Incorporation of Block Copolymer Micelles into Multilayer Films for Use

Oct 28, 2008 - Gamble Technical Centre, Longbenton, Newcastle upon Tyne, NE12 9TS, United Kingdom. ReceiVed July 25, 2008. ReVised Manuscript ...
1 downloads 0 Views 423KB Size
13328

Langmuir 2008, 24, 13328-13333

Incorporation of Block Copolymer Micelles into Multilayer Films for Use as Nanodelivery Systems Timothy Addison,*,† Olivier J. Cayre,† Simon Biggs,† Steven P. Armes,‡ and David York§ The Institute of Particle Science and Engineering, UniVersity of Leeds, Leeds, LS2 9JT, United Kingdom, Department of Chemistry, UniVersity of Sheffield, Sheffield, S3 7HF, United Kingdom, and Procter & Gamble Technical Centre, Longbenton, Newcastle upon Tyne, NE12 9TS, United Kingdom ReceiVed July 25, 2008. ReVised Manuscript ReceiVed September 10, 2008 This work demonstrates the potential application of stimulus responsive block copolymer micelles as triggerable delivery systems for use within multilayer films. Cationic, pH-responsive micelles of poly[2-(dimethylamino)ethyl methacrylate-block-poly(2-(diethylamino)ethyl methacrylate)] (PDMA-PDEA) were deposited on anionic polystyrene latex particles. The charge reversal of the surface and the amount of adsorbed polymer were monitored by zeta potential measurements and colloidal titrations, respectively. Prior to adsorption, the PDMA-PDEA micelles were loaded with a hydrophobic dye, and UV-vis spectroscopy was used to determine the amount of dye encapsulated within a monolayer of micelles. It was found that subtle chemical modification of the PDMA-PDEA diblock copolymer via permanent quaternization of the PDEA block results in micelles with tunable loading capacities. Multilayers of cationic micelles of partially quaternized PDMA-PDEA and anionic polyelectrolyte (poly(sodium 4-styrene sulfonate)) were deposited on the surface of polystyrene latex particles by sequential adsorption. UV-vis analysis of the dye present within the multilayer after the addition of each layer demonstrates that the micelles are sufficiently robust to retain encapsulated dye after multiple adsorption/washing cycles and can thus create a film that can be increasingly loaded with dye as more micelle layers are adsorbed. Multiple washing cycles were performed on micellar monolayers of PDMA-PDEA to demonstrate how such systems can be used to bring about triggerable release of actives. When performing several consecutive washing steps at pH 9.3, the micelle structure of the PDMA-PDEA micelles in the monolayer is retained, resulting in only a small reduction in the amount of encapsulated dye. In contrast, washing at pH 4, the structure of the micelle layers is severely disrupted, resulting in a fast release of the encapsulated dye into the bulk. Finally, if a sufficient number of micelle/homopolyelectrolyte layers are adsorbed, it is possible to selectively dissolve the latex template, resulting in hollow capsules.

In recent years, the layer-by-layer (LbL) technique has been employed as a robust method for depositing polyelectrolyte multilayer films onto a number of substrates.1 This technique has been used to develop highly functional coatings with a broad range of potential applications. Initially demonstrated on planar surfaces,2 the deposition process has also been carried out using colloidal substrates,3 notably on nanoparticles4 and emulsion droplets.5 Hollow capsules have been created by selective dissolution of a sacrificial colloidal template, leading to further interest in this area.6 Charged particles can be used instead of one (or both) of the polyelectrolyte layers. As a result, the number of possible applications for this technology has greatly increased. For example, enzymes and vesicles can be incorporated into these films, suggesting possible uses as biosensors7 and microreactors.8 The incorporation of “soft” particles such as block copolymer * To whom correspondence should be addressed. E-mail: pre1twa@ leeds.ac.uk. Telephone: +44 113 343 2392. Fax: +44 113 343 2781. † University of Leeds. ‡ University of Sheffield. § Procter and Gamble.

(1) Caruso, F. AdV. Mater. 2001, 13(1), 11–22. (2) Decher, G. Science 1997, 277(5330), 1232–1237. (3) Radtchenko, I. L.; Sukhorukov, G. B.; Leporatti, S.; Khomutov, G. B.; Donath, E.; Mo¨hwald, H. J. Colloid Interface Sci. 2000, 230(2), 272–280. (4) Caruso, F.; Mo¨hwald, H. Langmuir 1999, 15(23), 8276–8281. (5) Lu, G.; An, Z.; Li, J. Biochem. Biophys. Res. Commun. 2004, 315(1), 224–227. (6) Donath, E.; Sukhorukov, G. B.; Caruso, F.; Davis, S. A.; Mo¨hwald, H. Angew. Chem., Int. Ed. 1998, 37(16), 2202–2205. (7) Constantine, C. A.; Gattas-Asfura, K. M.; Mello, S. V.; Crespo, G.; Rastogi, V.; Cheng, T.-C.; DeFrank, J. J.; Leblanc, R. M. Langmuir 2003, 19(23), 9863– 9867. (8) Ghan, R.; Shutava, T.; Patel, A.; John, V.; Lvov, Y. Mater. Res. Soc. Symp. Proc. 2004, 782(Micro- and Nanosystems), 241–246.

micelles within LbL films has also been demonstrated. These micelles can encapsulate hydrophobic materials such as quantum dots,7 nanoparticles,9 and organic actives10 within their cores, which can lead to additional functionality. The use of stimulusresponsive block copolymers that can reversibly aggregate/ dissociate in response to an external trigger such as pH or temperature may provide new opportunities for triggered release of the encapsulated material.11 Recently, we prepared LbL films composed entirely of oppositely charged diblock copolymer micelles onto a colloidal silica template and showed that such films can be loaded with hydrophobic dyes.12 In the present study, we extend this work by producing LbL films composed of diblock copolymer micelles and homopolyelectrolytes. Moreover, we provide a more detailed investigation of the uptake, retention, and release of hydrophobic dye molecules by these multilayer films. Previously, it has been shown that micelles adsorbed at a planar interface can solubilize dyes. This was demonstrated initially using anionic diblock copolymer micelles combined with a cationic homopolyelectrolyte,13 and subsequently using oppositely charged micelles.14 More recently, Hammond and co-workers have demonstrated (9) Kim, B.-S.; Taton, T. A. Langmuir 2007, 23(4), 2198–2202. (10) Soo, P. L.; Luo, L.; Maysinger, D.; Eisenberg, A. Langmuir 2002, 18(25), 9996–10004. (11) Webber, G. B.; Sakai, K.; Wanless, E. J.; Armes, S. P.; Vamvakaki, M.; Bu¨tu¨n, V.; Biggs, S. In Preparation, Characterisation and Utilisation of Diblock Copolymer Micelle Thin-Films and Multilayers; 7th World Congress of Chemical Engineering, Glasgow, United Kingdom, 2005. (12) Biggs, S.; Sakai, K.; Addison, T.; Schmid, A.; Armes, S. P.; Vamvakaki, M.; Bu¨tu¨n, V.; Webber, G. AdV. Mater. 2007, 19(2), 247–250. (13) Ma, N.; Zhang, H.; Song, B.; Wang, Z.; Zhang, X. Chem. Mater. 2005, 17(20), 5065–5069. (14) Qi, B.; Tong, X.; Zhao, Y. Macromolecules 2006, 39(17), 5714–5719.

10.1021/la802396g CCC: $40.75  2008 American Chemical Society Published on Web 10/28/2008

Stimulus-ResponsiVe Micellar Films

slow, sustainable release of model hydrophobic drugs from surface-adsorbed micelles.15 LbL films allow various strategies to be exploited for delivery applications, as summarized by Hammond and co-workers.15 First, microcapsules can be fabricated and loaded with watersoluble actives, with the rate of diffusion across the LbL membrane providing the release profile.16 Another approach is to directly incorporate the active into the LbL film, with chemical degradation of the membrane over time providing the required release.17 These two methods generally provide relatively slow release rates but are not suitable for encapsulating hydrophobic substances. An approach using porous multilayers to take up hydrophobic actives via the use of an organic solvent with subsequent activated release in aqueous solutions has been demonstrated.18 However, this strategy is not always straightforward, as poor interaction and wetting of the multilayer by the organic solvent can cause problems with uptake/release of the actives. By using block copolymer micelles, we demonstrate a more robust method for encapsulating hydrophobic actives into an LbL-assembled film or capsule. As far as we are aware, this report is the first study demonstrating the uptake/triggered release of hydrophobic molecules from block copolymer micelles adsorbed onto colloidal templates. Block copolymer micelles are of considerable interest as the chemistry and properties of both the core and coronal blocks can be tailored to provide highly versatile and tuneable systems. Judicious monomer selection can also allow the design of stimulus-responsive micelles. Such systems may have potential applications in drug delivery,19 agrochemical formulations,20 and personal care products.21 Furthermore, it has been shown that water-soluble block copolymer micelles can provide a suitable local environment for the loading of hydrophobic actives such as dyes and drugs. Such copolymers can undergo reversible micellization in response to changes in pH,22 temperature,23 or ionic strength.24 This reversible micellar self-assembly in aqueous media provides a flexible route for the design of materials that exhibit interesting uptake/release behavior. Typically, block copolymers only form micelles above a certain critical micelle concentration (cmc). If the copolymer concentration drops below the cmc on dilution, this can result in micelle dissociation and premature release of the active(s). This problem can be mitigated by surface immobilization of the micelles. We have shown that surface-adsorbed pH-responsive diblock copolymer micelles can resist dissociation.25 In this particular case, micelles are only formed in bulk solution above pH 8. At low pH, where the micelles would normally dissociate in bulk solution, the structure of the surface-adsorbed micelles is disrupted, but relatively little desorption occurs. Upon replacing the bulk solution (15) Nguyen, P. M.; Zacharia, N. S.; Verploegen, E.; Hammond, P. T. Chem. Mater. 2007, 19(23), 5524–5530. (16) Antipov, A. A.; Sukhorukov, G. B. AdV. Colloid Interface Sci. 2004, 111(1-2), 49–61. (17) Vazquez, E.; Dewitt, D. M.; Hammond, P. T.; Lynn, D. M. J. Am. Chem. Soc. 2002, 124(47), 13992–13993. (18) Thierry, B.; Kujawa, P.; Tkaczyk, C.; Winnik, F. M.; Bilodeau, L.; Tabrizian, M. J. Am. Chem. Soc. 2005, 127(6), 1626–1627. (19) Lavasanifar, A.; Samuel, J.; Sattari, S.; Kwon, G. S. Pharm. Res. 2002, 19(4), 418–422. (20) Manteca, Z. I.; Chrisstoffels, L.; Berghaus, R.; Stierl, R. Preparation of block copolymer adjuvants for agrochemical formulations. International Patent WO 2005018321, 2005. (21) Rodriguez-Hernandez, J.; Checot, F.; Gnanou, Y.; Lecommandoux, S. Prog. Polym. Sci. 2005, 30(7), 691–724. (22) Webber, G. B.; Wanless, E. J.; Armes, S. P.; Tang, Y.; Li, Y.; Biggs, S. AdV. Mater. 2004, 16(20), 1794–1798. (23) Ishida, N.; Biggs, S. Langmuir 2007, 23(22), 11083–11088. (24) Ishida, N.; Biggs, S. Macromolecules 2007, 40(25), 9045–9052. (25) Webber, G. B.; Wanless, E. J.; Armes, S. P.; Biggs, S. Faraday Discussions 2005, 128, 193–209.

Langmuir, Vol. 24, No. 23, 2008 13329

with an aqueous alkaline solution (thus providing suitable conditions for micelle formation), partial reformation of the original micelle morphology is observed. This behavior has been likened to an “open-close” mechanism and confirms the potential of such micelle monolayers for targeted release of actives. LbL films comprising oppositely charged block copolymer micelles can be self-assembled onto a planar substrate.11 Atomic force microscopy (AFM) studies within a LbL film indicate that both types of diblock copolymer micelles retain their core-shell structures after polyelectrolyte complexation.26 This leads to the possibility of highly flexible films that can be increasingly loaded with actives as additional layers are added. By using block copolymer micelles that respond to pH, release of actives can be induced by pH cycling. Furthermore, specific layers can be selectively loaded, creating a much slower, diffusion-based release profile. Working with a colloidal silica template, we have constructed LbL films composed solely of block copolymer micelles.12 As with planar templates, these films are sufficiently robust to allow encapsulation of dyes within the hydrophobic micelle cores. As the number of dye-encapsulated micelle layers was increased, stronger absorption due to the presence of the encapsulated dye was observed. This earlier work suggests that hollow, stimulus-responsive capsules loaded with actives can be constructed from binary mixtures of oppositely charged micelles. However, a better understanding of the behavior of dye-loaded micelles within LbL films is desirable. For example, it is likely that dye exchange can occur between adsorbed micelles and those in bulk solution. There is also the possibility of dye exchange between micelles within adjacent layers, especially as such layers are known to interpenetrate.26 Combining two or more different types of micelles within an LbL film may offer the potential for more sophisticated delivery systems, as individual micelle layers can respond to different external stimuli. On the other hand, in some cases it may be preferable to encapsulate actives within only one type of micelle or an individual micelle layer within the film. This could provide a greater degree of control over release profiles. Because of the inherent complexity of binary micelle multilayers, it is instructive to examine LbL films composed of a single type of diblock copolymer micelle loaded with a specific hydrophobic dye, paired with a simple, oppositely charged, homopolyelectrolyte. In this study, LbL films are prepared using cationic diblock copolymer micelles (comprising poly[2-(dimethylamino)ethyl methacrylate-block-poly(2-(diethylamino)ethyl methacrylate)] (PDMA-PDEA)) combined with an anionic polyelectrolyte (poly(sodium 4-styrene sulfonate) (NaPSS)) and an anionic polystyrene latex template. The weakly basic PDMA-PDEA copolymer is molecularly dissolved at low pH, but forms welldefined PDEA-core micelles in alkaline solution.27 The unimerto-micelle transition occurs over a relatively narrow range (pH 7-9), which depends on both the molecular weight and block composition of the copolymer. Bu¨tu¨n et al. showed that the PDMA blocks of these copolymers can be selectively quaternized by reaction with alkyl halides to produce permanently cationic blocks. This derivatization ensures that the coronal PDMA blocks retain their cationic character and remain colloidally stable in alkaline media. The copolymer precursor used in this study was PDMA97PDEA24, with the subscripts referring to the mean degree of polymerization for each block. The PDMA block of this copolymer was reacted with methyl iodide to afford differing (26) Smith, E. G.; Webber, G. B.; Sakai, K.; Biggs, S.; Armes, S. P.; Wanless, E. J. J. Phys. Chem. B 2007, 111(20), 5536–5541. (27) Lee, A. S.; Gast, A. P.; Bu¨tu¨n, V.; Armes, S. P. Macromolecules 1999, 32(13), 4302–4310.

13330 Langmuir, Vol. 24, No. 23, 2008

Addison et al.

Figure 1. Zeta potential vs pH curves obtained for (a) bare 1 µm polystyrene latex particles (triangles), (b) latex particles coated with one layer of 50qPDMA-PDEA (squares), and (c) latex particles coated with alternating layers of 50qPDMA-PDEA and NaPSS (circles).

Figure 2. Zeta potential as a function of layer number for 50qPDMAPDEA/NaPSS-coated PS latex particles prepared using the LbL technique.

Table 1. Effect of Varying the Mean Degree of Quaternization of the PDMA Block on the Adsorbed Amounts (mg m-2) and Zeta Potentials (mV) for Cationic PDMA-PDEA Micelle Monolayers Adsorbed onto Anionic 1 µm Polystyrene Latex Particles at pH 9.3 degree of quaternization of PDMA block (%)

adsorbed amount of polymer (mg m-2)

zeta potential (mV)

0 10 50 100

4.30 ( 0.1 3.86 ( 0.1 3.62 ( 0.1 2.97 ( 0.1

5 18 29 33

mean degrees of quaternization (10%, 50%, and 100%) giving rise to a range of diblock precursors that are denoted as 10qPDMAPDEA, 50qPDMA-PDEA, and 100qPDMA-PDEA, respectively. The nonquaternized precursor is denoted as 0qPDMA-PDEA.28 The critical micellization pH for PDMA-PDEA ranges from pH to 7.0 to pH 8.8,25 depending on the degree of quaternization of the DMA residues and also the block copolymer composition. Our experiments were conducted at a pH of approximately 9.3. This allows for some degree of pH drift, while still maintaining a solution pH well above 8.8. First, the zeta potential vs pH curve for 1 µm polystyrene latex was determined (see Figure 1). This latex is strongly anionic in alkaline solution and hence acts as a suitable colloidal template for the adsorption of an initial layer of cationic PDMA-PDEA micelles. Full details of the deposition process can be found in the experimental section. It is important to note that the micelle-coated latex was purified by a centrifugation/redispersion cycle, with the supernatant containing excess micelles and noncomplexed dye being removed. Zeta potential studies indicated surface charge reversal, which confirmed successful micelle adsorption (see Table 1). As the mean degree of quaternization of the PDMA block is increased, the surface charge of the coated particles becomes more cationic, as expected. Colloidal titrations were used to analyze the copolymer concentrations of the supernatant solutions and thus determine the amount of copolymer adsorbed at the surface of the latex (see Table 1). It is clear that lower adsorbed amounts are obtained at higher degrees of quaternization. This is consistent with the adsorption behavior of PDMA-PDEA on planar surfaces.25 Zeta potentials and acid titrations were conducted on latex particles coated with a monolayer of 50qPDMA-PDEA and a bilayer of 50qPDMA-PDEA/NaPSS to confirm the electrostatic adsorption of the layers. These titration data provide more insight regarding the nature of the particle surface than single measurements because charge reversal does not fully confirm a successful coating. For example, adsorption of the (28) Bu¨tu¨n, V.; Armes, S. P.; Billingham, N. C. Macromolecules 2001, 34(5), 1148–1159.

Figure 3. UV-vis spectra recorded for chrysoidin dye-loaded PDMAPDEA micelle monolayers adsorbed onto 1 µm polystyrene latex particles at pH 9.3: 0q (filled squares), 10q (open circles), 50q (filled triangles), 100q (open squares), reference (line). The inset shows a digital photograph of the recovered dye from the monolayers of PDMA-PDEA: (from left to right) 0q, 10q, 50q, 100q, and a reference.

second polymer layer, which in this case is the homopolyelectroyte NaPSS may facilitate desorption of the first layer of surface adsorbed PDMA-PDEA micelles, resulting in an unsuccessful coating but an apparent charge reversal of the particle. Figure 1 gives pH titration data for latex coated with a single layer of 50qPDMA-PDEA and latex coated with single layers of 50qPDMA-PDEA and NaPSS. When the outer layer is 50qPDMA-PDEA, the zeta potential is still reasonably cationic even at basic pH; this is due to the permanent cationic charge on the quaternized DMA residues. This cationic charge increases at low pH due to protonation of the nonquaternized DMA residues. When the outer layer of the coated latex is NaPSS, the latex has a strong anionic charge that is largely independent of pH. To demonstrate the deposition of multiple 50qPDMA-PDEA/ NaPSS layers onto the latex, zeta potentials were determined after each deposition of either a homopolyelectrolyte or micelle layer. Inspecting Figure 2, the surface charge is reversed after each successful coating cycle, as expected. The following experiment was designed to allow the amount of material that can be encapsulated within the hydrophobic cores of the adsorbed PDMA-PDEA micelles to be determined. Chrysoidin was chosen as a suitable dye for this experiment because of its moderate solubility in water, which allows for a high loading of the block copolymer micelles. Its high affinity for tetrahydrofuran (THF) provides a convenient means of directly monitoring the dye present within the LbL film. Figure 3 shows the absorption spectra of THF solutions in which dye-loaded PDMA-PDEA monolayers were dissolved. The absorbance, and thus the amount of dye adsorbed within the micelle monolayer, is reduced as the mean degree of quaternization of the copolymer is systematically increased. This behavior can be partly attributed to the lower

Stimulus-ResponsiVe Micellar Films

Langmuir, Vol. 24, No. 23, 2008 13331

Table 2. Effect of Mean Degree of Quaternization of the PDMA Block on the Amount of Chrysoidin Dye within PDMA-PDEA Micelle Monolayers Adsorbed onto 1 µm Polystyrene Latex at pH 9.3 mean degree of quaternization of PDMA block (%)

adsorbed amount of chrysoidin dye (mg m-2)

mass of dye (kg) per micelle × 10-21

0 10 50 100

0.080 0.051 0.037 0.025

1.2 0.275 0.125 0.086

absorbance observed for higher levels of quaternization, but will also be influenced by the chemical nature of the micelles. Summarized within the Supporting Information is a table showing previously published data obtained for PDMA-PDEA micelles. Clearly, nonquaternized micelles have significantly larger hydrodynamic diameters and aggregation numbers, which in turn indicate significantly larger hydrophobic cores and higher loading capacities. Furthermore, the proposed model for PDMA-PDEA adsorption25 suggests that increasing the coronal charge density of the micelles results in a more flattened morphology on surface adsorption. Although a dye-loaded micelle may possibly exhibit greater structural stability, it is reasonable to assume that it will still undergo some degree of flattening during interfacial adsorption. It is likely that this deformation will reduce the size of its hydrophobic core, which will result in a reduced dye capacity. Comparing these absorption spectra with a calibration curve constructed for chrysoidin in THF, the adsorbed amount of dye per unit area of latex can be calculated. By combining this result with the copolymer adsorption data (see Table 1), the amount of encapsulated dye present per surface-adsorbed micelle can be estimated (see Table 2). The hydrodynamic diameters of the PDMA-PDEA micelles given as Supporting Information suggest that unquaternized micelles have a theoretical volume close to an order of magnitude greater than that of fully quaternized micelles. These data qualitatively agree with the difference in loading capacities of the micelles calculated from the amount of dye incorporated within the micelle monolayers on the particle surface, as shown in Table 2. Multilayer films were readily prepared by sequential adsorption of six layers of cationic 50qPDMA-PDEA micelles and anionic NaPSS. Chrysoidin dye was present in the micelle solutions but not in the NaPSS solutions. Samples were removed after addition of each layer for analysis. The resulting UV-vis spectra from these samples have been plotted in Figure 4, which shows a large increase in the amount of dye present within the film after adsorption of each 50qPDMA-PDEA micelle layer followed by

Figure 4. UV-vis spectra recorded for LbL multilayer films of 50qPDMA-PDEA/PSSS loaded with chrysoidin dye: one layer (+), two layers (open triangles), three layers (filled triangles), four layers (open squares), five layers (filled squares), six layers (open circles).

Figure 5. UV-vis spectra illustrating the effect of multiple washing cycles (pH 9.3) on the amount of chrysoidin dye present within a monolayer of 50qPDMA-PDEA micelles adsorbed onto 1 µm polystyrene latex: unwashed (filled squares), one wash (open circles), two washes (filled triangles), three washes (open squares), four washes (open triangles).

Figure 6. Reduction in the adsorbed amount of 50qPDMA-PDEA diblock copolymer micelles due to desorption after multiple washing cycles performed at pH 4 (open circles) and pH 9.3 (filled squares).

a significant reduction in dye content after adsorption of a NaPSS layer. Loss of dye may be due to two mechanisms. First, upon exposure to the NaPSS solution, adsorbed micelles could be removed from the surface of the latex. Alternatively, dye may be gradually lost from the micelle cores due to its partition into the bulk solution, with the micelle layer remaining intact during NaPSS adsorption. To further investigate these observations, an experiment was conducted whereby the amount of dye present within a 50qPDMA-PDEA monolayer was monitored by UVvis spectroscopy during multiple wash cycles. To retain the micellar morphology of the PDMA-PDEA layer, 10 mM KNO3 solution at pH 9.3 was used to perform each wash cycle. The resulting spectra (see Figure 5) indicate that some fraction of the dye partitions into the bulk solution after each wash cycle, which is consistent with the results shown in Figure 4. To examine the possibility of in situ micelle desorption, colloidal titrations were used to detect residual copolymer present within the wash solutions. These data are plotted in Figure 6 and confirm that, after an initial rinse off, only a very small amount of copolymer desorption occurs from the latex surface. The results shown in Figure 5 and 6 provide good evidence that the loss of dye from the PDMA-PDEA micelles within the LbL coating is largely due to partitioning of chrysoidin between the micelles and bulk solution, with minimal micelle desorption. Despite such dye partitioning, Figure 4 suggests that it is still possible to obtain an LbL film with an increasing dye loading as more micelle layers are added.

13332 Langmuir, Vol. 24, No. 23, 2008

Figure 7. UV-vis spectra illustrating the affect of multiple washing cycles (pH 4) on the amount of chrysoidin dye present within a 50qPDMA-PDEA micelle monolayer adsorbed onto 1 µm polystyrene latex: unwashed (filled squares), one wash (open circles), two washes (filled triangles).

As discussed previously, surface-adsorbed PDMA-PDEA copolymer micelles can undergo a partially reversible change in morphology upon switching the external solution pH. The suggested mechanism25 involves a transition from a flattened micelle structure to a polymer brush on cycling between alkaline and acidic solution. In principle, such morphological changes can be employed to trigger the release of hydrophobic actives from the micelle cores. Figure 7 shows the amount of dye present within an adsorbed monolayer of 50qPDMA-PDEA copolymer micelles after rinsing with KNO3 at pH 4. Previous data obtained on planar surfaces29 indicates that this loss of dye may be partially attributed to desorption of the 50qPDMA-DEA copolymer due to pH cycling. However, given that a modest amount of dye is still observed after two washes, it suggests that the original micelle structure of the monolayer is retained, at least to some degree. The solubility of chrysoidin in water was found to be pH dependent, with a higher solubility recorded at pH 4 as opposed to weakly basic conditions. In both weakly acidic and weakly basic conditions, however, the solubility of the dye is sufficiently high to suggest a very large driving force toward solubilization of the dye within the bulk. The increased capacity of the micelle monolayer to retain the dye with multiple washes at pH 9.3 therefore indicates that structural changes of the micelles play a significant role in the increased release of the dye at pH 4. These data serve to illustrate the potential of surface-adsorbed micelles for the encapsulation of hydrophobic actives. By varying the mean degree of quaternization of the DMA residues, it is possible to produce copolymer micelles with variable dye capacities. Furthermore, the responsive nature of the micelle cores provide a means to induce the release of encapsulated material by simply switching the solution pH. Given a sufficient number of adsorbed PDMA-PDEA/NaPSS layers, the latex template can be subsequently dissolved to produce hollow capsules. In terms of its low polydispersity, high surface charge and minimal surface roughness, polystyrene latex is a good choice of template. However, its dissolution requires the use of THF, which in turn results in destruction of the core-shell structure of the micelles. Alternative templates that can be dissolved at high pH are currently being evaluated so that the micelles retain their morphology and pH-responsive character within the hollow capsule membrane. The AFM and scanning electron microscopy (SEM) images shown in Figure 8 suggest that PDMA-PDEA micelles can be used to form multilayer (29) Sakai, K.; Smith, E. G.; Webber, G. B.; Baker, M.; Wanless, E. J.; Buetuen, V.; Armes, S. P.; Biggs, S. Langmuir 2006, 22(20), 8435–8442.

Addison et al.

capsules.16 Such LbL membranes seem to be sufficiently robust to survive dissolution of the template cores. In summary, we have demonstrated that a binary mixture of an anionic polyelectrolyte (NaPSS), and cationic PDMA-PDEA micelles can be used to construct LbL films, with the micelles retaining their structure after a number of adsorption cycles. Adjusting the degree of quaternization of the PDMA-PDEA micelles allows the amount of dye contained within each layer to be controlled and, although a degree of partitioning takes place between the micelles and bulk solution, the dye uptake of these LbL films increases as more layers are deposited. The stimulus-responsive nature of complexed PDMA-PDEA micelles is also demonstrated, since an external acidic environment leads to faster dye release. However, the chrysoidin dye used in these experiments is moderately soluble in water and will therefore be less inclined to remain within the micelle cores upon exchange of the bulk phase. By adjusting the pH of the bulk solution and the extent of micelle loading, it is possible to tune both the release profiles and the final dye “payload”. Although not shown in this study, employing dyes with different hydrophobicities should allow further control over the release profiles. In addition, we are not limited to the use of one type of micelle, which may provide further opportunities to tailor the coating and adsorb binary (or more complex) mixtures of actives within the cores. Finally, we have demonstrated that, if a sufficient number of layers are added, the polystyrene latex template can be selectively dissolved using THF, resulting in a hollow capsule. Unfortunately, dissolution of the latex cores results in the loss of micelle structure within the membrane. However, according to Antipov et al.,30 it should be possible to produce such capsules using calcium carbonate particles as sacrificial templates. This should allow core removal using a dilute aqueous EDTA solution at a solution pH that is well above the critical micellization pH, thus preserving the micelle structure.

Experimental Section Materials. The sources of chemicals were as follows: KOH (reagent grade, g90%), HNO3 (70% ACS reagent), poly(sodium 4-styrenesulfonate) (Mn ) 70 000), chrysoidin G C.I. 11270 (Basic Orange 2)-Aldrich; KNO3 (ReagentPlus, g99.0%), THF (>99%)Sigma; poly(potassium vinyl sulfate) (1/400 N), toluidine blue indicator solution-Wako Pure Chemicals; and 1.00 µm polystyrene latex (specific surface area ) 5.71 m2/g)-Bangs Laboratories, Inc. All chemicals were used as received with no further purification except for the latex, which was washed twice with pH-adjusted electrolyte solution (pH 9, 10 mM KNO3) The water used in all experiments was prepared from a Millipore “Milli-Q Ultrapure Water Purification System” with a resistivity of 18.2 MΩ cm-1 (25 °C), and a total organic carbon concentration (TOC) of less than 5 ppb. Copolymer Synthesis and Characterization. The PDMA-PDEA diblock copolymer was synthesized using group transfer polymerization. Full details of this synthesis may be found elsewhere.31 The number-average molecular weight (Mn) and polydispersity index (PDI) of this copolymer were determined to be 17 700 g mol-1 and 1.08, respectively, as measured by gel permeation chromatography with a THF eluent at a flow rate of 1.0 mL min-1 using poly(methyl methacrylate) calibration standards. On the basis of the 1H NMR spectra, the mean degrees of polymerization were calculated to be 97 for the PDMA block and 24 for the PDEA block. This precursor copolymer was then used to make further copolymers containing quaternized tertiary amine residues on the PDMA chains. This was achieved using a (sub)stoichometric amount of MeI, as described (30) Antipov, A. A.; Shchukin, D.; Fedutik, Y.; Petrov, A. I.; Sukhorukov, G. B.; Mo¨hwald, H. Colloids Surf., A: Physicochem. Eng. Aspects 2003, 224(1-3), 175–183. (31) Bu¨tu¨n, V.; Armes, S. P.; Billingham, N. C. Polymer 2001, 42(14), 5993– 6008.

Stimulus-ResponsiVe Micellar Films

Langmuir, Vol. 24, No. 23, 2008 13333

Figure 8. AFM (left), and SEM (right) images of the micrometer-sized 50qPDMA-PDEA/NaPSS capsules obtained after THF dissolution of the polystyrene latex cores.

previously.28 The mean degrees of quaternization of the PDMA blocks were assessed using 1H NMR and found to be 10, 50, and 100 mol %, respectively. Preparation of Chrysoidin-Doped Micelle Solution. The chrysoidin-doped PDMA-PDEA micelle solutions were prepared using degassed 10 mM KNO3 solution, which was adjusted to pH 9.3 using dilute aqueous KOH. Excess chrysoidin was added to the copolymer micelle solution, which was given 24 h to equilibrate under constant stirring. The resulting stock solution was filtered using a 0.2 µm disposable membrane filter prior to making up aqueous PDMA-PDEA solutions. The pH of each copolymer solution was adjusted and left for an additional 12 h to equilibrate prior to measurements. Preparation of Micelle/NaPSS-Coated Polystyrene Latex. Copolymer micelles were deposited onto PS latex templates using the LbL technique. An aqueous solution of PDMA-PDEA (10 mL, 1000 ppm) at pH 9.2-9.3 was added to an aqueous latex dispersion at the same pH (10 mL, 1500 ppm) and stirred for 30 min to allow for full coverage of the particles to take place. The sample was then centrifuged at 2000g for 20 min, and the supernatant was removed and discarded. If washing of the coated particles was required, it was performed by redispersion in 20 mL of pH-adjusted 10 mM KNO3 solution (pH 9.3) followed by further centrifugation and supernatant removal. After the formation of the first layer, the zeta potential was determined to confirm that surface charge reversal had indeed occurred. An aqueous solution of NaPSS (10 mL, 1000 ppm) at pH 9.2-9.3 was added to the redispersed latex particles to develop the second layer, with the addition of small volumes of dilute KOH and KNO3 to maintain the pH above 9.0. The separation/replacement/ redispersion protocol was repeated in order to coat particles with multiple micelle/polyelectrolyte layers. Formation of Multilayer Capsules. In order to create hollow capsules, the latex particles were coated with six PDMA-PDEA/ NaPSS pairs. The coated particles were then washed three times with pH-adjusted deionized water at pH 9.3 to remove the excess (co)polymer. Dissolution of the latex cores was achieved by addition of these particles (10 mg) to a small volume of THF (5 mL). After allowing 5 min for dissolution, the hollow capsules were centrifuged gently (500 g, 20 min). The supernatant was removed, and the capsules were redispersed in fresh THF. An additional two washing cycles were performed, with the capsules eventually being redispersed in pH-adjusted deionized water at pH 9.3. UV-vis Spectroscopy. All spectra were recorded using an Agilent 8453E UV-vis spectrophotometer. The coated latexes were dissolved in THF and left for 24 h to allow the chrysoidin dye to completely

leach out of the micelles into the THF solution. High-speed centrifugation (8000 g for 20 min) enabled analysis of the supernatants while avoiding unwanted scattering by residual polymer or capsules. All spectra were referenced against a background of pure THF. To calculate the (unknown) dye concentration, a comparison was made between the absorption observed at 421 nm and a calibration curve. Colloidal Titration. The copolymer concentration in the supernatant was determined by titration against 1/400 M poly(potassium vinyl sulfate) aqueous solution in the presence of 1 mL of 0.1 M HNO3, and two drops of toluidine blue indicator solution. At the end point, the color of the solution changes dramatically from blue to light pink with precipitation being observed. The unknown copolymer concentration of the supernatant solution, and thus the extent of copolymer adsorption, can then be determined by comparing the absorbance of the unknown solution to that of a calibration curve. Zeta Potential Measurements. A Malvern NanoSeries Zetasizer Nano-ZS (Malvern Instruments) equipped with a He-Ne laser source (wavelength 633 nm, power 4.0 mW) was used for measuring the zeta potential of the latex before and after micelle-coating. Autotitration involved addition of small volumes of dilute KOH and HNO3 solutions to give zeta potential versus pH curves. All measurements were performed in 10 mM KNO3 electrolyte. A typical data point represents the average of at least 100 individual measurements. Scanning Electron Microscopy. SEM images were obtained using a LEO 1530 Field Emission Gun SEM instrument operating at 3 kV. Atomic Force Microscopy. AFM samples were prepared by pipetting a drop of water-dispersed capsules onto a freshly cleaved mica surface. The instrument used was a Veeco Bioscope II, and all images were collected in air using tapping mode. A POINTPROBE Silicon SPM Probe was used, with a typical resonance frequency of 320 kHz and a force constant of 42 N m-1.

Acknowledgment. The EPSRC and Procter and Gamble are thanked for providing T.A. with an Industrial Case Award. Dr C. D. Vo is thanked for the quaternization of the copolymer sample. Supporting Information Available: Published data summarizing the hydrodynamic diameters, zeta potentials, and aggregation numbers of PDMA-PDEA micelles. This material is available free of charge via the Internet at http://pubs.acs.org. LA802396G