Angewandte
Communications
Chemie
Table 1: Design of SPDO ligands for the asymmetric cross-coupling
via asymmetric oxidative cross-coupling remains a significant
task to be accomplished. Inspired by pioneerꢀs work together
with our successful SPD (spirocyclic pyrrolidine) catalysts in
asymmetric catalysis,[6f,12,13] herein, we present our efforts
toward this challenging cross-coupling of 2-naphthylamine
and 2-naphthol to synthesize the enantioenriched 3,3’-disub-
stituted NOBINs utilizing our novel CuI/SPDO system
(Scheme 1c).
reaction.[a]
Based on the reported literatures,[4d,6b,f] a large enough gap
of the electronic property between two coupling partners is
necessary to achieve this cross-coupling. Considering that the
NH2 group is more electron-rich than the OH group, 2-
naphthylamine with C3-EDG is proposed as the more
electron-rich partner to preferably form the radical species,
and the 2-naphthol with electron-deficient C3-EWG is
expected to be activated to couple with the radical.[6b]
Obviously, the NH2 group of 2-naphthylamine would lead to
some undesired side-reactions and catalyst deactivation.[4d,14]
Therefore, searching for a flexible ligand that more actively
coordinate with CuI ions than the NH2 group and reasonably
designing substrates are essential to efficiently achieve this
cross-coupling.
Inspired by our previous successful cross-coupling of 2-
naphthols, the 2-naphthylamine 1a and 2-naphthol 2a were
thus recommended as model substrates. From the results in
Table 1, we found that an initial test of this cross-coupling
with CuBr/L1, fortunately, could afford the desired NOBIN
3aa after careful examinations of experiment conditions (see
SI), but only with moderate outcome (65% yield, 54% ee)
and a 20% of homo-BINOL formed. Subsequent ligand
modification by varying the substituentꢀs property and
location was performed (Table 1).[15] Firstly, changing the
configuration of SPD-backbone or introducing gem-dimeth-
yls at C5 of oxazoline moiety reversed the absolute config-
uration of 3aa (L1 vs. L2 and L3). Delightedly, further
modification at the C4 of oxazoline part could largely
influence on the reaction outcome (L4–L17). Although the
aliphatic substituents would only lead to low ee data (0.5–
31% ee, L4–L6), the aromatic ones bearing mono electron-
withdrawing Cl-atom at ortho-, meta- or para-position of the
phenyl could slightly influence on the enantioselectivity, with
the m-Cl-Ph (L9) yielding the best 63% ee (L7–L9). Thus, we
speculated that the p···p interaction may be contribute to the
good result. The following examinations were focused on the
meta-positionꢀs substituents of phenyl. Notably, both EWG-
and EDG-substituents at C3 or C-3,5 could afford the better
enantioselectivities (60–81% ee, L10–L17), with L17 bearing
strong EWGs (3,5-diCF3) gave the best ee (81%). To reduce
the homo-coupling of 2a as much as possible, then 2a was
added in three batches in 20 h and the good result (86% ee,
75% yield) could be obtained, moreover the homo-BINOL
amount reduced to only < 9%.[6d,f]
Ligand
ee (Yield) [%][b]
Ligand
ee (Yield) [%][b]
L1
L2
L3
L4
L5
L6
L7
L8
L9
54 (65)
À27 (63)
À10 (56)
31 (55)
10 (58)
0.5 (52)
50 (55)
59 (62)
63 (64)
L10
L11
L12
L13
L14
L15
L16
L17
L17[c]
60 (61)
64 (55)
67 (65)
66 (65)
67 (64)
71 (67)
65 (61)
81 (65)
86 (75)
[a] The reactions were conducted using 1a (2.0 equiv), 2a (0.1 mmol,
1.0 equiv), CuBr (0.05 equiv), L1–L17 (0.06 equiv) in THF (1.0 mL) at r.t.
under air. [b] Yield of isolated product. [c] CuBr (0.1 equiv), L17
(0.12 equiv) were used.
and location of the substituent at phenyl could affect the
reaction efficiency. For the weak EWG or EDG substituent at
meta-position, the reactions gave the good enantioselectiv-
ities (84–85% ee) and yields (70–76%) (3ab–3ad). While the
strong EWG (CF3) substituent lead to quite different out-
comes (3ae–3ag), the meta-substituent yielded the pretty
good result (3ae, 91% ee, 73% yield), and reversely the
ortho-substituent gave the poor result (3af, 65% ee, 58%
yield) because of the steric hindrance. In the second group of
examples (3ah–3ap), investigation was focused on the 3,5-
disubstituents. These moieties gave the corresponding prod-
ucts in good results (90% ee, 70% yield). Interestingly, the
hetero disubstituents with EWG (CF3) and EDG (OMe)
(3ah) gave high ee value (91%) as well as good yield (74%),
while the homo di-EDG (3al) gave the better yield (75%)
and di-EWG (3an–3ap) switched to the better ee values
(91%). The absolute configuration of cross-coupling products
was assigned as (S) through the X-ray analysis of 3aj. For the
aliphatic methyl ester (3aq), though a medium 70% ee was
observed initially in THF, delightedly it could be promoted to
96% after the solvent was replaced with MeCN. However,
when other functional groups 3-Br and 3-PhCO substituents
were introduced at C3 position, the poor results were
obtained. The substrate containing -COPh at 3-substituted
position led to moderate outcome (3ar, 49% ee, 62% yield),
while the substrate 3-Bromo-2-naphthol afforded poor result
With the optimal conditions in hand, the substrates 2 was
firstly explored by varying the substituents at C3, C6 and C7.
The results are listed in Scheme 2 and showed that most
examples worked well to generate the NOBINs 3aa–3ax with
good to high enantioselectivities (up to 96% ee), moderate to
good yields (up to 80%).[16] In the first group of examples
(3ab–3ag) with C3-phenyl esters, both electronic property
ꢀ 2020 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 7061 –7065