Development of Solar Cells Based on Synthetic Near-Infrared Absorbing Purpurins: Observation of Multiple Electron Injection Pathways at Cyclic Tetrapyrrole–Semiconductor Interface

Purpurin sensitizers with and without the central zinc, ZnP and H₂P, have been synthesized and used in dye-sensitized solar cells. Both the sensitizers readily formed aggregates on the semiconductor surface. The DFT and TD-DFT calculations suggest that the major difference between the two sensitizers is ascribable to the energy levels of their four molecular orbitals. With biased potential in the solid-state photovoltaic diodes, the photoresponse of ZnP and H₂P started from −2 and −2.5 V, respectively, and the observed difference is in agreement of the difference of calculated LUMO energy level for the two sensitizers. ZnP gave much better photovoltaic performance than H₂P, when TiO₂ electrode and 4-tert-butylpyridine (TBP)-free electrolyte were employed. The decrease of photocurrent of ZnP-based solar cell in TBP-containing electrolyte is attributed to the change in energy level of the electron acceptor, while that of H₂P-based solar cell in TBP-containing electrolyte is ascribed to the change of electron donor state. The replacement of TiO₂ with SnO₂ substantially improved the photocurrent of solar cells because the electron injection from LUMO orbital of the dye sensitizers becomes favorable. A clear observation of photocurrent generation from the dye aggregate suggests that the photon-generating excitons can diffuse over the dye aggregate and finally reach the semiconductor surface. TBP in electrolyte can disturb the dye aggregation, and this will reduce the possibility of exciton annihilation in dye layer, which was supported by the sub-picosecond time-resolved absorption spectra.