(Eq. 28) they combine to the well-known binuclear hydroxy-
bridged complex TpCuII(OH) 2 CuIITp ( 42 ). Generally, Cu(II)
complexes have available low-energy LF and LMCT states.
While Cu(II) LF states are not reactive, LMCT states initiate
the reduction of Cu(II) to Cu(I) and oxidation of ligands (48,49).
As a reductive elimination of H 2 O 2 , Eq. (6) finds a logic explana-
tion in this way. Reductive elimination of two hydroxide ligands
yielding H 2 O 2 has been reported previously (50,51). Another
observation related to our work is also quite important. The pho-
tolysis of TpCuIIOCuIITp in the presence of dimethyl sulfide or
cyclohexene is accompanied by oxygen transfer to these organic
substrates ( 52 ).
In summary, the photocatalysisEq. (30) or (31) can be consid-
ered as proof of principle for photocatalytic splitting of CO 32
or CO 2 itself (Eq. 3 or 4). However, in order to improve this pho-
tocatalysis, a few shortcomings of the present system should be
mentioned. Of course, alkyl carbonates or amides (urea) must
be replaced by CO 32 . Since carbonate complexes of Cu(I) are
known ( 36 ), the instability of CuTp/CO 32 mixtures is probably
associated with the Tp ligand. Accordingly, Tp* should be rep-
laced by another ligand that forms also stable Cu(I) and Cu(II)
complexes but does not facilitate a decomposition in the presence
of carbonate and the photooxidation of the spectator ligand as it
takes place in the case of copper phosphines (see above).
Finally, the present system needs UV light, but this is certainly
not an inherent energy requirement. An intermolecular or
intramolecular long-wavelength sensitization might circumvent
this problem.
IV. Dinitrogen SplittingThe lack of reactivity of dinitrogen which complicates its chem-
ical conversion has been a challenge to chemists for many
decades (53,54). This difficulty is based on the extreme stability
of the nitrogen–nitrogen triple bond. The huge energy difference
between HOMO and LUMO (23 eV) makes N 2 rather redox inert.
Moreover, the conversion of N 2 to simple species, such as ammo-
nia or nitride, requires the transfer of six electrons. Such multi-
electron transfer processes are generally associated with large
activation barriers. Nevertheless, the reduction of N 2 to NH 3
occurs in nature through the utilization of the enzyme nitroge-
nase as catalyst. This conversion also takes place in the
Haber–Bosch process; however, extreme conditions are required.
However, this catalysis occurs only under extreme conditions.
360 ARND VOGLER AND HORST KUNKELY