Photothermal superhydrophobic and corrosion resistant titania microstructured surface for ice prevention and removal Authors: Huamei Zhong, Zhifeng Hu, Chengjie Xiang, Chun Yang, Ruzhu Wang Published: 2026 ABSTRACT AbstractIce accretion severely compromises titanium alloy components in maritime environments. Conventional anti-icing coatings provide passive protection but often fail at the metal–coating interface under thermal and mechanical stresses, highlighting the need for in-situ engineered icephobic surfaces. Herein, we report an in-situ fabricated titanium surface integrating superhydrophobicity, photothermal conversion, and corrosion resistance through electrochemical treatment and vacuum annealing. The hierarchical micro–nano architecture with three-dimensional nanotubes establishes a robust superhydrophobic state. The reduced bandgap of titanium oxide enhances light absorption and thermal confinement, enabling rapid heating to 86 °C under 1 sun illumination. A stable oxide layer ensures excellent corrosion resistance after 120 h immersion in NaCl solution. Owing to these synergistic features, the surface demonstrates outstanding anti-icing and de-icing performance for pure and saline water, including outdoor conditions, achieving ice sliding within 235 s under 0.3 sun illumination. The surface retains superhydrophobicity after prolonged UV exposure, chemical corrosion, and mechanical abrasion, and exhibits anti-biofouling capability after 72 h immersion. This work provides a durable multifunctional strategy for mitigating ice formation on titanium components in complex environments.Subject terms: Nanoscale materials, Energy science and technology Introduction Ice formation is a ubiquitous natural phenomenon; however, it poses substantial challenges to maritime engineering1, including ocean vessels, oil rigs and offshore wind turbines, resulting in severe economic and operational consequences2. Titanium alloys (i.e., Ti-6Al-4V, also known as TC4) are widely employed as critical structural materials in marine applications3 owing to their exceptional strength-to-weight ratio, thermal stability, and fatigue resistance4. However, these safety-critical components frequently encounter severe icing challenges during low-temperature operations, particularly under high-humidity marine environments or altitude-induced supercooled water droplet conditions, necessitating advanced surface engineering solutions. Consequently, the development of effective anti-icing and de-icing strategies for titanium-based substrates is of significant scientific and engineering interest. Conventional de-icing techniques, including mechanical de-icing, hot air de-icing, chemical de-icing, and electrothermal de-icing, are often associated with inherent drawbacks such as low efficiency, incomplete ice removal, high energy consumption, and system complexity5. To mitigate ice formation at its origin, various passive anti-icing strategies have been explored, including superhydrophobic surfaces6, lubricating surfaces7, low interfacial toughness surfaces8, and so on. Among these, superhydrophobic surfaces have garnered considerable attention due to their capability to repel impinging droplets9 and induce spontaneous jumping of condensed droplets prior to freezing10. Despite their promising attributes in anti-icing performances, conventional superhydrophobic surfaces suffer from performance degradation under extreme environmental conditions, primarily due to diminished water repellency at low temperatures with high humidity and the deterioration of micro/nanostructures caused by repeated icing-deicing cycles11. To address these limitations, recent research has focused on the integration of photothermal functionality with superhydrophobic surfaces, thereby achieving a synergistic combination of passive anti-icing and active de-icing capabilities. Despite well-established fabrication of superhydrophobic titanium surfaces (e.g., laser ablation12, plasma etching13), in-situ construction strategies for photothermal superhydrophobic titanium-based substrates remain conspicuously underdeveloped. Current methodologies predominantly rely on coating techniques to impart photothermal functionality. Representative works include candle soot/silica/PDMS hybrid coatings demonstrating solar-activated deicing14, and TiN nanoparticle-modified architectures achieving subzero icephobicity through plasmonic heating15. Nevertheless, the practical application of these coatings remains constrained by their weak interfacial adhesion to rigid substrates, leading to coating delamination, detachment, and compromised long-term anti-icing performance16, while insufficient durability necessitates frequent maintenance of TC4 components. Additionally, TC4 exhibits inherent susceptibility to atmospheric corrosion and chloride-induced degradation17,18, particularly in marine environments. These combined challenges underscore the necessity for in-situ integration of photothermal superhydrophobicity with intrinsic corrosion resistance, providing a substrate-synchronized solution for extreme-condition anti-icing systems. To address these aforementioned challenges, we design an in-situ corrosion-resistant photothermal superhydrophobic TC4-based surface (CRPS-TC4) fabricated via two-step anodic oxidation, cathodic reduction and vacuum annealing. CRPS-TC4 maintains a superhydrophobic state, achieving a water contact angle (WCA) of 162.78° and a sliding angle (SA) of 3.25°. Also, the absorption of CRPS-TC4 increases to 90% because of the reduced bandgap of titanium oxide and the construction of micro-nano structures, resulting in the temperature increase to 86.44 °C from 25.39 °C under 1 sun illumination. Additionally, owing to the superhydrophobicity and stable titanium oxide structures, CRPS-TC4 exhibits exceptional corrosion resistance even after immersion in NaCl solution. The synergetic effects of hydrophobicity, photothermal conversion and anti-corrosion enable CRPS-TC4 to effectively prevent icing in both pure water and seawater while showing highly efficient de-icing capabilities. Moreover, the surface retains robust superhydrophobicity after prolonged UV light, chemical treatments, and mechanical abrasion, while also exhibiting excellent anti-biofouling performance after 72 h of immersion, demonstrating excellent durability and stability. Results Fabrication and characterizations of CRPS-TC4 As shown in Fig. 1, CRPS-TC4 was synthesized through a two-step electrochemical anodization process, followed by cathodic reduction, vacuum annealing and fluorination treatment. Microstructures on TC4 can well be adjusted by anodization using different electrolyte solutions. After the anodization in the electrolyte solutions containing NaOH and H₂O₂, the surfaces featured randomly oriented, irregularly shaped micrometer-sized pits (Supplementary Fig. 1). When NH₄F and ethylene glycol (EG) was used as the electrolyte solutions, nanotubes are uniformly formed on the substrate, exhibiting ordered, homogeneous, and vertically tubular structures (Supplementary Fig. 2)19. Therefore, to construct hierarchical micro-nano structures on TC4, the substrate was sequentially subjected to the two aforementioned anodization processes, followed by air annealing, yielding O-TC4. Scanning electronic microscope (SEM) of O-TC4 reveal that nanotubes with diameters of approximately 70 nm were successfully aligned along the normals of micro-pits, leading to the anisotropic three-dimensional nanotubes (Supplementary Fig. 3). Sequential anodization in the aforementioned electrolytes was followed by cathodic reduction and vacuum annealing to produce R-TC4. Finally, fluorination was applied to R-TC4 to achieve a superhydrophobic surface, forming CRPS-TC4. CRPS-TC4 exhibits a darker appearance (Supplementary Fig. 4), while the underlying micro/nanostructures remain well preserved. As shown in Fig. 2A–D and Supplementary Figs. 5 and 6, three-dimensional nanotubes are uniformly distributed within the irregular micro-pits, forming a hierarchical structure with an overall thickness of approximately 70 μm. Schematic of the manufacturing process of CRPS-TC4. The fabrication of CRPS-TC4 involves a sequential multi-step surface engineering strategy, including two-step anodic oxidation, cathodic reduction, vacuum annealing, and final fluorination treatment. The micro morphologies and superhydrophobicity of CRPS-TC4. A–C SEM images of CRPS-TC4 showing micro-nano structures. Red arrows denote different directions of nanotubes. D AFM image of CRPS-TC4 showing the surface roughness. Red arrows indicate the direction of growth for the nanotubes. E The WCA (162°) and SA (3°) of CRPS-TC4. F Schematic image of sealed and open air pockets. G Single water droplet adhesion test on CRPS-TC4 under the temperature of 25 °C and RH of 55%. H The coalescence-induced jumping of condensed droplets on CRPS-TC4 under the temperature of 5 °C and RH of 55%. Figure 2E shows that the WCA (Supplementary Table 1) of CRPS-TC4 reaches 162.78°, which is higher than that of the other fluorinated samples (Supplementary Figs. 7–9). This enhancement of WCA is attributed to the construction of hierarchical micro-nanostructures, enhancing surface roughness20 of 2.13 μm after two-step anodization process (Supplementary Fig. 10), as also evidenced by the atomic force microscope (AFM) images (Fig. 2D) showing the irregularly shaped micrometer-scale protrusions, combined with prominent nanoscale tubes of CRPS-TC4. Moreover, after fluorination, the hydrophobic PFOTS molecules deposit on the micro-nano structures confined the three-phase (gas–liquid–solid) contact lines, preventing water intrusion into the structures. Air is trapped within nanostructures (sealed air pockets) and between nanostructures (open air pockets) (Fig. 2F), thereby increasing liquid-air contact area21. The SA of CRPS-TC4 (Fig. 2E) is remarkably low at 3.25°, in contrast to the high SA of 45.75° for the fluorinated sample with homogeneous nanotubes (Supplementary Fig. 9C), indicating a strong adhesive force caused by the ordered TiO₂ nanotube arrays. Densely packed, ordered, and homogeneous nanotubes primarily trap air within their interiors (Fig. 2F), resulting in stronger negative pressure caused by volume changes in sealed air pockets22. For CRPS-TC4, the irregular shape and three-dimensional nanotubes enhance air movement and reduce the formation of sealed air pockets. SA of the sample with micro-pits is 7° (Supplementary Fig. 7C), while that of SO-TC4 (Supplementary Fig. 8C) with same hierarchical micro-nano structures is also reduced to 4°. This improvement of SA is attributed to the introduction of nanotubular structures with the second anodization. The high WCA and low SA suggest the excellent superhydrophobicity of CRPS-TC4. Resultantly, the droplet of 5 μL could not detach from the needle to wet CRPS-TC4 because of the extremely small adhesive force of CRPS-TC4 (Fig. 2G and Supplementary Movie 1) even at low temperatures like 5 °C (Supplementary Fig. 11) and −5 °C (Supplementary Fig. 12). Furthermore, the condensed droplets on CRPS-TC4 (Fig. 2H, Supplementary Movie 2 and Supplementary Fig. 13) and SO-TC4 (Supplementary Fig. 14, Supplementary Movie 3) tend to jump out through droplet coalescences and the impact droplets can be easily rebound from CRPS-TC4 (Supplementary Fig. 15). Photothermal performance of CRPS-TC4 TiO2 is a wide-bandgap semiconductor (bandgap 3.0 −3.2 eV), limiting its ability to absorb sunlight across the full spectral range23. While conventional photothermal coatings exhibit strong light-to-heat conversion, they often suffer from poor stability and challenges in large-scale fabrication16. To address these issues, an integrated fabrication approach is adopted, where TiO₂ nanotubes were first formed via two-step anodic oxidation and then subjected to cathodic reduction to narrow the bandgap. Followed by vacuum annealing, the stable black TixOy layer can be formed, enhancing both optical absorption and structural stability. As shown in Fig. 3A, the X-ray diffraction (XRD) patterns reveal that anatase is formed after annealing, as evidenced by the characteristic diffraction peaks at 2θ values of 25.2°, 37.8°, and 48.1°, corresponding to the (101), (004), and (105) planes, respectively (JCPDS card no. 21-1272).To further analyze the presence of Ti3+ and oxygen vacancies, X-ray photoelectron spectroscopy (XPS) was conducted to examine the surface changes in the Ti 2p and O 1 s regions (Fig. 3B, C and Supplementary Fig. 16). The Ti 2p spectrum of CRPS-TC4 reveals peaks corresponding to Ti4+ at 458.24 eV and 464.27 eV and Ti3+ at 457.6 eV and 463.29 eV, respectively, consistent with the Ti 2p3/2 and Ti 2p1/2 peaks (Fig. 3B)24. In contrast, the spectrum of SO-TC4 displays only Ti4+ at 458.61 eV and 464.29 eV (Supplementary Fig. 16A). As illustrated in Fig. 3C, the O 1 s spectrum of CRPS-TC4 exhibits a broader peak at 529.1 eV for Ti2O3 with an additional shoulder at 530.22 eV for TiO2 and an additional component at 531.71 eV attributed to non-lattice oxygen, which is notably different from SO-TC4 (Supplementary Fig. 16B)25. These shifts to lower binding energy suggest a modification in the bonding environment, indicating the formation of Ti3+ and oxygen vacancies in CRPS-TC426. The photothermal performances of CRPS-TC4. A XRD pattern of SO-TC4 and CRPS-TC4 showing the formation of anatase. XPS spectral of CRPS-TC4: Ti 2p (B) and O 1 s (C) showing the presence of Ti3+ and oxygen vacancies of CRPS-TC4. D VIS-IR absorption spectral of SO-TC4 and CRPS-TC4 showing stronger absorption across the 300–800 nm range of CRPS-TC4. E Band gaps based on Tauc plots of SO-TC4 and CRPS-TC4 showing the reduction in bandgap width of CRPS-TC4. F UV-VIS-IR absorptance of TC4, SO-TC4 and CRPS-TC4 showing the enhancement of solar absorptance of CRPS-TC4. Temperature evolution of CRPS-TC4, SO-TC4, and TC4 over three illumination on–off cycles under G 1 sun and H 0.5 sun illumination. I The temperature increases of CRPS-TC4 under different sun illuminations. To demonstrate the increasing absorption of CRPS-TC4, UV–vis reflectance spectra was employed to investigate its absorption and bandgap characteristics. As shown in Fig. 3D, SO-TC4 primarily absorbs ultraviolet light, while CRPS-TC4 exhibits strong absorption across the 300–800 nm range after electrochemical reduction. This enhancement is attributed to the presence of Ti3+ defects, which introduces a continuous vacancy band of electronic states just below the conduction band edge of CRPS-TC4 (Supplementary Fig. 17)27. The corresponding optical bandgaps (Fig. 3E) reveal a significant difference between the CRPS-TC4 (∼2.88 eV) and SO-TC4 (∼3.09 eV). The reduction in bandgap width following electrochemical treatment facilitates an extended photon response into the visible light region, thereby improving solar energy utilization28. Consistent with the change of bandgap width, Supplementary Fig. 4 shows that CRPS-TC4 appears blacker, whereas SO-TC4 exhibits a light gray coloration, further confirming the enhanced solar absorption. The photothermal performances of CRPS-TC4, SO-TC4, and TC4 under varying solar intensities were then investigated. As shown in Fig. 3F, CRPS-TC4 demonstrates absorptance ranging from 80% to 95% across the 280–2500 nm spectrum, where solar radiation is most intense. In comparison, SO-TC4 and TC4 exhibit absorptance in the ranges of 70%–95% and 30%–75%, respectively. The increased absorptance of SO-TC4 is primarily attributed to the micro-nanostructures, enhancing light trapping, while the additional ∼20% improvement (more than 600 nm) in CRPS-TC4 is due to the narrowed bandgap induced by cathodic reduction. Consequently, the photothermal performance of CRPS-TC4 is significantly enhanced. Under 1 sun illumination, the surface temperature of CRPS-TC4 rapidly increases to 86.44 °C, which is significantly higher than that of SO-TC4 (78.47 °C) and bare TC4 (51.24 °C) (Fig. 3G). Similarly, under 0.5 sun illumination, the surface temperatures of CRPS-TC4, SO-TC4, and TC4 rise to approximately 61.99 °C, 54.36 °C, and 37.56 °C, respectively (Fig. 3H). These results clearly demonstrate the critical role of solar intensity in governing the photothermal performance of the surfaces, as further corroborated by the illumination-dependent temperature evolution shown in Fig. 3I and Supplementary Fig. 18. Notably, even under weak illumination of 0.35 sun, CRPS-TC4 is still capable of reaching a temperature of 53.02 °C (Fig. 3I), highlighting its excellent photothermal efficiency under low-light conditions. Anti-corrosion effect of CRPS-TC4 To evaluate the potential marine applications of CRPS-TC4, the corrosion resistance was assessed by analyzing the anodic polarization curves, as shown in Fig. 4A. The corrosion current density (Icorr) and corrosion potential (Ecorr) are determined using Tafel analysis, and the results are summarized in Supplementary Table 2. Notably, the corrosion current density of CRPS-TC4 (0.62 μA·cm-2) is nearly 1 order of magnitude lower than that of TC4 (7.12 μA·cm-2). Even after immersed in NaCl solution for 24 h, Icorr of CRPS-TC4 increases slightly but remains substantially lower than that of TC4. Moreover, both the initial and 24-hour-immersed CRPS-TC4 samples exhibits higher Ecorr compared to TC4, further corroborating the enhanced anti-corrosion performance. The corrosion inhibition efficiency can be calculated: η=icorr0−icorricorr0×100%29, where icorr0 and icorr are the current densities of TC4 and CRPS-TC4, respectively. The corrosion inhibition efficiency of CRPS-TC4 almost remains at 91% even after 24 h of immersion, demonstrating the excellent durability of CRPS-TC4. The anti-corrosion performances of CRPS-TC4. A Dynamic polarization curves of CRPS-TC4 and TC4. B Nyquist plots of CRPS-TC4 and TC4; the inset highlights the smaller capacitive loop radius of TC4 and TC4 after 24 h immersion. Bode plots of CRPS-TC4 and TC4: C impedance magnitude and D phase angle. E Dynamic polarization curves of CRPS-TC4 after different immersion periods. F Nyquist plots of CRPS-TC4 after different immersion durations (the inset shows the equivalent circuit model used for fitting the EIS data of CRPS-TC4). The data points in B–D and F correspond to the experimental results, while the solid lines are the corresponding fitted curves. Electrochemical impedance spectroscopy (EIS) was further performed to elucidate the corrosion resistance of CRPS-TC4 (Fig. 4B–D). Figure 4B reveals that the impedance arc radius associated with charge transfer30 during the corrosion process on CRPS-TC4 is significantly larger than that of TC4, even after 24 h immersion, suggesting improved corrosion resistance of CRPS-TC4 compared to TC4. This is also supported by Fig. 4C, which shows that the magnitude of impedance |Z| at 0.01 Hz for both CRPS-TC4 and TC4 are 4.25×106 Ω·cm2 and 9840 Ω·cm², respectively. After 24 h of immersion, the |Z| at 0.01 Hz for CRPS-TC4 almost remains nearly 3 orders magnitude higher than that of TC4, demonstrating its excellent anti-corrosion stability. The phase angle is close to −90° for CRPS-TC4 in the mid-to-high frequency region (Fig. 4D), showing the characteristics of capacitor behavior31. Figure 4F depicts the equivalent circuit models used to fit the impedance spectra of CRPS-TC4 and untreated TC432, with the corresponding fitting parameters summarized in Supplementary Table 3. Rs is the solution resistance, Rct is the charge transfer resistance, Rf is the corrosion resistance induced by air cushion and micro-nano structures, CPEct is the constant phase element of the double layer and CPEf is the constant phase element of the air cushion and micro-nano structures33. The fitting results indicate that the Rct value of CRPS-TC4 is significantly higher than that of TC4, forming a stronger barrier at the solid-liquid interface and effectively impeding the migration and diffusion of corrosive ions, thereby substantially reducing the likelihood of localized corrosion. Morphological analysis (Supplementary Fig. 19) further reveals that CRPS-TC4 remained structurally intact after 24 h of immersion, whereas TC4 exhibited deepened cracks and the emergence of corrosion pits. Although the native oxide film on TC4 provides a certain degree of corrosion protection, inherently thin and porous structures allow corrosive liquids to penetrate into the surface34. When TC4 is immersed in a 3.5 wt% NaCl aqueous solution, its superhydrophilic nature facilitates rapid wetting by the electrolyte, forming a continuous liquid film. The presence of chloride (Cl⁻) ions accelerates migration and diffusion along the liquid film, leading to preferential corrosion at weak points such as grain boundaries. In contrast, CRPS-TC4 demonstrates significantly enhanced corrosion resistance due to two key factors: (1) the superhydrophobic surface, which prevents electrolyte adhesion and minimizes liquid-solid contact through the air cushion effect, and (2) the stable anatase oxide layer, serving as a protective barrier. The above findings highlight CRPS-TC4 as a promising candidate for corrosion-resistant marine applications. As shown in Fig. 4E, F, potentiodynamic polarization and EIS measurements were conducted after 0, 24 h, 72 h, and 120 h of immersion in 3.5 wt% NaCl solution. Overall, the corrosion resistance of CRPS-TC4 exhibits a non-monotonic evolution with immersion time, characterized by a slight degradation at the initial stage followed by a gradual recovery upon prolonged immersion. Specifically, the minor decrease in corrosion resistance observed during the first 24 h can be attributed to initial electrolyte penetration into the surface micro/nanostructures, which temporarily increases the electrochemically active area and facilitates interfacial charge transfer, as evidenced by an increase in corrosion current density and a reduction in the impedance radius35. With prolonged immersion from 24 h to 120 h, the formation and accumulation of corrosion products and hydrated oxide layers partially block electrochemically active sites36, leading to a gradual recovery of polarization resistance and impedance from 24 h to 120 h. Accordingly, the corrosion current density further decreases from 24 h to 72 h and 120 h, while the corresponding EIS results show a pronounced increase in both the impedance radius and the low-frequency impedance modulus (Supplementary Fig. 20), demonstrating the sustained corrosion-protection capability of CRPS-TC4 over extended immersion37. Anti-icing and de-icing performances of pure water To evaluate the anti-icing performance, the icing delay time of CRPS-TC4, SO-TC4, and TC4 was examined under ambient conditions of −23 °C and relative humidity (RH) of 55%. The process of droplet freezing is generally well understood and can be divided into three stages: nucleation, ice crystal growth, and freezing38. The time at which the droplet becomes opaque is considered the nucleation time, while the time when the ice crystal growth reaches its peak is considered the freezing time of pure water. Comparative analysis reveals that CRPS-TC4 exhibits a significant delay in ice nucleation initiation (358 s, Fig. 5A) compared to TC4 (26 s, Fig. 5C). Similarly, ice crystal growth time demonstrates substantial prolongation from 38 s on TC4 to 192 s on CRPS-TC4. The significant delay in icing on CRPS-TC4 stems from synergistic thermal and interfacial engineering. The total heat transfer resistance is Rtotal=Rt+Ra⋅RnRa+Rn+Rw (Fig. 5E), where Rt, Ra, Rn and Rw are thermal resistances of the titanium substrate, air, nanotubues and water droplet, respectively. Rtotal is systematical enhanced through: (1) the hierarchical titania nanotube architecture with lower thermal conductivity than the pure TC4 substrate39, (2) the gas-phase insulation provided by micro-nano confined air cushions, and (3) minimized liquid-solid contact due to PFOTS monolayers. Additionally, the reinforced surface roughness elevated nucleation energy barrier, thus prolonging the nucleation time. During icing, the air trapped in the micro- and nanoscale structures is transferred to the droplet, floating and accumulating at the top and on the outer surface of the icing droplet40, as evidenced by the formation of air bubbles during the icing process on CRPS-TC4 (Fig. 5A). Due to their similar structures, SO-TC4 also exhibits an icing delay time of approximately 476 s (Supplementary Fig. 21A). Ambient humidity significantly influences droplet freezing time by affecting vapor condensation, frost-bridge formation, and heterogeneous ice nucleation. As shown in Supplementary Fig. 22, reducing the RH to 15% increased the freezing time of the bare TC4 substrate to 486 s, suggesting that lower humidity can moderate icing. Besides, CRPS-TC4 demonstrated remarkable anti-icing performance, with the icing delay time extending to 3682 s, showcasing its excellent anti-icing capability. The freezing delay ratio, defined as the ratio of the freezing time of the designed material to that of the bare substrate, is introduced to evaluate and compare the freezing delay efficiency of recent superhydrophobic materials across different experimental setups. Figure 5H reveals that CRPS-TC4 exhibits a superior freezing delay compared to other reported materials (Supplementary Table 4), resulting from hierarchical micro-nano structures, three-dimensional nanotubes and superhydrophobic PFOTS monolayers. The anti-icing and de-icing performances of pure water on CRPS-TC4. A The anti-icing process of 30 μL-pure water on CRPS-TC4. B The photothermal de-icing process of 30 μL-pure water on CRPS-TC4 under 0.5 sun illumination. C The anti-icing process of 30 μL-pure water on TC4. D The photothermal de-icing process of 30 μL-pure water on TC4 under 0.5 sun illumination. All experiments are conducted under ambient conditions of -23 °C and RH of 55%. Scale bars in A–D 1 mm. Red lines and cycles in A–C represent the solid-liquid interface and the air bubble. E Schematic diagram of anti-icing process and the thermal resistance of CRPS-TC4, Rt, Ra, Rn and Rw are thermal resistances of Ti, air, three-dimensional nanotubes and water, respectively. Active de-icing process of 30 μL frozen water droplets on F CRPS-TC4 and G TC4 with the inclination angle of 15° at -10 °C under 0.1 sun. Scale bars in F, G 1 mm. H The comparison of anti-icing performances across various experimental setups. The data of numbers 1-19 are extracted from refs. 44–62. I The comparison of melting time of photothermal de-icing. The data of numbers 1-11 are extracted from: 1-363–65, 4-548,49, 6-766,67, 857, 9-1168–70. ★ in H, I represents the experimental results of our work. The photothermal de-icing performances of CRPS-TC4 and TC4 were systematically investigated under 0.5 sun illumination (Fig. 5B, D). Notably, the icing droplet on CRPS-TC4 exhibits complete melting within 130 s, while SO-TC4 requires 150 s to melt ice (Supplementary Fig. 21B). In stark contrast, TC4 fails to initiate melting and accumulates frost instead. This performance highlights the exceptional photothermal conversion efficiency of CRPS-TC4, accelerating de-icing effects. As illustrated in Fig. 5I (Supplementary Table 5), while some surfaces used in previous studies exhibited a deicing efficiency within 150 s, CRPS-TC4 demonstrates enhanced photothermal deicing capability even under lower illumination (only 0.5 sun). The superior photothermal performance arises not only from the enhanced three-dimensional microstructures but also from the reduced bandgap of titanium oxide. To characterize the photothermal superhydrophobic effect, de-icing tests were conducted with droplets at a tilt angle of 15° under 0.1 sun illumination. As shown in Fig. 5F, Supplementary Figs. 23 and 24, frozen droplets slid off the CRPS-TC4 surface after only 75 s and 158 s at −5 °C and −10 °C, respectively. In addition, SO-TC4 also enabled droplet removal, with sliding times of 136 s and 214 s under the same conditions. In stark contrast, droplets on the bare TC4 surface remained frozen and continued to accumulate frost (Fig. 5G and Supplementary Fig. 24). These results highlight the high photothermal efficiency of CRPS-TC4 combined with its robust superhydrophobicity, enabling effective droplet removal even under weak illumination and sub-zero temperatures. The anti-frosting performance of CRPS-TC4 was also evaluated at −23 °C under 0.35 sun illumination. CRPS-TC4 exhibits a prolonged condensation process, with no condensed droplets larger than 10 μm until 360 s (Supplementary Fig. 25A and Supplementary Movie 4). Droplets on hydrophilic TC4 quickly condense and merge, forming continuous liquid film as early as 60 s (Supplementary Fig. 25B and Supplementary Movie 5). The accumulation of liquid film accelerates the frost formation, with freezing initiating at 278 s and the entire surface covered by frost at 480 s. In contrast, on CRPS-TC4, droplet remain distinct without forming liquid film even after 4200 s, showing a frost-free state. This is attributed to the excellent superhydrophobic and photothermal effects of CRPS-TC4, which facilitate condensate droplet bouncing (Fig. 2H) and elevate the surface temperature, respectively. For SO-TC4 with relatively low solar-to-heat conversion, freezing initiates from around 1200 s (Supplementary Fig. 26 and Supplementary Movie 6). However, clearly segmented frozen areas around individual droplets are observed, rather than forming a frost layer, which is similar to cascade freezing of supercooled droplets41. This behavior arises from the combined influences of surface superhydrophobicity and the size effect of nanotubes. With a ~ 70 nm diameter blind-hole morphology, the nanotubes restrict water vapor penetration due to their comparable size to the mean free path of water molecules (around 85 nm)42. Additionally, their hydrophobic nature further inhibits ice and frost formation within the nanotubes. Resultantly, CRPS-TC4 demonstrates excellent long-term ice resistance. Even after 20 freezing-thawing cycles, CRPS-TC4 retains its superhydrophobic properties (Supplementary Fig. 27), with no significant change in the anti-icing performance (Supplementary Fig. 28). Overall, CRPS-TC4 can effectively delay icing in the absence of illumination and prevent ice formation or promote ice melting under sunlight. Durability of CRPS-TC4 in practical environments In practical environments, anti-icing surfaces are often exposed to saline contamination, subzero temperatures, and weak solar illumination. To evaluate the applicability of CRPS-TC4 under such conditions, the freezing behavior of saline droplets was investigated, with the nucleation time defined as the moment when the droplet became opaque. On CRPS-TC4, the nucleation time reached 780 s (Supplementary Fig. 29A), representing a delay of 732 s compared with bare TC4 (Supplementary Fig. 29B). A similarly prolonged nucleation time was observed on SO-TC4 (778 s, Supplementary Fig. 29C). Temperature evolution analysis (Supplementary Fig. 30) shows that saline droplets nucleate at lower temperatures than pure water, leading to extended nucleation times on these surfaces. In contrast, droplets on TC4 exhibited significant frost growth after 600 s, indicating inferior anti-icing performance. Under 0.5 sun illumination, CRPS-TC4 and SO-TC4 effectively suppressed saline droplet freezing (Fig. 6A and Supplementary Fig. 31), whereas rapid nucleation occurred on TC4 within 73 s due to insufficient thermal compensation (Fig. 6B). The enhanced anti-icing behavior originates from the photothermal effect of CRPS-TC4, which increases the heat flux transferred from the substrate to the droplet (Qc) beyond the heat loss to the ambient environment (Qa) (Fig. 6C). Durability performance of CRPS-TC4 in real-world and multifactor environments. The photothermal anti-icing process of 30 μL-saline water on CRPS-TC4 (A) and TC4 (B) under 0.5 sun illumination. All experiments are conducted under ambient conditions of −23 °C and RH of 55%. Scale bars in A, B: 1 mm. C Schematic diagram of photothermal anti-icing process and the heat transfer of a droplet on CRPS-TC4, Qc and Qa are the thermal energy transferred from the substrate to the droplet and thermal energy lost from the droplet to the ambient air, respectively. D WCAs and SAs of the surface over 30 days of continuous sunlight UV radiation, measured at 5-day intervals. E WCAs and SAs of the surface after 12 h of immersion in solutions with different pH values (pH=3, 5, 9, 11), deionized water (DI), alcohol, and seawater. F Surface coverage percentage of Chlorella adhesion on CRPS-TC4 and TC4 after distinct immersion durations. For D–F, scatter points represent individual measurements. Solid lines (D, E) and bar heights (F) indicate the mean values, with error bars denoting the standard deviation (SD) from four independent measurements for (D, E) and three for (F). The anti-icing process of 100 μL water on CRPS-TC4 (G) and TC4 (H) under ambient conditions of -11 °C and RH of 57% conducted in Lanzhou. Scale bars in G, H 0.5 cm. The photothermal de-icing process of ice block on CRPS-TC4 (I) and TC4 (J) under ambient conditions of -7 °C and RH of 50% conducted in Lanzhou. TC4 and CRPS-TC4 samples were placed on a sloped roof with an inclination angle of 15° of house model in all outdoor experiments (G–J). Scale bars in I, J 1 cm. Consistently, photothermal de-icing tests (Supplementary Fig. 32) show that saline ice on CRPS-TC4 begins melting within 50 s and fully transitions to liquid by 90 s under 0.5 sun irradiation, while no apparent ice recession is observed on bare TC4. Durability evaluations further confirm the practical robustness of CRPS-TC4. The surface retains superhydrophobicity after 20 freezing–thawing cycles (WCA ≈ 161°, SA ≈ 6°, Supplementary Fig. 33) and maintains stable photothermal anti-icing performance with ice-free durations exceeding 30 min (Supplementary Fig. 34). Collectively, these results demonstrate that CRPS-TC4 offers reliable anti-icing and de-icing performance under realistic conditions involving salinity, low temperatures, and limited solar irradiation, highlighting its strong potential for practical applications. To evaluate the durability of CRPS-TC4, we tested its UV resistance, chemical robustness, and mechanical strength. As shown in Fig. 6D, the sample was exposed outdoors in Singapore for 30 days, during which WCA and RA were measured every five days. CRPS-TC4 maintained stable values with only minor fluctuations, indicating that the monolayer remains intact under prolonged UV exposure. This is consistent with literature showing PFOTS’ high UV stability due to its strong Si–O–Ti bonding35,43. Additionally, after 12 h of immersion in solutions with pH values of 3, 5, 9, and 11, as well as deionized water, ethanol, and natural seawater, all samples retained excellent superhydrophobicity (Fig. 6E) with no significant degradation. After 24 h of immersion in natural seawater, no obvious changes in surface appearance were observed, indicating the coating structure remained intact (Supplementary Fig. 35). To systematically evaluate the mechanical robustness of CRPS-TC4, water jetting, sand impact, and sandpaper abrasion tests were conducted. As shown in Supplementary Fig. 36, CRPS-TC4 exhibited excellent abrasion resistance. Even after 2000 mL of water jetting, 1000 g of sand impact, and 60 abrasion cycles, the surface maintained robust superhydrophobicity, with WCAs consistently exceeding 150° and SAs remaining below 10°, indicating strong resistance to mechanical wear. For anti-biofouling performance, CRPS-TC4 showed significantly lower algae adhesion than the bare TC4 surface. After 24 h, the surface coverage on CRPS-TC4 was ~1%, compared to ~5% on TC4 (Fig. 6F and Supplementary Fig. 37). After 72 h, CRPS-TC4 maintained ~2% coverage, while TC4’s coverage increased to ~11%, demonstrating superior fouling resistance. These results demonstrate that CRPS-TC4 possesses excellent chemical, UV, and mechanical durability, highlighting its potential for practical applications in harsh environments where both chemical exposure and mechanical abrasion are unavoidable. To further evaluate the practical applicability of the developed CRPS-TC4 surface, outdoor experiments were conducted under natural environmental conditions, where complex and fluctuating factors such as ambient temperature, solar irradiation, and humidity are inevitably present. The winter outdoor performance of CRPS-TC4 was systematically investigated in Lanzhou, Gansu, P.R. China (Supplementary Fig. 38). Under static conditions without sunlight at −11 °C and RH of 57%, CRPS-TC4 exhibited excellent passive anti-icing performance. The freezing time of a 100 μL water droplet was only 238 s on bare TC4 (Fig. 6H), whereas it was significantly prolonged to 819 s on CRPS-TC4 (Fig. 6G). Moreover, the ice block formed on CRPS-TC4 began to slide after 50 s, reached the bottom edge of the sample at 141 s, and completely detached from the surface at 235 s (Fig. 6I) under 0.3 sun and -7 °C. In contrast, no observable sliding behavior was detected on the bare TC4 surface even after 10 min (Fig. 6J). Corresponding infrared thermography revealed that the edge temperature of CRPS-TC4 increased from −7.8 °C to −0.8 °C, while the overall surface temperature reached 2.3 °C at the moment of ice detachment (Supplementary Fig. 39A). By comparison, the surface temperature of TC4 remained nearly unchanged at approximately −9.5 °C throughout the entire test (Supplementary Fig. 39B). These results demonstrate that CRPS-TC4 retains effective de-icing capability under real outdoor conditions with only 0.3 sun solar illumination. Furthermore, even under extremely weak illumination (0.1 sun) at 3 °C, CRPS-TC4 preserved its superhydrophobicity and remained capable of repelling water droplets (Supplementary Fig. 39C), highlighting its robustness and practical applicability in cold and low-light environments. Discussion This study presents a multifunctional surface fabricated on a titanium-based substrate via in-situ surface modification, integrating anti-corrosion, photothermal conversion, and superhydrophobic properties for advanced anti-icing applications. The engineered surface features hierarchical micro/nano architectures and three-dimensional nanotube arrays, constructed through a two-step anodic oxidation process followed by fluorination. This architecture endows CRPS-TC4 with excellent superhydrophobicity, characterized by a WCA of 162° and a SA of 3°. Owing to the hierarchical micro/nano structures and the reduced bandgap of titanium oxide induced by electrochemical treatment and vacuum annealing, CRPS-TC4 exhibits enhanced solar absorption and achieves a surface temperature increase of up to 61 °C under 1 sun illumination. Meanwhile, the synergy between superhydrophobicity and a stable TiₓOy layer provides CRPS-TC4 with robust corrosion resistance, enabling high corrosion inhibition efficiency after 120 h of immersion in NaCl solution. Furthermore, CRPS-TC4 exhibits outstanding mechanical durability and chemical stability. The surface maintains superhydrophobicity after mechanical wear (including water jetting, sand impact and sandpaper abrasion) and prolonged chemical exposure under both acidic and alkaline conditions, indicating strong resistance to mechanical and chemical degradation. Benefiting from above combined functionalities, CRPS-TC4 shows significantly enhanced anti-icing performance, with an icing delay time nearly ten times longer than that of the bare TC4 substrate. The superhydrophobic surface enables efficient melting of both pure and brine ice within 130 s under 0.5 sun illumination, while rapid ice sliding can be achieved within 158 s even under weak illumination of 0.1 sun. Beyond laboratory-scale evaluations, the practical performance of CRPS-TC4 is further validated through outdoor experiments under natural winter conditions. The ice block on CRPS-TC4 initiates sliding after 50 s and completely detaches from the surface within 235 s under 0.3 sun illumination at −7 °C. These results confirm the robustness and reliability of CRPS-TC4 under realistic environmental conditions. Overall, the design of CRPS-TC4 provides valuable insights into the development of robust, titanium-based anti-icing and de-icing surfaces for practical maritime applications. Methods Sample preparation The TC4 substrate (3 × 3 × 0.5 cm³) was first ultrasonically cleaned in detergent, acetone, ethanol, and deionized water for 10 minutes each to remove any contaminants. After drying at 80 °C, the cleaned TC4 was anodized in an electrochemical cell, where a 304 stainless steel electrode served as the counter electrode, and the TC4 acted as the working electrode. To create micro-nano structures on the TC4 substrate, two different electrochemical solutions were used. The first aqueous solution contained 1.5 mol L-1 NaOH (98%, Aladdin) and 1 vol% H₂O₂ (30%, Sinopharm Chemical Reagents), with anodization performed at 30 V for 120 minutes. The second aqueous solution consisted of ethylene glycol (EG, anhydrous, 99.8%, Sigma-Aldrich) with 0.3 wt% ammonium fluoride (NH₄F, 98%, Sigma-Aldrich) and 2 vol% deionized water, where anodization was conducted at 60 V for 120 minutes. Following the two-step anodic oxidation, the sample was rinsed with deionized water and dried at 80 °C. O-TC4 was then obtained by annealing the sample in air at 450 °C for 120 minutes, with a heating rate of 5 °C min⁻¹. Sequential anodization in both electrolytes was followed by a reduction step, where the sample was subjected to a constant potential of 100 V for 1 minute in the same EG and NH₄F solution used during the second anodization step. To fabricate R-TC4, the sample was subsequently annealed under vacuum at 450 °C for 120 minutes, with the same heating rate of 5 °C min⁻¹. To achieve superhydrophobicity, samples at different stages, including after the first and second anodization steps, as well as O-TC4 and R-TC4, were modified by immersion in n-hexane (99%, Adamas) containing 5 vol% Triethoxy-1H,1H,2H,2H-perfluorooctylsilane (PFTOS, 98%, Adamas) for 30 minutes. This was followed by drying at 110 °C for 30 minutes, yielding SO-TC4 and CRPS-TC4, respectively. Sample characterization The surface and cross-sectional morphologies of the samples were examined using correlative Raman imaging and scanning electron microscopy (RISE, TESCAN, Czech Republic). Surface roughness was measured with an atomic force microscope (MPF-3D, Oxford Instruments, the United Kingdom). WCA and SA measurements were conducted using a contact angle goniometer (DSA 100, KRÜSS GmbH, Germany). Droplets of 10 μL were placed at four different locations on the testing surface to measure the WCA and SA at room temperature. The phase composition of the samples was identified using an XRD (Aeris XRD, Malvern Panalytical, Netherlands) with Cu Kα radiation (λ = 1.54178 Å). XPS was performed using a Theta probe (ESCALAB QXi, Thermo Fisher Scientific, the United States) with 300 W Al Kα X-ray (hν = 1486.6 eV). UV-vis diffuse reflectance absorption spectra were recorded using a UV-visible spectrophotometer (Lamda 950, PerkinElmer, the United States). Additionally, reflectance spectra across the wavelength range of 0.28-2.5 μm were collected using a UV-visible-NIR spectrometer (Lambda 950, PerkinElmer, the United States). The absorptance of the samples was calculated by subtracting the reflectance from 100%, given the absence of transmittance (samples thicker than 1 mm). For single droplet adhesion test, a single water droplet with a volume of 5 μL was gently deposited onto the sample surface at different temperatures using a microsyringe. Due to the stronger adhesion force between the droplet and the needle tip compared with that between the droplet and the sample surface, the droplet followed the needle tip upon withdrawal and could not remain on the surface. For photothermal experiments, the samples were suspended in the center of a plastic box, ensuring that their top surfaces were exposed to the air. To minimize temperature fluctuations caused by air convection, the plastic box was sealed with saran wrap. The samples were illuminated by a solar illuminator (UHE-NS−100, Sciencetech Instruments, Canada) to simulate varying sunlight conditions. Thermocouples were placed on the top of the samples to monitor the temperature changes. For electrochemical measurements, they were carried out through an electrochemical workstation (VSP-3e, Huayang Keyi, China) with 3.5 wt% NaCl aqueous solution as the corrosive electrolyte. Three-electrode testing system was employed during the electrochemical tests. The platinum plate, Ag/AgCl electrode, and samples were used as the counter electrode, reference electrode, and working electrode, respectively. The test system was stabilized for 1 hour before initiating testing. The dynamic polarization tests were conducted at a scan rate of 1 mV/s relative to the open circuit potential (OCP) across a potential range of −0.5 V to +0.5 V and the electrochemical impedance spectroscopy (EIS) was carried out within a frequency range of 105 Hz to 10−2 Hz. Anti-frosting/anti-icing/de-icing experiments The room temperature was maintained at 25 ± 1 °C, with a RH of 55 ± 5%. Liquid ethanol, kept at -35 °C, was circulated in the inner chamber to regulate the experimental temperature. The temperature at the bottom of the samples was measured to be -23 °C. After recording anti-icing processes, the icing droplets on the samples were exposed to 0.5 sun illumination to measure photothermal de-icing capacities. The icing and de-icing processes were recorded from the side using a high-speed camera (M220M-64G, Hefei Zhongke Junda Vision Technology Co., Ltd, China). The frosting process was captured using an optical microscope. A 30 μL water droplet was deposited onto a surface maintained at −30 °C and allowed to freeze for 10 min. The surface was then tilted to an inclination angle of 15°, after which the temperature was increased to −10 °C or −5 °C. Simultaneously, the surface was irradiated under 0.1 sun illumination, and the ice melting and sliding behavior of the droplet was recorded. Durability tests For UV radiation from sunlight testing, the surface was exposed to natural outdoor sunlight in Singapore for 30 days (from 3 November 2025 to 3 December 2025). Every five days, the WCAs and SAs of 5 μL were measured at four randomly selected locations on the surface. For chemical stability tests, the surface was immersed for 1 h in different solutions with pH values of 3, 5, 9, and 11 (adjusted using hydrochloric acid and sodium hydroxide), as well as in deionized water and ethanol, to evaluate its acid and alkali resistance. The samples were immersed under laboratory-controlled conditions in natural seawater collected from the East Coast of Singapore (1.30°N, 103.93°E). After immersion, WCAs and SAs of 5 μL were measured at four randomly selected locations on the surface. For anti-biofouling test, chlorella was selected as a representative model microorganism to investigate the anti-biofouling adhesion performance of CRPS-TC4. The Chlorella cells were inoculated into a modified ammonium mineral salts medium and were cultured at 25 °C. The medium contained NH₄Cl (8 g L⁻¹), MgSO₄·7H₂O (0.5 g L⁻¹), Na₂HPO₄·12H₂O (10.7 g L⁻¹), KH₂PO₄ (0.27 g L⁻¹), CaCl₂·6H₂O (0.2 g L⁻¹), ferric sodium EDTA (0.005 g L⁻¹), and a trace element stock solution (1 mL L⁻¹; composed of Na₂EDTA·H₂O, FeSO₄·7H₂O, H₃BO₃, ZnSO₄·7H₂O, MnCl₂·4H₂O, CoCl₂·6H₂O, CuSO₄·5H₂O, NiCl₂·6H₂O, and Na₂MoO₄·2H₂O). The samples were then immersed in the cultured Chlorella solution and incubated for 24 h and 72 h in a constant-temperature incubator maintained at 25 °C. After incubation, the samples were removed and gently rinsed three times with DI water to eliminate weakly attached cells. The adhesion of Chlorella on the sample surfaces was subsequently observed using an optical microscope. The anti-biofouling performance was quantitatively evaluated by calculating the surface coverage ratio of adhered Chlorella using ImageJ software. For mechanical durability test, the sample was mounted on a tilted plate with an inclination angle of 30°, and water droplets or sand particles were allowed to impact the surface from a height of 50 cm, as schematically illustrated in Supplementary Fig. 36D, E After exposure to different volumes of water or varying masses of sand, WCAs and SAs of 5 μL were measured to monitor the evolution of surface wettability. For the anti-abrasion test, the sample was placed face-down on 2000-grit SiC sandpaper under a constant load of 50 g and subjected to horizontal reciprocating abrasion (shown in Supplementary Fig. 36F). An abrasion distance of 40 cm, consisting of 20 cm forward and 20 cm backward motion, was defined as 1 abrasion cycle, WCAs and SAs of 5 μL were measured to monitor the evolution of surface wettability. Outdoor experiments For passive anti-icing test, in an outdoor environment without sunlight (-11 ± 1 °C, RH of 57 ± 5%), TC4 and CRPS-TC4 samples were placed on a sloped roof model with an inclination angle of 15°. The freezing process of a 100 μL water droplet on both bare TC4 and CRPS-TC4 surfaces was recorded using a CCD camera to analyze the droplet freezing behavior. For photothermal de-icing experiments, an ice block measuring 2.5 cm in length, 2.5 cm in width, and 3 cm in height was pre-frozen. The ice block was then placed on TC4 and CRPS-TC4 surfaces and allowed to adhere under outdoor, sunless conditions (-7 ± 2 °C, RH of 50 ± 5%, and 0.3 sun) for 10 minutes. Subsequently, the samples with the adhered ice were positioned on the same inclined roof model (15° slope). The melting process of the ice was recorded to evaluate the de-icing performance of each surface. Supplementary information Supplementary Information Description of Additional Supplementary File Supplementary Movie 1: Supplementary Movie 2: Supplementary Movie 3: Supplementary Movie 4: Supplementary Movie 5: Supplementary Movie 6: Transparent Peer Review file Source data Source data Acknowledgements The authors acknowledge the financial support from the National Natural Science Foundation of China (Grant No. 52106101, C.X.), the China Postdoctoral Science Foundation (Grant Nos. 2022T150402, C.X., 2021M702100, C.X, and 2023M742230, Z.Hu.), the Fundamental Research Funds for the Central Universities, and the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 51521004, R.W.). Author contributions R.W. and C.Y. supervised and led this research. H.Z., C.X., and Z.Hu. designed the experiments and processed the data. H.Z. performed the experiments. H.Z. and Z.Hu. prepared the original manuscript. C.X., C.Y., and R.W. revised the manuscript and optimized the experimental procedures. All authors discussed the experiments and results and approved the final manuscript. Peer review Peer review information Nature Communications thanks Yongmei Zheng, Lingwei Ma and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available. Data availability The data supporting the findings of the study are included in the main text and Supplementary Information files Source data are provided with this paper. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Huamei Zhong, Zhifeng Hu. Contributor Information Chun Yang, Email: mcyang@ntu.edu.sg. Ruzhu Wang, Email: rzwang@sjtu.edu.cn. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-71606-6. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary File Supplementary Movie 1: Supplementary Movie 2: Supplementary Movie 3: Supplementary Movie 4: Supplementary Movie 5: Supplementary Movie 6: Transparent Peer Review file Source data Data Availability Statement The data supporting the findings of the study are included in the main text and Supplementary Information files Source data are provided with this paper.