Surface Chemistry and Rheology of Polysulfobetaine-Coated Silica

An Asylum Research (Santa Barbara, CA) MFP-. 3D atomic force microscope was used to measure the forces between a silicon nitride tip and a borosilicat...
0 downloads 0 Views 169KB Size
Download PDF
Langmuir 2007, 23, 7587-7593

7587

Surface Chemistry and Rheology of Polysulfobetaine-Coated Silica Pierre Starck,† Wade K. J. Mosse,† Nathan J. Nicholas,† Marisa Spiniello,† Johanna Tyrrell,† Andrew Nelson,‡ Greg G. Qiao,† and William A. Ducker*,† Department of Chemical and Biomolecular Engineering, UniVersity of Melbourne, ParkVille, Victoria 3010, Australia, and Australian Nuclear Science and Technology Organization (ANSTO), New Illawarra Road, Menai, New South Wales 2234, Australia ReceiVed March 6, 2007. In Final Form: April 26, 2007 We have measured the viscosity of suspensions of colloidal silica particles (d ) 300 nm) and the properties of silica surfaces in solutions of a polymer consisting of zwitterionic monomer groups, poly(sulfobetaine methacrylate), polySBMA. This polymer has potential use in modifying surface properties because the polymer is net uncharged and therefore does not generate double-layer forces. The solubility of the polymer can be controlled and varies from poor to good by the addition of sodium chloride salt. Ellipsometry was used to demonstrate that polySBMA adsorbs to silica and exhibits an increase in surface excess at lower salt concentration, which is consistent with a smaller area per molecule at low salt concentration. Neutron reflectivity measurements show that the adsorbed polymer has a thickness of about 3.7 nm and is highly hydrated. The polymer can be used to exercise considerable control over suspension rheology. When silica particles are not completely covered in polymer, the suspension produces a highly viscous gel. Atomic force microscopy was used to show this is caused by bridging of polymer between the particles. At higher surface coverage, the polymer can produce either a high or very low viscosity slurry depending on the sodium chloride concentration. At high salt concentration, the suspension is stable, and the viscosity is lower. This is probably because the entrainment of many small ions renders the polymer film highly hydrophilic, producing repulsive surface forces and lubricating the flow of particles. At low salt concentrations, the polymer is barely soluble and more densely adsorbed. This produces less stable and more viscous solutions, which we attribute to attractive interactions between the adsorbed polymer layers.

Introduction Polymers and surfactants are employed to adsorb onto solidliquid interfaces in a number of applications, including paints, ceramics, pharmaceuticals, and natural lubricants. The adsorption controls the interparticle forces, stability, and rheology of the colloidal dispersions. Attraction of particles leads to aggregation, sedimentation, and high viscosities, whereas repulsion results in stable suspensions that have low viscosities. The rheology also depends on the volume fraction of solids, the size and shape of the particles, and the interparticle forces. In this study we describe how a linear zwitterionic polymer adsorbs to silica and affects the stability and rheology of silica particles in aqueous solution. As shown in Figure 1, each monomer of the poly(sulfobetaine methacrylate), (polySBMA), possesses both a positive and a negative charge (i.e., is zwitterionic), but the polymer as a whole bears no net charge. The strongly dipolar structure of the zwitterionic lateral groups leads these polymers to display unusual solubility behavior in aqueous solution known as the antipolyelectrolyte effect. In contrast to conventional polyanions and cations, polysulfobetaines are known to undergo an increase in solubility upon the addition of salt.1,2 In addition, the polymer solutions are known to increase in viscosity with addition of salt, which has been interpreted as an increase in molecular volume.1 The exact mechanism of the antipolyelectrolyte effect has not been demonstrated, but a reasonable hypothesis is that the salt screens the attractive forces between the positive ions on one monomer and negative ions on another. * To whom correspondence should be addressed. E-mail: wducker@ unimelb.edu.au. † University of Melbourne. ‡ Australian Nuclear Science and Technology Organization (ANSTO). (1) Salamone, J. C.; Volksen, W.; Olson, A. P.; Israel, S. C. Polymer 1978, 19, 1157-1162. (2) Hart, R.; Timmerman, D. J. Polym. Sci. 1958, 28, 638-640.

Figure 1. Polymer synthesis: (A) monomer SBMA and (B) polymer polySBMA.

However, at least in the case of polySBMA, light scattering is consistent with a roughly constant polymer radius when the viscosity increases due to added salt.3 Our interest in the zwitterionic polymers stems from two areas: (1) the ability to produce steric and hydration forces without electrostatic forces (these short-range hydration and steric repulsions may allow lubrication of particles and surfaces)4-6 and (2) the antipolyelectrolyte effect potentially offers a degree of control of interfacial and colloidal properties that has not yet been explored. Dobrynin et al. have recently published a review on the properties of polyampholyte solutions and their interactions with polyelectrolytes and surfaces.7 According to theory, the adsorption (3) Kato, T.; Kawaguchi, M.; Takahashi, A.; Onabe, T.; Tanaka, H. Langmuir 1999, 15, 4302-4305. (4) Ducker, W. A.; Clarke, D. R. Colloids Surf., A 1994, 93, 275-292. (5) Ducker, W. A.; Luther, E. P.; Clarke, D. R.; Lange, F. E. J. Am. Ceram. Soc. 1997, 80, 575-583. (6) Briscoe, W. H.; Titmuss, S.; Tiberg, F.; Thomas, R. K.; McGillivray, D. J.; Klein, J. Nature 2006, 444, 191-194. (7) Dobrynin, A. V.; Colby, R. H.; Rubinstein, M. J. Polym. Sci., Part B-Polym. Phys. 2004, 42, 3513-3538.

10.1021/la700642d CCC: $37.00 © 2007 American Chemical Society Published on Web 06/05/2007

7588 Langmuir, Vol. 23, No. 14, 2007

of polyampholytes is due to the polarization of chains in the external electric field created by the charged surface.8-13 In many polyampholytes, polarization can occur on the scale of the entire polymer through conformational change in which anionic monomers separate from cationic monomers. For polySMBA the polarization is limited by the fact that each monomer unit has both a cationic and an anionic unit. Polymers and surfactants with zwitterionic groups have been studied previously by a number of groups. Ducker et al. have investigated the effect of (zwitterionic) choline surfactants on interparticle forces, rheology, and particle packing of silicon nitride slurries near the isoelectric point. Surface forces measurements showed that adsorption of surfactant produced a shortrange repulsive force that dramatically reduced the friction between the solids.4 The repulsive force also greatly reduced the viscosity of slurries of particles.5 Several groups have investigated the adsorption of zwitterionic polymers. Kato et al. found that the surface excess of sulfobetaine polymer on silica surfaces decreases as a function of salt concentration.3 Sulfobetaine and carboxybetaine, polyCBMA, polymers have been used as superlow fouling agents.14,15 In these studies, polySBMA and polyCBMA brushes were polymerized from a self-assembled monolayer (SAM) terminated with initiators on a gold and silica surfaces via the surface-initiated atom-transfer radical polymerization (ATRP) method. An enzyme-linked immunosorbent assay (ELISA) was used to measure fibrinogen adsorption on these surfaces. These results showed that these zwitterionic polymers resist nonspecific protein adsorption as effectively as poly(ethylene glycol) or poly(methacryloyloxyethylphosphorylcholine). Moreover, the formation of the initial SAM is important to control polymerization and to achieve very low nonspecific protein adsorption. In the current work, we characterize the adsorption of polySBMA to silica using ellipsometry and neutron reflectivity. We then describe measurement of the surface forces using atomic force microscopy (AFM) and then examine the stability and rheology of silica suspensions. For experimental simplicity, we use a variety of silica-like materials. Silicon wafers with a thin layer of native oxide are used for ellipsometry and neutron reflectivity studies. In the AFM measurements we use silicon nitride tips, which also form a native silicon oxide layer, borosilicate glass slides, and borosilicate glass particles. For the study of particles, we use silica particles directly. Experimental Section Materials. All solutions were prepared in water purified by an Easypure UV system (Barnstead, Dubuque, IA) with charcoal, deionizing, UV light, and a 0.2 µm filter stages. The water has a specific resistivity of 18.2 MΩ cm and a surface tension of 72 mJ/ m2. Sodium chloride (NaCl, 99.99%) supplied by Aldrich was baked overnight in an oven at 500 °C to remove organic contamination. 1,3-Propane sultone (98%), ammonium persulphate (ACS grade, 99+%), sodium bromide (NaBr, 99+%), and acetone were also (8) Dobrynin, A. V.; Rubinstein, M.; Joanny, J. F. Macromolecules 1997, 30, 4332-4341. (9) Dobrynin, A. V.; Rubinstein, M.; Joanny, J. F. J. Chem. Phys. 1998, 109, 9172-9176. (10) Dobrynin, A. V.; Obukhov, S. P.; Rubinstein, M. Macromolecules 1999, 32, 5689-5700. (11) Dobrynin, A. V. Phys. ReV. E 2001, 6305. (12) Dobrynin, A. V.; Zhulina, E. B.; Rubinstein, M. Macromolecules 2001, 34, 627-639. (13) Zhulina, E.; Dobrynin, A. V.; Rubinstein, M. Eur. Phys. J. E 2001, 5, 41-49. (14) Zhang, Z.; Chen, S. F.; Chang, Y.; Jiang, S. Y. J. Phys. Chem. B 2006, 110, 10799-10804. (15) Zhang, Z.; Chao, T.; Chen, S. F.; Jiang, S. Y. Langmuir 2006, 22, 1007210077.

Starck et al. purchased from Aldrich and used without further purification. 2-(Dimethylamino)ethyl methacrylate was supplied by Aldrich and was purified by passing through a plug of inhibitor remover (Aldrich). Sulfobetaine methacrylate (SBMA) monomer was synthesized using a modified method described by McCormick et al.16 Sodium hydroxide (NaOH, 99%), acetonitrile (ACN, 99+%), and methanol (MeOH, 98%) were purchased from Merck, hydrochloric acid was from BDH, and 1,3-dinitrobenzene from Ajax Chemicals. The silica micropowder used in this study (Fuso Chemical Co. Ltd., Japan) consisted of nearly monodisperse spherical particles of diameter 0.30 µm (range: 0.27-0.34 µm), with a density of 2200 kg/m3, a maximum moisture of 0.5%, and a surface area of 12.15 ( 0.05 m2/g. The average hydrodynamic diameter of the silica particles was determined using dynamic light scattering (Malvern Instruments, HPPS). Polymer Synthesis. SBMA monomer was prepared from 2-(dimethylamino)ethyl methacrylate and 1,3-propane sultone in ACN, with a small amount of 1,3-dinitrobenzene added. The mixture was blanketed with argon and heated at 50 °C for 18 h. After being cooled, the zwitterionic product was filtered, thoroughly washed several times with cold acetonitrile, and dried in vacuo to give the monomer as a white powder in quantitative yield. PolySBMA polymer was synthesized by conventional free-radical polymerization (Figure 1). Zwitterionic monomer, SBMA (17.33 g, 0.062 mol), was dissolved in water (175 mL) with ammonium persulfate (APS, 0.45 g) added as the radical initiator. The mixture was flushed with argon (30 min) and heated at 60 °C for 18 h. After being quenched with water, the volume was reduced by 75% and the residue was then precipitated twice into acetone/MeOH (10:1). The product was filtered and dried in vacuo to give polySBMA as a white solid with a molecular weight of Mw ) 220 000 g/mol, a polydispersity of ∼2, and a density d ) 1.34 ( 0.05 g/cm3. The polymer was then dissolved in water (5.1 g of polySBMA in 40 mL of H2O) and neutralized to pH 7.3 with 1 M sodium hydroxide. The subsequent solution was again precipitated in acetone/MeOH (10:1). Aqueous size exclusion chromatography (SEC) was used to estimate the molecular weight of polySBMA. This was based on a Universal Calibration of linear poly(ethylene glycol) polymer standards using Viscotek OmniSEC 4.2 software. The SEC set up comprised a Waters 510 HPLC pump, 717 Plus Autosampler, 410 differential refractometer, and a Viscotek T50A differential viscometer. The mobile phase was 80% 0.5 M aqueous NaBr and 20% ACN with three Waters Ultrahydrogel columns of 250 Å, 2000 Å, and a blended pore size column, at a flow rate of 1 mL/min. SEC analysis and UV-vis spectroscopy, Varian, were used to determine the equilibrium concentration of polymer once steric stabilization was achieved. The density of the polymer was established in 50 mM NaCl using Anton Paar DMA 5000 densitometer. Methods. Ellipsometry. A phase modulation ellipsometer (Beaglehole Industries, Wellington, New Zealand) with a 632.8 nm laser was used to observe the adsorption of polySBMA to silicon wafers by measurement of Im(r) (the imaginary component of the reflectivity) at the Brewster angle.17,18 The surface of the silicon was cleaned in oxygen plasma for 30 s at 300 mTorr (Plasma Prep II, SPI Supplies, West Chester, PA); this resulted in the formation of a thin silica layer on the surface of the wafer. The wafer was then mounted on a platform encased by a cylindrical glass cell and aligned in a salt solution. After the Brewster angle was determined, the polymer solution was injected and the change in Im(r) monitored over time. Adsorption was deemed to have reached equilibrium when the signal was constant for at least 15 min. Adsorption was measured at 23 ( 0.5 °C. RefractiVe Index Measurements. The refractive index increment was obtained by measuring the refractive index of polySBMA (16) Johnson, K. M.; Poe, G. D.; Lochhead, R. Y.; McCormick, C. L. J. Macromol. Sci.-Pure Appl. Chem. 2004, A41, 587-611. (17) Mao, M.; Zhang, J. H.; Yoon, R. H.; Ducker, W. A. Langmuir 2004, 20, 1843-1849. (18) Russev, S. C.; Arguirov, T. V.; Gurkov, T. D. Colloids Surf., B 2000, 19, 89-100.

Polysulfobetaine-Coated Silica solutions (1-50 mg/mL) in aqueous 50 mM NaCl using a Bellingham and Stanley Limited refractometer. The refractometer was calibrated with ethanol and water. The refractive index increment was 0.14 L/g. At each concentration, the refractive index was measured five times to obtain an accurate average reading. Neutron Reflectometry. Neutron reflectivity measurements were performed in order to determine the thickness, d, and hydration of the polymer layer adsorbed to silica. Experiments were carried out at the HIFAR facility of the Australian Nuclear Science and Technology Organization (ANSTO), Australia. We measured the change in intensity of the specularly reflected beam as a function of incident angle (θ). Neutron reflectivity data were collected over the Q range 0.007-0.3 Å-1 with a ∆Q/Q resolution of 5% on the X172 reflectometer using thermal neutrons (λ ) 2.43 Å).19 A silicon wafer with a native oxide layer mounted in a fluid cell was exposed to a 2 mg/mL polySBMA solution prepared at a salt concentration of 50 mM NaCl and pH 6. The reflectivity was measured in both D2O with a scattering length density, Fsolv) 6.36 × 10-6 Å-2 and H2O (Fsolv ) -0.57 × 10-6 Å-2) solutions. The use of multiple contrasts results in multiple independent data sets, thus improving the determination of the properties of the layer. For radiation reflected at angles below the critical angle for total external reflection (θc), a plateau in intensity exists where R ) 1. For incident angles greater than θc, R rapidly decreases in intensity (as ∼Qz-4 where Q ) 4π sin θ/λ). Zeta Potential Measurements. The zeta potential of silica particles as a function of pH were performed using a zetasizer (Malvern Instruments). For those measurements, silica particles were prepared at 5 vol % and the pH was adjusted using 0.1 M HCl and NaOH solutions. The surface area of silica particles was determined using a surface area analyzer BET gas adsorption. AFM Studies. An Asylum Research (Santa Barbara, CA) MFP3D atomic force microscope was used to measure the forces between a silicon nitride tip and a borosilicate glass plate in aqueous salt and polymer solutions. AFM measures the change in endslope of a cantilever while the cantilever is translated normal to the solidliquid interface. The translation is measured using a linear variable differential transducer (LVDT). We convert the endslope to the deflection as described previously.20 The separation is the sum of the deflection and the translation. The AFM probes (NP, Veeco, Santa Barbara, CA, silicon nitride, square pyramid tip, nominal stiffness of 0.12 N/m) were cleaned under ultraviolet light prior to use. In experiments where a colloid probe was used, a 20 µm diameter borosilicate glass sphere was glued to a 0.58 N/m cantilever. The borosilicate glass slides (12-544-12, Fischer Scientific) have an rms roughness of 0.26 nm over a 5 µm area and were cleaned in oxygen plasma for 30 s using a Plasma Prep II (SPI Supplies, West Chester, PA) prior to use. Fluid exchange made use of a custom designed open fluid cell, consisting of an Asylum cantilever mount modified to hold two lengths of PEEK tubing (Valco Instruments, Houston, TX). This enabled the exchange of fluids within a liquid capillary formed between the surface and the cantilever mount. Determination of the Equilibrium Polymer Concentrations in Silica Dispersions. Owing to the large area of the solid-liquid interface in suspensions, the equilibrium concentration of polymer is less than the starting concentration because of adsorption. The equilibrium polymer concentration was determined using UV-vis and SEC techniques. For both techniques a calibration curve was first plotted using different diluted polymer concentrations (from 0.2 to 1 mg/ mL polySBMA). Using the UV-vis technique, the absorbance was recorded at a single wavelength of 212 nm and then plotted as a function of polymer concentrations, whereas using SEC, the total refractive index area of the peak calculated between 25.73 and 29.54 mL of retention volume baseline was established to plot the calibration curve. After equilibrium, suspensions of 3 and 5 wt % polySBMA/g (19) James, M.; Nelson, A.; Schulz, J. C.; Jones, M. J.; Studer, A. J.; Hathaway, P. Nucl. Instr. Methods Phys. Res., Sect. A 2005, 536, 165-175. (20) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Langmuir 1992, 8, 18311836.

Langmuir, Vol. 23, No. 14, 2007 7589

Figure 2. Adsorption of polySBMA from solution to silica. SiO2 were centrifuged for 5 min at 7500g using a Avanti 30 centrifuge (Beckman), after which the supernatant were diluted (10 and 100 times, respectively) and analyzed using SEC and UV-vis spectroscopy. In all experiments, a known ratio of polymer to particle weight was used. Rheology. Silica dispersions were prepared at 35 vol % using electrolyte solution (20, 50, 100 mM NaCl). The suspensions were sonicated for 8 min with a high-intensity sonic probe (Branson Sonifier 250) and then put at rest for at least 24 h, after which the pH was adjusted to 6 by using 0.1 M NaOH. Samples of the equilibrated suspensions (8 mL ) 11.36 g) were added into glass vials that contained 2 mL of polySBMA solution of equivalent pH and ionic strength required to produce the final desired polymer concentration. Samples were prepared at 0.01, 0.1, 0.3, 0.6, 1, 3, and 5 wt % of polySBMA/g of silica. The final volume fraction of silica was 28 vol %. The mixtures were shaken for 24 h and then transferred to the rheometer. Rheometric Scientific SR5 and ARES rheometers with the parallel plate geometry (PP40) at a gap of 1.5 mm were used to study the flow behavior of the mixtures. The samples were presheared for 2 min at a constant stress (or rate) and then left at rest for a further 2 min before each shearing experiment. The apparent viscosity was measured at a constant temperature of 25 °C for applied stresses (or rate) ranging between 0.025 and 20 Pa (10-1000 s-1). A delay time of 60 s elapsed during each applied stress to achieve steady-state flow conditions. A solvent trap was used to minimize evaporation.

Results and Discussion Solubility. The polySBMA was soluble at NaCl concentrations greater than about 20 mM when the polymer concentration was in the range 1-100 mg/mL. In 20 mM NaCl solutions, the polymer is slightly insoluble, which makes ellipsometry and AFM experiments problematic, but still allows measurement of particle rheology. The surface chemistry experiments were performed in the vicinity of 1 mg/mL polymer, which is equivalent to about 4 mM monomers. Adsorption to Flat Plates. The surface excess of PolySBMA is about 1-2 mg/m2 at the interface between silica and 0.1-2 mg/mL polymer solutions. The results are shown in Figure 2 and are consistent with those of Kato et al.3 Increasing NaCl concentration leads to a decrease in surface excess, the opposite trend to that normally observed for charged polymers. We hypothesize that the cause of the decrease in surface excess on addition of salt is that each polymer molecule occupies a larger surface area, and therefore, the total number of molecules at the interface is lower. At a fixed salt concentration, the amount adsorbed increases as the pH decreases to 2 (Figure 3). The

7590 Langmuir, Vol. 23, No. 14, 2007

Starck et al.

Figure 3. Effect of pH on the adsorption of polySBMA to silica from a 1 mg/mL polySBMA at 100 mM NaCl solution.

sulfonate group has a pK of 1.6,21 so the polymer becomes increasingly net cationic as the pH approaches 2. We interpret the increase in adsorption as being due to net attractive interactions between the cationic groups of the polymer and the silica. Below pH 2, the silica is no longer anionic, so the adsorbed amount falls. Film Thickness on Flat Plate. Neutron reflectivity measurements are useful for determining the thickness of adsorbed layers. The measured neutron reflectivity from the interface between silicon (native oxide) and a 2 mg/mL aqueous polySBMA in a 50 mM NaCl solution is shown in Figure 4B. The symbols represent the observed reflectivity data, while the solid lines are the calculated reflectivity profiles determined for the proposed structural model. The model used the Parratt formalism22 and was simulated with the Motofit analysis package.23 The data were fitted as log(R) vs Q with corrections for a linear background resolution smearing and Gaussian roughness at each interface. The model includes a native SiO2 layer (3 ( 1 Å thick) between the Si wafer and the polymer film. The density of a bulk solution of polySBMA was measured at d ) 1.34 ( 0.05 g/cm3, resulting in a calculated scattering length density of 1.03 × 10-6 Å-2. Figure 4A shows that the reflectivity of the polymer-coated wafer is significantly less than that of the bare wafer in the region 0.07-0.12 Å-1, which allows us to determine the thickness. An excellent fit to the measured data was obtained using a singlelayer model, based on polymer film of uniform density. The best fit was for a film of polySBMA with a thickness layer of δ ) 3.7 ( 0.2 nm that is 89% ( 2% water. Particle Zeta Potential. The primary determinant of colloidal properties in aqueous solution is usually the surface charge. Figure 5 shows the zeta potential for the 300 nm diameter silica particles in deionized water. The natural pH of the suspensions of silica particles was 6.5. We added either NaOH or HCl to alter the pH to obtain measurements in the range of 2-10. As is usual for silica, the zeta potential is negative over the range of pH 3-10, which indicates negative charge sites on the particles. Table 1 shows the effect of NaCl concentration on the zeta potential at pH 6. The zeta potential decreases as the salt concentration (21) Bordwell, F. G.; Algrim, D. J. Org. Chem. 1976, 41, 2507-2508. (22) Parratt, L. G. Phys. ReV. 1954, 95, 359-369. (23) Nelson, A. http://motofit.sourceforge.net/, 2005.

Figure 4. (A) Reflectivity of a silicon wafer in D2O with a layer of polySBMA adsorbed from 50 mM NaCl at pH 6 and without the polymer. (B) Reflectivity curves of the polySBMA adsorbed layers at pH 6 and 50 mM NaCl in two different contrasts.

Figure 5. Zeta potential of silica particles as a function of pH in H2O.

increases. The addition of 1 mg/mL polySBMA to the 50 and 100 mM NaCl solutions slightly decreases the zeta potential. Surface Forces. The forces between an AFM tip or a colloidal particle and a surface were measured in order to interpret the rheological measurements. We would like to use the AFM to obtain the thickness of the polymer film, but in general, this technique does not allow measurement of the thickness of the layers because the zero of separation is not known. We used evanescent wave-AFM, EW-AFM24 to measure the separation between the tip and sample, which showed that the AFM tip

Polysulfobetaine-Coated Silica

Langmuir, Vol. 23, No. 14, 2007 7591

Figure 6. Force on approach between an AFM tip and a borosilicate glass surface in salt solution, with and without 2 mg/mL polySBMA. Inset shows forces in polymer solution on a ln scale. Table 1. Zeta Potential of Silica Particles as a Function of Salt Concentration at pH 6.5 NaCl concn (mM)

zeta potential (mV)

0 20 50 50 + polySBMA 100 100 + polySBMA

-49 -39 -23 -18 -13.5 -10.6

could penetrate though the polymer layer, and therefore, it was valid to use conventional AFM to measure the film thickness. Figure 6 shows the forces in salt solution, both in the presence and the absence of polymer. In the absence of polymer, there is a long-range repulsive force between the surfaces. This is consistent with a double-layer force. Double-layer forces decay approximately exponentially, with a characteristic decay length (Debye length) that decreases as the salt concentration increases. The measured decay length in 50 mM salt was 1.5 nm, compared to the Debye length of 1.35 nm. When the salt concentration is increased to 500 mM, the Debye length was shorter (measured, 0.5 nm; theoretical, 0.43 nm). In addition, the magnitude of the double-layer force decreases, which showed a decrease in potential with salt concentration, consistent with the zeta potential measurements. When polySBMA (2 mg/mL) is added to the salt solutions, the range of repulsive force increases. Remembering that the zeta potential hardly changes on addition of polymer, this effect is attributed entirely to steric repulsion (caused by the presence of the polymer) that has a longer range than the double-layer force. The force is still roughly exponential, with a decay length of 1.8 nm (see inset to Figure 6). The neutron reflectivity data indicates a 3.7 nm average film thickness on each surface in 50 mM salt solution. The observed forces are consistent with a similar film thickness, as the forces approach zero at a little more than twice the average film thickness. The forces in polymer solution are quite insensitive to the concentration of salt. We had expected to observe an increased range of the force in high salt resulting from chain swelling, but this was not observed. Because the polymer barely affects the zeta potential, the steric force contribution can be estimated from the difference between the double-layer force measured in the absence of the polymer and the force in the presence polymer. Clearly, the polymer has a much larger effect on the forces in 500 mM salt than in 50 mM salt. A single polymer chain can potentially adsorb onto two surfaces, producing an attractive force when the surfaces are (24) McKee, C. T.; Mosse, W. K. J.; Ducker, W. A. ReV. Sci. Instrum. 2006, 77.

Figure 7. AFM force measurements upon withdrawal between two borosilicate surfaces in 0.03 g/L polySBMA in 50 mM NaCl. After 1 s in contact the withdrawal is continuous; after 30 s in contact the sphere motion is discontinuous, with jumps indicated by the arrows. The zero of separation was defined to be where there was a steep repulsive force. 1 s data increased by 0.2 nN for clarity.

separated. This effect was studied using AFM, but with colloidal silica spheres attached to the AFM probe (colloid probe AFM).25 Polymer-bridging forces show a characteristic feature in AFM experiments where the attractive force decreases in magnitude abruptly when the force exceeds the attachment strength. The presence of these events was found to depend on two factors: the concentration and the time surfaces were in contact. No bridging forces were observed at or below a concentration of 0.01 mg/mL. At or above 0.03 mg/mL we were able to routinely observe bridging forces (see Figure 7), but only when the solids were left in contact for some time. For example, when the surfaces were in contact for less than 1 s, we seldom saw bridging, and when they were left in contact for 30 s, bridging was routinely observed. The dependence of bridging on the time in contact is reasonable considering that a dangling polymer loop or tail will take time to pass through a conformation that is near an attractive site on the other solid. Adsorption of polySBMA to Particles. The chemical potential of the polymer depends on the equilibrium concentration of polymer rather than the initial concentration. The adsorption of polymer to the silica particles decreases the concentration of free polymer in solution, so to interpret studies of particle suspensions, we were obliged to measure the equilibrium concentration. For a 50 mM NaCl solution, at concentrations of 1 wt % polySBMA/g of silica or less, we cannot detect any free polymer in solution, which means that there is less than 0.05 mg/mL of polymer in solution (our detection resolution limit). At 3 wt % polymer, we have about 10 mg/mL, and at 5 wt % polymer, we have about 40 mg/mL. Reference to Figure 2 (the adsorption isotherm) tells us that this means that above 3 wt %, we have enough polymer to reach the plateau in adsorption, whereas, at lower initial concentrations, we have very little polymer in solution, and probably incomplete surface coverage. To check for self-consistency, we have also calculated the expected equilibrium concentration after adsorption, based on the measured surface area of the particles and the known adsorption to a flat plate (from ellipsometry). At 3 wt % polymer, we estimate about a solution concentration of 10 mg/mL and at 5 wt % polymer, we estimate about 30 mg/mL. There is not enough polymer to form a layer at 1 wt %. (25) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Nature 1991, 353, 239241.

7592 Langmuir, Vol. 23, No. 14, 2007

Figure 8. Viscosity of silica dispersions (28 vol %) at pH 6 as a function of shear rate and ionic strength.

Figure 9. Viscosity of 28% silica dispersions as a function of shear rate for a range of polySBMA concentrations in 20 mM NaCl at pH 6 and 25 °C.

Rheology. Suspensions of bare silica particles at 20, 50, and 100 mM NaCl and pH 6 exhibit a decrease in viscosity with increasing shear rate as shown in Figure 8. This shear-thinning is consistent with some weak coagulation of the particles:26 an increase in shear reduces the viscosity by breaking up particle aggregates. The dynamic viscosity also increases with salt concentration. This is consistent with salt increasing coagulation by screening of the double-layer forces, as observed in the force measurements. Before proceeding to a study of the effect of the polymer on suspension rheology, we measured the effect of the polymer alone (no particles) on the fluid viscosity. At a concentration of 100 mg/mL polySBMA (well above the equilibrium polymer concentration in all suspensions), the viscosity was about 2.5 mPa s for 20-50 NaCl solutions, which is much less than the viscosity of the silica suspensions (>10 mPa s). That is, without the particles, the polymer has little effect on solution viscosity. As an aside, we note that the viscosity is slightly greater in 50 mM NaCl than in 20 mM, an effect which others have used to hypothesize that the radius of gyration of the polymer increases with added salt.1 (26) Starck, P.; Vincent, B. Langmuir 2006, 22, 5294-5300.

Starck et al.

Figure 10. Viscosity of the silica dispersions as a function of shear rate for a range of polySBMA concentrations in 50 mM NaCl at pH 6 and 25 °C.

Figure 11. Viscosity of silica dispersions as a function of shear rate for a range of polySBMA concentrations in 100 mM NaCl at pH 6 and 25 °C.

We know that the polymer adsorbs to silica, and alters the surface forces, so we expect that the polymer will modify the rheology of silica suspensions. These data are shown in Figures 9-11 for 20, 50, and 100 mM NaCl, respectively. In each case the polymer concentration is the initial polymer concentration, not the equilibrium concentration. Some of the trends are more easily seen in Figure 12, where the viscosity has been plotted as a function of polymer concentration at a constant shear rate of 100 s-1. In Figure 12, we see that the suspension viscosity increases gradually as the polySBMA concentration increases in the range 0.01-0.7% polymer/g of silica, and the suspensions become shear thinning (Figures 9-11). The viscosity increases dramatically in the range 0.7-1% polymer: the suspension is strongly aggregated (a thick paste), and the viscosities are only estimates in this region. The viscosity is very high in 1% polymer. We attribute these changes in the flow behavior of suspensions to changes in surface forces between the particles as follows. Up to about 1 wt % polymer, we know from the measurement of equilibrium concentration and the plate adsorption studies that there is insufficient polymer to reach saturation adsorption on the particles. Thus, there are available sites for polymer adsorbed on one particle to adsorb on a second particle during

Polysulfobetaine-Coated Silica

Langmuir, Vol. 23, No. 14, 2007 7593

In 50 and 100 mM salt solutions, the particles become stable and the viscosity is low when there is sufficient polymer to coat the particles. In fact, the viscosity is lower for the polymercoated particles than for the naked particles. We hypothesize that in these higher salt concentrations, there is sufficient polymer coverage and adsorption strength, yet a good enough solvent quality to produce steric forces that stabilize the particles (i.e., prevent them for adhering to each other). The large number of charged groups on the polymer, and entrained ions makes the polymer hydrophilic, and this hydrophilic layer provides a repulsive force and lubrication for the particles to easily slide past each other. This hypothesis is supported by the AFM measurements of a repulsive force that has a longer range than the double-layer force. This work is also consistent with previous work showing that addition of ionic surfactants and polymers can be used to stabilize and decrease the viscosity of particles dispersions.28,30 Figure 12. Effect of polySBMA concentration on the viscosity of Silica dispersions (28 vol %) at various ionic strengths (as indicated) at pH 6 and at a shear rate of 100 s-1 at 25 °C.

a collision (bridging), which leads to an attractive force, particle flocculation, and an increase in viscosity and shear-thinning. Bridging flocculation has been observed by several authors in the past between polyelectrolytes and silica or latex particles.27-29 Shear-thinning means that the aggregated system breaks down into smaller spherical aggregates or particles that layer under shear (rate or stress), thereby decreasing the viscosity. The AFM measurements show that the bridging takes some time (>1 s in the AFM) to occur, so it is likely that the shear-thinning behavior experienced during rheological experiments is not reversible. At 3 wt % polymer and above, there is sufficient polymer to reach saturation coverage on the particles. In this concentration range, the stability and rheology depend on the salt concentration. In 20 mM NaCl, the particles in polymer solution do not restabilize when the saturation coverage is reached, although the viscosity does decrease. We suggest that the viscosity decreases because of the diminished bridging opportunities when the polymer coating is more complete. However, the instability of suspension remains because of the poor solvent quality, which might lead to attractive forces between polymer-coated particles. (27) Cawdery, N.; Milling, A.; Vincent, B. Colloids Surf., A 1994, 86, 239249. (28) Whitby, C. P.; Scales, P. J.; Grieser, F.; Healy, T. W.; Kirby, G.; Lewis, J. A.; Zukoski, C. F. J. Colloid Interface Sci. 2003, 262, 274-281. (29) Liang, W.; Tadros, T. F.; Luckham, P. F. Langmuir 1994, 10, 441-446.

Conclusions The adsorption of zwitterionic polymers to silica particles provides a good range of control of suspension behavior. At low concentrations, the suspension can be destabilized through bridging flocculation, leading to a thick paste. When saturation coverage of the polymer is reached, the polymer provides steric repulsion that stabilizes the particles and produces a low suspension viscosity, even when the surface potential is very low (10 mV). The viscosity of the suspension is even lower than in the polymer-free suspension. The strong repulsive force and stabilizing effect in high salt solution should be useful to produce low viscosity and low friction in physiological saline and salt-water applications. The dependence of solvent quality on the salt concentration provides an additional degree of control; the suspension is unstable and viscous at low salt concentrations. Acknowledgment. This research was supported by the Australian Research Council and the Australian Institute of Nuclear Science and Engineering. The authors thank Associate Professor George Franks for useful discussions and assistance with rheological equipment and Dr. Tiziana Russo for initial polymer synthesis work. Finally, we would like to thank the Particulate Fluids Processing Center. LA700642D (30) Panya, P.; Wanless, E. J.; Arquero, O. A.; Franks, G. V. J. Am. Ceram. Soc. 2005, 88, 540-546.