Organic Letters
Letter
yields compared with those with electron-donating groups (2d
and 2e). In addition, the performance decreased in the order of
para (2d/2h/2l) > ortho (2b/2f/2j) > meta (2c/2g/2k)
regardless of the electronic property, suggesting that both the
nucleophilicity and the stabilization of the transient α-carbon
nucleophile play a crucial role in the cyclization. On the
contrary, the steric hindrance of the Ar1 has nearly no
influence. Various disubstituted phenyl ring such as p-F/m-Me
(2n), p-OMe/m-OMe (2q), p-Me/m-Me (2r), and p-Me/o-
Me (2s) were tolerated and provided the corresponding 1H-
indenes in 72−82% yields. The reductive cyclization also
proceeded smoothly when the Ar1 was replaced by either
naphthyl derivatives (2o and 2p) or heteroaromatic thiophene
(2t). Finally, the functionalities of the phenyl group (Ar2) at
the bridge with such a protocol were examined. Substituents
with not only electron-withdrawing (2u, 2v, 2y, and 2z)
functionalities but also electron-donating groups (2w and 2x)
were found to be tolerated. We also investigated the Michael−
Aldol cycloreduction of 1a on a scale of 4.24 mmol, which
provided 2a in an acceptable yield of 71% under the optimum
Ar1 and Ar2 were tolerated, wherein neither an electronic nor a
steric effect has been remarkably detected.
The potential for performing an intermolecular Michael−
Aldol reductive coupling was also evaluated under standard
conditions, taking the reaction of benzaldehyde (4) with
chalcone (5) as an example (Table 2, entry 1). Unfortunately,
Table 2. Cross Experiments on the Intermolecular
a
Michael−Aldol Reductive Coupling
b
yield (%)
entry
4 (x)
5 (y)
ligand (z)
6
7
8
1
2
3
4
5
6
7
8
9
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0)
(0)
(0)
(0)
(0.1)
(0.02)
(0)
(0.1)
(0.02)
(0.1)
(0.1)
(0.1)
(0.1)
(0)
(0)
(0)
44
38
6
58
8
0
0
0
0
On the basis of our previous studies10 and the present
results, this low-valent Co-catalyzed Michael−Aldol cyclo-
reduction may proceed via the famous Meerwein−Ponndorf−
Verley reduction route.15 To identify whether it follows a
concerted or stepwise fashion, the poisoning studies using
TEMPO as an inhibitor were performed (Table S3). To our
surprise, the formation of 2a was significantly suppressed with
the addition of TEMPO, whereas the yield of 3a increased. We
rationally suspected that the hydrogen transfer leading to the
Michael−Aldol cycloreduction product (2) proceeds via the
transiency Co−H species. In contrast, the oxa-Michael route to
3 occurs mainly in a concerted fashion. These two paths are in
competition. Having clarified the selectivity to each cascade
under the present protocol, we turned our attention to the
optimum conditions leading to 3 and its substrate tolerance
(Scheme 3). As a result, the best yield of 3a (82%) was
obtained with the addition of 4 equiv of TEMPO. In addition,
enone-tethered aldehydes with functionalities at both sites of
dtbpx (15)
dtbpx (15)
(0)
dtbpx (15)
dcpp (15)
dcpe (15)
0
0
0
1
14
0
0
0
0
84
98
0
a
b
Reaction conditions: toluene (0.5 mL), under Ar. Yields were
1
determined by H NMR using CH2Br2 as the internal standard.
the desired two-component cascade addition product was not
detected. Instead, only the reduction of 4 partially proceeded,
and a catalytic amount of 5 was sufficient but necessary for
such a transformation (entries 2 and 3). The addition of dtbpx
(15 mol %) did not improve the reactivity (entry 4 vs entry 1).
Meanwhile, a stoichiometric amount of 5 was required to
obtain a comparative efficiency (entry 4 vs entry 5). This is due
to the preferred coordination to Co(I) with dtbpx over 5,
whereas the latter functions as the active site for the reduction
of 4. To ascertain whether the reductive cyclization leading to
2 proceeds via the typical Michael−Aldol coupling cascade and
whether the varied reactivities under study differ from the
reactivity of the Michael reduction of enones, chalcone (5) was
subjected to the optimum conditions (entry 6). To our
surprise, 5 remained unchanged. In addition, the reduction of 5
proceeded poorly in the presence of dtbpx (entry 7), dppb,
xantphos, and dppp (Table S2) but readily occurred with dcpp
and dcpe (entries 8 and 9). Collectively, a high efficiency in the
Michael reduction of enones did not promise the formation of
1H-indene 2 over dihydroisobenzofuran 3. The aldehydes were
preferentially reduced in the presence of an enone function-
ality, most likely functioning as the ligand.
Scheme 3. Co-Catalyzed oxa-Michael cascade Leading to
a
Dihydroisobenzofurans (3)
Deuterium labeling experiments with the purpose of
obtaining insights into the route leading to such a Michael−
Aldol cycloreduction cascade were conducted (Table 3).
Consequently, the benchmark reaction was carried out with
(CD3)2CDOH (i-PrOH-d7) and (CD3)2CDOD (i-PrOH-d8),
respectively. The five-membered ring of 2a−d is significantly
deuterated, delivering a full deuteration at the methylene (Ha−
Hb) with a high percentage (>60%). In addition, the deuterium
content is almost the same with two H surrogates. This
suggests that the tertiary CH in i-PrOH, rather than the
hydroxyl group, functions as a H source in the formation of 2a.
a
Reaction conditions: toluene (0.5 mL), under Ar, isolated yield.
3875
Org. Lett. 2021, 23, 3873−3878