Fabrication of tunable Large-Area Moth-Eye nanostructures for antireflection and hydrophobicity via anodic aluminum oxide templates and nanoimprint lithography

Abstract

In this study, a comprehensive process–structure–function roadmap was established for bio-inspired functional surfaces. We systematically controlled the pore diameter, interpore distance, pore depth, and pore aspect ratio of anodic aluminum oxide (AAO) master templates via multistep anodization. These templates were then replicated in durable nickel–cobalt alloy working molds through electroforming, and their nanostructures were transferred to polycarbonate films using nanoimprint lithography. Our findings highlighted the critical influence of pre-anodization, electrolyte type (oxalic acid for an ~ 100 nm interpore distance; phosphoric acid for ~ 400 nm), anodization potential, and time on the AAO structures. We also identified 100 A/m² as the optimal current density for achieving high-aspect-ratio structures under intense anodization. The polymer film replicas obtained using these precisely controlled templates showed significantly enhanced functional properties: the average surface reflectance decreased from 10.85% to a minimum of 3.5%, transmittance increased from 80.1 to 92.3%, and water contact angles improved from 91.46 to 138.69°. Thus, a higher structural aspect ratio is crucial for enhanced hydrophobic performance, consistent with the Cassie–Baxter model. In summary, this research provides an efficient, controllable method for manufacturing high-performance bio-inspired functional surfaces and, more critically, establishes direct correlations between anodization parameters and the resulting optical and wetting properties, offering key guidance for material design.

Supplementary Information

The online version contains supplementary material available at 10.1186/s11671-025-04373-w.

Introduction

Nanotechnology leverages materials’ unique, size-dependent nanoscale properties to drive innovation across diverse fields, including clean energy [1, 2], electronics [3], and biomimetics [4]. Unlike resolution-limited top-down methods, bottom-up self-assembly offers scalable, low-cost, uniform large-area patterning through the construction of complex structures from fundamental units [5]. Aluminum anodization, a bottom-up self-assembly technique, is used to create highly ordered nanoporous anodic aluminum oxide (AAO) structures through a dynamic electrochemical equilibrium. AAO nanofabrication templates are versatile due to their straightforward, cost-effective preparation and precise control over structural parameters, such as pore size and aspect ratio. Their tunable structure enables a wide range of applications, such as high-sensitivity chemicals and biosensors [6–12], hydrogen generation catalyst support [13], filtration membranes [14], photonic crystals and plasmonic devices [15, 16], surface-enhanced Raman spectroscopy [17], drug delivery [18], energy storage [19], biomedical scaffolds [20], and seawater desalination [21]. This study primarily focuses on using AAO templates with tunable pore dimensions and aspect ratios to fabricate polymeric nanostructures.

AAO templates have a distinct bilayer structure: a dense barrier layer between the aluminum substrate and a porous layer containing high-density, typically ordered hexagonal nanopores. Their large specific surface area benefits sensing and catalysis, whereas their abundant pore–wall hydroxyl groups facilitate surface functionalization [22]. Key AAO dimensions, namely, the pore diameter, interpore distance, pore depth, and template thickness, are precisely controlled by the anodization parameters, including the electrolyte type/concentration (e.g., sulfuric, oxalic, and phosphoric acids), applied voltage, current density, time, and temperature [5]. Electrolyte differences influence pore morphology, an increase in voltage typically enlarges the pore diameter and interpore distance, and the anodization time determines the pore depth. Two-step anodization, developed by Masuda et al. [23], guides pore growth through preformed dimple patterns, benefiting highly ordered arrangements. The pore diameter can be adjusted further through post-anodization chemical etching. Beyond standard cylindrical pores, noncylindrical AAO pores can be generated using various techniques, such as the use of wavy channels from dynamic voltage modulation [24], controlled electrochemical anodization using cell cultures [25], dual-diameter pores from altered electrolytes [26], and anisotropic elliptical nanopores via physical stretching [18]. High-frequency excitation can also reduce the brittleness of AAO structures [27], and their optical properties and corrosion resistance can be modified by incorporating different materials [17]. Theoretical frameworks have been developed to explain and predict AAO’s self-organization and arrangement [28, 29].

Nanoimprint lithography (NIL), proposed by Chou et al. [30], is a simple, cost-effective, high-throughput nanopatterning method that involves mechanically pressing a mold into a resist. Primary NIL techniques include thermal and UV NIL; in thermal NIL, thermoplastic polymers are hot-pressed above their glass transition temperatures () [30, 31], whereas UV NIL involves curing photosensitive resists through UV exposure, enabling room-temperature operation, low pressures, and rapid filling [32]. However, AAO substrates are inherently fragile and generally unsuitable for direct, high-repeatability nanoimprinting. Therefore, electroforming is often used to replicate AAO nanostructures in more robust nickel–cobalt (Ni–Co) metal molds for high-volume manufacturing.

AAO templates are promising tools for fabricating bio-inspired functional surfaces, especially antireflective (e.g., moth-eye structures) and superhydrophobic (e.g., lotus leaf surfaces) applications, due to their tunable nanostructures, material compatibility, and versatility. Nanoimprinting, or transfer methods, replicate these intricate subwavelength or hierarchical micro/nanostructures on polymer surfaces to achieve specific functions, such as high optical transmittance and self-cleaning properties. Research on superhydrophobic surfaces includes studies on the PDMS modification of anodic aluminum [33], enhancing hydrophobicity by replicating vertically aligned porous nanostructures via thermal NIL [34], improving hydrophobicity and friction using cobalt-based metal organic frameworks synthesized on anodic aluminum [35], and using hierarchical micro/nanoporous AAO stamps to create superhydrophobic polycarbonate (PC) and polymethyl methacrylate films (water contact angles [WCAs] of 153° and 151°, respectively) [36]. As for antireflection applications, Lim et al. used AAO templates to create embossed and engraved polymer moth-eye nanostructures for antireflective coatings [37], Pashchanka and Cherkashinin provided insights into light absorption in additive-free black porous anodic alumina [38], Yanagishita et al. produced anodic porous alumina with tapered pores on curved surfaces to expand the applications of moth-eye nanostructures [39], Lan et al. proposed methods for aluminum oxide optical thin films with varying refractive indices [40], and Zhang et al. studied boundary slip in nanoscale flows to optimize the antireflective performance of subwavelength moth-eye nanostructures [41].

Challenges persist despite the significant progress in fabricating functional AAO nanostructures for bio-inspired antireflective and superhydrophobic surfaces. These include precisely controlling pore sizes and morphologies, overcoming material and process limitations (e.g., aluminum purity, large-scale production, uniform deposition within high-aspect-ratio pores, and interfacial bonding), and fully understanding structure–function correlations. Comprehensive systematic data on the influence of anodization parameters (e.g., voltage, electrolyte type/concentration, time, and temperature) on key AAO nanopore dimensions (diameter, depth, and aspect ratio), which are useful for optimizing specific biomimetic functions (antireflection and superhydrophobicity), are lacking. This deficiency limits the precise design and reproducible manufacturing of AAO structures for specific applications. Furthermore, the effect of the anodization time of the initial layer (porous) on the thickening kinetics and properties of the subsequent layer (barrier) needs clarification.

This study addresses these gaps by systematically investigating the influence of anodization parameters on the nanopore dimensions and aspect ratios of AAO templates. AAO templates with diverse structural characteristics are prepared by controlling the key process parameters (multistep anodization is used with a cooling system to manage heat during high-voltage anodization). Their morphology is characterized via scanning electron microscopy (SEM). Subsequently, these nanostructures are replicated on polymer surfaces via nanoimprinting (e.g., hot embossing or UV-cured imprinting). Finally, by evaluating the antireflective (optical transmittance/reflectance) and superhydrophobic (contact angle) properties of the replicated polymer films, we aim to establish clear correlations between the anodization parameters, AAO template structure, and final biomimetic functional properties. This research provides experimental evidence and theoretical guidance for the efficient, controllable fabrication of high-performance bio-inspired polymeric nanostructures. The study covers the selection of specific aluminum materials, systematic modification of key anodization parameters, characterization of AAO template morphologies, exploration of polymer nanostructure replication, and functional evaluation of the final samples.

Materials and methods

Chemicals and reagents

Anodization experiments were conducted using three types of aluminum sheets differing in size and purity, with dimensions chosen to reflect distinct potential applications. Sample #A was a 2 cm × 2 cm × 0.127 mm annealed high-purity aluminum foil (99.9995%), with its geometry representing that of biomedical testing chips. Sample #B was a 14.5 cm × 10 cm × 0.5 mm annealed and single-side polished high-purity aluminum foil (99.9%), sized for potential use in mobile phone screens. Both samples #A and #B were purchased from Strem Chemicals (Taiwan). Sample #C was a large 30 cm × 22 cm × 0.6 mm AL1070 alloy substrate (99.70%) with single-side polishing, representing a scale suitable for automotive panels and displays, and was acquired from Fu-Lu Trading Co., Ltd. (Taiwan). The selection of these materials and their geometries was based on a systematic evaluation of process feasibility, from an ideal benchmark to an industrially viable substrate. Initially, the high-purity aluminum foil (Sample #A) was chosen as a benchmark to validate our anodization protocol under optimal conditions. Subsequently, we investigated the significantly more cost-effective AL1070 alloy (Sample #C) to assess the process's practical applicability and scalability. As our findings revealed that the performance difference between the high-purity and lower-purity substrates was minimal, the AL1070 alloy was incorporated as a viable material for a cost-effective process. Sample #B, an intermediate-purity material, was included to provide a broader comparison. This approach allowed us to assess our anodization process not only across a spectrum of material purities but also on substrates with geometries relevant to distinct, scalable industrial applications. All chemicals for the anodization were obtained from SIGMA-ALDRICH (MA, USA), including 99.5% oxalic acid dihydrate ((COOH)₂•2H₂O), 99.0% crystallized phosphoric acid (H₃PO₃), 95% anhydrous ethyl alcohol (C₂H₅OH), 70% perchloric acid (HClO₄), 99% chromium(VI) oxide (CrO₃), 85% phosphoric acid (H₃PO₄), and acetone.

Instrumentation

Figure 1 illustrates the experimental setup used to fabricate AAO in this study. The core system (Fig. 1b) was a temperature-controlled DC-powered electrochemical cell, whose temperature was precisely maintained by a circulating chiller (HCS-820, HC Scientific & Instrument Co., Taiwan). The aluminum foil workpiece was secured in a custom-made, multipart fixture to ensure stable, uniform reaction conditions (Fig. 1c). The internal copper plate and sealing structure were designed to define the reaction area precisely and uniformize the current and electric field distributions. The aluminum substrates were subjected to chemical etching pretreatment in a heated, stirred solution (Fig. 1a) to obtain clean, uniform starting surfaces before anodization.

Fig. 1.

Experimental setup for AAO template preparation: a pre-anodization etching system; b anodization system, including DC power supply and HCS-820 circulating chiller; c detailed view of custom clamping fixture used for aluminum workpiece

The circulating chiller in the core system (HCS-820) had a temperature range from − 20 °C to 100 °C. Due to the large size of the AAO samples, an external cooling bath was integrated into the experimental system to cool the aluminum sheets and fixtures effectively, ensuring optimal low-temperature anodization conditions for improved pore ordering and reduced dielectric breakdown (Fig. 1b). A smaller cooling circulator (LB-20, Strider Tech Co.,Ltd., Taiwan) was also used when applicable.

A custom-designed polytetrafluoroethylene (PTFE) jig (Liako Technology Corp., Taiwan) was used to secure the aluminum samples and precisely define the reaction area during anodization (Fig. 1c). PTFE was selected for its chemical inertness to strong acids, bases, and organic solvents and its wide operating temperature range (− 196 °C–250 °C). The jig’s design stabilized the fixation of the aluminum sheet, preventing agitation-induced distance variations between the anode and cathode. A waterproof O-ring and a specialized, waterproof square rubber plate were used to prevent electrolyte leakage to the aluminum backside, ensuring a uniform electric field distribution on the active surface. The jig comprised an outer shell, an internal sample slot, and a locking back cover.

A power supply (GPR-20H15HD, GW Instek) with an output voltage range of 0–200 V and current range of 0–15 A was used for constant-voltage anodization experiments. The cutoff voltage was preset, and the current was controlled to achieve the desired experimental voltage. A stirring hot plate (Super NUOVA series, Bestgen Biotech Corp., Taiwan) was used for pre-anodization etching (Fig. 1a). This instrument provided controlled heating for the etching solutions and ensured a uniform concentration distribution via magnetic stirring. A mechanical stirrer (BLG-2D, Jiuh Hsing Instrument Co., Ltd., Taiwan) operating at 0–1500 rpm was used for general stirring during anodization to dissipate heat and maintain electrolyte homogeneity.

A vacuum hot press (GTP1501, Genstek Automation Co., Ltd., Taiwan) was utilized for hot embossing. Polymer films of uniform thickness were pressed into molds, heated above their softening point, subjected to mechanical or vacuum pressure, and then cooled to solidify patterned products. A customized UV curing system (UV 500-A, Great Lighting Co., Ltd., Taiwan) with a maximum UV intensity of 80–100 mW/cm² and a peak wavelength of 365 nm was used for UV nanoimprinting. An AAO master mold was prepared, and a UV adhesive was evenly applied and cured using this high-intensity system for complete solidification.

Fabrication

Nanostructured replicas were fabricated via four stages: aluminum substrate pretreatment, anodization, electroforming mold preparation, and nanoimprint transfer (Fig. 2). In this comprehensive nanofabrication pathway, AAO was used as an intermediate template to create functional polymer surfaces with specific nanostructures.

Fig. 2.

Schematic illustration of multistage fabrication of tunable large-area moth-eye nanostructures: preprocessing, re-anodization, electroforming, and molding

Preprocessing

Initially, highly ordered dimple arrays were formed on the aluminum substrates through pretreatment steps, including electropolishing and etching. Electropolishing ensured surface flatness for uniform nanopore formation. The aluminum sheets were cleaned with acetone, ethanol, and deionized (DI) water and then immersed in a 1:3.5 (v/v) perchloric acid–anhydrous ethanol polishing solution stirred at 650 rpm for 2 min. Samples were then extensively rinsed with DI water and subsequently soaked for 25 min to ensure the complete removal of any residual perchloric acid from the electropolishing solution, which could otherwise interfere with the uniformity of the subsequent anodization process. The sheets were then dried with nitrogen gas. Pre-anodization was conducted at 1 °C using oxalic acid or phosphoric acid to form dense nanoporous structures with diameters of ~ 100 or ~ 400 nm, respectively. Prior to applying voltage, the aluminum sheets were immersed in the 1 °C electrolyte solution for 30 min while being stirred. This pre-immersion step was crucial for allowing the system to reach a stable state, ensuring both thermal equilibrium between the substrate and the electrolyte and a uniform chemical concentration distribution at the interface. This procedure is essential for achieving a stable and uniform initial current distribution, which is critical for the homogenous nucleation of ordered nanopores. The samples were then pre-anodized according to the parameters in Tables 1 and 2 for 1–10 h. After anodization, the samples were rinsed with DI water and dried, and the anodic surfaces appeared light yellow. The oxide layer on the backside of the AAO template, at the oxide–aluminum interface, was removed through chromic acid etching. A chromic acid solution in an acrylic holder was heated to 55 °C and stirred at 600 rpm for 6 h. The samples were then rinsed with DI water, immersed for 5–10 min, and dried.

Table 1.

Interpore distances for AAO molds after various pre-anodization processes

No	Acid Type	Voltage (V)	Interpore Distance (nm)	Standard Deviation (%)
A1	0.3 M oxalic	50	131.6	5.09
A2	0.3 M oxalic	80	240.3	15.7
A3	0.3 M oxalic	100	248.8	12.81
A4	0.3 M oxalic	110	254.2	4.03
A5	0.3 M oxalic	120	265.5	3.95
A6	0.3 M oxalic	140	282.5	4.07
B1	0.12 M phosphoric	100	240.2	13.63
B2	0.12 M phosphoric	120	251.5	10.24
B3	0.12 M phosphoric	140	383.4	2.56
B4	0.12 M phosphoric	175	412.7	1.75
C1	0.3 M oxalic + 0.12 M phosphoric (mixture ratio 5:1)	50	250.6	6.03
C2	0.3 M oxalic + 0.12 M phosphoric (mixture ratio 2:1)	50	258.3	7.24
C3	0.3 M oxalic + 0.12 M phosphoric (mixture ratio 1:1)	50	260.8	5.18
C4	0.3 M oxalic + 0.12 M phosphoric (mixture ratio 1:2)	50	265.1	11.24
C5	0.3 M oxalic + 0.12 M phosphoric (mixture ratio 1:5)	50	269.4	12.95
D1	0.03 M oxalic	120	281.1	13.66
D2	0.3 M oxalic	120	262.2	3.82
D3	3 M oxalic	120	141.4	10.68

Table 2.

AAO molds with different sizes for different pre-anodization, re-anodization, and pore-widening parameters

No	Size (cm × cm × cm)	Pre-Anodization			Re-Anodization
		Acid Type	Voltage (V)	Time (h)	Acid Type	Voltage (V)	Re-Anodization (s)/Pore Widening (m)	Cycle Repetitions
E1	2 × 2 × 0.0127	0.3 M oxalic	110	1	0.12 M phosphoric	140	15/30	16
E2	2 × 2 × 0.0127	0.3 M oxalic	110	1	0.12 M phosphoric	140	30/30	16
F1	14.5 × 10 × 0.05	0.3 M oxalic	50	3	N/A
F2	14.5 × 10 × 0.05	0.3 M oxalic	50	3	0.3 M oxalic	50	10/14	5
F3	14.5 × 10 × 0.05	0.12 M phosphoric	170	8	N/A
F4	14.5 × 10 × 0.05	0.12 M phosphoric	170	8	0.12 M phosphoric	170	10/14	20
G1	30 × 22 × 0.06	0.12 M phosphoric	140	10	0.12 M phosphoric	140	20/20	10
G2	30 × 22 × 0.06	0.12 M phosphoric	150	10	0.12 M phosphoric	150	20/20	10
G3	30 × 22 × 0.06	0.12 M phosphoric	160	10	0.12 M phosphoric	160	20/20	10

Note: All re-anodization processes were performed under a constant applied voltage. The specific voltage for each experiment was selected to achieve an approximate target current density (~ 10 mA/cm² for E and F-series; ~ 2.27 mA/cm² for G-series to ensure comparable growth conditions

Re-anodization and electroforming

Except for samples A and E, which underwent only pre-anodization, all samples underwent 5–20 cycles of re-anodization. After pre-anodization and etching, the aluminum substrates were re-anodized in potentiostatic (constant voltage) mode. A primary objective of our experimental design was to achieve a specific target current density to ensure comparable film growth conditions across different samples. Based on the relationships established in our preliminary study, we selected the appropriate constant applied voltage for each experiment (detailed in Table 2) that was expected to yield a result close to our target current density. The process was conducted at a constant 1 °C with magnetic stirring at 550 rpm for 10–30 s. The anodized sheets were then pore-widened in 0.43 M phosphoric acid at room temperature, which was stirred at 500 rpm for 14, 20, or 30 min. This cycle was repeated 5–20 times, followed by rinsing with DI water and drying. Electroforming was conducted to convert the fragile AAO templates into robust, durable Ni–Co alloy metal molds suitable for repetitive imprinting. This process was performed by Hui Fang Precision Forming Technology Co., Ltd. (Taiwan). The electroforming was carried out in galvanostatic (constant current) mode at an operating temperature of 45 °C. The electrolyte bath contained cobalt sulfite (CoSO₃), with high-purity nickel beads serving as the dissolving anode. A constant current density in the range of 0.7–1.3 A/dm² was applied to the AAO master mold, which functioned as the cathode. A high-hardness Ni–Co alloy with a specified Rockwell C hardness of 50 HRC was chosen over pure nickel (typically ~ 30 HRC) for its superior mechanical properties. While the AAO master template is itself a very hard material (typically 2.9–5 GPa on the Vickers scale [42]), it is inherently brittle and unsuitable for production-level imprinting. The electroformed Ni-Co mold, in contrast, provides an ideal combination of high surface hardness and metallic toughness, making it a robust and durable tool for high-fidelity pattern transfer. The AAO master mold served as the cathode for Ni–Co deposition. Once sufficiently thick, the resulting metal mold was peeled from the AAO master mold, yielding a high-fidelity inverse Ni–Co structure.

Figure 2 illustrates distinct electroforming paths. In frontside electroforming (frontside EP), the Ni–Co alloy is deposited from the opening side of the AAO template, completely filling its pores. Upon demolding, this process yields a negative Ni–Co mold with a nanopillar array. Conversely, backside electroforming (backside EP) is performed from the backside of the AAO template after the barrier layer is removed. Demolding results in a positive Ni–Co mold, which features a nanopore array that mirrors the original AAO template’s surface structure.

Nanoimprint lithography

Nanostructured patterns were transferred to polymers via hot embossing and UV imprinting. For hot embossing, polycarbonate (PC) films were used. PC film and the Ni-Co mold were clamped in a vacuum hot press, heated to 190 °C for 25 min (300 kg pressure for the first 10 min). After molding, the system was cooled, and the mold separated from the PC film. Resulting hot-embossed replicas exhibited smaller pore diameters and depths than the original master mold due to surface tension limitations. For UV imprinting, UV adhesives (Models: SB8820, SB8927-25 K, SB9278-HV, Sunwheel Materials) were used. The process involved: (1) Demolding pre-treatment: A fluorinated mold release agent was sprayed on the template and dried. (2) Resin coating: UV adhesive was applied to the master mold, covered with a glass substrate, and pressed with 5 kg for 10 min. (3) Curing: The UV-coated template was exposed to 80–100 mW/cm² (365 nm) UV light for 5 min after a 10-min preheat. (4) Demolding: After curing, the template was removed, the glass substrate detached, and the UV adhesive separated. The template was then cleaned (1 min acetone soak, ethanol, DI water rinse, nitrogen dry). UV-imprinted replicas often showed slightly larger pore diameters and depths than the original master mold due to external tensile stress during demolding, which can cause minor fracturing or elongation of surface structures.

Characterization of anodized aluminum molds and nanoimprinted replicas

Surface morphology

We characterized the surface morphologies of the AAO molds and nanoimprinted replicas using field-emission SEM (UltraPlus series, Zeiss, Germany) and focused ion beam (FIB) microscopy (JIB-4601F). The software ImageJ (National Institutes of Health, USA) was used for a quantitative analysis of the AAO nanopore structural parameters from the SEM images. This process involved scale calibration, image binarization (setting a grayscale threshold to isolate pores), and identifying and labeling individual pores using the particle analysis function. This yielded equivalent diameter and centroid coordinates for each pore. All data were then exported to Microsoft Excel for statistical analysis.

We calculated the average pore diameter and its standard deviation directly from the measured pore diameters. The average interpore distance was determined by averaging the center-to-center distances from each pore to its six nearest neighbors and then averaging these local interpore distances across all pores. The standard deviation of the interpore distance served as a key metric for evaluating the regularity of the hexagonal array, with a target of less than 5% for well-ordered structures. Finally, wall thickness and porosity were calculated as

1

where D_int is the interpore distance, D_p is the pore diameter, and t_w is the wall thickness.

2

where P is porosity.

Reflectance and hydrophobicity

SEM analysis indicated that the hot-embossed PC replicas exhibited more uniform structures and higher replication rates than the UV-imprinted replicas. Consequently, the former replicas were chosen for the optical testing of the moth-eye nanostructures. We measured reflectance and transmittance using UV–vis spectrophotometry (U3900-H, Hitachi, Japan) with tungsten and deuterium lamp sources, characterizing samples at wavelengths of 400–700 nm. We also assessed the degradation of the average reflectance after 1 and 7 days of exposure to natural light. The FIB images revealed that the UV-imprinted replicas possessed rougher surface structures and thus suited hydrophobic surface characterization. Thus, these replicas were used to measure WCAs using an imaging contact angle goniometer (1000B, First Ten Angstroms, USA). The WCA was defined and calculated as.

3

where θ is the surface contact angle of the material without nanostructures, θ_CB is the effective contact angle predicted using the Cassie–Baxter model for the nanostructured surfaces, and f( −) is the ratio of the solid–liquid interfacial contact area to the projected area. For nanostructured surfaces, f( −) is usually less than 1, and an increase in surface roughness typically enhances hydrophobicity. In cases where water fully penetrates the nanostructure pores, the Cassie–Baxter model is superseded by the Wenzel model. Research on moth-eye nanostructures for hydrophobic applications is also relevant [43, 44]. In general, a water contact angle (WCA) of below 10° signifies superhydrophilicity, while a WCA in the range of 10–90° indicates hydrophilicity. Surfaces with a WCA above 90° are considered hydrophobic. Superhydrophobicity represents a special case, typically defined by a WCA exceeding 150° along with a low sliding angle.

Results and discussion

Pre-anodization results

We fabricated highly ordered, symmetric hexagonal pore arrays with high aspect ratios through pre-anodization (for initial ordering) and then multistage re-anodization. The electrolyte type during pre-anodization significantly affected the final interpore distance. To optimize pre-anodization, we systematically adjusted the oxalic and phosphoric acid concentrations and applied voltage (Table 1). The rationale for this systematic approach was as follows:Groups A and B were designed to establish the fundamental relationship between anodization voltage and the resulting interpore distance and regularity for two standard, single-component electrolytes: oxalic acid and phosphoric acid, respectively. These experiments aimed to identify optimal voltage windows that yield highly ordered pore arrays (i.e., low standard deviation) in each system.Group C explored the potential for fine-tuning the interpore distance by using a mixed-acid electrolyte at a fixed voltage. By varying the mixture ratio of oxalic and phosphoric acids, we investigated the synergistic effects on the self-ordering process.Group D was conducted to isolate the influence of electrolyte concentration. We selected an optimal voltage (120 V) identified from the Group A experiments and varied the oxalic acid concentration to determine its specific impact on pore morphology and regularity.This structured investigation provided the foundational data necessary for selecting the optimized conditions used in the subsequent stages of our study. The pore morphology was characterized via SEM (Fig. 3), whereas ImageJ was used to calculate the standard deviation of the interpore distance to assess ordering, with the target being below 5%. It should be noted that while some experimental parameters (e.g., A4-A6, C1-C5, and D2) resulted in structures with very thin pore walls or high disorder, leading to a less-defined appearance in top-down images, cross-sectional and elemental analyses confirmed that an aluminum oxide film was formed in all cases. These results are crucial for defining the boundaries of the stable processing window. This iterative process was implemented to identify optimized AAO interpore distance parameter sets. Typically, oxalic acid electrolytes produce mildly anodized pores with interpore distances of 90–140 nm, such as in sample A1 (131.6 nm; Table 1) and sample F1 (129 nm; Table 3). We conducted hard anodization (HA) to achieve precise pore diameters (200–300 nm) and high aspect ratios. Unlike mild anodization, HA requires high voltages and current densities. The strong electric field in HA rapidly forms aluminum oxide, which dissolves in the electrolyte, resulting in high-aspect-ratio structures. This rapid growth incorporates more electrolyte anions, creating a denser, typically gray–black oxide layer. In this study, a 100 A/m² current density yielded the most stable and uniform AAO structures at high voltages. However, as the current density approached 150 A/m², the risk of localized dielectric breakdown significantly increased due to charge concentration at surface irregularities. Furthermore, HA generates substantial Joule heat, necessitating low operation temperatures (below 0 °C) and the use of an enhanced cooling system to prevent melting, sintering, or electrolyte degradation and to control oxide growth. Moreover, improper temperature control or substrate nonuniformity can cause localized dielectric breakdown and pore collapse, which is a key challenge of this technique.

Fig. 3.

SEM images of AAO templates after various pre-anodization processes

Table 3.

Interpore distances, pore diameters, pore depths, pore aspect ratios, pore densities, wall thicknesses, and porosities of AAO molds with different sizes

No	Structure	Interpore Distance (nm)	Pore Diameter (nm)	Standard Deviation (%)	Pore Depth (nm)	Pore Aspect Ratio	Pore Density (μm−2)	Wall Thickness (nm)	Wall Thickness (nm)	Porosity (%)
E1	Columnar	332	248	1.7	423	1.7	31.4	84	42	67.7
E2	Columnar	364	340	0.7	1900	5.6	62.4	24	12	84.7
F1	Hemispherical	129	121	3.3	24	0.2	39.1	8	4	85.1
F2	Columnar	123	75	3.1	130	1.7	82.6	42	24	55.3
F3	Hemispherical	465	461	4.7	115	0.25	32.7	4	2	89.9
F4	Columnar	318	272	0.7	413	1.5	27.6	46	23	77.6
G1	Columnar	235	173	3.5	493	2.85	38.5	62	31	66.8
G2	Columnar	329	197	1.8	733.7	3.7	34.3	132	66	54.3
G3	Columnar	386	222	2.5	2452	11	27.53	164	82	52.2

For samples A1–A6 (Table 1), 1 h anodization with 0.3 M oxalic acid consistently showed a positive correlation between the pore diameter and applied voltage. Higher voltages (110, 120, and 140 V) resulted in significantly lower standard deviations of the pore arrangement, indicating improved uniformity and consistency. This is likely because such high voltages promote rapid oxide layer growth, forming a dense, low-impurity barrier layer; this layer increases resistivity and the critical voltage, thereby stabilizing the anodization current and promoting the formation of regular structures [45, 46]. Conversely, below 110 V, variations in barrier layer thickness and oxygen evolution caused severe voltage/current fluctuations, reducing pore uniformity. The relationship formula of the ionic current, voltage, and barrier layer thickness was consistent with the literature [47]. Using these optimized parameters, we consistently fabricated AAO structures with interpore distances of 250–285 nm in oxalic acid electrolytes at 110–140 V, achieving the target structural standard deviation (within 5%).

This trend extended to the 1.2 M phosphoric acid electrolyte system (samples B1–B4; Table 1). At 160–195 V, phosphoric acid electrolytes generally produced AAO interpore distances of 405–500 nm, such as in sample B4 (412.7 nm; Table 1) and sample F3 (465 nm; Table 3). We successfully controlled the interpore distance by leveraging its positive correlation with the anodization voltage, observing improved pore uniformity at above 140 V. Interestingly, the oxide growth rate with phosphoric acid was lower than that with oxalic acid under similar conditions.

Finally, we explored mixed electrolytes (samples C1–C5; Table 1). In the case of 3 h anodization at 0 °C and 50 V using 2 L of electrolytes prepared by mixing 0.3 M oxalic acid and 1.2 M phosphoric acid in various ratios, simply adjusting the ratio was insufficient for effective interpore distance control (200–300 nm, standard deviation > 5%). Separate 1 h anodization experiments with varying oxalic acid concentrations at 0 °C and 120 V revealed that decreasing the concentration increased both the interpore distance and pore diameter (samples D1–D3; Table 1). This is because oxidation is prioritized over dissolution at higher concentrations, leading to smaller pores. However, only at the 0.3 M concentration was the standard deviation maintained within 5%; deviations significantly reduced pore regularity. Collectively, these results define the operational window, limits, and material dependence of HA, offering critical guidance for the stable production of large-diameter, high-uniformity AAO.

It is important to contextualize this discussion of pore regularity. The long-range order of single-step, long-term anodization process is related to film thickness, anodization time or voltage. However, our methodology is based on a multistage re-anodization technique. In this approach, the ordered pore arrangement created in a preceding step acts as a template to guide the nucleation and growth during subsequent anodization. Consequently, the stability of the initial pore pattern, governed by the electrochemical conditions in Table 1, becomes more critical to the final order than achieving great film thickness in a single run. While the film thickness for the screening experiments in Table 1 was not individually measured, it is estimated to be in the 50–500 nm range, and the final pore depths for optimized structures are reported in Table 3. Our discussion, therefore, focuses on identifying the parameters that yield a stable and regular initial structure, which is the necessary foundation for our multistage fabrication process.

Re-anodization results

During re-anodization, aluminum oxide growth was consistently constrained by the hemispherical dimples formed during pre-anodization and etching, which altered the expected interpore distance. Through this constraint, pre-anodization and etching significantly improved the pore arrangement ordering. Pre-anodization notably changed the interpore distance, which was primarily determined by the pit spacing from the initial oxalic acid anodization; this result aligned with Masuda et al.’s findings on nanoimprint-guided pore spacing [23]. Although pre-anodization added complexity and slightly influenced the standard deviation of the interpore distance, the standard deviation remained within our target of 5%. For instance, when an AAO template was pre-anodized using the parameters of sample A5 (Table 1) and then re-anodized using the parameters of samples B3 and B4, the resulting interpore distances were 267.8 and 345.5 nm, with corresponding standard deviations of 3.5–4.7%, respectively. Furthermore, longer pre-anodization times enhanced the regularity of the nanoporous structures after re-anodization. This was attributed to the gradual regularization of the dimple array at the metal–oxide interface, although this improvement had a saturation point.

The combined re-anodization and phosphoric acid pore widening successfully transformed the etched hemispherical dimples into conical pore structures. Building on Masuda et al.’s work on two-step anodization [23], the current work shows that repeated anodization is a common strategy to improve pore regularity. Beyond the cycle number, the duration of each anodization step is crucial for structural morphology. Hence, we systematically investigated the influence of the anodization and pore-widening times on the pore aspect ratio using 0.12 M phosphoric acid at 140 V (Figs. 4 and 5). Anodization for 25 s yielded a pore depth of 615 nm, which was approximately 2.18 times that obtained from 5 s treatment, confirming the considerable effect of anodization time on pore depth. Longer anodization also led to better pore arrangement ordering and gradual pore-widening, highlighting the complex interplay of process parameters. By contrast, the pore-widening time had a more limited effect on pore diameter. Based on these observations, multiple AAO process parameter sets (Table 2) were designed for comparative analysis.

Fig. 4.

Cross-sectional SEM images of AAO molds showing effect of re-anodization time: a 5 s, b 10 s, c 15 s, d 20 s, and e 25 s. Top-down views of AAO molds showing effect of pore-widening time: f 30 min, g 90 min, h 120 min, i 180 min, j 210 min, and k 240 min

Fig. 5.

Experimental results for re-anodization time, AAO pore depth, and AAO pore geometry: (a) relationship between re-anodization time and pore depth and (b) relationship between pore-widening time and pore diameter

Initial SEM observations and structural analysis of the AAO master molds (Fig. 6 and Table 3) showed that the AAO pore structures typically formed close-packed hexagonal arrays. However, higher anodization potentials, albeit increasing pore diameters, reduced the regularity of this hexagonal arrangement. This irregularity stemmed from heat generation on the electrolyte surface at high anodization voltages. Such irregularity can be mitigated by limiting the current and enhancing the cooling system to dissipate heat, thereby improving the pore arrangement and physical properties of the porous alumina.

Fig. 6.

SEM images of AAO molds: (E1, E2, F1, F2, F3, F4, G1, G2, and G3) showing various pore morphologies; (H) SEM image of an AAO mold exhibiting multiple smaller subpores

The critical potential needed for stable porous AAO growth depends on the electrolyte’s chemical stability or the acidic electrolyte’s dissolution capability, which is controlled by the acid strength, acid concentration, and solution temperature. The use of oxalic acid (e.g., samples F1 and F2; Tables 2 and 3) achieved the optimal nanopore arrangement at 50 V, yielding hexagonal structures with cup–conical or hemispherical pores (~ 100 nm interpore distance, depending on the number of repeated re-anodization cycles). However, the dimples formed by pre-anodization and etching caused a nonuniform electric field distribution during re-anodization, making high current densities challenging; thus, higher voltages are not recommended for re-anodization in this context. Re-anodization experiments at 120 V led to a sudden voltage drop to 70 V, resulting in complex, porous AAO films with multiple smaller subpores at the bottom of main pores (sample H; Fig. 6). This hierarchical, dendritic subpore structure deviates from that of conventional single-diameter channels. Its formation was primarily attributed to the drastic decrease in the anodization voltage during re-anodization, which halted the growth of large pores. The electric field then redistributed from the larger, hemispherical bottoms, spontaneously concentrating in new, smaller regions. These areas became highly reactive sites, leading to the nucleation and growth of new, smaller pores via preferential aluminum oxide dissolution.

The use of phosphoric acid (e.g., samples F3, F4, and G1–G3; Tables 2 and 3) resulted in cup–conical or hemispherical pores (~ 400 nm interpore distance, depending on the number of repeated re-anodization cycles). Through multiple anodization and pore-widening steps, the pore depths exceeded 2450 nm. This operation succeeded because of the use of thick, high-purity aluminum sheets and rigorously consistent control of all parameters in each repeated step, keeping the local temperature below 1 °C. As for the mixed (oxalic and phosphoric acid) electrolytes (e.g., samples E1 and E2; Tables 2 and 3), cup–conical pores with diameters of ~ 200 or ~ 300 nm could be produced (depending on the formal anodization time), with controllable pore depths from 400 to 1900 nm. An increase in anodization time enhanced the oxide layer height without affecting the regularity of long pore channels.

Theoretically, the porous AAO film’s interpore distance, barrier layer thickness, and lattice size linearly correlate with the applied anodization potential. A higher voltage intensifies the electric field at the metal–electrolyte interface, accelerating ion migration and promoting oxide growth. This was confirmed by the data for AAO templates G1–G3 (Fig. 7), showing substantial increases in pore diameter, interpore distance, and barrier layer thickness with voltage. However, a thickened barrier layer compromises conductivity and current efficiency. Conversely, porosity and pore density slightly decreased with the increase in voltage. The pore aspect ratio increased with voltage, sharply rising between 150 and 160 V (Table 3), potentially due to changes in the dynamic balance between oxidation and dissolution rates within this voltage range.

Fig. 7.

Interpore distances, pore diameters, and barrier layer thicknesses versus re-anodization voltage for AAO molds G1–G3

Maintaining a uniform current density, electrolyte temperature, and ion distribution in large-area anodization is challenging. Theoretically, larger reaction areas can cause uneven current distributions, leading to localized electrical breakdown. This compromises the regularity of pore self-assembly and even causes irregular aluminum surface deformation due to internal stress. We aimed to mitigate these problems by using thick, high-purity aluminum substrates and slowing the current increase to reduce breakdown risks and enhance structural integrity. The experimental results showed that under identical process parameters, increasing the anodization area slightly increased the AAO film’s pore diameter, interpore distance, and porosity while slightly decreasing its pore density and lattice thickness, consistent with a previous study [48]. Therefore, despite optimization, reaction area changes still influenced the final pore parameters, so current density and voltage stability should be controlled to fabricate uniform large-area AAO structures.

A further consideration is the ultimate resolution achievable with our optimized large-area fabrication platform. While this study focused on controlling nanostructures with interpore distances in the 100–400 nm range for specific biomimetic functions, the underlying anodization technique is capable of producing much finer features. The literature extensively reports that by employing electrolytes such as sulfuric acid at lower anodization voltages (e.g., 15–25 V), highly ordered nanopores with diameters as small as 25 nm and interpore distances of ~ 50–70 nm can be achieved [49, 50]. Although exploring these higher resolutions was beyond the scope of the current work, the process control strategies we developed—particularly for thermal management and current uniformity over large areas—provide a robust foundation for future efforts to scale up the fabrication of these high-resolution AAO templates for applications in nanoelectronics, data storage, or advanced filtration.

Geometric dimensions of hot-embossed and UV-imprinted replicas

We compared the average diameters and heights of the AAO master molds with those of the hot-embossed and UV-imprinted replicas (Tables 3 and 4). Visual confirmation of the replicated nanostructures is provided by the surface and cross-sectional Scanning Electron Microscopy (SEM) images, which have been included in Figures S1–S2. The hot-embossed replicas consistently had smaller average column diameters, defined as the diameter at the base of the tapered structure, and heights than the AAO master molds’ pore diameters and depths. This difference arose from limitations in the filling depth of the polymer into the nanopores and minor shrinkage due to demolding stress. By contrast, the UV-imprinted replicas generally exhibited larger average column diameters and heights than the AAO master molds’ pore diameters and depths. This deviation was likely due to a combination of UV curing shrinkage, stress interaction with the substrate, and operational forces during demolding, which could have induced the partial fracturing or stretching deformation of the nanostructures.

Table 4.

Nanostructure widths and heights, aspect ratios, and transfer rates of hot-embossed and UV-imprinted replicas

No	Hot-Embossed Replicas					UV-Imprinted Replicas
	Column Diameter (nm)	Column Height (nm)	Aspect Ratio	Column Density (μm−2)	Transfer Rate (%)	Column Diameter (nm)	Column Height (nm)	Aspect Ratio	Column Density (μm−2)	Transfer Rate (%)
E1	223	396	1.8	31	98.6	268	312	1.2	26.5	84.3
E2	312	1658	5.3	59.9	95.9	385	2081	5.4	38.7	61.9
F1	103	31	0.3	38.4	98.2	141	84	0.6	31.7	80.9
F2	66	105	1.6	81	98.1	121	170	1.4	71.6	86.8
F3	380	188	0.5	31.8	97.2	631	368	0.56	25.4	77.8
F4	243	397	1.6	23.4	84.7	523	588	1.1	18.9	68.2
G1	142	388	2.7	37.5	97.6	171	228	1.3	36.6	95.3
G2	158	523	3.3	33.5	97.9	262	733	2.8	29.2	85.2
G3	191	1218	6.37	22.4	81.2	401	2381	5.9	14.4	52.4

We assessed the replication fidelity by defining a "transfer rate," calculated as the ratio of the areal density of nanostructures on the replica to the areal density of nanopores on the AAO master mold. This ratio was quantified by analyzing feature counts from the grayscale density distributions of top-down SEM images (Figures S1-S2) of the master and replica molds using ImageJ, with the final rates presented in Table 4. Transfer rate analysis revealed that hot embossing molding typically achieved transfer rates exceeding 95%, whereas UV curing obtained 60%–87% transfer rates. This seemingly counter-intuitive result, where the low-viscosity UV resin process yields a lower transfer rate, is attributed primarily to challenges during the demolding step. We hypothesize that the strong interfacial adhesion between the cross-linked polymer nanostructures and the walls of the high-aspect-ratio AAO nanopores, coupled with the inherent brittleness of the cured resin, results in a significant number of nanostructures fracturing or delaminating from the substrate upon separation. Although hot embossing demonstrated higher replication fidelity, UV imprinting remains a viable technique for specific applications due to its rapid processing and robustness to high temperatures or pressures.

Reflectance of hot-embossed molds

We transferred geometric patterns from nine Ni–Co electroformed molds to the surface of commercially available, general-purpose polycarbonate (PC) sheets via hot embossing. Reflectance and transmittance were measured using UV–vis spectrophotometry (U3010, Hitachi, Japan) across the visible light spectrum (400–700 nm). The original unstructured PC board exhibited an average reflectance of 10.85% and an average transmittance of 80.1% in this range (Table 5). This baseline reflectance is higher than that of an ideal optical-grade surface, likely due to minor surface roughness inherent to the commercial substrate. The full spectral data for all samples are provided in Fig. 8. The unstructured PC sheet shows a relatively flat and high reflectance across the visible spectrum. In contrast, the nanostructured sample G2 demonstrates a strong broadband antireflective effect, with its reflectance suppressed across all wavelengths and dropping below 3% near 700 nm. This clear wavelength dependence is a characteristic feature of such diffractive nanostructures. While the effect varies with wavelength, the average reflectance remains a useful metric for comparing overall performance. Based on this metric, sample G2, with the lowest average reflectance of 3.5%, represents the most effective antireflective surface produced in this study. Conversely, transmittance generally increased (Table 5), with F4 achieving the highest average transmittance (92.3%). It is important to contextualize this result with state-of-the-art antireflective surfaces, such as the exceptional < 1% reflectance achieved by Yanagishita et al. [51]. Several factors contribute to this performance difference. First, the research objectives and scope diverge; our study’s primary goal was to establish a process-structure–function roadmap for multifunctional surfaces in the 100–400 nm size range, while the work by Yanagishita et al. [51] utilized smaller ~ 70 nm nanoarrays, which are highly effective for suppressing visible light reflection. Second, our results were achieved on the aforementioned general-purpose PC substrate with its high initial reflectance of 10.85%. Despite these differences, the significant relative reduction from 10.85% down to 3.5% demonstrates the powerful antireflective effect of our fabricated nanostructures. The high-fidelity injection molding process used in the reference study is a powerful technique, and integrating it with our controllable fabrication of AAO templates represents a promising avenue for future research to achieve even higher performance.

Table 5.

Average reflectance and average transmittance of original PC board and hot-embossed replicas after different durations of exposure to natural light

No	Ave. Reflectance			Ave. Transmittance
	0 day%	1 day%	7 days%	0 day%	1 day%	7 days%
E1	6.6	6.57	6.74	87.4	87.38	87.18
E2	4.8	4.80	4.85	89.5	89.54	89.43
F1	4.8	4.80	4.84	89.5	89.53	89.46
F2	5.4	6.51	6.52	88.3	86.15	86.14
F3	5.0	4.90	5.03	88.5	89.33	89.10
F4	3.7	3.48	3.50	92.3	92.29	92.25
G1	4.9	4.94	4.98	89.9	90.06	89.72
G2	3.5	3.51	3.60	90.9	90.98	90.86
G3	4.3	4.20	4.31	90.3	90.03	89.95

Fig. 8.

a Reflectance spectra and b transmittance spectra of hot-embossed molds: (E1, E2, F1, F2, F3, F4, G1, G2, and G3)

PC properties are generally stable under natural light for less than a month, but prolonged exposure can cause photoaging. The short-term optical stability of the hot-embossed PC replicas was assessed by exposing them to sunlight for 1 and 7 days. All replicas showed only minor changes in average reflectance (within ± 0.2%; Table 5), with E1 showing the largest change (a 0.17% increase). Similarly, the changes in average transmittance after 7 days of exposure were within ± 0.3% (Table 5), indicating a limited effect of short-term exposure on optical properties. The lack of structural changes in the PC material itself within one month meant its inherent influence on light reflection and transmission was negligible.

WCAs of UV-imprinted molds

We measured the WCAs of nine UV-imprinted replicas using a contact angle meter (Fig. 9). All replicas demonstrated a significant increase in hydrophobicity relative to the original PC board’s 91.46°. The updated WCAs of replicas E1–G3 were 121.03°, 127.59°, 106.59°, 112.74°, 116.95°, 122.13°, 125.33°, 128.81°, and 138.69°. An increase in the structural aspect ratio was a critical factor influencing the replicas’ hydrophobicity. Structures with higher aspect ratios (e.g., E2, G2, and G3) exhibited significantly higher WCAs than those with smaller aspect ratios. This observation aligned with the Cassie–Baxter model, which states that a greater aspect ratio promotes air entrapment within nanostructures, reducing the actual contact area between the liquid droplet and the solid surface. This reduction lowers the interfacial tension, thereby increasing the apparent contact angle. With a maximum WCA of 138.69°, sample G3 demonstrated highly hydrophobic behavior.

Fig. 9.

WCA measurements of (I) original PC board and replicas (E1, E2, F1, F2, F3, F4, G1, G2, and G3)

To properly position these results, a comprehensive comparison with numerous key studies is provided in the Supplementary Information (Table S1). As detailed therein, many state-of-the-art hydrophobic surfaces achieve WCAs exceeding 150°. These excellent results are often accomplished through specific techniques not employed in our current study, such as fabricating sub-100 nm nanoarrays (e.g., Guo et al. [52]), creating complex hierarchical structures [53], or applying chemical surface modifications [33, 54]. Our study, in contrast, was primarily focused on establishing precise process control for nanostructures within a specific 100–400 nm size range, with a secondary goal of evaluating the resulting multifunctional properties. Therefore, our achieved WCA of 138.69° represents a strong hydrophobic performance for a non-hierarchical, unmodified surface that was co-optimized for both wetting and optical properties (e.g., antireflection) [33, 44]. This result highlights a successful balance within the common design trade-offs of multifunctional materials and serves as a foundation for future work incorporating more advanced surface engineering techniques.

Conclusion

This study successfully demonstrated precise control of the nanostructured pores (pore diameter, interpore distance, pore depth, and pore aspect ratio) of high-aspect-ratio, large-area, highly ordered AAO templates. This was achieved through systematic manipulation of the anodization parameters, specifically by combining pre-anodization (for initial ordering) with multistage re-anodization. Our findings confirmed that pre-anodization critically determined the pore ordering and final interpore distance, with the oxalic acid and phosphoric acid electrolytes yielding ~ 100 and ~ 400 nm, respectively. The subsequent re-anodization time significantly modulated pore depth and regularity. The anodization potential directly influenced pore diameter, interpore distance, and barrier layer thickness, with the pore aspect ratio rapidly increasing in the high-voltage HA regime. A current density of 100 A/m² proved optimal for uniform, stable AAO growth. For large-area processing challenges, such as current breakdown and substrate nonuniformity, success hinged on using thick, high-purity aluminum sheets, carefully controlling and slowly ramping the reaction current, and integrating an enhanced cooling system for effective heat dissipation. A specialized clamping jig ensured precise control of the electrochemical and thermal conditions, laying a solid foundation for polymer nanostructure replication.

We successfully replicated the nanostructures of these controllable AAO templates on polymer films using nanoimprinting and assessed the biomimetic functionalities of the replicas. All nanostructured polymer replicas exhibited significantly reduced surface reflectance (to a minimum of 3.5 from 10.85% for pristine PC) and generally enhanced transmittance (to a maximum of 92.3%). Short-term sunlight exposure (7 days) negligibly affected these optical properties. As for surface wettability, all nanostructured replicas showed notable increases in WCA from the pristine PC’s 91.46°. Notably, some samples exhibit significantly enhanced hydrophobic properties, such as G3 (WCA = 138.69°). The results confirmed that the structural aspect ratio was a critical factor for hydrophobicity, with higher-aspect-ratio structures exhibiting greater WCAs due to enhanced air entrapment, consistent with the Cassie–Baxter model. In summary, this research not only elucidated the mechanisms by which anodization parameters can be used to control AAO nanostructures but also successfully transferred these structures to polymer surfaces, achieving excellent antireflective and hydrophobic properties. Our findings systematically establish the correlation between process parameters, AAO structures, and polymer surface functions, providing crucial experimental and theoretical guidance for the efficient, controllable design and fabrication of high-performance polymeric nanostructures with biomimetic functions.

Supplementary Information

Below is the link to the electronic supplementary material.

Abberivations

AAO

Anodic aluminum oxide

DI

Deionized

FIB

Focused ion beam

HA

Hard anodization

NIL

Nlithography

PC

Polycarbonate

PDMS

Polydimethylsiloxane

PTFE

Polytetrafluoroethylene

SEM

Scanning electron microscopy

T_g

Glass transition temperature

UV

Ultraviolet

UV–vis

Ultraviolet–visible

WCA

Water contact angle

D_int

Interpore distance

D_p

Pore diameter

AR

Aspect ratio

PD

Pore depth

PD

Pore density

t_w

Wall thickness

P

Porosity

BR

Barrier layer

TR

Transfer rate

HB

Hot embossing

UVI

UV imprinted

EF

Electroforming

PI

Pre-anodization

RI

Re-anodization

R

Reflectance

T

Transmittance

H

Hydrophobicity

SH

Superhydrophobicity

Author contributions

Hsuan-Hao Hung: Data curation, Formal analysis, Investigation, Validation, Writing–original draft and Writing – review & editing, Shi-Kai Lina: Formal analysis, Investigation, Ru-Xue Lina: Formal analysis, Investigation, Tzu-Ning Huang: Formal analysis, Investigation, Chia-Che Wu: Conceptualization, Funding acquisition, Project administration, Supervision, Writing–original draft and Writing – review & editing.

Funding

This research is also supported (in part) by National Science and Technology Council (NSTC) of Taiwan under grant numbers 112–2221-E-005–087, 113–2622-E-005–005, 113–2221-E-005–058, 114–2221-E-005–047-MY2.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent to Publish declaration

Not applicable.

Competing interests

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.