8316
Langmuir 2007, 23, 8316-8322
Pickering Miniemulsion Polymerization Using Laponite Clay as a Stabilizer Stefan A. F. Bon* and Patrick J. Colver Centre for Interfaces and Materials, Department of Chemistry, UniVersity of Warwick, CoVentry CV4 7AL, U.K. ReceiVed April 19, 2007. In Final Form: May 23, 2007 Solid-stabilized, or Pickering, miniemulsion polymerizations using Laponite clay discs as stabilizer are investigated. Free radical polymerizations are carried out using a variety of hydrophobic monomers (i.e., styrene, lauryl (meth)acrylate, butyl (meth)acrylate, octyl acrylate, and 2-ethyl hexyl acrylate). Armored latexes, of which the surfaces of the particles are covered with clay discs, are obtained, as confirmed by scanning electron microscopy (FE-SEM) and atomic force microscopy (AFM). Overall polymerization kinetics of the Pickering miniemulsion polymerizations of styrene were investigated via gravimetry. Comparison with the bulk polymerization analogue clearly shows compartmentalization. Moreover, retardation effects up to intermediate monomer conversions are observed; they are more prominent for the smaller particles and are ascribed to the Laponite clay. A model is presented that allows for the prediction of the average particle size of the latexes produced as a function of the amounts of monomer and Pickering stabilizers used. It shows that under specific generic conditions the number of clay discs used correlates in a linear fashion with the total surface area of the latex particles. This is a direct result of the reversibility of the Laponite clay disc adhesion process under the emulsification conditions (i.e., sonication) used.
Introduction Emulsion and miniemulsion polymerization are commercially important processes for manufacturing polymer particles of submicrometer dimensions, commonly made as a stable dispersion in water. To warrant the stability of the polymer latex during and after the polymerization process, surfactants of low molecular weight are often used, for example, sodium dodecyl sulfate and sodium dialkyl sulfosuccinates (aerosol series). Industrially, polymeric stabilizers and protective agents, such as poly(vinyl alcohol), are additionally employed, for example, to obtain good freeze-thaw stability in coatings. Sometimes these two types of stabilizers can be generated in situ, for example, in soap-free emulsion polymerizations or reactions involving hydrophilic comonomers such as (meth)acrylic acid. We are interested in the synthesis of polymer latexes via (mini)emulsion polymerization using different strategies to obtain stable colloidal systems. Appealing approaches were undertaken by Armes et al., who reported the surfactant-free synthesis of poly(methyl methacrylate)-silica nanocomposite particles in aqueous alcohol media at ambient temperature using a silica nanosol as a stabilizer in the absence of auxiliary co-monomers.1,2 Lewis and co-workers described that a stable colloidal system of negligible charged silica spheres could be obtained upon addition of a critical volume fraction of highly charged nanoparticles of hydrous zirconia. These nanoparticles did not adsorb onto the silica spheres, and the stabilization mechanism was referred to as haloing.3 Our attention was drawn to the pioneering work of Ramsden4 and Pickering,5 who showed that emulsions could be stabilized by solid particles. Hildebrand et al.6 suggested that the * Corresponding author. E-mail:
[email protected]. Tel: +44 (0)2476 574009. Web: www.stefanbon.eu. (1) Percy, M. J.; Armes, S. P. Langmuir 2002, 18, 4562-4565. (2) Schmid, A.; Fujii, S.; Armes, S. P. Langmuir 2006, 22, 4923-4927. (3) Tohver, V.; Smay, J. E.; Braem, A.; Braun, P. V.; Lewis, J. A. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 8950-8954. (4) Ramsden, W. Proc. R. Soc. 1903, 72, 156-164. (5) Pickering, S. U. J. Chem. Soc. Trans. 1907, 91, 2001-2021. (6) Finkle, P.; Draper, H. D.; Hildebrand, J. H. J. Am. Chem. Soc. 1928, 45, 2780-2788.
reason for this was that the particles were partially wettable by the two phases involved and thus would adhere to the interface. Moreover, they postulated that the type of emulsion produced by a solid powder is determined by the angle of contact of the interface with the solid. “In order for the powder to remain in the interface the angle must be finite, and unless the angle is 90°, the interface will be on one side or the other of the points of contact of the particles, and its tension will cause the film to be concave on that side.” Pieranski demonstrated by solely taking into account interfacial tensions and correctly neglecting gravity that micrometer-sized polystyrene spheres at the water/air interface are in essence trapped, with an energy barrier for reentry into the water phase of multiple orders of magnitude larger than the thermal energy, kBT.7 Solid-stabilized, or Pickering, emulsions and foams are widely applied in commercial areas such as food, cosmetics, petrochemicals, oil refining, and ore purification. Recent developments include responsive Pickering emulsifiers, such as magnetic,8 thermal, and pH-driven,9,10 and the focus on creating new materials and supracolloidal structures using liquid-liquid or liquid-gas interface-driven self-assembly of solid particles. Binks recently demonstrated that via phase inversion particle-stabilized airwater systems could lead to air-in-water foams or water-in-air stabilized powders or vice versa.11 The latter material, water droplets stabilized by particles, was tentatively named “liquid marbles”.12 Velev et al. described the synthesis of hollow supracolloidal structures via the assembly of a polystyrene latex on the interface of emulsion droplets.13 These permeable structures, created by fully covering emulsion droplets with colloidal building blocks, were named colloidosomes by Dins(7) Pieranski, P. Phys. ReV. Lett. 1980, 45, 569-572. (8) Melle, S.; Lask, M.; Fuller, G. G. Langmuir 2005, 21, 2158-2162. (9) Ngai, T.; Behrens, S. H.; Auweter, H. Chem. Commun. 2005, 3, 331-333. (10) Fujii, S.; Read, E. S.; Binks, B. P.; Armes, S. P. AdV. Mater. 2005, 17, 1014-1018. (11) Binks, B. P.; Murakami, R. Nat. Mater. 2006, 5, 865-869. (12) Aussillous, P.; Que´re´, D. Nature 2001, 411, 924-927. (13) Velev, O. D.; Furusawa, K.; Nagayama, K. Langmuir 1996, 12, 23742384.
10.1021/la701150q CCC: $37.00 © 2007 American Chemical Society Published on Web 06/29/2007
Pickering Miniemulsion Polymerization
Langmuir, Vol. 23, No. 16, 2007 8317
Table 1. Pickering Miniemulsion Polymerizations of Styrene Stabilized by Laponite Clay
expt
m0/ g
m M/ g
mhex/ g
mI/ g
mPS/ g
doil/ nm
C0/ mg g-1
Cexcess/ mg g-1
calcd cexcess/ mg g-1
calcd doil/ nm
100.2 100.8 100.5 101.0 100.4 99.5 100.0
PJC-1-037 PJC-1-042 PJC-1-046 PJC-1-047 PJC-1-048 PJC-1-049 PJC-1-052
0.502 0.255 0.249 0.355 0.693 1.501 1.004
10.003 10.186 10.009 10.031 9.928 10.027 10.014
0.404 0.423 0.402 0.404 0.405 0.401 0.401
0.049 0.066 0.046 0.053 0.045 0.074 0.051
8.40 7.13 7.46 8.13 8.34 8.52 8.11
495.7 658.3 643.2 607.9 391.5 234.8 287.7
5.01 2.53 2.48 3.52 6.90 15.08 10.04
3.02 1.25 1.11 1.96 4.41 10.80 6.72
3.03 1.16 1.12 1.91 4.47 10.65 6.84
499.3 615.4 647.1 587.9 400.9 227.0 298.6
102.0 100.0 101.2 100.4 100.0 100.6 100.3 99.9 99.9
PJC-1-021 PJC-1-024 PJC-1-025 PJC-1-027 PJC-1-028 PJC-1-040 PJC-1-043 PJC-1-045 PJC-1-050
1.508 0.253 0.507 0.354 0.701 0.995 0.348 0.505 0.248
9.966 10.025 10.021 10.016 10.030 10.077 10.014 10.047 10.009
0.809 0.810 0.842 0.808 0.820 0.803 0.833 0.817 0.800
0.050 0.052 0.047 0.052 0.052 0.050 0.048 0.047 0.045
9.86 8.26 8.70 7.39 8.02 9.57 8.46 8.44 7.41
244.3 846.5 449.3 589.5 317.1 294.0 594.2 461.0 830.3
14.78 2.53 5.01 3.52 7.01 9.89 3.47 5.06 2.48
9.94 1.31 2.62 1.94 3.84 5.92 1.69 2.78 1.35
9.59 1.10 2.82 1.79 4.21 6.20 1.75 2.86 1.07
227.9 724.0 489.9 538.1 359.0 316.5 615.0 477.9 663.9
mwater/ g series I
series II
Table 2. Additional Pickering Miniemulsion Polymerizations of Styrene Stabilized by Laponite Clay
series III
mwater/ g 99.6 101.2 100.1 100.6 99.7 99.6
expt
m0/ g
m M/ g
mhex/ g
mI/ g
mPS/ g
doil/ nm
C0/ mg g-1
Cexcess/ mg g-1
calcd Cexcess/ mg g-1
calcd doil/ nm
PJC-1-058 PJC-1-059 PJC-1-060 PJC-1-061 PJC-1-062 PJC-1-063
1.032 0.498 0.512 0.504 1.004 0.505
5.009 5.001 7.515 2.647 5.036 7.500
0.206 0.200 0.320 0.128 0.203 0.300
0.044 0.046 0.042 0.037 0.044 0.056
4.32 3.87 6.01 1.97 4.37 6.10
226.9 245.1 344.4 182.0 224.7 370.1
10.36 4.93 5.12 5.01 10.07 5.07
8.12 3.08 3.05 3.71 7.78 3.12
7.08 2.97 3.11 3.03 6.86 3.08
155.4 231.0 355.0 119.1 160.3 362.6
Table 3. Pickering Miniemulsion Polymerizations of Various Monomers Stabilized by Laponite Clay
series IV
mwater/ g
expt
m0/ g
m M/ g
mhex/ g
mI/ g
mPS/ g
doil/ nm
101.8 99.4 98.0 101.8 98.5 98.8
PJC-1-115(LMA) PJC-1-116(BMA) PJC-1-118(LA) PJC-1-119(OA) PJC-1-121(BA) PJC-1-126(2-EHA)
0.505 0.511 0.250 0.509 0.508 0.500
2.492 2.502 2.491 2.478 2.568 2.532
0.110 0.108 0.105 0.107 0.106 0.107
0.058 0.055 0.052 0.048 0.054 0.050
2.03 2.04 2.03 2.02 2.09 2.06
209.2 183.9 285.9 223.3 197.4 222.1
more et al.14 In retrospect, pioneering work by Hohenstein15,16 and Wiley17-19 and co-workers demonstrates that these solidstabilized emulsion droplets can be used as polymerization vessels and therefore that Pickering suspension polymerizations are possible. We recently reported the preparation of supracolloidal interpenetrating polymer network reinforced capsules with raspberry core-shell morphologies using micrometer-sized colloidosomes of poly(methyl methacrylate-co-divinylbenzene) microgels20 and titanium dioxide nanoparticles21 as suspension polymerization reaction vessels. Russell et al. reported on supracolloidal capsules of quantum dots using both radical and ring-opening metathesis polymerization techniques to cross-link vinyl benzyl and norbornene functional quantum dots, respectively.22,23 (14) Dinsmore, A. D.; Hsu, M. F.; Nikolaides, M. G.; Marquez, M.; Bausch, A. R.; Weitz, D. A. Science 2002, 298, 1006-1009. (15) Hohenstein, W. P. Polym. Bull. 1945, 1, 13-16. (16) Hohenstein, W. P.; Mark, H. J. Polym. Sci. 1946, 1, 127-145. (17) Wiley, R. M. J. Colloid Sci. 1954, 9, 427-437. (18) Wiley, R. M. Clay thickened suspension polymerization process with plug flow. US Patent 2,886,559, 1959. (19) Wiley, R. M. Quiescent suspension polymerization. US Patent 2,932,629, 1960. (20) Bon, S. A. F.; Cauvin, S.; Colver, P. J. Soft Matter 2007, 3, 194-199. (21) Chen, T.; Colver, P. J.; Bon, S. A. F. AdV. Mater. 2007, DOI://10.1002/ adma.200602447. (22) Lin, Y.; Skaff, H.; Boker, A.; Dinsmore, A. D.; Emrick, T.; Russell, T. P. J. Am. Chem. Soc. 2003, 125, 12690-12691. (23) Skaff, H.; Lin, Y.; Tangirala, R.; Breitenkamp, K.; Boker, A.; Russell, T. P.; Emrick, T. AdV. Mater. 2005, 17, 2082-2086.
One could say that miniemulsion polymerization is the nanoanalogue of the more traditional suspension polymerization process, a key difference being the submicrometer dimensions of the miniemulsion droplets. The use of nanosized Pickering stabilizers would therefore in principle allow successful miniemulsion polymerization to be carried out. This was confirmed by our feasibility work on Pickering miniemulsion polymerizations of styrene using nanosized Laponite clay disks as stabilizer, thereby creating Laponite clay-armored latex particles of polystyrene.24 Pickering stabilizers had been employed successfully before in (mini)emulsion polymerization, but always in conjunction with conventional surfactants and/or stabilizing comonomers.25,26 Recently, van Herk et al. reported the preparation of polymer-clay nanocomposite latex particles by inverse Pickering (mini)emulsion polymerization stabilized with hydrophobic montmorillonite platelets.27 Herein we report on our studies on Pickering miniemulsion polymerization using Laponite clay as a stabilizer, thereby taking a mechanistic approach. We describe the Pickering miniemulsion polymerization kinetics of styrene and the morphologies of the (24) Colver, P. J.; Cauvin, S.; Bon, S. A. F. Macromolecules 2005, 38, 78877889. (25) Choi, Y. S.; Xu, M.; Chung, I. J. Polymer 2005, 46, 531-538. (26) Tiarks, F.; Landfester, K.; Antonietti, M. Langmuir 2001, 17, 57755780. (27) Voorn, D. J.; Ming, W.; van Herk, A. M. Macromolecules 2006, 39, 2137-2143.
8318 Langmuir, Vol. 23, No. 16, 2007
Bon and ColVer
synthesized latexes. We demonstrate that this system can be applied to a variety of hydrophobic monomers. Moreover, a semiempirical relationship will be presented that allows for the prediction of the average particle size of the latexes produced as a function of the amounts of monomer and Pickering stabilizers used. To the best of our knowledge, this is the first mechanistic study on Pickering miniemulsion polymerization.
Results and Discussion Laponite RD clays disks can successfully be used as solid stabilizers in the Pickering miniemulsion polymerization of styrene.24 These discotic platelets can exist in pure water as individual colloids with a lateral diameter of ca. 25-35 nm and approximately 1 nm in thickness.2829 Laponite RD is a synthetic trioctahedric hectorite clay composed of two tetrahedral silica sheets and a central octahedral magnesia sheet. Its chemical formula can be expressed as [Si8(Mg5.45Li0.4)O20(OH)4]Na0.7, and it has a density of 2570 kg m-3. The disks have an overall negative charge with the rim being amphoteric. Sodium chloride is added to induce slight colloidal instability leading to clay particle flocculation. This greatly enhances the capacity of the clay to allow Pickering stabilization of oils in water.30 In other words, the addition of salt compresses the double layer31 and lowers the zeta potential, thereby reducing electrostatic repulsion, inducing possible clay flocculation, and thus increasing its partitioning to the oil-water interface. We performed four series of Pickering miniemulsion polymerization experiments in which we varied the number of Laponite clay disks (series I and II; Table 1) and the amount and type of monomer (series III and IV, respectively; Tables 2 and 3). The sodium chloride concentration was kept constant at ca. 0.1 mol L-1 in all experiments. The initiator used was 2,2′-azobis(2,4dimethyl valeronitrile) (V-65), which was premixed with the monomer. An oil-soluble initiator proved essential in previous work.24 The miniemulsions were prepared via sonication (Experimental Section). This high-powered homogenization step ensures that the clay platelets can redistribute and thus are not permanently trapped at the monomer-water interface so that it is possible to create Pickering-stabilized monomer droplets of submicrometer size. One interesting point is related to the fact that at a salt concentration of 0.1 mol L-1 Laponite clay disks will flocculate.29,30,32 Therefore, it is important first to disperse the discotic platelets in water prior to the addition of the sodium chloride. This flocculation of the clay nanoparticles leads to an increase in the overall viscosity. Upon addition of monomer and subsequent shear through sonication, parts of the clay platelets will be confined to the monomer-water interface. This leads to a lower overall viscosity of the miniemulsion in comparison to that of the aqueous clay dispersion in the NaCl solution. The prepared miniemulsions were degassed and polymerized overnight at 51 °C. The latexes were stable initially, but upon storage they tended to flocculate and phase separate into a clear upper aqueous layer and a lower turbid layer containing the polymer latex. Upon dialysis, carried out to remove the NaCl, the latexes were easily redispersed into indefinitely stable colloidal dispersions. Morphology of Latexes Made via Pickering Miniemulsion Polymerization. Pickering stabilizers adhere themselves to the (28) Balnois, E.; Durand-Vidal, S.; Levitz, P. Langmuir 2003, 19, 66336637. (29) Mongondry, P.; Tassin, J. F.; Nicolai, T. J. Colloid. Interface Sci. 2005, 283, 397-405. (30) Ashby, N. P.; Binks, B. P. Phys. Chem. Chem. Phys. 2000, 2, 5640-5646. (31) Tawari, S. L.; Koch, D. L.; Cohen, C. J. Colloid Interface Sci. 2001, 240, 54-56. (32) Ruzicka, B.; Zulian, L.; Ruocco, G. Langmuir 2006, 22, 1106-1111.
Figure 1. FEGSEM images of (a) a Laponite-armored polystyrene latex made via Pickering miniemulsion polymerization (scale bar ) 100 nm). (b) Film formed from a Laponite-armored polystyrene latex at 230 °C (scale bar ) 400 nm).
surface of the emulsion droplets. This warrants the stability of the emulsion. The morphologies of the latexes obtained after polymerization therefore are anticipated to be armored polymer colloids whose surfaces are covered with Laponite clay disks. Figure 1a shows the FEGSEM image of a group of Laponitearmored polystyrene spheres made via Pickering miniemulsion polymerization. Note that the fine structure (sub-10 nm) is the result of the sputtered gold layer. We should also mention that excess amounts of Laponite clay were observed in all samples. Figure 1b is the FEGSEM image of a film formed from these Pickering polystyrene latexes at 230 °C. This now more clearly shows the armored structure of the individual latex particles that formed the film after limited polymer-polymer interdiffusion. Tapping mode AFM (Figure 2) carried out on a single large Laponite clay-armored polystyrene particle clearly reveals that the Laponite disks lie flat on the surface of the particle. This behavior is predicted from theoretical studies on acicular and discotic Pickering stabilizers.33,34 Control of Particle Size in Pickering Miniemulsion Polymerization. We performed two series of Pickering miniemulsion polymerizations of styrene by varying the quantity of Laponite nanoparticles from 0.25 to 1.5 wt % with respect to water, at a constant sodium chloride concentrations of 0.1 mol L-1 and a constant monomer to water weight ratio of approximately 0.1. (See Table 1 for details.) In these sets of experiments, both 4 and 8 wt % (with respect to styrene) hexadecane as the hydrophobe were employed. The reason for varying the amount of clay was to investigate its influence on the particle size distributions of the resulting latexes. In conventional emulsion polymerizations that use low-molecularweight surfactants such as sodium dodecyl sulfate in concentrations above the critical micelle concentration (cmc), the number of particles generated (Npart) and therefore the size of the individual (33) Dong, L.; Johnson, D. T. Langmuir 2005, 21, 3838-3849. (34) Nonomura, Y.; Komura, S.; Tsujii, K. J. Phys. Chem. B 2006, 110, 1312413129.
Pickering Miniemulsion Polymerization
Langmuir, Vol. 23, No. 16, 2007 8319
Figure 3. Schematic representation of the 2D square lateral packing of the Laponite disks on a flat surface.
microemulsions whereby monomer cores were surrounded by surfactant molecules. Analogous to these, we have developed a basic model to predict the average particle size of our Pickeringstabilized latexes. There are two factors that we have to take into account. First, we have to realize that potentially not all solid particles (i.e., Laponite RD clay pellets) are adhered to the interface of the miniemulsion droplets or latex particles. The overall mass balance for the solid particles therefore is
C0 ) Csurf + Cexcess
Figure 2. Tapping mode AFM images (250 nm × 250 nm) obtained from the surface mapping of a single large Laponite-armored polystyrene latex sphere. The top image is the height (10 nm full scale), the center image is the amplitude, and the bottom image is the phase. (See Supporting Information file S1 for the selection of the single sphere.)
particles show a strong dependence on the amount of surfactant used (concentration of surfactant, [S]). The straightforward Smith-Ewart model predicts a dependence of Npart ∝ [S]0.6, which means that the radius of the particles shows a dependence with an exponent of -0.2.35 More elaborate models allowing, for example, aqueous-phase kinetics and compartmentalization predict different exponents. Antonietti et al.36 and Wu37,38 discussed simple models to predict the size of spherical (35) Smith, W. V.; Ewart, R. H. J. Chem. Phys. 1948, 16, 592-599. (36) Antonietti, M.; Bremser, W.; Muschenborn, D.; Rosenauer, C.; Schupp, B. Macromolecules 1991, 24, 6636-6643. (37) Wu, C. Macromolecules 1994, 27, 298-299. (38) Wu, C. Macromolecules 1994, 27, 7099-7102.
(1)
in which C0 is the overall concentration of solid particles in water in g g-1, with m0/mwater, Csurf is the concentration of solid particles adhered to the oil-water interface (i.e., to the monomer droplets or polymer particles, with respect to the amount of water phase, or msurf /mwater), and Cexcess is the excess concentration of solid particles that remain in the continuous phase, in the present case water, with mexcess/mwater. Note that we assume that the solid particles are added to the continuous phase and not to the tobe-dispersed phase prior to preparation of the (mini)emulsions. We also assume that the energy barrier to enter the dispersed phase is too high for the solid particles to overcome and thus that there are no particles present in this phase. Second, we have to come up with an expression that describes the surface coverage of the droplets. We assume hereby (i) that the liquid-liquid interface is “fully” covered, (ii) that the monomer droplets/polymer particles and the solid Laponite disks are uniform in size, and (iii) that the dimensions of the Laponite clay discs are negligible with respect to the size of the monomer droplets/polymer particles. The latter assumption ignores curvature and thus geometrical constraints. For simplicity we will assume here a 2D square lateral packing of the Laponite disks (Figure 3). This means that the disks lie flat on the surface. The latter is plausible from theoretical studies on acicular33 and discotic particles34 and in our case is confirmed experimentally (Figure 2.). The packing can easily be changed to different arrangements, such as hexagonal or random. The interfacial area of one monomer droplet/polymer latex equals
aoil ) πdoil2
(2)
with doil being the diameter of the droplets/polymer latex. The effective area covered by one Laponite clay disk equals
apart ) dpart2 in which dpart is the diameter of a Laponite clay disk.
(3)
8320 Langmuir, Vol. 23, No. 16, 2007
Bon and ColVer
The total number of monomer droplets/polymer particles can be expressed as
moil Foil Noil ) π 3 d 6 oil
(4)
in which moil is the combined amount of monomer/polymer and hexadecane and Foil is the combined density of monomer/polymer and hexadecane. The total number of Laponite clay disks adhered to the liquidliquid interface can be calculated from
4 Npart )
( ) msurf Fpart
πdpart2h
(5)
in which Fpart is the density of Laponite RD and h is the height of the disks. When we assume full coverage, the following relationship holds:
Noilaoil ) Npartapart
(6)
Substitution of eqs 2-5 into eq 6 and isolation of the only unknown parameter, which is the actual number of Laponite clay disks adhered to the interface, yields
( )( )
3π Fpart h m msurf ) 2 Foil doil oil
( )( )
mhd mpsty Fhd + F mhd + mpsty mhd + mpsty psty
(11)
The reason that these two sets show slightly dissimilar linear
(8)
(9)
To answer this, we need to know the diameter of the monomer droplets and/or polymer latex particles stabilized with Laponite clay disks. We measured the particle size of dialyzed Laponitearmored latexes with dynamic light scattering. The results are given in Table 1. When we insert the experimental data from Table 1 into eq 9 with h ) 1.0 nm, Fpart ) 2570 kg m-3, Foil as a combined value of the densities of polystyrene (Fpsty ) 1090 kg m-3), and hexadecane (Fhd ) 770 kg m-3), their fractional contributions corrected for overall monomer conversion, xM, calculated via
Foil )
Series I with 4 wt % hexadecane (with fit r2 ) 0.999):
(7)
The question now is, how does this expression behave under experimental conditions? In other words, how does the diameter of the Pickering monomer droplets or polymer latex correlate with the added overall concentration of Laponite clay. Can we express Csurf as a function of C0? In generic form
Cexcess ) C0 - f (C0)
With values for moil being the sum of mhd and mpsty and mpsty being mstyxM, we can construct a plot of C0 versus Cexcess (Figure 4). From this Figure, it can be observed that there is apparently linear behavior for the two series of experiments carried out using two different levels of hexadecane (i.e., 4 and 8 wt %), and thus f(C0) can be expressed as first-order polynomial functions in C0.
Cexcess ) (1 - 0.2438)C0 - 7.543 × 10 -4
The factor of (3π/2) changes if one assumes a different surface packing. For example, it becomes x3π when we assume hexagonal packing. Combination with the mass balance from eq 1 yields our final expression:
3π Fpart h C Cexcess ) C0 2 Foil doil oil
Figure 4. Calculated excess concentration of solid particles that remain in the continuous phase (Cexcess) vs the overall concentration of solid particles in water (C0) in g g-1 (series I, 3; series II, ×; series III, 2). The dotted lines are eqs 11 and 12.
(10)
Series II with 8 wt % hexadecane (with fit r2 ) 0.992): Cexcess ) (1 - 0.3076)C0 - 6.468 × 10 -4
(12)
behavior may originate in differences between the interfacial tensions. The evident linear correlation implies that for the current range of experimental conditions the partitioning of the Laponite clay platelets between the continuous water phase and the oilwater interface is a constant. In other words, the number of Laponite clay nanoparticles used dictates how much interface is created! This means that for a specified amount of monomer (moil) the average particle size of the Pickering-stabilized emulsion droplets obtained after emulsification via in the present case of sonication will have a fixed dependent value. It is important to realize that during the emulsification process the adhesion of the particles to the oil-water interface is reversible, as a direct result of the high energy input via sonication. This reversibility of the adhesion process allows for the partitioning of the Laponite clay platelets to reach equilibrium. This is directly reflected in the linear behavior of the plots of C0 versus Cexcess. The average diameter of the Pickering emulsion droplets and thus of the resulting latexes can be predicted and calculated by first obtaining values for Cexcess using eq 11 or 12 and then calculating doil from eq 8. The calculated results are given in Table 1 and show good correlation with the measured values obtained from DLS. Noteworthy is that for entries 2 and 9 in series II (i.e., large particles), a deviation is observed that can be ascribed to a more polydisperse particle size distribution, thereby overestimating the DLS data and the possible influences of gravity on the time scale of the DLS measurements.
Pickering Miniemulsion Polymerization
Langmuir, Vol. 23, No. 16, 2007 8321
Figure 5. Monomer conversion (xM) versus time (min) for Pickering miniemulsion polymerizations of styrene stabilized with Laponite clay (series I, Table 1).
To check the robustness of our findings, we varied the amount of styrene used and kept the amount of Laponite clay constant in a third series of experiments (Table 2). To our initial surprise entries III-1, III-4, and III-5 deviated from the expected linear relationship (eq 11). However, a closer look at the input values reveals that for these three experiments the ratio of Laponite clay discs to monomer and hexadecane is large. The calculated values for the diameters of the latexes for these experiments are 155, 120, and 160 nm for entries III-1, III-4, and III-5, respectively (using eqs 11 and 8). However, experimental values of 227, 182, and 225 nm were obtained. The predicted particle sizes are so small that one of the assumptions made in our model (i.e., (iii) that the dimensions of the Laponite clay discs are negligible with respect to the size of the monomer droplets/polymer particles) no longer holds. The curvature of the droplets now becomes an important factor, which can no longer be neglected. We believe that this effect results in an underestimation of the true experimental values for the particle diameters. Rate of Polymerization in Pickering Miniemulsion Polymerization of Styrene. The overall rates of polymerization for the Pickering miniemulsion polymerizations carried out in series I-III were monitored by determining the monomer conversion (xM) as a function of time using gravimetry. Figure 5 shows the monomer conversion versus time for series I. (For raw data series I-III, see Supporting Information.) As one can clearly see, the overall rate of polymerization is larger for smaller particles. This is the direct result of the compartmentalization of the system. In short, this means that two growing polymer chains cannot undergo bimolecular termination if they are present in two separate particles. In other words, they are compartmentalized, which results in an overall higher radical concentration and thus a higher rate of polymerization. When we assume that the rate of polymerization is first order in monomer concentration, the following expression holds
∫ [R] dt )
- ln(1 - xM) kp
(13)
in which xM is the monomer conversion determined gravimetrically, [R] is the overall radical concentration in mol dm-3, kp is the rate coefficient of propagation for the monomer (i.e., in the present case, styrene with a kp(324.15 K) ) 247.1 dm3mol-1s-1, which is the IUPAC recommended value.)39 As a comparison, we carried out a bulk polymerization for which eq 13 could be (39) Buback, M.; Gilbert, R. G.; Hutchinson, R. A.; Klumperman, B.; Kuchta, F.-D.; Manders, B. G.; O’Driscoll, K. F.; Russell, G. T.; Schweer, J. Macromol. Chem. Phys. 1995, 196, 3267-3280.
Figure 6. Ratios of the values obtained from eq 13 for the Pickering miniemulsion polymerizations and those obtained from eq 14 for the ordinary bulk polymerization of styrene (i.e., φ) as a function of monomer conversion.
approximated at low conversion with the following linear relationship
∫ [R] dt )
-ln(1 - xM) ) 2.723 × 10 -8t kp
(14)
with t being time in s. The ratios of the values obtained from eq 13 for the Pickering miniemulsion polymerizations and those obtained from eq 14 for the ordinary bulk polymerization of styrene, tentatively named φ, are plotted in Figure 6 as a function of monomer conversion. The obtained graphs for φ show that typically values increase to reach a certain plateau value at intermediate monomer conversion with a further increase at high monomer conversion. This ratio (i.e., φ) now clearly shows the effect of compartmentalization. A theoretical value of 1 would agree with ordinary bulk kinetics. Plateau values of 7.17, 5.74, 4.43, 3.45, 1.76, 1.54, and 1.34 are obtained for particle diameters (DLS) of 234.8, 287.7, 391.5, 495.7, 643.2, 607.9, and 658.3 nm, respectively. A clear increase is seen when the average particle size goes down. The onset behavior to these plateau values was rather unexpected and clearly indicates that there is an inhibition and/ or a retardation effect. An obvious possible explanation would be the presence of oxygen as a result of improper degassing of the system prior to polymerization. We repeated some of our experiments with different degassing times (20 min, 1 h, and 2 h) and found exactly the same behavior within experimental error, thereby ruling out oxygen as the inhibition/retardation source. From Figure 6, it clearly can be observed that this behavior becomes more pronounced and extents to higher values of monomer conversion for decreasing particle sizes. The only difference in the recipes for the Pickering miniemulsion polymerizations (series I, Table 1) is that various amounts of clay are used. A likely cause for the onset behavior shown in Figure 6, therefore, could be the presence of Laponite clay. The polymerization reaction and thus the presence of radical species are primarily confined to the Pickering-stabilized particles/ emulsion droplets. It seems plausible to assume that the Laponite clay discs in direct contact with and thus at the surface of the particles/emulsion droplets can have an influence. When oilsoluble initiators are used, the desorption of radical species (exit) becomes more pronounced for small particles.40 A radical species that wants to leave the particle has to cross the Laponite-covered (40) Asua, J. M.; Sudol, E. D.; El-Aasser, M. S. J. Polym. Sci., Part A: Polym. Chem. 1989, 27, 3903-3913.
8322 Langmuir, Vol. 23, No. 16, 2007
interface. It potentially could be terminated by reaction with the Laponite clay disc. This would explain why the observed onset behavior is more pronounced for smaller particles. At monomer conversions exceeding 50%, we clearly see an increase in φ. This can directly be ascribed to the Trommsdorff or gel effect.41 The linear relationship used to express the bulk polymerization of styrene in order to calculate φ is valid only to moderate monomer conversion (eq 14). Not taking into account this effect of enhanced diffusion limitation for termination for the bulk polymerization system leads to the observed increased values for φ. Pickering Miniemulsion Polymerization of Various Monomers. Besides the Pickering miniemulsion polymerizations of styrene, we carried out reactions using different monomers. Pickering miniemulsion polymerizations using Laponite clay discs as stabilizer and with lauryl (meth)acrylate, butyl (meth)acrylate, octyl acrylate, and 2-ethyl hexyl acrylate as monomers were all successful (Table 3). One common characteristic of all these monomers is their hydrophobicity. This appeared to be a crucial factor for success because reactions performed with monomers that have higher water solubility, such as methyl acrylate or methyl methacrylate, were only partially successful under the current experimental conditions.
Conclusions We investigated solid-stabilized, or Pickering, miniemulsion polymerization using Laponite clay discs as the stabilizer. We showed that Pickering miniemulsion polymerization was successful for a variety of hydrophobic monomers (i.e., styrene, lauryl (meth)acrylate, butyl (meth)acrylate, octyl acrylate, and 2-ethyl hexyl acrylate). The Laponite-stabilized miniemulsion polymerizations yielded armored latexes, of which the surfaces of the particles were covered with clay discs. Overall polymerization kinetics of the Pickering miniemulsion polymerizations of styrene showed compartmentalization. Moreover, retardation effects up to intermediate monomer conversions were observed; they were more prominent for the smaller particles and were ascribed to the Laponite clay. A model was presented that allows for the prediction of the average particle size of the latexes produced as a function of the amounts of monomer and Pickering stabilizers used. It shows that under specific generic conditions the number of clay discs used correlates in a linear fashion with the total surface area of the latex particles. This is a direct result of the reversibility of the Laponite clay disc adhesion process under the emulsification conditions used (i.e., sonication). Experimental Section Materials. All monomers were purchased from Aldrich or subsidiary companies at 99% or greater purity and were passed through an alumina column before use in order to remove inhibitors. (41) Trommsdorff, E.; Kohle, H.; Lagally, P. Makromol. Chem. 1948, 1, 169198.
Bon and ColVer n-Hexadecane was purchased from Aldrich, and sodium chloride was purchased from BDH, both at reagent-grade purity. Ammonia was purchased from Fisher at S.G. 0.88 (35%) concentration in water. All were used as supplied. The clay used was Laponite RD and was kindly donated by Rockward Additives Ltd. All initiators used were kindly donated by Wako Initiators and were used as supplied. Apparatus. pH measurements were performed using a Knick 765 calimetic pH meter. The miniemulsion used to make the colloidosomes was formed using a shear force created by a Branon 450 W digital sonifier after being mixed using an Ika Werka Ultra Turrax T25 basic set at 24 000 rpm. Particle size distributions were measured via dynamic light scattering using a Malvern Instruments zetasizer 3000HSa and a zeta sizer 4700 set to 90° after being dispersed using an ultrasound bath made by Kerry ultrasonics limited. The Contin algorithm was used to calculate the intensity mean, which is the value stated for the diameter in all cases. Scanning electron microscopy was performed on a Zeiss supra 55VP FEGSEM. All AFM images were recorded using tapping mode AFM (Digital Instruments, Multimode). Typical Recipe for Pickering Miniemulsion Polymerization. Laponite RD (1.0 g 10 wt %) was added to deoxygenated H2O (100 mL) and sonicated for 4.5 min at 70% amplitude with a 30 s wait every minute. After the first minute interval, NaCl (0.57 g, 0.1 mol dm-3) was added to the sonicating suspension. To a separate beaker, styrene (10.0 g, 0.1 mol, 8.3% solids), hexadecane (0.4 g, 4 wt %), and V-65 (2,2′-azobis(2,4-dimethyl valeronitrile)) (0.05 g, 0.2 × 10-4mol, 0.5 wt %) were mixed and then poured into the clay suspension during agitation by the Ultra Turrax. The emulsion was mixed until there was no visible organic layer. The emulsion was then subjected to ultrasound for 6.5 min at 70% amplitude with a 30 s wait every minute with a maximum temperature setting at 40 °C in order to prevent early polymerization. The resulting emulsion was poured into a 250 mL round-bottomed flask, which was sealed using a rubber seal and bubbled through with N2 for 20 min. The reaction mixture was then heated to 51 °C and gently stirred. Gravimetric analysis was performed by sampling an exact amount (approximately 2 mL) of the polymerizing mixture and depositing it into a foil dish of known mass. The dish was then heated to 120 °C to remove any monomer and hexadecane under vacuum for 48 h to determine the content of solids. Monomer conversion was obtained by taking into account the amounts of clay and salt in the system. After 2 days, the resulting latex was dialyzed in distilled water made up to pH 10 by the addition of concentrated ammonia solution.
Acknowledgment. We thank Neil Wilson for help with AFM measurements and Steve York for help with FE-SEM measurements. We thank the EPSRC for funding (P.J.C.). Supporting Information Available: AFM images of a film of Laponite-armored polystyrene spheres made via Pickering miniemulsion polymerization. Styrene conversion versus time data for Laponite claystabilized miniemulsion polymerizations, series I-III. (See also Tables 1 and 2.) This material is available free of charge via the Internet at http://pubs.acs.org. LA701150Q
