Inner Morphology

The inner film morphology is investigated with GISAXS since x-rays can penetrate the whole sample. This advanced scattering technique offers the possibility to probe the full film volume over a macroscopic sample area and therefore the structural information of the probed sample is obtained in the nano- and meso-scale with a high statistical relevance.

The GISAXS experiments are performed at the Elettra SAXS beamline as described in

Figure 7.2: Power spectral density (PSD) functions obtained from the SEM data. Orange and blue curves are extracted from the small-pore and large-pore mesoporous titania films, respectively. The orange curve is shifted along the intensity axis for clarity of the presenta-tion. The peaks highlighted by black arrows indicate the brighter rings of the FFT patterns, whereas the weaker peaks are marked by gray arrows showing the position of the outer rings.

section 3.1.6. For these measurements the mesoporous titania films are produced on silicon substrates. A incident angle of 0.43^◦ and a sample-detector distance of 1.96 m are selected to obtain the desired q range.

Figure 7.3 shows the 2D GISAXS data of the small-pore sample and the large-pore sample. The horizontal black stripe at qz=0.8 nm⁻¹ and the vertical one at qy = 0.4 nm⁻¹ are the inter-module gaps of the detector. To protect the detector from oversatura-tion, the specular x-ray beam is shielded by a circular beamstop. The diffuse scattering maximum observable is called Yoneda peak, which is located at the critical angle of ti-tania. Directly from the 2D GISAXS data it can be seen that two distinct Bragg rods are positioned symmetrical around the center of qy=0 nm⁻¹ in lateral direction for both samples, indicating ordered arrangements in the film volume. However, the position of Bragg rods for both film differs. In the small-pore sample the Bragg rods reside sym-metrically at qy = ± 0.27 nm⁻¹, while two Bragg rods are closer to the Yoneda peak in the large-pore sample, located at qy = ± 0.13 nm⁻¹. The smaller q values in the large-pore sample suggest a larger predominant structural length scale in real space as compared to the small-pore sample. For quantitative analysis, horizontal line cuts at the Yoneda peak position are performed along the qy direction. For better statistics, the line cuts integrate five pixels which are centered at the Yoneda peak. The horizontal line cuts allow for extracting structural information of the probed titania samples such

Figure 7.3: 2D GISAXS data of the a) small-pore mesoporous titania film and of the b) large-pore mesoporous titania film. The intensity scale for the scattering signal is shown at the bottom of the images. The circular black area is the beamstop shielding the specular peak. The horizontal black stripe at q_z=0.8 nm⁻¹ and the vertical one at q_y = 0.4 nm⁻¹ correspond to the inter-module gaps of the used detector.

as domain size and domain center-to-center distance. The horizontal line cuts of both films are illustrated in figure 7.4a. The sharp and narrow peaks in both curves derive from the Brogg rods in the 2D GISAXS data for both samples. However, both curves demonstrate weak higher-order peaks, which are not observed in the 2D GISAXS data due to the relative low scattering intensity. The existence of higher-order peaks implies long-range lateral ordering of the mesopores inside both films. To obtain quantitative information about titania nanostructures, the line cuts are fitted with the same model as described in section 5.1.2. From the data modeling, two titania domain radii (form factors) and the corresponding center-to-center distances (structure factors) are extracted for both samples and displayed in figure 7.4b and 7.4c. For the small-pore sample, the small-sized titania structures have a domain size of (3.9 ±0.1) nm and a domain center-to-center distance of (23.5 ± 0.2) nm. According to equation 5.1, the pore size between small-sized structures is calculated to be (15.7 ± 0.2) nm. A characteristic domain size of (4.7 ±0.1) nm and the corresponding center-to-center distance of (22.1 ±0.3) nm are observed for the large-sized titania nanostructures, yielding a pore size of (12.7±0.3) nm.

For thelarge-poresample, titania domian radii of (4.9±0.1) nm and (7.2±0.1) nm with center-to-center distances of (50.0 ± 0.5) nm and (54.0 ± 3.1) nm are extracted for the small- and larger-sized titania structures, respectively, which gives rise to the mesopore sizes of (40.2 ±0.5) nm and (39.6±3.1) nm. Compared to the results obtained from the PSD analysis of the SEM data, the center-to-center distances of both films obtained from

Figure 7.4:a) Horizontal line cuts obtained from the 2D GISAXS data for titania films with different pore sizes.The orange and blue curves represent the mesoporous small-pore and large-pore samples, respectively. The grey lines show the fits to the data. The curves are shifted along the intensity axis for clarity of the presentation. Extracted characteristic length scales: b) titania domain radii and c) corresponding center-to-center distance. Triangles indicate small-sized structure and circles indicate large-sized structure. The red crosses in c) are the center-to-center distances calculated from SEM images.

SEM and GISAXS evaluation are quite similar, suggesting that the surface morphology is reproduced within the bulk of the sample.

In general, it is found that both small-pore and large-pore titania films demonstrate ordered mesostructural arrays in the sample surface and inside the films. As intended, the average pore sizes on these two samples are different. The low molecular weight PS-b-PEO leads to smaller titania nanostructures and mesopores, while the bigger titania nanostructures and mesopores are present in the film using the high molecular weight polymer template.