Titania-based solar cells were fabricated by the use of six pheophorbide sensitizers having bacteriochlorin,
chlorin, and porphyrin macrocycles, to which the carboxyl or allylcarboxyl group is directly attached for
binding and electron injection to TiO2. Because of structural similarity, these sensitizers are expected to have
similar physical properties except for electronic-absorption and redox properties: Concerning the former, the
state energies, molar extinction coefficients (ε), oscillator strengths ( f ), and transition-dipole moments (μ)
of the Soret and Q
y
absorptions were determined, whereas concerning the latter, one electron-oxidation potential
(E
ox). Alternatively, the performance of pheophorbide-sensitized solar cells, including the incident photon-to-current conversion efficiency (IPCE), short-circuit current density (J
sc), open-circuit voltage (V
oc), and
solar energy-to-electricity conversion efficiency (η) was determined. It was found that the J
sc and η values
increased with the increasing Q
y
absorption and with the decreasing one electron-oxidation potential (in other
words, with the increasing electron-ejection potential). To explain the clear and strong dependence, we built
possible models for the pheophorbide-to-TiO2 electron injection and tried to fit the J
sc value in terms of the
Q
y
absorption and the E
ox value by the use of empirical equations. After a number of fitting trials, two successful
models of reasonable fitting emerged: One, a model of parallel electron injection, that is, electron injection
upon Q
y
excitation and redox electron transfer in the ground state, and the other, a model of electron injection
simply via the excited state, in which both the Q
y
absorption and the Q
y
-state one electron-oxidation potential
can contribute. These models as well as the future strategies in revealing the real mechanism of electron
injection are discussed.