Ordered mesoporous titania thin films with ultra-large mesopores .122

5.4.2. Ordered mesoporous titania thin films with ultra-large

Figure 5.8. Various routes for preparation of mesoporous titania films by PS-b-PEO block copolymer.

Figure 5.9. (a) Photos of two precursor solutions which is prepared by dissolving PS-b-PEO in THF only (Route I), and adding 0.5 mL ethanol into the polymer solution containing PS-b-PEO and THF (Route II). (b) top-surface SEM image of as-prepared film prepared by Route II. (d, e) top-surface SEM images of titania films calcined at 350°C prepared by (d) Route Iand (e)Route II, respectively. (c, f) in-plane XRD profiles of (c) as-prepared and (f) calcined film prepared by Route II, respectively.

On the contrary, in the latter Route II, very ordered mesoporous structure with hexagonal arrangement was observed on the film surface (Figures 5.9e, 5.10). The average pore size calculated from SEM image was around 28 nm. Ethanol is well miscible with THF and water even in the presence of titanium based species, so the addition of ethanol benefits for the formation of isotropic solution to promote the formation of stable micelles. The mixed solution with ethanol appeared transparent (Figure 5.9a, right), expecting that all the components, such as polymers and inorganic species, were well dispersed in the solution. SEM observation of the corresponding as-prepared film showed well-ordered spherical PS-b-PEO micelles with TiO2 frameworks with good coverage over the substrate (Figure 5.9b). The size and periodicity of these spheres were well consistent with the mesopores observed in the calcined film (Figure 5.9e). Therefore, I conclude that the orderly arranged

spherical PS-b-PEO micelles play the role of the template for hexagonally arranged ultra-large mesopores. Even after the calcination, in-plane diffraction peaks were well retained and their positions were not changed (Figures 5.9c, f), indicating the mesostructural ordering along the direction parallel to the substrate preserved well without notable thermal shrinkage. The pore-to-pore distance calculated from d-spacing was 43 nm, which was close to the value (around 47 nm) observed by SEM (Figures 5.9b, e). N2 adsorption-desorption data showed Type IV isotherms typical for mesoporous material (Figure 5.11). The pore size calculated by BJH method was around 27 nm (inset ofFigure 5.11), which was close to the value observed in SEM images.

Figure 5.10. (a) Top-surface SEM image of as-prepared film prepared by dissolving PS-b-PEO in THF only (Route I). (b) Top-surface SEM image of calcined film prepared by adding 0.5 mL ethanol into the polymer solution containing PS-b-PEO and THF (Route II). The corresponding FFT diagram is shown in an inset image.

(a) (b)

Figure 5.11. Nitrogen adsorption-desorption isotherm of mesoporous titania films prepared by Route II followed by calcined at 450°C. Inset image is a pore size distribution calculated by a BJH method.

Interesting phenomenon is that the mixing process of PS-b-PEO with organic solvents (THF and ethanol) plays an important role for getting the long-range ordered mesoporous titania film. When PS-b-PEO was directly dissolved in a mixed solvent of THF and ethanol (Route III in Figure 5.8), the long-range ordering was much poorer than that of the film prepared by Route II (first completely dissolving PS-b-PEO in THF and then adding ethanol), although the role of spherical micelles as templates was obviously presented by observing the top surface of the as-prepared film (Figure 5.12). The reason is still unclear, but we can consider the following possibility. After subsequent addition of ethanol (as selective solvent for the PEO blocks) into molecularly dissolved PS-b-PEO in THF, unimer-aggregate transition of the polymers is thought to occur subsequently to some extent. These aggregates could be well-dispersed in the solution, because such aggregates are formed by moderately assembling the molecularly dissolved polymers. In contrast, in the case of Route III with mixed solvent solution, the unimer-aggregate transition proceeds before the complete unimer formation of the polymers, probably resulting in the poorly organized assembly of polymer micelles.

Figure 5.12. Top-surface SEM images of (a) as-prepared film and (b and c) calcined film at 350°C prepared by dissolving PS-b-PEO in the mixed solvent of THF and ethanol (Route III).

Hydrolysis rate of Ti species is another factor which has direct influence on morphology of mesoporous materials. It also affects the interaction and self-assembly process of polymer micelles. Controlled the hydrolysis is very important to obtain the targeted structures. As we know, TiCl4 is highly susceptible towards hydrolysis in water but the presence of sufficient HCl results in slowing down the hydrolysis process. It is known that the hydrolysis of TiCl4depends on the amount of HCl. When TiCl4 was dissolved into diluted HCl solution and then mixed with polymer solution (Route IV in Figure 5.8), the resulting solution became semi-transparent and cloudy.

In the as-prepared film, I could observe full bowl-like structures on the top surface and sphere-like structure embedded in the film (Figure 5.13a). The diameters of bowl- and sphere-like structures ranged from 200 nm to 600 nm, with almost fixed wall thickness of around 40 nm. After the calcination, bowls filled with nanoparticles together with spheres were observed (Figure 5.13b). I propose that the vesicular

structures instead of spherical micelles are formed as shown in Figure 5.13c. The wall thickness of around 40 nm was close to the double size of hydrophobic PS blocks, that is, the hydrophobic core size of normal micelles. The bowl-like structure can be regarded as a semi-vesicular structure (i.e., the vesicles are present on the top surface), while the some sphere-like structures were also observed inside the films.⁵ The Ti species attached to the outer surface of the bowl-like and vesicle-like structures and the solution filled in the space between two hollow polymer aggregates, thereby forming continuous framework after the calcination (Figure 5.13b). Meanwhile, the Ti species also attached to the inner surface of the hollow aggregates, resulting in the shell-like structures. The fine TiO2 nanoparticles irregularly deposited inside the vesicles were also observed, which was transformed from the Ti species incorporated in the hollow spaces.

Figure 5.13.Top-surface SEM images of (a) as-prepared film and (b) calcined film at 450°C prepared by hydrolysis of TiCl4in the diluted HCl solution (diluting conc. HCl aq ȝ/ ZLWK ȝ/ ZDWHU Route IV). (c) Schematic diagram of unimer;

spherical micelle and vesicle of PS-b-PEO block copolymers.

Such bowl-like and vesicular structures have been often seen in the previous report on mesoporous TiO2 films.⁵ In this case, long-time sonication was applied to transform bowl-like and vesicle-like structures into spherical micelles for well-ordered and uniform mesoporous structures in the films. The well ordering of PS-b-PEO-templated mesopores was also achieved by acetylacetone (AcAc) ligand assisted evaporation-induced self-assembly.⁶The AcAc reacted with the Ti alkoxide to form AcAc coordinated Ti complex. The resulted retard of hydrolysis and condensation speed of the complex favors formation of ordered structures. But in this case, the pore size was only 13 nm. In our study, well-ordered mesoporous TiO2 films can be simply synthesized without any such external and complicated treatments.

Titanium tetraisopropoxide (TTIP) or titanium tetra-n-butoxide (TTNB) of the same mole amount of Ti as TiCl4 was also used as Ti sources instead of TiCl4 (Route Vin Figure 5.8). By controlling the hydrolysis in concentrated HCl, similar ordered mesoporous structures were obtained in both cases without sonication or other treatment, as shown in Figure 5.14.

Figure 5.14. Top-surface SEM image of mesoporous titania films prepared by using (a) TTIP and (b) TTNP as Ti sources. These films are calcined ones at 350°C.

The pristine framework with closely embedded spherical micelles underwent crystal transformation from amorphous titania to crystallized anatase. Thermal stability of mesoporous structure at elevated temperature was checked by calcining the as-prepared films at different temperatures (Figures 5.15,5.16). When the calcination temperature reached 350°C, the crystallization of the pore walls started. Lattice

fringes were clearly observed in the pore walls (Figure 5.17)and electron diffraction patterns could be assigned to be polycrystalline anatase phase (Figure 5.18). With increase in the calcination temperatures, the crystallization of framework proceeds, accompanying with mass immigration. As a result, the mesopore ordering was deteriorated gradually, and some pores in a neighborhood fused to form wormhole-like mesopores (Figure 5.15). It was found that the mesoporosity was partially maintained up to 600°C (Figure 5.15). In previous studies, Pluronic P123 has been mostly utilized to synthesize mesoporous titania films.^15,16,20 According to these literatures, the original mesostructures (in the as-prepared films) are totally destroyed when the calcination temperature reaches to 500°C.^21,22 Such increase in calcination temperature induces the expansion of frameworks by robust anatase crystallization, leading to the complete collapse of the original mesostructures.

Figures 5.15.SEM images of mesoporous titania films prepared by Route IIfollowed by calcined at various temperatures: (a) 450, (b) 500, (c) 550, and (d) 600°C, respectively.

Figure 5.16. Low-magnified TEM images of mesoporous titania films prepared by Route II followed by calcined at various temperatures: (a) 350, (b) 450, (c) 500, (d) 550, and (e) 600°C, respectively.

Figure 5.17. High-magnified TEM images of mesoporous titania films prepared by Route II followed by calcined at various temperatures: (a) 350, (b) 450, (c) 500, (d) 550, and (e) 600°C, respectively.

Figure 5.18. Electron diffraction patterns of mesoporous titania films prepared by Route II followed by calcined at various temperatures: (a) 350, (b) 450, (c) 500, (d) 550, and (e) 600°C, respectively.

To clearly understand this reason, I measured TG curves of the used polymer PS-b-PEO in this study (Figure 5.19a) and one commonly used commercially available Pluronic P123 (Figure 5.19B). The Pluronic P123 showed that the sharp weight loss was at around 180°C and the completely burning out was over 250°C, which was much lower than the crystallization temperature of titania (~350°C), and also demonstrated in our previous study.¹⁴ Thus, at temperatures higher than 250°C, there were no templates in the films. Therefore, the crystal growth was not suppressed and the original mesoporous structure easily collapsed as increasing the calcination temperature. However, TG curve of the used polymer in this study (Figure 5.19a) showed only 40 wt% was lost at 300°C and the template was totally removed at around 450°C. Therefore, the remained polymer derivatives, e.g., carbon matrix, can act as a support to prevent the rapid crystal growth, resulting in the retention of the original mesoporous structure.

Figure 5.19. TG-DTA curves of (a) PS-b-PEO and (b) P123 in the range of room temperature to 500°C with a ramping rate of 1 °C·min^-1in air.

5.5. Conclusion

In conclusion, solvent engineering combined with controlling hydrolysis of the precursor have been developed to prepare ordered mesoporous alumina and titania thin films with ultra-large mesopores templated by diblock copolymer PS(18000)-b-PEO(7500). Economic inorganic salts AlCl3 or TiCl4 as aluminum or titanium feedstock were successfully developed to prepare ordered mesoporous alumina or titania thin film with ultra-large mesopores of 35 or 28 nm respectively. In these procedures, ethanol was selected as the co-solvent to improve the miscibility of THF and water in presence of inorganic salt. I found that the amount of ethanol has a significant influence on the formation, morphologies and aggregation of micelles.

Addition of ethanol could greatly change the aggregation behavior of PS-b-PEO.

Without adding ethanol, the PS-b-PEO molecularly dispersed in THF seriously aggregated after mixing with aqueous Ti precursor solution, failing in serving as the structure directing agent for mesopores. The pore size can be tuned into macropore scale by simply adjusting the amount of co-solvent ethanol. This procedure are expected to developed for other metal oxide thin films with ordered large-size pores by employing diblock copolymer PS-b-PEO with relative low molecular weight as the template. Under the typical condition with suitable ethanol amounts, no significant changes were observed in the micelle morphology when the amount of PS-b-PEO in solution varies within a long range. On the other hand, the pore connectivity was

highly improved with increase of PS-b-PEO amount. The obtained metal oxide thin films enable many emerging applications such as biotechnology with large-sized targeted species, catalysis, gas sensing, optics and photovoltaics.

References

1 K. Yu, A. Eisenberg, Macromolecules, 1996,29, 6359.

2 K. Yu, A. Eisenberg, Macromolecules, 1998,31, 3509.

3 Y. Mai, A. Eisenberg, Acc. Chem. Res.,2012,45, 1657.

4 Y. J. Cheng, J. S. Gutmann, J. Am. Chem. Soc.,2006,128, 4658.

5 Y. Zhang, H. Li, Y. Q. Liu, J. Wang,Chem. Commun., 2012,48, 8538.

6 D. Feng, W. Luo, J. Zhang, M. Xu, R. Zhang, H. Wu, Y. Lv, A. M. Asiri, S. B.

Khan, M. M. Rahman, G. Zheng, D. Zhao, J. Mater. Chem. A,2013,1, 1591.

7 M. Kuemmel, J. Allouche, L. Nicole, C. Boissière, C. Laberty, H. Amenitsch, C.

Sanchez, D. Grosso,Chem. Mater.,2007,19, 3717.

8 D. Chandra , T. Ohji , K. Kato, T. Kimura, Phys. Chem. Chem. Phys., 2011, 13, 12529.

9 D. Chandra, M. Bekki, M. Nakamura, S. Sonezaki, T. Ohji, K. Kato, T. Kimura,J.

Mater. Chem.,2011,21, 5738.

10 T. Brezesinski, M. Antonietti, B. M. Smarsly, Adv. Mater.,2007,19, 1074.

11 D. Fattakhova-Rohlfing, M. Wark, T. Brezesinski, B. M. Smarsly, J. Rathouský, Adv. Funct. Mater.,2007,17, 123.

12 T. Coquil, C. Reitz, T. Brezesinski, E. J. Nemanick, S. H. Tolbert, L. Pilon, J.

Phys. Chem. C,2010,114, 12451.

13 E. Ortel, A. Fischer, L. Chuenchom, J. Polte, F. Emmerling, B. M. Smarsly, R.

Kraehnert, Small,2012,8, 298.

14 M. B. Zakaria, N. Suzuki, N. L. Torad, M. Matsuura, K. Maekawa, H. Tanabe, Y.

Yamauchi, Eur. J. Inorg. Chem.,2013,2013, 2330.

15 Y. Yamauchi, N. Suzuki, L. Radhakrishnan, L. Wang, Chem. Rec.,2010,9, 321.

16 J. L. Vivero-Escoto, Y. D. Chiang, K. C. W. Wu, Y. Yamauchi, Sci. Technol. Adv.

Mater.,2012,13, 013003.

17 Y. S. Seo, M. W. Kima, D. H. Ou-Yang, D. G. Peiffer, Polymer,2002,43, 5629.

18 Y. Zhang, W. Lin, R. Jing, J. Huang, J. Phys. Chem. B,2008,112,16455.

19 L. Zhang, A. Eisenberg, Macromolecules,1999,32, 2239.

20 C. W. Wu, T. Ohsuna, M. Kuwabara, K. Kuroda, J. Am. Chem. Soc., 2006, 128, 4544.

21 H. Oveisi, S. Rahighi, X. Jiang, Y. Nemoto, A. Beitollahi, S. Wakatsuki, Y.

Yamauchi,Chem. Asian J.,2010,5, 1978.

22 T. Kimura, Y. Yamauchi, N. Miyamoto, Chem. Euro. J.,2011,17, 4005.