of 300 to 500 um were used to impact the superhydrophobic coating directly from a height of 40 cm (Fig. 3a), and the corresponding impact energy was 1.1 × 10− 4 J/particle. As shown in Fig. 3b and c, the superhydrophobic silica shell were strong enough to fully withstand the impact of sand particles in 410 s. As the impact time increases, holes were eventually formed in the impacted area, and the silica shells were partially destroyed (Fig. 3d). In spite of this, the SEM image of the partially destroyed superhydrophobic coating indicated that the microscopic morphology of the cavities had hardly changed (Fig. 3d). Due to the self-similarity of the silica shell, this coating remained superhydrophobicity (CA = 151.8°) until the silica shell was completely removed from the substrate after continuous impact. The mechanical durability of the superhydrophobic coating mainly depended on the sand grains size, the sand impinging amount per time unit, the thickness of the silica shell, and the height of fall. And it would increase with the thickness of the silica shell, but at the expense of the coating’s transparency.
Oil-water separation. This superhydrophobic coating can be also used for oil-water separation due to its durable superhydrophobicity under oil. Thus, we utilized the strategy mentioned above to prepare an oil-water separation device with the stainless-steel mesh as substrate. Dichloromethane-water mixture was used as the representative to study the oil-water separation efficiency and oil flow rate, the results were shown in Figs. 4 and 5.
Both sides of the stainless-steel mesh were coated by the supehydrophobic silica shell as the oil-water separation mesh, and then the as-prepared mesh was imbedded into a split type filter. The filter was finally placed on a flask as shown in Fig. 4a. Following, the same volume (5 mL) of blue-dyed oil and red-dyed water was uniformly mixed and poured into the filter rapidly. Driven by gravity only, blue-dyed oil penetrated through the mesh immediately and red-dyed water was prevented (Fig. 4b and c). The oil-water mixture was completely separated; this was because the durable superhydrophobicity of the silica shell coated mesh even immersed under oil. The hydrophobic-oleophilic shell held a stable oil layer on its roughness surface, which constructed a 3-phase interface with water and prevented water from penetrating through the separation mesh (Fig. 4d). In addition, after oil-water separation course, the superhydrophobicity of the modified mesh was tested again. Water droplets did not wet or even contaminate the surface, demonstrating the reusability of the superhydrophilic surface.
According to the Cassie and Baxter theory 31, the three-phase interface is necessary for forming superhydrophobic surface. During the oil-water separation, the water-oil-solid system also constructed a three-phase interface and repelled water off, which only allowed the oil through. Ultimately, the oil-water mixture was successfully separated only dependent on gravity by utilizing the feature of this superhydrophobic coating, indicating the low energy consumption and easy operation of this strategy.
Due to the frequent offshore oil spills, it is very necessary to spread the oil-water separation application to different oil-water mixtures. We chose chloroform, toluene, gasoline and dichloromethane as the oil phase to blend with water. The separation efficiency was investigated to quantitatively evaluate separation property of our device, which was defined as the water content in oil after separation and measured by Karl Fischer moisture titrator. There were large differences of the separation efficiencies between different oil-water mixtures, this was mainly because water showed different solubility in different oil. The water contents in oil after separation were tested and shown in Fig. 5a, the separated gasoline, toluene, chloroform and dichloromethane presented water contents of 133 ppm, 262 ppm, 148 ppm and 77 ppm, respectively. Additionally, we also calculated the flow rate of different oil during separation through measuring its passing time, the results have been shown in Fig. 5b. The toluene-water and gasoline-water mixtures showed the peak flow rates of 0.991 and 1.4 m3·m− 2·h− 1, which indicated the oil kinds influenced the penetrating rate significantly. For chloroform, its mixture needed longer time for stratification after mixing thoroughly. Therefore, it showed a relatively low flow rate of 0.586 m3·m− 2·h− 1. After that, the gasoline-water mixture was used as a representative and continuously separated for 6 times to investigate the cycle stability of this separation device, and the separation efficiencies and flow rates were studied (Fig. 5b). For each separation cycle, the water content of the separated gasoline was below 300 ppm. In addition, the flow rate stabilized at about 1.0 m3·m− 2·h− 1, which verified the availability of this separation device. To sum up, the results indicated that this superhydrophobic coating has exhibited excellent comprehensive performance, not only in robustness and transparency, but also in oil-water separation applications.