Fabrication of a micro-nanostructured superhydrophobic aluminum surface with excellent corrosion resistance and anti-icing performance

Abstract

In this paper, a superhydrophobic aluminum (Al) surface with a hierarchical micro-nanostructure was successfully fabricated via anodization followed by surface modification using 1H,1H,2H,2H-perfluorodecyltriethoxysilane (FAS). The as-prepared superhydrophobic surface (SHS) displays a water contact angle (CA) as high as 156.0 ± 0.7° and a sliding angle (SA) as low as 2.5 ± 1.4°. The surface is found to have good mechanical stability against sand erosion and chemical durability in a series of solutions. The SHS also demonstrates a decrease of 3 orders of magnitude in the corrosion current density (J_corr) and significant positive shift of 0.93 V in the corrosion potential (E_corr). In addition, the SHS shows a delay in icing time and temperature, as well as low ice adhesion at 0.04 ± 0.02 MPa. The facile process and the achieved surface morphology provide an attractive way towards anti-wetting surfaces for corrosion protection and anti-icing applications for other structural materials.

---

Introduction

Aluminum (Al) is one of the most popular metals for modern industry. It has been extensively used in many fields, including marine, aerospace, transportation and so on.¹ Corrosion on the Al surface is an important issue under harsh environment conditions, particularly in marine environments when exposed to salty water or moisture. It is well known that Al develops a dense native oxide layer on the surface which can avoid further corrosion in dry and non-salty conditions. However, once it is in contact with wet or/and salty environments, the oxide layer can be penetrated and thus the corrosion reactions will be triggered.²

Ice accumulation can cause great economic loss and even disasters.^3,4 For example in the aerospace industry, ice accretion may cause deteriorating aerodynamics or even fatal accidents. The electrical power outages as a result of ice or wet snow accretion have occurred in many countries and led to huge economic loss.^5,6 Over the past several decades, two main strategies have been developed to deal with this issue, i.e., active and passive icing/deicing methods.^7,8 The active icing/deicing method can be considered as a traditional approach to remove ice, which includes thermally melting the ice, manual deicing, mechanical vibration and so forth.⁹ However, these methods require energy input and even introduce environmental contamination when chemicals are used. On the other hand, the passive method has an advantage because of its environmental friendliness, low cost and low energy consumption. Icephobic coatings belong to the latter strategy by which the water freezing time is delayed and icing temperature is reduced while ice adhesion is typically much lower when compared with the surface without the coating.¹⁰

In many practical applications, it is ideal to achieve multifunctional properties, including good corrosion resistance, icephobicity, and even self-cleaning. Though these properties may sound quite diversified, a fundamental requirement is water repellency of the surface, since close contact with water is a prior condition for corrosion to occur and for ice to form strong adhesion with the substrate. Making the surface superhydrophobic at a wide range of temperatures will, hypothetically, lead to anti-corrosion and anti-icing performance.

Inspired by the lotus effect where the water droplets can easily roll off and take away the contamination on leaves, a superhydrophobic surface (SHS) with a water contact angle (CA) higher than 150° and a sliding angle (SA) lower than 10° has aroused enormous interest of many researchers due to its unique properties, involving water repellency,^11–14 self-cleaning,^15–19 corrosion resistance,^20–22 anti-icing,^23–26 oil–water separation,^27–29 etc. SHS can be fabricated either by modifying a rough surface using low-surface-energy materials or roughening a hydrophobic surface.^30,31 However, for the same roughness value, different surface geometries can cause complete changes in wetting behavior. Therefore, investigation of the effect of all surface geometrical parameters, such as height, spacing and width on SHS behavior is necessary in this case.^32,33 In recent years, a great number of techniques have been reported to fabricate SHSs, such as sol–gel,^34,35 template,³⁶ chemical etching,^37,38 laser fabrication,³⁹ hydrothermal reaction,⁴⁰ electrospinning⁴¹ and anodization.^42–44 For example, Crick et al.⁴⁵ prepared a superhydrophobic photocatalytic surface through a single-step aerosol assisted chemical vapor deposition (AACVD) process. Bae et al.⁴⁶ obtained a superhydrophobic Al alloy surface by a wire electrical discharge machining (WEDM) approach. Saleema et al.⁴⁷ achieved a superhydrophobic Al alloy surface by chemical etching in a base medium containing fluoroalkyl-silane. However, most of the methods have potential weakness such as complicated procedure, time-consuming, poor scalability and high cost, which makes the preparation process not convenient with limited production scalability. In comparison, the anodization method is easy to control, relatively fast and reproducible. Moreover, it is also able to create various uniform surface morphologies over large surface area on metal substrates.⁴⁸ So far, although a great number of processes have been made to construct SHSs on Al substrate, only limited superhydrophobic Al surfaces with excellent corrosion resistance^17,49–51 and anti-icing property^52–54 have been achieved. Wang et al.⁴³ fabricated a superhydrophobic Al surface by combining anodization and low-temperature plasma treatment with trichlorooctadecyl-silane while Liu et al.⁵⁵ created SHSs on Al alloy via anodizing and polymeric coating. However, the corrosion resistance was not mentioned in the above research. Barkhudarov et al.⁵⁶ obtained superhydrophobic films as corrosion inhibitors on Al surfaces. Although the researchers did the work on the corrosion resistance, they mainly focused on the change in thickness reduction of superhydrophobic Al film and did not pay attention to the corrosion potential (E_corr) and corrosion current density (J_corr). Yin et al.⁵⁷ prepared a superhydrophobic coating on Al alloy surface with no significant improvement in corrosion resistant performance. He et al.⁵⁸ and Feng et al.² constructed SHSs on Al and its alloy substrates to improve their corrosion resistance, nevertheless, the aforementioned surfaces have not shown remarkable improvement as compared to the substrates. As a result, achieving excellent corrosion resistant superhydrophobic Al surface is still a great challenge. Besides, not all the prepared SHSs can display good or excellent enduring anti-icing property on Al substrate. Zhang et al.⁵⁹ prepared the superhydrophobic alumina films on Al foils, but they did not take into account the durability of anti-icing function. Yang et al.⁶⁰ and Song et al.⁶¹ produced superhydrophobic Al surfaces which can delay the ice formation effectively. However, the endurance test of icephobic property was not mentioned in the research. Zou et al.⁶² obtained a superhydrophobic Al surface with almost the same ice adhesion as that of the as-received Al. Bharathidasan et al.¹⁰ prepared a superhydrophobic Al surface with the ice adhesion higher than that of hydrophobic surfaces. Kulinich et al.⁶³ reported superhydrophobic Al surfaces with significant increasing ice adhesion during icing/deicing cycles. Farhadi et al.⁶⁴ produced superhydrophobic Al surfaces whose surface asperities were gradually damaged and even lost the superhydrophobicity after multiple icing/deicing cycles, indicating significant deterioration of the anti-icing performance. Therefore, it is an urgent demand to fabricate a superhydrophobic Al surface which is of great corrosion resistance and anti-icing property. Meanwhile, the surface should remain functional under harsh mechanical environment.

However, systematic work about all the properties together on the superhydrophobic Al surface has been rarely reported.

Herein, we reported a facile way to fabricate a hierarchically micro-nanostructured superhydrophobic Al surface through the anodization process together with surface modification. The synergy of combination hierarchical structure and low surface energy material is the key to SHS. The as-prepared SHS with a CA of 156.0 ± 0.7° and a SA of 2.5 ± 1.4° showed excellent corrosion resistance and anti-icing property. Meanwhile, the mechanical stability, chemical durability and weathering resistance as well as the self-cleaning effect were also examined.

Experimental

Sample preparation

The Al plates (purity: 99.9%, thickness: 0.05 mm, Art Friend & Buona Vista Pte Ltd, Singapore) were mechanically polished using 1000 grid sandpaper and cut into small pieces with the size of 2.5 cm × 2.5 cm. The small pieces of Al plates were ultrasonically cleaned with 5 wt% NaOH, 5 wt% HNO₃ and the deionized (DI) water in sequence to get rid of grease and other contaminants. Subsequently, the anodization was applied to fabrication of the samples after the cleaned Al plates were dried in air. The Al plate and the lead (Pb) plates were taken as the anode and cathodes with 2.5 cm distance during the double cathode anodization process. The cleaned Al plate was anodized in 15 wt% H₂SO₄ electrolyte under a constant voltage from 10 V to 22 V for 1 h. In order to avoid the excessive heat generated by the drastic anodized reaction, the electrolyte was magnetically stirred and cooled down with an ice-water bath to keep the reaction temperature at 25 °C. The anodized Al samples were next ultrasonically rinsed with DI water and dried. Finally, these samples were treated by ultrasonication for 30 min in 1.0 wt% methanol solution of FAS [1H,1H,2H,2H-perfluorodecyltriethoxysilane, (OC₂H₅)₃Si(CH₂)₂(CF₂)₇CF₃, 97%] and then heated at 100 °C for 1 h in the oven. All chemicals (Sigma-Aldrich, USA) were analytical grade and used as received without any further purification.

Sample characterization

The contact angle (CA) and sliding angle (SA) were measured by a contact angle measuring instrument (OCA 20, DataPhysics Corporation, Germany) at five different positions on each sample surface with 4 μL DI water droplet.

The surface morphology and elemental analysis were characterized by a field emission scanning electron microscope (FESEM, JEOL JSM-6340F, Japan) with attached energy dispersive X-ray spectrum (EDS). The roughness was analyzed by an atomic force microscope (AFM, Asylum Research Cypher S, Oxford Instruments Company, USA) with the scan area of 10 μm × 10 μm in the tapping mode.

The mechanical stability of the as-prepared samples was evaluated by a micro-sandblaster equipment (Comco Inc.-Lth Machinery Pte Ltd, Singapore). Sands (63 μm SiO₂ particles) were used to impinge the sample surface from a height of 15 cm under a pressure of 30 kPa for 30 s, 60 s and 90 s, respectively. The weight loss and the change in both CA and SA of the samples were measured after the sandblast test.

The weathering resistance test was conducted using a UV/condensation weathering instrument (ATLAS Material Testing Technology LLC, USA) by performing the standard of ASTM G154-12a. Eight UV lamps (295–400 nm) with power intensity of 0.77 W m⁻² for each one were used in the weathering chamber. The samples placed inside the chamber were exposed to UV for 4 h at 60 ± 3 °C, followed by water spray and condensation for 4 h at 50 ± 3 °C. The procedure was repeated over a period of 7 days.

The electrochemical corrosion test was performed by an electrochemical workstation (CHI 750C, Shanghai Chenhua Instrument Corporation) using a classical three-electrode system with a sample of 1.0 cm² exposed area as the working electrode (WE), a saturated calomel electrode (SCE) as the reference electrode (RE) and a platinum (Pt) sheet as the counter electrode (CE). Both the potentiodynamic polarization curves and the electrochemical impedance spectroscopy (EIS) measurements were tested in 3.5 wt% NaCl solution (pH = 6.0) at room temperature after the samples were immersed in the corrosive medium for 1 h to obtain a steady open circuit potential (OCPT). Prior to potentiodynamic polarization curves measurement, the EIS measurement were performed in the frequency range between 10⁻² Hz and 100 kHz with a sinewave amplitude of 5 mV. The experimental EIS spectra were interpreted on the basis of equivalent electrical analogs using the program Zview2.0 to obtain the corresponding fitting parameters.

The ice adhesion was tested using a self-designed adhesion device including an ice chamber, a sample stage and an air cylinder combined with a force gauge on the tip of the piston (see ESI, Fig. S1†). The column with an inner diameter of 18 mm was filled with DI water, covered by the samples, turned upside down and then placed on the sample stage in the ice chamber at −10 °C for at least 24 h to ensure complete ice formation. During the shear test, after the piston touched the ice column with a constant speed, its push force was increased at a rate of 12.5 N s⁻¹ supplied by a compressed gas tank located outside until the ice was fractured and removed from the sample surface. The corresponding push force (F) can be recorded and the contact area (A = πr² = π9² mm²) between ice and sample surface is considered to be a constant value, so the ice adhesion (τ) can be calculated by the following equation:

τ = F/A (1)

After each icing/deicing test, the sample was returned to room temperature and dried in air before the next cycle.

In order to evaluate the self-cleaning ability of the samples, an artificial dirt mixture composed of 85 wt% nano-clay, 10 wt% SiO₂ particles (1–5 μm), 2 wt% salts, 2 wt% oil, and 1 wt% carbon black was prepared by thorough mixing and then dispersing in water with a concentration of 15 g L⁻¹. After being immersed in the artificial dirty solution for 1 min, the contaminated samples were dried overnight and heated at 40 °C for 30 min before the water spray cleaning. The change of the lightness values which stand for the luminance on the sample surfaces was measured by a spectrophotometer (Elektro Physik, Germany). The lightness of the pristine and contaminated surfaces was recorded as L₁ and L₂, respectively. The L₂ was decreased after the sample was immersed in the dirty solution, so the percentage of dirt accumulation can be calculated by

(L₁ − L₂)/L₁ × 100 (2)

Results and discussion

CA and SA under different anodization voltages

To evaluate the hydrophobic properties of different samples, both the CA and SA were measured in terms of variation of anodization voltages, as illustrated in Fig. 1. Without anodization, the Al surface modified by FAS only has a CA of 120.9 ± 1.9°. After anodization, the CA of samples is increased while a decreasing trend is seen when the anodization voltage exceeds 20 V. It is worthy to note that the CA increases slowly when the voltage is increased from 10 V to 16 V. The CA increases to 143.7 ± 1.7° at the voltage of 18 V. The maximum CA of 156.0 ± 0.7° and the minimum SA of 2.5 ± 1.4° are obtained when the voltage reaches 20 V. However, the CA has small decrease (153.3 ± 0.9°) and the SA increases slightly (5.0 ± 1.0°) under 22 V.

Fig. 1 CA and SA of sample surfaces modified by FAS after different anodization voltages for 1 h.

It is evident that the increasing anodization voltage more than 20 V could not increase the superhydrophobicity and the optimal anodization voltage to prepare the SHS is 20 V.

Surface morphology and chemical composition of SHS

The surface morphology is shown in Fig. 2. As can be seen from Fig. 2(a), (b) and (e), the Al surface modified by FAS alone is relatively flat while the anodized Al surfaces are quite rough and compactly covered by a great number of micro-scale protuberances with different structures. From the magnified FESEM images, it is clear that the rough structure of the anodized Al surface at the voltage of 20 V is composed of numerous nanopores with average diameter about 40 nm [Fig. 2(c) and (d)]. When the anodization voltage further increases to 22 V, numerous intertwined nanowires are formed and cover the nanopores layer at the bottom on Al surface, as shown in Fig. 2(f). The observation evidently shows that two types of hierarchical micro-nanostructures, i.e., micro-nanopore and micro-nanowire types were constructed on the Al surface. However, the structure change from nanopores to nanowires can reduce the amount of nanoporosity, which may lead to less air trapped inside the surface layer. Consequently, the CA tends to decrease when the anodization voltage increases to 22 V.

Fig. 2 FESEM images of samples with FAS treatment: non-anodized (a); anodized Al surface under the voltage of 20 V (b–d) and 22 V (e–f) at different magnifications.

It is well known that both the surface structure and chemical composition are the key factors to affect the fabrication of SHS. Fig. 3 shows the EDS spectra of Al surface modified by FAS, anodic Al oxide (AAO) surface and SHS. It can be seen that the AAO surface is mainly composed of Al, O and S which comes from sulfate ions (SO₄²⁻) incorporated into the AAO films during anodization in sulfuric acid,^65,66 as shown in Fig. 3(b). In contrast, the modified Al surface and SHS show the existence of elements C, F and Si, as shown in Fig. 3(a) and (c), indicating that the FAS film is self-assembled on the two sample surfaces successfully.⁶⁷ The FAS consists of polar end-groups of –Si(OC₂H₅)₃ and a long hydrophobic chain of –(CH₂)₂(CF₂)₇CF₃. First, the Si–OC₂H₅ polar groups are hydrolyzed to form silanols (Si–OH) which act as highly reactive intermediates. Then the FAS, i.e., silane molecule could be strongly anchored onto the Al surface by the dehydration condensation reaction of Si–OH with the functional groups (Al–OH) on the Al surface to generate a self-assembled monolayer.⁶⁸ Meanwhile, the Si–OH groups can also form a grafted polysiloxane via a dehydration reaction with each other.⁶⁹ So the hydrophobic groups of –CF₂ and –CF₃ at the other end of the molecule can effectively decrease the free energy of the as-prepared sample surfaces.

Fig. 3 EDS spectra: (a) Al surface modified by FAS; (b) AAO surface; (c) SHS.

Combining Fig. 2(a) and 3(a), although the Al surface modified by FAS has a low-surface-energy layer, it is too flat to construct the SHS and thus can only exhibit limited degree of hydrophobicity. Fig. 4 shows the AFM images of both AAO surface and SHS. Their corresponding root-mean-square roughness (R_q) was 788.3 nm and 286.3 nm, respectively. As can be seen in Fig. 4(a), the AAO surface is rough enough, but it lacks the low-surface-energy materials, showing the superhydrophilicity. Comparing Fig. 4(a) and (b), the roughness of the SHS is decreased obviously, suggesting that FAS might have partially filled or covered the nanopores. From the surface roughness, it can be found that the surface with the biggest roughness does not show the highest CA, demonstrating that although the roughness is necessary for achieving a SHS, the CA of the surface is not simply related to the roughness value only; it is also related to the detailed micro/nanostructures as well as the surface chemistry. From the current study, it can be concluded that the synergistic effect of hierarchical micro-nanostructure and the chemical composition is the crucial point to construct the SHS.

Fig. 4 AFM images: (a) AAO surface; (b) SHS.

Wettability of SHS

The surface wettability can be described by the Wenzel⁷⁰ and Cassie–Baxter^71,72 theories which derive from Young–Dupre equation. In the Wenzel model, the CA for a hydrophilic surface which has an intrinsic CA lower than 90° decreases when its roughness increases. The capillary effect can occur and cause the water to suck into the grooves on the hydrophilic rough surface. The CA on the bare Al surface is 52.6 ± 1.4°, indicating the hydrophilic nature of Al. With the increasing anodization voltage, the Al surface becomes more and more rough. When the voltage reaches 20 V, it is rough enough so that the droplets fill in such structure and the CA drops rapidly to 0°, which exhibits perfect superhydrophilicity (see ESI, Video S1†). However, the CA increases to 156.0 ± 0.7° with the SA of 2.5 ± 1.4° after the AAO surface was modified by FAS. The formation of SHS can be elucidated by the Cassie–Baxter model. In this model, the water droplets could not penetrate into the micro-nanostructure and can easily roll off from the surface because of the enclosed air between the solid surface and the water droplet (see ESI, Video S2†).

Accordingly, the CA on the SHS can be expressed by Cassie–Baxter equation:⁷²

cos θ_γ = f₁ cos θ₁ + f₂ cos θ₂ (3)

where θ_γ is the apparent CA; f₁ and f₂ are the area fractions of solid surface and air in the composite surface; θ₁ and θ₂ are the corresponding intrinsic CAs. Given that f₁ + f₂ = 1 and θ₂ = 180°, eqn (3) can be modified as follows:

cos θ_γ = f₁(cos θ₁ + 1) − 1 (4)

According to the above CAs on the smooth bare Al surface modified by FAS and SHS, the f₁ can be calculated to be 0.1777, which indicates that 82.23% of the surface is occupied by the air/water interface when the water droplets contact SHS.

Mechanical stability

In order to thoroughly investigate the mechanical stability of the SHS, the micro-sandblasting test was carried out using a micro-sandblaster equipment and the abrasive gun is placed directly above the samples with the aid of washer (diameter = 1.76 cm) (see ESI, Fig. S2†).

There is no obvious change on the SHS surface as compared to the bare Al surface after the micro-sandblast test from a macroscopic perspective (see ESI, Fig. S3†). The changes in CA and SA as well as weight loss were measured to evaluate the erosion resistance of the SHS. With increasing sandblasting time, the CA decreases slightly while the SA shows an increasing trend, as shown in Fig. 5(a). Although the SA increases from 2.5 ± 1.4° to 22.4 ± 0.7°, 34.3 ± 2.8° and 35.8 ± 1.8° after sandblasting for 30 s, 60 s and 90 s, the CA can still maintain above 150°, showing good mechanical stability. Besides, the corresponding weight loss is increased from 0.08 mg cm⁻² to 0.12 mg cm⁻² and no longer changed after sandblasting for 60 s, as presented in Fig. 5(b).

Fig. 5 (a) The effect of sandblasting time on the CA and SA of SHS; (b) the change in weight loss of the SHS for different sandblasting time.

The surface morphology of SHS before and after micro-sand erosion test was compared to further study the change in surface structure. It seems that the surface became more and more smooth with the increasing micro-sandblasting time and finally the micro-nanostructure was mostly destroyed to generate a new surface structure, as presented in Fig. 6. It is reported that the mere micro- or nanostructure can lead to the superhydrophobic phenomenon, but it is difficult for water droplets on the surface to roll down. Only the surface with complex hierarchical micro/nanostructure can form both high CAs and low SAs.⁷³ Therefore, although the hierarchical structure was damaged after test, it was found that the nanostructure was still maintained [Fig. 6(b)], which was responsible for the continued superhydrophobicity while the SA was increased due to the loss of the microstructure. Despite of that, the phenomena suggest that the SHS has good mechanical damage resistance.

Fig. 6 FESEM images of the SHS after micro-sand erosion treatment for various time: (a) 30 s; (b) magnification of (a); (c) 60 s; (d) 90 s.

Chemical durability

For outdoor application, it is of great significance to evaluate the durability of SHS under different severe environments. The water droplets with different pH values were adjusted by HCl and NaOH and dripped onto the surface for 5 s before test. It is clear that the CA on the SHS is greater than 150° over a wide range of pH values from 1 to 13 while it has a slight decrease when the pH = 0 (149.6 ± 1.4°) and pH = 14 (147.9 ± 0.9°), as shown in Fig. 7. This indicates that the SHS has high hydrophobicity not just for DI water but also for corrosion liquids, such as acidic and basic solutions. That is to say, the as-prepared SHS can be highly durable against chemical corrosion and has potential in corrosive conditions.

Fig. 7 The relationship between the pH value and CA on the SHS.

Weathering resistance

Upon exposure to UV lights and condensation for 7 days, the CA on the hydrophobic Al surface, namely, modified Al surface after different anodization voltages from 10 V to 18 V, shows a significant decrease while it is worthy to note that the SHS is highly stable and remains superhydrophobicity with a CA as high as 154.9 ± 0.5°, as shown in Fig. 8.

Fig. 8 The change of CA on different samples before and after weathering resistance test.

The surface morphology of Al surface modified by FAS and SHS after weathering resistance test is shown in Fig. 9. From Fig. 9(a), it can be clearly seen that there are lots of cracks on the Al surface modified by FAS and the hydrophobic FAS layer is peeled off under continuous UV irradiation together with water spray, which means the water can quickly penetrate into the surface to corrode the Al substrate. On the contrary, we found that there is no significant difference on the SHS in Fig. 9(b), as compared with Fig. 2(b). This implies that the SHS has a substantial improvement on weathering resistance for Al substrate.

Fig. 9 FESEM images for the surface structure after weathering resistance test for 7 days: (a) Al surface modified by FAS; (b) SHS.

Furthermore, the EDS composition analysis (see ESI, Fig. S4†) shows that the elements F and Si are still preserved on SHS despite small loss of FAS molecules. The residual low-surface-energy layer contents combined with the maintained rough structure can still keep superhydrophobicity on the SHS, showing excellent weathering resistance.

Corrosion resistance

To investigate the protective performance of the SHS, the potentiodynamic polarization curves of the bare Al, AAO and SHS were measured in 3.5 wt% NaCl solution, as shown in Fig. 10. The corrosion current density (J_corr), corrosion potential (E_corr) and polarization resistance (R_p) obtained by Tafel extrapolation are shown in Table 1. A lower J_corr and a higher E_corr correspond to a better corrosion resistance.⁶⁹ It can be seen that J_corr of AAO (1.124 × 10⁻⁸ A cm⁻²) is reduced by 1 order of magnitude as compared to that of the bare Al (3.060 × 10⁻⁷ A cm⁻²). However, it is found that the corresponding E_corr shifts more negatively from −0.629 V to −1.053 V, suggesting its limited protection for the bare Al. The SHS, on the other hand, achieves the lowest J_corr and highest E_corr among the samples. The J_corr of SHS (8.493 × 10⁻¹¹ A cm⁻²) is reduced by more than 3 orders of magnitude and the E_corr undergoes a considerably positive shift of 0.93 V compared with bare Al, which indicates an effective anti-corrosion improvement. Besides, a higher R_p also indicates better corrosion resistant performance.⁷⁴ It can be found that the SHS has a much higher R_p, which is 7.47 × 10³ times that of the bare Al and 159 times that of AAO, confirming the excellent potential to protect Al against corrosion.

Fig. 10 Potentiodynamic polarization curves measured in 3.5 wt% NaCl solution for the bare Al surface, AAO surface and SHS.

Table 1 Electrochemical parameters of potentiodynamic polarization curves for the samples

Samples	Ecorr (V vs. SCE)	Jcorr (A cm^−2)	Rp (Ω cm^2)
Al	−0.629	3.060 × 10^−7	8.451 × 10^4
AAO	−1.053	1.124 × 10^−8	3.982 × 10^6
SHS	0.301	8.493 × 10^−11	6.315 × 10^8

---

The corrosion resistance of the as-prepared SHS was further explored by the electrochemical impedance spectroscopy (EIS). As represented in Fig. 11, the Nyquist plots for bare Al surface and SHS are quite different, so the corresponding equivalent circuits applied in such two surfaces are also different. The fitting results of relevant parameters can be obtained accordingly. As shown in Fig. 12(a), the equivalent circuit represents the electrochemical behavior of the bare Al surface and shows one time constant. In this circuit, R_s is the solution resistance, R_ct means the charge transfer resistance, and C_dl is the electrical double layer capacitance. In the case of superhydrophobic film, the equivalent circuit represents the electrochemical behavior of SHS and should consider two time constants in the corresponding impedance spectrum, as shown in Fig. 12(b). R_c can be ascribed to the solution resistance within the pores of the porous sub-layer, which mainly depends on the pore dimensions. C_c is assigned to the capacitance of a surface film, which depends on different factors such as film thickness, film defect and film structure, etc. R_ct||C_dl shows the impedance associated with the interface reactions and the solid state conduction in the barrier layer. The values of the R_ct for different samples also reveal their difference in anticorrosion performance.⁷⁵ The fitting results prove that the SHS has a much higher R_ct of 3.845 × 10⁸ Ω cm² compared with the untreated Al substrate (4.342 × 10⁴ Ω cm²). It is obviously that the R_ct values increased rapidly when the hydrophilic Al surface became superhydrophobic. The result agrees well with that of potentiodynamic polarization curves measurement and indicates that the superhydrophobicity plays an important role in improving the corrosion resistance of the bare Al substrate.

Fig. 11 Nyquist plots of the bare Al surface and SHS in 3.5 wt% NaCl solution.

Fig. 12 Equivalent circuits of different studied system (a) bare Al surface; (b) SHS.

It can be seen from the Bode plots [Fig. 13] that the SHS possesses the highest impedance as compared to the bare Al substrate and AAO in all frequencies. In particular at low frequency end, the impedance value of SHS can reach up to 3.99 × 10⁸ Ω cm², which is 1 to 4 orders of magnitude higher than that of AAO (2.99 × 10⁷ Ω cm²) and bare Al (4.34 × 10⁴ Ω cm²). The result further implies that the SHS has more advantage than the AAO surface in protecting the underlying Al substrate.

Fig. 13 Bode plots of the bare Al surface, AAO surface and SHS in 3.5 wt% NaCl solution.

From the obtained results, it can be concluded that the AAO is easy to penetrate by the chloride ions (Cl⁻) in the 3.5 wt% NaCl solution so as to fail in improving the corrosion resistance. However, once the AAO was modified by FAS, it shows excellent anti-corrosion performance, further indicating the synergy of hierarchical micro-nanostructure and the modification of the hydrophobic materials. The corrosion resistance mechanism for the SHS can be ascribed to the followings: the air trapped in the micro-/nanopores on the surface acts as a passivation layer so that the corrosive ions (Cl⁻) cannot penetrate into or arrive at the bare Al substrate easily.⁷⁶ As a result, the corrosion rate of the SHS slows down greatly. Another important factor why the SHS can improve the anti-corrosion of Al substrate is the “capillarity” effect. As discussed in the previous literature, the water transport against gravity is easy on porous surface with a CA higher than 150°. Therefore, the NaCl solution can be pushed out from the pores by the Laplace pressure and the Al substrate could be perfectly protected from corrosion.⁷⁷

To further illustrate the effectiveness of the treated SHS in the current work, we made a comparison with the previous reported anti-corrosion superhydrophobic Al and its alloy surfaces. Feng et al.⁷⁸ produced a superhydrophobic Al alloy surface with a J_corr of 5.01 × 10⁻⁵ A cm⁻² whereas that of the bare Al alloy is 7.26 × 10⁻⁴ A cm⁻², showing that the J_corr is reduced by 1 order of magnitude. Saleema et al.⁴⁷ fabricated a hydrophobic surface and a superhydrophobic surface on Al alloy, however, the J_corr and R_p shows no significant difference in their corrosion performance. Yin et al.⁷⁹ and He et al.⁵⁸ prepared a superhydrophobic surface on Al as the corrosion protective layer, but it was found that the E_corr undergoes a positive shift for only 0.2 V and the J_corr was about 10⁻⁷ A cm⁻², indicating limited corrosion resistance improvement. The above results indicate that our SHS has a much lower J_corr and a much higher E_corr, i.e., much better anti-corrosion improvement.

Anti-icing behavior of the SHS

The icing process of water droplets on sample surfaces was studied using a cooling apparatus of the contact angle measurement system, including a small chamber with a sample stage fixed on the bottom, recirculating cooling water and a K-type thermocouple. The temperature was measured using the thermocouple and the recirculating cooling water was used to provide cooling for the sample stage. To maximize the isolation of heat from atmosphere, the inner walls of the chamber were covered with adiabatic foam except the two small windows on the side walls for observation by a CCD camera. Besides, the thermally conductive silicone grease was used to fix the substrates on the sample stage to ensure better thermal conduction. The temperature was decreased at 5 °C min⁻¹ from room temperature to 0 °C and 1 °C min⁻¹ from 0 °C till the water starts to freeze during the icing process.

It can be seen that the freezing processes of the water droplets on bare Al surface and SHS are different, as shown in Fig. 14. The water droplets on the two samples start to freeze at −12.3 °C and −24.0 °C respectively, and their corresponding ice growth processes last 2 s and 23 s. It is obvious that the droplet on the SHS has a much lower icing temperature and takes a longer time to completely freeze even under quite a lower temperature. The droplet on the sample surface can be considered as a system of solid–liquid–air three-phase interface. It gains heat from air and loses heat to the cold surface in the form of contact heat conduction and thermal radiation. The net heat loss per unit time can be expressed as:

Δh = h_l + h_l′ − h_g − h_g′ (5)

where h_l and h_l′ are the heat loss through contact heat conduction and thermal radiation to the sample surface per unit time; h_g and h_g′ are the heat gains through contact heat conduction and thermal radiation per unit time from the air. According to thermodynamics, the time for the water droplet to freeze completely on the surface can be expressed as:^80,81where ρ_w and C_P are the water density and the heat capacity of pure water at constant pressure, respectively; T_o is the initial temperature of the droplet and T_s is the sample surface temperature. eqn (5) and (6) indicate that a smaller Δh can cause a larger Δt. Due to a higher percentage of trapped air in the SHS, the liquid–solid contact area was reduced and thus less heat would be lost from the water droplet to the surface. This can explain why the water droplet can be retarded to turn into ice on the SHS and it needs a much lower temperature to start to freeze comparing with the bare Al surface.

Fig. 14 Optical images of the icing processes with 4 μL droplet on bare Al surface and SHS.

To systematically evaluate the anti-icing property of the samples, the ice adhesion is tested under −10 °C after they are kept at the same temperature for 24 h (see ESI, Fig. S5†). As shown in Fig. 15, the ice adhesion significantly decreased from 1.14 ± 0.11 MPa, 1.02 ± 0.28 MPa, 0.20 ± 0.10 MPa to 0.04 ± 0.02 MPa when a surface changes from AAO (superhydrophilic), bare Al (hydrophilic), FAS modified Al (hydrophobic) to SHS (superhydrophobic). It's worth noting that although the superhydrophilic surface and SHS have a similar hierarchical micro-nanostructure, the ice adhesion strength on the AAO surface is 29 times higher than that on the SHS because of the modification of FAS, which means both the hierarchical structure and hydrophobic FAS make great contribution to water repellence even at subzero temperature. The different values can be ascribed to the surface wettability. Owing to the corresponding superhydrophilicity and hydrophilicity, the AAO and bare Al surfaces can be fully and partially wetted and thus when the water freezes, the stronger bonding forms between the ice and the two sample surfaces through the mechanical interlocking effect,⁶³ leading to the higher ice adhesion. For the Al surface modified by FAS, although the ice adhesion is decreased drastically due to its lower-surface-energy layer, it is not low enough. In contrast, the SHS has the lowest ice adhesion among all the samples due to the trapped air inside the SHS which can be used as “air cushion” to maintain water repellence at low temperatures and contribute to the anti-icing property. First, it can prevent the water from contact with the concave parts of the surface, and thus the water can only stay on the top of convex parts. According to the classical nucleation theory,^82–84 heterogeneous nucleation is more difficult on a convex surface than on a smooth or concave surface. As a result, the ice forms more difficultly on the SHS. Besides, it can reduce the real contact area between ice and sample surfaces. A large percentage of trapped air makes a larger liquid-air contact, which forms a smaller liquid–solid contact and hence greatly minimizes the mechanical anchoring effect. Additionally, it is in favor of forming loose ice structure on the interface when water freezes. The volume of water expands upon freezing, so the entrapped air would be compressed, resulting in a higher pressure than the atmosphere, which can cause a counter-force on the ice column so as to reduce the force needed to remove the ice.

Fig. 15 Ice adhesion of different samples at −10 °C. The insets are the residual ice after ice adhesion test on the corresponding samples.

The insets in Fig. 15 clearly display that there is a large fraction of residual ice adhered to the Al and AAO surfaces after the ice adhesion test. Conversely, a relative small amount is left on the Al surface modified by FAS and little can be seen on the SHS. This phenomenon can further prove that the SHS exhibits evident anti-icing property and can effectively reduce the ice accumulation.

The effect of the number of icing/deicing cycles on CA and SA of SHS is also an important consideration for practical applications as it is related to how much of the original superhydrophobicity is retained after ice removal. As indicated in Fig. 16, it can be seen that the CA decreased slightly from 156.0 ± 0.7° to 151.9 ± 0.8°, 151.5 ± 0.7° and 152.4 ± 0.5°, and the SA increased from 2.5 ± 1.4° to 5.0 ± 3.3°, 5.0 ± 1.2° and 4.8 ± 2.3° after 3, 9 and 11 cycles. This could be caused by the partial damage of the asperities on the surface during shearing of the ice block⁸⁵ (see ESI, Fig. S6†). As a consequence, the damaged asperities lead to a small decrease in CA and an increase in SA on the SHS. Although the CA of the SHS after 11 cycles was a little higher than that after 9 cycles, it is considered to be reasonable and acceptable in view of the measurement error in the experiment. Generally, the change trend of CA coincides with the increasing of icing/deicing cycles. Meanwhile, it is clear that the CA remains above 150° and the SA keeps below 10° after 11 icing/deicing cycles, which suggests that the micro-nano hierarchical structure still mostly survived. That is to say, the icing/deicing test has no significant influence on the CA and SA of SHS. Therefore, the as-prepared SHS after repeated test still has a low adhesion with water droplet and possesses superior superhydrophobicity (see ESI, Video S3†).

Fig. 16 The change of CA and SA before and after multiple icing/deicing cycles.

In comparison, the SHS is more effective in anti-icing than the previously reported superhydrophobic Al and its alloy surfaces. Wang et al.⁸⁶ produced an icephobic superhydrophobic Al plate surface with the ice adhesion increasing nearly about 50% from 0.22 MPa to 0.34 MPa after 8 icing/deicing cycles. The corresponding CA was decreased from 164.4° to 150.6° and SA was increased from 0.84° to 51.4° after 20 cycles. The values might demonstrate that the repeating icing/deicing cycles created damage to the superhydrophobic surface, and thus the surface lost part of their hydrophobicity. Farhadi et al.⁶⁴ created an anti-icing superhydrophobic surface on Al alloy with gradually increased ice adhesion of 0.11 MPa after multiple ice-breaking repeat, which is nearly twice the initial value (0.055 MPa). Accordingly, the CA measured is below 150°, which indicates that the surface loses its superhydrophobicity.

The observed results show that the SHS herein has a robust anti-icing property, which is much better than that of Al surface modified by FAS, and far better than that of bare Al and AAO surfaces.

Self-cleaning effect

Fig. 17 shows the dirt accumulation on different samples. It is noted that the SHS presents the lowest extent of dirt accumulation only about 0.18% compared to that of bare Al surface and AAO surface, indicating its outstanding self-cleaning effect. The detailed calculation is shown in Table S1.†

Fig. 17 Dirt accumulation on various samples after immersion in the dirty solution for 1 min followed by drying.

Fig. 18 shows the images of the samples before and after immersion in the artificial dirty solution as well as water spray cleaning. The images correlate very well with the percentage of dirt accumulation. As can be seen from Fig. 18(b) and (e), plenty of dirt adhered to the bare Al and AAO surfaces. In contrast, the SHS has little dirt on it, as shown in Fig. 18(h). After water spray cleaning and drying in air, both Al and AAO surfaces are still covered by a large amount of dirt, as shown in Fig. 18(c) and (f). However, the SHS seems to be as clean as it used to be [Fig. 18(i)]. This might be attributed to the joint action of high capillary forces induced by water droplets and weak adhesion of the dirt to the SHS.⁸⁷ Therefore, it can be concluded that our as-prepared SHS can protect Al substrates from contamination in practical application.

Fig. 18 Self-cleaning process of the samples: (a–c) bare Al surface; (d–f) AAO surface; (g–i) SHS. (a) (d) (g): pristine samples before test; (b) (e) (h): immersed samples after immersion in the dirty solution for 1 min followed by drying; (c) (f) (i): clean samples after water spray and drying.

Conclusions

In summary, the SHS has been fabricated on Al surface through a process combining both anodization and chemical modification. The highest CA of 156.0 ± 0.7° and lowest SA of 2.5 ± 1.4° can be achieved when the anodization voltage reaches 20 V. The synergistic effect of hierarchical micro-nanostructure and low-surface-energy FAS is the key to realize superhydrophobicity. The SHS possesses a CA above 150° after micro-sandblast test and exposure to UV/condensation condition for 7 days, showing good mechanical stability and weathering resistance. The high hydrophobicity for different pH solutions shows the SHS is highly durable against chemical corrosion. In the meantime, the smaller J_corr and more positive E_corr as well as higher R_ct show the excellent corrosion resistance of SHS. Furthermore, the much lower icing temperature of water droplet and the extremely low ice adhesion on SHS indicate the excellent anti-icing property. The developed method could provide a straightforward and effective route to fabricate SHS on various metal substrates for a great number of engineering applications.

Acknowledgements

The authors would like to thank a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) and the PhD Abroad Short-term Visiting Project of Nanjing University of Aeronautics and Astronautics as well as the Project of NUAA-UT Group Joint Laboratory of Advanced Electronic Materials.

References

1. G. Momen and M. Farzaneh, Appl. Surf. Sci., 2014, 299, 41–46.
2. L. Feng, Y. Che, Y. Liu, X. Qiang and Y. Wang, Appl. Surf. Sci., 2013, 283, 367–374.
3. Q. Fu, X. Wu, D. Kumar, J. W. C. Ho, P. D. Kanhere, N. Srikanth, E. Liu, P. Wilson and Z. Chen, ACS Appl. Mater. Interfaces, 2014, 6, 20685–20692.
4. Y. Shen, J. Tao, H. Tao, S. Chen, L. Pan and T. Wang, Langmuir, 2015, 31, 10799–10806.
5. E. A. Cherney, IEEE Trans. Power Appar. Syst., 1980, 99, 46–52.
6. M. Farzaneh and J. F. Drapeau, IEEE Trans. Power Delivery, 1995, 10, 1038–1051.
7. O. Parent and A. Ilinca, Cold Reg. Sci. Technol., 2011, 65, 88–96.
8. S. Chernyy, M. Järn, K. Shimizu, A. Swerin, S. U. Pedersen, K. Daasbjerg, L. Makkonen, P. Claesson and J. Iruthayaraj, ACS Appl. Mater. Interfaces, 2014, 6, 6487–6496.
9. Z. Zuo, R. Liao, C. Guo, Y. Yuan, X. Zhao, A. Zhuang and Y. Zhang, Appl. Surf. Sci., 2015, 331, 132–139.
10. T. Bharathidasan, S. V. Kumar, M. S. Bobji, R. P. S. Chakradhar and B. J. Basu, Appl. Surf. Sci., 2014, 314, 241–250.
11. Y. Lai, X. Gao, H. Zhuang, J. Huang, C. Lin and L. Jiang, Adv. Mater., 2009, 21, 3799–3803.
12. J. Huang, Y. Lai, F. Pan, L. Yang, H. Wang, K. Zhang, H. Fuchs and L. Chi, Small, 2014, 10, 4865–4873.
13. A. V. Singh, A. Rahman, N. V. G. Sudhir Kumar, A. S. Aditi, M. Galluzzi, S. Bovio, S. Barozzi, E. Montani and D. Parazzoli, Mater. Des., 2012, 36, 829–839.
14. S. H. Li, J. Y. Huang, M. Z. Ge, S. W. Li, T. L. Xing, G. Q. Chen, Y. Q. Liu, K. Q. Zhang, S. S. Al-Deyab and Y. K. Lai, Mater. Des., 2015, 85, 815–822.
15. D. Kumar, X. Wu, Q. Fu, J. W. C. Ho, P. D. Kanhere, L. Li and Z. Chen, Appl. Surf. Sci., 2015, 344, 205–212.
16. Y. Lai, Y. Tang, J. Gong, D. Gong, L. Chi, C. Lin and Z. Chen, J. Mater. Chem., 2012, 22, 7420–7426.
17. S. Peng, D. Tian, X. Yang and W. Deng, ACS Appl. Mater. Interfaces, 2014, 6, 4831–4841.
18. S. Zheng, C. Li, Q. Fu, W. Hu, T. Xiang, Q. Wang, M. Du, X. Liu and Z. Chen, Mater. Des., 2016, 93, 261–270.
19. B. Zhang, Y. Li and B. Hou, RSC Adv., 2015, 5, 100000–100010.
20. S. Zheng, C. Li, Q. Fu, M. Li, W. Hu, Q. Wang, M. Du, X. Liu and Z. Chen, Surf. Coat. Technol., 2015, 276, 341–348.
21. Z. She, Q. Li, Z. Wang, C. Tan, J. Zhou and L. Li, Surf. Coat. Technol., 2014, 251, 7–14.
22. Y. Liu, H. Cao, Y. Chen, S. Chen and D. Wang, RSC Adv., 2016, 6, 2379–2386.
23. Q. T. Fu, E. J. Liu, P. Wilson and Z. Chen, Phys. Chem. Chem. Phys., 2015, 17, 21492–21500.
24. Y. Shen, J. Tao, H. Tao, S. Chen, L. Pan and T. Wang, RSC Adv., 2015, 5, 32813–32818.
25. M. A. Sarshar, C. Swarctz, S. Hunter, J. Simpson and C. H. Choi, Colloid Polym. Sci., 2013, 291, 427–435.
26. M. Ruan, W. Li, B. Wang, B. Deng, F. Ma and Z. Yu, Langmuir, 2013, 29, 8482–8491.
27. C. Wang, T. Yao, J. Wu, C. Ma, Z. Fan, Z. Wang, Y. Cheng, Q. Lin and B. Yang, ACS Appl. Mater. Interfaces, 2009, 1, 2613–2617.
28. J. Y. Huang, S. H. Li, M. Z. Ge, L. N. Wang, T. L. Xing, G. Q. Chen, X. F. Liu, S. S. Al-Deyab, K. Q. Zhang, T. Chen and Y. K. Lai, J. Mater. Chem. A, 2015, 3, 2825–2832.
29. S. Yu and Z. Guo, RSC Adv., 2015, 5, 107880–107888.
30. Z. She, Q. Li, Z. Wang, L. Li, F. Chen and J. Zhou, ACS Appl. Mater. Interfaces, 2012, 4, 4348–4356.
31. D. Kumar, X. Wu, Q. Fu, J. W. C. Ho, P. D. Kanhere, L. Li and Z. Chen, Mater. Des., 2015, 86, 855–862.
32. W. Li and A. Amirfazli, Adv. Colloid Interface Sci., 2007, 132, 51–68.
33. W. Li and A. Amirfazli, J. Colloid Interface Sci., 2005, 292, 195–201.
34. X. Wu, Q. Fu, D. Kumar, J. W. C. Ho, P. Kanhere, H. Zhou and Z. Chen, Mater. Des., 2016, 89, 1302–1309.
35. Q. Zhao, L. Y. L. Wu, H. Huang and Y. Liu, Mater. Des., 2016, 92, 541–545.
36. L. Feng, Y. Song, J. Zhai, B. Liu, J. Xu, L. Jiang and D. Zhu, Angew. Chem., Int. Ed., 2003, 42, 800–802.
37. B. Qian and Z. Shen, Langmuir, 2005, 21, 9007–9009.
38. X. Gao, W. Tong, X. Ouyang and X. Wang, RSC Adv., 2015, 5, 84666–84672.
39. C. Dong, Y. Gu, M. Zhong, L. Li, K. Sezer, M. Ma and W. Liu, J. Mater. Process. Technol., 2011, 211, 1234–1240.
40. J. Li, Z. Jing, Y. Yang, Q. Wang and Z. Lei, Surf. Coat. Technol., 2014, 258, 973–978.
41. V. A. Ganesh, A. S. Nair, H. K. Raut, T. T. Yuan Tan, C. He, S. Ramakrishna and J. Xu, J. Mater. Chem., 2012, 22, 18479–18485.
42. C. Jeong and C.-H. Choi, ACS Appl. Mater. Interfaces, 2012, 4, 842–848.
43. H. Wang, D. Dai and X. Wu, Appl. Surf. Sci., 2008, 254, 5599–5601.
44. N. Tasaltin, D. Sanli, A. Jonáš, A. Kiraz and C. Erkey, Nanoscale Res. Lett., 2011, 6, 1–8.
45. C. R. Crick, J. C. Bear, A. Kafizas and I. P. Parkin, Adv. Mater., 2012, 24, 3505–3508.
46. W. G. Bae, K. Y. Song, Y. Rahmawan, C. N. Chu, D. Kim, D. K. Chung and K. Y. Suh, ACS Appl. Mater. Interfaces, 2012, 4, 3685–3691.
47. N. Saleema, D. Sarkar, D. Gallant, R. Paynter and X. Chen, ACS Appl. Mater. Interfaces, 2011, 3, 4775–4781.
48. T. Darmanin, E. T. De Givenchy, S. Amigoni and F. Guittard, Adv. Mater., 2013, 25, 1378–1394.
49. F. Zhang, L. Zhao, H. Chen, S. Xu, D. G. Evans and X. Duan, Angew. Chem., Int. Ed., 2008, 47, 2466–2469.
50. C. Liu, F. Su and J. Liang, RSC Adv., 2014, 4, 55556–55564.
51. C. Jeong, J. Lee, K. Sheppard and C. H. Choi, Langmuir, 2015, 31, 11040–11050.
52. W. Li, X. Zhang, J. Yang and F. Miao, J. Colloid Interface Sci., 2013, 410, 165–171.
53. P. Kim, T.-S. Wong, J. Alvarenga, M. J. Kreder, W. E. Adorno-Martinez and J. Aizenberg, ACS Nano, 2012, 6, 6569–6577.
54. N. Wang, D. Xiong, Y. Deng, Y. Shi and K. Wang, ACS Appl. Mater. Interfaces, 2015, 7, 6260–6272.
55. W. Liu, Y. Luo, L. Sun, R. Wu, H. Jiang and Y. Liu, Appl. Surf. Sci., 2013, 264, 872–878.
56. P. M. Barkhudarov, P. B. Shah, E. B. Watkins, D. A. Doshi, C. J. Brinker and J. Majewski, Corros. Sci., 2008, 50, 897–902.
57. B. Yin, L. Fang, J. Hu, A. Q. Tang, J. He and J. H. Mao, Surf. Interface Anal., 2012, 44, 439–444.
58. T. He, Y. Wang, Y. Zhang, Q. Lv, T. Xu and T. Liu, Corros. Sci., 2009, 51, 1757–1761.
59. Y. Zhang, X. Yu, H. Wu and J. Wu, Appl. Surf. Sci., 2012, 258, 8253–8257.
60. J. Yang and W. Li, J. Alloys Compd., 2013, 576, 215–219.
61. M. Song, Y. Liu, S. Cui, L. Liu and M. Yang, Appl. Surf. Sci., 2013, 283, 19–24.
62. M. Zou, S. Beckford, R. Wei, C. Ellis, G. Hatton and M. A. Miller, Appl. Surf. Sci., 2011, 257, 3786–3792.
63. S. A. Kulinich, S. Farhadi, K. Nose and X. W. Du, Langmuir, 2011, 27, 25–29.
64. S. Farhadi, M. Farzaneh and S. A. Kulinich, Appl. Surf. Sci., 2011, 257, 6264–6269.
65. S. Li, J. Wang, Y. Li, X. Zhang, G. Wang and C. Wang, Superlattices Microstruct., 2014, 75, 294–302.
66. J. Y. Wang, C. Li, C. Y. Yin, Y. H. Wang and S. L. Zheng, Surf. Coat. Technol., 2014, 258, 615–623.
67. S. Peng, X. Yang, D. Tian and W. Deng, ACS Appl. Mater. Interfaces, 2014, 6, 15188–15197.
68. I. Lee and R. P. Wool, Thin Solid Films, 2000, 379, 94–100.
69. W. Xu, J. Song, J. Sun, Y. Lu and Z. Yu, ACS Appl. Mater. Interfaces, 2011, 3, 4404–4414.
70. R. N. Wenzel, Ind. Eng. Chem., 1936, 28, 988–994.
71. R. E. Johnson and R. H. Dettre, J. Phys. Chem., 1964, 68, 1744–1750.
72. A. B. D. Cassie and S. Baxter, Trans. Faraday Soc., 1944, 40, 546–551.
73. L. Jiang and L. Feng, Bioinspired intelligent nanostructured interfacial materials, Chemical Industry Press, Beijing, 2007.
74. Y. Huang, D. K. Sarkar, D. Gallant and X. G. Chen, Appl. Surf. Sci., 2013, 282, 689–694.
75. Y. Liu, X. Yin, J. Zhang, S. Yu, Z. Han and L. Ren, Electrochim. Acta, 2014, 125, 395–403.
76. A. M. A. Mohamed, A. M. Abdullah and N. A. Younan, Arabian J. Chem., 2015, 8, 749–765.
77. T. Liu, S. Chen, S. Cheng, J. Tian, X. Chang and Y. Yin, Electrochim. Acta, 2007, 52, 8003–8007.
78. L. Feng, H. Zhang, Z. Wang and Y. Liu, Colloids Surf., A, 2014, 441, 319–325.
79. Y. Yin, T. Liu, S. Chen, T. Liu and S. Cheng, Appl. Surf. Sci., 2008, 255, 2978–2984.
80. P. Guo, Y. Zheng, M. Wen, C. Song, Y. Lin and L. Jiang, Adv. Mater., 2012, 24, 2642–2648.
81. M. Wen, L. Wang, M. Zhang, L. Jiang and Y. Zheng, ACS Appl. Mater. Interfaces, 2014, 6, 3963–3968.
82. N. H. Fletcher, J. Chem. Phys., 1958, 29, 572–576.
83. M. J. Jamieson, C. E. Nicholson and S. J. Cooper, Cryst. Growth Des., 2005, 5, 451–459.
84. S. J. Cooper, C. E. Nicholson and J. Liu, J. Chem. Phys., 2008, 129, 124715.
85. H. J. Ensikat, A. J. Schulte, K. Koch and W. Barthlott, Langmuir, 2009, 25, 13077–13083.
86. Y. Wang, J. Xue, Q. Wang, Q. Chen and J. Ding, ACS Appl. Mater. Interfaces, 2013, 5, 3370–3381.
87. F. Su and K. Yao, ACS Appl. Mater. Interfaces, 2014, 6, 8762–8770.

---

Footnote

† Electronic supplementary information (ESI) available: Video S1: a water droplet of 4 μL is dripped onto the AAO surface and spreads all over the surface, indicating its superhydrophilicity; Video S2: the water droplets of 10 μL can roll away easily after dripped onto the SHS; Video S3: the SHS shows the surface has an extremely low water adhesion after repeated icing/deicing processes. See DOI: 10.1039/c6ra13447e

---