Nanostructure of a Poly(acrylic acid) Brush and Its Transition in the

Jul 7, 2009 - ... transition between the carpet layer only and carpet + brush layer structures, ... available by participants in Crossref's Cited-by L...
0 downloads 0 Views 3MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

)

Nanostructure of a Poly(acrylic acid) Brush and Its Transition in the Amphiphilic Diblock Copolymer Monolayer on the Water Surface† Hideki Matsuoka,*,‡ Yoshiko Suetomi,‡ Ploysai Kaewsaiha,§ and Kozo Matsumoto

Department of Polymer Chemistry, Kyoto University, Kyoto 615-8510, Japan, and §Faculty of Science and Technology, Suan Sunandha Rajabhat University, 1U-Thong Nok Road, Wachira, Dusit Bangkok 10300, Thailand. Present address: Molecular Engineering Institute (MEI), Kinki University, 11-6 Kayanomori, Iizuka, Fukuoka 820-8555, Japan )

‡

Received April 25, 2009. Revised Manuscript Received June 10, 2009 The nanostructure and its transition of in a poly(acrylic acid) (PAA) brush in the water surface monolayers of poly(hydrogenated isoprene)-b-poly(acrylic acid) with different block lengths and block ratios were investigated by X-ray reflectivity as a function of surface pressure (brush density) and salt concentration in the subphase. The PAA brush showed the same behavior after salt addition as did the poly(methacrylic acid) (PMAA) brush, which was investigated previously. The brush chains expanded and then shrunk after passing the maximum with increasing added salt concentration. This behavior could be explained by the change in electric charges on the PAA brush chains as was observed on the PMAA brush. The PAA brush chains showed a critical brush density, where there was a transition between the carpet layer only and carpet þ brush layer structures, as did the PMAA and poly(styrene sulfonic acid) (PSS) brushes. The critical brush density was about 0.4 chains nm-2, which was higher than that of the PSS brush, a strong acid brush, and was close to that of the PMAA brush, a weak acid brush. However, the critical brush density of the PAA brush was independent of the hydrophilic chain length whereas that of the PMAA brush decreased with increasing PMAA chain length. In addition, the PAA brush had a thicker carpet layer than the PSS and PMAA brushes. Hence, the mechanism of PAA brush formation was predicted to be different from that of not only the PSS brush (strong acid brush) but also the PMAA brush.

Introduction The polymer brush, in which polymer chains are densely packed and tethered on the surface by one chain end, is a contemporary novel tool for surface modification in the areas of nanotechnology, biomaterials, lubricants, and so forth.1 The polymer brush shows very unique characters because of its special situation with respect to polymer chains.2,3 The polyelectrolyte brush might be a more complicated but interesting system because all of the brush chains are ionic.4-6 Recently, Raviv et al.7 reported that the friction between ionic brushes was very low. They estimated the friction coefficient between two surfaces by applying the SFA technique for nonionic brushes, polyelectrolyte adsorbed layers, and polyelectrolyte brushes. Extremely low friction was detected for polyelectrolyte brush surfaces. We have been investigating the nanostructure of the polymer monolayer at the air/water interface and the polyelectrolyte brush † Part of the “Langmuir 25th Year: Molecular and macromolecular selfassemblies” special issue. *To whom correspondence should be addressed. E-mail: matsuoka@star. polym.kyoto-u.ac.jp.

(1) Advincula, R. C., Brittain, W. J., Caster, K. C., R€uhe, J., Eds.; Polymer Brushes; Wiley: New York, 2004. (2) Milner, S. T. Science 1991, 251, 905. (3) Halperin, A.; Tirrell, M.; Lodge, T. P. Adv. Polym. Sci. 1992, 100, 31. (4) Zhulina, E. B.; Borisov, O. V.; Priamitsyn, V. A. J. Colloid Interface Sci. 1990, 137, 495. (5) Tran, Y.; Auroy, P.; Lee, L. T. Macromolecules 1999, 32, 8952. (6) Ahrens, H.; F€orster, S.; Helm, C. A. Phys. Rev. Lett. 1998, 81, 4172. (7) Raviv, U.; Giasson, S.; Kampf, N.; Gohy, J. F.; Jerome, R.; Klein, J. Nature (London) 2003, 425, 163. (8) Matsuoka, H.; Mouri, E. Nanostructure of Ionic Amphiphilic Block Copolymer Monolayer at an Air/Water Interface. In Encyclopedia of Nanoscience and Nanotechnology; Schwarz, J. A., Contescu, C., Putyera, K., Eds.; Marcel Dekker: New York, 2004; pp 2519-2529. (9) Mouri, E.; Matsumoto, K.; Matsuoka, H. J. Polym. Sci., Part. B 2003, 41, 1921–1928.

13752 DOI: 10.1021/la901466h

in it by X-ray and neutron reflectivity (XR, NR) techniques.8-16 We synthesized various ionic amphiphilic diblock copolymers with different compositions and block lengths and block ratios by utilizing living anionic polymerization and living radical polymerization. The polymers were spread on the water surface to form a polymer monolayer on the water. The hydrophobic block forms the hydrophobic layer on the water surface. The hydrophilic block (i.e., the polyelectrolyte block) forms an adsorbed “carpet” layer or “carpet þ brush” layer depending on the conditions under the water surface.9,10,14 The carpet layer is thought to be formed to stabilize the interface (i.e., to reduce interfacial energy between the hydrophobic layer and water subphase). These two kinds of hydrophilic layer structures-the carpet-only structure and carpet þ brush double-layer structure-are not unique to the system but a transition between them is observed depending on parameters such as brush density,11,12 hydrophilic chain length,10,15 salt concentration,16 and so forth. Helm et al.6 reported the existence of the carpet layer in addition to the brush layer, and the details of the transition such as the critical brush density have been clarified by our systematic studies. A strong advantage of the water surface system for the brush study is that the nanostructure of the brush can be investigated as a function of brush density continuously by compression and expansion of the (10) Mouri, E.; Furuya, Y.; Matsumoto, K.; Matsuoka, H. Langmuir 2004, 20, 8062. (11) Matsuoka, H.; Furuya, Y.; Kaewsaiha, P.; Matsumoto, K. Langmuir 2005, 21, 6845. (12) Matsuoka, H.; Furuya, Y.; Kaewsaiha, P.; Matsumoto, K. Macromolecules 2007, 40, 766. (13) Mouri, E.; Kaewsaiha, P.; Matsumoto, K.; Matsuoka, H.; Torikai, H. Langmuir 2004, 20, 10604. (14) Kaewsaiha, P.; Matsumoto, K.; Matsuoka, H. Langmuir 2004, 20, 6754. (15) Kaewsaiha, P.; Matsumoto, K.; Matsuoka, H. Langmuir 2007, 23, 20. (16) Kaewsaiha, P.; Matsumoto, K.; Matsuoka, H. Langmuir 2007, 23, 7065.

Published on Web 07/07/2009

Langmuir 2009, 25(24), 13752–13762

Matsuoka et al.

Article

Langmuir-Blodgett trough installed in the XR and NR instruments specially designed for liquid surface study. So far, we have systematically studied two polyelectrolyte brush systems in the water surface polymer monolayer. One is the poly(methacrylic acid) (PMAA) brush,9-13 and the second is the poly(styrenesulfonic acid) brush (PSS).14-16 The former is a weak acid brush, and the latter is a strong acid brush. Novel, unique features found in a preliminary study were the existence of “the critical brush density” and “the critical salt concentration”. The critical brush density is the brush density where the hydrophilic layer shows a transition between the carpet-only structure and the carpet þ brush layer structure. Both the PMAA and PSS brushes had a critical brush density, but the hydrophilic chain length dependence of the critical brush density was different. The critical brush density for PMAA decreased with increasing hydrophilic chain length whereas that for PSS was not affected. The critical brush density for PSS was not influenced by salt concentration. These observations suggest that the PMAA and PSS brushes are formed by a different mechanism. The critical salt concentration is the salt concentration below which the brush nanostructure is not influenced by salt addition. The PSS brush16 has a critical salt concentration, but the PMAA brush does not.13 To clarify whether these differences are simply due to the difference between weak and strong acids, in this study we examined the mechanism of the formation of the poly(acrylic acid) brush in detail. Poly(isoprene)-b-poly(acrylic acid) diblock copolymers were synthesized by living radical polymerization. The nanostructure of its monolayer on the water surface was investigated by an in situ X-ray reflectivity (XR) technique as a function of the hydrophilic chain length, the brush density (i.e., the area per molecule), and added salt concentration. The behavior of the PAA brush was different from that of the PMAA and PAA brushes, which suggested that the brush formation mechanism was unique for PAA.

Experimental Section Materials. t-Butyl acrylate was purchased from Wako Chemical (Osaka, Japan). The polymerization inhibitor was removed with 1 M NaOH(aq) and washed with pure water, followed by distillation at 6.9  104 Pa with calcium hydride twice at 60 °C after removing the water moiety with sodium sulfate. Isoprene was a product of Tokyo Kasei (Tokyo, Japan) and was distilled twice at 40 °C with calcium hydrate under an Ar atmosphere. Water used for sample and solution preparation was ultrapure water obtained from a Milli-Q system (18.2 MΩ cm). The initiator, AIBN, was obtained from Wako. Other chemicals were standard grade and used as received. Synthesis of DEPN. The mediator of living radical polymerization, N-tert-butyl-N-[1-diethylphosphono-(2,2-dimethylpropyl)] nitroxide (DEPN), was synthesized as reported previously.17 Block Copolymer Synthesis. Poly(tert-butryl acrylate) (PtBuA) was synthesized by living radical polymerization with DEPN as a mediator.18 A 200:2.5:1 mixture by molar ratio of t-butyl acrylate (tBuA), DEPN, and AIBN in a Schlenk tube was degassed by freezing three times and was polymerized under an Ar atmosphere at 120 °C for 4 h (conversion 63%, Scheme 1). Unreacted monomers were removed under vacuum and dissolved in tetrahydrofran (THF). Polymers were precipitated from a cooled water/methanol mixture (1/3 v/v) and dried under vacuum. Colorless, transparent, glassy PtBuA was obtained. The block (17) Benoit, D.; Chaplinski, V.; Braslau, R.; Hawker, C. J. J. Am. Chem. Soc. 1999, 121, 3904. (18) Benoit, D.; Grimaldi, S.; Robin, S.; Finet, J. P.; Tordo, P.; Gnanou, Y. J. Am. Chem. Soc. 2000, 122, 5929. (19) Benoit, D.; Harth, E.; Fox, P.; Waymouth, R. M.; Hawker, C. J. Macromolecules 2000, 33, 363.

Langmuir 2009, 25(24), 13752–13762

copolymer was obtained as follows.19 PtBuA, isoprene (Ip), and DEPN were mixed in a Schlenk tube in a molar ratio of 1:2000:0.43 and degassed three times by freezing (Scheme 2). Polymerization proceeded at 120 °C for 2 h under an Ar atmosphere. Polymers were dissolved in a minimum amount of THF and precipitated twice from a water/methanol mixture (1/3). After being vacuum dried, transparent, highly viscous PIp-b-PtBuA was obtained. The carbon double bond in the PIp unit was hydrogenated by using p-toluene sulfanil hydrazide (TSH).20,21 The block copolymer, 4 wt equiv of TSH, and p-xylene were put into an eggplant flask (0.01 g/mL) and refluxed for 3 h at 130 °C. The solvent was removed by evaporation, and the resultant polymer was dissolved in methanol. This solution was dialyzed against pure methanol to remove unreacted TSH. Methanol was removed by evaporation, and colorless, transparent, viscous hydrogenated block copolymer PIp-h2-b-PtBuA was obtained. Finally, the hydrolysis of tBu units was performed as follows.22-25 The polymer was dissolved in a toluene/acetic acid mixture (8/2 v/v) with 0.1 equiv of methane sulfonic acid as a catalyst (0.01 g/mL) and refluxed for 3 h at 110 °C under an Ar atmosphere (Scheme 3). The polymer obtained after solvent evaporation was dissolved in a minimum amount of THF. An equivalent volume of water was added stepwise, and then the solution was dialyzed against pure water. The purified solution was lyophilized to obtain a solid polymer. NMR Measurements. Proton nuclear magnetic resonance (1H NMR) measurements were carried out with a JEOL 400WS (JEOL, Japan). The typical solution concentration was 1.7 wt %. GPC Measurements. Gel permeation chromatography (GPC) was performed with a Jasco 880-PU system with a Jasco 930 RI detector and two Shodex KF-804 L polystyrene gel columns. The eluent was THF, and polystyrene standards (MW=700, 4000, 17 500, 49 000, and 152 000) were used.

π-A (Surface Pressure-Area per Molecule) Isotherm Measurements. A Langmuir-Blodgett (LB) trough (130 mm  60 mm, depth 5 mm, Teflon coating) and its FSD-220 controller were products of the USI system (Fukuoka, Japan). The polymer was dissolved in an ethanol/chloroform mixture (1/9 v/v) to be 1 mg/mL and was spread dropwise on the water surface via a microsyringe. After allowing 30 min for solvent evaporation, π-A was measured at a barrier moving speed of 0.010 mm/s at ambient temperature. X-ray Reflectivity (XR). XR was measured with an RINTTTR-MA (Rigaku Corp., Tokyo, Japan), which measures the water surface monolayer as a function of surface pressure. Details of the instruments26 and data analysis27 are fully described elsewhere. All of the XR measurements were performed under specular conditions. The typical accumulation times were 10 and 20 s for lower- and higher-angle regions, respectively.

Results and Discussion Polymer Characterization. An example of 1H NMR and GPC results for the PtBuA homopolymer is shown in the Supporting Information as Figure S1. The GPC profile is unimodal and gives a polydispersity index of Mw/Mn = 1.15-1.19, which is satisfactorily monodisperse and means that the reaction proceeded in a living manner. The conversion was estimated (20) H€unig, S.; M€uller, H. R.; Thier, W. Angew. Chem., Int. Ed. Engl. 1965, 4, 271. (21) Mango, L. A.; Lenz, R. W. Makromol. Chem. 1973, 163, 13. (22) Tran, A.; Prud’homme, J. Macromolecules 1977, 10, 149. (23) Huang, H.; Kowalewski, T.; Wooley, K. L. J. Polym. Sci., Part A 2003, 41, 1659. (24) Murthy, K. S.; Ma, Q.; Remsen, E. E.; Kowalewski, T.; Wooley, K. L. J. Mater. Chem. 2003, 13, 2785. (25) Pan, D.; Turner, J. L.; Wooley, K. L. Macromolecules 2004, 37, 7109. (26) Matsuoka, H.; Mouri, E.; Matsumoto, K. Rigaku J. 2001, 18, 54–67. (27) Mouri, E.; Matsumoto, K.; Matsuoka, H.; Torikai, N. Langmuir 2005, 21, 1840–1847.

DOI: 10.1021/la901466h

13753

Article

Matsuoka et al. Scheme 1. Synthesis of PIp-b-PtBuA

Scheme 2. Hydrogenation of PIp-b-PtBuA

Scheme 3. Hydrolysis of PIp-h2-b-PtBuA

by a comparison of the NMR charts of the polymer and monomer. An example of the 1H NMR spectra and GPC chart for the block copolymer before hydrogenation is shown in the Supporting Information as Figure S2. All of the NMR peaks could be assigned as shown in the Figure, and the shift of the GPC profile via the second polymerization guarantees the formation of the block copolymer. A 1H NMR chart after hydrogenation is shown in Figure S3 in the Supporting Information. Signal e from the proton at the double bond decreased remarkably, and the methyl proton signal at the hydrogenated part appeared at 0.7 ppm (peak g), which means that the hydrogenation reaction proceeded successfully. The degree of hydrogenation was calculated from the area of peak e and the total area of the other peaks, which was found to be in the range of 85-99% for the polymers synthesized. Figure S4 in the Supporting Information shows the 1H NMR chart after a hydrolysis reaction. Signal c in Figure S3, which is for the protons of the t-butyl unit, is not observed in Figure S4, and a new peak was detected at 12.5 ppm, which can be assigned to the proton of carboxylic acid (peak h). The degree of hydrolysis was evaluated from the area ratio of peak h and other peaks and was found to be more than 90%. The characteristics of the polymers thus estimated are summarized in Table 1. Three block copolymers thus obtained have almost the same hydrophobic chain length (degree of polymerization (DP) is 76-81) but have different hydrophilic chain (poly(acrylic acid)) lengths (DP = 30, 54, 77). Salt Concentration Dependence of the π-A Isotherm. The π-A isotherms of the spread monolayer on the water surface for three polymers are shown in Figure 1. The surface pressure π increased smoothly by compression, which indicates uniform, stable monolayer formation. The salt (NaCl) concentration in the water subphase was changed from 0 to 2.0 M. For all of the polymers investigated, the π-A isotherm shifted toward smaller A with increasing salt concentration but came back to the larger-A region when the salt concentration was higher than 0.10 M (as indicated by the arrows in the Figure). This observation should mean that some transition occurs under the 0.10 M NaCl condition. Similar behavior was also observed for the monolayer 13754 DOI: 10.1021/la901466h

Table 1. Characteristics of (Ip-h2)m-b-(AA)n sample

ma:nb

Mnc

Mw/Mnd

a b c

81:30 7900 1.36 79:54 9400 1.26 76:77 10 900 1.28 a Determined by 1H NMR of the parent copolymer in CDCl3. b Determined by GPC of the parent homopolymer (with THF as an eluent with polystyrene standards). c Calculated from m and n. d Determined by GPC.

having a PMAA brush.13 The results of detailed analysis and discussion will be given later with the results of XR analysis. One may note that some isotherms do not start from zero surface pressure but from a relatively higher surface pressure. Although we did not try to measure isotherms in large-A regions because we concentrated on smaller-A regions where the structural transition of monolayers occurs, this phenomenon might be interesting. The XR technique is not a suitable method for investigating what happens in these large-A regions, but other methods such as AFM or Brewster angle microscopy (BAM) in addition to further π-A analysis including hysteresis studies might bring us some information, which is another goal. Hydrophobic Chain Length Dependence of the π-A Isotherm. Figure 2 is a comparison of the π-A isotherms for three polymers. As described in the previous section, three polymers have almost similar hydrophobic chain lengths but different hydrophilic chain lengths. Hence, Figure 6 shows the effect of the hydrophilic chain length on the π-A isotherm. In a relatively higher surface pressure region (i.e., at smaller A values), the surface pressure π is higher for longer hydrophilic chain polymers at the same A value. The same A value means the same hydrophilic layer thickness on the water surface. Hence, this observation reflects the effect of the hydrophilic chain on the surface pressure of the monolayer. It is quite natural and understandable that a longer hydrophobic PAA chain produces stronger steric and electrostatic repulsions, resulting in a higher surface pressure π. Surface Pressure (Brush Density) Dependence of the PAA Brush Nanostructure. Figures 3-5 show the XR profiles (left) and density profiles (right) for the monolayers of three polymers Langmuir 2009, 25(24), 13752–13762

Matsuoka et al.

Article

Figure 2. π-A isotherms for (Ip-h2)m-b-(AA)n monolayers with different hydrophilic chain lengths.

Figure 1. (a) π-A isotherms for the (Ip-h2)81-b-(AA)30 monolayer on NaCl aqueous solutions of different concentrations (solvent: EtOH/CHCl3 (1/9 v/v)). (b) π-A isotherms for the (Ip-h2)79-b(AA)54 monolayer in NaCl aqueous solutions of different concentrations (solvent: EtOH/CHCl3 (1/9 v/v)). (c) π-A isotherms for the (Ip-h2)76-b-(AA)77 monolayer on NaCl aqueous solutions of different concentrations (solvent: EtOH/CHCl3 (1/9 v/v)).

as a function of surface pressure. The surface pressure was controlled by the compression of the LB trough, hence the surface pressure dependence is equivalent to the brush density (number of Langmuir 2009, 25(24), 13752–13762

hydrophilic chains in a unit area) dependence. The density profiles were evaluated by model fitting XR profiles with two-box or three-box models, as in our previous studies.8-16,27 The structural parameters obtained by the fitting are summarized in Tables S2-S4. In the XR profiles, clear Kiessig fringes were observed at higher surface pressure conditions for all three polymer monolayers. This indicates uniform, smooth monolayer formation. By compression, the fringe position shifted toward smaller q (scattering vector, q=(4π sin θ)/λ, where θ is the reflection and incident angle and λ is the X-ray wavelength (1.5406 A˚)), which indicates an increase in monolayer thickness. In general, XR profiles at low surface pressure were well reproduced by the two-box model consisting of hydrophobic and carpet layers. However, for those at higher surface pressure, it was necessary to use the three-box model in which the brush layer formation under the carpet layer is newly taken into account.9,10 This means that the monolayer shows a transition from the carpet-only structure to the carpet þ brush layer structure by compression (i.e., with increasing brush density). This trend is clearly seen in the density profiles in Figures 3-5. The density profiles normal to the surface for the monolayer on the water surface evaluated by box model fitting of XR profiles are shown on the right of these Figures. Z is the depth, and Z=0 at the water surface. On the left is the air phase, and on the right is the water subphase. δ is the difference from unity of the real part of the refractive index (i.e., the refractive index n = 1 - δ - iβ), which is equivalent to the electron density. On the water, there is a uniform hydrophobic chain layer that has a constant density. The thickness of this layer increased with compression. In the water subphase at low surface pressure, there is only a carpet layer just beneath the water surface with a thickness is about of 30 A˚. By compression, the brush layer is formed under the carpet layer and its thickness increased with compression, whereas the carpet layer thickness was not largely influenced. The density inside the brush also increased with compression. The nanostructure parameters thus evaluated are summarized in Tables S1-S3 in the Supporting Information. The thickness of each layer and the surface pressure are plotted as functions of brush density in Figure 6 for these three polymer monolayers. With increasing brush density via compression, the hydrophobic layer thickness increased accordingly. However, the carpet layer thickness is almost independent of the brush density. This observation itself is, in principle, the same as those for the PMAA9,10,27 and PSS15 cases studied previously, but it might be DOI: 10.1021/la901466h

13755

Article

Matsuoka et al.

Figure 3. XR profiles (left) and density profiles (right) for the (Ip-h2)81-b-(AA)30 monolayer on the water surface at different surface pressures.

Figure 4. XR profiles (left) and density profiles (right) for the (Ip-h2)79-b-(AA)54 monolayer on the water surface at different surface pressures.

noted that the carpet layer thickness is slightly larger for the PAA layer. The carpet layer thickness was about 15 and 10 A˚ for the PMAA and PSS brush layers, respectively, whereas it is 20-30 A˚ for the present PAA brush layer. We propose that carpet layer formation reduces the interfacial free energy by avoiding direct contact between the hydrophobic layer and water subphase. In this sense, a slightly thicker PAA carpet layer can be attributed to the higher hydrophilicity of PAA chain compared to that of PSS and PMAA chains. From Figure 6, the critical brush densities, where the transition from the carpet-only structure to the carpet þ brush layer structure occurs, can be estimated to be 0.47, 0.53, and 0.45 chains nm-2 for 81:31, 79:54, and 76:77 polymers, respectively, although the exact values are difficult to estimate. These values are larger than those for a PSS brush and very close to those for a PMAA brush, which is also a weak acid brush. (28) Zhu, J.; Eisenberg, A.; Lennox, R. B. J. Am. Chem. Soc. 1991, 113, 5583.

13756 DOI: 10.1021/la901466h

It has been reported that some block copolymer forms 2D surface micelles on the water surface28 instead of a uniform hydrophobic layer on the water and a brush layer under the water. This phenomena is observed when the polymer is composed of a rather rigid hydrophobic block whose glass-transition temperature (Tg) is high, such as polystyrene. It is also well known that the polystyrene homopolymer forms nanoparticles on the water surface.29 However, this is not the case when the hydrophobic segment is a low-Tg polymer. This is the reason that we chose hydrogenated isoprene as the hydrophobic block. This point was already confirmed in our series of study for poly (hydrogenated isoprene)-b-poly(styrene sulfonate) monolayer systems.14-16 Uniform layer formation in the present study is also obvious from XR fitting parameters in Tables S1-S3 (and also in Tables S4-S9 that appear later) in the Supporting (29) Kumaki, J. Macromolecules 1986, 19, 2258.

Langmuir 2009, 25(24), 13752–13762

Matsuoka et al.

Article

Figure 5. XR profiles (left) and density profiles (right) for the (Ip-h2)76-b-(AA)77 monolayer on the water surface at different surface pressures.

Figure 6. Thickness of each layer as a function of brush density for the (Ip-h2)m-b-(AA)n monolayer on the water surface: m:n = (a) 81:30, (b) 79:54, and (c) 76:77.

Information. The roughness between air and hydrophobic layer is only about 3 A˚, which is sufficiently smooth and flat. For a 2Dmicelle system, it was reported to be about 10 A˚. If one is willing confirm this situation, then AFM may be a suitable methods whereas BAM does not have high enough spatial resolution. However, for the AFM measurement, the monolayer should be deposited on the solid substrate and dried. We have demonstrated in our previous study30 how largely different the dried sample AFM image is from the actual wet sample structure. Salt Concentration Dependence of the PAA Brush Nanostructure. Figure 7 shows the salt concentration in the subphase dependence of XR profiles and density profiles for three polymer monolayers at a constant surface pressure (21 mN/m). The (30) Kago, K.; Matsuoka, H.; Yoshitome, R.; Mouri, E.; Yamaoka, H. Langmuir 1999, 15, 4295.

Langmuir 2009, 25(24), 13752–13762

structural parameters evaluated by box model fitting are shown in Tables S4-S6 in the Supporting Information. The NaCl concentration was changed from 0 to 2 M. For all cases, the Kiessig fringe shifted toward the higher-q region and then came back to smaller q with increasing salt concentration. (indicated by solid straight lines in the Figure for clarity) The change in this shifting behavior occurred at 0.01 M salt, which is the same salt concentration as for the transition observed in the π-A study. As is clear from the density profiles in Figure 7, the hydrophobic layer thickness decreases with increasing salt concentration up to 0.10 M and then increases. This means that the monolayer expanded with increasing salt up to 0.1 M and then shrunk beyond 0.10 M salt concentration, where the hydrophobic layer thickness increases, which means that this is a layer compression process under the constant surface pressure condition. As fully discussed in our previous report on a PMAA brush on the water DOI: 10.1021/la901466h

13757

Article

Matsuoka et al.

Figure 7. (a) XR profiles (left) and density profiles (right) for the (Ip-h2)81-b-(AA)30 monolayer on aqueous solutions of different NaCl concentrations at 21 mN/m. (b) XR profiles (left) and density profiles (right) for the (Ip-h2)79-b-(AA)54 monolayer on aqueous solutions of different NaCl concentrations at 21 mN/m. (c) XR profiles (left) and density profiles (right) for the (Ip-h2)76-b-(AA)77 monolayer on aqueous solutions of different NaCl concentrations at 21 mN/m.

13758 DOI: 10.1021/la901466h

Langmuir 2009, 25(24), 13752–13762

Matsuoka et al.

Article

Figure 8. Salt effect on the (Ip-h2)m-b-(AA)n (m:n=(a) 81:30, (b) 79:54, and (c) 76:77) monolayer on the water surface at the same brush densities ((a) 0.70, (b) 0.53, and (c) 0.50 nm-2).

surface,13 the origin of this observation can be attributed to the change in the degree of dissociation of the weak acid group (COOH) on the brush chains. Under a low-salt condition, the degree of dissociation of the weak acid increases, hence the number of ionic charges on the brush chain increases.31 The electrostatic repulsion between brush chains is enhanced, hence the monolayer is expanded. However, the shielding effect of small salt ions on the electrostatic interaction (repulsion) also increased with increasing salt concentration. Beyond 0.10 M, because the shielding effect is more pronounced, the monolayer shrinks again. The brush layer thickness appears to decrease up to 0.10 M salt and then increases while the carpet layer thickness is almost constant. However, it should be noted that this behavior is not simply due to an electrostatic effect but also is largely affected by the brush density change. In this experiment, to maintain constant surface pressure, the brush density (or A value) changes. The constant carpet layer with different A values indicates a change in the portion of hydrophilic chains contributing to brush layer formation. Hence, this observation is not simple but is a superimposed effect of the salt concentration change and brush density change. To avoid the complex situation above, we performed another salt concentration dependence experiment (i.e., while maintaining a constant brush density) whose results are shown in Figure 8. In this experiment, we kept the brush density constant at three different values (0.70, 0.53, and 0.50 nm-2) and plotted the carpet thickness and brush layer thickness as functions of salt concentration. The surface pressure values were also plotted. Needless to say, the hydrophobic layer thickness is constant in this experiment. We can see the same trend in Figure 8 regardless of the brush density value. The carpet layer thickness is almost constant, but the brush layer thickness and surface pressure are maximized at 0.10 M salt concentration although the behavior of the brush layer thickness at low surface pressure is not very clear. This observation can be clearly interpreted by the concept described in the previous section: with increasing salt concentration, the number of charges on the brush chains increases, which results in the stretching of the brush layer and increasing surface pressure due to the enhancement of electrostatic repulsion between brush chains. In conclusion, the salt concentration effect consists of two factors: (1) an increase in the charge number on the brush and (2) an electrostatic shielding effect. The competition of these two (31) Castellan, G. W. Physical Chemistry, 3rd ed.; Addison-Wesley: Reading, MA, 1983; Chapter 16.

Langmuir 2009, 25(24), 13752–13762

Figure 9. Effect of salt on PAA brush layers.

effects, which act in opposite directions, produces the transition under the 0.10 M salt condition. This interpretation is also supported by the change in surface pressure, which shows a maximum at the same salt concentration. The concept of this interpretation is shown in Figure 9 schematically. Zhulina et al. predicted this phenomenon theoretically.32 We observed the same trend for the PMAA brush previously13 and for the PAA brush in this study. Currie et al. reported a similar trend,33 and Kurihara et al.34also observed an increase in repulsive force by salt addition via a surface force measurement (SFA). Hence, this behavior should be common to the weakly charged polyelectrolyte brush. The salt effect on the carpet-only structure under a low surface pressure condition was also investigated (Figure 10 and Tables S7-S9 in the Supporting Information). No structural change by salt has been reported for the carpet layers for the PMAA and PSS brush systems.13,16 However, a change in the XR profile can be observed in Figure 10. Careful fitting analysis revealed that this change is due to hydrophobic layer thickness with constant carpet layer thickness. This means that the situation for the carpet layer, probably the charged state, is affected by salt ions. As discussed previously, the PAA carpet layer is thicker than the PSS and PMAA carpet layers. Hence, the PAA carpet layer should be in a swollen state, which is more easily influenced by salt addition, whereas the PMAA and PSS carpet layers are almost in a bulk state in which the carpet layer density is close to that in the bulk state. This observation is understandable because the PAA (32) Zhulina, E. B.; Birstein, T. M.; Borisov, O. B. Macromolecules 1995, 28, 1491. (33) Currie, E. P. K.; Sieval, A. B.; Fleer, G. J.; Cohen-Stuart, M. A. Langmuir 2000, 16, 8324. (34) Kurihara, K.; Kunitake, T.; Higashi, N.; Niwa, M. Langmuir 1992, 8, 2087.

DOI: 10.1021/la901466h

13759

Article

Matsuoka et al.

Figure 10. XR profiles for the (Ip-h2)m-b-(AA)n monolayer on aqueous solutions of different NaCl concentrations: m:n = (a) 81:30, (b) 79:54, and (c) 76:77.

Figure 11. Layer thickness of the (Ip-h2)m-b-(AA)n monolayer on the water surface: (left) PIp-h2 layer, (center) carpet layer, (right) brush layer.

chain should be more hydrophilic than the PSS and PMAA chains. Although the difference between PAA and PMAA is only one methyl group on the main chain, it is well known that this small difference produces a large difference in behavior.35 As will be discussed later, the nature of the PAA chain is one of the important factors that greatly influence the brush formation mechanism. Hydrophilic Chain Length Dependence of the PAA Brush Nanostructure. The effect of hydrophilic chain length on the nanostructure of the PAA brush was investigated. Each layer thickness for three polymers having different hydrophilic chain (PAA) lengths is shown in Figure 11 as a function of surface pressure (i.e., the brush density). In this plot, the white zone is the carpet-only structure region, and the shadowed zone is the carpet/ brush double-layer structure region. It is clear that the carpet layer thickness is almost independent of surface pressure regardless of the PAA length. However, the thickness of both the hydrophobic and brush layers increased with compression and was influenced by the PAA length. Also, it is interesting that the thicker the brush (35) Takahashi, A.; Kato, T.; Nagasawa, M . J. Phys. Chem. 1967, 71, 2001.

13760 DOI: 10.1021/la901466h

layer, the thinner the hydrophobic layer. It is quite natural to observe a thicker brush layer under higher surface pressure conditions for longer PAA chain systems, but the hydrophobic layer (PIp-h2 layer) is thinner for longer PAA polymers. This means that a longer PAA monolayer is more expanded at the same surface pressure. This observation is understandable if we take the electrostatic and steric repulsion effects between brush chains into account. The longer the brush chains, the stronger the repulsion, which results in a thinner hydrophobic layer (i.e., monolayer expansion). Salt Concentration Dependence of the Critical Brush Density. To elucidate the PAA brush formation mechanism, in Figure 12 we plotted the critical brush density as a function of salt concentration and compared the results with those obtained previously for the PSS brush having the same hydrophobic chain. As shown in Figure 12 (right), the critical brush density was not influenced by salt addition up to the critical salt concentration (0.2 M in this case) for the PSS brush, but then it increased. The existence of a critical salt concentration could be explained by the difference between the ion concentration inside and outside of the brush regions.16,32 This argument supports the fact that the Langmuir 2009, 25(24), 13752–13762

Matsuoka et al.

electrostatic interaction is a prior factor for brush formation for the PSS brush, a strong ionic brush. However, as seen in Figure 12 (left), the critical brush density is not influenced by salt addition. This observation means that the electrostatic interaction does not play a major role in the brush formation of PAA chains, which may form an annealed brush. One more point is the absolute value of the critical brush density. The chain brush density for PSS is around 0.1 brush chains/nm2 whereas that for the PAA brush is around 0.40.5, much higher values. This fact is in agreement with the origin of brush formation discussed above. Because the electrostatic interaction, which is strong and long-ranged, is the main factor for PSS brush formation, the PSS brush is formed at lower chain brush density. It should be recalled that the critical brush density for PMAA, a weak acid brush, was also in the range of 0.2-0.5, which is comparable to that of the PAA brush studied here.

Figure 12. Critical brush density variation for the (Ip-h2)76-b(AA)77 (left) and (Ip-h2)215-b-(SS)31 (right) monolayers as a function of salt concentration.

Article

Hydrophilic Chain Length Dependence of the Critical Brush Density. The critical brush density for the PAA brush was plotted as a function of the hydrophilic chain length in Figure 13 together with those for the PMAA brush11,12 and PSS brush15 studied previously. The critical brush density for PMAA decreased from a relatively high value (about 0.5 chains/nm2) to about 0.2 chains/nm2 with increasing brush chain length. This tendency was explained by the interfacial stabilization of the water/hydrophobic layer interface by the adsorption of hydrophilic chains. For PMAA, it is assumed that the brush is formed after the interface is stabilized. Because the adsorption of longer hydrophilic chains is more effective when the brush density is the same, the longer hydrophilic chain forms a brush layer at a low brush density more easily. However, the critical brush density of the PSS brush was not influenced by the brush chain length, and its absolute value is low (about 0.1 chains/nm2). This trend is understandable if we consider the electrostatic interaction (repulsion) between hydrophilic chains to be the major factor for brush formation rather than interfacial stabilization. This is reasonable because the PSS brush is strongly ionic but the PMAA brush is weakly ionic. As shown in Figure 13 (left), the critical brush density for the PAA brush as a function of hydrophilic chain length is different from those of the PMAA brush and PSS brush; it is almost constant at higher values of about 0.4. This observation is reminiscent of the fact that the brush formation mechanism for the PAA brush is different not only from that of the PSS brush but also from that of the PMAA brush.

Figure 13. Critical brush density as a function of the DP of the hydrophilic chain: (left) (Ip-h2)m-b-(AA)n, (center) (Et2SB)m-b-(MAA)n, (right) (Ip-h2)m-b-(SS)n).

Figure 14. Possible mechanisms for brush formation. Langmuir 2009, 25(24), 13752–13762

DOI: 10.1021/la901466h

13761

Article

Possible Mechanism of Brush Formation for Polyelectrolytes. From the discussion of the mechanism of brush formation for PAA obtained in this study together with those for PSS and PMA brushes, it is plausible that the three brushes have different brush formation mechanisms although they are all anionically charged polyelectrolyte brushes. Possible mechanisms are schematically shown in Figure 14. Because the chain brush density of the PMA brush depends on the brush chain length, the dominant factor is thought to be the stabilization of the hydrophobic layer/water interface.11,12 When the interface is satisfactory stabilized by the adsorption of hydrophilic chains, the brush layer forms. The chain brush density of the PSS brush does not depend on the hydrophilic chain length but is influenced by the added salt concentration. This means that the major factor of brush formation for PSS is an electrostatic interaction between ionic hydrophilic chains. Strong repulsion between sulfonic acid groups might be more predominant than interfacial stabilization. This is reasonable because PSS is strongly ionic and PMA is a weak acid. Brush formation behavior for PAA was different not only from that of the PSS brush but also from that of the PMA brush although PMA and PAA are both weak acids. The difference between PAA and PMA should come from the methyl group on the PMA’s main chain because this is only the difference. It is well known that PMA and PAA homopolymers exhibit quite different behavior in solution despite small difference in chemical structure.36 A representative example is the conformational change brought about by the change in the degree of dissociation predicted by pH titration for PMA, which does not occur for PAA. The origin of this phenomenon has been attributed to the hydrophobic effect of methyl groups on the PMA’s main chain. The contribution for hydrophobicity from one methyl group cannot be overlooked. This might also be the case here, at least for different brush formation behavior. The PAA brush layer should be more hydrophilic than the PMA brush layer. This is supported by our observation of PAA carpet layer thickness: the PAA carpet layer is thicker (about 20-30 A˚) than either the PMA (about 15 A˚) or PSS (about 10 A˚) carpet layer. This means that the PAA carpet layer’s ability to (36) Sugai, S.; Nitta, K. Biopolymers 1973, 12, 1363.

13762 DOI: 10.1021/la901466h

Matsuoka et al.

stabilize the hydrophobic layer/water interface is lower. Furthermore, the chain brush density is not affected by salt addition. This observation suggests that the PAA brush is formed by a steric effect in the PAA layer before the interfacial stabilization is completed.

Conclusions The nanostructure and formation mechanism of the PAA brush in the poly(hydrogenated isoprene)-b-poly(acrylic acid) monolayer on a water surface were studied by X-ray reflectometry. The diblock copolymers were synthesized by living radical polymerization with DEPN as the mediator. The PAA brush in the polymer monolayer also forms a carpet/brush double-layer structure and showed a carpet-only/carpet þ brush structure transition, as was the case for PMA and PSS brushes. The brush thickness showed a maximum with increasing added salt concentration as was observed for the PMA brush previously. This phenomenon, which can be attributed to the change in the degree of dissociation of carboxylic acid groups, can be regarded as common to weak acid brushes. The critical brush density for this transition and its dependence on the hydrophobic chain length and added salt concentration were investigated. The observed trends for the PAA brush were different not only from those of the PSS brush but also from those of the PMA brush. The PAA, PMA, and PSS brushes are thought to be formed by a different mechanism, and a steric effect is assumed to be the dominant factor for the formation of the PAA brush. Acknowledgment. This work was supported by a grant-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, to whom our sincere gratitude is due (A15205017 and B19350058). This work was also supported by the 21st Century COE Program, COE for a United Approach to New Materials Science. Supporting Information Available: GPC charts and NMR spectra for synthesized polymers and XR profile fitting parameters. This material is available free of charge via the Internet at http://pubs.acs.org.

Langmuir 2009, 25(24), 13752–13762