pubs.acs.org/Langmuir © 2009 American Chemical Society
Contact Angle Assessment of Hydrophobic Silica Nanoparticles Related to the Mechanisms of Dry Water Formation Laurent Forny,† Khashayar Saleh, Renaud Denoyel,‡ and Isabelle Pezron* Universit e de Technologie de Compi egne, EA 4297 Transformations Int egr ees de la Mati ere Renouvelable, B.P. 20529, 60205 Compi egne Cedex, France. †Present address: Nestl e Research Center, Department of Food Science and Technology, Vers-chez-les-blanc, 1000 Lausanne 26, Switzerland. ‡ Laboratoire Chimie Provence, CNRS-Universit es de Aix-Marseille I,II et III, Centre de Saint J er^ ome, Avenue Escadrille-Normandie-Niemen, 13397 Marseille Cedex 20, France Received July 27, 2009. Revised Manuscript Received September 23, 2009 Dry water is a very convenient way of encapsulating a high amount of aqueous solutions in a powder form made of hydrophobic silica nanoparticles. It was demonstrated in previous studies that both solid and liquid interfacial properties influence the quality of the final product resulting occasionally in mousse formation. To explain this behavior, contact angles of silica nanoparticles have been measured for water and water/ethanol solution by means of liquid intrusion experiments. It was found that the quality of the final product correlates with the contact angle, i.e., contact angle close to 105° leads to mousse formation whereas a slightly higher value of approximately 118° allows dry water formation. The proposed explanation was based on the energy of immersion and adhesion defined as the energy needed for a spherical particle to respectively penetrate into the liquid or attach at the liquid/air interface. Significantly lower energy of immersion calculated for lower contact angle might account for particle penetration into the liquid phase during processing, leading to continuous network aggregation, air entrapment, and finally mousse formation.
Introduction “Dry water” was first mentioned in a patent published in 1964.1 This powder contains up to 98% by weight of water but still exhibits dry flow properties. It may be produced at industrial scale by high shear mixing of hydrophobic fumed silica and water. The structure of individual grains of dry water has been studied by electronic microscopy after freeze fracture or water sublimation.2 It consists in micrometric water droplets, of few hundred micrometers, surrounded by a network of branched hydrophobic fumed silica nanoparticles. The water phase may be released by evaporation or under mechanical stress. The final product is a very convenient carrier system since various active agents may be incorporated into the water phase. Surprisingly, little attention has been devoted to this system until the mid 1990s, and the rapid growth of the cosmetic industry.3-5 Still, only few scientific studies were dedicated to better describe the process, even if potential applications are probably numerous in the food, pharmaceutical, and other industries.6,7 The influence of mixing characteristics on the quality of the final product, which can be obtained as a powder but also as a mousse or a solid suspension, was studied in a previous paper.8 The main parameters were shown to be the energetic contribution to the process and the affinity between the solid and the liquid *Corresponding author. E-mail:
[email protected]; tel: 33(0) 3 44 23 46 18; fax: 33 (0) 3 44 23 19 80. (1) Br€unner, H.; Schutte, D.; Schmitz, F.-T. German Patent DE1467023, 1964. (2) Forny, L.; Pezron, I.; Saleh, K.; Guigon, P.; Komunjer, L. Powder Technol. 2007, 171, 15–24. € (3) Hasenzahl, S.; Gray, A.; Walzer, E.; Braunagel, A. SOFW-J. 2005, 131, 1–8. (4) Dampeirou, C. International Patent WO2005034917, 2005. (5) Lahanas, K. M.; Vrabie, N.; Santos, E.; Miklean, S. U.S. Patent US6290941B1, 2001. (6) Binks, B. P.; Murakami, R. Nat. Mater. 2006, 5, 865–869. (7) Wang, W. X.; Bray, C. L.; Adams, D. J.; Cooper, A. L. J. Am. Chem. Soc. 2008, 130, 11608–11609. (8) Forny, L.; Saleh, K.; Pezron, I.; Komunjer, L.; Guigon, P. Powder Technol. 2009, 189, 263–269.
Langmuir 2010, 26(4), 2333–2338
phases. Experiments were performed with different grades of hydrophobic silica, whose hydrophobic nature was originally characterized by the manufacturer with a methanol wettability method.3 When mixed with pure water in a blender, which is considered as a high shear process, hydrophilic fumed silica produces a suspension, moderately hydrophobic fumed silica produces a mousse, and only highly hydrophobic fumed silica produces a powder. In addition, water-soluble additives can be easily encapsulated into powders as long as they do not significantly reduce the surface tension of the aqueous solution.8 Any change in the liquid/solid affinity may affect the quality of the product obtained. As an example, when blending different water/ ethanol mixtures with highly hydrophobic fumed silica, a powder form is obtained only if the surface tension remains higher than 50 mN 3 m-1. Below 50 mN 3 m-1, using for instance a 92/8 w/w water/ethanol mixture, a mousse is obtained, and below 38 mN 3 m-1, using an 80/20 w/w water/ethanol mixture, a suspension is formed. These observations show that the physicochemistry of the system plays a major role in the process, and it became necessary to better characterize the solid/liquid affinity by an accurate measurement of their contact angle. Our main objective was then to find the appropriate method to evaluate the solid/liquid contact angles in order to better understand the mechanisms leading to the formation of dry water during the process. Contact angles of liquids on finely divided solids are difficult to determine, even more when dealing with water interacting with hydrophobic solids.9,10 Direct methods, as the sessile drop method on compressed tablets, present some drawbacks. Wetting properties are influenced not only by the surface chemistry of the solid but also by the surface roughness resulting from the (9) Chibowski, E.; Perea-Carpio, R. Adv. Colloid Interface Sci. 2002, 98, 245– 264. (10) Lazghab, M.; Saleh, K.; Pezron, I.; Guigon, P.; Komunjer, L. Powder Technol. 2005, 157, 79–91.
Published on Web 10/19/2009
DOI: 10.1021/la902759s
2333
Article
Forny et al.
Figure 1. Picture of a water droplet deposited on a rough hydrophobic fumed silica surface obtained by powder compaction.
Figure 3. Experimental setup used for intrusion experiments.
Figure 2. Top: Typical shape of a water droplet deposited on (a) a wax surface before crystallization (smooth) and (b) on the same wax surface after crystallization at 50 °C for 2 h (formation of a rough fractal structure). Bottom: scanning electron microscopy (SEM) pictures of the solid surfaces.
fractal structure of the particles as well as the compression conditions. Figure 1 shows the typical shape of a water droplet deposited on a rough tablet made of fumed silica. This effect can be also observed on certain waxes before and after crystallization, i.e. on smooth amorphous versus fractal crystalline surfaces as shown in Figure 2.11 Contact angle on rough surfaces appears to be close to 180°, whereas contact angle on a smooth surface is much lower. This phenomenon is known as the lotus effect, and the related surfaces are referred to as superhydrophobic.12 Therefore, contact angles measured by the sessile drop method on compressed fumed silica tablets are much higher than the expected values on similar smooth surfaces and do not allow discrimination between the different grades of particles. Indirect determinations of water contact angle on hydrophobic divided solids have been performed with other techniques. Inverse gas chromatography has been used to characterize surface properties through the interaction of the surface with different gaseous phases. Contact angles are calculated from surface free energy values, which require a model for the polar interactions, for instance, by introducing the acid-base contributions.13 Immersion microcalorimetric methods are not straightforward for nonwetting liquids, for which the enthalpy of immersion is extrapolated at zero concentration from values obtained with a binary solution dispersing the particles.14,15 (11) Minami, T.; Mayama, H.; Nakamura, S.; Yokojima, S.; Shen, J.-W.; Tsujii, K. Soft Matter 2008, 4, 140–144. (12) de Gennes, P. G.; Brochard-Wyart, F. ; Quere, D. Capillarity and Wetting Phenomena: Drops, Bubbles, Pearls, Waves; Springer: New York, 2004; Chapter 1. (13) Elzer, F. M. Characterization of surface free energies and surface chemistry of solids. In Contact Angle Wettability and Adhesion; Mittal, K.L., Ed.; VSP: Utrecht, The Netherlands, 2003; Vol. 3, pp 219-264. (14) Yan, N.; Maham, Y.; Masliyah, J. H.; Gray, M. R.; Mather, A. E. J. Colloid Interface Sci. 2000, 228, 1–6. (15) Spagnolo, D. A.; Maham, Y.; Chuang, K. T. J. Phys. Chem. 1996, 100, 6626–6630.
2334 DOI: 10.1021/la902759s
The method that we found the most appropriate to overcome the various encountered difficulties consists in water intrusion experiments in tablets made of fumed silica16 (see Figure 3). In this technique, similar to mercury porosimetry, the minimum pressure required for the water phase to penetrate the porous solid is measured and related to the solid/liquid contact angle. High pressure conditions force the liquid to explore the entire solid surface, preventing the physical artifact observed with the sessile drop technique, and, in consequence, the results obtained are directly representative of the interactions between water and the solid surface. Resulting contact angle values were then used to calculate interaction energies between the aqueous phase and the solid, which were correlated to the structure of the product obtained during “dry water” processing.
Material and Methods Material. Fumed silica particles from the Aerosil range were supplied by Evonik Degussa (Germany). Aerosil R812S and Aerosil R972, which are pure nanoscale silicon dioxide aggregates grafted by hexamethyl-disilazane and dimethyl-dichlorosilane respectively, were used in this study. Further description and characteristics of these particles may be found elsewhere.2,3 Millipore Milli-Q water and pure ethanol (99.8% grade, Carlo Erba) were used for liquid intrusion experiments. Sample Preparation. Tablets of fumed silica were prepared using a cylindrical stainless steel die of 1 cm in diameter. A pressure of 250 MPa was applied for 10 min using a manual hydraulic press (Socachim-XRF Scientific, Belgium). When the pressure was removed, expansion of the tablets occurred, leading to its fragmentation in smaller pieces. These fragments were kept in a closed plastic container before liquid intrusion experiments or nitrogen adsorption analysis. The porosities calculated from the porous volume assessed by nitrogen adsorption are 66% and 72% for tablets made of Aerosil R812S and R972, respectively. These values are rather high if compared with other data of the literature on silica. The specific surface areas of the pellets were found to be in good agreement with the manufacturer’s specifications (Table 1). Therefore, we believe that the particles were not sintered during compaction as is usually observed with other types of particles. This might be that due to the presence of the grafting layer and specific shapes of the particles; there is strong relaxation of the structure. Nevertheless, our objective here was not to study the influence of preparation conditions on the pellet density, but simply to prepare a mesoporous solid that can be analyzed both by gas adsorption and water intrusion. Setup for Liquid Intrusion. The homemade experimental setup for liquid intrusion was composed of a syringe pump (Model 100DM, manufactered by Isco, USA) connected to a high-pressure stainless steel cell. This pump allows the control of both the intruded volume and the pressure in the range of 0-700 bars. The high-pressure cell is placed into the thermopile of a (16) Gomez, F.; Denoyel, R.; Rouquerol, J. Langmuir 2000, 16, 4374–4379.
Langmuir 2010, 26(4), 2333–2338
Forny et al.
Article
Table 1. Specific Surface Area, Mean Pore Size, and Total Porous Volume Obtained from Sorption Isotherm Analysis on Tablets Made of Aerosil R812S and Aerosil R972 Used for Liquid Intrusion Experiments total porous BETa specific surface mean BJHb pore width/nm volume/cm3 3 g-1 area/m2 3 g-1 Aerosil R812S 214 Aerosil R972 124 a BET = Brunauer-Emmett-Teller. Halenda.
16 0.92 37 1.2 b BJH = Barrett-Joyner-
Tian-Calvet type microcalorimeter, which allows recording heat exchange during the intrusion of the liquid. About 0.5 g of tablets was introduced into the high-pressure cell and placed into the microcalorimeter maintained at 25 °C. The sample was then evacuated (1 Pa) during at least 12 h. Liquid was then introduced by means of the syringe pump. First a large amount of liquid was injected in order to fill the dead volume and reach a pressure of approximately 60 kPa. Intrusion was then carried out by steps of 40 μL. The intrusion was stopped at high pressure (above 10 MPa) after almost vertical rise of the quasi-equilibrium pressure. Extrusion was performed with the same steps of 40 μL until the pressure reached 0.1 MPa. Raw data had to be corrected from the effects of liquid compressibility. This was achieved by carrying out blank compression experiments in an empty cell. The corrected volume is indeed the intruded volume minus the compression volume of the whole.
Contact Angle and Pore Size Distribution Determination. Contact angles were assessed from intrusion measurement of nonwetting liquid into the pores of tablets. This method was described by Gomez et al.16 It is based on the Laplace-Washburn equation that gives the pressure drop ΔP required for the intrusion of a nonwetting liquid of surface tension γ in cylindrical pores of radius r: 2γ cos θ ΔP ¼ - 3 r
ð1Þ
Assuming that the pores of the tablets are cylindrical and that the saturating vapor pressure of the liquid can be neglected as compared to intrusion pressure (moreover, there is no air in the system since the tablets are initially evacuated), the external pressure p applied upon intrusion is then equal to ΔP. Provided that the mechanical equilibrium is reached, this equation can be applied to a simple intrusion experiment where the pressure p is continuously monitored together with the intrusion volume to assess stepwise the pore size distribution. Mean contact angle is then obtained by fitting this distribution with the pore size distribution assessed by nitrogen adsorption, as described below. The surface tension γ was measured with the Wilhelmy plate technique (Kr€ uss K100, Hamburg, Germany). Calorimetric Analysis of the Intrusion Experiment. The calorimetric data recorded during the intrusion experiment may be postanalyzed to assess the volumetric pore size distribution. The complete analysis was described by Gomez et al.16 The route considers that the isothermal change dUT,V in internal energy of the capillary system (liquid þ porous solid) can be split into a bulk and an interfacial contribution and focuses the attention on the second term. The bulk contribution is calculated from the work of compression: Z Wcomp ¼ - p 3 dV ð2Þ and of the corresponding heat of compression Z m 3 h 3 dP Qcomp ¼ Langmuir 2010, 26(4), 2333–2338
where p is the intrusion pressure, V is the intruded volume, m is the mass of the bulk phase and h is its specific enthalpy of compression, as known from the tables (for water, h = -63 mJ 3 g-1 3 MPa-1). After correcting the experimental work and heat of intrusion from these bulk contributions, the remaining change in internal energy only results from the interfacial contribution:
ð3Þ
dUT;V ¼
X DU DA
dA T, V
ð4Þ
P where encompasses the various types of interfaces (i.e., solidvapor, solid-liquid, and liquid-vapor). In the intrusion/extrusion process, it is assumed that the total area of the liquid/vapor interface does not change, and that only the solid/vapor interface is converted into a solid/liquid interface of identical area. The variation of internal energy dU may then be expressed as a function of the area variation dA, by introducing a parameter, β, which depends on the surface tension of the liquid γ and the contact angle θ: dU ¼ β 3 dA
ð5Þ
where β ¼ T 3 cos θ
Dγ D cos θ þ T 3γ -γ 3 cos θ DT DT
ð6Þ
Equation 5 can be applied along a reversible path. It shows that the interfacial internal energy is proportional to the wetted surface area, provided that the interfacial tensions are independent of pressure. The surface must also be homogeneous, i.e., the contact angle is assumed to be the same everywhere in the solid. Then, between two equilibrium states, one can write ΔU ¼ β 3 ΔA
ð7Þ
Consequently, the term β may be derived from ΔUtotal, which is the total internal energy change on wetting (assessed from the measurements of pressure, volume, and heat), and from ΔAtotal, which is the total internal area of the pores accessible to the wetting liquid (assessed by applying the BET method to nitrogen adsorption measurement). Once β has been calculated, the pore-size distribution may be assessed stepwise assuming cylindrical pores of radius r: dV dU ¼ r 2β
ð8Þ
This pore size distribution may be compared to the pore size distribution directly obtained from the Laplace-Washburn equation to cross-check the accuracy and the homogeneity of the contact angle.
Surface Area and Pore Size Distribution by Nitrogen Adsorption. Physisorption of nitrogen was carried out at 77 K using an ASAP 2010 apparatus (Micromeritics, USA). Before adsorption, samples were evacuated during 10 h at 80 °C at a pressure lower than 10-3 Pa. The surface areas and mesopore size distributions were calculated using the BET and the BJH methods, respectively. The BJH method was applied to the desorption branch of the adsorption isotherm.17
Results Experimental results of water intrusion/extrusion in both Aerosil R812S and Aerosil R972 tablets are shown in Figure 4. (17) Rouquerol, F.; Luciani, L.; Llewellyn, P.; Denoyel, R.; Rouquerol, J. Tech. Ing. 1994, P1050, 1–24.
DOI: 10.1021/la902759s
2335
Article
Figure 4. Intrusion/extrusion pressure at 25 °C as a function of net volume of water intruded into tablets of Aerosil R812S (þ) and Aerosil R972 ().
Forny et al.
Figure 6. Normalized cumulative distributions of pore volume obtained from Laplace-Washburn equation (circle) and from BJH analysis of sorption isotherms (square) for Aerosil R812S (filled symbol) and Aerosil R972 (open symbol) assuming contact angles of 118° and 105°, respectively.
Pressure is plotted versus the corrected amount of intruded water per gram of solid, i.e., taking into account water compressibility. Initially, pressure rises significantly up to a plateau. This corresponds to the filling of both the external porosity and the macropores. Afterward, the pressure remains almost stable, which corresponds to the progressive filling of the mesopores. Finally, all the porosity is filled, and the pressure rises quickly. During extrusion phase, most liquid remains entrapped in the sample as it is sometimes observed in hydrophobic pores.18 Slight crossover of intrusion-extrusion curves may be related to the accuracy of the syringe pump. To be able to calculate the contact angle, pore size distribution was assessed by means of nitrogen adsorption analyses. In Figure 5, the sorption isotherms of Aerosil R812S and Aerosil R972 show that capillary condensation occurs in a narrow range of pressure, which indicates a narrow pore size distribution. The BET areas reported in Table 1 are in good agreement with specifications from the manufacturer (220 ( 25 m2 3 g-1 for Aerosil R812S and 110 ( 20 m2 3 g-1 for Aerosil R972). This may indicate that no structural deterioration of the initial powder has occurred during the tablet preparation. Mean BJH pore size is also reported in Table 1. Both the BJH method and Laplace-Washburn equation assume those pores are cylindrical. This is a rough approximation
for the present porous systems that are made of compacted aggregates.2 Nevertheless, for the present purpose, the most important point is to use the same hypothesis for the two approaches. A good agreement between pore size obtained by gas adsorption and mercury intrusion porosimetry has been observed in the case of well-defined monodisperse silica spheres.19 It would be necessary in the future to confirm this statement for more complex particles such as fumed silica particles. Tablets made of Aerosil R812S were found to have much smaller pores than tablets made of Aerosil R972. This may be due to different sizes of aggregate (approximately 270 nm for Aerosil R812S versus 360 nm for Aerosil R972) as well as different sizes of sintered primary particles forming the aggregate (7 nm for Aerosil R812S versus 16 nm for Aerosil R972).2 Normalized cumulative pore size distributions recalculated from eq 1 and intrusion experiments data were fitted to the normalized cumulative pore size distributions obtained from nitrogen adsorption by means of the contact angle (Figure 6). Contact angle was found to be 118° for Aerosil R812S and 105° for Aerosil R972. Systematic reproducibility analysis could not be done in the frame of this study. However, if the pore size distribution obtained by water intrusion is plotted for various values of the contact angle, there is an overlap between the two distributions for contact angles ranging between 112° and 121° for R812S and between 102° and 109° for R972. The contact angle measured in the same way between 8% ethanol solution and Aerosil R812S was found to be approximately 108°. Measured contact angles with water are in agreement with the supplier’s specification, i.e., Aerosil R812S is more hydrophobic than Aerosil R972.3 Very few studies tried to assess contact angles of fumed silica since classical methods are not well adapted to hydrophobic nanoparticles. Yan et al.14 used a calorimetric method based on the enthalpy of immersion in n-propanol/water mixtures to aid in dispersing the nanoparticles. The contact angle values they obtain, for Aerosil grades which are similar but not exactly the same, i.e., R812 and R974, are slightly higher (118° and 117°, respectively). In the case of the calorimetric method, the enthalpy of immersion is obtained by extrapolation of the results to zero n-propanol concentration and therefore, the way of assessing the contact angle is not as straightforward as for the intrusion method. Cumulative pore volume distributions obtained from the Laplace-Washburn equation were also compared with pore size
(18) Lefevre, B.; Saugey, A.; Barrat, J. L.; Bocquet, L.; Charlaix, E.; Gobin, P. F.; Vigier, G. J. Chem. Phys. 2004, 120, 4927–4938.
(19) Giesgche, H.; Unger, K. K.; M€uller, U.; Esser, U. Colloids Surf. 1989, 37, 93–113.
Figure 5. Nitrogen sorption isotherms measured on tablets made of Aerosil R812S (þ) and Aerosil R972 () compressed for 10 min in a die under a pressure of 250 MPa. The volume of nitrogen adsorbed V a per gram of sample ms is expressed as a function of the relative pressure p/p0.
2336 DOI: 10.1021/la902759s
Langmuir 2010, 26(4), 2333–2338
Forny et al.
Article
Figure 7. Cumulative distributions of pore volume obtained from the Laplace-Washburn equation (O) and from calorimetric data analysis () for Aerosil R812S (a) and Aerosil R972 (b) assuming contact angles of 118° and 105°, respectively.
variation of energy is called the energy of adhesion noted ΔEadh. Observing that the surface Σ1 is initially a liquid/vapor interface and that surface Σ2 is initially a solid/vapor interface of surface tension γSV and becomes a solid/liquid interface of surface tension γSL, the following equation, derived from the Young equation (γSV - γSL = γLV 3 cos θ) can be obtained: Figure 8. Schematic representation of a spherical particle (a) located outside of a liquid phase, (b) attached at the liquid/air interface, or (c) fully immersed into the liquid. Change from a to b relates to the energy of adhesion ΔEadh, and that from a to c relates to the energy of immersion ΔEimm.
distributions obtained from calorimetric analysis of the intrusion experiments (Figure 7). The second method consists in calculating the variation of superficial internal energy by adding the measured heat of intrusion and the calculated work of intrusion. According to eq 5 and surface areas obtained from BET analysis (Table 1), the values of β are 0.0122 and 0.042 J 3 m-2 for R812S and R972, respectively. It is then possible to assess stepwise the pore volume distributions according to eq 8. The more striking result is that the two pore volume distributions are very close to each other. This method, which is a minimization procedure applied to the whole pore size distribution, shows that all the porosity can be described with only one value of the contact angle, which proves the homogeneity of the solid as well as the consistency of the contact angle values derived from the intrusion experiments. It is also worthwhile to notice that the total porous volumes are in reasonable agreement with the values obtained from the analysis of the nitrogen sorption isotherms: 0.88 and 0.99 cm3 g-1 are obtained by water intrusion (from the LaplaceWashburn equation in Figure 7) against 0.92 and 1.2 cm3 g-1 by gas adsorption (Table 1), for R812S and R972, respectively. As already observed, lower values are often measured by water intrusion.16,18 Although not fully understood, this effect may be due to the existence of small pores or roughness not wetted by water. After considering different methods, water intrusion seems to be an effective and straightforward method to assess the contact angle of hydrophobic fumed silica particles.
Discussion On the basis of the contact angle determination described above, and in order to better understand the mechanisms responsible for dry water formation, we propose to study the theoretical interaction energy between a liquid of surface tension γLV and a model particle of radius R and surface area Σ0. The particle attaches to the liquid/vapor interface in an equilibrium position depending on the contact angle θ (Figure 8). The corresponding Langmuir 2010, 26(4), 2333–2338
ΔEadh ¼ γSL 3 Σ2 - ðγLV 3 Σ1 þ γSV 3 Σ2 Þ ¼ -γLV 3 πR2 3 ðcos θ þ 1Þ2 ð10Þ
The energy of adhesion appears to be negative, which means that any particle will spontaneously attach to the interface of any liquid. This is easily observed experimentally by bringing into contact a water droplet with highly hydrophobic fumed silica particles. In the same way, it is possible to evaluate the energy of immersion ΔEimm, which corresponds to the energy needed to immerse the particle from the air into the liquid: ΔEimm ¼ γSL 3 Σ0 - γSV 3 Σ0 ¼ -γLV 3 4πR2 3 cos θ
ð11Þ
ΔEimm becomes positive when θ > 90°, meaning that immersion is not spontaneous, and energy is required to immerse the particle. Figure 9 shows the variation of the energies of adhesion and immersion as a function of the contact angle (normalized by πR2) in the case of water and 8% ethanol solution. It may be noticed that the adhesion and immersion energies are almost identical when the contact angle is lower than 45°. Above 45°, the attachment to the surface is always energetically more favorable than full immersion. It means that, if a single particle, initially immerged, meets the liquid/vapor interface, it could remain attached to it if no energy is brought to the system. However, if the contact angle is below 45°, very little perturbation (i.e., waves or agitation) of the interface might be enough to immerse the particle again. Similar reasoning might be applied to any liquid shapes and non sharp solids. The calculation would become much more challenging, but the general findings would remain the same. From these considerations about the solid/liquid interactions, we can now try to explain the transformations occurring during dry water processing as described earlier in the Introduction. The shell-like structure of dry water indicates that the mechanism of formation involves the coating of micrometric water droplets. Similar observations were reported by P. Aussillous with larger water droplets.20,21 This coating is based on spontaneous (20) Aussillous, P. PhD Thesis, Universite de Paris 6, 2002. (21) Aussillous, P; Quere, D. Nature 2001, 411, 924.
DOI: 10.1021/la902759s
2337
Article
Forny et al. Table 2. Surface Tension, Contact Angle, and Normalized Energies of Adhesion and Immersion for Different Solid/Liquid Combinations solid/liquid system R812S/water R972/water R812S/ethanol 8%
surface tension contact ΔEadh/πR2 ΔEimm/πR2 angle θ (mJ 3 m-2) (mJ 3 m-2) γ (mJ 3 m-2) 72.8 72.8 50
118 105 108
-20 -40 -24
137 75 62
level as the value obtained with Aerosil R972 in pure water (Table 2). If even more hydrophilic particles are used, a suspension is obtained, in which the particles are completely immersed.
Conclusion Figure 9. Variation of the normalized energies ΔE/πR2 of adhesion (solid line) and immersion (dotted line) as a function of the contact angle θ for pure water and 8% ethanol solution in water.
adhesion of silica particles at the water/air interface, as discussed earlier in this section. In high shear processes, the formation of primary large water droplets is attributed to turbulent flow that takes place at high mixing rates.8 These large water droplets are then divided into more stable micrometric ones as result of high shearing conditions. During the mixing process, moderately hydrophobic particles such as Aerosil R972, exhibiting lower energy of immersion than Aerosil R812 (Table 2), will penetrate more easily into the liquid phase. If they come into interaction with other particles, they may aggregate into a continuous phase and form a network entrapping air and leading to the mousse formation. Similar results were obtained with highly hydrophobic particles (e.g., Aerosil R812S) when adding 8% of ethanol to the water phase. Immersion energy is reduced and reaches the same
2338 DOI: 10.1021/la902759s
Contact angles of two different range of hydrophobic fumed silica have been assessed using water intrusion experiments. This method allows a straightforward determination of the contact angle, as water is forced to penetrate the particle bed under high pressure conditions and explore the entire solid surface. In that way, the results obtained are directly representative of the interactions between water and the solid surface. The results obtained show that even a slight difference in contact angle may lead to very different product formation in high shear process (dry water, mousse, or suspension). Significant difference in immersion energy of the particles may explain the different behavior. Acknowledgment. This project, carried out in partnership with CLC Technologie (Beauvais) and Centre de Valorisation des Glucides (Amiens), was financially supported by the P^ole Regional Genie des Procedes (Region Picardie, France), which is gratefully acknowledged.
Langmuir 2010, 26(4), 2333–2338
