Voltage-Driven Ion Flux Promotes Emulsification at the Water|Oil Interface

Abstract

Emulsions are critical across a vast range of industries. Generally, emulsion synthesis is a complicated chemical process, requiring many mixed-phase systems. Here, we demonstrate that the flux of ions across the oil|water interface induces emulsification. Ion flux is achieved by a voltage-driven process, where an anode and a cathode are placed in each phase. When a current density of 2 mA/cm² is reached across the interface, emulsification occurs. We demonstrate that emulsification can be tuned to occur in both phases, depending on the ions present. Droplet sizes are on the order of hundreds of nm and are stable for over an hour even without purposefully added surfactant. We demonstrate qualitative control over droplet size and charge based on salt content, current densities, and polarity of the interface. The process is 1,000 times less energetic than ultrasonication. Our results introduce a robust and low-energy means of nanodroplet dispersion without the use of more than two phases and complex phase-transfer agents.

Emulsions are used in the production and manufacturing of foods, pharmaceuticals, and cosmetics.^1–5 For example, common items such as butter, face creams, and gasoline require the use of emulsions at some stage of their use or development.^{5, 6} Emulsions consist of the mixture of two immiscible phases (e.g., oil and water), where the droplet phase is termed the dispersed phase and the other phase the continuous phase. While some emulsions have been shown to form naturally or spontaneously, most emulsions used today depend on the use of mechanical, chemical, or thermodynamic means of emulsification.^7–9 The use of different forms of emulsification provides the benefit of allowing the user to produce a mixture that best fits their desired needs, whether it be industrial applications or food preparation, though these are not mutually exclusive.¹⁰ Mechanical means of emulsification, for example, can be seen in industrial and research settings and can include the use of highspeed mixing, sonication, microfluidics, membranes, and high pressures to achieve homogeneous emulsification.^{5, 11} Other non-mechanical emulsification methods rely on the use of thermodynamic qualities to achieve emulsification, for example taking advantage of temperature, phase inversions, and condensation.^10–13 Each of these emulsification techniques are highly energy intensive.

Several groups have shown that emulsification at the liquid|liquid interface occurs under various experimental conditions likely due to differences in chemical potential across the phase boundary.^{9, 14–17} Recently, our group demonstrated that the reason for emulsification at the water|oil interface was due to solute flux through the interface.¹⁸ We achieved this by using phase-transfer agents that would promote ions crossing the liquid|liquid interface. A validation of this work, and the topic of the current report, is to demonstrate that controlled flux via electrophoresis at the water|oil interface similarly induces emulsification. We term this process electro-emulsification. This technique differs from previous reports (e.g., Watanabe, Takhistov)^19–21 using electrocapillary emulsification in that it is more general and does not require a capillary, surfactants, or complex phase-transfer agents.

Results and Discussion:

The transfer of ions across liquid|liquid interfaces has previously been linked by our group to the spontaneous formation of droplets between two immiscible solutions.¹⁸ In these studies, the interfacial flux of ions across said interface was shown to promote droplet formation when in the presence of a phase-transfer agent. The observed behavior was attributed to an increase in surface tortuosity at the boundary due to the flux of ions across the interface, as well as the presence of a newly formed antagonistic salt that could stabilize said curved interfaces.^{22, 23} In fact, the efficacy of the antagonistic salt dictates in what phase droplets are stable and can form emulsions. In this prior article, the high degree of flux was attributed to differences in chemical potential across a water|oil interface that caused the marked partitioning of a chloroaurate anion (AuCl₄⁻) when in the presence of a tetrabutylammonium salt.²⁴ However, it stands to reason that if differences in chemical potential across boundaries can cause spontaneous emulsification, so, too, could electric potential differences. Thus, we hypothesize that applying a potential difference across an interface high enough to drive faradaic reactions on each electrode would drive ion flux through the interface, creating an emulsion. Here, we demonstrate that ionic flux across an oil|water interface induces emulsification. Figure 1 shows a schematic representation of the experiment. As with gel electrophoresis, ion movement can be convoluted by bubble formation. To rule out any influence of bubble formation, the cell was designed such that bubbles would be released upward in each phase. The oxidation and reduction of water occur in the water phase; however, the reactions occurring in the oil phase are less well-characterized (although bubble formation is always observed).

Figure 1:

Experimental setup for the current induced emulsification of a two-phase system. A cathode and anode are connected to platinum (r = 1 mm) wires submerged in the aqueous (top) and organic (bottom) phases respectively, as shown in panel A. Subsequent application of a current will drive reactions in both phases promoting the flux of species across the Liquid|Liquid interface to maintain electroneutrality, shown in panel B. This flux of species across the interface will generate an emulsion over time, as shown in panel C.

As shown in Figure 1, an anode and a cathode are placed in each phase, a voltage is applied, and a current is measured. As we apply a constant current and drive reactions in each phase, ions will transfer across the water|oil interface to maintain electroneutrality. Over time, and at sufficiently high currents, the transfer of species across the interface will lead to fluxification (flux-induced emulsification), and the formation of an emulsion will be observed at or near the liquid|liquid boundary. The voltage that is generally measured during these experiments is on the order of 30 V. Thus, the total power of the emulsification method (P=iV=.002 A * 30 V) is 0.09 W. Through optimization, we have found that for our system emulsification only occurs when a 3 mA threshold is reached. In the cell used, the cross-sectional area is ~1.5 cm² (neglecting the curvature of the phase). As stated above, we previously implicated flux in the emulsification of the water|oil interface. This new experiment, where flux can be tuned by simply tuning voltage, provides further validation that solute flux through an interface is responsible for emulsification. Recently, Girault and co-workers showed stochastic electrochemistry evidence of ionosomes at the liquid|liquid interface.²⁵ Our results are complementary in that solute flux can bring in enough solvent molecules into a new phase to comprise an ionosome.

Figure 2 demonstrates the visually observed emulsion formation after using the system previously described in Figure 1. Here, an aqueous top layer containing 0.1 M [NBu₄][Cl] was pipetted over an organic 1,2 -Dichloroethane (DCE) layer containing 0.1 M [NBu₄][PF₆] in a glass cuvette. A 5 mA current was then applied over the course of 5 minutes, causing emulsions to spontaneously form, as can be seen by comparing the before and after images. The aqueous and organic layers can both be observed to decrease in their clarity, evidenced by an increase in light scattering, when shining a green (532 +/− 10 nm) laser through each respective phase. An observed difference in emulsion formation is observed between both layers, which can be attributed to the difference in the dissolution of each phase within each other; this is to say that the DCE will be able to dissolve water at a much larger extent than water is able to dissolve DCE.²⁶ No spontaneous emulsification was observed 5 minutes before experimentation, nor was emulsification observed at any point when the solutions were not exposed to anodic/cathodic currents, as shown in Figure S1. Thus, the application of a rather large current is necessary to drive enough ionic flux to bring in solute molecules and nucleate droplets of a new phase.

Figure 2:

Glass cell images of DCE droplets formed in the aqueous phase before and after a 5 mA current is applied on a two-phase system over the course of 5 min. A 3 mL aqueous top layer containing 0.1 M [NBu₄][Cl] was pipetted over 3 mL of DCE containing 0.1 M [NBu₄][PF₆], a Pt wire (r = 1 mm) was submerged in each phase to drive the beforementioned current, with the cathode and anode corresponding to the aqueous and the organic phase, respectively.

We were also interested in how stable the droplets are that form from this method. Figure 3 shows the Dynamic Light Scattering (DLS) measurements for DCE droplets for the system described and shown in Figure 2. An aqueous top layer containing 0.1 M [NBu₄][Cl] was pipetted over 3 mL of DCE containing 0.1 M [NBu₄][PF₆] and emulsification was performed, as previously described. After 5 minutes of applying a constant current of 5 mA, the aqueous top layer was transferred to a cuvette with a Dip-Cell for DLS measurements and zeta potential acquisition, as shown in Figure 3A. Figure 3A shows DLS measurements taken ~10 min after emulsification, where the mean peak size can be observed for the droplets in solution, with an average size of 660 nm in radius. The zeta potential for these droplets can also be observed, with an average zeta potential of −32.7 mV; this suggests that a sufficient level of repulsion is occurring between droplets preventing coalescence and allowing for good overall stability, as can be seen in Figure 3B. Figure 3B shows the measurement of the peak mean size of droplets for 4 different time points over the course of an hour. No trend was observed in droplet size over the course of the measurements and the size distributions remained fairly constant.

Figure 3:

Dynamic light scattering measurements of DCE droplets formed in the aqueous phase when a 5 mA current is applied on a two-phase system over the course of 5 min. A 3 mL aqueous top layer containing 0.1M [NBu₄][Cl] was pipetted over 3 mL of DCE containing 0.1 M [NBu₄][PF₆], a Pt wire (r = 1 mm) was submerged in each phase to drive the beforementioned current, with the cathode and anode corresponding to the aqueous and the organic phase, respectively. Measurements for the zeta potential and mean peak size can be seen in (A), while timed stability measurements are shown in (B). DLS measurements were taken 3 times, for 60 runs at 1.64s/run, with a 120s equilibration time, and an equilibration temperature of 25 °C, as per instrument suggestion. A 1 cm glass cuvette was used to hold solutions within the instrument.

To show versatility and reproducibility, several different experimental conditions were tested. These included variations in the composition of each of the two phases and which phase the cathodic and anodic connection was placed in. Table 1 demonstrates the results for multiple variations of water in oil emulsions. Both the average diameter of droplets and the zeta potential are reported. For example, we observe an average droplet diameter of 1209 nm when we have the cathode placed in the aqueous layer containing 0.1 M [NBu₄][Cl] and the anode in the organic DCE layer containing 0.1 M [NBu₄][PF₆]. By comparison, when we switch the phase of the cathode and anode, a notable decrease in the average droplet size was observed, with an averaged diameter of 568.3 nm. Additionally, with the switched positioning of the cathode and anode, a more negative zeta potential was observed, with the prior case at −70.96 mV, vs. the latter case at −32.7 mV. We expect a negative zeta-potential under conditions of no surfactant because of hydrogen bonding networks at the interface.²⁷ We conclude from these results that the opposite polarity changes the composition of the droplets sufficiently to promote stability.

Table 1:

Observed Emulsions Under Different Experimental Conditions.^‡

Aqueous Phase†	Organic Phase†	Average Size (d. nm) N=3*	Zeta Potential (mV)
[Nbu4][Cl]	[Nbu4][PF6]	1209	−70.96
[Nbu4][Cl]	[Nbu4][PF6]	568.3	−32.7
None	[Nbu4][PF6]	954.1	−1.446
None	[Nbu4][PF6]	2211	−41.52
1M NaOH	[Nbu4][PF6]	1901	−155.3
1M NaOH	[Nbu4][PF6]	1039	−17.9
1M H2SO4	[Nbu4][PF6]	2990	−34.76
1M H2SO4	[Nbu4][PF6]	1635	−40.14
None	[Nbu4][TPB]	125.9	−57.08
None	[Nbu4][Br]	1635	70.76
None	[NHx4][ClO4]	3477	−6.031

^‡Different variations of voltage-induced emulsification were performed to test for reproducibility and generalizability.

^†Variation in the positioning of the cathode (Red) and anode (Black) was performed to test variation emulsion formation and can be seen within the table based on the color of each species, with red corresponding to the cathode and black to the anode.

^*See Supplementary Information Figure S2 for size distributions.

In the absence of an aqueous electrolyte, emulsification was still observed, likely due to the auto-ionization of water. However, lower zeta potentials and larger droplet diameters were observed, indicating that the overall stability of the droplets may be affected by the ions transferring. Such results indicate that the ions that are transferring are acting as stabilizing agents for the observed droplets. When testing emulsification under acidic or basic conditions, no notable time difference was observed to achieve emulsification. Differences were observed in the values obtained for the zeta potentials of the observed droplets. When the aqueous layer (cathodic) contained 1 M NaOH and the DCE (anodic) layer contained 0.1 M [NBu₄][PF₆] a zeta potential of −155 mV was observed. For comparison, a similar system containing 1 M H₂SO₄ had a zeta potential of about −40 mV. Finally, different organic salts were compared to elucidate the effects of anion hydrophobicity on observed emulsion behavior. When we compare the results obtained for tetrabutylammonium tetraphenylborate ([NBu₄][TPB]), tetrabutylammonium hexafluorophosphate ([NBu₄][PF₆]), and tetrabutylammonium bromide ([NBu₄][Br]), we observe a trend in the average droplet size and the observed zeta potential. Average size and zeta potential both appear to increase with increased hydrophilicity, indicating a clear correlation between the flux of the anions across the interface and observed emulsion behavior.^{28, 29} The behavior for the emulsification of droplets in different organic solvents was tested and showed good reproducibility, as shown in Figure S3. Our results are a notable departure from current industrial methods, where the energy requirements for emulsification are on the order of hundreds of watts. Our electro-emulsification method stands in stark contrast, requiring only ~0.1 watts, representing a technique that is a thousand times more energy efficient.

Finally, our results add another form of validation to our previously published model.¹⁸ In that model, we suggested that ions are driven across the water|oil interface. The ions also carry solvent molecules with them, as predicted by Benjamin.²² We related the perturbation to the water|oil interface to a meandering river that forms oxbow lakes over time.³⁰ If solute flux across the interface is sufficiently fast, droplets of the new phase can nucleate and grow into a coherent dispersed phase. This model assumes that droplets are forming actively in both phases at all given points in time. However, key to the model is a kinetic suggestion that droplets must be sufficiently stable to allow for phase dispersion. If droplets are unstable, an emulsion will not form.

Conclusion:

We have presented a phenomenon, voltage-induced emulsification, for the synthesis of emulsions. This new synthetic tool takes advantage of solute flux through a liquid|liquid interface. We demonstrated that one can make oil emulsions in water and water emulsions in oil by tuning the ions in the respective phases, the voltages and currents used to drive ion transfer, and the polarity of the voltage application. Emulsions were synthesized without the purposeful addition of surfactant molecules. The voltage-driven emulsification strategy is 10³ times less energy-intensive than ultrasonication and offers a practical pathway to industrially scalable emulsion synthesis without the need of complex phase-transfer agents.

Materials and Methods:

All Solutions were prepared fresh prior to experimentation. Aqueous solutions were made with Millipore GenPure ultra-pure water (18.20 MΩ*cm⁻¹), and organic solutions were prepared with different solvents as needed including DCE, Chloroform, and methylene chloride. Organic salts used during this study include tetrabutylammonium perchlorate, tetrabutylammonium hexafluorophosphate, tetrabutylammonium bromide, tetrahexylammonium perchlorate, and tetrabutylammonium tetraphenylborate. Aqueous salts included tetrabutylammonium chloride. All solvents and salts were obtained from Sigma-Aldrich and were of analytical grade. Acid solutions were prepared based on concentrated sulfuric acid stock solutions, while base solutions were prepared just prior to experimentation using sodium hydroxide, both were obtained from Fisher Scientific. All reagents used here were of analytical grade and required no further purification.

Electro-Emulsification Studies:

All experiments were performed using bi-phasic systems consisting of immiscible solutions (i.e., 1,2-Dichloroethane|Water). Before experimentation a platinum wire (r =1 mm) was sealed within a cell while maintaining an electrical connection, this allowed for external connection without perturbation of the liquid|liquid interface. An organic bottom layer containing the desired organic salt (i.e. tetrabutylammonium perchlorate, tetrabutylammonium hexafluorophosphate, tetrabutylammonium bromide, tetrahexylammonium perchlorate, and tetrabutylammonium tetraphenylborate) was then pipetted into the cell followed by an aqueous top layer, this order was done to limit mixing before experimentation. A secondary platinum wire (r =1 mm) was then submerged into the aqueous top layer to complete the circuit and allow for current to flow. Upon submersion, a current was then applied across the system using a GW Instek GPD-2303S DC power supply. Currents were maintained at 5mA for a total duration of 5 min unless specified. Solutions were allowed to rest for an additional 5 minutes before additional experimentation or measurement.

Dynamic Light Scattering (DLS) and Zeta Potential Measurements:

Measurements for droplet sizes were performed using a Malvern Zetasizer Pro instrument from Malvern Panalytical, Worcester, UK. In these measurements, the resulting top layer and bottom layer from the prior section were allowed to rest for 5 minutes and were then transferred to a 1 cm glass cuvette. After transfer, the cuvette was then inserted into the DLS instrumentation and was allowed to perform dynamic light scatter measurements to determine the size and intensity of the droplets formed during electro-emulsification. DLS measurements were taken 3 times, unless specified, for 60 runs at 1.64s/run, with a 120s equilibration time, and an equilibration temperature of 25 °C, as per instrument suggestion. Zeta potential measurements were performed intermittently between each DLS measurement using a ZEN1002 dip cell kit from Malvern Panalytical, Worcester, UK.