Organic Letters
Letter
phosphoric acid (CPA)-catalyzed enantioselective formal (3 +
2) cycloaddition of quinone diimides and enecarbamates that
provides simple and practical access to 2,3-disubstituted
indolines (Scheme 1).
Table 1. Optimization of the Catalytic Enantioselective
Synthesis of Indolines
a
Scheme 1. Catalytic Enantioselective Formal (3 + 2)
Cycloaddition with Quinone Derivatives
b
c
de
,
entry
cat
2
solvent
yield (%)
dr
ee (%)
1
2
3
4
5
6
7
8
9
4a
4b
4c
4d
4e
4d
4e
4d
4e
4e
2a
2a
2a
2a
2a
2a
2a
2a
2a
2b
toluene
toluene
toluene
toluene
toluene
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
93
91
66
92
91
98
93
72
93
94
5:1
5:1
5:1
5:1
5:1
6:1
6:1
6:1
6:1
6:1
(+) 34
(+) 32
(+) 43
(+) 98
(−) 98
(+) 98
(+) 98
(+) 98
(−) 98
(−) 94
We previously reported a CPA-catalyzed enantioselective (3
+ 2) cycloaddition reaction of quinones with ene- and
thioenecarbamates.7a The reaction has a broad scope, offering
2,3-disubstituted 2-amino-2,3-dihydrobenzofurans with high
enantioselectivity. The diastereoselectivity is highly dependent
on the protecting group on the nitrogen atom of the
dipolarophile: the thiocarbamates gave high selectivity in
favor of the trans isomer, while the reactions with the
enecarbamates were poorly diastereoselective.
f
f
10
a
General conditions: 1a (0.10 mmol), 2a (0.12 mmol), and Cat (5
b
mol %) in solvent (0.1 M) at rt for 16 h. Yields of isolated pure
c
1
products after chromatography. Determined by H NMR analysis.
d
(+) refers to the (2S,3S) absolute configuration and (−) to the
On the basis of our previous work,7a,13 we initially
investigated the reaction between p-quinone di-p-toluenesulfo-
nimide (1a) and (E)-thioenecarbamate 2a in the presence of a
CPA catalyst (Table 1).14 Pleasingly, the desired 2,3-
disubstituted 2-aminoindoline 3a was obtained in high yield
albeit, with moderate selectivity, by the use of 5 mol % 4a as
the catalyst in toluene at room temperature for 16 h (entry 1).
To increase both the diastereo- and enantioselectivity of this
reaction, a series of CPA catalysts 4 were evaluated (entries 1−
5). Catalyst 4d bearing 2,4,6-triisopropylphenyl groups at the
3- and 3′-positions of BINOL as well as catalyst 4e derived
from SPINOL and having 1-naphthyl groups at the 6- and 6′-
positions provided the desired product with much better
enantioselectivity (98% ee; entries 4 and 5), but the
diastereomeric ratio was not significantly influenced by the
catalyst choice. A slight increase in the trans diastereoselectivity
was observed with CH2Cl2 as the solvent (entry 6 vs 4). On
one side, reducing the catalyst loading of 4d to 2.5 mol % led
to a reduction in yield (entry 8 vs 6). On the other side, no
activity loss was noted when 2.5 mol % 4e was used (entry 9 vs
7), thereby providing conditions that were selected for further
investigations. When we conducted the reaction with
enecarbamate 2b, the corresponding cycloadduct 3b was
formed with similar diastereo- and enantioselectivity as 2a,
thus indicating (contrary to our previous work)7a that the
protecting group of 2 did not have too much influence on the
stereoselectivity. Other solvents such as CHCl3, CH3CN,
THF, and toluene were evaluated to improve the diaster-
higher diastereomeric ratio was found only in THF (8:1), the
yield was reduced substantially.15 All attempts to separate the
diastereomers were unsuccessful.
(2R,3R) configuration as assigned by analogy to the X-ray crystal
e
Determined by HPLC analysis on a chiral stationary phase. With
2.5 mol % 4.
worked well for selected protecting groups bonded to the
nitrogen of 2, including benzyloxy, prop-2-yn-1-yl, and tert-
butoxy carbamates. A range of linear β-substituted ene- and
thioenecarbamates 2 reacted successfully, affording the
corresponding 2-aminoindolines 3c−k in high yields with
excellent enantioselectivities. An additional benzyl ether
functional group was well-tolerated on the enecarbamate,
providing cycloadduct 3f with good enantioselectivity. The
enecarbamate bearing an ethylbenzene side chain also
furnished a high level of enantioselectivity (3g, 99% ee). To
our satisfaction, the ene- and thioenecarbamate substituted
with an isopropyl group at the β-position furnished
compounds 3h and 3i with excellent diastereoselectivity
(10:1 to 14:1) possibly revealing a significant substrate control
effect. However, a slight decrease in enantioselectivity was
observed for the thioenecarbamate compared with the
enecarbamate. Surprisingly, performing the reaction with an
unsubstituted cyclopropyl side chain instead of an isopropyl
group led to a drastic reduction in diastereoselectivity, but the
enantioselectivity was kept high (3j; 2:1 dr, 98% ee). Only
trace of product was formed with the enecarbamate bearing a
phenyl group at the 2-position (data not shown). The presence
of an enantiopure predefined chiral center on the enecarba-
mate moiety did not interfere with catalyst induction. Hence,
compound 3k was prepared in 93% yield with 95% ee and a
satisfactory 4:1 diastereoselectivity relative to the two newly
formed chiral centers.
Having identified the optimum conditions (Table 1, entry
9), we explored the generality of the reaction by employing
different enecarbamates (Scheme 2). In general, the reactions
Alternative quinone diimines bearing various protecting
groups and substituents were also investigated. On the one
hand, dimethylsulfonimide uneventfully participated in the
443
Org. Lett. 2021, 23, 442−448