Temperature- and Light-Responsive Blends of Pluronic F127 and Poly

Langmuir 2007, 23, 11475-11481

11475

Temperature- and Light-Responsive Blends of Pluronic F127 and Poly(N,N-dimethylacrylamide-co-methacryloyloxyazobenzene) Carmen Alvarez-Lorenzo,*,† Smeet Deshmukh,‡ Lev Bromberg,‡ T. Alan Hatton,‡ Isabel Sa´ndez-Macho,§ and Angel Concheiro‡ Departamento de Farmacia y Tecnologı´a Farmace´ utica, Facultad de Farmacia, UniVersidad de Santago de Compostela, 15782-Santiago de Compostela, Spain, Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, and Departamento de Quı´mica Fı´sica, Facultad de Farmacia, UniVersidad de Santiago de Compostela, 15782-Santiago de Compostela, Spain ReceiVed July 2, 2007. In Final Form: August 29, 2007 Photoresponsive poly(N,N-dimethylacrylamide-co-methacryloyloxyazobenzene) (DMA-MOAB) and temperatureresponsive Pluronic F127 (F127) copolymers were blended to obtain systems responsive to both stimuli that are potentially useful for pharmaceutical formulations. The random DMA-MOAB copolymer undergoes a trans to cis isomerization when irradiated by 366 nm light, which modifies both the air-water interfacial behavior and the self-associative properties of the copolymer. Under dark conditions the azobenzene groups of DMA-MOAB in the trans conformation self-associate and the interactions with F127 are minimal. The cis conformation of the azobenzene groups of the DMA-MOAB copolymer is relatively more hydrophilic than the trans conformation, which causes the copolymer micelles to dissociate upon irradiation, allowing the unimers to form mixed micelles with the F127. This causes the sol-gel transition temperature of the DMA-MOAB:F127 blend to be 10 °C lower upon irradiation at 366 nm compared to that for the dark conditions. It has been found that F127 (10-12 wt %):DMA-MOAB (5-6 wt %) aqueous solutions have at body temperature a low viscosity when equilibrated in the dark and undergo a sol-gel transition when irradiated. Such a transition strongly alters the diffusion of solutes such as methylene blue within the solutions. This light-induced interaction between the azobenzene moieties of DMA-MOAB and F127 micelles disappears when hydroxypropyl-β-cyclodextrin (HPβCD) is added to the medium. In the presence of HPβCD, the cis-azobenzene groups are hosted in the cyclodextrin cavities and the mixed micelles are not formed. Therefore, changes in HPβCD concentration could be used to modulate the response of the copolymer blends to light.

Introduction Stimuli-sensitive materials find a wide range of applications in the biomedical field, ranging from cell scaffolds to site-specific drug delivery systems.1 In particular, temperature-responsive polymers may provide liquid systems with in-situ-gelling ability to form solid drug depots once in a contact with the living tissues.2 The temperature responsiveness is due to conformational changes caused by the entropy-driven association of moieties on the polymer chains with a temperature-dependent water solubility.3 The appearance of associations with resulting aggregates acting as physical cross-links between polymer chains lead to a gelation or to a complete phase separation.4 Pluronic or poloxamer copolymers, consisting of poly(ethylene oxide) (PEO) and poly(propylene oxide) (PPO) blocks (PEO-b-PPO-b-PEO), are prevalent representatives of temperature-responsive copolymers used as pharmaceutical excipients. In aqueous solution, the copolymer molecules aggregate forming spherical micelles with PPO core and a hydrated PEO shell. If the concentration is high enough, an increase in temperature causes an ordered packing of the micelles and their entanglement, inducing gelation of the * To whom correspondence should be addressed. Fax: 34981547148. E-mail: [email protected]. † Departamento de Farmacia y Tecnologı´a Farmace ´ utica, Facultad de Farmacia, Universidad de Santago de Compostela. ‡ Department of Chemical Engineering, Massachusetts Institute of Technology. § Departamento de Quı´mica Fı´sica, Facultad de Farmacia, Universidad de Santiago de Compostela. (1) Packhauser, C. B.; Schnieders, J.; Oster, C. G.; Kissel, T. Eur. J. Pharm. Biopharm. 2004, 58, 445-455. (2) Ruel-Garie´py, E.; Leroux, J-C. Eur. J. Pharm. Biopharm. 2004, 58, 409426. (3) Tanaka, T. ACS Symp. Ser. 1992, No. 480, 1-21. (4) Chevillard, C.; Axelos, M. A. V. Colloid Polym. Sci. 1999, 275, 537-545.

system.5 A number of studies on the performance of Pluronic F127 (F127), which is approved for intravenous, inhalation, oral solution, suspension, ophthalmic, and topical formulations,6 as a micellar carrier of drugs or as a component of in situ gelling systems have been reported.7 It is known that the polymerpolymer and polymer-water interactions can be altered by the addition of cosolvents or other polymeric components, which strongly modify the sol-gel transition temperature and the gel strength and, as a consequence, any property dependent on the micro- and macroviscosity of the system, such as drug diffusion/ release.5,8-10 Photoresponsive systems have found a wide range of applications as sensors or actuators in various fields.11 Some recent papers have proved the interest of these systems as membranes able to control the transfer of ions or the flow of monatomic gases or liquid fluids through microchannels.12-15 The develop(5) Dumortier, G.; Grossiord, J. L.; Agnely, F.; Chaumeil, J. C. Pharm. Res. 2006, 23, 2709-2728. (6) Rowe, R.; Sheskey, P.; Weller, P.J. Handbook of Pharmaceutical Excipients, 4th ed.; Pharmaceutical Press and American Pharmaceutical Association: London, Washington, DC, 2003; p 447. (7) Kabanov, A. V.; Batrakova, E. V.; Alakhov, Yu, V. J. Controlled Release 2002, 82, 189-212. (8) Gilbert, J. C.; Richardson, J. L.; Davies, M. C.; Palin, K. J. J. Controlled Release 1987, 5, 113-118. (9) Eeckman, F.; Amighi, K.; Moe¨s, J. Int. J. Pharm. 2001, 222, 259-270. (10) Pisal, S. S.; Paradkar, A. R.; Mahadik, K. R.; Kadam, S. S. Int. J. Pharm. 2004, 207, 37-45. (11) Barrett, C.; Mermut, O. PMSE Prepr. 2005, 92, 51-52. (12) Rosario, R.; Gust, D.; Hayes, M.; Jahnke, F.; Springer, J.; Garcia, A. A. Langmuir 2002, 18, 8062-8069. (13) Kameda, M.; Sumaru, K.; Kanamori, T.; Shinbo, T. J. Appl. Polym. Sci. 2003, 88, 2068-2072. (14) Sumaru, K.; Ohi, K.; Takagi, T.; Kanamori, T.; Shinbo, T. Langmuir 2006, 22, 4353-4356.

10.1021/la7019654 CCC: $37.00 © 2007 American Chemical Society Published on Web 10/05/2007

11476 Langmuir, Vol. 23, No. 23, 2007

ment of suitable new materials and the study of the photoregulation of the permeation of greater molecules may open new possibilities for photoresponsiveness in drug delivery. Light can be externally applied to the body to switch on and off drug release at a specific site, offering a potential for controlling the release that can be difficult to achieve using other stimuli.16 Research in this field has been mainly focused on copolymer micelles or liposomes that can be dissociated by the action of light.17-21 The photoinduced conformational changes are typically due to the photoisomerization of a dye molecule such as azobenzene bonded to the polymer. The trans-cis isomerization of azobenzene groups alters the polarity and conformation of the polymer chains in a rapid and reversible fashion. The self-association of the azobenzene groups that occurs in the trans conformation is lost when, upon exposure to UV light, the relatively more hydrophilic cis conformation is adopted.22 This causes a decrease in the viscosity, which can facilitate the diffusion of target solutes.23 Implementation of temperature-induced gel systems with photocontrolled release capability may enhance the performance of formulations that can be administered as a free flowing fluid that gels in situ due to the temperature change; the viscosity changes upon exposure to light enable external tuning of drug delivery from the depot using an adequate light source. Hence, this work is aimed at development of temperature- and lightresponsive solutions based on blends of F127 and poly(N,Ndimethylacrylamide-co-methacryloyloxyazobenzene) (DMAMOAB) copolymer. The ability of the DMA-MOAB copolymer to change conformation upon irradiation in aqueous solutions has been recently reported.24 Herein, we are interested in exploring the effect of conformational changes in DMA-MOAB in response to changes in irradiation wavelength on the temperature-responsive behavior of pluronic micellar solutions. DMA-MOAB is amphiphilic, with aqueous solubility depending on the content of the hydrophobic MOAB groups. With a MOAB content of around 20 mol % of the polymer, most of the azobenzene groups can find their way into the pluronic micelles. In this paper, the lightresponsive properties of the DMA-MOAB copolymer in the absence and presence of F127 were evaluated through changes in the electronic absorption spectra, the π-A isotherms, and the size of aggregates formed upon irradiation. Further, gelling and probe diffusion properties of aqueous solutions containing blends of both copolymers were characterized. Since it is known that cyclodextrins may form complexes with azobenzene groups, altering their self-associative behavior,25 the effects of the hydroxypropyl-β-cyclodextrin addition to the DMA-MOAB solutions were also investigated. (15) Garcia, A.; Marquez, M.; Cai, T.; Rosario, R.; Hu, Z.; Gust, D.; Hayes, M.; Vail, S. A.; Park, C-D. Langmuir 2007, 23, 224-229. (16) Jiang, J.; Tong, X.; Morris, D.; Zhao, Y. Macromolecules 2006, 39, 46334640. (17) Bisby, R. H.; Mead, C.; Morgan, C. C. Biochem. Biophys. Res. Commun. 2000, 276, 169-173. (18) Kishor Mal, N.; Fujiwara, M.; Tanaka, Y.; Taguchi, T.; Matsukata, M. Chem. Mater. 2003, 15, 3385-3394. (19) Szczubialka, K.; Nowakowska, M. Polymer 2003, 44, 5269-5274. (20) Eastoe, J.; Vesperinas, A.; Donnewirth, A-C.; Wyatt, P.; Grillo, I.; Heenan, R. K.; Davis, S. Langmuir 2006, 22, 851-853. (21) Levrand, B.; Herrmann, A. FlaVour Fragrance J. 2006, 21, 400-409. (22) Tong, X.; Wang, G.; Soldera, A.; Zhao, Y. J. Phys. Chem. B 2005, 109, 20281-20287. (23) Lee, C. T., Jr.; Smith, K. A.; Hatton, T. A. Macromolecules 2004, 37, 5397-5405. (24) Deshmukh, S.; Bromberg, L.; Smith, K. A.; Hatton, T. A. Polym. Mater. Sci. Eng. 2006, 51, 878-880. (25) Zheng, P.; Hu, X.; Li, L.; Tam, K. C.; Gan, L. H. Macromol. Rapid Commun. 2004, 25, 678-682.

AlVarez-Lorenzo et al.

Materials and Methods Materials. Methacryloyl chloride (97%), 4-hydroxyazobenzene (98%), N,N-dimethylacrylamide (99%), 2,2′-azobis(isobutyronitrile) (AIBN, 98%), methylene blue (MB), and Pluronic F-127 (F127; PEO99-PPO69-PEO99) were from Sigma-Aldrich Chemical Co. (St. Louis, MO) and used as received. Triethylamine (99%) was obtained from J.T. Baker (Phillipsburg, NJ). HPβCD (degree of substitution 4.6, Mr ) 1310 Da) was from Janssen Pharmaceutica (Beerse, Belgium). Ultrapure water was obtained by reverse osmosis (MilliQ, Millipore Iberica SA, Madrid, Spain). All other reagents, solvents and gases were obtained from commercial sources and were of the highest purity available. DMA-MOAB Copolymer Synthesis. Synthesis of the DMAMOAB random copolymer has been reported previously.24 In brief, trans-4-methacryloyloxyazobenzene (MOAB) was prepared by condensation of 4-hydroxyazobenzene (50 mmol) and methacryloyl chloride (150 mmol) in THF, with triethylamine acting as a promoter and 2,6-di-tert-butyl-p-cresol as free-radical inhibitor. After purification, MOAB (0.1 g, 3.8 mmol) was copolymerized with DMA (0.154 mL, 15 mol) via free-radical polymerization using AIBN (5 mg, 2 wt %) in THF (0.15 mL). The solution was maintained at 70 °C for 18 h and then at 90 °C for 2 h. The resulting copolymer was dialyzed for 48 h against deionized water using a 3500 Da MWCO membrane (Spectra/Por, Spectrum Labs, Rancho Dominguez, CA) and then lyophilized. DMA-MOAB was finally purified by suspending it in THF for 3 days and removing the undissolved residue by filtration using GHP membrane filters (0.45 µm, Waters Corp., Milford, MA). The dissolved fraction was dried under vacuum. The 1H NMR spectra of copolymer solutions in CD2Cl2 (Bruker AMX400 spectrometer, Billerica, MA) indicated that the MOAB molar content was 20 mol % (based on the ratio of the protons in the azobenzene moiety to those of the methyl group of DMA). The weight-average molecular weight was of 11.5 kDa estimated by gel permeation chromatography (GPC) carried out using THF as the mobile phase.24 Preparation of F127:DMA-MOAB Blends and Aqueous Solutions. To obtain a homogeneous blend, DMA-MOAB (10 g) and F127 (10 or 20 g) were dissolved in 100 mL of ethanol and the solvent was removed under vacuum. To prepare solutions with different total copolymers concentration, keeping constant the F127: DMA-MOAB 1:1 or 2:1 weight ratio, adequate amounts of the dried mixture were added to cold water under magnetic stirring and then stored at 4 °C for a complete dissolution. HPβCD was added in different proportions to some of these solutions. UV Absorption Spectra. The UV absorption spectra (Agilent 8453 spectrophotometer, Bo¨blingen, Germany) of diluted solutions of DMA-MOAB in water (approximately 0.012 wt %), with or without F127, HPβCD, or MB, were recorded at 20 °C before and after irradiation at 366 nm using an 8 W lamp (Camag, Muttenz, Switzerland). The UV-stability of MB in solution was confirmed by maintaining MB solutions under 366 nm for several hours and then recording the UV spectra. Langmuir Monolayers. The π-A isotherms were recorded on a NIMA surface balance (Coventry, U.K.) equipped with a single barrier NIMA 601 hydrophobic Teflon-made trough, of total area 550 cm2, placed on an antivibration table. Prior to experiments, the trough was thoroughly cleaned and the surface was checked for impurities. Monolayers were prepared by spreading 50 ( 0.2 µL (Hamilton microsyringe) of a solution in chloroform of DMAMOAB or of F127:DMA-MOAB 2:1 weight ratio mixtures (1 mg/ mL) on the aqueous subphase maintained at 20 ( 0.1 °C. After spreading, the monolayers were left for 10 min to ensure solvent evaporation and afterward were compressed using a barrier speed of 15 cm2/min. The surface pressure was measured with the accuracy of ( 0.1 mN/m, using a Wilhelmy plate made from chromatography paper (Whatman Chr1, Brentford, U.K.) as a pressure sensor. The monolayer stability was verified by monitoring the change in surface pressure while holding the area constant. To evaluate the effect of the light wavelength on the monolayer at the air-water interface, a light source 366 nm was used (8 W, Camag, Muttenz, Switzerland).

F127:DMA-MOAB Blends Before the experiments the samples were irradiated for 30 min, and the irradiation was maintained during the monolayer compression. Additionally, the reflection spectrum of the monolayer was recorded with RefSpec (Nanofilm Technology NFT, Go¨ttingen, Germany). The difference in reflectivity, ∆R, between the monolayer-covered water surface and the bare water surface was determined. Dynamic Light Scattering (DLS). DLS measurements were performed using an ALV-5000F optical system equipped with CW diode-pump Nd:YAG solid-state laser (400 mW) operated at 532 nm (Coherent Inc., Santa Clara, CA). The intensity scale was calibrated against scattering from toluene. DMA-MOAB solutions (0.1 wt %), with or without F127, were filtered (Millipore 0.45 µm, Carringtwohill, Ireland) into the quartz cell (previously washed with condensing acetone vapor) and maintained at visible light or irradiated at 366 nm for 10 min. Solutions containing F127:DMA-MOAB: HPβCD 2:1:1 weight ratios were also analyzed. The diffusion coefficient was deduced from the standard second-order cumulant analysis of the autocorrelation functions measured at 90° angle. The experiments were carried out in triplicate, and the apparent hydrodynamic radius (rh,app) of the micelles was estimated from the apparent diffusion coefficients. Oscillatory Rheometry. The evolution of the storage or elastic (G′) and the loss or viscous (G′′) moduli of F127:DMA-MOAB 2:1 weight ratio solutions from 5 to 45 °C were evaluated in triplicate at 0.5 rad/s, using a temperature ramp of 2 °C/min, in a Rheolyst AR-1000N rheometer (TA Instruments, Newcastle, U.K.) equipped with an AR2500 data analyzer, a Peltier plate, and a quartz plate geometry (4 cm diameter). The sample was removed from the fridge (4 °C) and put directly on the Peltier plate of the rheometer and assayed under dark conditions. The temperature was then returned to 5 °C, and the sample was irradiated at 366 nm (8 W, Camag, Muttenz, Switzerland) for 10 min. The measurements began again with the lamp still on. Additional experiments were carried out with copolymer solutions containing also MB or HPβCD. The temperature at which G′ and G′′ cross over has been arbitrarily chosen as the gelation temperature (Tgel). MB Diffusion Assays. Samples of 1 mL of DMA-MOAB or of F127:DMA-MOAB 2:1 weight ratio aqueous solutions with various total copolymer content, and containing 0.02 mg of MB, were placed in the donor chambers of glass Franz-Chien diffusion cells (VidraFoc, Valencia, Spain). The temperature was set at 37 °C, and the surface available for diffusion was 0.785 cm2. A 0.20 µm cellulose acetate filter (Albet, Barcelona, Spain) or dialysis tubing (cut off 12 000 Da, Sigma, St. Louis, MO) was used to separate the donor and receptor compartments. The receptor compartment was filled with 5.5 mL of water and stirred gently with a small magnetic bar. Samples were taken from the sampling port to measure the absorbance at 665 nm and then returned to the recipient cell. Four donor cells were irradiated at 366 nm, four were under visible light, and four were covered to attain dark conditions.

Results and Discussion Blends of DMA-MOAB and F127 dissolved in water or organic solvents such as tetrahydrofuran or chloroform form clear solutions even at DMA-MOAB concentrations exceeding 10 wt %. Electronic absorption spectra of the blend solutions stored in the dark exhibited a peak at 335 nm due to the π f π* transitions of the azobenzene25,26 (Figure 1). When the DMAMOAB solutions were exposed to 366 nm light, the absorbance peak became less intense and shifted from 335 to 346 nm, while a weak peak at 430 nm slightly increased; this corresponds to the n f π* transition in azobenzene.25,26 When the solutions were left in the dark, the original spectrum recovered completely over time. These results illustrate the reversibility of the photoisomerization of DMA-MOAB copolymer. Analogous trans-cis isomerization of the azobenzene moiety of DMA(26) Cicciarelli, B. A.; Hatton, T. A.; Smith, K. A. Langmuir 2007, 23, 47534764.

Langmuir, Vol. 23, No. 23, 2007 11477

Figure 1. UV-visible spectra of DMA-MOAB (0.012 wt %) solutions in water at 20 °C before irradiation (a), after irradiation for 30 min at 366 nm (b), after irradiation for 30 min at 366 nm and 30 min in the dark (c), and after irradiation for 30 min at 366 nm and 3 days in the dark (d). The wavelengths of two peaks of each spectrum are indicated on the plot.

Figure 2. UV-visible spectra at 20 °C of F127:DMA-MOAB 2:1 aqueous solutions containing methylene blue before (a) and after irradiation at 366 nm for 10 min (b). The spectrum of methylene blue aqueous solution is also shown for reference (c).

MOAB was observed in the presence of F127 or MB (Figure 2). Regarding the kinetics of photoisomerization of DMA-MOAB in aqueous medium, it has been observed that the rate constants for the trans to cis conversion exhibited a sigmoid dependency on polymer concentration, with values ranging from 0.12 s-1 when the copolymer stays in its unimeric form (i.e., at concentration e0.005 wt %) to 0.017 s-1 when aggregated hydrophobic domains exist (g0.1 wt %), the rates for the reverse cis to trans conversion being very similar.24 This means that at any DMA-MOAB concentration the isomerization occurs on the time scale of minutes. The dark-adapted state is obtained under conditions where there is no irradiation at all; this is the lowest energy state when all the azobenzene moieties are in the trans form. A photostationary state under visible irradiation corresponds to a mixture of both isomers. On the other hand, a photostationary state obtained under UV irradiation corresponds to a large excess (>90% conversion) of the cis isomer. Blends of the amphiphilic DMA-MOAB copolymer and F127 exhibited excellent aqueous solubility, which afforded both irradiation- and temperature-sensitivity of the aqueous solutions. We observed that a blend composed of F127 and DMA-MOAB at a 2:1 weight ratio enables preparation of highly concentrated aqueous solutions with up to 10 wt % DMA-MOAB with a pronounced trans-cis isomerization. To gain an insight into the repercussion of such an isomerization on the macroscopic behavior of the DMA-MOAB solutions and blends with F127,

11478 Langmuir, Vol. 23, No. 23, 2007

AlVarez-Lorenzo et al.

Figure 4. π-A diagrams of F127:DMA-MOAB 2:1 weight ratio before and after being irradiated at 366 nm, at 20 °C, using water as the subphase.

Figure 3. (a) π-A diagrams of DMA-MOAB without irradiation and when irradiated at 366 nm. The sample was irradiated for 30 min before the experiment, and the irradiation was maintained during the monolayer compression. The insert shows the reflection spectra of the monolayers compressed at π ) 20 mN/m. (b) Proposed conformations of DMA-MOAB molecules at the air/water interface.

their interfacial and bulk properties were evaluated under various light conditions. DMA-MOAB and F127:DMA-MOAB Isotherms at the Air-Water Interface. The DMA-MOAB with 20 mol % MOAB content is an amphiphilic copolymer with a critical micelle concentration (cmc) of ca. 0.1% as estimated by surface tension measurements.24 To gain insight into the interfacial behavior of this polymer, evolution of the surface pressure as a function of the area available/molecule was evaluated using water as the subphase (Figure 3a). The pattern of the equilibrium surface pressure (π) isotherms was dependent on the irradiation conditions. The compression rate did not modify the profile. When the experiment was carried out in the dark, the isotherm consisted of an expanded liquid phase at low pressures. In the pressure range 16-21 mN/m a transition region is observed, 21 mN/m probably being the collapse pressure. When the isotherm was recorded after irradiating the DMA-MOAB solution for 30 min at 366 nm, greater pressures were observed in the expanded state region; i.e. when the area available/molecule is large (>300 Å2/molecule), the surface pressure exerted by the film of cis DMA-MOAB molecules is greater than the one exerted by the trans conformer. For a given pressure, a greater area is occupied by each DMA-MOAB molecule owing to the lateral expansion of the confined monolayer when the trans-cis isomerization occurs. In the absence of light, the molecules are in a closely packed, planar trans conformation, with the azobenzene groups preferentially oriented perpendicular to the water interface. When in the bent cis conformation under irradiation at 366 nm, the azobenzene moieties come in contact with the water surface due to the increased polarity and are oriented parallel to the water surface. Similar behavior has been observed previously for azobenzene surfactants26-28 and for a block copolymer comprised of azobenzene-containing methacrylate and poly(ethylene ox(27) Kim, I.; Rabolt, J. F.; Stroeve, P. Colloids Surf., A 2000, 171, 167-174. (28) Shang, T.; Smith, K. A.; Hatton, T. A. Langmuir 2003, 19, 1076410773.

ide).29 An overall picture of the physical conformations of the DMA-MOAB adsorbed at the air/water interface is shown in Figure 3b. The plots of the dark and UV samples cross over at 200 Å2/molecule, and below this area the changes in pressure are minor. In the condensed state the surface is saturated and a decrease in the area available/molecule may force some MOABDMA molecules to move from the interface to the bulk. The reflection spectra of the monolayers correlated well with the occurrence of a trans-cis isomerization. The insert in Figure 3a shows the reflection spectra (∆R) obtained for monolayers compressed at π ) 20 mN/m. The spectrum of the nonirradiated monolayer shows a maximum at 325 nm, corresponding to the trans conformation. The intensity of this peak decreased when the monolayer was previously irradiated for 30 min, while the weakest peak at 430 nm slightly increased. The original spectrum was almost recovered when the irradiated monolayer was left in the dark for 30 min. The π-A diagrams of mixed monolayers only revealed small changes as a function of irradiation conditions (Figure 4), probably because the high content of F127 hides the effect of isomerization on the surface pressure. In fact, the isotherm resembles that of F127 itself, which forms expanded monolayers at low surface pressures and undergoes a transition to a condensed film when further compressed.30 DLS. The effects of the trans-cis isomerization on the aggregation of DMA-MOAB molecules among themselves and with F127 were followed using dynamic light scattering. The weight-average hydrodynamic radius of the micelles formed in the dilute aqueous solutions of the DMA-MOAB copolymer, kept in the dark prior to the DLS experiments, was measured to be 42 nm (Figure 5a). When irradiated at 366 nm, the solutions contained no aggregates with radius exceeding 1 nm (Figure 5b). Compared to the solution equilibrated in the dark, the irradiated sample showed a lower number of counts at time 0 after being exposed to 366 nm for 10 min. When the irradiation was interrupted, the counts progressively increased and reached the value initially observed for the nonirradiated sample after 15 min (Figure 6). This indicates that the irradiation causes the number of micelles/aggregates to decrease. When irradiation is stopped, the system recovers the initial trans conformation. This fast recovery was also observed for F127:DMA-MOAB 1:1 weight ratio solutions but not in the case of F127:DMA-MOAB 2:1 weight ratio systems. This suggests that once DMA-MOAB chains undergo the trans to cis isomerization and the azobenzene (29) Kadota, S.; Aoki, K.; Nagano, S.; Seki, T. Colloids Surf., A 2006, 284285, 535-541. (30) O’Connor, S.; Gehrke, S. H.; Retzinger, G. S. Langmuir 1999, 15, 25802585.

F127:DMA-MOAB Blends

Figure 5. Intensity fraction distribution of the apparent hydrodynamic radius for 0.05% DMA-MOAB solution (a, b) and F127: DMA-MOAB:HPβCD 2:1:1 solution (c, d) under different light conditions. In sample c, the peak at 0.6 nm, which corresponds to free HPβCD units, disappeared after irradiation at 366 nm for 10 min (d), mostly due to complexation of the azobenzene groups of DMA-MOAB with HPβCD.

Figure 6. Evolution of the count number as a function of time for DMA-MOAB solution (0.1%) irradiated at 366 nm for 10 min once the irradiation is stopped. The increase in counts is related to the progressive self-aggregation of the azobenzene groups as they recover the trans conformation.

groups dissociate under 366 nm light, interactions with the F127 micelles make the recovery of the initial conformation under dark conditions slower. The greater the F127:DMA-MOAB ratio, the more difficult may be the stacking of the transazobenzene groups. It has been previously reported that inclusion complex formation between azobenzene and β-CD is more favorable for the trans conformer than for the cis conformer.25,31 The DLS analysis of F127:DMA-MOAB:HPβCD 2:1:1 weight ratio solutions stored at dark showed two populations with diameters (31) Takashima, Y.; Nakayama, T.; Miyauchi, M.; Kawaguchi, Y.; Yamaguchi, H.; Harada, A. Chem. Lett. 2004, 33, 890-891.

Langmuir, Vol. 23, No. 23, 2007 11479

centered at 0.6 and 7 nm, respectively (Figure 5c). The first can be attributed to the free HPβCD units.32 After irradiation at 366 nm, the peak at 0.6 nm disappeared, which indicates that as self-association of the azobenzene groups decreases, complexation with cyclodextrin units becomes more favorable (Figure 5d). This means that the self-association of the azobenzene groups at trans conformation makes the complexation with HPβCD difficult despite of being energetically more favorable than when they are at cis conformation. The isomerization to the cis conformation, with the consequent rupture of the aggregates, was needed to enable the cyclodextrin to complexate the azobenzene group. These findings also suggested a preferential interaction of DMA-MOAB with HPβCD compared to F127, which was confirmed in subsequent experiments. Rheometry. The temperature-responsiveness of pluronic solutions in the presence of a photoresponsive copolymer able to alter its conformation as a function of light wavelength has not yet been evaluated. Blends of DMA-MOAB with F127 were prepared to enhance the solubility of DMA-MOAB up a concentration high enough to enable the microscopic structural changes be transmitted to the macroscopic scale. The solutions were characterized rheologically under different light and temperature conditions. The F127:DMA-MOAB 2:1 solutions exhibited a temperature-responsive behavior owing to the presence of F127, with its PPO segments possessing temperature-dependent aqueous solubility (Figure 7). The aggregation pattern of DMA-MOAB is not influenced by temperature significantly. Nevertheless, the presence of the DMA-MOAB copolymer caused a notable decrease in the onset of gelation temperature (Tgel) of the F127 solutions. Additionally, DMA-MOAB provided the system with photoresponsive gelation; both Tgel and rheological behavior being strongly dependent on the irradiation conditions. For example, a 10 wt % F127 does not form gel at any temperature nor when 5% DMA-MOAB is added and the system is maintained at dark. By contrast, gel formation occurs when irradiated at 366 nm. Thus, the 10% F127:5% DMA-MOAB system only undergoes the sol-gel transition at around 37 °C when irradiated (Figure 7a). A 15 wt % F127 solution does not form a gel at 37 °C but does so when 7.5 wt % DMA-MOAB is added to the system at this temperature; this mixture has a Tgel of 27 °C under dark conditions and of 17 °C when the system is exposed to 366 nm light (Figure 7b). In the case of 20 wt % F127 and 10 wt % DMA-MOAB solution, Tgel is 15 °C in the absence of irradiation and 10 °C if irradiated at 366 nm (Figure 7c), while Tgel for 20 wt % F127 alone is 23.4 °C. Therefore, the shifts in Tgel are more remarkable when most of the azobenzene groups in the DMA-MOAB copolymer chains are in their more polar (less hydrophobic), cis state (upon UV irradiation) than when the azobenzene groups are primarily in their more hydrophobic, trans conformation (in the dark). These findings are likely due to the less persistent formation of DMAMOAB aggregates upon UV-irradiation, the cis-azobenzene groups becoming free to incorporate in the pluronic micelles and to form mixed micelles. The formation of the mixed micelles dramatically impacts the Tgel. On the other hand, the persistence of the DMA-MOAB aggregates in the dark diminishes the availability of the azobenzene moieties for incorporation in the pluronic micelles, and thus Tgel changes only to a small extent. Further, the more pronounced incorporation of the DMA-MOAB segments in the hydrophobic core of the F127 micelles upon UV-irradiation increases the density of the physical cross-links, (32) Wu, A.; Shen, X.; He, Y. J. Colloid Interfacial Sci. 2006, 297, 525-533.

11480 Langmuir, Vol. 23, No. 23, 2007

Figure 7. Evolution of the storage (G′, open symbols) and loss (G′′, full symbols) of F127:DMA-MOAB 2:1 weight ratio aqueous solutions with various total contents in copolymers, stored in the dark or after UV-irradiation.

Figure 8. UV-visible spectra of F127:DMA-MOAB:HPβCD 2:1:1 aqueous solutions after being stored in the dark or after irradiation at 366 nm for 10 min at 20 °C. The irradiated samples were stored again in the dark, and the spectra were recorded after 1, 5, 8, and 30 h. The MOAB-DMA concentration was 0.012%.

thereby increasing G′ when the majority of the azobenzene groups are in their cis conformation (see specifically Figure 7a,b). To test the hypothesis of the hydrophobic interaction of the aromatic groups of the DMA-MOAB with the F127 micelles, additional experiments were carried out by adding HPβCD to the solutions. It has been reported previously that the complexation

AlVarez-Lorenzo et al.

Figure 9. Methylene blue (MB) diffusion from F127:DMA-MOAB 2:1 weight ratio solutions, through a 0.20 µm pore size membrane under different light conditions. The concentration of DMA-MOAB in the donor receptor was 5% (a), 6% (b), and 7.5% (c), respectively.

of the azobenzene groups with HPβCD prevents their selfassociation.25 The DLS data indicated that such complexation indeed occurs. We have also previously observed that this cyclodextrin can form pseudorotaxanes with F127 although the changes in micellization are not large.33 The addition of HPβCD up 3.75 wt % to a DMA-MOAB 7.5 wt %-F127 15 wt % system did not significantly affect the influence of light on Tgel. By contrast, 7.5 wt % HPβCD caused the system to lose the light-induced gelation; i.e., Tgel remained constant regardless of the irradiation conditions. It is important to note that HPβCD did not prevent the F127:DMA-MOAB 2:1 solutions from undergoing the trans to cis isomerization and that the isomerization was reversible, as can be seen in the UV-spectra (Figure 8). However, azobenzene groups in the cis conformation may form complexes with HPβCD, thereby inhibiting their interactions with the F127 micelles. In a F127:DMA-MOAB:HPβCD 2:1: 0.5 weight ratio solution, there are 0.23 units of cyclodextrin/ MOAB group, and thus, a sufficient number of azobenzene groups remain free to interact with the F127 micelles. Therefore, the (33) Rodriguez-Perez, A.; Rodriguez-Tenreiro, C.; Alvarez-Lorenzo, C.; Concheiro, A.; Torres-Labandeira, J. J. J. Nanosci. Nanotechnol. 2006, 6, 31793186.

F127:DMA-MOAB Blends

system maintains its light-responsive Tgel. The lack of lightresponsiveness of F127:DMA-MOAB:HPβCD 2:1:1 weight ratio systems can be explained by the greater proportion of MOAB groups (almost 50%) that form complexes with the cyclodextrin units. Thus, the Tgel of the system becomes independent of the irradiation conditions when there are insufficient free azobenzene groups to interact with the F127 micelles. These results indicate that the effect of light on the Tgel of the blend system is due to its influence on the extent of the hydrophobic interactions between both copolymers. MB Diffusion Assays. The light- and temperature-responsiveness of the F127:DMA-MOAB blend solutions prompted an evaluation of the extent to which the conformational changes alter the diffusion of a hydrophilic solute. MB is expected to diffuse in the aqueous regions of the polymer matrix, while avoiding the hydrophobic cores. Further, MB can be easily quantified in the receptor medium because its absorption spectrum does not overlap with that of DMA-MOAB (as can be seen in Figure 2). MB, which is a photosensitizing agent effective toward dermal infections caused by antibiotic-resistant bacteria and yeasts,34 remains stable under irradiation at 366 nm. The diffusion profiles are shown in Figure 9. As expected, the higher the concentration of copolymers in the system, the lower the diffusion rate owing to the larger microviscosity. The diffusion rate of MB was not affected by the irradiation conditions when included in 10 wt % F127:5 wt % DMA-MOAB (Figure 9a) or 15 wt % F127:7.5 wt % DMA-MOAB (Figure 9c) systems. However, for the 12 wt % F127:6 wt % DMA-MOAB system, the MB diffusion rate under dark conditions was significantly faster than when the experiment was carried out at visible light or under 366 nm irradiation (Figure 9b). This is explained by the rheological behavior of the F127:DMA-MOAB solutions as a function of light and temperature (Figure 7). At the temperature of the donor cell, 30 °C, the solution containing 10 wt % F127 and 5 wt % (34) Wainwright, M.; Phoenix, D. A.; Laycock, S. L.; Wareing, D. R. A.; Wright, P. A. FEMS Microbiol. Lett. 1998, 160, 177-181.

Langmuir, Vol. 23, No. 23, 2007 11481

DMA-MOAB remains as a low viscosity system regardless of the light conditions. By contrast, the system consisting of 15 wt % F127 and 7.5 wt % DMA-MOAB is in the gel state at this temperature independent of the light conditions. Therefore, a dramatic effect of the irradiation on MB diffusion can only be seen for the 12 wt % F127:6 wt % DMA-MOAB system that undergoes a sol-gel transition at 30 °C when irradiated at 366 nm (Figure 9b). The increase in the number of physical crosslinks can cause increased hindrance to diffusion of MB leading to a slower diffusion rate. The trans to cis isomerization rate, under 366 nm light, is fast enough to enable an effective control of the rheological and diffusional properties of F127:DMAMOAB blends for a wide range of practical purposes, including drug delivery.

Conclusions The light-induced trans-cis isomerization of DMA-MOAB strongly alters the interaction of this copolymer with F127 micelles and, as a consequence, modifies the sol-gel temperature of the system. Therefore, it is possible to prepare a liquid system of low viscosity in the dark and to increase the viscosity rapidly when light is applied at body temperature. The addition of a cyclodextrin able to form inclusion complexes with the azobenzene groups of DMA-MOAB can serve to modulate the light-responsiveness of the systems and, consequently, also its temperatureresponsiveness. The dual light- and temperature-responsiveness of the DMA-MOAB:F127 blend is maintained when MB is added to the system, MB diffusion being dependent on both stimuli. Acknowledgment. This work was financed by the Ministerio de Educacio´n y Ciencia and FEDER (Grant SAF2005-01930) and the Xunta de Galicia (Grant PGIDIT06PXIC20303PN) of Spain and by the Singapore-MIT Alliance (SMA). We are grateful to P. Taboada for help with DLS experiments and to Janssen Pharmaceutica for providing free samples of hydroxypropyl-β-cyclodextrin. LA7019654