Supramolecular Complex of [60]Fullerene-Grafted Polyelectrolyte and

Jul 20, 2007 - ... electron microscopy were employed to study the supramolecular complexation between fullerene end-capped PAAs and nonionic surfactan...
0 downloads 0 Views 864KB Size
Download PDF
8798

Langmuir 2007, 23, 8798-8805

Supramolecular Complex of [60]Fullerene-Grafted Polyelectrolyte and Surfactant: Mechanism and Nanostructures Chang Wang, Palaniswamy Ravi, and Kam Chiu Tam* Singapore-MIT Alliance, School of Mechanical and Aerospace Engineering, Nanyang Technological UniVersity, 50 Nanyang AVenue, Singapore 639798 ReceiVed March 1, 2007. In Final Form: May 23, 2007 Water-soluble, pH-responsive mono- and di-[60]fullerene end-capped poly(acrylic acid)s (PAA-C60 and C60PAA-C60) were synthesized using the atom transfer radical polymerization technique. Isothermal titration calorimetry, dynamic light scattering, UV-vis spectroscopy, and transmission electron microscopy were employed to study the supramolecular complexation between fullerene end-capped PAAs and nonionic surfactant, polyethylene glycol (910) tert-octylphenyl ether, also known as Triton X100 (TX100) at different pH values. At pH < 4, TX100 bound specifically to C60 domains driven by hydrophobic and π-π interactions between TX100 and fullerene molecules. The binding was exothermic, and the magnitude of the interaction decreased gradually with increasing pH. The amount of polymer-bound TX100 was proportional to the fullerene content, which was ∼1.3 and ∼2.5 mM for 5 mM (concentration of carboxylic groups) PAA-C60 and C60-PAA-C60, respectively. Morphological transformations resulting in the formation of polymer/surfactant complex (PSC) precipitates in the course of binding were observed for both polymers. The PSC of PAA-C60 possessed a dense spherical structure, whereas the PSC of C60-PAA-C60 possessed a lamellar stacking structure. The PSC precipitates resolubilized in excess amounts of TX100 to form stable aggregates.

Introduction Fullerene-containing amphiphilic polyelectrolytes have attracted significant fundamental and industrial interest in recent years because of their interesting and complex behaviors. These polymers are water-soluble, stimuli-responsive, and biocompatible, and they also retain their electrophilic and photosensitive properties of fullerene. Moreover, the amphiphilic nature of the polymers allows them to self-assemble into micelles with different structures of well-defined shape and size, where the micellization is normally reversible and the conformation is tunable by controlling the environmental stimuli. Thus, fullerene-containing polyelectrolytes find promising applications as nanoreactors, drug delivery vehicles, nanotemplates, and photon-harvesting devices.1-14 A number of studies have focused on the synthesis and selfassembly behaviors of fullerene-containing amphiphilic polymers in terms of polymer architecture, hydrophilic/lipophilic balance (HLB), pH, temperature, ionic strength, and solvent property.1,2,4-13 * To whom correspondence should be addressed. Current address: Department of Chemical Engineering, University of Waterloo. Phone: 519888-4567 x38339; Fax: 519-746-4979; E-mail: [email protected]. (1) Yang, J.; Li, L.; Wang, C. Macromolecules 2003, 36, 6060. (2) Song, T.; Dai, S.; Tam, K. C.; Lee, S. Y.; Goh, S. H. Polymer 2003, 44, 2529. (3) Song, T.; Dai, S.; Tam, K. C.; Lee, S. Y.; Goh, S. H. Langmuir 2003, 19, 4798. (4) Dai, S.; Ravi, P.; Tan, C. H.; Tam, K. C. Langmuir 2004, 20, 8569. (5) Tan, C. H.; Ravi, P.; Dai, S.; Tam, K. C.; Gan, L. H. Langmuir 2004, 20, 9882. (6) Tan, C. H.; Ravi, P.; Dai, S.; Tam, K. C. Langmuir 2004, 20, 9901. (7) Ravi, P.; Dai, S.; Tan, C. H.; Tam, K. C. Macromolecules 2005, 38, 933. (8) Teoh, S. K.; Ravi, P.; Dai, S.; Tam, K. C. J. Phys. Chem. B 2005, 109, 4431. (9) Ravi, P.; Wang, C.; Dai, S.; Tam, K. C. Langmuir 2006, 22, 7167. (10) Ederle´, Y.; Mathis, C. Macromolecules 1997, 30, 2546. (11) Ederle´, Y.; Mathis, C. Macromolecules 1999, 32, 554. (12) Nuffer, R.; Ederle´, Y.; Mathis, C. Synth. Met. 1999, 103, 2376. (13) Audouin, F.; Renouard, T.; Schmaltz, B.; Nuffer, R.; Mathis, C. Polymer 2005, 46, 8519. (14) Cardullo, F.; Diederich, F.; Echegoyen, L.; Habicher, T.; Jayaraman, N.; Leblanc, R. M.; Stoddart, J. F.; Wang, C. Langmuir 1998, 14, 1955.

Mathis and co-workers reported the synthesis of anionic star-, palm-tree- and dumbbell-shaped fullerene-containing polymer using anionic polymerization and atom transfer radical polymerization.10-13 The fabrication of an ultrathin film using fullerenecontaining polyelectrolytes via electrostatic self-assembly was reported by Cao and co-workers.15 A similar route to construct a thin film using fullerene-containing polyelectrolytes through layer-by-layer sequential adsorption was also demonstrated by Durstock et al.16 Yang and co-workers reported the synthesis and aggregation behavior of poly(acrylic acid)-C60 in aqueous solution, which formed a core-shell structure that controlled the photoconductive properties of the polymer. Over the last couple of years, we have focused on the synthesis and self-assembly behavior of a series of well-defined fullerene containing polyacids, including poly(methacrylic acid)-b-C60 (PMAA-b-C60) and mono- and di-fullerene-end-capped poly(acrylic acid)s (C60PAA and C60-PAA-C60).5,6,9 The polymers self-assembled into large compound micelles in an aqueous medium. Similar to crosslinked PAA microgel particles, the polymeric micelles swell or deswell in response to pH, where the swelling ratio was inversely proportional to the fullerene content.6,9 We also demonstrated the formation and manipulation of nano- to microscale fractal patterns using micelles of PMAA-b-C60 and PAA-b-C60 in aqueous solutions in the presence of a salt.5,6 However, the impact of surfactants on the micellar structure, morphology, and conformation of the polymers is less extensively studied.17-20 In fact, polymer/surfactant interaction itself is an important research field that is gaining increasing attention because of its complex behaviors and potential applications in detergency, pharmaceutical, cosmetic, and agricultural formula(15) Cao, T.; Wei, F.; Yang, Y.; Huang, L.; Zhao, X.; Cao, W. Langmuir 2002, 18, 5186. (16) Durstock, M. F.; Tayler, B.; Spry, R. J.; Chiang, L.; Reulbach, S.; Heitfeld, K.; Baur, J. W. Synth. Met. 2001, 116, 373. (17) Burker, S. E.; Eisenberg, A. Langmuir 2001, 17, 8341. (18) Guldi, D. M. J. Phys. Chem. A 1997, 101, 3895. (19) Williams, R. M.; Crielaard, W.; Hellingwerf, K. J.; Verhoeven, J. W. J. R. Neth. Chem. Soc. 1996, 115 (1), 72. (20) Wang, C.; Ravi, P.; Tam, K. C. Langmuir 2006, 22, 2927.

10.1021/la700600r CCC: $37.00 © 2007 American Chemical Society Published on Web 07/20/2007

Supramolecular Polyelectrolyte/Surfactant Complex

tions, food processing, and electronic miniaturization. Most of the studies focused on the interactions between oppositely charged polymers and surfactants, where the strong electrostatic interaction is clearly observable and the surfactants bind to polymer at concentrations several orders of magnitude below the critical micelle concentration (cmc).21-24 The electrostatic interaction is reinforced by the hydrophobic aggregation of alkyl chains of bound surfactants, resulting in the formation of a polymer/ surfactant complex (PSC).21-24 For uncharged or weakly charged polymer/surfactant systems, forces other than electrostatic interaction, such as hydrophobic interaction,21,25-28 ion-dipole interaction,25,26 metal coordination or conjugation,29-33 and hydrogen bonding,34,35 are responsible for the binding and complexation behavior. The surfactant-induced complexation is governed by the characteristics of the surfactant (HLB, cmc, type of hydrophilic head), as well as the polymer architecture, charge density, overall hydrophobicity, chain flexibility, etc.21-35 It significantly affects the solubility, optical, electrical, and anisotropic properties of the polymer and leads to the construction of nano- to microscale supramolecular assemblies of various structures, such as spherical,20 cylindrical,29,30 lamellar,29,31,32 hexagonal,29,32,33 3-dimensional network,20,35 or even more complicated structures. Moreover, the noncovalent nature of the polymer/surfactant supramolecular assembly makes it a versatile way to prepare 2-dimensional or 3-dimentional functional nanostructures that respond to external conditions.29-33 The current study is an extension of our previous work on the interaction between a fullerene containing amphiphilic polymer and an aromatic nonionic surfactant. This study focuses on the assembly of a nanostructure through supramolecular complexation between fullerene end-capped poly(acrylic acid)s and nonionic surfactant polyethylene glycol (9-10) tert-octylphenyl ether, as well as the effects of polymer architecture and charge density on the complexation mechanism. Polyethylene glycol (9-10) tert-octylphenyl ether, also known as Triton X100 (TX100), was chosen because it possessed strong affinity with fullerenes and carbon nanotubes. It was reported that fullerene molecules can be successfully dispersed in micellar solutions of TX100.18,36 A recent study showed that the functionalization of single-walled carbon nanotubes (SWNTs) using TX100 significantly reduced nonspecific adsorption of protein to the surface of nanotubes.37 Intrigued by the strong affinity between TX100 and fullerene/ SWNTs and motivated by potential applications of nanomaterials in drug delivery and miniature biological devices, we investigated (21) Goddard, E. D. In Interactions of Surfactants with Polymers and Proteins; Goddard, E. D., Ananthapadmanabhan, K. P., Eds.; CRC Press: Boca Raton, FL, 1993. (22) Hayakawa, K.; Kwak, J. C. J. Phys. Chem. 1982, 86, 3866. (23) Ehtezazi, T.; Govender, T.; Stolink, S. Pharm. Res. 2000, 17, 871. (24) Fundin, J.; Hansson, P.; Brown, W.; Lidegran, I. Macromolecules 1997, 30, 1118. (25) Dai, S.; Tam, K. C.; Jenkins, R. D. J. Phys. Chem. B 2001, 105, 10189. (26) Dai, S.; Tam, K. C.; Li, L. Macromolecules 2001, 34, 7049. (27) Bloor, D. M.; Wan-Yunus, W. M. Z.; Wan-Badhi, W. A.; Li, Y.; Holzwarth, J. F.; Wyn-Jones, E. Langmuir 1995, 11, 3395. (28) Wang, G.; Olofsson, G. J. Phys. Chem. 1995, 99, 5588. (29) Ikkala, O.; Brinke, G. Science 2002, 295, 29. (30) Wei, Z.; Laitinen, T.; Smarsly, B.; Ikkala, O.; Faul, C. F. J. Angew. Chem., Int. Ed. 2005, 44, 751. (31) Vilkman, M.; Kosonen, H.; Nyka¨nen, A.; Ruokolainen, J.; Torkkeli, M.; Serimaa, R.; Ikkala, O. Macromolecules 2005, 38, 7793. (32) Ikkala, O.; Brinke, G. Chem. Commun. 2004, 2133, 2131. (33) Valkama, S.; Lehtonen, O.; Lappalainen, K.; Kosonen, H.; Castro, P.; Repo, T.; Torkkeli, M.; Serimaa, R.; Brinke, G.; Leskela¨, M.; Ikkala, O. Macromol. Rapid Commun. 2003, 24, 556. (34) Yoshida, K.; Dubin, P. L. Colloids Surface, A 1999, 147, 161. (35) Wang, C.; Tam, K. C. J. Phys. Chem. B 2004, 108, 8976. (36) Bensasson, R. V.; Blenvenue, E.; Dellinger, M.; Leach, S.; Patrick, S. J. Phys. Chem. 1994, 98, 3492. (37) Shim, M.; Wong, N. S. K.; Chen, R. J.; Li, Y.; Dai, H. Nano Lett. 2002, 2 (4), 285.

Langmuir, Vol. 23, No. 17, 2007 8799

the binding of TX100 to mono- and di-[60]fullerene end-capped telechelic poly(acrylic acid)s and the unusual morphological transformations in the course of binding and compared them in this paper. The investigation was corroborated by isothermal titration calorimetric (ITC), UV-vis spectroscopic, dynamic light scattering (DLS), and transmission electron microscopic (TEM) measurements. This study provides new insights into the mechanism of morphological transformation in the course of complexation between a fullerene-containing polymer and TX100, and it demonstrates a simple and pH-reversible route to tune nanostructures of fullerene containing amphiphilic copolymers. Experimental Section Materials. Well-defined water-soluble, pH-responsive mono- and di-[60]fullerene end-capped telechelic poly(acrylic acid)s were synthesized via the atom transfer radical polymerization (ATRP) technique using group protecting chemistry.9,38 The detailed information on the synthetic procedures and polymer characterizations can be found in our previous papers.9,20 The nonionic surfactant Triton X100, which possesses a chemical structure of

was kindly provided by Dow Chemicals. Gel permeation chromatography performed on the surfactant demonstrated a fairly narrow molecular weight distribution with a PDI of less than 1.1. Standard solutions (1 M) of HCl and NaOH from Merck were used to adjust the pH and the ionic strength. Isothermal Titration Calorimetry. The microcalorimetric measurements were carried out using a Microcal isothermal titration calorimeter. This power compensation, differential instrument was described in detail by Wiseman et al.39,40 It has a reference and a sample cell of approximately 1.35 mL, and both cells are insulated by an adiabatic shield. The titration was carried out at 25.00 ( 0.02 °C, by injecting surfactant solution from a 250 µL syringe into the sample cell filled with polymer solution (the concentrations of both surfactant and polymer solutions are known). The syringe is tailored-made such that the tip acts as a blade-type stirrer to ensure an optimum mixing efficiency at 400 rpm. The heat evolved or absorbed by each injection in the course of titration was directly measured, producing the raw heat signal, also known as cell feedback (CFB). Integration of the CFB gave the differential enthalpy curve. UV-Vis Spectroscopy. An Agilent 8453 UV-vis spectrophotometer was used to measure the absorption of 5 mM (monomer concentration) polymer solutions with varying TX100 concentrations ranging from 0 to 0.8 mM at 25 °C. Dynamic Light Scattering. DLS studies were conducted using a Brookhaven BI-200SM goniometer and BI-9000AT digital correlator equipped with an argon ion laser (λ ) 488 nm) at 298 K. The time correlation function of the scattered intensity G2(t), which is defined as G2(t) ) I(t) I(t + ∆t), where I(t) is the intensity at time t and ∆t is the lag time, was analyzed using the inverse Laplace transformation technique (REPES for our case) to produce the distribution function of decay times. Thus, the apparent hydrodynamic radius can be determined from the decay rate via the Stokes-Einstein equation Rh ) kTq2/6πηΓ, where k is the Boltzmann constant, q is the scattering vector (q ) {4πn sin(θ/2)}/λ, where n is the refractive index of the solvent, θ is the scattering angle, and λ is the wavelength of the incident laser light in vacuum), η is the solvent viscosity, and Γ is the decay rate. Several measurements were performed for a sample to obtain an average hydrodynamic radius, and the variation in the Rh values was found to be small. (38) Zhou, P.; Chen, G. Q.; Hong, H.; Du, F. S.; Li, Z. C.; Li, F. M. Macromolecules 2000, 33, 1948. (39) Wiseman, T.; Williston, S.; Brandts, J. F.; Lin, L. Anal. Biochem. 1989, 179, 131. (40) Jelesarov, I.; Bosshard, H. R. J. Mol. Recognit. 1999, 12, 3.

8800 Langmuir, Vol. 23, No. 17, 2007

Figure 1. Differential enthalpy curves obtained from titrating 20 mM C12H25EO9 into (]) 5 mM PAA at R ) 0, (0) 5 mM PAA85b-C60 at R ) 0, and ([) deionized water.

Wang et al.

Figure 2. Differential enthalpy curves obtained from titrating 20 mM TX100 into (4) 5 mM PAA at R ) 0, (0) 5 mM PAA at R ) 1, (]) 5 mM PAA85-b-C60 at R ) 0, (O) 5 mM PAA85-b-C60 at R ) 1, and ([) deionized water.

Transmission Electron Microscopy. TEM studies were conducted using a JEOL JEM-2010 transmission electron microscope operating at 200 kV. A copper grid precoated with carbon was placed on a filter paper. A drop of polymer solution was placed onto the copper grid and allowed to dry overnight.

Results and Discussion Overview on the Interaction between Nonionic Surfactants and Fullerene End-Capped PAA. ITC and dynamic light scattering studies were carried out to study and compare the binding interactions of two nonionic surfactants, i.e., polyoxyethylene 9 lauryl ether (C12H25EO9) and aromatic nonionic surfactant TX100 with PAA and fullerene end-capped PAA at different degrees of ionization R (R ) 0 corresponds to the unionized state of the polyacid, and R ) 1 represents the fully ionized state of the polyacid). The differential enthalpy curves obtained from isothermal calorimetric titrations of 20 mM C12H25EO9 into solutions of 5 mM (refer to the concentration of acrylic acid) PAA85-b-C60 and PAA at R ) 0 were plotted in Figure 1, together with the dilution curve obtained from the titration of C12H25EO9 into water. The dilution curve exhibited an exothermic sigmoidal profile, characterizing the demicellization of C12H25EO9 in water, where the cmc determined from the differential enthalpy curve was 0.23 mM. The titrations of C12H25EO9 into PAA85-b-C60 and PAA at R ≈ 0 demonstrated identical enthalpy profiles: the enthalpy curve exhibited a sigmoidal transition over a broader range of surfactant concentration with a much gentler slope. This indicated that C12H25EO9 may bind to aggregates of unionized polyacids, and the grafting of fullerene has a negligible effect on the binding interaction because C12H25EO9 was not able to bind on the fullerene domains of PAA85-b-C60. Moreover, dynamic light scattering demonstrated that the hydrodynamic radius Rhapp of 5 mM PAA85-b-C60 at R ≈ 0 was independent of C12H25EO9 concentration, which confirmed that the addition of C12H25EO9 to PAA85-b-C60 did not produce a structural reorganization or conformational change in the polymeric particles. The interactions between aromatic TX100 and PAA and fullerene end-capped PAA at different R values were characterized using ITC. The logarithm of the differential enthalpy curves obtained from the titrations of 20 mM TX100 into 5 mM PAA85b-C60 and PAA at R ) 0 and R ) 1 were plotted together with the dilution curve in Figure 2. The cmc determined from the exothermic sigmoidal transition on the dilution curve was 0.3

Figure 3. Differential enthalpy curves obtained from titrating 20 mM TX100 into 5 mM PAA85-b-C60 at (]) R ) 0, ([) R ) 0.1, (4) R ) 0.25, (2) R ) 0.4, (O) R ) 0.5, (b) R ) 0.7, and (0) R ) 1.

mM, which agreed with the literature value.41 The enthalpy profile corresponding to the titration of TX100 to PAA at R ) 0 suggested that TX100 may bind to hydrophobic domains of randomly coiled aggregates of un-ionized PAA to form mixed micelles, which is similar to interaction between C12H25EO9 and un-ionized PAA and fullerene end-capped PAA. With increasing TX100 concentration, the binding curve gradually approached and merged with the dilution curve, indicating that the aggregates of PAA were saturated and stabilized by bound TX100 micelles. When R ) 1, the enthalpy profile for TX100 in the presence of PAA was identical to the dilution curve, suggesting that there was no interaction between fully dissociated PAA and TX100. This was expected because fully dissociated PAA is hydrophilic and was not able to induce binding with nonionic surfactants. It should be noted that the enthalpy curve obtained from the titration of TX100 into PAA85-b-C60 at R ≈ 0 displayed a discontinuous trend on the enthalpy profile represented by an exothermic hump between two inflection points from a TX100 concentration of ∼0.45 to a TX100 concentration of ∼0.8 mM as marked in the figure. This unique feature suggested that the (41) Dai, S.; Tam, K. C. Colloids Surf., A 2003, 229, 157.

Supramolecular Polyelectrolyte/Surfactant Complex

Langmuir, Vol. 23, No. 17, 2007 8801

Figure 5. Dependencies of Rhapp on the TX100 concentration for 5 mM PAA85-b-C60 (O) and C60-b-PAA83-b-C60 (b) at R ) 0.

Figure 4. UV-vis spectra obtained in 5 mM PAA85-b-C60 (a) and C60-b-PAA83-b-C60 (b) with addition of TX100 at different concentrations. The insets show the red shift of the fullerene absorption band with an increase of the TX100 concentration.

mechanism of binding interaction between TX100 and PAA85b-C60 was different from that of binding of C12H25EO9 to PAA85b-C60. On the basis of our previous microcalorimetric studies on the interactions between polymers and surfactants, a discontinuous hump on the enthalpy curve is normally attributed to the structural reorganization of polymeric particles induced by the binding of surfactants.35,42,43 We believed that the aromatic character of the hydrophobic moiety of TX100 and fullerene grafted on PAA enhanced the binding interaction, allowing the micelles of TX100 to partition to the fullerene domains of the aggregates of PAA85b-C60 rather than binding unselectively to the hydrophobic parts of the polymer aggregates. This specific binding induced polymer/ surfactant complexation and disrupted the micellization process of TX100, resulting in the exothermic hump observed on the enthalpy curve. On the other hand, when R ) 1, the enthalpy profile corresponding to the binding exhibited a sigmoidal curve that was nearly identical to the dilution curve, suggesting there was negligible interaction between TX100 and fully ionized PAA85-b-C60. Interaction between TX100 and Fullerene End-Capped PAA. Figure 3 shows the enthalpy curves obtained from calorimetric titrations of 20 mM TX100 into 5 mM PAA85-b-C60 at different R values. When R ) 0, the enthalpic binding curve (42) Wang, C.; Tam, K. C. Langmuir 2002, 18, 6484. (43) Wang, C.; Tam, K. C. J. Phys. Chem. B 2004, 108, 8976.

Figure 6. Dependencies of (O) Rhapp and ([) differential enthalpy on the TX100 concentration for 5 mM PAA85-b-C60 at R ) 0.

deviated from the dilution curve at the first injection of TX100, suggesting that the binding took place at an extremely low surfactant concentration; thereafter the binding curve gradually approached the dilution curve with increasing TX100 concentration. The binding was exothermic, and the negative enthalpy was attributed to the energy evolved from the hydrophobic binding of TX100 to the interchain hydrophobic domains within the polymer aggregates. When the TX100 concentration reached 0.58 mM, the enthalpy exhibited a negative slope, resulting in the exothermic hump observed on the enthalpy curve. The hump reached its exothermic maximum of -6.58 kJ/mol at a TX100 concentration of ∼1.4 mM; thereafter the enthalpy curve merged with the dilution curve at TX100 concentration of ∼3.9 mM, signaling the end of the binding interaction. As discussed earlier, the exothermic hump was believed to correspond to the structural reorganization of the polymer aggregates induced by the binding of TX100. It was interesting to note that the exothermic hump was only observed when both the polymer (PAA85-b-C60) and surfactant (TX100) possessed phenyl rings, implying that the structural reorganization was present only when TX100 bound specifically to fullerene groups on the polymer. With the increase in R, the difference between the binding curve and the dilution curve became less significant, and the amplitude of the exothermic hump decreased and completely disappeared when R exceeded 0.25, which suggested that the binding interaction was weakened and the induced structural reorganization ceased when the

8802 Langmuir, Vol. 23, No. 17, 2007

Wang et al.

Figure 7. Particle size distributions and micrographs demonstrating the morphology change of un-ionzied PAA85-b-C60 induced by the binding of TX100: (a) CTX100 ) 0 mM; (b) CTX100 ) 1 mM; (c) CTX100 ) 1.7 mM; (d) CTX100 ) 2 mM.

carboxylic acids on the PAA85-b-C60 were gradually neutralized. This was expected because the hydrophobicity of the polymer decreased with increasing R, which reduced the affinity between the polymer and surfactant and weakened the binding driven by hydrophobic interaction. To further confirm the partitioning of TX100 into the fullerene domains of the polymers, the binding of TX100 to 5 mM PAA85b-C60 and C60-b-PAA83-b-C60 at R ) 0 was characterized using UV-vis spectroscopy. The UV-vis spectra obtained in PAA85b-C60 and C60-b-PAA83-b-C60 solutions with different amounts of TX100 are shown in Figures 4a and 4b, respectively. The insets in the figures depict the red shift of the fullerene absorption band of the polymer with increasing TX100 concentration. The fullerene content of PAA85-b-C60 demonstrated absorption bands at 256 and 334 nm, and they red-shifted with increasing in TX100 concentration as shown in Figure 4a. The band at 256 nm shifted to 274 nm, and the band at 334 nm was barely detectable when the TX100 concentration reached 0.5 mM, at which the mixture polymer/surfactant precipitates were produced. For C60-b-PAA83-

b-C60, the fullerene exhibited absorption bands at 261 and 333 nm, which were significantly stronger than those of PAA85-bC60 because of the much higher fullerene content possessed by the di-C60 end-capped PAA. The bands red-shifted to 266 and 337 nm, respectively, with increasing TX100 concentration to 0.8 mM as shown in Figure 4b. These phenomena suggested that TX100 bound specifically to the fullerene domains and enhanced the π-π conjugation between TX100 and fullerene and within fullerene itself, resulting in the reduction of the spacing between fullerene molecules as indicated by the red shifts of the fullerene absorption bands. According to the experimental observations obtained from ITC and UV-vis spectroscopy, we concluded that the interaction between TX100 and fullerene-end-capped PAA was initiated by hydrophobic force; however, it is the excellent compatibility and π-π conjugation between the phenyl group of the surfactant and fullerene moieties that localized TX100 molecules in the vicinity of fullerene domains, which triggered the specific binding and induced the observed structural reorganization.9,18,29,30,32

Supramolecular Polyelectrolyte/Surfactant Complex

Langmuir, Vol. 23, No. 17, 2007 8803

Figure 8. Particle size distributions and micrographs demonstrating the morphology change of un-ionized di-end-capped C60-b-PAA83-b-C60 induced by the binding of TX100: (a) CTX100 ) 0 mM; (b) CTX100 ) 0.7 mM; (c) CTX100 ) 1.7 mM; (d) CTX100 ) 3.5 mM.

Structural Reorganization in the Course of Binding. Dynamic light scattering and TEM studies were carried out to examine the structural reorganization of the polymers induced by the polymer/surfactant complexation. The decay rates (Γ) determined from DLS exhibited a linear relationship with q2, which confirmed that the distribution functions were caused by translational diffusion of polymer aggregates or the polymer/ surfactant complex. The hydrodynamic radii (Rhapp) of 5 mM PAA85-b-C60 and C60-b-PAA83-b-C60 at R ) 0 determined from dynamic light scattering were plotted against the TX100 concentration in Figure 5. Both polymers formed a large compound micelle (LCM) in the absence of surfactant with a hydrodynamic radius of ∼120 nm for PAA85-b-C60 and ∼100 nm for C60-b-PAA83-b-C60.4,5,9 The particle size of C60-b-PAA83b-C60 was slightly smaller than that of PAA85-b-C60 because of its larger fraction of fullerene, resulting in a stronger hydrophobic association and thus a more compact micellar structure.9 The dependence of the particle sizes on the TX100 concentration for the two polymers exhibited similar trends: Rhapp remained essentially constant until the TX100 concentration reached a

critical value, where Rhapp increased sharply and the scattering intensity also increased by several orders of magnitude, signaling the commencement of polymer/surfactant complexation that triggered a morphological change and precipitation of polymeric micelles; thereafter, Rhapp leveled off over a certain range of TX100 concentration. It then decreased rapidly at higher TX100 concentration, the scattering intensity also decreased drastically, and the solution became transparent again, which corresponded to the resolubilization of the precipitated PSC. Figure 6 shows the comparison of the dependence of Rhapp and enthalpy on the concentration of TX100 for PAA85-b-C60 at R ) 0. It is interesting to note that the maximum particle size region and the exothermic hump observed on the enthalpy curve occurred within the same concentration regime of TX100 (from 0.4 to 1.7 mM). This furthermore confirmed that the exothermic hump corresponded to the structural reorganization resulting in the precipitation of PSC. Although the particle sizes of the two polymers exhibited similar trends with respect to the TX100 concentration, the onset of complexation, size of the PSC, and TX100 concentration regime

8804 Langmuir, Vol. 23, No. 17, 2007

corresponding to polymer/surfactant complexation were significantly different. For PAA85-b-C60, Rhapp increased sharply from ∼134 to ∼510 nm at CTX100 ≈ 0.4 mM, whereas the Rhapp of C60-b-PAA83-b-C60 increased from ∼98 to ∼344 nm at CTX100 ≈ 1 mM. The complexation of C60-b-PAA83-b-C60 occurred at higher TX100 concentration, and the size of the PSC was smaller than that of PAA85-b-C60. This is caused by a higher proportion of fullerene, which resulted in a stronger hydrophobic association of fullerene moieties, requiring more surfactant molecules to induce the structural transformation of the polymeric micelles. With further addition of TX100, Rhapp decreased rapidly from ∼680 to ∼250 nm at CTX100 ≈ 1.6 mM and from ∼608 to ∼96 nm at CTX100 ≈ 3 mM for PAA85-b-C60 and C60-b-PAA83-b-C60 respectively. Apparently, the binding-induced complexation of di end-capped C60-b-PAA83-b-C60 took place over a broader range of TX100 concentration. This suggests that the polymer with higher fullerene content binds more TX100 molecules, which furthermore confirmed the selective binding of TX100 to fullerene domains. In excess amounts of surfactant, the complexes for both polymers were dissociated by polymer-bound TX100 micelles with their hydrophilic head groups extending outward, resulting in resolubilization and reduction in the particle size and scattering intensity. A series of TEM images corresponding to different regimes as marked in Figure 5 are summarized in Figure 7, showing the morphology changes of 5 mM un-ionized PAA85-b-C60 with increasing TX100 concentration. A schematic representing the association of the complex is correspondingly shown in Figure 9. [Note: It should be stressed that the schematics shown in Figures 9 and 10 are meant to illustrate the possible aggregation phenomenon, and quantitative assessment of these structures will require SAXS or SANS studies.] Figures 7a and 9a show the LCMs of un-ionized PAA85-b-C60 with a radius of 100 nm, which is consistent with the particle size obtained from light scattering measurements. When the TX100 concentration was increased to 1.0 mM, spherical and dense particles with a smooth surface and much larger size were observed, characterizing the formation of a precipitated PSC induced by surfactant binding. The hydrophobic binding is strengthened by the good compatibility between the hydrophobic segment of TX100 and fullerene; thus, the surfactant micelles are able to encapsulate the fullerene clusters within the LCMs. The selective binding may expand the fullerene clusters and release some fullerene moieties. The “free” polymer chains together with bound TX100 molecules may associate and act as physical cross-linkers that enhance interparticle association between the polymeric micelles, resulting in the formation of a flocculated PSC as shown in Figures 7b and 9b. The radii of the PSCs vary from 450 to 600 nm, which is in good agreement with the size determined from dynamic light scattering. With further addition of TX100 to 1.7 mM, more surfactant micelles are bound on the fullerene clusters within the LCMs with their hydrophilic head groups extending outward to enhance the overall hydrophilicity of the particles. Thus, the dense precipitates of the PSC disintegrated into smaller aggregates with a broad size distribution and irregular shapes as shown in Figures 7c and 9c. When the concentration of TX100 exceeded 2 mM, the complex dissociated completely into smaller particles of radius ranging from 60 to 100 nm. The polymeric micelles were coated and stabilized by a wetting layer of bound TX100 as shown in the micrograph (Figures 7d and 9d). These results demonstrated that the addition of TX100 significantly influenced the self-assembly of fullerene-containing polymers and is an effective way to tune the size, shape, and solubility of the polymeric micelles.

Wang et al.

Figure 9. Mechanism of the binding and induced morphological change of the PAA85-b-C60/TX100 system at R ) 0 in an aqueous medium: (a) CTX100 ) 0 mM, large compound micelle of PAA85b-C60; (b) CTX100 ) 0.5 mM, selective binding of TX100 to fullerene domains, which releases polymer chains; (c) 0.5 mM < CTX100 < 1.7 mM, precipitated polymer/surfactant complex; (d) CTX100 > 1.7 mM, LCMs stabilized by a wetting layer of TX100 in an excess amount of surfactant.

Figure 8 shows a series of micrographs together with particle size distributions depicting the morphology change of 5 mM di end-capped C60-b-PAA83-b-C60 induced by the binding of TX100 (the micrographs where the samples were taken are marked in Figure 5). Correspondingly, a schematic depicting the probable aggregation behavior is shown in Figure 10. The polymer formed LCMs in the absence of TX100 as shown in Figures 8a and 10a, where an abundance of spherical particles with a smooth surface

Supramolecular Polyelectrolyte/Surfactant Complex

Langmuir, Vol. 23, No. 17, 2007 8805

polymeric clusters from the compound micelles. The “free” fullerene moieties with bound TX100 together with the un-ionized PAA blocks tend to associate with each other driven by hydrophobic interaction,24,35 which induced the formation of a flocculated PSC. However, the morphology of the PSC was completely different from that of the mono end-capped PAA85b-C60/TX100 system; the PSC was arranged into a lamellar structure with a dimension of ∼500 nm and a thickness of ∼10 nm as demonstrated in Figures 8c and 10c. The aggregation of fullerene pendants at both sides of the PAA blocks was stabilized and solubilized by TX100 molecules; this expanded the PAA segments and promoted the formation of sheet-like structures, which are similar to the lamellar stack structure observed in amphiphilic block copolymers and complexes between oligomeric amphiphiles and polymers.29,30,32 When the concentration of TX100 reached 3.5 mM, the precipitated PSCs were resolubilized as more TX100 molecules were bound to the polymer with their hydrophiles pointing outward, and the PSC dissociated and the structure rearranged into a spherical structure which possessed lower interfacial energy and was more stable in an aqueous medium (as shown in Figures 8d and 10d). The radii of the spherical particles varied from ∼90 to ∼120 nm. The morphology of fullerene-grafted PAA/TX100 systems is responsive not only to the TX100 concentration, but also to pH. Our previous study has demonstrated that when R exceeded 0.2 (pH > 6), the PSC was dissociated and the binding completely ceased.20 This finding is in agreement with the observation depicted in Figure 3, showing that the structural reorganization ceased when R was greater than 0.25. As discussed earlier, neutralization of PAA enhanced the hydrophilicity of the polymer and considerably reduced the hydrophobic attraction between the polymers and TX100, indicating that an overall hydrophobicity of the polymer backbone is still crucial in inducing surfactant binding, and this was discussed previously.20

Conclusions

Figure 10. Mechanism of the binding and induced morphological change of the C60-b-PAA83-b-C60/TX100 system at R ) 0 in an aqueous medium: (a) CTX100 ) 0 mM, large compound micelle of PAA85-b-C60; (b) CTX100 ) 0.7 mM, selective binding of TX100 to fullerene domains, which releases polymer chains; (c) 0.7 mM < CTX100 < 2.5 mM, precipitated polymer/surfactant complex; (d) CTX100 > 2.5 mM, LCMs stabilized by a wetting layer of TX100 in an excess amount of surfactant.

and uniform size were observed. The micellar radius ranges from ∼90 to ∼100 nm, which was consistent with the size determined from dynamic light scattering. With the addition of 0.7 mM TX100 to the polymer solution, small polymer clusters with a diameter of ∼10 nm were released from the LCMs caused by the specific binding of TX100 to the fullerene domains within the polymeric micelles, which disrupted the compound micelles and slightly reduced their size to ∼90 nm (Figures 8b and 10b). With a further increase in TX100 concentration to 1.7 mM, the C60-b-PAA83-b-C60/TX100 mixture formed a precipitated complex as shown in Figures 8c and 10c. The continuous partitioning of TX100 into the fullerene domains released additional smaller

In summary, we observed the unique specific binding of TX100 to fullerene domains within the large compound micelles of PAA85-b-C60 and C60-b-PAA83-b-C60, which induced a structural transformation of polymeric micelles and opened possibilities to tune the morphology and solubility of C60-containing polymers using an aromatic nonionic surfactant. The polymers both existed as large compound micelles and formed PSC precipitates in the presence of TX100. Finally the precipitates were resolubilized by a wetting layer of bound TX100 molecules. However, the morphology of the PSC was different, depending on the polymer architecture: mono end-capped PAA85-b-C60 formed a dense spherical PSC with a radius of ∼600 nm, whereas the PSC produced from the di end-capped C60-b-PAA83-b-C60 possessed a lamellar structure with a dimension of ∼500 nm and a thickness of ∼10 nm. We believe this research work opens a new and simple approach to directly control the assembly of fullerenecontaining polymers using aromatic surfactants, which may find important applications in drug delivery, nanoscale patterning, miniaturized biological devices, and many others. Acknowledgment. We acknowledge the financial support provided by the Nanyang Technological University and Singapore-MIT Alliance (SMA). LA700600R