FORMATION AND STABILITY OF EMULSIONS: EFFECT OF FORMATION AND STABILITY OF EMULSIONS: EFFECT OF SURFACTANT-PARTICLE INTERACTIONS AND PARTICLE SHAPE SURFACTANT-PARTICLE INTERACTIONS AND PARTICLE SHAPE

Formation and stability of emulsions is one of the important topics in the field of colloids and interfacial science. Surfactants and colloidal particles are often used to stabilize emulsions. Surfactants are amphiphilic molecules; they minimize the energy required for the emulsion formation by reducing oil-water interfacial tension. Colloidal particles are not amphiphilic, but partially wettable particles favors the adsorption at oil-water interface with a desorption energy well above thermal energy. With sufficient coverage at the interface, they act as barriers against droplet coalescence and enhance the emulsion stability. In this work, the response of particle-stabilized (Pickering) emulsions to the addition of different surfactant solutions and the stability of surfactant stabilized emulsions to the addition of particle suspensions were studied. There were different end points for emulsion droplets and different particle release modes for Pickering emulsions depending upon the interactions between surfactants and particles, surfactant-particle ratio, and mixing conditions. The effect of particle shape on the formation of Pickering emulsions is also studied. It is found that the inter-particle interactions and particle shape play major role in determining the microstructure and final stability of the emulsions. The combinations of optical, confocal, and Cryogenic scanning electron microscopy were used to determine the final stability and structure of the emulsions. Abstract: We use carboxyl-terminated, negatively charged, carbon black (CB) particles suspended in water to create CB-stabilized octane-in-water emulsions, and examine the consequences of adding aqueous anionic (SOS, SDS), cationic (OTAB, DTAB) and nonionic (Triton X-100) surfactant solutions to these emulsions. Depending upon the amphiphile’s interaction with particles, interfacial activity and bulk concentration, some CB particles get displaced from the octane-water interfaces, and are replaced by surfactants. The emulsions remain stable through this exchange. Particles leave the octane-water interfaces by two distinct modes that depend on the nature of particle-surfactant interactions. Both happen over time scales of the order of seconds. For anionic and nonionic surfactants that bind to the CB through hydrophobic interactions, individual particles or small agglomerates stream away steadily from the interface. the Zeta potentials of CB These results are the first systematic observations of different particle release modes Abstract: As a model for understanding how surfactant-stabilized emulsions respond to the addition of interacting and non-interacting particles, we investigated the response of dodecane-in-water emulsions stabilized by SDS(anionic), CTAB(cationic) and Triton X-100(non-ionic) surfactants to the addition of an aqueous suspension of negatively charged fumed silica particles. The stability of the emulsion droplets and the concentration of surfactants/particles at the oil-water interfaces are sensitive to surfactant-particle interactions, mixing conditions and the particle concentration in the bulk. Addition of the particle suspension to the SDS-stabilized emulsions showed no effect on emulsion stability. The emulsion droplets coalesce when fumed silica particles were added to emulsions stabilized by Triton X-100. Depending on the concentration of silica particles in the suspension, the addition of fumed silica particles to CTAB-stabilized emulsions resulted in droplet coalescence and phase separation of oil and water, or formation of particle-coated droplets. Vigorous (vortex) mixing allows the particles to breach the oil-water interfaces, and the particles help stabilize emulsions. While we have examined a specific particle suspension and a set of three surfactants, these observations can be generalized for other surfactant-particle mixtures. Abstract: Here, we studied the effect of particle shape and inter-particle interaction on the formation and stability of bromohexadecane-in-water emulsions stabilized with spherical and fumed silica particles with similar hydrodynamic diameter. Emulsions were prepared at two different NaCl concentrations 0.1mM and 50mM. We found that the particle shape and inter-particle interactions have strong influence on the creaming behavior and microstructure of the emulsions. At 0.1mM NaCl, there is sedimentation of emulsion droplets stabilized with spherical silica particles and creaming of emulsion droplets stabilized with fumed silica particles. Increasing salt concentration to 50mM lead to the flocculation of emulsion droplets stabilized with spherical silica particles whereas, emulsions stabilized with fumed silica formed a gel like structure. All the emulsions have shown shear thinning behavior. The emulsions stabilized with fumed silica particles have higher viscosity and were yielding at higher strains when compared with emulsions stabilized with spherical silica particles. The degree of shear thinning and yielding has increased with an increase in salt concentration.

The basis for calculation of the free energy difference (we ignore entropic effects) between a surfactant-and a particle-stabilized emulsion. The ground state is an oil droplet with particles and surfactants in the aqueous phase. ∆E surf is the energy difference between the ground state and a state where only surfactants are at the oilwater interface. ∆E part is the energy difference between the ground state and a state where only particles are at the oil-water interface. The sign of ∆E surf -∆E part is the energy difference between a surfactant-and a particle-stabilized drop. R is the radius of the drop and is the contact angle measured through the aqueous phase…………12    showing partial phase separation of oil and water phase after the addition of fumed silica particles b) 0.5 wt% fumed silica particles; inset showing particle-coated emulsion droplets. Confocal fluorescence microscope images of the CTAB-stabilized emulsion droplets after gentle mixing with (c) 0.05 wt% fumed silica particles. The particles do not adsorb at oil water interfaces, but distribute uniformly in the aqueous phase d) 0.5wt% fumed silica particles, showing the formation of particle coated droplets along with the particle networks between droplets. Scale bars = 50µm…….45  particles at oil-water interfaces. SDS-stabilized emulsion drops in presence of (a) 0.05 wt% fumed silica; no particles at oil-water interface. Scale bar = 3µm (b) 0.5 wt% fumed silica. Particles at oil-water interface. Scale bar = 1µm. Triton X-100-stabilized emulsion drops in presence of (c) 0.05 wt% fumed silica; no particles at oil-water interface. Scale bar = 3µm and (d) 0.5 wt% fumed silica; the emulsion droplet is stabilized by both particles and surfactant. Scale bar = 1µm…………………………47 Where, is the change in interfacial area, is interfacial tension, is change in entropy and T is the temperature of the system. In most cases, i.e.
is positive. So, in the absence of any stabilizing mechanism emulsions will become unstable. Surfactants and colloidal particles are often used to stabilize emulsions.  Colloidal particles are not amphiphilic in nature but particles that are partially wettable in each of two immiscible phases will favor locating at the liquid-liquid interfaces. 3,4,5 Unlike surfactants they do not reduce the oil-water interfacial tension, but strongly get adsorbed at the oil-water interface. However, adsorption of the particles on the oilwater interface is a slow process 6, 7 and needs to be enhanced by mixing. The energy, required to remove a single spherical particle from an oil-water interface is given by , Where, r is the radius of the particle, γ o/w is the oil−water interfacial tension, and θ is the three-phase contact angle made by the particle at the oil-water interface. For a 10nm particle, and = 50mN/m, ΔE is ~10 3 kT for 35˚<θ< 145˚. Therefore, thermal fluctuations will be insufficient to remove a particle from the interface if the contact angle is within this range. Once at the interfaces, these particles contribute to electrostatic, steric or rheological barriers against droplet coalescence and effectively stabilize emulsions. 8,9 Wettability of the particles dictates the nature of the emulsion (O/W or W/O) being formed. 10 Recently there is lot of interest in using surfactants and colloidal particles together for emulsion formation. It is driven by a notion that surfactants decrease the oil-water interfacial tension hence lower the particle adoption energy at the interfaces 11,12 or they will modify the wettability 10 of the particles and promote their adsorption at the interfaces. The synergy between particle and surfactant mixtures has been exploited to make particle-stabilized emulsions. 13 Addition of surfactants to a particle-stabilized emulsion or addition of colloidal suspensions to a surfactant-stabilized emulsion are different class of experiments, as they allows the amphiphile to adsorb on the liquid-liquid interfaces as well as on the particles in a controlled way. Here, we studied the effect of addition of surfactant solutions to the stability of the particle-stabilized emulsions and the effect of addition of fumed silica suspensions to the stability of the surfactant stabilized emulsions. We also looked the effect of particle shape on the formation and stability of Pickering emulsions. The interactions between the colloidal particles are carefully controlled and the subsequent effects on the emulsion formation and stability are studied. The combination of optical, confocal, and cryogenic scanning electron microscopy were used to determine the final stability and structure of the emulsions. individual particles or small agglomerates stream away steadily from the interface.
Cationic surfactants bind strongly to the carboxylate groups, reduce the magnitude of the surface potential, and cause the CB particles to agglomerate into easily visible chunks at the droplet interfaces. These chunks then leave the interfaces at discrete intervals, rather than in a steady stream. For the longer chain cationic surfactant, DTAB, the particle ejection mode reverts back to a steady stream as the concentration is increased beyond a threshold. This change from chunks of particles leaving intermittently to steady streaming is because of the formation of a surfactant bilayer on the particles that reverses the particle surface charge and makes them highly hydrophilic. The charge reversal also suppresses agglomeration. Zeta potentials of CB particles measured after exposure to surfactant solutions support this hypothesis.
These results are the first systematic observations of different particle release modes from oil-water interfaces produced by variations in interactions between surfactants and particles. They can be generalized to other particle-surfactant systems and exploited for materials synthesis.

Introduction:
The ability of surfactants to lower liquid-liquid interfacial tensions is a key property not have to be amphiphilic, or Janus-like, to locate at oil-water interfaces. Instead, a particle with partial wettability in both immiscible liquid phases can reside preferentially at oil-water interfaces. These interfacially active particles can stabilize emulsions. In addition, the potential to take advantage of the particle shape, size, surface characteristics, as well as other intrinsic properties allows particle-stabilized emulsions to have functionalities that are difficult to replicate using surfactants. 22 The energy required to displace a spherical particle from a liquid-liquid (designated as oil-water in our case) interface into one of the surrounding liquid phases is given by , where r is the radius of the particle, γo/w is the oil-water interfacial tension and θ is the three phase contact angle measured through either the oil or water. For r = 100nm, = 50mN/m and  = 90, Eq. (1) gives ∆E~10 5 k B T. Therefore, once a partially wettable particle is at the interface, it cannot leave spontaneously. This is one of the important distinguishing features of particle-stabilized emulsions, allowing, among other things, for them to remain stable even when the dispersed phase is at a very low volume fraction.
During the formation of a particle-stabilized emulsion, fresh oil-water interfaces must be covered with an adequate number of particles to stabilize the droplets within a time scale that is less than that for drop coalescence. Because breaching of particles into oil-water interfaces is slow 7 , this process needs to be enhanced by mixing. The particles can be charged, providing repulsive interactions between drops, they can provide steric barriers, and increased interfacial viscosity that suppresses thinning of the intervening liquid during approach of drops, thus resisting coalescence. 5,8,9,23 In many practical situations it will be a combination of surfactants and particles that will provide the optimum characteristics for the emulsion, the surfactants often providing the low interfacial tension to facilitate drop formation, and the particles providing enhanced stability 11 . The synergy between particle and surfactant mixtures has been exploited to make particle-stabilized emulsions. 13,14,15,16 Surfactant-particle interactions can be tuned by varying the charge on the head group, the tail length and concentration of the amphiphile, 12,17,18 with potentially useful consequences on emulsion behavior. 10,24,25,26 The ability to tune particle surface characteristics using surfactants has been exploited for porous materials synthesis. 27,28 In all of these experiments, the particles were modified with surfactants prior to the formation of emulsions.
Addition of surfactants to a particle-stabilized emulsion is a different class of experiments, as it allows the amphiphile to adsorb on the liquid-liquid interfaces as well as on the particles in a controlled way. Binkset al. 13  The key distinguishing features of the work reported here are comprehensive sets of experiments that utilize optical microscopy to carefully monitor changes to a charged particle-stabilized emulsion upon addition of surfactants that interact with the particles either through hydrophobic or ion binding. We examine final states and transients, and support our observations using a simplified thermodynamic analysis, as well as zeta potential measurements and confocal microscopy. Our analysis and experiments suggest modes for particle displacement from these interfaces that have not been observed previously.
In order to establish the framework for our observations, we analyze two cases shown in Figure 2.1, and determine the free energy difference between a surfactant-and particle-stabilized emulsion drop, ΔE surf -ΔE part . If this quantity is positive, a surfactant stabilized emulsion would be more stable than a particle stabilized one.
Therefore, from energetic considerations, addition of surfactants would cause particles to be displaced from interfaces as the system seeks a lower energy state.

Figure 2.1:
The basis for calculation of the free energy difference (we ignore entropic effects) between a surfactant-and a particle-stabilized emulsion. The ground state is an oil droplet with particles and surfactants in the aqueous phase. ∆E surf is the energy difference between the ground state and a state where only surfactants are at the oilwater interface. ∆E part is the energy difference between the ground state and a state where only particles are at the oil-water interface The sign of ∆E surf -∆E part is the energy difference between a surfactant-and a particle-stabilized drop. R is the radius of the drop and is the contact angle measured through the aqueous phase.
For this simplified analysis, we assume no interactions between particles and surfactants, and ignore entropic contributions. Under these conditions, . ( Here R is the drop radius, r the particle radius (particles are assumed to be spheres in this analysis), γ o/w is the interfacial tension of the bare oil/water interface, γ surf is the interfacial tension of the surfactant-laden oil water interface, and n is the number of particles at the interface. If the area fraction of the interface covered by particles is ϕ, and assuming R >> r, Substituting Eq.(4) into the expression for gives the condition (5) for a surfactant-stabilized drop to have a lower free energy than a particle-stabilized one. For = 90°, this criterion simplifies to that is, the fractional change in oil-water interfacial tension upon addition of surfactant must be greater than the fractional surface coverage of the interface by particles.
Therefore, addition of a surfactant to a particle-stabilized emulsion can cause particles at an oil-water interface to get displaced if the inequality in Eq.(5) or Eq. (6) is satisfied.
Eqs. (1), (5) and (6) will need to be modified if the particles are not spherical, or the surfactant adsorbs on particles spontaneously in addition to occupying the oil-water interfaces. For fractal particles, as is the case in our experiments as well as those done with fumed silica, the cusps on the particles cause them to get pinned at the liquidliquid interfaces. 23,30,31 Thus, for an equivalent size, the energy barrier for a fractal particle to leave the interface will be greater than that for a spherical particle given by Eq. (1). If the surfactant interacts with the particles and adsorbs on them spontaneously, this exothermic process will cause the free energy change to be greater than ΔE surf , and the displacement of particles will be energetically more favorable than the case with no particle-surfactant interactions. In addition, this adsorption could change the contact angle , with concomitant consequences that can be understood using Eq.(5).
When a surfactant solution is added to a particle-stabilized emulsion, the response will therefore depend upon the ability of surfactant molecules to lower the interfacial tension, as well as the interactions between the particles and surfactants. Local variations in oil-water interfacial tension, and the Marangoni forces that result, can also aid the displacement of particles from the interfaces. We do not quantify this phenomenon. We also note that it is likely that after addition of surfactants to a particle-stabilized emulsion, the final sample has both particles as well as surfactants at the oil-water interfaces.
In this work, we report the behavior of carboxyl-terminated carbon black-stabilized Acros Organics. Hydrochloric acid(HCl, 37wt%) was obtained from Sigma Aldrich.
The surfactant solutions were prepared with water obtained from a Millipore Milli Q system.

Sample preparation
A 0.015wt% CB dispersion was used to prepare the emulsions. The zeta potential of the carbon black particles was measured to be -61.3mV. The pH of the carbon black dispersion was adjusted to 3.2 with HCl to protonate some of the surface carboxylate groups, thus rendering the particles partially hydrophobic. The particle zeta potential at this pH is -10.2mV. The viscosity of the suspension increases significantly because these partially hydrophobic particles form a network in the aqueous medium. To form the 'base' emulsion, 0.2ml of octane were added to 2ml of the CB particle dispersion and vortexed at 3000 rpm for 2 min.
The  Insight into adsorption of surfactants on the particles and the consequences of this on the behavior of the emulsions was obtained by monitoring the zeta potential of the carbon black particles in water at different concentrations of added of surfactants. The zeta potentials were measured using a Malvern Zetasizer Nano-ZS instrument.  To establish a potential end point for samples after addition of surfactants, we successfully prepared stable octane-in-water emulsions using SDS, Triton X-100 and DTAB at their respective CMC concentrations. We were unable to create stable emulsions with SOS and with OTAB because of the low surface activity of these short chain surfactants.

Effect of surfactants on CB-stabilized emulsions:
For SOS, a threshold concentration of ~10mM had to be exceeded before we noticed any impact. At concentrations just above the threshold, addition of SOS shows no obvious changes (    Interestingly, we captured a drop-drop coalescence event in this system, as shown in    and they are not released into the aqueous phase. The reduced particle surface potential also leads to particle aggregation on the oil droplet interface. With increasing surfactant concentration the CB particles start to become hydrophilic within complete bilayer formation, and the particles get ejected as small clusters from the droplet interfaces. The energy of detachment of a particle from the interface scales as the square of its size. The irregular morphology of these large agglomerates also causes them to be pinned strongly at the interfaces. The particle release kinetics is therefore much slower, and the clusters leave from the oil droplet at irregular intervals. The coalescence of emulsion droplets when a DTAB solution is added is a consequence of reduced electrostatic repulsion between drops because of particle charge neutralization and detachment. When the surfactant concentration is increased further, the CB particles becomes very hydrophilic because of complete bilayer formation at the particle surfaces. 1 The particles then assume a positive charge, and this repulsive interaction suppresses interparticle aggregation. The increased hydrophilicity promotes the displacement of particles into the aqueous phase. We note that our results are similar to those obtained by Subramanian et al., 37,38 who observed polystyrene particles getting ejected as singlets and small agglomerates from air-water interfaces when particle stabilized foams were exposed to different surfactants.

Zeta potential measurements:
We support these observations and explanations by monitoring the zeta potentials of carbon black particles in presence of anionic and cationic surfactants at pH 3.2. The results are shown in Figure 2.9.   and their consequence on particle ejection from octane-water interfaces.

Conclusions:
We studied the effect of addition different surfactant solutions on CB-stabilized oil-inwater emulsions. We show conditions for which the displacement of particle from the oil droplet interfaces is thermodynamically favorable. The details of particle ejection are complex, and are strongly influenced by particle-surfactant interactions, the surface activity as well as the concentration of the surfactant. For anionic and nonionic surfactants, which interact with CB through hydrophobic binding, the particles are released in steady streams from the oil droplet interfaces. Cationic surfactants cause CB particle clustering. Clusters of particles then get released intermittently from the interfaces. When the cationic surfactant concentration is increased further, the mechanisms of particle release changes to a steady stream of particles because of bilayer formation on the particles. The interfacial properties of surfactant molecules and the change in wettability of particles in the presence of surfactants play a major role on particle desorption from the oil-water interfaces.

Acknowledgements:
This work was supported by grants from the National Science Foundation (CBET   allows the particles to breach the oil-water interfaces, and the particles help stabilize emulsions. While we have examined a specific particle suspension and a set of three surfactants, these observations can be generalized for other surfactant-particle mixtures.

Introduction:
Surfactants minimize the energy required for the emulsion formation by reducing the oil-water interfacial tension, and they hinder the coalescence of the dispersed phase by forming electrostatic or steric barriers around droplet surfaces. The hydrophilic to hydrophobic balance of the surfactant molecule dictates whether an oil-in-water(O/W) or water-in-oil(W/O) emulsion is formed. 1 Colloidal particles that are partially wettable in each of two immiscible phases will favor locating at the liquid-liquid interfaces. 2,3,4 The energy, required to remove a single spherical particle from an oil-water interface is given by , where r is the radius of the particle, γ o/w is the oil−water interfacial tension, and θ is the three-phase contact angle made by the particle at the oil-water interface. For a 10nm particle, and = 50mN/m, ΔE is ~10 3 kT for 35˚<θ< 145˚. Therefore, thermal fluctuations will be insufficient to remove a particle from the interface if the contact angle is within this range. Once at the interfaces, these particles contribute to electrostatic, steric or rheological barriers against droplet coalescence and effectively stabilize emulsions. 5,6,7 Interactions between surfactants and particles have been exploited for the formation of stable emulsions. 8,9,10 Surfactants decrease the oil-water interfacial tension allowing more interfaces to be created during mixing, 11,12 or they interact with particles, modify their wettability 13,14 and affect their adsorption energy at the interfaces (Eq. 1). In emulsions made with both surfactants and particles, the surfactant to particle ratio and surfactant-particle interactions play a major role in determining the final balance of particles and surfactants at the oil-water interfaces, 15,16 as well as the type of emulsion that is formed. 17,18 The addition of surfactants to particle-stabilized emulsions can result in desorption of particles from oil-water interfaces. 15,19 The extent of desorption is The presence of surfactant molecules on the interface can create barriers for particle adsorption on those interfaces. The magnitude of these barriers can be estimated using DLVO theory. For repulsive electrostatic interactions between the surfactant and the particles, the magnitude of these adsorption barriers vary between ~10 -11 -10 -9 N, depending on the size of the particles and the charges on the oil-water interface and on the particles. 23 The force exerted on a 100nm colloidal particles due to Brownian motion is of the order of ~10 -14 N, 24 insufficient to spontaneously overcome this repulsive barrier. During vigorous (e.g. vortex mixing) mixing, the force on the particles varies between 10 -11 -10 -8 N, 23 thus allowing particles to breach the oil-water interfaces.
In this work, we investigate the consequence of controlled addition of an aqueous suspension of particles to surfactant-stabilized emulsions. The addition of the suspension is followed either by gentle shaking or by more vigorous (vortex) mixing.
In accordance with our estimates of forces, we show that mixing conditions can make a difference to the final state of the emulsions. We chose anionic, cationic, and nonionic surfactants, and used negatively charged fumed silica in an aqueous suspension.
The changes to the emulsion are monitored visually, and by a range of techniques including optical, confocal, and cryogenic scanning electron microscopy. We identified four different end states for emulsion droplets that depend on surfactantparticle interactions, particle concentrations and mixing conditions. These results are generic and can apply to other systems with similar particle-surfactant interactions.

Materials:
Sodium

Sample preparation and analysis:
Dodecane-in-water emulsions were made with surfactant concentrations at about two times the CMC for each surfactant. The fumed silica suspensions did not contain surfactants. To form the "base" emulsion, 0.2 mL of dodecane was added to 1 mL of the surfactant solution and vortexed at 3000 rpm for 1 min. The emulsions were diluted with a volume of particle suspension equal to the volume of the aqueous phase in the emulsion. The final surfactant concentration is therefore at their corresponding CMC values. The final silica particle concentration was varied between 0.05 wt% to 1wt%. The emulsions and the particle suspensions were mixed in two different ways.
The first was gentle shaking to avoid foaming or creation of any new oil−water interfaces. This gentle shaking also minimizes convective transport of the particles.
For the second case, they were vortexed at 3000rpm for 1min. The samples were allowed to rest for 24h after mixing, after which they were analyzed using a range of techniques.
Brightfield optical microscopy images were processed with Image-J to obtain average droplet sizes and size distributions. Silica particles were labeled with 0.5µM of Rhodamine B for confocal fluorescence microscopy on a Zeiss LSM 700. At the concentration we used for labeling, we do not expect the Rhodamine B to affect emulsion properties. 25 Cryogenic Scanning Electron Microscopy (Gatan Alto 2500 cryopreparation system attached to a Zeiss Sigma VP field emission scanning electron microscope) was used to look at the fine structure around the emulsion droplets. Zeta potentials were measured (Malvern Zetasizer Nano-ZS) to get additional insights on surfactant-particle interactions.

Results and Discussion
:   interactions between the particles and the surfactant covered droplets will result in no particles at these interfaces, and therefore a minimal effect on the emulsions. For emulsions stabilized with Triton X-100, we observe an increase in droplet size with an increase in silica concentration (Fig. 3.2(c)). Non-ionic surfactants with ethoxylated groups can form hydrogen bonds with hydroxyl groups on silica surfaces. 9, 26 As the particle concentration increases, more surfactant get adsorbed on particle surfaces, depleting surfactant from the interfaces. The loss of the stabilizing amphiphile results in droplet coalescence. 22 Strong attractive electrostatic interactions dominate between silica particles and CTAB. We observed a partial phase separation of oil and water at 0.05 wt% of silica particles. At low particle concentrations, surfactant adsorption on particle surfaces depletes surfactant from the interfaces and results in droplet coalescence and an increase in average droplet size (Fig. 3.2(d)). As the particle concentration increases, they bind to the surfactant-stabilized droplets and form particle coated emulsions that are stable.  Figure 3.3 shows zeta potentials of fumed silica particles after they are added to a SDS solution. We observed an increase in zeta potential (less negative) of the particles with an increase in particle concentration. We observed a similar behavior when NaCl is added to the fumed silica suspensions, suggesting that the rise in zeta potential is due to reduced overall dissociation of silanol groups at higher particle concentrations, followed by Na + ( counterion) binding on silica surfaces. There is little or no adsorption of the anionic surfactant moiety on silica particles. We observe no change to the zeta potential of fumed silica particles in the presence of Triton X-100 (beyond the change observed when fumed silica is added to water). Fumed silica particles attain a negative charge through the dissociation of silanol groups on the silica surfaces. Triton X-100 interacts with the particles through the formation of hydrogen bonds with undissociated silanol groups, and therefore does not alter the zeta potential of silica particles. 27 Silica particles are positively charged in the presence of CTAB due to the formation of surfactant bilayers on the surface of the particles. Since CTAB can adsorb strongly on fumed silica, we further examined this system under gentle mixing conditions visually as well as with confocal microscopy, and show results in Fig. 3.5.

Figure 3.5:
Optical microscopic images of CTAB-stabilized emulsion droplets after gentle mixing with fumed silica suspensions (a) 0.05 wt% fumed silica particles; inset showing partial phase separation of oil and water phase after the addition of fumed silica particles b) 0.5 wt% fumed silica particles; inset showing particle-coated emulsion droplets. Confocal fluorescence microscope images of the CTAB-stabilized emulsion droplets after gentle mixing with (c) 0.05 wt% fumed silica particles. The particles do not adsorb at oil water interfaces, but distribute uniformly in the aqueous phase d) 0.5 wt% fumed silica particles, showing the formation of particle coated droplets along with the particle networks between droplets. Scale bars = 50µm.
At low particle concentrations, surfactants get depleted from the oil-water interfaces and the emulsion partially destabilizes (Fig. 3.5(a). As the concentration of particles increases, more particles get attached to the droplet surfaces leading to the formation of particle-coated droplets (Fig. 3.5(b)). The particle layers on the droplet surfaces hinder droplet coalescence. Fig. 3.5(c) is a confocal microscope image of the emulsion at low particle concentrations. The particles are distributed quite uniformly in the aqueous phase. The presence of alkyl chains on the silica surfaces increases the particle hydrophobicity 17 and creates hydrophobic patches on the silica surface.
Exposure of these silica particles to water is not energetically favorable and the attractive van der Waals interactions then cause particle chaining in the bulk at high particle concentrations 28 and connections with particles located at the droplet surfaces.
This results in particle networks between the emulsion droplets ( Fig. 3.5(d)).  Cryo-SEM images of the emulsion droplets are shown in Fig. 3.7. At low particle concentrations, we do not see particles at the oil water interface (Fig. 3.7(a)).
However, we see silica at the oil-water interfaces as the particle concentration is increased (Fig. 3.7(b)). For emulsions stabilized with Triton X-100, no particles are observed at the oil-water interfaces at low particle concentration (Fig 3.7(c)). As the particle concentration increases, more particles stabilize the emulsion (Fig 3.7(d)).
Depletion of surfactant in the bulk and a change in particle wettability due to surfactant adsorption enhances particle adsorption at the dodecane-water interface.  As the concentration of the particles increases, more particles locate at the interface ( Fig. 3.8(c)). The attractive interactions between the silica particles in the bulk and particles around the emulsion droplets lead to the formation of three-dimensional networks between the emulsion droplets at high particle concentrations ( Fig. 3.8(d)).
All the excess particles in the continuous phase get incorporated in these networks resulting in an increased viscosity and thickness of the emulsion phase. 29 There are four different end states for emulsion droplets depending on surfactantparticle interactions, particle concentration and mixing conditions (Figure 3.9). Weak interactions (hydrophobic and hydrogen bonding) between surfactants and colloidal particles lead to the formation of surfactant-stabilized emulsions (Fig. 3.9(a)).
Attractive interactions between the particles and surfactants lead to the coalescence of emulsion droplets or formation of particle coated droplets depending on the surfactantto-particle concentration when there is no mixing in the system (Fig. 3.9(b)).
Vortexing results in formation of particle-stabilized emulsions or emulsions stabilized with both surfactants and particles depending on the relative concentration of the particles and surfactant molecules (Figs. 3.9(c), (d)).

Conclusions:
We studied the effect of the addition of negatively charged fumed silica particles on surfactant-stabilized emulsions. Depending on surfactant-particle interactions and mixing conditions, we observed four different end states for the emulsion. Addition of these fumed silica particles had no effect on the emulsion stability when the interactions between surfactants and particles were repulsive. We observed a phase

Acknowledgements:
This work was supported by the Gulf of Mexico Research Initiative and by the RI Consortium for Nanoscience and Nanotechnology. We thank Evonik Corporation for providing fumed silica particles.

Abstract:
Here, we studied the effect of particle shape and inter-particle interaction on the formation and stability of bromohexadecane-in-water emulsions stabilized with spherical and fumed silica particles with similar hydrodynamic diameter. Emulsions were prepared at two different NaCl concentrations 0.1mM and 50mM. We found that the particle shape and inter-

Introduction:
The ability of the colloidal particles to get strongly adsorb at oil-water interfaces makes them potential alternatives to surfactants for stabilizing emulsions. 1 Unlike surfactants, the adsorption of colloidal particles onto the oil-water interfaces is not spontaneous. 2, 3 However, once a partially wettable particle is placed at the oilwater interface, it gets kinetically trapped and thermal fluctuations will be insufficient to displace it from the oil-water interface. With sufficient coverage at the interface, these particles act as electrostatic, steric or mechanical barriers against droplet coalescence. 4,5 Particle shape and inter-particle interactions also play significant role in determining the micro-structure and stability of particle stabilized emulsions. 6 Madivala et. al. 7 showed that ellipsoidal polystyrene particles above a critical aspect ratio are capable of forming stable emulsion, even when spherical and lower aspect ratio particles with the same wetting properties do not produce an emulsion. San-Miguel et. al. 8 showed that roughness on the colloidal particle surface enhance the stability of the emulsions as long as there is a homogeneous wetting of particle surface by oil and water phases. It is argued that anisotropic and rough particles gets pinned at the oil-water interfaces and lead to a significant deformation of the oil-water interface when compared with smooth spherical particles. 6 This results in strong adsorption of these particles at oil-water interfaces when compared with smooth spherical particles of same size 9 which leads to an increase in emulsion stability. Interfacial coverage of the particles on the emulsion droplet interface also dependon particle shape. 7,10,11 Particle shape and inter-particle interactions also influence the microstructure and the viscosity of the particle suspension, which in turn affect the structure and properties of the emulsions. Silanized fumed silica particle can form volume filling networks at concentrations much lower than the spherical silica particles of same hydrodynamic size when the interactions between the particles are attractive. 12 These networks will have a huge influence on the creaming behavior and stability of the emulsion droplets. 13 Here we formed Pickering emulsions using colloidal particles with repulsive and attractive interactions and systematically compared the effect of particle shape on the formation and microstructure of the emulsions. The combination of optical, cryogenic scanning electron microscopy (Cryo-SEM), and rheological measurements were used to determine the microstructure and stability of the emulsions.

Materials and methods:
Mono-dispersed spherical silica particles (210±10 nm) were purchased from Fiber Optics Inc. Fumed silica particles (Aerosil 816) were provided by Evonik Corporation, which were fractal in nature with a primary particle size of ~12nm. They NaCl and attractive at 50mM NaCl concentration. The zeta potential of the spherical and fumed silica particles were -48.2mV and -45.6mV respectively at 0.1mM NaCl.
At 50mM NaCl, there is a rapid flocculation of spherical silica particles and a rise in viscosity for fumed silica suspensions, suggesting attractive interactions between particles. Bromohexadecane in water (1:1) emulsions were prepared by vortex mixing the oil and water phase at 3000 rpm for 2min. 2 wt% silica dispersions were used to prepare the emulsions. The volume fraction of the oil phase separated due to coalescence was less than 0.5% in all the cases. The emulsions were analyzed with bright field optical microscopy and the images were processed with Image-J to obtain average droplet size and size distributions. Cryogenic Scanning Electron Microscopy (Gatan Alto 2500 cryo-system attached to a Zeiss Sigma field emission scanning electron microscope) is used to look at the fine structure around the emulsion droplets.
An AR2000ex rheometer with concentric double wall cylindrical geometry is used for doing rheological measurements. The samples were pre-sheared at 1s -1 for 30seconds to remove any shearing history before doing the measurements.

Results and discussion:
Where, D and d p are the diameter of the droplet and particle, respectively, ϕ is the fractional coverage of the particles at the droplet interface, and ρ p , ρ w , and ρ o are the densities of the particles, water, and oil respectively. For spherical silica particles, the calculated diameter of the neutrally buoyant droplet is ~576µm. The average diameter of the emulsion droplets formed after vortex mixing was less than the calculated diameter of the neutrally buoyant droplet. Therefore, the effective density of the emulsions droplets will be higher than the continuous phase, 14   Fumed silica particles forms networks of particles at higher salt concentration due to attractive inter-particle interactions. The emulsion droplets gets trapped in between these networks resulting in a gel like structure, which might have resulted in no observable creaming of the emulsion phase. Cryo-SEM images were used to determine the structure of the emulsion droplets. Figure 2  .2(c)). As we know fumed silica particles were fractal in nature, so the coverage of the particles on the droplet surfaces will be different when compared to smooth spherical silica particles. 17 At higher salt concentrations, we observed the networks of fumed silica particles in the bulk and an increase in thickness of particle layers around the droplet surface (figure 4.2(d)), suggesting the aggregation of fumed silica particles on the droplet surface and in the bulk due to attractive inter-particle interactions.  3(a)). As mentioned before interactions between particle dictates the interactions between emulsion droplets. Therefore, emulsions stabilized at low salt concentrations will have repulsive interactions between the droplets, whereas emulsions stabilized at high salt concentrations will have attractive interactions between the emulsion droplets. Repulsive emulsions behave as disordered elastic solids 18 and show shear thinning behavior. We observed similar behavior for emulsions stabilized with spherical silica particles at 0.1mm NaCl and there is a monotonic decrease in viscosity with increased shear rate (figure 4.3(a)). At higher salt concentrations, there was a discontinuity in the flow curve for emulsions stabilized with spherical silica particles at a shear rate of 1s -1 ( figure 4.3(a)). This is due to the progressive breakdown of aggregated emulsion droplets which results in a low viscosity continuous phase. 19 Fumed silica suspensions showed Newtonian behavior at low salt concentration and shear thinning behavior at high salt concentration. There is a monotonic decrease in viscosity with an increase in shear rate in case of emulsions stabilized with fumed silica particles at both the salt concentrations. The formation of gel like structure at high salt concentration results in more ordered structure and increased the viscosity of the particle suspensions and emulsions. This is due to the fact that the effective volume faction occupied by the fumed silica particles in the suspension is more when compared to spherical silica particles with similar hydrodynamic size. The effective volume fraction occupied by fractal particles can be estimated using following equation,

Rheology measurements:
Where, ø eff and ø o are the effective and actual volume fraction of the silica particles in the suspension. R is the diameter of the fractal particle and R o is radius of the primary particle. D f is the mass fractal dimension of the particles. D f for fumed silica particles is 2.17. 20 The effective volume fraction occupied by the fumed silica particles is ~11 times higher than that of the spherical silica particles with similar size, which results in higher viscosity and the attractive interactions between fumed silica particles results in the formation of the particle networks in the suspension resulting in shear thinning behavior at 50mM NaCl concentration.
Oscillatory strain experiments were performed at an oscillatory frequency of  This behavior suggests that fumed silica stabilized emulsions have much resistance to deformation when compared with emulsions stabilized with spherical particles. This is due to the fact that the fumed silica particles are fractal in nature and the edges of these particles gets pinned to the oil-water interface which resulted in strong adsorption at the interface when compared with spherical silica particles. 9 At 50mM NaCl, emulsion stabilized with spherical silica particles have a small region of LVR up to strain amplitudes γ<0.2%. The emulsion started to yield above 0.2% strain with a crossover at 50% strain amplitude (figure 4.4(a)). Flocculation of emulsion droplets gives some structure to the emulsions stabilized with spherical silica particles at 50mM NaCl. Therefore, the emulsion retains solid like character until the flocks get broken which results in a crossover at higher strains. For fumed silica stabilized emulsions, there is a significant increase in elastic and shear module and the LVR goes up to strain amplitudes γ<1% at higher salt concentration ( figure 4.4(b)). The gel like structure formed due to the formation of three dimensional particle networks between the emulsion droplets results in such behavior.

Conclusions:
We studied the effect of particle shape and inter-particle interactions on the microstructure and rheology of Pickering emulsions using spherical and fumed silica particles as emulsifiers. Attractive and repulsive emulsions were prepared by controlling the interactions between the silica particles in the bulk. The shape of the particles and inter-particle interactions strongly affect the creaming and rheological properties of the emulsions. We observed sedimentation and creaming for emulsion droplets stabilized with spherical and fumed silica particles respectively at 0.1mM NaCl. At 50mM NaCl, we observed flocculation in spherical silica stabilized emulsions, whereas emulsions stabilized with fumed silica particles formed a gel like structure. All the emulsions showed shear thinning behavior. The emulsions stabilized with fumed silica particles yielded at higher strains when compared with emulsions stabilized with spherical silica particles. The degree of shear thinning and yielding has increased with an increase in salt concentration.

Acknowledgement: SOME CAVEATS:
Droplet size distributions were used to support some of the observations that were presented in this dissertation. However, the reproducibility of these distributions depends on many parameters. The mixing conditions, type of mixer and vial used for emulsion formation and the extent of mixing will influence the final distribution of the emulsion droplets. Therefore, the reproducibility of these measurements will be very sensitive to the conditions used during the emulsion formation. However, most of these observations are qualitative and are reproducible phenomenon.