Material characterizations

Figures 4-1(a) and (b) show the plane-view SEM images of the as-prepared hydrothermally synthesized TiO₂ thin film and the porous film annealed at 600^oC for 3 hr (thereafter denoted as TiO2-3h), respectively. The thin film had a carpet-like morphology with collapsed thin sheets on the surface. According to the cross-sectional SEM image

39

shown in Fig. 4-1(c), the hydrothermalized thin film had a two-layer structure with the carpet-like layer on the top and an ill-compacted granular layer on the bottom. The top and bottom layers were ~260 nm and ~65 nm thick, respectively. It is not clear to us at the present time why the two different microstructure layers were formed during the one step hydrothermal treatment. The top layer was virtually of a porous structure, which was constructed by mutually overlaying collapsed sheets. A careful examination from the cross-sectional SEM image of Fig. 4-1(c) reveals that the pore walls were composed of

Figure 4-1 SEM images for (a) the as prepared porous TiO2 thin film, (b) porous TiO2-3h thin film, (c) cross-section of the porous TiO2-3h thin film, (d) Pt nanoparticles deposited on the porous TiO2-3h thin film, (e) Pt particles deposited on the blanket TiO2 thin film, (f) Pt islands on the Ti substrate, and (g) blanket Pt thin film on the Ti substrate.

40

nanosized grains connecting with one another, and with a wall thickness roughly the diameter of a grain, which was ~15-25 nm. Compared to the blanket Ti thin film before hydrothermalization, the porous morphology of the hydrothermal TiO2 thin film provided a much larger surface area for the Pt catalyst loading. Figure 4-1(d) shows the plane-view SEM image of the TiO2-3h thin film after the pulse electrodeposition of Pt catalyst (thereafter denoted as Pt/TiO₂-3h). Because the electrodeposited Pt nanoparticles were very small in size and well dispersed on the TiO2 pore structure as will be shown by TEM images, they are hardly observed in the SEM image. Also shown in Fig. 4-1 are the SEM images of electrodeposited Pt particles on the blanket TiO2 surface (Fig. 4-1 (e), Pt islands on the Ti thin film (Fig. 4-1(f) and the e-beam deposited Pt thin film (Fig. 4-1(g). Pt particles on the blanket TiO2 thin film (thereafter denoted as Pt/TiO2-tf) had a relatively uniform size distribution around ~15-20 nm and well dispersed on the surface. For Pt islands deposited on the Ti surface (denoted as Pt/Ti), islands with sizes mostly larger than 300 nm covered nearly the whole Ti substrate surface. On the other hand, the e -beam deposited Pt thin film completely capped the Ti substrate and will be denoted as the blanket-Pt electrode. These three samples will be used for comparison when we discuss later about the electrocatalytic activity of the Pt/TiO2-3h electrode.

TiO₂ is an n-type semiconductor, and the electrical conductivity of TiO₂ increases with the concentration of donor-like oxygen vacancies [56]. The conductivity may increase by more than 4 orders of magnitude when the oxygen/Ti stoichiometry ratio varies from 2.0 to 1.7 [57]. The electrical conductivity of a TiO2 film can be greatly improved by a reducing treatment, e.g. thermal anneal under vacuum conditions [58]. Because the as-synthesized porous TiO2 thin film had a relatively high electrical resistivity, Pt electrodeposition on the porous TiO₂ support was barely possible. To increase the conductivity of the TiO₂ support, the as-prepared porous thin film was thermally annealed in a vacuum of 1x10^-7 torr.

Figure 4-2 shows XRD spectra of the TiO₂ thin film annealed at various temperatures. For

41

the as-prepared TiO₂ thin film (curve a), the two diffraction peaks situated at 25.4^o and 48^o are due to the anatase (101) and (200) lattice planes, respectively. The peak at 55.5^o may result from the overlap of the two adjacent peaks due to the anatase (211) and (105) lattice planes. The three diffraction peaks were weak and broad, indicating poor crystallinity and/or finite size effect. At temperatures below 500^oC, the intensity of the three peaks decreased with the annealing temperature. This was more apparent for the (200) plane, which was nearly undetectable at 400^oC. As the annealing temperature reached 500^oC, the three peaks showed a reverse trend in the intensity variation, i.e. the peak intensity increased with the temperature. For the sample annealed at 600^oC for one hour, the anatase (101) peak became narrower with an obvious increase in the peak height. When the annealing time increased to three hours, the anatase (101) peak grew much shaper and stronger, and the anatase (200) peak became clearly detectable with a narrow width. The XRD analysis suggested that the thermal annealing at 600^oC in vacuum improved the crystallinity of the anatase phase and made the porous TiO2 thin film preferentially (101)

Figure 4-2 X-ray diffraction spectra of the porous TiO₂ thin film annealed at various temperatures for one hour: (a) as-prepared, (b) 300^oC, (c) 400^oC (d) 500^oC (e) 600^oC and (f) 600^oC for three hours.

42

oriented. The better crystallinity of the anatase phase seemed to play an important role in improving the electrical conductivity of the porous TiO2 thin film. Pt nanoparticles could be easily electrodeposited on the TiO₂ thin film annealed at temperatures at 600^oC, and electrochemical measurements could be performed without difficulty. On the contrary, no electrochemical operation could be effectively conducted for samples annealed at temperatures below 500^oC.

Figure 4-3(a) shows the bright-field TEM image of electrodeposited Pt nanoparticles on the Pt/TiO2-3h electrode. Nanoparticles with a dark contrast are the electrodeposited Pt catalyst, as revealed by the lattice fringe of the Pt(111) plane (~0.23 nm) in the high

Figure 4-3 (a) TEM and (b) HRTEM images of Pt nanoparticles on the TiO₂-3h support. The lattice fringes labeled by A and B are due to the anatase (101) and the Pt(111) lattice planes, respectively.