Organic Letters
Letter
Scheme 1. Transition-Metal-Catalyzed Ring-Opening
Hydroacylation for the Synthesis of γ,δ-Unsaturated
Ketones
Table 1. Optimization Studies for the Rhodium-Catalyzed
Synthesis of γ,δ-Unsaturated Ketone 3aa
a
b
entry
variation from the standard conditions
none
without L1
L2 instead of L1
L3 instead of L1
yield (%)
1
2
3
4
5
6
7
8
73
33
25
64
62
45
53
48
59
64
55
70
60
58
L4 instead of L1
100 mol % of L1 instead of 25 mol % of L1
50 mol % of L1 instead of 25 mol % of L1
10 mol % of L1 instead of 25 mol % of L1
1.5 equiv of 2a instead of 2.5 equiv of 2a
PPh3 instead of (p-Me-C6H4)3P
without K2CO3
p-xylene instead of toluene
0.1 M instead of 0.2 M
120 °C instead of 140 °C
9
10
11
12
13
14
cyclopropane ring fragmentation and isomerization is predom-
inant over reductive elimination;7 and (3) control of the
selectivity would be a difficult task in order to obtain high
yields of the desired linear γ,δ-unsaturated ketones. In
continuation of our research on rhodium(I)-catalyzed
regioselective C−C bond formation reactions,10 herein we
describe the development of a new strategy that highlights the
chelation assistance of a catalytic amount of N,N-dimethylme-
thacrylamide (L1), enabling the first example of the
intermolecular hydroacylation of ACPs by the proximal C−C
bond cleavage.
a
Unless otherwise noted, each reaction was run with 1a (0.2 mmol, 1
equiv) and 2a (2.5 equiv) in 1 mL of toluene at 140 °C for 24 h.
b
Isolated yield. COD = 1,5-cyclooctadiene.
Finally, the ring-opening hydroacylation proved to be less
efficient when it was performed at a lower temperature (entry
14).
The generality of the rhodium-catalyzed intermolecular ring-
expanding hydroacylation with regard to the alkylidenecyclo-
propane partner was subsequently investigated utilizing the
optimized conditions (Scheme 2). The methyl-, tert-butyl-, and
methoxy-substituted benzylidenecyclopropanes all reacted
smoothly with salicylaldehyde 1a under rhodium(I) catalysis
to provide isolated 61−78% yields of γ,δ-unsaturated ketones
(3ab−3ae). Significantly, a variety of alkylidenecyclopropane
derivatives bearing electron-neutral or electron-deficient
groups expressed moderate to good reactivity to produce the
desired ketones (3af−3ak). Of note, 2-furyl-, 3-benzofuryl-,
and 2-pyridinyl-substituted components (3al−3an) were also
compatible with the reaction conditions for the hydroacylation,
highlighting the robustness of the protocol. It was found that
the symmetrical diphenylmethylenecyclopropane underwent
the coupling reaction smoothly to afford the anticipated
compound 3ao in 54% yield. Unfortunately, the substrate
bearing a long-chain alkyl group (2p) was not allowed for this
transformation. Moreover, the substrate substituted with
methyl and phenyl groups on the double bond was also
evaluated in the coupling reaction, but the product was isolated
in low yield (3aq).
Our attention was then turned to the coupling reactions of
chelating aldehydes with ACPs, and the results are illustrated
in Scheme 3. Good functional group compatibility for the
synthesis of linear γ,δ-unsaturated ketones was again observed,
with methyl (3ba), benzyloxy (3ca), chloro (3da), trifluor-
omethyl (3ea), nitro (3fa), and ester (3ga) groups all being
well-tolerated. Electron-donating salicylaldehydes reacted with
halo-substituted partners smoothly to furnish the correspond-
ing products 3hh and 3ii in 69% and 67% yield, respectively.
Initially, we investigated the possibility of such a conversion
by examining the model reaction between salicylaldehyde 1a
and freshly prepared benzylidenecyclopropane 2a under
rhodium(I) catalysis (Table 1). In light of the bidentate
coordination of α,β-unsaturated compounds to the rhodium(I)
intermediates,10b,11 we added a commercially available material
to the mixture to promote the reaction. To our delight, the
[Rh(COD)Cl]2/(p-Me-C6H4)3P/K2CO3 catalytic system in
combination with 25 mol % of N,N-dimethylmethacrylamide
(L1) delivered 73% yield of the linear trans-γ,δ-unsaturated
ketone 3aa (entry 1). The absence of L1 proved very
detrimental to the yield of the desired ring-opening hydro-
acylation product (entry 2). It is noteworthy that 2-
ethylacrolein (L2) that could promote the hydroacylation of
1,3-dienes with salicylaldehydes10b was much less efficient
(entry 3), demonstrating the importance of acrylamides in
promoting the transformation. Switching L1 to other
homologues such as N,N-dimethylacrylamide (L3) and
methacrylamide (L4) gave slightly lower yields of the product
(entries 4 and 5). The influences of the loading of L1 on the
ring-opening hydroacylation were also examined, and all led to
low conversions (entries 6−8). In addition, decreasing the
loading of benzylidenecyclopropane 2a to 1.5 equiv lowered
the yield to 59% (entry 9). Replacing (p-Me-C6H4)3P with
PPh3 still afforded the homoallylic ketone in 64% yield (entry
10). An experiment carried out without the base showed a
poorer yield for 3aa (entry 11). The hydroacylation performed
in p-xylene produced a yield comparable to that achieved in
toluene (entry 12). However, running the reaction at a lower
concentration resulted in inferior yield of 3aa (entry 13).
B
Org. Lett. XXXX, XXX, XXX−XXX