J. Phys. Chem. B 2010, 114, 10357–10367
10357
Noncovalent Polymerization and Assembly in Water Promoted by Thermodynamic Incompatibility Karen A. Simon, Preeti Sejwal, Eric R. Falcone, Erik A. Burton, Sijie Yang, Deepali Prashar, Debjyoti Bandyopadhyay, Sri Kamesh Narasimhan, Nisha Varghese, Nemal S. Gobalasingham,† Jason B. Reese,† and Yan-Yeung Luk* Department of Chemistry and Department of Biomedical and Chemical Engineering, Syracuse Biomaterials Institute, Syracuse UniVersity, Syracuse, New York 13244 ReceiVed: April 7, 2010; ReVised Manuscript ReceiVed: July 1, 2010
This work studies the phase separations between polymers and a small molecule in a common aqueous solution that do not have well-defined hydrophobic-hydrophilic separation. In addition to poly(acrylamide) (PAAm) and poly(vinyl alcohol) (PVA), poly(vinyl pyrrolidone) (PVP) also promotes liquid crystal (LC) droplet formation by disodium cromoglycate (5′DSCG) solvated in water. In the presence of these polymers, the concentration of 5′DSCG needed for forming LC droplets is substantially lower than that needed for forming an LC phase by 5′DSCG alone. To define the concentration ranges that 5′DSCG molecules form liquid crystals (either as droplets or as an isotropic-LC mixture), we constructed ternary phase diagrams for 5′DSCG, water, and a polymer - PVA, PVP, or PAAm. We discovered that PAAm with high molecular weight promotes LC droplet formation by 5′DSCG more effectively than PAAm with low molecular weight. At the same weight percentage, long-chain PAAm can cause 5′DSCG to form LC droplets in water, whereas short-chain PAAm does not. Poly(vinyl pyrrolidone) (PVP), which has functional groups that are more dissimilar to 5′DSCG than PVA and PAAm, promotes LC droplet formation by 5′DSCG more effectively than either of the other two polymers. Additionally, small angle neutron scattering data revealed that the assembly structure of 5′DSCG promoted by the presence of PVA is similar to the thread structure formed by 5′DSCG alone. Together, these results reveal how noncovalent polymerization can be promoted by mixing thermodynamically incompatible molecules and elucidate the basic knowledge of nonamphiphilic colloidal science. 1. Introduction Hydrophobic-hydrophilic separations govern many phenomena in conventional colloidal chemistry and novel fabrication methods,1-4 but entirely water-soluble nonamphiphilic molecules can also exhibit phase separations in water.5-8 The earliest known example of water-in-water emulsions consisted of two biomacromolecules, a protein and a polysaccharide, and was reported by Beijerinck in 1896. Mixing gelatin with either agar or dextrin generated microscopic droplets dispersed in a common aqueous solution.9,10 Water-in-water emulsions from two synthetic polymers were reported years later (1976) when ethylene oxide-2-vinylpyridinium chloride block copolymer was added to stabilize an aqueous emulsion of poly(ethylene glycol) (PEG) and poly(2-vinylpyridinium chloride).11 Solution polymerization of sodium acrylate in the presence of PEG also produced an emulsion system of water-solvated poly(acrylic acid) droplets dispersed in a continuous aqueous solution containing PEG.12 For these systems, when mixed in the common solvent water, each polymer preferably solvates with its own type, and the solution separates to form two water-based phases in contact with each other.5 Clark and co-workers discovered recently that short cDNA duplex oligomers stack into columns and phaseseparate into liquid crystal drops, whereas unpaired, noncomplementary strands remain isotropic in solution.13 Whereas these water-in-water emulsions are generated by mixing structurally different polymers, we discovered in 2007 that some small nonamphiphilic molecules, such as disodium * Corresponding author. E-mail:
[email protected]. † Undergraduate participant.
cromoglycate (5′DSCG), can also exhibit phase separation with a nonionic polymer when mixed in water.14 This observation is rather surprising, given the high entropy associated with small molecules. The present work elucidates the effect of functional groups and molecular weights of polymers on the formation of water-in-water emulsions, and characterizes the mechanism for promoting the self-assembly of small nonamphiphilic molecules in water. The understanding of the molecular organization in the liquid crystal phase formed by disodium cromoglycate (5′DSCG) and similar fused aromatic molecules (Scheme 1) in water was developed with intrigue and controversy. Disodium cromoglycate (5′DSCG) was synthesized in the 1960s and developed into a potent antiallergic drug.15-17 The liquid crystal phase of 5′DSCG in water was first observed in 1971,17 and a simple model was proposed in 1973 by the discoverers.18 In 1980, a rather blunt attack on the proposed model was reported without any experimental results,19 and a new “chimney” model was proposed for the liquid crystal formed at high concentration of 5′DSCG in water. In the following year, the discoverers gave a rather gracious response.20 Unknown to many, the “chimney” model attacking the original model was retracted by its author.21 Nonetheless, this obsolete model is still sometimes cited in the literature today. In addition to 5′DSCG, the assembly behavior has been studied for other nonamphiphilic molecules, including Sunset Yellow, Blue7, and Violet20.22-27 The original “H” type of stack model for the nematic phase of these liquid crystals is a continuous subject for discussion.17,18,21,28 For instance, a “J”-
10.1021/jp103143x 2010 American Chemical Society Published on Web 07/27/2010
10358
J. Phys. Chem. B, Vol. 114, No. 32, 2010
Simon et al.
SCHEME 1
SCHEME 2
type stack model was proposed29-32 in which the molecules stack on an offset position by some fixed distance (Å) from each other instead of directly on top of each other (“H” type). In the light that the chemical shift of the NMR signals of the aromatic protons shift upfield rather than downfield when 5′DSCG assemble,33 most likely the aromatic rings are not directly stacked on each other (otherwise, the proton signal will shift downfield), but stack with an offset position such that the aromatic protons are on top of another aromatic ring. These chemical shifting phenomena are due to the shielding and deshielding effect of the ring current.32,34 In a recent study, we discovered that instead of the stacking model,19-21,35,36 molecules of 5′DSCG self-assemble in water to form threads that are connected by salt bridges (Scheme 2).37,38 When the concentration of 5′DSCG reaches 11-12 wt % at ambient temperature, these thread assemblies further form a highly birefringent liquid crystal phase. Below 11 wt %, the solution remains isotropic. Below, we summarize six inconsistencies to the general stacking model based on several experimental results and evidence that supported a thread assembly
model, in which molecules are connected linearly by salt bridges positioned in desolvating microenvironments by aromatic rings.37,104 First, the stacking model actually resembles a smectic liquid crystal phase or a discotic columnar phase rather than nematic liquid crystal phase, but the liquid crystal formed by 5′DSCG appears and is commonly agreed to be nematic.39 Second, the stacking model does not take into consideration the importance of the counter inorganic ions on the assembly, but only considers the π-π interactions to be the driving force for assembly. We discovered that by changing the sodium counter cations of 5′DSCG to lithium or potassium cations, the liquid crystal phase is almost completely eliminated. Third, whether a liquid crystal phase exists for a nonamphiphilic mesogen (fused aromatic molecules with structure similar to 5′DSCG) is highly sensitive to the molecular details. For example, 9 out of 11 DSCG derivatives with relatively small and systematic structural variation do not exhibit liquid crystal phase.37 This result strongly contradicts the stacking model, which suggests that stacking is the primary attribute for assembly, and the molecular details may not matter as the stacking model was applied to all completely different structures such as 5′DSCG and Sunset Yellow dye. How could small structural changes completely remove the liquid crystal phase instead of simply altering the liquid crystal properties, such as transition temperature (as is the case for thermotropic liquid crystals) or concentration need for forming liquid crystal phase (as is the case for conventional lyotropic liquid crystals)? Furthermore, a small structural variation of 5′DSCG leads to a wide range of different assemblies. For example, 7′DSCG (Scheme 1) in water forms only crystals, not liquid crystals;37 5′DSCG readily forms liquid crystals, and pure crystals are difficult to obtain. It is unlikely that stacking is the only defining attribute for the formation of both liquid crystals and crystals. Fourth, molecular stacking does not necessarily lead to liquid crystal formation. Many fused aromatic molecules exhibit stacking based on the broadening and shifting of the proton NMR signals,37,40-43 but do not form liquid crystals. Fifth, small angle neutron scattering of 5′DSCG in water matches best with a thin rod model with negligible cross section diameter. This dimension is not consistent with the molecular stacking model, which has a large and nonspherical cross section area, but is consistent with a thread model. Sixth, and most importantly, a single molecule, 5′DSCG-diviol (Scheme 1), in this series of cromoglycate structures can exhibit different assembly structures (liquid crystal and crystal) under slightly different conditions, a phenomenon of polymorphism well-known for pharmaceuticals.37,44-46 It is unlikely that
Noncovalent Polymerization and Assembly in Water different aggregates formed by the same molecule are of the same H-type of stacking. Recently, a series of hypothesis-based and discovery-driven experiments has led us to the thread model for this class of liquid crystals that is consistent and supports the observation described above.37 This thread model suggests that the divalently charged molecules are connected by salt bridges linearly, and the salt bridges are stabilized in water by a desolvating microenvironment created by the aromatic rings in the adjacent threads.37,104 Two hypothesis-based experiments support this model. First, the presence of monocharged molecules with aromatic rings almost completely eliminated the liquid crystal phase formed by 5′DSCG in water. Whereas the thread model requires two charges on the molecules to make the connections for elongating the threads, the stacking model requires only the presence of stacking moiety. The result suggests that the monocharged molecules break the threads, instead of inserting into the stack assembly and, thus, supports the thread model.37 Second, adding a DSCG derivative that has two charges, which does not form liquid crystals on its own, the liquid crystal phase of 5′DSCG in water is rendered.37 This result also supports a thread model. In addition, Sunset Yellow dye and 5′DSCG demix and phase-separate in water. Because stacking is based on π-π interactions, different aromatic moieties should have a high affinity for stacking because of better polarity match. As such, this observation also contradicts with the stacking model. Finally, the stacking model makes it difficult (if not impossible) to rationalize the molecular arrangement for a uniform alignment of this class of liquid crystal because the H-type of molecular assembly is similar to smectic or discotic columnar phase rather than a nematic phase. Recently, we demonstrate a planar uniform alignment of 5′DSCG liquid crystals on self-assembled monolayers on nanostructured gold films.38 The molecular organization for such a uniform planar alignment is fully explained by using the thread model. We also note that the H-type of stacking model provokes contradictions when applied to different molecules that form this type of nonamphiphilic lyotropic liquid crystals. For instance, a recent theoretical study suggests that Sunset Yellow dye forms an H-type assembly with head-to-tail (antiparallel) stacking arrangement.47 The structure of Sunset Yellow dye does not have a 2-fold symmetry, and thus this head-to-tail (antiparallel) stacking is possible for H-type arrangement. The disodium cromoglycate (5′DSCG) molecule consists of two identical chromonyl groups and, thus, has a 2-fold symmetry, and this head-to-tail stacking is not possible for 5′DSCG while forming H-type assembly. This situation becomes more puzzling because 5′DSCG forms the liquid crystal phase much easier than Sunset Yellow dye, with a lower weight percentage and a higher birefringence.25 Thus, it is questionable whether different molecules really form the same H-type assembly. The term “chromonic” liquid crystal was proposed in 1984 as “derived from the bis(chromone) structure in sodium cromoglycate (SCG)”.48 Chromone is a general chemical term describing the structural moiety of a specific type of fused aromatic rings and, thus, is rigorous and applicable for DSCG molecules and its derivatives, which contain “chromone” moieties. However, for other dye molecules (such as Sunset Yellow dye) that also exhibit a liquid crystal phase, the term “chromonic” is meaningless and irrelevant. For these reasons, we use “nonamphiphilic lyotropic liquid crystals”, a more generic term to distinguish this class of liquid crystal from the
J. Phys. Chem. B, Vol. 114, No. 32, 2010 10359 more commonly known lyotropic liquid crystals that are formed by part water-soluble and part water-insoluble, amphiphilic molecules. Interestingly, when nonionic polymers such as poly(vinyl alcohol) or poly(acrylamide) is added to an isotropic 5′DSCG solution, say 6 wt %, the 5′DSCG molecules are “squeezed” to form water-based liquid crystal droplets, in which the 5′DSCG mesogens align either parallel or perpendicular to the interface, and the polymer is solvated in the carrier phase, which is also water-based.14 In another recent study, we demonstrated uniform alignment of liquid crystals formed by hydrated molecules of 5′DSCG and Sunset Yellow over a large surface area by using self-assembled monolayers of functionalized alkanethiols supported by anisotropic gold films with nanometer-scale topography.38 This uniform alignment of liquid crystal can be interpreted by the thread assembly rather than by the stacking assembly. In addition, the azimuthal direction of the threads is either parallel or perpendicular to the direction of gold deposition, depending on whether the number of atoms in the alkanethiols of the SAMs is odd or even. This odd-even effect creates a twisted assembly of liquid crystals that can also be readily explained by the thread assembly, but not the stacking model. Lavrentovich and co-workers recently also discovered birefringent assembly involving 5′DSCG and positively charged small molecules, such as multivalent spermine, in water. Novel tactoid and toroid assemblies were observed, the molecular organization of these assembly appears to be consistent with columnar stacking of 5′DSCG.49 We believe that the electrostatic associations between the positive and negative charges of spermine and 5′DSCG likely change the thread assembly of 5′DSCG without additives (Scheme 2). On the basis of our survey of different water-soluble polymers in the previous work, it appears that the structural dissimilarity between the polymer and the 5′DSCG molecule is the primary attribute for causing the phase separation and forming of waterin-water emulsions (Scheme 2).14,37 The 5′DSCG molecule is rich in aromatic rings and charged groups of sodium carboxylates, whereas the polymers contain only the functional groups of hydrogen bond donors and acceptors. Polycations do not support water-in-water emulsions, but cause precipitation. Polyanions do not support a stable emulsion, but give a mixture of liquid crystal and isotropic phases. These discoveries suggest that a thermodynamic incompatibility can exist between polymers and small molecules, both entirely soluble in water. This proposition raises a series of interesting questions in the context of a nonamphiphilic colloidal science. For instance, are there other nonionic functional groups in a polymer, other than hydroxyl and amide groups, that can also support this waterin-water emulsion? How does the molecular weight of a polymer influence its ability to promote LC droplet formation by 5′DSCG molecules? Promoted by the presence of a proper nonionic polymer, what is the lowest concentration of 5′DSCG that can exhibit birefringence through water-in-water emulsion? Does the presence of the nonionic polymers promote the assembly of small molecules or already formed assembly, or both, to form liquid crystal droplets? What contributes to the stability of this emulsion system rather than letting the solution separate into two aqueous layers, one enriched with 5′DSCG and the other with polymer? To address the first four questions listed above in this work, we examined the ability of poly(vinyl pyrrolidone) (PVP), another nonionic water-soluble polymer, to support the formation of a water-in-water emulsion. We then constructed ternary phase diagrams for systems containing 5′DSCG, a polymer and
10360
J. Phys. Chem. B, Vol. 114, No. 32, 2010
Simon et al.
Figure 1. (A) Micrograph images of 6.43 wt % 5′DSCG and 10.9 wt % PVP (mw ∼ 40 000) in water viewed between crossed polarizers. Scale bar ) 76 µm (B) Enlarged images of the droplets at different orientations relative to the crossed polarizers. Scale bar ) 35 µm. Schematic representation of the droplet is shown on the right - the tubes represent the molecular threads formed by 5′DSCG in water (See Scheme 2).
water, and traced out the boundary between an isotropic solution, solutions containing birefringence (either in the form of droplets or entire phase), and solutions containing precipitates. Four phase diagrams were constructed for two different molecular weights of poly(acrylamide) (mw ∼ 1500 and ∼10 000), poly(vinyl alchohol) (PVA, mw ∼ 95 000) and poly(vinyl pyrrolidone) (PVP, mw ∼ 40 000). We also use small angle neutron scattering (SANS) to reveal the emergence of a 5′DSCG assembly and the formation of a water-in-water emulsion that is promoted by the presence of PVA. These studies elucidate how thermodynamic incompatibility can give rise to and control the nonrandom behavior of a mixed polymer and small molecules. 2. Materials and Methods Materials and Equipment. Disodium cromoglycate (5′DSCG) was purchased from MP Biomedicals (Solon, OH). Poly(acrylamide) (PAAm, mw ∼ 1500 and mw ∼ 10 000, both 50 wt % in water), poly(vinyl alcohol) (PVA, mw ) 89 000-98 000, 99+%), and poly(vinyl pyrrolidone) (PVP, mw ∼ 40 000) were purchased from Sigma-Aldrich (St. Louis, MO). Fisherfinest glass microscope slides were purchased from Fisher Scientific (Fairlawn, NJ). Deionized water with a resistivity of 18.2 MΩcm was used for the preparation of all of the lyotropic liquid crystal samples. Birefringence of samples was measured by plane polarized light in transmission mode on an Olympus BX51 polarizing microscope. Micrographs were taken using an Olympus C-5060 wide zoom digital camera. Sample Preparation. Targeted mass of 5′DSCG, polymer, and water were mixed in a vial, sealed with Parafilm and stored overnight. To eliminate the effect of aging, the samples were heated briefly with a heat gun, mixed using a vortex, and allowed to cool for an hour prior to examination of phase behavior. Samples with undissolved components, after heating followed by cooling in ambient temperature for 1 h, were classified as precipitate. Samples without precipitate were examined under a polarizing microscope for birefringence. Optical Cell Assembly and Measurement of Birefringence. Samples that appeared homogeneous were sandwiched between two glass microscope slides with a hole-punched Saran wrap
as a spacer.14 Following heating and settling for an hour, samples were mixed on a vortex and pipeted into the center of the hole of the Saran wrap that is on top of one glass slide. The other glass slide was used to cover the sample, and the optical cell was secured with four binder clips to prevent evaporation. The assembled optical cell was immediately examined under a polarizing microscope. Samples that appeared dark between crossed polarizers were classified as isotropic; those that transmitted light between crossed polarizers were classified as birefringent. 3. Results and Discussions When two different polymers are mixed together in a common solvent, the resulting mixture can become cloudy and eventually separate into two distinct layers.5 Such phase-separated systems have been widely observed in mixtures of two different biopolymers, such as proteins and polysaccharides.6,8,50-63 This phase separation is often observed for polymers;11,12 in contrast, mixing two independently soluble small molecules often results in a homogeneous solution.8 For both macromolecular and small molecular systems, the spontaneity of mixing (free energy of mixing) is governed by the entropy-enthalpy balance. For small molecules, the entropy of mixing dominates the enthalpic effects. For macromolecules, the molecules have less freedom to move, reducing significantly the entropy and making the interaction more enthalpy-driven.64-67 Recently, we discovered that certain nonionic polymers can phase-separate with the small molecule 5′DSCG in a common aqueous solution.14 Poly(vinyl pyrrolidone) Promotes Liquid Crystal Droplet Formation by 5′DSCG in Water. Compared to polycations that may form a charge complex with 5′DSCG or to polyanions that may mix well with 5′DSCG, PVP possesses functional groups that are dissimilar to 5′DSCG and are unlikely to form strong noncovalent interactions with functional groups on 5′DSCG. For this reason, PVP is also a polymer that may support stable water-in-water emulsions of 5′DSCG in water. Figure 1 shows optical micrographs of a sample mixture composed of 6.43 wt % 5′DSCG and 10.9 wt % PVP (mw ∼ 40 000) with ∼42-µm-thick spacers. The sample exhibited
Noncovalent Polymerization and Assembly in Water
J. Phys. Chem. B, Vol. 114, No. 32, 2010 10361
Figure 2. Ternary phase diagrams of water-in-water emulsion systems composed of water, 5′DSCG, and PAAm having average molecular weights of (A) ∼1500 and (B) ∼10 000. Solid circles indicate birefringence is observed. Open triangles indicate precipitate formation, and open squares indicate isotropic phase.
birefringent droplets when viewed under crossed polarizers. The birefringent droplets are spherical and exhibited a cross image in each droplet, which did not change as the sample was rotated with respect to the crossed polarizers. These results suggest that the presence of PVP promotes the formation of droplets with radial configuration, in which threads of 5′DSCG are perpendicular to the droplet surface, with a single point defect at the center of the droplet. We note that a crossed birefringent image in a droplet can also be consistent with a concentric droplet configuration, in which the liquid crystals align parallel to the droplet surface and perpendicular to a disinclination (line defect) through the center of the droplet. We eliminated the possibility of this concentric configuration for our droplets here by one argument and one experiment. First, for concentric configuration, statistically, there must be different droplet orientations. Some will align such that the disclination line is parallel to the light path, which gives crossed birefringent images. For those droplets not oriented in the light path, different birefringence will show. In our results, we only observe droplets with crossed images, indicating that the concentric configurations are highly unlikely. Second, we implemented air outlets between the glass slides of the optical cell that allowed the liquid crystal samples to flow and tumble. If the droplets are, indeed, of concentric configurations, the cross image will change as the droplet tumbles. The video of a flowing droplet did not show change in the birefringent image. We also note that similar experiments performed with bipolar droplets do show a change in the birefringence as the droplets flow and tumble. For the functional groups in all three polymers (PAAm, PVA, and PVP), PVP consists of only hydrogen bond acceptors whereas PAAm and PVA contain both hydrogen bond acceptors and donors. Comparing to 5′DSCG molecules, which contain both a hydrogen bond donor and multiple acceptors, we judge that PVP is more structurally distinct from 5′DSGC than PAAm or PVA. These results and observations are consistent with the notion that functionally distinct molecules can be thermodynamically incompatible even though each independently can be solvated in the same solvent. In this case, PVP excludes 5′DSCG and promotes 5′DSCG to form noncovalent polymers (see below),37 and to form liquid crystal droplets. Long Polymer Promotes More Effectively Liquid Crystal Droplet Formation of 5′DSCG. With these three polymers, we use ternary phase diagrams to characterize the concentration
range over which an LC phase is formed, as well as to study the effect of functional groups and molecular weights of the polymers on promoting liquid crystal formation by 5′DSCG. Figure 2 shows two ternary phase diagrams consisting of 5′DSCG, water, and poly(acrylamides) (PAAm) of different molecular weights. A larger birefringent region was observed for the ternary system containing PAAm of higher molecular weight (mw ∼ 10 000) (Figure 2A) than the system containing PAAm of lower molecular weight (mw ∼ 1500) (Figure 2B). The larger area of birefringence in the phase diagram suggests that the longer polymer promotes the liquid crystal formation of 5′DSCG more than the shorter polymer. This result further suggests that the noncovalent polymerization of 5′DSCG is promoted more effectively by the longer PAAm than the shorter one (see below). To understand and illustrate the effect of the molecular weight of PAAm, Figure 3 shows the overlay of the two ternary phase diagrams from Figure 2. The left axis containing only DSCG and water shows the phase behavior of the system in the absence of a polymer. The curves in the phase diagram merged at ∼10 wt % 5′DSCG and 90 wt % water, which was consistent with the literature reported lower limit, by which 5′DSCG forms liquid crystal phases in water (∼11 wt %).17 We identified three regions that exhibited different phase behaviors for the two polymers (labeled as regions I, II, III in Figure 3). The tie line (the gray line) in Figure 3 connecting the left and the bottom axes shows the phase behavior of the ternary system as the mass content (wt%) of the polymer was increased while the concentration of 5′DSCG was kept constant at 6 wt %. On this tie line, ∼10 wt % of PAAm with high molecular weight is needed for the sample to show birefringence, but ∼12 wt % of PAAm with low molecular weight is needed for observing the birefringence. Appearance of birefringence was observed as the mass content (wt %) of polymers was increased. These observations suggest that as the mass content of the polymer increased, 5′DSCG was excluded more effectively. When excluded from polymers, 5′DSCG molecules were “squeezed” together to induce formation of thread-like assemblies, which further formed liquid crystal phases. We note that inside the birefringent regions for both polymers, as the concentration of 5′DSCG was increased, the samples transitioned from LC droplets (water-in-water emulsion) to a mixture of LC phase and isotropic phase and eventually to an
10362
J. Phys. Chem. B, Vol. 114, No. 32, 2010
Simon et al.
Figure 3. Overlay of two phase diagrams (Figure 2) of regions showing birefringence for two mixtures. Dashed boundary encloses birefringent samples containing 5′DSCG and poly(acrylamide) (mw ∼ 1500) in water, and the solid boundary encloses birefringent samples containing 5′DSCG and poly(acrylamide) (mw ∼ 10 000). Representative samples in region I containing 7.21 wt % 5′DSCG, 9.16 wt % PAAm (mw ∼ 10 000); in region II containing 4.62 wt % 5′DSCG, 17.23 wt % PAAm (mw ∼ 10 000); and in region III containing 14.43 wt % 5′DSCG, 1.84 wt % PAAm (mw ∼ 10 000); (inset) containing 14.51 wt % 5′DSCG, 1.84 wt % PAAm (mw ∼ 1500), viewed between crossed polarizers are shown on the right. Scale bar ) 76 µm. The tie line represents samples with increasing PAAm mass content while the 5′DSCG concentration remains constant (∼6 wt %).
LC phase over the entire sample. Overlaying the two birefringent boundaries for the two phase diagrams reveals different phase behaviors in three regions. Below, we describe these three regions (I, II, and III), and relate the different phase characteristics for the high and low molecular weights of PAAm to the molecular behavior or self-assembly. In region I, the mass content (wt %) of poly(acrylamide) is low. With relatively low mass content of PAAm (∼7-10 wt %), long PAAm (mw ∼ 10 000) induced the liquid crystal droplet formation by 5′DSCG, but short PAAm (mw ∼ 1500) did not. Because monomers of acrylamide do not cause waterin-water emulsion, this result is consistent with the notion that the larger the polymer, the more incompatible it is with 5′DSCG molecules and, thus, the stronger the ability the polymer has to “squeeze” 5′DSCG into solvating with its own type and forming liquid crystal phases. In region II, the mass content of PAAm is high (above 13 wt % PAAm). The long (high molecular weight) PAAm was able to induce assembly and liquid crystal formation of 5′DSCG at a lower concentration than that by short (low molecular weight) PAAm (Figure 3 region II). Increasing the mass content of PAAm further beyond this region caused the formation of a precipitate, which contained both PAAm and 5′DSCG molecules as determined by proton NMR (see the Supporting Information). In the extra area of birefergence in the phase diagram created by long PAAm, the long PAAm is able to promote a stable water-in-water emulsion, whereas the short polymer already caused precipitation. In contrast, a low mass content of the short polymer in region I retained the isotropic solutions. The concentration range (polymer and 5′DSCG) for stable waterin-water emulsion (droplets) promoted by long PAAm in this region II (low concentration of 5′DSCG, high polymer content) is larger than that in region I (higher concentrations of 5′DSCG, low polymer content).
In region III, the concentrations of 5′DSCG are high; highly hydrated liquid crystal phases are formed by the high concentration of 5′DSCG without needing the polymer to promote liquid crystal droplet formation.17 In this region, being deficient of water due to high concentration of 5′DSCG, the solution is closer to saturation, and molecules (either 5′DSCG or the polymer) are more prone to precipitate. As a result, the presence of the polymer in this region disrupts the formation of liquid crystal of 5′DSCG. Conceivably, shorter polymers are better solvated in the 5′DSCG solution than the larger polymers. We observed that, in this region, the shorter polymers retained an LC phase over a broader range of concentrations than the longer polymers did. The precipitate caused by the long PAAm in this region is a mixture of 5′DSCG molecules and long PAAm with a mass ratio of about 4.86, determined by integrating the proton peaks in the NMR spectra of the precipitant sample. As long polymer chains can provide a longer range of disruption than short polymers, these results are consistent with the long polymer’s being more effective at attenuating the already formed liquid crystal phase of 5′DSCG than the short polymer. It is interesting to ask whether there exists a reversed emulsion in which the droplet contains predominantly the polymer and the carrier phase is 5′DSCG liquid crystal. This exploration is a topic of our ongoing research. We tabulate the observations for these three regions for the two PAAm with different molecular weights (Table 1). Overall, a longer polymer of PAAm is more capable of inducing phase separation and of causing liquid crystal formation of 5′DSCG than shorter polymer of PAAm. This effect is manifested in regions I and II in the ternary phase diagram. Thermodynamic Incompatibility Promotes Noncovalent Polymerization of 5′DSCG. Because birefringent droplets are observed at concentrations of 5′DSCG significantly lower than that needed for liquid crystal formation in the absence of these
Noncovalent Polymerization and Assembly in Water
J. Phys. Chem. B, Vol. 114, No. 32, 2010 10363
TABLE 1: Effect of Molecular Weights of Polymer on the Ternary Phase Behavior of 5′DSCG, Poly(acrylamide) (mw ∼ 10 000 and ∼1500) and Water a
composition observations
region
I
II
III
low wt % 5′DSCG, low wt % of PAAm long PAAm promotes LC droplets formation short PAAm renders the solution isotropic
low wt % of 5′DSCG high wt % of PAAm long PAAm promotes LC droplets formation short PAAm causes precipitationb
high wt % 5′DSCG, low wt % of PAAm long PAAm causes precipitationc short PAAm renders LC phase
See Figure 3. b Mass ratio of 5′DSCG to PAAm (mw ∼ 1500) is 3.81 (mole ratio is 11.11) for the sample consisting of 4.60 wt % 5′DSCG, 17.25 wt % PAAm, and 78.15 wt % H2O. c Mass ratio of 5′DSCG to PAAm (mw ∼ 10 000) in the precipitate is 4.86 (mole ratio is 94.30) for the sample consisting of 14.43 wt % 5′DSCG, 1.84 wt % PAAm, and 83.73 wt % H2O. a
Figure 4. (A) Schematic representation of the isodesmic assembly of 5′DSCG and the formation of tangential liquid crystal droplets (water-inwater emulsion) promoted by the presence of a long polyacrylamide. Noncovalent polymers inside the droplets align parallel to the surface of the elongated tangential droplets. (B) The same mass content of short polymers close to the boundaries of a birefringent region retains the isotropic solution (high mass content, Figure 3, region I) or causes precipitation (low mass content, Figure 3, region II).
polymers, the presence of these nonionic polymers clearly promotes liquid crystal formation by 5′DSCG in water. An important question emerges from this observation: Does the presence of these polymers also promote the assembly of 5′DSCG molecules in water before a liquid crystal phase is formed, or does the presence of these polymers promote only liquid crystal formation from the already formed assemblies of 5′DSCG? Unlike amphiphilic molecules that form assembly with specific shapes (such as spherical micelles), we believe that 5′DSCG molecules, being nonamphiphilic, form an assembly (or polymerize noncovalently) in water without a critical aggregation concentration. This assembly process, without a critical aggregation, is known to be “isodesmic” assembly,23,27,68-70 which exists for both biological71,72 and synthetic molecules.73 For 5′DSCG molecules, Clark and co-workers carried out theoretical studies that verify the assembly of 5′DSCG to be isodesmic.27 Here, we demonstrate that as long as the assembly is isodesmic in nature, the presence of structurally dissimilar polymers will also promote the noncovalent polymerization of 5′DSCG, in addition to promoting the liquid crystal formation. Figure 4A shows a schematic representation of the molecules before and after formation of liquid crystal droplets promoted by the presence of polymer. Because the assembly of 5′DSCG is considered isodesmic,23,27 the solution contains a mixture of monomers, dimers, trimers, oligomers, and polymers (all noncovalent) in equilibrium, but the liquid crystal droplets contain only noncoalvent polymers aligned in a preferred direction. Thus, the formation of the liquid crystal droplets must involve noncovalent oligomerization and polymerization, followed by liquid crystal formation. This result is important because it suggests that the thermodynamic incompatibility can
occur at the molecular level between the polymers and the 5′DSCG molecules, in addition to at the macromolecular level between the covalent polymer and the noncovalent polymer formed by 5′DSCG. The schemes in Figure 4 also show the effect of molecular weights of PAAm that caused different phase behaviors, which occur at areas close to the boundaries of the birefringent regions. With the same mass content (wt %) of PAAm, short polymers retain isotropic solution (region I, Figure 3) or cause precipitation (region II, Figure 3), but long polymers induce stable water-in-water emulsions (region I and II, Figure 3). Small Angle Neutron Scattering (SANS) Shows ThreadLike Assembly Induced by Nonionic Polymers. To study the effect of a nonionic polymer on promoting the assembly of 5′DSCG, SANS was used to characterize the shape of the molecular assembly as well as the liquid crystal formation at low concentrations of 5′DSCG in the presence of a nonionic polymer, poly (vinyl alcohol), PVA. Figure 5 shows the SANS of 5′DSCG in D2O in the presence and absence of PVA. The SANS data of 12 wt % 5′DSCG and 6 wt % 5′DSCG in D2O was reported earlier37 and is included in this discussion for comparison. The bend in q ∼ 0.11 Å with a slope of -1 is characteristic of the rod-shaped assembly 5′DSCG in D2O.37 For the LC sample of 5′DSCG in D2O (12 wt % 5′DSCG), a characteristic peak at q ∼ 0.11 Å is observed, suggesting an intense correlated sizes in the LC phase.37 We examined SANS for samples containing 5′DSCG below and above the nematic concentration (12 wt %) with and without PVA. Figure 5 shows the SANS data for sample containing 6 wt % of 5′DSCG and 8 wt % of PVA, and 3 wt % DSCG and 8 wt % of PVA. Data of SANS for 12 wt % and 6 wt % of
10364
J. Phys. Chem. B, Vol. 114, No. 32, 2010
Simon et al.
Figure 5. SANS of a solution containing 5′DSCG and samples containing 5′DSCG and PVA (mw ∼ 95 000) in D2O. Plot of intensity versus q (in log scale) at different concentrations of 5′DSCG and mass contents of PVA are shown in the graph. SANS of samples without PVA were adopted from our earlier work.37 Optical micrographs of samples between crossed polarizers are shown on the right.
5′DSCG without PVA was also included. At 6 wt % of 5′DSCG with no polymer, the SANS did not show the characteristic peak at q ∼ 0.11 Å, suggesting that the LC phase is not formed. When PVA is present (3 wt % of 5′DSCG and 8 wt % of PVA), the SANS show a different trace of ln I(q) versus ln q, which is characteristic of the presence of the polymer. This sample (3 wt % of 5′DSCG and 8 wt % of PVA) is isotropic. Interestingly, when the 5′DSCG concentration was increased from 3 to 6 wt % in the presence of 8 wt % PVA, the characteristic peak corresponded to 5′DSCG assembly, and the LC phase formation37 appeared in the background trace dominated by the character of the polymer. This peak coincided with the peak for 12 wt % 5′DSCG without PVA, which was entirely a liquid crystal phase. At the same time, the sample (6 wt % of 5′DSCG and 8 wt % of PVA) exhibited radial liquid crystal droplets (Figure 5, inset). These results suggest that at 6 wt % of 5′DSCG, the presence of 8 wt % of PVA was able to induce assembly that showed the characteristic SANS corresponding to the thread assembly.37 The common characteristic peak at q ∼0.11 Å suggests that the assembly structures for 12 wt % 5′DSCG only and that promoted by PVA (6 wt % 5′DSCG and 8 wt % PVA) are likely of the same shape. Functional Group Dissimilarity of Polymers Promotes Liquid Crystal Formation by 5′DSCG in Water. To study the effect of different polymers on promoting liquid crystal formation of 5′DSCG, we also examined the ternary phase diagrams for PVA and PVP and compared them to the ternary phase diagrams involving the two PAAm. Overall, PVP (mw ∼ 40 000) supported the largest area in the phase diagram that showed birefringence. In the birefringent regions where the concentration of 5′DSCG is low (corresponding to regions I and II in Figure 3), PVP (mw ∼ 40 000) promotes LC formation with a boundary in the phase diagram larger than that by PVA
(mw ∼ 95 000) in this region, similar to that by the long PAAm. We note that both the structures and the molecular weights of PVP and PVA are different, but because the molecular weight of PVP is lower than that of PVA, and yet still promotes the liquid crystal droplet formation more effectively than PVA, it suggests that the functional groups of PVP is more effective at promoting liquid crystal formation of 5′DSCG (see Figure 6). At the region where the concentration of 5′DSCG is high (corresponding to region III in Figure 3), the birefringent boundary for PVP is significantly larger than that for PVA. Because the molecular weight of PVP is smaller than PVA, we cannot determine whether the disruption of liquid crystal formation is due to structural (or functional group) difference or due to low molecular weight. However, comparing these regions of PVP and PAAm (corresponding to region III in Figure 3), the boundary of birefringence due to PVP is still much larger than that of PAAm (both high and low molecular weight). Because the molecular weight of PVP is larger than both of the PAAm, considering the effect of molecular weight alone, PVP should cause a smaller birefringent area in region III than PAAm. Therefore, this result suggests that the stronger tendency for liquid crystal to form caused by PVP compared to other polymers is likely due to the difference in structural features (or functional groups) between the polymers and 5′DSCG. Coincidentally, PVA contains hydroxyl (-OH group), and PVP does not, which suggests that PVA is more structurally similar to 5′DSCG than PVP. On the basis of these observations, we believe that the more structurally different the polymer is with 5′DSCG, the stronger the ability that the polymer will have to support stable water-in-water emulsions. This interpretation is consistent with the notion that two dissimilar molecules are
Noncovalent Polymerization and Assembly in Water
J. Phys. Chem. B, Vol. 114, No. 32, 2010 10365
Figure 6. Phase diagrams of ternary systems composed of water, 5′DSCG, and either (A) PVA (mw ∼ 95 000) or (B) PVP (mw ∼ 40 000). Filled circles indicate the appearance of birefringence. Open triangles indicate precipitate formation and open squares indicate isotropic phase.
thermodynamically incompatible and will phase-separate, even though both are entirely soluble in water. Many other studies of ternary systems composed of water; oil; and either oligo(ethylene glycol)-derivatized surfactants,74,75 short chain alcohols,76,77 or ionic surfactants, such as sodium 1,4-bis(2-ethylhexyl)sulfosuccinate,78 were supported by the fundamental principle of hydrophobic-hydrophilic separations. One unique aspect of this phase study is that no welldefined hydrophobic-hydrophilic separation is involved in our system. Conventional colloidal systems that involve hydrophobic-hydrophilic separation have provided many explorations and applications in research and industry, ranging from templated synthesis,79-86 light polarizing device,87 catalysis,88 drug delivery,89-92 and biosensing,93-95 to oxygen carrier systems for artificial blood, and liquid breathing.96-101 We believe that nonamphiphilic colloidal chemistry in entirely aqueous solutions will likely form the basis for a new discipline of study, and create opportunities for research and applications that have a more stringent requirement for biocompatibility, such as for protein crystallization and for biomaterials fabrications. 4. Conclusions We have shown that water-in-water emulsions due to thermodynamic incompatibility in a common solvent (water) are not limited to polymers, but are also applicable to small, nonamphiphilic molecules connected by specific interactions; in this case, salt bridges stacked on aromatic rings.104 Here, we report another polymer, PVP, that when mixed with 5′DSCG in water also gives rise to stable water-in-water emulsions. Because of this thermodynamic incompatibility, the nonionic polymers with a rather broad range of molecular weights (∼1500 to ∼95 000) are capable of promoting thread assembly (noncovalent polymerization) and liquid crystal formation by 5′DSCG in water. Ternary phase diagrams comparing three polymers (PVP, PAAm, and PVA) were used to study the effects of different functional groups and molecular weights on promoting liquid crystal formation of 5′DSCG. Our results showed that polymers with higher molecular weight are more capable of promoting liquid crystal formation of 5′DSCG than that of lower molecular weight: at low concentration of 5′DSCG, the presence of long PAAm induces liquid crystal phases, whereas short PAAm
does not. However, given two different polymers (such as PVP vs PAAm) in which one induces 5′DSCG liquid crystal phase over a wider range of polymer mass content (wt %) than the other one, the former appears to be structurally more distinct from 5′DSCG than the latter. This result supports the notion that being structurally different is a requirement and a property for two molecules to be thermodynamically incompatible. Although this notion may seem obvious, it is not true for most of the small molecules that are soluble in a common solvent. These phase separations in a common solvent, together with the ability of 5′DSCG molecules to preserve protein activity102,103 and to sequester as many as 260 molecules of water per 5′DSCG,105 provides an opportunity to explore using 5′DSCG or similar nonamphiphilic mesogens to exclude and crystallize proteins. Acknowledgment. We thank the Chemistry Department of Syracuse University, Syracuse Center of Excellence, for a CARTI award supported by the U.S. Environmental Protection Agency (Grants nos. X-83232501-0), NSF-CMMI (no. 0727491), and NSF-CAREER (no. 0845686) for financial support. K. A. Simon was supported by a fellowship from Syracuse Biomaterials Institute. We thank Ryan Gerecht for early participation in this project. We also thank Professor Jerry Goodisman for fruitful discussions. Supporting Information Available: Proton NMR spectra of the collected precipitates from aqueous samples found in regions II and III for calculations of relative amounts of 5′DSCG and PAAm. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Kerker, M., Ed.; Colloid and Interface Science, Academic Press: San Diego, 1976; Vol. 2: Aerosols, Emulsions, and Surfactants. (2) Brake, J. M.; Daschner, M. K.; Luk, Y.-Y.; Abbott, N. L. Science 2003, 302, 2094. (3) Poulin, P.; Stark, H.; Lubensky, T. C.; Weitz, D. A. Science 1997, 275, 1770. (4) Han, Y.; Cheng, K.; Simon, K. A.; Lan, Y.; Sejwal, P.; Luk, Y.Y. J. Am. Chem. Soc. 2006, 128, 13913. (5) Tolstoguzov, V. Crit. ReV. Biotechnol. 2002, 22, 89. (6) Tolstoguzov, V. J. Therm. Anal. Calorim. 2000, 61, 397. (7) Clark, A. H.; Gidley, M. J.; Richardson, R. K.; Ross-Murphy, S. B. Macromolecules 1989, 22, 346.
10366
J. Phys. Chem. B, Vol. 114, No. 32, 2010
(8) Kasapis, S.; Morris, E. R.; Norton, I. T.; Gidley, M. J. Carbohydr. Polym. 1993, 21, 249. (9) Beijerinck, M. W. Z. Chem. Ind. Kolloide 1911, 7, 16. (10) Beijerinck, M. W. Centr-Bl. f. Bakter. u. Parasitenk. 1896, 2, 698. (11) Ossenbach-Sauter, M.; Riess, G. C. R. Hebd. Seances Acad. Sci., Ser. C 1976, 283, 269. (12) Hosoda, Y.; Ueshima, T.; Ishihara, S.; Imamura, K. Contemp. Top. Polym. Sci. 1984, 4, 575. (13) Nakata, M.; Zanchetta, G.; Chapman, B. D.; Jones, C. D.; Cross, J. O.; Pindak, R.; Bellini, T.; Clark, N. A. Science 2007, 318, 1276. (14) Simon, K. A.; Sejwal, P.; Gerecht, R. B.; Luk, Y.-Y. Langmuir 2007, 23, 1453. (15) Cox, J. S. G. Nature (London, U.K.) 1967, 216, 1328. (16) Cox, J. S. G.; Beach, J. E.; Blair, A. M. J. N.; Clarke, A. J.; King, J.; Lee, T. B.; Loveday, D. E. E.; Moss, G. F.; Orr, T. S. C.; et al. AdV. Drug Res. 1970, 5, 115. (17) Cox, J. S. G.; Woodard, G. D.; McCrone, W. C. J. Pharm. Sci. 1971, 60, 1458. (18) Hartshorne, N. H.; Woodard, G. D. Mol. Cryst. Liq. Cryst. 1973, 23, 343. (19) Lydon, J. E. Mol. Cryst. Liq. Cryst. 1980, 64, 19. (20) Hartshorne, N. H.; Woodard, G. D. Mol. Cryst. Liq. Cryst. 1981, 64, 153. (21) Lydon, J. Liq. Cryst. Today 2007, 16, 13. (22) Prasad, S. K.; Nair, G. G.; Hegde, G.; Jayalakshmi, V. J. Phys. Chem. B 2007, 111, 9741. (23) Horowitz, V. R.; Janowitz, L. A.; Modic, A. L.; Heiney, P. A.; Collings, P. J. Phys. ReV. E: Stat., Nonlinear, Soft Matter Phys. 2005, 72, 041710/1. (24) Park, H.-S.; Kang, S.-W.; Tortora, L.; Nastishin, Y.; Finotello, D.; Kumar, S.; Lavrentovich, O. D. J. Phys. Chem. B 2008, 112, 16307. (25) Nastishin, Y. A.; Liu, H.; Schneider, T.; Nazarenko, V.; Vasyuta, R.; Shiyanovskii, S. V.; Lavrentovich, O. D. Phys. ReV. E: Stat., Nonlinear, Soft Matter Phys. 2005, 72, 041711/1. (26) Edwards, D. J.; Jones, J. W.; Lozman, O.; Ormerod, A. P.; Sintyureva, M.; Tiddy, G. J. T. J. Phys. Chem. B 2008, 112, 14628. (27) Maiti, P. K.; Lansac, Y.; Glaser, M. A.; Clark, N. A. Liq. Cryst. 2002, 29, 619. (28) Lydon, J. Curr. Opin. Colloid Interface Sci. 2004, 8, 480. (29) Harrison, W. J.; Mateer, D. L.; Tiddy, G. J. T. J. Phys. Chem. 1996, 100, 2310. (30) Gonzalez-Rodriguez, D.; Janssen, P. G. A.; Martin-Rapun, R.; De Cat, I.; De Feyter, S.; Schenning, A. P. H. J.; Meijer, E. W. J. Am. Chem. Soc. 2010, 132, 4710. (31) Eckhardt, H.; Bose, A.; Krongauz, V. A. Polymer 1987, 28, 1959. (32) Merino, G.; Heine, T.; Seifert, G. Chem.sEur. J. 2004, 10, 4367. (33) Ding, X.; Stringfellow, T. C.; Robinson, J. R. J. Pharm. Sci. 2004, 93, 1351. (34) Gomes, J. A. N. F.; Mallion, R. B. Chem. ReV. 2001, 101, 1349. (35) Attwood, T. K.; Lydon, J. E. Mol. Cryst. Liq. Cryst., Lett. 1986, 4, 9. (36) Kostko, A. F.; Cipriano, B. H.; Pinchuk, O. A.; Ziserman, L.; Anisimov, M. A.; Danino, D.; Raghavan, S. R. J. Phys. Chem. B 2005, 109, 19126. (37) Wu, L.; Lal, J.; Simon, K. A.; Burton, E. A.; Luk, Y.-Y. J. Am. Chem. Soc. 2009, 131, 7430. (38) Simon, K. A.; Burton, E. A.; Cheng, F.; Varghese, N.; Falcone, E. R.; Wu, L.; Luk, Y.-Y. Chem. Mater. 2010, 22, 2434. (39) Lydon, J. Curr. Opin. Colloid Interface Sci. 1998, 3, 458. (40) Kraft, A.; Osterod, F.; Froehlich, R. J. Org. Chem. 1999, 64, 6425. (41) Berl, V.; Huc, I.; Khoury, R. G.; Krische, M. J.; Lehn, J.-M. Nature (London, U.K.) 2000, 407, 720. (42) Zhao, D.; Moore, J. S. J. Org. Chem. 2002, 67, 3548. (43) Nakade, H.; Jordan, B. J.; Xu, H.; Han, G.; Srivastava, S.; Arvizo, R. R.; Cooke, G.; Rotello, V. M. J. Am. Chem. Soc. 2006, 128, 14924. (44) Vishweshwar, P.; McMahon, J. A.; Oliveira, M.; Peterson, M. L.; Zaworotko, M. J. J. Am. Chem. Soc. 2005, 127, 16802. (45) Cabri, W.; Ghetti, P.; Pozzi, G.; Alpegiani, M. Org. Process Res. DeV. 2007, 11, 64. (46) Bond, A. D.; Boese, R.; Desiraju, G. R. Angew. Chem., Int. Ed. 2007, 46, 618. (47) Chami, F.; Wilson, M. R. J. Am. Chem. Soc. 2010, 132, 7794. (48) Attwood, T. K.; Lydon, J. E. Mol. Cryst. Liq. Cryst. 1984, 108, 349. (49) Tortora, L.; Park, H.-S.; Kang, S.-W.; Savaryn, V.; Hong, S.-H.; Kaznatcheev, K.; Finotello, D.; Sprunt, S.; Kumar, S.; Lavrentovich, O. D. Soft Matter 2010, in press. (50) Lundin, L.; Norton, I. T.; Foster, T. J.; Williams, M. A. K.; Hermansson, A. M.; Bergstrom, E. Spec. Publ., R. Soc. Chem. 2000, 251, 167. (51) Tolstoguzov, V. In Water Management in the Design and Distribution of Quality Foods, International Symposium on the Properties of Water in Foods, Helsinki, Finland, May 30-June 4, 1998 1999, p 199.
Simon et al. (52) Tolstoguzov, V. Int. ReV. Cytol. 2000, 192, 3. (53) Semenova, M. G.; Savilova, L. B. Food Hydrocolloids 1998, 12, 65. (54) Tolstoguzov, V. Food Hydrocolloids 1997, 11, 181. (55) Antonov, Y. A.; Lashko, N. P.; Glotova, Y. K.; Malovikova, A.; Markovich, O. Food Hydrocolloids 1996, 10, 1. (56) Semenova, M. G.; Pavlovskaya, G. E.; Tolstoguzov, V. B. Food Hydrocolloids 1991, 4, 469. (57) Tolstoguzov, V. B. Food Hydrocolloids 1988, 2, 195. (58) Antonov, Y. A.; Kiknadze, E. V. Nahrung 1987, 31, 57. (59) Zhuravskaya, N. A.; Kiknadze, E. V.; Antonov, Y. A.; Tolstoguzov, V. B. Nahrung 1986, 30, 601. (60) Zhuravskaya, N. A.; Kiknadze, E. V.; Antonov, Y. A.; Tolstoguzov, V. B. Nahrung 1986, 30, 591. (61) Grishchenkova, E. V.; Antonov, Y. A.; Braudo, E. E.; Tolstoguzov, V. B. Nahrung 1984, 28, 15. (62) Suchkov, V. V.; Grinberg, V. Y.; Tolstoguzov, V. B. Carbohydr. Polym. 1981, 1, 39. (63) Dickinson, E.; Semenova, M. G. J. Chem. Soc., Faraday Trans. 1992, 88, 849. (64) Flory, P. J.; Rehner, J., Jr. J. Chem. Phys. 1943, 11, 521. (65) Flory, P. J.; Rehner, J., Jr. J. Chem. Phys. 1943, 11, 512. (66) Bergfeldt, K.; Piculell, L.; Linse, P. J. Phys. Chem. 1996, 100, 3680. (67) Sieglaff, C. L. J. Polym. Sci. 1959, 41, 319. (68) Henderson, J. R. Phys. ReV. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top. 1997, 55, 5731. (69) Henderson, J. R. J. Chem. Phys. 2000, 113, 5965. (70) Smulders, M. M. J.; Nieuwenhuizen, M. M. L.; de Greef, T. F. A.; van der Schoot, P.; Schenning, A. P. H. J.; Meijer, E. W. Chem.sEur. J. 2010, 16, 362. (71) Nyrkova, I. A.; Semenov, A. N.; Aggeli, A.; Bell, M.; Boden, N.; McLeish, T. C. B. Eur. Phys. J. B 2000, 17, 499. (72) Kentsis, A.; Borden, K. L. B. Curr. Protein Pept. Sci. 2004, 5, 125. (73) Nieuwenhuizen, M. M. L.; de Greef, T. F. A.; van der Bruggen, R. L. J.; Paulusse, J. M. J.; Appel, W. P. J.; Smulders, M. M. J.; Sijbesma, R. P.; Meijer, E. W. Chem.sEur. J. 2010, 16, 1601. (74) Mathis, G.; Leempoel, P.; Ravey, J. C.; Selve, C.; Delpuech, J. J. J. Am. Chem. Soc. 1984, 106, 6162. (75) Blin, J. L.; Grignard, J.; Zimny, K.; Stebe, M. J. Colloids Surf., A 2007, 308, 71. (76) Waernheim, T.; Sjoeblom, E.; Henriksson, U.; Stilbs, P. J. Phys. Chem. 1984, 88, 5420. (77) Khmelnitsky, Y. L.; Van Hoek, A.; Veeger, C.; Visser, A. J. W. G. J. Phys. Chem. 1989, 93, 872. (78) Lynch, I.; Cornen, S.; Piculell, L. J. Phys. Chem. B 2004, 108, 5443. (79) Imhof, A.; Pine, D. J. Nature 1997, 389, 948. (80) Yang, P.; Deng, T.; Zhao, D.; Feng, P.; Pine, D.; Chmelka, B. F.; Whitesides, G. M.; Stucky, G. D. Science 1998, 282, 2244. (81) Velev, O. D.; Kaler, E. W. AdV. Mater. 2000, 12, 531. (82) Velev, O. D.; Tessier, P. M.; Lenhoff, A. M.; Kaler, E. W. Nature 1999, 401, 548. (83) Velev, O. D.; Lenhoff, A. M.; Kaler, E. W. Science 2000, 287, 2240. (84) Gin, D. L.; Pecinovsky, C. S.; Bara, J. E.; Kerr, R. L. Struct. Bonding (Berlin) 2008, 128, 181. (85) Tam-Chang, S.-W.; Mahinay, D.; Huang, L. AdV. Mater. Res. 2008, 47-50, 165. (86) Tam-Chang, S.-W.; Helbley, J.; Carson, T. D.; Seo, W.; Iverson, I. K. Chem. Commun. 2006, 503. (87) Gin, D. L.; Pecinovsky, C. S.; Bara, J. E.; Kerr, R. L. Struct. Bonding (Berlin) 2008, 128, 181. (88) Egger, H.; Sottmann, T.; Strey, R.; Valero, C.; Berkessel, A. Tenside, Surfactants, Deterg. 2002, 39, 17. (89) Patil, Y. B.; Toti, U. S.; Khdair, A.; Ma, L.; Panyam, J. Biomaterials 2009, 30, 859. (90) Astete, C. E.; Sabliov, C. M. Part. Sci. Technol. 2006, 24, 321. (91) Lee, P. J.; Langer, R.; Shastri, V. P. Pharm. Res. 2003, 20, 264. (92) Vallet-Regi, M.; Ramila, A.; del Real, R. P.; Perez-Pariente, J. Chem. Mater. 2001, 13, 308. (93) Xu, X.; Tian, B.; Kong, J.; Zhang, S.; Liu, B.; Zhao, D. AdV. Mater. 2003, 15, 1932. (94) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710. (95) Shiyanovskii, S. V.; Schneider, T.; Smalyukh, I. I.; Ishikawa, T.; Niehaus, G. D.; Doane, K. J.; Woolverton, C. J.; Lavrentovich, O. D. Phys. ReV., E: Stat. Nonlinear Soft Matter Phys. 2005, 71, 020702. (96) Fluorine and Health: Molecular Imaging, Biomedical Materials and Pharmaceuticals, 1st ed.; Tressaud, A., Haufe, A., Eds.; Elsevier Science: Amsterdam, The Netherlands, 2008, p 447. (97) Riess, J. G.; Krafft, M. P. Biomaterials 1998, 19, 1529.
Noncovalent Polymerization and Assembly in Water (98) Milius, A.; Greiner, J.; Riess, J. G. Colloids Surf. 1992, 63, 281. (99) Zarif, L.; Greiner, J.; Pace, S.; Riess, J. G. J. Med. Chem. 1990, 33, 1262. (100) Stebe, M. J.; Serratrice, G.; Delpuech, J. J. J. Phys. Chem. 1985, 89, 2837. (101) Chubb, C.; Draper, P. Proc. Soc. Exp. Biol. Med. 1987, 184, 489. (102) Luk, Y.-Y.; Jang, C.-H.; Cheng, L.-L.; Israel, B. A.; Abbott, N. L. Chem. Mater. 2005, 17, 4774.
J. Phys. Chem. B, Vol. 114, No. 32, 2010 10367 (103) Cheng, L.-L.; Luk, Y.-Y.; Murphy, C. J.; Israel, B. A.; Abbott, N. L. Biomaterials 2005, 26, 7173. (104) Thompson, S. E.; Smithrud, D. B. J. Am. Chem. Soc. 2002, 124, 442. (105) Cox, J. S. G.; Woodard, G. D.; McCrone, W. C. J. Pharm. Sci. 1971, 60, 1458.
JP103143X