Freezing-induced wetting transitions on superhydrophobic surfaces

Abstract

Supercooled droplet freezing on surfaces occurs frequently in nature and industry, often adversely affecting the efficiency and reliability of technological processes. The ability of superhydrophobic surfaces to rapidly shed water and reduce ice adhesion make them promising candidates for resistance to icing. However, the effect of supercooled droplet freezing-with its inherent rapid local heating and explosive vaporization-on the evolution of droplet-substrate interactions, and the resulting implications for the design of icephobic surfaces, are little explored. Here we investigate the freezing of supercooled droplets resting on engineered textured surfaces. On the basis of investigations in which freezing is induced by evacuation of the atmosphere, we determine the surface properties required to promote ice self-expulsion and, simultaneously, identify two mechanisms through which repellency falters. We elucidate these outcomes by balancing (anti-)wetting surface forces with those triggered by recalescent freezing phenomena and demonstrate rationally designed textures to promote ice expulsion. Finally, we consider the complementary case of freezing at atmospheric pressure and subzero temperature, where we observe bottom-up ice suffusion within the surface texture. We then assemble a rational framework for the phenomenology of ice adhesion of supercooled droplets throughout freezing, informing ice-repellent surface design across the phase diagram.

Keywords: Applied physics; Fluid dynamics.

Fig. 1. Freezing-induced droplet dynamics on superhydrophobic surfaces.

a–f, Synchronized side- (a,c,e) and bottom- (b,d,f) view image sequences of water droplets freezing through evaporative cooling in a dry, low-pressure environment with different outcomes. a,b, Impalement: penetration of the meniscus into the texture characterized by a low final contact angle and full substrate wetting (the dark area in b; red arrows illustrate the spreading direction of the penetrated liquid). The inset in a is a micrograph of the transparent superhydrophobic micropillar surface (scale bar, 100 μm). c,d, Expulsion: spontaneous de-wetting of the droplet (the receding contact line is marked by red arrows). e,f, Suffusion: freezing on top of the texture characterized by a high final contact angle, followed by volumetric expansion into the texture (the dark area in f; red arrows indicate the initial regions of substrate wetting). Scale bars: a,c,e, 2 mm; b,d,f, 500 μm. Employed surfaces: a,b, D6; c,d, D4; e,f, D1 (see Supplementary Table 1 for details).

Fig. 2. Microtexture topography and freezing characteristics alter freezing outcomes.

a, Schematic of a droplet resting on a superhydrophobic surface (not to scale) during recalescence, introducing β, s, d, h and the contact angle (θ). In the schematic, blue represents supercooled water and the grey area shows the progression of the freezing front. b, Φ versus s for water droplets in a low-pressure environment. Outcomes are differentiated by colour (red, impalement; blue, expulsion; green, suffusion) for two pillar heights (h = 25 µm and 40 µm). c, Bar chart of Φ for each s as a function of β (N = 249, n ≥ 19 experiments per data point). Employed surfaces: D1 to D6 (see Supplementary Table 1 for details). Source data

Fig. 3. Force scaling analysis of freezing-induced wetting transition mechanisms.

F_c/F_r versus F_r/F_a for the micropillar surfaces with constant diameter (D1 to D6), the micropillar surfaces with constant pitch (S1 to S4) and the spray-coated glass (C1) and mesh (C2), as well as the spray-coated micropillar surfaces (D1* and D6*) evaluated for the low-pressure conditions. The pie charts indicate the probabilities of the three different outcomes (red, impalement; blue, expulsion; green, suffusion), while the centre point of the pie charts locates the respective samples in the force ratio map evaluated for an initial water droplet volume of V = 10 μl. The background shading serves as a guide to the eye to identify regions of the three outcomes. N = 651, n ≥ 10. V = 10 μl for 461 experiments, while V ∈ [2 20] μl for the remaining 190 experiments. The effect of V is minor (see ‘Droplet size’ in the Supplementary Information and Supplementary Fig. 16). See Supplementary Tables 1 and 3 for further details. The inset is a schematic of a droplet on a micropillar surface with the dominant forces labelled. Source data

Fig. 4. Freezing on superhydrophobic surfaces at ambient pressure.

a,b, Synchronized side- (a) and bottom- (b) view image sequences of a water droplet freezing in a cold, dry environment at atmospheric pressure (red circles mark the approximate location of the contact line post-recalescence) on a superhydrophobic PDMS texture [d, s, h] = [10, 50, 40] µm (identifier D2 in Supplementary Table 1). c, Enlarged view between the pillars underneath the droplet for each timestep (for the regions of interest marked in b). d, Schematic of the bottom-up suffusion mechanism responsible for surface failure from condensation filling, coalescence (black arrow) and freezing. Water, ice slush and solid ice are represented by blue shading, light blue shading and hatching, respectively. Scale bars: a, 1 mm; b, 300 µm; c, 100 µm. All experiments (N = 24) showed bottom-up suffusion. Employed surfaces: D2 and D5 with pillar heights of 25 and 40 μm (n = 6) (see Supplementary Table 1 for details).