Effects of pH and Salt Concentration on Oil-in-Water Emulsions

Jan 31, 2006 - For continuous emulsions (prepared at high pH with the pH then decreased abruptly or progressively), demulsification takes place rapidl...
0 downloads 8 Views 520KB Size
Download PDF
2050

Langmuir 2006, 22, 2050-2057

Effects of pH and Salt Concentration on Oil-in-Water Emulsions Stabilized Solely by Nanocomposite Microgel Particles Bernard P. Binks,*,† Ryo Murakami,† Steven P. Armes,‡ and Syuji Fujii‡ Surfactant & Colloid Group, Department of Chemistry, UniVersity of Hull, Hull HU6 7RX, and Department of Chemistry, Dainton Building, UniVersity of Sheffield, Sheffield S3 7HF, United Kingdom ReceiVed NoVember 9, 2005. In Final Form: January 3, 2006 Aqueous dispersions of lightly cross-linked poly(4-vinylpyridine)/silica nanocomposite microgel particles are used as a sole emulsifier of methyl myristate and water (1:1 by volume) at various pH values and salt concentrations at 20 °C. These particles become swollen at low pH with the hydrodynamic diameter increasing from 250 nm at pH 8.8 to 630 nm at pH 2.7. For batch emulsions prepared at pH 3.4, oil-in-water (o/w) emulsions are formed that are stable to coalescence but exhibit creaming. Below pH 3.3, however, these emulsions are very unstable to coalescence and rapid phase separation occurs just after homogenization (pH-dependent). The pH for 50% ionization of the pyridine groups in the particles in the bulk (pKa) was determined to be 3.4 by acid titration measurements of the aqueous dispersion. Thus, the charged swollen particles no longer adsorb at the oil-water interface. For continuous emulsions (prepared at high pH with the pH then decreased abruptly or progressively), demulsification takes place rapidly below pH 3.3, implying that particles adsorbed at the oil-water interface can become charged (protonated) and detached from the interface in situ (pH-responsive). Furthermore, at a fixed pH of 4.0, addition of sodium chloride to the aqueous dispersion increases the degree of ionization of the particles and batch emulsions are significantly unstable to coalescence at a salt concentration of 0.24 mol kg-1. The degree of ionization of such microgel particles is a critical factor in controlling the coalescence stability of o/w emulsions stabilized by them.

Introduction Recently there has been growing interest in emulsions stabilized by colloidal particles, so-called Pickering emulsions.1-3 The energy of attachment of a single particle of intermediate wettability at the oil-water interface can be very high relative to the thermal energy kT, so particles once at the interface can be considered as effectively irreversibly adsorbed.4 The effectiveness of the particulate emulsifier depends on the particle wettability, particle size, particle shape, particle concentration, and interparticle interactions.1,2 These particulate emulsifiers offer a number of potential advantages over conventional surfactants such as imparting improved stability against coalescence and a reduced rate and extent of creaming/sedimentation owing to the enhanced viscosity of the continuous phase.1,2 Inorganic particles such as silica, carbon black, barium sulfate, and calcium carbonate have been widely used as particulate emulsifiers.5 However, the use of organic latex particles is much less common. Recently, tailormade sterically stabilized polystyrene latex particles were synthesized and used to stabilize emulsions of oil and water.6-9 The tertiary amine methacrylate-based steric stabilizer was designed to be stimulus-responsive; the wettability of the particle * To whom correspondence should be addressed. E-mail: b.p.binks@ hull.ac.uk (B.P.B.). † University of Hull. ‡ University of Sheffield. (1) Binks, B. P. Curr. Opin. Colloid Interface Sci. 2002, 7, 21. (2) Aveyard, R.; Binks, B. P.; Clint, J. H. AdV. Colloid Interface Sci. 2003, 100-102, 503. (3) Tambe, D. E.; Sharma, M. M. AdV. Colloid Interface Sci. 1994, 52, 1. (4) Binks, B. P.; Lumsdon, S. O. Langmuir 2000, 16, 8622. (5) Binks, B. P.; Lumsdon, S. O. Langmuir 2000, 16, 2539. (6) Binks, B. P.; Murakami, R.; Armes, S. P.; Fujii, S. Angew. Chem., Int. Ed. 2005, 44, 4795. (7) Amalvy, J. I.; Armes, S. P.; Binks, B. P.; Rodrigues, J. A.; Unali, G.-F. Chem. Commun. 2003, 1826. (8) (a) Amalvy, J. I.; Unali, G.-F.; Li, Y.; Granger-Bevan, S.; Armes, S. P.; Binks, B. P.; Rodrigues, J. A.; Whitby, C. P. Langmuir 2004, 20, 4345. (b) Read, E. S.; Fujii, S.; Amalvy, J. I.; Randall, D. P.; Armes, S. P. Langmuir 2004, 20, 7422. (9) Fujii, S.; Randall, D. P. Armes, S. P.Langmuir 2004, 20, 11329.

at the oil-water interface can be changed by tuning the temperature, pH, or salt concentration. Temperature-induced phase inversion from oil-in-water (o/w) to water-in-oil (w/o) emulsions with increasing temperature6 and demulsification of o/w emulsions in situ with decreasing pH9 can be realized. In principle, certain types of microgel particles could also act as a stimulus-responsive particulate emulsifier. Such particles comprise a cross-linked latex which is swollen in a good solvent.10-12 In an aqueous medium, the transition between swollen and nonswollen particles can be triggered by adjusting the dispersion temperature, pH, or salt concentration. Poly(Nisopropylacrylamide), PNIPAM, is the most widely studied polymer for the preparation of temperature-sensitive microgel particles.10-12 pH-sensitive microgel particles usually comprise either a weak polybase, e.g., poly(4-vinylpyridine) or poly(2vinylpyridine), or a weak polyacid, e.g., poly(acrylic acid) or poly(methacrylic acid), which is often copolymerized with NIPAM or styrene.13-19 These microgel particles are swollen either below (polybase) or above (polyacid) the pKa of the ionizable groups. We are aware of only two literature examples in which microgel particles act as a particulate emulsifier. Ngai et al. synthesized PNIPAM-stat-poly(methacrylic acid) copolymer microgel particles cross-linked with N,N′-methylenebisacrylamide.20 The microgel particles were swollen by either lowering the temperature or increasing the pH of the aqueous dispersion. At high pH, both (10) Murray, M. J.; Snowden, M. J. AdV. Colloid Interface Sci. 1995, 54, 73. (11) Saunders, B. R.; Vincent, B. AdV. Colloid Interface Sci. 1999, 80, 1. (12) Pelton, R. AdV. Colloid Interface Sci. 2000, 85, 1. (13) Ferna´ndez-Nieves, A.; Ferna´ndez-Barbero, A.; Vincent, B.; de las Nieves, F. J. Langmuir 2001, 17, 1841. (14) Kim, K. S.; Vincent, B. Polym. J. 2005, 37, 565. (15) Kratz, K.; Hellweg, T.; Eimer, W. Colloids Surf., A 2000, 170, 137. (16) Pinkrah, V. T.; Snowden, M. J.; Mitchell, J. C.; Seidel, J.; Chowdhry, B. Z.; Fern, G. R. Langmuir 2003, 19, 585. (17) Hoare, T.; Pelton, R. Langmuir 2004, 20, 2123. (18) Loxley, A.; Vincent, B. Colloid Polym. Sci. 1997, 275, 1108. (19) Rooney, M. T. V.; Seitz, W. R. Anal. Commun. 1999, 36, 267. (20) Ngai, T.; Behrens, S. H.; Auweter, H. Chem. Commun. 2005, 331.

10.1021/la053017+ CCC: $33.50 © 2006 American Chemical Society Published on Web 01/31/2006

pH and Salt Concentration Effects on o/w Emulsions

swollen particles at 25 °C and nonswollen ones at 60 °C act as an efficient emulsifier for o/w emulsions. However, no stable emulsions could be obtained at pH 2, irrespective of the temperature. In our recent paper, Fujii et al. described the synthesis of poly(4-vinylpyridine)/silica, P4VP/SiO2, nanocomposite microgel particles and their use as a pH-responsive particulate emulsifier of water and several oils.21 At high pH (8-9), the emulsions were stable to coalescence but exhibited creaming/ sedimentation. o/w emulsions were formed with methyl myristate and n-dodecane, whereas a w/o emulsion was preferred with 1-undecanol. In contrast, these nanocomposite microgel particles proved to be ineffective emulsifiers at pH 2-3, with macroscopic phase separation occurring immediately after emulsion formation. Thus, these particles exhibit pH-dependent emulsification. Furthermore, the addition of HCl to a methyl myristate-in-water emulsion originally prepared at pH 8-9 caused rapid and complete demulsification in situ. Thus, these particles also exhibit pHresponsive behavior. The main aim of this study is to explore in more detail the behavior of both the pH-dependent and pH-responsive emulsions of methyl myristate in water identified in ref 21, and determine the critical pH required for their demulsification by employing three different emulsification protocols. The properties of these emulsions are elucidated using conductivity, optical microscopy, and light diffraction measurements and are correlated with the known properties of the nanocomposite microgel particles in aqueous solution in the absence of oil. Ionizable colloidal particles in water are significantly affected by the addition of salt,22-24 since the presence of salt affects the degree of ionization. We have therefore also investigated the effect of salt concentration, at fixed pH, on both the aqueous dispersions and the emulsions prepared from them to understand more fully the impact of particle ionization on emulsifier efficiency. Experimental Section Materials. P4VP/SiO2 nanocomposite microgel particles were prepared by statistical copolymerization of 4-vinylpyridine with a bifunctional cross-linker in the presence of an ultrafine (20 nm diameter) hydrophilic silica sol in an aqueous medium.21 Ammonium persulfate and ethylene glycol dimethacrylate (EGDMA) were employed as the free radical initiator and cross-linker, respectively. The particles were lightly cross-linked; 1.0 wt % EGDMA was employed on the basis of the amount of 4-vinylpyridine monomer. Cross-linking of P4VP with EGDMA is essential to prevent the P4VP chains from dissolving at low pH.25 Details of the synthesis have been reported elsewhere.21 The milky white dispersion of P4VP/ SiO2 particles obtained was carefully purified by centrifugationredispersion cycles, with each successive supernatant being decanted and replaced with doubly distilled water until no excess silica sol remained, as observed by transmission electron microscopy. Thermogravimetric analysis indicated a mean silica content of approximately 35 wt %, which is consistent with the silica content reported for the non-cross-linked P4VP/SiO2 microgel particles.21,25 This fact suggests that the lightly cross-linked microgel particles should have the so-called “currant bun” particle morphology that confers dual surface character at or above neutral pH, i.e., segregated nanodomains of hydrophilic silica and hydrophobic P4VP chains.26-30 (21) Fujii, S.; Read, E. S.; Binks B. P.; Armes, S. P. AdV. Mater. 2005, 17, 1014. (22) Stone-Masui, J.; Watillion, A. J. Colloid Interface Sci. 1975, 52, 479. (23) Schulz, S. F.; Gisler, T.; Borkovec, M.; Sticher, H. J. Colloid Interface Sci. 1975, 52, 479. (24) Binks, B. P.; Rodrigues, J. A. Angew. Chem., Int. Ed. 2005, 44, 441. (25) Fujii, S.; Blanazs, A.; Read, E. S.; Armes, S. P.; Binks B. P.; Murakami, R. Langmuir, to be submitted for publication. (26) Barthet, C.; Hickey, A. J.; Cairns, D. B.; Armes, S. P. AdV. Mater. 1999, 11, 408. (27) Percy, M. J.; Amalvy, J. I.; Barthet, C.; Armes, S. P.; Greaves, S. J.; Watts, J. F.; Wiese, H. J. Mater. Chem. 2002, 12, 697.

Langmuir, Vol. 22, No. 5, 2006 2051 All water used in this study was first passed through an Elga reverse osmosis unit and then through a Milli-Q reagent water system. Methyl myristate (99%, Aldrich) was used as the oil and was columned twice through basic aluminum oxide. Hydrochloric acid (AR grade, Fisher Scientific) used for adjusting the pH and sodium chloride (AnalaR, BDH) were used without further purification. Methods. Preparation of Aqueous Dispersions of Microgel Particles. (i) Aqueous Dispersions of Particles at Various pH Values in the Absence of Added Salt. An aqueous dispersion (1.0 wt %) of P4VP/SiO2 microgel particles was prepared by diluting the original 7.96 wt % aqueous dispersion, at pH 8-9, using Milli-Q water. The pH was then lowered by adding a small volume of aqueous HCl. (ii) Aqueous Dispersions of Particles at pH 4.0 in the Presence of Salt. Solid NaCl was placed in a glass vessel, and approximately 5 mL of a 1.0 wt % aqueous dispersion of P4VP/SiO2 microgel particles was added. The mixture was stirred with a magnetic stirrer for 1 h. The pH of the mixture was adjusted to 4.0 using aqueous HCl, and the mixture was stirred for a further hour. All preparations were carried out at room temperature. Characterization of Aqueous Dispersions of Microgel Particles. To determine the pKa value of the P4VP/SiO2 particles, 5.0 mL of the aqueous dispersion was titrated with aqueous HCl. The pH was monitored using a pH meter (Hydrus 400, FisherBrand) equipped with a pH electrode (FB68791) at room temperature. Calibration was carried out using buffer solutions of pH 4, 7, and 10. Dynamic light scattering of a 1.0 wt % aqueous microgel dispersion was carried out using a Malvern Nano ZS ZEN3600 instrument under a N2 atmosphere at a scattering angle of 173°. Three measurements were made at each pH, with 30 min being allowed for equilibration. ζ potentials were calculated from the measured electrophoretic mobilities determined using the Malvern Nano ZS ZEN3600 instrument. Measurements were averaged over 20 runs using dilute dispersions (0.01 wt %) at different pH values. The dispersion was diluted with aqueous Na2SO4 to provide a suitable background electrolyte concentration to minimize electrical double layer effects. The measurements were carried out at 20 °C. Preparation of Emulsions. In this study we employed three different emulsification protocols. (i) Protocol 1 (Batch Emulsion). Equal volumes (5.0 mL) of oil and aqueous dispersion at different pH values or salt concentrations were placed in a glass vessel (inner volume 14 mL) at room temperature. The two phases were kept in a thermostated bath at 20 °C and then homogenized at 13000 rpm for 2 min with an Ultra Turrax T25 homogenizer (1 cm head) at 20 °C. (ii) Protocol 2 (Continuous Emulsion with pH Changes after Homogenization). Equal volumes (32 mL) of methyl myristate and 1 wt % aqueous dispersion at pH 8.1 were placed in a glass vessel (inner volume 120 mL) at room temperature. The two phases were kept in a thermostated bath at 20 °C and then homogenized at 13000 rpm for 4 min (twice as long due to the increased volume) with an Ultra Turrax T25 homogenizer (1 cm head) at 20 °C. The emulsion was subdivided into glass vessels of 8 mL each whose pH was individually adjusted rapidly by addition of a small volume of aqueous HCl to the continuous phase. (iii) Protocol 3 (Continuous Emulsion). Equal volumes (5.0 mL) of oil and 1 wt % aqueous dispersion at pH 8.1 were placed in a glass vessel (inner volume 14 mL) at room temperature. The two phases were kept in a thermostated bath at 20 °C and then homogenized at 13000 rpm for 2 min with an Ultra Turrax T25 homogenizer (1 cm head) at 20 °C. The pH of the emulsion immediately after preparation was 7.5 and was progressively lowered by adding a small volume of aqueous HCl with gentle shaking. Characterization of Emulsions. The emulsion type was inferred by observing whether a drop of the emulsion dispersed when added to a small volume of water or oil. The stabilities of emulsions at 20 (28) Percy, M. J.; Barthet, C.; Lobb, J. C.; Khan, M. A.; Lascelles, S. F.; Vamvakaki, M.; Armes, S. P. Langmuir 2000, 16, 6913. (29) Agarwal, G. K.; Titman, J. J.; Percy, M. J.; Armes, S. P. J. Phys. Chem. B 2003, 107, 12497. (30) Amalvy, J. I.; Percy, M. J.; Armes, S. P.; Leite, C. A. P.; Galembeck, F. Langmuir 2005, 21, 1175.

2052 Langmuir, Vol. 22, No. 5, 2006

Binks et al.

°C were assessed by monitoring the movement of the oil-emulsion and water-emulsion interfaces. For water-continuous emulsions, the upward movement of the water-emulsion boundary was used as a measure of the stability to creaming, and the position of the oil-emulsion interface was used as an indicator of coalescence. The conductivities of aqueous NaCl, aqueous dispersions, and emulsions were measured using a Jenway 4510 conductivity meter with Pt/Pt black electrodes. The pH of an emulsion was monitored by a Hydrus 400 pH meter (FisherBrand). Small samples of emulsion/dispersion were placed in a hemocytometer cell (Weber Scientific) and viewed with a Nikon Labophot microscope fitted with a QICAM 12-bit Mono Fast 1394 camera (QImaging). Photographs of vials containing emulsions were taken with a Dimage Xg digital camera (Konica Minolta). All images were processed with Image-Pro Plus version 5.1 software (Media Cybernetics). Volume-weighted droplet size distributions of o/w emulsions were measured using a Malvern MasterSizer 2000 instrument at room temperature. The emulsion droplets were diluted in water at the same pH and NaCl concentration as those in the original aqueous dispersion and circulated through the dispersion unit. The optical unit was cleaned between samples by being rinsed with 2-propanol and pure water several times. At least two measurements were made on separate samples for each system.

Figure 1. Degree of ionization of basic groups, e.g., pyridine, versus pH - pKa calculated using eq 4.

Results and Discussion The nanocomposite microgel particles used here become increasingly cationic upon lowering the pH of the aqueous dispersion. This is mainly due to protonation of the pyridine groups on the P4VP chains but also due in part to protonation of the anionic silanol groups on the surface of the silica sol. More importantly, the particles also become much more hydrophilic and swollen with water. It is worthwhile to consider the relationship between the pH of the aqueous dispersion and the degree of ionization of the polymerized pyridine groups in the particles. Let us consider the following equilibrium for 4VP

-C5H4NH+ T -C5H4N + H+ and define the proton dissociation constant of a pyridine group Ka as

Ka ) {[-C5H4N][H+]}/[-C5H4NH+]

(1)

and the degree of ionization of a pyridine group R as

R ) [-C5H4NH+]/{[-C5H4N] + [-C5H4NH+]}

(2)

Combining eqs 1 and 2, we obtain the Henderson-Hasselbach equation

pH - pKa ) log[(1 - R)/R]

(3)

R ) 1/(1 + 10pH-pKa)

(4)

or

Figure 1 shows the relationship between R and pH calculated on the basis of eq 4. It is found that the R value depends critically on the pH - pKa difference, especially when the values of pH and pKa lie close together. Since emulsion stability depends crucially on the wettability of the particles at the interface,1 it was expected that the behavior of emulsions stabilized with these P4VP/SiO2 microgel particles would change drastically at pH values close to the pKa of the particles in the bulk aqueous dispersion. Aqueous Particle Dispersions. Effect of pH. Values of the ζ potential and hydrodynamic diameter of the nanocomposite

Figure 2. ζ potential (4) and hydrodynamic diameter (9) of P4VP/ SiO2 microgel particles versus pH for an aqueous dispersion at 20 °C in the absence of salt. The dotted line indicates a pKa of 3.4. At each pH, three measurements of size were made, standard deviation 10 nm.

microgel particles are plotted against the pH of the aqueous dispersion in Figure 2. These ζ potential data show a classical “S” shape, being positive at low pH and negative at high pH. The isoelectric point of the particles is estimated to be around 6, which is practically identical to the value reported for noncross-linked P4VP/SiO2 microgel particles.27 The particles possess one cationic (pyridine) and two anionic (silanol and sulfate) ionizable surface groups. The sulfate groups, originating from the persulfate initiator, are ionized over a wide range of pH because the pKa values for sulfuric acid in water are -3 and 2.31 Similarly, the silica sols used to prepare these P4VP/SiO2 particles remain anionic over a wide range, displaying negative ζ potentials ranging from -30 mV at pH 2.5 to -56 mV at pH 9.1.27 Hence, the P4VP/SiO2 particles are anionic at high pH owing to their ionized silanol and sulfate groups, but become cationic at low pH due to protonation of the pyridine groups. The hydrodynamic diameter of the nonprotonated “hard” sphere microgel particle is almost constant (at about 250 nm) at pH > 8. At around pH 3.4 (which corresponds approximately to the pKa value of the P4VP chains; see later), the diameter increases dramatically with decreasing pH. This is attributed to swelling of the cationic microgel particles with water. Just either side of their isoelectric

pH and Salt Concentration Effects on o/w Emulsions

Figure 3. Optical micrographs of 1.0 wt % aqueous dispersions of P4VP/SiO2 microgel particles at various solution pH values in the absence of salt: (a) 3.0, (b) 3.4, (c) 5.5, (d) 7.5, (e) 4.0. (f) is for pH 4.0 in the presence of 0.24 mol kg-1 NaCl. The scale bar is the same on each micrograph.

Figure 4. Titration curves obtained for 1.0 wt % aqueous dispersions of P4VP/SiO2 microgel particles using aqueous HCl at various NaCl concentrations: (1) 0, (2) 0.010, (3) 0.020, (4) 0.040, (5) 0.060, (6) 0.10, (7) 0.14, (8) 0.22, (9) 0.26 mol kg-1.

point (pH 6), the microgel particles are too flocculated to allow estimation of their primary diameter. Figure 3 shows optical microscopy images of 1.0 wt % aqueous dispersions at different pH values. At pH 7.5 (d), there is no significant observable matter because the discrete particles are too small (about 250 nm). Highly flocculated particles are observed around the isoelectric point (c) as expected. On lowering the pH further, slightly cationic particles remain weakly flocculated (b). Because the particles are highly cationic and swollen below their pKa value of 3.4 (a), they are barely visible since the refractive index difference between the aqueous medium and the microgel particles is very small.18,19 Effect of Salt Concentration. Figure 4 shows the titration curves of 5.0 mL of 1.0 wt % aqueous dispersions of the P4VP/SiO2 particles at various molalities of NaCl (mNaCl) using HCl. In all cases, an inflection point can be observed as the pyridine groups

Langmuir, Vol. 22, No. 5, 2006 2053

Figure 5. (a) pKa of the P4VP/SiO2 microgel particles estimated from Figure 4 against NaCl molality. (b) Estimated degree of ionization of pyridine groups in the P4VP/SiO2 microgel particles as a function of NaCl molality using eq 4 and pKa values shown in (a) at pH 4.

become protonated on addition of acid. The pH value obtained for a given amount of HCl increases with increasing mNaCl. The pKa value for each curve was estimated by determining the slopes of the different parts of the curve and estimating the pH at which the slope is a minimum. In Figure 5a the dependence of the pKa on mNaCl is shown. At mNaCl ) 0, the pKa is 3.4, but addition of NaCl increases this value. It is expected that the particles are nonswollen above the pKa vs mNaCl curve while cationic and swollen below it. Furthermore, the degree of ionization of the P4VP chains in the particles as a function of mNaCl at a fixed pH 4 can be evaluated using eq 4 and the pKa values given in Figure 5a. As seen in Figure 5b, addition of salt enhances ionization due to screening of the charges; R increases sharply from a value of ∼0.2 at mNaCl ) 0 to more than 0.8 at 0.05 mol kg-1 NaCl. It has been reported that microgel particles containing ionizable groups become deswollen in the presence of salt because an increase in ionic strength decreases the Debye screening length and hence reduces the repulsive electrostatic forces between charged groups.14,15 However, our light scattering measurements gave a hydrodynamic diameter of the microgel particles of 700 nm at pH 4.0 and mNaCl ) 0.2 mol kg-1, compared with ∼300 nm in the absence of salt. In addition, microscopy observations showed that at this pH particles are flocculated in the absence of salt, Figure 3e, but barely visible and hence probably swollen at mNaCl ) 0.24 mol kg-1, Figure 3f. Therefore, the P4VP chains are cationic and swollen at low pH and high mNaCl. It is also noticed that the molality of NaCl corresponding to R ) 0.5 is estimated to be 0.01 mol kg-1 at this pH. Emulsions of Oil and Water. Effect of pH. Figure 6 shows the conductivities of 1.0 wt % aqueous dispersions of the P4VP/ SiO2 particles and also of the emulsions obtained from the homogenization of methyl myristate and an aqueous dispersion (1:1 by volume) using protocol 1 at different pH values. The conductivity of the aqueous dispersion monotonically increases on lowering the pH. At 3.4 e pH e 8.9, the conductivity of the emulsion is much higher than that of pure methyl myristate and is comparable to that of the aqueous dispersion. In addition, drop tests indicated that these emulsions were water-continuous. Thus, o/w emulsions are obtained at all pH values. However, at pH 3.3, complete phase separation of the emulsion was observed less than 3 min after homogenization.

2054 Langmuir, Vol. 22, No. 5, 2006

Figure 6. Conductivities of a 1.0 wt % aqueous dispersion of P4VP/ SiO2 microgel particles and methyl myristate-in-water emulsions prepared from the dispersion using protocol 1 immediately after preparation as a function of pH. The vertical dotted line indicates the boundary between stable and unstable emulsions (with respect to coalescence).

Binks et al.

Figure 8. Optical micrographs of o/w emulsions prepared using protocol 1 immediately after preparation at various pH values of the 1.0 wt % aqueous particle dispersion: (a) 3.0, (b) 3.4, (c) 5.5, (d) 7.5. The scale bar is the same on each micrograph.

Figure 7. Average drop diameters of o/w emulsions as a function of pH for freshly prepared emulsions using protocol 1. Sizes determined using light diffraction (0, arithmetic mean diameter; 9, median diameter) are shown. The standard deviation of the mean is in the range 3-5 µm. The dashed line indicates the boundary between stable and unstable emulsions.

The average drop diameter of o/w emulsions prepared using protocol 1 immediately after homogenization is plotted against pH in Figure 7. At 4 e pH e 8.9, a distinct shallow minimum occurs at around pH 5-6. This pH corresponds approximately to the isoelectric point, where the P4VP/SiO2 particles are known to be highly flocculated (Figure 3c). Thus, the flocculated particles appear to stabilize finer oil drops. Significant flocculation of the emulsion drops occurs at 3.4 e pH < 4 (see later). If the flocs are stable to dilution, the apparent diameter measured is larger than that of individual drops. The close correspondence of the mean and median diameters reflects the unimodal log-normal size distributions in these samples. In Figure 8, optical micrographs of the emulsions prepared using protocol 1 and diluted with water at the same pH as that of the aqueous dispersion are shown immediately after homogenization. At relatively high pH (c, d), drops are discrete, spherical, and simple (not multiple). At 3.4 e pH < 4 (b), drops are also spherical and simple but appear flocculated. Below pH 3.3 (a), smaller drops are observed (the sample was taken from the

Figure 9. Long-term stability of o/w emulsions stored at 20 °C: emulsions prepared using protocol 1 (a) or protocol 2 (b) after 6 months and protocol 3 (c) after 48 h. All emulsions are of the o/w type. fo (4) relates to coalescence, and fw ([) relates to creaming. The dotted line indicates the boundary between stable and unstable emulsions.

aqueous phase after phase separation), although the emulsions rapidly phase separated. However, these drops disappeared from the aqueous phase within 24 h. The long-term stability (after 6 months) of o/w emulsions prepared using protocol 1 is evaluated in terms of the fractions of resolved oil, fo, due to coalescence and resolved water, fw, due to creaming. In Figure 9a, the parameter fo (or fw) is defined as the volume of oil (or water) resolved relative to the initial volume of oil (or water). The emulsions at and above pH 3.4 creamed to a similar extent at all pH values but were completely stable to coalescence. On the other hand, at pH 3.3, emulsions are very unstable to coalescence and complete phase separation occurred

pH and Salt Concentration Effects on o/w Emulsions

within 3 min of preparation. This distinct boundary between stable and unstable emulsions lies at around pH 3.4. It can be recalled that the pH of this boundary is identical to the pKa of the particles in the bulk aqueous dispersion (Figure 5a). Therefore, once particles become sufficiently ionized, they become more hydrophilic, swell, and no longer adsorb at the oil-water interface, leading to significant coalescence of bare oil drops. Parts b and c of Figure 9 show the stability of continuous emulsions prepared using protocols 2 and 3, respectively. Here, the pH of the emulsion was monitored instead of that of the aqueous dispersion. Equating the pH of the dispersion with that of the emulsion is valid since the pH of emulsions prepared by protocol 1 was practically identical to that of the aqueous dispersions prior to homogenization. Although the microgel particles are already adsorbed in protocols 2 and 3, an abrupt decrease in emulsion pH starting from around 8 is effected for protocol 2 whereas a progressive reduction occurs with protocol 3. However, there is no significant difference between the results using protocol 2 or 3, so the demulsification is independent of the manner in which the emulsion pH is adjusted. In both protocols, demulsification again occurs rapidly below pH 3.4, implying that the initially adsorbed particles become swollen cationic microgels and then detach from the oil-water interfaces in situ; i.e., both the emulsifier and the emulsion are pHresponsive. The critical pH for demulsification of continuous emulsions prepared using protocols 2 and 3 is practically identical to the pKa value of the P4VP/SiO2 particles in the aqueous dispersion. This behavior differs from that of hexadecane-in-water emulsions stabilized by poly[2-(dimethylamino)ethyl methacrylate-blockmethyl methacrylate] (PDMA-b-PMMA)-stabilized polystyrene particles, in which the emulsion prepared at high pH (8) was not significantly demulsified by adding HCl to the emulsion.8 In both systems, a part of the particles is immersed in the oil phase. However, in the present system, the nanocomposite microgel particles can be swollen by water, and then the surface of the particles immersed in the oil phase as well as the core can be protonated by lowering the pH. It is suggested that the difference in the demulsification behavior between the PV4P/SiO2 microgel particle system and the PDMA-b-PMMA-stabilized polystyrene particle system is due to the difficulty of achieving high degrees of protonation of PDMA groups in the oil phase.21 Effect of Salt Concentration. In the absence of salt, all emulsions prepared below the pKa of the particles, i.e., above R ) 0.5, were very unstable and rapidly phase separated. On the other hand, the addition of salt increases the pKa value, and the degree of ionization at a fixed pH is also increased. It was therefore expected that the P4VP/SiO2 particles would become more charged and hydrophilic in the presence of added NaCl, leading to coalescence instability of the emulsions. Figure 10 shows the conductivities of a 1.0 wt % aqueous dispersion of P4VP/SiO2 particles and the corresponding methyl myristate-in-water emulsion plotted against mNaCl at pH 4. The conductivity of an aqueous NaCl solution alone is included as a reference. The conductivity of the aqueous dispersion increases with increasing mNaCl and is practically identical to that of the aqueous NaCl solution, implying that at this pH the particles make a negligible contribution to the measured conductivity. At mNaCl ) 0.20 mol kg-1, the conductivity of the emulsion is much larger than that of methyl myristate and is approximately half that of the aqueous dispersion due to the obstruction effect of the oil drops. In addition, drop tests confirmed the existence of o/w emulsions at all mNaCl values. At mNaCl ) 0.22 mol kg-1 it was not possible to measure the conductivity of the emulsion

Langmuir, Vol. 22, No. 5, 2006 2055

Figure 10. Conductivity versus NaCl molality at 20 °C: (2) 1.0 wt % aqueous dispersion at pH 4.0, (4) aqueous NaCl at pH 4.0, (9) o/w emulsion at pH 4.0.

Figure 11. Drop diameter of o/w emulsions versus NaCl molality in the aqueous dispersion for batch emulsions of equal volumes of methyl myristate and a 1.0 wt % aqueous dispersion of P4VP/SiO2 microgel particles at pH 4. Sizes were determined by light diffraction (0, arithmetic mean diameter; 9, median diameter). The error bar for the mean was estimated from the standard deviation of two sets of data.

due to the significant coalescence which occurred at this mNaCl. At mNaCl ) 0.24 mol kg-1 rapid phase separation occurred within 3 min of homogenization, as expected. The mean diameter of drops in the o/w emulsions immediately after preparation is plotted against mNaCl in Figure 11. This graph can be divided into three regions: below 0.055 mol kg-1, between 0.055 and 0.23 mol kg-1, and above 0.23 mol kg-1, corresponding to emulsions that are stable to coalescence, emulsions increasingly unstable to coalescence, and complete phase separation, respectively. The smallest drop diameters of 30 µm appear in the first region where emulsions are flocculated by low salt concentration (see later). These values reflect those of the flocs rather than discrete drops. In the light scattering measurements, emulsions were diluted and sheared in the dispersion unit. However, after the measurement, both discrete and flocculated oil drops were observed; the diameters of the flocs ranged from 20 to 50 µm as judged by optical microscopy. Above 0.05 mol kg-1, the mean drop diameter increases gradually with increasing mNaCl as a result of coalescence and reaches an average diameter of 160 µm, after which stable emulsions cannot be prepared. Optical micrographs of fresh emulsions diluted with water at pH 4 using the same mNaCl as that for the aqueous dispersion are

2056 Langmuir, Vol. 22, No. 5, 2006

Binks et al.

Figure 12. Optical micrographs of o/w emulsions at various NaCl molalities at pH 4: (a) 0, (b) 0.02, (c) 0.05, (d) 0.06, (e) 0.22, (f) 0.24 mol kg-1. The scale bar is the same on each micrograph. Optical micrographs of emulsions at a NaCl molality of 0.04 mol kg-1 recorded at different focal plane positions: (g) top, (h) middle, (i) bottom. The scale bar is the same on each micrograph.

shown in Figure 12. Oil drops in the emulsion prepared in the absence of salt are discrete, spherical, and simple (a). At 0.01 < mNaCl e 0.05 mol kg-1 (b, c), the drops are simple but flocculated. On the basis of these latter micrographs, the degree of flocculation appears to decrease with increasing mNaCl. Addition of salt causes two competing effects: it increases the shielding of ionized groups and also increases the degree of ionization of pyridine groups on the surface of the particles. It seems that the latter effect becomes more important with increasing mNaCl. At intermediate mNaCl (0.06-0.22 mol kg-1), oil drops were not flocculated and increased in size with increasing mNaCl (d, e). At mNaCl ) 0.24 mol kg-1 (f), small drops were observed, although this sample was extracted from the aqueous phase after phase separation. However, these drops were no longer visible after 1 month. Moderately flocculated emulsions at low mNaCl (around 0.04 mol kg-1) displayed an interesting phenomenon; see Figure 12g-i. These images were recorded for the same sample but different focal plane positions for the central oil drop marked with a dotted circle. In (h), focusing on the equator of the large drop, it appears to be closely surrounded by a ring of many smaller oil drops, suggesting that one bridging particle monolayer might serve to enable adherence of oil drops stabilized by microgel particles without coalescence. Such bridging, leading to a stable configuration, has been reported in various systems including a latex-particle-coated water drop approaching a flat oil-water interface,32,33 vertical emulsion films of oil,34 and a w/o emulsion stabilized by hydrophobic silica particles.35 In this case, we may be observing it in water films between oil drops. Finally, Figure 13a summarizes the stability of the emulsions with respect to both creaming and coalescence on standing for 3 months as a function of salt concentration. The emulsion without (31) Albert, A.; Serjeant, E. P. Ionization Constants of Acids and Bases; Methuen: London, 1962; Chapter 8. (32) Ashby, N. P.; Binks, B. P.; Paunov, V. N. Chem. Commun. 2004, 436. (33) Stancik, E. J.; Fuller, G. G. Langmuir 2004, 20, 4805. (34) Horozov, T. S.; Aveyard, R.; Clint, J. H.; Neumann, B. Langmuir 2005, 21, 2330. (35) Horozov, T. S.; Binks, B. P. Angew. Chem., Int. Ed. 2006, 45, 773.

Figure 13. (a) Long-term stability of emulsions stored at 20 °C after 3 months versus NaCl molality in the aqueous dispersion of P4VP/SiO2 microgel particles at pH 4. All emulsions are of the o/w type. fo (4) relates to coalescence, and fw ([) relates to creaming. (b) Appearance of vials stored at 20 °C after 3 months containing emulsions of methyl myristate and a 1.0 wt % of aqueous dispersion of P4VP/SiO2 particles for different mNaCl values (mol kg-1) at pH 4.

pH and Salt Concentration Effects on o/w Emulsions

added salt creams but is stable to coalescence as already shown in Figure 9a. At 0.01 e mNaCl e 0.05 mol kg-1, the flocculated emulsions are also completely stable to coalescence and the extent of creaming passes through a minimum value. Emulsions at mNaCl ) 0.01 and 0.02 mol kg-1 are more flocculated than those at mNaCl ) 0.04 and 0.05 mol kg-1 as shown in Figure 12. This minimum in creaming is thus attributed to the formation of a gel-like network structure in the highly flocculated emulsions. At intermediate salt concentrations (0.06 e mNaCl e 0.22 mol kg-1), emulsions become unstable to coalescence and both creaming and coalescence increase with increasing mNaCl. Coalescence occurred during the first few minutes and then stopped completely. Further increases in drop diameter were not observed, and the residual emulsions were stable against coalescence for at least 3 months. This phenomenon is called limited coalescence and is typical for particle-stabilized emulsions.36 At high salt concentration, mNaCl ) 0.24 mol kg-1, the system undergoes macroscopic phase separation within 3 min. The appearance of the vials after 3 months of storage at 20 °C containing emulsions of methyl myristate and 1.0 wt % P4VP/ SiO2 particles at various NaCl concentrations is shown in Figure 13b. The instability to coalescence with increasing salt concentration is clearly evident from these data. In addition, the aqueous phase is transparent in the region where emulsions are flocculated (0.01 e mNaCl e 0.05 mol kg-1), indicating most of the particles are incorporated into the creamy white emulsion phase.37 (36) Arditty, S.; Whitby, C. P.; Binks, B. P.; Schmitt, V.; Leal-Calderon, F. Eur. Phys. J. E 2003, 11, 273.

Langmuir, Vol. 22, No. 5, 2006 2057

The salt concentration scan revealed that emulsions stabilized using these microgel P4VP/SiO2 particles become progressively unstable to coalescence, with phase separation eventually occurring on increasing the salt concentration. This is in line with predictions made from Figure 5b regarding the degree of ionization of the P4VP chains.

Conclusions Methyl myristate-in-water emulsions stabilized with lightly cross-linked poly(4-vinylpyridine)/silica nanocomposite microgel particles were studied as a function of pH and salt concentration. The pH scan revealed that the critical pH for phase separation and demulsification of these emulsions is 3.4, which corresponds to the pKa value of the particles in the bulk. Hence, the particles are no longer adsorbed at oil-water interfaces in their swollen microgel form. The addition of salt leads to an increase in the degree of ionization of pyridine groups in the P4VP chains, increasing the hydrophilic character of the particles and hence inducing coalescence of oil drops in water. It can be concluded that the degree of ionization of these nanocomposite particles is crucial for controlling the coalescence stability of oil-in-water emulsions stabilized by them. Acknowledgment. This work was funded by the EPSRC, U.K. (Grants GR/S69283 and GR/S69276). S.P.A. is the recipient of a Royal Society/Wolfson Research Merit award. LA053017+ (37) Binks, B. P.; Lumsdon, S. O. Phys. Chem. Chem. Phys. 1999, 1, 3007.