3.1. Synthesis and Identification of the FPSZs
FPSZs with low surface energy were synthesized via hydrolysis–condensation reactions of OPSZ with FAS-17. The typical FPSZ synthesis reactions are presented in
Scheme 1B, and the synthesis steps were as follows. The methylhydrogensilane moieties (CH
3–Si–H) of OPSZ reacted rapidly with a small amount of water to form silanol groups (CH
3–Si–OH). The reactions involving the CH
3–Si–OH groups proceeded via two primary pathways: (1) hydrolysis–condensation with the methoxysilane groups of FAS-17 and (2) condensation of the CH
3–Si–OH groups. For reaction pathway (2), the cross-linking of the OPSZ chains increased its molecular weight, and then, the OPSZ chains reacted with FAS-17 to form an FPSZ with a basic structure ((3) in
Scheme 1B).
The
1H-NMR spectra of FAS-17, OPSZ, initial mixture of the reactants, and FPSZ
17.3 (
Figure 2) confirmed the formation of FPSZs via the reaction of FAS-17 with OPSZ. Moreover, the spectra were used to determine the chemical structures and content of fluorinated moieties of the fabricated FPSZs. The
1H-NMR spectra of the OPSZ and FAS-17 mixture (
Figure 2C) was compared with that of FPSZ
17.3 (
Figure 2D). The characteristic peaks in the
1H-NMR spectrum of the OPSZ and FAS-17 mixture matched those in the spectra of FAS-17 and OPSZ (
Figure 2A,B, respectively). No significant changes were observed in the positions of the characteristic peak of the trimethoxy group (–O–CH
3, 3.6 ppm) of FAS-17 and those of the hydrogen silane (Si–H, 4.35–5 ppm) and triethoxy groups (–O–CH
2–CH
3, 3.84 ppm and –O–CH
2–CH
3, 1.24 ppm) of OPSZ.
However, the positions of the characteristic peaks of the –O–CH3 and methylene (CH2–CF2) peaks in the 1H-NMR spectrum of FPSZ17.3 were shifted compared with those of the corresponding peaks in the 1H-NMR spectrum of FAS-17. The single –O–CH3 peak in the 1H-NMR spectrum of FAS-17 was split over the range of 3.45–3.6 ppm in the 1H-NMR spectrum of FPSZ17.3. The peak of CH2–CF2 was broadened and shifted to the left. The integration ratio of the CH2–CF2 and –O–CH3 peaks of 2:9 in the spectrum of FAS-17 was changed to 2:4.3 in the spectrum of FPSZ, indicating that half of the –O–CH3 groups participated in the reaction. In addition, the integral value of Si–H was lower than that of –Si–CH3 in the CH3–Si–H groups of FPSZ. Moreover, no significant changes were observed in the peak positions of the water-hydrolyzable N–H and triethoxysilane groups. These changes in peak positions indicated that hydrolysis and condensation with a small amount of water were the primary reactions between the CH3–Si–H groups of OPSZ and the trimethoxy groups of FAS-17. Therefore, FAS-17 was introduced into the backbone of OPSZ via hydrolysis and condensation reactions.
The weight contents of the fluoroalkyl (R
f) moieties of the FPSZs were calculated by comparing the integrals of the –CH
2–CF
2 (2.1 ppm) and –Si–CH
3 (0.1 ppm) peaks in the
1H-NMR spectrum of FPSZ
17.3 (
Figure 2D), and the results are summarized in
Table 1. The proportion of R
f moieties grafted onto OPSZ increased with increasing FAS-17 feeding ratio, indicating that the R
f content of FPSZ can be well controlled by adjusting the FAS-17 feeding ratio. The conversion rates of all the reforming reactions were confirmed to be higher than 80%, indicating that this was a suitable method for introducing R
f moieties into the OPSZ backbone.
3.2. Fabrication of Superhydrophobic Al Surface with Hierarchical Micro-Nanostructure
The synthesized FPSZs were coated onto Al surfaces with micro-nano hierarchical structures to fabricate superhydrophobic surfaces. The typical method for fabricating superhydrophobic Al surfaces with hierarchical structures is presented in
Figure 3. To obtain hierarchical structures on the Al surface, Al plates were etched with a HCl solution to generate a rectangular-shaped micro-roughness. Next, the micro-rough Al plates were immersed in boiling water, and nanostructures grew on the surfaces to obtain the final micro-nano hierarchical structure. Subsequently, the hierarchically rough Al plates were immersed in FPSZ dilute solutions, followed by drying to obtain superhydrophobic Al surfaces.
The scanning electron microscopy (SEM) images of the micro-nanostructured Al surfaces are presented in
Figure 4. After etching with the HCl solution, the Al surface was rough and comprised rectangular-shaped microscale patterns with lengths in the range of 0.2–1.0 μm (
Figure 4A,B). This microstructure was formed because of selective etching of vulnerable dislocation inside the Al crystals [
30] Upon immersing the etched plate with a microscale pattern in boiling water, the Al surface reacted with water to form crystalline boehmite (AlO(OH)), resulting in a nano-flake structure [
25,
31]. The SEM images of the Al surface after immersion in boiling water (
Figure 4C,D) revealed that a 20 nm thick × 200 nm long flower petal-shaped nanostructure was well formed and densely distributed on individual microscale Al spheres. The superhydrophilicity of the hierarchical micro-nano structure was attributed to the hydrophilicity of the surface, and a material with a low surface energy was required to convert this surface into a superhydrophobic surface. FPSZs were primarily fabricated as coating materials to enhance the superhydrophobicity and anti-icing properties of Al surfaces. To evaluate the properties of the fabricated FPSZs, hierarchically micro-nanostructured Al samples were coated with FPSZs using an immersion method.
After the hierarchical micro-nano structured Al plates were coated with FPSZs, they were subjected to water–oil wettability measurements, and the results are summarized in
Table 1. To compare the effect of fluorinated functional groups, unmodified OPSZ and OPSZ–FAS-17 mixtures were also coated on hierarchical Al plates, and the water wettability of these surfaces was evaluated. The water CA/SA values on the OPSZ- and OPSZ–FAS-17 mixture-coated surfaces were 167°/4° and 169°/7°, respectively. The superhydrophobicity of the OPSZ-coated surface was ascribed to the methyl groups in the backbone, and the distribution of the CA/SA values was even over the entire measurement area. For the oil CA measurements using hexadecane, the OPSZ-coated surface was wetted because of the presence of the lipophilic methyl groups of OPSZ. Furthermore, the water CA of the OPSZ–FAS-17 mixture-coated surface was higher than that of the OPSZ-coated surface; however, the sliding properties of the OPSZ–FAS-17 mixture-coated surface were inferior to those of the OPSZ-coated surface. Moreover, the error ranges of the CA and SA of the OPSZ–FAS-17 mixture-coated surface were three and eight times larger, respectively, than those of the OPSZ-coated surface. This indicated that the OPSZ–FAS-17 mixture did not form an even coating over the entire measurement area because both OPSZ and FAS-17 formed Al–O–Si bonds [
32] with the Al surface. The OPSZ layer did not form easily in the areas where amphiphobic FAS-17 reacted first with the Al surface; therefore, the Al surface was not evenly coated. Upon placing hexadecane drops onto the analyzed surfaces, hexadecane wetted most of the surface; however, the oil droplets presented CAs of approximately 50° over small areas.
The errors ranges of the CAs and SAs of the FPSZ-coated surfaces were smaller than those of the OPSZ–FAS-17 mixture-coated surface. In addition, the water CAs and SAs of the FPSZ-coated surfaces increased and gradually decreased, respectively, by increasing the content of R
f moieties introduced into the OPSZ backbone. Because the surface energies of fluorocarbons comprising –CF
3 and –CF
2 functional groups are lower than those of the corresponding hydrocarbons, increasing the fluorinated content of FPSZ increased the hydrophobicity of the coated surface. When the R
f content of FPSZ reached 17.3 wt%, the water CA and SA of the coated surface increased to 175° and decreased to 1.6°, respectively, and the CA and SA values were unchanged even at higher R
f contents. Moreover, the number of Si–H functional groups of the FPSZ backbone was decreased by the introduction of excess R
f moieties, which hindered the formation of a dense coating layer. As for the oil CA, hexadecane wetted the surfaces of the samples coated with FPSZs with R
f contents of up to 12.7 wt%, and the oil CA was 50° for the FPSZ
17.3-coated sample. The dependence of the oil CA on the R
f content was similar to that reported in our previous paper; however, the method used to introduce R
f moieties into the backbone of OPSZ in this study was simpler [
21].
3.3. Anti-Icing Properties
The primary parameters for evaluating the anti-icing performance of the fabricated surfaces were the ice nucleation rate and ice adhesion strength. To exhibit excellent anti-icing properties, surfaces must present low ice adhesion strength and excellent ability to inhibit ice nucleation [
33,
34]. In our previous study, we confirmed that FPSZs are good anti-icing materials [
21]. The ice nucleation rates of the FPSZ-coated Al surfaces in this study indicated that the FPSZs fabricated herein presented remarkable potential as anti-icing materials (
Figure 5). Improvement in icephobicity was expected with increasing the R
f content of FPSZs, which was confirmed by the delays in condensed water crystallization [
27]. The changes in the crystallization temperature of water droplets were determined using the DSC curves obtained by continuously lowering the temperature, and the crystallization temperature was defined as the onset temperature of the freezing peak. When water freezes, latent heat is released and the temperature of the droplet is raised, resulting in the formation of the slanted peaks. According to the DSC curves (
Figure 5), the freezing temperatures of water droplets on the superhydrophobic OPSZ- and FPSZ-coated surfaces were lower than −19.5 °C, whereas that on a flat Al surface was −11.6 °C. Because abundant air pockets formed between water droplets and the micro-nano structured Al surfaces coated with low-surface-energy materials, heat transfer and ice formation were delayed. As the R
f content of FPSZ was increased, the freezing point of the water droplets gradually decreased and reached the lowest value of −22.0 °C for FPSZ
41.5. In addition, the freezing temperature of FAS-17, with an R
f group content of 100 wt%, was the lowest among all the samples (−22.3 °C). The decrease in the freezing temperature of water with increasing R
f content was consistent with the increase in the water/oil Cas. The freezing temperature of water on the FPSZ-coated surfaces decreased with increasing R
f content because the surface energy of the R
f group was significantly lower than that of the methyl groups of OPSZ. Therefore, the contact area between water and the coating layer was minimized, and heat transfer was reduced. Moreover, crystallization times were calculated when the DSC curves in
Figure 5 was plotted as a function of time. These are in the range of 24 to 33 s, but any consistent pattern was not found for all samples.
DSC is a useful method for comparing the ice-forming temperatures of superhydrophobic surfaces under low RH conditions because measurements are performed with a nitrogen atmosphere. However, as the RH increases, the wettability of some superhydrophobic surfaces changes to a “Wenzel state,” which deteriorates the anti-icing performance of the surfaces [
27]. The freezing of water droplets on the surfaces at 70 ± 5% RH and −19.0 °C was recorded, and images were captured at different times (
Figure 6). The freezing times of water droplets on the OPSZ-, FPSZ
17.3-, FPSZ
25.1-, and FAS-17–coated Al surfaces were 334, 4914, 2657, and 2726 s, respectively. The freezing delay time of the water droplets on the FPSZ
17.3-coated Al surface was the longest. The freezing delay time of the FAS-17-coated Al surface was significantly shorter than that of the FPSZ
17.3-coated Al surface and was similar to that of FPSZ
25.1-coated Al surface. These results were closely related to the increase in ice adhesion strength on the surface of the FAS-17-coated Al with increasing RH reported in our previous paper [
27]. The substrate–liquid interface of the superhydrophobic surface changed from the “Cassie state” to the Wenzel state because of the increase in RH; moreover, the contact area was widened. The FPSZ
17.3-coated surface presented the best freezing delay effect because the change in solid–liquid interface with RH was negligible compared with those of the other surfaces.
The freezing delay time indicates how long ice nucleation is delayed at the water–solid interface. The anti-icing properties of surfaces can also be evaluated using ice adhesion strength measurements. Ice adhesion strength measurements consist of directly measuring the force required to remove ice frozen on a solid surface, and low ice adhesion strength values translate into excellent anti-icing properties. The ice adhesion strengths of the OPSZ- and FPSZ-coated Al plates are presented in
Figure 7. The initial ice adhesion strengths of the FPSZ-coated Al plates were lower than that of the OPSZ-coated Al plate. Although the OPSZ- and FPSZ-coated Al plates presented superhydrophobic properties in wettability tests, the ice adhesion strengths of the FPSZs were relatively superior to that of the OPSZ-coated Al plate because of the low surface energy of the R
f groups, which significantly affected ice adhesion strength [
21]. Ice adhesion strength decreased with increasing R
f content up to 17.3 wt%; however, it increased significantly at R
f contents higher than 17.3 wt%. This was attributed to the decrease in solubility of FPSZs with high R
f content and the decrease in coating–substrate compatibility. Furthermore, the ice adhesion strength of the FAS-17-coated Al plate with the highest R
f content (100%) was 140 kPa, which was higher than that of the FPSZ
17.3-coated Al plate (78 kPa). As the R
f content was increased, a dense R
f layer formed at the air–solid interface. The dense layer prevented the effective exposure of the surface Al–OH groups to water vapor, and the effect was maximized at an R
f content of 17.3 wt%. However, at R
f contents >17.3 wt%, the number of Si–H functional groups of FPSZ capable of self-condensation was relatively low; therefore, the coating layer became thinner. In addition, the FPSZs with extremely high R
f contents presented low solubility and high adhesion to the substrate [
21]. This led to the formation of a thin, non-uniform coating layer, which deteriorated the icephobicity of the coating and presented a wide error range. Unlike FPSZs, FAS-17 formed a uniform monolayer coating; however, the coating layer was thin. Consequently, the ice adhesion strength of the FAS-17-coated surface was intermediate compared with those of the FPSZ-coated surfaces. To obtain an FPSZ-coated superhydrophobic surface with low ice adhesion, abundant air pockets should form under the water droplets, and the contact surface area of the water droplets with the nanostructured surface should be minimized. Upon compressing the air pockets trapped in structures, they act as stress concentrators and exert a counter force on the ice column, leading to a decrease in ice adhesion strength [
15,
16]. The air pockets minimize the contact surface area between the ice and the substrate even during water freezing. Superhydrophobic Al surfaces can become mechanically damaged, and the middle of the ice column on the surface can be broken if the number of the air pockets at the interface is insufficient because of incomplete coating coverage or the lack of a dense coating layer [
12,
35]. In other words, water vapor that diffused into the air pockets during ice formation condensed at the hydrophilic sites of the imperfect superhydrophobic surface; therefore, the water droplets formed “Wenzel ice” [
14,
17]. To minimize the formation of Wenzel ice, the interactions between the hydrophilic Al–OH groups at the Al surface and water should be minimized by maintaining sufficient distance between them. Thick coating layers appear to minimize these interactions; however, the hierarchical nanostructured pores can be filled with coating material, which suppresses superhydrophobicity. Thickness control at the molecular level is required to simultaneously achieve a sufficiently thick coating layer and superhydrophobicity.
The surface energy of the R
f groups of FPSZs is lower than that of the CH
3–Si groups of OPSZ and the R
f chains arrange densely at the solid–air interface above a certain R
f content [
36,
37]. The thickness of the coating layer increased at the molecular level because of the surface orientation of the R
f groups. The XPS data in
Table 2 indicate the surface-covering effect caused by the orientation of the R
f groups after FPSZ coating. XPS analysis can be used to identify the atoms present in a 10 nm thick layer beneath the surface. Our results indicated that the amount of Al in the substrate decreased as the thickness of the coating layer increased. The XPS-determined Al contents of the OPSZ- and FPSZ
17.3-coated Al samples were 16.8% and 1.4%, respectively. By contrast, the Al content of the FAS-17-coated Al sample was 27.6%, indicating that the FPSZ
17.3 coating layer was thicker than the OPSZ and FAS-17 coating monolayers. In addition, as the R
f content of the FPSZ layer increased, the Al 2p values were decreased until the R
f reached 17.3 and were then increased at contents higher than 17.3. The coating layers of FPSZs with high R
f contents (>17.3) were relatively non-uniform compared to that of FPSZ
17.3, and non-uniformity was closely related to the rapid increase in ice adhesion strength.
3.4. Durability of the Anti-Icing Surfaces
Superhydrophobic surfaces exhibit low ice adhesion strength and delay the freezing of water droplets on them; however, surface durability is not proportional with surface wettability (CA and SA) or initial ice adhesion strength. A hierarchically rough surface is typically weak under physical impact because of the fragility of the micro-nano surface structure. During freezing, water expansion against the substrate causes damage to the hierarchical structure such that ice adhesion strength increases and surface durability deteriorates. Therefore, it is critical to improve the mechanical stability and durability of superhydrophobic surfaces for practical applications. In this study, the robustness and durability of the superhydrophobic surfaces were studied by evaluating the ice adhesion strength of the icephobic coatings after multiple icing–melting cycles (
Figure 7). The adhesion between the OPSZ-coated surface and the ice column was remarkably strong. After 40 icing–melting cycles the ice column broke, and ice fragments remained on the surface. The damage to the OPSZ-coated surface and loss of icephobic properties after multiple icing–melting cycles were associated with the change in wetting state that occurred when the water in the column froze [
38]. During repeating icing–melting cycles, the water vapor inside the surface pores underwent diffusion and condensation repeatedly, causing the wettability state to gradually change from Cassie–Baxter to Wenzel [
6]. When water freezes in the Wenzel state, the volume of Wenzel water expands inside the pores. The force generated via this volume expansion causes structural damage to the superhydrophobic surface, resulting in a sharp increase in ice adhesion strength after multiple icing–melting cycles. To improve the durability of superhydrophobic surfaces, it is critical to prevent water from penetrating and condensing into the surface pores during icing–melting cycles. In addition, the structure should present adequate strength to withstand the force caused by the volume expansion that occurs during ice formation. The ice-adhesion strength of the FPSZ-coated samples increased after 40 icing–melting cycles, and intermittent cohesive failures were observed for the FPSZ
4.3- and FPSZ
41.5-coated samples. Conversely, the ice adhesion strength of the FPSZ
17.3-coated sample after 40 icing–melting cycles (141 kPa) was 80.7% higher than the initial value. The ice adhesion strengths of the FPSZ
8.2-, FPSZ
12.7-, FPSZ
25.1-, and FAS-17-coated samples after 40 icing–melting cycles were 95.3%, 100%, 92.4%, and 82.0% higher, respectively, than the initial values. Moreover, these values were higher than that of the FPSZ
17.3-coated sample. Therefore, the surface durabilities of these samples were inferior to that of the FPSZ
17.3-coated sample. Consequently, we concluded that FPSZ
17.3 was a suitable coating material, featuring a good balance between the excellent hardness of OPSZ and the low surface energy of FAS-17. Unlike the FAS-17 monolayer coating, the relatively thick FPSZ
17.3 coating layer can protect the Al surface from direct physical impact owing to its excellent mechanical properties. FPSZ
17.3, which is a polysilazane-based material, formed a hard coating layer through self-condensation and formed chemical bonds with the Al substrate surface. The hard coating physically protected the nanostructure of the physically weak superhydrophobic Al surface. As expected, the nanoscale structure of the FAS-17 coating (
Figure S1) was almost completely damaged after sand impact, whereas the structure of the FPSZ
17.3 coating was less damaged. Unlike the durability of the FAS-17 coating, that of the FPSZ
17.3 coating was improved after 40 icing–melting cycles. The initial ice formation temperature of FAS-17 (−22.3 °C) increased to −20.7 °C after 40 icing–melting cycles (
Figure 8). The freezing temperature of water droplets on the FPSZ
17.3-coated surface after 40 icing–melting cycles was almost unchanged, indicating the superior durability of FPSZ
17.3 relative to that of FAS-17. The surface coated with FPSZ
17.3 showed anti-icing performance comparable to other systems described in the literature [
19,
20].