- Email: [email protected]
Formation and characteristion of surfactantless oil-in-water and water-in-oil emulsions
Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
Formation and characteristion of surfactantless oil-in-water and water-in-oil emulsions Katherine M. Knight, Daren J. Caruana ∗ Department of Chemistry, University College London, 20 Gordon St., London WC1H 0AJ, United Kingdom Received 12 July 2006; received in revised form 8 December 2006; accepted 28 December 2006 Available online 11 January 2007
Abstract A liquid–liquid interface in the form of a surfactantless oil-in-water or water-in-oil emulsion has been created and characterised. The emulsions, composed of 1,2-dichloroethane (DCE) and D2 O were created using a condensation method and this technique has been optimised to give the conditions needed to produce reproducible emulsions with comparable drop volumes and number densities. These have been measured using optical microscopy and light scattering techniques to characterise the emulsions and to investigate how the drop volume, and hence the solubilities of the two phases, are dependent on temperature. The destabilisation of the emulsion, due to the sedimentation of DCE (ρ = 1.235 g cm−3 ) or creaming of D2 O (ρ = 1.105 g cm−3 ), and the coalescence of the drops, has also been measured qualitatively using turbidity measurements. © 2007 Elsevier B.V. All rights reserved. Keywords: Liquid–liquid interface; Emulsions; Condensation; Surfactant-free
1. Introduction Mixtures of two or more liquids that are not miscible in each other are inherently unstable and will, after some time, separate into two phases. This phase separation can be reduced by the use of stabilisers, such as surfactants or polymers, to produce a dispersed emulsion system, and in some cases may produce thermodynamically stable microemulsions. There are many examples of oil-in-water emulsions stabilised by using ionic or non-ionic surfactants [1–3] and the basic principles of destabilisation processes have been developed on the basis of investigation of these systems. Emulsions formed in this way are technologically very important in the food, chemical, cosmetic and nanotechnology industries. For example, enzyme catalysed reactions of substrates that are oil soluble, using enzyme-modified oil-in-water emulsions, are of interest to the pharmaceutical industry [4]. However, with increasing pressure of strict environmental legislation controlling the use of chemical agents, the use of amphiphiles is becoming undesirable in industrial processes. As a consequence, the development of alternative methods or agents to stabilise emulsions, that can
∗
Corresponding author. Tel.: +44 207679 4527; fax: +44 207679 7463. E-mail address: [email protected] (D.J. Caruana).
0927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2006.12.081
replace traditional surfactants and are cheaper and environmentally benign, is becoming a priority in emulsion research. For this reason, surfactantless strategies for stabilising oil-in-water emulsions are receiving a great deal of attention owing to the technological benefits and scientific importance. Abe and co-workers have described a method of producing sub micron sized oil drops in water by sonication of binary mixtures of immiscible species, such as benzene/n-hexadecane in water [5] or oleic acids and various esters in water [6]. The emulsions produced appear to be stable for longer than days, however the precise physical basis of the stabilisation is unclear. Pashley [7] reported a new method of stabilising an oil-in-water emulsion by degassing a solution containing dodecane or squalane derivatives in water, and producing an emulsion by mechanical agitation of the solution. It was proposed that by removing dissolved gases from the solution, the hydrophobic interaction between the oil drop was reduced and so stabilises the emulsion [8]. Unlike the liquid–solid and liquid–air interfaces, studies of the structure of the interface between two immiscible liquids presents a more challenging experimental and technical problem [9,10]. Most experimental approaches used to study either the structure of the interface or the dynamics, rely on the formation of a continuous interface between the immiscible liquids, held apart by their different densities. The technique described in
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
353
Fig. 1. Schematic diagram of a DCE drop in water (A) and a water drop in DCE (B) showing the dynamics of solvent exchange, schemes 1 and 2.
this paper, provides a method of producing a stable ‘clean’ interface which can be investigated using light, or neutron scattering techniques [11,12]. We report a procedure for creating surfactantless emulsions of DCE-in-D2 O or D2 O-in-DCE. These are created using a condensation technique, [13] essentially controlling the solubility, of the two partially immiscible phases in each other (equilibriums 1 and 2 in Fig. 1A and B), by manipulating the temperature. The stability and characterisation of the emulsions has been investigated and monitored using turbidity, optical microscopy and light scattering measurements, and the drop volumes and number densities have been shown to be reproducible and comparable between the two emulsions. 2. Experimental section 2.1. Materials Any H2 O, used in the experiments and for cleaning was from a Milli-Q Gradient 18.2 M cm (at 25 ◦ C) water system. Glassware was thoroughly cleaned by boil washing in a 2 l beaker filled with dilute sulphuric acid (0.1 M, BDH; Analar grade with a minimum assay of 98.0%) for 2 h to remove contamination from detergents. The glassware was then thoroughly rinsed using deionised water several times and dried in an oven at 65 ◦ C for 12 h. The emulsion was composed of 1,2dichloroethane (DCE) (Sigma; >99% spectrophotometric grade) and D2 O (Aldrich; 99.9% atom D). All chemicals were used as received. 2.2. Procedure for creating the emulsions 10 cm3 of the chosen continuous phase and 1 cm3 of the emulsion sub-phase were mixed in a 25 cm3 stoppered glass conical flask. This was then heated to 65 ◦ C, with gentle stirring, in a 500 ml glass beaker water-bath on a hot-plate for 1 h. The flask was then put into a water-bath (Grant Ltd., 6; Grant Instruments, UK) set at 15 ◦ C and left to cool for 30 min. The emulsion formed immediately as the solution was cooled. These were found to be the most effective methods of heating and cooling due to efficient heat exchange. After the solution had cooled for 30 min, a sample of the emulsion was taken for analysis using a variable volume pipette (Volac 200–1000 l) with a modified tip. The pipette tip was cut, approximately 5 mm from the end, to increase the aperture surface area and so reduce the shear forces when pipetting the solution. The sample was removed with care to avoid withdraw-
ing large drops (1–2 mm) of suspended emulsion phase. These were always present in the bulk solution due to the excess volume of sub-phase used in the formation of the emulsion. The emulsion was then studied using either turbidity measurements with a UV–vis spectrophotometer to look at stability, or an optical microscope to measure the average volume and number density of the drops. 2.3. Stability measurements As a measure of stability with time, the turbidity of the emulsion was monitored using a UV–vis spectrophotometer (Agilent 8453 UV–vis Spectroscopy System; Agilent Technologies, Germany) with a thermostatted cell holder set at 15 ◦ C. The absorbance spectra between 190 and 1100 nm were recorded every 10 s, and even though the processes of sedimentation and coagulation could not be quantified separately, this gave a qualitative measure of the turbidity of the sample. 10 m or 2 mm pathlength Far UV quartz cuvettes (Hellma UK Ltd.) were used in the spectrophotometer and were filled to the top to prevent any DCE evaporating out of the emulsion into the headspace at the top of the cuvette. 2.4. Optical microscope experiments The emulsion drop volume and number density were determined using Optical Microscopy (Jenalab; Carl Zeiss, Germany), with transmission illumination, connected to a video camera (TK-1085E; JVC, Japan) with PC data collection (Presto Video Works Ver 4.1 Rev 6; NewSoft Technology Corp., Taiwan). The temperature was maintained and controlled with the use of a peltier device (Thermo Electric Cooler Type DT 1069; Marlow Industries Inc., USA), positioned on the stage of the microscope. This was connected to a variable current supply temperature selector (built in-house; R. Waymark) and the copper plate above the peltier device, allowed for efficient heat transfer between the peltier device and the microscope slide. A sample of the emulsion was put on the microscope slide on the temperature controller and the slide temperature was recorded using a thermocouple at the end of each experiment. This allowed for any changes in temperature due to extra heating from the microscope lamp. The thermocouple was also fixed to the top of the slide to compensate for any differences between the temperature selector and the emulsion sample, and for all sizing experiments the selector was set at 15 ± 0.2 ◦ C. Using the video camera and data capture software, images of the emulsion could be recorded and from these the average
354
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
drop volume was calculated. This was done using a calibrated size standard of 5.0 ± 0.05 m Polymer Microspheres in water (Duke Scientific Corporation, CA) and an image of these can be seen in Fig. 2A. Due to the nature of the emulsion, the diffraction pattern of individual drops changed according to the focal plane (Fig. 2D–I) and so, for accuracy, the sizing images were taken when the drops appeared with the same diffraction pattern (Fig. 2G). To calculate an average volume, three different images of each sample were recorded and analysed and a typical image of an emulsion can be seen in Fig. 2B. From these images, approximately 3–9 individual drops for each sample were in the same focal plane and therefore measurable. An error was then calculated from the standard deviation of the different drop measurements.
The number density of the emulsion was determined using a haemocytometer (non-metallised Improved Neubauer; Hawksley & Sons Ltd., UK) with a grid of 1/400 mm2 × 0.1 mm depth etched onto the surface. By placing a sample of the emulsion on the haemocytometer, the number of drops per square could be counted under the microscope and then scaled to give a value for drops per cm3 . To obtain an average of six values for the number density per cm3 , three images of each sample were taken (Fig. 2C) and the standard deviation was used as the error. 2.5. Light scattering experiments To determine the emulsion drop volume and size distribution, a Mastersizer 2000 was used at room temperature, 25 ◦ C. An
Fig. 2. Typical images from optical microscopy experiments measuring drop volume (B) and number density (using the haemocytometer) (C). An image of the microspheres used for size calibration is also included (A). All images were taken with the same magnification. (D–I) are images at different focal planes through an emulsion drop, and show some of the different diffraction patterns observed. The diffraction pattern in (G) was chosen as showing the sharpest interface and only drops seen with this pattern were measured.
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
355
emulsion was created, following the procedure described, composed of 100 cm3 H2 O and 1 cm3 DCE. Also included were, 0.1 M tetraethylammonium chloride (TEACl) in the water and 1 × 10−3 M tetraethylammonium tetraphenylborate (TEATPB) in the DCE phase to help stabilise the emulsion [14]. The emulsion was also added to the Mastersizer ‘hot’, at 65 ◦ C, rather than being cooled first. This was to make sure that the emulsion formed at the same temperature as the equipment and avoided any destabilisation caused when introducing the sample. A larger volume of the aqueous phase was also used so that the emulsion could be added directly to the Mastersizer without prior dilution. This was because the sizer employed a flow-through system which would mix the sample with an aqueous medium and dilute the sample for analysis. Stirring and dilution have been seen to destabilise the emulsion due to an increase in shear [14] and so, by initially using a larger volume of water to make the emulsion, this would eliminate the dilution stage and allow the propeller to be run at its lowest setting. For these experiments, this setting was 1500 rpm. 3. Results and discussion 3.1. Emulsion formation Emulsions can be created either by dispersion – breaking down the bulk material via shaking or stirring; or condensation of the discontinuous phase – building up the drops [15]. Dispersion methods generally form emulsions with an inhomogeneous and uncontrolled size distribution, with the discontinuous phase maintained with the use of surfactants. Condensation methods, however, allow for molecular clusters to form which grow via molecular diffusion to produce drops with a homogeneous size distribution. Emulsions can be made using this method by heating a mixture of oil and water, where the solubility of the dispersed phase, e.g. oil, in the continuous liquid, e.g. water, will increase, creating a saturated solution. This is then cooled rapidly, thereby decreasing the solubility and causing the oil to condense from the water as small homogeneous drops [16]. The solubility of DCE in a large excess of water, and water in a large excess of DCE, are shown in Fig. 3A and B and fitted with a three parameter exponential increase equation, y = −0.45 + 0.075e0.03T and y = 0.747 + 0.057e0.03T for DCE-in-water and water-in-DCE, respectively [16]. Using these equations, Table 1 shows the weight percent of dissolved DCE or water at different temperatures, and the calculated weight percent present as drops when the emulsion is heated to 65 ◦ C and cooled to temperatures in the range 55–5 ◦ C. The emulsion drops form initially via a process of nucleation where a cluster of molecules aggregate to make a critical nucleus of dimensions 1–10 nm, where the change in Gibbs freeenergy with respect to cluster size is zero [15]. The cluster size of this critical nucleus has been shown to be in the order of four molecules for the nucleation of water [17]. The extent of nucleation depends on the saturation of the oil in water; when the temperature of the system is decreased rapidly, this causes the concentration of the aqueous dissolved oil to suddenly go above the critical supersaturation point, causing
Fig. 3. Solubility plots of DCE-in-H2 O (A) determined between 23 and 73 ◦ C, and H2 O-in-DCE (B) determined between 20 and 70 ◦ C, from Ref. [16]. The solid lines in both graphs are regression fits using a three parameter exponential increase equation included in the plots.
the oil to spontaneously nucleate to reduce the concentration to below the critical point. Once this concentration is reached, no more nucleation will occur and instead the existing clusters will grow via molecular diffusion until the excess aqueous dissolved oil is depleted [15]. Because the nucleation is spontaneous and the nucleation rate is very fast, the resulting emulsion will initially have a monodisperse size distribution. After this time there would be some destabilisation from Ostwald ripening for an emulsion produced using this technique. In turbidity experiments using the UV–vis spectrophotometer, a sharp peak with a scattering >3.5 a.u. was seen at 195 nm when the emulsion was scanned from 190 to 1100 nm, shown in Fig. 4. This sharp peak occurred for all emulsion samples Table 1 Showing the solubility weight percent of DCE in an excess of water and water in an excess of DCE at different temperatures [16] and the calculated weight percent present as drops when the emulsion is heated to 65 ◦ C and cooled to temperatures in the range 55–5 ◦ C Temperature (◦ C)
65 55 45 35 25 15 5
Solubility (wt.%) DCE-in-H2 O
H2 O-in-DCE
1.25 1.11 1.01 0.94 0.88 0.85 0.82
0.46 0.32 0.23 0.16 0.11 0.08 0.06
DCE drops (wt.%)
H2 O drops (wt.%)
– 0.14 0.24 0.31 0.37 0.40 0.43
– 0.14 0.23 0.30 0.35 0.38 0.40
356
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
Fig. 4. Graph showing the turbidity vs. wavelength from 190 to 300 nm for an emulsion containing DCE-in-D2 O (including sucrose for stability), illustrating the sharp ‘scattering’ peak seen at 195 nm. The different plots correspond to measurements taken at different times; (—) initial scan, (- - -) after 10 min, and (· · ·) after 1 h.
that were placed into the spectrophotometer at 65 ◦ C and then allowed to cool to 15 ◦ C in the beam, while the turbidity was measured. The magnitude of this peak was seen to be at the saturation limit of the UV–vis absorbance over the first 10 min while the emulsion formed, and this then decreased slowly as the emulsion destabilised. This peak was also present in a sample of D2 O saturated with DCE at 15 ◦ C and is thought to correspond to the clusters of DCE molecules dissolved in the aqueous phase. The saturated maximum and then steady decrease in the scattering of this peak would then be due to the aqueous dissolved DCE molecules forming clusters and then growing to form emulsion drops, illustrating the link between the turbidity of an emulsion and its number density. The absorbance of this peak is still quite high after 1 h due to the presence of DCE still dissolved in the aqueous phase at 15 ◦ C. To determine the length of time needed to create a saturated solution of DCE-in-water at 65 ◦ C, emulsions were formed using the procedure described earlier but with different recorded solubilisation times. D2 O and sucrose were used in these experiments to reduce the rate of sedimentation and hence improve the stability of the emulsion, and this will be described fully in a later publication [14]. All emulsions were cooled to 15 ◦ C for 30 min. A sample from each emulsion was then taken and the average drop volume and number density were measured as outlined before. These were used to calculate the total weight percent of DCE present as condensed drops and, using Table 1 this should be a value of approximately 0.40 wt.% for a saturated heated emulsion sample. Graphs of the number density of the drops per cm3 and the weight percent of the DCE condensed as drops versus solubilisation time are shown in Fig. 5A and B, respectively. From Fig. 5A and B, it can be seen that both the number density and DCE weight percent reach a plateau after 1 h of solubilisation time. From Fig. 5B, it can also be seen that the weight percent of DCE condensed as drops is approximately 0.4 wt.% after 1 h. This supports the literature data seen in Table 1 and verifies that the heated emulsion sample is saturated after 1 h. The turbidity of the emulsion was also measured (Fig. 5C) for
Fig. 5. Number density (A), DCE weight percent (B) and turbidity measured at 550 nm (C) vs. solubilisation time for an emulsion of DCE-in-D2 O only. The different symbols used in (C) correspond to two different experimental data sets and each data point shown for all plots, was determined from a different emulsion created as described in the experimental section. Sucrose was included in these experiments to reduce the rate of sedimentation and improve stability, and will be discussed later.
emulsions heated for different times, and by plotting this with the emulsion number density from Fig. 5A, the fitted regression line in Fig. 6 shows a linear trend, illustrating the link between the turbidity and number density of the emulsion. A similar result would be expected for an emulsion composed of water-in-DCE, due to the data in Table 1. Even though the solubility of water in an excess of DCE is less than that of DCE in an excess of water, the calculated solubility weight percent condensed as emulsion drops is comparable between the two emulsion systems formed by heating to 65 ◦ C and cooling to temperatures in a range of 55–5 ◦ C. 3.2. Emulsion characterisation 3.2.1. Light scattering experiments Surfactant stabilised emulsions are conventionally characterised using light scattering techniques involving analysis using
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
357
Table 2 Comparing the measured drop volume, number density and calculated weight percent of emulsion drops for DCE-in-D2 O (stabilised with sucrose) and D2 O-in-DCE emulsions Emulsion system
Drop volume (×10−12 cm3 )
Number density (×106 drops cm−3 )
Weight percent emulsion drops
D2 O-in-DCE DCE-in-D2 O (+sucrose)
22 ± 3 (3.5 ± 0.1 m) 25 ± 5 (3.6 ± 0.2 m)
150 ± 15 140 ± 12
0.33 ± 0.05 0.35 ± 0.06
These were measured after 30 min cooling at 15 ◦ C and all other conditions were as described in the experimental.
Fig. 6. Graph plotting the emulsion turbidity vs. number density, for a DCEin-D2 O only emulsion, to illustrate the linear relationship between them. The equation for the fitted regression line is y = 24x − 58.
Mie Theory, using the dependence of the complex scattering from a large sphere, on the scattering angle [15]. To determine the emulsion drop volume and size distribution, a Mastersizer 2000 was used and, although this still accelerated the rate of destabilisation due to shear, this gave an interesting result for the drop volume, seen in Fig. 7. From Fig. 7, two peaks, at 3.6 and 12.3 m belonging to two different populations of drop diameters, can be seen. The peak at 3.6 m corresponds to the size of the emulsion drops, whereas the larger peak at 12.3 m was believed to be from larger coalesced drops formed by the stirring of the solution. Using the measured percentage volume for each diameter, a ratio of 6:1 is calculated for drops at 3.6:12.3 m, showing that the smaller drops are more prevalent in solution, even though the peak height for the larger drops is much greater than that for the small drops.
DCE-in-D2 O (including sucrose), and D2 O-in-DCE. These were heated to 65 ◦ C and cooled to 15 ◦ C, and samples were taken for analysis using optical microscopy as outlined. The drop volume and number density were measured for each emulsion and the results are shown in Table 2, as is the calculated weight percent of drops. From Table 2, it can be seen that there is no significant difference in the drop volume and number density for emulsions containing water-in-oil or oil-in-water formed using a condensation technique. The calculated weight percent of emulsion drops also corresponds well with the expected value determined from the literature in Table 1. Although the weight percent in these emulsions appears numerically small, the emulsions are very dense. If the average separation distance between each drop is calculated, using l = N−1/3 , where N is the number density of drops per m3 , a value of 15 m is determined. This is in the same order of magnitude as the drop diameter of 3.75 ± 0.25 m at 15 ◦ C. 3.2.3. Stability measurements To measure the stability of the oil-in-water emulsion, a mixture of DCE and D2 O (including 1.5 M sucrose for stability), was prepared as described, and a sample of the emulsion was put on a microscope slide on the Peltier device, set at 15 ◦ C. The initial drop volume and number density were measured whilst the flask containing the bulk emulsion solution was maintained at 15 ◦ C in the water bath for subsequent samples. Measurements of the initial drop volume were recorded as outlined, and then fresh samples from the bulk emulsion were taken every hour for 5 h as a measure of the emulsion stability. The change in drop volume with time are shown in Fig. 8, and it can be seen that the drop volume increased linearly.
3.2.2. Optical microscope experiments To accurately measure the emulsion drop size and number density emulsions were created, as described, containing
Fig. 7. Graph showing emulsion drop percentage volume vs. drop diameter (m) for an emulsion containing DCE-in-H2 O, measured at 25 ◦ C using a Mastersizer 2000.
Fig. 8. Change in drop volume with time for a DCE-in-D2 O (stabilised with sucrose) emulsion system.
358
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
The number density of the emulsion was also measured every hour in this experiment; however, due to the high vapour pressure of DCE, as every sample was removed, the DCE in the headspace of the vessel was also released. As a result the number density, and thus the volume fraction of DCE in the emulsion, was reduced falsely by a destabilising factor other than sedimentation or coalescence. This was also seen to be true for the drop volume, and so an experiment was performed where the drop volume and number density were measured initially and then again after 20 h. The flask was not reopened during this time and the results show that, for an emulsion of DCE-in-D2 O (including sucrose), heated to 65 ◦ C and cooled to 15 ◦ C, the initial drop volume was 25 ± 5 × 10−12 cm3 and after 20 h, this increased to 180 ± 4 × 10−12 cm3 . The initial number density was found to be 140 ± 12 × 106 drops cm−3 and after 20 h and this decreased to 7 ± 6 × 106 drops cm−3 (this error seems to be large due to extensive emulsion destabilisation, resulting in only 3 ± 3 drops being counted per square of the Haemocytometer). The value for the drop volume after 20 h is similar to that obtained after 7 h by extrapolating Fig. 8, and this suggests that by opening the flask and releasing the headspace every hour, the rate of destabilisation of the emulsion was accelerated. Fig. 8, still shows, qualitatively, how the emulsion destabilises over time and from calculations using the data for the drop volume measured initially and after 20 h, the volume is seen to double after 7 h for a DCE-in-D2 O (including sucrose) emulsion, instead of after 2 h as suggested by Fig. 8. Attempts were also made to measure the drop volume every hour, for a D2 O-in-DCE emulsion, to see qualitatively how the emulsion destabilised over time. Difficulties were encountered, however, due to the high vapour pressure of the DCE and these were similar to the problems experienced for the DCE-inD2 O emulsion. This was because the DCE would evaporate into the headspace of the flask and every time a measurement was taken, this would be released. Because the continuous phase of the emulsion was DCE, this caused the emulsion to decrease in total volume every time the flask was opened and caused rapid destabilisation of the emulsion. There were also problems encountered when the measurements were taken under the microscope, due to the DCE evaporation from the glass slide. To further investigate the stability of emulsions composed of DCE-in-water and water-in-DCE, emulsions were created as described using H2 O or D2 O in an excess of DCE, and DCE in an excess of D2 O, and samples were taken for analysis using turbidity measurements maintained at 15 ◦ C. The turbidity measurements shown in Fig. 9, have been normalised for easier qualitative comparison, due to the emulsions only being heated until they reached 65 ◦ C, approximately 20 min, rather than stirring them for 1 h at 65 ◦ C. This meant that the DCE was not fully saturated with the aqueous solution when the emulsion was cooled to 15 ◦ C and the initial turbidity values are 1.26, 1.41 and 0.99 au for DCE-in-D2 O, H2 O-in-DCE and D2 O-in-DCE, respectively. From Fig. 9, it can be seen that the water-in-oil systems were significantly more stable than an emulsion containing oil-in-water, with the change in the turbidity with time being increased from −172.6 × 10−6 a.u. s−1 for DCE-in-D2 O, to −30.2 × 10−6 a.u. s−1 for a D2 O-in-DCE system.
Fig. 9. Graph comparing normalised turbidity at 550 nm vs. time for emulsions containing (A) DCE-in-D2 O, (B) H2 O-in-DCE and (C) D2 O-in-DCE. The emulsions were formed at 15 ◦ C and this temperature was maintained for the turbidity measurements. The original turbidity values were 1.26, 1.41 and 0.99 a.u., and the rate of change of turbidity with time for the plots are −172.6, −138.3 and −30.2 × 10−6 a.u. s−1 for the DCE-in-D2 O, H2 O-in-DCE and D2 O-in-DCE emulsions, respectively.
The difference in stability between the H2 O-in-DCE and D2 O-in-DCE emulsions can be explained by considering the relative densities of the drop. The velocity v, of the sedimentation of a spherical DCE drop with radius r, can be obtained by equating the force due to gravity, g, and the viscous drag of the drop through the media. 4 3 πr (ρd − ρm )g = 6πηrv 3
(1)
rearranging gives, v=
2r2 (ρd − ρm )g 9η
(2)
where η is the viscosity of the DCE solution (0.84 mPa s), and ρd and ρm are the densities of the drop and medium, respectively. Therefore, the sedimentation velocities of D2 O and H2 O drops in DCE medium are −1.53 × 10−4 and −2.61 × 10−4 cm s−1 , respectively, a negative velocity indicates creaming. Hence the destabilisation of the H2 O is faster than that of D2 O drops in DCE. 3.2.4. The effect of temperature on the emulsions Due to the nature of the emulsion system and the way that it is formed, the drop volume is sensitive to temperature. Using the temperature controller, a sample of a DCE-in-D2 O (including sucrose for stability) emulsion, created as described, was cooled to 15 ◦ C for 30 min and then a sample was taken. The drop volume was measured at intervals as the emulsion was heated in recorded increments from 15 to 30 ◦ C, this was then converted to weight percent using the measured initial number density of 140 ± 12 × 106 drops cm−3 and was plotted in Fig. 10. A water-in-oil emulsion was then created composed of 10 cm3 DCE and 1 cm3 D2 O and a sample was taken for analysis using optical microscopy by placing on a slide on the microscope temperature controller, maintained at 15 ◦ C. Once the drop volume measurement had been taken the temperature, of the water bath
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
Fig. 10. Graph showing the weight percent of emulsion drops vs. the temperature of the system for emulsions containing: (䊉) DCE-in-D2 O (stabilised with sucrose) and () D2 O-in-DCE. The solid lines are regression fits using the three parameter exponential decay equations (䊉) y = 0.114 + 2.39e−0.162T and () y = 0.101 + 2.23e−0.146T .
and microscope temperature controller, was increased to 20 ◦ C and a second sample was taken for analysis. This was repeated in 5 ◦ C increments until 30 ◦ C (this was the highest temperature that could be measured accurately before the emulsion destabilised from the DCE evaporating out of the flask) and the results for the calculated drop weight percent, using the initial number density of 150 ± 15 × 106 drops cm−3 for the D2 O-inDCE emulsion, are also included in Fig. 10. This shows that the weight percent of the emulsion drops, and hence also drop volume, decreases with increasing temperature as expected, for both the oil-in-water and water-in-oil emulsions. The data were also regression fitted using a three parameter exponential decay equation of the form y = y0 + ae−bT . A different technique was used to heat the water-in-oil emulsion because of the volatile nature of the DCE and the length of time of the experiment. When the microscope-method was attempted, as used for the oil-in-water emulsion, the sample evaporated from the slide on the temperature controller after one measurement, approximately 2 min, making it impossible to accurately measure the drop volume for more than one temperature using this technique. The regression lines (using equation y = y0 + ae−bT ) from Fig. 10 may be compared using ratios of coefficients with those determined for the solubility of DCE in an excess of water, or water in an excess of DCE, in Fig. 3. The ratio of the coefficients a:b, are 2.56 and 1.67 for DCE-in-water and water-in-DCE, respectively, for the literature solubility data given in Fig. 3. For the emulsion data, the ratio a:b for DCE-in-water and waterin-DCE are 14.76 and 10.14, respectively, from Fig. 10. The numerical values are different because this is only an empirical comparison of parameters that have similar physical meaning under different experimental conditions, however comparing in a similar way the two ratios obtained for the water-in-DCE 0.173 (2.56/14.76) and DCE-in-water 0.165 (1.67/10.14) the values are almost equal. This indicates that the change in drop volume with temperature follows the same trend as the solubilities of DCE and water, as expected. This temperature effect described in Fig. 10 was also reversible and to illustrate this, an emulsion of DCE-in-D2 O was
359
Fig. 11. Turbidity at 550 nm vs. time for temperature cycling starting from 15 ◦ C (A), 25 ◦ C (B), 35 ◦ C (C), 25 ◦ C (D) and 15 ◦ C (E). This is for an emulsion of DCE-in-D2 O, stabilised with sucrose and the potential determining salt LiTPFB, using a 10 mm pathlength cell.
created as outlined earlier using sucrose and stabilising salts to improve stability [14]. A sample was put into a 10 mm pathlength cuvette and analysed using the UV–vis spectrophotometer set at 15 ◦ C. Fig. 11 shows the turbidity versus time of the emulsion sample as the temperature was cycled starting from 15 to 35 ◦ C and then down to 15 ◦ C again. As the temperature was increased, the drop volume decreased accordingly, and thus there was a corresponding decrease in turbidity. The reverse was observed on cooling from 35 to 15 ◦ C. The turbidity at 25 ◦ C, seen at points (B) and (D) in Fig. 11, shows no change after heating and cooling, indicating that there was no significant change in the emulsion. There was, however, a small decrease in the turbidity of the emulsion at 15 ◦ C (E), at the end of the experiment compared to the start, and this was probably due to convection currents destabilising the emulsion during the experiment. The reversibility is simply due to the DCE dissolving into the water as the temperature is increased from 15 to 35 ◦ C, hence decreasing the emulsion drop volume. This dissolved DCE then acts as a reservoir, facilitating the growth of the drops as the temperature of the emulsion is the changed from 35 to 15 ◦ C. The rate at which the temperature was altered was also important and this was maintained at a slow rate of approximately 1–2 ◦ C min−1 during the experiment. This was because, if the temperature was decreased too quickly (approximately 1–2 ◦ C s−1 ), the drop volume would increase slightly, but a secondary emulsion would also start to nucleate and grow. 4. Conclusions Reproducible emulsions were formed using a Condensation Technique, with the manipulation of the solubilities of DCE and water. This technique was optimised and a procedure was created for the formation of surfactantless DCE-in-water and water-in-DCE emulsions. This involved heating a mixture of DCE and water to 65 ◦ C for 1 h with stirring and then cooling rapidly in a water bath set at 15 ◦ C for 30 min. The emulsion was then characterised using optical microscopy and UV–vis spectrophotometry and was found to have an initial number density
360
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
of 140 ± 12 × 106 or 150 ± 15 × 106 drops cm−3 and an initial average drop volume of 25 ± 5 × 10−12 or 22 ± 3 × 10−12 cm3 for DCE-in-D2 O or D2 O-in-DCE emulsions, respectively. The weight percent of the condensed emulsion drops was also similar for both systems, irrespective of their differing solubilities. For the DCE-in-D2 O emulsion, the drop volume almost doubled after 7 h in a sealed container, and was shown to be very unstable. Turbidity measurements determined that a water-in-DCE emulsion was more stable relative to an emulsion containing DCE-in-D2 O. This was due to the different viscosities of the DCE and D2 O and because this restricted the motion of the D2 O drops in DCE when compared with DCE drops in D2 O. The emulsion drops were also found to be very sensitive to external shear and this was shown when using conventional light scattering techniques. The emulsions were seen to be sensitive to changes in temperature, with the drop volume being inversely proportional to an increase in temperature. This phenomenon was shown to be reversible and that the drop volume could be controlled using temperature changes. This was only dependent on the temperature being changed by 1–2 ◦ C min−1 , rather than 1–2 ◦ C s−1 . If the temperature change occurred too quickly, a secondary emulsion would start to nucleate and grow from the excess dissolved in the continuous phase, rather than this adding volume to the original emulsion drops. Overall, techniques, using a condensation method, optical microscopy and turbidity measurements, were developed and optimised to create and characterise a surfactantless oil-in-water or water-in-oil emulsion with a reproducible drop volume and number density. The drop volume could then be accurately controlled using temperature manipulation. Acknowledgements The Authors would like to thank: financial support from the EPSRC (GR/R12084/01); Prof. Dame Julia Higgins and Imperial College Analytical Services; Prof. David Williams from Unipath; Jim Stevenson and Dick Waymark from UCL Technical Services.
References [1] J. Weiss, D.J. McClements, Influence of Ostwald ripening on rheology of oil-in-water emulsions containing electrostatically stabilized droplets, Langmuir 16 (2000) 2145–2150. [2] B.P. Binks, W.-G. Cho, P.D.I. Fletcher, D.N. Petsev, Stability of oil-inwater emulsions in a low interfacial tension system, Langmuir 16 (2000) 1025–1034. [3] A.S. Kabalnov, A.V. Pertzov, E.D. Shchukin, Ostwald ripening in emulsions. I. Direct observations of Ostwald ripening in emulsions, J. Colloid Int. Sci. 118 (1987) 590–597. [4] A. Hickel, C.J. Ragke, H.W. Blanch, Hydroxynitrile lyase adsorption at liquid/liquid interfaces, J. Mol. Catal. B 5 (1998) 349–354. [5] K. Kamogawa, M. Matsumoto, T. Kobayashi, T. Sakai, H. Sakai, M. Abe, Dispersion and stabilising effects of n-hexadecane on tetralin and benzene metastable droplets in surfactant-free conditions, Langmuir 15 (1999) 1913–1917. [6] K. Kamogawa, H. Akatsuka, M. Matsumoto, S. Yokoyama, T. Sakai, H. Sakai, M. Abe, Surfactant-free O/W emulsion formation of oleic acid and its esters with ultrasonic dispersion, Colloid Surf. A 180 (2001) 41–53. [7] R.M. Pashley, Effect of degassing on the formation and stability of surfactant-free emulsions and fine Teflon dispersions, J. Phys. Chem. B 107 (2003) 1714–1720. [8] G.R. Burnett, R. Atkin, S. Hicks, J. Eastoe, Surfactant-free “emulsions” generated by freeze-thaw, Langmuir 20 (2004) 5673–5678. [9] J. Strutwolf, A.L. Barker, M. Gonsalves, D.J. Caruana, P.R. Unwin, D.E. Williams, J.R.P. Webster, Probing liquid/liquid interfaces using neutron reflection measurements and scanning electrochemical microscopy, J. Electroanal. Chem. 483 (2000) 163–173. [10] A. Zarbakhsh, J. Bowers, J.R.P. Webster, A new approach for measuring neutron reflection from a liquid/liquid interface, Meas. Sci. Technol. 10 (1999) 738–743. [11] T. Cosgrove, J.H.E. Hone, A.M. Howe, R.K. Heenan, A small-angle neutron scattering study of the structure of gelatin at the surface of polystyrene latex particles, Langmuir 14 (1998) 5376–5382. [12] E. Staples, J. Penfold, I.J. Tucker, Adsorption of mixed surfactants at the oil-water interface, Phys. Chem. B 104 (2000) 606–614. [13] K.G. Marinova, R.G. Alargova, N.D. Denkov, O.D. Velev, D.N. Petsev, I.B. Ivanov, R.P. Borwankar, Charging of oil–water interfaces due to spontaneous adsorption of hydroxyl ions, Langmuir 12 (1996) 2045–2051. [14] K.M. Knight, A surfactantless emulsion as a model for the liquid–liquid interface, Ph.D. Thesis, University College London, 2005. [15] D.H. Everett, Basic Principles of Colloid Science, Royal Society of Chemistry Paperbacks, London, 1998. [16] H. Stephen, T. Stephen, Solubilities of Inorganic and Organic Compounds, vol. 1, Pergamon Press, Oxford, 1963. [17] J. Merikanto, H. Vehkam¨aki, E. Zapadinsky, Monte Carlo simulations of critical cluster sizes and nucleation rates of water, J. Chem. Phys. 121 (2004) 914–924.
Formation and characteristion of surfactantless oil-in-water and water-in-oil emulsions Katherine M. Knight, Daren J. Caruana ∗ Department of Chemistry, University College London, 20 Gordon St., London WC1H 0AJ, United Kingdom Received 12 July 2006; received in revised form 8 December 2006; accepted 28 December 2006 Available online 11 January 2007
Abstract A liquid–liquid interface in the form of a surfactantless oil-in-water or water-in-oil emulsion has been created and characterised. The emulsions, composed of 1,2-dichloroethane (DCE) and D2 O were created using a condensation method and this technique has been optimised to give the conditions needed to produce reproducible emulsions with comparable drop volumes and number densities. These have been measured using optical microscopy and light scattering techniques to characterise the emulsions and to investigate how the drop volume, and hence the solubilities of the two phases, are dependent on temperature. The destabilisation of the emulsion, due to the sedimentation of DCE (ρ = 1.235 g cm−3 ) or creaming of D2 O (ρ = 1.105 g cm−3 ), and the coalescence of the drops, has also been measured qualitatively using turbidity measurements. © 2007 Elsevier B.V. All rights reserved. Keywords: Liquid–liquid interface; Emulsions; Condensation; Surfactant-free
1. Introduction Mixtures of two or more liquids that are not miscible in each other are inherently unstable and will, after some time, separate into two phases. This phase separation can be reduced by the use of stabilisers, such as surfactants or polymers, to produce a dispersed emulsion system, and in some cases may produce thermodynamically stable microemulsions. There are many examples of oil-in-water emulsions stabilised by using ionic or non-ionic surfactants [1–3] and the basic principles of destabilisation processes have been developed on the basis of investigation of these systems. Emulsions formed in this way are technologically very important in the food, chemical, cosmetic and nanotechnology industries. For example, enzyme catalysed reactions of substrates that are oil soluble, using enzyme-modified oil-in-water emulsions, are of interest to the pharmaceutical industry [4]. However, with increasing pressure of strict environmental legislation controlling the use of chemical agents, the use of amphiphiles is becoming undesirable in industrial processes. As a consequence, the development of alternative methods or agents to stabilise emulsions, that can
∗
Corresponding author. Tel.: +44 207679 4527; fax: +44 207679 7463. E-mail address: [email protected] (D.J. Caruana).
0927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2006.12.081
replace traditional surfactants and are cheaper and environmentally benign, is becoming a priority in emulsion research. For this reason, surfactantless strategies for stabilising oil-in-water emulsions are receiving a great deal of attention owing to the technological benefits and scientific importance. Abe and co-workers have described a method of producing sub micron sized oil drops in water by sonication of binary mixtures of immiscible species, such as benzene/n-hexadecane in water [5] or oleic acids and various esters in water [6]. The emulsions produced appear to be stable for longer than days, however the precise physical basis of the stabilisation is unclear. Pashley [7] reported a new method of stabilising an oil-in-water emulsion by degassing a solution containing dodecane or squalane derivatives in water, and producing an emulsion by mechanical agitation of the solution. It was proposed that by removing dissolved gases from the solution, the hydrophobic interaction between the oil drop was reduced and so stabilises the emulsion [8]. Unlike the liquid–solid and liquid–air interfaces, studies of the structure of the interface between two immiscible liquids presents a more challenging experimental and technical problem [9,10]. Most experimental approaches used to study either the structure of the interface or the dynamics, rely on the formation of a continuous interface between the immiscible liquids, held apart by their different densities. The technique described in
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
353
Fig. 1. Schematic diagram of a DCE drop in water (A) and a water drop in DCE (B) showing the dynamics of solvent exchange, schemes 1 and 2.
this paper, provides a method of producing a stable ‘clean’ interface which can be investigated using light, or neutron scattering techniques [11,12]. We report a procedure for creating surfactantless emulsions of DCE-in-D2 O or D2 O-in-DCE. These are created using a condensation technique, [13] essentially controlling the solubility, of the two partially immiscible phases in each other (equilibriums 1 and 2 in Fig. 1A and B), by manipulating the temperature. The stability and characterisation of the emulsions has been investigated and monitored using turbidity, optical microscopy and light scattering measurements, and the drop volumes and number densities have been shown to be reproducible and comparable between the two emulsions. 2. Experimental section 2.1. Materials Any H2 O, used in the experiments and for cleaning was from a Milli-Q Gradient 18.2 M cm (at 25 ◦ C) water system. Glassware was thoroughly cleaned by boil washing in a 2 l beaker filled with dilute sulphuric acid (0.1 M, BDH; Analar grade with a minimum assay of 98.0%) for 2 h to remove contamination from detergents. The glassware was then thoroughly rinsed using deionised water several times and dried in an oven at 65 ◦ C for 12 h. The emulsion was composed of 1,2dichloroethane (DCE) (Sigma; >99% spectrophotometric grade) and D2 O (Aldrich; 99.9% atom D). All chemicals were used as received. 2.2. Procedure for creating the emulsions 10 cm3 of the chosen continuous phase and 1 cm3 of the emulsion sub-phase were mixed in a 25 cm3 stoppered glass conical flask. This was then heated to 65 ◦ C, with gentle stirring, in a 500 ml glass beaker water-bath on a hot-plate for 1 h. The flask was then put into a water-bath (Grant Ltd., 6; Grant Instruments, UK) set at 15 ◦ C and left to cool for 30 min. The emulsion formed immediately as the solution was cooled. These were found to be the most effective methods of heating and cooling due to efficient heat exchange. After the solution had cooled for 30 min, a sample of the emulsion was taken for analysis using a variable volume pipette (Volac 200–1000 l) with a modified tip. The pipette tip was cut, approximately 5 mm from the end, to increase the aperture surface area and so reduce the shear forces when pipetting the solution. The sample was removed with care to avoid withdraw-
ing large drops (1–2 mm) of suspended emulsion phase. These were always present in the bulk solution due to the excess volume of sub-phase used in the formation of the emulsion. The emulsion was then studied using either turbidity measurements with a UV–vis spectrophotometer to look at stability, or an optical microscope to measure the average volume and number density of the drops. 2.3. Stability measurements As a measure of stability with time, the turbidity of the emulsion was monitored using a UV–vis spectrophotometer (Agilent 8453 UV–vis Spectroscopy System; Agilent Technologies, Germany) with a thermostatted cell holder set at 15 ◦ C. The absorbance spectra between 190 and 1100 nm were recorded every 10 s, and even though the processes of sedimentation and coagulation could not be quantified separately, this gave a qualitative measure of the turbidity of the sample. 10 m or 2 mm pathlength Far UV quartz cuvettes (Hellma UK Ltd.) were used in the spectrophotometer and were filled to the top to prevent any DCE evaporating out of the emulsion into the headspace at the top of the cuvette. 2.4. Optical microscope experiments The emulsion drop volume and number density were determined using Optical Microscopy (Jenalab; Carl Zeiss, Germany), with transmission illumination, connected to a video camera (TK-1085E; JVC, Japan) with PC data collection (Presto Video Works Ver 4.1 Rev 6; NewSoft Technology Corp., Taiwan). The temperature was maintained and controlled with the use of a peltier device (Thermo Electric Cooler Type DT 1069; Marlow Industries Inc., USA), positioned on the stage of the microscope. This was connected to a variable current supply temperature selector (built in-house; R. Waymark) and the copper plate above the peltier device, allowed for efficient heat transfer between the peltier device and the microscope slide. A sample of the emulsion was put on the microscope slide on the temperature controller and the slide temperature was recorded using a thermocouple at the end of each experiment. This allowed for any changes in temperature due to extra heating from the microscope lamp. The thermocouple was also fixed to the top of the slide to compensate for any differences between the temperature selector and the emulsion sample, and for all sizing experiments the selector was set at 15 ± 0.2 ◦ C. Using the video camera and data capture software, images of the emulsion could be recorded and from these the average
354
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
drop volume was calculated. This was done using a calibrated size standard of 5.0 ± 0.05 m Polymer Microspheres in water (Duke Scientific Corporation, CA) and an image of these can be seen in Fig. 2A. Due to the nature of the emulsion, the diffraction pattern of individual drops changed according to the focal plane (Fig. 2D–I) and so, for accuracy, the sizing images were taken when the drops appeared with the same diffraction pattern (Fig. 2G). To calculate an average volume, three different images of each sample were recorded and analysed and a typical image of an emulsion can be seen in Fig. 2B. From these images, approximately 3–9 individual drops for each sample were in the same focal plane and therefore measurable. An error was then calculated from the standard deviation of the different drop measurements.
The number density of the emulsion was determined using a haemocytometer (non-metallised Improved Neubauer; Hawksley & Sons Ltd., UK) with a grid of 1/400 mm2 × 0.1 mm depth etched onto the surface. By placing a sample of the emulsion on the haemocytometer, the number of drops per square could be counted under the microscope and then scaled to give a value for drops per cm3 . To obtain an average of six values for the number density per cm3 , three images of each sample were taken (Fig. 2C) and the standard deviation was used as the error. 2.5. Light scattering experiments To determine the emulsion drop volume and size distribution, a Mastersizer 2000 was used at room temperature, 25 ◦ C. An
Fig. 2. Typical images from optical microscopy experiments measuring drop volume (B) and number density (using the haemocytometer) (C). An image of the microspheres used for size calibration is also included (A). All images were taken with the same magnification. (D–I) are images at different focal planes through an emulsion drop, and show some of the different diffraction patterns observed. The diffraction pattern in (G) was chosen as showing the sharpest interface and only drops seen with this pattern were measured.
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
355
emulsion was created, following the procedure described, composed of 100 cm3 H2 O and 1 cm3 DCE. Also included were, 0.1 M tetraethylammonium chloride (TEACl) in the water and 1 × 10−3 M tetraethylammonium tetraphenylborate (TEATPB) in the DCE phase to help stabilise the emulsion [14]. The emulsion was also added to the Mastersizer ‘hot’, at 65 ◦ C, rather than being cooled first. This was to make sure that the emulsion formed at the same temperature as the equipment and avoided any destabilisation caused when introducing the sample. A larger volume of the aqueous phase was also used so that the emulsion could be added directly to the Mastersizer without prior dilution. This was because the sizer employed a flow-through system which would mix the sample with an aqueous medium and dilute the sample for analysis. Stirring and dilution have been seen to destabilise the emulsion due to an increase in shear [14] and so, by initially using a larger volume of water to make the emulsion, this would eliminate the dilution stage and allow the propeller to be run at its lowest setting. For these experiments, this setting was 1500 rpm. 3. Results and discussion 3.1. Emulsion formation Emulsions can be created either by dispersion – breaking down the bulk material via shaking or stirring; or condensation of the discontinuous phase – building up the drops [15]. Dispersion methods generally form emulsions with an inhomogeneous and uncontrolled size distribution, with the discontinuous phase maintained with the use of surfactants. Condensation methods, however, allow for molecular clusters to form which grow via molecular diffusion to produce drops with a homogeneous size distribution. Emulsions can be made using this method by heating a mixture of oil and water, where the solubility of the dispersed phase, e.g. oil, in the continuous liquid, e.g. water, will increase, creating a saturated solution. This is then cooled rapidly, thereby decreasing the solubility and causing the oil to condense from the water as small homogeneous drops [16]. The solubility of DCE in a large excess of water, and water in a large excess of DCE, are shown in Fig. 3A and B and fitted with a three parameter exponential increase equation, y = −0.45 + 0.075e0.03T and y = 0.747 + 0.057e0.03T for DCE-in-water and water-in-DCE, respectively [16]. Using these equations, Table 1 shows the weight percent of dissolved DCE or water at different temperatures, and the calculated weight percent present as drops when the emulsion is heated to 65 ◦ C and cooled to temperatures in the range 55–5 ◦ C. The emulsion drops form initially via a process of nucleation where a cluster of molecules aggregate to make a critical nucleus of dimensions 1–10 nm, where the change in Gibbs freeenergy with respect to cluster size is zero [15]. The cluster size of this critical nucleus has been shown to be in the order of four molecules for the nucleation of water [17]. The extent of nucleation depends on the saturation of the oil in water; when the temperature of the system is decreased rapidly, this causes the concentration of the aqueous dissolved oil to suddenly go above the critical supersaturation point, causing
Fig. 3. Solubility plots of DCE-in-H2 O (A) determined between 23 and 73 ◦ C, and H2 O-in-DCE (B) determined between 20 and 70 ◦ C, from Ref. [16]. The solid lines in both graphs are regression fits using a three parameter exponential increase equation included in the plots.
the oil to spontaneously nucleate to reduce the concentration to below the critical point. Once this concentration is reached, no more nucleation will occur and instead the existing clusters will grow via molecular diffusion until the excess aqueous dissolved oil is depleted [15]. Because the nucleation is spontaneous and the nucleation rate is very fast, the resulting emulsion will initially have a monodisperse size distribution. After this time there would be some destabilisation from Ostwald ripening for an emulsion produced using this technique. In turbidity experiments using the UV–vis spectrophotometer, a sharp peak with a scattering >3.5 a.u. was seen at 195 nm when the emulsion was scanned from 190 to 1100 nm, shown in Fig. 4. This sharp peak occurred for all emulsion samples Table 1 Showing the solubility weight percent of DCE in an excess of water and water in an excess of DCE at different temperatures [16] and the calculated weight percent present as drops when the emulsion is heated to 65 ◦ C and cooled to temperatures in the range 55–5 ◦ C Temperature (◦ C)
65 55 45 35 25 15 5
Solubility (wt.%) DCE-in-H2 O
H2 O-in-DCE
1.25 1.11 1.01 0.94 0.88 0.85 0.82
0.46 0.32 0.23 0.16 0.11 0.08 0.06
DCE drops (wt.%)
H2 O drops (wt.%)
– 0.14 0.24 0.31 0.37 0.40 0.43
– 0.14 0.23 0.30 0.35 0.38 0.40
356
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
Fig. 4. Graph showing the turbidity vs. wavelength from 190 to 300 nm for an emulsion containing DCE-in-D2 O (including sucrose for stability), illustrating the sharp ‘scattering’ peak seen at 195 nm. The different plots correspond to measurements taken at different times; (—) initial scan, (- - -) after 10 min, and (· · ·) after 1 h.
that were placed into the spectrophotometer at 65 ◦ C and then allowed to cool to 15 ◦ C in the beam, while the turbidity was measured. The magnitude of this peak was seen to be at the saturation limit of the UV–vis absorbance over the first 10 min while the emulsion formed, and this then decreased slowly as the emulsion destabilised. This peak was also present in a sample of D2 O saturated with DCE at 15 ◦ C and is thought to correspond to the clusters of DCE molecules dissolved in the aqueous phase. The saturated maximum and then steady decrease in the scattering of this peak would then be due to the aqueous dissolved DCE molecules forming clusters and then growing to form emulsion drops, illustrating the link between the turbidity of an emulsion and its number density. The absorbance of this peak is still quite high after 1 h due to the presence of DCE still dissolved in the aqueous phase at 15 ◦ C. To determine the length of time needed to create a saturated solution of DCE-in-water at 65 ◦ C, emulsions were formed using the procedure described earlier but with different recorded solubilisation times. D2 O and sucrose were used in these experiments to reduce the rate of sedimentation and hence improve the stability of the emulsion, and this will be described fully in a later publication [14]. All emulsions were cooled to 15 ◦ C for 30 min. A sample from each emulsion was then taken and the average drop volume and number density were measured as outlined before. These were used to calculate the total weight percent of DCE present as condensed drops and, using Table 1 this should be a value of approximately 0.40 wt.% for a saturated heated emulsion sample. Graphs of the number density of the drops per cm3 and the weight percent of the DCE condensed as drops versus solubilisation time are shown in Fig. 5A and B, respectively. From Fig. 5A and B, it can be seen that both the number density and DCE weight percent reach a plateau after 1 h of solubilisation time. From Fig. 5B, it can also be seen that the weight percent of DCE condensed as drops is approximately 0.4 wt.% after 1 h. This supports the literature data seen in Table 1 and verifies that the heated emulsion sample is saturated after 1 h. The turbidity of the emulsion was also measured (Fig. 5C) for
Fig. 5. Number density (A), DCE weight percent (B) and turbidity measured at 550 nm (C) vs. solubilisation time for an emulsion of DCE-in-D2 O only. The different symbols used in (C) correspond to two different experimental data sets and each data point shown for all plots, was determined from a different emulsion created as described in the experimental section. Sucrose was included in these experiments to reduce the rate of sedimentation and improve stability, and will be discussed later.
emulsions heated for different times, and by plotting this with the emulsion number density from Fig. 5A, the fitted regression line in Fig. 6 shows a linear trend, illustrating the link between the turbidity and number density of the emulsion. A similar result would be expected for an emulsion composed of water-in-DCE, due to the data in Table 1. Even though the solubility of water in an excess of DCE is less than that of DCE in an excess of water, the calculated solubility weight percent condensed as emulsion drops is comparable between the two emulsion systems formed by heating to 65 ◦ C and cooling to temperatures in a range of 55–5 ◦ C. 3.2. Emulsion characterisation 3.2.1. Light scattering experiments Surfactant stabilised emulsions are conventionally characterised using light scattering techniques involving analysis using
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
357
Table 2 Comparing the measured drop volume, number density and calculated weight percent of emulsion drops for DCE-in-D2 O (stabilised with sucrose) and D2 O-in-DCE emulsions Emulsion system
Drop volume (×10−12 cm3 )
Number density (×106 drops cm−3 )
Weight percent emulsion drops
D2 O-in-DCE DCE-in-D2 O (+sucrose)
22 ± 3 (3.5 ± 0.1 m) 25 ± 5 (3.6 ± 0.2 m)
150 ± 15 140 ± 12
0.33 ± 0.05 0.35 ± 0.06
These were measured after 30 min cooling at 15 ◦ C and all other conditions were as described in the experimental.
Fig. 6. Graph plotting the emulsion turbidity vs. number density, for a DCEin-D2 O only emulsion, to illustrate the linear relationship between them. The equation for the fitted regression line is y = 24x − 58.
Mie Theory, using the dependence of the complex scattering from a large sphere, on the scattering angle [15]. To determine the emulsion drop volume and size distribution, a Mastersizer 2000 was used and, although this still accelerated the rate of destabilisation due to shear, this gave an interesting result for the drop volume, seen in Fig. 7. From Fig. 7, two peaks, at 3.6 and 12.3 m belonging to two different populations of drop diameters, can be seen. The peak at 3.6 m corresponds to the size of the emulsion drops, whereas the larger peak at 12.3 m was believed to be from larger coalesced drops formed by the stirring of the solution. Using the measured percentage volume for each diameter, a ratio of 6:1 is calculated for drops at 3.6:12.3 m, showing that the smaller drops are more prevalent in solution, even though the peak height for the larger drops is much greater than that for the small drops.
DCE-in-D2 O (including sucrose), and D2 O-in-DCE. These were heated to 65 ◦ C and cooled to 15 ◦ C, and samples were taken for analysis using optical microscopy as outlined. The drop volume and number density were measured for each emulsion and the results are shown in Table 2, as is the calculated weight percent of drops. From Table 2, it can be seen that there is no significant difference in the drop volume and number density for emulsions containing water-in-oil or oil-in-water formed using a condensation technique. The calculated weight percent of emulsion drops also corresponds well with the expected value determined from the literature in Table 1. Although the weight percent in these emulsions appears numerically small, the emulsions are very dense. If the average separation distance between each drop is calculated, using l = N−1/3 , where N is the number density of drops per m3 , a value of 15 m is determined. This is in the same order of magnitude as the drop diameter of 3.75 ± 0.25 m at 15 ◦ C. 3.2.3. Stability measurements To measure the stability of the oil-in-water emulsion, a mixture of DCE and D2 O (including 1.5 M sucrose for stability), was prepared as described, and a sample of the emulsion was put on a microscope slide on the Peltier device, set at 15 ◦ C. The initial drop volume and number density were measured whilst the flask containing the bulk emulsion solution was maintained at 15 ◦ C in the water bath for subsequent samples. Measurements of the initial drop volume were recorded as outlined, and then fresh samples from the bulk emulsion were taken every hour for 5 h as a measure of the emulsion stability. The change in drop volume with time are shown in Fig. 8, and it can be seen that the drop volume increased linearly.
3.2.2. Optical microscope experiments To accurately measure the emulsion drop size and number density emulsions were created, as described, containing
Fig. 7. Graph showing emulsion drop percentage volume vs. drop diameter (m) for an emulsion containing DCE-in-H2 O, measured at 25 ◦ C using a Mastersizer 2000.
Fig. 8. Change in drop volume with time for a DCE-in-D2 O (stabilised with sucrose) emulsion system.
358
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
The number density of the emulsion was also measured every hour in this experiment; however, due to the high vapour pressure of DCE, as every sample was removed, the DCE in the headspace of the vessel was also released. As a result the number density, and thus the volume fraction of DCE in the emulsion, was reduced falsely by a destabilising factor other than sedimentation or coalescence. This was also seen to be true for the drop volume, and so an experiment was performed where the drop volume and number density were measured initially and then again after 20 h. The flask was not reopened during this time and the results show that, for an emulsion of DCE-in-D2 O (including sucrose), heated to 65 ◦ C and cooled to 15 ◦ C, the initial drop volume was 25 ± 5 × 10−12 cm3 and after 20 h, this increased to 180 ± 4 × 10−12 cm3 . The initial number density was found to be 140 ± 12 × 106 drops cm−3 and after 20 h and this decreased to 7 ± 6 × 106 drops cm−3 (this error seems to be large due to extensive emulsion destabilisation, resulting in only 3 ± 3 drops being counted per square of the Haemocytometer). The value for the drop volume after 20 h is similar to that obtained after 7 h by extrapolating Fig. 8, and this suggests that by opening the flask and releasing the headspace every hour, the rate of destabilisation of the emulsion was accelerated. Fig. 8, still shows, qualitatively, how the emulsion destabilises over time and from calculations using the data for the drop volume measured initially and after 20 h, the volume is seen to double after 7 h for a DCE-in-D2 O (including sucrose) emulsion, instead of after 2 h as suggested by Fig. 8. Attempts were also made to measure the drop volume every hour, for a D2 O-in-DCE emulsion, to see qualitatively how the emulsion destabilised over time. Difficulties were encountered, however, due to the high vapour pressure of the DCE and these were similar to the problems experienced for the DCE-inD2 O emulsion. This was because the DCE would evaporate into the headspace of the flask and every time a measurement was taken, this would be released. Because the continuous phase of the emulsion was DCE, this caused the emulsion to decrease in total volume every time the flask was opened and caused rapid destabilisation of the emulsion. There were also problems encountered when the measurements were taken under the microscope, due to the DCE evaporation from the glass slide. To further investigate the stability of emulsions composed of DCE-in-water and water-in-DCE, emulsions were created as described using H2 O or D2 O in an excess of DCE, and DCE in an excess of D2 O, and samples were taken for analysis using turbidity measurements maintained at 15 ◦ C. The turbidity measurements shown in Fig. 9, have been normalised for easier qualitative comparison, due to the emulsions only being heated until they reached 65 ◦ C, approximately 20 min, rather than stirring them for 1 h at 65 ◦ C. This meant that the DCE was not fully saturated with the aqueous solution when the emulsion was cooled to 15 ◦ C and the initial turbidity values are 1.26, 1.41 and 0.99 au for DCE-in-D2 O, H2 O-in-DCE and D2 O-in-DCE, respectively. From Fig. 9, it can be seen that the water-in-oil systems were significantly more stable than an emulsion containing oil-in-water, with the change in the turbidity with time being increased from −172.6 × 10−6 a.u. s−1 for DCE-in-D2 O, to −30.2 × 10−6 a.u. s−1 for a D2 O-in-DCE system.
Fig. 9. Graph comparing normalised turbidity at 550 nm vs. time for emulsions containing (A) DCE-in-D2 O, (B) H2 O-in-DCE and (C) D2 O-in-DCE. The emulsions were formed at 15 ◦ C and this temperature was maintained for the turbidity measurements. The original turbidity values were 1.26, 1.41 and 0.99 a.u., and the rate of change of turbidity with time for the plots are −172.6, −138.3 and −30.2 × 10−6 a.u. s−1 for the DCE-in-D2 O, H2 O-in-DCE and D2 O-in-DCE emulsions, respectively.
The difference in stability between the H2 O-in-DCE and D2 O-in-DCE emulsions can be explained by considering the relative densities of the drop. The velocity v, of the sedimentation of a spherical DCE drop with radius r, can be obtained by equating the force due to gravity, g, and the viscous drag of the drop through the media. 4 3 πr (ρd − ρm )g = 6πηrv 3
(1)
rearranging gives, v=
2r2 (ρd − ρm )g 9η
(2)
where η is the viscosity of the DCE solution (0.84 mPa s), and ρd and ρm are the densities of the drop and medium, respectively. Therefore, the sedimentation velocities of D2 O and H2 O drops in DCE medium are −1.53 × 10−4 and −2.61 × 10−4 cm s−1 , respectively, a negative velocity indicates creaming. Hence the destabilisation of the H2 O is faster than that of D2 O drops in DCE. 3.2.4. The effect of temperature on the emulsions Due to the nature of the emulsion system and the way that it is formed, the drop volume is sensitive to temperature. Using the temperature controller, a sample of a DCE-in-D2 O (including sucrose for stability) emulsion, created as described, was cooled to 15 ◦ C for 30 min and then a sample was taken. The drop volume was measured at intervals as the emulsion was heated in recorded increments from 15 to 30 ◦ C, this was then converted to weight percent using the measured initial number density of 140 ± 12 × 106 drops cm−3 and was plotted in Fig. 10. A water-in-oil emulsion was then created composed of 10 cm3 DCE and 1 cm3 D2 O and a sample was taken for analysis using optical microscopy by placing on a slide on the microscope temperature controller, maintained at 15 ◦ C. Once the drop volume measurement had been taken the temperature, of the water bath
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
Fig. 10. Graph showing the weight percent of emulsion drops vs. the temperature of the system for emulsions containing: (䊉) DCE-in-D2 O (stabilised with sucrose) and () D2 O-in-DCE. The solid lines are regression fits using the three parameter exponential decay equations (䊉) y = 0.114 + 2.39e−0.162T and () y = 0.101 + 2.23e−0.146T .
and microscope temperature controller, was increased to 20 ◦ C and a second sample was taken for analysis. This was repeated in 5 ◦ C increments until 30 ◦ C (this was the highest temperature that could be measured accurately before the emulsion destabilised from the DCE evaporating out of the flask) and the results for the calculated drop weight percent, using the initial number density of 150 ± 15 × 106 drops cm−3 for the D2 O-inDCE emulsion, are also included in Fig. 10. This shows that the weight percent of the emulsion drops, and hence also drop volume, decreases with increasing temperature as expected, for both the oil-in-water and water-in-oil emulsions. The data were also regression fitted using a three parameter exponential decay equation of the form y = y0 + ae−bT . A different technique was used to heat the water-in-oil emulsion because of the volatile nature of the DCE and the length of time of the experiment. When the microscope-method was attempted, as used for the oil-in-water emulsion, the sample evaporated from the slide on the temperature controller after one measurement, approximately 2 min, making it impossible to accurately measure the drop volume for more than one temperature using this technique. The regression lines (using equation y = y0 + ae−bT ) from Fig. 10 may be compared using ratios of coefficients with those determined for the solubility of DCE in an excess of water, or water in an excess of DCE, in Fig. 3. The ratio of the coefficients a:b, are 2.56 and 1.67 for DCE-in-water and water-in-DCE, respectively, for the literature solubility data given in Fig. 3. For the emulsion data, the ratio a:b for DCE-in-water and waterin-DCE are 14.76 and 10.14, respectively, from Fig. 10. The numerical values are different because this is only an empirical comparison of parameters that have similar physical meaning under different experimental conditions, however comparing in a similar way the two ratios obtained for the water-in-DCE 0.173 (2.56/14.76) and DCE-in-water 0.165 (1.67/10.14) the values are almost equal. This indicates that the change in drop volume with temperature follows the same trend as the solubilities of DCE and water, as expected. This temperature effect described in Fig. 10 was also reversible and to illustrate this, an emulsion of DCE-in-D2 O was
359
Fig. 11. Turbidity at 550 nm vs. time for temperature cycling starting from 15 ◦ C (A), 25 ◦ C (B), 35 ◦ C (C), 25 ◦ C (D) and 15 ◦ C (E). This is for an emulsion of DCE-in-D2 O, stabilised with sucrose and the potential determining salt LiTPFB, using a 10 mm pathlength cell.
created as outlined earlier using sucrose and stabilising salts to improve stability [14]. A sample was put into a 10 mm pathlength cuvette and analysed using the UV–vis spectrophotometer set at 15 ◦ C. Fig. 11 shows the turbidity versus time of the emulsion sample as the temperature was cycled starting from 15 to 35 ◦ C and then down to 15 ◦ C again. As the temperature was increased, the drop volume decreased accordingly, and thus there was a corresponding decrease in turbidity. The reverse was observed on cooling from 35 to 15 ◦ C. The turbidity at 25 ◦ C, seen at points (B) and (D) in Fig. 11, shows no change after heating and cooling, indicating that there was no significant change in the emulsion. There was, however, a small decrease in the turbidity of the emulsion at 15 ◦ C (E), at the end of the experiment compared to the start, and this was probably due to convection currents destabilising the emulsion during the experiment. The reversibility is simply due to the DCE dissolving into the water as the temperature is increased from 15 to 35 ◦ C, hence decreasing the emulsion drop volume. This dissolved DCE then acts as a reservoir, facilitating the growth of the drops as the temperature of the emulsion is the changed from 35 to 15 ◦ C. The rate at which the temperature was altered was also important and this was maintained at a slow rate of approximately 1–2 ◦ C min−1 during the experiment. This was because, if the temperature was decreased too quickly (approximately 1–2 ◦ C s−1 ), the drop volume would increase slightly, but a secondary emulsion would also start to nucleate and grow. 4. Conclusions Reproducible emulsions were formed using a Condensation Technique, with the manipulation of the solubilities of DCE and water. This technique was optimised and a procedure was created for the formation of surfactantless DCE-in-water and water-in-DCE emulsions. This involved heating a mixture of DCE and water to 65 ◦ C for 1 h with stirring and then cooling rapidly in a water bath set at 15 ◦ C for 30 min. The emulsion was then characterised using optical microscopy and UV–vis spectrophotometry and was found to have an initial number density
360
K.M. Knight, D.J. Caruana / Colloids and Surfaces A: Physicochem. Eng. Aspects 301 (2007) 352–360
of 140 ± 12 × 106 or 150 ± 15 × 106 drops cm−3 and an initial average drop volume of 25 ± 5 × 10−12 or 22 ± 3 × 10−12 cm3 for DCE-in-D2 O or D2 O-in-DCE emulsions, respectively. The weight percent of the condensed emulsion drops was also similar for both systems, irrespective of their differing solubilities. For the DCE-in-D2 O emulsion, the drop volume almost doubled after 7 h in a sealed container, and was shown to be very unstable. Turbidity measurements determined that a water-in-DCE emulsion was more stable relative to an emulsion containing DCE-in-D2 O. This was due to the different viscosities of the DCE and D2 O and because this restricted the motion of the D2 O drops in DCE when compared with DCE drops in D2 O. The emulsion drops were also found to be very sensitive to external shear and this was shown when using conventional light scattering techniques. The emulsions were seen to be sensitive to changes in temperature, with the drop volume being inversely proportional to an increase in temperature. This phenomenon was shown to be reversible and that the drop volume could be controlled using temperature changes. This was only dependent on the temperature being changed by 1–2 ◦ C min−1 , rather than 1–2 ◦ C s−1 . If the temperature change occurred too quickly, a secondary emulsion would start to nucleate and grow from the excess dissolved in the continuous phase, rather than this adding volume to the original emulsion drops. Overall, techniques, using a condensation method, optical microscopy and turbidity measurements, were developed and optimised to create and characterise a surfactantless oil-in-water or water-in-oil emulsion with a reproducible drop volume and number density. The drop volume could then be accurately controlled using temperature manipulation. Acknowledgements The Authors would like to thank: financial support from the EPSRC (GR/R12084/01); Prof. Dame Julia Higgins and Imperial College Analytical Services; Prof. David Williams from Unipath; Jim Stevenson and Dick Waymark from UCL Technical Services.
References [1] J. Weiss, D.J. McClements, Influence of Ostwald ripening on rheology of oil-in-water emulsions containing electrostatically stabilized droplets, Langmuir 16 (2000) 2145–2150. [2] B.P. Binks, W.-G. Cho, P.D.I. Fletcher, D.N. Petsev, Stability of oil-inwater emulsions in a low interfacial tension system, Langmuir 16 (2000) 1025–1034. [3] A.S. Kabalnov, A.V. Pertzov, E.D. Shchukin, Ostwald ripening in emulsions. I. Direct observations of Ostwald ripening in emulsions, J. Colloid Int. Sci. 118 (1987) 590–597. [4] A. Hickel, C.J. Ragke, H.W. Blanch, Hydroxynitrile lyase adsorption at liquid/liquid interfaces, J. Mol. Catal. B 5 (1998) 349–354. [5] K. Kamogawa, M. Matsumoto, T. Kobayashi, T. Sakai, H. Sakai, M. Abe, Dispersion and stabilising effects of n-hexadecane on tetralin and benzene metastable droplets in surfactant-free conditions, Langmuir 15 (1999) 1913–1917. [6] K. Kamogawa, H. Akatsuka, M. Matsumoto, S. Yokoyama, T. Sakai, H. Sakai, M. Abe, Surfactant-free O/W emulsion formation of oleic acid and its esters with ultrasonic dispersion, Colloid Surf. A 180 (2001) 41–53. [7] R.M. Pashley, Effect of degassing on the formation and stability of surfactant-free emulsions and fine Teflon dispersions, J. Phys. Chem. B 107 (2003) 1714–1720. [8] G.R. Burnett, R. Atkin, S. Hicks, J. Eastoe, Surfactant-free “emulsions” generated by freeze-thaw, Langmuir 20 (2004) 5673–5678. [9] J. Strutwolf, A.L. Barker, M. Gonsalves, D.J. Caruana, P.R. Unwin, D.E. Williams, J.R.P. Webster, Probing liquid/liquid interfaces using neutron reflection measurements and scanning electrochemical microscopy, J. Electroanal. Chem. 483 (2000) 163–173. [10] A. Zarbakhsh, J. Bowers, J.R.P. Webster, A new approach for measuring neutron reflection from a liquid/liquid interface, Meas. Sci. Technol. 10 (1999) 738–743. [11] T. Cosgrove, J.H.E. Hone, A.M. Howe, R.K. Heenan, A small-angle neutron scattering study of the structure of gelatin at the surface of polystyrene latex particles, Langmuir 14 (1998) 5376–5382. [12] E. Staples, J. Penfold, I.J. Tucker, Adsorption of mixed surfactants at the oil-water interface, Phys. Chem. B 104 (2000) 606–614. [13] K.G. Marinova, R.G. Alargova, N.D. Denkov, O.D. Velev, D.N. Petsev, I.B. Ivanov, R.P. Borwankar, Charging of oil–water interfaces due to spontaneous adsorption of hydroxyl ions, Langmuir 12 (1996) 2045–2051. [14] K.M. Knight, A surfactantless emulsion as a model for the liquid–liquid interface, Ph.D. Thesis, University College London, 2005. [15] D.H. Everett, Basic Principles of Colloid Science, Royal Society of Chemistry Paperbacks, London, 1998. [16] H. Stephen, T. Stephen, Solubilities of Inorganic and Organic Compounds, vol. 1, Pergamon Press, Oxford, 1963. [17] J. Merikanto, H. Vehkam¨aki, E. Zapadinsky, Monte Carlo simulations of critical cluster sizes and nucleation rates of water, J. Chem. Phys. 121 (2004) 914–924.