1. Introduction
In the past two decades, great attention has been paid to superhydrophobic coatings due to their exciting prospects for self-cleaning
[1],
[2], anti-corrosion
[3],
[4],
[5],
[6],
[7],
[8],
[9], anti-fouling
[10],
[11],
[12],
[13], droplet directional transport
[14],
[15],
[16],
[17], and oil–water separation
[18],
[19] applications, among others. Superhydrophobic coatings with good repellency to corrosive solutions can significantly decrease the interaction between the corrosive species and the substrate, endowing the coating with excellent corrosion resistance
[20],
[21],
[22],
[23],
[24]. However, this type of coating usually exhibits poor mechanical properties
[25],
[26],
[27]. This is because the construction of micro- and/or nano-structures on the surface is the key to achieving superhydrophobicity, but doing so results in a surface with high roughness. Thus, even a small mechanical load will lead to a high local pressure on the surface structure, eventually causing damage to the surface structure and weakening or eliminating the superhydrophobicity. This intrinsic weakness of superhydrophobic coatings has become a bottleneck restricting their practical long-term applications. Superhydrophobic coatings are prone to wear, erosion, and cracking under the continuous influence of the working environment, which includes particle impacts, raindrop erosion, and gas etching. Once a crack has been formed, moisture may accumulate in it, causing severe electrochemical corrosion (
Fig. 1(a))
[28],
[29],
[30].
To date, many strategies have been proposed to solve these challenges. Similar to the introduction of steel or fibers into concrete, fibers with a large aspect ratio have been used to enhance the mechanical strength of superhydrophobic coatings and the interaction between the dispersed phase and the continuous phase
[31],
[32]. Kim et al.
[31] used dispersed graphene liquid crystalline fibers to form a framework in adhesive polydopamine, enhancing the strength of the coating. In another strategy, a buffer structural layer was prepared on a coating to disperse impact energy; the buffer layer worked like the tendon tissue between bones in the human body
[33],
[34], minimizing stress at the interface through deformation
[35],
[36]. As a third strategy, a moisture-absorption layer such as a hydrogel can be introduced as part of a composite coating to enhance the coating’s durability. When the coating cracks, the hydrogel spontaneously fills in the gaps and blocks the penetration of external moisture, thereby avoiding or minimizing electrochemical corrosion. For example, Zhang et al.
[34] used a hydrogel mixed with SiO
2 nanoparticles (NPs) to improve the corrosion resistance and bonding strength of a coating.
In the present work, we constructed a biomimetic enamel-like coating by combining the three strategies described above. First, amylose hydrogels were obtained by cross-linking Ca(NO3)2∙4H2O and amylose in an aqueous solution. This viscous gel was then introduced into a coating to act as a filling agent for self-healing. When the top coating is seriously damaged, the moisture at the damaged site initiates the repair. By absorbing water, the prepared hydrogel can expand along the cracked surface and quickly fill the gap. The hydrogel not only heals cracks but also reduces the corrosion of the substrate. The erosion rate and corrosion potential of the completely repaired coating can be maintained at 40 nm∙s−1 and −0.21 V, respectively. At the same time, the soft and deformable hydrogel layer effectively buffers the impact pressure on the top coating. In comparison with a coating without this buffer structure, the stress value of the coating surface decreases by 52.3% under the impact of particles, based on an ABAQUS simulation.
The top layer, which is hierarchically structured, is constructed through multiple spin coatings and hot pressing. By controlling the content of lauric acid (LA)@TiO
2 and carbon nanotubes (CNTs) from 0.01 to 0.04 g, the viscoelasticity of the coating increases layer by layer from the top downwards. Inspired by the layered structure of dental enamel in the human body, this coating has a bionic tooth structure (
Figs. 1(b) and
(c)) that improves its energy dispersion under an impact loading, prevents cracks from propagating, and maintains its structural integrity. The hydrophobicity (contact angle (CA) = 152.1°; sliding angle (SA) = 4.4°) of the coating surface is improved by the LA and TiO
2 NPs, and the fibrous particles enhance the interfacial interaction between the dispersed phase and the continuous phase of the epoxy resin (EP). The protection strategy of the layered structure of dental enamel coating with the hydrogel layer (the DEA coating) is illustrated in
Figs. 1(c) and
(d).
2. Materials and methods
2.1. Materials
Amylopectin hydrate (from waxy corn) was purchased from Tokyo Chemical Industry Co., Ltd. (Japan). Multi-walled CNTs, LA (i.e., dodecanoic acid), and TiO2 (100 nm) were provided by Aladdin (China). EP and polyurethane (PU) were bought from Zhenjiang Danbao Resin Co., Ltd. (China) and Yichun Zhuoyue Chemical Co., Ltd. (China), respectively. Ca(NO3)2∙4H2O and anhydrous ethanol were obtained from Sinopharm Chemical Reagent Co., Ltd. (China), and deionized water was produced using a UPH-II-20T water purification system (Chengdu YouPu Biotechnology Co., Ltd., China). 316L steel samples were ultrasonically cleaned with acetone, ethanol, and deionized water for 15 min before use.
2.2. The rational robustness strategy
The sample preparation process is shown in
Fig. 2. The coating is composed of LA, CNTs, TiO
2 NPs, EP, and PU (the curing agent for the EP). First, 0.2 g LA is dispersed in ethanol for 10 min. Then, 300 μL of tetraethyl titanate is slowly added to the solution and stirred magnetically at 400 r∙min
−1 for 20 min. Subsequently, 200 μL of deionized water and 10 μL of 0.2 mol∙L
−1 HCl are introduced, and the mixture is stirred at 700 r∙min
−1 for an additional 20 min. The particle dispersion is centrifuged at 10 000 rad∙min
−1 for 3 h to obtain the modified TiO
2 NPs; the polar groups of the LA connect with the TiO
2 NPs by forming ester groups, while the non-polar groups at the tail grant the particles hydrophobicity. Next, a mixture of functional LA@TiO
2 and CNTs (mass ratio of LA@TiO
2 to CNTs is 1:2) is uniformly dispersed in water and then spray-coated onto semi-solidified EP. The process of spraying-coating is repeated four times, with the mass ratio of the particle mixture to EP changing in each layer—with layers of 1:10, 1:5, 3:10, and 2:5, respectively—to form a gradient structure. Finally, a superhydrophobic wear-resistant multi-layer bionic tooth structure is obtained via a simple process of hot-pressing.
The wear resistance of TiO2 versus SiO2 NPs and the effect of the mixing ratio of TiO2 NPs and CNTs on the mechanical stability of the coating are respectively compared in Tables S1 and S2 in Appendix A. It can be seen that, when the mass ratio of TiO2 to CNTs is 2:1, the sample exhibits the best wear resistance, while the superhydrophobicity is well maintained. The relationship between the weight of the hot-pressing method and the coating thickness on the hydrophobicity of the coating is given in Table S3 in Appendix A. The optimum hot-pressing weight is 500 g (49 kPa).
2.3. Damage repair
To prepare the amylose hydrogel, 16.7 g of Ca(NO
3)
2∙4H
2O and 10 g of starch were mixed in 50 mL of deionized water. The suspension was vigorously stirred for 2 h in a 70 °C water bath. EP was spin-coated on a 316L substrate at 2000 r∙min
−1 (40 s) and cured at 120 °C to isolate moisture from the substrate. Then, 0.2 g of the obtained transparent hydrogel was evenly coated on the 316L substrate, which was pretreated with EP. Finally, the bionic coating was bonded on the top to obtain an EP@LA@TiO
2@CNTs stratified dental-enamel-like (DE) coating with an amylose hydrogel layer underneath. The repair performance of polyacrylamide hydrogel (PAAm) versus the amylose hydrogel is compared in Fig. S1 in Appendix A. PAAm expands as a whole after absorbing water, so it would not effectively fill cracks and could even cause the coating to be lifted. In contrast, during the preparation of the amylose hydrogel, the chains of amylose change from an ordered granular structure to an amorphous paste and demonstrate better fluidity than PAAm
[37]. Therefore, the hydrogel can repair along the crack rather than absorbing water and expanding as a whole.
It should be noted that the amylose hydrogel has a low Young’s modulus and high Poisson’s ratio (a stress–strain curve and uniaxial tensile test are provided in Fig. S2 in Appendix A). Consequently, it assumes a pivotal role as a buffer through deformation, making it possible to achieve the energy-dispersion goals and thereby improving the buffering performance
[38],
[39]. According to the simulation results (Fig. S3 in Appendix A), the stress value of the coating surface was 52.3% lower under the impact of particles, based on an ABAQUS simulation, in comparison with the coating without a buffer structure. For a comprehensive understanding of the simulation, detailed information is provided in Section S1 in Appendix A, and Fig. S4 in Appendix A visually presents the boundary conditions and grid distribution.
2.4. Characterization
The surface morphology, energy dispersive spectrometry (EDS) mapping, and thickness of the coating were characterized using a scanning electron microscope (S-4800, Hitachi, Japan). Fourier-transform infrared (FTIR) spectroscopy was conducted using an Avatar 360 spectrometer (Thermo Fisher Scientific, USA). CAs and SAs were determined using an optical CA measuring device (OCA 25, DataPhysics Instruments GmbH, Germany) with a 5 µL droplet and 10 μL droplet, respectively. According to the American Society for Testing and Materials (ASTM) D3359 standard, the adhesion between the coating and substrate was evaluated using 3M tape. The flexibility of the coating was evaluated by repeatedly folding it from 0° to 90°, with each folding and unfolding process being counted as one cycle. The corrosion resistance (i.e., polarization curve and impedance spectroscopy) of the coating was determined using an electrochemical workstation (ModuLab XM, AMETEK Scientific Instruments, USA).
In addition, the mechanical durability of the coatings was evaluated by means of an erosion loop test based on the SY/T 7394–2017 standard
[40],
[41]. The experimental device is illustrated in Fig. S5(a) in Appendix A. The test sample was cut to 1 cm × 1 cm and was placed in the reserved groove shown in Fig. S5(b) in Appendix A. The average size of the sand particles used in the experiment was about 200 μm (Fig. S5(c) in Appendix A). The impact speed of the sand particles in the pipe was 20 m∙s
−1. As a commonly used material for the inner lining of bimetallic composite pipes in submarine pipeline systems, bare 316L stainless steel was chosen as the control in this experiment. Specific details on the erosion loop test are provided in Section S2 in Appendix A.
3. Results and discussion
3.1. Micro morphology
To verify the successful fabrication of the amylose hydrogel, amylose and the obtained hydrogel were tested via FTIR spectroscopy. The spectral band at 3000–3700 cm
−1 in both samples was attributed to the complex stretching vibration of free hydroxyl molecules and internal hydroxyl molecules on the starch. The peak at 1631.25 cm
−1 was attributed to the stretching and bending vibration of −OH in the starch gel network (
Fig. 3(a)). Characteristic peaks at 1148.68 and 1038.09 cm
−1 were attributed to the stretching vibration peaks of the C−O−C bond in the starch. The hydrogen bonding interaction was enhanced after the hydrogel formation, since NO
3− interacts with the −OH group of amyloses through hydrogen bonding, and the hydrogel showed a trend of shifting to a lower wavelength. The thicknesses of the four-layer DEA coating and the amylose hydrogel were 498.1 and 201.5 μm, respectively (
Figs. 3(b) and
(c)). As shown in
Fig. 3(d), the TiO
2 NPs were well dispersed on the coating surface due to their modification with LA, and no obvious gaps between the four layers of DEA coating, the hydrogel layer, and the 316L substrate were found in the cross-sectional morphology of the coating (
Figs. 3(d) and
(e); Fig. S6 in Appendix A). The four layers of the DEA coating could be clearly distinguished based on the distribution of titanium (Ti) (
Fig. 3(f)), as the particle content decreased from the top layer to the bottom layer, demonstrating the successful preparation of the bionic tooth-structure coating.
To illustrate the effectiveness of the patch repair strategy,
Figs. 3(g)–(i) demonstrate the hydrogel repairing process for cracks in the coating. The hydrogel completely fills the crack along its 50 μm width by absorbing the water, thereby preventing subsequent electrochemical corrosion of the coating and greatly extending the operation life of the coating.
3.2. Superhydrophobic stability
To determine the optimal layers of the DE coating, the wear resistance of the coating was characterized by means of sandpaper grinding. It can be seen that the wear resistance was positively correlated with the number of layers when the number was no more than three (
Fig. 4(a)). As the relative density of each layer decreased (density of CNTs, TiO
2, and EP are 2.10, 4.26, and 1.20 g∙cm
−3, respectively), the absorption rate of external energy was effectively improved. When the layers reached four, the CA remained stable for 800 cycles with 600-mesh sandpaper under a 49 kPa load (
Fig. 4(a), right). It should be noted that the mechanical strength of the DEA coating is remarkable in comparison with those of reported robust superhydrophobic coatings (Fig. S7(a) in Appendix A). The coating was able to withstand a scratch test using a 6H pencil (Fig. S8 in Appendix A). Furthermore, after 1500 repeated sand impact test cycles from a 0.5 m height (
Fig. 4(b)), both the CA and SA of the DEA coating showed negligible changes, and the surface morphology of the coating was almost the same before and after the sand impact (Fig. S9 in Appendix A).
A tape-peeling test of the samples before and after crack healing is depicted in
Fig. 4(c). After 30 peeling cycles, the CA (SA) of the original coating and the repaired coating decreased to 151.9° (4.2°) and 152.2° (4.5°), respectively. Moreover, there was no obvious separation at the boundary between the hydrogel and the coating (
Fig. 4(c), right), which can be attributed to the strong adherence of the amylose hydrogel. As illustrated in
Fig. 4(d), after 450 folding cycles from 0° to 90°, no delamination or cracks were observed, indicating that the coating had good flexibility and fatigue resistance. Due to the effectively enhanced bending resistance conferred on the composite coating by the CNTs
[42],
[43],
[44],
[45], the coating still maintained its superhydrophobicity, although creases appeared at the bending part (
Fig. 4(d), right).
In addition to the mechanical stability, the electrochemical corrosion behavior was investigated. Aside from the problem of periodic erosion in a submarine pipeline, the issue of widespread electrochemical corrosion in the pipeline cannot be ignored. To illustrate the feasibility of the amylose hydrogel’s repairing behavior, we investigated the corrosion of an amylose hydrogel layer, a 316L metal sheet, a damaged DE coating, a damaged DEA coating, a DEA coating that was 1/3 repaired, a DEA coating that was 2/3 repaired, a fully repaired DEA coating (where the degree of repair was controlled by the amount of water absorbed), and an intact DEA coating. As shown in
Fig. 4(e), a local corrosion phenomenon with a large cathode and a small anode formed at the crack, which caused the corrosion potential of the cracked coating to move to the left, implying that the corrosion rate increased significantly. Due to the joint effect of the narrow crack (with a width of 50 μm) and the water absorption of the hydrogel in the crack, the slope of the impedance spectrum and the corrosion potential of the DEA coatings with different degrees of repair were basically unchanged at −2.1 V (
Figs. 4(e) and
(f)). Thus, the repair of the coating is both necessary and effective for corrosion resistance. The corrosion potential of the coating shifted to the right by about 0.07–0.11 V, reflecting the excellent electrochemical corrosion resistance of the DEA coating. (A comparison of the corrosion potentials with references is provided in Fig. S10 in Appendix A.) In addition, to evaluate the chemical stability, the sample was immersed in solutions with different pH values (HCl, NaCl, NaOH) for 7 days (
Fig. 4(g); Fig. S11 in Appendix A); little effect on the CA and SA was found, and there was no obvious change in the coating surface before and after immersion (Fig. S12 in Appendix A), indicating that the coating exhibited excellent resistance against the different acidic and alkaline solutions.
3.3. Impact resistance
The impact resistance of the coating was tested via an erosion loop test based on the SY/T 7394–2017 standard. The micro morphologies of the bare 316L and the DEA-coated samples before and after erosion for 12 and 24 h are shown in
Figs. 5(a)–(c) and
(d)–(f), respectively. A large number of elongated scratches appeared on the surface of the samples, and the scratches became deeper and thicker as the impact time increased. The erosion process could be divided into four stages:
Stage I: The particles came into contact with the surface, and pressure was applied on the surface of the substrate.
Stage II: The particles slid and rolled on the substrate surface. The material underwent elastic and plastic deformation and erosion craters were formed. The craters elongated and deepened under the synergy of the sliding friction and rolling friction.
Stage III: The impact velocity of particles decreased to 0, and the substrate partially rebounded under the influence of elastic deformation, providing the particles with the initial velocity to leave the surface.
Stage IV: A height difference was created on the surface locations with erosion pits. According to the relative height, the lip (
Fig. 5(g), stage IV) was more likely to be impacted by the subsequent particles. Therefore, the scratches gradually elongated and deepened.
The erosion phenomenon is chiefly observed in regions where there are variations in the direction of fluid flow within the pipeline, specifically at bends. It is notable that the highest erosion rates are primarily concentrated around a particle impact angle of 22.5°
[46],
[47]. As shown in
Fig. 5(h), after the erosion, the fractured DE coating fell off in pieces, leading to a significant increase in the erosion rate, which fully demonstrates the necessity of repair. The erosion loop test demonstrated the buffering ability of the hydrogel layer, as the erosion rate of the DEA coating was 46.57% lower than that of the DE coating without the buffer structure. The erosion rate of the DEA coating was reduced by 57.6% and 68.1%, respectively, compared with those of the 316L and fractured DEA coating, confirming the superiority of the enamel structure and hydrogel repair strategy. Similar to the sand impact experiment, no peeling was found at the interface between the hydrogel and the coating, even under the simulated impact test (
Figs. 5(d) and
(e)). Furthermore, significantly less damage was generated on the coating surface after the erosion test for 24 h; thanks to the layered structure of the bionic enamel, the cushioning of the hydrogel layer, and the adhesion enhancement provided by the CNTs (
Fig. 5(f)), the energy generated by particles during impact was better dispersed, crack expansion was effectively suppressed, and the impact resistance of the coating was improved.
4. Conclusions
In summary, a bionic enamel coating with an amylose hydrogel layer was reported, along with its use to reduce erosion and corrosion in a gas transmission pipeline. Inspired by the layered structure of human tooth enamel, the coating is wear-resistant and can alleviate the surface damage caused by particle erosion. Based on the results from and analysis of an erosion loop test, the feasibility of the restored-crack strategy and the four-layered enamel structure with the layers having different viscoelasticities was determined. In addition, the mechanical properties of the coating were improved with the addition of EP, LA@TiO2, and CNTs. The corrosion potential of the coating was found to be maintained at −2.1 V after 24 h of erosion, thus demonstrating a significantly better erosion and corrosion resistance than those of other reported superhydrophobic coatings. The as-manufactured DEA coating also has other compelling attributes, including superhydrophobic stability after sanding, bending, scratching, and immersion in corrosive media. This study opens up a new route for a reliable strategy to solve the problems of hydrophobicity and wear resistance.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This study was supported by the National Natural Science Foundation of China (22375047, 51972063, and 22075046), the National Key Research and Development Program of China (2022YFB3804905, 2022YFB3804900, and 2019YFE0111200), the Natural Science Foundation for Distinguished Young Scholar of Fujian Province (2020J06038), the Natural Science Foundation of Fujian Province (2020J05098 and 2019J01256), 111 Project (D17005), and China Postdoctoral Science Foundation (2022M723497).
We thank L. Teng for critical reading and helpful comments of the manuscript, and National Engineering Research Center of Chemical Fertilizer Catalyst for use of the Hitachi S-4800 scanning electron microscope.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2024.03.024.
PDF
(2896KB)
