In Situ Optical Observation of Air-Mattress Recovery on Submerged Salvinia Leaf

Prior to each experiment, a liquid pressure of 6.8 atm was applied to collapse the air mattress on submerged Salvinia leaves. A digital camera was then employed to observe the recovery process. Here we used a syringe to inflate air into the wetted structures (Fig. 1 E and F). As soon as the air was inflated, a silvery film was visible on the leaf surface due to the reflection of light at the water–air interface, indicating that a thin air film emerged to cover the entire base of the leaf surface. As the replenished air volume increased, the air mattress grew thicker and thicker until the microstructures were completely filled with air (reflected by the gradual increase in silvery brightness; Movie S1), which exhibits a superwater-repellent capability. In addition to air inflation, other air replenishment methods, like regulating the pressure of surrounding water, can also recover the collapsed air mattress (SI Appendix, Fig. S1A). It is worth noting that, beyond superhydrophobicity, this superrepellency is still valid under high liquid pressures up to 7 atm in our experiments. The air-mattress recovery can be even achieved in flow with Reynolds number up to 5,000 (SI Appendix, Fig. S1B). Thus, this superrepellency is robust and expected to be able to endure extreme conditions. Moreover, the complete recovery of the air mattress is independent of the location of air inflation (SI Appendix, Fig. S1 C–E), and consequently, tiling a few Salvinia leaves together conduces to a large area of air mattress recovered over all of the leaves (SI Appendix, Fig. S1F). In addition, identical experiments were performed on a lotus leaf as control observations. In contrast to the Salvinia leaf, discrete bubbles, instead of a continuous air mattress, formed on the lotus leaf (SI Appendix, Fig. S2). As both Salvinia and lotus leaves are superhydrophobic, the superrepellent capability of water on Salvinia leaves is attributed to the unique hierarchical microstructures on the leaf surface, which are revealed below.

Microstructures on Salvinia Leaf Surface and the Details of the Air-Mattress Recovery at Microscale

The Salvinia leaf surface is covered with a dense hairy forest (Fig. 2A) owning three geometrical characteristics: 1) interconnected microgrooves on the epidermis, 2) long hair stems, and 3) eggbeater-shaped heads. The length of the hairy structures (including the hair stems and the eggbeater-shaped heads) l and that of the adjacent distance d are hundreds of micrometers, which are much smaller than the capillary length. Microgrooves covered with nanowax crystals (21) are formed at the joint of the neighboring convex epidermal cells and connected over the entire epidermis (Fig. 2 B and C). A confocal image in Fig. 2D shows the cross-sections of the microgrooves whose morphology can be approximated as a wedge with a half corner angle α = 17.2○ ± 6.1○. The measured equilibrium contact angle on the surface of the microgrooves θ = 146.1○ ± 7.8○ (see SI Appendix, Figs. S3 and S4 for details of the microstructure properties). Note that this large value of contact angle is due to the residual air that remained in the nanostructures to achieve superhydrophobicity (detailed explanations in SI Appendix).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

(A) SEM image of the S. molesta surface. The microstructures of the Salvinia surface consist of three parts, i.e., the eggbeater-shaped heads, the hair stems, and the ellipsoidal epidermal cells covering the base of the leaf. From the leaf edge to the center, the hair length l varies from 500 μm to 2,400 μm and the adjacent hair distance d varies from 350 μm to 850 μm. (B) Higher magnification of the yellow box in A. The hairy structures and the base of Salvinia surface are covered with epidermal cells. Interconnected grooves are formed between the adjacent cells. (C) A 3D confocal image of the epidermal cells and the microgrooves. (D) Cross-section of microgrooves at the red dotted box in C, showing the profile of the wedge-shaped grooves with the half corner angle α = 17.2○ ± 6.1○. (E) Air spreading along the grooves indicated by the brightness increase of the water–air interface in the dashed box. (F) Movement of the TCLs indicated by the movement of points 2 and 3 up along the hair stems, leading to the growth of the air mattress. (G) Pinning effect at the eggbeater. The eggbeater pins the early arrived TCL at point 5 to wait for the arrival of the later TCL at point 4, which ensures that the air mattress is completely recovered. (H) Schematics of three stages during air-mattress recovery. The numbers label the same positions as those in E, F, and G.

We zoomed in at the microscale to observe the recovery of the air mattress in detail (SI Appendix, Movie S2). Three stages corresponding to the three geometrical characteristics were captured, as shown in Fig. 2 E–G. First, immediately after inflation (Fig. 2E), air expanded spontaneously through the interconnected microgrooves until all microgrooves were fully filled with air. As the inflation was continuously conducted, the replenished air spilled from the microgrooves to form a thin air film covering the entire base of the leaf surface. The dashed box in Fig. 2E highlights the process of air expansion along one groove (point 1) by tracking the silvery reflection on the water–air interface. Second, based on the thin air film covering the leaf surface, the three-phase-contact lines (TCLs) on different hair stems slid vertically upward synchronously along the hair stems (movements of points 2 and 3 on different TCLs in Fig. 2F), which indicates an increase in the thickness of the entire air mattress. Third, due to the nonuniform sizes of the hairy structures, some TCLs arrived early and pinned (21) at the top of the hairy structures (point 5 in Fig. 2G), waiting for the other TCLs to continue to slide along the hairy structures (point 4 in Fig. 2G), until the air mattress was completely recovered. To show the process of air-mattress recovery on the Salvinia surface more clearly, the three stages of recovery are summarized in the schematics in Fig. 2H.

Mechanism of Air-Mattress Recovery

In principle, the realization of air-mattress recovery requires two consecutive steps, i.e., 1) the formation of seed air and 2) the spreading of air within the structures. The process of air-mattress recovery on Salvinia leaves can be characterized as the vertical growth of the continuous seed air film synchronously along the hairy structures, in which the transport of the replenished air in the wedge-shaped grooves to the entire base of the leaf surface creates the seed air film and the hairy structures (including the hair stems and the eggbeater-shaped heads) provide a frame for the spreading of air. The mechanism for air-mattress recovery on Salvinia leaves is addressed below.

A stable full expansion of air in wedge-shaped grooves is the premise to the recovery of the air mattress as the formation of the seed air will significantly influence the spreading process. A thermodynamic free-energy model (24) was used to analyze the stability of the air column in a wedge-shaped groove by introducing perturbations on the water–air interface (see Fig. 3A for the schematics and SI Appendix, SI Text for the detailed theory). The results show that the air column is stable for any perturbation (“full expansion” in Fig. 3B) when the half corner angle of the wedge (α) and the contact angle (θ) satisfyα+(π−θ)<π/2.[1]When Eq. 1 is not satisfied, the air column will lose its stability and dissociate into pieces (“semiexpansion” in Fig. 3B). Moreover, Eq. 1 demonstrates a gas wicking effect which is an inverse process of the classical interior-corner wetting phenomena depicted by the Concus–Finn (CF) condition (i.e., α + θ < π/2 and “no expansion” in Fig. 3B) (25⇓–27). Therefore, three regimes can be determined to describe the stability of the air column, i.e., full expansion (Eq. 1), no expansion (the CF condition), and semiexpansion (the unstable states in between), as shown in the phase diagram in Fig. 3C. Artificial wedge-shaped grooves with different corner angles and contact angles fabricated by 3D printing were used to verify the theory (Fig. 3 E–G), which shows a good agreement with the phase diagram (see SI Appendix, Figs. S5 and S6 for the properties of the artificial wedge-shaped grooves). For the Salvinia leaf with α = 17.2○ ± 6.1○ and θ = 146.1○ ± 7.8○, the experimental data corresponded to the mode of full expansion (Fig. 3C), which reveals the mechanism of the stable expansion of air in the interconnected microgrooves.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

(A) Schematics of the cross-section of the air column in the wedge-shaped groove, where the two straight lines represent the wedge corner, the arc represents the water–air interface, α is the half wedge corner angle, θ is Young’s contact angle, r is the radius of the arc, and φ is the azimuthal coordinate. (B) Schematics of air full expansion, semiexpansion, and no expansion along the wedge-shaped groove. (C) Phase diagram of the air expansion modes. The dots correspond to the experimental results in E, F, and G. (D) The air volume that can be inflated into the microstructures as a function of the duration of plasma treatment on the dried Salvinia surface. The dots are the experimental results from H, I, and J. (E–G) The 3D confocal images of the three modes of air expansion along the wedge-shaped groove corresponding to those in B. The blue color indicates the bulk water and the red color represents the interfaces. (H–J) Effect of wettability on the air-mattress recovery. The recovery capability of the air mattress decreases as the decrease of surface hydrophobicity which is tuned by the increase of the duration of plasma treatment t. t = 0, 21, and 30 s in H, I, and J, respectively. (K–M) The 3D confocal snapshots and 2D cross-sections of the local morphology of the water–air interface, showing the three modes of air expansion along the wedge-shaped grooves in H, I, and J, respectively.

To further verify the importance of full expansion of air in wedge-shaped grooves to the air-mattress recovery on the real Salvinia leaves, fresh leaves were dehydrated with a critical point dryer, which can maintain both microstructures and hydrophobicity unchanged, followed by a modification of the wettability using oxygen plasma treatment. As the surface hydrophobicity decreases (indicated by the increase of the plasma treatment time in Fig. 3D), the recovery capability of a continuous air mattress decreases (Fig. 3 H–J), which is reflected by the volume decrease of the inflated air in the microstructures in Fig. 3D. Confocal microscopy is employed to observe the local details of the water–air interface morphology in different inflated conditions, also demonstrating the three modes of air expansion along the wedge-shaped grooves (Fig. 3 K–M). The experimental results give strong evidence that full expansion of air (Fig. 3K) leads to a complete recovery of the air mattress (Fig. 3H), whereas semiexpansion (Fig. 3L) and no expansion (Fig. 3M) result in the semirecovery (Fig. 3I) and unrecovered conditions (Fig. 3J), respectively, which verifies the significance of the air expansion through the wedge-shaped grooves on the air-mattress recovery. Essentially speaking, the full expansion of air is achieved by the gas wicking effect, which ensures an efficient and spontaneous transport of air through the interconnected wedge-shaped grooves and makes full use of the replenished air to rapidly cover the entire base of the surface to form a continuous air film.

Based on the thin air film formed by the wedge-shaped grooves, the hairy structures will provide a frame for the air mattress to grow in thickness. As the hairy structures intersect the water–air interface, the resistance distributes along the circumference of the hairy structures in the vertical direction (Fslidepost in SI Appendix, Fig. S7). When the air pressure in the air mattress pair overcomes the sum of hydrostatic pressure pwater and the vertical resistance Fslidepost, the TCLs will start to slide vertically along the hairy structures. On the one hand, if the sizes of the hairy structures are ideally uniform, the entire water–air interface will rise uniformly until the air mattress is completely recovered. On the other hand, if the sizes of the hairy structures are nonuniform, some TCLs will arrive early and pin at the top of the shorter hairy structures to wait for the other TCLs to continue to slide along the longer hairy structures vertically until the air mattress is completely recovered (SI Appendix, Fig. S7), in which the maximum pinning force Fpin,maxtop at the top of the hairy structures should be larger than the sliding resistance Fslidepost to maintain a stable water–air interface. The pinning effect of the eggbeater-shaped heads can greatly increase Fpin,maxtop, which enhances the robustness of the air-mattress recovery. Therefore, no matter whether the hairy structures are uniform or not, the air-mattress recovery will proceed smoothly, indicating that the requirement for the configuration and distribution of the hairy structures is flexible (see SI Appendix, Fig. S8 for the air-mattress recovery on the other two species of Salvinia with different hairy structures). The hairy structures provide only a frame for the air mattress to grow, while the air-mattress recovery is dominated by the features of the wedge-shaped grooves on the base of Salvinia leaves. In contrast, if the air-mattress recovery is governed by the mechanism of regulating the hairy structures, a geometric constraint of hairy structures is requisite (15). Our investigations, however, demonstrate that the mechanism of air-mattress recovery on Salvinia leaves allows the hairy structures to grow beyond the geometric limitation (proved by the experimental results in SI Appendix, Figs. S9 and S10), which greatly extends the geometric design of microstructures for air-mattress recovery.

Air-Mattress Recovery on Artificial Salvinia Surface: Lessons from Salvinia Leaf

Inspired by the nature of Salvinia, we fabricated artificial Salvinia surfaces using 3D printing technology (Fig. 4A) (28). The artificial surface imitates the main features of the unique hierarchical structures on Salvinia leaves, i.e., 1) the wedge-shaped grooves on the base for the formation of seed air film and 2) the hairy structures for the spreading of air. The base of the artificial Salvinia surface was covered with half-cylindriform bulges to mimic the convex epidermal cells. Wedge-shaped grooves were formed between the adjacent cells, with α = 10○ and θ = 156○ ± 5.2○, which satisfied the full expansion condition (Fig. 3C). The artificial hairy structures are fabricated to support the air mattress growing in thickness (Fig. 4A), which includes the hair stem and the eggbeater-shaped head, with the hairy structures height la = 69 μm and the distance between two neighboring hairy structures da = 66 μm. Confocal microscopy was used to observe the recovery of the air mattress inflated by a microinjection system (Movie S3). The complete recovery of a continuous air mattress was successfully realized to exactly imitate the superrepellency of Salvinia leaves, as shown in Fig. 4 B and C from 2D and 3D perspectives, respectively. In addition, a control specimen was fabricated with the same structures as that of the biomimetic Salvinia surface but without microgrooves on the base (Fig. 4D). As expected, only bubbles formed instead of a continuous air mattress under the same experimental conditions (Fig. 4E). Moreover, experiments on artificial specimens with different hairy structures were also performed to further demonstrate that the requirement of the air-mattress recovery for the configuration and distribution of the hairy structures is flexible (SI Appendix, Figs. S9–S11).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

(A) SEM image of the 3D printed artificial Salvinia surface, duplicating the three main geometrical characteristics of the microstructures on Salvinia leaf, i.e., wedge-shaped grooves, hair stems, and eggbeater-shaped hair heads. Inset shows the details. (B) The 2D confocal images sequentially showing the air-mattress recovery by air inflation on the biomimetic Salvinia surface. The microscope was focused on the plane of the top of eggbeater heads. The blue color indicates water and the black color indicates that water is replaced by air. (C) The 3D confocal images sequentially showing the recovery process on the biomimetic Salvinia surface, including the initial state of air-mattress collapse, the air spreading along the wedge-shaped grooves, and the complete recovery of the air mattress. (D) SEM image of the control specimen with the same structures as those in A but without microgrooves on the base. Inset shows the details. (E) The 2D confocal images sequentially showing the formation of bubbles on the control specimen in contrast to a continuous air mattress on the biomimetic Salvinia surface. Inset shows the focal plane on the base of the specimen, which confirms the tip of the needle is placed deep into the microstructures to guarantee the direct contact of the inflated air with the specimen surface.