Article pubs.acs.org/JAFC
Stabilization of Vitamin E‑Enriched Nanoemulsions: Influence of Post-Homogenization Cosurfactant Addition Amir Hossein Saberi,† Yuan Fang,§ and David Julian McClements*,‡ †
Biopolymers and Colloids Laboratory, ‡Department of Food Science, University of Massachusetts Amherst, Amherst, Massachusetts 01003, United States § PepsiCo Global R&D, 100 Stevens Avenue, Valhalla, New York 10595, United States ABSTRACT: Oil-in-water nanoemulsions are being used in the food, beverage, and pharmaceutical industries to encapsulate, protect, and deliver lipophilic bioactive components, such as drugs, vitamins, and nutraceuticals. However, nanoemulsions are thermodynamically unstable systems that breakdown over time. We investigated the influence of posthomogenization cosurfactant addition on the thermal and storage stability of vitamin E acetate nanoemulsions (VE-nanoemulsions) formed from 10% oil phase (VE), 10% surfactant (Tween 80), 20% cosolvent (ethanol), and 60% buffer solution (pH 3). Addition of a nonionic cosurfactant (0.5% Tween 20) caused little change in droplet charge, whereas addition of anionic (0.5% SDS) or cationic (0.5% lauric arginate) cosurfactants caused droplets to be more negative or positive, respectively. Tween 20 addition had little impact on the cloud point of VE-nanoemulsions, but slightly decreased their isothermal storage stability at elevated temperatures (37 °C). Lauric arginate or SDS addition appreciably increased the cloud point, but did not improve storage stability. Indeed, SDS actually decreased the storage stability of the VE-nanoemulsions at elevated temperatures. We discuss these effects in terms of the influence of surfactants on droplet growth through Ostwald ripening and/or coalescence mechanisms. This study provides important information about the effect of cosurfactants on the stability of VE-nanoemulsions suitable for use in pharmaceutical and food products. KEYWORDS: Nanoemulsions, vitamin E, nutraceuticals, pharmaceuticals, stability
■
INTRODUCTION There is increasing interest in the utilization of nanoemulsions in the food, beverage, and pharmaceutical industries as they have a number of potential advantages over conventional emulsions for the oral delivery of lipophilic bioactive compounds. Nanoemulsions (radius 100 nm) due to their smaller droplet sizes: increased transparency, increased stability to gravitational separation and droplet aggregation, and increased oral bioavailability.1,2 In general, nanoemulsions can be prepared using two main methods: high-energy and low-energy approaches.3 High-energy methods use intense disruptive forces to intermingle and break up the oil and water phases and typically involve high-pressure homogenizers, microfluidizers, and ultrasound generators.1 In contrast, low-energy approaches rely on the spontaneous formation of nanoemulsions when specific changes in composition and/or environmental conditions occur.4 Several low-energy approaches have been developed to fabricate nanoemulsions including spontaneous emulsification (SE), phase inversion temperature (PIT), phase inversion composition (PIC), and emulsion inversion point (EPI) methods.4−6 The spontaneous emulsification method has considerable potential for many commercial applications since it simply involves metering an oil−surfactant mixture into an aqueous solution with constant agitation. Recently, we showed that the spontaneous emulsification method can be used for fabrication of vitamin E-enriched nanoemulsions with very fine oil droplets, i.e. r < 30 nm.7 The © 2014 American Chemical Society
size of the droplets could be further reduced by adding cosolvents such as glycerol,8 propylene glycol, or ethanol to the aqueous phase.9 Nanoemulsions with very fine droplets and high optical clarity could be produced by carefully controlling cosolvent type and concentration. However, the long-term stability of the vitamin E nanoemulsions decreased in the presence of cosolvent, especially at elevated storage temperatures.8,9 Nanoemulsions are thermodynamically unstable systems that may break down through a variety of physicochemical mechanisms such as gravitational separation, flocculation, coalescence, Ostwald ripening, and phase inversion.1,10,11 Clear identification of the dominant physicochemical mechanism(s) responsible for the instability of specific nanoemulsions is needed to determine the most efficient strategy to improve their stability. One of the main purposes of this study was therefore to identify the physiochemical mechanisms promoting instability in nanoemulsions produced using the spontaneous emulsification method in the presence of cosolvents. Emulsions stabilized by nonionic surfactants are highly prone to droplet coalescence when the temperature is close to their cloud point or phase inversion temperature.12,13 A correlation between the storage stability of emulsions and the phase inversion temperature (PIT) has been reported.14,15 Shinoda and Saito14 showed that O/W emulsions stabilized by nonionic Received: Revised: Accepted: Published: 1625
September 20, 2013 January 22, 2014 January 24, 2014 January 24, 2014 dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
surfactant, while the aqueous phase was a buffer solution (pH 3.0, 0.8% citric acid, 0.08% sodium benzoate) that contained ethanol. The initial nanoemulsions were prepared with the following composition: 10% vitamin E acetate, 10% Tween 80, 20% ethanol, and 60% buffer solution (pH 3). In these samples, the oil (10 g) and surfactant (10 g) were first mixed together and then the mixture was slowly poured into 80 g of aqueous phase over a 15-min period with continuous stirring. Surfactant Displacement. Surfactant displacement was carried out using a method described previously15 with a slight modification. Vitamin E nanoemulsions were prepared using the spontaneous emulsification method as described above and then diluted 2-fold using solutions of either nonionic (Tween 20) or ionic (SDS and/or lauric arginate) cosurfactants containing different cosurfactant concentrations. Particle Size and Charge Measurements. Particle size distributions were measured using a dynamic light scattering instrument (Zetasizer Nano ZS, Malvern Instruments, Malvern, UK). This instrument determines the particle size from intensity− time fluctuations of a laser beam (633 nm) scattered from a sample at an angle of 173°. Each individual measurement was an average of 13 runs. To avoid multiple scattering effects, samples were diluted ( × 500) before the particle size measurements using buffer solution (pH 3.0). The mean droplet radius (Z-average) was calculated from the particle size distribution. All measurements were conducted at ambient temperature. The ζ-potential of the droplets in the emulsions was measured using a particle electrophoresis instrument (NanoZS, Malvern Instruments, Malvern, UK). This instrument calculates the ζ-potential of the particles from measurements of the direction and velocity that the droplets moved in the applied electric field. Conversion of electrophoretic mobility measurements into ζ-potential values was conducted by the software using the Smoluchowsky model. To avoid the multiple scattering, nanoemulsion was diluted before the particle charge measurement using buffer solution (pH 3.0, 0.8% citric acid, 0.08% sodium benzoate). Turbidity Measurements and Cloud Point Calculation. The influence of temperature on nanoemulsion turbidity was determined using a UV−visible spectrophotometer with temperature scanning capabilities (Evolution Array, Thermo Scientific). The absorbance (A) at 600 nm was measured as the temperature was increased from 20 to 90 °C at 1 °C per minute. The turbidity increase was calculated as A(T) − A(T0), where A(T) is the turbidity at temperature T and A(T0) is the turbidity at the starting temperature, T0 (20 °C). The cloud point was determined by fitting the turbidity increase versus temperature data using graphical software (SigmaPlot, Systat Software Inc.). The cloud point was defined as the temperature at which the turbidity increase first reached 1 cm−1. Statistical Analysis. All experiments were carried out two or three times using two freshly prepared samples, and the results are reported as the calculated mean and standard deviation of these measurements. Statistical analysis was performed through subjection of data to analysis of variance (ANOVA) using statistics software (version 12.0; SPSS, Inc., Chicago, IL). Means were subjected to Duncan’s test and a P-value of 96% purity) was kindly donated by BASF (Ludwigshafen, Germany). Nonionic surfactants (Tween 80 and 20) and sodium dodecyl sulfate (SDS, 99% purity) were purchased from Sigma-Aldrich Co. (St. Louis, MO). Lauric arginate (LAE) was obtained from Vedeqsa (Vedeqsa Inc., New York, NY). Citric acid and sodium benzoate were provided by PepsiCo (Valhalla, NY). Ethanol (99%) was purchased from Fisher Scientific Co. (Springfield, NJ). Double distilled water was used in the preparation of all solutions and nanoemulsions. All concentrations are expressed as a mass percentage (w/w %) of whole emulsion. Nanoemulsion Preparation. Nanoemulsion formation was carried out using a method based on a spontaneous emulsification procedure described in the previous study.9 In brief, spontaneous emulsification was performed by addition of an organic phase to an aqueous phase while magnetically stirring (600 rpm) the system at 25 °C. The organic phase consisted of different amounts of VE and
RESULTS AND DISCUSSION
Effect of Cosurfactants on Thermal Stability. Knowledge of the thermal stability of a nanoemulsion-based delivery system is of particular importance as they are often exposed to elevated temperatures during their manufacture, storage, and utilization. Thermal stability measurements may also be useful as an accelerated test for predicting the stability of nanoemulsions during long-term storage.9 Therefore, oil-in-water nanoemulsions containing 10% vitamin E acetate (VE), 10% Tween 80, 20% ethanol, and 60% aqueous phase were prepared using the spontaneous emulsification method, and then diluted 2-fold using buffer solutions containing different concentrations of cationic (lauric arginate), ionic (SDS), or nonionic (Tween 1626
dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
20) cosurfactants. This led to nanoemulsions with a final composition of 5% VE, 5% Tween 80, 10% ethanol, and 0−1% cosurfactant. The dependence of the turbidity on temperature for buffer solution containing 5% (w/w) Tween 80 and 10% (w/w) ethanol and nanoemulsions containing a similar amount of Tween 80 and ethanol and different types and amounts of cosurfactants were determined during heating from 20 to 90 °C. The results for the buffer solution containing Tween 80 and ethanol and nanoemulsions containing different levels of lauric arginate are shown in Figure 2a and b, respectively. Similar
and 10% ethanol (Figure 2a). The turbidity of the buffer solution containing Tween 80 and ethanol did not change over the temperature range of 20−90 °C, suggesting that the cloud point was higher than the range of temperatures studied. However, a steep increase in turbidity was observed at 65 °C for the nanoemulsion (containing no cosurfactant). This result suggests that the presence of the oil phase appreciably decreased the cloud point, presumably by altering the optimum curvature of the system. The presence of the cosurfactants in the nanoemulsions also altered their cloud points, with steep increases in turbidity occurring at 66−74, 71−83, and 75−89 °C for Tween 20, lauric arginate, and SDS, respectively. The observed increase in turbidity upon heating can be related to the progressive dehydration of the polar head groups of the nonionic surfactant molecules with increasing temperature.24,25 The dehydration of the polar head groups of the nonionic surfactants alters their packing parameter (p = AT/AH), and thereby changes the optimum curvature of the surfactant monolayer. Here, AT and AH are the effective cross-sectional areas of the surfactant tail group and headgroup, respectively. At low temperatures, the packing parameter favors the formation of stable oil-in-water nanoemulsions (AH > AT). As the temperature increases, the headgroup becomes dehydrated (lower AH) and so the packing parameter moves toward unity, which leads to an ultralow interfacial tension and accelerated droplet coalescence.26 In addition, surfactant dehydration decreases the steric repulsion between the oil droplets thereby allowing the droplets to come closer together and coalesce. The turbidity increase versus temperature profiles were used to determine the cloud points of the nanoemulsions so that we could establish the influence of cosurfactant type and concentration on their thermal behavior. We defined the cloud point practically as the temperature at which the turbidity increase during heating first exceeded 1 cm−1. In general, the cloud point initially increased with increasing cosurfactant concentration, but then it decreased once a particular cosurfactant concentration was exceeded (Figure 3). The highest cloud points obtained for the three cosurfactants were ∼81, 76, and 69 °C at 0.5% SDS, 0.5% lauric arginate, and 0.25% Tween 20, respectively, which compared to a value of ∼67 °C measured in the absence of cosurfactant.
Figure 2. (a) Effect of temperature on the turbidity of surfactant solution containing 5% surfactant, 10% ethanol (pH 3.0) and nanoemulsion containing 5% vitamin E acetate, 5% surfactant, 10% ethanol, and 80% buffer solution (pH 3.0). (b) Effect of lauric arginate on the temperature dependence of the turbidity of nanoemulsions containing 5% oil phase (vitamin E acetate), 5% surfactant (Tween 80), 10% cosolvent (ethanol), and 80% buffer solution (pH 3).
results were also obtained for the nanoemulsions containing SDS or Tween 20 (data not shown). In general, the turbidity of the VE-nanoemulsions was relatively small at low temperatures, increased steeply above a particular temperature (the “cloud point”), and then remained high upon further heating (Figure 2b). Initially, we compared the cloud points of the nanoemulsion and the buffer solution both containing 5% Tween 80
Figure 3. Effect of cosurfactants on the cloud point of nanoemulsions containing 5% oil phase (vitamin E acetate), 5% surfactant (Tween 80), 10% cosolvent (ethanol), and 80% buffer solution (pH 3). 1627
dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
(Tween 20) caused little change in droplet charge (∼ −0.8 to −1.1 mV). The small negative charge on the droplet surfaces stabilized by the nonionic surfactants can be attributed to the presence of impurities such as free fatty acids in the surfactant or oil, or preferential adsorption of anions (such as hydroxyl ions) from the water.29 Addition of increasing amounts of cationic surfactant (lauric arginate) to the VE-nanoemulsions increased the positive charge on the droplets until the ζpotential reached a relatively constant value (∼ +9 mV). Conversely, adding increasing amounts of an anionic surfactant (SDS) to the VE-nanoemulsions increased the negative charge on the droplets until the ζ-potential reached a relatively constant value (∼ −11 mV). These results suggested that the ionic surfactants adsorbed to the surfaces of the droplets in the VE-nanoemulsions and altered their electrical characteristics, which would help account for the observed influence of ionic cosurfactants on the thermal stability of the nanoemulsions (Figure 3). Effect of Cosurfactants on Storage Stability. Commercial food and beverage products containing emulsion-based delivery systems are required to have good long-term stability during storage, transport, and utilization. Nanoemulsions are thermodynamically unstable systems that may undergo a variety of breakdown processes, including flocculation, coalescence, Ostwald ripening, and gravitational separation.1,10,28 As mentioned earlier, the two main instability mechanisms in VE-nanoemulsions containing cosolvents are droplet coalescence and Ostwald ripening.8,9 We therefore examined the possibility of using a surfactant displacement approach for increasing the storage stability of VE-nanoemulsions (containing 20% cosolvent) against droplet growth through these mechanisms. In this approach, VE-nanoemulsions were formed using 10% Tween 80 and 20% ethanol and then diluted (1:1) with aqueous solutions containing different cosurfactants. Three different VE-nanoemulsions were prepared with final concentrations of 5% VE, 5% Tween 80, 10% ethanol, 0.5% cosurfactant, and 79.5% buffer solution. This cosurfactant concentration was selected because it was close to the level where the highest cloud points were observed in the thermal stability studies for the two ionic surfactants (Figure 3). The influence of storage time and holding temperature on their mean droplet sizes was then measured (Figure 5). In the absence of cosurfactants (control), there was only a slight increase in the mean droplet radius over time for VEnanoemulsions stored at 5 °C for 1 month, but there was an appreciable increase for those stored at 20 and 37 °C (Figure 5): r increasing from ∼22 nm initially to ∼25 nm at 5 °C (about 9% increase), ∼30 nm at 20 °C (about 30% increase), and ∼48 nm at 37 °C (about 112% increase). The droplet size distribution of the VE-nanoemulsions containing no cosurfactant remained monomodal at all storage temperatures, but the peak maximum occurred at larger droplets sizes with increasing storage temperature (Figure 6a). These results indicate that the droplet growth rate increased with increasing holding temperature. Droplet growth may have occurred due to droplet coalescence (merging together of two or more droplets) or Ostwald ripening (diffusion of oil molecules between droplets), which is discussed in Potential Physicochemical Mechanisms. We found that the addition of the cosurfactants to the VEnanoemulsions did not improve their isothermal storage stability (Figure 5), even those that appreciably increased the cloud points (Figure 3). Indeed, some of the cosurfactants actually decreased the isothermal stability of the nanoemulsions
Significantly (p < 0.05) higher cloud points were observed in the presence of ionic cosurfactants than in the presence of the nonionic one. The addition of 0.25% Tween 20 to the VE-nanoemulsions slightly increased the cloud point, but the addition of higher amounts of this cosurfactant caused a slight decrease (Figure 3). Previous studies have reported that Tween 20 has a lower cloud point (∼ 79 °C) than Tween 80 (∼ 95 °C),27 which would account for the observed reduction in cloud point at higher Tween 20 levels. Tween 80 and Tween 20 both have similar polar head groups, but the nonpolar tail-groups are different. Tween 20 has a 12-carbon fully saturated hydrocarbon chain (C12:0) that would be expected to be linear, whereas Tween 80 has an 18-carbon unsaturated hydrocarbon chain (C 18:1 ) that would be expected to be kinked. Consequently, adsorption of Tween 20 to the surfaces of oil droplets initially stabilized by Tween 80 may have changed the optimum curvature of the surfactant monolayer, and thereby altered the PIT and cloud point of the nanoemulsions. The addition of ionic surfactants (SDS or lauric arginate) to the VE-nanoemulsions also caused an appreciable increase in the cloud point at relatively low levels (0−0.5%), but then caused an appreciable decrease at higher levels (Figure 3). This effect can be attributed to the ability of the ionic surfactants to adsorb to the droplet surfaces and alter the optimum curvature of the surfactant monolayer, as well as increasing the electrostatic repulsion between the oil droplets. Effect of Cosurfactants on Droplet Charge. Information about the ability of the cosurfactants to adsorb to oil droplet surfaces can be established using droplet charge measurements. In addition, droplet charge plays a key role in determining the stability of nanoemulsions to aggregation: if the droplet charge is sufficiently high, the droplets are prevented from aggregation due to electrostatic repulsion.28 We therefore measured the droplet charge of VE-nanoemulsions containing different concentrations of ionic and nonionic cosurfactants (Figure 4). The initial nanoemulsions stabilized by Tween 80 had a ζpotential that was slightly negative (∼ −0.8 mV). As expected, addition of different concentrations of nonionic cosurfactant
Figure 4. Effect of cosurfactant on the zeta potential of oil droplets in nanoemulsion containing 5% oil phase (vitamin E acetate), 5% surfactant (Tween 80), 10% cosolvent (ethanol), and 80% buffer solution (pH 3). 1628
dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
Figure 5. Effect of cosurfactant on mean particle diameter of nanoemulsions (5% vitamin E acetate), 5% surfactant (Tween 80), 10% cosolvent (ethanol), 0.5% cosurfactant, and 79.5% buffer solution) after storage at different holding temperatures for one month. Results are compared with controls (containing no cosurfactant) after 0 days (“fresh”) and one month storage. Means with different superscripts are significantly (P < 0.05) different.
at elevated holding temperatures, e.g., droplet growth was faster in the nanoemulsions containing SDS than in the control at 37 °C. Like the controls, the particle size distributions of the VEnanoemulsions in the presence of all three cosurfactants were monomodal at all storage temperatures and times, but with an increase in the location of the maximum at higher storage temperatures (e.g., see Figure 6b for nanoemulsions containing SDS). Potential Physicochemical Mechanisms. There are a number of possible mechanisms that might be responsible for the observed increase in mean particle size in the nanoemulsions during storage, including flocculation, coalescence, and Ostwald ripening.2,6,30 We do not believe that flocculation was important in the nanoemulsions used in this study since flocculated droplets were not observed in the systems using an optical microscope. In general, the growth of the individual droplets within a nanoemulsion may occur through either coalescence and/or Ostwald ripening mechanisms.31 In practice, it is often difficult to distinguish between these two mechanisms, and to precisely establish the relative importance of each mechanism in particular systems. In this section, we present a brief overview of the mechanisms underlying Ostwald ripening and coalescence, and discuss how they may be influenced by cosurfactant addition. Ostwald Ripening. Ostwald ripening (OR) is the process whereby larger droplets grow at the expense of smaller droplets due to diffusion of disperse phase molecules through the intervening continuous phase.32 The driving force for this mechanism is the higher chemical potential of the disperse phase within the smaller droplets due to their higher curvature.33 OR can be divided into two regimes that can be distinguished by the nature of the evolution of the particle size distribution (PSD) during storage: transient and steady state regimes.33 Initially, the shape of the PSD depends on the method used to prepare the nanoemulsion. In the transient regime, the shape of the PSD changes over time due to diffusion of oil molecules from the small to the large droplets. Eventually, the PSD attains a characteristic shape (determined
Figure 6. (a) Influence of storage temperature on particle size distribution of nanoemulsions produced using 5% oil phase (vitamin E), 5% surfactant phase (Tween 80), 10% ethanol, and 80% buffer solution immediately after formation (“Fresh”) and after storage at different temperatures for one month. (b) Influence of storage temperature on particle size distribution of nanoemulsions produced using 5% oil phase (vitamin E), 5% surfactant phase (Tween 80), 10% ethanol, 0.5% cosurfactant (SDS) and 79.5% buffer solution immediately after formation (“Fresh”) and after storage at different temperatures for one month.
by the size-dependence of droplet dissolution and growth processes), which signifies the beginning of the steady-state regime.33 The PSD then retains this characteristic shape over time, but it moves up the particle size axis as the average size of the droplets continues to increase. In the steady-state regime, the mean droplet radius (r) should vary with time (t) according to the following expression:33 r 3 − r0 3 = ωOR t =
8γVm 2c∞D t 9RT
(1)
Here, r0 is the initial droplet radius, ωOR is the Ostwald ripening rate, γ is the oil−water interfacial tension, Vm is the molar volume of the oil, c∞ is the solubility of the oil in the aqueous phase, D is the diffusion coefficient of the oil molecules through the aqueous phase, R is the gas constant, and T is the 1629
dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
with time, and therefore this data is not shown. The droplet growth rate (ω) for each of the nanoemulsions was calculated from the slope of r3 versus time by linear regression analysis (Table 1). There were linear relationships between r3 and time for VEnanoemulsions that contained no cosurfactant, lauric arginate, and Tween 20 at all storage temperatures, and for the VEnanoemulsions containing SDS at 5 and 20 °C. However, the relationship was not linear for the VE-nanoemulsions containing SDS at 37 °C (Figure 7). It is often assumed that a linear relationship between r3 and time and a time-invariant particle size distribution is indicative of OR, but this is not generally the case.31 Theoretical studies have shown that for certain types of coalescence mechanisms the PSD may also reach a constant shape at longer times, and that the cube of the droplet radius may also increase linearly with time, such as Brownian-motion induced coalescence, e.g., see Coalescence section below.34 At relatively high surfactant concentration (10% Tween 80) in the VE-nanoemulsions studied there may have been appreciable amounts of free surfactant micelles present that could transfer oil molecules between droplets through an Ostwald ripening mechanism35or promote droplet aggregation through a depletion mechanism.36 Further, we experimentally observed that the droplet growth rate increased with increasing temperature, and depended somewhat on cosurfactant addition (Table 1). A number of physical properties of the system may influence the OR rate as the temperature increases. First, the aqueous phase viscosity decreases with increasing temperature (and therefore D increases), which would increase the OR rate. However, this effect would be expected to be fairly small since the shear viscosity of water only changes from about 1.5 to 0.69 mPa s when the temperature is increased from 5 to 37 °C (CRC Handbook of Chemistry and Physics). Second, the interfacial tension of nonionic surfactant monolayers decreases with increasing temperature due to progressive dehydration of the surfactant head groups,37 which should decrease the OR rate. The change in interfacial tension with temperature may be appreciable, particularly near the phase inversion temperature (PIT) where the interfacial tension becomes ultralow, and therefore one would expect this effect to play an important role. Third, the solubility of the vitamin E acetate in the aqueous phase is likely to change with temperature. To our knowledge there are no published data on the solubility of vitamin E acetate in aqueous 20% ethanol solutions at different temperatures. We tried to measure this parameter by mixing excess vitamin E acetate with aqueous 20% ethanol solutions and storing the resulting systems at 5, 20, and 37 °C for 72 h. We then used UV−visible spectrophotometry to try to measure the concentration of vitamin E acetate present in the resulting solutions at saturation, using the published wavelength maximum (λ = 292 nm) and extinction coefficient (ε = 3260 M−1 cm−1) of this compound.38 However, the levels of vitamin
absolute temperature. This theory suggests that the OR rate should increase as the water-solubility of the oil increases, the interfacial tension increases, or the diffusion coefficient increases. Equation 1 indicates that the cube of the mean droplet radius (r3) should increase linearly with time for nanoemulsions undergoing Ostwald ripening. We therefore calculated r3 versus time profiles for nanoemulsions at different storage temperatures, e.g., the data at 20 and 37 °C are shown in Figure 7. At 5 °C, we saw little change in the particle size
Figure 7. Effect of cosurfactant and storage time on droplet radius cubed of nanoemulsion containing 5% oil phase (vitamin E), 5% surfactant phase (Tween 80), 10% cosolvent (ethanol), 0.5% cosurfactant, and 79.5% buffer solution at storage temperatures of 20 °C (a) and 37 °C (b).
Table 1. Calculated Growth Rates of Nanoemulsions Containing Different Cosurfactants (0.5%) at Various Storage Temperatures (Correlation Coefficients Are Shown in Brackets) growth rates (nm3 per day) storage temperature
control
+ LA
+ SDS
+ Tween 20
5 °C 20 °C 37 °C
106 (0.974) 526 (0.999) 3190 (0.998)
91 (0.964) 463 (0.997) 2840 (0.999)
104 (0.994) 656 (0.999) 13600 (0.948)
154 (0.953) 714 (0.995) 5780 (0.999)
1630
dx.doi.org/10.1021/jf404182b | J. Agric. Food Chem. 2014, 62, 1625−1633
Journal of Agricultural and Food Chemistry
Article
linearly with the cube of the droplet radius. For certain initial particle size distributions, this theory also predicts that the PSD should maintain a constant shape at longer storage times, and should just move up the particle size axis with time.34 The fact that changes in the mean particle size and PSD with time may be similar for OR and coalescence means that it is often difficult to distinguish between these two different growth mechanisms. To estimate the potential contribution of coalescence in the nanoemulsions we calculated the droplet growth rate using eq 2 and the physicochemical properties of the system (ϕ = 0.05: η = 0.001 Pa s; η′ = 2.95 Pa s) and assumed that every collision led to coalescence (E = 1). We estimated a droplet growth rate of 2400 nm3/day using eq 2, which is fairly close to the values measured at the higher temperatures. The relatively low growth rates measured at lower temperatures may have been because the collision efficiency was much less than unity, due to the presence of repulsion colloidal interactions operating between the droplets in the nanoemulsions (such as steric or electrostatic repulsion).24 These calculations suggest that coalescence may play an important role in determining the overall rate of droplet growth in the nanoemulsions during storage, particularly at elevated temperatures near the cloud point. Experimentally, we observed that the rate of droplet growth increased appreciably with temperature and depended somewhat on cosurfactant addition (Figure 7, Table 1). Previous studies have shown that the rate of droplet coalescence increases as the temperature moves toward the PIT of the system due to changes in the optimum curvature of the surfactant monolayer at the droplet surfaces.16 When the optimum curvature is close to unity the interfacial tension is very low, which leads to rapid droplet coalescence.26 For example, Shinoda and Saito14 showed that O/W emulsions stabilized by nonionic surfactants were relatively stable to droplet growth at storage temperatures well below the PIT, but were highly prone to coalescence at temperatures near the PIT. One might therefore expect that cosurfactants that increase the PIT would improve the coalescence stability of nanoemulsions at elevated temperatures. In addition, one would expect that cosurfactants that increased the repulsive interactions between droplets (decreased the collision efficiency) would reduce the coalescence rate.34 Nevertheless, we did not find any correlation between increasing effects of cosurfactant on the cloud point (and thereby PIT) and the corresponding storage stability. The origin of this phenomenon is currently unknown, and it would be useful in future studies to identify its molecular basis. In particular, it would be useful to establish the physicochemical origin of the differences in the impact of the different surfactants on the droplet growth rates. In this study, we investigated the effect of cosurfactants on the thermal and isothermal stability of VE-nanoemulsions (containing ethanol as cosolvent). These systems were highly unstable to droplet growth, which was attributed to droplet coalescence and/or Ostwald ripening. Addition of a nonionic cosurfactant (0.5% Tween 20) has little influence on the cloud point of VE-nanoemulsions, however addition of ionic surfactants (0.5% lauric alginate or SDS) caused an appreciable increase in the cloud point. On the other hand, addition of cosurfactants did not increase the long-term stability of VEnanoemulsions under isothermal conditions, particularly at elevated temperatures. Theoretical calculations suggested that both coalescence and Ostwald ripening may be important, with coalescence being particularly important at high storage
E acetate present were so small that we could not measure them, i.e., no maximum was observed at the expected wavelength and the UV-signal was close to the baseline. The solubility of vitamin E acetate in pure water reported in a chemical database (www.scifinder.com) was therefore used to provide a rough prediction of the expected rate of Ostwald ripening in the system: c∞ = 2 × 10−6 kg m−3 or 4.64 × 10−6 mol m−3 at 25 °C. The other parameters needed to calculate the OR rate using eq 1 were also taken from this database or measured in our laboratory: Vm = 0.431 kg mol−1, D = 3.9 × 10−10 m2 s−1, γ = 30 mJ m−2, R = 8.31 J K−1 mol−1, and T = 298.1 K. The droplet growth rate was estimated to be around 4.2 × 105 nm3/day, which is appreciably higher than the measured droplet growth rates particularly at the higher storage temperatures (Table 1). In addition, the above calculations have ignored the influence of ethanol on the predicted growth rate, and studies have shown that the solubility of vitamin E (αtocopherol) increases over 5-fold when the ethanol concentration is increased from 0 to 20% in aqueous ethanol solutions.39 There therefore seems to be a contradiction between the experimental measurements and the theoretical predictions made using the OR theory. There are a number of possible reasons for this discrepancy. First, some of the physicochemical parameters used in the calculations may have been incorrect. In particular, the solubility of the vitamin E acetate in the aqueous phase was predicted using a mathematical model from a chemical database, which may have been incorrect. Second, the influence of surfactant molecules adsorbed to the oil−water interface on the OR rate was ignored. Surfactants are known to decrease the interfacial tension (particularly near the PIT), which would reduce the OR rate considerably.40 Third, there may have been some highly hydrophobic impurities within the vitamin E acetate preparation used in this study that acted as ripening inhibitors.32,40 Coalescence. Coalescence is the process whereby two or more droplets come into close contact and then merge together.16,26 The rate at which this process occurs depends on the frequency and duration of encounters between droplets, as well as the colloidal and hydrodynamic interactions between them. The following expression has been derived to describe the increase in mean droplet radius with time due to Brownianmotion induced coalescence:34 r 3 − r0 3 = ωC t ∧
=
4.258r0 3N0kT (μ + 1)E ∧
t
μ(3μ + 2) ∧
≈
φkT (μ + 1)E ∧
μ(3μ + 2)
t (2)
Here, ωC is the coalescence droplet growth rate, N0 is the total number density of droplets in the initial system, ϕ is the disperse phase volume fraction, μ is the viscosity of the continuous phase, μ̂ is the dispersed-to-continuous phase viscosity ratio (μ′/μ), k is Boltzmann’s constant, and E is the collision efficiency. The collision efficiency depends on the nature of the colloidal and hydrodynamic interactions between the droplets and typically has a value