Chemistry Reference
In-Depth Information
both Ti-substituted and Ti-grafted MCM-41 materi-
als have been provided by Chen
et al
. [97]. They
investigated the epoxidation of cyclohexene with
hydrogen peroxide and
t
-BuOOH. With hydrogen
peroxide in methanol they found that the products
isolated were due to ring opening of the first-formed
epoxide (i.e. diols and methanol-based ring-opening
products). Even worse, Ti leaching was almost
complete within the first cycle. With
t
-BuOOH
(in hydrocarbon solvent) this behaviour was not
observed, epoxide was formed with excellent selec-
tivity (due, no doubt, to the much less nucleophilic
environment) and Ti leaching was not observed. This
behaviour is consistent with that found by Abben-
huis
et al
. [96] and suggests that hydrogen peroxide
can damage untreated Ti-containing materials.
Whether surface modification with silanes [91,92]
or by incorporation of polarity-modifying groups
[68,94] can enhance the stability of these catalysts
has not yet been clarified.
An interesting extension of this work is the one-
pot synthesis of isopulegol epoxide [98] (Fig. 7.19).
This two-step reaction involves an acid-catalysed
cyclisation of citronellal to isopulegol, followed by
epoxidation of the exocyclic double bond.
Excellent selectivity of >98% was found in the first
step and 90% for the second. An overall conversion
of 76% was achieved. One of the drawbacks of the
system is that the choice of solvent is made very dif-
ficult by the conflicting requirements of the first and
second steps. The optimum process was found to be
cyclisation catalysed by Ti-MCM-41 (prepared by
post-modification of MCM-41 by titanocene dichlo-
ride, followed by calcination [95,99]) in toluene, fol-
lowed by addition of
t
-BuOOH in acetonitrile. This
was found to be the best compromise between the
conflicting solvent requirements for both steps.
Unfortunately, it leads to a mixed solvent system at
the end of the reaction. The authors also demon-
strated that the reaction is heterogeneous and that
no catalytically active species are leached from the
catalyst.
An example of the efficient oxidation of bulky sub-
strates is provided by the oxidation of cholesterol
[100] in two different ways. The use of Ti-contain-
ing materials leads to the efficient epoxidation of
cholesterol, whereas using Zr in place of Ti leads to
the oxidation of a CHOH group to C=O. (Fig. 7.20).
A further related paper describes the use of Ti-
substituted micelle-templated aluminophosphate
materials as catalysts for the epoxidation of alkenes
using hydrogen peroxide in acetonitrile [101].
Another potential advantage of these larger pore
materials that has remained unexamined so far (with
the exception of some examples of the use of
t
-
BuOOH and one of cumene hydroperoxide) is that
the larger pores also will allow access to larger,
potentially chiral peroxides, which may find use in
systems where hydrogen peroxide does not function
effectively.
These materials represent an important step to-
wards extending the chemistry of TS-1 to larger sub-
strates. Of particular importance is the fact that
Ti-containing amorphous silicas have been reported
as being unsuitable for such chemistry, due to low
stability [102].
Pinnavaia demonstrated the utility of Mn com-
plexes supported in Al-MCM-41 as an oxidation cat-
alyst [103]. They exploited the charged framework
of the material to immobilise an Mn(bipy)
2
dication
on the walls, and showed that the resultant mater-
ial was an efficient oxidation catalyst for the oxi-
dation of styrene to epoxide, diol and further to
benzaldehyde, using various oxidants. The enhanced
Fig. 7.19
Conversion of citronellal into
isopulegol using Ti-MCM-41.
