Long term environmental durability of a superhydrophobic silicone nanofilament coating

Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

Long term environmental durability of a superhydrophobic silicone nanofilament coating Jan Zimmermann a , Felix A. Reifler b , Ulrich Schrade b , Georg R.J. Artus a , Stefan Seeger a,∗ a b

Institute of Physical Chemistry, University of Zurich, Winterthurerstr. 190, CH-8057 Zurich, Switzerland Empa, Swiss Federal Laboratories for Materials Testing and Research, Laboratory for Advanced Fibers, Lerchenfeldstrasse 5, CH-9014 St. Gallen, Switzerland Received 12 December 2006; received in revised form 8 February 2007; accepted 14 February 2007 Available online 20 February 2007

Abstract We report the results of extensive studies on the environmental durability of a transparent superhydrophobic silicone nanofilament coating developed in our lab. Both natural and artificial weathering test have been performed, revealing a true potential for outdoor applications. The coating retains its superhydrophobic and anti-reflective properties for at least 1 year of outdoor weathering, the longest time ever reported for a coating of this kind. Under artificial weathering the coating is impervious to (global) UV radiation but shows accelerated aging under the so called Acid Dew and Fog Test (ADF). Annealing the coating significantly improves its performance and durability. © 2007 Elsevier B.V. All rights reserved. Keywords: Superhydrophobic; Durability; Weathering; Silicone; Nanofilaments; Coating

1. Introduction Superhydrophobic surfaces are generating a lot of interest due to their potential for scientific and industrial application [1–3]. To date, many strategies to mimic the so called “Lotus Effect” have been published [4–17]. Aside from developing an inexpensive, easily applicable and versatile coating procedure, the greatest challenge remains the durability of the coatings [2,5]. Due to the high surface roughness required for the superhydrophobic effect, superhydrophobic surfaces are generally sensitive to abrasion. None the less, many non abrasive applications can be envisioned where the superhydrophobic effect offers significant benefits. Self cleaning facades or window panes are only the most prominent examples in this field. Naturally, these applications still require that the superhydrophobic effect is stable under the specific conditions that it is subjected to. In order to assess the potential for practical applications, a screening of its performance is important. In regard to the significant potential for superhydrophobic surfaces in outdoor applications, only few publications exist which evaluate a superhydrophobic coating in terms of envi-

∗

Corresponding author. Tel.: +41 44 6354451; fax: +41 44 6356813. E-mail address: [email protected] (S. Seeger).

0927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2007.02.033

ronmental durability. Nakajima et al. performed experiments on the durability of superhydrophobic boehmite films [18] for up to 75 days of outdoor exposure and 800 h under UV irradiation. Whereas static contact angles of their standard films decrease to 100◦ during the outdoor exposure, films doped with 2 wt% TiO2 showed an increased durability in terms of contact angle values. This was attributed to the photocatalytic effect of TiO2 that facilitates the removal of organic contaminants from the surface which are otherwise said to be a major factor in degradation of the surface under environmental conditions [5]. Additionally, the TiO2 doped films exhibited contact angles above 140◦ after 800 h of UV irradiation. Superhydrophobic films prepared by Sasaki et al. showed a decrease in contact angle from 158◦ to 149 ± 5◦ after 40 days of outdoor exposure [19]. Thieme et al. performed extensive artificial weathering tests on superhydrophobic surfaces as protective coatings on aluminium, with varying outcome in regards to wetting properties and corrosion protection [20]. Recently we have developed and reported a novel type of superhydrophobic surface coating comprising of silicone nanofilaments [21,22]. The coating procedure is cheap, versatile and applicable to a variety of materials and surface structures. Moreover the coating is transparent, even anti reflective. Studies on the chemical durability of the coating revealed excellent long

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

term durability in solvents and under neutral or mildly acidic aqueous conditions. The coating also retained its transparency [23]. Given these promising results, various weathering tests were performed in order to evaluate the coatings potential for long term applications, one of the major ambitions in the field of self cleaning coatings. Outdoor weathering was performed for a total duration of 12 months and artificial weathering for a total time of 4 weeks. All samples were evaluated in terms of surface functionality by contact and sliding angle measurements. Changes in surface structure were monitored by scanning electron microscopy and optical transparency by UV/vis spectroscopy. 2. Experimental 2.1. Sample preparation Details on the methods for sample preparation and annealing can be found in previous publications [21–23]. Briefly, cleaned glass coverslips (Menzel, Germany) were coated in a custom built gas phase reaction chamber at ambient temperature and an initial relative humidity of 40% with typically 300 ␮l Trichloromethylsilane (ABCR, Germany) overnight. After coating the substrates typically exhibit static contact angles of 160 ± 2◦ and sliding angles of 20 ± 4◦ for 10 ␮l water drops. Annealing was performed in a drying oven (Heraeus, Switzerland) at 200 ◦ C under ambient atmosphere overnight. Annealing generally improved the contact and sliding angle values by roughly 4–6◦ and 10–15◦ , respectively. 2.2. Outdoor weathering A coated and annealed glass slide (Ø 50 mm × 1 mm) was fixed in front of a webcam (Monacor, TVCCD-162SCOL) and mounted at 70◦ inclination on the weather side of our lab. Contact and sliding angles of the glass slide were measured at regular intervals and pictures taken during rain/snow to visualize the superhydrophobic effect. For reference, a second webcam with an uncoated glass slide was also mounted in the same way and measurements made simultaneously. Outdoor exposure was performed for a total duration of 12 months from September 2004 until September 2005. 2.3. Acid Dew and Fog Test The effect of acid rain on the coating was investigated using the so called “Acid Dew and Fog” test (ADF test) according to VDI 3958 Part 12 [24]. Test variant “C” was chosen to simulate conditions for Central Europe and degree of aggression I (pH 2.5) as recommended for paints and coatings. The samples were mounted in a Global UV testing chamber (systems Weiss type BAM, Weiss Umwelttechnik GmbH, Germany) by means of a custom-made sample holder. The 24 h ADF cycle comprised four distinct phases. 1. In a first step (spraying phase, duration: <5 min), the samples were sprayed with an acid solution (mixture of H2 SO4 , HNO3

235

and HCl in a mass ratio of 1:0.3:0.17, diluted with deionised water to a pH of 2.5) using a commercial sprayer as it is used in household applications. One sprayer burst at a distance of approximately 30 cm was applied for each group of 5 samples (=one sample holder) to cover the surface of the samples with little droplets (mean diameter: ∼60 ␮m), replicating the deposition of acid dew in the early hours of the morning under a clear sky. Due to the superhydrophobic nature of the sample surface, the resulting load (∼5 g/m2 ) was considerably lower than the load specified in VDI 3958 Part 12 (30–40 g/m2 ). 2. In the following first drying phase, the samples were kept at 35 ◦ C and 30% relative humidity (r.h.) for 9 h and afterwards at 60 ◦ C and <10% r.h. for 5 h. 3. In the following rain phase, the samples were sprinkled with deionised water at 35 ◦ C for 4 h. 4. Finally this was followed by a second drying phase at 60 ◦ C and 10% r.h. for 6 h. According to VDI 3958 Part 12, after every 5 complete ADF cycles, there followed 2 cycles with omission of the spraying phase. During steps 2 to 4, the substrates were subjected to UV radiation simulating the global radiation in the range of 290–450 nm (UV-A: 36 W m−2 ; UV-B: 40 W m−2 ; UV-C: 2.4 W m−2 ). The intensity of the UV radiation was measured with a UVRadiometer (Minolta Radio-Meter UM-1) using three different sensors for the UV-A (360–480 nm), UV-B (310–400 nm) and UV-C (220–390 nm) part of the spectrum, respectively. For the ADF test, a total number of 30 annealed and 10 non annealed samples were exposed for the recommended duration of 4 weeks. Six annealed and two non annealed samples were removed from the chamber after every week to monitor the contact angle and sliding angle development as well as for scanning electron microscopy studies. 2.4. Laboratory weathering Laboratory weathering was performed in a Ci65A Xenon Weather-Ometer (Atlas Material Testing Technology GmbH, Germany) according to ISO 9022-9:2000-09 [25] (degree of aggression 4) using the following operating parameters: water cooled xenon arc lamp (inner and outer filter: borosilicate glass); irradiation: 0.54 W m−2 nm−1 at 340 nm (corresponding to 625 W m−2 at 300–800 nm, which is in reasonable accordance with ISO 9022-9 requiring an irradiation of 623 W m−2 at 320–780 nm); black standard temperature: 78 ± 2 ◦ C; air temperature: 55 ± 2 ◦ C; relative humidity: 20 ± 3%. The samples were mounted on a custom made sample holder and placed at the height of the middle tier of the specimen rack into the exposure chamber. Two non annealed and three annealed samples were exposed in this fashion for a total time of 240 h. 2.5. Contact angle measurements Contact angle measurements were performed on a Contact Angle System OCA and included software (DataPhysics, Ger-

236

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

many). For static contact angle measurements, digital drop shape analysis was performed on a 10 ␮l sessile drop of deionised water using the Laplace Young fitting routine. Sliding angles were measured with the help of a home built tilting table immediately after the contact angle measurement. For the outdoor weathering experiment, the coated slide was evaluated at 10 random points. For the artificial weathering, each sample was evaluated at four designated positions with subsequent measurements at random control positions to exclude a possible bias. All measurements were performed at ambient conditions. Sliding angles instead of hysteresis were measured because of the experimental difficulties associated with measuring and analyzing advancing and receding contact angles at very high contact angle values (see also Refs. [5,21]). The limit of sliding angle measurements is 90◦ inclination. As we are primarily interested in the coatings ability to retain a low sliding angle, we forgo a further evaluation of the coatings properties in this regard once drops stay pinned on the surface. Consequently no sliding angle data are displayed for pinned drops. 2.6. Scanning electron microscopy Scanning electron microscopy was performed with a SUPRA 50VP (Zeiss, Germany). Samples were coated with 5 nm Platinum and analysed at 2 kV at a working distance of 2 mm using an In-Lens Detector. 2.7. UV/vis spectroscopy Optical transmittance was measured with a Lambda 900 UV/vis spectrometer (Perkin-Elmer) against air. 3. Results and discussion Benefits derived from a superhydrophobic coating in outdoor applications are two-fold. In a Cassie-Baxter type wetting regime at high static contact angles, the contact area between a water drop and a surface is minimized. This can reduce physical defects like water stains or any physico-chemical degradation such as corrosion. The primary benefit of superhydrophobic coatings towards outdoor applications however is their ability to “self-clean”. Water drops do not stick on the surface but roll off very easily. Due to the high surface roughness, dirt particles also show low adhesion and are effectively removed by the rolling drops [26–28]. It is therefore of primary importance to consider both the static contact angle as well as the dynamic behaviour of water drops when evaluating superhydrophobic surfaces in terms of application [1,29]. In our experiments we chose static contact angle and sliding angle measurements as indicators of the coating quality in terms of superhydrophobicity. For the outdoor experiment, these values are plotted in Fig. 1 as a function of time, along with the corresponding precipitation and temperature data [30]. The total precipitation of 1090 mm in the course of the 12-month outdoor exposure is conforming to the 30-year average of 1086 mm for the city of Zurich.

Fig. 1. Top: coating properties as a function of outdoor exposure time (, contact angle; 䊉, sliding angle; /, contact/sliding angle of a reference sample kept under ambient laboratory conditions). Middle: corresponding daily precipitation (bars) and integrated precipitation (line). Day 0 corresponds to 9 September 2004. Bottom: corresponding 7 days temperature average.

The sliding angle is much more sensitive than the contact angle, as was already noted in our studies on the chemical durability of the coating [23]. During the course of 12 months of outdoor exposure, the contact angle decreases to 148 ± 5◦ and the sliding angles increase to 69 ± 20◦ . A significant portion of this change occurs between day 50 and day 170. After 170 days, the contact and sliding angle values remain nearly constant. The scanning electron micrographs of the sample surface taken after 12 months of exposure (Fig. 2) show a strong, inhomogeneous change in surface structure. Compared to a freshly coated sample (Fig. 2A), the surface is inhomogeneously covered with spots and blotches of varying size and shape (Fig. 2B). Fig. 2C and D show magnifications of two distinct regions on the surface. A significant part

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

237

of the surface is virtually unchanged by the outdoor exposure (Fig. 2D) whereas the blotched areas are obviously contaminated (Fig. 2C). This is in accordance with the general opinion, that especially organic contaminants are a major problem in degradation of superhydrophobic surfaces under environmental conditions [5,19]. Obviously the self cleaning effect is not sufficient to remove these contaminants from the surface. Since the most pronounced degradation of surface properties coincides with a time of very little precipitation (winter season), we assume that in this time span contamination accumulates on the surface to an extent which can no longer be efficiently removed during the following rain period. Additionally, most of the precipitation in this period occurred in the form of snow which does not facilitate a self cleaning. At times of regular rainfall, fresh contaminants are more effectively removed from the surface and the coating properties are less affected (days 0–50, days 200–365). However, further investigations will be necessary to evaluate the explicit effect of regular rainfall on the coating properties during outdoor exposure. Also, given the coatings good stability towards detergent solution [23], removal of contaminants and regeneration of the surface by cleaning could be envisioned. Regardless of quantitative observations, the quality of the surface in respect to its extremely low water adhesion is still sufficient, even after 12 months of outdoor exposure. Fig. 3 shows images of the coated and uncoated glass slides taken by the

Fig. 2. Scanning electron micrographs of a silicone nanofilament coating after 12 months of outdoor exposure: (A) overview before exposure; (B) overview after 12 months of outdoor exposure; (C and D) magnification of areas indicated in (B). Fig. 3. Webcam images taken through an uncoated (top) and coated (bottom) glass slide during rainfall after 12 months of outdoor exposure.

238

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

webcam during rainfall at the end of the outdoor experiment (12 months). Raindrops adhere to the uncoated glass slide whereas the coated glass slide remains perfectly clear even though the coating properties have deteriorated. Raindrops are usually larger than 10 ␮l and possess kinetic energy when they come in contact with the surface. This explains why the coating is obviously still water repellent, although the sliding angle has increased significantly. The pinning forces between water and coating are large enough to pin a 10 ␮l sessile drop but are insignificant compared to the energy of falling raindrops. Snow adhesion was also observed to be significantly reduced for the coated substrate. The coated slide remained clear even during heavy snowfall whereas the uncoated slide was covered with snow and ice crystals. (For pictures taken during snowfall and videos taken with the webcam during rain see the Supplementary Material.) The anti-reflective properties observed for the silicone nanofilament coating [21] are only slightly affected by the outdoor exposure, in contrast to a significant decrease of transparency for the uncoated glass slide (Fig. 4). After 12 months of outdoor exposure, the coated glass slide shows an average of more than 5% higher transmittance in the visible range than the uncoated glass slide. The silicone nanofilaments are less affected by the outdoor weathering than the superhydrophobic coatings examined in related studies [5,18–20]. Naturally these comparisons are somewhat ambiguous because the coatings performance strongly depends on the environmental conditions and the setup of the outdoor experiments. For this reason we performed additional weathering tests under controlled laboratory conditions. Since no specific standards for the evaluation of the durability of superhydrophobic coatings under simulated environmental conditions exist to our knowledge, we chose procedures that are standards in related fields. For an assessment of UV durability ISO 9022-9 [25] was chosen, which is a standard for evaluating the influence of solar radiation on optics and optical instruments. For

Fig. 4. Transmittance of coated (solid lines) and uncoated (dotted lines) glass slides in the visible range. Upper curves indicate values before exposure, lower curves show values after 12 months of outdoor exposure, respectively.

Fig. 5. Contact and sliding angles as a function of ADF test duration. (䊉/) Sliding/contact angles of annealed samples and (/) sliding/contact angles of non annealed samples. For * please refer to the text.

the assessment of the durability against the combined influence of UV irradiation, temperature and (acid) rain or fog durability, the VDI 3958 guideline (ADF test) [24] was chosen. This test has proven to be convenient to simulate the effects of aggressive environmental conditions on automotive coatings with good correlation to outdoor weathering [31]. The VDI 3958 was given precedence over the ISO 11341 [32] that was used by Thieme et al. [20] since it accounts for acid precipitation and because the exact process of “moisturizing” the samples is not clearly defined in ISO 11341. Fig. 5 shows the development of contact and sliding angles of the annealed and non annealed samples during the ADF test. After 1 week the non annealed coatings show significant signs of deterioration. Contact angles have dropped below 150◦ and drops stick to the surface. Defects in the surface structure are clearly evident from the scanning electron micrographs (Fig. 6). At low magnification a multitude of circular regions, 10–20 ␮m in diameter, are distributed randomly on the whole surface (Fig. 6 top). Higher magnification (Fig. 6 bottom) reveals a slightly etched surface structure in these regions that is reminiscent of the changes observed during the coatings exposure to acidic solution [23]. We attribute these changes in surface structure to the workings of small acid drops that adhere to the surface after spraying with the acid solution in the first step of the ADF cycle. The drop size generated with the spray pump was determined by optical microscopy to be in the range of 30–70 ␮m. The corresponding contact diameter of the drops on the surface estimated with the spherical cap assumption and assuming a contact angle of 160◦ is in the range of 10–25 ␮m which corresponds well to the size of the deteriorated regions. Again we observe that an annealing step after coating significantly improves the durability of the silicone nanofilament coating [23]. Contact and sliding angles of the annealed samples indicate only a slight linear deterioration of the coating properties in the first 3 weeks of ADF testing. Contact angles drop to

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

239

Fig. 6. Scanning electron micrographs of non annealed samples after 1 week of ADF testing. Overview (top), and magnification of a degraded area (bottom).

Fig. 7. Scanning electron micrographs of an annealed coating after 4 weeks of ADF test at low (top) and high magnification (bottom).

153 ± 3◦ and sliding angles increase to 33 ± 8◦ . After 4 weeks the coating properties deteriorate to an extent that half of the samples show sliding angles of roughly 50◦ (indicated by * in Fig. 5) and on the other half of the samples drops remain pinned at 90◦ . Contact angles are now in the range of 147 ± 5◦ . All coatings exhibit a multitude of microscopic defect sites that become evident by little droplets of water adhering to the surface during rinsing. These defect sites however could not be identified by scanning electron microscopy (Fig. 7); no significant change in surface structure can be detected. The microscopically small etched regions observed in case of the non annealed samples cannot be observed on the annealed samples, presumably because the annealed coating is less susceptible to hydrolysis by acids than the non annealed coating [23]. The significantly improved resistance to artificial weathering of the annealed coating as opposed to the non annealed coating is most likely also due to the improved chemical durability. Annealing condensates residual OH-groups in the silicone leading to better hydrolytic stability and an increase in hydrophobicity [33]. The initial superhydrophobicity of the annealed coating is therefore superior to that of the non annealed coating, resulting in an improved initial resistance to weathering. In regard to their optical properties, all annealed coatings still show a higher transmittance in the visible range after 4 weeks of ADF test than freshly cleaned, uncoated glass slides. The non annealed samples develop a slight haze in the course of the ADF testing but the annealed samples stay perfectly clear.

Generally there is no indication of any structural damage on the coating caused by the continuous soft abrasive force of water drops acting on the surface during the rain phases of the ADF cycle. The laboratory weathering test performed in the WeatherOmeter according to DIN 9022-9 showed no influence of UV irradiation on the properties of the silicone nanofilament coating. Contact and sliding angle values remained virtually unchanged for both annealed and non annealed samples and scanning electron microscopy revealed no apparent change in surface structure. The optical properties also remained unchanged. 4. Summary and conclusion We have performed extensive weathering tests on a superhydrophobic silicone nanofilament coating under natural and artificial conditions. Under environmental conditions, contact angles remain near 150◦ and sliding angles below 90◦ for at least 12 months of outdoor exposure, the longest times ever reported for a coating of this kind. Both rain and snow do not adhere to the surface. The anti-reflective properties are maintained. Degradation through contamination appears to be the major factor under environmental conditions, but is not as significant as reported for other superhydrophobic coatings. Under laboratory conditions, non annealed coatings degrade within 1 week of artificial weathering according to the VDI 3958 Acid Dew and Fog Test. Acid precipitation is respon-

240

J. Zimmermann et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 302 (2007) 234–240

sible for an accelerated degradation of the surface properties. Annealing however significantly improves the durability of the coating and the annealed coatings resist at least 3 weeks of artificial weathering without a significant loss of superhydrophobic properties. The coating is unaffected by UV irradiation according to DIN ISO 9022-9. Given the exceptional durability of the superhydrophobic silicone nanofilament coating in outdoor and artificial weathering, along with its chemical stability [23] and range of additional interesting properties [21], we are confident that it constitutes a promising candidate for a variety of scientific and industrial applications. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.colsurfa.2007.02.033. References [1] [2] [3] [4] [5] [6] [7] [8] [9]

A. Nakajima, K. Hashimoto, T. Watanabe, Monatsh. Chem. 132 (2001) 31. R. Blossey, Nat. Mater. 2 (2003) 301. T.L. Sun, L. Feng, X.F. Gao, L. Jiang, Acc. Chem. Res. 38 (2005) 644. L. Feng, S.H. Li, Y.S. Li, H.J. Li, L.J. Zhang, J. Zhai, Y.L. Song, B.Q. Liu, L. Jiang, D.B. Zhu, Adv. Mater. 14 (2002) 1857. M. Callies, D. Qu´er´e, Soft Matter 1 (2005) 55. S. Shibuichi, T. Onda, N. Satoh, K. Tsujii, J. Phys. Chem. 100 (1996) 19512. M. Thieme, R. Frenzel, S. Schmidt, F. Simon, A. Hennig, H. Worch, K. Lunkwitz, D. Scharnweber, Adv. Eng. Mater. 3 (2001) 691. Y.Y. Wu, H. Sugimura, Y. Inoue, O. Takai, Chem. Vapor Depos. 8 (2002) 47. K.K.S. Lau, J. Bico, K.B.K. Teo, M. Chhowalla, G.A.J. Amaratunga, W.I. Milne, G.H. McKinley, K.K. Gleason, Nano Lett. 3 (2003) 1701.

[10] N.J. Shirtcliffe, G. McHale, M.I. Newton, C.C. Perry, Langmuir 19 (2003) 5626. [11] H.Y. Erbil, A.L. Demirel, Y. Avci, O. Mert, Science 299 (2003) 1377. [12] L. Feng, Y.L. Song, J. Zhai, B.Q. Liu, J. Xu, L. Jiang, D.B. Zhu, Angew. Chem. Int. Ed. 42 (2003) 800. [13] L. Jiang, Y. Zhao, J. Zhai, Angew. Chem. Int. Ed. 43 (2004) 4338. [14] H.M. Shang, Y. Wang, S.J. Limmer, T.P. Chou, K. Takahashi, G.Z. Cao, Thin Solid Films 472 (2005) 37. [15] W. Ming, D. Wu, R. van Benthem, G. de With, Nano Lett. 5 (2005) 2298. [16] J.T. Han, Y. Zheng, J.H. Cho, X. Xu, K. Cho, J. Phys. Chem. B 109 (2005) 20773. [17] S. Wang, L. Feng, L. Jiang, Adv. Mater. 18 (2006) 767. [18] A. Nakajima, K. Hashimoto, T. Watanabe, K. Takai, G. Yamauchi, A. Fujishima, Langmuir 16 (2000) 7044. [19] M. Sasaki, N. Kieda, K. Katayama, K. Takeda, A. Nakajima, J. Mater. Sci. 39 (2004) 3717. [20] M. Thieme, R. Frenzel, V. Hein, H. Worch, J. Corros. Sci. Technol. 6 (2003). [21] G.R.J. Artus, S. Jung, J. Zimmermann, H.-P. Gautschi, K. Marquardt, S. Seeger, Adv. Mater. 18 (2006) 2758. [22] S. Jung, G.R.J. Artus, J. Zimmermann, S. Seeger, Superhydrophobic Coating, WO2004113456, 2004. [23] J. Zimmermann, G.R.J. Artus, S. Seeger, Appl. Surf. Sci., 2007, in press. [24] VDI 3858 Part 12: Environmental simulation – effects of acidic precipitation on polymers – test methods, VDI Verein Deutscher Ingenieure, D¨usseldorf, 2004. [25] ISO 9022-9:1994-07-15, Optics and optical instruments – environmental test methods – Part 9: solar radiation, International Organization for Standardization, Geneva, 1994. [26] W. Barthlott, C. Neinhuis, Planta 202 (1997) 1. [27] R. F¨urstner, W. Barthlott, C. Neinhuis, P. Walzel, Langmuir 21 (2005) 956. [28] I.P. Parkin, R.G. Palgrave, J. Mater. Chem. 15 (2005) 1689. ¨ [29] W. Chen, A.Y. Fadeev, M.C. Hsieh, D. Oner, J. Youngblood, T.J. McCarthy, Langmuir 15 (1999) 3395. [30] Data supplied by MeteoSchweiz. [31] U. Schulz, Gefahrstoffe- Reinhaltung der Luft 1/2 (2004) 61. [32] ISO 11341:2004, Paints and varnishes – artificial weathering and exposure to artificial radiation – exposure to filtered xenon-arc radiation, International Organization for Standardization, 2004. [33] E.G. Rochow, Silicon and Silicones, Springer, Berlin, 1987.