Vitamin E-enriched nanoemulsions formed by emulsion phase inversion: Factors influencing droplet size and stability

Journal of Colloid and Interface Science 402 (2013) 122–130

Contents lists available at SciVerse ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Vitamin E-enriched nanoemulsions formed by emulsion phase inversion: Factors influencing droplet size and stability Sinja Mayer a, Jochen Weiss a, David Julian McClements b,⇑ a b

Dept. of Food Physics and Meat Science, Institute of Food Science and Biotechnology, University of Hohenheim, 70599 Stuttgart, Germany Biopolymers and Colloids Research Laboratory, Department of Food Science, University of Massachusetts, Amherst, MA 01003, USA

a r t i c l e

i n f o

Article history: Received 5 March 2013 Accepted 4 April 2013 Available online 22 April 2013 Keywords: Emulsions Nanoemulsions Low energy methods Microfluidization Phase inversion Emulsion phase inversion Natural surfactants

a b s t r a c t There is considerable interest in using nanoemulsions as delivery systems for lipophilic bioactive ingredients, such as oil-soluble vitamins. Nanoemulsions can be fabricated using either high-energy or low-energy methods, but the latter offer advantages in terms of low cost, higher energy efficiency, and simplicity of implementation. In this study, the emulsion phase inversion (EPI) method was used to produce food-grade nanoemulsions enriched with vitamin E acetate. The EPI method simply involves titrating water into a mixture containing oil and surfactant, which initially leads to the formation of a water-in-oil emulsion that then inverts into an oil-in-water emulsion. Oil composition, surfactant type, and surfactant-to-oil ratio (SOR) were all found to influence the particle size distribution of the systems produced. Nanoemulsions with a mean particle diameter of 40 nm could be produced at a final system composition of 2 wt% MCT, 8 wt% vitamin E acetate, and 20 wt% Tween 80. The EPI method was shown to be unsuitable for producing nanoemulsions from label-friendly surfactants, such as Quillaja saponin, whey protein, casein, and sucrose monoesters. The EPI method was more effective at producing nanoemulsions at high SOR than microfluidization, but much less effective at low SOR. Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction The food and beverage industries are interested in the development of colloidal delivery systems to encapsulate lipophilic functional ingredients, such as bioactive lipids (e.g. x-3 fatty acids), flavor oils (e.g. orange oil) or oil-soluble-vitamins (e.g. Vitamin E) [1,2]. These lipophilic components have different chemical and physical properties, which means that it is usually necessary to tailor a particular delivery systems for each type of functional ingredient [1]. In this study, we focused on the encapsulation of an oil-soluble vitamin (vitamin E). The term ‘‘Vitamin E’’ actually refers to a group of eight closely related fat-soluble vitamins [3], which includes four tocopherols (a, b, c, and d) and four tocotrienols (a, b, c, and d) [4,5]. All eight molecules have a chromanol ring and a phytol side chain [6]. The form of vitamin E with the highest abundance and biological activity is a-tocopherol [5,7]. This form of vitamin E is a lipophilic antioxidant which scavenges free radicals and protects membrane lipids from oxidation [5,7]. It has been postulated that the antioxidant activity of vitamin E may help prevent chronic diseases, such as cardiovascular diseases, arthrosclerosis and cancer [4]. There is therefore considerable interest in enriching various food, pharmaceutical, and cosmetic products ⇑ Corresponding author. Fax: +1 413 545 1262. E-mail address: [email protected] (D.J. McClements). 0021-9797/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcis.2013.04.016

with vitamin E [3]. Vitamin E (a-tocopherol) is highly susceptible to oxidation due to its polyunsaturated structure and therefore it is usually utilized in its esterified form: Vitamin E acetate (atocopherol acetate). Vitamin E acetate has a low water-solubility and therefore it is necessary to encapsulate it within colloidal delivery systems if it is going to be incorporated into aqueous-based products [8,9]. Emulsion-based delivery systems are particularly suited for the encapsulation of bioactive lipids [10]. The most commonly used emulsion-based delivery systems are emulsions, nanoemulsions and microemulsions, which can be distinguished by their particle dimensions and thermodynamic stability [10,11]. Microemulsions are thermodynamically stable systems that typically have particle radii <100 nm [12]. In contrast, emulsions and nanoemulsions are thermodynamically unstable systems with particle radii >100 nm and <100 nm, respectively. Differences in particle dimensions lead to differences in the functional performance of emulsion-based delivery systems. The smaller droplets in nanoemulsions means that they tend to have higher stability to gravitational separation, flocculation, and coalescence than equivalent conventional emulsions [11,13]. The intensity of light scattering by oil droplets decreases with decreasing droplet size, and so nanoemulsions are less turbid than conventional emulsions with similar oil contents. Finally, the bioavailability of encapsulated compounds has been reported to increase with reduced particle sizes [14–16].

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

Oil-in-water (O/W) emulsions and nanoemulsions can be formed by either high- or low-energy methods [17]. High-energy methods are currently the most widely used homogenization methods in the food industry. These methods utilize mechanical devices (‘‘homogenizers’’) that generate intense forces that disrupt and intermingle oil and water phases, such as microfluidizers, high pressure homogenizers, and sonicators [13,18–20]. These devices are relatively expensive to purchase and operate, and are often prone to mechanical problems [1,18,12,21]. Low-energy methods such as for example the phase inversion temperature (PIT), phase inversion composition (PIC), emulsion phase inversion (EPI) and spontaneous emulsification methods rely on the spontaneous formation of oil droplets in surfactant–oil–water mixtures under specific environmental conditions [17]. The advantage of low-energy approaches is that merely simple stirring is required, which makes these methods more energy efficient [12]. These methods allow energy costs to stay low, as they do not require the use of any expensive equipment [18]. It is, in part, due to this fact that researchers have been working on the design, development and improvement of new low-energy methods for more than 20 years [22]. One has to differentiate between nanoemulsions formed by self-emulsification or spontaneous emulsification methods and the formation by phase inversion [12]. Phase inversion means the conversion of an oil-in-water (O/W) emulsion to a water-in-oil (W/O) emulsion or vice versa [23,24]. Phase inversion approaches include phase inversion temperature (PIT), phase inversion composition (PIC) and emulsion phase inversion (EPI) methods [1]. These methods can be distinguished by the conditions used to spontaneously form the emulsions or nanoemulsions. One of the major objectives of the current study was to investigate the potential of using the EPI method to form emulsions and nanoemulsions containing vitamin E acetate. This method simply involves titrating an aqueous phase into an organic phase containing oil and a hydrophilic surfactant with constant stirring [22,23]. The EPI method is based on a catastrophic phase inversion that occurs after a certain amount of aqueous phase has been added to the organic phase [25]. As increasing amounts of water are added the system transitions from a water-in-oil (W/O) emulsion, to a multiple emulsion (O/W/O), and finally to an oil-inwater (O/W) emulsion [24]. The O/W/O emulsions formed at intermediate water contents is believed to be critical to the formation of small oil droplets in the O/W emulsions [24,26]. This O/W/O emulsion consists of small oil droplets trapped within larger water droplets that are dispersed within an oily continuous phase [10,27]. Another objective of this study was to determine whether stable emulsions and nanoemulsions could be produced using ‘‘label-friendly’’ emulsifiers, rather than the synthetic surfactants normally used to produce low-energy systems. This is important because the food and beverage industry is trying to replace many synthetic ingredients with more ‘‘label-friendly’’ alternatives [28]. We therefore investigated the possibility of using a number of more natural emulsifiers to form emulsions and nanoemulsions using the EPI method, including Quillaja saponin, whey protein, casein, and sucrose monopalmitate [15,28].

2. Materials and methods 2.1. Materials Oil phase: Medium chain triglycerides (MCT) (Miglyol 812, Warner Graham, Cockeysville, MD) and orange oil (SC020295, International Fragrances and Flavors, NJ) were used as carrier oils. Vitamin E acetate (VE) was kindly donated by BASF (Ludwigshafen, Germany).

123

Surfactants: A series of non-ionic surfactants (Tween 20, 40, 60, 80 and 85) were obtained from Sigma–Aldrich (St. Louis, MO). Whey protein isolate (WPI) was supplied by Davisco (Le Sueur, MN), sucrose monopalmitate (SMP) by Compass Foods Company (Singapore), casein (from bovine milk) was obtained from Sigma– Aldrich (St. Louis, MO), and Quillaja saponin (Q-Naturale powder or liquid) were supplied by Natural Starch LLC (Bridgewater, NJ). Water phase: Citric buffer (pH 3.0) was prepared using 0.8 wt% citric acid and 0.08 wt% sodium benzoate obtained from Sigma–Aldrich (St. Louis, MO). Double distilled water was used to prepare all solutions and emulsions. 2.2. Methods 2.2.1. Emulsion preparation 2.2.1.1. Simple mixing. In this approach, surfactant, oil, and water were added to a beaker and then stirred for 60 min on a stirring plate (500 rpm) at ambient temperature (ca. 25 °C). 2.2.1.2. Low energy method. Emulsion phase inversion (EPI) was used as a low-energy method to prepare emulsions and nanoemulsions. Initially, an organic phase containing carrier oil, surfactant and vitamin E was stirred for 30 min (500 rpm). Then the aqueous phase (citric buffer, pH 3) was titrated into the oil phase with constant stirring (500 rpm) at a flow rate of 20 drops per 10 s using a burette. 2.2.1.3. High energy method. Microfluidization was used as a highenergy method to prepare emulsions and nanoemulsions. Oil, surfactant and water were blended together for two minutes using a high-shear mixer to form a coarse emulsion (Barmix, Biospec Products, Bartlesville, OK). The coarse emulsions were then passed four times through a microfluidizer (M-110L, Microfluidcs, Newton, MA) at 12,000 psi to form fine emulsions. 2.2.1.4. Surfactant concentration. The influence of surfactant concentration was investigated by varying the surfactant-to-oil ratio (SOR). The total oil content in the final systems was held constant, while the SOR was varied by altering the amounts of surfactant and water content in the system:

SOR ¼ ms =mo

ð1Þ

mw ¼ 100  mo  ms

ð2Þ

Here, ms, mo and mw are the masses of surfactant, oil and water, respectively. 2.2.2. Particle size measurements Dynamic light scattering: Systems containing relatively small droplets (d < 3000 nm) were analyzed by dynamic light scattering (DLS). The particle size distribution, mean particle diameter, and polydispersity index (PDI) of samples were measured using a commercial DLS instrument (Zetasizer Nano ZS, Malvern Instruments, Malvern, UK). Static light scattering: Systems containing relatively large droplets (d > 1000 nm) were analyzed by static light scattering (SLS). The particle size distribution and mean particle diameter (d32) were measured using a commercial SLS instrument (Mastersizer 2000, Malvern Instruments Ltd., Malvern, Worcestershire, UK). A refractive index of 1.33 was used for the water phase and 1.43 for the oil phase. Samples were diluted in buffer solution prior to analysis to avoid multiple scattering effects. 2.2.3. Shear viscosity measurement Shear viscosity measurements were performed using a dynamic shear rheometer (Kinexus Rotaional Rheometer, Malvern Instru-

124

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

ments, Malvern, UK) equipped with a cup-and-bob measurement cell. All measurements were carried over a range of different shear rates (0.1–100 s1) at 25 °C. The apparent shear viscosity (g) at a fixed shear rate (10 s1) was also reported.

100

2500

SOR 1.0

Viscosity Conductivity Gel

2.2.5. Experimental design All measurements were performed on two or three freshly prepared samples. Then mean and standard deviations were calculated from this data.

2000

1

1500

0.1

1000

0.01

500

0.001

3. Results and discussion 3.1. Emulsion phase inversion point Initially, we examined the phase behavior of surfactant– oil–water (SOW) systems containing Tween 80 as the surfactant (S), MCT as the oil phase (O), and citric buffer solution as the water phase (W). These systems were prepared using different surfactant-to-oil ratios (SOR = 0.1, 0.2, 0.5 and 1) to determine the influence of surfactant-to-oil ratio (SOR) on the formation of small droplets using the EPI method. Oil and surfactant were mixed together and then water was titrated into this mixture while the system was constantly stirred, until the final water content was reached. The appearance, shear viscosity, and electrical conductivity of the samples were measured as increasing amounts of water were added to the system (Fig. 1). All the initial surfactant/oil mixtures had a transparent appearance, a low electrical conductivity (0 lS/cm), and a relatively low shear viscosity (0.03– 0.15 Pa s), with the viscosity increasing with surfactant content. In general, systems with different SOR levels exhibited qualitatively similar behavior as increasing amounts of water were added. Visually, they initially became more turbid, then a white sediment layer formed (presumably water droplets), then a white viscous or gel phase formed, and then a white cream layer formed (presumably oil droplets). The electrical conductivity remained relatively low when small amounts of water were added indicating that the continuous phase was oil, but then it increased steadily once a critical water level was exceeded, which is consistent with a W/O to O/W phase transition [29]. Initially, the shear viscosity increased steeply as the water content increased (due to an increase in the volume of water droplets in the W/O emulsion), then it reached a maximum value, and then it decreased with a further increase in water content (due to a decrease in the volume of oil droplets in the O/W emulsion). The highly viscous or gel-like material observed at intermediate water contents may be attributed to the formation of a liquid crystalline phase [29] and/or a multiple emulsion with a high disperse phase volume fraction. The water content where the electrical conductivity began to increase was somewhat lower than that where the maximum in the shear viscosity occurred, suggesting that the intermediate phase may have had a bicontinuous structure with some interconnected water regions that allowed ions to flow [29]. The change in electrical conductivity and shear viscosity with water content depended on the surfactant-to-oil ratio (Fig. 1). The water content at which the electrical conductivity first in-

Electrical Conductivity (µS/cm)

2.2.4. Conductivity measurements The electrical conductivity was measured to detect the phase inversion region for the O/W to W/O transition [8]. A mixture of oil (MCT), surfactant (Tween 80) and water (citric buffer pH 3) was prepared by simple mixing at ambient temperature. If there was no gel formation, the emulsions were continuously stirred using a magnetic stirrer. Afterwards, their electrical conductivity was measured using a portable conductivity meter (Accumet AP85; Fisher Scientific, Pittsburgh, PA).

Apparent Shear Viscosity (Pa s)

10

0 0

10

20

30

40

50

60

70

80

90

Water Content (%) Water Content:

Fig. 1a. Effect of increasing water concentration (citric buffer pH 3) on the appearance, shear viscosity (at 10 s1) and electrical conductivity of emulsions at a fixed surfactant-to-oil ratio of 1.0.

creased was 20%, 10%, 10% and 10% for SOR = 0.1, 0.2, 0.5, and 1, whereas the water content at which the maximum in the apparent viscosity occurred was 25%, 25%, 32% and 40% for SOR = 0.1, 0.2, 0.5, and 1. In addition, the magnitude of the apparent shear viscosity increased appreciably as the SOR increased, being 0.2, 3, 11 and >40 Pa s for SOR = 0.1, 0.2, 0.5, and 1. Indeed, visual observation of the samples showed that strong gels were formed at intermediate water contents for the mixtures with high SOR, whereas viscous liquids were formed for the systems with low SOR (Fig. 1). We also measured the influence of SOR on the mean droplet diameter of the oil-in-water emulsions/nanoemulsions formed at the end of the EPI titration process (Fig. 2). The size of the droplets produced using the EPI method were compared with those produced by simply mixing the surfactant, oil, and water together at their final concentrations. Our results show that nanoemulsions cannot be formed by simply mixing all the components together, which is in agreement with previous studies [29]. Indeed, all of the emulsions formed by simple mixing contained very large droplets (d > 40 lm) regardless of SOR. On the other hand, nanoemulsions could be formed using the EPI method under certain conditions (Fig. 2): d32  98, 15, 1.5, and 0.15 lm at SOR = 0.1, 0.2, 0.5, and 1, respectively. Thus nanoemulsions (d < 200 nm) could only be formed once a critical surfactant level (SOR = 1) was reached, which is consistent with previous studies [30]. These results suggest that the formation of a gel-like material at intermediate water contents may be important in the formation of fine oil droplets by the EPI method. Presumably, this gel-like material was a liquid crystalline phase or a multiple emulsion (O/W/O) with a high disperse phase volume fraction. Nevertheless, further studies are needed to determine the structural organization of the surfactant, oil, and water molecules within this intermediate

125

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

100

1

2500

SOR 0.5

2500 Viscosity

SOR 0.1

Viscosity

Conductivity

1

1500

0.1

1000

0.01

Apparent Shear Viscosity (Pa s)

2000

2000

0.1 1500

1000 0.01

Electrical Conductivity (µS/cm)

10

Elcetrical Conductivity (µS/cm)

Apparent Shear Viscosity (Pa s)

Conductivity

500

500 0.001

0 0

0.001

0

10

20

30

40

50

60

70

80

90

10

20

30

40

50

60

70

80

90

100

Water Content (%)

0 100

Water Content:

Water Content (%) Water Content:

Fig. 1d. Effect of increasing water concentration (citric buffer pH 3) on the appearance, shear viscosity (at 10 s1) and electrical conductivity of emulsions at a fixed surfactant-to-oil ratio of 0.1. Fig. 1b. Effect of increasing water concentration (citric buffer pH 3) on the appearance, shear viscosity (at 10 s1) and electrical conductivity of emulsions at a fixed surfactant-to-oil ratio of 0.5.

EPI

Simple mixing

2500

10

SOR 0.2

Viscosity Conductivity

Apparent Shear Viscosity (Pa s)

1500 0.1 1000

0.01

Electrical Conductivity (µS/cm)

2000 1

500

0.001 0

10

20

30

40

50

60

70

80

90

Mean Particle Diameter (µm)

100

10

1

0.1 0.1

0.2

0.5

1

SOR Fig. 2. Influence of surfactant-to-oil ratio (SOR) and preparation method (EPI versus simple mixing) on the mean particle diameter of systems containing 10% oil phase (MCT), and varying amounts of surfactant (Tween 80) and water phase (citrate buffer pH 3).

0 100

Water Content (%) Water Content:

phase, and to determine the relationship between this structure and the size of the droplets produced in the O/W emulsions after phase inversion.

3.2. Influence of oil phase composition on particle size

Fig. 1c. Effect of increasing water concentration (citric buffer pH 3) on the appearance, shear viscosity (at 10 s1) and electrical conductivity of emulsions at a fixed surfactant-to-oil ratio of 0.2.

In this series of experiments we examined the influence of oil phase composition (vitamin E acetate-to-MCT ratio) on the formation and stability of nanoemulsions produced using the EPI method. This information is important for the rational design of

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

3.3. Influence of SOR on particle size In this series of experiments we investigated the impact of surfactant-to-oil ratio (SOR) on the droplet size of vitamin-enriched nanoemulsions produced by the EPI method. The total oil concentration was kept constant at 10 wt% (2% MCT + 8% VE), but the surfactant (Tween 80) and water phase (citric buffer, pH 3.0) ratio was varied. The mean particle diameter decreased and the polydispersity increased when the SOR was raised from 0.5 to 2.0 (Fig. 4a). The main reason for the observed increase in polydispersity at high surfactant concentrations was the formation of a bimodal distribution, with a population of small droplets around 24 nm and another population of large droplets around 190 nm (Fig. 4b). The smallest droplet diameter (d  40 nm) was determined for the system with the highest surfactant concentration

0.6 120

Mean Particle Diameter (nm)

Diameter 100

0.5

PDI

0.4

80

0.3

60

Polydispersity Index

nanoemulsion-based delivery systems for lipophilic nutraceuticals, such as oil-soluble vitamins. For these experiments, we used 10 wt% surfactant (Tween 80), 10 wt% oil phase (MCT + VE), and 80 wt% water phase (citric buffer, pH 3). Nanoemulsions were produced using the EPI method by stepwise addition of the water phase to the surfactant/oil mixture with constant stirring. Nanoemulsions (d < 200 nm) could be formed at all oil compositions using the EPI method, however, the droplet size produced depended strongly on oil phase composition (Fig. 3). The smallest droplets were formed in the emulsions containing 2 wt% MCT and 8 wt% VE (d = 88 nm; PDI = 0.2), whereas larger and more polydisperse droplets were formed at higher or lower vitamin E levels. These results are in agreement with a recent study where droplets were produced using an alternative low-energy method: spontaneous emulsification [31]. Spontaneous emulsification (SE) involves adding a surfactant/oil mixture to water, rather than adding water to a surfactant/oil mixture as in the EPI method. In the SE study, a minimum droplet size (d = 55 nm; PDI = 0.12) was also observed at 2 wt% MCT and 8 wt% vitamin E acetate (10 wt% Tween 80). This suggests that the physicochemical mechanisms underlying droplet formation using the SE and EPI methods may be similar (Section 3.7).

0.2

40

0.1

20 0.5

1.0

1.5

2.0

SOR Fig. 4a. Effect of surfactant-to-oil ratio (SOR) on mean particle diameter and poydispersity index of nanoemulsions produced by the EPI method. Nanoemulsions were prepared using 10 wt% oil (MCT + VE), 5–20 wt% surfactant (Tween 80) and 85–70 wt% water phase (citric buffer pH 3).

70

SOR 2.0

60 50

Intensity (%)

126

1.5

40 30

1.0

20 145 0.25 10

0.5

0 0.23 125

115

0.21

10

Polydispersity Index

Mean Particle Diameter (nm)

135

105 0.19 95

Diameter PDI 0.17

85 0

2

4

6

8

10

% Vitamin E Acetate in Emulsion (w/w) Fig. 3. Effect of oil composition on mean particle diameter and poydispersity index of emulsions produced by EPI method. Emulsions were prepared using 10 wt% oil (MCT + VE), 10 wt% surfactant (Tween 80) and 80 wt% water (citric buffer pH 3).

100

1000

10000

Particle Diameter (nm) Fig. 4b. Effect of surfactant-to-oil ratio (SOR) on particle size distribution of emulsions produced by EPI method. Emulsions were prepared using 10 wt% oil (MCT + VE), 5–20 wt% surfactant (Tween 80) and 85–70 wt% water phase (citric buffer pH 3).

(SOR = 2.0). The majority of the oil droplets in this nanoemulsion were so small compared to the wavelength of light (k = 390– 750 nm) that they did not scatter light strongly and so the nanoemulsion was optically transparent (Fig 4a, inset). Other studies have also reported a decrease in droplet diameter with increasing surfactant concentration for nanoemulsions prepared using the EPI method [1,24]. The formation of O/W/O multiple emulsions at intermediate water contents has previously been reported to be an essential step in the production of O/W nanoemulsions containing small droplets [1,24]. The internal oil droplets within the O/W/ O emulsion are released when the system breaks down and form the oil droplets in the final O/W emulsion. Previous studies suggest

127

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

that high surfactant concentrations are needed to form these O/W/ O emulsions [24]. A possible reason for this phenomenon is that a process similar to that which occurs in spontaneous emulsification leads to the formation of fine oil droplets at the oil–water boundary in the multiple emulsions (Section 3.7). 3.4. Influence of synthetic surfactant type on particle size The purpose of these experiments was to examine the influence of a series of food-grade synthetic non-ionic surfactants on the formation of nanoemulsions by the EPI method (Tweens). The overall composition of the different systems was kept constant, but the surfactant type was varied: 10 wt% oil phase (2 wt% MCT + 8 wt% VE); 10% surfactant (Tween 20, 40, 60, 80, or 85); and 80% water phase (citric buffer, pH 3.0). The chemical formulae and Hydrophile–Lipophile Balance (HLB) numbers of these surfactants are

shown in Table 1. It has been reported that HLB numbers between 11 and 16 are optimum for stabilization of oil-in-water emulsions [24]. The smallest mean particle diameter was obtained in the systems prepared using Tween 80: d  87 nm, PDI = 0.21 (Fig. 5). Nanoemulsions with narrow size distributions could also be produced using Tween 40 (d  107 nm, PDI = 0.19) and Tween 60 (d  160 nm, PDI = 0.21), but the droplets formed were appreciably larger than those produced using Tween 80. Emulsions produced using Tween 85 contained the largest droplets (d  530 nm) and had the widest size distributions (PDI = 0.77). We did not find a correlation between the mean droplet diameter produced using the EPI method and the surfactant HLB number, e.g. Tween 80 and Tween 60 had similar HLB numbers but gave different droplet sizes. Recent studies using spontaneous emulsification (SE) to form nanoemulsions also reported that there was not a close correlation

Table 1 Chemical formula, HLB numbers and structures of synthetic non-ionic surfactants used in this study. Tween 20 and 40 have similar structures to Tween 60, but a shorter saturated tail. The chemical structures are highly schematic. Non-ionic surfactant

Chemical structure

HLB

TWEEN 20 TWEEN 40 TWEEN 60

Polyoxyethylen-20-sorbitan-monolaurate Polyoxyethylen-20-sorbitan-moopalmitate Polyoxyethylen-20-sorbitan-monostearate

16.7 15.5 14.9

TWEEN 80

Polyoxyethylen-20-sorbitan-monooleate

15

TWEEN 85

Polyoxyethylen-20-sorbitan-trioleate

11

128

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

600

200

Diameter

MCT

0.8

Orange oil

0.6

400

300 0.4 200

Particle Diameter (µm)

PDI

Polydispersity Index

Mean Particle Diameter (nm)

500

150

100

50 0.2 100

0 0

0

Q-Naturale Q-Naturale (liquid) (powder)

Tween 20 Tween 40 Tween 60 Tween 80 Tween 85

between droplet size and HLB number [31]. These findings suggest that other surfactant properties must be important in determining their effectiveness at forming nanoemulsions using low energy methods. The molecular geometry of surfactants is one of the most important parameters that may influence their ability to form small droplets [32]. The molecular geometry of a surfactant is characterized by the packing parameter (p), which is the ratio of the tail group area to the head group area: p = aT/aH. Differences in the packing of surfactants at oil–water boundaries influence interfacial properties such as surface tension and mobility, which are likely to play an important role in the spontaneous formation of oil droplets using low energy methods. Tween 60 has a saturated linear chain (C18:0), whereas Tween 80 has an unsaturated kinked one (C18:1). Thus, Tween 80 would be expected to have a higher packing parameter (due to a larger aT) than Tween 60. On the other hand, Tween 80 has a single unsaturated tail (monooelate) whereas Tween 85 has three unsaturated tails (triooleate) and so one would expect Tween 80 to have a smaller packing parameter (due to a lower aT) than Tween 85. Presumably, there is an optimum surfactant geometry required to promote the spontaneous formation of ultrafine droplets by low energy methods.

3.5. Effect of ‘‘label-friendly’’ surfactant type on particle size The food and beverage industry is trying to replace synthetic surfactants with more ‘‘label-friendly’’ alternatives [33,34]. We therefore examined whether various ‘‘label-friendly’’ surfactants could be used to form nanoemulsions using the EPI method: Q-Natural (liquid form); Q-Natural (powder form); whey protein isolate (WPI); casein; and sucrose monopalmitate (SMP). Otherwise, the overall composition of the different systems was kept constant: 10 wt% oil phase (2 wt% MCT + 8 wt% VE); 10% surfactant; and 80% water phase (citric buffer, pH 3.0). These systems were produced using the same EPI method used to produce nanoemulsions from synthetic surfactants, i.e., titration of the water phase into the surfactant/oil mixture with constant stirring at ambient temperature.

Casein

SMP

Fig. 6. Effect of natural surfactant type on mean particle diameter and polydispersity index of emulsions produced by EPI method. Emulsion were prepared using 10 wt% oil (MCT or orange oil), 10 wt% surfactant and 80 wt% water (citric buffer pH 3).

Microfluidizer

100000

EPI

Mean Particle Diameter (nm)

Fig. 5. Effect of synthetic surfactant type on mean particle diameter and polydispersity index of emulsions produced by EPI method. Emulsion were prepared using 10 wt% oil (MCT + VE), 10 wt% surfactant and 80 wt% water (citric buffer pH 3).

WPI

10000

1000

100

0

0.25

0.5

0.75

1

SOR Fig. 7. Effect of surfactant-to oil ratio (SOR) on the mean particle diameter of emulsions produced by the microfluidization method. Emulsions were prepared using 10 wt% oil (2 wt% MCT + 8 wt% VE), 1–10 wt% surfactant (Tween 80), and 80– 89% water phase (citric buffer pH 3). The diameter was measured by SLS for the EPI method, and DLS for the microfluidization method.

The influence of surfactant type on the size of the droplets produced by the EPI method is shown in Fig. 6. None of the surfactants tested was able to form nanoemulsions (d < 200 nm) using the EPI method. Indeed, the mean droplet diameters were all greater than 10 lm, which accounts for the very poor stability of these systems to gravitational separation (Fig. 6, inset). This result suggests that these surfactants did not have the molecular or physicochemical characteristics required to promote the spontaneous formation of oil droplets at oil–water boundaries, e.g., solubility characteristics, molecular geometry, interfacial mobility, or interfacial tension. In particular, all of the surfactants tested had poor oil solubility, and tended to form opaque colloidal suspensions when they were mixed with oil, rather than transparent solutions.

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

129

Fig. 8. Schematic diagram of the physicochemical events leading to the formation of small droplets in the emulsion phase inversion method.

3.6. Comparison of low-energy and high-energy homogenization methods In this section we compare the effectiveness of the low energy method (EPI) with a high-energy method (microfluidization) commonly used for producing nanoemulsions. The microfluidization method involved homogenizing coarse emulsions containing 10 wt% oil phase (2% MCT + 8% VE) and different Tween 80 concentrations (1, 2, 5, and 10 wt%). The influence of SOR on the mean droplet diameters for the emulsions produced using microfluidization and EPI were compared (Fig. 7). There was a decrease in mean droplet diameter with increasing SOR for both methods. However, the microfluidizer was able to produce relatively small droplets at low surfactant levels (d  280 nm at SOR = 0.1), whereas the EPI method was not (d > 100,000 nm at SOR = 0.1). On the other hand, both methods produced relatively small droplets at high surfactant levels (d < 160 nm at SOR = 1). This result suggests that high-energy methods are the only ones appropriate for producing nanoemulsions when low levels of surfactants are required, but either high- or low-energy methods can be used if high surfactant-tooil ratios are not a problem. Previous studies have also reported that high energy methods require a much lower surfactant concentration to produce fine emulsions than low energy methods [1,20]. The major advantage of using low energy methods for this purpose is that they are more energy efficient, inexpensive, and simple to implement than high energy methods. At the highest surfactant concentration studied (SOR = 1), the mean droplet diameter measured by dynamic light scattering was appreciably smaller for the EPI method (d = 90 nm, PDI = 0.22) than for microfluidization (d = 159 nm, PDI = 0.22). A possible reason for this observation is that vitamin E acetate is highly viscous (300 Pa s) and it is therefore difficult to disrupt by mechanical forces within a homogenizer [35]. It should be noted that there were some differences in the mean particle diameters measured by dynamic light scattering (DLS) and static light scattering (SLS). For example, the mean particle diameter in the EPI nanoemulsions (SOR = 1) was determined to be 90 nm by DLS but 147 nm by SLS. These variations can be attributed to differences in the physical principles underlying particle detection in the two particle sizing instruments. Dynamic light scattering is sensitive to particles from about 3 nm to 3 lm, whereas static light scattering is sensitive to particles from about 100 nm to 1000 lm. Consequently, some of the smaller droplets in the EPI nanoemulsions may not have been detected by SLS. 3.7. Similarity of spontaneous emulsification and EPI methods The influence of system composition on the droplet size of nanoemulsions produced using the EPI method in this study closely corresponds to that obtained using the SE method in a recent

study using a similar system: vitamin E acetate, MCT, and Tween 80 [31]. In both studies, a minimum droplet size was obtained at the same intermediate oil phase composition: 80% vitamin E acetate + 20% MCT. In addition, both studies showed that the mean droplet diameter decreased with increasing SOR, and that Tween 80 produced the smallest droplet sizes. These results suggest that there is a common physicochemical mechanism underlying both of these low energy methods. One possibility is that small oil droplets are spontaneously formed at the internal side of the oil–water boundary in the W/O emulsions initially created using the EPI method, leading to the formation of O/W/O emulsions (Fig. 8). In the spontaneous emulsification method the surfactant/oil mixture is progressively added to the water phase, but in the EPI method the opposite occurs. However, in both cases there will be a boundary formed between a water phase and an oil phase that initially contains the hydrophilic surfactant. Tiny oil droplets are then spontaneously formed when the hydrophilic surfactant moves out of the oil phase and into the water phase. In the EPI method this phenomenon will occur at the internal boundary of the water droplets in the W/O emulsion leading to the formation of an O/W/O emulsion (Fig. 8). This would suggest that small droplets can only be formed using the EPI method under conditions where they would also be formed using the SE method. Interestingly, we found that smaller droplets were formed by the SE method (d = 55 nm; PDI = 0.12) [31] than by the EPI method (d = 88 nm; PDI = 0.2) for systems with similar compositions (2 wt% MCT, 8 wt% VE, 10% Tween 80), which suggests that the SE method may be more suitable for nanoemulsion production. It should be noted that the mechanism proposed in Fig. 8 is likely to be over simplistic, since it does not take into account the formation of a bicontinuous microemulsion or lamellar liquid crystalline phases at intermediate SOR. 4. Conclusions In this study, we examined some of the major factors influencing the formation of vitamin-enriched nanoemulsions by a low-energy method: Emulsion phase inversion (EPI). The surfactant-to-oil ratio (SOR) was shown to play an important role in the phase inversion process, leading to changes in appearance, rheology, electrical conductivity, and droplet size of the SOW systems formed by titration of water into a surfactant/oil mixture. In addition, we showed that the size of the droplets produced using the EPI method was much smaller than that produced by simple mixing at high SOR levels. The composition of the surfactant–oil–water systems played a critical role in determining the size of the droplets in vitamin-enriched nanoemulsions produced using the EPI method. At a fixed surfactant level (SOR = 1), the smallest oil droplets (d  90 nm) were produced when the emulsions contained 8% vitamin E acetate

130

S. Mayer et al. / Journal of Colloid and Interface Science 402 (2013) 122–130

and 2% MCT, whereas larger droplets were formed at higher or lower vitamin levels. The droplet size decreased with increasing SOR, and optically transparent vitamin-enriched nanoemulsions with small droplet sizes (d = 41 nm) could be formed at sufficiently high surfactant levels (SOR = 2). From a group of synthetic non-ionic surfactants (Tween 20, 40, 60, 80, 85) tested, we found that Tween 80 gave the smallest droplets using the EPI method, which was attributed to its specific molecular characteristics, e.g., tail group and head group dimensions. It was not possible to form nanoemulsions using the EPI method from label-friendly surfactants (Quillaja saponin, sucrose monopalmitate, casein, or WPI), which was attributed to their poor oil solubility and interfacial characteristics. Finally, we compared the effectiveness of the EPI method with a commonly used high energy homogenization method (microfluidization). The high energy method could produce much smaller droplets at low surfactant levels (SOR < 1), but the EPI method could produce smaller droplets at higher surfactant levels. We also proposed a physicochemical mechanism to account for the close correlation between the droplet size produced by the EPI method and those produced using the spontaneous emulsification method. Our results will be useful for the rational design of nanoemulsion-based delivery systems to encapsulate oil-soluble vitamins in food, beverage, cosmetic, and pharmaceutical products. Acknowledgments This material is partly based upon work supported by grants from the United States Department of Agriculture (CREES, NRI Grant and NSF-EPI-USDA Grant) and the Massachusetts Department of Agricultural Resources. We also acknowledge funding from the University of Massachusetts (Hatch). We want to thank Amir Hossein Saberi for his support and contribution in theoretical and practical affairs. We also want to acknowledge Ernest Solvay Foundation for providing funding for Sinja Mayer.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

F. Ostertag, J. Weiss, D.J. McClements, J. Colloid Interface Sci. 388 (2012) 95. J. Rao, D.J. McClements, Food Hydrocolloids 29 (2012) 326. Y.C. Chiu, W.L. Yang, Colloids Surf., A 63 (1992) 311. R. Brigelius-Flohe, M.G. Traber, FASEB J. 13 (1999) 1145. M.G. Traber, L. Packer, Am. J. Clin. Nutr. 62 (1995) 1501. M. Gonnet, L. Lethuaut, F. Boury, J. Controlled Release 146 (2010) 276. Z. Cordero, D. Drogan, C. Weikert, H. Boeing, Crit. Rev. Food Sci. Nutr. 50 (2010) 420. J. Feng, Z. Wang, J. Zhang, Z. Wang, F. Liu, Colloids Surf., A 339 (2009) 1. T. Julianto, K.H. Yuen, a.M. Noor, Int. J. Pharm. 200 (2000) 53. D.J. McClements, E.a. Decker, J. Weiss, J. Food Sci. 72 (2007) R109. D.J. McClements, Soft Matter 7 (2011) 2297. C. Solans, I. Solé, Curr. Opin. Colloid Interface Sci. 17 (2012) 246. T. Tadros, P. Izquierdo, J. Esquena, C. Solans, Adv. Colloid Interface Sci. 108–109 (2004) 303. E. Acosta, Curr. Opin. Colloid Interface Sci. 14 (2009) 3. Q. Huang, H. Yu, Q. Ru, J. Food Sci. 75 (2010) R50. D.J. McClements, Ther. Delivery 3 (2012) 801. D.J. McClements, J. Rao, Crit. Rev. Food Sci. Nutr. 51 (2011) 285. N. Anton, J. Benoit, P. Saulnier, J. Controlled Release 128 (2008) 185. U. Einhorn-Stoll, M. Weiss, H. Kunzek, Nahrung 46 (2002) 294. Y. Yang, C. Marshall-Breton, M.E. Leser, A.A. Sher, D.J. McClements, Food Hydrocolloids 29 (2012) 398. D. Werner, Chem. Eng. Process. 34 (1995) 205. N. Anton, T.F. Vandamme, Int. J. Pharm. 377 (2009) 142. R.K. Thakur, C. Villette, J.M. Aubry, G. Delaplace, Colloids Surf., A 315 (2008) 285. F. Jahanzad, G. Crombie, R. Innes, S. Sajjadi, Chem. Eng. Res. Des. 87 (2009) 492. S. Sajjadi, Langmuir 22 (2006) 5597. S. Sajjadi, F. Jahanzad, M. Yianneskis, B.W. Brooks, Ind. Eng. Chem. Res. 42 (2003) 3571. N. Garti, C. Bisperink, Interface Sci. 3 (1998) 657. Y. Yang, D.J. McClements, Food Hydrocolloids 30 (2013) 712. A. Forgiarini, J. Esquena, C. Gonza, C. Solans, Langmuir 17 (2001) 2076. P. Fernandez, V. André, J. Rieger, A. Kühnle, Colloids Surf., A 251 (2004) 53. A.H. Saberi, Y. Fang, D.J. McClements, J. Colloid Interface Sci. 391 (2013) 95. J. Israelachvili, Y. Min, M. Akbulut, A. Alig, G. Carver, W. Greene, K. Kristiansen, E. Meyer, N. Pesika, K. Rosenberg, H. Zeng, Rep. Prog. Phys. 73 (2010). G. Özlem, G. Mazza, Crit. Rev. Food Sci. Nutr. 47 (2007) 231. C. Qian, D.J. McClements, Food Hydrocolloids 25 (2011) 1000. K. Ziani, Y. Fang, D.J. McClements, Food Chem. 134 (2012) 1106.