Article pubs.acs.org/Langmuir
Multivalent-Counterion-Induced Surfactant Multilayer Formation at Hydrophobic and Hydrophilic Solid−Solution Interfaces Jeffrey Penfold,*,†,‡ Robert K. Thomas,† Peixun Li,† Hui Xu,† Ian M. Tucker,§ Jordan T. Petkov,§ and Devinderjit S. Sivia∥ †
Physical and Theoretical Chemistry Laboratory, South Parks Road, Oxford OX1 3QZ, United Kingdom STFC, Rutherford Appleton Laboratory, Harwell, Didcot, OXON OX11 0QX, United Kingdom § Unilever Research and Development Laboratory, Port Sunlight, Quarry Road East, Bebington, Wirral CH62 4ZD, United Kingdom ∥ St. John’s College, Oxford OX1 3JP, United Kingdom ‡
S Supporting Information *
ABSTRACT: Surface multilayer formation from the anionic−nonionic surfactant mixture of sodium dodecyl dioxyethylene sulfate, SLES, and monododecyl dodecaethylene glycol, C12E12, by the addition of multivalent Al3+ counterions at the solid−solution interface is observed and characterized by neutron reflectivity, NR. The ability to form surface multilayer structures on hydrophobic and hydrophilic silica and cellulose surfaces is demonstrated. The surface multilayer formation is more pronounced and more well developed on the hydrophilic and hydrophobic silica surfaces than on the hydrophilic and hydrophobic cellulose surfaces. The less well developed multilayer formation on the cellulose surfaces is attributed to the greater surface inhomogeneities of the cellulose surface which partially inhibit lateral coherence and growth of the multilayer domains at the surface. The surface multilayer formation is associated with extreme wetting properties and offers the potential for the manipulation of the solid surfaces for enhanced adsorption and control of the wetting behavior.
■
INTRODUCTION Understanding the adsorption of surfactants at the solid− solution interface is important in the context of a number of important technological processes associated with soft lubrication, detergency, dyeing, mineral flotation, and enhanced oil recovery.1 The use of techniques such as neutron and X-ray reflectometry,2,3 ellipsometry,4 and atomic force microscopy5 has greatly enhanced our knowledge of the adsorption process at solid surfaces.6,7 In recent years the type of surfaces probed has expanded from the simple model surfaces such as silica, quartz, mica, and graphite to more complex model surfaces with different functionalities such as hydrophilic, hydrophobic and polymeric surfaces,6,7 amino-functionalized,8 carboxyl-terminated thiols,9 and surfaces that mimic the properties of the hair cuticle,10 the corneum stratum of the skin,11 and fabrics such as polyester and cotton.12−15 The patterns of surfactant adsorption at some standard interfaces are now well established.6,7 At the hydrophobic solid surface surfactants generally adsorb as a monolayer,6,7 similar to their adsorption at the more ideal hydrophobic air−water interface.2 At the hydrophilic surface and surfaces with other more specific functionalities. adsorption is usually cooperative and in the form of aggregates, flattened micellar structures, or fragmented bilayers or lamellae.6,7 More extensive and complex patterns of adsorption, in the form of surface multilayer structures, have been reported for a range of surfactant, © XXXX American Chemical Society
polymer−surfactant, and polyelectrolyte systems, predominantly at the air−water interface.16 At relatively high surfactant concentrations surface-induced micellar ordering in triblock copolymers17 and surface multilayer structures in dialkyl chain cationic surfactants18 have been observed at the solid−liquid and air−water interfaces. Tucker et al.19 demonstrated the formation of surface multilayers at the air−water interface for the anionic−nonionic surfactant mixture of sodium dodecyl benzenesulfonate, LAS, and monododecyl octaethylene glycol, C12E8, with the addition of Ca2+ ions. Petkov et al.20 also demonstrated similar surface structures in the sodium dodecyl dioxyethylene sulfate, SLES, and monododecyl dodecaethylene glycol, C12E12, mixture with the addition of Al3+ ions. In both of those latter cases the strong binding of the multivalent ions to the anionic headgroup resulted in the formation of the extended surface structures. Surface multilayer formation is extensively reported for surfactant−polyelectrolyte21 mixtures at the air−water interface. Polyelectrolyte adsorption onto hydrophilic surfaces can result in charge reversal.22−27 This enables the surfactant adsorption to be manipulated, for example, resulting in the adsorption of ionic surfactants onto surfaces with like charge.27 Charge reversal is also the Received: May 2, 2015 Revised: June 2, 2015
A
DOI: 10.1021/acs.langmuir.5b01610 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir
Table 1. Key Model Parameters from Modeling 1 mM 95/5 Mole Ratio SLES/C12E12/AlCl3 on Hydrophilic Silicona,b sample bare Si interface +SLES/C12E12 + 0.4 mM AlCl3
σs (±3 Å)
d1 (±1 Å)
ρ1(±0.1 × 10−6 Å−2)
σ1 (±3 Å)
d2 (±1 Å)
ρ2 (±0.1 × 10−6 Å−2)
σ2 (±3 Å)
d3 (±1 Å)
ρ3(±0.1 × 10−6 Å−2)
σ3 (±3 Å)
d4 (±1 Å)
ρ4(±0.1 × 10−6 Å−2)
σ4 (±3 Å)
9 9
15 15
2.5 2.5
12 10
31
2.8
5
34
3.8
5
17
0.7
5
a σn, dn, and ρn are the interfacial roughness, thickness, and scattering length density on the nth layer, and σs is the substrate roughness. bFor the scattering length density profiles for 0.6 and 1.0 mM AlCl3, see Figure 2.
Table 2. Key Model Parameters for 1 mM SLES/C12E12/AlCl3/NaCl Adsorption onto Cellulosea,b,c sample bare cellulose surface SLES/C12E12 + 0.0/3.0 mM AlCl3/NaCl SLES/C12E12/0.24/1.8 mM AlCl3/NaCl SLES/C12E12/0.36/1.2 mM AlCl3/NaCl SLES/C12E12/0.6/0.0 mM AlCl3/ NaCl bare cellulose surface +SLES/C12E12 /0.0/3.0 mM AlCl3/NaCl +SLES/C12E12/0.24/1.8 mM AlCl3/NaCl +SLES/C12E12/0.36/1.2 mM AlCl3/NaCl +SLES/C12E12 /0.6 0.0 mM AlCl3/NaCl
σs (±3 Å)
d1 (±1 Å)
ρ1(±0.1 × 10−6 Å−2)
ρ2 (±0.1 × 10−6 Å−2)
σ2 (±3 Å)
d3 (±1 Å)
ρ3(±0.1 × 10−6 Å−2)
σ3 (3 Å)
20 20
29 29
3.9 3.9
2.8 2.4
10 20
20
29
3.9
10
21
2.4
20
20
29
3.9
10
21
1.9
10
48
4.1
30
20
29
3.9
10
21
1.9
10
48
4.1
30
15 15
24 24
4.2 4.2
2.3 2.3
10 10
15
24
4.2
10
16
2.3
10
15
24
4.2
10
16
1.6
20
15
24
4.2
10
15
1.6
15
σ1 (±3 Å)
d2 (±1 Å)
(a) Hydrophobic Cellulose 10 21 10 21
(b) Hydrophilic Cellulose 10 16 10 16
a σn, dn, and ρn are the interfacial roughness, thickness, and scattering length density on the nth layer, and σs is the substrate roughness. bFor 0.6 mM AlCl3, the additional weak Bragg peak is not analyzed quantitatively. cFor 0.6 mM AlCl3, the additional weak Bragg peak is not analyzed quantitatively.
measurements on model hydrophilic and hydrophobic cellulose surfaces in order to extend the range and relevance of the functionalized surfaces studied. The results demonstrate for the first time surface multilayer formation at the solid interface from these relatively dilute surfactant solutions and are potentially significant for a wide range of applications.
mechanism responsible for the layer-by-layer formation of polyelectrolyte multilayer structures that are now being developed for exploitation in encapsulation and controlled release.28−31 Surface multilayer formation is observed in lung surfactants32−35 and is now recognized as an important aspect of their biological function. More generally interfacial multilayer structures are implicated in many aspects of biolubrication.36−38 The occurrence of surface surfactant multilayer structures at the air−water interface at relatively low surfactant concentrations and induced by multivalent counterions of polyelectrolytes is now well established.16,19−21 Closely associated with that structural signature is the observation of extreme and persistent wetting at the solid surface. The formation of such surface structures provides a great opportunity for enhancing the delivery of different benefit agents to surfaces and for controlling and manipulating the surface wetting and has significant implications for a wide range of potential applications. However, the formation of surface multilayer structures has not been convincingly demonstrated or characterized in these dilute systems at any solid−solution interface. In this article we have investigated, using NR, the conditions for the formation of surface multilayer structures in SLES/ C12E12 mixtures by the addition of Al3+ multivalent counterions. Measurements were made on both hydrophobic and hydrophilic silicon surfaces to establish the role of the nature of the surface and charge reversal in the structural evolution at the interface. These measurements were contrasted by similar
■
EXPERIMENTAL DETAILS
The neutron reflectivity measurements were made on the INTER39 and D1740 reflectometers at the ISIS and ILL neutron sources, respectively, using the white beam time-of-flight method. On INTER the measurements were made over a Q range (where Q, the wave vector transfer normal to the surface, is defined as Q = 4π sin θ/λ and λ and θ are the neutron wavelength and grazing angle of incidence) of 0.01 to 0.3 Å−1 using two angles of incidence 0.7 and 2.3° and a neutron wavelength range of 1 to 15 Å. On D17 a similar Q range was measured using two angles of incidence 0.8 and 3.0° and a neutron wavelength range of 2 to 20 Å. In both cases the associated Q resolution, ΔQ/Q, was ∼4%. The standard experimental arrangement for the liquid−solid interface was used,6 where the neutron beam was at grazing incidence by transmission through the crystalline silicon phase. On INTER the reflecting surface was horizontal, whereas on D17 it was in the vertical plane. The silicon, ⟨111⟩, was supplied by Crystran and polished to a surface roughness of nominally ∼5 Å rms. The illuminated area was 45 × 45 mm2, and the solvent gap was ∼0.2 mm. The data were normalized for the incident beam spectral distribution, detector efficiency, illumination, and run time and were established on an absolute scale using standard procedures.6,39,40 Specular neutron reflectivity provides information about the refractive index or neutron scattering length density in a direction B
DOI: 10.1021/acs.langmuir.5b01610 Langmuir XXXX, XXX, XXX−XXX
Article
Langmuir normal to the surface or interface. The neutron refractive index n(z) and the scattering length density ρ(z) are related by
n(z) = 1 −
λ 2ρ(z) 2π
(1)
where z is the direction normal to the interface. The variation in the reflectivity R(Q) with the wave vector transfer normal to the surface, Q, is related to ρ(z)2 and can be calculated using an adaptation of the optical matrix formulizm41 or the kinematic approximation.2 In neutron reflectivity H and D have very different scattering powers and hence by selective deuteration ρ(z) can be manipulated. This provides the sensitivity and selectivity required for the study of surfactant adsorption at the air−water2 and liquid−solid6 interfaces and has been extensively exploited. In this study the scattering length densities of the different components are ρSi = 2.1 × 10−6 Å−2, ρSiO2 = 3.5 × 10−6, ρD2O = 6.35 × 10−6, ρSLES = 0.4 × 10−6, and ρcellulose = 1.0 × 10−6. The NR data were analyzed using the optical matrix adaptation for neutron reflectivity,41 for the data where four or fewer layers can be used to describe the data. In this case each layer is characterized by a thickness d, a scattering length density ρ, and an interfacial roughness σ, as listed in Tables 1 and 2. The simplest model (fewest number of layers) consistent with the data was adopted and assessed by least squares. For the quantitative analysis of the multilayer structures a Bayesian/ maximum entropy method42 was used to establish the simplest model consistent with the data. In this case a freely variable scattering length distribution was used in which the constraints/limits associated with the scattering length density of the solvent, substrate, and adsorbate are included. The resulting output is then a scattering length distribution, as shown in Figure 2. Contact angle measurements were made using a Kruss DSA100 drop shape analyzer in order to quantify the changes in the wetting properties of a hydrophobic surface after surface multilayer formation and rinsing. Five microliters of water was used to create a droplet on the surface of a treated silicon wafer substrate. The silicon surfaces were oxidized and hydrophobized using the same hydrophobing agent as described below in the preparation of the cellulose surfaces and exposed to solutions which exhibited either monolayer or multilayer adsorption. The images were captured using the recommended manufacturer’s setting and adjusted for optimal focus. The contact angle, where possible, was evaluated from static images using the sessile drop, circle algorithm, and a manually guided fitted background. The video images were captured at a rate of 25 frames/s for a total of 100 frames. The SLES was synthesized and purified using procedures described elsewhere20 and supplied by the Oxford Isotope Facility.43 The C12E12 was custom synthesized at Unilever R&D Port Sunlight. High-purity water (Elga Ultrapure) and D2O (Aldrich) were used in rinsing and in the preparation of the solvents. Analytical grade (>99.9% purity) AlCl3 and NaCl were used. All of the glassware and sample cells were cleaned in alkali detergent (Decon 90), followed by copious washing in high-purity water. The surfaces of the silicon blocks used were prepared with four different surfaces: hydrophilic and hydrophobic silicon, hydrophobic and hydrophilic cellulose. The hydrophilic silicon surface was prepared by exposure to a “mild piranha” treatment,6 followed by hydration in water, to produce a thin oxide layer of a well-defined density, thickness, and hydrophilicity. The hydrophobic silicon surface was prepared by the deposition of 1,1,1,3,3,3-hexamethyldisilazane (Merck) from a dilute solution in chloroform to form a well-defined hydrophobic layer.44 The cellulose surfaces were prepared, as described in detail elsewhere,44 by repeated Langmuir−Blodgett, LB, deposition from a layer of trimethylsilyl cellulose, TMSC, spread from a chloroform solution in a concentration range of 0.5 to 0.8 mg/ mL to form a thin hydrophobic layer, ∼50 Å. A dipping/pulling rate of 5 mm/min at a constant surface pressure of 20 mN/m was used. Ten cycles produced a deposited cellulose thickness of ∼50 Å. The hydrophobic cellulose surface that was prepared was made hydrophilic by exposure to the vapor of concentrated HCl (10 wt %), which cleaves the surface methyl groups of the TMSC.
Figure 2. Scattering length density profiles from model fits for 1 mM 95/5 SLES/C12E12 in D2O: (a) 0.6 mM AlCl3 and (b) 1.0 mM AlCl3, as shown in Figure 1. The four different initial surfaces were all characterized in D2O prior to exposure to surfactant. The measurements with surfactant were all made for 1 and 2 mM solutions of 95/5 mol ratio SLES/C12E12 in D2O, with increasing amounts of AlCl3 or AlCl3/NaCl mixtures. As discussed later some initial measurements were made in AlCl3/NaCl mixtures to maintain constant ionic strength. Subsequent measurements were made in AlCl3 only, as the important factor was determined to be the AlCl3 concentration. The small amount of C12E12 was present to ensure that precipitation did not occur at the higher electrolyte concentrations,19 and none was observed. Each individual measurement took ∼60 min, and the solvent/solution exchanges were made using a LaChrom chromatography pump, with typically ∼10 to 20 mL exchanged between each individual measurement. From an illuminated area of 45 × 45 mm2 and a solvent gap of 0.2 mm, the sample volume was ∼0.2 mL. With an additional 0.2 mL in the plenum chambers at each end of the cell to ensure good mixing within the chambers, the total exchange volume was ∼0.8 mL. From the plug flow obtained with a low Reynolds number, fluid efficient exchange takes place and the volume is exchanged ∼10 to 20 times, with an estimated residue of
