Chemistry Reference
In-Depth Information
and does not catalyse epoxidation. Hydrolytic stabil-
ity therefore is enhanced by high negative charges
and lower-valent heteroatoms or substituents. Triva-
lent heteroatoms include B and Al, but the divalent
Zn and Co are very much more robust. As usual
there is a cost—higher negative charge inhibits cat-
alysis of electrophilic oxidations.
Since the early work, catalytic activity has been
shown for a variety of polyoxometallates, at least
some of which appear to act in the undissociated
form. Several substituted 'lacunary' Keggin com-
plexes have been studied [112], although the activ-
ity in epoxidations usually has been low for terminal
olefins. The 'sandwich' type has shown the greatest
promise for epoxidations in the form of the two
complexes [WZnMn
II
2
(ZnW
9
O
34
)
2
]
12-
[113] and
[(WZnRh
III
2
)(ZnW
9
O
34
)
2
]
10-
[114], the latter offering
lower H
2
O
2
decomposition and greater stability.
Keggin complexes with a single transition metal
substituent can give both radical and electrophilic
reactions, depending on the substituent, but
reported activities are not of great interest industri-
ally. A W-peroxo derivative of an intact singly sub-
stituted Keggin structure has been discovered now:
b
3
-[Co
II
O
4
)W
11
O
31
(O
2
)
4
]
10-
[115]. This appears, from
cyclohexenol oxidation results, to be a relatively
nucleophilic oxidant, as expected. Hydroxylations
of alkanes are catalysed by the substituted Keggin
structure [g-SiW
10
{Fe(OH)
2
}
2
O
38
]
6-
[116] and it is
established by nuclear magnetic resonance (NMR)
that the 1,2-Fe topomer (with vicinal Fe atoms) is
the main active component. Similar results are
reported for oxygen oxidations with H
5
PV
2
Mo
10
O
40
,
with the 1,2-V topomer being the best catalyst of
phenol and alcohol oxidation, among others [117].
Both of these catalysts seem to involve cooperation
between two one-electron oxidising species, as may
others with this feature [118]. Mixed Mo-V com-
plexes up to PV
6
Mo
6
catalyse phenol hydroxylation
by H
2
O
2
[119], and the dependence of
o
-/
p
- ratio
on V/Mo ratio may well be related to topomer inter-
actions. This phenomenon will be revisited in
Section 4.
An important development in the practical use
of polyoxometallates, aimed thus far at paper
pulp bleaching, is successful self-assembly—includ-
ing self-repair and self-reassembly after reaction,
even if dissociation occurs at an intermediate stage
[120]. This makes use of the thermodynamic stabil-
ity of the complexes under given conditions once an
effective catalytic structure can be matched to those
conditions. A lot of laborious research is needed to
achieve this match, but it is ultimately one of the
most valuable properties of polyoxometallates and
should ensure their adoption for many catalytic
processes in future. This same feature suggests
enormous potential in immobilised systems, as yet
largely untapped. A limitation of polyoxometallates
is their high equivalent weight as oxidising interme-
diates. For this reason, true catalytic cycles, rather
than stoichiometric generation/use/regeneration
loops, remain a key target.
3.4 Zeolitic and smectitic materials
This section addresses heterogeneous catalysts with
no homogeneous analogue, as distinct from immo-
bilised homogeneous catalysts. Some excellent criti-
cal reviews covering one or both areas have been
published recently [121]. For the most part, the cat-
alysts are based on peroxo-metal chemistry.
Titanium and other silicalites
The titanium-substituted aluminium-free silicalite
TS-1 with 5.5 Å channels (MFI structure, analogous
to ZSM-5), was found to catalyse many H
2
O
2
oxida-
tions [52-55,122]. After the first reports of tita-
nium silicalite in the early 1980s there was a huge
research effort worldwide to find the many analo-
gous materials believed to be waiting to be discov-
ered. However, as time has progressed, TS-1 itself
seems more and more unusual. Hence, this effort has
not abated, but a large amount of it has been applied
to finding out why TS-1 works so well, before being
in a position to make new breakthroughs [123-125].
This in itself has fuelled progress on the characteri-
sation techniques for such materials [126,127].
From the beginning at least three features poten-
tially were important: hydrophobic environment,
tetrahedral geometry and constrained reaction site.
All of these have turned out to be relevant. Work
on solvent dependence and on Ti/Si xerogels with
varying hydrophobicity [128] confirms the impor-
tance of the active site environment, while not
achieving comparable catalysis by the latter route. In
fact, TS-1 has been commercialised for the oxidation
of phenol to catechol/quinol and for in situ oxida-
tion of ammonia to hydroxylamine in the produc-
tion of caprolactam from cyclohexanone (via the
