Surfactant

pubs.acs.org/Langmuir © 2009 American Chemical Society

Effect of Graft Density on the Nonionic Bottle Brush Polymer/Surfactant Interaction Imre Varga,*,†,‡ Robert M
esz
aros,*,‡ Ricardas Makuska,^ Per M. Claesson,†,§ and Tibor Gil
anyi‡ †

Department of Chemistry, Surface and Corrosion Science, Royal Institute of Technology, Drottning Kristinas v€ ag 51, SE-100 44 Stockholm, Sweden, ‡Institute of Chemistry, E€ otv€ os University, 1117 Budapest, P
azm
any s. 1/A, Hungary, §YKI, Institute for Surface Chemistry, P.O. Box 5607, SE-114 86 Stockholm, Sweden, and ^Department of Polymer Chemistry, Vilnius University, Naugarduko 24, LT-03225 Vilnius, Lithuania Received April 27, 2009. Revised Manuscript Received August 2, 2009 The effect of graft density on the interaction of nonionic bottle brush polymers with an anionic surfactant (sodium dodecyl sulfate) was investigated. The graft density of 45 units long poly(ethylene oxide) (PEO) side chains was varied in a wide range (30, 50, 75, 90, and 100%) on a methacrylate type polymer backbone. The surfactant binding isotherms were determined by the potentiometric method in the presence of 0.1 M sodium bromide. It was found that due to the grafting of the PEO chains to a polymer backbone the surfactant binding becomes significantly suppressed. The amount of bound surfactant at the critical micelle concentration (cmc) decreases almost 2 orders of magnitude compared to the binding on a linear PEO having a similar molecular weight. The binding of the surfactant was found to occur in cooperative fashion, though the critical aggregation concentration (cac) of the binding was found surprisingly small. This result was interpreted in terms of the surfactant aggregation numbers that were found much smaller in the case of the bottle brush polymers than in the case of linear PEOs due to the steric crowding of the grafted PEO chains. To confirm the results of the binding isotherm measurements, steady-state fluorescence probe (pyrene) measurements as well as static and dynamic light scattering measurements were performed.

1. Introduction Over the past few years, comb polymers have received increased attention as biodegradable materials with added functionality in next-generation products. These features arise from their interesting associative behavior in solution1,2 as well as their unique characteristics at interfaces.3-9 Comb polymers with nonionic hydrophilic side chains are commonly applied in a wide range of industrial products such as surface coatings and inks. PEO grafted bottle brush polyelectrolyte additives have been used to sterically stabilize red blood cells and fibroblasts10 as well as inorganic suspensions.11,12 *Corresponding authors. E-mail: [email protected]. (I.V.); meszaros@ chem.elte.hu (R.M.). (1) Basak, P.; Nisha, C. K.; Manorama, S. V.; Maiti, S.; Jayachandran, K. N. J. Colloid Interface Sci. 2003, 262, 560–565. (2) Duschner, S.; Gr€ohn, F.; Maskos, M. Polymer 2006, 47, 7391. (3) van der Linden, C. C.; Leermakers, F. A. M.; Fleer, G. J. Macromolecules 1996, 29, 1000–1005. (4) Sartori, A.; Johner, A.; Viovy, J. L.; Joanny, J.-F. Macromolecules 2005, 38, 3432. (5) Striolo, A.; Jayaraman, A.; Genzer, J.; Hall, C. K. J. Chem. Phys. 2005, 123, 64710. (6) Naderi, A.; Iruthayaraj, J.; Vareikis, A.; Makuska, R.; Claesson, P. M. Langmuir 2007, 23, 12222–12232. (7) Olanya, G.; Iruthayaraj, J.; Poptoshev, E.; Makuska, R.; Vareikis, A.; Claesson, P. M. Langmuir 2008, 24, 5341–5349. (8) Naderi, A.; Olanya, G.; Makuska, R.; Claesson, P. M. J. Colloid Interface Sci. 2008, 323, 223–228. (9) Iruthayaraj, J.; Olanya, G.; Claesson, P. M. J. Phys. Chem. C 2008, 112, 15028. (10) Elbert, D. L.; Hubbell, J. A. Chem. Biol. 1998, 5, 177–183. (11) Kirby, G. H.; Lewis, J. A. J. Am. Ceram. Soc. 2004, 87, 1643–1652. (12) Yoshikawa, J.; Lewis, J. A. J. Am. Ceram. Soc. 2009, 92, S42–S49. (13) Elbert, D. L.; Hubbell, J. A. Macromolecules 1997, 30, 6947–6956. (14) Zhang, Z.; Ma, H.; Hausner, D. B.; Chilkoti, A.; Beebe, T. P., Jr. Biomacromolecules 2005, 6, 3388–3396. (15) Zhou, Y.; Liedberg, B.; Gorochovceva, N.; Makuska, R.; Dedinaite, A.; Claesson, P. M. J. Colloid Interface Sci. 2007, 305, 62–71.

Langmuir 2009, 25(19), 11383–11389

The adsorption of these types of comb polyelectrolytes can be also utilized to prepare protein resistant coatings13-15 as well as surfaces with good lubrication properties.16 Ionic surfactants can strongly interact with both neutral and oppositely charged polymers. The bulk association between linear macromolecules and charged amphiphiles has been extensively studied, and the phenomenon is relatively well understood. It has been shown17 that the ionic surfactant binding on neutral, flexible polymers is a cooperative interaction. This means that the surfactant molecules interact with the polymer chain in the form of surfactant aggregates. The polymer/surfactant complex is usually viewed as a “string of beads” in which the polymer chain connects micelle-like surfactant aggregates by wrapping around them. Because of the cooperative nature of the interaction, the surfactant binding starts at a well-defined surfactant concentration that is usually called the critical aggregation concentration (cac). Above the cac the bound amount rapidly increases, which is accompanied by a slow increase of the equilibrium (monomer) surfactant activity. The bound amount increases until either saturation occurs or the equilibrium surfactant activity reaches the critical micelle concentration (cmc), where free micelles form in the solution. The ionic surfactants interact in a similar manner with oppositely charged polyelectrolytes.18 However, because of the very strong interaction of the polymer and the surfactant, the cac is usually orders of magnitude smaller than in the case of neutral polymers and an associative phase separation takes place as charge neutralization is approached. In access surfactant the (16) Pettersson, T.; Naderi, A.; Makuska, R.; Claesson, P. M. Langmuir 2008, 24, 3336–3347. (17) Kwak, J. C. T. Polymer-Surfactant Systems; Surfactant Science Series; Marcel Dekker: New York, 1998; Vol. 77. (18) Holmberg, K.; J€onsson, B.; Kronberg, B.; Lindman, B. Surfactants and Polymers in Aqueous Solution, 2nd ed.; John Wiley & Sons: New York, 2002.

Published on Web 09/09/2009

DOI: 10.1021/la901499x

11383

Article

resolution of the complexes can take place due to the overcharging of the complexes. On the other hand, there are only a few studies on the complexation between comb polymers and surfactants. For instance, Middleton et al. revealed that the interaction between sodium dodecyl sulfate (SDS) and the likely charged PEO grafted sodium methacrylate comb polymers is negligible.19 By contrast, recent studies of Bastardo et al.,20 Nisha et al.,21,22 and Bokias and coworkers23,24 demonstrated a strong interaction between amphiphiles and oppositely charged comb polyelectrolytes with neutral side chains. The authors revealed that in the case of moderate charge density of the backbone no precipitation occurs in a wide surfactant concentration range due to the formation of a sterically stabilized dispersion of the hydrophobic polyelectrolyte/surfactant aggregates. A similar behavior was observed in the case of charged bottle brush polyelectrolytes and oppositely charged polyelectrolytes, where the formation of water-soluble molecular complexes could be observed independently of the polyelectrolyte mixing ratio.25-29 According to our knowledge, the effect of the architecture of neutral comb polymers on their association with ionic surfactants has not been explored yet. The present paper focuses on the interaction between SDS and a novel type of neutral comb polymers30,31 with PEO side chains. The investigated bottle brush polymers are composed of poly(ethylene oxide) methyl ether methacrylate [(PEO)45MEMA] and 2-hydroxyethyl methacrylate [HEMA] units (see Figure 1). HEMA:(PEO45)MEMA-X represents the general abbreviation of the used bottle brush polymers. The subscript 45 refers to the number of ethylene oxide units in the side chain, and X denotes the molar percentage of HEMA:(PEO45) units (the graft density) in the copolymer. The PEO homopolymers are known to interact strongly with SDS.32-36 This interaction exhibits a strong dependence on molecular weight.32-36 The association between PEO and SDS occurs in the bulk only if the polymer molecular weight exceeds 1000 g/mol. In the molecular weight range 1000 < Mw < 8000, the critical aggregation concentration (cac) decreases with increasing Mw. Above Mw ≈ 8000, however, the cac is not affected by the size of the PEO molecules.36 We investigate the effect of PEO graft density on the association between SDS and HEMA:(PEO45)MEMA-X comb (19) Middleton, H.; English, R. J.; Williams, P. A.; Broze, G. Langmuir 2005, 21, 5174. (20) Bastardo, L. A.; Iruthayaraj, J.; Lundin, M.; Dedinaite, A.; Vareikis, A.; Makuska, R.; van der Wal, A.; Fur
o, I.; Garamus, V. M.; Claesson, P. M. J. Colloid Interface Sci. 2007, 312, 21–33. (21) Nisha, C. K.; Basak, P.; Manorama, S. V.; Maiti, S.; Jayachandran, K. N. Langmuir 2003, 19, 2947–2955. (22) Nisha, C. K.; Manorama, S. V.; Kizhakkedathu, J. N.; Maiti, S. Langmuir 2004, 20, 8468–8475. (23) Balomenou, I.; Bokias, G. Langmuir 2005, 21, 9038–9043. (24) Tsolakis, P.; Bokias, G. Macromolecules 2006, 39, 393–398. (25) Shovsky, A.; Varga, I.; Makuska, R.; Claesson, P. M. Langmuir 2009, 25, 6113–6121. (26) Sotiropoulou, M.; Cincu, C.; Bokias, G.; Staikos, G. Polymer 2004, 45, 1563. (27) Sato, A.; Choi, S. W.; Hirai, M.; Yamayoshi, A.; Moriyarna, R.; Yamano, T.; Takagi, M.; Kano, A.; Shimamoto, A.; Maruyama, A. J. Controlled Release 2007, 122, 209–216. (28) Sato, Y.; Moriyama, R.; Choi, S. W.; Kano, A.; Maruyama, A. Langmuir 2007, 23, 65–69. (29) Choi, S. W.; Kano, A.; Maruyama, A. Nucleic Acids Res. 2008, 36, 342–351. (30) Naderi, A.; Iruthayaraj, J.; Pettersson, T.; Makuska, R.; Claesson, P. M. Langmuir 2008, 24, 6676–6682. (31) P
eron, N.; Campbell, R. A.; Nylander, T.; Vareikis, A.; Makuska, R.; Gil
anyi, T.; M
esz
aros, R. J. Phys. Chem. B 2008, 112, 7410–7419. (32) Schwuger, M. J. J. Colloid Interface Sci. 1973, 43, 491. (33) da Silva, R. C.; Loh, W.; Olofsson, G. Thermochim. Acta 2004, 417, 295. (34) Dai, S.; Tam, K. C. J. Phys. Chem. B 2001, 105, 10759. (35) Bernazzani, L.; Borsacchi, S.; Catalano, D.; Gianni, P.; Mollica, V.; Vitelli, M.; Asaro, F.; Feruglio, L. J. Phys. Chem. B 2004, 108, 8960. (36) M
esz
aros, R.; Varga, I.; Gil
anyi, T. J. Phys. Chem. B 2005, 109, 13538.

11384 DOI: 10.1021/la901499x

Varga et al.

Figure 1. Structural formula of the PEO45MEMA and the HEMA units of the comb polymers.

polymers. Binding isotherms are used to monitor the variation of surfactant binding with the molecular architecture of the comb polymers. We present a detailed thermodynamic analysis of the binding isotherms in order to shed light to the possible mechanism of surfactant binding on nonionic bottle brush polymers. Furthermore, steady-state fluorescence probe (pyrene) measurements as well as static and dynamic light scattering measurements were performed to confirm the results of the binding isotherm measurements as well as to monitor the conformation changes of the bottle brush polymers in the formed complexes.

2. Experimental Details 2.1. Materials. The sodium dodecyl sulfate sample (Aldrich) was twice recrystallized from a 1:1 benzene/ethanol mixture. The critical micelle concentration (cmc) of the recrystallized SDS was 8.2 mM without added salt and 1.4 mM in 0.1 M NaBr, as determined by surface tension measurements at 25 C. The NaBr (Reanal) was analytical grade. Double distilled water was used for the preparation of the solutions. The poly(ethylene oxide) methyl ether methacrylate [(PEO)45MEMA] and its copolymer with 2-hydroxyethyl methacrylate [HEMA] were synthesized by free-radical copolymerization. Details of the synthetic procedure as well as the characterization of these polymers have been reported elsewhere.31 The synthesis resulted in close to random copolymers with high polydispersity.31 It should be noted that the prepared polymers lost their water solubility when the mole fraction of the (PEO)45MEMA monomer decreased below 0.3. The polymers were dialyzed against distilled water for 5 days, changing the solvent two times per day. The dialyzed solutions were concentrated with a rotating evaporator; then the polymers were extracted with chloroform and dried in a Petri dish at room temperature for 2-3 days. Finally, the products were dried in a vacuum oven at 60 C. Some physical characteristics of the HEMA:(PEO45)MEMA-X polymers (X = 100, 90, 75, 50, 30) are summarized in Table 1. 2.2. Methods. Potentiometric Measurements. The emf values of the Ag/AgBr/0.1 M NaBr/membrane/0.1 M NaBr, cPEO, cSDS/AgBr/Ag galvanic cell were determined by means of a Radelkis research pH-meter at 25.00 ( 0.05 C. The preparation of the surfactant selective electrode membrane was based on the substitution of a fraction of the chlorine atoms bound to the poly(vinyl chloride) backbone by trimethylamine.37 The details of functionalization of the PVC and the construction of the SDS selective membrane electrodes have been described elsewhere.37 10 cm3 medium (0.1 M NaBr and cp polymer) was placed into the measuring cell and titrated with a SDS stock solution (of the same NaBr and polymer concentration). The equilibrium emf values were read at each titration step. The measured emf values were converted into surfactant monomer concentration by means of a self-calibration procedure. In the absence of the polymer the emf vs log(cSDS) function was found to be linear up to the cmc (with a slope of 58.6 mV). In the presence of polymer the emf vs log(cSDS) function is linear only below the cac. This linear part was used as a calibration curve to calculate the normal potential and the response of the electrode in the case of each individual (37) Varga, I.; M
esz
aros, R.; Szak
acs, Z.; Gil
anyi, T. Langmuir 2005, 21, 6154– 6156.

Langmuir 2009, 25(19), 11383–11389

Varga et al.

Article Table 1. Characteristic Data of the Investigated HEMA:PEO45MEMA-X Polymers

polymera

Mwb (kDa)

Rgb (nm)

dhb (nm)

Rg/rh

nSDS/nPEOc

l/Rg,PEOd

(PEO45)MEMA-100 330 31.9 27.4 2.3 0.2 0.18 360 32.4 31.4 2.1 0.2 0.21 HEMA:(PEO45)MEMA-90 630 37.5 39.5 1.9 0.5 0.25 HEMA:(PEO45)MEMA-75 1100 45.0 50.1 1.8 0.7 0.38 HEMA:(PEO45)MEMA-50 1040 46.5 51.2 1.8 1.0 0.63 HEMA:(PEO45)MEMA-30 a The mole percentage of PEO45MEMA monomer in the comb polymers is given at the end of the name. b The weight-averaged molecular weight (Mw) and the radius of gyration (Rg) of the bottle brush polymers were determined by static light scattering. The hydrodynamic diameter (dh) was measured by DLS. c The average number of bound surfactant molecules per PEO side chain. d The average spacing of two neighboring PEO brushes along the polymer backbone (l) normalized to the radius gyration of an individual PEO side chain (Rg,PEO).

Figure 2. (a) The emf curves measured in the absence of and the presence of 0.2% linear PEO in 0.1 M NaBr and (b) the corresponding surfactant binding isotherms. potentiometric titration in order to determine the equilibrium surfactant monomer concentration ce above the cac. Using this approach, any shift in the normal potential and in the response of the electrode could be eliminated, which resulted in high accuracy and reproducibility ((0.01 mM) of the measurements.36 A typical set of measurements are shown in Figure 2A. Determination of the Surfactant Binding Isotherms. The exact binding isotherm of an ionic surfactant on a nonionic polymer (B(ce)) can be calculated from the expression38,39 c0 ¼ ce Æe æþBmpol þcmic y

ð1Þ

where B is the amount of surfactant bound to unit mass of the polymer, c0 is the total surfactant concentration, and ce is the equilibrium monomer surfactant concentration in a polymer-free (38) Gil
anyi, T.; M
esz
aros, R.; Varga, I. Prog. Colloid Polym. Sci. 2001, 117, 141–144. (39) Gil
anyi, T.; Varga, I.; M
esz
aros, R.; Filipcsei, G.; Zrinyi, M. Langmuir 2001, 17, 4764–4769.

Langmuir 2009, 25(19), 11383–11389

reference system. cmic is the concentration of the micelles in monomer unit, mpol is the polymer concentration (grams of polymer in unit volume), y = eΦ/kT is the dimensionless electric potential, and the brackets denote volume averaging. The factor Æeyæ takes into account that in the presence of macroions (ionic surfactant aggregates) the distribution of small ions (e.g., surfactant molecules) is not uniform in the solution.38-40 In 0.1 M NaBr due to the high ionic strength the Æeyæ factor becomes practically equal to one.39,40 Therefore, the difference between the total and the equilibrium surfactant concentrations gives the sum of the micelle and polymer bound surfactant concentrations (c0 - ce = mpolB(ce) þ cmic(ce)). If the measurements are restricted to surfactant concentrations below the cmc, then the binding isotherm can be determined directly (B(ce) = {c0 - ce}/mpol). Steady-State Fluorescent Measurements. Pyrene fluorescenct spectra were measured on a Hitachi F4500 fluorometer. The excitation wavelength was 320 nm, and the pyrene concentration was 5  10-7 M in each sample to minimize the probes effect on the polymer/surfactant interaction. The polymer concentration was 0.1% in the measured samples, and measurements were performed in 0.1 M NaBr. Static and Dynamic Light Scattering Measurements. Static and dynamic light scattering measurements were performed with a Brookhaven Instruments device, which consists of a BI-200SM goniometer and a BI-9000AT digital autocorrelator. A watercooled argon-ion laser, Lexel 95 model 2, was used as light source. The laser was used at a wavelength of 488 nm and emitted a vertically polarized light. The autocorrelator was used in a “multi tau” mode; i.e., the time axis was logarithmically spaced to span the required correlation time range. The autocorrelation functions were measured at an angle of 90 in 218 channels using a 100 μm pinhole size. The measured autocorrelation functions were analyzed by the CONTIN and the second-order cumulant methods. The samples were found to be polydisperse, having a wide monomodal size distribution. Despite the polydispersity of the samples, the mean hydrodynamic diameter (the first cumulant of the second-order cumulant expansion) was used for the presentation of the changes in the hydrodynamic size distribution. This could be done because the complex formation did not affect the size distribution function, as indicated by the CONTIN analysis. At finite concentrations and q values, an apparent diffusion coefficient is obtained (Dapp); by using the Einstein-Stokes equation, the apparent hydrodynamic radius of the complexes could be determined Rh ¼

kT 6πηDapp

ð2Þ

where k is the Boltzmann constant, T the absolute temperature, η the viscosity of the medium, Rh the hydrodynamic radius, and Dapp is the apparent diffusion coefficient. Static light scattering measurements were performed at 25 C at 11 different angles in the angular range 40 e θ e 140. The excess Rayleigh ratios (ΔR(θ,c)) were determined by standard (40) Gil
anyi, T.; Varga, I. Langmuir 1998, 14, 7397.

DOI: 10.1021/la901499x

11385

Article

Varga et al.

procedures using benzene as the reference with the absolute Rayleigh ratio (Rbenzene = 3.86  10-5 cm-1):
 Iðθ, cÞsolution -Iðθ, cÞsolvent nsolution 2 RðθÞbenzene ΔRðθ, cÞ¼ Iðθ, cÞbenzene nbenzene ð3Þ where n is the refractive index and I denotes scattering intensities corrected for the detector dead time reflections and scattering volume. The static light scattering data measured for the polymer solutions were analyzed by the Zimm method, yielding the weight-averaged molecular weight (Mw), the z-mean of the square of the radius of gyration (Rg,z2), and the second virial coefficient (A2) as the result of the analysis:
!
 16π2 n0 2 Rg 2 Kc 1 2 θ ¼ þ2Bc 1þ sin ΔRðθ, cÞ Mw 2 3λ0 2

ð4Þ

where λ0 is the wavelength of the laser beam in vacuum, n0 is the refractive index of the solvent, and K is the optical constant given as K = 4π2n02(dn/dc)2/NAλ04, with dn/dc being the refractive index increment of the polymer (determined with an interferometric differential refractometer, Optilab DSP from Wyatt Technology); c is the polymer concentration in mg/mL, and NA is Avogadro’s number. In the case of the bottle brush/SDS mixtures, the mean-squared radius of gyration ÆRg2æ was determined by the Guinier equation (5) from the angular dependence of the scattering intensity measured at finite polymer concentrations (0.1%): ln ΔRðqÞ¼ ln ΔRðq0 Þ -q2

ÆRg 2 æ 3

ð5Þ

where q is the scattering vector defined as q = (4πn0/λ0) sin(θ/2).

3. Results and Discussion Previous investigations indicated36 that the interaction of PEO and SDS has pronounced polymer molecular weight dependence. Namely, below a characteristic molecular weight (MPEO = 8000) the polymer chain can accommodate only a single surfactant aggregate. As the polymer chain length decreases, the number of interacting polymer segments involved in the surfactant aggregate also decreases. As a consequence, the formed surfactant aggregates become more similar to the free surfactant micelles: their critical aggregation concentration (cac) and surfactant aggregation number increase and approach that of a free micelle. If the PEO chain is short enough (Mw < 1000), PEO/SDS complex formation does not take place before the cmc of the surfactant is reached. The PEO side chains of the investigated bottle brush polymers have a molecular weight of 2000. PEO with this chain length is expected to interact with SDS with a relatively high cac. To test these expectations, the interaction of the nongrafted PEO and SDS was first investigated. In Figure 2A the emf curves measured in the absence and the presence of 0.2% PEO (Mw = 2000) are presented. For comparison, the emf curve measured for PEO with a molecular weight of 8000 is also presented. In agreement with previous results,36 the PEO with a molecular weight of 2000 interacts with SDS, and as expected the interaction starts at larger SDS concentration than in the case of the larger molecular weight polymer. The surfactant binding isotherms calculated from the measured emf curves are plotted in Figure 2b. The binding isotherm has a sigmoid shape in the case of the larger molecular weight PEO. 11386 DOI: 10.1021/la901499x

Figure 3. Surfactant binding isotherms measured for the HEMA:(PEO45)MEMA-X polymers.

Surfactant binding starts at the cac (1.15 mM), steeply increases, and levels off at ∼9 mM bound SDS/g PEO before the cmc is reached. In the case of the smaller molecular weight PEO (Mw = 2000) the cac is shifted to a considerably larger SDS concentration. Furthermore, the binding isotherm does not exhibit a sigmoid shape, indicating that the binding does not reach saturation before the cmc. Despite the lack of saturation, the lower molecular weight PEO binds the surfactant in a much larger amount. These results are in good agreement with previous investigations36 and serve as a reference to investigate how the grafting of the PEO chains to a nonionic polymer backbone affects their interaction with SDS. Figure 3a presents the surfactant binding isotherms measured for SDS on the investigated bottle brush polymers. The most striking feature of the measured binding isotherms is the almost complete lack of SDS binding compared to the free PEO, which seems to hold for all PEO brush densities in the investigated range. However, the surfactant binding is not zero on an absolute scale as demonstrated by Figure 3b. The presence of a critical aggregation concentration (cac) indicates that the surfactant binds in a cooperative manner (in the form of surfactant aggregates) to the brush polymers. Below this concentration there is no binding of surfactant to the polymers. As the surfactant concentration exceeds the cac, the surfactant starts to bind to the polymer. As the total surfactant concentration increases, the bound amount increases, which is accompanied by the increase of the equilibrium (monomer) surfactant activity. Finally, the equilibrium surfactant activity reaches the critical micelle concentration (cmc), and free micelles are formed in the solution. The fact that the measured binding isotherms do not have a sigmoid shape but increase Langmuir 2009, 25(19), 11383–11389

Varga et al.

Article

steeply before the cmc indicates that the surfactant binding does not reach saturation at the cmc. This can partly explain that the bottle brush polymers show almost 2 orders of magnitude smaller binding at the cmc than the PEO chains when they are not grafted to the methacrylate backbone. Finally, it should be noted that the binding capacity of the bottle brush polymers (their binding at cmc) systematically decreases as the graft density of the PEO brushes increases. In the case of the large molecular weight linear PEO molecules, the cac of the polymer/surfactant interaction was found independent of the molecular weight. However, as the PEO molecular weight becomes smaller than Mw ≈ 8000, the cac starts to increase. In the case of the bottle brush polymers an opposite trend can be observed. The surfactant starts to bind at a lower concentration (∼0.5 mM) to the bottle brush polymer than to the large molecular weight linear PEOs. According to the widely used pseudophase separation model of the polymer/surfactant interaction,41 the cac is related to the standard free energy change of the polymer/surfactant complex formation (ΔG0 = RT ln xcac). This implies that the driving force of the complex formation is considerably larger in the case of the brush polymers than in the case of the free PEOs. However, this conclusion has to be treated with caution, since the pseudophase separation model is strictly valid only in the case of infinite surfactant aggregation number. A more realistic approach for the description of the polymer/surfactant interaction is provided by the mass action model.41-43 The underlying assumption of this model is that surfactant aggregates of constant size bind on the polymer along the chain. A major advantage of this model is that it explicitly treats the surfactant aggregation number (the cooperativity of the binding). As it is demonstrated in Figure 4a, the aggregation number has a profound effect on the shape of the binding isotherm. For large aggregation numbers (m) the binding isotherm increases steeply. However, as m decreases, the binding isotherm becomes less steep, and as a consequence the surfactant concentration where binding “starts” (cac) also decreases. As is indicated by the mass action model, what remains invariant to the aggregation number at a given binding driving force is the concentration where half-saturation occurs. In practice, this means that while m is large enough, thus the cac is equal to the concentration of half-saturation within the experimental error, the cac provides a good measure of the binding driving force. However, if m is small (e.g., m < 30), then its use can give rise to the significant overestimation of the driving force value. The above arguments imply that in the case of the bottle brush polymer/SDS interaction the observed shift of the cac to lower surfactant concentration can be related either to the larger binding driving force or to the formation of considerably smaller surfactant aggregates in the brush polymers than along the linear PEO chains. The answer to this question can be given with the further analysis of the measured binding isotherms. As it has been shown42,44 if the binding approaches zero (B f 0), then the aggregation number is equal to the initial slope of the binding isotherm plotted in a log-log representation: log B ¼ m log xe þconst

ð6Þ

It should be emphasized that this approach relies only on the assumption that the surfactant binding is the result of surfactant (41) Evans, D. F.; Wennerstr€om, H. The Colloidal Domain - Where Physics, Chemistry, Biology and Technology Meets, 2nd ed.; Wiley-VCH: New York, 1999. (42) Gilanyi, T.; Wolfram, E. Colloids Surf. 1981, 3, 181. (43) Nagarajan, R. Colloids Surf. 1985, 13, 1. (44) Gil
anyi, T. J. Phys. Chem. B 1999, 103, 2085.

Langmuir 2009, 25(19), 11383–11389

Figure 4. (a) Effect of surfactant aggregation number on the onset of surfactant binding at constant binding driving force. The curves are calculated by the mass action model neglecting counterion binding.42 (b) A log-log representation of the measured surfactant binding isotherms.

aggregate formation along the polymer chain and remains valid in the initial binding range even if the aggregation number changes with increasing binding.44 The binding isotherms measured for the brush polymers and for the PEO2000 are plotted in a log-log scale in Figure 4b. As shown by the data in the figure the initial slope of the binding isotherms measured for the brush polymers are practically identical, and their value (∼4) is much smaller than the value observed for PEO2000 (∼40). Though the very small numerical value should be treated with care, these results clearly indicate that the surfactant aggregates formed in the brush polymers has a much smaller aggregation number than the aggregates formed on the linear PEO2000. As a consequence, the observed significant decrease of the cac values is related to the formation of the rather small aggregates; thus, the driving force of the surfactant binding could only be determined correctly from the surfactant concentration where half saturation occurs (x1/2). Unfortunately, the binding isotherms of the brush polymers do not reach saturation before the cmc, which means that the surfactant concentrations where half-saturation occurs, and thus the driving force values, cannot be estimated correctly from the experimental data. To confirm the results of the binding isotherm measurements, steady-state fluorescence measurements were also performed. Pyrene was used as a fluorescent probe, which preferentially solubilizes in the hydrophobic domains of the investigated system, and the ratio of the third to the first vibronic peaks (I3/I1) in the fluorescence spectra is sensitive to the polarity of the local environment of the probe molecule.45 Figure 5a shows the observed I3/I1 ratio as a function of the overall SDS concentration in the absence and in the presence of linear PEO. As it is indicated (45) Kalyanasundaram, K.; Thomas, J. K. J. Am. Chem. Soc. 1977, 99, 2039.

DOI: 10.1021/la901499x

11387

Article

Figure 5. Variation of the I3/I1 pyrene ratios as a function of SDS concentration at constant polymer concentration (1 g/dm3) in the presence of 100 mM NaBr: (a) measured for linear PEOs, Mw = 2000 and 8000; (b) measured for the HEMA:(PEO45)MEMA-X polymers. The horizontal dotted lines indicate the I3/I1 values measured after the first sharp rise for PEO2000 (orange), PEO8000 (green), and free micelles (blue).

by the figure, the introduction of a small amount of SDS into the investigated systems did not affect the observed I3/I1 ratio, which indicates that the surfactant molecules are present in monomer form in these systems. However, the initial plateau is followed by a sharp increase that takes place in a narrow surfactant concentration range. The surfactant concentrations, where the increase is observed, are in good agreement with the onset of surfactant binding (cac) found by surfactant selective electrode. This reveals that the probe does not affect considerably the investigated polymer/surfactant interaction. In the case of the pure surfactant, a second plateau is reached after the observed steep increase of the I3/I1 ratio. This indicates that the formed micelles provide a hydrophobic environment for the fluorescent probe that is not affected by the surfactant concentration increase above the cmc (micelles of similar size/ aggregation number form). However, in the presence of PEO the initial steep increase is followed by further moderate increase in a wide SDS concentration range. This shows that after the formation of the initial surfactant aggregates the pyrene molecules experience an increasingly hydrophobic environment with increasing SDS concentration. This observation can be interpreted in terms of the surfactant aggregation number measurements performed by fluorescence probe methods,46-48 trace probe electrolyte,40,49 and binding isotherm measurements.36 (46) Zana, R.; Lianos, P.; Lang, J. J. Phys. Chem. 1985, 89, 41. (47) Francois, J.; Dayantis, J.; Sabbadin, J. Eur. Polym. J. 1985, 21, 165. (48) van Stam, J.; Almgren, M.; Lindblad, C. Prog. Colloid Polym. Sci. 1991, 84, 13. (49) Varga, I.; Gilanyi, T. Prog. Colloid Polym. Sci. 2001, 117, 136.

11388 DOI: 10.1021/la901499x

Varga et al.

These measurements indicated that the aggregation number of the formed surfactant aggregates significantly increases as the equilibrium surfactant concentration increases between the cac and cmc. This implies that the larger aggregates provide a more hydrophobic environment for the probe molecules. This interpretation is also consistent with the fact that the I3/I1 ratio shows a larger initial increase in the presence of the PEO2000 than in the presence of the PEO8000, since previous aggregation number measurements36 indicated that the smaller the molecular weight of the PEO the larger surfactant aggregates form in this polymer molecular weight range. In Figure 5b, the measured I3/I1 ratios are plotted as a function of the overall SDS concentration in the presence of the bottle brush polymers. Similarly to the PEO/SDS systems, the measured curves start with an initial plateau that is followed by increasing I3/I1 values. This, in agreement with the presented binding isotherms, indicates the cooperative nature of the association. However, the observed I3/I1 values remain much smaller in this case than it was observed in the presence of the linear PEOs. This indicates that the surfactant aggregates formed in the bottle brush polymers provide much less hydrophobic environment for the probe molecule than the aggregates formed along the linear PEO chains. This observation is in good agreement with the presented binding measurements that indicated much lower aggregation numbers, thus the formation of less hydrophobic moieties in the case of the bottle brush polymers. Though the I3/I1 ratio shows a continuous increase with increasing SDS concentration, its value remains smaller than the values observed for PEO/SDS systems, indicating that the size of the surfactant aggregates remains small compared to the aggregates formed in the linear PEO/SDS systems. Finally, it should also be noted that the curves measured for the different brush density polymers practically coincide with each other, implying that similar surfactant aggregates form in the investigated bottle brush polymers independently of their graft density. This observation is also in good agreement with the results of the binding study that indicated the formation of similar small surfactant aggregates in each of the bottle brush polymers. To rationalize the brush density independence of the formed surfactant aggregates, we made an estimation of the average spacing of the PEO brushes along the methacrylate backbone. This was done by assuming that the PEO brushes are distributed uniformly along the backbone, and the distance was calculated along the C-C bonds using a bond length of 1.54 A˚. Finally, the calculated average distance between the PEO brushes was normalized by the radius of gyration of an individual PEO side chain (16 A˚50). The calculated values are summarized in Table 1. The calculations indicate that the average spacing of the PEO45MEMA brush monomers along the backbone is much smaller in the investigated polymers than would be required to avoid the overlap of the PEO coils. As a consequence, the PEO side chains are expected to swell into a more brushlike conformation and form a corona of high PEO segment concentration around the methacrylate backbone that provides only a limited space for the surfactant aggregate formation independently of the brush density. This further implies that only a limited number of small surfactant aggregates can form, presumably in the least crowded regions of the PEO corona of the brush polymers. As the graft density and consequently the PEO segment density of the corona decrease, the number of surfactant aggregates slowly increases, giving rise to a moderate increase of the surfactant binding capacity. However, the binding capacity remains much (50) Kawaguchi, S.; Imai, G.; Suzuki, J.; Miyahara, A.; Kitano, T.; Ito, K. Polymer 1997, 38, 2885.

Langmuir 2009, 25(19), 11383–11389

Varga et al.

Article

on three selected bottle brush polymers (100, 75, and 30% brush densities) as a function of the SDS concentration. The results of the dynamic light scattering measurements are plotted in Figure 6a. As is indicated by the figure, the surfactant binding does not have a noticeable effect on the hydrodynamic size of the bottle brush polymers, which is in agreement with the stiff nature of the polymers. To gain further insight into the conformation changes associated with the surfactant binding, the radii of gyration of the three selected bottle brush polymers were determined at three selected SDS concentrations (0 mM, to characterize the polymer; 0.4 mM, which is just before the cac; 1.4 mM, which that is close to maximum binding). The radii of gyration were determined from the measured Gunier plots. As a characteristic example, the Gunier plots determined for the least grafted bottle brush polymer (30%) are plotted in Figure 6b. As is indicated in the figure, the introduction of a small amount of SDS (cSDS < cac) does not have any effect on the scattering of the bottle brush polymer. However, when the system contains enough surfactant and binding takes place (cac < cSDS < cmc), a significant increase of the scattering intensity can be observed, while the slope of the Gunier plot (the radius of gyration of the polymer) remains practically unchanged. These results clearly indicate that despite the surfactant binding that gives rise to the increasing scattering of the system, the average mass distribution of the polymer remains unchanged due to its stiff nature. Figure 6. (a) Hydrodynamic diameter (dh) of the HEMA:(PEO45)MEMA-X polymers (X = 100, 75, and 30%) as a function of SDS concentration. (b) Guinier plots for the HEMA:(PEO45)MEMA-30 polymer in the presence of 0, 0.4, and 1.4 mM SDS. Solid lines represent the least-squares fits.

lower than that of a free PEO2000 chain even in the case of the least grafted brush polymer (30%) in the investigated range. This is clearly indicated by the fact that the average number of bound surfactant per a PEO brush is only ∼1 even in the case of the least grafted (30%) bottle brush polymer (see Table 1). The characteristic ratios, F = Rg/Rh, evaluated from the light scattering data of the bottle brush polymers are summarized in Table 1. The Rg/Rh ratio provides information about the global shape of the scattering particles. It can be shown51 that in the case of a random coil structure F can take a value from 1.3 to 1.5, whereas F is expected to be larger than 2 for stiff rodlike particles. As it is indicated by the calculated F values, the investigated bottle brush polymers have a stiff rodlike structure. This observation is in agreement with previous small-angle neutron scattering (SANS) and small-angle X-ray scattering measurements, which indicated that PEO45MEMA:METAC-X bottle brush polymers (where METAC denotes methacryloxyethyltrimethylammonium chloride) adopt rodlike conformation in the solution phase.20,52 To investigate how the rodlike conformation of the bottle brush polymers are affected by the surfactant binding, we performed static and dynamic light scattering measurements (51) Konishi, T.; Yoshizaki, T.; Yamakawa, H. Macromolecules 1991, 24, 5614– 5622. (52) Dedinaite, A.; Bastardo, L. A.; Oliveira, C. L. P.; Skov Pedersen, J.; Claesson, P. M.; Vareikis, A.; Makuska, R. Proc. Baltic Polym. Symp. 2007, 112– 117.

Langmuir 2009, 25(19), 11383–11389

4. Conclusions The effect of graft density on the interaction of nonionic bottle brush polymers (HEMA:(PEO45)MEMA-X) with sodium dodecyl sulfate was investigated by means of surfactant binding isotherm measurements as well as steady-state fluorescence probe (pyrene) measurements and static and dynamic light scattering. Our results indicate that due to the steric crowding of the PEO chains grafted to the polymer backbone, the surfactant binding becomes significantly suppressed. The amount of bound surfactant at the cmc is almost 2 orders of magnitude smaller than the binding on a linear PEO and slightly increases with decreasing graft density. The surfactant binds in cooperative fashion to the bottle brush polymers, though the formed surfactant aggregates have a much smaller aggregation number than in the case of the linear PEOs, which gives rise to a considerable decrease of the critical aggregation concentration (cac). As it is indicated by the light scattering data, the surfactant binding does not have a considerable effect on the rodlike conformation of the bottle brush polymers. Acknowledgment. This work is supported within the 6th European Community RTD Framework Program by the Marie Curie Research Training Network SOCON (MRTN-CT-2004512331) and by the Marie Curie Intra-European Fellowships PE-NANOSTRUCTURES (24997) as well as by the Hungarian Scientific Research Fund, OTKA K68027 and K68434. R.M. and I.V. are Bolyai J
anos fellows of the Hungarian Academy of Sciences, which is gratefully acknowledged. P.C. acknowledges financial support from the Swedish Research Council, VR.

DOI: 10.1021/la901499x

11389