Angewandte
Chemie
Table 1: Preliminary assays of nickel-catalyzed cycloadditions of 1a.
Table 2: Nickel-catalyzed [3+2+2] cycloadditions of 1a and alkenes.
Entry[a] MVK
[equiv]
[Ni] (mol%)
Solvent
T
5aa/
Yield[c]
Entry[a]
Alkene
5/6 (ratio)[b]
5aa/6aa
Yield[c]
80
[8C] 6aa[b]
1
2
(71:29)
(76:24)
(80:20)
1
2
3
4
5
6
7
8
9
10
0
0
2
2
2
10
10
10
10
10
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2] (10)
[Ni(cod)2]/
toluene RT
toluene 90
toluene RT
toluene 40
toluene 90
toluene 40
DMF 40
dioxane 40
THF
toluene 40
–
–
0[d]
0[d]
40[e]
50[f]
58[g]
80
49
73
75
0
5ab/6ab
5ac/6ac
70
74:26
64:36
72:28
71:29
67:33
67:33
65:35
–
3
51
4
5
6
5ad/6ad
(83:17)
40[d]
–
–
–
–
–
40
–
[a] Conditions: 1a (0.2m in toluene), 10 mol% [Ni(cod)2], 10 equiv of
alkene unless otherwise noted, for 3 h. [b] Based on the 1H NMR
spectrum of the crude reaction mixture. [c] Combined yield of 5a and 6a
after full conversion. [d] 1.2 equiv of the alkene was employed.
PPh3(10)
[a] Conditions: 1a (0.2m in toluene), 10 mol% [Ni(cod)2], MVK (0–
10 equiv) for 3–12 h. [b] Based on the H NMR spectrum of the crude
1
reaction mixture. [c] Combined yield of 5aa and 6aa after full conversion,
unless otherwise noted. [d] Compound 1a was recovered after 12 h.
[e] 88% conversion. [f] 66% conversion. [g] 65% conversion. DMF=
N,N-dimethylformamide, THF=tetrahydrofuran.
and efficiency of the cycloaddition, we evaluated the reac-
tivity of derivative 1b, which features an electron-withdraw-
ing ester substituent on the alkyne. The formation of [3+2+2]
adducts of type 5 was now found to be clearly favored, and the
competitive [3+2] cycloadducts (6) were not detected
(Table 3, entries 1–4). However, the transformations suffered
from moderate conversion,[12] probably because of catalyst
inhibition by the product.[13] In any case, when phenyl vinyl
sulfone was used as an intermolecular two carbon (2C)
partner, the [3+2+2] cycloadduct could be obtained in up to
68% yield (Table 3, entry 4).
In search of a compromise to favor the [3+2+2] annealing
process while avoiding the potential deactivation of the
catalyst, we analyzed the performance of enynes equipped
with other activating substituents on the alkyne group.
Remarkably, substrate 1c, which bears a CH2OAc substitu-
ent, provided full conversion using 10 mol% of catalyst, and
gave higher [3+2+2]/[3+2] ratios (80:20) than those obtained
from the cycloaddition of the nonactivated analogue 1a
(71:29; Table 3, entry 5 vs. Table 1, entry 6).
Replacing the acetyl residue in 1c with a tert-butyl-
dimethylsilyl group (1d) led to improved selectivity without
eroding the catalyst efficiency. Indeed, the cycloaddition
between 1d and methyl vinyl ketone at 408C, in the presence
of [Ni(cod)2] (10 mol%), exclusively gave the desired
[3+2+2] cycloaddition product 5da, which was isolated in
an excellent 89% yield (Table 3, entry 6). The cycloaddition
of 1d was also carried out with other alkenes, including ethyl
acrylate, acrolein, and phenyl vinyl sulfone. In all of these
cases, the reaction was complete after 12 hours and the
[3+2+2] cycloadducts were obtained in good to excellent
yields (Table 3, entries 7–9).[14] Importantly, the presence of a
germinal diester in the connecting chain of 1d is not
mandatory, as the cycloaddition of the ether (1e) or N-tosyl
derivatives (1 f) led to comparable yields and also complete
selectivities in favor of the desired [3+2+2] adducts (Table 3,
entries 10–13).
of [Ni(cod)2] (10 mol%), even when heated in toluene at
908C (Table 1, entries 1 and 2). However, 1a did react at
room temperature with the same nickel complex when it was
treated with 2 equivalents of methyl vinyl ketone (MVK),
thereby providing a 74:26 mixture of the [3+2+2] and [3+2]
cycloadducts 5aa and 6aa, which were isolated in a 40%
combined yield (Table 1, entry 3). Raising the temperature to
408C, or even to 908C, led only to a marginal increase in the
efficiency of the process (Table 1, entries 4 and 5). Con-
versely, increasing the amount of methyl vinyl ketone up to
10 equivalents allowed for full conversions after 3 hours at
408C, and cycloadducts 5aa and 6aa were isolated in a good
80% combined yield (ratio 71:29; Table 1, entry 6).[9,10] The
reaction could also be performed in other solvents, such as
N,N-dimethylformamide, dioxane, or tetrahydrofuran; how-
ever, the products were obtained in somewhat lower yields
and selectivities (Table 1, entries 7–9). Interestingly, the
addition of an external ligand, such as PPh3, completely
inhibited the cycloaddition (Table 1, entry 10), whereas other
sources of nickel(0), such as [NiCl2(PPh3)2]/Et2Zn or Ni-
(acac)2/Et2Zn proved to be ineffective.[10]
The electronic and structural requirements of the alkene
component were studied next (Table 2). Gratifyingly, cyclo-
addition reactions of 1a with ethyl acrylate, acrolein, or
phenyl vinyl sulfone were also efficient, thus providing the
desired [3+2+2] adducts 5ab–5ad with moderate selectivities
and good overall yields (Table 2, entries 1–4). The presence of
the electron-withdrawing group on the alkene seems to be
mandatory, as the reactions with nonactivated alkenes, such as
styrene or methyl-5-hexenoate, led to recovery of 1a (Table 2,
entries 5 and 6).[11]
Envisaging that the electronic properties of the alkyne
could play an important role in determining the selectivity
Angew. Chem. Int. Ed. 2010, 49, 9886 –9890
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim