Effect of Concentration and Addition of Ions on the Adsorption of

Aug 24, 2010 - Neutron Reflectometry for Studying Corrosion and Corrosion Inhibition. Mary Wood , Stuart Clarke. Metals 2017 7 (8), 304 ...
0 downloads 0 Views 988KB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Effect of Concentration and Addition of Ions on the Adsorption of Aerosol-OT to Sapphire Maja S. Hellsing* and Adrian R. Rennie Department of Physics and Astronomy, Uppsala University, Box 516, 751 20, Uppsala, Sweden

Arwel V. Hughes ISIS Facility, Rutherford Appleton Laboratory, Didcot, OX11 0QX, United Kingdom Received May 17, 2010. Revised Manuscript Received August 11, 2010 Aerosol-OT (sodium bis 2-ethylhexyl sulfosuccinate or NaAOT) adsorbs to hydrophilic sapphire solid surfaces. The structure of the formed bilayer has been determined over the concentration range 0.2-7.4 mM NaAOT. It was found that the hydrocarbon tails pack at maximum packing limit at very low concentrations, and that the thickness of the bilayer was concentration-independent. The adsorption was found to increase with concentration, with the surfactant molecules packing closer laterally. The area per molecule was found to change from 138 ( 25 to 51 ( 4 A˚2 over the concentration range studied, with the thickness of the layer being constant at 33 ( 2 A˚. Addition of small amounts of salt was found to increase the surface excess, with the bilayer being thinner with a slightly larger area per molecule. Addition of different salts of the same valency was found to have a very similar effect, as had the addition of NaOH and HCl. Hence, the effects of adding acid or base should be considered an effect of ionic strength rather than an effect of pH. Adsorption of NaAOT to the sapphire surface that carries an opposite charge to the anionic surfactant is similar in many respects to the adsorption reported previously for hydrophilic and hydrophobic silica surfaces. This suggests that the adsorption of NaAOT to a surface is driven primarily by NaAOT self-assembly rather than effects of electrostatic attraction to the interface.

Introduction and Background Understanding the factors that control adsorption of molecules to interfaces and how they pack in self-assembled structures is crucial to the design of optimum formulations in many practical applications of surfactants such as detergents, wetting agents, and personal care products. Interactions of surfactants will usually depend on a balance of contributions that arise from enthalpic terms such as electrostatic charge or dipolar interactions, as well as the entropy of particular packing arrangements. It is therefore interesting to explore how systematic changes in particular contributions alter adsorption. There is a considerable amount of literature with detailed information about the structure of a range of nonionic1-3 and cationic4-6 surfactants adsorbed at solid/ liquid interfaces that has benefited from the high resolution and sensitivity to composition that can be obtained from neutron reflection measurements. Our present study aims to provide data on an anionic surfactant, Aerosol-OT (NaAOT, sodium bis(2ethylhexyl)sulfosuccinate), at the surface of Al2O3. There have been fewer studies of anionic surfactants, but interesting comparisons can be made as NaAOT is known to adsorb to a range of different hydrophilic and hydrophobic interfaces. In particular, we report on the structure and surface excess at the (0001) sapphire surface for concentrations up to about three times the *Corresponding author. [email protected]. (1) Lee, E. M.; Thomas, R. K.; Cummins, P. G.; Staples, E. J.; Penfold, J.; Rennie, A. R. Chem. Phys. Lett. 1989, 162, 196–202. (2) Penfold, J.; Tucker, I.; Thomas, R. K. Langmuir 2005, 21, 6330–6336. (3) Levitz, P. E. C. R. Geoscience 2002, 334, 665–673. (4) Rennie, A. R.; Lee, E. M.; Simister, E. A.; Thomas, R. K. Langmuir 1990, 6, 1031–1034. (5) Parida, S. K.; Dash, S.; Patel, S.; Mishra, B. K. Adv. Colloid Interface Sci 2006, 121, 77–110. (6) Golub, T. P.; Koopal, L. K.; Sidorova, M. P. Colloid J. 2004, 66, 38–43.

Langmuir 2010, 26(18), 14567–14573

critical micelle concentration in pure water and describe the effects of added salts and changes in pH. NaAOT is a branched double-tail, anionic surfactant that readily forms a lamellar phase at a low concentration.7 The critical micelle concentration in pure water8 is 2.5 mM, and the lower boundary of the lamellar phase7 is approximately 50 mM at 25 °C. NaAOT has previously been found to adsorb to hydrophobic silica surfaces prepared with a monolayer of octadecyl trichlorosilane,9,10 and a coverage of about 0.9 mg m-2 at and above the critical micelle concentration was found. This is broadly similar to the adsorption behavior at the air/water interface that is also a hydrophobic boundary.8,10-12 The adsorption isotherm of NaAOT to R-Al2O3 particles has been reported and shows adsorption increasing until a plateau of 2 mg m-2 is reached at the critical micelle concentration.13 At high concentrations, NaAOT does form a lamellar structure even at a hydrophilic silica surface.14 The present paper is concerned only with concentrations below the lamellar phase boundary, and so, these structures will not be discussed, but results of a study of high concentrations are to be found in a forthcoming paper.15 (7) Rogers, J.; Winsor, P. A. J. Colloid Interface Sci. 1969, 30, 247–257. (8) Li, Z. X.; Lu, J. R.; Thomas, R. K.; Penfold, J. J. Phys. Chem. B 1997, 101, 1615–1620. (9) Fragneto, G.; Li, Z. X.; Thomas, R. K.; Rennie, A. R.; Penfold, J. J. Colloid Interface Sci. 1996, 178, 531–537. (10) Li, Z. X.; Lu, J. R.; Fragneto, G.; Thomas, R. K.; Binks, B. P.; Fletcher, P. D. I.; Penfold, J. Colloids Surf., A 1998, 135, 277–281. (11) Li, Z. X.; Lu, J. R.; Thomas, R. K.; Penfold, J. Prog. Colloid Polym. Sci. 1995, 98, 243–247. (12) Li, Z. X.; Lu, J. R.; Thomas, R. K. Langmuir 1997, 13, 3681–3685. (13) Esumi, K.; Takaku, Y.; Otsuka, H. Langmuir 1994, 10, 3250–3254. (14) Li, Z. X.; Weller, A.; Thomas, R. K.; Rennie, A. R.; Webster, J. R. P.; Penfold, J.; Heenan, R. K.; Cubitt, R. J. Phys. Chem. B 1999, 103, 10800–10806. (15) Hellsing M. S.; Rennie A. R.; Hughes A. V. Manuscript in preparation.

Published on Web 08/24/2010

DOI: 10.1021/la101969p

14567

Article

Hellsing et al.

The charge on a hydrophilic sapphire surface depends on the pH of the adjacent aqueous phase. The isoelectric point where the surface carries no net charge depends on the particular surface for crystalline samples. For small colloidal particles that may have amorphous or hydrated surfaces, various different values are reported. Measurements of streaming potential or zeta potential have been used extensively.16,17 More recently, sum-frequency spectroscopy18 has been used to directly determine surface groups, and for the (0001) surface of sapphire, an isoelectric point of pH 6.3 is reported. This is not very different from typical values of about pH 7 determined from zeta potential measurements.16 Kershner et al.16 report that the surface potential varies from about 25 mV to -30 mV as the pH increases from 3 to 11.

molecules. Specular neutron reflection cannot distinguish between a structure that has patches of surfactant that are separated by regions with no adsorption and a layer with uniform coverage at a lower density, provided the lateral patches or gaps are small. The technique is sensitive only to the average scattering length density in the direction normal to the interface. Different molecules and isotopes have different scattering length densities. Hydrogen and deuterium have very different values for b as shown in Table 1, and this is often used for “contrast matching” or “contrast variation”, where, for example, the refractive index of the solvent can be matched to the substrate. Under these conditions, the measured signal comes almost exclusively from the material adsorbed at the surface. The surface excess is defined as

Neutron Reflection - Interpretation of Data Specular neutron reflection is a useful technique for studying buried interfaces on molecular scales. The incoming neutron beam is reflected at the interface and the intensity is measured. In the case of specular reflection, the angle of the incident beam, θ, is equal to the angle of the reflected beam. The ratio of the intensity of the reflected neutron beam to that of the incident beam is given as a function of the momentum transfer, Q, normal to the reflecting surface

Γ ¼ M=ðA N A Þ

Q ¼ ð4π=λÞ sin θ where λ is the incident neutron wavelength. Measurements can be made with reflectometers in monochromatic mode by varying the angle or in time-of-flight mode with a polychromatic pulsed neutron beam such that the wavelength of each detected neutron is determined. In this study, to collect data over a wide range of Q, data were collected in time-of-flight mode at several angles. Neutron scattering is a nuclear interaction, and information is derived via the scattering length density of the material, given by X F ¼ N i bi where Ni is the number density of the element, and bi is the coherent neutron scattering length. The scattering length density, F, determines the neutron refractive index, and the reflectivity can be calculated from the profile of refractive index as a function of the thickness of the adsorbed layer. The scattering length density is related to the volume fraction of each component in the layer by F ¼ js Fs þ jw Fw and js þ jw ¼ 1 where js and jw are the volume fractions of surfactant and water and Fs and Fw their scattering length densities. If a surfactant molecule has a total scattering length bs (from the sum of b for all component atoms), the area per molecule, A, in an interfacial layer of thickness, t, is A ¼ bs =ðtjs Fs Þ It is important to recognize that the area, A, calculated in this way represents an average over the entire surface rather than providing information directly about the local lateral packing of (16) Kershner, R. J.; Bullard, J. W.; Cima, M. J. Langmuir 2004, 20, 4101–4108. (17) Franks, G. V.; Meagher, L. Colloids Surf., A 2003, 214, 99–110. (18) Zhang, L.; Tian, C.; Waychunas, G. A.; Shen, Y. R. J. Am. Chem. Soc. 2008, 130, 7686–7694.

14568 DOI: 10.1021/la101969p

where M is the molecular mass of the material, A is the area per molecule and NA is Avogadro’s constant. Making measurements with materials with known different hydrogen/deuterium compositions allows ambiguities associated with the loss of phase information common in scattering measurements to be overcome. This permits unambiguous determination of parameters such as the thickness and composition of an adsorbed layer. The optical matrix method of Abeles19 can be used to calculate neutron reflection R(Q). In the present study, we have used a model of three layers to represent a surfactant bilayer where the structural parameters are the thickness of head and tail regions and area per molecule.20,21 When considering the packing of surfactants, it is useful to think about the area occupied by a molecule, Am in half of the bilayer. The total surface excess for both parts of the bilayer is described by Γ ¼ 2M=ðAm N A Þ with Am ¼ 2A The equations for surface excess and area per molecule can be applied to individual layers in a multilayer structure and to parts of individual molecules such as heads and tails. Modeling and fits to the reflectivity data were made with the computer programs bike, mbike,22 and spots23 by Rennie. In contrast to rudimentary programs for analysis of neutron reflection data that simply fit structures with a defined number of layers, the programs bike and mbike permit constraints to be imposed on the structural models such as stoichiometry of heads and tails, as well as symmetry for a bilayer structure if necessary. The molecular volumes are defined as in Table 1, and the parameters that are minimized are the average area per molecule, thickness of each region, and background, if appropriate.

Experimental Section Sodium bis(2-ethylhexyl)sulfosuccinate (98% purity), NaAOT, was obtained from Aldrich and then purified by liquid-liquid extraction using the procedure of Li et al.,12 where NaAOT was dissolved in pure water. Impurities were removed by extraction with heptane and the product was obtained by freezedrying from the aqueous phase. NaCl obtained from Fluka (g99.5% purity), NaNO3 obtained from Sigma (>99% purity), NaOH (AnalaR) obtained from VWR, and HCl (37% solution, (19) Abeles, F. Ann. Phys. 1950, 11, 307–309 & 310-314. (20) Penfold, J.; Thomas, R. K. J Phys.: Condens. Matter 1990, 2, 1369–1412. (21) Lu, J. R.; Thomas, R. K.; Penfold, J. Adv. Colloid Interface Sci. 2000, 84, 143–304. (22) Rennie A. R. http://material.fysik.uu.se/Group_members/adrian/bike. htm. (23) Rennie A. R. http://material.fysik.uu.se/Group_members/adrian/spots. htm.

Langmuir 2010, 26(18), 14567–14573

Hellsing et al.

Article Table 1. Materials Used and Values of Molecular Parametersa

name

formula

formula mass/g mol-1

Sapphire Al2O3 NaAOT C20H37NaO7S Heavy water D2O Water H2O DHO 50% D2O NaAOT tail C8H15 NaAOT head C4H7NaO7S a Note that there are two tails per AOT molecule.

102.0 444.6 20.0 18.0 19.0 111.0 222.6

normapureAR, Prolabo) were all used as received. Water was obtained from a Millipore system and D2O from EURISOTOP, CEA, Saclay, and from Sigma Aldrich. The sapphire surface was obtained from PiKem and cleaned by “dilute piranha” as described by Turner24 with a concentration of 5:4:1 of water, concentrated sulfuric acid, and 30% hydrogen peroxide25 at a temperature near 80 °C for 15 min, followed by extensive rinsing by Millipore water. The other parts of the cell and connecting tubing were cleaned with Decon 90 followed by extensive rinsing. A more detailed description of the cell and its use for small-angle scattering and grazing incidence scattering, as well as reflection, will be published in a forthcoming paper.15 The neutron reflection experiments were performed at the ISIS Facility Rutherford Appleton Laboratory, U.K., on the SURF horizontal surface neutron reflectometer26 and at the Institut Laue Langevin, France, on D17 vertical surface neutron reflectometer.27 On both instruments, the angle and size of the incident beam were defined by two sets of motorized slits before the sample. The slits were chosen so that the illuminated area of the sample was approximately constant for all measurements and set to be about 35  35 mm2. The data from both instruments were regrouped into constant intervals of dQ/Q of 5% that approximate the chosen resolution of the experiment. Stock solutions of surfactant and salts were prepared and injected into the reflection cell along with the desired amount of solvent with an HPLC pump. All experiments were carried out at 298 K by circulating water from a thermostat through the cell mount. Data were normalized to an absolute scale of reflectivity by measuring the direct beam transmitted through the sapphire with the slits used for measurements of reflectivity at the lowest angle using standard software at each facility. A few data sets from SURF had to be corrected due to an error in the motor position on one of the slits after the sample, resulting in an incorrect opening being set and a subsequent slight attenuation of the beam. This attenuation results in a slight drop in overall intensity but does not affect the shape of the profile, and so, the data are easily corrected by multiplying with an appropriate scale factor. The scale factor was determined by comparing these data to previous unaffected runs. For data measured with D17, the background was estimated by integrating measured intensity from areas on the position sensitive detector adjacent to that of the specular peak and was subtracted from the reflectivity curves. Data from SURF were measured with a single detector and modeled assuming a Q independent background.

Results and Discussion Structure and Adsorption Isotherm in Pure Water. Neutron specular reflectivity profiles were measured for hydrogenous (24) Turner, S. F. PhD Thesis, Cambridge University, United Kingdom, 1998. (25) Note: Work with this strong oxidizing acid must be performed in a hood and protective clothing must be worn. (26) Penfold, J.; Richardson, R. M.; Zarbakhsh, A.; Webster, J. R. P.; Bucknall, D. G.; Rennie, A. R.; Jones, R. A. L.; Cosgrove, T.; Thomas, R. K.; Higgins, J. S.; Fletcher, P. D. I.; Dickinson, E.; Roser, S.; McLure, I. A.; Hillman, A. R.; Richards, R. W.; Burgess, A. N.; Simister, E. A.; White, J. W. J. Chem. Soc., Faraday Trans. 1997, 93, 3899–3917. (27) Cubitt, R.; Fragneto, G. Appl. Phys. A: Mater. Sci. Process. 2002, 74, 329– 331.

Langmuir 2010, 26(18), 14567–14573

F/10-6 A˚-2

density/g cm-3

volume/A˚3

b/fm

5.75 0.64 6.35 -0.56 2.90 -0.14 2.03

4.0 1.16 1.11 1.00 1.05 0.92 1.57

636 30 30 30 200 236

40.60 19.05 -1.68 8.69 -2.76 47.85

Figure 1. Reflectivity data for solutions of NaAOT in D2O with increasing concentration are shown in (a) and indicate the increasing adsorption to sapphire. Each successive data set is multiplied by 10 for clarity. The continuous lines are the model fits. (b) Simultaneous fit (continuous lines) of the bilayer model fit to data measured with two contrasts of water for 1.2 mM NaAOT.

NaAOT in two water contrasts (see Figure 1) in order to establish adsorption and determine the structure and quantity of the adsorbed layer. Measurements were made increasing the concentration of NaAOT. This was followed by rinsing and a further measurement with pure D2O in order to establish reversibility of the adsorption. NaAOT was found to adsorb to sapphire surface as shown by the data in Figure 1, forming a bilayer at the interface. The structure of the bilayer was determined using reflectivity profiles for DOI: 10.1021/la101969p

14569

Article

Hellsing et al.

Table 2. Parameters Describing the Adsorbed Layer of NaAOT at Different Concentrationsa NaAOT/mM thead/ A˚

ttail/ A˚

Am/ A˚2

nwater per head Γ/mg m-2

0.2 138 ( 23 121 1.1 ( 0.2 0.7 10 ( 2 15 -0/þ2 63 ( 6 16 2.3 ( 0.1 1.2 10 ( 2 15 -0/þ2 60 ( 6 14 2.5 ( 0.1 2.5 10 ( 2 15 -0/þ2 57 ( 6 12 2.6 ( 0.1 7.4 10 ( 2 15 -0/þ2 51 ( 9 9 2.9 ( 0.2 a For the lowest concentration (0.2 mM), it was not possible to reliably determine values to characterize the internal structure in the adsorbed layer. The thickness of the hydrophobic region, ttail, cannot be less than 15 A˚ to maintain a realistic density. The optimized fit is 15 A˚, and so, the lower estimate of the uncertainty is taken as zero.

hydrogenous NaAOT in two different contrasts of solvent that is plotted in Figure 1b. The thickness and the area give a volume that was constrained to be large enough to contain at least two surfactant molecules. Any space not filled by surfactant is occupied by solvent. The parameters for the fitted model are given in Table 2. These are constrained by the molecular volumes given in Table 1. In the model for a bilayer of NaAOT, the thicknesses of the inner and outer head regions are equal and the bilayer is symmetric. For measurements with solutions in D2O, that are close to contrast match of the scattering length density to the sapphire substrate, the measured signal comes almost exclusively from the adsorbed material. The signal is dominated by the large contrast between the hydrogenous hydrocarbon tails and both the substrate and the solvent that are similar in scattering length density. This showed clearly that the tails pack in a thin layer of about 15 ( 2 A˚. The fitted scattering length density corresponds to a density of hydrocarbon tails of 0.92 g cm-3. This is a very high density for amorphous or liquid hydrocarbons and was observed even for low NaAOT concentrations. Figure 1 shows the error bars that arise from statistics of counting in these measurements. In practice, the largest errors arise from other sources such as the precision of the slit opening and the monitor response. This gives an estimated uncertainty of about (10%. As discussed previously,28 the assumptions made in the treatment of the background could be the most significant source of error for data measured when there is a low surface excess. Over the concentration range of this study, 0.2 to 7.4 mM NaAOT, the area per two molecules in the bilayer changed from 138 ( 25 to 51 ( 4 A˚2, and this corresponds to the surface excess increasing from 1.1 to 2.9 mg m-2 as shown in Figure 2. As the concentration of NaAOT was increased, the surfactant molecules pack laterally more closely as shown schematically in Figure 3. The total thickness of the layer remains constant at 33 ( 2 A˚, which corresponds to hydrated head regions of thickness 10 A˚. At the highest concentration, there are nine water molecules associated with each head region. Details of the structural changes that occur at high concentrations are difficult to resolve from data that is restricted both by the limit of maximum Q that is observable and by experimental uncertainty. The parameters in Table 2 imply that a small fraction of the tails would need to be in the headgroup regions rather than in the densely packed hydrocarbon region for the model of molecular packing to apply. This would amount to less than 5%, or under half a CH2 per tail, in a slightly more polar environment at the highest concentration. This detail would give insignificant changes to the reflectivity in the measured range of Q. In comparison with the total estimated experimental error of about 10% in most parameters, this effect is (28) Kwaambwa, H. M.; Hellsing, M.; Rennie, A. R. Langmuir 2010, 26, 3902– 3910.

14570 DOI: 10.1021/la101969p

Figure 2. Surface excess of NaAOT at the sapphire surface. The critical micelle concentration according to Li et al.8 is indicated with an arrow.

Figure 3. Schematic diagram of adsorbed layer of NaAOT and the effect of increasing concentration.

small and there is a relatively large uncertainty in the details of the mixing of different parts of the molecule, whereas the presence of a thin, densely packed region of hydrocarbon is unambiguous. The dimensions found for the NaAOT layer are similar to those that were observed for adsorption of a monolayer at a hydrophobic surface.9 The critical micelle concentration of NaAOT is 2.5 mM (see, for example, Li et al.8) and the concentration at which the plateau in the adsorption isotherm shown in Figure 2 is observed is in reasonable agreement with this value. It has been reported from neutron reflection12 and surface tension29 measurements that there is a high surface excess even for low concentrations at the air/NaAOT solution interface. However, the excess is only about 1 mg m-2, and although one might expect twice that amount for an interface where the surfactant packs as a bilayer, the present study indicates a packing of molecules at the highest concentrations that is about 50% denser than at the air/solution interface, as seen by the plateau in Figure 2. Although specular neutron reflection does not provide direct information about lateral packing of molecules in an adsorbed surface layer, some inferences can be drawn in this experiment. The scattering length density that is fitted to the experimental data of the “tail” or hydrocarbon-rich region above about 1 mM concentration is very low and corresponds to a hydrocarbon density of about 0.92 g cm-3. There is no space for water within this region, and so, it is apparent that at these concentrations the adsorbate forms a continuous, uniform layer. In studies of some other surfactants, only incomplete monolayers or bilayers have been observed even at high concentrations. For example, studies of the adsorption of the nonionic surfactant hexaoxyethylene glycol monododecyl ether (C12EO6) to silica30 found bilayers with (29) Datwani, S. S.; Stebe, K. J. Langmuir 2001, 17, 4287–4296. (30) McDermott, D. C.; Lu, J. R.; Lee, E. M.; Thomas, R. K.; Rennie, A. R. Langmuir 1992, 8, 1204–1210.

Langmuir 2010, 26(18), 14567–14573

Hellsing et al.

Article

Figure 4. Effect of addition of NaCl to a solution of 0.5 mM NaAOT. Reflectivity curves for solutions in D2O against the sapphire substrate. Each of the curves with salt is shifted by 1 for clarity.

only 75% at most of the maximum packing density at the critical micelle concentration. There are theoretical ideas supported by experimental data that suggest that aggregation or formation of admicelles at a surface can occur at low coverage in a wide range of systems,31 but our present study does not allow us to make any detailed statement about the lateral structure for the one concentration (0.2 mM) of NaAOT in water without complete coverage. The measurements that were made with pure D2O after the adsorption of NaAOT showed that, within the accuracy of the technique, all surfactant could be removed by rinsing. Adsorption at the sapphire interface is therefore completely reversible. This contrasts with the adsorption at the alkyl layer grafted on silica9 where the surfactant was not removed by rinsing with water. A small air bubble was observed for one measurement with a horizontal surface for the lowest surfactant concentration (0.2 mM NaAOT Figure 1). This could be included in the model of the interface using the program spots23 and it was found that the bubble covered just 5% of the total interface. This accounts for a slightly larger estimated error for this data point in Figure 2. Effect of Added Electrolyte. The reflectivity profiles of NaAOT with varying concentrations of NaCl were measured and are shown in Figure 4. These data show the effect of ionic strength on adsorption of NaAOT. It was found that the different ions such as NaNO3 shown in Figure 5 had very similar effects. Parameters for the model fits to these data are shown in Table 3. Surface tension measurements reported by other authors32 suggest that the nature of the ion is also not important at the airsolution interface. Increasing the electrolyte concentration causes the NaAOT layer thickness to decrease from 33 ( 2 A˚ to 23 ( 2 A˚, and the surface excess was found to increase from 0.9 ( 0.2 mg m-2 for pure 0.5 mM NaAOT to 1.8 ( 0.2 mg m-2 with addition of 1 mM salt (see Figure 6). Increasing the concentration of salt from 0.5 mM to 5.0 mM did not change the surface excess significantly, as shown in Figure 6. The change in the critical micelle concentration of NaAOT with added salt has previously (31) Israelachvili, J. Langmuir 1994, 10, 3774–3781. (32) Umlong, I. M.; Ismail, K. J. Colloid Interface Sci. 2005, 291, 529–536.

Langmuir 2010, 26(18), 14567–14573

Figure 5. Comparison of the effect of 1 mM salt and acid or base at the same ionic strength on the adsorption of NaAOT. The reflectivity data shown are for 0.5 mM solutions in D2O. For comparison, the data for a clean substrate and a solution without salt are also shown. The curves for NaAOT with added electrolytes are similar, and so, successive curves are shifted by a factor of 10 for clarity. The continuous lines show the models described by the parameters in Table 3. Table 3. Parameters Used to Model Adsorbed Layer of 0.5 mM NaAOT with Different Concentrations of Various Saltsa salt

[salt]/mM

thickness/A˚

Am/ A˚2

nwater

Γ/mg m-2

NaCl 0.0 33 ( 2 166 ( 30 26 0.9 ( 0.2 NaCl 0.5 23 ( 2 72 ( 7 4 2.0 ( 0.2 NaCl 1.0 23 ( 2 80 ( 9 6 1.8 ( 0.2 NaCl 5.0 23 ( 2 70 ( 7 3 2.3 ( 0.2 0.5 24 ( 3 120 ( 20 27 1.2 ( 0.3 NaNO3 1.0 23 ( 3 74 ( 7 4 2.0 ( 0.3 NaNO3 5.0 24 ( 3 74 ( 7 8 2.0 ( 0.3 NaNO3 HCl 0.5 22 ( 3 73 ( 7 3 2.0 ( 0.2 HCl 1.0 23 ( 3 67 ( 7 2 2.2 ( 0.2 NaOH 0.5 23 ( 3 70 ( 7 3 2.1 ( 0.2 NaOH 1.0 23 ( 3 70 ( 7 3 2.1 ( 0.2 a In all cases, the tail region had a thickness of 10 or 11 A˚ and the remainder of the total thickness is attributed to the head region. The distribution between inner and outer head regions was not very sensitive, but no improvements to the fits were obtained by allowing the difference in thickness to exceed 2 A˚.

been studied by measurement of the surface tension,12 and it was found that NaCl had very little effect on the surface tension at the concentrations comparable with those used in the present study. The adsorption of 0.5 mM NaAOT with 0.5 mM NaNO3 is lower than that seen for the other solutions with added salt. There is no clear explanation for this effect with nitrate being less than chloride ions. The measurements after rinsing indicated that the surface was clean of surfactant prior to the injection of both this solution and the comparable solution with NaCl, and so, there was no residual NaAOT from previous experiments. The interpretation of the structure and the amount of adsorbate is made by model fitting to the full reflectivity curve, but it is useful to consider whether this change in reflectivity is significant with respect DOI: 10.1021/la101969p

14571

Article

Figure 6. Surface excess for solutions of 0.5 mM NaAOT as a function of the concentration of added salt.

Figure 7. Surface excess of 0.5 mM NaAOT as a function of pH derived from neutron reflection data. The points at pH 3.5 and pH 9 correspond to addition of 1 mM HCl and 1 mM NaOH, respectively. The reflectivity curves for these points can be compared with those for other salts at this concentration that are shown in Figure 5.

to both random counting errors and systematic errors. The reflectivity for the solution with 0.5 mM NaNO3 is more than 5 times larger in the region of Q around 0.05 A˚-1, and so, the distinction of these amounts of adsorbed surfactant is quite clear. Effects of pH. Although the concentration of hydrogen ions for solutions in D2O might be described by the term pD, for ready comparison with other literature, we will use pH. The solution of 0.5 mM NaAOT in D2O had a pH of 6. The effect of changing pH on the adsorption of NaAOT to sapphire was studied by adding small amounts of acid or base. Neutron reflectivity data for 0.5 mM solutions of NaAOT with increasing concentrations of both NaOH and HCl were measured in a series of experiments. Examples of these data are shown in Figure 5. The surface excess as a function of pH calculated from fits to the complete series of reflectivity data at different pH is shown in Figure 7. At pH 6 to 7, the isoelectric point, the sapphire surface will carry no net charge. If there were an effect of pH, it would be expected that adsorption increases at higher concentrations of HCl as the sapphire surface becomes oppositely charged to the anionic surfactant. However, it is seen that the effect of addition of NaOH and HCl was very similar to the effect of addition of equal concentrations of either NaCl or NaNO3. There is no significant change in the adsorbed 14572 DOI: 10.1021/la101969p

Hellsing et al.

amount or structure of the adsorbed layer with change in pH. The effect of changing pH on adsorption to sapphire can therefore be considered simply as an effect of increased ionic strength. Surface tension measurements12 have shown that the addition of HCl lowered the surface tension of aqueous solutions by 10 mN m-1. However, at the air/solution interface, the area per molecule without addition of acid or salt is significantly larger (78 A˚2) at the critical micelle concentration than seen for the packing of molecules in bilayers in the present study at the sapphire surface. This leaves free space for the molecules to pack more closely on addition of electrolyte. At the sapphire surface, the hydrocarbon tails are already at a maximum density at the critical micelle concentration. The absence of changes in adsorption of the anionic surfactant NaAOT to sapphire with pH is confirmation that the driving force for the adsorption of NaAOT is self-assembly of the hydrophobic tails rather than an effect of electrostatic attraction to the interface. In this respect, the adsorption has similarities to that seen at hydrophobic interfaces, although the molecular packing is different. The hydroxide ions that are present in the basic solution have approximately the same effect as the chloride ions when NaCl was added. There was, as described above, a small difference in adsorbed amount of surfactant from that observed with added nitrate ions. The change in layer thickness with added salt that was observed and is reported in Table 3 is suggestive of an effect of ionic association that reduces the repulsion between the opposite sides of the bilayer. This would depend largely on the sodium cations. The effects of different anions in the Hofmeister series have been described for micelles of sodium dodecyl sulfate,33 but there is apparently no clear effect on micellar structure for NaAOT.29 Packing of surfactant molecules in a planar structure such as at a flat surface could give rise to different effects. The significant feature of the data presented in Figures 5 and 7 is that different salts as well as adding either acid or base to NaAOT have broadly similar effects. There are small, systematic differences between the model curves and the data at large values of Q that were not apparent in the data for solutions without added salt. The former were measured on the D17 reflectometer and the latter with the SURF reflectometer. There is a difference between the two instruments in how the higher Q region is treated. On D17, the background (as measured around the area that contains the specular peak) is subtracted from the data, whereas on SURF, where a single detector is used which does not afford an independent measurement of the background, the usual practice is to leave the data uncorrected and add a flat background to the simulations. The difficulty for samples that contain micelles in the bulk is that they give rise to small angle scattering that leads to a nonflat background for the reflectivity. This complicates significantly the interpretation of the data at high Q. This problem arises regardless of whether it is the data or the simulation that is corrected, and so, deviations between simulations and data above 0.15 A˚-1 should be treated with caution. As mentioned previously,28,34 background subtraction can be a significant source of errors in model fits. These can be quantified and allowance made in the stated uncertainty. The models presented were constrained to the molecular parameters for the surfactant and to fit the data up to Q = 0.15 A˚-1. No variable background is included in the fits that might obscure the difference. Unconstrained multiple layer models with a variable background could undoubtedly provide curves closer to the data. However, the comparison of effects of salts, acid, and base and the conclusions about the thin (33) Ikeda, S.; Hayashi, S.; Imae, T. J. Phys. Chem. 1981, 85, 106–112. (34) Lu, J. R.; Thomas, R. K. Nucl. Instrum. Methods A 1995, 354, 149–163.

Langmuir 2010, 26(18), 14567–14573

Hellsing et al.

Article

dense layer of NaAOT are rather insensitive to the data in the region of high Q. For these reasons, it is undesirable to draw further conclusions about details of the internal structure of the adsorbed layers.

Conclusions Sodium AOT show an adsorption isotherm with an increase in adsorption to sapphire with a plateau near the critical micelle concentration. The structure of the adsorbed layers of AOT is unusual in that the hydrocarbons pack extremely densely in a thin, well-defined layer even at very low concentrations. The density of packing of the hydrocarbon chains is higher than that found in many other surfactant bilayers. Addition of salt is known to change the critical micelle concentration, and this is reflected in the measurements of adsorption with added NaCl. Comparing the adsorption from a solution of 0.5 mM NaAOT to 0.5 mM NaAOT with 0.5 mM NaCl, the

Langmuir 2010, 26(18), 14567–14573

adsorbed amount at the interface is doubled with added salt. The effect of various different monovalent salts is very similar. In particular, the effect of pH is not significant. The effects of adding NaOH or HCl at the same concentration as a neutral salt are about the same as adding sodium chloride or sodium nitrate. This study suggests that opposite charge is not a significant factor in driving adsorption of NaAOT to sapphire from aqueous solutions. Rather, the tendency of AOT to associate in planar structures such as lamellae is dominant. This may account for the similar structures that have been observed previously at other interfaces such as that of hydrophobic silica. Acknowledgment. We are grateful to the Institut Laue Langevin, Grenoble, France and the ISIS Facility, Chilton, Oxfordshire, U.K., for allocations of beam time for these measurements. The Swedish Research Council provided partial support for this work.

DOI: 10.1021/la101969p

14573