5686
2009, 113, 5686–5689 Published on Web 04/03/2009
Formation Process of Bilayer Gel Structure in a Nonionic Surfactant Solution Youhei Kawabata,* Akimi Matsuno, Tomoaki Shinoda, and Tadashi Kato Department of Chemistry, Tokyo Metropolitan UniVersity, Hachioji, Tokyo 192-0397, Japan ReceiVed: December 11, 2008; ReVised Manuscript ReceiVed: February 6, 2009
We investigated the structure and formation process of lamellar domains below the Krafft temperature in two nonionic surfactants, the C16E7/water and C16E6/water systems, using an optical microscope and small-angle X-ray scattering. We found that vesicles and long leek-like lamellar domain structures are formed in the C16E7 and C16E6 systems, respectively. This large difference between the lamellar domain structures of the systems can be explained by the elastic properties of bilayers in the structural formation process. In a binary system consisting of an ionic surfactant and water, crystalline hydrated solids are deposited when the temperature decreases below the Krafft temperature, which corresponds to the melting point of the hydrophobic parts of surfactants and represents temperature-dependent equilibrium between dissolved monomers (surfactants) and the hydrated solids.1,2 In general, the hydrated solids are considered to coexist with excess water and a lamellar phase where hydrophobic tails of bilayers are a “solid-like” gel (all-trans). In this Krafft transition, metastable structures have been reported. For bilayer structures on the nanometer scale, Sasaki has found metastable crystalline lamella (Lβ) in the Krafft transition of aqueous cetylpyridinium chloride solutions.3 With the addition of alcohol to the binary system, the mixtures are sometimes strongly turbid throughout the sample without macroscopic phase separation, and they are highly viscous, a property that is applied to the cosmetic industry.4 From the viewpoint of this viscoelastic property, Yamagata et al. have shown time-dependent viscoelasticities in a cetyltrimethylammonium chloride/cetyl alcohol/water system.5 These types of nonequilibrium phenomena are probably caused by structural transformations over a wide range of length scales from sub-nm to µm, i.e., the transition from micelles to bilayers, the growth of the Lβ domains by stacking of bilayers, and phase separation into the Lβ phase and excess water. Therefore, to clarify the formation mechanism of the metastable structures, it is important to investigate the change in structure at nm to µm scales. We have found that solutions similar to the ionic surfactant solutions described above could be obtained in nonionic surfactant solutions, the C16E7/water system and C16E6/water system (C16H33(OC2H4)nOH). Below the Krafft temperature in both systems, aqueous solutions of C16E7 or C16E6 are also turbid without macroscopic phase separation and show different viscoelasticities in each system. Small-angle X-ray scattering (SAXS) experiments revealed that the structures of the C16E7 system are different from those of the C16E6 system despite the slight difference in hydrophilic length between these two * To whom correspondence should be addressed. E-mail: youheik@ tmu.ac.jp.
10.1021/jp810911y CCC: $40.75
surfactants. This implies that the formation processes of the Lβ phase in the two systems are different. In this study, to clarify mechanisms, the formation mechanisms of the hydrated solid phase in each system, we have performed time-resolved SAXS and optical microscope experiments after the temperature (T) quench to the hydrated solid phase of the C16E7/water and C16E6/water systems. We report that large differences between the two systems are found in the transition process at the nm to µm scale. Samples were prepared by mixing C16E6 or C16E7 with D2O. C16E6 and C16E7 were purchased from Nikko Chemicals, Inc., in crystalline form and used without further purification. Deuterated water (D2O of 99.9 at%-D) purchased from Isotec, Inc., was used after nitrogen bubbling to avoid oxidation of the ethylene oxide group of surfactants.6 The surfactant concentrations were fixed at 10 wt %. The Krafft temperatures of the C16E6 and C16E7 systems are 26 and 12 °C, respectively. The temperature was quenched from 28 to 20 °C in the C16E6 system and from 16 to 8 °C in the C16E7 system. The actual cooling rate was set to 4.5 °C/min for both experiments. We observed the time evolution of the structures after T-quench using an optical microscope and SAXS. For optical microscope observations, we used a BX-51 (Olympus Inc.). Temperature was controlled by using a hotstage FP-90 instrument (Mettler Inc.). The sample thickness was less than 150 µm. The microscope images were taken with a digital video camera. Time-resolved SAXS experiments were performed using the synchrotron radiation SAXS spectrometer installed in BL-15A at the Photon Factory (PF) of the High Energy Accelerator Research Organization (KEK), Tsukuba. The scattering X-rays were detected with a CCD camera. The scattering vector q (q ) 4π sin θ/λ, where 2θ is the scattering angle and λ is the X-ray wavelength) range is from 0.05 to 0.25 Å-1. The energy resolution (∆E/E) is 3 × 10-3. The scattering intensity was measured every 30 s with an exposure time of 5 s. The sample cell is made of copper with Kapton windows, and the temperature of the sample was controlled using the DTA/SAXS instrument with a precision of (0.1 K.7 Figure 1a shows an optical microscope image in the C16E7 system at 1200 s after T-quench. Round structures can be clearly 2009 American Chemical Society
Letters
Figure 1. (a) Microscope image of phase contrast mode at 1200 s after T-quench in the C16E7 system. Round structures can be clearly observed; their average size is about 10 µm. Under crossed Nichols, the Maltese cross is observed only in the husk part. From this result, we can imagine the round structures as illustrated. (b) A plate-like lamellar domain closes gradually to form a vesicle. Numbers indicate the elapsed time after T-quench. The bar in the figure corresponds to 5 µm. Combining the results of the SAXS and microscope experiments, we can illustrate the formation process of lamellar domains as described in i-iv. From i to ii, micellar structures can not keep their formations and collapse because hydrophobic tails extend; the intensity due to the micellar structures become weak, and monomers whose hydrophobic tails are extended aggregate to form bilayers and plate-like domains. After the plate-like lamellar domains form, they finally close gradually and form vesicles (iii to iv).
observed. As shown in the polarizing microscope image in Figure 1b, only the shell part of the round structures glittered. Therefore, these are vesicles that have lamellar structures in the shell part, as described in the schematic picture (Figure 1a). In general, multilamellar vesicles, or onions, where the inside of the vesicles is clogged with bilayers, are often found in surfactant or lipid solutions. However, the above-mentioned hollow vesicles have never been reported. Furthermore, we observed the growth process of these vesicles using the microscope, as shown in Figure 1b. Before forming a vesicle, a “plate-like” lamellar domain was formed. It closed slowly and finally became spherical. This type of vesicle formation process resembles the “standard” pathway,8 which is found in the case of unilamellar vesicles.9-13 According to this “standard” pathway, amphiphilic molecules first self-assemble into spheres and rods (micelles). They transform themselves to bilayers whose hydrophobic tails are in a liquid state, and finally, the bilayers bend around and close up to form vesicles. In our case, however, the hydrophobic tails are extended at first, followed by the formation of the plate-like multilamellar domains, where the hydrophobic tails of bilayers are interdigitated (gel state, Lβ), and then, they surround excess water. The type of formation process in Figure 1b is driven by the rim
J. Phys. Chem. B, Vol. 113, No. 17, 2009 5687
Figure 2. (a) Polarizing microscope image at 300 s after T-quench in the C16E6 system. Worm-like lamellar domain structures can be found; their thickness is ∼1 µm. The bilayers are considered to be stacked as a long leek (illustration). (b) Bright field images of the worm-like lamellar domains at 120 and 1800 s. Although the structure in the C16E7 system transforms dramatically after T-quench, the morphology of the structure does not change much in the C16E6 system.
energy of the disk-like bilayers and/or lamellar domains, as reported in previous studies.14 We also examined lamellar domains in the C16E6 system, for comparison with those in the C16E7 system. Figure 2a is a microscope image of the C16E6 system. We can observe wormlike lamellar domain structures, which are connected with each other to form a network structure. According to the polarizing microscope image, light is transmitted from strings, as shown in Figure 2a. Bilayers would be stacked in the worm-like lamellar domains as a long leek, shown in Figure 2a. Figure 2b shows typical images at 120 and 1800 s after T-quench. Although the structure in the C16E7 system transforms dramatically but slowly after T-quench, worm-like lamellar domains form rapidly in the C16E6 system, and they do not change much with time. It is clear that the structures in the C16E6 system are quite different from those in the C16E7 system. This structural difference emerges in each system’s viscoelastic properties. In the C16E7 system, the hydrated solid is viscous, while it is elastic in the C16E6 system. The elastic property of the C16E6 system is due to the entangled network structures of the worm-like lamellar domains. On the other hand, as vesicles are dispersed in the C16E7 system, which can be classified as a so-called dispersion gel, it is viscous, and its elastic property is much less than that in the C16E6 system. Furthermore, to clarify this large difference between the lamellar domain structures in these two systems, we investigated the formation process of bilayers of Lβ coexisting with excess water. Figure 3 shows the evolution of SAXS profiles in these two systems. The numbers in the figure correspond to the elapsed time after T-quench. The broad peaks at 0 s correspond to the form factor of micelles. After T-quench, Bragg peaks due to the lamellar structures of the Lβ phase appear and sharpen
5688 J. Phys. Chem. B, Vol. 113, No. 17, 2009
Letters
Figure 3. Time evolution of the SAXS profiles in the C16E7/water and C16E6/water systems. The broad peaks are due to micellar structures at 0 s. After T-quench, Bragg peaks corresponding to lamellar structures of the Lβ phase appear. Numbers indicate the elapsed times from T-quench. Lines are fitting results to eq 1.
with time. To quantify the formation process, we have analyzed the SAXS profiles using the equation
Itotal ) aImicelle + bIlamellar +
c q2Ξ2 + 1
(1)
Itotal is the total scattering intensity, Imicelle is the intensity due to the micellar structures that exist for a while after T-quench, and Ilamellar is the intensity corresponding to the lamellar structures. a, b, and c are the prefactors of each term, and a/(a + b) corresponds to the volume fraction of the micellar phase. The prefactor c is the scattering intensity due to inhomogeneity of the structures at g100 Å scale. Here, to express the increase of the low-q intensity in the C16E7 system, we introduced an Ornstein-Zernike function in which Ξ is the correlation length. For Imicelle, we used the scattering function for ellipsoidal micellar structures15 with the structure factor S(q) ) 1. For Ilamellar, we applied the equation that Nallet et al. proposed to explain the scattering data for lamellar bilayer structures:16
Ilamellar )
2π P(q) S(q) d q2
(2)
Here, q is the scattering vector, P(q) the form factor for bilayers, S(q) the structure factor for a lamellar structure, and d the repeat distance between bilayers. P(q) is written as
P(q) )
[ {(
)}
δh 4 F sin q + δe w 2 q2
( )]
+ Fh sin q
δh 2
2
(3)
where Fw is the scattering length density difference between the hydrophilic segment and water and Fh is the scattering length density difference between the hydrophilic segment and the hydrophobic tail, δh is the hydrophobic tail length, and δe is the hydrophilic segment length. In this analysis procedure, we especially focused on the parameters d and δh. Figure 4a shows the time evolution of the repeat distance d of the lamellar structure (Lβ), which coexists with excess water in the hydrated solid phase. The broad horizontal line indicates d at the phase boundary of Lβ in each system. The repeat distance of the C16E6 system achieves its equilibrium state soon
Figure 4. Time evolution of the repeat distance d (A), the hydrophobic length δh (B), and the ratio a/(a + b) (C) (see eq 1 and the text) in the C16E7/water and C16E6/water systems. Time 0 corresponds to the start of T-quench. The broad horizontal line in part A indicates the repeat distances at the phase boundary of Lβ. The dashed line in part B corresponds to the value obtained at the Lβ phase. The repeat distance d in C16E7/water tends to be away from the equilibrium values; the hydrophobic length δh is near equilibrium at about 2000 s, when it suddenly decreases. Coincidently, the ratio of a/(a + b), corresponding to the fraction of micelles, decreases to 0 around 2000 s.
after T-quench. On the other hand, the repeat distance of the C16E7 system becomes larger than that at the phase boundary of Lβ. Figure 4b indicates the time evolution of the hydrophobic segment length δh in both systems. The horizontal dashed line corresponds to the segment length around the phase boundary of Lβ (in other words, the equilibrium value). The segment length δh in the C16E6 system gradually goes to its equilibrium value, ∼21 Å, via the local minimum around 1000 s. In the C16E7 system, on the other hand, the segment length suddenly decreases around 2000 s and deviates from its equilibrium value, although it retains the equilibrium value until that point. These imply that the lamellar structure formed in the C16E7 system is not the equilibrium state, while that in the C16E6 system is an equilibrium one. Next, we focus on the coefficients a and b in eq 1. Figure 4c shows the time dependence of the ratio a/(a + b), corresponding to the volume fraction of the micellar phase. It decreases with time and reaches 0 after 2000 s. Therefore, until 2000 s, micelle structures are collapsed by the extension of hydrophobic tails; alternatively, bilayers form followed by the growth of lamellar domains. After 2000 s, micelles make no further contribution to vesicle formations. In the C16E6 system, on the other hand, the ratio a/(a + b) becomes 0 immediately after T-quench. This could indicate that free energy barriers from micelles to lamellar transitions are small in the C16E6 system. Combining these SAXS results with the microscope observations, we can illustrate the formation process of the lamellar domains and vesicle structures, as shown in Figure 1. Just after T-quench, micelles cannot keep their shape and collapse because hydrophobic tails extend; the intensity due to the micellar
Letters structures becomes weak (i), and monomers whose hydrophobic tails are extended aggregate to form bilayers and plate-like domains. After the formation of the plate-like lamellar domains, they finally close gradually and form vesicles. The growth process from i to ii is relatively fast, while the formation process from iii to iv is slow. We have considered that this large structural difference at the µm scale might be due to the formation process of bilayers and lamellar domains, especially between 0 and 1000 s, where lamellar domains begin to be formed. As shown in Figure 4, in the C16E6 system, the hydrophobic tails shrink to about δh ) 16 Å. In the C16E7 system, on the other hand, the hydrophobic tails are extended to δh ∼22 Å until 2000 s after T-quench. If the rigidity of bilayers depends only on the length of the hydrophobic tails, in the C16E7 system, the rigidity of bilayers becomes about 2.5 times larger than that in the C16E6 system after about 1000 s using the relationship κ ∼ δh3. Then, the persistence length of bilayers, ξ ∼ exp(4κ/kBT), in the C16E7 system is about 400 times larger than that in the C16E6 system, if the bending modulus in the C16E6 system is κ ) 1kBT. This may cause the difference in the µm scale structures observed by the optical microscope; hollow vesicles with large diameter (∼20 µm) are formed in the C16E7 system, whereas the C16E6 system exhibits worm-like lamellar domains with a diameter (∼1 µm) much smaller than that in the C16E7 system. In this study, we discovered the morphologies of the lamellar domains of hydrated solids below the Krafft temperature in the C16E7/water and C16E6/water systems, and investigated the structure formation process at the nm to µm scale using optical microscope and SAXS. Vesicles and worm-like lamellar domain structures were found in the C16E7 and C16E6 systems, respectively. This large difference in the structures of these systems
J. Phys. Chem. B, Vol. 113, No. 17, 2009 5689 can be explained in terms of the elastic properties of bilayers in the initial stage after T-quench. Acknowledgment. We thank Prof. H. Tanaka at the Institute of Industrial Science for fruitful discussion, and Prof. M. Imai of Ochanomizu University for the SAXS measurements. This work was supported by a Grant-in-Aid on Priority Area “Soft Matter Physics” from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT). Our SAXS experiments at KEK were performed under approval of the Photon Factory Advisory Committee (Proposal No. 2006G066). Supporting Information Available: Description of how the scattering function of the micellar phase was used. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16)
Shinoda, K.; et al. J. Phys. Chem. 1989, 93, 7216. Kunieda, H.; Shinoda, K. J. Phys. Chem. 1976, 80, 2468. Sasaki, S. J. Phys. Chem. B 2007, 111, 8453. Akatsuka, H.; et al. J. Colloid Interface Sci. 2006, 302, 341. Yamagata, Y.; Senna, M. Langmuir 1999, 15, 4388. Minewaki, K.; et al. Langmuir 2001, 17, 1864. Yoshida, H.; et al. Thermochim. Acta 1995, 264, 173. He, X. H.; Schmid, F. Phys. ReV. Lett. 2008, 100, 137802. Leng, J.; et al. Europhys. Lett. 2002, 59, 311. Weiss, T. M.; et al. Phys. ReV. Lett. 2005, 94, 038303. Weiss, T. M.; et al. Langmuir 2008, 24, 3759. Egelhaaf, S. U.; Schurtenberger, P. Phys. ReV. Lett. 1999, 82, 2804. Nieh, M.-P.; et al. Langmuir 2005, 21, 6656. Iwashita, Y.; Tanaka, H. Phys. ReV. Lett. 2007, 98, 145703. Zulauf, M.; et al. J. Phys. Chem. 1985, 89, 3411. Nallet, F.; et al. J. Phys. II 1993, 3, 487.
JP810911Y
