We want to develop an artificial photosynthetic water splitting device enabling the separate evolution of H2 and O2 in two different aqueous phases. This approach avoids the evolution of a potentially explosive H2/O2 gas mixture together with the development of a gas separation facility required to capture H2 from the mixture. We are thus attempting to develop a two-electrode system for solar-light water splitting with the anode only subjected for photo-driven water oxidation to uptake electrons and protons, as nature does. Our target device converts the electrons transferred to the cathode directly to H2 without light illumination, as is the case for the Calvin cycle where CO2 is converted into glucose as a dark reaction. An outstanding feature also lies in the high specific surface areas of both electrodes due to the mesoporous nature of the TiO2 films adopted as the electrode materials.