Angewandte
Communications
Chemie
À
substrates for transition-metal-catalyzed N C cross-coupling
in good yields with moderate to good ee values (entries 6–10),
except ligand L7, which resulted in the yield decreasing
sharply (entry 11). Moreover, monophosphine L8 was also
effective, giving 4a in 70% yield and 63% ee (entry 12).
Furthermore, the catalyst loading could be lowered to
2.5 mol% to allow completion of the reaction with only
a slight decrease in the yield and ee value (entry 13).
reactions not only because they coordinate strongly to the
transition metal and cause catalyst deactivation but also
because they undergo tautomerization and E/Z isomerization
(6 possible isomers).[18] These issues make enantioselective
control more complex and challenging.
We began optimizing the reaction conditions using
amidine 3a, which was easily prepared by a two-step reaction,
as the model substrate (Table 1). Condensation of 2-bromoa-
niline (1a) with TFA in CCl4 in the presence of triphenyl-
phosphine afforded a trifluoroacetimidoyl chloride. Subse-
quent addition/elimination with substituted aniline 2a pro-
duced amidine 3a in a good yield. The reaction of 3a in the
presence of Pd(OAc)2 (5 mol%), (S)-BINAP L1 (7.5 mol%),
and K2CO3 (1.2 equiv) in toluene at 608C afforded the desired
product 4a in 82% yield and 75% ee (entry 1). Slightly lower
enantioselectivity was obtained when the Br substituent in the
substrate was replaced by I (entry 2). To our delight, when
KOH was used instead of K2CO3, 4a was formed in 93% yield
and 89% ee (entry 3). Further screening revealed that the
best result was obtained when Cs2CO3 was used, with 4a
obtained in 98% yield and 92% ee (entry 4). A comparable ee
was obtained when NaOH was used as the base, although the
yield was slightly lower (entry 5). Next, the effect of different
ligands was tested. Most of the biphosphine ligands resulted
With the optimal conditions in hand, we next examined
the substrate scope of the amidine (Scheme 2). A wide range
of amidines were suitable, affording the desired biaryl
atropisomers in good to excellent yields and enantioselectiv-
ities. First, the effect of substituents at the C1-position of the
amidine moiety was examined. Slightly lower enantioselec-
tivity was obtained when CF3 was replaced by CF2Cl (4b) or
C2F5 (4c) on the substrate. However, when alkyl- and aryl-
substituted amidines were used under identical reaction
conditions, the enantioselectivities were reduced significantly
(4d, 4e), thus demonstrating the unique fluorine effect of the
CF3-bearing substrate. The racemization barriers (DG°) of 4a
and 4d were measured to be 34.3 and 31.5 kcalmolÀ1,
respectively (see the Supporting Information for details),
which may indicate that partial thermal racemization of the
product can be excluded as the reason for the decrease in the
ee value with 4d. To investigate the effect of substituent R on
the phenyl ring of the 2-bromophenyl moiety, we carried out
the reaction of 3 with a series of amidines with meta and/or
para substituents on the 2-bromophenyl moiety. The sub-
stituents included Me (4 f, 4n), t-Bu (4k), MeO (4r), halogen
(4g-4i, 4p, 4q), CF3 (4j, 4o), OCF3 (4l, 4s), CN (4m), and
CO2Me (4t). The reactions smoothly furnished the coupling
products with good to excellent yields (63–98%) and
enantioselectivities (86–93%). The absolute configuration
of 4r (> 99% ee after recrystallization) was determined by X-
ray crystallographic analysis.[19] In general, the ee values for
substrates bearing electron-withdrawing substituents were
higher than those for substrates bearing electron-donating
substituents (4 f vs. 4j and 4o vs. 4r). Furthermore, ortho-F-2-
bromophenylamidine also underwent transformation to form
biaryl atropisomer 4u in good yield and enantioselectivity. As
expected, disubstituted substrates underwent the title reac-
tion to give 4v–4x with excellent ee values. The effect of
substituents (R1) on the aniline moiety was also examined.
The ortho-alkyl group in the aniline moiety could be a simple
methyl or ethyl group, although the yields slightly decreased
when the MeO group was not introduced at the 4-position (4y
and 4z). Additional groups adjacent to the 2-methyl group
affected neither the yield nor the enantioselectivity, with
methyl (4ae, 4ag) or methoxy groups (4aa, 4af) at different
positions affording the products in 90–92% ee. In addition,
a substrate with a free hydroxy group was compatible in this
reaction, with 4ah obtained in 93% yield and 86% ee. The
ortho groups in the aniline structure can be efficiently
extended to a naphthyl (4ab) or (tetrahydro)naphthyl (4ad)
scaffold to achieve good enantioselectivity. Furthermore, this
reaction is also compatible with halogen-containing sub-
strates and gives the corresponding product 4ai containing
a sensitive free NH2 group in high yield and 89% ee.
Moreover, the reaction is also compatible with quinoline
and indole heterocyclic substrates. However, when indole
Table 1: Optimization of the reaction conditions.[a]
Entry
Ligand
Base
X
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
L1
L1
L1
L1
L1
L2
L3
L4
L5
L6
L7
L8
L1
K2CO3
K2CO3
KOH
Br
I
82
84
93
98
92
90
85
82
80
78
10
70
80
75
68
89
92
92
89
88
84
70
62
–
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Cs2CO3
NaOH
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
9
10
11
12
13[d]
63
90
[a] Reaction conditions: 1 (0.1 mmol), Pd(OAc)2 (5.0 mol%), ligand
(7.5 mol%), and base (1.2 equiv) in solvent (1.0 mL) at 608C for 18 h.
[b] Yield of isolated product. [c] Determined by chiral HPLC. [d] Pd(OAc)2
(2.5 mol%), ligand (3.75 mol%), reaction time: 36 h.
2
ꢀ 2021 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 1 – 6
These are not the final page numbers!