Organic Letters
Letter
our continuous efforts on the addition reactions of imines and
unsaturated imines,8 we envisioned that the opposite chemo-
selectivity, i.e., the organometal species reacted with alkynes
first rather than imines, might be realized under a carefully
optimized asymmetric reaction system.
Table 1. Optimization of the Reaction Conditions
The carbon−carbon triple bond is a versatile functional
group that has abundant derivatization possibilities.9 The
transition-metal-catalyzed hydroalkynylation of internal al-
kynes with terminal alkynes has been extensively studied for
its perfect atom economy in the synthesis of functionalized
alkynes.10 Due to the wealth of carbo- and heterocycles in
natural products, it is of great importance to develop efficient
asymmetric annulation protocols; however, the synthesis of
enantioenriched heterocycles via transition-metal-catalyzed
alkynylative tandem reactions of alkynes is quite limited.11
Consequently, we proposed the rhodium-catalyzed enantiose-
lective alkynylative cyclization of N-(o-alkynylaryl)imines, even
though several challenges were encountered in this process
(Scheme 1d). The first issue is the unconventional chemo-
selectivity: the rhodium acetylide complex should undergo
cross addition with aryl alkynes rather than imines. The second
issue is the enantioselectivity: a suitable catalytic system should
be found to effectively distinguish the prochiral imines during
the intramolecular addition of alkenyl rhodium. If successful,
this could be an efficient path to the construction of
enantioenriched 3-methyleneindolines.12
With these considerations in mind, we initiated our
investigation by searching for a suitable chiral ligand for the
alkynylative cyclization reaction of alkynyl imine 1a and
(triisopropylsilyl)acetylene 2a using [Rh(COD)Cl]2 as the
catalyst and MgSO4 as the additive to prevent hydrolysis of the
imine substrates. A screening of frequently used ligands (L1−
L4) indicated that the efficiency was extremely low for our
desired reaction (Table 1, entries 1−4), and the major side
product was the N-substituted indole 3a′ derived from the
opposite chemoselectivity. (S)-SYNPHOS (L5) could pro-
mote this transition with a higher yield but showed almost no
enantioselectivity (entry 5). To our delight, the desired
product 3a was obtained with a much higher yield and
moderate enantioselectivity when (R,R)-Ph-BPE (L6) was
applied as the chiral ligand (entry 6). Then, we examined the
solvent effects and found that tetrahydrofuran was privileged in
the promotion of both reactivity and enantioselectivity (71%
yield, 77% ee, entry 7). After evaluation of a series of bases,
Cs2CO3 was proved to be vital to this transformation; other
bases such as its homologue K2CO3 only led to 14% yield and
27% ee (entry 8). This finding suggested that the cesium ion
may play a role in the enantioselective addition step via
coordination with imine. Decreasing the reaction temperature
resulted in an enhancement of enantioselectivities but a large
erosion of the yields (entries 9−11). Applying 1.5 equiv of
Cs2CO3 gave an improved yield to 60% (entry 12). Notably, in
the absence of MgSO4, the reaction was obviously inhibited to
give 8% yield (entry 13), and the addition of 4 Å molecular
sieve led no reaction as well (entry 14). These results indicated
that MgSO4 played an important role in the cascade reaction
indeed, but not the initially supposed role of dehydration. We
suspected that a trace amount of hydrate water in the MgSO4
may accelerate the reaction, so 20 equiv water was applied
instead of MgSO4. Gratifyingly, the addition of free water
significantly promoted the reactivity and afforded the product
3a with satisfactory yield and enantioselectivity (83% yield and
91% ee, entry 15). However, applying a stoichiometric amount
a
b
b
c
entry
L
solvent
T (°C) conv (%) yield (%) ee (%)
1
2
3
4
5
6
7
L1
L2
L3
L4
L5
L6
L6
L6
L6
L6
L6
L6
L6
L6
L6
L6
1,4-dioxane
1,4-dioxane
1,4-dioxane
1,4-dioxane
1,4-dioxane
1,4-dioxane
THF
THF
THF
THF
THF
60
60
60
60
60
60
60
60
40
30
25
30
30
30
30
30
76
56
35
55
77
100
100
45
92
87
53
85
18
<5
100
38
15
13
6
27
24
22
5
19
47
62
71
14
61
56
24
60
8
2
50
77
27
82
83
86
88
d
8
9
10
11
e
12
THF
THF
THF
THF
f
13
g
14
h
15
83
27
91
i
16
THF
a
b
Reactions were performed under an Ar atmosphere. Determined by
c
1H NMR analysis of unpurified mixtures. Determined by HPLC
d
e
analysis. K2CO3 was applied instead of Cs2CO3. 1.5 equiv of
Cs2CO3 was used. Mg2SO4 was not added. 50 mg of 4 Å MS was
applied instead of Mg2SO4. 20 equiv of H2O was applied instead of
Mg2SO4. 1 equiv of H2O was applied instead of Mg2SO4.
f
g
h
i
of water to the model reaction only led to 27% NMR yield,
which indicates that the equivalent of water has an important
influence on the reaction.
With the optimal reaction conditions established, the scope
of the N-(o-alkynylaryl)imines 1 was examined (Scheme 2).
We found that substrates derived from various aryl aldehydes
worked well in this reaction. The para-substituted and meta-
substituted phenyl substrates, including those with electron-
deficient substituents (NO2, F, Cl, Br) as well as electron-rich
substituents (Me, CH2Cl, Ph), were all converted smoothly
into the corresponding chiral indolines with moderate to good
yields and high ee values (52−80% yield, 77−92% ee, 3a−3i).
Pyridine-substituted alkynyl imine also gave the desired
cyclization product 3j, albeit with slightly lower ee value
(76% ee). A variety of substrates bearing electron-withdrawing
and electron-donating groups at the 4/5-position of the aryl
nucleus of N-(o-alkynylaryl)imines underwent the reaction
with high efficiency (62−82% yield, 82−94% ee, 3k−3o). As
for the alkynyl motif in the alkynyl imines, the substrate with
an electron-donating group (Me) at the 4-position of
phenylacetylene reacted smoothly under the standard con-
ditions to afford the corresponding chiral indoline with 67%
yield and 86% ee (3r). However, substrate with an electron-
4824
Org. Lett. 2021, 23, 4823−4827