From a mechanistic point of view, the cytochrome P450
enzymes are monooxygenases( 150 ) and therefore they release
one molecule of water in every catalytic turnover. As the oxygen
atom of the formed H 2 O molecule is originating from O 2 reduc-
tion, it has been tempting to exploit the fundamental principles
of microscopic reversibility and detailed balancing (157,158),
and to photochemically drive the catalytic reaction pathways in
a reverse direction, thus resulting in a novel type of bioinspired
water oxidation process. In fact, it could be shown that the biomi-
metic multielectron transfer photocatalysts described above are
able to induce an OO bond formation at the metal-oxo site with
water as the donor substrate ( 111 ). In this context, it is impor-
tant to keep in mind that a quite similar mechanism at a single
metal-oxo site has been proposed for the first step in photosyn-
thetic water oxidation (5,148), which among other possible
pathways becomes more and more plausible with the advent of
the crystal structure of oxygen-evolving photosystem II at an
atomic resolution ( 149 ).
B. EMERGINGTHEORETICALFRAMEWORKS
Once the required functional features for a certain substrate
transformation have been defined, individual building blocks
such as the structural motifs presented in the previous section
have to be combined in a synergistic manner to create an
operating photocatalytic system. The choice of suitable metal
centers according to their relative atomic energy levels ( 159 ),
and the fine-adjustment of the different components organized
in the coordination sphere of the active site is a very crucial stage
of development in biomimetic photochemistry. As already
discussed before, a rough guideline for potentially useful com-
binations of the desired fragments can be derived from electro-
chemical and spectroscopic data, which may serve to predict
the predominant orbital parentage and to estimate the relative
energetic ordering of the lowest excited states (Fig. 6). When
the first experiments at this stage of development have led to
promising results, further optimization of such systems can be
guided by an arsenal of emerging theoretical models and
unifying concepts.
Interestingly, there is a current renaissance of classical bond-
theoretical models which are able to illustrate the crucial molec-
ular features of complex systems qualitatively by an interpreta-
tion of more or less localized fragment orbitals. Straightforward
arguments based, for example, on fundamental aspects of ligand
PHOTOSENSITIZATION AND PHOTOCATALYSIS 267