A facile synthesis of a highly stable superhydrophobic nanofibrous film for effective oil/water separation

Abstract

An oil/water separation film with excellent durability and stable recyclability is highly desired for the treatment of oil containing effluents, like industrial oily wastewater. Herein, a novel superhydrophobic nanofibrous film (SNF) was fabricated by growing TiO₂ on a collagen fiber membrane (CFM), followed by surface coating of vinyl triethoxysilane. The as-prepared SNF was well characterized by SEM, SEM-EDS and XPS. Due to the unique structure, the SNF featured superhydrophobicity with a water contact angle (WCA) of 153°. When applied in the separation of oil/water mixtures, SNF exhibited a high separation efficiency of ∼97% in 5 circles, and it showed more effective oil/water separation efficiency compared to the commercial PTFE film. SNF was also effective for the separation of gasoline/water or diesel oil/water mixtures with varied volume ratios, and all the corresponding separation efficiencies were higher than 97% even after 5 cycles. In addition, SNF could still maintain superhydrophobicity with WCA around 150° after being treated by high temperature (120 °C), ultraviolet irradiation, water or organic solvent immersion for 24 h, manifesting an excellent durability.

---

1. Introduction

The ever-increasing industrial oily wastewater has received considerable attention owing to its potential risk to aquatic environmental systems. In this regard, great efforts have been dedicated to separate oil from oily wastewater. Conventional strategies, such as centrifugation, depth filtration and sedimentation, have been developed for the separation of immiscible oil/water mixtures. More recently, high-performance separation of oil/water mixtures has been realized using superhydrophobic/superoleophobic materials.^1–7 For the superhydrophobic materials with a water contact angle (WCA) higher than 150°, they are able to repel water from entering into the materials while allowing the oil to effectively penetrate through the superhydrophobic materials. In this way, effective and fast separation of oil/water mixtures is achieved.

Currently, the fabrication of superhydrophobic materials are generally achieved by coating low surface energy molecules on a rough surface of substrate,^8,9 which is more likely controlled by the hierarchical micro/nanostructure.¹⁰ Up to now, various methods have been proposed for constructing hierarchically rough surface, such as crystallization control,¹¹ chemical vapour deposition,¹² electrochemical deposition,¹³ lithography,¹⁴ and spray-coating method.^15,16 Huang and co-workers prepared a superhydrophobic Fe–N coated silicon substrate by chemical vapour deposition with aligned carbon nanotubes on it.¹⁷ Im and co-workers developed superhydrophobic microlens arrays on flexible substrates from PDMS using 3D diffusion lithography techniques.¹⁸ Compared with these artificial structures,^17–19 the nature mother has actually accomplished a series of fantastic superhydrophobic structural morphologies that present multi-scale hierarchical roughness from nanoscale to microscale after billions of years of evolution. For example, the surface of the lotus leaf is featured with a hierarchical rough surface, where micrometer-sized papillae has nanometer-size branch like protrusions.^8,20 Although learning from nature has long been a source of bio-inspiration for human beings,^21–25 it is still essentially important to discover other novel and fine structures from nature, which have more effective superhydrophobic properties.

Herein, a superhydrophobic nanofibrous film (SNF) was fabricated using collagen fiber membrane (CFM) processed from cattle hide as the natural template because of the unique hierarchically fibrous structure of CFM, from nanoscale to microscale. CFM is a typical 3D and robust nanostructure film, which features to a structural hierarchy based on the self-assembly of nanofibrous collagen. More specifically, tropocollagen first assembles together to form protofibril and then to microfibrils,^26,27 and these 1D nanofibrous microfibrils self-assemble into collagen fiber strand and further to 3D collagen fiber network. The 3D fiber network can provide enough hierarchical roughness from nanoscale to micrometer-scale. On the other hand, there are abundant void spaces existing inside the CFM, which are derived from the weave of microfibrils, and can serve as air bags to enhance the hydrophobic property when using CFM to fabricate SNF.

Although CFM has hierarchically rough surface, a high surface energy is involved in CFM because of its abundant polar groups such as carboxyl groups, hydroxyl groups and amino groups. Hence, it is necessary to utilize these hydrophilic groups as reactive sites to bind low surface tension materials, gaining the desired superhydrophobicity. To achieve this goal, TiO₂ nanoparticles were first grown on CFM surface via the hydrolysis of tetrabutyl titanate (TBT). Then, silane coupling agent (vinyl triethoxysilane, VTEO) is chemically bond onto the surface of CFM with the help of pre-formed TiO₂ nanoparticles. As a result, superhydrophobic SNF can be prepared owing to the existing of hierarchically micro/nanostructured surface with a low surface energy.

In the present investigation, we prepared a superhydrophobic nanofibrous film (SNF) based on above idea using CFM processed from cattle hide as the template. The obtained SNF was well characterized by SEM and SEM-EDS. The preparation mechanism of SNF was also elucidated according to XPS analysis. Subsequently, the wetting properties of SNF were systematically investigated by measuring the contact angle of water or liquid drop on its surface. The performances of SNF in oil/water separation were also tested in different conditions, and compared with commercial PTFE film. The durability of SNF was also investigated. Our experimental results demonstrate that the SNF exhibits excellent hydrophobicity, high separation efficiency and reusability in oil/water separation, which provides a facile route for realization of superhydrophobic surface based on natural biotemplate.

2. Experimental section

2.1. Materials

Cattle split leather was used as the CFM, which was supplied by Ruixing Leather Co., Ltd (Haining, China). All of the chemicals were analytical-grade reagents and used as received without further purification. Hydrophobic PTFE (average pore size, 0.6 μm) membrane was purchased from Laisheng filtration equipment factory (Haining, China).

2.2. Fabrication of SNF

CFM with diameter of 9 cm was rinsed with deionized water and ethanol thoroughly and then dried in the air. Then, 10 mL of 10 mM tetrabutyl titanate (TBT) solution (ethanol and toluene in 1 : 1 volume ratio) was infiltrated by CFM, and maintained at 70 °C for 90 min in an oven. Subsequently, the resultant TiO₂–CFM was immersed into 50 mL of toluene solution containing 10 mM vinyl triethoxysilane (VTEO) for 24 h. After that, the film was dried in vacuum at 30 °C, and the corrected film was denoted as SNF. For comparison, CFM was directly immersed in toluene solution containing 10 mM VTEO for 24 h. After vacuum dried at 30 °C, the resultant film was denoted as VTEO–CFM. For the SNF applied in oil/water separation, 100 mM VTEO was employed for its preparation.

2.3. Characterizations

The surface morphology of the samples was observed by field emission scanning electron microscope (FESEM, Hitachi 4700, Japan) operating at 5.0 kV. The elemental composition was determined by energy dispersive X-ray spectroscopy (EDS, INCA X-MAX 50, Oxford Instruments, U.K.). The chemical states of the element on the surface of the samples were recorded by X-ray photoelectron spectrum (XPS, Shimadzu ESCA-850, Japan) measurements with Al-Kα X-rays. Water contact angle (WCA) and oil contact angle (OCA) were tested on a Contact Angle System (OCA20/6, Dataphysics, Germany) at ambient temperature. A 5 μL of liquid droplet was suspended with needle tube and brought in contact with film using a computer controlled device, and the average contact angle value was obtained by measuring more than ten different positions for one sample.

3. Results and discussion

3.1. Surface morphology and chemical composition analysis

Fig. 1 is the schematic illustration showing the preparation mechanism from CFM to SNF and the corresponding SEM images at different steps. As shown in Fig. 1a and b, the deposition of TiO₂ nanoparticles onto CFM was carried out before the coating of silane coupling agent. This step aimed to strength the bonding effect between collagen fiber and the subsequent coated silane coupling agent. In general, a poor durability of superhydrophobic surfaces could be encountered when they are exposed to harsh conditions, such as corrosive substances and strong ultraviolet irradiation, which may even cause a permanent loss of superhydrophobic surface. Since the silane coupling agents are chemically bond with collagen fiber by using TiO₂ as bridge, the hydrophobic of the finally prepared SNF is therefore robust enough, and is able to exhibit long-term superhydrophobicity and durability against harsh environmental conditions. As shown in Fig. 1b, the original geometrical morphology of CFM is well retained in TiO₂–CFM, including nanofiber, collagen fiber and collagen fiber bundles. Compared with CFM, TiO₂–CFM is indeed featured by a rougher nanofiber surface with deposited TiO₂ nanoparticles with an average diameter of 25 nm (Fig. S1 in ESI†). As well reviewed by Sun et al., the hierarchical micro/nanostructures are essential for achieving superhydrophobic surface with WCA larger than 150°,¹⁰ because when a rough surface contacts with water, air trapped in the voids may minimize the contact area between water droplet and the material surface, which contribute greatly to the increase of hydrophobicity.^28–30 In this regard, the deposition of TiO₂ on CFM is crucial for achieving superhydrophobic SNF.

Fig. 1 Schematic illustration showing the preparation mechanism from CFM to SNF, and the corresponding SEM images.

To confirm this, VTEO–CFM was prepared by direct coating of silane coupling agent on CFM. As shown in Fig. S2 (ESI†), little difference is observed in geometrical morphology between CFM and VTEO–CFM. As for VTEO–CFM, the roughness-promoted hydrophobicity is less intensive than SNF. Actually, a direct coating of VTEO on CFM just provides a contact angle of 141° (Fig. 4c). The following SEM-EDS analysis further reveals that without the deposition of TiO₂ nanoparticles onto CFM, the amount of coated silane coupling agent is very limited, which renders VTEO–CFM much poor hydrophobicity as compared with SNF.

SEM-EDS test was carried out to analyze the elemental composition and corresponding distribution in SNF. The mapping images of Ti and Si share a similar shape, which is well matched with the surface morphology of SNF, as shown in Fig. 2a–c. Analogous phenomenon is also observed in Fig. 2e and f. These results manifest that both the TiO₂ and silane coupling agent are homogeneously dispersed on the whole surface and the cross-section of SNF. In contrast, the intensity of Si signal on the mapping image of Si is very weak both on the surface and the cross-section of VTEO–CFM (Fig. S3 in ESI†). It indicates that very limited amount of silane coupling agent is coated on the surface of CFM without the pre-deposition of TiO₂. The obvious difference of Si content in SNF and VTEO–CFM suggests that the pre-deposition of TiO₂ on CFM is indeed critical for achieving both surface roughness and bond robustness between collagen fiber and the silane coupling agent.

Fig. 2 SEM images and SEM-EDS mapping images for the surface of SNF (a–c), and for the cross-section of SNF (d–f). The scale bar is 0.3 mm.

XPS analysis was carried out to further investigate the preparation mechanism of SNF. Fig. 3a is C 1s XPS spectrum of CFM, where the C 1s peak is fitted by three peaks, C–C/C–H (284.8 eV), C–O/C–N (286.1 eV) and C=O (287.8 eV).³¹ After the deposition of TiO₂, the peak intensities of C–O/C–N and C=O are decreased, which suggest that the functional groups contained in CFM, such as –COOH and –NH₂, have reacted with Ti precursors, leading to a stable deposition of TiO₂ nanoparticles on CFM. Accordingly, the intensity of O 1s peak of TiO₂–CFM is increased at 532.4 eV owing to the formation of Ti–O (Fig. 3b–e), and a new peak attributed to N–O appears at 402.5 eV. The presence of Ti 2p XPS peak at 458.8 eV (Fig. S4a in ESI†) confirms the existing of TiO₂ in TiO₂–CFM. For SNF, the survey scan XPS spectrum show the presence of Si (Fig. S4b in ESI†), and the peak intensity of O 1s at 532.6 eV is further increased owing to the surface coverage of Si on SNF (Fig. 3h).

Fig. 3 C 1s (a), O 1s (b) and N 1s (c) core level XPS spectra of CFM surface, C 1s (d), O 1s (e) and N 1s (f) XPS spectra of TiO₂–CFM surface, C 1s (g), O 1s (h) and N 1s (i) XPS spectra of SNF surface.

3.2. Water wettability

The water wettability of CFM, TiO₂–CFM, VTEO–CFM and SNF was systematically investigated, by measuring water contact angle (WCA) of a 5 μL water droplet on the surface of these films. Although CFM is somewhat hydrophobic with a WCA of 133° (inset in Fig. 4a), a lager water droplet (50 μL of water labelled by methyl orange) wetted the surface completely in a short time (Fig. 4a), which resulted from the existence of abundant polar groups in CFM. Hence, with the unique hierarchical fibrous structure, CFM can just perform temporary hydrophobicity. Compared with CFM, TiO₂–CFM exhibits a higher temporary WCA of 137° with 5 μL water droplet (inset in Fig. 4b), which is attributed to the enhanced roughness from TiO₂ deposition. However, a 50 μL water droplet still permeated into TiO₂–CFM (Fig. 4b) due to the hydrophilic nature of TiO₂.

Fig. 4 Photographs of water droplets (50 μL, labelled by methyl orange) on CFM (a), TiO₂–CFM (b), VTEO–CFM (c), SNF (d). Insets in each figure show the corresponding images of static water droplets (5 μL) on different film surfaces.

For SNF, a 5 μL water droplet could sit on its surface with a WCA at 153.1° (inset in Fig. 4d). When a larger volume of orange water (50 μL) is placed on SNF, a nearly spherical water droplet is still steadily stayed on the surface for extended periods of time (Fig. 4d). These facts confirm the superhydrophobic nature of SNF surface. On the contrary, the WCA of VTEO–CFM is only 141° (inset image in Fig. 4c), and a 50 μL water droplet could remain on the film surface (Fig. 4c). These differences in hydrophobicity suggest that besides the surface coverage of silane coupling agent with low surface energy, the enhancement of roughness by TiO₂ deposition is essentially important for SNF to achieve superhydrophobicity. Considering the surface structure of SNF, the observed hydrophobic property of SNF can be attributed to the formation of air cushion beneath the Si coated TiO₂-nanofiber network. As shown in Fig. 2a, the microstructure of SNF features to a hierarchically fibrous structure and contains many cavities. When water droplets were dropped on SNF, these cavities can trap air between the water droplets and SNF surface to enhance the hydrophobic property, like on porous surface. Accordingly, we employed Cassie–Baxter model to describe the wetting behaviours of water droplet on SNF, which is expressed by the following Cassie–Baxter equation:³²

cos θ_r = f₁ cos θ − f₂ (1)

where θ_r is the contact angle on the superhydrophobic SNF surface, f₁ is the area fraction of SNF in contact with liquid, θ is the contact angle on TiO₂ surface, f₂ is the area fraction of air on SNF surface, respectively.

Based on the wettability measurements of SNF and TiO₂, the θ_r and θ are 153.1° and 0°, respectively. Thus, the values of f₂ and f₁ are determined to be 0.95 and 0.05 according to eqn (1), which suggests that 95% surface area of SNF is covered by air, and only 5% surface area of SNF is contacted with water. In contrast, the values of f₂ and f₁ are determined to be 0.89 and 0.11 for VTEO–CFM. Obviously, the surface area of VTEO–CFM contacted with water is much larger (11%) than that of SNF (5%). Hence, the air pocket created by the deposition of TiO₂ nanoparticles onto fibers is the main reason responsible for the hydrophobicity of SNF.

Then all the films were put on the surface of aqueous solution. CFM, TiO₂–CFM and VTEO–CFM were all wetted by water and sank into water gradually (Fig. S5a–c in ESI†) except SNF (Fig. S5d in ESI†). In contrast, SNF kept to float on water surface after releasing of the external force, and no water uptake was observed. This phenomenon indicates that a Cassie–Baxter state was indeed formed on SNF surface owing to the existing of air pockets, and the trapped air guarantees the surface unwetted when put on the surface of aqueous solution. Silver mirror like phenomenon can be also observed on SNF when it was submerged in water (Fig. S5e in ESI†), which was the Cassie–Baxter nonwetting behaviour owing to a continuous air layer between the superhydrophobic surface and water,³³ thus getting a total reflectance of light at the air layer trapped on this special surface.³⁴ These results demonstrated that the excellent superhydrophobicity of SNF.

Besides, other silane coupling agents were also used as surface modifying agents, including γ-aminopropyltriethoxysilane and trimethoxy(octadecyl)silane, which also ensured TiO₂–CFM the superhydrophobicity with the water contact angle of 151.3° and 152.7°, respectively (Fig. S6 in ESI†).

3.3. Performance of SNF in oil/water separation

One attractive application of SNF is the oil/water separation. Hence, SNF was used as membrane in oil/water separation, and different organic solvents were employed to simulate the oil, including n-hexane, octane and dodecane. The separation process was carried out at atmospheric pressure and room temperature. To get a clear observation for the separation, oil (the organic solvent) was coloured by sudan IV and mixed with water coloured by methylene blue. When the mixture was poured onto the filter funnel, the water remained on the upper surface of SNF while the oils with lower surface tension quickly spread, penetrated and passed through SNF, and was collected in the container below SNF. Fig. 5a shows the already separated dodecane/water mixture, where methylene blue coloured water was retained at the upper surface of SNF, while the sudan IV coloured dodecane was collected in the container.

Fig. 5 Application of SNF for the separation of (a) dodecane/water mixture and (b) diesel oil/water mixture, where the water is coloured by methyl blue while the dodecane is coloured by sudan IV.

The separation efficiency of SNF for oil/water mixture is evaluated by the mass ratio of the water before mixture and after separation, which is calculated according to eqn (2):³⁵

η = m₀/m₁ × 100% (2)

where m₀ and m₁ are the mass of the water after and before the separation process, respectively.

Based on our experimental results, n-hexane/water mixture (Fig. 6a), octane/water mixture (Fig. S7a in ESI†) and dodecane/water mixture (Fig. S7b in ESI†) with varied volume ratios from 1 : 1 to 4 : 1 could be separated by using SNF as the membrane, and all the separation efficiencies were higher than 97%. The recyclability of SNF was also investigated, and we found that the decreases of separation efficiency were all less than 3% even after 5 cycles. Furthermore, we also applied SNF for the separation of gasoline/water mixture and diesel oil/water mixture. Fig. 5b shows the already separated diesel oil/water mixture. Gasoline/water mixture and diesel oil/water mixture with varied volume ratios from 1 : 1 to 1 : 20 (Fig. 6b) could be separated at atmospheric pressure by our SNF with all the separation efficiencies higher than 97%, even after 5 cycles (Fig. S8 in ESI†). Hence, these results demonstrated an outstanding reusability of SNF in oil/water separation. For comparison, CFM was directly coated with VTEO without pre-deposition of TiO₂ nanoparticles. When VTEO–CFM was applied for oil/water separation, the water gradually penetrated through the VTEO–CFM, and mixed with oil at the bottom of the collector. Hence, VTEO–CFM cannot be applied for effective separation of oil/water mixture, demonstrating the necessary of pre-deposition of TiO₂ nanoparticles on CFM.

Fig. 6 Separation efficiency of SNF for n-hexcane/water mixture in 5 cycles (a), separation efficiency of SNF for gasoline/water mixture and diesel/water mixture (b), WCA of SNF after different harsh treatments for 24 h (c), stored at 120 °C (1), UV irradiation (2), immersed in deionized water (3), n-hexane (4), octane (5) and dodecane (6).

To further study the separation ability of SNF, its water intrusion pressure and oil flux were investigated. The bearable height of water column that SNF can support determines the water intrusion pressure. The intrusion pressure (P) was calculated by the eqn (3):^16,36,37

P = ρgh (3)

where ρ is the density of water, g is the acceleration of gravity, and h is the bearable height of water column SNF can support. As shown in Fig. S9 in ESI,† the average bearable height was determined to be 75.0 cm, and the corresponding intrusion pressure was about 7.4 kPa.

In addition, the oil flux of SNF was investigated under a fixed column of gasoline, which was calculated by the eqn (4):^1,2

F = V/St (4)

where V is the volume of gasoline that permeates through the membrane (set as 0.1 L), S is the area of SNF, and t is the required time for the permeation of 0.1 L of gasoline. The gasoline permeates through SNF with an average flux of 2.4 L m⁻² min⁻¹.

The wettability is sensitive to the environmental change,³⁸ so we investigated SNF wettability against harsh environments by measuring the WCA after treated by storing at 120 °C, UV irradiation (365 nm) or immersing in different liquids for 24 h, respectively. As shown in Fig. 6c, SNF maintained its wonderful hydrophobicity with WCA around 150° after different treatments, and this manifests an outstanding stability of the superhydrophobic coating on SNF.

We also compared the oleophilicity of SNF and a commercial hydrophobic PTFE membrane (WCA = 141.0°, pore size = 0.6 μm). Dynamic wetting behaviours of dodecane adsorption on the surface of SNF and PTFE were studied with a time evolution (see Fig. 7a and b). Once oil (dodecane) contacted with SNF, it is rapidly spread out and the oil contact angle (OCA) quickly reduced to 0°, and then penetrated SNF thoroughly in 18 ms. However, the oil permeation time was extended to 2860 ms on PTFE membrane, which is about 150 times as much as that of SNF, and an initial OCA at 34.2° (Fig. 7b) could be even observed on PTFE.

Fig. 7 Photographs of dynamic measurements of dodecane adsorption on the surface of SNF (within 18 ms) (a), PTFE (within 2860 ms) (b). Separation time of SNF and PTFE for dodecane/water separation (c).

We recorded the separation time of dodecane/water mixture by using SNF and PTFE, respectively (Fig. 7c). For the same volume of dodecane/water mixture with ratio of 5 : 5 or 8 : 2, the separation time required by SNF is 0.53 min or 1.17 min, much shorter than those of PTFE (22.5 min or 24.3 min). SNF owns a robust nanostructure on the surface (Fig. 2a) and a hierarchically fibrous structure in cross-section (Fig. 2d), while PTFE just owns a much flatter surface (Fig. S10a–c in ESI†) without a hierarchically fibrous structure (Fig. S10d–f in ESI†). Thus, it is deduced that the structure differences between SNF and PTFE are responsible for their different oleophilicity and separation efficiency.

4. Conclusions

In summary, a robust superhydrophobic nanofibrous film (SNF) was facilely fabricated by collagen fiber film as the biotemplate. This novel SNF film features to a hierarchical microstructure and low surface energy, and thus can be used as a high-performance membrane for the separation of a variety of oil/water mixtures. Moreover, this SNF is highly stable to sustain its superhydrophobic properties in harsh environmental conditions. The approach developed in this work may also be expanded for the fabrication of other superhydrophobic organic–inorganic hybrid materials.

Acknowledgements

This work was supported by National Natural Science Foundation of China (51507107) and the Science and Technology Fund for Distinguished Young Scholars of Sichuan Province (2016JQ0002).

References

1. J. Li, L. Yan, H. Li, W. Li, F. Zha and Z. Lei, J. Mater. Chem. A, 2015, 3, 14696–14706.
2. J. Li, L. Yan, Y. Zhao, F. Zha, Q. Wang and Z. Lei, Phys. Chem. Chem. Phys., 2015, 17, 6451–6457.
3. J. Li, D. Li, Y. Yang, J. Li, F. Zha and Z. Lei, Green Chem., 2016, 18, 541–549.
4. L. Y. Liu, C. Chen, S. Y. Yang, H. Xie, M. G. Gong and X. L. Xu, Phys. Chem. Chem. Phys., 2016, 18, 1317–1325.
5. S. Yu and Z. Guo, RSC Adv., 2015, 5, 107880–107888.
6. C. Tan, Q. Li, Y. Li, C. Zhang and L. Xu, RSC Adv., 2016, 6, 53813–53820.
7. W. Ma, Q. Zhang, D. Hua, R. Xiong, J. Zhao, W. Rao, S. Huang, X. Zhan, F. Chen and C. Huang, RSC Adv., 2016, 6, 12868–12884.
8. L. Feng, S. H. Li, Y. S. Li, H. J. Li, L. J. Zhang, J. Zhai, Y. L. Song, B. Q. Liu, L. Jiang and D. B. Zhu, Adv. Mater., 2002, 14, 1857–1860.
9. M. Zhang, C. Y. Wang, S. L. Wang and J. Li, Carbohydr. Polym., 2013, 97, 59–64.
10. T. L. Sun, L. Feng, X. F. Gao and L. Jiang, Acc. Chem. Res., 2005, 38, 644–652.
11. X. He, T. Ge, Z. Hua, J. Zhou, J. Lv, J. Zhou, Z. Liu and J. Shi, ACS Appl. Mater. Interfaces, 2016, 8, 7118–7124.
12. J. Cherusseri, R. Sharma and K. K. Kar, Carbon, 2016, 105, 113–125.
13. Y. Liu, S. Li, J. Zhang, J. Liu, Z. Han and L. Ren, Corros. Sci., 2015, 94, 190–196.
14. J. Li, D. Li, Y. Yang, J. Li, F. Zha and Z. Lei, Green Chem., 2016, 18, 541–549.
15. F. Wang, X. Zhang, L. Zhang, M. Cao, Y. Lin and J. Zhu, Dyes Pigm., 2016, 130, 202–208.
16. J. Li, L. Yan, H. Li, J. Li, F. Zha and Z. Lei, RSC Adv., 2015, 5, 53802–53808.
17. L. Huang, S. P. Lau, H. Y. Yang, E. S. P. Leong, S. F. Yu and S. Prawer, J. Phys. Chem. B, 2005, 109, 7746–7748.
18. M. Im, D. H. Kim, J. H. Lee, J. B. Yoon and Y. K. Choi, Langmuir, 2010, 26, 12443–12447.
19. J. Li, R. Kang, X. Tang, H. She, Y. Yang and F. Zha, Nanoscale, 2016, 8, 7638–7645.
20. W. Barthlott and C. Neinhuis, Planta, 1997, 202, 1–8.
21. L. P. Lee and R. Szema, Science, 2005, 310, 1148–1150.
22. F. Xia and L. Jiang, Adv. Mater., 2008, 20, 2842–2858.
23. N. Huebsch and D. J. Mooney, Nature, 2009, 462, 426–432.
24. M. J. Liu, Y. M. Zheng, J. Zhai and L. Jiang, Acc. Chem. Res., 2010, 43, 368–377.
25. K. S. Liu and L. Jiang, Nanoscale, 2011, 3, 825–838.
26. W. Friess, Eur. J. Pharm. Biopharm., 1998, 45, 113–136.
27. A. Steplewski, V. Hintze and A. Fertala, J. Struct. Biol., 2007, 157, 297–307.
28. R. N. Wenzel, Ind. Eng. Chem., 1936, 28, 988–994.
29. A. B. D. Cassie, Discuss. Faraday Soc., 1948, 3, 11–16.
30. S. Herminghaus, Europhys. Lett., 2000, 52, 165–170.
31. R. Flamia, G. Lanza, A. M. Salvi, J. E. Castle and A. M. Tamburro, Biomacromolecules, 2005, 6, 1299–1309.
32. A. B. D. Cassie and S. Baxter, Trans. Faraday Soc., 1944, 40, 546–551.
33. I. A. Larmour, S. E. J. Bell and G. C. Saunders, Angew. Chem., Int. Ed., 2007, 46, 1740–1742.
34. X. T. Zhu, Z. Z. Zhang, B. Ge, X. H. Men, X. Y. Zhou and Q. J. Xue, J. Colloid Interface Sci., 2014, 432, 105–108.
35. R. Gao, Q. Liu, J. Wang, J. Y. Liu, W. L. Yang, Z. Gao and L. H. Liu, Appl. Surf. Sci., 2014, 289, 417–424.
36. Z. Xue, S. Wang, L. Lin, L. Chen, M. Liu, L. Feng and L. Jiang, Adv. Mater., 2011, 23, 4270–4273.
37. X. Li, D. Hu, K. Huang and C. Yang, J. Mater. Chem. A, 2014, 2, 11830–11838.
38. H. Li, Y. S. Li and Q. Z. Liu, Nanoscale Res. Lett., 2013, 8, 183–188.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra17832d

---