Coronamicroparticle Arrays with Stable Superamphiphobicity

ABSTRACT

Springtails use reentrant nanostructures to repel both water and oil liquids, inspiring scientists to construct reentrant nanostructures for superamphiphobic surfaces. However, the packing density of the reentrant nanostructures related to their liquid repelling performance is generally limited on a planar surface. Packing reentrant nanostructures on curved surfaces can greatly increase the packing density, however fabrication of reentrant nanostructures on curved surfaces is a great challenge. Here, we develop a method to fabricate reentrant nanostructure arrays on polystyrene (PS) microsphere surfaces, giving rise to the formation of coronamicroparticles with the morphology resembling corona virus. The critical step to fabricating the coronamicroparticles is that a silver nanomesh is transferred onto the PS microspheres to form a conformal coating layer. Thermal evaporation of gold through the silver nanomesh formed gold nanodisk arrays on the PS microspheres. Consequent reactive ion etching of the gold nanodisk‐protected PS microspheres gives rise to the formation of coronamicroparticle arrays. The morphology of the coronamicroparticles can be simply adjusted by the size of the PS spheres and the reactive ion etching time. The coronamicroparticle arrays demonstrate stable superamphiphobicity, suggesting potential applications in self‐cleaning, liquid transportation, and sensing fields.

Controlling the interaction between the solid surfaces and liquids is an important task in diverse fields. We develop a simple method to prepare reentrant nanostructures on spherical microparticles to mimic the cross species of the coronaviruses and the springtails. The coronamicroparticle arrays show stable superamphiphobic properties resulting from the closely packed reentrant nanostructure arrays over the microparticle surfaces.

1. Introduction

Controlling the interaction between solid surfaces and liquids is important in diverse fields, including heat transfer [1, 2, 3], 3D printing [4, 5, 6], drug delivery [7, 8, 9], and sensing fields [10, 11, 12, 13]. The extreme case is the superamphiphobic surfaces that can almost completely repel both water and oil [14, 15, 16, 17]. Many plants and animals have evolved special surface structures to repel water or even oil [18, 19, 20]. The well‐known examples are the lotus leaves [21, 22] employing micro/nanostructures to repel water and the springtails [23, 24] using reentrant nanostructures to repel both water and oil. These examples inspired scientists to construct diverse reentrant nanostructures to achieve superamphiphobic surfaces [25, 26, 27, 28, 29, 30, 31]. Owing to the nanometer scale and special structure of the reentrant nanostructures, the sophisticated fabrication method of the reentrant nanostructures has to rely on photo‐ or electron beam‐lithography technique [32, 33, 34]. These lithography approach can only prepare rigid reentrant nanostructures on flat silicon surfaces [1, 28, 29]. Increasing the packing density of reentrant nanostructures needs to prepare reentrant nanostructures on curved surfaces. Jung, et al. first prepared reentrant nanostructures on polystyrene (PS) membranes, and consequent thermal treatment created wrinkled PS membranes covered by reentrant nanostructures [35, 36]. The wrinkled PS membranes covered by reentrant nanostructures showed both high static repellency and pressure resistance [35, 36]. Coronaviruses are surrounded by coronas to enhance or prohibit their interactions with surrounding environment [31]. The coronavirus structure inspires us to design coronamicroparticles surrounded by coronas‐like reentrant nanostructures to realize superamphiphobicity. However, existing methods are inapplicable to prepare reentrant nanostructures on microparticle surfaces. In this study, we demonstrate fabrication of coronamicroparticle arrays using monolayer colloidal crystal (MCC) templates self‐assembled from PS spheres. The critical step is that we realize transferring of a freestanding silver nanomesh film conformally onto the MCC template formed by PS microspheres. Thermal evaporation of gold through the nanomesh covered MCC template results in the formation of a hexagonally packed gold nanodisk array over each PS microspheres after peeling off the nanomesh mask. Reactive ion etching (RIE) treatment of the gold nanodisks‐protected PS microspheres generates coronamicroparticle arrays. The coronamicroparticle arrays show stable superamphiphobicity. Our method is very general and is applicable to prepare ordered reentrant nanostructures on curved surfaces of any materials that can be etched by plasma, pushing forward the development of the surface nanopatterning methods. Our method opens a new avenue toward engineering reentrant nanostructures on curved surfaces with promising applications in metamaterials, plasmonics, and biomedical fields.

2. Results and Discussion

2.1. Fabrication Process of the Coronamicroparticle Arrays

To emulate the coronavirus surface, we developed a high‐throughput technique to prepare hexagonally arranged reentrant nanostructures on PS microspheres by combining the nanosphere lithography technique and the RIE method (Figure 1a). First, an MCC template composed of hexagonally arranged 1 µm PS spheres was prepared by a self‐assembly process (Figure 1b and Figure S1) [37, 38]. The area of the MCC template could easily reach square centimeters, mainly limited by the surface area of water. Then, oxygen plasma etching was used to reduce the size of the PS spheres into desired sizes (Figure 1c). Subsequently, 100 nm‐thick Ag film was thermally evaporated onto the loosely packed PS sphere arrays. After dissolving the PS spheres by ultrasonic cleaning in dichloromethane, an Ag nanomesh with highly ordered nanohole arrays was obtained (Figure 1d) [39, 40, 41]. The Ag nanomesh was coated with a self‐assembled monolayer (SAM) of octadecanethiol by immersing into a 30 mM octadecanethiol ethanol solution for 1 h to enhance its surface hydrophobicity. The Ag nanomesh on the glass substrate was inserted slowly into a NH₄F aqueous solution at an angle of 45° to slowly peel off the nanomesh from the glass substrate. The intact freestanding Ag nanomesh was floated at the water/air interface. An MCC template composed of 5 µm PS spheres was used to pick up the freestanding Ag nanomesh at the water surface (Figure 1e,f and Figure S2). The Ag nanomesh conformally attached onto the underneath PS microspheres (Figures S3 and S4). Consequently, 50 nm‐thick Au was thermally evaporated onto the Ag nanomesh‐covered MCC template, giving rise to the formation of Au nanodisks on the PS microspheres (Figure S5). The Ag nanomesh was peeled off using a piece of scotch tape. The RIE process was carried out to etch the PS microspheres protected by Au nanodisks. The undercut of the Au nanodisk covered PS area created reentrant nanostructures, resulting in the formation of coronamicroparticles (Figure 1g and Figure S6). The oblique and the cross‐sectional scanning electron microscopy (SEM) images revealed that the coronamicroparticles mimic the structure of the springtail and the coronavirus surfaces (Figure 1h,i). A magnified SEM image indicated that the coronal nanostructure was composed of Au nanodisk head and PS nanopillar neck (inset of Figure 1h). We designed the average diameter (d ∼ 500 nm) of the head and the distance (p ∼ 1 µm) between the adjacent head of the coronal nanostructures to well match the dimensions of the springtails. By controlling the plasma etching time, the maximum height of the PS nanopillar can reach about 1.1 µm (Figure S7). The spherical surface determined that the height of the nanopillar decreased from the top to the equator of the PS microsphere (Figure 1i). The coronamicroparticle arrays were then chemically modified by a thin fluorosilane layer (FOTS) realized by a molecular vapor deposition process of 1H,1H,2H,2H‐Perfluorodecyltrichlorosilane.

FIGURE 1.

Fabrication process of the springtail and coronavirus‐inspired coronamicroparticle arrays. (a) Schematic of the steps used to fabricate coronamicroparticle arrays. First step: oxygen plasma treatment of the PS sphere array. Second step: thermal evaporation of Ag over the non‐closely packed PS sphere array and removing the PS spheres. Third step: Ag nanomesh was transferred onto PS microsphere array. Fourth step: thermal evaporation of Au and removing the Ag nanomesh. Last step: oxygen plasma treatment of the Au nanodisk protected PS microspheres and surface modification. (b) SEM image of 1 µm PS sphere arrays. (c) The SEM image of PS sphere arrays with the diameter of 500 nm after oxygen plasma treatment. (d) SEM image of Ag nanomesh. (e) SEM image of 5 µm‐sized PS sphere arrays. (f) Ag nanomesh covered on the 5 µm PS sphere arrays. The red dotted line indicates the outline of the PS spheres underneath the Ag nanosmesh. (g–i) The SEM images of the coronamicroparticle arrays from the top view, the oblique view, and the cross‐sectional view, respectively. Inset in (h): Magnified observation of the reentrant nanostructures.

It is easy to control the morphology of the coronamicroparticles. For example, the mesh size of the silver nanomesh determined the size of the gold nanodisks and the inter‐nanodisk distance, as well as the number of reentrant nanostructures on a single microsphere substrate. The mesh size can be easily controlled by the reactive ion etching time of the PS nanospheres. The size of the PS microsphere substrate determined the number of the reentrant nanostructures and the entire size of the coronamicroparticles. The height of the reentrant nanostructures is determined by the reactive ion etching time of the PS microspheres. The size of the PS spheres is available from less than 50 nm and more than 10 µm, allowing us to prepare coronamicroparticles of various morphologies.

2.2. Superamphiphobicity of the Coronamicroparticle Arrays

Our artificial springtail surface showed outstanding static superamphiphobicity (Figure 2a and Figure S8). To quantitatively evaluate the wetting properties of the obtained coronamicroparticle arrays, we used three liquids including deionized (DI) water (σ = 71.9 mN/m), glycerol (σ = 63.3 mN/m), and ethylene glycol (σ = 46.5 mN/m) to compare the liquids repellency of the PS microsphere arrays and the coronamicroparticle arrays (Figure 2b). The coronamicroparticle arrays were prepared by constructing reentrant nanostructures on microsphere arrays (Figure 2c). Benefiting from the microscale structures, the static contact angles of PS sphere arrays for the three liquids were greater than 120°. In contrast, the designed coronamicroparticle arrays exhibited excellent superamphiphobicity with static contact angles of 155.3°, 150.9°, and 146.6° for water, glycerol, and ethylene glycol, respectively (Figure 2b). This enhancement in liquid repellency was attributed to the reentrant nanostructures with stable air pockets and low solid fraction [18, 35].

FIGURE 2.

Wetting characterization of PS microsphere arrays and coronamicroparticle arrays. (a) Liquid repellency of coronamicropaticle arrays. (b) Static contact angels of water, glycerol, and ethylene glycol on PS microsphere arrays and coronamicroparticle arrays. The data are the average value of 5 independent experiments. (c) Schematic images of PS microsphere arrays and coronamicroparticle arrays. (d) Rolling angle of water on six superamphiphobic samples and each sample was tested five times. (e) Self‐cleaning tests of coronamicroparticle arrays and PS microsphere arrays by water droplets. The sand particles have a size of 20 to 300 µm. (f, g) The water droplet was dropped onto the coronamicroparticle arrays and PS microsphere arrays. Inset in f and g) Magnified observations.

Droplet adhesion performance of coronamicroparticle arrays was further evaluated. The rolling angle of the water droplet on PS microsphere arrays was 34.9°. In contrast, the rolling angle of a water droplet on the coronamicroparticle arrays was only 3.8° (Figure S9 and Movie S1). Six samples were prepared and each sample was tested five times to demonstrate their super‐slippery surface characteristics (Figure 2d). The oil‐repellency of the samples was characterized by measuring the roll‐off angles of glycerol and ethylene glycol on the six superamphiphobic samples. Each sample was also tested five times, showing excellent sliding ability (Figure S10). The liquid‐repellent coronamicroparticle arrays have broad applications. As an example, we demonstrated the self‐cleaning performance of the coronamicroparticle arrays. To simulate the real environment as much as possible, we chose dust particles with sizes ranging from 20 to 300 µm collected directly from the roadside as the pollution source and placed the tested samples at an angle of 30° (Figure 2e and Figure S11). Water droplets dropped vertically and bounced on the surface of the coronamicroparticle arrays. The droplet easily took the dust particles away and finally rolled off the substrate (orange circle in Figure 2f and Movie S2). We performed the self‐cleaning experiments for more than 50 times with the self‐cleaning performance unchanged. In comparison, the water droplet directly adhered to the PS microsphere arrays and contaminated the surface (Figure 2g and Movie S3).

2.3. Theoretical Wettability Model of the Coronamicroparticle Arrays

Contact angle analysis of the coronamicroparticle arrays supported the Cassie‐Baxter model [16, 42]. Microscopic pockets of air were trapped underneath the water droplets, forming a composite solid‐liquid‐air interface (Figure 3a). This composite interface could increase the macroscopic contact angle and reduce the contact angle hysteresis, enabling the water droplets to roll off easily to clean the surface. We estimated the apparent contact angles (θ*) of different liquids on coronamicroparticle arrays using Equation (1) obtained from the Cassie‐Baxter model:

$$\cos {\theta }^{\ast }=f_{sl}\left(1+\cos {\theta }_i\right)-1$$ (1)

where θ _i is the intrinsic contact angle of the smooth Au‐coated silicon substrate, and f_sl is the solid‐liquid contact fraction, i.e., the ratio of the solid to the total contact area between the liquid droplet and the underneath solid structure. The value of θ _i for water, glycerol, and ethylene glycol is 98°, 84°, 65°, respectively (Figure 3b). The f_sl for the coronamicroparticle structures could be estimated by Equation (2) [43]:

$$f_{sl}=\frac{19\pi r^2}{2\pi R^2}=9.5{\left(\frac{r}{R}\right)}^2$$ (2)

where r is the radius of the Au head and R is the radius of the PS sphere. Because the Au head can repel the droplet sitting on the coronal structures, f_sl equals to the ratio of the area of all Au heads on a PS sphere to the half surface area of a PS sphere (Figure 3c), which approximately deduces Equation (2).

FIGURE 3.

Surface wettability model for coronamicroparticle arrays. (a) The schematic diagram of liquid droplet resting on coronamicroparticle arrays. (b) The intrinsic contact angle of water, glycerol, and ethylene glycol on smooth Au‐coated Si wafer after vapor deposition of 1H,1H,2H,2H‐Perfluorodecyltrichlorosilane. (c) The number of reentrant structures on each PS sphere. (d) The theoretical and experimental value of the static contact angle of water, glycerol, and ethylene glycol on coronamicroparticle arrays.

Based on Equations (1) and (2), as well as the structural parameters of the coronamicroparticle arrays, the calculated θ*of different liquids on the coronamicroparticle arrays were shown as the orange histogram on the left, and the corresponding experimental θ* were shown as green histogram on the right (Figure 3d and Table S1). The calculated values of contact angles were similar to the experimentally observed ones, indicating the suitability of the Cassie‐Baxter model for the coronamicroparticle arrays.

We obtained reentrant nanostructures of different heights by RIE of the gold nanodisk protected PS microspheres for different times. We found that the height of the reentrant nanostructures showed negligible impact on the surface wettability (Figure S12). This is understandable considering that the liquid droplet only has a direct contact with the top gold nanodisks. Therefore, the diameter of the gold nanodisks should play important roles in influencing the wettability. We therefore controlled the diameter of the gold nanodisk by etching the 1 µm PS spheres for different times (Figure S13). When etching the 1 µm PS spheres for 4.5, 6, and 7.5 min, nanodisks with a diameter of 650, 500, and 240 nm were obtained, respectively. When the nanodisk size decreases, f _sl becomes smaller. As expected, the liquid‐repellent performance of the coronamicroparticles obtained using the nanodisks prepared by etching the PS spheres for 6 min is better than that of the 4.5 min. However, further etching for 7.5 min reduced the size of the PS spheres to less than half of the original size, creating extremely rough surfaces. Therefore, uniform reentrant nanostructures were failed to be obtained and the liquid‐repellent performance was greatly reduced. Overall, we can simply vary the morphology of the coronamicroparticles. The morphology has important influence on the wettability of the coronamicroparticles.

2.4. Excellent Pressure Resistance of the Coronamicroparticle Arrays

To study the pressure resistance of the coronamicroparticle arrays under real conditions, we observed the bouncing process of a droplet on the coronamicroparticle arrays. The time‐lapsed images of the water droplets released from a height of 5 cm impacting on the coronamicroparticle arrays and the PS microsphere arrays were captured (Figure 4a). According to the photographs, the impacting droplet spread to the maximum spreading diameter and then lifted without any residues, which implied high resistance to pressure. In contrast, the water droplet exhibited strong adhesion to the surface of the PS microsphere arrays and partially rebounded.

FIGURE 4.

Droplet bouncing tests. (a) Selected snapshots of a bouncing droplet on coronamicroparticle arrays and PS microsphere arrays. (b) The plots of maximum spreading as a function of We for PS microsphere arrays and coronamicroparticle arrays. (c, d) Contact time analysis for PS microsphere arrays and coronamicroparticle arrays. (e) Bouncing behaviors of ethylene glycol droplet for coronamicroparticle arrays.

The outcome of the impacting droplet was affected by a variety of test conditions, including liquid properties (temperature, surface tension, etc.) and kinetic parameters (droplet volume, descent height, etc.) [44]. Dimensionless parameter Weber number (We) was employed to investigate droplet impact process under different experimental parameters: We  =  ρV_i ² d ₀/σ, where V_i is the impact velocity, ρ is the density, and σ is the surface tension [35, 45]. The higher the impact pressure of the droplet, the greater the value of We. The water spreading and bouncing‐off behaviors as a function of We changing from ∼ 32 to ∼ 191 was studied (Figure 4b). In all tests, the droplets remained complete rebounding on the coronamicroparticle arrays exhibiting high stability against pressure. For high values of We, the droplet hit on the surface with extremely high impact velocity, resulting in fragmentation and the formation of many secondary droplets (hollow symbols in the plots of Figure 4a and Movie S4). While PS microsphere arrays did yield a loss of resistance, which implied partial rebounding and deposition behaviors as We was increased. The complete rebounding indicated that the air pockets of coronal arrays did not experience permeation under high pressure, different from those of the PS microsphere arrays. We attributed this remarkable improvement to the stable liquid pinning points provided by the reentrant nanostructures. The maximum spreading factor D_max * (D_max * = d_max  /d ₀, where d_max is the maximum diameter of the spread droplet and d ₀ is the initial diameter of the droplet) for coronal arrays was greater than that for PS mcirosphere arrays, indicating more effective drag reduction and lower contact angle hysteresis for coronal arrays (Figure S14).

The value of contact time t_c , the time between initial impact and bouncing off, of water on the coronamicroparticle arrays and the PS microsphere arrays was further measured. The results indicated the overall affinity between the liquid and the surface. The spreading factor D* (D* = d_t  /d ₀, where d_t is the spreading diameter at time t) of water (We ∼ 32) as a function of time was measured (Figure 4c). The droplet receded right after reaching the maximum spreading diameter (t = 2.16 ms, D_max * = 1.92) and completely rebounded off the coronamicroparticle arrays (t_c = 11.7 ms), in a sharp contrast to the deposition phenomenon on the PS microsphere arrays. As reported in previous studies [46], the contact time was independent of We. The ratio of contact time t_c to inertial‐capillary timescale τ ($\tau =\sqrt{\rho {d_0}^3/4\sigma }$) on the macroscopically untextured surfaces is all close to the minimum value of 2.2 (Figure 4d and Table S2). These phenomena further confirmed the outstanding water repellency of the coronamicroparticle arrays.

The artificial coronamicroparticle arrays also showed certain pressure resistance to ethylene glycol when We was less than 120 (Figure 4e). The dynamic droplet bouncing results proved the stable droplet repellency of the coronamicroparticle arrays for fast removing droplets. The droplets would not reach the gap between the coronal nanostructures and could easily move on the surface.

To evaluate the durability of the coronamicroparticle arrays in practical applications, chemical stability tests (hydrochloric acid immersion and ultraviolet irradiation) and mechanical durability tests (sandblasting abrasion) were conducted. For the chemical stability tests, the samples were either immersed in a 6 wt.% hydrochloric acid solution or irradiated with ultraviolet light at a wavelength of 365 nm. Three groups were set for each treatment, with durations of 0 h, 24 h, and 72 h, respectively. The SEM images and rolling angle data of the samples under different treatment conditions are shown in Figures S15 and S16, respectively. The results indicated that after immersing the structure into hydrochloric acid immersion or ultraviolet irradiation for several hours, the structure and performance of the coronamicroparticle arrays remained stable (Figure S15b, e). The unchanged superhydrophobic properties further confirmed this conclusion (Figure S16). After 72 h of treatment, some damage to the microstructure occurred with slight influence on the superhydrophobic performance. The mechanical durability test was carried out using a sandblasting abrasion method. 50 g of sand powders were poured onto the prepared sample surface from a height of 10 cm. After sandblasting treatment, several water droplets were dropped onto the surface. Owing to the strong repellency of the surface, the water droplets exhibited a significant self‐cleaning effect. After repeating the sandblasting cycle for 0, 30, and 50 times, the surface morphology and the roll‐off angle of the samples were shown in Figures S17 and S18, respectively. The results demonstrated that the samples still maintained excellent superhydrophobic performance after sandblasting abrasion treatment.

3. Conclusions

In summary, we developed a general strategy to prepare reentrant nanostructures on curved surfaces, allowing us to better mimic the cross species of the springtails and the coronaviruses. Inspired by the coronavirus surfaces surrounded by coronals, we engineered the coronamicroparticles armored by hexagonally arranged reentrant nanostructures using a freestanding Ag nanomesh capable of conformally attaching onto any curved surfaces. The coronamicroparticle arrays demonstrated outstanding and stable liquid repellency benefiting from the high density of the reentrant nanostructures. Our method can be extended to prepare ordered reentrant nanostructures on curved surfaces of any materials that can be etched by RIE method, overcoming the planar limitations of traditional lithography methods. The availability of PS spheres from less than 50 nm to more than 10 µm allowed one to control the morphology of the coronamicroparticles, pushing forward the development of the reentrant nanostructures in practical fields and opening up new application opportunities of the reentrant nanostructures in sensing and other fields [47, 48, 49, 50].

4. Experimental Section

4.1. Preparation of the MCC Templates

MCC templates composed of hexagonally arranged PS spheres were prepared by a self‐assembly process [37, 38, 39]. The glass slide (1.5 cm × 1.5 cm) was ultrasonically rinsed successively by deionized water and ethanol for 15 min each. Then, oxygen plasma treatment was performed for 10 min to make the glass slide superhydrophilic. The glass slide was placed at the mid‐bottom of a crystallizing dish (with the diameter of 10 cm), and then DI water was added carefully to the dish around the glass to level the edge of the upper surface of the glass. Monodisperse PS colloidal sphere (10 wt.% aqueous dispersion) was diluted with an equal volume of ethanol. Then the dispersion was dropped on the top of the glass, which spread freely to cover nearly the whole water surface. Within a few seconds, the MCC template was formed on the water surface. A few microliters of sodium dodecyl sulfate solution (2 wt.%) was dropped onto the water surface to vary the surface tension. The MCC template was packed closer and then quickly picked up by a piece of silicon.

4.2. Preparation of the Ag Mesh [39]

The 1 µm PS spheres were etched for 6 min, unless otherwise specified, within a plasma etching machine (Samco, RIE‐10NR) at a power of 50 W in an oxygen flow of 50 sccm. The chamber pressure within the machine was 20 Pa. An ultrathin Ag layer (∼100 nm) was thermally deposited onto the plasma‐etched PS sphere array using a thermal evaporator at a deposition rate of 0.1 nm/s under a vacuum of 5×10⁻⁴ Pa. After ultrasonic treatment in dichloromethane for 10 min, the PS spheres were completely removed and the Ag mesh with highly ordered nanohole arrays was obtained. Then, the Ag mesh was coated with a self‐assembled monolayer of octadecanethiol from a 30 mM solution in ethanol for 1 h to enhance its surface hydrophobic properties. 5 mL of 40 wt.% HF and1.5 g NH₄F were added into 50 mL of deionized water [39]. The mesh attached to the glass slide was immersed into the etching solution for 5 min and was then inserted slowly into deionized water at an angle of ∼45°, leaving the intact mesh on the water/air interface.

4.3. Preparation of Coronamicroparticle Arrays

The Ag mesh could be picked up by another 5 µm PS spheres. The PS spheres were heated at 110°C for 1 min to ensure a strong bond between the PS spheres and the silicon substrate. Then, Au layer (∼ 50 nm) was thermally evaporated onto the Ag nanomesh, using the similar method described above. The Ag nanomesh was peeled off using a piece of scotch tape, leaving the Au nanodisks covering the PS spheres. Finally, the coronal arrays were fabricated by plasma etching the sample for 12 min in the plasma etching machine at a power of 50 W in an oxygen flow of 50 sccm.

4.4. Fluorosilanization Process

The PS arrays and coronal arrays were modified by a thin fluorosilane layer (FOTS) realized by a molecular vapor deposition process of 1H,1H,2H,2H‐Perfluorodecyltrichlorosilane. The deposition process was performed for 1 h in vacuum.

4.5. Measurements and Characterization

The morphology of the samples was observed by a scanning electron microscopy (SU8010, HITACHI). The apparent contact angle measurements were performed using a Dataphysics OCA 20 goniometer. A 3.5 µL droplet generated at a speed of 0.5 µL/s was applied onto the substrate. 7 µL droplet was used for the measurement of roll‐off angles. At least five independent measurements were performed and the values were averaged for all of the data reported here. Self‐cleaning tests were taken by adhering the samples onto a glass slide using double‐sided tape. Real dust was spread onto the substrate. Bouncing off behaviors after drop impact on fabricated surfaces were captured by a high‐speed camera (HG‐LE, REDLAKE MASD). 7 µL water droplet and 5 µL ethylene glycol were used for pressure resistance tests. The height of the droplet was controlled from 5 cm to 30 cm in the case of water (We from ∼32 to ∼191) to get different hydrodynamic pressure (We). On the other hand, ethylene glycol was controlled finely from 5 cm to 15 cm (We from ∼50 to ∼149).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: smtd70801‐sup‐0001‐SuppMat.docx.

Supporting File 2: smtd70801‐sup‐0002‐MovieS1.mp4.

Supporting File 3: smtd70801‐sup‐0003‐MovieS2.mp4.

Supporting File 4: smtd70801‐sup‐0004‐MovieS3.mp4.

Supporting File 5: smtd70801‐sup‐0005‐MovieS4.mp4.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.