High-speed Videography of Water Droplets Impacting Polydimethylsiloxane Microarrays Reveals Interesting Behaviour at Small Scales
The wetting of microstructured flat surfaces is affected both by the properties of water and the properties of the surface, namely the arrangement, shape, height and surface area of the micropillars. Microstructured surfaces as models are studied extensively to understand how topographical or chemical heterogeneities influence wetting phenomena. Such phenomena include super-hydrophobicity, ‘fakir’ droplets, electrowetting, and more. The same apparatus in the Dynamic Microfluidics laboratory can also be used to study the wetting of surfaces other than polydimethylsiloxane, such as metals, leaves, crystals and other polymers.
Research | Physics| Aimee Lew
Water, in scales large and small, in speeds high and low, remains an active focus of physical inquiries. Dynamic microfluidics, the study of water at small scales and high speeds, sheds important insights on the interaction between fluids and surfaces at the microscopic level. The University of Auckland’s Dynamic Microfluidics laboratory conducts various research projects on water behaving with plastic, crystalline, metallic and organic surfaces, as well as milk-drying experiments in air. In 2022 I experimented with water droplets on samples of polydimethylsiloxane (PDMS). Exactly how water wets different surfaces, and thus how a surface becomes hydrophobic or hydrophilic, has important implications for materials sciences, chemical engineering, and food sciences.
Background
It is entirely possible to put a drop of water on both a hydrophobic (water-repelling) and hydrophilic (water-attracting) surface and have it stay there. How then, is hydrophobicity and its counterpart quantified? In fluid dynamics, wetting and wetness is defined through the contact angle that water makes with a surface. The contact angle is measured from the flat surface along the tangent of a water droplet at the point of contact. A surface with a contact angle less than 90° is hydrophilic. A surface with a contact angle greater than 90° is hydrophobic.
There are two ways a surface’s hydrophobicity can be altered: chemical or structural. Frequently these two techniques are used in tandem with each other to design surfaces with manifold wetting properties, fit for manifold purposes. Chemical hydrophobicity comes from using substances like oils, waxes, acrylics, and other polymers that repel water by the very nature of their chemical makeup. Structural hydrophobicity comes from patterning and texturing a substance so that the equilibrium contact angle (the contact angle when water sits on a flat surface with homogeneous surface chemistry), and thus the wetting behaviour, changes. Whether water droplets spread or not on rough surfaces can be reduced to energy considerations, namely from surface tension (surface area of the droplet) and gravity (mass of the droplet). Roughening a surface, all other things equal, increases the surface area of the material by introducing hills and valleys, divots and folds. If a flat surface was originally hydrophilic, then as the surface area increases, water spreads even more. If a flat surface was originally hydrophobic, then an increased surface area makes spreading/wetting less energetically favourable. Subsequently, roughness intensifies the original behaviour of a material. Hydrophobic surfaces become more hydrophobic, and vice versa.
Figure 1: Diagram of water droplets on four different surfaces, with their contact angles marked. Proceeding from left to right, the surfaces are superhydrophilic, hydrophilic, hydrophobic, and superhydrophobic. Artwork by the author.
Methods
In my experiments, water droplets are dropped onto slides of polydimethylsiloxane (PDMS). The independent variables are the microstructure of the array and the height of droplet release. The dependent variable is the wetting behaviour, with droplet volumes and release heights controlled electronically. The water droplets are dispensed onto the PDMS sample by an automatic syringe pump controlled by a computer programme. The volume of each droplet was kept constant at five microlitres. The syringe mount height is adjustable by raising or lowering, by computer, its vertical stage. High-speed videography records the behaviour of the water immediately after impact.
A lateral camera (foreground) records a side view of a falling droplet. A second camera (right) focuses on a mirror which directs overhead light, yielding a bottom view of a falling droplet. A green laser light backlights both cameras to increase the contrast between a light background and a dark water droplet. Photron FastCam Viewer is the software used to capture and process the high-speed footage. Trials were conducted at low droplet heights (less than 20 cm above the PDMS sample) to limit their acceleration under gravity. At slower droplet speeds, the effects of splashing and motion blur are mitigated.
Figure 2: Experimental apparatus. Photography by the author.
The ‘rough’ surfaces used are PDMS micropillar arrays. Micropillar arrays are surfaces with posts in a regular grid, made using lithography. The dimension of the grid is determined by properties such as the arrangement (the length and width between each post; the array can be square, rectangular, even triangular); the shape of the posts (the top of the posts can be circular, square, hexagonal); the height or ‘pitch’ or the posts; and the surface area Φ of the posts. Microstructured surfaces such as these are used to model different combinations of surface chemistry and topography. In experiments they reveal numerous wetting phenomena on heterogeneous surfaces, like super-hydrophobicity [1], ‘fakir’ droplets [2], and electrowetting [3].
Two samples of PDMS were used. The shape of the pillar cross-section in this experiment is square. Both samples have 21.2μm pitch (pillar height) and surface area Φ = 20μm. They are both microarrays with pillars arranged in a rectangular grid. The M2A sample is a rectangular 40μm x 60μm array. The M2B sample is a rectangular 60μm x 80μm array. Samples were cleaned with water, dried with nitrogen gas, and stored in sealed containers.
Figure 3: Schematic of rectangular array with circular micropillars. Artwork by the author.
In the microscopic world of these arrays, numerous combinations and designs are possible, each presenting a new line of inquiry for researchers. Adjusting the properties of the grid, pillars, or material can bring about new surfaces with new wetting behaviour. The wetting of microarrays and its applications are widespread and ever-developing. It builds better understanding and control over how water interacts with different materials at the microscopic and nanoscopic level. Applications are interdisciplinary, spanning biomedical engineering, materials science, nanotechnology, and more. For example, self-cleaning surfaces utilise hydrophobic materials, while probes are deposited on DNA chips through miniscule and precisely-placed hydrophilic spots [5]. While wetting in different regimes continues to be explored, turn a new eye to raindrops on the car windshield, or shower spray on the wall. There’s a lot of potential in things so tiny. Thanks to Geoff Willmott for his supervision and guidance.
Results
Conclusion
The images displayed were obtained with the M2B sample, which best demonstrated the wetting phenomena. Increasing the height from which the water droplets are released increases the droplet’s inertia to surface tension ratio (also called the Weber number). The spread factor, a ratio of the droplet’s original diameter to its diameter at the maximum spread, increases with increasing height. At the origin of the impact, a microbubble of air forms as incoming water compresses and traps the ambient gas. Fringing this origin point is an ‘impact region’ — appearing as darker, dense spots — which also traps air between the pillars of the microarray, forming a partially-wetting area that often takes on geometric shapes [4].
Another feature observed is the formation of fingers from droplets released at larger heights, and thus travelling faster upon impact, along one axis of the rectangular array. See Figure 3(m). This follows from the fact that in a rectangular array, pillars are clustered more closely in one direction than the other. Water travelling perpendicular to the axis with more tightly-spaced pillars faces more resistance, and while water travelling along this axis can extend into lobes and fingers at the edge of the droplet. Another interesting observation were triangular wetted regions at the border of the droplets at intermediate release heights. See Figure 3(i), where triangular shapes appear at the top and bottom of the droplet’s profile. These regions also only appear along the axis that the fingers form, which indicates they could be a result of the same fluid velocity and fluid pressure differences.
Figure 4: Radial images of water droplet impacts. Down the vertical axis, images advance through time. Across the horizontal axis, the release height of the droplet increases. Image by the author.
Aimee Lew - BSc/BA, Physics, Politics, International Relations, Chinese
Aimee is a Creative Officer and the resident illustrator for the Scientific. She is a fourth-year BA/BSc student majoring in Physics and has previously researched in optics, microfluidics, and physical acoustics projects. In 2023 she will be on exchange at the University of California, Los Angeles.