Angewandte
Communications
Chemie
Table 1: Summary of reaction optimization.
optimal catalyst C and 460 nm light, out of the six
amines we screened (I–VI), only Hꢀnigꢁs base
(DIPEA, III) gave synthetically useful yields of
products 4–6. When using triethylamine I (Table 1,
entry 1, reaction partners 2a and 3a) or amines II or
IV–VI, full conversion was generally achieved within
48 h. However, complex mixtures of undefined
products were formed that contained only up to
10% of the products 4–6 in varying ratios. Entries 2
and 3 of Table 1 show the results of the reaction of
substrate 2a with methyl acrylate (3a) in the
presence of catalyst C and DIPEA (III). When
10 equivalents of the alkene were used along with
1 mol% of the photocatalyst and 5 equivalents of the
amine base (entry 3), the relative amount of the
desired product 4a was maximized and it could be
isolated in 29% yield as a single diastereoisomer.
Using the N-Boc-protected substrate 2b under the
same conditions led to a lower yield of product 4b
(entry 4), and as observed in the previous experi-
ments, catalysts D and E again gave inferior results
(entries 5 and 6). When using substrate 2a, catalyst
C, and amine III but with N,N-dimethyl acrylamide
(3b) as the alkene component, 5c and 6c were the
only reaction products (entry 7). Finally, we were
pleased to observe that using acrylonitrile (3c, added
in three portions over a 72 h period) under otherwise
similar reaction conditions led to the isolation of the
(4+2)-cyclization/1,4-addition product 4d with
a chemical yield as high as 80%. Compound 4d
was obtained with C4a,C9a-cis configuration as a 5:1
mixture of two separable C1 epimers, with the major
isomer being C1,C9a-trans configured.
No. 2 (Pg)
3 (equiv)
Cat.
Amine
Ratio 5/6/4
Yield
[%][b]
(mol%) (equiv)
[%][a]
1
2a (Ts) 3a (5)[c]
C (1)
I (5)
0:64:36
6a <5%
4a <5%
n.d.
5a 14%
6a 17%
4a 29%
5b 14%
6b 20%
4b 26%
n.d.
2
3
2a (Ts) 3a (5)[c]
2a (Ts) 3a (10)[c]
C (1)
C (1)
III (5)
III (5)
16:52:32
35:17:48
4
2b
3a (10)[c]
C (1)
III (5)
21:30:49
(Boc)
5
6
7
2a (Ts) 3a (10)[c]
2a (Ts) 3a (10)[c]
2a (Ts) 3b (10)[c]
D (1)
E (5)
C (1)
III (5)
III (5)
III (5)
46:40:14[d]
n.d.
n.d.
6a trace
5c 9%
6c 39%
4d 80%
8
2a (Ts) 3c
C (1)
III (3ꢀ5)
5:5:90
(3ꢀ10)[e]
General conditions: c=0.10m in MeCN, irradiation using blue LEDs (460 nm, 12
W) for catalysts C,D and a compact fluorescent lamp (CFL) (370 nm, 36 W) for
catalyst E. dtbbpy=4,4’-di-tert-butyl-2,2’-bipyridine, ppy=2-phenylpyridine.
[a] Ratio determined by 1H-NMR analysis. [b] Yield of isolated product. [c] Reaction
time 48 h. [d] 30% conversion. [e] Reaction time 72 h.
Utilizing the optimal reaction conditions,
a number of hexahydro-1H-carbazoles (4) could be
synthesized from substrates 2, with yields ranging
evaluated a variety of possible reaction conditions (Table 1).
The radical initiators A and B were tested, as well as the
photocatalysts C–E, which were used under irradiation and in
combination with the six tertiary amines I–VI. These initial
studies soon established that substrate 2a could successfully
be converted into mixtures of three major (4+2)-cyclization
products, namely the aromatic tetrahydrocarbazole 5, the
hexahydrocarbazole 6, and the valuable C4a-substituted
hexahydrocarbazole product 4, which we consequently
aimed for. Notably, products 4–6 were observed only when
using the alkenes 3a–c, whereas alkenes 3d–f underwent
polymerization under different reaction conditions and gave
at best only traces of any addition or cyclization products.
In terms of the reaction conditions, initiation of the
reaction failed when using Et3B/O2 (A),[7] and reduction of
from 29–80% (Figure 2). The stereochemical assignments,
which were initially based on NMR analysis, were confirmed
by single-crystal X-ray analysis of compound 4a and the
C1 epimer of compound 4k.[10] When acrylonitrile (3c) was
used as the alkene, N-tosyl derivatives were generally
obtained in higher yields than N-Boc or N-methyl derivatives
(examples 4d–f). On the other hand, employing substrates 2,
which possess different substituents on the aromatic ring, led
to quite variable yields (products 4g–k). In addition, the
hexahydrodibenzofuran derivative 7 was sucessfully prepared
from its corresponding 3-(2-iodoethyl)benzofuran precursor.
While all products 4 and 7 showed cis-fused B and C rings, the
C1 epimeric ratio varied from 8:1 to 1.5:1, and the C1,C9a-
trans-configured isomer predominated in all cases. Notably,
all C1 epimers were easily separable by column chromatog-
raphy, and in addition, the C1 configuration could be readily
equilibrated. When pure 1-epi-4d was treated with LDA and
reprotonated at low temperature (ꢀ788C, THF, then addition
of MeOH), it was converted into the thermodynamically
more stable product 4d with 75% selectivity.
ꢀ
the C I bond of substrate 2a predominated when using
TMS3SiH (B) with 370 nm irradiation.[8] Employing the
strongly reducing photocatalyst N-phenylphenothiazine
(E)[9] in combination with amine bases and 370 nm light led
ꢀ
to C I bond reduction accompanied by N-detosylation.
Between the IrIII-based catalysts C and D (which were excited
with blue LED light), Ir(dtbbpy)(ppy)2PF6 (C) gave superior
results compared to Ir(ppy)3 (D).[6a] Finally, when using the
A proposed mechanism for the novel radical (4+2)-
cyclization/1,4-addition cascade leading to products 4 is
depicted in Scheme 2. Initially, the excited photocatalyst C
2
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!