Order and Orientation Control of Mesoporous Silica Films on

Order and Orientation Control of Mesoporous Silica Films on Conducting Gold Substrates Formed by Dip-Coating and Self-Assembly: A Grazing Angle of ...
0 downloads 0 Views 331KB Size
10112

Langmuir 2005, 21, 10112-10118

Order and Orientation Control of Mesoporous Silica Films on Conducting Gold Substrates Formed by Dip-Coating and Self-Assembly: A Grazing Angle of Incidence Small-Angle X-ray Scattering and Field Emission Scanning Electron Microscopy Study† Michael P. Tate, Brian W. Eggiman, Jonathan D. Kowalski, and Hugh W. Hillhouse* School of Chemical Engineering, Purdue University, West Lafayette, Indiana 47907 Received March 4, 2005. In Final Form: April 28, 2005 Grazing-angle of incidence small-angle X-ray scattering (GISAXS) and high-resolution field emission scanning electron microscopy have been used to characterize the mesophase symmetry, orientation, and long-range order in PEO20-PPO70-PEO20 (Pluronic P123) templated mesoporous silica thin films on conducting gold substrates as a function of silica-to-ethylene oxide (Si/EO) block ratio and relative humidity (RH). The films are formed by dip-coating followed by evaporation-induced self-assembly under tightly controlled RH. The general evolution of the mesophase follows the trends that are expected based on shape factors due to swelling of the PEO block. However, changes in orientation of the nanostructure relative to the substrate and the degree of long-range order are found to depend on Si/EO ratio. These effects are likely due to the dynamics of evaporation and self-assembly. Generally, at Si/EO ratios lower than 3.29, the films contained regions where the nanostructure was not oriented relative to the plane of the substrate. However, for Si/EO ratios greater than 3.62, conditions were found where the nanostructure of the film was highly oriented relative to the plane of the substrate. This is true over the range of RH studied, independent of the nanostructure symmetry. For low Si/EO ratios at the highest RH levels, the films were composed of a mixture of spherical and cylindrical pores. At high Si/EO ratios and high RH levels, the films had a highly oriented R-3m nanostructure but displayed streaking perpendicular to the substrate in the Bragg spots on GISAXS patterns. This streaking is interpreted as faulting along planes parallel to the substrate.

Introduction 1-4

mesoporous thin films of inSurfactant-templated sulating silica5-8 have many promising potential applications ranging from low-k dielectrics9 to templates for the electrodeposition of ultra-small-diameter nanowires for thermoelectric devices.10 Further, the synthesis of these self-assembled films has been extended to semiconducting metal oxides such as titania11,12 and tin oxide.13,14 These thin film compositions have potential applications in * Corresponding author. E-mail: [email protected]. † Part of the Bob Rowell Festschrift special issue. (1) Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc. 1992, 114, 10834-10843. (2) Monnier, A.; Schuth, F.; Huo, Q.; Kumar, D.; Margolese, D.; Maxwell, R. S.; Stucky, G. D.; Krishnamurty, M.; Petroff, P.; Firouzi, A.; Janicke, M.; Chmelka, B. F. Science 1993, 261, 1299-1303. (3) Huo, Q.; Margolese, D.; Cielsa, U.; Feng, P.; Gler, T. E.; Sieger, P.; Leon, R.; Petroff, P. M.; Schuth, F.; Stucky, G. D. Nature 1994, 368, 317-321. (4) Firouzi, A.; Kumar, D.; Bull, L. M.; Besier, T.; Sieger, P.; Huo, Q.; Walker, S. A.; Zasadzinski, J. A.; Glinka, C.; Nicol, J.; Margolese, D.; Stucky, G. D.; Chmelka, B. F. Science 1995, 267, 1138-1143. (5) Yang, H.; Kuperman, A.; Coombs, N.; MamicheAfara, S.; Ozin, G. A. Nature 1996, 379, 703-705. (6) Hillhouse, H. W.; Okubo, T.; vanEgmond, J. W.; Tsapatsis, M. Chem. Mater. 1997, 9, 1505-1507. (7) Trau, M.; Yao, N.; Kim, E.; Xia, Y.; Whitesides, G. M.; Aksay, I. A. Nature 1997, 390, 674-676. (8) Lu, Y. F.; Ganguli, R.; Drewien, C. A.; Anderson, M. T.; Brinker, C. J.; Gong, W. L.; Guo, Y. X.; Soyez, H.; Dunn, B.; Huang, M. H.; Zink, J. I. Nature 1997, 389, 364-368. (9) Baskaran, S.; Liu, J.; Domansky, K.; Kohler, N.; Li, X. H.; Coyle, C.; Fryxell, G. E.; Thevuthasan, S.; Williford, R. E. Adv. Mater. 2000, 12, 291-294. (10) Hillhouse, H. W.; Tuominen, M. T. Microporous Mesoporous Mater. 2001, 47, 39-50.

photovoltaics and sensors. However, the engineering needed to develop such films for the desired applications is still in its infancy. One of the most promising routes to mesoporous thin film formation is via evaporation-induced self-assembly (EISA).8,15-20 In this method a dilute liquid thin film is spread on a substrate by dip-coating (or spincoating, spray-coating, etc). Upon evaporation of the solvent, the precursors become concentrated and selfassemble into a nanostructured thin film. With proper control of coating solution composition, coating solution history, and details of the environmental conditions during dipping and evaporation, thin films of a targeted phase may be reproducibly synthesized. Careful postsynthesis treatment and calcination then result in a mesoporous thin film. The synthesis of these films can be quite sensitive to various parameters such as the time elapsed since (11) Alberius, P. C. A.; Frindell, K. L.; Hayward, R. C.; Kramer, E. J.; Stucky, G. D.; Chmelka, B. F. Chem. Mater. 2002, 14, 3284-3294. (12) Crepaldi, E. L.; Soler-Illia, G.; Grosso, D.; Cagnol, F.; Ribot, F.; Sanchez, C. J. Am. Chem. Soc. 2003, 125, 9770-9786. (13) Yuliarto, B.; Zhou, H. S.; Yamada, T.; Honma, I.; Asai, K. Chem. Lett. 2003, 32, 510-511. (14) Urade, V. N.; Hillhouse, H. W. J. Phys. Chem. B 2005, 109 (21), 10538-10541. (15) Zhao, D.; Yang, P.; Melosh, N.; Feng, J.; Chmelka, B. F.; Stucky, G. D. Adv. Mater. 1998, 10, 1380-+. (16) Zhao, D. Y.; Yang, P. D.; Margolese, D. I.; Chmelka, B. F.; Stucky, G. D. Chem. Commun. 1998, 2499-2500. (17) Brinker, C. J.; Lu, Y. F.; Sellinger, A.; Fan, H. Y. Adv. Mater. 1999, 11, 579-+. (18) Grosso, D.; Balkenende, A. R.; Albouy, P. A.; Ayral, A.; Amenitsch, H.; Babonneau, F. Chem. Mater. 2001, 13, 1848-1856. (19) Gibaud, A.; Grosso, D.; Smarsly, B.; Baptiste, A.; Bardeau, J. F.; Babonneau, F.; Doshi, D. A.; Chen, Z.; Brinker, C. J.; Sanchez, C. J. Phys. Chem. B 2003, 107, 6114-6118. (20) Grosso, D.; Cagnol, F.; Soler-Illia, G.; Crepaldi, E. L.; Amenitsch, H.; Brunet-Bruneau, A.; Bourgeois, A.; Sanchez, C. Adv. Funct. Mater. 2004, 14, 309-322.

10.1021/la050595h CCC: $30.25 © 2005 American Chemical Society Published on Web 06/30/2005

Mesoporous Silica Films

mixing the coating solution and the environmental conditions such as relative humidity (RH).19,21 Also, effects that may be attributable to the substrate surface composition and cleanliness may also be observed. One hallmark in the development of these films was the report by Stucky and co-workers on the use of nonionic triblock copolymers of poly(ethylene oxide)poly(propylene oxide)-poly(ethylene oxide) (PEO-PPOPEO) to template the nanostructure.11,15 This was an important development due to the fact that these templates are inexpensive, environmentally benign, and commercially available with many choices of hydrophilic and hydrophobic block lengths. The various block lengths along with the initial composition of the reacting species allows for custom tailoring of the pore sizes and nanostructure. With such templates two-dimensional (2D) hexagonal, lamellar, and cubic films have been reported. However, upon solvent evaporation, the films experience a compression perpendicular to the substrate. The manner in which this compression occurs and how it changes the symmetry are quite important since in some cases it has been shown to contribute to changes in the pore topology.12 As a result of the compression upon drying, the 2D hexagonal phase (which has plane group p6mm symmetry in bulk powder) distorts and is explained by a centered 2D lattice described by the plane group c2mm. Further, there are several possible 3D cubic phases that may experience symmetry breaking upon drying. If the cubic phase has Im-3m symmetry, the resulting thin films have been shown to have 3D face-centered orthorhombic symmetry described by space group Fmmm.22 Most recently though, a new structure of mesoporous silica film was reported by Hillhouse and co-workers23 that has 3D rhombohedral symmetry described by space group R-3m. The films are oriented such that the (111) plane is parallel to the substrate and are conjectured to result from a (111) oriented face-centered cubic phase. Further, the authors showed that under certain conditions extremely wellordered phases may be synthesized that yield grazing angle of incidence small-angle X-ray scattering (GISAXS) patterns populated with dozens of diffraction spots. A key issue that must be addressed in order to develop these self-assembled films for most applications is control of order and orientation of the nanopores on substrates of technological interest (gold, platinum, silicon, etc.). For many of the devices we desire, it is necessary to create films where the mesopores are accessible to the vapor or liquid phase above the film. This is an absolute necessity to develop these types of films for thermoelectric, photovoltaic, or data storage devices. Several other technologies have been developed that yield nanoporous thin films with accessible pores. Techniques such as nuclear track etching,24,25 anodic electrochemical etching,26-28 and selfassembly of a pure diblock copolymers (followed by selective removal of one block)29 all yield accessible nanopores. However, thin films templated by self(21) Cagnol, F.; Grosso, D.; Soler-Illia, G.; Crepaldi, E. L.; Babonneau, F.; Amenitsch, H.; Sanchez, C. J. Mater. Chem. 2003, 13, 61-66. (22) Falcaro, P.; Grosso, D.; Amenitsch, H.; Innocenzi, P. J. Phys. Chem. B 2004, 108, 10942-10948. (23) Eggiman, B. W.; Tate, M. P.; Hillhouse, H. W. Submitted for publication. (24) Price, P. B.; Walker, R. M. J. Appl. Phys. 1962, 33, 3407-&. (25) Guillot, G.; Rondelez, F. J. Appl. Phys. 1981, 52, 7155-7164. (26) Furneaux, R. C.; Rigby, W. R.; Davidson, A. P. Nature 1989, 337, 147-149. (27) Masuda, H.; Yamada, H.; Satoh, M.; Asoh, H.; Nakao, M.; Tamamura, T. Appl. Phys. Lett. 1997, 71, 2770-2772. (28) Choi, J.; Sauer, G.; Nielsch, K.; Wehrspohn, R. B.; Gosele, U. Chem. Mater. 2003, 15, 776-779.

Langmuir, Vol. 21, No. 22, 2005 10113

assembled amphiphiles are unique among the various technologies due to the size range of the pores and the simplicity. Most other techniques run into limitations in achieving pore sizes below 10-20 nm. For films assembled by amphiphiles, the most robust range of pore sizes is from 2 to 10 nm. There have recently been a few reports in the literature addressing the accessibility of amphiphile-templated thin films. Lu and Brinker electrodeposited palladium within the pores of PEO20-PPO70-PEO20 (Pluronic P123) templated hexagonal phase mesoporous silica thin films.30 However, in this case (as is typical with the hexagonal phase) the mesopores in the film run parallel to the substrate. Thus, direct access to the mesopores via mesopores is not available. Access to the mesopores is only obtained through micropores running perpendicular to the substrate. Similarly, material deposited in the mesopores is not directly addressable from above the film. In most materials templated by PEO-containing triblock copolymers, the microporosity can be controlled to some extent. Davidson and co-workers showed that a microporous corona exists around the mesopores templated by P123 and may be tailored by hydrothermal treatments.31 While controlling this phenomenon is critical for developing applications, ultimately, direct access to the mesopores is desired. Zhao and co-workers investigated the effects of adding mesitylene to swell the hydrophobic cores of Pluronic F127 (PEO106-PPO70-PEO106) templated mesoporous silica powders. On the basis of nitrogen adsorption experiments, they demonstrated not only that mesitylene does swell the core but also that this results in opening connections between the cages resulting in accessibility.32 However, these experiments were performed on powders and have yet to be shown in mesoporous silica thin films. Toward the goal of accessibility in films, Okubo and coworkers demonstrated direct access to mesopores at the surface of F127 templated films by optimizing the preparation of the coating solution and following the synthesis with a thermal treatment.33,34 While the mesopores had direct access to the vapor space above, an ordered hexagonal phase was formed below the top surface of the film and direct access to the substrate did not exist. However, to develop nanowire arrays, it is precisely this direct access to the mesopores from both the vapor and the substrate that is desired. To obtain 3D access to the mesopore structure, researchers have focused on PEO-PPO-PEO templates that favor cubic phases in the binary triblock/solvent system. These triblock copolymers typically have large PEO segments (such as F127). The larger the PEO block, the more readily cubic phases are formed due to the increased size of the “headgroup”. However, at the same time the distance between the hydrophobic cores is increased, decreasing the likelihood of interconnections. In this report we focus on using a triblock copolymer template that has a smaller PEO block length (P123) to (29) Thurn-Albrecht, T.; Steiner, R.; DeRouchey, J.; Stafford, C. M.; Huang, E.; Bal, M.; Tuominen, M.; Hawker, C. J.; Russell, T. Adv. Mater. 2000, 12, 787-791. (30) Wang, D. H.; Zhou, W. L.; McCaughy, B. F.; Hampsey, J. E.; Ji, X. L.; Jiang, Y. B.; Xu, H. F.; Tang, J. K.; Schmehl, R. H.; O’Connor, C.; Brinker, C. J.; Lu, Y. F. Adv. Mater. 2003, 15, 130-+. (31) Imperor-Clerc, M.; Davidson, P.; Davidson, A. J. Am. Chem. Soc. 2000, 122, 11925-11933. (32) Fan, J.; Yu, C. Z.; Gao, T.; Lei, J.; Tian, B. Z.; Wang, L. M.; Luo, Q.; Tu, B.; Zhou, W. Z.; Zhao, D. Y. Angew. Chem. Int. Ed. 2003, 42, 3146-3150. (33) Naik, S. P.; Yamakita, S.; Ogura, M.; Okubo, T. Microporous Mesoporous Mater. 2004, 75, 51-59. (34) Naik, S. P.; Yamakita, S.; Sasaki, Y.; Ogura, M.; Okubo, T. Chem. Lett. 2004, 33, 1078-1079.

10114

Langmuir, Vol. 21, No. 22, 2005

Tate et al.

identify regions of phase space where direct mesopore access from the surface to the substrate may exist. While the cubic phase for P123 exists in a narrower region of phase space, it is hypothesized that mesopore accessibility may be easier to obtain in these films due to the decreased distance between hydrophobic cores. The accessibility of the mesopore structure to the region above the film is probed directly by imaging with a high-resolution field emission microscope (FESEM) while the long-range order (LRO), orientation, and phase topology are illuminated by GISAXS. We identify each region of phase space (composition and humidity) by its symmetry, degree of LRO, and its orientation with respect to the substrate. In addition, these studies are carried out on gold substrates such that optimized conditions can be found for electrochemical deposition of nanowires. It is shown in particular that the combination GISAXS and FESEM provides a powerful characterization tool to engineer these mesoporous materials. 2. Experimental Methods 2.1. Synthesis of Mesoporous Silica Thin Films. P123 triblock copolymer template was obtained from BASF and used as received. The other reagents used during synthesis were all ACS reagent grade and include tetraethyl orthosilicate (TEOS) from Aldrich, hydrochloric acid (HCl) from Aldrich, and ethanol from EM Science. Gold-coated substrates were prepared by thermal evaporation of a 15 nm adhesion layer of Ti followed by thermal evaporation of 150 nm of gold onto borosilicate glass “cover” slides (60 mm × 22 mm × 0.18 mm) obtained from VWR Scientific. Just prior to dip coating, the gold-coated substrates were cleaned in a 1 wt % Alconox solution at 65 °C. The substrates were then thoroughly rinsed with copious quantities of deionized ultrafiltrated water and allowed to air-dry. A range of silica-to-ethylene oxide (Si/EO) block molar ratios was chosen for investigation (5.21 > Si/EO > 2.43). For each Si/EO ratio, a surfactant precursor solution was prepared by first dissolving 2.4, 2.99, 3.15, 3.31, 3.48, 4.15, 4.5, 4.84, 5.19, or 5.91 g of P123 in 12.84 g of ethanol. The solution was sealed and set to stir at room temperature for at least 24 h to equilibrate prior to use. A silica precursor solution was prepared by first mixing 19.19 g of ethanol and 9.48 g of dilute HCl/water with 18.22 g of TEOS. Immediately after addition of TEOS, the solution was sealed and stirred for exactly 20 min at room temperature (∼21 °C). At the end of 20 min the appropriate P123 precursor solution was added immediately. The solution was resealed and stirred in a cold water bath at 5 °C. After 10 min of stirring, the solution was quickly transferred to an open mouth beaker inside the humidity-controlled dip-coating chamber. The solution level was within 1 mm of the top of the beaker to ensure the RH above the liquid was as close as possible to the overall RH of the chamber. The RH inside the dip-coating chamber was controlled within 2% of the RH set point by forced flow of a controlled mixture of dry and water-saturated air streams. Set points investigated were from 10% to 85% RH. The dip-coating speed was 60 mm/ min for all dips. Two minutes after dipping, each film was transferred to a controlled RH chamber (at the same RH as the dipping chamber) and aged for at least 12 h. The films were then calcinated at 400 °C for 4 h with a heating and cooling ramp of 1 °C/min. 2.2. GISAXS and FESEM Characterization. GISAXS patterns were collected from the as-synthesized and calcined samples using a Molecular Metrology X-ray scattering instrument that incorporates a three-pinhole camera, microfocus X-ray source, an Osmic MaxFlux confocal X-ray optic, and a gas-filled 2D multiwire detector at a camera length of 1424 mm. The scattering patterns were calibrated in q-space using an isotropic silver behenate powder standard. For illustrative purposes the log intensity (arbitrary scale) is presented in each GISAXS pattern. The films were mounted on a two-axis goniometer that allowed for 180° rotation of the sample about a vertical axis, perpendicular to the beam (see Figure 1a). Due to the presence of an intense specular beam at grazing angles of incidence, aluminum strips were used to attenuate the scattering along the

Figure 1. (a) Experimental setup for GISAXS data collection. (b) Plan view of the experimental setup showing the different diffraction and refraction events that result in spots in the GISAXS patterns. The effects of refraction are exaggerated in the figure for illustrative purposes. specular plane, blocking the region of -0.05 < qz < 0.2 Å-1 and -0.015 < qx < 0.015 Å-1. Lattice constants were fitted to the 2D GISAXS data using NANOCELL,35 a program that simulates full 2D SAXS spot patterns for any given angle of incidence from nanomaterials that have a single domain of orientation, are oriented with respect to specific axes (but sample all orientations along other axes), or are completely free to sample all orientations. The predicted spot patterns from NANOCELL are overlaid on the GISAXS data. Due to the high intensity of the source and low thickness of the substrates, the main beam along with diffracted beams can be transmitted through the substrate. Both reflection and refraction events can alter the path of these beams and introduce shifts on where they strike the detector. All spots on the left side of the GISAXS pattern have experienced a shift, as well as spots below the substrate “horizon” on the right side of the pattern; see Figure 1b. High-resolution FESEM images were collected (top view) from calcined films using a Hitachi S-4800. The accelerating voltage was 3 kV with a working distance of 8 mm. The edge of each sample received a layer of carbon paint (applied with a small brush) to reduce charging effects.

3. Results and Discussion 3.1. Synthesis and Characterization of Ordered Films. To investigate the sensitivity of the structure, order, and orientation of P123 templated films to changes in RH and composition, ranges of RH and polymer concentrations were chosen that span the regions reported for the highly ordered rhombohedral R-3m films reported by Hillhouse and co-workers23 and the 2D hexagonal phases by Stucky and co-workers.15 To predict the resulting (35) Tate, M. P.; Hamilton, B. D.; Eggiman, B. W.; Wei, T. C.; Kowalski, J. D.; Hillhouse, H. W. Submitted for publication.

Mesoporous Silica Films

Langmuir, Vol. 21, No. 22, 2005 10115

Figure 2. (a) GISAXS pattern of an as-synthesized R-3m film with the (111) oriented perpendicular to the gold substrate. (b) GISAXS pattern of a calcined R-3m film with LRO. The overlay of simulated spots is from a NANOCELL simulation with lattice constants a ) 112 Å and R ) 86°. (c) FESEM micrograph with FFT inset of LRO R-3m structure with showing the 6-fold symmetry of the [111] direction.

phase, Alberius used a parameter (Φ) defined as the volume of the polymer over the volume of the polymer plus the volume of the inorganic.11 When the synthesis is carried out near the isoelectric point of the metal oxide, this parameter may be used to estimate the phase of mesoporous film based on the phase of the binary aqueous phase diagram of the amphiphile. This very useful synthesis parameter, which is surfactant specific, captures composition effects from the coating sol but does not capture effects due to processing conditions that can dramatically alter the mesostructure. In particular, incorporating the effect of humidity is problematic. Grosso and co-workers demonstrated a “modulable steady state” for cetyltrimethylammonium bromide/silica systems in which the mesostructure, shortly after dipping, could be reversibly altered by changing the RH.21 Later, Gibaud and co-workers showed that some time after dipping the lattice vector components parallel to the substrate become pinned while the lattice vector components perpendicular to the substrate remain changeable.36 It was also discovered that under certain circumstances, the rate of evaporation of water from the film greatly affects the order of the films.19 This is expected to become an issue whenever the drying time becomes comparable to the time scale needed for condensation reactions to form a percolating interconnected network (particularly important for thicker films). Thus, the synthesis parameters that are used to predict phase topology can be overwhelmed by processing conditions (RH and evaporation rates) that force the structure into a different phase and control the LRO and orientation. To illuminate these competing effects both Si/EO and RH were varied. Characterizing the mesophase topology, orientation, and order of mesoporous thin films presents a major challenge using standard characterization techniques. Theta-theta geometry 1D X-ray diffraction (XRD) (widely available and typically used for powders) provides only a very limited view of reciprocal space for thin films since the powder approximation is not valid. This makes the proper identification of the mesostructure symmetry and film orientation very difficult. Also, due to the small angles involved, pole-figure texture analysis becomes difficult. Further, on gold substrates the critical angle for total reflection is much larger than that on silica (0.35° compared to 0.17°) and results in a much more intense background in a theta-theta diffractometer scan that overwhelms the Bragg diffraction at low angles. This makes identification of phases on gold much more difficult using standard powder XRD techniques. This difficultly in identification of mesostructured thin films on gold substrates may have hampered earlier research attempts

because films with mesostructure may have been falsely identified as having no mesostructure. The other workhorse characterization technique, transmission electron microscopy (TEM) of microtomed film sections (or fragments scraped from the substrate) is limited to use on well-ordered samples since the image contrast roughly represents the changes in average electron density along the beam path through the sample. In films where the mesopores are not oriented over the thickness of the sample, the contrast is lost and the mesostructure cannot be imaged (note: this can be improved by preparing thinner samples by ion milling or other techniques). To overcome these limitations, we have used the combination of GISAXS and FESEM to characterize the mesophase topology, orientation, and order. In all the films synthesized it was found that many domains exist that have different rotational orientations with respect to the substrate normal. As a result, reciprocal space consists of rings instead of points.12,37 Further, we note that since the distances between the real space lattice points of the nanostructure are large, the reciprocal space lattice is small, and the Ewald sphere may be considered to be flat. As a result, the pattern on the detector may be considered to first approximation as a planar slice through reciprocal space. At high Si/EO ratios and/or high RH an R-3m phase with LRO is favored while at moderate Si/EO ratios and lower RH, a c2mm phase with medium-range order (MRO) that is highly oriented relative to the substrate is favored. GISAXS and FESEM data for a typical synthesis yielding the oriented LRO R-3m structure are shown in Figure 2. While the films have no orientational order relative to the substrate normal, they are highly oriented with respect to rotation about any axis in the plane of the substrate. NANOCELL simulations of the Bragg spot patterns show that the R-3m structure is oriented with the (111) plane parallel to the substrate with a ) 112 Å and R ) 86°. As a result, the first peak in powder XRD scans of these films is correctly indexed as the (111). These results on gold substrates are qualitatively similar to those found on silica substrates. The nanostructure symmetry, degree of LRO, and orientation follow the same trends; however slight differences in the location of the transitions are present. Those results are not presented here. In a study on silica films templated with F127,34 some syntheses yielded film morphologies that displayed a (36) Gibaud, A.; Dourdain, S.; Gang, O.; Ocko, B. M. Phys. Rev. B 2004, 70. (37) Hayward, R. C.; Alberius, P. C. A.; Kramer, E. J.; Chmelka, B. F. Langmuir 2004, 20, 5998-6004.

10116

Langmuir, Vol. 21, No. 22, 2005

Tate et al.

Figure 3. (a) GISAXS pattern of a calcined MRO c2mm film with strong orientational order about any axis parallel to the plane of the gold substrate. The overlay of simulated spots is from a NANOCELL simulation with lattice constants a ) 122 Å and b ) 140 Å. (b) FESEM micrograph of MRO c2mm structure with the (10) oriented perpendicular to the substrate.

Figure 4. Schematic showing the relative orientation of the (a)c2mm structure and (b) R-3m structure with respect to the substrate. For the films shown in Figures 2 and 3, in both cases the films are highly oriented about the x-axis but sample many orientations about the z-axis.

mixture of spherical and cylindrical objects in top-view FESEM images while cross-sectional TEM images showed that layers closer to the substrate were actually an oriented hexagonal or c2mm phase. Here, this possibility is ruled out by the GISAXS data. Many c2mm cells were chosen to try to explain all or parts of the GISAXS pattern in Figure 2b. In all cases the c2mm cell could only be made to fit one (independent) peak. As a result, the film shown in Figure 2c is a pure phase material. Analogously, GISAXS and FESEM data for a typical synthesis yielding the oriented c2mm structure are shown in Figure 3 along with the NANOCELL simulation overlay. The MRO c2mm structure contains rotational freedom about an axis perpendicular to the substrate but is highly oriented relative to an axis in the plane of the substrate. It is oriented with the (10) plane parallel to the substrate with a ) 122 and b ) 140 Å. A schematic showing the order and orientation of the films is shown in Figure 4. The results of all the syntheses on gold substrates are shown in Table 1.

3.2. Effect of Relative Humidity. The trends in mesophase structure may be rationalized in the context of shape factors. In this model, the effective area of the headgroup is compared with the volume of the surfactant to control the curvature of micellar structures.38 It takes the form v/a0lc where v is the volume of the hydrophobic portion, a0 is the effective area of the headgroup, and lc is the critical chain length. This model predicts spherical aggregates for shape factors less than 1/3, cylindrical aggregates from 1/3 to 1/2, and lamellar or bicontinuous structures as it increases to 1. For the polymeric systems discussed here, the effective area of the headgroup is difficult to estimate since water and silica associate with (and swell) the PEO block. However, the model may be used to understand trends. At low RH the effective area of the headgroup is smaller, tending to favor structures composed of cylindrical aggregates. This leads to the c2mm structure (distorted 2D hexagonal) or disordered phases with only short-range order (SRO). At higher RH, more water is retained in the films while the silica network condenses. This swells the PEO blocks and increases the curvature of the hydrophilic/hydrophobic interface, tending to favor structures composed of spherical micelles. This packing of globular or spherical micelles results in formation of the R-3m structure after compression. The trend of increasing humidity is shown in Figure 5. The well-defined and oriented structures follow the expected evolution based on the shape factor. However, for regions intermediate to these and at the highest RH, the films show some structural disorder. At the highest RH, the order in the films change resulting in streaking of the Bragg peaks normal to the substrate in the GISAXS pattern, Figure 5g. This streaking can be understood in a couple of ways. One possibility is that there is faulting in planes that are parallel to the substrate. This will stretch out reciprocal space points perpendicular to the faulting plane and cause streaks to appear in the GISAXS pattern.39 This faulting can potentially be explained by slowed evaporation. At very high humidity, the rate of evaporation of water from the film is greatly reduced. This causes a disruption of the smooth and even drying line across the surface of the film and may result in the generation of lateral stresses in the plane of the substrate that cause the faulting. A previous report has shown that streaking in the GISAXS pattern of thermally treated mesoporous TiO2 films was the result of pore merging and resulted in opening accessibility to the substrate.12 Such an explanation may also be a (38) Israelachvili, J. Intermolecular and Surface Forces, 2nd ed.; Academic Press: London, 1992. (39) Giacovazzo, C., Ed. Fundamentals of Crystallography; Oxford University Press: New York, 1992.

Mesoporous Silica Films

Langmuir, Vol. 21, No. 22, 2005 10117

Table 1. The Nanophase Topology and Degree of Order for P123 Templated Mesoporous Silica Thin Films on Gold Substrates with Changes in the Si/EO Ratio and RHa Si/EO ratio

9 ( 2% RH

30 ( 2% RH

5.21

MRO c2mm oriented

MRO c2mm oriented

4.24 3.97 3.79 3.62

MRO c2mm oriented

MRO c2mm not oriented

40 ( 2% RH

LRO R-3m oriented LRO R-3m oriented MRO R-3m oriented MRO R-3m oriented

3.29

SRO c2mm not oriented

3.01

MRO c2mm not oriented MRO c2mm not oriented SRO c2mm not oriented SRO c2mm not oriented

2.81 2.60 2.43

50 ( 2% RH

60 ( 2% RH

LRO R-3m oriented LRO R-3m oriented MRO R-3m oriented MRO R-3m oriented SRO c2mm not oriented

SRO R-3m not oriented

MRO c2mm not oriented

mixture not oriented mixture not oriented mixture not oriented

SRO c2mm not oriented

70 ( 2% RH

85 ( 2% RH

LRO R-3m streaking oriented

LRO R-3m streaking oriented

LRO R-3m streaking oriented LRO R-3m streaking oriented

LRO R-3m streaking oriented LRO R-3m streaking oriented

SRO R-3m not oriented mixture not oriented

a LRO, MRO, and SRO represent long-, medium-, and short-range order, respectively. Streaking represents reciprocal space lattice points that have broadened in the direction perpendicular to the substrate. Additionally, experimental points labeled as mixture contain spherical micelles and cylindrical rods. These points only occur at low Si/EO ratios and high RH.

Figure 5. Trend in nanophase topology and orientation with respect to the substrate upon increases in the RH at Si/EO ) 3.62. The GISAXS patterns (a, c, e, and g) are from samples at 9%, 40%, 60%, and 70% RH, respectively. The FESEM micrographs (b, d, f, and h) correspond to the GISAXS patterns to their left.

potential candidate here; however, high-resolution TEM images would need to be collected to support this. 3.3. Effect of Si/EO Ratio. The general trend of decreasing the Si/EO ratio is shown in Figure 6 and may

Figure 6. General trend in nanophase topology and orientation with respect to the substrate upon changes in the Si/EO ratio. The GISAXS patterns are for Si/EO ratios of 4.24, 3.29, 2.81, and 2.43 for (a), (b), (c), and (d), respectively. The FESEM micrographs (b), (d), (f), and (h) correspond to the GISAXS patterns to their left. Note: All samples were prepared at 40% RH, except (c) and (d) which were prepared at 60% RH.

also be partially understood in terms of the shape factor. At high Si/EO ratios the effective area of the headgroup is larger since it is swelled by the associated silica

10118

Langmuir, Vol. 21, No. 22, 2005

Figure 7. (a) GISAXS pattern and (b) FESEM micrograph of films at low Si/EO ratios and high RH.

oligomers. As a result, we observe an oriented LRO R-3m structure due to the high curvature of the micellar building blocks. Upon decreasing the Si/EO ratio, we observe transition to a c2mm structure due to decreasing effective area of the headgroup and hence decreasing curvature. However, the more dramatic change observed with decreasing the Si/EO is a break down of orientational order, as evidenced from the rings in the GISAXS patterns in parts c, e, and g of Figure 6 compared to those in Figure 6a. The presence of a ring (as opposed to spots) indicates that domains with at least SRO sample many rotational states about an axis that is within the plane of the substrate. For the film in Figure 6c, local 6-fold symmetry is observed in the top-view FESEM image shown in Figure 6d, while a ring is observed in the GISAXS data (Figure 6c). This is explained by the presence of an ordered surface layer and an underlying disordered phase below the surface. For c2mm films this type of disorder was previously reported8 and is present in films shown in Figure 6e-h. Due to the oriented c2mm layers at the surface (and perhaps at the substrate), films with this structure do not have accessible mesopores. However, for the R-3m phase shown in parts c and d of Figure 6 this is not the case owing to the 3D nature of the phase and represents a very interesting region of composition/RH space due to the possibility of having mesopore access to both the vapor and the liquid phases. This loss of orientational order below the surface is systematic for lower Si/EO values over the RH range investigated. For Si/EO ratios greater than 3.62, RH conditions can be found that result in nanostructures that are highly oriented relative to the plane of the substrate. Ultimately, at low Si/EO ratios and high RH, a mixed and disordered phase results, Figure 7, and appears very similar in the FESEM micrographs to those reported by Okubo and co-workers for F127 templated films.34 4. Conclusions In this report we identify regions of differing nanophase topology, orientation, and order in mesoporous silica films on gold as a function of Si/EO ratio and RH using the combination of GISAXS and FESEM data. The results range from an oriented LRO R-3m mesostructure at 60%

Tate et al.

RH and 4.24 Si/EO ratio to an oriented MRO c2mm mesostructure at 9% RH and 3.62 Si/EO. In terms of the evolution of the well-ordered phases in response to changes in parameters that swell the headgroup (increasing RH and increasing Si/EO), the shape factor model provides a rational basis to understand the trends. However, the changes in order and orientation observed are not explained by any such geometric model and are likely due to the relative rates of assembly, condensation, and evaporation. Thus, the current study serves to illuminate the conditions needed to obtain various degrees of order in the films. Additionally, at very high RH, streaking of the reciprocal space lattice points occurs in the direction perpendicular to the substrate. The streaking may be explained by faulting of the planes parallel to the substrate. The faulting occurs due to an increase in the lateral stresses during drying at very high RH. The top-view FESEM images of the R-3m phase are tantalizing in that it appears that the highly ordered mesopores are open to the space above the film (and perhaps the substrate below). However this is not conclusively proven by the data shown above. Below the surface the “pores” could be more cagelike. Electrochemical studies are underway to address this issue and directly measure accessibility to the substrate. However, if it is found that the mesostructure is more cagelike, there are rational strategies that may be pursued to open windows between the cages. These include the use of hydrothermal treatments and the use of various organic additives or salts. Regardless of the accessibility of the mesopores in the highly ordered R-3m phase, there is most likely accessibility in some of the more disordered phases. During the transition from R-3m to c2mm or to mixed phases there is a transition from spherical micelles to cylindrical rods. This suggests that there are regions where pores merge and orientational order is lost. The combination of these two effects should provide direct access to the mesopores from the substrate and liquid/vapor phase above the film. Samples such as those shown in Figure 7 are examples of this type of order and morphology. Acknowledgment. The authors wish to acknowledge financial support from the National Science Foundation under the CAREER Award (0134255-CTS) and the Purdue Research Foundation (PRF-6904009). In addition, the NSF funded (MRI program award 0321118-CTS) facility for In-situ X-ray Scattering from Nanomaterials and Catalysts was used to collect GISAXS data. Additionally, the authors thank the Microstructural Analysis Facility at the School of Materials Engineering, Purdue University for FESEM usage and BASF for providing the triblock copolymer templates. LA050595H