Introduction

In recent years, superhydrophobic microstructured surfaces, for example, lotus leaves [1], moth eyes [2], and so on [3,4,5,6], have become a topic of much interest based on their potential functions, such as self-cleaning [7,8,9], water manipulation [10,11,12,13], water separation [14, 15], and so on [16, 17]. There are two famous equations expressing superhydrophobicity: Wenzel’s equation [18] and the Cassie–Baxter equation [19]. Wenzel’s equation (1), applicable to wetted rough surfaces, is as follows:

$${\mathrm{cos}}\,\theta _{\mathrm{w}} = R\;{\mathrm{cos}}\,\theta,$$
(1)

where θw is the water contact angle (WCA) on a wetted rough surface, θ is the WCA on a flat material surface, and R is the roughness factor defined as the ratio of the true and apparent surfaces of the material. This equation simply indicated that WCAs are enhanced by surface roughness. Accordingly, fractal surfaces consisting of hydrophobic materials show remarkably large WCAs [20]. When a surface consists of two or more materials, the Cassie–Baxter equation (2) is adopted:

$${\mathrm{cos}}\,\theta _{\mathrm{c}} = f_1\,{\mathrm{cos}}\,\theta _1 + f_2\,{\mathrm{cos}}\,\theta _2,$$
(2)

Here, θc is the WCA of the Cassie–Baxter state, f1 and f2 are fractions of each surface composing the material, and θ1 and θ2 are the WCAs on flat surfaces of those materials. In most cases, for a superhydrophobic surface having a large surface area, one material is air (the WCA of air is regarded as 180°), and the Cassie–Baxter equation can be altered as shown below (3):

$${\mathrm{cos}}\,\theta _{\mathrm{c}} = - 1 + f_1\left( {{\mathrm{cos}}\,\theta _{1}+1} \right).$$
(3)

This equation indicates that a reduction in the water contact area (f1) is important for superhydrophobic surfaces and that, in many instances, spiky nano- and microstructures are prepared from various materials [21]. According to the equations, how to prepare a large surface area from hydrophobic materials is important for the preparation of superhydrophobic materials, and a large number of papers have been published [22,23,24,25,26].

As mentioned above, there have been many reports on artificial superhydrophobic surfaces; however, a number of unresolved problems remain. First, superhydrophobic materials are usually prepared from stiff and fragile materials, so it is difficult to apply them to daily use, as the surface nano- and microstructures are easily destroyed by contact, resulting in a loss of superhydrophobicity. Therefore, poly(dimethyl siloxane), which is a typical soft hydrophobic material, has been used due to its easy handling, processability, transparency, and flexibility [27,28,29]. However, although this silicone elastomer is a useful material, it is relatively expensive for daily uses or in paints and is also rather brittle without composite materials. We therefore have focused on vulcanized rubber, which is a common durable material used in tires, rubber bands, and so on [30,31,32]. Furthermore, the characteristics of vulcanized rubber can be adjusted through changes to its compositions by the inclusion of carbon black, silica, sulfur, and so on, and various compositions have already been developed for various purposes over its long history. Here, we first describe how to prepare vulcanized rubber microstructures as simply as possible and thereafter demonstrate their surface functions with a focus on the wettability generated by the microstructures [33]. Furthermore, we attempt to understand the changes in the wettability of the vulcanized rubber microstructures by theoretical analysis.

Materials and methods/experimental procedure

The unvulcanized rubber compositions used in this experiment are listed in Table 1. The unvulcanized rubbers were placed on single-crystalline silicon molds with finely patterned microstructures prepared by conventional photolithographic techniques [34, 35], then compressed at 5 MPa, and finally heated at 180 °C for 10 min (H300-15 Compact Hot-Press System, AS ONE Corporation, Japan) to vulcanize the material. After that, the vulcanized rubbers were simply removed from the silicon molds and observed using laser microscopy (OLS4000, Olympus Corporation, Japan) and a field-emission scanning electron microscope (FE-SEM, 5.00 kV, 80 µA, JSM-7800F, JEOL, Japan) after sputtering with Pt to a thickness of approximately 6 nm (30 mA, 80 s, JEC-3000FC, JEOL, Japan).

Table 1 Composition of the rubber compounds used in this research
Full size table

Water contact angles were measured using a WCA analyzer (FAMAS, Drop Master, Kyowa Interface Science, Japan) with 1.5 or 5 µL ultrapure water droplets, and the behavior of 5 µL water droplets dropped from a height of 2 cm onto the pillared vulcanized rubber surface with different degrees of elongation was recorded by a high-speed camera (EXILIM, EX- FH20, CASIO, Japan) at a frame rate of 1200 fps.

Results and discussions

Surface observations of the vulcanized rubber microstructures

Figure 1a, b shows FE-SEM images of the silicon mold used in this research. Hexagonally arranged tapered micropores (surface diameter of ca. 6.1 µm, bottom diameter of ca. 2.5 µm, and a depth of ca. 6.3 µm) are shown. Figure 1c–e shows FE-SEM images of the vulcanized rubber surfaces prepared by using a silicon mold with tapered micropores. Spiky vulcanized rubber microstructure arrays prepared from the three types of vulcanized rubber listed in Table 1 were successfully formed on the surface. This result suggests that the rubber composition did not have a significant effect on the structural transfer, as the compositions used in this experiment are generally used in tires and were not designed specifically for this research. Therefore, it is possible to obtain vulcanized rubber microstructures with a wide range of compositions.

Fig. 1
figure 1

Field-emission scanning electron microscope (FE-SEM) images of the silicon mold and the microstructured vulcanized rubber surfaces. a Top and b cross-sectional view of the silicon mold. Spiky vulcanized rubber structures prepared from the c Rubber 1, d Rubber 2, and e styrene-butadiene rubber (SBR) compounds listed in Table 1. f Cross-sectional view of a split sample before removal of the vulcanized rubber from the silicon mold

Full size image

Detailed FE-SEM observations revealed that the sides of the spikes were jagged, reflecting the shape of the silicon mold, while the tops of the spikes were smooth. According to the FE-SEM image shown in Fig. 1f, a transverse sample of the silicon mold and vulcanized rubber before their separation, the vulcanized rubber almost fully filled the mold micropores, although there was a little space between the bottom of the mold micropores and the vulcanized rubber. We speculated that this space was caused by compressed air and that it is the reason why the tops of the spikes were smooth and enabled easy removal of the vulcanized rubber from the silicon mold without the use of release agents (wettability differences were also important for easy peeling of the hydrophobic vulcanized rubber off the hydrophilic silicon mold with oxide surface layers). This result indicates that the formation of vulcanized rubber microstructures is feasible, but we should be somewhat concerned about the existence of trapped air that inhibits exact structure transfer from the mold to the rubber surface, although the trapped compressed air might also aid in easy removal.

Surface wettability and pattern rearrangement of the spiky vulcanized rubber by stretching

We successfully prepared vulcanized rubber microstructures with the intention of enhancing surface wettability. Before WCA measurements of the spiky vulcanized rubber surfaces, the WCAs on flat rubber surfaces composed of Rubber 1 (Table 1) were measured before and after vulcanization. The WCAs on the flat rubber surfaces were ca. 106° regardless of vulcanization, indicating that the vulcanization process did not affect surface wettability. However, the WCAs on the flat rubber surfaces were slightly higher than those on a natural rubber surface (ca. 90°–100°) [31]. This difference may have been caused by surface roughness generated by the addition of carbon black or silica. The vulcanized rubber surfaces maintained their hydrophobicity after the vulcanization process, and the WCAs on the spiky vulcanized rubber surfaces were measured with 1.5 µL of ultrapure water. As a result, the WCAs on the surfaces were ca. 158°, indicating that the spiky vulcanized rubber surfaces showed superhydrophobicity.

We inferred the state of a water droplet on the spiky rubber surface using Wenzel’s and the Cassie–Baxter equation. According to Wenzel’s equation (1), with a WCA on a flat vulcanized rubber surface of 106° and an R of 2.57, as measured using a laser microscope, the theoretical value of the WCA on the spiky vulcanized rubber surface was 121°, which differs significantly from the experimental value. On the other hand, the theoretical WCAs on spiky vulcanized rubber surfaces were 161° based on the Cassie–Baxter equation with a vulcanized rubber fraction f of 0.077, which was calculated as a water droplet on the top of the 2.55 µm diameter spikes, and a hexagonally arranged spike periodicity of 8.73 µm. To compare the theoretical WCAs obtained by the Cassie–Baxter equation and the experimental value, spiky vulcanized rubber surfaces were assessed in the Cassie–Baxter state.

One of the characteristics of vulcanized rubber is its elasticity. We examined structural and WCA changes induced by stretching the microstructured vulcanized rubber. Here, we chose Rubber 1, which is softest vulcanized rubber among the compounds used in this research. In this experiment, the spikes were arranged hexagonally; therefore, the surface had six-fold symmetry belonging to p6m in the plane crystallographic group, and there are two elongation directions, (1, 1) and (1, 0), to the equilateral triangle of the basic grid. Figure 2a–h shows three-dimensional (3D) laser microscope images taken before and after stretching the microstructured vulcanized rubber. When the microstructured vulcanized rubber was stretched in the (1, 1) direction (Fig. 2a–d), the spike arrangement changed from hexagonal to a linear pattern, and in the case of the (1, 0) direction, the spike arrangement changed to a slanted linear pattern (Fig. 2e–h). This arrangement difference was caused by deformation of underlying parts of the spikes. In the case of the (1, 1) elongation direction, the deformations were not affected by the spike arrangement because the arrangement had already been transformed before deformation of the underlying parts of the spikes. In the case of the (1, 0) elongation direction, once periods of spikes were elongated (Fig. 2f), vertical spike periods to the elongation directions were decreased (Fig. 2g, h). At this moment, pointed ellipse-shaped underlying parts interfered with each other, and line patterns were slanted to escape stress. Figure 2i shows graphs of the changes in the structural parameters by elongation in the (1, 1) direction. In Fig. 2i, LP represents the line period, which is the width between the newly created lines by stretching; PP represents the pillar (spike) period of the compression of distance induced by elongation; and PH represents the pillar (spike) height. According to the graph, changes in the LP were proportional to the degree of elongation, and those in the PP were inversely proportional to the degree of elongation. However, although the PH seemed to decrease slightly, in comparison to the other structural parameters, PH was relatively unchanged. In this vulcanized rubber, the spiky structures seemed to be formed on the surface of the vulcanized rubber sheet so that when the vulcanized rubber was elongated, the bottom of the vulcanized rubber sheet stretched initially and the spiky structures were not affected. This fact indicates that we can transform the spike arrangement by stretching without destroying or deforming the spiky structures. Figure 2j shows the graph of changes in the WCAs and water sliding angles (WSAs) induced by stretching. The WCAs were not changed, as the rubber fraction f was also not changed significantly; however, the WSAs decreased slightly with increases in the degree of elongation. Unfortunately, the linearly patterned spiky vulcanized rubber surfaces did not show any water sliding directional anisotropy, as these surfaces were too water repellent to affect the direction of water sliding. According to these results, vulcanized rubber surfaces can realize superhydrophobicity, and their microstructure arrangements can be deformed without destroying the microstructures. Furthermore, although spiky vulcanized rubber surfaces were too water repellent, the results suggested the possibility of wettability control through stretching of the vulcanized rubber microstructure arrangement.

Fig. 2
figure 2

Laser microscope images and graphs of spiky vulcanized rubber surfaces with different degrees of elongation. Laser microscope three-dimensional (3D) images of spiky vulcanized rubber surfaces elongated in the a–d (1, 1) and e–h (1, 0) directions. Inserted images are the top views of each surface. i Graph of changes in structural parameters induced by stretching. PH, LP, and PP represent spike period, line period, and spike height, respectively. Inserted images indicate the LP and PP. j Graph of water contact angle (WCAs) and water sliding angles (WSAs). Inserted images show photographs of a 1.5 µL water droplet on the spiky vulcanized rubber surface for each degree of elongation

Full size image

Vulcanized rubber pillar arrays and wettability control

To control the surface wettability of vulcanized rubber, another type of microstructure was prepared by using a silicon mold with hexagonally arranged perpendicular micropores (ca. 5.4 µm in diameter and ca. 9.6 µm in depth) [36]. Figure 3 shows laser microscope images and graphs of the vulcanized rubber surface prepared using the silicon mold with perpendicular micropores. The vulcanized rubber surface had hexagonally arranged micropillar structures of ca. 5.3 µm in diameter and ca. 9.5 µm in height, forming a structure that mirrored the silicon mold. This arrangement was also controlled by elongation without changing the pillar structure (Fig. 3e). The WCAs on the vulcanized rubber pillar surfaces were ca. 156°, and the WSAs were ca. 15°; therefore, the WCAs and WSAs were almost constant in this elongation range (Fig. 3f). The WCA value closely matched the theoretical value of 158° obtained using the Cassie–Baxter equation with an f &value of 0.10 (only the top of the pillars were in contact with the water). Although these microstructured vulcanized rubber surfaces appeared to be merely superhydrophobic surfaces, when water droplets were dropped onto the surface, the water repellent behavior of the surfaces changed depending on the degree of elongation.

Fig. 3
figure 3

Laser microscope images and graphs of vulcanized rubber micropillar surfaces with different degrees of elongation. The elongation direction was (1, 1). a–d Three-dimensional (3D) laser microscope images of the pillar surfaces at different degrees of elongation ranging from 100 to 250%. Inserted images are the top views of each surface. All bars represent 10 µm. e Graph of the changes in structural parameters induced by stretching. Inserted images also indicate the line period (LP) and spike height (PP). f Graph of water contact angles (WCAs) and water sliding angles (WSAs). Inserted images show photographs of a 5.0 µL water droplet on the vulcanized rubber micropillar surface for each degree of elongation

Full size image

Figure 4 shows time-sequence photographs obtained by a high-speed camera during the application of 5 µL water droplets from a height of 2 cm onto the pillars of the vulcanized rubber surfaces with different degrees of elongation. In the case of 100% elongation (before stretching the rubber), the water droplet bounced off the surface, while at 150% elongation, the surface was partially wetted and the water droplet was split in two (white arrows in Fig. 4). At over 200% elongation, the microstructured rubber surfaces were wetted, and the water droplets adhered to the surface. These phenomena indicated that the wetting state was changed from a Cassie–Baxter state to a Wenzel state by transformation of the microstructure arrangement of the vulcanized rubber.

Fig. 4
figure 4

Sequential photographs taken using a high-speed camera at 1200 fps during the dropping of 5.0 µL water droplets onto the vulcanized rubber micropillar surface

Full size image

Theoretical analysis of the water behavior on vulcanized rubber pillar arrays with different degrees of elongation

We next tried to understand this phenomenon by a simple theory focusing on wettability transition and water penetration among the pillars. In this discussion, we present a semiquantitative theory in which the order of the parameters is valid, as we would like to obtain an overview of this quite variable phenomenon by comparing the magnitude relationship among factors. The driving factor is the potential energy of the water droplet, and the suppression factors are excess surface energy and frictional energy, as explained below.

When a water droplet was dropped onto the microstructured vulcanized rubber surface, the water droplet was highly deformed from a sphere to a pancake- shape, so the excess surface energy of the water droplet from the sphere \(\Delta E_{\mathrm{surf}}\) appears. Furthermore, to realize a Wenzel state, water should penetrate among the hydrophobic pillar structures. In doing so, frictional energy Efric is generated by the viscosity of the water. If the sum of \(\Delta E_{\mathrm{surf}}\) and Efric is equal to the potential energy mgh of a water droplet, the relationship among the three energies can be described as follows:

$$mgh = \Delta E_{\mathrm{surf}} + E_{\mathrm{fric}}.$$
(4)

Here, mgh was ca. 1.0 µJ. When the shape of the water droplet was deformed from a sphere (5 µL, radius of 1.06 mm) to a pancake, its surface area ∆S was estimated to double, and it became 1.41 × 10–5 m2. \(\Delta E_{\mathrm{surf}}\) is the multiple of the surface tension of water ywater and ∆S, and its value was calculated as ca. 1.0 µJ, which is a similar value to the potential energy of water; therefore, Eq. (4) well explains the deformation of the dropped water droplets.

Next, we focused on Efric due to the water viscosity. Figure 5 shows schematic representations of Efric on the patterned surface. When a water droplet was deformed to a pancake shape (Fig. 5a), the water penetrated one triangular grid hole consisting of three pillars (Fig. 5b), and frictional energy was generated (Fig. 5c). Here, considering efric, which is the frictional energy generated by the intrusion of water into one triangular grid hole, and the hole number N that came into contact with the bottom of the pancake-shaped water droplet, the energy required to transition from a Cassie–Baxter to a Wenzel state can be obtained. This energy is equal to Efric:

Fig. 5
figure 5

Schematic representations of the friction energy on a patterned rubber surface. a A water droplet spreading on the surface. b A triangle lattice made up of three pillars (left) and its corresponding pore (right). c Velocity gradient of the water in the pore. Reproduced from the literature [36]

Full size image

According to previous literature reports [37], the frictional energy rate \({\dot e_{\mathrm{fric}}}\) for one hole (radius of pore: Rpore; depth of pore: PH) per unit time is expressed as below:

$$\dot e_{\mathrm{fric}} = {\int} {{\int}_0^{R_{\mathrm{pore}}} {{\mathrm{d}}r^2{\int}_0^{PH} {{\mathrm{d}}z\eta \left( {\frac{{{\mathrm{d}}v_z\left( r \right)}}{{{\mathrm{d}}r}}} \right)^2 = \eta v_{\mathrm{center}}^2 \times PH} } },$$
(5)

where r is \(x\) y and η are viscosity coefficients (calculated as 10−3 Pa·s); Rpore (expressed as \(2R_{\mathrm{pore}} = (PP - D)/\sin \theta _{\mathrm{pp}}\), \({\mathrm{d}}v_z/{\mathrm{d}}r\)) is the gradient of the velocity of the water flowing into the hole; vcenter is the velocity of the hole center (\(mgh = \left( {1/2} \right)mv_{\mathrm{center}}^2\)); and the gradient operates as \(v_{\mathrm{center}}/R_{\mathrm{pore}}\).

On the other hand, N is given as

$$N = \frac{{\Delta S}}{{\left( {PP \times LP} \right)/2}}.$$
(6)

The combination of Eqs. (5) and (6) allows the energy-dissipation rate \(\dot E_{\mathrm{fric}}\) due to the friction generated when water penetrates the whole gaps among pillars in contact with the bottom of a water droplet to be estimated

$$\dot E_{\mathrm{fric}} = \dot e_{\mathrm{fric}} \times N = \eta \times (2gh) \times PH \times N.$$
(7)

By setting the time taken for energy dissipation as τ,

$$E_{\mathrm{fric}} = \eta \times (2gh) \times PH \times N \times \tau.$$
(8)

Table 2 shows the calculated values for the friction energies. By setting Efric for an elongation of 150%, which was the transition condition from a Cassie–Baxter to a Wenzel state, as the criterion, the friction energy for 100% elongation was found to be 20% larger than that for 150% elongation. This result means that the penetration of water into the gaps among the vulcanized rubber micropillars was difficult because of the large friction energy and that wetting increased with an increase in the degree of elongation, in other words, with a decrease in friction energy. Moreover, because we can observe water droplet deformation by a high-speed camera with a 1 ms time resolution, τ was assumed to be 1 ms, and the Efric for an elongation of 100% was calculated as 0.52 µJ. This value is half the potential energy of the water droplet (ca. 1.0 µJ); thus, friction energy could not be ignored, and a Cassie–Baxter state was maintained when the elongation was 100%. To summarize this section, by considering friction energy during the penetration of water into the microstructures, we can obtain a general understanding of the dynamic wetting behavior from a phenomenological perspective, making it possible to design wettability-controllable surfaces regulated by the degree of elongation.

Table 2 Estimations of energy-dissipation rate Éfric and friction energy Efric from a Cassie–Baxter state to a Wenzel state on the microstructured vulcanized rubber with different degrees of elongation (reproduced from the literature [36])
Full size table

Conclusion

In this report, we described a simple and easy preparation of microstructured vulcanized rubber surfaces by using silicon molds, which were processed by conventional photolithographic techniques. Furthermore, the microstructured vulcanized rubber surfaces showed novel functions generated by the surface microstructures. The spiky vulcanized rubber surface showed superhydrophobicity, with WCAs of ca. 160°. The WCAs well matched the theoretical values derived by using the Cassie–Baxter equation. The arrangement of the spikes could be rearranged from a hexagonal to a linear pattern without structural changes to the spikes by stretching, and in this case, the WCAs were constant regardless of the degree of elongation. This fact indicates that we can rearrange only the arrangement of the microstructures formed on vulcanized rubber surfaces. We also prepared vulcanized rubber micropillar surfaces to control surface wettability by stretching, and a demonstration and theoretical explanation of the water droplet behavior upon dropping onto the surface are presented. The results revealed that the surface wettability of the vulcanized micropillar array surface was dependent on the degree of elongation. When the elongation was 100%, the water droplet bounced off the surface, while water droplets could be made to adhere to the surface by increasing the elongation degree. In other words, this change in wettability represented a transition from a Cassie–Baxter to a Wenzel state. We also provided a general explanation of this quite variable phenomenon by considering the surface energy of a water droplet and the friction energy generated during water penetration into the gaps between the vulcanized micropillars.

We believe that these results can lead to the development of a new field. Generally, vulcanized rubbers are used for bulk products, such as tires, so surface microtexturing had not been considered. First, highly viscous unvulcanized rubber was not considered able to run into a micromold. However, we have demonstrated the formation of microstructured vulcanized rubber by simple hot-press molding. In this report, we showed only the functions related to wettability, but functions generated by surface microstructures not only include superwettability but also affect friction [38,39,40], adhesion [41], coloration [42,43,44,45], anti-reflection [2, 23, 46], and so on [47]. Thus, we consider that our results will stimulate the development of new fields and, in the near future, allow the production of excellent products, as vulcanized rubber is a tough, inexpensive material and is used by many people belonging to a wide variety of fields.