Shear-Induced Topology Changes in Liquid ... - ACS Publications

Sep 15, 2007 - G. Montalvo , R. Pons , G. Zhang , M. Díaz , and M. Valiente ... in the Salt-Free Catanionic Surfactant Systems Containing Deoxycholic...
0 downloads 0 Views 579KB Size
10518

Langmuir 2007, 23, 10518-10524

Shear-Induced Topology Changes in Liquid Crystals of the Soybean Lecithin/DDAB/Water System Gemma Montalvo,*,† Mercedes Valiente,† and Ali Khan‡ Quı´mica Fı´sica, UniVersidad de Alcala´ , E-28871 Alcala´ de Henares (Madrid), Spain, and Physical Chemistry 1, Lund UniVersity, P.O. Box 124, SE-22100 Lund, Sweden ReceiVed May 25, 2007. In Final Form: July 4, 2007 The viscoelastic behavior of the two different liquid crystalline lamellar phases and the liquid crystalline cubic phase of the mixed soybean lecithin/DDAB system in water was studied through rheology, with mechanical parameters studied as a function of composition. The swollen or diluted lamellar region is formed by vesicles, and its characteristic flow curve presents two-power law regions separated by a region where viscosity passes through a maximum. Yield stress and shear-dependent flow behavior were also observed. The microstructure suffers transformation under shear stress, and rheological response shifts from thixotropic to antithixotropic loops. Similar rheological behavior has been observed for samples in the collapsed or concentrated lamellar region, at the water-rich corner of the phase diagram. Vesicle formation may therefore occur by shearing the initial stacked and open bilayers. However, concentrated lamellar samples in the water-poor part of the phase diagram are less sensitive to shear effects and show plastic behavior and thixotropy. All lamellar samples manifest high elasticity. The dynamic responses of both lamellar topologies, i.e., vesicles and open bilayers, are comparable and exhibit an infinite relation time. The bicontinuous cubic, liquid crystalline phase is highly viscous. Its dynamic response cannot be modeled by a Maxwell model.

1. Introduction Lecithin is a fully biocompatible substance which constitutes the lipid matrix of many biological membranes. The amphiphilic character of lecithin induces very complex behavior in mixtures with water,1 hydrocarbons,2,3 or surfactants. Many different structures can be formed, depending on the system’s thermodynamic and structural parameters, e.g., temperature, composition, or the length of an aliphatic chain in a surfactant molecule. Among basic structures that form at a low concentration of amphiphilics are spherical micelles, cylinders, and plane surfaces. It is important to note the great ability of lecithin to form vesicles in water. At higher concentrations, different liquid crystalline phases are formed, and these may be hexagonal, lamellar, or cubic. Lecithin is widely used in everyday life as a surfactant, while mixtures of lecithin in water or oils may be found in human and animal food, medicine, cosmetics, and pharmaceutics.4,5 In many of these applications, it is important to control the flow behavior and viscoelastic properties of the system. For this reason, the rheological aspects of such systems should be studied, even more so if we consider that they are sensitive to association phenomena. Viscoelastic behavior can usually be determined by measuring the storage modulus G′(ω) and the loss modulus G′′(ω) over a wide range of frequencies and temperatures. The storage modulus is proportional to energy storage in a deformation cycle, reflecting a certain degree of elasticity in the system, whereas the loss modulus is proportional to dissipation or loss of energy as heat in a deformation cycle, reflecting a certain degree of viscosity. * To whom correspondence should be addressed: Telephone: + 34 91 885 4671. Fax: + 34 91 885 4763. E-mail: [email protected]. † Universidad de Alcala ´. ‡ Lund University. (1) Bergenståhl, B.; Fontell, K. Prog. Colloid Polym. Sci. 1983, 68, 48. (2) Angelico, R.; Ceglie, A.; Olsson, U.; Palazzo, G. Langmuir 2004, 60, 619. (3) Angelico, R.; Ceglie, A.; Olsson, U.; Palazzo, G. Langmuir 2000, 16, 2124. (4) Gunstone, F. D., Harwood, J. L., Padley, F. E., Ed. The Lipid Book; Chapman and Hall: London, 1994. (5) Wendel, A. John Willey & Sons: New York, 1995; Vol. 15.

Shear flow is known to have a strong influence on the structure of complex fluids due to a mechanical deformation that can alter the structure. Although lamellar liquid crystalline phases are rather different in their molecular structure and composition, similar shear orientation effects have been observed. In some surfactant systems, shear-induced lamellae alignment was found at both low6,7 and high shear rates.6 At intermediate shear rates, however, multilamellar vesicle formation was observed by Roux and colleagues in the systems AOT/brine and SDS/pentanol/ dodecane/water.6 The shear-induced vesicles have also been reported at different critical shear rates in several systems involving nonionic,7,8 ionic,8,9 and zwitterionic10 surfactants and block copolymer in solution (in water and butanol).11 The presence of a defective lamellar structure, characterized by water holes in surfactant lamellae, favors the formation of elongated vesicles that are disordered in the direction of shear flow.8,12,13 Rheological experiments alone are not sufficient for understanding the rheological properties, and so a combination of small-angle light- and neutron scattering has been employed in these systems.8,12,13 A transition from parallel lamellae to perpendicular lamellae, i.e., with the normal layer along the vorticity direction being observed in several surfactant13-15 and block copolymer systems16,17 at increasing shear rates. The (6) Diat, O.; Roux, D.; Nallet, F. J. Phys. II Fr. 1993, 3, 1427. (7) Lukaschek, M.; Mu¨ller, S.; Hansenhindl, A.; Grabowski, D. A.; Schmidt, C. Colloid Polym. Sci. 1996, 274, 1. (8) Richtering, W. Prog. Colloid Polym. Sci. 1997, 104, 90. (9) Goldszal, A.; Jaimeson, A. M.; Mann, J. A.; Polak, J.; Rosenblatt, C. J. Colloid Interface Sci. 1996, 180, 261. (10) Bergmeier, M.; Hoffmann, H.; Thuning, C. J. Phys. Chem. B 1997, 101, 5767. (11) Zipfel, J.; Lindner, P.; Tsianou, M.; Alexandridis, P.; Richtering, W. Langmuir 1999, 15, 2599. (12) La¨uger, J.; Weigel, R.; Berger, K.; Hiltrop, K.; Richtering, W. J. Colloid Interface Sci. 1996, 181, 521. (13) Zipfel, J.; Lindner, P.; Richtering, W. Prog. Colloid Polym. Sci. 1998, 110, 139. (14) Berghausen, J.; Zipfel, J.; Lindner, P.; Richtering, W. Europhys. Lett. 1998, 43, 683. (15) Pedenfold, J.; Staples, E.; Hammouda, B. Langmuir 1996, 12, 3122. (16) Chen, Z.-R.; Kornfield, J. A.; Smith, S. D.; Grothaus, J. T.; Satkowski, M. M. Science 1997, 277, 1248.

10.1021/la701539f CCC: $37.00 © 2007 American Chemical Society Published on Web 09/15/2007

Rheological Properties of Lecithin/DDAB/H2O System

reorientation process is accompanied by a decrease in viscosity.14 Apparently, surfactant mesophases and polymer melts can show similar behavior under shear, indicating that the perpendicularto-parallel transition might be a general feature of complex fluids with lamellar morphology. The phase behavior of the lecithin/didodecyldimethylammonium (DDAB)/water system has been presented previously.18 The cationic and double-tailed surfactant DDAB has applications as a disinfecting agent and immunosuppressant agent and as a drug to control organ transplant rejection.19 The triangular phase diagram shows a large collapsed lamellar phase originating from the water-poor part of the DDAB/water system to a corner of pure lecithin. Electrostatic interactions may be responsible for the huge swelling. A swollen lamellar phase appears in the waterrich part of the phase diagram, at lecithin contents lower than 12 wt %, formed by a polidispersity population of multilamellar vesicles, some of which are nearly 5 µm in size. The two lamellar phases coexist in equilibrium without any macroscopic phase separation, and only by using the small-angle X-ray scattering (SAXS) technique can the phase boundaries be drawn. There is also a bicontinuous type Ia3d cubic liquid crystalline phase occupying a small area in the water-poor part of the phase diagram. An understanding of plane bilayer behavior in lamellar phases is essential for its implications in membrane protein reconstitution. Consequently, both the rich phase behavior of bilayers and their structural relationship to biological membranes have motivated studies from many disciplines and, in particular, inspired the studies of this ternary system. The microstructural parameters of these phases are well established;18 however, a detailed study of the mechanical properties of the lamellar membranes was not made. In this paper, the rheological properties of the lecithin/ DDAB/water system are investigated as a function of both lipid and surfactant concentrations. The correlation between structural behavior and rheological properties is also examined. 2. Experimental Section 2.1. Materials and Sample Preparation. Pure soybean lecithin (1,2-diacyl-sn-3-phosphatidylcholine) with the trade name Epikuron 200 was obtained from Lucas Meyer GmbH (Hamburg, Germany). It contains about 2.5% water, with the molecular weight of lecithin being 773.1 The main component is a C-18 acid with two double bounds. Unsaturation is >84%. High-purity didodecyldimethylammonium bromide (DDAB) was obtained from Chemtronica, Tokyo Kasei. Both compounds were used as supplied, and Millipore filtered water was used as solvent. Samples were prepared by weighing the appropriate amounts of each component in small screw-cap tubes. The samples were centrifuged on both sides for about 30 min per side, a few times a day for several days, until the samples were considered well mixed. The lamellar samples were then left to equilibrate at 25 °C. 2.2. Rheometry. Rheological measurements were performed with a Carri-Med CSL 100 controlled-stress rheometer using cone-plate geometry for the lamellar samples. The acrylic cone has an angle of 1° and a diameter of 4 cm. In the case of cubic samples, parallel steel plate-plate geometry with a gap of 1000 µm and a plate diameter of 2 cm was used. The temperature of the system was controlled at 25.0 °C by a Peltier system. In addition, a humidification chamber containing wetted sponges was used to prevent the evaporation of the sample during measurement. Sample compression during loading was done at the lowest rate to minimize perturbation. Shear-rate-dependent viscosity was determined in flow experiments by applying a logarithmic series of increasing stress for 20 min. Viscosity values (η) were calculated as the ratio of shear stress (17) Maring, D.; Weisner, U. Macromolecules 1997, 30, 660. (18) Montalvo, G.; Khan, A. Lanmguir 2002, 18, 8330. (19) Ashman, R. B.; Ninham, B. W. Mol. Inmunol. 1985, 22, 609.

Langmuir, Vol. 23, No. 21, 2007 10519

Figure 1. Phase diagram of the lecithin/DDAB/water system, at 25 °C. The plotted dashed lines correspond to different series studied at a constant lecithin/DDAB ratio (1-3); at a constant lecithin/water ratio (4 and 5); and constant water (6 and 7). Symbols correspond to samples studied for: (O) Lam1 or swollen lamellar region; (9) Lam2 or collapsed lamellar region; (∆) Q, cubic phase. to shear rate. The measurements were carried out after 10 min of thermal equilibration on the plate of the rheometer. Repetitive cycles of increasing and decreasing stresses were done in order to evaluate the shear effect on the viscosity of the lamellar phases. Structural parameters such as the dynamic storage modulus (G′), the loss modulus (G′′), and complex viscosity (η*) were recorded as a function of angular frequency ω (ω ) 2πf) in the range 0.00510 Hz and at constant stress. All of these measurements were carried out under the regimen of linear viscoelasticity; i.e. the material parameters are independent of the applied stress. G′ and G′′ are the two components of the complex modulus G* (ratio of stress amplitude to strain amplitude) and are connected to complex viscosity through eq 1: η* )

xG′2 + G′′2 ω

(1)

3. Results and Discussion The lecithin/DDAB/water system was previously studied in the entire range of concentrations at 25 °C (Figure 1). The two distinct lamellar liquid crystalline phases and the bicontinuous cubic liquid crystalline phase were studied from microscopic perspective.18 In this paper we present a general characterization of the rheological properties of all phases through flow and oscillatory experiments. All of the samples studied are plotted on the phase diagram (Figure 1). 3.1. Flow Measurements. 3.1.1. Swollen Lamellar Liquid Crystalline Phase, Lam1. According to previous SAXS and cryoTEM experiments, the swollen lamellar phase is formed by polydispersed multilamellar vesicles.18 The extension of this region is relatively small. For this reason, only a small range of concentration is suitable for the study of flow properties. These samples require a minimum stress to initiate the flow, which is called yield-stress value, σo. Above yield-stress value, the viscosity curves of all the measured samples present two-power law regions separated by a region where viscosity passes through a maximum centered at a critical shear rate, γ˘ c, ca. 20 or 30 s-1 (Figure 2). The slopes of both regions are nearly the same. A similar flow curve trend has been extensively observed in other lamellar liquid crystalline phases containing other sorts of surfactants.12,20,21 (20) Candau, F.; Jimenez Regalado, E.; Selb, J. Macromolecules 1998, 31, 5550. (21) Montalvo, G.; Rodenas, E.; Valiente, M. J. Colloid Interface Sci. 1998, 202, 232.

10520 Langmuir, Vol. 23, No. 21, 2007

Figure 2. Viscosity as a function of shear rate for the swollen lamellar samples.

Yield-stress values (σo) are relative to different kinds of interactions between the molecules which confer a threedimensional internal structural network. σo values were calculated from the viscosity curves by simple extrapolation to the zero shear rate at the limit for low shear rates. The values are in the range of 6.2-7.9 Pa for all sample compositions in this region, high enough to prevent the rising of bubbles trapped in the sample. These values are 1 order of magnitude greater than those in the binary DDAB/water system at similar surfactant content.22 At a fixed lecithin/DDAB constant ratio (Series 3 of Figure 1) there is a tendency to increase the yield-stress value by decreasing the water content. Obviously, higher yield-stress values are related to a stronger lamellar network structure. In accordance with reported data in the binary DDAB/water system, electrostatic repulsion between surfaces determines the swelling for the diluted lamellar phase.23 The presence of lecithin may reduce the repulsive forces between the cationic head group of DDAB. On the other hand, a lower amount of water corresponds with smaller distances between amphiphilic bilayers, which favors a stronger lamellar network. The rheological behavior of the lyotropic liquid crystalline phase is often strongly dependent on shear history. In order to evaluate the shear effect on the viscosity, each sample was subjected to a cycle of increasing and decreasing stresses, applied consecutively for 20 min each. For the ascending stress ramp, viscosity decreases until it reaches the maximum described above, whereas single power-law behavior is observed in the descending stress ramp. An antithixotropic loop at high shear rates changes into a thixotropic loop at low shear rates, in a fashion similar to that reported for the swollen lamellar phase of the binary DDAB/ water system.22 This behavior could be due to the microstructure’s evolving and altering throughout the duration of the applied stress. Studies reported previously in the literature, combined with rheology and with microscopic information techniques (RheoSAXS), should be considered in order to gain a full understanding of the rheological flow results. Differing and occasionally permanent structures may occur at a critical stress given by the maximum of viscosity.6,13,20,21,24-26 Thus, the full flow curve (22) Soltero, J. F. A.; Bautista, F.; Pecina, E.; Puig, J. E.; Manero, O.; Proverbio, Z.; Schulz, P. C. J. Colloid Polym. Sci. 2000, 278, 37. (23) Khan, A.; Jo¨nsson, B.; Wennerstro¨m, H. J. Phys. Chem. 1985, 82, 5180. (24) Herve´, P.; Roux, D.; Bellocq, A.-M.; Nallet, F.; Gulik-Krzywicki, T. J. Phys. II Fr. 1993, 3, 12255. (25) Richtering, W.; Weigel, R.; La¨uger, J. Optical InVestigation of Lyotropic Liquid Crystals Under Shear; XIIth International Congress on Rheology, Quebec, 1996. (26) Zipfel, J.; Nettesheim, F.; Lindner, P.; Le, T. D.; Olsson, U.; Richtering, W. Europhys. Lett. 2001, 53, 335.

MontalVo et al.

Figure 3. Viscosity as a function of shear rate for the collapsed lamellar samples in the region DDAB/water (wt %) > 1. The dashed line represents DDAB/water (wt %) ) 1; the shaded area indicates samples with a power-law flow curve.

(Figure 2) can be explained by several steps: (i) at low shear rate, the vesicles are orientated in flow direction; (ii) at the critical shear rate, γ˘ c, larger multilayer vesicles could be formed, or the vesicles size may decrease, which implies that the number of dispersed vesicles increases. Both facts may lead to an increase of viscosity; (iii) above the γ˘ c, the vesicles may be deformed or destroyed under the shear rate. Dislocations or defects result in an unstable flow situation at high shear rate, as can be observed in the bumpiness of the flow (Figure 2). The formation of smaller vesicles by shearing has already been described in the literature,7 in particular for defective lamellar structures with an excess of water such as in this studied system.8 For that reason, a decrease in the vesicle size by shearing in regimen ii may be a reasonable hypothesis. That situation may also explain the maximum of viscosity, while the change from antithixotropic to thixotropic behavior may be evidence that those induced vesicles relax. The presence of DDAB seems to favor a more rapid vesicle relaxation. 3.1.2. Collapsed Lamellar Liquid Crystalline Phase, Lam2. Samples in this region have the classical lamellar structure, with stacked lamellar bilayers separated by water, according to previous 2H NMR and SAXS experiments.18 Lam is a huge phase in 2 which samples exhibit two different flow behaviors, depending on sample composition. The line DDAB/water (wt %) ) 1 delimits two regions according to differing flow behaviors. For samples in the DDAB/water (wt %) > 1 region, above a yield-stress value, viscosity decreases monotonically with an increase in shear rate (Figure 3). This flow behavior may correspond to a classic lamellar structure orientated to the streamline.27 For these samples, when an increasing/decreasing stress ramp cycle is performed, a large thixotropic area is obtained. The structure does not recover after 30 min of relaxation. Samples at the DDAB/water (wt %) < 1 region, in the waterrich part of the Lam2 phase, exhibit the same type of flow curve described before for samples in the swollen lamellar region (previous section 3.1.1.). Above a yield-stress value, viscosity decreases and later goes through a maximum at a critical shear rate near 20-30 s-1 (Figure 4). For all of these samples, if the flow curve is recorded at different experiment times of less than 20 min, there is a shift in the critical shear rate at which maximum viscosity appears. In cyclic experiments of increasing and decreasing logarithmic stresses for 40 min, a thixotropic-to-anthitixotropic trend was obtained, in the same way that this occurred in the swollen lamellar phase. (27) Montalvo, G.; Valiente, M.; Rodenas, E. Langmuir 1996, 12, 5202.

Rheological Properties of Lecithin/DDAB/H2O System

Langmuir, Vol. 23, No. 21, 2007 10521

Figure 4. Viscosity as a function of shear rate after running for (9) 2 min and (0) 20 min; T ) 25 °C. (Insert) Dashed line represents DDAB/water (wt %) ) 1; shaded area corresponds to samples with a maximum in the flow curve.

Together, all of these flow experiments may help to correlate macroscopic behavior with the microstructure of the collapsed lamellar samples. Evidence given in the literature indicates that unilamellar or multilamellar vesicles can be prepared from the classic stacked lamellar phase by shear, resulting in higher viscosity.12,21,28,29 The maximum in the viscosity curve and its shift (Figure 4) signify the formation of shear-induced vesicles, as structural changes are related to the applied stress values but also to the time during which those stresses are applied. On the other hand, the antithixotropic loop is also attributed to the formation of vesicles, changing to a thixotropic loop due to the relaxation of those vesicles, resulting in the classic stacked lamellar structure or smaller vesicles. In the binary DDAB/water system, thixotropic behavior was only observed for the collapsed lamellar region,22 which is a good argument for assuming that the heterogeneous composition of the bilayer is responsible for a defective lamellar structure that favors vesicle formation by shearing. The presence of defects in the lamellar bilayer as a consequence of the high water content in this part of the collapsed region, and/or electrostatic interactions, may play a part as well.21,30 Still more evidence of a shear-induced structure is obtained by means of an optical microscope with crossed polarizers. Microphotographs showed distinct appearances before and after the flow curve was collected (Figure 5). The result is in agreement with previous observations for DDAB/water22 and AOT/water9 samples. Although both pictures indicate some differences, both correspond to the same sample of a lamellar structure. Thus, the samples suffer a topological transformation from the classic lamellar structure to vesicles, rather than a liquid crystalline phase transition. 3.1.3. Effect of Composition on Yield-Stress Values in Collapsed Liquid Crystalline Lamellar Phase. Yield-stress values (σo) were calculated from the viscosity curves by a simple extrapolation to the zero shear rate. The yield-stress values show good reproducibility from the different sample additions and are composition dependent (Figure 6). At a constant lecithin/water (28) Panizza, P.; Roux, D.; Vuillaume, V.; Lu, C.; Cates, M. Langmuir 1996, 12, 248. (29) Escalante, J.; Hoffmann, H.; Horbascheck, K. J. Phys. Chem. B 2000, 104, 10144. (30) Gulik-Krzywicki, T.; Dedieu, J. C.; Roux, D.; Degert, C.; Laversanne, R. Langmuir 1996, 12, 2668.

Figure 5. Optical textures in polarized light (magnification × 100). Sample of 27 wt % lecithin and 22 wt % DDAB composition, observed at equilibrium (above) and after running a flow experiment for 20 min (below).

Figure 6. Yield-stress values (σo) for samples in the collapsed lamellar liquid crystalline region. (O) Series 4; (B) Series 5. Solid lines guide the eyes. Dotted line corresponds to the DDAB/water ) 1 composition.

ratio, yield-stress values decrease by increasing surfactant content and reach a minimum near 25-30 wt % DDAB, which corresponds to equimolar composition in the bilayer. Above that amount of DDAB, σo values increase significantly. The samples present more resistance to flow when they are formed by a higher lecithin-to-water ratio (Series 4) with a fixed DDAB content.

10522 Langmuir, Vol. 23, No. 21, 2007

Figure 7. Representative profiles of storage modulus G′ (opened symbols), loss modulus G′′ (closed symbols), and complex viscosity modulus η* (crossed symbols) in the linear viscoelastic regime as a function of the angular frequency. Different symbols correspond to different runs.

However, the values remain in the same order of magnitude. A decrease in yield-stress values occurs with the replacement of lecithin by DDAB, and a constant amount of water. It should be attributed to less structured and more defective samples when these are richer in DDAB. 3.2. Oscillatory Measurements. Oscillatory results show good reproducibility in different runs and different sample additions. Rheogram type is identical for both swollen (vesicle) and collapsed (mostly open and stacked bilayers) lamellar regions, and all the samples behave as shown in Figure 7. Under linear viscoelastic conditions, rheograms indicate that these samples are more elastic than viscous, as the storage modulus (G′) is higher by about 1 order of magnitude than the loss modulus (G′′) in the entire range of the frequency investigated. The moduli do not cross over, resulting in an infinite structural relaxation time. This is in close agreement with the existence of yield-stress values in the flow response. This type of behavior is called gel and is comparable to results obtained in the lamellar phases of both of the described topologies (vesicles and stacked bilayers) of various surfactant systems.12,21,31-35 The complex viscosity modulus decreases strongly in a linear fashion over the entire frequency range (Figure 7). In all samples, these slopes are close to -1, thus corresponding to gels.35,36 Similar values have been reported for the binary DDAB/water system22 and for lamellar samples of different types of surfactants.21,32,33 The G′ modulus follows the relation G′ ∝ ωn with values for the exponent n being lower than 0.1 for all the lamellar samples. It is therefore nearly frequency independent, as is typical of gel behavior.36-38 Some authors found a frequency-independent storage modulus for multilamellar vesicle systems,34,35,39,40 as well as for open and extended bilayers.12,27,31,32 Similar behavior (31) Robles-Vazquez, O.; Corona-Galva´n, S.; Soltero, J. F. A.; Puig, J. E.; Tripodi, S. B.; Valle´s, E.; Manero, O. J. Colloid Interface Sci. 1993, 160, 65. (32) Warriner, H. E.; Davidson, P.; Slack, N. L.; Schellhorn, M.; Eiselt, P.; Idziak, H. J.; Schmidt, H.-W.; Safinya, C. R. J. Chem. Phys. 1997, 107, 3707. (33) Ne´meth, Z.; Hasa´lsz, L.; Pa´linka´s, J.; Bo´ta, A.; Hora´nyi, T. Colloid Surf., A 1998, 145, 107. (34) Bergmeier, M.; Gradzielski, M.; Hoffmann, H. J. Phys. Chem. 1998, 102, 2838. (35) Abdel-Rahem, R.; Gradzielski, M.; Hoffmann, H. J. Colloid Interface Sci. 2005, 288, 570. (36) Te Nijenhuis, K. ThermoreVesible Networks: Viscoelastic Properties and Structure of Gels; Springer-Verlag: Berlin, 1997. (37) Clark, A. H.; Ross-Murphy, S. B. AdV. Polym. Sci. 1987, 83, 57. (38) Chronakis, I. S.; Egermeyer, M.; Piculell, L. Macromolecules 2002, 35, 4113. (39) Hoffmann, H.; Munkert, U.; Thuning, C.; Valiente, M. J. Colloid Interface Sci. 1994, 16, 217.

MontalVo et al.

Figure 8. Plot of storage modulus (open, G′), loss modulus (closed, G′′) and complex viscosity modulus (crossed, η*) versus DDAB (wt %). (4) Series 3 in the swollen lamellar phase; (0) Series 1 in the collapsed lamellar phase. Values of the moduli correspond at f ) 0.1 Hz. The solid lines correspond to the linear fitting.

in the two distinct lamellar topologies could be due to the fact that, in both cases, the bilayers are equally “flat” at a local level of observation. On the other hand, the loss modulus of the different mixtures was frequency dependent. At low frequencies, G′′ remained almost constant (i.e., plateau) and then increased progressively with increasing frequencies of oscillation (Figure 7). The plateau may be associated with a relaxation process occurring over long time scales. 3.2.1. Effect of Composition on Viscoelastic Parameters for Lamellar Phases. The effect of composition on the viscoelastic parameters (i.e., η*, G′, G′′) was studied at f ) 0,1 Hz for Series 1 and 3 (see Figure 1), which contain constant lecithin/DDAB ratios. For a log-linear plot, all of the parameters grow linearly with an increase in the DDAB content (Figure 8). Such growth is nearly the same for all the moduli with a slope near 0.09, independent of the kind of lamellar phase. Therefore, the dependence of viscous and elastic parameters on DDAB concentration does not distinguish between collapsed (Lam2) and swollen (Lam1) lamellar phases. However, different behavior is obtained when data are analyzed as function of the molar ratio between the amphiphile components and water (R) (Figure 9). In this situation, the dependence of all moduli with R is nearly the same again, but there is a dramatic change from Lam1 to Lam2 phases. The slope from swollen lamellar phase (Series 3) is about 10 times greater than for collapsed lamellar phase (Series 1). The slopes are 145 and 14, respectively. That means lecithin concentration changes the viscous and elastic properties to a greater extent in the swollen lamellar phase than in the collapsed lamellar phase. Thus, for the Lam1 all of the moduli undergo a very sharp increase although the variation in composition is only 3 wt % lecithin. At constant lecithin/water ratio (Figure 1, Series 4 and 5), all the mechanical moduli have the same composition dependence. For this reason, we only present results on the basis of the storage modulus. By increasing surfactant content, the mechanical modulus decreases and passes through a minimum near 30 wt % DDAB (Figure 10). A further addition of surfactant produces an increase in the mechanical values. The values are slightly higher for the series at a higher lecithin/water ratio (Series 4). Both trends are well in agreement with the composition dependence of yield-stress values obtained in flow experiments and with the antithixotropic-to-thixotropic behavior transition. (40) Hoffmann, H.; Thuning, C.; Schmeidel, P.; Munkert, V. Langmuir 1994, 10, 3272.

Rheological Properties of Lecithin/DDAB/H2O System

Figure 9. Plot of storage modulus (open, G′), loss modulus (closed, G′′), and complex viscosity modulus (crossed, η*) versus molar ratio between amphiphile and water (R): (4) Series 3 in the swollen lamellar phase; (0) Series 1 in the collapsed lamellar phase. Values of the moduli correspond at f ) 0.1 Hz. The solid lines correspond to the linear fitting.

Figure 10. Storage modulus (G′) versus surfactant content for the series at constant lecithin/water ratio: (O) Series 4; (B) Series 5. Solid lines guide the eyes. Dotted line corresponds to the DDAB/ water ) 1 composition.

Figure 11. Plot of storage modulus (G′, open symbols) and loss modulus (G′′, closed symbols) versus the molar fraction between DDAB (D) and lecithin (L) in the bilayer, for the series with constant water (wt %): (4) 20; (0) 40. Solid lines are the linear fitting. The corresponding slopes are given in the plot.

In SAXS experiments we also obtained evidence of a change in lamellar structure, surpassing 30-40 DDAB (wt %). Thus, the X-ray pattern displays three reflection Bragg peaks at lower than 30 wt % DDAB content, instead of the single first-order

Langmuir, Vol. 23, No. 21, 2007 10523

Figure 12. Representative profiles of storage modulus (0, G′), loss modulus (9, G′′), and complex viscosity modulus (crossed squares, η*) in the linear viscoelastic regime, as a function of angular frequency (ω) for the cubic liquid crystalline phase. Straight lines correspond to the Maxwell model.

Bragg peak obtained at a higher amount of DDAB.18 Sample compositions below 30% DDAB are found in the water-rich part of the Lam2 region and manifest shear-induced structures in flow experiments. Taking into account all of this evidence, the rheological behavior observed could be related to the presence of bilayer defects for samples in the water-rich part of the collapsed lamellar phase. There are already smaller and more frequencydependent G′ and G′′ values reported for the defective lamellar phase in the SDS/decanol/water system.12 For Series 6 and 7 (Figure 1) total amphiphile and water weight ratio is nearly constant. Since water molecules are located between the amphiphile bilayers, according to ideal behavior, the effect on viscoelastic properties of replacing lecithin by DDAB can be studied. At a constant amount of water (Figure 11, only G′ and G′′), it is observed a good trend of a linear reduction of all the mechanical moduli as the molar fraction of DDAB in the bilayer (xD ) nD/(nL + nD)) is increased, in a log-linear representation. The slope is different for G′ and G′′ moduli for each series with a constant amount of water. Moreover, both moduli are larger in the series with a lower amount of water, and these values decrease more significantly with xD. Higher mechanical parameter values should be related to more structured samples. The electrostatic effect should be significant in lecithin/DDAB mixtures, since there is no direct relationship between higher concentration and greater elasticity, as has been described for nonionic systems.33 3.2.2. Cubic Liquid Crystalline Phase. The cubic phase is highly viscous, and it was not possible to measure its viscosity by flow experiments due to experimental limitation with the rheometer. Thus, we focus on oscillatory behavior, since it shows very interesting properties. The storage modulus reaches a plateau at high frequency, while the loss modulus surpasses a maximum with angular frequency (Figure 12). The complex viscosity modulus tends to plateau at low angular frequency before dropping linearly to the bottom. Where G′(ω) ) G′′(ω) occurs, taking place at a critical crossover frequency of ω ) 1/τ, a structural relaxation time (τ) of 35 s-1 may be estimated. However, viscoelastic behavior is not a typical feature of the Maxwell model (fitting given in the Figure 12). Some authors have pointed out that the cubic phases behave like quasi-Maxwellian fluid in a manner similar to that described for the lecithin system.41-44 The higher value of the instantaneous elastic modulus is close to other values available in the literature.27,41,44,45

10524 Langmuir, Vol. 23, No. 21, 2007

In general, the existence of a finite structural time is related to a bicontinuous cubic structure, while the cubic phase made up of individual sphere aggregates has infinite relaxation time.45 In self-diffusion NMR experiments, a bicontinuous microstructure of Ia3d space group for the lecithin/DDAB/water system has already been demonstrated.18 The same dynamic response was obtained for the cubic phase of the glucopone/water/hydrocarbon system, which has the same Ia3d space group. Although relaxation time is lower, it is possible to establish a good correlation between the microscopic determination and rheological behavior.

Summary and Conclusions The mechanical properties of the soybean lecithin/DDAB/ water system have been characterized by rheology. Flow experiments indicate that the bilayer microstructure suffers transformation under shear stress in the case of samples in the water-rich part of the phase diagram (DDAB/water < 1), for both lamellar phases. However, collapsed lamellar samples in the water-poor part of the phase diagram (DDAB/water > 1) are less sensitive to shear effects and show a plastic behavior. The (41) Radiman, S.; Toprakcioglu, C.; McLeish, T. Langmuir 1994, 10, 61. (42) Jones, J.; McLeish, T. Langmuir 1995, 11, 785. (43) Jones, J.; McLeish, T. Langmuir 1999, 15, 7495. (44) Siddig, M. A.; Radiman, S.; Muniandy, S. V.; Jan, L. S. Colloids Surf., A 2004, 236, 57. (45) Gradzielski, M.; Hoffmann, H.; Panitz, J.-C.; Wokaun, A. J. Colloid Interface Sci. 1995, 103, 169.

MontalVo et al.

change in flow behavior from the water-rich to water-poor parts of the phase diagram is also observed in yield-stress values. The presence of a high amount of water may introduce defects in the classical lamellar structure, which may favor the formation of induced vesicles by shearing. Moreover, the dynamic response of both lamellar topologies, i.e., vesicles and the classic lamellar structure of stacked bilayers, is comparable and exhibits an infinite relaxation time, as corresponds to a gel behavior. We assume that molecules found in similar environments due to large multilamellar vesicles (over 5 µm of size)18 are equal to flat bilayers at a local level. The mechanical moduli are composition dependent. The cubic liquid crystalline phase is highly viscous. Its dynamic response cannot be modeled by a Maxwell model. The existence of finite relaxation times is well in agreement with the bicontinuous microstructure, deduced from self-diffusion NMR experiments.18 Acknowledgment. This work was supported by the Ministerio de Ciencia y Tecnologı´a of Spain (MAT 2004-4793-C02-02) and by the UniVersity of Alcala´ (UAH GC2006-008). Supporting Information Available: Table listing detailed composition of the samples; figures of yield-stress determination (i) and tixotropic behavior of both Lam1 (ii) and Lam2 (ii) phases. This material is available free of charge via the Internet at http://pubs.acs.org. LA701539F