The etching process was expressed as previous work [45]. The whole fabrication process of this patterned slippery surface was shown in Figure 1. The transparent tape was functional for the protection to avoid pristine copper to be etched which resulted in a patterned surface, the yellow part reflected the pristine copper, and the black part meant the oxide copper after being etched. Figures 1(a)–1(c) show the detailed morphologies of the modified oxide copper and the section view of the patterned surface. The oxide copper showed a spherical shape, and the section view displayed a concave-convex shape. The oxide copper was in the lower part. Besides, the width ratio between the yellow and black parts is 1 : 1. Furthermore, the images of bubble and water contact angles on different types of surfaces (superhydrophobic surfaces, slippery surfaces, original surfaces, and PSS) were shown in Figure 2, Figure S1, and Table S1. In addition, the black part was modified to be superhydrophobic and aerophilic with the water and bubble contact angles of 155\pm 2^{°} and 60\pm 2^{°}; the pristine copper is hydrophilic and aerophobic of water and bubble contact angles of 80\pm 2^{°} and 130\pm 2^{°}. For the ideal surface, the water contact angle ({\theta }_w) in air and the bubble contact angle ({\theta }_b) in water complementary resulted from almost the same three-phase contact line. Additionally, the effect of buoyancy for bubble was analogous of gravity for water droplet. According to Young’s equation, the values of {\theta }_w and {\theta }_b can be calculated as

where {\gamma }_{\text{SV}}, {\gamma }_{\text{LV}}, and {\gamma }_{\text{SL}} represented the surface energy of solid-vapor interface, liquid-vapor interface, and solid-liquid interface, respectively. Based on equation (1) and (2), it is easy to derive {\theta }_b={180}^{°}-{\theta }_w. After lubricant PFPE was infused, the oil layer had substituted the solid substrate which resulted in the change of three phases. Therefore, the water (θ_wo) and bubble (θ_bo) contact angles were 110\pm 2^{°} and 53\pm 2^{°} with oil layer, which can be expressed as

where {\gamma }_{\text{OV}} and {\gamma }_{\text{OL}} are the surface energy of oil-vapor interface and oil-liquid phase. Besides, θ_wo and θ_wb were still complementary, {\theta }_{wo}+{\theta }_{wb}={180}^{°}. In addition, the patterned surface displays anisotropic wettability that resulted from the wettability difference between the yellow and black parts; the bubble contact angle is 126\pm 2^{°} and 70\pm 2^{°} in the x and y directions on the treated part with 1 mm width. From these results, it can be concluded that the treated part remains aerophilic for better bubble capture and transportation.

As for the whole process of bubble directional manipulation, the first step is capturing the bubbles. As shown in Figure 3(a), the pristine copper sheet cannot capture bubbles. Bubbles bounced to the upper position continuously to escape after it contacted with the inclined samples since the pristine copper is aerophobic. Moreover, the directional manipulation of underwater bubbles cannot be realized on the superhydrophobic oxide copper. From Figure 3(b), the silver mirror-like phenomenon was observed immediately on the superhydrophobic sample after being immersed in water; there was a bubble layer on the sample. Additionally, although the substrate can capture bubbles efficiently, bubbles stay in the original position to expand the air layer instead of moving upward, which was shown in Figure 3(b). Thus, it is hard to realize the directional transportation of underwater bubbles on superhydrophobic substrates, while in previous work the directional transportation of bubbles has been accomplished by a slippery surface with shape gradient structure [25].

Therefore, based on those researches, the slippery surface fabricated by the superhydrophobic substrate with infused oil layer was conducive to capture and transport bubbles [25, 34–37, 39]. In this work, we have fabricated patterned slippery surface (PSS) with different widths to achieve the goal of directional manipulation. As shown in Figure 4 and Movie S1, the PSS-2 was inclined in water to observe the bubble movement. Then, releasing a single bubble with the volume of about 15 μL in the slippery stripe, it was restricted by the boundary on the treated stripes and moved from the bottom up with the shape of ellipse (Figure 4(a)). Besides, releasing a single bubble with the same volume in the untreated stripe, the bubble bounced on the stripe to the upper location firstly, then it moved horizontally to the slippery area that resulted from the touch between the bubble and the slippery stripe. Then, the bubble transported to the top on the slippery stripe after horizontal movement (Figure 4(b)). From the single bubble motion on the PSS-2, it can be concluded that the combined surface can realize the directional manipulation of underwater bubbles on the treated stripes. Further, the same phenomenon had been observed in samples of PSS-1 and PSS-3 as shown in Fig. S2 and S3 with the same volume of bubble. Bubbles appeared in different shapes in treated stripes with different widths that resulted from the tendency that bubble would cover more area on the treated stripes. The transportations were also shown in Movie S1 and S2. However, there were some special phenomena in the directional manipulation in the samples PSS-1 and PSS-3 with the single bubble with the volume of 15 μL. As shown in Figure 4(c), the bubble covered three stripes (treated, untreated, and treated) during the whole upward transportation with the first contacted position of the untreated stripe, while in the sample of PSS-3 the bubble moved from the bottom up in the untreated part instead of a horizontal movement to the treated stripe to the top, which was shown in Figure 4(d). The special transportations were also shown in Movie S1. It can be ascribed to the relationship of bubble volume and stripe width. In PSS-1, the bubble tended to move horizontally to the treated part after being released on the untreated part, but the stripe width was much narrower compared to the bubble diameter that resulted in the touch between the bubble and two treated stripes. Then, the bubble and three stripes achieve a balance for directional movement. Nevertheless, the sample width of PSS-3 was too large for this bubble which resulted in the bouncing of the bubble from bottom up on the untreated stripe. From those movements of 15 μL bubble on the PSS samples, it could be concluded that the directional movement was mainly decided by the treated stripe which offered the capacity of capturing bubbles and the transport area. Furthermore, the horizontal movement depended on the contact of the bubble and the treated part. The treated part is also aerophilic with the oil layer for bubbles containing lower free energy than the pristine copper part that resulted in better capturing of the bubbles. Moreover, the added oil layer can make the substrate more regular and polisher to cover defects of this substrate. Based on this oil layer, bubbles can move entirely instead of bulk of movement with part of it remaining in the initial position under the driving force of buoyancy. In addition, the driving force for the bubble directional transportation included buoyancy (F_{\text{BP}}) and Laplace force difference (F_{\text{LP}}), and the resistance was mainly generated by the adhesive force (F_{\text{AD}}) and the sample defects (F_{\text{DE}}), which was shown in Figure 5. Furthermore, the buoyancy was generated by a part of the bubble which was the unbounded region, F_{\text{BP}} can be calculated as [46, 47]

where \rho is the density of water, g is the gravitational acceleration, and V_P is the effective volume of the bubble (the left part of bubble in Figure 5(b)). Moreover, there is a Laplace force difference (F_{\text{LP}}) generated by the curved bubble surface; it can be calculated as [47]

where {\gamma }_{\text{water}} is the surface tension of water, R_1 and R_2 are the radii of bubble in the front and back sides, and A_1 and A_2 are the projected area of the bubble at the front and back sides. In addition, in those inclined samples, the adhesion belonged to the lateral adhesion (F_{\text{LA}}) which its direction is against the directional transportation of bubbles on a solid substrate [46–49]. Thus, the resultant force (F_{\text{RT}}) in the moving direction can be calculated as

where \theta is the inclined angle of the surface in water. Based on equation (6), the movement was affected by several factors. F_{\text{BP}} and F_{\text{LP}} mainly depended on the inclined angle; F_{\text{LA}} had been reduced by the oil layer that resulted in the directional transportation. From above experiments, it can be confirmed that the PSS samples with straight treated stripes can realize efficiently the directional transportation of bubbles. Further, the samples also reduced the modified area in the total surface but with the same capacity of capturing bubbles compared to the whole modified surface which was beneficial for the environment. Hence, the patterned surface with the oil layer is an ideal candidate for capturing and transporting bubbles underwater.

Based on the fabricated method of the above samples with straight pathway, we also had fabricated the patterned slippery surface with the treated part displayed letter “C” (PSS-C). The width of the treated path was 4 mm. As shown in Figure 6(a), the bubble with a volume of 15 μL moved from the bottom up in a curve track which was the same as the treated path. The boundary line had played an important role in preventing the bubble escape from the treated path. Thus, the treated path on PSS samples offered the transported tracks for bubbles after bubble releasing. We also fabricated the patterned slippery surface with the treated part displayed letter “M” (PSS-M). There were some turning points in this sample. Therefore, the single bubble with the volume of about 15  μL was released on the PSS-M to detect the ability of directional transportation on this sample, which was shown in Figure 6(b) and Movie S2. The width of “M” was 3 mm, and the boundary lines were parallel and straight. The single bubble displayed a circular shape and moved along the treated path. Then, the bubble turned right and left when it reached the turning point of letter “M” in order to remain on the treated part all the time. Moreover, the shape of the bubble changed largely when going through the turning point. The transported track was also the same as the treated path, the letter “M.” With the letters “M” and “C” on the substrate, we can control the directional transportation of bubbles underwater more precisely in ideal directions and paths. Although those PSS samples (PSS-1, PSS-2, PSS-3, PSS-C, and PSS-M) could realize the directional transportations, bubbles could only move vertically from the lower position to the higher position driven by buoyancy force. Nonetheless, this facile method of fabricating the PSS samples offered the opportunity for spontaneous directional transportation of bubbles underwater. Thus, we fabricated the PSS samples with wedge shapes for spontaneous directional transportation of bubbles. There were four equal wedges on the copper sheet with the apex angle of 7°. As shown in Figure 7(a) and Movie S2, the bubble moved from the narrow tip to the wide end after being released on the wedge spontaneously. Besides, the bubble moved faster in the early time and slowed down to stagnation. Moreover, this PSS-wedge sample could also realize antibuoyancy movement. As shown in Figure 7(b) and Movie S2, the bubble moved downward spontaneously after being released on the inclined sample. The driving force on this sample of the PSS-wedge was the Laplace force difference (F_{\text{LP}}^{\prime }) no matter how the bubble moved. As shown in Figure 7(c), F_{\text{LP}}^{\prime } can be calculated as [34, 50]

where {\gamma }_{\text{oil}} is the surface tension of the oil layer, \alpha is the apex angle, V is the bubble volume, and r_1 and r_2 are the radii of the bubble in the tip and end positions. For the movement in Figure 7(b), the buoyancy became the resistance. Thus, the transported distance was much shorter than that in the inclined PSS-wedge sample. Hence, the patterned slippery surface can be designed with serval special shapes on it by covering the pristine copper sheet for further directional transportation of bubbles.

Bubble capture and collection have attracted much attention in recent years. Though there were many materials investigated for bubble collection, simplifying the fabrication methods for those materials was still a challenge. However, the PSS samples were fabricated via a facile etching method and it could efficiently capture bubbles and transport them along the treated area. Therefore, we designed a sample (PSS-Y) based on this method to achieve the goal of bubble collection. This sample also included two parts, the treated and untreated parts. As shown in Figure 8(a), the treated part was combined by two areas, the collect area and the transport area. Moreover, the transport area was 4 mm wide and 20 mm long which showed a rectangle, while another area showed a trapezoid with the upper and lower widths of 4 mm and 20 mm, respectively, and the height was 20 mm. The total process of bubble collection could be divided into three steps: attach on the collect area, move on to the transport area, and escape from this sample. We first examined the movement of two individual bubbles on this PSS-Y sample. As shown in Figure 8(b) and Movie S3, two bubbles with the volume of about 15 μL moved upward along the boundary line in the lower part. Then, the two bubbles merged into one when they moved to the joint between the straight and lower part. Finally, the merged bubble transported on the stripe to the end along the treated straight path. From the transportation of bubbles on the sample of PSS-Y, this special PSS samples could collect bubbles by the two parts. Based on the phenomenon of two bubbles on PSS-Y, we have released multiple bubbles on PSS-Y in different positions. From Figure 8(c) and Movie S3, all bubbles were captured by the collect area and moved along the treated part. Furthermore, bubbles might coalesce in the joint and move upward to escape this sample in the straight treated path. In this sample of PSS-Y, the trapezoid part had amplified the contact area between bubbles and sample which resulted in larger collect area, and the straight part restricted the leaving position of bubbles for bubble collection.

All reagents are obtained as listed. Pentadecafluorooctanoic acid (PFOA) was obtained from Saen Chemical Technology Co., Ltd.; hydrochloric acid (HCl) was obtained from Xilong Scientific. Ammonium persulfate (98%, (NH₄)₂S₂O₈), sodium hydroxide (96%, NaOH), ethyl alcohol, and acetone were purchased from Tianjin Rionlon Pharmaceutical Science & Technology Development Co., Ltd. Copper foils were supplied by Sinopharm Chemical Reagent Co., Ltd. The transparent tape was purchased from a local supermarket. The DuPont Krytox perfluoroalkylpolyether (PFPE, >95.0%) GPL 103 lubricant was purchased from the Chemours Company. The bubbles were generated by the syringe of air bubble. Deionized (DI) water was purified using a ModuPure system and used throughout the study.

Copper sheet (20\text{mm}\times 40\text{mm}) was cleaned several times by acetone, ethanol, dilute hydrochloric acid, and deionized water under consecutive ultrasonication for about 40 minutes. Put the tapes (1\text{mm}\times 40\text{mm}, 2\text{mm}\times 40\text{mm}, and 3\text{mm}\times 40\text{mm}) on the substrate line by line with the spacing of 1 mm, 2 mm, or 3 mm after drying the substrate in an oven at 60°C. Then, put the samples into a mixed etching solution of 100 mL (2 mol/L NaOH, 0.1 mol/L (NH₄)₂S₂O₈) for 3 h to obtain CuO on the uncovered surface. After that, peel the tape to obtain the patterned surface which was combined by pristine copper and etched copper. Dry the surface by N₂ and modify it by PFOA water solution (0.01 mol/L) to obtain superhydrophobicity of the treated part. At last, an amount of PFPE was applied on the patterned surfaces to realize the patterned slippery surfaces with different stripe widths (PSS-1, PSS-2, and PSS-3).

Field emission scanning electron microscope (FESEM, JSM-6701F) with Au-sputtered specimens was served to obtain detailed morphologies of the surface. The contact angles of water and bubble on serval samples were measured by a JC2000D goniometer (Zhongchen Digital Equipment Co. Ltd., Shanghai, China). All photographs and videos were captured by a mobile phone and a digital camera (Sony camera, DSCHX200). The average contact angles were measured in the same sample in five different positions.

Those bubble manipulations included two types, the buoyancy-driven and the self-driven. Moreover, the positions of those samples were different. The samples for buoyancy-driven were fixed underwater vertically with an inclined angle of 40°, while the others were horizontally placed in the container. Bubbles were generated by the syringes.