Introduction

Superhydrophobic surfaces that mimic lotus leaves have been extensively studied over the past several decades, and various micro/nanostructures have been developed to create bio-inspired superhydrophobic surfaces.1, 2, 25 Practical applications of non-wetting surfaces, however, have been impeded by the limited stability of their underwater superhydrophobicity.26, 32 it was found that the MP structures were not sufficient for the formation of a continuous gas interlayer, as only 70% of the layer was regenerated. Thus, a modified model of the hierarchical ZnO NR/Si MP structures needed to completely regenerate the underwater superhydrophobicity was developed.

The modified equation, derived in Supplementary Information 6, defines the lower and upper bounds for stable regeneration of underwater superhydrophobicity.

where L is bare Si MP pitch, H is bare Si MP height, d is ZnO NR length, φ is the gas fraction, and θp and θb are the water CAs of the ZnO NRs and the bottom surface between the MPs, respectively. This model gives the NR length (d) required to regenerate a stable gas interlayer in water and presents lower(dmin)/upper(dmax) bounds, which are shown in Figure 5a. The water CA was fixed to 152° for the superhydrophobic ZnO NRs (θp) and to 114° for the hydrophobic Si MPs (θb).

Figure 5
figure 5

(a) Boundaries of the ZnO NR length required to allow gas bubbles to fill the interfacial region of the surface and form a stable gas interlayer theoretically calculated using equation (1). ● denotes the experimental results that involved the successful formation of a stable gas interlayer from the generated gas bubbles. All three ● data points are within the theoretically acceptable zone. X denotes the results of experiments in which a stable gas interlayer was not formed from the generated gas bubbles. All three X data points are outside the theoretically acceptable zone. (b) Three-dimensional schematic image of the ZnO/Si HRs. Inset image of b depicting magnified the ZnO NRs on top of the Si MPs in the ZnO/Si HRs. (c) Relative intensity transitions of the ZnO/Si HRs for several wetting and dewetting cycles.

Full size image

Figure 5a shows those conditions in the plot of the NR length (d) vs the gas fraction (φ). The lower bound of d (dmin) in Figure 5a was derived from the minimum ZnO/Si HRs height/pitch ratio required to keep the water meniscus from contacting the bottom surface of the substrates. When MP pitch is 30 or 50 μm and MP height is 50 μm for the bare MPs, the d value is greater than the lower bound. The growth of ZnO NRs on the Si MPs results in an increase in the ZnO/Si HRs height/pitch ratio, and thus, d is always greater than the lower bound for the ZnO/Si HRs. However, when MP pitch is 100 μm and φ is greater than 0.9, it is possible that d might not be greater than the lower bound, which is plotted as a black line in Figure 5a. When the NR length is shorter than the lower bound, the generated gas bubbles cannot grow, and they rupture.

The upper bound of d (dmax) in the model was derived to ensure that the growth of the gas bubbles in the lateral direction is more favorable than that in the normal direction, meaning the lateral movements of the hydrogen bubbles are faster than the vertical movements. The upper bounds on d for MP pitch=30, 50 and 100 μm are plotted as red, green and blue lines, respectively, in Figure 5a. The upper bound shifts upward as the pitch increases because a wider pitch allows the gas bubbles to spread along longer ZnO NRs. In addition, the pressure against lateral growth decreases with increasing pitch. When the ZnO NR length exceeds the upper bound, the gas bubbles do not grow laterally and instead escape from the surface due to decreased spacing between MPs. Depending on the MP pitch, the d values that allow gas bubbles to be captured and retained fall between dmin and dmax.

The ZnO/Si HRs experimental data collected for various NR lengths and MP pitches are included in Figure 5a to test the proposed model for regenerating underwater superhydrophobicity. When ZnO/Si HRs with a NR length of 8 μm and MP pitches of 30, 50 or 100 μm were used, the underwater superhydrophobicity was successfully regenerated, and, thus, the corresponding data points are marked with an O. They fall within the acceptable zone of the proposed model. By contrast, when ZnO/Si HRs with a longer NR length of 16 μm were employed, the underwater superhydrophobicity could not be regenerated, and, thus, the corresponding data points are marked with an X. These data points are located in the gas bubble diffusion zone (outside the acceptable zone for each MP pitch condition). These results demonstrate that the proposed model satisfactorily predicts the regeneration of underwater superhydrophobicity for the experimental conditions tested in this study. It should be noted that the ZnO NRs played a crucial role in forming a stable, continuous gas interlayer because the MP structure with no NRs had water on its surface, and the gas bubbles formed between each post could not merge to form a continuous gas layer. The ZnO NRs in the ZnO/Si HRs provided attachment points for the formation of a continuous gas interlayer, and the Laplace pressure of the ZnO NRs aided in the removal of the water layer on the surface (inset image of Figure 5b).49 The details about Laplace pressure on the ZnO NRs are presented in the Supplementary Information 7.

Figure 5c shows the changes in the relative intensity for the superhydrophobic ZnO/Si HRs over five wetting (losing the gas interlayer) and dewetting (regenerating the gas interlayer) cycles. The wetting process required a few days due to the high stability of the underwater superhydrophobicity, whereas regeneration occurred within a few seconds. Reproducible trends and constant maximum intensity values were observed, meaning that the underwater superhydrophobicity was completely regenerated by the PEC reaction over several cycles. Also static water CAs of each state during the wetting–dewetting cycles were measured (Supplementary Figure S7). These results demonstrate the reproducibility of our regeneration method.

Conclusion

In summary, the lotus leaf effect (superhydrophobicity) and artificial leaf effect (solar water splitting) were combined for the first time to develop a method for completely regenerating the underwater superhydrophobicity of ZnO/Si hierarchical surfaces using photocatalytic hydrogen generation. The rationally designed n/p junction in the ZnO/Si HRs allowed for high gas interlayer stability in water as demonstrated by a long plastron lifetime and also highly reproducible regeneration of the underwater superhydrophobicity during cycling due to the unique ability of the surface to capture and retain a stable gas layer. A model was developed to define the optimal geometry of the surface structures for the formation of a stable, continuous gas interlayer on the superhydrophobic surface. The proposed model was consistent with our experimental results. This study presents a new possible solution to the continuing problem of the limited stability of underwater superhydrophobicity and broadens the applicability of superhydrophobic surfaces.

scheme 1

An illustration of underwater superhydrophobicity regeneration system.