Journal of the American Chemical Society
Article
the dihedral angle in the chiral backbone decreases [(R)-
BINAP (73.49°), (R)-MeO-BIPHEP (68.56°), and (R)-
SegPhos (64.99°)],19 the diastereoselectivity increases but
the yield drops by more than half. Although it is difficult to rule
out the electronic effects of the ligands, the dihedral angle of
the axially chiral ligands seems to play a crucial role in
reactivity and diastereoselectivity. Higher diastereoselectivity
was obtained with the chiral ligand possessing the smaller
dihedral angle, which could exert much better asymmetric
induction. Although it is difficult to understand the low
reactivity of this reaction with (R)-SegPhos L3, the underlying
reason might be the three-dimensional space incompatibility
between the chiral Pd complex and the chiral Cu complex.
Then, the ligand with higher steric hindrance (R)-DTBM-
SegPhos L4 was applied in this reaction. Interestingly, the
complementary diastereomer (R,S)-3aa′ was obtained in 48%
yield and 1:6 dr and >99% ee (entry 4). This result suggested
that the bulky steric environment of the (R)-DTBM-SegPhos
L4 could overwrite the stereocontrol ability of the chiral Cu
catalyst and directly control both stereocenters of the allenyl
esters and the nucleophiles. Subsequently, isopropyl-substi-
tuted Phosferrox (R,Rp)-L6 instead of tert-butyl-substituted
Phosferrox (R,Rp)-L5 was subjected to this reaction (entry 5 vs
entry 3). The yield was obviously improved with comparable
enantioselectivity and diastereoselectivity. Furthermore, the
diastereoselectivity increased rapidly when the reaction
temperature was decreased to −10 °C (entry 6). To our
generality of enantio- and diastereodivergency was demon-
strated by the synthesis of all four stereoisomers of 3aa using
all four combinations of SegPhos L3 and Phosferrox L6.
Gratifyingly, almost perfect stereoselectivities (>20:1 dr and
>99% ee) with high reactivity were achieved using the same set
of starting materials and under identical reaction conditions,
which verified the efficiency of the stereodivergent bimetallic
catalysis for the stereocontrol of the 1,3-nonadjacent stereo-
centers.
Having established the optimized reaction conditions, we
explored the specific performance of our bimetallic catalytic
system in the DyKAT of racemic allenylic acetates in a
diastereodivergent manner (Table 2). For most of the
substrates, there is no obvious difference in catalytic activity
and stereoinduction of the two catalyst combinations. First, for
the substrates bearing the synthetically attractive tertiary amine
and alcohol groups (1b−1h), the substitution reaction
proceeded smoothly to afford the desired products (3ba-
3ha) with 1,3-nonadjacent axial and central chirality in high
yields, and with excellent diastereoselectivities (10:1 to >20:1
dr) and enantioselectivities (96% to >99% ee). In general,
sterically bulky substituent of 1,3-disubstituted allenes together
with the elaborate sterically hindered ligand DTBM-SegPhos
are indispensable for achieving high enantioselectivity due to
the loosening nature of the allene axial chirality spreading over
three carbon atoms.6d,e,g Accordingly, we were very curious
whether our bimetallic catalytic system is suitable for substrates
substituted with much less sterically bulky group. To our
delight, the allenylic esters bearing a linear 1°-alkyl group (1i−
1n) reacted with imine ester to afford products (3ia−3na)
with high diastereoselectivities (up to 9:1 dr) and excellent
enantioselectivities (up to >99% ee). Furthermore, the
substrates with a secondary alkyl group (1o) afforded the
desired products (R,S)-3oa and (S,S)-3oa with moderate
diastereoselectivities (12:1 dr and 4:1 dr, respectively) and
complete enantiocontrol. To achieve satisfactory diastereose-
lectivity of the products (R,S)-3ia−3oa, Pd/(R)-DTBM-
SegPhos and Cu/(S,Sp)-L6 were employed. However, the
three-dimensional space incompatibility between the Pd/(R)-
DTBM-SegPhos and the Cu/(S,Sp)-L6 led to relatively low
yields. Additionally, 1p bearing a planar phenyl group is also
compatible with this Pd/Cu dual catalytic system, giving the
corresponding products (R,S)-3pa and (S,S)-3pa with
satisfactory results (82% yield, >20:1 dr, >99% ee and 50%
yield, 9:1 dr, >99% ee).
Next, the scope of the nucleophile was investigated in a
diastereodivergent manner to explore the generality and
limitations of our dual catalytic system. As described in
Table 3, the influence of the substituents on the aromatic ring
was initially investigated, and it was apparent that the position
and electronics of the substituted groups exerted a limited
effect, affording the desired products in high yields with
excellent stereoselectivities (up to >20:1 dr and >99% ee). For
example, both electron-donating and electron-withdrawing
groups at the para position of the phenyl ring were well
tolerated and delivered the corresponding products 3ab−3ae
in high yields, with 12:1 to >20:1 dr and >99% ee. As shown
by examples 3af−3aj, substituents at the meta and ortho
positions were also compatible as well. Furthermore, reactions
involving naphthyl-, piperonyl, and furyl-substituted substrates
also proceeded well to deliver their desired products (3ak−
3am) in high yields and with excellent stereoselectivities. The
t
delight, the imino esters substituted with Et, Bn, and Bu are
also compatible with this dual catalysis system, providing the
corresponding products in high yields and with high to
excellent diastereoselectivities and enantioselectivities (entries
7−9, for 2a, 80% yield, >20:1 dr, >99% ee). Subsequently,
control experiments were conducted to gain insight into the
cooperative interplay of the bimetallic catalytic system (entries
10−12). The substituted product was obtained in only 14%
yield in the absence of the Cu catalyst and the dr decreased
rapidly from >20:1 to 1:2. No reaction occurred in the absence
of the Pd catalyst. These results suggested that the combined
use of two chiral metal catalysts appears to be important for
the reactivity and stereochemical control of this reaction.
After establishing the feasibility of an enantio- and
diastereoselective synthesis of 1,3-nonadjacent axial and central
stereocenters, we then speculated if the present Pd/Cu
catalytic system could furnish products with full control over
the absolute and relative configuration of both axial and central
chirality (Scheme 3). Indeed, by simply changing (R,Rp)-L6,
the ligand for Cu, to (S,Sp)-L6 while keeping the same
enantiomer of (R)-SegPhos L3, (R,S)-3aa was formed
successfully in high yield with excellent diastereoselectivity
(>20:1 dr) and enantioselectivity (>99% ee). Next, the
Scheme 3. Synthesis of All Four Stereoisomers of 3aa
C
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX