Silica Composite

Feb 26, 2013 - Polymer/silica composite films, stable to calcination, were produced using ... were probed by time-resolved neutron reflectivity measur...
0 downloads 0 Views 604KB Size
Download PDF
Article pubs.acs.org/Langmuir

Robust Ordered Cubic Mesostructured Polymer/Silica Composite Films Grown at the Air/Water Interface Bin Yang, James A. Holdaway, and Karen J. Edler* Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, United Kingdom S Supporting Information *

ABSTRACT: Polymer/silica composite films, stable to calcination, were produced using catanionic surfactant mixtures (hexadecyltrimethylammonium bromide (CTAB) and sodium dodecyl sulfate (SDS)) and polymers (polyethylenimine (PEI) or polyacrylamide (PAAm)) at the air/water interface. Film formation processes were probed by time-resolved neutron reflectivity measurements. Grazing incidence X-ray diffraction (GID) measurements indicate that the mesophase geometry of the interfacial films could be controlled to give lamellar, 2D hexagonal, and several cubic phases (Pn3̅m, Fm3̅m, and Im3̅m) by varying the polyelectrolyte molecular weight, polyelectrolyte chemical nature, or the cationic:anionic surfactant molar ratio. On the basis of GID results, a phase diagram for the catanionic surfactant/polyelectrolyte/TMOS film system was drawn. These films can be easily removed from the interface and mesoporous silica films which retain the film geometry can be obtained after calcination; moreover, this film preparation method provides a simple way to impart polymer functionality into the mesostructured silica wall, which means these films have potential applications in a variety of fields such as catalysis, molecular separation, and drug delivery.

■

INTRODUCTION Mesoporous films are of great interest because of their wide potential applications in chemical sensors, separation, and catalysis. Evaporation-driven self-assembly methods such as dip and spin-coating have been extensively investigated to form supported films.1 However, the growth of films at the air/ solution interface has also been under development since the spontaneous formation of mesostructured free-standing silica films was reported in 1996.2,3 This method is efficient to form self-supporting films with a thickness ranging from nanometers to micrometers; it is less sensitive to the synthesis environment and also a useful geometry to understand the formation mechanisms of surfactant templated self-assembly processes.4 Silica films with two-dimensional (2D) hexagonal2,3 and cubic5 mesostructures grown using surfactant templates at the air/water interface have previously been reported; however, the purely inorganic films are brittle, which limits their practical application. Polymer/surfactant interactions have been noted to produce precipitates with mesomorphic structures,6 and these have recently been applied as templates for porous silicas. Polyelectrolyte/surfactant complexes generally form because of the Coulombic interaction between the charged functional groups of the polyelectrolyte and the oppositely charged surfactant; however, the studies using these complexes as templates have produced powders which lack a high degree of long-range order, regular morphology, and thermal stability.7,8 Recently, we reported hydrothermally stable mesoporous films with a well-ordered 2D hexagonal mesostructure, templated using CTAB/PEI complexes which form films at high pH.9 The polyelectrolyte was incorporated into the silica © XXXX American Chemical Society

wall while the surfactant templated the mesostructure in these films. Employing polyelectrolytes as part of the template for inorganic films not only makes the films stronger, more flexible, and more resistant to cracking but also introduces polymer functionality into the inorganic wall, which will be advantageous for potential applications in a variety of fields such as catalysis, molecular separation, and drug delivery. These films formed through a combined interaction between the PEI, CTAB, and silica precursor. Hydrogen bonding between the amine groups in the polyamine chains and the quaternary ammonium on the surfactant with the oxygen adjacent to silicon in the precursor facilitate the silica hydrolysis and condensation, driving formation of mesostructured particles. With continuous evaporation from the solution surface, particle combination and aggregation at the interface are responsible for the film formation.10 The pore orientation in these 2D hexagonal films is however largely parallel to the interface, preventing diffusion through the membrane. Robust films with a three-dimensionally connected cubic phase structure are more attractive for potential applications in molecular separation, diffusion, and absorption. Preparation of such structures was therefore the main goal of this study. Mixing cationic surfactant and anionic surfactant results in catanionic surfactant solutions, in which the two oppositely charged groups are distributed within mixed micelles. The micellar properties can be tailored by adjusting the competition Received: January 11, 2013 Revised: February 24, 2013

A

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

between the various molecular interactions (van der Waals, hydrophobic, electrostatic, etc.), resulting in a variety of structures, such as catanionic vesicles, salts, and micelles.11−14 Catanionic complexes are reported to interact with hydrophobically modified biocompatible polymers,15 DNA or charged polymers,16 and salts. We have used them to prepare polymer films with a range of highly ordered mesophases at the air/water interface using water-soluble polymers and catanionic surfactant mixtures.17,18 Here we report the use of these films as templates for mesoporous silica. In the present work, we have employed two polymers, which interact differently with the catanionic surfactant mixture in order to controllably synthesize robust polymer/silica composite films with a variety of ordered mesostructures at the air/ solution interface. The molar ratios of cationic to anionic surfactant, the polymer, and its concentration and molecular weight are critical for control of film mesostructure. By varying these factors, phase transitions are induced from lamellar, to 2D-hexagonal, to bicontinuous cubic, and to micellar cubic. These mesoporous silica films have improved resistance to cracking, and this film preparation method provides a simple way to impart polymer functionality into the mesostructured silica walls, which will be of use to further functionalize the membranes with active species for encapsulation or catalysis applications.

■

Figure 1. Free-standing catanionic surfactant/polymer/silica films during preparation. From left to right: film growth on surface of solution in a Petri dish, wet film just after removal from the solution interface onto a mesh pulled up from underneath the film; the same film after drying in air; calcined silica film taken from the portion of film indicated in the previous picture. The side of each cell in the plastic mesh is 1 cm. Diamond Light Source, UK. An X-ray wavelength of 0.0827 nm and a 3.2 m flight tube were used with the RAPID 2D SAXS detector, giving a Q range of 0.7−8 nm−1. The experiments were conducted in several steps. First, the mixed surfactant solution was circulated for 4 min, then the polymer solution was added, and solution circulated for another 10 min. The silicate precursor TMOS then was introduced, and the data were collected for 10 min (30 patterns at 20 s per pattern; TMOS was added at the third pattern). Finally, the collection time for each pattern was changed to 100 s, and another 10 patterns were collected. Neutron reflectivity measurements of the films grown at the air/ water interface were performed on the SURF20 (Target Station I) and INTER21 instruments (Target Station II) at the ISIS Pulsed Neutron Source facility within the Rutherford Appleton Laboratories, UK. Both instruments have been established for the study of surfaces using specular neutron reflectivity. For Target Station I, neutrons of wavelengths 0.55−6.8 Å in pulses at 50 Hz were used to give a Q range of 0.048−1.1 Å−1. The incident angles for the reflectivity experiments were 1.5° and 0.5°. The INTER instrument (Target Station II) operates at a lower frequency of 10 Hz, and the incident angle used for the reflectivity experiment was 2.3°, with data being collected between 0.035 and 0.326 Å−1. On SURF patterns could only be collected every 15 min while on INTER only 1.5 min was required to obtain each pattern. For the neutron reflectivity experiments, the solution was prepared by pouring a film forming solution in D2O into a 4 × 15 cm PTFE trough to provide a meniscus above the edge of the trough. The scattering was collected on this air/solution interface at 25 °C. Mature mesostructured polymer/silica interfacial films were characterized in situ by grazing incidence X-ray diffraction (GID) on beamline ID10B22 at ESRF, using a wavelength of 1.55 Å. Samples were contained in Teflon troughs of the same dimensions used in the neutron reflectivity experiments. GID patterns were recorded at an incident angle equal to the angle of the first-order diffraction peak measured in an initial reflectivity pattern. After growth, films were harvested from the interface on an open plastic mesh (with holes 1 cm/side) and dried. SAXS patterns from dry films were collected on a PANalytical/Anton Paar SAXSess system with a PW 3830 generator. Samples were supported in sticky tape, and a sticky tape background was subtracted from the data before analysis. Thermogravimetric analysis of the air-dried films was done using a PerkinElmer TGA 7 thermogravimetric analyzer with Puris software for data collection in terms of percentage mass and sample temperature. Dry free-standing films were cut out of the mesh holes to provide samples of 1.5−3 mg of film material and measured under nitrogen flow in the temperature range from room temperature to 800 °C at a heating rate of 5 °C min−1. Nitrogen adsorption and desorption from films at 77 K were measured using BELSORP absorption measuring apparatus. Film samples were prepared by cutting out the dry film from holes in the supporting mesh and removing the template either by calcination or by washing with ethanol. Approximately 0.1 g of films was used for the adsorption measurement. Material from ∼30 films, made with identical reagent concentrations and conditions, was required to provide

EXPERIMENTAL SECTION

Hyperbranched polyethylenimine (MW ∼ 750 000 Da, denoted LPEI, or ∼2000 Da, denoted SPEI) and polyacrylamide (MW ∼ 10 000 Da, PAAm) as 50 wt % solutions in water, cetyltrimethylammonium bromide (CTAB, 99%), sodium dodecyl sulfate (SDS, 98.5%), and tetramethoxysilane (TMOS, 99.9%) were purchased from SigmaAldrich, and all chemicals were used without further purification. Ultrapure Milli-Q water (18.2 MΩ cm resistance) or D2O (SigmaAldrich) was used as the solvent. To prepare silica films using catanionic surfactant mixtures with polymers, a CTAB solution (74 mM) and a SDS solution (92.5 mM) were initially mixed. A stock aqueous polymer solution (either PEI or PAAm solution) at 100 g/L was diluted with water and added to the catanionic surfactant solution to give a total solution volume of 20 mL with a final CTAB concentration of 37 mM but variable SDS and polymer concentrations. TMOS (84.7 mM for all samples) was added to this solution. This solution was stirred for 30 s and then poured into a 6 cm diameter polystyrene dish over a piece of plastic mesh. To determine the role of the catanionic surfactant mixture and polymer in film formation and structure, we varied the molar ratio of CTAB:SDS (from 2 to 8), final polymer concentration (10−40 g/L), and the type of polymer. No pH adjustment was made to the solutions used, so the solution pH for PEI containing solutions was between 10.5 and 11.2, arising from the normal pH of the PEI in aqueous solution, while PAAm containing solutions were between pH 5 and 5.5. At these pH values the polymers are essentially uncharged, as is required for film formation to be observed.17−19 No significant alteration of the pH due to the addition of the surfactants, prior to addition of the silica precursor, was observed. Films were removed from the solution surface by drawing the plastic mesh up from the subphase after film formation had occurred at the solution surface. Films were dried in air at room temperature. To improve mesostructure retention after removing the template, the dry silica/surfactant/polymer film was exposed to a TMOS atmosphere in an oven at 40 °C for 1 day, and then the film was calcined at 600 °C for 6 h or washed with ethanol to remove the surfactant template while retaining the polymer in the silicate walls. The process of film formation and harvesting is shown pictorially in Figure 1. Synchrotron time-resolved small-angle X-ray scattering (SAXS) experiments were performed using a circulating solution in a capillary flow cell10 on the Non-Crystalline Diffraction Beamline (I22) at the B

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 2. GID patterns and line profiles at Qxy = 0.007 Å−1 from GID patterns collected from interfacial catanionic surfactant/SPEI/silica films in situ at the solution interface, showing typical mesophases. (A, D:a) CTAB:SDS = 8 and 20 g/L SPEI; (B, D:b) CTAB:SDS = 4 and 10 g/L SPEI; (C, D:c) CTAB:SDS = 4 and 30 g/L SPEI.

Figure 3. GID patterns and corresponding line profiles at Qxy = 0.007 Å−1 collected from catanionic surfactant/LPEI/silica films in situ at the air− solution interface, showing typical mesophases. (A, C:a) CTAB:SDS = 4 and 10 g/L LPEI; (B, C:b) CTAB:SDS = 2 and 20 g/L LPEI. sufficient material for measurement. Samples were degassed under vacuum overnight at ∼200 °C for calcined samples and 75 °C for 48 h for washed samples23 before measurement. Measurement was performed with an equilibration period of 300 s at each partial pressure.

before and after calcination as indicated by the ellipse) but remained continuous, without cracking even after calcination at 600 °C for 6 h. This suggests that the polymer/catanionic template improved the film thickness and strength compared to silica films templated using surfactant alone, which are usually fragile and tend to fracture into powders during calcination. Film Structure at the Air/Water Interface. The formation of a catanionic surfactant/polymer/silica film depends on the interaction between the inorganic species and all of the organic species present. Phase transitions can therefore be achieved either by variation of the cationic and anionic surfactant molar ratios or the polymer concentration. The phase diagram was investigated using GID to determine the in situ film structures. Typical patterns obtained from the catanionic surfactant/SPEI templated silica films showed strong diffraction rings spanning the entire Qxy range measured (Figure 2A−C). Since diffraction rings rather than spots are observed, the mesostructures in the films have no preferred orientation within the films. The peaks are also relatively sharp, indicating the presence of large domains of ordered mesostructures within the films, although the amount of amorphous material between such domains in the films cannot be easily quantified. For comparison, line profiles of the data

■

RESULTS Films were formed by initially mixing the catanionic surfactant mixture and polymer solutions of either high or low molecular weight PEI or PAAm before adding the inorganic precursor TMOS. This film forming solution was then poured into a dish to allow film growth on the solution surface (Figure 1, first panel). For CTAB/SDS molar ratios of 2 or below, LPEI concentrations higher than 30 g/L, or SPEI concentrations between 10 and 40 g/L, only precipitation was observed and no film formation occurred. At all other concentrations films formed at the air/solution interface within a few minutes. As shown in Figure 1, these free-standing films were white and smooth, robust enough to be removed from the solution interface on an open mesh, and became transparent when they were dry. The film thickness, measured using a micrometer, ranges from 50 ± 5 to 350 ± 5 μm. Films shrank slightly (in Figure 1 the final two images show the same piece of film C

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 4. Phase diagrams for interfacial catanionic surfactant/PEI/silica films at the air−solution interface (left: LPEI; right SPEI). P = precipitate, L = lamellar phase. Dots correspond to solutions compositions investigated during this set of GID experiments.

along Qxy = 0.007 Å−1 are also plotted (Figure 2D). Structural assignments and unit cell dimensions for the line profiles are given in the Supporting Information (Table SI1 for SPEI and Table SI2 for LPEI films). When films were prepared with a CTAB: SDS molar ratio of 8 and 20 g/L SPEI (Figure 2A,D: a), the GID patterns correspond to a 2D hexagonal structure with unit cell dimension of 52 Å. This is smaller than previously observed for SPEI/CTAB templated silica films at similar polymer and surfactant concentration (60 Å).9 The smaller unit cell is due to the fact that SDS in the CTAB/SDS films will reduce the charge on the catanionic micelles, allowing closer packing of the cylindrical micelles in the catanionic surfactant templated films. The shorter average tail length in the CTAB/SDS mixed micelles will also reduce the micelle diameter, decreasing the dspacing. When films were prepared with 10 g/L SPEI at relatively high CTAB:SDS molar ratio of 4 (Figure 2B,D: b), a bicontinuous cubic structure Pn3m ̅ with a unit cell of 63 Å was the dominant phase formed in the films. However, when the SPEI concentration was increased to more than 30 g/L (Figure 2C,D: c), the GID pattern can be described as a Fm3̅m cubic with a unit cell dimension of 91 Å, combined with a lamellar phase, having a unit cell of 37 Å. An equivalent lamellar mesophase was previously reported to occur at the surface of catanionic surfactant solutions at a molar ratio of CTAB:SDS = 6:4 with no polymer17,18 and is also observed in SAXS data of the mixed surfactant solution before polymer is added (discussed below, see also Supporting Information Figure SI3A). This suggests that as well as regions where polymer, surfactants, and silica all participate in formation of the composite film, other areas contain catanionic surfactant only, perhaps due to incorporation of CTAB/SDS precipitate particles from the subphase into the film. Since the d-spacing is identical to that observed for solutions of only the two surfactants (with no polymer or silica added), the CTAB/SDS particles giving this signal must not contain either polymer or surfactant intercalated into the lamellae as this would be expected to increase the d-spacing. Figure 3 shows GID patterns and corresponding line profiles collected on air/water interfacial films for typical mesophases found in silica films synthesized with CTAB/SDS mixtures and LPEI. Structure assignments and unit cell dimensions for the line profiles are given in Table SI2. For films synthesized with a

catanionic molar ratio of 4 and 10 g/L LPEI, a Pn3̅m phase was obtained (Figure 3A,C: a), with a unit cell size similar to that of films synthesized with low concentrations of SPEI at a similar CTAB:SDS ratio. When the CTAB:SDS molar ratio is reduced to 2 and the LPEI concentration is 20 g/L (Figure 3B,C: b), three distinct peaks were assigned to the (110), (200), and (220) reflections of a Im3m ̅ phase with a unit cell of 52 Å. One reflection (211) of this mesostructure is missing, probably due to differential drying in the vertical direction compared to the in-plane direction, causing lattice distortion.24 Phase diagrams for the CTAB/SDS with SPEI or LPEI films can be sketched as a result of these studies (Figure 4). When PAAm is used in place of PEI, a typical GID pattern of the interfacial film (CTAB:SDS = 2 and 30 g/L PAAm) is shown in Figure 5. A sharp peak at 0.17 Å−1 corresponds to the

Figure 5. GID pattern and corresponding line profile at Qxy = 0.007 Å−1 collected from catanionic surfactant/PAAm/silica films in situ at the air−solution interface ((A, B): CTAB:SDS = 2 and 30 g/L PAAm), showing the typical lamellar mesophases.

lamellar CTAB/SDS particles included in the film, as mentioned above. Apart from that sharp peak, two broad peaks at 0.08 and 0.16 Å−1 were observed for PAAm films at all CTAB:SDS molar ratios, suggesting a lamellar phase with a unit cell of 78 Å. Subphase Solutions Investigated by Time-Resolved SAXS. Aggregation processes in catanionic surfactant/polymer/silica film forming solutions were studied by time-resolved SAXS to probe the formation mechanism of the films (Figure D

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 6. Time-resolved SAXS pattern of the subphase of the silica/polymer film forming solutions. TMOS was added at the third pattern (1 min after data collection began).

unambiguously assign this to a specific cubic phase. All of these cubic phases can be formed by the twisting of lamellar sheets, so could be formed through perturbation of the initial lamellar phase CTAB/SDS particles as SPEI and TMOS infiltrate into the particles. Another possible explanation of the two peaks could be the coexistence of two lamellar phases: one at 0.17 Å−1 from the lamellar phase of the initial CTAB/SDS particles and the one at 0.14 Å−1 arising from a CTAB/SDS lamellar phase with intercalated silica. Variation of the CTAB:SDS molar ratio alters the final mesostructure of the subphase particles as well as that in the films. At a molar ratio of CTAB:SDS = 4 and 30 g/L SPEI (Figure 6B) the SAXS patterns show the typical mixed surfactant lamellar phase (Figure SI3C) prior to TMOS addition at the third pattern. In this case, three distinct peaks at 0.12, 0.14, and 0.16 Å−1 appeared about 13 min after the addition of silicate species, corresponding to the (110), (111), and (200) reflections of a Pn3m ̅ cubic phase with a unit cell of 74 Å. The retardation of mesostructure formation is probably due to higher viscosity of the solution induced by the addition of larger amounts of SDS, giving a higher overall surfactant concentration in the mixture. Considering the final mesostructures found in solution at these two different CTAB:SDS molar ratios, when more SDS is present within the catanionic surfactant micelle it will neutralize the positive CTAB headgroup charge and allow the surfactant headgroups to sit closer together, resulting in a smaller effective headgroup area. This should give a larger surfactant packing parameter, g (where g = v/a0lc25 and a0 is the effective hydrophobic to hydrophilic interfacial area, v is the hydro-

6). Initially, SAXS patterns of the mixed surfactant-only solutions were taken. Precipitation was observed immediately after the two surfactants were mixed. Patterns from these solutions at different CTAB:SDS molar ratios all show one strong diffraction peak at 0.17 Å−1 (Figure SI3A), so the molar ratio has little effect on the mesophase formed in the precipitate formed only by the two surfactants. The d-spacing calculated from the peak position is around 37 Å, similar to the d-spacing observed at the surface of equivalent mixed CTAB/SDS solutions where a lamellar phase was observed. This suggests that the CTAB/SDS mixture forms lamellar particles, which in some cases can be segregated to the interface or incorporated into the films.17,18 In all cases, addition of the polymer did not cause a change in the observed scattering pattern (Figure SI3). After the scattering from the CTAB/SDS solution and then the surfactant/polymer solution had been measured, TMOS was added to the circulating solution. For a solution containing SPEI (30 g/L) with CTAB:SDS at a ratio of 8, the silicate precursor TMOS was added at the third pattern after SPEI was added (Figure 6A). Two sharp peaks at 0.14 and 0.17 Å−1 appear around 2 min after this addition. These peaks can be indexed to a cubic phase as either the (110), (111) reflections of a Pn3̅m primitive cubic mesostructure with a unit cell of 63 Å or the (111), (200) reflections of a Fm3̅m face-centered cubic mesostructure with a unit cell of 78 Å, or the (211), (220) reflection of a Ia3d̅ body-centered cubic mesophase with a unit cell of 109 Å. In other words, CTAB/SDS/SPEI/TMOS particles with cubic mesostructures were formed 60 s after the introduction of the silicate precursor, but with only two peaks visible it is impossible to E

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 7. Time-resolved neutron reflectivity patterns of CTAB/SDS/polymer/silica films formed with (A) CTAB:SDS = 4 and 30 g/L SPEI and (B) CTAB:SDS = 8 and 40 g/L SPEI. (C) CTAB:SDS = 4 and 10 g/L LPEI collected on INTER and (D) CTAB:SDS = 2 and 30 g/L PAAm collected on SURF with 15 min time resolution. Gaps in the time-resolved data are due to beam off periods when the neutron source ceased operation momentarily.

phobic tail volume, and lc is the critical tail length), and thus a mesophase with a smaller micellar curvature. In the case of the solution composed of CTAB:SDS = 8 and 30 g/L SPEI, the face-centered cubic phase of Fm3̅m is therefore more likely to be correct. Otherwise, if the micelle packing in the particle did not change and both solutions had a Pn3̅m cubic structure, neutralizing the CTAB charge by increasing the SDS concentration should reduce the charge of the catanionic surfactant micelles, leading to a closer packing and a smaller dspacing, which is opposite to the observed results. For most concentrations of the CTAB/SDS/LPEI silica film forming solutions, the solution was too viscous to be circulated so it was impossible to conduct a time-resolved experiment. SAXS patterns of the one CTAB/SDS/LPEI film forming solution which was sufficiently fluid to measure (CTAB:SDS = 8 and 30 g/L LPEI TMOS) are shown in Figure 6C. Similar to the SAXS patterns of CTAB/SDS with SPEI, the initial CTAB/ SDS/LPEI solution showed the peak arising from the CTAB:SDS lamellar phase (Figure SI3D). About 12 min after TMOS addition, increasing intensity in the low Q range (below Q = 0.1 Å−1) indicates the formation of a larger-scale structure, probably formation of disordered silica/surfactant/polymer particles that are too big to resolve in the Q range used. A bump appears at 0.14 Å−1 around 10 min after adding TMOS (Figure 6C). This bump grew in intensity with time, suggesting that a less organized mesophase forms in the solution, probably within the large composite particles. The appearance of the peak in the subphase scattering occurs later than those measured for the CTAB/SDS/SPEI solutions. LPEI is a more branched polymer with a much higher molecular weight; thus,

the polyelectrolyte/silica network is more viscous around the micelles, making their rearrangement more difficult and slower. When PAAm is used as the film-forming polymer instead of PEI, the SAXS patterns from the subphase solutions prior to TMOS addition (Figure SI3E) indicate there is no long-range ordering, as no peaks were observed. After TMOS was added, the scattered intensity at the lowest angles increased due to formation of large particles, outside the range of the measured scattering pattern, and a broad bump developed in the scattering patterns with time. The final SAXS patterns of the surfactant/PAAm/silica solution (Figure 6D) contain one broad peak around 0.1 Å−1, suggesting a less ordered mesostructure with a repeat unit of 63 Å. The CTAB:SDS molar ratio and PAAm concentration had no great effect on structure or d-spacing in these solutions. Film Growth Studies. In order to study the film formation process, time-resolved neutron reflectivity data were collected on SURF (15 min resolution) or INTER (1.5 min resolution) as shown in Figure 7. For films prepared with a CTAB:SDS molar ratio of 4 with 30 g/L SPEI (total surfactant concentration 46 mM), two sharp diffraction peaks at 0.09 and 0.18 Å−1 appeared within the first 4 min (Figure 7A, peak positions have an error of ±0.01 Å−1). These two peaks become less intense, and another diffraction peak around 0.12 Å−1 grew up with time. At the beginning, the first two peaks were related to a lamellar structure with a dspacing of 70 Å. However, the third peak does not appear directly related to the initial two lamellar peaks. GID experiments on well-developed films from this solution (Figure 2C,D: c) confirmed a Fm3m ̅ cubic phase with a first-order peak F

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 8. (a) SAXS patterns for a dry free-standing silica film before and after template removal; this silica film was synthesized with a molar ratio of CTAB:SDS = 4 and 30 g/L LPEI (A, as-synthesized film; B, film with template washed out using ethanol; C, silica film after calcination). (b) TGA profiles collected on films as in (a) (dashed line) as-synthesized film and (solid line) film after extraction of the surfactant component of the template in ethanol. (c) Nitrogen adsorption−desorption isotherms collected on films as in (a) where the template was removed by calcination or washing with ethanol.

position at 0.12 Å−1. These time-resolved results may indicate a phase transformation from lamellar multilayers to Fm3̅m phase occurring inside the interfacial film during growth and drying at the interface. For films prepared with a CTAB:SDS molar ratio of 8 with 40 g/L SPEI, initial neutron reflectivity data collected on SURF (not shown) showed one peak at Q = 0.13 Å−1 appearing in the first scan which did not seem to change with time. The final pattern at 207 min, collected for a longer time to obtain better signal-to-noise ratio, showed two small peaks at Q = 0.13 and 0.18 Å−1. Neutron data collected on INTER (Figure 7B) for the same system also showed one sharp peak at Q = 0.13 Å−1 in the first scan. This peak intensity decreased with time while a second sharp peak at Q = 0.16 Å−1 appeared at about 1 h. These two peaks may be indexed as two reflections (110), (111) of a primitive cubic Pn3m ̅ mesophase with d-spacing of 68 Å. However, both diffraction peaks became broad, and the second peak position moved gradually to 0.18 Å−1 in the last measurement (at about 90 min), to give a pattern similar to those initiatelly observed on SURF. In this case, these two peaks can be related as the (110), (200) reflections of a primitive cubic or body-centered cubic Im3̅m phase. The INTER data indicate the mesophase transformation between two different cubic mesophases occurs slowly upon continuous evaporation and film drying with time. High water content in silica−surfactant films is thought to favor a highly curved mesophase.26,27 The presence of water at the micelle−silica interface promotes intercalation of water between the headgroups, pushing them apart by steric hindrance, which results in a higher effective headgroup area, decreasing the packing parameter and increasing micellar curvature. Continuous evaporation processes in dip-coated silica−surfactant films induce a transformation from a micellar cubic phase to a 2D hexagonal phase. Here, losing water around the micellar headgroup in the interfacial film leads to smaller effective headgroup areas and a decrease in micellar curvature. This observation confirms the role of water content on the final interfacial film mesophase. The effect of lower total surfactant concentrations on film growth was also investigated. At the higher surfactant concentrations used, precipitate tended to settle out as solutions were poured into the trough; thus, lower total

concentrations were used at the same molar ratios to prevent this from occurring. Neutron reflectivity patterns of films formed with total surfactant concentrations of 23 and 12 mM (compared to the sample CTAB:SDS = 4 with 30 g/L SPEI, total surfactant concentration 46 mM for the standard preparation) are shown in Supporting Information Figure SI4. When the films were prepared with a total surfactant concentration of 23 mM, solutions were evenly white. No diffraction peaks were seen in the time-resolved reflectivity patterns during the first 40 min, but a peak at ∼0.12 Å−1 grew with time. Reflectivity patterns of the well-developed films reveal three diffraction peaks at 0.12, 0.20, and 0.24 Å−1, which correspond to (100), (110), and (200) reflections, indicating an ordered 2D hexagonal mesostructure with a unit cell dimension of 70 Å. Films still grew when the total surfactant concentration was decreased to 12 mM. These solutions were also opaque. Timeresolved neutron reflectivity patterns showed only two diffraction peaks at 0.09 and 0.18 Å−1 in the first pattern, and these peaks became sharper with time, corresponding to lamellar multilayers with a unit cell of 70 Å. Thus, the film mesostructure is also sensitive to the total concentration of components as well as the molar ratios in solution. Silica films templated by CTAB/SDS/LPEI take a much shorter time to grow on the interface (as observed by the eye) than those prepared with CTAB/SDS/SPEI, especially for films synthesized with high CTAB:SDS molar ratio or LPEI concentration. However, for some cases, with 30 or 40 g/L LPEI, no diffraction peaks could be observed as these films exhibited macroscopic surface roughening, preventing any neutron reflection when the films were grown in the Teflon troughs. Time-resolved neutron reflectivity data with 1.5 min resolution were collected for films prepared with CTAB:SDS molar ratio 4 with 10 g/L LPEI (Figure 7C) on INTER. For this film no obvious peaks were observed at the beginning, but a peak started to appear at Q = 0.13 Å−1 after about 42 min and continued to grow in intensity, up to the last measurement (at 120 min). The film mesostructure therefore appears much later than the film itself, around 42 min, and continues to become more ordered as the film matures. Nevertheless, the structure cannot be determined by only one peak. G

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Thick films were also formed when PAAm was used as the polymer component. Slower time-resolved neutron reflectivity patterns with 15 min resolution were collected for films formed with CTAB:SDS molar ratio 2 and 30 g/L PAAm (Figure 7D) on SURF. For films prepared with PAAm, diffraction peaks were not visible in the first 30 min but had started to appear at Q = 0.08 Å−1 in the 45 min scan. Two diffraction peaks at 0.08 and 0.16 Å−1 were observed at about 1 h and continued to grow in intensity with time. These two peaks indicate a lamellar mesophase with a d-spacing of 78 Å, suggesting that in this case the polymer and silica are intercalated between the CTAB/SDS bilayers unlike the 37 Å lamellar phase observed in the PEI films where the d-spacing was equivalent to that found in precipitates containing only the two surfactants. Dried Free-Standing Polymer/Silica Hybrid Films. These catanionic surfactant templated silica/polymer films are sufficiently robust to be removed from the air/water interface to form free-standing films. SAXS patterns of the dried films showing typical mesophases are given in Figure SI5 (structure assignment and unit cell dimensions are in Table SI6). Most of the as-synthesized film mesostructures were retained after drying, although the unit cell dimension became smaller due to the removal of the water, causing compaction of the unit cell. The organic composite in the PEI-containing films was removed by calcination, leaving open pores but with some loss of the long-range order of the structure, as shown in Figure 8a. The d-spacing shrinks by 7.4 Å during calcination, from 43.3 Å in the as-prepared materials to 35.9 Å in the calcined materials. However, the SAXS data show the d-spacing was maintained when the surfactant component of the template was washed out using ethanol. Films synthesized from more dilute CTAB−SDS solutions were white and smooth and were still robust enough to be removed from the air−solution interface onto an open mesh, although the dry film thickness decreases with a decrease of total solution concentration. SAXS data from the dried films grown at low concentrations (Figure SI7) suggest that films prepared with 46 mM total surfactant concentration have a Fm3̅m mesostructure with a unit cell of 72 Å. When the concentration was decreased to 23 mM, three diffraction peaks at 0.15, 0.27, and 0.30 Å−1 were indexed as the (100), (110), and (200) reflections of an ordered 2D hexagonal structure with a unit cell of 47 Å. Only a broad peak at 0.14 Å−1 was obtained for the dry films prepared with a total surfactant concentration 12 mM, suggesting a less ordered mesophase. Figure 8b shows the TGA profiles collected on an assynthesized CTAB/SDS/LPEI/silica film (CTAB:SDS = 4 and 30 g/L LPEI, SAXS profile shown in Figure 8a) and an equivalent film where the template was extracted using ethanol. The as-synthesized film shows three steps of weight loss: the first step below 150 °C is mainly due to the removal of residual water, the second step starts from 150 °C and is suggested to be the point where decomposition of surfactant begins,28,29 and the last sharp step of weight loss starts at 220 °C due to the decomposition of PEI.23 When the film was first extracted using ethanol, the second step of weight loss disappears, implying that the surfactant has been largely removed from the film. On the other hand, the weight loss starting at 220 °C reveals that about 40 wt % PEI was retained in the mesostructured film after washing. Figure 8c provides the nitrogen adsorption isotherm collected on the same silica films with the template removed either by calcination or washing. When the template was removed by calcination, the nitrogen adsortion isotherm

includes mesopore adsorption at intermediate partial pressures, while at the highest relative pressure adsorption is still rising, suggesting the presence of macropores, possibly between film particles or arising from the incorporation of large CTAB−SDS precipitate particles in the films, as mentioned above, which would leave macropores after calcination. The calculated BET pore volume, surface area, and mesopore diameter were 0.56 cm3/g, 478.8 m2/g, and 5.3 nm, respectively. For films where the surfactant component of the template was removed by washing with ethanol, the nitrogen adsorption isotherm is catagorized as type III, with minimal nitrogen adsorption observed. This can be explained by three possible scenarios: (1) the template has not been completely removed; (2) the mild degassing protocol used (75 °C for 48 h) was insufficient to remove tightly bound water and CO2 already adsorbed to the PEI molecules in the pores and walls of this material; or (3) there is weakly bound multilayer adsorption on the exterior of the surface of the film due to PEI occupying the pores of the film, and consequently fewer adsorption sites are available for nitrogen molecules. If we assume most of the surfactant template was removed by washing, there is about 40 wt % PEI left in the film from the TGA measurements shown in Figure 8b, which will be located in both the pores and the walls of the materials and which will restrict access of nitrogen into the pores. It has been shown by Son et al.23 that meaningful information on the textural preoperties cannot be obtained from nitrogen adsorption−desorption isotherms if PEI loading in mesoporous silica is sufficiently high to block access to the pores so this and the mild degassing may both have contributed to the lack of observed nitrogen adsorption in the ethanolwashed films.

■

DISCUSSION We have successfully employed a catanionic surfactant/polymer film forming system to synthesize silica films at the air/water interface. The introduction of polymer into the reaction system allows electrostatic and hydrophobic interactions between the catanionic surfactants, polymer, and the inorganic precursors to control the packing parameter of the surfactant, and these are key factors in the formation and transitions of the highly ordered mesophases. In these systems even without silica addition, film formation occurs due to a combination of phase separation and evaporation, causing drying and concentration of the species at the solution surface.10 Micelles trapped in the surface layer which formed initially at the solution−air interface evolve to give the final ordered mesostructure within a polymer hydrogel matrix. With addition of silica, the evolution from disorder to order occurs independently in the particles formed rapidly in the subphase solution and in the phase-separated layer at the air−solution interface, which evolves more slowly, meaning the structures can differ between the two phases.9 In the template solution, before the silicate precursor was added, only lamellar phase mixed CTAB/SDS particles were formed, even in the presence of polymer, but composite particles and films with a range of different liquid crystalline ordering formed after silica precursor addition, which is similar to the evolution observed in CTAB/polymer templated films.10 However, the mesophases found in the films and in the subphase were enriched in this case because of the relatively more complicated interactions between the various polymers and the two surfactants. Generally, a mixture of cationic and anionic surfactants with an excess of either surfactant will form mixed catanionic H

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

The interaction between catanionic surfactant mixtures and PEI can lead to a significant change in the micelle interfacial curvature, which is responsible for the observed transitions in the polymer/silica film mesophase with film composition. This interaction with PEI means that, contrary to expectation, as the CTAB:SDS ratio decreases (i.e., the amount of anionic surfactant increases), for a constant SPEI concentration, the micellar structures observed increase in curvature (e.g., at 30 g/ L SPEI the films contain an Fm3̅m structure for a CTAB:SDS ratio of 4, but a 2D hexagonal structure when the CTAB:SDS ratio is 8). Adding SDS causes the catanionic micelles to become more compact and more hydrophobic due to charge neutralization of the CTAB,30 which will improve the hydrophobic interactions between both surfactants and PEI; thus, PEI can pack more closely around the headgroup of the micelle. A bigger headgroup area leads to a smaller packing parameter and higher micelle curvature, resulting in a phase transition from 2D hexagonal, to bicontinuous cubic Pn3̅m or Im3̅m, to micellar cubic Fm3̅m as more SDS is added to the system. Also, an increase of the polymer concentration has the same trend for constant catanionic surfactant molar ratios. As PEI concentration increases, more polyelectrolyte chains will pack around the catanionic micelle head groups, increasing its interfacial area at the micelles surface, which also leads to a higher micelle curvature, resulting in phase transitions from 2D hexagonal to a cubic Im3m ̅ or Pn3m ̅ mesophases to micellar cubic Fm3̅m at different initial catanionic surfactant molar ratios. This trend in phase transitions and the surface curvature of the micellar assembly are thus consistent with previous reports for surfactant templated silicas.32−34 When the polymer/silica hybrid films were prepared with PAAm, the final mesophases of the interfacial films are always lamellar structures, independent of the surfactant molar ratio or polymer concentration. Compared with PEI, PAAm is thought to have little interaction with either CTAB or SDS in solution or toward coadsorption at the interface; thus, the lamellar phase probably occurs due to intercalation of the silica and polymer into the precipitated lamellar catanionic surfactant phase aligned at the interface. The overall solution concentration of the components also has a large effect on the mesostructures both in the subphase solution and in the films. When the total surfactant concentration of film forming solutions was decreased from 46 to 23 mM, films formed with an initial Fm3̅m cubic mesophase transferred to 2D hexagonal mesophase upon drying, a transition which has also been observed by Cagnol previously.35 However, when the film forming solution concentration is too dilute, only lamellar multilayers formed at the interface. Scheme 1 shows the proposed interfacial mesophase transformation with the continuous dilution of the film forming solution. The dominant interaction between the CTAB and PEI is the dipole−cation interaction17 while SDS shows strong affinity to PEI due to both electrostatic and hydrophobic interactions. The polymer is thought to wrap around the micelles, but PEI is more likely to move into the solution and become less packed around the micelle headgroup when the total surfactant and polymer concentrations decrease; thus, as steric hindrance from the PEI is reduced, the electrostatic interaction between the CTAB and SDS will bring them closer. This result in a smaller effective headgroup, bigger g value, and a decrease of the curvature; thus, diluting the film-forming

micelles, which have larger aggregation numbers than those of micelles containing a uniform charged species. An excess of the cationic surfactant (CTAB) results in a larger size, less compact, and more polar mixed catanionic aggregates,30 but as the amount of anionic surfactant is increased, the micelles become more compact and neutralization of the headgroup charges leads to lower headgroup areas, suggesting lower micelle curvatures should be observed. When soluble polymers (LPEI and SPEI) are introduced to the mixed surfactant solution, the interaction between the polymer and catanionic surfactant induces mesophase selfassembly. PEI is almost neutral at the pH range used for this experiment (>3% charged between 9 and 10).19,31 Our earlier work shows that the dominant interaction between PEI and the surfactant CTAB is a dipole−cation interaction, where the polymer amine groups interact with the charged CTAB quaternary ammonium group.19 There is also a strong affinity between anionic surfactant SDS and PEI. The DS− anion can also interact electrostatically with the remaining positively charged amino groups of the polymer. Under basic conditions, however, PEI is essentially a neutral polymer, so the ethylene group in PEI also gives the polymer a hydrophobic character. SDS binds to linear PEI at high pH because the hydrophobic interaction is much larger than electrostatic interaction, and this will also occur in the hyperbranched PEI/SDS complexes used here.31 Thus, the complex interaction between the polymer PEI and catanionic surfactant mixture can be used to cause formation of a range of mesostructures in the films that grow after the addition of silica. From the previous investigations of CTAB/PEI templated silica films, those formed with LPEI take a longer time to develop into a relatively less ordered structure.9 Here again, when time-resolved SAXS was used to investigate the solution below the film, the mesostructure of the particles in the CTAB/ SDS/LPEI mixture takes a much longer time to appear compared to the mesostructured particles in the solution with SPEI. Particles in film-forming solutions with LPEI display a less ordered mesostructure while the particles in the filmforming solution with SPEI mainly display a Pn3̅m structure over a range of polymer concentrations and surfactant molar ratios. Similar behavior happens for the mesostructure of the film formed on the surface of these solutions due to interfacial phase separation. Films synthesized with LPEI take more than 1 h to develop diffraction peaks while the film prepared with SPEI contain an ordered mesostructure within 15 min, and this ordering continues to develop until the final highly ordered structure is formed in less than 2 h. Polyelectrolyte molecular weight and chemical nature as well as the cationic−anionic surfactant molar ratio can be used to control the mesophase geometry of the catanionic surfactant/ polymer templated films. Free-standing thick silica films with lamellar, 2D hexagonal, and several cubic mesostructures were obtained. The mesophase generated in the film can be described by the surfactant packing parameter, g, mentioned above. This packing parameter is used to describe the surface curvature of the micelles and the micelle shape.25,32 A high packing parameter yields a low interfacial curvature. Surface curvature in microemulsions has been observed to increase in the following order for mesophases: lamellar−bicontinuous cubic Ia3̅d−2D hexagonal−3D hexagonal−micellar cubic Im3̅m−micellar cubic Pn3̅m−micellar cubic Pm3̅n−face-centered micellar cubic Fm3̅m.33 A similar order of transitions is observed in these films. I

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

concentration solutions. This material is available free of charge via the Internet at http://pubs.acs.org.

Scheme 1. Total Solution Component Concentration Drives the Film Mesophase Transition at the Air−Solution Interface

■

AUTHOR INFORMATION

Corresponding Author

*Tel +44(0)1225 384192, Fax +44(0)1225 386231, e-mail k. [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge the European Synchrotron Radiation Facility for provision of synchrotron radiation facilities, and we thank the beamline scientists Dr. Oleg Konovalov and Dr. Alexei Vorobiev for their assistance with X-ray reflectivity and grazing incidence diffraction on ID10B (Tröika II, experiment CH3068). ISIS is gratefully acknowledged for provision of neutron scattering facilities and beamtime (experiments RB910127, RB910125, and RB820117). We thank the beamline scientists Dr. Arwel Hughes and Dr. Maximilian Skoda on SURF and Dr. John Webster and Dr. Stephen Holt on SURF and INTER at ISIS, RAL. The beamline scientists Dr. Nick Terrill and Dr. Claire Pizzey on beamline I22 at Diamond Light Source Ltd Didcot, Oxfordshire, are thanked for assistance with the timeresolved SAXS (experiment SM1002). We thank Dr. Sean Rigby for kind discussions on the nitrogen adsoption and desorption data. We also thank the University of Bath and the ORSAS scheme for funding.

solution causes the film structure transition to 2D hexagonal and then a lamellar phase at the interface.

■

CONCLUSIONS Catanionic surfactant and polymer mixtures were mineralized to synthesize robust silica/surfactant/polymer composite films. This method provides an easy way to synthesize mesoporous silica films with a variety of mesostructures grown at the air/ water interface. Interfacial film formation processes were probed by neutron reflectivity, and well-developed film mesostructures were measured by GID. The subphase was also studied by time-resolved SAXS. Polymer molecular weight and chemical nature as well as the cationic−anionic surfactant molar ratio can be used to control the mesophase geometry of the catanionic surfactant/polymer templated films. Lamellar, 2D hexagonal, and several cubic mesostructures were obtained, and a phase diagram in terms of the catanionic molar ratio and polyelectrolyte concentration was drawn for PEI templated films. These films are thick and robust enough to be easily removed from interface, and the film morphology is retained even after calcination. Silica/polymer composite films can also be prepared from very dilute surfactant and polyelectrolytes solutions at the air/water interface. Neutron reflectivity shows a mesophase transition from Fm3̅m to lamellar in the films when the film formation solutions were diluted. This film preparation method provides a simple way to impart polymer functionality into the mesostructured silica wall; we are currently investigating these aspects of the films to develop them into useful membranes for future applications, such as heavy ion removal or CO2 absorption.

■

■

REFERENCES

(1) Sanchez, C.; Boissiere, C.; Grosso, D.; Laberty, C.; Nicole, L. Design, synthesis, and properties of inorganic and hybrid thin films having periodically organized nanoporosity. Chem. Mater. 2008, 20 (3), 682−737. (2) Yang, H.; Coombs, N.; Sokolov, I.; Ozin, G. A. Free-standing and oriented mesoporous silica films grown at the air-water interface. Nature 1996, 381 (6583), 589−592. (3) Aksay, I. A.; Trau, M.; Manne, S.; Honma, I.; Yao, N.; Zhou, L.; Fenter, P.; Eisenberger, P. M.; Gruner, S. M. Biomimetic pathways for assembling inorganic thin films. Science 1996, 273 (5277), 892−898. (4) Edler, K. J.; Yang, B. Formation of mesostructured thin films at the air-liquid interface. Chem. Soc. Rev. 2013, DOI: 10.1039/ C2CS35300H. (5) Fernandez-Martin, C.; Edler, K. J.; Roser, S. J. Evolution of nonionic surfactant templated silicate films at the air-liquid interface. J. Mater. Chem. 2008, 18 (11), 1222−1231. (6) Antonietti, M.; Conrad, J.; Thuenemann, A. Polyelectrolytesurfactant complexes: A new type of solid, mesomorphous material. Macromolecules 1994, 27 (21), 6007−6011. (7) Pantazis, C. C.; Pomonis, P. J. Mesostructure design via poly(acrylic acid)-C(n)TAB complexes: A new route for SBA-1 mesoporous silica. Chem. Mater. 2003, 15 (12), 2299−2300. (8) Pang, J. B.; Na, H.; Lu, Y. F. Effect of ionic polymer on cetyltrimethyl ammonium bromide templated synthesis of mesoporous silica. Microporous Mesoporous Mater. 2005, 86 (1−3), 89−95. (9) Yang, B.; Edler, K. J. Free-standing ordered mesoporous silica films synthesized with surfactant-polyelectrolyte complexes at the air/ water interface. Chem. Mater. 2009, 21 (7), 1221−1231. (10) Yang, B.; Jaber, R.; Edler, K. J. Silica-surfactant-polyelectrolyte film formation: evolution in the subphase. Langmuir 2012, 28 (22), 8337−8347. (11) Kaler, E. W.; Murthy, A. K.; Rodriguez, B. E.; Zasadzinski, J. A. N. Spontaneous vesicle formation in aqueous mixtures of single-tailed surfactants. Science 1989, 245 (4924), 1371−1374.

ASSOCIATED CONTENT

S Supporting Information *

A summary of structure assignments and unit cell parameters from GID patterns for catanionic surfactant/polymer/silica films with typical mesophases for SPEI and LPEI measured in situ at the solution interface; time-resolved SAXS patterns of the surfactant−polymer composite solutions, in situ reflecivity experiments for surfactant/polymer/silica films formed on low concentration solutions, structure assignments, and unit cell parameters from SAXS patterns for dry catanionic surfactant− polyelectrolyte−silica films with typical mesophases as well as SAXS patterns for dried free-standing films from lowJ

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(12) Kaler, E. W.; Herrington, K. L.; Murthy, A. K.; Zasadzinski, J. A. N. Phase-behavior and structures of mixtures of anionic and cationic surfactants. J. Phys. Chem. 1992, 96 (16), 6698−6707. (13) Brasher, L. L.; Herrington, K. L.; Kaler, E. W. Electrostatic effects on the phase-behavior of aqueous cetyltrimethylammonium bromide and sodium octyl sulfate mixtures with added sodium bromide. Langmuir 1995, 11 (11), 4267−4277. (14) Brasher, L. L.; Kaler, E. W. A small-angle neutron scattering (SANS) contrast variation investigation of aggregate composition in catanionic surfactant mixtures. Langmuir 1996, 12 (26), 6270−6276. (15) Lee, J. H.; Gustin, J. P.; Chen, T. H.; Payne, G. F.; Raghavan, S. R. Vesicle-biopolymer gels: Networks of surfactant vesicles connected by associating biopolymers. Langmuir 2005, 21 (1), 26−33. (16) Antunes, F. E.; Marques, E. F.; Gomes, R.; Thuresson, K.; Lindman, B.; Miguel, M. G. Network formation of catanionic vesicles and oppositely charged polyelectrolytes. Effect of polymer charge density and hydrophobic modification. Langmuir 2004, 20 (11), 4647−4656. (17) Edler, K. J.; Wasbrough, M. J.; Holdaway, J. A.; O’Driscoll, B. M. D. Self-assembled films formed at the air/water interface from CTAB/ SDS mixtures with water soluble polymers. Langmuir 2009, 25 (7), 4047−4055. (18) O’Driscoll, B. M. D.; Nickels, E. A.; Edler, K. J. Formation of robust, free-standing nanostructured membranes from catanionic surfactant mixtures and hydrophilic polymers. Chem. Commun. 2007, 10, 1068−1070. (19) O’Driscoll, B. M. D.; Milsom, E.; Fernandez-Martin, C.; White, L.; Roser, S. J.; Edler, K. J. Thin films of polyethylenimine and alkyltrimethylammonium bromides at the air/water interface. Macromolecules 2005, 38 (21), 8785−8794. (20) Penfold, J.; Richardson, R. M.; Zarbakhsh, A.; Webster, J. R. P.; Bucknall, D. G.; Rennie, A. R.; Jones, R. A. L.; Cosgrove, T.; Thomas, R. K.; Higgins, J. S.; Fletcher, P. D. I.; Dickinson, E.; Roser, S. J.; McLure, I. A.; Hillman, A. R.; Richards, R. W.; Staples, E. J.; Burgess, A. N.; Simister, E. A.; White, J. W. Recent advances in the study of chemical surfaces and interfaces by specular neutron reflection. J. Chem. Soc., Faraday Trans. 1997, 93 (22), 3899−3917. (21) Webster, J.; Holt, S.; Dalgliesh, R. INTER the chemical interfaces reflectometer on target station 2 at ISIS. Physica B 2006, 385−386, 1164−1166. (22) Smilgies, D.-M.; Boudet, N.; Struth, B.; Konovalov, O. Troika II: a versatile beamline for the study of liquid and solid interfaces. J. Synchrotron Radiat. 2005, 12 (3), 329−339. (23) Son, W. J.; Choi, J. S.; Ahn, W. S. Adsorptive removal of carbon dioxide using polyethyleneimine-loaded mesoporous silica materials. Microporous Mesoporous Mater. 2008, 113 (1−3), 31−40. (24) Toombes, G. E. S.; Finnefrock, A. C.; Tate, M. W.; Ulrich, R.; Wiesner, U.; Gruner, S. M. A re-evaluation of the morphology of a bicontinuous block copolymer-ceramic material. Macromolecules 2007, 40 (25), 8974−8982. (25) Israelachvili, J. Intermolecular & Surface Forces, 2nd ed.; Academic Press: San Diego, CA, 1991. (26) Grosso, D.; Cagnol, F.; Soler-Illia, G. J. D. A.; Crepaldi, E. L.; Amenitsch, H.; Brunet-Bruneau, A.; Bourgeois, A.; Sanchez, C. Fundamentals of mesostructuring through evaporation-induced selfassembly. Adv. Funct. Mater. 2004, 14 (4), 309−322. (27) Besson, S.; Gacoin, T.; Ricolleau, C.; Jacquiod, C.; Boilot, J.-P. Phase diagram for mesoporous CTAB−silica films prepared under dynamic conditions. J. Mater. Chem. 2003, 13 (2), 404−409. (28) Reichinger, M.; Schmidt, W.; van den Berg, M. W. E.; Aerts, A.; Martens, J. A.; Kirschhock, C. E. A.; Gies, H.; Grunert, W. Alkene epoxidation with mesoporous materials assembled from TS-1 seeds - is there a hierarchical pore system? J. Catal. 2010, 269 (2), 367−375. (29) Sicard, L.; Llewellyn, P. L.; Patarin, J.; Kolenda, F. Investigation of the mechanism of the surfactant removal from a mesoporous alumina prepared in the presence of sodium dodecylsulfate. Microporous Mesoporous Mater. 2001, 44, 195−201.

(30) Chakraborty, H.; Sarkar, M. Optical spectroscopic and TEM studies of catanionic micelles of CTAB/SDS and their interaction with a NSAID. Langmuir 2004, 20 (9), 3551−3558. (31) Winnik, M. A.; Bystryak, S. M.; Chassenieux, C.; Strashko, V.; Macdonald, P. M.; Siddiqui, J. Study of interaction of poly(ethylene imine) with sodium dodecyl sulfate in aqueous solution by light scattering, conductometry, NMR, and microcalorimetry. Langmuir 2000, 16 (10), 4495−4510. (32) Zhao, D. Y.; Huo, Q. S.; Feng, J. L.; Chmelka, B. F.; Stucky, G. D. Nonionic triblock and star diblock copolymer and oligomeric surfactant syntheses of highly ordered, hydrothermally stable, mesoporous silica structures. J. Am. Chem. Soc. 1998, 120 (24), 6024−6036. (33) El-Safty, S. A.; Hanaoka, T. Microemulsion liquid crystal templates for highly ordered three-dimensional mesoporous silica monoliths with controllable mesopore structures. Chem. Mater. 2004, 16 (3), 384−400. (34) El-Safty, S. A.; Hanaoka, T. Fabrication of crystalline, highly ordered three-dimensional silica monoliths (HOM-n) with large, morphological mesopore structures. Adv. Mater. 2003, 15 (22), 1893− 1899. (35) Cagnol, F.; Grosso, D.; Soler-Illia, G. J. D. A. S.; Crepaldi, E. L.; Babonneau, F.; Amenitsch, H.; Sanchez, C. Humidity-controlled mesostructuration in CTAB-templated silica thin film processing. The existence of a modulable steady state. J. Mater. Chem. 2003, 13 (1), 61−66.

K

dx.doi.org/10.1021/la4001329 | Langmuir XXXX, XXX, XXX−XXX