Chemistry Reference
In-Depth Information
atom of the nicotinamide group should not be able to bind to the Mo, and therefore this
group was replaced by a nitrogen-free benzamide group (
L43d
). No signifi cant change
in the yield and selectivity was observed, which supports this hypothesis (entries 4 and
5). In contrast, replacing both picolinamides by benzamides (
L43e
) led to a very poor
ligand in terms of rate and selectivity (entry 6).
Steric factors also play a signifi cant role as illustrated with ligand
L43f
, where one of
the picoline rings was replaced by a quinoline (entry 7). The selectivities were satisfac-
tory, but the yield was low. This is in good agreement with the results described by
Moberg et al. (Table 8B.31, entries 7 and 8) [215]. A fourfold slower rate was observed,
when the picolinamide was replaced by a pivalamide (
L43g
), but the yield and selectivi-
ties were high (entry 8). In principle, the cyclohexane ring system can be replaced by an
acyclic structure such as in ligand
L43h
, although the regioselectivity was slightly lower
(entry 9).
Ko
ovský et al. proposed that one chiral center in the backbone might suffi ce to
determine the sense of wrapping of the metal by the ligand, creating a similar chiral
environment. To prove this hypothesis, they introduced ligand
L43i
. Although yield and
selectivity were lower compared with
L43a
or
L43h
, the selectivities were still in the 90s.
Little differences were observed between results obtained with the branched and the
linear substrates (entries 10 and 11).
For the improvement of the effi ciency of the ligand, the phenyl substituent in the
backbone was replaced by a benzyl group (
L43j
) and by an isopropyl group (
L43k
).
Not unexpected, the sterically less demanding benzyl group resulted in a slight drop in
selectivity (entries 12 and 13), while the isopropyl ligand
L43k
gave results comparable
to the Trost ligand
L43a
(entries 14 and 15). These results clearly indicate that
C
2
-
symmetric ligands are not required to obtain good enantioselectivities in this reaction.
č
8B.4.1.2.2. Bisoxazoline Ligands
In 1999, Glorius and Pfaltz introduced another suc-
cessful class of ligands for Mo-catalyzed allylic alkylations. These ligands
L45 - L47
(Fig.
8B.24) contain the same cyclohexylamide backbone as the Trost ligands, but the pyridine
rings were replaced by oxazoline rings, containing additional stereogenic centers in these
heterocyclic units [219] .
These ligands were fi rst tested with the branched and linear phenyl-substituted allylic
substrates to allow a comparison of the results with those obtained with the Trost ligand
L43a
. The bisoxazolines
L45
and
L46b
were found to induce similar levels of enantiose-
lectivity as
L43a
; however, the amount of branched product was lower and the reaction
was slower (Table 8B.33, entries 1-3). The corresponding branched substrate gave sig-
nifi cantly lower ee (entry 4), an observation that was also made with ligand
L43a
(Table
8B.32 , entry 2).
SS
SS
RR
O
O
O
O
O
O
NH
HN
NH
HN
NH
HN
O
O
L46
,
L4 7
S
S
S
S
N
N
N
N
N
N
O
O
O
O
a
R=Ph
b
R=Pr
c
R=
i
-Pr
S
S
R
R
R
R
L45
L46
L47
Figure 8B.24.
Bisoxazolines used as ligands in Mo-catalyzed allylic alkylations.
