Chemistry Reference
In-Depth Information
after a single batch reaction potentially economic for
fine chemicals. There is always, however, some com-
petition between substrate oxidation and catalyst
degradation, which is more serious for less-reactive
substrates and masked in much literature work by
the use of large excesses of substrate to H
2
O
2
. In very
electron-deficient Mn porphyrins (such as poly-
b-nitro [70] or poly-b-chloro [71]), the role of the
axial ligand is diminished and it may be that the
M-O-O-H species acts directly on the substrate, as
has been shown for similar iron systems. The
catalytic potential of such systems remains to be
established.
For iron porphyrins, imidazole is not very useful
in homogeneous systems because the inactive
trans
-
dicoordinated complex is readily formed. Accord-
ingly, homolytic O-O cleavage is more of a factor.
Moreover, it appears that when oxygen transfer from
the oxene does occur, it is in two discrete one-
electron steps, as shown, for example, by isomerisa-
tion of
cis
-stilbene to the
trans
-epoxide, and by the
greater tendency to hydroxylate aryl-substituted
olefins along with the epoxidation. This is likely to
be due to the absence of Fe(V) in the resonance
equilibria:
R
R
N
N
N
R
Fig. 11.10
Tri-
N
-alkyl triazacyclononane.
major factor is lack of control over the relative
surface affinity for reactants (and products). In Mn
systems, co-catalyst access also often is needed—
tethering of the co-factor is not a promising approach
owing to its own degradation and to the dual role
of axial ligand and proton donor required (although
these functions can be performed by different
molecules).
Recent work with phthalocyanines [76] has
revealed new catalytic potential, a key advantage
being their cheap and simple preparation by incor-
porating transition metals from all three rows of the
periodic table. They can be prepared readily in zeolite
cavities and on other supports [77]. Indeed, com-
plexes thus encapsulated can have a longer lifetime
with H
2
O
2
than they do in solution: e.g. Fe phthalo-
cyanine in zeolite Y, which catalyses cyclohexene
epoxidation [78]. They appear on the whole,
however, to be more suited to C-H oxidation reac-
tions than to epoxidation.
Another relative newcomer as a ligand is sym-
triazacyclononane (tacn) and its
N
,
N
¢,
N
≤-trimethyl
analogue (tmtacn) (see Fig. 11.10). These form stable
complexes with all first-row transition metals, and
the manganese complexes in particular offer elec-
trophilic oxygen-transfer catalysis with H
2
O
2
[79]
and radical hydroxylation of saturated hydrocarbons
in acetic acid solution [80]. Once again, ligand degra-
dation is not negligible over long periods, and leads
to a loss of selectivity as Mn is liberated in different
forms. The consequences of this for catalysis of stain
bleaching in domestic laundry are well known. Salen
(Schiff-base type) ligands are another class that have
not achieved much success with H
2
O
2
owing to their
easy oxidisability. By contrast, hypervalent metal
complexes with tetragonal amide ligands can be
p
+
·Fe(IV) = O ´ pFe(IV)—O· ´ p
+
· Fe(III) - O·
(cf: pMn(V) = O ´ p
+
·Mn(IV) = O ´ pMn(IV) - O·)
allowing significant radical character arising from the
single-bonded species. Recent mechanistic work also
has shown, however, that O-O cleavage is not nec-
essary to obtain a catalytic system—the Fe-O-O-H
intermediate can transfer oxygen directly to sub-
strates [72]. Where the oxene is formed, iron por-
phyrins have been shown to exhibit 'oxo-hydroxo
tautomerism' in the presence of water, which is a
favoured axial ligand. All in all, iron porphyrins may
have the same catalytic potential as manganese, with
less dependence on co-catalysts, but fewer studies
on relevant substrates have been reported. Water-
soluble Mn [73] and Fe [72a,b,74] porphyrins have
been prepared, mainly by substitution of
meso
-
phenyl substituents with charged groups such as
quaternary ammonium or sulfonate.
Attempts to immobilise porphyrins on solid sup-
ports [75], whether by physical or chemical att-
achment, have met with limited success to date.
Although site isolation should improve catalyst life
by preventing face to edge contact, this is counter-
balanced by the loss of activity usually seen. A likely
