The dye solar cell, is a novel photoelectrochemical solar cell presenting unique advantages, as the use of low cost materials and the potential simplicity of manufacturing. However, a liquid electrolyte is actually required for the transport of the photogenerated positive charges. A major breakthrough is the replacing of the liquid electrolyte by a solid-state hole conducting material – the aim of this work. First, the metal/TiO2 junction was investigated, with either silver or gold as rectifying contact. Sensitization was demonstrated, as action spectra were measured using different dyes. The evaporated metal could not enter the pores of the nanocrystalline TiO2 to contact all the dye molecules, hence only low photovoltaic performances were obtained. Furthermore, these metal/TiO2 junctions showed a high sensitivity to humidity and to forward bias polarization. The use of an electrically conducting polymer, poly(bithiophene), was studied in a metal/polymer junction and in TiO2/polymer devices with flat and nanocrystalline TiO2 electrodes. The photoelectric characterizations indicated that electrochemically reduced poly(bithiophene) was able to inject electrons into TiO2, i.e. sensitizing TiO2, and it acted as a hole conducting layer. Short-circuit current densities of 2-3 mA/cm2 were measured at an irradiation of 1.2 W/cm2. The open-circuit voltages were in the 250-350 mV range, depending on the light intensity and temperature. The use of a transparent hole conducting material, CuI, which is a wide bandgap p-type semiconductor, was demonstrated in a nanocrystalline junction made of highly porous sensitized TiO2 and solution-cast CuI. A short-circuit current of 1 mA/cm2 was achieved under one sun simulated light; the open-circuit voltage was 600 mV and the light-to-power conversion efficiency was 0.3 %. The peak quantum efficiencies of 8 % in the UV, and 3 % in the visible, showed that the nanocrystalline junction was working inefficiently, probably due to the poor electrical contact between the TiO2 nanocrystallites and the CuI crystals. Furthermore, the sensitizer was not tuned to inject holes into the valence band of CuI. As conclusion, a controlled-capillary pore based sensitized device is suggested. In this concept, all the internal surface of the TiO2 electrode coated with sensitizer is contacted with the transparent hole conducting layer, hence improving the photovoltaic properties of the device.