🔬 physics

Droplet Spreading on Nanowall-Textured Hydrophobic Surfaces: A Nonlinear Reduced-Order Model

This paper presents a nonlinear one-degree-of-freedom reduced-order model that successfully captures the radial deformation and maximum spread factors of water droplets impacting both smooth and nanowall-textured hydrophobic silicon surfaces, demonstrating that the textured surface significantly reduces spreading compared to smooth surfaces while identifying nonlinear stiffness and pinning magnitude as the primary controlling factors.

Original authors: Mubarak Yakubu, Najib Aminu Ismail

Published 2026-10-05

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Mubarak Yakubu, Najib Aminu Ismail

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

When a drop of water hits a surface, it does not simply splash and stop; it performs a brief, complex dance of energy. The drop's forward motion crashes into the surface, flattening the liquid into a wide, thin disk before surface tension—the same force that makes water bead up—tries to pull it back into a sphere. Whether the drop spreads out wide or stays small depends on a tug-of-war between that initial speed and the resistance of the surface it hits. On a perfectly smooth, water-repelling surface, the drop can spread out significantly. But when that surface is covered in tiny, microscopic structures, like a forest of nanoscale walls, the story changes. These textures can trap air, pin the edge of the drop in place, or create extra friction, all of which alter how far the water spreads. Understanding this behavior is crucial for everything from designing self-cleaning windows to improving inkjet printing, yet predicting exactly how a drop will react to such intricate surfaces remains a difficult challenge for scientists.

In a recent study, researchers set out to simplify this complex physical problem into a manageable model. They focused on water droplets, specifically 20 microliters in volume, falling from a height of 20 millimeters onto two types of silicon surfaces: one that was smooth and hydrophobic (water-repelling), and another covered in nanowalls with a spacing of just 130 nanometers. The goal was not to simulate every single molecule of water or the exact shape of the air trapped between the walls, which would require immense computing power. Instead, the team built a "reduced-order" model. Think of this as creating a simplified map of a city that shows only the main roads and traffic lights, ignoring every side street and building, just to predict how long a trip will take. Their model treated the entire droplet as a single, moving point that could stretch and shrink, governed by a few key rules about how the surface resists that stretching.

The researchers first tested their model on the smooth surface. They found that if they treated the water drop like a simple spring that resists stretching in a straight line, the model predicted the drop would spread far too much, far beyond what experiments actually show. To fix this, they added a rule that the surface gets stiffer the more the drop stretches, much like a rubber band that becomes harder to pull the further you stretch it. With this adjustment, the model's prediction for the smooth surface matched the experimental data almost perfectly, showing a maximum spread factor of about 2.25. This means the drop's diameter grew to roughly 2.25 times its original size before it stopped expanding.

Next, the team applied this same model to the textured surface with the nanowalls. Here, they introduced a new element: a "pinning" force. This represents the microscopic walls grabbing the edge of the water drop, making it harder for the drop to spread. The model calculated that this extra resistance would significantly limit the spread. The result was a maximum spread factor of 1.810, meaning the drop on the textured surface stayed much smaller than on the smooth one. This outcome aligned with the general trends observed in previous experiments, confirming that the texture successfully held the drop back. The researchers also checked how the model behaved when they changed the drop's size or the height from which it fell. In every case, the model correctly predicted that the drop would spread further if it hit the surface harder, but it would always spread less on the textured surface than on the smooth one.

However, the authors were very clear about what their model could not do. Because they simplified the drop into a single moving point, their model cannot predict what happens after the drop stops spreading. It cannot tell us if the drop will bounce off the surface, how high it might jump, or exactly when it will lift off. It also does not account for the complex physics of air getting trapped under the drop or the water seeping into the tiny gaps between the nanowalls. The model is strictly a tool for understanding the maximum width the drop reaches during the initial impact. The researchers also noted that while their model worked well for the specific conditions they tested, the way they calculated the "pinning" force was an estimate based on a fixed reference, and it might not perfectly capture every nuance of different surface textures.

Ultimately, this work provides a clear, computationally cheap way to understand the limits of how far a water drop will spread on a rough, water-repelling surface. By stripping away the complex details of fluid dynamics and focusing on the main forces at play, the researchers created a reliable tool that captures the essential behavior of the system. They demonstrated that while the drop's speed and the surface's stiffness determine how far it spreads, the tiny texture of the surface acts as a brake, keeping the drop smaller. This insight helps scientists and engineers design better surfaces for specific needs, knowing that adding microscopic texture is an effective way to control how liquids behave, even if the full picture of the drop's motion remains too complex for this simple model to solve.

Technical Summary: Droplet Spreading on Nanowall-Textured Hydrophobic Surfaces

Problem Statement
Droplet impact on non-wetting surfaces involves a complex transient conversion of translational kinetic energy into interfacial energy, internal motion, and dissipation. While classical scaling laws (e.g., $\beta_{max} \sim We^{1/4}$ ) summarize maximum deformation trends, they do not uniquely identify how surface textures modify the underlying force and energy balances. Specifically, surface textures introduce competing effects: changes in equilibrium wetting geometry (Wenzel vs. Cassie–Baxter states), energy barriers that arrest the contact line (pinning), and potential liquid penetration into texture that alters dissipation and recoverable energy.

The specific setting addressed is the impact of water droplets on hydrophobized smooth versus nanowall-textured silicon surfaces, based on experimental data from Yilbas et al. [11]. The challenge is to determine the simplest radial dynamical system capable of reproducing the observed direction and condition dependence of maximum spreading without resorting to computationally expensive, fully resolved multiphase simulations (which would require resolving length scales spanning four orders of magnitude).

Methodology
The authors formulate a nonlinear one-degree-of-freedom (1-DOF) reduced-order model acting as a lumped radial surrogate. The model is not a resolved volume-of-fluid or phase-field simulation.

Governing Dynamics: The system tracks an effective radial extent $q(t)$ , with a reference value $q_0 = D_0/2$ . The equation of motion is:
$m\ddot{q} = -k_{cap}(q - q_0)\left[1 + \kappa_s \left(\frac{q - q_0}{q_0}\right)^2\right] - c_{visc}\dot{q} - F_{pin}\tanh\left(\frac{\dot{q}}{v_s}\right)$
- Mass ( $m$ ): Set equal to the total droplet mass.
- Capillary Stiffness ( $k_{cap}$ ): Fixed by the inertial–capillary timescale, not fitted.
- Nonlinear Stiffness: A cubic term ( $\kappa_s$ ) provides amplitude-dependent stiffening to prevent unrealistic compliance at large deformations.
- Viscous Damping: Modeled with a Reynolds-number-dependent coefficient.
- Kinetic Resistance: A regularized term opposes motion, approaching a magnitude $F_{pin}$ at high speeds.
Hybrid Pinning Rule: Since the regularized kinetic force vanishes at zero velocity (preventing static pinning), an explicit moving-to-stuck event is imposed. The trajectory terminates when the velocity drops below a capture band ( $v_{stick}$ ) and the capillary driving force is insufficient to overcome a static capacity threshold ( $\eta_s F_{pin}$ ).
Pinning Closure: The pinning magnitude $F_{pin}$ is estimated using a pore-scale force model scaled by a fixed reference encounter count. Notably, under the assumption that the hydraulic diameter equals the texture spacing ( $D_H = 2a$ ), the dimensionless pinning magnitude becomes independent of both spacing and droplet size.
Calibration and Validation:
- The cubic coefficient $\kappa_s$ is fitted to a single graphical data point (smooth surface, $V=20\,\mu$ L, $h=20\,$ mm).
- The textured surface response is predicted using fixed parameters, with no further fitting.
- Sensitivity Analysis: Local deterministic sensitivity is performed on closure parameters ( $\kappa_s, F_{pin}, \zeta_0, v_s$ ) and numerical settings to distinguish physical controls from numerical artifacts.

Key Contributions

Model Hierarchy: The paper establishes a hierarchy of models (linear unpinned, nonlinear unpinned, nonlinear pinned) to isolate the specific roles of nonlinear stiffness and pinning resistance.
Explicit Hybrid Dynamics: It introduces a mathematically stated moving-to-stuck rule to handle static pinning, acknowledging that smooth regularization alone cannot sustain a pinned state.
Identifiability and Transparency: The work explicitly classifies parameters as literature values, reconstructed inputs, fitted coefficients, or assumed closures. It demonstrates that agreement at one condition cannot validate all closures simultaneously.
Scope Definition: The model is rigorously bounded to predicting the maximum radial spread factor ( $\beta_{max}$ ). It explicitly excludes predictions of vertical rebound, contact time, wetting-state transitions, or physical lift-off.

Results

Base Case ( $V=20\,\mu$ L, $h=20\,$ mm):
- The linear unpinned model overpredicts spread ( $\beta_{max} = 3.955$ ).
- The nonlinear unpinned model (calibrated) matches the smooth surface target ( $\beta_{max} = 2.253$ ).
- The nonlinear textured model predicts a reduced spread ( $\beta_{max} = 1.810$ ), consistent with the qualitative trend of reduced spreading on nanowall surfaces observed by Yilbas et al.
Condition Transfer: With parameters fixed, the model predicts that maximum spread increases monotonically with the reconstructed Weber number ( $We$ ) for both surfaces. The textured surface consistently yields a lower $\beta_{max}$ than the smooth surface across the investigated volume and release height ranges.
Sensitivity:
- Nonlinear stiffness ( $\kappa_s$ ) and pinning magnitude ( $F_{pin}$ ) are identified as the principal controls on maximum spread.
- Pinning magnitude is the dominant control specific to the textured surface.
- Numerical parameters (solver tolerance, event localization, kinetic smoothing speed) have negligible effects on the predicted maximum spread at the reported precision.
- Endpoint Time: The time at which the trajectory terminates is highly sensitive to the static capacity factor ( $\eta_s$ ) and pinning magnitude near switching thresholds, indicating that endpoint time is a state-classification diagnostic rather than a robust physical prediction.

Significance and Claims
The paper modestly claims to provide a reproducible, computationally inexpensive radial surrogate that defines what can and cannot be inferred from available maximum-spread information.

It successfully reproduces the qualitative trend that nanowall textures reduce maximum spreading compared to smooth surfaces.
It clarifies that the reduction in spread is driven by the combined effects of outward pinning work and the hybrid static-capacity rule.
The authors emphasize that the model does not predict vertical rebound, normal restitution coefficients, or physical contact times. The "zero" radial recovery proxy for the textured case is a classification convention for radial non-return, not a prediction of zero experimental rebound height.
The work serves as an explicit baseline for future two-coordinate or resolved multiphase models, highlighting the limitations of single-coordinate approaches in capturing complex wetting-state transitions and energy recovery mechanisms.

Technical Summary: Droplet Spreading on Nanowall-Textured Hydrophobic Surfaces

More like this