In Situ GISAXS Study of the Formation of Mesostructured Phases

formation mechanism within the confined space of the AAM pores is a direct result of this study. ... in the past few years.5,6 The EISA approach was r...
0 downloads 0 Views 457KB Size
Download PDF
5018

Langmuir 2008, 24, 5018-5023

In Situ GISAXS Study of the Formation of Mesostructured Phases within the Pores of Anodic Alumina Membranes Barbara Platschek, Ralf Ko¨hn, Markus Do¨blinger, and Thomas Bein* Department of Chemistry and Biochemistry, UniVersity of Munich, Butenandtstrasse 5-13 (E), 81377 Munich, Germany ReceiVed April 14, 2007. In Final Form: September 28, 2007 The formation and subsequent transformations of mesostructured silica within the confined tubular environment of anodic alumina membrane (AAM) channels [porous alumina membrane (PAM) channels] were investigated for the first time in situ with grazing incidence small-angle X-ray scattering (GISAXS) techniques, in combination with ex situ transmission electron microscopy (TEM) of the same samples. A better understanding of the mesostructure formation mechanism within the confined space of the AAM pores is a direct result of this study. Three different surfactants were used as the structure-directing agents in acid-catalyzed silica synthesis solutions. With ionic cetyltrimethylammonium bromide acting as the structure-directing agent, a columnar hexagonal structure with mesopores oriented parallel to the AAM channels was observed to form directly from the beginning of the synthesis. In samples synthesized with the nonionic surfactants Brij 56 and Pluronic P123, a circular hexagonal structure was found to form first; here, the mesopores are aligned around the circumference of the AAM channels. The circular structure subsequently transforms directly into a columnar hexagonal (P123 surfactant), or a mixture of columnar hexagonal and a new curved lamellar phase with lamellae oriented parallel to the walls of the AAM channels (Brij 56 surfactant). These transformations occur after complete solvent evaporation and therefore differ from a simple evaporation-induced phase formation. The existence of a previously postulated lamellar phase could be proven by GISAXS and TEM investigations.

Introduction Periodic mesoporous materials are of great interest for a variety of applications such as catalyst supports, molecular sieves, separation membranes, hosts for conductive nanostructures, or potential gain media in laser technology.1-4 The preparation of mesoporous silica thin films on various flat substrates by evaporation-induced self-assembly (EISA) has been established in the past few years.5,6 The EISA approach was recently used on porous anodic alumina membranes (AAMs) [porous alumina membranes (PAMs)] to synthesize mesostructured silica within the vertical channels of the membranes.7-11 At about the same time, the synthesis of mesoporous silica inside AAMs via a solgel method was developed.12-15 Furthermore, the formation of other mesoporous oxides and mesoporous carbon replicas in AAMs was reported.16-19 The unique combination of anisotropic * Corresponding author. Phone: +49-89-2180-77623. FAX: +49-892180-77622. E-mail: [email protected]. (1) Hoffmann, F.; Cornelius, M.; Morell, J.; Froeba, M. Angew. Chem., Int. Ed. 2006, 45, 3216. (2) Loerke, J.; Marlow, F. AdV. Mater. 2002, 14, 1745. (3) Petkov, N.; Stock, N.; Bein, T. J. Phys. Chem. B 2005, 109, 10737. (4) Ye, B.; Trudeau, M. L.; Antonelli, D. M. AdV. Mater. 2001, 13, 561. (5) Brinker, C. J.; Lu, Y.; Sellinger, A.; Fan, H. AdV. Mater. 1999, 11, 579. (6) Grosso, D.; Babonneau, F.; Albouy, P.-A.; Amenitsch, H.; Balkenende, A. R.; Brunet-Bruneau, A.; Rivory, J. J. Chem. Mater. 2002, 14, 931. (7) Yamaguchi, A.; Uejo, F.; Yoda, T.; Uchida, T.; Tanamura, Y.; Yamashita, T.; Teramae, N. Nat. Mater. 2004, 3, 337. (8) Wu, Y.; Cheng, G.; Katsov, K.; Sides, S. W.; Wang, J.; Tang, J.; Fredrickson, G. H.; Moskovits, M.; Stucky, G. D. Nat. Mater. 2004, 3, 816. (9) Ku, A. Y.; Taylor, S. T.; Loureiro, S. M. J. Am. Chem. Soc. 2005, 127, 6934. (10) Platschek, B.; Petkov, N.; Bein, T. Angew. Chem., Int. Ed. 2006, 45, 1134. (11) Yoo, S.; Ford, D. M.; Shantz, D. F. Langmuir 2006, 22, 1839. (12) Yang, Z.; Niu, Z.; Cao, X.; Yang, Z.; Lu, Y.; Hu, Z.; Han, C. C. Angew. Chem., Int. Ed. 2003, 42, 4201. (13) Lu, Q.; Gao, F.; Komarneni, S.; Mallouk, T. E. J. Am. Chem. Soc. 2004, 126, 8650. (14) Yao, B.; Fleming, D.; Morris, M. A.; Lawrence, S. E. Chem. Mater. 2004, 16, 4851. (15) Wang, D.; Kou, R.; Yang, Z.; He, J.; Yang, Z.; Lu, Y. Chem. Commun. 2005, 166.

spatial constraints and micellar interactions with the curved AAM channel walls leads to the formation of new and unusual mesophase morphologies with different pore orientations.8,14 In the case of an encapsulated two-dimensional (2D)-hexagonal mesophase, the mesostructure pores are either aligned parallel along the AAM channels (“columnar hexagonal” orientation), or they assume a “circular hexagonal” orientation perpendicular to the AAM channels. Both orientations were found to coexist under certain conditions, depending on the composition of the precursor mixtures as well as synthesis conditions, e.g., the humidity during the evaporation process.10 This behavior illustrates the high sensitivity and flexibility of the system. Here we report an in situ grazing incidence small-angle X-ray scattering (GISAXS) study, aimed at a detailed understanding of the formation process of the mesostructured material inside the confined environment of the AAM channels. To our knowledge, this is the first time that the formation of structured material within the ordered pores of a substrate was investigated by in situ GISAXS. The formation of different structure types in mesoporous films or powders has been well studied in the past. However, the existence of two different orientations of one and the same structure has not been observed in films on flat substrates and is related to the unique confined geometry of the porous substrate investigated here. The energetic difference between the circular and the columnar hexagonal structure seems to be quite small because the production of pure phases remains a challenge. The formation mechanism of the two orientations is unknown, and, until now, no information could be obtained on whether both (16) Cott, D. J.; Petkov, N.; Morris, M. A.; Platschek, B.; Bein, T.; Holmes, J. D. J. Am. Chem. Soc. 2006, 128, 3920. (17) Ku, A. Y.; Taylor, S. T.; Heward, W. J.; Denault, L.; Loureiro, S. M. Microporous Mesoporous Mater. 2006, 88, 214. (18) Lai, W. H.; Shieh, J.; Teoh, G.; Hon, M. H. Nanotechnology 2006, 17, 110. (19) Xiao, Y.; Li, L.; Li, Y.; Fang, M.; Zhang, L. Nanotechnology 2005, 16, 671.

10.1021/la7010937 CCC: $40.75 © 2008 American Chemical Society Published on Web 04/10/2008

Formation of Mesostructured Phases within AAMs

Langmuir, Vol. 24, No. 9, 2008 5019

Figure 1. Geometry for the in situ GISAXS experiments. The incident X-rays hit the sample almost parallel to the planar AAM surface, the primary beam is fully shadowed by a beam stop. The samples were mounted on a balance recording the weight loss.

orientations are formed from the beginning of the reaction or whether one phase is developed via conversion of the other phase. A detailed understanding of how the mesostructured material is formed inside the AAM channels will provide us with greater control over the structure and orientation of the synthesized products and hence the adaptation of the system for future applications. Experimental Methods Sol Preparation. A 2.08 g (0.01 mol) portion of tetraethyl orthosilicate (TEOS, Aldrich) was mixed with 3 g of 0.2 M HCl(aq), 1.8 g of H2O, and 5 mL of EtOH and heated at 333 K for 1 h to accomplish acid-catalyzed hydrolysis-condensation of the silica precursor. For the preparation of cetyltrimethylammonium bromide (CTAB)-containing deposition mixtures, this solution was mixed with 0.947 g (2.6 mmol) of CTAB dissolved in 10 mL of EtOH. The decaethylene glycol hexadecyl ether (Brij 56)-containing samples were prepared by adding 1.81 g (2.65 mmol) of Brij 56 dissolved in 30 mL of ethanol. For the preparation of the poly(ethylene oxide)20-poly(propylene oxide)70-poly(ethylene oxide)20 (P123)containing solutions, the prehydrolyzed silica was mixed with 15 mL of a 5 wt % ethanolic solution of P123 (0.13 mmol). The AAMs (47 mm, Anodisc, Whatman) with average pore diameters of 120200 nm and thicknesses of approximately 60 µm were mounted on a balance and soaked with the prepared precursor mixtures by distributing 0.75 mL over the whole membrane surface immediately before starting the in situ measurements. During the EISA process, the ambient conditions were kept at 35-40% relative humidity and 299 K, respectively. Transmission Electron Microscopy (TEM). The TEM images were obtained with a JEOL 2011 transmission electron microscope operating at 200 kV. The samples were prepared after the in situ experiments by dimple grinding followed by argon ion polishing. GISAXS. The measurements were performed at the 1-BM beamline at the APS at ANL, Argonne, IL (for the geometry of the measurement see Figure 1). The sample-detector distance was 1040 mm, and the wavelength of the primary beam was 1.24 Å. The angle between the primary beam and the sample substrate was 0.175° (2θ ) 0.35°). He-filled tubes between the final beamline slit and the sample and between the sample and the detector were used to avoid additional scattering in air. Every 3 min, a diffraction pattern was recorded. After every measurement, a background image of the detector was recorded. The time intervals between the measurements were limited by the detector readout speed. Evaluation of Data. The intensity of the reflections was determined for each frame by integrating the area within the rectangles depicted in Figure 2 along qx or qz, respectively (insets). The area under the peaks was fitted to obtain the intensities of the respective diffraction spots. The vertically oriented channels of the amorphous alumina membrane lead to a strong background in the qx-direction. Thus, in case of the reflections in the qx-direction, the diffuse scattering

Figure 2. Diffraction pattern of the sample synthesized with Brij 56 as the structure-directing agent, recorded 45 min after the initiation of the experiment. The scale bar is given in q ) 2π/d. The rectangles mark the integrated areas for the determination of the 10r and 01r intensities (for indexation, see Figure 3). Corresponding profiles after first integration along qx or qz are given as insets. The comparatively weak, sharp reflections correspond to a hexagonally structured layer on top of the AAM (s denotes surface), the broader, bright reflections correspond to a circular hexagonal structure formed inside the AAM channels (p denotes pore). The reflections from structures formed on the top layer or within the channels are positioned on a ring representing the same 2θ value, but occur at different azimuthal angles. The in-plane (01r) reflections are superposed by diffuse scattering from the AAM. As the channels of the AAM are aligned along the z-axis, the diffuse scattering is mainly observed in the qx-direction, and it becomes weaker with increasing distance from the primary beam. resulting from the alumina membrane was fitted and subtracted prior to calculating the area under the peak. If two reflections with different q-values were observed in the qx-direction, corresponding to hexagonal and lamellar phases (see results), they were fitted separately. For two samples, additional formation of structured material on top of the membranes was observed. These surface films were usually scratched off after complete synthesis and were not further investigated. The usually much weaker but sharper reflections of the surface structures occur at different azimuthal angles than the ones from phases within the AAM channels; the patterns do not interfere (Figure 2).

5020 Langmuir, Vol. 24, No. 9, 2008

Figure 3. (a-d) Sketch of the observed structures within the AAM channels and description of their intensity-maxima in reciprocal space, and their respective diffraction pattern corresponding to the 10-type reflections of a normal hexagonal (a-c) or lamellar (d) lattice. The circular hexagonal structure forming rings (a) or spirals (b) results in two 01r and two visible 10r reflections (the other two -10r reflections are obscured by the sample holder). The index “r” denotes “ring” for reflections from ring-shaped intensity maxima, indexation is referring to a normal hexagonal lattice.21 The columnar hexagonal structure (c) and the tubular lamellar phase (d) both result in two reflections in the horizontal plane of the primary beam. In the case of the columnar phase, the indexing of one reflection as 01 defines a coordinate system, hence the opposed ip reflection is indexed 0-1. In contrast, in the case of the lamellar phase, both ip reflections stem from the same ring-shaped intensity maxima and therefore are both indexed as 01r.

Results Three different systems have been investigated, using either the ionic surfactant CTAB, the nonionic triblock copolymer P123, or the nonionic diblock copolymer Brij 56 as structure-directing agents (for details, see Experimental Methods). An overview of the observed structures inside the AAM channels, and their respective diffraction patterns in the given diffraction geometry are depicted in Figure 3. The cylindrical geometry of the circular hexagonal phase, forming either rings or spirals (Figure 3a,b), is also represented in reciprocal space. The intersection of the ring-shaped intensity maxima with the Ewald sphere gives an X-ray diffraction pattern showing two reflections in the horizontal plane of the primary beam and two out-of-plane reflections, all corresponding to 10-type reflections of a normal hexagonal lattice20,21 (01r and 10r according to Figure 3a,b; the index r (20) Marlow, F. F.; Leike, I.; Weidenthaler, C.; Lehmann, C. W.; Wilczok, U. AdV. Mater. 2001, 13, 307. (21) Smarsly, B.; Gibaud, A.; Ruland, W.; Sturmayr, D.; Brinker, C. J. Langmuir 2005, 21, 3858.

Platschek et al.

denotes “ring” for reflections from ring-shaped intensity maxima). The columnar hexagonal structure (Figure 3c) exhibits only two reflections (01 and 0-1) in the horizontal plane of the primary beam. In this case, the rotational symmetry in reciprocal space results from numerous crystal-like domains rotated against each other around the z-axis, parallel to the pore channels (Figure 3c). The tubular lamellar structure depicted in Figure 3d shows ring-shaped intensity maxima in reciprocal space that reflect the ring shape of the silica sheets and that result in two 01r reflections. In an ideal model, the lamellar sheets show increasing curvature when located closer to the center, finally forming one tube. In summary, the observation of reflections out of the horizontal plane of the primary beam (10r) provides evidence for the presence of the circular hexagonal structure only. In contrast, the reflections in the qx-direction can result from the 01r reflections, or the 01and 0-1 reflections of the circular hexagonal, the tubular lamellar, or the columnar hexagonal phase, respectively. The tubular lamellar structure postulated in previous reports8,15 is not easily confirmed for the following reasons. In the given diffraction geometry, both lamellar and columnar hexagonal phases only show reflections along qx. Still, it is possible to distinguish the lamellar phase from the hexagonal phases by their first-order reflections since they appear at different 2θ values. In a lamellar structure, the lattice constant matches the d-spacing, but in hexagonal structures the d-spacing is smaller by a factor of sin(120°). However, this approach is limited when the d value becomes larger, as the intensity maxima will move closer together until it becomes impossible to separate them. Moreover, it will always be necessary to show diffraction data from a hexagonal and a lamellar phase to prove their existence. TEM is another powerful tool to characterize the mesoporous silica-AAM-composite materials. Nevertheless, when the samples are viewed normal to the membrane surface (plan-view), the circular hexagonal and tubular lamellar structures look identical, as do cross-sectional views from the columnar hexagonal and the lamellar structure. To truly confirm the existence of a lamellar phase, it is necessary to combine either the diffraction pattern with plan-view TEM images, or plan-view and cross-sectional micrographs. Here we show for the first time diffraction data that sufficiently demonstrate the formation of a tubular lamellar phase inside the channels of an AAM. In order to evaluate the in situ data, the temporal evolution of the integrated intensities of the spots corresponding to the structures inside the AAM channels was analyzed for each sample (Figure 4a-c).22 Furthermore, the weight-loss related to the evaporation of the solvent as a function of time is displayed in Figure 4d. A list of the d values and respective lattice constants of each sample is given in Table 1.

Discussion CTAB. Using CTAB as an ionic surfactant, intrachannel mesostructure was generated within 50 min, when most of the solvent had already evaporated and the critical micelle concentration (cmc) was exceeded (Figure 4a,d). The intensity of the 01 reflection increases strongly within the following 10 min, reflecting increasing order of the structure. In the following period, the slope levels off, almost reaching a plateau after 100 min. At this time, the sample has almost reached its final weight, suggesting that solvent evaporation is mostly completed. The intensity of the 10r reflection is extremely weak, leading to the conclusion that no significant amount of the circular hexagonal (22) Doshi, D. A.; Gibaud, A.; Goletto, V.; Lu, M.; Gerung, H.; Ocko, B.; Han, S. M.; Brinker, C. J. J. Am. Chem. Soc. 2003, 125, 11646.

Formation of Mesostructured Phases within AAMs

Langmuir, Vol. 24, No. 9, 2008 5021

Figure 4. Intensities of the 10r, 01r, or 01 diffraction spots plotted versus time for the samples synthesized with CTAB (a), Brij 56 (b), and P123 (c) as structure-directing agents. The diffraction patterns of the sample synthesized with Brij 56 (b) show two distinguishable reflections in the qx-direction assigned to the columnar and circular hexagonal and the tubular lamellar phase. The weight loss for each sample is plotted versus time (d). Mesostructure was observed during the time period indicated by filled symbols. Table 1. List of d-Spacings and Respective Lattice Constants for the Samples after ca. 60 min sample

d-spacing (nm)

CTAB Brij 56 P123

4.2 5.6/6.3 10.4

lattice constant (nm) 4.8 6.5/6.3 12.0

phase has been formed. TEM images confirm that the main phase is columnar hexagonal (Figure 5d). These results provide a better understanding of the formation mechanism of the mesostructure within the AAM pores. In a recent article, the formation of a similar hexagonal structure in the direct vicinity of the pore wall was ascribed to the adsorption of cationic CTA to the alumina wall.7 As a result of this adsorption, the hexagonal structure was supposed to form even before the cmc is reached within the entire pore. However, as demonstrated in our present study, the formation of structure starts very late in the evaporation process. Therefore, we propose that the cmc is reached first, followed by structure formation in the vicinity of the alumina wall, which acts as a heterogeneous nucleation center. From there, the structure formation into the center of the pore continues rapidly, leading to the highly ordered hexagonal structure as depicted in the TEM in Figure 5d. The mesostructures made with the nonionic surfactants P123 or Brij 56 exhibit 10r reflections of considerable intensity (Figure 4b-c). However, as the structures evolve, in both cases, the intensity of the 10r reflections does not reach a plateau, but instead decreases significantly with time after showing an early maximum, indicating the metastability of the circular hexagonal phase. In both samples produced with nonionic surfactants, the reflections along qx show increasing intensity, even after the intensity maximum of the respective 10r reflections has been passed. These

results strongly support our earlier hypothesis10 that the circular hexagonal phase is kinetically favored but that it does not represent the thermodynamically stable phase in these mesostructured systems. Brij 56. When using Brij 56 as the structure-directing agent, all three mesophase orientations are observed in the diffraction pattern (Figure 4b). The reflections in qx-direction corresponding to a larger d-spacing are assigned to the tubular lamellar phase (01r), and the reflections corresponding to the smaller d-spacing are assigned to the columnar and circular hexagonal phases (01r/ 01 or 01r/0-1). The ratio of their d values is 1.0:0.9, which is in good agreement with the theoretically expected value of sin(120°). In the first 20 min of detectable structure formation, the intensity of the 01r/01 reflection progresses in the same manner as the intensity of the 10r reflections. In the following 30 min of the synthesis, the intensity of the 10r reflections vanishes completely, while the intensity of the 01r/01 reflection increases to finally reach a plateau after 95 min following the beginning of the experiment. Hence, the circular hexagonal structure forms first and is then transformed into the columnar structure. The tubular lamellar structure forms after 45 min, roughly at the same time as the columnar structure starts to form. At the end of the synthesis, the circular structure has completely transformed into the columnar and lamellar phases. This is confirmed by TEM images, showing a well-preserved columnar phase and a degraded tubular lamellar phase (the latter was observed to collapse after irradiation by the electron beam) (Figure 5 e). It should be noted that the columnar phase most often resides in the center of the AAM channels, surrounded by the lamellar phase. The silica/surfactant lamellae are more and more curved when located closer to the center of the AAM channels (see sketch of structure in Figure 3d). Apparently, the mesostructure

5022 Langmuir, Vol. 24, No. 9, 2008

Platschek et al.

Figure 5. (a,b,c). Diffraction patterns from different samples. (a) Sample synthesized with CTAB and recorded after 70 min, (b) synthesized with Brij 56 and recorded after 57 min, and (c) synthesized with P123 and recorded after 63 min. The pattern in panel a shows mainly the formation of the columnar structure. The pattern in panel b shows the 01r reflections closer to the primary beam that correspond to a lamellar structure, as well as additional reflections along qx from the columnar and circular hexagonal structures (01r/0-1 or 01r/01). The 10r reflections corresponding to the circular hexagonal structure are also present in this pattern. The pattern in (c) shows the existence of a circular and a columnar hexagonal phase. The strong superposition of the reflections along qx with the primary beam made it impossible to confirm or exclude a possible second maximum corresponding to a lamellar phase. Instead, the formation of a lamellar phase was excluded with high probability from the corresponding TEM images (see text). (d,e,f) Corresponding plan-view transmission electron micrographs showing the plane perpendicular to the AAM-channel axis. The columnar hexagonal mesophase was found in the sample synthesized with CTAB (d); the sample synthesized with Brij 56 as surfactant (e) shows a mixture of the columnar hexagonal and the lamellar structure, (where the lamellar phase collapsed when irradiated with electrons). This example is representative for most cases, showing the lamellar phase located at the AAM channel wall surrounding the columnar hexagonal phase in the center of the channels. The sample synthesized using P123 shows a mixture of the circular and the columnar hexagonal phase (f). In contrast to the lamellar phase in the sample synthesized with Brij 56, no degraded lamellar phase was observed.

Formation of Mesostructured Phases within AAMs

avoids these high curvatures by forming a columnar hexagonal structure in the center of the AAM channels. Pluronic P123. Mesostructure synthesis using P123 as a template results in the first detectable structure after 42 min, and the intensity curves of the 10r and 01r/01 reflections show the same shape for up to 60 min (Figure 4c). In the following 30 min, the intensity of the 10r reflections strongly decreases, while the intensity of the 01r/01 reflection reaches a plateau. After 90 min, the intensity of all reflections decreases simultaneously, which suggests a general decrease in the mesophase order. The lamellar and the columnar hexagonal structure could not be distinguished in this case, as the corresponding reflections for P123 are located too close to the primary beam, where the superposition with the diffuse scattering from the AAM is relatively strong (compare Figure 5c). In combination with the short distance between the first-order reflections of both structures, it was impossible to distinguish more than one maximum. However, the intensities of both the 10r and 01r/01 reflections remain at a high level, thus the circular phase as well as the columnar hexagonal or the lamellar phase are present after the synthesis is completed. The coexistence of the circular and the columnar hexagonal phase within this sample was confirmed by TEM (Figure 5f). We did not observe any collapsed or distorted phase within this sample, indicating the absence of a lamellar phase. For all three structure-directing agents, the start of the structure formation is correlated with the weight loss. All samples have roughly the same weight of 0.6 g (i.e., 0.75 mL of precursor solution distributed on the AAMs) at the initiation of the experiment. For the sample synthesized with Brij 56, the solvent evaporation is faster than that for the samples synthesized with P123 and CTAB, respectively. This is assumed to originate from the different solvation energies of the surfactants. The same trend is observed for the start of the structure formation. While the Brij 56 sample shows the first reflections after 35 min, the mesostructure of the P123 sample occurs after 42 min, and the mesostructure formation in the CTAB sample starts after 50 min. At these respective times, the remaining weight of the

Langmuir, Vol. 24, No. 9, 2008 5023

samples was around 25% of their initial weight. For both samples synthesized with the nonionic surfactants P123 or Brij 56, it should be noted that, at the point where the transformations between ordered phases start, the final sample weight is already reached. Hence, these transformations are independent of ongoing solvent evaporation.

Conclusions In summary, we have performed GISAXS experiments to investigate in situ the formation and subsequent transformations of mesostructured silica inside the confined tubular environment of AAM channels. New insights were gained regarding the formation mechanism of mesostructures within the confined space of the AAM pores. Using ionic CTAB as the structure-directing agent, the columnar hexagonal structure was observed to form directly from the beginning. In the samples synthesized with the nonionic surfactants, a circular hexagonal structure was found to form first and to directly transform into the columnar hexagonal (P123) or a mixture of the columnar hexagonal and the tubular lamellar phase (Brij 56). The methodology used here allows us to directly observe phase transformations in these confined mesoporous systems; they start after complete evaporation of the solvents and are either partial (P123) or complete (Brij 56). The existence of a previously postulated tubular lamellar phase could be proven by GISAXS and TEM investigations. This work demonstrates the importance of in situ structural studies for the creation and structural control of hierarchically ordered porous nanostructures in confined environments and the understanding of the mechanisms involved. Acknowledgment. The SFB 486 of the Deutsche Forschungsgemeinschaft is acknowledged for financial support. The authors thank the APS (Argonne National Laboratory, Argonne, IL) for allocating beam time, and especially Jan Ilavsky, Xuefa Li, Jin Wang, and Carlee Ashley for their support during the experiments. LA7010937