The dye-sensitized solar cell (DSC) is expected to be the next-generation device for energy conversion.1 One of the issues in the DSC is the internal resistivity. In conventional DSCs, as shown in Fig. 1(a), porous semiconductors are used in order to increase the number of adsorbed dye molecules on the semiconductor surface. In this case, although sufficient light absorption is available, the porous structure degrades the crystalline quality, suppressing electron mobility, and makes the carrier path in the semiconductor narrow, increasing the internal resistivity of DSCs. To solve this problem, we proposed the waveguide-type sensitized solar cell2based on thin-film semiconductors shown in Fig. 1(b). In this cell, guided lights in the thin films are used to generate photocurrents instead of normally-incident lights. In such “guided light” configuration, the lights pass through a substantial number of dye molecules to enhance light absorption, consequently, enhance photocurrents. A schematic illustration of the proposed integrated sensitized solar cell film is shown in Fig. 2. The waveguide-type sensitized solar cells are embedded in a light beam collecting film consisting of two stacked films; a film with tapered vertical waveguides and a film with planar waveguides having vertical mirrors. Incident lights are collected by the tapered vertical waveguides into the planar waveguides through the vertical mirrors, and are guided to the waveguide-type sensitized solar cells. Following advantages are expected in the integrated sensitized solar cell film. -Reduction of semiconductor material consumption. -Availability of optical circuit functions. -Light-weight and flexible characteristics. In the present work, fabrication and evaluation of the tapered vertical waveguide films, and sensitization for ZnO thin films were investigated. A fabrication process of the tapered vertical waveguide films by the built-in mask method3is shown in Fig. 3. A photopolymer layer on an Al built-in mask is exposed to tilted ultra-violet (UV) beams from different directions to make cured regions having beveled walls. After development, by removing the removable layer, a tapered vertical waveguide film is obtained. In Fig. 4(a), a photograph for a top view of a tapered vertical waveguide film is presented. Light beam collecting efficiency of the tapered vertical waveguide is shown in Fig. 4(b). The efficiency was evaluated by introducing white light of a halogen lamp from the upper side and measuring the power of the light that transmits through the Al built-in mask window. The tapered vertical waveguide with a taper angle of 30othat is placed in front of the Al built-in mask increases the light beam collecting efficiency by 2~3 times, confirming the light beam collecting capability of the tapered vertical waveguides. The efficiency will be improved by optimizing the structure and the fabrication process. For sensitization of ZnO thin films, to date, we have investigated photocurrents generated in vacuum-evaporated ZnO thin films with two-dye stacked structures2 grown on the surface by the liquid-phase molecular layer deposition (LP-MLD)2,4,5, and demonstrated that guided lights propagated in the vacuum-evaporated ZnO thin films enhance the dye sensitization capability by 4~10 times compared to the normally-incident lights. One concern about the vacuum-evaporated ZnO thin films is that the film quality is not high, which limits the solar energy conversion efficiency. To improve the film quality of ZnO, we attempt to apply ZnO thin films deposited by the sputtering to the semiconductor instead of the vacuum-evaporated ZnO. Figs. 5(a) and (b) show two proposed sensitization mechanisms for the sputtered ZnO thin films. In Fig. 5(a), a two-dye stacked structure is grown using LP-MLD by sequentially providing p-type dye (rose bengal: RB) and n-type dye (crystal violet: CV) on sputtered ZnO thin film surfaces. This structure is expected to widen the photocurrent spectra as shown in Fig. 5(c) because electrons excided in the p-type dye are injected into ZnO directly, and electrons excided in the n-type dye are injected into ZnO through the p-type dye. In the model shown in Fig. 5(b), the two-dye stacked structure is replaced with a ZnO/Cr2O3 multi-layer, which is regarded as a multiple quantum well consisting of Cr2O3quantum wells with different widths. This structure is expected to widen the photocurrent spectra, too. Experiments on these attempts are under way. References 1 B. O’Regan and M. Gratzel, Nature 353, 737 (1991). 2 T. Yoshimura, H. Watanabe, and C. Yoshino, J. Electrochem. Soc., 58, 51 (2011). 3 T. Yoshimura, M. Ojima, Y. Arai, and K. Asama, IEEE J. Select. Topics in Quantum Electron., 9, 49 (2003). 4T. Yoshimura, Japanese Patent, Tokukai Hei3-60487 (1991) [in Japanese]. 5 T. Yoshimura, Japanese Patent, Tokukai 2009-60487 (2009) [in Japanese]. Figure 1