Langmuir 2007, 23, 4837-4841
4837
Particle Bridging between Oil and Water Interfaces Hui Xu, Mauricio Lask, John Kirkwood, and Gerald Fuller* Department of Chemical Engineering, Stanford UniVersity, Stanford, California 94305-5025 ReceiVed December 12, 2006. In Final Form: February 5, 2007 Particle bridging between a water drop and a flat oil-water interface has been observed when the drop is brought into contact with the interface, leading to the formation of a dense particle monolayer of disc shape (namely, particle disc) that prevents the drop from coalescing into the bulk water phase. Unlike previous observations where particles from opposite interfaces appear to register with each other before bridging, the present experiment demonstrates that the particle registry is not a necessity for bridging. In many cases, the particles from one of the interfaces were repelled away from the contact region, leaving behind the particles from the other interface to bridge the two interfaces. This is confirmed by particle bridging experiments between two interfaces covered with different sized particles, and between a particle-covered interface and a clean interface. The dynamics associated with the growth of the particle disc due to particle bridging follows a power law relationship between the radius of the disc and time: r ∝ t0.32(0.03. A scaling analysis assuming capillary attraction as the driving force and a hydrodynamic resistance leads to the power law r ∝ t1/3, in good agreement with the experiment. In addition, we found that binary mixtures of two different sized particles can undergo phase segregation driven by the particle bridging process.
1. Introduction Solid-stabilized emulsions, termed Pickering emulsions,1 have received much interest in recent years, partly because of their importance in a range of practical applications2-7 and partly because of their uniqueness in the stabilization mechanism against drop coalescence in comparison to ordinary surfactants.8-15 Unlike surfactant molecules that stabilize emulsion drops in a thermodynamic way, particles are thought to stabilize emulsions in a mechanical (kinetic) fashion.16-18 In most cases, particles are adsorbed at the interface irreversibly if the particle size is not (1) Pickering, S. U. Emulsions. J. Chem. Soc. 1907, 91, 2001-2021. (2) Binks, B. P.; Lumsdon, S. O. Pickering emulsions stabilized by monodisperse latex particles: Effects of particle size. Langmuir 2001, 17, 4540-4547. (3) Binks, B. P. Macroporous silica from solid-stabilized emulsion templates. AdV. Mater. 2002, 14, 1824-1827. (4) Binks, B. P.; Whitby, C. P. Silica particle-stabilized emulsions of silicone oil and water: Aspects of emulsification. Langmuir 2004, 20, 1130-1137. (5) Fujii, S.; Read, E. S.; Binks, B. P.; Armes, S. P. Stimulus-responsive emulsifiers based on nanocomposite microgel particles. AdV. Mater. 2005, 17, 1014-1018. (6) Tarimala, S.; Dai, L. L. Structure of microparticles in solid-stabilized emulsions. Langmuir 2004, 20, 3492-3494. (7) Melle, S.; Lask, M.; Fuller, G. G. Pickering emulsions with controllable stability. Langmuir 2005, 21, 2158-2162. (8) Aveyard, R.; Clint, J. H.; Horozov, T. S. Aspects of the stabilisation of emulsions by solid particles: Effects of line tension and monolayer curvature energy. Phys. Chem. Chem. Phys. 2003, 5, 2398-2409. (9) Aveyard, R.; Clint, J. H. Liquid droplets and solid particles at surfactant solution interfaces. J. Chem. Soc., Faraday Trans. 1995, 91, 2681-2697. (10) Binks, B. P.; Lumsdon, S. O. Stability of oil-in-water emulsions stabilised by silica particles. Phys. Chem. Chem. Phys. 1999, 1, 3007-3016. (11) Binks, B. P. Particles as surfactantssSimilarities and differences. Curr. Opin. Colliod Interface Sci. 2002, 7, 21-41. (12) Vignati, E.; Piazza, R.; Lockhart, T. P. Pickering emulsions: Interfacial tension, colloidal layer morphology, and trapped-particle motion. Langmuir 2003, 19, 6650-6656. (13) Tambe, D. E.; Sharma, M. M. Effect of colloidal particles on fluid-fluid interfacial properties and emulsion stability. AdV. Colloid Interface Sci. 1994, 52, 1-63. (14) Tadros, Th. V.; Vincent, B. Encyclopedia of Emulsion Technology, Vol. 1; Dekker: New York, 1983. (15) Denkov, N. D.; Ivanov, I. B.; Kralchevsky, P. A.; Wasan, D. T. A possible mechanism of stabilization of emulsions by solid particles. J. Colloid Interface Sci. 1992, 150, 589-593. (16) Levine, S.; Bowen, B. D. Capillary interaction of spherical-particles adsorbed on the surface of an oil-water droplet stabilized by the particles. 1. Colloids Surf., A 1991, 59, 377-386. (17) Levine, S.; Bowen, B. D. Capillary interaction of spherical particles adsorbed on the surface of an oil/water droplet stabilized by the particles. 3. Effective interfacial tension. Colloids Surf., A 1993, 70, 33-45.
too small (say, a few nanometers).19 The adsorption energy is on the order of thousands to millions of thermal energy (kT) depending on the size of the particles and the contact angle of the particles with the interface.20,21 The particles adsorbed at the interface behave as a mechanical barrier that prevents the emulsion drops from coming into contact thus protecting them against coalescence. One of the scenarios envisioned by Tadros and Vincent14 as well as Denkov et al.15 is that the particles bridge the interfaces when two drops come into close proximity, forming a dense particle monolayer that prevents or retards a thin film of the second phase between the drops from draining. This hypothesis of the mechanism for particle stabilization was recently confirmed by direct experimental observations.22-26 In these experiments, particle-covered interfaces were brought into contact in a controlled manner. When the two interfaces were close enough, the particles at both interfaces arranged in such a way that the particles at one interface matched the interstices of particle lattices on the opposing interface, followed by a rapid growth of a dense particle aggregate of disc shape that resisted further coalescence between the two interfaces. The particle aggregate was found to be only a monolayer thick that spanned the two interfaces, which is consistent with the bridging configuration. A model experiment (18) Levine, S.; Bowen, B. D. Capillary interaction of spherical-particles adsorbed on the surface of an oil-water droplet stabilized by the particles. 2. Colloids Surf., A 1992, 65, 273-286. (19) Lin, Y.; Skaff, H.; Emrick, T.; Dinsmore, A. D.; Russell, T. P. Nanoparticle assembly and transport at liquid-liquid interfaces. Science 2003, 299, 226-229. (20) Levine, S.; Bowen, B. D.; Partridge, S. J. Stabilization of emulsions by fine particles. 1. Partitioning of particles between continuous phase and oil-water interface. Colloids Surf. 1989, 38, 325-343. (21) Xu, H.; Melle, S.; Golemanov, K.; Fuller, G. G. Shape and buckling transitions in solid-stabilized drops. Langmuir 2005, 21, 10016-10020. (22) Stancik, E. J.; Kouhkan, M.; Fuller, G. G. Coalescence of particle-laden fluid interfaces. Langmuir 2004, 20, 90-94. (23) Stancik, E. J.; Fuller, G. G. Connect the drops: Using solids as adhesives for liquids. Langmuir 2004, 20, 4805-4808. (24) Ashby, N. P.; Binks, B. P.; Paunov, V. N. Bridging interaction between a water drop stabilised by solid particles and a planar oil/water interface. Chem. Commun. 2004, 436-437. (25) Horozov, T. S.; Aveyard, R.; Clint, J. H.; Neumann, B. Particle zips: Vertical emulsion films with particle monolayers at their surfaces. Langmuir 2005, 21, 2330-2341. (26) Horozov, T. S.; Binks, B. P. Particle-stabilized emulsions: A bilayer or a bridging monolayer? Angew. Chem., Int. Ed. 2006, 45, 773-776.
10.1021/la063593l CCC: $37.00 © 2007 American Chemical Society Published on Web 03/23/2007
4838 Langmuir, Vol. 23, No. 9, 2007
Xu et al.
demonstrated that such particle discs produced an adhesive force per unit area as high as 300 N/m2,23 far more than the Laplace pressure of the millimeter-sized bridging drops under study. Even more interestingly, the growth of the particle disc was found to obey a universal power law: r ∝ t1/3, where r is the disc radius and t is time.22 It is worthwhile noting that currently the particle bridging was only observed between water drops in oil while the opposite casesbridging between oil drops in an aqueous phases has not been observed so far. Particle bridging plays an important role in the long-term stability of particle-stabilized emulsions and foams.26 The experimental confirmation of such bridging phenomenon paves a way for a thorough understanding of this problem. Although a great deal of information has been retrieved through recent studies, there are still many questions that remain to be answered, specifically, for instance, is it necessary for the particle lattices at the opposite interfaces to register with each other before bridging occurs, how does the particle size affect the bridging, what is the underlying physics that governs the power law growth of the particle disc, and so on. The present study is aimed at these questions by using controlled experiments that will probe the particle bridging at various conditions involving particles of different sizes. In addition, theoretical understanding of the bridging dynamics is advanced by a simple scaling analysis. 2. Experimental Section Spherical polystyrene particles with a mean diameter of 3.1 and 1.0 µm and a surface charge density of 9.1 and 7.5 µC/cm2, respectively, were acquired in the form of a surfactant-free, aqueous dispersion from Interfacial Dynamics Corp. The contact angle of the particles with the oil-water interface was determined to be about 130° in a previous study.22 Through dilution with deionized water (specific resistance, 18.2 MΩ‚cm; Millipore) and isopropyl alcohol (Mallinckrodt), a working dispersion consisting of (5.0-10.0) × 108 particles/mL in a 20% isopropyl alcohol solution was created. Note that isopropyl alcohol is included as an aid to spreading the particles at interfaces. Sonication of this dispersion for 5 min prior to injection at interfaces helped to ensure the absence of aggregates. The interfaces themselves were created between deionized water (Millipore) and decane (Fisher). Prior to use in experiments, the decane was passed three times through a column of aluminum oxide (Fisher) to remove any polar contaminants that might have been present. The experimental setup being utilized to study the particle bridging is shown in Figure 1A, which is analogous to that used by Stancik et al.22 on the drop coalescence experiment. A glass capillary was used to create a water droplet in the oil phase near the flat interface. A volume of the working dispersion was injected onto the flat interface and/or the surface of the droplet to attain the desired interfacial particle concentrations. After the system was allowed to come to equilibrium, the droplet was brought into contact with the flat interface via a xyz stage with motion controlled by micrometerdriven actuators. Particle bridging was observed from the angle perpendicular to the flat interface using a Nikon Eclipse TE300 inverted microscope. Images were recorded via CCD cameras (Hamamatsu) attached to the microscopes. Metamorph software from Universal Imaging Corp. was used to perform the image analysis.
3. Results and Discussions The surface concentration of the particles at the oil-water interface was in a range of 2-20% (area/area) at equilibrium. At these surface concentrations, the particles attain a hexagonal packing at the interfaces due to a dipole-dipole repulsion between the charged particles27,28 (or unscreened Coulombs), as shown (27) Pieranski, P. Two-dimensional interfacial colloidal crystals. Phys. ReV. Lett. 1980, 45, 569-572. (28) Hurd, A. J. The electrostatic interaction between interfacial colloidal particles. J. Phys. A 1985, 18, L1055-L1060.
Figure 1. (A) Setup designed to view the particle bridging process. (B) Schematic illustration of the paricle bridging and subsequent disc formation. (C-F) Experimental observation of the radial growth of particle disc as time elapses.
in Figure 1C. The drop, which is about 1 mm in diameter, was brought down to the flat interface at a controlled speed that could be as low as 0.5 µm/s. Once the bridging occurs, the motion of the translation stage holding the drop is arrested. When the two interfaces are close enough, one can see the particles at both interfaces, as evidenced by the appearance of doublets or triplets that are due to the overlapping of the particle lattices at the two interfaces (in the experiment shown in Figure 1 identical particles of 3 µm in diameter are applied to both interfaces). The interesting observation here is that when the drop was brought into closer proximity with the flat surface, instead of finding registry with each other, the particles on the drop surface were repelled away from the contact area (as can be seen from Figure 1D, the doublets and triplets were depleted out of the center region), while the particles at the flat interface were driven into the center, causing an increase of the local surface concentration there. At a critical moment, a few particles successfully bridged the two interfaces and a rapid growth of a dense particle disc occurred (Figure 1E,F). It was obvious that the disc was exclusively fed from the particles at the flat interface in this case. We repeated these experiments numerous times, and in the majority cases similar phenomena were observed. We did observe particle registry as reported in previous studies,23,25 but that happened very rarely. To confirm this finding, two control experiments were conducted. In one experiment, the flat interface was covered with 1 µm sized particles, while the drop surface was covered with 3 µm sized particles (Figure 2A). In another experiment, a clean drop was brought down to the flat interface covered with 3 µm sized particles (Figure 3A). In the first experiment, the surface concentration of the small particles at flat interface is (29) Aveyard, R.; Clint, J. H.; Nees, D.; Paunov, V. N. Compression and structure of monolayers of charged latex particles at air/water and octane/water interfaces. Langmuir 2000, 16, 1969-1979.
Particle Bridging between Oil and Water Interfaces
Langmuir, Vol. 23, No. 9, 2007 4839
Figure 4. Radial growth of the particle disc as a function of time. The solid line is a linear fitting of the data that yields a power exponent of 0.32 ( 0.03.
Figure 2. (A) Configuration of the experiment in which the water drop is covered with large (3 µm) particles, while the flat interface is covered with small (1 µm) particles. (B-E) Observation of the disc formation that starts with the small particles followed by the large particles at a later stage.
Figure 5. Schematic representation of the geometry of the particle disc and its surroundings during the disc growth process.
Figure 3. (A) Configuration of the experiment in which a clean water drop was brought into contact with the interface covered with 3 µm particles. (B-D) Observation of the formation and growth of the particle disc.
about 5 times higher than that of large particles at the drop surface. The individual small 1 µm particles are hardly visible under this optical magnification (×100), which is intended to allow a relatively large field of view for observing the whole bridging process. When the small particles are present at higher surface concentration, the depletion of larger particles was normally observed as shown in Figure 2B. In this case, the small particles bridged the interfaces at first, followed by a rapid growth of the dense particle disc that is made exclusively of the small particles (Figure 2C,D). At a later stage, the large particles at the drop surface were able to bridge the interfaces as well and join the disc on its perimeter (Figure 2E). This experiment verifies that particle registry between two interfaces is not a necessity for the particle bridging and disc formation, and also, the particle bridging is independent of particle size for the micrometer-sized particles studied here. The next experiment went a step further by leaving the drop surface free of particles, thus eliminating the particle-particle
interactions between the two interfaces, which might be responsible for the particle registry. Again, particle bridging and the formation of a particle disc were observed (Figure 3B-D), which further confirms that the occurrence of particle bridging and subsequent disc formation are not a consequence of particle registry (a similar experiment on particle bridging between a particle-covered drop and a clean flat interface was also noted24). This experiment also suggests that the particles at one interface alone are sufficient to protect the drop from coalescence. In all the cases discussed above, the growth of the particle discs obeyed a universal power law: r ∝ t0.32(0.03 (r is the radius of the disc and t is time), which confirmed the result of the previous study.22 One typical example is shown here in Figure 4. As proposed by Denkov et al.,15 particle bridging results in a deformation of both interfaces that the particles span, which induces lateral capillary attraction between particles, leading to the particle aggregation. However, the correlation between such capillary interaction and the growth kinetics of the particle disc has not been previously explained. Here, a scaling analysis is performed to recover the observed power law relation. We start with a geometric analysis of the particle disc and its surroundings. A top view of this geometry is illustrated in Figure 5, as motivated by the experiments (Figures 1-3). There are two distinct surface concentrations associated with this problem: the surface concentration of the particle disc, c, and the surface concentration of the surrounding monolayer, c0. c is much greater than c0. It is noted that the formation of the particle disc consumes particles from the surrounding monolayer,
4840 Langmuir, Vol. 23, No. 9, 2007
Xu et al.
leading to a particle depletion zone between the disc and surrounding monolayer. This is simply a result of mass (particle) conservation. The relation between r (the radius of the disc) and R (the sum of r and the gap width L) can be easily derived from the principle of particle conservation:
R)
x
c r c0
(1)
The growth of the disc is at the expense of the surrounding monolayer. In other words, the particles are transferred from the surrounding monolayer to the disc. It is reasonable to assume that this particle transport process dictates the growth rate of the particle disc. The driving force of the transport, as discussed above, will be assumed here to be the lateral capillary force between particles and the disc: Fc ∝ l-1, as a first approximation, where l is the distance between a traveling particle and the center of the particle disc,30 which falls between R and r, i.e., R g l g r. The transport will be resisted by hydrodynamic friction force, Ff ) -ζu, where ζ is friction factor and u is the particle velocity. Another resisting force may come from the repulsive electrostatic interactions between particles and the disc. However, this force, Fe ∝ -l-4, 27,28,31 is too short-ranged and too small to compete with the capillary force Fc arising from the deformation of two interfaces33 and can be safely omitted in this particular case. The particle convection considered here is still in the viscous regime, where inertial effects are negligible.32 Thus, the sum of the forces vanishes:
Fc + F f ) 0
(2)
Figure 6. (A) Configuration of the experiment in which a binary mixture of 3 and 1 µm particles was applied to the flat interface while no particles were applied to the surface of the water drop. (B-E) The observation of the disc formation and growth, and the phase segregation at a later stage in which multiple satellite discs composed of small particles were formed.
Therefore, the disc radius increases by
δr ) r′ - r )
This leads to
u∝
1 ζl
(3)
which is an instaneous velocity of a particle during translation. The total time that a particle moves from the edge of the surrounding monolayer to the disc can be found by integrating over the infinitesimal time steps:
δt )
∫dt ) - ∫Rr dlu ∝ 2ζ(R2 - r2) )
ζ(c - c0) 2 r 2c0
(4)
During the time scale of δt, one ring of particles from the surrounding monolayers is assumed to be added into the disc (this assumption is based on the equal probability of the particle bridging along the edge of the surrounding monolayer), causing the disc radius growth. The new radius, r′, after time δt, is calculated from the particle conservation relation: 2πR(1/c01/2)c0 ) π(r′)2c - πr2c. Replacing R with r by using the relation from eq 1, one obtains
r′ )
x
2r + r2 xc
(5)
(30) Here one considers the entire disc as a large cylindrical “particle”, whose interaction with a small particle has the usual form of capillary interaction, as can be found in the following article: Kralchevsky, P. A.; Nagayama, K. Capillary interactions between particles bound to interfaces, liquid films and biomembranes. AdV. Colloid Interface Sci. 2000, 85, 145-192. (31) Aveyard, R.; Binks, B. P.; Clint, J. H.; Fletcher, P. D. I.; Horozov, T. S.; Neumann, B.; Paunov, V. N.; Annesley, J.; Botchway, S. W.; Nees, D.; Parker, A. W.; Ward, A. D.; Burgess, A. N. Measurement of long-range repulsive forces between charged particles at an oil-water interface. Phys. ReV. Lett. 2002, 88, 246102.
x
2r 1 + r2 - r ≈ xc xc
(6)
Knowing δt and δr, one can now proceed to calculate the growth rate of the particle disc:
r˘ )
2c0 1 dr δr ) ∝ dt δt ζxc(c - c ) r2 0
(7)
By simply solving this first-order differential equation, one obtains the final relation between r and t:
r ∝ kt1/3
(8)
where k ) [6c0/ζc1/2(c - c0)]1/3 is a constant for a specific system. Note that the constant of integration has been dropped since there is no disc initially, i.e., r ) 0 at t ) 0. This scaling argument offers a strong support to the hypothesis that capillary force drives the particle aggregation and likely captures the physical origin of the universal power law that governs the growth kinetics of the particle discs. This study also considered the bridging of binary mixtures of particles of different sizes. In this experiment, the flat interface was covered with a binary mixture of 1 and 3 µm sized particles (in a number ratio of 2-20), while the drop surface was left clean for simplicity (Figure 6A). When the drop was brought down to the flat interface, the large particles bridged the interfaces first, as expected, and formed a dense disc (Figure 6B,C). However, inevitably some small particles were also carried into the disc though they might not bridge the interfaces at this moment. (32) Loudet, J. C.; Alsayed, A. M.; Zhang, J.; Yodh, A. G. Capillary interactions between anisotropic colloidal particles. Phys. ReV. Lett. 2005, 94, 018301. (33) The electrostatic force could be as important as capillary force when the particles are close to each other, which might determine the packing density of the particle disc (i.e., the surface concentration c).
Particle Bridging between Oil and Water Interfaces
The depletion of large particles in regions of the surrounding monolayer (Figure 6C) suggested that the large particles were more concentrated in the disc than in the surrounding monolayer, an indication of phase segregation. A more pronounced phase segregation occurred at a later stage as the drop was further pushed down toward the flat interface. The force of pressing the drop against the interface created stresses in the particle disc that eventually developed small cracks (Figure 6D). As the cracks grew, the oil film sandwiched between the drop and the flat interface appeared to be stretched in the crack regions. The film could be stretched into such a thickness that the bridging of small particles became possible. This eventually led to the following experimentally observed phenomenon: at one point the small particles trapped inside the disc rushed to the center of cracks forming daughter discs within the mother disc (Figure 6E). Such satellite-like discs were composed purely of small particles; that is, a phase segregation had occurred. The phase segregation induced by particle bridging provides an intriguing way to manipulate particle assembly in mixed monolayers. In addition, it may offer insights into controlling the stability of Pickering emulsions involving polydisperse particles.
4. Conclusions A systematic study of particle bridging between a water drop and a flat interface has been carried out. From the experiments,
Langmuir, Vol. 23, No. 9, 2007 4841
we demonstrated that the particle registry between opposite interfaces was not necessary for the bridging to occur. The particle bridging is independent of the particle sizes provided that the contact angles of the particles with the interface are comparable and larger than 90° (into the oil phase). In addition, the particles from one interface alone were able to bridge the interfaces thus protecting the drop from coalescence. The particle bridging resulted in rapid growth of particle discs which obeyed a universal power law. The scaling analysis based on the assumption that capillary force drove the particle aggregation yielded a relation of r ∝ t1/3, which recovered the experimental law of disc growth and confirmed the validity of a previous hypothesis of Tradros and Vincent and Denkov et al.14,15 Binary mixtures of particles of different sizes undergo phase segregation driven by particle bridging, forming satellite-like discs embedded in a mother disc. It should be noted that the conclusions drawn from this study apply only to water-in-oil emulsions since particle bridging has not been observed for oil-in-water cases. This could imply that the stability mechanism of water-in-oil and oil-in-water emulsions might be different, which certainly needs further study. Acknowledgment. We acknowledge the National Science Foundation (NSF) and Unilever Co. LA063593L
