Reverse electron transfer from TiO2 to I2 in nanocrystalline TiO2 film electrodes with coadsorbed bipyridine and biquinoline ruthenium complexes

Reverse electron transfer from TiO2 to I-2 in nanocrystalline TiO2 films with coadsorbed [NBu4](2)[cis-Ru(Hdcbiq)(2)(NCS)(2)] (1; [NBu4](+)) tetrabutylammonium cation; H(2)dcbiq) 4,4'-dicarboxy-2,2'-biquinoline) and [NBu4](2)[cis-Ru(Hdcbpy)(2)(NCS)(2)] (2; H(2)dcbpy) 4,4'-dicarboxy-2,2'-bipyridine) was investigated. The apparent electron lifetime decreased when 1 was coadsorbed with 2 on TiO2. The incident monochromatic photon-to-current conversion efficiency and the open-circuit photovoltage decreased substantially even when a small amount of 1 was coadsorbed with 2 on TiO2. The reduction of 1 on TiO2 was spectroelectrochemically observed at a potential that was more positive than the energy of the conduction band edge of the TiO2 films. These results suggest that the injected electron from photoexcited complex 2 reacted with I-2 through 1. Reverse electron transfer from TiO2 to I-2 depended not only on the electron density but also on the amount of 1 adsorbed on TiO2. Therefore, we assumed that the injected electron was trapped by 1 on TiO2 and then reacted with I-2. If an injected electron diffused randomly in TiO2, the relative cross-section (sigma(trap)) of electron trapping by 1 on TiO2 would be expected to be 1 and to be the same as that of trapping by 2. However, the sigma(trap) value was much larger than expected, which indicates that the electron in TiO2 predominantly reacted with I-2 through 1.