W

Oct 29, 2010 - Phone: (860) 486-3760. Fax: (860) ... Formation of a normal (not temporary) W/O/W multiple emulsion via the one-step method as a result...
0 downloads 3 Views 784KB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Relationship between Rheological Properties and One-Step W/O/W Multiple Emulsion Formation Jacqueline M. Morais,†,‡ Pedro A. Rocha-Filho,‡ and Diane J. Burgess*,† †

Department of Pharmaceutical Sciences, School of Pharmacy, University of Connecticut, Storrs, Connecticut 06269, United States, and ‡Department of Pharmaceutical Sciences, College of Pharmaceutical Sciences of Ribeir~ ao Preto, University of S~ ao Paulo, Ribeir~ ao Preto, 14.040-903, Brazil Received May 14, 2010. Revised Manuscript Received October 11, 2010 Formation of a normal (not temporary) W/O/W multiple emulsion via the one-step method as a result of the simultaneous occurrence of catastrophic and transitional phase inversion processes has been recently reported. Critical features of this process include the emulsification temperature (corresponding to the ultralow surface tension point), the use of a specific nonionic surfactant blend and the surfactant blend/oil phase ratio, and the addition of the surfactant blend to the oil phase. The purpose of this study was to investigate physicochemical properties in an effort to gain a mechanistic understanding of the formation of these emulsions. Bulk, surface, and interfacial rheological properties of adsorbed nonionic surfactant (CremophorRH40 and Span80) films were investigated under conditions known to affect W/O/W emulsion formation. Bulk viscosity results demonstrated that CremophorRH40 has a higher mobility in oil compared than in water, explaining the significance of the solvent phase. In addition, the bulk viscosity profile of aqueous solutions containing CremophorRH40 indicated a phase transition at around 78 ( 2 °C, which is in agreement with cubic phase formation in the Winsor III region. The similarity in the interfacial elasticity values of CremophorRH40 and Span80 indicated that canola oil has a major effect on surface activity, showing the significance of vegetable oil. The highest interfacial shear elasticity and viscosity were observed when both surfactants were added to the oil phase, indicating the importance of the microstructural arrangement. CremophorRH40/Span80 complexes tended to desorb from the solution/solution interface with increasing temperature, indicating surfactant phase formation as is theoretically predicted in the Winsor III region. Together these interfacial and bulk rheology data demonstrate that one-step W/O/W emulsions form as a result of the simultaneous occurrence of phase-transition processes in the Winsor III region and explain the critical formulation and processing parameters necessary to achieve the formation of these normal W/O/W emulsions.

1. Introduction Multiple emulsions are complex heterogeneous systems in which simple emulsions (O/W or W/O) are further dispersed in oil or water to form W/O/W or O/W/O emulsions, respectively, in the presence of two stabilizing surfactants, one hydrophilic and the other lipophilic.1-6 The surfactants rearrange at the interface of the dispersed droplets to expose their hydrophilic moieties to the water phase and their hydrophobic moieties to the oil phase.7,8 A novel one-step method for the formation of W/O/W multiple emulsions containing nonionic surfactants (CremophorRH40 and Span80) and canola oil has recently been reported.6,9,10 Stable and normal multiple emulsions formed only at 78 ( 2 °C, corresponding to the ultralow surface tension point.6 Accordingly, further physicochemical characterization of this unique emulsification process is warranted. The interfacial properties of fluid interfaces are intimately linked to the kinetics of film formation and rupture during the *Corresponding author. Phone: (860) 486-3760. Fax: (860) 486-4998. E-mail: [email protected]. (1) Florence, A. T.; Whitehill, D. Int. J. Pharm. 1982, 11, 277. (2) Fox, C. Cosmet. Toiletries 1986, 101, 101. (3) Morrison, I. D.; Ross, S. Colloidal Dispersions: Suspensions, Emulsions, and Foams; John Wiley & Sons: New York, 2002; p 420. (4) Vasudevan, T. V.; Naser, M. S. J. Colloid Interface Sci. 2002, 256, 208. (5) Vasiljevic, D.; Vuleta, G.; Primorac, M. Int. J. Cosmet. Sci. 2005, 27, 81. (6) Morais, J. M.; Rocha-Filho, P. A.; Burgess, D. J. Langmuir 2009, 25, 7954. (7) Myers, D. Surfactant Science and Technology; VCH: New York, 1988; Vol. 23, p 81. (8) Holmberg, K.; J€onsson, B.; Kronberg, B.; Lindmam, B. Surfactants and Polymers in Aqueous Solution, 2nd ed.; John Wiley & Sons: New York, 2002; p 451. (9) Morais, J. M.; Santos, O. D. H.; Nunes, J. R. L.; Zanatta, C. F.; RochaFilho, P. A. J. Dispersion Sci. Technol. 2008, 29, 63. (10) BR Patent, no. 0180700074452, Rio de Janeiro, Brazil, Nov. 2007.

17874 DOI: 10.1021/la103358n

emulsification process.10,11 These properties include the (i) surface activity; (ii) surface concentration; (iii) lateral mobility of adsorbed molecules; (iv) interactions between adsorbed molecules; (v) ability of the molecules to change conformation, (vi) film structure; (vii) film thickness; and (viii) film topography.13-16 All of these properties affect the interfacial rheology, whereas some also affect the interfacial tension and charge. Interfacial rheology is closely related to film structure and is affected by processes (such as temperature, surface-active molecule concentration, and pressure) that disturb film equilibrium.13,15,17 The composition and structure of adsorbed surfactant layers also determine the behavior and stability of emulsions.11,12,18-21 These provide stabilization via the fluidic (Gibbs-Marangoni), structural-mechanical, and electrostatic mechanisms.15,22 The ability of the liquid film to recover from deformation upon collision can (11) Tadros, T. F. Colloids Surf., A 1994, 91, 39. (12) Tadros, T. F. Adv. Colloid Interface Sci. 2004, 108-109, 227. (13) Van den Tempel, M. J. Non-Newtonian Fluid Mech. 1977, 2, 205. (14) Warburton, B. In Techniques in Rheological Measurements; Collyer, A. A., Ed.; Chapman & Hall: London, 1993. (15) Wild, P. Interfacial Rheology in Industry; SISC - Interfacial Rheology & Industry, Camtel Ltd.: Royston, England, 2007; p 1. (16) Shaw, M. T.; MacKnight, W. J. In Introduction to Polymer Viscoelasticity, 3rd ed.; John Wiley & Sons: New York, 2005; p 7. (17) Bos, M. A.; van Vliet, T. Adv. Colloid Interface Sci. 2001, 91, 437. (18) Junginger, H. E. In Emulsions: A Fundamental and Practical Approach; Sjoblom, J., Ed.; NATO ASI Series C; Kluwer Academic Publishers: Dordrecht, The Netherlands,1992; Vol. 363, p 189. (19) Barnes, H. A. Colloids Surf., A 1994, 91, 89. (20) Opawale, F. O.; Burgess, D. J. J. Pharm. Pharmacol. 1998, 50, 965. (21) Opawale, F. O.; Burgess, D. J. J. Colloid Interface Sci. 1998, 197, 142. (22) Schramm, L. L.; Stasiuk, E. N.; Marangoni, D. G. Annu. Rep. Prog. Chem., Sect. C 2003, 99, 3.

Published on Web 10/29/2010

Langmuir 2010, 26(23), 17874–17881

Morais et al.

determine the coalescence rate and therefore indicates the long-term stability of the system. To achieve long-term stability, the film must diminish external disturbances and resist rupture. For this, the important qualities are the chemical composition and the structural organization of the film because these dictate the behavior under stress as well as the electrostatic repulsion. In addition, excess surfactant molecules in the bulk assist stabilization via the Gibbs-Marangoni effect.11,22 Surface or interfacial rheology defines the relationship among the applied stress, deformation, and rate of deformation at a fluid interface.19,23 There are two primary methods of measuring the interfacial rheological properties of adsorbed (soluble) and deposited (insoluble) layers of surface-active materials at an interface: (i) shear rheology and (ii) dilatational rheology.24-26 Dilatational interfacial rheology is sensitive to the adsorption/ desorption kinetics of emulsification. Interfacial shear rheology is useful in the study of adsorbed layers of polymers and mixtures of polymers and surfactants as well as insoluble monolayers and generates information about interactions between molecules at the interface that are not apparent from dilatational rheology.24 Interfacial rheology provides a quantitative characterization of both the elastic (G0 ) and viscous (G00 ) components of the film.13,14,27 Interfacial elasticity reflects the ability of the film to restore its interfacial tension after stress and thus is a measure of the capacity of the layer to resist deformation. Interfacial viscosity is a measure of the speed of the relaxation process that restores equilibrium after a disturbance and hence represents the capacity of the layer to adapt to deformation and the resistance of a material to flow under stress.5,16,17,25,28 Viscoelasticity is a combination of viscous and elastic properties in a material with the relative combination of each being dependent on the time, temperature, stress, and strain rate. In purely elastic materials, the stress and strain occur in phase; therefore, the response of one occurs simultaneously with the other. However, in purely viscous materials, there is a phase difference between stress and strain, where strain lags stress by a 90° (π/2 radians) phase lag.16 Temperature (thermal motion) contributes to the deformation of polymeric materials, favoring some conformations over others. Accordingly, viscoelastic properties change with changes in temperature.16,29 Another important effect of temperature on polymeric materials is changes in the hydrophilicity of the polymer backbone. Polymers containing PEO groups are particularly susceptible to changes in hydrophilicity as a result of temperature change.30,31 The surface viscoelasticity of polymeric films is known to be effected by compound hydrophilicity.29 The aim of this work was to evaluate bulk and surface and interfacial rheological properties of surfactant solutions containing CremophorRH40 and Span80 (alone or as a blend) as a function of temperature and hydrophilic surfactant concentration in order to understand the system microstructure during the emulsification process (i.e., surfactant molecular rearrangement and possible inter- and intramolecular interactions). Bulk viscosity was investigated in order to determine whether any viscosity changes occurred (23) Idson, B. Cosmet. Toiletries 1978, 93, 23. (24) Bantchev, G. B.; Schwartz, D. K. Langmuir 2003, 19, 2673. (25) Maldonado-Valderrama, J.; Martı´ n-Rodriguez, A.; Galvez-Ruiz, M. J.; Miller, R.; Langevin, D.; Cabrerizo-Vı´ lchez, M. A. Colloids Surf., A 2008, 323, 116. (26) Maestro, A.; Ortega, F.; Monroy, F.; Kr€agel, J.; Miller, R. Langmuir 2009, 25(13), 7393. (27) Py, C.; Rouviere, J.; Loll, P.; Taelman, M. C.; Tadros, T. F. Colloids Surf., A 1994, 91, 215. (28) Lippacher, A.; Muller, R. H.; Mader, K. Eur. J. Pharm. Biopharm. 2004, 58, 561. (29) Kim, C.; Yu, H. Langmuir 2003, 19, 4460. (30) Saito, H.; Shinoda, K. J. Colloid Interface Sci. 1967, 24, 10. (31) Shinoda, K.; Friberg, S. E. Emulsions and Solubilization; John Wiley & Sons: New York, 1986.

Langmuir 2010, 26(23), 17874–17881

Article Table 1. Evaluation of the Bulk Viscosity of Surfactant Solutions as a Function of Temperature canola oil þ Span80 (24.2% w/w) canola oil þ CremophorRH40 (25.8% w/w) canola oil þ CremophorRH40 Span80 (50.0% w/w) water þ CremophorRH40 (25.8% w/w)

around the critical emulsification temperature. This information is used to facilitate an understanding of the one-step W/O/W process.

2. Experimental Section 2.1. Chemicals. Surfactant solutions were prepared using the following components: canola oil (commercial grade) purchased from Bunge (St. Louis, MO); PEG-40 hydrogenated castor oil (MW = 2589) (CremophorRH40) kindly donated by BASF (Florham Park, NJ); PEG octylphenyl ether (MW = 647) (TritonX100), PEG20 sorbitan monooleate (MW=1310) (Tween80) and sorbitan monooleate (MW=428.61) (Span80) purchased from Sigma (St Louis, MO); and distilled and deionized water (Nanopure Ultrapure, Barnstead). 2.1.i. Bulk Viscosity Study. The bulk viscosity values of the surfactant solutions were determined as a function of temperature using a nonlinear method. A rotational rheometer (DV-III, Brookfild Instruments Stoughton, MA) with a 00 cylindrical spindle and UL adapter for small samples was used. Temperature control was achieved using a circulating water bath with ramp function control (Haake Phoenix, P1 B5, ThermoFisher Scientific, Waltham, MA). A total of 15 g of sample was prepared. The total amount of surfactant blend (CremophorRH40 and Span80) was kept at 50.0% (w/w). CremophorRH40 was used at 25.8% (w/w) and Span80 was used at 24.2% (w/w) in order to reproduce the HLB value of the W/O/W multiple emulsions (9.3). All measurements were conducted over the temperature range of 25 to 85 ( 0.5 °C with a temperature rate increment of 1 °C/min. The viscosity values were recorded after each ∼5 °C increase. Before each time point measurement, the sample temperature was verified using an external thermometer. The experiments were performed only after the sample temperatures were in agreement with the water bath temperatures ((0.5 °C between the water bath and sample was considered to be appropriate). To determine the flow properties of the system, preliminary flow curve were constructed for all samples in the critical temperature range (78 ( 2 °C). The shear rate value used for all of the oily solutions was 6.12/s (5 rpm). For an aqueous solution containing CremophorRH40, a 15.91/s shear rate (13 rpm) was used. The solutions investigated are described in Table 1. 2.1.ii. Surface and Interfacial Shear Rheology Studies. For the surface and interfacial shear rheology measurements, both the interfacial elasticity and viscosity were determined and measurements were performed as a function of time (time sweep). A CIR A-100 oscillating stress interfacial rheometer and a Pt/Ir du No€ uy ring were used.32 A high-resolution displacement of the sensor was used to monitor the amplitude of motion of the ring, and automatic analysis of the generated signal provided the dynamic surface rigidity modulus (surface elasticity, G0 s) (mN 3 m). Calibration was carried out prior to each measurement. For rheometer calibration, the ring was set to oscillate in air using a series of standard inertia bars of known weight. A pure canola oil/air interface and a water/pure canola oil interface were used in all experiments as references at the oily solution/air interface and oily solution/water interface, respectively. The instrument was operated in a normalized resonance (NR) mode (g1 Hz). The amplitude was maintained at 5000 mrad/s and an oscillatory frequency of 3.0 Hz. Temperature control was achieved using a circulating water bath with ramp function control (Haake Phoenix, P1 B5, ThermoFisher Scientific Waltham, MA). A total of 25 g of each sample was prepared. The total amount of surfactant blend was maintained at (32) Camtel Ltd. 2000. Emulsion and Foam Stability Testing Using the Interfacial Shear Rheometer CIR-100: Royston, England, 1997; p 1.

DOI: 10.1021/la103358n

17875

Article

Morais et al.

Table 2. Surface Shear Elasticity and Viscosity of Surfactant Solutions as a Function of Time (s) at 25 and 50 ( 0.1 °C canola oil þ Span80 (24.2% w/w) canola oil þ CremophorRH40 (25.8% w/w) canola oil þ CremophorRH40 (25.8% w/w) þ Span80 (24.2% w/w)

50.0% w/w in order to reproduce not only the surfactant/oil phase ratio (1:1) but also the HLB value (9.3) in which W/O/W multiple emulsions were formed. The components were homogenized before the test. For samples containing only Span80, the use of a glass rod was sufficient. Conversely, samples containing CremophorRH40 or a surfactant blend in canola oil required heating to approximately 40 ( 0.5 °C using a hot plate. All measurements were performed three times, and the means and standard deviations are reported. The interfacial elasticity (G0 s) (mN/m) and interfacial viscosity 0 (η s) (mN 3 s/m) are defined as described in eqs 1 and 233 G0 s ¼ g f I4π2 ð f 2 - f 2 0 Þ

ð1Þ

Figure 1. Bulk viscosity (cPs) as a function of temperature (°C). The experiments were conducted in triplicate (n = 3), and all data were reproducible. Representative graphs are shown, but error bars are not included because the time points varied with the experiment as a result of the incremental temperature increase. (The results are under the statistically significance limit.)

1 η0s ¼ gf INCð - X0 Þ X

ð2Þ

Table 3. Interfacial Shear Elasticity and Viscosity of Surfactant Solutions as a Function of Time (s) at 25 and 50 ( 0.1 °C

where G0 s is the interfacial elasticity, gf is the geometry factor, I is the moment of inertia, f2 is the interfacial resonance frequency of the sample, f20 is the interfacial resonance frequency of the reference, η0 s is the interfacial viscosity, NC is the number of cycles of integration, X is the mean amplitude of the sample interface, and X0 is the mean amplitude at the reference interface. gf is defined as a function of the radius of the ring (eq 3) gf ¼ 4πðR1 - R2 Þ

ð3Þ

where R1 is the radius of the Pt/Ir ring (13 mm) and R2 is the radius of the sample cell (1.9265). The phase angle (δ) or the phase shift relative to the strain was calculated as follows (eq 4): tan δ ¼

lost energy ðη0s Þ stored energy ðG0s Þ

ð4Þ

G0 is the stress measured at the maximum strain divided by the strain amplitude (γ0), which is the strain reached at ωt = π/2. However, G00 is the stress at zero strain (i.e., the strain at ωt = 0 divided by γ0.) Thus, the stress at zero strain is the result of the sample responding to the strain rate as would a purely viscous material (δ = 90°). As the strain reaches its peak at ωt = π/2, the strain rate approaches zero. The sample at this point must therefore be responding only to strain, as would an elastic material (δ = 0°). Viscoelastic materials exhibit behavior that is intermediate between the behavior of these two types of materials, exhibiting some lag in strain.16 The canola oil/air interface was used as a simplified model to investigate the surface properties of the oil-phase-containing surfactants. The surface rheology of different solutions of lipoand/or hydrophilic surfactants in canola oil (Table 2) was investigated (at 25 and 50 ( 0.1 °C) In addition, different concentrations of CremophorRH40 in aqueous solution (0.5, 1.0, 5.0, 10.0, 15.0, 20.0, and 25.8% w/w or 0.0019, 0.039, 0.0193, 0.0386, 0.0579, 0.0772, and 0.0996 mol/L) were investigated at the aqueous/air interface. The surface rheology of 25.8% w/w aqueous solutions of TritonX100 and Tween80 was also investigated (0.3988 and 0.1969 mol/L, respectively). These experiments were conducted at 25 ( 0.1 °C. At higher temperatures, an aqueous solution of CremophorRH40 formed a gel; consequently, it was not possible to measure surface rheology above 25 °C. Surfactant canola oil solutions/pure water and surfactant canola oil solutions/hydrophilic aqueous solutions were used as (33) Sheriff, M.; Warburton, B. Polymer 1974, 15, 253.

17876 DOI: 10.1021/la103358n

canola oil þ Span80 (24.2% w/w)/pure water canola oil þ CremophorRH40 (25.8% w/w)/pure water canola oil þ CremophorRH40(25.8% w/w) þ Span80 (24.2% w/w)/ pure water canola oil þ Span80 (24.2% w/w)/water þ CremophorRH40 (25.8% w/w)

a simplified model to investigate the rheological properties at the interface during the emulsification process. For interfacial rheology analysis, the following samples were evaluated (at 25 ( 0.1 and 50 ( 0.1 °C): (i) canola oil containing a lipo- and/or hydrophilic surfactants/water interface and (ii) canola oil containing a lipophilic surfactant/water-containing hydrophilic surfactant interface (Table 3).

3. Results 3.1. Bulk Viscosity Study. The bulk viscosity of all oily surfactant solutions decreased with increasing temperature (Figure 1). At 40 ( 0.5 °C, samples containing Span80, CremophorRH40, and a blend of both surfactants exhibited viscosity values of 45.43 ( 1.2503, 116.33 ( 1.1547, and 104.8 ( 5.6320 cP, respectively. At 80 ( 0.5 °C, viscosity values for samples containing Span80, CremophorRH40, and the surfactant blend were 11.63 ( 1.4844, 27.63 ( 6.2931, and 23.6 ( 1.4177, respectively. Comparing viscosity values at low and high temperatures using the t test (paired samples to assume equal or unequal variances), the following p values were obtained for Span 80, CremophorRH40, and blend surfactant samples, respectively: 0.000017, 0.001128, and 0.001063. Accordingly, the bulk viscosity decreases significantly for all samples with increasing temperature. All of the oily surfactant solutions exhibited Newtonian flow behavior (data not shown). The effect of temperature on the bulk viscosity of aqueous solutions containing CremophorRH40 was different from that of the oily surfactant solutions. A significant decrease in viscosity from 32.97 ( 10.46 cP at 40 ( 0.5 °C to 7.73 ( 1.6166 cP at 70 ( 0.5 °C (p value = 0.019377) was observed. The viscosity then remained constant in the range of (70-78) ( 0.5 °C until the system reached 80 ( 0.5 °C, and at this temperature, the viscosity increased sharply to 22.3 ( 2.2 cP. With further increases in temperature, the viscosity remained constant. All CremophorRH40 solutions also showed Newtonian flow behavior (data not shown). 3.2. Surface (Air/solution Film) Shear Rheology of Oily Solutions. The behavior of the surface and interfacial films as a function of time is not presented because a plateau was reached at approximately 90 to 120 s with no further changes until the end of Langmuir 2010, 26(23), 17874–17881

Morais et al.

Article

Table 4. Phase Angle of the Air/Solution Interface at Low and High Temperatures after 43 Minutesa

Table 6. Surface Viscosity (η0 s) (mN 3 s/m) of the Air/Oily Solution Interface after 43 Minutesa surface viscosity (η0 s) (mN 3 s/m)

δ (deg) surfactant oil solutions

25 ( 0.1 °C

54 ( 0.1 °C

p value

CO þ Span80 CO þ CremophorRH40 CO þ blend

83.7005 ( 2.1105 78.8970 ( 2.1492 88.8812 ( 0.2775

63.0242 ( 6.0238 8.8878 ( 1.8239 59.5003 ( 8.7409

0.009402 0.000048 0.028454

a

surfactant oil solutions

25 ( 0.1 °C

50 ( 0.1 °C

p value

CO þ Span80 134.7934 ( 10.192 55.3453 ( 4.8135 0.003155 CO þ CremophorRH40 73.4798 ( 2.717 33.3708 ( 1.9664 0.001212 CO þ surfactant blend 255.5503 ( 27.332 81.8962 ( 7.7603 0.005053 a

Each value is the mean of three measurements (n = 3) and the standard deviation ((SD).

Each value is the mean of three measurements (n = 3) and the standard deviation ((SD). CO represents canola oil.

Table 5. Surface Elasticity (G0 s) (mN/m) of the Air/Oily Solution Interface after 43 Minutesa

Table 7. Surface Rheology of the Air/Aqueous Solution Interface after 43 Minutes at 25 ( 0.1 °Ca

surface elasticity (G0 s) (mN/m) surfactant oil solutions CO þ Span80 CO þ CremophorRH40 CO þ surfactant blend

surfactant aqueous solution

25 ( 0.1 °C

50 ( 0.1 °C

p value

189.5407 ( 64.7406 184.770 ( 34.013

95.0716 ( 18.6349 12.29 ( 3.0212

0.05434 0.011099

539.9737 ( 99.4913

57.0559 ( 17.3906

0.013283

a

Each value is the mean of three measurements (n = 3) and the standard deviation ((SD). CO represents canola oil.

the experiment (43 min). Accordingly, the mean values for the plateau region are shown in Tables 4-10 and are discussed below. All samples exhibited viscoelastic properties with viscousdominant behavior (δ g 45°) at 25 ( 0.1 °C. In addition, all samples exhibited some decrease in the phase angle (δ) with increase temperature. Phase angle (δ) values at low and high temperatures were compared using the t test (two paired samples used to assume equal or unequal variances). At 50( 0.1 °C, the oily sample containing Span80 and the surfactant blend showed decreases from 83.70 ( 2.11 to 63.02 ( 6.02° ( p value = 0.009402) and from 88.88 ( 0.28 to 59.50 ( 8.74° ( p value = 0.028454), respectively. However, samples containing CremophorRH40 exhibited the most considerable decrease in the δ value, from 78.90 ( 2.15 to 8.89 ( 1.82° ( p value=0.000048), altering the system properties to elastic-dominant (δ e 45°) (Table 4). Surface elasticity values are presented in Table 5. At 25 ( 0.1 °C, samples containing the surfactant blend had the highest surface elasticity (539.97 ( 99.49 mN/m). Span80 and CremophorRH40 surface elasticity values were 189.54 ( 64.74 and 184.77 ( 34.01, respectively. Surface elasticity values decreased as the temperature increased for all samples. At 50 ( 0.1 °C, Span80, CremophorRH40, and surfactant blend solutions exhibited surface elasticities of 95.07 ( 18.63, 12.29 ( 3.02, and 57.06 ( 17.39 mN/m, respectively. Comparing elasticity values at low and high temperatures using the t test (two paired samples used to assume equal or unequal variances), the following p values were obtained: 0.05434, 0.011099, and 0.013283. Accordingly, only samples containing Span80 did not show a significant decrease in surface elasticity. Surface viscosity values are presented in Table 6. At 25 ( 0.1 °C, blend surfactant solutions had the highest surface viscosity (255.5503 ( 27.332 mN 3 s/m). Span80 and CremophorRH40 surface viscosity values were 134.7934 ( 10.192 and 73.4798 ( 2.717, respectively. At 50 ( 0.1 °C, Span80, CremophorRH40, and surfactant blend solutions exhibited surface viscosity values of 55.3453 ( 4.813, 33.3708 ( 1.966, and 81.8962 ( 7.76 mN 3 s/m, respectively. Comparing elasticity values at low and high temperatures using the t test (two paired samples used to assume equal variances), the following p values were obtained: 0.003155, 0.001212, and 0.005053. Accordingly, only the samples containing the surfactant blend did not Langmuir 2010, 26(23), 17874–17881

elasticity (G0 s) (mN/m)

viscosity (η0 s) (mN 3 s/m)

CremophorRH40 383.1871 ( 77.7078 55.3766 ( 3.0667 TritonX100 487.2812 ( 82.094 98.7992 ( 8.2189 Tween80 339.2285 ( 3.0101 54.7158 ( 4.3985 a Each value is the mean of three measurements (n = 3) and the standard deviation ((SD).

show a significant decrease in surface viscosity as a function of increasing temperature. 3.3. Surface (Air/solution Film) Shear Rheology of Aqueous Solutions. Because the samples containing CremophorRH40 in canola oil exhibited a significant reduction in δ values as a function of temperature, this surfactant was also investigated in the aqueous phase at different concentrations (w/w %). Aqueous solutions containing CremophorRH40 were sensitive to temperature increases, changing from a homogeneous solution at 25 ( 0.1 °C to a heterogeneous gel-sol mixture at 50 ( 0.1 °C. At high temperature, the presence of the gel did not allow free rotation of the ring; therefore, the study was terminated. The elasticity values decreased as a function of increasing concentration; conversely, the opposite behavior was observed for the surface viscosity values. The results of the surface elasticity and viscosity are shown in Figure 2. In addition, considering the individual behavior of CremophorRH40, not only in oily solution as a function of temperature but also in aqueous solution as a function of concentration, other hydrophilic surfactants (TritonX100 and Tween80) were also evaluated in an attempt to understand the surface rheological properties as a function of the chemical structure (Table 7). All hydrophilic surfactants in aqueous solution exhibited predominantly viscous behavior (δ g 45°). The δ values for CremophorRH40 (82.6588 ( 1.1017°) and Tween80 (81.7472 ( 0.612°) were statistically similar ( p value=0.173321), whereas that for TritonX100 (86.7535 ( 0.3212°) was significantly higher than that for CremophorRH40 ( p value=0.003625). TritonX100 exhibited the highest surface elasticity and viscosity, 487.2812 ( 82.094 mN/m and 98.7992 ( 8.2189 mN 3 s/m, respectively. CremophorRH40 and Tween80 had almost identical surface viscosity values (55.3766 ( 3.0667 and 54.7158 ( 4.3985, respectively, p value = 0.406782) and surface elasticity (383.1871 ( 77.7078 and 339.2285 ( 3.01, respectively, p value = 0.43079). CremophorRH40 and TritonX100 had similar surface elasticity values ( p value = 0.092972); however, for surface viscosity, TritonX100 had a significantly higher value ( p value=0.000037). 3.4. Interfacial (Solution/Solution Film) Shear Rheology of Oily Solutions. Interfacial rheology studies were carried out for two different interfacial models: (i) pure water/canola oil containing either lipo- or hydrophilic surfactants or a blend of both and (ii) hydrophilic surfactants in aqueous solution/canola oil containing a lipophilic surfactant. All samples demonstrated predominant viscous behavior (δ g 45°) at both 25 and 50 ( 0.1 °C. DOI: 10.1021/la103358n

17877

Article

Morais et al. Table 8. Phase Angle of the Solution/Solution Interface at Low and High Temperatures after 43 Minutesa δ (deg) surfactant solutions

25 ( 0.1 °C

54 ( 0.1 °C

p value

CO þ Span80/water 89.9007 ( 0.0152 89.8407 ( 0.0102 0.030334 CO þ CremRH40/water 87.9233 ( 0.4905 84.4322 ( 1.4253 0.035099 CO þ blend/water 89.9925 ( 0.0002 89.9987 ( 0.0005 0.004332 CO þ Span80/water þ CremRH40 88.8485 ( 0.1351 87.6222 ( 0.2733 0.010821 a Each value is the mean of three measurements (n = 3) and the standard deviation ((SD). CO represents canola oil, and CremRH40 represents CremophorRH40.

Table 9. Surface Elasticity (G0 s) (mN/m) of the Solution/Solution Interface after 43 Minutesa elasticity (G0 s) (mN/m) surfactant solution

25 ( 0.1 °C

50 ( 0.1 °C

p value

CO þ Span80/water 3610.74 ( 205.87 3783.92 ( 160.66 0.2121 CO þ CremRH40/water 358.6023 ( 72.305 516.8711 ( 69.748 0.005763 b 18739.11 ( 384.68 13313.91 ( 544.364 0.004462 CO þ blend/water CO þ Span80/water þ CremRH40 746.169 ( 68.194 511.495 ( 71.654 0.023692 a Each value is the mean of three measurements (n = 3) and the standard deviation ((SD). CO represents canola oil, and CremRH40 represents CremophorRH40. b Measurements were made after 600 s because of ring movement restriction.

Table 10. Surface Viscosity (η0 s) (mN 3 s/m) of the Solution/Solution Interface after 43 Minutesa viscosity (η0 s) (mN 3 s/m) surfactant solution

25 ( 0.1 °C

50 ( 0.1 °C

p value

CO þ Span80/water 421.2885 ( 52.122 247.858 ( 8.806 0.029596 CO þ CremRH40/water 198.336 ( 57.94 55.7283 ( 22.109 0.019584 CO þ blend/water 1062.49 ( 3.1408 993.2 ( 0 0.000684 CO þ Span80/water þ CremRH40 175.114 ( 7.748 124.1196 ( 4.3822 0.000757 a Each value is the mean of three measurements (n = 3) and the standard deviation ((SD). CO represents canola oil, and CremRH40 represents CremophorRH40. b Measurements were made after 600 s because of ring movement restriction.

Figure 2. Surface rheological properties of CremophorRH40 in aqueous solution as a function of concentration (w/w %): (a) surface elasticity (mN/m) and (b) surface viscosity (mN 3 s/m). Each value is the mean of three measurements (n = 3) and the standard deviation ((SD).

Phase angle (δ) values at low and high temperatures were compared using the t test (two paired samples used to assume equal or unequal variances). Although all samples showed significant changes in the phase angle, viscous-dominant (δ g 45°) rheological behavior was maintained (Table 8). Interfacial elasticity values are presented in Table 9. At 25 ( 0.1 °C, Span80 in canola oil/pure water had higher interfacial elasticity (3610.74 ( 205.87 mN/m) than CremophorRH40 in canola oil/pure water (385.6023 ( 71.3052 mN/m). The surfactant blend in canola oil/pure water had the highest surface elasticity (18739.11 ( 384.68 mN/m). However, it is noteworthy that samples of canola oil containing surfactant blend/pure water 17878 DOI: 10.1021/la103358n

formed a gel-like structure and the measurements were terminated after 600 s (point at which the ring stops rotating freely). Moreover, the addition of CremophorRH40 to water and Span80 to canola oil resulted in an interfacial elasticity value of only 746.169 ( 68.194 mN/m. At 50 ( 0.1 °C, samples of canola oil containing (i) Span80/pure water, (ii) CremophoRH40/pure water, (iii) a surfactant blend, and (iv) Span80/water containing CremophorRH40 had the following surface elasticity values: 3783.92 ( 160.66, 516.8711 ( 69.748, 13313.91 ( 544.364, and 511.495 ( 71.654 mN/m, respectively. Accordingly, the interfacial elasticity of canola oil samples containing Span80/pure water did not change significantly with increasing temperature. Samples of Langmuir 2010, 26(23), 17874–17881

Morais et al.

canola oil containing the surfactant blend/pure water and Span80/water containing CremophorRH40 exhibited a significant decrease in surface elasticity as a function of increasing temperature. However, the surface elasticity for canola oil samples containing CremophorRH40/pure water increased significantly as the temperature increased. Interfacial viscosity values are presented in Table 10. At 25 ( 0.1 °C, canola oil containing surfactant blend/pure water had the highest interfacial viscosity value (1062.49 ( 3.1408 mN 3 s/m). In addition, canola oil containing Span80/pure water had a higher interfacial viscosity (421.2885 ( 152.32 mN 3 s/m) than the CremophorRH40 sample (188.336 ( 57.94 mN 3 s/m). Samples of canola oil containing Span80/water containing CremophorRH40 had the smallest interfacial viscosity value (175.114 ( 7.748 mN 3 s/m). A statistically significant decrease in the interfacial viscosity as a function of increasing temperature was observed for all samples.

4. Discussion The formation of stable and normal W/O/W multiple emulsions using the one-step method requires specific conditions, such as (i) a critical emulsification temperature; (ii) the use of a surfactant or surfactant blends with specific HLB values; (iii) a specific surfactant blend/oil phase ratio; and (iv) the addition of the surfactant blend to the oil phase.6,9,10 Specifically for the emulsion system investigated, normal W/O/W formed spontaneously under the following conditions: (i) a critical emulsification temperature of 78 ( 2 °C; (ii) the use of a nonionic surfactant blend (CremophorRH40 and Span80) with an HLB value of 9.3; (iii) a surfactant blend/oil phase in a ratio of 1:1 (w/w); and (iv) the addition of the surfactant blend to the oil phase.9,10 Changes in interfacial and/or bulk rheological properties of polymeric and/or monomeric surfactant solutions can be correlated to inter- and intramolecular rearrangements and interactions (e.g., fatty acid chains of the hydrophobic moieties and hydrogen bonds between polar heads and the continuous phase) at the interface during the emulsification process.8,22 To understand the one-step W/O/W emulsification process further, the rheological properties of the surfactant solutions were evaluated with respect to temperature, surfactant concentration, surfactant ratio, and surfactant solvent phase. 4.1. Bulk Viscosity. The bulk viscosity of the surfactant solutions decreased as a function of increasing temperature for all oily surfactant solution samples (Figure 1). However, no abrupt decrease or other atypical behavior was observed at the critical emulsification temperature. Aqueous surfactant solutions showed a more gradual decrease when compared to the oily samples. Because viscosity is defined as the measure of friction between adjacent layers in a fluid and is directly related to the resistance of these layers to flow, decreases in the bulk viscosity are usually related to increases in the bulk molecular mobility.8,19,23 Consequently, the migration of the surfactant molecules from the bulk to the interface during the emulsification process is assisted by increased molecular mobility. The samples containing CremophorRH40 in canola oil showed a faster decrease in bulk viscosity with increasing temperature when compared to the samples containing CremophorRH40 in aqueous solution; therefore, it appears that the hydrophilic surfactant has higher mobility in the oil media and consequently can reach the interface more rapidly. The slower mobility of CremophorRH40 in water may be responsible for the absence of multiple droplet formation when CremophorRH40 is added to the water phase. Lin and collaborators have already suggested the importance of this behavior in multiple droplet formation by the one-step emulsification process.34 (34) Lin, T. J.; Kurihara, H.; Ohta, H. J. Soc. Cosmet. Chem. 1975, 26, 121.

Langmuir 2010, 26(23), 17874–17881

Article

The change in the bulk viscosity behavior of the CremophorRH40 aqueous solution at approximately 82 °C (sharp increase) suggests a surfactant phase transition from a micellar solution to a cubic crystalline phase (thermal gelation).8,22 The occurrence of the cubic phase in the region of the critical emulsification temperature (78 ( 2 °C) is indicative of the occurrence of the Winsor III phase and consequent microemulsion and multiple emulsion formation.6 4.2. Surface (Air/Solution Film) Shear Rheology of Oily Solutions. Surface/interfacial shear rheology reflects changes in film properties during shearing and provides a direct measurement of the mechanical strength of the adsorbed layer.17,35 The surface shear rheology of all films reached a plateau within 90 to 120 s; such behavior is expected for solutions of low-molecular-weight surfactants as explained by the Gibbs-Marangoni effect.17,22,32 At low and high temperature evaluation, phase angle values (δ) above 45° indicated viscous-dominant behavior for all samples (air/solution film), which can be explained by the formation of a liquidlike film by low-molecular-weight surfactants.16,24 However, at 50 ( 0.1 °C, canola oil solutions containing only CremophorRH40 changed to elastic-dominant behavior (8.89 ( 1.82°), indicating the formation of a solidlike film under thermal influence (Table 4). At 25 ( 0.1 °C, canola oil samples containing Span80 or CremophorRH40 had similar low shear elasticity values (Table 5). Low shear elasticity values are usually expected for solutions containing low-molecular-weight molecules because they have reversible adsorption at the interface and do not have the ability to form loops and tails that protrude from and entangle at the interface. The similarity in the interfacial elasticity values of CremophorRH40 (MW=2589 and HLB=14.1) and Span80 (MW=428.61, HLB=4.3) was not expected and indicated that the surfactant molecular weight and chemical structure did not affect the interfacial shear elasticity. Consequently, it appears that canola oil has a major effect on the surface activity. Canola oil contains 94.4-99.1% triglycerides, 0.4-1.2% free fatty acids, and 700-1200 ppm of tocopherols.36 Triglycerides are surface-active and are expected to interact with the surfactants at the interface and in the subsurface layers. Surface/interfacial shear elasticity is related to intra- and intermolecular interactions not only within the interface but also within the subphase. These interactions contribute positively or negatively to the film strength.17,24 It noteworthy that stable multiple emulsions were obtained by the one-step method only when vegetable oils were used, whereas mineral oil produced limited-stability systems. This also indicates the role of the triglycerides in the formation and stability of these systems. The highest surface shear elasticity values were observed when both surfactants were added to canola oil. This synergistic effect on the surface shear elasticity can be interpreted in terms of cooperative binding between the polymeric and monomeric surfactants. This result also helps to explain why both surfactants must be used to achieve one-step multiple emulsion formation. Similarly, the surface/interfacial shear viscosity can be correlated to the interactions between the surfactants and canola oil. The highest viscosity was achieved with the surfactant blend in canola oil. Enhanced surface shear viscosity supports the notion of a mixed of lipo-/hydrophilic surfactant film.22,37 At 50 ( 0.1 °C, a decrease in surface shear elasticity values was observed for all oily samples. However, canola oil solutions containing either CremophorRH40 or the surfactant blend were (35) Burgess, D. J.; Sahin, N. O. J. Colloid Interface Sci. 1997, 189, 74. (36) Ying, C. F.; deMan, J. M. Can. Inst. Food Sci. Technol. 1989, 22, 222. (37) Regismond, S. T. A.; Gracie, K. D.; Winnik, F. M.; Goddard, E. D. Langmuir 1997, 13, 5558.

DOI: 10.1021/la103358n

17879

Article

more susceptible to increasing temperature when compared to samples containing only Span80. Increasing temperature leads to an increase in molecular mobility, which may be favorable or unfavorable for molecular interactions depending on the types of interactions.28 This decrease in the surface shear elasticity indicates a reduction in intermolecular interactions/surface activity as the molecular mobility increases. This result can be explained on the basis of CremophorRH40 hydrophilicity changes. Shinoda and collaborators postulated that the hydrophilicity of nonionic polyethylene oxide surfactants decreases with increasing temperature.7,30,31 Accordingly, the interfacial activity of CremophorRH40 (PEO groups) should decrease as the temperature increases and the hydrophobic interactions between the surfactants and the canola oil should increase in the bulk phase. 4.3. Surface (Air/Solution Film) Shear Rheology of Aqueous Solutions. The surface shear viscosity of CremophorRH40 solutions increased as a function of concentration as expected, as a result of the increased number of molecules at the interface (Figure 2b). However, the surface shear elasticity of the CremophorRH40 aqueous solutions decreased as a function of increasing concentration (Figure 2a). This may be explained by considering the complexity of the chemical structure of CremophorRH40. The hydrophobic moiety of CremophorRH40 has three triglyceride chains; therefore, it is difficult to pack at the interface because of the steric restrictions of the tails anchored at the liquid/air interface. In addition, the tail groups also contain hydrophilic moieties (PEO groups). Consequently, as the concentration increases, it is expected to be more favorable for these molecules to migrate into the bulk phase. The surface rheological properties of TritonX100 and Tween80 were also evaluated in order to correlate these properties to the surfactant chemical structure (Table 7). All three surfactants exhibited δ g 45°, indicating dominant-viscous behavior. It is noteworthy that TritonX100 had the highest dominant-viscous aspect and the highest surface shear viscosity, which can be correlated to its lower hydrophilicity/HLB value (13.5) when compared to those of Tween80 (15) and CremophorRH40 (14.1). Tween80 and CremophorRH40 have similar phase angles (δ) and surface viscosity values. All three surfactants exhibited similar surface shear elasticity values that may be related to their low molecular weight and similar HLB values. 4.4. Interfacial (Solution/Solution Film) Shear Rheology of Oily Solutions. The interfacial shear rheology of all films reached a plateau within a few minutes, and this behavior is in accordance with the Gibbs-Marangoni effect.15,17,32 All films showed viscous-dominant behavior both at low and high temperatures. In all cases, the interfacial shear elasticity values were higher than the surface values. These results suggest more extensive molecular interactions at the interface of solution/solution films than in air/solution films. These interactions are possibly due to surfactant multilayer formation and/or interfacial associations with micelles (aqueous subphase) and inverse micelles (oil phase) on both sides of the interface.20-22 Samples of canola oil containing Span 80/pure water had higher values when compared to samples of canola oil containing CremophorRH40. This behavior can be explained by the high solubility (high HLB=14.1) of CremophorRH40 in water, which results in its migration from the interface to the aqueous subphase, decreasing its surface activity. In addition, the highest interfacial shear elasticity value was observed when both surfactants were added to the oil phase, showing evidence of synergism between the lipo- and hydrophilic surfactants. The surface shear viscosity values were also the highest when both surfactants were 17880 DOI: 10.1021/la103358n

Morais et al.

added to the oil phase. These data are in agreement with the well-known principle in emulsion preparation that a blend of lipoand hydrophilic surfactants results in synergistic lowering of surface tension and is able to form stronger films than single surfactants.37 In addition, high interfacial viscosity contributes to emulsion long-term stability by reducing the rate of droplet coalescence.22 When the surfactants where placed in different solvent phases (i.e., Span80 in canola oil and CremophorRH40 in water), the interfacial shear elasticity was significantly lower than when both surfactants were added to canola oil. When the hydrophilic surfactant is added to the aqueous phase, it is less likely to migrate from the aqueous subphase to the solution/solution interface and disrupt the equilibrium of Span80 molecules at the interface. These results show the superior mechanical strength of the solution/solution film formed when both surfactants were added to the oil phase and help to explain why this parameter is critical in multiple droplet formation and its long-term stability by the onestep process. After being heated, samples of canola oil containing Span80 at the pure water interface did not show any significant change in interfacial shear elasticity. However, samples of canola oil containing CremophorRH40 at the pure water interface did show a significant increase in interfacial elasticity. Oxyethylene-based nonionic surfactants are temperature-sensitive and become less hydrophilic with increasing temperature, changing their affinity from the water to the oil phase. Therefore, it is speculated that some of the monomers/micelles migrated from the aqueous subphase to the solution/solution interface, increasing the interfacial shear elasticity. Thermal denaturation of the PEO groups of the CremophorRH40 backbone should also be considered. Samples of canola oil containing Span80/water containing CremophorRH40 and samples of canola oil containing the surfactant blend/pure water showed a decrease in interfacial shear elasticity values. These results indicate that the CremophorRH40/ Span80 complex tends to desorb from the solution/solution interface, decreasing the surface activity as the temperature increases. These results are indicative of decreasing molecular interactions (adhesion forces) between surfactants with both the continuous phase (aqueous phase) and the dispersed phase (canola oil). Accordingly, it is reasonable to speculate that at higher temperatures (critical emulsification temperature) this would eventually lead to the formation of a third phase (surfactant phase) as described by the Winsor III nomenclature. In the Winsor III region, the ultralow surface tension value assists multiple droplet formation by the one-step process. Because of equipment limitations, it was not possible to evaluate the interfacial rheology at temperatures above 50 °C. It has been previously shown that the presence of a polymeric surfactant (CremophorRH40) is necessary for the formation of multiple emulsions via the one-step process.6 This can be explained by the stronger primary and secondary films formed by polymeric surfactants when compared to monomeric surfactants as well as the greater physical stability of the emulsion formed using polymeric surfactants.1,5,38,39 In this study, the role of the polymeric surfactant and the superiority of the simultaneous use of two different types of surfactants (polarity, molecular weight, and chemical structure) for one-step multiple emulsion formation are established. The results explain the need for polymeric surfactants and surfactant blends for the (38) Hameyer, P.; Jenni, K. R. Cosmet. Toiletries 1996, 111, 39. (39) Kanouni, M.; Rosano, H. L.; Naouli, N. Adv. Colloid Interface Sci. 2002, 99, 229.

Langmuir 2010, 26(23), 17874–17881

Morais et al.

Article

formation of stable (not temporary) multiple droplets by the one-step emulsification process.

5. Conclusions The characterization of bulk properties and molecular interactions at the solution/air and solution/solution interfaces facilitated an understanding of multiple emulsion formation via the one-step method. Microstructural changes characterized via surface elasticity and viscosity were directly related to critical parameters necessary for one-step multiple emulsion formation, such as (i) the critical emulsification temperature; (ii) the use of a nonionic surfactant blend (CremophorRH40 and Span80) with a HLB value of 9.3; (iii) a surfactant blend/oil phase at a ratio of 1:1 (w/w); and (iv) the addition of the surfactant blend to the oil

Langmuir 2010, 26(23), 17874–17881

phase. In addition, the bulk viscosity profile showed a phase transition in the ultralow interfacial tension temperature range. The interfacial and bulk rheology data support the previous conclusion that normal (not temporary) W/O/W emulsions formed by the one-step method are a result of the simultaneous occurrence of phase-transition processes in the Winsor III region.6 These data provide mechanistic insight into multiple droplet formation via the one-step method and will assist research and development scientists in the process and composition design of such multiple emulsion preparations. Acknowledgment. We thank CAPES (MEC-Brasil) for the financial support of this work by Programa de Doutoramento com Estagio no Exterior (PDEE).

DOI: 10.1021/la103358n

17881