Superomniphobic Surfaces with Improved Mechanical Durability: Synergy of Hierarchical Texture and Mechanical Interlocking

Abstract

Due to their unique functionality, superomniphobic surfaces that display extreme repellency toward virtually any liquid, have a wide range of potential applications. However, to date, the mechanical durability of superomniphobic surfaces remains a major obstacle that prevents their practical deployment. In this work, a two-layer design strategy was developed to fabricate superomniphobic surfaces with improved durability via synergistic effect of interconnected hierarchical porous texture and micro/nano-mechanical interlocking. The improved mechanical robustness of these surfaces was assessed through water shear test, ultrasonic washing test, blade scratching test, and Taber abrasion test.

Graphical Abstract

A simple two-layer design strategy is developed to enable the fabrication of robust superomniphobic surfaces. The interconnected hierarchical porous texture of fluorinated fumed silica particles and the micro/nano-mechanical interlocking between the particle layer and the polyurethane based adhesive layer results in the improved mechanical robustness of superomniphobic surfaces.

Superomniphobic surfaces that exhibit extreme repellency towards virtually any liquid have attracted increasing interest due to their great potential for a wide variety of applications such as chemical shielding,^[1] carbon dioxide absorption,^[2] water energy harvesting^{[3, 4]} and paper-based microfluidics.^[5] In the past few years, the noticeable advances in understanding the design principles of superomniphobic surfaces have enabled rapid development of such surfaces.^[6–12] To date, various techniques have been used to fabricate superomniphobic surfaces, which possess unique functionalities such as transparency,^{[13, 14]} reversible morphological transformation,^[15] tunable wettability^[16] and weight bearing capability.^[17] In spite of noticeable progress in design and fabrication of superomniphobic surfaces,^[18–21] the mechanical robustness of these surfaces remains a major obstacle that impedes their practical applications. This is because the delicate micro/nano-textures of superomniphobic surfaces are susceptible to external mechanical abrasion and wear, which leads to rapid degradation of superomniphobicity.^[22] In this work, we developed a design strategy that enables the fabrication of superomniphobic surfaces with improved durability via synergistic effect of interconnected hierarchical porous texture and micro/nano-mechanical interlocking. We demonstrate the improved mechanical robustness of our superomniphobic surfaces through water shear test, ultrasonic washing test, blade scratching test, and Taber abrasion test.^[22]

To minimize the overall free energy, a liquid droplet contacting a textured solid surface can adopt either the Wenzel state^[23] or the Cassie-Baxter state^[24] and exhibits an apparent contact angle θ*. In the Wenzel state, the liquid droplet completely permeates the surface texture. In the Cassie-Baxter state, air pockets remain trapped in the surface texture underneath the liquid droplet. This leads to high apparent contact angle and low contact angle hysteresis$\Delta {\theta }^{*}={\theta }_{adv}^{*}-{\theta }_{rec}^{*}$, where ${\theta }_{adv}^{*}$ and ${\theta }_{rec}^{*}$ are the apparent advancing and apparent receding contact angles, respectively. Low Δθ* in turn results in low roll off angle ω (i.e., the minimum angle by which the surface must be tilted relative to the horizontal for the droplet to roll off).^{[25, 26]} Consequently, the Cassie-Baster state is desired for designing super-repellent surfaces.^[27–32] Superomniphobic surfaces, which exhibit apparent contact angle θ* > 150° and roll off angle ω < 10 with both high surface tension liquids (e.g., water) and low surface tension liquids (e.g., oils and alcohols), can be designed by combining a surface chemistry possessing low solid surface energy with re-entrant surface texture (i.e., convex or overhang texture).^{[6, 9]}

In this work, we developed a simple two-layer design strategy to fabricate superomniphobic surfaces that continue to repel high and low surface tension liquids even after various harsh mechanical durability tests. The first layer of our superomniphobic surface was a polyurethane based adhesive, which was applied uniformly onto a glass substrate via spin coating. Immediately after spin coating the polyurethane based adhesive, a layer of fluorinated fumed silica particles was sprayed onto the adhesive layer (Figure 1a and see the Experimental Section). We chose to use the polyurethane based adhesive because of its unique capability of expansion during the curing process,^[33–35] which allowed the adhesive to penetrate into the porous structure of fluorinated fumed silica particles, form strong mechanical interlocking and firmly hold the particles. We chose to use fumed silica particles not only because of the ease of modifying their surface chemistry to impart low surface energy material (e.g., fluorinated silane), but also because of their interconnected hierarchical porous structure with re-entrant texture (i.e., micron-scale agglomerates consisting of interconnected nano-sized silica particles, Figure 1a). It is worth noting that the micron-scale particle agglomerates are formed through weak interactions (e.g., van der Waals forces) between nano-scale particle aggregates.^[36] The low solid surface energy and re-entrant texture of fumed silica particles layer imparted superomniphobicity to the surfaces, while the synergistic effect of interconnected hierarchical porous surface texture and strong micro/nano-mechanical interlocking between the particle layer and adhesive layer resulted in the improved mechanical robustness of our superomniphobic surfaces.

Figure 1.

(a) Schematic illustrating the fabrication of robust superomniphobic surface and scanning electron microscope (SEM) image showing a representative fumed silica agglomerate consisting of interconnected nano-sized silica particles. (b) and (c) Apparent contact angles and roll off angles, respectively, of water and n-hexadecane droplets (~8 μl) on the surfaces fabricated with adhesive layer of different thickness. SEM images showing the morphology of surfaces fabricated with adhesive layer of (d) 0 μm, (e) 27 μm, and (f) 78 μm. Jets of (g) water and (h) n-hexadecane bouncing on the robust superomniphobic surfaces.

The thickness h of adhesive layer plays an important role in determining the wettability and mechanical robustness of the surfaces. When the adhesive layer is too thin (i.e., h is too small), there is insufficient mechanical interlocking between the adhesive layer and the particle layer, leading to surface with poor mechanical robustness. When the adhesive layer is too thick (i.e., h is too high), most of sprayed particles completely penetrate into the adhesive layer, and therefore, the surface coverage of the fluorinated fumed silica particles is insufficient to render the surface superomniphobic. In order to investigate the influence of adhesive layer thickness h on the superomniphobicity as well as the mechanical robustness of the surfaces, we fabricated a series of samples with systematically varied h by controlling the spin coating speed (see the Experimental Section and Supporting Information, Section 1). Our results indicate that increasing the adhesive layer thickness h leads to decreasing apparent contact angles and increasing contact angle hysteresis (Figure 1b), which in turn results in increasing roll off angles (Figure 1c) for both water (a representative liquid with high surface tension, γ_lv ≈ 72.1 mN m⁻¹) and n-hexadecane (a representative liquid with low surface tension, γ_lv ≈ 27.5 mN m⁻¹). For our surfaces, when h < 40 μm, the fluorinated fumed silica particles layer is sufficiently exposed above the adhesive layer to render the surface superomniphobic (Figure 1d and 1e). When h > 40 μm (e.g., h = 78 μm; Figure 1f), the layer of fluorinated fumed silica particles begin to completely penetrate into the adhesive layer. Consequently, the fluorinated fumed silica particles layer are not sufficiently exposed above the adhesive layer to render the surface superomniphobic. Further, our superomniphobic surfaces fabricated with h < 20 μm possess poor mechanical durability due to insufficient mechanical interlocking between the adhesive layer and the particle layer (Figure 1d). The particle layers of these surfaces can be easily removed when the surfaces are subjected to a water jet impact, resulting in the rapid loss of superomniphobicity. In contrast, our superomniphobic surfaces fabricated with 20 μm < h < 40 μm (e.g., h = 27 μm) exhibit good mechanical durability due to the strong micro/nano-mechanical interlocking formed between the adhesive layer and the particle layer (i.e., the polyurethane based adhesive penetrated into the porous structure of fumed silica particles and firmly held the particles, Figure 1e). Liquid jets of water and n-hexadecane with an impact velocity ~1 m s⁻¹ can be effectively repelled by these superomniphobic surfaces without causing any damage to the surfaces (Figure 1g and 1h, Movies S1–S3, and Supporting Information, Section 2). The impact velocity (~1 m s⁻¹) of liquid jets generated using a syringe pump is equivalent to an impact velocity of liquid droplets released from a height of ~5 cm. The surfaces sustained their superomniphobicity even after scratching and cutting with a razor blade (Movies S4 and S5).

To evaluate the mechanical robustness of our superomniphobic surfaces more quantitatively, we conducted shear tests by subjecting the surfaces to continuous shear flow in an aqueous environment. The water shear flow was generated by a motor-driven rotating propeller, which was fully immersed in a water bath. The superomniphobic surface, which was submerged in the water and mounted vertically, was placed at a distance of 10 cm from the center of the propeller. The tangential velocity of water flow and the corresponding shear stress in the vicinity of the superomniphobic surfaces were approximately 6.6 m s⁻¹ and 40 Pa (see the Supporting Information, Section 3), respectively. The surfaces sustained their superomniphobicity for up to 50 hours of water shear test, as evidenced by the virtually unaltered apparent contact angles (Figure 2a) and roll off angles (Figure 2b) of both water and n-hexadecane.

Figure 2.

(a) and (b) Apparent contact angles and roll off angles, respectively, of water and n-hexadecane droplets (~8 μl) on the superomniphobic surface subjected to water shear tests. (c) and (d) Apparent contact angles and roll off angles, respectively, of water and n-hexadecane droplets (~8 μl) on the superomniphobic surface subjected to ultrasonic washing tests. Droplets of (e) water and (f) n-hexadecane bouncing on the superomniphobic surface after ultrasonic washing test.

To assess the mechanical durability of our superomniphobic surfaces under harsher conditions, we further conducted ultrasonic washing tests. The superomniphobic surface was immersed in an ethanol-filled ultrasonic tank, which was operated at a frequency of 40 kHz and a power of 110 W. Ethanol was used as the solvent because of its extremely low surface tension (γ_lv ≈ 22.1 mN m⁻¹), which enabled complete wetting of the superomniphobic surface. Therefore, the entire structure of the superomniphobic surface experienced the ultrasonic washing test. During the ultrasonic washing process, the local shear stress exerted on the surface due to the implosion of cavitation bubbles generated by the ultrasonic bath can reach a few MPa.^[37] Our surface remained superomniphobic, even at such high shear stress, after 60 min ultrasonic washing. Both water and n-hexadecane droplets still exhibited high apparent contact angles (Figure 2c) and low roll off angles (Figure 2d) on the washed surfaces. The sustained superomniphobicity is further evidenced by the bouncing water and n-hexadecane droplets on the surfaces (Figure 2e and 2f, Movies S6 and S7). Our results indicate that the apparent contact angles slightly decrease and roll off angles slightly increase within the first 20 minutes of ultrasonic washing. This is due to the removal of some excess fluorinated fumed silica particles that are not firmly bonded with the adhesive layer. The contact angles and roll off angles remain unaltered after 20 minutes washing, which suggests that the mechanical interlocking between the particle layer and the adhesive layer is sufficiently strong to withstand ultrasonic washing.

We further assessed the mechanical robustness of our superomniphobic surfaces in an accelerated manner by conducting linear Taber abrasion tests. A 2.5 cm-by-5 cm superomniphobic surface, which was firmly attached to the spline shaft of a linear Taber abraser, was brought in contact with a 2500-grit sandpaper (see the Experimental Section). The normal load applied on the superomniphobic surface was ~140 g, which corresponds to an average pressure of ~1 kPa. The superomniphobic surface was abraded against the sandpaper at a constant velocity of 2.5 cm s⁻¹. During one abrasion cycle, the surface moved back and forth on the sandpaper with a total travel distance of 5 cm. Since both the superomniphobic surface and the sandpaper are textured, we anticipate that the local shear stress experienced by the superomniphobic surface is much higher than 1 kPa.^[38] Our results indicate that the apparent contact angles decrease (Figure 3a) and roll off angles increase (Figure 3b) with the increasing abrasion cycles for both water and n-hexadecane. This gradual deterioration of superomniphobicity is induced by the loss of fluorinated fumed silica particles in the process of abrasion. Within 200 abrasion cycles (i.e., the travel distance of the surface on the sandpaper is 10 m), the roll off angles of water and n-hexadecane remain less than 10°, which indicates the mechanical robustness of our superomniphobic surfaces. The sustained superomniphobicity is due to the interconnected hierarchical porous structure of fumed silica particles and the strong micro/nano-mechanical interlocking between the particle layer and adhesive layer. The hierarchical structure of particle layer consists of a coarser texture (i.e., micron-scale agglomerates of interconnected nano-sized silica particles) and a finer texture (i.e., nano-size silica particles). During the abrasion tests, the hierarchical texture is gradually damaged (Figure 3c–3e). As the abrasion increased to 500 cycles (i.e., the travel distance of the surface on the sandpaper is 25 m), further deterioration of surface texture (Figure 3e) resulted in increased roll off angles (Figure 3b). However, even after 500 abrasion cycles, both water and n-hexadecane are completely removed from the surface, without any residue (Movie S8). It is worth noting that higher normal load or lower grit sandpaper leads to more rapid deterioration of superomniphobicity. When our surface was abraded against the 2500-grit sandpaper under a normal load of ~600 g (corresponding to an average pressure of ~5 kPa), it completely lost repellency against n-hexadecane (i.e., n-hexadecane droplet can no longer roll off from the surface) after ~45 abrasion cycles (see Supporting Information, Section 4). When our surface was abraded against a 320-grit sandpaper under a normal load of ~140 g (corresponding to an average pressure of ~1 kPa), it lost superomniphobicity after ~25 abrasion cycles (see Supporting Information, Section 4).

Figure 3.

(a) and (b) Apparent contact angles and roll off angles, respectively, of water and n-hexadecane droplets (~8 μl) on the superomniphobic surface after linear Taber abrasion tests. SEM images showing the morphology of (c) as-prepared superomniphobic surface, (d) after 200 abrasion cycles, and (e) after 500 abrasion cycles.

In summary, we developed a simple two-layer design strategy that enables the fabrication of robust superomniphobic surfaces. The interconnected hierarchical porous texture of fluorinated fumed silica particles and the micro/nano-mechanical interlocking between the particle layer and the polyurethane based adhesive layer resulted in the improved mechanical robustness of our superomniphobic surfaces, which was assessed through water shear test, ultrasonic washing test, blade scratching test, and Taber abrasion test.

Experimental Section

Functionalization of fumed silica particles:

Fumed silica particles (600 mg of 7 nm, Sigma-Aldrich) were dispersed in hexane (20 ml) contained in a glass vial. Heptadecafluoro-1,1,2,2-tetrahydrodecyl trichlorosilane (0.6 ml, Gelest) was then added into the particle suspension. The glass vial was sealed and placed on a shaker at room temperature for three days to obtain a suspension of fluorinated fumed silica particles.

Fabrication of robust superomniphobic surfaces:

The 2.5 cm-by-5 cm glass substrates were thoroughly washed with acetone, ethanol, and deionized water in sequence. The substrates were then dried using nitrogen. A polyurethane-based adhesive (shear strength and tensile strength are ~17–24 MPa and ~27–41 MPa, respectively, Gorilla Glue, Inc.) was spin coated on the cleaned glass substrate. The thickness of the adhesive layer was tuned by adjusting the spin coating speed (see Supporting Information, Section 1) and was measured using an optical profilometer (Zygo Zescope). The suspension of fluorinated fumed silica particles was sprayed onto the adhesive layer using an air brush (Paasche) immediately after the spin coating. The spray pressure and distance between the surface and the air brush were 30 psi and 15 cm, respectively. Particle suspension (30 ml) was used to prepare the particle layer of each sample. Subsequently, the surface was stored at room temperature for one day to ensure the complete curing of the adhesive layer. No further treatment was involved in the fabrication of the superomniphobic surfaces.

Contact angle and roll-off angle measurements:

The contact angles and roll-off angles of liquid droplets on the fabricated surfaces were measured using a contact angle goniometer (Ramé-Hart 260). The advancing contact angle and receding contact angle were measured by increasing and decreasing, respectively, the volume of a droplet on the surface using a syringe (Gilmont). The roll-off angles were measured by slowly tilting the surface until the droplet (~ 8 μl) rolled off from the surface. At least five measurements were performed on each surface.

Morphology Characterization:

The surface morphology was imaged using a scanning electron microscope (SEM; JEOL 6500F) at 5 kV.

Bouncing of droplets:

High-speed movies of bouncing droplets were imaged using a high speed camera (Photron Fastcam SA3) at 2000 frames per second.

Assessment of mechanical durability via abrasion tests:

A Taber linear abraser (Taber Industries, Model 5750) was used to conduct abrasion tests. The abrasion tests were performed at 2.5 cm s⁻¹ with 2500-grit sandpaper under a pressure of ~1 kPa.