Chemistry Reference
In-Depth Information
diethylzinc, leading to non-enantioselective alkyla-
tion. Control over these factors can be achieved and
the final catalyst was shown to be equally selective
to the homogeneous system.
Hutchings
et al
. [159,160] have published details
of an Mn-MCM-41-based catalyst that displays
good selectivity in the epoxidation of stilbenes. Their
catalyst is prepared using a different approach to that
described above. They prepare Al-MCM-41 and
exchange the cationic charge-balancing cations in
the material with Mn(II) ions, followed by calcina-
tion to give Mn-exchanged MCM-41. This material
then is interacted with a chiral salen ligand, the Mn
is oxidised to Mn(III) and the resultant complex is
used as a catalyst for the epoxidation of stilbene with
iodosylbenzene (Fig. 7.38). Although the homoge-
neous equivalent of the catalyst gives predominantly
the
trans
-epoxide (in an enantiomeric excess of
78%), the immobilised catalyst gives slightly more of
the
cis
-isomer, with the enantiomeric excess of the
trans
-isomer being almost that of the homogeneous
system (70%). The
cis
-product is obviously a
meso
-
compound. Selective silanisation of different sites
within the catalyst indicated that sites on the exter-
nal (out-of-pore) sites led to essentially identical
behaviour to the homogeneous catalyst, whereas
the internal sites gave substantially more
cis
-epoxide
product. Interestingly, the iodobenzene co-product
was found to be a poison of the heterogeneous cat-
alyst but not of the homogeneous version.
A similar type of complex has been prepared by
Kim & Shin [60]. They used a different approach to
the immobilisation of the catalyst, which involved a
stepwise construction of the ligand system on the
surface followed by complexation of the metal (Fig.
7.39). Reaction with styrene and methylstyrene gave
epoxides in good conversions and with enantiomeric
excesses typically slightly superior to those found in
solution with equivalent complexes.
A very recent example of a supported ferrocenyl-
Pd complex illustrates the enormous potential
of this class of materials compared with both homo-
geneous and amorphous heterogeneous materials.
A collaborative project between the groups led by
Thomas and Johnson [59] has provided a chiral fer-
rocenylphosphine anchored onto the walls of MCM-
41. The external surface of the material first was
passivated with a bulky silane to ensure that all the
catalytic sites were situated within the material's
pores. The chiral ferrocenylphosphine then was
attached to the walls of the support by reaction of a
pendant amine with a surface-bound bromoalkane.
This surface-bound ligand system then was treated
with a Pd complex to generate the active catalyst.
The reaction studied was the allylic substitution of
cinnamyl acetate with benzylamine, which can lead
to a linear amine (achiral) or to allylic substitution
(a chiral product) (Fig. 7.40).
The dependence of the reaction path on the nature
of the catalyst is remarkable. With a homogeneous
palladium catalyst the only product is the thermo-
dynamically favoured linear substitution product.
With an amorphous silica support there is a little
(2%) of the allylic substitution product, with a 43%
enantiomeric excess reported. Astonishingly, the
allylic product is favoured (51% selectivity) over the
linear with the MCM-41-based catalyst, which is an
enormous change in regioselectivity. Furthermore,
enantioselectivity with the (
S
)-isomer of the catalyst
is >99% but with the (
R
)-isomer it is 93%. This is a
H
H
NN
+
Mn-Al-MCM-41
catalyst
OH HO
PhlO
O
Ph
cis and trans
Fig. 7.3
8
Preparation and use of an
Mn-exchanged Al-MCM-41 catalyst in
epoxidations.
Ph
Ph
Ph
