Film-Forming Microgels for pH-Triggered Capture and Release

Aug 19, 2008 - 4, B1904DPI, La Plata, Argentina and Grupo Materiales Poliméricos-LIMF, ... UniVersidad Nacional de La Plata, Calle 116 y 48, B1900TAG...
0 downloads 0 Views 1MB Size
Download PDF
10228

Langmuir 2008, 24, 10228-10234

Film-Forming Microgels for pH-Triggered Capture and Release Paul A. FitzGerald,†,| Javier I. Amalvy,‡ Steven P. Armes,§ and Erica J. Wanless*,† School of EnVironmental and Life Sciences, UniVersity of Newcastle, Callaghan, New South Wales, 2308, Australia, Grupo Materiales Polime´ricos, Instituto de InVestigaciones Fisicoquı´micas Teo´ricas y Aplicadas (INIFTA), (UNLP - CCT La Plata-CONICET), UniVersidad Nacional de La Plata, Diag. 113 y 64 CC 16 Suc. 4, B1904DPI, La Plata, Argentina and Grupo Materiales Polime´ricos-LIMF, Área Departamental Meca´nica, Facultad de Ingenierı´a, UniVersidad Nacional de La Plata, Calle 116 y 48, B1900TAG, La Plata, Argentina, and Department of Chemistry, Dainton Building, UniVersity of Sheffield, Brook Hill, Sheffield S3 7HF, U.K. ReceiVed June 5, 2008. ReVised Manuscript ReceiVed July 16, 2008 The pH-responsive behavior for a series of lightly cross-linked, sterically stabilized poly(tertiary amine methacrylate)based latexes adsorbed onto mica and silica was investigated using in situ tapping mode AFM at room temperature. The adsorbed layer structure was primarily determined by the glass transition temperature, Tg, of the latex: poly[2-(diethylamino)ethyl methacrylate]-based particles coalesced to form relatively featureless uniform thin films, whereas the higher Tg poly[2-(diisopropylamino)ethyl methacrylate] latexes retained their original particulate character. Adsorption was enhanced by using a cationic poly[2-(dimethylamino)ethyl methacrylate] steric stabilizer, rather than a nonionic poly(ethylene glycol)-based stabilizer, since the former led to stronger electrostatic binding to the oppositely charged substrate. Both types of adsorbed latexes acquired cationic microgel character and swelled appreciably at low pH, even those that had coalesced to form films. Fluorescence spectroscopy was used to study the capture of a model hydrophobic probe, pyrene, by these adsorbed latex layers followed by its subsequent release by lowering the solution pH. The repeated capture and release of pyrene through several pH cycles was also demonstrated. Since these poly(tertiary amine methacrylate) latexes are readily prepared by aqueous emulsion polymerization and adsorption occurs spontaneously from aqueous solution, this may constitute an attractive route for the surface modification of silica, mica and other oxides.

Introduction Microgels are lightly cross-linked polymer latex particles of colloidal dimensions that swell substantially when exposed to changes in external stimuli such as pH, temperature or salt.1-8 Temperature-sensitive microgels are typically based on poly(N-isopropylacrylamide) (PNIPAM).5,9-18 Various classes of pH* To whom correspondence should be addressed. E-mail: erica.wanless@ newcastle.edu.au. † University of Newcastle. ‡ Universidad Nacional de La Plata. § University of Sheffield. | Present address: School of Chemistry, University of Sydney, Sydney, NSW 2006, Australia.

(1) Ma, G. H.; Fukutomi, T. Macromolecules 1992, 25, 1870–1875. (2) Rodriguez, B. E.; Wolfe, M. S.; Fryd, M. Macromolecules 1994, 27, 6642– 6647. (3) Saunders, B. R.; Crowther, H. M.; Vincent, B. Macromolecules 1997, 30, 482–487. (4) Saunders, B. R.; Vincent, B. AdV. Colloid Interface Sci. 1999, 80, 1–25. (5) Pelton, R. AdV. Colloid Interface Sci. 2000, 85, 1–33. (6) Kuckling, D.; Vo, C. D.; Wohlrab, S. E. Langmuir 2002, 18, 4263–4269. (7) Hoare, T.; Pelton, R. Macromolecules 2004, 37, 2544–2550. (8) Amalvy, J. I.; Wanless, E. J.; Li, Y.; Michailidou, V.; Armes, S. P.; Duccini, Y. Langmuir 2004, 20, 8992–8999. (9) Bradley, M.; Ramos, J.; Vincent, B. Langmuir 2005, 21, 1209–1215. (10) Deng, Y.; Pelton, R. Macromolecules 1995, 28, 4617–4621. (11) Gan, D.; Lyon, L. A. J. Am. Chem. Soc. 2001, 123, 7511–7517. (12) Kawaguchi, H.; Fujimoto, K.; Mizuhara, Y. Colloid Polym. Sci. 1992, 270, 53–57. (13) Morris, G. E.; Vincent, B.; Snowden, M. J. J. Colloid Interface Sci. 1997, 190, 198–205. (14) Pelton, R. H.; Chibante, P. Colloids Surf. 1986, 20, 247–256. (15) Tam, K. C.; Wu, X. Y.; Pelton, R. H. J. Polym. Sci., Part A: Polym. Chem. 1993, 31, 963–969. (16) Wu, X.; Pelton, R. H.; Hamielec, A. E.; Woods, D. R.; McPhee, W. Colloid Polym. Sci. 1994, 272, 467–477. (17) Ho¨fl, S.; Zitzler, L.; Hellweg, T.; Herminghaus, S.; Mugele, F. Polymer 2007, 48, 245–254. (18) Schmidt, S.; Motschmann, H.; Hellweg, T.; von Klitzing, R. Polymer 2008, 49, 749–756.

responsive microgels based on methacrylic acid,2,3,19,20 4vinylpyridine, 2-vinylpyridine,1,21 2-(diethylamino)ethyl methacrylate or 2-(diisopropylamino)ethyl methacrylate have been reported.8 A number of applications for stimulus-responsive microgels have been suggested, such as viscosity modifiers, delivery vehicles, biosensors and ‘smart’ particulate emulsifiers.2,15,22-24 Recent research has focused on microgel activity at interfaces because of the growing potential for functional coatings in controlled uptake/release applications,25 micro-optics,26 stimulusresponsive membranes,27 ultralow friction surfaces,28 and drug delivery.29 For example, Serpe et al. reported multilayers of anionic PNIPAM-co-AA microgel and cationic poly(allylamine hydrochloride) (PAH) spin-coated films, which were used for the thermally controlled uptake and release of doxorubicin.25 Lipid bilayer-coated poly(NIPAM-co-N-vinylimidazole) microgels were synthesized by Kazakov et al. and then adsorbed onto mica to study their deformation using tapping mode atomic force (19) Tan, B. H.; Tam, K. C.; Lam, Y. C.; Tan, C. B. Langmuir 2005, 21, 4283–4290. (20) Tan, B. H.; Tam, K. C.; Lam, Y. C.; Tan, C. B. AdV. Colloid Interface Sci. 2005, 113, 111–120. (21) Dupin, D.; Fujii, S.; Armes, S. P.; Reeve, P.; Baxter, S. M. Langmuir 2006, 22, 3381–3387. (22) Dupin, D.; Armes, S. P.; Connan, C.; Reeve, P.; Baxter, S. M. Langmuir 2007, 23(13), 6903–6910. (23) Fujii, S.; Armes, S. P.; Binks, B. P.; Murakami, R. Langmuir 2006, 22(16), 6818–6825. (24) Tsuji, S.; Kawaguchi, H. Langmuir 2008, 24(7), 3300–3305. (25) Serpe, M. J.; Yarmey, K. A.; Nolan, C. M.; Lyon, L. A. Biomacromolecules 2005, 6, 408–413. (26) Serpe, M. J.; Kim, J.; Lyon, L. A. AdV. Mater. 2004, 16, 184–187. (27) Zhang, K.; Huang, H.; Yang, G.; Shaw, J.; Yip, C.; Wu, X. Y. Biomacromolecules 2004, 5, 1248–1255. (28) Gong, J. G. Soft Matter 2006, 2, 544–552. (29) Nolan, C. M.; Serpe, M. J.; Lyon, L. A. Biomacromolecules 2004, 5, 1940–1946.

10.1021/la8017425 CCC: $40.75  2008 American Chemical Society Published on Web 08/19/2008

Microgels for pH-Triggered Capture and Release

Langmuir, Vol. 24, No. 18, 2008 10229

Table 1. Aqueous Solution Properties of the Four pH-Responsive Sterically-Stabilized Latexes Used in This Study diameter (nm)

latex type

steric stabilizer

critical swelling pH

isoelectric point

pH 9

pH 4

PDPA PDEA PDPA PDEA P2VP

PEGMA PEGMA 10qPDMASt PDMASt PEGMA

5.2-5.4 6.3-6.6 5.2-5.5 6.5-7.0 4.0-4.5

6.8 7.6 9.3 9.2 6.5

315 280 1300b 455 380

1150 920 >2800 1230 1300

a The behaviour of a high Tg poly(2-vinylpyridine) (P2VP) latex is included as a comparison. The synthesis and bulk solution behavior of this P2VP latex is reported elsewhere.43 b A local minimum in the intensity-average particle diameter was observed at approximately pH 6, which is the diameter cited here.8

microscopy (TMAFM) in air.30 After incubation for 2 h, the partially dehydrated particles became flattened and formed a uniform, homogeneous structure, which the authors described as a ‘nanofilm’. Monolayer particle films were prepared by Nerapusri et al. by adsorbing swollen PNIPAM-based microgels onto poly(ethylene imine)-coated silicon wafers, followed by drying to obtain films.31 The rehydrated films could be reversibly swollen and deswollen by cycling either the solution pH or temperature. Absorption of a cationic surfactant within these films was explored as a prelude to controlled uptake/release applications.32 Previously we reported that 380 nm diameter poly(2vinylpyridine) particles, stabilized with methoxy-capped poly(ethylene glycol) monomethacrylate (PEGMA), adsorb to both mica and silica in their latex form at pH 5 to produce randomly packed particulate monolayers that cover the entire substrate.33 This layer can subsequently be reversibly cycled between a swollen microgel-type layer which adopts localized hexagonal packing, and a collapsed, hexagonally close-packed layer of partially flattened latex particles that exposes a significant fraction of the underlying substrate. The initial layer formation was best explained by modifying the standard Random Sequential Adsorption (RSA) model (often invoked to explain particle adsorption) to include flattening and deformation of the particles after striking the surface. The absence of film formation and instead the retention of particulate character throughout the pH swelling-collapse cycle was attributed to the high glass transition temperature (Tg) of poly(2-vinylpyridine) (Tg ≈ 105 °C)34 in its nonswollen latex form. The rearrangement and desorption of many particles during the swelling process was attributed to the inability of the nonionic PEGMA stabilizer to anchor the particles sufficiently strongly to the surface so as to resist the lateral repulsive forces induced by neighboring particles. We postulated that an electrostatically bound steric stabilizer should be more strongly adsorbed and hence would minimize lateral movement and desorption. This hypothesis was tested in this current work. Herein we report the pH-responsive behavior of adsorbed, lightly cross-linked sterically stabilized poly(tertiary amine methacrylate)-based latexes. We have recently described the synthesis and aqueous solution properties of such particles, which are summarized for convenience in Table 1.8 We decided to examine adsorption of these particles at the solid/aqueous solution interface, primarily because their relatively low glass transition temperatures were expected to promote genuine interfacial film (30) Kazakov, S.; Kaholek, M.; Kudasheva, D.; Teraoka, I.; Cowman, M. K.; Levon, K. Langmuir 2003, 19, 8086–8093. (31) Nerapusri, V.; Keddie, J. L.; Vincent, B.; Bushnak, I. A. Langmuir 2006, 22, 5036–5041. (32) Nerapusri, V.; Keddie, J. L.; Vincent, B.; Bushnak, I. A. Langmuir 2007, 23(19), 9572–9577. (33) FitzGerald, P. A.; Dupin, D.; Armes, S. P.; Wanless, E. J. Soft Matter 2007, 3, 580–586.

formation. Moreover, the critical swelling pH values for such latexes lie close to neutral pH. This is significantly higher than that required for the poly(2-vinylpyridine) latexes,21 which only swell under relatively acidic conditions (around pH 4.1). We also used fluorescence spectroscopy to investigate the capacity of these adsorbed poly(tertiary amine methacrylate) latex/microgel layers to take up (and subsequently release) a model hydrophobic compound, pyrene, during pH cycling.

Experimental Section Materials. HNO3, KNO3 and KOH (all Analytical grade) were obtained from Aldrich. All water was passed through a Millipore Milli Q filtration unit before use. Silica particles (Silica 400G, Commercial Minerals, Unimin Australia Ltd.) were fractionated using a Warman Cyclosizer (Warman International Ltd.) to produce particles with an average diameter of 15 µm. Solution pH adjustments were made incrementally with HNO3 or KOH, ensuring that the pH never exceeded the target value. Lightly cross-linked (1%), sterically stabilized poly[2-(diethylamino)ethyl methacrylate] (PDEA) and poly[2-(diisopropylamino)ethyl methacrylate] (PDPA) latexes were synthesized by aqueous emulsion polymerization, as reported previously.8 Four particular latexes were selected for this study (Table 1): (i) PEGMA-stabilized PDPA, (ii) PEGMA-stabilized PDEA, (iii) 10qPDMASt-stabilized PDPA and (iv) PDMASt-stabilized PDEA, where ‘PEGMA’ is a monomethoxy-capped poly(ethylene glycol) methacrylate macromonomer (Mn ) 2000; Mw/Mn ) 1.10, i.e., n ≈ 45) and ‘PDMASt’ and ‘10qPDMASt’ are styrene-capped poly[2-(dimethylamino) ethyl methacrylate] macromonomers (n ≈ 50; ‘10q’ indicates that 10% of the DMA residues are quaternized with CH3I).8 Tg by DSC. Although not a first-order phase transition, the glass transition temperature (Tg) is accompanied by dramatic physical changes in specific volume, viscosity and heat capacity34 and is often used to empirically designate a polymer as either ‘soft’ (low Tg) or ‘hard’ (high Tg) at a particular temperature. Glass transition temperatures were measured using Differential Scanning Calorimetry (DSC). Samples were prepared in their latex form by drying at least two pH units above their critical swelling pH (i.e., approximately pH 7 for 10qPDMASt-stabilized PDPA and approximately pH 9 for PDMASt-stabilized PDEA). DSC data were collected using either a TA instruments 2920 Modulated DSC or a Mettler Toledo DSC 1 (FRS5 sensor; temperature range: -150 to 700 °C) with cryogenic cooling. AFM. In situ imaging was performed using a Digital Instruments, Nanoscope IV Atomic Force Microscope (AFM), in a room maintained at 20 °C. Images were collected using silicon nitride tips (Nanoprobe, Veeco, USA) with a nominal spring constant of 0.06 N · m-1. The tips were irradiated with ultraviolet light (∼254 nm), for 30-60 min immediately prior to use (using a Nanosciences Inc. Bioforce instrument). Images were collected in 10 mM KNO3 solution using tapping mode (TM-AFM) in a Digital Instruments fluid tapping cell. The muscovite mica substrates were cleaved immediately prior to use to provide a clean surface. Silicon wafer substrates (purchased from Silicon Valley Microelectronics, CA, with a well-defined 115 nm oxide layer) were prepared by ultraviolet irradiation for 30-60 min, then soaking in 10 wt % KOH for 10 min, followed by rinsing with copious amounts of water and then dried under a clean, dry, dust-free nitrogen stream. Latexes were adsorbed onto either a muscovite mica or silica substrate from a 10 mM KNO3 solution within the AFM fluid tapping cell for 0.5-2 h, before rinsing the cell with 10 mM KNO3 solution at the same pH to remove nonadsorbed particles. This protocol ensured that only the stable adsorbed layer was imaged. Indeed, because the AFM laser signal is scattered by colloidal particles, imaging was only possible after removal of the excess particles. Fluorescence Sample Preparation and Measurements. Samples for the fluorescence capture and release experiments were (34) Brandrup, J.; Immergut, E. H.; Grulke, E. A.; Abe, A.; Bloch, D. R. Polymer Handbook, 4th ed.; John Wiley and Sons: New York, 2005.

10230 Langmuir, Vol. 24, No. 18, 2008 prepared as follows. The silica particles were washed first with ethanol, then with 10 wt % KOH and finally with copious amounts of water. After each wash the particles were allowed to settle under gravity and the clear supernatant was removed. Excess latex (1-10 wt % solution) was added dropwise to the silica particles dispersed in 10 mL of 10 mM KNO3 and allowed to adsorb for at least 2 h at ambient temperature. The latex-coated silica particles were settled by centrifugation at 3000 rpm for 1 min. The supernatant containing the nonadsorbed latex particles was removed and the latex-coated silica particles were redispersed in fresh 10 mM KNO3. This process was repeated until the supernatant was clear (i.e., contained no free latex particles). One drop of a methanolic solution of pyrene was added to the purified latex-coated silica particles so as to produce a final pyrene concentration of 2-6 µM (i.e., not exceeding the known aqueous solubility of pyrene in water of 6 µM).35 The fluorescence spectrum (in the region of 350-500 nm) was recorded to ensure that there was no excimer formation, which would indicate a pyrene concentration greater than its aqueous solubility.36 Emission spectra were recorded between 350 - 450 nm as a function of solution pH using a Carey Eclipse fluorescence spectrophotometer (Varian) at an excitation wavelength of 330 nm, using excitation and emission slit widths of 20 and 2.5 nm, respectively, and a scan rate of 150 nm/min. Some spectra were averaged over ten scans to increase the signal-to-noise ratio to an acceptable level. Note that the pH of the supernatant was changed in situ after the addition of the pyrene. The supernatant was never turbid, either after pyrene addition or between pH changes, indicating that no significant latex or microgel desorption occurred.

Results Our investigation focused on the physicochemical behavior of two poly(tertiary amine methacrylate)-based latexes (i.e., PDPA and PDEA) prepared using either nonionic or cationic steric stabilizers, i.e., PEGMA vs PDMASt (or 10qPDMASt). Both types of particles were studied at two different pH values, with the higher pH corresponding to the nonswollen latex form and the lower pH corresponding to the swollen microgel form. Both pH values were chosen to be below the isoelectric point so that the particles had positive zeta potentials and were thus electrostatically adsorbed onto the anionic mica or silica substrate. The Tg values of poly(tertiary amine methacrylates) lie close to room temperature.34 For example, PDEA homopolymer has a Tg of 16-24 °C, PDMA homopolymer has a Tg of approximately 19 °C and poly(tert-butylaminoethyl methacrylate) has a Tg of 33 °C. Although there is no literature data available for the Tg of PDPA homopolymer, its physical appearance at room temperature is similar to that of PDEA homopolymer. Thus the first four latexes listed in Table 1 were initially expected to form films under ambient conditions. However, the Tg of the 10qPDMASt-stabilized PDPA latex was measured to be 22 ( 2 °C, but the Tg of the PDMASt-stabilized PDEA latex was measured to be -5 ( 3 °C. Nevertheless, both latexes were expected to be relatively soft at the temperatures accessed during the AFM experiments. PEGMA-Stabilized PDPA Latex Adsorbed onto Mica. AFM images for the PEGMA-stabilized PDPA particles adsorbed onto mica are shown in Figure 1, with corresponding representative normal force curves. These particles were initially adsorbed in their latex form at pH 6.0 (Figure 1a), acquired swollen microgel character at pH 3.0 (Figure 1b), and returned to their original deswollen latex form at pH 9.0 (Figure 1c). Initial adsorption leads to a disordered layer of latex particles with an average particle diameter of 313 ( 10 nm, which is comparable to the intensity-average diameter of 315 nm measured (35) Schwarz, F. P. J. Chem. Eng. Data 1977, 22, 273–277. (36) Kalyanasundaram, K.; Thomas, J. K. J. Am. Chem. Soc. 1977, 99(7), 2039–2044.

FitzGerald et al.

in bulk solution by DLS (Table 1). The small steep repulsion in the corresponding approach force curve is indicative of contact with relatively hard particles,33 while the multiple adhesive events in the retraction force curve (occurring at ∼40, 200 and 300 nm separation) have previously been ascribed to polymer chains bridging the surface and tip, followed by subsequent release when the tip is removed sufficiently far from the surface.37 On lowering the solution pH, the PEGMA-stabilized PDPA particles become protonated and form a fully swollen microgeltype layer that completely covers the substrate. However, this swollen layer retains some residual particulate character, with a large increase in the average diameter up to 2000 ( 500 nm (measured as the average center-to-center distance between neighboring particles). This substantial degree of swelling leads to strong lateral repulsive forces, which causes the majority of the particles to be expelled from the surface. For example, the number of adsorbed particles per unit area is ∼6 µm-2 for the initial layer shown in Figure 1a, but only ∼0.5 µm-2 for the swollen layer observed in Figure 1b. Thus, less than 10% of the originally adsorbed particles remain adsorbed after in situ acidinduced swelling on the mica substrate. The swollen particle diameter is significantly larger than the 1150 nm diameter observed in bulk aqueous solution, but is consistent with highly charged particles (with a zeta potential of ζ ) +54 mV in aqueous solution)8 spreading onto the highly anionic mica surface due to electrostatic attractions.17,18 The normal force curve (Figure 2b) becomes purely repulsive with no adhesive events during tip retraction from the surface. The extent of this tip-surface interaction is also significantly increased, with the force extending beyond 600 nm. The force profile can be fitted using F ) A exp(-x/L),38 where x is the separation distance between the tip and the surface, L is a characteristic decay length and A is a constant. Between 200 and 600 nm (where the data quality is best), the measured decay length is 173 ( 10 nm, which is significantly larger than the Debye length, κ-1, of 3.1 nm calculated for a purely electrostatic repulsion in 10 mM KNO3 using Debye-Hu¨ckel theory39. This difference can be attributed to an electrosteric interaction between the tip and the swollen microgel layer on the surface. This interpretation is consistent with previously published force interaction studies on adsorbed polyelectrolyte layers.40-42 We initially attempted to deswell the PDPA surface layer by returning to pH 6.0, but this was unsuccessful. Relatively slow deswelling was, however, expected given that timescales of many hours are required for complete deswelling of PDPA microgels in bulk aqueous solution.8 However, flushing the AFM cell with 10 mM KNO3 at pH 9.0 (Figure 1c) led to the formation of a highly ordered, partially deswollen particle layer within the twenty minute period required to exchange solution and begin recollecting data. This suggests that the relatively slow interfacial deswelling is simply due to insufficient hydroxide ions, which are required to deprotonate the PDPA chains. Similarly retarded kinetics have been recently reported for the rate of deswelling of poly(2vinylpyridine) latexes in aqueous solution by Dupin et al.43,44 On deswelling, the average particle diameter decreases from (37) Hugel, T.; Grosholz, M.; Clausen-Schaumann, H.; Pfau, A.; Gaub, H.; Seitz, M. Macromolecules 2001, 43(4), 1039–1047. (38) Israelachvili, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic Press: San Diego, 1992. (39) Hunter, R. J. Foundations of Colloid Science, 2nd ed.; Oxford University Press. London, 2001. (40) Biggs, S. Langmuir 1995, 11(1), 156–162. (41) Biggs, S.; Proud, A. D. Langmuir 1997, 13, 7202–7210. (42) Feiler, A.; Plunkett, M. A.; Rutland, M. W. Langmuir 2003, 19, 4173– 4179. (43) Dupin, D.; Rosselgong, J.; Armes, S. P.; Routh, A. F. Langmuir 2007, 23(7), 4035–4041.

Microgels for pH-Triggered Capture and Release

Langmuir, Vol. 24, No. 18, 2008 10231

Figure 1. AFM images (top row) and corresponding force curves (bottom row) for PEGMA-stabilized PDPA particles adsorbed onto mica from aqueous solution. (a) submonolayer coverage of adsorbed particles at pH 6.0 (the zeta potential, ζ, of these particles in aqueous solution at this pH is ζ ) +16 mV);8 (b) swollen microgel-type layer at pH 3.0 (ζ > +54 mV) and (c) deswollen particles at pH 9.0 (ζ ) -19 mV). Solid lines are the extension force curves (i.e., tip moving toward surface) and the dashed lines are the retraction force curves (i.e., tip moving away from the surface after contact).

Figure 2. AFM images (top row) and corresponding example force curves (bottom row) for (a) a film of coalesced PEGMA-stabilized PDEA particles adsorbed onto mica at pH 7.5 (the zeta potential, ζ, of these particles in aqueous solution at this pH is ζ ) +5 mV);8 (b) swollen microgel-type layer at pH 3.0 (ζ ) +47 mV); (c) partially deswollen film at pH 7.5 (ζ ) +5 mV). Solid lines are the extension force curves and the dashed lines are the retraction force curves. The force curves show (a) the initial layer, (b) an ∼200-400 nm swollen layer, and (c) a partially deswollen layer at pH 7.5.

Figure 3. 10qPDMASt-stabilized PDPA particles adsorbed onto planar silica from aqueous solution. (a) Initial isolated adsorbed particles at pH 6.3 (the zeta potential of these particles in aqueous solution is ζ ) +38 mV);8 (b) swollen microgel-type layer at pH 3.9 (ζ ) +41 mV) and (c) deswollen particles at pH 6.3 (ζ ) +38 mV).

2000 to 900 ( 50 nm, although the average center-to-center distance remains around 2000 nm. This indicates that, although the particles deswell significantly, this volumetric change is not (44) Yin, J.; Dupin, D.; Li, J.; Armes, S. P.; Liu, S. Langmuir 2008, ASAP article, DOI: 10.1021/la8014282.

accompanied by lateral surface movement. This interpretation is supported by the retention of hexagonal symmetry. The force curves (Figure 1c) obtained for the deswollen particles confirm a return to ‘hard particle’ behavior, with short-range repulsive forces and a polymeric adhesive interaction.

10232 Langmuir, Vol. 24, No. 18, 2008

FitzGerald et al.

Figure 4. (a) Film of coalesced PDMASt-stabilized PDEA particles adsorbed onto planar silica from aqueous solution at (a) pH 8.5 (the zeta potential of this latex in aqueous solution at this pH is ζ ) +12 mV);8 (b) swollen microgel-type film formed at pH 3.8 (ζ ) +46 mV); (c) deswollen film produced after returning to pH 8.5 (ζ ) +12 mV).

The initially adsorbed latex and the deswollen microgel layers are particulate in nature, with bare substrate clearly discernible between the adsorbed particles. Individual swollen microgel particles can also be observed, as has been reported by Ho¨fl et al.,17 although in this case almost the entire surface appears to be covered owing to the longer adsorption time. In summary, PEGMA-stabilized PDPA latex adsorbs onto mica at pH 6 and exhibits pH-responsive character during pH cycling. However, the latex Tg appears to be too high to allow substantial particle coalescence to form a film, as evidenced by the presence of bare substrate. PEGMA-Stabilized PDEA Latex Adsorbed onto Mica. It was anticipated that the lower Tg PEGMA-stabilized PDEA latex might be sufficiently soft to allow film formation following adsorption. Typical AFM images obtained for PEGMA-stabilized PDEA latex adsorbed onto mica at pH 7.5 are shown in Figure 2a. The aqueous solution above this adsorbed layer was then switched to pH 3.0 (see Figure 2b), before being returned to pH 7.5 (Figure 2c). In their latex form, these particles only exhibit cationic surface charge between pH 6.6 and 7.6. The initial adsorbed layer varied widely in texture, possibly because of the rather narrow pH window that was available for deposition (Table 1). Despite these experimental difficulties, the adsorbed lightly cross-linked PEGMA-stabilized PDEA particles clearly coalesce to form a film with no exposed underlying substrate, as shown in Figure 2a. The addition of acid leads to a swollen PEGMA-stabilized PDEA layer (see Figure 2b). This microgel layer has significantly lower surface roughness than the initial adsorbed latex layer. Moreover, it appears to be continuous rather than particulate in nature, since there is no evidence for the discrete features that are observed within the adsorbed swollen PEGMA-stabilized PDPA layer. The corresponding force curve (Figure 2b) shows the anticipated large increase in the effective layer thickness to ∼200-400 nm due to electrosteric repulsion within the swollen film, as discussed for Figure 1b. When the pH is returned to 7.5 (Figure 2c) this PDEA microgel film, like the particulate PDPA layer, does not immediately deswell. This behavior was also expected given the relatively slow kinetics of deswelling previously reported for PDEA microgels in aqueous solution.8 The deswelling is accompanied by an increase in film texture, however, the significant extent of the interaction force curve (Figure 2c) suggests only partial deswelling. In summary, although both PDPA and PDEA particles are electrostatically adsorbed, only the latter has a sufficiently low Tg to allow true film formation at ambient temperature. The PDPA particles are only weakly adsorbed, as evidenced by the relatively large fraction that is expelled from the surface during

acid-induced swelling. On the other hand, the relatively narrow pH window available for the electrostatic deposition of the PEGMA-stabilized PDEA particles is a limitation. To address this problem, both PDPA and PDEA latexes were synthesized using cationic steric stabilizers, and their adsorption onto silica and mica was studied. 10qPDMASt-Stabilized PDPA Latex Adsorbed on Silica and Mica. Very similar adsorption, swelling and deswelling behavior was observed on both silica and mica under otherwise identical conditions. Thus only the silica results are presented here. Also, force curves obtained on both substrates are qualitatively similar to those in Figure 1 (data not shown). The 10qPDMASt-stabilized PDPA latex was adsorbed onto silica (Figure 3a) and mica at pH 6.3. At this solution pH, the zeta potential of the particles is ζ ) +38 mV (as compared to ζ ) +16 mV for the PEGMA-stabilized PDPA latex). The pH was then adjusted to pH 3.9 to induce swelling (Figure 3b), before being returned to pH 6.3 to deswell the layer (Figure 3c). The PDMA steric stabilizer chains were 10% quaternized using methyl iodide in order to increase the electrosteric stabilization of these particles at higher pH. Like the PEGMA-stabilized PDPA latex discussed above, the 10qPDMASt-stabilized PDPA latex adsorbs as discrete particles, rather than forming a film (Figure 3a). The relatively poor quality of the image (cf. Figure 1a) is attributed to tip convolution, due to the greater height of these micrometre-sized particles (see Table 1). The 10qPDMASt-stabilized PDPA layer swells to form a compressible layer that has a considerably higher degree of disorder than that obtained for the PEGMA-stabilized PDPA particles. These AFM images were somewhat easier to obtain than those shown in Figure 1, which suggests that the cationic 10qPDMASt stabilizer chains lead to stronger latex adsorption onto the oppositely charged substrate than the nonionic PEGMA chains, as expected. Once adsorbed, the former particles are less mobile and thus are unable to relax to reduce the strong lateral interactions generated during acid-induced swelling. When the pH is returned to pH 6.3, this particulate adsorbed layer does not immediately deswell. However, frequent flushing of the AFM cell with electrolyte for more than 24 h eventually led to deswelling to produce discrete particles (Figure 3c). It is likely that faster deswelling could be achieved by working at more alkaline pH, as reported in Figure 1. For both the initial layer and the deswollen layer, bare substrate is clearly present between the adsorbed particles, as found for the PEGMA-stabilized PDPA particles. It is noteworthy that, unlike the local ordering observed for the latter system, the 10qPDMASt-stabilized PDPA particles retain their original disordered surface structure. This confirms the stronger electrostatic binding of the cationic PDMA stabilizer chains onto the anionic silica, which in turn allows the adsorbed

Microgels for pH-Triggered Capture and Release

Figure 5. Fluorescence studies of pyrene uptake/release by adsorbed PDPA and PDEA layers on 15 µm silica particles. Closed circles show the effect of pH cycling between pH 8.3 and 8.5 (latex form) and pH 4.0 (microgel form) for a 10qPDMASt-stabilized PDPA layer. Open circles show the effect of pH cycling between pH 8.3 and 8.5 (latex form) and pH 4.0 (microgel form) for a PDMASt-stabilized PDEA layer.

Langmuir, Vol. 24, No. 18, 2008 10233

The intensity ratios (I1/I3) calculated from the first and third fluorescence emission peaks for pyrene are shown in Figure 5 as a function of solution pH for the two latex-coated silica particle systems. The I1/I3 ratio is directly related to the polarity of the environment experienced by the pyrene. For example, this ratio is typically 1.8 for pure water46 but only 0.6 for n-hexane.36 The I1/I3 ratios for the deswollen microgel layers are relatively low for both systems: around 1.4 for the PDMASt-stabilized PDEA film and 1.07 for the 10qPDMASt-stabilized PDPA particles. Thus Figure 5 confirms that pyrene partitions into both types of adsorbed polymer layer at alkaline pH. The lower I1/I3 ratio observed for the PDPA layer indicates that it provides a less polar (i.e., more hydrophobic) environment for pyrene. On lowering the solution pH to 4.0, the poly(tertiary amine methacrylate) chains become protonated and hence highly hydrophilic. Thus the I1/I3 ratio increases to around 1.7 in both cases, resulting in almost complete release of the pyrene into the aqueous phase. On cycling to pH 8.3-8.5, the I1/I3 ratios return to their original values. Thus the pyrene is completely recaptured by the deswollen layers, which are now in their hydrophobic latex form. Subsequent pH cycling confirms excellent reversibility in the pyrene release and capture in the absence of any significant desorption of the polymer layer which is bound to the silica via the PDMA steric stabilizer chains. Interestingly, the morphology of the adsorbed pH-responsive polymer does not appear to be an important factor, although it is possible that this parameter may influence the kinetics of pyrene uptake/release.

particles to better resist the lateral repulsive interactions generated during acid-induced swelling. PDMASt-Stabilized PDEA Latex Adsorbed on Silica and Mica. Finally, latex particles comprising a cationic steric stabilizer (PDMASt) with the lower Tg PDEA were adsorbed onto silica, see Figure 4. Initial adsorption was conducted at pH 8.5, followed by acid-induced swelling at pH 3.8 prior to returning to the original pH. Once again the behavior on silica is essentially the same as on mica, and only the silica results are presented. Latex adsorption led to homogeneous film formation with no evidence for discrete particles over the entire substrate. This suggests efficient particle coalescence. On addition of acid, this latex film swelled appreciably and acquired cationic microgel character (Figure 4b). On returning to pH 8.5, the film deswelled rapidly (Figure 4c) as expected from the more rapid deswelling kinetics observed for this sample in solution,8 and fully recovered its original morphology. Of the four latexes studied in this investigation, these PDMAStstabilized PDEA particles produced the most robust film with respect to both morphology and pH cycling. The dry thickness of the initially adsorbed coalesced latex film was measured to be only 9 nm using a Nanofilm EP3 imaging ellipsometer (fitted with a 532 nm laser and using a film refractive index of 1.51734). This represents highly efficient spreading of the 455 nm diameter latex particles (Table 1) upon adsorption. From the in situ AFM force measurements, the acid-swollen film extends approximately 100-200 nm from the substrate, indicating partially hindered swelling by the extensively flattened adsorbed particles as has been reported by Ho¨fl et al.17 However, the potential film volume that is available to sequester or release probe molecules far exceeds that previously reported for rather less robust PDMA-b-PDEA diblock copolymer micellar coatings of equivalent chemical functionality.45 Pyrene Uptake-Release Experiments. Since the PDMAStstabilized PDEA latex formed the most robust film, this system was chosen for uptake/release studies of pyrene using fluorescence spectroscopy. The 10qPDMASt-stabilized PDPA latex was also studied so as to allow comparison between a genuine film-forming microgel layer and a particulate microgel layer. A particulate silica substrate was selected so as to increase the available surface area and hence the amount of pyrene that is taken up/released by the adsorbed poly(tertiary amine methacrylate) layer.

The adsorbed layer structure and pH-responsive behavior of the four poly(tertiary amine methacrylate) layers on mica and silica are summarized in Table 2, and the behavior of the two PDMA-stabilized latexes is summarized schematically in Figure 6. As a comparison, the behavior of a previously characterized layer of PEGMA-stabilized poly(2-vinylpyridine) particles is also included in Table 2.33 The Tg of this latter system is around 105 °C, which is far too high to allow film formation to occur in alkali pH values at ambient temperature. Thus these particles are a good model for understanding the behavior of particulate adsorbed layers. In contrast, the four types of poly(tertiary amine methacrylate) examined in this study have glass transition temperatures that are either close to or below ambient temperature. Comparison of the initial adsorbed layers (shown schematically in Figure 6) shows that the two types of PDPA-based latexes form patchy particulate layers, with plenty of exposed substrate between adsorbed particles. This was somewhat unexpected: it was originally thought that the Tg of the PDPA chains would be sufficiently low to allow at least partial coalescence of the adsorbed particles. In contrast, both types of PDEA-based latexes formed good-quality films that completely coat both mica and silica. This suggests that the film-forming behavior of these particles is primarily determined by the significantly lower Tg of the PDEA chains and is essentially independent of the nature of the steric stabilizer. In addition, the PDMASt-stabilized PDEA film proved to be very robust with respect to pH cycling, which may be related to the stronger binding of the cationic PDMASt stabilizer chains to the anionic substrates compared to the nonionic PEGMA chains. Note that the low level of quaternisation of the 10qPDMA stabilizer is not expected to significantly influence the adsorption of the 10qPDMA-stabilized PDPA particles at pH 6.3, since unquaternised PDMA chains should have a similar charge under these conditions. Thus a direct comparison can be made between

(45) 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.

(46) Ananthapadmanabhan, K. P.; Goddard, E. D.; Turro, N. J.; Kuo, P. L. Langmuir 1985, 1(3), 352–355.

Discussion

10234 Langmuir, Vol. 24, No. 18, 2008

FitzGerald et al.

Table 2. Summary of the Adsorbed Layer Structure, Surface Coverage and Film-Forming Behaviour of pH-Responsive Latexes Adsorbed onto Mica from Aqueous Solution latex type

initial adsorbed layer

swollen layer

final deswollen layer

PEGMA-stabilized PDPA

• submonolayer coverage of randomly adsorbed particles • exposed substrate

• swollen microgel-type layer • lateral rearrangement and desorption of particles • no exposed substrate

• highly ordered layer of deswollen particles • substrate re-exposed

PEGMA-stabilized PDEA

• film of coalesced particles • no exposed substrate

• swollen microgel-type film • no exposed substrate

• partially deswollen film

10qPDMASt-stabilized PDPAa • submonolayer coverage of randomly adsorbed particles • exposed substrate

• swollen microgel-type layer • no exposed substrate

PDMASt-stabilized PDEAa

• swollen microgel-type film • no exposed substrate

PEGMA-stabilized P2VPa,b

• film of coalesced particles • no exposed substrate

• no exposed substrate • disordered layer of deswollen particles • substrate re-exposed • deswollen film

• no exposed substrate • structurally disordered monolayer • swollen microgel-type layer • highly ordered, deswollen particles of adsorbed particles • lateral rearrangement and desorption • exposed substrate • substrate re-exposed of particles • no exposed substrate

a Adsorption onto both silica and mica, which are qualitatively the same for these systems. b The adsorption behavior for this P2VP latex has been reported elsewhere.33

Figure 6. Schematic representation of (a) 10qPDMASt-stabilized PDPA latex and (b) PDMASt-stabilized PDEA latex adsorbed onto silica. The behavior of PEGMA-stabilized PDPA latex is qualitatively the same as that shown in (a), except there is a significant rearrangement of particles in their swollen microgel state. Similarly, the behavior for the PEGMA-stabilized PDEA latex is qualitatively the same as shown in (b).

the adsorption behavior of the two PDMA-stabilized latexes. The final deswollen state of each adsorbed layer (achieved either through exposure to excess hydroxide or long timescales) confirms that both of the PDPA adsorbed layers return to their original particulate state, exposing the underlying surface. However, deswelling is incomplete because the PDPA particles do not recover their original dimensions. This hysteresis was also observed for PEGMA-stabilized poly(2-vinylpyridine) particles33 and is presumably due to residual electrostatic binding between the anionic substrate and the cationic polymer chains. In contrast, the PDMASt-stabilized PDEA layer readily returns to its original collapsed film morphology, which still completely covers the substrate. Thus the chemical nature of the adsorbed particles determines their final deswollen state.

Conclusions Four lightly cross-linked, sterically stabilized poly(tertiary amine methacrylate)-based latexes were adsorbed onto both mica and silica from aqueous solution. Using cationic (rather than nonionic) steric stabilizer chains ensures strong electrostatic

adsorption onto the oppositely charged substrate. The PDPAbased latexes did not undergo particle coalescence. However, in the case of the lower Tg PDEA-based particles, lowering the solution pH below their critical swelling pH leads to the formation of relatively featureless, highly swollen cationic microgel films, as judged by in situ AFM studies. These films have a much greater capacity for sequestration of hydrophobic molecules than coatings based on adsorbed micelles. Indeed, the capture and subsequent acid-induced release of a model probe (pyrene) within these adsorbed layers on silica particles is demonstrated and this behavior is shown to be reversible over several pH cycles. Acknowledgment. The ARC is thanked for Research Grant No. DP0555908. The authors also thank S. Ata for donating the silica particles, E. Smith, T. Hunter and P. Yates for ancillary physical measurements, and J. Zank for access to the Mettler DSC. S.P.A. is the recipient of a five-year Royal Society Wolfson Research Merit Award. J.I.A. is a member of CICPBA. LA8017425