Proximal Adsorption at Glass Surfaces: Ionic Strength, pH, Chain

378

Langmuir 2004, 20, 378-388

Proximal Adsorption at Glass Surfaces: Ionic Strength, pH, Chain Length Effects William J. Lokar and William A. Ducker* Department of Chemistry, Virginia Tech, Blacksburg, Virginia 24061 Received July 16, 2003. In Final Form: October 13, 2003

The adsorption and desorption of pyridinium chloride surfactants on borosilicate glass are studied as a function of the separation between two glass-solution interfaces. Both the adsorption and desorption change exponentially with the separation; the decay is equal to the solution Debye length. Changes in adsorption are smaller at pH 1.8 (near the point of zero charge of glass) than at pH 6. These results are consistent with an electrostatic cause for the changes in adsorption. The magnitude of the adsorption regulation, however, depends on the length of the alkyl chain and the surface excess of the surfactant. Therefore, proximal adsorption in this system depends on the coupling between the long-range electrostatic forces and the short-range chain-chain interactions. The equation of state for the surfactant on a regulating surface is discussed with respect to changes in intersurface separation.

Introduction Surfactant adsorption onto surfaces is important in the study of colloidal stability in solution, which has applications to products such as paints and food. The usual starting point for predicting the stability of colloidal suspensions is DLVO theory.1,2 This theory considers only electrostatic double-layer and van der Waals forces and considers them as independent entities. As a consequence of its simplicity, DLVO theory often fails to accurately describe experimental results. This has led many people to include “extra” forces in their theoretical interpretation. An additional difficulty with the practice of DLVO theory is that the input parameters, the surface charge, and the Hamaker “constant” are sometimes determined for isolated surfaces. When surfaces interact, however, average and local surface properties can be altered by the force fields contributed by the approaching surface. For example, the interaction between two surfaces can change the adsorption to the surfaces. This effect has been termed “proximal adsorption”.3 Recently, researchers have shown that the adsorption of surfactant to solid/aqueous surfaces is very rapid (it is usually transport limited).4,5 Thus, adsorption can change on the time scale of particle collisions. The most important change when particles collide is usually the change in the surface charge which is brought about by adsorption or desorption of ions (charge regulation).6 In surface force studies of individual particles or surfaces, which are immersed in a large reservoir, the reservoir maintains a constant chemical potential of each adsorbate. When these surfaces are brought into close proximity under equilibrium conditions, the surface excess of the adsorbate must change in order to keep the chemical * To whom correspondence may be addressed. E-mail: [email protected]. (1) Verwey, E. J.; Overbeek, J. T. G. Theory of the Stability of Lyophobic Colloids; Elsevier: Amsterdam, 1948. (2) Derjaguin, B. V.; Landau, L. D. Acta Phys.-Chim. URSS 1941, 14, 633. (3) Subramanian, V.; Ducker, W. J. Phys. Chem. B 2001, 105, 1389. (4) Clark, S.; Ducker, W. A. J. Phys. Chem. B 2003, 107, 9011. (5) Atkin, R.; Craig, V. S. J.; Biggs, S. Langmuir 2001, 17, 6155. (6) Ninham, B. W.; Parsegian, V. A. J. Theor. Biol. 1971, 31, 405.

potential of adsorbed species the same as the fixed chemical potential in bulk solution. Therefore, predictions of the changes in surface excess during the collision depend on how the surface excess is related to the chemical potential at the surface. At this point there is a divergence between traditional treatments that examine a single surface and those that examine two surfaces. Traditionally, researchers who study charge regulation of two interacting surfaces include only terms for the surface entropy and the surface electrical potential.7 In contrast, researchers who study surfactant adsorption to isolated surfaces usually need to include an interaction between the hydrophobic alkyl chains8 (e.g., in the Frumkin adsorption isotherm9). Because the calculation of the theoretical force between surfaces rests on assumptions about the adsorption to the surfaces, it is useful to know the extent of adsorption when two surfaces approach each other (the proximal adsorption). In the absence of direct experimental values of the adsorption during a collision, the adsorption can be determined using the formulation of Hall10 and Ash et al.11 They each derived a Maxwell relation between the change in force, F, with chemical potential, µ, and the change in adsorbed amount on both surfaces, nT, with separation, s. Changes in the surface energy, γ, for an object of area, A, result from the change in force between two surfaces and the change in surface excess of an adsorbed species, nT. At constant temperature and pressure

-A dγ ) F ds +

∑j njT dµj

(1)

where the summation is over all species that maintain equilibrium between bulk and adsorbed states. Since dγ is an exact differential, a Maxwell relation can be obtained (7) Israelachvili, J.; Pashley, R. M. J. Colloid Interface Sci. 1984, 98, 500. (8) Somasundaran, P.; Fuerstenau, D. W. J. Phys. Chem. 1966, 70, 90. (9) Prosser, A. J.; Franses, E. I. Colloids Surf., A 2001, 178, 1. (10) Hall, D. G. J. Chem. Soc., Faraday Trans. 2 1972, 68, 2169. (11) Ash, S. G.; Everett, D. H.; Radke, C. J. Chem. Soc., Faraday Trans. 2 1973, 69, 1256.

10.1021/la035288v CCC: $27.50 © 2004 American Chemical Society Published on Web 12/19/2003

Proximal Adsorption at Glass Surfaces

Langmuir, Vol. 20, No. 2, 2004 379

from eq 1 under conditions of adsorption equilibrium

( ) ∂F ∂µi

( )

∂niT ) ∂s p,T,s,µj

(2)

p,T,µi,µj

For a Z:Z electrolyte, eq 2 takes the form

( ) ∂F ∂µ(

) p,T,s,µj

(

)

∂(n+T + n-T) ∂s

p,T,µi,µj

(3)

Equation 3 describes the sum of the adsorptions of both ions in terms of the average chemical potential µ( ) (µ+ + µ-)/2. If other ions are present, and can adsorb or desorb from the surface, then n+ is not necessarily equal to n-. In the current work, our primary interest is to study the adsorption of a surfactant cation, so we have held µnearly constant through the addition of salt. This returns us to the conditions of eq 2. Considering the interaction between two infinite flat sheets, if we divide through by a unit area over which the adsorption is uniform and integrate with respect to s, eq 2 becomes

∆Γ ) Γi(s) - Γi(∞) ) -

( )

1 ∂Ea 2 ∂µi

(4)

p,T,µj,s

where Γ is the surface excess per unit area of one surface and Ea is the interaction energy per unit area of one infinite flat sheet interacting with another infinite flat sheet. The prefactor of 1/2 arises because one-half of the adsorption is assigned to each surface. It is difficult to arrange the parallel alignment of infinite flat sheets, so in practice forces are usually either measured between a sphere and a flat sheet or measured between two spheres. The Derjaguin approximation12 can be used to convert the experimental forces to Ea. For a sphere of radius R interacting with an infinite flat sheet

Ea ) F/2πR

(5)

Equation 5 holds when both the range of the force and the separation between surfaces are much smaller than R. The power of applying eq 4 is that it does not depend on the choice of a particular model for the surface. Therefore, it can be used to test the predictions of various calculations. A number of authors have used the idea of proximal adsorption to interpret force data in surfactant solution.13-19 Studies of the cationic surfactant C12TABr in both the presence and absence of added electrolyte show a strong change in adsorption of the surfactant as the particles approach.18,19 Results show a region where the change in adsorption is constant regardless of the surface potential, suggesting that “proximal adsorption” is not exclusively driven by electrostatics. Furthermore, the attractive force obtained in surfactant solution (sometimes (12) Derjaguin, B. V. Kolloid-Z. 1934, 69, 155. (13) Podgornik, R.; Parsegian, V. A. J. Phys. Chem. 1995, 99, 9491. (14) Pethica, B. A. Colloids Surf., A 1995, 105, 257. (15) Yaminsky, V.; Jones, C.; Yaminsky, F.; Ninham, B. W. Langmuir 1996, 12, 3531. (16) Yaminsky, V. V.; Ninham, B. W.; Christenson, H. K.; Pashley, R. M. Langmuir 1996, 12, 1936. (17) Christenson, H. K.; Yaminsky, V. V. Colloids Surf., A 1997, 129-130, 67. (18) Subramanian, V.; Ducker, W. J. Phys. Chem. B 2001, 105, 1389. (19) Lokar, W. J.; Ducker, W. A. Langmuir 2002, 18, 3167.

attributed to long-range hydrophobic force) was found to be of much shorter range in electrolyte solution and was proposed to be of electrical origin. Apparent deviations from DLVO theory were explained as errors in estimating the surface adsorption regulation of surfactant. This error probably arises from neglect of surface density dependent short-range hydrophobic effect.18,19 This study aims to gain a further understanding of the contributions to proximal sorption on charge-regulating surfaces. Experiments are conducted at different ionic strengths and pH conditions to gain further insight into the double-layer contribution to proximal adsorption. Experiments with different alkyl chain lengths are used to understand the importance of chain-chain interactions. We have studied cetylpyridium chloride (CPC) and dodecylpyridinium chloride (DPC) because the existence of a UV chromophore facilitates the measurement of adsorption isotherms at infinte separation. Forces between silica surfaces in CPC were previously measured by Pashley et al. in the presence of 0.1 M NaCl.20 They report the presence of long-range (up to 40 nm) attractive forces. In general, we find steplike features in force-distance data for surfactant solutions with this concentration of salt. These steps may be associated with the presence of small bubbles on the surface of the silica.21 Therefore, we have focused on measurements in more dilute salt solutions where there is more confidence that the interacting surfaces are glass-solution interfaces. The adsorption of CPC and DPC to suspensions of silica particles has been studied by Koopal and co-workers.22-24 They find that the functional form of the adsorption depends on the electrolyte concentration. At low concentrations of salt (1 mM), the adsorption shows two plateaus whereas at high concentrations (100 mM) the adsorption shows a steep rise. At low surfactant concentrations, salt hinders the adsorption by screening and competing with attractive electrostatic interactions between the silica and the surfactant. At high surfactant concentrations the salt screens the repulsive interactions between surfactant molecules. These two effects lead to a “common intersection point”: when the net surface charge (silica plus surfactant) is close to zero, the isotherms at different salt concentrations intersect. The sessile contact angle is also maximum at this point. Experimental Section Reagents. Cetylpyridinium chloride and dodecylpyridinium chloride (Aldrich, Milwaukee, WI) were recrystallized three times from a 10:1 mixture of HPLC grade acetone and methanol. There was no minimum in a plot of surface tension vs concentration. Water was prepared by an EASYpure UV system (model D7401, Barnstead Thermolyne Co., Dubuque, IA). The water had a conductivity of 18.3 MΩ/cm at 25 °C and a surface tension of 72.3 mN/m at 22.5 °C. By the time that we finished preparing solutions, CO2 from the air had equilibrated in the solution, so all our experiments were at pH ≈ 6, unless otherwise noted. NaOH (Fischer Scientific, Pittsburgh, PA) and KCl (Aldrich, Milwaukee, WI) were each roasted in air at 300 °C to decompose organic impurities. HCl was (EM Science, Gibbstown, NJ) was used as received. Solid Substrates. Careful selection of substrates can increase the usefulness of force measurements. Two parameters that are (20) Craig, V. S. J.; Ninham, B. W.; Pashley, R. M. Langmuir 1998, 14, 3326. (21) Carambassis, A.; Jonker, L. C.; Attard, P.; Rutland, M. W. Phys. Rev. Lett. 1998, 80, 5357. (22) Goloub, T. P.; Koopal, L. K.; Bijsterbosch, B. H.; Sidorova, M. P. Langmuir 1996, 12, 3188. (23) Koopal, L. K.; Goloub, T. P.; de Keizer, A.; Sidorova, M. P. Colloids Surf., A 1999, 151, 15. (24) Goloub, T. P.; Koopal, L. K. Langmuir 1997, 13, 673.

380

Langmuir, Vol. 20, No. 2, 2004

important to us are substrate smoothness (to allow measurement of short-range forces) and similarity between the substrate sphere and flat plate (to simplify fitting to DLVO theory). Our earlier experiment measurements have been conducted between a sodalime glass and a silica plate.18,19 Here we use a borosilicate glass plate (Duke Scientific, Palo Alto, CA) and a borosilicate glass sphere (Fischer Sci., Pittsburgh, PA). Both had a root mean square roughness of approximately 0.3 nm over (400 nm)2, but the largest asperities in the same area were about 1.3 nm. The surface area of the particles was determined to be 3.85 ( 0.20 m2/g by BET analysis. The same batch of glass spheres was used for atomic force microscopy (AFM), adsorption, and zeta potential measurements, but the treatments were slightly different. For adsorption and zeta-potential measurements, the glass spheres were placed in water and then titrated to pH 6 using HCl. The spheres were then dried for the adsorption isotherms so that the exact mass of particles was known. For AFM measurements, the plate and the cantilever-mounted sphere were cleaned with an argon/water cold plasma at 20 W for 4 min. The plates were completely wet by water after the cleaning procedure. Surface Forces. Surface force measurements were conducted with a Digital Instruments Nanoscope III atomic force microscope at 25 ( 1 °C at a scan rate of 1 Hz (0.1-0.5 µm s-1).25,26 Forces were measured using triangular silicon cantilevers (k ∼1.4 N/m, Park Scientific, Sunnyvale, CA). A sphere was glued to the cantilever using Epon 1004 resin (Shell Chemical Co.) The diameter of individual spheres was measured with (0.5 µm error using an Olympus BH-2 light microscope with a video caliper. In our experiments 2 µm < R < 4 µm, and the forces are nonzero for s < 30 nm, so the Derjaguin approximation is appropriate when the sphere remains undeformed. Changes to the shape of the adsorbed layer are generally more facile than changes to the glass support, and these may change the geometry during the interaction. The spring constant of the cantilevers was calibrated from the loaded and unloaded resonant frequencies.27 The error in the spring constant was (7%, giving a total systematic error of about 13% in F/R. At the start of each experiment, the AFM cell was then flushed thoroughly with water. If a monotonic repulsion was not obtained, 40 mM NaOH solution was injected into the flow cell and equilibrated for 5-10 min. The cell was then flushed with water until a monotonic repulsion appeared. NaOH probably dissolves the first few layers from both the plate and sphere, resulting in fresh, clean surfaces for subsequent experimentation. We used the absence of an attractive minimum in the force between the sphere and plate as a test for a hydrophilic surface. Surfaces were only used when there was no attractive minimum. The AFM measurement does not directly reveal the separation between the sphere and the plate. The separation is assumed to be constant when the slope of the cantilever deflection vs piezo displacement graph is linear and nonzero. This separation defines s0, and all force-separation plots show the separation relative to s0. We are assuming that, in the linear region, the spring is much more deformable than the sphere, the plate, or the thin film between them. Because we interpreted our results in terms of an equilibrium model, it was important to establish that the measurements were at adsorption equilibrium. We established this by noting that the forces were (a) the same when the concentration was increased and decreased, (b) the same on approach and separation, and (c) the same when the speed was increased or decreased by a factor of 10. There were three circumstances when our results did not meet these criteria. The forces were not reversible during mechanical instabilities of the tip, and the force for concentrated surfactant solutions (e.g., Figure 9, 1.0 mM) was speed dependent at 10 Hz just before the mechanical stability. Most important, the force was not reproducible when the concentration was decreased from above the force minimum in large increments. (25) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Nature 1991, 353, 239. (26) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Langmuir 1992, 8, 1831. (27) Cleveland, J. P.; Manne, S.; Bocek, D.; Hansma, P. K. Rev. Sci. Instrum. 1993, 64, 403.

Lokar and Ducker

Figure 1. Electrical potential of a cetylpyridinium-selective electrode vs a silver/ silver chloride reference electrode: O, pH 1.8; 9, pH 6 and 1 mM KCl; ×, pH 6 and 10 mM KCl. This suggests that there may be a slow adsorption or rearrangement on a longer time scale. Adsorption to “Isolated” Particles. Adsorption to suspensions of particles was measured by the solution depletion method. An accurately known mass of approximately 20 mg of particles was added to 4 mL of a known concentration of surfactant and salt solution and allowed to equilibrate for 48 h. During the equilibration the solutions were shaken and sonicated periodically to redisperse the particles. At the end of this period, the pH was measured. In all cases it was in the range 6.0 ( 0.2, so no further adjustments of the pH were made. Finally, the particles were removed by sedimentation and the solution was analyzed by a UV spectrometer (Ocean Optics Inc., Dunedin, FL) at 260 nm. The final concentrations were determined by reference to a calibration plot of adsorbance versus known concentration. Zeta Potential, ζ. Measurements were conducted with a Zeta Meter 3.0+ (Zeta-Meter Inc., Staunton, VA). Approximately 0.01 g of particles was added to 50 mL solutions of surfactant and KCl and equilibrated for 48 h. The electrophoretic mobility, u, was measured in both forward and reverse directions and averaged over several particles. The zeta-potential was then calculated from the Hemholtz-Smoluchowski equation,28 where it is assumed that the particle size is much greater than the size of the double layer

ζ ) ηu/r0

(6)

where η is the visocity of the solution, r is the relative permittivity of the solution, and 0 is the permittivity of a vacuum. Contact Angles. Contact angles were measured using a First Ten Angstroms model FT125 (Portsmouth, VA). The plates were treated as for AFM experiments, then a solution droplet was placed on the surface for ∼10 min prior to measuring the receding angle Chemical Potential. Surfactant selective electrode measurements were performed to obtain both the critical micelle concentration (cmc) and the chemical potential of the surfactant monomer above the cmc in each system.29 Electrodes selective for the cetylpyridinium and dodecylpyridinium ions were fabricated using the procedure of Cutler et al.29 A silver/silver chloride reference electrode (Orion, Beverly, MA) was used to complete the circuit.

Results The results focus on data for CPC in 10 mM KCl at pH 6. Additional data are provided at pH 1.8, 1 mM KCl, or for a shorter surfactant chain. Figure 1 shows the electrical potential of a CP+-selective electrode as a function of surfactant concentration. The (28) Evans, D. F.; Wennerstro¨m, H. The Colloidal Domain: Where Physics, Chemistry, Biology, and Technology Meet, 2nd ed.; WileyVCH: New York, 1999. (29) Cutler, S. G.; Meares, P.; Hall, D. G. J. Electroanal. Chem. 1977, 85, 145.


Langmuir, Vol. 20, No. 2, 2004 381

Figure 2. Adsorption of CPC to borosilicate glass particles in the presence of 10 mM KCl at pH ) 6 ( 0.2. The plateau adsorption occurs at a value of 2.2 molecules/nm2. Table 1. cmc Values for CPC and DPC at Various Solution Conditions surfactant

conditions

cmc/mM

CPC CPC CPC DPC

1 mM KCl, pH 6 10 mM KCl, pH 6 6 pH 1.8 10 mM KCl, pH 6

0.68 0.20 0.14 12.4

change in slope reveals the cmc of the surfactants in various salt solutions, which are summarized in Table 1. The adsorption and surface potential of silica in the presence of surfactant has been studied previously.22-24,30-35 However, there is often considerable variation between results obtained for different particles. Here we report results for particles that are particularly suited to AFM force measurements. Figure 2 shows the adsorption of CP+ on the glass spheres as a function of surfactant concentration in the presence of 10 mM KCl. Adsorption rises slowly at low concentrations as expected and shows a steep rise before leveling off near the cmc. The isotherm shows similar behavior to isotherms for CPC on silica at 10 mM salt and similar pH.24,36 Figure 3 shows the zeta potential of the glass spheres in CPC/10 mM KCl solutions. The zeta potential follows approximately the same pattern as the isotherm in Figure 2. This is expected because it is the adsorption of the cationic surfactant that changes the net charge on the surface. The greatest slope in the adsorption occurs at approximately the charge neutralization point of the particles, ≈0.02 mM CPC. Figure 4 shows the receding contact angle of aqueous solutions of CPC on the same glass material that was used for the force measurements. As is typical for surfactant solutions on hydrophilic surfaces, the contact angle passes through a maximum.23 The initial rise in contact angle is usually interpreted as being a result of adsorption of surfactant with the alkyl chains facing the solution, and the decline in contact angle at higher concentrations is usually interpreted as being the result of adsorption with the headgroups facing the solution. In (30) Wa¨ngnerud, P.; Olofsson, G. J. Colloid Interface Sci. 1992, 153, 392. (31) Fleming, B. D.; Biggs, S.; Wanless, E. J. J. Phys. Chem. B 2001, 105, 9537. (32) Favoriti, P.; Treiner, C. Langmuir 1998, 14, 7493. (33) Somasundaran, P.; Healy, T. W.; Fuerstenau, D. W. J. Phys. Chem. 1964, 68, 3562. (34) Esumi, K.; Matoba, M.; Yamanaka, Y. Langmuir 1996, 12, 2130. (35) Pashley, R. M.; Karaman, M. E.; Craig, V. S. J.; Kohonen, M. M. Colloids Surf., A 1998, 144, 1. (36) Favoriti, P.; Mannebach, M. H.; Treiner, C. Langmuir 1996, 12, 4691.

Figure 3. Zeta-potential of borosilicate glass spheres as a function of CPC concentration in 10 mM KCl at pH ) 6 ( 0.2. The zeta potential is zero at approximately 0.02 mM CPC, indicating that the surface charge on the silica plus the charge from adsorbed Cl- is compensated by adsorption of CP+ and K+ ions from solution.

Figure 4. Receding contact angles for CPC solution droplets on a borosilicate glass plate at pH ) 6 ( 0.2. The lines are merely aids to enhance discrimination of data sets.

our results, the addition of KCl is seen to increase the slope in the contact angle. This is consistent with a delay in adsorption caused by screening of the electrostatic interaction between the silica surface and the surfactant. The addition of HCl causes a reduction in maximum contact angle, which is consistent with less adsorption of surfactant. The proton is known to have a higher affinity for silica than K+.16 Figures 5 and 6 show the force as a function of CPC concentration in 10 mM KCl (Debye length, κ-1 ) 3.0 nm). Note that the forces at all measurable concentrations follow a very simple form: there is no crossing of curves for different concentrations. At concentrations less than 0.005 mM, the force changes little with concentration aside from a small increase in force with the initial addition of surfactant. This is consistent with the very small levels of adsorption shown in Figure 2. From 0.005 to 0.01 mM, the force decreases with concentration, and from 0.01 mM to the cmc, the force increases with concentration. In broad outline, these trends are consistent with the zeta potential measurements. Since both surfaces are very similar, the force should be greater when the absolute surface potential is greater. Note, however, that the minimum in force (0.01 mM) is not exactly at the zero of zeta potential (0.02 mM). Furthermore, at the attractive minimum, the force is much greater in magnitude than the predicted van der Waals force. Above the cmc there is little change because the chemical potential of surfactant is relatively insensitive to changes in concentration. The forces in 1 mM KCl (Figures 7-9) are similar to those in 10 mM KCl, except that they are longer ranged,

382


Figure 5. Forces between glass surfaces in 10 mM KCl as a function of CPC concentration. The forces in 0.001 and 0.005 mM are the same within experimental error and are slightly greater than that in 0 mM CPC. From 0.008 to 0.01 mM CPC, the force decreases as the concentration increases. The Debye length is κ-1 ) 3.0 nm. The van der Waals force, shown for comparison, was calculated using the Hamaker constant for silica-water-silica of 0.8 × 10-20 J (Hough, D. B.; White, L. R. Adv. Colloid Interface Sci. 1980, 14, 3-41). The arrow on the data measured in 0.01 mM CPC indicates that there is a mechanical instability in the measurement spring. This occurs when the force at which the cantilever spring constant equals the gradient in force. Over this range, the sphere undergoes nonequilibrium motion until s ) s0.

Lokar and Ducker

Figure 8. Forces between glass surfaces in 1 mM KCl as a function of CPC concentration. The force decreases from 0.005 to 0.01 mM CPC. The Debye length is κ-1 ) 9.5 nm.

Figure 9. Forces between glass surfaces in 1 mM KCl as a function of CPC concentration. Above 0.01 mM CPC the force increases as the surfaces are brought together. A mechanical instability occurs for separations of less than s - s0 ) 4 nm. The Debye length varies as the surfactant concentration approaches the KCl concentration.

Figure 6. Forces between glass surfaces in 10 mM KCl as a function of CPC concentration. Forces increase with increasing CPC concentration up to the cmc. A mechanical instability occurs for separations of less than s - s0 ) 4 nm. The Debye length is κ-1 ) 3.0 nm.

Figure 10. Forces between glass surfaces in as a function of CPC concentration at pH 1.8. Forces are virtually unchanged for CPC concentrations below 0.002 mM. Between 0.001 and 0.002 mM CPC, the force increases slightly. The force then decreases as the concentration of CPC is increased. The Debye length is κ-1 ) 2.4 nm.

Figure 7. Forces between glass surfaces in 1 mM KCl as a function of CPC concentration. Forces do not vary within experimental error for nonzero surfactant concentrations below 0.005 mM. The Debye length is κ-1 ) 9.5 nm.

as expected (κ-1 ) 9.5 nm). The minimum in force occurs at the same CPC concentration for both salt concentrations.

We also measured the forces at pH ) 1.8 (Figures 10 and 11). At this pH, the glasslike surfaces are known to have only a small charge.22 In the absence of CPC, the force is weak and short range, as expected when the surface potential is small and the decay length is short (κ-1 ) 2.4 nm). The surfactant, however, can exert a strong influence on the force. The overall behavior is similar to that observed in 10 mM KCl: as the surfactant concentration is increased, the force first decreases and then increases.


Figure 11. Forces between glass surfaces in as a function of CPC concentration at pH 1.8. Forces increase with increasing CPC concentration above 0.06 mM. A mechanical instability occurs for separations of less than s - s0 ) 4 nm. The Debye length is κ-1 ) 2.4 nm.

Langmuir, Vol. 20, No. 2, 2004 383

Figure 14. Interaction energy, Ea, measured by AFM compared to the energy calculated using DLVO theory at constant electrical potential. The experimental forces are at 0.008 and 0.04 mM CPC in 10 mM KCl and 0.005 mM CPC in 1 mM KCl. The DLVO fits shown are for -52 mV/κ-1 ) 2.9 nm, +67 mV/κ-1 ) 2.9 nm, and -80 mV/κ-1 ) 9.45 nm The plane of charge for +67 mV has been shifted 3.5 nm toward the bulk solution.

cmc, the repulsive force in DPC is also much smaller than that in CPC, leading to the prediction that CPC would be much better at stabilizing colloidal silica particles. The additional repulsive barrier in CPC solution arises from a combination of a higher surface potential and an extra short range repulsive force. The higher potential originates in the higher packing parameter, which produces a greater average density of surfactant. The additional short-range force was approach-speed dependent and may be due to the activation energy to remove the surfactant. Analysis Figure 12. Forces between glass surfaces in 10 mM KCl as a function of DPC concentration. The force in 0.1 mM is slightly greater than that in 0 mM DPC. From 0.1 to 0.8 mM DPC the force decreases as the concentration increases. The Debye length is κ-1 ) 3.0 nm. The arrows on the data show the force at which the cantilever spring constant equals the gradient in force. At this point the sphere undergoes nonequilibrium motion until s ) s0.

Figure 13. Forces between glass surfaces in 10 mM KCl as a function of DPC concentration. Forces increase with increasing DPC concentration up to the cmc. A mechanical instability occurs for separations of less than s - s0 ) 4 or 5 nm. Note that the Debye length is not constant in these data.

The small increase in repulsive force at low surfactant concentrations is more pronounced at pH ) 1.8. Figures 12 and 13 show the forces as a function of DPC (the 12 carbon analogue of CPC) concentration in 10 mM KCl. The forces are again very similar to those observed in CPC but shifted to higher surfactant concentrations because of the greater surfactant solubility. Above the

Force Analysis. The experimental forces were converted to Ea using eq 5. The surface potentials were determined by fitting the data to DLVO theory using an exact numerical solution to the Poisson-Boltzmann equation and a Hamaker constant of 0.8 × 10-20 J (silica/ water/silica).37 The potentials were fitted at large separation where the effect of adsorption regulation is small. At concentrations where the force increases with surfactant concentration, the forces were too large to fit to DLVO theory if the plane of charge were placed at s0, as discussed previously.18,19 Results from neutron reflectivity38 as well as our contact angle measurements suggest that there are headgroups facing solution. The steep rise in the force curve at about 5 nm also suggests that the surfactant film is at least 2.5 nm thick on each surface (at the cmc), so we shifted the plane of charge to s > s0. The plane of charge may move under an applied force, but we did not introduce this additional freedom into our fit. The adsorption of surfactant will also change the van der Waals force. The Hamaker constant for alkane/water/alkane is similar to that of silica/water/silica,37 so we made an approximate calculation of the van der Waals force by simply moving the zero of the van der Waals force to the same position as the zero for charge. The DLVO forces shown in Figure 14 were calculated using a constant potential boundary condition, which in this case fits the data quite well down to small separations in 10 mM salt. In 1 mM salt the measured force is less repulsive than the fitted force for s - s0 < 3 nm. Three examples of fits are shown in Figure 14. Table 2 sum(37) Hough, D. B.; White, L. R. Adv. Colloid Interface Sci. 1980, 14, 3. (38) Fragneto, G.; Thomas, R. K.; Rennie, A. R.; Penfold, J. Langmuir 1996, 12, 6036.

384


Lokar and Ducker

Table 2. Fitted DLVO Parametersa [surfactant] (mM)

[KCl] (mM)

ψ0a (mV)

κ-1 (nm)

0 0.0005 0.001 0.002 0.005 0.008 0.02 0.04 0.06 0.1 0.2 0.4 0.8 0 0.0001 0.0005 0.001 0.005 0.02 0.05 0.1 0.2 0.4 0.8 1.0

10 10 10 10 10 10 10 10 10 10 10 10 10 1 1 1 1 1 1 1 1 1 1 1 1

-58 -60 -60 -60 -62 -52 +45 +67 +79 +93 +118 +123 +120 -92 -97 -102 -102 -80 +73 +106 +125 +138 +154 +174 +176

10 10 10 10 10 10 10

-85 -95 -75 +40 +52 +56 +56

0 0.1 0.2 4.0 8.0 12.0 16.0

shift dist (nm)

σ0a (charges/nm2)

CPC 2.9 2.9 2.9 3.0 2.9 2.9 2.9 2.9 2.75 2.9 2.9 3.0 2.85 9.45 9.45 9.45 9.45 9.45 9.6 9.3 9.1 8.9 8.0 7.2 7.4

0 0 0 0 0 0 2.7 3.5 3.4 3.3 2.8 2.7 2.5 0 0 0 0 0 3.0 3.3 3.8 3.9 3.6 3.0 2.6

-0.10816 -0.11317 -0.11317 -0.10969 -0.11893 -0.09321 +0.07740 +0.13349 +0.18342 +0.23361 +0.38830 +0.41486 +0.41126 -0.07022 -0.07786 -0.08625 -0.08625 -0.05453 +0.04611 +0.09507 +0.14234 +0.18850 +0.28785 +0.47457 +0.48031

DPC 3.1 2.8 3.0 2.6 2.2 2.1 2.1

0 0 0 3.8 3.6 3.6 3.6

-0.18489 -0.25216 -0.15418 +0.07475 +0.12287 +0.14227 +0.14227

a The signs of the potentials and charges are inferred from the trends in the force data.

marizes the DLVO parameters, including the shift distance, for all the force data. Note that the shift and the potential are coupled. It is interesting to note that the potentials fitted to double-layer theory are consistently greater than the ζ potentials measured on the particles (Figure 3) under identical conditions. This is the opposite trend to that observed by Scales et al.39 The zeta potential is the potential measured at the plane of shear, so one would expect a lower zeta potential when the Debye length is only 3 nm. If we choose a plane of shear at s - s0 ) 3 nm, the ζ potential fits well to the fitted surface potential at all concentrations. Adsorption Analysis. Values of Ea at a given surfactant chemical potential were converted to ∆Γ using eq 4. For solutions below the cmc, µsurfactant was obtained from Debye-Hu¨ckel theory, and for solutions above the cmc, µsurfactant was determined from the electrochemical measurements presented in Figure 1. The activity-potential relationship above the cmc was determined by comparing the measured potential below the cmc to Debye-Hu¨ckel theory. Equation 4 requires a constant chemical potential of all species except the surfactant. In our experiments we have used an excess of KCl to maintain a constant µCl-. Therefore, eq 4 can be applied specifically to the adsorption behavior of the surfactant cation. This approximation becomes worse when the concentration of surfactant (39) Johnson, S. B.; Drummond, C. J.; Scales, P. J.; Nishimura, S. Langmuir 1995, 11, 2367.

Figure 15. Changes in surface excess for CP+ as a function of separation for low (below the pzc) concentrations of CPC in 10 mM KCl. The Debye length is κ-1 ) 3.0 nm.

Figure 16. Changes in surface excess for CP+ as a function of separation for higher (below the cmc) concentrations of CPC in 10 mM KCl. The Debye length is κ-1 ) 3.0 nm.

approaches the concentration of added salt (in this paper, for concentrated DPC and for concentrated CPC in 1 mM KCl). Utilization of eq 4 requires measurement of forces at a variety of concentrations, but at the same separation. Because of thermal drift in the AFM, the separation (s s0) is only known after s0 is established from the measurement. So measurements must be interpolated to obtain a data set at a constant separation. We interpolated our data at 0.2 nm intervals using a linear interpolation. The derivatives in eq 4 were determined from interpolated curves of Ea versus chemical potential. We used either a cubic or smoothing spline, whichever most accurately represented the experimental data. The changes in surface excess as a function of separation at various surfactant concentrations and solution conditions are presented in Figures 15-20. It is difficult to determine the exact concentration at which the minimum in force occurs, so there is a relatively large error in determining the concentration at which there is no proximal adsorption. The changes in surface excess presented in Figures 1520 appear to decay exponentially. Some sample plots in Figure 21 confirm this observation. Figure 22 is a plot of the fitted decay length of proximal adsorption (ln(∆Γ) vs s - s0) as a function of the known solution Debye length. The decay length is approximately equal to the solution Debye length, which confirms the strong electrostatic contribution to proximal adsorption of these charged surfactants. Figure 22 includes data from many experiments including those conducted at pH ) 1.8. Therefore, even though adsorption of H+ reduces the magnitude of adsorption, it does not appear to have changed the mechanism.


Langmuir, Vol. 20, No. 2, 2004 385

Figure 17. Changes in surface excess for CP+ as a function of separation for low (below the pzc) concentrations of CPC in 1 mM KCl. The Debye length is κ-1 ) 9.5 nm.

Figure 20. Changes in surface excess for DP+ as a function of separation for various concentrations of DPC in 10 mM KCl. The Debye length varies according to surfactant concentration.

Figure 18. Changes in surface excess for CP+ as a function of separation for higher (below the cmc) concentrations of CPC in 1 mM KCl. The Debye length varies.

Figure 21. Examples of changes in surface excess displayed on a semilogarithmic scale. The proximal adsorption is exponential with respect to separation.

Figure 19. Changes in surface excess for CP+ as a function of separation for various concentrations of CPC at pH 1.8. The Debye length is κ-1 ) 2.4 nm.

Figure 22. Decay length for proximal adsorption (b) and desorption (4) plotted versus the solution Debye length. The solid line has a slope of 1. The decay lengths correlate well with the Debye length. The proximal adsorption and desorption used for the fits correspond to regions where the proximal sorption is a maximum.

Discussion Basic Features of Proximal Adsorption. Figures 15-18 show proximal adsorption for CP+ in the presence of 10 mM and 1 mM KCl. These measurements will be interpreted in terms of the chemical potential of the surfactant at the surface. In our experiments, the surfactant at the interface is in equilibrium with a much larger amount of surfactant in bulk solution, so we will discuss the limit of a constant chemical potential of surfactant during changes in separation between the particles. We will use a form of the chemical potential that includes contributions due to entropy of mixing, the local potential, ψs, acting on the charged headgroup, and short-range interactions of magnitude RTA(θ), where θ is fractional coverage of surfactant on the surface. Following

Prosser and Franses’ description of adsorption at the airwater interface,9 we will use the following form for isolated surfaces

µads ) µads0 + RT ln

(1 -θ θ) + zFψ + RTA(θ) s

(7)

which we have adapted by explicitly noting the dependence of ψ on the surface separation with the subscript s. Except for the last term, this is the form traditionally used to describe charge regulation during surface-surface interactions. Because the last term contains all densitydependent chemical-specific short-range interactions, the

386


Figure 23. Differences between the proximal adsorption (∆Γ) calculated from Ea(µ) using eq 4 and the change in the number of charges (∆σ/e) obtained by fitting solutions to the PoissonBoltzmann equation at constant electrical potential to Ea(s s0): (O) 0.008 mM CPC in 10 mM KCl; (0) 0.005 mM CPC in 1 mM KCl; (s) 0.2 mM DPC in 10 mM KCl; (×) 0.8 mM C12TABr in 20 mM KBr. The curves for the surfactants with a 12 carbon chain (DPC and C12TABr) are considerably smaller in magnitude than those for the 16 carbon chain (CPC). The data for C12TABr are taken from Lokar, W. J.; Ducker, W. A. Langmuir 2002, 18, 3167-3175.

functional form is difficult to write down. Prosser and Franses describe the Frumkin treatment where A(θ) is negative and linear and is used to account for attractive interactions between the alkyl chains (the hydrophobic effect). Preliminary calculations of A(θ) from our experimental data suggest that this is a reasonable assumption up to the charge neutralization point but that ∂A(θ)/∂θ varies at higher surface densities. This is reasonable as the surfactant forms complex and perhaps even layered shapes at solid-liquid interfaces (micelles, bilayers, etc.) and we would not expect the same extent of chain-chain interactions at all concentrations. Furthermore, the existence of micelles on the surface suggests that there should also be repulsive terms in A(θ) that are important at high surface densities. Our inability to write the exact functional form of A(θ) at this point necessarily makes the following discussion qualitative, but in all cases we will consider only A(θ) < 0. CPC in 10 mM KCl. There is a low concentration regime where there is little surfactant adsorbed, and the force does not change as surfactant is added, and therefore the adsorption does not change as the surfaces approach each other. This is easily explicable because the surfactant density is so low that the last term in eq 7 is negligible. During the approach of another surface, the charge on both surfaces can be regulated by adsorption of the much more numerous K+ ions. At higher concentrations (10-2-10-1 mM), the steep rise in the adsorption isotherm (Figure 2) indicates that A(θ) is now significant. At the same surface excess where the density is very responsive to chemical potential, the density is also responsive to the approach of another surface. Figure 15 shows that the surfactant adsorption increases by ∼50% as the surfaces approach 0.008 mM CPC. Recall that in Figure 14 we showed that the force for this condition fit well to a double-layer force calculated at constant potential with no A(θ) term. In fact, the actual adsorption of CP+ is much greater than that suggested by the good fit to the constant potential boundary condition. Figure 23 shows the difference between the known surfactant adsorption and the fitted charge adsorption. Clearly, the actual proximal adsorption of surfactant is far in excess of what is required simply to regulate the

Lokar and Ducker

Figure 24. Adsorption as a function of concentration for various separations. The isotherm at infinite separation is for particles only, and the adsorption at other separations was obtained by adding the proximal adsorption for the particle-plate interaction. The plateau decreases with decreasing separation. The height of the peak below [CPC] ) 0.01 mM increases as the separation is decreased.

surface charge. Similar deviations were observed in previous work.18,19 We propose that the additional adsorption is driven by the chain-chain interaction, which is known to be important in adsorption of surfactants to isolated surfaces.8 Figure 23 shows that at a lower salt concentration, the deviation extends to greater separations. The deviation is so bad at this long Debye length that it is not possible to even fit the surface forces at small separation (Figure 14). When the measured force is much more attractive than the calculated force using DLVO theory, there is a temptation to conclude that there is an additional long-range attractive force. However, the proximal adsorption analysis shows us that the actual surfactant adsorption is much greater than that required to maintain constant electrical potential. Unless this discrepancy is made up by adsorption of Cl- or desorption of K+, the surface charge is smaller than required by the fit to constant electrical potential, and the double-layer force is weaker than was predicted by conventional charge regulation models. The need to invoke an additional longrange force is replaced by a need to more accurately model the regulation of surfactant adsorption at constant chemical potential. In future work we plan to introduce into DLVO theory a more realistic equation of state for surfactant adsorption that includes the short-range interactions between alkyl chains. Here we can only state qualitatively that the longrange electrostatic forces are coupled to the short-range hydrophobic interactions between alkyl chains via the surfactant molecule in which groups experiencing these forces are covalently bound together. Neglect of this coupling can lead to errors of interpretation. At surfactant concentrations greater than 0.2 mM, the surfaces are positively charged and the electrostatic contribution to eq 7 is now positive. If the surfaces were to approach each other without adsorption of any ions, the potential would increase, so cations desorb to reduce this effect. The magnitude of desorption of the CP+ is shown in Figure 16. The adsorption change is almost independent of concentration, even though the potential changes over a wide range (see Table 2). Thus A(θ) must be increasing with concentration to compensate. Similar features are observed in 1 mM KCl. Total Adsorption as a Function of Separation. Adsorption isotherms as a function of separation are shown in Figure 24 (data from Figures 2, 15, and 16 and force of adhesion). The concentration of the force minimum (0.01


Figure 25. Change in surface excess of CP+ and the change in charge required to maintain a constant potential interaction for two surfactant concentrations in 10 mM KCl.

mM) defines the common intersection point of the proximal adsorption. (The slight displacement to higher concentration in Figure 24 arises from the error in fitting to data at a finite number of concentrations.) For two identical surfaces interacting only through classical DLVO forces, the intersection point should occur when each surface is net neutral, and therefore it would coincide with the common intersection point observed for isolated surfaces when the salt concentration is varied.18 Indeed, we do observe the same intersection point for 1 and 10 mM salt, but not at the concentration when the particle is neutral. Furthermore, this intersection point is moved by a change in pH: it is at 0.06 mM at pH 1.8. Approach of a second surface is a bit like changing the pH for an isolated surface: the second surface changes the local potential, causing adsorption or desorption of the charged surfactant. This effect is large just below the intersection point where the dense alkyl chains and the negative potential conspire to enhance adsorption. Even greater adsorption occurs at ca. zero separation, where the alkyl chains from the two surfaces can overlap. The “zero separation” data above the intersection point suggests that we are not able to remove all the surfactant from the surface. However, details of the adsorption at zero separation must be viewed with some skepticism because (1) it is more difficult to achieve equilibrium in a thin film, (2) the adhesion measurement is necessarily dynamic, and (3) the steep forces introduce large errors. At [CPC] > 0.01 mM, the surfactant monomers desorb as the surfaces come together. This proximal desorption appears to level off to a constant value as can been seen in Figure 16 and to some extent Figures 18-20 at different conditions. Figure 25 shows that the measured proximal adsorption of surfactant is at about the level required to maintain a constant electrical potential in 10 mM salt. The trends are the same for other conditions studied. It is somewhat surprising that the agreement is so good, and the agreement may be fortuitous. We expected that the Cl- counterions would play a large role in regulating the charge. CPC in 1 mM KCl. The general features of proximal adsorption are the same in 1 mM KCl as in 10 mM KCl. As the surfaces approach, there is adsorption below 0.01 mM and desorption at higher concentrations. The changes in adsorption simply occur at larger separations. This phenomenon has been seen previously for C12TABr solutions with19 and without added salt.18 Eliminating the Double Layer: pH ) 1.8. Doublelayer effects can be minimized at high salt or at low surface potentials. In practice, we find that in solutions where the KCl concentration exceeds 50 mM there is a high

Langmuir, Vol. 20, No. 2, 2004 387

probability of experiencing steps in the force curve, which have been attributed to bubble formation on the surfaces.21 Therefore, we examined the adsorption in pH ) 1.8 solutions, near the point of zero charge (pzc). The Debye lengths in 10 mM KCl and ∼15 mM HCl are similar, so comparison of the two allows us to see how the magnitude of the surface potential affects proximal adsorption. By comparing Figures 15 and 19, we see that the change in adsorption at low concentrations is greatly reduced for CP+ near the pzc of the glass. Thus, noncontact proximal adsorption is small when the surface charge is small, independent of whether the small surface charge is obtained using H+ or CP+. This reinforces the idea that electrostatic forces are the only long-range forces that affect adsorption. Figure 11 shows that, at pH ) 1.8, surfactant adsorption leads to a significant repulsive force when [CPC] > 0.1 mM. At this concentration, there is a net surface charge arising from the charged surfactant headgroups. The protons in solution are not potential-determining ions for charged surfactant headgroups, so they do not eliminate the surface potential. The existence of the surface potential leads to proximal desorption. We would expect protons to have a effect over a larger range of surfactant concentrations for surfactants with a pH-dependent headgroup charge (such as a carboxylic acid headgroup). Effect of Chain Length. Comparison of results for DPC and CPC at the same ionic strength (10 mM KCl) allow us to estimate the effect of the alkyl chain length on proximal adsorption. The forces in DPC show the same trends as in CPC, except that similar features occur at elevated concentrations due to greater solubility of DPC. As a consequence, the proximal adsorption has a similar form for the two surfactants (see Figures 15, 16, and 20). The principal difference is that the magnitude of proximal adsorption is smaller for the shorter-chain surfactant. At surfactant concentrations above the force minimum, the surface potentials are much greater for CPC (because the surfactant is more densely packed).18 This greater potential should drive greater desorption from the surface. Figure 23 also shows that the deviation between the charge regulation model (which neglects A(θ)) and the actual adsorption (model independent) is significantly smaller for DPC than for CPC, confirming that neglect of the alkyl chain interactions is the cause of the deviation. In addition, Figure 23 shows that results for DPC in 10 mM KCl on borosilicate glass are nearly identical to those for C12TABr in 20 mM KBr on silica. The two surfactants have different headgroups, but specific headgroup-surface interactions do not appear to be important; it is the alkyl chain that causes the deviation. Proximal Desorption of Surfactant in Very Dilute Surfactant Solutions. Below the force minimum, surfactant usually adsorbs as the surfaces approach each other. This is reasonable, because it lowers electrostatic potential. In earlier work, we noted that, in very dilute solution, the surfactant actually desorbs as the surfaces approach even though the surfactant and the surface have opposite charges.19 In the current work, this effect is seen as an increase in the force on addition of surfactant (Figure 5, 0.001 mM; Figure 8, 0.001 mM; Figure 10, 0.002 mM; and Figure 12, 0.1 mM). These deviations are indeed small in most cases but have been consistent in all measurements. A possible explanation is as follows. Table 2 shows that addition of surfactant to solution increases the magnitude of the fitted net charge of an isolated surface. The energy contribution of an alkyl chain to surfactant adsorption on silica is thought to be about -1 kT per methylene unit.33

388


Perhaps an adsorbing alkyl chain displaces more than one small cation (H+ or K+) thereby causing an increase in net charge. On approach of a second surface, the doublelayer force could be minimized by again replacing the surfactant with more than one small cation. Long-Range Hydrophobic Forces? We clearly observe net attractive forces that are much greater in magnitude than van der Waals forces. The simplest explanation for these forces is an attractive double-layer force caused by opposite charges on the two surfaces. This is suggested by the discrepancy between the surfactant concentration for neutral particles (∼0.02 mM) and the maximum contact angle for the plate (∼0.01 mM). However, we were not able to fit the functional form of the measured force to the exact numerical solutions of the Poisson-Boltzmann equation at constant potential. If we fit an exponential curve to the data, the decay length is shorter than the solution Debye lengths (e.g., 1 mM, κ-1 ) 9.5 nm, fitted decay length 1.8 nm; 10 mM, κ-1 ) 3.0 nm, fitted decay length 1.6 nm). More complex explanations of this attractive force include (1) an attraction caused by the high energy of water in the thin film between the hydrophobic alkyl chains40 or (2) the existence of patches of positive and negative charges on each surface and correlations between them.41 A mechanism for charge patch formation is the clustering of cationic surfactant molecules on the surface that overcompensate the negative silica charge in one region and leave undercompensated silica charge in another region. Conclusions In 1 mM or 10 mM KCl solutions, the force between glass surfaces decreases at all finite separations when the concentration of surfactant is progressively increased to 0.01 mM CPC and then increases as further surfactant is added up to the cmc. Similarly, the solution contact angle increases and then decreases with concentration. The most attractive force occurs at the same concentration as the maximum receding contact angle. Analysis using the Hall equation shows that, when the separation between the glass surfaces is decreased, the surfactant adsorbs below 0.01 mM and desorbs above 0.01 mM. As expected, the cationic surfactant adsorbs when the surfaces have a significant negative potential and desorbs when the surfaces have a significant positive potential. (40) Israelachvili, J. N.; Pashley, R. M. J. Colloid Interface Sci. 1984, 98, 500. (41) Tsao, Y.-H.; Evans, D. F.; Wennerstro¨m, H. Science 1993, 262, 547.

Lokar and Ducker

The change in adsorption with distance is exponential and correlates very well with the calculated Debye length. This suggests that the long-range driving force for proximal adsorption is the electrostatic potential. Proximal adsorption (at s > s0) of surfactant is very low when the surface charge is low. The maximum in proximal adsorption corresponds approximately to the maximum slope of the contact angle with respect to concentration. The magnitude of the proximal adsorption depends on the chain length of the surfactant and therefore on the short-range interactions between surfactant molecules. Because ionic surfactants have both charged and hydrophobic groups, they provide a means by which a longrange force (electrostatics) can couple to short-range forces (e.g., local ordering of water around hydrophobic chains). We compared the actual surfactant adsorption during a collision to the level required to satisfy a charge regulation model, which neglects density dependent attractive terms. At surfactant concentrations below the force minimum (0.01 mM), surfactant proximal adsorption is much greater than necessary to fit conventional charge regulation. Above the force minimum, the charge carried by the desorbing CP+ ions is similar to the change in charge fitted to the charge regulation model. When two isolated surfaces are near the hemimicelle concentration, contact (s ∼ s0) between the two surfaces causes a dramatic increase in surfactant adsorption. We hypothesize that this adsorption is driven by the hydrophobic effect. We see no evidence for long-range nonelectrostatic forces causing proximal adsorption at large separations between the surfaces. The largest net attractive forces measured are much larger than van der Waals forces and are measurable up to about 7 nm separation between the surfaces. The maximum attraction does not occur at the same concentration at which the ζ potential for the particle is zero; it occurs where the contact angle on the plate is a maximum. The surface excess of surfactant is about 33% of the plateau value at the force minimum. Acknowledgment. This work is based on research supported by the National Science Foundation Grant CHE-0203987. We thank Dr. Matthew Eick at Virginia Tech for the BET analysis of our glass spheres, Dr. Mark Edwards for use of the Zeta Meter, and Mark Rutland for useful discussions. LA035288V

Proximal Adsorption at Glass Surfaces: Ionic Strength, pH, Chain

Recommend Documents