6638
Langmuir 1997, 13, 6638-6643
Adsorption of Mixed Cationic and Nonionic Surfactants at the Hydrophilic Silicon Surface from Aqueous Solution: Studied by Specular Neutron Reflection J. Penfold,*,† E. J. Staples,‡ I. Tucker,‡ and L. J. Thompson‡ ISIS Facility, CLRC, Rutherford Appleton Laboratory, Chilton, Didcot, OX11 0QX Oxon, U.K., and Port Sunlight Laboratory, Unilever Research, Quarry Road, East Bebington, Wirral, U.K. Received May 5, 1997. In Final Form: September 15, 1997X Specular neutron reflection has been used to determine the structure and composition of the mixed cationic and nonionic surfactants of hexadecyltrimethylammonium bromide, C16TAB, and hexaethylene glycol monododecyl ether, C12E6, adsorbed at the hydrophilic silica solid/water interface. Measurements have been made at two different values of pH (2.4 and 7.0), where the relative affinity for the adsorption of the two surfactants at the interface is dramatically altered. At high pH (pH ) 7.0) the composition of the adsorbed bilayer is dominated by C16TAB and the structure of the bilayer is nonuniform: that is, the surfactant layer immediately adjacent to the solid silica surface is richer in C16TAB. At the lower pH (pH ) 2.4), where the affinity for the surface of C12E6 is much greater relative to that of C16TAB, the surface composition is closer to the bulk solution composition and the structure of the adsorbed bilayer is essentially uniform.
Introduction The study of mixed surfactant adsorption is of considerable interest because the mechanisms underlying the synergistic behavior frequently exploited in the commercial applications of surfactant mixtures are not well understood. The adsorption of surfactants, and particularly mixed surfactants, is important in a wide range of technological and industrial applications, which include detergency, fabric conditioning, dyeing, mineral flotation, colloidal stability, and surface modification. In recent years specular neutron reflection has emerged as a powerful technique for the study of surfactant adsorption at the air/water interface.1 Isotopic labeling, through H/D isotopic substitution, enables the structure and adsorbed amount to be determined.2-4 For mixtures particularly this allows both the surface composition and structure5-9 to be determined. It provides a selectivity not directly available in other techniques, such as ellipsometry10 and X-ray reflectivity,11 and information complementary to the optical techniques of sum frequency and second harmonic generation.12 In the study of surfactant adsorption at the liquid/solid interface techniques such as fluorescence decay,13 solution depletion,14 ellipsom†
Rutherford Appleton Laboratories. Unilever Research. X Abstract published in Advance ACS Abstracts, November 15, 1997. ‡
(1) Penfold, J.; Thomas, R. K. J. Phys. Condens. Matter 1990, 2, 1369. (2) Simister, E. A.; Thomas, R. K.; Penfold, J.; Aveyard, R.; Binks, B. P.; Fletcher, P. D. I.; Lu, J. R.; Sokolowski, A. J. Phys. Chem. 1992, 96, 1383. (3) Simister, E. A.; Lee, E. M.; Thomas, R. K.; Penfold, J. J. Phys. Chem. 1992, 96, 1373. (4) Lu, J. R.; Li, Z. X.; Smallwood, J.; Thomas, R. K.; Penfold, J. J. Phys. Chem. 1995, 99, 8233. (5) Staples, E. J.; Thompson, I.; Tucker, I.; Penfold, J.; Thomas, R. K.; Lu, J. R. Langmuir 1993, 9, 1651. (6) Penfold, J.; Staples, E. J.; Thompson, I.; Tucker, I.; Hines, J.; Thomas, R. K.; Lu, J. R. Langmuir 1995, 11, 2496. (7) Penfold, J.; Staples, E. J.; Thompson, I.; Tucker, I. Colloids Surf. 1995, 102, 127. (8) Penfold, J.; Staples, E. J.; Cummins, P.; Tucker, I.; Thompson, L.; Thomas, R. K.; Simister, E. A.; Lu, J. R. J. Chem. Soc., Faraday Trans. 1996, 92, 1773. (9) Penfold, J.; Staples, E. J.; Cummins, P.; Tucker, I.; Thomas, R. K.; Simister, E. A.; Lu, J. R. J. Chem. Soc., Faraday Trans. 1996, 92, 1547. (10) Tiberg, F. J. Chem. Soc., Faraday Trans. 1996, 92, 531. (11) Russell, T. P. Mater. Sci. Rep. 1990, 5, 171. (12) Bain, C. D.; Davies, P. B.; Ward, R. N. Langmuir 1994, 10, 2000.
S0743-7463(97)00468-X CCC: $14.00
etry,10 X-ray reflectivity,11 and atomic force microscopy15 either provide little or no structural information or do not have the selectivity required to study mixtures. Brinck and Tiberg16 have used ellipsometry to provide indirectly information about the adsorption of mixed nonionic surfactants at the silica/water interface; where the total adsorbed amounts at different solution compositions have been measured. The ellipsometric measurements in particular offer the opportunity to obtain kinetic as well as equilibrium information. More recently, Thomas and co-workers17-21 have demonstrated that specular neutron reflection can provide detailed information on the adsorption of surfactants at the liquid/solid interface and the structure of nonionic surfactants17,18,20 and cationic surfactants19,21 have been determined at both the hydrophilic and hydrophobic solid interfaces. In contrast, there has to date been only limited neutron reflectivity measurements on surfactant mixtures adsorbed at the liquid/solid interface.22 Much of the early work in this area has been hampered by the lack of reproducibility of adsorption of the nonionic surfactants at the hydrophilic silica solid surface.18,23 The amount adsorbed can vary markedly, depending upon the detailed surface treatment and hence on the density of hydroxyl groups on the surface. It is, however, well established that the adsorption of nonionic surfactants is strongly pH dependent;17 for pH values g9.0 the nonionic surfactants desorb, whereas at low pH (∼2) they always strongly adsorbed. We have used this in the work reported (13) Levitz, P.; Van Damne, H. J. Phys. Chem. 1986, 90, 1302. (14) Hough, H. B.; Rendell, H. M. In Adsorption from solution at the liquid/solid interface; Parfitt, G. D., Rochester, G. H., Eds.; Academic Press: London, 1983. (15) Mawie, S.; Cleveland, J. P.; Gaub, H. E.; Stucky, G. D.; Hausma, P. K. Langmuir 1994, 10, 4409. (16) Brinck, J.; Tiberg, F. Langmuir 1996, 12, 5042. (17) Lee, E. M.; Thomas, R. K.; Cummins, P. G.; Staples, E. J.; Penfold, J.; Rennie, A. R. Chem. Phys. Lett. 1989, 162, 196. (18) McDermott, D. C.; Lu, J. R.; Lee, E. M.; Thomas, R. K.; Rennie, A. R. Langmuir 1992, 8, 1204. (19) McDermott, D. C.; McCarney, J.; Thomas, R. K.; Rennie, A. R. J. Colloid Interface Sci. 1994, 162, 304. (20) Fragnetto, G.; Lu, J. R.; McDermott, D. C.; Thomas, R. K.; Rennie, A. R.; Gallagher, P. D.; Satija, S. K. Langmuir 1996, 12, 477. (21) Fragnetto, G.; Thomas, R. K.; Rennie, A. R.; Penfold, J. Langmuir 1996, 12, 6036. (22) McDermott, D. C.; Kanelleas, D.; Thomas, R. K.; Rennie, A. R.; Satija, S. K.; Majkrzak, L. F. Langmuir 1993, 9, 2404. (23) Thomas, R. K.; Rennie, A. R.; Penfold, J.; Staples, E. J., unpublished results.
© 1997 American Chemical Society
Adsorption of Mixed Cationic and Nonionic Surfactants
here to adjust the relative sensitivity of the two different components in a binary surfactant mixture in order to provide insight into the role of specific interactions with the solid surface in the partitioning of mixtures at the solid interface. We have used specular neutron reflection to study the composition and structure of the mixed cationic/nonionic surfactants adsorbed at the hydrophilic solid silica/water interface. The measurements were made for an equimolar mixture of the cationic and nonionic surfactants of C16TAB and C12E6 at a concentration of 10-4 M (∼cmc of the mixture), in 0.1 M NaBr, and at two different values of pH (2.4 and 7.0). A series of different neutron reflectivity measurements were made with different isotopically labeled combinations of the surfactants and solvent. The resulting reflectivity profiles have been analyzed using models calculated for the exact optical matrix formulation.24,25 Neutron Reflectivity Specular neutron reflection gives information about inhomogeneities normal to an interface or surface, and its theory has been described in detail elsewhere.1 The basis of a neutron reflection measurement is that the variation in specular reflection with κ (the wave vector transfer defined as κ ) 4π sin θ/λ, where θ is the glancing angle of incidence and λ is the neutron wavelength) is simply related to the composition or concentration profile in the direction normal to the interface. In the kinematic approximation26 the specular reflectivity, R(κ), is given by
R(κ) )
16π2 |F(κ)|2 κ2
(1)
where F(κ) is the one-dimensional Fourier transform of F(z), the average scattering length density profile in the direction normal to the interface,
F(κ) )
∫-∞∞F(z) exp(iκz) dz
(2)
∑i ni(z)bi
(3)
and
F(z) )
where ni is the number density profile of species i and bi is its scattering length. In the context of surfactant adsorption, the key feature of the neutron reflectivity method is that the neutron scattering properties of H and D are sufficiently different that for studying adsorption at both the air/liquid and liquid/solid interfaces H/D isotopic substitution can be used to manipulate the neutron refractive index profile at that interface. This is particularly important for surfactant mixtures where, by selective deuteration, particular components or fragments can be highlighted or isolated. It is this selectivity that makes the neutron reflectivity method so powerful. The effectiveness of the method in determining the surfactant structure depends on being able to combine reflectivity profiles from solutions of the same chemical but different isotopic compositions. This, of course, assumes that there is no isotopic dependence of the structure or adsorbed (24) Heavens, O. S. Optical properties of thin films; Butterworth: London, 1955. (25) Penfold, J. In Neutron, X-ray and light scattering; Lindner, P., Zemb, T., Eds.; Elsevier: New York, 1991. (26) Crowley, T. L.; Lee, E. M.; Simister, E. A.; Thomas, R. K. Physica B 1991, 173, 143.
Langmuir, Vol. 13, No. 25, 1997 6639
amounts and this has now been well established27 for a range of systems, including that reported here. In terms of structure, the reflectivity profiles can be analyzed by, principally, two different methods. In the first method, a structural model is assumed for the interface and the reflectivity calculated exactly using the optical matrix method.24,25 The model is then optimized to fit simultaneously the reflectively profiles for the different isotopically labeled combinations. Alternatively, a more direct method, based on the kinematic approximation,3 is used to analyze the reflectivity profiles from the different isotopic combinations by separating out the contributions from the different components of the interfacial layer. This latter approach has now been extensively applied to the determination of the structure of the adsorbed surfactant layer at the air/water interface and has, for example, been used to determine the structure of the C16TAB monolayer at the air/water interface to a resolution of two methylene groups.4 For the study of adsorption at the liquid/solid interface, where the structure is more complicated and it is difficult to isolate the contributions from all the individual components, the former approach based on model fitting (using a least squares criterion) and the optical matrix method of calculation presently provides in practice the only satisfactory route to analyze the data. Experimental Details The specular neutron reflection measurements were made on the SURF reflectometer27,28 at the ISIS pulsed neutron source, using the “white beam time of flight” method. Measurements were made in the κ range of 0.012-0.4 Å-1 using a wavelength band of 1.0-7.0 Å and three different glancing angles of incidence, 0.35, 0.8, and 1.8°. A sample geometry that is now well established for liquid/solid studies is used: that is, the neutron beam is incident at glancing angles at the liquid/solid interface by transmission through the crystalline silicon upper phase. The sample cell has been previously described and was developed originally for Poiseuille shear flow measurements at grazing incidence29 but was used here for static measurements. The polished silicon single (〈111〉 face) crystal was obtained from Semiconductor Processing, Boston, and was used without any additional surface treatment. The illuminated area was 30 × 60 mm, and the resolution in κ, ∆κ/κ, was ∼4%. The data were normalized for the incident beam spectral distribution, detector efficiency, and the silicon transmission using standard procedures and established on an absolute reflectivity scale by reference to the direct beam intensity.21 The contribution to the reflectivity at the silicon/solution interface from the thin oxide layer at the silicon surface is significant and must be included in any model calculations. It is, hence, important to characterize the nature of the oxide layer at the silicon surface and measurements (in the absence of surfactant) of the silicon/water interface with different water contrasts (D2O and H2O index matched to silicon, cmSi) have been made. Reflectivity measurements were made for the equimolar mixture of C16TAB/C12E6 in 0.1 M NaBr at a concentration of 10-4 M (above the cmc of the mixture) and at two different pH values, 2.4 and 7.0 (adjusted by the addition of HCl). Measurements were made for hC16hTAB/hC12hE6, dC16TAB/hC12hE6, and hC16hTAB/dC12hE6 in D2O and dC16dTAB/dC12hE6, dC16dTAB/ hC12hE6, and hC16hTAB/dC12hE6 in H2O at pH ) 2.4 and for dC16dTAB/hC12hE6 in D2O and cmSi at pH ) 7.0. Additional measurements for 10-4 M C16TAB and C12E6 in 0.1 M NaBr as single components were also made. For C16TAB the isotopic (27) Lu, J. R.; Lee, E. M.; Thomas, R. K.; Penfold, J.; Flitsch, S. L. Langmuir 1995, 9, 1352. (28) Bucknall, D. G.; Penfold, J.; Webster, J. R. P.; Zarbakhsh, A.; Richardson, R.M.; Rennie, A. R.; Higgins, J. S.; Jones, R. A. L.; Fletcher, P. D. I.; Roser, S.; Dickinson, E. Proc. ICANS-XIII, PSI Proc. 1995, 95-02, 123. (29) Penfold, J.; Staples, E.; Tucker, I.; Fragnetto, G. Physica B 1996, 221, 325.
6640 Langmuir, Vol. 13, No. 25, 1997
Penfold et al.
Figure 1. Neutron reflectivity profiles for the system Si/SiO2/ water (side A) where the isotropic composition of the water is (b) D2O, (O) H2O, and (4) cmSi. The solid lines are profiles calculated using the parameters in Table 1. Table 1. Parameters Used To Fit the Oxide Layer on Silicona ×10-6 Å-2 oxide (side A) oxide (side B)
t ((2 Å)
F D 2O
FH2O
Fsi
20.0 20.0
5.5 5.4
1.9 3.2
2.9 3.5
a d is the thickness of the layer, and F is the scattering length si density of the oxide layer in water of composition i.
combinations of hC16hTAB in D2O and cmSi and dC16dTAB in H2O and cmSi were measured. For C12E6 the isotopic combinations of hC12hE6 in D2O and cmSi and dC12hE6 in H2O and cmSi were measured at pH ) 7.0, and hC12hE6/D2O and dC12hE6/H2O, at FH ) 2.4. All the measurements were made at a temperature of 30 °C, unless otherwise stated, and in 0.1 M NaBr. The cmc of the mixture was previously measured by surface tension to be 5.8 × 10-5 M.8 hC12hE6 was obtained from Nikkol and purified by chromatography,27 and the other isotopes of C12E6 and C16TAB were synthesized by R. K. Thomas at Oxford: the details of the preparation, purification, and characterization are given elsewhere.26,27 High-purity water was used for all the measurements (Elga Ultrapure) and the D2O was obtained from Fluorochem. All glassware and the sample cell were cleaned using alkaline detergent (Decon 90) followed by copious washing in ultrapure water, rinsing to nonfoaming end points.
Results and Discussion Surface (Oxide Layer) Characterization. Figure 1 shows the specular neutron reflectivity from the silicon/ SiO2/water interface, without surfactant, and for three different water contrasts (or refractive indices), D2O, H2O and cmSi. The solid lines are calculations assuming that the oxide is a thin layer of uniform density. Both sides of the silicon block were polished, and the two different sides were used to study the adsorption of the mixed surfactants at the different values of pH. Reflectivity measurements to characterize the nature of the oxide layer on both sides were made, and the results are summarized in Table 1. The thickness of the oxide layer on each side was within error the same (20 ( 2 Å), but the composition of the two oxide layers was slightly different. The variation in scattering length density of the oxide layer with solvent contrast is indicative of a solvated oxide layer. The composition of the layer can be estimated from
FSiO2(1 - φ) + Fs(φ) ) Fi
Figure 2. Neutron reflectivity profiles for 10-4 M C12E6 adsorbed at the Si/SiO2/water interface: (b) hC12hE6/D2O; (O) dC12hE6/H2O; (3) dC12hE6/cmSi; (4) hC12hE6/cmSi. The solid lines are profiles calculated using the bilayer model described in the main text and the parameters in Table 2.
respectively; φ is the volume fraction of solvent in the hydrated oxide layer. For side A the volume fraction of solvent, φ, is ∼24%, whereas for side B it is ∼32%. Although the differences between the two oxide layers are not substantial, in terms of the subsequent analysis of the surfactant adsorption they are significant and emphasize the need to properly characterize the bare interface. The parameters describing the oxide layer are consistent with other recent measurements20,21 and are typical of the native oxide layer that forms on the 〈111〉 face of silicon. The parameters in Table 1 used to describe the oxide layer at the interface are included in the subsequent analysis of the surfactant adsorption data. Adsorbed Surfactant Layers of C12E6 and C16TAB. Prior to the measurements of the adsorption of the C16TAB/C12E6 mixture, the adsorptions of C12E6 and C16TAB alone were measured. Figure 2 shows the specular neutron reflectivity for 10-4 M C12E6 at the Si/SiO2/water interface in 0.1 M NaBr (side A) for different surfactant and solvent contrasts. As a consequence of using partially labeled C12E6 (dC12hE6), it was not possible to fit all the different contrasts measured for C12E6 using a single layer of uniform density and a more detailed model is necessary. It has been well established by reflectivity and SANS measurements30,31 on nonionic surfactant adsorption and reflectivity measurements on cationic surfactant19 that the nature of the adsorbed layer is a fragmented bilayer. The adsorbed layer can then be described by three layers, a layer of thickness d1 adjacent to the solid surface containing headgroups and the associated hydration, a layer of thickness d2 containing the hydrocarbon chains interpenetrating (or overlapping) from both sides of the bilayer, and a layer of thickness d3 adjacent to the fluid phase containing headgroups and hydration. In many cases, where the structure of the bilayer is not sufficiently well ordered, an additional parameter, fc, is required, which allows for some intermixing between the headgroup and alkyl chain regions. fc is then the fraction of alkyl chains in the headgroup region. The model can then be described by three thicknesses, d1, d2, and d3, the area per molecule in the bilayer, A, fc, and the fractional coverage, f (the fraction of the surface covered by bilayer patches). From the known molecular volumes and scattering lengths, the scattering length densities of each of the layers can then be estimated.
(4)
where FSiO2, Fs, and Fi are the scattering length densities of SiO2, water, and of the layer (from the model fit),
(30) Cummins, P. G.; Staples, E. J.; Penfold, J. J. Phys. Chem. 1990, 94, 3740. (31) Cummins, P. G.; Staples, E. J.; Penfold, J. J. Phys. Chem. 1990, 94, 8092.
Adsorption of Mixed Cationic and Nonionic Surfactants
Langmuir, Vol. 13, No. 25, 1997 6641
Table 2. Parameters Used To Fit the C12E6 Bilayer, Adsorbed at the Si/SiO2/Solution Interface, Using the Model Described in the Text block/conditions
d1
Å d2
d3
fc
A (Å2)
f
Γ (×10-10 mol cm-2)
side A, pH ) 7.0 side B, pH ) 7.0 side B, pH ) 2.4
23 ( 5 14 ( 3 12 ( 1
22 + 5 27 ( 5 20 ( 2
16 ( 3 26 ( 5 14 ( 2
0.0 0.0 0.2
65 ( 5 48 ( 4 34 ( 2
0.15 ( 0.02 0.09 ( 0.02 0.5 ( 0.02
0.77 0.62 4.88
Figure 3. Neutron reflectivity profiles for (b) Si/SiO2/D2O and for 10-4 M hC12hE6/D2O/0.1 M NaBr at pH ) 7.0 and (4) at pH ) 2.4.
Measurements of the adsorption of C12E6 alone at a concentration of 10-4 M and in 0.1 M NaBr were also made on side B at a pH of 7.0 and 2.4, but with only limited contrasts (at pH ) 7.0, hC12hE6/D2O, and at pH ) 2.4, hC12hE6/D2O and dC12hE6/H2O). The results of the model fits for all three measurements are summarized in Table 2: the refinable parameters, A, d1, d2, d3, fc, and f have been adjusted to give the best fit consistent with the measurements for each of the isotopically labeled combinations. The adsorption of C12E6 at the hydrophilic Si/ SiO2/water interface at pH ) 7.0 is much lower than has been previously reported. This is in part due to the presence of 0.1 M NaBr but is also consistent with other observations regarding the variability in the adsorbed amount being dependent upon the specific nature of the surface treatment.18,23 As expected, the reduction in pH (to pH ) 2.4) results in a marked increased in the adsorbed amount. This is clearly seen in Figure 3 where the reflectivity profiles for the bare interface (Si/SiO2/D2O) and hC12hE6/D2O at pH ) 7.0 and 2.4 are plotted. The reflectivity for hC12hE6/D2O at pH ) 7.0 is only slightly different from that of the bare interface, but at pH ) 2.4 there is a marked difference consistent with a substantial increase in the adsorbed amount. At pH ) 2.4 the adsorbed amount and thickness of the headgroup and alkyl chain regions are consistent with other measurements.17,19-21 At the lower coverage (pH ) 7.0) the headgroup and alkyl chain regions appear to be systematically thicker. However, at this level of coverage the adsorbed layer is much more disordered than at higher coverages, and the uncertainty in these values is correspondingly much larger. Hence the changes are probably not necessarily associated with a real change in the structure of the adsorbed surfactant bilayer. Measurements of the adsorption of C16TAB alone were made at a concentration of 10-4 M in 0.1 M NaBr for the isotopically labeled combinations of hC16hTAB in D2O and cmSi and dC16dTAB in H2O and cmSi (see Figure 4). For these labeled combinations a single uniform layer is sufficient to describe the adsorbed bilayer (see Table 3), giving rise to a bilayer of thickness ∼32 ( 2 Å and a fractional coverage ∼0.73 ( 0.04 and corresponding to an adsorbed amount of 6.8 × 10-10 mol cm-2, consistent with
Figure 4. Neutron reflectivity profiles for 10-4 M C16TAB adsorbed at the Si/SiO2/solution interface in 0.1 M NaBr: (b) hC16hTAB/D2O; (9) hC16hTAB/cmSi; (O) dC16dTAB/H2O; (2) dC16dTAB/cmSi. The solid lines are profiles calculated using the parameters in Table 3 for a single layer of uniform density. Table 3. Parameters Used To Fit the C16TAB Adsorbed Layer at the Si/SiO2/Solution Interface isotopic combination
d (Å)
F (×10-6 Å-2)
f
hC16hTAB/D2O hC16hTAB/cmSi dC16dTAB/H2O dC16dTAB/cmSi
32.0 ( 2 32.0 32.0 32.0
1.3 3.2 4.8 5.9
0.69 0.71 0.74 0.77
a d is the adsorbed layer thickness, F is the scattering length density, and f is the fractional coverage (estimated from the known molecular volumes and scattering lengths of the surfactant and solvent in conjunction with d and F).
Table 4. Parameters Used To Fit the C16TAB Bilayer Adsorbed at the Si/SiO2 Solution Interface (As Described in the Text) d1
Å d2
d3
f
fc
A (Å2)
Γ (×10-10 mol cm-2)
4.5
22.9
5.4
0.73
0.0
39.0
6.2
other measurements.19 McDermott et al.19 have, in comparision, reported a thickness of 34 Å for C16TAB adsorbed onto quartz. The data can equally well be fitted by a bilayer model similar to that used for C12E6 (see Table 4) and give values consistent with the simple single uniform layer model. The dominant interaction between the nonionic surfactants and the silica surface is due to hydrogen bonding between the ether oxygens of the ethylene oxide groups on the silica surface,31 whereas the interaction for the cationic surfactant is electrostatically driven. Mixed Surfactant Adsorption. Figure 5 shows the specular reflectivity profiles for the adsorption at the Si/ SiO2 interface from an equimolar mixture of C16TAB/C12E6 at 10-4 M in 0.1 M NaBr and at a pH of 2.4: the profiles in Figure 5a are for different isotopically labeled combinations of the two surfactants in D2O and in H2O in Figure 5b. Similar data were measured for the same mixture, but at a pH of 7.0 and with a different combination of labeled compounds (see Experimental Section for details for the combinations measured).
6642 Langmuir, Vol. 13, No. 25, 1997
Figure 5. Neutron reflectivity profiles for the mixed surfactant C16TAB/C12E6 at 10-4 M in 0.1 M NaBr at the Si/SiO2 interface for (a) (in D2O) (b) hC16hTAB/hC12hE6, (O) dC16dTAB/hC12hE6, and (4) hC16hTAB/dC12hE6 and (b) (in H2O) (b) dC16dTAB/ dC12hE6, (O) dC16dTAB/hC12hE6, and (4) hC16hTAB/dC12hE6. The solid lines are profiles calculated for the bilayer model described in the text and parameters summarized in Table 5.
The reflectivity profiles in Figure 5a,b for the two different solution contrasts (D2O and H2O) show significant changes with the different isotopically labeled surfactant combinations. An analysis of the profiles in terms of a single layer of uniform density does not provide a set of parameters for the different contrasts, which represents a consistent description of the adsorbed mixed surfactant layer. The simplest model that is consistent with all the data is a fragmented interpenetrating bilayer, similar to that used for C12E6 alone. The parameters in Table 5 give an indication of the spread in values from the fits to the different isotropically labeled combinations. The procedure adopted was to refine the individual model parameters (d1, d2, d3, fd (see later discussion), f, fc, Ac, An) for each profile on a least squares criterion to give the minimum spread in the model parameters for the reflectivities for all the different isotopically labeled combinations that were measured. The mean values and their errors are then summarized in Table 6 for the measurements at pH ) 2.4 and 7.0. The main structural features are that at pH ) 2.4 the bilayer is essentially symmetrical, the two headgroup regions are within error identical, and the alkyl chain region is less than the dimension of a fully extended chain. This is in part due to fraction of alkyl chains that are in the headgroup region, fc ∼0.2, which stems primarily from constraints arising from the disparity in headgroup size between C16TAB and C12E6. At pH ) 7.0 the structure of the bilayer is not symmetrical, although the alkyl chain region and headgroup layer adjacent to the solvent are of similar dimensions to those at pH ) 2.4; the headgroup region adjacent to the solid phase (Si/SiO2) is significantly
Penfold et al.
thinner (7.4 Å compared to 13.8 Å). This asymmetry results from the two surfactant species being unevenly distributed across the bilayer at the higher value of pH. The layer adjacent to the solid phase has a higher mole fraction of cationic surfactant than the outer layer. An additional model parameter has been introduced to accommodate this asymmetry in distribution, and fd is a weighting factor that describes that asymmetry in distribution across the bilayer. At pH ) 2.4 fd is unity, the surfactant composition is the same in both halves of the bilayer, whereas at pH ) 7.0 fd ∼0.82, indicating that there is ∼20% more cationic surfactant than nonionic surfactant in the layer adjacent to the solid phase and, consequently, the outer layer is richer in nonionic. The reduction in pH from 7.0 to 2.4 reduces the charge density on the silica surface and at pH ) 2.4 is close to the zero point of charge.32 Wa¨gnerud and Olofsson have shown that the adsorption of C16TAB is also strongly pH dependent,33 and although the adsorption decreases with decreasing pH, there is still some adsorption at pH ) 2.4. However, at low pH the adsorption mechanism is substantially dominated by C12E6. C16TAB is largely present due to the entropy of mixing. At high pH C16TAB has a strong affinity for the surface, while the presence of surface charge reduces the possibility of hydrogen bonding for C12E6. The fractional coverage of the fragmented bilayer is simillar at both values of pH and at ∼0.8 is close to the maximum value observed for C12E6 and C16TAB alone.17-22 At pH ) 2.4 the total adsorbed amount is slightly less than that at pH ) 7.0, but in both cases the amount adsorbed is significantly larger than obtained for single surfactants. At pH ) 7.0, where the affinity of C12E6 for the surface is significantly reduced, the surface composition is rich in C16TAB (mole percent of C16TAB at the interface is ∼70%). Whereas at pH ) 2.4, when the affinity of C12E6 for the surface is greatly enhanced and that of C16TAB is reduced, the surface composition is close to the bulk solution composition (mole percent of C16TAB ∼50%). McDermott et al.22 have used specular neutron reflection to study the adsorption of C16TAB/C12E6 mixed surfactants at the crystalline quartz interface in the absence of electrolyte. Measurements were made at concentrations gcmc of the mixture and for solution compositions of mole fractions of C16TAB of 0.55 and 0.92. Surface compositions, corresponding to C16TAB mole fractions of 0.05 and 0.77, respectively, were obtained for measurements with limited isotropically labeled combinations. From regular solution theory (using an interaction parameter, β ∼ -2.1, calculated from the cmc values)34 the predicted micelle compositions were 0.26 and 0.78 mole fraction of C16TAB. However, it is difficult to make a detailed comparison with our results because the measurements of McDermott et al.22 were made in the absence of electrolyte and using a different solid interface. Measurements of the composition at the air/liquid interface by neutron reflectometry8 and of the micelle composition by SANS8 have shown that the C16TAB/C12E6 mixed surfactants exhibit essentially ideal mixing. At concentrations . cmc the micelle and surface compositions should reflect the bulk composition, and this is found for many of the systems investigated,5-7 including the mixed surfactants C16TAB/C12E6.8 For nonideal mixing and even in the case of ideal mixing, if the cmc values of the two components are sufficiently different, the surface and micelle compositions can vary markedly in the region of (32) Iler, R. K. The Chemistry of silica; Wiley-Interscience: New York, 1979. (33) Wa¨gnerud, P.; Olofsson, G. J. Colloid Interface Sci. 1992, 153, 392.
Adsorption of Mixed Cationic and Nonionic Surfactants
Langmuir, Vol. 13, No. 25, 1997 6643
Table 5. Parameters Used To Fit the C16TAB/C12E6 Bilayer Adsorbed at the Si/SiO2/Solution Interface (As Described in the Text) Å2
Å isotopic combination
pH
d1
d2
d3
fd
f
fc
Ac
An
hC16hTAB/hC12hE6/D2O dC16dTAB/hC12hE6/D2O hC16hTAB/dC12hE6/D2O dC16dTAB/dC12hE6/H2O dC16dTAB/hC12hE6/H2O hC16dTAB/dC12hE6/H2O hC16hTAB/hC12hE6/D2O dC16dTAB/hC12hE6/D2O dC16hTAB/hC12hE6/H2O hC16dTAB/hC12hE6/cmSi dC16dTAB/hC12hE6/cmSi
2.4 ′′ ′′ ′′ ′′ ′′ 7.0 ′′ ′′ ′′ ′′
14.7 13.9 11.8 14.1 14.2 13.9 7.4 6.8 7.5 6.1 9.3
13.6 14.4 13.9 11.4 14.8 12.8 11.4 11.8 16.5 13.7 15.1
10.8 11.1 7.4 14.0 15.8 14.0 10.6 6.3 15.7 15.3 12.9
1.0 ′′ ′′ ′′ ′′ ′′ 0.85 0.53 0.97 0.88 0.89
0.78 0.95 0.87 0.74 0.77 0.75 0.86 1.0 0.78 0.83 0.78
0.2 ′′ ′′ ′′ ′′ ′′ 0.11 0.09 0.10 0.10 0.05
57.5 61.3 59.8 58.1 49.7 53.4 41.4 39.4 37.1 33.9 39.8
63.3 60.2 71.5 53.2 48.5 58.6 83.2 104.0 78.1 82.0 112.0
a A and A are the areas/molecule of C TAB and C E and f is the weighing factor that describes any asymmetry in the composition c n 16 12 6 d within the bilayer (see text for more explanation).
Table 6. Mean Values of Parameters from Model Fits Summarized in Table 5 Å2
pH
d1
Å d2
2.4 7.0
13.8 ( 1 7.4 ( 1
13.6 ( 1 13.7 ( 2
d3
fd
f
fc
Ac
An
Γ (×10-10 mol cm-2)
mol % C16TAB
12.2 ( 2 12.2 ( 3.5
1.0 0.82 ( 0.15
0.81 ( 0.08 0.85 ( 0.08
0.2 ( 0.05 0.09 ( 0.02
56.6 ( 4 38.3 ( 2.6
59.2 ( 7 91.9 ( 13.5
9.29 ( 0.10 10.44 ( 0.10
0.51 ( 0.05 0.71 ( 0.05
the cmc, consistent with the predictions of regular solution theory.34 For C16TAB/C12E6 at 10-4 M and in 0.1 M NaBr the surface composition at the air/solution interface is 60 mol % C16TAB/40 mol % C12E6 (the cmc of the mixture is ∼6 × 10-5 M), and at 3 × 10-4 M is 54 mol % C16TAB, close to the bulk composition. The compositions measured here for the adsorption at the hydrophilic Si/SiO2 solid interface from an equimolar mixture C16TAB/C12E6 at pH 7.0 and 2.4 are significantly different and evidence of the role of the specific surface interaction with the surfactants in determining the surface composition. In particular, the change in pH dramatically alters the surface composition, the decrease in pH resulting in a surface more rich in the nonionic surfactant. Conclusions The structure and composition of the mixed surfactant bilayer of C16TAB/C12E6 adsorbed at the hydrophilic silica solid/solution interface have been determined by specular neutron reflectivity. The change in the solution pH from 7.0 to 2.4 has a dramatic effect on both the structure and composition of the adsorbed layer, in circumstances where the relative affinity of the two surfactants for the surface has substantially altered. At a pH of 7.0 the surface composition is significantly different from the bulk composition. Previously, we have (34) Holland, P. M. Colloids Surf. 1986, 19, 171.
found by neutron reflectivity and small angle neutron scattering8 that the micelle and surface compositions are close to the bulk composition and consistent with ideal mixing. At a pH of 2.4 the surface composition is now close to the bulk composition. The application of theories, such as regular solution theory,34 would not directly predict such changes and will require modification to take into account the specific interaction of the surfactants with the solid surface, which clearly dominate the surface properties of the mixture. The results and analysis also illustrate that extreme care should be taken in the interpretation of such data. It is essential to make measurements using a range of contrasts, differently isotopically labeled combinations of surfactants and solvent. Measurements with limited isotopic combinations and simpler models (for example, single layer of uniform density) can be misleading. Indeed, it is not possible to decouple completely the determination of the composition and the structure without a very specific set of isotopically labeled combinations. Acknowledgment. We would particularly like to acknowledge the provision of the deuterium-labeled surfactants by Bob Thomas and his group at Oxford and Adrian Rennie for many informative discussions on the adsorption of surfactants at the liquid/solid interface. LA970468O
