Spontaneous dewetting transitions of droplets during icing & melting cycle

Abstract

Anti-icing superhydrophobic surfaces have been a key research topic due to their potential application value in aviation, telecommunication, energy, etc. However, superhydrophobicity is easily lost during icing & melting cycles, where the water-repellent Cassie-Baxter state turns to the sticky Wenzel state. The reversible transition during icing & melting cycle without external assistance is challenging but vital for reliable anti-icing superhydrophobic performance, such a topic has rarely been reported. Here we demonstrate a spontaneous Wenzel to Cassie-Baxter dewetting transition during icing & melting cycle on well-designed superhydrophobic surfaces. Bubbles in ice droplets rapidly impact the micro-nano valleys under Marangoni force, prompting the continuous recovery of air pockets during melting processes. We establish models to confirm the bubbles movement broadens the dewetting conditions greatly and present three criteria for the dewetting transitions. This research deepens the understanding of wettability theory and extends the design of anti-icing superhydrophobic surfaces.

Fig. 1. Schematic illustration of fabrication strategies and the topologies of four types of micro-nanostructured surfaces.

a Double-scale periodical microcones with dense nanoparticles (denoted as MCNP) by laser. b Single-scale periodical microcones (SMC) by laser. c Double-scale random microbumps with dense nanoparticles (MBNP) by laser. d Irregular micro-nanostructure by chemical etching (IMN). SEM images with different magnifications e of MCNP, f of SMC, g of MBNP, and h of IMN. Scale bars are marked. i Surface roughness of four surfaces. j Contact angles (CA) and sliding angles of four surfaces at room temperature (15 °C and a humidity of 20%). Contact angles of four surfaces are marked. Data are mean ± s.d. from at least three independent measurements.

Fig. 2. States, contact diameters, and contact angles (CA) of droplets on the four hydrophobic surfaces during icing and melting processes.

a–c On the SMC surface. d–f On the IMN surface. g–i On the MBNP surface. j–l On the MCNP surface. Orange dashed box marks the recovered air pockets in j. Blue represents the icing processes while orange denotes the melting processes. Black dashed lines mark icing zones and melting zones. Circulation directions are presented by blue arrows (icing direction) and orange arrows (melting direction). The CA 150° boundaries of superhydrophobicity are indicated with black dashed lines. Scale bars and temperatures corresponding to different states are marked. The scale bars of a, d, g and j are 500 μm, 500 μm, 500 μm, and 300 μm, respectively. The fluctuations of the correlation evaluations can be found in the Methods section.

Fig. 3. The contact diameter recovery rate (CDRR), the contact angle recovery rate (CARR), contact angle, and sliding angle changes of droplets on the four hydrophobic surfaces after the icing & melting cycle.

a and b CDRR and CARR of droplets on the four surfaces after icing and melting. c Comparison of the original contact angles (CA), contact angles after one cycle, and the contact angle differences on the four hydrophobic surfaces. d Comparison of the original sliding angles (SA), the sliding angles after one cycle and the sliding angle differences on the four hydrophobic surfaces. The CDRR (δ_r) and CARR (δ_a) represent the CB state recovery extents of droplets. They can be calculated by ${\delta }_{\mathrm{r}}=1-\frac{D_{\mathrm{f}}-D_0}{D_0}$ and ${\delta }_a=1-\frac{{\theta }_0-{\theta }_{\mathrm{f}}}{{\theta }_0}$, where D_f and θ_f, D₀ and θ₀ denote the final contact diameter and contact angle after melting, the original contact diameter and angle, respectively. Schematics are inserted in a and b. The CA 150° boundary of superhydrophobicity is indicated with the black dashed line. Sliding angles of four surfaces after one cycle are marked. Data are mean ± s.d. from at least three independent measurements.

Fig. 4. Icing process of a droplet on the MCNP surface.

a Schematic illustration of icing process. b Direct observation from the side view. Proof of bubbles forming in the icing process is shown via the magnified picture. Orange dashed lines indicate the icing front. Bubbles and air pockets are marked with orange dashed boxes and red dashed boxes, respectively. c Direct observation from the top view. The magnified observations from the top view for the disappearance of the underneath air pockets are shown in the 1st, 2nd, and 3rd pictures. The zones with lighter contrast in the magnified pictures indicate air pockets, while those with darker contrast indicate the pinned zones. d Proof of bubbles movement. The bubble is marked with an orange dashed box. The moving directions of bubbles are marked with red arrows. e Observations for the original air pockets at 15 °C and ice penetration in the micro-nanostructure at −15 °C. Ice permeation modes are marked with blue arrows. f The changes of contact diameters and angles of the droplet on the MCNP surface with the temperature decrease. Three stages of the icing process are marked by black dashed lines, which indicate the cooling and condensation stage, icing stage and frosting stage, respectively. Scale bars, times and temperatures corresponding to different states are marked. The scale bars of b–d are 300 μm, 300 μm, and 70 μm, respectively. The fluctuations of the correlation evaluations can be found in the Methods section.

Fig. 5. Melting process of an ice droplet on the MCNP surface.

a Schematic illustration of melting process. b Direct observation from the side view. The orange dashed line indicates the melting front. Blue curved arrow represents the inclining direction of the ice droplet. Red arrows indicate the bubble movement directions. Bubbles and air pockets are marked with orange dashed boxes and red dashed boxes, respectively. c Direct observation from the top view. The magnified observation for the bottom air pockets is shown in the 4th picture. d Proof of bubbles downward movement and squeezing into the bottom micro-nano valleys sequentially. e Magnified observations for the recovery of the bottom air pockets with the increase of temperature. A schematic of bubbles movement during the melting stage is depicted. f The changes of contact diameters and contact angles of droplets with the temperature increase. Three stages of the melting process are marked by black dashed lines, which are the warming stage, melting stage, and stabilizing stage, respectively. Temperatures, times and scale bars corresponding to different states are marked. The scale bars of b–d are 300 μm, 300 μm, and 100 μm, respectively. The fluctuations of the correlation evaluations can be found in the Methods section.

Fig. 6. Analysis of the interfacial thermal resistances R_i in different wetting modes and experimental results for icing delay time and melting delay time on the four surfaces.

a Phase diagram of interfacial thermal resistances R_i in different wetting modes. The local magnified diagram is inserted in the center, and different wetting modes are marked. b Schematic showing different wetting modes and their equivalent thermal resistances. Droplet temperature (T_d), substrate temperature (T_s) and different equivalent thermal resistances are marked. c Median icing delay time on the four surfaces. The icing delay time is recorded from the beginning of the icing process (15 °C) to the recalescence. The side views of ice droplets on the four surfaces are presented. The bubbles of ice droplets are also marked with orange dashed boxes. d Median melting delay time on the four surfaces. The melting delay time is recorded from 0 °C to the end of melting. The side views of ice droplets during melting processes are also presented. Melting fronts are marked with orange dashed lines. For each type of surface, 10 repeated experiments are conducted to guarantee the accuracy of the results.

Fig. 7. Theoretical analysis for the energy changes during dewetting transitions.

a Three energy states of droplets on different surfaces. b Original phase diagram describing different energy states on surfaces with different micro-nanostructures. Background colors correspond to the distributions of each state, respectively. Yellow denotes the monostable CB zones while orange and blue denote the metastable Wenzel and metastable CB zones, respectively. The black dashed line represents the superhydrophobic conditions, which are described in Supplementary Fig. 15 and Supplementary Discussion 4. The optimal zone for surface design is emphasized by the dashed arrows. The experimental result in this work is marked with the purple point. c Phase diagram after bubbles impacting. It is corresponding to the condition that bubbles contribute to the recovery of 3% of the total air pockets volume (ε_r = 0.03). The extended optimal zone under bubbles impact is emphasized. It can be observed that the droplets on the MCNP surface become the monostable CB state under bubbles impact.

Fig. 8. Effects of different microcones heights on the recovery of CB state.

a State changes of droplets on the MCNP surfaces with different microcones heights after an icing & melting cycle. b Morphology evolutions of micro-nanostructure with different microcones heights. By adjusting laser processing parameters (Supplementary Table 2), we achieve precisely tunable heights and pitches of the micro-nanostructure on the MCNP surfaces. Two kinds of surfaces are fabricated: the surfaces with different microcones heights and constant pitch and the surfaces with different microcones pitches and constant height (Supplementary Figs. 20–22 and Supplementary Method 2). In the figure, microcones pitch is fixed at 35 μm while microcones heights vary from 35 μm to 55 μm. c–e Effects of different microcones heights on the contact diameter recovery rates (CDRR), the contact angle recovery rates (CARR), contact angles (CA), and sliding angles (SA), respectively. The CA 150° and SA 10° boundaries of superhydrophobicity are indicated with the black dashed line. Data are mean ± s.d. from at least three independent measurements.

Fig. 9. Effects of different microcones pitches on the recovery of CB state.

a State changes of droplets on the MCNP surfaces with different microcones pitches after an icing & melting cycle. b Morphology evolutions of micro-nanostructure with different microcones pitches. Microcones height is fixed at 45 μm while microcones pitches vary from 25 μm to 45 μm. c–e Effects of different microcones pitches on the contact diameter recovery rates (CDRR), the contact angle recovery rates (CARR), contact angles (CA), and sliding angles (SA), respectively. The CA 150° and SA 10° boundaries of superhydrophobicity are indicated with the black dashed line. Data are mean ± s.d. from at least three independent measurements.

Fig. 10. Robustness tests of the W2C transition.

a The $\frac{D}{V^{\frac{1}{3}}}$ changes of the same droplet on the same location with the temperature decrease after 1–5 cycles. It corresponds to the icing process. D represents the contact diameters of droplets while V represents the droplet volumes. b The $\frac{D}{V^{\frac{1}{3}}}$ change of the same droplet on the same location with increasing temperature after 1–5 cycles. It corresponds to the melting process. c Evolutions of surface micro-nano structures after multiple cycles. The formed nanopits are noted by the red circle. Cycle times and scale bars are marked. d Evolutions of contact angles and sliding angles of the same droplet after multiple cycles. e The changes of the contact diameter recovery rates (CDRR) and the contact angle recovery rates (CARR) of droplets on the MCNP surface with the increase of cycle times. Data are mean ± s.d. from at least three independent measurements.