2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Li et al. Sci China Chem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
donor-acceptor carbenes [12a], carbene acetates [12b] and
carbene malonates [12c]. Our study began with the screening
of a number of BOX and TOX ligands. As shown in Table 1,
BOX ligands L1–L4 bearing different chiral backbones were
initially examined, and L-tert-Leucine derivatives L4 proved
to be the optimal ligand, giving the desired cyclopropane 3a
in 56% yield with 88/12 dr and 79% ee (entry 5). Based on
our previous study on the complex structures of TOX/nickel
groups swing towards the copper center, which may force the
coordinated parent oxazoline groups and allow a more sui-
table space to accommodate the substrates and release the
products. However, in this reaction, when TOX ligand L5
Figure 1 Useful biologically active compounds bearing 1,1-dimethyl
cyclopropane scaffold (color online).
and SaBOX ligand L6 were employed, both the yields and
enantioselectivity declined (entries 6 and 7).
Based on the previously reported works, the bridge angle
of the two oxazolines in a chiral BOX ligand has significant
effects on the enantioselectivity of a reaction. For example,
Davies et al. [14a] found that the increasing bridge angles of
the chiral BOX ligands led to the increasing enantioselec-
tivity in the In-BOX/Cu(II) catalyzed asymmetric Diels-
Alder reaction. Denmark and coworkers [14b,14c] also de-
monstrated that when bulkier substituents were installed on
the bridge carbon of the chiral BOX ligands, the corre-
sponding bridge angles were smaller, and the enantioselec-
tivities of the asymmetric reactions promoted by these chiral
BOX ligands were increased, for some substrates in the en-
antioselective additions of organolithium to imines. In our
recent studies on the catalytic asymmetric [2+2+2] tandem
cyclization reactions, we utilized a strong Thorpe-Ingold
effect to modify the chiral ligands, which by means of in-
creasing the steric hindrance of the substituents on the bridge
Scheme 1 Trisubstituted olefins in asymmetric 1,1-dimethyl cyclopro-
pantion.
carbon of chiral BOX ligand, the enantioselectivity of the
reaction could be improved [14d]. In this cyclopropanation,
we noticed that, in comparison of L2a with L2b, L2b with a
methods are specific to one or one type of olefins. Moreover,
due to the high sensitivity of metallocarbenes to the steric
hindrance and geometry of the alkenes, for more steric bulky
trisubstituted olefins bearing different aryl groups, the en-
antioselective cyclopropanation has rarely been realized yet.
Accordingly, in consideration of the versatility of 1,1-di-
methyl cyclopropanes in drugs development, a new chiral
catalyst system, that is highly efficient on both reactivity and
enantioselectivity, as well as with abroad substrate scopes, is
still in eager demand. Here, we report our efforts on this
subject.
In a metal complex involved asymmetric catalysis, the
chiral ligands usually proved to be crucial for both the re-
activity and the stereo selectivity. Recently, we have devel-
oped a series of side arm modified chiral bisoxazoline
ligands [11], which were successfully applied in the en-
antioselective cyclopropanation of terminal olefins and 1,2-
disubstituted olefins with different copper carbenes such as
cyclopropyl group led to an obviously drop on both the yield
and ee value (entry 2 vs. entry 3, Table 1). These results
suggested that the reaction was quite sensitive to the steric
demand of R1 and R2, and that drove us to the screening of
ligands with different R1 and R2 groups. With L7 and L8,
containing less hindered R1 and R2 groups, 38% and 50%
yields with 77/23 dr as well as 73% ee and 72% ee were
obtained respectively (entries 8 and 9). Inspiringly, 91/9 dr
and 87% ee were achieved with a diethyl substituted ligand
L9 (entry 10). Further increasing the steric hindrance with
diisopropyl substituents (L10) allowed the current reaction
to give an 89/11 dr with 91% enantioselectivity (entry 11).
However, with diisobutyl substituted ligand L11, the stereo
control of this reaction was destroyed (entry 12). To our
delight, both the yield and enantioselectivity of the desired
cyclopropane could be increased in terms of lowering the
reaction temperature to 0 °C, and gave 76% yield, 95/5 dr
and 95% ee (entry 13), probably due to the lower temperature