ARTICLE pubs.acs.org/Langmuir
A Neutron Reflectivity Study of Surfactant Self-Assembly in Weak Polyelectrolyte Brushes at the SapphireWater Interface Mauro Moglianetti,† John R. P. Webster,‡ Steve Edmondson,§,|| Steven P. Armes,§ and Simon Titmuss*,^ †
Department of Chemistry, Physical & Theoretical Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, OX1 3QZ, U.K. ‡ ISIS Facility, Rutherford Appleton Laboratory, Chilton, Didcot, OX11 0QX, U.K. § Department of Chemistry, University of Sheffield, Brook Hill, Sheffield, S3 7HF, U.K. ^ School of Physics & Astronomy, University of Edinburgh, Edinburgh, EH9 3JZ, U.K. ABSTRACT: Poly(2-(dimethylamino)ethyl methacrylate) (PDMAEMA) brushes grown by surface-initiated polymerization from a polyanionic macroinitiator adsorbed at the sapphirewater interface have been used as a substrate to study the interaction between the weak polyelectrolyte PDMAEMA and the oppositely charged surfactant sodium dodecyl sulfate (SDS) with neutron reflectivity. At pH 3, multilayered structures are formed in which the interlayer separation (∼40 Å) is comparable to the dimensions of a SDS bilayer or micelle. The number of repeating layers that form depends on brush thickness, ranging from three layers in a relatively thin brush (5 nm dry thickness) to 15 layers in a relatively thick brush (17 nm dry thickness). In the 5 nm brush, addition of 0.01 mM SDS leads to brush deswelling, and the distinct layered structure only forms when the SDS concentration reaches 1 mM, with the brush reswelling slightly at 5 mM SDS. In the thicker (11 and 17 nm) brushes, distinct layered structures form at 0.1 mM SDS, in which the molar SDS/DMAEMA ratio is greater than unity. Exposing the 17 nm brush/SDS complex to 1 M NaNO3 results in the complete removal of the surfactant and recovery of the bare brush structure. At pH 9, there is significant surfactant uptake by the brush, but no multilayer structures are formed. The brush presents a high concentration of DMAEMA segments that are localized to within 5001000 Å of the sapphire interface. At pH 9 the high local concentration of hydrocarbon segments in the brush screens the hydrophobic tails of the surfactants from the unfavorable interaction with water, leading to significant surfactant uptake by the brush. At pH 3 the high local concentration of charges inside the brush additionally screens the repulsive interactions between the surfactant headgroups, making surfactant uptake even more favorable, leading to the formation of multilayered surfactant aggregates confined within the brush.
1. INTRODUCTION The interactions between polymers and surfactants in aqueous solutions have been well-studied due to their importance for technological formulations ranging from pharmaceuticals to personal care products.14 Complex formation and interfacial behavior often play an important role in such applications, hence understanding these processes is of practical as well as fundamental importance.5 Polymer/surfactant mixtures have been classified as either strongly or weakly interacting, with the strongly interacting mixtures comprising a polyelectrolyte and an oppositely charged surfactant, while the weakly interacting mixtures typically comprise a neutral polymer and an ionic surfactant.6 We have previously studied the structures formed at the airwater interface of mixtures of the weak polybase PDMAEMA and the anionic surfactant SDS, both when the polymer is cationic (pH 3) and essentially neutral (pH 9).7 Multilayer structures were formed from mixtures containing 0.1 and 1 mM SDS at pH 3 and 9, respectively, in which the interlayer separation is comparable to the dimensions of a SDS bilayer or micelle. At concentrations of surfactant close to the bulk critical r 2011 American Chemical Society
micelle concentration for SDS, the polymer is displaced from the airwater interface by a monolayer of surfactant. The interfacial structure of such polymer/surfactant mixtures is determined by a subtle balance between the bulk association properties and the interfacial behavior of each component. Experimentally, it is difficult to distinguish between complexes formed directly at the interface from complexes initially formed in bulk and subsequently adsorbed at the interface. Multilayer structures comprising ∼30 layers, with an overall dimension of ∼100 nm or more, could be associated with precipitated phases that assemble at the airwater interface. Polymer brushes grown from a solidliquid interface by surfaceinitiated atom transfer radical polymerization (SI-ATRP) provide a high local concentration of polymer chains (φpoly ∼ 0.25) confined to the vicinity of the solidliquid interface. Grafting the polymer to the solidliquid interface fixes the interfacial excess of this Received: January 17, 2011 Revised: March 3, 2011 Published: March 17, 2011 4489
dx.doi.org/10.1021/la200211x | Langmuir 2011, 27, 4489–4496
Langmuir
ARTICLE
component and allows its interaction with increasing surfactant concentration to be studied without the complications of either variation of polymer coverage or adsorption of polymersurfactant complexes formed by bulk association. Previous experimental studies of polymer brush/surfactant systems have focused on characterizing the degree of surfactant binding as a function of surfactant concentration.810 However, these studies provide no direct evidence for the structure of the interfacial layers, although in one case a lamellar structure has been proposed for a fully loaded polyacid brush in which all the counterions have been exchanged for cationic surfactant.10 In this study, we grow four PDMAEMA brushes of different dry thicknesses (referred to as ad) at a grafting density of σ ∼ 0.14 nm2 (determined in ref 11 from the surface excess of polymer and the brush swelling ratios) from a polyelectrolytic ATRP macroinitiator electrostatically adsorbed at the sapphire D2O interface. Specular neutron reflectivity is used to interrogate the structures formed by this model weak polybase brush and the surfactant SDS in both acidic and alkaline media. These measurements provide both the structure and the stoichiometry of the interfacial complex that is formed.
2. EXPERIMENTAL SECTION 2.1. Materials and Sample Preparation. 2.1.1. Materials. DMAEMA was purchased from Aldrich and purified by passing it through a basic alumina column (150 mesh) also from Aldrich. Cu(I)Br, Cu(II)Br2, 1,1,4,7,10,10-hexamethyltriethylenetetramine (HMTETA), and 2-propanol were purchased from Aldrich and used as received. A-Plane (1120) sapphire wafers, for use in the AFM measurments, were purchased from MTI Corp. A-Plane synthetic sapphire blocks (60 70 15 mm), for use in the neutron reflectivity measurements, were obtained from Dr. D. Styrkas. They were lapped and polished by PI-KEM Ltd. and finally polished for 20 min using a 0.1 μm diamond slurry in the Physical & Theoretical Chemistry Laboratory at Oxford. These sapphire substrates were cleaned using a mild piranha treatment (incubation for 1520 min in a mixture of concentrated sulfuric acid, hydrogen peroxide, and ultrapure water in a ratio of 4:1:5 at 80 C) and then subjected to an Ar/H2O plasma for a few minutes to remove any
Figure 1. Anionic macroinitiator with an overall degree of esterification of 82% (36% hydroxy groups esterified with BIBB and 46% esterified with SBA) as judged by 1H NMR spectroscopy. The macroinitiator structure shown is approximate, and random esterification of the hydroxy groups on the PGMA chains is expected.
remaining interfacial contamination and to promote the formation of hydroxyl groups at the interface. Hydrogenous SDS was obtained from BDH and purified by methods previously described.12 The solution pH was varied using HCl and NaOH (Analytical Reagents) and measured using a standard pH meter. 2.1.2. Surface Inititiated Atom Transfer Radical Polymerization. SIATRP was performed from the anionic polyelectrolyte, illustrated in Figure 1, after this macroinitiator had been electrostatically adsorbed at the sapphirewater interface. Briefly, the macroinitiator was prepared via a convenient one-pot synthesis, in which a poly(glycerol monomethacrylate) homopolymer precursor is sequentially esterified using first 2-bromisobutyryl bromide (BIBB) followed by an excess of 2-sulfobenzoic acid cyclic anhydride (SBA) to provide the anionic groups that promote its adsorption at the sapphirewater interface.13 Macroinitiator was electrostatically adsorbed at the sapphirewater interface from a 0.1 wt % solution at pH 4 for 12 h. This was confirmed by X-ray reflectivity and ellipsometry measurements (not shown), which consistently indicated a 1 ( 0.5 nm layer at the interface and also by an increase in the rms surface roughness observed using tapping mode AFM, from 0.06 nm, prior to adsorption, to 0.18 nm, after adsorption. These measurements are consistent with the adsorbed mass of 1.03 ( 0.06 mg/m2 observed for this macroinitiator at an aminated silicon oxynitridewater interface.13 SI-ATRP was conducted under a nitrogen atmosphere in a reaction medium comprising 50% DMAEMA monomer and 50% of a 9:1 v/v propan-2-ol/water mixture, with Cu(I)Br and Cu(II)Br2 as activator and deactivator, respectively, and employing HMTETA ligand to stabilize and solubilize the copper complex ([DMAEMA]:[Cu(I)Br]: [Cu(II)Br2]:[HMTETA] = 239:1:0.1:0.39).14,15 The polymerization time ranged from 30 to 60 min, and the resulting brushes (referred to as ad in Table 1) were characterized by ellipsometry, X-ray reflectivity, and neutron reflectivity measurements.11 Measurements made at pH 3, 7, and 9 showed the surface excess of polymer in each brush to be constant to within 15%. Kinetic studies confirmed an approximately linear increase in the dry brush thickness with polymerization time for approximately 100 min. FT-IR spectra recorded for sapphire substrates functionalized using this protocol display prominent bands at 2819 and 2769 cm1, which can be assigned to CH stretches of the dimethylamino groups. 2.2. Neutron Reflectivity. Specular neutron reflectivity measurements were made in time-of-flight mode using the SURF reflectometer at ISIS,16 with the neutron beam incident through the sapphire and a solution of the appropriate concentration of SDS in D2O as the subphase.17 Reflectivity profiles presented in this work are plots of the specular reflectivity with respect to the momentum transfer Q = 4π sin θ/λ, where θ is the glancing angle of incidence and λ is the neutron wavelength. Measurements were made at glancing angles of incidence of 0.1, 0.25, 0.7, and 1.5, covering a Q-range from 3.3 103 to 0.6 Å1. The reflected intensity was normalized to the incident beam spectral distribution and detector efficiency and established on an absolute reflectivity scale with a resolution ΔQ/Q = 5%. The reflectivity at Q > 0.3 Å1 is dominated by a sample-dependent background, which arises primarily from the incoherent scattering from the bulk solution. This was accounted for by the inclusion of a constant background reflectivity in the model calculations. The MOTOFIT package18 was used to fit the reflectivity profiles to model scattering length density
Table 1. Grafting Densities, Degrees of Polymerization, and Corresponding Chain Molecular Weights sample
dry thickness (nm)
γ (Å)
ΓDMAEMA (1025 mol Å2)
σ (nm2)
N
Mw (kg/mol)
a
5
47
3.5 ( 0.3
0.13 ( 0.02
155
24 ( 5
b c
11 17
100 142
7.4 ( 0.7 10.4 ( 1.0
0.12 ( 0.02 0.14 ( 0.02
443 430
70 ( 16 68 ( 15
d
17
167
12.4 ( 1.2
0.18 ( 0.03
434
68 ( 15
4490
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir
ARTICLE
Table 2. Surface Excess of SDS in the Composite Structure Formed at pH 3 in the 5 nm Dry Thickness Brush (a) (ΓDMAEMA = 3.5 ( 0.3 1025 mol Å2) [SDS] (mM) ΓSDS (1025 mol Å2) nSDS/nDMAEMA nbilayer ASDS (Å2) 0.01
0.25 ( 0.03
0.07
1
131
0.1
0.45 ( 0.05
0.13
2
148
1
0.57 ( 0.06
0.16
2
116
5
1.2 ( 0.1
0.35
3
81
φðzÞ ¼
Table 3. Surface Excess of SDS in the Composite Structure Formed at pH 3 in the 11 nm Dry Thickness Brush (b) (ΓDMAEMA = 7.4 ( 0.7 1025 mol Å2) [SDS] (mM) ΓSDS (1025 mol Å2) nSDS/nDMAEMA nbilayer ASDS (Å2) 0.0
0.1
13.3 ( 1.4
1.80
10
25
0.2
13.7 ( 1.4
1.86
10
24
1
14.9 ( 1.5
2.03
11
24
5
13.7 ( 1.4
1.86
13
32
10
14.3 ( 1.5
1.94
14
32
0.01
0.0
Table 4. Surface Excess of SDS in the Composite Structure Formed at pH 3 in the 17 nm Dry Thickness Brush (c) (ΓDMAEMA = 10.4 ( 1.0 1025 mol Å2) [SDS] (mM) ΓSDS (1025 mol Å2) nSDS/nDMAEMA nbilayer ASDS (Å2) 0.01
0.05 ( 0.005
0.005
0.1 1
13.3 ( 1.3 14.5 ( 1.4
1.28 1.39
15 15
37 34
profiles. The model brush profiles were typically constructed from three to five layers, each characterized by a thickness, scattering length density, and a Gaussian roughness. The first stage of the fitting procedure used a trial-and-error approach to determine the number of layers required to give a reasonable fit between the measured and calculated reflectivities. The scattering length density profiles were then optimized using a genetic algorithm in which the layer thicknesses, scattering length densities, and roughnesses were varied to minimize the χ2 between the measured and calculated reflectivities. To account for the Bragg peaks observed in the reflectivity from brushes b and c and for the shoulder observed in the reflectivity for brush a, it is necessary to incorporate a layered structural element into the scattering length density profile. The position of the Bragg peak (Q = 0.16 Å1) constrains the interlayer repeat to 40 Å, which is comparable to the dimensions of a SDS bilayer, and the width of the peak constrains the fitting of the number of repeated layers, denoted nbilayer and tabulated in Tables 24. The repeating unit is split into two sublayers of thickness ∼5 and ∼35 Å, with typical solvent penetrations of 55% and 16%, which can be associated with the headgroup and tail regions of the bilayer structure, respectively. In addition to determining the dimensions and number of the repeat units, we also determine the position of this layered structural element within the brush structure. The scattering length density profile of the brush region, F(z), is given by the following expression FðzÞ ¼ φDMAEMA ðzÞFDMAEMA þ φSDS ðzÞFSDS þ ð1 φDMAEMA ðzÞ φSDS ðzÞÞFD2 O
DMAEMA and SDS, respectively, and φDMAEMA(z) and φSDS(z) are the corresponding volume fraction profiles. As the scattering length densities of DMAEMA (FDMAEMA = 0.8 106 Å219) and SDS (FSDS = 0.4 106 Å2) are approximately equal and very much less than the scattering length density of D2O, FD2O = 6.35 106 Å2, then to a very good approximation the volume fraction profile of SDS/DMAEMA can be determined from
ð1Þ
where, FDMAEMA and FSDS are the scattering length densities of
FD2 O FðzÞ FD2 O FDMAEMA
ð2Þ
where φ(z) = φDMAEMA(z) þ φSDS(z). From the volume R fraction profile, the total surface excess of polymer and surfactant, γ= ¥ 0 φ(z) dz, can be evaluated. Anchoring the polymer layer ensures that the total amount of polymer at the interface is conserved throughout the experiments, as demonstrated by the recovery of the bare brush reflectivity profile following exposure of the surfactantloaded brush to a 1 M solution of NaNO3 (Figure 2c). Subtraction of the surface excess of polymer determined in the bare brush measurements11 (γDMAEMA) reported in Table 1 enables the surface excess of SDS to be determined, as reported in TablesR 25. The brush thickness is charR 2022 ¥ acterized as L = [2 ¥ 0 zφ(z) dz]/[ 0 φ(z) dz].
3. RESULTS AND DISCUSSION Four brushes (ad) were prepared with dry thicknesses in the range 517 nm and were characterized by a combination of ellipsometry/X-ray reflectivity and neutron reflectivity measurements to determine the total surface excesses of PDMAEMA, chain grafting densities, and corresponding chain molecular weights summarized in Table 1.11 Neutron reflectivity measurements were then made following exposure of these PDMAEMA brushes to solutions at the appropriate SDS concentrations in D2O at pH 3 (brushes ac) and pH 9 (brush d). The neutron reflectivity profiles measured for each of the four brushes with and without surfactant are shown in Figures 2 and 3 and the corresponding volume fraction profiles are shown in Figures 4 and 5. Tables 25 give the surface excess of SDS as a function of the SDS concentration for each of the brushes ad. For brushes ac, where distinct layered structures form, the tables also show the number of bilayers, nbilayer, and the area per SDS, ASDS, in each layer of the nbilayer bilayers. 3.1. Five Nanometer Dry Thickness Brush at pH 3. It is clear from Figure 2a that exposure of the 5 nm dry thickness brush to 0.01 mM SDS results in a significant change in the neutron reflectivity. The corresponding volume fraction profile for the brush shown in Figure 4a indicates that, in addition to a modest SDS uptake, there is a substantial redistribution of the polymer toward the sapphire interface, resulting in a reduction of the swollen thickness of the brush layer from L = 261 to 127 Å, which is also lower than the bare brush thickness at pH 7 (LpH7 = 168 Å). PDMAEMA is a weak polybase (the pKa for PDMAEMA chains is 7.07.5 in dilute aqueous solution23), so at pH 3 the acidbase equilibrium will favor protonated DMAEMA segments over uncharged segments. Local electroneutrality is ensured by the presence of OH counterions, which in the osmotic regime are confined within the brush.24 Although confined within the brush, these counterions are mobile and so exert an osmotic pressure. At pH 3 the charge density is sufficiently high for the counterion osmotic pressure to exceed that due to the excluded volume of the DMAEMA segments, 4491
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir
ARTICLE
Table 5. Surface Excess of SDS in the Composite Structure Formed at pH 9 in the 17 nm Dry Thickness Brush (d) (ΓDMAEMA= 12.4 ( 1.2 1025 mol Å2) [SDS] (mM)
ΓSDS (1025 mol Å2)
nSDS/nDMAEMA
1
4.6 ( 0.5
0.37
5
6.7 ( 0.7
0.55
10
5.3 ( 0.5
0.43
Figure 3. Specular neutron reflectivity profiles obtained from the PDMAEMA brush (d) at the sapphireD2O interface (as in Table 1) at pH 9. The reflectivity profile for the bare brush is on the correct absolute scale, whereas those with added SDS are offset by a constant scaling factor of 10.
Figure 2. Specular neutron reflectivity profiles obtained from the PDMAEMA brushes at the sapphireD2O interface (ac as in Table 1) at pH 3. In each case, the reflectivity profile for the bare brush is on the correct absolute scale, whereas those with added SDS are offset by a constant scaling factor of 10.
causing greater swelling of the bare brush at pH 3 compared to that observed at either pH 7 or pH 9.11 Exposure of the brush to 0.01 mM SDS results in a partial exchange of the OH counterions by this anionic surfactant DS.
Given that the fraction of charged DMAEMA segments at pH 3 in comparable brushes has been estimated as f ∼ 0.4,22 nSDS/nDMAEMA = 0.07 implies that 17.5% of the counterions are exchanged by DS. This exchange results in a brush deswelling, which is analogous to the pronounced volumetric reduction observed when complexes are formed between polyelectrolyte gels and oppositely charged surfactants.25 In both cases the deswelling is due to the lower osmotic pressure within the polyelectrolyte brush (or gel), which is caused by the replacement of some of the mobile OH counterions by DS. The surfactant ions are likely to adsorb onto the polymer chains as a result the hydrophobic interaction between the hydrocarbon tails of the surfactant and the polymer backbone. This binding means that the adsorbed surfactant does not contribute to the counterion osmotic pressure. The decrease in the layer thickness below that observed for the brush at pH 7 and 9 (neutral brush regime) means that there must be a further reduction in the internal osmotic pressure beyond that caused by the (partial) removal of the counterion contribution. This effect has been considered theoretically for the interaction between flexible polymers and surfactants.26 The affinity between the polymer and the surfactant induces an attractive correlation between the polymer segments that competes with the bare segmentsegment repulsion. This has the effect of reducing the second virial coefficient of the polymer. In the case of a polymer brush, such as studied here (a), this annealing of the excluded volume leads to the observed deswelling.27 4492
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir
ARTICLE
Figure 5. Interfacial volume fraction profiles determined for the PDMAEMA brushes (d as in Table 1) at pH 9 in the absence of SDS (dotted line) and following incubation with the indicated concentration of SDS.
Figure 4. Interfacial volume fraction profiles determined for the PDMAEMA brushes (ac as in Table 1) at pH 3 in the absence of SDS (dotted line) and following incubation with the indicated concentration of SDS.
As the surfactant concentration is increased, a shoulder develops in the reflectivity profile at Q = 0.15 Å1, which can be associated with the emergence of distinct layers in the volume fraction profile. The overall brush layer thickness remains approximately constant until the SDS concentration reaches 5 mM, at which the layered structure becomes particularly
prominent and the overall thickness increases to 150 Å. At this point, it can be estimated that approximately 90% of the counterions have been exchanged by surfactant. At this surfactant loading there are insufficient DMAEMA residues that remain undecorated by surfactant molecules for the surfactant-induced attractive correlation to make a significant enough contribution to the effective second virial coefficient of the polymer to overcome the increase in layer thickness necessary to accommodate the doubling of the surfactant load. The widths of the inner and outer peaks in the volume fraction profile are 28 and 38 Å, respectively. The latter is comparable to the dimensions of a typical SDS bilayer or micelle. If these prominent peaks are assumed to be bilayer-like structures, and it is further assumed that all the SDS is incorporated into these nbilayers, then it is possible to estimate a lower limit for the area (ASDS) occupied by each SDS in the structure. The evolution of the volume fraction profile suggests that these layers, which can be assigned as surfactant aggregates, assemble from the interior region of the brush, where the polymer segment concentration is highest. 3.2. Eleven Nanometer Dry Thickness Brush at pH 3. In this case, 0.01 mM SDS causes no visible change to the reflectivity profile, suggesting that any surfactant uptake or conformational change in the brush layer is negligible. The volume fraction profile that gives the best fit to the measured reflectivity is characterized by a layer thickness that is 7% less than the 569 Å observed for the bare brush. Although this is within the experimental uncertainty on the measurement, it does suggest that there may be a weak uptake of surfactant, as for brush a. However, the osmotic pressure from the larger number of undecorated segments of the higher molecular weight chains forming brush b resists the collapse that was observed for brush a, such that the unperturbed brush profile dominates the reflectivity. At 0.1 mM SDS a Bragg peak is observed at Q = 0.16 Å1, which is characteristic of a multilayer structure comprising 10 layers with a repeat distance of 40 Å. As the SDS concentration is increased to 1 and 10 mM, the number of layers that best fit the Bragg peak increases to 11 and 14, respectively, although there is little change in the total amount of SDS, which corresponds to an area per molecule of ∼29 ( 4 Å2. Since additional layers, which 4493
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir lead to a slight decrease in the width of the Bragg peak, form without an increase in the surfactant uptake, this suggests that there may be a structural transition. The development of the aggregate structure is accompanied by a small increase in the layer thickness from L = 570 Å in the absence of SDS to L ∼ 670 ( 20 Å in the presence of SDS concentrations in excess of 0.1 mM. In this structure the number of DS exceeds the number of charged DMAEMA segments. The excess negative charge must be balanced by counterions, which exert an osmotic pressure that is responsible for the observed increase in brush layer thickness. In most previous studies of surfactant binding by dense polymer brushes, the ratio of bound surfactant to polymer monomer does not exceed charge stoichiometry.8,9,28 To our knowledge, the only exception is for poly(acrylic acid) brushes, which display a second uptake step provided that the area per chain exceeds some minimum value, which in the case of a C12 surfactant is 8 nm2,10 comparable to the chain area in this brush (σ1 = 8.3 nm2). The estimated lower limit of the area per SDS in each layer of the 10 bilayer structure remains approximately constant at ∼25 Å2 for solution concentrations in the range 0.11 mM. Assuming the molecular volume of SDS to be ν = 417 Å3 and the hydrophobic chain length to be l = 17 Å, an area per molecule of 25 Å2 implies a packing ratio, ν/Al ∼ 1, which is suggestive of a lamellar phase.29 By analogy to complex formation with gels, Pyshkina et al. postulated the existence of a lamellar mesophase and increased chain stretching for the polyelectrolyte/surfactant complexes formed within poly(acrylic acid) brushes.10 We comment that addition of SDS at the cmc to layers of a strong cationic polyelectrolyte adsorbed at mica interfaces in a surface force apparatus leads to the development of oscillatory forces between the interfaces with a periodicity that is the same as the 40 Å bilayer repeat distance observed in this study.30 3.3. Seventeen Nanometer Dry Thickness Brush at pH 3. A Bragg peak is again observed at Q = 0.16 Å1 when the SDS concentration reaches 0.1 mM. The peak is slightly narrower and has a greater amplitude compared to that for brush b, corresponding to an increase in the number of layers to 15, with the same 40 Å periodicity. Increasing the SDS concentration to 1 mM leads to a slight increase in surfactant uptake, although the area per SDS in each layer of the 15 bilayers is slightly higher than for brush b, suggesting a less well-ordered structure. Flushing this structure with 30 mL of pure D2O, which will reduce the SDS concentration to e0.1 mM, does not disrupt the multilayer structure, as the reflectivity profile remains unchanged. This suggests that the multilayer structure may represent a trapped phase, as observed for the adsorption of aggregates from mixtures of SDS and cationic polyelectrolytes at solidliquid interfaces.5,30 Tran and Auroy reported that complexation between polyanionic brushes and oppositely charged surfactant was irreversible, attributing this to the hydrophobic interaction between the surfactant tails and hydrophobic backbone segments of the polyelectrolyte brush.28 Subsequent exposure to 1 M NaNO3 disrupts the multilayer structure, resulting in the loss of the Bragg peak from the reflectivity profile and the total surface excess returning to that observed for the bare brush prior to SDS adsorption, indicating complete desorption of the surfactant. The brush thickness in the presence of salt (L = 553 Å) is approximately 15% less than that of the bare brush at pH 3 prior to exposure to surfactant (L = 641 Å). This suggests that at 1 M NaNO3 brush c is in the salted regime.31 This is reasonable as the average concentration of DMAEMA segments within this brush
ARTICLE
layer is 1.6 M in the absence of added salt and 1.9 M in the presence of salt; moreover, not all the segments will be charged even at pH 3.22 Comparable behavior was observed for the quaternized analogue of these brushes grafted from the silica water interface at a similar density (σ = 0.123 nm2), with the crossover between the osmotic and salted regimes occurring at a NaCl concentration of 0.7 M.22 A larger deswelling (25%) was observed for sodium polystyrene sulfonate brushes grafted to well-defined gold nanoparticles in the presence of 1 M NaCl,32 which is a signature of the greater role played by the counterion osmotic pressure (as compared to the segment excluded volume) in these less densely grafted polyanionic brushes. A recent theoretical study of the influence of the addition of 1:1 electrolyte on complexes of a cationic surfactant and an oppositely charged polyelectrolyte suggested that the fraction of condensed surfactant drops discontinuously to zero at a critical salt concentration of 0.44 M, with the surfactant being replaced by the counterions of the added salt,33 consistent with our observation that the brush layer ultimately ends up in the salted regime. Although this effect is attributed to a screening of the attractive interactions between the polyelectrolyte and the surfactant headgroup and the repulsive interactions between surfactant headgroups in micellar aggregates, which is responsible for the well-documented decrease in the bulk critical micelle concentration with increasing salt,3436 the discontinuous nature of the surfactant release is dependent on the hydrophobic contribution to the interaction energy. At 1 M 1:1 electrolyte, the Debye screening length is 3.1 Å, and one would expect the bulk cmc to be lower than the 0.6 mM reported by Tanford at a NaCl concentration of 0.4 M34 and the 0.3 mM obtained by simulation at a salt concentration of 0.6 M.37 At such a high salt concentration, the contribution of electrostatic effects to the free energy of the system will be greatly reduced. 3.4. Seventeen Nanometer Dry Thickness Brush at pH 9. The behavior observed at pH 9 is clearly very different from that observed at pH 3, with no change in the reflectivity profiles observed before the SDS concentration reaches 1 mM. The reflectivity profiles develop a more prominent fringe around Q = 0.2 Å1, particularly at 5 mM SDS, which can be associated with the increase in the volume fraction of the spike layer adjacent to the sapphire interface that can be observed in Figure 5d. The oscillatory modulation at Q ∼ 0.010.02 Å1 of the bare brush reflectivity profile commences at a lower Q in the presence of 5 mM SDS and extends to higher Q, corresponding to the outward shift of the local maximum in the volume fraction profile. Table 5 indicates a surface excess of surfactant that, while lower than for the multilayer structure of brush c, is still of a comparable order of magnitude. There is a significant increase in the layer thickness from L = 536 Å in the bare brush case to L = 880 Å at 5 mM SDS. This is in marked contrast to the more modest increase from L = 640 Å for the bare brush c to L = 780 Å for brush c in the presence of 1 mM SDS. At pH 9, the DMAEMA segments are essentially uncharged and uptake of DS is driven by the more favorable environment that the polymer brush, which comprises a significant volume fraction of hydrocarbon, offers to the hydrophobic tails, which will become bound to the polymer chains. To maintain electroneutrality, the negatively charged DS must be accompanied by their Naþ counterions, which will be mobile through the brush layer, leading to an osmotically induced swelling of the brush/surfactant composite. The onset of the brush/surfactant interaction occurs at a SDS concentration of 1 mM. By analogy with the critical aggregation 4494
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir
ARTICLE
concentration (cac) defined for bulk polyelectrolyte/surfactant mixtures, the lowering of the chemical potential of the surfactant inside the brush at pH 9 (relative to a reference state of a SDS micelle) can be estimated using38
Æ[SDS]æbrush (M)
ð3Þ
to be 1.4kBT. 3.5. Driving Force for Multilayer Formation. Multilayer structures have been reported to form from mixtures of polyelectrolytes and oppositely charged surfactants at both the airwater and solidwater interfaces.7,30,39,40 However, these complex structures have generally been attributed to the formation of hydrophobic complexes in the bulk which then adsorb at the interface, in processes that have been compared to precipitation due to charge neutralization.5 As there is no free polymer in the reservoir in these experiments, then clearly a different mechanism must be responsible for the formation of well-ordered surfactant aggregates within the brush layer. The PDMAEMA brushes studied here present a high concentration of DMAEMA (typical volume fraction, φDMAEMA ∼ 0.2) segments that are localized to within ∼5001000 Å of the sapphire interface. This corresponds to a concentration of 1.5 M, which, assuming that a fraction f∼0.4 of the monomers are charged at pH 3, implies a concentration of charged segments [DMAEMAþ] ∼ 0.6 M. As the SDS/DMAEMA molar ratio in the thicker brushes (b and c) exceeds 1, the mechanism leading to the formation of the multilayer structures must involve more than just counterion exchange, in which the mobile OH conterions of the protonated DMAEMA segments are replaced by DS, which condense on the hydrophobic polymer backbones. We suggest that the high concentration of charged DMAEMA segments screen the repulsive interactions between the surfactant headgroups that would otherwise raise the chemical potential of a surfactant monomer within an aggregate above that of the free surfactant monomer at 0.1 mM, thus preventing aggregate formation. Such screening is analogous to the lowering of the cmc of a bulk surfactant solution caused by the addition of electrolyte.34 At pH 3, the onset concentration for aggregate formation inside the brush is 0.1 mM. Using eq 3 it can be estimated that the chemical potential of the surfactant aggregated inside the brush is lowered by 4.4kBT relative to that in a SDS micelle. The additional stabilization resulting from the screening of the headgroup repulsions by the field of charged polymer segments in the brush can thus be estimated as 3kBT. The average surfactant concentrations confined within the brush layer are shown in Table 6. It is clear that, for SDS concentrations at and above 0.1 mM for pH 3 and at 1 mM and above for pH 9, the average concentration of SDS inside the brush is well above the cmc of SDS. At 21 C, these average concentrations sit within the micellar region of the SDS/water phase diagram.41 However, it is also well-known that screening of headgroup interactions can lead to a transition from spherical micelles to cylindrical micelles,42 and hexagonal phases comprising cylindrical micelles have been observed for bulk mixtures of polyelectrolytes and oppositely charged surfactants, whereby small changes in the alkyl chain length of the surfactant and the polymer architecture can induce structural transformations from hexagonal cylindrical micellar to lamellar phases.4345 Cylindrical micelles have also been induced to crystallize into a hexagonal phase at the airwater interface of a semidilute
[SDS]solution (mM)
a
0.01
0.2
0.1
0.4
1 5 10
b
c
d
0
0
1.9
1.7
0.5
2.8
1.9
0.5
0.8
2.1
0.8
2.1
0.6
solution of a strong cationic polyelectrolyte (∼0.1 M monomer charges) and ≈0.1 mM SDS, in the presence of 0.1 M NaCl,40 with salt screening of electrostatic interactions being suggested as the driving force for the transition to cylindrical aggregates.
4. CONCLUSIONS Polymer brushes provide a convenient method of systematically exploring the interactions between strongly interacting polyelectrolytes and surfactants. PDMAEMA brushes of moderate grafting density exhibit significant uptake of the anionic surfactant SDS from aqueous solutions with concentrations at and above 0.1 mM at pH 3 and 1 mM at pH 9, which are the same surfactant concentrations at which we observed multilayer formation at the airwater interface of bulk mixtures of the free polymer and surfactant.7 This behavior is very different from that observed in the absence of PDMAEMA brushes, where no more than 89% of a single bilayer is formed at the sapphirewater interface at a SDS concentration of 7 mM.46 At pH 3, multilayered surfactant aggregates form within the brushes, with a periodic repeat that is consistent with lamellae of SDS bilayers or a hexagonal phase of cylindrical SDS micelles. The number of repeating layers varies systematically with the brush layer thickness. At pH 3, we suggest that the brush presents a sufficiently high concentration of charged segments to screen the repulsive headgroup interactions between surfactant molecules, promoting aggregate formation at a surfactant concentration that is nearly 2 orders of magnitude lower than the bulk cmc of SDS. At pH 9, this electrostatic screening is absent, but the uncharged hydrocarbon segments provide a more favorable environment for the hydrophobic tails of the surfactant, which provides the driving force for surfactant uptake at higher surfactant concentrations. The bare brush is less swollen than at pH 3, meaning that there is insufficient free volume to incorporate the dense surfactant phases that lead to multilayer formation. ’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]. Present Addresses
)
cac Δμ ¼ kB T ln cmc
Table 6. Average Concentration of SDS Confined within the Brushes, Æ[SDS]æbrush = ΓSDS/L, at pH 3 (brushes ac) and pH 9 (brush d)
Department of Materials, Loughborough University, Loughborough, LE11 3TU, U.K.
’ ACKNOWLEDGMENT M.M. acknowledges MRTN-CT-2004-512331 (SOCON) for the award of an Early Stage Researcher Fellowship. S.T. thanks 4495
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
Langmuir the Royal Society for a University Research Fellowship. We thank ISIS (Rutherford Appleton Laboratory, STFC) for the award of beam-time. S.P.A. and S.E. thank EPSRC for a Platform Grant.
’ REFERENCES (1) Langevin, D. Adv. Colloid Interface Sci. 2009, 147148, 170–177. (2) Bain, C. D.; Claesson, P. M.; Langevin, D.; Meszaros, R.; Nylander, T.; Stubenrauch, C.; Titmuss, S.; von Klitzing, R. Adv. Colloid Interface Sci. 2010, 155, 32–49. (3) Goddard, E. D. J. Colloid Interface Sci. 2002, 256, 228–235. (4) Brown, M. A.; Hutchins, T. A.; Gamsky, C. J.; Wagner, M. S.; Page, S. H.; Marsh, J. M. Int. J. Cosm. Sci. 2010, 32, 193–203. (5) Nylander, T.; Samoshina, L.; Lindman, B. Adv. Colloid Interface Sci. 2006, 123126, 105–123. (6) Taylor, D. J. F.; Thomas, R. K.; Penfold, J. Adv. Colloid Interface Sci. 2007, 132, 69–110. (7) Moglianetti, M.; Li, P.; Malet, F. L. G.; Armes, S. P.; Thomas, R. K.; Titmuss, S. Langmuir 2008, 24, 12892–12898. (8) Ishikubo, A.; Mays, J.; Tirrell, M. Ind. Eng. Chem. Res. 2008, 47, 6426–6433. (9) Konradi, R.; R€uhe, J. Macromolecules 2005, 38, 6140–6151. (10) Pyshkina, O.; Sergeyev, V.; Zezin, A.; Kabanov, V.; Gage, D.; Stuart, M. C. Langmuir 2003, 19, 2000–2006. (11) Moglianetti, M.; Webster, J. R. P.; Edmondson, S.; Armes, S. P.; Titmuss, S. Langmuir 2010, 26, 12684–12689. (12) Lu, J. R.; Marrocco, A.; Su, T. J.; Thomas, R. K.; Penfold, J. J. Colloid Interface Sci. 1993, 158, 303–316. (13) Edmondson, S.; Vo, C. D.; Armes, S. P.; Unali, G. F. Macromolecules 2007, 40, 5271–5278. (14) Zhang, X.; Xia; Matyjaszewski, K. Macromolecules 1998, 31, 5167–5169. (15) Xu, C.; Wu, T.; Drain, C. M.; Batteas, J.; Fasolka, M. J.; Beers, K. L. Macromolecules 2006, 39, 3359–3364. (16) Penfold, J.; et al. J. Chem. Soc., Faraday Trans. 1997, 93, 3899–9917. (17) Solution exchange was effected by injecting approximately 40 mL of the new (typically higher concentration) solution into the cell formed by clamping the sapphire substrate to a Teflon bath (volume 8 mL) while stirring continuously with a magnetic stirrer; the cell temperature was maintained at 21 C. (18) Nelson, A. J. Appl. Crystallogr. 2006, 39, 273–276. (19) An, S. W.; Thirtle, P. N.; Thomas, R. K.; Baines, F. L.; Billingham, N. C.; Armes, S. P.; Penfold, J. Macromolecules 1999, 32, 2731–2738. (20) Mir, Y.; Auroy, P.; Auvray, L. Phys. Rev. Lett. 1995, 75, 2863. (21) Tran, Y.; Auroy, P.; Lee, L. T. Macromolecules 1999, 32, 8952–8964. (22) Sanjuan, S.; Perrin, P.; Pantoustier, N.; Tran, Y. Langmuir 2007, 23, 5769–5778. (23) B€ut€un, V.; Billingham, N. C.; Armes, S. P. Polymer 2001, 42, 5993–6008. (24) Zhulina, E. B.; Borisov, O. V. J. Chem. Phys. 1997, 107, 5952–5967. (25) Hansson, P. Curr. Opin. Colloid Interface Sci. 2006, 11, 351–362. (26) Diamant, H.; Andelman, D. Macromolecules 2000, 33, 8050–8061. (27) Currie, E. P. K.; Fleer, G. J.; Stuart, M. A. C.; Borisov, O. V. Eur. Phys. J. E 2000, 1, 27–40. (28) Tran, Y.; Auroy, P. Eur. Phys. J. E 2001, 5 (1), 65–79. (29) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc., Faraday Trans. 1976, 72, 1525–1568. (30) Dedinaite, A.; Claesson, P. M.; Bergstr€om, M. Langmuir 2000, 16, 5257–5266. (31) Zhulina, E. B.; Klein Wolterink, J.; Borisov, O. V. Macromolecules 2000, 33, 4945–4953. (32) Jia, H.; Grillo, I.; Titmuss, S. Langmuir 2010, 26, 7482–7488. (33) Diehl, A.; Kuhn, P. S. Phys. Rev. E 2009, 79, 011805.
ARTICLE
(34) Tanford, C. Hydrophobic Effect; Wiley and Sons: New York, 1980. (35) Rosen, M. J. Surfactants and Interfacial Phenomena; Wiley: New York, 1978. (36) Newbery, J. E. Colloid Polym. Sci. 1979, 257, 773–775. (37) Jusufi, A.; Hynninen, A. P.; Haatja, M.; Panagiotopoulos, A. Z. J. Phys. Chem. B 2009, 113, 6314–6320. (38) Cabane, B.; Duplessix, R. J. Phys. (Paris) 1982, 43, 1529–1542. (39) Penfold, J.; Tucker, I.; Thomas, R. K.; Zhang, J. Langmuir 2005, 21, 10061–10073. (40) Vaknin, D.; Dahlke, S.; Travesset, A.; Nizri, G.; Magdassi, S. Phys. Rev. Lett. 2004, 93, 218302. (41) Kekicheff, P.; Grabielle-Madelmont, C.; Ollivon, M. J. Colloid Interface Sci. 1989, 131, 112–132. (42) Mazer, N. A.; Benedek, G. B.; Carey, M. C. J. Phys. Chem. 1976, 80, 1075–1084. (43) Zhou, S.; Hu, H.; Burger, C.; abd Chu, B. Macromolecules 2001, 34, 1772–1778. (44) Bergstr€om, M.; Kjellin, U. R. M.; Claesson, P. M.; Pedersen, J. S.; Nielsen, M. M. J. Phys. Chem. B 2002, 106, 11412–11419. (45) Claesson, P. M.; Bergstr€om, M.; Dedinaite, A.; Kjellin, M.; Legrand, J.-F.; Grillo, I. J. Phys. Chem. B 2000, 104, 11689–11694. (46) Follows, D. I. Ph.D. Thesis, University of Oxford, 2005.
4496
dx.doi.org/10.1021/la200211x |Langmuir 2011, 27, 4489–4496
