Diblock Copolymer

Sep 29, 2009 - ... Science, University of Minnesota, Minneapolis, Minnesota 55455 ... Can Zhou , Gilman E. S. Toombes , Matthew J. Wasbrough , Marc A...
3 downloads 0 Views 3MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

Path-Dependent Morphologies in Oil/Water/Diblock Copolymer Mixtures Sangwoo Lee, Manickam Adhimoolam Arunagirinathan, and Frank S. Bates* Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, Minnesota 55455 Received July 21, 2009. Revised Manuscript Received September 11, 2009 Amphiphilic block copolymer surfactants have been studied primarily in binary or pseudobinary systems, where they form a variety of dispersions, including micelles and vesicles, characterized by extremely low critical micelle concentrations (CMCs). In this study, a poly(1,2-butadiene-b-ethylene oxide) (OB) diblock copolymer (Mn=33 kg/mol, and 72 wt % PEO) was combined with water (W) and 1,5-cyclooctadiene (C) in binary (OB/W and OB/C) and ternary (OB/W/C and OB/C/W) mixtures, where the order of components indicates the mixing sequence. The resulting morphologies were characterized by small-angle X-ray scattering (SAXS) and cryogenic scanning electron microscopy (cryo-SEM). Addition of 1,5-cyclooctadiene to premicellized OB in water (OB/W) leads to swelling of the initially spherical hydrophobic micelle cores followed by micelle fusion leading to three-dimensional network structures. Combining water with premixed OB in 1,5-cyclooctadiene (OB/C) transforms a periodic sheet-like lamellar microstructure into a complex state of phase separation built on water-swollen poly(ethylene oxide) bilayers, which are manifested as uniformly sized small vesicles. Thus, two entirely different microstructures, which are stable for at least a month, have been prepared from a single three-component formulation. This pronounced nonequilibrium phase behavior, attributed to the non-ergodic character of amphiphilic block copolymers, may offer new processing strategies for the preparation of desirable self-assembled structures.

1. Introduction Amphiphilic compounds, including surfactants, soaps, and lipids, are ubiquitous in nature and provide the enabling ingredients in many manufactured items of commerce.1,2 These materials have been studied extensively for more than a century, yet this rich and interdisciplinary subject continues to present unanticipated phenomena that bridge the physical and life sciences. For approximately a decade, we have been exploring the structure and properties of amphiphilic block copolymers when dispersed in water.3-5 With the name “macromolecular surfactants”, various similarities and differences have been documented between block copolymer-based and conventional, low-molecular weight, nonionic surfactants. Increasing the surfactant molecular weight by one or more orders of magnitude (from ∼300-800 to >2000 g/mol) results in a variety of thermodynamic and dynamic *To whom correspondence should be addressed. E-mail: [email protected]. edu. (1) Laughlin, R. G. The Aqueous Phase Behavior of Surfactants; Academic: San Diego, 1994. (2) Myers, D. Surfactant Science and Technology, 3rd ed.; Wiley: Hoboken, NJ, 2006. (3) Jain, S.; Bates, F. S. Science 2003, 300, 460. (4) Jain, S.; Bates, F. S. Macromolecules 2004, 37, 1511. (5) Jain, S.; Dyrdahl, M. H. E.; Gong, X.; Scriven, L. E.; Bates, F. S. Macromolecules 2008, 41, 3305. (6) Aniansson, E. A. G.; Wall, S. N.; Almgren, M.; Hoffmann, H.; Kielmann, I.; Ulbricht, W.; Zana, R.; Lang, J.; Tondre, C. J. Phys. Chem. 1976, 80, 905. (7) Kahlweit, M. J. Colloid Interface Sci. 1982, 90, 92. (8) Dormidontova, E. E. Macromolecules 1999, 32, 7630. (9) Lund, R.; Willner, L.; Richter, D.; Dormidontova, E. E. Macromolecules 2006, 39, 4566. (10) Won, Y.-Y.; Davis, H. T.; Bates, F. S. Macromolecules 2003, 36, 953. (11) Cerritelli, S.; Fontana, A.; Velluto, D.; Adrian, M.; Dubochet, J.; De Maria, P.; Hubbell, J. A. Macromolecules 2005, 38, 7845. (12) Jakobs, B.; Sottmann, T.; Strey, R.; Allgaier, J.; Richter, D. Langmuir 1999, 15, 6707. (13) Endo, H.; Allgaier, J.; Gompper, G.; Jakobs, B.; Monkenbusch, M.; Richter, D.; Sottmann, T.; Strey, R. Phys. Rev. Lett. 2000, 85, 102. (14) Gompper, G.; Richter, D.; Strey, R. J. Phys.: Condens. Matter 2001, 13, 9055.

Langmuir 2010, 26(3), 1707–1715

changes to the dispersions.6-16 Perhaps the most dramatic effects derive from a drastic reduction in the critical micelle concentration (CMC), which virtually eliminates the exchange of block copolymers between micelles, a condition termed a non-ergodic state.17-19 Increasing the molecular weight also facilitates the formation of microstructured branches at intermediate concentrations resulting in disordered network morphologies that display soft yet solid mechanical properties.20,21 Spherical and wormlike cylindrical micelles, and vesicles (known as polymersomes), can be produced with prescribed hydrophobic domain dimensions that can be exploited for drug delivery, medical imaging, and various other applications.22-24 Notwithstanding these distinguishing differences, amphiphilic block copolymers share certain structural features with nonionic surfactants, most notably monodisperse self-assembled nanoscale morphologies.25,26 Recently, we expanded our program to include three components: water, amphiphilic block copolymer, and oil. Despite a vast number of studies on binary and pseudobinary block copolymer solutions (references listed above), only a few studies dealing with (15) Frank, C.; Strey, R.; Schmidt, C.; Stubenrauch, C. J. Colloid Interface Sci. 2007, 312, 76. (16) Foster, T.; Sottmann, T.; Schweins, R.; Strey, R. J. Chem. Phys. 2008, 128, 064902. (17) Wilhelm, M.; Zhao, C. L.; Wang, Y.; Xu, R.; Winnik, M. A.; Mura, J. L.; Riess, G.; Croucher, M. D. Macromolecules 1991, 24, 1033. (18) Jada, A.; Siffert, B.; Riess, G. Colloids Surf., A 1993, 75, 203. (19) Alexandridis, P.; Lindman, B. Amphiphilic block copolymers: Self-assembly and applications; Elsevier: Amsterdam, 2000. (20) Jain, S.; Gong, X.; Scriven, L. E.; Bates, F. S. Phys. Rev. Lett. 2006, 96, 138304. (21) Jain, S. Ph.D. Thesis, University of Minnesota, Minneapolis, MN, 2005. (22) Blanazs, A.; Armes, S. P.; Ryan, A. J. Macromol. Rapid Commun. 2009, 30, 267. (23) Torchilin, V. P. Curr. Pharm. Biotechnol. 2000, 1, 183. (24) Glass, R.; M€oller, M.; Spatz, J. P. Nanotechnology 2003, 14, 1153. (25) Won, Y.-Y.; Davis, H. T.; Bates, F. S. Science 1999, 283, 960. (26) Th€unermann, A. F.; Kubowicz, S.; von Berlepsch, H.; M€ohwald, H. Langmuir 2006, 22, 2506.

Published on Web 09/29/2009

DOI: 10.1021/la902671r

1707

Article

Figure 1. Generalized nonionic surfactant ternary phase diagram based on ref 40.

the emulsification of oil and water by block copolymers have been reported.27-30 Our intention was to extend the analogies established in the two-component limit to the well-studied threecomponent mixtures involving nonionic surfactant, oil, and water.31-35 Many authors have shown a rich phase behavior for such systems, which depends critically on the precise architecture of the surfactant (and frequently cosurfactant) and may include a host of single-phase ordered morphologies, a disordered bicontinuous microemulsion phase, and the coexistence of two and three phases.36-39 A hypothetical yet representative isothermal three-component phase diagram, presented by Davis et al. and reproduced in Figure 1, illustrates this complex array of equilibrium structural features.40 The work presented here was motivated by theoretical calculations and experimental results that anticipate bicontinuous microemulsions in oil and water mixtures containing very small amounts of diblock copolymer. Jakob and co-workers have demonstrated that substitution of 10% PEP-PEO diblock copolymer for conventional surfactant in a three-component water/oil (C10H12)/surfactant (C10E4) mixture halves the amphiphile concentration required to achieve a bicontinuous microemulsion (see Figure 1).12,13 In the limit of only block copolymer surfactant, the (27) Riess, G.; Nervo, J.; Rogez, D. Polym. Eng. Sci. 1977, 17, 634. (28) Omarjee, P.; Hoerner, P.; Riess, G.; Cabuil, V.; Mondain-Monval, O. Eur. Phys. J. E 2001, 4, 45. (29) Hadjiyannakou, S. C.; Vamvakaki, M.; Patrickios, C. S. Polymer 2004, 45, 3681. (30) Kyriacou, M. S.; Hadjiyannakou, S. C.; Vamvakaki, M.; Patrickios, C. S. Macromolecules 2004, 37, 7181. (31) Kahlweit, M.; Strey, R. Angew. Chem., Int. Ed. 1985, 24, 654. (32) Kahlweit, M.; Strey, R.; Busse, G. Phys. Rev. E 1993, 47, 4197. (33) Alexandridis, P.; Olsson, U.; Lindman, B. Macromolecules 1995, 28, 7700. (34) Alexandridis, P.; Olsson, U.; Lindman, B. Langmuir 1998, 14, 2627. (35) Svensson, B.; Olsson, U.; Alexandridis, P. Langmuir 2000, 16, 6839. (36) Janert, P. K.; Schick, M. Macromolecules 1997, 30, 137. (37) Bates, F. S.; Maurer, W. W.; Lipic, P. M.; Hillmyer, M. A.; Almdal, K.; Mortensen, K.; Fredrickson, G. H.; Lodge, T. P. Phys. Rev. Lett. 1997, 79, 849. (38) Washburn, N. R.; Lodge, T. P.; Bates, F. S. J. Phys. Chem. B 2000, 104, 6987. (39) Lee, J. H.; Ruegg, M. L.; Balsara, N. P.; Zhu, Y.; Gido, S. P.; Krishnamoorti, R.; Kim, M.-H. Macromolecules 2003, 36, 6537. (40) Davis, H. T.; Bodet, J. F.; Scriven, L. E.; Miller, W. G. Microemulsions and Their Precursors. In Physics of Amphiphilic Layers; Langevin, D., Boccara, N., Eds.; Springer: Heidelberg, Germany, 1987; Vol. 21, pp 310-327.

1708 DOI: 10.1021/la902671r

Lee et al.

bicontinuous microemulsion window is expected to shift down to compositions containing 99%), a model oil, was obtained from Sigma. All solvents were used as received. Mixtures. Two different procedures were employed in preparing symmetric ternary mixtures, each containing 3 wt % OB29, 45 wt % COD, and 52 wt % water; this composition represents equal volume fractions of COD and water. OB-29 was added to water (5.5 wt %, termed OB-29/W) in a glass vial and vigorously mixed (vortexed) for several hours followed by incubation for 1 week at room temperature. The OB-29/W dispersions are somewhat opalescent (bluish) but relatively clear and homogeneous at room temperature. OB-29 was also added to COD (6.3 wt %, termed OB-29/C) in glass vials and mixed vigorously for 3 h. During the first few minutes of mixing, the specimen was heated with hot air (>45 °C) and cooled several times. Differential (41) Hillmyer, M. A.; Bates, F. S. Macromolecules 1996, 29, 6994.

Langmuir 2010, 26(3), 1707–1715

Lee et al.

Figure 2. Appearance of binary and ternary samples. (a) OB-29/ W and OB-29/C with opalescent (bluish) and slightly grayish appearances. (b) The OB-29/W/C sample displays a clear top, a translucent bluish bottom, and a milky and opaque middle layer. (c) The OB-29/C/W sample generates a small clear fluid layer at the bottom, and a large even milky layer through most of the specimen (left). Addition of hydrophilic 5 nm gold particles reveals two layers in the milky layer (right). scanning calorimetry measurements revealed that the PEO blocks melt in COD at approximately 42 °C (see below). Above this temperature, the diblock copolymer quickly disperses in the organic solvent and forms a clear and homogeneous mixture. Cooling this mixture to room temperature leads to a slightly gray and somewhat translucent state due to crystallization of the dispersed and microphase-separated PEO blocks. Reheating above 45 °C quickly clarifies the specimen. Over the period of several days to 1 week at room temperature, the block copolymer settles into a solvated mat toward the bottom of the vial due to the higher-density crystalline PEO domains; this inhomogeneity is quickly reversed by gentle stirring (Figure 2a). Ternary mixtures were prepared via addition of the specified amount of COD to the OB-29/W dispersion, and water to the OB29/C dispersion, yielding the ternary three-component mixtures denoted OB-29/W/C and OB-29/C/W, respectively; this notation identifies the order of mixing. The three-component mixtures were frozen with liquid nitrogen, evacuated, sealed in glass vials, heated to 45 °C, vortexed for 3 h, and then slowly cooled and stored quiescently at room temperature for at least 1 month. Size exclusion chromatography and 1H NMR experiments showed no evidence of polymer degradation following these procedures. Also, there is no evidence of PEO crystallinity in either the OB29/C/W or OB-29/W/C mixtures as determined by DSC42,43 (see below). After several hours, the OB-29/W/C specimens began to show evidence of macroscopic phase separation evidenced by the development of three fluid layers, a clear top layer, a translucent bluish bottom layer, and a milky and opaque middle layer. These regions continued to consolidate over the course of several days, reaching a final steady state between 1 day and ∼1 week. Figure 2b shows a picture of this three-layer state within a sealed vial after 1 month at room temperature. The OB-29/C/W specimens behaved very differently compared to the OB-29/W/C mixtures. With the exception of a small (ca. 2% by volume) portion of relatively clear fluid at the bottom of the vials, these specimens remained entirely milky and opaque even after standing quiescently for >3 months at room temperature as shown in Figure 2c (left side). To better assess the distribution of water within these opaque specimens, we used a 0.01 wt % aqueous dispersion of hydrophilic 5 nm colloidal gold nanoparticles (Sigma) in forming an OB-29/C/W mixture. This colloidal gold has a distinctive red color in aqueous dispersions, thereby (42) Bogdanov, B.; Mihailov, M. J. Polym. Sci., Part B: Polym. Phys. 1985, 23, 2149. (43) Bogdanov, B.; Mihailov, M.; Popov, A.; Uzov, C. Acta Polym. 1986, 37, 628.

Langmuir 2010, 26(3), 1707–1715

Article providing a method for tracing the movement of water in the mixtures. In this case, the milky specimen revealed two layers, a reddish bottom layer and a slightly pink, milky, top layer, thus establishing the location of the bulk water. A picture showing one of these specimens is included in Figure 2c (right side). DSC. Differential scanning calorimetry (DSC) experiments were conducted with the undiluted OB diblock copolymer along with the binary and ternary mixtures to assess the presence of poly(ethylene oxide) crystallinity. Samples (5-10 mg) were placed in a hermetic aluminum DSC pan, mounted in a Q1000 TA Instruments calorimeter, heated to 80 °C at a rate of 10 °C/min, cooled to -100 °C, and then heated again at a rate of 10 °C/min to 80 °C. The reported DSC traces were acquired during the final heating step. X-ray Scattering. The structures of bulk portions of the binary and ternary mixtures were characterized using synchrotron-based small-angle X-ray scattering (SAXS) performed at the DND-CAT (beamline 5I-D) at the Advanced Photon Source (APS) located at Argonne National Laboratory (Argonne, IL). Data were obtained during two sessions using a Mar CCD area detector and a sample to detector distance of 6.53 m and X-ray wavelengths (λ) of 0.729 and 0.886 A˚. Liquid specimens were held at room temperature in quartz capillary tubes during the SAXS measurements. Two-dimensional data were normalized by the data acquisition time and azimuthally averaged to the onedimensional form of intensity (arbitrary units) versus the magnitude of the scattering wavevector [q=4πλ-1 sin(θ/2)]. Wide-angle X-ray scattering (WAXS) measurements also were obtained at the APS with an area detector, a sample to detector distance of 0.134 m, and a wavelength (λ) of 0.729 A˚. WAXS results are presented in two-dimensional (area detector) form. Cryo-SEM. Microstructural morphology was characterized by cryogenic scanning electron microscopy (cryo-SEM) using a Hitachi S-900 field emission gun scanning electron microscope operated at an accelerating voltage of 3 keV. Portions of the binary and ternary mixtures, removed from specific locations in the storage tubes, were transferred with a glass capillary to brass freezer hats (200 μm deep), which were then sandwiched together and frozen with streams of liquid nitrogen delivered at 2100 bar over a period of 9 ms. This procedure vitrifies water (and other liquids) in the specimen without crystallization. The frozen sandwiches were fractured in liquid nitrogen, and the exposed surfaces were etched at a pressure of approximately 10-4 Pa to remove water and/or oil down to a prescribed depth, leaving behind only block copolymer.44 Etching rates are controlled by the specimen temperature, which must be carefully regulated. Sublimation of vitrified water is known to occur at approximately 0.2, 2, and 20 nm/s at -110, -100, and -90 °C, respectively.45 We determined that vitrified COD sublimes at roughly one-fifth the rate of glassy water over the same temperature range. Two- and three-component samples were etched at -102, -95, or -90 °C to enhance the SEM images obtained from the mixtures. Approximately 10 nm of platinum was coated on the surfaces of the etched specimens at -130 °C to minimize charging followed by cooling to -172 °C for imaging in the microscope.

3. SAXS Modeling We have modeled our SAXS results using IGOR Pro software marketed by WaveMetrics. A description of this product along with examples can be found in the literature.46 Briefly, after a specific particle form factor (e.g., hard sphere, cylindrical micelle, vesicle, etc.) has been selected, the model permits adjustment of a host of structural parameters, including particle size (e.g., sphere radius) and size distribution, interfacial profile, and corona size (44) Schatten, H.; Payley, J. Biological Low-Voltage Scanning Electron Microscopy; Springer: New York, 2008. (45) Ma, Y. Ph.D. Thesis, University of Minnesota, Minneapolis, MN, 2002. (46) Ilavsky, J.; Jemian, P. R. J. Appl. Crystallogr. 2009, 42, 347.

DOI: 10.1021/la902671r

1709

Article

Figure 3. DSC curves obtained from OB-29, OB-29/W, OB-29/C, OB-29/W/C, and OB-29/C/W (top to bottom). Melting of eutectic PEO crystals is denoted by a star.42,43 Calculated percent crystallinities for the PEO in OB-29 and OB-29/C are 74 and 33%, respectively. Addition of water eliminates PEO crystallinity in all specimens.

and spatial distribution function.47-49 The scattering amplitude factors, used to establish the scattering contrast for PB, PEO, COD, and water, were fixed on the basis of the number of electrons per unit volume for each component at room temperature; the scattering amplitude factor for PB is nearly identical to that of COD. A background scattering correction and an overall intensity scale factor are also added to the calculated q-dependent intensity. Actual scattering data can modeled in two ways: allowing the program to optimize a result with multiple adjustable parameters or building a solution through the sequential addition of variables beginning with the most sensitive ones while using a minimum number of adjustable parameters. We have characterized our data using the second approach as summarized in the following sections.

4. Results and Analysis The OB-29 diblock copolymer employed in this study is a very strong amphiphile in the presence of water and oil because of the high selectivity of poly(butadiene) (PB) and poly(ethylene oxide) (PEO) for COD and water, respectively. Bulk PB is an amorphous liquid, while PEO is a semicrystalline solid at room temperature. A DSC trace obtained from bulk (undiluted) OB-29 exhibits a PEO melting temperature of 65 °C and a crystallinity of approximately 74% (Figure 3), consistent with prior studies dealing with bulk PEO-based block copolymers.50,51 Adding COD to the OB diblock lowers the melting temperature to ∼40 °C and reduces the crystallinity to ∼33%, most likely reflecting the effects of swelling of the B blocks with the good COD solvent. The integrity of the PEO crystals present in the OB-29/C specimen was further established using WAXS measurements. Two sharp scattering (47) Burge, R. F.; Draper, J. C. Acta Crystallogr. 1967, 22, 6. (48) Won, Y.-Y.; Davis, H. T.; Bates, F. S.; Agamalian, M.; Wignall, G. D. J. Phys. Chem. B 2000, 104, 7134. (49) Bang, J.; Jain, S.; Li, Z.; Lodge, T. P.; Pedersen, J. S.; Kesselman, E.; Talmon, Y. Macromolecules 2006, 39, 1199. (50) Schmalz, H.; Knoll, A.; M€uller, A. J.; Abetz, V. Macromolecules 2002, 35, 10004. (51) Huang, P.; Zheng, J. X.; Leng, S.; Van Horn, R. M.; Jeong, K.-U.; Guo, Y.; Quirk, R. P.; Cheng, S. Z. D.; Lotz, B.; Thomas, E. L.; Hsiao, B. S. Macromolecules 2007, 40, 526.

1710 DOI: 10.1021/la902671r

Lee et al.

Figure 4. Two-dimensional WAXS powder pattern from OB29/C. Bragg reflections are assigned to the first two peaks on the basis of the indexing reported in the literature.52

rings located at q values consistent with the documented lattice spacings for PEO were recorded as shown in Figure 4.52 Adding water to OB-29 or the OB-29/C binary mixture eliminates all PEO crystallinity as evidenced by the DSC traces for OB-29/W, OB-29/C/W, and OB-29/W/C (Figure 3). In the following sections, we describe experimental results that illuminate the associated binary mixture morphologies and the morphological changes induced by mixing OB-29/W with COD and OB-29/C with water. 4.1. Binary Mixtures. The two-component OB-29/W and OB-29/C specimens prepared in this project form homogeneous, translucent, and opalescent mixtures as illustrated in Figure 2a. We have characterized the molecular-scale morphologies using a combination of SAXS and cryo-SEM. Representative SAXS data are presented in Figure 5 for the OB-29/W and OB-29/C binary mixtures at concentrations of 5.5 and 6.3 wt % block copolymer, respectively. Each trace displays very different features, suggestive of very different underlying microstructures. Extensive work by our group with numerous OB specimens leads us to anticipate disordered spherical micelles for the OB-29/W mixture. A numerical fit using a spherical B core with a radius (R) of 20.5 nm and a PEO corona quantitatively accounts for these SAXS data as illustrated by the solid curve in Figure 5. The model fitting procedure is most sensitive to the core radius, which sets the position of the rounded peaks beginning at q = 0.03 A˚ -1, followed by the core polydispersity (ΔR/R=0.08), which damps the amplitude of the peaks, and corona profile and extension as described previously.48 The key adjustable parameters are largely uncoupled, providing a high degree of confidence in the quoted values. This sphere assignment was confirmed with cryo-SEM imaging. Figure 6 shows a nearly monodisperse array of spherical objects with roughly the dimensions identified by SAXS analysis. Quantitative determination of the spherical micelle size by cryoSEM is somewhat complicated by the presence of a 10 nm platinum over layer and the consequences of etching away the continuous (i.e., water) matrix, leaving some ambiguity in discriminating the core versus corona. Nevertheless, the geometric

(52) Huang, P.; Zhu, L.; Cheng, S. Z. D.; Ge, Q.; Quirk, R. P.; Thomas, E. L.; Lotz, B.; Hsiao, B. S.; Liu, L.; Yeh, F. Macromolecules 2001, 34, 6649.

Langmuir 2010, 26(3), 1707–1715

Lee et al.

Figure 5. SAXS patterns obtained from OB-29/W (O) and OB29/C (0). On the basis of form factor analyses (;), OB-29/W is characterized by spherical micelles with a core radius of 20.5 nm. The calculated scattering response associated with OB-29/C indicates 6.6 nm thick PEO bilayers that are organized in a onedimensional (lamellar) array with a 45 nm periodicity (inset).

Figure 6. Representative cryo-SEM micrograph from OB-29/W. A spherical OB micelle geometry is clearly identified.

shape of the dispersion is definitive, and agreement between the two methods is excellent. The SAXS pattern obtained from the OB-29/C specimen was less easily characterized. Therefore, we first turn to the cryo-SEM results to assist in selecting a particle form factor. Figure 7 illustrates four images, two (Figure 7a,b) obtained from the 6.3 wt % mixture and two (Figure 7c,d) taken from a 5.0 wt % solution. All four pictures indicate an exfoliated leaflike morphology consistent with a dispersion of semicrystalline PEO lamellae stabilized by COD-swollen PB blocks. Careful examination of Figure 7d reveals a layered substructure with a spacing Langmuir 2010, 26(3), 1707–1715

Article

Figure 7. Representative cryo-SEM micrographs obtained from OB-29/C. Micrographs a and b were taken from a 6.3 wt % dispersion, and micrographs c and d were taken from 5 wt % dispersions All these images are consistent with a layered morphology. Panel d clearly shows a lamellar substructure to the large-scale sheets.

appropriate for the expected dimensions of swollen OB-29 lamellae. Caution should be exercised in interpreting these pictures. Swollen semicrystalline lamellae are likely to buckle and agglomerate when the solvent is sublimed, which is how we characterize these images. Nevertheless, this finding provides an invaluable clue regarding the modeling of the associated SAXS result. We have used a flat bilayer model (crystalline PEO stabilized by swollen PB) to account for the scattering data from OB-29/C for q g 0.05 A˚ -1 in Figure 5.47 The minimum at q = 0.10 A˚ -1 and the maximum at q = 0.13 A˚ -1 and strong upturn in intensity for q < 0.10 A˚-1 were the principal fitting factors, which yield a PEO layer thickness of 6.6 nm. At lower q values (90% solvent. 4.2. Ternary Mixtures. In contrast to the binary mixtures, the ternary systems are complex and challenging to characterize. We have examined dozens of compositions, prepared using two mixing protocols: adding water to OB-29/C and adding oil to OB29/W. Here we focus on two systems each containing 3 wt % OB29, 45 wt % COD, and 52 wt % water, corresponding to equal volume fractions of COD and water. The results obtained from these mixtures are representative of our overall findings and permit us to draw important conclusions regarding the behavior of block copolymers as nonionic surfactants. 4.2.1. OB/W/C. Adding COD to OB-29/W with vigorous mixing leads to qualitative changes in the dispersed structure. Figure 8 shows three SAXS patterns taken from small portions of the OB-29/W/C specimen located in the top, middle, and bottom layers of the phase-separated mixture (see Figure 2c). The top layer produced a featureless and relatively low-intensity scattering pattern, consistent with the absence of block copolymer in the oil DOI: 10.1021/la902671r

1711

Article

Figure 8. SAXS patterns obtained from the top (0), middle (O), and bottom (4) layers of OB-29/W/C (see Figure 2b). Solid lines show the form factor fit with a spherical micelle model with a core radius R of 31 nm. The featureless scattering pattern associated with the top layer indicates an absence of block copolymer (i.e., pure COD).

phase; this point was verified by size exclusion chromatography (SEC) measurements, which established that the concentration was below the detectable limit. SAXS data acquired from the middle and bottom layers are virtually indistinguishable in form, although the intensity recorded from the former is approximately one-third of that from the latter. This result suggests a two-phase state, e.g., pure COD (top layer) and an aqueous dispersion (bottom layer) with a mixed middle region. The SAXS data from the middle and bottom layers can be quantitatively modeled for q > 0.01 A˚ using the same approach described for OB-29/W as shown in Figure 8, where R=31 nm and ΔR/R=0.07. Therefore, assuming spherical symmetry, the radial dimension of the hydrophobic PB cores has increased by 51%; i.e., the cores have swollen 345% by volume by imbibing oil. Interparticle interference produces a pair of relatively broad peaks at the lowest q values, characteristic of disordered nanoparticle dispersions in the proximity of an ordering transition. Cryo-SEM analysis reveals a remarkably different morphology in the middle layer than anticipated by the SAXS results. Figure 9 displays four images taken from this region of the OB-29/W/C sample. A highly interconnected network morphology is plainly evident, one that resembles network structures identified earlier in higher-concentration solutions of OB diblock copolymer in water.5 Similar images were obtained throughout the material taken from the middle layer, with a surprising absence of swollen spherical micelles. We have shown previously that random networks can produce SAXS powder patterns that may closely resemble a cylindrical form factor. By analogy, the intriguing networks found in Figure 9 may be characterized by a form factor similar to a sphere. We hasten to point out that bicontinuous morphologies were recorded over OB-29 concentrations between 1 and 10 wt % (not shown here), and in each case, the SAXS patterns obtained from the middle and bottom layers were nearly identical in form. Formation of a three-dimensional network from swollen spherical micelles would be a logical structural transition. The effective fraction of PEO falls from 72 to ∼44% (based on the spherical assumption). This should drive the (53) Won, Y.-Y.; Brannan, A. K.; Davis, H. T.; Bates, F. S. J. Phys. Chem. B 2002, 106, 3354.

1712 DOI: 10.1021/la902671r

Lee et al.

Figure 9. Representative cryo-SEM micrographs obtained at two magnifications from OB-29/W/C. Contrast in these images was created by subliming water and COD from the specimen. These pictures reveal a three-dimensional network morphology.

Figure 10. SAXS patterns obtained from the top (0), middle (O), and bottom (4) layers of OB-29/C/W (see Figure 2c). The featureless trace from the bottom layer indicates the absence of block copolymer. The solid curves represent calculated form factor scattering from bilayer-based vesicles with a core (water-swollen PEO) layer thickness of 14.7 nm.

interfacial curvature toward a more neutral value, which would support a bicontinuous (network) phase.5,53 4.2.2. OB/C/W. Switching the order of addition of solvent to the OB polymer leads to completely different results. Macroscopically, the OB-29/C/W mixture contains two nearly equivalent large layers and a small bottom phase as illustrated in Figure 2. Use of gold nanoparticles makes evident the fact that the middle layer is rich in water. Figure 10 displays SAXS data acquired from each of these three regions. The lowest (water) layer produced no identifiable structure, and SEC analysis showed a nearly immeasurable amount of block copolymer (