1948
J. Phys. Chem. B 2009, 113, 1948–1953
Factors Controlling the Stability of a Kinetically Hindered Lamellar-Lamellar Transition Gemma C. Shearman,† Stephane Ugazio,‡ Laurent Soubiran,‡ John Hubbard,‡ Oscar Ces,† John M. Seddon,† and Richard H. Templer*,† Department of Chemistry, Imperial College London, Exhibition Road, South Kensington, London SW7 2AZ, United Kingdom, and UnileVer Research and DeVelopment, Port Sunlight, Quarry Road East, Bebington, Wirral CH63 3JW, United Kingdom ReceiVed: September 9, 2008; ReVised Manuscript ReceiVed: December 19, 2008
We show that we can manipulate the stability of a metastable gel phase, either to enhance its transitory nature or to “lock” it in. Using simple additives such as salt and fatty alcohol we were able to examine both the long-range effect, acting between charged bilayers, and short-range effects on the metastability. We found that the addition of salt to the cationic surfactant diethanolamine ester dimethyl ammonium chloride destabilized the gel phase, and at high concentrations it was able to decrease the length of time taken for the gel phase to revert to a hydrated solid “coagel” phase by an order of magnitude. The growth of the coagel phase was also found to be affected by increasing salt concentration, changing from needle-like (1D) to spherical growth. In contrast to the marked destabilization of the gel phase by salt, the addition of 1-octadecanol was found to prolong the lifetime of the gel phase almost indefinitely by disrupting the short-range packing between the surfactant molecules. This suggests that counterion binding plays a major role in the stability of metastable lamellar gel phases. Introduction Self-assembling amphiphiles that form multilamellar stacks when mixed with water range from biologically relevant phospholipids to synthetic double-chained surfactants, which are used in a myriad of industrial applications from hair care products to ore flotation and emulsion stabilizers.1 The most widespread self-assembled lyotropic structure is the fluid lamellar, LR, phase, although there are other lamellar phases, such as the gel, Lβ, phase, which is a so-called “rotator” phase with long axis rotation, but little or no lateral diffusion. Nonlamellar phases for example, the inverse hexagonal, HII, phase and the inverse bicontinuous cubic phases (QIIP, QIID, QIIG), are also commonly observed. Thermodynamically unstable gel phases formed by cooling fluid lamellar phases below the Krafft boundary have been reported both for ionic surfactant/water systems2-4 as well as for some uncharged5 and zwitterionic systems such as the phosphatidylcholines,6 phosphatidylethanolamines,7 and phosphatidylglycerols.8 Over time, these metastable phases undergo a transition to a “coagel” phase (also described as a hydrated crystal phase), where the stability of the gel phase has been shown to be dependent on many factors, ranging from the temperature and concentration5 through to the counterion for charged surfactant systems.3 The fact that these gel phases are relatively long-lived has been attributed to the onset of significant steric repulsion at the bilayer surface arising from increased motion of the polar headgroups of the amphiphiles. The osmotically driven collapse of lamellar stacks, albeit from a fluid-lamellar to condensed fluid-lamellar (LR f LR′) phase with a small water layer thickness, has been observed for the synthetic double-chained surfactant didodecyldimethylammonium bromide.9 However, the transition was not seen for either * To whom correspondence should be addressed. E-mail: r.templer@ imperial.ac.uk. † Imperial College London. ‡ Unilever Research and Development.
chloride or iodide counterions, where for the former case only the swollen LR phase was seen, and for the latter only the collapsed LR′ phase. This observation was subsequently successfully modeled by assuming that only a certain fraction of the counterions dissociate from the charged surfactant bilayers.10 Further, Harries et al. have also predicted that the effect of high salt concentrations is to reduce or eliminate the phase transition.10 The electrolyte concentration has long been known to affect the behavior and properties of many different amphiphilic systems, including glycerolipids found in cell membranes. Charged membrane systems are particularly affected by the addition of electrolytes, as the interbilayer forces can be substantially altered. For bilayers comprising a cationic surfactant the electrostatic interactions between the positively charged surfaces can extend a great distance when separated by pure water, that is, the Debye length is long, which can result in highly “swollen” phases where the distance between neighboring bilayers can extend over tens or even hundreds of nanometers.9 However, if negatively charged counterions are added, they will act as a screen, causing the range of electrostatic interactions to diminish11 and the interlamellar spacing to decrease as illustrated by Eccleston et al. for cetrimide, a typical cationic surfactant, mixed with fatty alcohol.12 If the concentration of counterions is high enough, these forces will become effectively negligible and the Debye length so short that the interactions between the bilayers will be dominated by hydration forces.13 Structural parameters of amphiphiles such as the area per headgroup have been found to be affected by the addition of salt.14,15 Aroti et al.15 have argued that both ion screening and ion binding have roles in influencing the electrostatic interactions and, subsequently, the structural parameters. We have previously reported the existence of a metastable gel phase for three structurally similar dichained cationic surfactants and were able to show that there was a possible link between the degree of counterion dissociation from the inter-
10.1021/jp807998d CCC: $40.75 2009 American Chemical Society Published on Web 01/27/2009
Stability of Metastable Lamellar Gel Phases
Figure 1. The chemical structure of diethanolamine ester dimethyl ammonium chloride (DEEDMAC).16
facial region of the bilayers (and its associated Debye length) and the instability of the gel phase, which converted to a coagel phase via a nucleation and growth mechanism. Here, we examine the effect of the addition of an electrolyte to one of the three surfactants previously studied, diethanolamine ester dimethyl ammonium chloride (DEEDMAC) (structure shown in Figure 1), on the stability of the gel phase to test our hypothesis that the stability of the gel phase is related to the repulsive forces between the bilayers. If true, then screening the bilayer repulsions would be expected to lead to more rapid coagel formation. In addition, we also investigate the effect that the addition of an uncharged amphiphile such as a fatty alcohol to DEEDMAC has on the kinetics of the system. Experimental Section Materials. DEEDMAC was purified at Unilever Research and Development, Port Sunlight and subsequently dissolved in chloroform/methanol (v/v 3:1) then lyophilized under vacuum to remove any excess water. We assume that the crystal monohydrate form was obtained after lyophilization.16 The 1-octadecanol was obtained from Fluka AG, and the sodium chloride from Sigma Aldrich, both with a purity g99.9%. HPLC grade 99.9% purity chloroform was purchased from SigmaAldrich, and both Analar grade methanol and HPLC grade water were from BDH. DEEDMAC/1-octadecanol mixtures were prepared by codissolution in chloroform/methanol (v/v 3:1) followed by lyophilization described above to remove the solvents. Samples were hydrated by the addition of weighed amounts of solution, specifically either a NaCl solution or pure water, then sealed and heated above their melting (Lβ f LR) transition. The samples were then immediately extruded through a narrow hole (1.0 mm) by centrifugation a number of times until homogenized. Conveniently, the centrifugation also ensures the elimination of any trapped air bubbles, and the inherent shearing action guaranteed that, after preparation, the samples were in the coagel phase. All samples were then left at room temperature to equilibrate for at least 24 h before any measurements were run. X-ray Diffraction. X-ray measurements, both wide- and small-angle, were made on a GX-20 rotating anode generator (Enraf-Nonius, Netherlands). The Cu KR X-ray beam was focused by double-mirror Franks optics to a point (∼190 µm diameter). Diffraction was then recorded at one of two different distances away from the sample by a 2D image-intensified charge-coupled device (CCD) detector. Both the automated acquisition of data as well as data analysis were performed using software known as “TV4” developed by S. Gruner (Cornell University, U.S.A.) and E. F. Eikenberry (Paul-Scherrer Institute, Switzerland) and adapted for use at Imperial College London. The samples were held within a copper sample holder, where the temperature was controlled by Peltier thermoelectric heater/ coolers and determined by a platinum resistance (Pt100) temperature sensor to an overall accuracy of ( 0.03 °C. All beamline elements, that is, the sample chamber, optics chamber,
J. Phys. Chem. B, Vol. 113, No. 7, 2009 1949 and the flight tube, were evacuated to reduce background scattering. Calibrations were performed using silver behenate (d001 ) 58.38 Å), and the precision of subsequent lattice spacing measurements was determined to be within 1 Å. For the kinetic measurements, each sample was initially heated to 75 °C, held at that temperature for 10 min, and X-ray diffraction images were taken to confirm that the sample was in the fluid lamellar phase. This ensured that any nucleation sites within the sample were destroyed. The sample was then cooled to 25 °C, within a 5 min time frame. Successive images were then captured over a predetermined length of time, dependent on the NaCl concentration or molar ratio of fatty alcohol. Differential Scanning Calorimetry. A Perkin-Elmer DSC7 differential scanning calorimeter was used to determine the phase transition temperatures of the coagel-gel and gel-fluid lamellar transitions. The temperatures of transitions were taken as the peaks, rather than the onsets as previously done,16 since extrapolation would have been necessary for the gel-fluid lamellar transition measured for the first heating scan. The calibrant used was indium (mp 156.60 °C; ∆Hfusion ) 28.51 J/g). Each sample (of weight ∼5 to 10 mg) underwent a heating and cooling cycle and then a further heating scan, between 0 and 80 °C, at a rate of 10 °/min. Reference pans consisting of a known quantity of water were used throughout the series of measurements, and the error was estimated to be ( 1 °C. Results and Discussion Variation in Salt Concentration. As mentioned earlier, the introduction of negatively charged ions, by the addition of salt, to positively charged membranes has a profound effect on the interbilayer forces by screening the electrostatic repulsive interactions. This can have repercussions for amphiphilic phase behavior, such as the shifting of transition temperatures. Typically, the gel-fluid lamellar transition temperature is expected to increase upon the addition of salt; however, for some cationic surfactants the transition temperature has instead been reported to have remained constant17 or even decreased18 with added salt, illustrating that the screening argument may be too simplistic and that other considerations such as headgroup hydration, structural changes, and inefficient electrostatic screening should not be neglected. Similarly, the addition of increasing amounts of NaCl to ascorbyl-6-O-dodecanoate (ASC12), a single-chained anionic surfactant, reduced its coagel-gel transition temperature.19 To examine the effect of addition of an electrolyte on the coagel-gel and gel-fluid lamellar phase transition temperatures of DEEDMAC, we performed DSC scans as a function of NaCl concentration. NaCl was chosen due to the common counterion, which eliminated any complications from the possible exchange of ions at the hydrocarbon-water interface. The hydration of each sample was kept constant at 66 wt %. In each case, the first heating scan from 0 to 80 °C showed two peaks, with the peak at the lower temperature corresponding to the coagel-gel transition temperature and the higher temperature peak to the gel-fluid lamellar transition. Further heating scans only showed the gel-fluid lamellar transition, due to the relatively long-lived metastability of the gel phase, as previously described16 in further detail and shown in Figure 2. No enthalpic observations of the coagel-gel transition will be made here as it was impossible to ensure that complete conversion into the coagel phase had occurred during the preparation of the sample before the first heating scan, although each sample was inspected
1950 J. Phys. Chem. B, Vol. 113, No. 7, 2009
Figure 2. The DSC trace of DEEDMAC/brine ([NaCl] ) 10 mM) showing both initial heating and cooling curves (solid line) as well as the second heating curve (dotted line) where the peaks and troughs denote endothermic and exothermic transitions respectively, and the regions A, B, and C represent the coagel, gel, and fluid lamellar phases, respectively, in which the sample exists.
Figure 3. The second heating DSC curve of DEEDMAC/brine at three different NaCl concentrations, illustrating the lack of variation in the gel-fluid lamellar phase transition temperature.
Figure 4. The temperature difference between the two peaks that represent the coagel-gel and gel-fluid lamellar transitions as a function of the NaCl concentration. The horizontal dotted line represents the temperature difference obtained in pure water, i.e. [NaCl] ) 0 M.
visually during equilibration to check that there were no translucent regions indicative of the gel phase. The transition temperature for the gel-fluid lamellar phase was found to be 58 °C ((1 °C), and was not found to vary with NaCl concentration (Figure 3), which is in agreement with previous findings reported by Nascimento et al. for similar dichained cationic surfactants.17 However, the coagel-gel phase transition temperature was found to vary significantly, and the difference in temperature (∆Τdiff, Figure 2) between the two transitions with respect to the NaCl concentration is plotted in Figure 4, removing any inaccuracies caused by the variation in weights and therefore reducing the error. From Figure 4, it is clear that by decreasing the NaCl concentration of the solution the difference between the two phase transition temperatures becomes more marked. Because the temperature of the gel-fluid lamellar phase transition is
Shearman et al. invariant with the NaCl concentration, this shows that the transition temperature for the coagel-gel transition decreases with decreasing NaCl concentration. This can be understood by considering the number of ions in the system. By decreasing the NaCl concentration, the total number of chloride ions is decreased, which in turn means that fewer ions are absorbed at the interface. This stabilizes the gel phase, meaning that the transition at which it converts into the gel phase is reduced. These findings are at odds with those found for ASC12, but this can be simply explained by the difference in nature of the headgroups. ASC12 is an anionic surfactant, whereas DEEDMAC is cationic. Lo Nostro et al.19 have shown that anions absorb strongly onto anionic surfaces, a surprising finding, however they also showed that this destabilized the anionic surface, resulting in a lower coagel-gel phase transition temperature with increased ionic strength. Here, we have the opposing situation, with anions absorbing onto the cationic surface, stabilizing it instead. Kinetics. The metastability of the DEEDMAC gel phase has been reported previously16 and during the course of these studies we observed visually that the addition of salt to the translucent gel phase appeared to hasten the collapse into the opaque coagel phase. To test this hypothesis, we examined the kinetic stability of the gel as a function of the NaCl concentration using X-ray scattering. Each sample was initially held at a temperature above that of the gel-fluid lamellar transition, and the interlamellar spacing was found to be relatively constant at 115 Å with respect to the NaCl concentration, with the exception of those samples with [NaCl] g 0.5 M, which were found not to melt into the fluid lamellar phase. All samples that were found to undergo a gel-fluid lamellar phase transition were then quenched to 25 °C, and successive SAXS/WAXS images were taken as a function of time, an example of which is shown in Figure 5a. Although no small-angle data was used to determine any kinetic information due to overlapping peaks of the gel and coagel phases, the combined SAXS/WAXS patterns revealed that the reduction of the water layer between lamellar sheets, giving a collapsed phase with a repeat spacing of approximately 55 Å, did not necessarily occur concurrently with the reduction of rotational movement within the chain region, as indicated by the appearance of more than one sharp peak within the WAXS region. For high salt concentrations (>10 mM), the disparity was marked with a sharp peak of low d-spacing appearing immediately after quenching to 25 °C in the SAXS region, but no sharp peaks were visible in the WAXS pattern for over 2 h. With decreased NaCl, SAXS experiments revealed that the pattern of the metastable gel phase was less distinct and only a diffuse ring could be seen immediately after cooling to 25 °C (Figure 6b), which is indicative of a lack of periodic structure implying that some regions of the sample have less water between the bilayer sheets than other areas. However, at low salt concentrations a more homogeneous SAXS diffraction pattern was observed (Figure 6c). It is plausible that the number and frequency of nucleation sites is a factor in this behavior, with a large number of nucleation sites occurring immediately after cooling for high salt concentrations but few forming for the low salt samples. The exact determination of the kinetic relaxation times for varying salt concentrations was found to be impossible due to the difficulty in reproducing the same conditions for each sample. For example, the prevalence of nucleation sites on the glass capillary surface for certain samples induced a much faster collapse to the coagel phase than might be expected. The irreproducibility of the kinetics for similar systems such as
Stability of Metastable Lamellar Gel Phases
J. Phys. Chem. B, Vol. 113, No. 7, 2009 1951
Figure 7. The time constant, τ, for the gel-coagel phase transition (filled squares; left axis) together with the dimensionality, n, of the growing coagel domains (empty circles; right axis) as a function of the logarithm of the NaCl concentration.
peak chosen for analysis lies at ∼0.26 Å-1 (indicated by the arrow). For each sample, the intensity profile with respect to time of this WAXS peak was normalized and fitted as a sigmoidal curve following Kolmorov-Avrami kinetics of crystallization21 (Figure 5b), expressed as:
X(t) ) 1 - exp[-(Kt)n+1]
Figure 5. (a) A series of X-ray diffraction patterns for DEEDMAC/ brine (where [NaCl] ) 40 mM), showing the broad wide-angle gel peak at ∼0.24 Å-1 gradually disappearing and two sharp coagel peaks at ∼0.22 Å-1 and ∼0.26 Å-1 appearing over time; and (b) the normalized intensity of the WAXS peak at ∼0.26 Å-1 and its Kolmogorov-Avrami best fit (solid line), described later.
Figure 6. Three SAXS images obtained for DEEDMAC/brine samples shortly after quenching from the fluid lamellar phase down to 25 °C, where the NaCl concentrations are (a) 0.4, (b) 0.01, and (c) 0.0001 M.
DODMAC has been documented previously,20 and DEEDMAC was found to be equally problematic. Another issue arose for a number of samples, where the broad gel peak overlapped slightly with the sharp coagel peaks in the coexistence region. Therefore, any samples that were found to have anomalously low values or alternatively had any overlapping WAXS peaks were excluded from analysis. As mentioned earlier, the kinetics of the gel-coagel phase transition were determined using the two WAXS peaks that were unique to the coagel phase. Figure 5a depicts the changing X-ray pattern for DEEDMAC/NaCl solution over time with [NaCl] ) 40 mM, where the WAXS
(1)
where X(t) here refers to the normalized intensity, n is the effective dimensionality of the growing domains, and K is the rate constant of the transformation. The time constant for the phase transformation, τ, is simply the inverse rate constant, K-1, and the error was estimated to be (10% due to the possibility of pre-existing nucleation sites. τ is plotted in Figure 6 as a function of the NaCl concentration, together with the dimensionality n. It should be noted that eq 1, the basic Avrami equation, was felt to be sufficient to fit the intensity profiles for each of the samples, rather than a modified model,22 since no obvious changes in the growth kinetics (i.e., kinks in the sigmoidal curve) were observed during the course of the phase transitions. From Figure 7, it is clear that there is a trend for an increasing time constant for the gel-coagel phase transformation with less salt. Interestingly, however, there also seems to be a trend for an increasing dimensionality for higher salt concentrations, that is, the growth of the coagel domains goes from rod-like at low NaCl concentrations, through to plate-like, and finally spherical at high salt concentrations. Previously,16 we found that the dimensionality of the coagel domain growth was equal to 1 for DEEDMAC in pure water, by fitting the change in the lattice parameter of the gel phase over time, corroborating our current results in the low salt concentration regime. It has been reported recently by Chodankar et al.23 that the rate at which the protein lysozyme crystallized from solution was vastly increased upon the increase of salt in the solution whereas the Avrami exponent, n, for this crystallization process was independent of the salt concentration, and needle-like growth of the protein crystals was routinely observed. For the DEEDMAC system, the dependence of the Avrami exponent on the salt concentration suggests that the increased screening of charges for neighboring bilayers at higher NaCl concentrations encourages interbilayer coupling, which in turn would promote the growth of crystalline domains across bilayers, that is, three-dimensional crystal growth, rather than being confined within a single bilayer. A similar effect has been seen for a monoglyceride/water system where the dimensionality of the growing domains was found not only to depend on the concentration of the monoglyceride but also the storage temperature,5 explained by a change in
1952 J. Phys. Chem. B, Vol. 113, No. 7, 2009
Shearman et al.
Figure 8. A DSC scan showing the first heating and cooling curves for (a) DEEDMAC/1-octadecanol (xC18OH ) 0.29) and (b) DEEDMAC, both in 66 wt % water.
bilayer spacing or radial growth rate respectively. Kinoshita et al.24 have argued that the decrease in initial growth rate of the subgel phase from that of metastable gel for dimyristoylphosphatidylglycerol with increased salt concentration was due to morphological changes of the bilayers, which change from flat to curved bilayer sheets, rather than an effect of an increased number of Na+ ions, which they expected to have the opposite effect. For the DEEDMAC system, we do see an increase in domain growth for an increased NaCl concentration, suggesting that, in this case, the electrostatic screening effect of the Clions is likely to be the dominant reason for the rate of coagel formation increasing with salt concentration. The regulation of the surface charge density of the amphiphilic bilayers is therefore key to controlling the coagel formation kinetics. These results point to a coupling between long- and short-range effects. The dependence of the dimensionality of the coagel growth on the Debye length suggests that local interactions between the DEEDMAC molecules may be important, and in order to test this further, we examined the effect that the addition of uncharged amphiphiles to the surfactant bilayers had on the kinetics. Addition of 1-Octadecanol. The addition of a single fatty alcohol into an amphiphilic monolayer will alter the local environment in which it is situated, and therefore the phase behavior of the amphiphile may also vary. Indeed, long-chain fatty alcohols have been reported to extend the region over which lyotropic liquid crystalline phases can be found, whereas the opposite has been found for short-chain alcohols.25 An additional benefit of using a long-chain fatty alcohol such as 1-octadecanol rather than a short-chain alcohol is that there would be little or no chain disordering, a common phenomenon caused by short-medium chain-length fatty alcohols.26 By choosing a fatty alcohol that had exactly the same chain length and saturation as the surfactant to which it would be added, we hoped to minimize any significant variations in the lateral forces within the chain region. This hypothesis was supported, as we did not observe any change in the phases formed, that is, during the course of the experiments, only the coagel, gel, and fluid lamellar phases were observed. However, it is likely that the addition of a fatty alcohol into the surfactant bilayer will cause the headgroup region to be altered profoundly due to the different sizes, charges, etc. of the headgroups, and this in turn will have repercussions for the steric, hydration, and electrostatic interactions. DEEDMAC samples with fatty alcohol compositions between 8 and 62 mol % were prepared and examined as previously, both by DSC and X-ray. A typical DSC scan of DEEDMAC/1-octadecanol (xC18OH ) 0.29) is shown in Figure 8 together with DEEDMAC in 66 wt % water, where in both cases two peaks are seen in the heating curves, whereas only one can be seen in each of the cooling curves. The gel-fluid-
Figure 9. The variation of the SAXS pattern over time for DEEDMAC/ 1-octadecanol (xC18OH ) 0.24) in 66 wt % water.
lamellar transition was found to have decreased slightly upon the addition of fatty alcohol. Interestingly, previous studies of the influence of fatty alcohols on the chain-melting transition temperature of double-chained cationic amphiphile have shown that at low fatty alcohol concentrations the transition temperature is reduced, but at high concentrations the situation is reversed, although no explanation was hypothesized for this behavior.27 X-ray measurements revealed that as opposed to the addition of NaCl, discussed above, that caused the metastable gel to relax into the stable coagel form, over the time-scale examined (up to 90 h) complete relaxation into the coagel phase was not observed (i.e., a significant proportion of the Lβ phase was always in existence) for any of the samples with 1-octadecanol (where the molar percentage of 1-octadecanol varied from 15.2 to 50%). Therefore, no kinetic information was able to be extracted from the samples studied, suggesting that the DEEDMAC/1-octadecanol samples were significantly more stabilized in the gel phase than the pure DEEDMAC samples. As an illustration of this, Figure 9 shows the variation in SAXS pattern over time for DEEDMAC/1-octadecanol (xC18OH ) 0.24), where the small-angle Bragg peaks of the gel phase gradually decrease over 70 h but do not entirely disappear. Samples with high molar ratios of fatty alcohol were visually observed to remain translucent, even after many months of storage. We suggest that the gel phase can be considered essentially as stable for these systems, since DEEDMAC will, over such a time-scale, hydrolyze to a certain degree in the presence of water. Therefore, the 1-octadecanol appears to have disrupted the local packing and average distance between DEEDMAC molecules, lowering the ability of the counterions to bind to the positively charged headgroups by forming ion pairs between neighboring ammonium groups. Conclusions The effect of altering the electrolyte concentration has been found to have a profound effect on the rate of collapse of the metastable gel phase of DEEDMAC. The time taken for this phase change to occur can be reduced in some cases by over an order of magnitude, dependent on the salt concentration, and hence also the interbilayer interactions. The reverse was found to be true for the addition of a fatty alcohol to DEEDMAC, where for a high 1-octadecanol content, the gel phase persists for many months due to the disruption of the local packing of DEEDMAC molecules. However, more generally, the kinetics of transformation from the gel to coagel phase occur over a
Stability of Metastable Lamellar Gel Phases relatively long time-scale, even for very high salt concentrations, implying that the energy barrier to overcome in order to extrude water from between the bilayers is high. Acknowledgment. We thank Unilever plc. for funding G. C. S. as well as the U.K. Engineering and Physical Sciences Research Council (EPSRC) for the Platform grant GR/S77721. References and Notes (1) Chemistry and Technology of Surfactants; Fern, R., Ed.; Blackwell Publishing: Oxford, 2006. (2) Kodama, M.; Kunitake, T.; Seki, S. J. Phys. Chem. 1990, 94, 1550. (3) Kodama, M.; Tsujii, K.; Seki, S. J. Phys. Chem. 1990, 94, 815. (4) Laughlin, R. G.; Munyon, R. L.; Fu, Y. C.; Fehl, A. J. J. Phys. Chem. 1990, 94, 2546. (5) Cassin, G.; de Costa, C.; van Duynhoven, J. P. M.; Agterof, W. G. M. Langmuir 1998, 14, 5757. (6) Chen, S. C.; Sturtevant, J. M.; Gaffney, B. J. Proc. Natl. Acad. Sci. U. S. A. 1980, 77, 5060. (7) Seddon, J. M.; Harlos, K.; Marsh, D. J. Biol. Chem. 1983, 258, 3850. (8) Koynova, R. Chem. Phys. Lipids 1997, 89, 67. (9) Dubois, M.; Zemb, T.; Fuller, N.; Rand, R. P.; Parsegian, V. A. J. Chem. Phys. 1998, 108, 7855. (10) Harries, D.; Podgornik, R.; Parsegian, V. A.; Mar-Or, E.; Andelman, D. J. Chem. Phys. 2006, 124, 224702.
J. Phys. Chem. B, Vol. 113, No. 7, 2009 1953 (11) Israelachvili, J. N. Intermolecular and Surface Forces: With Applications to Colloidal and Biological Systems, 2nd ed.; Academic Press: London, 1992. (12) Eccleston, G. M.; Behan-Martin, M. K.; Jones, G. R.; TownsAndrews, E. Int. J. Pharm. 2000, 203, 127. (13) Rand, R. P.; Parsegian, V. A. Biochim. Biophys. Acta 1989, 988, 351. (14) Dubois, M.; Zemb, T. Langmuir 1991, 7, 1352. (15) Aroti, A.; Leontidis, E.; Dubois, M.; Zemb, T. Biophys. J. 2007, 93, 1580. (16) Shearman, G. C.; Ugazio, S.; Soubiran, L.; Hubbard, J.; Ces, O.; Seddon, J. M.; Templer, R. H. J. Colloid Interface Sci. 2009, 331, 463. (17) Nascimento, D. B.; Rapuano, R.; Lessa, M. M.; Carmona-Ribeiro, A. M. Langmuir 1998, 14, 7387. (18) MacDonald, R. C. Biophys. J. 1999, 77, 2612. (19) Lo Nostro, P. Langmuir 2003, 19, 9583. (20) Laughlin, R. G.; Munyon, R. L.; Fu, Y. C.; Emge, T. J. J. Phys. Chem. 1991, 95, 3852. (21) Yang, C. P.; Nagle, J. F. Phys. ReV. A 1988, 37, 3993. (22) Narine, S.; Humphrey, K.; Bouzidi, L. J. Am. Oil Chem. Soc. 2006, 83, 913. (23) Chodankar, S.; Aswal, V. K.; Kohlbrecher, J.; Hassan, P. A.; Wagh, A. G. Physica B: Condensed Matter 2007, 398, 164. (24) Kinoshita, M.; Kato, S.; Takahashi, H. Chem. Phys. Lipids 2008, 151, 30. (25) Zana, R. AdV. Colloid Interface Sci. 1995, 57, 1. (26) Miller, K. W.; Firestone, L. L.; Alifimoff, J. K.; Streicher, P. Proc. Natl. Acad. Sci. U. S. A. 1989, 86, 1084. (27) Klijn, J. E.; Engberts, J. B. F. N. Org. Biomol. Chem. 2004, 2, 1789.
JP807998D
