Droplet bouncing on topological nonwetting surfaces via laser fabrication

Abstract

Purpose

The purpose of this paper is to show that the droplet impact phenomenon is important for the advancement of industrial technologies in many fields such as spray cooling and ink jet printing. Droplet bouncing on the nonwetting surfaces is a special phenomenon in the impact process which has attracted lots of attention.

Design/methodology/approach

In this work, the authors fabricated two kinds of representative nonwetting surfaces including superhydrophobic surfaces (SHS) and a slippery liquid-infused porous surface (SLIPS) with advanced UV laser processing.

Findings

The droplet bouncing behavior on the two kinds of nonwetting surfaces were compared in the experiments. The results indicate that the increasing Weber number enlarges the maximum droplet spreading diameter and raises the droplet bounce height but has no effect on contact time.

Originality/value

In addition, the authors find that the topological SHS and SLIPS with the laser-processed microwedge groove array produce asymmetric droplet bouncing with opposite offset direction. Microdroplets can be continuously transported without any additional driving force on such a topological SLIPS. The promising method for manipulating droplets has potential applications for the droplet-based microfluidic platforms.

Citation

Zhuang, K., Xiao, J. and Yang, X. (2022), "Droplet bouncing on topological nonwetting surfaces via laser fabrication", Journal of Intelligent Manufacturing and Special Equipment, Vol. 3 No. 2, pp. 192-203. https://doi.org/10.1108/JIMSE-05-2022-0008

Introduction

Droplet impact is an extremely common phenomenon in nature such as stone erosion and natural rainfall (Tropea and Marengo, 1999; Josserand and Thoroddsen, 2016). It is also easily found in engineering applications like water jet processing, spray cooling, ink jet printing, etc. (Khan et al., 2010; Hu et al., 2021; Cheng et al., 2016). Understanding the droplet impact phenomenon on different surfaces is important for the advancement of industrial technologies in these fields, as the droplet impact can involve complex and different solid–liquid–air interactions depending on the environmental and interfacial conditions (Xu et al., 2005; Yarin, 2005). The possible outcomes of droplet impact on a solid surface were first summarized by Rioboo et al. (2001), which were classified into these six categories – deposition, prompt splash, corona splash, receding break-up, partial rebound and complete rebound. Droplet bouncing is a relatively special phenomenon in the above context because of the stringent requirements on solid surface wettability, morphology, droplet characteristics and even environmental conditions. With the development of 3D micro-nano imaging techniques such as scanning electron microscopy in recent decades, the principle of the lotus leaf effect has been revealed and the micro-nanotextured superhydrophobic surfaces (SHS) inspired by the lotus leaf have been prepared by various surface treatment techniques (Balu et al., 2009; Mertaniemi et al., 2011; Mumm et al., 2009). Water droplets on SHS are in a classical Cassie wetting state with triple-phase interfaces in which the air pockets were trapped in the rough micro-nano structure. Thus, the apparent contact angle of water droplets on such surfaces exceeds 150°, and the falling droplets can roll off the SHS with significant bouncing which lead to a small roll angle of less than 2°. Owing to the small energy dissipation caused by the underlying trapped air pockets, droplet impact on SHS results in elastic rebound, with two parts of droplet deformation including spreading and retraction; during the former process the kinetic energy of the droplet is transformed into surface energy and stored; when the maximum spreading diameter is reached the droplet retracts and reverts to a sphere, bouncing upward and releasing surface energy. The droplet bounce phenomenon on SHS has attracted wide attention from researchers due to the great potential in numerous biological and chemical applications such as self-cleaning, anti-icing, thermal devices and microfludics (Bhushan and Jung, 2011; Yu et al., 2021; Liu et al., 2019; Asmatulu et al., 2011; Feng et al., 2008; Liu and Kim, 2014; Liu et al., 2015; Liu et al., 2014).

However, SHS have some insurmountable problems in practical applications such as high adhesion due to high temperature and high pressure environments where droplets on SHS can change from Cassie to Wenzel wetting state, and general SHS is difficult to resist physical damage due to the poor mechanical properties. The researchers have, therefore, proposed another nonwetting surface called slippery liquid-infused porous surface (SLIPS) that mimics pitcher plants (Wong et al., 2011), which also has a micro-nanotextured substrate but differs from SHS in that the SLIPS is infused with a low surface energy lubricant into a rough substrate forming a stable lubricant layer that can exclude various immiscible liquids (Yong et al., 2017; Huang et al., 2019). Due to the extremely low shear sliding resistance between the lubricant layer and the surface droplets, the droplets on the SLIPS also have an extremely low sliding angle of less than 2°. The promising SLIPS is capable of resisting ice adhesion and self-cleaning, and even rapidly self-healing cracks caused by physical damage (Li et al., 2021; Zhang et al., 2018; Jiao et al., 2019; Dai et al., 2015). Recent research has made progress in the study of droplet bouncing dynamics on the SLIPS. Liu et al. found that the complete rebound phenomenon similar to that of droplets on SHS occurs on the SLIPS. And they concluded that the thin lubricant film interface on the SLIPS leads to a difference in droplet impact behavior from that of SHS (Hao et al., 2015). Despite the impressive progress in droplet bouncing research based on SHS and the SLIPS, few studies have been conducted to visually compare and differentiate droplet impact phenomena on these two surfaces under the same experimental conditions.

In this paper, SHS with clustered spherical micro-nanotextured substrates were prepared by a simple method combined with advanced UV laser processing and chemical surface modification, then further infused with lubricant to fabricate SLIPS. The impact dynamic behavior of water droplets on SHS and the SLIPS including the maximum spreading diameter, contact time and bounce height were compared under different conditions of Weber number. The key parameters of the different droplet impact behaviors were also calculated and theoretically analyzed. Further processing of asymmetric microwedged groove array topologies on SHS and the SLIPS revealed asymmetric bouncing behavior of droplets with opposite direction, which helps to further understand the different static and dynamic wetting characteristics of droplets on SHS and the SLIPS. In addition, droplets on the topological SLIPS with wedge-shaped groove array can achieve stable and spontaneous transport without mass loss for engineering applications such as precise microdroplet manipulation, medical detection and enhanced condensation heat transfer.

Materials and methods

Surface fabrication

The whole surfaces of prepolished copper sheets (30 mm× 30 mm ×2 mm, purchased from Suzhou Metal Material Manufacturer (China)) were first milled by using a UV laser marking machine (KY-M-UV3L, Wuhan Keyi, China) with high-speed of 1,500 mm s⁻¹. The laser intensity of ∼12.0 J cm⁻¹ was held constant during the ablating process. Then, the samples were immersed in 1 wt% ethanol solution with fluoroalkylsilane [FAS, C₈F₁₃H₄Si(OCH₂CH₃)₃] for 90 min to lower the surface energy and the SHS can be obtained after 30 min drying at 120 °C. The homogeneous SLIPS can then be obtained by infusing 3M Fluorinert^TM FC-40 (density ρ = 1.85 g cm⁻³) on the as-prepared SHS which was tilted at ∼45° and held still for 3 h to drain away extra floating oil 3M Fluorinert^TM FC-40 (surface tension γ = 16 mN m⁻¹) purchased from Shenzhen China Fluorine Technology Co., Ltd. The thickness of the 3M Fluorinert^TM FC-40 was ∼20 μm, which was controlled by precision balance.

Characterization

Micro morphology of as-prepared surfaces was analyzed using a scanning electron microscope (SEM, Ultra-60, Zeiss, Germany) and a digital microscope (VHX-600, Keyence, Japan). Static and dynamic contact angles were characterized with a goniometer (Rame-Hart 290, USA). Sliding angles were measured with a vertically mounted rotary table. Images and videos of the droplet impact process were taken by a high-speed digital camera with 10,000 ftp recording function ((i-SPEED 726R, iX Cameras, UK)). The mass of remaining lubricant was measured with the precision electronic balance (JB5374-91, Mettler Toledo, Shanghai).

Results and discussion

Construction of nonwetting surfaces (SHS and SLIPS)

There are two keys to construct an SHS: one is a rough substrate with a micro-nano structure, and the other is to modify the substrate with a low surface energy. A nanosecond pulsed UV laser was used in following experiments to perform a rapid ablation for the whole surface on the metallic copper sheet as shown in Figure 1a, which is a simple and efficient texturing technique to build a large-area homogeneous substrate with micro-nano structure. The scanned copper surface undergone a series of physical structuring process of melting, evaporation/boiling and phase explosion (Figure 1d) (He et al., 2020; Yoo et al., 2000). Then, liquid copper sputtered by phase explosion eventually recondensed into a dense layer of spherical particulate structure (Figure 1c) which could suspend the droplets on the surface with a small liquid–solid interface. The homogeneous SHS could be obtained after the low surface chemical modification, on which the apparent contact angle and sliding angle of 10 μL water droplet are relatively 169° and 2° (Figure 1a and b). To fabricate a SLIPS, the prepared SHS was infused with 3M Fluorinert^TM FC-40 and a stable thin oil layer generated on the upper surface of the substrate. A 10 μL water droplet dispensed on the on the as-prepared SLIPS had an apparent contact angle of 95° with a large footprint and a sliding angle of 2° due to an ultralow detaining force between the liquid–liquid contact interface (Figure 1a and b). As seen in Figure 1c, the dense layer of spherical particulate structure can improve stability of Cassie wetting state on the SHS, while, here can provide enough capillary force for the lubricant to imbibe the textures and generate a uniform lubricant layer (Xue et al., 2006). Although water droplets have excellent dynamic sliding properties on both SHS and the SLIPS (with a sliding angle of only 2°), significant distinctions were shown in static wettability due to the contact interface being solid–liquid and liquid–liquid respectively. The difference of the apparent contact angles on SHS and the SLIPS was nearly 74°, so that the liquid–liquid contact droplet footprint on the SLIPS is large and the droplets are as seemed cap-shaped while on the SHS resemble spherical shapes suspended on pinpoints, with an actual small solid–liquid contact surface. It can be seen in Figure 1e that the 10 μL water droplets formed arrays of “NU” and “AA” on SHS and the SLIPS, respectively. The SLIPS had an oily sheen, and the droplets have an overall cap shape, unlike the almost complete spherical shape on the SHS, which led to the different droplet bouncing phenomena in the following experiments.

Droplet bouncing on the nonwetting surfaces (SHS and the SLIPS)

To investigate the droplet impact dynamics on the SHS and SLIPS, the prepared SHS and SLIPS were placed on a horizontal platform and the droplets were released from a predetermined height by a microinjector and then impacted vertically on the surfaces, while the entire water droplet impact process was recorded at 10,000 fps with a high-speed camera. Different impact phenomena such as complete bouncing and partial bouncing were investigated in detail for different Weber numbers. Weber number is a dimensionless quantity used to compare kinetic energy with surface energy, which is defined as follows:

(1)We=ρv02d0γ

where ρ, γ is the density and surface tension of the liquid droplet. The initial diameter of droplet d₀ was always 2.7 mm in the following experiments while the initial impact velocity v₀ was dependent on the released height according to the following free-fall equation:

(2)v0=2gh0

where g and h₀ are gravitational acceleration and released height of droplet.

As seen in Figure 2a and b, the droplets impacted on the SHS and SLIPS with significant complete rebound at the condition of the initial impact velocity v₀ = 0.5 mm s⁻¹ and We = 9. The whole impact process can be divided into three stages: spreading, retraction and bouncing. During the spreading process on SHS, the droplet expanded radially in stacked layers, with the bottom edge spreading outward until the droplet relaxed to an equilibrium state in the shape of a pancake and reaching a maximum diameter D_max of ∼4.0 mm within the time of 3.6 ms. Here, the concept of spreading factor β was introduced which can be expressed as follows:

(3)β=Dmaxd0

As seen in Figure 2c, all the data in the process of spreading and retraction collapsed into a curve, showing the spreading factor β as a function of contact time. In the retraction process, β decreased and returned to 0 upon rebound. Correspondingly, the pancake-shaped droplet then released the surface and internal kinetic energy converted from the initial kinetic energy (Clanet et al., 2004), and the boundaries on both sides began to contract toward the center, eventually detaching from the SHS at 14 ms to perform a complete rebound. The main dissipation on SHS and the SLIPS during droplet spreading progress occurs between the droplet and the unescaped air cushion at the bottom (Hao et al., 2015; de Ruiter et al., 2015), so that the spreading phenomena and maximum spreading diameters of droplets (We = 9) are similar on SHS and the SLIPS. However, as the droplet reaches the D_max = 4.0 mm on the SLIPS and started to retract, the droplet needed detach away from the oil film interface to undergo viscous dissipation of E_μ (Keiser et al., 2020) while the SHS is nonviscous and the droplet overcomes much less resistance during retraction than E_μ. Therefore, the droplet detached from the SLIPS at 19 ms which is longer than that on the SHS. The viscous dissipation between the liquid–liquid interface acted as a damper, dragging the droplets out of a long tail during droplet retraction on the SLIPS and resulting in a lower bounce height of H_max = 3.7 mm, compared to that of 6.4 mm on SHS (Figure 2a and b).

The droplet impact process on the SHS and SLIPS was shown in Figure 3a and b when the conditions of v₀ = 1.0 mm s⁻¹ and We = 37 were applied. As the initial impact velocity increased, the droplet spreading phase became intense, with the droplet quickly compressing as a bowler hat shape at 1 ms. The droplet, then, was squeezed into a thin pancake shape, eventually reaching a maximum spreading diameter of 6.23 mm and 6.27 mm on the SHS and SLIPS, respectively. During the retraction process, the droplet on the SHS retracted to form a baseball bat and started to recede upward as a whole due to the low adhesion caused by the strong water repellency on the surface. While the droplet on the SLIPS was stretched in a nearly cylindrical shape at the bottom because of the high viscous force between water–lubricant interface until it breaks away at 20.8 ms eventually the turbulence within the droplet during the bouncing process contributed to the multiple splitting behavior on the top of the stretched droplet, with a droplet on the SHS bouncing up to the H_max = 11.75 mm that was higher than that of 6.0 mm on the SLIPS. Note that, the contact time τ was independent of We during the droplet impact process on the SLIPS, which was consistent with previous studies on SHS. The contact time depends on the inertia and capillarity forces of the droplet, which was interactions with the contact surface and dissipitation (Muschi et al., 2018). It was the water–lubricant contact interface on the SLIPS contributed to a contact time of nearly 20 ms greater than the 13 ms caused by the water–solid contact interface on the SHS (Figure 3c).

However, the different phenomena of droplet impact on the SHS and SLIPS was shown in Figure 4a and b when the impact droplet was at the condition of We = 233. With the impact initial velocity of 2.5 mm s⁻¹, the droplet spread in a dome-like straw hat at 0.6 ms with a fragmentation tendency, then followed by an intense splitting of many small droplets at the outer edge on the SHS, and eventually the droplet completely crashed into numerous microdroplets. While on the SLIPS the droplet remained relatively intact after obvious splashing when the droplet expanded to its maximum diameter. The damping effect of the lubricant layer on the SLIPS increased the interfacial dissipation and dissipated the high initial kinetic energy of the impact droplet, which resulted in such different spreading processes. Broken droplets on the SHS, then, undergone partial retraction and lost significant bounce height due to the volume loss; the main droplet on the SLIPS performed a claw-like recombination and rebound. As seen in Figure 4c, the spreading maximum diameter of a 10 μL droplet increases with Weber number (under a threshold of We = 110 where the droplet fragmentation does not occur). The experiment results of D_max are corresponded to a classical scaling law as follows:

(4)Dmax∼d0We14

This means β∝We14, as shown in Figure 4d. The scaling law gives good quantitative agreement with the experiments’ data for the substrates of SHS and the SLIPS. Meanwhile, the increasing Weber number also contributes to enlarge another key experiment parameter, maximum bouncing height H_max on the SHS and SLIPS as seen in Figure 4e.

Droplet directional bouncing on the topological nonwetting surfaces with wedged array

Previous studies have revealed that the droplet presented asymmetric bouncing on the SHS with specific topological macrotexture (Liu et al., 2016). Interestingly, we found that the asymmetric bouncing behaviors were totally different with completely opposite offset direction on the topological SHS and SLIPS. To evaluate how the surface topology affects the droplet bouncing behavior, the copper sheets were preprocessed with a three-dimension array of microwedge grooves structure with laser ablation as shown in Figure 5a, with the single wedge groove measuring 15 mm long, 0.2 mm at wide end, 0.03 mm at narrow end and 0.3 mm deep. During the laser ablation of microwedge grooves, the cross-sectional profile of the groove resembles a Gaussian distribution curve, so that the narrow end of the wedge has a V-shaped cross-section, while the wide end has an inverted trapezoid shape in the presence of partially molten material (Figure 5b). At the same time there is also a layer of micro-nano spherical particle structure inside the entire groove surface similar to that in Figure 1c, so that the laser processed copper surface has a wide range of submicro-micro-nano composite structures. Then, the topological SHS can be obtained by chemically modifying the processed copper surface. A 10 μL droplet can be suspended on the microwedge groove array structure of the topological SHS to form a macroscopic Cassie wetting state (Figure 5c). When a 10 μL droplet with We = 9 impacted the center area of the microwedge groove array, multiple complete droplets bouncing similar to Figure 2a could be observed, but the difference was that each bouncing process performed a slight deflection toward the wide end of the microwedge groove. As seen in Figure 5d, the droplet had a slight offset Δx = 1.5 mm to the right after the second bounce compared to its initial position. While on the topological SLIPS infused with 3M Fluorinert^TM FC-40 the droplet bounced only once and moves to the narrow end of the microwedge groove with bumps. As seen in Figure 5e, the slight offset Δx = 1.2 mm was in the complete opposite direction to the offset in Figure 5d.

The asymmetric bouncing motion in the opposite direction of the offset was more pronounced in the droplet impact phenomenon with 0.3 mm at wide end for each single wedge groove. As seen in Figure 6a, a 10 µl droplet with the initial velocity of 0.5 m s⁻¹ impacted the center of the microwedge groove array on the topological SHS and the droplet bounced strongly and deflected to the right, followed by a secondary bouncing that decreased in height but still maintains an obvious rightward offset Δx = 4.6 mm. The leftward offset Δx = 2.8 mm occurred on the topological SHS with only once faint bouncing and a bumpy sliding process. Compared to the plain SHS, the microwedge groove array topology changed the contact interface upon droplet impact on the topological SHS. As seen in Figure 6b, the bottom of the droplet was trapped in the wedge grooves by the strong impact and the confined footprint was squeezed forming an asymmetric wedge shape. An internal Laplace pressure ΔP was produced in the impact droplet which can be expressed as follows (Xue et al., 2006):

(5)ΔP≈2γc(1Rt−1Rh)

where γc is the surface tension of the contact interface; Rt, Rh represent the radii of curvature at the tail and head of the squeezed footprint. The wedge-shaped tab contact area, due to the small actual solid–liquid interface of the droplet on the SHS, generated no radial driving forces. In contrast, the droplet impact surface on the topological SLIPS, the lubricant–droplet contact interface has a damper effect to alleviate the droplet groove zone extrusion; the main extrusion occurs at the lubricant–droplet contact interface of the wedge-bump surface. The squeezed footprints were, therefore, shown in Figure 6b, with the direction of the internal Laplace pressure differential force pointing toward the narrow end of the groove. Thus, the topologically induced Laplace pressure ΔP in opposite directions during the droplet impact on the SHS and SLIPS resulted in the asymmetric bouncing as shown in Figure 6a. Actually, the directional bouncing of droplets on the topological SHS is difficult to manipulate because droplets roll away from the surface very easily after multiple bouncing. However, on the topological SLIPS the directional droplet bouncing becomes weak due to the damping effect of the lubricant layer and can used for the stable directional droplet transport. As seen in Figure 6a and c a 10 µl droplet suspended from a microinjector can be continuously transported for 12 mm within 3.5 s on the topological SLIPS with the microwedge groove array.

Conclusion

We introduced and investigated the droplet bouncing behavior with different Weber number on the SHS and SLIPS. The experiment results indicate that the increasing Weber number enlarges the maximum droplet spreading diameter and raises the droplet bounce height but has no effect on contact time. Droplet contact time is shorter on SHS than on the SLIPS with the former having a larger maximum bouncing height. There is no difference in maximum spread diameter between the droplet bouncing on the SHS and SLIPS. In addition, the topological SHS and SLIPS with the laser-processed microwedge groove array produce asymmetric droplet bouncing with opposite offset direction, which was induced by the internal Laplace pressure of squeezed footprint. Microdroplets can be continuously transported without any additional driving force on such a topological SLIPS. The promising method for manipulating droplets has potential applications for the droplet-based microfluidic platforms.

The TOC graphic

Figures

Figure 1

Construction of nonwetting surfaces (SHS and SLIPS)

Figure 2

Complete bouncing of droplets with We = 9 on SHS and the SLIPS

Figure 3

Complete bouncing of droplets with We = 37 on SHS and the SLIPS

Figure 4

Crushing and splashing of droplets with We = 233 on SHS and the SLIPS

Figure 5

Droplet directional bouncing on the topological nonwetting surfaces with wedged array

Figure 6

Mechanisms of asymmetric droplet bouncing and droplet directional transport