Particle-Assisted Wetting - Langmuir (ACS Publications)

May 7, 2003 - In these wetting layers, the particles adsorb to and partially penetrate ... For a more comprehensive list of citations to this article,...
0 downloads 0 Views 184KB Size
4950

Langmuir 2003, 19, 4950-4952

Particle-Assisted Wetting Hui Xu† and Werner A. Goedel*,†,‡ Organic and Macromolecular Chemistry, OC III, University of Ulm, 89069 Ulm, Germany, and Polymer Research, Department of Polymer Physics, BASF Aktiengesellschaft, J542S, 67056 Ludwigshafen, Germany Received November 12, 2002. In Final Form: March 14, 2003 The wetting of a solid surface by a liquid is dramatically impeded if either the solid or liquid is decorated by particles. Here, we show that in the case of contact between two liquids the opposite effect can occur; mixtures of a hydrophobic liquid and suitable particles form wetting layers on a water surface, though the liquid alone is nonwetting. In these wetting layers, the particles adsorb to and partially penetrate through the liquid/air and/or liquid/water interface. This formation of wetting layers can be explained by the reduction in the total interfacial energy due to the replacement of part of the fluid/fluid interfaces by the particles. Furthermore, one can observe wetting layers of a thickness considerably larger than the particle diameter. This indicates that, in addition to their surfactant-like properties, particles adsorbed to an interface can compensate unfavorable long-range interactions.

The wetting or dewetting of surfaces by a liquid is a fascinating phenomenon of great importance for scientific and technological problems ranging from the self-protection of living organisms1 to the production of coatings and microstructures.2 The wetting of smooth surfaces generally is explained by an interplay of long-range and short-range forces3 and can be influenced by manipulating these forces, for example, by changing the refractive indices of the liquids involved,4 introducing surface-defective sites,5 or adding surfactants.6 Wetting is further impeded if the interfaces under consideration are decorated with particles. From recent publications on the so-called “lotus effect”, it is known that smooth solid surfaces bearing small wax particles are extremely hydrophobic.1 In a variation of this theme, it was shown that coating liquid droplets with small solid particles tremendously reduces their tendency to wet a smooth surface.7 Here, we report the surprising discovery of the opposite effect: mixtures of water-insoluble liquids and suitably hydrophobized particles form a wetting layer on a water surface, though the liquid alone does not wet it. These observations are shown as electron microscopy images in Figure 1 using polymerizable liquids, which can be photopolymerized and transferred to solid substrates before imaging. Like most organic liquids, the liquids used here do not wet a water surface but rather form lenses, as is shown in Figure 1a for the liquid trimethylolpropane trimethacrylate (TMPTMA; for the chemical structure, see * Corresponding author. † University of Ulm. ‡ BASF Aktiengesellschaft. (1) Barthlott, W.; Neinhuis, C. Planta 1997, 202, 1-8. (2) Gau, H.; Herminghaus, S.; Lenz, P.; Lipowsky, R. Science 1999, 283, 46-49. (3) Diedrich, S. In Phase Transition and Critical Phenomena; Domb, C., Lebowitz, J., Eds.; Academic Press: London, 1988; Vol. 12. Schick, M. In Liquids at Interfaces; Charvolin, J., Ed.; Elsevier: Amsterdam, 1989. Blockhuis, E. M.; Widom, B. Curr. Opin. Colloid Interface Sci. 1996, 1, 424-429. (4) Ragil, K.; Meunier, J.; Broseta, D.; Indekeu, J.; Bonn, D. Phys. Rev. Lett. 1996, 77, 1532-1535. (5) Sun, R.-D.; Nakajima, A.; Fujishima, A.; Watanabe, T.; Hashimoto, K. J. Phys. Chem. B 2001, 105, 1984-1990. (6) Hill, R. M. Curr. Opin. Colloid Interface Sci. 1998, 3, 247-254. (7) Aussillous, P.; Quere, D. Nature 2001, 411, 924-927.

Figure 1a). The hydrophobized silica particles used in our investigation form stable monolayers on a water surface, as is shown in Figure 1b for the methacrylate-coated silica particles. If we apply mixtures of the liquid TMPTMA and methacrylate-coated silica colloids to a water surface, we observe the formation of a uniform mixed layer in which the particles adhere to the water surface and the organic liquid fills the volume between the particles (see Figure 1c,d). In these layers, the particles are visible from both the top and bottom of the mixed wetting layers at a liquidto-particle weight ratio of 0.3-0.5 (Figure 1d,e). The particles are completely covered by the liquid and adhere only to the bottom interface if the liquid-to-particle weight ratio is higher than 0.5 (Figure 1f). We observe further variations of this behavior if we change the liquid or the surface coating of the particles. For example, if we choose, instead of TMPTMA, the liquid pentaerythritol triacrylate (PETA; for the chemical structure, see Figure 1g), we can obtain mixed layers with particles being adsorbed simultaneously to the top and bottom interfaces (Figure 1g). On the other hand, if we change the surface coating of the particles from methacrylate groups to oligoisobutene chains, the particleassisted wetting does not occur anymore; instead, we observe the coexistence between lenses of the liquid and irregular aggregates of the particles on the water surface (Figure 1h-i). Given the fact that particles can prevent the wetting of a solid surface,7 it seems at first glance puzzling that we observe the contrary; here, particles assist the wetting of a surface. However, one major difference between our case and former investigations is the fact that we investigate here the wetting of a fluid interface. Particles have a strong tendency to adsorb to fluid/fluid interfaces. They easily form stable colloidal monolayers8,9 and have been used as stabilizers for oil/water emulsions.10-13 Part of the driving (8) Aveyard, R.; Clint, J. H.; Nees, D.; Paunov, V. N. Langmuir 2000, 16, 1969-1979. (9) Ghezzi, F.; Earnshaw, J. C. J. Phys.: Condens. Matter 1997, 9, 517-523. (10) Pickering, S. U. J. Chem. Soc. 1907, 91, 2001-2021. (11) Velev, O. D.; Furusawa, K.; Nagayama, K. Langmuir 1996, 12, 2374-2384.

10.1021/la026833f CCC: $25.00 © 2003 American Chemical Society Published on Web 05/07/2003

Wetting

Langmuir, Vol. 19, No. 12, 2003 4951

Figure 2. Schematic representation of the different possible scenarios involved in the particle-assisted wetting: (a) coexistence of liquid lenses and “dry” particles at high liquid/particle contact angles, (b) complete incorporation of the particles into the liquid at low contact angles, and (c-f) different scenarios of particle-assisted wetting at intermediate contact angles.

Figure 1. Scanning electron microscopy images of the organic liquid, silica colloids, and their mixtures after application to a water surface, UV-irradiation, and transfer to mica substrates. (a) Lenses formed by the organic liquid TMPTMA (the chemical structure is depicted above part a). (b) Monolayer formed by methacrylate-coated silica colloids. (c) Top view of a wetting layer formed from TMPTMA mixed with methacrylate-coated silica colloids. The crack (indicated by a white arrow) was intentionally included to show the existence of the film on the surface. (d) A magnified picture of part c. (e) Bottom view of the same mixed layer. (f) Side view of a thick wetting layer formed from methacrylate-coated silica particles and TMPTMA (to increase the contrast, the silica particles have been etched away with hydrofluoric acid). (g) Wetting layer formed from a mixture of methacrylate-coated silica particles and the organic liquid PETA (the chemical structure is depicted above part g). (h) Structures formed by a mixture of TMPTMA and silica colloids coated with oligoisobutene chains (the chemical structure of the coating is depicted above the image). (i) A magnified image of the particle aggregates in part e (indicated by a white box). The scale bar is 50 µm for parts a, c, and h and 2 µm for parts b, d-g, and i.

force for this adsorption is the fact that the particles partially replace the fluid/fluid interface and, thus, reduce the total interfacial energy of the system. In absence of particles, the energy change per unit surface, ∆Espreading/a, associated with the formation of a thick wetting layer, is equivalent to the negative value of the spreading coefficient S and can be calculated from the difference between the interfacial tension of the surface to be covered γ0 and the combined interfacial tensions of the top and bottom interfaces of the wetting layer: ∆Espreading/a ) -S ) (γtop + γbottom) - γ0. For most organic liquids at the water surface, S is negative (∆Espreading is positive), and wetting does not occur.14 For example, for the liquid TMPTMA, one can estimate S ) -0.1 × 10-3 N/m. The adsorption of particles to one of the interfaces gives rise to a gain in energy per particle, which has its maximum value at a contact angle of 90° [for a hexagonally closed packed layer of particles, ∆Eparticle adsorption/a ) -(π/x12)γbottom(1 - sin |90° - Θ|)2].13 While a more detailed mathematical description currently is in preparation,15 the basic results can be summarized schematically in Figure 2: if the contact angles of the particles with the liquid/air interface and with the liquid/water interface are close to 180°, the liquid forms lenses on the water surface that are separate from the particles (Figure 2a). If both contact angles are close to 0°, the liquid will form lenses, which completely incorporate the particles (Figure 2b). If the contact angles at one or both of the interfaces are at intermediate values, particle-assisted wetting can occur (Figure 2c-f), and the particles adhere preferentially to the interface having the contact angle closest to 90° (Figure 2e-f). If the volume of the liquid is smaller than the volume of the particles (Figure 2c) or if the particle/liquid contact angles at both interfaces are similar (Figure 2d), the particles adhere to both interfaces of the wetting film. In our experiments, we observe four of the six scenarios depicted in Figure 2 (parts a, c, d, and f). From the scanning force microscopy images of the corresponding structures (images not shown), we can estimate the contact angles at the top and bottom interfaces of the wetting layers. For the system of (12) Binks, B. P.; Lumsdon, S. O. Phys. Chem. Chem. Phys. 1999, 1, 3007-3016. (13) Binks, B. P.; Lumsdon, S. O. Langmuir 2000, 16, 8622-8631. (14) Davies, J. T.; Rideal, E. K. Interfacial Phenomena; Academic Press: London, 1961; pp 21-25. (15) Goedel, W. A. Europhysics Lett. 2003, 62 (4), 607-613.

4952

Langmuir, Vol. 19, No. 12, 2003

TMPTMA/methacrylate-coated particles shown in Figure 1c-f, the contact angle at the bottom interface (50°) is significantly larger than the contact angle at the top interface (25°). This is in good agreement with the preferential adsorption of the particles to the bottom interface observed in Figure 1f. Furthermore, using the equation given previously, one can calculate the change in interfacial energy due to the adsorption of the particles to the bottom interface, ∆Eparticle adsorption/a ) -8 × 10-3 N/m. This gain in interfacial energy more than compensates for the unfavorable spreading coefficient of S ) -0.1 × 10-3 N/m. For the system of PETA/methacrylate-coated particles shown in Figure 1g, we obtain rather similar contact angles at the bottom and top interfaces (23 and 20°, respectively), in good agreement with the adsorption of the particles to both interfaces. Until now, we have only discussed the surfactant-like action of the particles. However, the formation of a wetting layer is a function of both short- and long-range forces. The long-range forces give rise to an either attractive or repulsive force between the two interfaces of a wetting layer. The force per area, the disjoining pressure Π, is given by the negative first derivative of the free energy with respect to the layer thickness. In simple systems, it is proportional to the Hamaker constant A and the third power of the layer thickness D of the considered system (Π ) -A/6πD3).16 A positive Hamaker constant indicates attraction between the two interfaces of a thin layer and eventually dewetting; a negative Hamaker constant indicates repulsion between the interfaces and stable wetting layers. The (nonretarded) Hamaker constant for the system of water/TMPTMA/air is positive (+1.6 × 10-20 J), indicating the instability of a wetting layer of TMPTMA on a water surface. Thus, at first glance, one would not expect the formation of a wetting layer that substantially exceeds the dimensions of the particles, such as the one depicted in Figure 1f. On the other hand, the (nonretarded) Hamaker constant for the system of silica/TMPTMA/air is negative (-9.4 × 10-22 J). Thus, in the system investigated here, one obtains a competition between attractive and repulsive long-range interactions. If we represent our particles as a platelet of silica of thickness d ) 300 nm on top of a bulk-water subphase, we can estimate the long-range interactions in a first approximation using expressions derived from ref 17 [π ) -∂Φ(D)/ ∂D; Φ(D) ) -Asilica/TMPTA/air/12πD2 - (Awater/TMPTA/air - Asilica/ 2 TMPTA/air)/12π(D + d) ] for thin and thick TMPTMA layers on top of the silica.18 These calculations yield repulsive disjoining pressures. Thus, the particles adsorbed to the liquid/liquid interface influence the long-range interactions in favor of wetting. In summary, we have demonstrated that particles can assist the wetting of a liquid by a second liquid. This phenomenon is due to the “surfactant-like” interactions of particles adsorbed to fluid/fluid interfaces and is further enhanced as a result of a favorable modification of the long-range interactions caused by the adsorbed particles. Particle-assisted wetting has been demonstrated here using a combination of an organic liquid and water, but it might as well be applied to a large variety of liquids and (16) Israelachvili, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic Press: London, 1991, pp 176-212. (17) Seemann, R.; Herminghaus, S.; Jacobs, K. J. Phys.: Condens. Matter 2001, 13, 4925-4938. (18) With an increasing layer thickness, the influence of retardation effects on the Hamaker constants has to be taken into account. See Russel, W. B.; Saville, D. A.; Schowalter, W. R. Colloidal Dispersions; Cambridge University Press: New York, 1989; pp 153-156.

Xu and Goedel

interfaces such as complex metal or polymer alloys or to problems such as flotation in oil/water mixtures. The concept might even be further extended to include deformable solids, for example, by replacing either the “water” phase or the particles by a soft or elastic substance. Thus, particle-assisted wetting reveals a new perspective in colloid and surface science and could open up a variety of new applications. Experimental Section The photopolymerizable organic liquids were mixtures of either TMPTMA (Aldrich) or PETA (Aldrich) and 3-5 wt % of the photoinitiator benzoin isobutyl ether (Aldrich). Silica colloids19 (Si-TPM, 320-nm diameter, 6% polydispersity, coated by treatment with methacryloxypropyltrimethoxysilane; SAP230, 140nm diameter, 11% polydispersity, coated by treatment with oligoisobutene amphiphiles) were obtained as dispersions from the Utrecht Colloid Synthesis Facility, The Netherlands. Mixtures of the organic liquid and colloid dispersions with chloroform and ethanol (the final solution contained 20% ethanol and approximately 1 wt % of the organic liquid and particles) were applied to a Langmuir trough filled with water (equipment was identical to that described in ref 20). After the evaporation of the solvent and lateral compression to a surface pressure of 15-20 × 10-3 N/m, the structures were solidified by photopolymerization and transferred to mica substrates using horizontal transfer.21 Electron microscopy images were obtained using a Zeiss DSM 962 scanning electron microscope. Scanning force microscopy images were recorded with a NanoScope III (Digital Instruments, Santa Barbara, CA) operated in the tapping mode at a resonance frequency of about 320 kHz. The measurements were performed at ambient conditions using Si probes with a spring constant of ∼50 N/m. The nonretarded Hamaker constants (Awater/TMPTMA/air ) +1.6 × 10-20 J, Asilica/TMPTMA/air ) -9.4 × 10-22 J) were calculated, as was described in ref 16, from the refractive indices of the compounds (supplier’s data) and static dielectric constant (for water and silica, literature data were used; for TMPTMA,  ) 4.57 was measured with equipment described in ref 22). Retardation effects were calculated according to ref 18. The interfacial tensions of TMPTMA against air and water were measured with a Kru¨ss interfacial tension analyzer. The interfacial tension of water in the presence of TMPTMA was measured with a KSV Sigma70 tensiometer.

Acknowledgment. We thank Carlos van Kats, Judith Wijnhoven, Albert Philipse, and Alfons van Blaaderen (Utrecht Colloid Synthesis Facility, Utrecht University, The Netherlands) for providing the silica colloids and valuable discussions. We also thank Professor Paul Walther (Center for Electron Microscopy, University of Ulm) for helping us image our samples with highresolution scanning electron microscopy, B. Stoll for measuring the dielectric constant of TMPTMA, and A. Ding for measuring the interfacial tensions. The investigations have been conducted in the Laboratory for Organic and Macromolecular Chemistry of the University of Ulm, and the support by Martin Mo¨ller is greatly appreciated. H.X. thanks the Alexander von Humboldt Foundation for a fellowship. W.A.G. thanks R. Iden and H. Auweter for support. This work was funded by the Deutsche Forschungsgemeinschaft through the SFB “Hierarchic Structure Formation and Function of OrganicInorganic Nano Systems” and the SPP “Wetting and Structure Formation at Interfaces”. LA026833F (19) Philipse, A. P.; Vrij, A. J. Colloid Interface Sci. 1989, 128, 121136. Pathmamanoharan, C. Colloids Surf. 1988/89, 34, 81-88. (20) Goedel, W. A.; Heger, R. Langmuir 1998, 14, 3470-3474. (21) Araki, T.; Oinuma, S. I.; Iriyama, K. Langmuir 1991, 7, 738744. (22) Heinrich, W.; Stoll, B. Colloid Polym. Sci. 1985, 263, 873-878.