Polystyrene-b-poly(acrylic acid) Vesicle Size Control Using Solution

Fusion and fission of molecular assemblies of amphiphilic polypeptides generating small vesicles from ..... Young Hie Kim , Warren T. Ford , Thomas H...
0 downloads 0 Views 150KB Size
Download PDF
3894

Langmuir 2004, 20, 3894-3900

Polystyrene-b-poly(acrylic acid) Vesicle Size Control Using Solution Properties and Hydrophilic Block Length Amira Choucair, Christine Lavigueur, and Adi Eisenberg* Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal, Quebec, H3A 2K6, Canada Received October 14, 2003. In Final Form: March 2, 2004 Polymeric vesicles have attracted considerable attention in recent years, since they are a model for biological membranes and have versatile structures with several practical applications. In this study, we prepare vesicles from polystyrene-b-poly(acrylic acid) block copolymer in dioxane/water and dioxane/ THF/water mixtures. We then examine the ability of additives (such as NaCl, HCl, or NaOH), solvent composition, and hydrophilic block length to control vesicle size. Using turbidity measurements and transmission electron microscopy (TEM) we show that larger vesicles can be prepared from a given copolymer by adding NaCl or HCl, while adding NaOH yields smaller vesicles. The solvent composition (ratio of dioxane to THF, as well as the water content) can also determine the vesicle size. From a given copolymer, smaller vesicles can be prepared by increasing the THF content in the THF/dioxane solvent mixture. In a given solvent mixture, vesicle size increases with water content, but such an increase is most pronounced when dioxane is used as the solvent. In THF-rich solutions, on the other hand, vesicle size changes only slightly with the water concentration. As to the effect of the acrylic acid block length, the results show that block copolymers with shorter hydrophilic blocks assemble into larger vesicles. The effect of additives and solvent composition on vesicle size is related to their influence on chain repulsion and aggregation number, whereas the effect of acrylic acid block length occurs because of the relationship among the block length, the width of the molecular weight distribution, and the stabilization of the vesicle curvature.

1. Introduction In recent years, polymeric vesicles have attracted the attention of several researchers, both in the academic and the industrial communities. Polymeric vesicles, which are spherical bilayers, have been prepared from a variety of materials, including amphiphilic coil-coil type diblock copolymers,1-10 peptide-based coil-rod diblocks,11-14 symmetric ABA type triblock copolymers,15-17 and, more recently, asymmetric ABC triblocks.18,19 The resemblance * To whom all correspondence should be addressed. (1) van Hest, J. C. M.; Delnoye, D. A. P.; Baars, M. W. P. L.; van Genderen, M. H. P.; Meijer, E. W. Science 1995, 268, 1592-1595. (2) Zhang, L.; Eisenberg, A. Science 1995, 268, 1728-1731. (3) Ding, J.; Liu, G. Macromolecules 1997, 30, 655-657. (4) Holder, S. J.; Sommerdijk, N. A. J. M.; Williams, S. J.; Nolte, R. J. M.; Hiorns, R. C.; Jones, R. G. Chem. Commun. 1998, 1445-1446. (5) Yu, K.; Bartels, C.; Eisenberg, A. Macromolecules 1998, 31, 93999402. (6) Discher, B. M.; Won, Y.-Y.; Ege, D. S.; Lee, J. C. M.; Bates, F. S.; Discher, D. E.; Hammer, D. A. Science 1999, 284, 1143-1146. (7) Maskos, M.; Harris, J. R. Macromol. Rapid Commun. 2001, 22, 271-273. (8) Luo, L.; Eisenberg, A. J. Am. Chem. Soc. 2001, 123, 1012-1013. (9) Harris, J. K.; Rose, G. D.; Bruening, M. L. Langmuir 2002, 18, 5337-5342. (10) Gravano, S. M.; Borden, M.; von Werne, T.; Doerffler, E. M.; Salazar, G.; Chen, A.; Kisak, E.; Zasadzinski, J. A.; Patten, T. E.; Longo, M. L. Langmuir 2002, 18, 1938-1941. (11) Cornelissen, J. J. L. M.; Fischer, M.; Sommerdijk, N. A. J. M.; Nolte, R. J. M. Science 1998, 280, 1427-1430. (12) Kukula, H.; Schlaad, H.; Antonietti, M.; Foerster, S. J. Am. Chem. Soc. 2002, 124, 1658-1663. (13) Vriezema, D. M.; Hoogboom, J.; Velonia, K.; Takazawa, K.; Christianen, P. C. M.; Maan, J. C.; Rowan, A. E.; Nolte, R. J. M. Angew. Chem., Int. Ed. 2003, 42, 772-776. (14) Checot, F.; Lecommandoux, S.; Klok, H. A.; Gnanou, Y. Eur. Phys. J. E 2003, 10, 25-35. (15) Schillen, K.; Bryskhe, K.; Mel’nikova, Y. S. Macromolecules 1999, 32, 6885-6888. (16) Nardin, C.; Hirt, T.; Leukel, J.; Meier, W. Langmuir 2000, 16, 1035-1041. (17) Chen, X. L.; Jenekhe, S. A. Macromolecules 2000, 33, 46104612. (18) Stoenescu, R.; Meier, W. Chem. Commun. 2002, 3016-3017.

of synthetic vesicles to biological membranes makes them an attractive model for cells and organelles,20,21 and their ability to offer both hydrophobic and hydrophilic domains suitable for incorporation opens the door for their applications as microreactors,13,20,22 environmental toxin sequesters, and drug delivery vehicles.23-26 In our group, vesicles have been prepared from a variety of amphiphilic block copolymers, such as polystyrene-bpoly(acrylic acid),2,27,28 polystyrene-b-poly(ethylene oxide),29,30 and polystyrene-b-poly(4-vinyl pyridine).31 Typically, such vesicles are prepared by first dissolving the copolymer in a solvent common for both blocks, such as dioxane, tetrahydrofuran (THF), or dimethylformamide (DMF). Water, which is a poor solvent for the hydrophobic block, is then added dropwise to induce self-assembly. (19) Liu, F.; Eisenberg, A. J. Am. Chem. Soc. 2003, 125, 1505915064. (20) Nardin, C.; Widmer, J.; Winterhalter, M.; Meier, W. Eur. Phys. J. E 2001, 4, 403-410. (21) Hammer, D. A.; Discher, D. E. Annu. Rev. Mater. Res. 2001, 31, 387-404. (22) Chiu, D. T.; Wilson, C. F.; Karlsson, A.; Danielsson, A.; Lundqvist, A.; Stromberg, A.; Ryttse, F.; Davidson, M.; Nordholm, S.; Orwar, O.; Zare, R. N. Chem. Phys. 1999, 247, 133-139. (23) Dufes, C.; Schatzlein, A. G.; Tetley, L.; Gray, A. I.; Watson, D. G.; Olivier, J.-C.; Couet, W.; Uchegbu, I. F. Pharm. Res. 2000, 17, 12501258. (24) Brown, M. D.; Schaetzlein, A.; Brownlie, A.; Jack, V.; Wang, W.; Tetley, L.; Gray, A. I.; Uchegbu, I. F. Bioconjugate Chem. 2000, 11, 880-891. (25) Lee, J. C. M.; Bermudez, H.; Discher, B. M.; Sheehan, M. A.; Won, Y.-Y.; Bates, F. S.; Discher, D. E. Biotechnol. Bioeng. 2001, 73, 135-145. (26) Photos, P. J.; Bacakova, L.; Discher, B.; Bates, F. S.; Discher, D. E. J. Controlled Release 2003, 90, 323-334. (27) Zhang, L.; Eisenberg, A. J. Am. Chem. Soc. 1996, 118, 31683181. (28) Shen, H.; Eisenberg, A. J. Phys. Chem. B 1999, 103, 94739487. (29) Yu, K.; Eisenberg, A. Macromolecules 1996, 29, 6359-6361. (30) Yu, K.; Eisenberg, A. Macromolecules 1998, 31, 3509-3518. (31) Luo, L.; Eisenberg, A. Angew. Chem., Int. Ed. 2002, 41, 10011004.

10.1021/la035924p CCC: $27.50 © 2004 American Chemical Society Published on Web 04/09/2004

Size Control of PS-b-PAA Vesicles

The formation of spherical micelles marks the onset of self-assembly, and as more water is added, a morphological transition from spheres to rods and then to vesicles commonly occurs. Several studies have focused on determining the factors that control the self-assembly of vesicles. It is well-known that the formation of any given morphology, including vesicles, is determined by a balance among three main forces: core-chain stretching, corona-chain repulsion, and interfacial tension.32,33 Factors that alter the above balance were, therefore, used to control the formation of vesicles. Such factors include the hydrophobic-to-hydrophilic relative block length of the copolymer,9,11,27,34,35 the polymer concentration,9,33,34 the polydispersity of the hydrophilic block,36 the solvent nature,37-39 and the water content,34,40 as well as the presence of additives such as surfactants,41 homopolymer,27 or ions.42 Other studies investigated the physicochemical properties of vesicles. Shen et al.28 and Luo et al.43 showed that vesicles prepared from polystyrene-b-poly(acrylic acid) in dioxane/water mixtures are equilibrium structures that increase in diameter as the water content in the solvent mixture increases. The change in size was reversible, since a decrease in vesicle diameter could be induced by reducing the water content. The mechanism of increase in size involves the fusion of vesicles, while the decrease in size occurs by vesicle fission.43 The kinetics of fusion of polystyrene-b-poly(acrylic acid) vesicles was also investigated.44 The results showed that, under the investigated conditions, the average relaxation times of vesicle fusion range between 10 and 700 s, depending on factors such as the solvent composition, the initial polymer concentration, and the acrylic acid block length.44 With a better understanding of the thermodynamic properties and the kinetic aspects of polystyrene-b-poly(acrylic acid) vesicles, determining the factors that control their size became of interest. Being able to prepare vesicles with controllable sizes is essential for constructing application-specific vesicles and for optimizing their performance in any given application. While several studies have focused on controlling the size of spherical micelles by using polymers of varying block lengths,40,45 by adding surfactants,46 or by utilizing external stimuli (such as pH,47-50 ionic strength,42,51-53 or the solvent nature54), (32) Zhang, L.; Yu, K.; Eisenberg, A. Science 1996, 272, 1777-1779. (33) Zhang, L.; Eisenberg, A. Macromolecules 1999, 32, 2239-2249. (34) Shen, H.; Eisenberg, A. Macromolecules 2000, 33, 2561-2572. (35) Won, Y.-Y.; Brannan, A. K.; Davis, H. T.; Bates, F. S. J. Phys. Chem. B 2002, 106, 3354-3364. (36) Terreau, O.; Luo, L.; Eisenberg, A. Langmuir 2003, 19, 56015607. (37) Yu, Y.; Eisenberg, A. J. Am. Chem. Soc. 1997, 119, 8383-8384. (38) Yu, Y.; Zhang, L.; Eisenberg, A. Macromolecules 1998, 31, 11441154. (39) Riegel, I. C.; Samios, D.; Petzhold, C. L.; Eisenberg, A. Polymer 2003, 44, 2117-2128. (40) Zhang, L.; Eisenberg, A. Polym. Adv. Technol. 1998, 9, 677699. (41) Burke, S. E.; Eisenberg, A. Langmuir 2001, 17, 8341-8347. (42) Zhang, L.; Eisenberg, A. Macromolecules 1996, 29, 8805-8815. (43) Luo, L.; Eisenberg, A. Langmuir 2001, 17, 6804-6811. (44) Choucair, A.; Kycia, A.; Eisenberg, A. Langmuir 2003, 19, 10011008. (45) Stepanek, M.; Podhajecka, K.; Tesarova, E.; Prochazka, K.; Tuzar, Z.; Brown, W. Langmuir 2001, 17, 4240-4244. (46) Zheng, Y.; Davis, H. T. Langmuir 2000, 16, 6453-6459. (47) Groenewegen, W.; Egelhaaf, S. U.; Lapp, A.; van der Maarel, J. R. C. Macromolecules 2000, 33, 3283-3293. (48) Gohy, J.-F.; Willet, N.; Varshney, S.; Zhang, J.-X.; Jerome, R. Angew. Chem., Int. Ed. 2001, 40, 3214-3216. (49) Ravi, P.; Wang, C.; Tam, K. C.; Gan, L. H. Macromolecules 2003, 36, 173-179. (50) Dai, S.; Ravi, P.; Tam, K. C.; Mao, B. W.; Gan, L. H. Langmuir 2003, 19, 5175-5177.

Langmuir, Vol. 20, No. 10, 2004 3895

similar investigations involving vesicles are only relatively new. In a study by Gnanou et al.14 vesicles were prepared from poly(butadiene)-b-poly(γ-L-glutamic acid) block copolymer, and the pH-induced transition from R-helix to random coil conformation of the peptide was used to change the radius of the formed vesicles from ca. 150 nm at pH ) 7 to ca. 100 nm at pH ) 3. The effect of added salt on vesicle size was also investigated. The addition of ca. 0.5 M NaCl reduces electrostatic repulsion among the poly(γ-L-glutamic acid) block by screening the charges and causes a decrease in the vesicle radius from 152 to ca. 130 nm.14 In another study, Wang et al.55 used the polymer molecular weight to control the vesicle size. Reducing the molecular weight of the vesicle-forming palmitoyl glycol chitosan block copolymer by approximately 10-fold caused a decrease in the average vesicle diameter from ca. 500 to 200 nm.55 In the present study, we exploit the polyelectrolyte nature of polystyrene-b-poly(acrylic acid) block copolymers to manipulate the electrostatic repulsion among the chains and, consequently, to control the aggregation number and the vesicle size. To vary the strength of electrostatic repulsion, vesicles were prepared in the presence of additives (such as NaCl, HCl, or NaOH) or in different solvent mixtures (dioxane and THF at different ratios). We also examine the effect of the hydrophilic block length on the vesicle size. Three block copolymers with the same polystyrene block length, but varying lengths of the poly(acrylic acid) block, were used for this purpose. 2. Experimental Section 2.1. Materials. Vesicles were prepared from three polystyreneb-poly(acrylic acid) block copolymers: PS310-b-PAA28, PS310-bPAA36, and PS310-b-PAA45, where the numbers refer to the number-average degree of polymerization. The block copolymers were synthesized by sequential anionic polymerization and have a polydispersity of 1.03, as determined by size exclusion chromatography using polystyrene standards. The details of synthesis and characterization procedures are given in previous publications.43,56,57 1,4-Dioxane and tetrahydrofuran (HPLC grade) were purchased from Fisher Scientific. NaOH and HCl were purchased as solutions (0.1004 and 0.0959 N, respectively) and NaCl as a solid (99.999% purity) from Aldrich. Water, deionized to a resistivity of 18 MΩ cm was used in all of the experiments. 2.2. Procedure. 2.2.1. Turbidity Diagrams. The polymer solutions used to construct the turbidity diagrams were prepared in the following way: the desired amount of polymer was dissolved in dioxane (or in a dioxane/THF mixtures when the effect of solvent composition was examined) and left to stir overnight. All polymer solutions had an initial concentration of 0.5% (w/w). Deionized water was added dropwise to each solution at a rate of ca. 0.2% (w/w) per minute. After the addition of approximately 1% (w/w) of water, the solution was left to stir for about 10 min. The turbidity was then followed for 2 min (or until the readings were constant within the noise level), and the signal was averaged and plotted as the turbidity of the solution at that water content. Turbidity diagrams constructed in the presence of additives (HCl, NaCl, and NaOH) were prepared by first adding the desired amount of the salt, acid, or base (as aqueous solution) to the polymer-solvent mixture and then adding deionized water dropwise as described above. The cycle of water addition and (51) Lee, A. S.; Buetuen, V.; Vamvakaki, M.; Armes, S. P.; Pople, J. A.; Gast, A. P. Macromolecules 2002, 35, 8540-8551. (52) Choucair, A.; Eisenberg, A. Eur. Phys. J. E 2003, 10, 37-44. (53) Foerster, S.; Hermsdorf, N.; Boettcher, C.; Lindner, P. Macromolecules 2002, 35, 4096-4105. (54) Lin, Y.; Alexandridis, P. Langmuir 2002, 18, 4220-4231. (55) Wang, W.; McConaghy, A. M.; Tetley, L.; Uchegbu, I. F. Langmuir 2001, 17, 631-636. (56) Zhong, X. F.; Varshney, S. K.; Eisenberg, A. Macromolecules 1992, 25, 7160-7167. (57) Hautekeer, J. P.; Varshney, S. K.; Fayt, R.; Jacobs, C.; Jerome, R.; Teyssie, P. Macromolecules 1990, 23, 3893-3898.

3896

Langmuir, Vol. 20, No. 10, 2004

Choucair et al.

turbidity measurement was continued until the change in turbidity with water content became very small (ca. 1% increase). The concentrations of additives reported in Figures 1-3 represent their concentrations in the initial polymer/organic solvent mixture, prior to the addition of deionized water. All turbidity measurements were performed using a Cary 50 UV-visible spectrophotometer. Absorbance readings were recorded every 0.02 min at λ) 650 nm. At this wavelength the polymer has minimal absorption, and any attenuation of light is due to scattering from the aggregates. 2.2.2. Transmission Electron Microscopy (TEM). To prepare samples for TEM, about 30 µL of a polymer solution was added to an excess of water (ca. 10-fold dilution) in order to kinetically freeze the aggregates. A drop of this solution was then placed on a copper grid, previously coated with a thin film of carbon, and left to air-dry. A JEOL JEM-2000 FX electron microscope, operating at an acceleration voltage of 80 keV, was used to examine the grids. TEM pictures were taken using a multiscan CCD camera, and the aggregate size was determined using Sigma Scan 4.0 image sizing program. 2.2.3. Electrophoretic Mobility. A microelectrophoresis apparatus (Rank Brothers Ltd, MK II) was used to measure the electrophoretic mobility of block copolymer aggregates present in solutions of different dioxane/THF ratios. All solutions had the same water content [ca. 20% (w/w)]. The velocity of the aggregates moving under the influence of the applied electric field was evaluated by measuring the time taken by the aggregates to cover a given distance (100 µm in this case). The electrophoretic mobility was then determined using the following equation:

EM )

velocity of aggregates applied electric field

3. Results and Discussion 3.1. Effect of HCl and NaCl on Vesicle Size. Measuring the turbidity as a function of water content is often used to follow the self-assembly and the morphological transitions of block copolymer aggregates in solution.28,58 The turbidity of the solution yields also a semiquantitative indication of the aggregate size, since larger structures scatter more light. In Figure 1 we show the turbidity diagram of a 0.5% (w/w) polystyrene310-bpoly (acrylic acid)36 solution prepared in the absence of additives, along with the diagrams obtained in the presence of different concentrations of HCl (Figure 1a) and NaCl (Figure 1b). First, we note that the onset of self-assembly, which is characterized by an increase in turbidity, shows no detectable shift in the presence of HCl or NaCl [the critical water content (cwc) = 10% (w/w)]. However, in the vesicle region of the diagram, which occurs at water concentrations between ca. 15% and 40% (w/w) (the presence of vesicles over that part of the turbidity diagram was confirmed using transmission electron microscopy), the turbidity of block copolymer solutions containing HCl or NaCl is larger than that of the additivefree solution (Figure 1a,b). The higher turbidity indicates that the addition of the acid or the salt results in the formation of larger vesicles. The effect of HCl and NaCl on vesicle size occurs through their influence on the strength of corona chain repulsion and, consequently, on the aggregation number. The addition of HCl results in the protonation of the carboxylic groups of poly(acrylic acid). Electrostatic repulsion among corona chains is consequently reduced. A similar effect is obtained by adding NaCl, which would reduce electrostatic repulsion by shielding the charges along the partially ionized poly(acrylic acid) chains. The lower electrostatic repulsion among the corona chains allows the aggregation (58) Chen, L.; Shen, H.; Eisenberg, A. J. Phys. Chem. B 1999, 103, 9488-9497.

Figure 1. (a) Turbidity diagrams of 0.5% (w/w) solutions of PS310-b-PAA36 in dioxane, prepared in the presence of different concentrations of HCl. (b) Turbidity diagrams of 0.5% (w/w) solutions of PS310-b-PAA36 in dioxane, prepared in the presence of different concentrations of NaCl.

number to increase, and results in an increase in the vesicle size. A similar effect of salt on the aggregation number of micelles prepared from polyelectrolyte block copolymer was reported by Armes et al.51 and Forster et al.53 The increase in turbidity caused by the addition of HCl was not a function of the amount of additive used, since the three different concentrations of the acid (64, 460, and 1270 µM) caused a similar increase in turbidity (Figures 1a). In the case of NaCl (Figure 1b), the increase was also independent of the salt concentrations used (2.7, 5.0, and 10.0 mM). These observations indicate that even the smallest concentration used was sufficient to protonate the ionic sites of poly(acrylic acid) in the case of HCl or to shield the charges in the case of NaCl, to a point where electrostatic repulsion became unimportant. Moreover, it was observed that the addition of either HCl or NaCl had a similar effect on vesicle size, as was confirmed from transmission electron microscopy data (Figure 2). It is important to note that the initial rise in turbidity of the salt-containing solutions between water concentrations of ca. 1% and 7% (w/w) (Figure 1b) is not a result of any anomalous micellization or aggregation but is simply due to the low solubility of NaCl in the dioxanerich solution. When the water content exceeds ca. 7% (w/w), the salt dissolves completely and the turbidity drops. The absence of any aggregates in this region was confirmed

Size Control of PS-b-PAA Vesicles

Figure 2. Effect of HCl, NaCl, and NaOH on the size of vesicles present in 0.5% (w/w) solutions of PS310-b-PAA36. All solutions were prepared in dioxane. The number of vesicles counted from TEM pictures range between 42 and 365. See Supporting Information for more details.

Figure 3. Turbidity diagrams of 0.5% (w/w) solutions of PS310b-PAA36 in dioxane, prepared in the presence of different concentrations of NaOH.

by examining samples of the solution using transmission electron microscopy. 3.2. Effect of NaOH on Vesicle Size. In Figure 3, we show the change in turbidity as a function of water content for 0.5% (w/w) solutions of polystyrene310-b-poly(acrylic acid)36 in dioxane, prepared in the presence of different concentrations of NaOH. The presence of ca. 16 µM NaOH causes a shift in the onset of self-assembly toward a higher water concentration. The critical water content, as determined from the initial rise in turbidity, increases from ca. 10% for the additive-free polymer solution to ca. 12% (w/w) in the presence of 16 µM NaOH. When higher concentrations of NaOH were used (33, 67, and 335 µM), the polymer sample precipitated out of solution at water concentrations of ca. 6% (w/w). Precipitation occurs, most probably, because the highly charged PS-polyelectrolyte diblock formed under these conditions is not soluble in the dioxane-rich solution. Moreover, it was observed that, in the presence of 16 µM NaOH, the turbidity in the vesicle region of the diagram (i.e. between 15% and 40% (w/w) water content) is lower than that of the additive-free polymer solution, indicating that, in the presence of NaOH, smaller vesicles are formed. The effect of additives on the cwc occurs through their influence on the degree of poly(acrylic acid) chain repulsion, which is one of the factors determining the onset of

Langmuir, Vol. 20, No. 10, 2004 3897

micellization. When water is added to a block copolymer dissolved in a good solvent, the interactions between the hydrophobic polystyrene block and the solvent mixture become less favorable, driving the chains to aggregate. In the presence of electrostatic repulsion among the hydrophilic chains, higher water concentrations become required to bring about the aggregation of the chains and the formation of stable aggregates. With the addition of NaOH, poly(acrylic acid) is ionized. As a result, the strength of electrostatic repulsion among its chains increases, causing the observed shift in the value of the critical water content. On the other hand, the addition of HCl or NaCl, which reduce the strength of corona chain repulsion, does not cause any detectable shift in the onset of self-assembly compared to the additive-free block copolymer solution. The absence of a shift might be due to the small degree of dissociation of poly(acrylic acid) at water concentrations around the cwc (ca. 10% w/w water and 90% dioxane) and, consequently, to the fact that electrostatic repulsion plays only a minor role in determining the onset of micellization. Under these conditions, reducing electrostatic repulsion among corona chains even further (by adding HCl or NaCl) would cause only a small perturbation to the system and, therefore, to the cwc. The increase in the electrostatic repulsion among the corona chains in the presence of NaOH is also responsible for the formation of smaller vesicles. With higher repulsion, the aggregation number decreases, causing the observed decrease in vesicle size. Figure 2 shows the average diameter of vesicles present in the additive-free block copolymer solution and those present in solutions containing HCl, NaCl, or NaOH. The results, obtained using transmission electron microscopy, show that, at a given water content, vesicles prepared in the presence of HCl or NaCl are larger than those present in the additivefree polymer solution. On the other hand, vesicles prepared in the presence of NaOH have smaller diameter. The data for Figure 2 (the average vesicle diameter, the number of counted vesicles, and the standard deviation, as determined using TEM) are given in Table 1 of the Supporting Information. For comparison, the size of vesicles prepared in the presence of 16 µM NaOH was also determined using dynamic light scattering (DLS). The results show that the vesicle size determined using DLS is comparable to, but slightly larger than, the one obtained using TEM. A summary and discussion of these results is included in the Supporting Information Figure 2 also shows that vesicles prepared in the presence of NaOH do not increase in size as the water content increases from 15% to 40% (w/w), contrary to the typical increase commonly observed in additive-free block copolymer solutions.28,44 Adding water to a solution of polystyrene-b-poly(acrylic acid) vesicles reduces the solvent quality for the polystyrene chains, which constitute the vesicle wall, and causes an increase in the interfacial tension. In response, and in order to reduce the total interfacial area, vesicle fusion occurs, causing a decrease in the total number of vesicles and an increase in vesicle size.43 At the same time, adding water, which has a higher dielectric constant than dioxane (water ) 78.5 and dioxane ) 2.2),59 is accompanied by an increase in the degree of ionization of poly(acrylic acid), and in the negative charge density of the vesicle corona chains. The increase in vesicle size with water content is, therefore, governed by a balance between two opposing factors: first, the increase in the interfacial tension, which favors the fusion of vesicles and, (59) Weast, R. C. Handbook of Chemistry and Physics, 60th ed.; CRC Press: Boca Raton, FL, 1979-1980.

3898

Langmuir, Vol. 20, No. 10, 2004

Figure 4. Turbidity diagrams for 0.5% (w/w) solutions of PS310b-PAA28 prepared in THF/dioxane mixtures of different ratios.

therefore, the increase in their size, and second, the increase in the electrostatic repulsion among the poly(acrylic acid) chains, which opposes an increase in the aggregation number and vesicle size. In the absence of NaOH, and as previous studies have shown, vesicles do increase in size upon water addition, indicating that the increase in interfacial tension with water content dominates the above force balance. However, in the presence of NaOH, in which case the poly(acrylic acid) chains are deprotonated and certainly more negatively charged, corona chain repulsion seems to counterbalance the interfacial tension effect, hindering, as a result, an increase in the aggregation number and in vesicle size. In addition to the above argument, it is also conceivable that the presence of NaOH prevents the increase in vesicle size with water content due to kinetic reasons. Adding NaOH, as discussed earlier, increases the negative charge density of the vesicle surface by deprotonating poly(acrylic acid). The resulting increase in intravesicle repulsion might reduce the collision efficiency between vesicles, slowing down the kinetics of vesicle fusion (which is the mechanism through which vesicles increase in size).44 To determine if such is the case, the turbidity of NaOHcontaining solutions was measured after approximately 1 year, and no significant increase in turbidity was observed, indicating that the vesicles did not change in size. Considering that the fusion of vesicles occurs with relaxation times between ca. 10 and 700 s,44 observing no change in turbidity after 1 year indicates that it is unlikely that the presence of NaOH prevents the increase in vesicle size with water content due to kinetic reasons. 3.3. Effect of Solvent Nature and Composition on Vesicle Size. The nature of the solvent used to dissolve the block copolymer, whether a single solvent or a solvent mixture, and the water content in solution are factors that control the formation, as well as the size of block copolymer vesicles. In Figure 4, we show turbidity diagrams for 0.5% (w/w) solutions of polystyrene310-b-poly(acrylic acid)28 prepared in mixtures of dioxane/THF at different ratios. The onset of micellization is clearly a function of the solvent composition. As the dioxane content in the solvent mixture increases from 25% to 50%, 75%, and 100% (w/w), the critical water content, cwc, determined from the initial increase in turbidity, shifts from 22% to 17%, 14%, and then to 8% (w/w), respectively. After self-assembly, the turbidity of all four solutions continues to increase with water content, with the incremental change in turbidity increasing with the dioxane content

Choucair et al.

Figure 5. Change in vesicle size as a function of water content and solvent composition (dioxane/THF mixtures). All vesicles were prepared from 0.5% (w/w) PS310-b-PAA28 block copolymer.

(Figure 4). Moreover, comparing the turbidity of the four solutions at any given water concentration above the cwc shows that, as the solvent mixture becomes richer in dioxane, the turbidity becomes higher, indicating, semiquantitatively, that larger aggregates are present in solutions with progressively higher dioxane content. Transmission electron microscopy, TEM, was used to determine the range of water concentrations over which vesicles exist in each of the above solutions and to estimate the average vesicle diameter. The results show that when the block copolymer is prepared in dioxane, vesicles exist in solutions with water concentrations between 14% and 21% (w/w). The average vesicle diameter increases from ca. 290 to 440 nm over that water range. At water contents higher than 21%, vesicles coexist with large compound micelles. Using a solvent composition of 75% dioxane/ 25% THF, vesicles exist in solutions at water concentrations between 15% and 40% (w/w), with the average diameter increasing from ca. 150 to 165 nm over that water content. In a 50:50 dioxane/THF mixture, vesicles coexist with spheres at water concentrations between 17% and 40% (w/w) and increase in diameter only slightly, from ca. 110 to 130 nm. When the block copolymer solution is prepared in a 25% dioxane/75% THF solvent mixture, vesicles coexist with spheres in solutions at water concentrations between 25% and 40% (w/w), and the average diameter of vesicles (ca. 70 nm) does not seem to increase with water content. Figure 5 summarizes the above results and shows a three-dimensional plot of vesicle size as a function of water content and solvent composition. Consistent with the conclusions drawn from turbidity measurements, the results given in Figure 5 show that larger vesicles can be prepared by increasing the dioxane content in the solvent mixture and that the increase in vesicle size with water content becomes progressively smaller as the THF content in the solvent mixture increases. In addition to its effect on vesicle size, the solvent composition also determines the morphology of the aggregates. Figure 6 shows representative TEM pictures of aggregates present in solutions of 0.5% (w/w) polystyrene310-b-poly(acrylic acid)28 as a function of the composition of the dioxane/THF solvent mixture. While vesicle are the only morphology present when the solvent consists of 100% or 75% (w/w) dioxane, mixtures of spheres and vesicles coexist as the dioxane content decreases and the solvent becomes richer in THF. In fact, when the copolymer

Size Control of PS-b-PAA Vesicles

Figure 6. Representative TEM pictures of aggregates present in 0.5% (w/w) solutions of PS310-b-PAA28. All solutions are at 40% (w/w) water content, except for the solution in 100% dioxane, which is at 21% w/w.

is dissolved in 100% THF, water addition does not result in the formation of vesicles, and spheres are the only morphology observed, even at water concentrations as high as 40% (w/w). As shown in Figure 6, increasing the THF content in the solvent mixture causes the formation of smaller vesicles and then a shift in morphology from vesicles, to mixtures of vesicles and spheres, and finally to spheres. The solvent nature and composition determine the type of the copolymer-solvent interactions and, consequently, affect the degree of swelling of the hydrophobic polystyrene block, as well as the degree of ionization of a polyelectrolyte such as poly(acrylic acid). Since the solubility parameters of THF (δ ) 18.6 MPa1/2) and dioxane (δ ) 20.5 MPa1/2) are both close to that of styrene (δ ) 19.0 MPa1/2),60 the degree of swelling of the polystyrene chains is expected to be comparable when either one of these solvents, or their mixtures, is used. However, the dielectric constant of THF (THF ) 7.5) is higher than that of dioxane ( dioxane ) 2.2).59 Therefore, when THF is used as the common solvent instead of dioxane, or when its content in the solvent mixture increases, the degree of ionization of poly(acrylic acid), at a given water content, should increase. Consequently, the strength of electrostatic repulsion among its chains is also expected to increase. To follow the change in the degree of ionization of PAA, the electrophoretic mobility of PS310-b-PAA28 aggregates was measured as a function of the dioxane/THF ratio in the solvent mixture. The results show that as the THF content increases from 0 to 50 and then to 100% (w/w), the measured electrophoretic mobility increases from (3.2 ( 0.5) × 10-9 to (5.3 ( 0.3) × 10-9 and then to (8.6 ( 1.6) × 10-9 m2 V-1s-1, respectively. Since the electrophoretic mobility is proportional to the surface charge density of the aggregates,61,62 the observed increase in the electrophoretic mobility reflects the enhanced degree of ionization of poly(acrylic acid) with the THF content. As explained earlier, with an increase in the strength of electrostatic repulsion among the hydrophilic chains, larger water concentrations become required to induce micellization. Since increasing the THF content in the solvent mixture causes an increases in the electrostatic (60) Brandrup, J.; Immergut, E. H. Polymer Handbook, 3 ed.; John Wiley and Sons: New York, 1989. (61) Shaw, D. J. Introduction to Colloid and Surface Chemistry, 4th ed.; Butterworth-Heinemann: Oxford, 1992. (62) Van de Ven, T. G. M. Colloidal Hydrodynamics; Academic Press: London, 1989.

Langmuir, Vol. 20, No. 10, 2004 3899

Figure 7. Effect of poly(acrylic acid) block length on vesicle size. All solutions were prepared in dioxane and had an initial polymer concentration of 0.5% (w/w). The number of vesicles counted from TEM pictures range between 42 and 190 vesicles.

repulsion among the poly(acrylic acid) chains, the value of the cwc shifts toward higher water concentrations, as the solutions become progressively richer in THF (Figure 4). A similar effect of the solvent nature on the value of the cwc was observed by Riegel at al. using di- and triblock copolymers of styrene and 5-(N,N-diethylamino) isoprene.39 The increase in electrostatic repulsion with THF content also favors the formation of smaller structures, with smaller aggregation numbers. The increase in electrostatic repulsion explains the formation of smaller vesicles and then the shift in morphology from vesicles, to mixtures of vesicles and spheres, and finally to spheres (Figure 6) that occurs as the THF content in the initial solvent mixture increases. As to the increase in vesicle size with water content, and as discussed previously, it is governed by a balance between two opposing forces: the increase in interfacial tension, which favors the increase in vesicle size, and the increase in poly(acrylic acid) chain repulsion, which opposes an increase in the aggregation number and, therefore, in vesicle size. When the block copolymer aggregates are prepared in THF-rich solutions, the higher electrostatic repulsion among poly(acrylic acid) chains seems to dominate the above force balance and oppose an increase in the aggregation number. As a result, the size of vesicles prepared in THF-containing solutions increases only slightly with water content, as shown in Figure 5 and as reflected in the turbidity diagrams given in Figure 4. 3.4. Effect of Poly(acrylic acid) Block Length on Vesicle Size. In the previous sections, we showed that by controlling solution parameters, such as the ion content or the solvent composition, a given block copolymer can be used to prepare vesicles of different diameters. In this section, we examine the effect of the hydrophilic block length on vesicle size. Three polymer samples with the same polystyrene block length, but different lengths of the poly(acrylic acid) block, were used: PS310-b-PAA28, PS310-b-PAA36, and PS310-b-PAA45. Vesicles were prepared from each of the above block copolymers, and the average vesicle diameter was determined using transmission electron microscopy. All solutions were prepared in dioxane as a common solvent, at an initial polymer concentration of 0.5% (w/w). Figure 7 shows the effect of poly(acrylic acid) block length on the size of vesicles prepared at different water concentrations (the error bars correspond

3900

Langmuir, Vol. 20, No. 10, 2004

to the population standard deviation). For all three block copolymers, vesicles increase in size with water content. However, at a given water concentration, block copolymers with shorter poly(acrylic acid) block length self-assemble into larger vesicles. The effect of poly(acrylic acid) block length on vesicle size can be understood when the relationship between the hydrophilic block length and the stabilization of the vesicle curvature is considered. Previously, Luo et al.8 showed that polystyrene-b-poly(acrylic acid) vesicles are equilibrium structures stabilized by the preferential segregation of short poly(acrylic acid) chains to the interior surface of the vesicle and of longer chains to the exterior surface. By increasing the block length of poly(acrylic acid), and for a given value of the polydispersity index (PDI), the molecular weight distribution becomes wider, meaning that the difference in length between the longest chains and the shortest chains increases. The preferential segregation of shorter poly(acrylic acid) chains (i.e the chains with smaller coil volume) to the interior surface of the vesicle is, therefore, improved, favoring the formation of smaller vesicles. Terreau et al.36 also reported a decrease in the size of PS-b-PAA vesicles upon broadening the molecular weight distribution of the poly(acrylic acid) block. Figure 1 of the Supporting Information shows a plot of the molecular weight distribution of poly(acrylic acid) chains as a function of the block length.36,63 In addition to improved chain segregation, an increase in the poly(acrylic acid) block length would enhance the steric repulsion among the chains. The higher repulsion favors a decrease in the radius of curvature and contributes to the formation of smaller vesicles from copolymers with longer PAA block length. 4. Summary and Conclusions We prepared vesicles from polystyrene-b-poly(acrylic acid) diblock copolymer and used solution properties (such as the ion content or the solvent composition) to control the vesicle size. We showed that a block copolymer with a given block length can be used to prepare vesicles with a variety of sizes by tuning certain solution properties. (63) Nguyen, D.; Zhong, X.-F.; Williams, C. E.; Eisenberg, A. Macromolecules 1994, 27, 5173-5181.

Choucair et al.

For instance, while polystyrene310-b-poly(acrylic acid)36 self-assembles in dioxane/water mixtures into ca. 280 nm vesicles at 40% (w/w) water concentration, the same copolymer at the same water content forms vesicles as large as ca. 400 nm when prepared in the presence of HCl or NaCl and as small as 120 nm in the presence on NaOH. As a general trend, it was observed that the addition of HCl or NaCl leads to the formation of larger vesicles, while NaOH can be used to obtained smaller vesicles. The solvent composition was another factor used to control vesicle size. By preparing polystyrene310-b-poly(acrylic acid)28 vesicles in dioxane/THF mixtures of different ratios, we learned that smaller vesicles can be obtained by increasing the THF content in the solvent mixture. In a given solvent mixture, vesicles increase in size with water content, but such an increase is most pronounced when dioxane is used as the solvent. On the other hand, vesicle size changes only slightly with water content in THF-rich solutions. Therefore, by selecting the appropriate solvent, one can prepare vesicles whose size is tunable by, or resistant to, changes in the water concentration. The ability of additives and solvent composition to control vesicle size occurs through their effect on the electrostatic repulsion among the poly(acrylic acid) chains and, therefore, on the aggregation number. In addition to solution properties, we have also determined the effect of the poly(acrylic acid) block length on vesicle size. An increase in the hydrophilic block length resulted in the formation of smaller vesicles. Since longer blocks have a wider molecular weight distribution, the segregation of shorter hydrophilic chains to the inside of the vesicle and longer ones to the outside improves, favoring the formation of smaller vesicles. Acknowledgment. We would like to thank Dr. T. G. M. van de Ven, for use of the microelectrophoresis apparatus, as well as the National Science and Engineering Research Council (NSERC), for financial support. Supporting Information Available: A plot of the molecular weight distribution of poly(acrylic acid) block as a function of block length, the detailed data of Figure 2, and the size of vesicles prepared in the presence of NaOH, as determined using dynamic light scattering (DLS). This material is available free of charge via the Internet at http://pubs.acs.org. LA035924P